Myelin: Biology and Disorders

Contributors

Mihaela Anitei Department of Neuroscience, University of Connecticut Medical School, Farmington, Connecticut 06030-3401 Jack Antel Department of Neurology and Neurosurgery, McGill University, Montreal, Quebec, Canada H3A 2B4 D. L. Arnold Magnetic Resonance Spectroscopy Laboratory, Brain Imaging Center, Montreal Neurological Institute, Montreal, Quebec, Canada H3A 2B4 Edgardo J. Arroyo Department of Neurology, The University of Pennsylvania Medical Center, Philadelphia, Pennsylvania 19104 Lakshmi Bangalore Department of Neurology and PVA/EPVA Center for Neuroscience Research, Yale University School of Medicine, New Haven, Connecticut 06510 Rashmi Bansal Department of Neuroscience, University of Connecticut Medical School, Farmington, Connecticut 06030-3401 Susan C. Barnett Division of Clinical Neuroscience, University of Glasgow, G61 1BD Glasgow, United Kingdom A. Baron-Van Evercooren The French Institute of Health and Medical Research (INSERM), CHU Pitie´ Salpeˆtrie`re, 75634 Paris, France Amit Bar-Or Neuroimmunology Unit, Montreal Neurological Institute, Montreal, Quebec, Canada H3A 2B4 Hugo J. Bellen Howard Hughes Medical Institute, Department of Molecular and Human Genetics, Division of Neuroscience, Department of Molecular and Cell Biology, Program in Developmental Biology, Baylor College of Medicine, Houston, Texas 77030 Manzoor A. Bhat Cardiovascular Research Institute, Department of Medicine, Mount Sinai School of Medicine, New York, New York 10029 A. J. Bieber Department of Neurology and Program in Molecular Neuroscience, Mayo Medical and Graduate Schools, Rochester, Minnesota 55905 W. F. Blakemore Department of Clinical Veterinary Medicine, University of Cambridge, Cambridge, United Kingdom Peter E. Braun Department of Biochemistry, McGill University, Montreal, Quebec, Canada H3G 1Y6 Michael Brenner Department of Neurobiology, University of Alabama at Birmingham, Birmingham, Alabama 35294-0021 Peter J. Brophy Department of Preclinical Veterinary Sciences, University of Edinburgh, EH9 1QH Edinburgh, Scotland Anthony T. Campagnoni Neuropsychiatric Institute, University of California Medical School, Los Angeles, California 90024

Myelin Biology and Disorders, Volume 1

xvii

Copyright 2004, Elsevier Science (USA). All rights reserved.

xviii

CONTRIBUTORS

Celia W. Campagnoni Neuropsychiatric Institute, University of California Medical School, Los Angeles, California 90024 Alberto Cifelli Center for Functional Magnetic Resonance Imaging of the Brain, Department of Clinical Neurology, University of Oxford, John RadcliVe Hospital, OX3 9DU Oxford, United Kingdom Alastair Compston Department of Neurology, University of Cambridge Clinical School, Addenbrooke’s Hospital, CB2 2QQ Cambridge, United Kingdom Andre Dautigny Laboratory of Neurogenetics, National Center for ScientiWc Research (CNRS), UMR 7624, 75252 Paris, France Maria Laura Feltri Department of Biological and Technological Research (DIBIT), San RaVaele ScientiWc Institute, 20132 Milan, Italy Robin J. M. Franklin Department of Clinical Veterinary Medicine and Brain Repair Center, University of Cambridge, CB3 OE5 Cambridge, United Kingdom Charles Vrench-Constant Cambridge Center for Brain Repair, University of Cambridge, CB2 2PY Cambridge, United Kingdom James Y. Garbern Department of Neurology, Wayne State University School of Medicine, Detroit, Michigan 48201 John Georgiou Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada M5G 1X5 James E. Goldman Department of Pathology and the Center for Neurobiology and Behavior, Columbia University College of Physicians and Surgeons, New York, New York 10032 Steven A. Goldman Department of Neurology and Neuroscience, Cornell University Medical College, New York, New York 10021 Alexander Gow Center for Molecular Medicine and Genetics, Departments of Pediatrics and Neurology, Wayne State University School of Medicine, Detroit, Michigan 48301 John W. GriYn Department of Neurology, The John Hopkins University School of Medicine, Baltimore, Maryland 21287-7608 Ian R. GriYths Institute of Comparative Medicine, Glasgow, United Kingdom Michel Gravel Department of Biochemistry, McGill University, Montreal Quebec, Canada H3G 1Y6 Rebecca J. Hardy MRC National Survey of Health and Development, Department of Epidemiology and Public Health, University College London Medical School, WC1E 6BT London, United Kingdom Lynn D. Hudson National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland 20892-4160 K. R. Jessen Department of Anatomy and Developmental Biology, University College London, WC1E 6BT London, United Kingdom Richard T. Johnson Department of Neurology, John Hopkins School of Medicine, Baltimore, Maryland 21287 John A. Kamholz Department of Neurology, Wayne State University, Detroit, Michigan 48201 Grahame J. Kidd Department of Neuroscience, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, Ohio 44195 Daniel A. Kirschner Biology Department, Boston College, Chestnut Hill, Massachusetts 02467-3811 Kleopas A. Kleopa The Cyprus Institute of Neurology and Genetics, 1683 Nicosia, Cyprus Hans Lassmann Division of Neuroimmunology, Brain Research Institute, University of Vienna, A 1090 Wien, Austria

CONTRIBUTORS

Christopher Linington Neuroimmunology Department, Max Planck Institute, Martinsried, Germany Fred D. Lublin Corinne Goldsmith Dickinson Center for Multiple Sclerosis, Mount Sinai Medical Center, New York, New York 10032 Eugene O. Major Laboratory of Molecular Medicine and Neuroscience, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland 20892 Cecilia B. Marta Department of Neuroscience, University of Connecticut Medical School, Farmington, Connecticut 06030-3401 P. M. Matthews Magnetic Resonance Spectroscopy Laboratory, Brain Imaging Center, Montreal Neurological Institute, Montreal, Quebec, Canada H3A 2B4 Albee Messing Department of Pathobiological Sciences, University of Wisconsin School of Veterinary Medicine, Madison, Wisconsin 53706 Robert Miller Department of Neuroscience, Case Western Reserve University, School of Medicine, Cleveland, Ohio 44106 R. Mirsky Department of Anatomy and Developmental Biology, University College London, WC1E 6BT London, United Kingdom Hugo W. Moser Neurogenetics Research, Kennedy Krieger Institute, Baltimore, Maryland 21205 Klaus-Armin Nave Max Planck Institute of Experimental Medicine, Go¨ttingen, Germany Pedro Pasik Department of Neurology, Mount Sinai School of Medicine, New York, New York 10029 Tauba Pasik Department of Neurology, Mount Sinai School of Medicine, New York, New York 10029 David L. Paul Department of Molecular Cell Biology, Weizmann Institute of Science, 76100 Rehovot, Israel Elior Peles Department of Molecular Cell Biology, Weizmann Institute of Science, 76100 Rehovot, Israel Steven E. PfeiVer Department of Neuroscience, University of Connecticut Medical School, Farmington, Connecticut 06030-3401 Danielle Pham-Dinh The French Institute of Health and Medical Research (INSERM), U 546, University of Paris, 75013 Paris, France James M. Powers Department of Pathology and Laboratory Medicine, University of Rochester School of Medicine, Rochester, New York 14627 Mahendra Rao Laboratory of Neurosciences, Gerontology Research Center, National Institute on Aging, National Institutes of Health, Baltimore, Maryland 21224 Richard Reynolds Department of NeuroinXammation, Division of Neuroscience, Imperial College Faculty of Medicine, Charing Cross Campus, W68 RF London, United Kingdom John C. Roder Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada M5G 1X5 M. Rodriguez Departments of Neurology, Immunology, and Program in Molecular Neuroscience, Mayo Medical and Graduate Schools, Rochester, Minnesota 55905 Neeta S. Roy Department of Neurology and Neuroscience, Cornell University Medical College, New York, New York 10021 Steven S. Scherer Department of Neurology, The University of Pennsylvania Medical Center, Philadelphia, Pennsylvania 19104 Karen L. Schulze Howard Hughes Medical Institute, Department of Molecular and Human Genetics, Division of Neuroscience, Department of Molecular and Cell Biology, Program in Developmental Biology, Baylor College of Medicine, Houston, Texas 77030

xix

xx

CONTRIBUTORS

Kazim Sheikh Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, Maryland 21287 Diane L. Sherman Department of Preclinical Veterinary Sciences, University of Edinburgh, EH9 1 QH Edinburgh, Scotland Ueli Suter Institute for Cell Biology, Eth Zurich, 8093 Zurich, Switzerland Kunihiko Suzuki Brain and Developmental Research Center, University of North Carolina, Chapel Hill, North Carolina 27599 Peter K. Stys Department of Medicine, Division of Neuroscience, University of Ottowa, Ottowa, Canada K1Y 4E9 Christopher M. Taylor Department of Neuroscience, University of Connecticut Medical School, Farmington, Connecticut 06030-3401 Bruce D. Trapp Department of Neuroscience, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, Ohio 44195 Michael B. Tropak Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada M5G 1X5 Stephen G. Waxman Department of Neurology and PVA/EPVA Center for Neuroscience Research, Yale University School of Medicine, New Haven, Connecticut 06510, and Rehabilitation Research Center, VA Connecticut Healthcare System, Westhaven, Connecticut 06516 Martha S. Windrem Department of Neurology and Neuroscience, Cornell University Medical College, New York, New York 10021 Lawrence Wrabetz Department of Biological and Technological Research (DIBIT), San RaVaele ScientiWc Institute, 20132 Milan, Italy

Pierre Morell, Ph.D. (1941-2003)

It is a sad coincidence that Pierre Morell, a friend of everybody in the Weld of myelin, died a tragically early death on July, 15, 2003, just when this new book on myelin was being readied for publication. His name is inextricably associated with the biochemistry and biology of myelin. The two editions of the monograph, Myelin, he edited Wrst in 1977 and then in 1984, have been read by experts and beginners alike as the standard reference source. The Wrst edition of the Morell Myelin book was published when Pierre was only 35 years old, and underscores his precocious development as a scientist and his organizational skills. An entire generation of biochemists, neuroscientists, and toxicologists around the world have beneWted from the clarity, rigor and creativity Pierre brought to his life’s work on the biology of myelin. The present new and massively expanded book on myelin was planned as the successor to his book, since a fresh update of the ‘‘Morell Myelin book’’ was needed after twenty years. Pierre began his life in turmoil, or perhaps in a more positive mode, in Sturm und Drang. In 1941, his parents were among the beneWciaries of the now famous Japanese Consulate oYcial in Kaunas, Lithuania, who deWed instructions from his government and kept issuing transit visas through Japan for thousands of Jewish refugees. Pierre always boasted that he traveled through Japan in utero. After several months in Japan, his parents landed in Dominican Republic where Pierre was born. Another of Pierre’s declaration was that, luckily not having been born within US territory, he could not run for the US Presidency. Eventually, the Morells settled in New York city, where his father, Anatol Morell, was Professor of Biochemistry at Albert Einstein College of Medicine for many years. He was a distinguished biochemist who, together with Gil Ashwell, discovered one of the earliest known carbohydrate receptors, the Ashwell-Morell hepatic galactose receptor. Pierre graduated from the famed Bronx High School of Science, which produced so many distinguished scholars, scientists and Nobel Lauriates. Pierre obtained his undergraduate degree in chemistry from Columbia University and then a Ph.D. in biochemistry working on nucleic acids under Julie Murmur at Albert Einstein College of Medicine. Unlike many of us who started with other areas of biochemistry and eventually moved to molecular biology, Pierre started with molecular biology and moved on to other areas. His introduction to ‘‘neuro’’ came while he was working as a postdoc in the laboratory of Norm Radin, where he started on projects involving biosynthesis of the most characteristic myelin lipid, galactosylceramide. Almost immediately he corrected the then textbook notion that galactosylceramide synthesis occurred Wrst by galactosylation of sphingosine to psychosine and then by its acylation to galactosylceramide and showed deWnitively that galactosylceramide synthesis occurs through ceramide. Moving back to Einstein as Assistant Professor in Neurology in 1969 Pierre pushed further the enzymological characterization of UDP-gal-ceramide galactosyltransferase. Pierre was hooked on myelin research for life. In 1973, he moved to North Carolina where he would remain for the rest of his

Myelin Biology and Disorders, Volume 1

xxi

Copyright 2004, Elsevier Science (USA). All rights reserved.

xxii

Pierre with his parents.

career. His work took many new directions. Pierre was among the Wrst to recognize the importance of understanding myelin as the sum of its biochemical, genetic and molecular properties. His talents as a biochemist combined with his intellectual gifts as a neurobiologist produced keen insights into how the molecular constituents of myelin are synthesized, transported and assembled, and how these processes might be compromised by disease or environmental hazards. Pierre was among the earliest workers to use mutant mice, such as quaking and jimpy, as models to demonstrate how abnormalities in single genes can compromise the function of myelin. His eVorts provided new insights into our understanding of demyelinating diseases like multiple sclerosis as well as the hazards of a wide array of toxic substances for the integrity of the nervous system. Pierre’s family, his parents, his wife Bonnie and his two children, David and Sharon, lost the most important person in their lives. We in the myelin research community have lost a most wonderful leader, colleague and friend. We can only feel fortunate that we all knew Pierre, his scientiWc accomplishment, his devotion to young generations of neuroscientists and his mischievous smile, twinkles in his eye, and witty remarks in all aspects of life. Kunihiko Suzuki

Preface

Cajal, Achu´carro, Rı´o Hortega, and the Early Exploration of Neuroglia Pedro Pasik and Tauba Pasik

The subject matter of this collection, namely the details of myelin and myelin diseases, must be played within a neuroglia scenario, given the fact that myelin is a neuroglial derivative. The concept of neuroglia evolved over a period of about 100 years during which the major cell types were characterized and the stage was set for investigations of their biology that led to our contemporary ideas of their roles in nervous system function. Since Santiago Ramo´n y Cajal was the central Wgure in this development, the period may be subdivided into a pre-Cajalian, Cajalian, and post-Cajalian stages. Each of these epochs is driven by the discovery of particular staining methods that contributed signiWcantly to conceptual issues. In this preface, we attempt to trace the evolution of neuroglial typology, which was completed only after the work of Pı´o del Rı´o Hortega. A major point will be to clarify the so-called Cajal-Rı´o Hortega controversy, which has tainted the issue with some bitterness probably derived from the dynamics of interpersonal relations. We shall try to remain outside such argument as much as possible and refer strictly to data that are in the public domain. Most of them, however, have been published in Spanish and are therefore of limited access to the contemporary scientiWc community.

PRE-CAJALIAN PERIOD The methods utilized during this period included the dissociation of tissue, either by mechanical means with needles or by treatment with softening agents of the intercellular material such as chromic acid, weak solutions of potassium dichromate, or iodinated serum, among others. Using one of these methods, Deiters (1865) discovered the neuroglia as a distinct component of the nervous tissue, recognizing that, contrary to the neuron, the glial cell exhibits a single class of processes. He interpreted them as connective tissue cells. Years earlier, Virchow (1846, p. 248) (Fig. 1) had mentioned the existence of a layer of tissue beneath the ependyma of the cerebral ventricles, which he considered as a sort of connective substance. He later repeated and conWrmed the observation, accepting ‘‘the fact that a soft stroma, which is a kind of connective tissue substance, pervades and holds together the central nervous system elements everywhere’’ (Virchow, 1854, p. 138). Finally, ‘‘This connective tissue substance, which is in the brain, spinal cord and higher sensory nerves, is a kind of kitt (neuroglia) in which the nervous elements are planted’’ (Virchow, 1856, p. 890). As correctly pointed out by Somjen (1988), Virchow used the term ‘‘(nerven) kitt,’’ which corresponds better to putty than to glue. In fact, the term ‘‘neuroglia’’ derives

Myelin Biology and Disorders, Volume 1

xxiii

Copyright 2004, Elsevier Science (USA). All rights reserved.

xxiv

PREFACE

FIGURE 1 Rudolph L. K. Virchow (1821–1902) drawing (circa 1849), at the age of 25, was Wrst to mention the presence of a special sticky component of the nervous tissue while working in the autopsy room of the Charite´ Hospital in Berlin, for which he later coined the term ‘‘neuroglia.’’ Courtesy of the University of California at Berkeley.

from archaic Greek, gloia, meaning something sticky or clammy, and not glue as commonly stated. Soon after, Bidder and KupVer (1857) recognized that the cells described by Virchow had Wlamentous appendages, which Ranvier (1873) considered as connective tissue Wbers. The concept of neuroglia as a special component of the nervous tissue, distinct from connective tissue, was suggested by Ko¨lliker (1863), who considered the formation of interstitial plexi as the joining and intercrossing of neuroglial processes. The concept became an established fact only after Boll (1874) demonstrated the diVerences between glia and connective tissue with regard to both staining properties of dissociated tissue, as well as the diVerent histogenesis. At about the same time, other investigators introduced methods of hardening the tissue, which enabled relatively thin sections to be obtained and viewed either without additional treatment or after staining with carmine or basic anilines. Two such techniques are still in use today. The Nissl method, originally with methylene blue (Nissl, 1885), clearly shows the diVerential characteristics of neurons and various neuroglial types mostly on the basis of the nuclear structure. Weigert developed both a technique for staining myelin with hematoxylin (Weigert, 1884) and one for neuroglial Wbrils with methyl violet (Weigert, 1890, 1895). Not all, however, contributed to progress. Weigert, for instance, concluded wrongly that Wbrils were not components of the glial soma and would pass from one to another cell with no apparent terminal tips. This view supported the results of Ranvier (1882, 1893), but subsequently was convincingly refuted by Robertson (1897) and Cajal (see Cajalian Period). A major progress was the introduction of metallic stains, particularly silver chromate that gave rise to the famous black reaction (Golgi, 1872, 1873). Using these reagents, Golgi eventually demonstrated the lack of anastomoses between neuroglial Wbrils, some of the

CAJALIAN PERIOD

morphologic variations in glial cells of the gray and white matter (Golgi, 1886), and the close relations of neuroglial processes with blood vessels (Golgi, 1894). Several other studies contributed signiWcantly to the development of the neuroglia concept in this period. Ko¨lliker (1891, 1896) was Wrst to clearly distinguish two types of glial cells, both in the gray matter of the spinal cord, but it was Lenhosse´k (1891) who showed their precise distribution in the gray and white matters. Contemporary studies included those of Retzius (1891, 1893, 1894), with magniWcent lithographs of both human and nonhuman mammalian brain, and by Andriezen (1893a, 1893b), who coined the terms in use today: ‘‘Wbrous’’ and ‘‘protoplasmic’’ neuroglial cells. It should be noted that Cajal had already envisioned such a distinction (Cajal, 1888a) (see Cajalian Period). The early authors who thought neuroglia was connective tissue had no problem to considering it a mesoderm derivative (Eichhorst, 1875; Golgi, 1872; Ko¨lliker, 1867; Ranvier, 1873; Virchow, 1846, 1854). Later, however, Ranvier (1882) compared the neuroglia of the spinal cord with the retinal Mu¨ller Wbers and considered both to be primitive, undiVerentiated neuroepithelium. His (1890), among others, claimed them to be of mixed ectoderm and mesoderm origin. Andriezen (1893a) was more speciWc, arguing that his neuroglia Wber cell (Wbrous neuroglia) was epiblastic, whereas his protoplasmic neuroglia cell was mesoblastic. An exclusively ectoderm origin was suggested by Boll (1874) and Vignal (1884, 1889) and strongly supported by the Wndings of Nansen (1886) in hagWsh. It was Cajal again that provided the Wnal answer (see Cajalian Period). We should mention that the presence of nodes of Ranvier in central nerve fibers was noted by Tourneaux and Le Goff (1875) in the spinal cord, although their existence was not accepted by Ranvier (1873, 1878) himself, or even by Ko¨lliker (1896). Cajal (1888b) was again first to demonstrate them convincingly in the cerebral lobes of the electric ray (see Cajalian Period).

CAJALIAN PERIOD Cajal came into the glia research picture with his initial attempts with the Golgi technique. After Wrst learning about this method in 1887, he soon published a study on the nervous system of birds (Cajal, 1888a). He demonstrated clear diVerences between the neuroglia in the granule cell layer of the cerebellum and in the underlying white matter, thus preceding the observations of Andriezen (1893a, 1893b) and Ko¨lliker (1896). In the same year, he also demonstrated for the Wrst time the presence of nodes of Ranvier in Wbers of the brain white matter (Cajal, 1888b) and later in the spinal cord and cerebellum (Cajal 1889a, 1889b). He discovered as well that the nodes were the site of origin of axon collaterals and bifurcations (Cajal 1889b, 1890a). More detailed descriptions of nodes of Ranvier were provided with the use of the Ehrlich method (metylene blue) in Cajal’s hands (Cajal, 1896). Cajal deWnitively established the origin of neuroglia from the ectoderm. He demonstrated that all neuroglial elements are just displaced and modiWed epithelial cells, observing all gradations of this transformation in silver chromate preparations of avian and mammalian embryos (Cajal, 1889a, 1890b). These Wndings were amply conWrmed by several investigators (Ko¨lliker, 1891; Lenhosse´k, 1891; Retzius, 1893; Sala y Pons, 1894), and later by Mu¨ller (1900) in a comprehensive comparative and histogenetic study. We must now immerse ourselves into Cajal’s opera magna: the Texture of the Nervous System of Man and the Vertebrates (Cajal, 1999–2002, edited translation). The Wrst volume was originally published in separate fascicles. The chapter on neuroglia appeared in 1897, the chapter on neuroglia of the gray matter of the spinal cord in 1898, and Wnally the growth and development of the ependyma and neuroglia were explored within the chapter on Histogenesis of the spinal cord in 1899. It is in this volume where he referred to neuroglial (stellate) cells as astrocytes, a term that had been introduced earlier by Lenhosse´k (1895, p. 180). The extensive treatment of the subject in the analytical descriptive style followed by the synthetic interpretation, so characteristic of Cajal’s approach, shines in these chapters, and the resulting wealth of information and ideas is indeed remarkable.

xxv

xxvi

PREFACE

Starting with the typology, he clearly distinguished the Wbrous, protoplasmic, and epithelial neuroglial cells, and also accepted the existence of many intermediate forms. He proved beyond doubt that neuroglial Wbers are products of the cytoplasm, that is, intracellular, conWrming and making reference to the observations of Robertson (1897) (see PreCajalian Period). Cajal extended his initial studies on the origin of neuroglia, conclusively demonstrating its ectodermic derivation. All the transitions from the spongioblast of His to the astrocytes are shown in superb drawings of the spinal cord of the chick, mouse, cat, and human embryos. The intermediate stages are represented by the displaced epithelial cell or astroblast with the still present radial peripheral process reaching the pia, and the young neuroglial cell, which, having lost the radial process, is still elongated in the radial direction (Cajal, 1999–2002, Vol. I, Figs. 253–259). Texture is sprinkled with information about the neuroglia of practically every structure of the nervous system. Particularly notable is the treatment of the spinal cord in Volume I, the cerebellum and retina in Volume II, and the cerebral cortex and olfactory bulb in Volume III, which attest to Cajal’s unsurpassable breadth and depth of knowledge. Furthermore, the neuroglia of several components of the nervous system were examined across many diverse vertebrate species. Finally, Cajal expounded his ideas on the possible functions of the neuroglia of the gray matter (Cajal, 1897). He Wrst oVered a rather Werce rebuttal of Golgi’s nutritional theory (Golgi, 1886), which was based on the claimed direct contacts of dendrites with capillaries and neuroglial elements, and of Weigert’s Wlling-in theory, for which neuroglia would have just a passive role of Wlling the spaces left by neuronal somata and their processes (Weigert, 1895). He then sided (for the time being) with his brother P. Ramo´n’s idea of the insulating role of neuroglia in the gray matter with a number of arguments that are still valid today. Lugaro (1907) accepted the supporting and insulating role of the neuroglia, but in addition proposed the novel possibility of antitoxic function, which by migration, engulWng, and histolysis would impede deleterious substances from reaching the neurons, as well as neutralize regressive products of neuronal metabolism. Furthermore, Lugaro utilized Cajal’s neurotropic theory to suggest a possible role of the neuroglia in determining neuronal topographic relations and connections. Although the supporting arguments were indirect, they posed interesting questions that would guide the development of future ideas on the function of the neuroglia (discussed later). Cajal encountered major diYculties when considering the role of the neuroglia of the white matter. He argued that various facts do not support a similar insulating function since neuroglial cells are very abundant in white matter, where myelin already protects against the spread of nerve impulses. Therefore it remained an open question to be answered in what we call the post-Cajalian period (see Post-Cajalian Period). This problem relates to Cajal’s views on the nature of myelin. It is interesting that after long references to opinions on various features of the myelin sheath, Cajal only expresses reservations to them with no clear stand of his own (Cajal, 1999, pp. 232–234). By 1910, Cajal (Fig.2) had extracted as much information as he could from the Golgi silver chromate method and his own variants, particularly the double impregnation. His newly devised reduced silver nitrate method, a transparent technique as opposed to the opaque Golgi one, contributed signiWcantly to the information on neuronal structure. These contributions formed the main addition to the original Textura incorporated into the French edition, the Histologie. However, being basically a neuroWbrillary stain, the silver nitrate did not provide much information on neuroglia. This period is dominated by Cajal’s Wercest Wghts against the neoreticularists (Bethe, Apa´thy, Hensen-Held) and by his studies on nerve regeneration and the chemotaxis or neurotropic hypothesis. By this time, Cajal had been awarded the Nobel Prize. He was the director of the Laboratory of Biologic Research with his own journal, and the president of both the Junta de Ampliacio´n de Estudios (a mixture of National Science Foundation, the National Institutes of Health, and graduate school) and of the Spanish Biological Society (in perpetuity). He was, indeed, a towering Wgure and a very busy man.

CAJALIAN PERIOD

FIGURE 2 Santiago Ramo´n y Cajal (1852–1934; photo circa 1912), the founder of modern neurobiology, contributed major fundamental observations in glia research throughout his entire career. Courtesy of the heirs of S. R. y Cajal.

FIGURE 3 Nicola´s Achu´carro (1880–1918; photo circa 1910) established in a few years the basis of neuroglia histopathology and raised the category of this tissue, adducing its possible secretory role as the basis for certain mental disorders. Courtesy of La Gran Enciclopedia Vasca, Bilbao.

xxvii

xxviii

PREFACE

At about the time that the neuroglia became the focus of Cajal’s laboratory, a new star appeared in his laboratory. He was Nicola´s Achu´carro (Fig. 3), whose background was indeed impressive. He had extensive training in Simarro’s laboratory (where Cajal saw for the Wrst time a Golgi preparation), and he had spent time with Pierre Marie and Babinski in Paris, Lugaro and Tanzi in Messina and Florence, Kraepelin and Alzheimer in Munich. He was then hired to direct the histopathology laboratory in the Government Hospital for the Insane in Washington, D.C., and Wnally settled in Madrid as chief of a psychiatry hospital service. Achu´carro’s Wrst publications in French, German, and English dealt with the histopathology of neuroglia in human disease and after experimental lesions. He found that the abnormal cells discovered in cases of general paresis and named rod cells (Sta¨bchenzellen) by Nissl (1899) were abundant in the stratum radiatum of the hippocampal gyrus. Moreover, he observed that during the repair phase after experimental stab wounds of the brain, there were many granulo-adipose cells, equivalent to the reticulated cells of Nissl, or Gitterzellen, as well as intermediate forms between the latter and the rod cells. Achu´carro (1908, 1909) thought that all of these cells were forms of neuroglia with phagocytic properties and of probably of dual ectoderm and mesoderm origin. He interpreted the shape of rod cells as an adaptation to the environment of the damaged neurons. This idea had been already advanced by Cerletti (1905), who proposed that neuroglia around the soma of pyramidal cells were expressed as satellite cells whereas those against the dendrites became rod cells. Soon after, Achu´carro published results of a major study he carried out in Alzheimer’s laboratory, on the histopathology of rabies with very elegant illustrations of rod cells with the Nissl and Cajal’s reduced silver nitrate methods (Achu´carro, 1910a). He reported similar Wndings in cases of senile dementia with the Bielschowsky method. In that report he described the various stages of the neuronal neuroWbrillar tangles, the comparable changes in the glia surrounding the neurons as well as glial components of senile plaques. He proposed the idea of an encrustation of a pathologic product upon diVerent structures: glia, pericellular reticulum, and intracellular neuroWbrils (Achu´carro, 1910b). He then concentrated on the repair process after stab wounds of the hippocampal gyrus. A clear distinction appeared between the elongated and rod cells, on the one hand, and the Wbrous astrocytes, on the other. Whereas the latter related to blood vessels and did not accumulate products of degeneration, the former depended on the orientation of neuron processes and did enclose degeneration material (Achu´carro, 1910c). This view was supported by a next case of general paresis where he observed with the reduced silver nitrate method that rod cells containing silver stained inclusions are attached to pyramidal cell dendrites (Achu´carro, 1910d). He considered the rod cells as a special form of phagocytic cells, the shape of which adapted to the degenerating neuronal structures. It is of interest that in this article he states that the scanty protoplasm of rod cells blends into the neuroglial syncytium. The notion of this syncytium or diVuse interstitial net was popular among investigators at the time (Hardesty, 1904; Held, 1903, 1909), but not with Cajal. After his return from the United States, Achu´carro joined Cajal’s teaching and research endeavors. His next publication was on cases of arsenic poisoning, sleeping sickness, and experimental injections of sporotrichosis in the rabbit, using Cajal and Bielschowsky methods with Wxation variants of his own. He found an abundant neuroglial proliferation with many elongated cells and even ameboid cells in the white matter, which he generally called interstitial cells, with no evidence of a syncytial formation (Achu´carro, 1911a). He greatly praised Cajal’s method and anticipated that some selective variants of it ‘‘will allow the diVerentiation of the various neuroglial classes,’’ a prediction that was to be fulWlled a decade later by his most notable disciple, Rı´o Hortega, during the post-Cajalian period. About this time, Achu´carro (1911b) developed his own modiWcation of the Bielschowsky method using tannin as a mordant, and he applied it to the brain of a general paresis case (Achu´carro, 1911c). According to one of his biographers (Vitoria Ortiz, 1977), Achu´carro gave a series of invited courses at Fordham University in New York and was awarded a doctorate honoris causa at Yale University. The latter, however, cannot be veriWed at Yale archives, but the

CAJALIAN PERIOD

list at Fordham does contain a Sc.D. awarded to Achu´carro in 1912. On his return, he became chief of an independent histopathology laboratory under the auspices of Cajal and the Junta de Ampliacio´n de Estudios (mentioned earlier). Although located at the Museum of Natural History, the work of his laboratory was intimately related to Cajal’s. Achu´carro published extensively in Cajal’s journal and was considered by Cajal as one of his outstanding disciples. By that time, Cajal (1912) had introduced a variant of his reduced silver nitrate method, changing the Wxative to formalin-uranyl nitrate. With this modiWcation he obtained complete staining of protoplasmic astrocytes with all their processes forming complicated plexi but ending freely, in what he called a polygenic plexus as opposed to Held’s diVuse plexus or syncytium. At this time, Achu´carro (1913a, 1913b) started to study the normal human cerebral cortex with Cajal’s and his own tannin-silver method and obtained very elegant impregnations of protoplasmic astrocytes with enclosed granules, variously interpreted as mitochondria or as special gliosomes (Fieandt, 1910). He also clearly showed the absence of anatomoses, which reinforced his and Cajal’s views that Held’s neuroglial syncytium hypothesis was incorrect. Furthermore, he oVered a new view on the structure of the perivascular end-feet, which were found not to form a continuous membrane. Finally, Achu´carro advocated a multiple function for the neuroglia. In addition to phagocytosis and scar formation in pathologic conditions and the supporting and, in some locations, insulating role under normal conditions, the normal glia could have a secretory function. This idea had been introduced earlier by Nageotte (1910), but Achu´carro enlarged on it to include the possibility that the presumed secretion could form a communication between the nervous system and the blood vessels, as if the neuroglia were an endocrine gland. At the same time, Cajal had made a major breakthrough with the development of a new technique that was extremely selective for impregnation of the glia of both the white and gray matter (Cajal, 1913a, 1916, 1920b). It consisted of treating the tissue with a mixture of gold chloride and mercury chloride or corrosive sublimate, in short a gold-sublimate method. Already then Cajal noticed that the technique never stained neurons or connective tissue such as the vascular adventitia. He became convinced that the majority of the socalled satellite elements that remained unstained were not glial in nature. This communication was followed by a comprehensive monograph on the neuroglia of the human brain (Cajal, 1913b). It contains a thorough discussion of his previous views on the uniWed ectodermic origin of both protoplasmic and Wbrous astrocytes, the absence of a neuroglial syncytium, and the presence of perivascular end-feet in all astrocytes. In addition, Cajal oVered a generalized concept of the satellite cell, which included those protoplasmic astrocytes applied to the soma and dendrites of cortical neurons, as well as Schwann cells and intracapsular cells of sensory and autonomic ganglia. He asserted that such elements form a kind of symbiosis with the neurons from which both partners derive trophic and functional beneWts. A most signiWcant issue of this publication is the description of what he designated as the ‘‘third element’’ of the nervous system. Also named as apolar or adendritic cells, they were not neurons and, since they remained unstained by the gold-sublimate method, they were not (astro)glia. Although not fully appreciated at the time, these were not a single cell type, but a mixture of microglia and oligodendrocytes, among others (see PostCajalian Period). They were located around neurons (‘‘nonglial perineuronal satellites’’), near glial cells (‘‘satellites of the glia’’), along blood vessels (‘‘perivascular satellites’’), and in the white matter, where they were very abundant forming series of cellular packets or columns. Cajal argued extensively about the nonglial nature of the third element, which he tended to consider, albeit provisionally, as of mesoderm origin. He also correlated the regional abundance of apolar cells with the richness of particular areas in myelinated Wbers, oVering with great caution various reasons for their homology with Schwann cells. Finally, he observed cell divisions of the apolar elements as well as phagocytic phenomena in experimental brain wounds. Cajal’s third element had been seen earlier and variously named as naked nuclei, round nuclei, cuboid cells, pre-ameboid cells, and indiVerent cells. They were interpreted either as special cells with processes unstainable with the techniques for neuroglia (Held, Alzheimer,

xxix

xxx

PREFACE

Fieandt, Jakob) or as cells with no processes, of germinal or undiVerentiated nature, diVerent than astrocytes (Bevan Lewis, Nissl, Robertson, Bonome, Rosenthal). For review of these antecedents, see Rı´o Hortega (1920). Cajal’s discoveries on neuroglia were Wrst conWrmed outside Spain by the Hungarian histologist SchaVer (1915). The gold-sublimate method was successfully applied to the neuropathology of general paresis and senile dementia (Achu´carro and Gayarre, 1914a, 1914b). It was again clear that this method failed to stain the rod cells in the stratum radiatum of the hippocampal gyrus, which were so characteristic of general paresis and experimental rabies. These cells were then interpreted as having a mesoderm origin coming from blood vessel walls, as Alzheimer (1904) had suggested, as well as migratory properties. The method provided no evidence for the participation of neuroglia in the Alzheimer lesions of senile dementia. Near the end of the Cajalian period, Achu´carro (1915a, 1915b) published two major articles using both the gold-sublimate and tannin-silver methods. In one he described the distribution of the classic types of neuroglia (astroglia) in the human, monkey, dog, and rabbit hippocampal and dentate gyri, elaborating so-called gliotectonic maps, which followed quite closely the cytoarchitectonics of these regions, at least in the primate. There were also some preliminary observations on the development of neuroglia in a 3-month-old infant and 10- to 90-day-old kittens. The second paper deals extensively with evolutionary aspects of neuroglia in submammalian species, with special emphasis on the perivascular end-feet. These end-feet first appear in phylogeny in reptiles, although neuroglia in these animals is generally still of the ependymal type. This arrangement is preceded in Wsh by pial feet of similar structure and in amphibians by simple insertion of processes of ependymal cells on blood vessels. Finally in birds, the so-called autonomous neuroglia, which derives from the ependymal neuroglia, gives rise to fully developed vascular feet extending from processes of such cells, in all similar to those found in mammals. Regarding the neuroglial and non-neuroglial satellites (Cajal’s terminology) the former are seen only in birds where glial cells are totally independent of the ependyma (autonomous neuroglia). The non-neuroglial satellites or third elements are observed for the Wrst time in reptiles within large white commissures and the optic chiasm. Finally, Achu´carro could not refrain from elaborating on a possible role in psychiatric diseases (he was a practicing psychiatrist in addition to a histologist and histopathologist) and speculated that protoplasmic neuroglia might play a double role: as an important component for nervous system function locally and as an endocrine gland inXuencing processes of aVective nature (i.e., emotions). It was his last publication. After a long illness, he died 3 years later at the age of 37. One other signiWcant contribution using the Cajal and the Achu´carro methods dealt with the neuroglia of the cerebellum (Fan˜anas, 1916). A protoplasmic astrocyte was discovered in the molecular layer, as well as two types of satellites to Purkinje cells, namely the classic astrocytes and the adendritic or third elements. The latter were also seen around displaced Golgi cells and were particularly abundant between nerve Wbers of the white matter. The author of these studies was Cajal’s son who, in the tradition of his father, eventually chose to be known by his mother’s last name. In summary, the methods of this period were essentially metallic stains, with progressive selectivity for neuroglia: silver chromate, double impregnation, platinum chloride, reduced silver nitrate, tannin-silver, ammoniacal silver, and gold chloride-corrosive sublimate, with various Wxatives, such as formalin-uranyl nitrate and formalin-ammonium bromide. These methods yielded a remarkable number of deWnitive observations in Cajal’s hands and those of his disciples, particularly Achu´carro. They allowed a sharp deWnition of the two classic neuroglial types (protoplasmic and Wbrous astrocytes) as well as intermediate forms, and to determine the intracellular nature of glioWbrils, the absence of a neuroglial syncytium, and the ectodermic origin of neuroglia. The investigators of this period deWned the special features of the glia in various structures of the nervous system, the presence of nodes of Ranvier in central Wbers, and gave an extensive description of vascular end-feet. Furthermore, their theoretical interpretations included a generalized concept of satellite cells and the probable functions of neuroglia: support, insulation and secretion, the latter acting as an interface between neurons and the bloodstream. It was demonstrated that the rod cells and gitter cells found in human pathology and in the repair process after

POST-CAJALIAN PERIOD

experimental lesions were of glial nature. Lastly, they established the presence of a ‘‘third element’’ (non-neuronal, nonglial) in the nervous tissue, which was later known to be a mixture of cell types that included microglia, oligodendrocytes, and others. They did not fully recognize the oligodendrocyte as a component of the third element or assign to it its critical role in myelin formation.

POST-CAJALIAN PERIOD Putting together the fundamental discoveries of the previous period, it becomes evident that in addition to the already classic types of glia, Wbrous and protoplasmic, there was another element diVerent in morphology and staining properties. Although it had been seen before and given various descriptive names (see Cajalian Period), Cajal brought it to the foreground and speculated on its nature. There were indeed discrepant issues. His goldsublimate method did not reveal the existence of processes; therefore he designated these cells as apolar or adendritic cells. Some were neuronal satellites in the gray matter; others were exceedingly abundant in the white matter. In any event, what had the ‘‘third element’’ to do with the rod cells, gitter cells, and intermediate forms found in pathologic cases? Was it of ectoderm or mesoderm origin, or both? What, if any, was its relationship to myelin? All of these questions were subjects of speculations and very cautious opinions. They were Wnally answered with pristine clarity with the development of the new silver carbonate method (Rı´o Hortega, 1917). In the words of Rı´o Hortega (1949): ‘‘Those who invent no techniques rarely succeed in making discoveries of importance.’’ In the best Cajalian tradition, he doggedly tried modiWcation after modiWcation of methods to selectively stain one or the other cell type. Thus, while working full time in Achu´carro’s laboratory, which had moved to the same building as Cajal’s, but functioned with some autonomy, Rı´o Hortega (1916b) developed four variants of Achu´carro’s method. The Wrst and third were excellent to stain the Wbrous astroglia and glial Wbers, and the fourth was good for impregnating the protoplasmic neuroglia. In this way he demonstrated that in both invertebrates and vertebrates the neuroglial protoplasm is made of a Wbrillar plexus almost the same as the protoplasmic reticule. He believed that the glioWbrils originated by progressive diVerentiation of preexisting Wbers of the reticule (Rı´o Hortega, 1916b), a view previously advanced by Cajal (see Cajalian Period). That same year, he reported the neuroglial changes occurring in the human brain after ischemic infarcts with a wide range of survival times (Rı´o Hortega, 1916a). He noted that, in addition to the more classic alterations of the astroglia, there was an increased number of ‘‘third elements,’’ which frequently contained pigment accumulation. These cells showed phagocytic capacity even at the earliest stages (8 hours post-infarct), so that all transitional forms to the gitter cells could be observed. These Wndings proved that such type of cells, of most probable mesoderm origin as Cajal had suggested, preexisted in the nervous system and needed not arrive from the outside. This was an important antecedent to his later discoveries. The new silver carbonate method with various timings of the formalin-ammonium bromide Wxative introduced by Cajal (1913a, 1916, 1920b), however, allowed him to sharpen the characteristics of classic neuroglia, particularly the pathologic alterations such as the hypertrophied neuroglia, ameboid cells, and moniliform changes in glial Wbers. So far, Rı´o Hortega kept within the framework of Cajal’s and Achu´carro’s thinking. He would soon revolutionize that thinking. It took, however, some time, torn as he was between publishing his Wndings and the respect for Cajal’s description and interpretations of the ‘‘third element.’’ Finally, four communications to the Spanish Biological Society appeared as unusually long articles in the bulletin of that Society (Rı´o Hortega, 1919a, 1919b, 1919c, 1919d), which were later essentially reprinted as a single publication in Cajal’s journal (Rı´o Hortega, 1920). First, he demonstrated that the so-called apolar or adendritic cells had processes that, because of their nature, did not stain with the gold-sublimate method. Furthermore, two diVerent types were included in the third element on the bases of size, nuclear appearance

xxxi

xxxii

PREFACE

and predominant distribution in the gray and white matter, naming them microglia and interfascicular (oligodendro-) glia, respectively. Rı´o Hortega then concentrated on the detailed description of the microglia in all vertebrate classes including the human, stained selectively by precise timing of heated formalin-bromide Wxative and heated ammoniacal silver carbonate. Although the morphology of microglia varies as an adaptation to surrounding tissue, a general characteristic is the absence of glioWbrils and gliosomes. Next, utilizing human and animal neuropathologic cases, as well as experimental lesions in the rabbit, dog, and pigeon, he outlined in minute detail the successive steps in microglia changes. The normal cells increase in number by migration and proliferation, develop phagocytic properties, and change their conWguration to become rod cells of varied shapes or granulo-adipose cells (gitter cells), depending in great part on the route of migration and Wnal location. Thus, rod cells appear along Wber bundles and apical dendrites of cortical neurons, whereas granulo-adipose cells require space such as in ischemic infarcts. Finally, he oVered ample evidence regarding the mesoderm origin of microglia and introduced the ill-fated term of mesoglia. He concluded that the microglia was the third element of the nervous tissue. The year 1920 was tumultuous in the master-disciple relationship. Cajal had been guiding and constantly inspiring Rı´o Hortega with advice and support, and many of his predictions and interpretations as well as technical nuances permeated Rı´o Hortega’s work. Moreover, the fact that he accepted the full paper in his journal in spite of the serious objections to his own concepts about the third element attests to Cajal’s openmindedness. While the article was in press, there was an exchange of letters (Rı´o Hortega, 1986) in which Rı´o Hortega, who had stopped attendance to the laboratory, requested that some action be taken against a so-called concierge (a mixture of technical helper and janitor) who was interfering with his work by limiting the number of hours he had access to the laboratory. Cajal’s answer ignored the issue, but he, in turn, complained about the crowding of the laboratory by Rı´o Hortega’s students. He suggested that Rio Hortega move his setup to another temporary location with the promise of a large facility in a future building. Rı´o Hortega reiterated his complaint in more detail, regretfully accepted Cajal’s decision, and continued his absence from the laboratory. Finally Cajal apparently recalled him with assurances that the concierge would apologize or be Wred. Rı´o Hortega returned to the laboratory, but not for long. After the summer vacation, Cajal Wred him deWnitively on the basis of accusations by third parties, with many reminders of the beneWts derived from his association. The tone of the letter transpires in the Wnishing line: ‘‘Your ex-friend and ex-protector addresses you for the last time.’’ This was followed by Rı´o Hortega’s rebuttal to each accusation and beneWt reminder, and a request for witnesses of his alleged transgressions, ending with: ‘‘Your always devoted and thankful disciple.’’ Cajal, then, softened a little, obtained a working place in the Students Residence,1 allowed the use of the library, renewed his promises for the future, and ended with ‘‘Your true comrade and friend.’’ The tension had apparently subsided when two publications appeared soon after Rı´o Hortega’s (Cajal, 1920a, 1920b). In the Wrst one, which Cajal considered a critique more than a descriptive study, he introduced the notion of Robertson’s priority (1900a) in the description of so-called mesoglial cells and equated these cells to the interstitial cells of Achu´carro (see Cajalian Period). He accepted the existence of the microglia with the characteristics fully reported by Rı´o Hortega, giving elegant examples of his own in the gray and white matters of the cerebellum, cerebrum, and spinal cord. Cajal, however, insisted on and showed the presence of still quite a number of apolar or adendritic components of his third element in the same regions of the gray matter (dwarf satellites) and white matter (interfascicular cells of Rı´o Hortega). He cast doubts on the constancy of the Wne and delicate projections of the latter cells, as had been described by Rı´o Hortega. 1 The Students Residence (Residencia de Estudiantes), far from being just a students’ dormitory as its name would imply, was a kind of small college, with teaching and modest research facilities for both science and the humanities. It was frequented and even lodged personalities such as the philosopher Miguel de Unamuno and the poet Juan Ramo´n Jime´nez. The Histology laboratory was actually created by Achu´carro.

POST-CAJALIAN PERIOD

The article ended with an extensive critical discussion on the relative value of negative Wndings, pointing out in this case that the lack of staining of his dwarf satellite cells separated them from the microglial and astroglial satellites. The second article (Cajal, 1920b) described the variant of the Bielschowsky method that he used to stain the microglia. The procedure does not diVer much from Rı´o Hortega’s, since the critical features of heating the Wxative as well as the silver solution remain the same, and only sodium hydroxide is used instead of sodium carbonate to obtain the ammoniacal silver solution. Moreover, Cajal argued that heating of the silver carbonate releases CO2 and what remains is the same silver oxide of his variant. In any event, he admitted that the results were not so selective as those of Rı´o Hortega’s method. Unfortunately, Cajal’s indication of Robertson’s priority was based on a reference by Cerletti (1907–1908) to a single-page communication of that author with no illustrations. In fact, Robertson (1899) had developed an impregnation method with platinum dichloride that showed the existence throughout the dog and human cortex and white matter of nonglial, non-neuronal cells. A full description appeared in his textbook (Robertson, 1900b). Here he stated that these cells ‘‘commonly have three to six delicate processes which branch dichotomously, sometimes dividing three or four times,’’ with no connections to blood vessels or other structures. Few of them had no processes at all. The almost 20 Wgures oVered, some of which had appeared earlier in the technical paper of 1899, leave no doubt that Robertson’s cell has nothing to do with Rı´o Hortega’s microglia. What becomes somewhat confusing, however, is Robertson’s opinion, with no corresponding illustrations, that in pathologic conditions his cells lose their branches, proliferate, act as phagocytes, become granular cells, and eventually ‘‘amyloid bodies.’’ It was on this basis that he regarded them as of mesoderm origin, therefore the term ‘‘mesoglia’’; unfortunately the same name proposed later by Rı´o Hortega for his microglia. The designation was, therefore, the only common feature of the two cells. Once in his own laboratory at the Students Residence, Rio Hortega worked frantically, but the resulting publications no longer appeared in Cajal’s journal. In fact, Rı´o Hortega (1921a) stated that although the observations were made at the Cajal Institute up to the summer of 1920, because of events beyond his control he felt forced to delay the publication. In any event, he Wrmly established the mesoderm origin of microglia arising in the perinatal period and showed that they were derived from the pia, tela choroidea, and adventitia of large blood vessels, without ruling out some leucocytes as a source. In addition, he described the various shapes adopted by the microglial cells in their accommodation to the interstices of the nervous tissue during normal development: globular, ameboid, pseudopodic, and Wnally ramiWed. Furthermore, he recounted in minute detail the migration and distribution of the microglia in every single region of the gray and white matter of the cerebrum, cerebellum, brain stem, and spinal cord of newborn and adult primates, carnivora, lagomorpha, and rodents. This very important publication ends with a discussion of Robertson’s mesoglial cells and presents convincing evidence that such elements are not microglia but examples of oligodendroglia (discussed later), and therefore of ectoderm origin. Lastly, there is a refutation of Cajal’s claim of Robertson’s priority for the discovery of mesoglia and an expression of conWdence that Cajal’s impartial reexamination of Rı´o Hortega’s and Robertson’s preparations would lead him to Rı´o Hortega’s conclusions. In a letter to Rı´o Hortega (1986 p. 120), Cajal hoped that after a careful review of the problem he would be able to fully recognize and proclaim Rı´o Hortega’s priority of the microglia discovery. The same year, Rı´o Hortega came out with a full description of what he called glia of scarce radiations or oligodendroglia (Rı´o Hortega, 1921b). By reducing the initial Wxation time and using a seven-fold increase in the strength of the silver carbonate solution, he selectively impregnated this new cell type. In addition to a description of the morphology, the report showed that it was present in huge numbers in all regions of the central nervous system, although with a strong predominance in the white matter. They were frequently grouped near neurons (neuronal satellites), along blood vessels (vascular satellites), and formed long series along nerve Wbers (interfascicular glia). Furthermore, these cells rapidly increased in number at about the time of birth, when they were numerous in the white

xxxiii

xxxiv

PREFACE

matter and contained large gliosomes, coinciding with a period of active myelinogenesis. The morphology of these cells actually resembled those described by Robertson (1900a, 1900b) as mesoglia, and, therefore, Rı´o Hortega was quick to point out the errors of Robertson’s interpretation. Soon after Rı´o Hortega (1922) credited Cajal (1913b) with the idea of similarity between neuronal satellites, intracapsular cells of ganglia, and the apolar cells of the white matter, and argued that the latter are the same as his interfascicular glia (i.e., a type of oligodendroglia). Furthermore, he suggested that the membrane indicated by Cajal around central myelinated Wbers is in fact a derivative of oligodendroglia cells, with functions probably similar to the Schwann cells of peripheral myelinated Wbers. Soon after Rı´o Hortega’s, a second communication to the Biological Society by Castro and Lorente de No´ (1923), young disciples of Cajal, appeared, challenging Rı´o Hortega’s conclusions. They agreed, however, with the view that oligodendroglia is another type of neuroglia of ependymal origin based on data from the olfactory bulb. Apparently, this adversarial position occurred frequently when some investigators of the Cajal Institute discussed Rı´o Hortega’s work, at times using terms ‘‘way beyond objectivity and ethical conduct’’ (Ochoa, 1986 p. 30). In a Festschrift dedicated to von Monakow, Cajal (1923a) updated technical details on his methods for staining neuroglia, namely gold/sublimate, ammoniacal silver oxide with formalin/bromide Wxation, and formalin/uranyl nitrate. As examples, he distinguished between astrocytes, an ameboid cell of probable leucocytic nature with rare processes (microglial satellite cells of Rı´o Hortega), and a third element still unstained by any method that had little protoplasm, no processes, and was diVerent from microglia and neuroglia. Although he recognized that results with silver carbonate were more constant in staining the ameboid cells, he ignored the possibility that the oligodendroglia (Rı´o Hortega, 1921b, 1922) might represent one component of his third element. It is of interest, however, that in making comments on his own article of 1920, Cajal not only added a question mark after the name of Robertson in the title, but also considered the microglia as one of the most valuable contributions of the Spanish school. Yet he attributed the initial discovery to Achu´carro, albeit in pathologic cases, and recognized the merit of Rı´o Hortega in revealing the general occurrence of microglia, its various shapes in the cerebrum, the developmental stages, and the mesoderm origin (Cajal, 1923b, p. 336). The year 1924 was one of deWnitions and conWrmations. Rı´o Hortega (1924) (Fig. 4) clariWed the nomenclature by considering the microglia, with its characteristic morphology, mesodermic origin, and migratory and phagocytic properties, as the true third element of the nervous tissue. The oligodendroglia, of ectodermic origin, sedentary attributes, and probable myelinogenic properties ought to be considered as a variety of neuroglia. The article ended paying ‘‘homage of admiration to Cajal, always our Master, whose prodigious intuition about the third element oriented our investigations, thus allowing to extend its knowledge.’’ Ample recognition of microglia cells outside Spain was Wrst provided by Metz and Spatz (1924) who, although not subscribing to its mesoderm origin, designated them as ‘‘Hortega’s cells,’’ a name that later succumb to the general rejection of eponyms. Finally, PenWeld (1924), working in Rı´o Hortega’s laboratory, fully conWrmed the original description of oligodendroglia. Adding ethyl alcohol as a mordant, this investigator emphasized the resemblance of oligodendroglia to astroglia, except that the former never developed Wbers and had no perivascular end-feet. The importance of this publication lies in its being the Wrst description of these cells in the English language, thus contributing signiWcantly to broadening the availability of information of both microglia and oligodendroglia to the scientiWc community. On returning to the United States, PenWeld adapted Rı´o Hortega’s methods to systematic neuropathology at the Presbyterian Hospital in New York, which included the description of acute swelling of oligodendrocytes in toxic states (PenWeld and Cone, 1926). He eventually collaborated with Rı´o Hortega in a study on experimental brain wounds and scar formation, where they showed the early and prolonged phagocytic activity of microglia and the astrocytic reactions of amitotic division and rapid development of Wbers arranged radially about the wound (Rı´o Hortega and PenWeld, 1927). In the meantime, Rı´o Hortega (1925) developed a speciWc technique that revealed the granulations described by Fieandt, Cajal, and Achu´carro (see Cajalian Period), which are

POST-CAJALIAN PERIOD

FIGURE 4 Pı´o del Rı´o Hortega (1882–1945; photo circa 1924) discovered the microglia and oligodendroglia, of mesodermic and ectodermic nature, respectively, consolidating the phagocytic role of the former and the myelinogenic function of the latter. Courtesy of Editores Asociados, Mexico.

present in all astrocytes and oligodendrocytes, and at times showed a distinction between mitochondria and gliosomes, thus resurrecting the idea of a neuroglial secretory role. He called special attention to the abundance of gliosomes in oligodendrocytes, particularly in the early post-natal period that coincided with active myelinogenesis. In the same year, Cajal (1925) published an exhaustive study of changes in cerebral and cerebellar neuroglia in general paresis that included the disappearance of Fan˜anas cells and their possible transformation into Golgi epithelial cells. It is here were he amply recognized the exactness of Rı´o Hortega’s Wndings, and in a footnote (p. 181), Cajal stated: ‘‘We initially believed, based on a summary by Cerletti, that Robertson was Wrst in the discovery of normal microglia; but today, better informed, we have changed our opinion. The majority of the mesoglial cells described by the English (sic) author, are a variety of macroglia designated by Rı´o Hortega as oligodendroglia.’’ In a short communication, Cajal (1926) gave further technical details on the staining methods used in the previous investigation and Wnally wrote a personal letter to Rı´o Hortega (1986, p. 147) where he stated: ‘‘My judgement concerning the priority of discovery of those two glial types has changed radically since I read Robertson’s book. . . . I had been mislead by the critical references of the Italian histologists.’’ He then closes with an expression of ‘‘love and admiration from your devoted friend and comrade.’’ The deWnitive publication on oligodendroglia is undoubtedly that of Rı´o Hortega, which appeared in 1928. Here again, in the best Cajalian tradition, he developed yet another modiWcation of the Golgi method, this time adding chloral hydrate to the potassium dichromate/formalin Wxative, thus obtaining excellent, although not exclusive, impregnations of oligodendroglia. The report is amply illustrated, not only with drawings,

xxxv

xxxvi

PREFACE

but also for the Wrst time with abundant photomicrographs. It shows the morphological and structural features, as well as the varieties of oligodendrocytes with gradations from satellite cells in the gray matter, through three other cell types in the white matter, ultimately approaching the appearance of the Schwann cells of peripheral myelinated Wbers. He also reported their pathologic alterations and conWrmed the ectoderm origin shared with the astroglia, as shown previously by PenWeld (1924). Most important, Rı´o Hortega demonstrated that oligodendrocytes acquire close relationships with nerve Wbers forming delicate wrappings of laminar, fenestrated, or reticulated appearance, reinforced by rings and interannular trabecules. He concluded that all of these structures are morphologically equivalent to those present in the cells and sheaths of Schwann of peripheral nerve Wbers. Although circumstantial, the evidence pointed to the myelogenic role of the oligodendrocytes. This publication contains also the following statement: ‘‘It is our wish that Cajal, our master, see in the investigations on oligodendroglia the desire to Wnd out the scientiWc truth with no detriment to the prestige of his school, and the humble contribution to the solution of a problem of his preferred interest. And that the venerated master see in this publication, which conWrms one of his most astute intuitions, an homage of loving respect.’’ Soon after, Cajal sent a letter (Rı´o Hortega, 1986, p. 146) containing the following: I have quickly glanced over your new work on Oligodendroglia and became convinced, not only of the reality of this glia type, but on the many morphologies that it adopts. Particularly interesting are the connections of the processes with myelinated axons. It always seemed to me that there must be something in the neural centers to enclose or contain the myelin. Otherwise, it is inconceivable how it does not squeeze through the interstices of the nervous structures.

Do these statements close the occasionally bitter but always mutually respectful exchange between master and disciple? It is diYcult to say. On the one hand, to overcome the language barrier, Cajal published a much needed French translation of his classic monograph of 1913 (Cajal, 1932a, 1932b), with some footnotes added. One of them (pp. 436–437) reads: Regarding the form and signiWcance of my enigmatic apolar cells (third element), Rı´o Hortega has brilliantly elucidated the problem in a series of magniWcent works between 1920 and 1930. Microglia is the third element of the gray matter, and oligodendroglia represents the third element of the white matter.’’

On the other hand, the term oligodendroglia is not mentioned at all in a later edition of Cajal’s textbook of Histology (Cajal, 1933). By this time, both PenWeld (1932) and Rı´o Hortega (1932) had published extensive reviews in English, the latter including the detailed distribution of microglia in various structures of the central nervous system, from the cerebral cortex to the spinal cord. Cajal died in 1934 at the age of 82, forever recognized as the founder of modern neurobiology. The vacuum was Wlled in part by his direct disciples, particularly Tello, Castro, and Lafora, but the center of gravity for glia research stayed with Rı´o Hortega. He became director of the Cancer Institute in Madrid and eventually was given ample facilities in a new building at the periphery of the University City. It was one of the Wrst hits of bombing by Hitler’s LuftwaVe. Rı´o Hortega left for Valencia, then moved to Paris, and Wnally to Oxford where he was invited to organize a neuropathology laboratory. He stayed for 3 years during which he was awarded a D. Sc. honoris causa. One of his lectures on microglia appeared in The Lancet (Rı´o Hortega, 1939) and received an editorial comment to the eVect that ‘‘A great service has been done by Wilder PenWeld in making the English-speaking people more familiar with the recent Spanish work, but it is even better that we are able to publish a lecture given in this country by Hortega himself.’’ In the ensuing years, glia research intensiWed, but with rare exceptions, not much was added to the classic descriptions of the two new types, microglia and oligodendroglia. It is noteworthy, however, to mention the work of Kershman (1939), which deWned in the human embryo certain constant sources of the migrating youngest forms of microglia, namely at points where the choroid plexus is attached to the brain, and around large blood

POST-CAJALIAN PERIOD

vessels near tracts and beneath the meninges. In fact, he replicated in human material, the already classic histogenetic Wndings of Rı´o Hortega (1921a). Rı´o Hortega’s exile continued in Buenos Aires, where under the auspices of the Spanish Cultural Institute and the help of a distinguished neuropathologist, Moise´s Polak, he established a Laboratory of Normal and Pathologic Histology, published his own journal, and, in a few years, developed a distinctive school. Major publications came out in 1942, all in the new Archivos de Histologia Normal y Patolo´gica. The Wrst was the most complete account of neuroglia to that date, where in addition to the factual observations on central astroglia and oligodendroglia, he proposed so-called histophysiologic concepts of both, introducing the terms of ‘‘angiogliona’’ and ‘‘neurogliona,’’ respectively (Rı´o Hortega, 1942). ‘‘Angiogliona’’ referred to the constant association of astrocytes with neurons, on the one hand, and blood vessels, on the other, with possible multiple functions: mechanical, trophic, secretory, and antitoxic. Of particular interest is the possible ‘‘secretory’’ function, the product of which could act on the neuron, contributing to its normal function and being responsible, by over or under secretion, for neurologic and psychiatric alterations of diYcult explanation at the time. The concept of ‘‘neurogliona’’ was based on the constant association of the oligodendroglia with central myelinated Wbers. It was also attributed multiple functions, but in this case, mechanical, trophic, and most important myelinogenic. He was cautious enough to state that although observations supported the formation of myelin by oligodendrocytes, either directly or by supplying nerve Wbers with needed materials, they were not deWnitive. He concluded that the oligodendroglia is a variety of macroglia, symbiotically associated with nerve Wbers, forming central ‘‘neuroglionas,’’ which are homologous to the peripheral ‘‘neuroglionas’’ represented by the Schwann cells. The concept was therefore extended to include what he called peripheral neuroglia—that is, the Schwann cells and the intracapsular and extracapsular cells of sensory and autonomic ganglia. Such cells could be reduced to highly diVerentiated forms of oligodendroglia, all representing symbiotic arrangements of oligodendrocytes and neurites with similar hypothetical functions. Rı´o Hortega’s description of the Schwann cell is indeed perfect: a unit closed at the level of the nodes of Ranvier, with a reticular protoplasm, a system of intramyelin funnels, and perhaps a deep protoplasmic lamina in immediate contact with the axon. The original reports of Cajal and Olo´riz (1897) and Lenhosse´k (1907) concerning the presence of intracapsular cells in spinal ganglia were updated with the demonstration of the oligodendroglial nature of these elements and their arrangement either perineuronal, fully enveloping the neurons, and the intracapsular, forming spiral wrappings around the axons (Rı´o Hortega, Polak and Prado, 1942). Similar Wndings applied to the autonomic ganglia (Rı´o Hortega and Prado, 1942), where, in addition, these investigators described spiral wrappings around neuronal dendrites. These Wndings conWrmed the earlier predictions of Castro (1923, 1937), who after detailed studies of sympathetic ganglia considered the perineuronal and peridendritic elements as neuroglial and perhaps oligodendroglial in nature. Finally, Rı´o Hortega (1942–1943) provided a thorough review of the silver carbonate method and all of its variants as applied in normal and pathologic histology, and this was followed by an extensive description and discussion of pathologic alterations of neuroglial cells (Rı´o Hortega, 1943–1945). It was his last publication. He died in 1945 at the age of 63. In summary, the silver carbonate technique and the continuing use of variants of the Golgi and Bielschowsky methods dominated the post-Cajalian period. The preeminent Wgure was Pı´o del Rı´o Hortega, who single-handed discovered the microglia of mesoderm origin, and the oligodendroglia, a new form of ectoderm neuroglia central to the formation of myelin in the central nervous system. Many of his reasonings, however, reXected Cajal’s teachings and some of Achu´carro’s thoughts, as can be found in the review presented earlier of the Cajalian period of glia research, but undoubtedly, it took the ingenuity of Rio Hortega to bring the various issues into focus. This development emulates Cajal’s indisputable priority in formulating the neuron doctrine, in spite of some antecedents being available at the time. What remains of the Cajal-Rı´o Hortega controversy about the ‘‘third element’’ is the sense, so many times repeated in the history of science, of an

xxxvii

xxxviii

PREFACE

overreacting teacher and a pupil anxious for recognition. What is rather unusual is the human level of these personal interactions, which, with very rare exceptions, were always courteous and imbued with love and respect for each other. Let this be an example for present and future generations.

References Achu´carro, N. (1908). Sur la formation de cellules a` battonnet (Sta¨bchenzellen) et d’autres elements similaires dans le syste`me nerveux central. Trab Lab Invest Biol 6, 97–122. Achu´carro, N. (1909). Cellules allonge´es et Sta¨bchenzellen, cellules ne´vrogliques et cellules granulo-adipeuses a` la corne d’Ammon du lapin. Trab Lab Invest Biol 7, 201–216. Achu´carro, N. (1910a). Zur Kenntnis der pathologischen Histologie des Zentralnervensystems bei Tollwut. Histol histopathol Arb 3, 143–199. Achu´carro, N. (1910b). Some pathological Wndings in the neuroglia and the ganglion cells of the cortex in senile conditions. Bull Gov Hosp Insane 2, 81–90. Achu´carro, N. (1910c). Elongated cells, Sta¨bchenzellen, neuroglia cells, and fat-granular cells in the Ammon’s horn of the rabbit. Bull Gov Hosp Insane 2, 91–103. Achu´carro, N. (1910d). Algunos datos relativos a la naturaleza de las ce´lulas en bastoncito de la corteza cerebral humana, obtenidos con el me´todo de Cajal. Trab Lab Invest Biol 8, 169–176. Achu´carro, N. (1911a). Neuroglia y elementos intersticiales patolo´gicos del cerebro, impregnados por los me´todos de reduccio´n de la plata o por sus modiWcaciones. Trab Lab Invest Biol 9, 161–180. Achu´carro, N. (1911b). Nuevo me´todo para el estudio de la neuroglia y del tejido conjuntivo. Bol Soc Espan˜ Biol 1, 139–141. Achu´carro, N. (1911c). Darstellung von neugebildeten Fasern des Gefa¨ssbindegewebes in der Hirnrinde eines Falles von progressiver Paralyse, durch eine neue Tannin-Silbermethode. Z ges Neurol Psychiat 7, 375–383. Achu´carro, N. (1913a). La estructura de la neuroglia en la corteza cerebral. Bol Soc Espan˜ Biol, 3, 27–29. Achu´carro, N. (1913b). Notas sobre la estructura de la neuroglia y en particular de la neuroglia de la corteza cerebral humana. Trab Lab Invest Biol 11, 187–217. Achu´carro, N. (1915a). Contribucio´n al estudio gliotecto´nico de la corteza cerebral. El asta de Ammon y la fascia dentata. Trab Lab Invest Biol 12, 229–272. Achu´carro, N. (1915b). De l’e´volution de la ne´vroglie, et spe´cialement de ses relations avec l’appareil vasculaire. Trab Lab Invest Biol 13, 169–212. Achu´carro, N., and Gayarre, M. (1914a). La corteza cerebral en la demencia paralı´tica con el nuevo me´todo del oro y sublimado de Cajal. Trab Lab Invest Biol 12, 1–38. Achu´carro, N., and Gayarre, M. (1914b). Contribucio´n al estudio de la neuroglia en la corteza de la demencia senil y su participacio´n en la alteracio´n de Alzheimer. Trab Lab Invest Biol 12, 67–83. Alzheimer, A. (1904). Histologischen Studien zur diVerential Diagnose der progressiver Paralyse. Histol histopathol Arb 1, 18–314. Andriezen, W. L (1893a). The neuroglia elements in the human brain. Brit Med J 3, 227–230. Andriezen, W. L (1893b). On a system of Wbre-celles surrounding the blood-vessels of the brain of man and mammals and its physiological signiWcance. Internat Monatsschr Anat Physiol 10, 532–540. Bidder, F. H., and KupVer, C. (1857). ‘‘Untersuchungen u¨ber die Textur des Ru¨ckenmarks und die Entwickenlung seiner Formelemente.’’ Breitkopf and Ha¨rtel, Leipzig, pp. 121. Boll, F. (1874). Die Histologie und Histogenese der nervo¨sen Centralorgane. Arch Psychiat Nervenkr 4, 1–136. Cajal, S. R. y (1888a). Nota sobre la estructura de los tubos nerviosos del lo´bulo cerebral ele´ctrico del torpedo. Rev Trimest Histol Norm Patol 1, 49–55. Cajal, S. R. y (1888b). Estructura de los centros nerviosos de las aves. Rev Trimest Histol Norm Patol 1, 1–10. Cajal, S. R. y (1889a). Contribucio´n al estudio de la estructura de la me´dula espinal. Rev Trimest Histol Norm Patol 1, 79–106. Cajal, S. R. y (1889b). Sobre las Wbras nerviosas de la capa granulosa del cerebelo. Rev Trimest Histol Norm Patol 1, 107–118. Cajal, S. R. y (1890a). Sur l’origine et les ramiWcations des Wbres nerveuses de la moelle embryonaire. Anat Anz 5, 85–95, 111–119. Cajal, S. R. y (1890b). A quelle e´poque apparaissent les expansions des cellules nerveuses de la moelle e´pinie`re du poulet? Anat Anz 5, 609–613; 631–639. Cajal, S. R. y (1896). El azul de metileno en los centros nerviosos. Rev Trimest Microg 1, 151–203. Cajal, S. R. y (1897). Algo sobre la signiWcacio´n Wsiolo´gica de la neuroglia. Rev Trimest Microg 2, 33–47. Cajal, S. R. y (1912). Fo´rmula de Wjacio´n para la demostracio´n fa´cil del aparato de Golgi y apuntes sobre la disposicio´n de dicho aparato en la retina, en los nervios y algunos estados patolo´gicos. Trab Lab Invest Biol 10, 209–220. Cajal, S. R. y (1913a). Sobre un nuevo proceder de impregnacio´n de la neuroglia y sus resultados en los centros nerviosos del hombre y animales. Trab Lab Invest Biol 11, 219–237. Cajal, S. R. y (1913b). Contribucio´n al conocimiento de la neuroglia del cerebro humano. Trab Lab Invest Biol 11, 255–315.

ed5 au6

POST-CAJALIAN PERIOD

Cajal, S. R. y (1916). El proceder del oro sublimado para la coloracio´n de la neuroglia. Trab Lab Invest Biol 14, 155–162. Cajal, S. R. y (1920a). Algunas consideraciones sobre la mesoglı´a de Roberston y Rı´o Hortega. Trab Lab Invest Biol 18, 109–127. Cajal, S. R. y (1920b). Una modiWcacio´n del me´todo de Bielschowsky para la impregnacio´n de la neuroglia comu´n y mesoglia y algunos consejos acerca de la te´cnica del oro-sublimado. Trab Lab Invest Biol 18, 129–141. Cajal, S. R. y (1923a). Quelques me´thodes simples pour la coloration de la ne´vroglie. Arch Suiss Neurol Psychiat 13, 187–193. Cajal, S. R. y (1923b). ‘‘Recuerdos de mi Vida: Historia de mi Labor Cientı´Wca.’’ (Reprinting of 3rd ed. Alianza Editorial, Madrid, 1981), pp. 370. Cajal, S. R. y (1925). Contribution a` la connaissance de la ne´uroglie ce´re´brale et ce´re´be´lleuse dans la paralysie ge´ne´rale progressive, avec quelques indications techniques sur l’impregnation argentique du tissu nerveux pathologique. Trav Lab Rech Biol Univ Madrid 23, 151–216. Cajal, S. R. y (1926). Algunas precisiones sobre el proceder de formolbromuro y plata amoniacal, para la coloracio´n de la glia y microglia patolo´gica, singularmente de la para´lisis general. Bol Soc Espan˜ Biol 16, 111–115. ´ tudes sur la ne´vroglie (macroglie). I Partie. Indications techniques. Trav Lab Rech Biol Cajal, S. R. y (1932a). E Univ Madrid 27 (append), 377–388. ´ tudes sur la ne´vroglie (macroglie). II Partie. Contribution a` la connaissance de la ne´vroglie Cajal, S. R. y (1932b). E (macroglie) du cerveau humain. Trav Lab Rech Biol Univ Madrid 27 (append), 388–474. Cajal, S. R. y (1933). ‘‘Histology’’ (revised by J. F. Tello-Mun˜oz; translation of 10th ed. by M Ferna´n-Nu´n˜ez). Wood, Baltimore, pp. 738. Cajal, S. R. y (1999–2002). ‘‘Texture of the nervous system of man and the vertebrates’’ (annotated and edited translation by P. Pasik and T. Pasik). Springer, Wien, New York, Barcelona, Vol. 1, pp. 631;Vol. 2, pp. 667; Vol. 3, pp. 663. Cajal, S. R. y, and Olo´riz, F. (1897). Los ganglios sensitivos craneales de los mamı´feros. Rev Trimest Microg 2, 129–152. Castro, F. de (1923). Sobre la Wna anatomı´a de los ganglios simpa´ticos de algunos mamı´feros de gran talla. Bol Soc Espan˜ Biol 10, 92–97. Castro, F. de (1937). Sur la re´ge´ne´ration fonctionnelle dans le sympathique. Une refe´rence espe´ciale sur la constitution des synapses. Trab Inst Cajal Invest Biol 31, 271–345. Castro, F. de, and Lorente de No´, R. (1923). A propo´sito de la homologı´a entre la glia de escasas radiationes y la ce´lula de Schwann y endocapsulares. Bol Soc Espan˜ Biol 10, 63–68. Cerletti, U. (1905). Sopra alcuni rapporti tra le ‘‘cellule a bastoncello’’ (Sta¨bchenzellen), e gli elementi nervosi nella paralisi progressiva. Riv Sper Freniat 31, 483–495. Cerletti, U. (1907–1908). Estudi recenti sull’istogenesi dell nevroglia. Ann Ist Psychiatr R Univ Roma 4, 221–234. Deiters, O. F. K. (1865). ‘‘Untersuchungen u¨ber Gehirn und Ru¨ckenmark des Menschen und der Sa¨ugethiere’’ (M. Schultze, ed.). Vieweg, Braunschweig, pp. 318. Eichhorst, H. (1875). Ueber die Entwicklung des menschlichen Ru¨ckenmarkes und seiner Formelemente. Arch pathol Anat Physiol klin Med 64, 425–475. Fan˜anas, J. R. y (1916). Contribucio´n al estudio de la neuroglia del cerebelo. Trab Lab Invest Biol 14, 163–179. Fieandt, H. (1910–1911). Eine neue Methode zur Darstellung des Gliagewebes, nebst Beitrage zur Kenntnis des Baues und der Anordnung der Neuroglia des Hundehirns. Arch mikr Anat 76, 125–209. Golgi, C. (1872). Contribuzione alla Wna anatomia degli organi centrali del sistema nervoso. Riv Clin 2, 38–46. Golgi, C. (1873). Sulla struttura della sostanza grigia del cervello. Gaz Med Ital Lomb 6, 244–246. Golgi, C. (1886). ‘‘Sulla Wne anatomia degli organi centrali del sistema nervoso.’’ Hoepli, Milano, pp. 214. Golgi, C. (1894). ‘‘Untersuchungen u¨ber den feineren Bau des centralen und peripherischen Nervensystems.’’ Fischer, Jena, pp. 272. Hardesty, I. (1904). On the development and nature of the neuroglia. Am J Anat 3, 229–268. ¨ ber den Bau der Neuroglia und u¨ber die Wand der Lymphgefa¨sse in Haut und Schleimhaut. Held, H. (1903). U Abhandl math-physis Klasse ko¨nigl sa¨chs Gesellsch Wissensch 28, 199–318. Held, H. (1909). Ueber die Neuroglia marginalis der menschlichen Grosshirnrinde. Monatssch Psychiat Neurol 26, 360–416. His, W. (1890). Histogenese und Zusammenhang der Nervenelemente. Arch Anat Entwicklungsgesch 95–117. Kershman, J. (1939). Genesis of microglia in the human brain. Arch Neurol Psychiat 41, 24–50. Ko¨lliker, A. v (1863). ‘‘Handbuch der Gewebelehre des Menschen,’’ 4th ed. Engelmann, Leipzig, pp. 730. Ko¨lliker, A. v (1867). ‘‘Handbuch der Gewebelehre des Menschen,’’ 5th ed. Engelmann, Leipzig, pp. 749. Ko¨lliker, A. v (1891). Zur feineren Anatomie des centralen Nervensystems. Zweiter Beitrag. Das Ru¨ckenmark. Z wissensch Zool 51, 1–54. Ko¨lliker, A. v (1896). ‘‘Handbuch der Gewebelehre des Menschen.’’ 6th ed, Vol II. Nervensystem des Menschen und der Thiere. Engelmann, Leipzig, pp. 874. Lenhosse´k, M. v (1891). Zur Kenntniss der Neuroglia des menschlichen Ru¨ckenmarkes. Verhandl d anat Gesellsch 5, 193–221. Lenhosse´k, M. v (1895). ‘‘Der feinere Bau des Nervensystems in Lichte neuester Forschungen.’’ 2nd ed. Fischer, Berlin, pp. 409. Lenhosse´k, M. v (1907). Zur Kenntnis der Spinalganglienzellen. Arch Mikr Anat 69, 245–263. Lugaro, E. (1907). Sulle funzioni della nevroglia. Riv Patol Nerv Ment 12, 225–233.

xxxix

xl

PREFACE

Metz, A., and Spatz, H. (1924). Die Hortega’schen Zellen (= das sogenannte ‘‘dritte Element’’) und u¨ber ihre funktionelle Bedeutung. Z ges Neurol Psychiat 89, 138–170. Mu¨ller, E. (1900). Studien uber Neuroglia. Arch Mikr Anat 60, 11–62. Nageotte, J. (1910). Phe´nome`nes de se´cre´tion dans le protoplasma des cellules ne´vrogliques de la substance grise. C R Se´an Soc Biol Paris 68, 1068–1069. Nansen, F. (1886). The structure and combination of the histological elements of the central nervous system. Bergens Museums Aarsberetning 27–215. Nissl, F. (1885). Ueber die Untersuchungsmethodender Grosshirnrinde. Tagebl 58 Versamm Dtsch Naturf Aerzte Strassburg 506–507. Nissl, F. (1899). Ueber einige Beziehungen zur Nervenzellenerkrankungen und gliosen Erscheinungen bei verschiedenen Psychosen. Arch Psychiat 32, 656–676. Ochoa, S. (1986). Recuerdo de Don Pı´o. In Rı´o Hortega P d. ‘‘El Maestro y Yo.’’ Cons Sup Invest Cient, Madrid, (posthumous memoirs) 27–33. PenWeld, W. (1924). Oligodendroglia and its relation to classical neuroglia. Brain 47, 430–452. PenWeld, W. (1932). Neuroglia, normal and pathological. In ‘‘Cytology and Cellular Pathology of the Nervous System’’ (W. PenWeld, ed.), Vol 2, 421–479. Hoeber, New York. PenWeld, W., and Cone, W. (1926). A speciWc type of neuroglia change, acute swelling of oligodendroglia. Arch Neurol Psychiat 16, 131–153. Ranvier, L.-A. (1873). Sur les e´le´ments conjonctifs de la moelle e´pinie`re. Compt Rend Hebd Se´an Acad Sc 77, 1299–1302. Ranvier, L.-A. (1878). ‘‘Lec¸ons sur l’histologie du syste´me nerveux.’’ Vol 2. Savy, Paris, pp. 380. Ranvier, L.-A. (1882). De la ne´vroglie. Compt Rend Hebd Se´an Acad Sc 94, 1536–1539. Ranvier, L.-A. (1893). De la neuroglie. Arch Physiol Norm Pathol 1 (ser 3), 177–185. Retzius, G. (1891). Ueber den Bau der OberXa¨cheschicht der Grosshirnrinde beim Menschen und bei den Sau¨gethieren. Biologiska Fo¨reningens Fo¨rhandlingar 3, 90–102. Retzius, G. (1893). Studien u¨ber Ependym und Neuroglia. Biol Untersuch NF 5, 9–26. Retzius, G. (1894). Die Neuroglia des Gehirns beim Menschen und bei Sa¨ugethieren. Biol Untersuch NF 6, 1–28. Rı´o Hortega, P. del (1916a). Contribution a` l’e´tude d’histopathologie de la ne´vroglie. Ses variation dans la ramollissement cerebral. Trab Lab Invest Biol 14, 1–34. Rı´o Hortega, P. del (1916b). Nuevas reglas para la coloracio´n constante de las formaciones conectivas, por el me´todo de Achu´carro. Trab Lab Invest Biol 14, 181–188. Rı´o Hortega, P. del (1917). Noticia de un nuevo y fa´cil me´todo para la coloracio´n de la neuroglia y del tejido conjuntivo. Trab Lab Invest Biol 15, 367–378. Rı´o Hortega, P. del (1919a). El ‘‘tercer elemento’’ de los centros nerviosos. I.– La microglı´a en estado normal. Bol Soc Espan˜ Biol 9, 68–82. Rı´o Hortega, P. del (1919b). El ‘‘tercer elemento’’ de los centros nerviosos. II.– Intervencio´n de la microglı´a en los procesos patolo´gicos. Bol Soc Espan˜ Biol 9, 91–103. Rı´o Hortega, P. del (1919c). El ‘‘tercer elemento’’ de los centros nerviosos. III.– Naturaleza probable de la microglı´a. Bol Soc Espan˜ Biol 9, 108–120. Rı´o Hortega, P. del (1919d). El tercer elemento de los centros nerviosos. Poder fagocitario y movilidad de la microglia. Bol Soc Espan˜ Biol 9, 154–166. Rı´o Hortega, P. del (1920). Estudios sobre la neuroglia. La microglı´a y su transformacio´n en ce´lulas en bastoncito y cuerpos gra´nulo-adiposos. Trab Lab Invest Biol 18, 37–82. Rı´o Hortega, P. del (1921a). Histoge´nesis y evolucio´n normal, e´xodo y distribucio´n regional de la microglı´a. Mem R Soc Espan˜ Hist Nat, 11: 213–268. Rı´o Hortega, P. del (1921b). La glı´a de escasas radiaciones (oligodendroglı´a). Bol R Soc Espan˜ Hist Nat 21, 63–92. Rı´o Hortega, P. del (1922). Son homologables la glı´a de escasa radiaciones y la ce´lula de Schwann? Bol Soc Espan˜ Biol 10, 25–28. Rı´o Hortega, P. del (1924). Lo que debe entenderse por tercer elemento de los centros nerviosos. Bol Soc Espan˜ Biol 11, 33–35. Rı´o Hortega, P. del (1925). Condrioma y granulaciones especı´Wcas de las ce´lulas neuro´glicas. Bol R Soc Espan˜ Hist Nat 25, 34–55. Rı´o Hortega, P. del (1928). Tercera aportacio´n al conocimiento morfolo´gico e interpretacio´n funcional de la oligodendroglı´a. Mem R Soc Espan˜ Hist Nat 14, 5–122. Rı´o Hortega, P. del (1932). Microglia. In ‘‘Cytology and Cellular Pathology of the Nervous System’’ (W. PenWeld, ed.), Vol 2, 483–534. Hoeber, New York. Rı´o Hortega, P. del (1939). The microglia. Lancet 236, 1023–1026. Rı´o Hortega, P. del (1942). La neuroglia normal. Conceptos de angiogliona y neurogliona. Arch Histol Norm Patol 1, 5–71. Rı´o Hortega, P. del (1942–43). El me´todo del carbonato arge´ntico. Revisio´n general de sus te´cnicas y aplicaciones en histologı´a normal y patolo´gica. Arch Histol Norm Patol 1–2, 165–231. Rı´o Hortega, P. del (1943–45). Ensayo de clasiWcacio´n de las alteraciones celulares del tejido nervioso. II. Alteraciones de las ce´lulas neuro´glicas. Arch Histol Norm Patol 2, 5–100. Rı´o Hortega, P. del (1949). Art and artiWce in histological science. Texas Rep Biol Med 7, 363–390 (Transl by E W Wolfe and G M Butler from the original Arte y artiWcio de la ciencia histolo´gica, Residencia 1933, 4, 191–206). Rı´o Hortega, P. del (1986). ‘‘El Maestro y Yo.’’ Cons Sup Invest Cient, Madrid, pp. 151 (posthumous memoirs).

POST-CAJALIAN PERIOD

Rı´o Hortega, P. del, and PenWeld, W. (1927). Cerebral cicatrix. Reaction of neuroglia and microglia to brain wounds. Johns Hopkins Hosp Bull 41, 278–303. Rı´o Hortega, P. del, Polak, M., and Prado, J. M. (1942). Investigaciones sobre la neuroglia de los ganglios sensitivos. Arch Histol Norm Patol 1, 234–275. Rı´o Hortega, P. del, and Prado, J. M. (1942). Investigaciones sobre la neuroglia de los ganglios simpa´ticos. Arch Histol Norm Patol 1, 83–138. Robertson, W. F. (1897). The normal histology and pathology of neuroglia (in relation specially to mental disease). J Ment Sci 43, 733–752. Robertson, W. F. (1899). On a new method of obtaining a black reaction in certain tissue-elements of the central nervous system (platinum method). Scottish Med Surg J 4, 23–30. Robertson, W. F. (1900a). A microscopic demostration of the normal and pathological histology of mesoglia cells. J Ment Sci 46, 724. Robertson, W. F, (1900b). ‘‘A Text-Book of Pathology in Relation to Mental Diseases.’’ Clay, Edinburgh, pp. 380. Sala y Pons, C. (1894). ‘‘La Neuroglia de los Vertebrados. Estudios de Histologı´a Comparada.’’ Madrid, pp. 44. SchaVer, K. (1915). Zur Kenntnis der normalen und pathologischen Neuroglia. Z ges Neurol Psychiat 30, 1–41. Somjen, G. G. (1988). Nervenkitt: Notes on the history of the concept of neuroglia. Glia 1, 2–9. Tourneaux, F., and Le GoV, R. (1875). Note sur les e´tranglements des tubes nerveux de la moelle e´piniere. J Anat Physiol Norm Pathol Homme Anim 11, 403–404. Vignal, W. (1884). Sur le de´veloppement des e´le´ments de la moelle des mammiferes. Arch Physiol Norm Pathol 3–4 (ser 3), 177–238, 364–421. ´ le´ments du Syste`me Nerveux Ce´re´bro-spinal; Nerfs Pe´riphe´riques; Vignal, W. (1889). ‘‘De´veloppement des E Moelle; Couches Corticales du Cerveau et du Cervelet.’’ Masson, Paris, pp. 214. Virchow, R. (1846). Ueber das granulirte Ansehen der Wandungen der Gehirn-ventrikel. Allg Z Psychiat psychgericht Med 3, 242–250. Virchow, R. (1854). Ueber eine im Gehirn und Ru¨ckenmark des Menschen aufgefundene Substanz mit der chemischen Reaction der Cellulose. Arch pathol Anat Physiol klin Med 6, 135–138. Virchow, R. (1856). ‘‘Gesammelte Abhandlungen zur wissenschaftlichen Medizin.’’ Meidinger, Frankfurt, pp. 1024. Vitoria Ortiz, M. (1977). ‘‘Vida y Obra de Nicola´s Achu´carro.’’ La gran enciclopedia vasca, Bilbao, pp. 511. Weigert, C. (1884). Ausfu¨hrliche Beschreibung der in No 2 dieser Zeitschrift erwa¨hnten neuen Fa¨rbungsmethode fu¨r das Centralnervensystem. Fortschr Med 2, 190–191. Weigert, C. (1890). Bemerkungen u¨ber das Neurogliageru¨st des menschlichen Centralnervensystems. Anat Anz 5, 543–551. Weigert, C. (1895). ‘‘Beitra¨ge zur Kenntnis der normalen menschlichen Neuroglia.’’ Weisbrod, Frankfurt, pp. 149.

xli

C H A P T E R

1 Structure of the Myelinated Axon Bruce D. Trapp and Grahame J. Kidd

INTRODUCTION In his 1928 description in Degeneration and Regeneration of the Nervous System (Ramon y Cajal, 1928), Ramon y Cajal stated, ‘‘Myelin is an organ which is an adjunct of the axon, and as such, it is entirely foreign to Schwann cells, from which it is perfectly distinct in well Wxed and stained preparations.’’ This was one of the few misinterpretations in Ramon y Cajal’s classic light microscopic observations. Ramon y Cajal’s appreciation for the intricate relationships between the various components of the nervous system remains at the forefront of modern science. His major limitation in determining the source of myelin was merely technical. It was then, and is still today, impossible to resolve the cellular boundaries of the axon and myelin at the light microscopic level. It was not until the advent of the transmission electron microscope in the late 1950s that fundamental questions regarding the microstructure of the myelinated axons could be addressed. Initially, transmission electron microscopy (TEM) was a science unto itself. As the methods and electron microscopes improved, myelin ultrastructure was eloquently described. Myelin was unequivocally demonstrated to be a spiraled extension of the Schwann cell plasma membrane in the PNS (Geren, 1954) and of the oligodendrocyte plasma membrane in the CNS (Bunge et al., 1962). Furthermore, oligodendrocytes were shown to form multiple myelin internodes, while Schwann cells formed but a single internode. Eventually, other concepts emerged. Myelin internodes were shown to consist of ultrastructurally distinct domains, and these domains were not identical in the PNS and CNS. These concepts are fundamental to current EM research that focuses on the molecular architecture of myelin internodes and on pathological changes that result from inherited and acquired diseases of myelin. While TEM is no longer a science unto itself, it is currently the only road to many of the most momentous questions of modern molecular science. This chapter describes the molecular ultrastructure of myelin with a focus on normal function and pathological changes that result from myelin disease. For historical perspectives, the reader is directed to Peters et al., 1991 and Rosenbluth, 1999. A discussion of the role of myelin in saltatory conduction is presented in Chapter 5.

MYELIN INTERNODE ORGANIZATION To appreciate the ultrastructure of the myelin internode, one must understand its threedimensional organization. In simple morphological terms, the myelin internode can be divided into two domains, compact myelin and noncompact myelin. Figure 1.1A is a diagram of a teased myelinated PNS axon observed and drawn by Ramon y Cajal (Ramon y Cajal, 1928). This myelin internode is 0.5 mm long and surrounds an axon

Myelin Biology and Disorders, Volume 1

3

Copyright 2004, Elsevier (USA). All rights reserved.

4

1. STRUCTURE OF THE MYELINATED AXON

FIGURE 1.1 Myelin internodes in the peripheral nervous system (PNS). Drawings by Ramon y Cajal of osmic acid-Wxed (A) and silver impregnated (B, negative image) myelinated fibers from the PNS. Compact myelin membranes surround the axon (A, black) and comprise the majority of the myelin internode. To illustrate the membrane expansion that occurs during myelination, the size of a Schwann cell before myelination (C) is compared to a Schwann cell and it ‘‘unrolled’’ myelin internode (D). Cytoplasmic domains at the outer surface of the internode (B, D) are contiguous with the cytoplasmic channels that surround (paranodes and inner mesaxon) and traverse (Schmidt-Lanterman incisures) the compact myelin (D). Panels A and B have been reproduced from Ramon y Cajal, 1928, with permission.

with a diameter of 2.8 mm’s. The black, osmic acid–stained compact myelin surrounds the axon. The Schwann cell nucleus is positioned at the middle of the internode. Translucent cytoplasmic channels, called Schmidt-Lanterman incisures, radially traverse the compact myelin. Inconspicuous in osmicated nerves, cytoplasmic channels at the outer surface of the internode are highlighted as white areas in Figure 1.1B. Figure 1.1C shows the relative size of a Schwann cell before myelination. The sheath in Figure 1.1A is unrolled in Figure 1.1D and schematically illustrates the compact myelin (black) and cytoplasmic domains (white) of the myelin internode. Compact myelin comprises the majority of the myelin internode. Nonompact myelin provides cytoplasmic continuity between myelin forming cells and various regions of the myelin internode. Since the majority of myelin components are synthesized in the perinuclear cytoplasm of myelin-forming cells, cytoplasmic channels at the outer or abaxonal surface are needed to expand and maintain the myelin internode. These channels contain microtubules and other cytoskeletal components for transport and stability and mitochondria for energy. Also, in some areas, they contain smooth endoplasmic reticulum and free polysomes for local membrane component biosynthesis. In addition, membranes of noncompact myelin serve special functions that are reXected by unique molecular compositions. The cytoplasmic channels or paranodal loops at the

MYELIN INTERNODE FUNCTION

lateral ends of the internode are a major site of myelin-axon adhesion. The membranes of the inner or adaxonal surface are in direct contact with the axons. Their cytoplasmic channels may transmit axonal signals that regulate myelin formation and help determine the length and thickness of the myelin internode. Schmidt-Lanterman incisures traverse compact PNS myelin and connect outer and inner regions of the internode. Oligodendrocytes diVer from Schwann cells in that they have the ability to form multiple myelin internodes and do so by extending a single process to each internode. The number of internodes formed by each oligodendrocyte is regulated by axons (Blakemore, 1981; Fanarraga et al., 1998). Oligodendrocytes in CNS Wber tracts with small diameter axons myelinate more axons than those with large diameter axons. This suggests that oligodendrocytes have the ability to ‘‘read’’ the future diameter of axons as all axons are of similar diameter (<1 mm) when myelination is initiated. CNS internodes surrounding small diameter axons are shorter and thinner than those surrounding large diameter axons. Each oligodendrocyte, therefore, has the potential to synthesize similar amounts of myelin whether it is multiple small internodes or a few large internodes. Figure 1.2A is a cartoon of the relationship between an oligodendrocyte and its myelin internodes. One sheath is unrolled to highlight the compact myelin (black) and cytoplasmic domains (white). As in PNS myelin internodes, noncompact myelin surrounds the exterior of the CNS internode. Schmidt-Lanterman incisures, however, are not a prominent feature of CNS internodes (Blakemore, 1969).

MYELIN INTERNODE FUNCTION Myelin promotes rapid and repetitive communication between neurons and it modulates the maturation and survival of axons. We have a good understanding of how myelin

FIGURE 1.2 Myelin internodes in the central nervous system (CNS). Oligodendrocytes (OL) form multiple myelin internodes in the CNS. One internode is unrolled and schematically depicts compact myelin and the cytoplasmic domains of the paranodal loops and outer and inner tongue processes. A cutaway view of a nodal region illustrates the terminal ends of adjacent myelin internodes and the ‘‘naked’’ axonal membrane or node where Naþ channels are concentrated. Electrophysiologically, the myelin internode insulates the axonal surface between nodes of Ranvier (B) and facilitates rapid propagation of the action potential, which jumps from node to node by saltatory conduction.

5

6

1. STRUCTURE OF THE MYELINATED AXON

facilitates neuronal communication, but have little knowledge of how it modulates axonal integrity. Neurons communicate by depolarizing the ‘‘electrical potential’’ of their axons. This occurs by channel-mediated exchange of Naþ in an energy dependent manner (see Chapter 5). Unmyelinated axons generate action potentials along their entire surface at a speed that is proportional to their radius. Rapid unmyelinated axonal communication would demand energy and space requirements that are not conducive to evolution of the large, complex mammalian brain. It was advantageous, therefore, to evolve a mechanism to propagate neuronal communication through thin axons in an energy eYcient manner. Myelin-forming cells serve this function by producing a series of discontinuous insulation units called internodes along individual axons. Each myelin internode can be divided into two ultrastructurally and functionally distinct domains: compact myelin and the paranodal loops. Compact myelin inhibits ion exchange during nerve conduction by having a low capacitance and high resistance. The paranodes, which demarcate the longitudinal ends of each internode, facilitate ion exchange at the node of Ranvier, a small ‘‘bare patch’’ of axon that separates each successive internode. The nerve impulse jumps from node to node by saltatory conduction (Fig. 1.2B). Mammalian unmyelinated nerves have conduction velocities of less than 1 m/sec. Myelinated axons of the same diameter conduct 10 times faster and the largest Wbers conduct up to 100 m/sec. For unmyelinated axons to achieve similar velocities to myelinated axons of the same size, they would need to be about 100 times thicker.

COMPACT MYELIN Most of the myelin internode consists of compact myelin that appears by TEM as a lamellar structure of alternating dark and light lines that spiral around the axon (Fig. 1.3). Although compact myelin has a very simple ultrastructure, it is often diYcult to comprehend. Compact myelin ultrastructure is best understood if it is considered a spiraled cellular process or sheet that contains two plasma membranes and no cytoplasm. Membranes visualized by conventional TEM have an internal translucent core and two electron dense lines or leaXets at the extracellular and intracellular surface (Alberts et al., 1994). Figure 1.4A depicts a cross section of the theoretical myelin membrane sheet with cytoplasm (green). The lipid bilayer of the surface membrane is disproportionately increased in size: the red line represents the extracellular leaXet, while the dark green line represents the cytoplasmic leaXet. This process is spiraled in Figure 1.4B and spiraled and compacted in Figure 1.4C. In the compacted spiral, apposing extracellular leaXets (called intraperiod or intermediate lines) are separated by 2.0 nm in the CNS and 2.5 nm in the PNS (Fig. 1.4D). The cytoplasmic leaXets appear fused and form the major dense lines or intraperiod lines. The spiral membranes of compact myelin have a periodicity (distance from dense line to dense line) of approximately 13 to 14 nm when Wxed with aldehydes and embedded in epoxy resins (Fernandez-Moran and Finean, 1957; Kirschner and Hollingshead, 1980) and is slightly greater in PNS myelin than in CNS myelin. In situ, however, these distances are slightly larger (17 to 18 nm) when measure by X-ray diVraction (Kirschner and Sidman, 1976). Since the major function of compact myelin is insulation, it need not be biochemically complex or molecularly dynamic. Ultrastructure would predict membrane adhesion as the major function of compact myelin molecules. Furthermore, adhesion molecules cannot occupy large extracellular or cytoplasmic areas and it would be most eYcient if they have a slow turnover rate. All of these predictions have been conWrmed and are discussed in detail in other chapters of this book. A high lipid composition (~75%), imparts a high buoyancy to homogenized myelin membranes upon sucrose gradient subfractionation and separates it from other membrane fractions with the more common 1 to 1 lipid:protein ratios (Norton and Poduslo, 1973). Because of its abundance, invariable structure and unique biochemical composition, myelin was the Wrst CNS membrane to be isolated and molecularly characterized. Abundant proteins include P0 protein, the myelin basic proteins (MBP), P2 protein and PMP-22 in PNS myelin (GreenWeld et al., 1973), and proteolipid protein (PLP) and MBP in CNS myelin (Lees and BrostoV, 1984). Immunocytochemistry

COMPACT MYELIN

FIGURE 1.3 Ultrastructure of myelinated axons in the peripheral (A) and the central (B) nervous systems. In both PNS and CNS, axons (Ax) are surrounded by electron-dense compact myelin (m). The myelin spirals are bounded by the inner (periaxonal) and outer (abaxonal) mesaxons (arrows). In PNS myelinated Wbers (A), organelle-rich cytoplasmic channels (*) occupy approximately 50% of the outer circumference, and are separated by nonchannel regions. Basal lamina (bl) surrounds the entire abaxonal surface, and extracellular collagen Wbrils (c) separate adjacent Wbers. Cytoplasmic channels of the CNS internode include inner and outer tongue processes (B, *). Unlike the PNS, no visible extracellular matrix separates myelinated Wbers, and where compact sheaths lie adjacent to each other, a common intraperiod line forms. Panel B has been reproduced from Tabira et al., 1978, with permission. Scale bars ¼ 0.5 mm.

conWrmed an enrichment of P0, MBP, PMP-22, and P2 protein in compact PNS myelin and PLP and MBP in compact CNS myelin (Hartman et al., 1982; Snipes et al., 1992; Sternberger et al., 1978; Trapp et al., 1979). Amino acid composition and biochemical properties identiWed P0, PMP-22, and PLP as integral membrane proteins and MBP and P2 protein as extrinsic membrane proteins. P0, PMP-22, and PLP are positioned to maintain the periodicity of both the extracellular and cytoplasmic spacing of myelin membranes, while MBP and P2 could inXuence cytoplasmic leaXet fusion. A major focus of myelin research in the 1990s was the development of technology to eliminate these myelin proteins in mice. One unexpected Wnding of these studies is that signiWcant amounts of multilammellar membranes form in the absence of P0 (Giese et al., 1992), PLP (Duncan et al., 1988, 1989), MBP (Privat et al., 1979), or MAG (Li et al., 1994; Montag et al., 1994), PLP/MBP (StoVel et al., 1997), or PLP/MBP/MAG (Uschkureit et al., 2000). Spiral expansion of myelin membranes, therefore, does not depend on these molecules. Electron microscopy, however, has convincingly established that PLP and MBP are required for normal spacing of compact CNS myelin (Duncan et al., 1988; Privat et al.,

7

8

1. STRUCTURE OF THE MYELINATED AXON

FIGURE 1.4 Schematics of compact myelin periodicity. Compact myelin can be viewed as a cell process (A). The process grows spirally (B) around an axon, and during myelin compaction (C), the cytoplasm (green) is excluded. Cytoplasmic membrane leaXets (dark green) fuse to form the major dense lines (D, MDL), while extracellular leaXets (red) of adjacent lamellae become closely apposed to form the intraperiod lines (D, IPL). The orientation of three proteins that play a dominant role in compact myelin integrity are depicted. For simplicity, both CNS and PNS proteins are shown here; see the text for their individual distributions.

1979), while P0 is essential for normal spacing of PNS compact myelin (Giese et al., 1992). When PLP and P0 are removed, compact myelin membrane spacing is often variable. While the mistargeting of other proteins may play a role in this variability, it is most likely due to poor Wxation. Interpreting compact myelin ultrastructure is not always straightforward. Larger Wbers are prone to poor Wxation even in the most experienced hands and can have split and disrupted lamellae. Mechanical damage during preparation can also produce artifacts. Great care must be taken in ascribing abnormal myelin ultrastructure to pathology. Fixation is a special challenge in PLP-null and P0-null mutants since the majority of compact myelin proteins, which are the substrates for glutaraldehyde Wxation, are missing. Individual chapters in this volume are dedicated to each myelin protein. Salient features of individual proteins as they relate to compact myelin structure are brieXy summarized.

P0 Protein P0 constitutes as much as 70% of the total proteins in PNS myelin. It is a type I transmembrane glycoprotein with a single extracellular immunoglobulin-like domain, one transmembrane domain, a cytoplasmic C-terminal, and an apparent molecular weight of 30 kD (Lemke and Axel, 1985). When transfected into nonadherent cells in vitro, the extracellular domain of P0 protein mediated homophilic, but not heterophilic plasma membrane cell adhesion (D’Urso et al., 1990; Filbin et al., 1990). The molecular inter˚ resolution of actions responsible for P0 homophilic adhesion have been delineated by 1.9A P0 extracellular domain crystals (Shapiro et al., 1996). The extracellular domains of P0 interact as cis-linked tetramers that bind through hydrogen bonds to P0 tetramers in

COMPACT MYELIN

FIGURE 1.5 Alterations in the periodicity of compact myelin due to the absence of proteolipid protein (PLP) and the presence of 2’3’-cyclic nucleotide 3’-phosphodiesterase (CNP) (D). The periodicity of compact CNS myelin is compared in wild-type (A) and PLP deWcient md (B) rats and in wild-type (C) and CNP overexpressing (D) mice. The absence of PLP results in the apparent ‘‘fusion’’ of the extracellular leaXets of compact myelin. Overexpression of CNP by oligodendrocytes results in mistargeting of CNP to compact myelin where it prevents fusion of cytoplasmic leaXets of compact myelin (D). Arrowheads in D mark transition from normal to abnormal major dense lines. Lines between panels A and B and panels C and D align major dense lines. Ax ¼ axon. Reproduced from Duncan et al., 1989 (A and B), and Yin et al., 1997 (C and D), with permission. Scale bars ¼ 50 nm.

opposite orientation on the opposing plasma membrane. Cis tetramerization occurs through polar interactions that may be stabilized by trans interactions. The intraperiod line of PNS myelin, therefore, is stabilized by P0 homophilic interaction in the same and opposing compact myelin membranes. Spacing between the extracellular leaXets of compact PNS myelin is altered in P0-null mice (Giese et al., 1992), but the major dense line is ultrastructurally similar to wild-type mice.

Myelin Basic Protein MBP is localized at the cytoplasmic surface of compact CNS and PNS myelin. It is a family of alternatively spliced, highly charged extrinsic membrane proteins (GreenWeld et al., 1973; Zeller et al., 1984), which bind negatively charged lipids, especially phosphatidylserine residues. In MBP-deWcient shiverer mice, the cytoplasmic leaXets of compact CNS myelin are not fused (Privat et al., 1979). In contrast, the major dense line of PNS myelin is unaltered. MBP, therefore, maintains the major dense line of CNS but not PNS myelin. It has been hypothesized that the positively charged cytoplasmic domain of P0 protein can maintain the major dense line of PNS myelin.

Proteolipid Protein Proteolipid protein (PLP) is the most abundant CNS myelin protein. It is a hydrophilic protein with four membrane-spanning domains and a MW of 26 kD. In rodents with PLP mutations and little or no PLP in compact myelin (Duncan et al., 1988, 1989), the extracellular leaXets of compact CNS myelin are fused (Figs. 1.5 A and 1.5B). In PLP-null mice, the extracellular leaXets of many CNS myelin lamellae were separated by distances greater than 2 nm (GriYths et al., 1998). These diVerences in MD rat and PLPnull mice may reXect diVerences in Wxation or tissue processing. PLP lacks the homophilic

9

10

1. STRUCTURE OF THE MYELINATED AXON

binding potential described for P0 protein. It maintains the 2.0 nm spacing between extracellular leaXets of compact CNS myelin by electrostatic interactions with myelin lipids. As described later in this chapter, PLP may be essential for long-term survival of some CNS axons.

Peripheral Myelin Protein-22 Peripheral myelin protein-22 (PMP-22) is present in compact PNS myelin (Snipes et al., 1992). It is a hydrophilic protein with four transmembrane domains and a MW of 22 kD. While absence has little eVect on compact myelin ultrastructure, PMP-22-null mice initially exhibit a severe hypomyelination (Adlkofer et al., 1995). PMP-22 may have additional roles in Schwann cells that are related to its function as a growth arrest protein (Bosse et al., 1994; Zoidl et al., 1995).

P2 Protein P2 is an extrinsic membrane protein that is enriched in compact PNS myelin. P2 has also been detected in CNS myelin of rabbits and humans (Trapp et al., 1983), but not in mice or rats. In the CNS of those species that contain P2 protein, it is abundant in spinal cord and brain stem myelin but not in cerebral cortex myelin. The reasons for the species diVerences and caudal to rostral gradient in CNS P2 expression are unknown. P2 protein can induce an autoimmune PNS demyelinating disease when injected subcutaneously with complete Freunds adjuvant (Kadlubowski et al., 1980). P2 protein has not been eliminated from the mouse genome and P2 mutations in rodent or humans have not been identiWed. P2 belongs to a family of proteins that function in fatty acid transport (Uyemura et al., 1984). It may also participate in fusion of the major dense line of compact PNS myelin.

2’,3’-Cyclic Nucleotide 3’-Phosphodiesterase (CNP) CNP is concentrated on the cytoplasmic side of noncompact myelin membranes (Braun et al., 1988; Trapp et al., 1988). As its name implies, CNP is capable of catalyzing the hydrolysis of 2’,3’ nucleotides into their corresponding 2’ nucleotides (Drummond et al., 1962; Sprinkle et al., 1978). A CNS substrate for this enzymatic activity has not been identiWed. When overexpressed in mice, CNP is mistargeted to compact CNS myelin membranes (Figs. 1.5C and 1.5D). These membranes are deWcient in MBP and do not form major dense lines (Yin et al., 1997). CNP, therefore, may help target MBP to wildtype compact myelin by inhibiting MBP binding to noncompact myelin.

Myelin Protein Dosage Considering all the complex roles of compact myelin proteins, it is not surprising that their stoichiometry is as important as their presence. Alterations in myelin protein gene dosage often have chronic rather that acute manifestations. One allele of the human PMP-22 gene causes hereditary neuropathy with liability to pressure palsy (HNPP). Three copies of the human PMP-22 gene cause Charcot-Marie-Tooth disease type 1A. Duplication of the PLP gene (two alleles, PLP is on the X chromosome) causes the inherited CNS hypomyelination disorder Pelizaeus Merzbacher disease. Alterations in PMP-22 and PLP protein dosage are the major cause of inherited diseases of myelin.

NONCOMPACT MYELIN When comparing the ultrastructure of PNS (Fig. 1.3A) and CNS (Fig. 1.3B) myelinated Wbers, one is struck by the similar appearances of both the compact myelin and the axon.

NONCOMPACT MYELIN

While the periodicity of compact CNS and PNS myelin diVers, they still remain diYcult to distinguish without morphometric comparisons. CNS and PNS myelinated Wbers can be easily distinguished by the presence of a basal lamina around PNS, but not CNS, myelinated Wbers. They also diVer by the distribution and ultrastructure of cytoplasmic channels at the outer margin of the internode. CNS Wbers contain a single external or outer tongue process of oligodendrocyte cytoplasm, which occupies 5 to 20% of the outer circumference. This cytoplasmic channel longitudinally traverses the entire internode and connects oligodendrocyte perinuclear cytoplasm and paranodal cytoplasm. When viewed in cross section, the outer tongue process usually contains transversely sectioned microtubules, occasional mitochondria, smooth endoplasmic reticulum, and vesicles. The remainder of the outer membrane of CNS internodes appears identical to all other compact myelin lamellae: they form a major dense line and can form an intraperiod line with closely apposed neighboring internodes. The external surface of PNS myelin Wbers is surrounded by a basal lamina. In addition, the endoneurial or extracellular space of peripheral nerves is conspicuously abundant and contains collagen Wbrils. Basal lamina and collagen are reliable criteria to distinguish PNS and CNS myelinated Wbers. The Schwann cell basal lamina contains laminin-2 and collagens II and IV (Carey et al., 1983; Cornbrooks et al., 1983). The outer or abaxonal membrane of the PNS internode diVers molecularly and structurally from the compact myelin. The outer surface of each PNS internode contains multiple cytoplasmic channels (Fig. 1.6A). A major channel that extends the entire length of the internode contains the outer mesaxon, which is comprised of the membranes that connect the abaxonal Schwann cell membrane and compact myelin. The cytoplasmic surface of the Schwann cell membrane that forms the channels contain CNP (Trapp et al., 1988) and the microWlament components, Wlamentous actin (F-actin), spectrin, and ankyrin (Trapp et al., 1989b). Schwann cell channel membranes are also rich in cavaeoli, which contain caveolin (Mikol et al., 1999); cavaeoli are not a prominent feature of CNS internodes. The surface membrane of the nonchannel regions of the outer surface of PNS myelin internodes are separated from compact myelin by about 10 nm and is molecularly characterized by the presence of dystroglycan-related protein 2 and periaxin (Sherman et al., 2001), but not by CNP, microWlaments or caveolin (Fig. 1.6B). Disruption of the molecular organization of the abaxonal membrane can cause dysmyelination. Periaxin-null mice have a peripheral neuropathy characterized by dysmyelination, demyelination, and remyelination (Gillespie et al., 2000; Guilbot et al., 2001). The abaxonal nonchannel regions of the PNS myelinated internodes were conspicuously absent in these mice. Normal Schwann cell myelination, therefore, requires the development of specialized abaxonal transport conduits. Despite their diVerences, the inner surface of both PNS and CNS myelin internodes are separated from the axon by a 12 to 14 nm periaxonal space. The periaxonal membranes contain MAG, CNP, and microWlaments (Trapp et al., 1988 1989b). MAG is also present in the inner mesaxon membrane, which connects compact myelin and the periaxonal membrane. The periaxonal membranes of all PNS Wbers and developing CNS Wbers are separated from the compact myelin by a gap of at least 10 nm. Portions of the periaxonal membrane of mature CNS Wbers can form a major dense line with the compact myelin lamellae. A single cytoplasmic channel is present at the inner surface of both CNS and PNS internodes. These channels longitudinally traverse the entire internode and are distinguished from the remainder of the periaxonal region by increased cytoplasm, occasional organelles, and the presence of the inner mesaxon. Compact and noncompact myelin membranes are contiguous with each other and with the plasma membrane of myelin-forming cells. However, they diVer ultrastructurally and have distinct nonoverlapping protein compositions. Many noncompact myelin proteins serve specialized functions and are discussed later in this chapter. Others have more generalized functions and are enriched in all or some noncompact membranes. The distribution of proteins in the myelin internode was initially inferred from light microscopic immunocytochemistry. Figures 1.7A and 1.7B compare the light microscopic distribution of P0 protein, the major structural protein of compact PNS myelin, with that of MAG. Based on these and other comparisons, Trapp and Quarles (Trapp and Quarles,

11

12

1. STRUCTURE OF THE MYELINATED AXON

FIGURE 1.6 Cytoplasmic channels at the outer perimeter of PNS myelin internodes. In freeze fracture preparations (A), longitudinally and radially oriented cytoplasmic channels are raised above nonchannel regions, and the surface of the outer mesaxon (A, arrows) appears as a furrow in the large longitudinal channel. Numerous cavaeoli (A, arrowheads) pit the channel surface. Membranes in channel and nonchannel regions diVer in molecular composition (B). Channel membranes are rich in actin-associated cytoskeletal components, cavaeolin, and CNP, while nonchannel membranes contain dystroglycan/sarcoglycan/periaxin-complexes that include dystrophin-related protein 2 (DRP-2). A basal lamina surrounds the entire surface of PNS internodes. Panel A has been reproduced from Mugnaini et al., 1977, with permission. Scale bars ¼ 1 mm.

1984) proposed that MAG was enriched in the periaxonal membrane, as well as in the membranes of Schmidt-Lanterman incisures, paranodal loops, and inner and outer mesaxons. In addition, they noted that all MAG-containing membranes apposed other membranes by 12 to 14 nm and did not form the major dense lines characteristic of compact myelin. MAG, therefore, is a molecular marker for the membranes that demarcate the cytoplasmic domains within the PNS internode. In CNS myelin, MAG is enriched in the periaxonal membrane and the CNS inner mesaxon, but not in paranodal membranes or outer mesaxons (Trapp et al., 1989a). Because the size of the immuno-complexes is approximately 10 times larger than that of a single membrane (Trapp and Quarles, 1984), a comparison of P0 and MAG distributions in 1 mm thick serial section is only suggestive of nonoverlapping distributions. However, electron microscopic immunocytochemistry, utilizing ultrathin cryosections and smaller gold-labeled immunoreagents (Trapp et al., 1989a), clearly demonstrated the nonoverlapping distributions of P0 and MAG inferred from light microscopic studies. As the molecular structure of myelin is better characterized, the need for electron microscopic immunocytochemistry will increase. This is especially the case for molecules enriched in paranodal regions. Paranodal loops contain multiple subdomains that are diYcult to resolve at the light microscopic level. Two general approaches, pre-embedding (staining before sectioning) and post-embedding (staining cut sections) are successfully used to localize antigens ultrastructurally. Each approach has its advantages and disadvantages. The trade-oV is between tissue Wxation and penetration of immunoreagents. Preembedding methods utilize tissue slices (30 to 100 mm thick) (Butt et al., 1999) or teased nerve Wbers Wxed with low concentrations (0.1 to 0.5%) of glutaraldehyde. Glutaraldehyde is needed for optimal Wxation and it irreversibly cross-links proteins. Too much glutaraldehyde Wxation will inhibit penetration of immunoreagents into the tissue. Glutaraldehyde can alter the antigenicity of proteins, but this is usually the exception rather than the

NODES OF RANVIER

FIGURE 1.7 Light microscopic (A, B) and electron microscopic (C, D) distributions of P0 (A, C) and MAG (B, D) in PNS myelin. By light microscopy, the distributions of P0 and MAG appear diVerent in the same Wbers (arrows), but the section thickness and light microscopic resolution limit interpretation. Immunogold labeling of frozen sections indicates that P0 is concentrated in compact membranes and excluded from incisural (*), periaxonal, and abaxonal membranes (C). MAG (D) is concentrated in incisures and periaxonal membrane (arrowheads) but not in compact myelin or at the cell surface. Reproduced from Trapp and Quarles, 1982 (A and B), and GriYn et al., 1993 (C and D), with permission. Scale bars ¼ 10 mm (A, B) and 0.2 mm (C, D).

rule. The use of post-embedding methods avoids the penetration limitation of stringent (2 to 5%) glutaraldehyde Wxation. Immunostained ultrathin cryosections (Trapp et al., 1989a) or etched embedded sections (Scherer et al., 1995b) provide the best ultrastructure, and antigens exposed at the surface are accessible to immunoreagents. While post-embedding methods are technically more demanding than pre-embedding techniques, both approaches have determined the ultrastructural distribution of myelin proteins.

NODES OF RANVIER CNS and PNS nodes serve identical functions, and are morphologically similar but not ultrastructurally identical (Fig. 1.8). From a functional and ultrastructural point of view, three regions can be distinguished: node, paranode, and juxtaparanode (Fig. 1.9A). Each region contributes to saltatory conduction and is maintained by complex molecular interactions that have been compared with those operating at the neural and immunological synapse (Pedraza et al., 2001). We are just beginning to understand the complex molecular interactions at nodes of Ranvier; interactions in which many participants still remain to be identiWed.

Nodes The node is a small stretch of bare axonal plasma membrane (axolemma) that separates myelin segments along individual nerve Wbers. Easily identiWed in longitudinal sections, nodes separate the terminal paranodal loops of adjacent myelin internodes (Figs. 1.8A and

13

14

1. STRUCTURE OF THE MYELINATED AXON

FIGURE 1.8 Nodes of Ranvier in longitudinal (A, B) and transverse (C, D) orientations from PNS (A, C) and CNS (B, D). The nodal axonal (Ax) membrane extends between the arrowheads (A, B), covering approximately 1 mm, and bulges outward in these relatively small myelinated Wbers. To either side, the paranodal loops of the terminating internodes indent the axon and demarcate the paranodal region. In transverse section (C), the PNS nodal membrane is surrounded by numerous microvilli (mv), while CNS nodes (D) may be surrounded by glial processes or simply by amorphous extracellular matrix material. The nodal axolemma contains an electron dense undercoating (C, D small arrowheads). Panels A, B, and D have been reproduced from Peters et al., 1991, and panel C from Boyle et al., 2001, with permission. Scale bars ¼ 0.5 mm (A, B) and 0.2 mm (C, D).

1.8B). Nodal length is related to the diameter of the axon and can vary from less than 1 mm in the small Wbers of the optic nerve to more than 5 mm in the large Wbers of the spinal cord (Fraher, 1973; McDonald and Ohlrich, 1971). Nodal axon diameter usually diVers from paranodal and internodal, ranging from a conical bulge in small Wbers of the CNS to constriction in large Wbers of the PNS. When constricted, the density of microtubules, vesicles, and neuroWlaments increases. NeuroWlament spacing is reduced and reXects, in part, reduced neuroWlament phosphorylation and reduced neuroWlament sidearm extension (Mata et al., 1992).

NODES OF RANVIER

FIGURE 1.9 Molecular composition and structure of nodal regions in longitudinal (A, C, D) and transverse orientation (B). Three regions (nodal, paranodal, and juxtaparanodal) can be deWned morphologically and each has a unique molecular composition (A). Nodal axolemma contains high densities of Naþ channels and adhesion molecules such as Nr-CAM and neurofascin 186 (NF186), and an actin cytoskeleton enriched in ankyrin G. In the paranodal region, axoglial junctions involve complexes of axonal Caspr and contactin and glial neurofascin 155 (NF155). Adjacent paranodal loops are stabilized through E-cadherin/ß-catenin–based adherens junctions, and spacing is regulated by the presence of MAG. In transverse section (B), the paranodal loops appear as noncompacted membranes containing substantial cytoplasm. In longitudinal sections (C), the septate junctions (arrows) appear as punctate structures spanning the cleft between the axonal surface and paranodal loops. Contactin-null mice lack septate junctions (D, arrows). Juxtaparanodal axolemmae are enriched in Kþ channels and caspr2. Reproduced from Trapp and Kidd, 2000 (A), Bosio et al., 1998 (C), and Boyle et al., 2001 (D), with permission. Scale bars ¼ 0.5 mm (B) and 0.1 mm (C, D).

The nodal axolemma is the functional ultimate of myelin formation as it contains voltage-gated Naþ channels and thereby is directly responsible for saltatory conduction. The channels were initially visualized by freeze-fracture electron microscopy as 10 nm diameter axolemmal particles, at a density of 1000 to 1500/mm2 (Rosenbluth, 1976). Although initial estimates suggested many more channels were present, the observed particle density matched the electrophysiological estimates of nodal Naþ channel concentration (Chiu, 1980; Waxman and Ritchie, 1985). IdentiWcation of nodes in cross section (Figs. 1.8C and 1.8D) is not as easy. One must rely on more subtle characteristics of the nodal axolemma and surrounding structures. One ultrastructural characteristic is a dense 20 nm thick coating beneath the axolemma. This coating is similar to that at the initial axon segment and correlates with the presence of high concentrations of Naþ channels (Caldwell et al., 2000; Ellisman and Levinson, 1982). It represents a specialized submembranous cytoskeleton that is connected to a multimolecular complex in the nodal axolemma (Fig. 1.9A). In addition to voltage-gated Naþ channels, membrane proteins enriched in the nodal axolemma include an isoform of Naþ/ Kþ-ATPase, neurofascin-186 and Nr-CAM (Ariyasu and Ellisman, 1987; Davis et al., 1996; Tait et al., 2000). The submembranous cytoskeleton contains spectrin, f-actin, and two alternately spliced isoforms of the ankyrinG (270 and 480 kD) (Kordeli et al., 1990, 1995). The dense undercoating is separated from the nodal axolemma by 10 nm; it helps to stabilize Naþ channels at the node. AnkyrinG may also associate with Naþ channels in the neuronal soma and assist in targeting Naþ channels to the node (Kordeli et al., 1995). Although PNS and CNS nodes can appear similar by TEM, each is surrounded by distinctly diVerent immediate environments. The nodal axolemma is surrounded by Schwann cell microvilli in the PNS and by glial cell processes in the CNS. The microvilli, which extend from the outer border of the adjacent myelin internodes, are

15

16

1. STRUCTURE OF THE MYELINATED AXON

best visualized in cross sections, where they appear as thin processes (Fig. 1.8C). Nodal microvilli are enriched in actin Wlaments and members of the ERM family of actin associated proteins (Kordeli et al., 1990; Melendez-Vasquez et al., 2001; Scherer et al., 2001). The Schwann cell basal lamina is continuous across the node but does not surround individual microvilli. The glial cell processes surrounding CNS nodes originate from astrocytes (Hildebrand and Waxman, 1984) and possibly oligodendrocyte progenitor cells (Butt et al., 1999). These CNS glial processes do not, however, closely adhere to all regions of the nodal axolemma. Where absent, granular-appearing extracellular matrix, which may bind ions and participate in saltatory conduction, surrounds the CNS nodal axolemma. Little is known about the molecular characterization of the glial processes at CNS nodes.

Paranodal Regions Paranodal loops are the lateral ends of the myelin internode. They consist of a longitudinally oriented spiral of Schwann cell cytoplasm that is tightly tethered to the axon and to each other by junctional complexes. Paranodal ultrastructure is best observed in longitudinal orientation where they appear on each side of the node as a series of Schwann cell cytoplasmic pockets (Figs. 1.8A and 1.8B). As compact myelin approaches the node, the major dense line ‘‘opens up’’ to accommodate cytoplasm. The extracellular leaXets of neighboring paranodal loops oppose each other by 12 to 14 nm. In this respect, they are similar to SchmidtLanterman incisures, and in transverse orientation (see Fig. 1.9B), they may be mistaken for incisures. They diVer from incisures in that they usually contain more cytoplasm and organelles. In smaller Wbers, paranodal loops are arranged in an orderly manner with the inner compact myelin lamellae terminating Wrst and the outer lamellar terminating last, directly adjacent to the node. Paranodal loop alignment in larger diameter Wbers follows a similar but less organized pattern, as some loops do not reach the axonal surface. The paranodal membrane that apposes the axolemma is slightly curved, but separated from the axolemma by a constant distance of 2.5-3.0 nm (Peters et al., 1991). The axolemma in paranodal regions, therefore, appears as a series of shallow troughs with each corresponding to a terminal loop. Between paranodal loops and axons are evenly spaced densities that appear to connect the two membranes (Fig. 1.9C). The densities are each about 15 nm long and 25 to 30 nm apart and because of their orientation were originally referred to as transverse bands. Based on a variety of experimental approaches, it is now clear that these densities represent septate-like junctions and are the major adhesive apparatus between myelin and axons. Paranodal membranes that oppose the axolemma contain large intramembranous particles that appose with a double row of smaller intramembranous particles on the axolemma (Wiley and Ellisman, 1980). Recent studies have begun to characterize the molecular composition of axoglial septate junction (Fig. 1.9A). The membrane proteins, contactin and Caspr, are enriched in paranodal axolemma in both the CNS and PNS (Einheber et al., 1997; Menegoz et al., 1997; Rios et al., 2000). They function as heterodimers that may be assembled prior to insertion in the axolemma. Paranodal loops contain an isoform of neurofascin (NF155), which co-localizes with and possibly binds to contactin and Caspr (Charles et al., 2002; Tait et al., 2000). These molecules have homologues associated with septate junctions in invertebrates, and axoglial junctions are abnormal or missing in Caspr and contactin-null mice (Fig. 1.9D) (Bhat et al., 2001; Boyle et al., 2001). Clustering of Naþ channels at the PNS node is not grossly altered in contactin-null mice. Shaker type Kv1.1 and Kv1.2 potassium channels, which are enriched in the juxtaparanodal axolemma and absent from paranodal axolemma in wild-type mice, are detected in paranodal axolemma in mice deWcient in CGT (Dupree et al., 1998), Caspr (Bhat et al., 2001), or contactin (Boyle et al., 2001). The paranodal axoglia junction, therefore, appears to play a role in spatially separating Naþ and Kþ channel distribution at the node of Ranvier.

Juxtaparanodes The juxtaparanodal region, is deWned as the 10 to 15 nm of the myelin internode that is adjacent to the paranodal loops. Other than this spatial relationship to the paranodes,

SCHMIDT-LANTERMAN INCISURES

FIGURE 1.10 Ultrastructure of Schmidt-Lanterman incisures. Drawing by Ramon y Cajal (A) shows arrangement of ‘‘funnelshaped’’ incisures in a teased PNS myelinated Wber. When incisures are sectioned longitudinally and examined by electron microscopy (B), the spiraled cytoplasmic channel transverses the compact myelin at an angle, but incisure membranes are in register with each other. When sectioned in transverse orientation (C), they appear as noncompacted membrane spirals (*) that are contiguous with compact myelin. Ax ¼ axon. Arrow in panel A depicts plane of section for panel C. Reproduced from Ramon y Cajal, 1928 (A), Mugnaini et al., 1977 (B), and Trapp and Quarles, 1982 (C), with permission. Scale bars ¼ 0.5 mm (A) and 1 mm (B).

TEM characteristics of the juxtaparanode are indistinguishable from the remainder of the myelin internode. Freeze fracture electron microscopy, however, identiWed clusters of intramembranous particles in the juxtaparanodal axolemma (Rosenbluth, 1976). These particles correspond to the distribution of delayed rectifying Kþ channel. They have a similar size (10 nm) to that predicted for Kþ channels that function in myelinated Wbers as mixed tetramers of the Kv1.1 and Kv1.2 subunits. Immunocytochemistry has localized Kv1.1, Kv1.2, and their associated subunit b2 to juxtaparanodal regions (Rasband et al., 1998). In addition, a homologue of Caspr, Caspr 2, is enriched in the juxtaparanodal axolemma (Poliak et al., 1999). Potassium channels are thought to play a role in saltatory conduction by ‘‘dampening’’ the excitability of the node. This may be important in facilitating rapid, repetitive Wring of the action potential.

SCHMIDT-LANTERMAN INCISURES As part of the Wrst reports on the morphology of PNS myelinated axons, Schmidt and Lanterman independently described a series of funnel-shaped clefts or incisures that interrupted the dense appearance of myelin (Figs. 1.1A and 1.10A). These clefts, called Schmidt-Lanterman incisures, traverse the thickness of the myelin internode at an average angle of 9 degrees. They are abundant between the perinuclear region and

17

18

1. STRUCTURE OF THE MYELINATED AXON

paranodes and their total number within individual internodes increases with myelin sheath thickness and internodal length. The largest Wbers in rat sciatic nerve have approximately 25 incisures. They are present during early stages of myelination and in remyelinated internodes. Their existence as a normal feature of the PNS internode or as a byproduct of poor Wxation or mechanical stress sheering defect was initially debated (Peters et al., 1991). However, electron microscopy established ultrastructural criteria for their identiWcation as Schwann cell cytoplasmic channels that are conXuent with the cytoplasm of the outer and inner margins of the internode (Robertson, 1958). When examined ultrastructurally in longitudinal orientation, Schmidt-Lanterman incisures appear as a series of cytoplasmic openings of the compact myelin that are in register with each other (Robertson, 1958; also see Ghabriel and Allt, 1981; Hall and Williams, 1970; Peters et al., 1991) (Fig. 1.10B). Because they traverse the compact myelin at an angle, only portions of each incisure are viewed in cross section of myelinated Wber (Fig. 1.10C). They appear as spiral wraps of Schwann cell membranes, which, depending on the plane of section, can be located at the outer, middle, or inner aspects of the compact myelin. Whether viewed in longitudinal or transverse orientation, the membranes of Schmidt-Lanterman incisures are distinguished from compact myelin by two ultrastructural criteria: the presence of Schwann cell cytoplasm (the major dense lines are not fused) and apposition of extracellular leaXets by 12 to14 nm. The cytoplasm can contain microtubules, intermediate Wlaments, and small vesicles. Incisure membranes contain MAG and adherens-like junctions at their outer margins. The functions of Schmidt-Lanterman incisures are not fully understood. It has been suggested that they play a role in myelin formation. To date, little information exists that supports their role in transporting proteins or lipids required for myelin formation or turnover. It has also been suggested that the incisures are dynamic structures that can rapidly form and migrate along the internode. This hypothesis was based on static images of partially formed incisures (Mugnaini et al., 1977). In vivo observation, however, has not conWrmed translocation of incisures along PNS internodes (Hall and Williams, 1970). This is because individual incisures can dilate and contract when subjected to hypotonic or hypertonic solutions (Hall and Williams, 1970) and incisural changes are among the Wrst steps in myelin internode fragmentation after axonal transection (Ghabriel and Allt, 1979). One function of Schmidt-Lanterman incisures appears to be that they act as a beneWcial ‘‘shortcut’’ for communication between the outer and inner aspects of the myelin internode. It has been estimated for large PNS internodes that the radial pathway through incisures is 1000 times shorter than traversing the cytoplasmic channels at the outer margins of the myelin internode (Balice-Gordon et al., 1998). Radial diVusion of small ions through incisures has now been documented (Balice-Gordon et al., 1998). However, such diVusion does not occur through the cytoplasm of the funnel-shaped incisures but by gap junction-mediated diVusion through incisure membranes. These data were somewhat surprising, as classical gap junctions have not been identiWed ultrastructurally as part of Schmidt-Lanterman incisures. Incisures do contain the gap junction protein connexin-32 (Cx-32) and pharmacological blockers of gap junction function inhibited radial diVusion of ions (Balice-Gordon et al., 1998; Scherer et al., 1995a). Disruption of this radial pathway may contribute to the inherited PNS dysmyelination resulting from Cx-32 mutations in individuals with X-linked Charcot-Marie-Tooth disease. Radial diVusion, however, appeared normal in Cx-32-null mice (Balice-Gordon et al., 1998). Connexin 29 is also present in incisures and juxtaparanodal regions (Altevogt et al., 2002; Li et al., 2002), but appears unable to form channels alone (Altevogt et al., 2002). Other gap junction proteins such as Cx-26 or Cx-43 may also exist in incisures and compensate for the loss of Cx-32. In either event, further characterization of junctional connections between Schmidt-Lanterman incisure membranes will be required to fully understand the function of these cytoplasmic domains. Incisure-like structures have been identiWed in electron micrographs of larger CNS myelin internodes (Blakemore, 1969). They diVer from PNS incisures in that they do not contain MAG. The lack of molecular markers speciWc for these CNS structures has hindered their identiWcation and quantiWcation. It is generally assumed, however, that they are not prominent features of CNS internodes.

RADIAL COMPONENT OF CNS MYELIN

RADIAL COMPONENT OF CNS MYELIN An ultrastructural feature present in CNS but not PNS compact myelin is the radial component. In transverse orientation, the radial component appears as registered lines of junctional complexes that extend partially or totally across the thickness of the myelin sheath (Fig. 1.11). These intralamellar junctional complexes resemble tight junctions in that they have reduced or no spacing between extracellular leaXets and what appears to be an increase in the electron lucent core of the membrane bilayer. Radial components are also in register when the myelin sheath is visualized in longitudinal orientation and they extend from paranode to paranode. Upon freeze fracture analysis, intramembranous particles typical of tight junctions in other tissues have distributions identical to the radial components identiWed in thin sections. The major dense line shows little variation in the radial component. It should be noted that the radial component is not optimally visualized in tissue sections routinely processed for electron microscopy (glutaraldehyde Wxed epoxyembedded sections stained with uranyl acetate and lead citrate). It was Wrst identiWed in osmic acid Wxed sections stained with potassium permanganate (Peters, 1961) and eloquently displayed in sections embedded in the water-soluble embedding media, Durcupan (Tabira et al., 1978). Radial components are abundant and randomly distributed in developing myelin internodes. In the mature myelin internode, radial components are most common beneath the outer tongue process. Since the outer and inner tongue processes are often in the same quadrant, radial components connect these cytoplasmic channels. While there are multiple radial components in this area, they can appear at other locations and the radial components of adjacent sheath appear to align with each other. Radial component assembly appears to require the tight junction protein, claudin11 (Gow et al., 1999; Morita et al., 1999). What is the function of the radial component and why are they not found in PNS myelin? The answer may lie in diVerences in the major structural protein of PNS and CNS myelin. PLP does not possess homophilic binding properties of P0 protein. PLP helps maintain the 2.0 nm spacing between the extracellular leaXets of compact CNS myelin through electrostatic interactions with myelin lipids. Certain pathological conditions, such as hexachlorophene (HCP) or triethyl tin intoxication, induce white matter edema and large intramyelinic vacuoles that result from splitting of the extracellular leaXets of compact CNS myelin. These myelin vacuoles remain tethered to the axons in the area of the inner tongue process where tight junctions predominate (Tabira et al., 1978) (Fig. 1.11D). Transient storage of extracellular water between the relatively inert compact myelin lamellae may minimize nodal dysfunction. Despite the large vacuoles, demyelination is rare in acute HCP intoxication and the lesions are reversible (Tabira et al., 1978). The radial component, therefore, appears to help maintain CNS myelin sheath integrity and limit demyelination in certain toxin-induced pathological conditions that result in white matter edema.

MYELINATION MODULATES THE MATURATION AND SURVIVAL OF AXONS It has been recognized for decades that axons regulate the myelinating phenotype of oligodendrocyte and Schwann cells. Conversely, myelination also inXuences axonal maturation and survival. We have little knowledge of the molecules directly responsible for either eVect. Electron microscopy, however, has deWned axonal parameters that are aVected by myelination and myelin pathology. Because axonal pathology and degeneration are major causes of neurological disability associated with inherited diseases of myelin, axonal pathology is a therapeutically important aspect of current myelin research. Myelination increases the diameter of axons (Windebank et al., 1985) and helps directs Naþ channels to the node of Ranvier (Pedraza et al., 2001). It is important to note, however, that axonal caliber is an inherent property of axons. Myelination is required for each axon to attain its maximal diameter. This eVect is important functionally because

19

20

1. STRUCTURE OF THE MYELINATED AXON

FIGURE 1.11 Radial component of CNS myelinated Wbers. In transversely orientated electron micrographs (A), the radial component consists of thin translucent bands in which the compact myelin outer leaXets are fused (arrows). These bands traverse compact myelin lamellae from the inner (*) to the outer (**) mesaxons. The radial components of adjacent internodes may be aligned (B, arrows). Schematic arrangement of radial components (C, arrows). Radial components extend both transversely between lamellae and longitudinally along the internode. In hexachlorophene intoxication (D), vacuolation involving the minor dense lines does not disrupt the radial components. Ax, axon. Panels A, B, and D have been reproduced from Tabira et al., 1978, with permission. Scale bars ¼ 0.5 mm (A) and 0.1 mm (B).

axonal caliber is a determinant of conduction velocity (Arbuthnott et al., 1980) (see Chapter 5). Myelination results in signiWcant increases in the number and phosphorylation state of neuroWlaments, the major component of the axonal cytoskeleton (de Waegh et al., 1992; Hsieh et al., 1994; Sanchez et al., 1996). Phosphorylation increases the negative charge and lateral extension of side arms from the 10-nm thick neuroWlaments. This side arm extension, in turn, increases neuroWlament spacing and axonal caliber. Nonmyelinated axons and nonmyelinated regions of myelinated Wbers (nodal axons and initial segments of DRG axons) (de Waegh et al., 1992; Hsieh et al., 1994) have fewer neuroWlaments, low neuroWlament phosphorylation, little neuroWlament sidearm extension, close neuroWlament spacing, and small diameters. The ultimate dimensions of the axons and myelin are determined by a complex interplay between axons and the myelinating cells. In normal nerves, the ratio of axonal diameter and myelin thickness is called the g-ratio and remains remarkably consistent across a wide variety of axonal diameters; in humans the g-ratio is approximately 0.6 (Friede and Bischhausen, 1982). Internodal length also correlates with axonal diameter and myelin thickness (Friede and Bischhausen, 1980), although some variation between Wber popula-

MYELINATION MODULATES THE MATURATION AND SURVIVAL OF AXONS

tions is normally observed. As noted earlier, the axonal diameter is an intrinsic property of the axon under normal tissue conditions, but is not achieved without myelination (Windebank et al., 1985). In addition, many extrinsic factors have been shown to aVect axonal caliber, including mechanical compression of the nerve, alterations of impulse conduction frequency, toxins, mutations aVecting the axonal cytoskeleton, and aberrant myelination. It is not known how myelinating cells ‘‘read’’ the properties of the axon in order to match myelin production to axonal dimensions. In animals with aberrantly small, neuroWlament-deWcient axons, oligodendrocyte myelination was proportional to axonal diameter, while Schwann cell internodes were disproportionately thick (Elder et al., 2001). These data suggest that Schwann cells and oligodendrocytes require or utilize diVerent axonal cues to regulate myelin sheath thickness. As previously mentioned, axonal loss is a major cause of permanent neurological disability in inherited and acquired diseases of myelin. Axonal loss in inherited diseases of myelin often occurs as a late-onset manifestation of dysmyelination. Slow nerve conduction velocities, a physiological indication of dysmyelination, are often present early when the patient shows little or no neurological disability. With age, conduction velocities remain the same, but conduction block, an indication of axonal loss, correlates with disease onset and progression. These and other observations have lead to the hypothesis that axons require glial-derived extrinsic trophic support for their survival. A variety of dysmyelinating animal models support this hypothesis and provide early clues to potential mechanisms. The underlying molecular mechanisms responsible for axonal pathology in many inherited diseases of myelin are compounded by chronic dysmyelination, remyelination, and ‘‘gain of function mutations.’’ In contrast, late-onset axonal pathology and degeneration have been described in mice deWcient in MAG (Yin et al., 1998), PLP (GriYths et al., 1998), P0 (Giese et al., 1992), Cx32 (Anzini et al., 1997; Scherer et al., 1998), and complex gangliosides (Sheikh et al., 1999). The mechanisms by which these ‘‘loss of function’’ mutations cause chronic axonal pathology still remain to be fully elucidated. Axonal pathologies that precede axonal degeneration include atrophy, swelling, reduced axonal transport, an altered axonal cytoskeleton, and misregulation of gene expression by the dysmyelinated neuron. The mechanisms by which loss of MAG and PLP induce axonal pathology diVer. Based on its periaxonal location and biochemical properties, MAG was initially considered a potential receptor for an axon ligand that initiated and regulated myelination. Normal myelination in MAG-deWcient mice (Li et al., 1994; Montag et al., 1994) ruled out an essential role for MAG during myelination. In vitro, MAG was subsequently shown to inhibit neurite outgrowth (Geiger et al., 1982; Mukhopadhyay et al., 1994) and cause growth cone collapse (Li et al., 1996). MAG function appears more akin to an axonal ligand that regulates the axonal cytoskeleton (Yin and Trapp, 1997). In support of the hypothesis, MAG-deWcient mice exhibit progressive axonal atrophy that was accompanied by reduced neuroWlament phosphorylation and spacing (Yin et al., 1998). These axonal changes were most prominent in paranodal regions and, to a lesser degree, beneath Schmidt-Lanterman incisures. Because the Wbers myelinate normally, the axonal atrophy resulted in collapse of the myelin sheath and tomacula formation (Fig. 1.12). Tomacula and axonal atrophy are also prominent in mice and humans that contain one allele of the PMP-22 gene (Adlkofer et al., 1997). Like MAG-deWcient mice, these tomacula manifest late in life. While neuroWlament spacing has not been quantiWed in PMP-22 mutants, micrographs published support neuroWlament dephosphorylation and increased neuroWlament density as a contributor to tomacula formation. It is possible, therefore, that the absence of MAG and PMP-22 directly or indirectly modulates axonal kinase or phosphatase activity. Reduced phosphorylation renders neuroWlaments more susceptible to enzymatic degradation (Pant, 1988) and may eventually lead to the axonal degeneration that subsequently results from those mutations. Like MAG deWcient and PMP-22 mutant mice, PLP-deWcient mice initially develop normally, with myelin surrounding CNS axons. Yet this brief stage of apparent normality does not linger. Between 6 and 8 weeks of age, small diameter axons develop focal axonal swellings at distal paranodes that may reXect deWcits in retrograde axonal transport (GriYths et al., 1998). By 12 months, axons with ovoids begin to degenerate, and by 18 months, neurological phenotypes develop. These experimental data support the concept

21

22

1. STRUCTURE OF THE MYELINATED AXON

FIGURE 1.12 Pathologies that precede axonal degeneration in fibers with myelin protein gene defects. Electron micrographs of paranodal regions of control (A) and MAG-deficient (B, C) myelin internodes. In longitudinal orientation, paranodal axonal caliber is much greater in control (A) than in MAG-deficient (B) fibers. In transverse orientation, thick redundant myelin (C, arrowheads) partially surrounds an axon (C, arrow) with remarkably small caliber. (D–I) Schematic representation comparing myelin and axonal pathology in MAG- and PLP-deficient fibers. In MAG deficient mice, paranodes develop normally (D, G). With time, paranodal axonal calibers shrink (E, H) and result in myelin sheath collapse and tomacula formation (F, H). MAG-deficient mice develop a late onset axonal degeneration preceded by this paranodal axon atrophy, reduced neurofilament spacing (H), and tomacula formation. PLP-deficient mice develop late-onset paranodal axonal swellings, caused by apparent defects in retrograde axonal transport (I). Reproduced from Yin et al., 1998 (A through F), and Trapp et al., 1999 (G through I), with permission. Scale bars ¼ 0.5 mm.

that neurological disability in PLP-deWcient mice begins when a threshold of axonal loss is attained. Null mutations in the human PLP gene also result in axonal degeneration and neurological disability that worsens with age (Garbern et al., 2002). Similar to the PLP-null mice, little demyelination is actually associated with the axonal degeneration, which was most prominent in the longest CNS axons. A similar pathobiology of axonal loss has been proposed for permanent neurological disability in patients with multiple sclerosis (Trapp et al., 1999), Charcot Marie-Tooth disease (Scherer, 1999) and other inherited or acquired diseases of myelin. Disruption of glial-axon interactions is likely to be the underlying cause of axonal degeneration, which can occur in the absence of demyelination and in the presence of normal axonal ensheathment by myelinating cells. A major challenge for myelin research, therefore, is to obtain a better understanding of the myelin-axon interactions that maintain normal axonal function and integrity.

SUMMARY AND FUTURE CONSIDERATIONS As illustrated in this chapter, myelinated axons form a complex structure that exempliWes functional cell-cell interactions. While our general knowledge of the molecular compos-

SUMMARY AND FUTURE CONSIDERATIONS

ition of myelin internodes is substantial, we know little about the molecular mechanisms responsible for spiral wrapping of myelin membranes or for axon-myelin-forming cell communication. Future research will identify additional myelin molecules that are essential to the process of myelination and to axonal function. Electron microscopy will continue to be an essential and irreplaceable tool for elucidating the function of individual molecules and may help identify novel therapeutic targets for the treatment of human myelin disease.

Acknowledgments The authors thank Vikki Pickett and Brian Trapp for editorial assistance and Rosalia Yacubova for help preparing the micrographs. Dr. Trapp is supported by the National Institutes of Health grants NS35058, NS38186, NS29818, and NS38667.

References Adlkofer, K., Frei, R., Neuberg, D. H. H., Zielasek, J., Toyka, K. V., and Suter, U. (1997). Heterozygous peripheral myelin protein 22-deWcient mice are aVected by a progressive demyelinating tomaculous neuropathy. J. Neurosci. 17, 4662–4671. Adlkofer, K., Martini, R., Aguzzi, A., Zielasek, J., Toyka, K. V., and Suter, U. (1995). Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deWcient mice. Nature Genet. 11, 274–280. Alberts, B., Bray, D., Lewis, J., RaV, M., Roberts, K., and Watson, J. D. (1994). ‘‘Molecular Biology of the Cell.’’ Garland Publishing, New York and London. Altevogt, B. M., Kleopa, K. A., Postma, F. R., Scherer, S. S., and Paul, D. L. (2002). Connexin29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J Neurosci. 22, 6458–6470. Anzini, P., Neuberg, D. H. H., Schachner, M., Nelles, E., Willecke, K., Zielasek, J., Toyka, K. V., Suter, U., and Martini, R. (1997). Structural abnormalities and deWcient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin 32. J. Neurosci. 17, 4545–4551. Arbuthnott, E. R., Boyd, I. A., and Kalu, K. U. (1980). Ultrastructural dimensions of myelinated peripheral nerve Wbres in the cat and their relation to conduction velocity. J. Physiol. 308, 125–157. Ariyasu, R. G., and Ellisman, M. H. (1987). The distribution of (NaþþKþ)ATPase is continuous along the axolemma of unensheathed axons from spinal roots of ‘‘dystrophic’’ mice. J. Neurocytol. 16, 239–248. Balice-Gordon, R. J., Bone, L. J., and Scherer, S. S. (1998). Functional gap junctions in the schwann cell myelin sheath. J. Cell Biol. 142, 1095–1104. Bhat, M. A., Rios, J. C., Lu, Y., Garcia-Fresco, G. P., Ching, W., Martin, M. S., Li, J., Einheber, S., Chesler, M., Rosenbluth, J., Salzer, J. L., and Bellen, H. J. (2001). Axon-glia interactions and the domain organization of myelinated axons requires neurexin iv/caspr/paranodin. Neuron 30, 369–383. Blakemore, W. F. (1969). Schmidt-Lantermann incisures in the central nervous system. J. Ultrastruct. Res. 29, 496–498. Blakemore, W. F. (1981). Observations on myelination and remyelination in the central nervous system. In ‘‘Development in the Nervous System’’ (D. R. Garrod and J. D. Feldman, eds.), pp. 289–308. Cambridge University Press, Cambridge. Bosio, A., Bussow, H., Adam, J., and StoVel, W. (1998). Galactosphingolipids and axono-glial interaction in myelin of the central nervous system. Cell Tissue Res. 292, 199–210. Bosse, F., Zoidl, G., Wilms, S., Gillen, C. P., Kuhn, H. G., and Mu¨ller, H. W. (1994). DiVerential expression of two mRNA species indicates a dual function of peripheral myelin protein PMP22 in cell growth and myelination. J. Neurosci. Res. 37, 529–537. Boyle, M. E., Berglund, E. O., Murai, K. K., Weber, L., Peles, E., and Ranscht, B. (2001). Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30, 385–397. Braun, P. E., Sandillon, F., Edwards, A., Matthieu, J.-M., and Privat, A. (1988). Immunocytochemical localization by electron microscopy of 2’, 3’-cyclic nucleotide 3’-phosphodiesterase in developing oligodendrocytes of normal and mutant brain. J. Neurosci. 8, 3057–3066. Bunge, M. B., Bunge, R. P., and Pappas, G. D. (1962). Electron microscopic demonstration of connections between glia and myelin sheaths in the developing mammalian central nervous system. J. Cell Biol. 12, 448–453. Butt, A. M., Duncan, A., Hornby, M. F., Kirvell, S. L., Hunter, A., Levine, J. M., and Berry, M. (1999). Cells expressing the NG2 antigen contact nodes of Ranvier in adult CNS white matter. Glia 26, 84–91. Caldwell, J. H., Schaller, K. L., Lasher, R. S., Peles, E., and Levinson, S. R. (2000). Sodium channel Na(v)1.6 is localized at nodes of ranvier, dendrites, and synapses. Proc Natl Acad Sci U. S. A 97, 5616–5620. Carey, D. J., Eldridge, C. F., Cornbrooks, C. J., Timpl, R., and Bunge, R. P. (1983). Biosynthesis of type IV collagen by cultured rat Schwann cell. J. Cell Biol. 97, 473–479.

23

24

1. STRUCTURE OF THE MYELINATED AXON

Charles, P., Tait, S., Faivre-Sarrailh, C., Barbin, G., Gunn-Moore, F., Denisenko-Nehrbass, N., Guennoc, A. M., Girault, J. A., Brophy, P. J., and Lubetzki, C. (2002). Neurofascin is a glial receptor for the paranodin/Casprcontactin axonal complex at the axoglial junction. Curr. Biol. 12, 217–220. Chiu, S. Y. (1980). Asymmetry currents in the mammalian myelinated nerve. J Physiol 309, 499–519. Cornbrooks, C. J., Carey, D. J., McDonald, J. A., Timpl, R., and Bunge, R. P. (1983). In vivo and in vitro observations on laminin production by Schwann cells. Proc. Natl. Acad. Sci. USA 80, 3850–3854. D’Urso, D., Brophy, P. J., Staugaitis, S. M., Gillespie, C. S., Frey, A. B., Stempak, J. G., and Colman, D. R. (1990). Protein zero of peripheral nerve myelin: Biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron 2, 449–460. Davis, J. Q., Lambert, S., and Bennett, V. (1996). Molecular composition of the node of Ranvier: IdentiWcation of ankyrin-binding cell adhesion molecules neurofascin (mucinþ/third FNIII domain-) and NrCAM at nodal axon segments. J. Cell Biol. 135(5), 1355–1367. de Waegh, S. M., Lee, V. M., and Brady, S. T. (1992). Local modulation of neuroWlament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68, 451–463. Drummond, G. I., Lyer, N. T., and Keith, J. (1962). Hydrolysis of ribonucleoside 2’, 3’ cyclic phosphates by a diesterase from brain. J. Biol. Chem. 237, 3535–3539. Duncan, I. D., Hammang, J. P., Goda, S., and Quarles, R. H. (1989). Myelination in the jimpy mouse in the absence of proteolipid protein. Glia 2, 155–160. Duncan, I. D., Hammang, J. P., and Trapp, B. D. (1988). Abnormal compact myelin in the myelin-deWcient rat: Absence of proteolipid protein correlates with a defect in the intraperiod line. Proc. Natl. Acad. Sci. USA 84, 6287–6291. Dupree, J. L., Coetzee, T., Blight, A., Suzuki, K., and Popko, B. (1998). Myelin galactolipids are essential for proper node of Ranvier formation in the CNS. J. Neurosci. 18, 1642–1649. Einheber, S., Zanazzi, G., Ching, W., Scherer, S., Milner, T. A., Peles, E., and Salzer, J. L. (1997). The axonal membrane protein Caspr, a homologue of neurexin IV, is a component of the septate-like paranodal junctions that assemble during myelination. J. Cell Biol. 139, 1495–1506. Elder, G. A., Friedrich, V. L., Jr., and Lazzarini, R. A. (2001). Schwann cells and oligodendrocytes read distinct signals in establishing myelin sheath thickness. J. Neurosci. Res. 65, 493–499. Ellisman, M. H., and Levinson, S. R. (1982). Immunocytochemical localization of sodium channel distributions in the excitable membranes of Electrophorus electricus. Proc. Natl. Acad. Sci. U. S. A 79, 6707–6711. Fanarraga, M. L., GriYths, I. R., Zhao, M., and Duncan, I. D. (1998). Oligodendrocytes are not inherently programmed to myelinate a speciWc size of axon. J. Comp Neurol. 399, 94–100. Fernandez-Moran, H., and Finean, J. B. (1957). Electron microscope and low-angle x-ray diVraction studies of the nerve myelin sheath. J. Biophys. Biochem. Cytol. 3, 725–748. Filbin, M. T., Walsh, F. S., Trapp, B. D., Pizzey, J. A., and Tennekoon, G. I. (1990). The role of myelin P0 protein as a homophilic adhesion molecule. Nature 344, 871–872. Fraher, J. P. (1973). A quantitative study of anterior root Wbres during early myelination. II. Longitudinal variation in sheath thickness and axon circumference. J. Anat. 115, 421–444. Friede, R. L., and Bischhausen, R. (1980). The precise geometry of large internodes. J. Neurol. Sci. 48, 367–381. Friede, R. L., and Bischhausen, R. (1982). How are sheath dimensions aVected by axonal caliber and internodal length? Brain Res. 235, 335–350. Garbern, J. Y., Yool, D. A., Moore, G. J., Wilds, I. B., Faulk, M. W., Klugmann, M., Nave, K. A., Sistermans, E. A., van der Knaap, M. S., Bird, T. D., Shy, M. E., Kamholz, J. A., and GriYths, I. R. (2002). Patients lacking the major CNS myelin protein, proteolipid protein 1, develop length-dependent axonal degeneration in the absence of demyelination and inXammation. Brain 125, 551–561. Geiger, B., Rosen, D., and Berke, G. (1982). Spatial relationships of microtubule-organizing centers and the contact area of cytotoxic T lymphocytes and target cells. J. Cell Biol. 95, 137–143. Geren, B. B. (1954). The formation from the Schwann cell surface of myelin in the peripheral nerves of chick embryos. Exp. Cell Res. 7, 558–562. Ghabriel, M. N., and Allt, G. (1979). The role of Schmidt-Lanterman incisures in Wallerian degeneration. II. An electron microscopic study. Acta Neuropathol.(Berl) 48, 95–103. Ghabriel, M. N., and Allt, G. (1981). Incisures of Schmidt-Lanterman. Prog. Neurobiol. 17, 25–58. Giese, K. P., Martini, R., Lemke, G., Soriano, P., and Schachner, M. (1992). Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71, 565–576. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., Cottrell, D. F., Tait, S., Garry, E. M., Wallace, V. C., Ure, J., GriYths, I. R., Smith, A., and Brophy, P. J. (2000). Peripheral demyelination and neuropathic pain behavior in periaxin-deWcient mice. Neuron 26, 523–531. Gow, A., Southwood, C. M., Li, J. S., Pariali, M., Riordan, G. P., Brodie, S. E., Danias, J., Bronstein, J. M., Kachar, B., and Lazzarini, R. A. (1999). CNS myelin and Sertoli cell tight junction strands are absent in Osp/ claudin-11 null mice. Cell 99, 649–659. GreenWeld, S., BrostoV, S., Eylar, H. H., and Morell, P. (1973). Protein composition of myelin of the peripheral nervous system. J. Neurochem. 20, 1207–1216. GriYn, J. W., Kidd, G., and Trapp, B. D. (1993). Interactions between axons and Schwann cells. In ‘‘Peripheral Neuropathy’’ (P. J. Dyck and P. K. Thomas, eds.) pp. 317–330. W. B. Saunders Company, Philadelphia.

SUMMARY AND FUTURE CONSIDERATIONS

GriYths, I., Klugmann, M., Anderson, T., Yool, D., Thomson, C., Schwab, M. H., Schneider, A., Zimmermann, F., McCulloch, M., Nadon, N., and Nave, K.-A. (1998). Axonal swellings and degeneration in mice lacking the major proteolipid of myelin. Science 280, 1610–1613. Guilbot, A., Williams, A., Ravise, N., Verny, C., Brice, A., Sherman, D. L., Brophy, P. J., LeGuern, E., Delague, V., Bareil, C., Megarbane, A., and Claustres, M. (2001). A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10, 415–421. Hall, S. M., and Williams, P. L. (1970). Studies of the ‘‘incisures’’ of Schmidt and Lanterman. J. Cell Sci. 6, 767–791. Hartman, B. K., Agrawal, H. C., Agrawal, D., and Kalmback, S. (1982). Development and maturation of central nervous system myelin: Comparison of immunocytochemical localization of proteolipid protein and basic protein in myelin and oligodendrocytes. Proc. Natl. Acad. Sci. USA 79, 4217–4220. Hildebrand, C., and Waxman, S. G. (1984). Postnatal diVerentiation of rat optic nerve Wbers: Electron microscopic observations on the development of nodes of Ranvier and axoglial relations. J. Comp Neurol. 224, 25–37. Hsieh, S.-T., Kidd, G. J., Crawford, T. O., Xu, Z., Trapp, B. D., Cleveland, D. W., and GriYn, J. W. (1994). Regional modulation of neuroWlament organization by myelination in normal axons. J. Neurosci. 14, 6392–6401. Kadlubowski, M., Hughes, R. A. C., and Gregsen, N. A. (1980). Experimental allergic neuritis in the Lewis rat: Characterization of the activity of peripheral myelin and its major basic protein, P2. Brain Res. 184, 439–454. Kirschner, D. A., and Hollingshead, C. J. (1980). Processing for electron microscopy alters membrane structure and packing in myelin. J. Ultrastruct. Res. 73, 211–232. Kirschner, D. A., and Sidman, R. L. (1976). X-ray defraction study of myelin structure in immature and mutant mice. Biochim. Biophys. Acta 448, 73–87. Kordeli, E., Davis, J. D., Trapp, B. D., and Bennett, V. (1990). An isoform of ankyrin is localized at nodes of Ranvier in myelinated axons of central and peripheral nerves. J. Cell Biol. 110, 1341–1352. Kordeli, E., Lambert, S., and Bennett, V. (1995). AnkyrinG. A new ankyrin gene with neural-speciWc isoforms localized at the axonal initial segment and node of Ranvier. J. Biol. Chem. 270, 2352–2359. Lees, M. B., and BrostoV, S. W. (1984). Proteins of myelin. In ‘‘Myelin’’ (P. Morrell, ed.), pp. 197–224. Plenum Press, New York. Lemke, G., and Axel, R. (1985). Isolation and sequence of a cDNA encoding the major structural protein of peripheral myelin. Cell 40, 501–508. Li, C., Tropak, M. B., Gerial, R., ClapoV, S., Abramow-Newerly, W., Trapp, B., Peterson, A., and Roder, J. (1994). Myelination in the absence of myelin-associated glycoprotein. Nature 369, 747–750. Li, M., Shibata, A., Li, C., Braun, P. E., McKerracher, L., Roder, J., Kater, S. B., and David, S. (1996). Myelinassociated glycoprotein inhibits neurite/axon growth and causes growth cone collapse. J. Neurosci. Res. 46, 404–414. Li, X., Lynn, B. D., Olson, C., Meier, C., Davidson, K. G., Yasumura, T., Rash, J. E., and Nagy, J. I. (2002). Connexin29 expression, immunocytochemistry and freeze-fracture replica immunogold labelling (FRIL) in sciatic nerve. Eur. J Neurosci. 16, 795–806. Mata, M., Kupina, N., and Fink, D. J. (1992). Phosphorylation-dependent neuroWlament epitopes are reduced at the node of Ranvier. J. Neurocytol. 21, 199–210. McDonald, W. I., and Ohlrich, G. D. (1971). Quantitative anatomical measurements on single isolated Wbres from the cat spinal cord. J. Anat. 110, 191–202. Melendez-Vasquez, C. V., Rios, J. C., Zanazzi, G., Lambert, S., Bretscher, A., and Salzer, J. L. (2001). Nodes of Ranvier form in association with ezrin-radixin-moesin (ERM)-positive Schwann cell processes. Proc. Natl. Acad. Sci. U. S. A 98, 1235–1240. Menegoz, M., Gaspar, P., Le Bert, M., Galvez, T., Burgaya, F., Palfrey, C., Ezan, P., Arnos, F., and Girault, J. A. (1997). Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron 19, 319–331. Mikol, D. D., Hong, H. L., Cheng, H. L., and Feldman, E. L. (1999). Caveolin-1 expression in Schwann cells. Glia 27, 39–52. Montag, D., Giese, K. P., Bartsch, U., Martini, R., Land, Y., Blu¨thmann, H., Karthigasan, J., Kirschner, D. A., Wintergerst, E. S., Nave, K.-A., Zielasek, J., Toyka, K. V., Lipp, H.-P., and Schachner, M. (1994). Mice deWcient for the myelin-associated glycoprotein show subtle abnormalities in myelin. Neuron 13, 229–246. Morita, K., Sasaki, H., Fujimoto, K., Furuse, M., and Tsukita, S. (1999). Claudin-11/OSP-based tight junctions of myelin sheaths in brain and Sertoli cells in testis. J Cell Biol. 145, 579–588. Mugnaini, E., Olsen, K. K., Schnapp, B., and Friedrich, V. L., Jr. (1977). Distribution of Schwann cell cytoplasm and plasmalemmal vesicles (caveolae) in peripheral myelin sheaths. An electron microscopic study with thin sections and freeze-fracturing. J. Neurocytol. 6, 647–668. Mukhopadhyay, G., Doherty, P., Walsh, F. S., Crocker, P. R., and Filbin, M. T. (1994). A novel role for myelinassociated glycoprotein as an inhibitor of axonal regeneration. Neuron 13, 757–767. Norton, W. T., and Poduslo, S. E. (1973). Myelination in rat brain: Method of myelin isolation. J. Neurochem. 21, 749–757. Pant, H. C. (1988). Dephosphorylation of neuroWlament proteins enhances their susceptibility to degradation by calpain. Biochem. J. 256, 665–668. Pedraza, L., Huang, J. K., and Colman, D. R. (2001). Organizing principles of the axoglial apparatus. Neuron 30, 335–344. Peters, A. (1961). A radial component of central myelin sheaths. J. Biophys. Biochem. Cytol. 11, 733–735.

25

26

1. STRUCTURE OF THE MYELINATED AXON

Peters, A., Palay, S. L., and Webster, H. d. (1991). ‘‘The Fine Structure of the Nervous System: Neurons and their Supporting Cells.’’ Oxford University Press, New York. Poliak, S., Gollan, L., Martinez, R., Custer, A., Einheber, S., Salzer, J. L., Trimmer, J. S., Shrager, P., and Peles, E. (1999). Caspr2, a new member of the neurexin superfamily, is localized at the juxtaparanodes of myelinated axons and associates with Kþ channels. Neuron 24, 1037–1047. Privat, A., Jacque, C., Bourre, J. M., Duponey, P., and Baumann, N. (1979). Absence of the major dense line in myelin of the mutant mouse ‘‘Shiverer’’. Neurosci. Lett. 12, 107–112. Ramon y Cajal, S. (1928). ‘‘Degeneration and Regeneration of the Nervous System’’ (R. M. May, trans.). Oxford University Press, London. Rasband, M. N., Trimmer, J. S., Schwarz, T. L., Levinson, S. R., Ellisman, M. H., Schachner, M., and Shrager, P. (1998). Potassium channel distribution, clustering, and function in remyelinating rat axons. J. Neurosci. 18, 36–47. Rios, J. C., Melendez-Vasquez, C. V., Einheber, S., Lustig, M., Grumet, M., Hemperly, J., Peles, E., and Salzer, J. L. (2000). Contactin-associated protein (Caspr) and contactin form a complex that is targeted to the paranodal junctions during myelination. J. Neurosci. 20, 8354–8364. Robertson, J. D. (1958). The ulstrastructure of Schmidt-Lantermann clefts and related shearing defects of the myelin sheath. Journal of Biophysical and Biochemical Cytology 4, 39–44. Rosenbluth, J. (1976). Intramembranous particle distribution at the node of Ranvier and adjacent axolemma in myelinated axons of the frog brain. J. Neurocytol. 5, 731–745. Rosenbluth, J. (1999). A brief history of myelinated nerve Wbers: One hundred and Wfty years of controversy. J Neurocytol. 28, 251–262. Sanchez, I., Hassinger, L., Paskevich, P. A., Shine, H. D., and Nixon, R. A. (1996). Oligodendroglia regulate the regional expansion of axon caliber and local accumulation of neuroWlaments during development independently of myelin formation. J. Neurosci. 16, 5095–5105. Scherer, S. (1999). Axonal pathology in demyelinating diseases. Ann. Neurol. 45, 6–7. Scherer, S. S., Deschenes, S. M., Xu, Y. T., Grinspan, J. B., Fischbeck, K. H., and Paul, D. L. (1995a). Connexin32 is a myelin-related protein in the PNS and CNS. J. Neurosci. 15, 8281–8294. Scherer, S. S., Xu, T., Crino, P., Arroyo, E. J., and Gutmann, D. H. (2001). Ezrin, radixin, and moesin are components of Schwann cell microvilli. J. Neurosci. Res. 65, 150–164. Scherer, S. S., Xu, Y. T., Bannerman, P. G., Sherman, D. L., and Brophy, P. J. (1995b). Periaxin expression in myelinating Schwann cells: Modulation by axon-glial interactions and polarized localization during development. Development 121, 4265–4273. Scherer, S. S., Xu, Y. T., Nelles, E., Fischbeck, K., Willecke, K., and Bone, L. J. (1998). Connexin32-null mice develop demyelinating peripheral neuropathy. Glia 24, 8–20. Shapiro, L., Doyle, J. P., Hensley, P., Colman, D. R., and Hendrickson, W. A. (1996). Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17, 435–449. Sheikh, K. A., Sun, J., Liu, Y., Kawai, H., Crawford, T. O., Proia, R. L., GriYn, J. W., and Schnaar, R. L. (1999). Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proc Natl Acad Sci U. S. A 96, 7532–7537. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J. (2001). SpeciWc disruption of a Schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron 30, 677–687. Snipes, G. J., Suter, U., Welcher, A. A., and Shooter, E. M. (1992). Characterization of a novel peripheral nervous system myelin protein (PMP-22/SR13). J. Cell Biol. 117, 225–238. Sprinkle, T. J., Zaruba, M. E., and McKhann, G. M. (1978). Activity of 2’, 3’-cyclic nucleotide 3’-phosphodiesterase in regions of rat brain during development: Quantitative relationship to myelin basic protein. J. Neurochem. 30, 309–314. Sternberger, N. H., Itoyama, Y., Kies, M. K., and Webster, H. d. (1978). Immunocytochemical method to identify basic protein in myelin-forming oligodendrocytes of newborn rat CNS. J. Neurocytol. 7, 251–263. StoVel, W., Boison, D., and Bussow, H. (1997). Functional analysis in vivo of the double mutant mouse deWcient in both proteolipid protein (PLP) and myelin basic protein (MBP) in the central nervous system. Cell Tissue Res. 289, 195–206. Tabira, T., Cullen, M. J., Reier, P. J., and Webster, H. d. (1978). An experimental analysis of interlamellar tight junctions in amphibian and mammalian CNS myelin. J. Neurocytol. 7, 489–503. Tait, S., Gunn-Moore, F., Collinson, J. M., Huang, J., Lubetzki, C., Pedraza, L., Sherman, D. L., Colman, D. R., and Brophy, P. J. (2000). An oligodendroctye cell adhesion molecule at the site of assembly of the paranodal axo-glial junction. J. Cell Biol. 150, 657–666. Trapp, B. D., Andrews, S. B., Cootauco, C., and Quarles, R. H. (1989a). The myelin-associated glycoprotein is enriched in multivesicular bodies and periaxonal membranes of actively myelinating oligodendrocytes. J. Cell Biol. 109, 2417–2426. Trapp, B. D., Andrews, S. B., Wong, A., O’Connell, M., and GriYn, J. W. (1989b). Co-localization of the myelinassociated glycoprotein and the microWlament components f-actin and spectrin in Schwann cells of myelinated Wbers. J. Neurocytol. 18, 47–60. Trapp, B. D., Bernier, L., Andrews, S. B., and Colman, D. R. (1988). Cellular and subcellular distribution of 2’, 3’ cyclic nucleotide 3’ phosphodiesterase and its mRNA in the rat nervous system. J. Neurochem. 51, 859–868. Trapp, B. D., Itoyama, Y., MacIntosh, T. D., and Quarles, R. H. (1983). P2 protein in oligodendrocytes and myelin of the rabbit central nervous system. J. Neurochem. 40, 47–54.

SUMMARY AND FUTURE CONSIDERATIONS

Trapp, B. D., and Kidd, G. J. (2000). Axo-glial septate junctions. The maestro of nodal formation and myelination? J Cell Biol. 150, F97-F100. Trapp, B. D., McIntyre, L. J., Quarles, R. H., Sternberger, N. H., and Webster, H. d. (1979). Immunocytochemical localization of rat peripheral nervous system myelin proteins: P2 protein is not a component of all peripheral nervous system myelin sheaths. Proc. Natl. Acad. Sci. USA 76, 3552–3556. Trapp, B. D., and Quarles, R. H. (1982). Presence of the myelin-associated glycoprotein correlates with alterations in the periodicity of peripheral myelin. J. Cell Biol. 92, 877–882. Trapp, B. D., and Quarles, R. H. (1984). Immunocytochemical localization of the myelin-associated glycoprotein: Fact or artifact? J. Neuroimmunol. 6, 231–249. Trapp, B. D., RansohoV, R. M., Fisher, E., and Rudick, R. A. (1999). Neurodegeneration in multiple sclerosis: Relationship to neurological disability. The Neuroscientist 5, 48–57. Uschkureit, T., Sporkel, O., Stracke, J., Bussow, H., and StoVel, W. (2000). Early onset of axonal degeneration in double (plp
/
mag
/
) and hypomyelinosis in triple (plp
/
mbp
/
mag
/
) mutant mice. J. Neurosci. 20, 5225–5233. Uyemura, K., Yoshimura, K., Suzuki, M., and Kitamura, K. (1984). Lipid binding activities of the P2 protein in peripheral nerve myelin. Neurochem. Res. 9, 1509–1514. Waxman, S. G., and Ritchie, J. M. (1985). Organization of ion channels in the myelinated nerve Wber. Science 228, 1502–1506. Wiley, C. A., and Ellisman, M. H. (1980). Rows of dimeric-particles within the axolemma and juxtaposed particles within glia, incorporated into a new model for the paranodal glial-axonal junction at the node of Ranvier. J. Cell Biol. 84, 261–280. Windebank, A. J., Word, P., Bunge, R. P., and Dyck, P. J. (1985). Myelination determines the caliber of dorsal root ganglion neurons in culture. J. Neurosci. 6, 1563–1567. Yin, X., Crawford, T. O., GriYn, J. W., Tu, P.-H., Lee, V. M. Y., Li, C., Roder, J., and Trapp, B. D. (1998). Myelin-associated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J. Neurosci. 18, 1953–1962. Yin, X., Peterson, J., Gravel, M., Braun, P. E., and Trapp, B. D. (1997). CNP overexpression induces aberrant oligodendrocyte membranes and inhibits MBP accumulation and myelin compaction. J. Neurosci. Res. 50, 238–247. Yin, X., and Trapp, B. D. (1997). Myelin-associated glycoprotein. Ligand and/or receptor. In ‘‘Cell Biology and Pathology of Myelin: Evolving Biological Concepts and Therapeutic Approaches’’ (Juurlink, B. H. J., Doucette, J. R., Devon, R. M., Schreyer, D. J., and Verge, V. K. M., eds.), pp. 27–36. Plenum Press, New York. Zeller, N. K., Hunkeler, M. J., Campagnoni, A. T., Sprague, J., and Lazzarini, R. A. (1984). Characterization of mouse myelin basic protein messenger RNAs with a myelin basic protein cDNA clone. Proc. Natl. Acad. Sci. USA 81, 18–22. Zoidl, G., Blass-Kampmann, S., D’Urso, D., Schmalenbach, C., and Mu¨ller, H. W. (1995). Retroviral-mediated gene transfer of the peripheral myelin protein PMP22 in Schwann cells: Modulation of cell growth. EMBO J. 14, 1122–1128.

27

C H A P T E R

2 Cell Biology of Myelin Assembly Bruce D. Trapp, Steven E. Pfeiffer, Mihaela Anitei, and Grahame J. Kidd

INTRODUCTION The process of myelination, carried out by oligodendrocytes and Schwann cells in the central and peripheral nervous systems, respectively, is one of the most signiWcant milestones in vertebrate evolution. By permitting rapid nerve transmission through thin nerve Wbers, it promoted the development of a compact nervous system capable of integrating the sensory, motor, and cognitive functions of humans. The process of myelination also represents one of the best examples of cell-cell cooperativity. The functional diVerentiation of myelin-forming cells is governed by, and dependent on, physical contact with another cell, speciWcally with long, specialized processes (axons) of the neuron. Neither axons nor myelin-forming cells functionally diVerentiate to completion without the other. Thus, although myelin membrane production is initiated independently of neuronal contact (oligodendrocytes produce signiWcant amounts of myelin-like membrane in neuron-free culture), in vivo the quantity and stability of myelin is strongly enhanced (Singh and PfeiVer, 1985), presumably as a result of the extensive contact with axons. As oligodendrocytes (OLs) enter terminal diVerentiation and contact neurons, they begin to produce myelin membrane on a remarkable scale (>104 mm2 myelin membrane surface area/cell/day; (PfeiVer et al., 1993). The cell biological hallmark of myelination, therefore, is membrane biosynthesis to produce a unique extension of the oligodendrocyte and Schwann cell plasma membrane that spirally wraps around axons. As for epithelial cells, this membrane is not a simple extension of the myelin-forming cell plasma membrane; it consists of multiple domains that are molecularly, morphologically, and functionally distinct. In this chapter, we discuss some of the mechanisms by which myelin-forming cells expand and polarize their surface membranes into multiple domains. Although relatively little is yet known about the molecular interactions responsible for sorting, targeting, and stabilizing proteins and lipids to these domains, the molecular and ultrastructural characterization of myelin membrane domains is substantial, providing a baseline ripe for elucidating the mechanisms and molecular interactions that regulate myelin assembly.

VESICULAR TRANSPORT AND MEMBRANE BIOGENESIS IN POLARIZED CELLS Functional polarity of cell surface membranes is a fundamental property of all eukaryotic cells. Integral membrane proteins are synthesized and processed in the rough endoplasmic reticulum (RER) and Golgi apparatus and transported intracellularly in membrane vesicles that ultimately dock and fuse with the plasma membrane (‘‘exocytosis’’) (Mellman and Warren, 2000; Mostov et al., 2000). This results in addition of vesicle membranes to the

Myelin Biology and Disorders, Volume 1

29

Copyright 2004, Elsevier (USA). All rights reserved.

30

2. CELL BIOLOGY OF MYELIN ASSEMBLY

FIGURE 2.1 Schematic of protein targeting pathways in a simple polarized epithelial cell (A). Proteins exit the trans-Golgi apparatus and are directly inserted into either apical or basolateral surfaces. Some proteins use an indirect pathway by being delivered to the basolateral surface and then transcytosed to the apical surface. Myelin proteins have targeting sequences that are recognized by simple polarized epithelial cells (B). When transfected into MDCK cells, the S-isoform of MAG is targeted to the apical surface (B, green), while transfected P0 protein is enriched on the basolateral surface (B, red).

plasma membrane and the exocytosis of material inside the vesicle. SpeciWc components mediate each stage: coat proteins facilitate vesicle budding (e.g., COPs), small GTPbinding proteins (e.g., Rab proteins) mediate vesicle transport and fusion at speciWc steps, tethering complexes initiate targeting (e.g., exocyst complex), and soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins (e.g., synaptobrevins/VAMPs) mediate vesicle docking/fusion. An analysis of whole genome sequences predicts 35 SNAREs, 60 Rabs, and 53 coat complex subunits in the human (Bock et al., 2001). While controversy remains, certain docking and fusion components of the SNARE complex may concentrate in lipid rafts (Chamberlain et al., 2001; Ikonen, 2001; Lang et al., 2001; Taylor et al., 2002; van Meer, 1989). The mechanisms by which these pathways operate, in particular how speciWcity is achieved, is a forefront of molecular cell biology and of critical importance to our understanding of myelin biogenesis. The prototype and best-studied polarized cells are the monolayers of cuboidal or columnar-shaped epithelial cells that line the surface and cavities of all organs. The plasma membranes of these epithelia are divided into apical and basolateral domains separated by tight junctions (Mellman and Warren, 2000; Mostov et al., 1992). The apical domain faces the ‘‘outside world’’ or a lumen, while the basolateral domain is connected to adjoining cells and underlying connective tissue. These membrane domains have specialized functions, which are reXected by diVerent protein and lipid constituents. Epithelial cells use two basic mechanisms to target molecules to the correct surface (Fig. 2.1): a direct pathway (Weimbs et al., 1997) and an indirect pathway (Mostov et al., 2000). In both pathways, new membrane proteins are translated in the RER and post-translationally modiWed in the Golgi apparatus.

VESICULAR TRANSPORT AND MEMBRANE BIOGENESIS

In the ‘‘direct’’ pathway, apical and basolateral proteins are packaged into separate vesicles in the trans-Golgi network (TGN) and directly delivered to the appropriate membrane domain. The ‘‘indirect’’ pathway sends molecules from the TGN to the basolateral surface where they are removed by endocytosis and then translocated to the apical surface by transcytosis. Endocytosed proteins can also be recycled back to their surface of origin or degraded in the endosomal/lysosomal compartment. Utilization of the direct or indirect pathway varies from cell type to cell type and is governed by structural properties or speciWc amino acid sequences of the surface membrane protein. As discussed later, it appears that the ‘‘direct’’ pathway may predominate in myelinating cells. Simple epithelial cell lines, such as Mardin-Darby Canine Kidney (MDCK) cells, provide excellent in vitro systems for investigating mechanisms of surface membrane polarization and protein targeting. MDCK cells share molecular sorting and targeting machinery with other polarized epithelial cells. Most membrane proteins contain apical or basolateral targeting information that operates in MDCK cells. Much of the current knowledge about protein sorting and targeting signals has been generated by expressing mutated, truncated, and chimeric proteins in MDCK cells. Basolateral sorting signals usually reside in the cytoplasmic domain of transmembrane proteins. Many of these signals are encoded in short (2 to 10 residue) sequences, while others involve tertiary structures such as a type I b-turn (Matter and Mellman, 1994). The best-described apical targeting signal is the glycosylphosphatidylinositol (GPI) anchor (Simons and Ikonen, 1997). GPIlinked proteins segregate with glycosphingolipids and cholesterol into membrane rafts (Ikonen, 2001; Lisanti et al., 1990; van Meer, 1989). Transmembrane domain properties also promote raft association of some apically targeted proteins (Barman and Nayak, 2000). Other apical membrane sorting signals have been identiWed, including transmembrane or cytoplasmic domain amino acids and N- and O-linked oligosaccharides (Benting et al., 1999; Rodriguez-Boulan and Gonzalez, 1999) that may function by binding to lectins or by aVecting protein tertiary structure. Translocation of TGN-derived membrane vesicles to speciWc surface domains of large cells such as neurons occurs in an energy-dependent manner utilizing motors, including members of the kinesin, dynein, and myosin superfamilies that bind to both vesicles and microtubules (MTs) or microWlaments (Hirokawa, 1998; Kamal and Goldstein, 2000). MTs are polar filaments that have a plus and minus end, and organize biosynthetic organelles and transport cargo-laden vesicles. MT motors preferentially transport toward either the plus (kinesin) or minus (dynein) ends of MTs, while myosin mediates short-range transport on actin microWlaments. The distribution of MT plus and minus ends within particular cytoplasmic domains of cells is therefore important for protein targeting and membrane and organelle transport (Cole and Lippincott-Schwartz, 1995). A well-described example is in the neuron, for which all axonal MTs point their plus ends toward the process termini (Black and Baas, 1989), while MTs in proximal dendrites have a mixed polarity. Once a TGN-derived vesicle reaches its destination, it incorporates into the membrane by docking and fusing with the target membrane (Rothman and Warren, 1994; Scales et al., 2000). Correct binding of a v-SNARE on the vesicle membrane with a cognate t-SNARE on the target membrane mediates and directs fusion of the membrane bilayers. This mechanism may also inhibit mistargeting of vesicles to inappropriate surface membranes. t-SNARE isoforms of the syntaxin family are enriched at speciWc membrane domains of a variety of polarized epithelial cell lines. In MDCK cells, for example, syntaxin-3 is localized to the apical surface, syntaxin-4 to the basolateral surface and syntaxin-2 to both surfaces (Low et al., 1996) (Fig. 2.1). Oligodendrocytes in culture express several v-SNAREs (VAMP-2, 4,7), t-SNAREs (syntaxin-2, 3, 4), and other members of the complex, such as SNAPs (25, 23, 29). The ‘‘basolateral’’ cognate pair VAMP-2/syntaxin-4 is up-regulated in mature OLs (Madison et al., 1999), and the ‘‘apical’’ VAMP-7/syntaxin-3 is also expressed. In cultured OLs, syntaxin-4 is localized to myelin-like membranes, syntaxin-3 is expressed in minor quantities in the cell body, while syntaxin-2 is ubiquitous (de Vries and Hoekstra, 2000). While SNARE-SNARE interactions can mediate both direct and indirect protein targeting in polarized epithelial cells, and are crucial for vesicle docking and fusion,

31

32

2. CELL BIOLOGY OF MYELIN ASSEMBLY

other protein interactions also facilitate correct pairing. Tethering complexes share a conserved N-terminal coiled-coil domain that might mediate interactions with Rab and SNARE proteins or with other docking/fusion components. The exocyst (Sec6/8) was one of the Wrst tethering complexes to be described. The exocyst concentrates at sites of membrane addition in yeast (TerBush and Novick, 1995), neurons (Kee et al., 1997), and is involved in basolateral targeting in epithelial cells (GrindstaV et al., 1998). In the active zones of presynaptic nerve terminals, the cortical active zone (CAZ) is composed of a family of multidomain proteins that play fundamental roles in deWning neurotransmitter release sites. These include CASK, Rim (Rab3a interacting molecule), the structurally related, large (450/530 kDa) proteins Bassoon, Piccollo/Aczonin, and Munc-13 and -18 (Garner et al., 2000). Several of these molecules have been identiWed in oligodendrocytes, including Munc-13, exocyst, Rim, and Bassoon (Madison et al., 1999) (Anitei and PfeiVer, unpublished). CASK is of interest in oligodendrocytes by virtue of its interaction with syndecan-2 (Hsueh et al., 1998), a developmentally regulated HSPG in oligodendrocytes (Bansal et al., 1996), and with Caspr-2, a component of the neuro-glial junction (Spiegel et al., 2002). Rim is of interest for myelin biogenesis because of its association with Rab3a, which is strongly up-regulated as oligodendrocytes enter terminal diVerentiation (Madison et al., 1996); Piccollo also interacts with a Rab3a eVector, PRA-1 (Fenster, et al., 2000). Do myelin proteins have targeting sequences that operate in polarized epithelial cell lines? The simple answer is yes. It is too early, however, to predict the value of this approach for elucidating molecular speciWcity of myelin protein sorting and targeting mechanisms. When transfected into MDCK cells, the large isoform of the myelin-associated glycoprotein (L-MAG) was targeted to both apical and basolateral domains, while the small isoform, S-MAG, was targeted to apical domains (Minuk and Braun, 1996). The myelin oligodendrocyte glycoprotein (MOG) was targeted to basolateral domains (KroepX and Gardinier, 2001) and a potential basolateral sorting signal was identiWed at its C terminus. PLP, a compact myelin component, was transported to the apical MDCK domain (KroepX and Gardinier, 2001). The PLP N-terminal (13 amino acids) is suVicient to transport LacZ to the myelin sheath (Wight et al., 1993). Protein-zero (P0) cDNAs have been transfected into nonadherent cell lines to study adhesive properties (D’Urso et al., 1990; Filbin et al., 1990). Although targeting mechanisms were not a focus of these studies, it is interesting to note that P0 transfection into HeLa cells induced surface membrane polarization, which was mediated by homophilic trans P0 adhesion. MDCK cell lines permanently transfected with P0, S-MAG and L-MAG have been recently established (Kidd et al., 2002). P0 was targeted to the basolateral surface (Fig. 2.1B). Growing P0 transfected and nontransfected MDCK cells together did not alter the basolateral distribution of P0, indicating that intracellular site speciWc targeting and retention are responsible for P0 steady-state distribution. This observation, plus electron microscopic analysis, indicates that P0 trans homophilic adhesions do not contribute to nor stabilize P0 in the lateral membranes of adjoining MDCK cells (Kidd and Trapp, unpublished results). In addition, deletion of the cytoplasmic domain results in apical targeting of P0 in MDCK cells. Overall, these observations indicate that targeting signals can be identiWed for myelin proteins in MDCK cells. However, once identiWed, it will be necessary to develop strategies to determine whether such signals actually operate during myelination. Basolateral targeting of both P0 and MOG, which are enriched in quite diVerent domains of the myelin internodes, indicates that the surfaces of simple polarized epithelial cell lines cannot be directly compared to the membrane surfaces of myelin-forming cells.

MYELIN ASSEMBLY The following reviews key steps in myelin assembly. For each event under consideration, processes that are shared by oligodendrocytes and Schwann cells, such as myelin basic protein (MBP) mRNA translocation and pathologies caused by myelin protein misfolding, are described for the cell that has generated the most data. Processes unique to either

SCHWANN CELL DIFFERENTIATION

FIGURE 2.2 Schwann cell diVerentiation. Schwann cells originate in the neural crest (A) and migrate along axons to the periphery. Before myelination, each Schwann cell surrounds multiple small axons (B). Myelination is preceded by Schwann cell division (C). One daughter cell isolates a single axon (D) and myelinates it (D’), while the other surrounds developing axons (E) and continues to divide (C) until all appropriate axons are myelinated. Many small diameter axons are not myelinated and are ensheathed by nonmyelinating Schwann cells (F).

Schwann cells or oligodendrocytes are highlighted. In addition, sites of synthesis of myelin proteins and their distribution in the myelin internode are described. With the exception of MBP mRNA translocation, little is known about the molecules or the molecular interactions that are responsible for sorting and targeting proteins to various regions of the myelin internode. Therefore, we can only speculate about these processes based on precedents established in other cells.

SCHWANN CELL DIFFERENTIATION Schwann cells originate in the neural crest and expand as they migrate to peripheral locations in association with axons. Once they reach their Wnal destination, each Schwann cell surrounds several small diameter axons in a polyaxonal pocket (Fig. 2.2). Axons that are destined to be myelinated undergo a maturation characterized by an increase in diameter. As axons mature, the Schwann cell segregates a single axon from the polyaxonal pocket. The Schwann cell then divides: one daughter cell maintains the polyaxonal pocket; the other establishes a one-to-one relationship with the segregated axon and myelinates it (Webster, 1971). Schwann cell myelination, therefore, is regulated by axons and its initiation is closely related to cell division. It is unknown if Schwann cell division and initiation of the myelinating phenotype are regulated by diVerent molecular signals. The processes of axonal segregation, Schwann cell division and myelination continue until all appropriate axons are myelinated. Small diameter sensory axons (C-Wbers) remain ensheathed by Schwann cells but are not myelinated. Schwann cells in peripheral nerves have two major phenotypes: those that ensheath multiple axons (unmyelinated Wbers) and those that myelinate single axons. All Schwann cells have the potential to form myelin. They only do so upon induction by appropriate peripheral axons. While the molecular nature of this axonal signal is poorly understood, the following observations indicate that it operates at the level of gene transcription. Schwann cells of unmyelinated Wbers do not express detectable levels of myelin protein

33

34

2. CELL BIOLOGY OF MYELIN ASSEMBLY

mRNAs (Martini and Schachner, 1986; Trapp et al., 1981; Winter et al., 1982). P0 and MBP accumulation and the levels of their respective mRNAs parallel the rate of myelin formation during development. Removal of axons by transection (Trapp et al., 1988b) or cellular dissociation (Brockes et al., 1980; Mirsky et al., 1980) results in rapid decreases in myelin protein gene expression. Following nerve transection, sciatic nerve axons regenerate among distal stump Schwann cells, these Schwann cells re-express myelin protein genes and myelinate, whether they come from previously myelinated or unmyelinated nerves (Aguayo et al., 1976a, 1976b; Politis et al., 1982).

SCHWANN CELL MYELINATION Initially, myelinating Schwann cells polarize their surface membranes into two domains. The abaxonal plasma membrane is exposed to, and interacts with, the external environment or endoneurial Xuid. As part of its maturation, it directs the production of the basal lamina, which is a prerequisite for myelination (Bunge et al., 1986; Eldridge et al., 1987). The second domain, the adaxonal or periaxonal membrane, is in direct contact with the axon and is enriched in myelin-associated glycoprotein (MAG) (Trapp et al., 1984, 1989a). The periaxonal membrane radially and longitudinally ensheaths the axon and plays a role in directing axonal Naþ channels to nodes (Pedraza et al., 2001). Spiral growth of the myelin internode is initiated by the expansion of mesaxon membranes. This membrane connects the periaxonal and abaxonal membranes and partially forms two potential ‘‘leading edges’’ for spiral wrapping at the periaxonal and abaxonal surface. These develop into the inner and outer mesaxons of more mature internodes (Peters et al., 1991). It is still not established if spiral expansion occurs by rotation of one or both ‘‘leading edges.’’ In either event, mesaxon membranes expand by rotating upon themselves or upon the axon. The membranes of the mesaxon spiral also contain MAG and are ultrastructurally characterized by a 12 to 14 nm gap between their extracellular leaXets (Trapp and Quarles, 1982). Schwann cell cytoplasm separates cytoplasmic leaXets of mesaxon membranes but contains few organelles. Once several spiral turns are formed, the cytoplasm between mesaxon membranes is eliminated, and the major dense line of compact myelin is formed. Simultaneously, the 12 to 14 nm spacing between the extracellular leaXets decreases to the 2 nm spacing of compact myelin. P0 protein, the major structural protein of compact PNS myelin, has not been detected in early mesaxon wraps during normal myelination. It is possible, however, that P0 is present in the mesaxon membranes at levels undetectable by current methods or that the transition to compact myelin occurs so fast that intermediate stages are not detectable. Conversion of mesaxon membranes to compact myelin during normal myelination involves two molecular events: the removal of MAG and the addition of P0 protein (Trapp, 1988) (Fig. 2.3). While the molecular mechanisms responsible for the segregation of MAG-enriched and P0-enriched membranes are unknown, we discuss three possible mechanisms. One hypothesis proposes that MAG is removed from mesaxon membranes by endocytosis. Although this hypothesis is supported by the presence of clathrin coated pits on mesaxon membranes during early stages of myelination (Trapp et al., 1989a), MAG does not accumulate in endosomes during early stages of Schwann cell myelination. Therefore, endocytosis does not appear to play a role in the removal of S-MAG from Schwann cell mesaxon membranes. The situation may be diVerent in the CNS. Two MAG transcripts, L-MAG and S-MAG, are produced by alternate splicing of the cytoplasmically deposed C-terminus of MAG (Arquint et al., 1987; Lai et al., 1987; Salzer et al., 1987), with Schwann cells expressing the smaller isoform (S-MAG). In contrast, oligodendrocytes diVerentially express both isoforms during development (Quarles et al., 1973; Tropak et al., 1988); L-MAG dominates during early stages of CNS myelination, whereas L-MAG and S-MAG are found at similar concentrations in the mature CNS. In contrast to S-MAG in Schwann cells, L-MAG is enriched in endosomes during early stages of CNS myelination and thus could play a role in membrane remodeling (Bo¨ et al., 1995; Trapp et al., 1989a).

SCHWANN CELL MYELINATION

FIGURE 2.3 Schematic representation of the orientation and membrane distribution of P0 protein and MAG during conversion of mesaxon membranes to compact myelin. MAG and P0 are shown only in the inner membrane bilayers. Initial mesaxon wraps contain MAG but little P0 protein (A). The bulk of MAGs extracellular domain maintains a constant spacing of 12 to 14 nm between apposing mesaxon membranes. The spacing between the extracellular leaXets of apposing mesaxon membranes containing MAG and P0 is dictated by the larger MAG molecule (B). The extracellular spacing of compact myelin is mediated via homotypic interactions between apposing P0 molecules (C). The close apposition of the cytoplasmic leaXets of compact myelin may occur by homotypic interactions between the cytoplasmic domain of P0 or by heterotypic interactions involving P0 and acidic lipids. Reproduced from Trapp, 1988, with permission.

In a second hypothesis, note is made of several proteolytic cleavage sites in the cytoplasmic domain of MAG (Malfroy et al., 1985). Proteolysis at these sites releases the transmembrane and extracellular domains of MAG from the membrane (Sato et al., 1982). This degraded MAG product (d-MAG) is soluble and can be detected in pathological conditions of the CNS white matter (Moller et al., 1987). However, d-MAG has not been detected during early stages of normal PNS myelination. Therefore, it seems unlikely that proteolytic removal of MAG participates in conversion of mesaxon membranes to compact myelin. As a third hypothesis, in the absence of solid evidence to support endocytic or proteolytic removal of MAG from mesaxon membranes, one can speculate that the conversion of mesaxon membranes to compact myelin involves a combination of lateral diVusion of MAG and the establishment of speciWc docking sites for the P0-enriched membrane vesicles within mesaxon membranes. This could occur by correct pairing of a v-SNARE on the vesicle membrane with its cognate t-SNARE on the target membrane (Gaisano et al., 1996; Peng et al., 1997; Weimbs et al., 1997). SNARE complex proteins, including the cognate pair vamp-2 and syntaxin-4 were recently detected in oligodendrocytes (Madison et al., 1999), but have yet to be investigated during Schwann cell myelination. In contrast to normal development, MAG and P0 protein co-localize in the same membrane in the dysmyelinating mouse mutants Trembler J (Heath et al., 1991) and Quaking (Trapp, 1988). These mice have deWcits in converting mesaxon membranes to compact myelin. Several observations concerning these studies are worth noting. First, the deWcits in converting mesaxon membranes to compact myelin occur in the setting of remyelination and not during initial myelination. Second, the deWcits occur in a proximal to distal gradient in the peripheral nervous system with ventral roots more severely aVected than sciatic nerves. Third, the deWcits are caused by mutations in the genes for PMP-22 (trembler mouse) (Suter et al., 1992) and Quaking (Ebersole et al., 1996; Hardy et al., 1996), indicating that neither P0 nor MAG is directly aVected. Fourth, while the defects that prevent mesaxon membrane conversion to compact myelin in Trembler J and Quaking diVer, both appear to involve the failure to remove MAG from the mesaxon membrane. In both mutants, regions of PNS myelin internodes appeared normal and the distributions of P0 and MAG were identical to wild type nerves. Both mutants also exhibited multiple wraps of Schwann cell membranes that contained P0 and MAG. The extracellular leaXets of these membranes were always separated by 12 to 14 nm.

35

36

2. CELL BIOLOGY OF MYELIN ASSEMBLY

FIGURE 2.4 Comparison of P0 distributions (red dots) in wild-type and P0-overexpressing Schwann cell surface membranes. During normal myelination, P0 is directly targeted to compact myelin and not detectable in periaxonal, mesaxon, or abaxonal membranes. Extra copies of the P0 gene cause mistargeting of P0 to periaxonal, mesaxon, and abaxonal surfaces. Early mesaxon (box) expansion does not occur due to obligate homophilic adhesion of mesaxon membranes.

Thus, when P0 and MAG are in the same spiraled membrane, the larger extracellular domain of MAG dominates the extracellular spacing between the membranes (Fig. 2.3). The removal of MAG, therefore, is absolutely necessary for the conversion of mesaxon membranes to compact myelin. In these same Quaking membranes, major dense lines did not form; the cytoplasmic leaXets of Trembler J P0/MAG-positive membranes appeared fused but they were twice as thick as those found in normal compact myelin. Trembler J P0/ MAG-positive membranes were located between normal appearing mesaxon and compact myelin membranes, raising the possibility that this may be an intermediate stage in the conversion of mesaxon membranes to compact myelin during normal myelination. Neither P0 nor MAG is essential for spiral growth of PNS myelin membranes. Their presence, however, dictates the spacing of these membranes and modulates the overall stability and eventual viability of the myelin internode. As discussed earlier, initial stages of Schwann cell myelination involve the polarization of abaxonal and adaxonal Schwann cell membranes. One way of looking at compact myelin formation is to consider it a third Schwann cell surface membrane domain that diVerentiates in a timely manner. In support of this hypothesis, MAG is targeted to periaxonal and mesaxon membranes prior to signiWcant detection of P0 protein expression. This suggests that compact myelin production proceeds by a series of maturation events that may be governed, in part, by coordinated expression and targeting of P0 protein. This hypothesis was tested by introducing extra copies of the P0 gene into mice (Wrabetz et al., 2000). In peripheral nerves of these mice, signiWcant levels of P0 protein were detected during initial Schwann cell ensheathment of axons (Yin et al., 2000). In mouse lines with the highest P0 transgene copy number, the spiral wrapping of mesaxon membranes was halted during its initial turn (Fig. 2.4). Early and high-level expression of P0 protein, therefore, arrests myelination during initial stages of spiral wrapping. A signiWcant degree of surface membrane polarization and appropriate protein targeting occurred in these Schwann cells. Basal lamina was formed on their abaxonal surface and MAG was targeted to, and highly enriched in, the periaxonal membrane. The targeting of the early and overexpressed P0 protein diVered signiWcantly from that in wild-type Schwann cells. Over-expressed P0 protein did not accumulate in Golgi membranes. Furthermore, as in wild-type myelinating Schwann cells, P0 and MAG were sorted into separate transport vesicles in the TGN. P0-enriched transport vesicles did not accumulate in the Schwann cell cytoplasm. In contrast to wild-type mice, however, these transport vesicles were targeted to abaxonal, adaxonal, and mesaxon membranes. The presence of P0 protein in these mem-

ROLE OF THE CYTOSKELETON IN CELL DIFFERENTIATION

branes produced abnormal phenotypes, but only where P0-positive membranes apposed each other. P0 and MAG were present in the periaxonal membrane, which displayed a normal ultrastructure, apposing the P0-negative axolemma by 12 to 14 nm. The abaxonal membrane contained signiWcant levels of P0 protein and a basal lamina. P0 in a single membrane, therefore, does not appear to have a dramatic eVect on other protein constituents. However, when P0-positive mesaxon or abaxonal membranes closely appose each other, MAG and basal lamina are excluded, resulting in close opposition (2 nm) of extracellular leaXets. These data are consistent with earlier in vitro studies in which P0 transfection into nonadherent cells induced obligate P0-mediated adhesion, reorganization of the submembranous cytoskeleton, and adherens-like junctions at the transition between adherent and nonadherent membrane domains (D’Urso et al., 1990; Filbin et al., 1990). The conversion of mesaxon membranes to compact myelin appears to be governed by the obligate homophilic trans-adhesion of P0 protein. X-ray crystallography of the extracellular domain of P0 has provided details of the adhesion. Tetrameric P0 complexes homophilically bind to P0 tetramers in opposite orientation on the apposing plasma membranes (Shapiro et al., 1996). An initial step in the conversion of mesaxon membranes to compact myelin, therefore, is the insertion of P0 tetramers in apposing mesaxon membranes. Somehow, these tetramers must Wnd each other and adhere through hydrogen bonds. Critical to this trans association is the possibility that tryptophan residues (Trp-28), oriented at the outer surface of the folded immunoglobulin domain, intercalate into the apposing membranes and assist the homophilic P0 contacts in establishing the close intermembrane apposition (Shapiro et al., 1996). If intercalation of Trp-28 occurs, it may require or induce trans binding of P0 tetramers and possibly other lipid and protein interactions unique to compact myelin membranes. The 2 nm apposition of the extracellular leaXets of normal compact myelin, of P0-positive mesaxon and abaxonal membranes in P0-overexpressing mice, and of adherent plasma membranes of cells transfected with P0 (D’Urso et al., 1990; Filbin et al., 1990) occurs in continuous stretches. This obligate P0-induced 2 nm apposition and exclusion of proteins with large extracellular domains raise the possibility that expansion of compact myelin domains includes trans binding of P0 tetramers for adhesion and cis binding of P0 tetramers for exclusion of molecules that inhibit adhesion. Arrest of myelination in P0 over-expressing mice indicates that mesaxon membranes undergo maturation before they can appropriately handle P0 protein, convert to compact myelin, or continue to spiral. Early conversion to compact myelin is likely to exclude molecules responsible for spiral wrapping of mesaxon membranes.

ROLE OF THE CYTOSKELETON IN CELL DIFFERENTIATION AND MYELIN BIOGENESIS MTs consist of heterodimers of a and b tubulin, which polymerize into cylindrical Wlaments with an external diameter of about 24 nm (Dustin, 1984). Each MT can reach up to 700 mm’s in length, and they are polarized, with a plus and a minus end. MTs are inherently unstable and oscillate between phases of elongation and collapse. MT growth is asymmetrical: subunits are preferentially added or removed at the plus or ‘‘fast growing’’ end. In vivo, the minus end is initially bound to an MT-nucleating structure, such as the centrosome. MT asymmetry is important to MT-based transport, because the motors that translocate organelles and vesicles move preferentially toward either the plus or minus end of the MT.

Schwann Cells Myelination represents a formidable challenge for the Schwann cell. The size and the complex geometry of PNS myelin internodes require specialized mechanisms for coordinating the synthesis, sorting, and targeting of membrane components to sites

37

38

2. CELL BIOLOGY OF MYELIN ASSEMBLY

FIGURE 2.5 Axons regulate the distribution of Schwann cell microtubules. In Schwann cells without axonal contact (A), microtubules radiate from the centrosome. In Schwann cells of myelinated Wbers (B), MTs are dispersed in perinuclear cytoplasm. In degenerating internodes (C), the Schwann cell centrosome is re-established as an MTorganizing center, and multiple MT-organizing centers and MT bundles form along the myelin internode. Reproduced from Kidd et al., 1994, with permission.

millimeters away from their points of synthesis. MTs play a central role in this coordination by organizing organelle distribution and by providing the substrate for intracellular translocation of membrane vesicles destined for the myelin internode (Kidd and Trapp, 1995). Schwann cell MT networks are dynamically regulated by axons (Trapp et al., 1995). Specialized MT distributions and conWgurations mediate site-speciWc targeting of membrane proteins and organelle distribution in the myelinating Schwann cells. Major determinants of MT organization are the location, the number of the MT-nucleating sites, and the orientation of MT-minus ends within the cytoplasmic domains of the myelin internode (Fig. 2.5). As neonatal Schwann cells migrate, their MTs are nucleated from a single site that is associated with the centrosome. As in other migrating cells, the perinuclear MT-organizing center (MTOC) is located between the nucleus and the cell surface so that many MTs are oriented with their plus ends toward the direction of migration (Euteneuer and Schliwa, 1992). The organization and nucleation sites of MTs change to meet the demands of myelin biosynthesis. A major function of MTs in Schwann cells is the transport of myelin and of other membrane proteins that are post-translationally modiWed in the Golgi apparatus. Golgi membranes are tethered to the minus ends of the MTs (Thyberg and Moskalewski, 1985) facilitating the rapid association of the membrane vesicles budding from the Golgi apparatus with the MT-based translocation machinery. As myelination begins, MT minus ends are dispersed throughout the Schwann cell perinuclear cytoplasm so that multiple Golgi stacks become tethered to individual MT networks (Kidd et al., 1996). As part of their myelinating phenotype, Schwann cells rearrange MT networks by changing the location or the mechanisms of MT nucleation. The sites of MT nucleation were determined by reversibly depolymerizing myelinating Schwann cell MTs in vivo (Kidd et al., 1994). Based on the reappearance of MTs, they are generated initially in the perinuclear cytoplasm from multiple sites. MT nucleation or MT self-assembly along the myelin internode did not occur. Thus, it appears that MTs are nucleated in the myelinating Schwann cell perinuclear cytoplasm, and that stable MT units are transported along the internode as demonstrated for axons and dendrites. As in proximal dendrites, MTs located along the outer perimeter of the myelin internode are of a mixed polarity, with 75% having plus ends oriented toward paranodes (Kidd et al., 1994). Schwann cells and dendrites may share mechanisms for transporting MTs of mixed polarity or for rotating MTs 1808, because there are no peripheral MT-nucleating sites in either cell.

ROLE OF THE CYTOSKELETON IN CELL DIFFERENTIATION

The dynamic nature of Schwann cell MTs is further demonstrated by their organization during early stages of Wallerian degeneration (Fig. 2.5C). Following axonal transection, Schwann cells form multiple MTOCs and MT bundles (Kidd et al., 1996). One MTOC is associated with the centrosome. Other MTOCs appear Wrst in the perinuclear cytoplasm and then at sites along the internode, indicating that axonal transection promotes transcription, and transport of MT-nucleating material along the internode. These MTOCs are located between degenerating myelin ovoids. Lysosomes and endosomes become concentrated at these MTOCs through association with MT minus ends (Matteoni and Kreis, 1987) and participate in myelin degradation. This is the only example described for MTOC-based peripheral targeting in Schwann cells. MTs can be disassembled by a variety of methods, including MT depolymerizing drugs (colchicine, colcemid, nocodazole), low temperature, and hypertonic shock. In one study, the MT depolymerizing drug colchicine was used to reversibly depolymerize myelinating Schwann cell MTs in rat sciatic nerve (Trapp et al., 1995). Two distinct but overlapping functions for MTs during myelination were identiWed. As described in other cells (Lee et al., 1989; Rogalski and Singer, 1984; Terasaki et al., 1986), Schwann cell MT disassembly disrupted the organization of Golgi membranes, endoplasmic reticulum (ER) and intermediate Wlaments (Fig. 2.6). Although the organization of RER and Golgi membranes was disrupted, these organelles continued to synthesize P0, MAG and laminin. These proteins accumulated in the Schwann cell perinuclear cytoplasm (Fig. 2.7), indicating that MTs are essential for their transport to sites along the myelin internode. MTs, therefore, organize the distribution of the major membrane synthesizing organelles (RER and Golgi apparatus) in the Schwann cell perinuclear cytoplasm. They are also essential for the normal distribution of the organelles that are present in the cytoplasmic channels, which extend along the outer perimeter of the myelin internode. MT disassembly segregated these transport channels into organelle-free and organelle-enriched domains. The organelle-rich domains were further organized into a central core of smooth ER (SER)

FIGURE 2.6 Microtubules help organize myelinating Schwann cell organelles. Normal distributions of endoplasmic reticulum (A) and intermediate Wlaments (B) are disrupted by microtubule disassembly (C, endoplasmic reticulum; D, intermediate Wlaments). Rough endoplasmic reticulum forms irregular proWles (E, arrows), and aggregates of small vesicles (E, arrowheads) accumulate near distended Golgi cisternae (G). Reproduced from Trapp et al., 1995, with permission. Scale bars ¼ 5 mm (A-D) and 0.2 mm (E).

39

40

2. CELL BIOLOGY OF MYELIN ASSEMBLY

FIGURE 2.7 Microtubule (MT) depolymerization disrupts targeting of Schwann cell proteins. Comparisons of P0 (A, B) and laminin (C, D) distributions in normal myelinated Wbers (A, C) and after colchicine treatment (B, D) indicate substantial accumulation of both glycoproteins in the Schwann cell cytoplasm in the absence of MTs. As shown schematically (E), P0, MAG, and laminin are processed in the perinuclear Golgi apparatus, sorted into separated carrier vesicles in the trans Golgi network, transported along the internode on MTs, and directly inserted into the appropriate surface membrane. Panels A through D have been reproduced from Trapp et al., 1995, with permission. Scale bars ¼ 5 mm.

(Fig. 2.6C) that was surrounded by intermediate Wlaments (Fig. 2.6D), mitochondria, endosomes, and lysosomes. The interdependence between MT and SER distributions is likely to be very important for myelin membrane assembly, since many lipids enriched in compact myelin are synthesized in the SER of the cytoplasmic channels (Gould and Sinatra, 1981). MTs, therefore, help coordinate the transport of Golgi-derived membrane vesicles with the local lipid synthesis by extending or distributing the SER throughout the cytoplasmic channels. P0 and MAG distributions in the Schwann cell perinuclear cytoplasm, following MT disassembly, have also provided insights into how these molecules are sorted and targeted to diVerent membrane domains (Fig. 2.7E). Electron microscopic immunocytochemistry detected P0 and MAG in separate carrier vesicles that accumulated near the trans Golgi network (Trapp et al., 1995). Furthermore, P0-rich carrier vesicles fused and formed compact myelin-like membrane whorls, while MAG-rich carrier vesicles formed mesaxon-like membrane whorls. MT disassembly did not result in mistargeting of P0 or MAG to surface membranes. These data are consistent with the hypothesis that MAG and P0 are sorted into separate transport vesicles during synthesis, transported along the myelin internode by MTs and then inserted directly into the appropriate membrane domains. It is likely, therefore, that P0 and MAG contain speciWc sorting signals. It is possible that formation of cis tetramers occurs during P0 synthesis. This would be an eVicient mechanism for obtaining P0-rich membrane vesicles while simultaneously excluding other proteins from being targeted to compact myelin. Mechanisms must also be available to stabilize proteins in their speciWc membrane domains. Similar to polarized epithelial cells, adherens-like junctions separate the abaxonal membrane from the outer mesaxons and restrict lateral diVusion between these membrane domains. The transition from compact myelin to mesaxons, Schmidt-Lanterman incisures, and paranodal loops is abrupt and is not demarcated by junctional complexes. While the precise nature of the diVusion barrier operating at this interface is unknown, it is likely to involve the tight adhesion of compact myelin membranes (mediated by P0 protein) and the association of a submembranous cytoskeleton (Trapp et al., 1989b) within the noncompact regions of the internode. Other adhesive interactions such as in the septate junctions between paranodal loops and the axolemma also prevent lateral diVusion of both axolemmal and myelin internode proteins.

ROLE OF THE CYTOSKELETON IN CELL DIFFERENTIATION

Oligodendrocytes The size and the complex geometry of oligodendrocytes and their multiple myelin internodes require specialized mechanisms for coordinating the synthesis, sorting, and targeting of myelin components. As for Schwann cells, MTs help organize oligodendrocyte synthetic machinery and provide the substrates for intracellular translocation of myelin gene products. In contrast to Schwann cells that form one myelin internode, oligodendrocytes must coordinate the delivery of membrane components to multiple myelin internodes. The oligodendrocyte does this, in part, by developing several ‘‘biosynthetic stations’’ within their perinuclear cytoplasm. These can be viewed in light microscopy as cytoplasmic bulges from which one or more main processes emerge. Each of these main trunks can subsequently branch into several Wner processes that form CNS myelin internodes (Sternberger et al., 1979). In contrast to Schwann cells, where transport occurs in a bidirectional, linear manner from a centrally located perinuclear region, oligodendrocytes must initially coordinate myelin component delivery radially. Once myelin components reach the CNS myelin internode, they are then delivered linearly and bidirectionally. It is likely, therefore, that oligodendrocytes utilize specialized MT organizations not found in Schwann cells. As discussed later, this hypothesis is supported by an oligodendrocyte-speciWc MT-related gene defect in the taiep rat. In contrast to Schwann cells where most MT-related data has been obtained in vivo, most studies of oligodendrocyte MT function have been conducted in vitro. Unlike Schwann cells, oligodendrocytes in vitro express myelin proteins, extend multiple processes, and form myelin-like sheets. PuriWed oligodendrocytes also provide a means of readily identifying microtubule-associated proteins (MAPs). Therefore, much more is known about oligodendrocyte MAPs than Schwann cell MAPs. One point must be stressed regarding oligodendrocyte MT data obtained in vitro. Most studies have been performed in the absence of axons. If axons play a major role in regulating MT function in oligodendrocytes as demonstrated in Schwann cells (Kidd et al., 1996), some caution must be taken in extrapolating data to in vivo myelination. Many MT functions are shared by Schwann cells and oligodendrocytes. Treatment of isolated oligodendrocytes (Benjamins and Nedelkoska, 1994; Sato et al., 1986) or brain slices (Bizzozero et al., 1982) with MT depolymerizing or stabilizing drugs reduces or halts incorporation of myelin components into myelin membranes. As discussed earlier, MTs are essential for translocation of MBP mRNA (Carson et al., 1997). MTs in oligodendrocyte progenitor cells are nucleated and attached to a centrosomal MTOC (Simpson and Armstrong, 1999), while diVerentiated oligodendrocytes have a dispersed noncentrosomal MT network (Lunn et al., 1997; Simpson and Armstrong, 1999; Song et al., 2001). MTs in the larger oligodendrocyte processes have a mixed polarity (Song et al., 2001) similar to that described in Schwann cells and dendrites (approximately 80% with distal plus ends). MTs in smaller processes appear to be oriented with their plus ends distally. Both plus and minus end MT motors have the potential to operate in both anterograde and retrograde transport. Nonlinear regression analysis of MT disassembly in cultured oligodendrocytes have identiWed two labile MT populations distinguished by short (4.2 min) and long (389 min) half lives (Song et al., 2001). The more stable MT populations were enriched in oligodendrocyte processes. Since all MTs appear to be nucleated in perinuclear cytoplasm, it would make sense that MTs in processes and along the myelin internode have longer half-lives. Stability of MTs is increased by specialized b-tubulin isoforms or MAPs. While MAPs have not been associated with the stable MT population in oligodendrocytes, various developmentally regulated MAPs are expressed by oligodendrocytes in vitro. These include MAP1B (Fischer et al., 1990), MAP4 (Vouyiouklis and Brophy, 1995), MAP2C (Richter-Landsberg and Gorath, 1999; Vouyiouklis and Brophy, 1995), and various isoforms of tau (LoPresti et al., 1995; Muller et al., 1997). The gene that encodes MAP2C, a major MAP in oligodendrocytes, undergoes diVerential splicing from a protein with 3 MT binding domains to one with 4 MT binding domains during oligodendrocyte diVerentiation in vitro (Richter-Landsberg and Gorath, 1999). Tau proteins are expressed by both immature and mature oligodendrocytes and are enriched at branching points and

41

42

2. CELL BIOLOGY OF MYELIN ASSEMBLY

at the ends of oligodendrocyte processes. Tau isoforms with more MT binding domains also appear to increase with oligodendrocyte maturation (Richter-Landsberg and Gorath, 1999). Both MAP2c and Tau, therefore, may induce MT stability or specialized MT functions during oligodendrocyte myelination. Such functions remain to be identiWed. It has been postulated that oligodendrocytes may also express unique MAPs. Indirect evidence from studies of the dysmyelinating mutant rat, taiep, provides compelling support for this hypothesis (Duncan et al., 1992; Lunn et al., 1997; Song et al., 2001). Taiep rats develop progressive neurological phenotypes due to CNS-speciWc dysmyelination (Duncan et al., 1992). Morphological phenotypes include accumulation and bundling of MTs in oligodendrocytes and reversal of MT polarity in oligodendrocyte processes (Song et al., 2001). Since the phenotype appears oligodendrocyte-autonomous and is autosomal recessive, it is likely to be due to a genetic defect in an oligodendrocyte-speciWc MAP, which indirectly reduces myelin protein processing, myelin formation, and myelin internode maintenance. Levels of myelin protein mRNAs were unaVected. MT bundling or reversal of MT may contribute to these phenotypes. Interestingly, reversal of MT depolarization by nocodazole, restored targeting of PLP in vitro, suggesting accumulation of a mutated protein, which binds to polymerized MTs. IdentiWcation of the gene defect in taiep should identify the Wrst oligodendrocyte speciWc MAP.

OLIGODENDROCYTE DIFFERENTIATION AND INITIATION OF MYELINATION Oligodendrocyte myelination diVers from Schwann cell myelination in several ways. Oligodendrocytes have the potential to myelinate multiple axons. Proteolipid protein (PLP) is the major structural protein of CNS myelin. While axons regulate CNS myelination, the sequence of events in oligodendrocyte diVerentiation (Fig. 2.8) diVers from that of Schwann cells. Oligodendrocytes are produced by a progenitor cell that originates in

FIGURE 2.8 Schematic summary of oligodendrogenesis in vivo. Oligodendrocyte progenitor cells (OPCs) originate in subventricular zones (SVZ) and populate the CNS as early progenitors. Late oligodendrocyte progenitor cells give rise to premyelinating oligodendrocytes that either myelinate axons or die through programmed cell death. Cells with phenotypic characteristics of OPCs are abundant in the adult mammalian brain. Some of the phenotypic markers for each stage are listed. Adapted from Trapp et al., 1997, with permission.

OLIGODENDROCYTE DIFFERENTIATION AND INITIATION OF MYELINATION

FIGURE 2.9 DiVerentiation of myelinating oligodendrocytes. Confocal images of premyelinating (A), transitional premyelinating (B), and myelinating oligodendrocytes (C). Oligodendrocytes initially extend multiple processes (A). Many but not all ensheath and myelinate axons. Premyelinating cells express DM-20 (red). As myelination proceeds, proteolipid protein is also expressed (B, C, yellow) and targeted only to compact myelin (B, C, arrowheads). Reprinted from Trapp et al., 1997, with permission. Scale bars ¼ 10 mm.

specialized regions of the subventricular zone. Early oligodendrocyte progenitor cells (OPCs) express the platelet derived growth factor receptor a (PDGFaR) and the sulfated proteoglycan, NG2 (Nishiyama et al., 1991, 1999). They migrate, divide, and populate the nervous system during early development and establish a network of stellate cells that cover most of the CNS parenchyma. Each cell establishes a microdomain within the neuropil, and contact inhibition with neighboring OPCs may control their distribution. These early OPCs diVerentiate into a late OPC stage (or oligodendroblasts) that appears to be committed to oligodendrogenesis. Late OPCs express the tetraspan protein CD9 (Terada et al., 2002) and the POA antigen recognized by the monoclonal antibody O4, but do not express sulfatide or galactosylcerebroside (Bansal et al., 1992). Late OPCs are postmigratory but still proliferative (Warrington et al., 1993). Once oligodendrogenesis is complete, stellate NG2-, PDGFaR-positive cells remain as a major glial component of the adult mammalian CNS, apparently providing a potential pool of quiescent progenitors that can be tapped later for repair of demyelinated axons. OPCs diVerentiate into oligodendrocytes in a temporal and spatial sequence that precedes myelination by a few days (Trapp et al., 1997; Ueda et al., 1999). Newly produced oligodendrocytes, referred to as premyelinating oligodendrocytes, are characterized by distinct morphological and molecular criteria. First, premyelinating oligodendrocytes radially and symmetrically extend multiple processes (Fig. 2.9A). Second, they express many but not all myelin proteins. Myelin proteins detected in premyelinating oligodendrocytes include DM-20, MAG, 2’, 3’-cyclic nucleotide 3’-phosphodiesterase (CNP), and MBP. PLP and MOG are not expressed at levels that have been detected by current methods. All myelin proteins so far detected in premyelinating oligodendrocytes appear evenly distributed throughout the entire cell. This highlights a major diVerence between oligodendrocyte and myelinating Schwann cell diVerentiation. Oligodendrocytes appear to be constitutively produced in the developing CNS and express many myelin protein genes before genes that encode targeting machinery. Axons, however, play a role in regulating or modulating oligodendrocyte number, but they are not essential for oligodendrocyte production. Following transection of optic nerve or dorsal root rhizotomy, oligodendrocytes survive and express myelin protein genes (Kidd et al., 1990; McPhilemy et al., 1990) (Fig. 2.10). In the absence of axons in vitro, OPCs produce oligodendrocytes that can generate myelin-like membranes. This is in stark contrast to the PNS, where Schwann cell expression of myelin proteins requires axonal contact. Premyelinating oligodendrocytes in vivo have a limited life span (2 to 3 days in rodent brain) and one of two fates: programmed cell death or myelination. It has been estimated that as many as 50% of newly produced oligodendrocytes die (Barres et al., 1992a, 1992b). The incidence of premyelinating oligodendrocyte programmed cell death, however, remains to be established because current estimates were based on unconWrmed 1-hour clearance times of dying cells.

43

44

2. CELL BIOLOGY OF MYELIN ASSEMBLY

FIGURE 2.10 Axons diVerentially regulated myelin gene expression in the PNS (A-E) and CNS (F-H). Following transection of peripheral nerves, myelin protein mRNA levels rapidly (by 5 days) decrease to undetectable levels in Schwann cells undergoing Wallerian degeneration. Panels A through D compare P0 (A, B) and MBP (C, D) mRNA distributions determined by in situ hybridization in intact (A, C) and 5d transected (B, D) sciatic nerves. QuantiWed by slot blot hybridization (E), both mRNAs are undetectable by 5d post-transection. In contrast, oligodendrocytes continue to express myelin protein mRNAs 40d after axonal degeneration. Panels F and G show in situ hybridization distributions of PLP (F) and MBP (G) mRNAs in spinal cord dorsal columns with intact axons (right of dotted line) and 40d after contralateral axon transection (left). As quantiWed by Northern blots (H), oligodendrocytes in axon-free optic nerves express signiWcant myelin protein mRNA levels. Reproduced from Trapp et al., 1988b (A through E), and Kidd et al., 1990 (F through H), with permission. Scale bars ¼ 50 mm (AD) and 100 mm (F, G).

DiVerentiation of a premyelinating oligodendrocyte to a myelinating oligodendrocyte is characterized by further morphological and molecular changes. Most important to the process of myelination is the maturation of the oligodendrocyte from a nonpolarized to a polarized cell. As myelination commences, the oligodendrocyte begins to target myelin proteins to speciWc membrane domains (Trapp et al., 1997). MAG is selectively targeted to periaxonal membranes, PLP to compact myelin, CNP to noncompact regions of the myelin internode, and MBP mRNA is transported along oligodendrocyte processes. Additional proteins enriched in the paranodal (e.g., NF-155) and juxtaparanodal regions, for which specialized targeting mechanisms must also develop, are under current investigation. This implies that a major inducer of myelination is not myelin protein gene expression but rather the development of specialized transport and targeting pathways for molecules enriched in myelin and other membrane domains.

OLIGODENDROCYTE MYELINATION

The initiation of the myelinating phenotype may also involve the expression of the myelinspeciWc proteins PLP and MOG, neither of which has been detected in premyelinating oligodendrocytes. When detected, PLP appears only in compact myelin, suggesting that the 35 amino acids that distinguish PLP and its spliced variant, DM-20, contain signals that target PLP exclusively to compact myelin. Not all processes radially extending from premyelinating oligodendrocytes form myelin sheaths, and those that do not appear to be reabsorbed into the cell. The subsequent maturation of the surface membranes of the myelinating oligodendrocyte perikaryon and of the processes extending to myelin internodes includes the removal of MAG, MBP, and DM-20 and the addition or retention of other molecules.

OLIGODENDROCYTE MYELINATION How does the CNS myelin internode expand? Little is currently known about the molecular events that accompany early stages of CNS myelination. Oligodendrocytes extend multiple processes that radially and longitudinally ensheath axons. MAG is restricted to the periaxonal membrane of CNS myelin internodes and to the small stretch of inner tongue process membrane that connects periaxonal and compact myelin. Compact myelin usually begins to form during the initial one or two spirals of the oligodendrocyte process. Molecules that are speciWc or enriched in the early wraps (such as MAG in the PNS mesaxon membranes) have not been identiWed during CNS myelination.

Proteolipid Protein The molecular interactions responsible for targeting PLP to compact myelin are unknown. Alterations in PLP processing, however, are a major cause of inherited diseases of CNS myelin. These conditions are caused by mutations in the PLP gene and by PLP overexpression due to gene duplication (Hudson et al., 1989; Inoue et al., 1999; Nave et al., 1987). This topic is discussed in detail in other chapters. Several aspects of the PLP processing are relevant to the present chapter and are brieXy described. In addition, general concepts may apply to PMP-22 biosynthesis and to inherited dysmyelinating diseases in the PNS (Snipes et al., 1992). PLP is an intrinsic membrane protein with four transmembrane domains. A C-terminal truncation of PLP in Jimpy (JP) mice (Macklin et al., 1987; Nave et al., 1987) and a PLP point mutation in JimpyMSD (MSD) mice (Gencic and Hudson, 1990) causes protein misfolding and accumulation in RER (Gow et al., 1994, 1998). Misfolded proteins are usually degraded in the lysosomal or proteosomal systems. The overload of the proteosomal degradation pathway by misfolding of the proteins with multiple transmembrane domains (e.g., ion channels, tetraspan proteins) can cause cell death in a variety of cell types. The JP and MSD mutations kill oligodendrocytes only after a threshold of mutated protein accumulates during myelination. Over-expression of PLP can also be toxic and kill oligodendrocytes by a similar mechanism (Kagawa et al., 1994; Readhead et al., 1994). It is interesting to note that mutated or over-expressed P0 proteins do not accumulate in the ER or Golgi apparatus, nor appear toxic to myelinating Schwann cells (Yin et al., 2000). Toxic accumulation, therefore, appears to be restricted to myelin proteins with multiple transmembrane domains. Existing data support the concept that misfolding of the PLP peptide is responsible for the failure of processing. This implies that certain amino acid motifs exposed at the cytoplasmic or extracellular domains of PLP contain RER sorting and processing sequences that are disrupted or lost in the mutated and misfolded protein. Accumulation of over-expressed PLP in the RER may tax or saturate the availability of molecules that help fold or move PLP through the RER.

Myelin Associated Glycoprotein (MAG) MAG is a type I integral membrane glycoprotein comprised of a single transmembrane domain, a cytoplasmic C-terminal, and an extracellular N-terminal with Wve

45

46

2. CELL BIOLOGY OF MYELIN ASSEMBLY

immunoglobulin-like domains. Two developmentally regulated MAG polypeptides with molecular masses of 72 kD (L-MAG) and 67 kD (S-MAG) are produced by diVerential splicing of a primary RNA transcript (Arquint et al., 1987; Lai et al., 1987; Salzer et al., 1987). S-MAG is missing the 44 C-terminal amino acids of L-MAG and contains a Cterminal decapeptide not present in L-MAG. L- and S-MAG, therefore, diVer in their cytoplasmic domains, which often contain targeting or sorting sequences. MAG is selectively enriched in the periaxonal membrane of CNS myelin internodes and to the small stretch of inner tongue process membrane that connects periaxonal and compact myelin (Sternberger et al., 1979; Trapp, 1988). This restricted distribution diVers from that in PNS myelin internodes where MAG is also enriched in mesaxons, Schmidt-Lanterman incisures, and paranodal loop membranes. In the CNS, L-MAG is more abundant than S-MAG during the early stages of myelination, whereas L- and S-MAG are present at approximately equal amounts in mature CNS myelin internodes. Schwann cells express mostly S-MAG at all stages of myelination. Immunocytochemical studies investigated the possibility that L-MAG may contain amino acid sequences that aVect its targeting or sorting during early stages of CNS myelination. In contrast to mature CNS internodes and PNS myelin internodes, L-MAG was detected in endosomes during early stages of CNS myelination (Trapp et al., 1989a). L-MAG-enriched endosomes were present in inner tongue processes, paranodal loops, outer tongue processes, oligodendrocyte processes extending to myelin internodes, and oligodendrocyte perinuclear cytoplasm (Fig. 2.11). These data indicate that L-MAG, but not S-MAG, contains an endocytic signal. Since endocytosed MAG was not targeted to other membrane domains and remained in endosomes that were transported back to the oligodendrocyte cell body, it is most likely that the endocytosed L-MAG is degraded in the endosomal/lysosomal system. It is also possible that L-MAG endocytosis is mediated by binding to a ligand present in the periaxonal space or axolemma. This could serve as one mechanism by which axons inXuence oligodendrocyte myelination. A selective abnormality in the distribution of L-MAG, but not S-MAG occurs during early stages of CNS myelination in the hypomyelinating-dysmyelinating mouse mutant Quaking (Bo¨ et al., 1995). As in wild-type mice, L-MAG was inserted into the periaxonal membrane of developing quaking CNS internodes and removed by receptor-mediated endocytosis. In contrast to wild-type mice, however, the amount of L-MAG detected in endosomes was increased 5-fold. In addition, endocytosed L-MAG was not replaced, resulting in a dominance of S-MAG in adult quaking CNS internodes. The dramatic reduction of L-MAG in mature quaking CNS internodes may disturb axon-oligodendrocyte signaling and contribute to the hypo/dysmyelination.

FIGURE 2.11 L-MAG is endocytosed from CNS periaxonal membranes. Immunocytochemical distribution of MAG in endosomes present in periaxonal (A) outer tongue processes (B) and perinuclear regions (C) of oligodendrocytes. MAG is only detected in periaxonal membranes (arrowheads) of CNS myelin internodes. Endocytosed L-MAG is replaced in part by S-MAG. En ¼ endosomes, Ax ¼ axon, Nu ¼ nucleus of oligodendrocyte. Reproduced from Bo¨ et al., 1995, with permission. Scale bars ¼ 0.2 mm.

OLIGODENDROCYTE MYELINATION

The presence of an endocytic signal in the C-terminal of L-MAG is likely to reside in the 44 amino acids that are not present in S-MAG (Bo¨ et al., 1995). Three consensus motifs (tyrosine-based, dileucine and diacidic) are present in the L-MAG-speciWc domain and may potentially serve as endocytosis signals, but have not as yet been tested for endocytic activity.

2’, 3’-cyclic nucleotide 3’ phosphodiesterase CNP is enriched in myelin-forming cells and, as its name implies, catalyzes the hydrolysis of 2’, 3’-cyclic nucleotides into their corresponding 2’-nucleotides (Sprinkle, 1989; Tsukada and Kurihara, 1992). A substrate for this enzymatic activity has not been identiWed in the nervous system. CNP is excluded from compact myelin and enriched at the cytoplasmic surface of all other surface membranes of myelinating oligodendrocytes (Braun et al., 1988; Trapp et al., 1988a). While the precise function of CNP is unknown, studies support the possibility that it associates with the submembranous cytoskeleton where it participates in membrane expansion and migration. When transfected into nonmyelinating cells, they produce Wlopodia and extend processes (De Angelis and Braun, 1997). CNP is an extrinsic membrane protein, which is translated on free polysomes located in the oligodendrocyte perinuclear cytoplasm (Trapp et al., 1988a). The vast majority of CNP, however, posttranslationally associates with membranes by isoprenylation (Braun et al., 1991). CNP has not been linked to inherited diseases of myelin. Premyelinating oligodendrocytes in transgenic mice that over-express CNP form aberrant periaxonal membranes at the myelin internodes and at the contact point between oligodendrocyte processes and the myelin internode. The abnormal membrane expansion culminates in large intramyelinic vacuoles in mature CNP over-expressing mice (Gravel et al., 1996). These ‘‘gain of function’’ changes produced by CNP over-expression are consistent with the hypothesis that CNP is part of a molecular complex that regulates myelin membrane expansion. CNP shares structural features with the Rho family of GTP-binding proteins and may operate mechanistically in a similar manner (Bussey, 1996; Takai et al., 1995; Zigmond, 1996). Ultrastructural immunocytochemistry identiWed mistargeting of CNP and a failure to form major dense lines in CNP-overexpressing mice (Yin et al., 1997). Although MBP mRNA levels are similar in wild-type and CNP over-expressing mice, MBP protein is reduced by approximately 30%, and myelin compaction is inhibited (Gravel et al., 1996). CNP over-expression, therefore, appears to post-transcriptionally reduce MBP levels. Since MBP and CNP both bind to the cytoplasmic leaXets of surface membranes, CNP in compact myelin may inhibit MBP binding and the formation of major dense lines. Cytoplasmic unbound MBP may regulate MBP translation or turnover. It is likely that CNP plays a similar role during normal myelination by prohibiting the binding of MBP to noncompact myelin membranes (i.e., outer tongue processes and paranodal loops).

Myelin Basic Protein and mRNA Translocation Since the RER and Golgi apparatus are conWned to oligodendrocyte perinuclear cytoplasm, all integral membrane proteins are likely to be synthesized there and then delivered to myelin via membrane vesicles originating in the trans Golgi network. Vesicular transport, however, is not the only mechanism by which myelin proteins are targeted to the myelin internode. It is well established that mRNA translocation to the outer tongue process of myelin internodes is the mechanism by which MBP is delivered to the myelin internode (Barbarese et al., 1999; Farina and Singer, 2002). Initial indications that MBP mRNA may be transported to and translated at sites close to compact myelin were based on three observations. First, biochemical studies demonstrated incorporation of MBP into compact myelin within minutes of its synthesis. In contrast, a delay of 30 minutes occurred between synthesis of PLP and P0 and their insertion into compact myelin (Benjamins et al., 1978; Rapaport and Benjamins, 1981). Second, MBP mRNA (but not PLP mRNA) was enriched 20-fold in RNA extracts from isolated myelin when compared to RNA extracts of

47

48

2. CELL BIOLOGY OF MYELIN ASSEMBLY

whole brain homogenates (Colman et al., 1982). Third, in situ hybridization using cDNA probes detected MBP mRNA throughout white matter (Trapp, 1988). In contrast, PLP and P0 cDNAs were enriched in the perinuclear regions of the myelinating cells (Fig. 2.12). The hypothesis that MBP mRNA transport is an active process was supported by the observation that other myelin protein mRNAs translated on free polysomes (CNP and P2 protein) were conWned to the perinuclear regions of the oligodendrocytes (Gillespie et al., 1990; Trapp et al., 1988a). Although a variety of mechanisms contribute to mRNA targeting, two aspects appear universal. Translocated mRNA transcripts contain cis-acting elements that direct their localization by binding to speciWc trans-acting factors. With a few exceptions, all cis-acting sequences reside in the 3’ untranslated regions (UTRs) of the transported mRNA. An 11 nucleotide RNA transport signal (RTS) is suVicient to direct MBP mRNA translocation from the oligodendrocyte cell body into oligodendrocyte processes (Munro et al., 1999). The trans-acting factor is a heterogeneous nuclear riboprotein, hnRNP A2, which binds to the RTS (Hoek et al., 1998) and is essential for MBP mRNA targeting in vitro (Munro et al., 1999). hnRNP A2 appears to associate with MBP mRNA in the nucleus and is part of the cytoplasmic granules, which contain MBP mRNA, but not other nontransported mRNAs (Ainger et al., 1993, 1997). hnRNP A2 may also modulate other functions including splicing, RNA stabilization, and translation regulation (Caceres and Kornblihtt, 2002; Kwon et al., 1999). The intracellular movement of MBP mRNA has been visualized in vitro following injection of Xuorescently tagged MBP mRNA into oligodendrocytes (Ainger et al., 1993). The MBP mRNA is incorporated into granules that also contain hnRNP A2 (Munro et al., 1999), ribosomal RNA, the elongation factor EF1-a, and arginyl-transfer RNA synthetase (Barbarese et al., 1995). These particles translocated at an average rate of 12 mm/min, which is consistent with MT-mediated transport (Carson et al., 1997). The MBP mRNA granules that associate with MTs are detergent insoluble and their translocation is inhibited by MT depolymerization and treatment with kinesin heavy chain and hnRNP A2 antisense oligonucleotides (Barbarese et al., 1999; Carson et al., 1997; Munro et al., 1999). To date, all data support the plus-end anterograde MT-based translocation of MBP mRNA in translation-competent ribonucleoprotein granules. Another 340nt, cis-acting, 3’ UTR element termed the RNA localization signal (RLS) is thought to play a role in anchoring the MBP mRNA granule in the outer tongue process of the myelin internode (Ainger et al., 1993, 1997). The trans-acting factor for this cis element has not been identiWed and the speciWcity of the cis domain may depend on secondary structure of the 340nt element (Palacios and Johnston, 2001). It is possible that this interaction is essential for translational activation of the MBP mRNA. Why are MBP mRNAs translocated to the myelin internode? The most likely explanation is related to the high aVinity of MBP for acidic lipids. This binding is relatively nonspeciWc and MBP coating of oligodendrocyte organelles may be detrimental to normal function. Translation of MBP, therefore, close to its site of insertion into compact myelin would be one mechanism to assure appropriate and exclusive targeting to myelin. It has also been suggested that MBP mRNA may be transported with other mRNAs. MBP mRNA is not the only translocated mRNA in oligodendrocytes. Carbonic anhydrase II and myelin-associated oligodendrocyte basic protein mRNAs are also distributed diVusely in white matter. Synchronous translation of these mRNAs could provide a mechanism for regulating the stoichiometry of diVerent protein constituents in the developing myelin internode. Oligodendrocytes should continue to provide a model system for elucidating mechanisms and functions of mRNA translocation that are utilized by all cells.

SUMMARY AND PERSPECTIVES Polarized membrane assembly is a complex process, requiring the coordinated synthesis, transport, and sorting of proteins and lipids. Studies of simpler model systems such as polarized epithelial cells and synaptic exocytosis have provided valuable insights into

SUMMARY AND PERSPECTIVES

FIGURE 2.12 Spatial segregation of myelin protein mRNAs at the CNS-PNS boundary (A-D). PLP mRNA (A), visualized as bright silver grains, is restricted to the CNS and is clustered around oligodendrocyte nuclei. P0 mRNA (B) is restricted to the PNS and is concentrated around Schwann cell nuclei. In contrast, MBP mRNA (C) is diVusely distributed over myelinated Wbers in both CNS and PNS, indicating that MBP mRNAs are transported into the myelinating cell processes. Immunostaining for MBP (D, staining is black) demonstrates the higher concentration of MBP in oligodendrocyte myelin compared with PNS. MBP mRNA traYcking pathway in an oligodendrocyte (E). Mouse oligodendrocyte in culture (green) was microinjected with MBP mRNA (appears yellow). The RNA assembled into granules, which are transported along microtubules in the processes, eventually becoming localized to the distal, Xattened membranous sheets. The molecular components involved at each step in the pathway are illustrated schematically in the overlay diagram. Reproduced from Trapp et al., 1987 (A through D), and Barbarese et al., 1999 (E), with permission. Scale bars ¼ 100 mm.

myelin biogenesis. Myelin production is exceptional, however, in the magnitude of membrane biogenesis, the complex domain structure of the myelin membranes, and the need to coordinate membrane biosynthesis with axonal properties. Thus, it seems likely that Schwann cells and oligodendrocytes will also utilize novel components or mechanisms. Concepts of myelin assembly will continue to evolve as imaging and allied technologies

49

50

2. CELL BIOLOGY OF MYELIN ASSEMBLY

advance, and as new mechanisms of membrane organization, cell adhesion and signal transduction, and intracellular protein and lipid traVicking come to light.

Acknowledgments The authors thank Vikki Pickett for editorial assistance and Rosalia Yacubova for help in preparing the Wgures. Supported by research grants from the National Institutes of Health; NS35058 (BDT, NS38186 (BDT), NS29818 (BDT) and NS38667 (BDT), NS10861 (SP), and NS41078 (SP).

References Aguayo, A. J., Charron, L., and Bray, G. M. (1976a). Potential of Schwann cells from unmyelinated nerves to produce myelin: A quantitative ultrastructural and autoradiographic study. J. Neurocytol. 5, 565–573. Aguayo, A. J., Epps, J., Charron, L., and Bray, G. M. (1976b). Multipotentiality of Schwann cells in cross anastomosed and grafted myelinated and unmyelinated nerves: Quantitative microscopy and radioautography. Brain Res. 104, 1–20. Ainger, K., Avossa, D., Diana, A. S., Barry, C., Barbarese, E., and Carson, J. H. (1997). Transport and localization elements in myelin basic protein mRNA. J Cell Biol. 138, 1077–1087. Ainger, K., Avossa, D., Morgan, F., Hill, S. J., Barry, C., Barbarese, E., and Carson, J. H. (1993). Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes. J. Cell Biol. 123, 431–441. Arquint, M., Roder, J., Chia, L.-S., Down, J., Wilkinson, O., Bayley, H., Braun, P., and Dunn, R. (1987). Molecular cloning and primary structure of myelin-associated glycoproteins. Proc. Natl. Acad. Sci. USA 84, 600–604. Bansal, R., Kumar, M., Murray, K., and PfeiVer, S. E. (1996). Developmental and FGF-2-mediated regulation of syndecans (1–4) and glypican in oligodendrocytes. Mol. Cell Neurosci. 7, 276–288. Bansal, R., Stefansson, K., and PfeiVer, S. E. (1992). Proligodendroblast antigen (POA), a developmental antigen expressed by A007/04-positive oligodendrocyte progenitors prior to the appearance of sulfatide and galactocerebroside. J. Neurochem. 58, 2221–2229. Barbarese, E., Brumwell, C., Kwon, S., Cui, H., and Carson, J. H. (1999). RNA on the road to myelin. J Neurocytol. 28, 263–270. Barbarese, E., Koppel, D. E., Deutscher, M. P., Smith, C. L., Ainger, K., Morgan, F., and Carson, J. H. (1995). Protein translation components are colocalized in granules in oligodendrocytes. J Cell Sci 108, 2781–2790. Barman, S., and Nayak, D. P. (2000). Analysis of the transmembrane domain of inXuenza virus neuraminidase, a type II transmembrane glycoprotein, for apical sorting and raft association. J Virol. 74, 6538–6545. Barres, B. A., Hart, I. K., Coles, H. S., Burne, J. F., Voyvodic, J. T., Richardson, W. D., and RaV, M. C. (1992a). Cell death in the oligodendrocyte lineage. J. Neurobiol. 23, 1221–1230. Barres, B. A., Hart, I. K., Coles, H. S. R., Burne, J. F., Voyvodic, J. T., Richardson, W. D., and RaV, M. C. (1992b). Cell death and control of cell survival in the oligodendrocyte lineage. Cell 70, 31–46. Benjamins, J. A., Iwata, R., and Hazlett, J. (1978). Kinetics of entry of proteins into the myelin membrane. J. Neurochem. 31, 1077–1085. Benjamins, J. A., and Nedelkoska, L. (1994). Maintenance of membrane sheets by cultured oligodendrocytes requires continuous microtubule turnover and Golgi transport. Neurochem Res. 19, 631–639. Benting, J. H., Rietveld, A. G., and Simons, K. (1999). N-Glycans mediate the apical sorting of a GPI-anchored, raft-associated protein in Madin-Darby canine kidney cells. J. Cell Biol. 146, 313–320. Bizzozero, O. A., Pasquini, J. M., and Soto, E. F. (1982). DiVerential eVect of colchicine upon the entry of proteins into myelin and myelin related membranes. Neurochem Res. 7, 1415–1425. Black, M. M., and Baas, P. W. (1989). The basis of polarity in neurons. TINS 12, 211–214. Bo¨, L., Quarles, R. H., Fujita, N., Bartoszewicz, Z., Sato, S., and Trapp, B. D. (1995). Endocytic depletion of L-MAG from CNS myelin in quaking mice. J. Cell Biol. 131(6), 1811–1820. Bock, J. B., Matern, H. T., Peden, A. A., and Scheller, R. H. (2001). A genomic perspective on membrane compartment organization. Nature 409, 839–841. Braun, P. E., De Angelis, D., Shtybel, W. W., and Bernier, L. (1991). Isoprenoid modiWcation permits 2’, 3’-cyclic nucleotide 3’-phosphodiesterase to bind to membranes. J. Neurosci. Res. 30, 540–544. Braun, P. E., Sandillon, F., Edwards, A., Matthieu, J.-M., and Privat, A. (1988). Immunocytochemical localization by electron microscopy of 2’, 3’-cyclic nucleotide 3’-phosphodiesterase in developing oligodendrocytes of normal and mutant brain. J. Neurosci. 8, 3057–3066. Brockes, J. P., RaV, M. C., Nishiguchi, D. J., and Winter, J. (1980). Studies on cultured rat Schwann cells. III. Assays for peripheral myelin proteins. J. Neurocytol. 9, 67–77. Bunge, R. P., Bunge, M. B., and Eldridge, C. F. (1986). Linkage between axonal ensheathment and basal lamina production by Schwann cells. Annu. Rev. Neurosci. 9, 305–328. Bussey, H. (1996). Rho returns: Its targets in focal adhesions. Science 273, 203.

SUMMARY AND PERSPECTIVES

Caceres, J. F., and Kornblihtt, A. R. (2002). Alternative splicing: Multiple control mechanisms and involvement in human disease. Trends Genet. 18, 186–193. Carson, J. H., Worboys, K., Ainger, K., and Barbarese, E. (1997). Translocation of myelin basic protein mRNA in oligodendrocytes requires microtubules and kinesin. Cell Motil. Cytoskeleton 38, 318–328. Chamberlain, L. H., Burgoyne, R. D., and Gould, G. W. (2001). SNARE proteins are highly enriched in lipid rafts in PC12 cells: Implications for the spatial control of exocytosis. Proc Natl Acad Sci U. S. A 98, 5619–5624. Cole, N. B., and Lippincott-Schwartz, J. (1995). Organization of organelles and membrane traVic by microtubules. Curr. Opin. Cell Biol. 7, 55–64. Colman, D. R., Kreibich, G., Frey, A. B., and Sabatini, D. D. (1982). Synthesis and incorporation of myelin polypeptide into CNS myelin. J. Cell Biol. 95, 598–608. D’Urso, D., Brophy, P. J., Staugaitis, S. M., Gillespie, C. S., Frey, A. B., Stempak, J. G., and Colman, D. R. (1990). Protein zero of peripheral nerve myelin: Biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron 2, 449–460. De Angelis, D., and Braun, P. E. (1997). 2’, 3’-cyclic nucleotide 3’-phosphodiesterase binds to actin-based cytoskeletal elements in an isoprenylation independent manner. J. Neurochem. 67, 943–951. de Vries, H., and Hoekstra, D. (2000). On the biogenesis of the myelin sheath: Cognate polarized traVicking pathways in oligodendrocytes. Glycoconj. J 17, 181–190. Duncan, I. D., Lunn, K. F., Holmgren, B., Urba-Holmgren, R., and Brignolo-Holmes, L. (1992). The Taiep rat; a myelin mutant with an associated oligodendrocyte microtubulin defect. J. Neurocytol. 12, 870–874. Dustin, P. (1984). ‘‘Microtubules.’’ Springer-Verlag, Berlin. Ebersole, T. A., Chen, Q., Justice, M. J., and Artzt, K. (1996). The quaking gene product necessary in embryogenesis and myelination combines features of RNA binding and signal transduction proteins. Nat. Genet. 12, 260–265. Eldridge, C. F., Bunge, M. B., Bunge, R. P., and Wood, P. M. (1987). DiVerentiation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. J. Cell Biol. 105, 1023–1034. Euteneuer, U., and Schliwa, M. (1992). Mechanism of centrosome positioning during the wound response in BSC1 cells. J. Cell Biol. 116, 1157–1166. Farina, K. L., and Singer, R. H. (2002). The nuclear connection in RNA transport and localization. Trends Cell Biol. 12, 466–472. Fenster, S. D., Chung, W. D., Zhai, R., Cases-Langhoff, C., Voss, B., Garner, A. M., Kaempf, U., Kindler, S., Guldenfinger, E. D., and Garner, C. C. (2000). Piccolo, a zinc finger protein structually related to Bassoon. Neuron 25, 203–214. Filbin, M. T., Walsh, F. S., Trapp, B. D., Pizzey, J. A., and Tennekoon, G. I. (1990). The role of myelin P0 protein as a homophilic adhesion molecule. Nature 344, 871–872. Fischer, I., Konola, J., and Cochary, E. (1990). Microtubule associated protein (MAP1B) is present in cultured oligodendrocytes and co-localizes with tubulin. J Neurosci. Res. 27, 112–124. Gaisano, H. Y., Ghai, M., Malkus, P. N., Shev, E. E., Bouquillon, A., Bennett, M. K., and Trimble, W. S. (1996). Distinct cellular locations of the syntaxin family of proteins in rat pancreatic acinar cells. Mol. Biol. Cell 7, 2019–2027. Garner, C. C., Kindler, S., and GundelWnger, E. D. (2000). Molecular determinants of presynaptic active zones. Curr. Opin. Neurobiol. 10, 321–327. Gencic, S., and Hudson, L. D. (1990). Conservative amino acid substitution in the myelin proteolipid protein of jimpymsd mice. J. Neurosci. 10, 117–124. Gillespie, C. S., Trapp, B. D., Colman, D. R., and Brophy, P. J. (1990). Distribution of myelin basic protein and P2 mRNA’s in rabbit spinal cord oligodendrocytes. J. Neurochem. 54, 1556–1561. Gould, R. M., and Sinatra, R. S. (1981). Internodal distribution of phosphatidylcholine biosynthetic activity in teased peripheral nerve Wbres: An autoradiographic study. J. Neurocytol. 10, 161–167. Gow, A., Friedrich, V. L., Jr., and Lazzarini, R. A. (1994). Many naturally occurring mutations of myelin proteolipid protein impair its intracellular transport. J. Neursci. Res. 37, 574–583. Gow, A., Southwood, C. M., and Lazzarini, R. A. (1998). Disrupted proteolipid protein traVicking results in oligodendrocyte apoptosis in an animal model of Pelizaeus-Merzbacher disease. J. Cell Biol. 140, 925–934. Gravel, M., Peterson, J., Yong, V. W., Kottis, V., Trapp, B., and Braun, P. E. (1996). Overexpression of 2’, 3’-cyclic nucleotide 3’-phosphodiesterase in transgenic mice alters oligodendrocyte development and produces aberrant myelination. Molec. Cell. Neurosci. 7, 453–466. GrindstaV, K. K., Yeaman, C., Anandasabapathy, N., Hsu, S. C., Rodriguez-Boulan, E., Scheller, R. H., and Nelson, W. J. (1998). Sec6/8 complex is recruited to cell-cell contacts and speciWes transport vesicle delivery to the basal-lateral membrane in epithelial cells. Cell 93, 731–740. Hardy, R. J., Loushin, C. L., Friedrich, V. L., Jr., Chen, Q., Ebersole, T. A., Lazzarini, R. A., and Artzt, K. (1996). Neural cell type-speciWc expression of QKI proteins is altered in quaking viable mutant mice. J Neurosci. 16, 7941–7949. Heath, J. W., Inuzuka, T., Quarles, R. H., and Trapp, B. D. (1991). Distribution of P0 protein and the myelinassociated glycoprotein in peripheral nerves from Trembler mice. J. Neurocytol. 20, 439–449. Hirokawa, N. (1998). Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science 279, 519–526.

51

52

2. CELL BIOLOGY OF MYELIN ASSEMBLY

Hoek, K. S., Kidd, G. J., Carson, J. H., and Smith, R. (1998). hnRNP A2 selectively binds the cytoplasmic transport sequence of myelin basic protein mRNA. Biochemistry 37, 7021–7029. Hsueh, Y. P., Yang, F. C., Kharazia, V., Naisbitt, S., Cohen, A. R., Weinberg, R. J., and Sheng, M. (1998). Direct interaction of CASK/LIN-2 and syndecan heparan sulfate proteoglycan and their overlapping distribution in neuronal synapses. J Cell Biol. 142, 139–151. Hudson, L. D., Puckett, C., Berndt, J., Chan, J., and Gencic, S. (1989). Mutation of the proteolipid protein gene PLP in a human x chromosome-linked myelin disorder. Proc. Natl. Acad. Sci. USA 86, 8128–8132. Ikonen, E. (2001). Roles of lipid rafts in membrane transport. Curr. Opin. Cell Biol. 13, 470–477. Inoue, K., Osaka, H., Imaizumi, K., Nezu, A., Takanashi, J., Arii, J., Murayama, K., Ono, J., Kikawa, Y., Mito, T., ShaVer, L. G., and Lupski, J. R. (1999). Proteolipid protein gene duplications causing PelizaeusMerzbacher disease: Molecular mechanism and phenotypic manifestations. Ann. Neurol. 45, 624–632. Kagawa, T., Ikenaka, K., Inoue, Y., Kuriyama, S., Tsujii, T., Nakao, J., Nakajima, K., Aruga, J., Okano, H., and Mikoshiba, K. (1994). Glial cell degeneration and hypomyelination caused by overexpression of myelin proteolipid protein gene. Neuron 13, 427–442. Kamal, A., and Goldstein, L. S. (2000). Connecting vesicle transport to the cytoskeleton. Curr. Opin. Cell Biol. 12, 503–508. Kee, Y., Yoo, J. S., Hazuka, C. D., Peterson, K. E., Hsu, S. C., and Scheller, R. H. (1997). Subunit structure of the mammalian exocyst complex. Proc Natl Acad Sci U. S. A 94, 14438–14443. Kidd, G. J., Andrews, S. B., and Trapp, B. D. (1994). Organization of microtubules in myelinating Schwann cells. J. Neurocytol. 23, 801–810. Kidd, G. J., Andrews, S. B., and Trapp, B. D. (1996). Axons regulate the distribution of Schwann cell microtubules. J. Neurosci. 16, 946–954. Kidd, G. J., Brand, S. L., Low, S. H., Weimbs, T., and Trapp, B. D. (2002). Myelin protein targeting in MDCK cells. Mol. Biol. Cell 11, 489a. 2002. Kidd, G. J., Hauer, P. E., and Trapp, B. D. (1990). Axons modulate myelin protein messenger RNA levels during central nervous system myelination in vivo. J. Neurosci. Res. 26, 409–418. Kidd, G. J., and Trapp, B. D. (1995). Cytoskeletal organization of myelinating Schwann cells. In ‘‘Neuroglia’’ (H. Kettenmann and B. R. Ransom, eds.), pp. 634–649. Oxford University Press, New York. KroepX, J. F., and Gardinier, M. V. (2001). IdentiWcation of a basolateral membrane targeting signal within the cytoplasmic domain of myelin/oligodendrocyte glycoprotein. J Neurochem 77, 1301–1309. Kwon, S., Barbarese, E., and Carson, J. H. (1999). The cis-acting RNA traVicking signal from myelin basic protein mRNA and its cognate trans-acting ligand hnRNP A2 enhance cap-dependent translation. J Cell Biol. 147, 247–256. Lai, C., Brow, M. A., Nave, K.-A., Noronha, A. B., Quarles, R. H., Bloom, F. E., Milner, R. J., and SutcliVe, J. G. (1987). Two forms of 1B236/myelin-associated glycoprotein (MAG), a cell adhesion molecule for postnatal neural development, are produced by alternative splicing. Proc. Natl. Acad. Sci. USA 84, 4337–4341. Lang, T., Bruns, D., Wenzel, D., Riedel, D., Holroyd, P., Thiele, C., and Jahn, R. (2001). SNAREs are concentrated in cholesterol-dependent clusters that deWne docking and fusion sites for exocytosis. EMBO J 20, 2202–2213. Lee, C., Ferguson, M., and Chen, L. B. (1989). Construction of the endoplasmic reticulum. J. Cell Biol. 109, 2045–2055. Lisanti, M. P., LeBivic, A., Saltiel, A. R., and Rodriguez-Boulan, E. (1990). Preferred apical distribution of glycosyl-phosphatidylinositol (GPI) anchored proteins: A highly conserved feature of the polarized epithelial cell phenotype. J. Memb. Biol. 113, 155–167. LoPresti, P., Szuchet, S., Papasozomenos, S. C., Zinkowski, R. P., and Binder, L. I. (1995). Functional implications for the microtubule-associated protein tau: Localization in oligodendrocytes. Proc. Natl. Acad. Sci U. S. A 92, 10369–10373. Low, S. H., Chapin, S. J., Weimbs, T., Komuves, L. G., Bennett, M. K., and Mostov, K. E. (1996). DiVerential localization of syntaxin isoforms in polarized Madin-Darby canine kidney cells. Mol. Biol. Cell 7, 2007–2018. Lunn, K. F., Baas, P. W., and Duncan, I. D. (1997). Microtubule organization and stability in the oligodendrocyte. J Neurosci. 17, 4921–4932. Macklin, W. B., Gardinier, M. V., King, K. D., and Kampf, K. (1987). An AG-GG transition at a splice site in the myelin PLP gene in jimpy mice results in the removal of an exon. FEBS Lett. 223, 417–421. Madison, D. L., Krueger, W. H., Cheng, D., Trapp, B. D., and PfeiVer, S. E. (1999). SNARE complex proteins, including the cognate pair VAMP-2 and syntaxin-4, are expressed in cultured oligodendrocytes. J. Neurochem. 72, 988–998. Madison, D. L., Kruger, W. H., Kim, T., and PfeiVer, S. E. (1996). DiVerential expression of rab3 isoforms in oligodendrocytes and astrocytes. J Neurosci. Res. 45, 258–268. Malfroy, B., Bakhit, C., Bloom, F. E., SutcliVe, J. G., and Milner, R. J. (1985). Brain-speciWc polypeptide 1B236 exists in multiple molecular forms. Proc. Natl. Acad. Sci. U. S. A 82, 2009–2013. Martini, R., and Schachner, M. (1986). Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve. J. Cell Biol. 103, 2439–2448. Matteoni, R., and Kreis, T. E. (1987). Translocation and clustering of endosomes and lysosomes depends on microtubules. J. Cell Biol. 105, 1253–1265. Matter, K., and Mellman, I. (1994). Mechanisms of cell polarity: Sorting and transport in epithelial cells. Curr. Opin. Cell Biol. 6, 545–554.

SUMMARY AND PERSPECTIVES

McPhilemy, K., Mitchell, L. S., GriViths, I. R., Morrison, S., Deary, A. W., Sommer, I., and Kennedy, P. G. (1990). EVect of optic nerve transection upon myelin protein gene expression by oligodendrocytes: Evidence for axonal inXuences on gene expression. J. Neurocytol. 19, 494–503. Mellman, I., and Warren, G. (2000). The road taken: Past and future foundations of membrane traVic. Cell 100, 99–112. Minuk, J., and Braun, P. E. (1996). DiVerential intracellular sorting of the myelin-associated glycoprotein isoforms. J Neurosci. Res. 44, 411–420. Mirsky, R., Winter, J., Abney, E. R., Pruss, R. M., Gavrilovic, J., and RaV, M. C. (1980). Myelin-speciWc proteins and glycolipids in rat Schwann cells and oligodendrocytes in culture. J. Cell Biol. 84, 483–494. Moller, J. R., Yanagisawa, K., Brady, R. O., Tourtellotte, W. W., and Quarles, R. H. (1987). Myelin-associated glycoprotein in multiple sclerosis lesions: A quantitative and qualitative analysis. Ann. Neurol. 22, 469–474. Mostov, K., Apodaca, G., Aroeti, B., and Okamoto, C. (1992). Plasma membrane protein sorting in polarized epithelial cells. J. Cell Biol. 116, 577–583. Mostov, K. E., Verges, M., and Altschuler, Y. (2000). Membrane traVic in polarized epithelial cells. Curr. Opin. Cell Biol. 12, 483–490. Muller, R., Heinrich, M., Heck, S., Blohm, D., and Richter-Landsberg, C. (1997). Expression of microtubuleassociated proteins MAP2 and tau in cultured rat brain oligodendrocytes. Cell Tissue Res. 288, 239–249. Munro, T. P., Magee, R. J., Kidd, G. J., Carson, J. H., Barbarese, E., Smith, L. M., and Smith, R. (1999). Mutational analysis of a heterogeneous nuclear ribonucleoprotein A2 response element for RNA traVicking. J Biol. Chem. 274, 34389–34395. Nave, K.-A., Bloom, F., and Milner, R. J. (1987). A single nucleotide diVerence in the gene for myelin proteolipid protein deWnes the jimpy mutation in mouse. J. Neurochem. 49, 1873–1877. Nishiyama, A., Chang, A., and Trapp, B. D. (1999). NG2þ glial cells: A novel glial cell population in the adult brain. J. Neurolpathol. Exp. Neurol. 58, 1113–1121. Nishiyama, A., Dahlin, K., Prince, J., Johnstone, S., and Stallcup, W. B. (1991). Primary structure of NG2, a novel membrane-spanning proteoglycan. J. Cell Biol. 114, 359–371. Palacios, I. M., and Johnston, D. S. (2001). Getting the message across: The intracellular localization of mRNAs in higher eukaryotes. Annu. Rev. Cell Dev. Biol. 17, 569–614. Pedraza, L., Huang, J. K., and Colman, D. R. (2001). Organizing principles of the axoglial apparatus. Neuron 30, 335–344. Peng, X. R., Yao, X., Chow, D. C., Forte, J. G., and Bennett, M. K. (1997). Association of syntaxin 3 and vesicleassociated membrane protein (VAMP) with Hþ/K(þ)-ATPase-containing tubulovesicles in gastric parietal cells. Mol. Biol. Cell 8, 399–407. Peters, A., Palay, S. L., and Webster, H. d. (1991). ‘‘The Fine Structure of the Nervous System: Neurons and their Supporting Cells.’’ Oxford University Press, New York. PfeiVer, S. E., Warrington, A. E., and Bansal, R. (1993). The oligodendrocyte and its many cellular processes. Trends Cell Biol. 3, 191–197. Politis, M. J., Sternberger, N. H., Ederle, K., and Spencer, P. S. (1982). Studies on the control of myelinogenesis. IV. Neuronal induction of Schwann cell myelin-speciWc protein synthesis during nerve Wber regeneration. J. Neurosci. 2, 1252–1266. Quarles, R. H., Everly, J. L., and Brady, R. O. (1973). Myelin-associated glycoprotein: A developmental change. Brain Res. 58, 506–509. Rapaport, R. N., and Benjamins, J. A. (1981). Kinetics of entry of P0 protein into peripheral nerve myelin. J. Neurochem. 37, 164–171. Readhead, C., Schneider, A., GriViths, I., and Nave, K.-A. (1994). Premature arrest of myelin formation in transgenic mice with increased proteolipid protein gene dosage. Neuron 12, 583–595. Richter-Landsberg, C., and Gorath, M. (1999). Developmental regulation of alternatively spliced isoforms of mRNA encoding MAP2 and tau in rat brain oligodendrocytes during culture maturation. J. Neurosci. Res. 56, 259–270. Rodriguez-Boulan, E., and Gonzalez, A. (1999). Glycans in post-Golgi apical targeting: Sorting signals or structural props? Trends Cell Biol. 9, 291–294. Rogalski, A. A., and Singer, S. J. (1984). Associations of elements of the Golgi apparatus with microtubules. J. Cell Biol. 99, 1092–1100. Rothman, J. E., and Warren, G. (1994). Implications of the SNARE hypothesis for intracellular membrane topology and dynamics. Curr. Biol. 4, 220–233. Salzer, J. L., Holmes, W. P., and Colman, D. R. (1987). The amino acid sequences of the myelin-associated glycoproteins: Homology to the immunoglobulin gene superfamily. J. Cell Biol. 104, 957–965. Sato, C., Schriftman, M., and Larocca, J. N. (1986). Transport of sulfatides towards myelin eVect of colchicine monensin and calcium on their intracellular translocation. Neurochemistry International 9, 265–272. Sato, S., Quarles, R. H., and Brady, R. O. (1982). Susceptibility of the myelin-associated glycoprotein and basic protein to a neutral protease in highly puriWed myelin from human and rat brain. J. Neurochem, 39, 97–105. Scales, S. J., Chen, Y. A., Yoo, B. Y., Patel, S. M., Doung, Y. C., and Scheller, R. H. (2000). SNAREs contribute to the speciWcity of membrane fusion. Neuron 26, 457–464. Shapiro, L., Doyle, J. P., Hensley, P., Colman, D. R., and Hendrickson, W. A. (1996). Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17, 435–449. Simons, K., and Ikonen, E. (1997). Functional rafts in cell membranes. Nature 387, 569–572.

53

54

2. CELL BIOLOGY OF MYELIN ASSEMBLY

Simpson, P. B., and Armstrong, R. C. (1999). Intracellular signals and cytoskeletal elements involved in oligodendrocyte progenitor migration. Glia 26, 22–35. Singh, H., and PfeiVer, S. E. (1985). Myelin-associated galactolipids in primary cultures from dissociated fetal rat brain: Biosynthesis, accumulation, and cell surface expression. J Neurochem 45, 1371–1381. Snipes, G. J., Suter, U., Welcher, A. A., and Shooter, E. M. (1992). Characterization of a novel peripheral nervous system myelin protein (PMP-22/SR13). J. Cell Biol. 117, 225–238. Song, J., Goetz, B. D., Baas, P. W., and Duncan, I. D. (2001). Cytoskeletal reorganization during the formation of oligodendrocyte processes and branches. Mol. Cell Neurosci. 17, 624–636. Spiegel, I., Salomon, D., Erne, B., Schaeren-Wiemers, N., and Peles, E. (2002). Caspr3 and caspr4, two novel members of the caspr family are expressed in the nervous system and interact with PDZ domains. Mol. Cell Neurosci. 20, 283–297. Sprinkle, T. J. (1989). 2’, 3’-cyclic nucleotide 3’-phosphodiesterase, an oligodendrocyte-Schwann cell and myelinassociated enzyme of the nervous system. CRC Crit. Rev. Neurobiol. 4, 235–301. Sternberger, N. H., Quarles, R. H., Itoyama, Y., and Webster, H. d. (1979). Myelin-associated glycoprotein demonstrated immunocytochemically in myelin and myelin-forming cells of developing rats. Proc. Natl. Acad. Sci. USA 76, 1510–1514. Suter, U., Welcher, A. A., Ozcelik, T., Snipes, G. J., Kosaras, B., Francke, U., Billings-Gagliardi, S., Sidman, R. L., and Shooter, E. M. (1992). Trembler mouse carries a point mutation in a myelin gene. Nature 356, 241–244. Takai, Y., Sasaki, T., Tamaka, K., and Nakanishi, M. (1995). Rho as a regulator of the cytoskeleton. Trends in Biochem. Sci. 20, 227–231. Taylor, C. M., Coetzee, T., and PfeiVer, S. E. (2002). Detergent-insoluble glycosphingolipid/cholesterol microdomains of the myelin membrane. J Neurochem 81, 993–1004. Terada, N., Baracskay, K., Kinter, M., Melrose, S., Brophy, P. J., Boucheix, C., Bjartmar, C., Kidd, G., and Trapp, B. D. (2002). The tetraspanin protein, CD9, is expressed by progenitor cells committed to oligodendrogenesis and is linked to beta1 integrin, CD81, and Tspan-2. Glia 40, 350–359. Terasaki, M., Chen, L. B., and Fujiwara, K. (1986). Microtubules and the endoplasmic reticulum are highly interdependent structures. J. Cell Biol. 103, 1557–1568. TerBush, D. R., and Novick, P. (1995). Sec6, Sec8, and Sec15 are components of a multisubunit complex which localizes to small bud tips in Saccharomyces cerevisiae. J Cell Biol. 130, 299–312. Thyberg, J., and Moskalewski, S. (1985). Microtubules and the organization of the Golgi complex. Exp. Cell Res. 159, 1–16. Trapp, B. D. (1988). Distribution of the myelin-associated glycoprotein and P0 protein during myelin compaction in Quaking mouse peripheral nerve. J. Cell Biol. 107, 675–685. Trapp, B. D., Andrews, S. B., Cootauco, C., and Quarles, R. H. (1989a). The myelin-associated glycoprotein is enriched in multivesicular bodies and periaxonal membranes of actively myelinating oligodendrocytes. J. Cell Biol. 109, 2417–2426. Trapp, B. D., Andrews, S. B., Wong, A., O’Connell, M., and GriVin, J. W. (1989b). Co-localization of the myelinassociated glycoprotein and the microWlament components f-actin and spectrin in Schwann cells of myelinated Wbers. J. Neurocytol. 18, 47–60. Trapp, B. D., Bernier, L., Andrews, S. B., and Colman, D. R. (1988a). Cellular and subcellular distribution of 2’, 3’ cyclic nucleotide 3’ phosphodiesterase and its mRNA in the rat nervous system. J. Neurochem. 51, 859–868. Trapp, B. D., Hauer, P., and Lemke, G. (1988b). Axonal regulation of myelin protein mRNA levels in actively myelinating Schwann cells. J. Neurosci. 8, 3515–3521. Trapp, B. D., Itoyama, Y., Sternberger, N. H., Quarles, R. H., and Webster, H. d. (1981). Immunocytochemical localization of Po protein in Golgi complex membranes and myelin of developing rat Schwann cells. J. Cell Biol. 90, 1–6. Trapp, B. D., Kidd, G. J., Hauer, P. E., Mulrenin, E., Haney, C., and Andrews, S. B. (1995). Polarization of myelinating Schwann cell surface membranes: Role of microtubules and the trans-Golgi network. J. Neurosci. 15, 1797–1807. Trapp, B. D., Moench, T., Pulley, M., Barbosa, E., Tennekoon, G., and GriVin, J. W. (1987). Spatial segregation of mRNA encoding myelin-speciWc proteins. Proc. Natl. Acad. Sci. USA 84, 7773–7777. Trapp, B. D., Nishiyama, A., Cheng, D., and Macklin, W. (1997). DiVerentiation and death of premyelinating oligodendrocytes in developing rodent brain. J. Cell Biol. 137, 459–468. Trapp, B. D., and Quarles, R. H. (1982). Presence of the myelin-associated glycoprotein correlates with alterations in the periodicity of peripheral myelin. J. Cell Biol. 92, 877–882. Trapp, B. D., Quarles, R. H., and Suzuki, K. (1984). Immunocytochemical studies of quaking mice support a role for the myelin-associated glycoprotein in forming and maintaining the periaxonal space and periaxonal cytoplasmic collar of myelinating Schwann cells. J. Cell Biol. 99, 594–606. Tropak, M. B., Johnson, P. W., Dunn, R. J., and Roder, J. C. (1988). DiVerential splicing of MAG transcripts during CNS and PNS development. Mol. Brain Res. 4, 143–155. Tsukada, Y., and Kurihara, T. (1992). 2’, 3’-cyclic nucleotide 3’-phosphodiesterase: Molecular characterization and possible functional signiWcance. In ‘‘Myelin: Biology and Chemistry’’ (R. E. Martensen, ed.), pp. 449–480. CRC Press, Boca Raton, FL. Ueda, H., Levine, J. M., Miller, R. H., and Trapp, B. D. (1999). Rat optic nerve oligodendrocytes develop in the absence of viable retinal ganglion cell axons. J. Cell Biol. 146, 1365–1374. van Meer, G. (1989). Lipid traVic in animal cells. Annu. Rev. Cell Biol. 5, 247–275.

SUMMARY AND PERSPECTIVES

Vouyiouklis, D. A., and Brophy, P. J. (1995). Microtubule-associated proteins in developing oligodendrocytes: Transient expression of a MAP2c isoform in oligodendrocyte precursors. J Neurosci. Res. 42, 803–817. Warrington, A. E., Barbarese, E., and PfeiVer, S. E. (1993). DiVerential myelinogenic capacity of speciWc developmental stages of the oligodendrocyte lineage upon transplantation into hypomyelinating hosts. J. Neurosci. Res. 34, 1–13. Webster, H. d. (1971). The geometry of peripheral myelin sheaths during their formation and growth in rat sciatic nerves. J. Cell Biol. 48, 348–367. Weimbs, T., Low, S. H., Chapin, S. J., and Mostov, K. E. (1997). Apical targeting in polarized epithelial cells: There’s more aXoat than rafts. Trends Cell Biol. 7, 393–399. Wight, P. A., Duchala, C. S., Readhead, C., and Macklin, W. B. (1993). A myelin proteolipid protein-LacZ fusion protein is developmentally regulated and targeted to the myelin membrane in transgenic mice. J. Cell Biol. 123, 443–454. Winter, J., Mirsky, R., and Kadlubowski, M. (1982). Immunocytochemical study of the appearance of P2 in developing rat peripheral nerve: Comparison with other myelin components. J. Neurocytol. 11, 351–362. Wrabetz, L., Feltri, M. L., Quattrini, A., Imperiale, D., Previtali, S., D’antonio, M., Martini, R., Yin, X., Trapp, B. D., Zhou, L., Chiu, S. Y., and Messing, A. (2000). P0 glycoprotein overexpression causes congenital hypomyelination of peripheral nerves. J Cell Biol. 148, 1021–1034. Yin, X., Kidd, G. J., Wrabetz, L., Feltri, M. L., Messing, A., and Trapp, B. D. (2000). Schwann cell myelination requires timely and precise targeting of P0 protein. J Cell Biol. 148, 1009–1020. Yin, X., Peterson, J., Gravel, M., Braun, P. E., and Trapp, B. D. (1997). CNP overexpression induces aberrant oligodendrocyte membranes and inhibits MBP accumulation and myelin compaction. J. Neurosci. Res. 50, 238–247. Zigmond, S. H. (1996). Signal transduction and actin Wlament organization. Curr. Opin. Cell Biol. 8, 66–73.

55

C H A P T E R

3 The Transport, Assembly, and Function of Myelin Lipids Christopher M. Taylor, Cecilia B. Marta, Rashmi Bansal, and Steven E. Pfeiffer

INTRODUCTION Lipids are ‘‘the center, life and chemical soul of all bioplasm.’’ (Thudichum, 1884).

This dramatic, insightful statement, uttered well over 100 years ago, continues to gain importance for neuroscience in general, and myelin structure and function in particular. Since lipids constitute one-half of brain tissue dry weight, it is not surprising that lipid biochemistry and neurochemistry have evolved together. In fact, many complex lipids, including gangliosides, cerebrosides, sulfatides, and phosphoinositides, were Wrst discovered in brain, where they are highly enriched compared to other tissues. Traditionally, lipids have been assigned two functions in the body, as repositories of energy in storage fat (mainly triglycerides) and as structural components of cell membranes. Later, many messenger functions were recognized for both nonmembrane (e.g., steroid hormones, eicosanoids) and membrane lipids (e.g., inositides, phosphatidylcholine) that have important roles in signal transduction across biological membranes. Since the brain has almost no triglycerides, what is the primary function of lipid membrane components? Among the complex lipids, sphingolipids are both ‘‘modulators for transmembrane signaling and mediators for cellular interactions’’ (Hakomori, 1990). Sphingolipids form specialized structures, mediate cell-cell and cell-substratum interactions, modulate the behavior of cellular proteins and receptors, and participate in signal transduction (Merrill, 2002). Sphingolipids are virtually absent from mitochondria and very low in the endoplasmic reticulum (ER); in contrast, they constitute 20 to 35 mol % of plasma membrane lipids. Further polarization occurs in the plasma membrane. This can be observed in polarized epithelial cells; the ratios of sphingolipids:glycerophospholipids:sterols in the apical and basolateral membranes of intestinal epithelial cells are 38:29:33 and 19:56:25, respectively (van Meer and Lisman, 2002). In myelin, these ratios are 28:44:28 (Norton and Cammer, 1984; Morell et al., 1994; also see Table 3.1), and interesting diVerences may be expected within subcellular domains of myelinating cells and the myelin sheath (discussed later). Oligodendrocytes (OLs) produce dramatic amounts of myelin membrane estimated at as much as 5,000–50,000 mm2 of myelin surface area per cell per day during the period of active myelin assembly (PfeiVer et al., 1993). This must, of course, be accomplished by an equally impressive synthesis and transport of myelin lipids, in particular galactosylceramide and its sulfated analogue sulfatide. Generally, in most cells the delivery of newly synthesized lipids and proteins to the plasma membrane is balanced by internalization through endocytosis, resulting in a steady state amount of total plasma membrane. In contrast, in the assembly of myelin, this membrane recycling must be at least substantially attenuated to allow for the elongation and

Myelin Biology and Disorders, Volume 1

57

Copyright 2004, Elsevier (USA). All rights reserved.

58

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

TABLE 3.1 Composition of Rat CNS Myelin and Brain Substance Total Proteina

Myelin

Whole brain

29.5

56.9

Total Lipid

70.5

37.0

Cholesterol

27.3

23.0

Cerebroside

23.7

14.6

a

7.1

4.8

Total Galactolipid

Sulfatide

31.5

21.3

Total Phospholipid

44.0

57.6

a Protein and lipid Wgures in percent dry weight; all others in percent total lipid weight. For further quantitative analyses see Norton and Cammer, (1984).

growth of the membrane (Kra¨mer et al., 2001; Madison and PfeiVer, 1997; Madison et al., 1999; Watanabe et al., 1999; see also Trapp et al., Section I, Chapter 2 of this volume). Myelin provides a rich opportunity for investigating the roles of lipids in cell biology and physiology. The dry mass of both central nervous system (CNS) and peripheral nervous system (PNS) myelin is characterized by a high lipid (70 to 85%) and low protein (15 to 30%) content compared with other biological membranes where the proportion of proteins exceeds that of lipids (Fig. 3.1). The composition of brain myelin is largely conserved among mammalian species, although interesting diVerences have been noted; for example, rat myelin has less sphingomyelin than bovine or human myelin. Regional variations are also in evidence; for example, spinal cord myelin has a higher lipid-toprotein ratio than brain myelin from the same species and its lipid composition is distinct from that of gray matter (the interested reader is urged to further consult Morell and Quarles, 1999, for an excellent discussion of myelin lipids). Although there are no ‘‘myelin-speciWc’’ lipids per se, myelin is substantially enriched for certain lipids (Table 3.1). In particular, galactocerebroside (galactosylceramide, GalCer, GalC) is highly enriched in myelin. During active myelination, the concentration of galactocerebroside in brain is directly proportional to myelin accumulation, and the rate of cerebroside synthesis is a good measure of the rate of myelination (Muse et al., 2001; Singh and PfeiVer, 1985). Up to 20% of galactocerebroside in myelin is converted to a 3’-hydroxyl sulfate form, sulfatide. In view of the enrichment of galactocerebroside in oligodendrocytes and myelin, the idea that it may be essential for oligodendrocyte diVerentiation and for the structure and function of myelin seemed reasonable. However, a mouse deWcient in UDPgalactose:ceramide galactosyltransferase, an enzyme necessary for galactocerebroside and sulfatide biosynthesis, has surprisingly normal myelin (Coetzee et al., 1996a, 1996b, 1998a); nevertheless, signiWcant diVerences in structure and changes in axonal conduction velocity are present, in addition to alterations in oligodendrocyte diVerentiation (see the discussion that follows). A mouse deWcient in cholesterol, in contrast, experiences massive hypomyelination (Saher et al., 2002; also see the discussion that follows). Gangliosides, a diverse class of glycosphingolipids composed of a hydrophobic ceramide moiety modiWed by the addition of a hydrophilic carbohydrate chain with one or more sialic acids, constitute 0.1 to 0.3% of myelin lipids. Gangliosides are ubiquitously expressed, but are quantitatively enriched in the exoplasmic leaflet of the plasma membranes in the nervous system (Svennerholm, 1980). Myelin is greatly enriched relative to other brain membranes in monosialoganglioside GM1 and in certain species including human, sialosylgalactosylceramide GM4. Increasingly, the functional importance of gangliosides is being recognized. Myelin has large quantities of cholesterol and ethanolamine-containing plasmalogens as well as lecithin; in contrast, sphingomyelin is a relatively minor component. Additional complexities within families of lipids are possible through variations in ceramide alkyl chain length, hydroxylation, and desaturation. Finally, extraction of myelin with acidiWed organic solvents demonstrates the presence of polyphosphoinositides, including triphosphoinositides and diphosphoinositides (4 to 6% and 1 to 1.5%, respectively, of total myelin phos-

LIPID SYNTHESIS

FIGURE 3.1 Lipid and protein composition of myelin. Compared to other cell membranes, myelin has elevated levels of lipids, in particular glycosphingolipids. Cholesterol (Chol), phospholipids (PL), Glycosphingolipids (GSL). The most abundant proteins in myelin are proteolipid protein (PLP) and its alternative splicing isoform DM20, and myelin basic protein (MBP), which together account for approximately 80% of myelin proteins. See text and Table 3.1 for quantiWcation.

phorus). Quantitatively minor lipid components of myelin include at least three fatty acid esters of cerebroside and the glycerol-based galactosyldiglycerides diacylglycerylgalactoside and monoalkylmonoacylglycerylgalactoside. Some long chain alkanes are also present. Peripheral and central nervous system myelin lipids are qualitatively similar, but there are some quantitative diVerences. For example, compared to CNS myelin, PNS myelin has less cerebroside and sulfatide, considerably more sphingomyelin, and in some species the ganglioside LM1 (sialosyl-lactoneotetraosylceramide) (Ogawa-Goto et al., 1992). The following sections review some of the features of myelin lipid synthesis, structure and of particular interest, function. Clearly, myelin can no longer be seen as a relatively inert insulation material for axons that simply assists in nerve conduction. Rather, myelin is part of an intricate three-way communication network between the environment, the myelin sheath, and the axolemmal membrane (Menon et al., 2003). It is clear that lipids play crucial roles in these events.

LIPID SYNTHESIS Sphingolipids The biosynthesis of sphingolipids is carefully orchestrated both through transcriptional regulation and cellular compartmentalization. Not only are the Wnal products of these biosynthetic pathways of major importance in cellular physiology, but, in addition, many biosynthetic intermediates are themselves highly bioreactive. Finally, sphingolipids are being increasingly implicated in pathology, including Batten’s disease and lysosomal storage diseases such as Krabbe’s disease and MLD (Dawson and Cho, 2000). ed1

59

60

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

FIGURE 3.2 Biosynthesis of sphingolipids. Ceramide is synthesized from the condensation of palmitoylCoA and serine. Ceramide is highly biologically active itself and is also the common precursor of all classes of sphingolipids. Ceramide can be converted to galactosylceramide and sulfatide, catalyzed by ceramide galactosyltransferase (CGT) and sulfotransferase (CST), respectively; to sphingomyelin catalyzed by sphingomyelin synthase (5); to lactosylceramide (the precursor of all gangliosides except GM4) catalyzed by the sequential action of ceramide glucosyltransferase (GlcT) and galactosyltransferase I (GalTI), or degraded by ceramidase (6) to sphingosine, which can be reversibly converted to the signaling molecule sphingosine-1-phosphate by sphingosine kinase (7)/sphingosine 1-phosphate phosphatase (8). A to D indicate enzymes for which knockout mice have been developed: (A) Ceramide galactosyltransferase, (B) sulfotransferase, (C) ceramide glucosyltransferase, (D) galactosyltransferase I. Enzymes catalyzing sphingolipid biosynthesis are (1) serine palmitoyltransferase, (2) 3-ketosphinganine reductase, (3) (dihydro)ceramide synthase, (4) dihydroceramide desaturase, (5) sphingomyelin synthase, (6) ceramidase, (7) sphingosine kinase, (8) sphingosine-1phosphate phosphatase, and (9) sialyltransferase. Insert shows the structure of ceramide with the reactive head group to which sugars are added to the right, and the carbon chain that inserts into a leaXet of the lipid bilayer on the left.

The investigation of the synthesis and regulated transport of lipids continues to be a major area of focus, evidenced by a number of excellent recent reviews (Maier et al., 2001; Kolter et al., 2002; Merrill, 2002; Pomorski et al., 2001; Smith and Merrill, 2002; van Meer and Lisman, 2002). The basic principles appear to be common among cell types; a brief review of some of the salient points is provided here (IUPAC nomenclature is used in the following sections; Chester, 1998). A striking feature of sphingolipid synthesis is the subcellular compartmentalization of individual metabolic events. Sphingolipid synthesis is initiated in the ER with the condensation of L-serine and palmitoyl-CoA to 3-ketosphinganine (Serine Palmitoyltransferase), followed by a reduction of this product to the sphingoid base sphinganine (Fig. 3.2). Mutations in Serine Palmitolytransferase-1 can lead to hereditary sensory neuropathy

LIPID SYNTHESIS

type I, the most common hereditary disorder of peripheral sensory neurons (Bejaoui et al., 2001; Dawkins et al., 2001); alterations of de novo sphingolipid synthesis, such as occurs in response to fumonisins that target (Dihydro)ceramide Synthase, can lead to a variety of lipotoxic diseases (Unger, 2002). Sphinganine is then modiWed on the cytosolic side of the ER by the addition of a fatty acid [from fatty acyl-CoA; (Dihydro)ceramide Synthase to form dihydroceramide. This is followed by the formation of a double bond in the sphinganine carbon chain (Dihydroceramide Desaturase, with NADPH) to form ceramide, a highly biologically active molecule (Hannun and Obeid, 2002) implicated in a variety of critical cell physiology, including stress responses, senescence, and apoptosis. SpeciWc targets of ceramide action are being identiWed, such as the ceramide-activated protein phosphatases—for example, PP1 and PP2A (Chalfant and Hannun, 2002), cathepsin D (Heinrich et al., 2000), and others (Hannun and Obeid, 2002). Ceramide also can be formed by the breakdown of complex sphingolipids through the activity of sphingomyelinases and glycosidases. Some of these activities may depend on the inclusion of ceramide into lipid rafts (Kolesnick, 2002; also see the discussion that follows). Ceramide is the common precursor for all major glycosphingolipids (Hannun and Obeid, 2002) and sphingomyelin by the addition of either a sugar (Ceramide Glucosyltransferase, Ceramide Galactosyltransferase), or phosphocholine (Sphingomyelin Synthase), respectively (Figs. 3.3 and 3.4; also see the discussion that follows) (Ichikawa and Hirabayashi, 1998). In addition, ceramide can be converted to sphingosine by removal of the fatty acid chain (Ceramidase). Sphingosine in turn can be converted reversibly to the signaling molecule sphingosine-1-phosphate (Sphingosine Kinase/Sphingosine-1-Phosphate Phosphatase) (Fig. 3.2) (Spiegel and Milstien, 2002), both an intracellular mediator and an extracellular agonist for the Edg family of G-protein receptors (S1PRs). Of the Wve known S1P receptors, S1P5/Edg-8 is expressed primarily by oligodendrocytes or Wbrous astrocytes in rat brain (Im et al., 2000), where it mediates anti-proliferative eVects (Spiegel and Milstien, 2002). Ceramide is used for the synthesis of galactocerebroside (Fig. 3.2, Fig. 3.3) after being flipped to face the lumen of the ER (Kolter et al., 2002; van Meer and Lisman, 2002). This is catalyzed by Ceramide Galactosyltransferase, a type I transmembrane protein (in contrast to most glycosyltransferases that are type II transmembrane proteins) containing a C-terminal ER retention signal (Schulte and StoVel, 1993; Sprong et al., 1998). Subsequently, in the lumen of the Golgi, a portion of the GalC is converted to GalC-3-sulfate (‘sulfatide’) by the addition of sulfate activated as PAPS (3’-phosphoadenosine 5’phosphosulfate) to the 3’-OH group in the galactose residue, catalyzed by a speciWc Sulfotransferase (Fig. 3.3) (Tennekoon et al., 1983; Sundaram and Lev, 1992; Tennekoon et al., 1983). Both GalC and sulfatide are highly enriched in oligodendrocytes, Schwann cells and myelin and, as discussed later, serve as key markers for oligodendrocyte diVerentiation and may serve a signaling function. Early work indicated that although sulfatide is transported to the plasma membrane by vesicular transport through the Golgi, GalC can bypass the Golgi to reach the plasma membrane (Brown et al., 1993; Townsend et al., 1984). Mice deWcient in either GalC and sulfatide (Ceramide Galactosyltransferase-null, Fig. 3.2A; Coetzee et al., 1996a; Coetzee et al., 1996b) or sulfatide alone (Sulfotransferasenull, Fig. 2B; Honke et al., 2002) show enhanced diVerentiation of oligodendrocytes, are able to produce substantial amounts of myelin that nevertheless develop signiWcant defects with age, and have serious defects in myelin-axolemmal interactions at the paranodal junctions (discussed later). Finally, less polar, acylated derivatives of GalC, 3-O-acetylsphingosine GalC with nonhydroxy or hydroxy fatty N-acylation, are also abundant in myelin (Dasgupta et al., 2002). After its biosynthesis on the cytosolic side of ER membranes, ceramide is also transferred to the Golgi (Fig. 3.3). The transport of ceramide both across and among membranes is a complex topic of current investigation, one that appears to involve both vesicular and nonvesicular transport mechanisms, possibly through the intervention of glycolipid transfer proteins (Fukasawa et al., 1999; Funakoshi et al., 2000; Funato and Riezman, 2001; Kok et al., 1998; Lin et al., 2000; van Meer and Lisman, 2002; Yasuda et al., 2001). The participation of one or more members of the ATP-binding cassette (ABC)

61

62

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

FIGURE 3.3 Topography of glycosphingolipid synthesis. Glycosphingolipid synthesis begins with the formation of ceramide (narrow end of trapezoid indicates head group of ceramide) in the endoplasmic reticulum (ER). Ceramide can be Xipped from the outer leaXet of the ER bilayer to the inner leaXet (see also Fig. 3.11) where galactose is added in the lumen to form GalC. GalC is then shuttled (via vesicular transport) to the cis Golgi. GalC can be converted to sulfatide (SUL) in the lumen of the trans Golgi. Ceramide is shuttled (via vesicular transport) to the cis Golgi where glucose can be added to the ceramide head group on the cytosolic side of the membrane, producing GlcC. GlcC can be Xipped from the outer leaXet of the Golgi to the inner leaXet where a galactose is added in the lumen to form LacC. LacC can then be modiWed by the addition of sugars to form complex gangliosides in the lumen of the Golgi (see Fig. 3.4). Alternatively, ceramide can be Xipped from the outer leaXet of the Golgi to the inner leaXet where a phosphatidylcholine (PC) is added in the lumen of the Golgi to form sphingomyelin (SM). Lipids are then shipped to the plasma membrane (PM) via vesicular transport. There is evidence that some lipids, such as ceramide, may translocate from the ER to the Golgi by a nonvesicular mechanism (see text).

family of transmembrane transporters of endogenous lipids seems likely (Borst et al., 2000; Broccardo et al., 1999; Raggers et al., 2000). In particular, myelin-forming oligodendrocytes express one such transporter, the 260 kDa ABCA2 protein, in a developmentally regulated manner (Tanaka et al., 2003). In addition, there are a plethora of so-called flippases, translocases, and scramblases that participate in the transport of lipids across membranes (Pomorski et al., 2001).

LIPID SYNTHESIS

FIGURE 3.4 Biosynthesis of gangliosides. Lactosylceramide is converted to the diVerent types of gangliosides, catalyzed by sialyltransferases (SAT) I (E), II (F), III, IV and V, and galactose N-acetyltransferase (GalNAcT) (G), and galactosyltransferase II (GalT II). Gangliosides are abbreviated using Svennerholm’s nomenclature. Ganglio series 0-, a-, b-, and c- are indicated. E to G indicate enzymes for which knockout mice have been developed. Ceramide is represented by a trapezoid (the narrow end indicates the reactive head group of ceramide).

Ceramide is then converted on the cytosolic side of the Golgi to glucocerebroside (GlcCer) by the transfer of a glucose residue from UDP-glucose to ceramide, catalyzed by the type III transmembrane protein Glucosyltransferase (Fig. 3.2, Fig. 3.3) (Ichikawa et al., 1996; Paul et al., 1996). In addition to glycosphingolipid synthesis, ceramide glucosylation may serve a function in eVect by removing excess cellular ceramide (Ichikawa and Hirabayashi, 1998). A critical next step is the translocation of GlcCer across the Golgi membrane, probably by an energy-dependent translocator (Burger et al., 1996; Lannert et al., 1998) or by the MDR1 P-glycoprotein, an ATP-binding cassette transporter involved in multidrug resistance (Fig. 3.3; Lala et al., 2000). In the lumen of the Golgi, LacCer is synthesized by the transfer of a galactose residue from UDP-galactose to GlcCer (Fig. 3.2, Fig. 3.3; Galactosyltransferase I; Lannert et al., 1994). The glycosyltransferases require nucleotide-activated sugars that are transported into the lumen of the organelles by antiporters (Hirschberg et al., 1998). Ceramide is converted to sphingomyelin (Sphingomyelin Synthase) by the addition of phosphocholine in the lumen of the Golgi (Fig. 3.3). It appears that as much as 50% of sphingomyelin also may be synthesized at the level of the plasma membrane (Vos et al., 1995). With the exception of GM4 (3’-sialyl-GalCer) (Fig. 3.2) (Ledeen et al., 1973), an important component of myelin, the complex acidic gangliosides are synthesized by the stepwise addition of sugar and sialic acid residues to LacCer to produce the hematoside GM3, GD3 and GT3 (Fig. 3.4). These in turn serve as precursors for the synthesis of the complex gangliosides of the a-, b-, and c-series with one, two, or three sialic acid residues linked to the 3-position of the inner galactose moiety, respectively. The transferases that catalyze the formation of LacCer, GM3, and GD3 are highly substrate speciWc.

63

64

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

In contrast, the enzymes catalyzing the subsequent stepwise glycosylations are much less speciWc, and as a result only a few glycosyltransferases are required. As noted by Kolter and SandhoV (Cane et al., 1998; Kolter et al., 2002), this synthetic pathway occurs by a process of ‘‘combinatorial biosynthesis.’’ In spite of the potential for a dramatic variety of products, in fact, only a few glycolipid ‘‘series’’ are produced. It appears that many of the glycosyltransferases that catalyze this stepwise transfer of monosaccharide sugars from sugar nucleotides to speciWc glycolipid acceptors are part of a protein complex co-localized in the Golgi (Giraudo et al., 2001; Bieberich et al., 2002; Kapitonov and Yu, 1999). The activities of these glycosyltransferases are believed to be regulated by changes in phosphorylation state by protein kinases (Ariga and Yu, 1998; Yu and Bieberich, 2001). Complex glycosphingolipids are unable to return to the cytosolic face of the Golgi, requiring vesicular transport in order to exit the Golgi. It is by this route that most of the major gangliosides and sphingolipids are sorted to the exoplasmic leaflet of the plasma membrane (Farrer and Benjamins, 1992). Nevertheless, there is evidence that some of the cytosolic GlcCer can be transported to the cytosolic leaflet of the plasma membrane by a ‘‘direct,’’ nonvesicular route (Fig. 3.3). In contrast to the membrane-bound nature of ganglioside synthesis, glycosidases responsible for ganglioside catabolism are soluble enzymes that act on membrane-bound substrates and require activator proteins (Kolter et al., 1999). In myelin, a membrane-bound sialidase can modify the gangliosides (Yu, 1994; Riboni et al., 1997). The functional signiWcance of gangliosides is under consideration through the application of genetically engineered mice with defects in speciWc biosynthetic pathways (Figs. 3.2, 3.4) (Furukawa et al., 2001; Horinouchi et al., 1995; Kolter et al., 2002). Mice deWcient in Ceramide Glucosyltransferase (GlcT; Fig. 3.2C) begin to die as early as embryonic day 7.5, indicating that GlcC and higher-order glycolipids are important for development (note, however, that the accumulation of ceramide in such animals could also lead to deleterious eVects). Subsequently, strains of mice have been developed that are deWcient in speciWc classes of gangliosides found especially in the brain. As shown in Figures 3.2 and 3.4, mice also have been developed that are deWcient in Galactosyltransferase I (Fig. 3.2D: GalTI), Sialyltransferase I (Fig. 3.4E: SAT I), Sialyltransferase II (Fig. 3.4F: SAT II), and GalNActransferase (Fig. 3.4G: GalNAcT). GalNAcT-null animals lacking the major gangliosides GM2, GD2, GM1a, GD1b, GD1a, GT1a, GT1b, and GQ1b, and SAT II-null (Fig. 3.4F) animals lacking GD3, GD2, GD1b, GT1b, and GQ1b had only subtle impairment of brain function. However, in the case of GalNAcT-null mice this included axonal degeneration and defects in myelination (Sheikh et al., 1999). SAT II-null mice develop normally in spite of the absence of GD3 and the b-series gangliosides; careful investigation of myelin has apparently not yet been undertaken with most of these mutants. Double GalNAcT and SAT II null mice have been prepared by cross breeding to produce mice with only GM3; these mice are prone to sound-induced lethal seizures (Kawai et al., 2001). SAT I-null (Fig. 3.4E) mice are unable to form GM3, and thus none of the ganglio-series gangliosides present in the nervous system (0-series gangliosides are of course still formed); these mice experience altered insulin-mediated regulation (Tagami et al., 2002). Finally, a mouse doubly deWcient in GalNAcT and SAT I has been produced that cannot form any glycolipid of the ganglio series and that have highly elevated brain levels of LacCer; the investigation of these mice is underway (Kolter et al., 2002). Myelin glycosphingolipids may also be involved in the traYcking and formation of lipid rafts, in which glycosphingolipids play major roles along with cholesterol and sphingomyelin (see Maier et al., 2001, the discussion that follows, and the chapter by Trapp et al., Section I, Chapter 2 of this volume). Fluorescent analogues of ceramide have been useful tools for the study of lipid metabolism and transport along the exocytic pathway, and to assess the extent and mode of membrane internalization by endocytosis, of many cell types. This approach has been applied successfully to the analyses of in vivo lipid metabolism in adult rat CNS myelin (Di Biase et al., 1991), oligodendrocytes (Vos et al., 1995; Watanabe et al., 1999), and in myelinating neuron-Schwann cell co-cultures (Bilderback et al., 1997). These studies demonstrate that some internalization does occur in oligodendrocytes, at least in culture,

LIPID SYNTHESIS

and have provided evidence for diVerential internalization and sorting of speciWc lipids. For example, upon inserting fluorescently-labeled analogues of lactosylceramide (LacCer) and sulfatide into the plasma membrane of mature oligodendrocytes in culture, there is a marked diVerence in their internalization, with LacCer being relatively poorly internalized compared to sulfatide.

Cholesterol Cholesterol constitutes about 30% of myelin lipids (Tab. 3.1, Fig. 3.1) (Benjamins and Smith, 1984). The prevailing view has been that cholesterol is poorly imported into the brain, and therefore it must be synthesized by oligodendrocytes (Jurevics and Morell, 1995). Cholesterol is synthesized in the ER and transported through the Golgi to the plasma membrane to form a gradient of cholesterol concentration, lowest in the ER (only about 0.5 to 1% of cellular cholesterol) and highest in the plasma membrane (60 to 80% of cellular cholesterol; Liscum and Munn, 1999). Not surprisingly, the concentration of cholesterol in myelin is high. The synthesis of cholesterol levels in cells must be tightly regulated insofar as cholesterol is toxic. This is accomplished by the transcriptional regulation of enzymes involved in the cholesterol synthesis pathway (see Fig. 3.5), cellular uptake, and storage of cholesterol in esteriWed form, and cellular eVlux (Simons and Ikonen, 2000). Many of these regulatory modes are sensitive to the concentration of cholesterol in the ER via feedback mechanisms. The mechanisms by which cholesterol is transported in the cell are not well understood and may include a combination of vesicular transport, cytoplasmic cholesterol carriers, and membrane-membrane contact (MaxWeld and Wustner, 2002). Cholesterol is a critical element in the assembly and integrity of myelin. For example, patients with phenylketonuria (PKU), in which phenylalanine (Phe) levels are signiWcantly elevated, exhibit white matter abnormalities, including hypomyelination of untreated individuals and demyelination in some patients even upon treatment coupled with axonal pathology (Dyer et al., 1996; Huttenlocher, 2000). These lesions correlate in size and number with the level of Phe in the blood (Thompson et al., 1993). Both Phe-sensitive and -insensitive oligodendrocytes have been reported (Dyer et al., 1996). An intriguing insight comes from studies indicating that elevated Phe levels moderately inhibit 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase, the key regulatory enzyme in the cholesterol biosynthetic pathway (Fig. 3.5) (Shefer et al., 2000). It may be that insuYcient or damaged myelin is the primary pathological event in PKU, and that this secondarily causes neuronal dysfunction (Dyer, 2000) by disruption of the myelinaxolemmal complex. This view is strongly supported by the recent development of a knockout of the squalene synthase gene (see Fig. 3.5; carefully chosen to be downstream of farnesyl pyrophosphate, thus leaving unperturbed several other important biochemical pathways) that, using crelox technology, is restricted to myelin forming cells (Saher et al., 2002). Unlike a general knockout of this enzyme that resulted in embryonic lethality (Tozawa et al., 1999), these mice survive, but they develop tremors at about 10 postnatal days and have spinal cords nearly devoid of myelin. Reduced, but present, levels of myelin in the subcortical white matter and cortex are attributed to exogenous uptake of cholesterol. This mutant once again emphasizes the critical importance of lipids for myelin biogenesis, integrity and function, insofar as it produces a pathological phenotype far more advanced than that of most knockouts targeting myelin proteins.

Phospholipids Besides their role as structural components of biomembranes, phospholipids have a very interesting function as signal transduction messengers, such as diacylglycerol (DAG) and arachidonic acid, and phosphoinositides, which have been implicated as messengers for Ca2þ homeostasis and protein phosphorylation. However, in contrast with other brain

65

66

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

FIGURE 3.5 Biosynthesis of cholesterol. Enzymes catalyzing each step are (1) HMG-CoA reductase, (2) mevalonate-5-phosphotransferase, (3) phosphomevalonate kinase, (4) pyrophosphomevalonate decarboxylase, (5) prenyl transferase, (6) prenyl transferase, (7) squalene synthase, (8) squalene epoxidase, and (9) squalene oxidocyclase.

lipids, they do not appear to have distinct roles in myelin compared with other biological membranes. The pathways for phospholipid biosynthesis are illustrated in Figure 3.6. Briefly, dihydroxyacetone phosphate (DHAP) is reduced to glycerol-3-phosphate (Gro-3P) and then doubly acylated to form phosphatidic acid (PtdOH). DHAP can also be acylated then alkylated to form alkyl DHAP in which the acyl group of acyl DHAP is replaced by a long chain alcohol. This alkyl ether is the precursor for the ether lipids, including plasmalogens, of which ethanolamine plasmalogens are major lipids of myelin (see AgranoV et al., 1999). PtdOH can be hydrolyzed to form DAG, the precursor of phosphatidyl-serine (PtdSer), -ethanolamine (PtdEth), and –choline (PtdCho), or it can be converted into the liponucleotide CDP-DAG (CMP-PtdOH) that reacts with inositol to form phosphatidylinositol (precursor of phosphatidyl-inositol phosphate) or with Gro-3P, which is then converted to cardiolipin. In general, enzymes catalyzing the biosynthesis of phospholipids are bound to the ER, with the exception of those catalyzing the synthesis of cardiolipin and ether lipids localized in the mitochondria and peroxisomes, respectively.

LIPID SYNTHESIS

67

FIGURE 3.6 Biosynthesis of phospholipids. Dihydroxyacetone phosphate (DHAP) can be reduced to glycerophosphate (Gro-3P) or acylated (Acyl DHAP) to serve as a precursor for the synthesis of ether lipids. Phosphatidic acid (PtdOH) is converted to diacyl glycerol (DAG), which is then converted into phosphatidylcholine (PtdCho) and phosphatidylethanolamine (PtdEtn). PtdCho and PtdEtn can be interconverted and also converted to phosphatidylserine (PtdSer). PtdOH is also converted to the liponucleotide CDP-DAG (CMP-PtdOH), which can be converted to cardiolipin or inositides (PtdIns). The enzymes catalyzing phospholipid biosynthesis are (1) glycerophosphate dehydrogenase, (2) dihydroxyacetone phosphate acyltransferase, (3) sn-glycerol-3-phosphate acyltransferase, (4) acyl/alkyl dihydroxyacetone phosphate reductase, (5) alkyl dihydroxyacetone phosphate synthase, (6) 1-acyl glycerol-3-phosphate acyltransferase, (7) phosphatidate phosphohydrolase, (8) diacylglycerol cholinephosphotransferase, (9) diacylglycerol ethanolaminephosphotransferase, (10) phosphatidylethanolamine N-methyl transferase and phosphatidyl-N-methylethanolamine Nmethyl transferase, (11) phosphatidylethanolamine:serine transferase, (12) phosphatidylserine decarboxylase, (13) phosphatidate cytidyltransferase, (14) phosphatidylinositol synthase, (15) CDP-DAG:glycerol-3-phosphate phosphatidyltransferase, (16) phosphatidylglycerol phosphatase, (17) cardiolipin synthase, (18) phosphatidylinositol-4-kinase, (19) phosphatodylinositol-4-phosphate 5-kinase, and (20) phosphatidylcholine:ceramide cholinephosphotransferase. ModiWed from AgranoV et al., 1999.

Turnover of Lipids in Myelin (Morell and Quarles, 1999) In general, lipid turnover in brain is measured by injecting a radioactive metabolic precursor and following the loss of radioactivity as a function of time. However, reutilization of the precursor has to be considered in order to avoid misinterpretations of lipid halflife. For example, acetate and glycerol are both precursors of glycerolipids, however, radioactive acetate, but not glycerol, can be reutilized after these lipids are degraded. In the case of myelin, another complication is that the rate of metabolic turnover of newly synthesized proteins and lipids is dependent on the age of the animal; radioactive precursors incorporated into myelin of young animals during the period of rapid myelin formation are more stable than when incorporated into older animals, for which newly

68

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

FIGURE 3.7 Model of oligodendrocyte lineage. Commonly used stage-speciWc markers for each in stage oligodendrocyte development are indicated, with an emphasis on lipid antigens. Antibodies identifying some markers are shown in parentheses. Also indicated are three stages at which lineage progression can be reversibly inhibited (e.g., R-mAb blocks progression at the Pre-GalC stage) and periods of migratory and proliferative ability. ModiWed from ‘‘Trends in Cell Biology,’’ Vol. 3, S. E. PfeiVer, A. E. Warrington, and R. Bansal, ‘‘The Oligodendrocyte and Its Many Cellular Processes,’’ pp. 191–197, Copyright (1993), with permission from Elsevier Science.

synthesized lipids apparently enter a metabolically unstable compartment. For example, glycerophospholipids have a half-life of approximately 1 month in young animals and a week or less older animals. Cholesterol, GalC and sulfatide are even more stable in young animals, with half-lives of several months.

ROLE OF GLYCOSPHINGOLIPIDS IN THE REGULATION OF THE INTERFACE BETWEEN OLIGODENDROCYTE PROLIFERATION/TERMINAL DIFFERENTIATION Oligodendrocytes are the end product of a well-characterized developmental lineage distinguished by distinct phenotypic stages under the control of both intrinsic and extrinsic factors (PfeiVer et al., 1993). Mechanisms that regulate lineage progression are of considerable current general interest, including the roles of critical growth factors and the regulation of cell cycle progression (reviewed in Casaccia-BonneWl and Liu, 2003). Antibodies against speciWc glycolipids have been cornerstones in the identiWcation of speciWc stages of the oligodendrocyte lineage in culture and in vivo both in normal and remyelinating animal models (Bansal et al., 1989; Hardy and Friedrich, 1996; PfeiVer et al., 1993; Warrington and PfeiVer, 1992; Warrington et al., 1993; and Fig. 3.7). They continue to be highly reliable markers for this type of study. For example, antisera to the ganglioside GD3, GT3 and its O-acylated homologue are among the most important markers of the early progenitor stage (O-2A) (Eisenbarth et al., 1979). The next stage of the lineage, the pro-oligodendroblast (Pro-OL) or late progenitor stage, is identiWed by cellular reactivity with O4 and A007, IgM monoclonal antibodies (Sommer and Schachner, 1981, 1982) that recognize the surface sulfated antigen prooligodendroblast antigen (POA) (Bansal et al., 1992; Gard and PfeiVer, 1990; Knapp, 1991), in the absence of staining with R-mAb or O1 (discussed later). The biochemical identity of POA is uncertain. Although these antibodies also react with sulfatide (discussed later), Pro-OLs do not synthesize sulfatide (Bansal and PfeiVer, 1994a). Therefore, POA must be another molecule. However, when sulfation is blocked, either biochemically (Bansal and PfeiVer, 1994a) or genetically (Bansal et al., 1999), staining of Pro-OLs by O4/A007 is eliminated, supporting the view that POA is sulfated.

ROLE OF GLYCOSPHINGOLIPIDS

In particular, the decision to cease proliferation and initiate terminal diVerentiation is a critical point in cellular diVerentiation (Fig. 3.7). Here is the knife-edge of the decision of whether to continue life as a proliferative progenitor, or rather to cease proliferation and become a mature myelin-producing cell. Current views suggest that these two processes, while temporally related and often considered to be opposite sides of a coin, may in fact be under separate control mechanisms. In the oligodendrocyte lineage this decision is made at the Pre-GalC stage of the lineage, recognized by immunoreaction with the IgG3 monoclonal antibody R-mAb that reacts with both GalC and sulfatide (Bansal et al., 1989; Ranscht et al., 1982). This stage, with the immunophenotype R-mAbþ/O1
(Bansal and PfeiVer, 1992) is situated between the late progenitor stage (O4þ/O1
) and the immature oligodendrocyte stage (O1þ/MBP
). Glycosphingolipids appear to play an important role in managing this transition (discussed later). The onset of terminal diVerentiation (Immature OL stage) is identiWed by the synthesis and transport of GalC and sulfatide to the surface. GalC was one of the earliest markers to be used to identify OLs in a seminal paper by RaV et al. (RaV et al., 1978; RaV et al., 1979) that established the utility of using antisera against speciWc glycolipids as markers for oligodendrocytes. The staining of these cells is routinely carried out using the IgM monoclonal antibodies O1 (Sommer and Schachner, 1981), which identiWes GalC, and O4/A007 (noted earlier and see Figs. 3.8A and 3.8B), which identify sulfatide at this stage in the lineage. Later stages are further identiWed by the sequential addition of myelin speciWc proteins, against a continuing reactivity with O1, O4, and R-mAb. It has been proposed that glycosphingolipids, in particular GalC and sulfatide, play functional roles in the regulation of OL terminal diVerentiation by acting as sensors/ transmitters of environmental information (Bansal et al., 1988; Bansal and PfeiVer, 1989, 1994b). A number of lines of evidence strongly support this idea (Boggs and Wang, 2001; Diaz et al., 1978; Dorfman et al., 1979; Dyer and Benjamins, 1988a; Dyer and Benjamins, 1990, 1991; Jungalwala, 1994; Ranscht et al., 1987). First, GalC and sulfatide are synthesized and transported to the outer leaflet of the OL plasma membrane at this developmental interface (Fig. 3.7) (Hardy and Reynolds, 1991; PfeiVer et al., 1993; RaV et al., 1978). Second, exposure of OL progenitors to anti-GalC/ sulfatide (R-mAb) or anti-sulfatide (O4) antibodies (but not anti-GalC [O1]; also, not antisurface antigen 1A9, -HNK1, -cholesterol, or -NCAM antibodies) leads to the arrest of OL lineage progression, precisely at this point (Fig. 3.7) (Bansal and PfeiVer, 1989; Bansal et al., 1999). This block is eVected at the level of mRNA, and is complete, nontoxic, and rapidly and accurately reversible; upon removal of the antibody, the cells resume normal lineage progression into terminal diVerentiation in #12h (Krueger et al., 1997). Unlike the inhibition of progenitor diVerentiation by FGF2 (bFGF), R-mAb blocked cells are poorly proliferative. It is thought that R-mAb and O4 may mimic an endogenous ligand that is part of the regulatory machinery of OL diVerentiation. There is ample precedence for antibody-mimicking of normal ligands, resulting in transmembrane signaling (reviewed in Dyer, 1993): for example, anti-integrin activates integrin signaling (Coppolino et al., 1995); anti-EGF and -insulin receptor induce eVects normally mediated by their ligands (Jacobs et al., 1978; Schreiber et al., 1981); anti-b-adrenergic receptor stimulates adenyl cyclase (Couraud et al., 1981); anti-syndecan-1 induces clustering of syndecan-1 and reorganization of actin Wlaments (Carey et al., 1994); and anti-GM1 and -CD59 (NCAM) induce an accumulation of actin and tyrosine-phosphorylated proteins in glycosphingolipid clusters (Harder and Simons, 1999). It seems likely that sulfatide is the key molecule in this regulation of diVerentiation by glycosphingolipids since, as indicated earlier, treatment of enriched cultures of OLs with anti-sulfatides (O4, R-mAb), not anti-GalC (O1), inhibits OL terminal diVerentiation (Bansal et al., 1999). Further, negatively charged sulfatide (but not GalC) interacts with key adhesion and extracellular matrix proteins (e.g., tenascin-R/janusin/J-1, laminin and thrombospondin are secreted by OLs or astrocytes and bind to sulfatide; Pesheva et al., 1997; Vos et al., 1994). Finally, treatment of Schwann cells in culture with R-mAb results in inhibition of myelin formation (Owens and Bunge, 1990).

69

70

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

FIGURE 3.8 Sulfatide expression in oligodendrocytes. Mature oligodendrocytes were live-cell immunostained with anti-sulfatide. (A) Rat oligodendrocyte early progenitors were transplanted into the forebrain of a mutant mouse unable to make myelin and analyzed after 21 days by live staining of 250mm sections (Warrington and PfeiVer, 1992). Adapted from ‘‘Trends in Cell Biology,’’ Vol. 3, S. E. PfeiVer, A. E. Warrington, and R. Bansal, ‘‘The Oligodendrocyte and Its Many Cellular Processes,’’ pp. 191–197, Copyright (1993), with permission from Elsevier Science. (B) Mature mouse oligodendrocytes grown in culture were live immunostained.

Third, in mice lacking the enzyme ceramide, galactosyltransferase (the CGT-null model; Bosio et al., 1996; Coetzee et al., 1998b; Dupree et al., 1998; Dupree and Popko, 1999; StoVel and Bosio, 1997), GalC and its product sulfatide are absent while glucocerebroside (GlcCer) is somewhat elevated. Although the ultrastructure of compact myelin in CGTnull mice is not dramatically altered, the animals exhibit severe tremors and hind limb paralysis, and the paranodal loops are structurally defective (Fig. 3.9). Of particular

REGULATION OF OLIGODENDROCYTE PHYSIOLOGY BY GANGLIOSIDES

FIGURE 3.9 Optic nerve abnormalities in CGT-null mice. In CGT-null (CGT-KO) mice the normal adherence of the paranodal loops is altered. Immunostaining of wild type (WT; left) or CGT-null (CGT-KO; right) optic nerve demonstrates severe loss of organization of the paranodal protein caspr-1 (green) and juxtaparanodal potassium channels (red), whereas nodal sodium channel (blue) localization is relatively unaVected (images kindly contributed by Dr. M. N. Rasband, Department of Neuroscience, University of Connecticut Medical School).

interest to present considerations, terminal diVerentiation of OLs in these animals is enhanced without aVecting OL progenitor proliferation or survival (Bansal et al., 1999; nevertheless, myelin formation per se is retarded; Marcus et al., 2000). Recently, mice have been developed that are unable to synthesize sulfatide due to the absence of sulfotransferase activity (the CST-null model; note that GalC is expressed in these mice; Honke et al., 2002). These mice exhibit similar (but not as severe) characteristics as CGT-null mice, again suggesting that sulfatide may be the most important of these GSLs (Bansal et al., 1999). Here again, terminal diVerentiation was enhanced compared to wild-type littermates (Hirahara et al., 2003). This result is consistent with the concept derived from antibody perturbation (noted earlier) that sulfatide may be the critical lipid player. A summary model (Fig. 3.10) has been proposed (Bansal et al., 1999) in which sulfatide interacts with external ligands to negatively regulate OL terminal diVerentiation. According to this model, anti-GalC/sulfatide antibodies mimic an external ligand to continuously activate this negative regulation pathway; in contrast, the absence of GalC/ sulfatide in CGT-null and sulfatide in CST-null mice precludes this regulation. The mechanism of this glycosphingolipid-mediated regulation has not been identiWed. Pioneering studies by Dyer and Benjamins oVer some suggestions, however. These investigators demonstrated that treatment of cultured OLs with anti-GalC triggers a cascade of events that include the redistribution of membrane surface GalC over internal domains of MBP, the disruption of microtubules and microWlaments within the myelin-like sheets, and the influx of extracellular Ca2þ (also seen upon treatment with anti-sulfatide) (Dyer and Benjamins, 1988b, 1989, 1990, 1991; Dyer et al., 1994). This association with MBP may be functional, for upon antibody treatment, MBP becomes dephosphorylated and detergent insoluble through its association with the cytoskeleton. Consistent with these results, exposure of myelinated axons with O1 or O4 IgM antibodies leads to a distinctive dysmyelination in vivo (Rosenbluth et al., 1996, 1997) and in vitro (Rosenbluth and Moon, 2003) that is characterized by a widening of the myelin period. Recent indications that GSLs are intimately involved in the initiation of signal transduction through the formation of lipid rafts (noted later) oVer intriguing possibilities for investigation (Popko, 2000).

REGULATION OF OLIGODENDROCYTE PHYSIOLOGY BY GANGLIOSIDES Gangliosides are involved in cell-to-cell interactions, regulation of cell growth, apoptosis, neuritogenesis, and diVerentiation of cells including oligodendrocytes (Hakomori et al., 1998; Yim et al., 1994). Many gangliosides, and less complex sphingolipids such as sphingosine, sphingosine-1-phosphate, and ceramide, are important intracellular

71

72

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

FIGURE 3.10 Proposed model for GalC and sulfatide function in oligodendrocytes. SpeciWc antiglycosphingolipid antibodies, presumably mimicking an external ligand, either directly or indirectly activate negative regulatory pathways, thus inhibiting terminal diVerentiation of oligodendrocyte progenitors. In the studies of oligodendrocyte diVerentiation in the CGT-null and CST-null mice, the model suggests that the absence of sulfatide precludes the negative regulation, thus allowing terminal diVerentiation to proceed more eYciently. PL: phospholipid; GalC: galactocerebroside; SUL: sulfatide.

modulators of protein kinases (Mahoney and Schnaar, 1997; Spiegel and Milstien, 2002). Gangliosides are not homogeneously distributed on the cell surface; together with glycosylphosphatidyl-inositol (GPI)-anchored proteins, neutral glycosphingolipids, sphingomyelin, and cholesterol, they segregate into membrane domains (lipid rafts; discussed later), for which gangliosides, particularly GM1 and GM3, have been used as ‘‘raft markers’’ (Hakomori et al., 1998). Within these microdomains, gangliosides are thought to interact with key transmembrane receptors involved in cell adhesion and signaling events (Hakomori, 2002). Gangliosides can down-regulate speciWc intrinsic growth factor tyrosine kinase receptors by diVerent exquisite mechanisms; for example, binding of GM3 to epidermal growth factor (EGF) receptor results in the inhibition of its tyrosine kinase, but only when N-glycans in the receptor are fully processed to complex-type structure, suggesting a tight carbohydrate-carbohydrate interaction (Hakomori, 2002). Other gangliosides can up-regulate receptor kinases; for example, the binding of GM1 to the nerve growth factor receptor in PC12 cells activates its tyrosine kinase and promotes cell survival (also discussed later; Mutoh et al., 1995).

Ganglioside Expression as a Function of OL Development The major gangliosides found in vivo in adult human OLs and myelin are GM4 and GM1. Ganglioside composition undergoes signiWcant changes during OL development. Morphological diVerentiation of OLs is accompanied by increases in GM2 and GM3 expression, while GD3 is replaced by more complex gangliosides in the mature CNS (Cochran et al., 1982; Vartanian et al., 1992; Yu et al., 1988). As noted earlier, gangliosides are useful markers for identifying speciWc stages of the oligodendrocyte lineage (Fig. 3.7); in particular monoclonal antibody A2B5 recognizes developmentally regulated surface antigens of early rat glial cell precursors, identiWed as GT3 and O-acetyl GT3, both of which are downregulated as the cells diVerentiate to mature OLs (Farrer and Quarles, 1999). Both qualitative and quantitative diVerences are present in the composition of gangliosides and glycosphingolipids in the PNS compared to the CNS. One diVerence is the abundance in the PNS of neolacto-series gangliosides with the backbone structure Galb1,4GlcNAcb-

REGULATION OF OLIGODENDROCYTE PHYSIOLOGY BY GANGLIOSIDES

1,3Galb1,4Glc1,1Cer. The ceramide (fatty acid and sphingosine base) compositions of PNS gangliosides are also diVerent, having greater percentages of long-chain fatty acids and dehydrosphingosines (Ogawa-Goto and Abe, 1998).

GD3 and Cell Migration GD3 is enriched in a variety of neural cell types including reactive glia, gliomas, undiVerentiated neurons, and OL precursor cells (Seyfried and Yu, 1985). Whereas GD3þ progenitor cells are highly motile, the migratory capacity of OLs committed to terminal diVerentiation is strongly reduced (PfeiVer et al., 1993; Warrington et al., 1993). Tenascin-R promotes the adhesion and diVerentiation of O4þ late OL progenitors by a sulfatidemediated autocrine mechanism. However, when GD3 is expressed, an interaction occurs between tenascin and gangliosides that activates a signaling mechanism leading to an inhibition of focal adhesion kinase (FAK) tyrosine phosphorylation and integrindependent cell adhesion to extracellular matrix proteins such as Wbronectin (Probstmeier et al., 1999). This responsive mechanism appears to be common to various cell types expressing disialogangliosides. Acetylated GD3 (9-O-acetyl GD3) is also involved in neuronal migration. This ganglioside is expressed along the route of tangential migration of neurons into the olfactory bulb during development, and blocking 9-O-acetyl GD3 reduces the rate of neuronal locomotion by half (Santiago et al., 2001). The mechanism by which ganglioside expression leads to neuronal migration is still not well-deWned. However, it was suggested that speciWc gangliosides could interact with broadly distributed protein-based adhesion systems or that homophilic interactions between gangliosides located on adjacent membranes could regulate movement of subsets of cells. It is still a matter of speculation if acetylated GD3 regulates migration of glial progenitors as well.

GM3 and Cell Differentiation GD3 is the principal ganglioside labeled when OL progenitors are incubated with [14C]galactose. However, as cells diVerentiate over time in culture, GM3 becomes the predominantly labeled ganglioside. Thus, there is a switch in OL progenitors from GD3þ nonadhesive, highly migratory cells into GM3þ, integrin-responsive adhesive, nonmigratory cells; gangliosides are major players in the regulation of these processes. Moreover, exogenous GM3 but not GM1, GM2, GD3, or GD1a, enhances diVerentiation of oligodendrocytes (Yim et al., 1994), suggesting that GM3 may play an important role in this process as well. Gangliosides, in particular GM1, also promote diVerentiation of other cell types such as neurons, where they favor extension of neurites (Leskawa et al., 1995).

Gangliosides and Cell Survival Ceramide induces apoptosis in many cell types, including OLs; while the mechanism is not clear, activation of certain mitogen-activated protein kinases and caspase-3 is required (Brogi et al., 1997; Casaccia-BonneWl et al., 1996; D’Souza et al., 1996; Hida et al., 1998; Larocca et al., 1997; Scurlock and Dawson, 1999). Increased levels of ceramide can be triggered in OLs by nerve growth factor binding to mature OLs expressing the p75 neurotrophin receptor, tumor necrosis factor or IL-1b. In contrast to ceramide, certain gangliosides can promote cell survival, some by reversing ceramide-induced apoptosis.

Gangliosides, Axo-glial Communication and Nerve Regeneration Myelin inhibits axonal regeneration (Woolf and Bloechlinger, 2002). Although this may regulate unwanted nerve sprouting in the mature nervous system, it severely limits recovery from injury. The major brain gangliosides GD1a and GT1b have been implicated in both the maintenance of myelin stability and the control of nerve regeneration (Vyas and Schnaar, 2001). Both functional roles may be mediated via speciWc interactions of GD1a

73

74

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

and GT1b on axons with myelin-associated glycoprotein (MAG), a sialic acid binding protein on the periaxonal myelin membrane the binds preferentially to the terminal sequence of GD1a and GT1b, NeuAc3Galb3GalNAc (Crocker et al., 1996; Kelm et al., 1994; Vinson et al., 2001; Vyas et al., 2002). Supporting these observations, GM2/GD2 synthase knockout mice (Fig. 3.3G) lacking GD1a and GT1b have phenotypic features similar to those of MAG-deWcient animals (Sheikh et al., 1999). The binding of MAG to apposing gangliosides in the outer leaflet of an axon generates an intracellular inhibitory signal in the neuron. The molecular mechanism by which this occurs is unclear. The recently discovered Nogo receptor (neuronal receptor for Nogo, another myelin inhibitor of axonal regeneration) is a GPI-linked protein and is also implicated as a binding partner for MAG (Domeniconi et al., 2002; Fournier et al., 2001; Liu et al., 2002). Both GPI-linked proteins and gangliosides are thought to reside in membrane rafts, which are enriched in intracellular signaling molecules (discussed later).

Gangliosides and Diseases Sphingolipids are internalized and transported to late endosomes and lysosomes where they are degraded in a stepwise fashion, culminating in the cleavage of ceramide to fatty acid and sphingosine. Each step in the degradation process is catalyzed by a speciWc enzyme, which in some cases are assisted by helper proteins termed ‘‘saposins,’’ activator proteins that present the substrate to the hydrolytic enzyme. Mutations in either a lysosomal hydrolase or an activator protein can lead to defective hydrolysis and intracellular accumulation of lipids (see Section IV for general review of these diseases). Defective ganglioside degradation is a common feature among a group of these diseases called gangliosidosis. Ganglioside accumulation in these pathologies is highly cytotoxic (e.g., psychosine in Krabbe’s disease), leading to an inflammatory process that may participate in the precipitous loss of neurons found in individuals with these disorders (Suzuki, 1998; Tohyama et al., 2001). Gangliosides also have been implicated in many neuropathies including HIV-associated demyelinating neuropathy characterized by high levels of antibodies against gangliosides (Ariga et al., 2001; Fredman and Lekman, 1997; Petratos and Gonzales, 2000). Antibodies against GM1 and GD1b are frequently detected in multifocal motor neuropathy, IgM paraproteinemic neuropathy and Guillain-Barre syndrome, while antibodies against GD1a are also present in Guillain-Barre syndrome and multiple sclerosis (Kusunoki et al., 1993; Mata et al., 1999; Miyazaki et al., 2001). The occurrence and the role of these auto-antibodies to gangliosides are mostly unresolved, but the existing data suggest that among the multitude of enhanced B-cell responses occurring in autoimmune diseases such as the ones mentioned earlier, those directed to gangliosides are very common and should be evaluated in the future for treatment of these diseases.

GLYCOSPHINGOLIPID/CHOLESTEROL RAFTS For many years, the cell membrane was viewed as a two-dimensional solvent composed of various lipids into which integral membrane and transmembrane proteins were dissolved. According to this ‘‘fluid mosaic’’ model (Singer and Nicolson, 1972), the membrane can exist in two states. In the gel state, lipid acyl chains are essentially frozen, allowing little or no lateral mobility. In the liquid disordered state the whole bilayer is fluid. The change of state is dependent on temperature and lipid composition. More recently, biophysicists have identiWed a third phase. In this phase, called the liquid ordered phase, glycosphingolipids and cholesterol form lateral assemblies, becoming lipid ‘‘rafts’’ within a sea of glycerolipids (Brown and London, 1998a; Simons and Ikonen, 1997; Simons and Toomre, 2000). The degree of order present from raft to raft can be diVerent, depending on the composition of lipids present. Thus, depending on the tightness of lipid packing, proteins may or may not interact with the raft environment. It is generally believed that the transient stability that arises from the interactions of proteins and lipids

GLYCOSPHINGOLIPID/CHOLESTEROL RAFTS

allows some proteins to partition in and out of these domains while others are excluded (Fig. 3.11 A). While the determining factors that lead to raft association are still uncertain, it appears that lipid properties (e.g., head group, acyl chains and three-dimensional structure) play a role. N-glycosylation of proteins also appears to be an important factor in raft association (Benting et al., 1999).

FIGURE 3.11 Protein partitioning into rafts. (A) Membrane proteins can (a) ‘‘partition’’ between raft and nonraft domains (e.g., myelin oligodendrocyte glycoprotein), (b, c, d) be exclusively raft-associated (e.g., GPI-linked proteins (b) such as contactin/NCAM120, certain tetraspan proteins (c) such as oligodendrocyte speciWc protein OSP, and transmembrane raft proteins (d) ), or (e) be localized entirely outside raft microdomains (e.g., myelin associated glycoprotein). GalC: galactocerebroside; Chol: cholesterol; PL: phospholipids. Model drawn by Dr. Tim Coetzee, National Multiple Sclerosis Society. (B) Atomic force microscopy reveals sphingomyelin/cholesterol rafts (orange) protruding from a dioleoylphosphatidylcholine background (black) in a mica-supported lipid bilayer. The SNARE protein syntaxin 1A (yellow/white peaks) is shown to be eYciently excluded from the rafts. Image was generously contributed by D.E. Saslowsky, R.M. Henderson, and J.M. Edwardson, Department of Pharmacology, University of Cambridge, U.K. (C) Induced protein partitioning into rafts can initiate cell signaling. In this model, a protein that is mostly localized outside raft microdomains, after being cross-linked (by an antibody in this example), can become preferentially partitioned into rafts, where it interacts with other signaling complex proteins (e.g., kinases and phosphatases) that are also selectively recruited into rafts, leading to the activation of signaling pathways. Arrows indicate relative partition coeYcients.

75

76

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

Certain classes of proteins preferentially exist within rafts of the plasma membrane, including GPI-anchored proteins, doubly acylated proteins such as the Src family tyrosine kinases, the alpha subunit of heterotrimeric G-proteins, and many transmembrane proteins, including receptors, which tend to be palmitoylated. Furthermore, multiple types of lipid microdomains, or ‘‘glycosynapses,’’ have been predicted (Hakomori, 2002; van Meer, 2002). This provides the cell with a powerful tool with which to regulate signaling at the level of the cell surface (discussed later). Though they are sometimes used interchangeably, the term ‘‘raft’’ and ‘‘caveolae’’ are biochemically distinct (Hakomori, 2002; Simons and Toomre, 2000). The former is dependent on glycosphingolipids and cholesterol, whereas the latter is dependent on the structural protein caveolin. There have been several studies identifying caveolae in Schwann cells (Mikol et al., 2002). However, the presence of caveolae in oligodendrocytes has not been reported. We will therefore restrict our discussion to glycosphingolipid rafts.

Definition and Biochemistry of Rafts The high degree of order present within rafts can be used to advantage in their biochemical isolation. At cold temperatures, the degree of order within rafts becomes suYciently high that they are unlikely to be solubilized by nonionic detergents such as Triton X-100 (TX100) (Brown and Rose, 1992). These domains can therefore be isolated as detergentinsoluble glycosphingolipid/cholesterol-enriched microdomains: DIGs; also referred to as GEMs (glycolipid-enriched domains), DIMs (detergent-insoluble membranes), and DRMs (detergent-resistant membranes). Conversely, the degree of order of rafts decreases as temperature increases. Therefore, rafts become soluble at 378C when nonionic detergents are used. Since the degree of order in a raft is often cholesterol-dependent, rafts can be disrupted by agents that deplete (e.g., methyl b-cyclodextrin; Ledesma et al., 1998), sequester (e.g., saponin, Taylor et al., 2002; Wlipin, Vereb et al., 2000), or inhibit the synthesis of cholesterol (e.g., lovastatin Keller and Simons, 1998). Lastly, due to the high content of lipids in detergent-insoluble domains, they have a low density, making them highly buoyant on a sucrose density gradient (Brown and Rose, 1992). This isolation distinguishes insolubility due to lipid rafts versus cytoskeletal interactions or protein aggregation. Therefore, four traditional biochemical criteria for DIG association are (1) insolubility at 48C in a nonionic detergent, (2) solubility at 378C in a nonionic detergent, (3) solubility in a nonionic detergent at 48C following cholesterol extraction/perturbation, and (4) floating to low density on sucrose density gradients. The association of several CNS myelin proteins with DIGs was investigated (Arvanitis et al., 2002; Kim and PfeiVer, 1999). Of the proteins tested, only MOG and CNP (proteins restricted to the cytoplasm containing noncompact myelin) satisfy all four criteria mentioned earlier for DIG association (class I). PLP and MBP (myelin proteins present in the compact myelin, which contains little or no cytoplasm) and the noncompact myelin protein MAG are nearly entirely soluble in TX-100 at 48C, suggesting that these proteins are not associated with DIGs (class II). Cx32 is detergent-insoluble in TX-100 at 48C and floats to low density on a sucrose gradient but does not solubilize at 378C or following saponin pretreatment (class III). Lastly, OSP is detergent-insoluble at 48C, floats to low density on a sucrose gradient, solubilizes at 378C, but fails to solubilize following saponin pretreatment (class IV). Although PLP is not part of detergent-insoluble complexes from myelin membranes when extracted with TX-100 at 48C, PLP is insoluble in the zwitterionic detergent CHAPS and associates closely with cholesterol (as elegantly evidenced by its association with photoactivatable cholesterol in OLs) (Simons et al., 2000). Depletion of cholesterol or inhibition of sphingolipid synthesis in OLs, or suppression of galactocerebroside and sulfatide synthesis (CGT-null mutant mice, noted earlier), all disrupt the association of PLP with the CHAPS insoluble fraction. It was proposed that the association of PLP with these lipids is critical for the sorting of PLP and the assembly of CNS myelin. Subsequent work showed that overexpression of PLP in oligodendrocytes leads to an accumulation of PLP in late endosomes/lysosomes along with a sequestration of cholesterol in these cellular compartments (Simons et al., 2002).

GLYCOSPHINGOLIPID/CHOLESTEROL RAFTS

Lipid rafts are enriched in GPI-anchored proteins. The role of rafts in traYcking of GPI-anchored proteins to the myelin sheath has been investigated (Kra¨mer et al., 1997). GPI-anchored proteins in OLs are associated with glycosphingolipid/cholesterol enriched DIGs, and all GPI-anchored proteins studied were found in a fraction enriched in myelin, including F3/contactin, neuronal cell adhesion molecule (NCAM-120), 5’-nucleotidase, and CD55/DAF (decay accelerating factor). Domains enriched in GPIanchored proteins from OLs and myelin also include Fyn and Lyn kinases (Kra¨mer et al., 1999). Moreover, Fyn kinase is tightly associated with NCAM-120 and F3/contactin, suggesting a crucial role for this signaling cascade in initiating myelination. The investigation of myelin DIG proteins was recently extended to peripheral myelin, for which the established peripheral proteins PMP22, P0, and plasmolipin were shown to be present in detergent-insoluble complexes (Hasse et al., 2002). In addition, MAL and CD59 (NCAM-120), previously shown to be in OL/myelin DIGs, are also in Schwann cell/ peripheral myelin DIGs (Erne et al., 2002) (discussed later). A careful investigation of the eVects of diVerent detergents and conditions on the association of myelin proteins with DIGs was performed (Taylor et al., 2002). It was concluded that although the DIG isolation procedure appears straightforward, the lipid/ detergent interactions are complex. As noted by one lipid raft expert, ‘‘even deWned biochemical criteria for the isolation of DIGs do not guarantee that the resulting DIGs existed as lateral domains before detergent addition’’; furthermore, ‘‘some aspects of DIG isolation are still not understood, such as the composition and fate of the cytosolic surface of the sphingolipid/cholesterol rafts in the outer leaflet’’ (Gerrit van Meer, personal communication). Therefore, one must use caution when interpreting results from DIG analyses (Brown and London, 1998b; Claas et al., 2001; Drevot et al., 2002; Madore et al., 1999; Simons et al., 2000; Taylor et al., 2002).

Imaging of Rafts The raft hypothesis has remained subject to skepticism. The biochemical criteria for isolating rafts as detergent-insoluble glycosphingolipid domains were believed by some to create ‘‘rafts’’ as an artifact. Many, but not all, skeptics have been won over by convincing imaging studies (Friedrichson and Kurzchalia, 1998; Varma and Mayor, 1998; Vyas et al., 2001). These in vitro imaging studies took advantage of fluorescence resonance energy transfer (FRET) as well as biochemical crosslinking to visualize raft proteins clustering on the cell surface. Further imaging studies supporting the raft hypothesis include using photonic force microscopy to measure the size of rafts in living cells, visualizing rafts involved in IgE receptor signaling by electron microscopy, and visualizing the incorporation of a GPI-anchored protein (placental alkaline phosphatase, PLAP) into sphingomyelin rafts in supported lipid bilayers using atomic force microscopy (Pralle et al., 2000; Saslowsky et al., 2002; Wilson et al., 2000) (see Fig. 3.11B for an example of a protein excluded from sphingomyelin rafts). More recently, the role that lipid modiWcation plays in partitioning proteins into rafts was investigated using live cell FRET imaging. It was determined that acyl, but not prenyl, modiWcation promotes clustering into lipid rafts (Melkonian et al., 1999; Zacharias et al., 2002). This is particularly interesting because PLP, which is heavily acylated, may not be raft associated by traditional raft criteria (Simons et al., 2000), whereas CNP, which is anchored to the membrane by isoprenylation, is raft associated, emphasizing the need for FRET analysis in Schwann cells and/or OLs.

Functional Significance of Rafts: Protein Trafficking The lipid raft hypothesis gained impetus from early studies of cell polarity in epithelial cells (Simons and van Meer, 1988). It was noted that the lipid compositions of the apical and basolateral membranes are diVerent. Lipid rafts were identiWed as ‘‘dynamic assemblies of cholesterol and sphingolipids’’ in the outer leaflet of the plasma membrane of epithelial cells. Glucosylceramide and sphingomyelin are synthesized (in the lumens of the ER and

77

78

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

Golgi, respectively) from a common precursor, ceramide, yet sort from the trans Golgi network (TGN) to the apical and basolateral membranes, respectively, of polarized epithelial cells. This suggests that sphingolipids within the same membrane leaflet sort laterally into distinct microdomains before being shipped to the plasma membrane, leading to the hypothesis that ‘‘rafts’’ may be functionally important in budding of transport vesicles responsible for apical sorting of lipids and proteins from the TGN to the plasma membrane (Holthuis et al., 2001). It follows from this hypothesis that rafts functioning in apical sorting are formed in the TGN because of the cis to trans gradient of cholesterol in the Golgi (Orci et al., 1981). Furthermore, perturbation of cholesterol homeostasis aVects the sorting and accumulation of several proteins and lipids, likely due to alterations in membrane fluidity and therefore modulation of molecular sorting; it is therefore not surprising that cholesterol homeostasis is well regulated by the cell (Neufeld et al., 1999; Puri et al., 1999; Simons and Ikonen, 2000). Due to hydrogen bonding and van der Waals interactions, cholesterol prefers to partition into glycosphingolipid domains. However, glycosphingolipids are not required for raft formation (Ostermeyer et al., 1999). In the absence of glycosphingolipids, sphingomyelin and cholesterol are suYcient to form rafts. Indeed, upregulation of sphingomyelin synthesis induces axonal sorting of the GPI-linked protein Thy-1 (Ledesma et al., 1999). Alternatively, sphingolipid accumulation can occur in the absence of cholesterol (Brown, 1998). Following the example of the apical/basolateral sorting experiments, it was later shown that there is an enrichment of axonal proteins in rafts, whereas dendritic proteins are deenriched (Ledesma et al., 1998). More recently, investigators have transfected epithelial cells with the myelin proteins MOG and PLP to investigate the potential role of rafts in the sorting of these proteins. It was determined that although an apical sorting pathway does exist for PLP, MOG follows a basolateral pathway (Kroepfl and Gardinier, 2001a, b) (see Trapp et al., Section I, Chapter 2 of this volume). Although PLP and MOG are both traYcked to the myelin sheath, they exist in two distinct regions, the compact myelin and the noncompact myelin, respectively. Therefore, diVerent ‘‘rafts’’ may target proteins to diVerent domains. The relevance of traYcking of myelin proteins in polarized epithelial cells, while provocative and capable of generating interesting hypotheses, will need to be tested further in OLs and Schwann cells. Further evidence supporting the existence of sorting pathways in oligodendrocytes has come from investigations of the traYcking of transfected proteins known to sort apically or basolaterally in epithelial cells (de Vries et al., 1998). Hemagglutinin (HA), which sorts to the apical membrane in polarized epithelial cells and is present in detergent insoluble complexes, was conWned to the OL cell body plasma membrane; conversely, vesicular stomatitis virus-G (VSVG) protein, which sorts to the basolateral domain of polarized epithelial cells, was directed to the myelin-like membranes. On the basis of these studies, the existence of ‘‘apical and basolateral sorting pathways’’ in OLs has been proposed (de Vries and Hoekstra, 2000). However, the lipid composition of myelin more resembles that of the apical domains of polarized cells, and most of the major myelin proteins reach the myelin sheath in a raft-independent manner. Emerging ideas on the mechanisms of lipid sorting and transport will be pivotal in elucidating the role of rafts in these processes (Maier et al., 2001). Using a diVerential screening approach for oligodendrocyte gene expression, a rat myelin and lymphocyte protein (rMAL) was identiWed and cloned (Schaeren-Wiemers et al., 1995a, 1995b). Independently, the same protein was identiWed as a developmentally regulated proteolipid protein called MVP17 (myelin vesicular protein, 17kDa) that is present in detergent-insoluble complexes from OLs and myelin, providing the Wrst published work identifying rafts in oligodendrocytes and myelin (Kim et al., 1995; Kim and PfeiVer, 2002). MVP17/rMAL is upregulated during terminal diVerentiation of OLs. This protein is important for vesiculation in other cell lines, in which over-expression of MVP17/rMAL leads to an increase of apical traYcking vesicles, whereas decreased expression reduces apical transport, leaving basolateral transport unaltered (Cheong et al., 1999; Puertollano et al., 1999). MVP17/rMAL is believed to be a component of the cell sorting/ vesicle traYcking machinery in OLs, and it colocalizes with MBP and PLP in the compact

GLYCOSPHINGOLIPID/CHOLESTEROL RAFTS

myelin (Frank et al., 1998). MVP17/rMAL is thought to interact directly with sulfatide, supporting the view that rafts may be important for the recruitment of components of the sorting machinery (CaduV et al., 2001; Frank et al., 1998; Frank, 2000).

Functional Significance of Rafts: Signal Transduction Evidence from recent raft experiments has lead to a burgeoning conviction for their role as platforms for the initiation of signal transduction (Zajchowski and Robbins, 2002). Lipid rafts containing a given set of proteins can change their size and composition in response to intra- or extracellular stimuli. This favors speciWc protein-protein interactions (involving proteins such as growth factor receptors, integrins, and cell adhesion molecules), providing ‘‘activation centers’’ for signal transduction: upon ligand binding receptors move into lipid rafts where the concentration of signaling molecules leads to signal ampliWcation. This ampliWcation occurs via exclusion of negative modulators (e.g., phosphatases) and inclusion of positive modulators (e.g., kinases) resulting in downstream signaling. The ability of rafts to amplify signaling events is likely to be a key factor in their involvement in many diseases (Simons and Ehehalt, 2002). Examples of signaling pathways involving lipid rafts abound, including receptors for FcRI, T-and B-cells, EGF, Insulin, EphrinB1, Neurotrophin, GDNF, Hedgehog, H-Ras, integrins, and eNOS (Simons and Toomre, 2000). The interactions that drive raft assembly are dynamic and reversible. Raft clusters can be disassembled by negative modulators or by removal of raft components from the cell surface by endocytosis. Individual rafts can coalesce to form raft clusters upon crosslinking. Antibodies and lectins provide an established, powerful experimental method for such cross-linking in living cells, whereupon raft and nonraft components separate into micron-sized patches (Harder et al., 1998). For example, antibody crosslinking of MOG in oligodendrocytes leads to a major redistribution of this protein into detergent-insoluble fractions (TX-100, 48C; Fig. 3.11C), resulting in the initiation of speciWc signaling changes and cell morphology that may have important implications for pathological contexts such multiple sclerosis (Marta et al., 2003). The movement and behavior of raft clusters can be influenced by cytoskeletal interactions and second messengers (e.g., PtdIns4,5P2), which help organize actin and tubulin assemblies on the cytoplasmic surface (Czech, 2000; Holowka et al., 2000; Maekawa et al., 2001; Rozelle et al., 2000). Cytoskeleton-stabilized membrane domains, or ‘‘membrane skeletons,’’ also resist disruption by nonionic detergents such as Triton X-100 and contain cholesterol (Nebl et al., 2002). These higher density domains also provide sites at which speciWc signal transduction molecules can concentrate (Holowka et al., 2000; Valensin et al., 2002). Thus, future studies on the functional roles of either low density lipid rafts or cytoskeletal associated membrane domains should consider them as two diVerent but related entities. Rafts are likely to be important in signaling between the axon and glia at the inner mesaxonal loop and at the paranodal loops. Indeed, this could explain why the major compact myelin proteins, PLP and MBP, are soluble in TX-100 at 48C, whereas many noncompact myelin proteins, such as MOG, CNP, OSP, and NCAM-120 are insoluble. The association of PLP with the glycosphingolipids GalC and sulfatide may have more to do with structural integrity (negative charges on sulfatide, positive charges on PLP) than with signal transduction. However, this could be a secondary function, at least temporally, since PLP is an important molecule in integrin signaling in OLs (Gudz et al., 2002). There also appears to be a link between rafts and the extracellular matrix. Tenascin-R, which is expressed by OLs during myelination, binds directly to cell surface sulfatide and is present in rafts (Kappler et al., 2002). This interaction induces gene expression of myelin proteins as well as an increase in its own expression, thus making tenascin-R an intrinsic autocrine factor leading to terminal diVerentiation. Future studies of glycosphingolipid rafts in oligodendrocytes, Schwann cells, and myelin will provide insight into the cellular mechanisms employed by the myelinating cells to traYc proteins, regulate signaling, and initiate myelination. This knowledge will contribute to the therapies being developed to encourage remyelination in pathological contexts.

79

80

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

PROSPECTUS Lipids are clearly cornerstones of cell biology. In myelin this position extends to basic issues of myelin structure, myelin function mediated by endogenous signal transduction, and myelin-axon signaling. Evolving new technical and conceptual advances promise to grease the road of lipid research. In this chapter, we have sought to emphasize aspects of generic lipid research that are likely to be important focal points for myelin ‘‘lipidomics’’ in the present and near future. Areas of particular promise include analyses of the mechanisms by which lipid rafts regulate lipid and protein sorting on the one hand and signal transduction on the other. This is expected to have important implications for the bidirectional signaling occurring between the myelin membrane and the axolemma. Further, additional mutants in which the genetic modiWcation is targeted to myelin forming cells can be expected to oVer important new insights. Finally, evolving technologies for high resolution imaging will provide novel avenues for studying the role of lipids in intracellular traYcking. Only then can we expect to fully understand the lipid ‘‘soul’’ stated so clearly by Thudichum more than 100 years ago.

Acknowledgments We sincerely thank Professors Pierre Morell (University of North Carolina, Chapel Hill, North Carolina), Joyce Benjamins (Wayne State Medical School, Detroit, Michigan), Ronald Schnaar (Johns Hopkins University, Baltimore, Maryland), and Gerrit van Meer (Institute of Biomembranes, Utrecht, The Netherlands) for truly excellent suggestions for improvements and corrections to this chapter. We are pleased to acknowledge support from the following research grants: National Institutes of Health NS10861 (SP), NS41078 (SP), NS38878 (RB), and NS45440 (CT); National Multiple Sclerosis Society FG1423 (CM).

References Agranoff, B. W., Benjamins, J. A., and Hajra, A. K. (1999). Lipids. In ‘‘Basic Neurochemistry: Molecular, Cellular and Medical Aspects’’ (Siegel, G. J., et al.), pp. 47–67. Lippincott-Raven Publishers, Philadelphia. Ariga, T., Miyatake, T., and Yu, R. K. (2001). Recent studies on the roles of antiglycosphingolipids in the pathogenesis of neurological disorders. J. Neurosci. Res. 65, 363–370. Ariga, T., and Yu, R. K. (1998). The role of globo-series glycolipids in neuronal cell differentiation–a review. Neurochem. Res. 23, 291–303. Arvanitis, D. N., Yang, W., and Boggs, J. M. (2002). Myelin proteolipid protein, basic protein, the small isoform of myelin-associated glycoprotein, and p42MAPK are associated in the Triton X-100 extract of central nervous system myelin. J. Neurosci. Res. 70, 8–23. Bansal, R., Gard, A. L., and Pfeiffer, S. E. (1988). Stimulation of oligodendrocyte differentiation in culture by growth in the presence of a monoclonal antibody to sulfated glycolipid. J. Neurosci. Res. 21, 260–267. Bansal, R., and Pfeiffer, S. E. (1989). Reversible inhibition of oligodendrocyte progenitor differentiation by a monoclonal antibody against surface galactolipids. Proc. Natl. Acad. Sci. USA. 86, 6181–6185. Bansal, R., and Pfeiffer, S. E. (1992). Novel stage in the oligodendrocyte lineage defined by reactivity of progenitors with R-mAb prior to O1 anti-galactocerebroside. J. Neurosci. Res. 32, 309–316. Bansal, R., and Pfeiffer, S. E. (1994a). Inhibition of protein and lipid sulfation in oligodendrocytes blocks biological responses to FGF-2 and retards cytoarchitectural maturation, but not developmental lineage progression. Dev. Biol. 162, 511–524. Bansal, R., and Pfeiffer, S. E. (1994b). Regulation of gene expression in mature oligodendrocytes by the specialized myelin-like membrane environment: antibody perturbation in culture with the monoclonal antibody R-mAb. Glia 12, 173–179. Bansal, R., Stefansson, K., and Pfeiffer, S. E. (1992). Proligodendroblast antigen (POA), a developmental antigen expressed by A007/O4-positive oligodendrocyte progenitors prior to the appearance of sulfatide and galactocerebroside. J. Neurochem. 58, 2221–2229. Bansal, R., Warrington, A. E., Gard, A. L., Ranscht, B., and Pfeiffer, S. E. (1989). Multiple and novel specificities of monoclonal antibodies O1, O4, and R-mAb used in the analysis of oligodendrocyte development. J. Neurosci. Res. 24, 548–557. Bansal, R., Winkler, S., and Bheddah, S. (1999). Negative regulation of oligodendrocyte differentiation by galactosphingolipids. J. Neurosci. 19, 7913–7924. Bejaoui, K., Wu, C., Scheffler, M. D., Haan, G., Ashby, P., Wu, L., de Jong, P., and Brown, R. H., Jr. (2001). SPTLC1 is mutated in hereditary sensory neuropathy, type 1. Nat. Genet. 27, 261–262.

PROSPECTUS

Benjamins, J. A., and Smith, M. E. (1984). Metabolism of Myelin. In ‘‘Myelin’’ (Morell, P.), pp. 225–258. Plenum Press, New York. Benting, J. H., Rietveld, A. G., and Simons, K. (1999). N-Glycans mediate the apical sorting of a GPI-anchored, raft-associated protein in Madin-Darby canine kidney cells. J. Cell. Biol. 146, 313–320. Bieberich, E., MacKinnon, S., Silva, J., Li, D. D., Tencomnao, T., Irwin, L., Kapitonov, D., and Yu, R. K. (2002). Regulation of ganglioside biosynthesis by enzyme complex formation of glycosyltransferases. Biochemistry 41, 11479–11487. Bilderback, T. R., Chan, J. R., Harvey, J. J., and Glaser, M. (1997). Measurement of the rate of myelination using a fluorescent analogue of ceramide. J. Neurosci. Res. 49, 497–507. Boggs, J. M., and Wang, H. (2001). Effect of liposomes containing cerebroside and cerebroside sulfate on cytoskeleton of cultured oligodendrocytes. J. Neurosci. Res. 66, 242–253. Borst, P., Zelcer, N., and van Helvoort, A. (2000). ABC transporters in lipid transport. Biochimica et Biophysica Acta 1486, 128–144. Bosio, A., Binczek, E., and Stoffel, W. (1996). Functional breakdown of the lipid bilayer of the myelin membrane in central and peripheral nervous system by disrupted galactocerebroside synthesis. Proc. Natl. Acad. Sci. USA. 93, 13280–13285. Broccardo, C., Luciani, M., and Chimini, G. (1999). The ABCA subclass of mammalian transporters. Biochim. Biophys. Acta 1461, 395–404. Brogi, A., Strazza, M., Melli, M., and Costantino-Ceccarini, E. (1997). Induction of intracellular ceramide by interleukin-1 beta in oligodendrocytes. J. Cell. Biochem. 66, 532–541. Brown, D. A., and London, E. (1998a). Functions of lipid rafts in biological membranes. Annu. Rev. Cell Dev. Biol. 14, 111–136. Brown, D. A., and London, E. (1998b). Structure and origin of ordered lipid domains in biological membranes. J. Membr. Biol. 164, 103–114. Brown, D. A., and Rose, J. K. (1992). Sorting of GPI-anchored proteins to glycolipid-enriched membrane subdomains during transport to the apical cell surface. Cell 68, 533–544. Brown, M. C., Besio Moreno, M., Bongarzone, E. R., Cohen, P. D., Soto, E. F., and Pasquini, J. M. (1993). Vesicular transport of myelin proteolipid and cerebroside sulfates to the myelin membrane. J. Neurosci. Res. 35, 402–408. Brown, R. E. (1998). Sphingolipid organization in biomembranes: what physical studies of model membranes reveal. J. Cell. Sci. 111, 1–9. Burger, K. N., van der Bijl, P., and van Meer, G. (1996). Topology of sphingolipid galactosyltransferases in ER and Golgi: transbilayer movement of monohexosyl sphingolipids is required for higher glycosphingolipid biosynthesis. J. Cell Biol. 133, 15–28. Caduff, J., Sansano, S., Bonnet, A., Suter, U., and Schaeren-Wiemers, N. (2001). Characterization of GFP-MAL expression and incorporation in rafts. Microsc. Res. Tech. 52, 645–655. Cane, D. E., Walsh, C. T., and Khosla, C. (1998). Harnessing the biosynthetic code: combinations, permutations, and mutations. Science 282, 63–68. Carey, D. J., Stahl, R. C., Cizmeci-Smith, G., and Asundi, V. K. (1994). Syndecan-1 expressed in Schwann cells causes morphological transformation and cytoskeletal reorganization and associates with actin during cell spreading. J. Cell Biol. 124, 161–170. Casaccia-Bonnefil, P., Carter, B. D., Dobrowsky, R. T., and Chao, M. V. (1996). Death of oligodendrocytes mediated by the interaction of nerve growth factor with its receptor p75. Nature 383, 716–719. Casaccia-Bonnefil, P., and Liu, A. (2003). The relationship between cell cycle molecules and onset of oligodendrocyte differentiation. J. Neurosci. Res. in press. Chalfant, C. E., and Hannun, Y. A., The role of serine/threonine protein phosphatases in ceramide signaling, in Ceramide Signaling, Futerman, A. H., Editor. 2003, Landes Bioscience: Georgetown. Cheong, K. H., Zacchetti, D., Schneeberger, E. E., and Simons, K. (1999). VIP17/MAL, a lipid raft-associated protein, is involved in apical transport in MDCK cells. Proc. Natl. Acad. Sci. USA 96, 6241–6248. Chester, M. A. (1998). IUPAC-IUB Joint Commission on Biochemical Nomenclature (JCBN). Nomenclature of glycolipids—recommendations 1997. Eur. J. Biochem. 257, 293–298. Claas, C., Stipp, C. S., and Hemler, M. E. (2001). Evaluation of prototype transmembrane 4 superfamily protein complexes and their relation to lipid rafts. J. Biol. Chem. 276, 7974–7984. Cochran, F. B., Jr., Yu, R. K., and Ledeen, R. W. (1982). Myelin gangliosides in vertebrates. J. Neurochem. 39, 773–779. Coetzee, T., Dupree, J. L., and Popko, B. (1998a). Demyelination and altered expression of myelin-associated glycoprotein isoforms in the central nervous system of galactolipid-deficient mice. J. Neurosci. Res. 54, 613–622. Coetzee, T., Fujita, N., Dupree, J., Shi, R., Blight, A., Suzuki, K., and Popko, B. (1996a). Myelination in the absence of galactocerebroside and sulfatide: normal structure with abnormal function and regional instability. Cell 86, 209–219. Coetzee, T., Li, X., Fujita, N., Marcus, J., Suzuki, K., Francke, U., and Popko, B. (1996b). Molecular cloning, chromosomal mapping, and characterization of the mouse UDP-galactose:ceramide galactosyltransferase gene. Genomics 35, 215–222. Coetzee, T., Suzuki, K., and Popko, B. (1998b). New perspectives on the function of myelin galactolipids. Trends Neurosci. 21, 126–130. Coppolino, M., Leung-Hagesteijn, C., Dedhar, S., and Wilkins, J. (1995). Inducible interaction of integrin alpha 2 beta 1 with calreticulin. Dependence on the activation state of the integrin. J. Biol. Chem. 270, 23132–23138.

81

82

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

Couraud, P. O., Delavier-Klutchko, C., Durieu-Trautmann, O., and Strosberg, A. D. (1981). ‘‘Antibodies raised against beta-adrenergic receptors stimulate adenylate cyclase’’. Biochem. Biophys. Res. Commun. 99, 1295–1302. Crocker, P. R., Kelm, S., Hartnell, A., Freeman, S., Nath, D., Vinson, M., and Mucklow, S. (1996). Sialoadhesin and related cellular recognition molecules of the immunoglobulin superfamily. Biochem. Soc. Trans. 24, 150–156. Czech, M. P. (2000). PIP2 and PIP3: complex roles at the cell surface. Cell 100, 603–606. D’Souza, S. D., Alinauskas, K. A., and Antel, J. P. (1996). Ciliary neurotrophic factor selectively protects human oligodendrocytes from tumor necrosis factor-mediated injury. J. Neurosci. Res. 43, 289–298. Dasgupta, S., Levery, S. B., and Hogan, E. L. (2002). 3-O-acetyl-sphingosine-series myelin glycolipids: characterization of novel 3-O-acetyl-sphingosine galactosylceramide. J. Lipid Res. 43, 751–761. Dawkins, J. L., Hulme, D. J., Brahmbhatt, S. B., Auer-Grumbach, M., and Nicholson, G. A. (2001). Mutations in SPTLC1, encoding serine palmitoyltransferase, long chain base subunit-1, cause hereditary sensory neuropathy type I. Nat. Genet. 27, 309–312. Dawson, G., and Cho, S. (2000). Batten’s disease: clues to neuronal protein catabolism in lysosomes. J. Neurosci. Res. 60, 133–140. de Vries, H., and Hoekstra, D. (2000). On the biogenesis of the myelin sheath: cognate polarized trafficking pathways in oligodendrocytes. Glycoconj. J. 17, 181–190. de Vries, H., Schrage, C., and Hoekstra, D. (1998). An apical-type trafficking pathway is present in cultured oligodendrocytes but the sphingolipid-enriched myelin membrane is the target of a basolateral-type pathway. Mol. Biol. Cell 9, 599–609. Di Biase, A., Argiolas, L., Confaloni, A., and Salvati, S. (1991). In vivo metabolism of fluorescent ceramide in central nervous system myelin of adult rats. Neurochem. Res. 16, 551–554. Diaz, M., Bornstein, M. B., and Raine, C. S. (1978). Disorganization of myelinogenesis in tissue culture by antiCNS antiserum. Brain Res. 154, 231–239. Domeniconi, M., Cao, Z., Spencer, T., Sivasankaran, R., Wang, K., Nikulina, E., Kimura, N., Cai, H., Deng, K., Gao, Y., He, Z., and Filbin, M. (2002). Myelin-associated glycoprotein interacts with the Nogo66 receptor to inhibit neurite outgrowth. Neuron 35, 283–290. Dorfman, S. H., Fry, J. M., and Silberberg, D. H. (1979). Antiserum induced myelination inhibition in vitro without complement. Brain Res. 177, 105–114. Drevot, P., Langlet, C., Guo, X. J., Bernard, A. M., Colard, O., Chauvin, J. P., Lasserre, R., and He, H. T. (2002). TCR signal initiation machinery is pre-assembled and activated in a subset of membrane rafts. EMBO J. 21, 1899–1908. Dupree, J. L., Coetzee, T., Blight, A., Suzuki, K., and Popko, B. (1998). Myelin galactolipids are essential for proper node of Ranvier formation in the CNS. J. Neurosci. 18, 1642–1649. Dupree, J. L., and Popko, B. (1999). Genetic dissection of myelin galactolipid function. J. Neurocytol. 28, 271–279. Dyer, C. A. (1993). Novel oligodendrocyte transmembrane signaling systems. Investigations utilizing antibodies as ligands. Mol. Neurobiol. 7, 1–22. Dyer, C. A. (2000). Comments on the neuropathology of phenylketonuria. Eur. J. Pediatr. 159 Suppl 2, S107–108. Dyer, C. A., and Benjamins, J. A. (1988a). Antibody to galactocerebroside alters organization of oligodendroglial membrane sheets in culture. J. Neurosci. 8, 4307–4318. Dyer, C. A., and Benjamins, J. A. (1988b). Redistribution and internalization of antibodies to galactocerebroside by oligodendroglia. J. Neurosci. 8, 883–891. Dyer, C. A., and Benjamins, J. A. (1989). Organization of oligodendroglial membrane sheets: II. Galactocerebroside: antibody interactions signal changes in cytoskeleton and myelin basic protein. J. Neurosci. Res. 24, 212–221. Dyer, C. A., and Benjamins, J. A. (1990). Glycolipids and transmembrane signaling: antibodies to galactocerebroside cause a calcium influx in oligodendrocytes. J. Cell Biol. 111, 625–633. Dyer, C. A., and Benjamins, J. A. (1991). Galactocerebroside and sulfatide independently mediate Ca2þ responses in oligodendrocytes. J. Neurosci. Res. 30, 699–711. Dyer, C. A., Kendler, A., Philibotte, T., Gardiner, P., Cruz, J., and Levy, H. L. (1996). Evidence for central nervous system glial cell plasticity in phenylketonuria. J. Neuropathol. Exp. Neurol. 55, 795–814. Dyer, C. A., Philibotte, T. M., Wolf, M. K., and Billings-Gagliardi, S. (1994). Myelin basic protein mediates extracellular signals that regulate microtubule stability in oligodendrocyte membrane sheets. J. Neurosci. Res. 39, 97–107. Eisenbarth, G. S., Walsh, F. S., and Nirenberg, M. (1979). Monoclonal antibody to a plasma membrane antigen of neurons. Proc. Natl. Acad. Sci. U.S.A. 76, 4913–4917. Erne, B., Sansano, S., Frank, M., and Schaeren-Wiemers, N. (2002). Rafts in adult peripheral nerve myelin contain major structural myelin proteins and myelin and lymphocyte protein (MAL) and CD59 as specific markers. J. Neurochem. 82, 550–562. Farrer, R. G., and Benjamins, J. A. (1992). Entry of newly synthesized gangliosides into myelin. J. Neurochem. 58, 1477–1484. Farrer, R. G., and Quarles, R. H. (1999). GT3 and its O-acetylated derivative are the principal A2B5-reactive gangliosides in cultured O2A lineage cells and are down-regulated along with O-acetyl GD3 during differentiation to oligodendrocytes. J. Neurosci. Res. 57, 371–380. Fournier, A. E., GrandPre, T., and Strittmatter, S. M. (2001). Identification of a receptor mediating Nogo-66 inhibition of axonal regeneration. Nature 409, 341–346. Frank, M. (2000). MAL, a proteolipid in glycosphingolipid enriched domains: functional implications in myelin and beyond. Prog. Neurobiol. 60, 531–544.

PROSPECTUS

Frank, M., van der Haar, M. E., Schaeren-Wiemers, N., and Schwab, M. E. (1998). rMAL is a glycosphingolipidassociated protein of myelin and apical membranes of epithelial cells in kidney and stomach. J. Neurosci. 18, 4901–4913. Fredman, P., and Lekman, A. (1997). Glycosphingolipids as potential diagnostic markers and/or antigens in neurological disorders. Neurochem. Res. 22, 1071–1083. Friedrichson, T., and Kurzchalia, T. V. (1998). Microdomains of GPI-anchored proteins in living cells revealed by crosslinking. Nature 394, 802–805. Fukasawa, M., Nishijima, M., and Hanada, K. (1999). Genetic evidence for ATP-dependent endoplasmic reticulum-to-Golgi apparatus trafficking of ceramide for sphingomyelin synthesis in Chinese hamster ovary cells. J. Cell Biol. 144, 673–685. Funakoshi, T., Yasuda, S., Fukasawa, M., Nishijima, M., and Hanada, K. (2000). Reconstitution of ATP- and cytosol-dependent transport of de novo synthesized ceramide to the site of sphingomyelin synthesis in semiintact cells. J. Biol. Chem. 275, 29938–29945. Funato, K., and Riezman, H. (2001). Vesicular and nonvesicular transport of ceramide from ER to the Golgi apparatus in yeast. J. Cell Biol. 155, 949–959. Furukawa, K., Takamiya, K., Okada, M., Inoue, M., and Fukumoto, S. (2001). Novel functions of complex carbohydrates elucidated by the mutant mice of glycosyltransferase genes. Biochim. Biophys. Acta 1525, 1–12. Gard, A. L., and Pfeiffer, S. E. (1990). Two proliferative stages of the oligodendrocyte lineage (A2B5þO4
and O4þGalC-) under different mitogenic control. Neuron 5, 615–625. Giraudo, C. G., Daniotti, J. L., and Maccioni, H. J. (2001). Physical and functional association of glycolipid N-acetylgalactosaminyl and galactosyl transferases in the Golgi apparatus. Proc. Natl. Acad. Sci. U.S.A. 98, 1625–1630. Gudz, T. I., Schneider, T. E., Haas, T. A., and Macklin, W. B. (2002). Myelin proteolipid protein forms a complex with integrins and may participate in integrin receptor signaling in oligodendrocytes. J. Neurosci. 22, 7398–7407. Hakomori, S. (1990). Bifunctional role of glycosphingolipids. Modulators for transmembrane signaling and mediators for cellular interactions. J. Biol. Chem. 265, 18713–18716. Hakomori, S., Yamamura, S., and Handa, A. K. (1998). Signal transduction through glyco(sphingo)lipids. Introduction and recent studies on glyco(sphingo)lipid-enriched microdomains. Ann. N. Y. Acad. Sci. 845, 1–10. Hakomori, S. I. (2002). Inaugural Article: The glycosynapse. [erratum appears in Proc Natl Acad Sci USA 2002 Mar 5;99(5):3356.]. Proc. Natl. Acad. Sci. U.S.A. 99, 225–232. Hannun, Y. A., and Obeid, L. M. (2002). The Ceramide-centric universe of lipid-mediated cell regulation: stress encounters of the lipid kind. J. Biol. Chem. 277, 25847–25850. Harder, T., Scheiffele, P., Verkade, P., and Simons, K. (1998). Lipid domain structure of the plasma membrane revealed by patching of membrane components. J. Cell Biol. 141, 929–942. Harder, T., and Simons, K. (1999). Clusters of glycolipid and glycosylphosphatidylinositol-anchored proteins in lymphoid cells: accumulation of actin regulated by local tyrosine phosphorylation. Eur. J. Immunol. 29, 556–562. Hardy, R., and Reynolds, R. (1991). Proliferation and differentiation potential of rat forebrain oligodendroglial progenitors both in vitro and in vivo. Development 111, 1061–1080. Hardy, R. J., and Friedrich, V. L., Jr. (1996). Oligodendrocyte progenitors are generated throughout the embryonic mouse brain, but differentiate in restricted foci. Development 122, 2059–2069. Hasse, B., Bosse, F., and Muller, H. W. (2002). Proteins of peripheral myelin are associated with glycosphingolipid/cholesterol-enriched membranes. J. Neurosci. Res. 69, 227–232. Heinrich, M., Wickel, M., Winoto-Morbach, S., Schneider-Brachert, W., Weber, T., Brunner, J., Saftig, P., Peters, C., Kronke, M., and Schutze, S. (2000). Ceramide as an activator lipid of cathepsin D. Adv. Exp. Med. Biol. 477, 305–315. Hida, H., Takeda, M., and Soliven, B. (1998). Ceramide inhibits inwardly rectifying Kþ currents via a Ras- and Raf-1-dependent pathway in cultured oligodendrocytes. J. Neurosci. 18, 8712–8719. Hirahara, Y., Bansal, R., Honke, K., Ikenaka, K., and Wada, Y. (2003). Sulfatide is a negative regulator of oligodendrocyte differentiation: development in sulfatide-null mice. Glia in press. Hirschberg, C. B., Robbins, P. W., and Abeijon, C. (1998). Transporters of nucleotide sugars, ATP, and nucleotide sulfate in the endoplasmic reticulum and Golgi apparatus. Annu. Rev. Biochem. 67, 49–69. Holowka, D., Sheets, E. D., and Baird, B. (2000). Interactions between Fc(epsilon)RI and lipid raft components are regulated by the actin cytoskeleton. J. Cell Sci. 113, 1009–1019. Holthuis, J. C., Pomorski, T., Raggers, R. J., Sprong, H., and Van Meer, G. (2001). The organizing potential of sphingolipids in intracellular membrane transport. Physiol. Rev. 81, 1689–1723. Honke, K., Hirahara, Y., Dupree, J., Suzuki, K., Popko, B., Fukushima, K., Fukushima, J., Nagasawa, T., Yoshida, N., Wada, Y., and Taniguchi, N. (2002). Paranodal junction formation and spermatogenesis require sulfoglycolipids. Proc. Natl. Acad. Sci. U.S.A. 99, 4227–4232. Horinouchi, K., Erlich, S., Perl, D. P., Ferlinz, K., Bisgaier, C. L., Sandhoff, K., Desnick, R. J., Stewart, C. L., and Schuchman, E. H. (1995). Acid sphingomyelinase deficient mice: a model of types A and B Niemann-Pick disease. Nat. Genet. 10, 288–293. Huttenlocher, P. R. (2000). The neuropathology of phenylketonuria: human and animal studies. Eur. J. Pediatr. 159(S2), S102–106. Ichikawa, S., and Hirabayashi, Y. (1998). Glucosylceramide synthase and glycosphingolipid synthesis. Trends Cell Biol. 8, 198–202. Ichikawa, S., Sakiyama, H., Suzuki, G., Hidari, K. I., and Hirabayashi, Y. (1996). Expression cloning of a cDNA for human ceramide glucosyltransferase that catalyzes the first glycosylation step of glycosphingolipid synthesis. Proc. Natl. Acad. Sci. U.S.A. 93, 4638–4643.

83

84

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

Im, D. S., Heise, C. E. Ancellin, N., O’Dowd, B. F., Shei, G. J., Heavens, R. P., Rigby, M. R., Hla, T., Mandala, S., McAllister, G., George, S. R., and Lynch, K. R. (2000). Characterization of a novel sphingosine 1-phosphate receptor, Edg-8. J. Biol. Chem. 275, 14281–14286. Jacobs, S., Chang, K. J., and Cuatrecasas, P. (1978). Antibodies to purified insulin receptor have insulin-like activity. Science 200, 1283–1284. Jungalwala, F. B. (1994). Expression and biological functions of sulfoglucuronyl glycolipids (SGGLs) in the nervous system–a review. Neurochem. Res. 19, 945–957. Jurevics, H., and Morell, P. (1995). Cholesterol for synthesis of myelin is made locally, not imported into brain. J. Neurochem. 64, 895–901. Kapitonov, D., and Yu, R. K. (1999). Conserved domains of glycosyltransferases. Glycobiology 9, 961–978. Kappler, J., Baader, S. L., Franken, S., Pesheva, P., Schilling, K., Rauch, U., and Gieselmann, V. (2002). Tenascins are associated with lipid rafts isolated from mouse brain. Biochem. Biophys. Res. Commun. 294, 742–747. Kawai, H., Allende, M. L., Wada, R., Kono, M., Sango, K., Deng, C., Miyakawa, T., Crawley, J. N., Werth, N., Bierfreund, U., Sandhoff, K., and Proia, R. L. (2001). Mice expressing only monosialoganglioside GM3 exhibit lethal audiogenic seizures. J. Biol. Chem. 276, 6885–6888. Keller, P., and Simons, K. (1998). Cholesterol is required for surface transport of influenza virus hemagglutinin. J. Cell Biol. 140, 1357–1367. Kelm, S., Pelz, A., Schauer, R., Filbin, M. T., Tang, S., de Bellard, M. E., Schnaar, R. L., Mahoney, J. A., Hartnell, A., Bradfield, P., and et al. (1994). Sialoadhesin, myelin-associated glycoprotein and CD22 define a new family of sialic acid-dependent adhesion molecules of the immunoglobulin superfamily. Curr. Biol. 4, 965–972. Kim, T., Fiedler, K., Madison, D. L., Krueger, W. H., and Pfeiffer, S. E. (1995). Cloning and characterization of MVP17: a developmentally regulated myelin protein in oligodendrocytes. J. Neurosci. Res. 42, 413–422. Kim, T., and Pfeiffer, S. E. (1999). Myelin glycosphingolipid/cholesterol-enriched microdomains selectively sequester the non-compact myelin proteins CNP and MOG. J. Neurocytol. 28, 281–293. Kim, T., and Pfeiffer, S. E. (2002). Subcellular localization and detergent solubility of MVP17/rMAL, a lipid raftassociated protein in oligodendrocytes and myelin. J. Neurosci. Res. 69, 217–226. Knapp, P. E. (1991). Studies of glial lineage and proliferation in vitro using an early marker for committed oligodendrocytes. J. Neurosci. Res. 30, 336–345. Kok, J. W., Babia, T., Klappe, K., Egea, G., and Hoekstra, D. (1998). Ceramide transport from endoplasmic reticulum to Golgi apparatus is not vesicle-mediated. Biochem. J. 333, 779–786. Kolesnick, R. (2002). The therapeutic potential of modulating the ceramide/sphingomyelin pathway. J. Clin. Invest. 110, 3–8. Kolter, T., Doering, T., Wilkening, G., Werth, N., and Sandhoff, K. (1999). Recent advances in the biochemistry of glycosphingolipid metabolism. Biochem. Soc. Trans. 27, 409–415. Kolter, T., Proia, R. L., and Sandhoff, K. (2002). Combinatorial ganglioside biosynthesis. J. Biol. Chem. 277, 25859–25862. Kra¨mer, E. M., Klein, C., Koch, T., Boytinck, M., and Trotter, J. (1999). Compartmentation of Fyn kinase with glycosylphosphatidylinositol-anchored molecules in oligodendrocytes facilitates kinase activation during myelination. J. Biol. Chem. 274, 29042–29049. Kra¨mer, E. M., Koch, T., Niehaus, A., and Trotter, J. (1997). Oligodendrocytes direct glycosyl phosphatidylinositol-anchored proteins to the myelin sheath in glycosphingolipid-rich complexes. J. Biol. Chem. 272, 8937–8945. Kra¨mer, E. M., Schardt, A., and Nave, K. A. (2001). Membrane traffic in myelinating oligodendrocytes. Microsc. Res. Tech. 52, 656–671. Kroepfl, J. F., and Gardinier, M. V. (2001a). Identification of a basolateral membrane targeting signal within the cytoplasmic domain of myelin/oligodendrocyte glycoprotein. J. Neurochem. 77, 1301–1309. Kroepfl, J. F., and Gardinier, M. V. (2001b). Mutually exclusive apicobasolateral sorting of two oligodendroglial membrane proteins, proteolipid protein and myelin/oligodendrocyte glycoprotein, in Madin-Darby canine kidney cells. J. Neurosci. Res. 66, 1140–1148. Krueger, W. H., Gonye, G. E., Madison, D. L., Murray, K. E., Kumar, M., Spoerel, N., and Pfeiffer, S. E. (1997). TPO1, a member of a novel protein family, is developmentally regulated in cultured oligodendrocytes. J. Neurochem. 69, 1343–1355. Kusunoki, S., Chiba, A., Tai, T., and Kanazawa, I. (1993). Localization of GM1 and GD1b antigens in the human peripheral nervous system. Muscle Nerve 16, 752–756. Lala, P., Ito, S., and Lingwood, C. A. (2000). Retroviral transfection of Madin-Darby canine kidney cells with human MDR1 results in a major increase in globotriaosylceramide and 10(5)- to 10(6)-fold increased cell sensitivity to verocytotoxin. Role of p-glycoprotein in glycolipid synthesis. J. Biol. Chem. 275, 6246–6251. Lannert, H., Bunning, C., Jeckel, D., and Wieland, F. T. (1994). Lactosylceramide is synthesized in the lumen of the Golgi apparatus. FEBS Lett. 342, 91–96. Lannert, H., Gorgas, K., Meissner, I., Wieland, F. T., and Jeckel, D. (1998). Functional organization of the Golgi apparatus in glycosphingolipid biosynthesis. Lactosylceramide and subsequent glycosphingolipids are formed in the lumen of the late Golgi. J. Biol. Chem. 273, 2939–2946. Larocca, J. N., Farooq, M., and Norton, W. T. (1997). Induction of oligodendrocyte apoptosis by C2-ceramide. Neurochem. Res. 22, 529–534. Ledeen, R. W., Yu, R. K., and Eng, L. F. (1973). Gangliosides of human myelin: sialosylgalactosylceramide (G7) as a major component. J. Neurochem. 21, 829–839.

PROSPECTUS

Ledesma, M. D., Brugger, B., Bunning, C., Wieland, F. T., and Dotti, C. G. (1999). Maturation of the axonal plasma membrane requires upregulation of sphingomyelin synthesis and formation of protein-lipid complexes. EMBO J. 18, 1761–1771. Ledesma, M. D., Simons, K., and Dotti, C. G. (1998). Neuronal polarity: essential role of protein-lipid complexes in axonal sorting. Proc. Natl. Acad. Sci. USA 95, 3966–3971. Leskawa, K. C., Maddox, T., and Webster, K. A. (1995). Effects of ethanol on neuroblastoma cells in culture: role of gangliosides in neuritogenesis and substrate adhesion. J. Neurosci. Res. 42, 377–384. Lin, X., Mattjus, P., Pike, H. M., Windebank, A. J., and Brown, R. E. (2000). Cloning and expression of glycolipid transfer protein from bovine and porcine brain. J. Biol. Chem. 275, 5104–5110. Liscum, L., and Munn, N. J. (1999). Intracellular cholesterol transport. Biochim. Biophys. Acta 1438, 19–37. Liu, B. P., Fournier, A., GrandPre, T., and Strittmatter, S. M. (2002). Myelin-associated glycoprotein as a functional ligand for the Nogo-66 receptor. Science 297, 1190–1193. Madison, D. L., Krueger, W. H., Cheng, D., Trapp, B. D., and Pfeiffer, S. E. (1999). SNARE complex proteins, including the cognate pair VAMP-2 and syntaxin-4, are expressed in cultured oligodendrocytes. J. Neurochem. 72, 988–998. Madison, D. L., and Pfeiffer, S. E. (1997). Do secretory pathway SNARE proteins mediate myelinogenesis? In ‘‘Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches’’ (Juurlink, B. H. J., Devon, R. M., Doucette, J. R., Schreyer, D. J., Verge, V. M. L.), pp. 145–155. Plenum Press, New York. Madore, N., Smith, K. L., Graham, C. H., Jen, A., Brady, K., Hall, S., and Morris, R. (1999). Functionally different GPI proteins are organized in different domains on the neuronal surface. EMBO J. 18, 6917–6926. Maekawa, S., Morii, H., Kumanogoh, H., Sano, M., Naruse, Y., Sokawa, Y., and Mori, N. (2001). Localization of neuronal growth-associated, microtubule-destabilizing factor SCG10 in brain-derived raft membrane microdomains. J. Biochem. (Tokyo) 129, 691–697. Mahoney, J. A., and Schnaar, R. L. (1997). Multivalent ganglioside and sphingosine conjugates modulate myelin protein kinases. Biochim. Biophys. Acta 1328, 30–40. Maier, O., Ait Slimane, T., and Hoekstra, D. (2001). Membrane domains and polarized trafficking of sphingolipids. Semin. Cell Dev. Biol. 12, 149–161. Marcus, J., Dupree, J. L., and Popko, B. (2000). Effects of galactolipid elimination on oligodendrocyte development and myelination. Glia 30, 319–328. Marta, C. B., Taylor, C. M., Coetzee, T., Kim, T., Winkler, S., Bansal, R., and Pfeiffer, S. E. (2003). Antibody cross-linking of myelin oligodendrocyte glycoprotein leads to its rapid repartitioning into detergent-insoluble fractions, and altered protein phosphorylation and cell morphology. J Neurosci 23, 5461–5471. Mata, S., Lolli, F., Soderstrom, M., Pinto, F., and Link, H. (1999). Multiple sclerosis is associated with enhanced B cell responses to the ganglioside GD1a. Mult. Scler. 5, 379–388. Maxfield, F. R., and Wustner, D. (2002). Intracellular cholesterol transport. J. Clin. Invest. 110, 891–898. Melkonian, K. A., Ostermeyer, A. G., Chen, J. Z., Roth, M. G., and Brown, D. A. (1999). Role of lipid modifications in targeting proteins to detergent-resistant membrane rafts. Many raft proteins are acylated, while few are prenylated. J. Biol. Chem. 274, 3910–3917. Menon, K. K., Rasband, M. N., Taylor, C. M., Brophy, P. J., Bansal, R., and Pfeiffer, S. E. (2003). The myelinaxolemmal complex:biochemical dissection and the role of galactosphingolipids. J. Neurochem. in press. Merrill, A. H., Jr. (2002). De novo sphingolipid biosynthesis: a necessary, but dangerous, pathway. J. Biol. Chem. 277, 25843–25846. Mikol, D. D., Scherer, S. S., Duckett, S. J., Hong, H. L., and Feldman, E. L. (2002). Schwann cell caveolin1 expression increases during myelination and decreases after axotomy. Glia 38, 191–199. Miyazaki, T., Kusunoki, S., Kaida, K., Shiina, M., and Kanazawa, I. (2001). Guillain-Barre syndrome associated with IgG monospecific to ganglioside GD1b. Neurology 56, 1227–1229. Morell, P., and Quarles, R. H. (1999). Myelin Formation, Structure and Biochemistry. In ‘‘Basic Neurochemistry: Molecular, Cellular and Medical Aspects’’ (Siegel, G. J., et al.), pp. 69–93. Lippincott-Raven Publishers, Philadelphia. Morell, P., Quarles, R. H., and Norton, W. T. (1994). Molecular, Cellular, and Medical Aspects. In ‘‘Basic Neurochemistry’’ (Siegel, G. J., et al.), pp. 117–143. Raven Press, New York. Muse, E. D., Jurevics, H., Toews, A. D., Matsushima, G. K., and Morell, P. (2001). Parameters related to lipid metabolism as markers of myelination in mouse brain. J. Neurochem. 76, 77–86. Mutoh, T., Tokuda, A., Miyadai, T., Hamaguchi, M., and Fujiki, N. (1995). Ganglioside GM1 binds to the Trk protein and regulates receptor function. Proc. Natl. Acad. Sci. USA 92, 5087–5091. Nebl, T., Pestonjamasp, K. N., Leszyk, J. D., Crowley, J. L., Oh, S. W., and Luna, E. J. (2002). Proteomic Analysis of a Detergent-resistant Membrane Skeleton from Neutrophil Plasma Membranes. J. Biol. Chem. 277, 43399–43409. Neufeld, E. B., Wastney, M., Patel, S., Suresh, S., Cooney, A. M., Dwyer, N. K., Roff, C. F., Ohno, K., Morris, J. A., Carstea, E. D., Incardona, J. P., Strauss, J. F., 3rd, Vanier, M. T., Patterson, M. C., Brady, R. O., Pentchev, P. G., and Blanchette-Mackie, E. J. (1999). The Niemann-Pick C1 protein resides in a vesicular compartment linked to retrograde transport of multiple lysosomal cargo. J. Biol. Chem. 274, 9627–9635. Norton, W. T., and Cammer, W. (1984). Isolation and Characterization of Myelin. In ‘‘Myelin’’ (Morell, P.), pp. 147–195. Plenum Press, New York. Ogawa-Goto, K., and Abe, T. (1998). Gangliosides and glycosphingolipids of peripheral nervous system myelins– a minireview. Neurochem. Res. 23, 305–310.

85

86

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

Ogawa-Goto, K., Funamoto, N., Ohta, Y., Abe, T., and Nagashima, K. (1992). Myelin gangliosides of human peripheral nervous system: an enrichment of GM1 in the motor nerve myelin isolated from cauda equina. J. Neurochem. 59, 1844–1849. Orci, L., Montesano, R., Meda, P., Malaisse-Lagae, F., Brown, D., Perrelet, A., and Vassalli, P. (1981). Heterogeneous distribution of filipin–cholesterol complexes across the cisternae of the Golgi apparatus. Proc. Natl. Acad. Sci. U.S.A. 78, 293–297. Ostermeyer, A. G., Beckrich, B. T., Ivarson, K. A., Grove, K. E., and Brown, D. A. (1999). Glycosphingolipids are not essential for formation of detergent-resistant membrane rafts in melanoma cells. methyl-beta-cyclodextrin does not affect cell surface transport of a GPI-anchored protein. J. Biol. Chem. 274, 34459–34466. Owens, G. C., and Bunge, R. P. (1990). Schwann cells depleted of galactocerebroside express myelin-associated glycoprotein and initiate but do not continue the process of myelination. Glia 3, 118–124. Paul, P., Kamisaka, Y., Marks, D. L., and Pagano, R. E. (1996). Purification and characterization of UDPglucose:ceramide glucosyltransferase from rat liver Golgi membranes. J. Biol. Chem. 271, 2287–2293. Pesheva, P., Gloor, S., Schachner, M., and Probstmeier, R. (1997). Tenascin-R is an intrinsic autocrine factor for oligodendrocyte differentiation and promotes cell adhesion by a sulfatide-mediated mechanism. J. Neurosci. 17, 4642–4651. Petratos, S., and Gonzales, M. E. (2000). Can antiglycolipid antibodies present in HIV-infected individuals induce immune demyelination? Neuropathology 20, 257–272. Pfeiffer, S. E., Warrington, A. E., and Bansal, R. (1993). The oligodendrocyte and its many cellular processes. Trends Cell Biol. 3, 191–197. Pomorski, T., Hrafnsdottir, S., Devaux, P. F., and van Meer, G. (2001). Lipid distribution and transport across cellular membranes. Semin. Cell Dev. Biol. 12, 139–148. Popko, B. (2000). Myelin galactolipids: mediators of axon-glial interactions? Glia 29, 149–153. Pralle, A., Keller, P., Florin, E. L., Simons, K., and Horber, J. K. (2000). Sphingolipid-cholesterol rafts diffuse as small entities in the plasma membrane of mammalian cells. J. Cell Biol. 148, 997–1008. Probstmeier, R., Michels, M., Franz, T., Chan, B. M., and Pesheva, P. (1999). Tenascin-R interferes with integrindependent oligodendrocyte precursor cell adhesion by a ganglioside-mediated signalling mechanism. Eur. J. Neurosci. 11, 2474–2488. Puertollano, R., Martin-Belmonte, F., Millan, J., de Marco, M. C., Albar, J. P., Kremer, L., and Alonso, M. A. (1999). The MAL proteolipid is necessary for normal apical transport and accurate sorting of the influenza virus hemagglutinin in Madin-Darby canine kidney cells. J. Cell Biol. 145, 141–151. Puri, V., Watanabe, R., Dominguez, M., Sun, X., Wheatley, C. L., Marks, D. L., and Pagano, R. E. (1999). Cholesterol modulates membrane traffic along the endocytic pathway in sphingolipid-storage diseases. Nat. Cell Biol. 1, 386–388. Raff, M. C., Fields, K. L., Hakomori, S. I., Mirsky, R., Pruss, R. M., and Winter, J. (1979). Cell-type-specific markers for distinguishing and studying neurons and the major classes of glial cells in culture. Brain Res. 174, 283–308. Raff, M. C., Mirsky, R., Fields, K. L., Lisak, R. P., Dorfman, S. H., Silberberg, D. H., Gregson, N. A., Leibowitz, S., and Kennedy, M. C. (1978). Galactocerebroside is a specific cell-surface antigenic marker for oligodendrocytes in culture. Nature 274, 813–816. Raggers, R. J., Pomorski, T., Holthuis, J. C., Kalin, N., and van Meer, G. (2000). Lipid traffic: the ABC of transbilayer movement. Trac 1, 226–234. Ranscht, B., Clapshaw, P. A., Price, J., Noble, M., and Seifert, W. (1982). Development of oligodendrocytes and Schwann cells studied with a monoclonal antibody against galactocerebroside. Proc. Natl. Acad. Sci. U.S.A. 79, 2709–2713. Ranscht, B., Wood, P. M., and Bunge, R. P. (1987). Inhibition of in vitro peripheral myelin formation by monoclonal anti-galactocerebroside. J. Neurosci. 7, 2936–2947. Riboni, L., Viani, P., Bassi, R., Prinetti, A., and Tettamanti, G. (1997). The role of sphingolipids in the process of signal transduction. Prog. Lipid Res. 36, 153–195. Rosenbluth, J., Liang, W. L., Liu, Z., Guo, D., and Schiff, R. (1996). Expanded CNS myelin sheaths formed in situ in the presence of an IgM antigalactocerebroside-producing hybridoma. J. Neurosci. 16, 2635–2641. Rosenbluth, J., Liang, W. L., Schiff, R., and Dou, W. K. (1997). Spinal cord dysmyelination induced in vivo by IgM antibodies to three different myelin glycolipids. Glia 19, 58–66. Rosenbluth, J., and Moon, D. (2003). Dysmyelination induced in vitro by IgM antisulfatide and antigalactocerebroside monoclonal antibodies. J. Neurosci. Res. 71, 104–109. Rozelle, A. L., Machesky, L. M., Yamamoto, M., Driessens, M. H., Insall, R. H., Roth, M. G., Luby-Phelps, K., Marriott, G., Hall, A., and Yin, H. L. (2000). Phosphatidylinositol 4,5-bisphosphate induces actin-based movement of raft-enriched vesicles through WASP-Arp2/3. Curr. Biol. 10, 311–320. Saher, G., Lappe-Siefke, C., Ishibashi, S., and Nave, K.-A. (2002). Dysmyelination of mice caused by CREmediated inactivation of cholesterol biosynthesis in oligodendrocytes, Program No. 728.6 2002 Abstract Viewer/Itinerary Planner. Society for Neuroscience Online. Santiago, M. F., Berredo-Pinho, M., Costa, M. R., Gandra, M., Cavalcante, L. A., and Mendez-Otero, R. (2001). Expression and function of ganglioside 9-O-acetyl GD3 in postmitotic granule cell development. Mol. Cell. Neurosci. 17, 488–499. Saslowsky, D. E., Lawrence, J., Ren, X., Brown, D. A., Henderson, R. M., and Edwardson, J. M. (2002). Placental alkaline phosphatase is efficiently targeted to rafts in supported lipid bilayers. J. Biol. Chem. 277, 26966–26970.

PROSPECTUS

Schaeren-Wiemers, N., Schaefer, C., Valenzuela, D. M., Yancopoulos, G. D., and Schwab, M. E. (1995a). Identification of new oligodendrocyte- and myelin-specific genes by a differential screening approach. J. Neurochem. 65, 10–22. Schaeren-Wiemers, N., Valenzuela, D. M., Frank, M., and Schwab, M. E. (1995b). Characterization of a rat gene, rMAL, encoding a protein with four hydrophobic domains in central and peripheral myelin. J. Neurosci. 15, 5753–5764. Schreiber, A. B., Lax, I., Yarden, Y., Eshhar, Z., and Schlessinger, J. (1981). Monoclonal antibodies against receptor for epidermal growth factor induce early and delayed effects of epidermal growth factor. Proc. Natl. Acad. Sci. U.S.A. 78, 7535–7539. Schulte, S., and Stoffel, W. (1993). Ceramide UDPgalactosyltransferase from myelinating rat brain: purification, cloning, and expression. Proc. Natl. Acad. Sci. U.S.A. 90, 10265–10269. Scurlock, B., and Dawson, G. (1999). Differential responses of oligodendrocytes to tumor necrosis factor and other pro-apoptotic agents: role of ceramide in apoptosis. J. Neurosci. Res. 55, 514–522. Seyfried, T. N., and Yu, R. K. (1985). Ganglioside GD3: structure, cellular distribution, and possible function. Mol. Cell. Biochem. 68, 3–10. Shefer, S., Tint, G. S., Jean-Guillaume, D., Daikhin, E., Kendler, A., Nguyen, L. B., Yudkoff, M., and Dyer, C. A. (2000). Is there a relationship between 3-hydroxy-3-methylglutaryl coenzyme a reductase activity and forebrain pathology in the PKU mouse? J. Neurosci. Res. 61, 549–563. Sheikh, K. A., Sun, J., Liu, Y., Kawai, H., Crawford, T. O., Proia, R. L., Griffin, J. W., and Schnaar, R. L. (1999). Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proc. Natl. Acad. Sci. U.S.A. 96, 7532–7537. Simons, K., and Ehehalt, R. (2002). Cholesterol, lipid rafts, and disease. J. Clin. Invest. 110, 597–603. Simons, K., and Ikonen, E. (1997). Functional rafts in cell membranes. Nature 387, 569–572. Simons, K., and Ikonen, E. (2000). How cells handle cholesterol. Science 290, 1721–1726. Simons, K., and Toomre, D. (2000). Lipid Rafts and Signal Transduction. Nat. Rev. Mol. Cell Biol. 1, 31–39. Simons, K., and van Meer, G. (1988). Lipid sorting in epithelial cells. Biochemistry 27, 6197–6202. Simons, M., Kra¨mer, E. M., Macchi, P., Rathke-Hartlieb, S., Trotter, J., Nave, K. A., and Schulz, J. B. (2002). Overexpression of the myelin proteolipid protein leads to accumulation of cholesterol and proteolipid protein in endosomes/lysosomes: implications for Pelizaeus-Merzbacher disease. J. Cell Biol. 157, 327–336. Simons, M., Kra¨mer, E. M., Thiele, C., Stoffel, W., and Trotter, J. (2000). Assembly of myelin by association of proteolipid protein with cholesterol- and galactosylceramide-rich membrane domains. J. Cell Biol. 151, 143– 154. Singer, S. J., and Nicolson, G. L. (1972). The fluid mosaic model of the structure of cell membranes. Science 175, 720–731. Singh, H., and Pfeiffer, S. E. (1985). Myelin-associated galactolipids in primary cultures from dissociated fetal rat brain: biosynthesis, accumulation, and cell surface expression. J. Neurochem. 45, 1371–1381. Smith, W. L., and Merrill, A. H., Jr. (2002). Sphingolipid metabolism and signaling minireview series. J. Biol. Chem. 277, 25841–25842. Sommer, I., and Schachner, M. (1981). Monoclonal antibodies (O1 to O4) to oligodendrocyte cell surfaces: an immunocytological study in the central nervous system. Dev. Biol. 83, 311–327. Sommer, I., and Schachner, M. (1982). Cell that are O4 antigen-positive and O1 antigen-negative differentiate into O1 antigen-positive oligodendrocytes. Neurosci. Lett. 29, 183–188. Spiegel, S., and Milstein, S. (2002). Sphingosine 1-phosphate, a key cell signaling molecule. J. Biol. Chem. 277, 25851–25854. Sprong, H., Kruithof, B., Leijendekker, R., Slot, J. W., van Meer, G., and van der Sluijs, P. (1998). UDPgalactose:ceramide galactosyltransferase is a class I integral membrane protein of the endoplasmic reticulum. J. Biol. Chem. 273, 25880–25888. Stoffel, W., and Bosio, A. (1997). Myelin glycolipids and their functions. Curr. Opin. Neurobiol. 7, 654–661. Sundaram, K. S., and Lev, M. (1992). Purification and activation of brain sulfotransferase. J. Biol. Chem. 267, 24041–24044. Suzuki, K. (1998). Twenty five years of the ‘‘psychosine hypothesis’’: a personal perspective of its history and present status. Neurochem. Res. 23, 251–259. Svennerholm, L. (1980). Ganglioside designation. Adv. Exp. Med. Biol. 125, 11. Tagami, S., Inokuchi Ji, J., Kabayama, K., Yoshimura, H., Kitamura, F., Uemura, S., Ogawa, C., Ishii, A., Saito, M., Ohtsuka, Y., Sakaue, S., and Igarashi, Y. (2002). Ganglioside GM3 participates in the pathological conditions of insulin resistance. J. Biol. Chem. 277, 3085–3092. Tanaka, Y., Yamada, K., Zhou, C. J., Ban, N., Shioda, S., and Inagaki, N. (2003). Temporal and spatial profiles of ABCA2-expressing oligodendrocytes in the developing rat brain. J. Comp. Neurol. 455, 353–367. Taylor, C. M., Coetzee, T., and Pfeiffer, S. E. (2002). Detergent-insoluble glycosphingolipid/cholesterol microdomains of the myelin membrane. J Neurochem 81, 993–1004. Tennekoon, G., Zaruba, M., and Wolinsky, J. (1983). Topography of cerebroside sulfotransferase in Golgienriched vesicles from rat brain. J. Cell Biol. 97, 1107–1112. Thompson, A. J., Tillotson, S., Smith, I., Kendall, B., Moore, S. G., and Brenton, D. P. (1993). Brain MRI changes in phenylketonuria. Associations with dietary status. Brain 116, 811–821. Thudichum, J. L. W. (1884). A Treatise on the Chemical Constitution of the Brain. Bailliere, Tindall, and Cox, London.

87

88

3. THE TRANSPORT, ASSEMBLY, AND FUNCTION OF MYELIN LIPIDS

Tohyama, J., Matsuda, J., and Suzuki, K. (2001). Psychosine is as potent an inducer of cell death as C6-ceramide in cultured fibroblasts and in MOCH-1 cells. Neurochem. Res. 26, 667–671. Townsend, L. E., Benjamins, J. A., and Skoff, R. P. (1984). Effects of monensin and colchicine on myelin galactolipids. J. Neurochem. 43, 139–145. Tozawa, R., Ishibashi, S., Osuga, J., Yagyu, H., Oka, T., Chen, Z., Ohashi, K., Perrey, S., Shionoiri, F., Yahagi, N., Harada, K., Gotoda, T., Yazaki, Y., and Yamada, N. (1999). Embryonic lethality and defective neural tube closure in mice lacking squalene synthase. J. Biol. Chem. 274, 30843–30848. Unger, R. H. (2002). Lipotoxic diseases. Annu. Rev. Med. 53, 319–336. Valensin, S., Paccani, S. R., Ulivieri, C., Mercati, D., Pacini, S., Patrussi, L., Hirst, T., Lupetti, P., and Baldari, C. T. (2002). F-actin dynamics control segregation of the TCR signaling cascade to clustered lipid rafts. Eur. J. Immunol. 32, 435–446. van Meer, G. (2002). Cell biology. The different hues of lipid rafts. Science 296, 855–857. van Meer, G., and Lisman, Q. (2002). Sphingolipid transport: rafts and translocators. J. Biol. Chem. 277, 25855– 25858. Varma, R., and Mayor, S. (1998). GPI-anchored proteins are organized in submicron domains at the cell surface. Nature 394, 798–801. Vartanian, T., Szuchet, S., and Dawson, G. (1992). Oligodendrocyte-substratum adhesion activates the synthesis of specific lipid species involved in cell signaling. J. Neurosci. Res. 32, 69–78. Vereb, G., Matko, J., Vamosi, G., Ibrahim, S. M., Magyar, E., Varga, S., Szollosi, J., Jenei, A., Gaspar, R., Jr., Waldmann, T. A., and Damjanovich, S. (2000). Cholesterol-dependent clustering of IL-2Ralpha and its colocalization with HLA and CD48 on T lymphoma cells suggest their functional association with lipid rafts. Proc. Natl. Acad. Sci. U.S.A. 97, 6013–6018. Vinson, M., Strijbos, P. J., Rowles, A., Facci, L., Moore, S. E., Simmons, D. L., and Walsh, F. S. (2001). Myelinassociated glycoprotein interacts with ganglioside GT1b. A mechanism for neurite outgrowth inhibition. J. Biol. Chem. 276, 20280–20285. Vos, J. P., Giudici, M. L., van der Bijl, P., Magni, P., Marchesini, S., van Golde, L. M., and Lopes-Cardozo, M. (1995). Sphingomyelin is synthesized at the plasma membrane of oligodendrocytes and by purified myelin membranes: a study with fluorescent- and radio-labelled ceramide analogues. FEBS Lett. 368, 393–396. Vos, J. P., Lopes-Cardozo, M., and Gadella, B. M. (1994). Metabolic and functional aspects of sulfogalactolipids. Biochim. Biophys. Acta 1211, 125–149. Vyas, A. A., Patel, H. V., Fromholt, S. E., Heffer-Lauc, M., Vyas, K. A., Dang, J., Schachner, M., and Schnaar, R. L. (2002). Gangliosides are functional nerve cell ligands for myelin-associated glycoprotein (MAG), an inhibitor of nerve regeneration. Proc. Natl. Acad. Sci. U.S.A. 99, 8412–8417. Vyas, A. A., and Schnaar, R. L. (2001). Brain gangliosides: functional ligands for myelin stability and the control of nerve regeneration. Biochimie 83, 677–682. Vyas, K. A., Patel, H. V., Vyas, A. A., and Schnaar, R. L. (2001). Segregation of gangliosides GM1 and GD3 on cell membranes, isolated membrane rafts, and defined supported lipid monolayers. Biol. Chem. 382, 241–250. Warrington, A. E., Barbarese, E., and Pfeiffer, S. E. (1993). Differential myelinogenic capacity of specific developmental stages of the oligodendrocyte lineage upon transplantation into hypomyelinating hosts. J. Neurosci. Res. 34, 1–13. Warrington, A. E., and Pfeiffer, S. E. (1992). Proliferation and differentiation of O4þ oligodendrocytes in postnatal rat cerebellum: analysis in unfixed tissue slices using anti-glycolipid antibodies. J. Neurosci. Res. 33, 338–353. Watanabe, R., Asakura, K., Rodriguez, M., and Pagano, R. E. (1999). Internalization and sorting of plasma membrane sphingolipid analogues in differentiating oligodendrocytes. J. Neurochem. 73, 1375–1383. Wilson, B. S., Pfeiffer, J. R., and Oliver, J. M. (2000). Observing FcepsilonRI signaling from the inside of the mast cell membrane. J. Cell Biol. 149, 1131–1142. Woolf, C. J., and Bloechlinger, S. (2002). Neuroscience. It takes more than two to Nogo. Science 297, 1132–1134. Yasuda, S., Kitagawa, H., Ueno, M., Ishitani, H., Fukasawa, M., Nishijima, M., Kobayashi, S., and Hanada, K. (2001). A novel inhibitor of ceramide trafficking from the endoplasmic reticulum to the site of sphingomyelin synthesis. J. Biol. Chem. 276, 43994–44002. Yim, S. H., Farrer, R. G., Hammer, J. A., Yavin, E., and Quarles, R. H. (1994). Differentiation of oligodendrocytes cultured from developing rat brain is enhanced by exogenous GM3 ganglioside. J. Neurosci. Res. 38, 268–281. Yu, R. K. (1994). Development regulation of ganglioside metabolism. Prog. Brain Res. 101, 31–44. Yu, R. K., and Bieberich, E. (2001). Regulation of glycosyltransferases in ganglioside biosynthesis by phosphorylation and dephosphorylation. Mol. Cell. Endocrinol. 177, 19–24. Yu, R. K., Macala, L. J., Taki, T., Weinfield, H. M., and Yu, F. S. (1988). Developmental changes in ganglioside composition and synthesis in embryonic rat brain. J. Neurochem. 50, 1825–1829. Zacharias, D. A., Violin, J. D., Newton, A. C., and Tsien, R. Y. (2002). Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science 296, 913–916. Zajchowski, L. D., and Robbins, S. M. (2002). Lipid rafts and little caves. Compartmentalized signalling in membrane microdomains. Eur. J. Biochem. 269, 737–752.

C H A P T E R

4 Functional Organization of the Nodes of Ranvier Steven S. Scherer, Edgardo J. Arroyo, and Elior Peles

INTRODUCTION (FIGS. 4.1–4.4)

au2

au1

Myelinated axons are completely covered by myelin sheaths except at nodes of Ranvier, small gaps (less than 1 mm in length) that are directly exposed to the extracellular milieu. By increasing the resistance and especially by reducing the capacitance, myelin reduces current Xow across the internodal axonal membrane (Blight, 1985; Funch and Faber, 1984), thereby facilitating saltatory conduction at nodes (Hille, 2001; also see Chapter 5 by Waxman and Bangalore). As shown in Figure 4.1, owing to their diVerential staining, Ramo´n y Cajal (1928) deduced that nodes, paranodes, and incisures contain diVerent molecular components. He and his contemporaries recognized an acellular sheath (‘‘the sheath of Schwann’’) that extended across PNS nodes, the reduction of the axonal caliber at nodes, as well as specializations around the node (‘‘cementing disc of Ranvier’’) and the paranode (‘‘spinous bracelets of Nageotte’’). His observations presaged later ultrastructural studies, which, together with recent investigations of the organization of molecular constituents provide new insights on the molecular architecture of the node, are the subjects of this review.

MOLECULAR SPECIALIZATIONS OF THE PNS AND CNS NODAL AXOLEMMA (FIGS. 4.2–4.6)

au3

By electron microscopy (EM), the nodal axolemma is more electron-dense and contains a higher density of intramembranous particles (1200/m2)—the voltage-gated Naþ channels (Nav channels)—than the internodal axolemma (Rosenbluth, 1995). The Xow of Naþ through these channels is essential for the propagation of action potentials (Hille, 2001); also see Chapter 5 by Waxman and Bangalore). The Nav1.1-Nav1.9 a subunits belong to a gene family (designated SCN1A-SCN9A in mammals; Goldin et al., 2000). The actual channel is composed of a single a subunit, a large, glycosylated, polytopic protein comprised of four homologous domains (Goldin et al., 2000). Each kind of Nav channel has distinct electrophysiological properties, including its sensitivity to tetrodotoxin (TTX). Nav1.6, a TTX-sensitive channel, appears to be the main one expressed in mature nodes of the CNS and PNS (Arroyo et al., 2002; Caldwell et al., 2000; Tzoumaka et al., 2000). Nav1.9 is mainly associated with unmyelinated sensory axons, but has also been reported in some PNS nodes (Fjell et al., 2000), and Nav1.2 and Nav1.8 are found at many CNS nodes (Arroyo et al., 2002). In both the PNS and the CNS, Nav1.6 replaces Nav1.2 in some, but not all, nodes (Arroyo et al., 2002; Boiko et al., 2001; Kaplan et al., 2001; Rasband and Trimmer, 2001).

Myelin Biology and Disorders, Volume 1

89

Copyright 2004, Elsevier (USA). All rights reserved.

90

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

FIGURE 4.1 Ramo´n y Cajal’s depiction of the nodal region. (A and B) ‘‘Node of nerve Wbres. (A) Axonic impregnation often found in diluted and quick-acting silver solutions. (B) Impregnation with silver, after Wxation in formol-pyridinemanganese. a, spinous bracelets of Nageotte; b, disc of Ranvier; c, Schwann’s membrane; d, axon. (C) ‘‘Schematic drawing of the nerve Wbre at the level of the node. a, Wne oblique neuroWbrils; b, cementing discs; c, longitudinal stout neuroWbril; d, neurilemma; e, region of the axon which corresponds to the node.’’ Figures and quotations are from (Ramo´n y Cajal, 1928), with permission of Oxford University Press.

FIGURE 4.2 Schematic ultrastructure of the nodal and paranodal regions of a myelinated PNS axon. This drawing illustrates the transcellular bridges between Schwann cell microvilli and the nodal axolemma, which contains a high density of large particles (Nav channels). The paranodal axolemma has rows of particles in register with the rows of particles in the paranodal loops; together these form the septate-like junctions. ModiWed from Ichimura and Ellisman, 1991, with permission of Kluwer Academic Press.

Mature channels are composed of one a and two b subunits (Fig. 4.5). Three genes (SCN1B-SCN3B) encode b subunits (b1-3), which are type I transmembrane domain proteins with a single extracellular immunoglobulin (Ig) domain (Catterall, 2000; Isom, 2002). The Ig domain of b2 is closely related to one of the Ig domains of contactin, a glycosylphosphatidylinositol (GPI)-anchored cell adhesion molecule (CAM) of the Ig superfamily (Isom et al., 1995). b1 and b3 are more closely related to each other than to b2. Both b1 and b2 have been localized to nodes (Chen et al., 2002a; RatcliVe et al., 2001); whether b3 is localized to nodes remains to be determined, but its mRNA is widely expressed by CNS and PNS neurons (Qu et al., 2001; Shah et al., 2000). Because Nav1.6 is the most common a subunit at nodes, it is their likely partner, but this remains to be shown directly; whether b1, b2, and b3 interact with Nav1.2, Nav1.8, or Nav1.9 at nodes also remains to be determined. b2 is required for normal density of nodal Nav channels (Chen et al., 2002a). A variety of functions have been proposed for b subunits. In transfected cells, b subunits increase the delivery to the cell membrane, and may alter their electrophysiological characteristics, of some but not all kinds of a subunits (Catterall, 2000; Meadows et al., 2001). The Ig domain of b subunits enables them to act as CAMs and appears to mediate

MOLECULAR SPECIALIZATIONS OF THE PNS AND CNS NODAL AXOLEMMA

FIGURE 4.3 ‘‘Autotypic’’/ ‘‘reXexive’’ tight, gap, and adherens junctions in the PNS myelin sheath. (A) A myelinating Schwann cell has been ‘‘unrolled’’ to reveal its trapezoidal shape; the two lateral edges deWne the paranodes; the outside edge deWnes the outer mesaxon; the inside edge deWnes the inner mesaxon. The nodal microvilli are associated with the outermost aspect of the myelin sheath. Noncompact myelin is found in the paranodal region and in incisures. Tight junctions are depicted as two continuous lines; these form a circumferential belt and are also found in incisures. Gap junctions are depicted as ovals; these are found between the rows of tight junctions. Adherens junctions are depicted as ‘‘x’’s. ModiWed from Kleopa and Scherer, 2002, with permission of W.B. Saunders. (B) Schematic representation of the proteins of compact and noncompact myelin. Compact myelin contains P0, peripheral myelin protein 22 kDa (PMP22), and myelin basic protein (MBP); noncompact myelin contains Ecadherin, myelin-associated glycoprotein (MAG), DM20, Cx32, Cx29, claudin-1, and claudin-5. ModiWed from Arroyo and Scherer, 2000, with permission of Springer-Verlag.

complex interactions with various nodal components: b1 and b2 subunits can interact with the extracellular matrix molecules tenascin-C and tenascin-R (Srinivasan et al., 1998; Xiao et al., 1999); b1 and b3 (but not b2) can interact in cis with neurofascin and Nr-CAM (RatcliVe et al., 2001), both of which are found at nodes (see below); b1 (but not b2) can interact with contactin and phosphacan, a secreted isoform of receptor protein tyrosine phosphatase b (RPTPb; Kazarinova-Noyes et al., 2001; RatcliVe et al., 2000). Furthermore, b1 and b2 can also interact homophilically and their cytoplasmic domains can recruit ankyrinG (Malhotra et al., 2000). In addition, the interaction of Nav channels with ankryrinG at nodes could also be mediated by the a subunit (Bouzidi et al., 2002). Although it is not clear which of these potential interactions play important roles in the formation of nodes, nodal Nav channels are complexed with b subunits and CAMs that are also linked to the axonal cytoskeleton through ankyrinG (Bennett et al., 1997). Ankyrins are adaptor proteins that link many intrinsic membrane proteins to the spectrin cytoskeleton (Bennett and Baines, 2001). Two splice variants of ankyrinG, 270 and 480 kDa, colocalize with Nav channels at initial segments and nodes (Kordeli et al., 1990; Kordeli et al., 1995). These isoforms are distinguished by their membrane-binding domain composed of ANK repeats, a spectrin-binding domain, and a serine/threonine-rich domain (Zhang and Bennett, 1996). AnkyrinG 270/480 kDa interacts with the cytoplasmic domains of Nav1.2 (Bouzidi et al., 2002), NF186 and Nr-CAM, two Ig CAMs that are localized at the nodes (Davis and Bennett,

91

92

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

FIGURE 4.4 Schematic depiction of the node, paranode, juxtaparanode, and internode. Both panels depict the CNS and PNS myelin sheaths (meeting at a node) surrounding a single myelinated axon. The myelin sheaths have been bisected, revealing the surface of the axon. The relative intensity of the red and green indicates the relative abundance of the indicated proteins. (A) Wild-type pattern. This drawing illustrates that axonal membrane of nodes, paranodes, juxtaparanodes, and internodes are each characterized by their expression of a diVerent set of molecules, in both the CNS and the PNS. ModiWed from Arroyo and Scherer, 2000, used with permission of Springer-Verlag. (B) Altered axonal organization in cgt- and cst-null mice. In each of these mutants, septate-like junctions are missing. Further, the components of septate-like junctions (NF-155, Caspr, and contactin) are mislocalized in a diVuse pattern, and the axonal proteins that are normally largely excluded from the paranodal region—Kv1.1, Kv1.2, Kvb2, and Caspr2—are apposed to the nodal region. Similar anatomical changes (missing septate-like junctions, retraction of oligodendrocyte glial loops) have also been observed in contactin- and Caspr-null mice.

1994; Davis et al., 1993, 1996; Srinivasan et al., 1988). NF186 and Nr-CAM share an intracellular epitope (FIGQY) whose tyrosine must be dephosphorylated for them to bind to ankyrinG (Davis et al., 1993; Garver et al., 1997; Zhang and Bennett, 1998). Bennett and colleagues (Bennett et al., 1997; Davis et al., 1996; Lambert et al., 1997) have proposed that NF186 and NrCAM have heterophilic interactions in trans with other CAMs on the microvilli (Fig. 4.5), in accord with the ultrastructural data showing tethering of the microvilli to the nodal axolemma (Ichimura and Ellisman, 1991; Raine, 1982), as depicted in Figure 4.2.

MOLECULAR SPECIALIZATIONS OF THE PNS AND CNS NODAL AXOLEMMA

93

FIGURE 4.5 Possible cis and trans interactions in the PNS nodal region. This schematic drawing depicts the molecular organization of nodes, paranodes, and juxtaparanodes. At nodes, the ANK domains of ankryinG are depicted encircling the cytoplasmic aspect of the Nav channel a subunit (Michaely et al., 2002); the Nav channels b subunits as well as tenascin-R and tenascin-C are depicted as interacting with the globular domain of ankryinG, which in turn interacts with spectrin bIVS1. The extracellular domains of b2 subunits may interact with tenascin-R, tenascin-C, and NF186. Nr-CAM, NF186, and b subunits may interact in trans with CAMs on the Schwann cell microvilli. At paranodes, Caspr and contactin multimers interact in trans with NF155. At juxtaparanodes, TAG-1 is depicted as dimers that interact homophilically in trans (Freigang et al., 2000; Kunz et al., 2002). On the Schwann cell side, TAG-1 may also participate in the formation of a complex that consists of Cx29 and other proteins. On the axon side, TAG1 inteacts with Caspr2 (Poliak et al., 2003), which also interacts with a multimeric PDZ domain protein that links Caspr2 and tetramers of Kv1.1/ Kv1.2 channels (Poliak et al., 1999). Protein 4.1B links the cytoplasmic tail of Caspr and Caspr2 to the spectrin cytoskeleton. Homotypic gap junctions comprised of Cx29 or Cx32 link the paranodal membranes of the myelin sheath; the paranodal loops contain Kir4.1, an inwardly rectifying Kþ channel, and the adaxonal juxtaparanodal membrane contains Cx29, putative hemichannels.

In keeping with the idea that ankyrinG is an adaptor protein, inactivation of the ankyrinG gene in the cerebellum results in the failure of Nav1.6, NF186, Nr-CAM, and spectrin bIVS1 to cluster in the initial segments of Purkinje cells (Jenkins and Bennett, 2001), which accordingly have a diminished ability to initiate axon potentials (Zhou et al., 1998a). Although initial segments and nodes share many molecular characteristics, for unknown reasons, the nodal membranes of Purkinje cells are less aVected, as they have clusters of Nav1.6, neurofascin, Nr-CAM, and spectrin bIVS1 (Jenkins and Bennett, 2001). AnkyrinG also interacts with spectrin, likely the splice variant of spectrin IV (bIVS1) that is speciWcally localized to initial segments and nodes (Berghs et al., 2000;

94

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

FIGURE 4.6 ERM proteins in Schwann cell microvilli. A–D: A confocal reconstruction of a teased myelinated Wber from a rat sciatic nerve, labeled with a pan-ERM antiserum (A), a mouse monoclonal antibody against Nav channels (B), and a rat monoclonal antibody against a phosphorylated epitope of neuroWlament heavy (NF-H, C); the merged image is shown in panel D. The insets in panels A and B show the superimposed NF-H staining. At the node (double arrowheads), note that the ERMs form a larger diameter disk than do Nav channels. (E) A single 0.5 mm thick optical section of a node of Ranvier; taken from a section of an unWxed ventral root, double-labeled with a rabbit antiserum against ezrin (red) and a mouse monoclonal antibody against Nav channels (green). Note that the ring of ERMs is larger than the ring of Nav channels. Scale bar: A-D, 10 mm; E, 1 mm. From Scherer et al., 2001, with permission of Wiley-Liss.

Koenig and Repasky, 1985; Trapp et al., 1989b). Proof of a role for spectrin IV in this regard comes from the analysis of quivering mice, in which recessive mutations in the spectrin 4 gene cause altered ion channel distributions in myelinated axons (Komada and Soriano, 2002; Parkinson et al., 2001). Other molecules have been localized to the nodal axolemma. An isoform of Naþ/KþATPase was localized to CNS nodes in goldWsh, and PNS nodes in mammals (Ariyasu et al., 1985; Mata et al., 1991; Schwartz et al., 1981; Vorbrodt et al., 1982; Wood et al., 1977). The high concentrations of Naþ/Kþ-ATPase is in keeping with the physiological function of the nodal membrane. Mammalian nodes have fast, intermediate, and slow Kþ conductances (Reid et al., 1999; Safronov et al., 1993); two of these Kþ a subunits, Kv3.1b and KCNQ2, have now been identiWed (Devaux et al., 2003a, b), but their functional importance remains to be determined.

NODAL SPECIALIZATIONS IN THE PNS (FIGS. 4.2–4.6) In the PNS, the extracellular matrix in the nodal region can be selectively stained with methylene blue and a variety of metal salts—the ‘‘nodal gap substance’’ (Hess and Young, 1952; Landon and Langley, 1971; Quick and Waxman, 1977). Hyaluronidase treatment abolishes staining by some kinds of metal salts (Landon and Langley, 1971), and hyaluronic acid and the hyaluronate-binding domain of versican are localized to the nodal gap (Abood and Abul-Haj, 1956; Apostolski et al., 1994; Delpech et al., 1982). The nodal gap is also stained by GriVonia simplicifolia-B4 isolectin and peanut agglutinin (PNA), lectins that recognize terminal a- and b-D-galactose, respectively (Apostolski et al., 1994; Corbo et al., 1993; Streit et al., 1985). The actual molecule that GriVonia simplicifolia-B4 isolectin

NODAL SPECIALIZATIONS IN THE CNS

recognizes in the nodal gap is not known. PNA binds to terminal galactosyl b1-3 N-acetylgalactosamine [Gal(b1-3)GalNAc], an epitope that is found on many glycoproteins including versican (Apostolski et al., 1994; Delpech et al., 1982). PNA also binds to glycolipids, including the gangliosides GM1, asialo GM1, and GD1b (Latov, 1990). GD1b and asialo GM1 have not been localized to nodes (Kusunoki et al., 1993; Kusunoki et al., 1997). Whether GM1 is present at nodes is not resolved, as cholera toxin (which binds to GM1) labels nodes (Corbo et al., 1993; Ganser et al., 1983; Goodyear et al., 1999; Sheikh et al., 1999; Thomas et al., 1991), whereas antibodies against GM1 typically do not (Gong et al., 2002; Molander et al., 1997; Sheikh et al., 1999). The lateral borders of the Schwann cell cytoplasm have microvilli. By freeze-fracture EM, their membranes appear similar to the outer (abaxonal) membrane of myelinating Schwann cells (Blanchard et al., 1985; Devor et al., 1993; Ritchie et al., 1990; Waxman and Black, 1987). Nevertheless, Kir2.1 and Kir2.3, two inwardly rectifying Kþ channels are enriched in the microvilli; these have been proposed to redistribute Kþ (Mi et al., 1996). Two other kinds of Kþ channels, Kv1.5 and Slo1, are enriched in the abaxonal Schwann cell membrane, but not in the microvilli per se (Mi et al., 1999; Mi et al., 1996). The proximal portions of microvilli, even those from diVerent Schwann cells, appear to be connected to one another by tight junctions (Berthold and Rydmark, 1983), and claudin-2 is speciWcally localized to this region (Poliak et al., 2002). Like microvilli in other tissues, nodal microvilli contain F-actin (Trapp et al., 1989b; Zimmermann, 1996), ezrin, radixin, and moesin (the deWning members of the ERM family of proteins), as well as ezrin-binding protein (Hayashi et al., 1999; Melendez-Vasquez et al., 2001; Scherer et al., 2001). ERM proteins bind to actin Wlaments mainly via their C-termini and can associate with a number of diVerent integral membrane proteins via their N-termini. It remains to be determined what integral membrane proteins are associated with ERM proteins in Schwann cell microvilli and whether their phosphorylation regulates their function in this setting (Hayashi et al., 1999). Both Nr-CAM and neurofascin can have homophilic and heterophilic interactions with a variety of CAMs (Volkmer et al., 1996, 1998). The nodal basal lamina and/or the nodal gap have also been reported to be enriched in several extracellular matrix proteins and CAMs, including tenascin-R, tenascin-C, NG2, N-CAM, and L1 (DaniloV et al., 1986, 1989; Martin et al., 1990, 2001; Mege et al., 1992; Mirsky et al., 1986; Rieger et al., 1986). It is possible that some of these molecules may be associated with the Schwann cell microvilli rather than the extracellular matrix and/or the basal lamina.

NODAL SPECIALIZATIONS IN THE CNS (FIGS. 4.4 AND 4.7) The axolemma of CNS nodes is thought to be similarly organized as those in the PNS. The main diVerence between CNS and PNS nodes is that processes from astrocytes (Black and Waxman, 1988; Butt et al., 1994) and oligodendrocyte progenitor cells (Butt et al., 1999) appose the nodal axolemma. These processes are not as well developed as the Schwann cell microvilli, especially for small axons (Bjartmar et al., 1994; Raine, 1984), but they may be important for clustering Nav channels, as clusters of Nav channels are associated with astrocytic processes, even in demyelinating diseases (Arroyo et al., 2002; Rosenbluth, 1985; Rosenbluth et al., 1985). These astrocytic processes express the HNK1 epitope, as well as tenascin-R (Bartsch et al., 1993; Vrench-Constant et al., 1986), which can bind to Nav channels (Srinivasan et al., 1998), and alter their electrophysiological properties (Xiao et al., 1999). These interactions may be the reason that myelinated optic axons in mice lacking tenascin-R have slowed conduction velocities (Weber et al., 1999). The extracellular matrix of CNS nodes also contains versican, which is made by oligodendrocytes (Asher et al., 2002; Delpech et al., 1982), and Bral1, a brain-speciWc link protein that binds to versican, which is made by neurons (Oohashi et al., 2002). CNS nodes also contain RPTPb, which can interact with tenascin-R and both the a and the b subunits of Nav channels (Milev et al., 1998; Weber et al., 1999; Xiao et al., 1997).

95

96

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

FIGURE 4.7 The ultrastructure of the paranode. (A) A transmission electron micrograph showing a CNS paranode in longitudinal section. The terminal loops of the myelin sheath (note the Xuting of the axonal membrane in the upper aspect of the image) are connected to the axonal membrane by septate-like junctions. From Hirano and Dembitzer, 1982, with permission of Kluwer Academic Press. (B) A transmission electron micrograph showing lanthanum outlining the space between the septate-like junctions of CNS axons. The arrows indicate the locations of the spaces between adjacent glial loops; these spaces are larger than the spaces between adjacent rows of septate-like junctions. From Hirano and Dembitzer Hirano and Dembitzer, 1982, with permission of Kluwer Academic Press. (C) A longitudinal freeze-fracture EM through a CNS node and its paranodes. ‘‘In the middle of the Weld is the A face of the nodal axon (NdA), which is studded with randomly distributed membrane particles. On each side of this are the A faces of the paranodal axolemma (AlA). The one at the bottom shows the scalloping (arrows) producted by the pockets of paranodal cytoplasm that are visible at the sides (P). The fracture plane then passes across the axon (Ax) so that the B face (AlB) of the axolemma on the other side comes into view.’’ From Peters et al., 1991, with permission of W.B. Saunders.

The signiWcance of these interactions are not clear, however, as the distribution of nodal Nav channels, as well as conduction velocity of CNS myelinated axons, is normal in RPTPb-deWcient mice (Harroch et al., 2000). Both tenascin-R and RPTPb also interact with contactin (Peles et al., 1995; Pesheva et al., 1993), which, in contrast to the PNS, is found at CNS nodes (Rios et al., 2000). Nodes contain a high MW form of contactin,

SPECIALIZATIONS AT PARANODES

97

FIGURE 4.8 Paranodal gap junctions in the CNS myelin sheath. (A) ‘‘Junctions at the glial loops, diagram. The fracture plane inside the glial membrane has been illustrated as a plane which may be viewed from either side to disclose the glial P face and E face. The interglial junctions consist of junctions between consecutive gyres in the helix of glial loops (GL). One or two lines of tight junctions (TJ) follow the margin on the interglial P face (GL-GL PF). These ridges may consist of tightly arranged rows of particles or a series of bars of varying length. Occasional gap junctions (GJ) appear between the lines of tight junctions.’’ (B) ‘‘Glial-glial junctions in rat spinal cord. The fracture has passed obliquely through the axon (Ax) at the junction of the last turn of the glial loop seen as the axonal E face (Ax EF) and the node of Ranvier (n R). Several glial loops which have piled up at the margin of the node without contact with the axolemma show lines of tight junctions (TJ) and dense particle accumulations representing gap junctions.’’ Figures and quotations are from Sandri et al., 1982, with permission of Elsevier Press.

whereas paranodes contain a low MW form (Rios et al., 2000). The Wnding that Caspr induces a high to low MW switch and reduced levels of nodal contactin (Gollan et al., 2003) is in keeping with the higher levels of nodal contactin in Caspr-null mice (Bhat et al., 2001). Contactin regulates the surface expression of Nav channels (Kazarinova-Noyes et al., 2001; Liu et al., 2001), and mice lacking contactin have reduced nodal Nav channel clusters (Kazarinova-Noyes and Shrager, 2002).

SPECIALIZATIONS AT PARANODES (FIGS. 4.2–4.9) At paranodes, the lateral edge of the myelin sheath spirals around the axon, forming the axoglial junctions. There are also ‘‘reXexive’’/‘‘autotypic’’ junctions between the paranodal loops themselves, including tight junctions, gap junctions, and adherens junctions (Arroyo and Scherer, 2000; Scherer and Arroyo, 2002; Spiegel and Peles, 2002). In the PNS, these reXexive junctions are also localized to incisures as well as to the inner and outer mesaxons; whether CNS ‘‘incisures’’ have any of these specializations remains to be demonstrated. Paranodal glial and axonal membranes have distinct molecular specializations (Girault and Peles, 2002). MAG is enriched between the glial loops in the PNS but not in the CNS (Trapp et al., 1989, 1989b). Glial loops are enriched in inward rectifying Kþ channel Kir4.1 (Chen et al., 2002b; Wilson and Chiu, 1990), and Kir4.1-null mice develop profound CNS demyelination (Neusch et al., 2001). PNS paranodes are enriched in oligodendrocyte-myelin glycoprotein (Apostolski et al., 1994), and the gangliosides GD1b and

98

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

FIGURE 4.9 Localization of Caspr and Kv1.1 in the paranode and internode. This is a confocal reconstruction of a teased myelinated Wber from an adult mouse sciatic nerve, labeled with a rabbit antiserum against Caspr (red), a mouse monoclonal antibody against Kv1.1 (green), and a rat monoclonal antibody against NF-H (blue). For clarity, the labeling from each antibody is shown separately in panels A–E; panels G and H show the merged images; panels B, D, G, and H are enlargements. Note the separation of Caspr and Kv1.1 staining at the paranode (large arrows) and juxtaparanode (large arrowheads), and that the spiral of Caspr staining in the juxtaparanodal region Wlls a void in the Kv1.1 staining. In the internodal region, the double line of Kv1.1 staining Xanks the single line of Caspr staining. NF-H staining is diminished at the node (double arrowheads). Scale bars: 5 mm. From Arroyo et al., 1999, with permission of Kluwer Academic Press.

GQ1b (Chiba et al., 1993; Kusunoki et al., 1993). Finally, there are P2 ATP receptors in the paranodal membrane (Grafe et al., 1999)

Tight Junctions

au4

Tight junctions are specialized cell-cell contact sites between cells that act as a selective permeability barrier (gate function) and as an intramembranous fence, restricting the movement of lipids and proteins from speciWc membrane domains (Mitic and Anderson, 1998). In myelinating Schwann cells and oligodendrocytes, freeze fracture EM reveals tight junction strands comprised of linear rows of intermembranous particles between adjacent membranes of the paranodal loops, inner mesaxons, and outer mesaxons (Mugnaini and Schnapp, 1974; Sandri et al., 1982; TetzlaV, 1982). Tight junctions in the paranodal loops prevent large molecules from entering the potential space between the layers of the myelin sheath (Hall and Williams, 1971; MacKenzie et al., 1984b; Revel and Hamilton, 1969; Tabira et al., 1978). These junctions may also stabilize newly formed wraps of myelin during development. In PNS myelin sheaths, tight junctions are also found in incisures (Sandri et al., 1982) and between apposed microvilli (Berthold and Rydmark, 1983). In CNS myelin sheaths, tight junctions are found in the so-called ‘‘radial component,’’ a radial stack of tight junctions that typically extends through a sector of compact myelin (Chapter 1 by Trapp and Kidd). In epithelial cells, tight junctions contain large protein complexes, consisting of transmembrane proteins (claudins, occludin, JAM, and Crumb), which are linked to the actin

SPECIALIZATIONS AT PARANODES

cytoskeleton through an array of cytoplasmic adaptors (Tsukita et al., 2001). Claudins are integral membrane proteins with four transmembrane domains that are exclusively localized at tight junctions and are both suYcient and necessary for their formation (Furuse et al., 1998; Tsukita and Furuse, 2000). In the PNS, myelinating Schwann cells contain several claudins that are diVerentially localized in regions of noncompact myelin— claudin-5 is found at the paranodes, claudin-1 in incisures, and claudin-2 is located at junctions formed between nodal microvilli (Poliak et al., 2002). Similar to epithelial cells, tight junctions between the paranodal loops also contain three multi-PDZ domain proteins MUPP1, ZO-1 and ZO-2, as well as the adapter protein Par-3 (Parmantier et al., 1999; Poliak et al., 2002). Claudin-11 is present in tight junctions that are mainly found at the radial component (Gow et al., 1999; Morita et al., 1999). These tight junction strands in CNS myelin are absent in mice lacking claudin-11; presumably this is somehow related to their slowed conduction velocities (Gow et al., 1999).

Gap Junctions

au5

Gap junction-like plaques are enclosed between tight junction strands in both the PNS and the CNS (Fig. 4.8). Gap junctions have not been detected by transmission EM, likely because of their small size (Wagner and Kachar, 1995). Connexin29 (Cx29) and Cx32 likely form the gap junctions in myelinating Schwann cells (Altevogt et al., 2002; BergoVen et al., 1993; Chandross et al., 1996; Li et al., 2002; Scherer et al., 1995). Although oligodendrocytes also express Cx29 and Cx32, these connexins have not been consistently localized to paranodes (Li et al., 1997; Scherer et al., 1995). Unlike Schwann cells, oligodendrocytes also have large gap junction plaques on their somata and outer/abaxonal cell membrane; these are coupled to astrocytic gap junctions containing Cx30 and Cx43 (Rash et al., 2001). A role for gap junctions in the myelin sheath was not established until it was discovered that mutations in the gene encoding Cx32, GJB1, cause X-linked CMT (Chapter 39 by Wrabetz et al.). Dye transfer studies in living myelinated Wbers provide functional evidence that gap junctions mediate a radial pathway of diVusion across incisures (Balice-Gordon et al., 1998). A radial pathway would be advantageous as it provides a much shorter pathway (up to 1000fold), owing to the geometry of the myelin sheath. Disruption of this radial pathway may be the reason that GJB1 mutations cause CMTX. However, the pathway and the rate of 5,6carboxyXuorescein diVusion in Gjb1/cx32-null mice did not appear to be diVerent than in wild type mice (Balice-Gordon et al., 1998), implying that another connexin(s) forms functional gap junctions in PNS myelin sheaths. This connexin now appears to be Cx29, which is localized in incisures, but not necessarily in the same gap junction plaques (Altevogt et al., 2002). What molecules diVuse through these gap junctions is not known.

Adherens Junctions These junctions, commonly referred to as ‘‘desmosome-like’’ junctions in the older literature, are found in mesaxons, paranodes, and incisures (Fannon et al., 1995; Hall and Williams, 1970). In paranodes and incisures, these form a series of radially arranged junctions that typically span many layers, most prominently in the outer layers of the myelin sheath (Fig. 4.3). Adherens junctions contain E-cadherin, a Ca2þ-dependent CAM that forms ‘‘strand dimers’’ that bind homophilically in trans, thereby linking apposing membranes (Shapiro et al., 1995; Takeichi, 1990). The cytoplasmic domain of E-cadherin binds a- and b-catenin (Fannon et al., 1995), which link E-cadherin to the actin cytoskeleton (Nagafuchi et al., 1993). Conditionally deleting E-cadherin in myelinating Schwann cells results in the loss of adherens junctions, but remarkably, does not cause any other structural alterations in the myelin sheath (Young et al., 2002).

Septate-Like Junctions In addition to the autotypic junctions mentioned previously, a specialized septate-like junction is formed between the axon and the paranodal loops of oligodendrocytes or

99

100

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

myelinating Schwann cells. In freeze-fracture EM, the paranodal loops of the myelin sheath contain rows of large particles that are in register with a double row of smaller particles on the axolemma (Ichimura and Ellisman, 1991; Sandri et al., 1982; Thomas et al., 1993). These particles correspond to the so-called ‘‘terminal bands’’ seen by transmission EM, and have been more recently termed septate-like junctions, as they resemble invertebrate septate junctions (Einheber et al., 1997). Septate junctions may function similarly to vertebrate tight junctions, forming intercellular junctions that prevent the diVusion of small molecules and ions. Septate-like junctions, however, do not prevent the diVusion of lanthanum or even microperoxidase (molecular mass 5 kDa) into the periaxonal space, and thus are not ‘‘tight’’ in the conventional sense (Feder, 1971; Hirano et al., 1969; MacKenzie et al., 1984a). An intrinsic membrane glycoprotein called Caspr (contactin-associated protein; also known as paranodin) was originally found in a neuronal complex with contactin that bound to the glial RPTPb (Peles et al., 1995; 1997b). Casprs are transmembrane molecules that form a distinct subgroup within the neurexin superfamily, a polymorphic family of proteins involved in cell adhesion and intercellular communication (Baumgartner et al., 1996; Bellen et al., 1998; Missler et al., 1998; Peles et al., 1997a; Poliak et al., 1999; Ullrich et al., 1995; Ushkaryov et al., 1992). There are Wve human genes in the Caspr family (Caspr-Caspr5; Peles et al., 1997b; Poliak et al., 1999; Spiegel et al., 2002), two in Drosophila (neurexin IV and axotactin; Baumgartner et al., 1996; Poliak et al., 1999; Yuan and Ganetzky, 1999), and two in C. elegans (intexin and Nlr; Haklai-Topper and Peles, unpublished). Caspr proteins contain a variety of subdomains also found in proteins that have been implicated in synaptogenesis, axonal guidance, and target recognition (Peles et al., 1997b). Their extracellular region consist of a mosaic of domains implicated in mediating protein-protein interactions, including a discoidin and a Wbrinogen-like domains, EGF motifs, and several regions with homology to the G domain of laminin A, thought to mediate cell adhesion. As shown in Figures 4.5 and 4.9, Caspr is localized to the paranodal axolemma in myelinated Wbers of the PNS and CNS (Einheber et al., 1997; Menegoz et al., 1997; Peles et al., 1997b). Caspr forms heterodimers in cis with contactin (Peles et al., 1997b), and the two are colocalized at paranodes, across from NF155, a glial isoform of neurofascin that is located at the paranodal loops (Einheber et al., 1997; Menegoz et al., 1997; Rios et al., 2000; Tait et al., 2000). The paranodal accumulation of Caspr is composed of a series of spirals, one spiral for each spiral of the myelin sheath; the paranode is thus not a uniform domain. During the myelination of sensory axons by Schwann cells in vitro, a loose spiral of Caspr staining, corresponding to the turns of the forming paranodal loops, consolidates into a tight helical coil of the mature paranode (Pedraza et al., 2001). These junctions appear relatively late during myelination, Wrst generated closer to the nodes by the most outer paranodal loop and then forming gradually as additional loops are attached to the axon (Einheber et al., 1997). Both Caspr and contactin are essential for the formation of the paranodal junction. In the absence of either one, the ultrastructure of the paranodes is severely altered: the gap between glial and the axonal membranes is increased, and the septa themselves are absent (Bhat et al., 2001; Boyle et al., 2001). Just as the eYcient export of Caspr from the endoplasmic reticulum to the plasma membrane of transfected cells requires contactin (Faivre-Sarrailh et al., 2000), Caspr is retained neuronal cell bodies and does not reach the axons in contactin-deWcient mice (Boyle et al., 2001). Conversely, Caspr is necessary to maintain contactin at the paranodes (Bhat et al., 2001; Gollan et al., 2002), where the Caspr/contactin complex mediates axon-glia contact by binding in trans to NF155 (Charles et al., 2002; Tait et al., 2000). If the absence of either contactin or Caspr, NF155 is not localized to paranodes and septate-like junctions are not formed (Bhat et al., 2001; Boyle et al., 2001; Marcus and Popko, 2002; Poliak et al., 2001). Thus, septatelike junctions contain a tripartite adhesion complex and likely additional proteins (Gollan et al., 2003; Marcus et al., 2002; Poliak et al., 2001). The intracellular regions of Caspr and Caspr2 contain a juxtamembrane sequence that binds protein 4.1B, which is present at paranodes and juxtaparanodes (Ohara et al., 2000;

SPECIALIZATIONS AT JUXTAPARANODES

au6

Parra et al., 2000). Protein 4.1B could thus immobilize Caspr (and hence contactin) to the cytoskeleton (Gollan et al., 2002). Consistent with this notion, protein 4.1B is abnormally distributed along peripheral myelinated axons of mice lacking either contactin and galactolipids, both of which lack paranodal Caspr (Gollan et al., 2002; Poliak et al., 2001). In these mutants, the position of protein 4.1B is strongly correlated with those of Caspr and Caspr2, suggesting that they determine its localization. Further, the cytoplasmic tail of Caspr is required for stabilizing the Caspr/contactin complex at paranodes, as a Caspr mutant lacking this domain is not properly maintained at paranodes (Gollan et al., 2002). Thus, Caspr appears to serve as a transmembrane scaVold that stabilizes the Caspr/ contactin adhesion complex at septate-like junctions by connecting it to the axonal cytoskeleton via protein 4.1B. This mechanism closely resembles the function of Drosophila neurexin IV, which recruits Coracle (the homologoue of protein 4.1) to septate junctions (Baumgartner et al., 1996; Lamb et al., 1998). Further discussion regarding the morphological and molecular similarity of the paranodal and invertebrate septate junctions can be found in the Chapter 8 by Bellen and Schulze.

SPECIALIZATIONS AT JUXTAPARANODES (FIGS. 4.3–4.10) In the region extending 10 to 15 mm from the paranode, the axolemma contains clusters of 5 to 6 particles in freeze-fracture EM (Miller and Da Silva, 1977; Rosenbluth, 1976; Stolinski et al., 1981; Stolinski et al., 1985; Tao-Cheng and Rosenbluth, 1984). The distribution of these juxtaparanodal particles corresponds to the distribution of delayed rectifying Kþ channels (Chiu and Ritchie, 1980), subsequently identiWed as Kv1.1 and Kv1.2 and their associated b2 subunit (Arroyo et al., 1999; Gulbis et al., 1999; Mi et al., 1995; Rasband et al., 1998; Vabnick and Shrager, 1998; Wang et al., 1993; Zhou et al., 1998b). Both Kv1.1 and Kv1.2 can freely mix in varying proportions to form tetramers, the functional channels (Hopkins et al., 1994). Because Kv1.1/Kv1.2 channels appear to be concealed under the myelin sheath in adult nerves (Chiu and Ritchie, 1981, 1982; Corrette et al., 1991), Kþ channel blockers have little eVect (Chiu and Ritchie, 1980; Hildebrand et al., 1994; Kocsis et al., 1983). In developing nerves, demyelinated nerves, and regenerated nerves, however, Kþ channel blockers have more profound eVects, probably related to an abnormal distribution or number of Kv1.1/Kv1.2 channels (Bostock et al., 1981; Bowe et al., 1989, 1994; Rasband et al., 1998; Vabnick and Shrager, 1998). Caspr2 is localized to the juxtaparanodes of myelinated Wbers in both the CNS and the PNS, colocalizing with Kv1.1/1.2/b2 (Arroyo et al., 2001; Poliak et al., 1999, 2001). In addition to a protein 4.1B binding site, Caspr2, like Kv1.1 and Kv1.2, has an intracellular PDZ binding domain. At the juxtaparanodal region, Caspr2 physically associates with Kv1.1, Kv1.2, and their Kvb2 subunit (Poliak et al., 1999). This association is mediated by the C-terminal region of Caspr2 and is likely mediated by a PDZ protein with multiple PDZ domains (Poliak et al., 1999), akin to ones that cluster of ion channels and other membrane proteins (Sheng and Sala, 2001). So far, only one such PDZ protein, PSD95, has been found at the juxtaparanodal region (Baba et al., 1999), but it is not required for the formation of Caspr2/Kv1.1/Kv1.2 complexes or for their clustering in myelinated axons (Rasband et al., 2002). Finally, TAG-1, a GPI-linked CAM that is closely related to contactin, is a component of the axonal juxtaparanodes and associates with Caspr2/ Kv1.1/Kv1.2 complexes at this site (Poliak et al., 2003). The localization of Caspr2 and TAG-1 at juxtaparanodes is interdependent, and both are essential for clustering of Kv1.1/ Kv1.2 channels (Poliak et al., 2003), suggesting that they are primary constituents of the mechanism by which glial cells control the localization of ion channels at the juxtaparanodal axonal membrane. There are complementary specializations in the juxaparanodal (adaxonal) membrane of myelinating Schwann cells (Fig. 4.11). Cx29 is localized in the juxtaparanodal region of myelinating Schwann cells and oligodendrocytes (Altevogt et al., 2002; Li et al., 2002); this distribution appears to correspond to gap junction-like rosettes of particles in freeze-fracture EM (Li et al., 2002; Stolinski and Breathnach, 1982; Stolinski et al., 1981,

101

102

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

FIGURE 4.10 The localization of Cx29 in PNS and CNS myelin sheaths. (A) These are images of a teased myelinated Wber from an adult rat, immunostained for Cx29 and Kv1.2. The node is marked by apposed arrowheads. Note that Cx29 staining does not overlap with that of Kv1.2 at paranodes and incisures (red staining in the merged image). Cx29 and Kv1.2 appear to overlap completely (yellow color in the merged image) at the juxtaparanode, the inner mesaxons (arrows), and the inner most aspect of incisures (the band that is perpendicular to the long axis of the axon). (B) These are images of a longitudinal section of adult mouse spinal cord, immunostained for Cx29, and Caspr. Cx29 is localized to the juxtaparanodal region of small (arrowheads) but not large myelinated axons, as demarked by the sizes of Casprpositive paranodes. ModiWed from Altevogt et al., 2002, with permission of the Society for Neuroscience.

1985). These Wndings raise the possibility that Cx29 may form hemi-channels in the adaxonal membrane; how such channels would function presents an interesting challenge for further work. In addition to Cx29, TAG-1 has also been found in the juxtaparanodal region of myelinating Schwann cells and oligodendrocytes (Traka et al., 2002). Because TAG-1 can interact homophilically (Felsenfeld et al., 1994), there may be an adhesion complex consisting of a glial TAG-1 molecule and an axonal Caspr2/TAG-1 heterodimer. Thus, it appears that two members of the Caspr family play a role in the organization of the nodal environs by two diVerent mechanisms: Caspr is involved in the generation of a barrier-like structure at the paranodal junction (discussed later), restricting the movement of ion channels and associated proteins from beneath the compact myelin, whereas Caspr2 serves as a scaVold that maintains Kþ channels in the juxtaparanodal region.

THE INTERNODAL REGION (FIGS. 4.3, 4.4, 4.9–11) The internodal axonal membrane of the PNS is organized in concert with its myelin sheath. This was Wrst revealed by freeze-fracture EM, which showed that internodes had similar kinds of intramembranous particles as the juxtaparanodal region (Miller and Da Silva, 1977; Stolinski and Breathnach, 1982; Stolinski et al., 1981, 1985). As shown in Figure 4.11, the axonal membrane apposing the internal mesaxon and incisures has paranodaland juxtaparanodal-type particles, which have been termed ‘‘juxta-mesaxonal’’ and ‘‘juxta-

DEVELOPMENTAL ASSEMBLY OF THE NODAL REGION

FIGURE 4.11 Internodal staining organization of myelinated PNS axons. This schematic drawing depicts the organization of the axolemma and the adaxonal Schwann cell membrane, as seen by freeze-fracture electron microscopy. On both apposed membranes, a single row of intramembranous particles is Xanked by a double row of intramembranous particles. The identity of the intramembranous particles is indicated. ModiWed from Stolinski et al., 1985, with permission of Elsevier Press.

incisural’’ (Peles and Salzer, 2000; Scherer and Arroyo, 2002). In accord with these Wndings, the internodal membranes of PNS axons have a tripartite strand, consisting of a central strand of Caspr/contactin staining, Xanked by strands of Kv1.1/1.2/b2/Caspr2immunoreactivity (Arroyo et al., 1999; Rios et al., 2000). More recently, it was found that NF155 (Tait et al., 2000), Cx29 (Altevogt et al., 2002), and TAG-1 (Traka et al., 2002) are localized in a complementary distribution on the adaxonal membrane of myelinating Schwann cells. Although one might presume that the same trans molecular interactions that link these molecules in the paranode and juxtaparanode also do so in internodal region, septate-like junctions are not features of the internodal region. Furthermore, despite the marked reduction in the juxtaparanodal accumulation of Kþ channels observed in Caspr2- and TAG-1-deWcient mice, the internodal localization of these channels appeared normal, suggesting that diVerent mechanisms control their localization at the juxtaparanodal region and along the internodes (Poliak et al., 2003). In myelinated CNS axons, Caspr appears to appose the inner mesaxon near the paranode (Menegoz et al., 1997), although it juxta-mesaxonal staining is seldom extends into the internodal region (Arroyo et al., 1999). Unlike the PNS, however, Kv1.1, Kv1.2, and Kvb2 are not localized in a ‘‘juxta-mesaxonal’’ pattern (Arroyo et al., 2001), and Cx29 and TAG-1 are not localized in a mesaxonal pattern (Altevogt et al., 2002; Tait et al., 2000; Traka et al., 2002). In keeping with the idea that the incisures of CNS myelin sheaths are not equivalent to their PNS counterparts, NF155, Cx29, and TAG-1 do not appear to localized in CNS incisures (Altevogt et al., 2002; Tait et al., 2000; Traka et al., 2002), and juxta-incisural staining of Caspr, Kv1.1, Kv1.2, and NF155 has not been seen (Arroyo et al., 2001). Thus, if the internodal localization of axonal proteins is dictated by the myelin sheath, then oligodendrocytes appear to lack the ability to organize the internodal domains myelinated axons.

DEVELOPMENTAL ASSEMBLY OF THE NODAL REGION Myelinating glial cells organize the axonal membrane in the developing PNS and CNS (Rasband et al., 1999a, 1999b; Rasband and Shrager, 2000; Rasband and Trimmer, 2001; Vabnick and Shrager, 1998; Vabnick et al., 1999). Nav channels accumulate at the ends of developing myelin sheaths, and as two adjacent internodes elongate, these clusters fuse to form a node of Ranvier. Kv1.1 and Kv1.2 are subsequently excluded from the nodal axolemma and sequestered beneath the myelin sheath by the developing paranode, resulting in decreased sensitivity to Kþ channels blockers (Bowe et al., 1994; Devaux et

103

104

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

al., 2002; Vabnick and Shrager, 1998; Vabnick et al., 1999). The distribution of Nav channels and Kv1.1/Kv1.2 during remyelination of peripheral nerves appears to follow the same sequence of events (Rasband et al., 1998). In myelinating co-cultures, Schwann cells are required for the formation of nodes (Ching et al., 1999), and the localization of a nodal marker (Nav channels) appears before that of a paranodal marker (Caspr; (Einheber et al., 1997; Lustig et al., 2001). Based on the sequential appearance of antigens in immunostained developing nerves, Bennett and colleagues (Bennett et al., 1997; Davis et al., 1996; Lambert et al., 1997) proposed that neurofascin and Nr-CAM organize PNS nodes, recruiting ankryinG, which, in turn, recruits Nav channels. They were the Wrst to propose that neurofascin and NrCAM have heterophilic interactions with other CAMs on Schwann cell microvilli as depicted in Figure 4.5B (Bennett et al., 1997). The most compelling evidence for this idea comes from experiments in which Fc-Nr-CAM fusion protein was added to myelinating co-cultures (Lustig et al., 2001). The fusion protein clustered axonal neurofascin, and inhibited the focal accumulation of Nav channels and ankyrinG. The Wnding that microvilli are co-localized with clusters of Nav channels in developing nerves (Melendez-Vasquez et al., 2001) suggests that microvilli may be the source of an endogenous ligand for axonal neurofascin. In the CNS, there may be a diVerent temporal order. In the initial segments of Purkinje cells, ankyrinG and spectrin bIVS1 appear before Nav1.6, Nr-CAM, and neurofascin (Jenkins and Bennett, 2001), but it is possible that a diVerent Nav channel appears before Nav1.6 (Boiko et al., 2001; Kaplan et al., 2001). In developing nodes of the optic nerve, ankyrinG appears before spectrin bIVS1, Nav1.6, Nr-CAM, and neurofascin (Jenkins and Bennett, 2002). In another study of the developing optic nerve, paranodal Caspr staining preceeded that of Nav channels (Rasband et al., 1999a). However, the paranodal accumulation of Caspr or neurofascin155, or even the presence of septate-like junctions, are not required for the molecular assembly of nodes (Arroyo et al., 2002; Bhat et al., 2001; Jenkins and Bennett, 2002; Mathis et al., 2001; Popko, 2000). Although nodal specializations can develop in vitro without direct physical interaction with oligodendrocytes (Kaplan et al., 1997), whether this is the case in vivo is not known. Nodal specializations fail to develop in mice in which oligodendrocytes are ablated (Mathis et al., 2001). Glial ensheathment alone, and not myelin sheaths per se, appears to be suYcient for the formation node-like clusters, and these clusters probably persist for days after the death of oligodendrocytes (Arroyo et al., 2002; Jenkins and Bennett, 2002; Mathis et al., 2001).

FUNCTIONAL ATTRIBUTES OF THE NODAL REGION The Node Whereas their functional role in promoting saltatory conduction via clustered nodal Nav channels is well established, the diversity of the Nav family raise interesting questions. Why do certain nodes have Nav1.2, Nav1.6, Nav1.8, and/or Nav1.9? One presumes that their diVerent electrophysiological characteristics have been optimized by natural selection, but what are these attributes? Are these channels interchangeable? The most important insights to date are based on the loss of function mutations of the genes encoding Nav1.6 and Nav1.8. A recessively inherited mutation in the gene encoding Nav1.6 causes motor endplate disease (med). These mice have severe motor impairments and die of respiratory paralysis around 20 days post-natal, presumably because myelinated motor axons fail to conduct (Burgess et al., 1995). Motor axons, which are myelinated within a few days after birth, initially have nodal Nav1.2 (Rasband and Trimmer, 2001), but these presumably disappear during post-natal development. Not all myelinated axons depend on Nav1.6, as the conduction velocity in ‘‘mixed’’ nerves (containing myelinated sensory and motor axons) of med mice is minimally slowed (Duchen and Stefani, 1971; Rieger et al., 1984). Nav1.8-null mice have blunted pain responses (Laird et al., 2002), but this behavior is largely mediated by unmyelinated axons (Novakovic et al., 1998b).

FUNCTIONAL ATTRIBUTES OF THE NODAL REGION

The Paranode Forming septate-like junctions appears to be a crucial function of paranodes. Even though myelin sheaths are formed, septate-like junctions are absent in contactin- and Caspr-null mice, as well as UDP-galactose ceramide galactosyltransferase (cgt) and cerebroside sulfotransferase (cst)-null mice (Bhat et al., 2001; Bosio et al., 1996; Boyle et al., 2001; Coetzee et al., 1996, 1998; Honke et al., 2002; Ishibashi et al., 2002). Whereas it seems plausible that the absence of contactin or Caspr would result in the loss of septate-like junctions, it remains to be determined why they are absent in cst- and cgt-null mice. CGT and CST catalyze the sequential synthesis of galactocerebroside and sulfatide, which are glycolipids found in the myelin sheath. One possibility is that galactocerebroside and sulfatide are essential components of lipid rafts that transport NF155 to non-compact myelin. Alternatively, sulfatides may be ligands for the Caspr/contactin complex, localizing them to paranodes. The lack of septate-like junctions leads to a profound reorganization of the axonal membrane (Fig. 4.4B). In cgt- and cst-null mice, contactin and Caspr are not restricted to the paranode, but are more diVusely localized in the internodal membrane, and NF155 is not restricted to the paranodal loops (Dupree et al., 1999; Honke et al., 2002; Ishibashi et al., 2002; Poliak et al., 2001). In contrast to Nav channels, which are still concentrated in the nodes in all the aforementioned mutants, Kv1.1, Kv1.2, Caspr2, and TAG-1 are mislocalized to the paranodal axonal membrane (Bhat et al., 2001; Boyle et al., 2001; Coetzee et al., 1996; Traka et al., 2002). Mislocalized Kv1.1 and Kv1.2 is at least partially responsible for the slowed axonal conduction found in contactin-, Caspr, and cgt-null mice, and presumably accounts for the pronounced eVects of Kþ channel blockers in these mutants (Bhat et al., 2001; Boyle et al., 2001; Coetzee et al., 1996). Thus, sequestering Kv1.1/Kv1.2 away from the node is an essential function of the paranode.

The Juxtaparanode In spite of the robust phenotype of contactin-, Caspr-, cgt-, and cst-null mice, the function of juxtaparanodal Kv1.1/Kv1.2 channels is not well understood. Since they are electrically isolated from the node, they may not facilitate repolarization, but they may dampen excitability (Chiu, 1991), as Kþ channel blockers cause repetitive activity in developing nerve (Vabnick et al., 1999). Similarly, in spite of the striking reduction in the accumulation of Kþ channels at the juxtaparanodal region in Caspr2-null mice, there is no apparent change in nerve conduction in both adult and developing nerves (Poliak et al., 2003). Mice lacking the Kv1.1 gene (Kcna1) reveals abnormal impulse generators near the neuromuscular junctions (Smart et al., 1998; Zhou et al., 1998b), and KCNA1 mutations in humans cause a form of familial episodic ataxia that is associated with ectopic impulse generators in peripheral nerve (Adelman et al., 1995; Browne et al., 1994; Brunt and Van Weerden, 1990; Maylie et al., 2002; Van Dyke et al., 1975; Zerr et al., 1998). The striking similarity in the distribution of axonal Kv1.1/Kv1.2 channels and adaxonal Cx29 hemichannels raises the possibility of a Kþ exchange between axons and their myelin sheaths. It has been proposed that myelinating Schwann cells siphon Kþ that accumulates in the periaxonal space during neural activity (Chiu, 1991; Konishi, 1990). Paranodal Kir4.1 channels and perhaps juxtaparanodal Cx29 hemichannels could provide the channels for siphoning Kþ. Once inside the adaxonal cytoplasm of myelinating Schwann cell, Kþ could diVuse through gap junctions in the paranodes and incisures (Balice-Gordon et al., 1998). Just as spatial buVering of Kþ has been considered an important function of astrocytes in the CNS (Orkand et al., 1966), it may be an essential role of the molecular specializations of myelin sheaths. Although this proposed movement of Kþ remains to be shown, it has been hypothesized to be the trigger for the paranodal Ca2þ transients in Schwann cells that accompany the propagation of action potentials (Lev-Ram and Ellisman, 1995).

105

106

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

THE NODAL REGION AND DISEASE

au7 au8

Although their clinical features may vary, demyelination is a common denominator of the diVerent kinds of inherited and acquired demyelinating diseases discussed elsewhere in this book. Yet how demyelination aVects the organization of the nodal region has received surprisingly little attention. In animal models of acute demyelination in peripheral nerve, Shrager and colleagues have shown that nodal Nav channels and juxtaparanodal Kv1.1/Kv1.2 channels disperse and only reform after remyelination has begun (Dugandzija-Novakovic et al., 1995; Novakovic et al., 1996, 1998a; Rasband et al., 1998). If myelinating Schwann cells dictate the topography of the nodal region, this would account for many of the alterations found animal models of inherited demyelinating neuropathies, in which nodes, paranodes, and juxtaparanodes appear to be continuously reorganized by myelinating Schwann cells (Koszowski et al., 1998; Menichella et al., 2001; Neuberg et al., 1999). How demyelination alters the nodal architectures is less well studied in the CNS. Acute demyelination results in the failure to form properly clustered Nav channels as well as Caspr at paranodes (Mathis et al., 2000). Plpjimpy mice and Plpmyelin-deWcient rats (see Chapter 48 by Suter and Wrabetz) are animal models of severe Pelizaeus-Merbacher disease (see Chapter 37 by Hudson et al.). In these animals, oligodendrocytes ensheathe axons then die, creating node-like clusters of Nav channels that are not associated with paranodes (Arroyo et al., 2002; Jenkins and Bennett, 2002; Mathis et al., 2001). As in the mice lacking septate-like junctions described earlier, Kv1.1/Kv1.2 channels abut these node-like clusters of Nav channels. In contrast, in Mbpshiverer mice, which have severe CNS dysmyelination without oligodendrocytes cell death, paranodes are highly disorganized, and Kv1.1/Kv1.2 abut nodal clusters of Nav channels (Rasband and Trimmer, 2001; Rasband et al., 1999b; Tait et al., 2000). These data raise the possibility that the severity of the demyelinating disease may be related more to the degree of disorganization of the axonal membrane than to other, better known characteristics of these diseases. Various molecular components of myelinated axons appear to be the targets of certain autoimmune disorders. Antibodies to various components of the myelin sheath have been associated with diVerent neuropathies (Eurelings et al., 2001; Kwa et al., 2001; Lawlor et al., 2002; Ritz et al., 2000; Saperstein et al., 2001; Yan et al., 2001). In addition to these proteins, antibodies to various gangliosides have been implicated as causing various forms of acute and chronic demyelinating neuropathies: GM1 in multifocal motor neuropathy, GD1a in acute motor axonal neuropathy, GD1b in acute and chronic sensory demyelinating neuropathies, and GQ1b in Miller Fisher syndrome (Willison and Yuki, 2002). While it is plausible that other molecular components of the nodal region are the targets of autoantibodies, the only syndrome that has been recognized to date is acquired neuromyotonia (also known as Isaac’s syndrome) and a more pervasive variant, Morvan’s syndrome (Hart et al., 1997; Hayat et al., 2000; Liguori et al., 2001). In these diseases, there are ectopic impulse generators within peripheral nerves, caused autoantibodies against Kv1.1 or Kv1.2, that presumably interfere with their function.

CONCLUSION Although the role of the myelin sheath in supporting saltatory conduction has been surmised for 50 years, recent advances in the molecular anatomy of myelinated axons highlight several functional issues. One of the most surprising is that PNS myelin sheaths do not mainly act as electrical insulators of axons. The radial pathway for ion diVusion formed by gap junctions is not compatible with this common view; reducing capacitance is the likely role of myelin sheaths. Second, the intricate organization of the nodal, paranodal, juxtaparanodal, and internodal axonal domains relate closely to the organization of the myelin sheath itself. Although the details are still incomplete, the emerging evidence indicates that a molecular dialogue between myelinating glial cells and axons

CONCLUSION

au9 results in the local diVerentiation of their complementary cellular domains. Finally, the organization of the axonal membrane is disrupted by a number of mutations and disease processes that aVect myelin sheaths, with functional consequences that are beginning to make sense.

Acknowledgments Steven S. Scherer and Edgardo J. Arroyo have been supported by the National Institutes of Health (NS37100, NS34528, and NS08075), the Juvenile Diabetes Foundation, and the Charcot-Marie-Tooth Association. Elior Peles is supported by the National Multiple Sclerosis Society (RG3102), the Israeli Academy of Science, and the U.S.-Israel Binational Science Foundation. We thank our colleagues for their many contributions.

References Abood, L. G., and Abul-Haj, S. K. (1956). Histochemistry and characterization of hyaluronic acid in axons of peripheral nerve. J. Neurochem. 1, 119–125. Adelman, J. P., Bond, C. T., Pessia, M., and Maylie, J. (1995). Episodic ataxia results from voltage-dependent potassium channels with altered functions. Neuron 15, 1449–1454. Altevogt, B. M., Kleopa, K. A., Postma, F. R., Scherer, S. S., and Paul, D. L. (2002). Cx29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J. Neurosci. 22, 6458–6470. Apostolski, S., Sadiq, S. A., Hays, A., Corbo, M., Suturkova-Milosevic, L., ChaliV, P., Stefansson, K., LeBaron, R. G., Ruoslahti, E., Hays, A. P., and Latov, N. (1994). IdentiWcation of Gal(b1–3)GalNAc bearing glycoproteins at the nodes of Ranvier in peripheral nerve. J. Neurosci. Res. 38, 134–141. Ariyasu, R. G., Nichol, J. A., and Ellisman, M. H. (1985). Localization of sodium/potassium adenosine triphosphatase in multiple cell types of the murine nervous system with antibodies raised against the enzyme from kidney. J. Neurosci. 5, 2581–2596. Arroyo, E. J., and Scherer, S. S. (2000). On the molecular architecture of myelinated Wbers. Histochem. Cell Biol. 113, 1–18. Arroyo, E. J., Xu, T., Grinspan, J., Lambert, S., Levinson, S. R., Brophy, P. J., Peles, E., and Scherer, S. S. (2002). Genetic dysmyelination alters the molecular architecture of the nodal region. J. Neurosci. 22, 1726–1737. Arroyo, E. J., Xu, T., Poliak, S., Watson, M., Peles, E., and Scherer, S. S. (2001). Internodal specializations of myelinated axons in the CNS. Cell Tiss. Res. 305, 53–66. Arroyo, E. J., Xu, Y.-T., Zhou, L., Messing, A., Peles, E., Chiu, S. Y., and Scherer, S. S. (1999). Myelinating Schwann cells determine the internodal localization of Kv1.1, Kv1.2, Kvb2, and Caspr. J. Neurocytol. 28, 333–347. Asher, R. A., Morgenstern, D. A., Shearer, M. C., Adcock, K. H., Pesheva, P., and Fawcett, J. W. (2002). Versican is upregulated in CNS injury and is a product of oligodendrocyte lineage cells. J. Neurosci. 22, 2225–2236. Baba, H., Akita, H., Ishibashi, T., Inoue, Y., Nakahira, K., and Ikenaka, K. (1999). Completion of myelin compaction, but not the attachment of oligodendroglial processes triggers Kþ channel clustering. J. Neurosci. Res. 58, 752–764. Balice-Gordon, R. J., Bone, L. J., and Scherer, S. S. (1998). Functional gap junctions in the Schwann cell myelin sheath. J. Cell Biol. 142, 1095–1104. Bartsch, U., Pesheva, P., RaV, M., and Schachner, M. (1993). Expression of janusin (J1–160/180) in the retina and optic nerve of the developing and adult mouse. Glia 9, 57–69. Baumgartner, S., Littleton, J. T., Broadie, K., Bhat, M. A., Harbecke, R., Lengyel, J. A., Chiquet-Ehrismann, R., Prokop, A., and Bellen, H. J. (1996). A Drosophila neurexin is required for septate junction and blood-nerve barrier formation and function. Cell 87, 1059–1068. Bellen, H. J., Lu, Y., Beckstead, R., and Bhat, M. A. (1998). Neurexin IV, Caspr and paranodin -novel members of the neurexin family: Encounters of axons and glia. Trends Neurosci. 21, 444–449. Bennett, V., and Baines, A. J. (2001). Spectrin and ankyrin-based pathways: Metazoan inventions for integrating cells into tissues. Physiol. Rev. 81, 1353–1392. Bennett, V., Lambert, S., Davis, J. Q., and Zhang, X. (1997). Molecular architecture of the specialized axonal membrane at the node of Ranvier. Soc. Gen. Physiol. 52, 107–120. Berghs, S., Aggujaro, D., Dirkx, R., Maksimova, E., Stabach, P., Hermel, J. M., Zhang, J. P., Philbrick, W., Slepnev, V., Ort, T., and Solimena, M. (2000). beta IV spectrin, a new spectrin localized at axon initial segments and nodes of Ranvier in the central and peripheral nervous system. J. Cell Biol. 151, 985–1001. BergoVen, J., Scherer, S. S., Wang, S., Oronzi-Scott, M., Bone, L., Paul, D. L., Chen, K., Lensch, M. W., Chance, P., and Fischbeck, K. (1993). Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262, 2039–2042. Berthold, C.-H., and Rydmark, M. (1983). Electron microscopic serial section analysis of nodes of Ranvier in lumbosacral spinal roots of the cat: Ultrastructural organization in Wbres of diVerent sizes. J. Neurocytol. 12, 475–505.

107

108

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

Bhat, M. A., Rios, J. C., Lu, Y., Garcia-Fresco, G. P., Ching, W., St Martin, M., Li, J. J., Einheber, S., Chesler, M., Rosenbluth, J., Salzer, J. L., and Bellen, H. J. (2001). Axon-glia interactions and the domain organization of myelinated axons requires Neurexin IV/Caspr/Paranodin. Neuron 30, 369–383. Bjartmar, C., Karlsson, B., and Hildebrand, C. (1994). Cellular and extracellular components at nodes of Ranvier in rat white matter. Brain Res. 667, 111–114. Black, J. A., and Waxman, S. G. (1988). The perinodal astrocyte. Glia 1, 169–183. Blanchard, C. E., Mackenzie, M. L., Sikri, K., and Allt, G. (1985). Filipin-sterol complexes at nodes of Ranvier. J. Neurocytol. 14, 1053–1062. Blight, A. R. (1985). Computer stimulation of action potentials and afterpotentials in mammalian myelinated axons: The case for a lower resistance myelin sheath. Neuroscience 15, 13–31. Boiko, T., Rasband, M. N., Levinson, S. R., Caldwell, J. H., Mandel, G., Trimmer, J. S., and Matthews, G. (2001). Compact myelin dictates the diVerential targeting of two sodium channel isoforms in the same axon. Neuron 30, 91–104. Bosio, A., Binczek, E., and StoVel, W. (1996). Functional breakdown of the lipid bilayer of the myelin membrane in central and peripheral nervous system myelin by disrupted galactocerebroside synthesis. Proc. Natl. Acad. Sci. USA 93, 13280–13285. Bostock, H., Sears, T. A., and Sherratt, R. M. (1981). The eVects of 4-aminopyridine and tetrathylammonium ions on normal and demyelinated mammalian nerve Wbers. J. Physiol. 313, 301–315. Bouzidi, M., Tricaud, N., Giraud, P., Kordeli, E., Caillol, G., Deleuze, C., Couraud, F., and Alcaraz, G. (2002). Interaction of the Nav1.2a subunit of the voltage-dependent sodium channel with nodal ankyrinG—In vitro mapping of the interacting domains and association in synaptosomes. J. Biol. Chem. 277, 28996–29004. Bowe, C. M., Hildebrand, C. H., Kocsis, J. D., and Waxmen, S. G. (1989). Morphological and physiological properties of neurons after long-term axonal regeneration: Observations on chronic and delayed sequalae of peripheral nerve injury. J. Neurol. Sci. 91, 259–292. Bowe, C. M., Johansson, C. S., Hildebrand, C., and Evans, N. H. (1994). Functional properties and nodal spacing of myelinated Wbers in developing rat mental and sural nerves. Dev. Brain Res. 79, 186–194. Boyle, M. E. T., Berglund, E. O., Murai, K. K., Weber, L., Peles, E., and Ranscht, B. (2001). Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30, 385–397. Browne, D. L., Gancher, S. T., Nutt, J. G., Brunt, E. R. P., Smith, E. A., Kramer, P., and Litt, M. (1994). Episodic ataxia/myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNA1. Nat. Genet. 8, 136–140. Brunt, E. R. P., and Van Weerden, T. W. (1990). Familial paroxysmal kinesigenic ataxia and continuous myokymia. Brain 113, 1361–1382. Burgess, D. L., Kohrman, D. C., Galt, J., Plummer, N. W., Jones, J. M., Spear, B., and Meisler, M. H. (1995). Mutation of a new sodium channel gene, Scn8a, in the mouse mutant ‘motor endplate disease’. Nat. Genet. 10, 461–465. Butt, A. M., Duncan, A., and Berry, M. (1994). Astrocyte associations with nodes of Ranvier: Ultrastructural analysis of HRP-Wlled astrocytes in the mouse optic nerve. J. Neurocytol. 23, 486–499. Butt, A. M., Duncan, A., Hornby, M. F., Kirvell, S. L., Hunter, A., Levine, J. M., and Berry, M. (1999). Cells expressing the NG2 antigen contact nodes of Ranvier in adult CNS white matter. Glia 26, 84–91. Caldwell, J. H., Schaller, K. L., Lasher, R. S., Peles, E., and Levinson, S. R. (2000). Sodium channel NaV1.6 is localized at nodes of Ranvier, dendrites, and synapses. Proc. Natl. Acad. Sci. USA 97, 5616–5620. Catterall, W. A. (2000). From ionic currents to molecular mechanisms: The structure and function of voltagegated sodium channels. Neuron 26, 13–25. Chandross, K. J., Kessler, J. A., Cohen, R. I., Simburger, E., Spray, D. C., Bieri, P., and Dermietzel, R. (1996). Altered connexin expression after peripheral nerve injury. Mol. Cell. Neurosci. 7, 501–518. Charles, P., Tait, S., Faivre-Sarrailh, C., Barbin, G., Gunn-Morre, F., Denisenko-Nehrbass, N., Guennoc, A. M., Girault, J. A., Brophy, P. J., and Lubetzki, C. (2002). Neurofascin is a glial receptor for the paranodin/Casprcontactin axonal complex at the axoglial junction. Curr. Biol. 12, 217–220. Chen, C., Bharucha, V., Chen, Y., Westenbroek, R. E., Brown, A., Malhotra, J. D., Jones, D., Avery, C., Gillespie, P. J. I., Cazen-Gillespie, K. A., Kazarinova-Noyes, K., Shrager, P., Saunders, T. L., Macdonald, R. L., Ransom, B. R., Scheuer, T., Catterall, W. A., and Isom, L. L. (2002a). Reduced sodium channel density, altered voltage dependence of inactivation, and increased susceptibility to seizures in mice lacking sodium channel b2-subunits. Proc. Natl. Acad. Sci. USA 99, 17072–17077. Chen, S., Ren, Y.-Q., and Hillman, D. E. (2002b). Expression of the Kir4.1 channel at axonal paranodes and synapses during cerebellar deveopment and into adulthood. Soc. Neurosci., 546.513. Chiba, A., Kusunoki, S., Obata, H., Machinami, R., and Kanazawa, I. (1993). Serum anti-GQ1b IgG antibody is associated with ophthalmoplegia in Miller Fisher syndrome and Guillain-Barre syndrome—clinical and immunohistochemical studies. Neurology 43, 1911–1917. Ching, W., Zanazzi, G., Levinson, S. R., and Salzer, J. L. (1999). Clustering of neuronal sodium channels requires contact with myelinating Schwann cells. J. Neurocytol. 28, 295–301. Chiu, S. Y. (1991). Functions and distribution of voltage-gated sodium and potassium channels in mammalian Schwann cells. Glia 4, 541–558. Chiu, S. Y., and Ritchie, J. M. (1980). Potassium channels in nodal and internodal axonal membrane of mammalian myelinated Wbres. Nature 284, 170–171. Chiu, S. Y., and Ritchie, J. M. (1981). Evidence for the presence of potassium channels in the paranodal region of acutely demyelinated mammalian single nerve Wbres. J. Physiol. 313, 415–437.

CONCLUSION

Chiu, S. Y., and Ritchie, J. M. (1982). Evidence for the presence of potassium channels in the internode of frog myelinated nerve Wbres. J. Physiol. 322, 485–501. Coetzee, T., Dupree, J. L., and Popko, B. (1998). Demyelination and altered expression of myelin-associated glycoprotein isoforms in the central nervous system of galactolipid-deWcient mice. J. Neurosci. Res. 54, 613–622. Coetzee, T., Fujita, N., Dupree, J., Shi, R., Blight, A., Suzuki, K., Suzuki, K., and Popko, B. (1996). Myelination in the absence of galactocerebroside and sulfatide: Normal structure with abnormal function and regional instability. Cell 86, 209–219. Corbo, M., Quattrini, A., Latov, N., and Hays, A. P. (1993). Localization of GM1 and Gal(b1–3)GalNAc antigenic determinants in peripheral nerve. Neurology 43, 809–814. Corrette, B., Repp, H., Dreyer, F., and Schwarz, J. (1991). Two types of fast Kþ channels in rat myelinated Wbres and their sensitivity to dendrotoxin. PXu¨gers Arch. 418, 408–416. DaniloV, J. K., Crossin, K. L., Ancon-Raymond, M., Murawsky, M., Reiger, F., and Edelman, G. M. (1989). Expression of cytotactin in the normal and regenerating neuromuscular system. J. Cell Biol. 108, 623–635. DaniloV, J. K., Levi, G., Grumet, M., Reiger, F., and Edelman, G. M. (1986). Altered expression of neuronal cell adhesion molecules induced by nerve injury and repair. J. Cell Biol. 103, 929–945. Davis, J. Q., and Bennett, V. (1994). Ankyrin binding activity shared by the neurofascin/L1/NrCAM family of nervous system cell adhesion molecules. J. Biol. Chem. 269, 27163–27166. Davis, J. Q., Lambert, S., and Bennett, V. (1996). Molecular composition of the node of Ranvier: IdentiWcation of ankyrin-binding cell adhesion molecules neurofascin (mucinþ third FNIII domain-) and NrCAM at nodal axon segments. J. Cell Biol. 135, 1355–1367. Davis, J. Q., McLaughlin, T., and Bennett, V. (1993). Ankyrin-binding proteins related to nervous system cell adhesion molecules: Candidates to provide transmembrane and intercellular connections in adult brain. J. Cell Biol. 121, 121–133. Delpech, A., Firard, N., and Delpech, B. (1982). Localization of hyaluronectin in the nervous system. Brain Res. 245, 251–257. Devaux, J., Alcaraz, G., Grinspan, J., Bennett, V., Joho, R., Crest, M., and Scherer, S. S. (2003a). Kv3.1 is a novel component of CNS nodes., 23, 4509–4518. Devaux, J., Gola, M., Jacquet, G., and Crest, M. (2002). EVects of Kþ channel blockers on developing rat myelinated CNS axons: IdentiWcation of four types of Kþ channels. J. Neurophysiol. 87, 1376–1385. Devaux, J. J., Kleopas, A. K., Cooper, E. C., Bennett, V., Scherer, S. S. (2003b). A novel role for KCNQ2 at nodes of Ranvier. Soc Neurosci Abstr 368.8.3651. Devor, M., Govrin-Lippmann, R., and Angelides, K. (1993). Naþ channel immunolocalization in peripheral mammalian axons and changes following nerve injury and neuroma formation. J. Neurosci. 13, 1976–1992. Duchen, L. W., and Stefani, E. (1971). Electrophysiological studies of neuromuscular transmission hereditary ‘‘motor end-plate disease’’ of the mouse. J. Physiol. 212, 535–548. Dugandzija-Novakovic, S., Koszowski, A. G., Levinson, S. R., and Shrager, P. (1995). Clustering of Naþ channels and node of Ranvier formation in remyelinating axons. J. Neurosci. 15, 492–503. Dupree, J. L., Girault, J. A., and Popko, B. (1999). Axo-glial interactions regulate the localization of axonal paranodal proteins. J. Cell Biol. 147, 1145–1151. Einheber, S., Zanazzi, G., Ching, W., Scherer, S. S., Milner, T. A., Peles, E., and Salzer, J. L. (1997). The axonal membrane protein Caspr/Neurexin IV is a component of the septate-like paranodal junctions that assemble during myelination. J. Cell Biol. 139, 1495–1506. Eurelings, M., Moons, K. G. M., Notermans, N. C., Sasker, L. D., DeJager, A. E. J., Wintzen, A. R., Wokke, J. H. J., and VandenBerg, L. H. (2001). Neuropathy and IgM M-proteins—Prognostic value of antibodies to MAG, SGPG, and sulfatide. Neurology 56, 228–233. Faivre-Sarrailh, C., Gauthier, F., Denisenko-Nehrbass, N., Le Bivic, A., Rougon, G., and Girault, J.-A. (2000). The glycosylphosphatidyl inositol-anchored adhesion molecule F3/contactin is required for surface transport of paranodin/contactin-associated protein (Caspr). J. Cell Biol. 149, 491–501. Fannon, A. M., Sherman, D. L., Ilyina-Gragerova, G., Brophy, P. J., Friedrich, V. L., and Colman, D. R. (1995). Novel E-cadherin mediated adhesion in peripheral nerve: Schwann cell architecture is stabilized by autotypic adherens junctions. J. Cell Biol. 129, 189–202. Feder, N. (1971). Microperoxidase—an ultrastructural tracer of low molecular weight. J. Cell Biol. 51, 339–343. Felsenfeld, D. P., Hynes, M. A., Skoler, K. M., Furley, A. J., and Jessell, T. M. (1994). TAG-1 can mediate homophilic binding, but neurite outgrowth on TAG-1 requires an L1-like molecule and beta 1 integrins. Neuron 12, 675–690. Vrench-Constant, C., Miller, R. H., Kruse, J., Schachner, M., and RaV, M. C. (1986). Molecular specialization of astrocyte processes at nodes of Ranvier in rat optic nerve. J. Cell Biol. 102, 844–852. Fjell, J., Hjelmstrom, P., Hormuzdiar, W., Milenkovic, M., Aglieco, F., Tyrrell, L., Dib-Haji, S., Waxman, S. G., and Black, J. A. (2000). Localization of the tetrodotoxin-resistant sodium channel NaN in nociceptors. Neuroreport 11, 199–202. Freigang, J., Proba, K., Leder, L., Diederichs, K., Sonderegger, P., and Welte, W. (2000). The crystal structure of the ligang binding module of axonin-1/TAG-1 suggests a zipper mechanism for neural cell adhesion. Cell 101, 425–433. Funch, P. G., and Faber, D. S. (1984). Measurement of myelin sheath resistances: Implications for axonal conduction and pathophysiology. Science 225, 538–540. Furuse, M., Sasaki, H., Fujimoto, K., and Tsukita, S. (1998). A single gene product, claudin-1 or -2, reconstitutes tight junction strands and recruits occludin in Wbroblasts. J. Cell Biol. 143, 391–401.

109

110

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

Ganser, A. L., Kirschner, D. A., and Willinger, M. (1983). Ganglioside localization on myelinated nerve Wbres by cholera toxin binding. J. Neurocytol. 12, 921–938. Garver, T. D., Ren, Q., Tuvia, S., and Bennett, V. (1997). Tyrosine phosphorylation at a site highly conserved in the L1 family of cell adhesion molecules abolishes ankyrin binding and increases lateral mobility of neurofascin. J. Cell Biol. 137, 703–714. Girault, J. A., and Peles, E. (2002). Development of nodes of Ranvier. Curr. Opin. Neurobiol. 12, 476–485. Goldin, A. L., Barchi, R. L., Caldwell, J. H., Hofmann, F., Howe, J. R., Hunter, J. C., Kallen, R. G., Mandel, G., Meisler, M. H., Netter, Y. B., Noda, M., Tamkun, M. M., Waxman, S. G., Wood, J. N., and Catterall, W. A. (2000). Nomenclature of voltage-gated sodium channels. Neuron 28, 365–368. Gollan, L., Sabanay, H., Poliak, S., Berglund, E. O., Ranscht, B., and Peles, E. (2002). Retention of a cell adhesion complex at the paranodal junction requires the cytoplasmic region of Caspr. J. Cell Biol. 157, 1247–1256. Gollan, L., Salomon, D., and Peles, E. (2003). Complex interactions at the axoglial junction: Contactin mediates binding of Caspr to the glial isoform of neurofascin. (submitted),. Gong, Y., Tagawa, Y., Lunn, M. P. T., Laroy, W., HeVerLauc, M., Li, C. Y., GriYn, J. W., Schnaar, R. L., and Sheikh, K. A. (2002). Localization of major gangliosides in the PNS: Implications for immune neuropathies. Brain 125, 2491–2506. Goodyear, C. S., O’Hanlon, G. M., Plomp, J. J., Wagner, E. R., Morrison, I., Veitch, J., Cochrane, L., Bullens, R. W. M., Molenaar, P. C., Conner, J., and Willison, H. J. (1999). Monoclonal antibodies raised against Guillain-Barre syndrome-associated Campylobacter jejuni lipopolysaccharides react with neuronal gangliosides and paralyze muscle-nerve preparations. J. Clin. Invest. 104, 697–708. Gow, A., Southwood, C. M., Li, J. S., Pariali, M., Riordan, G. P., Brodie, S. E., Dania, J., Branstein, J. M., Kachar, B., and Lazzarini, R. A. (1999). CNS myelin and Sertoli cell tight junctions strands are absent in Osp/ claudin-11 null mice. Cell 99, 649–659. Grafe, P., Mayer, C., Takigawa, T., Kamleiter, M., and Sanchez-Brandelik, R. (1999). Confocal calcium imaging reveals an ionotropic P2 nucleotide receptor in the paranodal membrane of rat Schwann cells. J. Physiol. 515, 377–383. Gulbis, J. M., Mann, S., and MacKinnon, R. (1999). Structure of a voltage-dependent Kþ channel beta subunit. Cell 97, 943–952. Hall, S. M., and Williams, P. L. (1970). Studies on the ‘incisures’ of Schmidt and Lanterman. J. Cell Sci. 6, 767–791. Hall, S. M., and Williams, P. L. (1971). The distribution of electron-dense tracers in peripheral nerve Wbres. J. Cell Sci. 8, 541–555. Harroch, S., Palmeri, M., Rosenbluth, J., Custer, A., Okigaki, M., Shrager, P., Blum, M., Buxbaum, J. D., and Schlessinger, J. (2000). No obvious abnormality in mice deWcient in receptor protein tyrosine phosphatase b. Mol. Cell. Biol. 20, 7706–7715. Hart, I. K., Waters, C., Vincent, A., Newland, C., Beeson, D., Pongs, O., Morris, C., and Newsom-Davis, J. (1997). Autoantibodies detected to express Kþ channels are implicated in neuromyotonia. Ann. Neurol. 41, 238–246. Hayashi, K., Yonemura, S., Matsui, T., Tsukita, S., and Tsukita, S. (1999). ImmunoXuorescence detection of ezrin/radixin/moesin (ERM) proteins with their carboxyl-terminal threonine phosphorylated in cultured cells and tissues. Application of a novel Wxation protocol using trichloroacetic acid (TCA) as a Wxative. J. Cell Sci. 112, 1149–1158. Hayat, G. R., Kulkantrakorn, K., Campbell, W. W., and Giuliani, M. J. (2000). Neuromyotonia: Autoimmune pathogenesis and response to immune modulating therapy. J. Neurol. Sci. 181, 38–43. Hess, A., and Young, J. Z. (1952). The nodes of Ranvier. Proc. R. Soc. Lond. B 140, 301–320. Hildebrand, C., Bowe, C. M., and Remahl, I. N. (1994). Myelination and myelin sheath remodelling in normal and pathological PNS nerve Wbres. Prog. Neurobiol. 43, 85–141. Hille, B. (2001). ‘‘Ion Channels of Excitable Membranes.’’ Sinauer Associates, Sunderland, MA. Hirano, A., Becker, N. H., and Zimmerman, H. M. (1969). Isolation of the periaxonal space of the central myelinated nerve Wber with regard to the diVusion of peroxidase. J. Histochem. Cytochem. 17, 512–516. Hirano, A., and Dembitzer, H. M. (1982). Further studies on the transverse bands. J. Neurocytol. 11, 861–866. Honke, K., Hirahara, Y., Dupree, J., Suzuki, K., Popko, B., Fukushima, K., Fukushima, J., Nagasawa, T., Yoshida, N., Wada, Y., and Taniguchi, N. (2002). Paranodal junction formation and spermatogenesis require sulfoglycolipids. Proc. Natl. Acad. Sci. USA 99, 4227–4232. Hopkins, W. F., Allen, M. L., Mouamed, K. M., and Tempel, B. L. (1994). Properties of voltage-gated Kþ currents expressed in Xenopus oocytes by mKv1.1, mKv1.2 and their heteromultimers as revealed by mutagenesis of the dendrotoxin-binding site in Kv1.1. PXu¨gers Arch. 428, 382–390. Ichimura, T., and Ellisman, M. H. (1991). Three-dimensional Wne structure of cytoskeletal-membrane interactions at nodes of Ranvier. J. Neurocytol. 20, 667–681. Ishibashi, T., Dupree, J. L., Ikenaka, K., Hirahara, Y., Honke, K., Peles, E., Popko, B., Suzuki, K., Nishino, H., and Baba, H. (2002). A myelin galactolipid, sulfatide, is essential for maintenance of ion channels on myelinated axon but not essential for initial cluster formation. J. Neurosci. 22, 6507–6514. Isom, L. L. (2002). the role of sodium channels in cell adhesion. Frontiers Biosci. 7, 12–23. Isom, L. L., Ragsdale, D. S., De Jongh, K. S., Westenbroek, R. E., Reber, B. F., Scheuer, T., and Catterall, W. A. (1995). Structure and function of the beta 2 subunit of brain sodium channels, a transmembrane glycoprotein with a CAM motif. Cell 83, 433–442.

CONCLUSION

Jenkins, S. M., and Bennett, V. (2001). AnkyrinG coordinates assembly of the spectrin-baseed membrane skeleton, voltage-gated sodium channels, and L1 CAMs at Purkinje neuron initial segments. J. Cell Biol. 155, 739–745. Jenkins, S. M., and Bennett, V. (2002). Developing nodes of Ranvier are deWned by ankyrin-G clustering and are independent of paranodal axoglial adhesion. Proc. Natl. Acad. Sci. USA 99, 2303–2308. Kaplan, M. R., Cho, M.-H., Ullian, E. M., Isom, L. L., Levinson, S. R., and Barres, B. A. (2001). DiVerential control of clustering of the sodium channels Nav1.2 and Nav1.6 at developing CNS nodes of Ranvier. Neuron 40, 105–119. Kaplan, M. R., Meyer-Franke, A., Lambert, S., Bennett, V., Duncan, I. D., Levinson, S. R., and Barres, B. A. (1997). Induction of sodium channel clustering by oligodendrocytes. Nature 386, 724–728. Kazarinova-Noyes, K., Malhotra, J. D., McEwen, D. P., Mattei, L. N., Berglund, E. O., Ranscht, B., Levinson, S. R., Schachner, M., Shrager, P., Isom, L. L., and Xiao, Z.-C. (2001). Contactin associates with Naþ channels and increases their functional expression. J. Neurosci. 21, 7517–7525. Kazarinova-Noyes, K., and Shrager, P. (2002). Molecular constituents of the node of Ranvier. Mol. Neurobiol. 26, 167–182. Kleopa, K. A., and Scherer, S. S. (2002). Inherited Neuropathies. Neurol. Clin. N. Am. 20, 679–709. Kocsis, J. D., Ruiz, J. A., and Waxman, S. G. (1983). Maturation of mammalian myelinated Wbers: Changes in action-potential characteristics following 4-aminopyridine application. J. Neurophysiol. 50, 449–463. Koenig, E., and Repasky, E. (1985). A regional analysis of a-spectrin in the isolated Mauthner neuron and in isolated axons of the goldWsh and rabbit. J. Neurosci. 5, 705–714. Komada, M., and Soriano, P. (2002). beta IV-spectrin regulates sodium channel clustering through ankyrin-G at axon initial segments and nodes of Ranvier. J. Cell Biol. 156, 337–348. Konishi, T. (1990). Voltage-gated potassium currents in myelinating Schwann cells in the mouse. J. Physiol. 431, 123–139. Kordeli, E., Davis, J., Trapp, B., and Bennett, V. (1990). An isoform of ankyrin is localized at nodes of Ranvier in myelinated axons of central and peripheral nerves. J. Cell Biol. 110, 1341–1352. Kordeli, E., Lambert, S., and Bennett, V. (1995). AnkyrinG: A new ankyrin gene with neural-speciWc isoforms localized at the axonal initial segment and node of Ranvier. J. Biol. Chem. 270, 2352–2359. Koszowski, A. G., Owens, G. C., and Levinson, S. R. (1998). The eVect of the mouse mutation claw paw on myelination and nodal frequency in sciatic nerves. J. Neurosci. 18, 5859–5868. Kunz, B., Lierheimer, R., Rader, C., Spirig, M., Ziegler, U., and Sonderegger, P. (2002). Axonin-1/TAG1 mediates cell-cell adhesion by a cis-assisted trans-activation. J. Biol. Chem. 277, 4551–4557. Kusunoki, S., Chiba, A., Tai, T., and Kanazawa, I. (1993). Localization of GM1 and GD1b antigens in the human peripheral nervous system. Muscle Nerve 16, 752–756. Kusunoki, S., Mashiko, H., Mochizuki, N., Chiba, A., Arita, M., Hitoshi, S., and Kanazawa, I. (1997). Binding of antibodies against GM1 and GD1b in human peripheral nerve. Muscle Nerve 20, 840–845. Kwa, M. S. G., van Schaik, I. N., Brand, A., Baas, F., and Vermeulen, M. (2001). Investigation of serum response to PMP22, connexin 32 and P0 in inXammatory neuropathies. J. Neuroimmunol. 116, 220–225. Laird, J. M. A., Souslova, V., Wood, J. N., and Cervero, F. (2002). DeWcits in visceral pain and referred hyperalgesia in Nav1.8 (SNS/PN3)-null mice. J. Neurosci. 22, 8352–8356. Lamb, R. S., Ward, R. E., Schweizer, L., and Fehon, R. G. (1998). Drosophila Coracle, a member of the protein 4.1 superfamily, has essential structural functions in the septate junctions and developmental functions in embryonic and adult epithelia cells. Mol. Cell. Biol. 9, 3505–3519. Lambert, S., Davis, J. Q., and Bennett, V. (1997). Morphogenesis of the node of Ranvier: Co-clusters of ankyrin and ankyrin-binding integral proteins deWne early developmental intermediates. J. Neurosci. 17, 7025–7036. Landon, D. N., and Langley, O. K. (1971). The local chemical environment of nodes of Ranvier: A study of cation binding. J. Anat. 108, 419–432. Latov, N. (1990). Antibodies to glycoconjugates in neurologic disease. Clinical Aspects Autoimmunity 4, 18–29. Lawlor, M. W., Richards, M. P., DeVries, G. H., Fisher, M. A., and Stubbs, E. B. (2002). Antibodies to Lperiaxin in sera of patients with peripheral neuropathy produce experimental sensory nerve conduction deWcits. J. Neurochem. 83, 592–600. Lev-Ram, V., and Ellisman, M. H. (1995). Axonal activation-induced calcium transients in myelinating Schwann cells, sources, and mechanisms. J. Neurosci. 15, 2628–2637. Li, J., Hertzberg, E. L., and Nagy, J. I. (1997). Connexin32 in oligodendrocytes and association with myelinated Wbers in mouse and rat brain. J. Comp. Neurol. 379, 571–591. Li, X., Lynn, B. D., Olson, C., Meier, C., Davidson, K. G. V., Yasumura, T., Rash, J. E., and Nagy, J. L. (2002). Connexin29 expression, immunocytochemistry and freeze-fracture replica immunogold labelling (FRIL) in sciatic nerve. Eur. J. Neurosci. 16, 795–806. Liguori, R., Vincent, A., Clover, L., Avoni, P., Plazzi, G., Cortelli, P., Baruzzi, A., Carey, T., Gambetti, P., Lugasesi, E., and Mntagna, P. (2001). Morvan’s syndrome: Peripheral and central nervous system and cardiac involvement with antibodies to voltage-gated potassium channnels. Brain 124, 2417–2426. Liu, C.-j., Dib-Hajj, S. D., Black, J. A., Greenwood, J., Lian, Z., and Waxman, S. G. (2001). Direct interaction with contactin targets voltage-gated sodium channel NaV1.9 to the cell membrane. J. Biol. Chem. 276, 46553–46561. Lustig, M., Zanazzi, G., Sakurai, T., Blanco, C., Levinson, S. R., Lambert, S., Grumet, M., and Salzer, J. L. (2001). Nr-CAM and neurofascin interactions regulate ankyrin G and sodium channel clustering at the node of Ranvier. Curr. Biol. 11, 1864–1869.

111

112

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

MacKenzie, M. L., Ghabriel, M. N., and Allt, G. (1984a). Nodes of Ranvier and Schmidt-Lanterman incisures: An in vivo lanthanum tracer study. J. Neurocytol. 13, 1043–1055. MacKenzie, M. L., Shorer, Z., Ghabriel, M. N., and Allt, G. (1984b). Myelinated nerve Wbres and the fate of lanthanum tracer: An in vivo study. J. Anat. 138, 1–14. Malhotra, J. D., Kazen-Gillespie, K., Hortsch, M., and Isom, L. L. (2000). Sodium channel beta subunits mediate homophilic cell adhesion and recruit ankyrin to points of cell-cell contact. J. Biol. Chem. 275, 11383–11388. Marcus, J., Dupree, J. L., and Popko, B. (2002). Myelin-associated glycoprotein and myelin galactolipids stabilize developing axo-glial interactions. J. Cell Biol. 156, 567–577. Marcus, J., and Popko, B. (2002). Galactolipids are molecular determinants of myelin development and axo-glial organization. BBA Gen. Subjects 1573, 406–413. Martin, S., Levine, A. K., Chen, Z. J., Ughrin, Y., and Levine, J. M. (2001). Deposition of the NG2 proteoglycan at nodes of Ranvier in the peripheral nervous system. J. Neurosci. 21, 8119–8128. Martini, R., Schachner, M., and Faissner, A. (1990). Enhanced expression of the extracellular matrix molecule J1/ tenascin in the regenerating adult mouse sciatic nerve. J. Neurocytol. 19, 601–616. Mata, M., Fink, D. J., Ernst, S. A., and Segal, G. J. (1991). Immunocytochemical demonstration of Naþ, KþATPase in internodal axolemma of myelinated Wbers of rat sciatic and optic nerves. J. Neurochem. 57, 184–192. Mathis, C., Denisenko-Nehrbass, N., Girault, J. A., and Borrelli, E. (2001). Essential role of oligodendrocytes in the formation and maintenance of central nervous system nodal regions. Development 128, 4881–4890. Mathis, C., Hindelang, C., LeMeur, M., and Borrelli, E. (2000). A transgenic mouse model for inducible and reversible dysmyelination. J. Neurosci. 20, 7698–7705. Maylie, B., Bissonnette, E., Virk, M., Adelman, J. P., and Maylie, J. G. (2002). Episodic ataxia type 1 mutations in the human Kv1.1 potassium channel alter hKv beta 1-induced N-type inactivation. J. Neurosci. 22, 4786–4793. Meadows, L., Malhotra, J. D., Isom, L. L., and Ragsdale, D. S. (2001). The intracellular segment of the sodium channel b-subunit is required for its eYcient association with the channel a-subunits. J. Neurochem. 76, 1871–1878. Mege, R. M., Nicolet, M., Pincon-Raymond, M., Murawsky, M., and Rieger, F. (1992). Cytotactin is involved in synaptogenesis during regeneration of the frog neuromuscular system. Dev. Biol. 149, 381–394. Melendez-Vasquez, C. V., Rios, J. C., Zanazzi, G., Lambert, S., Bretscher, A., and Salzer, J. L. (2001). Nodes of Ranvier form in association with ezrin-radixin-moesin (ERM)-positive Schwann cell processes. Proc. Natl. Acad. Sci. USA 98, 1235–1240. Menegoz, M., Gaspar, P., Le Bert, M., Galvez, T., Burgaya, F., Palfrey, C., Ezan, P., Arnos, F., and Girault, J.-A. (1997). Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron 19, 319–331. Menichella, D. M., Arroyo, E. J., Awatramani, R., Xu, T., Baron, P., Vallat, J.-M., Balsamo, J., Lilien, J., Scarlato, G., Kamholz, J., Scherer, S. S., and Shy, M. E. (2001). Protein zero is necessary for cadherinmediated adherens junction formation in Schwann cells. Mol. Cell. Neurosci. 18, 606–618. Mi, H., Harris-Warrick, R. M., Deerinck, T. J., Inman, I., Ellisman, M. H., and Schwartz, T. L. (1999). IdentiWcation and localization of Caþþ-activated Kþ channels in rat sciatic nerve. Glia 26, 166–175. Mi, H. Y., Deerinck, T. J., Ellisman, M. H., and Schwarz, T. L. (1995). DiVerential distribution of closely related potassium channels in rat Schwann cells. J. Neurosci. 15, 3761–3774. Mi, H. Y., Deerinck, T. J., Jones, M., Ellisman, M. H., and Schwarz, T. L. (1996). Inwardly rectifying Kþ channels that may participate in Kþ buVering are localized in microvilli of Schwann cells. J. Neurosci. 16, 2421–2429. Michaely, P., Tomchick, D. R., Machius, M., and Anderson, R. G. W. (2002). Crystal structure of a 12 ANK repeat stack from human ankyrinR. EMBO J.. 21, 6387–6396. Milev, P., Chiba, A., Haring, M., Rauvala, H., Schachner, M., Ranscht, B., Margolis, R. K., and Margolis, R. U. (1998). High aYnity binding and overlapping localization of neurocan and phosphacan protein-tyrosine phosphatase-z/b with tenascin-R, amphoterin, and the heparin-binding growth-associated molecule. J. Biol. Chem. 273, 6998–7005. Miller, R. G., and Da Silva, P. P. (1977). Particle rosettes in the periaxonal Schwann cell membrane and particle clusters in the axolemma of rat sciatic nerve. Brain Res. 130, 135–141. Mirsky, R., Jessen, K. R., Schachner, M., and Goridis, R. (1986). Distribution of the adhesive molecules N-CAM and L1 on peripheral neurons and glia in adult rats. J. Neurocytol. 15, 799–815. Missler, M., Fernandez-Chacon, R., and Sudhof, T. C. (1998). The making of neurexins. J. Neurochem. 71, 1339–1347. Mitic, L. L., and Anderson, J. M. (1998). Molecular architecture of tight junctions. Annu. Rev. Physiol. 60, 121–142. Molander, M., Berthold, C.-H., Persoon, H., Anderson, K., and Fredman, P. (1997). Monosialoganglioside (GM1) immunoXuorescence in rat spinal roots studied with a monoclonal antibody. J. Neurocytol. 26, 101–111. Morita, K., Sasaki, H., Fujimoto, K., Furuse, M., and Tsukita, S. (1999). Claudin-11/OSP-based tight junctions of myelin sheaths in brain and Sertoli cells in testis. J. Cell Biol. 145, 579–588. Mugnaini, E., and Schnapp, B. (1974). Possible role of zonula occludens of the myelin sheath in demyelinating conditions. Nature 251, 725–727. Nagafuchi, A., Tsukita, S., and Takeichi, M. (1993). Transmembrane control of cadherin-mediated cell-cell adhesion. Semin. Cell Biol. 4, 175–181. Neuberg, D. H. H., Sancho, S., and Suter, U. (1999). Altered molecular architecture of peripheral nerves in mice lacking the peripheral myelin protein 22 or connexin32. J. Neurosci. Res. 58, 612–623.

CONCLUSION

Neusch, C., Rozengurt, N., Jacobs, R. E., Lester, H. A., and Kofuji, P. (2001). Kir4.1 potassium channel subunit is crucial for oligodendrocyte development and in vivo myelination. J. Neurosci. 21, 5429–5438. Novakovic, S. D., Deerinck, T. J., Levinson, S. R., Shrager, P., and Ellisman, M. H. (1996). Clusters of axonal Naþ channels adjacent to remyelinating Schwann cells. J. Neurocytol. 25, 403–412. Novakovic, S. D., Levinson, R., Schachner, M., and Shrager, P. (1998a). Disruption and reorganization of sodium channels in experimental allergic neuritis. Muscle Nerve 21, 1019–1032. Novakovic, S. D., Tzoumaka, E., McGivern, J. G., Haraguchi, M., Sangameswaran, L., Gogas, K. R., Eglen, R. M., and Hunter, J. C. (1998b). Distribution of the tetrodotoxin-resistant sodium channel PN3 in rat sensory neurons in normal and neuropathic conditions. J. Neurosci. 18, 2174–2187. Ohara, R., Yamakawa, H., Nakayama, M., and Ohara, O. (2000). Type II brain 4.1 (4.1B/KIAA0987), a member of the protein 4.1 family, is localized to neuronal paranodes. Mol. Brain Res. 85, 41–52. Oohashi, T., Hirakawa, S., Bekku, Y., Rauch, U., Zimmermann, D. R., Su, W. D., Ohtsuka, A., Murakami, T., and Ninomiya, Y. (2002). Bral1, a brain-speciWc link protein, colocalizing with the versican V2 isoform at the nodes of Ranvier in developing and adult, mouse central nervous systems. Mol Cell Neurosci 19, 43–57. Orkand, P. M., Nicholls, J. G., and KuZer, S. W. (1966). EVect of nerve impulses on the membrane potential of glial cells in the central nervous sytem of amphibia. J. Neurophysiol. 29, 788–806. Parkinson, N. J., Olsson, C. L., Hallows, J. L., McKee-Johnson, J., Keogh, B. P., Noben-Trauth, K., Kujawa, S. G., and Tempel, B. L. (2001). Mutant beta-spectrin 4 causes auditory and motor neuropathies in quivering mice. Nat. Genet. 29, 61–65. Parmantier, E., Lynn, B., Lawson, D., Turmaine, M., Namini, S. S., Chakrabarti, L., McMahon, A. P., Jessen, K. R., and Mirsky, R. (1999). Schwann cell-derived desert hedgehog controls the development of peripheral nerve sheaths. Neuron 23, 713–724. Parra, M., Gascard, P., Walensky, L. D., Gimm, J. A., Blackshaw, S., Chan, N., Takakuwa, Y., Berger, T., Lee, G., Chasis, J. A., Snyder, S. H., Mohandas, N., and Conboy, J. G. (2000). Molecular and functional characterization of protein 4.1B, a novel member of the protein 4.1 family with high level, focal expression in brain. J. Biol. Chem. 275, 3247–3255. Pedraza, L., Huang, J. K., and Colman, D. R. (2001). Organizing principles of the axoglial apparatus. Neuron 30, 335–344. Peles, E., Joho, K., Plowman, G. D., and Schlessinger, J. (1997a). Close similarity between Drosophila neurexin IV and mammalian Caspr protein suggests a conserved mechanism for cellular interactions. Cell 88, 745–746. Peles, E., Nativ, M., Campbell, P. L., Sakurai, T., Martinez, R., Lev, S., Clary, D. O., Schilling, J., Barnea, G., Plowman, G. D., Grumet, M., and Schlessinger, J. (1995). The carbonic anhydrase domain of receptor tyrosine phosphatase b is a functional ligand for the axonal cell recognition molecule contactin. Cell 82, 251–260. Peles, E., Nativ, M., Lustig, M., Grumet, M., Martinez, R., Plowman, G. D., and Schlessinger, J. (1997b). IdentiWcation of a novel contactin-associated transmembrane receptor with multiple domains implicated in protein-protein interactions. EMBO J. 16, 978–988. Peles, E., and Salzer, J. L. (2000). Molecular domains of myelinated Wbers. Curr. Opin. Neurobiol. 10, 558–565. Pesheva, P., Gennarini, G., Goridis, C., and Schachner, M. (1993). The F3/11 cell adhesion molecule mediates the repulsion of neurons by the extracellular matrix glycoprotein J1–160/180. Neuron 10, 69–82. Peters, A., Palay, S. L., and Webster, H. d. (1991). ‘‘The Fine Structure of the Nervous System.’’ Oxford University Press, New York. Poliak, S., Gollan, L., Martinez, R., Custer, A., Einheber, S., Salzer, J. L., Trimmer, J., Shrager, P., and Peles, E. (1999). Caspr2, a new member of the neurexin superfamily is localized at the juxtaparanodes of myelinated axons and associates with Kþ channels. Neuron 24, 1037–1047. Poliak, S., Gollan, L., Salomon, D., Berglund, E. O., Ranscht, B., and Peles, E. (2001). Localization of Caspr2 in myelinated nerves depends on axon-glia interactions and the generation of barriers along the axon. J. Neurosci. 21, 7568–7575. Poliak, S., Matlis, S., Ullmer, C., Scherer, S. S., and Peles, E. (2002). Distinct claudins and associated PDZ proteins form diVerent autotypic tight junctions in myelinating Schwann cells. J. Cell Biol. 159, 361–371. Poliak, S., Salomon, S., Elhanany, H., Sabanay, H., Kiernan, B., Pevny, L., Stewart, C. L., Xu, X., Chiu, S., Shrager, P., Furley, A. W., and Peles, E. (2003). Juxtaparanodal clustering of Shaker-like Kþ channels in myelinated axons depends on Caspr2 and TAG-1. J. Cell Biol. 162, 1149–1160. Popko, B. (2000). Myelin galactolipids: Mediators of axon-glial interactions? Glia 29, 149–153. Qu, Y., Curtis, R., Lawson, D., Gilbride, K., Ge, P., DiStefano, P. S., Silos-Santiago, I., Catterall, W. A., and Scheuer, T. (2001). DiVerential modulation of sodium channel gating and persistent sodium currents by the b1, b2, and b3 subunits. Mol. Cell. Neurosci. 28, 570–580. Quick, D. C., and Waxman, S. G. (1977). Ferric ion, ferrocyanide, and inorganic phosphate as cytochemical reactants at peripheral nodes of Ranvier. J. Neurocytol. 6, 555–570. Raine, C. S. (1982). DiVerences between the nodes of Ranvier of large and small diameter Wbres in the P. N. S. J. Neurocytol. 11, 935–947. Raine, C. S. (1984). On the association between perinodal astrocytic processes and the node of Ranvier in the C. N. S. J. Neurocytol. 13, 21–27. Ramo´n y Cajal, S. (1928). ‘‘Degeneration and Regeneration of the Nervous System’’ (R. M. May, ed.). Oxford University Press, Oxford. Rasband, M., Trimmer, J. S., Schwarz, T. L., Levinson, S. R., Ellisman, M. H., Schachner, M., and Shrager, P. (1998). Potassium channel distribution, clustering, and function in remyelinating rat axons. J. Neurosci. 18, 36–47.

113

114

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

Rasband, M. N., Park, E. W., Zhen, D., Arbuckle, M. I., Poliak, S., Peles, E., Grant, S. G. N., and Trimmer, J. S. (2002). Clustering of neuronal potassium channels is independent of their interaction with PSD-95. J. Cell Biol. 159, 663–672. Rasband, M. N., Peles, E., Trimmer, J. S., Levinson, S. R., Lux, S. E., and Shrager, P. (1999a). Dependence of nodal sodium channel clustering on paranodal axoglial contact in the developing CNS. J. Neurosci. 19, 7516–7528. Rasband, M. N., and Shrager, P. (2000). Ion channel sequestration in central nervous system axons. J. Physiol. 525, 63–73. Rasband, M. N., and Trimmer, J. S. (2001). Developmental clustering of ion channels at and near the node of Ranvier. Develop. Biol. 236, 5–16. Rasband, M. N., Trimmer, J. S., Peles, E., Levinson, S. R., and Shrager, P. (1999b). Kþ channel distribution and clustering in developing and hypomyelinated axons of the optic nerve. J. Neurocytol. 28, 319–331. Rash, J. E., Yasumura, T., Dudek, F. E., and Nagy, J. I. (2001). Cell-speciWc expression of connexins and evidence of restricted gap junctional coupling between glial cells and between neurons. J. Neurosci. 21, 1983–2000. RatcliVe, C. F., Qu, Y., McCormick, K. A., Tibbs, V. C., Dixon, J. E., Scheuer, T., and Catterall, W. A. (2000). A sodium channel signaling complex: Modulation by associated receptor protein tyrosine phosphatase b. Nat. Neurosci. 3, 437–444. RatcliVe, C. F., Westenbroek, R. E., Curtis, R., and Catterall, W. A. (2001). Sodium channel beta 1 and beta 3 subunits associate with neurofascin through their extracellular immunoglobulin-like domain. J. Cell Biol. 154, 427–434. Reid, G., Scholz, A., Bostock, H., and Vogel, W. (1999). Human axons contain at least Wve types of voltagedependent potassium channels. J. Physiol. 518, 681–696. Revel, J.-P., and Hamilton, D. W. (1969). The double nature of the intermediate dense line in peripheral nerve myelin. Anat. Rec. 163, 7–16. Rieger, F., DaniloV, J. K., Pincon-Raymond, M., Crossin, K. C., Grumet, M., and Edelman, G. M. (1986). Neuronal cell adhesion molecules and cytotactin are colocalized at the node of Ranvier. J. Cell Biol. 103, 379–391. Rieger, F., Pincon-Raymond, M., Lombet, A., Ponzio, G., Lazdunski, M., and Sidman, R. L. (1984). Paranodal dysmyelination and increase in tetrodotoxin binding sites in the sciatic nerve of the motor end-plate disease (med/med) mouse during postnatal development. Dev. Biol. 101, 401–409. Rios, J. C., Melandez-Vasquez, C. V., Einheber, S., Lustig, M., Grumet, M., Hemperly, J., Peles, E., and Salzer, J. L. (2000). Contactin-associated protein (Caspr) and contactin form a complex that is targeted to the paranodal junctions during myelination. J. Neurosci. 20, 8354–8364. Ritchie, J. M., Black, J. A., Waxman, S. G., and Angelides, K. J. (1990). Sodium channels in the cytoplasm of Schwann cells. Proc. Natl. Acad Sci. USA 87, 9290–9294. Ritz, M. F., Lechner-Scott, J., Scott, R. J., Fuhr, P., Malik, N., Erne, B., Taylor, V., Suter, U., Schaeren-Wiemers, N., and Steck, A. J. (2000). Characterisation of autoantibodies to peripheral myelin protein 22 in patients with hereditary and acquired neuropathies. J. Neuroimmunol. 104, 155–163. Rosenbluth, J. (1976). Intramembranous particle distribution at the node of Ranvier and adjacent axolemma in myelinated axons of the frog brain. J. Neurocytol. 5, 731–745. Rosenbluth, J. (1985). Intramembranous particle patches in myelin-deWcient rat axons. Neurosci. Lett. 62, 19–24. Rosenbluth, J. (1995). Glial membranes and axoglial junctions. In ‘‘Neuroglia’’ (H. Kettenmann and B. R. Ransom, eds.), pp. 613–633. Oxford University Press, New York. Rosenbluth, J., Tao-Cheng, J.-H., and Blakemore, W. F. (1985). Dependence of axolemmal diVerentiation on contact with glial cells in chronically demyelinated lesions of cat spinal cord. Brain Res. 358, 287–302. Safronov, B. V., Kampe, K., and Vogel, W. (1993). Single voltage-dependent potassium channels in rat peripheral nerve membrane. J. Physiol. 460, 675–691. Sandri, C., Van Buren, J. M., and Akert, K. (1982). Membrane morphology of the vertebrate nervous system. Prog. Brain Res. 46, 201–265. Saperstein, D. S., Katz, J. S., Amato, A. A., and Barohn, R. J. (2001). Clinical spectrum of chronic acquired demyelinating polyneuropathies. Muscle Nerve 24, 311–324. Scherer, S. S., and Arroyo, E. J. (2002). Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7, 1–12. Scherer, S. S., Descheˆnes, S. M., Xu, Y.-T., Grinspan, J. B., Fischbeck, K. H., and Paul, D. L. (1995). Connexin32 is a myelin-related protein in the PNS and CNS. J. Neurosci. 15, 8281–8294. Scherer, S. S., Xu, T., Crino, P., Arroyo, E. J., and Gutmann, D. H. (2001). Ezrin, radixin, and moesin are components of Schwann cell microvilli. J. Neurosci. Res. 65, 150–164. Schwartz, M., Ernst, S. A., Siegel, G. J., and AgranoV, B. (1981). Immunocytochemical localization of (Naþ, Kþ)ATPase in the goldWsh optic nerve. J. Neurochem. 36, 107–115. Shah, B. S., Stevens, E. B., Gonzalez, M. I., Bramwell, S., Pinnock, R. D., Lee, K. S., and Dixon, A. K. (2000). b3, a novel auxiliary subunit for the voltage-gated sodium channel, is expressed preferentially in sensory neurons and is upregulated in the chronic constriction injury model of neuropathic pain. Eur. J. Neurosci. 12, 3985–3990. Shapiro, L., Fannon, A. M., Kwong, P. D., Thompson, A., Lehmann, M. S., Grubel, G., Legrand, J.-F., Als-Nielsen, J., Colman, D. R., and Hendrickson, W. A. (1995). Structural basis of cell-cell adhesion by cadherins. Nature 374, 327–337. Sheikh, K. A., Deerinck, T. J., Ellisman, M. H., and GriYn, J. W. (1999). The distribution of ganglioside-like moieties in peripheral nerves. Brain 122, 449–460.

CONCLUSION

Sheng, M., and Sala, C. (2001). PDZ domains and the organization of supramolecular complexes. Annu. Rev. Neurosci. 24, 1–29. Smart, S. L., Lopantsev, V., Zhang, C. L., Robbins, C. A., Wang, H., Chiu, S. Y., Schwartzkroin, P. A., Messing, A., and Tempel, B. L. (1998). Deletion of the Kv1.1 potassium channel causes epilepsy in mice. Neuron 20, 809–819. Sobko, A., Peretz, A., Shirihai, O., Etkin, S., Cherepanova, V., Dagan, D., and Attali, B. (1998). Heteromultimeric delayed-rectiWer Kþ channels in Schwann cells: Developmental expression end role in cell proliferation. J. Neurosci. 18, 10398–10408. Spiegel, I., and Peles, E. (2002). Cellular junctions of myelinated nerves. Mol. Membrane Biol. 19, 95–101. Spiegel, I., Salomon, D., Erne, B., Schaeren-Wiemers, N., and Peles, E. (2002). Caspr3 and Caspr4, two novel members of the Caspr family are expressed in the nervous system and interact with PDZ domains. Mol. Cell. Neurosci. 20, 283–297. Srinivasan, Y., Elmer, L., Davis, H., Bennett, V., and Angelides, K. (1988). Ankyrin and spectrin associate with voltage-dependent sodium channels in brain. Nature 333, 177–180. Srinivasan, Y., Schachner, M., and Catterall, W. (1998). Interaction of voltage-gated sodium channels with the extracellular matrix moleculres tenascin-C and tenascin-R. Proc. Natl. Acad. Sci. USA 95, 15753–15757. Stolinski, C., and Breathnach, A. S. (1982). Freeze-fracture replication of mammalian peripheral nerve—a review. J. Neurol. Sci. 57, 1–28. Stolinski, C., Breathnach, A. S., Martin, B., Thomas, P. K., King, R. M. H., and Gabriel, G. (1981). Associated particle aggregates in juxtaparanodal axolemma and adaxonal Schwann cell membrane of rat peripheral nerve. J. Neurocytol. 10, 679–691. Stolinski, C., Breathnach, A. S., Thomas, P. K., Gabriel, G., and King, R. M. H. (1985). Distribution of particle aggregates in the internodal axolemma and adaxonal Schwann cell membrane of rodent peripheral nerve. J. Neurol. Sci. 67, 213–222. Streit, W. J., Schulte, B. A., Spicer, S. S., and Balentine, J. D. (1985). Histochemical localization of galactosecontaining glycoconjugate at peripheral nodes of Ranvier in the rat. J. Histochem. Cytochem. 33, 33–39. Tabira, T., M. J., C., Reier, P. J., and Webster, H. d. (1978). An experimental analysis of interlamellar tight junctions in amphibian and mammalian C. N. S. myelin. J. Neurocytol. 7, 489–503. Tait, S., Gunn-Moore, F., Collinson, J. M., Huang, J., Lubetzki, C., Pedraza, L., Sherman, D. L., Colman, D. R., and Brophy, P. J. (2000). An oligodendrocyte cell adhesion molecule at the site of assembly of the paranodal axo-glial junction. J. Cell Biol. 150, 657–666. Takeichi, M. (1990). Cadherins: A molecular family important in selective cell-cell adhesion. Annu. Rev. Biochem. 59, 237–252. Tao-Cheng, J.-H., and Rosenbluth, J. (1984). Extranodal particle accumulations in the axolemma of myelinated frog optic axons. Brain Res. 308, 289–300. TetzlaV, W. (1982). Tight junction contact events and temporary gap junctions in the sciatic nerve Wbres of the chicken during Wallerian degeneration and subsequent regeneration. J. Neurocytol. 11, 839–858. Thomas, F. K., Trajaborg, W., Nagy, C., Santoro, M., Sadiq, S. A., Latov, N., and Hays, A. P. (1991). Experimental autoimmune neuropathy with anti-GM1 antibodies and immunoglobulin deposits at the nodes of Ranvier. Acta Neuropathol. 82, 378–383. Thomas, P. K., Berthold, C.-H., and Ochoa, J. (1993). Microscopic anatomy of the peripheral nervous system. In ‘‘Peripheral Neuropathy’’ (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low, and J. F. Poduslo, eds.), pp. 28–91. W. B. Saunders, Philadelphia. Traka, M., Dupree, J. L., Popko, B., and Karagogeos, D. (2002). The neuronal adhesion protein TAG-1 is expressed by Schwann cells and oligodendrocytes and is localized to the juxtaparanodal region of myelinated fibers. J. Neurosci. 11, 3016–3024. Trapp, B. D., Andrew, S. B., Cootauco, C., and Quarles, R. (1989a). The myelin-associated glycoprotein is enriched in multivesicular bodies and periaxonal membrane of actively myelinating oligodendrocytes. J. Cell Biol. 109, 2417–2425. Trapp, B. D., Andrews, S. B., Wong, A., O’Connell, M., and GriYn, J. W. (1989b). Co-localization of the myelinassociated glycoprotein and the microWlament components, F-actin and spectrin, in Schwann cells of myelinated nerve Wbers. J. Neurocytol. 18, 47–60. Tsukita, S., and Furuse, M. (2000). Pores in the wall: Claudins constitute tight junction strands containing aqueous pores. J. Cell Biol. 149, 13–16. Tsukita, S., Furuse, M., and Itoh, M. (2001). Multifunctional strands in tight junctions. Nat. Rev. Mol. Cell Biol. 2, 285–293. Tzoumaka, E., Tischler, A. C., Sangameswaran, L., Eglen, R. M., Hunter, J. C., and Novakovic, S. D. (2000). DiVerential distribution of the tetrodotoxin-sensitive rPN4/NaCh6/Scn8a sodium channel in the nervous system. J. Neurosci. Res. 60, 37–44. Ullrich, B., Ushkaryov, Y. A., and Sudhof, T. C. (1995). Cartography of neurexins: More than 1000 isoforms generated by alternative splicing and expressed in distinct subsets of neurons. Neuron 14, 497–507. Ushkaryov, Y. A., Petrenko, A. G., Geppert, M., and Sudhof, T. C. (1992). Neurexins—synaptic cell surface proteins related to the a-latrotoxin receptor and laminin. Science 257, 50–56. Vabnick, I., and Shrager, P. (1998). Ion channel redistribution and function during development of the myelinated axon. J. Neurobiol. 37, 80–96. Vabnick, I., Trimmer, J. S., Schwarz, T. L., Levinson, S. R., Risal, D., and Shrager, P. (1999). Dynamic potassium channel distributions during axonal development prevent aberrant Wring patterns. J. Neurosci. 19, 747–758.

115

116

4. FUNCTIONAL ORGANIZATION OF THE NODES OF RANVIER

Van Dyke, D. H., Griggs, R. C., Murphy, M. J., and Goldstein, M. N. (1975). Hereditary myokymia and periodic ataxia. J. Neurol. Sci. 25, 109–118. Volkmer, H., Leuschner, R., Zacharias, U., and Rathjen, F. G. (1996). Neurofascin induces neurites by heterophilic interactions with axonal NrCAM while NrCAM requires F11 on the axonal surface to extend neurites. J. Cell Biol. 135, 1059–1069. Volkmer, H., Zacharias, U., Norenberg, U., and Rathjen, F. G. (1998). Dissection of complex molecular interactions of neurofascin with axonin-1, F11, and tenascin-R, which promote attachment and neurite formation of tectal cells. J. Cell Biol. 142, 1083–1093. Vorbrodt, A. W., Lossinsky, A. S., and Wisneiwski, H. M. (1982). Cytochemical localization of ouabain-sensitive, Kþ-dependent p-nitro-phenylphosphatase (transport ATPase) in the mouse central and peripheral nervous systems. Brain Res. 243, 225–234. Wagner, R., and Kachar, B. (1995). Linear and tight junctions assemblies between endothelial cell in the eel rete mirabile. Anat. Rec. 242, 545–552. Wang, H., Kunkel, D. D., Martin, T. M., Schwartkroin, P. A., and Tempel, B. L. (1993). Heteromultimeric Kþ channels in terminal and juxtaparanodal regions of neurons. Nature 365, 75–79. Waxman, S. G., and Black, J. A. (1987). Macromolecular structure of the Schwann cell membrane. Perinodal microvilli. J. Neurol. Sci. 77, 23–34. Weber, P., Bartsch, U., Rasband, M. N., Czaniera, R., Lang, Y., Bluethmann, H., Margolis, R. U., Levinson, S. R., Shrager, P., Montag, D., and Schachner, M. (1999). Mice deWcient for tenascin-R display alterations of the extracellular matrix and decreased axonal conduction velocities in the CNS. J. Neurosci. 19, 4245–4262. Willison, H. J., and Yuki, N. (2002). Peripheral neuropathies and anti-glycolipid antibodies. Brain 135, 2591–2625. Wilson, G. F., and Chiu, S. Y. (1990). Ion channels in axon and Schwann cell membranes at paranodes of mammalian myelinated Wbers studied with patch clamp. J. Neurosci. 10, 3263–3274. Wood, J. G., Jean, D. H., Whitaker, J. N., McLaughlin, B. J., and Albers, R. W. (1977). Immunocytochemical localization of the sodium, potassium activated ATPase in knifeWsh brain. J. Neurocytol. 6, 571–581. Xiao, Z.-c., Bartsch, U., Margolis, R. K., Rougon, G., Taylor, J., Montag, D., and Schachner, M. (1997). Isolation of a tenascin-R binding protein from mouse brain. J. Biol. Chem. 272, 32092–32101. Xiao, Z. C., Ragsdale, D. S., Malhotra, J. D., Mattei, L. N., Braun, P. E., Schachner, M., and Isom, L. L. (1999). Tenascin-R is a functional modulator of sodium channel beta subunits. J. Biol. Chem. 274, 26511–26517. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D. (2001). P0 protein is a target antigen in chronic inXammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50, 286–292. Young, P., Boussadia, O., Berger, P., Leone, D. P., Charnay, P., Kemler, R., and Suter, U. (2002). E-cadherin is required for the correct formation of autotypic adherens junctions of the outer mesaxon but not for the integrity of myelinated Wbers of peripheral nerves. Mol. Cell. Neurosci. 21, 341–351. Yuan, L. L., and Ganetzky, B. (1999). A glial-neuronal signaling pathway revealed by mutations in a neurexinrelated protein. Science 283, 1343–1345. Zerr, P., Adelman, J. P., and Maylie, J. (1998). Episodic ataxia mutations in Kv1.1 alter potassium channel function by dominant negative eVects or haploinsuYciency. J. Neurosci. 18, 2842–2848. Zhang, X., and Bennett, V. (1996). IdentiWcation of O-linked N-acetylglucosamine modiWcation of ankyrinG isoforms targeted to nodes of Ranvier. J. Biol. Chem. 271, 31391–31398. Zhang, X., and Bennett, V. (1998). Restriction of 480/270-kD ankyrinG to axon proximal segments requires multiple ankyrinG-speciWc domains. J. Cell Biol. 142, 1571–1581. Zhou, D. X., Lambert, S., Malen, P. L., Carpenter, S., Boland, L. M., and Bennett, V. (1998a). AnkyrinG is required for clustering of voltage-gated Na channels at axon initial segments and for normal action potential Wring. J. Cell Biol. 143, 1295–1304. Zhou, L., Zhang, C. L., Messing, A., and Chiu, S. Y. (1998b). Temperature-sensitive neuromuscular transmission in Kv1.1 null mice: Role of potassium channels under the myelin sheath in young nerves. J. Neurosci. 18, 7200–7215. Zimmermann, H. (1996). Accumulation of synaptic vesicle proteins and cytoskeletal specializations at the peripheral node of Ranvier. Microsc. Res. Technique 34, 462–473.

C H A P T E R

5 Electrophysiologic Consequences of Myelination Stephen G. Waxman and Lakshmi Bangalore

INTRODUCTION Several million people around the world are aVected by demyelinating diseases, which include acquired diseases such as multiple sclerosis (MS) and Guillain-Barre disease, as well as hereditary neurodegenerative disorders such as the leukodystrophies. A common underlying motif in all these disorders is the slowing or block of axonal conduction associated with the lack of myelin around axons, due to either damage to the myelin (demyelination) or improper formation of myelin (dysmyelination). Thus, many of the signs and symptoms of demyelinating diseases, such as weakness, visual loss, somatosensory loss, and bladder and bowel malfunctions, reXect the abnormal conduction of nerve impulses along demyelinated or dysmyelinated axons. This chapter discusses the molecular substrates for impulse conduction and the physiology of impulse conduction in myelinated axons, and the pathophysiology of demyelinated axons in which impulse conduction is impaired. Conduction abnormalities following demyelination are attributable not only to changes in the passive electrical properties due to myelin loss, but also due to alterations in active electrogenic properties reXecting the molecular organization within the axonal membrane (Waxman, 1999). In this chapter, the role of demyelination in symptom production, and the role of molecular plasticity within neurons and glia in the remission of clinical deWcits is reviewed in the context of the physiology of impulse conduction. Restoration of conduction by pharmacological modulation as well as other current strategies for the development of new therapeutic interventions are also discussed.

THE MYELINATED NERVE FIBER Axons in vertebrates are unique in being surrounded by a myelin sheath. The myelin sheath, produced by Schwann cells in the PNS and oligodendrocytes in the CNS, is wrapped around the axons in segments (internodes) in which the glial membrane spirals around the axon with extrusion of the cytoplasmic contents and compaction until adjacent cytoplasmic and external faces of the membrane directly appose each other, so that the myelin acquires a high resistance and low capacitance. The number of concentric layers of myelin wrapped around the axon depends on the size of the Wber; large Wbers may have as many as 50 concentric spirals of myelin, and small Wbers may have as few as 2 or 3. This compacted layered structure of myelin endows the myelinated Wber with properties that

Myelin Biology and Disorders, Volume 1

117

Copyright 2004, Elsevier (USA). All rights reserved.

118

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

reduce the current Xow, especially the capacitative current Xow, across the internodal axonal membrane and therefore focus depolarizations produced by action potentials upstream along the Wbers on the periodic interruptions in the myelin, between adjacent internodes. These periodic interruptions in the myelin sheath, termed nodes of Ranvier, are present at distances ranging from less than 100 mm (small-diameter Wbers) to slightly over 1 mm (larger-diameter Wbers). While Schwann cells myelinate single axons, individual oligodendrocytes can produce myelin sheaths around as many as 100 axons (Bjartmar et al., 1994). The oligodendrocyte cell body, unlike a Schwann cell, usually does not surround its myelin sheaths and is connected to its myelin sheath via thin cytoplasmic bridges (Bunge et al., 1961; Hirano, 1968). This suggests a tenuous link between the genomic and biosynthetic machinery within the oligodendrocyte cell body and the myelin, an arrangement that has been interpreted as explaining the scarcity of remyelination within the CNS. There is evidence for local synthesis of myelin in distal parts of the oligodendroglial processes close to the myelin sheaths, where polyribosomes are present and focal axon-oligodendrocyte interactions appear to regulate properties of the myelin sheath such as thickness and the internode distance (Waxman and Sims, 1984; Waxman et al., 1988). The distal ends of the oligodendrocyte processes undoubtedly depend, however, on the oligodendrocyte cell body for maintenance. As a result of their structural specialization, myelinated Wbers conduct impulses in a rapid and discontinuous or saltatory manner, in contrast to nonmyelinated Wbers in which impulses are conducted slowly in a continuous manner (Huxley and Sta¨mpXi, 1949; Tasaki, 1959). The impulse conduction velocity in myelinated Wbers is approximately proportional to the Wber diameter, compared to nonmyelinated Wbers in which conduction velocity is proportional to diameter1/2. Thus, above a critical diameter of approximately 0.2 mm, at which the conduction velocity-diameter relationships for nonmyelinated and myelinated Wbers intersect, myelinated Wbers conduct impulses more rapidly than nonmyelinated Wbers of the same diameter (Waxman and Bennett, 1972) (Fig. 5.1). Consistent with the adaptive role of myelin in increasing conduction velocity, neurons with axonal diameters of more than 1.0 mm in the PNS and of more than 0.2 mm in the CNS are generally myelinated (Waxman and Bennett, 1972).

MEMBRANE ARCHITECTURE OF THE MYELINATED AXON The membrane architecture of myelinated axons is highly specialized and organized to support saltatory conduction. Nonmyelinated axons have uniform properties along their length (Black et al., 1981) and conduct action potentials in a continuous manner. In contrast, the structure of the axon membrane of myelinated Wbers is extremely heterogeneous with foci of strategically placed voltage-sensitive ion channels (Fig. 5.2). Notably, at the node of Ranvier, where action potential electrogenesis takes place in normal myelinated axons, voltage-dependent sodium (Naþ) channels cluster at a high density (approximately 1000/mm2) (Ritchie and Rogart, 1977; Waxman, 1977). The density of Naþ channels at the internodal axon membrane under the myelin sheath is much lower (<25/mm2) (Ritchie and Rogart, 1977; Shrager, 1989; Waxman, 1977) and cannot reliably support conduction under most circumstances. Nearly a dozen sodium channels have been identiWed so far, each with a unique molecular structure superimposed on an invariant overall motif of four similar domains containing six membrane-spanning segments. It is not known at this time how many types of sodium channels are present in myelinated axons. One sodium channel subtype, Nav 1.6, is known to be expressed at high levels at most nodes in the PNS and CNS (Fig. 5.3) (Caldwell et al., 2000), although it is also present along nonmyelinated axons where it participates in action potential electrogenesis (Black et al., 2002). At some nodes, other sodium channel subtypes (e.g., Nav1.2, Nav1.8) are present, but this is less common than Nav1.6.

MEMBRANE ARCHITECTURE OF THE MYELINATED AXON

FIGURE 5.1 Relation between conduction velocity (Y-axis) and diameter (X-axis), shown for both myelinated (dashed line) and nonmyelinated (solid line) axons. Note that above a diameter of 0.2 mm, at which the two relationships intersect, myelinated axons conduct more rapidly than nonmyelinated axons of the same size. ModiWed from Waxman and Bennett, 1972.

FIGURE 5.2 Working model of ion channel organization of the myelinated Wber gNaþ ¼ sodium channels; gKs ¼ fast potassium channels. Sodium channels are clustered at the node of Ranvier. In contrast, fast potassium channels, responsible for repolarization of the action potential, are present in the axon membrane beneath the myelin.

FIGURE 5.3 Fluorescent image of a node of Ranvier in a teased Wber from the sciatic nerve, merged with a scanning transmission diVerential contrast image to show the myelin sheath. Intense Nav1.6 immunostaining at the node (red) is bounded by labeling (green) for Caspr, a transmembrane protein that forms heterodimers with contactin in the paranode. Scale bar ¼ 5 m. From Black et al., 2002.

119

120

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

The development of Nav1.6 – rich nodes along mature myelinated axons is the end point of a complex process. During early stages of development of myelinated axons, Nav1.2 (which is present along nonmyelinated axons) is present along premyelinated axons and then is initially clustered at immature nodes. However, as the Wbers mature, Nav1.6 replaces Nav1.2 as the major nodal sodium channel (Boiko et al. 2001). The role of glial cell-derived factors or glial contact in the development of nodal sodium channel clusters remains controversial, with some studies suggesting that channel clustering is dependent on glial cell contact (see, e.g., Ching et al. 1999; Rasband et al. 1999) and other studies suggesting that channel clustering can occur independent of glial cell contact (Deerink et al. 1997). Kaplan et al. (2001) have suggested that a soluble oligodendrocyte-derived factor can induce clustering of Nav1.2 but not of Nav1.6. Myelinated Wbers also express multiple potassium channels. On the basis of intra-axonal recordings, three types of potassium channels can be distinguished: ‘‘fast’’ Kþ channels, ‘‘slow’’ Kþ channels, and an inward rectiWer that is permeable to both Kþ and Naþ. The distribution of fast Kþ channels is complementary to that of the Naþ channels that are clustered in the nodal axon membrane. Fast Kþ channels, which can be blocked by externally applied 4-aminopyridine (4-AP), are expressed at relatively low densities in the nodal axon membrane and are present in highest density in the axon membrane under the myelin (Chiu and Ritchie, 1980, 1981; Eng et al., 1988; Foster et al., 1982; Kocsis et al., 1982; Ritchie et al., 1981). It has been suggested on the basis of a voltage clamp study that the density of fast Kþ channels is highest in the paranode, decreasing to one-sixth of the paranodal density in the node and internode (Ro¨per and Schwarz, 1989). This study, however, may not have accurately diVerentiated between paranodal and juxtaparanodal channels. Molecular analysis indicates that Kv1.1 and Kv1.2 potassium channel subunits, which combine as heterotetramers, are present with the juxtaparanodal part of the axon membrane (Rasband et al. 1998; Vabnick and Shrager 1998; Wang et al. 1993) where they appear to prevent hyperexcitability or re-entry phenomena (Zhou et al. 1998). The high density of Naþ channels at the nodal axon membrane and the shortage of these channels in the axon membrane under the myelin, combined with the expression of fast Kþ channels in the axon membrane under the myelin, have important implications for axonal pathophysiology. Following demyelination, the low density of Naþ channels in the exposed internodal axon membrane results in a capacitative but nonelectrogenic load because a low density of inward Naþ current is insuYcient for action potential electrogenesis, and this interferes with the conduction of action potentials. Moreover, fast Kþ channels are unmasked when the overlying myelin is lost and tend to clamp the demyelinated axon membrane close to the Kþ equilibrium potential Ek, further impeding action potential conduction (Chiu and Ritchie, 1981). Naþ, Kþ -ATPase molecules are also present in the axon membrane at the node (Ariyasu and Ellisman, 1987; Wood et al., 1977) where they maintain ionic homeostasis and produce electrogenic hyperpolarization (Gordon et al., 1990). It has been demonstrated, using intra-axonal recordings, that experimentally demyelinated PNS axons are hyperpolarized due to Naþ, Kþ -ATPase activity (Bostock and Grafe, 1985). Ouabain, an inhibitor of Naþ, Kþ -ATPase, reduces the hyperpolarization and can lead to restoration of conduction. On the basis of these Wndings, Bostock and Grafe (1985) suggested that pharmacological blockade of Naþ, Kþ -ATPase activity might improve conduction of high-frequency impulse trains in demyelinated axons. The heterogeneous distribution of molecules along the axon also has important implications for the recovery of conduction following demyelination. The molecular reorganization within the demyelinated axon membrane, with the acquisition of a higher-than-normal density of Naþ channels in some demyelinated regions, which develop the capability to support action potential conduction, is described later. The distribution of channels within the myelinated Wber also suggests the possibility of pharmacologically blocking the fast Kþ channels that are unmasked by demyelination, thereby increasing the safety factor for conduction in demyelinated axons.

ACTION POTENTIAL CONDUCTION IN MYELINATED AXONS

ACTION POTENTIAL CONDUCTION IN MYELINATED AXONS As mentioned earlier, myelinated Wbers are constructed to support saltatory conduction in which action potentials leap from one node to another in a discontinuous or saltatory manner. An action potential begins with an inXux of sodium ions due to the transient opening of the sodium channels, followed by closing and inactivation of the sodium channels and the opening of potassium channels and exit of potassium ions, thus reversing the depolarization of the nodal axon membrane. Additional sodium channels are activated as the positive charge from the inXux of sodium ions depolarizes a series of downstream nodes, thus enabling the action potentials to move from node to node. Even though, in principle, this system might support bidirectional propagation or echoing of impulses, action potential conduction is unidirectional in nature, partly because sodium channels close soon after activation and remain refractory to additional sodium ions for a short time. This endows the axon with a refractory period for transmission and insures that, in normal Wbers, impulse conduction occurs in a unidirectional manner without echoing. The insulating action of myelin around normal myelinated Wbers facilitates the movement of action currents from each active node of Ranvier, through the relatively lowresistance axoplasm, to the subsequent node (Fig. 5.4A). Following excitation of a given node, mammalian myelinated axons at 378C reach a threshold for excitation at the next node after approximately 20 msec (internodal conduction time) (Rasminsky and Sears, 1972). Thus, the impulse travels along the axon by jumping sequentially from node to node in a saltatory manner with a high conduction velocity. The safety factor, a measure of reliability of conduction, is the ratio between current available to stimulate a node of Ranvier and the current required to stimulate the node). In normal myelinated Wbers there is a high degree of reliability for impulse conduction with a safety factor of around 5–7 (Tasaki, 1959).

IMPAIRED ACTION POTENTIAL CONDUCTION IN DEMYELINATED AXONS Following damage to myelin, the density of action current is reduced as a result of capacitative and resistive shunting (see Fig. 5.4B). In axons in which the safety factor is reduced but still >1.0, there is an increase in charging time for the nodal membrane such that it will take longer than normal for the axon to reach threshold, and conduction velocity is therefore reduced. In axons with severe demyelination, the safety factor can be reduced to values <1.0, and the threshold will not be reached; hence conduction will fail (Waxman, 1982). Demyelination is not an all-or-none process,and can range from minimal damage to the myelin (or even isolated loss of the seal between myelin and axon without damage to the myelin itself) to loss of the entire myelin sheath. Demyelinated axons therefore display a spectrum of conduction abnormalities that are negative in the jacksonian sense, including slowed, desynchronized, or blocked conduction (see Fig. 5.5B throuth 5.5E). Conduction abnormalities that are positive in the jacksonian sense also occur and are presumably the result of ectopic impulse generation or abnormal cross-talk between demyelinated axons (see Fig. 5.5F through 5.5.I). As shown in Figure 5.5, conduction abnormalities in demyelinated axons include decreased conduction velocity, failure to transmit high-frequency trains of action potentials, or total conduction block (McDonald, 1963; McDonald and Sears, 1970). These abnormalities are conWned to the zone of demyelination in focally demyelinated Wbers, with apparently normal conduction in regions proximal and distal to the area of demyelination. The internodal conduction time in demyelinated ventral root Wbers can be increased to nearly 500 msec compared with approximately 20 msec in normal Wbers (Rasminsky and Sears, 1972). Thus, the conduction velocity is reduced because the impulse takes longer to travel along the Wber (see Fig. 5.5B).

121

122

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

FIGURE 5.4 Diagrammatic representation of current Xow associated with conduction through normally (A) myelinated and (B) demyelinated regions of an axon. The action potential is conducted from left to right (arrow). This idealized diagram represents the myelin as a perfect insulation. Dashed arrows illustrate current Xow resulting from an action potential that is located at the cross-hatched node. Current is lost in demyelinated regions as a result of capacitative and resistive shunting.

Depending on the degree of damage to myelin, the increase in internodal conduction time can vary between individual demyelinated internodes along single Wbers and between neighboring demyelinated Wbers in a nerve or CNS tract. Loss of synchrony in CNS tracts or temporal dispersion, in which diVerent Wbers exhibit diVerent degrees of conduction slowing, is a result of variably decreased conduction velocity and reXects the span of internodal conduction times seen in the demyelinated axons (see Fig. 5.5C). Temporal dispersion can produce clinical abnormalities by disrupting functions that require synchronous discharge and may account, for example, for the loss of deep tendon reXexes in the course of demyelinating polyneuropathies. Demyelinated Wbers also display conduction block. Conduction block can be frequencyrelated, with high-frequency impulse trains failing to propagate while low-frequency trains conduct reliably (see Fig. 5.5D); or conduction failure can be complete, with single action potentials failing to propagate beyond the zone of demyelination (see Fig. 5.5E). Hyperpolarization of the axon membrane, due to electrogenic pump (Naþ, Kþ -ATPase) activity, can contribute to conduction block of high-frequency impulse trains (Bostock and Grafe, 1985). In addition, increased intracellular Naþ concentration at the ‘‘driving node’’ (Rasminsky and Sears, 1972) and depolarization of demyelinated axons due to increases in extracellular Kþ concentration (Brismar, 1981) may contribute to block of high-frequency impulse trains. Demyelination also results in positive conduction abnormalities, as illustrated in Figures 5.5F through 5.5I. Ectopic action potential generation (see Fig. 5.4F) has been observed, for example, in demyelinated axons in cat dorsal columns (Smith and McDonald, 1980). DiVerences in the accommodative properties of demyelinated sensory versus motor axons (Bowe et al., 1987) are consistent with the more frequent occurrence of paresthesia (a positive sensory abnormality), compared with extraneous motor activity, in MS patients. Increased mechanosensitivity (Smith and McDonald, 1980) has also been observed in CNS demyelinated Wbers (see Fig. 5.5G) and probably accounts for clinical phenomena such as Lhermitte’s symptom (Vollmer et al., 1991). In addition, abnormal Wber-Wber excitation (cross-talk) may occur between abnormally myelinated axons (see Fig. 5.5H). This cross-talk, which has also been termed ephaptic transmission, occurs at frequencies up to 70 Hz in the dysmyelinated spinal cord axons of the dy-dy mouse, although in this system it appears to occur preferentially in a unidirectional manner in this model system, from bare Wbers to myelinated Wbers (Rasminsky, 1981).

IMPAIRED ACTION POTENTIAL CONDUCTION IN DEMYELINATED AXONS

123

FIGURE 5.5 Conduction abnormalities in demyelinated axons. Demyelinated regions of the axon are diagrammatically shown as dashed lines. Cell bodies are located to the left and axon terminals to the right. The direction of normal conduction is indicated by the arrow; see text for further explanation.

Impulse reXection or echoing may occur in some demyelinated Wbers (Burchiel, 1980) and may result in extraneous sensory or motor activity such as paresthesia, pain, or tonic spasms. Because the reXected (antidromic) impulses can collide with and abolish orthodromic impulses, impulse reXection can also interfere with normal impulse traYc. Positive conduction abnormalities may interact by positive feedback, with ectopic activity being ampliWed by, for example, impulse reXection or cross-talk (Rasminsky, 1981). The high packing density of demyelinated axons in some plaques, with no or few intervening glial elements, may contribute to cross-talk and abnormal discharge. Block of Naþ channels, without demyelination, can also produce conduction block or a decrease in axonal conduction velocity (Stys et al., 1992a; Yokota et al., 1994) and the literature contains several reports of ‘‘blocking factors’’ which were proposed to contribute to clinical deWcits in demyelinated diseases but could not be reproduced. Several reports have contended that acute paralysis in Guillain-Barre syndrome may be due to a Naþ channel-blocking factor in the cerebrospinal Xuid (Brinkmeier et al., 1992); and that the cerebrospinal Xuid of MS patients also contains the same factor that can impair Naþ channel function (Brinkmeier et al., 1993). A more recent report identiWed this purported blocking factor as a pentapeptide with the sequence Gln-Tyr-Asn-Ala-Asp (QYNAD) and it was reported to block sodium channels at concentrations as low as 10 mM, by shifting the steady-state inactivation to more negative potentials (Brinkmeier et al. 2000). However, Cummins et al. (2003) have recently demonstrated that QYNAD has no sodium channel

124

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

blocking eVects even at concentrations as high as 500 mM. Notably, this latter study examined the eVects of QYNAD on the Nav1.6 channel (which is heavily expressed in myelinated axons) and in the Nav1.2 channel (which is expressed in dysmyelinated and demyelinated axons) and did not reveal block, even at concentrations as high as 500 mM (50-fold higher than the blocking concentration originally reported by Brinkmeier et al. 2000). Thus, at this time, there has been no deWnitive demonstration of a Naþ channel blocking factor that contributes to the pathophysiology of Guillain-Barre syndrome or MS.

LABILITY OF ACTION POTENTIAL CONDUCTION IN DEMYELINATED AXONS The clinical status of patients with MS can be labile, changing even on an hour-to-hour basis. Although poorly understood, this lability appears to have a basis in the pathophysiology of demyelinated axons. Several studies have conWrmed that there can be a striking sensitivity to temperature in MS patients, with clinical status worsening as temperature increases (Guthrie, 1951; Watson, 1959; Waxman and Geschwind, 1983). Increased susceptibility to thermally induced, reversible conduction block has been demonstrated in demyelinated axons (Davis and Jacobson, 1971). The underlying mechanism for conduction block in experimentally demyelinated Wbers includes a reduced time integral of inward (depolarizing) nodal current that occurs because sodium channels open and close more quickly at higher temperatures, as a result the temperature-dependence of channel kinetics (Rasminsky, 1973). Reversible conduction block can occur even with small increases in temperature (0.58 C) in some Wbers. It has been demonstrated that with the lowering of temperature conduction block in demyelinated axons can be temporarily overcome and the action potential prolonged (Sears and Bostock, 1981). Schauf and Davis (1974), in an analysis of thermal sensitivity in demyelinated axons, concluded that as the safety factor is decreased from its normal value of 5–6, conduction should become increasingly labile. In axons with safety factors only slightly greater than 1.0, conditions that marginally decrease it (i.e., conditions that might have little, if any, eVect on conduction in normal Wbers) would be expected to lead to transient conduction failure. Conversely, in demyelinated axons with safety factor slightly less than 1.0, conditions that increase safety factor would be expected to restore conduction. Therefore, changes in metabolic status, as well as alterations in temperature would be expected to temporarily alter the conduction properties of demyelinated Wbers. This has, in fact, been demonstrated in human subjects with MS. For instance, alterations in acid-base balance, which can be produced by oral phosphate or intravenous bicarbonate administration or by hyperventilation, can produce reversible alterations in the neurological status of patients with MS (Becker et al., 1974; Davis et al., 1970). These observations in humans illustrate the sensitivity of demyelinated Wbers to change in metabolic status. The lability of action potential conduction in demyelinated axons suggests that pharmacological modiWcation in channel kinetics, or in other membrane properties, may enhance action potential conduction. This has provided part of the basis for the development of drugs that improve safety factor in demyelinated axons.

RELATIONSHIP BETWEEN CONDUCTION ABNORMALITIES AND CLINICAL STATUS In patients with MS, clinicopathological and neuroimaging studies have demonstrated that demyelinated plaques can be present without necessarily producing clinical deWcits. For example, in some MS patients with demyelination of the total cross-section of the optic nerves for more than 1 cm, useful visual acuity is still retained (Ulrich and Groebke–Lorenz, 1983; Wisniewski et al., 1976). As indicated by evoked potential studies, demyelinating lesions can produce slowed conduction, without necessarily leading to

RECOVERY OF IMPULSE CONDUCTION IN DEMYELINATED AXONS

clinical deWcit. For instance, in MS patients who have never experienced visual diYculties, or after recovery from optic neuritis, visual evoked responses may be delayed by more than 40 ms (Halliday et al., 1973). Parallel improvements in visual acuity and amplitude of the visual evoked response generally accompany recovery from acute optic neuritis; however, recovery of the prolonged evoked response latency usually is not a prerequisite for clinical recovery. Subclinical slowing of conduction along demyelinated axons is not a rare phenomenon. A study on evoked potentials (Hume and Waxman, 1988) in MS suspects who subsequently developed clinically deWnite MS (McAlpine criteria) showed that a signiWcant proportion of evoked potential abnormalities (59% of the visual evoked potential abnormalities that were observed; 41% somatosensory; 50% auditory) were clinically silent. The observations that asymptomatic demyelinated plaques are not uncommon in MS patients, and that increases in conduction latency do not necessarily produce clinical abnormalities, are not entirely unexpected, since the input-output functions of some neural systems are relatively robust, even in the face of changes in latency that can be measured in milliseconds. SigniWcant changes in conduction velocity occur with physiological changes in body temperature since the Q10 for conduction velocity is greater than 1.5 in CNS axons (Moore et al., 1978; Stys et al., 1992; Swadlow et al., 1981). Even though conduction velocity—and thus conduction time—is dependent on temperature, most neural systems continue to function eYciently, even in poikilothermic species, as temperature changes within the physiological range. Conduction block prevents the transmission of action potentials that encode information and is correlated with clinical abnormalities. Because neural information can be encoded by the pattern, or frequency, of impulses transmitted along axons in some CNS tracts, the ability of myelinated axons to carry trains of impulses at speciWc frequencies, or in speciWc patterns, may be an important determinant of clinical status. Loss of even a fraction of impulses within a train, as a result of frequency-related conduction block, can degrade neural messages. A somatosensory evoked potential study showed that impaired frequency following in MS patients was quantitatively correlated with degree of clinical involvement (Namerow, 1968b). Total conduction block produces signiWcant loss of function unless alternative conduction pathways are available. As indicated earlier, increased temperature can produce conduction block in demyelinated axons. Consistent with this, MS patients are often sensitive to temperature, with symptoms becoming more pronounced as body temperature rises (Guthrie, 1951; Namerow, 1968a; Watson, 1959; Waxman and Geschwind, 1983). Increased temperature is correlated with increased conduction velocity as well as decreased safety factor in CNS axons in situ. The worsening of clinical status in MS patients as temperature rises provides support for the idea that conduction block is a more important determinant of clinical status that decreased conduction velocity.

RECOVERY OF IMPULSE CONDUCTION IN DEMYELINATED AXONS Clinical remissions in patients with MS appear to have at least a partial basis in the restoration of impulse conduction along demyelinated axons. Recovery of secure conduction in some chronically demyelinated axons can be inferred from clinicopathological observations. Functional visual acuity is retained or recovered in some MS patients with extensive demyelination (>1 cm) of the total cross section of the optic nerves (Ulrich and Groebke-Lorenz, 1983; Wisniewski et al., 1976). In such cases, recovery of function can not be attributed to mechanisms such as synaptic plasticity, utilization of alternative conduction pathways, or remyelination. This indicates that at least a subpopulation of the demyelinated axons must have retained or recovered the ability to conduct action potentials.

CONTINUOUS CONDUCTION Early electrophysiological studies of demyelinated axons revealed slow conduction velocities similar to those of nonmyelinated axons (Cragg and Thomas, 1964; Lehmann and Ule,

125

126

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

1964) and were interpreted as suggesting that in demyelinated axons, action potentials might be conducted in a continuous manner similar to propagation in premyelinated or nonmyelinated axons. Longitudinal current analysis in single, undissected normal nerve Wbers has conWrmed saltatory conduction, with excitability apparently conWned to the nodes of Ranvier (Rasminsky and Sears, 1972). On the contrary, a continuous mode of action potential conduction has been observed in some demyelinated axons. Action potentials can propagate continuously along extensive regions of the demyelinated axon, as observed in rat ventral root Wbers exposed to diphtheria toxin where the inward membrane current can travel continuously along demyelinated axons segments for distances greater than the average internode distance (Bostock and Sears, 1976; Bostock and Sears, 1978). Beginning as early as 4 to 6 days after demyelination, continuous conduction occurs with a velocity of 1/20 to 1/40 of the normal, saltatory conduction velocity for these Wbers. Thus, at least in some regions, demyelinated internodal axon membrane can develop widespread excitability that supports continuous conduction. Some axons can conduct over continuous lengths of demyelination exceeding 2 mm (several internodes) in length (Felts et al., 1977) as demonstrated by recordings from axons within experimentally demyelinated lesions within the spinal cord and subsequent ultrastructural analysis. Considering what we know about the molecular architecture of axons, how can continuous conduction occur following demyelination? Computational modeling indicates that in some small-caliber demyelinated axons, the density of preexisting Naþ channels in the demyelinated region may be large enough to support conduction of at least single action potentials (Hines and Shrager, 1991; Waxman and Brill, 1978). Indeed, in some smalldiameter (<0.5 mm) axons that lack myelin, action potential conduction can be supported by low Naþ channel densities (<40/mm2) (Ritchie et al., 1976; Strichartz et al., 1984; Tang et al., 1979), at least if temperature is lowered. In the smallest axons with a diameter of <0.25 mm, the conduction of action potentials can be supported by Naþ channel densities of < 10/mm2 (Waxman et al., 1989). These Wndings may apply to small-diameter axons that have been demyelinated or to axons in which the diameter has decreased following demyelination. The diameter of some demyelinated axons is, in fact, reduced in regions where myelin is lost (Prineas and Connell, 1978; Smith et al., 1983), possibly as a result of decreased neuroWlament phosphorylation and increased neuroWlament packing density in demyelinated regions (Brady, 1993; De Waegh et al., 1992). Nevertheless, these results may not apply to larger axons, because of their lower input impedance. Because of this, reorganization of the demyelinated axon membrane through the acquisition of a higherthan-normal Naþ channel density may be required for restoration of conduction. There is experimental evidence that indicates that this membrane reorganization does, in fact, take place. Early ultrastructural studies of experimentally demyelinated axons demonstrated the development of regions of axon membrane with cytochemical (Foster et al., 1980) and freeze-fracture (Black et al., 1987) properties similar to those of nodal membrane. Further evidence for the acquisition of increased Naþ channel densities in the demyelinated axon membrane has been provided by immunocytochemical studies. Studies on Wsh lateral line nerves revealed relatively high densities of Naþ channels in previously internodal regions (England et al., 1990). Similarly, immunocytochemical observations 14 to 21 days after injection of the demyelinating toxin doxorubicin suggest the expression of sodium channels at newly formed nodes along mammalian remyelinated axons (Dugandzija-Novakovic et al., 1995). These early studies, however, utilized generic sodium channel antibodies that did not diVerentiate between channel subtypes. More recent studies using subtype-speciWc antibodies have begun to provide evidence supporting the suggestion (Waxman 1987) that the acquisition of higher-than-normal densities of Naþ channels in persistently demyelinated axon regions may represent a dediVerentiation of the axon membrane (discussed later). Even though the mature internodal membrane appears to be incapable of secure conduction (Ritchie and Rogart, 1977; Waxman, 1977), the premyelinated axon membrane, (including regions destined to develop into internodal membrane), is electrically excitable (Foster et al., 1982; Waxman et al.,

CONTINUOUS CONDUCTION

FIGURE 5.6 Changes in Nav1.6 and Nav1.2 immunostaining in EAE. Representative digital images of section of optic nerve from control (A–B) and EAE (E—F), showing immunostaining for Nav1.6 and Nav1.2, and for Caspr. Red indicates Nav1.6 immunostaining in panels A, C, and E; red indicates Nav1.2 in B, D, and F; green indicates Caspr. A and B demonstrate nodes with Nav1.6 (A) and less commonly Nav1.2 (B) positive immunostaining (yellow arrowheads) in control optic nerve. C through E and D through F demonstrate a pattern of progressive demyelination and nodal disruption in EAE. Some nodes in EAE show weak and diVuse Caspr immunostaining (inset, C). There is a loss of nodes with Nav1.6 immunostaining (C and E) and increased frequency of nodes with Nav1.2 immunostaining (D and F). Demyelinated axonal proWles with diVuse Nav1.6 (E, arrow) and, more commonly, Nav1.2 (D, F, arrows) immunostaining are seen in EAE. ModiWed from Craner et al., 2003.

1989). Studies on the developing internodal axon during normal ontogenesis suggest that suppression of Naþ channel expression reduces excitability after axons are covered by myelin (Black et al., 1986). This ensures that conduction is not compromised due to a premature loss of sodium channels during development. Molecular studies have recently provided evidence that there is, in fact, a dediVerentiation of the axon membrane in some demyelinated axons. Using subtype-speciWc antibodies for immunocytochemical analysis of demyelinated CNS axons in experimental allergic encephalomyelitis, Craner et al. (2003) observed a signiWcant switch from Nav1.6 to Nav1.2 expression, with a reduction in the frequency of Nav1.6-positive nodes and an increase in the frequency of Nav1.2-positive nodes (Figs. 5.6A and 5.B). There was also evidence for a diVuse distribution of Nav1.2 channels and, in a smaller number of axons, of Nav1.6 channels which extended for more than 30 mm (i.e. as far as the axons could be followed within sections) in some demyelinated axons (Figs. 5.6D through 5.6F). The diVuse pattern of Nav1.2 immunostaining is similar to the pattern of Nav1.2 expression along premyelinated axons (Boiko et al. 2001) and may provide a substrate for continuous conduction of impulses. The physiological implications of a reversion to Nav1.2 expression (rather than Nav1.6) is not fully understood. There is some evidence from patch clamp studies (Zhou and Goldin 2002) suggesting that diVerent functional properties may allow Nav1.6 channels to maintain higher Wring rates. Thus, substitution of Nav1.2 for Nav1.6 may permit conduction to continue in demyelinated axons, but with less reliability in terms of sustained high-frequency Wring.

Nonuniform Conduction Following Demyelination In some demyelinated axons, a diVerent mode of impulse propagation, nonuniform conduction, may also contribute to recovery of conduction. During nonuniform conduction,

127

128

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

the action potential propagates between isolated foci of inward current generation (termed ‘‘phi-nodes’’), which are scattered aggregations of Naþ channels that are established before remyelination. These aggregations or hot spots may represent the precursors of nodes in Wbers that are about to remyelinate. In ventral root axons demyelinated with lysophosphatidylcholine, phi-nodes develop several days before remyelination (Smith et al., 1982). Initially phi-nodes appear as isolated foci of inward current spaced at 100 to 200 mm (i.e., the distance between the nodes that will form during demyelination). Phi-nodes appear to support conduction through the lesion in some Wbers studied several days before remyelination. Electron microscopy has demonstrated an ultrastructural correlate of phi-nodes, specialized regions of the axon membrane similar to the node of Ranvier, with a dense cytoplasmic undercoating that extends for several micrometers long (Blakemore and Smith, 1983). Freeze-fracture studies have demonstrated a counterpart for these axon membrane specializations. Intramembranous E-face particles of a size similar to those in the nodal axon are found within patches of the axon membrane along axons demyelinated with ethidium bromide (Rosenbluth and Blakemore, 1984) and along axons that have been deprived of myelination by irradiation during gliogenesis (Black et al., 1986), and appear to represent clusters of sodium channels. Consistent with the notion that each patch is the precursor of a node of Ranvier, the number of intramembranous particles in each patch is also approximately the same as at a mature node.

MOLECULAR BASIS FOR MEMBRANE REORGANIZATION IN DEMYELINATED AXONS It has been argued that spread of preexisting Naþ channels by diVusion alone is not suYcient to support action potential conduction through demyelinated regions (Hines and Shrager, 1991). Patch-clamp recordings at former nodes of Ranvier in demyelinated axons reveal sharp gradients in channel density suggesting that Naþ channels do not diVuse away from the node in large numbers following demyelination (Shrager, 1989). There is also a signiWcant increase in Naþ channel concentration per weight of tissue (approximately three fold at 21 to 28 days after demyelination) as determined by radioimmunoassays, providing some evidence for the insertion of new Naþ channels into the demyelinated axon membrane of the Wsh lateral line nerve (England et al., 1991). Consistent with the notion that new Naþ channels are inserted into the demyelinated axon membrane, quantitative autoradiography with [3H]saxitoxin (STX) in demyelinated white matter in MS patients has revealed a fourfold increase in STX-binding sites, compared with normal white matter, (Moll et al., 1991). In addition, as noted earlier, Craner et al. (2003) have observed the expression of Nav1.2 channels (which are present in only small numbers in normal myelinated axons) along extensive regions of demyelinated axons, suggesting the production of new channels. The origin of the Naþ channels destined for the demyelinated axon membrane has not yet been identiWed. It is likely that some channels are synthesized within the neuronal cell body and translocated to the demyelinated regions by axonal transport. Craner et al. (2003) noted up-regulated expression of Nav1.2 mRNA in neuronal cell bodies giving rise to Nav1.2-expressing axons in experimental allergic encephalomyelitis, a Wnding that strongly suggests channel synthesis within the neuronal cell body. Axonal transport of Naþ channels has been demonstrated in peripheral nerves (Lombet et al., 1985). In some cases (e.g., deployment of Nav1.6 to demyelinated axons), the change in channel expression appears to be adaptive. However, increased synthesis and development of sodium channels in demyelinated neurons does not necessarily imply that the appropriate types of sodium channels are always added to the neuronal membrane. Glial cells have been postulated to be another possible source of new sodium channels (Bevan et al., 1985; Gray and Ritchie, 1985). Naþ channel-rich axon membrane at the node is contacted by the perinodal astrocyte processes and Schwann cell microvilli in a very

A CHANNELOPATHY IN MS?

speciWc manner (Hildebrand, 1971; Raine, 1984; Waxman and Black, 1984). Glial cell processes are apposed to demyelinated axons at sites of Naþ channel clustering (Black et al., 1984; Bray et al., 1979; Rosenbluth, 1979; Rosenbluth, 1985; Rosenbluth and Blakemore, 1984), and the two specializations (glial contact and Naþ channel clustering) are usually precisely juxtaposed, suggesting that glial cell contact may be necessary for Naþ channel clustering in the axon membrane. In addition, there is evidence that astrocytes express voltage-sensitive Naþ channels (Barres et al., 1989; Bevan et al., 1985; Black et al., 1994a; Nowak et al., 1987; Sontheimer et al., 1992; Sontheimer et al., 1991) as well as mRNAs for neuronal-type Naþ channel a- and b-subunits (Black et al., 1994b; Oh et al., 1994; Oh and Waxman, 1994). It should be stressed, however, that there has been no demonstration of Naþ channel transfer from glial cells to axons. Increased numbers of Naþ channels have been observed along the demyelinated axons in demyelinated Wsh lateral line nerves despite injection of doxorubincin which kills Schwann cells. On this basis it has been argued that glial synthesis of Naþ channels is unlikely (England et al., 1991). Alternative functions for astrocyte Naþ channels such as providing a return pathway for Naþ ions that maintains and modulates astrocyte Naþ and Kþ -ATPase activity have been proposed (Sontheimer et al., 1994). It has also been suggested (Waxman, 1992) that astrocytes may secrete extracellular molecules that participate in the targeting or anchoring of Naþ channels at sites of channel aggregation.

A CHANNELOPATHY IN MS? Nearly a dozen distinct sodium channel genes encode distinct channels. Small variations within the amino acid sequences of the diVerent sodium channels result in diVerent kinetics, voltage-dependencies, and pharmacological proWles. At least eight sodium channel genes are known to be expressed in the CNS (Black et al., 1996). Maladaptive changes in the pattern of sodium channel gene expression (an acquired channelopathy) have been observed following axonal transection. Axotomized dorsal root ganglion neurons express abnormal combinations of sodium channel mRNA (Dib-Hajj et al., 1996, 1998; Waxman et al., 1994) that are associated with abnormal patterns of sodium current expression. Axonal transection of spinal sensory neurons results in downregulation of the expression of sodium channels Nav1.8 (sensory-neuron-speciWc, SNS) and Nav1.9 (NaN) and a decrease of their currents (Dib-Hajj et al., 1999; Sleeper et al., 2000). The downregulation of Nav1.8 and Nav1.9 is accompanied by increased expression of the sodium channel Nav1.3 (type III) (Black et al., 1999a; Waxman et al., 1994) and a simultaneous emergence of a rapidly-repriming TTXsensitive sodium current (Cummins and Waxman, 1997) attributable to Nav 1.3 (Cummins et al., 2001). These changes in channel expression contribute to hyperexcitability of dorsal root ganglion cells that underlies neuropathic pain. Recent evidence suggests that an acquired channelopathy may occur in association with demyelinating disorders (Black et al., 1999b, 2000, Waxman, 2001; Westenbroek et al., 1992). Abnormal expression of sodium channel Nav1.8 (SNS), normally expressed only in spinal sensory and trigeminal neurons, is detected as shown in Figure 5.7, in cerebellar Purkinje cells of patients with progressive MS who had displayed cerebellar deWcits on examination (Black et al., 2000). This upregulation of Nav1.8 is also detected in the Purkinje cells of Taiep rats in which myelin is lost because of a mutation aVecting oligodendrocytes (Black et al., 1999b) and in mice with chronic relapsing experimental allergic encephalomyelitis (CR-EAE), an inXammatory model of MS (Black et al., 2000). Sodium channel Nav1.8 contributes a substantial fraction of the inward transmembrane current during action potential electrogenesis within the cells in which it is normally present (Renganathan et al., 2001). It has not yet been deWnitively determined whether the expression of Nav1.8 within the cerebellum in demyelinating disorders is adaptive (contributing to restoration of conduction along demyelinated Purkinje cell axons) or maladaptive (perturbing the pattern of impulse activity in Purkinje cells). Transfection of Purkinje cells with Nav1.8 results in substantially altered patterns of impulse activity, in favor of the idea that the expression of Nav1.8 in Purkinje cells is maladaptive (Renganathan et al., 2003).

129

130

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

FIGURE 5.7 Expression of sodium channel Nav1.8/SNS is upregulated within cerebellar Purkinje cells within brains obtained at post-mortem from MS patients. Panels at left show in situ hybridization with Nav1.8-speciWc antisense riboprobes, and demonstrates the lack of SNS mRNA in control cerebellum (c) and its presence in Purkinje cells in post-mortem tissue from two patients with MS (a, b). No signal is present following hybridization with sense riboprobe (d). Panels on the right show immunostaining with antibody directed against Nav1.8, and illustrate lack of Nav1.8 protein in control cerebellum (g, arrow indicates Purkinje cell) and its presence in MS (e, f). ModiWed from Black et al., 2000.

IMPEDANCE MISMATCHING IN DEMYELINATED AXONS As indicated earlier, restoration of impulse conduction in chronically demyelinated axons is correlated with the development of higher-than-normal Naþ channel density in the bared axon membrane. However, the development of high Naþ channel density in demyelinated axon regions does not, in itself, ensure reliable impulse conduction. Impedance mismatch reXects the passive properties of the nerve Wber and is due to electrical loading that occurs (largely as a result of increased membrane capacitance) at sites of axonal inhomogeneity, such as the transition between myelinated and demyelinated regions. At these sites, impedance mismatch can increase the total charge movement (and, thus, the charging time) needed for action potential propagation (Waxman, 1978). Impedance mismatch can delay or prevent the action potential from invading a demyelinated region and thus can be a signiWcant factor causing conduction failure, even if there is a very high density of Naþ channels. Figure 5.8A shows computer-simulated action potentials (Waxman and Brill, 1978) in a Wber in which one internode (between nodes D1 and D4) has totally demyelinated, under the assumption that the demyelinated axon membrane has developed a high Naþ channel density (similar to that at nodes). Conduction failure occurs at the junction between normal and demyelinated axon regions (D1), despite the high Naþ channel density in the demyelinated area, because impedance mismatch prevents threshold from being reached. Impedance mismatch must be overcome in order for action potentials to successfully invade the demyelinated zone. Mechanisms for overcoming impedance mismatching include development of relatively short myelinated segments proximal to the demyelinated area (Waxman and Brill, 1978); decrease in axon diameter of the demyelinated region relative to the proximal myelinated segment (Sears and Bostock, 1981), development of an increased Naþ channel density at the node upstream to the demyelinated area, and development of a specialized transition zone with relatively high Naþ channel densities or relatively low Kþ channel densities at the border of the region of demyelination (Waxman and Wood, 1984). As shown in Figure 5.8B, these factors can overcome impedance

PHARMACOLOGICAL RESTORATION OF IMPULSE CONDUCTION IN DEMYELINATED AXONS

mismatch, thus facilitating the invasion of impulses into demyelinated axons where they can propagate if Naþ channel density is high enough.

PHARMACOLOGICAL RESTORATION OF IMPULSE CONDUCTION IN DEMYELINATED AXONS The use of pharmacological agents as therapeutic strategies to improve the safety factor and thereby overcome conduction block is suggested by the hyperpolarizing eVects of fast Kþ channels and Naþ, Kþ -ATPase in demyelinated axons. Blockade of fast Kþ channels, resulting in an increased inward transmembrane current along with prolonged action potential duration after, would be expected to produce greater depolarization of the axon membrane at sites of demyelination. And blockade of Naþ, Kþ -ATPase would be expected to attenuate the pump-mediated hyperpolarization, bringing the membrane

FIGURE 5.8 (A) Computer simulations showing conduction through a focally demyelinated axon. Even in the presence of an adequate density of sodium channels in the demyelinated zone (D1-D4), conduction can fail as a result of impedance mismatch. (B) Interposition of two short internodes, just proximal to the demyelinated zone, provides impedance matching that facilitates conduction into, and then through, the demyelinated zone. From Waxman and Brill, 1978.

131

132

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

FIGURE 5.9 EVect of 4-AP on conduction block in demyelinated Wber. (A) Experimental design permits examination of conduction through normal (stimulation at S1) or demyelinated (stimulation at S2) regions. (B) Stimulation of S1 leads to an action potential, but conduction block occurs following S2 stimulation, when the demyelinated zone is interposed between the stimulation and recording site. (C) Application of 4-AP leads to overcoming of the conduction block. NS, normal solution. ModiWed from Targ and Kocsis, 1986.

closer to threshold. In fact, drugs that block fast Kþ channels (such as 4-aminopyridine [4-AP]) have been shown to reverse conduction block and prolong the action potential in demyelinated spinal roots (Bostock et al., 1981), sciatic nerve (Targ and Kocsis, 1985, 1986) and posterior column axons (Blight, 1989). 4-AP has a relatively speciWc eVect on increasing the duration of action potentials in demyelinated axons because the fast Kþ channels that are sensitive to 4-AP are usually covered by the myelin in normal axons. Focal application of 4-AP results in the reversal of conduction block in experimentally demyelinated axons, as shown in Figure 5.9 (Targ and Kocsis, 1985). In this experiment, stimulating electrodes were located on both ends of a focally demyelinated nerve, and recordings from single axons were obtained with a microelectrode on one side of the lesion (Fig. 5.9A). The response to stimulation of the axon proximal to the lesion, so that the conduction path did not include the demyelinated zone, could be recorded as a propagated action potential. In contrast, following stimulation on the contralateral side of the lesion in the absence of 4-AP, so that conduction had to cross the zone of demyelination, there was conduction failure at the site of demyelination so that an action potential could not be recorded (Fig. 5.9B). The nerve was then exposed to 4-AP, which blocks fast Kþ channels. Following exposure to 4-AP, conduction proceeded with a high safety factor so that the action potential traveled successfully, albeit with an increased latency reXecting decreased conduction velocity, through the demyelinated axon (Fig. 5.9C). Clinical studies have documented transient improvement in motor function, reduction in scotoma and improved critical Xicker fusion in MS patients following treatment with 4-AP and with the related drug 3, 4 diamininopyridine, (3,4-DAP) (Davis et al., 1990; Stefoski et al., 1987). Additional clinical studies are under way to fully determine the therapeutic potential and side eVect proWle of these agents in MS and spinal cord injury. The results available to date demonstrate, at a minimum, the rationale for the use of pharmacological approaches to reverse conduction block in demyelinated axons.

RESTORATION OF CONDUCTION THROUGH REMYELINATION

Inhibition of Naþ, Kþ-ATPase by pharmacological drugs has also been studied as a strategy to overcome conduction block in demyelinated axons. It has been reported that systemic digitalis can reverse the latency increase and improve frequency following, as studied by somatosensory evoked potentials, in rats with focal demyelinated lesions of the spinal cord (Kaji and Sumner, 1989a). Increased conduction velocity and a reversal of conduction block have also been reported in demyelinated axons following exposure to ouabain (Kaji and Sumner, 1989b). MS patients treated with intravenous digoxin (0.22 mg/kg body weight) displayed small degrees of clinical improvement concurrent with changes in evoked potentials (Kaji et al., 1990). All these studies suggest that blockade of Naþ, Kþ -ATPase activity may provide an alternate approach to restoring conduction in patients with demyelinating disorders. It is conceivable that inhibition of Naþ, Kþ-ATPase could be used in parallel with blockade of fast Kþ channels.

RESTORATION OF CONDUCTION THROUGH REMYELINATION Computer simulations have suggested that remyelination with even thin, or short, myelin segments can provide a shield against capacitative and resistive current loss and should thus promote conduction through previously demyelinated Wbers, provided that remyelinated nodes of Ranvier develop membrane properties similar to those in normal Wber (Funch and Faber, 1984; Koles and Rasminsky, 1972; Waxman and Brill, 1978). Remyelinated PNS Wbers conduct at velocities close to those predicted from their internode distances, suggesting the development of near normal Naþ channel densities at the newly formed nodes (Brill et al., 1977). And newly formed nodes along remyelinated axons develop normal cytochemical properties, suggesting that they acquire a high density of sodium channels (Weiner et al., 1980). Indeed, newly formed nodes along remyelinated axons appear to develop a high density of Naþ channels as indicated by studies on experimentally demyelinated-remyelinated axons in sciatic nerve, which display an increase in STX binding that is proportional to the increase in nodal membrane area imposed by the shorter spacing between remyelinated nodes (Ritchie, 1982). Conduction velocity is restored to near normal and the ability to conduct high-frequency impulse trains is improved in demyelinated peripheral nerves following remyelination (Smith and McDonald, 1980). Similarly, conduction is enhanced in experimentally demyelinated CNS axons following remyelination. Even though the remyelinated Wbers are surrounded by abnormally thin and short myelin sheaths, the increase in the refractory period of transmission can be at least partially reversed following remyelination (Smith et al., 1981; Smith et al., 1983). In rats with experimental allergic encephalomyelitis, clinical recovery is correlated with remyelination in both the PNS and CNS (Stanley et al., 1991). Remyelination of CNS axons as well as enhancement of conduction can be achieved by both oligodendrocytes and Schwann cells. Schwann cell–mediated remyelination of dorsal columns, for example, appears to be eVective in enhancing action potential conduction and also improves the capability of these Wbers to conduct the high-frequency impulse trains (Blight and Young, 1989; Felts et al., 1991). These observations provide a basis for transplantation of myelin-forming cells into the demyelinated CNS.

Transplantation of Myelin-Forming Cells Restoration of conduction in demyelinated axons through transplantation of exogenous, myelin-forming cells or their precursors is currently an area of active investigation. Transplanted Schwann cells and oligodendrocytes have been shown to form myelin with a relatively normal compact structure around demyelinated axons with the CNS (Duncan et al., 1988; Gout et al., 1988; Rosenbluth et al., 1989). 02-A glial progenitor cells can also diVerentiate and form myelin when transplanted into the CNS (Groves et al., 1993). However, as indicated earlier, the success of transplantation as a strategy for improving conduction depends not only on whether compact myelin has been formed, since

133

134

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

impedance mismatch, incomplete or patchy myelination, or failure to form nodes with adequate Naþ channel density can impair conduction. Early electrophysiological studies demonstrated that myelin-deWcient axons in the md rat spinal cord (Utzschneider et al., 1994) display a marked increase in conduction velocity, which approached myelinated control values, as a result of formation of myelin by transplanted glial cells from unaVected littermates. The remyelinated axons were able to initiate action potentials outside the transplant region, which could then invade and propagate into the region of demyelination, as well as propagate beyond the demyelinated region. Similar restoration in action potential conduction has been observed following transplantation of Schwann cells (Honmou et al., 1996), olfactory ensheathing cells (Imaizumi et al., 1998), as well as bone marrow stromal cells (Akiyama et al., 2002) to demyelinated lesions within the spinal cord of adult rats. These results suggest that action potential conduction in demyelinated and dysmyelinated axons can be enhanced following myelination by several types of exogenous transplanted cells. Further studies are needed and are under way.

AXONAL DEGENERATION AND NEUROPROTECTION IN DEMYELINATING DISEASES Although MS has traditionally been viewed as a demyelinating disease, recent attention has been focused on axonal degeneration in MS (Ferguson et al., 1997; Trapp et al., 1998; Waxman, 1998). There is evidence suggesting that axonal loss contributes to the acquisition of nonremitting, permanent deWcits in MS (Davie et al., 1995). Whether axonal loss is due to inXammation or some other cause such as loss of myelin or of a glial factor is not yet clear. Irrespective of this, it is known that axonal degeneration does not depend on excitotoxicity (Ransom et al., 1991). Stys et al. (1992a) showed that in anoxic white matter, Naþ inXux via persistently activated Naþ channels collapses the transmembrane gradient of Naþ in axons, thereby driving the Naþ/Ca2þ exchanger to operate in a ‘‘reverse’’ mode, importing deleterious concentrations of calcium that lead to axonal degeneration. As might be predicted from this, sodium channel blockers such as tetrodotoxin (Stys et al., 1992a), tertiary and quaternary anaesthetics (Stys et al., 1992b), mexitine (Stys and Lesiuk, 1996), and phenytoin (Fern et al., 1993) have a protective eVect on the isolated anoxic optic nerve, protecting axons so that they do not lose the ability to function. Recently, Lo et al. (2002) extended this to an in vivo model of neuroinXammation, experimental allergic encephalomyelitis (EAE), and showed that phenytoin has a protective eVect, substantially reducing the number of axons that degenerate. The target of phenytoin’s action (i.e., the channel subtype(s) that permit the damaging persistent sodium inXux) are not yet known. Nav1.6 is known to produce a persistent sodium current (Maurice et al., 2001; Raman et al., 1997; Smith et al., 1998). The heavy expression of Nav1.6 at nodes of Ranvier (Caldwell et al., 2000) and its presence along some demyelinated axons (Craner et al., 2003) suggest that it might participate in the axonal injury cascade. Consistent with this hypothesis, expression of Nav1.6 is down-regulated in many demyelinated axons (Craner et al., 2003), and demyelinated CNS axons are known to be relatively resistant to anoxic injury (Imaizumi et al., 1998). Other sodium channel isoforms can also, however, produce persistent currents. IdentiWcation of the responsible channels may facilitate the development of neuroprotective strategies that slow or prevent axonal degeneration in disorders such as MS.

OVERVIEW AND PROSPECTS As discussed in this chapter, multiple mechanisms contribute to the eYciency and speed of conduction in myelinated axons, to abnormal action potential conduction after demyelination, and to restoration of conduction in demyelinated and remyelinated axons. Con-

OVERVIEW AND PROSPECTS

duction within demyelinated axons is shaped by both passive and active properties of the Wber. Although the multifactorial determinants of altered axonal conduction after demyelination have made it more diYcult to study, they have provided multiple options for the development of therapies that may enhance the function of demyelinated axons. Among them are pharmacological manipulation of Kþ channels, promotion of Naþ channel expression in demyelinated axons, manipulation of cellular characteristics at the edge of the plaque, and transplantation of myelin-forming cells which may restore conduction in demyelinated axons. Restoration of action potential conduction in demyelinated axons may contribute to improvement of clinical status in a number of disorders including Guillain-Barre syndrome and MS. Given the pace of recent progress, this appears to be a realistic objective.

Acknowledgments Research in the authors’ laboratory has been supported in part by grants from the National Multiple Sclerosis Society and the Rehabilitation Research Service and Medical Research Service, Department of Veterans AVairs, and by gifts from the Eastern Paralyzed Association, Paralyzed Veterans of America, and the Nancy Davis Foundation.

References Akiyama, Y., Radtke, C., and Kocsis, J. D. (2002). Remyelination of the rat spinal cord by transplantation of identiWed bone marrow stromal cells. J Neurosci 22(15), 6623–6630. Ariyasu, R. G., and Ellisman, M. H. (1987). The distribution of (Naþ/Kþ)ATPase is continuous along the axolemma of unensheathed axons from spinal roots of ‘dystrophic’ mice. J Neurocytol 16, 239–248. Barres, B. A., Chun, L. L. Y., and Corey, D. P. (1989). Glial and neuronal forms of the voltage–dependent sodium channels: Characteristics and cell-type distribution. Neuron 2, 1375–1388. Becker, F. O., Michael, J. A., and Davis, F. A. (1974). Acute eVects of oral phosphate on visual function in multiple sclerosis. Neurology 24, 601–607. Bevan, S., Chiu, S. Y., Gray, P. T. A., and Ritchie, J. M. (1985). The presence of voltage-gated sodium, potassium and chloride channels in rat cultured astrocytes. Proc. Roy. Soc. Lond. B 225, 229–313. Bjartmar, C., Hildebrand, C., and Loinder, K. (1994). Morphological heterogeneity of rat oligodendrocytes: Electron microscopic studies on serial sections. Glia 11, 235–244. Black, J. A., Cummins, T. R., Plumpton, C., Chen, Y. H., Hormuzdiar, W., Clare, J. J., and Waxman, S. G. (1999a). Upregulation of a silent sodium channel after peripheral, but not central, nerve injury in DRG neurons. J Neurophysiol 82, 2776–2785. Black, J. A., Dib-Hajj, S., Baker, D., Newcombe, J., Cuzner, M. L., and Waxman, S. G. (2000). Sensory neuronspeciWc sodium channel SNS is abnormally expressed in the brains of mice with experimental allergic encephalomyelitis and humans with multiple sclerosis. Proc Natl Acad Sci USA 97, 11598–11602. Black, J. A., Dib-Hajj, S., McNabola, K., Jeste, S., Rizzo, M. A., Kocsis, J. D., and Waxman, S. G. (1996). Spinal sensory neurons express multiple sodium channel alpha-subunit mRNAs. Mol Brain Research 43, 117–131. Black, J. A., Fjell, J., Dib-Hajj, S., Duncan, I. D., O’Connor, L. T., Fried, K., Gladwell, Z., Tate, S., and Waxman, S. G. (1999b). Abnormal expression of SNS/PN3 sodium channel in cerebellar Purkinje cells following loss of myelin in the taiep rat. NeuroReport 10, 913–918. Black, J. A., Foster, R. E., and Waxman, S. G. (1981). Freeze-fracture ultrastructure of rat CNS and PNS nonmyelinated axolemma. J Neurocytol 10, 981–993. Black, J. A., Renganathan, M., and Waxman, S. G. (2002). Sodium channel Nav1.6 is expressed along nonmyelinated axons and it contributes to conduction. Molec Brain Res. 105, 19–28. Blakemore, W. F., and Smith, K. J. (1983). Node–like axonal specializations along demyelinated central nerve Wbers: Ultrastructural observations. Acta Neuropathol 60, 291–296. Black, J. A., Waxman, S. G., and Hildebrand, C. (1984). Membrane specialization and axo-glial association in the rat retinal nerve Wber layer: Freeze-fracture observations. J Neurocytol 13, 417–430. Black, J. A., Waxman, S. G., Sims, T. J., and Gilmore, S. A. (1986). EVects of delayed myelination by oligodendrocytes and Schwann cells on the macromolecular structure of axonal membrane in rat spinal cord. J Neurocytol 15, 745–762. Black, J. A., Waxman, S. G., and Smith, M. E. (1987). Macromolecular structure of axonal membrane during acute experimental allergic encephalomyelitis in rat and guinea pig spinal cord. J Neuropath Exp Neurol 46, 167–184. Black, J. A., Westenbroek, R., Ransom, B. R., Catterall, W. A., and Waxman, S. G. (1994a). Type II sodium channels in spinal cord astrocytes in situ: Immunocytochemical observations. Glia 12, 219–227.

135

136

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

Black, J. A., Yokoyama, S., Waxman, S. G., Oh, Y., Zur, K. B., Sontheimer, H., Higashida, H., and Ransom, B. R. (1994b). Sodium channel mRNAs in cultured spinal cord astrocytes: In situ hybridization in identiWed cell types. Brain Res Mol Brain Res 23, 235–245. Blight, A. R. (1989). EVect of 4-AP on the axonal conduction block in chronic spinal cord injury. Brain Res Bull 22, 47–52. Blight, A. R., and Young, W. (1989). Central axons in injured cat spinal cord recover electrophysiological function following remyelination by Schwann cells. J Neurol Sci 91, 15–34. Boiko, T., Rasband, M. N., Levinson, S. R., Caldwell, J. H., Mandel, G., Trimmer, J. S., and Matthews, G. (2001). Compact myelin dictates the diVerential targeting of two sodium channel isoforms in the same axon. Neuron 30, 91–104. Bostock, H., and Grafe, P. (1985). Activity–dependent excitability changes in normal and demyelinated rat spinal root axons. J Physiol (Lond) 365, 239–257. Bostock, H., and Sears, T. A. (1976). Continuous conduction in demyelinated mammalian nerve Wbres. Nature 263, 786–787. Bostock, H., and Sears, T. A. (1978). The internodal axon membrane: Electrical excitability and continuous conduction in segmental demyelination. J Physiol Lond 280, 273–301. Bostock, H., Sears, T. A., and Sherratt, R. M. (1981). The eVects of 4–aminopyridine and tetraethylammonium ions on normal and demyelinated mammalian nerve Wbers. J Physiol (Lond) 313, 301–315. Bowe, C. M., Kocsis, J. D., Targ, E. F., and Waxman, S. G. (1987). Physiological eVects of 4-aminopyridine on demyelinated mammalian motor and sensory Wbers. Ann Neurol 22, 264–268. Brady, S. T. (1993). Axonal dynamics and regeneration. In ‘‘Neuroregeneration,’’ pp. 7–36. Raven Press, New York. Bray, G. M., Cullen, M. J., Aguayo, A. J., and Rasminsky, M. (1979). Node-like areas of intramembranous particles in the unensheathed axons of dystrophic mice. Neurosci Letts 13, 203–208. Brill, M. H., Waxman, S. G., Moore, J. W., and Joyner, R. W. (1977). Conduction velocity and spike conWguration in myelinated Wbers: Computed dependence on internode distance. J Neurol Neurosurg Psychiat 40, 769–774. Brinkmeier, H., Aulkemeyer, P., Wollinsky, K. H., and Rudel, R. (2000). An endogenous pentapeptide acting as a sodium channel blocker in inXammatory autoimmune disorders of the central nervous system. Nat Med 6, 808–811. Brinkmeier, H., Wollinsky, K. H., Hulser, P-J., Seewald, M. J., Mehrkens H. H., and Kornhuber, H. H. (1992). The acute paralysis in Guillain-Barre syndrome is related to Naþ channel blocking factor in the cerebrospinal. Eur J of Physiol 421, 552–557. Brinkmeier, H., Wollinsky, K. H., Seewald, M. J., Hulser, P. J., Mehrkens, H. H., Kornhuber, H. H., and Rudel, R. (1993). Factors in the cerebrospinal Xuid of multiple sclerosis patients interfering with voltage-dependent sodium channels. Neurosci Letts 156, 172–175. Brismar, T. (1981). SpeciWc permeability properties of demyelinated rat nerve Wbers. Acta Physiol Scand 113, 167–176. Bunge, M. B., Bunge, R. P., and Pappas, G. D. (1961). Electron microscope demonstration of connections between glia and myelin sheaths in the developing mammalian central nervous system. J Cell Biol 12, 448–453. Burchiel, K. (1980). Abnormal impuse generation in focally demyelinated trigeminal roots. J Neurosurg 53, 674–683. Caldwell, J. H., Schaller, K. L., Lasher, R. S., Peles, E., and Levinson, S. R. (2000). Sodium channel Nav1.6 is localized at nodes of ranvier, dendrites, and synapses. Proc Natl Acad Sci USA 97, 5616–5620. Ching, W., Zanazzi, G., Levinson, S. R., and Salzer, J. L. (1999). Clustering of neuronal sodium channels requires contact with myelinating Schwann cells. J Neurocytology 28, 295–301. Cragg, B. G., and Thomas, P. K. (1964). Changes in nerve conduction in experimental allergic neuritis. J Neurol Neurosurg Psychiatry 27, 106–115. Chiu, S. Y., and Ritchie, J. M. (1980). Potassium channels in nodal and internodal axonal membrane in mammalian myelinated Wbers. Nature 284, 170–171. Chiu, S. Y., and Ritchie, J. M. (1981). Evidence for the presence of potassium channels in the paranodal region of acutely demyelinated mammalian nerve Wbres. J Physiol Lond 313, 415–437. Craner, M. J., Lo, A. C., Black, J. A., Waxman, S. G. (2003). Abnormal sodium channel distribution in optic nerve axons in a model of inXammatory demyelination. Brain, in press. Cummins, T. R., Aglieco, F., Renganathan, M., Herzog, R. I., Dib-Hajj, S. D., and Waxman, S. G. (2001). Nav1.3 sodium channels: Rapid repriming and slow closed-state inactivation display quantitative diVerences after expression in a mammalian cell line and in spinal sensory neurons. J Neurosci 21, 5952–5961. Cummins, T. R., Renganathan, M., Stys, P. K., Herzog, M. D., Scarfo, B. S., Horn, R., Dib-Hajj, S. D., and Waxman, S. G. (2003). The pentapeptide QYNAD does not block voltage-gated sodium channels. Neurology 60, 224–230. Cummins, T. R., and Waxman, S. G. (1997). Downregulation of tetrodotoxin-resistant sodium currents and upregulation of a rapidly repriming tetrodotoxin-sensitive sodium current in small spinal sensory neurons after nerve injury. J Neurosci 17, 3503–3514. Davie, C. A., Barker, G. J., Webb, S., Toft, P. S., Thompson, A. J., Harding, A. E., McDonald, W. I., and Miller, D. H. (1995) Persistent functional deWcit in multiple sclerosis and autosomal dominant cerebellar ataxia is associated with axon loss. Brain 118, 1583–1592.

OVERVIEW AND PROSPECTS

Davis, F. A., Becker, F. O., Michael, J. A., and Sorensen, E. (1970). EVect of intravenous sodium bicarbonate, diodium edetate, and hyperven tilation on visual and oculomotor signs in multiple sclerosis. J Neurol Neurosurg Psychiatry 33, 723–732. Davis, F. A., and Jacobson, S. (1971). Altered thermal sensitivity of injured cat demyelinated nerve. J Neurol Neurosurg Psychiatry 34, 551–561. Davis, F. A., Stefoski, D., and Rush, J. (1990). Orally administered 4–aminopyridine improves clinical signs in multiple sclerosis. Ann Neurol 27, 186–192. Deerink, T. J., Levinson, S. R., Bennett, G. V., and Ellisman, M. H. (1997). Clustering of voltage-sensitive sodium channels on axons is independent of direct Schwann cell contact in the dystrophic mouse. J Neurosci 17, 5080–5088. De Waegh, S. M., Lee, V. M., and Brady, S. T. (1992). Local modulation of neuroWlament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68, 451–463. Dib-Hajj, S., Black, J. A., Felts, P., and Waxman, S. G. (1996). Down-regulation of transcripts for Na channel alpha-SNS in spinal sensory neurons following axotomy. Proc Natl Acad Sci USA 93, 14950–14954. Dib-Hajj, S. D., Fjell, J., Cummins, T. R., Zheng, Z., Fried, K., LaMotte, R., Black, J. A., and Waxman, S. G. (1999). Plasticity of sodium channel expression in DRG neurons in the chronic constriction injury model of neuropathic pain (in process citation). Pain 83, 591–600. Dib-Hajj, S. D., Tyrrell, L., Black, J. A., and Waxman, S. G. (1998). NaN, a novel voltage-gated Na channel, is expressed preferentially in peripheral sensory neurons and down-regulated after axotomy. Proc Natl Acad Sci USA 95, 8963–8968. Dugandzija-Novakovic, S., Koszowski, A. G., Levinson, S. R., and Shrager, P. (1995). Clustering of Naþ channels and node of Ranvier formation in remyelinating axons. J Neurosci 15, 492–503. Duncan, I. D., J. P., H., and Gilmore, S. A. (1988). Schwann cell myelination of the myelin deWcient rat spinal cord following X-irradiation. Glia 1, 233–239. Eng, D. L., Gordon, T. R., Kocsis, J. D., and Waxman, S. G. (1988). Development of 4-AP and TEA sensitivities in mammalian myelinated nerve Wbers. J Neurophysiol 60, 2168–2179. England, J. D., Gamboni, F., and Levinson, S. R. (1991). Increased numbers of sodium channels form along demyelinated axons. Brain Res 548, 334–337. England, J. D., Gamboni, F., Levinson, S. R., and Finger, T. E. (1990). Changed distribution of sodium channels along demyelinated axons. Proc Natl Acad Sci USA 87, 6777–6780. Felts, P., Baker, T. A., and Smith, K. J. (1977). Conduction in segmentally demyelinated mammalian central axons. J Neurosci 17, 7267–7277. Felts, P., and Smith, K. J. (1991). Conduction properties of central nerve Wbers remyelinated by Schwann cells. Brain Res 574, 178–192. Ferguson, B., Matyszak, M. K., Esiri, M. M., et al. (1997). Axonal damage in acute multiple sclerosis lesions. Brain 120, 393–399. Fern, R., Ransom, B. R., Stys, P. K., and Waxman, S. G. (1993). Pharmacological protection of CNS white matter during anoxia: Actions of phehytoin, carbamazepine and diazepam. J Pharmacol Exper Ther 266, 1549–1555. Foster, R. E., Connors, B. W., and Waxman, S. G. (1982). Rat optic nerve: Electrophysiological, pharmacological, and anatomical studies during development. Dev Brain Res 3, 361–376. Foster, R. E., Whalen, C. C., and Waxman, S. G. (1980). Reorganization of the axonal membrane of demyelinated nerve Wbers: Morphological evidence. Science 210, 661–663. Funch, P. G., and Faber, D. S. (1984). Measurement of myelin sheath resistances: Implications for axonal conduction and pathophysiology. Science 225, 538–540. Gordon, T. R., Kocsis, J. D., and Waxman, S. G. (1990). Electrogenic pump (Naþ/Kþ-ATPase) activity in rat optic nerve. Neuroscience 37, 829–837. Gout, O., Gansmuller, A., Baumann, N., and Gumpel, M. (1988). Remyelination by transplanted oligodendrocytes of a demyelinated lesion in the spinal cord of the adult shiverer mouse. Neurosci Lett 87, 195–199. Gray, P. T., and Ritchie, J. M. (1985). Ion channels in Schwann and glial cells. Trends Neurosci 8, 411–415. Groves, A. K., Barnett, S. C., Franklin, R. J. M, et al. (1993). Repair of demyelinated lesions by transplantation of puriWed 0–2A progenitor cells. Nature 363, 453–456. Guthrie, T. C. (1951). Visual and motor changes in patients with multiple sclerosis. Arch Neurol Psychiatry 65, 437. Halliday, A. M., McDonald, W. I., and Mushin, J. (1973). The visual evoked response in the diagnosis of multiple sclerosis. Brit Med Journal 4, 661–664. Hildebrand, C. (1971). Ultrastructural and light-microscopic studies of the developing feline spinal cord white matter. I. The nodes of Ranvier. Acta Physiol Scand Suppl 364, 81–101. Hines, M., and Shrager, P. (1991). A computational test of the requirements for conduction in demyelinated axons. Restorative Neurol Neurosci 3, 81–93. Hirano, A. (1968). A conWrmation of the oligodendroglial origin of myelin in the adult rat. J Cell Biol 38, 637–340. Honmou, O., Felts, P. A., Waxman, S. G., and Kocsis, J. D. (1996). Restoration of normal conduction properties in demyelinated spinal cord axons in the adult rat by transplantation of exogenous Schwann cells. J Neurosci 16, 3199–3208. Hume, A. L., and Waxman, S. G. (1988). Evoked potentials in suspected multiple sclerosis: Diagnostic value and prediction of clinical course. J Neurol Sci 83, 191–210.

137

138

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

Huxley, A. F., and Sta¨mpXi, R. (1949). Evidence for saltatory conduction in peripheral myelinated nerve Wbres. J Physiol (London) 108, 315–339. Imaizumi, T., Kocsis, J. D., and Waxman, S. G. (1998). Resistance to anoxic injury in rat spinal cord following demyelination. Brain Res. 779: 292–296. Imaizumi, T., Lankford, K. L., Waxman, S. G., Green, C. A., and Kocsis, J. D. (1998). Transplanted olfactory ensheathing cells remyelinate and enhance axonal conduction in the demyelinated dorsal columns of the rat spinal cord. J Neurosci 18, 6176–6185. Kaji, R., Happel, L., and Sumner, A. J. (1990). EVect of digitalis on clinical symptoms and conduction variables in patients with multiple sclerosis. Ann Neurol 28, 582–584. Kaji, R., and Sumner, A. (1989a). EVects of digitalis on CNS demyelinative conduction block. Ann Neurol 25, 159–166. Kaji, R., and Sumner, A. J. (1989b). Ouabain reverses conduction disturbances in single demyelinated nerve Wbers. Neurology 39, 1364–1368. Kaplan, M. R., Cho, M-H., Ullian, E. M., Isom, L. L., Levinson, S. R., and Barres, B. A. (2001). DiVerential control of clustering of the sodium channels Nav1.2 and Nav1.6 at developing CNS nodes of Ranvier. Neuron 30, 105–119. Kocsis, J. D., Waxman, S. G., Hildebrand, C., and Ruiz, J. A. (1982). Regenerating mammalian nerve Wbres: Changes in action potential waveform and Wring characteristics following blockage of potassium conductance. Proc R Soc Lond B 217, 277–287. Koles, Z. J., and Rasminsky, M. (1972). A computer simulation of conduction in demyelinated nerve Wbres. J Physiol Lond 227, 351–364. Lehmann, H. J., and Ule, G. (1964). Electrophysiological Wndings and structural changes in circumscript inXammation of peripheral nerves. Prog Brain Res 6, 169–173. Lo, A. C., Black, J. A., and Waxman, S. G. (2002). Neuroprotection of axons with phenytoin in experimental allergic encephalomyelitis. NeuroReport 13, 1909–1912. Lombet, A., Laduron, P., Mourre, C., Jacomet, Y., and Lazdunski, M. (1985). Axonal transport of the voltage–dependent Na channel protein identiWed by its tetrodotoxin–binding site in rat sciatic nerves. Brain Res 345, 153–158. Maurice, N., Tkatch, T., Meisler, M., Sprunger, L. K., and Sumeier, D. J. (2001). D1/D5 Dopamine receptor activation diVerentially modulates rapidly inactivating and persistent sodium currents in prefrontal cortex pyramidal neurons. J. Neurosci 21(7): 2268–2277. McDonald, W. I. (1963). The eVects of experimental demyelination on conduction in peripheral nerve: A histological and electrophysiological study. Electrophysiological observations. Brain 86, 501–524. McDonald, W. I., and Sears, T. A. (1970). The eVects of experimental demyelination on conduction in the central nervous system. Brain 93, 583–598. Moll, C., Mourre, C., Lazdunski, M., and Ulrich, J. (1991). Increase of sodium channels in demyelinated lesions of multiple sclerosis. Brain Res 556, 311–316. Moore, J. W., Joyner, R. W., Brill, M. H., Waxman, S. G., and Najar-Joa, M. (1978). Simulations of conduction in uniform myelinated Wbres: Relative sensitivity to changes in nodal and internodal parameters. Biophys J 21, 147–161. Namerow, N. S. (1968a). Circadian temperature rhythm and vision in multiple sclerosis. Neurology 18, 417–422. Namerow, N. S. (1968b). Somatosensory evoked responses in multiple sclerosis patients with varying sensory loss. Neurology 18, 1197–1204. Nowak, L., Ascher, P., and Berwald-Netter, Y. (1987). Ionic channels in mouse astrocytes in culture. J Neurosci 7, 101–109. Oh, Y., Black, J. A., and Waxman, S. G. (1994). The expression of rat brain voltage-sensitive Naþ channel mRNAs in astrocytes. Brain Research Molecular Brain Res 23, 57–65. Oh, Y., and Waxman, S. G. (1994). The beta 1 subunit mRNA of the rat brain Naþ channel is expressed in glial cells. Proc Natl Acad Sci USA 91, 9985–9989. Prineas, J., and Connell, F. (1978). The Wne structure of chronically active multiple sclerosis plaques. Neurology 28, 68–75. Raine, C. S. (1984). On the association between perinodal astrocyte processes and the node of Ranvier in the CNS. J Neurocytol 13, 21–27. Raman, I. M., Sprunger, L. K., Meisler, M. H., and Bean, B. P. (1997). Altered subthershold sodium currents and disrupted Wring patterns in Purkinje neurons of Scn8a mutant mice. Neuron 19:881–891. Rasband, M., Trimmer, J. S., Schwarz, T. L., Levinson, S. R., Ellisman, M. H., Schachner, M., and Shrager, P. (1998). Potassium channel distribution, clustering, and function in remyelinating rat axons. J. Neurosci 18, 36–47. Rasband, M. N., Peles, E., Trimmer, J. S., Levinson, S. R., Felux, S., and Schrager, P. (1999). Dependence of nodal sodium channel clustering on paranodal axoglial contact in the developing CNS. J Neurosci 19, 7516–7528. Rasminsky, M. (1973). The eVects of temperature on conduction in demyelinated single nerve Wbers. Arch Neurol 28, 287–292. Rasminsky, M. (1981). Hyperexcitability of pathologically myelinated axons and positive symptoms in multiple sclerosis. In ‘‘Basic and Clinical Electrophysiology’’ (S. G. Waxman and J. M. Ritchie, eds.), pp. 289–297. Raven Press, New York.

OVERVIEW AND PROSPECTS

Rasminsky, M., and Sears, T. A. (1972). Internodal conduction in undissected demyelinated nerve Wbers. J Physiology (London) 227, 323–350. Ransom, B. R., Stys, P. K., and Waxman, S. G. (1990). Anoxic injury of CNS white matter: Protective eVect of ketamine. Neurology 40, 1399–1404. Renganathan, M., Cummins, T. R., and Waxman, S. G. (2001). Contribution of Nav1.8 Sodium Channels to Action Potential Electrogenesis in DRG Neurons. J Neurophysiol 86, 629–640. Renganathan, M., Gelderblom, M., Black, J. A., and Waxman, S. G. (2003). Expression of Nav 1.8 sodium channels perturbs the Wring patterns of cerebellar Purkinje cells. Brain Res 959, 235–242. Ritchie, J. M. (1982). Sodium and potassium channels in regenerating and developing mammalian myelinated nerves. Proc Roy Soc B 215, 273–287. Ritchie, J. M., Rang, H. P., and Pellegrino, R. (1981). Sodium and potassium channels in demyelinated and remyelinated mammalian nerve. Nature 294, 257–259. Ritchie, J. M., and Rogart, R. B. (1977). The density of sodium channels in mammalian myelinated nerve Wbers and the nature of the axonal membrane under the myelin sheath. Proc Natl Acad Sci USA 74, 211–215. Ritchie, J. M., Rogart, R. B., and Strichartz, G. (1976). A new method for labeling saxitoxin and its binding to non–myelinated Wbres of the rabbit vagus, lobster walking, and garWsh olfactory nerves. J Physiol (Lond) 261, 477–494. Ro¨per, J., and Schwarz, J. R. (1989). Heterogeneous distribution of fast and slow potassium channels in myelinated rat nerve Wbers. J Physiol (Lond) 416, 93–110. Rosenbluth, J. (1979). Aberrant axon-Schwann cell junctions in dystrophic mouse nerves. J Neurocytol 8, 655–672. Rosenbluth, J. (1985). Intramembranous particle patches in myelin-deWcient rat mutant. Neurosci Letts 62, 19–24. Rosenbluth, J., and Blakemore, W. F. (1984). Structural specializations in cat of chronically demyelinated spinal cord axons as seen in freeze-fracture replicas. Neurosci Letts 48, 171–177. Rosenbluth, J., Hasegawa, M., and SchiV, R. (1989). Myelin formation in myelin-deWcient rat spinal cord following transplantation of normal fetal spinal cord. Neurosci Lett 97, 35–40. Schauf, C. L., and Davis, F. A. (1974). Impulse conduction in multiple sclerosis: A theoretical basis for modiWcation by temperature and pharmacological agents. J Neurol Neurosurg Psychiat 37, 152–161. Sears, T. A., and Bostock, H. (1981). Conduction failure in demyelination: Is it inevitable? In ‘‘Demyelinating Diseases: Basic and Clinical Electrophysiology’’ (S. G. Waxman and J. M. Ritchie, eds.), pp. 357–375. Raven Press, New York. Shrager, P. (1989). Sodium channels in single demyelinated mammalian axons. Brain Res 483, 149–154. Sleeper, A. A., Cummins, T. R., Dib-Hajj, S. D., Hormuzdiar, W., Tyrrell, L., Waxman, S. G., and Black, J. A. (2000). Changes in expression of two tetrodotoxin-resistant sodium channels and their currents in dorsal root ganglion neurons after sciatic nerve injury but not rhizotomy. J Neurosci 20, 7279–7289. Smith, K. J., Blakemore, W. F., and McDonald, W. I. (1981). The restoration of conduction by central remyelination. Brain 104, 383–404. Smith, K. J., Blakemore, W. F., and McDonanld, W. I. (1983). Central remyelination restores secure conduction. Nature 280, 395–396. Smith, K. J., Bostock, H., and Hall, S. M. (1982). Saltatory conduction precedes remyelination in axons demyelinated with lysophosphatidyl choline. J Neurol Sci 54, 13–31. Smith, K. J., and Hall, S. M. (1980). Nerve conduction during peripheral demyelination and remyelination. J Neurol Sci 104, 383–404. Smith, K. J., and McDonald, W. I. (1980). Spontaneous and mechanically evoked activity due to central demyelinating lesions. Nature 286, 154–155. Smith, M. E., Kocsis, J. D., and Waxman, S. G. (1983). Myelin protein metabolism in demyelination and remyelination in the sciatic nerve. Brain Res 270, 37–44. Smith, M. R., Smith, R. D., Plummer, N. W., Meisler, M. H., and Goldin, A. L. (1998). Functional analysis of mouse Scn8a sodium channel. J. Neurosci. 18: 6093–6102. Sontheimer, H., Black, J. A., Ransom, B. R., and Waxman, S. G. (1992). Ion channels in spinal cord astrocytes in vitro. I. Transient expression of high levels of Naþ and Kþ channels. J Neurophysiol 68, 985–1000. Sontheimer, H., Fernandez-Marques, E., Ullrich, N., Pappas, C. A., and Waxman, S. G. (1994). Astrocyte Naþ channels are required for maintenance of Naþ/K(þ)-ATPase activity. J Neurosci 14, 2464–2475. Sontheimer, H., Ransom, B. R., Cornell-Bell, A. H., Black, J. A., and Waxman, S. G. (1991). Naþ-current expression in rat hippocampal astrocytes in vitro: Alterations during development. J Neurophysiol 65, 3–19. Stanley, G. P., and Pender, M. P. (1991). Pathophysiology of chronic relapsing experimental allergic encephalomyelitis. Brain 114, 1827–1853. Stefoski, D., Davis, F. A., Faut, M., and Schauf, C. L. (1987). 4-Aminopyridine improves clinical signs in multiple sclerosis. Ann Neurol 21, 71–77. Strichartz, G. R., Small, R. K., and Pfenninger, K. H. (1984). Components of the plasma membrane of growing axons. III. Saxitoxin binding to sodium channels. J Cell Biol 98, 1444–1452. Stys, P. K., and Lesiuk, H. (1996). Correlation between electrophysiological eVects of mexiletine and ischemic protection in CNS white matter. J Neurosci 71, 27–36. Stys, P. K., Ransom, B. R., and Waxman, S. G. (1992b). Tertiary and quaternary local anesthetics protect CNS white matter from anoxic injury at concentrations that do not block excitability. J Neurophysiol 67, 236–240.

139

140

5. ELECTROPHYSIOLOGIC CONSEQUENCES OF MYELINATION

Stys, P. K., Waxman, S. G., and Ransom, B. R. (1992a). Ionic mechanisms of anoxic injury in mammalian CNS white matter: Role of Naþ channels and Naþ-Ca2þ exchanger. J Neurosci 12, 430–439. Swadlow, H. A., Waxman, S. G., and Weyand, T. G. (1981). EVects of variations in temperature on impulse conduction along nonmyelinated axons in the mammalian brain. Exper Neurol 71, 383–389. Tang, C. M., Strichartz, G. R., and Orkand, R. K. (1979). Sodium channels in axons and glial cells of the optic nerve of Necturus maculosa. J Gen Physiol 74, 629–642. Targ, E. F., and Kocsis, J. D. (1985). 4-Aminopyridine leads to restoration of conduction in demyelinated rat sciatic nerve. Brain Res 328, 358–361. Targ, E. F., and Kocsis, J. D. (1986). Action potential characteristics of demyelinated rat sciatic nerve following application of 4-aminopyridine. Brain Res 363, 1–9. Tasaki, I. (1959). Conduction of the nerve impulse. In ‘‘American Physiological Society. Handbook of Physiology.’’ Section 1, Neurophysiology, Volume 1, J. Field, H. W. Magoun, and V. E. Hall (Eds), Am Physiol Soc, pp. 75–121. Trapp, B. D., Peterson, J., RansohoV, R. M., et al. (1998). Axonal transection in the lesions of multiple sclerosis. New Engl J Med 338, 278–285. Ulrich, J., and Groebke–Lorenz, W. (1983). The optic nerve in multiple sclerosis: A morphological study with retrospective clinicopathological correlation. Neurol Ophthalmol 3, 149–159. Utzschneider, D. A., Archer, D. R., Kocsis, J. D., Waxman, S. G., and Duncan, I. D. (1994). Transplantation of glial cells enhances action potential conduction of amyelinated spinal cord axons in the myelin-deWcient rat. Proc Natl Acad Sci USA 91, 53–57. Vabnick, I., and Shrager, P. (1998). Ion channel redistribution and function during development of the myelinated axon. J Neurobiol 37, 80–96. Vollmer, T. L., Brass, L. M., and Waxman, S. G. (1991). Lhermitte’s sign in a patient with herpes zoster. J Neurol Sci 106, 153–157. Wang, H., Kunkel, D. D., Martin, T. M., Schwartkroin, P. A., and Tempel, B. L. (1993). Heteromultimeric Kþ channels in terminal and juxtaparanodal regions of neurons. Nature 365, 75–79. Watson, C. W. (1959). EVects of lowering body temperature on the symptoms and signs of multiple sclerosis. New Engl J Med 261, 1252–1259. Waxman, S. G. (1977). Conduction in myelinated, unmyelinated, and demyelinated Wbers. Arch Neurol 34, 585–590. Waxman, S. G. (1978). Prerequisites for conduction in demyelinated Wbers. Neurology, 28, 27–34. Waxman, S. G. (1982). Current concepts in neurology: membranes, myelin and the pathophysiology of multiple sclerosis. New Engl J Med 306, 1529–1533. Waxman, S. G. (1987). Molecular organization of the cell membrane in normal and pathological axons: relation to glial contact. In ‘‘Glial-Neuronal Communication in Development and Regeneration’’ (H. Althaus and W. Seifert, eds.), pp. 711–736. Springer-Verlag, Heidelberg, Germany. Waxman, S. G. (1992). The perinodal astrocyte: Functional and developmental considerations. In ‘‘Biology and Pathobiology of Astrocyte-Neuron Interactions’’ (S. FedoroV, R. Doucette, and B. H. Juurlink, eds.). Plenum, New York. Waxman, S. G. (1998). Demyelinating disease: New pathological insights, new therapeutic targets. New Engl J Med 338, 323–325. Waxman, S. G. (1999). Multiple sclerosis as a neuronal disease. Arch Neurol. 57: 22–24. Waxman, S. G. (2001). Transcriptional channelopathies: an emerging class of disorders. Nat Rev Neurosci 2, 652–659. Waxman, S. G., and Bennett, M. V. L. (1972). Relative conduction velocities of small myelinated and nonmyelinated Wbers in the central nervous system. Nature New Biology 238, 217–219. Waxman, S. G., and Black, J. A. (1984). Freeze-fracture ultrastructure of the perinodal astrocyte and associated glial junctions. Brain Res 308, 77–87. Waxman, S. G., Black, J. A., Kocsis, J. A., and Ritchie, J. M. (1989). Low density of sodium channels supports action potential conduction in axons of neonatal rat optic nerve. Proc Natl Acad Sci 86, 1406–1410. Waxman, S. G., and Brill, M. H. (1978). Conduction through demyelinated plaques in multiple sclerosis: computer simulations of facilitation by short internodes. J Neurol Neurosurg Psychiat 41, 408–417. Waxman, S. G., and Geschwind, N. (1983). Major morbidity related to hyperthermia in multiple sclerosis. Ann Neurol 13, 348. Waxman, S. G., Kocsis, J. D., and Black, J. A. (1994). Type III sodium channel mRNA is expressed in embryonic but not adult spinal sensory neurons, and is reexpressed following axotomy. J Neurophysical 72, 466–470. Waxman, S. G., and Sims, T. J. (1984). SpeciWcity in central myelination: evidence for local regulation of myelin thickness. Brain Res 292, 179–185. Waxman, S. G., Sims, T. J., and Gilmore, S. A. (1988). Cytoplasmic membrane elaborations in oligodendrocytes during myelination of spinal motoneuron axons. Glia 1, 286–291. Waxman, S. G., Sontheimer, H., Black, J. A., Minturn, J. E., and Ransom, B. R. (1992). Dynamic aspects of sodium channel expression in astrocytes. In ‘‘Neural Regeneration’’ (F. J. Seil, ed.). Raven Press, New York. Waxman, S. G., Utzschneider, D. A., and Kocsis, J. D. (1994). Enhancement of action potential conduction following demyelination: Experimental approaches to restoration of function in multiple sclerosis and spinal cord injury. Progress in Brain Res 100, 233–244.

OVERVIEW AND PROSPECTS

Waxman, S. G., and Wood, S. L. (1984). Impulse conduction in inhomogeneous axons: EVects of variation in voltage-sensitive ionic conductances on invasion of demyelinated axon segments and preterminal Wbers. Brain Res 294, 111–122. Weiner, L. P., Waxman, S. G., Stohlman, S. A., and Kwan, A. (1980). Remyelination following viral-induced demyelination: Ferric ion-ferrocyanide staining of nodes of Ranvier within the CNS. Ann. Neurol., 8, 580–583. Westenbroek, R. E., Noebels, J. L., and Catterall, W. A. (1992). Elevated expression of type II Naþ channels in hypomyelinated axons of shiverer mouse brain. J Neurosci 12, 2259–2267. Wisniewski, H. M., Oppenheimer, D., and McDonald, W. E. (1976). Relation between myelination and function in MS and EAE. J Neuropathol Exp Neurol 35, 327. Wood, J. G., Dou, H. J., Whitaker, J. N., McLaughlin, B. J., and Albers, R. W. (1977). Immunocytochemical localization of the sodium, potassium activated ATPase in knifeWsh brain. J Neurocytology 6, 571–581. Yokota, T., Saito, Y., and Miyatake, T. (1994). Conduction slowing without conduction block of compound muscle and nerve action potentials due to sodium channel block. J Neurol Sci 124, 220–224. Zhou, W., and Goldin, A. L. (2002). Functional diVerences between the Nav1.6 and Nav1.2 sodium channels. Soc Neurosci Abstr, 834–844. Zhou, L., and Zhang, C. L., Messing, A., Chiu, S. Y. (1998). Temperature-sensitive neuromuscular transmission in Kv1.1 null mice: role of potassium channels under the myelin sheath in young nerves. J Neurosci 18, 7200–7215.

141

C H A P T E R

6 Remyelination through Engraftment A. Baron-Van Evercooren and W. F. Blakemore

HISTORY The Wrst experiments showing that it was possible to achieve remyelination of CNS axons following the transplantation of myelin-forming cells were undertaken by investigators interested in axon regeneration. Thus, Santiago Ramon y Cajal and his pupil Francisco Tello demonstrated that CNS axons could enter segments of peripheral nerve implanted into the CNS (Cajal, 1928; Le Gros Clark, 1943; Tello, 1911). Once the CNS axons entered the implanted nerve, they became myelinated by the transplanted Schwann cells. In 1977 Blakemore showed that demyelinated CNS axons became myelinated by Schwann cells after segments of peripheral nerve were placed over areas of demyelination in the spinal cord (Blakemore, 1977), and Duncan and colleagues demonstrated that cultured Schwann cells could myelinate demyelinated CNS axons in animals with defective myelination (Duncan et al., 1981). Madeleine Gumpel and colleagues then, for the Wrst time, demonstrated myelination by CNS cells by implanting fragments of embryonic CNS into the developing CNS of mice with defective myelination resulting from an absence of MBP expression (Gumpel et al., 1983). The demonstration that transplant-mediated oligodendrocyte remyelination could be achieved in situations where endogenous myelination or remyelination was inhibited quickly followed (Blakemore and Crang, 1988; Duncan et al., 1988b). These papers in the 1980s established the basic experimental models that were to dominate all subsequent studies examining transplant-mediated myelination. These early workers appreciated the therapeutic potential of glial cell transplantation for diseases characterized by a failure of remyelination, such as multiple sclerosis and the leucodystrophies, and this provided the impetus for the subject being considered in this chapter to a point where clinical applications are beginning to be attempted. Although much of the work to be reported in this chapter is driven by examining and evaluating the therapeutic potential of achieving remyelination by transplantation of myelin-forming cells, these studies also illustrate the power of glial cell transplantation for examining aspects of glial cell biology during development and in pathological situations. Remyelination by engraftment has been the subject of a number of reviews (Baron-Van Evercooren, 1994; BaronVan Evercooren et al., 1997; Blakemore et al., 1990a, 1990b, 1995a, 1995c; Blakemore and Crang, 1992; Blakemore and Franklin, 1991, 2000; Brustle, 1998; Bulte et al., 2002; Chari and Blakemore, 2002b; Decker et al., 2001; Duncan, 1996; Duncan and Archer, 1994; Duncan and Milward, 1995; Duncan et al., 1997; Franklin, 2002a; Franklin and Blakemore, 1990, 1993, 1995, 1997, 1998, 1999; Gumpel et al., 1985, 1988; Kocsis, 1999; Lachapelle, 1995; Lachapelle et al., 1989, 1994b; Liblau et al., 2001; Nait-Oumesmar et al., 1998; Smith and Blakemore, 2000b; Vescovi and Snyder, 1999; Zhang and Duncan, 2000).

Myelin Biology and Disorders, Volume 1

143

Copyright 2004, Elsevier (USA). All rights reserved.

144

6. REMYELINATION THROUGH ENGRAFTMENT

THE PRINCIPLES OF GLIAL CELL TRANSPLANTATION EXPERIMENTS The normal glial populations of the CNS and PNS have an intrinsic replacement capacity; therefore, two major issues must be addressed when undertaking and interpreting glial cell transplantation studies. First, there is the question of identifying the transplanted cells and, second, there is the question of having a clear knowledge of the nature of the cells being transplanted. The Wrst issue has led to the development of various experimental models and the use of cellular markers to track the fate of transplanted cells. The second has led to the development of diVerent cell isolation and expansion procedures that result in enrichment for cells of a particular phenotype. It must also be appreciated that although one may document results of glial cell transplantation in terms of presence of myelin-forming oligodendrocytes or Schwann cells; it is likely that in the case of oligodendrocytes, and in certain instances with Schwann cells that one is examining the consequences of transplanting precursors rather than the diVerentiated cells. In interpreting transplantation experiments it also has to be appreciated that during development and within pathological tissue there is an opportunity for introduced cells to integrate into tissue that is generating new cells. However, in normal mature tissue, cell numbers are stable and strictly controlled. Therefore the transplanted cells will be supernumerary and under such circumstances they will be exposed to homeostatic mechanisms that regulate cell numbers. In the short term, some cells may survive when transplanted into normal tissue, perhaps because the injection procedure and the introduction of cells induces a reactive environment that can support supernumerary cells, or cell numbers may be suYcient to sustain survival by autrocrine/paracrine signaling. However, any claim of long-term survival and especially migration of cells from the point of implantation in adult animals has to be viewed with considerable caution, since it indicates that either cells have been incorporated into some intrinsic regenerative system (such as the rostral migratory stream), or the transplantation procedure has created an abnormal environment that allows supernumerary cells to survive, or the introduced cells are abnormal (i.e., the cells have properties associated with neoplasia), or some intrinsic property of the cells allows them to escape the normal homeostatic mechanisms that control cell numbers in adult tissue (as appears to be the case with cells derived from embryos). Early reports stressed mainly the anatomical integration potential of the transplanted cells and to date there is little if any functional data on the potential beneWts of transplantmediated myelination. This is mainly a consequence of the nature of the animal models used. It has to be noted that data obtained with species other than primates may also need to be interpreted with caution in view of the many diVerences between species in terms of the biology of the transplanted cells as well as the size and anatomical organization of the CNS environment in which they are placed. It is therefore important that therapeutic applications should not be considered without the demonstration of clinical recovery using models that mimic potential clinical situations where this procedure may be undertaken.

TRACKING THE FATE OF TRANSPLANTED CELLS A major issue that must be addressed when carrying out and interpreting glial cell transplantation studies is the identiWcation of the transplanted cells. Initial oligodendrocyte transplantation experiments looked for the diVerentiation of transplanted cells by introducing cells from normal animals into the so-called myelin mutants. These animals include the MBP-deWcient Shiverer mouse (Friedman et al., 1986; Gumpel et al., 1983, 1989; Kohsaka et al., 1986), the various PLP mutants: the myelin-deWcient rat (Duncan et al., 1988b; Rosenbluth et al., 1990), the Jimpy mouse (Lachapelle et al., 1991), and the shaking pup (Archer et al., 1997). All the PLP mutants show such severe hypomyelination that any myelin formation observed following transplantation can be related to transplanted cells. Immunostaining for PLP can be used if

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

TRACKING THE FATE OF TRANSPLANTED CELLS

necessary to distinguish host PLP-deWcient myelin from the PLP-positive myelin produced by transplanted cells (Duncan et al., 1988b). MBP positive normal myelin is easily recognized in the Shiverer mouse either by immunocytochemistry (Lachapelle et al., 1984) (Fig. 6.1C) or by electron microscopy (Gansmuller et al., 1986). The transgenic approach has and will result in further hypomyelinated animals becoming available that can be used as transplant recipients. An example is the MAG and Fyn knockout mouse (Ader et al., 2001). While these models oVer the possibility to clearly assay the Wnal product of transplanted myelin-forming cells, they do not allow one to track the fate and distribution of all the cells introduced into the host environment. Various strategies have been used to distinguish transplanted cells from host cells and nearly all impose limitations on the type of studies that can be undertaken. Transplanted cells can be labeled in vitro prior to transplantation (exogenous labeling) using vital dyes such as fast-blue (Espinosa de los Monteros et al., 1993) and bisbenzimide (Hoechst dyes 33258 and 33342) (Gansmuller et al., 1991) for subsequent identiWcation using Xuorescence microscopy, or carbocyanine, which after photoactivation can also be visualized by electron microscopy (Baron-Van Evercooren et al., 1996). Recent advances have been made in labeling cells with superparamagnitic iron

FIGURE 6.1 The shiverer model. (A) PSA-NCAM immunoselected neonatal mouse precursors are expanded as spheres grown in bFGF/EGF medium; (B) bFGF/FGF expanded spheres express PSA-NCAM (green) and start to express the oligodendrocyte progenitor marker GD3 (red) when primed for 6 days in N1B104; (C) oligospheres derived from the transgenic MBP Lac Z mice express the transgene (b-Gal, red) and MBP (green) 30 days after transplantation in the corpus callosum of the neonate shiverer brain; (D) higher magniWcation allows to visualize the b-Galþ oligodendrocyte cell body and processes which are surrounded with MBPþ myelin internodes; (E) the myelin domain of b-Galþ oligodendrocytes are visualized in toluidine blue counterstained semithin sections; (F) the b-Galþ deposits and compaction of myelin in ultrathin sections conWrm the donor origin of the myelin sheath.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

145

146

6. REMYELINATION THROUGH ENGRAFTMENT

oxide, which allows cells to be detected by a noninvasive procedure such as MRI (Bulte et al., 1999, 2001, 2002; Franklin et al., 1999). With further development, these novel techniques hold the prospect of tracing cells longitudinally in the same animal and oVer the possibility of monitoring the fate of cells following implantation into human subjects. However, these labels, as well as the use of the s-phase markers such as tritiated thymidine and BrdU suVer the disadvantage of dilution by repeated division of cells. There is also the possibility that some dyes, notably the Hoechst dyes 33258 and 33342, can be taken up by host cells even after careful pretransplantation washing of cells (Iwashita et al., 2000a). In addition to direct labeling of cells at the time of implantation, there can also be direct transfer of label to host cells and macrophages when cells die. Thus, exogenous labels should be used with caution and not used to track the fate of individual cells. It is generally accepted that transfer of label is not a problem encountered with tritiated thymidine and BrdU. A strategy that overcomes the dilution and transfer problems associated with in vitro labeling of transplanted cells is the use of genetic markers (endogenous labels). These may take the form of genes such as green Xuorescent protein (GFP) (Herrera et al., 2001; Tamiki et al., 2002; Windrem et al., 2002) and bacterial Lac-Z introduced directly in the cells (Franklin et al., 1995; Groves et al., 1993; Learish et al., 1999; Tontsch et al., 1994), or deriving tissue plugs (Friedrich and Lazzarini, 1993), fragments (Lachapelle et al., 1994a, 2002a), or cells from transgenic animals (Mitome et al., 2001b; Nakatsuji and Miller, 2001; Vitry et al., 1999) (Fig. 6.1). One can also utilize inherent genetic mismatches between the grafted cells and host to identify transplanted cells. Such systems are male to female transplantation utilizing repeat sequences on the Y-chromosome as the marker (Harvey et al., 1992; O’Leary and Blakemore, 1997b). Another option is to use species mismatches for certain antigens (glutathione-S-transferase (GSTp), proliferating cell nuclear antigen (PCNA), human mitochondrial antigen or Alu repeat transcripts when human cells are transplanted in the rodent CNS (Brustle et al., 1998; ReubinoV et al., 2001; Zhang et al., 2001). An alternative strategy to the use of markers to identify transplanted cells, which can be adopted in adult animals, is to make demyelinating lesions and abolish the inherent repair capacity of the tissue into which cells are introduced. This can be achieved by exposing the lesion-containing tissue to 40 Gy of X-irradiation (see the section on models). It is also possible to evaluate the consequences of transplantation simply by documenting fundamental morphological diVerences between transplanted and untransplanted animals, however, without the use of cell identiWcation markers it is impossible to know whether the transplanted cells are contributing directly or indirectly to the observed diVerences.

EXPERIMENTAL MODELS Neonatal Animals and Myelin Mutants The developing nervous system provides an excellent system into which cells can be introduced in order to assess their myelin generating potential since it is generating myelinating cells at this time and therefore the introduced cells will be exposed to signals that are optimal for controlling their diVerentiation into myelin-forming cells. However, it has to be appreciated that cells introduced into the developing nervous system will be in competition with endogenous cells for axons to myelinate and that in terms of therapeutic applications the environment is permissive for myelination and in no way resembles the situation that would be found in chronic disease states. A major advantage of introducing cells into the embryonic or neonatal CNS is that immunosuppression is not required to prevent rejection. The recipients can either be normal animals, in which case cell markers have to be employed to identify transplanted cells, or animals in which a myelin protein fails to be expressed (myelin mutants) allowing myelin formed by transplanted normal cells to be

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

EXPERIMENTAL MODELS

readily identiWed. The normal practice is to inject cells into the CNS of E19 embryos (Brustle et al., 1998) or newborn rat and mouse pups (Gumpel et al., 1983). MyelindeWcient rats are normally injected between 6 to 8 days, the time when clinical symptoms can be Wrst recognized (Archer et al., 1994). Transplantation of adult animals is also possible when the mutants have an extended life span (the shaking pup has been injected up to 32 weeks of age (Archer et al., 1997) and demyelinating lesions can be combined with transplantation in the adult shiverer mouse (Gout et al., 1988). The following paragraphs will describe only those models that have so far been used for transplantation studies. MBP Mutants The shiverer mouse The shiverer mouse played a major role in the development of oligodendrocyte transplantation studies. It was Madeleine Gumpel and her colleagues who Wrst recognized the value of this myelin mutant (Gumpel et al., 1983). The shiverer mouse has a deletion of seven of the nine exons of the myelin basic protein gene. This results in a failure to produce myelin basic protein (MBP) and an absence of the major dense line at the ultrastructural level. Thus, the myelin produced as a consequence of transplantation can be readily demonstrated by immunocytochemistry using antibodies to MBP and by electron microscopy. Shiverer mice have a normal or increased number of oligodendrocytes and processes from these cells associate with axons. Myelin sheaths form around the majority of axons but they are thin and uncompacted. Although these animals have a normal life span, the fact that oligodendrocytes associate with axons means that these animals cannot be used as recipients after they are more than several weeks old unless demyelination is induced (Gout et al., 1988). The Long Evans shaker(les) rat The les rat is another MBP mutant (Kwiecien et al., 1998); in this case, MBP fails to be produced due to disruption of the MBP gene caused by the insertion of a retrotransposon into a noncoding region (intron 3) of the gene (O’Connor et al., 1999). The pathology of this animal is very similar to that of the shiverer mouse. Transplantation of young animals (up to 5 weeks of age) results in some transplantmediated myelination; however, like the shiverer mouse although there is failure to produce normal myelin sheaths, oligodendrocytes processes associate with axons and this may provide an explanation why extensive myelin sheath formation by transplanted cells does not occur in these animals (WFB, personal observations). A problem with these animals that limits their usefulness as transplant recipients is that they suVer from seizures that result in spinal fractures. PLP Mutants The myelin-deWcient (md) rat This mutant used by the laboratories of Duncan (Duncan et al., 1988b) and Rosenbluth (Rosenbluth et al., 1989) has also played a signiWcant role in the development of oligodendrocyte transplantation studies. The md rat is an X-linked mutant that has a point mutation in the PLP gene that results in death of oligodendrocytes as they begin myelinating axons (Boison and StoVel, 1989). This results in an almost complete absence of myelin sheaths and thus any signiWcant myelin formation observed following transplantation can be ascribed to the transplanted cells or their progeny. Moreover, the myelin sheaths that are formed in this mutant are PLP negative and thus antibodies to PLP can be used to speciWcally identify myelin sheaths produced by the transplanted cells. A disadvantage of this mutant is its short life span (approximately 24 to 30 days and thus the long-term consequences of transplantation cannot be explored using these animals. There is, however, a longer-lived strain that survives until 75 to 80 days (Duncan et al., 1995). The jimpy (jp) mouse The jp mouse is another PLP mutant that has a base pair substitution in the fourth intron of the PLP gene. In this mutant, most oligodendrocytes die as they commence myelination, but a few survive, and synthesize small amounts of myelin (Farkas-Bargeton et al., 1972). The myelin sheath formed shows ultrastructural abnormalities such as poor compaction, the absence of a continuous intraperiod line, and

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

147

148

6. REMYELINATION THROUGH ENGRAFTMENT

failure to stain with antibodies to PLP (Nussbaum and Mandel, 1973). If myelin is formed, it has a patchy distribution throughout the entire CNS (Lachapelle et al., 1990). In addition to aVecting myelin and oligodendrocytes, the mutation is associated with a complex syndrome that includes axonal and neuronal abnormalities (Gotow et al., 1999; Rosenfeld and Friedrich, 1983) and astrocytosis (SkoV, 1976). Studies of transplantation in the jp brain are also limited by the life span of these mutants (35 days). The shaking(sh) pup The shaking pup has a point mutation in the PLP gene (Nadon et al., 1990). Although not as severely myelin deWcient as the md rat, these animals have a unique role to play in the evaluation of glial cell transplantation in a clinical context as they can with careful nursing be kept alive for one year or more. This allows transplants to be made into environments that more closely mimic an area of chronic demyelination and allows for long survival times post transplantation. The drawback of this model is the expense, time, and eVort required to rear these animals which has limited the number of experiments undertaken so far. Despite these problems, a number of important results have been obtained (Archer et al., 1997). The Wrst is that transplant-mediated remyelination can be extensive when embryo-derived cells are used in young animals. Another important observation is that transplant mediated remyelination has persisted for up to 28 weeks in the absence of immunosuppression, even though donor and recipients were not inbred (in some cases donors were not even littermates). The data so far obtained would indicate that the extent of myelination achieved is signiWcantly reduced when cells are injected into older recipients. Transgenic animals over-expressing the PLP gene Ikenaka and colleagues designed a transgenic mouse over-expressing four copies of the wild-type PLP gene (Kagawa et al., 1994). Homozygous mice have severe hypomyelination and a short life span similar to jimpy mice. However, heterozygous mice with three copies of the PLP gene have a longer life span. While these mice have a normal myelination process, demyelination occurs after 3 to 4 months. These animals thus constitute a valuable tool to assay transplantation in the spontaneously demyelinated adult CNS (Espinosa de los Monteros et al., 2001). Mag-fyn Knockout Mouse Myelin-associated glycoprotein (MAG) is functionally involved in the initiation of myelination and fyn, a nonreceptor tryrosine kinase of the src family, is a downstream signaling molecule of MAG. The null mouse for MAG and fyn is hypomyelinated. The myelin sheaths that are formed are abnormal and do not contain MAG (BiYger et al., 2000). Thus, myelin produced by transplanted normal cells can be detected using antibodies to MAG (Ader et al., 2001). These animals have a normal life span.

Induced Lesions in Adult Animals A number of procedures have been used to create demyelinating lesions that are amenable to transplantation studies. The use of diVerent toxins and location of the lesions lend themselves to the analysis of diVerent parameters associated with remyelination. Since endogenous remyelination is inXuenced by factors such as the location and the nature of the demyelinating processes, the diVerent lesions will be described in some detail paying particular attention to features that may inXuence analysis of transplantation experiments. It must be appreciated that all demyelinating lesions in experimental animals are small compared to demyelinating lesions in man and nearly all will undergo spontaneous remyelination, therefore unless this spontaneous repair process is abolished the transplanted cells will be in competition with host-derived cells for axons to myelinate. Focal Lesions Induced by Gliotoxic Agents The production of focal areas of demyelination in known locations by injecting chemicals into white matter tracts, which speciWcally destroys glial cells and their myelin without killing axons, allows one either to inject cells directly into the area of demyelination or to

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

EXPERIMENTAL MODELS

place cells at a distance from the area of demyelination to examine whether cells are recruited into the area of demyelination. The demyelinating agents that have been used so far are lysolecithin, ethidium bromide, 6-aminonicotinamide, antibodies to oligodendrocyte related molecules, cholera toxin B-subunit conjugated to saporin. and bacterial endotoxin. The locations used have included the dorsal funiculi of the spinal cord (usually at the thoraco-lumbar level), the caudal cerebellar peduncle, the corpus callosum, the optic nerve, and subcortical white matter. With most toxins, increasing the volume injected, or increasing the concentration, enhances axon degeneration and with some toxins hemorrhage occurs. It should be noted that with all toxins, axon degeneration in the corpus callosum can be extensive due to the small size of the axons. Lysolecithin (lysophosphatidyl choline) This membrane solubilizing agent was Wrst used as a demyelinating agent by Hall (1972) and has been used in numerous studies in mouse, rat, rabbit, and cat. It is normally used as a 1% solution made up in saline. At this strength, 1ml injected into a dorsal funiculus of the spinal cord of a rat, produces an ellipsoid shaped area of demyelination which extends over 3 to 8 mm (depending on the speed of injection and bore of the injecting needle) and involves approximately 20 to 35% of the cross-sectional area of the dorsal funiculus. It has a particular toxicity for myelin but spares some oligodendrocytes and possibly their progenitors within the area of demyelination. There is some loss of axons and astrocytes around the point of injection. The majority of the demyelinated axons are remyelinated by oligodendrocytes. However, some axons, particularly those in the center of the area of demyelination are remyelinated by Schwann cells. If lesions are large, or there is excessive axon degeneration, Schwann cell remyelination can be extensive. In young animals, lysolecithin lesions in the spinal cord of all species remyelinate rapidly, however, remyelination in old animals occurs more slowly (Gilson and Blakemore, 1993). Following lysolecithin-induced demyelination in the rabbit, remyelination is incomplete even after 6 months (Blakemore, 1978). In the caudal cerebellar peduncle, remyelination is faster after demyelination with lysolecithin than with ethidium bromide or Gal-C antibodies (WoodruV and Franklin, 1999). When injected into the optic nerve of monkeys, remyelination is poor in contrast to lesions made in the spinal cord (Lachapelle et al., 2000). Lysolecithin lesions have been induced in the shiverer mouse spinal cord to examine remyelination following cell transplanted into and at a distance from the area of demyelination (Gout et al., 1988). Ethidium bromide Ethidium bromide is used either as a 0.05 or 0.1% solution in saline and 1ml injected into the dorsal funiculus of the rat spinal cord results in much larger area of demyelination than can be induced with lysolecithin. Lesions can involve almost all of the dorsal funiculi and extend 4 to 8 mm longitudinally. This DNA intercalating agent kills both astrocytes and oligodendrocytes (Blakemore, 1982) as well as oligodendrocyte progenitors (Sim et al., 2002) throughout the area of demyelination. In young animals (under 3 months), remyelination is rapid and undertaken mainly by oligodendrocytes (60 to 80% of demyelinated axons), while in old animals (> 5 months), remyelination can be slow and the percentage of axons remyelinated by Schwann cells is increased (up to 80%) (Fig. 6.2). Ethidium bromide has also been used to produce demyelinating lesions in the caudal cerebellar peduncles of rats where the lesions have a cross-sectional area of 0.6 mm2 when 4ml of 0.01% ethidium bromide is injected (WoodruV and Franklin, 1999). The extent of Schwann cell remyelination in lesions made at this site is less than in spinal cord lesions. Ethidium bromide lesions made in the cervical spinal cord of rats have been used for functional studies (JeVery and Blakemore, 1997). The toxin has also been used to produce demyelinating lesions in the spinal cords of cats (Blakemore, 1982) and pigs (WFB unpublished observations), and in both species the predominant remyelinating cells are Schwann cells. Anti-galactocerebroside antibodies and complement Galactocerebroside is a major sphingolipid constituent of myelin and antibodies to these molecules are used to identify oligodendrocytes in tissue culture. Large volumes (4ml) are required to induce

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

149

150

6. REMYELINATION THROUGH ENGRAFTMENT

FIGURE 6.2 Injections of ethidium bromide into white matter results in death of oligodendrocytes, astrocytes and oligodendrocyte progenitor cells and when injected into tissue that has been exposed to 40Gy of X-irradiation, endogenous remyelination is totally abolished provided a long length (usually 4 cm) of spinal cord in exposed to X-irradiation. (A) The appearance of remyelination 2 months after injection of ethidium bromide into the dorsal funiculus of a normal 5 month old rat. Axons are remyelinated by both Schwann cells and oligodendrocytes. Oligodendrocyte remyelination (O) can be distinguished from normal myelination (N) because the myelin sheaths that surround the axons are thinner than normal for the axons they surround and during the early stages of remyelination do not stain as intensely as normal myelin sheaths. Schwann cell myelination (S) can be distinguished from oligodendrocyte myelination because the axons are not as closely packed as oligodendrocyte myelinated axon and often a nucleus can be detected next to the myelin sheath. (B) The appearance of an ethidium bromide induced area of demyelination made in X-irradiated white matter. The demyelinated axons lie clumped together and there is no evidence of remyelination by either oligodendrocytes or Schwann cells. The only cell present between the demyelinated axons are debris Wlled macrophages (M) and an occasional astrocyte process. Because of the absence of glial cells in the area of demyelination (other than macrophages) any cell found in the area of demyelination following transplantation can be deemed to be transplant-derived. Bar ¼ 50 mm. Toluidine blue stained one micron plastic sections.

demyelination with this preparation (Keirstead and Blakemore, 1997). The areas of demyelination are shorter in length but involve a greater area of the dorsal funiculi than those induced using lysolecithin. Remyelination is usually undertaken by oligodendrocytes unless the lesions are large, in which case axons in the center of the lesions are remyelinated by Schwann cells. Some oligodendrocytes survive within the area of demyelination. The main advantage of antibody-mediated demyelination over lysolecithin is that it is more axon sparing. X-irradiation Although X-irradiation does not result in demyelination, it provides a valuable tool for creating lesions devoid of endogenous remyelinating capacity that can then be used to investigate the remyelinating potential of glial cell preparations (Figures 6.2 and 6.3). X-irradiating the rat spinal cord with a single dose of 40 Gy of X-irradiation inhibits remyelination by depleting tissue of its endogenous oligodendrocyte progenitor (OPC) population (Hinks et al., 2001). When demyelinating lesions are made in tissue exposed to 40 Gy of X-irradiation, there is no remyelination by host cells (Blakemore and Patterson, 1978). The nonrepairing nature of demyelinating lesions made in long lengths of X-irradiated tissue can be rigorously tested by showing that no remyelination is present in animals killed 2 months after X-irradiation (Blakemore et al., 2002). In addition, the injection of cell preparations that can cause recruitment of host cells fail to do so when injected into X-irradiated lesions (Blakemore et al., 1995b; Crang et al., 1991). Furthermore, there is a direct correlation between the nature of repair observed and the composition of the cells introduced into this environment (Blakemore and Crang, 1992; Fig. 6.3).

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

EXPERIMENTAL MODELS

FIGURE 6.3 The appearance of X-irradiated ethidium bromide lesions following the injection of oligodendrocyte precursors A and Schwann cells B. One micron toluidine blue stained resin sections provide a useful way of identifying and distinguishing between oligodendrocyte and Schwann cell remyelination and assessing the extent of remyelination. (A) Oligodendrocyte remyelination can be distinguished from normal CNS myelination because the myelin sheaths that surround axons are thinner than normal. (B) Schwann cell myelination can be distinguished from oligodendrocyte myelination because the axons are not as closely packed as oligodendrocyte myelinated axon and often a nucleus can be detected next to the myelin sheath. Some demyelinated axons are still present (D) and in such areas the early stages of ensheathment can be seen. N ¼ normal tissue and M ¼ myelin debris Wlled macrophages. Bar ¼ 40 mm.

When xenogenic cells are used, survival is only seen when animals are immunosuppressed (Crang and Blakemore, 1991). This nonrepairing, cell-free lesion therefore provides an ideal system in which to assess the diVerentiation potential of diVerent cell preparations without the need to resort to the use of cell markers to identify the transplanted cells (Akiyama et al., 2001; Franklin et al., 1995; Groves et al., 1993; Sasaki et al., 2001). It also provides a system in which it is possible to directly compare the relative remyelinating capacity of endogenous and transplanted neonatal OPCs (Blakemore et al., 2000b), and by further manipulations, to model large areas of OPC-depleted demyelination, by creating a zone of OPC depletion between a source of transplanted OPCs and an area of demyelination that must be repopulated before OPCs can interact with demyelinated axons. Since this repopulation takes time, it delays the interaction of OPCs and demyelinated axons and allows examination of the consequences for remyelination of dissociating remyelination from demyelination (Blakemore et al., 2002). Diffuse Lesions There are a number of other models of demyelination that could be used to explore the potential of cell transplantation in which the areas of demyelination are more diVusely distributed through the CNS. These include experimental allergic encephalomyelitis

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

151

152

6. REMYELINATION THROUGH ENGRAFTMENT

(EAE), demyelination induced by intrathecal injection of Cholera toxin B-subunit conjugated to saporin, cuprizone intoxication, and various viral induced demyelinating diseases. To date, these have seldom been used to assay cell therapy in view of their more complex pathophysiological features, the dispersion of lesions and the fact that they all show endogenous remyelination. However, they could oVer the possibility, with greater availability of robust marker systems like cells derived from GFP and lacZ expressing transgenics, to allow investigators to examine critical issues in cell therapy such as the ability of transplanted cells to gain access to lesions dispersed throughout the CNS, the resistance of the transplanted cells to inXammation and recurrent episodes of demyelination as well as their ability to sustain clinical recovery. Experimental allergic encephalomyelitis (EAE) Demyelination can be induced by sensitizing animals to oligodendrocyte and myelin-related molecules by injection of adult CNS homogenate, puriWed myelin, or myelin proteins. EAE can be induced in several species including monkeys (reviewed in Van der Goes and Dijkstra, 2001). Clinical signs of EAE and extent of demyelination varies from species to species and according to the procedure of immunization. The models that are monophasic (acute EAE or passive transfer EAE) show little demyelination, while those with remission and exacerbation (chronic relapsing EAE or antibody-dependent EAE) have demyelinating lesions that are widely dispersed through the CNS with a prevalence for the spinal cord. Despite the fact that some of these models reXect very closely the pathogenesis of MS, their complexity may explain their infrequent use for cell transplantation studies. So far these models have reported the potential of survival and migration of the transplanted cells as well as their ability to improve clinical recovery (Pluchino et al., 2003; Tourbah et al., 1997). Nonmyelinated Tracts (Eye) Although myelinated in the optic nerve, the intraretinal segments of ganglion cell axons are unmyelinated in rodents and thus provide targets for glial cell transplantation experiments. In some species (dog, man, rabbit, chicken), bundles of these axons are myelinated by oligodendrocytes while in other species Schwann cell myelination has been reported to occur occasionally (Bussow, 1978; Wyse and Spira, 1981). Schwann cell myelination of ganglion cell axons has been observed following traumatic lesions to the back of the eye in young, but not older rats (Perry and Hayes, 1985). Attempts to achieve myelination through Schwann cell transplantation in mature cats resulted in extremely limited myelination (Blakemore et al., 1986). Transplant-mediated oligodendrocyte myelination, which increases with time post-implantation and sometimes covers up to 25% of the retina, has been achieved following transplantation of oligodendrocyte precursor enriched cultures into eyes of neonatal rats (Huang et al., 1991; Laeng et al., 1996) and EGF expanded neural precursors into neonatal mice (Ader et al., 2000). To date there have been few studies in which oligodendrocyte precursors have been introduced into the retina of mature animals. However, a recent study introduced EGF/FGF expanded neural precursors into normal and dystrophic retina of 6-month-old mice and after 1 month some evidence of oligodendrocyte diVerentiation and axon association was detected (Pressmar et al., 2001). The same cells introduced into 2-month-old animals that were allowed to survive for 6 months produced signiWcant amounts of myelin.

CONSEQUENCES OF TRANSPLANTING DIFFERENT CELL PREPARATIONS Some principles underlying transplantation studies were discussed in the section titled ‘‘The Principles of Glial Cell Transplantation Experiments.’’ It will become apparent in this section that the environment into which transplanted cells are introduced inXuences their survival and diVerentiation. Survival of the transplanted cells is excellent in the

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

CONSEQUENCES OF TRANSPLANTING DIFFERENT CELL PREPARATIONS

embryonic and neonatal CNS since this tissue is generating oligodendrocytes and thus transplanted cells are exposed to an environment that provides optimal trophic support. Although a similar situation exists in acutely demyelinating lesions, the signals that the cells are exposed to are quite diVerent and as a consequence the diVerentiation potential of cells may be altered.

Neural Stem Cells, Neural Precursors, and Oligodendroglia Cells Derived from the Immature Nervous System Cells of the oligodendrocyte lineage and their precursors were extensively studied Wrst in the developing brain and spinal cord of myelin mutants and later in various models of demyelination in the adult CNS. Myelination was Wrst achieved by transplantation of CNS fragments containing oligodendrocyte and neural precursors (Gumpel et al., 1983) and subsequently with mixed glial cell populations (Blakemore and Crang, 1988) and oligodendrocytes (Duncan et al., 1988b), puriWed populations of deWned cell stages of the oligodendroglial lineage (Warrington et al., 1993), spheres of oligodendrocyte precursors (oligospheres) (Avellana-Adalid et al., 1996), and neurospheres (Hammang et al., 1997; Milward et al., 1997) (Fig. 6.4). The oligosphere strategy provided a means to expand oligodendrocyte precursors from neural stem cells (Zhang et al., 1998a), neural precursors (Vitry et al., 1999), or expand pre-oligodendrocytes and oligodendrocytes from embryonic cells (Liu et al., 2000). Cryopreserved and long-term expanded oligodendrocyte precursors retain their capacity to myelinate axons (Archer et al., 1994; Avellana-Adalid et al., 1996). In addition, cells derived from neonatal rodents, canine (Archer et al., 1994; Zhang et al., 1998b), pig (Smith and Blakemore, 2000a), and cat (Targett and Blakemore, 1994) CNS can generate myelin-forming cells when grafted into the hypo- or demyelinated rodent CNS. Myelination has also be obtained after intra-CNS transplantation of the growthfactor-generated oligodendrocyte precursor cell line CG4 (Espinosa de los et al., 1997; Franklin et al., 1995; Tontsch et al., 1994). Conditionally immortalized oligodendrocytegenerating cell lines have produced variable results following transplantation. In some cases, cells show diVerentiation and myelin sheath formation, while in other instances cells continue to divide and do not make myelin sheaths (Barnett et al., 1993; Jung et al., 1995; Trotter et al., 1993). Transplantation of an immortalized multipotent neural precursor cell line into the immature CNS has also resulted in the generation of myelin-forming oligodendrocytes (Yandava et al., 1999). From the early studies it soon became apparent that the myelination potential of cell preparations following transplantation was highly dependent on its degree of enrichment in immature cells. Thus, fetal donor cells generated myelin more eVectively than neonatal and adult ones following transplantation in the neonatal CNS (Archer et al., 1994, 1997; Rosenbluth et al., 1990, 1993). Transplantation of selected cell stages of the oligodendroglial lineage into the neonate brain indicated that oligodendrocyte precursors were more eYcient in myelin formation than pre-oligodendrocytes or mature oligodendrocytes (Warrington et al., 1993). A greater complexity was revealed when neural precursors (cells capable of generating neurons in addition to glial cells) and pre-oligodendrocytes were transplanted into the adult CNS and into astrocyte free areas of acute demyelination. Although Zhang and colleagues considered that transplantation of oligospheres resulted in more myelin formation than transplantation of neurospheres (Zhang et al., 1998a), no diVerence in the myelin domain per grafted cell was found after transplantation of neural precursors and oligodendrocyte precursors by Vitry and colleagues (2001). Transplantation of cells that show a high degree of plasticity leads to the generation of cell types other than oligodendrocytes. Thus, when embryonic glial restricted precursors were transplanted in a myelin-deWcient background, they diVerentiated into both oligodendrocytes and astrocytes (Herrera et al., 2001). Multipotent neural precursors generate oligodendrocyte and astrocytes (Mitome et al., 2001; Vitry et al., 2001) when targeted to potential white matter regions but also neurons in neurogenic territories (Vitry et al., 2001). Although intraventricular injection of neural precursors led to widespread colonization of newborn

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

153

154

6. REMYELINATION THROUGH ENGRAFTMENT

FIGURE 6.4 The relationship between the starting point and type of cell preparation from which oligodendrocyte myelination has been achieved in rodents following transplantation of rodent cells (A) or human cells (B). Using rodent cells, the most extensive myelination is achieved following the introduction of oligosphere preparations. At present, it is not yet been possible to produce this preparation with human cells.

shiverer brains, only 28% of the transplanted cells were estimated to diVerentiate into oligodendrocytes, the remaining population having diVerentiated in astrocytes and neurons (Yandava et al., 1999). In contrast to their introduction into the developing nervous system, the transplantation of multipotent neural precursors into glial-free areas of acute demyelination does not result in oligodendrocyte generation unless signiWcant numbers of cells are committed to the oligodendrocyte lineage prior to transplantation (Smith and Blakemore, 2000a; WFB, personnel observations). Commitment of neural precursors to the oligodendrocyte lineage can be achieved for rodent, pig, and canine, but not human neural precursors by a short period in culture with B-104 neuroblastoma-conditioned medium (Avellana-Adalid et al., 1996; Learish et al., 1999; Smith and Blakemore, 2000a; Vitry et al., 1999; Zhang et al., 1998a, 1999). A further example of how the environment into which cells are introduced inXuences their diVerentiation is illustrated by Wnding that PSA-NCAM positive precursors that give rise to oligodendrocytes, astrocytes, and neurons when introduced into the developing brain (Vitry et al., 2001) generate oligodendrocytes and Schwann cells when introduced into the astrocyte-free areas of acute demyelination induced by EB injection (Keirstead et al., 1999). When the (pluripotent) embryonic stem cells are grafted in the brain, they can give rise to other tissue derivatives

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

CONSEQUENCES OF TRANSPLANTING DIFFERENT CELL PREPARATIONS

(Bru¨stle et al., 1997) unless primed in vitro to become neural precursors, in which case they diVerentiate into neurons, astrocytes, as well as myelin-forming oligodendrocytes when introduced into the developing brain (Bru¨stle et al., 1999, Liu et al., 2000). These observations suggest that if pluripotent and multipotent precursors are to be used as the source material for cell therapy of demyelinating diseases they should be committed ex vivo to the oligodendroglial lineage prior to implantation (Fig. 6.4). When transplanted cells are able to survive in host tissue they show distinct patterns of migration, and it is apparent that some cells show greater migration potential than others. Transplantation of fragments containing oligodendrocyte precursors and neural precursors into the developing shiverer CNS (Gumpel et al., 1983) showed that cells tended to follow axonal tracks and thus the success of their migration was found to be largely inXuenced by their point of implantation. However, other routes must also be involved since cells implanted in the forebrain have been found in the brain stem and cerebellum (Baulac et al., 1987; Gansmuller et al., 1986). The early Wndings obtained after transplantation in the myelin mutants were conWrmed in the neonate wild-type environment. However, it was noted that the extent of migration was less in normal animals than in the myelin mutants (Friedrich and Lazzarini, 1993; Hardy and Friedrich, 1996; Lachapelle et al., 1994a; Vitry et al., 2001). Transplantation in the neonate brain also showed that immature cells migrate more eYciently than diVerentiated ones since oligodendrocyte precursors migrate further than pre-oligodendrocytes (Warrington et al., 1993) and PSANCAMþ neural precursors migrate more eYciently than oligodendrocyte precursors (Vitry et al., 2001). Repeated delivery of neural precursor cells through the ventricle or the cisterna magna can result in extensive colonization of the neonate brain (Mitome et al., 2001). Intraventricular transplantation of neural precursor or oligodendrocyte precursor cells indicates that the migration of these cells is even more extensive in the embryonic brain leading to widespread colonization of the cerebral parenchyma (Brustle et al., 1999; Learish et al., 1999; Yandava et al., 1999). Transplantation of oligodendrocyte precursor in the spinal cord has shown that cells can migrate up to 3 to 4 mm in the md rat (Tontsch et al., 1994) and 20 mm in the shaking pup (Archer et al., 1997). It is now becoming clear that in the adult nervous system the migratory potential of transplanted neonatal and adult oligodendrocyte precursors, and embryo-derived neural and glial precursors may be diVerent. Neonatal oligodendrocyte precursor cells (O’Leary and Blakemore, 1997a) and the CG4 oligodendrocyte precursor cell line (Franklin et al., 1996a), both of which migrate extensively when introduced into the developing nervous system (discussed earlier), fail to migrate when introduced into the normal adult spinal cord. In contrast, embryo-derived glial-restricted precursors (Herrera et al., 2001) and embryo-derived neural precursors (Ader et al., 2001; Isacson et al., 1995; Mitome et al., 2001; Svendsen et al., 1997) show migration through normal tissue away from their point of implantation. However, the rate and extent of migration is less in adults than in neonates. When adult tissue is exposed to 40 Gy of X-irradiation, a procedure that removes the oligodendrocyte precursor cell population (Hinks et al., 2001), transplanted neonatal oligodendrocyte precursors colonize tissue, and they do so at a rate of 1.7 to 2.0 mm per week (Blakemore et al., 2002). Some studies have found that neonatal cells implanted in normal white matter at a distance from an area of demyelination are attracted to the area of demyelination with evidence that some of this migration takes place through normal tissue (Gout et al., 1988; Gout and Dubois-Dalcq, 1993; Vignais et al., 1993). Others have failed to demonstrate this (Franklin et al., 1996a) suggesting that a major route of migration may be along blood vessels or via CSF Wlled spaces (Baron-Van Evercooren et al., 1996) or that there are fundamental diVerences between experimental protocols. When the CG4 cell line was co-transplanted, or transplanted at a distance from cells of the B104 cell line, enhanced migration was observed, suggesting that soluble signals of neuronal origin can enhance migration (Espinosa de los Monteros et al., 2001; Milward et al., 2000). The presence of inXammation appears to promote colonization of normal adult tissue by the CG4 cell line (Tourbah et al., 1997) and neonate-derived neural precursors (Ben-Hur et al., 2003). Thus, one can speculate that environmental factors related to inXammation promote colonization of normal tissues while in the absence of

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

155

156

6. REMYELINATION THROUGH ENGRAFTMENT

inXammation the presence of endogenous oligodendrocyte precursors prevents colonization of normal tissue by neonatal and adult-derived cells but not embryo-derived cells. Cells Derived from the Adult Nervous System Oligodendrocytes Until recently it was considered that the cells that remyelinated axons in the adult CNS could be derived from both oligodendrocyte progenitors and oligodendrocytes (Chari and Blakemore, 2001b). Some of the evidence that mature oligodendrocytes could give rise to myelinating cells came from transplant experiments in which cell preparations considered to be either enriched or puriWed to contain oligodendrocytes were used. Thus, a cell preparation containing 85% GalC positive cells isolated from the 7day-old mouse brain resulted in areas of oligodendrocyte remyelination (Kohsaka et al., 1986). Likewise cells isolated from 4- to 6-week (Lubetzki et al., 1988) or 8-week-old rats (Duncan et al., 1992), transplanted into shiverer mice or md rat, respectively, resulted in oligodendrocyte remyelination and evidence of focal aggregations of oligodendrocyte cells, some of which were surrounded by ‘‘myelin-like’’ membranes (Duncan et al., 1992). Remyelination has also been observed following injection of cells derived from the brain of 4- to 5-month-old rats into X-EB lesions (Crang et al., 1998). However, when that oligodendrocyte preparation was X-irradiated with 40 Gy of X-irradiation, only oligodendrocytes and myelin sheet formation was observed. This result was interpreted as illustrating that oligodendrocytes isolated from the adult CNS were post-mitotic cells incapable of forming myelin sheaths. A similar behavior was observed following the introduction of oligodendrocytes isolated from the adult human brain (Targett et al., 1996). When cells isolated from the adult rat CNS using magnetic beads coated with antibodies to myelin oligodendrocyte glycoprotein (MOG) were transplanted into the X-EB lesions, the result was oligodendrocyte and Schwann cell remyelination (Blakemore et al., 2000a). In these experiments, MOG was chosen as a target to isolate the cells since it, like galactocerebroside expression used in previous studies, was considered to be a molecule whose expression was restricted to mature oligodendrocytes. However, recent studies have indicated that in the adult rat CNS, galactocerebroside and MOG expression is not restricted to mature oligodendrocytes but can also be expressed by oligodendrocyte progenitors (Li et al., 2002). This co-expression could well explain why preparations of cells considered to be mature oligodendrocytes on the basis of galactocerebroside expression could result in myelination following transplantation. Adult-derived neural precursors Adult-derived neural ‘‘stem cells’’ can be propagated in vitro either as neurospheres or as Xat cultures and adult-derived oligospheres have been transplanted into md rat pups resulting in an extensive myelin formation as would be expected following implantation of neonatal oligospheres (Zhang et al., 1999). Adult SVZ fragments containing neural stem cells and immature neural precursors form myelin when transplanted into the shiverer and wild-type neonate brain (Lachapelle et al., 2002a). In these experiments, the potential for myelin formation was somewhat restricted compared to that observed in the study of Zhang and colleagues (1999), however, priming SVZ donors with FGF enhanced the amount of myelin formed by the transplanted cells (Lachapelle et al., 2002a). Another recent study has shown that when adult neural precursors are infected intravenously or grafted intrathecally into mice with EAE, there is amelioration of the clinical disease and a decrease in the extent of demyelination (Pluchino et al., 2003).

Olfactory Glia Transplantation of the glial cells associated with the olfactory nerve and bulb results in myelination of demyelinated axons (Barnett et al., 2000; Franklin et al., 1996b; Imaizumi et al., 1998, 2000; Kato et al., 2000). The myelin formed contains P0 and the cell’s association with the axon resembles that of Schwann cells. To date all olfactory glial transplants have been conducted using the X-irradiated ethidium bromide lesion or models

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

157

CONSEQUENCES OF TRANSPLANTING DIFFERENT CELL PREPARATIONS

of spinal cord transaction. These are situations where demyelinated axons are present in astrocyte depleted tissue, and thus the remyelinating potential of these cells in astrocyte dominated areas of demyelination has not yet been tested. Issues relating to these cells and their potential advantage over peripheral nerve-derived Schwann cells are not described in this chapter and can be found in Chapter 14.

Schwann Cells Remyelination of demyelinated CNS axons by grafted Schwann cells has been achieved in the spinal cord (Baron-Van Evercooren et al., 1991; Blakemore, 1977; Duncan et al., 1981), cerebellar white matter (Shields et al., 2000), cortex, and optic disc (Blakemore et al., 1986). It has been demonstrated in several species including mice (Baron-Van Evercooren et al., 1992; Duncan et al., 1981), rat (Blakemore, 1977), cat (Blakemore and Crang, 1985), and monkeys (Lachapelle et al., 2000) with Schwann cells transplanted either as dissociated nerve fragments (Blakemore, 1977), enriched (Blakemore and Crang, 1985; Duncan et al., 1981), or highly puriWed and expanded Schwann cell preparations (Avellana-Adalid et al., 1998; Baron-Van Evercooren et al., 1991). Cross-species transplantation allowed the assay of the consequences of cryopreservation (Kohama et al., 2001), immortalization (BaronVan Evercooren et al., 1992; Boutry et al., 1992), and preparative procedures on the myelinating eYciency of cells from various species including primates (Avellana-Adalid et al., 1998; Brierley et al., 2001; Lankford et al., 2002). Remyelination by transplanted Schwann cells depends on appropriate axon signals since it only occurs if the previously myelinated axons are of suYcient size to be myelinated by Schwann cells and thus injection of Schwann cells into areas that are not myelinated by endogenous myelin-forming cells such as the hippocampus or Wmbria does not result in myelination (Brook et al., 1993). Schwann cells myelinate successfully when transplanted into or over gliotoxin induced areas of demyelination (Baron-Van Evercooren et al., 1992; 1993, Blakemore, 1977, 1980, 1984; Blakemore and Crang, 1985; Duncan et al., 1981; Honmou et al., 1996; Lachapelle et al., 2002b; Shields et al., 2000). In the spinal cord, implanted Schwann cells sustain stable myelin for up to 5 months after transplantation (Duncan et al., 1981) and following endogenous Schwann cell remyelination there is no evidence that Schwann cell remyelination is replaced by oligodendrocyte remyelination (Gilson and Blakemore, 2002), as has been suggested by some investigators (Jasmin et al., 2000). Of interest, the myelin formed by Schwann cells transplanted in the demyelinated spinal cord improves markedly the conduction of demyelinated axons (Honmou et al., 1996). When compared to grafted olfactory ensheathing cells in the mildly contused spinal cord, Schwann cells myelinate and sustain functional improvement more vigorously than olfactory ensheathing cells (Takami et al., 2002). The extent of remyelination achieved is related to the number of Schwann cells introduced (Iwashita and Blakemore, 2000) and their purity (Brierley et al., 2001). Albeit Schwann cells from neonate and young adult donors can achieve successful repair, this ability is to some extent reduced with old donors (Lankford et al., 2002). While few data exist on the repair potential of the more immature stages of the Schwann cell lineage, successful remyelination was achieved after transplantation of a conditionally immortalized Schwann cell precursor line (Lobsiger et al., 2001). Some studies indicate that transplanted Schwann cells are attracted to areas of demyelination (Baron-Van Evercooren et al., 1992, 1993, 1996), others have failed to conWrm this (Iwashita et al., 2000b). This potential to migrate may depend upon the environment in which cells are placed since Schwann cells migrate successfully along the meninges and the blood vessels, but fail to migrate in normal white matter tracks (Baron-Van Evercooren et al., 1992; Brook et al., 1993; Iwashita et al., 2000b; Langford and Owens, 1990; Shields 2000). The most important factor that may aVect Schwann cell remyelination is the presence of astrocytes within the area of demyelination. The close association between astrocyte loss and Schwann cell remyelination (Franklin and Blakemore, 1993; Shields et al., 2000), and the demonstration in transplantation experiments that the presence of (Blakemore et al., 1986), or reconstruction of an astrocyte environment by the co-transplantation of

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

au1

158

6. REMYELINATION THROUGH ENGRAFTMENT

astrocytes (Franklin et al., 1991) limits Schwann cell remyelination of demyelinated axons, suggests that astrocytes and Schwann cells are mutually exclusive cells (Franklin and Blakemore, 1993). However, migration and integration of Schwann cells transplanted in the CNS is favored by surgical procedures such as microtransplantation, which limits astrocyte reactivation (Li and Raisman, 1997). Thus, the role of astrocytes in limiting Schwann cell myelination in the CNS is not clear cut. Schwann cell myelination occurs in the older myelin-deWcient animals despite there being astrocyte hypertrophy (Duncan and HoVman, 1997). It also occurs when Schwann cells are grafted into lysolecithin induced or mild contusion lesions in the rodent spinal cord (Baron-Van Evercooren et al., 1992, Takami et al., 2002). Recent data shows that remyelination by transplanted autologous Schwann cells is successful in lysolecithin-induced lesions of the monkey spinal cord, which contain astrocytes, but fails in optic nerve lesions that are free of astrocytes (Lachapelle et al., unpublished results). The relationship between astrocytes and Schwann cells may be further complicated by competition with the host OPCs for demyelinated axons (Blakemore and Crang, 1989; Franklin et al., 1992). An interaction that may explain in part why Schwann cell transplantation fails in the newborn shiverer mouse (Baron-Van Evercooren et al., 1992) that contains viable but defectively myelinating oligodendrocytes, but can occur in the md rat where axon associating oligodendrocytes are absent (Duncan et al., 1988a). Optic nerve astrocytes slightly increase the remyelinating potential of transplanted Schwann cells (Langford et al., 2002). Thus, astrocytes may have properties that could both enhance as well as limit transplanted Schwann cells ability to remyelinate demyelinated axons (Blakemore and Crang, 1985; Franklin and Blakemore, 1993). In view of these controversies and the fact that phase one clinical trials are underway in MS patients, it is critical that the potential behavior of Schwann cells transplanted into areas of acute and chronic demyelination is resolved. Until recently authors did not take into account the possibility that Schwann cells could be generated from neural precursors. That this can occur is demonstrated by both in vitro and in vivo studies. In vitro, cloned CNS neural precursors can be shown to give rise to both oligodendrocytes and Schwann cells (Mujtaba et al., 1998). In vivo experiments using the X-irradiated ethidium bromide lesion have demonstrated that the introduction of FGF-2 expanded PSA-NCAM isolated glial restricted precursors devoid of Schwann cells results in both Schwann cell and oligodendrocyte remyelination (Keirstead et al., 1999) and introduction of cloned human neural precursors isolated from the adult human brain resulted in Schwann cell remyelination (Akiyama et al., 2001). These studies indicate that CNS precursors can give rise to Schwann cells and suggest that rather than astrocytes acting as a barrier that prevents diVerentiated Schwann cells from associating with demyelinated CNS axons, their presence or absence inXuences the fate of CNS precursors that enter areas of demyelination; in the presence of astrocytes they diVerentiate into oligodendrocytes, while in their absence they come under the inXuence of factors that will induce Schwann cell diVerentiation.

PLURIPOTENT STEM CELLS FROM OUTSIDE THE CNS Adult bone marrow contains mesenchymal stem cells and haematopoietic stem cells that can be diVerentiated into various cell types including neural cell types in vitro (Colter et al., 2001; Kohyama et al., 2001; Sanchez-Ramos et al., 200) and following incorporation into blastocyts (Jiang et al., 2002) or transfer into the brain (Brazelton et al., 2000; Kopen et al., 1999; Mezey et al., 2000; Nakano et al., 2001). While intraparenchymal transplantation of bone marrow CD 117þ cells into the neonatal mouse brain resulted in the generation of oligodendrocytes (Bonilla et al., 2002), injection of freshly isolated rat bone marrow (Sasaki et al., 2001) and puriWed mouse bone marrow stromal cells (Akiyama et al., 2002b) into the X-Eb lesion resulted in Schwann cell remyelination. Intravenous injection of rat bone marrow acutely fractionated for mononuclear cells into animals with X-EB lesions also resulted in Schwann cell and possibly oligodendrocyte remyelination (Akiyama et al., 2002a). Intravenous injection of olfactory glia and Schwann cells resulted in no

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

MYELINATION FOLLOWING TRANSPLANTATION OF CELLS DERIVED FROM HUMAN TISSUE

remyelination. An alternative to the use of bone marrow is cord blood, which is rich in stem cells. A recent study claims that signiWcant oligodendrocyte diVerentiation (10%) can be achieved in vitro when cells from human cord blood is co-cultured with cortical cells (Buzanska et al., 2002) and intraparenchymal transplantation of human umbilical cord blood cells into the newborn rodent brain led to the generation of a small population of neural cells including neurons (Zigova et al., 2002); however, so far the genesis of oligodendrocytes has not been described. Bone marrow and cord blood-derived stem cells in view of their greater accessibility and possibility of transfer to the CNS in a noninvasive mode, may thus oVer a promising alternative source for myelinating cells that could be used as autografts for myelin repair. However, these studies are still very preliminary and other studies have failed to conWrm these initial observations (Brazelton et al., 2000; Mezey et al., 2000; Wagers et al., 2002). As the use of bone marrow cells may have its own limitations, multipotent stem cells from skin (Toma et al., 2001) and muscle (RomeroRamos et al., 2002) have been identiWed. Their priming in culture generates neurospheres which can diVerentiate in oligodendrocytes, astrocytes and neurons. Albeit their ability to form functional oligodendrocytes after transplantation has to be demonstrated, these peripheral stem cells could be an attractive novel source of myelin-forming cells for autologous replacement in demyelinating diseases.

MYELINATION FOLLOWING TRANSPLANTATION OF CELLS DERIVED FROM HUMAN TISSUE Both human Schwann cells and oligodendrocytes can myelinate rodent axons in vitro, and thus it is possible to evaluate the remyelinating ability of human cells using the same in vivo models as those used to evaluate rodent cells. Embryonic and neonatal recipients tolerate xenografts, while adult recipients have either to be immunocompromised animals or be given immunosuppressive agents.

Neural and Oligodendrocyte Precursors In vitro, small numbers of oligodendrocytes can be generated from many diVerent human ‘‘stem’’ cells, including cells from the olfactory bulb (Pagano et al., 2000), embryonic stem cells (ReubinoV et al., 2001; Zhang et al., 2001), neural multipotent precursor cells derived from embryonic brain (Brannen and Sugaya, 2000; Murray and Dubois-Dalcq, 1997; Vescovi et al., 1999), and adult fresh or post-mortem brain (Arsenijevic et al., 2001; Palmer et al., 2001; Roy et al., 1999;). Unless immortalized (Villa et al., 2000), long-term expansion of human stem cells and oligodendrocyte generation appears limited and cells from embryonic spinal cord fail to generate any oligodendrocytes (Quinn et al., 1999). It is not possible to produce oligospheres or any type of preparations enriched in oligodendrocyte precursor from human tissue and when transplanted into newborn rodent CNS, neural progenitors derived from human embryonic stem cells that give rise to neurons and astrocytes, only generate a very limited numbers of oligodendrocytes (ReubinoV et al., 2001) and in some experiments none at all (Svendsen et al., 1998; Zhang et al., 2001). Similarly, only few oligodendrocyte precursors are obtained with nonhuman primate neurospheres, in vitro or after transplantation in the neonate mouse brain (Avellana-Adalid, unpublished observations). By contrast, solid grafts of fresh and frozen human embryonic brain and spinal cord (8 to 20 weeks) give rise to areas of oligodendrocyte myelin following transplantation into the newborn shiverer brain indicating that human fetal cells transplanted in the newborn mouse brain can be well tolerated and induced to diVerentiate into functional oligodendrocytes in a environment that is itself generating myelinating cells (Gumpel et al., 1987; Seilhean et al., 1996). However, the extent of myelin formation obtained using human CNS fragments was always lower than that seen with mouse CNS fragments (Seilhean et al., 1996). Oligodendrocyte generation was achieved when freshly dissociated and neurosphere cultures derived from human fetal brain (8-10/16 weeks) were injected in utero into the

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

159

160

6. REMYELINATION THROUGH ENGRAFTMENT

brains of wild-type embryonic rat pups (Brustle et al., 1998) and into the ventricles of newborn wild-type mice (Flax et al., 1998; Tamaki et al., 2002). In all cases, neurons and astrocytes were generated; however, the extent of oligodendrocyte generation was low and myelin was not identiWed. Interestingly, when cloned adult human neural precursor cells were transplanted into X-EB lesions, they gave rise to Schwann cells (Akiyama et al., 2001). Thus, while primate embryonic stem cells and neural stem cells can be cultured, their ability to give rise to substantial populations of myelinating oligodendrocytes to repair the adult CNS remains an open issue. Other strategies aimed at isolating oligodendroglial committed populations from the adult brain are being explored (Roy et al., 1999; Roy et al., 2000; Keyoung et al., 2001). Cells isolated from the adult brain have recently been shown to be able to diVerentiate into oligodendrocyte and astrocytes in vivo after their transplantation into lysolecithin-induced myelin lesions in immunosuppressed rats, however myelin sheath generation was not documented (Windrem et al., 2002). The myelin-forming behavior of A2B5 sorted cells from the adult and fetal brain has been compared following transplantation into the neonatal shiverer brain. It was found that adult-derived cells diVerentiated more rapidly into myelin-forming cells (by 4 weeks post transplantation) than fetus-derived cells (myelin not being seen until 8 weeks post transplantation). There were also diVerences in the extent of migration of the transplanted cells, with fetal cells showing a more widespread colonization of host brains than the adult-derived cells. Also, when the two cell types were implanted into lysolecithin lesions made in the corpus callosum of immunosuppressed rats, the adultderived cells were not observed to invade the adjacent normal tissue while this was seen with the fetus-derived cells. Recently, a protocol for obtaining and achieving in vitro expansion of human oligodendrocyte precursors has been described (Dietrich et al., 2002). However, the ability of these cells to diVerentiation into mature oligodendrocytes as well as their ability to generate myelinating cells in vivo has not yet been established.

Schwann Cells Unlike human oligodendrocytes, human and monkey Schwann cells can be puriWed and expanded successfully from embryonic, neonatal, or adult peripheral tissues. While initial culturing of human Schwann cells was hampered by low cell yields, senescence, and contaminating Wbroblasts, the cloning of human growth factor heregulin genes and the development of recombinant technology has allowed investigators to obtain substantial populations (range of 5
106 to 108 for 100 mg of tissue) of Schwann cells from adult human (Levi, 1996; Rutkowski et al., 1995), or monkey (Avellana-Adalid et al., 1998) peripheral nerve biopsies. Expansion of these adult primate Schwann cells does not alter their antigenic properties in vitro (Avellana-Adalid et al., 1998; Hanemann et al., 1998) nor their functionality in vivo since expanded primate Schwann cells promote axonal regrowth after transplantation into the injured PNS (Levi et al., 1994) and diVerentiate in myelin-forming cells after transplantation into acute lesions of the demyelinated mouse spinal cord (Avellana-Adalid et al., 1998). Moreover, cryopreservation (Kohama et al., 2001) and in vitro expansion (Imaizumi et al., 1999) does not alter the ability of adult primate Schwann cells transplanted into lesions of the rodent spinal cord to remyelinate and overcome conduction block. While some concerns have been raised over the potential for development of tumors after long term expansion of human Schwann cells, no tumors were reported when human Schwann cells expanded for 6 passages in vitro were transplanted into lesioned rodent PNS (Emery et al., 1999). These technical advances have allowed the demonstration of the feasibility of achieving successful remyelination using autologous Schwann cell transplantation in adult monkey spinal cord (Lachapelle et al., 2002b). While the resistance of these cells to inXammation as well as their ability to sustain clinical recovery remain to be addressed, these encouraging data, in addition to the fact that peripheral myelin in MS is spared, make Schwann cells attractive candidates for therapeutic transplantation.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

ASSESSMENT OF POTENTIAL EFFICACY

ASSESSMENT OF POTENTIAL EFFICACY To date, the consequences of transplantation have mainly been assessed morphologically and with a few exceptions there is little quantitative data available to compare the relative merits of diVerent cell preparations. This is surprising since most of the models lend themselves to quantitative analysis and such studies will be vital to the development of human therapies. While most early studies reported the potential for survival, migration, and myelination of transplanted cell populations, the group of Kocsis and Duncan combined to show that transplant-mediated myelination by oligodendrocytes establishes normal conduction properties to amyelinated or demyelinated axons (Utzschneider et al., 1994). Neonate rodent and adult human Schwann cells (Honmou et al., 1996, Kohama et al., 2001) and olfactory ensheathing cells are able to overcome conduction block (Imaizumi et al., 1998, 2000). For adult primate Schwann cells, these repair capabilities are not aVected by cyopreservation and expansion (Kohama et al., 2001). Functional data on the potential beneWcial consequences of glial cell transplantation have been very limited so far. Using ethidium bromide to demyelinate the corticospinal tract in the cervical spinal cord of rats and a beam walking task to assess motor function, JeVery and coworkers were able to show that demyelination was associated with a measurable functional deWcit that was ameliorated by remyelination (JeVery and Blakemore, 1997). Suppression of remyelination by X-irradiation resulted in persistent motor deWcits, which could be prevented by transplantation of oligodendrocyte progenitors provided extensive axonal degeneration had not occurred (JeVery et al., 1999). Yandava and colleagues claimed that intraventricular transplantation of an immortalized multipotent neural precursor cell line into newborn shiverer mice resulted in correction of tremor in some animals (Yandava et al., 1999). However, myelination in sites relevant to tremor was not demonstrated and Mitome and colleagues performing multiple injections of neurospheres that resulted in very widespread myelination in the same model noted no clinical improvement (Mitome et al., 2001). More recent reports indicate that neural stem cell transplantation can lead to improvement of clinical recovery in EAE (Pluchino et al., 2003). However, a direct link between clinical improvement and myelinating capabilities of the transplanted cell population has yet to be established. Functional data on Schwann cells and olfactory ensheathing cells are still lacking. However, a recent study comparing the repair capacities of Schwann cells and olfactory ensheathing cells in the moderately contused spinal cord in rodents indicates that Schwann cells sustain functional recovery more eYciently than olfactory ensheathing cells and that this may result from their ability to remyelinate demyelinated axons more eYciently (Takami et al., 2002).

TARGET DISEASES AND POTENTIAL CLINICAL EFFECTIVENESS The clinical situations where transplant-mediated remyelination would be beneWcial are multiple sclerosis, the leucodystrophies, and, possibly, spinal cord trauma. In multiple sclerosis, the areas of chronic demyelination are large and scattered throughout the neuraxis, while in the leucodystrophies the myelin loss is extensive. Therefore, to have any beneWt, introduced cells will have to be able to colonize large areas of tissue. This will require both cell division and migration (and temporary maintenance of progenitor phenotype in the case of oligodendrocyte progenitors). Moreover, cells will have to colonize both normal tissue and pathological tissue and, as currently envisioned, the latter would be situations where the endogenous remyelination had failed. This makes the assessment of the likely outcome of clinical transplantation into large areas of chronic demyelination diYcult and rather speculative. However, it is now becoming clear that transplantation of myelin-forming cells into areas of acute demyelination could have signiWcant clinical beneWt, especially if failure of remyelination is due to loss of oligodendrocyte precursors along with oligodendrocytes during demyelinating episodes

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

161

162

6. REMYELINATION THROUGH ENGRAFTMENT

(discussed in Chari and Blakemore, 2002b). Moreover, in addition to providing cells to generate myelin sheaths, it is also possible that introduction of large numbers of immature cells into areas of acute pathology may be neuroprotective since such cells are a source of many growth factors known to have trophic properties. In the context of MS and spinal trauma, this property could be extremely beneWcial since axon loss is probably of greater clinical signiWcance than demyelination (Takami et al., 2002). However, to date, most authors have concentrated on chronic MS lesions as potential targets and the relative merits of oligodendrocyte progenitors, Schwann cells, and olfactory glia in such situations have been discussed in a number of recent reviews (Blakemore and Franklin, 2000; Franklin, 2002a; Zhang and Duncan, 2000). Because of this, we will only comment on the current status of key issues associated with potential clinical eYcacy in the context of chronically demyelinated lesions since these are likely to be the Wrst targets for clinical transplantation studies.

Availability of Cells It is clear that signiWcant numbers of cells would be required for cell transplantation therapies, which would necessitate in vitro expansion of cells. This can currently be achieved for Schwann cells and olfactory glia and only recently for oligodendrocyte precursors. However, it is not yet possible to generate oligodendrocyte precursors in large numbers form ES cells and neurospheres via the oligosphere approach (Tab. 6.1).

Safety There is no indication from animal studies or following implantation of human fetal tissue into the human brain that engraftment of myelinogenic cells will carry a risk over and above that associated with the procedure required to introduce the cells. It is important to stress that the use of artiWcially immortalized cells may pose a risk and it will be important that such cells are adequately screened in experimental animals. In addition, if these cells are to be used it will also probably require the introduction of a suicide gene that would enable cells to be destroyed if there was evidence of uncontrolled proliferation.

Rejection The exact requirement for histocompatibility between grafted cells and the host has not yet been established. In the case of Schwann cells, autologous transplantation is possible (Blakemore and Crang, 1985; Lachapelle et al., 2000; 2002b). Although this may be theoretically possible with olfactory glia and perhaps oligodendrocyte precursors, at present there are no studies that would indicate that this is feasible.

Colonization The experimental studies already discussed would indicate that oligodendrocyte precursors do not colonize normal adult tissue. However, there is evidence that they can be dispersed through the CSF compartment and colonize demyelinating lesions. Neonatal oligodendrocyte precursors have considerably more potential to colonize oligodendrocyte precursordepleted tissue and eVect remyelination than endogenous oligodendrocyte precursors. Cells generated from embryo-derived neurospheres and embryo-derived glial-restricted precursor cells can colonize normal adult tissue, but it is not yet clear if they would be able to generate oligodendrocyte precursors in response to demyelinating lesions. Schwann cells migrate poorly through normal CNS and they are generally excluded from astrocyte deWned areas; however, under some circumstances, they appear capable of myelinating axons in tissue that contains astrocytes and able to be dispersed through the CSF compartment and in perivascular spaces. Although olfactory glia migrate over astrocytes in vitro, this potential remains to be resolved in vivo.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

163

TARGET DISEASES AND POTENTIAL CLINICAL EFFECTIVENESS

TABLE 6.1 Summary of Oligodendrocytes, Schwann Cells, and Olfactory Ensheathig Cell Properties in the Context of Cell Therapy for Demyelinating Disease such as Multiple Sclerosis rOPC Purification Fetal/Neonatal Adult Expansion Fetal/Neonatal Adult Trophic support to CNS axons Autograft

pOPC

rSC

pSC

rOEC

pOEC

þ

þ

þ

þ

þ

?

þ

þ

þ

þ

þ

þ

þ

?

þ

þ

þ

?

þ

?

þ

þ

þ

þ





þ

þ

þ

þ

?

?

þ

þ

þ

þ

Migration in adult Normal tissue

þ/

?

?

?

Remyelination

þ

þ

þ

þ

þ

þ

Recovery of axonal conduction

þ

þ

þ

þ

þ

þ

Recovery of clinical deficit

þ

?

?

?

?

?

Vulnerability to ongoing disease process

þ

þ









?

þ/

 r: rodent, p: primate; þ ¼ favorable evidence;  ¼ unfavorable evidence; þ/ ¼ conflicting data; ? not yet established.

Ability to Myelinate Axons With the recent demonstration of oligodendrocyte precursors in areas of chronic demyelination (Chang et al., 2000; Wolswijk, 1998) and our increased understanding of the factors that drive the normal remyelination process (Franklin, 2002b; Franklin et al., 2002), one can conclude that the introduction of oligodendrocyte precursors into areas of chronic demyelination would not result in extensive myelination as the environment is not conducive for remyelination. Also, one is drawn to the same conclusion with regard to Schwann cell implantation because of the astrocyte presence around the demyelinated axons. Although progressive increase in the extent of myelination with time has been reported in myelin mutants following the introduction of embryo-derived cells (Ader et al., 2001; Mitome et al., 2001a), it is likely that for myelinating-cell therapy to be clinically eVective, the targeted lesions will have to share the properties of the experimental lesions in which transplantation has been shown to be an extremely eVective means of achieving remyelination. It should also be kept in mind that in MS the transplanted oligodendrocyte precursors and their progeny may become new targets of the disease or be rejected. This would be most unlikely with Schwann cells or olfactory glia, since peripheral myelin is not a target in MS and autologous transplantation is possible.

Conclusion The foregoing may be regarded as a rather negative assessment of the potential of achieving remyelination by engraftment. However, it stems from the fact that most studies so far undertaken have mainly been concerned with establishing the myelinogenic potential of diVerent cell preparations. It is essential that such studies are conducted in physiological systems. There is therefore a great need for experimental models that mimic clinical lesions, in particular ones that address questions of size and chronicity, as well as more studies that make direct comparisons between diVerent cell preparations in the same model and documentation of how the same cells behave in diVerent models. Such studies are under way and when completed using a variety of diVerent cell types, they will provide us with much more secure data on the best choice of cell. We will then be better able to predict the au2

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

164

6. REMYELINATION THROUGH ENGRAFTMENT

outcome following introduction of myelin-forming cells into areas of demyelination that either are, or are likely to develop into areas of chronic demyelination.

Acknowledgments We sincerely thank our friends and colleagues who have made major contributions to the development of this Weld of research. In particular, we would like to thank Francois Lachapelle, Brahim Nait-Oumesmar, and Robin Franklin for helpful comments and discussions that helped formulate many of the views expressed in this chapter. We would also like to acknowledge the technical expertise of John Crang and Jennifer Gilson, who played major roles in the experiments carried out at Cambridge, and Virginia AvellanaAdalid, Sandrine Vitry, and Corrine Bachelin for their major input in experiments conducted in Paris.

References Ader, M., Meng, J. H., Schachner, M., and Bartsch, U. (2000). Formation of myelin after transplantation of neural precursor cells into the retina of young postnatal mice. Glia 30, 301–310. Ader, M., Schachner, M., and Bartsch, U. (2001). Transplantation of neural precursor cells into the dysmyelinated CNS mutant mice deWcient in the myelin-associated glycoprotein and Fyn trosine kinase. European J. Neurosci. 14, 561–566. Akiyama, Y., Honmou, O., Kato, T., Uede, T., Hashi, K., and Kocsis, J. D. (2001). Transplantation of clonal neural precursor cells derived from adult human brain establishes functional peripheral myelin in the rat spinal cord. Expt. Neurol. 167, 27–39. Akiyama, Y., Radtke, C., Honmou, O., and Kocsis, J. D. (2002a). Remyelination of the spinal cord following intravenously delivery of bone marrow cells. Glia 39, 227–234. Akiyama, Y., Radtke, C., and Kocsis, J. D. (2002b). Remyelination of the rat spinal cord by transplantation of identiWed bone marrow stromal cells. J. Neurosci. 22, 6623–6630. Archer, D. R., Cuddon, P. A., Lipsitz, D., and Duncan, I. D. (1997). Myelination of the canine central nervous system by glial cell transplantation: A model for repair of human myelin disease. Nature Med. 3, 54–59. Archer, D. R., Leve´n, S., and Duncan, I. D. (1994). Myelination by cryopreserved xenografts and allografts in the myelin-deWcient rat. Expt. Neurol. 125, 268–277. Arsenijevic, Y., Villemure, J. G., Brunet, J. F., Bloch, J. J., De´glon, N., Kostic, C., Zurn, A., and Aebischer, P. (2001). Isolation of multipotent neural precursors residing in the cortex of the adult human brain. Expt. Neurol. 170, 48–62. Avellana-Adalid, V., Bachelin, C., Lachapelle, F., Escriou, C., Ratzkin, B., and Baron-Van Evercooren, A. (1998). In vitro and in vivo behaviour of NDF-expanded monkey Schwann cells. Eur. J. Neurosci. 10, 291–300. Avellana-Adalid, V., Nait-Oumesmar, B., Lachapelle, F., and Baron-Van Evercooren, A. (1996). Expansion of rat oligodendrocyte progenitors into proliferative ‘‘oligospheres’’ that retain diVerentiation potential. J. Neurosci. Res. 45, 558–570. Barnett, S. C., Alexander, C. L., Iwashita, Y., Gilson, J. M., Crowther, J., Clark, L., Dunn, L. T., Papanastassiou, V., Kennedy, P. G. E., and Franklin, R. J. M. (2000). IdentiWcation of a human olfactory ensheathing cell that can eVect transplant-mediated remyelination of demyelinated CNS axons. Brain 123, 1581–1588. Barnett, S. C., Franklin, R. J. M., and Blakemore, W. F. (1993). In vitro and in vivo analysis of a rat bipotential O-2A progenitor cell line containing the temperature sensitive mutant gene of the SV 40 large T antigen. Eur. J. Neurosci. 5, 1247–1260. Baron-Van Evercooren, A. (1994). Future prospects for transplantation. Annal. Neurol. 31, S151–156. Baron-Van Evercooren, A., Avellana-Adalid, V., Ben Younes-ChennouW, A., Gansmuller, A., Nait-Oumesmar, B., and Vignais, L. (1996). Cell-cell interactions during the migration of myelin-forming cells transplanted in the demyelinated spinal cord. Glia 16, 147–164. Baron-Van Evercooren, A., Avellana-Adalid, V., Lachapelle, F., and Liblau, R. (1997). Schwann cell transplantation and myelin repair of the CNS. Mult. Scler, 3, 157–161. Baron-Van Evercooren, A., Duhamel-Clerin, E., Boutry, J. M., Hauw, J. J., and Gumpel, M. (1993). Pathways of migration of transplanted Schwann cells in the demyelinated mouse spinal cord. J. Neurosci. Res. 35, 428–438. Baron-Van Evercooren, A., Gansmuller, A., Clerin, E., and Gumpel, M. (1991). Hoechst 33342 a Suitable fluorescent marker for schwann cells after transplantation in the mouse spinal cord. Neurosci. Lett. 131, 241–244. Baron-Van Evercooren, A., Gansmuller, A., Duhamel, E., Pascal, F., and Gumpel, M. (1992). Repair of a Myelin lesion by schwann cells transplanted in the adult mouse spinal cord. J. Neuroimmunol. 40, 235–242. Baulac, M., Lachapelle, F., Gout, O., Berger, B., Baumann, N., and Gumpel, M. (1987). Transplantation of oligodendrocytes in the newborn mouse brain: Extension of myelination by transplanted cells. Anatomical study. Brain Res. 420, 39–47.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

165

TARGET DISEASES AND POTENTIAL CLINICAL EFFECTIVENESS

Ben-Hur, T., Einstein, O., Mizrachi-Kol, R., Ben-Menachem, O., Reinhartz, E., Karusis, D., Abramsky, O. (2003). Transplanted multipotential neural progenitors migrate into the inXamed white matter in response to experimental autoimmune encephalomyelitis. Glia 4, 73–80. BiYger, K., Bartsch, S., Montag, D., Aguzzi, A., Schachner, M., and Bartsch, U. (2000). Severe hypomyelination of the murine CNS in the absence of myelin-associated glycoprotein and Fyn tyrosine kinase. J. Neurosci. 20, 7430–7437. Blakemore, W. F. (1977). Remyelination of CNS axons by Schwann cells transplanted from the sciatic nerve. Nature 266, 68–69. Blakemore, W. F. (1978). Observations on remyelination in the rabbit spinal_cord following demyelination induced by lysolecithin. Neuropath. Appl. Neurobiol. 4, 47–59. Blakemore, W. F. (1980). The eVect of sub-dural nerve tissue transplanation on the spinal cord of the rat. Neuropath. Appl. Neurobiol. 6, 433–447. Blakemore, W. F. (1982). Ethidium bromide induced demyelination in the spinal cord of the cat. Neuropath. Appl. Neurobiol. 8, 365–375. Blakemore, W. F. (1984). Limited remyelination of CNS axons by Schwann cells transplanted into the subarachnoid space. J. Neurol. Sci 64, 265–276. Blakemore, W. F., Chari, D. M., Gilson, J. M., and Crang, A. J. (2002). Modelling large areas of demyelination in the rat reveals the potential and possible limitations of transplanted glial cells for remyelination in the CNS. Glia 38, 155–168. Blakemore, W. F., and Crang, A. J. (1985). The use of cultured autologous Schwann cells to remyelinate areas of persistent demyelination in the central nervous system. J. Neurol. Sci 70, 207–223. Blakemore, W. F., and Crang, A. J. (1988). Extensive oligodendrocyte remyelination following injection of cultured central nervous system cells into demyelinating lesions in adult central nervous system. Dev. Neurosci. 10, 1–10. Blakemore, W. F., and Crang, A. J. (1989). The relationship between type-1 astrocytes, Schwann cells and oligodendrocytes following transplanation of glial cell cultures into demyelinating lesions in the adult rat spinal cord. J. Neurocytol. 18, 519–528. Blakemore, W. F., and Crang, A. J. (1992). Transplantation of glial cells into areas of demyelination in the adult rat spinal cord. In ‘‘Neural Transplantation: A Practical Approach’’ (S. Dunnett and A. Bjorklund, eds.), pp. 105–122. Oxford University Press, Oxford. Blakemore, W. F., Crang, A. J., and Curtis, R. (1986). The interaction of Schwann cells with CNS axons in regions containing normal astrocytes. Acta Neuropathol. 71, 295–300. Blakemore, W. F., Crang, A. J., and Franklin, R. J. M. (1990a). The transplantation of glial cells into areas of primary demyelination. In ‘‘Cellular and Molecular Biology of Myelination’’ (G. Jeserich, H. H. Althaus, and T. V. Waehneldt, eds.). Nato ASI Series Vol. 33, pp. 171–184. Springer-Verlag, Berlin, Heidleberg. Blakemore, W. F., Crang, A. J., and Franklin, R. J. M. (1990b). Transplantation of glial cell cultures into areas of demyelination in the adult CNS. In ‘‘Neural Transplantation: From Molecular Basis to Clinical Application’’ (S. Dunnett and S.-J. Richards, eds.). Prog. Brain Res. Vol. 82, pp. 225–232. Elsevier, Amsterdam. Blakemore, W. F., Crang, A. J., and Franklin, R. J. M. (1995a). Transplantation of glial cells. In ‘‘Neuroglia’’ (B. R. Ransom and H. Kettenmann, eds.), pp. 869–882. Oxford University Press, New York. Blakemore, W. F., Crang, A. J., Franklin, R. J. M., Tang, K., and Ryder, S. (1995b). Glial cell transplants that are subsequently rejected can be used to inXuence regeneration of glial cell environments in the CNS. Glia 13, 79–91. Blakemore, W. F., and Franklin, R. J. M. (1991). Transplantation of glial cells in the CNS. TINS 14, 323–327. Blakemore, W. F., and Franklin, R. J. M. (2000). Transplantation options for therapeutic CNS remyelination. Cell Transplantation 9, 289–294. Blakemore, W. F., Franklin, R. J. M., and Noble, M. (1995c). Glial cell transplantation and the repair of demyelinating lesions. In ‘‘Glial Cell Development, Basic Principles and clinical relevance’’ (K. R. Jessen and W. D. Richardson, eds.), pp. 209–220. Bios ScientiWc Publishers, Oxford. Blakemore, W. F., Gilson, J. M., and Crang, A. J. (2000a). The remyelinating capacity of MOGþve cells isolated from the adult rat CNS. Brain Pathol. 10, 639. Blakemore, W. F., Gilson, J. M., and Crang, A. J. (2000b). Transplanted glial cells migrate over a greater distance and remyelinate demyelinated lesions more rapidly than endogenous remyelinating cells. J. Neurosis. Res. 61, 288–294. Blakemore, W. F., and Patterson, R. C. (1978). Suppression of remyelination in the CNS by x-irradiation. Acta Neuropathol. 42, 105–113. Boison, D., and StoVel, W. (1989). Myelin-deWcient rat: A point mutation in exon lll (A -> C, Thr -> Pro) of myelin proteolipid protein causes dysmyelination and oligodendrocyte death. EMBO Journal 8, 3295–3302. Bonilla, S., Alarcon, P., Villaverde, R., Aparicio, P., Silva, A., and Martinez, S. (2002). Haematopoietic progenitor cells from adult bone marrow diVerentiate into cells that express oligodendroglial antigens in the neonatal mouse brain. Eur. J. Neurosci. 15, 575–582. Brannen, C. L., and Sugaya, K. (2000). In vitro diVerentiation of multipotent human neural progenitors in serumfree medium. NeuroReport 11, 1123–1128. Brazelton, T. R., Rossi, F. M., Keshet, G. I., and Blau, H. M. (2000). From marrow to brain: Expression of neuronal phenotypes in adult mice. Science 290, 1775–1779.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

au3

166

6. REMYELINATION THROUGH ENGRAFTMENT

Brierley, C. M. H., Crang, A. J., Iwashita, Y., Gilson, J. M., Scolding, N. J., Compston, D. A. S., and Blakemore, W. F. (2001). Remyelination of demyelinated CNS axons by transplanted human Schwann cells: The deleterious eVects of contaminating Wbroblasts. Cell Transplantation 10, 305–315. Brook, G. A., Lawrence, J. M., and Raisman, G. (1993). Morphology and migration of cultured Schwann cells transplanted into the Wmbria and hippocampus in adult rats. Glia 9, 292–304. Brustle, O. (1998). Building brains: Neural chimeras in the study of nervous system development and repair. Brain Pathol. 9, 527–545. Brustle, O., Choudhary, K., Karram, K., Huttner, A., Murray, K., Dubois-Dalcq, M., and McKay, R. (1998). Chermeric brains generated by intraventricular transplantation of fetal human brain cells into embryonic rats. Nat. Biotechnol. 16, 1040–1044. Brustle, O., Jones, K. N., Learish, R. D., Karram, K., Choudhary, K., Wiestler, O. D., Duncan, I. D., and McKay, R. D. G. (1999). Embryonic stem cell-derived glial precursors: A source of myelinating transplants. Science 285, 754–756. Brustle, O., Spiro, C. A., Karram, K., Choudhay, K., Okabe, S., and McKay, R. D. G. (1997). In vitro-generated neural precursors participate in mammalian brain development. Proc. Natl. Acad. Sci. USA 94, 14809–14814. Bulte, J. W., Douglas, T., Zhang, S.-C., Strable, E., Lewis, B. K., Zywicke, H., Miller, B., Van Gelderen, P., Moskowitz, B. M., Duncan, I. D., and Frank, J. A. (2001). Magnetodendrimers allow endosomal magnetic labeling and in vivo tracking of stem cells. Nat. Biotechnol. 19, 1141–1147. Bulte, J. W., Duncan, I. D., and Frank, J. A. (2002). In vivo magnetic resonance tracking of magnetically labeled cells after transplantation. J. Cereb. Blood Flow Metab. 22, 899–907. Bulte, J. W., Zhang, S. C., Van Gelderen, P., Herynek, V., Jordan, E. K., Duncan, I. D., and Frank, J. A. (1999). Neurotransplantation of magnetically labeled oligodendrocyte progenitors: Magnetic resonance tracking of cell migration and myelination. Proc. Nat. Acad. Sci. USA 96, 15256–15261. Bussow, H. (1978). Schwann cell myelin ensheathing C. N. S. axons in the nerve Wbre layer of the cat retina. J. Neurocytol. 7, 207–214. Buzanska, L., Machaj, E. J. K., Zablocka, B., Pojda, Z., and Domanska-Janik, K. (2002). Human cord bloodderived cells attain neuronal and glial features in vitro. J. Cell Sci. 115, 2131–2138. Cajal, S. R. Y. (1928). ‘‘Degeneration and Regeneration in the Nervous System.’’ Oxford University Press, London. Chang, A., Nishiyama, A., Peterson, J., Prineas, J., and Trapp, B. D. (2000). NG2-positive oligodendrocyte progenitor cells in adult human brain and multiple sclerosis lesions. J. Neurosci. 20, 6404–6412. Chari, D. M., and Blakemore, W. F. (2002a). EYcient recolonisation of progenitor-depleted areas of the CNS by adult oligodendrocyte progenitor cells. Glia 37, 307–313. Chari, D. M., and Blakemore, W. F. (2002b). New insights into remyelination failure in MS: Implications for glial cell transplantation. Mult. Scler. 8, 271–277. Colter, D. C., Sekiya, I., and Prockop, D. J. (2001). IdentiWcation of a subpopulation of rapidly self-renewing and multipotential adult stem cells in colonies of human marrow stromal cells. Proc. Natl. Acad. Sci. U. S. A 98, 7841–7845. Crang, A. J., and Blakemore, W. F. (1991). Remyelination of demyelinated rat axons by transplanted mouse oligodendrocytes. Glia 4, 305–313. Crang, A. J., Franklin, R. J. M., Blakemore, W. F., Trotter, J., Schachner, M., Barnett, S. C., and Noble, M. (1991). Transplantation of normal and genetically-engineered glia into areas of demyelination. Annal. N. Y. Acad. Sci. 633, 563–565. Crang, A. J., Gilson, J., and Blakemore, W. F. (1998). The demonstration by transplantation of the very restricted remyelinating potential of post-mitotic oligodendrocytes. J. Neurocytol. 27, 541–553. Decker, L., Picard, N., Lachapelle, F., and Baron-Van Evercooren, A. (2001). Neural precursors and demyelinating diseases. Prog. Brain Res. 132, 175–184. Dietrich, J., Noble, M., and Mayer-Proschel, M. (2002). Characterization of A2B5þ glial precursor cells from cyropreserved human fetal brain progenitor cells. Glia 40, 65–77. Duncan, I. D. (1996). Glial cell transplantation and remyelination of the central nervous system. Neuropath. Appl. Neurobiol. 22, 87–100. Duncan, I. D., Aguayo, A. J., Bunge, R. P., and Wood, P. M. (1981). Transplantation of rat Schwann cells grown in tissue culture into mouse spinal cord. J. Neurol. Sci 49, 241–252. Duncan, I. D., and Archer, D. R. (1994). Transplantation of myelinating cells into the central nervous system. In ‘‘A Multidisciplinary Approach to Myelin Disease ll’’ (S. Salvati, ed), pp. 195–206. Plenum Press, New York and London. Duncan, I. D., Grever, W. E., and Zhang, S.-C. (1997). Repair of myelin disease: Strategies and progress in animal models. Mol. Med. Today 3, 554–561. Duncan, I. D., Hammang, J. P., and Gilmore, S. A. (1988a). Schwann cell myelination of the myelin deWcent rat spinal cord following x-irradiation. Glia 1, 233–239. Duncan, I. D., Hammang, J. P., Jackson, K. F., Wood, P. M., Bunge, R. P., and Langford, L. (1988b). Transplantation of oligodendrocytes, and Schwann cells into the spinal cord of the myelin deWcent rat. J. Neurocytol. 17, 351–360. Duncan, I. D., and HoVman, R. L. (1997). Schwann cell invasion of the central nervous system of myelin mutants. J. Anat 180, 35–49. Duncan, I. D., and Milward, E. A. (1995). Glial cell transplants: Experimental therapies of myelin diseases. Brain Pathol. 5, 301–310.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

au4

TARGET DISEASES AND POTENTIAL CLINICAL EFFECTIVENESS

Duncan, I. D., Nadon, N. L., HoVman, R. L., Lunn, K. F., Csiza, C., and Wells, M. R. (1995). Oligodendrocyte survival and function in the long-lived strain of the myelin deWcient rat. J. Neurocytol. 24, 745–762. Duncan, I. D., Paino, C., Archer, D. R., and Wood, P. M. (1992). Functional capacities of transplanted cellsorted adult oligodendrocytes. Dev. Neurosci. 14, 114–122. Emery, E., Li, X., Brunschwig, J.-P., Olson, L., and Levi, A. D. (1999). Assessment of the malignant potential of mitogen stimulated human Schwann cells. J. Peripheral Nerv. Sys. 4, 107–116. Espinosa de los Monteros, A., Baba, H., Zhao, P. M., Pan, T., Chang, R., de Vellis, J., and Ikenaka, K. (2001). Remyelination of the adult demyelinated mouse brain by grafted oligodendrocyte progenitors and the eVect of B-104 cografts. Neurochem. Res. 26, 673–682. Espinosa de los Monteros, A., Bernard, R., Tiller, B., Rouget, P., and de Vellis, J. (1993). Grafting of Fast Blue labeled glial cells into neonatal rat brain: DiVerential survival and migration among cell types. Int. J. Dev. Neurosci. 11, 625–639. Espinosa de los, M. A., Zhao, P., Huang, C., Pan, T., Chang, R., Nazarian, R., Espejo, D., and de Vellis, J. (1997). Transplantation of CG4 oligodendrocyte progenitor cells in the myelin-deWcient rat brain results in myelination of axons and enhanced oligodendroglial markers. J. Neurosci. Res. 50, 872–887. Farkas-Bargeton, E., Robain, O., and Mandel, P. (1972). Abnormal glial maturation in the white matter in Jimpy mice: An optical study. Acta Neuropathol. 21, 272–281. Flax, J. D., Aurora, S., Yang, C., Simonin, C., Wills, A. M., Billinghurst, L. L., Jendoubi, M., Sidman, R. L., Wolfe, J. H., Kim, S. U., and Synder, E. Y. (1998). Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign genes. Nat. Biotechnol. 16, 1033–1039. Franklin, R. J. M. (2002a). Remyelination of demyelinated CNS: The case for and against transplantation of central, peripheral and olfactory glia. Brain Res. Bull. 57, 827–832. Franklin, R. J. M. (2002b). Why does remyelination fail in multiple sclerosis. Nat. Rev. Neurosci. 3, 1–12. Franklin, R. J. M., Bayley, S. A., and Blakemore, W. F. (1996a). Transplanted CG4 cells (an oligodendrocyte progenitor cell line) survive, migrate and contribute to repair of areas of demyelination in X-irradiated and damaged spinal cord, but not in normal spinal cord. Expt. Neurol. 137, 263–276. Franklin, R. J. M., Bayley, S. A., Milner, R., Vrench-Constant, C., and Blakemore, W. F. (1995). DiVerentiation of the O-2A progenitor cell line CG-4 into oligodendrocytes and astrocytes following transplantation into gliadeWcent areas of CNS white matter. Glia 13, 39–44. Franklin, R. J. M., and Blakemore, W. F. (1990). The PNS-CNS regeneration dichotomy: A role for glial cell transplantation. J. Cell Sci. 95, 185–190. Franklin, R. J. M., and Blakemore, W. F. (1993). Requirements for Schwann cell migration within CNS environments: A viewpoint. Int. J. Dev. Neurosci. 11, 641–649. Franklin, R. J. M., and Blakemore, W. F. (1995). Glial cell transplantation and plasticity in the O-2A lineage– Implications for CNS repair. TINS 18, 151–156. Franklin, R. J. M., and Blakemore, W. F. (1997). Transplanting oligodendrocyte progenitors into the adult CNS. J. Anat 190, 23–33. Franklin, R. J. M., and Blakemore, W. F. (1998). Transplanting myelin-forming cells into the CNS–principles and practice. In ‘‘Techniques for the PuriWcation, Functional Evaluation and Transplantation of Brain Cells’’ (R. Rozental, ed.). Methods: A Companion to Methods in Enzymology, Vol. 16, pp. 311–319. Academic Press, New York. Franklin, R. J. M., and Blakemore, W. F. (1999). Myelinogenic cells into the central nervous system. In ‘‘Neural Transplantation Methods’’ (S. B. Dunnett, A. A. Boulton, and G. B. Baker, eds.), Neuromethods, Vol. 36, pp. 305–317. Humana Press, Totowa, NJ. Franklin, R. J. M., Blaschuk, K. L., Bearchell, M. C., Prestoz, L. L., Setzu, A., Brindle, K. M., and VrenchConstant, C. (1999). Magnetic resonance imaging of transplanted oligodendrocyte precursors in the rat brain. NeuroReport 10, 3961–3965. Franklin, R. J. M., Crang, A. J., and Blakemore, W. F. (1991). Transplanted type-1 astrocytes facilitate repair of demyelinating lesions by host oligodendrocytes in adult rat spinal cord. J. Neurocytol. 20, 420–430. Franklin, R. J. M., Crang, A. J., and Blakemore, W. F. (1992). Type-1 astrocytes fail to inhibit Schwann cell remyelination of CNS axons in the absence of cells of the O-2A lineage. Dev. Neurosci. 14, 85–92. Franklin, R. J. M., Gilson, J. M., Franceschini, L. A., and Barnett, S. C. (1996b). Schwann cell-like myelination following transplantation of an olfactory bulb-ensheathing cell line into areas of demyelination in the adult CNS. Glia 17, 217–224. Franklin, R. J. M., Zhao, C., and Sim, F. J. (2002). Astrocyte and macrophage interactions in promoting oligodendrocyte remyelination of the CNS. In ‘‘Glial Interfaces in the Nervous System’’ (H. Aldskogius and J. Fraher, eds), pp. 95–104. IOS Press, Amsterdam. Friedman, E., Nilavar, G., Carmel, P., Perlow, M., Spatz, L., and Latov, N. (1986). Myelination by transplanted fetal and neonatal oligodendrocytes in a dysmyelinating mutant. Brain Res. 378, 142–146. Friedrich, V. L., Jr., and Lazzarini, R. A. (1993). Restricted migration of transplanted oligodendrocytes or their progenitors, revealed by transgenic marker MbP. J. Neural. Transplant. Plast. 4, 139–146. Gansmuller, A., Clerin, E., Kruger, F., Gumpel, M., and Lachapelle, F. (1991). Tracing transplanted oligodendrocytes during migration and maturation in the Shiverer mouse brain. Glia. 4, 580–590. Gansmuller, A., Lachapelle, F., Baron van Evercooren, A., Hauw, J. J., Baumann, N., and Gumpel, M. (1986). Transplantation of newborn CNS fragments into the brain of shiverer mutant mice: Extensive myelination by transplanted oligodendrocytes. II. Electron microscopic study. Dev. Neurosci. 8, 197–207.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

167

168

6. REMYELINATION THROUGH ENGRAFTMENT

Gilson, J., and Blakemore, W. F. (1993). Failure of remyelinationn in areas of demyelination produced in the spinal cord of old rats. Neuropath. Appl. Neurobiol. 19, 173–181. Gilson, J. M., and Blakemore, W. F. (2002). Schwann cell remyelination is not replaced by oligodendrocyte remyelination following ethidium bromide induced demyelination. NeuroReport 13, 1205–1208. Gotow, T., Leterrier, J. F., Ohsawa, Y., Watanabe, T., Isahara, K., Shibata, R., Oikenaka, K., and Uchiyama, Y. (1999). Abnormal expression of neuroWlament protein in dysmyelinated axons located in the central nervous system of jimpy mutant mice. Eur. J. Neurosci. 11, 3893–3903. Gout, O., and Dubois-Dalcq, M. (1993). Directed migration of transplanted glial cells toward a spinal cord demyelinating lesion. Int. J. Dev. Neurosci. 11, 613–623. Gout, O., Gransmuller, A., Baumann, N., and Gumpel, M. (1988). Remyelination by transplanted oligodendrocytes of a demyelinated lesion in the spinal cord of the adult shiverer mouse. Neurosis. Lett. 87, 195–199. Groves, A. K., Barnett, S. C., Franklin, R. J. M., Crang, A. J., Mayer, M., Blakemore, W. F., and Noble, M. (1993). Repair of demyelinated lesions by transplantation of puriWed O-2A progenitor cells. Nature 362, 453– 455. Gumpel, M., Baumann, N., Raoul, M., and Jacque, C. (1983). Survival and diVerentiation of oligodendrocytes from neural tissue transplanted into new-born mouse brain. Neurosis. Lett. 37, 307–311. Gumpel, M., Gout, O., Lubetzki, C., Gransmuller, A., and Baumann, N. (1989). Myelination and remyelination in the central nervous system by transplanted oligodendrocytes using the shiverer model. Dev. Neurosci. 11, 132–139. Gumpel, M., Lachapelle, F., Gansmuller, A., Baron-Van Evercooren, A., Lombrail, P., Gout, O., and Baulac, M. (1988). Myelination capacity of transplanted oligodendrocytes. In ‘‘Multidisciplinary Approach to Myelin Diseases’’ (G. Serlupi and G. Crescenzi, eds.), pp. 343–352. Plenum, New York. Gumpel, M., Lachapelle, F., Gansmuller, A., Baulac, M., Baron-Van Evercooren, A., and Baumann, N. (1987). Transplantation of human embryonic oligodendrocytes into shiverer brain. Annal. N. Y. Acad. Sci. 495, 71–85. Gumpel, M., Lachapelle, F., Jacque, C., and Baumann, N. (1985). Central nervous tissue transplantation into mouse brain: DiVerentiation of myelin from transplanted oligodendrocytes. In ‘‘Neural Grafting in the Mammalian CNS’’ (A. Bjorklund and U. Stenevi, eds.), pp. 151–158. Elsevier, Amsterdam. Hall, S. M. (1972). The eVect of injections of lysophosphatidyl choline into white matter of the adult mouse spinal cord. J. Cell Sci. 10, 535–546. Hammang, J. P., Archer, D. R., and Duncan, I. D. (1997). Myelination following transplantation of EGFresponsive neural stem cells into a myelin-deWcient environment. Exp. Neurol. 147, 84–95. Hanemann, C. O., Rosenbaum, C., Kupfer, S., Wosch, S., Stoegbauer, F., and Muller, H. W. (1998). Improved culture methods to expand Schwann cells with altered growth behaviour from CMT1A patients. Glia 23, 89–98. Hardy, R. J., and Friedrich, V. L., Jr. (1996). Oligodendrocyte progenitors are generated throughout the embryonic mouse brain, but diVerentiate in restricted foci. Development 122, 2059–2069. Harvey, A. R., Fan, Y., Beilharz, M. W., and Grounds, M. D. (1992). Survival and Migration of Transplanted Male Glia in Adult Female Mouse Brains Monitored by a Y-Chromosome-SpeciWc Probe. Mol. Brain. Res. 12, 339–343. Herrera, J., Yang, H., Zhang, S.-C., Pro¨schel, C., Tresco, P., Duncan, I. D., Luskin, M. B., and Mayer-Proschel, M. (2001). Embryonic-derived glial-restricted precursor cells (GRP cells) can diVerentiate into astrocytes and oligodendrocytes in vivo. Expt. Neurol. 171, 11–21. Hinks, G. L., Chari, D. M., O’Leary, M. T., Zhao, C., Keirstead, H. S., Blakemore, W. F., and Franklin, R. J. M. (2001). Depletion of endogenous oligodendrocyte progenitors rather than increased availability of survival factors is a likely explanation for the enhanced survival of transplanted oligodendrocyte progenitors in X-irradiated compared to normal CNS. Neuropath. Appl. Neurobiol. 27, 59–67. Honmou, O., Felts, P. A., Waxman, S. G., and Kocsis, J. D. (1996). Restoration of normal conduction properties in demyelinated spinal cord axonsin the adult rat by transplantation of exogenous Schwann cells. J. Neurosci. 16, 3199–3208. Huang, P. P., Alliquant, B., Carmel, P. W., and Friedman, E. D. (1991). Myelination of the Rat Retina by Transplantation of Oligodendrocytes into Four-Day-Old Hosts. Exp. Neurol 113, 291–300. Imaizumi, T., Lankford, K. L., Baron-Van Evercooren, A., and Kocsis, J. D. (1999). Electrophysiological properties of spinal cord axons remyelinated by various cell transplantation conditions. J. Neurosci. 24, 737. Imaizumi, T., Lankford, K. L., Burton, W. V., Fodor, W. L., and Kocsis, J. D. (2000). Xenotransplantation of transgenic pig olfactory ensheathing cells promotes axonal regeneration in rat spinal cord. Nat. Biotechnol. 18, 949–953. Imaizumi, T., Lankford, K. L., Waxman, S. G., Greer, C. A., and Kocsis, J. D. (1998). Transplanted olfactory ensheathing cells remyelinate and enhance axonal conduction in the demyelinated dorsal columns of the rat spinal cord. J. Neurosci. 18, 6176–6185. Isacson, O., Deacon, T. W., Pakzaban, P., Galpern, W. R., Dinsmore, J., and Burns, L. H. (1995). Transplanted xenogeneic neural cells in neurodegenerative disease models exhibit remarkable axonal target speciWcity and distinct growth patterns of glial and axonal Wbres. Nature Med. 1, 1189–1194. Iwashita, Y., and Blakemore, W. F. (2000). Areas of demyelination do not attract signiWcant numbers of Schwann cells transplanted into normal white matter. Glia 31, 232–240. Iwashita, Y., Crang, A. J., and Blakemore, W. F. (2000a). Redistribution of bisbenzimide Hoechst 33342 Xurochrome from transplanted cells to host cells. NeuroReport 11, 1013–1016.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

au5

TARGET DISEASES AND POTENTIAL CLINICAL EFFECTIVENESS

Iwashita, Y., Fawcett, J. W., Crang, A. J., Franklin, R. J. M., and Blakemore, W. F. (2000b). Schwann cells transplanted into normal and x-irradiated adult white matter do not migrate extensively and show poor longterm survival. Expt. Neurol. 164, 280–302. Jasmin, L., Janni, G., Moallem, T. M., Lappi, D. A., and Ohara, P. T. (2000). Schwann cells are removed from the spinal cord after eVecting recovery from paraplegia. J. Neurosci. 20, 9215–9223. JeVery, N. D., and Blakemore, W. F. (1997). Locomotor deWcits induced by experimental spinal cord demyelination are abolished by spontaneous remyelination. Brain 120, 27–37. JeVery, N. D., Crang, A. J., O’Leary, M. T., Hodge, S. J., and Blakemore, W. F. (1999). Behavioural consequences of oligodendrocyte progenitor cell transplantation into experimental demyelinating lesions in rat spinal cord. Eur. J. Neurosci. 11, 1508–1514. Jiang, Y., Jahagridar, B. N., Reinhard, R. L., Schwartz, R. E., Keene, C. D., Ortiz-Gonzarlez, X., Reyes, M., Lenvik, T., Lund, T., Blackstad, M., Du, J., Aldrich, S., Lisberg, A., Low, W. G., Largaespada, D. A., and Verfalllle, C. M. (2002). Pluripotency of mesenchymal stem cells derived from adult marrow. Nature 41–49. Jung, M., Kramer, E., Grzenkowski, M., Tang, K., Blakemore, W. F., Aguzzi, A., Kaazaie, K., Chlichlia, K., Blankenfeld, G. v., Kettenmann, H., and Trotter, J. (1995). Lines of murine oligodendroglial precursor cells immortalised by an activated neu-tyrosine kinase show distinct degrees of interaction with axons in vitro and in vivo. Eur. J. Neurosci. 7, 1245–1265. Kagawa, T., Ikenaka, K., Inoue, Y., Kuriyama, S., Tsujii, T., Nakao, J., Nakajima, K., Aruga, J., Okano, H., and Mikoshiba, K. (1994). Glial cell degeneration and hypomyelination caused by overexpression of myelin proteolipid protein gene. Neuron 13, 427–442. Kato, T., Honmou, O., Uede, T., Hashi, K., and Kocsis, J. D. (2000). Transplantation of human olfactory ensheathing cells elicits remyelination of demyelinated rat spinal cord. Glia 30, 209–218. Keirstead, H. S., Ben-Hur, T., Rogister, B., O’Leary, M. T., Dubois-Dalcq, M., and Blakemore, W. F. (1999). PSA-NCAM positive CNS precursors generate both oligodendrocytes and Schwann cells to remyelinate the CNS following transplantation. J. Neurosci. 19, 7529–7536. Keirstead, H. S., and Blakemore, W. F. (1997). IdentiWcation of post-mitotic oligodendrocytes incapable of remyelination within the demyelinated adult spinal cord. J Neuropath. Exp. Neurol 56, 1191–1201. Keyoung, H. M., Roy, N. S., Benraiss, A., Louissaint, A. J., Suzuki, A., Hashimoto, M., Rashbaum, W. K., Okano, H., and Goldman, S. A. (2001). High-yield selection and extraction of two promotor-deWned phenotypes of neural stem cells from the fetal human brain. Nat. Biotechnol. 19, 843–850. Kocsis, J. D. (1999). Restoration of function by glial cell transplantation. J. Neurotrauma 16, 695–703. Kohama, I., Lankford, K. L., Preiningerova, J., White, F. A., Vollmer, T. L., and Kocsis, J. D. (2001). Transplantation of cryopreserved adult human Schwann cells enhances axonal conduction in demyelinated spinal cord. J. Neurosci. 21, 944–950. Kohsaka, S., Yoshida, K., Inoue, Y., Shinozaki, T., Takayama, H., Inoue, H., Mikoshiba, K., Takamata, K., Otani, M., Toya, S., and Tsakada, Y. (1986). Transplantation of bulk-separated oligodendrocytes into the brains of shiverer mutant mice: An immunohistochemical and electron microscopic studies on myelination. Brain Res. 372, 137–142. Kohyama, J., Abe, H., Shimazaki, T., Koizumi, A., Nakashima, K., Gojo, S., Taga, T., Okano, H., Hata, J., and Umezawa, A. (2001). Brain from bone: EYcient ‘‘meta-diVerentiation’’ of marrow stroma-derived mature osteoblasts to neurons with Noggin or a demethylating agent. DiVerentiation 68, 235–244. Kopen, G. C., Prockop, D. J., and Phinney, D. G. (1999). Marrow stromal cells migrate throughout forebrain and cerebellum, and they diVerentiate into astrocytes after injection into neonatal mouse brains. Proc. Nat. Acad. Sci. USA 96, 10711–10716. Kwiecien, J. M., O’Connor, L. T., Goetz, B. D., Delaney, K. H., Fletch, A. L., and Duncan, I. D. (1998). Morphological and morphometric studies of the dysmyelinating mutant, the Long Evans shaker rat. J. Neurocytol. 27, 581–591. Lachapelle, F. (1995). Glial transplants: An in vivo analysis of extrinsic and intrinsic determinants of dysmyelination in genetic variants. Brain Pathol. 5, 289–299. Lachapelle, F., Avellana-Adalid, V., Nait-Oumesmar, B., and Baron-Van Evercooren, A. (2002a). Fibroblast Growth Factor-2 (FGF-2) and Platelet-Derived Growth Factor AB (PDGF(AB) ) Promote Adult SVZDerived Oligodendrogenesis in Vivo. Mol. Cell Neurosci. 20, 390–403. Lachapelle, F., Bachelin, C., Moissonnier, P., Avellana-Adalid, V., and Baron-Van Evercooren, A. (2002b). Autologous Schwann cell (SC) engrafment in the demyelinated adult macaque CNS. Glia Suppl. May, p357. Lachapelle, F., Bachelin, C., Moissonnier, P., Avellana-Adalid, V., Fontaine, D., Seilhan, T. M., and Baron-Van Evercooren, A. (2000). Transplantation of autologous Schwann cells to repair demyelinating lesions in the optic nerve and spinal cord. J. Neurosci. 24, 1379. Lachapelle, F., Duhamel-Clerin, E., Gansmu¨ller, A., Baron-Van Evercooren, A., Villarroya, H., and Gumpel, M. (1994a). Transplanted transgenically marked oligodendrocytes survive, migrate and myelinate in the normal mouse brain as they do in the shiverer mouse brain. Eur. J. Neurosci. 6, 814–824. Lachapelle, F., Gansmuller, A., Baulac, M., Baron-Van Evercooren, A., Baumann, N., and Gumpel, M. (1989). Transplantation of oligodendrocytes in the CNS. Acta Cytobiol, Bulgarian Acad. Sci. 1, 24–30. Lachapelle, F., Gumpel, M., Baulac, M., Jacque, C., Duc, P., and Baumann, N. (1984). Transplantation of CNS fragments into the brain of shiverer mutant mice: Extensive myelination by implanted oligodendrocytes. Dev. Neurosci. 6, 325–334. Lachapelle, F., Gumpel, M., and Baumann, N. (1994b). Contribution of transplantations to the understanding of the role of the PLP gene. Neurochem. Res. 19, 1083–1090.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

169

170

6. REMYELINATION THROUGH ENGRAFTMENT

Lachapelle, F., Lapie, P., Campagnoni, A. T., and Gumpel, M. (1991). Oligodendrocytes of the Jimpy phenotype can be partially restored by environmental factors invivo. J. Neurosci. Res. 29, 235–243. Lachapelle, F., Lapie, P., Gansmuller, A., Villarroya, H., and Gumpel, M. (1990). Patchy myelination pattern in the Jimpy mouse brain–immunohistochemical study. Glia 3, 375–384. Laeng, P., Molthagen, M., Yu, E. G. X., and Bartsch, U. (1996). Transplantation of oligodendrocyte progenitor cells into the rat retina: Extensive myelination of retinal ganglion cell axons. Glia 18, 200–210. Langford, L. A., and Owens, G. C. (1990). Resolution of the pathway taken by implanted Schwann cells to a spinal cord lesion by prior infection with a retrovirus encoding beta-galactosidase. Acta Neuropathol. 80, 514–520. Lankford, K. L., Imaizumi, T., Honmou, O., and Kocsis, J. D. (2002). A quantitative morphometric analysis of rat spinal cord remyelination following transplantation of allogenic Schwann cells. J. Comp. Neurol. 443, 259–274. Le Gros Clark, W. E. (1943). The problem of neuronal regeneration in the central nervous sytem. II The insertion of peripheral nerve stumps into the brain. J. Anat 77, 251–259. Learish, R. D., Bru¨stle, O., Zhang, S. C., and Duncan, I. D. (1999). Intraventricular transplantation of oligodendrocyte progenitors into a fetal myelin mutant results in widespread formation of myelin. Annal. Neurol. 46, 716–722. Levi, A. D. O. (1996). Characterization of the technique involved in isolating Schwann cells from adult human peripheral nerve. J. Neurosci. Methods 68, 21–26. Levi, A. D. O., Gue´nard, V., Aebischer, P., and Bunge, R. P. (1994). The functional characteristics of Schwann cells cultured from human peripheral nerve after transplantation into a gap within the rat sciatic nerve. J. Neurosci. 14, 1309–1319. Li, G., Crang, A. J., Rundle, J. L., and Blakemore, W. F. (2002). Oligodendrocyte progenitor cells in the adult rat CNS express myelin oligodendrocyte glycoprotein (MOG). Brain Pathol. 463–471. Li, Y., and Raisman, G. (1997). Integration of transplanted cultured Schwann cells into the long myelinated Wber tracts of the adult spinal cord. Exp. Neurol.. 145, 397–411. Liblau, R., Fontaine, B., Baron, A., Wekerle, and H. Lassmann. (2001). Demyelinating diseases: From pathogenesis to repair strategies. Trends Neurosci. 24, 133–134. Liu, S., Qu, Y., Stewart, T. J., Howard, M. J., Chakraborty, S., Hollekamp, T. F., and Mc Donald, J. W. (2000).Embryonic stem cells diVerentiate into oligodendrocytes and myelinate in culture and after spinal cord transplantation. Proc. Nat. Acad. Sci. USA. 97 6126–6131. Lobsiger, C. S., Smith, P. M., Butchstaller, J., Schweitzer, B., Franklin, R. J. M., Suter, U., and Taylor, V. (2001). SpL201: A conditionally immortalized Schawn cell precursor line that generates myelin. Glia 36, 31–47. Lubetzki, C., Gansmuller, A., Lachapelle, F., Lombrail, P., and Gumpel, M. (1988). Myelination by oligodendrocytes isolated from 4–6-week-old rat central nervous system and transplantation into newborn shiverer brain. J. Neurol. Sci 88, 161–175. Mc Donald, J. W., Liu, X. Z., Qu, Y., Liu, S., Mickey, S. K., Turesky, D., Gottlieb, D. I., and Choi, D. W. (1999).Transplanted embryonic stem cells survive, diVerentiate and promote recovery in injured rat spinal cord. Nature Med. 5, 1410–1412. Mezey, E., Chandross, K. J., Harta, G., Maki, R. A., and McKercher, S. R. (2000). Turning blood into brain: Cells bearing neuronal antigens generated in vivo from bone marrow. Science 290, 1779–1782. Milward, E. A, Lundborg, G. C., Ge, B., Lipsitz, D., Zhao, M., and Duncan, I. D. (1997). Isolation and transplantation of multipotential populations of epidermal growth factor-responsive, neural progenitor cells from the canine brain. J. Neurosi. Res. 50, 862–871. Milward, E. A., Zhang, S. C., Zhao, M., Lundberg, C., Ge, B., Goetz, B. D., and Duncan, I. D. (2000). Enhanced proliferation and directed migration of oligodendroglial progenitors co-grafted with growth factor-secreting cells. Glia 32, 264–270. Mitome, M., Low, H. P., Van den Pol, A., Nunnari, J. J., Wolf, M. K., Billings-Gagliardi, S., and Schwartz, W. J. (2001a). Towards the reconstruction of central nervous system white matter using neural precursor cells. Brain 124, 2147–2161. Mujtaba, T., Mayer-Proschel, M., and Rao, M. S. (1998). A common neural progenitor for the CNS and PNS. Dev. Biol. 200, 1–15. Murray, K., and Dubois-Dalcq, M. (1997). Emergence of oligodendrocytes from human neural spheres. J. Neurosis. Res. 50, 146–156. Nadon, N. L., Duncan, I. D., and Hudson, L. D. (1990). A point mutation in the proteolipid protein gene of the shaking pup interrupts oligodendrocyte development. Development 110, 529–537. Nait-Oumesmar, B., Lachapelle, F., Decker, L., Avellana-Adalid, V., and Vitry, S. (1998). Candidate cells for remyelinating the central nervous system. Rev Neurol (Paris) 154, 592–599. Nakano, K., Migita, M., Mochizuki, H., and Shimada, T. (2001). DiVerentiation of transplanted bone marrow cells in the adult mouse brain. Transplantation 71, 1735–1740. Nakatsuji, Y., and Miller, R. H. (2001). Control of oligodendrocyte precursor proliferation mediated by densitydependent cell cycle protein expression. Dev. Neurosci. 23, 356–363. Nussbaum, J. L., and Mandel, P. (1973). Brain proteolipids in neurological mutant mice. Brain Res. 61, 295–310. O’Connor, L. T., Goetz, B. D., Kwiecien, J. M., Delaney, K. H., Fletch, A. L., and Duncan, I. D. (1999). Insertion of a retrotransposon in mbp disrupts mRNA splicing and myelination in a new mutant rat. J. Neurosci. 19, 3404–3413. O’Leary, M. T., and Blakemore, W. F. (1997a). Oligodendrocyte precursors survive poorly and do not migrate following transplantation into the normal adult central nervous system. J. Neurosci. Res. 48, 159–167.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

TARGET DISEASES AND POTENTIAL CLINICAL EFFECTIVENESS

O’Leary, M. T., and Blakemore, W. F. (1997b). Use of a rat Y-chromosome probe to determine the long term fate of glial cells transplanted into areas of central nervous system demyelination. J. Neurocytol. 26, 191–206. Pagano, S. F., Impagnatiello, F., Girelli, M., Cova, L., Grioni, E., Onofri, M., Cavallaro, M., Etteri, S., Vitello, F., Giombini, S., Solero, C. L., and Parati, E. A. (2000). Isolation and chaaracterization of neural stem cells from the adult human olfactory bulb. Stem Cells 18, 295–300. Palmer, T. D., Schwartz, P. H., Taupin, P., Kasper, B., Stein, S. A., and Gage, F. H. (2001). Cell culture. Progenitor cells from human brain after death. Nature 411, 42–43. Perry, V. H., and Hayes, L. (1985). Lesion-induced myelin formation in the retina. J neurocytol 14, 297–307. Pluchino, S., and Brambilla, E. (2002). Intrathecal administration of multipotential stem cells ameliorates disease progression in chronic experimental autoimmune disease. J. Neurol 249, I/19. Pluchino, S., Quattrini, A., Brambilla, E., Gritta, A., Salani, G., Dina, G., Galli, R., Del Carra, U., Amadio, S., Bergami, A., Furlan, R., Comi, G., Vescovi, A. L., and Martino, G. (2003). Injection of adult neurospheres induces recovery in a chronic model of multiple sclerosis. Nature 122, 688–694. Pressmar, S., Ader, M., Richard, G., Schachner, M., and Bartsch, U. (2001). The fate of heterotopically grafted neural precursor cells in the normal and dystrophic adult mouse retina. Invest Ophthalmol. Vis. Sci. 42, 3311–3319. Quinn, S. M., Walters, W. M., Vescovi, A. L., and Whittemore, S. R. (1999). Lineage restriction of neuroepithelial precursor cells from fetal human spinal cord. J. Neurosis. Res. 57, 590–602. ReubinoV, B. E., Itsykson, P., Turetsky, T., Pera, M. F., Reinhartz, E., Itzik, A., and Ben-Hur, T. (2001). Neural progenitors from human embryonic stem cells. Nat. Biotechnol. 19, 1134–1140. Rosenbluth, J., Hasegawa, M., and SchiV, R. (1989). Myelin formation in myelin-deWcient rat spinal cord following transplantation of normal fetal spinal cord. Neurosis. Lett. 97, 35–40. Rosenbluth, J., Hasegawa, M., Shirasaki, N., Rosen, C. L., and Liu, Z. (1990). Myelin formation following transplantation of normal fetal glia into myelin deWcent rat spinal cord. J. Neurocytol. 19, 718–730. Rosenbluth, J., Liu, Z., Guo, D., and SchiV, R. (1993). Myelin formation by mouse glia in myelin-deWcient rats treated with cyclosporine. J. Neurocytol. 22, 967–977. Rosenfeld, J., and Friedrich, V. L. (1983). Axonal swellings in jimpy mice: Does lack of myelin cause neuronal abnormalities. Neurosci. 10, 959–966. Roy, N. S., Wang, S., Harrison-Restelli, C., Benraiss, A., Fraser, R. A., Gravel, M., Braun, P. E., and Goldman, S. A. (1999). IdentiWcation, isolation, and promoter-deWned separation of mitotic oligodendrocyte progenitor cells from the adult human subcortical white matter. J. Neurosci. 19, 9986–9995. Roy, N. S., Wang, S., Ljiang, L., Kang, J., Bentaiss, A., Harrison-Restelli, C., Fraser, R. A., Couldwell, W. T., Kawaguchi, A., Okano, H., Nedergaard, M., and Goldman, S. A. (2000). In vitro neurogenesis by progenitor cells isolated from the adult human hippocampus. Nature Med. 6, 271–277. Rutkowski, J. L., Kirk, C. J., Lerner, M. A., and Tennekoon, G. I. (1995). PuriWcation and expansion of human Schwann cells in vitro. Nature Med. 1, 80–83. Sanchez-Ramos, J., Song, S., Cardozo-Pelaez, F., Hazzi, C., Stedeford, T., Willing, A., Freeman, T. B., Saporta, S., Janssen, W., Patel, N., Cooper, D. R., and Sanberg, P. R. (2000). Adult bone marrow stromal cells diVerentiate into neural cells in vitro. Exp. Neurol 164, 247–256. Sasaki, M., Honmou, O., Akiyama, Y., Uede, T., Hashi, K., and Kocsis, J. D. (2001). Transplantation of an acutely isolated bone marrow fraction repairs demyelinated adult rat spinal cord axons. Glia 35, 26–34. Seilhean, D., Gansmu¨ller, A., Baron-Van Evercooren, A., Gumpel, M., and Lachapelle, F. (1996). Myelination by transplanted human and mouse central nervous system tissue after long-term cryopreservation. Acta Neuropathol. 91, 82–88. Shields, S. A., Blakemore, W. F., and Franklin, R. J. M. (2000). Schwann cell remyelination is restricted to astrocyte-deWcient areas following transplantation into demyelinated rat brain. J. Neurosi. Res. 60, 571–578. Shields, S. A., Gilson J. M, Blakemore, W. F., and Franklin, R. J. M. (1999). Remyelination occurs as extensively but more slowly in old rats compared to young rats following glyotoxin-induced CNS demyelination. Glia 28, 77–83. Sim, F. J., Zhao, C., Penderis, J., and Franklin, R. J. M. (2002). The age-related decrease in CNS remyelination eYcency is attribable to an ompairment of both oligodendrocyte progenitor recruitment and diVerentiation. J. Neurosci. 22, 2451–2459. SkoV, R. P. (1976). Myelin deWcit in the jimpy mouse may be due to cellular abnormalities in astroglia. Nature 264, 560–562. Smith, P. M., and Blakemore, W. F. (2000a). Porcine neural progenitors require commitment to the oligodendrocyte lineage prior to transplantation in order to achieve signiWcant remyelination of demyelinated lesions in the adult CNS. Eur. J. Neurosci. 12, 2414–2425. Smith, P. M., and Blakemore, W. F. (2000b). he formation of new myelin in the CNS by transplantation. Neurosci. News 3, 60–67. Svendsen, C. N., Caldwell, M. A., ter-Borg, M. G., Rosser, A. E., Tyers, P., Karmiol, S., and Dunnett, S. B. (1997). Long-term survival of human central nervous sytem progenitor cells transplanted into a rat model of Parkinson’s disease. Expt. Neurol. 148, 135–146. Svendsen, C. N., ter Borg, M. G., Armstrong, R. J., Rosser, A. E., Chandran, S., Ostenfeld, T., and Caldwell, M. A. (1998). A new method for the rapid and long term growth of human neural precursor cells. J. Neurosci. Meth. 85, 141–152. Takami, T., Oudega, M., Bates, M. L., Wood, P. M., Kleitman, N., and Bunge, M. B. (2002). Schwann cell but not olfactory ensheathing glia transplants improve hindlimb locomotor performance in the moderately contused adult rat thoracic spinal cord. J. Neurosci. 22, 6670–6681.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

171

172

6. REMYELINATION THROUGH ENGRAFTMENT

Targett, M. P., and Blakemore, W. F. (1994). The use of xenografting to evaluate the remyelinating potenial of glial cell cultures. Eye 8, 238–244. Targett, M. P., Sussman, J., O’Leary, M. T., Compston, D. A. S., and Blakemore, W. F. (1996). Failure to achieve remyelination of demyelinated rat axons following transplantation of glial cells obtained from the adult human brain. Neuropath. Appl. Neurobiol. 22, 199–206. Tello, T. (1911). La inXuencia del neurotropismo en la regeneracion de los centros nervisos. Trab, Lab. Invest. Biol. Univ. Madr. 9, 123–159. Tontsch, U., Archer, D. R., Dubois-Dalcq, M., and Duncan, I. D. (1994). Transplantation of an oligodendrocyte cell line leading to extensive myelination. Proc. Natl. Acad. Sci. USA 91, 11616–11620. Tourbah, A., Linnington, C., Bachelin, C., Avellana-Adalid, V., Wekerle, H., and Baron-Van Evercooren, A. (1997). InXammation promotes survival and migration of the CG4 oligodendrocyte progenitors transplanted in the spinal gord of both inXammatory and demyelinated EAE rats. J. Neurosci. Res. 50, 853–861. Trotter, J., Crang, A. J., Schachner, M., and Blakemore, W. F. (1993). Lines of glial precursor cells immortalised with a temperature-sensitive oncogene give rise to astrocytes and myelin-forming oligodendrocytes on transplanation into demyelinated lesions in the central nervous system. Glia 9, 25–40. Utzschneider, D. A., Archer, D. R., Kocsis, J. D., Waxman, S. G., and Duncan, I. D. (1994). Transplantation of glial cells enhances action potential conduction of amyelinated spinal cord axons in the myelin-deWcient rat. Proc. Nat. Acad. Sci. USA 91, 53–57. Van der Goes, A., and Dijkstra, C. D. (2001). Models for demyelination. Prog. Brain Res. 132, 149–163. Vescovi, A., Parati, E. A., Gritti, A., Poulin, P., Ferrario, M., Wanke, E., Frolichsthal-Schoeller, P., Cova, L., Arcellana-Panlilo, M., Colombo, A., and Galli, R. (1999). Isolation and cloning of multipotential stem cells from the embryonic human CNS and establishment of transplantatable human neural stem cell lines by epigenetic stimulation. Expt. Neurol. 156, 71–83. Vescovi, A., and Synder, E. Y. (1999). Establishment and properties of neural stem cell clones: Plasticity in vitro and in vivo. Brain Pathol. 9, 569–598. Vignais, L., Nait Oumesmar, B., Mellouk, F., Gout, O., Labourdette, G., Baron-Van Evercooren, A., and Gumpel, M. (1993). Transplantation of oligodendrocyte precursors in the adult demyelinated spinal cord: Migration and remyelination. Int. J. Dev. Neurosci. 11, 603–612. Villa, A., Snyder, E. Y., Vescovi, A., and Martı´nez-Serrano, A. (2000). Establishment and properties of a growth factor-dependent, perpetual neural stem cell line from the human CNS. Expt. Neurol. 161, 67–84. Vitry, S., Avellana-Adalid, V., Hardy, R., Lachapelle, F., and Baron-Van Evercooren, A. (1999). Mouse oligospheres: From pre-progenitors to functional oligodendrocytes. J. Neurosci Res. 58, 735–751. Vitry, S., Avellana-Adalid, V., Lachapelle, F., and Baron-Van Evercooren, A. (2001). Migration and multipotentiality of PSA-NCAMþ neural precursors transplanted in the developing brain. Mol. Cell. Neurosci. 17, 983–1000. Wagers, A. J., Sherwood, R. I., Christensen, J. L., and Weissman, I. L. (2002). Little evidence for developmental plasticity of adult hematopoietic stem cells. Science 297, 2256–2259. Warrington, A. E., Barbarese, E., and PfeiVer, S. E. (1993). DiVerential myelinogenic capacity of speciWc developmental stages of the oligodendrocyte lineage upon transplantation into hypomyelinating hosts. J. Neurosci. Res. 34, 1–13. Windrem, M. S., Roy, N. S., Wang, J., Nunes, M., Benraiss, A., Goodman, R., McKhann II, G. M., and Goldman, S. A. (2002). Progenitor cells derived from the adult human subcortical white matter disperse and diVerentiate as oligodendrocytes within demyelinated lesions of the rat brain. J. Neurosci. Res. 69, 966–675. Wolswijk, G. (1998). Chronic stage multiple sclerosis lesions contain a relatively quiescent population of oligodendrocyte precursor cells. J. Neurosci. 18, 601–609. WoodruV, R. H., and Franklin, R. J. M. (1999). Demyelination and remyelination of the caudal cerebellar peduncle of adult rats following stereotaxic injection of lysolecithin, ethidium bromide and complement/antigalactocerebroside–a comparative study. Glia 25, 216–228. Wyse, J. P. H., and Spira, A. W. (1981). Ultrastructural evidence of a peripheral nervous system pattern of myelination in the avascular retina of the guina pig. Acta Neuropathol. 54, 203–210. Yandava, B. D., Billinghurst, L. L., and Snyder, E. Y. (1999). ‘‘Global’’ cell replacement is feasible via neural stem cell transplantation: Evidence from the dysmyelinated shiverer mouse brain. Proc. Nat. Acad. Sci. USA 96, 7029–7034. Zhang, S.-C., and Duncan, I. D. (2000). Remyelination and restoration of axonal function by glial cell transplantation. In ‘‘Functional Neural Transplantation II: Novel Cell Therapies for CNS Disorders’’ (S. Dunnett and A. Bjorklund, eds). pp. 515–533. Elsevier Science, B. V. Zhang, S.-C., Lundberg, C., Lipsitz, D., O’Connor, L. T., and Duncan, I. D. (1998a). Generation of oligodendroglial progenitors from neural stem cells. J. Neurocytol. 27, 475–489. Zhang, S.-C., Wernig, M., Duncan, I. D., Brustle, O., and Thomson, J. A. (2001). In vitro diVerentiation of transplantable neural precursors from embryonic stem cells. Nat. Biotechnol. 19, 1129–1133. Zhang, S. C., Ge, B., and Duncan, I. D. (1999). Adult brain retains the potential to generate oligodendroclial progenitors with extensive myelination capacity. Proc. Nat. Acad. Sci. USA 96, 4089–4094. Zhang, S. C., Lipsitz, D., and Duncan, I. D. (1998b). Self-renewing canine oligodendroglial progenitor expanded as oligospheres. J. Neurosci. Res. 54, 181–190. Zigova, T., Song, S., Willing, A. E., Hudson, J. E., Newman, M. B., Saporta, S., Sanchez-Ramos, J., and Sanberg, P. R. (2002). Human umbilical cord blood cells express neural antigens after transplantation into the developing rat brain. Cell Transplant. 11, 265–274.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

C H A P T E R

7 Remyelination by Endogenous Glia Robin J. M. Franklin and James E. Goldman

INTRODUCTION Demyelination in the central nervous system (CNS) may be followed by a spontaneous regenerative process termed remyelination. This process involves reinvesting demyelinated axons with new myelin sheaths (or internodes) and has been described in a number of experimental models as well as in naturally occurring demyelinating disease, most notably the acute lesions of multiple sclerosis (MS), the most widely occurring primary demyelinating disease in humans (Lassmann et al., 1997; Ludwin, 1988) (Fig. 7.1). Remyelination allows the axon to transmit action potentials by saltatory conduction, a property that is lost in demyelination (Smith et al., 1979) (Fig. 7.2A). Although lengths of demyelinated axon may acquire the ability to conduct by continuous or nonsaltatory conduction, this form of conduction is less rapid and more vulnerable to small changes in the environment, such as temperature, than the saltatory conduction permitted by remyelination. Moreover, some demyelinated axons do not conduct action potentials due to an imbalance between densities of sodium and potassium channels (Smith and McDonald, 1999). Several lines of evidence indicate that remyelination is an important contributor to recovery of functional deWcits that arise due to demyelination (JeVery and Blakemore, 1997) (Fig. 7.2B) and may also play a role in protecting axons from the atrophy to which they are vulnerable in inXammatory demyelinating disease (Kornek et al., 2000). Its unusual status as a spontaneous regenerative process in the CNS and its role in recovery of function make remyelination a particularly fascinating area of myelin biology from both purely scientiWc and clinical perspectives.

MORPHOLOGICAL FEATURES OF REMYELINATION A characteristic feature of the morphology of remyelination is that the new myelin sheath is invariably thinner and shorter than would be expected for the diameter of the axon. A degree of modeling of the sheath goes on for several months, but the original myelin sheath/axon relationship is never regained (Blakemore, 1974b; Ludwin and Maitland, 1984). This feature of remyelination accounts for the paler staining quality of areas of remyelination compared to areas of normal myelination observed using many standard histological stains for myelin such as luxol fast blue, solochrome cyanine, and toluidine blue. The thin myelin sheaths are especially apparent in transverse semithin plastic (resin) sections counterstained with toluidine blue and viewed by light or electron microscopy (Fig. 7.3). In normal myelination there is generally a close relationship between the axon diameter and the myelin internode length and thickness, both increasing with increasing axonal diameter (Murray and Blakemore, 1980). In remyelination this relationship

Myelin Biology and Disorders, Volume 1

173

Copyright 2004, Elsevier (USA). All rights reserved.

174

7. REMYELINATION BY ENDOGENOUS GLIA

* *

*

FIGURE 7.1 Section from the forebrain of a patient who died with multiple sclerosis, stained with luxol fast blue to reveal the areas of myelination in the subcortical white matter. The asterisks indicate three areas from which the myelin stain is absent and indicate three areas of demyelination. The arrows indicate ‘‘Shadow plaques’’ are visible where the demyelinated axons have undergone remyelination.

a

b

FIGURE 7.2 (A) Restoration of secure conduction following CNS remyelination. This Wgure shows saphenous nerve compound action potentials at diVerent times during a lysolecithin-induced demyelinating/remyelinating lesion in the dorsal funiculus of a cat. From Smith et al., 1979. (B) Recovery from demyelination-induced locomotor deWcits by remyelination. These Wgures are taken from a study in which ethidium bromide lesions are induced in the dorsal funiculus of rats trained to walk across a beam. Following demyelination the rats exhibit locomotor deWcits that are scored (the higher the score the greater the degree of deWcit). Following demyelination animals recover their locomotor function with a time course commensurate with remyelination of the lesion (a). The belief that this recovery is due to remyelination is lent support by the failure of recovery if remyelination is blocked by X-irradiation (b). From JeVery and Blakemore, 1997.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

MORPHOLOGICAL FEATURES OF REMYELINATION

a

b

c FIGURE 7.3 Endogenous oligodendrocyte remyelination in experimental models of demyelination. (A) PLP mRNA expressing oligodendrocytes detected by in situ hybridization in an area of remyelination following lysolecithin-induced demyelination in spinal cord white matter of an adult rat. Note the higher density of PLPþ oligodendrocytes within the remyelinated area (dotted outline) than in the surrounding normal white matter. (B) Extensive area of oligodendrocyte remyelination following ethidium bromide-induced demyelination of the deep cerebellar white matter of an adult rat, demonstrable by toluidine blue staining of a semithin resin section. (C) Electronmicrograph of oligodendrocyte remyelination in the murine TMEV model of demyelination and remyelination. An oligodendrocyte cell body is indicated (O). Image kindly provided by Dr. Moses Rodriguez. Scale bar represents approximately 30 mm in A and B and approximately 4 mm in C.

becomes less distinct, the internodal dimensions being considerably less with increasing axonal diameter (Blakemore and Murray, 1981; Miller and Rodriguez, 1995). The identiWcation of an abnormally thin myelin sheath remains the most reliable means of unequivocally demonstrating remyelination, particularly if it can also be demonstrated that demyelination has previously occurred. This is easier to demonstrate in experimental models than in clinical material. Other approaches that have been used to demonstrate remyelination include the expression of transcripts associated with myelin sheath formation such as MBP exon 2 (Capello et al., 1997; Jordan et al., 1990), which has relatively high levels of expression during myelin sheath formation compared to other MBP splice variants (Jordan et al., 1989). However, in a study on MBP mRNA expression during remyelination of toxin-induced demyelination in the rat it was found that the expression pattern of MBP exon 2 simply reXected the overall changes in all MBP transcripts (WoodruV and Franklin, 1999b), suggesting that at least in the rat MBP exon 2 expression proWles may not speciWcally indicate remyelination. It remains the case, however, that there are no surrogate markers that enable remyelinated tissue to be distinguished from normally myelinated tissue and that veriWcation relies on the demonstration of thin myelin sheaths. This becomes easier to recognize the larger the diameter of axon. With small diameter axons the diVerence in a normal myelin sheath and a thin sheath of remyelination can be very marginal and diYcult to distinguish even with morphometric analyses such as the measurement of g-ratios (the ratio of axonþmyelin sheath diameter to axon diameter).

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

175

176

7. REMYELINATION BY ENDOGENOUS GLIA

This makes identiWcation of remyelination easier in some white matter tracts than others. For example, remyelination is more readily identiWable following demyelination of the large diameter Wbers of the cerebellar peduncles (Blakemore, 1974b; WoodruV and Franklin, 1999a) or several of the spinal cord tracts (Blakemore et al., 1977; Hall, 1972; Miller and Rodriguez, 1995) than following demyelination of the small diameter Wbers of the corpus callosum (Mason et al., 2001a). The reason why the myelin sheath of remyelination should be diVerent from the myelin sheath of myelination has never been satisfactorily explained. A proposed explanation is that whereas in development myelinating oligodendrocytes associate with axons that are still in the dynamic phase of growth, in remyelination the myelinating cells engage a mature axon of established diameter (Franklin and Hinks, 1999). Thus, in developmental myelination the oligodendrocyte responds to the increasing area of an expanding axon by appropriately regulating the amount of myelin it generates, resulting in a close association between the axolemmal area beneath the myelin sheath and the length and thickness of the internode (Smith et al., 1982). By contrast, the remyelinating oligodendrocyte engages an axon whose diameter will remain largely static, the dimensions of the myelin sheath thus representing a sheath of default dimensions that are similar regardless of the diameter of the demyelinated axon. Conceptually attractive though this explanation might be, the situation is clearly more complex since axon diameter may change during demyelination and remyelination (Mason et al., 2001a) and is modulated by a complex interaction between both intrinsic and glia-induced factors (Colello et al., 1994, 1994; Sa´nchez et al., 1996). The concept that there exist separate populations of oligodendrocytes destined to make myelin sheaths of certain dimensions during development but not during remyelination seems unlikely given that oligodendrocytes make myelin sheaths appropriate for the axon they myelinate regardless of whether they are derived from a region of small or large diameter Wbers (Fanarraga et al., 1998).

REMYELINATION IN CLINICAL DISEASE Remyelination in MS The occurrence of remyelination in MS was initially suggested at the beginning of the previous century by Marburg, who identiWed thinly myelinated axons at the edge of lesions and inferred that they might represent a regenerative response on the basis of similar Wndings in myelin regeneration in peripheral nerves (Marburg, 1906). With the advent of electronmicroscopy and the realization from experimental models of demyelination that thin myelin sheaths are usually indicative of remyelination, more formal evidence of remyelination in MS lesions emerged (Perier and Gregoire, 1965; Suzuki et al., 1969). There is now clear evidence that remyelination can occur in MS lesions, and that this can occur extensively within a lesion (Prineas and Connell, 1979; Prineas et al., 1989, 1993a; Raine and Wu, 1993) (Fig. 7.1) (see Chapter 31). However, there is still much to be learned about the true extent of remyelination in MS patients, how this varies with the duration of the disease, and how it is aVected in diVerent subtypes of the disease (Lassmann et al., 1997; Lucchinetti et al., 2000).

Remyelination in Nonhuman Demyelinating Disease Although a wide variety of primary demyelinating disease has been described in a range of nonhuman species, there is little documented evidence of remyelination occurring, primarily because its presence or absence has not been speciWcally examined. Nevertheless, extensive remyelination has been reported in a white matter disease of silver foxes (Bunker and Friede, 1989) and in an unusual primary demyelinating condition of cattle in which large areas of remyelination occur that are predominantly mediated by Schwann cells (Palmer et al., 1991). In contrast, there are a plethora of examples of CNS remyelination in experimental models of demyelination, many of which have been speciWcally devised with a view to studying remyelination.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

REMYELINATION IN EXPERIMENTAL MODELS

REMYELINATION IN EXPERIMENTAL MODELS EAE The most widely used are the autoimmunity models, collectively referred to as experimental allergic encephalomyelitis (EAE). The pathology of EAE shares a number of features with MS and the model has been especially useful for studying the mechanisms of myelin injury. EAE is induced by sensitizing animals to myelin antigens, either actively by direct antigen exposure or passively by adoptive transfer of T cells. This results in inXammation within the CNS that is often accompanied by demyelination. Both acute and chronic relapsing forms of EAE can be induced. Acute forms are often followed by remyelination (Lampert, 1965). Although remyelination can occur in the chronic relapsing form of EAE (Raine et al., 1988; Snyder et al., 1975), it is also possible to produce lesions where remyelination fails or is limited, such as the model involving repeat exposure to a combination of myelin basic protein (MBP) speciWc T-cells and myelin-oligodendrocyte glycoprotein (MOG) antibody (Linington et al., 1992).

Viral Models Viral models are more widely used to study remyelination. Virulent strains of Theiler’s murine encephalomyelitis virus (TMEV) produce a biphasic disease in susceptible strains of mice. The Wrst phase of disease involves a mild and transient encephalitis, while the second phase results in multifocal demyelination (Dal Canto and Lipton, 1975; Lipton, 1975). Some strains of TMEV produce a disease in which successive waves of demyelination are followed by widespread remyelination by both oligodendrocytes and Schwann cells (Dal Canto and Barbano, 1984; Dal Canto and Lipton, 1980) (Fig. 7.3), while the Daniel’s strain of TMEV produces a progressive disease with superimposed acute inXammatory episodes in which remyelination is generally incomplete (Dal Canto and Lipton, 1975; Miller and Rodriguez, 1996). The A-59 and JHM strains of mouse hepatitis virus (MHV) also produce multifocal demyelination in mice, which may be followed by widespread remyelination (Herndon et al., 1977; Kristensson et al., 1986). Remyelination following demyelination induced by other viruses, including herpes simplex viruses types I and II, Ross-River virus, and Semliki Forest virus, has also been reported.

Toxin Models Remyelination commonly occurs following toxin-induced experimental remyelination. A widely used systemic chemical model involves feeding mice with a diet containing the copper-chelating agent cuprizone (Matsushima and Morell, 2001). Cuprizone causes demyelination in speciWc white matter regions of the brain, the location of which are determined by the dose. Low doses of cuprizone result in demyelination of the corpus callosum and related white matter tracts, while higher doses result in the cerebellar peduncles also becoming demyelinated. The exact mechanisms of action of cuprizone and why some regions are aVected more than others are not clear. Historically this lesion has been widely used in remyelination studies (Blakemore, 1974a, Johnson and Ludwin, 1981; Ludwin, 1979a, 1979b). Recently, there has been renewed interest in the cuprizone model as a means of studying demyelination and remyelination in transgenic mice (Arnett et al., 2001; Mason et al., 2001b). The two most frequently used chemical models that involve local administration or injection are the lysolecithin and ethidium bromide models (Fig. 7.3). Lysolecithin is a phospholipid that acts by solubilizing membranes via a general detergent eVect. It preferentially aVects myelin membranes, possibly due to the ability of nonmyelin membranes to resist its eVects by reacylation. Thus lysolecithin mainly causes demyelination by dissolving myelin sheaths, but can also lead to oligodendrocyte death. Remyelination following lysolecithin induced demyelination has been reported in a variety of species and following injection into diVerent white matter areas (Blakemore et al., 1977; Hall, 1972; Waxman et al., 1979; WoodruV and Franklin, 1999a). Ethidium bromide (EB)

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

177

178

7. REMYELINATION BY ENDOGENOUS GLIA

is a DNA-intercalating agent that prevents DNA replication and disrupts RNA and thus causes protein synthesis leading to cell death. It is a general cell toxin but can be used to induce demyelination by killing oligodendrocytes (as well as astrocytes and oligodendrocyte precursor cells—OPCs) when introduced into white matter at low concentrations either by direct injection or via injection into the subarachnoid or cisternal compartments (Blakemore, 1982; Reynolds and Wilkin, 1993; Sim et al., 2002b; WoodruV and Franklin, 1999a; Yajima and Suzuki, 1979). Like lysolecithin-induced demyelination, EB-induced demyelinating lesions generally undergo widespread remyelination mediated by both oligodendrocytes and Schwann cells, the extent to which each contributes being determined by the degree of astrocyte loss induced by the toxin (WoodruV and Franklin, 1999a). Remyelination also occurs following demyelination induced by the injection of complement and antibodies directed against oligodendrocyte antigens (Carroll et al., 1984; Keirstead and Blakemore, 1997; WoodruV and Franklin, 1999a).

Others Various forms of physical injury cause demyelination that is followed by remyelination. Indeed, the Wrst unequivocal demonstration of CNS remyelination involved such an approach. Using the technique of ‘‘CNS barbotage,’’ in which cerebrospinal Xuid is removed from and then reinjected into the cisterna magna, demyelination of the underlying white matter occurred that subsequently underwent remyelination (Bunge et al., 1961). Remyelination also occurs following demyelination induced by several models of traumatic injury (Blight and Young, 1989; Gledhill et al., 1973; GriYths and McCulloch, 1983; Olby and Blakemore, 1996) and following thermal injury (Sasaki and Ide, 1989).

CELLULAR BASIS OF CNS REMYELINATION When an area of white matter has been remyelinated, the number of oligodendrocytes in that region is greater than it was before demyelination (Prayoonwiwat and Rodriguez, 1993) (Fig. 7.3A). This suggests that there must be a means by which new oligodendrocytes are generated. Although oligodendrocytes may be able to survive within an area of demyelination, two lines of evidence indicate that it is unlikely that they are able to regenerate new myelin sheaths and contribute to remyelination. First, Gal-C expressing oligodendrocytes isolated from adult white matter do not form new myelin sheaths when transplanted into areas of demyelination (Targett et al., 1996). Second, post-mitotic oligodendrocytes that survive a demyelinating episode despite having lost their myelin sheaths do not contribute to subsequent remyelination (Keirstead and Blakemore, 1997). Although the cells identiWed in these studies were not proven to be myelinating oligodendrocytes prior to isolation or demyelination, the observations do suggest that remyelination is not eVected by cells expressing markers of the later stages of the oligodendrocyte lineage.

Origin of Remyelinating Cells The source of new oligodendrocytes in remyelination has been one of the longest running debates in remyelination research, but a consensus is now emerging that most, and probably all, remyelinating cells come from oligodendrocyte progenitor cells (OPCs) (Blakemore and Keirstead, 1999; Franklin, 1999; Ludwin and Bakker, 1988; Mason et al., 2000b; Wood and Bunge, 1991) (Fig. 7.4). The evidence for this claim is based on the following observations: First, proliferating cells that are likely to be OPCs (Horner et al., 2000) and can be labeled either by injecting a LacZ-expressing retrovirus into normal adult white matter or by labeling with tritiated thymidine or BrdU give rise to labeled remyelinating oligodendrocytes following induction of demyelination (Carroll and Jennings, 1994; Gensert and Goldman, 1997; Watanabe et al., 2002). Second, OPCs isolated from adult

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

179

CELLULAR BASIS OF CNS REMYELINATION

c

a

Normal appearing white matter

CNP (red) / NG2(green)

d

b

White matter lesion FIGURE 7.4 NG2þ cells in normal spinal cord and in MOG-EAE lesions at day 34 following induction. (A) The sparse but uniform distribution of NG2þ cells in the spinal cord white matter. (B) Focal increase in NG2þ cell number and immunoreactivity in a white matter inXammatory demyelinating lesion. (C) Reactive NG2þ OPCs in an area of demyelination. (D) Cell division in OPCs in lesioned areas is a common occurrence, as demonstrated by an NG2þ cell in metaphase. Scale bar represents 50 mm in A and B and 10 mm in C and D. Images kindly provided by Prof. Richard Reynolds.

CNS can remyelinate areas of demyelination following transplantation (Windrem et al., 2002; Zhang et al., 1999). Third, OPCs identiWed with a range of markers that include the growth factor receptor PDGF-aR (Redwine and Armstrong, 1998; Sim et al., 2002b), the proteoglycan NG2 (Cenci di Bello et al., 1999; Levine and Reynolds, 1999; Mason et al., 2000a; Watanabe et al., 2002) (Fig. 7.4), and the zinc Wnger transcription factor MyT1 (Sim et al., 2002b) have patterns of expression that are consistent with being the source of new oligodendrocytes, although for few of these markers is there unequivocal evidence that the cells they label become remyelinating oligodendrocytes (Dawson et al., 2000). These more recent studies using OPC markers substantiate earlier speculations on the ‘‘precursor’’ origin of remyelinating cells that were based on morphological criteria (Blakemore, 1972; Carroll et al., 1990; Ludwin, 1979a). Progenitor cells at even earlier stages in the oligodendrocyte lineage than the OPC are also likely to contribute to remyelination, because cells are recruited from the subventricular zone (where these early progenitors are found) when demyelination occurs nearby (Decker et al., 2002b; NaitOumesmar et al., 1999; Picard-Riera et al., 2002). A further level of complexity revealed by tissue culture studies is that the OPC population may be more heterogeneous than was previously thought (Gensert and Goldman, 2001). If a similar heterogeneity also occurs in vivo, then the possibility exists that diVerent progenitors may not necessarily respond to environmental cues or contribute to remyelination in the same manner (Mason and Goldman, 2002).

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

180

7. REMYELINATION BY ENDOGENOUS GLIA

Stages of Remyelination The important Wrst step in remyelination is to populate an area of demyelination with suYcient OPCs. These might come from cells already present within the demyelinated area, or they might be recruited from surrounding intact white matter (Carroll and Jennings, 1994; Franklin et al., 1997; Sim et al., 2002b). This initial recruitment phase involves OPC proliferation and migration. The contribution made by OPC migration to remyelination in MS is not clear, but it is likely to be determined by the extent to which OPCs survive within areas of demyelination, an issue that is itself not fully resolved. Once recruited, the OPCs diVerentiate into myelin-sheath-forming oligodendrocytes. During this diVerentiation phase, the OPCs diVerentiate into premyelinating oligodendrocytes that engage the demyelinated axon, before Wnally becoming mature oligodendrocytes as their sheet-like processes form spiral wraps around axons and the cytoplasmic contents of the wraps are extruded to form compacted myelin sheaths. Although generally thinner and shorter than the original myelin sheath, the composition of the myelin, the manner in which it is formed, and the transcriptional regulators involved are essentially similar to those in developmental myelination (Capello et al., 1997; Ludwin and Sternberger, 1984; Morell et al., 2000; Sim et al., 2000).

MEDIATORS OF CNS REMYELINATION The success of remyelination depends on OPCs being exposed to the appropriate environmental signals that mediate both the recruitment and diVerentiation phases. Regulation of the proliferation and migration of OPCs is one of the most widely studied areas of oligodendrocyte biology, although generally in the context of development (Richardson, 2001). Nevertheless, developmental studies provide valuable clues to the regenerative process. Most attention has focused on growth factors, some of which have both proliferative and pro-migratory eVects, while others seem to be solely mitogenic (Franklin et al., 2001). Other paracrine signaling molecules such as chemokines and cytokines (Robinson et al., 1998; Benveniste and Merrill, 1986) can also be mitogenic, and extracellular matrix molecules can aVect both proliferation and migration (Blaschuk et al., 2000; Garcion et al., 2001; Kiernan et al., 1996). An important point that highlights the likely complexity of the signaling environment for remyelination is that diVerent signaling mechanisms can interact with one another. For example, certain combinations of growth factors work synergistically (Jiang et al., 2001; Shi et al., 1998; Wolswijk and Noble, 1992), and in some instances in a manner that is dependent on the CNS region from which the cells are derived (Robinson and Miller, 1996). Further evidence of important interactions between diVerent signaling mechanisms has also been demonstrated between growth factors and the ECM (Baron et al., 2002; Colognato et al., 2002).

Mediators of Recruitment Comparatively little is known about which factors are actually involved in OPC recruitment following demyelination. A number of studies on growth factor expression during remyelination of experimental models have described patterns of expression consistent with PDGF and FGF, both of which have mitogenic and pro-migratory eVects on adult OPCs, being involved in the recruitment phase (Armstrong et al., 2002; Hinks and Franklin, 1999; Messersmith et al., 2000; Redwine and Armstrong, 1998; Tourbah et al., 1992). The correlation between the delay in PDGF expression and the delay in the accumulation of OPCs following demyelination that occurs when old animals are compared to young animals is consistent with PDGF mediating OPC recruitment (Hinks and Franklin, 2000; Sim et al., 2002b). A similar role has been assigned to FGF-2, since augmenting FGF-2 levels in EAE using a herpes viral vector approach results in an increase in NG2-expressing OPCs (RuYni et al., 2001), while FGF-2 enhances proliferation of OPCs isolated from areas of demyelination (Armstrong et al., 2002). However, the

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

MEDIATORS OF CNS REMYELINATION

OPC response to demyelination is the same in FGF-2
/
mice as in wild-type probably due to the presence of other mitogens compensating for the absence of FGF-2 (Armstrong et al., 2002). Similarly, studies on the expression patterns of the pro-migratory adhesion molecule PSA-NCAM indicate a likely involvement in OPC recruitment (Decker et al., 2002a; Nait-Oumesmar et al., 1995).

Mediators of Differentiation Less is known about the mechanisms by which OPCs diVerentiate into myelin-sheathforming oligodendrocytes than about their proliferation and migration. Nevertheless, some progress has been made in recent years in identifying factors that either promote or inhibit diVerentiation, elucidating the cell-intrinsic mechanisms that are associated with this process. In the context of MS, this is an area of oligodendrocyte biology that has acquired particular signiWcance with the realization that failure of diVerentiation might be a signiWcant contributor to remyelination failure (Chang et al., 2000, 2002; Wolswijk, 1998; also see the section titled ‘‘The Origin of Schwann Cells That Remyelinate CNS Axons’’). Inhibitors of diVerentiation may also play a crucial role in the recruitment phase by keeping OPCs in a state where they are still responsive to recruitment signals. Growth factors were among the Wrst regulators of OPC diVerentiation to be studied, although their precise roles are not clearly understood. IGF-I has been implicated in OPC diVerentiation primarily on the basis of its eVects on myelination (Carson et al., 1993; Goddard et al., 1999; Ye et al., 1995), and its expression proWle relative to oligodendrocyte markers in experimental models (Hinks and Franklin, 1999; Komoly et al., 1992; Mason et al., 2000a) and MS lesions (Gveric et al., 1999) is consistent with this role. However, IGF-I is also involved in OPC proliferation (Jiang et al., 2001), and might therefore be both mitogenic and enhance diVerentiation in a context-dependent manner (Rosenthal and Cheng, 1995). TGF-b1 inhibits mitogen-induced OPC proliferation and, as exit from the cell cycle is associated with diVerentiation (Barres and RaV, 1994), might be implicated in triggering diVerentiation (Mckinnon et al., 1993). Consistent with a role for both IGF-I and TGF-b1 in promoting OPC diVerentiation, the peak expression of both these growth factors is delayed in aged animals in which there is also an impairment of diVerentiation (Hinks and Franklin, 2000; Sim et al., 2002b). However, advancing the onset of IGF-I expression in old animals using IGF-I-expressing adenoviral vectors is not suYcient to bring forward the onset of diVerentiation (O’Leary et al., 2002). This result would be consistent with multiple factors being involved in controlling diVerentiation and highlights the diYculties of changing a process that involves many regulatory mechanisms by simply changing a single component. FGFs exert an inhibitory eVect on diVerentiation by preventing progression of OPCs into cells that express markers of a more mature stage of development (Armstrong et al., 2002; Goddard et al., 1999, 2001; Mckinnon et al., 1990). Similarly, the axonally-expressed neuregulin GGF-2 prevents diVerentiation of immature oligodendrocytes (Canoll et al., 1996). However, its eVects are clearly complex as it is also necessary for terminal diVerentiation of oligodendrocyte lineage cells into mature oligodendrocytes (Park et al., 2001). Recently, however, light has been shed on these apparent ambiguities with the demonstration that the diVerentiation-inhibiting eVects of GGF-2 are switched to a survival eVect by concurrent signaling by axonally expressed laminin via OPC-expressed integrins (Colognato et al., 2002). Several other mechanisms have also been implicated in regulating OPC diVerentiation. Particularly intriguing is the ability of neurotransmitters acting at glutamate and b-adrenergic receptors to either inhibit or promote OPC diVerentiation (Gallo et al., 1996; Ghiani et al., 1999). The notch-jagged signaling system has multiple roles in development, many of which involve inhibition of diVerentiation. This includes the diVerentiation of OPCs, which is blocked by axonally expressed jagged acting at notch-1 receptors on OPCs (Genoud et al., 2002; Givogri et al., 2002; Wang et al., 1998). The expression of the adhesion molecule PSA-NCAM by axons can also inhibit diVerentiation (Charles et al., 2000). Conversely, electrical activity in axons appears to promote myelination, implying that it promotes OPC diVerentiation (Demerens et al., 1996). The extracellular matrix, signaling through ß1 or ß5

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

181

182

7. REMYELINATION BY ENDOGENOUS GLIA

integrin, might also regulate the Wnal phase of diVerentiation by promoting myelin sheath formation (Blaschuk et al., 2000; Buttery and Vrench-Constant, 1999; Relvas et al., 2001). A role for integrin-mediated eVects on oligodendrocyte lineage cells is supported by evidence of expression of the integrin ligands vitronectin and tenascins-R and -C in chronic MS lesions (Gutowski et al., 1999; Sobel et al., 1995). The axonal expression of GGF-2, jagged, PSA-NCAM, and vitronectin, together with the eVects of electrical activity, suggest that the axon is unlikely to be a passive partner in remyelination. Indeed, given the critically important, but incompletely understood, oligodendrocyte-axon interactions required for myelin sheath formation, the integrity of the axon may constitute a fundamental determinant of remyelination. Some of the intracellular signaling mechanisms involved in regulating OPC diVerentiation have been identiWed. Critical events in OPC development are the accumulation of the cyclin-dependent kinase inhibitor p27 (Casaccia-BonneWl et al., 1997; Durand et al., 1997, 1998) and beta thyroid hormone receptor (TRb1) (Gao et al., 1998), which eventually reach a threshold level within the cell that stops proliferation and triggers the onset of diVerentiation. More recently, histone deacetylase activity within OPCs has been identiWed as necessary for their diVerentiation (Marin-Husstege et al., 2002). The regulation of some of these processes at the transcriptional level is also becoming clearer. The transcription factor Sox 10 plays a key role in the terminal diVerentiation of oligodendrocyte lineage cells (Stolt et al., 2002), a process that is also associated with declining levels of the bHLH proteins Id2, Id4 and Hes5 (Kondo and RaV, 2000a, 2000b; Wang et al., 2001), and the zinc-Wnger protein MyT1 (Armstrong et al., 1995; Sim et al., 2002b). In contrast, OPC diVerentiation is associated with increasing levels of AP-1 (FitzGerald and Barnett, 2000), the bHLH protein Mash1 (Kondo and RaV, 2000a), the homeodomian factor Gtx (Awatramani et al., 1997; Sim et al., 2000), and the peroxisome proliferator-activated protein delta (PPARd) (Saluja et al., 2001). Both Hes5, the expression of which is increased by activation of notch receptors (Wang et al., 1998), and Mash1 appear to exert their eVects on OPC diVerentiation by regulating the levels of TRb1. The recently described bHLH protein oligo1 may also play a role in OPC diVerentiation since the onset of expression of proteins expressed by mature oligodendrocytes during development is delayed in its absence (Lu et al., 2002).

Inflammation and Remyelination It is now clear that a multiplicity of environmental signals are involved in orchestrating remyelination. Each phase of remyelination is regulated by a complex pattern or matrix of signaling events in which the correct levels and temporal proWle of expression are also likely to be critical. A pro-recruitment environment must be maintained long enough for a lesion to become repopulated to the extent commensurate with complete remyelination, and only when repopulation is complete should the environment shift to one that supports diVerentiation. What are the critical triggering events that create an environment that is conducive to remyelination? Several lines of evidence implicate the inXammatory response to demyelination. Pathological evidence from MS tissue indicates that when remyelination occurs it does so in acute, active lesions characterized by a robust inXammatory response (Prineas et al., 1989, 1993a; Raine and Wu, 1993; Wolswijk, 2002). Although clinical evidence suggests that remyelination can be a slow and protracted process (Brusa et al., 2001), the process appears to progress most eYciently when it is linked to the inXammatory response. These pathological observations are supported by experimental evidence, particularly from toxin models of demyelination where the inXammatory response is a consequence of demyelination, and not its cause as is the case in immune-mediated demyelination. The macrophage emerges from these studies with more credit than is the case in immune-mediated models of demyelination, where their activation by the speciWc immune response renders them more damaging to tissue. In general, macrophages associated with demyelination are derived from both resident microglia and recruited haematogenous monocytes, and earlier studies drew attention to the correlation between their numbers and activity and the eYciency of remyelination (Graca and Blakemore, 1986;

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

WHY DOES REMYELINATION FAIL?

Ludwin, 1980). In fact, macrophage depletion impairs remyelination (Kotter et al., 2001), while addition of macrophages to a co-culture system of neurons and glia leads to increased synthesis of proteins associated with myelination (Loughlin et al., 1997). These observations Wt well with recent studies using knockout mice in which remyelination is impaired following cuprizone-induced demyelination in the absence of the pro-inXammatory cytokines Il-1b and TNF-a (Mason et al., 2001b; Arnett et al., 2001). These observations imply that while cytokines may not be directly involved in regulating the behavior of oligodendrocyte lineage cells during remyelination, they are nevertheless necessary to create the appropriate cellular and molecular milieu for the process to occur. As is often the case, there are two sides to the TNF-a story: this cytokine appears to have two opposite eVects, either promoting OPC survival or promoting cell death, depending on the receptor through which it signals. The presence of survival factors such as IGF-I in the lesion may prevent or at least ameliorate the cell death signaling. The precise functions of macrophages in remyelination have yet to be fully evaluated. One suggestion has been that their role in removing myelin debris is in itself helpful since contact with myelin can inhibit OPC diVerentiation (Robinson and Miller, 1999). However, there is no direct evidence for this, and it may be that this is simply a histologically obvious outcome of an appropriate macrophage response whose real contribution is in the signaling molecules they secrete. For example, macrophages within demyelinating lesions are a source of some of the growth factors that have been implicated in remyelination (Gveric et al., 1999; Hinks and Franklin, 1999). They also produce factors that activate astrocytes, which, in their activated state, produce remyelination-associated growth factors (Hinks and Franklin, 1999; Redwine and Armstrong, 1998). A picture is emerging of a situation in which activation of microglia and astrocytes, initially by demyelination and subsequently by each other, triggers a cascade of events that, as they unfold, create and regulate a pro-remyelinating environment in which multiple signaling mechanisms, including cytokines, growth factors, extracellular matrix, and adhesion molecules are all expressed in an appropriate manner.

WHY DOES REMYELINATION FAIL? Despite increasing understanding of the biology of remyelination, there is no clear picture why remyelination fails (Franklin, 2002). What is clear, however, is that there are likely to be many diVerent reasons. Remyelination might fail because of an inadequate provision of OPCs (recruitment failure) or because of a failure of recruited OPCs to diVerentiate into remyelinating oligodendrocytes (diVerentiation failure). As animals age, both of these phases of remyelination become more protracted (Sim et al., 2002b), leading to a decrease in remyelination eYciency (Shields et al., 1999), an outcome that in itself has signiWcant implications for the likelihood of recovery from new lesions during the course of a disease that can last for several decades (Franklin et al., 2002). In addition, it is obvious that remyelination will not occur if axons are lost, which takes place in demyelinating disorders. However, the mere presence of axons per se may not result in remyelination if the axons do not express appropriate diVerentiation signals to OPCs or if they express molecules that inhibit myelin sheath formation (noted earlier).

Failure of Recruitment Recruitment might fail if OPC numbers are depleted as part of the disease process, and recent studies demonstrating the presence of antibodies against OPC-expressed antigens indicate that at least in some patients this may be the case (Archelos et al., 1998; Niehaus et al., 2000). Moreover, if the recent demonstration that at least some adult OPCs in the rat express myelin oligodendrocyte protein (MOG) also applies to some human OPCs, then patients with antibodies against this protein may also become deWcient in OPCs (Li et al., 2002). Another possible cause of OPC depletion, about which there has been much

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

183

184

7. REMYELINATION BY ENDOGENOUS GLIA

speculation, is the concept that repeated episodes of demyelination and remyelination might lead to their exhaustion and hence to there being too few available for recruitment. In experimental models, repeated rounds of demyelination and remyelination lead to less eYcient remyelination (Johnson and Ludwin, 1981), and there is some evidence that the same may be true in MS (Prineas et al., 1993b). Although eventual depletion of OPCs for recruitment represents one explanation for this phenomenon, recent evidence suggests that OPCs have a very robust ability to repopulate areas from which they are missing (Chari and Blakemore, 2002a). Moreover, there appears to be no depletion of OPCs in and around areas of remyelination (Levine and Reynolds, 1999; Sim et al., 2002b). A recent study showed that if a suYcient period of time elapses between subsequent episodes of demyelination then remyelination is not impaired, neither is there a prolonged depletion of OPCs, indicating that the process of remyelination per se does not prevent remyelination occurring at the same site at a later time (Penderis et al., 2002). Whether protracted and sustained exposure to a demyelinating stimulus is suYcient to eventually exhaust the OPC population has not been established, but it is conceivable that in the face of relentless demand OPC depletion might contribute to poor remyelination (Linington et al., 1992, Ludwin, 1980). While OPCs proliferate perhaps indeWnitely in culture supplemented with growth factors, the proliferation potential of OPCs in vivo is not clear. The depletion of OPCs must depend on the equilibrium between proliferation (generating more OPCs) and diVerentiation (responding to signals to myelinate).

Failure of Differentiation In the past few years, it has become apparent that some chronically demyelinated lesions are not deWcient in OPCs, and that remyelination failure is associated not with a paucity of oligodendrocyte lineage cells but with their failure to diVerentiate into remyelinating oligodendrocytes (Chang et al., 2000; Maeda et al., 2001; Scolding et al., 1998; Wolswijk, 1998) (Fig. 7.5). Some lesions also contain proteolipid protein (PLP)-expressing premyelinating oligodendrocytes that have progressed to the stage of engaging demyelinated axons but have not progressed further to form compacted myelin sheaths (Chang et al., 2002). Whether these OPCs fail to diVerentiate because they do not receive the appropriate signals or are exposed to signals that block diVerentiation is not clear. However, two recent reports indicating expression of two diVerentiation inhibitors, PSA-NCAM and notch ligands provide circumstantial evidence in support of the latter observation (Charles et al., 2002; John et al., 2002).

The Dysregulation Hypothesis On the basis of the preceding overview of the mechanisms involved in achieving successful remyelination, a hypothesis can be proposed for remyelination failure in which the signaling environment becomes inappropriately regulated or ‘‘dysregulated’’ (Franklin, 2002). According to this hypothesis, there are no individual villains of the piece that are responsible for remyelination failure. Instead, the process fails because the complex and Wnely tuned mechanism by which it proceeds loses its precise coordination. For example, a lesion may fail to remyelinate because a pro-recruitment environment is not maintained for a suYciently long period or because the premature onset of a pro-diVerentiation environment arrests recruitment before the lesion becomes adequately populated by OPCs. If OPCs are recruited from outside an area of demyelination, the size of the lesion will clearly have a bearing on how much OPC repopulation is needed. Large lesions will require a more protracted recruitment phase and it has been calculated that it could take up to 5 months for a 2-cm lesion to become repopulated with OPCs (Chari and Blakemore, 2002b). As the acute inXammation associated with demyelination provides a powerful stimulus for OPC recruitment, the subsiding of the inXammatory response may paradoxically cut short the regenerative process, a potential reason for remyelination failure that should be considered in devising anti-inXammatory therapies for MS. If remyelination is incomplete when inXammation subsides, then the resulting lesion is likely to remain

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

CNS REMYELINATION BY SCHWANN CELLS

FIGURE 7.5 Remyelination failure due to inadequate diVerentiation of oligodendrocyte lineage cells within areas of demyelination. An O4þ OPC (green) within an area containing demyelinated, neuroWlamentþ axons (red) within the brain of a patient with MS of 46 years duration. Image kindly provided by Dr. Guus Wolswijk of The Netherlands Institute for Brain Research.

chronically demyelinated. These lesions are generally characterized by the presence of a dense matrix of astrocyte processes, sometimes referred to as a glial scar, and it is a longstanding view that this might cause remyelination failure by preventing remyelinating cells from gaining access to demyelinated axons. However, an alternative view is that the scar is a consequence of failed remyelination, appearing after the window of opportunity for remyelination has lapsed rather than being present at the outset and preventing it from occurring. Indeed, a case can be made that astrocytes generally perform a helpful role in remyelination; although there is evidence that they might impair OPC recruitment by restricting migration (Fok-Seang et al., 1995; Groves et al., 1993), there is also evidence that they produce a wide range of signaling molecules that support recruitment (Redwine and Armstrong, 1998). The scarring astrocytes within chronically demyelinated plaques are relatively quiescent, and it is unlikely that remyelination will occur in this environment. If, on the other hand, these cells reverted to the activated state they exhibit during inXammation, then conditions might become more conducive for remyelination. In this regard, it is interesting that tenascin-C, an inhibitor of OPC migration (Kiernan et al., 1996), is expressed by astrocytes in chronic MS lesions but not within active inXammatory lesions (Gutowski et al., 1999).

CNS REMYELINATION BY SCHWANN CELLS The phenomenon of Schwann cell remyelination in the CNS has been extensively documented in both naturally occurring demyelinating disease such as MS (Feigin and Ogata, 1971; Ghatak et al., 1973; Itoyama et al., 1985; Yamamoto et al., 1991) and in many diVerent forms of experimental demyelination including EAE (Raine et al., 1978; Snyder et al., 1975) and those induced by toxins (Blakemore, 1976, 1982; Harrison et al., 1972), viruses (Dal Canto and Lipton, 1980), and physical injury (Blight and Young, 1989; GriYths and McCulloch, 1983) (Fig. 7.6). Remyelination mediated by Schwann

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

185

186

7. REMYELINATION BY ENDOGENOUS GLIA

FIGURE 7.6 Schwann cell remyelination in the CNS. (A) Toluidine blue stained resin section of an area of remyelination following ethidium bromide-induced demyelination of the deep cerebellar white matter of an adult rat. Most of the axons have thin myelin sheaths characteristic of oligodendrocyte remyelination (oligo), giving rise to pale areas which contrast with the dark staining of normally myelinated axons present at the bottom of the image. The dotted line delineates an area of Schwann cell remyelination (SC) that has a staining quality distinct from that in the oligodendrocyte remyelinated area. Typically, Schwann cell remyelination occurs around blood vessels. (B) Although Schwann cell remyelination can be readily distinguished from oligodendrocyte remyelination by morphological criteria alone, it can clearly be distinguished by immunostaining with antibodies against P0, a peripheral myelin speciWc protein. (C) At the ultrastructural level, the distinction between Schwann cell and oligodendrocyte remyelination is very apparent. A Schwann cell myelinated axon can be seen at the top of the image. The cell body lies adjacent to the myelin sheath, which is surrounded by a rim of Schwann cell cytoplasm, giving rise to the characteristic signet ring appearance of Schwann cell remyelinated axons. Three oligodendrocyte remyelinated axons can be seen at the bottom of the image. Scale bar represents approximately 30 mm in A, approximately 25 mm in B, and approximately 1.5 mm in C.

cells can be distinguished from oligodendrocyte remyelination by a variety of criteria. In semithin plastic sections stained with toluidine blue, the myelin sheath has a dark staining hue by which it can be distinguished from myelin formed by myelinating or remyelinating oligodendrocytes (Fig. 7.6A). Since Schwann cells make a single myelin sheath, the cell body often rests adjacent to the myelinated axon, giving rise to a characteristic signet ring appearance. At the ultrastructural level, Schwann cell myelination has a number of features that make its identiWcation straightforward (Fig. 7.6C). The myelin sheath has a characteristic periodicity slightly greater than that associated with the central type myelin formed by oligodendrocytes and is surrounded by a distinct rim of cytoplasm. The surface of the myelinating cell is usually covered by a basal lamina. A number of markers, the most frequently used of which is the peripheral myelin speciWc protein P0, can be readily detected by immunohistochemistry (Fig. 7.6B). Recently, it has been shown that two POU-domain transcription factors, Oct-6 and Brn-2, are speciWcally associated with Schwann cell remyelination and not oligodendrocyte remyelination of the CNS (Sim et al., 2002a).

When Does Schwann Cell Remyelination of the CNS Occur? In early studies on toxin-induced demyelination, it was noted that Schwann cell remyelination occurred in areas from which astrocytes are absent, whereas oligodendrocyte remye-

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

au4

CNS REMYELINATION BY SCHWANN CELLS

lination predominated in areas in which astrocytes were present. Indeed, it was established that the degree of Schwann cell type remyelination is related to the degree to which astrocytes are lost in demyelinating lesions. For example, Schwann cell remyelination is less following lysolecithin-induced demyelination of the spinal cord white matter compared to EB-induced demyelination, since lysolecithin is relatively astrocyte-sparing compared to EB, a more nonspeciWc cell toxin that causes injury to astrocytes and oligodendrocytes in equal measure. The relationship between astrocytes and Schwann cell remyelination is seen in the more aggressive forms of MS (Itoyama et al., 1985) and also accounts for the distribution of Schwann cell remyelination following transplantation into areas of experimentally induced demyelination (Shields et al., 2000). As a result of these observations, the concept of Schwann cell myelination of CNS axons constituting a repositioning of the glia limitans has emerged (Franklin and Blakemore, 1993). The astrocytic processes of the glia limitans normally deWne the boundary between the CNS and the PNS. When the astrocytes of the glia limitans are damaged as well as oligodendrocytes, then the demyelinated axons become available for remyelination by Schwann cells that remyelinate up to the position of the newly established glia limitans. What results is a region containing axons that were previously within the CNS but now become PNS tissue by virtue of the glia that associate with them. Thus, in a strict sense, Schwann cell remyelination does not occur within the CNS. Although many experimental observations support this concept, others indicate that a more complex relationship between Schwann cells and a CNS glial environment. For example, in the shaking pup, an animal with severe hypomyelination arising from an Xlinked PLP gene mutation, Schwann cell myelination can be found in spinal cord white matter despite the presence of a normal or hypertrophic astrocyte population (Duncan and HoVman, 1997).

The Origin of Schwann Cells That Remyelinate CNS Axons It has always been assumed that when Schwann cell remyelination of CNS axons occurs, these cells are derived from Schwann cells of the PNS. More recently, however, it has been suggested that Schwann cells within the CNS may be derived from a multipotent progenitor population residing within the CNS (Keirstead et al., 1999; Akiyama et al., 2001). Whether Schwann cells remyelinating CNS axons are derived exclusively from one source or the other, or from a combination of both sources, has not been fully established. It is appropriate, however, to review the arguments for and against either a PNS or CNS origin for CNS remyelinating Schwann cells. The distribution of Schwann cell remyelination of the CNS frequently suggests a PNS origin. For example, Schwann cell remyelination is much more extensive following demyelination of spinal cord tracts close to the Schwann cell–myelinated axons of the spinal roots or meningeal nerves than it is when demyelination is induced within the deep cerebellar white matter remote from a peripheral nerve source (Franklin et al., 1991; WoodruV and Franklin, 1999a). Moreover, in the spinal cord the Schwann cell remyelination is most extensive closest to the surface (Duncan and HoVman, 1997; Raine et al., 1978). When Schwann cell remyelination is seen deep within CNS matter, it generally has a perivascular distribution. If these Schwann cells were derived from the PNS, then their probable source is the axons of the autonomic nervous system associated with the larger diameter blood vessels within the brain. These cells could access areas of demyelination following migration along perivascular spaces (Raine et al., 1978a), a favored route of migration for transplanted Schwann cells (BaronVan Evercooren et al., 1993, 1996). If Schwann cells are indeed derived from PNS tissue, their presence in the CNS can be accounted for as arising from a breach in the glia limitans allowing them access to axons previously denied when this structure was intact. The concept that Schwann cells myelinating within the CNS are derived from nearby PNS tissue has recently been challenged by the demonstration that multipotent neural progenitor cells from the CNS give rise to Schwann cells following transplantation into the CNS (Akiyama et al., 2001; Keirstead et al., 1999). These Wndings are consistent with earlier reports that embryonic spinal cord neuroepithelial cells can be induced to become either PNS or CNS glia (Mujtaba et al., 1998). It is possible therefore that spontaneous

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

187

188

7. REMYELINATION BY ENDOGENOUS GLIA

Schwann cell remyelination results from diVerentiation of a precursor cell present within the CNS along a PNS pathway. In this scenario, it is the presence or absence of astrocytes au1 that determines diVerentiation fate, multipotent precursor cells giving rise to Schwann cells in their absence but oligodendrocytes where their presence deWnes a CNS environment. One could also hypothesize that endothelial derived signals may also fervor this pathway, accounting for the perivascular distribution of Schwann cell remyelination deep within demyelinated white matter tracts. However, the mechanism of this regulation remains speculative at present.

PROSPECTS FOR THERAPEUTIC PROMOTION OF CNS REMYELINATION The failure of remyelination, which becomes an increasingly characteristic feature of multiple sclerosis as the disease progresses, is the basis of the current quest for methods by which remyelination might be promoted. Transplantation of myelinogenic cells is one approach in which much eVort has been invested and one that gets ever closer to clinical implementation (see Chapter 6, this volume). An alternative approach is to try and enhance or reactivate intrinsic remyelination.

Growth Factors The prospect that growth factors may provide a means of promoting remyelination has often been discussed (McMorris and Mckinnon, 1996; Wolswijk, 1999; WoodruV and Franklin, 1997). Although much remains to be learned about the precise roles of growth factors in remyelination, it is becoming clear from the existing experimental data that the process requires the sequential expression of diVerent growth factor combinations. Moreover, the timing of this sequence critically determines its rate and eYciency of remyelination. An implication of this is that the eVects of growth factor supplementation will be determined by the stage of remyelination at which they are administered and that inappropriate supplementation may actually impede remyelination. For example, if a lesion were failing to remyelinate because of a shortage of OPCs, then recruitment-associated growth factors would be beneWcial. On the other hand, diVerentiation-associated growth factors would be disadvantageous by inducing premature diVerentiation at the expense of the required OPC expansion. Conversely, if a lesion was replete with OPCs that were failing to diVerentiate, then diVerentiation-associated growth factors may allow remyelination to proceed to completion, while recruitment-associated growth factors would exacerbate the problem. The potential for inhibiting myelination by growth factor delivery has been demonstrated in developmental myelination (Goddard et al., 1999) and following demyelination in FGF-2
/
mice (Armstrong et al., 2002). With current technology it seems unlikely that it would be possible to know what growth factor strategy would be required for a particular plaque or for a particular patient without recourse to biopsy. It is noteworthy that the occasions in which growth factor therapy has been reported to enhance remyelination that the growth factors are delivered systemically and there is no clear indication whether the eVects are mediated directly or indirectly via, for example, modulating the inXammatory response (Cannella et al., 1998; Yao et al., 1995, 1996). In contrast to the eVects of systemically delivered IGF-I in an immune-mediated model of demyelination, when IGF-I is delivered directly into an area of demyelination there is no enhancement of repair (O’Leary et al., 2002).

Immunoglobulins Perhaps the most promising approach to promoting endogenous remyelination has been the use of immunoglobulins (Miller et al., 1996). Using the DA-TMEV model of chronic demyelination it was found, contrary to expectation, that when animals where administered spinal cord homogenate, there was an enhancement of remyelination instead of the

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

au2

PROSPECTS FOR THERAPEUTIC PROMOTION OF CNS REMYELINATION

predicted exacerbation in demyelination (Lang et al., 1984). This result was reminiscent of an earlier report that administration of myelin components after induction of EAE could stimulate repair (Traugott et al., 1982). It was subsequently shown that the same eVect could be induced by passive transfer of either antiserum or puriWed immunoglobulin, establishing the principle of autoantibody-mediated remyelination. Isolation of speciWc remyelination-enhancing monoclonal antibodies indicate that all are IgMs and many recognize epitopes expressed on the surface of cultured oligodendrocytes (Asakura et al., 1998; Miller et al., 1994). With a view to progressing this experimental data toward clinical therapy, two human monoclonal antibodies directed against oligodendrocyte surface antigens have been isolated that promote remyelination in the DA-TMEV model (Warrington et al., 2000) and following toxin-induced demyelination (Bieber et al., 2002). The mechanism by which autoantibody-mediated enhancement of remyelination operates is still unclear, and although the antibodies are directed against epitopes expressed by oligodendrocyte lineage cells, this does not imply that the eVect arises from a direct stimulation of these cells. Indeed, eliciting a change in the behavior of cultured oligodendrocyte lineage cells has proven diYcult (Stangel et al., 1999), suggesting that the eVects may be indirectly mediated.

Future Prospects At present there are no therapies used in demyelinating disease that speciWcally aim to enhance remyelination. This is in part because we do not fully comprehend how remyelination occurs. Nevertheless, advances in our understanding of the biology of oligodendrocyte lineage cells, especially those present in the adult CNS, together with an increasingly sophisticated use of animal models of remyelination, will undoubtedly lead to a more complete picture. The next few years are likely to witness an increasing understanding of the biology of remyelination and signiWcant strides toward identifying new therapeutic approaches.

References Akiyama, Y., Honmou, O., Kato, T., Uede, T., Hashi, K., and Kocsis, J. D. (2001). Transplantation of clonal neural precursor cells derived from adult human brain establishes functional peripheral myelin in the rat spinal cord. Exp. Neurol. 167, 27–39. Archelos, J. J., Trotter, J., Previtali, S., Weissbrich, B., Toyka, K. V., and Hartung, H. P. (1998). Isolation and characterization of an oligodendrocyte precursor-derived B-cell epitope in multiple sclerosis. Ann. Neurol. 43, 15–24. Armstrong, R. C., Kim, J. G., and Hudson, L. D. (1995). Expression of myelin transcription factor I (MyTI), a ‘‘zinc-Wnger’’ DNA-Binding protein, in developing oligodendrocytes. Glia 14, 303–321. Armstrong, R. C., Le, T. Q., Frost, E. E., Borke, R. C., and Vana, A. C. (2002). Absence of Wbroblast growth factor 2 promotes oligodendroglial repopulation of demyelinated white matter. J. Neurosci. 22, 8574–8585. Arnett, H. A., Mason, J., Marino, M., Suzuki, K., Matsushima, G. K., and Ting, J. P. Y. (2001). TNFa promotes proliferation of oligodendrocyte progenitors and remyelination. Nat. Neurosci. 4, 1116–1122. Asakura, K., Miller, D. J., Pease, L. R., and Rodriguez, M. (1998). Targeting of IgMkappa antibodies to oligodendrocytes promotes CNS remyelination. J. Neurosci. 18, 7700–7708. Awatramani, R., Scherer, S., Grinspan, J., Collarini, E., SkoV, R., OqHagan, D., Garbern, J., and Kamholz, J. (1997). Evidence that the homeodomain protein Gtx is involved in the regulation of oligodendrocyte myelination. J. Neurosci. 17, 6657–6668. Baron-Van Evercooren, A., Avellana-Adalid, V., Ben Younes-ChennouW, A., Gansmu¨ller, A., Nait-Oumesmar, B., and Vignais, L. (1996). Cell-cell interactions during the migration of myelin-forming cells transplanted in the demyelinated spinal cord. Glia 16, 147–164. Baron-Van Evercooren, A., Duhamel-Clerin, E., Boutry, J. M., Hauw, J. J., and Gumpel, M. (1993). Pathways of migration of transplanted Schwann cells in the demyelinated mouse spinal cord. J. Neurosci. Res. 35, 428–438. Baron, W., Shattil, S. J., and Vrench-Constant, C. (2002). The oligodendrocyte precursor mitogen PDGF stimulates proliferation by activation of alpha(v)beta3 integrins. EMBO J. 21, 1957–1966. Barres, B. A., and RaV, M. C. (1994). Control of oligodendrocyte number in the developing rat optic nerve. Neuron 12, 935–942. Benveniste, E. N., and Merrill, J. E. (1986). Stimulation of oligodendroglial proliferation and maturation by interleukin-2. Nature 321, 610–613.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

189

190

7. REMYELINATION BY ENDOGENOUS GLIA

Bieber, A. J., Warrington, A., Asakura, K., Ciric, B., Kaveri, S. V., Pease, L. R., and Rodriguez, M. (2002). Human antibodies accelerate the rate of remyelination following lysolecithin-induced demyelination in mice. Glia 37, 241–249. Blakemore, W. F. (1972). Obsevations on oligodendrocyte degeneration, the resolution of status spongiosus and remyelination in cuprizone intoxication in mice. J. Neurocytol. 1, 413–426. Blakemore, W. F. (1974a). Pattern of remyelination in the CNS. Nature 249, 577–578. Blakemore, W. F. (1974b). Remyelination of the superior cerebellar peduncle in old mice following demyelination induced by cuprizone. J. Neurol. Sci. 22, 121–126. Blakemore, W. F. (1976). Invasion of Schwann cells into the spinal cord of the rat following local injections of lysolecithin. Neuropath. App. Neurobiol. 2, 21–39. Blakemore, W. F. (1982). Ethidium bromide induced demyelination in the spinal cord of the cat. Neuropath. App. Neurobiol. 8, 365–375. Blakemore, W. F., Eames, R. A., Smith, K. J., and McDonald, W. I. (1977). Remyelination in the spinal cord of the cat following intraspinal injections of lysolecithin. J. Neurol. Sci. 33, 31–43. Blakemore, W. F., and Keirstead, H. S. (1999). The origin of remyelinating cells in the CNS. J. Neuroimmunol. 98, 69–76. Blakemore, W. F., and Murray, J. A. (1981). Quantitative examination of internodal length of remyelinated nerve Wbres in the central nervous system. J. Neurol. Sci. 49, 273–284. Blaschuk, K. L., Frost, E. E., and Vrench-Constant, C. (2000). The regulation of proliferation and diVerentiation in oligodendrocyte progenitor cells by alphaV integrins. Development 127, 1961–1969. Blight, A. R., and Young, W. (1989). Central axons in injured spinal cat spinal cord recover electrophysiological function following remyelination by Schwann cells. J. Neurol. Sci. 91, 15–34. Brusa, A., Jones, S. J., and Plant, G. T. (2001). Long-term remyelination after optic neuritis: A 2-year visual evoked potential and psychophysical serial study. Brain 124, 468–479. Bunge, M. B., Bunge, R. P., and Ris, H. (1961). Ultrastructural study of remyelination in an experimental lesion in the adult cat spinal cord. J. Biophys. Biochem. Cytol. 10, 67–94. Bunker, E., and Friede, R. L. (1989). Changes in the histograms of nerves resulting from growth and various modalities of damage. Acta Neuropath. 78, 521–527. Buttery, P. C., and Vrench-Constant, C. (1999). Laminin-2/integrin interactions enhance myelin membrane formation by oligodendrocytes. Mol. Cell. Neurosci. 14, 199–212. Cannella, B., Hoban, C. J., Gao, Y. L., Garcia-Arenas, R., Lawson, D., Marchionni, M., Gwynne, D., and Raine, C. S. (1998). The neuregulin, glial growth factor 2, diminishes autoimmune demyelination and enhances remyelination in a chronic relapsing model for multiple sclerosis. Proc. Natl. Acad. Sci. USA 95, 10100–10105. Canoll, P. D., Musacchio, J. M., Hardy, R., Reynolds, R., Marchionni, M. A., and Salzer, J. L. (1996). GGF/ neuregulin is a neuronal signal that promotes the proliferation and survival and inhibits the diVerentiation of oligodendrocyte progenitors. Neuron 17, 229–243. Capello, E., Voskuhl, R. R., McFarland, H. F., and Raine, C. S. (1997). Multiple Sclerosis: Re-expression of a developmental gene in chronic lesions correlates with remyelination. Ann Neurol. 41, 797–805. Carroll, W. M., and Jennings, A. R. (1994). Early recruitment of oligodendrocyte precursors in CNS remyelination. Brain 117, 563–578. Carroll, W. M., Jennings, A. R., and Mastaglia, F. L. (1984). Experimental demyelinating optic neuropathy induced by intra-neural injection of galactocerebroside antiserum. J. Neurol. Sci. 65, 125–135. Carroll, W. M., Jennings, A. R., and Mastaglia, F. L. (1990). The origin of remyelinating oligodendrocytes in antiserum-mediated demyelinative optic neuropathy. Brain 113, 953–973. Carson, M. J., Behringer, R. R., Brinster, R. L., and McMorris, F. A. (1993). Insulin-like growth factor I increases brain growth and central nervous system myelination in transgenic mice. Neuron 10, 729–740. Casaccia-BonneWl, P., Tikoo, R., Kiyokawa, H., Friedrich, V., Jr., Chao, M. V., and KoV, A. (1997). Oligodendrocyte precursor diVerentiation is perturbed in the absence of the cyclin-dependent kinase inhibitor p27Kip1. Genes Dev. 11, 2335–2346. Cenci di Bello, I., Dawson, M. R. L., Levine, J. M., and Reynolds, R. (1999). Generation of oligodendroglial progenitors in acute inXammatory demyelinating lesions of the rat brain stem is stimulated by demyelination rather than inXammation. J. Neurocytol. 28, 365–381. Chang, A., Nishiyama, A., Peterson, J., Prineas, J., and Trapp, B. D. (2000). NG2-positive oligodendrocyte progenitor cells in adult human brain and multiple sclerosis lesions. J. Neurosci. 20, 6404–6412. Chang, A., Tourtellotte, W. W., Rudick, R., and Trapp, B. D. (2002). Premyelinating oligodendrocytes in chronic lesions of multiple sclerosis. New Eng. J. Med. 346, 165–173. Chari, D. M., and Blakemore, W. F. (2002a). EYcient recolonisation of progenitor-depleted areas of the CNS by adult oligodendrocyte progenitor cells. Glia 37, 307–313. Chari, D. M., and Blakemore, W. F. (2002b). New insights into remyelination failure in multiple sclerosis: Implications for glial cell transplantation. Multiple Sclerosis 8, 271–277. Charles, P., Hernandez, M. P., StankoV, B., Aigrot, M. S., Colin, C., Rougon, G., Zalc, B., and Lubetzki, C. (2000). Negative regulation of central nervous system myelination by polysialylated-neural cell adhesion molecule. Proc. Natl. Acad. Sci. USA 97, 7585–7590. Charles, P., Reynolds, R., Seilhean, D., Rougon, G., Aigrot, M. S., Niezgoda, A., Zalc, B., and Lubetzki, C. (2002). Re-expression of PSA-NCAM by demyelinated axons: An inhibitor or remyelination in multiple sclerosis? Brain 125, 1972–1979.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

PROSPECTS FOR THERAPEUTIC PROMOTION OF CNS REMYELINATION

Colello, R. J., Pott, U., and Schwab, M. E. (1994). The role of oligodendrocytes and myelin on axon maturation in the developing rat retinofugal pathway. J. Neurosci. 14, 2594–2605. Colognato, H., Baron, W., Avellana-Adalid, V., Relvas, J. B., Evercooren, A. B., Georges-Labouesse, E., and Vrench-Constant, C. (2002). CNS integrins switch growth factor signalling to promote target-dependent survival. Nat. Cell Biol. 4, 833–841. Dal Canto, M. C., and Barbano, R. L. (1984). Remyelination during remission in Theiler’s virus infection. Am. J. Pathol. 116, 30–45. Dal Canto, M. C., and Lipton, H. L. (1975). Primary demyelination in Theiler’s virus infection. An ultrastructural study. Lab. Invest. 33, 626–637. Dal Canto, M. C., and Lipton, H. L. (1980). Schwann cell remyelination and recurrent demyelination in the central nervous system of mice infected with attenuated Theiler’s virus. Am. J. Pathol. 98, 101–122. Dawson, M. R. L., Levine, J. M., and Reynolds, R. (2000). NG-2-expressing cells in the central nervous system: Are they oligodendroglial progenitors? J. Neurosci. Res. 61, 471–479. Decker, L., Durbec, P., Rougon, G., and Evercooren, A. B. (2002a). Loss of polysialic residues accelerates CNS neural precursor diVerentiation in pathological conditions. Mol. Cell. Neurosci. 19, 225–238. Decker, L., Picard-Riera, N., Lachapelle, F., and Baron-Van Evercooren, A. (2002b). Growth factor treatment promotes mobilization of young but not aged adult subventricular zone precursors in response to demyelination. J. Neurosci. Res. 69, 763–771. Demerens, C., StankoV, B., Logak, M., Anglade, P., Allinquant, B., Couraud, F., Zalc, B., and Lubetzki, C. (1996). Induction of myelination in the central nervous system by electrical activity. Proc. Natl. Acad. Sci. USA 93, 9887–9892. Duncan, I. D., and HoVman, R. L. (1997). Schwann cell invasion of the central nervous system of the myelin mutants. J. Anat. 190, 35–49. Durand, B., Fero, M. L., Roberts, J. M., and RaV, M. C. (1998). p27Kip1 alters the response of cells to mitogen and is part of a cell-intrinsic timer that arrests the cell cycle and initiates diVerentiation. Curr. Biol. 8, 431–440. Durand, B., Gao, F. B., and RaV, M. (1997). Accumulation of the cyclin-dependent kinase inhibitor p27/Kip1 and the timing of oligodendrocyte diVerentiation. EMBO J. 16, 306–317. Fanarraga, M. L., GriYths, I. R., Zhao, M., and Duncan, I. D. (1998). Oligodendrocytes are not inherently programmed to myelinate a speciWc size of axon. J. Comp. Neurol. 399, 94–100. Feigin, I., and Ogata, J. (1971). Schwann cells and peripheral myelin within human central nervous tissues: The mesenchymal character of Schwann cells. J. Neuropathol. Exp. Neurol. 30, 603–612. FitzGerald, U. F., and Barnett, S. C. (2000). AP-1 activity during the growth, diVerentiation, and death of O-2A lineage cells. Mol. Cell Neurosci. 16, 453–469. Fok-Seang, J., Mathews, G. A., Vrench-Constant, C., Trotter, J., and Fawcett, J. W. (1995). Migration of oligodendrocyte precursors on astrocytes and meningeal cells. Dev. Biol. 171, 1–15. Franklin, R. J. M. (1999). Remyelination – a regenerative process in the CNS. The Neuroscientist 5, 184–191. Franklin, R. J. M. (2002). Why does remyelination fail in multiple sclerosis? Nat. Rev. Neurosci. 3, 705–714. Franklin, R. J. M., and Blakemore, W. F. (1993). Requirements for Schwann cell migration within CNS environments: A viewpoint. Int. J. Dev. Neurosci. 11, 641–649. Franklin, R. J. M., Crang, A. J., and Blakemore, W. F. (1991). Transplanted type-1 astrocytes facilitate repair of demyelinating lesions by host oligodendrocytes in adult rat spinal cord. J. Neurocytol. 20, 420–430. Franklin, R. J. M., Gilson, J. M., and Blakemore, W. F. (1997). Local recruitment of remyelinating cells in the repair of demyelination in the central nervous system. J. Neurosci. Res. 50, 337–344. Franklin, R. J. M., and Hinks, G. L. (1999). Understanding CNS remyelination – clues from developmental and regeneration biology. J. Neurosci. Res. 58, 207–213. Franklin, R. J. M., Hinks, G. L., WoodruV, R. H., and O’Leary, M. T. (2001). What roles do growth factors play in CNS remyelination? Prog. Brain Res. 132, 185–193. Franklin, R. J. M., Zhao, C., and Sim, F. J. (2002). Ageing and CNS remyelination. NeuroReport 13, 923–928. Gallo, V., Zhou, J. M., McBain, C. J., Wright, P., Knutson, P. L., and Armstrong, R. C. (1996). Oligodendrocyte progenitor cell proliferation and lineage progression are regulated by glutamate receptor-mediated Kþ channel block. J. Neurosci. 16, 2659–2670. Gao, F. B., Apperly, J., and RaV, M. (1998). Cell-intrinsic timers and thyroid hormone regulate the probability of cell-cycle withdrawal and diVerentiation of oligodendrocyte precursor cells. Dev. Biol. 197, 54–66. Garcion, E., Faissner, A., and Vrench-Constant, C. (2001). Knockout mice reveal a contribution of the extracellular matrix molecule tenascin-C to neural precursor proliferation and migration. Development 128, 2485–2496. Genoud, S., Lappe-Siefke, C., Goebbels, S., Radtke, F., Aguet, M., Scherer, S. S., Suter, U., Nave, K. A., and Mantei, N. (2002). Notch1 control of oligodendrocyte diVerentiation in the spinal cord. J. Cell Biol. 158, 709–718. Gensert, J. M., and Goldman, J. E. (1997). Endogenous progenitors remyelinate demyelinated axons in the adult CNS. Neuron 19, 197–203. Gensert, J. M., and Goldman, J. E. (2001). Heterogeneity of cycling glial progenitors in the adult mammalian cortex and white matter. J. Neurobiol. 48, 75–86. Ghatak, N. R., Hirano, A., Doron, Y., and Zimmerman, H. M. (1973). Remyelination in multiple sclerosis with peripheral type myelin. Arch. Neurol. 29, 262–267.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

191

192

7. REMYELINATION BY ENDOGENOUS GLIA

Ghiani, C. A., Eisen, A. M., Yuan, X. Q., DePinho, R. A., McBain, C. J., and Gallo, V. (1999). Neurotransmitter receptor activation triggers p27Kip1 and p21CIP1 accumulation and G1 cell cycle arrest in oligodendrocyte progenitors. Development 126, 1077–1090. Givogri, M. I., Costa, R. M., Schonmann, V., Silva, A. J., Campagnoni, A. T., and Bongarzone, E. R. (2002). Central nervous system myelination in mice with deWcient expression of Notch1 receptor. J. Neurosci. Res. 67, 309–320. Gledhill, R. F., Harrison, B. M., and McDonald, W. I. (1973). Demyelination and remyelination after acute spinal cord compression. Exp. Neurol. 38, 472–487. Goddard, D. R., Berry, M., and Butt, A. M. (1999). In vivo actions of Wbroblast growth factor-2 and insulin-like growth factor-I on oligodendrocyte development and myelination in the central nervous system. J. Neurosci. Res. 57, 74–85. Goddard, D. R., Berry, M., Kirvell, S. L., and Butt, A. M. (2001). Fibroblast growth factor-2 inhibits myelin production by oligodendrocytes in vivo. Mol. Cell Neurosci. 18, 557–569. Graca, D. L., and Blakemore, W. F. (1986). Delayed remyelination in the rat spinal cord following ethidium bromide injection. Neuropath. App. Neurobiol. 12, 593–605. GriYths, I. R., and McCulloch, M. C. (1983). Nerve Wbres in spinal cord impact injuries. Part 1. Changes in the myelin sheath during the initial 5 weeks. J. Neurol. Sci. 58, 335–349. Groves, A. K., Entwistle, A., Jat, P. S., and Noble, M. (1993). The characterization of astrocyte cell lines that display properties of glial scar tisue. Dev. Biol. 159, 87–104. Gutowski, N. J., Newcombe, J., and Cuzner, M. L. (1999). Tenascin-R and C in multiple sclerosis lesions: Relevance to extracellular matrix remodelling. Neuropath. App. Neurobiol. 25, 207–214. Gveric, D., Cuzner, M. L., and Newcombe, J. (1999). Insulin-like growth factors and binding proteins in multiple sclerosis plaques. Neuropath. App. Neurobiol. 25, 215–225. Hall, S. M. (1972). The eVect of injections of lysophosphatidyl choline into white matter of the adult mouse spinal cord. J. Cell Sci. 10, 535–546. Harrison, B. M., McDonald, W. I., and Ochoa, J. (1972). Remyelination in the central diphtheria toxin lesion. J. Neurol. Sci. 17, 293–302. Herndon, R. M., Price, D. L., and Weiner, L. P. (1977). Regeneration of oligodendrocytes during recovery from demyelinating disease. Science 195, 693–694. Hinks, G. L., and Franklin, R. J. M. (1999). Distinctive patterns of PDGF-A, FGF-2, IGF-I and TGF-beta1 gene expression during remyelination of experimentally-induced spinal cord demyelination. Mol. Cell. Neurosci. 14, 153–168. Hinks, G. L., and Franklin, R. J. M. (2000). Delayed changes in growth factor gene expression during slow remyelination in the CNS of aged rats. Mol. Cell. Neurosci. 16, 542–556. Horner, P. J., Power, A. E., Kempermann, G., Kuhn, H. G., Palmer, T. D., Winkler, J., Thal, L. J., and Gage, F. H. (2000). Proliferation and diVerentiation of progenitor cells throughout the intact adult rat spinal cord. J. Neurosci. 20, 2218–2228. Itoyama, Y., Ohnishi, A., Tateishi, J., Kuroiwa, Y., and Webster, H. D. (1985). Spinal cord multiple sclerosis lesions in Japanese patients: Schwann cell remyelination occurs in areas that lack glial Wbrillary acidic protein (GFAP). Acta Neuropathol. 65, 217–223. JeVery, N. D., and Blakemore, W. F. (1997). Locomotor deWcits induced by experimental spinal cord demyelination are abolished by spontaneous remyelination. Brain 120, 27–37. Jiang, F., Frederick, T. J., and Wood, T. L. (2001). IGF-I synergizes with FGF-2 to stimulate oligodendrocyte progenitor entry into the cell cycle. Dev. Biol. 232, 414–423. John, G. R., Shankar, S. L., ShaWt-Zagardo, B., Massimi, A., Lee, S. C., Raine, C. S., and Brosnan, C. F. (2002). Multiple sclerosis: Re-expression of a developmental pathway that restricts oligodendrocyte maturation. Nat. Med. 8, 1115–1121. Johnson, E. S., and Ludwin, S. K. (1981). The demonstration of recurrent demyelination and remyelination of axons in the central nervous system. Acta Neuropathol. 53, 93–98. Jordan, C., Friedrich, V., and Dubois-Dalcq, M. (1989). In situ hybridization analysis of myelin gene transcripts in developing mouse spinal cord. J. Neurosci. 9, 248–257. Jordan, C. A., Friedrich, V. L., De Ferra, F., Weismiller, D. G., Holmes, K. V., and Dubois-Dalcq, M. (1990). DiVerential exon expression in myelin basic protein transcripts during central nervous system (CNS) remyelination. Cell Mol. Neurobiol. 10, 3–18. Keirstead, H. S., Ben-Hur, T., Rogister, B., O’Leary, M. T., Dubois-Dalcq, M., and Blakemore, W. F. (1999). Polysialylated neural cell adhesion molecule-positive CNS precursors generate both oligodendrocytes and Schwann cells to remyelinate the CNS after transplantation. J. Neurosci. 19, 7529–7536. Keirstead, H. S., and Blakemore, W. F. (1997). IdentiWcation of post-mitotic oligodendrocytes incapable of remyelination within the demyelinated adult spinal cord. J. Neuropath. Exp. Neurol. 56, 1191–1201. Kiernan, B. W., Go¨tz, B., Faissner, A., and Vrench-Constant, C. (1996). Tenascin-C inhibits oligodendrocyte precursor cell migration by both adhesion-dependent and adhesion-independent mechanisms. Mol. Cell. Neurosci. 7, 322–335. Komoly, S., Hudson, L. D., Webster, H. d., and Bondy, C. A. (1992). Insulin-like growth factor I gene expression is induced in astrocytes during experimental demyelination. Proc. Natl. Acad. Sci. USA 89, 1894–1898. Kondo, T., and RaV, M. (2000a). Basic helix-loop-helix proteins and the timing of oligodendrocyte diVerentiation. Development 127, 2989–2998.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

193

PROSPECTS FOR THERAPEUTIC PROMOTION OF CNS REMYELINATION

Kondo, T., and RaV, M. (2000b). The Id4 HLH protein and the timing of oligodendrocyte diVerentiation. EMBO J. 19, 1998–2007. Kornek, B., Storch, M. K., Weissert, R., Wallstroem, E., SteVerl, A., Olsson, T., Linington, C., Schmidbauer, M., and Lassmann, H. (2000). Multiple sclerosis and chronic autoimmune encephalomyelitis: A comparative quantitative study of axonal injury in active, inactive, and remyelinated lesions. Am. J. Pathol. 157, 267–276. Kotter, M. R., Setzu, A., Sim, F. J., van Rooijen, N., and Franklin, R. J. M. (2001). Macrophage depletion impairs oligodendrocyte remyelination following lysolecithin-induced demyelination. Glia 35, 204–212. Kristensson, K., Holmes, K. V., Duchala, C. S., Zeller, N. K., Lazzarini, R. A., and Dubois-Dalcq, M. (1986). Increased levels of myelin basic protein transcripts gene in virus-induced demyelination. Nature 322, 544–547. Lampert, P. W. (1965). Demyelination and remyelination in experimental allergic encephalomyelitis. J. Neuropath. Exp. Neurol. 24, 371–385. Lang, W., Rodriguez, M., Lennon, V. A., and Lampert, P. W. (1984). Demyelination and remyelination in murine viral encephalomyelitis. Ann. N. Y. Acad. Sci. 436, 98–102. Lassmann, H., Bruck, W., Lucchinetti, C., and Rodriguez, M. (1997). Remyelination in multiple sclerosis. Multiple Sclerosis 3, 133–136. Levine, J. M., and Reynolds, R. (1999). Activation and proliferation of endogenous oligodendrocyte precursor cells during ethidium bromide-induced demyelination. Exp. Neurol. 160, 333–347. Li, G., Crang, A. J., Rundle, J. L., and Blakemore, W. F. (2002). Oligodendrocyte progenitor cells in the adult rat CNS express myelin oligodendrocyte protein (MOG). Brain Pathol. 12, 463–471. Linington, C., Engelhardt, B., Kapocs, G., and Lassman, H. (1992). Induction of persistently demyelinated lesions in the rat following the repeated adoptive transfer of encephalitogenic T cells and demyelinating antibody. J. Neuroimmunol. 40, 219–224. Lipton, H. L. (1975). Theiler’s virus infection in mice: An unusual biphasic disease process leading to demyelination. Infect. Immun. 11, 1147–1155. Loughlin, A. J., Copelman, C. A., Hall, A., Armer, T., Young, B. C., Landon, D. N., and Cuzner, M. L. (1997). Myelination and remyelination of aggregate rat brain cell cultures enriched with macrophages. J. Neurosci. Res. 47, 384–392. Lu, Q. R., Sun, T., Zhu, Z., Ma, N., Garcia, M., Stiles, C. D., and Rowitch, D. H. (2002). Common developmental requirement for Olig function indicates a motor neuron/oligodendrocyte connection. Cell 109, 75–86. Lucchinetti, C., Bruck, W., Parisi, J., Scheithauer, B., Rodriguez, M., and Lassmann, H. (2000). Heterogeneity of multiple sclerosis lesions: Implications for the pathogenesis of demyelination. Ann. Neurol. 47, 707–717. Ludwin, S. K. (1979a). An autoradiographic study of cellular proliferation in remyelination of the central nervous system. Am. J. Pathol. 95, 683–676. Ludwin, S. K. (1979b). The perineuronal satellite oligodendrocyte. A role in remyelination. Acta Neuropath. 47, 49–53. Ludwin, S. K. (1980). Chronic demyelination inhibits remyelination in the central nervous system. Lab. Invest. 43, 382387. Ludwin, S. K. (1988). Remyelination in the central nervous system and the peripheral nervous system. In ‘‘Functional Recovery in Neurological Diseases–Advances in Neurology 47’’ (S. G. Waxman, ed.), pp. 215–254. Raven Press, New York. Ludwin, S. K., and Bakker, D. A. (1988). Can oligodendrocytes attached to nuclei proliferate? J. Neurosci. 8, 1239–1244. Ludwin, S. K., and Maitland, M. (1984). Long-term remyelination fails to reconstitute normal thickness of central myelin sheaths. J. Neurol. Sci. 64, 193–198. Ludwin, S. K., and Sternberger, N. H. (1984). An immunohistochemical study of myelin proteins during remyelination in the central nervous system. Acta Neuropathol. 63, 240–248. Maeda, Y., Solanky, M., Menonna, J., Chapin, J., Li, W., and Dowling, P. (2001). Platelet-derived growth factoralpha receptor-positive oligodendroglia are frequent in multiple sclerosis lesions. Ann. Neurol. 49, 776–785. Marburg, O. (1906). Die sogenannte ‘akute Multiple Sklerose’. Jahrbuch Psychiatrie 27, 211–312. Marin-Husstege, M., Muggironi, M., Liu, A., and Casaccia-BonneWl, P. (2002). Histone deacetylase activity is necessary for oligodendrocyte lineage progression. J. Neurosci. 22, 10333–10345. Mason, J. L., and Goldman, J. E. (2002). A2B5(þ) and O4(þ) Cycling Progenitors in the Adult Forebrain White Matter Respond DiVerentially to PDGF-AA, FGF-2, and IGF-1. Mol. Cell. Neurosci. 20, 30–42. Mason, J. L., Jones, J. J., Taniike, M., Morell, P., Suzuki, K., and Matsushima, G. K. (2000a). Mature oligodendrocyte apoptosis precedes IGF-I production and oligodendrocyte progenitor accumulation and diVerentiation during demyelination/remyelination. J. Neurosci. Res. 61, 251–262. Mason, J. L., Langaman, C., Morell, P., Suzuki, K., and Matsushima, G. K. (2001a). Episodic demyelination and subsequent remyelination within the murine central nervous system: Changes in axonal calibre. Neuropath. App. Neurobiol. 27, 50–58. Mason, J. L., Suzuki, K., Chaplin, D. D., and Matsushima, G. K. (2001b). Interleukin-1beta promotes repair of the CNS. J. Neurosci. 21, 7046–7052. Mason, J. L., Ye, P., Suzuki, K., D’Ercole, A. J., and Matsushima, G. K. (2000b). Isulin-like growth factor1 inhibits mature oligodendrocyte apoptosis during primary demyelination. J. Neurosci. 20, 5703–5708. Matsushima, G. K., and Morell, P. (2001). The neurotoxicant, cuprizone, as a model to study demyelination and remyelination in the central nervous system. Brain Path. 11, 107–116. Mckinnon, R. D., Matsui, T., Dubois-Dalcq, M., and Aaronson, S. A. (1990). FGF Modulates the PDGF-Driven Pathway of Oligodendrocyte Development. Neuron 5, 603–614.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

au3

194

7. REMYELINATION BY ENDOGENOUS GLIA

Mckinnon, R. D., Piras, G., Ida, J. A., Jr., and Dubois-Dalcq, M. (1993). A role for TGF-b in oligodendrocyte diVerentiation. J. Cell Biol. 121, 1397–1407. McMorris, F. A., and Mckinnon, R. D. (1996). Regulation of oligodendrocyte development and CNS myelination by growth factors: Prospects for therapy of demyelinating disease. Brain Pathol. 6, 313–329. Messersmith, D. J., Murtie, J. C., Le, T. Q., Frost, E. E., and Armstrong, R. C. (2000). Fibroblast growth factor 2 (FGF2) and FGF receptor expression in an experimental demyelinating disease with extensive remyelination. J. Neurosci. Res. 62, 241–256. Miller, D. J., Asakura, K., and Rodriguez, M. (1996). Central nervous system remyelination clinical application of basic neuroscience principles. Brain Pathol. 6, 331–344. Miller, D. J., and Rodriguez, M. (1995). Spontaneous and induced remyelination in multiple sclerosis and the Theiler’s virus model of central nervous system demyelination. Microsc. Res. Tech. 32, 230–245. Miller, D. J., and Rodriguez, M. (1996). Spontaneous central nervous system remyelination in strain A mice after infection with the Daniel’s (DA) strain of Theiler’s virus. Acta Neuropathol. 91, 559–565. Miller, D. J., Sanborn, K. S., Katzmann, J. A., and Rodriguez, M. (1994). Monoclonal autoantibodies promote central nervous system repair in an animal model of multiple sclerosis. J. Neurosci. 14, 6230–6238. Morell, P., Barrett, C. V., Mason, J. L., Toews, A. D., Hostettler, J. D., Knapp, G. W., and Matsushima, G. K. (2000). Gene expression in brain during cuprizone-induced demyelination and remyelination. Mol. Cell. Neurosci. 12, 220–227. Mujtaba, T., Mayer-Pro¨schel, M., and Rao, M. S. (1998). A common neural progenitor for the CNS and PNS. Dev. Biol. 200, 1–15. Murray, J. A., and Blakemore, W. F. (1980). The relationship between internodal length and Wbre diameter in the spinal cord of the cat. J. Neurol. Sci. 45, 29–41. Nait-Oumesmar, B., Decker, L., Lachapelle, F., Avellana-Adalid, V., Bachelin, C., and Van Evercooren, A. B. (1999). Progenitor cells of the adult mouse subventricular zone proliferate, migrate and diVerentiate into oligodendrocytes after demyelination. Eur. J. Neurosci. 11, 4357–4366. Nait-Oumesmar, B., Vignais, L., Duhamel-Cle´rin, E., Avellana-Adalid, V., Rougon, G., and Baron-Van Evercooren, A. (1995). Expression of the highly polysialylated neural cell adhesion molecule during postnatal myelination and following chemically induced demyelination of the adult mouse spinal cord. Eur. J. Neurosci. 7, 480–491. Niehaus, A., Shi, J., Grzenkowski, M., Diers-Fenger, M., Archelos, J., Hartung, H.-P., Toyka, K., Bruck, W., and Trotter, J. (2000). Patients with active relapsing-remitting multiple sclerosis synthesize antibodies recognizing oligodendrocyte progenitor cell surface protein: Implications for remyelination. Ann. Neurol. 48, 362–371. O’Leary, M. T., Hinks, G. L., Charlton, H. M., and Franklin, R. J. M. (2002). Increasing local levels of IGF-I mRNA expression using adenoviral vectors does not alter oligodendrocyte remyelination in the CNS of aged rats. Mol. Cell. Neurosci. 19, 32–42. Olby, N. J., and Blakemore, W. F. (1996). Primary demyelination and regeneration of ascending axons in the dorsal funiculus of the rat spinal cord following photochemically-induced injury. J. Neurocytol. 25, 465–480. Palmer, A. C., Jackson, P. G., and Blakemore, W. F. (1991). A primary demyelinating disorder of young cattle. Neuropathol. Appl. Neurobiol. 17, 457–467. Park, S. K., Miller, R., Krane, I., and Vartanian, T. (2001). The erbB2 gene is required for the development of terminally diVerentiated spinal cord oligodendrocytes. J. Cell Biol. 154, 1245–1258. Penderis, J., Shields, S. A., and Franklin, R. J. M. (2003). Impaired remyelination and depletion of oligodendrocyte progenitors does not occur following repeated episodes of focal demyelination in the rat CNS. Brain. In press. Perier, O., and Gregoire, A. (1965). Electron microscopic features of multiple sclerosis lesions. Brain 88, 937–952. Picard-Riera, N., Decker, L., Delarasse, C., Goude, K., Nait-Oumesmar, B., Liblau, R., Pham-Dinh, D., and Evercooren, A. B. (2002). Experimental autoimmune encephalomyelitis mobilizes neural progenitors from the subventricular zone to undergo oligodendrogenesis in adult mice. Proc. Natl. Acad. Sci. USA 99, 13211–13216. Prayoonwiwat, N., and Rodriguez, M. (1993). The potential for oligodendrocyte proliferation during demyelinating disease. J. Neuropath. Exp. Neurol. 52, 55–63. Prineas, J. W., Barnard, R. O., Kwon, E. E., Sharer, L. R., and Cho, E.-S. (1993a). Multiple sclerosis: Remyelination of nascent lesions. Ann. Neurol. 33, 137–151. Prineas, J. W., Barnard, R. O., Revesz, T., Kwon, E. E., Sharer, L., and Cho, E.-S. (1993b). Multiple sclerosis: Pathology of recurrent lesions. Brain 116, 681–693. Prineas, J. W., and Connell, F. (1979). Remyelination in multiple sclerosis. Ann. Neurol. 5, 22–31. Prineas, J. W., Kwon, E. E., Goldenberg, P. Z., Ilyas, A. A., Quarles, R. H., Benjamins, J. A., and Sprinkle, T. J. (1989). Multiple sclerosis: Oligodendrocyte proliferation and diVerentiation in fresh lesions. Lab. Invest. 61, 489–503. Raine, C. S., Traugott, U., and Stone, S. H. (1978). Glial bridges and Schwann cell migration during chronic demyelination in the CNS. J Neurocytol. 7, 541–553. Raine, C. S., Wayne Moore, G. R., Hintzer, R., and Traugott, U. (1988). Induction of oligodendrocyte proliferation and remyelination after chronic demyelination. Lab. Invest. 59, 467–476. Raine, C. S., and Wu, E. (1993). Multiple sclerosis: Remyelination in acute lesions. J. Neuropath. Exp. Neurol. 52, 199–204. Redwine, J. M., and Armstrong, R. C. (1998). In vivo proliferation of oligodendrocyte progenitors expressing PDGF-alphaR during early remyelination. J. Neurobiol. 37, 413–428.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

PROSPECTS FOR THERAPEUTIC PROMOTION OF CNS REMYELINATION

Relvas, J. B., Setzu, A., Baron, W., Buttery, P. C., LaFlamme, S. E., Franklin, R. J. M., and Vrench-Constant, C. (2001). Expression of dominant-negative and chimeric subunits reveals an essential role for beta1 integrin during myelination. Curr. Biol. 11, 1039–1043. Reynolds, R., and Wilkin, G. P. (1993). Cellular reaction to an acute demyelinating/remyelinating lesion of the rat brain stem: Localisation of GD3 ganglioside immunoreactivity. J. Neurosci. Res. 36, 417–434. Richardson, W. D. (2001). Oligodendrocyte development. In ‘‘Glial Cell Development’’ (K. R. Jessen and W. D. Richardson, eds.), pp. 21–54. Oxford Univeristy Press, Oxford. Robinson, S., and Miller, R. (1996). Environmental enhancement of growth factor-mediated oligodendrocyte precursor proliferation. Mol. Cell. Neurosci. 8, 38–52. Robinson, S., and Miller, R. H. (1999). Contact with central nervous system myelin inhibits oligodendrocyte progenitor maturation. Dev. Biol. 216, 359–368. Robinson, S., Tani, M., Strieter, R. M., RansohoV, R. N., and Miller, R. H. (1998). The chemokine growthregulated oncogene-a promotes spinal cord oligodendrocyte precursor proliferation. J. Neurosci. 18, 10457– 10463. Rosenthal, S. M., and Cheng, Z. Q. (1995). Opposing early and late eVects of insulin-like growth factor I on diVerentiation and the cell cycle regulatory retinoblastoma protein in skeletal myoblasts. Proc. Natl. Acad. Sci. USA 92, 10307–10311. RuYni, F., Furlan, R., Poliani, P. L., Brambilla, E., Marconi, P. C., Bergami, A., Desina, G., Glorioso, J. C., Comi, G., and Martino, G. (2001). Fibroblast growth factor-II gene therapy reverts the clinical course and the pathological signs of chronic experimental autoimmune encephalomyelitis in C57BL/6 mice. Gene Ther. 8, 1207–1213. Saluja, I., Granneman, J. G., and SkoV, R. P. (2001). PPAR delta agonists stimulate oligodendrocyte diVerentiation in tissue culture. Glia 33, 191–204. Sasaki, M., and Ide, C. (1989). Demyelination and remyelination in the dorsal funiculus of the rat spinal cord after heat injury. J. Neurocytol. 18, 225–239. Sa´nchez, I., Hassinger, L., Paskevich, P. A., Shine, H. D., and Nixon, R. A. (1996). Oligodendroglia regulate the regional expansion of axon caliber and local accumulation of neuroWlaments during development independently of myelin formation. J. Neurosci. 16, 5095–5105. Scolding, N., Franklin, R., Stevens, S., HELDIN, C. H., Compston, A., and Newcombe, J. (1998). Oligodendrocyte progenitors are present in the normal adult human CNS and in the lesions of multiple sclerosis. Brain 121, 2221–2228. Shi, J. Y., Marinovich, A., and Barres, B. A. (1998). PuriWcation and characterization of adult oligodendrocyte precursor cells from the rat optic nerve. J. Neurosci. 18, 4627–4636. Shields, S. A., Blakemore, W. F., and Franklin, R. J. M. (2000). Schwann cell remyelination is restricted to astrocyte-deWcient areas following transplantation into demyelinated adult rat brain. J. Neurosci. Res. 60, 571–578. Shields, S. A., Gilson, J. M., Blakemore, W. F., and Franklin, R. J. M. (1999). Remyelination occurs as extensively but more slowly in old rats compared to young rats following gliotoxin-induced CNS demyelination. Glia 28, 77–83. Sim, F. J., Hinks, G. L., and Franklin, R. J. M. (2000). The re-expression of the homeodomain transcription factor Gtx during remyelination of experimentally-induced demyelinating lesions in young and old rat brain. Neurosci. 100, 131–139. Sim, F. J., Zhao, C., Li, W.-W., Lakatos, A., and Franklin, R. J. M. (2002a). Expression of the POU domain transcription factors SCIP/Oct-6 and Brn-2 is associated with Schwann cell but not oligodendrocyte remyelination of the CNS. Mol. Cell. Neurosci. 20, 669–682. Sim, F. J., Zhao, C., Penderis, J., and Franklin, R. J. M. (2002b). The age-related decrease in CNS remyelination eYciency is attributable to an impairment of both oligodendrocyte progenitor recruitment and diVerentiation. J. Neurosci. 22, 2451–2459. Smith, K. J., Blakemore, W. F., and McDonald, W. I. (1979). Central remyelination restores secure conduction. Nature 280, 395–396. Smith, K. J., Blakemore, W. F., Murray, J. A., and Patterson, R. C. (1982). Internodal myelin volume and axon surface area: A relationship determining myelin thickness? J. Neurol. Sci. 55, 231–246. Smith, K. J., and McDonald, W. I. (1999). The pathophysiology of multiple sclerosis: The mechanisms underlying the production of symptoms and the natural history of the disease. Phil. Trans. R. Soc. Lond. B 354, 1649–1673. Snyder, D. H., Valsamis, M. P., Stone, S. H., and Raine, C. S. (1975). Progressive demyelination and reparative phenomena in chronic experimental allergic encephalomyelitis. J. Neuropathol. Exp. Neurol. 34, 209–221. Sobel, R. A., Chen, M., Maeda, A., and Hinojoza, J. R. (1995). Vitronectin and integrin vitronectin receptor localization in multiple sclerosis lesions. J. Neuropathol. Exp. Neurol. 54, 202–213. Stangel, M., Compston, A., and Scolding, N. J. (1999). Polyclonal immunoglobulins for intravenous use do not inXuence the behaviour of cultured oligodendrocytes. J. Neuroimmunol. 96, 228–233. Stolt, C. C., Rehberg, S., Ader, M., Lommes, P., Riethmacher, D., Schachner, M., Bartsch, U., and Wegner, M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev. 16, 165–170. Suzuki, K., Andrews, J. M., Waltz, J. M., and Terry, R. D. (1969). Ultrastructural studies of multiple sclerosis. Lab. Invest. 20, 444–454.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

195

196

7. REMYELINATION BY ENDOGENOUS GLIA

Targett, M. P., Sussman, J., Scolding, N., O’Leary, M. T., Compston, D. A. S., and Blakemore, W. F. (1996). Failure to remyelinate rat axons following transplantation of glial cells obtained from adult human brain. Neuropath. App. Neurobiol. 22, 199–206. Tourbah, A., Baron van Evercooren, A., Oliver, L., Raulais, D., Jeanny, J. C., and Gumpel, M. (1992). Endogenous aFGF expression and cellular changes after a demyelinating lesion in the spinal cord of adult normal mice: Immunohistochemical study. J. Neurosci. Res. 33, 47–59. Traugott, U., Stone, S. H., and Raine, C. S. (1982). Chronic relapsing experimental autoimmune encephalomyelitis. treatment with combinations of myelin components promotes clinical and structural recovery. J. Neurol. Sci. 56, 65–73. Wang, S., Sdrulla, A., Johnson, J. E., Yokota, Y., and Barres, B. A. (2001). A role for the helix-loop-helix protein Id2 in the control of oligodendrocyte diVerentiation. Neuron 29, 603–614. Wang, S., Sdrulla, A. D., diSibio, G., Bush, G., Nofziger, D., Hicks, C., Weinmaster, G., and Barres, B. A. (1998). Notch receptor activation inhibits oligodendrocyte diVerentiation. Neuron 21, 63–75. Warrington, A. E., Asakura, K., Bieber, A. J., Ciric, B., Van, K., V, Kaveri, S. V., Kyle, R. A., Pease, L. R., and Rodriguez, M. (2000). Human monoclonal antibodies reactive to oligodendrocytes promote remyelination in a model of multiple sclerosis. Proc. Natl. Acad. Sci. USA 97, 6820–6825. Watanabe, M., Toyama, Y., and Nishiyama, A. (2002). DiVerentiation of proliferated NG2-positive glial progenitor cells in a remyelinating lesion. J. Neurosci. Res. 69, 826–836. Waxman, S. G., Kocsis, J. D., and Nitta, K. C. (1979). Lysophosphatidyl choline-induced focal demyelination in the rabbit corpus callosum. Light-microscopic observations. J. Neurol. Sci. 44, 45–53. Windrem, M. S., Roy, N. S., Wang, J., Nunes, M., Benraiss, A., Goodman, R., McKhann, G. M., and Goldman, S. A. (2002). Progenitor cells derived from the adult human subcortical white matter disperse and diVerentiate as oligodendrocytes within demyelinated lesions of the rat brain. J. Neurosci. Res. 69, 966–975. Wolswijk, G. (1998). Chronic stage multiple sclerosis lesions contain a relatively quiescent population of oligodendrocyte precursor cells. J. Neurosci. 18, 601–609. Wolswijk, G. (1999). Oligodendrocyte regeneration in the adult rodent CNS and the failure of this process in multiple sclerosis. Prog. Brain Res. 117, 233–247. Wolswijk, G. (2002). Oligodendrocyte precursor cells in the demyelinated multiple sclerosis spinal cord. Brain 125, 338–349. Wolswijk, G., and Noble, M. (1992). Cooperation between PDGF and FGF converts slowly dividing O-2Aadult progenitors to rapidly dividing cells with characteristics of O-2Aperinatal progenitor cells. J. Cell Biol. 118, 889–900. Wood, P. M., and Bunge, R. P. (1991). The origin of remyelinating cells in the adult central nervous system: The role of the mature oligodendrocyte. Glia 4, 225–232. WoodruV, R. H., and Franklin, R. J. M. (1997). Growth factors and remyelination in the CNS. Histol. Histopath. 12, 459–466. WoodruV, R. H., and Franklin, R. J. M. (1999a). Demyelination and remyelination of the caudal cerebellar peduncle of adult rats following stereotaxic injections of lysolecithin, ethidium bromide and complement/antigalactocerebroside – a comparative study. Glia 25, 216–228. WoodruV, R. H., and Franklin, R. J. M. (1999b). The expression of myelin protein mRNAs during remyelination of lysolecithin-induced demyelination. Neuropath. App. Neurobiol. 25, 226–235. Yajima, K., and Suzuki, K. (1979). Demyelination and remyelination in the rat central nervous system following ethidium bromide injection. Lab. Invest. 5, 385–392. Yamamoto, T., Kawamura, J., Hashimoto, S., and Nakamura, M. (1991). Extensive proliferation of peripheral type myelin in necrotic spinal cord lesions of multiple sclerosis. J. Neurol. Sci. 102, 163–169. Yao, D.-L., Liu, X., Hudson, L. D., and Webster, H. d. (1995). Insulin-like growth factor I treatment reduces demyelination and up-regulates gene expression of myelin-related proteins in experimental autoimmune encephalomyelitis. Proc. Natl. Acad. Sci. USA 92, 6190–6194. Yao, D. L., Liu, X., Hudson, L. D., and Webster, H. D. (1996). Insulin-like growth factor-I given subcutaneously reduces clinical deWcits, decreases lesion severity and upregulates synthesis of myelin proteins in experimental autoimmune encephalomyelitis. Life Sci. 58, 1301–1306. Ye, P., Carson, J., and D’Ercole, A. J. (1995). In vivo actions of insulin-like growth factor-I (IGF-I) on brain myelination: Studies of IGF-I and IGF binding protein-1 (IGFBP-1) transgenic mice. J. Neurosci. 15, 7344– 7356. Zhang, S. C., Ge, B., and Duncan, I. D. (1999). Adult brain retains the potential to generate oligodendroglial progenitors with extensive myelination capacity. Proc. Natl. Acad. Sci. USA 96, 4089–4094.

I. GLIAL CELL AND MYELIN FUNCTIONAL BIOLOGY

C H A P T E R

8 Invertebrate Glia Hugo J. Bellen and Karen L. Schulze

INTRODUCTION It is generally believed that there is no myelin in invertebrates, so most readers will wonder why a chapter is devoted to invertebrate glia in a book titled Myelin Biology and Disorders. We have learned much about the development of Xies in the past 10 years and by extension also about the biology of glia. The question for Xy biologists is, can our knowledge of Xy glia translate easily to vertebrate glia? This question is not readily answered, because of the vast morphological and functional diversity of glial cells and the recent quest for key proteins that specify glial cells and play a role in their diVerentiation. Many invertebrates have glial cells. Glial cells have been documented in arthropods such as Xies and crayWsh, annelids (earthworms and leeches), and mollusks including snails like Lymnaca, Aplysia, and squid (Carlson and Saint Marie, 1990; Hoyle, 1986; Lane, 1981; Pentreath, 1989). Some nematodes and Xatworms seem to have glia or glial-like cells, but they are diYcult to classify unambiguously as glial cells (Chanal and Labouesse, 1997). C. elegans has no true glial cells. We therefore anticipate that we will learn little of relevance to vertebrate glial biology from studies in nematodes. Furthermore, for any invertebrate species to reveal important molecular aspects of vertebrate glial biology, evolutionary conservation must apply. Two scenarios can be envisioned; in the Wrst case, one would assume that the common ancestor of vertebrates and invertebrates possessed glial cells. If the molecular pathways by which glia are speciWed, diVerentiate, and function in vertebrates and invertebrates have been evolutionarily conserved, then much of what we learn in Xies would be directly applicable to vertebrate glial biologists (Erwin and Davidson, 2002). Alternatively, if glial cells did not exist in the common ancestor of invertebrates and vertebrates, the need for glial cells to facilitate and enhance nervous system function would have induced or promoted the development of glial cells in the progeny of the common ancestor. Likewise, the absence of glia in C. elegans may be explained by this possibility, as these animals and other species may have never developed a need for glial cells. In this situation, a few molecules and pathways regulating glial function and development may show striking similarities between Xies and vertebrates; however, these pathways would not have been maintained through evolutionary conservation. Rather, they developed through evolutionary convergence. We would then anticipate that fewer pathways will show similarities in this latter scenario than in the Wrst scenario, and we would assume that many of the pathways and proteins used to develop glial cells and their speciWc functions would be diVerent. In this chapter, we will draw parallels between Xy glia and vertebrate glia wherever possible, and we will critically evaluate the body of knowledge summarized in this chapter in the discussion.

Myelin Biology and Disorders, Volume 1

199

Copyright 2004, Elsevier (USA). All rights reserved.

200

8. INVERTEBRATE GLIA

The structure and function of insect glia was reviewed by Carlson and Saint Marie (1990) from a historical perspective. Much of the work prior to 1985 through 1990 was done in the metathoracic ganglia of locusts (Hoyle, 1986), the Wrst optic ganglia of adult Diptera (Saint Marie and Carlson, 1985), and the ventral nerve cord and smaller nerves of a variety of arthropods, including cockroaches, grasshoppers, and crayWsh (Lane, 1981). However, most of our progress on the knowledge of invertebrate glia in the past 10 to 12 years has been made in Drosophila melanogaster, the invertebrate species that is most commonly used to discover the genes and proteins required for development and function. Many genes, proteins, and pathways have been shown to control Xy development, and these insights often provide a stepping stone to identify the molecules that may be required in vertebrate biology. This chapter is outlined to provide vertebrate glial biologists with a knowledge of Xy glial biology that will allow them to quickly access the Xy literature on this topic. The chapter also presents a general update of our current knowledge of glial biology in Xies, with an emphasis on the development of glia and their known roles in the nervous systems of Xies.

GLIA IN FLIES: TYPES, CLASSIFICATION, AND COMPARISON TO VERTEBRATE GLIA Embryonic and First Instar Glia CNS Glia Glial cells constitute about 10% of cells of the central nervous system (CNS) in the Xy embryo and Wrst instar larva (Fig. 8.1A). They have been carefully classiWed and described by Ito et al. (1995) using molecular markers that label speciWc subsets of glial cells. There are three major classes of glia: the surface-, cortex-, and neuropil-associated glia (Fig. 8.1C). About 60 glial cells for each segment of the nervous system or neuromere have been detected and described in late-stage embryos and Wrst-instar larvae. The surface-associated glia are critical for the insulation of the CNS from the hemolymph, although classiWcation of these large round Xat cells remains controversial. Unlike vertebrates, the nervous system of insects is submerged in hemolymph, and hence a barrier must be established between the neurons and this Xuid, as it often contains a relatively high concentration of Kþ ions (Hoyle, 1952). This high concentration of potassium ions does not permit neuronal function without insulation (Auld et al., 1995; Baumgartner et al., 1996; Bellen et al., 1998). The surface of the CNS is therefore covered by a perineurium. It consists of an acellular lamella (Scharrer, 1939) below which reside the subperineural glia (Figs. 8.1B and 8.1C). Some authors do not classify these large, round, and Xat cells as glia, whereas others do (Carlson and Saint Marie, 1990; Ito et al., 1995). In addition, channel glia line the channels that traverse the CNS from dorsal to ventral (Fig. 8.1C). The surfaceassociated glia are most similar to the ependymal cells in vertebrates that cover the outer surface of the nervous system. The cortex-associated glia are situated among the neural cell bodies in the cortex and are typically called cell body glia (Figs. 8.1B and 8.1C). There are typically 10 of these cells per neuromere. They are most similar to astrocytes but do not express glial Wbrillary acidic protein (GFAP; Eng and DeArmond, 1982). The neuropil-associated glia include several classes that are associated with axonal structures: the nerve root glia with the nerve roots; the interface or longitudinal glia with the longitudinal connectives (Figs. 8.1B and 8.1C); and the midline glia, the most common class, with the anterior and posterior commissures that connect the left and right longitudinal tracts of the CNS. The neuropil-associated glia are comparable to oligodendrocytes even though they do not form myelin. PNS Glia Two major nerves exit from each hemi-neuromere in the Drosophila embryo (Figs. 8.1C and 8.2). The PNS nerves of each segment carry sensory information gathered by PNS organs from the periphery to the CNS, and action potentials generated in the CNS

GLIA IN FLIES: TYPES, CLASSIFICATION, AND COMPARISON TO VERTEBRATE GLIA

201

FIGURE 8.1. Embryonic CNS glia. (A) Schematic drawing of the late embryonic CNS. A cross section at the level of the Wfth abdominal segment (line) is shown in part B. (B) The structures protruding in the foreground correspond to the neuropil, which mostly consists of axons and synapses. No glial cells are present in the neuropil in the embryo and Wrst instar larva. (C) Schematic drawing of part B showing the organization of many of the glial cells in cross section. These Wgures were adapted from Ito et al., 1995.

motorneurons to the muscles. Figure 8.2 depicts the anterior intersegmental nerve and segmental nerve as well as the peripheral neurons of a single abdominal segment. The neurons are subdivided in three major clusters (the muscles are not shown). Each nerve is associated with a few glial cells (hatched in Fig. 8.2): the exit glia and three to Wve peripheral glial cells (PG). These cells wrap around individual axons (see also Fig. 8.4) and can be compared to Schwann cells in vertebrates.

Adult Glia CNS Glia Adult glial cells have not been well categorized in Drosophila. In addition, no systematic description or classiWcation has been conducted and applied. Ramo´n y Cajal and Sanchez (1915) were the Wrst to describe adult insect glial cells in the optic system of Musca using Golgi staining procedures. Hence, many of the diVerent glial cell types described here are inferred to exist in Drosophila based on studies in other insects including Musca, bees, locusts, and Rhodnius. Here we will use the classiWcation established by Strausfeld (1976) and try to relate adult glia nomenclature to embryonic glia classiWcation (Fig. 8.3). Adult perineural glia are surrounded by a neural lemma and are thought to form the blood brain barrier. They are clearly comparable to the embryonic surfaceassociated glia.

202

8. INVERTEBRATE GLIA

FIGURE 8.2 Embryonic PNS glia. One abdominal hemi-segment of the Drosophila embryo is shown. Hatched cells represent peripheral glia. The shaded cells correspond to neurons. Abbreviations: PG, peripheral glia.

FIGURE 8.3 Forms of glia present in the adult CNS. This schematic diagram depicts the classes of glia present in the adult Drosophila brain. Adapted from Strausfeld (1976).

203

GLIA IN FLIES: TYPES, CLASSIFICATION, AND COMPARISON TO VERTEBRATE GLIA

FIGURE 8.4 PNS glia ensheathing sensory and motor axons. Diagram of a cross section of a peripheral nerve of an adult Xy. PNS glia (g) extend invaginations to enfold sensory and motor axons (center white circles). The light gray area corresponds to the peripheral glia. The outer white area is the cytoplasm of the perineural glia. The perineural glial cells (p) surrounding the complex are sealed with septate junctions. Adapted from Hertwick (1931).

The glia internal to the perineural glia have been categorized into four classes (Fig. 8.3). The cells labeled class 1 by Strausfeld (1976) probably correspond to the subperineural glia (surface-associated glia) described by Ito et al. (1995) in the embryonic nervous system. Class 1 neuroglia are tightly apposed to the perineural glia and are joined by septate junctions similar to those observed in epithelial cells in molluscs and Hydra (Flower, 1971). Class 2 neuroglia reside in the periphery of the brain rind. They probably correspond to the cortex-associated glia and some of the interface glia (neuropil-associated glia) for the longitudinal connectives. These glial cells partially ensheath the perikarya and neurites of neurons. Class 3 neuroglia include the nerve root, interface, and midline forms of neuropilar glia; these glia invade the synaptic neuropil. According to Ito et al. (1995), the embryonic ventral nerve cord does not have glia comparable to adult class 3 neuroglia; however, a few such cells have been observed in the embryonic brain. Finally, class 4 neuroglia include all glial cells that ensheath neurites, similar to oligodendrocytes. They correspond to neuropil-associated glia of embryos. PNS Glia The PNS glia consist of two cell types: perineural glia and PNS glia. As shown in Figure 8.4, the PNS glia wrap around individual sensory and motor axons. Perineural glia encircle the entire complex of axons and PNS glia and form septate junctions with themselves; this forms a tight seal, which is the basis of the barrier between the nerve and hemolymph. Some authors describe speciWc cells found in PNS organs as glia. PNS organs convey information akin to touch, stretch, or taste. Each PNS organ typically consists of four cells that are derived through a few divisions from a sensory organ precursor (SOPI; see Fig. 8.8B for cell lineage). The cell that is most closely associated with a neuron is sometimes called a glial cell. This nomenclature is substantiated by the observation that these cells are missing and sometimes transformed into neurons in embryos that lack the gene glial cells missing, a major molecular determinant for glial cells in Xies (Hosoya et al., 1995; Jones et al., 1995; Kania et al., 1995; Vincent et al., 1996). However, because of the special

au1

204

8. INVERTEBRATE GLIA

nature of these cells and their highly specialized function, we will not discuss the cells associated with PNS organs in this chapter.

THE ORIGIN OF GLIAL CELLS The Origin of Embryonic Glia Much of what we know about cellular lineages of glia in invertebrates comes from studies in Drosophila embryos. Almost all embryonic glial cells of the CNS are derived from neural stem cells in the ventral ectoderm: neuroglioblasts and glioblasts. Neuroglioblasts give rise to neurons and glial cells, whereas glioblasts only produce glial cells. As shown in Figure 8.5, the blast cells delaminate from ectodermal proneural clusters that are located lateral to the midline during germ band extension (Fig. 8.6A). These clusters express one of the proneural genes that encode bHLH proteins under the control of genes that are expressed early in development such as segment polarity genes (Hassan and Bellen, 2000; Jan and Jan, 1993; Skeath, 1999). A single cell in the proneural cluster expresses higher levels of proneural protein because of lateral inhibition mediated by the Notch pathway (discussed later) and becomes fated to develop as a neuroblast, neuroglioblast, or a glioblast. The

FIGURE 8.5 Origin of neuro-, neuroglio-, and glioblasts in the Drosophila embryo. Neuroblast, neuroglioblast, and glioblast lineages are derived from the embryonic ectoderm. Patches of cells are resolved by positional cues and exhibit increased expression of basic helix-loop-helix proteins (darkest gray cells in top panel). This expression determines which ectodermal cells are competent to undergo further levels of speciWcation to become blast cells (black, lower panels). Abbreviations: NB, neuroblast; GB, glioblast; NGB, neuroglioblast; GMC, ganglion mother cell.

THE ORIGIN OF GLIAL CELLS

FIGURE 8.6 Derivation of embryonic CNS glia and midline glia. (A) Embryonic CNS glia. Within the developing ventral nerve cord of an embryo during germ band extension, three rows of cells on either side of the midline form glioblasts, neuroglioblasts, or neuroblasts. (B) Midline glia. The subset of CNS glia called the midline glia are speciWed from the ventral mesoectodermal cells (black). These cells are located along the ventral periphery of the embryo at cellular blastoderm (shown in cross section). During gastrulation, these cells come together as a result of the invagination to form the mesoderm. Four cells on each side fuse and divide once to form 16 cells. The three anterior-most pairs will become midline glia, whereas the Wve posterior pairs develop into neurons.

neuroglioblast then divides to give rise to a ganglion mother cell that divides to form neurons and glia, whereas the glioblast produces only glia (see also Fig. 8.8). Each ganglion mother cell inherits a diVerent combination of factors to determine a unique pattern of neurons and glial cells. The majority of the glial cells are derived from glioblasts located in the most lateral row of blast cells (Fig. 8.6A), which are in proximity to the mesectodermal cells (Fig. 8.6B; Jones, 2001). Only 2 of the 30 blast precursors per hemi-segment are glioblasts, whereas 5 to 7 correspond to neuroglioblasts (Jones, 2001). The midline glia are derived from mesectodermal cells and their origin and fate in Drosophila have been extensively reviewed by Jacobs (2000). These glia are among the best characterized cells in the Xy. The mesectoderm consists of a single row of cells that forms the boundary between the ectoderm and the mesoderm (Fig. 8.6B). These cells migrate ventrally to form the midline of the Xy embryo, creating a boundary that is observed in all bilaterally symmetric organisms. The vertebrate tissues with equivalent function are the notochord and the Xoorplate. Like the mesectoderm, these vertebrate structures play mostly a patterning function after which they disappear (discussed later).

205

206

8. INVERTEBRATE GLIA

Four mesectodermal cells from each hemi-segment (eight cells) meet in the center. These cells divide once to form 16 cells. The six most anterior cells in each segment become midline glial cells, whereas the others give rise to neuronal cell lineages (Bossing and Technau, 1994; Dong and Jacobs, 1997; Klambt et al., 1991). Once these cells have played a major role in patterning, many die, and the remaining cells proliferate to form midline glial cells that ensheath the commissures. The precise lineage of subperineural glia is not known. They are probably of mesodermal origin, as they are absent in twist mutants, which lack mesodermal cells (Edwards et al., 1993). In summary, the majority of the CNS glial cells are derived from neurectodermal precursors. However, the midline glia arise from mesectoderm, and the subperineural glia from mesoderm. Hence, unlike neurons, glial cells are probably derived from the three diVerent germ layers. The vast majority of embryonic PNS glia are derived from glial cells born in the CNS. Six to eight of the PNS glia present in each embryonic segment are derived from two lateral neuroglioblasts (Schmidt et al., 1997). One PNS glial cell is born in the periphery (neurectoderm) and essentially remains in this position and wraps around the axons that reach this cell (Campbell et al., 1994; Halter et al., 1995). The other cells, initially born in the CNS, migrate to the CNS/PNS border where they Wrst form a coneshaped structure at the site where the growth cones of the motor neurons will exit the CNS (see Fig. 8.1C). Axons from the motor neurons and peripherally located PNS neurons that grow into the CNS contact these cells before their neurites venture into the periphery or the CNS, respectively (Sepp et al., 2000). This situation reminds us of the vertebrate CNS/PNS ventral transition zone and the dorsal root entry zone. These PNS glia always trail the leading tips of growth cones, similar to vertebrate glia (Carpenter and Hollyday, 1992; Sepp et al., 2000).

The Origin of Adult Glia Our knowledge about the origin of adult Xy CNS glia is quite fragmentary; most knowledge has been gathered in studies of the visual system by the Steller and Benzer labs. Their studies, which focused on the lamina, the Wrst optic station in the brain where the photoreceptors synapse with their postsynaptic targets, revealed that lamina glia derive from a lineage that is distinct from that of lamina neurons. In addition, the lamina glia are generated independently of the presence of the photoreceptor cells, whereas the lamina neurons require the input of photoreceptors in order to proliferate. However, the terminal diVerentiation of lamina glia depends on an inductive signal from the photoreceptor axons (Winberg et al., 1992). The origin of the lamina glia is not known, but Winberg et al. (1992) identify three periods of cell proliferation and suggest that glial precursors proliferate during these phases. The Wrst phase occurs in late Wrst and second instar larval brain. The second period occurs in the early third instar larva prior to invasion of the retinal axons into the optic lobe. Several lines of evidence also suggest that diVerent classes of lamina glia separate their lineages relatively early in development (Tix et al., 1997). The third period of glial proliferation occurs when the photoreceptors have innervated the lamina. No proliferation is observed after the Wrst day of pupal life, suggesting that most adult glia are born during larval stages. In contrast to adult CNS glia, the lineages of some adult PNS glia have been well deWned. Leg glia are derived partly from cells that migrate from the CNS into the leg imaginal disc and partly from sensory organ precursors (SOPs) that give rise to peripheral nervous system organs. All of these cells express a speciWc glial marker, Repo (Campbell et al., 1994). As shown in Figure 8.8B, the SOP divides to give rise to four cells that form an external sensory organ (neuron, socket, hair shaft, and sheath) and a glial cell (Gho et al., 1999). Both PNS and SOP-derived glial cells migrate along the axons and position themselves at regular intervals. Similarly, retinal basal glia migrate from their origin in the retinal stalk into the eye disc (Choi and Benzer, 1994).

MOLECULAR COMPONENTS REQUIRED FOR GLIAL DETERMINATION

Wing glial cells are derived from the wing disc proper, as removal of the wing from the body wall prior to glial formation does not produce a loss of wing glia. Wing glia diVerentiate from the lineages that produce the mechanoreceptors and gustatory receptors of the wing. These glial cells migrate along the axons that project from the periphery toward the CNS. Some of the glia diVerentiate early during the late third instar stage and some after puparium formation (Giangrande, 1994; Giangrande et al., 1993; Van de Bor et al., 2000).

MOLECULAR COMPONENTS REQUIRED FOR GLIAL DETERMINATION The molecular mechanisms by which glia arise, become speciWed, and diVerentiate have received much attention in the Xy Weld. We Wrst cover the cascades of molecular events determining CNS embryonic glia, particularly the midline glia (MG), the longitudinal glia, and the exit glia. We then discuss the pathway by which adult PNS glia become speciWed. Finally, we draw a parallel between all these pathways.

The Embryonic MG in the CNS The MG are derived from the mesectodermal cells (Fig. 8.6B). The identity of the mesectodermal cells (MECs) is established before gastrulation in ventrolateral cells that form the boundary between the mesoderm and the neurectoderm. The MEC region is composed of the cells that express Twist, a basic Helix-loop-Helix (bHLH) protein, but fail to express Snail, a zinc-Wnger transcription factor (Kosman et al., 1991; Rao et al., 1991). The MECs are further limited to a single row of cells on either side of the midline (Fig. 8.6B), which express the single-minded (Sim) protein (Crews et al., 1988; Nambu et al., 1991; Thomas et al., 1988). Sim is also a bHLH protein, which dimerizes with a ubiquitously expressed bHLH protein named Tango (Sonnenfeld et al., 1997; Wharton et al., 1994). Sim is critical for MEC development, because loss of function of sim causes a loss of the MECs, and ectopic expression of Sim in ectodermal cells produces extra MECs. Sim expression is dependent on Notch and Suppressor of Hairless activation (Morel and Schweisguth, 2000). The cells that express Sim also express Rhomboid, a protein that regulates epidermal growth factor receptor (EGFR) signaling through the processing of the EGFR ligand, Spitz (Tsruya et al., 2002; Urban et al., 2001). During gastrulation, the MECs migrate to the midline; the anterior MECs, which diVerentiate into MG, continue to express Rhomboid and Sim, whereas the posterior MECs diVerentiate into neurons and abort Sim and Rhomboid expression (Bier et al., 1990). Although EGFR signaling plays an important role in survival of MG and in signaling to neighboring neurons (discussed later), this pathway is not required for speciWcation or diVerentiation of MG. Unlike all other glial cells, the MG do not express two prototypical glial markers: Glial Cells Missing and Reverse potential (Repo). The MG, under the control of Sim, initiate the expression of early markers, which include Slit (Klambt et al., 1991; Rothberg et al., 1988), the Netrins, Wrapper (Noordermeer et al., 1998), and Dichaete (Soriano and Russell, 1998). In addition, the MG express tramtrack (ttk), a transcription factor whose expression precludes the MG from taking a neuronal fate (Giesen et al., 1997). Around embryonic stage 12, the cells become dependent on EGFR signaling for their survival. This event coincides with the expression of the transcription factors Drifter (a POU domain transcription factor; Anderson et al., 1995), and Pointed (an ETS factor; Giesen et al., 1997), which is a key target of the EGFR signaling pathway.

Other Embryonic CNS and PNS Glia Most of the embryonic glia, including the lateral glia, the exit glia, as well as cortex glia, are derived from stem cells that are originally speciWed in ventral neurectoderm. Their speciWcation and diVerentiation follows a stereotyped pathway that involves proneural genes,

207

208

8. INVERTEBRATE GLIA

FIGURE 8.7 Notch pathway in speciWcation of sensory organ precursors (SOPs). Early in development, proneural gene expression is maintained in both the presumptive SOP and in ectodermal cells. Later, proneural expression is up-regulated in the SOP, and this imbalance leads to further diVerentiation of these cells. See text for details.

Notch signaling, and expression of Glial Cells Missing followed by glial diVerentiation (Fig. 8.7). Prior to and during gastrulation, clusters of ventral ectodermal cells express proneural genes that encode bHLH proteins such as Scute and Achaete. The selection of a single neuroglioblast or glioblast precursor from this cluster of cells involves a complex interplay between proneural genes and genes of the Notch signaling pathway (Jan and Jan, 1993). Simplistically, the presence of Notch signaling means that changes in cell fate are suppressed. Therefore, suppression of the Notch signal in a speciWc cell or subset of cells allows a fate change to occur in these cells. Hence, to change from an ectodermal cell into a neuroglioblast or glioblast requires the down-regulation of Notch signaling in that cell, whereas Notch signaling is maintained in the other cells of the proneural cluster. If Notch signaling is lost in all cells of the proneural cluster, all cells of the cluster become neuroglioblasts or glioblasts. Notch signaling is required not only for this original decision but also for all later stages of cell fate speciWcation during cell divisions that give rise to diVerentiated neurons and glial cells. However, the Notch pathway is used in the speciWcation of diVerent glia in diVerent contexts; sometimes Notch must be down-regulated in order to specify some glial cells, whereas in few situations Notch is up-regulated (Umesono et al., 2002; Van de Bor et al., 2002). The generic Notch pathway has been extensively reviewed (Artavanis-Tsakonas et al., 1999; Gaiano and Fishell, 2002), and we will therefore only brieXy identify the key players (see Fig. 8.7). Expression of proneural genes within clusters of cells is regulated by a transcriptional cascade, which involves homeobox genes of the Iroquois complex (Cavodeassi et al., 2001). Proneural gene expression in ectodermal cells is seminal for these cells to adopt a neuronal or glial fate. Expression of proneural genes leads to activation of the E(spl) transcription factor and thereby a negative feedback loop onto proneural gene expression (Jennings et al., 1994). Hence, in the absence of input, proneural genes and E(spl) are maintained, expression levels of most of the Notch pathway components in each ectodermal cell are quite similar (Fig. 8.7, top), and cells adopt an ectodermal

MOLECULAR COMPONENTS REQUIRED FOR GLIAL DETERMINATION

fate. If E(spl) expression decreases, cells adopt a neural or glial fate. All cells express delta, the Notch ligand, which upon binding to Notch leads to the activation of a protease (not shown) that promotes cleavage of the Notch intracellular domain. The intracellular domain complexes with the Suppressor of Hairless protein to activate transcription of E(spl) and thereby decrease the level of proneural gene expression in the ectodermal cells (Fig. 8.7, bottom, Schweisguth and Posakony, 1992). This in turn leads to decreased Delta production (Hinz et al., 1994) and hence to decreased Notch signaling in the adjacent ectodermal cell. A second mechanism for generating decreased Notch signaling in an ectodermal cell is by down-regulation of the level of Notch protein in the membrane, a process that seems to be dependent in some cells on the Numb protein (Berdnik et al., 2002; Jafar-Nejad et al., 2002). In both scenarios, the end result is a down-regulation of E(spl) protein levels. In addition to the intercellular communication mediated by Notch signaling, a cell autonomous mechanism ensures that E(spl) activity is antagonized. For example, in the PNS, proneural genes induce the expression of senseless, which antagonizes E(spl) activity and synergizes with the proneural genes to elevate their expression in the SOP (Fig. 8.7, bottom; Nolo et al., 2000). The end result is a loss of expression of proneural proteins in most cells of the proneural cluster and a signiWcant enhancement of proneural protein expression in the SOPs. Similar programs probably exist for neuroblasts, neuroglioblasts, and glioblasts (Jones, 2001). How the neural stem cells generate multiple cell types in a stereotypic fashion is poorly understood. Neuroblasts and neuroglioblasts are known to express a combination of transcription factors (Cui and Doe, 1992; Kambadur et al., 1998). When the ganglion mother cells divide from the neuroblasts or neuroglioblasts, they inherit localized determinants such as Prospero and Numb. Numb antagonizes Notch signaling (Berdnik et al., 2002), whereas Prospero is a homeobox transcription factor (Vaessin et al., 1991) that is asymmetrically localized in the neuroblast. The ganglion mother cell inherits most Prospero, where it enters the nucleus and promotes the proper cell fate (Hirata et al., 1995; Knoblich et al., 1995). A common theme to all glial precursors, except the MG precursors, is expression of Glial Cells Missing (GCM, Hosoya et al., 1995; Jones et al., 1995), also named Glial Cells DeWcient (GLIDE, Vincent et al., 1996). GCM is the major player involved in glial cell determination in Drosophila, and therefore we will expand on its role in glial fate speciWcation. GCM functions as a context-dependent binary genetic switch. When GCM is lost in neural tissue, cells speciWed to form glia become neurons, whereas when GCM is ectopically expressed, presumptive neurons adopt a glial cell fate (Hosoya et al., 1995; Jones et al., 1995). The period during which gcm is expressed is critical for diVerentiation of the precursor into a glia or neuron, and gcm tightly autoregulates its expression in a positive fashion (Miller et al., 1998). As shown in Figure 8.8A, the pattern of gcm expression in embryonic CNS cells ultimately speciWes the fate of three types of precursor cells: glioblasts, neuroglioblasts type I, and neuroglioblasts type II. They are determined and selected from the ventral ectoderm as shown in Figure 8.5. For glioblasts, small groups of three to Wve ectodermal cells transiently express gcm RNA and protein prior to delamination of the single speciWed glioblast, which is the only cell to maintain gcm expression. Restriction of gcm expression to a single cell is mediated by the Notch signaling pathway, which prevents the cell from adopting a neural fate. The glioblasts then divide to produce only glial cells. The type 1 neuroglioblast (NGB, Fig. 8.8A) divides asymmetrically to form a glioblast and a neuroblast. In the NGB, gcm RNA and protein are expressed, but in the daughter cells, gcm expression is restricted to one cell, the presumptive glioblast (GB). The GB divides to generate glia, whereas the neuroblast (NB) produces daughter cells, which ultimately diVerentiate to form neurons. Type 2 neuroglioblasts generate a series of ganglion mother cells (GMC). The GMCs divide to form either two neurons or a neuron and a glia. Hence, GMCs that divide to form both neurons and glia express gcm, whereas the type 2 NGBs and GMCs that produce only neurons do not. Again, gcm expression in the GMC, which divides to form a neuron-glia pair, is controlled by the antagonistic functions of Notch and Numb. In the absence of Notch, these GMCs form two neurons, but in embryos lacking Numb, the GMCs form two glia (Udolph et al., 2001).

209

210

8. INVERTEBRATE GLIA

FIGURE 8.8 Glial precursors of the embryonic CNS and PNS. (A) CNS. Glia are derived from either glioblasts (GB) or neuroglioblasts (NGB). (B) PNS. Sensory organ precursors (SOP) in the periphery follow a pattern of divisions to produce neurons, glia, and other cells to form the sensory organs. In parts A and B, gray cells indicate Glial Cells Missing (GCM) expression, based on RNA patterns. GCM is only transiently expressed early in the diVerentiation of maturing glia.

In the PNS, glia are similarly speciWed by gcm, Notch and numb. In the lateral ectoderm, the sensory organ precursor (SOPI) delaminates and divides to form two daughter cells, SOPIIa and SOPIIb (Fig. 8.8B). The socket and shaft cells of the peripheral sensory organ derive from SOPIIa. SOPIIb divides to generate SOPIIIb, the precursor of the neuron and the sheath cell, and GPI, glial precursor I. gcm RNA and protein are only expressed in GPI and its progeny. With the expression of gcm, the GPI adopts a glial cell fate and undergoes two symmetrical divisions, which ultimately produce glia (Van de Bor et al., 2002). (Note that, as mentioned previously, GCM is also present in the sheath cell, which is why some authors also consider these cells glial cells.) To date, no data have been obtained that implicate the two vertebrate homologs Gcm1 or Gcm2 in glial development (reviewed by Jones, 2001). Gcm1
/
mice die early in development because of a failure of the placenta (Schreiber et al., 2000), and Gcm2
/
mice lack a parathyroid (Gunther et al., 2000). Low levels of Gcm1 have been detected in

DEVELOPMENTAL FUNCTIONS OF GLIA

the CNS (Kim et al., 1998), but since the embryos die early in development, mosaic mice need to be generated to determine if and when Gcm1 is required for glial cell development in mice.

DEVELOPMENTAL FUNCTIONS OF GLIA Nervous System Morphogenesis The glial cells most involved with the development of the nervous system are the midline glia (MG). The early function of the MG cannot be separated from the function of the mesectodermal cells from which they are derived. The two most important functions of the MG and mesectodermal cells are regulation of medial to lateral signaling and axon guidance. We recommend an extensive review written by Jacobs on the role of these cells in neural development (Jacobs, 2000). Loss of the precursor cells that give rise to MG, as observed in single minded (sim) mutant embryos, produces embryos that have a very severe patterning defect that results in fusion of lateral structures of the ectoderm and lateral portions of the nervous system (Crews et al., 1988). The precursors of the MG cells play a key role in EGFR signaling and control the formation and speciWcation of the adjacent neuroblasts along the dorsoventral axis (Menne et al., 1997; Skeath, 1998; Udolph et al., 1998). Hence, neural structures are lost in the absence of EGFR signaling from the midline. This is nicely illustrated with the phenotype associated with mutations in the genes of the spitz group, Wrst identiWed by Mayer and Nusslein-Volhard (1988). Loss of any one of these four genes causes similar phenotypes. They have now been shown to encode Spitz, a ligand for the EGFR; Rhomboid and Star, two proteins that are required for cleavage of Spitz; and Pointed, an ETS transcription factor that is required for transcriptional responses to EGFR signaling. In all cases, when these proteins are absent, more MG enter the apoptotic pathway, leading to a failure to separate and ensheath the commissures and aberrant guidance of midline axons (see Jacobs, 2000, for review). In summary, the MG play an important role in CNS morphogenesis; this role is similar to that observed for the notochord and the ventral Xoor plate in the spinal cord of vertebrates in patterning the CNS. For example, vertebrate homologues of Slit are expressed in the cells of the ventral Xoor plate, and these cells also play a major role in patterning and axon guidance (Jacobs, 2000; Stein and Tessier-Lavigne, 2001).

Axon Guidance Although some glial cells clearly play a role in axon guidance, other glial cells do not participate in this process. We will illustrate this dichotomy by contrasting the role of embryonic MG, which play a major role in axon guidance, with that of the peripheral wing glia, which are not involved in guidance. Although these two types of glia are at two extreme ends of the spectrum, other types of glia, such as exit glia, probably play some role in axon guidance that is not essential to the process. Many axons in the embryonic nervous system cross the midline once. These neurons must be guided to the midline, cross the midline, grow to the contralateral side, and then be instructed not to cross the midline again. Recent work has provided compelling evidence that the Slit protein is produced in the midline glia. The Slit ligand diVuses away from the MG and Wnds its receptor, Roundabout (Robo), which is expressed on the growth cone (Fig. 8.9). Slit functions as a repellent and thereby prevents the growth cones from crossing the midline. However, at speciWc developmental time points, certain neurons express the gene commisureless (comm). The Comm protein down-regulates the Robo protein from the growth cones in a temporally restricted fashion and shunts the Robo receptor to the lysosome for degradation. Hence, during the period in which Robo is down-regulated, the Slit ligand cannot repel the growth cones, and they are able to cross the midline. However, as they traverse the midline, these neurons cease expression of Comm, resulting

211

212

8. INVERTEBRATE GLIA

FIGURE 8.9 Molecular events governing how axons cross the midline in the embryonic CNS. Slit is secreted by the midline glia and serves as a repellent to the navigating axons. The receptor for Slit, Robo, is present in the axonal growth cone. Certain neurons speciWcally express Comm at a very tightly controlled time point. Comm directs Robo to the lysosome for degradation, thereby decreasing the amount of Robo in the growth cone and permitting midline crossing. Adapted from Keleman et al., 2002.

in the reappearance of Robo on the growth cone. Once they cross the midline they are therefore repelled again and will join the contralateral tracts. This simple and elegant model was primarily assembled on the basis of the expression pattern and mutant phenotypes associated with the loss and gain of function of these three proteins. These observations demonstrate the paramount role of the MG in patterning (Georgiou and Tear, 2002; Goodman, 1996; Jacobs, 2000; Keleman et al., 2002). At the other end of the spectrum are the PNS glia from the wing disc. These glial cells are born in the periphery and are derived from SOPs. These cells insulate PNS axons that provide a sensory input from the numerous mechano- and gustatory receptors (Fig. 8.4). The glial cells are born as the growth cones of the adjacent neurons grow along wing veins toward the CNS. Glial migration occurs in a distal to proximal fashion; that is, the glial cells migrate toward the CNS. These glial cells do not guide the neurons, but rather they attach to the developing axons and follow the growth cones. These glial cells seem to be unable to migrate in the absence of axonal tracts, and genetic manipulations that truncate the axons lead to a failure of the glial cells to migrate beyond the truncated end. In summary, in the case of the wing, the glial cells trail the neurons, and the cells that provide guidance information are associated with the wing vein cells, not the glia (Giangrande, 1994, 1995).

DEVELOPMENTAL FUNCTIONS OF GLIA

In most cases, the role of glial cells in axon guidance is more blurry, and glial cells are intertwined with neuronal growth cones in an incestuous relationship. Because numerous cell contacts are established between a single growth cone and other neuron cell bodies, axon tracts, and glial cells in the CNS during development, ablation experiments of speciWc glial cells have often revealed a phenotype that is less spectacular than anticipated. However, in most cases there are phenotypes associated with the loss of glial cells. The simplest explanation seems to be that there is quite a bit of redundancy built into the wiring path and the numerous cell contacts that emanate from growth cones help to reduce the number of mistakes that occur. Hence, in most cases genetic or molecular ablation experiments reveal a reduction in the number of faithful connections as well as a partial defasciculation, phenomena that are rare in wild-type embryos (Hidalgo and Booth, 2000; Schindelholz et al., 2001; Sepp et al., 2000; Sun et al., 2001).

Cell Division and Cell Death Two issues need to be addressed with regard to division and death. First, do glial cells promote/suppress cell division or cell death in neurons by releasing factors? Second, do glial cells depend on trophic factors of neurons or other cells for cell division and survival? The key issue is that there is a balance between the number of neurons and the number of glia. Since this balance can be achieved in numerous ways, the question is, how is it achieved? In vertebrates there is a reciprocal relationship: glial survival depends on trophic interactions from neurons, and neurons require trophic factors from glia. What is known about these interactions in Xies? Until quite recently, the absence of neurotrophins in Drosophila had led to the speculation that these factors might be a prerequisite for complex brain development (Jaaro et al., 2001). However, several publications have provided compelling evidence for the presence of trophic factors and their role in cell survival (reviewed by Hidalgo and Booth, 2000). The main diVerence between Drosophila and vertebrates is that the Drosophila trophic factors are not neurotrophins but rather EGFR ligands. Glia Are Required for Neuronal Survival Glial cells appear to be required for neuronal survival in adult Xies. One of the Wrst mutants illustrating this point was drop-dead. Mutations in drop-dead cause a severe defect in cortical glia, leading to a massive loss of neurons and premature death (Buchanan and Benzer, 1993). Based on their phenotypic analysis the authors favored the hypothesis that the glial cells fail to mature in these mutants, leading to a poor insulation of many neurons, and hence a dysfunctional nervous system. It is therefore likely that glial cells that fail to develop or function properly either directly or indirectly aVect neuronal survival. However, these data provide little insight into the mechanism by which glia support neuronal vitality. The role of glia in neuronal survival was further illustrated by the phenotype of the reverse polarity (repo) mutation. repo mutants were so named due to the reversal of polarity in their electroretinogram, a measurement of the depolarization of the photoreceptors in response to light. The repo gene is expressed in the optic lobe glia, and its loss causes abnormal development of these cells. The consequences are neuronal apoptosis in the optic lobe and in the retina, even though repo is not expressed in the retina. Hence, the death of the retinal neurons is likely due to disrupted neuron-glia interactions in the optic lobe (Xiong and Montell, 1995). Studies of the embryonic CNS by Booth et al. (2000) have provided compelling evidence that neuronal survival depends on the presence of glia. During axon guidance, growth cones contact many glial cells prior to reaching their Wnal targets. The glia function as guide posts and later enwrap the axons. The MG ensheath the commissures and the longitudinal glia envelop the longitudinal portions of the axons. The survival of pioneer neurons, which are the Wrst to establish the main trajectories, does not depend on glia; however, the survival of the remaining neurons is dependent on the presence of longitudinal glia.

213

214

8. INVERTEBRATE GLIA

Glial Survival Depends on Neuronal Interaction A well-described population of CNS glia that is partially lost because of programmed cell death is the MG. During embryogenesis, about 50% of the MG undergo cell death. These cells are expelled from their context and engulfed by circulating hemocytes. Sonnenfeld and Jacobs (1995) concluded that a likely explanation is that MG death may result from a competition for available axons. The key issue for this hypothesis is identiWcation of the trophic factor that is required for survival of the MG. As we indicated earlier, EGFR signaling is of great importance for early MG development. However, in addition to an early role for EGFR signaling, there is also a late role for EGFR signaling. In the absence of proper EGFR signaling, many MG undergo apoptosis (Jacobs, 2000). Spitz, a TGFalpha protein, is the ligand that is secreted by neurons to mediate MG survival. Glial cells compete for limited amounts of secreted Spitz to survive. Binding of Spitz to EGFR leads to the inactivation of Head Involution Defective (Hid), which normally induces apoptosis by blocking a caspase inhibitor (Bergmann et al., 2002). Recently, longitudinal glia have also been shown to depend on trophic factors secreted by neurons in the embryonic CNS. These glial cells are formed in superXuous numbers and their individual survival depends on the proximity of neighboring axons. Similarly to the MG, the longitudinal glial cells express EGFR, but the ligand required for their survival is not Spitz, but rather a neuregulin named Vein (Hidalgo et al., 2001). Vein is produced by pioneer neurons and maintains the survival of longitudinal glia. Hence, there is remarkable similarity with oligodendrocytes that require neuregulin for survival in the vertebrate optic nerve (Fernandez et al., 2000; Hidalgo, 2002). Glial Precursors Control Neuronal Proliferation In 1993, the Zipursky lab (Ebens et al., 1993) reported the cloning and phenotypic analysis of anachronism (ana). Ana encodes a secreted protein that is expressed speciWcally in a subclass of glial cells that seem to control the proliferation of neuroblasts. In ana mutants, quiescent postembryonic central brain and optic lobe neuroblasts enter S phase precociously. This observation suggests that glia play a negative role in regulating proliferation of neuronal precursor cells and thereby control the timing of postembryonic division of neuroblasts (Ebens et al., 1993).

GLIAL CELLS: LATE DIFFERENTIATION AND FUNCTION Compared to early glial development, much less is known in Drosophila about the precise function of the glial cells in adult animals. The main exception might be the presence and formation of an insulatory barrier, which is comparable to the blood brain barrier (BBB). Recently, a few genes have also been implicated in late diVerentiation and function of glial cells, and these will also be reviewed here.

Blood Brain Barrier (BBB) The BBB is an important component in the insulation of insect nervous systems because the hemolymph contains very high levels of potassium (Hoyle, 1952), and neuronal exposure to this concentration of potassium would render neuronal excitability ineVectual. Hence, most, if not all neurons are well insulated from the hemolymph. This insulatory barrier is provided by pleated septate junctions (pSJ). The molecular components of these junctions include Gliotactin and Neurexin IV (Tepass and Hartenstein, 1994). Neurexin IV is a member of the Neurexin superfamily (Baumgartner et al., 1996; Bellen et al., 1998). Gliotactin is most likely part of a protein complex that includes Neurexin IV (Auld et al., 1995). Loss of these proteins causes a failure of neuronal function. Several components of this protein complex are conserved in vertebrates and play a role in the formation of the septate-like junctions at the paranodal areas Xanking the nodes of Ranvier (Bellen et al., 1998; Bhat et al., 2001; Peles and Salzer, 2000). The protein components of septate junctions and septate-like junctions will form the topic of a separate chapter in this

au7

GLIAL CELLS: LATE DIFFERENTIATION AND FUNCTION

book, ‘‘The Neurexin Gene,’’ by Bhat (2003), and hence are not discussed further in this chapter.

Axonal Insulation Although the BBB is an insulatory barrier, it is by no means the only one. As mentioned before, the BBB is formed by perineural glia that do not contact the axons themselves (see Fig. 8.4). The glial cells that wrap around each individual axon also play a role in insulation. In addition, they also control or aVect growth of perineural glia (Yager et al., 2001), axonal conduction (Yuan and Ganetzky, 1999), neurotransmitter uptake (Faeder and Salpeter, 1970; Yager et al., 2001), and provide nutrients for neurons (Carlson and Saint Marie, 1990). Here we will summarize some of the observations that relate to the late diVerentiation and function of glial cells involved in axonal insulation. The repo gene is only expressed in glial cells and encodes a homeodomain protein. Loss of the Repo protein causes a variety of defects because it is required for terminal diVerentiation of glial cells. Null alleles cause embryonic death, a reduction in the number of glial cells, and disrupted fasciculation (Campbell et al., 1994; Halter et al., 1995). Repo is probably under direct transcriptional control of GCM as it has many binding sites for GCM (Akiyama et al., 1996). Repo is probably a late key diVerentiation gene for most glial cells in Drosophila. A gene that was shown to aVect glial ensheathment was the Drosophila FGF receptor, heartless (Shishido et al., 1997). In these mutants, the longitudinal glia do not Xatten and fail to wrap around the longitudinal axons, suggesting that an FGF like ligand is secreted by neurons to promote their ensheathment. A similar phenotype was also observed in locomotion defects (loco) mutants. Loco is also present in longitudinal glia, like Heartless, and its expression is regulated by GCM and Pointed. The loco gene encodes a regulator of G-protein signaling (RGS) protein (Granderath et al., 1999, 2000). Interestingly, loco also aVects the diVerentiation of subperineural glia and probably the BBB. One more player that has been implicated in neuronal ensheathment is the frayed (fray) gene. Fray encodes a serine/threonine kinase homologous to mammalian PASK. Proper function of the protein causes occasional bulges in the peripheral nerves, and the protein is expressed in the peripheral glia. The gene must play a role late in glial development, because in fray mutants, the glial processes elaborate and extend into the nerve and begin to surround the axons but fail to form the complete wraps. Interestingly, the mammalian homolog of Fray (PASK) rescues the Drosophila mutant phenotype, suggesting a conserved signaling pathway (Leiserson et al., 2000). Another molecular pathway that is thought to be conserved evolutionarily is considered to emanate from the peripheral glia to control the growth of perineural glia. This pathway is complex and is thought to involve at least Wve diVerent players. Although the evidence supporting this pathway is quite preliminary, it is displayed in Figure 8.10. The Amnesiac peptide is thought to be secreted by neurons and captured by its receptor in peripheral glia. This signal is proposed to activate a pathway that involves the NeuroWbromatosis 1 gene (NF1) in the peripheral glia. NF1 together with another very large and complex protein (Pushover) is thought to negatively regulate a growth signal secreted from the peripheral glia. This signal or trophic factor is proposed to promote perineural glia proliferation. In addition to this negative regulatory pathway, a positive pathway to promote release of the signal from peripheral glia is thought to be present as well. This signal, as depicted in Figure 8.10, would be under the positive control of neurotransmitters secreted by the neurons, which bind to receptors present in peripheral glia, and under the negative control of a neurotransmitter transporter, Inebriated (Yager et al., 2001). Although this model is highly speculative, it provides an interesting framework to further unravel signaling between neurons and glia.

Neurotransmitter Uptake One of the major functions of insect glia is probably to remove neurotransmitters secreted by neurons. The uptake of neurotransmitters by Drosophila glial cells was Wrst

215

216

8. INVERTEBRATE GLIA

FIGURE 8.10 Molecular pathway by which neurotransmitter release indirectly aVects the growth of perineual glia. A trophic factor secreted by the peripheral glia increases growth of the perineural glia. The release of this protein is modulated by two factors released from the neuron: a neurotransmitter (here Glu) and Amn. Amn acts through Push and NF1 to inhibit release of the trophic factor, whereas the neurotransmitter promotes release of the trophic factor. Ine and Eag diminish the eVects of the neurotransmitter. Abbreviations: Eag, Ether-a-go-go; Glu, glutamate (or other neurotransmitter); Amn, Amnesiac; Ine, Inebriated; GluR, glutamate receptor; AmnR, Amnesiac receptor; Push, Pushover; NF1, NeuroWbromatosis 1. Adapted from Yager et al., 2001.

demonstrated by Campos-Ortega (1974), who found that tritiated gamma-aminobutyric acid (GABA) was quickly taken up by the glia of the lamina in the optic system; this is similar to the absorption of glutamate by glial cells in cockroaches (Faeder and Salpeter, 1970). Surprisingly little more is known about neurotransmitter uptake in fruit Xies. It is assumed that acetylcholine esterase is secreted by glial cells based on the work in other insects (Smith and Treherne, 1965). Most recently, with the sequencing of the Xy genome (Adams et al., 2000), many neurotransmitter transporters have been discovered but their function and localization are unknown. Notable exceptions are the excitatory amino acid transporters (EAAT), EAAT1 and EAAT2, that mediate the high aYnity uptake of L-glutamate and aspartate, respectively (Soustelle et al., 2002). These proteins are expressed in discrete and partially overlapping subsets of glia, but not in neurons, and it is therefore likely that glial cells are involved in mopping up neurotransmitters in Xies. Other transporters have been identiWed in Drosophila, such as the GABA transporters (Neckameyer and Cooper, 1998) and a serotonin transporter (Galli et al., 1997), as well as transporters of unknown neurotransmitters (Inebriated, Huang and Stern, 2002); however, their expression and association with glial biology have not been documented.

Feeding Neurons Since glial cells are known for their ability to protect and insulate neurons, the question arises of how neurons get their energy sources and amino acids. This is especially true in insects as Xuid-borne ions and nutrients must have very limited access to the neurons since

DISCUSSION

the neuropil is not vascularized. It is therefore assumed that the insect glia is very eYcient at transporting ions, amino acids, glucose, and metabolites across these barriers. This issue has not been investigated in fruit Xies to any extent, but Carlson and his colleagues have argued that in the Musca visual system, gap junctions are responsible for this transport. Based on transmission electron microscopy studies they observed interglial reXexive gap junctions and proposed that these junctions played a paramount role in transport of ions, nutrients, and amino acids (Chi and Carlson, 1981; Saint Marie and Carlson, 1983).

Axonal Conductance In vertebrates, a role for glia in axonal conductance is well established (Kaplan et al., 1997; Pfrieger and Barres, 1997). However, defects in axonal conduction have typically been associated with neuronal dysfunction in fruit Xies. For example, kinesin mutants impair action potential propagation, probably because of a decrease in sodium and potassium channel density in axonal membranes (Gho et al., 1992). Similarly, mutations in neuronal sodium channels cause problems in axonal conductance (Loughney et al., 1989). It is assumed that mutations in some of the ‘‘wrapping’’ mutants discussed previously (e.g., loco, fray, heartless) would display defects in axon conductance, but this has yet not been shown, although repo mutants clearly display defects in their electroretinograms, suggesting a problem in axonal conductance or insulation (Xiong and Montell, 1995). Interestingly, Yuan and Gantezky (1999) isolated mutations in a gene, axotactin, that encodes a Neurexin-like protein and mutations in which cause a temperature induced paralysis. They demonstrated that in the absence of this glial speciWc protein, axonal conduction is completely lost at 388C, yet no ultrastructural defects were observed in glial cells. It is therefore likely that Axotactin is part of a glial-neuronal signaling pathway that is important in establishing electrical properties of neuronal membranes.

DISCUSSION The data summarized in this chapter suggest that there are many similarities as well as many diVerences between invertebrate and vertebrate glia. The major diVerences are the lack of myelin and its associated proteins, and the lack of Schwann cells or cells that wrap around axons multiple times (Carlson and Saint Marie, 1990). In addition, the role of GCM, the major determinant of glial identity in Drosophila, has yet to be determined in vertebrate glial cells, and current data suggest a very diVerent role for Gcm1 and 2 in vertebrates than the speciWcation of glial cell identity (Gunther et al., 2000; Kim et al., 1998). Finally, growth factors secreted by glia that are known to maintain neuronal viability appear to have a very diVerent molecular nature in Xies than in vertebrates (Hidalgo, 2002). The major similarities, on the other hand, include the existence of a variety of diVerent cells that have similar functions (Ito et al., 1995), and the presence of numerous molecular pathways that are evolutionarily conserved. An interesting and recurrent observation is that vertebrate homologs of several Drosophila proteins can substitute for the loss of the Drosophila homolog, thereby showing that the vertebrate proteins are at least able to perform the same or similar function as the Drosophila homolog in glial cells or their precursors. We predict that this will be the case for a substantial number of genes and hence will allow investigators to dissect the function of protein domains of vertebrate proteins in Xies and will also permit the manipulation of homologous genetic pathways in fruit Xies using sophisticated genetic screens. Hence, we surmise that the fruit Xy will play an even more prominent role in glial biology in the future than it has now. The other issue that remains to be discussed is the presence of glial cells in a common ancestor of Xies and vertebrates. Flies and human are more than 500 million years apart on the evolutionary scale. This is a huge time interval as it is thought that the common ancestor lived about 650 million years ago (Erwin and Davidson, 2002). Hence, several arguments can be made that glial cells may not have been present in the common ancestor

217

218

8. INVERTEBRATE GLIA

and that they developed independently after the evolutionary tree split, generating species that gave rise to Xies and vertebrates. Indeed, many species that have evolved from the common ancestor do not have glial cells. In addition, numerous proteins are not present in Xies that play a major role in glial biology of vertebrates (PMP22, P0, MBP; Scherer and Arroyo, 2002). Moreover, trophic pathways to maintain glial and neuronal survival are mediated by very diVerent proteins in Xies and vertebrates (Hidalgo et al., 2001). Finally, proteins that play a key role in the development of glia in Xies have not been studied in vertebrates with suYcient thoroughness to assess the extent by which they are similar. In summary, these data would suggest that the common ancestor may not have had glial cells and that much of the similarities that we observe are due to the presence of molecular pathways that existed in the common ancestor. These pathways were partly used and reused when glial cells developed independently in diVerent species. This interpretation would support that the idea that convergent evolution is at the base of the conservation of function of molecular pathways between Xies and vertebrates, a theory that is now gaining ground on the basis of other observations (Erwin and Davidson, 2002).

Acknowledgments We thank Koen Venken, Neal Boerkoel, Patrik Verstreken, and Hamed Jafar-Nejad for discussions. Our research is supported by the Howard Hughes Medical Institute, the National Institutes of Health, and the National Aeronautics and Space Administration.

References Adams, M. D., Celniker, S. E., Holt, R. A., Evans, C. A., Gocayne, J. D., Amanatides, P. G., Scherer, S. E., Li, P. W., Hoskins, R. A., Galle, R. F., et al. (2000). The genome sequence of Drosophila melanogaster. Science 287, 2185–2195. Akiyama, Y., Hosoya, T., Poole, A. M., and Hotta, Y. (1996). The gcm-motif: A novel DNA-binding motif conserved in Drosophila and mammals. Proc Natl Acad Sci USA 93, 14912–14916. Anderson, M. G., Perkins, G. L., Chittick, P., Shrigley, R. J., and Johnson, W. A. (1995). drifter, a Drosophila POU-domain transcription factor, is required for correct diVerentiation and migration of tracheal cells and midline glia. Genes Dev 9, 123–137. Artavanis-Tsakonas, S., Rand, M. D., and Lake, R. J. (1999). Notch signaling: Cell fate control and signal integration in development. Science 284, 770–776. Auld, V. J., Fetter, R. D., Broadie, K., and Goodman, C. S. (1995). Gliotactin, a novel transmembrane protein on peripheral glia, is required to form the blood-nerve barrier in Drosophila. Cell 81, 757–767. Baumgartner, S., Littleton, J. T., Broadie, K., Bhat, M. A., Harbecke, R., Lengyel, J. A., Chiquet-Ehrismann, R., Prokop, A., and Bellen, H. J. (1996). A Drosophila neurexin is required for septate junction and blood-nerve barrier formation and function. Cell 87, 1059–1068. Bellen, H. J., Lu, Y., Beckstead, R., and Bhat, M. A. (1998). Neurexin IV, caspr and paranodin–novel members of the neurexin family: Encounters of axons and glia. Trends Neurosci 21, 444–449. Berdnik, D., Torok, T., Gonzalez-Gaitan, M., and Knoblich, J. A. (2002). The endocytic protein alpha-Adaptin is required for numb-mediated asymmetric cell division in Drosophila. Dev Cell 3, 221–231. Bergmann, A., Tugentman, M., Shilo, B. Z., and Steller, H. (2002). Regulation of cell number by MAPKdependent control of apoptosis: A mechanism for trophic survival signaling. Dev Cell 2, 159–170. Bhat, M. A. (2004). The neurexin gene. In ‘‘Myelin Biology and Disorders’’ (R. Lazzarini, J. GriYn, H. Lassman, K-A. Nave, R. Miller, and B. Trapp. eds.). pp. Elsevier San Diego. Bhat, M. A., Rios, J. C., Lu, Y., Garcia-Fresco, G. P., Ching, W., St Martin, M., Li, J., Einheber, S., Chesler, M., Rosenbluth, J., Salzer, J. L., and Bellen, H. J. (2001). Axon-glia interactions and the domain organization of myelinated axons requires neurexin IV/Caspr/Paranodin. Neuron 30, 369–383. Bier, E., Jan, L. Y., and Jan, Y. N. (1990). Rhomboid, a gene required for dorsoventral axis establishment and peripheral nervous system development in Drosophila melanogaster. Genes Dev 4, 190–203. Booth, G. E., Kinrade, E. F., and Hidalgo, A. (2000). Glia maintain follower neuron survival during Drosophila CNS development. Development 127, 237–244. Bossing, T., and Technau, G. M. (1994). The fate of the CNS midline progenitors in Drosophila as revealed by a new method for single cell labelling. Development 120, 1895–1906. Buchanan, R. L., and Benzer, S. (1993). Defective glia in the Drosophila brain degeneration mutant drop-dead. Neuron 10, 839–850. Campbell, G., Goring, H., Lin, T., Spana, E., Andersson, S., Doe, C. Q., and Tomlinson, A. (1994). RK2, a glialspeciWc homeodomain protein required for embryonic nerve cord condensation and viability in Drosophila. Development 120, 2957–2966.

DISCUSSION

Campos-Ortega, J. A. (1974). Autoradiographic localization of 3H-gamma-aminobutyric acid uptake in the lamina ganglionaris of Musca and Drosophila. Z Zellforsch Mikrosk Anat 147, 415–431. Carlson, S. D., and Saint Marie, R. L. (1990). Structure and function of insect glia. Annu Rev Entomol 35, 597–621. Carpenter, E. M., and Hollyday, M. (1992). The location and distribution of neural crest-derived Schwann cells in developing peripheral nerves in the chick forelimb. Dev Biol 150, 144–159. Cavodeassi, F., Modolell, J., and Gomez-Skarmeta, J. L. (2001). The Iroquois family of genes: From body building to neural patterning. Development 128, 2847–2855. Chanal, P., and Labouesse, M. (1997). A screen for genetic loci required for hypodermal cell and glial-like cell development during Caenorhabditis elegans embryogenesis. Genetics 146, 207–226. Chi, C., and Carlson, S. D. (1981). The perineurium of the adult houseXy: Ultrastructure and permeability to lanthanum. Cell Tissue Res 217, 373–386. Choi, K. W., and Benzer, S. (1994). Migration of glia along photoreceptor axons in the developing Drosophila eye. Neuron 12, 423–431. Crews, S. T., Thomas, J. B., and Goodman, C. S. (1988). The Drosophila single-minded gene encodes a nuclear protein with sequence similarity to the per gene product. Cell 52, 143–151. Cui, X., and Doe, C. Q. (1992). ming is expressed in neuroblast sublineages and regulates gene expression in the Drosophila central nervous system. Development 116, 943–952. Dong, R., and Jacobs, J. R. (1997). Origin and diVerentiation of supernumerary midline glia in Drosophila embryos deWcient for apoptosis. Dev Biol 190, 165–177. Ebens, A. J., Garren, H., Cheyette, B. N., and Zipursky, S. L. (1993). The Drosophila anachronism locus: A glycoprotein secreted by glia inhibits neuroblast proliferation. Cell 74, 15–27. Edwards, J. S., Swales, L. S., and Bate, M. (1993). The diVerentiation between neuroglia and connective tissue sheath in insect ganglia revisited: The neural lamella and perineurial sheath cells are absent in a mesodermless mutant of Drosophila. J Comp Neurol 333, 301–308. Eng, L. F., and DeArmond, S. J. (1982). Immunocytochemical studies of astrocytes in normal development and disease. Adv Cell Neurobiol 3, 145–171. Erwin, D. H., and Davidson, E. H. (2002). The last common bilaterian ancestor. Development 129, 3021–3032. Faeder, I. R., and Salpeter, M. M. (1970). Glutamate uptake by a stimulated insect nerve muscle preparation. J. Cell. Biol. 46, 300–307. Fernandez, P. A., Tang, D. G., Cheng, L., Prochiantz, A., Mudge, A. W., and Raff, M. C. (2000). Evidence that axon-derived neuregulin promotes oligodendrocyte survival in the developing rat optic nerve. Neuron. 28, 81–90. Flower, N. E. (1971). Septate and gap junctions between the epithelial cells of an invertebrate, the mollusc Cominella maculosa. J Ultrastruct Res 37, 259–268. Gaiano, N., and Fishell, G. (2002). The role of Notch in promoting glial and neural stem cell fates. Annu Rev Neurosci 25, 471–490. Galli, A., Petersen, C. I., deBlaquiere, M., Blakely, R. D., and DeFelice, L. J. (1997). Drosophila serotonin transporters have voltage-dependent uptake coupled to a serotonin-gated ion channel. J Neurosci 17, 3401–3411. Georgiou, M., and Tear, G. (2002). Commissureless is required both in commissural neurones and midline cells for axon guidance across the midline. Development 129, 2947–2956. Gho, M., Bellaiche, Y., and Schweisguth, F. (1999). Revisiting the Drosophila microchaete lineage: A novel intrinsically asymmetric cell division generates a glial cell. Development 126, 3573–3584. Gho, M., McDonald, K., Ganetzky, B., and Saxton, W. M. (1992). EVects of kinesin mutations on neuronal functions. Science 258, 313–316. Giangrande, A. (1994). Glia in the Xy wing are clonally related to epithelial cells and use the nerve as a pathway for migration. Development 120, 523–534. Giangrande, A. (1995). Proneural genes inXuence gliogenesis in Drosophila. Development 121, 429–438. Giangrande, A., Murray, M. A., and Palka, J. (1993). Development and organization of glial cells in the peripheral nervous system of Drosophila melanogaster. Development 117, 895–904. Giesen, K., Hummel, T., Stollewerk, A., Harrison, S., Travers, A., and Klambt, C. (1997). Glial development in the Drosophila CNS requires concomitant activation of glial and repression of neuronal diVerentiation genes. Development 124, 2307–2316. Goodman, C. S. (1996). Mechanisms and molecules that control growth cone guidance. Annu Rev Neurosci 19, 341–377. Granderath, S., Bunse, I., and Klambt, C. (2000). gcm and pointed synergistically control glial transcription of the Drosophila gene loco. Mech Dev 91, 197–208. Granderath, S., Stollewerk, A., Greig, S., Goodman, C. S., O’Kane, C. J., and Klambt, C. (1999). loco encodes an RGS protein required for Drosophila glial diVerentiation. Development 126, 1781–1791. Gunther, T., Chen, Z. F., Kim, J., Priemel, M., Rueger, J. M., Amling, M., Moseley, J. M., Martin, T. J., Anderson, D. J., and Karsenty, G. (2000). Genetic ablation of parathyroid glands reveals another source of parathyroid hormone. Nature 406, 199–203. Halter, D. A., Urban, J., Rickert, C., Ner, S. S., Ito, K., Travers, A. A., and Technau, G. M. (1995). The homeobox gene repo is required for the diVerentiation and maintenance of glia function in the embryonic nervous system of Drosophila melanogaster. Development 121, 317–332. Hassan, B. A., and Bellen, H. J. (2000). Doing the MATH: Is the mouse a good model for Xy development? Genes Dev 14, 1852–1865.

219

220

8. INVERTEBRATE GLIA

Hertweck, H. (1931). Anatomie und Variabilitat des Nervensystems und der Sinnesorgane von Drosophila melanogaster (Meigen). Z Wiss Zool 139, 559–663. Hidalgo, A. (2002). Interactive nervous system development: Control of cell survival in Drosophila. Trends Neurosci 25, 365–370. Hidalgo, A., and Booth, G. E. (2000). Glia dictate pioneer axon trajectories in the Drosophila embryonic CNS. Development 127, 393–402. Hidalgo, A., Kinrade, E. F., and Georgiou, M. (2001). The Drosophila neuregulin vein maintains glial survival during axon guidance in the CNS. Dev Cell 1, 679–690. Hinz, U., Giebel, B., and Campos-Ortega, J. A. (1994). The basic-helix-loop-helix domain of Drosophila lethal of scute protein is suYcient for proneural function and activates neurogenic genes. Cell 76, 77–87. Hirata, J., Nakagoshi, H., Nabeshima, Y., and Matsuzaki, F. (1995). Asymmetric segregation of the homeodomain protein Prospero during Drosophila development. Nature 377, 627–630. Hosoya, T., Takizawa, K., Nitta, K., and Hotta, Y. (1995). glial cells missing: A binary switch between neuronal and glial determination in Drosophila. Cell 82, 1025–1036. Hoyle, G. (1952). High blood potassium in insects in relation to nerve conduction. Nature 169, 281–282. Hoyle, G. (1986). Glial cells of an insect ganglion. J Comp Neurol 246, 85–103. Huang, Y., and Stern, M. (2002). In vivo properties of the Drosophila inebriated-encoded neurotransmitter transporter. J Neurosci 22, 1698–1708. Ito, K., Urban, J., and Technau, G. M. (1995). Distribution, classiWcation, and development of Drosophila glial cells in the late embryonic and early larval ventral nerve cord. Roux Arch Dev Biol 204, 284–307. Jaaro, H., Beck, G., Conticello, S. G., and Fainzilber, M. (2001). Evolving better brains: A need for neurotrophins? Trends Neurosci 24, 79–85. Jacobs, J. R. (2000). The midline glia of Drosophila: A molecular genetic model for the developmental functions of glia. Prog Neurobiol 62, 475–508. Jafar-Nejad, H., Norga, K., and Bellen, H. (2002). Numb: ‘‘Adapting’’ notch for endocytosis. Dev Cell 3, 155–156. Jan, Y. N., and Jan, L. Y. (1993). The peripheral nervous system. In ‘‘The Development of Drosophila melanogaster’’ (M. Bate, and A. Martinez Arias, eds.), pp. 1207–1244. Plainview, Cold Spring Harbor Laboratory Press, New York. Jennings, B., Preiss, A., Delidakis, C., and Bray, S. (1994). The Notch signalling pathway is required for Enhancer of split bHLH protein expression during neurogenesis in the Drosophila embryo. Development 120, 3537–3548. Jones, B. W. (2001). Glial cell development in the Drosophila embryo. Bioessays 23, 877–887. Jones, B. W., Fetter, R. D., Tear, G., and Goodman, C. S. (1995). glial cells missing: A genetic switch that controls glial versus neuronal fate. Cell 82, 1013–1023. Kambadur, R., Koizumi, K., Stivers, C., Nagle, J., Poole, S. J., and Odenwald, W. F. (1998). Regulation of POU genes by castor and hunchback establishes layered compartments in the Drosophila CNS. Genes Dev 12, 246–260. Kania, A., Salzberg, A., Bhat, M., D’Evelyn, D., He, Y., Kiss, I., and Bellen, H. J. (1995). P-element mutations aVecting embryonic peripheral nervous system development in Drosophila melanogaster. Genetics 139, 1663–1678. Kaplan, M. R., Meyer-Franke, A., Lambert, S., Bennett, V., Duncan, I. D., Levinson, S. R., and Barres, B. A. (1997). Induction of sodium channel clustering by oligodendrocytes. Nature 386, 724–728. Keleman, K., Rajagopalan, S., Cleppien, D., Teis, D., Paiha, K., Huber, L. A., Technau, G. M., and Dickson, B. J. (2002). Comm sorts robo to control axon guidance at the Drosophila midline. Cell 110, 415–427. Kim, J., Jones, B. W., Zock, C., Chen, Z., Wang, H., Goodman, C. S., and Anderson, D. J. (1998). Isolation and characterization of mammalian homologs of the Drosophila gene glial cells missing. Proc Natl Acad Sci USA 95, 12364–12369. Klambt, C., Jacobs, J. R., and Goodman, C. S. (1991). The midline of the Drosophila central nervous system: A model for the genetic analysis of cell fate, cell migration, and growth cone guidance. Cell 64, 801–815. Knoblich, J. A., Jan, L. Y., and Jan, Y. N. (1995). Asymmetric segregation of Numb and Prospero during cell division. Nature 377, 624–627. Kosman, D., Ip, Y. T., Levine, M., and Arora, K. (1991). Establishment of the mesoderm-neuroectoderm boundary in the Drosophila embryo. Science 254, 118–122. Lane, N. J. (1981). Invertebrate neuroglia: Junctional structure and development. J Exp Biol 95, 7–33. Leiserson, W. M., Harkins, E. W., and Keshishian, H. (2000). Fray, a Drosophila serine/threonine kinase homologous to mammalian PASK, is required for axonal ensheathment. Neuron 28, 793–806. Loughney, K., Kreber, R., and Ganetzky, B. (1989). Molecular analysis of the para locus, a sodium channel gene in Drosophila. Cell 58, 1143–1154. Mayer, U., and Nusslein-Volhard, C. (1988). A group of genes required for pattern formation in the ventral ectoderm of the Drosophila embryo. Genes Dev 2, 1496–1511. Menne, T. V., Luer, K., Technau, G. M., and Klambt, C. (1997). CNS midline cells in Drosophila induce the diVerentiation of lateral neural cells. Development 124, 4949–4958. Miller, A. A., Bernardoni, R., and Giangrande, A. (1998). Positive autoregulation of the glial promoting factor glide/gcm. Embo J 17, 6316–6326. Morel, V., and Schweisguth, F. (2000). Repression by suppressor of hairless and activation by Notch are required to deWne a single row of single-minded expressing cells in the Drosophila embryo. Genes Dev 14, 377–388. Nambu, J. R., Lewis, J. O., Wharton, K. A., Jr., and Crews, S. T. (1991). The Drosophila single-minded gene encodes a helix-loop-helix protein that acts as a master regulator of CNS midline development. Cell 67, 1157–1167. Neckameyer, W. S., and Cooper, R. L. (1998). GABA transporters in Drosophila melanogaster: Molecular cloning, behavior, and physiology. Invert Neurosci 3, 279–294.

DISCUSSION

Nolo, R., Abbott, L. A., and Bellen, H. J. (2000). Senseless, a Zn Wnger transcription factor, is necessary and suYcient for sensory organ development in Drosophila. Cell 102, 349–362. Noordermeer, J. N., Kopczynski, C. C., Fetter, R. D., Bland, K. S., Chen, W. Y., and Goodman, C. S. (1998). Wrapper, a novel member of the Ig superfamily, is expressed by midline glia and is required for them to ensheath commissural axons in Drosophila. Neuron 21, 991–1001. Peles, E., and Salzer, J. L. (2000). Molecular domains of myelinated axons. Curr Opin Neurobiol 10, 558–565. Pentreath, V. (1989). Invertebrate glial cells. Comp Biochem Physiol C 93A, 77–83. Pfrieger, F. W., and Barres, B. A. (1997). Synaptic eYcacy enhanced by glial cells in vitro. Science 277, 1684–1687. Ramo´n y Cajal, S., and Sanchez, D. (1915). Contribucio´n al conocimiento de los centros nerviosos de los insectos. Trab Lab Invertebr Biol Univ Madrid 13, 1–168. Rao, Y., Vaessin, H., Jan, L. Y., and Jan, Y. N. (1991). Neuroectoderm in Drosophila embryos is dependent on the mesoderm for positioning but not for formation. Genes Dev 5, 1577–1588. Rothberg, J. M., Hartley, D. A., Walther, Z., and Artavanis-Tsakonas, S. (1988). slit: An EGF-homologous locus of D. melanogaster involved in the development of the embryonic central nervous system. Cell 55, 1047–1059. Saint Marie, R. L., and Carlson, S. D. (1983). Glial membrane specializations and the compartmentalization of the lamina ganglionaris of the houseXy compound eye. J Neurocytol 12, 243–275. Saint Marie, R. L., and Carlson, S. D. (1985). Interneuronal and glial-neuronal gap junctions in the lamina ganglionaris of the compound eye of the houseXy, Musca domestica. Cell Tissue Res 241, 43–52. Scharrer, B. C. J. (1939). The diVerentiation between neuroglia and connective tissue sheath in the cockroach (Periplaneta americana). J Comp Neurol 70, 77–78. Scherer, S. S., and Arroyo, E. J. (2002). Recent progress on the molecular organization of myelinated axons. J Peripher Nerv Syst 7, 1–12. Schindelholz, B., Knirr, M., Warrior, R., and Zinn, K. (2001). Regulation of CNS and motor axon guidance in Drosophila by the receptor tyrosine phosphatase DPTP52F. Development 128, 4371–4382. Schmidt, H., Rickert, C., Bossing, T., Vef, O., Urban, J., and Technau, G. M. (1997). The embryonic central nervous system lineages of Drosophila melanogaster. II. Neuroblast lineages derived from the dorsal part of the neuroectoderm. Dev Biol 189, 186–204. Schreiber, J., Riethmacher-Sonnenberg, E., Riethmacher, D., Tuerk, E. E., Enderich, J., Bosl, M. R., and Wegner, M. (2000). Placental failure in mice lacking the mammalian homolog of glial cells missing, GCMa. Mol Cell Biol 20, 2466–2474. Schweisguth, F., and Posakony, J. W. (1992). Suppressor of Hairless, the Drosophila homolog of the mouse recombination signal-binding protein gene, controls sensory organ cell fates. Cell 69, 1199–1212. Sepp, K. J., Schulte, J., and Auld, V. J. (2000). Developmental dynamics of peripheral glia in Drosophila melanogaster. Glia 30, 122–133. Shishido, E., Ono, N., Kojima, T., and Saigo, K. (1997). Requirements of DFR1/Heartless, a mesoderm-speciWc Drosophila FGF-receptor, for the formation of heart, visceral and somatic muscles, and ensheathing of longitudinal axon tracts in CNS. Development 124, 2119–2128. Skeath, J. B. (1998). The Drosophila EGF receptor controls the formation and speciWcation of neuroblasts along the dorsal-ventral axis of the Drosophila embryo. Development 125, 3301–3312. Skeath, J. B. (1999). At the nexus between pattern formation and cell-type speciWcation: The generation of individual neuroblast fates in the Drosophila embryonic central nervous system. Bioessays 21, 922–931. Smith, D. S., and Treherne, J. E. (1965). The electron microscopic localization of cholinesterase activity in the central nervous system of an insect, Periplaneta americana 1. J Cell Biol 26, 445–465. Sonnenfeld, M., Ward, M., Nystrom, G., Mosher, J., Stahl, S., and Crews, S. (1997). The Drosophila tango gene encodes a bHLH-PAS protein that is orthologous to mammalian Arnt and controls CNS midline and tracheal development. Development 124, 4571–4582. Sonnenfeld, M. J., and Jacobs, J. R. (1995). Apoptosis of the midline glia during Drosophila embryogenesis: A correlation with axon contact. Development 121, 569–578. Soriano, N. S., and Russell, S. (1998). The Drosophila SOX-domain protein Dichaete is required for the development of the central nervous system midline. Development 125, 3989–3996. Soustelle, L., Besson, M. T., Rival, T., and Birman, S. (2002). Terminal glial diVerentiation involves regulated expression of the excitatory amino acid transporters in the Drosophila embryonic CNS. Dev Biol 248, 294–306. Stein, E., and Tessier-Lavigne, M. (2001). Hierarchical organization of guidance receptors: Silencing of netrin attraction by slit through a Robo/DCC receptor complex. Science 291, 1928–1938. Strausfeld, N. J. (1976). ‘‘Atlas of an Insect Brain.’’ Springer-Verlag, Berlin; New York. Sun, Q., Schindelholz, B., Knirr, M., Schmid, A., and Zinn, K. (2001). Complex genetic interactions among four receptor tyrosine phosphatases regulate axon guidance in Drosophila. Mol Cell Neurosci 17, 274–291. Tepass, U., and Hartenstein, V. (1994). The development of cellular junctions in the Drosophila embryo. Dev Biol 161, 563–596. Thomas, J. B., Crews, S. T., and Goodman, C. S. (1988). Molecular genetics of the single-minded locus: A gene involved in the development of the Drosophila nervous system. Cell 52, 133–141. Tix, S., Eule, E., Fischbach, K. F., and Benzer, S. (1997). Glia in the chiasms and medulla of the Drosophila melanogaster optic lobes. Cell Tissue Res 289, 397–409. Tsruya, R., Schlesinger, A., Reich, A., Gabay, L., Sapir, A., and Shilo, B. Z. (2002). Intracellular traYcking by Star regulates cleavage of the Drosophila EGF receptor ligand Spitz. Genes Dev 16, 222–234.

221

222

8. INVERTEBRATE GLIA

Udolph, G., Rath, P., and Chia, W. (2001). A requirement for Notch in the genesis of a subset of glial cells in the Drosophila embryonic central nervous system which arise through asymmetric divisions. Development 128, 1457–1466. Udolph, G., Urban, J., Rusing, G., Luer, K., and Technau, G. M. (1998). DiVerential eVects of EGF receptor signalling on neuroblast lineages along the dorsoventral axis of the Drosophila CNS. Development 125, 3291– 3299. Umesono, Y., Hiromi, Y., and Hotta, Y. (2002). Context-dependent utilization of Notch activity in Drosophila glial determination. Development 129, 2391–2399. Urban, S., Lee, J. R., and Freeman, M. (2001). Drosophila rhomboid-1 deWnes a family of putative intramembrane serine proteases. Cell 107, 173–182. Vaessin, H., Grell, E., WolV, E., Bier, E., Jan, L. Y., and Jan, Y. N. (1991). prospero is expressed in neuronal precursors and encodes a nuclear protein that is involved in the control of axonal outgrowth in Drosophila. Cell 67, 941–953. Van de Bor, V., Heitzler, P., Leger, S., Plessy, C., and Giangrande, A. (2002). Precocious expression of the Glide/ Gcm glial-promoting factor in Drosophila induces neurogenesis. Genetics 160, 1095–1106. Van de Bor, V., Walther, R., and Giangrande, A. (2000). Some Xy sensory organs are gliogenic and require glide/ gcm in a precursor that divides symmetrically and produces glial cells. Development 127, 3735–3743. Vincent, S., Vonesch, J. L., and Giangrande, A. (1996). Glide directs glial fate commitment and cell fate switch between neurones and glia. Development 122, 131–139. Wharton, K. A., Jr., Franks, R. G., Kasai, Y., and Crews, S. T. (1994). Control of CNS midline transcription by asymmetric E-box-like elements: Similarity to xenobiotic responsive regulation. Development 120, 3563–3569. Winberg, M. L., Perez, S. E., and Steller, H. (1992). Generation and early diVerentiation of glial cells in the Wrst optic ganglion of Drosophila melanogaster. Development 115, 903–911. Xiong, W. C., and Montell, C. (1995). Defective glia induce neuronal apoptosis in the repo visual system of Drosophila. Neuron 14, 581–590. Yager, J., Richards, S., Hekmat-Scafe, D. S., Hurd, D. D., Sundaresan, V., Caprette, D. R., Saxton, W. M., Carlson, J. R., and Stern, M. (2001). Control of Drosophila perineurial glial growth by interacting neurotransmitter-mediated signaling pathways. Proc Natl Acad Sci USA 98, 10445–10450. Yuan, L. L., and Ganetzky, B. (1999). A glial-neuronal signaling pathway revealed by mutations in a neurexinrelated protein. Science 283, 1343–1345.

C H A P T E R

9 Neural Cell Specification during Development Mahendra Rao

NEURAL DIFFERENTIATION The nervous system is one of the earliest organ systems that diVerentiates from the blastula-stage embryo. The primitive neural tube forms by approximately the fourth week of gestation, and neurogenesis has commenced by the Wfth week. Development is quite prolonged and proceeds throughout embryogenesis, while myelination is not completed till late post-natal stages. Eight major stages in the development of the nervous system can be recognized (Cowan, 1979; Fig. 9.1). These are (1) induction of the neural plate, (2) localized/regional proliferation of undiVerentiated cells in diVerent brain regions, (3) migration of cells from the region in which they are generated to places they will Wnally reside, (4) aggregation of cells to form identiWable parts of the brain, (5) diVerentiation of neurons, (6) formation of connections with other neurons, (7) selective death of certain cells, (8) and elimination of connections. Glial diVerentiation parallels and overlaps neuronal diVerentiation and can be likewise examined in separable stages (Fig. 9.2). Like neuronal cells, glial cells are formed from the induced neuroectoderm. Initially formed glial precursors generated in the ventricular or subventricular zone migrate to their Wnal destinations using cues similar to those used by neurons. Precursor cells undergo maturation in situ, aggregate with newly formed neurons to form speciWc brain regions, and undergo selective apoptosis such that an appropriate neuron to glial ratio is maintained. Unlike neurons, however, glial processes are rarely long, the variety of glial cells that can be functionally distinguished are far fewer, and the process of glial communication is generally considered far simpler than the complex synapse that mediates most neuronal signaling. Glial progenitors are present throughout the parenchyma of the brain and likely divide at a slow rate in situ to provide cellular replacement of glial populations throughout life. Overall glial number is tightly regulated, and a constant proportion of neurons to support elements is maintained. Regulatory processes include regulation of proliferation and apoptosis (Fig. 9.1). Two major classes of glial cells oligodendrocytes and astrocytes that subserve distinct functions have been identiWed. Oligodendrocytes form the myelin sheaths that insulate axons and enable saltatory conduction allowing myelinated nerve Wbers to transmit impulses up to 10 times faster than nonmyelinated Wbers of the same diameter (Ritchie, J. M., 1984). Oligodendrocytes can be distinguished from astrocytes by morphology, antigenic expression, and response to growth factors and cytokines. Their development and maturation is described in detail in Chapter 11. Astrocytes make up 20 to 50% of the volume of most brain areas and appear to be a more heterogeneous class of cells that likely have many diVerent roles. The degree of

Myelin Biology and Disorders, Volume 1

223

Copyright 2004, Elsevier (USA). All rights reserved.

224

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

Neurons

Glia

Induction of neural Plate

Induction of neural Plate

Localized/regional proliferation

Localized/regional proliferation

Migration

Migration

Aggregation

Aggregation

Differentiation and maturation

Differentiation and maturation

Formation of connections

Matching cell number/myelination

Selective cell death

Ongoing replacement

Elimination of connections

Persistence of precursors

FIGURE 9.1 Stages of glial diVerentiation. Eight stages of neuronal development have been described. Glial development parallels neuronal development to a large extent, albeit with some diVerences. These are summarized.

astrocyte diversity in the CNS is unclear, though at least two kinds of astrocytes can be distinguished based on morphology (Wbrous and protoplasmic). A characteristic shared by all classes of astrocytes is their expression of glial Wbrillary acidic protein (GFAP) (Bignami, 1972) and the expression by a signiWcant proportion of S-100ß. Astrocyte development is described in Chapter 9. In addition to astrocytes and oligodendrocytes, several specialized glial cells have been identiWed. These include radial glia, Bergman glia, Muller glia, pituitary glia, olfactory ensheathing cells, and so on. These cells originate from diVerent regions of the brain and can be distinguished from the relatively more abundant astrocytes and oligodendrocytes. These cells together have been termed aldynoglia on the basis of their overall similarity and specialized functions (Gudino-Cabrera et al., 1999). Aldynoglia share characteristics with both astrocytes and oligodendrocytes and overall resemble Schwann cells (Tab. 9.1). Radial glia cells a prototypic aldynoglial cells provide a scaVold for guiding neurons to their ultimate destinations and are the Wrst glial population that can be distinguished from the proliferating neuroepithelium. The process of glial diVerentiation from the proliferating neuroepithelial cells in the ventricular zone overlaps and extends beyond neuronal diVerentiation and likely continues throughout life. Neuronal generation is essentially complete by birth with the exception of the cerebellum, which develops post-natally, and a limited amount of persistent neurogenesis in the olfactory bulb and the hippocampus. In contrast, while glial precursors cells have segregated from the neuroepitheTABLE 9.1

Properties of Specialized Glia Specialized glia

Estrogen receptor alpha immunoreactive Low aYnity NGFR (p75 ) immunoreactive) Low GFAP expression in culture and co-expression of vimentin High levels of PSANCAM Capable of reentry into the cell cycle Found in regions where spontaneous axonal regeneration occurs Nieto-Sampedro and colleagues have suggested that specialized glial share several properties though their lineage is distinct. These glial cells can be readily distinguished form astrocytes or oligodendrocytes and overall appear similar to Schwann cells (Gudino-Cabrera et al., 1999)

II. GLIAL CELL DEVELOPMENT

NEURAL INDUCTION AND THE CNS/PNS SEGREGATION

lium early in development, maturation and diVerentiation occur much later (see Chapter 12). Radial glial cells and oligodendrocyte-astrocyte precursor cells are among the Wrst glial cells that can be distinguished from the proliferating neuroepithelium, while astrocytes as identiWed by GFAP immunoreactivity develop later, and most astrocyte and oligodendrocyte development is predominantly post-natal with myelination of long tracts not being completed till age 12 in humans. A small turnover of astrocytes and oligodendrocytes occurs throughout life in all regions of the brain in both gray and white matter. In subsequent sections, the major stages of neural development will be brieXy reviewed with an emphasis on glial diVerentiation and the segregation of the glial lineage from the neuronal lineage.

NEURAL INDUCTION AND THE CNS/PNS SEGREGATION The earliest stage of segregation of glial cell development occurs at the time of neural induction. The neuroectoderm segregates from the ectoderm by a process of neural induction (Fig. 9.2). The initially formed neural plate undergoes a stereotypic set of morphogenetic movements to form a hollow tube by a process termed primary neurulation. Primary neurulation includes a process of diVerential proliferation, infolding, and fusion of the lateral margins. The caudalmost portion of the neural tube (somite 35 onwardNievelstein et al., 1993; Schoenwolf, 1984) is formed by a process of secondary neurulation wherein a solid cord of cells undergoes cavitation to form a hollow central tube that fuses with the neural tube formed by primary neurulation. In some species (amphibians), the entire tube is formed by secondary neurulation (Colas and Schoenwolf, 2001). At the time of infolding and neural tube formation, some neuroectodermal cells present at the dorsal margins are excluded from the neural tube. These cells derived from junctional ectoderm-neurectoderm form neural crest, and in cephalic regions laterally placed ectodermal cells form placodes. As development proceeds, the neural tube is completely

FIGURE 9.2 Neural induction. The peripheral nervous system is derived from neural crest cells, placodal precursors, late emigrating crest, and possibly VENT cells. The major pathways of migration and placodes in the cranial region are shown.

II. GLIAL CELL DEVELOPMENT

225

226

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

separated from the surface ectoderm while placodal cells and neural crest undergo an epithelial to mesenchymal transition to invade the developing mesoderm and localize to multiple regions to generate the peripheral nervous system and the craniofacial mesenchyme (Fig. 9.3). Peripheral glial cells including Schwann cells, enteric glia, and nonmyelinating PNS glia are derived from the neural crest and placodal precursor cells while oligodendrocytes, astrocytes, and most aldynoglia are derived from the neural tube. The segregation of the neural crest from the neural tube is essentially irreversible and this has been convincingly demonstrated in chick heterochronous transplant experiments. Neural crest cells when transplanted into the CNS fail to generate CNS derivatives in any abundance (see RuYns et al., 1998), and transplantation of Schwann cells into the CNS does not transdiVerentiate Schwann cells into oligodendrocytes or astrocytes. The neuroepithelium that will generate the CNS retains its ability to generate PNS cells at least for the Wrst few days of development and may retain its ability in the adult. In chick, the time period is approximately 24 hours as demonstrated by neural tube rotation experiments and dorsal tube ablation experiments and transplant experiments (Korade and Frank, 1996; Scherson et al., 1993). It is likely that this is true in rodents as well based on cultures of neuroepithelial cells, which can be shown to generate neural crest cells and PNS derivates either spontaneously in culture or after exposure to BMP. The role of wnt/ß-catenin signaling in the segregation of neural tube and neural crest lineages has been shown by a variety of investigators. Wnt receptors and wnts are expressed in the appropriate region, wnt antisense oligonucleotides down-regulate crest diVerentiation; overexpression enhances neural crest development (reviewed in Broner-Fraser, 2002). Thus, CNS and PNS glial cells have clearly distinct lineages, and the segregation of the PNS from the CNS has occurred early in development. PNS glia are derived from the neural crest stem cell and in the craniofacial region from placodal precursors. CNS precursors likely retain their ability to generate PNS cells for at least a short time period in vivo and after prolonged periods in culture (Mujtaba et al., 1999). Two late emigrating

FIGURE 9.3 Neural crest development. The peripheral nervous system is derived from neural crest cells, placodal precursors, late emigrating crest, and possibly VENT cells. The major pathways of migration and placodes in the cranial region are shown.

II. GLIAL CELL DEVELOPMENT

REGIONALIZATION

populations of cells that are derived from the neural tube and populate the PNS have been described. Late-emigrating crest and VENT (ventrally emigrating neural tube) cells have also been described that may contribute to PNS development (Sharma et al., 1995, Sohal et al., 1998). Some recent experiments have provided direct evidence that dividing cells harvested from the adult rodent brain may retain the ability to make PNS derivatives including Schwann cells. Mckay and colleagues (Tsai and McKay, 2000), for example, have shown that adult CNS-derived neural stem cells can generate smooth muscle and possible PNS neurons, while Blakemore and colleagues (Keirstead et al., 1999) have described the generation of myelinating Schwann cells from a CNS-derived NCAMþ glial progenitor cell. Overall, however, it is clear that in normal development the neural tube generates CNS glial cells (with the exception of the olfactory ensheathing glia) and does not contribute in any signiWcant fashion to PNS glial development.

REGIONALIZATION As neural development proceeds, the initially formed tube undergoes diVerential expansion and regionalization to form identiWable rostrocaudal regions that will generate the future subdivisions of the brain (Fig. 9.4). The anterior neural tube undergoes a dramatic expansion and can be delineated into three primary vesicles: the forebrain (prosencephalon), the midbrain (mesencephalon), and the hindbrain (rhombencephalon). DiVerential growth and further segregation lead to additional delineation of the prosencephalon into the telencephalon and diencephalon, and the rhombencephalon into the metencephalon and myelencephalon. The caudal neural tube does not undergo a similar expansion but does increase in size to parallel the growth of the embryo and undergoes further diVerentiation to form the spinal cord. Major derivatives of each secondary vesicle are listed (Fig. 9.5). A further subdivision of the forebrain (prosencephalon) into Wve prosomeres has been described (Rubenstein, 1998), as have further subdivisions of the hindbrain into rhombomeres (Glover, 2001). These regions can be distinguished from each other by homeobox gene expression and the future fates of that particular rhombomere or prosomeric region. It is important to emphasize that the overall pattern of the CNS is laid down early, and the process of regionalization has occurred at the stage of neuroepithelial proliferation before the onset of signiWcant neurogenesis and well before the onset of gliogenesis.

FIGURE 9.4 Subdivisions of the brain. The early neural tube is rapidly patterned into discrete domains. The caudal portions expand to form three primary vesicles, which undergo further subdivisons to form secondary vesicles, which subsequently develop into distinct regions of the brain.

II. GLIAL CELL DEVELOPMENT

227

228

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

FIGURE 9.5 Derivatives of the brain. The main subdivisions and the regions that develop are shown. The cerebellum develops as an extension of the rhombencephalon.

Each region of the brain has physical boundaries that restrict cellular movement and thus each region develops as relatively independent units (Fig. 9.5). The available information so far suggests that cellular diVerentiation in each region of the brain is similar, and identical strategies are used to regulate cell survival proliferation and diVerentiation even though development may be temporally segregated, and the relative ratio and phenotypes of diVerentiated cells may be distinct. We will use examples from the spinal cord or the cerebral cortex as these are the best available studies but suggest that development in other regions is similar as well. As the rostrocaudal pattern is developing, dorso-ventral segregation is also underway and is inXuenced by extrinsic agents The folding of the neural tube occurs at speciWc hinge points that separate the alar from the basal plate. Most sensory structures are derived from the alar plate, while most motor structures are derived from the basal plate. Ventralizing signals emanating from the Xoor plate in caudal regions and a Xoor plate equivalent in more ventral regions confer ventral identity to the developing neural tube (Fig. 9.6). One ventralizing candidate is shh hedgehog. Dorsal identity is maintained by antagonizing signals from the overlying ectoderm and potential dorsalizing signals include BMP, RA, and wnt. It is important to note that in general oligodendrocytes appear to arise predominantly from ventral regions, while astrocytes arise from more dorsal region (Lee et al., 2000; Schmechel et al., 1979; Voigt et al., 1989), suggesting that rostrocaudal identity may inXuence/bias cell fate. Several studies, though not all direct, have suggested that dorsoventral and rostrocaudal patterning occurs early and this pattern inXuences the fate of proliferating neuroepithelial cells. For example, Hox-1 misexpression in chicken embryos is suYcient to alter rhombomere3 identity (Bell et al., 1999). Likewise, misexpression of Pax-6, Eyeless, and other eyespeciWc genes is suYcient to induce ectopic eye formation (reviewed in Gehring, 1996). We have found that misexpression of fz-3 in Xenopus is suYcient to induce ectopic eyes at the midbrain/hindbrain junction (Rasmussen et al., 2000). Neural crest and placodal cells, which generate the PNS, are likewise prepatterned early (reviewed in Hunt et al., 1999). Transplanting presumptive placodes from one region to another will alter their identity only early in development (Baker et al., 1999, 2002). Similarly, not all crest cells can contribute to nodose ganglion or to cardiac crest even when transplanted back to the appropriate migratory pathway (see, for example, Harrison et al., 1995). Transplant experiments likewise have suggested that regional speciWcation occurs at the stem cell stage and that stem cells isolated from diVerent regions consequently diVer in their

II. GLIAL CELL DEVELOPMENT

THE EARLY NEUROEPITHELIAL TISSUE CONSISTS OF MULTIPOTENT STEM CELLS THAT ARE NOT RADIAL GLIA

229

FIGURE 9.6 Dorsoventral patterning. As regionalization proceeds patterning is also evident in the dorso-ventral axis. The folding of the neural tube occurs at speciWc hinge points that separate the alar from the basal plate. Most sensory structures are derived from the alar plate, while most motor structures are derived from the basal plate. Further segregation of the dorsal and ventral regions occurs, and an example of regionally expressed molecules is shown. As discussed in the text, while dorsoventral identity appears critical to the phenotype of neurons generated, it is not as clear in the case of glial cells.

diVerentiation bias. For example, progenitors from the hippocampus, but not from the cerebellum or midbrain, produce hippocampal pyramidal neurons (Shetty and Turner, 1998). Mesencephalic neural precursors (Ling, 1998) diVerentiate into predominantly dopaminergic neurons, while other precursor cell populations do not (reviewed in Svendsen and Rosser, 1995). MGE precursors diVer from LGE precursors in both their migration and diVerentiation ability (Wichterle et al., 1999). Thus, as embryonic development proceeds, stem cells themselves undergo regionalization to acquire distinct properties while maintaining their proliferative potential and ability to diVerentiate into multiple phenotypes. The data, while not conclusive, also suggest that regional identity is maintained through multiple passages in culture (Hitoshi et al., 2002; Nakagawa, 1996), suggesting that dediVerentiation and acquisition of an appropriate identity may not happen routinely or reliably if cells are transplanted into ectopic regions. It is clear that regional identity is established early in development, and it can aVect the fate choices that stem cells make and deWne the functional identity of neurons that are generated. It is not as clear, however, if regional identity is critical for astrocyte and oligodendrocyte function. Oligodendrocytes or late oligodendrocyte precursors from cranial or caudal brain regions appear similar in their ability to myelinate and survive after transplantation and astrocytes appear morphologically indistinguishable as well. Some recent data suggesting that astrocytes may be regionally distinct were provided by Dr. Gage and colleagues (Song et al., 2002). They showed that astrocytes from cranial but not cortical regions were capable of supporting neurogenesis from stem cells and that midbrain astrocytes appeared to bias cells toward a dopaminergic phenotype, suggesting that regional identity mat be important for glial cells as well. Overall the available evidence indicates that regionalization occurs early and patterns stem cells at the proliferating neuropeithelial cell stage. This patterning is relatively stable even in culture and biases the proliferation and diVerentiation of the stem cells. More diVerentiated cells retain their regional identity, and this regional identity is likely to be important for appropriate functions of astrocytes and oligodendrocytes as well.

THE EARLY NEUROEPITHELIAL TISSUE CONSISTS OF MULTIPOTENT STEM CELLS THAT ARE NOT RADIAL GLIA The initially formed neural tube consists of a single layer of cells that extend from the central canal to the external limiting membrane (Sauer, 1935). While a single layer of cells,

II. GLIAL CELL DEVELOPMENT

230

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

it appears pseudo-stratiWed with cell nuclei seen at diVerent levels. This pseudo-stratiWed epithelium is termed the neuropeithelium and the cells neuroepithelial cells. Neuropeithelial cells that are initially homogenous undergo diVerentiation to ultimately generate all the neurons and glial cells that comprise the adult brain (Fig. 9.7). The properties of the neuroepithelium have been characterized (Cai et al., 2002). Neuropeithelial cells appear homogenous despite the acquisition of rostrocaudal and dorsoventral identity (discussed earlier). Neuropeithelial cells have certain characteristic properties (Table 9.2). In particular, neuropeithelial cells lack the expression of radial glial markers such as RC1, RC2, and vimentin (Ying et al., 2002; Table 9.2). However, like later appearing radial glial cells, neuroepithelial cells span the entire width of the neural tube. Cells appear morphological indistinguishable from other progenitor cells, possess tight and gap junctions, and the pial margin appears to contain asymmetrically localized molecules including (Cai et al., 2002; Chen and McConnell, 1995). Cells at this stage have decondensed chromatin and are dividing rapidly and estimates range from 6 to 8 hours (reviewed in Sommer and Rao, 2000). At this stage, with the exception of the roof plate and the Xoor plate, cells appear as a homogeneous, rapidly proliferating population of cells that communicate with each other by gap and tight junctions. These cells have been termed neuroepithelial cells and can be distinguished from later appearing cells by a variety of criteria. A variety of evidence suggests that the neuroepithelial cells of the developing neural tube are stem cells. Stem cells have been deWned as cells that demonstrate self-renewal and retain the capacity to diVerentiate into one or many kinds of diVerentiated cells through a characteristic process of sequential diVerentiation that is dependent on the type of stem

TABLE 9.2 Properties of NEP Cells Ventricular zone derived neuropeithelial stem cells Hoechst and Rhodhamine levels

High (but BCRP1 expressed)

Telomerase activity and TERT expression

High

Nestin, Sox-2, Fz-9, FGFR4

Present

Absence of lineage speciWc markers

GFAP, CD44, S-100ß, Map-2, ß-111 tubulin , PSANCAM, RC1, RC2, A2B5, CNP, MBP are absent

Absent neurotransmitter response

Absence of glutamate, GABA, PACAP, VIP, responses.

Growth factor dependence

Respond to FGF but not to EGF or PDGF

Acutely isolated and cultured stem cells isolated from E10.5 neural tubes can be distinguished from other cells by the presence and absence of markers. Note that while telomerase expression is high other cells also express high telomerase activity.

FIGURE 9.7 The ventricular zone. Image of a semithin toludine blue-labeled section from the developing rodent neural tube at E10.5 and a high power TEM image are shown. Cells appear homogenous, extend from the central canal to the outer limiting membrane, and appear attached to each other by tight junctions and gap junctions.

II. GLIAL CELL DEVELOPMENT

THE EARLY NEUROEPITHELIAL TISSUE CONSISTS OF MULTIPOTENT STEM CELLS THAT ARE NOT RADIAL GLIA

cells and stage of development. Cells that fulWll the criteria of stem cells have been isolated from nearly all major tissues or organs and their ability to diVerentiate has been demonstrated in vitro and in vivo (reviewed in Cai and Rao, 2002). The ability to both self-renew and generate diVerentiated cells is likely due to asymmetric divisions at the single cell level or coordinated symmetric divisions that serve to maintain the population but not individual stem cells (discussed later). Stem cells are classiWed on the basis of their diVerentiation capacity, region of isolation, and degree of self-renewal. Experiments analyzing the self-renewal and diVerentiation characteristics of neuroepithelial cells have unequivocally shown that neuropepithelial cells are stem cells (Davis and Temple, 1994; Kalyani et al., 1997). Kalyani et al. (1997), for example, have shown that the majority of the neuroepithelium at embryonic day 10.5 (E10.5) in rat is comprised of multipotent stem cells. The authors found that no unipotent cells exist at this stage, though bipotential or tripotential cells are present (Fig. 9.8). Retroviral lineage tracing has suggested that at the early developmental stages multipotent stem cells are present, while at later stages colonies are phenotypically more restricted (see, for example, Levison and Goldman, 1997). Using cultures of acutely dissociated cells from diVerent embryonic ages, investigators have shown that most of the rat neuroepithelium at E10.5 is composed of multipotent stem cells (Cai et al., 2002; Kalyani et al., 1997) The earliest neural tube derived NSCs have been termed NEPs (neuropeithelial cells) and at E10.5 (in rats, E8.5 in mice) are the only stem cell population present. NEPs are FGF dependent, do not express detectable levels of EGFR, and are present within the ventricular zone along the entire rsotrocaudal axis. NEP cells are capable of self-renewal and generate neurons, astrocytes, and oligodendrocytes in vitro and in vivo (Cai et al., 2002; Kalyani et al., 1997; 1999; Mayer-Proschel et al., 2002; Mujtaba et al., 1999). NEP cells can be distinguished from other dividing cells in the nervous system by their high telomerase activity, expression of BCRP-1, nestin, Sox-2 and Fz-9, and the absence of lineage speciWc markers such as of E-NCAM, A2B5, GFAP, and other markers characteristic of neurons and glia (Lin-, Cai et al., 2002; Yongquan et al., 2002). Based on marker expression and clonal analysis, the percentage of NEPs rapidly declines and at E14.5 and comprises a fraction less than 10% of the neural tube population. The predominant populations of cells at this stage are NRPs (neuronal restricted precursors) and GRPs (glial restricted precursors), which constitute about 80% of the cells (Kalyani et al., 1998; Mayer-Proschel et al., 1998). In about an additional day of development, neuronal and radial diVerentiation can be observed. Neuroepithelial cells appear conWned to the ventricular zone and can be identiWed by the persistence of some of markers present at E 10.5 in rats (Fig. 9.9). The early born migrating neuronal progenitors and maturing neurons begin to express E-NCAM (Fig. 9.9) and are present in the overlying mantle (Fig. 9.9). The neurons likely migrate along a radial glial scaVold, which begins to appear at around the same time (Fig. 9.10) or utilize other cues to follow tangential paths to their ultimate location. Radial glial cells span the entire width of the neural tube and can be distinguished form other cells by the expression of radial glial markers. Neuroepithelial cells, while conWned to the ventricular zone, may undergo lateral migration and thus their progeny may be widely dispersed. Neuroepithelial cells may undergo both symmetrical and asymmetrical divisions. It is likely that at an early stage, symmetric divisions that amplify the number of stem cells present and enlarge the neural tube are the primary mode of cell division. During further development it is likely that asymmetric cell divisions where only one daughter cell is a selfrenewing stem cell become more predominant (Martens et al., 2000) and Wnally when the ventricular zone is diminishing in size perhaps symmetric divisions that generate two diVerentiated daughter cells are predominant (Fig. 9.10) The persistent neuropeithelial cells can be distinguished from the diVerentiating cells by the absence of markers characteristic of neurons or radial glial and their localization to the ventricular zone. The ventricular zone undergoes a further restriction in size till by birth it comprises of a single layer of cells termed the ependymal layer. Cells within the ependymal layer have characteristic features including cilia and may co-express both neuronal and glial markers. Cells in the ependymal layer divide slowly if at all and at least at later stages of development do not contribute to the ongoing neurogenesis and gliogenesis (Chiasson et al., 1999).

II. GLIAL CELL DEVELOPMENT

231

232

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

FIGURE 9.8 Clonal analysis. E14.5 dissociated neural tube cells were sorted by negative selection (absence of E-NCAM and A2B5) and grown at clonal density (50 to 100 cells per dish) or high density (10,000 cells/dish) in NEP-basal medium with 10% CEE. As a control, NCAMþ/A2B5þ cells were plated at clonal density as well. After 8 days in culture, CEE was withdrawn from the medium and cells were grown further for 5 days before being triple labeled for A2B5, bIII tubulin, and GFAP expression. Control cells, which were incubated only with FITC conjugated secondary antibody, determined the background Xuorescence (A). M1 was set according to the control to isolate the bottom 10% of the E-NCAM-/A2B5-population (B). The purity of the collected negative population was checked and was approximately 97% (C). One day after FACS, E-NCAM-/A2B5-cells (Lin-) were labeled with E-NCAM (red in D), A2B5 (red in E), and Brdu (green in F). Dapi counterstaining was used to analyze all cells (D and F). E-NCAM-/A2B5-cells plated at high density were able to diVerentiate into GalC, GFAP, and bIII tubulin positive cells when they were replated on PLL/LAM coated dishes for 5 days (G–I). The morphology of E-NCAM-/A2B5-clones (Lin-, J) is uniform and quite diVerent from that of E-NCAMþ/A2B5þ cells (Linþ, K–M). Phase pictures were taken 8 days after FACS. Two examples of clones of single E-NCAM-/A2B5-cells that generated GFAP (blue), GalC (green), and bIII tubulin (red) immunoreactive cells were shown (N–O). The scale bar for D–F and N–O is 50 mm, and the scale bar for G–I and J–M is 100 mm. Reproduced from Cai et al. (2002).

II. GLIAL CELL DEVELOPMENT

THE EARLY NEUROEPITHELIAL TISSUE CONSISTS OF MULTIPOTENT STEM CELLS THAT ARE NOT RADIAL GLIA

FIGURE 9.9 BRDU/Sox-2/NCAM. Ventricular zone–derived stem cells have characteristic properties. A recently described marker is Sox-2 (red). Double labeling with an antibody to PSANCAM (green) and Sox-2 shows that that most cells in the vz express Sox-2 but not diVerentiation markers such as PSANCAM. Photograph courtesy Dr. J. Cai, National Institute of Aging.

The earliest diVerentiated glial cell to be identiWed is the radial glial cell. Radial glial cells have a characteristic elongated bipolar morphology with a proximal process extending to the lumen and a distal process extending to the external limiting membrane (Fig. 9.11). Radial glial cells can be distinguished from the proliferating neuroepithelial cells by the expression of RC2. Neuroanatomical studies including three-dimensional reconstructions of electron microscopic sections have shown that the early born neurons use the radial glial scaVold to migrate to appropriate locations. This radial glial directed migration is a characteristic feature of the developing cortex and is likely not as important in more caudal brain regions where cortical lamination is not prominent (Rakic, 1971, 1972; Rakic et al., 1974). As migration of newly generated neurons is completed, radial glial cells can no longer be detected and their Wnal fate is open to debate. They may mature into astrocytes and as has been more recently suggested, radial glial cells maybe at least bipotential and capable of generating both neurons and astrocytes. Alternatively, radial glial cells may persist as a quiescent stem cell that is reactivated after injury. It is clear,

II. GLIAL CELL DEVELOPMENT

233

234

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

FIGURE 9.10 Stem cells in the ventricular zone. Neuroepithelial cells present in the ventricular zone (VZ) divide rapidly. Cells may undergo symmetric divisions (SD) to increase the pool of stem cells or asymmetric divisions AsD) to selfrenew and generate a diVerentiated progeny. Lateral migration (LM) of stem cells within the ventricular zone has been described. DiVerentiate cells migrate along a radial glia (RG) scaVold or tangentially to their Wnal domains. The total number of stem cells present depends on the balance between self-renewal, proliferation rates, cell death, and diVerentiation.

FIGURE 9.11 Radial glial cells identiWed by RC1 immunoreactivity develop around the time of neurogenesis and span the entire width of the neural tube. Individual cells can be studied in vitro and in vivo, and their transition from radial glial cells to astrocytes has been well documented. More recent evidence suggests a transition to a neuronal fate as well. No data on oligodendrocyte diVerentiation are available. Photograph courtesy Y. Luo, National Institute on Aging.

however, that neuroepithelial stem cells and radial glial cells are not synonymous. NEP cells likely generate radial glia and can be distinguished from them by the absence of radial glial markers and the presences of additional markers.

II. GLIAL CELL DEVELOPMENT

ADDITIONAL ZONES OF NEUROGLIOGENESIS

Thus, by midgestation dynamic changes have sculpted the initial uniform neural tube into speciWc regions and within each region the proliferating neuroepithelial stem cells have begun to generate diVerentiated progeny, which have begun to migrate away from the zone of proliferation. The ventricular zone is beginning to constitute an ever-smaller portion of the growing neural tube, and the cell cycle length of the cells with the ventricular zone is increasing. At least two diVerentiated populations of cells can be readily distinguished from the neuropithelial cells, and these include radial glial cells and neurons. Some early glial speciWc markers are beginning to be expressed in speciWc ventral domains (Fig. 9.12), though no markers of astrocytes or mature oligodendrocytes can be detected at this stage (Ying et al., 2002).

ADDITIONAL ZONES OF NEUROGLIOGENESIS AND AN ADDITIONAL STEM CELL POPULATION CAN BE DISTINGUISHED AT MID EMBRYONIC DEVELOPMENT As development progresses, the rapid proliferation of the neuropeithelium and the early diVerentiation of neurons and radial glial cells leads to an alteration of the simple neuroepithelial structure in the cortex. The earliest step involves the creation of a marginal zone where early born neurons exit the ventricular zone to form a preplate. The axons projecting from preplate neurons separate the ventricular zone from the preplate, and this region is termed the intermediate zone. As neurogenesis continues, the preplate is subdivided into the subplate and the marginal zone that will mature to form the appropriate cortical layers. The subplate neurons are a transient population that undergoes apoptosis. By mid-embryogenesis, the ventricular zone is much reduced in size and additional zones of mitotically active precursors can be identiWed. Mitotically active cells that accumulate adjacent to the ventricular zone have been termed the subventricular zone cells. This SVZ is later called the subependymal zone as the ventricular zone is reduced to a single layer of ependymal cells. The SVZ is prominent in the forebrain and can be identiWed as far back as the fourth ventricle. No SVZ can be detectable in more caudal regions of the brain, and if it exists it is likely a very small population of cells. An additional germinal matrix derived from the rhombic lip of the fourth ventricle called the external granule layer generates the granule cells of the cerebellum (Table 9.3). The SVZ cells are morphologically distinct from the ventricular zone stem cells (Table 9.4). Cells are small and compact and have little cytoplasm and few organelles. Cells do not display interkinetic movements characteristic of VZ cells and unlike VZ cells possess a single elongated cilium that extends to the ventricular lumen. In addition, SVZ cells express separate homeodomain markers and likely acquire regional cues that separate the forebrain SVZ into distinct domains that generate phenotypically distinct progeny. The SVZ in

FIGURE 9.12 Layering of the cortex. The transition from a single layer of neuropeithelium to a well-deWned multilayered cortex is shown. The ventricular zone is becoming reduced, an additional zone of proliferation, the subventricular zone, is becoming apparent in speciWc regions of the brain, and two classes of diVerentiated cells radial glia and neurons are clearly identiWable.

II. GLIAL CELL DEVELOPMENT

235

236

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

TABLE 9.3 Additional Zones of Neurogenesis. Zones of proliferation

Gliogenesis

SVZ

SVZ of the cortical areas is a mosaic of cells including stem cells and progenitors. The relative proportion varies in diVerent regions

MGE

The MGE appears early and it diVers in morphology from the SVZ and generates both astrocytes and oligodendrocytes

LGE

The LGE likely contributes to cortical astrocyte and oligodendrocyte generation

Svz of midbrain

It is likely that a SVZ derived from the VZ exists

EGL

The external granule layer generates the granule cells of the cerebellum. Possibly radial glia, and likely astrocytes and oligodendrocytes

Svz of caudal neural tube

No clear-cut SVZ is evident in the caudal neural tube

Several subdivisions of the subventriuclar zone can be identiWed. Cells in each of these regions are a heterogenous population of cells that include multipotent self renewing populations as well as more restricted progenitor cells. The SVZ’s generate most diVerentiated cells including astrocytes and oligodendrocytes from midgestation onwards

TABLE 9.4 Differences between VZ and SVZ Ventricular zone cells

Subventricular zone cells

PseudostratiWed neuroepithelium

Aggregation of proliferating cells

Homogenous population of cells

Heterogeneous population of cells

Interkinetic movement is characteristic

No interkinetic movement observed

Cell cycle is short

Longer and increases throughout development

FGF receptors expressed while EGF receptor not expressed till later

EGF and FGFR expressed

Ventricular zone disappears by birth and is replaced by a single layer of ependymal cells

Persists into adulthood in deWned regions

Absence of GFAP and absence of cilia

Stem cells in the svz likely express GFAP and possess a single cilium that extends towards the ventricle

Pyramidal projection neurons and other early born excitatory neuron populations

Primarily interneurons and other inhibitoy neuron populations

Radial glia arise from VZ cells

?? radial glia generated

Neural crest diVerentiates from vz cells

?? neural crest generation

Positional markers diVer from the overlying svz

Positional markers diVer

Table summarizes some diVerences between the ventricular zone and subventricular zone and may predict the properties of the stem cells derived from these regions.

mid-embryogenesis can be divided into distinct regions that include the cortical SVZ, the medial ganglion eminence and the lateral ganglion eminence (Table 9.3). Retroviral labeling of the SVZ cells has suggested regional heterogeneity and demonstrated that the SVZ consists of a mixture of stem and progenitor cells (Levison and Goldman, 1993; Luskin et al., 1993, 1998; Price and Thurlow, 1988; Williams and Price, 1995). Tritiated thymidine injections to kill actively dividing cells at later stages of subventricular zone development has suggested that approximately 1% of the cells are slowly dividing stem cells (Moorsehead et al., 1994). The ratio at earlier stages remains undetermined, though it is reasonable to assume that the proportion of stem cells is higher. Thus, during the second half of embryogenesis the predominant proliferating populations that are present are localized to the subventricular zone. Cells in the SVZ are regionally distinct, can be distinguished from VZ cells, and a subset of the cells are multipotent stem cells that are capable of generating neurons astrocytes and oligodendrocytes. The proportion of SVZ stem cells declines with development and in the adult multipotent stem cells are likely present only in regions of ongoing neurogenesis (for example, anterior SVZ and

II. GLIAL CELL DEVELOPMENT

ADDITIONAL ZONES OF NEUROGLIOGENESIS

TABLE 9.5 FGF and EGF Dependent Stem Cells have Distinct Properties EGF dependent neurospheres

FGF dependent NEP cells

Grow in suspension cultures.

Grow as adherent cells or in suspension cultures.

Present later (E14.5 onwards in mice) in embryonic development.

Present as early as E10.5 in rats and E8.5 in mice.

Absolute requirement of EGF for isolation.

Absolute requirement of FGF for isolation.

Express EGF-R.

Do not express EGF-R.

No EGF dependent neurospheres present in the spinal cord.

FGF dependent stem cell present in the spinal cord throughout development.

No evidence that neurospheres can generate radial glia, neural crest or PNS derivatives.

Can generate PNS derivatives and NCSC’s.

Low frequency of neuronal diVerentiation.

High frequency of neuronal diVerentiation.

The table summarizes the differences between EGF dependent neurospheres and FGF dependent NEP stem cells. Both cells are multipotent self-renewing cells that can differentiate into neurons astrocytes and oligodendrocytes. Nevertheless the two cell types differ in cytokine response, receptor expression, spatial and temporal distribution and in the cell types that they can generate. Note that a lineage relationship between the two stem cell types has been postulated. NCSC ¼ neural crest stem cell, EGF-R ¼ Epidermal growth factor receptor, EGF ¼ Epidermal growth factor, FGF ¼ Fibroblast growth factor.

the SVZ underlying the hippocampus). Additional stem cells may be present in other brain regions, and this is discussed in Chapter 10. Soon after the diVerentiation of early born neurons and radial glial cells and prior to the generation of NG2þ oligodendrocyte precursors, and GFAPþ astrocytes and around the time the SVZ can be clearly demarcated and an additional stem cell population can be isolated and propagated in culture. This stem cell population was Wrst described by Weiss and colleagues (Reynolds et al., 1992, Reynolds and Weiss, 1996) and well over a thousand papers have been published describing/using this cell. This second stem cell population has been termed the EGF-dependent stem cell. EGF-dependent, neurosphere-forming cells constitute a small population (1 to 3%) of all cells present in vivo (Fig. 9.13). EGF-dependent stem cells that grow in suspension culture constitute a fraction of the cells present in any sphere, and this fraction of cells has been shown to undergo self-renewal and diVerentiate into neurons, astrocytes, and oligodendrocytes in vivo and after transplantation. EGF-dependent stem cells can be isolated from the entire rostrocaudal axis from E14 onward, and EGF-dependent stem cells have been isolated from the adult and cadaveric tissue as well (Laywell et al., 1999; Reynolds and Weiss, 1996). Several lines of evidence suggest that in the developing embryo, the neurosphere-forming stem cell population likely resides in the subventricular zone and are distinct from the ventricular zone derived neural stem cell populations. Retroviral labeling experiments indicate that a subpopulation of cells within the SVZ are multipotent (noted earlier). High EGFR and EGF expression are seen in the SVZ but not in the early VZ (Burrows et al., 2000). Loss of EGFR expression does not alter the size of the VZ but aVects later neuronal and glial survival (Kornblum et al., 1998, Threadgill et al., 1995). It has been diYcult to isolate neurospheres with EGF alone from regions of the brain where an SVZ cannot be morphologically identiWed. EGF-dependent neurosphere-forming stem cells cannot be isolated at stages prior to the formation of the SVZ and microdissection experiments have shown that the region that contains the largest neurosphere-forming ability includes the SVZ and its immediate environs (reviewed in Moorsehead and Van der Kooy, 2001). However, while it is likely that the neurosphere-forming stem cell population arises from SVZ cells, the exact identity of the cell remains controversial. The SVZ is a heterogeneous population (Levison, 2002, review), and it is not clear which cell in the heterogeneous population equates to the neurosphere-forming stem cell population. DiVerent groups have suggested diVerent locations and diVerent properties. In the adult, neuro-

II. GLIAL CELL DEVELOPMENT

237

238

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

FIGURE 9.13 Neurosphere. Neural cells placed in deWned medium in the presence of high concentrations of EGF will form aggregates of dividing cells termed neurospheres. Such aggregates can be isolated from E 14.5 onward, and each sphere of cells contains a fraction of stem cells as assessed by self-renewal and the ability of individual cells to diVerentiate into neurons, astrocytes, and oligodendrocytes. Photograph courtesy J. Cai, National Institute on Aging.

sphere-forming stem cells may be localized to the ependyma, to type B cells in the SVZ, or type C cells in the SVZ (reviewed in Garcia-Verdugo et al., 1998; Alvarez-Byulla et al., 2000). Since the VZ is present earlier than the SVZ and the SVZ is known to arise from the VZ, it is reasonable to assume that neuropeithelial stem cells likely generate neurosphereforming neural stem cells. Such a lineage relationship has been elegantly demonstrated in a variety of ways by several investigators (reviewed in Rao, 1999; Fig. 9.14). FGF dependent stem cells can be isolated prior to isolation of EGF dependent stem cells. FGF and EGF in combination generate more neurospheres than either factor alone (Ciccolinin F., 2001; Repressa et al., 2001; Tropepe et al., 1999). EGF knockouts do not alter the size or prominence of the ventricular zone that contains FGF-dependent stem cells. In a technically diYcult set of experiments, Tropepe and colleagues constructed

FIGURE 9.14 VZ to SVZ transition. A schematic model of the transition from one class of neural stem cells, the VZ derived FGF dependent neuropeithelial stem cell, to the EGF dependent neurosphere-forming SVZ derived neural stem cell. Note that VZ cells can generate radial glia and neural crest and do so in normal development. Data on whether neurosphere-forming stem cells can generate these two populations remain elusive. Both stem cell populations persist in the adult and both populations can generate astrocytes and oligodendrocytes and likely generate diVerent classes of neurons.

II. GLIAL CELL DEVELOPMENT

MODELS OF STEM CELL DIFFERENTIATION

chimeras from Wbroblast growth factor receptor 1-null (FGFR1
) and normal animals and showed that EGF-dependent stem cells must go through a FGF dependent stage. These data argue that FGF-dependent cells are the precursors of the EGF-dependent neurosphere stem cells. These results are compelling but do not address at what stage the two lineages diverge and whether the transition from one stem cell to another is complete. Indeed, several groups have shown that both FGF- and EGF-dependent stem cells coexist in similar brain regions at later stages of development. Using population and statistical analyses, these investigators argued that these represent two distinct populations of cells (Table 9.5). The coexistence of these two populations suggests that not all FGF-dependent stem cells necessarily transform into an EGF-dependent stem cells; rather, both populations coexist. Thus, it appears that a subset of the FGF-dependent vz-derived stem cells mature to form SVZ derived EGF dependent neuropshere-forming stem cells, but both populations coexist at through later stages of development. It is also unclear how many stem cell markers and functional regulators are shared between EGF and FGF dependent stem cells and markers identiWed as being speciWc from FGF dependent stem cells have not been rigorously tested in the adult. DiVerences between FGF dependent ventricular zone–derived stem cell populations and EGF-dependent SVZderived neuropshere-forming stem cell populations have been described (reviewed in Rao, 1999), though it must be emphasized that the properties of EGF-dependent stem cells have been quite diYcult to assess as it is diYcult to isolate a relatively pure population of cells in a neurosphere type culture where true stem cells generally constitute a minority population. More recently, Dr. Bartlett and colleagues (Reitze et al., 2001) have suggested two potential markers that can be used to enrich for neural stem cells. That authors showed that low PNA (peanut agglutinin) and HSA (heat stable antigen) staining can be used to select for stem cell populations from neurosphere cultures. Weissman and colleagues have suggested that AC133 may be an additional neural stem cell marker (Uchida et al., 2000). Whether these markers label ventricular zone derived stem cells remains to be determined, though our results suggest that these markers are not speciWc at early developmental stages (Mayer-Proschel et al., 2002). Overall, the data indicate that at least two populations of stem cells are present at overlapping but distinct stages of neural development. These cells fulWll the criteria of stem cells but diVer from each other in the method and timing of isolation, the localization in vivo, and their cytokine response and their intrinsic biases in diVerentiation. Both populations can contribute to glial development and the relative importance of two potential lineages of glial cells remains obscure.

MODELS OF STEM CELL DIFFERENTIATION Overall, studies suggest that glial cells must be derived from the developing neuroepithelium or from the SVZ, and it is useful to consider possible models of segregation process. The process of stem cell diVerentiation into neurons, astrocytes, and oligodendrocytes could occur via a variety mechanisms and a variety of developmental models have been proposed (see Fig. 9.15). At one extreme, the neuroepithelium, while an apparently homogenous population of nestin immunoreactive dividing cells, may in fact consist of a mosaic of precommited stem cells. Each stem cell may give rise to multiple progeny of only one type (part A in Fig. 9.15). In this model, the multiplicity of cell types seen reXects the initial heterogeneity of the developing neuropeithelial population. However, a variety of experimental data would argue against this model of diVerentiation. Retroviral labeling studies in developing chick embryos have unequivocally shown that individual labeled cells can generate multiple phenotypes. No apparent restriction to either dorsal or ventral fates, or to a particular lineage, is evident based on these studies (Leber et al., 1990; Sanes, 1989). Even if cells were labeled just prior to their terminal divisions, clones containing two diVerent phenotypes were observed, clearly demonstrating the existence of a multipotent population. Essentially identical results were obtained with dye injection studies. Single cell

II. GLIAL CELL DEVELOPMENT

239

240

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

FIGURE 9.15 Models of segregation. Several schematic models of stem cell diVerentiation are shown. It is likely that no single model precisely describes normal stem cell diVerentiation. As discussed in the text, Model C, which incorporates intermediate or more restricted precursors, appears to best Wt the observations.

injection experiments in chick spinal cords generated labeled progeny that included neurons and glia, and also included PNS derivatives (Artinger et al., 1995; Bronner Fraser and Fraser, 1988, 1989). Retroviral labeling in rodent tissue has shown the existence of multipotent stem cells as well (see, for example, Leber and Sanes, 1990). Clonal analysis in culture and transplants of clonal cell population in vivo have shown that single cells can generate all major phenotypes in the brain. Thus, there is clear-cut evidence that multipotent stem cells exist and that individual cells can diVerentiate into neurons astrocytes and oligodendrocytes. Models B and C suggest that multipotent stem cells exist, but diVer from each other in the prediction of intermediate precursors. Model B suggests that terminal diVerentiation proceeds through an intermediate stage of proliferating precursors, which have more restricted developmental fates. These intermediate precursors might correspond to the restricted precursors identiWed in the lineage tracing experiments discussed earlier. Model A assumes no such restricted precursors exist and that a multipotential precursor may give rise to any two phenotypes when it undergoes terminal diVerentiation. Such a model has been invoked to explain retinal diVerentiation (Livesey and Cepko, 2001). We tend to favor model C as more representative of CNS development for several reasons. Perhaps the most important is the identiWcation of restricted precursors at later stages of development. Evidence for the presence of neural crest stem cells (see, for example, Stemple and Anderson, 1992), glial (oligodendrocyte-astrocyte) progenitor cells (see, for example, Pringle and Richardson, 1993; Warf et al., 1991), oligodendroblasts (reviewed in Lee et al., 2000), neuroblasts that can give rise to motoneurons as well as to other neurons (see, for example, Mayer-Proschel et al., 1997; Ray and Gage, 1994), and committed astrocyte precursors (Miller and Szigeti, 1991 and references therein) suggests the presence of multiple committed precursors in the developing brain. Model D diVers from models B and C in that it does not presuppose the existence of asymmetric cell divisions. Two kinds of symmetric divisions are proposed. A symmetric selfrenewing division where two daughter stem cells are generated and a symmetric diVerentiation division where two diVerentiated cells are generated. This model suggests that the population itself is self-renewing and mutlipotent but that any individual cell may not be. The overall balance is maintained by global regulation of the two kinds of symmetric

II. GLIAL CELL DEVELOPMENT

MULTIPOTENT NEURAL STEM CELLS GENERATE OLIGODENDROCYTE AND ASTROCYTES

divisions. This model does not presuppose the existence of an identiWable stage of intermediate progenitors or blast cells nor does it require them as in essence stem cells, which undergo a symmetric diVerentiated cell division behave like a blast cell/intermediate precursor population. This model could also explain how CNS stem cells diVerentiate into neurons, astrocytes, and oligodendrocytes. Model D does diVer from other models in the absence of any asymmetrical divisions. Such asymmetrical divisions where one daughter cell is retained in the ventricular zone while a diVerentiated cell migrates away from the ventricular zone have been observed in vivo and in culture. Dr. McConnell and colleagues (O’Rourke et al., 1995) have shown that both symmetric and asymmetric divisions occur during development and that asymmetrical divisions can be predicted based on the axis of cleavage suggesting an asymmetric distribution of molecules underlies the diVerent fates of daughter cells. Multipotent cells in the ventricular zone undergo both symmetrical and asymmetrical divisions. Cai et al. (1997b) used retroviral labeling to show that approximately 48% of labeled cells formed clusters located entirely within the ventricular zone suggesting self-renewal via symmetrical divisions. Approximately 20% of cells, however, appeared to generate cells in both the ventricular zone as well as in the mantle suggesting at least some asymmetrical divisions. Thus, model C Wts CNS stem cell proliferation and diVerentiation best. It is important, however, to emphasize that it is likely that stem cell diVerentiation can be modulated such that it changes with time. Symmetric divisions may be present early in development with asymmetric divisions at an intermediate stage and symmetric diVerentiated cell divisions at a later stage (noted earlier). An additional observation worth emphasizing is that stem cells isolated from multiple regions (both dorsal and ventral) of the brain can generate precursors that in turn generate oligodendrocytes and astrocytes, even though in normal development oligodendrocytes arise ventrally and astrocytes dorsally. In an elegant set of experiments, Chandross et al. (1999) isolated dorsal and ventral portions of the neural tube and showed that both tissues were capable of generating oligodendrocytes. These results are in contrast with observations in vivo and in explant cultures where oligodendrocytes develop only in ventral regions (see, for example, Miller, 1996; Richardson et al., 1997). Taken together these results imply that in vivo, extrinsic environmental signals bias the fate of multipotent stem cells. In the case of oligodendrocyte diVerentiation, these signals are likely ventrally located while in the case of astrocytes signals may be present more dorsally. This diVerence in competence versus normal fate is important in evaluating the results of various studies that analyze the diVerentiation of stem cells. Unless challenged, stem cells may not reveal their true competence. Overall we believe the data indicate that the process of segregation of a multipotent stem cell into multiple derivatives could occur via multiple pathways. A set of sequential asymmetric binary fate/trinary fate choices or symmetric diVerentiation steps can both explain CNS development. Development of the CNS, however, can be best described by the presence of a multipotent stem cell that undergoes both symmetric and asymmetric divisions to generate intermediate precursors/blast cells, which can be distinguished from stem cells in their more limited diVerentiation potential and self-renewal ability. These intermediate precursors or blast cells are likely present in the zones of proliferation that exist at that particular stage of development, and such restricted precursors have been isolated and characterized. The probability that a stem cell will self-renew or diVerentiate and which cell type it will generate depend on extrinsic cues that act to modulate the intrinsic bias of stem cells that is conferred by their positional temporal and regional identity.

MULTIPOTENT NEURAL STEM CELLS GENERATE OLIGODENDROCYTE AND ASTROCYTES BY GENERATING MORE RESTRICTED GLIAL PRECURSORS The unambiguous demonstration of multipotent stem cells in vitro and in vivo suggests that diVerentiated cells must be derived from an initially multipotent stem cell population. How

II. GLIAL CELL DEVELOPMENT

241

242

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

the process of diVerentiation occurs is only now being clariWed. As discussed earlier, it is likely that restricted precursors are Wrst generated from multipotent stem cells and these then undergo further progressive diVerentiation. This model of diVerentiation is conceptually similar to the model of hematopoietic stem cell diVerentiation (reviewed in Rao, 1999) and predicts a direct lineage relationship between multipotent ventricular zone neuroepithelial stem cells and/or SVZ-derived neurosphere-forming stem cells and more restricted precursor cell populations. Clonal analysis in vivo or in vitro, single cell labeling, visualization of development using heritable dye tracers or retroviral constructs, and/or analysis of cultured precursor cells are a necessary prerequisite to classifying glial precursor cells and establishing a lineage relationship between them. Several groups have undertaken such studies (Levison and Goldman, 1997; Luskin et al., 1993; Parnavelas, 1999; Price et al., 1992; Williams and Price, 1995) and shown that diVerentiated oligodendrocytes and astrocytes arise from dividing precursor cells. These glial precursors share the inability to generate neurons but can be distinguished from each other by antigenic criteria, developmental potential, and growth factor response. At least Wve diVerent types of glial precursor cells (Table 9.6) have been identiWed in the past few years (Grinspan et al., 1990; Mi and Barres, 1999; Rao and Mayer-Proschel, 1997; Seidman et al., 1997; Zhang et al., 1998). Three of the better characterized precursors are discussed next.

TABLE 9.6 Cell Type

Glial Precursor Cells in Rodents

Reference

Site of isolation

Cells that diVerentiate

Glial restricted tripotential precursor cell (GRP)

GRP cells (Rao and Mayer-Proschel, 1997; Rao et al., 1998)

Embryonic spinal cord

Oligo’s Type 1 and type 2 astrocytes in vitro and in vivo

Oligodendrocyte precursor

Oligosphere (Avellana-Adalid et al., 1996; Zhang et al., 1998)

Neonatal rat brain, neurospheres from adult striatum

Olig o’s Type 1 and type 2 astrocytes in vitro and in vivo

Precursor to the oligodendrocyte-type-2 astrocyte precursor (pre-O-2A)

Pre-O-2A (Ben-Hur et al., 1998; Grinspan et al., 1990; Hardy and Reynolds, 1991)

Perinatal cerebral white matter

Olig + astrocyte in vitro

Oligodendrocyte type-2 astrocyte precursor cell (O-2A)

O-2A perinatal (RaV et al., 1983a), O-2A (Behar et al., 1988; Dutly and Schwab, 1991; Fok-Seang and Miller, 1994; Levi et al., 1987)

Perinatal optic nerve, cerebellum, cerebral cortex, brain stem, spinal cord

Oligo’s and type 2 astrocytes in vitro and Olio’s in vivo.

O-2A adult (Engel and Wolswijk, 1996; Wolswijk and Noble, 1989) O-2A like cells (Levine et al., 1993)

Adult optic nerve, adult spinal cord, Adult cerebellum

Oligo’s and type 2 astrocytes in vitro and Olio’s in vivo.

Putative type
1 astrocyte precursor (Seidman et al., 1997) APC (Mi and Barres, 1999) Astrocyte precursor (Mayer-Proschel et al., 2002)

E16 mouse cerebellum

Astrocytes in vitro

Astrocyte Precursor cell

Developing optic nerve Neonatal spinal cord

A list of the glial precursor cells isolated from rodent tissue is provided and a reference to a recent publication describing the cells type is provided. Glial precursor cells can be broadly subdivided into those that generate both astrocytes and oligodnedrocytes in vitro and in vivo, those that generate oligodnedrocytes and those that generate astrocytes.

II. GLIAL CELL DEVELOPMENT

OLIGODENDROCYTE-TYPE 2 ASTROCYTE CELLS

OLIGODENDROCYTE-TYPE 2 ASTROCYTE CELLS Oligodendrocyte-type 2 astrocyte (O2A) cells, which represent one of the best-deWned glial precursors of the CNS, were initially isolated from the post-natal rat optic nerve and subsequently from the post-natal cortex and spinal cord. O2A cells have a default pathway of diVerentiation into oligodendrocytes, and this diVerentiation can be modulated by growth factors. In culture, O2A cells can also diVerentiate into type 2 astrocytes. Type 2 astrocytes diVer from the more common type 1 astrocyte in their expression of A2B5 immunoreactivity and the absence of Ran 2 immunoreactivity. O2A cells will not diVerentiate into neurons under any culture condition and upon transplantation will diVerentiate into myelinating oligodendrocytes. O2A cells thus represent glial-restricted precursor cells that can generate a subset of the glial population present in the adult brain.

ASTROCYTE-RESTRICTED PRECURSORS Another class of precursor cells restricted to glial diVerentiation is the astrocyte-restricted precursor (APC). Seidman et al. (1997) have described astrocyte restricted precursor cells isolated from the E16 mouse cerebellum that do not express glial Wbrillary acid protein (GFAP), an astrocyte marker, and are EGF dependent. Upon diVerentiation, the cells begin to express high levels of GFAP but do not diVerentiate into oligodendrocytes. APCs are not A2B5 immunoreactive and the astrocytes that diVerentiate appear to be type 1 astrocytes. Mi and Barres (1999) have independently provided unambiguous evidence for an astrocyte precursor cell or APC. The authors show that this cell expressed A2B5 immunoreactivity and expresses some, but not all, astrocytic markers. This cell can diVerentiate into astrocytes under appropriate culture conditions. The authors argue that since these APC cells do not default to an oligodendrocyte pathway of diVerentiation and but rather diVerentiate into type 1 and not type 2 astrocytes, they are clearly distinct from O2A cells. Likewise we have used CD 44 immunoreactivity to isolate an astrocyte restricted precursor cell population that will diVerentiate into astrocytes but not into other cells (Mayer-Proschel et al., 2002, Ying et al., 2002). Examining oligodendrocyte diVerentiation from spinal cords, Richardson and colleagues (Pringle et al., 1992) noted that while oligodendrocyte diVerentiation occurred in the ventral cord and required platelet-derived growth factor receptor (PDGFR)-alpha expression, astrocytes can be generated from dorsal spinal cord from cells that do not express PDGFR-alpha. Their results suggested the existence of an astrocyte precursor that is present in the dorsal spinal cord. Whether the diVerent group has identiWed a slightly diVerent precursor population or whether the same precursor cell has been isolated multiple times remains to be determined.

GRP CELLS ARE GLIAL-RESTRICTED PRECURSOR CELLS THAT CAN DIFFERENTIATE INTO OLIGODENDROCYTES AND TYPE 1 ASTROCYTES In a series of in vitro experiments, we identiWed and characterized a glial-restricted precursor (GRP) that is present in the developing neural tube (Rao and Mayer-Proschel, 1997, Rao et al., 1998). GRPs can be identiWed as early as E12 by their A2B5 and nestin immunoreactivity. Dividing glial precursors with similar properties can be identiWed as late as post-natal day 2 (Gregori et al., 2002; Power et al., 2002). GRP cells lack PDGFRalpha immunoreactivity (at least initially) and synthesize detectable levels of PLP/DM-20. GRPs arise ventrally from a restricted region of the proliferating neuroepithelium, although the potential to generate GRPs appears much more widespread. GRP cells generate two kinds of astrocytes in vitro and in vivo and thus diVer from O2A cells, generate only

II. GLIAL CELL DEVELOPMENT

243

244

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

one kind of astrocyte in vitro, and do not diVerentiate into astrocytes in vivo (Espinosa et al., 1993; Yang et al., 2000). GRP cells thus represent a distinct progenitor population that can be distinguished from the later appearing O2A (Oligodendroblast) cell. It is important to note that GRP cells represent the earliest identiWable glial precursor and can be isolated from the entire rostrocuadal axis. O2A and APCs can in general be isolated only at later developmental stages though prior to the appearance of fully diVerentiated cells. This sequence of appearance of progressively more restricted precursors suggests though does not prove that a lineage relationship may exist between them.

GLIAL PROGENITORS MAY BE LINEALLY RELATED GRPs can generate type-1 astrocytes, type 2 astrocytes, and oligodendrocytes, while O2A cells and APCs generate subsets of these populations. Further, all three classes of glial precursors are present at diVerent stages of development. It is therefore tempting to suggest a lineage relationship between these precursor cell types. A hypothetical relationship is schematized in Figure 9.16, and data on some lineage relationships are summarized later in this discussion. A lineage relationship between multipotent stem cells and more restricted precursors has been demonstrated. Work done by Rao and colleagues have clearly demar-

FIGURE 9.16

II. GLIAL CELL DEVELOPMENT

GLIAL PROGENITORS MAY BE LINEALLY RELATED

TABLE 9.7 Origins of Different Glia Cell type Radial Glia

Source VZ

Oligodendrocytes

Regionally generated from the VZ/SVZ

Optic nerve oligodendrocytes

Migrate into the optic nerve from the forebrain VZ/SVZ

Astrocytes

Both VZ and SVZ

OEG

Olfactory placode

Pitucytes

Midbrain SVZ

Muller glia

Optic cup

Bergman Glia

Cerebellar VZ

The origin of various glia are listed. Oligodendrocytes are thought to arise predominantly from ventral regions of the developing neural tube while astrocyte from more dorsal regions. However both ventral and dorsal regions are capable of generating both types of glia when exposed to appropriate signals. Note that the origin of OEG’s is distinct from that of all other CNS glial cells

cated the lineage relationship between multipotent NEP stem cells and many of their more restricted progeny (Mayer-Proschel et al., 1997; Mujtaba et al., 1998; Rao and MayerProschel, 1997; Rao et al., 1998). At least three intermediate precursors have been shown to arise from ventricular zone derived NEP stem cells. These include a glial restricted precursor (GRP) that generates oligodendrocytes and astrocytes. They have shown a direct lineage relationship between FGF-dependent, neuroepithelial stem cells (NEP) and glial restricted precursor cells (Rao and Mayer-Proschel, 1997; Rao et al., 1998). The authors grew NEP stem cells in culture and then diVerentiated them into A2B5 immunoreactive cells. When these A2B5þ cells were isolated the authors found that NEP-derived A2B5 immunoreactive cells diVerentiated into astrocytes and oligodendrocytes but not into neurons. NEP-derived glial cells appear morphologically and antigenically similar to GRP cells directly isolated from E13.5 neural tubes (though subtle diVerences do exist). Thus, a direct lineage relationship exists between multipotent NEP stem cells and GRP cells. This demonstration of a transition from an NEP cell to a GRP cell provides the Wrst evidence that restricted precursors are an intermediate stage between pluripotent stem cells and fully diVerentiated, post-mitotic cells. In analogous experiments, a similar lineage relationship between EGF-dependent neurosphere stem cells and oligodendroglial precursors (oligospheres) has been established. Duncan and colleagues showed that a glial-restricted precursor could be isolated from canine neurospheres or rat neurospheres by manipulating cultures conditions (Zhang et al., 1998, 1999). The authors reported that, unlike GRP cells, these oligospheres were initially A2B5 immunonegative and subsequently acquired A2B5 immunoreactivity. Cells could be maintained in culture for several months and oligodendroglial progenitors underwent selfrenewal and could generate astrocytes and oligodendrocytes. Transplanted oligospheres myelinate axons in vivo (Zhang et al., 1998), indicating that cultured precursor cells were functionally competent. Hence multipotent stem cells likely generate oligodendrocytes and astrocytes via a more restricted progenitor cells. Mayer-Proschel and colleagues (Gregori et al., 2002) have established a further lineage relationship between oligodendroblasts and glial restricted precursor cells. They have shown that the tripotential glial restricted precursor cell can generate a more restricted precursor that generates oligodendrocytes but not type 1 astrocytes. We have established that astrocyte precursor cells can be generated from GRP cells as well and can be distinguished from GRP cells by the expression of CD-44. Induction of the astrocyte precursor fate is regulated by Notch/delta signaling and by BMP and LIF/CNTF. Other more restricted glial precursors have been described (discussed earlier), but whether they can be generated from cultured stem cells remains to be proven. Indeed, retroviral lineage tracing studies and cell culture assays have suggested the existence of oligodendrocyte-neuron

II. GLIAL CELL DEVELOPMENT

245

246

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

and a neuron-astrocyte precursors as well (Hulspas et al., 1997; Luskin et al., 1988, 1993; Price et al., 1988; Turner and Cepko, 1987; Young and Levison, 1996;). However, no clear-cut evidence of a lineage relationship with precursor stem cells exists. Further, isolation experiments that would allow one to analyze those speciWc precursor populations in detail need to be performed. Nevertheless, the existence of additional intermediate precursors remains a distinct possibility. Overall, the results (summarized in Fig. 9.15) suggest that two populations of stem cells present in the nervous system can generate astrocytes and oligodendrocytes. Both populations of stem cells likely generate diVerentiated progeny via the generation of progressively more restricted precursor cells, and several such precursor cells have been identiWed. A lineage relationship between some precursor cells has been established, which is consistent with the temporal appearance of these precursors in vivo. Less is known about the origin of other specialized glial populations. It is clear that olfactory ensheathing glia likely arise from the olfactory placode and thus have a distinct peripheral origin when compared with all other CNS glial populations, and this distinct origin may account for their unique properties (Table 9.7). Muller glia of the retina arise from the otpic cup, which is an extension of the forebrain vesicle, and the process of stem cell diVerentiation in this region may be distinct from model C (Fig. 9.14), which appears to describe stem cell diVerentiation in the CNS. Interestingly, oligodendrocytes that populate the optic nerve do not arise of the optic stalk; rather they migrate into the optic stalk from the forebrain VZ/SVZ. Bergman glia of the cerebellum likely arise from the VZ of the fourth ventricle rather than the EGL layer of the cerebellum, though GFAPþ cells derived from the EGL may also contribute to the glial scaVold (Abraham et al., 2001; Moskovin et al., 1978; Rakic and Sidman, 1970; Seivers et al., 1994). Bergman glia serve as a scaVold for the granule cells that are derived from the EGL, and the cerebellum would represent one of the few regions where the glial scaVold has a diVerent origin than the neurons that utilize this scaVold. Radial glial cells, which are among the Wrst glial population to develop likely, arise from proliferating ventricular zone stem cells throughout the rostrocaudal axis. Retroviral labeling in the spinal cord and tectum has shown however that radial glial arise from a common progenitor that generates neurons and other astrocytes as well (Gray and Sanes, 1992). Whether SVZ cells can generate radial glia is open to debate. Some SVZ regions lack radial glia (reviewed in Levison et al., 2002), and lineage analysis has not shown radial glial cells arising from SVZ domains. Whether this represents an inability or lack of competence rather than simply lack of appropriate diVerentiation signals in vivo remains to be determined.

REGULATION OF THE STEM CELL TO GLIAL PRECURSOR TRANSITION As a general scheme, we have suggested that multiple stages can be identiWed in the process of diVerentiation of a mutlipotent stem cell to a diVerentiated cell. Much is known about the process of migration, maturation, survival, and apoptosis at later stages of development. These have been described elsewhere (reviewed in Collarini et al., 1991; McMorris and McKinnon, 1996) and will be discussed in chapters detailing astrocyte and oligodendrocyte diVerentiation (Chapter 11 and 12). In this chapter we have focused on the early stages of stem cell to glial precursor cell diVerentiation. As discussed in previous sections, stem cells undergo regionalization, proliferate via symmetrical and asymmetrical divisions, and diVerentiate into a particular precursor cell phenotype that is appropriate for that stage of development and for that speciWc brain region. Each of these aspects of stem cell development can be modulated by extrinsic factors, and some examples of diVerent mechanisms are summarized in Figure 9.17. The present discussion is by no means complete: some factors that may be integral in gliogenesis remain unidentiWed, and combinatorial or stage speciWc eVects of such factors, if any, have not been investigated.

II. GLIAL CELL DEVELOPMENT

ALTERATION OF REGIONAL IDENTITY

ALTERATION OF REGIONAL IDENTITY As summarized in Figure 9.18 and discussed earlier, stem cells acquire rostrocaudal and dorsoventral identity that speciWes which types of neurons are made and what proportion of glia will be generated. If a rhombomere identity is latered neural crets cell may not be generated (or will be destined to die) altering the number and types of glial cells generated. Likewise, if the midbrain is converted to retinal tissue by misexpresison of frizzled-3 (Rasmussen et al., 2001), then the pattern of stem cell division is altered and the properties of the glial cell generated are completely distinct. Such regionalizing molecules have not traditionally been considered as molecules that regulate glial cell number, but since they act early and can have profound eVects, it is important at least conceptually, to include them as molecules that regulate stem cell diVerentiation.

CHOICE BETWEEN PROLIFERATION AND DIFFERENTIATION These molecules and their roles are summarized in Figure 9.17. Factors that alter cell cycle length while maintaining cell division over the usual period of development (Fig. 9.16; Sommers and Rao, 2000) will increase total numbers of stem cells. Similarly factors that alter the total number of cell divisions will alter the total number of stem cells present. These factors could be components of an intrinsic clock that counts numbers of cell divisions or it could be global regulators that are present extrinsically and that regulate total number of cell divisions (RaV et al., 1985; RaV and Lillien, 1998). Other factors may act to inhibit diVerentiation without aVecting proliferation kinetics. These inhibitory factors can prolong the stem cells response to other stimulatory factors leading to an increase in the stem cell pool. The consequence of altering cell kinetics, proliferation rates, or inhibiting diVerentiation can be a reduction in the number of diVerentiated cells seen in the short term with a increase in the total number of diVerentiated cells seen at later stages provided, of course, additional compensatory mechanisms do not come into play. Factors that may act by either mechanism have been described. FGF and Wnt-1 appear to be important in maintaining neuroepithelial cell proliferation (Dickinson et al., 1994; Kalyani et al., 1997; Kilpatrick and Bartlett, 1993; Kitchens et al., 1994, Ray et al., 1993). Indeed, FGFs appear to be the only cytokines that promote mitosis of NEP cells. EGF, PDGF, NT-3, and a variety of other cytokines tested do not appear to stimulate NEP cell division (Deloulme et al., 1991; Groves, 1992; Kalyani et al., 1997, 1998b; Shihauddin et

FIGURE 9.17 Potential glial lineages. A hypothetical glial lineage is summarized. Data supporting this model of progressive restriction are discussed in the text. It is important to note that additional precursors such as neuron-astrocyte or neuron-oligodendrocyte precursors may exist.

II. GLIAL CELL DEVELOPMENT

247

248

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

FIGURE 9.18 Regulation of stem to glial precursor transition. Possible mechanisms by which extrinsic signals may alter the fate of stem cells are schematized. Evidence that each of these mechanisms operates in normal development is available, though much needs to be learned about how these extrinsic signals interact with the intrinsic machinery of the stem cell.

al., 1997; Weis et al., 1996b). FGFR1 and FGF2 knockouts lead to a signiWcant reduction in the thickness of the ventricular zone and a net decrease in the output of diVerentiated cells. Surprisingly, EGFR knockouts do not appear to aVect the size of the VZ/SVZ but do appear to aVect neuronal migration. EGF may be important in regulating the proliferation of SVZ derived neurosphere-forming stem cells. Infusion of EGF will cause cell proliferation and enhance the number of diVerentiated cells seen subsequently in the olfactory bulb, suggesting that EGF regulates SVZ cell proliferation. It has been suggested that BDNF may be important in regulating proliferation/diVerentiation of the striatal SVZ cells (Benraiss et al., 2001), but identiWcation of the exact stage that it acts remains to determined. Cell cycle regulators such as PTEN, p27/kip, and p16 may be important in regulating cell cycle length or cell cycle duration. PTEN knockouts will prolong self-renewal and reduce diVerentiation, while p21 may alter total cycle number (Groszer et al., 2001; Li et al., 2002; RaV et al., 1998). Both molecules will aVect the total number of stem and precursor cells, albeit by diVerent mechanisms and ultimately alter the output of glial cells.

II. GLIAL CELL DEVELOPMENT

CHOICE BETWEEN PROLIFERATION AND DIFFERENTIATION

Altering total number of cell divisions

e.g.: Pten. P21, FGF

Altering cell proliferation

e.g.: cell cycle regulatros, telomerase

Altering cell survival

Biasing between symmetrical and asymmetrical divisions

Biasing Fate

Altering positional identity

e.g.: caspases

e.g.: Notch, Numb, presenillin

e.g.: BMP, PDGF, HES, neuregulin

e.g.: homeobox genes

Altered output of glial progenitors

FIGURE 9.19 Factors regulating the process. A partial list of molecules that may regulate the process of stem cell proliferation and diVerentiation is provided. It is important to note that molecules may act at multiple stages and that their eVect in combinations may be diVerent from their eVect when used in isolation.

TABLE 9.8 Multipotent neural stem cells eg: Chalmers-Redman et al.,1997, Sah et al., 1997, Brustle et al., 1998, Flax et al., 1998, Svendsen et al., 1998, Vescovi et al.,1999, Carpenter et al.,2000, Piper et al., 2000, Mayer-Proschel et al., 2002 Neuron restricted precursors N-CAM immunoreactive HNRP cells eg: Piper et al., 2001 Alpha-I tubulin expressing fetal HNRP cells eg: Roy et al et al., 2001 Alpha-I tubulin expressing adult HNRP cells eg: Roy et al., 2001 Glial Restricted Precursors Oligodendrocyte precursors fetal eg: Zhang et al., 2000 Oligodendrocyte precursors adult eg: Scolding 1998, Roy et al., 1999 Bipotential oligodendrocyte-astrocyte precursors fetal eg: Mayer-Proschel et al., (2002) Astrocyte precursor cells eg: Loo et al., 1991, Mayer-Proschel et al., 2002 Restricted precursors are cells which have a limited ability to self-renew and are capable of generating fewer cell types than multipotent stem cells. Some examples multipotent stem cells and neuronal and glial restricted precursors derived from human tissue are listed.

Although FGF promotes NEP cell proliferation, it is not suYcient to completely prevent their diVerentiation; an additional factor is required (Kalyani et al., 1997). The identity of this additional factor remains unknown, though a soluble component present in chick embryo extract is suYcient to prevent diVerentiation (Kalyani et al., 1997). We have suggested that a member of the wnt family of growth factors may be important in

II. GLIAL CELL DEVELOPMENT

249

250

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

inhibiting NEP cell diVerentiation (unpublished results). Wnts are expressed at the appropriate developmental age (for review, see Dickinson et al., 1992; Hollyday et al., 1995; McMahon et al., 1992), and overexpression of wnts can promote NEP cell proliferation (Dickinson et al., 1994). Further, wnt-1/3a knockout or antisense knockdown experiments can interfere with neural tube development (Augustine et al., 1993, 1995; Ikeya et al., 1997), and conditonal overexpression of ß-catenin and downstream signaling molecules lead to a dramatic expansion of the cortex (Chen et al., 2002). Transcription factors that play a role in inhibiting diVerentiation have also been described. HES (a homologue of Drosophila enhancer of split gene) is a negative regulator of diVerentiation. Overexpression of HES inhibits spinal precursor cells from diVerentiating, and abolishing HES expression in transgenic mice results in premature diVerentiation (Ishibashi et al., 1994, 1995; reviewed in Kageyama et al., 1997). Another group of negative regulators, the Id family of genes, is also expressed in the spinal cord (see, for example, Nagata et al., 1994; reviewed in Kageyama et al., 1997). The Id subfamily of HLH proteins may also play an important role in inhibiting stem cell diVerentiation.

CHOICE OF SELF-RENEWAL VERSUS ASYMMETRIC DIVISION Symmetrical divisions that generate two stem-like daughter cells will double the pool of dividing cells, while symmetrical divisions that generate diVerentiated progeny will remove dividing cells and halve the dividing cell pool. Asymmetrical divisions will maintain the size of the pool while generating diVerentiated progeny. If a stable precursor pool needs to be maintained, then the balance between asymmetrical and symmetrical division must be tightly regulated. Several molecules that may be important in this process have been identiWed. These include asymmetrically distributed molecules such as Notch and Numb, as well as cytoskeletal components that regulate intracellular movement of molecules, and proteolytic enzymes that regulate asymmetrical distribution by degrading molecules in unwanted compartments. It has been suggested that notch and numb may be involved in a stem cells decision to generate either a daughter stem cell or a diVerentiated progenitor. Several groups have shown that asymmetric distribution of notch and numb protein will select for diVerentiation into a stem cell or a precursor cell (reviewed in Jan and Jan, 1998). Much remains to be learned about this class of molecules and how they interact with proliferation and diVerentiation signals. Nevertheless, it is clear that this class of molecules will be important in determining the ultimate proportion of progenitor and diVerentiated cells.

CHOICE BETWEEN ONE FATE OR ANOTHER Instructive factors that direct diVerentiation along a particular fate may aVect fates directly by instructing stem cell to diVerentiate along a particular pathway or indirectly by altering other fates or by depleting the stem cell pool. This is particularly true for astrocyte generation, as astrocytes arise later in development. Premature diVerentiation of stem cells into neurons or biasing diVerentiation into the oligodendrocyte fate will reduce the number of stem cells available that can respond to cues to diVerentiate into astrocytes. Thus, it is important to examine all fates when assessing the role of a particular diVerentiation factor. Several positive diVerentiation factors have also been described. DiVerentiation of NEP stem cells into NCSCs is promoted by BMP-2/4, and diVerentiation does not require cells to divide, indicating that BMP plays an instructive role in the diVerentiation process (Liem et al., 1995; Mujtaba et al., 1998). BMP-2/4 are expressed at the appropriate time and in the appropriate location to play such a role (Liem et al., 1995; Stark et al., 1997). Factors that bias diVerentiation of NEP cells into neuroblasts and glioblasts remain elusive and to our knowledge no single molecule that selectively acts to instruct stem cells to become neuroblasts or glioblasts has been identiWed. PDGF, however, may represent one such candidate molecule that biases cortical stem cells toward a

II. GLIAL CELL DEVELOPMENT

CHOICE BETWEEN DEATH AND SURVIVAL

neuronal fate (Johe et al., 1996; Williams et al., 1997). CNTF and LIF, acting via the JAKSTAT pathway (Bonni et al., 1997), have also been suggested to act as instructive molecules that bias stem cell fate (Richards et al., 1996). However, knockout and more recent culture experiments suggest that they do not act at this early stage of development but may be important at the intermediate precursor to mature cell transition. In a series of experiments, it has been shown that Shh is both necessary and suYcient to regulate motoneuron and oligodendrocyte generation (reviewed in Ericson et al., 1995). The response is dose dependent, and at lower doses, Shh promotes the generation of interneurons (Ericson et al., 1997). Whether shh acts instructively or simply increases the precursor pool remains to be determined. Nkx2.2, ngn-3, olig1/2 all appear to act at the glial precursor stage to alter its fate (Zhou et al., 2001). Absence of these genes or a combination of genes results is an absence of myelinating oligodnedrocytes (Fu et al., 2002; Lu et al., 2000; Qi et al., 2001; Zhou et al., 2000,). However, using markers that identify diVerent stages of development, it appears as if stem cells diVerentiate into glial precursors and further maturation is blocked. Thus, it is unlikely that these factors act as instructive molecules. Notch and its downstream mediators may act to instructively promote astrocyte diVerentiation. However, the eVect is seen at the glial precursor stage (Wu et al., 2002). Overexpression of the activated form of notch or Hes-1, a downstream regulator, does not cause stem cells to diVerentiate into astrocytes, while expression in glial precursors biases their fate to the astrocytic lineage.

CHOICE BETWEEN DEATH AND SURVIVAL Stem cells are an actively dividing population that will undergo several rounds of division over a period of 4 days in rodents. Thus, a single cell will potentially generate hundreds of diVerentiated cells. Loss of a single cell at an early stage can have dramatic consequences, and a variety of evidence indicates that regulation of death is an important mechanism in regulating overall cell number and consequently the output of diVerentiated cells. DiVerences between stem cell apoptosis and target mediated apoptosis at later stages has been described (reviewed in Kuan et al., 2000; Raoth and D’Sa C, 2001). It appears that Fas and fas ligand are not expressed early and do not play an active role in stem cell apoptosis. Bcl2 does not appear to be important as it is not expressed early in development (reviewed in Sommer and Rao, 2000). Caspases appear to be important as null mice show a dramatic increase in brain size. We have examined the apoptotic pathway in vz derived neural stem cells (Luo et al., 2002) and Wnd high levels of AIF, caspase 3, and caspase 8 suggesting that these molecules may be important downstream mediators of cell death. Thus, cell death plays an important role in regulating cell number prior to the initiation of synapses and target dependent cell death. The pathways activated at early stages are distinct from the pathways required for target mediated cell death and likely included caspases and AIF.

Multiplicity of Effects An important inference from these studies is that the eVect of a particular cytokine is stage speciWc, may be direct or indirect, and, when used in combination, may have a distinct eVects. For example, BMP-2 may dorsalize cells and promote NCSC generation from NEP cells, but may act on neuronal precursors to initiate neuronal diVerentiation and on GRPs to initiate astrocytic diVerentiation. Likewise, Shh ventralizes neural tube cells, promoting both motoneuron and oligodendrocyte diVerentiatiation, but may act on other cell types to promote mitosis (Flax et al., 1997; Jensen and Wallace, 1997; Kalyani, 1998a; reviewed in Parisi and Lin, 1998). Factors in combination, Shh and BMP-2 may induce a dopaminergic phenotype (Jordan, 1997; Ye et al., 1998) rather than promote glial diVerentiation. Similar synergistic aVects can be inferred for glial maturation studies as well. For example, BMP and CNTF in combination have a diVerent aVect than either factor alone. BMP promotes

II. GLIAL CELL DEVELOPMENT

251

252

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

CD44 expression, while LIF promotes GFAP expression with little CD44 induction. Together the factors act synergistically to generate astrocytes. Overall, as indicated earlier, the process of stem cell to diVerentiated cell transition can be regulated at multiple stages. Factors can alter stem cell kinetics, act at the progenitor cell stage, or aVect the survival or maturation of diVerentiated cells, which in turn can regulate precursor cell diVerentiation. Many candidates have been identiWed and the stage speciWc roles of each molecule is being identiWed. However, we remain far from a comprehensive model to explain how molecules act in concert to direct diVerentiation in an appropriate spatio-temporal pattern. Rodent versus Human CNS diVerentiation in humans is similar to that described for rat and mouse and extends through a signiWcant period of embryogenesis (Herschkowitz, 1988; Mrzljak et al., 1990). The neural pores have closed around week 4 of gestation and brain subdivisions appear around week 5. DiVerentiation of the neural tube into an outer mantle layer and an inner proliferative zone has occurred by week 6 and some neurons have already been born. Neurogenesis and gliogenesis proceed over the next several weeks and most, though not all, neuronal proliferation is completed by 8 to 10 weeks of gestation. Gliogenesis proceeds for longer time periods and multipotent stem cells and neuronal and glial precursors can be isolated from 10 to 18 weeks of gestation (Chalmers-Redman et al., 1997; Li et al., 1998; Svendsen et al., 1997; Tohyama et al., 1991; reviewed in Kalyani and Rao, 1998). The antigens expressed by these cells as well as their response to growth factors appear similar to those expressed in mouse and rat cells. In the adult, oligodendrocyte precursors have also been identiWed (see, for example, Armstrong et al., 1992; Scolding et al., 1999). The growth properties of these precursors, as well as the sequential acquisition of phenotypic markers, appears remarkably similar to that described in rodents. Many of the markers cross react across species and human stem cells appear to respond to rodent signals after transplantation. Many of the molecules that play a role in regionalization, proliferation, and diVerentiation appear similar and a functional analysis of many of these genes show a good cross species correlation. DiVerences that we and others have noted are a diVerence in growth factor requirements for human stem cells (Carpenter et al., 1999), the relative promiscuity of GFAP expression (Piper et al., 2000), diVerences in radial glial biology due to the dramatic expansion of the forebrain (Rakic, 1978), and diYculty in generating mature oligodendrocytes from precursor cells in culture (unpublished results). Overall, however, the process of stem cell to precursor cell diVerentiation appears similar. Multipotent cells are present early in development, glial restricted and neuronal restricted precursors can be isolated at later stages of development, and it is likely that most (although not all) that we learn from rodent stem cells can be extrapolated to human stem cell development as well.

SUMMARY We have described the development of the nervous system from the process of neural induction to the development of cell type speciWc progenitors. A process of CNS/PNS segregation, regionalization, VZ/SVZ formation, stem cell proliferation, and diVerentiation into intermediate precursors that subsequently undergo further maturation characterizes rodent and human neural development. Further maturation of glial precursors includes migration to appropriate locations, aggregation with neurons, and maintenance of overall cell number and a prolonged process of myelination. These processes will be discussed in subsequent chapters.

References Abraham, H., Tornoczky, T., Kosztolanyi, G., and Seress, L. (2001). Cell formation in the cortical layers of the developing human cerebellum. Int J Dev Neurosci 19, 53–62. Alvarez-Buylla, A., Herrera, D. G., and Wichterle, H. (2000). The subventricular zone: source of neuronal precursors for brain repair. Prog Brain Res 127, 1–11.

II. GLIAL CELL DEVELOPMENT

SUMMARY

Armstrong, R. C., Dorn, H. H., Kufta, C. V., Friedman, E., and Dubois-Dalcq, M. E. (1992). Pre-oligodendrocytes from adult human CNS. J Neurosci 12, 1538–1547. Artinger, K. B., Fraser, S., and Bronner-Fraser, M. (1995). Dorsal and ventral cell types can arise from common neural tube progenitors. Dev Biol 172, 591–601. Augustine, K., Liu, E. T., and Sadler, T. W. (1993). Antisense attenuation of Wnt-1 and Wnt-3a expression in whole embryo culture reveals roles for these genes in craniofacial, spinal cord, and cardiac morphogenesis. Dev Genet 14, 500–520. Augustine, K. A., Liu, E. T., and Sadler, T. W. (1995). Interactions of Wnt-1 and Wnt-3a are essential for neural tube patterning. Teratology 51, 107–119. Baker, C. V., and Bronner-Fraser, M. (2001). Vertebrate cranial placodes I. Embryonic induction. Dev Biol 232, 1–61. Baker, C. V., Stark, M. R., Marcelle, C., and Bronner-Fraser, M. (1999). Competence, speciWcation and induction of Pax-3 in the trigeminal placode. Development 126, 147–156. Bell, E., Wingate, R. J., and Lumsden, A. (1999). Homeotic transformation of rhombomere identity after localized Hoxb1 misexpression. Science 284, 2168–2171. Benraiss, A., Chmielnicki, E., Lerner, K., Roh, D., and Goldman, S. A. (2001). Adenoviral brain-derived neurotrophic factor induces both neostriatal and olfactory neuronal recruitment from endogenous progenitor cells in the adult forebrain. J Neurosci 21, 6718–6731. Bignami, A., Eng, L. F., Dahl, D., and Uyeda, C. T. (1972). Localization of the glial Wbrillary acidic protein in astrocytes by immunoXuorescence. Brain Res 43, 429–435. Bonni, A., Sun, Y., Nadal-Vicens, M., Bhatt, A., Frank, D. A., Rozovsky, I., Stahl, N., Yancopoulos, G. D., and Greenberg, M. E. (1997). Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway. Science 278, 477–483. Bronner-Fraser, M. (2002). Molecular analysis of neural crest formation. J Physiol Paris 96, 3–8. Bronner-Fraser, M., and Fraser, S. (1989). Developmental potential of avian trunk neural crest cells in situ. Neuron 3, 755–766. Bronner-Fraser, M., and Fraser, S. E. (1988). Cell lineage analysis reveals multipotency of some avian neural crest cells. Nature 335, 161–164. Burrows, R. C., Lillien, L., and Levitt, P. (2000). Mechanisms of progenitor maturation are conserved in the striatum and cortex. Dev Neurosci 22, 7–15. Cai, J., and Rao, M. (2002). Neural stem cell and aging. In Advances in Cell Aging and Gerontology. Stem cells-a cellular fountain of youth., M. P. Mattson, and G. Van Zant, eds. (Amsterdam, Elsevier), pp. 97–116. Cai, J., Wu, Y., Mirua, T., Pierce, J. L., Lucero, M. T., Albertine, K. H., Spangrude, G. J., and Rao, M. S. (2002). Properties of a Fetal Multipotent Neural Stem Cell (NEP cell). Dev Biol (In press). Cai, L., Hayes, N. L., and Nowakowski, R. S. (1997). Synchrony of clonal cell proliferation and contiguity of clonally related cells: production of mosaicism in the ventricular zone of developing mouse neocortex. J Neurosci 17, 2088–2100. Carpenter, M. K., Cui, X., Hu, Z. Y., Jackson, J., Sherman, S., Seiger, A., and Wahlberg, L. U. (1999). In vitro expansion of a multipotent population of human neural progenitor cells. Exp Neurol 158, 265–278. Chalmers-Redman, R. M., Priestley, T., Kemp, J. A., and Fine, A. (1997). In vitro propagation and inducible diVerentiation of multipotential progenitor cells from human fetal brain. Neuroscience 76, 1121–1128. Chandross, K. J., Cohen, R. I., Paras, P., Jr., Gravel, M., Braun, P. E., and Hudson, L. D. (1999). IdentiWcation and characterization of early glial progenitors using a transgenic selection strategy. J Neurosci 19, 759–774. Chenn, A., and McConnell, S. K. (1995). Cleavage orientation and the asymmetric inheritance of Notch1 immunoreactivity in mammalian neurogenesis. Cell 82, 631–641. Chenn, A., and Walsh, C. A. (2002). Regulation of cerebral cortical size by control of cell cycle exit in neural precursors. Science 297, 365–369. Chiasson, B. J., Tropepe, V., Morshead, C. M., and van der Kooy, D. (1999). Adult mammalian forebrain ependymal and subependymal cells demonstrate proliferative potential, but only subependymal cells have neural stem cell characteristics. J Neurosci 19, 4462–4471. Collarini, E. J., Pringle, N., Mudhar, H., Stevens, G., Kuhn, R., Monuki, E. S., Lemke, G., and Richardson, W. D. (1991). Growth factors and transcription factors in oligodendrocyte development. J Cell Sci Suppl 15, 117–123. Cowan, W. M. (1979). The development of the brain. Sci Am 241, 113–133. Davis, A. A., and Temple, S. (1994). A self-renewing multipotential stem cell in embryonic rat cerebral cortex. Nature 372, 263–266. Deloulme, J. C., Baudier, J., and Sensenbrenner, M. (1991). Establishment of pure neuronal cultures from fetal rat spinal cord and proliferation of the neuronal precursor cells in the presence of Wbroblast growth factor. J Neurosci Res 29, 499–509. Dickinson, M. E., Krumlauf, R., and McMahon, A. P. (1994). Evidence for a mitogenic eVect of Wnt-1 in the developing mammalian central nervous system. Development 120, 1453–1471. Dickinson, M. E., and McMahon, A. P. (1992). The role of Wnt genes in vertebrate development. Curr Opin Genet Dev 2, 562–566. Ericson, G., Minchew, P., and Wollenzien, P. (1995a). Structural changes in base-paired region 28 in 16 S rRNA close to the decoding region of the 30 S ribosomal subunit are correlated to changes in tRNA binding. J Mol Biol 250, 407–419.

II. GLIAL CELL DEVELOPMENT

253

254

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

Ericson, J., Briscoe, J., Rashbass, P., van Heyningen, V., and Jessell, T. M. (1997). Graded sonic hedgehog signaling and the speciWcation of cell fate in the ventral neural tube. Cold Spring Harb Symp Quant Biol 62, 451–466. Ericson, J., Muhr, J., Jessell, T. M., and Edlund, T. (1995b). Sonic hedgehog: a common signal for ventral patterning along the rostrocaudal axis of the neural tube. Int J Dev Biol 39, 809–816. Espinosa de los Monteros, A., Zhang, M., and De Vellis, J. (1993). O2A progenitor cells transplanted into the neonatal rat brain develop into oligodendrocytes but not astrocytes. Proc Natl Acad Sci U S A 90, 50–54. Flax, J. D., Aurora, S., Yang, C., Simonin, C., Wills, A. M., Billinghurst, L. L., Jendoubi, M., Sidman, R. L., Wolfe, J. H., Kim, S. U., and Snyder, E. Y. (1998). Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign genes. Nat Biotechnol 16, 1033–1039. Fu, H., Qi, Y., Tan, M., Cai, J., Takebayashi, H., Nakafuku, M., Richardson, W., and Qiu, M. (2002). Dual origin of spinal oligodendrocyte progenitors and evidence for the cooperative role of Olig2 and Nkx2.2 in the control of oligodendrocyte diVerentiation. Development 129, 681–693. Garcia-Verdugo, J. M., Doetsch, F., Wichterle, H., Lim, D. A., and Alvarez-Buylla, A. (1998). Architecture and cell types of the adult subventricular zone: in search of the stem cells. J Neurobiol 36, 234–248. Gehring, W. J. (1996). The master control gene for morphogenesis and evolution of the eye. Genes Cells 1, 11–15. Gibert, S. F. (1997). Developmental Biology. In (Sinauer Associates, Inc.). Glover, J. C. (2001). Correlated patterns of neuron diVerentiation and Hox gene expression in the hindbrain: a comparative analysis. Brain Res Bull 55, 683–693. Gray, G. E., and Sanes, J. R. (1992). Lineage of radial glia in the chicken optic tectum. Development 114, 271–283. Gregori, N., Proschel, C., Noble, M., and Mayer-Proschel, M. (2002). The tripotential glial-restricted precursor (GRP) cell and glial development in the spinal cord: generation of bipotential oligodendrocyte-type-2 astrocyte progenitor cells and dorsal-ventral diVerences in GRP cell function. J Neurosci 22, 248–256. Grinspan, J. B., Stern, J. L., Pustilnik, S. M., and Pleasure, D. (1990). Cerebral white matter contains PDGFresponsive precursors to O2A cells. J Neurosci 10, 1866–1873. Groszer, M., Erickson, R., Scripture-Adams, D. D., Lesche, R., Trumpp, A., Zack, J. A., Kornblum, H. I., Liu, X., and Wu, H. (2001). Negative regulation of neural stem/progenitor cell proliferation by the Pten tumor suppressor gene in vivo. Science 294, 2186–2189. Groves, R. W., Allen, M. H., and MacDonald, D. M. (1992). Abnormal expression of epidermal growth factor receptor in cutaneous epithelial tumours. J Cutan Pathol 19, 66–72. Gudino-Cabrera, G., and Nieto–Sampedro, M. (1999). Estrogen receptor immunoreactivity in Schwann-like brain macroglia. J Neurobiol 40, 458–470. Harrison, T. A., Stadt, H. A., Kumiski, D., and Kirby, M. L. (1995). Compensatory responses and development of the nodose ganglion following ablation of placodal precursors in the embryonic chick (Gallus domesticus). Cell Tissue Res 281, 379–385. Herschkowitz, N. (1988). Brain development in the fetus, neonate and infant. Biol Neonate 54, 1–19. Hitoshi, S., Tropepe, V., Ekker, M., and van der Kooy, D. (2002). Neural stem cell lineages are regionally speciWed, but not committed, within distinct compartments of the developing brain. Development 129, 233–244. Hollyday, M., McMahon, J. A., and McMahon, A. P. (1995). Wnt expression patterns in chick embryo nervous system. Mech Dev 52, 9–25. Hulspas, R., Tiarks, C., Reilly, J., Hsieh, C. C., Recht, L., and Quesenberry, P. J. (1997). In vitro cell densitydependent clonal growth of EGF-responsive murine neural progenitor cells under serum-free conditions. Exp Neurol 148, 147–156. Hunt, P., Clarke, J. D., Buxton, P., Ferretti, P., and Thorogood, P. (1998). Segmentation, crest prespeciWcation and the control of facial form. Eur J Oral Sci 106 Suppl 1, 12–18. Hynes, M., Porter, J. A., Chiang, C., Chang, D., Tessier-Lavigne, M., Beachy, P. A., and Rosenthal, A. (1995). Induction of midbrain dopaminergic neurons by Sonic hedgehog. Neuron 15, 35–44. Ikeya, M., Lee, S. M., Johnson, J. E., McMahon, A. P., and Takada, S. (1997). Wnt signalling required for expansion of neural crest and CNS progenitors. Nature 389, 966–970. Ishibashi, M., Ang, S. L., Shiota, K., Nakanishi, S., Kageyama, R., and Guillemot, F. (1995). Targeted disruption of mammalian hairy and Enhancer of split homolog-1 (HES-1) leads to up-regulation of neural helix-loophelix factors, premature neurogenesis, and severe neural tube defects. Genes Dev 9, 3136–3148. Ishibashi, M., Moriyoshi, K., Sasai, Y., Shiota, K., Nakanishi, S., and Kageyama, R. (1994). Persistent expression of helix-loop-helix factor HES-1 prevents mammalian neural diVerentiation in the central nervous system. Embo J 13, 1799–1805. Jan, Y. N., and Jan, L. Y. (1998). Asymmetric cell division. Nature 392, 775–778. Jensen, A. M., and Wallace, V. A. (1997). Expression of Sonic hedgehog and its putative role as a precursor cell mitogen in the developing mouse retina. Development 124, 363–371. Johe, K. K., Hazel, T. G., Muller, T., Dugich-Djordjevic, M. M., and McKay, R. D. (1996). Single factors direct the diVerentiation of stem cells from the fetal and adult central nervous system. Genes Dev 10, 3129–3140. Kageyama, R., Ishibashi, M., Takebayashi, K., and Tomita, K. (1997). bHLH transcription factors and mammalian neuronal diVerentiation. Int J Biochem Cell Biol 29, 1389–1399. Kalyani, A., Hobson, K., and Rao, M. S. (1997). Neuroepithelial stem cells from the embryonic spinal cord: isolation, characterization, and clonal analysis. Dev Biol 186, 202–223. Kalyani, A. J., Mujtaba, T., and Rao, M. S. (1999). Expression of EGF receptor and FGF receptor isoforms during neuroepithelial stem cell diVerentiation. J Neurobiol 38, 207–224.

II. GLIAL CELL DEVELOPMENT

SUMMARY

Kalyani, A. J., Piper, D., Mujtaba, T., Lucero, M. T., and Rao, M. S. (1998). Spinal cord neuronal precursors generate multiple neuronal phenotypes in culture. J Neurosci 18, 7856–7868. Kalyani, A. J., and Rao, M. S. (1998). Cell lineage in the developing neural tube. Biochem Cell Biol 76, 1051–1068. Keirstead, H. S., Ben-Hur, T., Rogister, B., O’Leary, M. T., Dubois-Dalcq, M., and Blakemore, W. F. (1999). Polysialylated neural cell adhesion molecule-positive CNS precursors generate both oligodendrocytes and Schwann cells to remyelinate the CNS after transplantation. J Neurosci 19, 7529–7536. Kilpatrick, T. J., and Bartlett, P. F. (1993). Cloning and growth of multipotential neural precursors: requirements for proliferation and diVerentiation. Neuron 10, 255–265. Kitchens, D. L., Snyder, E. Y., and Gottlieb, D. I. (1994). FGF and EGF are mitogens for immortalized neural progenitors. J Neurobiol 25, 797–807. Korade, Z., and Frank, E. (1996). Restriction in cell fates of developing spinal cord cells transplanted to neural crest pathways. J Neurosci 16, 7638–7648. Kornblum, H. I., Hussain, R., Wiesen, J., Miettinen, P., Zurcher, S. D., Chow, K., Derynck, R., and Werb, Z. (1998). Abnormal astrocyte development and neuronal death in mice lacking the epidermal growth factor receptor. J Neurosci Res 53, 697–717. Kuan, C. Y., Roth, K. A., Flavell, R. A., and Rakic, P. (2000). Mechanisms of programmed cell death in the developing brain. Trends Neurosci 23, 291–297. Laywell, E. D., Kukekov, V. G., and Steindler, D. A. (1999). Multipotent neurospheres can be derived from forebrain subependymal zone and spinal cord of adult mice after protracted postmortem intervals. Exp Neurol 156, 430–433. Leber, S. M., Breedlove, S. M., and Sanes, J. R. (1990). Lineage, arrangement, and death of clonally related motoneurons in chick spinal cord. J Neurosci 10, 2451–2462. Lee, J. C., Mayer-Proschel, M., and Rao, M. S. (2000). Gliogenesis in the central nervous system. Glia 30, 105–121. Levison, S. W. (2002). Prog Neurobiol In press. Levison, S. W., Chuang, C., Abramson, B. J., and Goldman, J. E. (1993). The migrational patterns and developmental fates of glial precursors in the rat subventricular zone are temporally regulated. Development 119, 611–622. Levison, S. W., and Goldman, J. E. (1997). Multipotential and lineage restricted precursors coexist in the mammalian perinatal subventricular zone. J Neurosci Res 48, 83–94. Li, L., Liu, F., Salmonsen, R. A., Turner, T. K., Litofsky, N. S., Di Cristofano, A., PandolW, P. P., Jones, S. N., Recht, L. D., and Ross, A. H. (2002). PTEN in neural precursor cells: regulation of migration, apoptosis, and proliferation. Mol Cell Neurosci 20, 21–29. Li, M., Pevny, L., Lovell-Badge, R., and Smith, A. (1998). Generation of puriWed neural precursors from embryonic stem cells by lineage selection. Curr Biol 8, 971–974. Liem, K. F., Jr., Tremml, G., Roelink, H., and Jessell, T. M. (1995). Dorsal diVerentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell 82, 969–979. Ling, Z. D., Tong, C. W., and Carvey, P. M. (1998). Partial puriWcation of a pramipexole-induced trophic activity directed at dopamine neurons in ventral mesencephalic cultures. Brain Res 791, 137–145. Liu, Y., Wu, Y., Lee, J. C., Xue, H., Pevny, L. H., Kaprielian, Z., and Rao, M. S. (2002). Oligodendrocyte and astrocyte development in rodents: An in situ and immunohistological analysis during embryonic development. Glia 40, 25–43. Livesey, F. J., and Cepko, C. L. (2001). Vertebrate neural cell-fate determination: lessons from the retina. Nat Rev Neurosci 2, 109–118. Lu, Q. R., Yuk, D., Alberta, J. A., Zhu, Z., Pawlitzky, I., Chan, J., McMahon, A. P., Stiles, C. D., and Rowitch, D. H. (2000). Sonic hedgehog-regulated oligodendrocyte lineage genes encoding bHLH proteins in the mammalian central nervous system. Neuron 25, 317–329. Luo, Y., Cai, J., Liu, Y., Xue, H., Chrest, F. J., Wersto, R. P., and Rao, M. S. (2002). Microarray Analysis of Selected Genes in Neural Stem and Progenitor Cells. J Neurochem (In press). Luskin, M. B. (1998). Neuroblasts of the postnatal mammalian forebrain: their phenotype and fate. J Neurobiol 36, 221–233. Luskin, M. B., Parnavelas, J. G., and BarWeld, J. A. (1993). Neurons, astrocytes, and oligodendrocytes of the rat cerebral cortex originate from separate progenitor cells: an ultrastructural analysis of clonally related cells. J Neurosci 13, 1730–1750. Luskin, M. B., Pearlman, A. L., and Sanes, J. R. (1988). Cell lineage in the cerebral cortex of the mouse studied in vivo and in vitro with a recombinant retrovirus. Neuron 1, 635–647. Martens, D. J., Tropepe, V., and van Der Kooy, D. (2000). Separate proliferation kinetics of Wbroblast growth factor-responsive and epidermal growth factor-responsive neural stem cells within the embryonic forebrain germinal zone. J Neurosci 20, 1085–1095. Mayer-Proschel, M., Kalyani, A. J., Mujtaba, T., and Rao, M. S. (1997). Isolation of lineage-restricted neuronal precursors from multipotent neuroepithelial stem cells. Neuron 19, 773–785. Mayer-Proschel, M., Liu, Y., Xue, H., Wu, Y., Carpenter, M. K., and Rao, M. S. (2002). Human neural precursor cells - an in vitro characterization. Clin Neurosci Res 2, 58–69. McMahon, A. P., Gavin, B. J., Parr, B., Bradley, A., and McMahon, J. A. (1992). The Wnt family of cell signalling molecules in postimplantation development of the mouse. Ciba Found Symp 165, 199–212; discussion 212–198. McMorris, F. A., and McKinnon, R. D. (1996). Regulation of oligodendrocyte development and CNS myelination by growth factors: prospects for therapy of demyelinating disease. Brain Pathol 6, 313–329.

II. GLIAL CELL DEVELOPMENT

255

256

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

Mi, H., and Barres, B. A. (1999). PuriWcation and characterization of astrocyte precursor cells in the developing rat optic nerve. J Neurosci 19, 1049–1061. Miller, R. H. (1996). Oligodendrocyte origins. Trends Neurosci 19, 92–96. Miller, R. H., and Szigeti, V. (1991). Clonal analysis of astrocyte diversity in neonatal rat spinal cord cultures. Development 113, 353–362. Morshead, C. M., Reynolds, B. A., Craig, C. G., McBurney, M. W., Staines, W. A., Morassutti, D., Weiss, S., and van der Kooy, D. (1994). Neural stem cells in the adult mammalian forebrain: a relatively quiescent subpopulation of subependymal cells. Neuron 13, 1071–1082. Morshead, C. M., and van der Kooy, D. (2001). A new ‘spin’ on neural stem cells? Curr Opin Neurobiol 11, 59–65. Moskovkin, G. N., Fulop, Z., and Hajos, F. (1978). Origin and proliferation of astroglia in the immature rat cerebellar cortex. A double label autoradiographic study. Acta Morphol Acad Sci Hung 26, 101–106. Mrzljak, L., Uylings, H. B., Van Eden, C. G., and Judas, M. (1990). Neuronal development in human prefrontal cortex in prenatal and postnatal stages. Prog Brain Res 85, 185–222. Mujtaba, T., Mayer-Proschel, M., and Rao, M. S. (1998). A common neural progenitor for the CNS and PNS. Dev Biol 200, 1–15. Mujtaba, T., Piper, D. R., Kalyani, A., Groves, A. K., Lucero, M. T., and Rao, M. S. (1999). Lineage-restricted neural precursors can be isolated from both the mouse neural tube and cultured ES cells. Dev Biol 214, 113–127. Nagata, Y., and Todokoro, K. (1994). Activation of helix-loop-helix proteins Id1, Id2 and Id3 during neural diVerentiation. Biochem Biophys Res Commun 199, 1355–1362. Nakagawa, Y., Kaneko, T., Ogura, T., Suzuki, T., Torii, M., Kaibuchi, K., Arai, K., Nakamura, S., and Nakafuku, M. (1996). Roles of cell-autonomous mechanisms for diVerential expression of region-speciWc transcription factors in neuroepithelial cells. Development 122, 2449–2464. Nievelstein, R. A., Hartwig, N. G., Vermeij-Keers, C., and Valk, J. (1993). Embryonic development of the mammalian caudal neural tube. Teratology 48, 21–31. O’Rourke, N. A., Sullivan, D. P., Kaznowski, C. E., Jacobs, A. A., and McConnell, S. K. (1995). Tangential migration of neurons in the developing cerebral cortex. Development 121, 2165–2176. Parisi, M. J., and Lin, H. (1998). The role of the hedgehog/patched signaling pathway in epithelial stem cell proliferation: from Xy to human. Cell Res 8, 15–21. Parnavelas, J. G. (1999). Glial cell lineages in the rat cerebral cortex. Exp Neurol 156, 418–429. Pellegrino, R. G., Spencer, P. S., and Ritchie, J. M. (1984). Sodium channels in the axolemma of unmyelinated axons: a new estimate. Brain Res 305, 357–360. Piper, D. R., Mujtaba, T., Rao, M. S., and Lucero, M. T. (2000). Immunocytochemical and physiological characterization of a population of cultured human neural precursors. J Neurophysiol 84, 534–548. Power, J., Mayer-Proschel, M., Smith, J., and Noble, M. (2002). Oligodendrocyte precursor cells from diVerent brain regions express divergent properties consistent with the diVering time courses of myelination in these regions. Dev Biol 245, 362–375. Price, J., and Thurlow, L. (1988). Cell lineage in the rat cerebral cortex: a study using retroviral-mediated gene transfer. Development 104, 473–482. Price, J., Williams, B., and Grove, E. (1992). The generation of cellular diversity in the cerebral cortex. Brain Pathol 2, 23–29. Pringle, N. P., Mudhar, H. S., Collarini, E. J., and Richardson, W. D. (1992). PDGF receptors in the rat CNS: during late neurogenesis, PDGF alpha-receptor expression appears to be restricted to glial cells of the oligodendrocyte lineage. Development 115, 535–551. Pringle, N. P., and Richardson, W. D. (1993). A singularity of PDGF alpha-receptor expression in the dorsoventral axis of the neural tube may deWne the origin of the oligodendrocyte lineage. Development 117, 525–533. Qi, Y., Cai, J., Wu, Y., Wu, R., Lee, J., Fu, H., Rao, M., Sussel, L., Rubenstein, J., and Qiu, M. (2001). Control of oligodendrocyte diVerentiation by the Nkx2.2 homeodomain transcription factor. Development 128, 2723–2733. RaV, M. C., Abney, E. R., and Fok-Seang, J. (1985). Reconstitution of a developmental clock in vitro: a critical role for astrocytes in the timing of oligodendrocyte diVerentiation. Cell 42, 61–69. RaV, M. C., Durand, B., and Gao, F. B. (1998). Cell number control and timing in animal development: the oligodendrocyte cell lineage. Int J Dev Biol 42, 263–267. RaV, M. C., and Lillien, L. E. (1988). DiVerentiation of a bipotential glial progenitor cell: what controls the timing and the choice of developmental pathway? J Cell Sci Suppl 10, 77–83. Rakic, P. (1971). Guidance of neurons migrating to the fetal monkey neocortex. Brain Res 33, 471–476. Rakic, P. (1972). Mode of cell migration to the superWcial layers of fetal monkey neocortex. J Comp Neurol 145, 61–83. Rakic, P. (1974). Neurons in rhesus monkey visual cortex: systematic relation between time of origin and eventual disposition. Science 183, 425–427. Rakic, P. (1978). Neuronal migration and contact guidance in the primate telencephalon. Postgrad Med J 54 Suppl 1, 25–40. Rakic, P., and Sidman, R. L. (1970). Histogenesis of cortical layers in human cerebellum, particularly the lamina dissecans. J Comp Neurol 139, 473–500. Rao, M. S. (1999). Multipotent and restricted precursors in the central nervous system. Anat Rec 257, 137–148.

II. GLIAL CELL DEVELOPMENT

SUMMARY

Rao, M. S., and Mayer-Proschel, M. (1997). Glial-restricted precursors are derived from multipotent neuroepithelial stem cells. Dev Biol 188, 48–63. Rao, M. S., Noble, M., and Mayer-Proschel, M. (1998). A tripotential glial precursor cell is present in the developing spinal cord. Proc Natl Acad Sci U S A 95, 3996–4001. Rasmussen, J. T., DeardorV, M. A., Tan, C., Rao, M. S., Klein, P. S., and Vetter, M. L. (2001). Regulation of eye development by frizzled signaling in Xenopus. Proc Natl Acad Sci U S A 98, 3861–3866. Ray, J., and Gage, F. H. (1994). Spinal cord neuroblasts proliferate in response to basic Wbroblast growth factor. J Neurosci 14, 3548–3564. Represa, A., Shimazaki, T., Simmonds, M., and Weiss, S. (2001). EGF-responsive neural stem cells are a transient population in the developing mouse spinal cord. Eur J Neurosci 14, 452–462. Reynolds, B. A., TetzlaV, W., and Weiss, S. (1992). A multipotent EGF-responsive striatal embryonic progenitor cell produces neurons and astrocytes. J Neurosci 12, 4565–4574. Reynolds, B. A., and Weiss, S. (1996). Clonal and population analyses demonstrate that an EGF-responsive mammalian embryonic CNS precursor is a stem cell. Dev Biol 175, 1–13. Richards, L. J., Kilpatrick, T. J., Dutton, R., Tan, S. S., Gearing, D. P., Bartlett, P. F., and Murphy, M. (1996). Leukaemia inhibitory factor or related factors promote the diVerentiation of neuronal and astrocytic precursors within the developing murine spinal cord. Eur J Neurosci 8, 291–299. Richardson, W. D., Pringle, N. P., Yu, W. P., and Hall, A. C. (1997). Origins of spinal cord oligodendrocytes: possible developmental and evolutionary relationships with motor neurons. Dev Neurosci 19, 58–68. Rietze, R. L., Valcanis, H., Brooker, G. F., Thomas, T., Voss, A. K., and Bartlett, P. F. (2001). PuriWcation of a pluripotent neural stem cell from the adult mouse brain. Nature 412, 736–739. Roth, K. A., and D’Sa, C. (2001). Apoptosis and brain development. Ment Retard Dev Disabil Res Rev 7, 261–266. Rubenstein, J. L., and Beachy, P. A. (1998). Patterning of the embryonic forebrain. Curr Opin Neurobiol 8, 18–26. RuYns, S., Artinger, K. B., and Bronner-Fraser, M. (1998). Early migrating neural crest cells can form ventral neural tube derivatives when challenged by transplantation. Dev Biol 203, 295–304. Sanes, J. R. (1989). Analysing cell lineage with a recombinant retrovirus. Trends Neurosci 12, 21–28. Sauer, M. E., and Walker, B. E. (1959). Radiographic study of interkinetic nuclear migration in the neural tube. Proc Soc Exp Biol Med 101, 557–560. Scherson, T., Serbedzija, G., Fraser, S., and Bronner-Fraser, M. (1993). Regulative capacity of the cranial neural tube to form neural crest. Development 118, 1049–1062. Schmechel, D. E., and Rakic, P. (1979). A Golgi study of radial glial cells in developing monkey telencephalon: morphogenesis and transformation into astrocytes. Anat Embryol (Berl) 156, 115–152. Schoenwolf, G. C., and Nichols, D. H. (1984). Histological and ultrastructural studies on the origin of caudal neural crest cells in mouse embryos. J Comp Neurol 222, 496–505. Scolding, N. J., Rayner, P. J., and Compston, D. A. (1999). IdentiWcation of A2B5-positive putative oligodendrocyte progenitor cells and A2B5-positive astrocytes in adult human white matter. Neuroscience 89, 1–4. Seidman, K. J., Teng, A. L., Rosenkopf, R., Spilotro, P., and Weyhenmeyer, J. A. (1997). Isolation, cloning and characterization of a putative type-1 astrocyte cell line. Brain Res 753, 18–26. Sharma, K., Korade, Z., and Frank, E. (1995). Late-migrating neuroepithelial cells from the spinal cord diVerentiate into sensory ganglion cells and melanocytes. Neuron 14, 143–152. Shetty, A. K., and Turner, D. A. (1998). In vitro survival and diVerentiation of neurons derived from epidermal growth factor-responsive postnatal hippocampal stem cells: inducing eVects of brain-derived neurotrophic factor. J Neurobiol 35, 395–425. Shihabuddin, L. S., Horner, P. J., Ray, J., and Gage, F. H. (2000). Adult spinal cord stem cells generate neurons after transplantation in the adult dentate gyrus. J Neurosci 20, 8727–8735. Shihabuddin, L. S., Ray, J., and Gage, F. H. (1997). FGF-2 is suYcient to isolate progenitors found in the adult mammalian spinal cord. Exp Neurol 148, 577–586. Sievers, J., Pehlemann, F. W., Gude, S., Hartmann, D., and Berry, M. (1994). The development of the radial glial scaVold of the cerebellar cortex from GFAP-positive cells in the external granular layer. J Neurocytol 23, 97–115. Sohal, G. S., Ali, M. M., Galileo, D. S., and Ali, A. A. (1998). Emigration of neuroepithelial cells from the hindbrain neural tube in the chick embryo. Int J Dev Neurosci 16, 477–481. Sommer, L., and Rao, M. (2002). Neural stem cells and regulation of cell number. Prog Neurobiol 66, 1–18. Song, H. J., Stevens, C. F., and Gage, F. H. (2002). Neural stem cells from adult hippocampus develop essential properties of functional CNS neurons. Nat Neurosci 5, 438–445. Stark, M. R., Sechrist, J., Bronner-Fraser, M., and Marcelle, C. (1997). Neural tube-ectoderm interactions are required for trigeminal placode formation. Development 124, 4287–4295. Stemple, D. L., and Anderson, D. J. (1992). Isolation of a stem cell for neurons and glia from the mammalian neural crest. Cell 71, 973–985. Svendsen, C. N., Caldwell, M. A., Shen, J., ter Borg, M. G., Rosser, A. E., Tyers, P., Karmiol, S., and Dunnett, S. B. (1997). Long-term survival of human central nervous system progenitor cells transplanted into a rat model of Parkinson’s disease. Exp Neurol 148, 135–146. Svendsen, C. N., and Rosser, A. E. (1995). Neurones from stem cells? Trends Neurosci 18, 465–467. Threadgill, D. W., Dlugosz, A. A., Hansen, L. A., Tennenbaum, T., Lichti, U., Yee, D., LaMantia, C., Mourton, T., Herrup, K., Harris, R. C., and et al. (1995). Targeted disruption of mouse EGF receptor: eVect of genetic background on mutant phenotype. Science 269, 230–234.

II. GLIAL CELL DEVELOPMENT

257

258

9. NEURAL CELL SPECIFICATION DURING DEVELOPMENT

Tohyama, T., Lee, V. M., Rorke, L. B., and Trojanowski, J. Q. (1991). Molecular milestones that signal axonal maturation and the commitment of human spinal cord precursor cells to the neuronal or glial phenotype in development. J Comp Neurol 310, 285–299. Tsai, R. Y., and McKay, R. D. (2000). Cell contact regulates fate choice by cortical stem cells. J Neurosci 20, 3725–3735. Turner, D. L., and Cepko, C. L. (1987). A common progenitor for neurons and glia persists in rat retina late in development. Nature 328, 131–136. Uchida, N., Buck, D. W., He, D., Reitsma, M. J., Masek, M., Phan, T. V., Tsukamoto, A. S., Gage, F. H., and Weissman, I. L. (2000). Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci U S A 97, 14720–14725. Voigt, T. (1989). Development of glial cells in the cerebral wall of ferrets: direct tracing of their transformation from radial glia into astrocytes. J Comp Neurol 289, 74–88. Warf, B. C., Fok-Seang, J., and Miller, R. H. (1991). Evidence for the ventral origin of oligodendrocyte precursors in the rat spinal cord. J Neurosci 11, 2477–2488. Wichterle, H., Garcia-Verdugo, J. M., Herrera, D. G., and Alvarez-Buylla, A. (1999). Young neurons from medial ganglionic eminence disperse in adult and embryonic brain. Nat Neurosci 2, 461–466. Williams, B. P., Park, J. K., Alberta, J. A., Muhlebach, S. G., Hwang, G. Y., Roberts, T. M., and Stiles, C. D. (1997). A PDGF-regulated immediate early gene response initiates neuronal diVerentiation in ventricular zone progenitor cells. Neuron 18, 553–562. Williams, B. P., and Price, J. (1995). Evidence for multiple precursor cell types in the embryonic rat cerebral cortex. Neuron 14, 1181–1188. Wu, Y. Y., Mujtaba, T., Han, S. S., Fischer, I., and Rao, M. S. (2002). Isolation of a glial-restricted tripotential cell line from embryonic spinal cord cultures. Glia 38, 65–79. Yang, H., Mujtaba, T., Venkatraman, G., Wu, Y. Y., Rao, M. S., and Luskin, M. B. (2000). Region-speciWc diVerentiation of neural tube-derived neuronal restricted progenitor cells after heterotopic transplantation. Proc Natl Acad Sci U S A 97, 13366–13371. Ye, W., Shimamura, K., Rubenstein, J. L., Hynes, M. A., and Rosenthal, A. (1998). FGF and Shh signals control dopaminergic and serotonergic cell fate in the anterior neural plate. Cell 93, 755–766. Young, G. M., and Levison, S. W. (1996). Persistence of multipotential progenitors in the juvenile rat subventricular zone. Dev Neurosci 18, 255–265. Zhang, S. C., Ge, B., and Duncan, I. D. (1999). Adult brain retains the potential to generate oligodendroglial progenitors with extensive myelination capacity. Proc Natl Acad Sci U S A 96, 4089–4094. Zhang, S. C., Lipsitz, D., and Duncan, I. D. (1998). Self-renewing canine oligodendroglial progenitor expanded as oligospheres. J Neurosci Res 54, 181–190. Zhou, Q., and Anderson, D. J. (2002). The bHLH Transcription Factors OLIG2 and OLIG1 Couple Neuronal and Glial Subtype SpeciWcation. Cell 109, 61–73. Zhou, Q., Wang, S., and Anderson, D. J. (2000). IdentiWcation of a novel family of oligodendrocyte lineagespeciWc basic helix-loop-helix transcription factors. Neuron 25, 331–343.

C H A P T E R

10 Progenitor Cells of the Adult Human Subcortical White Matter Neeta S. Roy, Martha S. Windrem, and Steven A. Goldman

OLIGODENDROCYTE PROGENITOR CELLS OF THE ADULT MAMMALIAN BRAIN Neural Progenitor Cells of the Adult Brain Over the past few decades, historic notions of the structural immutability and cellular constancy of the adult vertebrate brain have been largely dispelled. Neurogenesis was Wrst demonstrated in the rodent olfactory bulb and the hippocampus (Altman and Das, 1965, 1966; Kaplan and Hinds, 1977; Kaplan, 1985) and the songbird vocal control centers (Goldman and Nottebohm, 1983; Nottebohm, 1985). The phenomenon of adult neurogenesis has now been described throughout vertebrate phylogeny (Goldman, 1998), including monkeys (Gould et al., 1998) and humans (Eriksson et al., 1998; Kirschenbaum et al., 1994; Pincus et al., 1998). In all species yet examined, newly generated neurons seem to be generated from multipotential stem cells, the principal source of which appears to be the periventricular subependyma (SVZ) (Goldman et al., 1993; Lois and Alvarez-Buylla, 1993). In addition, restricted pools of mitotically competent but phenotypically biased neuronal progenitor cells appear to derive from these stem cell populations. These neuronally restricted pools include the anterior subventricular zone of the forebrain and its rostral extension through the olfactory subependyma, as well as the subgranular zone of the hippocampus, each of which give rise almost exclusively to neurons in vivo. However, persistent multipotential stem cells have been reported in cultures derived from each of these regions (Gage et al., 1998), suggesting that the apparent neuronal restriction of these progenitor populations may reXect not the inherent lineage capacity of the cells, so much as local environmental signals biasing toward neuronal diVerentiation (Seaberg and van der Kooy, 2002). Besides these persistent neuronal progenitors and multipotential neural stem cells, more restricted lineages of glial progenitor cells also persist in the adult brain, in both the residual ventricular zone (Levison and Goldman, 1993; Luskin, 1993), as well as dispersed throughout the subcortical and cortical parenchyma (Gensert and Goldman, 1996; Levine et al., 2001; Noble, 199 ; Reynolds and Hardy, 1997). Indeed, in contrast to the restricted distribution of neuronal progenitor cells and SVZ stem cells, oligodendrocyte progenitor cells (OPCs) seem to be extraordinarily widespread in the adult mammalian brain.

Myelin Biology and Disorders, Volume 1

259

Copyright 2004, Elsevier (USA). All rights reserved.

260

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

Oligodendrocyte Progenitors of the Normal Adult Rodent Brain The principal class of OPCs in adult rodents is a bipotential astrocyte-oligodendrocyte progenitor cell designated the O-2A progenitor, by virtue of its generation in vitro of oligodendrocytes and type 2 astrocytes, the latter comprising the traditionally recognized Wbrous astrocytes of the white matter. These cells were initially isolated from the optic nerves of perinatal rats, as O-2A progenitors (RaV et al., 1983b). In neonatal rats, OPCs are characterized by expression of the GD3 and GQ gangliosides, the latter recognized by the monoclonal antibody A2B5, which has been used to identify this cell population (Noble et al., 1992). Though similar progenitors were long ago reported in the adult optic nerve (Vaughn, 1969), the isolation of adult OPCs, or O-2AAdult progenitors, was only accomplished relatively recently in rodents (Ffrench-Constant and RaV, 1986; Wolswijk and Noble, 1989). These cells have since been isolated from the adult rat ventricular zone, spinal cord, cerebellum, and subcortical white matter (Engel and Wolswijk, 1996; Gensert and Goldman, 1996; Levine et al., 1993). Antigenic Recognition of Adult OPCs Little is known about the natural history of the adult OPC in normal adults. In histological sections of the adult rodent brain, OPCs have mainly been identiWed by their expression of both NG2 chondroitin sulfate proteoglycan (Levine et al., 1993; Nishiyama et al., 1997) and the platelet derived growth factor-alpha receptor (PDGF-aR). The expression of PDGF-aR and the NG2 epitope substantially overlaps in rats (Nishiyama et al., 1996; Pringle et al., 1992). Moreover, a persistent population of O4/NG2 co-expressing cells has been demonstrated in the adult rat cerebral cortex, eVectively bridging the antigenic gap between early and committed OPCs (Reynolds and Hardy, 1997). On the basis of these studies, NG2-immunoreactivity has been developed as a surrogate marker for parenchymal oligodendrocyte progenitor cells. In addition, adult-derived OPCs have several features that may allow them to be distinguished: Whereas the perinatal OPC utilizes vimentin as an intemediate Wlament and does not express the oligodendrocytic sulfatide recognized by Mab O4, its adult counterpart does not express vimentin, but does express O4 (Shi et al., 1998; Wolswijk and Noble, 1989; Wolswijk et al., 1991). These parenchymal OPCs are present in both gray and white matter, and exist in vivo as extensively branched cells. The NG2 population represents as many as 5–8% of all the cells in the adult rodent brain (Dawson et al., 2000); this is congruent with earlier estimates that 5% of all glia in the optic nerve may be progenitors (Vaughn and Peters, 1968). Turnover OPCs in the adult brain may include both slowly dividing cells in normal parenchyma and a quiescent cell population that responds only to injury or demyelination. In vivo studies of the adult cerebellar cortex reveal the presence of slowly dividing OPCs with a mitotic index of 0.2 to 0.3% (Levine et al., 1993). Nevertheless, OPCs seem to constitute the main cycling population of the adult brain parenchyma. Bromodeoxyuridine (BrdU) labeling of the intact spinal cords of 13- to 14-week-old rats has shown that 10% of all cells in the white matter incorporated BrdU, of which 70% expressed NG2. In animals maintained for 4 weeks after BrdU injection, BrdU-labeled astrocytes and oligodendrocytes were noted, indicating that the cycling NG2 cells would have generated both cell types (Horner et al., 2000). In studies using retroviral labeling to mark dividing cells, 35% of the cycling cells in the adult cortex co-labeled with NG2, and these were distinctly present as clusters. Furthermore, these NG2-positive clusters doubled in size every 3 months (Levison et al., 1999). Using similar retroviral labeling techniques, the presence of cycling cells that preferentially give rise to oligodendrocytes has been shown in both the subventricular zone (SVZ) and subcortical white matter of adult rats (Gensert and Goldman, 1996; Levison and Goldman, 1993). Lineage Potential Previous studies had concluded that the perinatal OPC has a limited life span in vivo, which was attributed to a pattern of ‘‘exhaustive’’ symmetrical division and diVerentiation in

II. GLIAL CELL DEVELOPMENT

au1

OLIGODENDROCYTE PROGENITOR CELLS OF THE ADULT MAMMALIAN BRAIN

oligodendrocytes (Temple and RaV, 1986). Yet OPCs now appear to be maintained throughout life. This suggests that at least a fraction of OPCs may arise through a selfrenewing, asymmetrical divisions, such that OPCs generate both diVerentiated progeny and themselves (Wren et al., 1992). Indeed, adult OPCs of both rodents and humans retain their ability to generate oligodendrocytes and astrocytes over several generations in vitro (Tang et al., 2000). It seems likely that perinatal OPCs are the source of adult OPCs. Using time-lapse microcinematography, it has been shown that ‘‘founder cells’’ exhibiting properties of perinatal OPCs eventually give rise to cells with the properties of adult OPCs (Wren et al., 1992). As noted, just as repetitive passage of perinatal OPCs gives rise to cells with adult OPC-like properties (Wolswijk et al., 1990), slowly dividing adult OPCs can respond to FGF and PDGF by assuming the more rapid expansion kinetics typical of perinatal OPCs. Together, these data argue that perinatal and adult OPCs constitute two points along the diVerentiation spectrum of a common lineage. Nonetheless, diversiWcation within that lineage may nonetheless have resulted in substantial phenotypic heterogeneity among adult OPCs (Gensert and Goldman, 2001). Humoral Control of Oligoneogenesis Adult and perinatal OPCs share many commonalities in their responses to humoral growth factors, but nonetheless exhibit diVerential responses to both neural mitogens and diVerentiation agents. These include, but are by no means limited to, the following: 1. Platelet derived growth factor. PDGF is perhaps the most prominent described oligotrophin and has been implicated in both the mitotic expansion of OPCs and their initiation of terminal lineage commitment (Hart et al., 1989a; Noble et al., 1988; RaV et al., 1988; Wolswijk et al., 1991). OPCs uniquely express high levels of PDGFa receptor, and au2 can be speciWcally identiWed on that basis (Ellison and de Vellis, 1994; Fruttiger et al., 1999; Hart et al., 1989b). In response to PDGF, both perinatal and adult OPCs enter the mitotic au3 cycle. However, cycling time diVers in the two cell populations, in that adult OPCs have a slow, 3- to 4-day cell cycle, whereas perinatal OPCs divide daily (Noble et al., 1988); (Wolswijk et al., 1991). In OPCs derived from the adult spinal cord, PDGF alone supports the slow mitotic expansion of OPCs, as the cells divide slowly and undergo asymmetrical division, generating a diVerentiated oligodendrocyte and another progenitor (Engel and Wolswijk, 1996). However, in the presence of PDGF and FGF, adult OPCs accelerate their cycle progression, dividing rapidly and apparently symmetrically to yield additional progenitors. They then assume the bipolar morphology and A2B5 immunoreactivity of oligodendrocyte progenitor cells, but fail to generate oligodendrocytes without downstream inductive diVerentiation. As a corollary to this ‘‘perinatalization’’ of adult-derived OPCs, cultures of perinatal OPCs expanded over long periods of time in the presence of PDGF alone develop the cyclicity of adult OPCs (Tang et al., 2000). These results suggest that in rodents at least, perinatal and adultderived OPCs represent points on a continuum of diVerentiative state, rather than discrete phenotypes 2. Fibroblast growth factor. FGF diVerentially regulates OPC proliferation and diVerentiation in culture and modulates gene expression of its own receptors in a developmental and receptor type-speciWc manner (Bansal et al., 1996). Most in vitro studies show that bFGF is a major mitogen for cells in the oligodendrocyte lineage (Besnard et al., 1989; Eisenbarth et al., 1979). It has been shown to stimulate the proliferation of late progenitors and inhibit their terminal diVerentiation (Bansal and PfeiVer, 1994; McKinnon et al., 1990). More important, it establishes the responsiveness to PDGF by up-regulating the expression of PDGF-aR (McKinnon et al., 1990). Most studies with adult OPCs show that bFGF is most mitogenic when used in combination with PDGF (Mason and Goldman, 2002; Tang et al., 2000). Recently it has been shown that OPCs maintained in the presence of bFGF eventually become resistant to replicative senescence (Tang et al., 2001). Besides its well-documented eVect on OPCs, bFGF also induces the down-regulation of myelin genes, such as myelin basic protein (MBP), in mature oligodendrocytes without reverting

II. GLIAL CELL DEVELOPMENT

261

262

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

au6

them to the progenitor phenotype or eVecting reentry into the cell cycle (Bansal and au4 PfeiVer, 1997; Grinspan et al., 1993). 3. Neurotrophin-3 (NT3). Whether NT3 has proliferative or diVerentiative eVect on OPCs is yet unresolved. One study indicated that NT3, speciWcally in combination with PDGF, is proliferative for post-natal OPCs both in vitro and in vivo (Barres et al., 1994b). au5 Other studies, however, found that NT3 is not proliferative for adult OPCs alone or in combination with PDGF and bFGF (Engel and Wolswijk, 1996; Ibarrola et al., 1996). Perhaps this diVerential response may be a function of the diVerent OPC-types that have been used for the two studies. In the contused adult spinal cord, NT3 has been shown to increase OPC proliferation and myelination (McTigue et al., 1998). A recent in vitro study with OPCs from adult spinal cord dissociates indicates that NT3 induced myelination and the proliferation of O4þ/O1-cells (Yan and Wood, 2000). 4. Neuregulin. The neuregulins are a family of soluble and transmembrane protein isoforms, of which glial growth factor 2 (GGF2) is a member (Adlkofer, 2000). The neuregulins act upon erbB receptors, in particular on the erbB2, 3, and 4 heterodimeric receptors (Buonanno and Fischbach, 2001). Perinatal OPCs divide in response to GGF provided cAMP levels are high, so that adenyl cyclasae and erbB stimulation may operate synergistically as glial progenitor mitogens (Shi et al., 1998). Canoll et al. observed a similar proliferative eVect on O4þ/O1
progenitors (Canoll et al., 1996). Adult OPCs respond to GGF2 as well, although their mitogenic activation by GGF2 appears to require the concurrent activation of the PDGF receptor, along with elevated cAMP. An interesting feature of neuregulins includes their induction of phenotypic reversion by diVerentiated oligodendrocytes (Canoll et al., 1999). OPCs produce neuregulins (Raabe et al., 1997) as well as respond to it (Shi et al., 1998). Since they express full-length neuregulin erbB receptors, OPCs may utilize neuregulins as an autocrine factor, as well as a neuronally derived oligotrophin (Fernandez et al., 2000). This is likely to obtain in the environment of the adult human white matter, from which oligodendrocytes have similarly been shown to produce neuregulins and express receptors to them (Cannella et al., 1999; Deadwyler et al., 2000). 5. Triiodothyronine. When OPCs derived from optic nerves or cerebral hemispheres are cultured in the presence of T3, they immediately stop dividing and diVerentiate into oligodendrocytes. In fact, the number of times an OPC can divide varies inversely with its concentration of T3, implicating T3 as an oligodendrocytic diVerentiation factor (Baas et al., 1997). T3 seems to play a major role in controlling the timing of OPC diVerentiation (Barres et al., 1994a). Accordingly, hypothyroid states have been associated with deWcits in early myelination in neonatal cretinism, which may reXect a failure in T3-mediated OPC expansion. 6. Insulin growth factor-1 (IGF-1). During development, high levels of IGF1 are observed just before active myelination commences (Bach et al., 1991; Carson et al., 1993). IGF-1 increases proliferation and survival, enhance diVerentiation, and modulate the expression of MBP in both OPCs and oligodendrocytes (Barres et al., 1992; McMorris and Dubois-Dalcq, 1988; Saneto et al., 1988).

Oligodendrocyte Progenitors of the Adult Human Brain The earliest evidence that the adult human brain harbors oligodendrocyte progenitors came from early studies of MS lesions. Histopathologically, these lesions were found to harbor regions of extensively remyelinated axons, as well as numerous free oligodendrocytes (Moore et al., 1985; Prineas and Connell, 1979; Prineas et al., 1984). Subsequent studies identiWed populations of immature cells expressing the neural carbohydrate epitope HNK1; these were postulated to comprise early oligodendroglia, although these early studies were unable to identify any deWnitive oligodendrocyte progenitor cell phenotype (Prineas et al., 1989; Wu and Raine, 1992). PDGF-aR expressing OPCs have been shown in both MS lesions and surrounding normal white matter (Scolding et al., 1998). These PDGF-aRþ cells were found to be more

II. GLIAL CELL DEVELOPMENT

OLIGODENDROCYTE PROGENITOR CELLS OF THE ADULT MAMMALIAN BRAIN

frequent in or near MS lesions compared to normal surrounding white matter (WM), and those near lesions were more often cycling, as revealed by immunoreactivity for Ki67, a marker of proliferation (Maeda et al., 2001). Corroborating these observations with another marker of phenotype, the NG2 chondroitin sulfate proteoglycan was demonstrated in both normal adult human WM and MS lesions. As in their rodent counterparts, human NG2þ cells were found to be extensively ramiWed. Cells morphologically similar to NG2þ cells were reported to express PDGF-aR as well, although co-expression of the two by a common phenotype has yet to be directly demonstrated. Premyelinating oligodendrocytes—deWned by their expression of proteolipid protein (PLP), and their contiguity with axons despite an absence of attendant ensheathment— have also been shown in such MS lesions (Chang et al., 2002). Interestingly, NG2þ cells are virtually absent from lesions lacking premyelinating oligodendrocytes. This suggests that NG2þ cells might be the source of these premyelinating oligodendrocytes. However, the NG2 chondroitin sulfate may not be speciWc to OPCs in the adult human brain, as microglial cells express or sequester high levels of NG2-IR (Pouly et al., 1999; also Nunes, Roy, and Goldman, unpublished observations). Indeed, in dissociates of both fetal and adult human brain tissue, most NG2þ cells were microglial (Pouly et al., 1999). To establish a more reliable marker of OPCs in adult human tissues, Scolding et al. thus assessed the phenotypic speciWcity of two cardinal markers of OPC phenotype in rodents, speciWcally the PDGF-a receptor and the A2B5 epitope represented by the GQ ganglioside. By scoring the incidence of both PDGF-aRþ and A2B5þ cells in tissue print preparations of adult human white matter, Scolding and colleagues determined that these markers recognize a common parenchymal progenitor cell population. On this basis, they were able to report the Wrst estimates of the incidence of oligodendrocyte progenitor cells in the human white matter (Scolding et al., 1999). Despite this wealth of histological assessment of parenchymal progenitor cells, relatively few studies have yet correlated the antigenic expression patterns of single parenchymal phenotypes with their lineage potential, either in vivo or in vitro. As a result, it remains unclear if the expression of markers such as GD3, NG2, A2B5, or PDGF-aR is speciWc to adult OPCs, or whether it instead is shared among diVerent, already discrete lineages at similarly early points in their phenotypic speciWcation. The uncertain lineage potential of histologically antigen-deWned oligodendrocyte progenitor cells has derived in part from an historic inability to identify or isolate these cells from human brain tissues. An early attempt to identify oligodendrocyte progenitors in dissociates of adult human brain (Kim et al., 1983) was followed by successful in vitro and in vivo demonstrations of immature oligodendroglia, which were termed pro-oligodendrocytes because of their post-mitotic state. These cells were deWned as being O4þ/A2B5-/GalC- (Armstrong et al., 1992). Pro-oligodendrocytes were further characterized and found to express the PDGFaR in tissue, where they were estimated to constitute 2% of the total cell population (Gogate et al., 1994). Subsequent studies of the adult human white matter in vitro revealed the presence of mitotic cells that could give rise to oligodendrocytes, though the identity of the precursor remained unclear (Roy et al., 1999; Scolding et al., 1995);.

Humoral Control of Adult Human Oligodendrocyte Progenitor Cells Human and rodent OPCs diVer not only in their antigenic expression patterns, as noted, but also as in their responses to humoral growth factors. Adult human OPCs do not proliferate in response to bFGF, PDGF, or IGF-1, each of which can act singly as a mitogen for rodent OPCs (Armstrong et al., 1992; Gogate et al., 1994; Prabhakar et al., 1995). Instead, in human OPCs, IGF-1 has been shown to increase the proportion of postmitotic pro-oligodendrocytes and to promote the maturation of these cells as oligodendrocytes (Armstrong et al., 1992). Human OPCs also seem to be mitotically unresponsive to astrocyte conditioned medium (Armstrong et al., 1992; Gogate et al., 1994; Prabhakar et al., 1995; Scolding et al., 1995). As noted previously, neuregulin supports the expansion of OPCs and is released by neurons in an activity-dependent manner that might allow the activity-dependent modulation of OPC expansion (Canoll et al., 1996). However, these

II. GLIAL CELL DEVELOPMENT

263

264

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

observations have yet to be veriWed as operative in human OPCs. Indeed, little data are available on the factor responsiveness of human OPCs, despite the overt clinical importance of establishing the optimal expansion and diVerentiation conditions for these cells. Rather, the study of their growth factor responsiveness, patterns of receptor expression, and likely paracrine interactions with other parenchymal cell populations have been impeded by the inability to identify and isolate OPCs from the adult human brain, and hence the lack of material for molecular and cellular analysis.

Isolation of Adult Human Oligodendrocyte Progenitor Cells To address the need for isolating enriched populations of adult OPCs, we used promoterspeciWed Xuorescent activated cell sorting (FACS) to identify and extract these cells from adult human brain tissue. Traditionally, FACS has been used to sort live cells on the basis of surface antigen expression, particularly in the hematopoetic system, in which FACS has been used to deWne and isolate the major stem cell and intermediate progenitor phenotypes generated during lymphopoesis and hematopoiesis. However, the application of FACS to the nervous system was stymied by the lack of identiWed surface antigens speciWc to stage or phenotype among neural cells. Yet in 1994, the green Xuorescence protein was Wrst identiWed as a live cell reporter of gene expression (ChalWe et al., 1994). By placing GFP under the transcriptional control of promoters regulating the expression of cell-speciWc genes, we were able target speciWc cell phenotypes for FACS isolation. We Wrst applied this approach to extracting neuronal progenitor cells from the fetal ventricular zone (Wang et al., 1998), by transducing ventricular zone cells with GFP placed under the control of the Ta1 tubulin promoter, an early neuronal regulatory sequence (Gloster et al., 1994; Miller et al., 1987, 1989). This approach has since allowed us to isolate neuronal progenitor cells from both the adult human ventricular zone (VZ) and hippocampus (Roy et al., 2000a, 2000b). In addition, by modifying our choice of promoters to those speciWcally active in even earlier neural progenitors, we were able to isolate less committed neural stem cells from both the adult and fetal human brain (Keyoung et al., 2001; Roy et al., 2000a, 2000b). The development of promoter-based FACS gave use the means to identify and then isolate oligodendrocyte progenitor cells from the adult human brain (Fig. 10.1). To this end, we used the early promoter for an early oligodendrocyte protein, 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNP) (Scherer et al., 1994; Vogel and Thompson, 1988). CNP protein is the earliest myelin-associated protein known to be expressed in developing oligodendrocytes. It is expressed by oligodendrocytes at all ontogenetic stages (Sprinkle, 1989; Weissbarth et al., 1981), including by newly generated cells of oligodendrocytic lineage within the subventricular zone and their mitotic precursors (Scherer et al., 1994; Yu et al., 1994). The 5’ regulatory region of the CNP gene includes two distinct promoters, P2 and P1, which encode for two CNP isoforms, CNP1 (46kDA) and CNP2 (48 kDa). These promoters are sequentially activated during development, with the more upstream P2 promoter (P/CNP2) directing transcription to immature oligodendrocytes and their progenitors (Gravel et al., 1998; O’Neill et al., 1997). On this basis, P/CNP2 was chosen to identify oligodendrocyte progenitors from adult human subcortical white matter (Roy et al., 1999). P/CNP2:hGFP was transfected into dissociate of adult human white matter, and following GFP expression 3 to 4 days later, the P/CNP2:GFPþ cells were isolated by FACS (Roy et al., 1999). These cells, maintained in serum-deWcient media supplemented with FGF2, PDGF, and NT-3, were bipolar, immunoreactive for A2B5, incorporated BrdU from their culture media, and developed into O4þ oligodendrocytic in vitro (Fig. 10.2). These data indicated that the P/CNP2:hGFP-deWned cells were mitotic oligodendrocyte progenitors. On this basis, P/CNP2:hGFPþ oligodendrocyte progenitors were extracted directly from adult human WM dissociates using FACS. We found that an average of 0.5 + 0.1% of all white matter cells directed P/CNP2:hGFP expression. Given a transfection eYciency of 13.5%, determined using the percentage of GFP expressing cells obtained with p/CMV:GFP for noncell type speciWc transfection, it could be estimated that over 4% of adult human subcortical WM are P/CNP2-deWned progenitors. Immediately after FACS, these P/CNP2:hGFP-separated cells were initially bipolar, and

II. GLIAL CELL DEVELOPMENT

265

OLIGODENDROCYTE PROGENITOR CELLS OF THE ADULT MAMMALIAN BRAIN

Adult human subcortical white matter 1 yr to >70 yrs

Papain dissociation 5-6 days

24 hrs

Transduce with pP/CNP2:hGFP Plate 5-6 days

A2B5 Magnetic separation

P/CNP2:hGFP expression

Flow Cytometry

Trypsinize

FACS separation

Immunocytochemistry

Differentiation Sorted cells

O4 GFAP βIII-tubulin/MAP2 BrdU

Immunocytochemistry

E-17 Xenograft

Primary sphere generation

Differentiation

O4 GFAP βIII-tubulin/MAP2 BrdU

Secondary sphere generation

FIGURE 10.1 Oligodendrocyte progenitor cells may be speciWcally targeted and isolated from the white matter. This schematic outlines basic strategies for isolating oligodendrocyte progenitor cells from the adult white matter, using either Xuorescence-activated cell sorting (FACS) or a higher-yield, less speciWc alternative immunomagnetic isolation (MACS).

expressed the early oligodendrocytic marker A2B5, but none of the diVerentiated markers O4, O1, or galactocerebroside; over half incorporated BrdU. When followed up to a month in culture, >80% of the PCNP2:hGFPþ cells become oligodendrocytes, progressing through a succession of A2B5, O4, and galactocerebroside expression, recapitulating the developmental sequence of antigenic expression (Noble, 1997). Thus, with this strategy not only was the existence of oligodendrocyte progenitors established in adult human white matter, but a method was developed to separate the progenitors in a form appropriate for engraftment and further analysis.

II. GLIAL CELL DEVELOPMENT

au7

266

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

FIGURE 10.2 Sorted human white matter progenitor cells typically mature as oligodendrocytes. (A–B) A representative sort of a human white matter sample, derived from the frontal lobe of a 42-year-old woman during repair of an intracranial aneurysm. This plot shows 50,000 cells (sorting events) with their GFP Xuorescence intensity plotted against forward scatter (a measure of cell size). Part A indicates the sort obtained from a nonXuorescent P/hCNP2:lacZtransfected control, while part B indicates the corresponding result from a matched culture transfected with P/ hCNP2:hGFP. (C–D) A bipolar A2B5þ/BrdUþ cell, 48 hours after FACS. (E–F) By 3 weeks post-FACS, P/ CNP2:hGFP-sorted cells developed multipolar morphologies and expressed oligodendrocytic O4 (red). These cells often incorporated BrdU, indicating their in vitro origin from replicating A2B5þ cells. (G–I) Matched phase (G, I) and immunoXuorescent (H, J) images of maturing oligodendrocytes, 4 weeks after P/CNP2:hGFP-based FACS. These cells expressed both CNP protein (H) and galactocerebroside (J), indicating their maturation as oligodendrocytes. Scale bar ¼ 20 mm. Taken from Roy et al., 1999; with permission.

Antigenicity of Oligodendrocyte Precursor Cells As described earlier, virtually all P/CNP2:hGP-deWned OPCs are immunoreactive for A2B5 (Roy et al., 1999). This permitted us to use A2B5-based sorting to increase the yield of isolated progenitors, to numbers suYcient for experimental transplantation. Although both immature neurons and glia express A2B5-immunoreactivity during development (Aloisi et al., 1992; Eisenbarth et al., 1979; Lee et al., 1992), the adult subcortical parenchyma is relatively devoid of young neurons, allowing A2B5 to be used as a selective marker of glial and oligodendrocyte progenitor cells (RaV et al., 1983a; Satoh et al., 1996; Scolding et al., au8 1999). The speciWc use of A2B5 as an antigenic surrogate for P/CNP2:hGFP-deWned OPCs has thus constituted a signiWcant practical advance. By extracting OPCs via A2B5-based surface-antigen based sorting, the limitations of transfection-based tagging, which include

II. GLIAL CELL DEVELOPMENT

OLIGODENDROCYTE PROGENITOR CELLS OF THE ADULT MAMMALIAN BRAIN

direct cytotoxicity as well as low eYciency, can be avoided entirely. As a result, the practical issue of acquiring suYcient numbers of viable OPCs to permit transcriptional and biochemical analysis, as well as engraftment studies, can now be eVectively addressed.

Multipotential Progenitors of the Adult Human White Matter Like their lower species counterparts, human OPCs may not be strictly dedicated or autonomously programmed to oligodendrocytic diVerentiation. When puriWed from adult human subcortical tissue, derived from surgically resected temporal lobe, white matter progenitor cells (WMPCs) give rise largely to oligodendrocytes. However, when grown under conditions of very low density, we noted that these cells also generate occasional neurons (Roy et al., 1999). On this basis, we asked whether the white matter progenitor cells of the adult human brain might actually constitute a type of multipotential neural progenitor or neural stem cell. We found that white matter progenitor cells, puriWed by FACS from the adult human brain, can indeed generate neurons as well as both major glial cell types—astrocytes and oligodendrocytes—when raised in culture under conditions of high purity and low density (Nunes et al., 2003). Under these conditions, the cells are eVectively removed from other cells, as well as from the proteins that other cells may secrete. Under these conditions, the sorted progenitor cells divide and expand as multipotential clones that generate neurons as readily as oligodendrocytes (Fig. 10.3). They can continue to divide and expand for several months in culture, dividing to increase their

FIGURE 10.3

au10

Adult human WMPCs give rise to multipotential neurospheres. (A) First-passage spheres generated from A2B5sorted cells 2 weeks post-sort. (B) First-passage spheres arising from P/CNP2:hGFP sorted cells, 2 weeks. (C) Second-passage sphere derived from an A2B5-sorted sample, at 3 weeks. (D) Once plated onto substrate, the primary spheres diVerentiated as bIII-tubulinþ neurons (red), GFAPþ astrocytes (blue), and O4þ oligodendrocytes (green). (E) Neurons (red), astrocytes (blue), and oligodendrocytes (green) similarly arose from spheres derived from P/CNP2:GFP-sorted WMPCs. (F–H) BrdU incorporation (blue) revealed that new neurons (F: bIII-tubulin in red; G: MAP2 in red) and oligodendrocytes (H: O4 in green) were both generated in vitro. (I) bIII-tubulinþ neurons (green) co-expressed neuronal Hu protein (Barami et al., 1995; Marusich et al., 1994) (red, yielding; yellow doublelabel). Nuclei counterstained with DAPI (blue). From Nunes et al. (2003). Scale: A–E, 100 mm; F–I, 24 mm.

II. GLIAL CELL DEVELOPMENT

267

268

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

numbers in the process. Moreover, upon xenograft to the developing fetal rat forebrain, adult human WMPCs can mature into neurons as well as oligodendrocytes and astrocytes in vivo, in a region- and context-dependent manner (Fig. 10.4). The nominally glial progenitor cell of the adult human white matter thus appears to constitute a multipotential neural progenitor. These cells appear to be typically restricted by their local brain environment to produce only oligodendrocytes and some astrocytes, in response to local environmental signals whose identities remain to be established. But when removed from the

FIGURE 10.4 WMPCs engrafted into fetal rats gave rise to neurons and glia in a site-speciWc manner. Sections from a rat brain implanted at E17 with A2B5-sorted WMPCs and sacriWced a month after birth. These cells were maintained in culture for 10 days prior to implant. (A–B) Nestinþ (red) progenitors and doublecortinþ (red) migrants, respectively, each co-expressing human nuclear antigen (hNA, green) in the hippocampal alvius. (C) CNPþ oligodendrocytes (red) that were found exclusively in the corpus callosum. (D) A low-power image of GFAPþ (green, stained with anti-human GFAP) astrocytes along the ventricular wall. (E) bIII-tubulinþ (green)/hNAþ (red) neurons migrating in a chain in the hippocampal alvius. (F) bIII-tubulinþ and MAP2þ (inset in part F) neurons in the striatum, adjacent to the RMS (antigens in green; hNA in red; yellow: double-stained human nuclei). (G) An Huþ/ hNAþ neuron in the septum. (H) An hNAþ (green)/GAD-67þ (red) striatal neuron. Insets in each Wgure show orthogonal projections of a high-power confocal image of the identiWed cell (arrow). From Nunes et al. (2003). Scale: A–E, 40 mm; F–H, 20 mm.

II. GLIAL CELL DEVELOPMENT

THERAPEUTIC POTENTIAL OF HUMAN OLIGODENDROCYTE PROGENITOR CELLS

environment of the brain and from other brain cells, these cells proceed to make all brain cell types, including neurons and glia, and remain able to do so for long periods of time in culture. This observation has precedent in lower species. Progenitor cells capable of giving rise to multiple lineages, including oligodendrocytes and neurons, have been consistently derived from the cortical and subcortical parenchyma as well as from the ventricular zone of embryos (Davis and Temple, 1994; Qian et al., 1997; Williams et al., 1991). Similar multipotential progenitors have shown to exist in early postnatal rat cortex (Marmur et al., 1998). A more recent study suggested that postnatal rat optic nerve derived O-2A progenitor cells could be ‘‘reprogrammed’’ to multipotential stem cells capable of generating neurons (Kondo and RaV, 2000). This was achieved by sequential exposure of O-2A progenitors to serum to induce astrocytic diVerentiation, followed by their expansion in the presence of bFGF in serum-free conditions. Constant mitogenic stimulation of adult rat forebrain parenchymal cells with FGF2 has been shown to result in the generation of neurons as well as astrocytes and oligodendrocytes (Palmer et al., 1995; Richards et al., 1992). Together, these observations of the multilineage potential of CNS glial progenitors suggest that the apparent lineage commitment of progenitors might depend on epigenetic factors. As a result, the nominally glial progenitors of the adult white matter may retain far more lineage plasticity and competence than traditionally appreciated. Adult subcortical P/ CNP2:hGFPþ progenitors, though competent to generate multiple cell types, may therefore be restricted to the oligodendrocytic lineage by virtue of the epigenetic bias imparted by their environment before their isolation. A corollary of the environmental restriction of WMPC phenotype is that other, nonwhite-matter-derived neural progenitors might similarly restrict to oligodendrocytic lineage when presented to the environment of the adult white matter. Indeed, several groups have reported that EGF-expanded murine neural stem cells diVerentiate as oligodendrocytes upon xenograft (Mitome et al., 2001); remarkably, in none of these models were substantial numbers of oligodendrocytes generated in vitro. Similarly, v-myc transformed neural stem cells transplanted to perinatal mice can diVerentiate as oligodendrocytes once recruited to the white matter (Yandava et al., 1999), but not otherwise, and never in vitro.

The Distribution and Heterogeneity of White Matter Progenitor Cells The persistence and sheer abundance of WMPCs in the adult human brain is striking: Over 3% of the white matter cell population may be sorted on the basis of CNP2:GFP-based FACS, and over half of these cells are mitotically active upon isolation (Roy et al., 1999). That being said, the extent to which this parenchymal progenitor cell population is homogeneous remains unclear; by limiting dilution analysis, only 0.2% of its cells are multipotential (Nunes et al., 2003). Nonetheless, the very existence of multipotential progenitors scattered throughout the white matter parenchyma forces us to reconsider our understanding of both the nature and incidence of neural stem cells in the adult brain and challenges our conception of the supposed rarity of adult neural progenitor and stem cells. In doing so, they point to an abundant and widespread source of cells, which may be used both as a target for pharmacological induction and as a cell type appropriate for therapeutic engraftment to the diseased adult brain.

THERAPEUTIC POTENTIAL OF HUMAN OLIGODENDROCYTE PROGENITOR CELLS The Natural History of Remyelination in the Adult CNS The existence of active remyelination in the adult human brain has been mainly derived from observations of MS lesions. However, it has been unclear whether that remyelination has been the result of local expansion of parenchymal OPCs or of the recruitment of distant OPCs to sites of acute demyelination. Moreover, the source and in resting phenotype of the

II. GLIAL CELL DEVELOPMENT

269

270

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

remyelinating cells has been unclear. To address these questions, Gensert and Goldman (1997) used a combination of retroviral labeling and lysolecithin-induced demyelination to show that normally cycling cells of the adult rodent WM can diVerentiate as myelinating oligodendrocytes (Gensert and Goldman, 1997). Interestingly, before the endogenous OPCs participated in remyelination, they proliferated locally. Similarly, mice infected with a demyelinating murine hepatitis virus exhibited almost a 14-fold increase in PDGF-aRþ OPCs in the lesion bed (Redwine and Armstrong, 1998). Other studies using rats with EAE or ethidium bromide lesions have shown that after remyelination, OPC numbers were stable (Levine and Reynolds, 1999). This in turn suggested that OPCs can undergo asymmetric division to replicate themselves while generating a diVerentiating oligodendrocyte. There appears to be limited survival of OPCs in demyelinated lesions; as a result, most remyelination may be accomplished by unaVected OPCs recruited from the lesion surround. Carroll et al. have shown that OPCs in regions adjacent to immunolytic lesions Wrst respond by dividing, followed by their migration into the lesion, and ultimate myelinogenesis (Carroll et al., 1998). Similar observations were made in the demyelinated adult spinal cord, where the population of NG2þ cells expanded signiWcantly in areas adjacent to demyelinating lesions. In this case though, the proliferating pool appeared unable to sustain its self-renewal, as NG2þ cells were depleted following remyelination (Keirstead et al., 1998). Using X-irradiation, Chari and Blakemore (2002) reported that locally recruited NG2þ and PDGF-a Rþ OPCs can repopulate depleted areas over distances of approximately 0.5 mm per week in the Wrst month. No secondary progenitor loss was observed in those surround regions from which progenitor cells were recruited, indicating dynamic replacement of the emigrants (Chari and Blakemore, 2002). However, the question of how far the progenitor population can migrate in intact tissue remains debatable, an issue of particular concern for remyelination strategies involving transplantation (Franklin and Blakemore, 1997). Complicating matters further, recent studies have reported an age-related decrease both in recruitment of OPCs and in their subsequent diVerentiation (Sim et al., 2002).

Candidate Cellular Vectors for Experimental Remyelination via Progenitor Implantation Progenitor cells capable of local cell genesis therefore persist throughout the subcortical white matter of the adult brain, where they might constitute a potential substrate for cellular replacement and local repair. However, several criteria must be considered when evaluated the transplantation potential of any progenitors. These include the ability of transplanted cells to survive in the host environment, to migrate accurately to the target lesion or tissue type, to generate myelin, to ensheath host axons, and to achieve a degree of myelination capable of functional reconstitution. To assess the myelinogenic potential of transplanted cells, a variety of cell types including multipotential stem cells and OPCs, derived from both animals and humans, have been tested in both developmentally dysmyelinated and experimentally demyelinated models of myelin loss. Neural Stem Cells and Progenitors from the Fetal Brain Human fetal brain cells have been found to have robust myelinogenic capacity in the shiverer mice (Gansmuller et al., 1986; Gumpel et al., 1987, 1989). Cells isolated from the rodent or human fetal forebrains at various gestational ages, and expanded in vitro under a variety of serum-free, factor-supplemented conditions, have been used as sources of transplantable cells (Ader et al., 2001; Brustle et al., 1998; Carpenter et al., 1999; Fricker et al., 1999; Hammang et al., 1997). However, there are potential risks to prolonged in vitro expansion, since the cells are not only exposed to exogenous mitogens, but also to autocrine factors in artiWcially high concentration, and to paracrine agents produced by the neurons and glia present within the initially mixed cultures. As a result, propagated stem or progenitor cells may not retain or reXect the lineage potential or diVerentiation competence of the native progenitor cells from which they derived. Two recent studies have highlighted the eVects of in vitro expansion of cells prior to transplant. Buchet et al. observed that freshly isolated cells

II. GLIAL CELL DEVELOPMENT

THERAPEUTIC POTENTIAL OF HUMAN OLIGODENDROCYTE PROGENITOR CELLS

proliferated longer and gave rise to very extended grafts before diVerentiating into neurons and glia while cells that were expanded prior to transplant showed poor proliferation and quick diVerentiated capacity (Buchet et al., 2002). In contrast, Englund et al. found that after 9 weeks of expansion, human fetal brain cells lost the capacity to diVerentiate and remained as undiVerentiated progenitors when transplanted into adult recipients (Englund et al., 2002). To circumvent the issue of paracrine eVects on deWned stem cells in mixed culture, several groups have developed methods of directly isolating neural stem cells from tissue, thereby preventing their in vitro exposure to diVerentiated cell products during either isolation or expansion (Keyoung et al., 2001; Uchida et al., 2000). Neural Stem Cells and Progenitors from Adult Brain Several studies describe the use of neural stem cells derived primarily from the adult rat and human VZ, and then propagated as neurospheres, as a potential source of myelinogenic cells (Akiyama et al., 2001; Kukekov et al., 1999; Zhang et al., 1999). As described earlier, the adult human white matter harbors an abudance of oligodendrocyte progenitors. By virtue of their abudance, these progenitors represent a potential cellular substrate for therapeutic transplantation. Nonetheless, only a few studies, constrained by the lack of any reliable method to isolate these cells, have attempted to assess the myelinogenic capacity of OPCs derived from the adult human white matter. In one such study (Targett et al., 1996), a crude cell preparation derived from adult human white matter was transplanted into the ethidium bromide-lesioned and radiosensitized, X-irradiated adult rat spinal cord. The transplanted oligodendrocytes survived in the demyelinated zone, associated with denuded host axons, and expressed myelin proteins. But the transplanted cells did not migrate or divide, nor was any myelination noted. The failure of these implanted oligodendrocytes to myelinate was attributed to the diminished regenerative potential of post-mitotic oligodendrocytes, and the lack of a permissive environment for remyelination within the rat lesion bed (Targett et al., 1996). Propagated Oligospheres Though neural stem cells have myelinogenic capacity, they also have the inherent capacity to generate neurons and astrocytes. The co-generation of astrocytes may not necessarily be deleterious, given their roles in both OPC proliferation and myelination (Blakemore, 1992; Franklin et al., 1991). However, the co-generation of neurons may be undesirable, given the potential generation of ectopic neuronal foci, which might conceivably act as epileptogenic foci. Thus, priming neural stem cells or OPCs toward oligodendrocytic diVerentiation prior to implant might be necessary to ensure the quantities and phenotypic homogeneity of oligodendrocyte progenitor cells that will be needed for clinical implantation. One approach to this goal has been the expansion of neural stem cells as neurospheres in the presence of oligodendrocyte-inducing agents. For instance, when rat cerebellum-derived neurospheres were propagated in the presence of conditioned medium from the neuroblastoma B104 line (B104/CM), oligodendrocytes were preferentially generated. The resultant ‘‘oligospheres’’ were capable of being exponentially expanded through several passages without phenotypic degradation and exhibited robust myelination on transplantation into the shiverer mice brain (Avellana-Adalid et al., 1996). Since then, several groups have used a similar strategy to generate oligospheres from neural precursor cells of the mouse, rat, and canine forebrains (Vitry et al., 1999; Zhang et al., 1998). Smith and Blakemore the compared the remyelinating capacity of cells isolated from porcine SVZ within hours after dissociation, to that exhibited by matched cells after growth in B104/CM as oligospheres. Whereas the freshly isolated SVZ cells remained undiVerentiated after xenograft, those expanded in B104/CM eVected signiWcant remyelination of demyelinated axons in vivo (Smith and Blakemore, 2000).

Human OPCs Integrate When Grafted to Demyelinated Foci of the Adult Rat Brain The remyelinating potential of adult human white matter-derived progenitors has been recently shown in lysolecithin-induced demyelinating lesions of adult rat corpus callosum

II. GLIAL CELL DEVELOPMENT

271

272

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

(Windrem et al., 2002). In this study, A2B5 expression by P/CNP2:hGP-deWned OPCs (Roy et al., 1999) provided the rationale for immunomagnetically selecting OPCs on the basis of A2B5 expression. Like P/CNP2:hGFPþ cells, A2B5-sorted cells generated largely oligodendrocytes when raised at high density in the presence of serum. In addition, immunomagnetic selection allowed their higher-yield acquisition, without the losses in viability and number associated with FACS separation. As a result, A2B5-antibody based immunomagnetic sorting increased the yield of extractable OPCs by over 5-fold. These A2B5-sorted white matter progenitors were transplanted into cyclosporine-immunosuppressed adult rats, 3 days after lysolecithin lesions. As previously described (Gensert and Goldman, 1997), these lesions yielded a discrete region of transcallosal demyelination, with mild local reactive astrocytosis within the demyelinated focus, and intact vasculature. When A2B5-sorted human OPCs were injected into these lesions, they migrated widely and rapidly; within 7 days of implantation, the cells had readily traversed the midline to inWltrate the furthest reaches of the demyelinated lesion beds, which often extended over 8 mm in breadth. The migration rate of the cells was hence at least 1 mm/day, or 50 mm per hour, within the lesion borders (Fig. 10.5). The engrafted adult A2B5-sorted progenitors diVerentiated rapidly, expressing CNP within 2 weeks and MBP within 3 weeks of implantation. These OPC-derived oligodendrocytes projected MBPþ lamellopodia and were associated with a branched array of myelinating Wbers, indicating the initiation of progenitorassociated myelinogenesis. Of note, many transplanted progenitor derived astrocytes were also observed in the lesions. With cyclosporine immunosuppression, the cells could survive at least 2 months in lysolecithin-demyelinated recipients. These Wndings suggested that the introduction of highly enriched preparations of progenitor cells derived from the adult human white matter might permit local remyelination.

Migratory Characteristics of human OPCs Adult human-derived OPCs engrafted into demyelinated brain remained restricted to regions of demyelination; they rarely extended into normal surrounding myelin (Fig. 10.5). Even the few cells that were typically noted to have inWltrated normal myelin appeared to have migrated along the extraluminal surfaces of penetrating blood vessels. Yet when lentiviral-GFP tagged A2B5-sorted progenitor cell pools from adult human white matter were implanted into intact subcortex of adult rats, the transplanted cells remained localized to the implant site and continued to be so even after 3 months (Windrem et al., 2002). These observations suggest strongly that normal adult white matter is non-permissive for the migration of adult-derived WMPCs, as has been observed in other studies (Iwashita et al., 2000). This restriction on migration may be similar at the molecular level to that observed toward axons, whose extension through normal white matter is suppressed by their expression of Nogo receptor, by which they respond to myelin-associated Nogo and MAG (myelin-associated glycoprotein) with repulsion and/or cessation of further advance (Grandpre and Strittmatter, 2001). That being said, the operative white matter signals that restrict progenitor cell migration have yet to be identiWed. Whatever its mechanism, normal myelin clearly retains cues suYcient to tonically impede WMPC inWltration; accordingly, demyelination appears to remove those cues, allowing the active invasion and dispersion of OPCs throughout regions of acute myelin loss. The characterization of the ligands providing these repulsive cues, and of their anticipated progenitor cell receptors, will likely constitute an important avenue of future study.

Myelin Construction by Perinatal Transplant-Based Therapy Several models of congenital dysmyelination have been used to assess the myelinogenic potential of animal and human-derived progenitors. The myelinogenic potential of implanted fetal human brain cells was Wrst noted in the shiverer mouse (Gumpel et al., 1987; Lachapelle et al., 1983). The myelinogenic potential of diVerent, stage-deWned phenotypes of oligodendrocyte progenitors, extracted so as to sample the engraftment eYcacy of diVerent stages of progenitor progression, have also been compared in shiverer mice. Using rat donor tissue, Warrington and PfeiVer established that the A2B5-deWned oligo-

II. GLIAL CELL DEVELOPMENT

THERAPEUTIC POTENTIAL OF HUMAN OLIGODENDROCYTE PROGENITOR CELLS

FIGURE 10.5 Implanted white matter progenitors migrated widely throughout the demyelinated callosum. Sorted adult human white matter progenitors were transplanted into lysolecithin-induced demyelinated lesions in the corpus callosa of adult rats. (A) Lysolecithin infusion (1 ml of 2% lysolecithin-V, delivered into the corpus callosum) yielded demyelinated plaques in the target white matter. In part A, the lesion is visible as a discoid region of myelin basic protein (MBP)-immunonegativity, surrounded by the otherwise MBPþ callosum (green). (B) Though denuded of myelin (MBP, blue), neuroWlamentþ axons (green) initially survived lysolecithin demyelination, 1 week after callosal lesion. The implanted progenitors (orange) have just immigrated to the lesion bed. Axonal spheroids were frequent within the lysolecithin-lesion bed, indicating some degree of early injury and transection. The ability of implanted progenitors to eVect repair is limited by the viability and integrity of the axonal cohort that one wishes to myelinate. (C) This low-power montage demonstrates the rapidity of long-distance migration by xenografted adult human white matter progenitors. These DiI-labeled human progenitor cells (red) were visualized 1 week after their implantation, by which point the cells extend throughout the demyelinated lesion, deWned by its loss of myelin basic protein (MBP)-immunoreactivity (green). The lesion was induced 3 days before 105 sorted, DiI-tagged (red) human progenitors were delivered in 2 ml. Within a week of implantation into this demyelinated callosum, the cells had traversed the midline. (D) A higher magniWcation image showing that the transplanted cells migrated throughout the demyelinated plaque, but not beyond its borders, except for occasional migrants that followed the parenchymal surfaces of blood vessels (arrow). The restriction of migration to demyelinated regions suggests that normal myelin impedes the migration of these cells. (E) Human white matter progenitor cells, identiWed as human nuclear antigenþ (HNA; green), occupied the MBP (green)-deWcient lysolecithin lesion, and expressed oligodendrocytic CNP (red) by 15 days after implantation. (F) A cluster of HNAþ human cells (green) associated with a plethora of donor-derived MBPþ, myelinating oligodendrocytic lamellopodia (red). (G) Lentiviral GFP-tagged human (green) MBPþ (red) oligodendrocytes in the lesion bed of a lysolecithin-injected rat callosum, 8 weeks after cell implantation. Besides the MBPþ cells (arrows), other human progenitor-derived cells were also present, which did not express MBP and which instead manifested astrocytic morphologies (arrowheads). Immunolabeling adjacent sections for human GFAP (red) revealed that many of GFP-tagged human progenitors had also given rise to astrocytes. From Windrem et al. (2002). Scale bars: A, 200 mm; B, 20 mm; D, 100 mm; E, 30 mm.

II. GLIAL CELL DEVELOPMENT

273

274

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

dendrocyte progenitor phenotype was more eYcient at migration and myelinogenesis in neonatal shiverers than the more mature O4-deWned oligodendrocyte (Warrington et al., 1993). Yandava et al. similarly achieved myelination within the shiverer brain, using the C17.2 line of transformed murine cerebellar progenitor cells, which act as neural stem cells after v-myc immortalization (Yandava et al., 1999). Similarly, fetal oligodendrocytes transplanted to the md rat remyelinated signiWcant portions of the postnatal spinal cord (Archer et al., 1994). Moreover, analogous studies in the shaking pup showed that fetal oligodendrocytes were able to engraft widespread regions of the shaking CNS, with graft survival of over 6 months. Although neonatal recipients fared best, adult recipients also exhibited graft oligodendrocyte survival and stable myelination (Archer et al., 1997). Duncan and colleagues then demonstrated that oligosphere-derived cells raised from the neonatal rodent subventricular zone could engraft another dysmyelinated mutant, the myelin-deWcient rat, upon perinatal intraventricular administration (Learish et al., 1999). The success of these approaches led then to the seminal work of Mitome and colleagues, who used EGF responsive primary neural progenitor cells, in tandem with a combination of ventricular and cisternal transplant, to achieve the widespread myelination of the shiverer brain (Mitome et al., 2001).

Human OPCs Can Myelinate Congenitally Dysmyelinated Brain On the basis of these studies, Windrem et al. investigated whether highly enriched populations of human progenitor cells, directly isolated from the brain, might be used for cellbased therapy of congenital dysmyelination. SpeciWcally, this study postulated that the eYciency of myelination might be improved by using puriWed OPCs, derived via selection so as to exclude astrocytes, microglia, and vascular derivatives from the implanted pool. It further postulated that such puriWed human OPCs, both adult-derived and taken from the fetal brain during its period of maximal oligoneogenesis, would be suYciently migratory and myelinogenic to mediate the widespread myelination of a perinatal host. To this end, A2B5-based FACS was used in conjunction with PSA-NCAM-dependent immunodepletion of neuronal derivatives, to prepare highly enriched dissociates of human OPCs, of both fetal and adult derivation. Both classes of human oligodendrocyte progenitor cells proved capable of widespread and high-eYciency myelination of the shiverer brain after perinatal xenograft. Indeed, the cells migrated so widely as to eVect myelination throughout the recipient brains (Fig. 10.6, unpublished data). The cells inWltrated widely throughout the presumptive white matter, ensheathed resident murine axons, and formed antigenically and ultrastructurally compact myelin. After implantation, the cells slowed their mitotic expansion with time and generated neither undesired phenotypes nor parenchymal aggregates. In this initial study, despite histologically extensive myelination in these animals, no change in the behavioral phenotype of the shi/shi recipients or any improvement in their neurological phenotype was evident. Nonetheless, the geographic extent of forebrain and diencephalic MBP expression evidenced by these animals, who received but a single perinatal intraventricular cell injection, suggested that combined cisternal and intraventricular delivery of donor progenitors might achieve remyelination throughout the rostral neuraxis, potentially spanning the entire brain. Besides demonstrating the myelinogenic capacity of the transplanted cells, studies in the dysmyelinated animal models indirectly indicate that congenital dysmyelination, even more so than adult demyelination, may be an appropriate target for CNS progenitor cell-based therapy. In particular, these studies aYrmed that the neonatal brain environment may be especially amenable to therapeutic remyelination. It is conducive to widespread migration and may continue to provide the instructive developmental cues necessary for region-speciWc diVerentiation.

Fetal and Adult OPCs Differ Despite the use of both fetal and adult-derived OPCs in experimental therapeutic models, no head-to-head comparison of the two phenotypes had ever been performed from

II. GLIAL CELL DEVELOPMENT

THERAPEUTIC POTENTIAL OF HUMAN OLIGODENDROCYTE PROGENITOR CELLS

275

FIGURE 10.6 Myelin basic protein (MBP) was widely expressed by human fetal OPCs implanted into neonatal homozygote shiverer mice. (A) This low-power view of the recipient Wmbria shows abundant Wber-associated MBP expression (green), 3 months after perinatal engraftment (MBP, green) Since shiverer homozygotes do not express immunoreactive MBP, all such signal must derive from donor progenitor cells. (B-D) Donor fetal OPCs, additionally validated as such by human nuclear antigen (hNA, red ), diVerentiated to express CNP protein (green in B) and MBP (green in C-D). In the 0.5 mm confocal optical section of D, MBP (green) is noted to surround the donor human nucleus (red ), as viewed in orthogonal planes. (E). A single donorderived MBPþ oligodendrocyte that has matured, 3 months after engraftment, to associate with multiple recipient axons. (F) An 0.2 mm optical section through a recipient corpus callosum shows engrafted human OPCs (hNA, blue) expressing MBP (red ), and surrounding native axons (neuroWlament, green). Arrows indicate examples of ensheathed axons, a higher magniWcation of which is shown in (G). Human OPCs enwrap native axons and reorganize the paranodal region to permit nodal formation. Caspr protein, an axonal paranodal marker, is expressed on unmyelinated axons between myelinated segments of axon, without invading the nodal region. (H-K) Optical sections through engrafted shiverer corpus callosum, showing donorderived MBP (green) and native axonal Caspr protein in red, indicating that donor OPCs develop not only myelin production and architecture, but permit nodes of Ranvier to form (anti-Caspr antibodies generously provided by Dr. M. Rasband). (L) Electron microscopy conWrmed that donorderived oligodendrocytes developed compact myelin, I that myelin produced by engrafted fetal human OPCs wrapped native axons to form compact sheathes with major dense lines (inset). Scale bars: A, 200mm; B-F, 20mm; H-K, 5 mm; L, 1 mm.

II. GLIAL CELL DEVELOPMENT

276

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

analogously acquired and maintained cells implanted into the same models at the same time. As a result, it was unclear if fetal-derived OPCs diVered from their counterparts derived from the adult human brain, with respect to either their migration competence, myelinogenic capacity, or the time courses thereof. To assess the relative advantages and disadvantages as therapeutic vectors of these two stage-deWned OPC phenotypes, newborn shiverer mice were implanted with either fetal or adult-derived OPCs, each isolated via A2B5-directed immunomagnetic sorting (IMS). The implanted neonatal mice were allowed to survive for 1 to 3 months, and their brains then sectioned and stained for MBP, GFAP, and anti-human nuclear antigen. By this means, it was determined that fetal and adultderived human OPCs diVered substantially in their respective time courses and eYcacy of myelinogenesis upon xenograft. Adult OPCs myelinated shiverer brain more rapidly than their fetal counterparts, achieving widespread and dense MBP expression by 4 weeks after xenograft. In contrast, substantial MBP expression by fetal OPCs was generally not observed until 12 weeks post-implant (Windrem et al., unpublished data). Besides myelinating more quickly than fetal OPCs, adult OPCs were found to give rise to myelinogenic oligodendrocytes in much higher relative proportions, and with much less astrocytic co-generation, than did fetal-derived OPCs. When assessed at the midline of the recipient corpus callosum, just over 10% of fetal hNA-deWned OPCs expressed MBP at 12 weeks, while virtually none had done so at 4 weeks. In contrast, almost 40% of adult OPCs expressed MBP by 4 weeks after xenograft into matched recipients. Thus, engrafted adult OPCs were at least four times more likely to mature as oligodendrocytes and develop myelin than their fetal counterparts. As another cardinal diVerence between fetal and adult OPCs, adult OPCs largely remained restricted to the host white matter, within which they generated almost entirely MBPþ oligodendrocytes. In contrast, fetal OPCs migrated into both gray and white matter, generating both astrocytes and oligodendrocytes in a contextdependent manner. Thus, both fetal and adult-derived OPCs were competent to remyelinate murine axons, but each had distinct advantages and disadvantages as potential vectors for cell therapy: Whereas fetal OPCs were highly migratory, they myelinated slowly and ineYciently. In contrast, adult-derived OPCs migrated over lesser distances, but they myelinated more rapidly and in higher proportions than their fetal counterparts. Together, these studies argued that while both fetal and adult human OPCs might provide eVective cellular substrates for remyelination, the choice of cellular source must be dictated not only by the availability of donor material, but also by the speciWc biology of the disease target.

A Caveat: Some Implanted Progenitors May Remain Undifferentiated A corollary of the multipotential nature of white matter progenitor cells is that when transplanted as nominally oligodendrocytic precursors, these cells might encounter local signals that instruct their maturation along alternative lineages. As a result, we need to be concerned about the possibility of their diVerentiation into undesired or functionally heterotopic phenotypes. This possibility is of further concern given the persistence of many implanted progenitors as undiVerentiated cells; these may remain able to respond to signals in the host tissue environment, not only at the time of implantation, but also long thereafter. As such, these cells might comprise a reservoir of implanted precursors, from which desired phenotypes might be later recruited upon injury or insult. On the other hand, they might just as well constitute potential sources of undesired cell types that might be ectopically generated and recruited in the tissue environment of an acutely injured focus. Such local production of undesired phenotypes might introduce not only ineYciency to transplant-based treatment strategies, but also frank danger. For instance, the production of neurons in a white matter lesion could generate an epileptogenic focus, just as the production of astrocytes in a more typically oligodendrocytic region might disrupt local ionic gradients and hence axonal transmission. These and many other untoward processes of heterotopic phenotypic maturation could more than oVset whatever beneWts might be gleaned from a therapeutic cell implant. As a result, it may prove advisable to initiate the phenotypic diVerentiation of these cells in vitro, prior to implantation, so as to limit

II. GLIAL CELL DEVELOPMENT

EXPERIMENTAL IMPLANTATION OF NON-CNS PROGENITOR CELL TYPES

the range of phenotypic choices available to the isolated progenitors to those appropriate for the intended region and disease target. Time will tell whether the possibility of heterotopic misdiVerentiation will mandate such in vitro priming steps.

EXPERIMENTAL IMPLANTATION OF NON-CNS PROGENITOR CELL TYPES A wide range of other potentially myelinogenic cell types have also been implanted into experimental models of de- and dysmyelination, with varying degrees of success.

Schwann Cells Schwann cells, the myelinating cells of the peripheral nervous system, have been considered as an attractive alternative to oligodendrocyte precursors for experimental transplantation. Schwann cells from several sources, including humans (Kohama et al., 2001), have been implanted in dysmyelinated shiverer mice (Baron-Van Evercooren et al., 1992), MD rats, and shaking pups (Duncan and HoVman, 1997). They have also been transplanted into lysolecithin (Baron-Van Evercooren et al., 1993; Duncan et al., 1981) and EB-X (Blakemore and Crang, 1985) demyelinated lesions in the brain and spinal cord. In all these systems, they have demonstrated varying degrees of myelination (Franklin and Barnett, 1997) with the myelin produced by these cells being of the PNS-variety as speciWed by the expression of P0. In some cases, functional reconstitution of saltatory conduction has also been shown (Felts and Smith, 1992; Honmou et al., 1996; Kohama et al., 2001). In addition, Schwann cells have been reported to improve axonal regeneration, which might be of importance in MS where axonal loss is a major part of the lesion pathology. Considering the relative ease of expanding human Schwann cells in culture (Rutkowski et al., 1995), it has been suggested that they might be appropriate cellular vectors for autologous transplants. Indeed, they have the added advantage of producing non-CNS myelin, which may be refractory to the immunological destruction in diseases like MS. However, like central oligodendrocyte progenitors, the migratory capacity of these cells is unclear. Some studies indicate that Schwann cells migrate satisfactorily over large distances to speciWc target sites (Franklin and Barnett, 1997), while others indicate that they are unable to migrate through normal white matter (Iwashita et al., 2000). In addition, Schwann cells seem to have a complex relationship with central astrocytes. After transplantation, Schwann cells are mainly found in areas devoid of astrocytes (Baron-van Evercooren et al., 1992; Blakemore and Patterson, 1975), and, moreover, they are excluded as astrocyte numbers increase with time (Shields et al., 2000).

Olfactory Ensheathing Cells (OEC) In nature, OEC’s ensheath small diameter axons of the peripheral olfactory epithelium neurons that project through the olfactory nerve into the olfactory bulb of the CNS. Unlike Schwann cells, these cells do not normally produce myelin. However, OECs from both animal (Franklin et al., 1996; Imaizumi et al., 1998) and human sources (Barnett et al., 2000; Kato et al., 2000) show remyelination with a peripheral pattern of myelin expression upon transplantation to demyelinated spinal cords. In some studies, a functional restoration of conduction has also been demonstrated (Imaizumi et al., 2000). OECs may have an advantage over Schwann cells, as they co-exist naturally with astrocytes in the olfactory bulb (Lakatos et al., 2000). Perhaps as a result, their interaction with astrocytes in not restrictive (Franklin and Barnett, 2000) in fact, they have been reported to support axonal regeneration through the astrocytic environment of a transected spinal cord (Ramon-Cueto et al., 1998). Nonetheless, their restoration of central axonal conduction remains inconclusive, as is the long-term fate of their remyelinated units. Whether these cells are capable of the contact-dependent and humoral support of neuronal function

II. GLIAL CELL DEVELOPMENT

277

278

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

normally exercised by central oligodendrocytes, or conversely, whether they are in turn supported by the axons with which they interact (Fernandez et al., 2000; Vartanian et al., 1997), similarly remain unknown.

Embryonic Stem Cells Myelination by in vitro conditioned mouse embryonic stem cells has been reported in both hypomyelinated MD rat E-17 fetuses and shiverer newborns, as well as in adult lysolecithin demyelinated lesions in adult rats (Brustle et al., 1999; Liu et al., 2000). More recent reports describe transplanted human ES cells sequentially cultured to induce neural stem cells capable of generating oligodendrocytes in a region-speciWc manner (ReubinoV et al., 2001). Though ES cells might represent a readily cultivable source of OPCs, the use of these cells is still limited by our inability to fully instruct all cells in the undiVerentiated population to the desired phenotype. Of greater concern is the persistent uncommitted progenitors within the implanted population, which may retain the latent capacity for undiVerentiated expansion and possibly tumorigenicity.

Mesenchymal and Marrow-Derived Stem Cells In addition to ES cells, mesenchymal and marrow-derived stem cells have been in focus as a source of neurally speciWed cells. Some controversial studies indicate that these cells may be capable of trans- or ectopic diVerentiation to neuroectodermal lineage (Mezey et al., 2000; Sanchez-Ramos et al., 2000). Of particular concern has been the lack of clear clonal evidence of neural speciWcation as well as recent reports indicting that cell fusion may explain some of observations of trans-diVerentiation (Terada et al., 2002; Ying et al., 2002). Nonetheless, a recent study, in which mouse bone marrow stromal cells were grafted into EB-X demyelinated spinal cord lesions, reported not only donor-cell derived histological remyelination, but also an improvement in conduction velocity (Akiyama et al., 2002). This work remains to be replicated by other groups. Should this study prove veriWable, its approach may open new avenues of stromal cell-based remyelination therapy.

DISEASE TARGETS FOR PROGENITOR-BASED THERAPEUTIC MYELINATION Congenital Dysmyelination Congenital diseases of myelination, such as periventricular leukomalacia (PVL), which may serve as an anatomic form fruste for the later development of cerebral palsy (Grow and Barks, 2002; Rezaie and Dean, 2002; Volpe, 2001) and the hereditary leukodystrophies and storage diseases, such as Krabbe’s and Tay Sachs disease, are leading causes of infant morbidity and mortality (reviewed by SchiVmann and BoespXug-Tanguy, 2001; Berger et al., 2001). As such, these may constitute feasible and attractive targets for therapeutic remyelination (Tate et al., 2001). Periventricular leukomalacia PVL describes a lesion of the periventricular white matter, associated with a failure in early myelination of the cerebral hemispheres. PVL appears to be a pathological concomitant to perinatal hypoxic-ischemic insult and may result from germinal matrix hemorrhage, sustained hypoxia, and excitotoxic injury, and most likely from combinations of these insults. PVL predicts the development of cerebral palsy in most cases (Volpe, 2001). Experimental models of hypoxic-ischemia in neonatal rats (Back et al., 2002; Levison et al., 2001) as well as studies of pediatric autopsies (Back et al., 2001) have suggested that the late oligodendrocyte progenitors of the forebrain SVZ comprise the predominant cell population lost in perinatal ischemic injury. This is in accord with our understanding of the natural history of oligodendrogliogenesis in humans (Grever et al., 1999; Rakic and Zecevic, 2003; Zhang et al., 2000), the developmental window for which corresponds to the period of ischemic vulnerability of the periventricular white matter.

II. GLIAL CELL DEVELOPMENT

DISEASE TARGETS FOR PROGENITOR-BASED THERAPEUTIC MYELINATION

Congenital leukodystrophies include an ever-expanding group of inherited diseases of myelin synthesis and metabolism. Although a diverse group, these may roughly be divided into lysosomal storage diseases, such as Krabbe’s globoid cell leukodystrophy (Wenger et al., 2000) and Tay Sachs diseases (Gravel et al., 1991); disorders of myelin synthesis, such as Pelizaeus-Merbacher disease (PMD) (Koeppen and Robitaille, 2002); and metabolic deWciencies leading to toxic demyelination, such as Canavan’s disease (Matalon and Michals-Matalon, 2000). Each of these disease categories is attended by extensive white matter involvement and clinical leukoencephalopathy, typically leading to severe neurological disability and death. As a group, the clinical leukodystrophies represent especially attractive targets for progenitor cell-based therapy, since the restoration of healthy oligodendrocytes in early perinatal development may be suYcient to permit myelination and hence to slow or prevent the development of the disease phenotype. In addition, eVective murine models of these diseases are available (Werner et al., 1998). Inherited diseases of the PLP and MBP genes are modeled by twitcher (Mikoshiba et al., 1985; Yoshimura et al., 1989) and shiverer mice (Roach et al., 1985), respectively. In addition, mutations of hexoseaminidase-B, modeling SandhoV’s and Tay-Sachs diseases (Kolter and SandhoV, 1998), and aspartoacylase, mimicking Canavan’s disease (Matalon et al., 2000), have been similarly employed. The availability of such genetically precise models of the childhood leukodystrophies is already greatly accelerating the process of developing experimental treatment strategies for these disorders.

Acquired Demyelination In adults, the diseases of acquired demyelination include later-onset leukodystrophies, such as metachromatic leukodystrophy and adrenoleukodystrophy, as well as vascular, inXammatory, and nutritional demyelinating syndromes (Baumann and Turpin, 2000; Berger et al., 2001; Desmond, 2002; Dichgans, 2002). The vascular demyelinations include hypertensive and diabetic leukoencephalopathies, which may both be due to chronic oligodendrocytic ischemic hypoxia (Dewar et al., 1999; Leys et al., 1999). Subcortical strokes, particularly those within the distributions of the forebrain lenticulostriate and thalamogeniculate arterial territories, are also prominent causes of vascular demyelination (Dichgans, 2002). The inXammatory demyelinations include multiple sclerosis, transverse myelitis (Kerr and Ayetey, 2002), optic neuritis (Cree et al., 2002; Eggenberger, 2001), and less commonly Schilder’s leukoencephalitis (Kotil et al., 2002), as well as postvaccinial (An et al., 2002; Konstantinou et al., 2001) and postinfectious leukoencephalitis (KleinschmidtDeMasters and Gilden, 2001; Rust, 2000). All of these syndromes of acquired demyelination are potential targets of therapeutic remyelination. Yet most attempts at cell-based remyelination in experimental animals have been made using acute chemical demyelinating insults, such as lysolecithin. In contrast to the availability of eVective animal models of congenital dysmyelination, the study of acquired demyelination has suVered from its lack of biologically appropriate, clinically reXective animal models. As a result, few adequate studies of cell-based remyelination of acquired, adult demyelinating lesions have been reported using any cellular vector. Those studies that have reported oligodendrocytic maturation and myelination by implanted oligodendrocyte progenitors have typically failed to demonstrate substantial axonal ensheathment, though this has likely reXected the loss of competent axons in these models, rather than any insuYciency on the part of the implanted progenitor cells. Indeed, the etiological complexity and manifold sequelae of demyelination in the adult brain argues against easy therapeutic intervention. As such, until improved models of acquired demyelinating disease are available, progress in cellbased therapy of adult demyelinating diseases will be necessarily slow. In contrast, the arguably simpler etiologies of congenital dysmyelination, their frequent lack of association with underlying systemic disease, and the persistent structural plasticity of the perinatal brain, together with the many eVective animal models for congenital dysmyelination, collaborate to make these diseases attractive targets for near-term intervention, both experimentally and clinically. Indeed, we may reasonably except the congenital leukodystrophies to be especially promising targets for cell-based therapeutic remyelination.

II. GLIAL CELL DEVELOPMENT

279

280

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

References

au9

Ader, M., Schachner, M., and Bartsch, U. (2001). Transplantation of neural precursor cells into the dysmyelinated CNS of mutant mice deWcient in the myelin-associated glycoprotein and Fyn tyrosine kinase. Eur J Neurosci 14, 561–566. Adlkofer, K., and C., L. (2000). Role of neuregulins in glial cell development. Glia 29, 104–111. Akiyama, Y., Honmou, O., Kato, T., Uede, T., Hashi, K., and Kocsis, J. D. (2001). Transplantation of clonal neural precursor cells derived from adult human brain establishes functional peripheral myelin in the rat spinal cord. Exp Neurol 167, 27–39. Akiyama, Y., Radtke, C., and Kocsis, J. D. (2002). Remyelination of the rat spinal cord by transplantation of identiWed bone marrow stromal cells. J Neurosci 22, 6623–6630. Aloisi, F., Giampaolo, A., Russo, G., Peschle, C., and Levi, G. (1992). Developmental appearance, antigenic proWle, and proliferation of glial cells of the human embryonic spinal cord: An immunocytochemical study using dissociated cultured cells. Glia 5, 171–181. Altman, J., and Das, G. D. (1965). Autoradiographic and histological evidence of postnatal hippocampal neurogenesis in rats. J. Comparative Neurology 124, 319–335. Altman, J., and Das, G. D. (1966). Autoradiographic and histological studies of postnatal neurogenesis. I. A longitudinal investigation of the kinetics, migration and transformation of cells incorporating tritiated thymidine in neonate rats, with special reference to postnatal neurogenesis in some brain regions. J. Comp. Neurol. 127, 337–390. An, S. F., Groves, M., Martinian, L., Kuo, L. T., and Scaravilli, F. (2002). Detection of infectious agents in brain of patients with acute hemorrhagic leukoencephalitis. J Neurovirol. 8, 439–446. Archer, D., Cuddon, P., Lipsitz, D., and Duncan, I. (1997). Myelination of the canine central nervous sytem by glial cell transplantation: A model for repair of human myelin disease. Nature Medicine 3, 54–59. Archer, D. R., Leven, S., and Duncan, I. D. (1994). Myelination by cryopreserved xenografts and allografts in the myelin-deWcient rat. Exp Neurol 125, 268–277. Armstrong, R. C., Dorn, H. H., Kufta, C. V., Friedman, E., and Dubois-Dalcq, M. E. (1992). Pre-oligodendrocytes from adult human CNS. J Neurosci 12, 1538–1547. Avellana-Adalid, V., Nait-Oumesmar, B., Lachapelle, F., and Baron-Van Evercooren, A. (1996). Expansion of rat oligodendrocyte progenitors into proliferative ‘‘oligospheres’’ that retain diVerentiation potential. J Neurosci Res 45, 558–570. Baas, D., Bourbeau, D., Sarlieve, L. L., Ittel, M. E., Dussault, J. H., and Puymirat, J. (1997). Oligodendrocyte maturation and progenitor cell proliferation are independently regulated by thyroid hormone. Glia 19, 324–332. Bach, M. A., Shen-Orr, Z., Lowe, W. L., Jr., Roberts, C. T., Jr., and LeRoith, D. (1991). Insulin-like growth factor I mRNA levels are developmentally regulated in speciWc regions of the rat brain. Brain Res Mol Brain Res 10, 43–8. Back, S. A., Han, B. H., Luo, N. L., Chricton, C. A., Xanthoudakis, S., Tam, J., Arvin, K. L., and Holtzman, D. M. (2002). Selective vulnerability of late oligodendrocyte progenitors to hypoxia-ischemia. J Neuroscience 22, 455–463. Back, S. A., Luo, N. L., Borenstein, N. S., Levine, J. M., Volpe, J. J., and Kinney, H. C. (2001). Late oligodendrocyte progenitors coincide with the developmental window of vulnerability for human perinatal white matter injury. J Neurosci 21, 1302–1312. Bansal, R., Kumar, M., Murray, K., Morrison, R. S., and PfeiVer, S. E. (1996). Regulation of FGF receptors in the oligodendrocyte lineage. Mol Cell Neurosci 7, 263–275. Bansal, R., and PfeiVer, S. E. (1994). Regulation of gene expression in mature oligodendrocytes by the specialized myelin-like membrane environment: Antibody perturbation in culture with the monoclonal antibody R-mAb. Glia 12, 173–179. Bansal, R., and PfeiVer, S. E. (1997). FGF-2 converts mature oligodendrocytes to a novel phenotype. J Neurosci Res 50, 215–228. Barami, K., Iversen, K., Furneaux, H., and Goldman, S. A. (1995). Hu protein as an early marker of neuronal phenotypic diVerentiation by subependymal zone cells of the adult songbird forebrain. J Neurobiol 28, 82–101. Barami, K., Zhao, J., Diaz, F. G., and Lyman, W. D. (2001). Comparison of neural precursor cell fate in second trimester human brain and spinal cord. Neurol Res 23, 260–266. Barnett, S. C., Alexander, C. L., Iwashita, Y., Gilson, J. M., Crowther, J., Clark, L., Dunn, L. T., Papanastassiou, V., Kennedy, P. G., and Franklin, R. J. (2000). IdentiWcation of a human olfactory ensheathing cell that can eVect transplant-mediated remyelination of demyelinated CNS axons. Brain 123, 1581–1588. Baron-Van Evercooren, A., Duhamel-Clerin, E., Boutry, J. M., Hauw, J. J., and Gumpel, M. (1993). Pathways of migration of transplanted Schwann cells in the demyelinated mouse spinal cord. J Neurosci Res 35, 428–438. Baron-Van Evercooren, A., Gansmuller, A., Duhamel, E., Pascal, F., and Gumpel, M. (1992). Repair of a myelin lesion by Schwann cells transplanted in the adult mouse spinal cord. J Neuroimmunol 40, 235–242. Barres, B. A., Hart, I. K., Coles, H. S., Burne, J. F., Voyvodic, J. T., Richardson, W. D., and RaV, M. C. (1992). Cell death and control of cell survival in the oligodendrocyte lineage. Cell 70, 31–46. Barres, B. A., Lazar, M. A., and RaV, M. C. (1994a). A novel role for thyroid hormone, glucocorticoids and retinoic acid in timing oligodendrocyte development. Development 120, 1097–1108. Barres, B. A., RaV, M. C., Gaese, F., Bartke, I., Dechant, G., and Barde, Y. A. (1994b). A crucial role for neurotrophin-3 in oligodendrocyte development. Nature 367, 371–375.

II. GLIAL CELL DEVELOPMENT

DISEASE TARGETS FOR PROGENITOR-BASED THERAPEUTIC MYELINATION

au11

Baumann, N., and Turpin, J. C. (2000). Adult-onset leukodystrophies. J Neurology 247, 751–759. Berger, J., Moser, H. W., and Forss-Petter, S. (2001). Leukodystrophies: Recent developments in genetics, molecular biology, pathogenesis and treatment. Current Opinion in Neurology 14, 305–312. Besnard, F., Perraud, F., Sensenbrenner, M., and Labourdette, G. (1989). EVects of acidic and basic Wbroblast growth factors on proliferation and maturation of cultured rat oligodendrocytes. Int J Dev Neurosci 7, 401–409. Blakemore, W. F. (1992). Transplanted cultured type-1 astrocytes can be used to reconstitute the glia limitans of the CNS: The structure which prevents Schwann cells from myelinating CNS axons. Neuropathol Appl Neurobiol 18, 460–466. Blakemore, W. F., and Crang, A. J. (1985). The use of cultured autologous Schwann cells to remyelinate areas of persistent demyelination in the central nervous system. J Neurol Sci 70, 207–223. Blakemore, W. F., and Patterson, R. C. (1975). Observations on the interactions of Schwann cells and astrocytes following X-irradiation of neonatal rat spinal cord. J Neurocytol 4, 573–85. Brustle, O., Choudhary, K., Karram, K., Huttner, A., Murray, K., Dubois-Dalcq, M., and McKay, R. D. (1998). Chimeric brains generated by intraventricular transplantation of fetal human brain cells into embryonic rats. Nat Biotechnol 16, 1040–4. Brustle, O., Jones, K. N., Learish, R. D., Karram, K., Choudhary, K., Wiestler, O. D., Duncan, I. D., and McKay, R. D. (1999). Embryonic stem cell-derived glial precursors: A source of myelinating transplants. Science 285, 754–756. Buchet, D., Buc-Caron, M. H., Sabate, O., Lachapelle, F., and Mallet, J. (2002). Long-term fate of human telencephalic progenitor cells grafted into the adult mouse brain: EVects of previous ampliWcation in vitro. J Neurosci Res 68, 276–283. Buonanno, A., and Fischbach, G. D. (2001). Neuregulin and ErbB receptor signaling pathways in the nervous system. Curr Opin Neurobiol. 11, 287–296. Cannella, B., Pitt, D., Marchionni, M., and Raine, C. S. (1999). Neuregulin and erbB receptor expression in normal and diseased human white matter. J Neuroimmunol 100, 233–242. Canoll, P. D., Kraemer, R., Teng, K. K., Marchionni, M. A., and Salzer, J. L. (1999). GGF/neuregulin induces a phenotypic reversion of oligodendrocytes. Mol Cell Neurosci 13, 79–94. Canoll, P. D., Musacchio, J. M., Hardy, R., Reynolds, R., Marchionni, M. A., and Salzer, J. L. (1996). GGF/ neuregulin is a neuronal signal that promotes the proliferation and survival and inhibits the diVerentiation of oligodendrocyte progenitors. Neuron 17, 229–243. Carpenter, M. K., Cui, X., Hu, Z. Y., Jackson, J., Sherman, S., Seiger, A., and Wahlberg, L. U. (1999). In vitro expansion of a multipotent population of human neural progenitor cells. Exp Neurol 158, 265–278. Carroll, W. M., Jennings, A. R., and Ironside, L. J. (1998). IdentiWcation of the adult resting progenitor cell by autoradiographic tracking of oligodendrocyte precursors in experimental CNS demyelination. Brain 121, 293–302. Carson, M., Behringer, R., Brinster, R., and McMorris, F. (1993). Insulin-like growth factor I increases brain growth and central nervous system myelination in transgenic mice. Neuron 10, 729–740. ChalWe, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green Xuorescent protein as a marker for gene expression. Science 263, 802–805. Chang, A., Tourtellotte, W. W., Rudick, R., and Trapp, B. D. (2002). Premyelinating oligodendrocytes in chronic lesions of multiple sclerosis. N Engl J Med 346, 165–173. Chari, D. M., and Blakemore, W. F. (2002). EYcient recolonisation of progenitor-depleted areas of the CNS by adult oligodendrocyte progenitor cells. Glia 37, 307–13. Cree, B. A., Goodin, D. S., and Hauser, S. L. (2002). Neuromyelitis optica. Seminars in Neurology 22, 105–22. Davis, A. A., and Temple, S. (1994). A self-renewing multipotential stem cell in embryonic rat cerebral cortex. Nature 372, 263–6. Dawson, M. R., Levine, J. M., and Reynolds, R. (2000). NG2-expressing cells in the central nervous system: Are they oligodendroglial progenitors? J Neurosci Res 61, 471–479. Deadwyler, G. D., Pouly, S., Antel, J. P., and Devries, G. H. (2000). Neuregulins and erbB receptor expression in adult human oligodendrocytes. Glia 32, 304–312. Desmond, D. W. (2002). Cognition and white matter lesions. Cerebrovascular Disease 13, 53–57. Dewar, D., Yam, P., and McCulloch, J. (1999). Drug development for stroke: Importance of protecting cerebral white matter. European Journal of Pharmacology 375, 41–50. Dichgans, M. (2002). Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy: Phenotypic and mutational spectrum. J Neurological Sciences 203–204, 77–80. Duncan, I. D., Aguayo, A. J., Bunge, R. P., and Wood, P. M. (1981). Transplantation of rat Schwann cells grown in tissue culture into the mouse spinal cord. J Neurol Sci 49, 241–252. Duncan, I. D., and HoVman, R. L. (1997). Schwann cell invasion of the central nervous system of the myelin mutants. J Anat 190, 35–49. Eggenberger, E. R. (2001). InXammatory optic neuropathies. Ophthalmol Clin North Am. 14, 73–82. Eisenbarth, G. S., Walsh, F. S., and Nirenberg, M. (1979b). Monoclonal antibody to a plasma membrane antigen of neurons. Proc Natl Acad Sci USA 76, 4913–4917. Ellison, J. A., and de Vellis, J. (1994). Platelet-derived growth factor receptor is expressed by cells in the early oligodendrocyte lineage. J Neurosci Res. 37, 116–128. Engel, U., and Wolswijk, G. (1996). Oligodendrocyte-type-2 astrocyte (O-2A) progenitor cells derived from adult rat spinal cord: In vitro characteristics and response to PDGF, bFGF and NT-3. Glia 16, 16–26.

II. GLIAL CELL DEVELOPMENT

281

282

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

Englund, U., Bjorklund, A., and Wictorin, K. (2002). Migration patterns and phenotypic diVerentiation of longterm expanded human neural progenitor cells after transplantation into the adult rat brain. Brain Res Dev Brain Res 134, 123–141. Eriksson, P., PerWlieva, E., Bjork-Eriksson, Alborn, A., Nordberg, C., Peterson, D., and Gage, F. (1998). Neurogenesis in the adult human hippocampus. Nature Medicine 4, 1313–1317. Felts, P. A., and Smith, K. J. (1992). Conduction properties of central nerve Wbers remyelinated by Schwann cells. Brain Res 574, 178–192. Fernandez, P. A., Tang, D. G., Cheng, L., Prochiantz, A., Mudge, A. W., and RaV, M. C. (2000). Evidence that axon-derived neuregulin promotes oligodendrocyte survival in the developing rat optic nerve. Neuron 28, 81–90. Ffrench-Constant, C., and RaV, M. C. (1986). Proliferating bipotential glial progenitor cells in adult rat optic nerve. Nature 319, 499–502. Franklin, R. J., and Barnett, S. C. (1997). Do olfactory glia have advantages over Schwann cells for CNS repair? J Neurosci Res 50, 665–672. Franklin, R. J., and Barnett, S. C. (2000). Olfactory ensheathing cells and CNS regeneration: The sweet smell of success? Neuron 28, 15–18. Franklin, R. J., and Blakemore, W. F. (1997). To what extent is oligodendrocyte progenitor migration a limiting factor in the remyelination of multiple sclerosis lesions? Mult Scler 3, 84–87. Franklin, R. J., Crang, A. J., and Blakemore, W. F. (1991). Transplanted type-1 astrocytes facilitate repair of demyelinating lesions by host oligodendrocytes in adult rat spinal cord. J Neurocytol 20, 420–430. Franklin, R. J., Gilson, J. M., Franceschini, I. A., and Barnett, S. C. (1996). Schwann cell-like myelination following transplantation of an olfactory bulb-ensheathing cell line into areas of demyelination in the adult CNS. Glia 17, 217–224. Fricker, R. A., Carpenter, M. K., Winkler, C., Greco, C., Gates, M. A., and Bjorklund, A. (1999). Site-speciWc migration and neuronal diVerentiation of human neural progenitor cells after transplantation in the adult rat brain. J Neurosci 19, 5990–6005. Fruttiger, M., Karlsson, L., Hall, A. C., Abramsson, A., Calver, A. R., Bostrom, H., Willetts, K., Bertold, C. H., Heath, J. K., Betsholtz, C., and Richardson, W. D. (1999). Defective oligodendrocyte development and severe hypomyelination in PDGF-A knockout mice. Development 126, 457–467. Gage, F. H., Kempermann, G., Palmer, T. D., Peterson, D. A., and Ray, J. (1998). Multipotent progenitor cells in the adult dentate gyrus. J Neurobiol 36, 249–266. Gansmuller, A., Lachapelle, F., Baron-Van Evercooren, A., Hauw, J. J., Baumann, N., and Gumpel, M. (1986). Transplantations of newborn CNS fragments into the brain of shiverer mutant mice: Extensive myelination by transplanted oligodendrocytes. II. Electron microscopic study. Dev Neurosci 8, 197–207. Gensert, J. M., and Goldman, J. E. (1996). In vivo characterization of endogenous proliferating cells in adult rat subcortical white matter. Glia 17, 39–51. Gensert, J. M., and Goldman, J. E. (1997). Endogenous progenitors remyelinate demyelinated axons in the adult CNS. Neuron 19, 197–203. Gensert, J. M., and Goldman, J. E. (2001). Heterogeneity of cycling glial progenitors in the adult mammalian cortex and white matter. J Neurobiol 48, 75–86. Gloster, A., Wu, W., Speelman, A., Weiss, S., Causing, C., Pozniak, C., Reynolds, B., Chang, E., Toma, J. G., and Miller, F. D. (1994). The T alpha 1 alpha-tubulin promoter speciWes gene expression as a function of neuronal growth and regeneration in transgenic mice. J Neurosci 14, 7319–7330. Gogate, N., Verma, L., Zhou, J. M., Milward, E., Rusten, R., O’Connor, M., Kufta, C., Kim, J., Hudson, L., and Dubois-Dalcq, M. (1994). Plasticity in the adult human oligodendrocyte lineage. J Neurosci 14, 4571–87. Goldman, S. A. (1998). Adult neurogenesis: From canaries to the clinic. J Neurobiol 36, 267–286. Goldman, S. A., Lemmon, V., and Chin, S. S. (1993). Migration of newly generated neurons upon ependymally derived radial guide cells in explant cultures of the adult songbird forebrain. Glia 8, 150–160. Goldman, S. A., and Nottebohm, F. (1983). Neuronal production, migration, and diVerentiation in a vocal control nucleus of the adult female canary brain. Proceedings of the National Academy of Sciences 80, 2390–2394. Gould, E., Tanapat, P., McEwen, B., Flugge, G., and Fuchs, E. (1998). Proliferation of granule cell precursors in the dentate gyrus of adult monkeys is diminished by stress. Proc. Natl. Acad. Sci. 95, 3168–3171. Grandpre, T., and Strittmatter, S. M. (2001). Nogo: A molecular determinant of axonal growth and regeneration. Neuroscientist 7, 377–386. Gravel, M., Di Polo, A., Valera, P. B., and Braun, P. E. (1998). Four-kilobase sequence of the mouse CNP gene directs spatial and temporal expression of lacZ in transgenic mice. J Neurosci Res 53, 393–404. Gravel, R. A., Triggs-Raine, B. L., and Mahuran, D. J. (1991). Biochemistry and genetics of Tay-Sachs disease. Can J Neurol Sci. 18, 419–423. Grever, W. E., Zhang, S., Ge, B., and Duncan, I. D. (1999). Fractionation and enrichment of oligodendrocytes from developing human brain. J Neurosci Res 57, 304–14. Grinspan, J. B., Stern, J. L., Franceschini, B., and Pleasure, D. (1993). Trophic eVects of basic Wbroblast growth factor (bFGF) on diVerentiated oligodendroglia: A mechanism for regeneration of the oligodendroglial lineage. J Neurosci Res 36, 672–80. Grow, J., and Barks, J. D. (2002). Pathogenesis of hypoxic-ischemic cerebral injury in the term infant: Current concepts. Clin Perinatol. 29, 585–602.

II. GLIAL CELL DEVELOPMENT

DISEASE TARGETS FOR PROGENITOR-BASED THERAPEUTIC MYELINATION

au12

Gumpel, M., Gout, O., Lubetzki, C., Gansmuller, A., and Baumann, N. (1989). Myelination and remyelination in the central nervous system by transplanted oligodendrocytes using the shiverer model. Discussion on the remyelinating cell population in adult mammals. Dev Neurosci 11, 132–139. Gumpel, M., Lachapelle, F., Gansmuller, A., Baulac, M., Baron van Evercooren, A., and Baumann, N. (1987). Transplantation of human embryonic oligodendrocytes into shiverer brain. Ann NY Acad Sci 495, 71–85. Hammang, J. P., Archer, D. R., and Duncan, I. D. (1997). Myelination following transplantation of EGFresponsive neural stem cells into a myelin-deWcient environment. Exp Neurol 147, 84–95. Hart, I. K., Richardson, W. D., Bolsover, S. R., and RaV, M. C. (1989a). PDGF and intracellular signaling in the timing of oligodendrocyte diVerentiation. J Cell Biol 109, 3411–3417. Hart, I. K., Richardson, W. D., Heldin, C. H., Westermark, B., and RaV, M. C. (1989b). PDGF receptors on cells of the oligodendrocyte-type-2 astrocyte (O-2A) cell lineage. Development 105, 595–603. Honmou, O., Felts, P. A., Waxman, S. G., and Kocsis, J. D. (1996). Restoration of normal conduction properties in demyelinated spinal cord axons in the adult rat by transplantation of exogenous Schwann cells. J Neurosci 16, 3199–3208. Horner, P. J., Power, A. E., Kempermann, G., Kuhn, H. G., Palmer, T. D., Winkler, J., Thal, L. J., and Gage, F. H. (2000). Proliferation and diVerentiation of progenitor cells throughout the intact adult rat spinal cord. J Neurosci 20, 2218–2228. Ibarrola, N., Mayer-Proschel, M., Rodriguez-Pena, A., and Noble, M. (1996). Evidence for the existence of at least two timing mechanisms that contribute to oligodendrocyte generation in vitro. Dev. Biol. 180, 1–21. Imaizumi, T., Lankford, K. L., and Kocsis, J. D. (2000). Transplantation of olfactory ensheathing cells or Schwann cells restores rapid and secure conduction across the transected spinal cord. Brain Res 854, 70–78. Imaizumi, T., Lankford, K. L., Waxman, S. G., Greer, C. A., and Kocsis, J. D. (1998). Transplanted olfactory ensheathing cells remyelinate and enhance axonal conduction in the demyelinated dorsal columns of the rat spinal cord. J Neurosci 18, 6176–6185. Iwashita, Y., Fawcett, J. W., Crang, A. J., Franklin, R. J., and Blakemore, W. F. (2000). Schwann Cells Transplanted into Normal and X-Irradiated Adult White Matter Do Not Migrate Extensively and Show Poor Long-Term Survival. Exp Neurol 164, 292–302. Kaplan, M. S. (1985). Formation and turnover of neurons in young and senescent animals: An electronmicroscopic and morphometric analysis. Ann NY Acad Sci 457, 173–192. Kaplan, M. S., and Hinds, J. W. (1977). Neurogenesis in the adult rat: Electron microscopic analysis of light radioautographs. Science 197, 1092–1094. Kato, T., Honmou, O., Uede, T., Hashi, K., and Kocsis, J. D. (2000). Transplantation of human olfactory ensheathing cells elicits remyelination of demyelinated rat spinal cord. Glia 30, 209–218. Keirstead, H. S., Levine, J. M., and Blakemore, W. F. (1998). Response of the oligodendrocyte progenitor cell population (deWned by NG2 labelling) to demyelination of the adult spinal cord. Glia 22, 161–170. Kerr, D. A., and Ayetey, H. (2002). Immunopathogenesis of acute transverse myelitis. Curr Opin Neurol. Current Opinion in Neurology 15, 339–347. Keyoung, H. M., Roy, N. S., Benraiss, A., Louissaint, A., Jr., Suzuki, A., Hashimoto, M., Rashbaum, W. K., Okano, H., and Goldman, S. A. (2001). High-yield selection and extraction of two promoter-deWned phenotypes of neural stem cells from the fetal human brain. Nat Biotechnol 19, 843–850. Kim, S. U., Sato, Y., Silberberg, D. H., Pleasure, D. E., and Rorke, L. B. (1983). Long-term culture of human oligodendrocytes. Isolation, growth and identiWcation. J Neurol Sci 62, 295–301. Kirschenbaum, B., Nedergaard, M., Preuss, A., Barami, K., Fraser, R. A., and Goldman, S. A. (1994). In vitro neuronal production and diVerentiation by precursor cells derived from the adult human forebrain. Cerebral Cortex 4, 576–589. Kleinschmidt-DeMasters, B. K., and Gilden, D. H. (2001). The expanding spectrum of herpesvirus infections of the nervous system. Brain Pathol. 11, 440–451. Koeppen, A. H., and Robitaille, Y. (2002). Pelizaeus-Merzbacher disease. J Neuropathol Exp Neurol. 61, 747–759. Kohama, I., Lankford, K. L., Preiningerova, J., White, F. A., Vollmer, T. L., and Kocsis, J. D. (2001). Transplantation of cryopreserved adult human Schwann cells enhances axonal conduction in demyelinated spinal cord. J Neurosci 21, 944–950. Kolter, T., and SandhoV, K. (1998). Glycosphingolipid degradation and animal models of GM2-gangliosidoses. J Inherit Metab Dis. 21, 548–563. Kondo, T., and RaV, M. (2000). Oligodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells. Science 289, 1754–1757. Konstantinou, D., Paschalis, C., Maraziotis, T., Dimopoulos, P., Bassaris, H., and Skoutelis, A. (2001). Two episodes of leukoencephalitis associated with recombinant hepatitis B vaccination in a single patient. Clin Infect Dis. 33, 1772–1773. Kotil, K., Kalayci, M., Koseoglu, T., and Tugrul, A. (2002). Myelinoclastic diVuse sclerosis (Schilder’s disease): Report of a case and review of the literature. Br J Neurosurg 16, 516–519. Kukekov, V. G., Laywell, E. D., Suslov, O., Davies, K., ScheZer, B., Thomas, L. B., O’Brien, T. F., Kusakabe, M., and Steindler, D. A. (1999). Multipotent stem/progenitor cells with similar properties arise from two neurogenic regions of adult human brain. Exp Neurol 156, 333–344. Lachapelle, F., Gumpel, M., Baulac, M., Jacque, C., Duc, P., and Baumann, N. (1983). Transplantation of CNS fragments into the brain of shiverer mutant mice: Extensive myelination by implanted oligodendrocytes. I. Immunohistochemical studies. Dev Neurosci 6, 325–334.

II. GLIAL CELL DEVELOPMENT

283

284

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

Lakatos, A., Franklin, R. J., and Barnett, S. C. (2000). Olfactory ensheathing cells and Schwann cells diVer in their in vitro interactions with astrocytes. Glia 32, 214–225. Learish, R. D., Brustle, O., Zhang, S. C., and Duncan, I. D. (1999). Intraventricular transplantation of oligodendrocyte progenitors into a fetal myelin mutant results in widespread formation of myelin. Ann Neurol 46, 716–722. Lee, S. C., Liu, W., Brosnan, C. F., and Dickson, D. W. (1992). Characterization of primary human fetal dissociated central nervous system cultures with an emphasis on microglia. Lab Invest 67, 465–76. Levine, J. M., and Reynolds, R. (1999). Activation and proliferation of endogenous oligodendrocyte precursor cells during ethidium bromide-induced demyelination. Exp Neurol 160, 333–347. Levine, J. M., Reynolds, R., and Fawcett, J. W. (2001). The oligodendrocyte precursor cell in health and disease. Trends Neurosci 24, 39–47. Levine, J. M., Stincone, F., and Lee, Y. S. (1993). Development and diVerentiation of glial precursor cells in the rat cerebellum. Glia 7, 307–321. Levison, S. W., and Goldman, J. E. (1993). Both oligodendrocytes and astrocytes develop from progenitors in the subventricular zone of postnatal rat forebrain. Neuron 10, 201–212. Levison, S. W., Rothstein, R. P., Romanko, M. J., Snyder, M. J., Meyers, R. L., and Vannucci, S. J. (2001). Hypoxia/ischemia depletes the rat perinatal subventricular zone of oligodendrocyte progenitors and neural stem cells. Developmental Neuroscience 23, 234–247. Levison, S. W., Young, G. M., and Goldman, J. E. (1999). Cycling cells in the adult rat neocortex preferentially generate oligodendroglia. J Neurosci Res 57, 435–446. Leys, D., Englund, E., Del Se, r. T., Inzitari, D., Fazekas, F., Bornstein, N., Erkinjuntti, T., Bowler, J. V., Pantoni, L., Parnetti, L., De Reuck, J., Ferro, J., and Bogousslavsky, J. (1999). White matter changes in stroke patients. Relationship with stroke subtype and outcome. European Journal of Neurology 42, 67–75. Liu, S., Qu, Y., Stewart, T. J., Howard, M. J., Chakrabortty, S., Holekamp, T. F., and McDonald, J. W. (2000). Embryonic stem cells diVerentiate into oligodendrocytes and myelinate in culture and after spinal cord transplantation. Proc Natl Acad Sci USA 97, 6126–6131. Lois, C., and Alvarez-Buylla, A. (1993). Proliferating subventricular zone cells in the adult mammalian forebrain can diVerentiate into neurons and glia. Proceedings of the National Academy of Sciences 90, 2074–2077. Luskin, M. B. (1993). Restricted proliferation and migration of postnatally generated neurons derived from the forebrain subventricular zone. Neuron 11, 173–189. Maeda, Y., Solanky, M., Menonna, J., Chapin, J., Li, W., and Dowling, P. (2001). Platelet-derived growth factoralpha receptor-positive oligodendroglia are frequent in multiple sclerosis lesions. Ann Neurol 49, 776–785. Marmur, R., Mabie, P. C., Gokhan, S., Song, Q., Kessler, J. A., and Mehler, M. F. (1998). Isolation and developmental characterization of cerebral cortical multipotent progenitors. Dev Biol 204, 577–591. Marusich, M., Furneaux, H., Henion, P., and Weston, J. (1994). Hu neuronal proteins are expressed in proliferating neurogenic cells. Journal of Neurobiology 25, 143–155. Mason, J. L., and Goldman, J. E. (2002). A2B5þ and O4þ Cycling progenitors in the adult forebrain white matter respond diVerentially to PDGF-AA, FGF-2, and IGF-1. Mol Cell Neurosci 20, 30–42. Matalon, R., Rady, P. L., Platt, K. A., Skinner, H. B., Quast, M. J., Campbell, G. A., Matalon, K., Ceci, J. D., Tyring, S. K., Nehls, M., Surendran, S., Wei, J., Ezell, E. L., and Szucs, S. (2000). Knock-out mouse for Canavan disease: A model for gene transfer to the central nervous system. J Gene Med. 2, 165–175. Matalon, R. M., and Michals-Matalon, K. (2000). Spongy degeneration of the brain, Canavan disease: Biochemical and molecular Wndings. Front Biosci. 5, D307-D311. McKinnon, R. D., Matsui, T., Dubois-Dalcq, M., and Aaronson, S. A. (1990). FGF modulates the PDGF-driven pathway of oligodendrocyte development. Neuron 5, 603–614. McMorris, F. A., and Dubois-Dalcq, M. (1988). Insulin-like growth factor I promotes cell proliferation and oligodendroglial commitment in rat glial progenitor cells developing in vitro. J Neurosci Res 21, 199–209. McTigue, D. M., Horner, P. J., Stokes, B. T., and Gage, F. H. (1998). Neurotrophin-3 and brain-derived neurotrophic factor induce oligodendrocyte proliferation and myelination of regenerating axons in the contused adult rat spinal cord. J Neurosci 18, 5354–5365. Mezey, E., Chandross, K. J., Harta, G., Maki, R. A., and McKercher, S. R. (2000). Turning blood into brain: Cells bearing neuronal antigens generated in vivo from bone marrow. Science 290, 1779–1782. Mikoshiba, K., Fujishiro, M., Kohsaka, S., and Okano, H. (1985). Takamatsu K, Tsukada Y. Disorders in myelination in the twitcher mutant: Immunohistochemical and biochemical studies. Neurochem Res. 10, 1129–1141. Miller, F. D., Naus, C. C., Durand, M., Bloom, F. E., and Milner, R. J. (1987). Isotypes of alpha-tubulin are diVerentially regulated during neuronal maturation. J Cell Biol 105, 3065–3073. Miller, F. D., TetzlaV, W., Bisby, M. A., Fawcett, J. W., and Milner, R. J. (1989). Rapid induction of the major embryonic alpha-tubulin mRNA, T alpha 1, during nerve regeneration in adult rats. J Neurosci 9, 1452–1463. Mitome, M., Low, H. P., van Den Pol, A., Nunnari, J. J., Wolf, M. K., Billings-Gagliardi, S., and Schwartz, W. J. (2001). Towards the reconstruction of central nervous system white matter using neural precursor cells. Brain 124, 2147–2161. Moore, G. R., Neumann, P. E., Suzuki, K., Lijtmaer, H. N., Traugott, U., and Raine, C. S. (1985). Balo’s concentric sclerosis: New observations on lesion development. Ann Neurol 17, 604–611. Nishiyama, A., Lin, X. H., Giese, N., Heldin, C. H., and Stallcup, W. B. (1996). Co-localization of NG2 proteoglycan and PDGF alpha-receptor on O2A progenitor cells in the developing rat brain. J Neurosci Res 43, 299–314.

II. GLIAL CELL DEVELOPMENT

DISEASE TARGETS FOR PROGENITOR-BASED THERAPEUTIC MYELINATION

Nishiyama, A., Yu, M., Drazba, J. A., and Tuohy, V. K. (1997). Normal and reactive NG2þ glial cells are distinct from resting and activated microglia. J Neurosci Res 48, 299–312. Noble, M. (1997). The oligodendrocyte-type 2 astrocyte lineage: In vitro and in vivo studies on development, tissue repair and neoplasia. In ‘‘Isolation, Characterization and Utilization of CNS Stem Cells’’ (Y. C. F. Gage, ed.), pp. 101–128. Springer-Verlag, Berlin. Noble, M., Murray, K., Stroobant, P., WaterWeld, M. D., and Riddle, P. (1988). Platelet-derived growth factor promotes division and motility and inhibits premature diVerentiation of the oligodendrocyte/type-2 astrocyte progenitor cell. Nature 333, 560–562. Noble, M., Wren, D., and Wolswijk, G. (1992). The O-2A(adult) progenitor cell: A glial stem cell of the adult central nervous system. Semin Cell Biol 3, 413–422. Nottebohm, F. (1985). Neuronal replacement in adulthood. Ann NY Acad Sci 457, 143–161. Nunes, M. C., Roy, N. S., Keyoung, H. M., Goodman, R., McKhann, G., Kang, J., Jiang, L., Nedergaard, M., and Goldman, S. (2003). IdentiWcation and isolation of multipotential neural progenitor cells from the subcortical white matter of the adult human brain. Nature Med. 9, 439–447. O’Neill, R. C., Minuk, J., Cox, M. E., Braun, P. E., and Gravel, M. (1997). CNP2 mRNA directs synthesis of both CNP1 and CNP2 polypeptides. J Neurosci Res 50, 248–257. Palmer, T. D., Ray, J., and Gage, F. H. (1995). FGF-2-responsive neuronal progenitors reside in proliferative and quiescent regions of the adult rodent brain. Mol Cell Neurosci 6, 474–486. Pincus, D. W., Keyoung, H. M., Harrison-Restelli, C., Goodman, R. R., Fraser, R. A., Edgar, M., Sakakibara, S., Okano, H., Nedergaard, M., and Goldman, S. A. (1998). Fibroblast growth factor-2/brain-derived neurotrophic factor-associated maturation of new neurons generated from adult human subependymal cells. Ann Neurol 43, 576–585. Pouly, S., Becher, B., Blain, M., and Antel, J. P. (1999). Expression of a homologue of rat NG2 on human microglia. Glia 27, 259–268. Prabhakar, S., D’Souza, S., Antel, J. P., McLaurin, J., Schipper, H. M., and Wang, E. (1995). Phenotypic and cell cycle properties of human oligodendrocytes in vitro. Brain Res 672, 159–169. Prineas, J. W., and Connell, F. (1979). Remyelination in multiple sclerosis. Ann Neurol 5, 22–31. Prineas, J. W., Kwon, E. E., Cho, E. S., and Sharer, L. R. (1984). Continual breakdown and regeneration of myelin in progressive multiple sclerosis plaques. Ann NY Acad Sci 436, 11–32. Prineas, J. W., Kwon, E. E., Goldenberg, P. Z., Ilyas, A. A., Quarles, R. H., Benjamins, J. A., and Sprinkle, T. J. (1989). Multiple sclerosis. Oligodendrocyte proliferation and diVerentiation in fresh lesions. Lab Invest 61, 489–503. Pringle, N. P., Mudhar, H. S., Collarini, E. J., and Richardson, W. D. (1992). PDGF receptors in the rat CNS: During late neurogenesis, PDGF alpha-receptor expression appears to be restricted to glial cells of the oligodendrocyte lineage. Development 115, 535–551. Qian, X., Davis, A. A., Goderie, S. K., and Temple, S. (1997). FGF2 concentration regulates the generation of neurons and glia from multipotent cortical stem cells. Neuron 18, 81–93. Raabe, T. D., Clive, D. R., Wen, D., and DeVries, G. H. (1997). Neonatal oligodendrocytes contain and secrete neuregulins in vitro. J Neurochem 69, 1859–1863. RaV, M. C., Abney, E. R., Cohen, J., Lindsay, R., and Noble, M. (1983a). Two types of astrocytes in cultures of developing rat white matter: DiVerences in morphology, surface gangliosides, and growth characteristics. J Neurosci 3, 1289–1300. RaV, M. C., Lillien, L. E., Richardson, W. D., Burne, J. F., and Noble, M. D. (1988). Platelet-derived growth factor from astrocytes drives the clock that times oligodendrocyte development in culture. Nature 333, 562–565. RaV, M. C., Miller, R. H., and Noble, M. (1983b). Glial cell lineages in the rat optic nerve. Cold Spring Harb Symp Quant Biol 48, 569–572. Rakic, S., and Zecevic, N. (2003). Early oligodendrocyte progenitor cells in the human fetal telencephalon. Glia 41, 117–127. Ramon-Cueto, A., Plant, G. W., Avila, J., and Bunge, M. B. (1998). Long-distance axonal regeneration in the transected adult rat spinal cord is promoted by olfactory ensheathing glia transplants. J Neurosci 18, 3803–3815. Redwine, J. M., and Armstrong, R. C. (1998). In vivo proliferation of oligodendrocyte progenitors expressing PDGFalphaR during early remyelination. J Neurobiol 37, 413–428. ReubinoV, B. E., Itsykson, P., Turetsky, T., Pera, M. F., Reinhartz, E., Itzik, A., and Ben-Hur, T. (2001). Neural progenitors from human embryonic stem cells. Nat Biotechnol 19, 1134–1140. Reynolds, R., and Hardy, R. (1997). Oligodendroglial progenitors labeled with the O4 antibody persist in the adult rat cerebral cortex in vivo. J Neurosci Res 47, 455–470. Rezaie, P., and Dean, A. (2002). Periventricular leukomalacia, inXammation and white matter lesions within the developing nervous system. Neuropathology 22, 106–132. Richards, L. J., Kilpatrick, T. J., and Bartlett, P. F. (1992). De novo generation of neuronal cells from the adult mouse brain. Proceedings of the National Academy of Sciences 89, 8591–8595. Roach, A., Takahashi, N., Pravtcheva, D., Ruddle, F., and Hood, L. (1985). Chromosomal mapping of mouse myelin basic protein gene and structure and transcription of the partially deleted gene in shiverer mutant mice. Cell 42, 149–155.

II. GLIAL CELL DEVELOPMENT

285

286

10. PROGENITOR CELLS OF THE ADULT HUMAN SUBCORTICAL WHITE MATTER

au13

Roy, N. S., Benraiss, A., Wang, S., Fraser, R. A., Goodman, R., Couldwell, W. T., Nedergaard, M., Kawaguchi, A., Okano, H., and Goldman, S. A. (2000a). Promoter-targeted selection and isolation of neural progenitor cells from the adult human ventricular zone. J Neurosci Res 59, 321–331. Roy, N. S., Wang, S., Harrison-Restelli, C., Benraiss, A., Fraser, R. A., Gravel, M., Braun, P. E., and Goldman, S. A. (1999). IdentiWcation, isolation, and promoter-deWned separation of mitotic oligodendrocyte progenitor cells from the adult human subcortical white matter. J Neurosci 19, 9986–9995. Roy, N. S., Wang, S., Jiang, L., Kang, J., Benraiss, A., Harrison-Restelli, C., Fraser, R. A., Couldwell, W. T., Kawaguchi, A., Okano, H., Nedergaard, M., and Goldman, S. A. (2000b). In vitro neurogenesis by progenitor cells isolated from the adult human hippocampus. Nat Med 6, 271–277. Rust, R. (2000). Multiple Sclerosis, acute disseminated encephalomyelitis and related conditions. Semin. Pediatr. Neurol. 7, 66–90. Rutkowski, J. L., Kirk, C. J., Lerner, M. A., and Tennekoon, G. I. (1995). PuriWcation and expansion of human Schwann cells in vitro. Nat Med 1, 80–83. Sanchez-Ramos, J., Song, S., Cardozo-Pelaez, F., Hazzi, C., Stedeford, T., Willing, A., Freeman, T. B., Saporta, S., Janssen, W., Patel, N., Cooper, D. R., and Sanberg, P. R. (2000). Adult bone marrow stromal cells diVerentiate into neural cells in vitro. Exp Neurol 164, 247–256. Saneto, R. P., Low, K. G., Melner, M. H., and de Vellis, J. (1988). Insulin/insulin-like growth factor I and other epigenetic modulators of myelin basic protein expression in isolated oligodendrocyte progenitor cells. J Neurosci Res 21, 210–219. Satoh, J. I., Tai, T., and Kim, S. U. (1996). DiVerential expression of gangliosides and galactolipids in fetal human oligodendrocytes and astrocytes in culture. Brain Res Dev Brain Res 93, 172–181. Scherer, S. S., Braun, P. E., Grinspan, J., Collarini, E., Wang, D. Y., and Kamholz, J. (1994). DiVerential regulation of the 2’,3’-cyclic nucleotide 3’-phosphodiesterase gene during oligodendrocyte development. Neuron 12, 1363–1375. SchiVmann, R., and BoespXug-Tanguy, O. (2001). An update on the leukodsytrophies. Current Opinion in Neurology 14, 789–794. Scolding, N., Franklin, R., Stevens, S., Heldin, C. H., Compston, A., and Newcombe, J. (1998). Oligodendrocyte progenitors are present in the normal adult human CNS and in the lesions of multiple sclerosis. Brain 121, 2221–2228. Scolding, N. J., Rayner, P. J., and Compston, D. A. (1999). IdentiWcation of A2B5-positive putative oligodendrocyte progenitor cells and A2B5-positive astrocytes in adult human white matter. Neuroscience 89, 1–4. Scolding, N. J., Rayner, P. J., Sussman, J., Shaw, C., and Compston, D. A. (1995). A proliferative adult human oligodendrocyte progenitor. Neuroreport 6, 441–445. Seaberg, R. M., and van der Kooy, D. (2002). Adult rodent neurogenic regions: The ventricular subependyma contains neural stem cells, but the dentate gyrus contains restricted progenitors. J Neurosci 22, 1784–1793. Shi, J., Marinovitch, A., and Barres, B. (1998). PuriWcation and characterization of adult oligodendrocyte precursor cells from the rat optic nerve. J. Neurosci. 18, 4627–4636. Shields, S. A., Blakemore, W. F., and Franklin, R. J. (2000). Schwann cell remyelination is restricted to astrocytedeWcient areas after transplantation into demyelinated adult rat brain. J Neurosci Res 60, 571–578. Sim, F. J., Zhao, C., Penderis, J., and Franklin, R. J. (2002). The age-related decrease in CNS remyelination eYciency is attributable to an impairment of both oligodendrocyte progenitor recruitment and diVerentiation. J Neurosci 22, 2451–2459. Smith, P. M., and Blakemore, W. F. (2000). Porcine neural progenitors require commitment to the oligodendrocyte lineage prior to transplantation in order to achieve signiWcant remyelination of demyelinated lesions in the adult CNS. Eur J Neurosci 12, 2414–2424. Sprinkle, T. J. (1989). 2’,3’-cyclic nucleotide 3’-phosphodiesterase, an oligodendrocyte-Schwann cell and myelinassociated enzyme of the nervous system. Crit Rev Neurobiol 4, 235–301. Tang, D. G., Tokumoto, Y. M., Apperly, J. A., Lloyd, A. C., and RaV, M. C. (2001). Lack of replicative senescence in cultured rat oligodendrocyte precursor cells. Science 291, 868–871. Tang, D. G., Tokumoto, Y. M., and RaV, M. C. (2000). Long-term culture of puriWed postnatal oligodendrocyte precursor cells. Evidence for an intrinsic maturation program that plays out over months. J Cell Biol 148, 971–984. Targett, M. P., Sussman, J., Scolding, N., O’Leary, M. T., Compston, D. A., and Blakemore, W. F. (1996). Failure to achieve remyelination of demyelinated rat axons following transplantation of glial cells obtained from the adult human brain. Neuropathol Appl Neurobiol 22, 199–206. Tate, B., Bower, K., and Snyder, E. (2001). Transplant herapy. In ‘‘Stem Cells and CNS Development’’ (M. Rao, ed.), pp. 291–306. Humana, Totowa, NJ. Temple, S., and RaV, M. C. (1986). Clonal analysis of oligodendrocyte development in culture: Evidence for a developmental clock that counts cell divisions. Cell 44, 773–739. Terada, N., Hamazaki, T., Oka, M., Hoki, M., Mastalerz, D. M., Nakano, Y., Meyer, E. M., Morel, L., Petersen, B. E., and Scott, E. W. (2002). Bone marrow cells adopt the phenotype of other cells by spontaneous cell fusion. Nature 416, 542–545. Uchida, N., Buck, D. W., He, D., Reitsma, M. J., Masek, M., Phan, T. V., Tsukamoto, A. S., Gage, F. H., and Weissman, I. L. (2000). Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci USA 97, 14720–14725.

II. GLIAL CELL DEVELOPMENT

287

DISEASE TARGETS FOR PROGENITOR-BASED THERAPEUTIC MYELINATION

Vartanian, T., Goodearl, A., Viehover, A., and Fischbach, G. (1997). Axonal neuregulin signals cells of the oligodendrocyte lineage through activation of HER4 and Schwann cells through HER2 and HER3. J Cell Biol 137, 211–220. Vaughn, J. E. (1969). An electron microscopic analysis of gliogenesis in rat optic nerves. Z Zellforsch Mikrosk Anat. 94, 293–324. Vaughn, J. E., and Peters, A. (1968). A third neuroglial cell type. An electron microscopic study. J Comp Neurol 133, 269–288. Vitry, S., Avellana-Adalid, V., Hardy, R., Lachapelle, F., and Baron-Van Evercooren, A. (1999). Mouse oligospheres: From pre-progenitors to functional oligodendrocytes. J Neurosci Res 58, 735–751. Vogel, U. S., and Thompson, R. J. (1988). Molecular structure, localization, and possible functions of the myelinassociated enzyme 2’,3’-cyclic nucleotide 3’-phosphodiesterase. J Neurochem 50, 1667–1677. Volpe, J. J. (2001). Neurobiology of periventricular leukomalacia in the premature infant. Pediatr Res. 50, 553–562. Wang, S., Wu, H., Jiang, J., Delohery, T. M., Isdell, F., and Goldman, S. A. (1998). Isolation of neuronal precursors by sorting embryonic forebrain transfected with GFP regulated by the T alpha 1 tubulin promoter. Nature Biotechnol 16, 196–201. Warrington, A. E., Barbarese, E., and PfeiVer, S. E. (1993). DiVerential myelinogenic capacity of speciWc developmental stages of the oligodendrocyte lineage upon transplantation into hypomyelinating hosts. J Neurosci Res 34, 1–13. Weissbarth, S., Maker, H. S., Raes, I., Brannan, T. S., Lapin, E. P., and Lehrer, G. M. (1981). The activity of 2’,3’cyclic nucleotide 3’-phosphodiesterase in rat tissues. J Neurochem 37, 677–680. Wenger, D. A., RaW, M. A., Luzi, P., and Datto, J. (2000). Costantino-Ceccarini E. Krabbe disease: Genetic aspects and progress toward therapy. Mol Genet Metab. 70, 1–9. Werner, H., Jung, M., Klugmann, M., Sereda, M., GriYths, I. R., and Nave, K. A. (1998). Mouse models of myelin diseases. Brain Pathol 8, 771–793. Williams, B. P., Read, J., and Price, J. (1991). The generation of neurons and oligodendrocytes from a common precursor cell. Neuron 7, 685–693. Windrem, M., Roy, N., Wang, J., Nunes, M., Benraiss, A., Goodman, R., McKhann, G., and Goldman, S. A. (2002). Progenitor cells derived from the adult human subcortical white matter disperse and diVerentiate as oligodendrocytes within demyelinated regions of the rat brain. J. Neurosci. Res. 69, 966–975. Wolswijk, G., Munro, P. M., Riddle, P. N., and Noble, M. (1991). Origin, growth factor responses, and ultrastructural characteristics of an adult-speciWc glial progenitor cell. Ann NY Acad Sci 633, 502–504. Wolswijk, G., and Noble, M. (1989). IdentiWcation of an adult-speciWc glial progenitor cell. Development 105, 387–400. Wolswijk, G., Riddle, P. N., and Noble, M. (1990). Coexistence of perinatal and adult forms of a glial progenitor cell during development of the rat optic nerve. Development 109, 691–698. Wren, D., Wolswijk, G., and Noble, M. (1992). In vitro analysis of the origin and maintenance of O-2Aadult progenitor cells. J Cell Biol 116, 167–176. Wu, E., and Raine, C. S. (1992). Multiple sclerosis. Interactions between oligodendrocytes and hypertrophic astrocytes and their occurrence in other, nondemyelinating conditions. Lab Invest 67, 88–99. Yan, H., and Wood, P. M. (2000). NT-3 weakly stimulates proliferation of adult rat O1(
)O4(þ) oligodendrocyte-lineage cells and increases oligodendrocyte myelination in vitro. J Neurosci Res 62, 329–335. Yandava, B., Billinghurst, L., and Snyder, E. (1999). Global cell replacement is feasible via neural stem cell transplantation: Evidence from the dysmyelinated shiverer mouse brain. Proc. Natl. Acad. Sci. 96, 7029–7034. Ying, Q. L., Nichols, J., Evans, E. P., and Smith, A. G. (2002). Changing potency by spontaneous fusion. Nature 416, 545–548. Yoshimura, T., Kobayashi, T., Mitsuo, K., and Goto, I. (1989). Decreased fatty acylation of myelin proteolipid protein in the twitcher mouse. J Neurochemistry 52, 836–841. Yu, W. P., Collarini, E. J., Pringle, N. P., and Richardson, W. D. (1994). Embryonic expression of myelin genes: Evidence for a focal source of oligodendrocyte precursors in the ventricular zone of the neural tube. Neuron 12, 1353–1362. Zhang, S. C., Ge, B., and Duncan, I. D. (1999). Adult brain retains the potential to generate oligodendroglial progenitors with extensive myelination capacity. Proc Natl Acad Sci USA 96, 4089–4094. Zhang, S. C., Ge, B., and Duncan, I. D. (2000). Tracing human oligodendroglial development in vitro. J Neurosci Res 59, 421–429. Zhang, S. C., Lundberg, C., Lipsitz, D., O’Connor, L. T., and Duncan, I. D. (1998). Generation of oligodendroglial progenitors from neural stem cells. J Neurocytol 27, 475–489.

II. GLIAL CELL DEVELOPMENT

au14

C H A P T E R

11 Oligodendroglial Lineage Robert Miller and Richard Reynolds

CHARACTERIZATION OF OLIGODENDROCYTE DEVELOPMENT IN VITRO Introduction The oligodendrocyte lineage represents one of the most intensely studied and best understood cell lineages in the vertebrate CNS. In part, this reXects the fact that the cells of the oligodendrocyte lineage can be grown extensively in culture, and their biochemical, morphological and molecular characteristics have been more accessible to identiWcation. As this information has been forthcoming it has been possible to go back into the more complex cellular environment of the intact CNS and validate the data obtained in simpler systems. Remarkably, the vast majority of major advances in understanding revealed by culture studies have proven to illuminate the development process or injury responses in the intact CNS. As with the other major classes of CNS neural cells the oligodendrocyte lineage is derived from cells of the neural tube. Presumably early in development, all neural tube cells have the capacity to generate oligodendrocytes, although during the course of normal embryogenesis only some cells manifest that potential. It seems likely these early neural tube cells have many characteristics of stem cells. That is, they undergo asymmetric divisions generating an identical cell and a sibling that is more restricted in either diVerentiative or proliferative potential. Currently there are no prospective isolation approaches for such cells in the CNS, only retrospective identiWcation of the progeny, although recent studies have begun to facilitate enrichment of putative neural stem cells in the adult CNS. As a result, this discussion is restricted only to those cells where the steps to a diVerentiated oligodendrocyte are reasonably well understood. During normal development there is a unidirectional progression of cells from immature to more mature states, and in the oligodendrocyte lineage this occurs in a series of overlapping phases that are demarcated by distinct morphological, biochemical, and behavioral characteristics. In general, younger animals contain more immature oligodendrocyte lineage cells than do older animals, although, as in most other lineages, individual cells within an organism, and even within a particular region of the CNS, mature and diVerentiate at diVerent rates. Indeed, the genesis of diVerentiated oligodendrocytes is extremely protracted and can be ongoing in the adult CNS. It is therefore critical to deWne maturational characteristics of individual cells rather than the overall maturation characteristics of the source animal.

Myelin Biology and Disorders, Volume 1

289

Copyright 2004, Elsevier (USA). All rights reserved.

290

11. OLIGODENDROGLIAL LINEAGE

Biochemical and Morphological Characteristics of Oligodendrocyte Lineage Cells in Vitro

au1

au2

The capacity to identify biochemical characteristics on the surface or in the cytoplasm of individual cells has proven enormously important in allowing rational cellular classiWcation. Pioneered in the hematopoetic system, the use of cell markers continues to be a cornerstone of cellular neuroscience. The usefulness of any marker depends on it being used in the appropriate context. In general, the rat CNS has the most veriWed markers, although this is likely to change with our increasing understanding and identiWcation of transcription factors and their targets. Much of the earliest work on the development of the oligodendrocyte lineage focused on the optic nerve, a region selected due to its relatively simple cytoarchitecture and lack of neuronal cell bodies (RaV et al., 1987). For example, cell cultures derived from rodent optic nerve contain exclusively glial cells with small contaminants from the meningies and blood vessels (RaV et al., 1987). The earliest well characterized cells of the oligodendrocyte lineage in the rodent optic nerve express cell surface antigens recognized by the monoclonal antibody A2B5 (RaV et al., 1984). These cells have a characteristic morphology in that they are either bipolar or unipolar with a large cell body and one or two major processes (Fig. 11.1). Several other antigenic characteristics have been ascribed to early oligodendrocyte precursors, including the expression of the NG2 antigen (Nishiyama et al., 1996), the embryonic form of polysialic acid containing neural cell adhesion molecule (E-NCAM or PSA-NCAM), and antigens recognized by a number of diVerent monoclonal antibodies including GD3. A major growth factor for oligodendrocyte precursors is platelet-derived growth factor (PDGF) (Noble et al., 1988; Richardson et al., 1988), and expression of the alpha receptor for PDGF (PDGFaR) is a further characteristic of oligodendrocyte precursors (Pringle et al., 1992). Cells with similar antigenic phenotypes have been identiWed in other regions of the CNS, including spinal cord, hindbrain, and forebrain, and precursors of the oligodendrocyte lineage clearly reside among this population. Unlike the optic nerve, however, in these more complex regions of the CNS, no single reagent allows for the unambiguous identiWcation of early oligodendrocyte precursors. Oligodendrocyte precursors undergo a number of structural and biochemical changes as they mature (PfeiVer et al., 1993). Structurally, the cells begin to develop a more complex morphology and frequently develop multiple processes (Fig. 11.1B), although they retain a relatively large cell body. Biochemically, maturing oligodendrocyte precursors begin to express antigens on their surface that bind the monoclonal antibody O4 (Bansal and PfeiVer, 1992; Bansal et al., 1992; Sommer and Schachner, 1981;). Among these antigens is the POA antigen (Bansal and PfeiVer, 1992) and galactosulfatide that subsequently becomes a structural component of myelin (Sommer and Schachner, 1981). Cells at this particular stage of development have been termed pro-oligodendrocytes denoting their progression toward diVerentiated oligodendrocytes (PfeiVer et al., 1993). Pro-oligodendrocytes also express a number of other characteristic antigens, including early myelin proteins (Fig. 11.2). The diVerentiation of oligodendrocyte precursors into oligodendrocytes is associated with the loss of expression of precursor antigens such as those recognized by mAbA2B5 (RaV et al., 1983, 1984) and the gain of expression of oligodendrocyte antigens such as galactocerebroside (RaV et al., 1978), a major glycolipid of myelin. Structurally, the morphology of oligodendrocytes becomes far more complex, with multiple cellular processes and a smaller phase dark cell body (Fig. 11.1). Under certain conditions, cultured oligodendrocytes will begin to express broad membranous sheets that have been suggested to represent the early stages of myelin formation. Continued maturation of oligodendrocytes results in elevated expression of the major myelin proteins such as myelin basic protein (MBP) and proteolipid protein (PLP) (Campagnoni, 1995; Lemke 1993).

Potential, Proliferation, and Migration of Oligodendrocyte Lineage Cells

au3

A remarkable feature of immature A2B5þ oligodendrocyte precursors is that they have the potential to generate more than one type of cell (Kondo and RaV, 2000; RaV et al., 1983).

CHARACTERIZATION OF OLIGODENDROCYTE DEVELOPMENT LINEAGE IN VITRO

291

FIGURE 11.1 Examples of rodent spinal cord cells at diVerent stages of development of the oligodendrocyte lineage in vitro. (A) Immature oligodendrocyte precursors labeled with mAb A2B5. These immature cells with a bipolar morphology are highly migratory and proliferate in response to their major mitogen PDGF. During development, A2B5þ cells diVerentiate constitutively into oligodendrocytes and can be induced to diVerentiate into type 2 astrocytes under the inXuence of serum or CNTF stimulation, and even into neurons in response to the correct stimulation. (B) A pro-olgiodendrocyte labeled with mAb O4. Compared to immature oligodendrocyte precursors, O4þ cells have a larger cell body and a more process-bearing morphology. The O4þ cells are less proliferative, responding primarily to FGF rather than PDGF. Furthermore, 04þ cells are considerably less migratory than A2B5þ cells, although they actively extend processes in vitro. (C) A newly diVerentiated oligodendrocyte labeled with mAb 01, which recognizes galactocerebroside as well as other myelin. Compared to oligodendrocyte precursors, diVerentiated oligodendrocytes have a more complex morphology with multiple cellular processes. DiVerentiated oligodendrocytes seldom divide in vitro, or under normal conditions in vivo, and are not migratory. As these cells mature in vitro, they extend wide processes or sheets of membrane that may correspond to uncompacted myelin sheaths in vivo. Bar ¼ XXmm?? in all Wgures.

FIGURE 11.2 Schematic representation of the major cellular stages in the development of myelinating oligodendrocytes in vitro. Like other lineages of the CNS, oligodendrocyte precursors (OPC) arise from stem cells that can develop into either neurons or type –1 astrocytes, depending on environmental signals. Oligodendrocyte precursors are bipotential in vitro, giving rise to type 11 astrocytes or pro-oligodendrocytes. In vivo, these cells mature to give rise to pro-oligodendrocytes or adult progenitor cells. During development, pro-oligodendrocytes diVerentiate into immature oligodendrocytes that either undergo cell death due to a lack of survival factor or mature into myelinating cells that myelinate several axonal segments depending on axonal diameter. The major molecule regulators of each stage in the development of oligodendrocytes is discussed in detail in the text.

While they constitutively diVerentiate into oligodendrocytes, under the inXuence of diVerent environmental signals these cells can give rise to the major classes of neural cells (Kondo and RaV, 2000). The Wrst evidence of the multipotential nature of the A2B5þ cells from the developing optic nerve was the Wnding that under the inXuence of serum these cells gave rise to a distinct population of astrocytes termed type 2 astrocytes (RaV et al., 1984). This bipotential nature led to these cells being termed oligodendrocyte-type-2 astrocyte (O-2A) progenitor cells (RaV et al., 1984). Type 2 astrocytes can be distinguished from other type 1 astrocytes, which express intermediate Wlaments composed of glial

292

11. OLIGODENDROGLIAL LINEAGE

au4

au5

Wbrillary acidic protein (GFAP) based on the process-bearing morphology and diVerent antigenic characteristics, including the short-term expression of mAb A2B5 binding (RaV et al., 1983). Subsequent studies have identiWed some environmental factors that can induce O2A progenitor cells to diVerentiate into type 2 astrocytes. Such factors include serum and more speciWcally ciliary neurotrophic factor (CNTF) (Hughes et al., 1988; Lillien et al., 1990). More recently it has become clear that O-2A progenitors can be induced to give rise not only to type 2 astrocytes but also to type 1 astrocytes and neurons, suggesting these cells have stem cell like properties (Kondo and RaV, 2000). Indeed, in regions of the CNS such as the spinal cord, mAb A2B5 labeling can be used to isolate cells with stem cell like properties (Rao et al., 1998; Rao and Mayer-Proschel, 1997). Since, however, it is unclear if A2B5þ cells manifest any potential diVerentiative diversity apart from the generation of oligodendrocyte in vivo (discussed later), they will be termed oligodendrocyte precursors in this chapter. Other markers identify bipotential oligodendrocyte precursors, including expression of NG2 (Levine and Card, 1998; Nishiyama et al., 1996). It is currently unclear whether A2B5þ and NG2 cells represent the same stage in the oligodendrocyte lineage or whether one is an earlier cell in the lineage. Circumstantial evidence suggests that NG2þ cells may generate A2B5þ cells and thus represent an earlier stage in the oligodendrocyte lineage. During development, the capacity of oligodendrocyte precursors to generate type 2 astrocytes is transient and is lost as the cells begin to express POA and other antigens recognized by mAbO4. While O4þ cells appear to be able to generate type 2 astrocytes under certain conditions, they may have to revert to a more immature cell phenotype to do so. The capacity to revert to more immature cells as well as the capacity to generate type 2 astrocytes is lost when oligodendrocyte precursors diVerentiate into oligodendrocytes. The growth factors that promote oligodendrocyte precursor proliferation are diVerent depending on the developmental stage of the cells (Fok-Seang and Miller, 1994; PfeiVer et al., 1993). In general, more immature oligodendrocyte precursors have a greater proliferative capacity than do more mature oligodendrocyte precursors. The major mitogen for immature A2B5þ oligodendrocyte precursors is platelet-derived growth factor (PDGF) (Noble et al., 1988; Richardson et al., 1988). This mitogenic response is mediated through the PDGF alpha receptor (PDGFaR) (Pringle et al., 1992), the expression of which overlaps with that of A2B5 immunoreactivity and therefore characterizes oligodendrocyte precursors in speciWc regions of the developing CNS. Immature oligodendrocyte precursors also respond to Wbroblast growth factor (FGF), although to a somewhat lesser extent (Fok-Seang and Miller, 1994), while a combination of PDGF and FGF promotes extended proliferation and inhibits diVerentiation (Bogler et al., 1990; Gard and PfeiVer, 1993; Mayer et al., 1993; McKinnon et al., 1990). Several other cytokines and growth factors inXuence the proliferation of immature oligodendrocyte precursors. These include the neurotrophin NT3 (Barres et al., 1994), insulin-like growth factor and the chemokine CXCL1 (Robinson et al., 1998). While not a mitogen in its own right, CXCL1 enhances the proliferative response of immature oligodendrocyte precursors to PDGF (Robinson et al., 1998; Wu et al., 2000), and this synergy is mediated through the chemokine receptor CXCR2 (Robinson et al., 1998; Tsai et al., 2000). In contrast to A2B5þ cells, O4þ prooligodendroblasts are largely refractory to PDGF but proliferate largely in response to bFGF (Fok-Seang and Miller, 1994). The response to bFGF is mediated through diVerent receptors depending on the maturity of the precursor cells (Bansal et al., 1996). Exposure of O4þ cells to FGF not only promotes proliferation but also modiWes cell fate. Initial studies suggested that treatment with FGF induced a reversion of the O4þ cells to a more immature phenotype (Grinspan et al., 1993); however, subsequent analyses indicates that rather than a reversion, FGF exposure induces a novel cell phenotype (Bansal and PfeiVer, 1997), whereby cells may become more susceptible to diVerentiation inducing signals. Upon diVerentiation into GCþ oligodendrocytes, cells largely drop out of the cell cycle and in the majority of cases become post-mitotic (PfeiVer et al., 1993). Not only do growth factors promote the proliferation of oligodendrocyte precursors, but inhibitory signals suppress the proliferation of oligodendrocyte precursors. For example, soluble signals such as TGFb (Louis et al., 1992; McKinnon et al., 1993) tend

CHARACTERIZATION OF OLIGODENDROCYTE DEVELOPMENT LINEAGE IN VITRO

au6

to inhibit the proliferation of oligodendrocyte precursors and can counteract the proliferative stimulation of growth factors. More localized signals can also inhibit the proliferation of oligodendrocyte precursors in response to growth factors (Nakatsujji and Miller, 2001; Zhang and Miller, 1996). Like many other cell types, oligodendrocyte precursors demonstrate contact dependent inhibition of proliferation (Wieser et al., 1990). This phenomenon is cell type speciWc and can signiWcantly truncate the number of cell divisions a precursor cell undergoes (Nakatsujji and Miller 2001). The ligand/receptor complexes mediating contact inhibition of proliferation are as yet uncharacterized; however, the cell cycle machinery utilized to inhibit cell proliferation is better understood and involves coordinate down-regulation of speciWc cyclins and an up-regulation of the cell cycle inhibitor p27 (Hengst and Reed 1996; Kato et al., 1997; Nakatsujji and Miller 2001). Similar molecular mechanisms are utilized during the cessation of cell proliferation that accompanies the transition from precursor cells to diVerentiated oligodendrocytes (discussed later). Oligodendrocyte precursors are highly migratory. In cultures of developing optic nerve, the most immature cells are the most migratory (Noble et al., 1988). For example, bipolar A2B5þ cells are extremely motile and are some of the most motile cells in the body (Noble et al., 1988). By contrast, O4þ cells are far less migratory and diVerentiated oligodendrocytes are largely non-motile (Warrington et al., 1993). The migration of immature oligodendrocyte precursors is inXuenced by a variety of environmental cues. Type 1 astrocytes stimulate both the proliferation and migration of A2B5þ cells (Noble et al., 1988). This activity probably reXects the synthesis and secretion of PDGF by these cells (Noble et al., 1988). Consistent with this hypothesis, PDGF promotes both proliferation and migration of oligodendrocyte precursors (Noble et al., 1988). Not only does PDGF promote oligodendrocyte precursor migration, but it may also act as a chemoattraction factor (Armstrong et al., 1990, 1991; Frost et al., 2000). For example, in Boyden chamber studies oligodendrocyte precursors migrated toward higher concentrations of PDGF (Armstrong et al., 1990). Likewise, bFGF may also act as a chemoattractant for oligodendrocyte precursors (Armstrong et al., 1990). Other guidance molecules such as the semaphorins and netrins (Sugimoto et al., 2001) also inXuence oligodendrocyte migration while the chemokine CXCL1 acts as a stop signal for migratory oligodendrocyte precursors (Tsai et al., 2000). The issues of migrational control of oligodendrocyte precursors are discussed in more detail in a later chapter.

Differentiation and Survival of Oligodendrocyte Precursors: Control of Cell Number Oligodendrocyte precursors diVerentiate into immature oligodendrocytes before they generate myelin (Figs. 11.1 and 11.2). This is a critical transition in the lineage and has many important cellular consequences. For example, as cells exit the cell cycle they become acutely dependent on survival factors (Barres et al., 1992, 1993). The mechanisms mediating the transition from precursor to diVerentiated cell are still not clearly understood. Both cell intrinsic and extrinsic signals appear to regulate this critical transition. In vitro, clonally related oligodendrocyte precursors derived from the rat optic nerve diVerentiate at approximately the same time and after similar numbers of divisions (Temple and RaV, 1985, 1986), suggesting diVerentiation is in part regulated through an intrinsic clock that measures the number of cell divisions or elapsed time (RaV et al., 1985; RaV and Lillien, 1988). The nature of the clock is unclear but appears to be comprised of several components including the cell cycle regulator p27 (Casaccia-BonneWl et al., 1997). Extrinsic signals also inXuence oligodendrocyte diVerentiation (Bogler et al., 1990; McKinnon et al., 1990). Axonal signals promote the diVerentiation of oligodendrocyte precursors (Payne and Lemmon, 1993), and considerable circumstantial evidence suggests that axons provide a signal that initiates the process of myelination (Trapp 1990). One potential candidate for an oligodendrocyte diVerentiation signal is Neuregulin 1 (Marchionni et al., 1993). Neuregulin is expressed by many myelinating axons and induces process outgrowth in oligodendrocytes (Vartanian et al., 1994). Neuregulins signal by binding to the ErbB receptors (Lemke, 1996). In the absence of ErbB2 signaling in vitro oligodendrocyte precursors fail to diVerentiate and

293

294

11. OLIGODENDROGLIAL LINEAGE

au7 au8

au9

remain as precursors (Park et al., 2001). In the intact CNS, treatment with high doses of Neuregulin induces a demyelination, possibly as a result of driving a de-diVerentiation of precursors. Axonal signals may not only promote oligodendrocyte diVerentiation, they may also inhibit diVerentiation. For example, many axons express delta and other notch ligands on their surface. Stimulation of the notch signaling pathway by these ligands appears to inhibit oligodendrocyte precursor diVerentiation (Wang et al., 1998). Soluble signals also regulate diVerentiation of oligodendrocytes. For example, exposure to thyroid hormone elevates the number of mature oligodendrocytes as well as directly enhancing the expression of myelin-speciWc genes (Barres et al., 1994). The inXuence of other signals such as retinoic acid varies depending on either the developmental stage of the cells or their origin. For example, retinoic acid promotes the diVerentiation of postnatal optic nerve oligodendrocyte precursors (Barres et al., 1994) while it inhibits the diVerentiation of embryonic rat spinal cord oligodendrocyte precursors (Noll and Miller, 1994). Although oligodendrocyte precursors are highly proliferative, mature diVerentiated oligodendrocytes do not proliferate extensively; for this reason, the regulation of oligodendrocyte precursor number is crucial for generating suYcient oligodendrocytes in the adult. The Wnal number of oligodendrocytes that develop in a speciWc region of the CNS is closely correlated with the number of axons that require myelination (Barres and RaV, 1994; Burne et al., 1996). This matching of cell number is accomplished both by regulation of cell proliferation and cell survival (Barres et al., 1992; Barres and RaV, 1994) (Fig. 11.1). As with neuronal lineages, oligodendrocyte precursors are produced in excess and apoptosis and programmed cell death remove extraneous cells (Barres et al., 1992). Newly formed oligodendrocytes depend on PDGF for survival both in vitro and in vivo, while more mature oligodendrocytes depend on insulin growth factor–1 (IGF-1) and the neurotrophin NT3 for survival (Barres et al., 1992, 1994). Increased understanding of the biology of oligodendrocyte development has revealed that, at least in vitro, cells of this lineage respond to an extremely broad range of diVerent molecular cues. Why such diverse signaling exists is unclear but may be a reXection of the multiple cellular interactions that impinge on the cells destined for an oligodendrocyte fate. The challenge for current and future studies is to discern which cues predominate in the developing intact CNS.

CHARACTERIZATION OF OLIGODENDROCYTE DEVELOPMENT IN VIVO Overview Much of the basic developmental biology of the oligodendrocyte lineage Wrst revealed in a variety of in vitro systems has proven to be directly applicable to understanding oligodendrocyte development in the intact CNS. Regional diVerences in the cytoarchitecture of the developing CNS as well as the detailed molecular control of oligodendrocyte development do, however, add considerable complexity to the system as well as providing critical insights into the fundamental pathways. For example, recent advances on understanding the transcriptional control of oligodendrocyte development stem directly from the localization of the earliest cells of the lineage in diVerent regions of the CNS. In many cases, the underlying molecular mechanisms controlling speciWc aspects of oligodendrocyte development are less clearly understood in the intact CNS due presumably to the existence of multiple regulators at each stage.

Oligodendrocyte Precursors Arise in Restricted Regions of the CNS during Development Although the majority of mature oligodendrocytes are located in white matter, the founder cells of the oligodendrocyte lineage arise early in development in restricted regions of the neural tube (Warf et al., 1991). Early oligodendrocyte development has been most exten-

CHARACTERIZATION OF OLIGODENDROCYTE DEVELOPMENT IN VIVO

FIGURE 11.3

au15

Developing oligodendrocytes in vivo. Transverse section of the ventral spinal cord labeled with antibodies to myelin basic protein. Individual diVerentiating oligodendrocytes are seen throughout the developing white matter surrounded by 1-6 myelin proWles. Bar ¼ XXmm.

sively studied in selected regions of the CNS such as the spinal cord and the optic nerve. (Fig. 11.3) In the rat optic nerve, tissue culture studies suggested the founder cells of the oligodendrocyte lineage originated in the brain or optic chiasm and migrated along the nerve during subsequent development (Small et al., 1987). For example, isolated cultures of the retinal end of embryonic and early postnatal rat optic nerve do not develop oligodendrocytes, while parallel cultures of the regions of the nerve closer to the optic chiasm do (Small et al., 1987). A similar migration of oligodendrocyte precursors occurs during development of the chick optic nerve where a source of at least a subset of optic nerve oligodendrocytes has been deWned in the Xoor of the third ventricle (Ono et al., 1997). In rodents, the migration of oligodendrocyte precursors is inhibited at junction between the nerve and the retina (Vrench-Constant et al., 1988), while in chick the cells continue into the retina where they myelinate retinal ganglion cell axons (Ono et al., 1998, 2001). In the spinal cord, the earliest oligodendrocyte precursors arise in the ventral ventricular zone after the majority of neurogenesis is complete (Ono et al., 1995; Pringle and Richardson 1993). Several independent approaches have localized the source of spinal cord oligodendrocytes to a domain of the neural tube dorsal to the Xoor plate of the neural tube including their localized proliferation (Noll and Miller, 1993), expression of growth factor receptors (Pringle and Richardson, 1993), and biochemical proWle (Ono et al., 1995; Timsit et al., 1995). Oligodendrocyte precursors can be identiWed by expression of the PDGFaReceptor (PDGFa-R) (Pringle et al., 1992) and in situ hybridization indicated PDGFa-Rþ cells are initially seen in the same region of developing spinal cord as cells expressing mRNA for the myelin genes CNP (2’,3’-cyclic-nucleotide 3’-phosphodiesterase) (Yu et al., 1994) and DM20, an isoform of the major myelin proteolipid protein (PLP) (Timsit et al., 1995) gene as well as in chick antigens recognized by the O4 monoclonal antibody (mAb) (Ono et al., 1995). These cells arise in the same ventricular domain that generates motor neurons and appear around the time the generation of motor neurons is complete (Ono et al., 1995; Pringle et al., 1992; Richardson et al., 1997). The localized origin of oligodendrocyte precursors is not restricted to the spinal cord. In more rostral regions of the CNS, oligodendrocyte precursors arise in speciWc regions of the ventricular and subventricular zone at particular stages of development (Ono et al., 1997). For example, cells in the ventricular mantle zone of the ventral diencephalon of the E13 rat express mRNA for the PDGFa-R (Pringle and Richardson 1993). During subsequent

295

296

11. OLIGODENDROGLIAL LINEAGE

au10

development, these cells appear to migrate into the developing thalamus and hypothalamus as well as to more dorsal regions including the developing cerebellum (Pringle and Richardson, 1993). While most regions of the CNS seem to be populated by precursor cells derived from a single region, in the telecephalon multiple oligodendrocyte precursor domains have been described (Spassky et al., 1998, 2001). These include the anterior penducle area and the olfactory bulb (Olivier et al., 2001; Spassky et al., 2001). It is currently unclear if the progeny of each of these domains are identical and whether they contribute oligodendrocytes to overlapping or nonoverlapping axon tracks.

Molecular Control of Oligodendrocyte Precursor Specification

au11 au12

The mechanisms by which neuroepithelial cells become speciWed to the oligodendrocyte lineage are best understood in the developing spinal cord where two general mechanisms may account for their ventral origin. It could be that cells in dorsal regions lack the intrinsic potential to generate oligodendrocytes regardless of extrinsic cues. Alternatively, both dorsal and ventral cells have the capacity to generate oligodendrocytes, but only ventral cells receive the appropriate cues. Several lines of evidence demonstrate that the ventral origin of spinal cord oligodendrocyte precursors is a reXection of local signaling. For example, local inXuences from the notochord, a ventrally located mesodermally derived structure, are required for the appearance of oligodendrocytes in adjacent spinal cord in both mouse (Pringle et al., 1996) and Xenopus embryos (Maier and Miller, 1997). Signals from the notochord are involved in formation of the dorsal/ventral axis in the developing CNS, which results in the subsequent speciWcation of distinct populations of spinal cord neurons (Jessell and Dodd, 1990; van Straaten et al., 1988, 1989). Transplantation of an additional notochord adjacent to the dorsal spinal cord at the appropriate stage in development resulted in the local induction of oligodendrocyte precursors in chick and Xenopus embryos (Maier and Miller, 1997; Orentas and Miller, 1996;), while co-culture of notochord and dorsal spinal cord results in oligodendrocyte induction in the spinal cord tissue (Orentas and Miller, 1996; Poncet et al., 1996; Pringle et al., 1996; Trousse et al., 1995). Many of the inductive properties of the notochord appear to be due to its production of the signaling molecule sonic hedgehog (Echelard et al., 1993; Roelink et al., 1994). Sonic hedgehog, the vertebrate homologue of the Drosophila pattern forming gene hedgehog, is localized to the notochord and adjacent Xoor plate (Roelink et al., 1994). In vitro, sonic hedgehog induces the development of Xoor plate and motor neurons in a concentration dependent manner (Roelink et al., 1994). In vitro, oligodendrocytes are induced at similar concentrations of Shh required for the induction of motor neurons (Pringle et al., 1996), suggesting that the development of these two cell types is closely linked (Richardson et al., 1997; Rowitch et al., 2002). In more rostral regions of the CNS, the expression of Shh and the appearance of oligodendrocytes is spatially and temporally closely linked (Davies and Miller, 2001; Nery et al., 2001; Spassky et al., 2001; Tekki-Kessaris et al., 2001). Furthermore, ectopic expression of Shh leads to concomitant local development of oligodendrocytes (Nery et al., 2001). Whether Shh is essential for the development of all rostral populations of oligodendrocytes is less clear. In cell cultures derived from Shh knockout animals, considerable numbers of oligodendrocytes develop, indicating that oligodendrocytes can arise in the absence of Shh signaling (Nery et al., 2001). It seems likely, however, that other members of the hedgehog family can substitute for Shh in its absence and blocking all hedgehog family member signaling with cyclopamine (Incardona et al., 1998) appears to block all oligodendrocyte development (Tekki-Kessaris et al., 2001). In vitro, the development of oligodendrocyte precursors is inhibited by exposure to members of the TGF& family (Mabie et al., 1997). SpeciWcally, bone morphogenetic proteins 2 and 4 inhibit the development of oligodendrocytes (Mabie et al., 1997, 2000). This appears to be in part a reXection of the commitment of cells to astrocyte lineages at the expense of the oligodendrocyte lineage (Mabie et al., 1997, 2000). Whether BMP signaling contributes to the spatial patterning of oligodendrocyte precursor induction in

CHARACTERIZATION OF OLIGODENDROCYTE DEVELOPMENT IN VIVO

the developing intact CNS is currently unclear although recent studies suggest that BMPs inhibit spinal cord oligodendrocyte development (Mekki-Dauriac et al., 2002) and it seems likely that the lack of dorsally derived oligodendrocytes in the spinal cord reXect active inhibition by BMPs. Indeed, if the BMP source was adjacent to the spinal cord, this hypothesis would explain why oligodendrocyte develop in isolated explants of dorsal spinal cord over time (Sussman et al., 2000). Additional, as yet uncharacterized, inhibitors of oligodendrocyte precursor development may also exist, since dorsal spinal cord contains an inhibitor of early oligodendrocyte development (Wada et al., 2000) functionally distinct from any known BMP (Wada et al., 2000). The generation of motor neurons and oligodendrocytes is closely linked (Richardson et al., 1997; Rowitch et al., 2002). Not only do they arise from the same region of the spinal cord and require the same concentration of Shh (Pringle et al., 1996), neurons and oligodendrocytes also arise from clonally related cells in vitro and in vivo (He et al., 2001; Leber et al., 1990; Leber and Sanes, 1991; Williams et al., 1991). The diVerentiation of neuropepithelial cells into a neuronal or glial fate is regulated by expression of distinct combinations of transcription factors (Kessaris et al., 2001). For example, the ventricular zone of the ventral spinal cord contains several speciWc cellular domains identiWed by diVerent transcription factor expression and generating distinct cell populations (Briscoe et al., 2001; Jessell, 2000). Oligodendrocyte precursors arise during later development from the motor neuron pool that is characterized by expression of the transcription factor Olig2 (Zhou et al., 2001; Zhou and Anderson 2002). Olig2 is thought to combine with the Basic Helix-Loop-Helix transcription factors neurogenin 1 and 2 to generate motor neurons. As development proceeds, however, the expression of neurogenins is down-regulated (Zhou et al., 2001) and this allows for an alteration in the distribution pattern of a more ventrally expressed transcription factor, Nkx2.2, into the Olig2 domains such that they overlap. Cells that express both transcription factors subsequently develop into spinal cord oligodendrocytes rather than motor neurons (Kessaris et al., 2001; Zhou et al., 2001). Several lines of evidence support such a model. For example, oligodendrocytes can be ectopically generated in other regions of the CNS by expression of both Olig2 and either Nkx2.2 or components of the Notch signaling pathway (Zhou et al., 2001). ConWrmation of the requirement and roles of the Olig transcription factors in the genesis of oligodendrocytes has come from targeted disruption of these genes. Olig 2 is required for the speciWcation of both motor neurons and oligodendrocytes, while Olig1 is required for the later development of oligodendrocytes, particularly in rostral regions of the CNS (Lu et al., 2002). In the absence of both Olig1 and 2, the cells that would normally give rise to motor neurons and oligodendrocyte generate a speciWc class of interneurons and surprisingly astrocytes (Zhou and Anderson, 2002). The simple model of a restricted motor neuron oligodendrocyte precursor will, however, require further reWnement since in the mouse CNS some early oligodendrocyte precursor cells arise outside the Nkx2.2þ domains of the CNS (Lu et al., 2000; Sun et al., 1998). It may be that there is more than one population of oligodendrocyte precursors, which diVer in the mechanism by which they become speciWed (Spassky et al., 1998, 2000) or some cells previously characterized as oligodendrocyte precursors are in fact astrocyte precursors or glial restricted precursors (Rao and Mayer-Proschel 1997).

Control of Oligodendrocyte Number in Vivo The functioning of the adult CNS depends critically on matching the correct number of oligodendrocytes to axons. In the developing nervous system, oligodendrocyte cell number is regulated by several independent mechanisms. The proliferation of oligodendrocyte precursors is mediated in large part by the availability of mitogens such as PDGF (Richardson et al., 1988). Thus, increasing levels of PDGF result in increasing numbers of oligodendrocyte precursors, at least in the spinal cord (Calver et al., 1998). This increase in precursor cells may also reXect increased cell survival, since PDGF is a strong survival factor cells of the oligodendrocyte lineage (Barres et al., 1992; Barres and RaV, 1994; Calver et al., 1998). The Wnal number of diVerentiated oligodendrocytes in any particular region of the CNS is a combination of precursor proliferative control, cell diVerentiation,

297

298

11. OLIGODENDROGLIAL LINEAGE

and regulation of cell survival (Barres and RaV, 1994). The diVerentiation of oligodendro-

au13 cytes is regulated by thyroid hormone (Barres et al., 1994) and the intrinsic redox state of individual cells (Smith et al., 2000), which also alters the potential for cell proliferation. Likewise, modulation of cell cycle regulators such as p27kip-1 and p21cip-1 inXuence oligodendrocyte development (Casaccia-BonneWl et al., 1997; Casaccia-BonneWl, 2000; Durand et al., 1997). In the absence of p27Kip-1 the number of oligodendrocytes is signiWcantly altered, although they diVerentiate on time (Casaccia-BonneWl et al., 1997). By contrast, in the absence of p21cip-1 the diVerentiation of oligodendrocytes is disrupted (Zezula et al., 2001), suggesting this molecule is required for their timely diVerentiation. The control of cell survival is a major component in determining oligodendrocyte number in the developing CNS. In the developing optic nerve as well as other regions of the CNS, the Wnal number of oligodendrocytes that ultimately diVerentiate appears to be regulated by competition for local survival factors (Barres and RaV, 1994) including PDGF and possibly axonally derived neuregulin (Park et al., 2001). A signiWcant number of newly generated oligodendrocytes die during normal development, indeed in the optic nerve this has been estimated to be as high as 50% (Barres et al., 1992). This extent of cell death can be reduced dramatically by increased expression of PDGF or insulin-like growth factor (Barres et al., 1992). Since the ultimate goal of oligodendrocytes is to myelinate all available axons during development, it seems likely that expression of oligodendrocyte survival signals will be either directly or indirectly controlled by the number of available axons (Barres and RaV, 1994; Burne et al., 1996). Evidence to support this hypothesis comes from the increased oligodendrocyte number in animals in which the number of retinal axons is increased as a result of inhibition of cell death while removal of axons results in decreases in the number of optic nerve oligodendrocytes (Burne et al., 1996). Not all oligodendrocyte precursors in the developing nervous system either diVerentiate or die. A signiWcant population of oligodendrocyte precursors persist as potential progenitor cells in the adult CNS and these enigmatic cells are discussed further in the following sections.

THE OLIGODENDROCYTE LINEAGE IN THE ADULT CNS The Existence of a Population of Glial Progenitors in the Adult CNS By late embryonic stages in the rodent, oligodendrocyte progenitors (OPCs), identiWed by their expression of either PDGFaRþ, NG2, or O4, have become evenly distributed throughout the presumptive gray and white matter of the developing brain and spinal cord (Nishiyama et al., 1996; Pringle et al., 1992; Reynolds and Hardy, 1997). (Fig. 4) As described earlier in this chapter, a combination of the action of soluble and contactdependent growth factor signals, together with an intrinsic timing mechanism, leads to the diVerentiation of the progenitors into myelin producing oligodendrocytes. It is clear that, unlike progenitors, the distribution of myelinating oligodendrocytes is not even throughout the CNS but reXects the diVerent tract speciWc requirements for myelin. There is evidence that newly formed oligodendrocytes are dependent on axonal signals for their survival and undergo apoptosis if they do not contact competent axons that require myelination. In this way oligodendrocyte number is matched to the requirement for myelin. Thus, the two major fates of oligodendrocyte progenitors are to diVerentiate into mature myelinating oligodendrocytes or to die by apoptosis. There is, however, considerable evidence to suggest that not all perinatal oligodendrocyte progenitors undergo these two fates. Here we review the evidence that a proportion of progenitors persist in the adult CNS in a phenotypically immature form and discuss the origins and functions of these cells.

In Vitro Studies A number of studies have demonstrated that cells with the phenotype and expected growth characteristics of oligodendroglial progenitors can be isolated from the optic nerve

THE OLIGODENDROCYTE LINEAGE IN THE ADULT CNS

FIGURE 11.4 Appearance of oligodendrocyte lineage cells in the adult CNS. NG2 expressing cells seen in the rat hippocampal gray matter (A) and in the corpus callosum (B). The multiple processed nature of these cells is clearly seen in the hippocampus (A, inset). NG2-expressing cells are seen to be intermingled with, but distinct from, OX-42þ microglia (C, cerebral cortex), GFAPþ astrocytes (D, corpus callosum), and CNPþ oligodendrocytes (E, cerebral cortex). A pair of NG2þBrdUþ cells are seen in the rat cerebral cortex 6 days after a 6-hour pulse of BrdU (F). OPCs can be isolated from the adult forebrain using O4 and A2B5 double immunopanning. (G) A bipolar A2B5þ cell 1 day after isolation. (H) A Weld of galatocerebroside expressing cells 3 days after isolation.

299

300

11. OLIGODENDROGLIAL LINEAGE

au14 (Vrench-Constant and RaV, 1986; Shi et al., 1998; Wolswijk and Noble, 1989) and cerebellum (Levine et al., 1993) of adult rats. These cells express the A2B5, O4, PDGF aR, and NG2 antigens and therefore have a more mature phenotype than the early perinatal progenitors that do not yet express O4 (Dawson et al., 2000). When grown in serum-free chemically deWned medium they diVerentiate rapidly into oligodendrocytes, and when grown in 10% fetal calf serum containing medium they take on an astrocyte phenotype, indicated by the expression of glial Wbrillary acidic protein (GFAP). Because of the similarity of these cells to the O-2A progenitor cells isolated from neonatal optic nerves (RaV et al., 1983), they were termed O-2Aadult progenitors (Wolswijk and Noble, 1989). A number of fundamental diVerences between perinatal and adult progenitors have been identiWed. Progenitor cells derived from perinatal optic nerves are highly motile (Small et al., 1987) with a cell cycle of about 18 hours (Noble et al., 1988; Temple and RaV, 1986) and can be induced to diVerentiate into either oligodendrocytes or astrocytes in less than 3 days (RaV et al., 1983). In contrast, adult OPCs migrate at a slower rate, have an average cell cycle of 65 hours, and also diVerentiate at a slower rate (Shi et al., 1998; Wolswijk and Noble, 1989). Reports that the morphology of progenitors isolated from the adult optic nerve diVered from those isolated from the developing optic nerve (Wolswijk and Noble, 1989) are not substantiated in other studies (Vrench-Constant and RaV, 1986; Shi et al., 1998). Later studies have identiWed two populations of OPCs in the adult rat brain: an A2B5þ O4
population that is bipotential and an O4þ population that appears committed to the oligodendrocyte lineage (Gensert and Goldman, 2001). These two populations diVer in their responsiveness to growth factors in vitro (Mason and Goldman, 2002). Although the precise relationship of these cells to one another and to those identiWed in vivo is not clear, it is clear that the mature CNS contains a population of cells with the potential to give rise to oligodendrocytes when isolated in culture (Fig. 11.4G,H).

In Vivo Morphological Studies The glial component of the adult CNS is generally thought to consist of astrocytes, oligodendrocytes, and microglia. However, there have been many reports over the past 40 years that a proportion of cells in the CNS do not fall easily into these categories. Although these cells have been largely ignored for a long time, there is increasing interest in these populations mainly because they express characteristics of immature cells and, therefore, may be important for regenerative purposes. Smart and Leblond (1961) identiWed a population of cells in the adult mouse corpus callosum that had the morphological characteristics of immature glia and were mitotically active. They were shown to have some ultrastructural features in common with oligodendrocytes and were classiWed as light and medium oligodendrocytes, precursors to the post-mitotic dark oligodendrocytes (Mori and Leblond, 1970). Vaughn and Peters (1968) identiWed cells in the developing and adult rat optic nerve lacking intermediate Wlaments and microtubules. They showed these cells to comprise 5% of all glia in the optic nerve and suggested that they might represent a multipotential stem cell. A population of cells with similar characteristics were found in the adult rat cortical gray matter (Reyners et al., 1982, 1986) and were called betaastrocytes. They were found to be mitotically active and were suggested to be glial precursor cells. However, until very recently it was not possible to unequivocally identify these cells as part of the oligodendrocyte lineage.

Characterization and Distribution of Oligodendrocyte Progenitors in the Adult CNS Antibodies to both the NG2 chondroitin sulphate proteoglycan and the PDGF a-receptor have been used extensively to identify OPCs in the developing CNS (reviewed in Levine et al., 2001) and more recently have been applied to studies of the mature CNS. In the adult rodent, these antibodies identify a widespread and numerous population of cells that have the phenotype of late oligodendroglial progenitors (Dawson et al., 2000; Fig. 11.4A,B). Although NG2-immunopositive, process-bearing cells present in the mature rat cerebellum were initially identiWed as smooth protoplasmic astrocytes on the

THE OLIGODENDROCYTE LINEAGE IN THE ADULT CNS

basis of their light microscope and ultrastructural characteristics (Levine and Card, 1987), later studies suggested that they corresponded to the in vivo counterparts of the optic nerve O-2Aadult progenitors described in vitro (Levine et al., 1993; Wolswijk and Noble, 1989). In all regions of the adult CNS, both gray and white matter, NG2 expression completely overlaps with the expression of the PDGF a-receptor (Dawson et al., 2000; Nishiyama et al., 1996; Reynolds et al., 2001). More compelling evidence that these cells represent OPCs comes from studies demonstrating that they also express the O4 antigen in vivo (Dawson et al., 2002; Reynolds and Hardy, 1997), but not antigens characteristic of more mature cells of the oligodendrocyte lineage, for example, CNP, galactocerebroside and myelin basic protein (Dawson et al., 2002; Levine et al., 1993; Reynolds and Hardy, 1997; Reynolds et al., 2001). However, there have been a number of recent reports that OPCs in the adult CNS may express low levels of galactocerebroside and myelin/oligodendrocyte glycoprotein (MOG) (Li et al., 2002; Shi et al., 1998). All this evidence taken together suggests that these cells are phenotypically part of the oligodendrocyte lineage and are closely related to the late NG2þ/PDGF-aRþ/O4þ progenitor seen during development. They appear to stop diVerentiating at the point in the lineage before they diVerentiate into post-mitotic young oligodendrocytes. However, this does not mean that their only function is that of becoming oligodendrocytes following demyelination or that they are all capable of such diVerentiation. A number of studies have looked at the relationship between NG2þ/PDGF-aRþ cells and the other major glial cell types of the mature CNS. Although when isolated into culture OPCs from the adult CNS appear to be bipotential, there is no evidence of expression of markers known to be speciWc for astrocytes, such as GFAP and S100a, (Dawson et al., 2000; Nishiyama et al., 1996) in vivo (Fig. 11.4D). In addition, in the normal resting CNS, NG2þ/PDGF-aRþ cells do not express microglial markers such as OX-42 (Fig. 11.4C) (Levine et al., 1993; Reynolds and Hardy, 1997) or GSA I-B4 lectin (Nishiyama et al., 1997). NG2þ cells are found in close association with MAP2þ neurons in the cerebral cortex but do not themselves express MAP2 (Reynolds et al., 2001). It remains to be seen whether they are capable of a fate other than oligodendrocyte following appropriate stimulation in vivo, but data available at present clearly demonstrate that in the resting situation they are phenotypically part of the oligodendrocyte lineage. Adult OPCs display a highly branched morphology that is very diVerent when compared to developing OPCs and one that does not Wt easily into our concept of immature progenitor cells (Dawson et al., 2000; Levine et al., 2001), although it may be closely related to their function. In gray matter areas, the process network is extensive and radial and reXects the particular spatial arrangement of groups or layers of neuronal cell bodies (Dawson et al., 2002). In white matter tracts, the cell bodies of OPCs are found among the rows of interfascicular oligodendrocytes in a regular pattern and extend processes both perpendicular and parallel to the bundles of axons (Berry et al., 2002; Butt et al., 1999). Within any one gray or white matter area there does not appear to be any morphological heterogeneity in the NG2þ OPC population. Any heterogeneity between CNS regions is likely to be a consequence of diVerences in the cytoarchitecture that determines the number and course of the processes. NG2þ and PDGF-aRþ OPCs are found in abundance in both gray and white matter regions throughout the entire CNS comprising between 3 to 9% of all cells depending on the region (Dawson et al., 2002). The gray:white matter ratio in OPC cell number of approximately 1:1.5 is surprising and demonstrates that the numerical density of these cells is not simply a reXection of the abundance of myelin. These results also reinforce the idea that they are likely to play other roles in the CNS, in addition to providing a reserve cell population for both physiological and pathological requirements for new myelin synthesis. When considering the role of these cells in remyelination, of particular importance is the consistent ratio of OPCs to oligodendrocytes of approximately 1:4 across the spinal cord in both gray and white matter, but only 1:1 in the cerebral cortex and hippocampus (Dawson et al., 2002), which has signiWcance for the eYciency and speed of remyelination in these diVerent areas.

301

302

11. OLIGODENDROGLIAL LINEAGE

The Origin of Oligodendrocyte Progenitors in the Adult CNS It is a matter of much debate why there should be such a large population of cells in the adult gray matter with the potential to become oligodendrocytes. The answer to this conundrum may lie in the developmental origin of these cells. During embryonic development in the rodent, PDGF-aRþNG2þ OPCs migrate out from the germinal zones and become evenly distributed in both prospective white and gray matter areas prior to any myelination (Nishiyama et al., 1996; Pringle et al., 1992). As myelination begins in the cerebral cortex, a wave of diVerentiation of OPCs can be observed spreading from the corpus callosum out toward the pial surface (Reynolds and Hardy, 1997). However, as this wave of diVerentiation occurs a proportion of the O4þNG2þ cells do not begin to express GalC and myelin proteins and remain among the immature oligodendrocytes (Reynolds and Hardy, 1997). The same pattern of diVerentiation can be seen in the optic nerve. Therefore, a population of immature O4þNG2þ cells would be expected to remain in all areas of the CNS irrespective of the density of myelinated Wbers. This may in fact explain the 1:1 ratio of OPCs to oligodendrocytes in most gray matter areas, because a smaller proportion of cells would be induced to diVerentiate into oligodendrocytes compared with the white matter. The lower degree of expansion of OPCs in the gray matter in response to growth factors would account for the overall lower numbers of oligodendroglial lineage cells. It is a common misconception that because PDGF-aRþNG2þ cells have the capability of becoming oligodendrocytes, then there should be more of them in the white matter than in the gray matter. Wren et al. (1992) used time-lapse microscopy to demonstrate that perinatal OPCs isolated from the rat optic nerve can give rise to adult progenitors over two or more cell divisions. The results suggest that both perinatal and adult type OPCs would be expected to coexist during the later stages of development and that the proportion of adult OPCs would increase as myelinogenesis proceeds. This is in agreement with studies on the optic nerve and cerebral cortex in vivo (Fulton et al., 1992; Reynolds and Hardy, 1997). More recent investigations, using transgenic mice in which EGFP expression is driven by the PLP gene promoter, have demonstrated two populations of NG2þ OPCs in the cerebral cortex from as early as postnatal day 1 (Mallon et al., 2002). One population expresses both NG2 and PLP gene promotor activity and is suggested to give rise to myelinating oligodendrocytes while the other population does not express PLP gene promoter activity and is suggested to persist into adulthood as a population of progenitor cells. The question of whether these two NG2þ populations represent two distinct lineages that diverge during early development of the oligodendrocyte lineage or whether the two populations represent two diVerent stages of the diVerentiation schedule of a single lineage remains to be answered. It is possible that the NG2þEGFPþ cells give rise to the NG2þEGFP
cells during postnatal development as suggested by the in vitro experiments of Wren et al. (1992). The two populations of cells described in the transgenic mice have similar morphology, distribution, and proliferative characteristics, and therefore it is likely that they are closely related. The preceding experiments raise the question of whether two populations of NG2þ OPCs continue to be present in the adult CNS, and therefore adult OPCs are a heterogeneous population. Evidence to date from studies conducted in vivo suggest that the NG2þPDGFaRþ population of cells is homogeneous with respect to immunological phenotype and radiation sensitivity (Dawson et al., 2000; Li et al., 2002), and it remains to be seen whether this extends to the expression of PLP gene promoter activity (Mallon et al., 2002). Although in vitro experiments suggest that there are several populations of OPCs in the adult brain (Gensert and Goldman, 2001; Mason and Goldman, 2002), it is diYcult to reconcile these data with the in vivo results. The majority of in vitro experiments have been conducted under conditions in which the immunological phenotype of adult OPCs in vivo is unlikely to have been preserved following isolation, and therefore the results are diYcult to interpret.

The Maintenance of Adult OPCs in an Undifferentiated State Adult OPCs have been suggested to represent a slowly dividing stem cell population in the rodent CNS (Wren et al., 1992), capable of generating oligodendroglial lineage cells

THE OLIGODENDROCYTE LINEAGE IN THE ADULT CNS

throughout life. NG2þ OPCs have been demonstrated to be a cycling population in all areas of the CNS (Dawson et al., 2000; Horner et al., 2000), although their ability to divide asymmetrically in vivo has received little attention and is still a matter of debate in vitro (Shi et al., 1998). It is suggested that OPCs may progressively change their properties during development, lengthening their cell cycle time, possibly as a result of p27 accumulation (Durand et al., 1997; Shi et al., 1998), and eventually attaining the properties of adult OPCs. The availability of neuronally derived growth factors, such as the neuregulins, would be expected to decline as myelination proceeds and might also contribute to slowing the rate of OPC proliferation. Continued production of PDGF by astrocytes might be expected to provide a continuous survival signal and also allow a slow rate of proliferation. The further diVerentiation of adult OPCs following myelination is suggested to be prevented by activation of the Notch-Jagged signaling pathway (Wang et al., 1998), thereby maintaining a pool of undiVerentiated progenitor cells. Notch receptors on OPCs would be activated by the Jagged ligand on nearby oligodendrocytes or at nodes of Ranvier. In addition, myelin has been demonstrated to inhibit the further diVerentiation of OPCs in vitro (Robinson and Miller, 1999). However, experiments on cells isolated from the fully mature CNS have yet to be performed (Wang et al., 1998). It remains to be demonstrated whether Notch becomes asymmetrically distributed following OPC division in the adult CNS, thus allowing the daughter cell to diVerentiate into an oligodendrocyte, while also maintaining OPC number.

A Mitotically Active Population of Cells in the Adult CNS Although adult OPCs divide slowly compared to perinatal cells, there are a signiWcant number of studies demonstrating that they represent the major cycling cell population in the adult CNS. In the rat spinal cord, up to 70% of cells incorporating BrdU after a single injection were found to express NG2 (Horner et al., 2000), whereas in both white and gray matter areas of the forebrain, 70 to 75% of BrdU incorporating cells were NG2þ after a 2-hour pulse, representing 1 to 4% of the OPC population (Fig. 11.4F) (Dawson et al., 2002). Although a number of studies of the characteristics of cycling cells in the adult rat CNS ex vivo have suggested that they do not express NG2, but rather are either O4þ or A2B5þ (Gensert and Goldman, 2001; Mason and Goldman, 2002), these conXicting results are most likely the result of technical diYculties associated with NG2 antigenicity. Six days after a BrdU pulse, labeled CNPþ oligodendrocytes are observed in vivo in both gray and white matter (Dawson et al., 2002; Wu et al., 2001), indicating that cycling progenitors give rise to small numbers of oligodendrocytes throughout adult life, although direct evidence that NG2þPDGF-aRþ OPCs give rise to oligodendrocytes in the normal adult CNS in response to physiological demands for new myelin is still lacking. Retroviral lineage studies also suggest that oligodendrocytes are generated from NG2þ cycling cells in the adult CNS (Levison et al., 1999). It remains possible that new OPCs and oligodendrocytes generated in the adult CNS undergo apoptosis, although TUNEL labeled OPCs or oligodendrocytes are only very rarely seen (Dawson et al., 2002). The recent report of no change in the number of PDGFaR-expressing OPCs in the rat cerebellar peduncles during aging (Sim et al., 2002) also suggests that the newly generated NG2þ cells do not remain as OPCs. Thus, it is likely that a small but consistent number of myelinating oligodendrocytes diVerentiate from non-myelinating NG2þ cycling cells during the course of adult life. However, the possibility that BrdU-labeled oligodendrocytes arise from BrdUþ NG2
cells cannot be ruled out at this stage.

The Function of Oligodendrocyte Progenitors in the Normal Adult CNS The very abundance in the adult white and gray matter of cells with the phenotype of OPCs, and their extensive process network, has been taken as evidence for a role in the adult CNS other than as oligodendrocyte progenitors. The presence of processes of NG2þ cells at the node of Ranvier (Butt et al., 1999) and at synapses (Bergles et al., 2000; Ong and Levine, 1999) is suggestive of a role in the modulation of neuronal activity. The presence of

303

304

11. OLIGODENDROGLIAL LINEAGE

AMPA receptors on OPCs, the surprise observation of glutaminergic synaptic input onto these cells in the hippocampus (Bergles et al., 2000), and their ability to transport glutamate (Domercq et al., 1999; Reynolds and Herschkowitz, 1987), also indicates a role in glutamate neurotransmission and homeostasis. However, it is not known whether this occurs throughout the CNS. The location of OPC processes also places the cells in an ideal position to monitor and respond to electrical activity in both white and gray matter, although little is known concerning the cell-cell contacts that OPCs make in the gray matter, other than at synapses. Thus, it is possible that electrical activity may play an indirect role in cell cycle regulation in OPCs. Therefore, there is accumulating evidence for multiple roles for NG2þ cells in the adult CNS, although this is not incompatible with their role as oligodendrocyte progenitors. There is little doubt that these cells have the capability of diVerentiating into oligodendrocytes in certain circumstances, both physiological and pathological, in response to a demand for new myelin.

The Response of Adult OPCs to Demyelination and Other Pathologies Thus, a population of slowly dividing OPCs is retained in the adult CNS, as a reservoir of cells that have the potential to generate new oligodendrocytes. Changes in the CNS environment following injury or myelin loss appear to enable adult OPCs to reacquire the rapidly dividing phenotype of neonatal OPCs so large numbers of oligodendrocytes can be replaced. Adult OPCs react to a variety of pathological insults to the CNS (for review, see Levine et al., 2001) by up-regulating NG2 expression and a thickening of their primary processes. Processes are then retracted and cell bodies become swollen. In the presence of demyelination (Cenci di Bello et al., 1999; Levine and Reynolds, 1999) or severe tissue damage (Levine, 1994), this is followed by proliferation. A consistent feature of the adult OPC response to demyelination is that it is highly restricted in space and is extremely rapid. Only OPCs in the area of demyelination and immediate border proliferate (Cenci di Bello, 1999; Reynolds et al., 2002; Watanabe et al., 2002). In response to inXammation, induced by passive transfer of activated MBP-speciWc T cells (Cenci di Bello et al., 1999) or herpes virus infection (Levine et al., 1998), OPCs show reactive changes but do not proliferate. These data suggest that OPC proliferation in demyelinated lesions in which OPCs are preserved is brought about in part by signals from the now bare axons, in combination with synergistic signaling from astrocytes and microglia/macrophages (Franklin, 2002). This reactive process appears to generate more than enough cells to carry out remyelination, which is rapid in nearly all animal models of demyelinating disease. The increase in the number of OPCs within demyelinated areas and subsequent decrease on remyelination, and the close physical association between the processes of reactive OPCs and demyelinated axons, strongly suggests adult OPCs are able to revert to a neonatal phenotype in order to undergo rapid proliferation and migration before diVerentiating into myelin forming cells. However, a direct demonstration that a reactive adult OPC can mature into a myelin-forming cell in vivo is still lacking. Destruction of both oligodendrocytes and myelin would be likely to both release OPCs from the block on diVerentiation mediated by Notch-Jagged signaling and also increase the availability of the axonally derived GGF. In vitro, adult OPCs can be stimulated to divide rapidly by a combination of PDGF and GGF (Shi et al., 1998). Therefore, rapid repopulation of demyelinated lesions by OPCs would be expected to occur in the presence of bare axons and reactive astrocytes. Following remyelination, numbers of NG2þ or PDGF-aRþ OPCs in the immediate area of demyelination return to approximately pre-lesion levels (Cenci di Bello et al., 1999) or remained elevated (Levine and Reynolds, 1999). This would be expected if OPCs were acting as a stem cell population, dividing asymmetrically in order to maintain a population of cells for future repair. This is in agreement with in vitro studies showing asymmetrical division of OPCs isolated from adult rat optic nerve (Wren et al., 1992). A small decrease in NG2þ OPC number has been noted at the lesion border following remyelination of lesions induced by the injection of GalC antibodies in the rat spinal cord (Keirstead et al., 1998), although ethidium bromide induced lesions in the same location did not result in depletion of OPCs from the areas surrounding the lesion (Chari and Blakemore, 2002).

THE OLIGODENDROCYTE LINEAGE IN THE ADULT CNS

Conclusions In conclusion, the adult CNS contains a substantial population of slowly dividing cells with the phenotype of oligodendroglial progenitors. These cells are perfectly situated to be able to rapidly respond to a variety of insults, and all the experimental evidence produced using animal models of demyelinating disease is highly suggestive of NG2þPDGF-aRþ OPCs as the source of new oligodendrocytes. Their number and location in the resting CNS is also suggestive of a number of additional roles that have yet to be explored in detail.

Acknowledgments Work described in this review carried out by the authors was funded by the Wellcome Trust, the U.K. and U.S. Multiple Sclerosis Societies, and the National Institutes of Health (NINDS) Grants # NS3800 and NS. We thank Dr. Mary Dawson for supplying mages of NG2-expressing cells.

References

au16

Armstrong, R., L. Harvath, et al. (1991). Astrocytes and O-2A progenitors migrate toward distinct molecules in a microchemotaxis chamber. Ann NY Acad Sci 633(520), 520–522. Armstrong, R. C., L. Harvath, et al. (1990). Type 1 astrocytes and oligodendrocyte-type 2 astrocyte glial progenitors migrate toward distinct molecules. J Neurosci Res 27(3), 400–407. Bansal, R., M. Kumar, et al. (1996). Regulation of FGF receptors in the oligodendrocyte lineage. Molecular and Cellular Neuroscience 7, 263–275. Bansal, R., and S. E. PfeiVer (1992). Novel stage in the oligodendrocyte lineage deWned by reactivity of progenitors with R-mAb prior to O1 anti-galactocerebroside. J Neurosci Res 32(3), 309–316. Bansal, R., and S. E. PfeiVer (1997). FGF2 converts mature oligodendrocytes to a novel phenotype. Journal of neuroscience reserach 50, 215–228. Bansal, R., K. Stefansson, et al. (1992). Proligodendroblast antigen (POA), a developmental antigen expressed by A007/O4-positive oligodendrocyte progenitors prior to the appearance of sulfatide and galactocerebroside. J Neurochem 58(6), 2221–2229. Barres, B. A., I. K. Hart, et al. (1992). Cell death and control of cell survival in the oligodendrocyte lineage. Cell 70(1), 31– 46. Barres, B. A., M. A. Lazar, et al. (1994a). A novel role for thyroid hormone, glucocorticoids and retinoic acid in timing oligodendrocyte development. Development 120(5), 1097–1108. Barres, B. A., and M. C. RaV (1994). Control of oligodendrocyte number in the developing rat optic nerve. Neuron 12(5), 935–942. Barres, B. A., M. C. RaV, et al. (1994b). A crucial role for neurotrophin-3 in oligodendrocyte development. Nature 367(6461), 371–375. Barres, B. A., R. Schmid, et al. (1993). Multiple extracellular signals are required for long-term oligodendrocyte survival. Development 118, 283–295. Bergles, D. W., Roberts, R. D. B., Somogyi, P., and Jahr, C. E. (2000). Glutamatergic synapses on oligodendrocyte precursor cells in the hippocampus. Nature 405, 187–191. Berry, M., Butt, A. M., Wilkin, G., and Perry, V. H. (2002). Structure and function of glia in the central nervous system. In GreenWeld’s Neuropathology (D. I. Graham and P. L. Lantos, eds.), 7th ed., pp. 75–121. Arnold, London. Bogler, O., D. Wren, et al. (1990). Cooperation between two growth factors promotes extended self-renewal and inhibits diVerentiation of oligodendrocyte-type-2 astrocyte (O-2A) progenitor cells. Proc Natl Acad Sci USA 87(16), 6368–6372. Briscoe, J., Y. Chen, et al. (2001). Molecular Cell 7, 1279–1291. Burne, J. F., J. K. Staple, et al. (1996). Glial cells are increased proportionally in trangenic optic nerve with increased number of axons. Journal of Neuroscience 16, 2064–2073. Butt, A. M., Duncan, A., Hornby, M. F., Kirvell, S. L., Hunter, A., Levine, J. M., Berry, M. (1999). Cells expressing the NG2 antigen contact Nodes of Ranvier in adult CNS white matter. Glia 26, 84–91. Calver, A. R., A. C. Hall, et al. (1998). Oligodendrocyte population dynamics and the role of PDGF in vivo. Neuron 20, 869–882. Casaccia-BonneWl, P. (2000). Cell death in the oligodendrocyte lineage: A molecular perspective of life/death decisions in development and disease. Glia 29, 124 –135. Casaccia-BonneWl, P., R. Tikoo, et al. (1997). Oligodendrocyte precursor diVerentiation is perturbed in the absence of the cyclin-dependent kinase inhibitor p27 Kip-1. Genes and Development 11, 2335–2346.

305

306

11. OLIGODENDROGLIAL LINEAGE

au17

Cenci di Bello, I., Dawson, M. R. L., Levine, J. M., and Reynolds, R. (1999). Generation of oligodendroglial progenitors in acute inXammatory demyelinating lesions of the rat brain stem is stimulated by demyelination rather than inXammation. J Neurocytol 28, 366–381. Chari, D. M., and Blakemore, W. F. (2002). EYcient recolonisation of progenitor depleted areas of the CNS by adult oligodendrocyte progenitor cells. Glia 37, 307–313. Davies, J. E., and R. H. Miller (2001). Local sonic hedgehog signaling regulates oligodendrocyte precursor appearance in multiple ventricular domains in the chick metencephalon. Developmental Biology 233, 5130525. Dawson, M. R. L., Levine, J. M., and Reynolds, R. (2000). NG2-expressing cells in the central nervous system: Are they oligodendroglial progenitors? J Neurosci Res 61, 471–479. Dawson, M. R. L., Polito, A., Levine, J. M., and Reynolds, R. (2003). NG2-expressing glial progenitor cells: An abundant and widespread population of cycling cells in the adult rat CNS. Mol. Cell. Neurosci. 24, 476–488. Domercq, M., Sanchez-Gomez, M. V., Areso, P., and Matute, C. (1999). Expression of glutamate transporters in rat optic nerve oligodendrocytes. Eur J Neurosci 11, 2226–2236. Durand, B., Gao, F. B., and RaV, M. C. (1997). Accumulation of the cyclin-dependent kinase inhibitor p27/Kip1 and the timing of oligodendrocyte diVerentiation. EMBO J 16, 306–317. Echelard, Y., D. J. Epstein, et al. (1993). Sonic hedgehog, a member of a family of putative signaling molecules is implicated in the regulation of CNS polarity. Cell 75, 1417–1430. Vrench-Constant, C., R. H. Miller, et al. (1988). Evidence that migratory oligodendrocyte-type-2 astrocyte (O-2A) progenitor cells are kept out of the rat retina by a barrier at the eye-end of the optic nerve. Journal of Neurocytology 17(1), 13–25. Fok-Seang, J., and R. H. Miller (1994). Distribution and diVerentiation of A2B5þ glial precursors in the developing rat spinal cord. J. Neurosci. Res. 37, 219–235. Franklin, R. J. M. (2002). Why does remyelination fail in multiple sclerosis? Nat Neurosci Rev 3, 705–714. Frost, E., R. Milner, et al. (2000). Migration assays for oligodendrocyte precursor cells. Methods in Molecular Biology 139, 265–278. Fulton, B. P., Burne, J. F., RaV, and M. C. (1992). Visualisation of O-2A progenitor cells in developing and adult rat optic nerve by quisqualate-stimulated cobalt uptake. J. Neurosci 12, 4816–4833. Gard, A. L., and S. E. PfeiVer (1993). Glial cell mitogens bFGF and PDGF diVerentially regulate development of O4þ GalC-oligodendrocyte progenitors. Developmental Biology 159, 618–630. Gensert, J. M., and Goldman, J. E. (2001). Heterogeneity of cycling glial progenitors in the adult mammalian cortex and white matter. J Neurobiol 48, 75–86. Grinspan, J., J. L. Stern, et al. (1993). Trophic eVect of basic Wbroblast growth factor (bFGF) on diVerentiated oligodendroglia: A mechanism for regeneration of the oligodendroglial lineage. Journal of Neuroscience Research 36, 672–690. He, W., C. Ingraham, et al. (2001). Multipotent stem cells from the basal forebrain contribute GABAergic neurons and oligodendrocytes to the cerebral cortex during development. Journal of Neuroscience 21, 8854–8863. Hengst, L., and S. I. Reed (1996). Translational control of p27 Kip-1 accumulation during the cell cycle. Science 271, 1861–1864. Horner, P. J., Power, A. E., Kempermann, G., Kuhn, H. G., Palmer, T. D., Winkler, J., Thal, L. J., and Gage, F. H. (2000). Proliferation and diVerentiation of progenitor cells throughout the intact adult rat spinal cord. J Neurosci 20, 2218–2228. Hughes, S. M., L. E. Lillien, et al. (1988). Ciliary neurotrophic factor induces type-2 astrocyte diVerentiation in culture. Nature 335(6185), 70–73. Incardona, J. P., W. Gassield, et al. (1998). The teratogenic veratrum alkaliod cyclopamine inhibits sonic hedgehog signal transduction. Development 125, 3353–3562. Jessell, T. M. (2000). Nature Review Genetics 1, 20–29. Jessell, T. M., and J. Dodd (1990). Floor plate-derived signals and the control of neural cell pattern in vertebrates. Harvey Lect 86(87), 87–128. Kato, A., H. Takahashi, et al. (1997). Inactivation of the cyclin D-dependent kinase in the rat Wbroblast cell line 3Y1, induced by contact inhibition,. J. Biol. Chem. 272, 8065–8070. Keirstead, H. S., Levine, J. M., and Blakemore, W. F. (1998). Response of oligodendrocyte progenitor cell population (deWned by NG2 labeling) to demyelination of the adult spinal cord. Glia 22, 161–170. Kessaris, N., N., and Pringle, et al. (2001). Ventral Neurogenesis and the Neuron-glial switch. Neuron 31, 677–680. Kondo, T., and M. C. RaV (2000). Oligodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells/. Science 289, 1754. Leber, S. M., S. M. Breedlove, et al. (1990). Lineage, arrangement, and death of clonally related motoneurons in chick spinal cord. J Neurosci 10(7), 2451–2462. Leber, S. M., and J. R. Sanes (1991). Lineage analysis with a recombinant retrovirus: Application to chick spinal motor neurons. Adv Neurol 56(27), 27–36. Lemke, G. (1993). The molecular genetics of myeliantion: An update. Glia 7, 263–271. Lemke, G. (1996). Neuregulins in development. Molecular and Cellular Neuroscience 7, 247–262. Levine, J. M. (1994). Increased expression of the NG2 chondroitin-sulphate proteoglycan after brain injury. J Neurosci 14, 4716–4730.

THE OLIGODENDROCYTE LINEAGE IN THE ADULT CNS

Levine, J. M., and Card, J. P. (1987). Light and electron microscope localisation of a cell surface antigen (NG2) in the rat cerebellum: Association with smooth protoplasmic astrocytes. J Neurosci 7, 2711–2720. Levine, J. M., Enquist, L. W., and Card, J. P. (1998). Activation of oligodendrocyte precursor cells in the neuropathological responses to alpha herpes virus infection of the central nervous system. Glia 23, 316–328. Levine, J. M., and Reynolds, R. (1999). Activation and proliferation of endogenous oligodendrocyte precursor cells during ethidium bromide-induced demyelination. Exp Neurol 160, 333–347. Levine, J. M., Reynolds, R., and Fawcett, J. W. (2001), The oligodendrocyte precursor cell in health and disease. Trends in Neurosciences 24, 39–47. Levine, J. M., Stincone, F., and Lee, Y-S. (1993). Development and diVerentiation of glial precursor cells in the rat cerebellum. Glia 7, 307–321. Levison, S. W., Young, G. M., and Goldman. J,E, (1999). Cycling cells in the adult rat neocortex preferentially generate oligodendroglia. J Neurosci Res 57, 435–446. Li, G., Crang, A. J., Rundle, J. L., Blakemore, W. F. (2002). Oligodendrocyte progenitor cells in the adult rat CNS express myelin oligodendrocyte glycoprotein (MOG). Brain Pathol 12, 463–471. Lillien, L. E., M. Sendtner, et al. (1990). Extracellular matrix-associated molecules collaborate with ciliary neurotrophic factor to induce type-2 astrocyte development. J Cell Biol 111(2), 635–44. Louis, J. C., D. Muir, et al. (1992). Autocrine Inhibition of Mitotic Activity in Cultured Oligodendrocyte-Type-2 Astrocyte (O-2A) Precurosr cells. Glia 6, 30–38. Lu, Q. R., T. Sun, et al. (2002). Common developmental requirement for oligfunction indicates a motor neuron/ oligodendrocyte connection. Cell 109, 75–86. Lu, Q. R., D. Yuk, et al. (2000). Sonic hedgehog regulated oligodendrocyte lineage genes encoding bHLH proteins in the mammalian central nervous system. Neuron 25, 317–329. Mabie, P. C., M. F. Mehler, et al. (1997). Bone morphogenitic proteins induce astroglial diVerentiation of oligodendroglial-astroglial progenitor cells. Journal of Neuroscience 17, 4112–4120. Maier, C. E., and R. H. Miller (1997). Notochord is essential for oligodendrocyte development in Xenopus spinal cord. Journal of Neuroscience Research 47, 361–371. Mallon, B. S., Shick, H. E., Kidd, G. J., and Macklin, W. B .(2002). Proteolipid promoter activity distinguishes two populations of NG2-positive cells throughout neonatal cortical development. J Neurosci 22, 876–885. Marchionni, M. A., A. D. J. Goodearl, et al. (1993). Glial growth factors are alternatively spliced erbB2 ligands expressed in the nervous system. Nature 362, 312–317. Mason, J. L., and Goldman, J. E. (2002). A2B5þ and O4þ cycling progenitors in the adult forebrain white matter respond diVerentially to PDGF-AA, FGF-2, and IGF-1. Mol Cell Neurosci 20, 30–42. Mori, S., and Leblond, C. P. (1970). Electron microscopic identiWcation of three classes of oligodendrocytes and a preliminary study of their proliferative activity in the corpus callosum of young rats. J Comp Neurol 139, 1–30. Mayer, M., O. Bogler, et al. (1993). The inhibition of oligodendrocyte diVerentiaion of O-2A progenitors caused by bFGF is overridden by astrocytes. Glia 8, 12–19. McKinnon, R. D., T. Matsui, et al. (1990). FGF modulates the PDGF-driven pathway of oligodendrocyte development. Neuron 5(5), 603–614. McKinnon, R. D., G. Piras, et al. (1993). A Role for TGF-B in Oligodendrocyte DiVerentation. Journal of Cell Biology 121(6), 1397–1407. Mehler, M.,.F., P. C. Mabie, et al. (2000). Developmental changes in progenitor cell responsivness to bone morphogenetic proteins diVerentially modulare progressive CNS lineage fate. Dev. Neuroscience 22, 74–85. Mekki-Dauriac, S., E. Agius, et al. (2002). Bone morphognenitic proteins negatively control oligodendrocyte precursor speciWcation in the chick spinal cord. Development 129, 5117–5130. Nakatsujji, Y., and R. H. Miller (2001). Control of oligodendrocyte precursor proliferation mediated by densitydependent cell cycle protein expression. Developmental Neuroscience 23, 356–363. Nery, S., H. Wichterle, et al. (2001). Sonic hedghog contributes to oligodendrocyte speciWcation in the mammalian forebrain. Development 128, 527–540. Nishiyama, A., X.-H. Lin, et al. (1996). Interaction between NG2 proteoglycan and PDGF aReceptor on O2A progenitor cells is required for optimal response to PDGF. Journal of Neuroscience Research 43, 315–330. Nishiyama, A., Yu, M., Drazba, J. A., and Tuohy, V. K. (1997). Normal and reactive NG2þ glial cells are distinct from resting and activated microglia. J Neurosci Res 48, 299–312. Noble, M., K. Murray, et al. (1988). Platelet-derived growth factor promotes division and motility and inhibits premature diVerentiation of the oligodendrocyte/type-2 astrocyte progenitor cell. Nature 333(6173), 560–562. Noll, E., and R. H. Miller (1993). Oligodendrocyte precursors originate at the ventral ventricular zone dorsal to the ventral midline in the embryonic rat spinal cord. Development 118, 563–573. Noll, E., and R. H. Miller (1994). Regulation of oligodendrocyte diVerentiation: A role for retinoic acid in the spinal cord. Development 120, 649–660. Olivier, C., I. Cobos, et al. (2001). Monofocal origin of telencephalic oligodedrocytes in the anterior entopenduncular ares of the chick embryo. Development 128, 1757–1769. Ong, W. Y., and Levine, J. M. (1999). A light and electronmicroscope study of NG2 chondroitin sulphate proteoglycan-positive oligodendrocyte precursor cells in the normal and kainate lesioned rat hippocampus. Neuroscience 92, 83–95. Ono, K., R. Bansal, et al. (1995). Early development and dispersal of oligodendrocyte precursors in the embyonic chick spinal cord. Development 121, 1743–1754. Ono, K., T. Tsumori, et al. (1998). Developmental appearance of oligodendrocytes in the embryonic chick retina. Journal of comparative Neurology 398, 309–322.

307

308

11. OLIGODENDROGLIAL LINEAGE

au18

au19

Ono, K., T. Tsumori, et al. (2001). Extensive proliferation of oligodendrocyte precursors in the parenchyma of the embryonic chick central nervous system. Developmental Biology 231, 77–86. Ono, K., Y. Yasui, et al. (1997). Focal ventricular origin and migration of oligodendrocyte precursors into the chick optic nerve. Neuron 19, 1–20. Orentas, D. M., and R. H. Miller (1996a). A novel form of migration of glial precursors. Glia 16, 27–39. Orentas, D. M., and R. H. Miller (1996b). The origin of spinal cord oligodendrocytes is dependent on local inXuences from the notochord. Developmental Biology 177, 43–53. Park, S. K., R. H. Miller, et al. (2001). The erbB2 gene is required for the development of terminally diVerentiated spinal cord oligodendrocyts. Journal of Cell Biology. 154, 1245–1258. Payne, H. R., and V. Lemmon (1993). Glial cells of O-2A lineage bind preferentially to N-cadherin and develop distinct morphologies. Developmental Biology 159, 595–607. PfeiVer, S. E., A. E. Warrington, et al. (1993). The oligodendrocyte and its many cellular processes. Trends in cell Biology 3, 191–197. Poncet, C., C. Soula, et al. (1996). Induction of oligodendrocyte progenitors in the trunk neural tube by ventralizing signals: EVects of notochord and Xoor plate grafts and sonic hedgehog. Mechanisms of Development 60, 13–32. Pringle, N. P., H. S. Mudhar, et al. (1992). PDGF receptors in the rat CNS: During late neurogenesis, PDGF alpha-receptor expression appears to be restricted to glial cells of the oligodendrocyte lineage. Development 115(2), 535–551. Pringle, N. P., and W. D. Richardson (1993). A singularity of PDGF alpha-receptor expression in the dorsoventral axis of the neural tube may deWne the origin of the oligodendrocyte lineage. Development 117(2), 525–533. Pringle, N. P., W. Yu, et al. (1996). Determination of neuroepithelial cell fate: Induction of the oligodendrocyte lineage by ventral midline cells and sonic hedgehog. Developmental Biology 177, 30–42. RaV, M. C., E. R. Abney, et al. (1983). Two types of astrocytes in cultures of developing rat white matter: DiVerences in morphology, surface gangliosidase, and growth characteristics. J Neurosci 3(6), 1289–1300. RaV, M. C., E. R. Abney, et al. (1985). Reconstitution of a developmental clock in vitro: A critical role for astrocytes in the timing of oligodendrocyte diVerentiation. Cell 42(1), 61–69. RaV, M. C., C. C. Ffrench, et al. (1987). Glial cells in the rat optic nerve and some thoughts on remyelination in the mammalian CNS. J Exp Biol 132(35), 35–41. RaV, M. C., and L. E. Lillien (1988). DiVerentiation of a bipotential glial progenitor cell: What controls the timing and the choice of developmental pathway? J Cell Sci Suppl 10(77), 77–83. RaV, M. C., R. H. Miller, et al. (1983). A glial progenitor cell that develops in vitro into an astrocyte or an oligodendrocyte depending on culture medium. Nature 303(5916), 390–396. RaV, M. C., R. Mirsky, et al. (1978). Galactocerebroside is a speciWc cell surface antigenic marker for oligodendrocytes in culture. Nature 274, 813–816. RaV, M. C., B. P. Williams, et al. (1984). The in vitro diVerentiation of a bipotential glial progenitor cell. EMBO J 3(8), 1857–1864. Rao, M., and M. Mayer-Proschel (1997). Glial restricted precursors are derive from multipotent neuroepithelial stem cells. Developmental Biology 188, 48–63. Rao, M., M. Noble, et al. (1998). A tripotential glial precursor cell is present in the developing spinal cord. Proc. Nat. Acad. Sci. 95, 3996–4001. Reynersm H., Gianfelici de Reynersm E., Maisinm J. R. (1982). The beta astrocyte: A newly recognised radiosensitive glial cell type in the cerebral cortex. J Neurocytol 11, 967–983. Reynersm H., Gianfelici de Reynersm E., Regniersm L., and Maisin, J. R. (1986). A glial progenitor cell in the cerebral cortex of the adult rat. J. Neurocytol. 15, 53–61. Reynolds, R., Cenci di Bello, I., Dawson, M., and Levine, J. M. (2001). The response of adult oligodendrocyte progenitors to demyelination in EAE. Prog. Brain Research 132, 165–174. Reynolds, R., Dawson ,M., Polito, A., Cenci di Bello, I., Papadopoulos, D., Pham-Dinh, D., and Levine, J. (2002). The response of NG2-expressing oligodendrocyte progenitors to demyelination in MOG-EAE and MS. J Neurocytol., 31, 523–536. Reynolds, R., and Hardy, R. (1997). Oligodendroglial progenitors labeled with the O4 antibody persist in the adult rat cerebral cortex in vivo. J Neurosci Res 47, 455–470. Reynolds, R., and Herschkowitz, N. (1987). Oligodendroglial and astroglial heterogeneity in mouse primary central nervous system culture as demonstrated by diVerences in GABA and D-aspartate transport and Immunocytochemistry. Brain Res 433, 13–25. Richardson, W. D., N. Pringle, et al. (1988). A role for platelet-derived growth factor in normal gliogenesis in the central nervous system. Cell 53(2), 309–319. Richardson, W. D., N. P. Pringle, et al. (1997). Origin of spinal cord oligodendrocytes: Possible developmental and evolutionary relationships with motor neurons. Developmental Neuroscience 19: 58–68. Robinson, S., and Miller, R. H. (1999). Contact with central nervous system myelin inhibits oligodendrocyte progenitor maturation. Dev. Biol. 216, 359–368. Robinson, S., M. Tani, et al. (1998). The chemokine growth related oncogene-a promotes spinal cord precursor proliferation. Journal of Neuroscience 18, 10457–10463. Roelink, H., A. Ausgsburger, et al. (1994). Floor plate and motor neuron induction by vhh-1, a vertebrate homolog of hedgehog expressed by the notochord. Cell 76, 761–775. Rowitch D, Lu et al. (2002). An oligarchy rules neural development. TINS 25, 417–422.

THE OLIGODENDROCYTE LINEAGE IN THE ADULT CNS

au20

Shi, J., Marinovich, A., and Barres, B. A. (1998). PuriWcation and characterisation of adult oligodendrocyte precursor cells from the rat optic nerve. J Neurosci 18, 4627–4636. Sim, F. J., Zhao, C., Penderis, J., and Franklin, R. J. M. (2002). The age-related decrease in CNS remyelination eYciency is attributable to an impairment of both oligodendrocyte progenitor recruitment and diVerentiation. J Neurosci 22, 2451–2459. Small, R. K., P. Riddle, et al. (1987). Evidence for migration of oligodendrocyte–type-2 astrocyte progenitor cells into the developing rat optic nerve. Nature 328(6126), 155–157. Smart, I., and Leblond, C. P. (1961). Evidence for division and transformations of neuroglia cells in the mouse brain, as derived from radioautography after injection of thymidine. J Comp Neurol 116, 349–368. Smith, J., E. Ladi, et al. (2000). Redox state is a central modulator of the balance between self-renewal and diVerentiation in a dividing glial precursor cell. Proc. Nat. Acad. Sci. 97, 10032–10037. Sommer, I., and M. Schachner (1981). Monoclonal antibodies (O1 to O4) to oligodendrocyte cell surfaces: An immunocytological study in the central nervous system. Dev Biol 83(2), 311–327. Spassky, N., C. Goujet-Zalc, et al. (1998). Multiple restricted origin of oligodendrocytes. Journal of Neuroscience 18(20), 8331–8343. Spassky, N., Heydon, et al. (2002). Sonic hedgehog-dependent emergence of oligodendrocytes in the telecephalon: Evidence for a source of oligodendrocytes in the olfactory bulb that is independent of PDGFRalpha signaling. Development 128, 4993–5004. Spassky, N., C. Olivier, et al. (2000). Single or multiple oligodendrocyte lineages: A controversy. Glia 29, 143–148. Sugimoto, Y., M. Taniguchi, et al. (2001). Guidance of glial pecursor cell migration by secreted cues in the developing optic nerve. Development 128, 3321–3330. Sun, T., N. P. Pringle, et al. (1998). Pax6 InXuences the Time and Site of Origin of Glial Precursors in the Ventral Neural Tube. Molecular and Cellular Neuroscience 12, 228–239. Sussman, C. R., K. L. Dyer, et al. (2000). Local control of oligodendrocyte developmental potential in the dorsal mouse spinal cord. J. Neurosci. Res. 59, 413–420. Tekki-Kessaris, N., R. WoodruV, et al. (2001). Hedgehog-dependent oligodendrocyte speciWcation in the telencephalon. Development 128, 2545–2554. Temple, S., and M. C. RaV (1985). DiVerentiation of a bipotential glial progenitor cell in a single cell microculture. Nature 313(5999), 223–225. Temple, S., and M. C. RaV (1986). Clonal analysis of oligodendrocyte development in culture: Evidence for a developmental clock that counts cell divisions. Cell 44(5), 773–779. Timsit, S., S. Martinez, et al. (1995). Oligodendrocytes originate in a restricted zone of the embryonic ventral neural tube deWned by DM-20 mRNA expression. Journal of Neuroscience 15, 1012–1024. Trapp, B. D. (1990). Myelin-associated glycoprotein: Location and potential functions. Ann NY Acad Sci 605(29), 29–43. Trousse, F. Giess, et al. ((1995). Notochord and Xoor plate stimulate oligodendrocyte diVerentiation in cultures of dorsal neural tube. J. Neurosci. Res. 41, 522–560. Tsai, H., E. Frost, et al. (2000). The inhibitory eVect on oligodendrocyte precursor migration by growth regulated oncogene-a (GRO-a) is mediated by CXCR2. J. Neurosci. Abs 26 part 2, 506.16. van Straaten, H. W., J. W. Hekking, et al. (1988). EVect of the notochord on the diVerentiation of a Xoor plate area in the neural tube of the chick embryo. Anat Embryol Berl 177(4), 317–324. van Straaten, H. W., J. W. Hekking, et al. (1989). EVect of the notochord on proliferation and diVerentiation in the neural tube of the chick embryo. Development 107(4), 793–803. Vartanian, T., G. Corfas, et al. (1994). A role for the acetylcholine receptor-inducing protein ARIA in oligodendrocyte development. Proc. Natl. Acad. Sci. 91, 11626–11630. Vaughn, J. E., and Peters, A. (1968). A third neuroglial cell type: An electron microscope study. J Comp Neurol 133, 269–288. Wada, T., T. Kagawa, et al. (2000). Dorsal spinal cord inhibits oligodendrocyte development. Developmental Biology 227, 42–55. Wang S, Sdrulla AD, diSibio G, Bush G, Nofziger D, Hicks C, Weinmaster, G., Barres, B. (1998). Notch receptor activation Inhibits oligodendrocyte diVerentiation. Neuron 21, 63–75. Warf, B. C., J. Fok-Seang, et al. (1991). Evidence for the ventral origin of oligodendrocyte precursors in the rat spinal cord. J Neurosci 11(8), 2477–2488. Warrington, A. E., E. Barbarese, et al. (1993). DiVerential myelinogenic capacity of speciWc stages of the oligodendrocyte upon transplantation into hypomyelinating hosts. Journal of Neuroscience Research 34, 1–13. Watanabe, M., Toyama, Y., and Nishiyama, A. (2002). DiVerentiation of proliferated NG2-positive glial progenitor cells in a remyelinating lesion. J Neurosci Res 69, 826–836. Wieser, R. J., D. Renauer, et al. (1990). Growth control in mammalian cells by cell-cell contacts. Environmental Health Perspectives 88, 251–253. Williams, B. P., J. Read, et al. (1991). The generation of neurons and oligodendrocytes from a common precursor cell. Neuron 7(4), 685–693. Wolswijk, G., and Noble, M. (1989). IdentiWcation of an adult-speciWc glial progenitor cell. Development 105, 387–400. Wren D., Wolswijk, G., Noble, M. (1992). In vitro analysis of the origin and maintenance of O-2Aadult progenitor cells. J Cell Biol 116, 167–176.

309

310

11. OLIGODENDROGLIAL LINEAGE

Wu, H. Y., Dawson ,M. R. L., Reynolds, R., Hardy, R. J. (2001). Expression of QKI proteins and MAP1B identiWes actively myelinating oligodendrocytes in adult rat brain. Mol Cell Neurosci 17, 292–302. Wu, Q., R. H. Miller, et al. (2000). Elevated levels of the chemokine GRO-1 correlate with elevated oligodendrocyte progenitor proliferation in the jimpy mutant. Journal of Neuroscience 20, 2609–2617. Yu, W. P., E. J. Collarini, et al. (1994). Embryonic expression of myelin genes: Evidence for a focal source of oligodendrocyte precursors in the ventricular zone of the neural tube. Neuron 12(6), 1353–1362. Zezula, J., P. Casaccia-BonneWl, et al. (2001). p21cip1 is required for the diVerentiation of olgiodednrocytes independnetly of cell cycle withdrawal. EMBO Rep. 2, 27–34. Zhang, H., and R. H. Miller (1996). Density-dependent feedback inhibition of oligodendrocyte precursor expansion. Journal of Neuroscience 16, 6886–6895. Zhou, Q., and D. J. Anderson (2002). The bHLH transcription factors Olig2 and Olig1 couple neuronal an glial subtypes speciWcation. Cell 109, 61–73. Zhou, Q., G. Choi, et al. (2001). The bHLH transcription factor Olig2 promotes oligodendrocyte diVerentiation in collaboration with Nkx2.2. Neuron 31, 791–807.

C H A P T E R

12 Astrocyte Lineage James E. Goldman

INTRODUCTION First described at the turn of the 20th century (Andriezen, 1893; Cajal, 1995), astrocytes, the star-shaped glia that populate the CNS, play many important roles in brain development and function. These include regulating extracellular ion and neurotransmitter concentrations, metabolizing neurotransmitters, maintaining the blood-brain-barrier, providing trophic support for neurons and oligodendrocytes, and guiding migrating neurons and glia during CNS development and perhaps serving stem-like or neurogenic functions in the developing and adult brain. To assert their many and varied eVects, astrocytes are positioned carefully within the CNS with respect to their interactions with blood vessels, the pial surface, neurons, and other astrocytes (Figure 12.1). A detailed understanding of astrocyte development must explain how these cells come to reside throughout the brain and how they form speciWc interactions with other cell types. Although our current knowledge of astrocyte development does not clarify exactly how astrocytes come to be organized within the CNS, we do know something about the origins of astrocytes, the paths over which astrocyte progenitors migrate to reach their Wnal destinations, and the molecular signals that act on immature neuroectodermal cells to induce astrocyte genes. Astrocytes are not arranged in random fashion throughout the CNS, but rather form a ‘‘matrix’’ in which cell bodies are separated in space (Chan-Ling and Stone, 1991; Levison and Goldman, 1993; Tout et al., 1993; and most recently and clearly in Bushong et al., 2002). Thus, astrocytes establish their own domains, with a modest degree of overlap between adjacent cells (Bushong et al., 2002). The overlap zones contain contacts between neighboring astrocytes, mediated by Cx43 gap junctions, which allow passage of calcium and other small molecules (reviewed in Ransom, 1995). The interactions of astrocytes with other cell types are critical to astrocyte functions. Thus, the basal laminae of blood vessels and the pial surfaces of the CNS are covered by ends of astrocyte processes (Peters et al., 1991). It is this interaction with endothelial cells that plays a major role in establishing the blood brain barrier (see Abbot, 2002, for recent review). Astrocytes invest neuronal cell bodies and dendrites, the Wne glial processes intimately enfolding individual synaptic endings (Hama et al., 1993; Kosaka and Hama, 1986; Ventura and Harris, 1999). The formation of these various types of cell contacts must be accounted for in a developmental model for astrocyte genesis.

IS ASTROCYTE DEVELOPMENT IDENTICAL IN DIFFERENT REGIONS OF THE CNS? No, this is unlikely to be the case. In particular, there appear to be two distinct pathways of astrocyte development. In the Wrst, astrocytes arise from radial glia. This pathway likely

Myelin Biology and Disorders, Volume 1

311

Copyright 2004, Elsevier (USA). All rights reserved.

312

12. ASTROCYTE LINEAGE

FIGURE 12.1 There are a variety of astrocyte forms in the CNS. This Wgure illustrates several astrocyte morphologies, including (A) astrocytes of the white matter (‘‘Wbrous astrocytes’’), (B) Bergmann glia of the cerebellar cortex, and (C,D) astrocytes in gray matter (‘‘protoplasmic astrocytes’’). Note the endfeet of astrocytes processes that abut blood vessels (arrows in A,C). Figures were captured from rat CNS that had been injected perinatally with a beta-gal retrovirus into the SVZ and allowed to mature for several weeks. Infected cells were visualized with the X-gal reagent.

occurs in all CNS regions. In the second, astrocytes arise from migratory progenitors in the subventricular zone (SVZ). This pathway occurs in forebrain and cerebellum, where SVZ cells migrate out into the parenchyma to diVerentiate into astrocytes, oligodendrocytes, and in the cerebellum and olfactory bulb, interneurons. Whether astrocyte genesis from SVZ occurs in spinal cord or brainstem is not clear. It may be that in those areas, transformation from radial glia constitutes the major astrocytogenic pathway. These two pathways must be kept in mind when reading studies of gliogenesis from diVerent areas of the CNS.

Some Astrocytes Arise from Radial Glia Radial glia, which arise from the ventricular zone (VZ) concurrently with neurogenesis, span the developing brain from ventricular surface to pia. These cells act as substrates along which cortical neurons migrate during early forebrain development. Recent studies suggest that radial glia give rise to neurons (Campbell and Gotz, 2002; Malatesta et al., 2000; Noctor et al., 2001) and thus have properties of multipotent CNS progenitors (discussed later). Radial glia share a number of characteristics with astrocytes: (1) the intermediate Wlament proteins, nestin and vimentin, and in primates, GFAP, as well (Benjelloun-Touimi etal., 1985; Dahl et al., 1981; HockWeld and McKay, 1985; LeVine and Goldman, 1988; Levitt et al., 1981; Tohyama et al., 1992); (2) glycogen storage (Khadim et al., 1988); (3) glutamate transporters (Yamada et al., 1998); (4) the aldolase isoform, zebrinll (Marshall and Goldman, 2002; Staugaitis et al., 2001); and (5) the attachments of end feet of the cells to the basal lamina at the pial surface. Radial glia, at least some proportion thereof, transform into astrocytes, largely after neuronal migration has ceased. This transformation was originally inferred from morphological studies that showed ‘‘intermediate’’ or ‘‘transitional’’ forms between radial glia and astrocytes in the early post-natal rodent brain and late gestational primate brain (Choi and Lapham, 1978; Misson et al., 1988, 1991b; Schmechel and Rakic, 1979b) (Figure 12.2).These ‘‘intermediates’’ exhibit a general radial orientation but in addition a more complex

II. GLIAL CELL DEVELOPMENT

IS ASTROCYTE DEVELOPMENT IDENTICAL

FIGURE 12.2 Transformation of radial glia into astrocytes. Radial glial cells (left) transform into a variety of astrocytes in gray and white matter with blood vessel and pial attachments (right). ML, molecular layer; CP, cortical plate; IZ, intermediate zone; VZ, ventricular zone. (From Cameron and Rakic, 1991)

shape, sprouting branches as they become more astrocytic. While these observations do not directly prove a transformation, they are consistent with it. The morphological studies are complicated by the fact that progenitors from the SVZ usually take a radial path as they migrate into the cortex, and some of these progenitors come to rest at the pial surface, with which they establish contact (Zerlin et al., 1995). The transformation of radial glia to astrocytes has been directly observed in two ways. First, radial glia were labeled in situ in the neonatal ferret brain by placing crystals of the Xuorescent dye, Dil, on the pial surface and then following the development of these cells into astrocytes (Voigt, 1989). Second, expressing activated Notch 1 in VZ cells of the embryonic mouse forebrain via a retroviral vector promotes their development into radial glia (Gaiano et al., 2000, and see the discussion that follows). The alkaline phosphatase reporter gene encoded by the vector allowed the investigators to trace the fates of radial glia into astrocytes in the post-natal forebrain.

II. GLIAL CELL DEVELOPMENT

313

314

12. ASTROCYTE LINEAGE

Little is known of the mechanisms by which radial glia transform into astrocytes. One can infer from the morphological changes—radial to a more complex, multiprocess shape—that a great deal of remodeling and new process growth is required. In addition, some radial glia must lose their attachments at the pial and ventricular surfaces and form new attachments to blood vessels.

Some ‘‘Radial Glia’’ Persist in the Adult Mammalian CNS Some glial cells with radial morphologies apparently persist into adulthood. These include the Bergmann glia of the cerebellar cortex, radial glia in the hippocampus, and radially oriented glia found in the brainstem and spinal cord (see, for example, Flament-Durand and Brion, 1983; King, 1966; Mori et al., 1990; Reichenbach, 1990). Bergmann glia maintain their connections to the pial surface, while many of the radial glia in the stem and cord maintain connections to the ventricular surface. These latter cells have been classically termed ‘‘tanycytes.’’ Some of these cells serve as migratory pathways in the post-natal CNS, one example of which is the well-known cerebellar granule cell migration from the external to the internal granule cells layer along Bergmann glial processes. Others have been presumed to carry signals of ions or small proteins from the cerebral spinal Xuid into the brain (Mori et al., 1990). Recent studies have suggested that some of the residual radial glia may have stem-like cell properties. That is, they retain the developmental plasticity of radial glia of the embryonic CNS. A subependymal layer of astrocytes (deWned as such by their expression of GFAP) remains into adulthood. These cells are likely to be derivatives of the immature neuroepithelial cells of the VZ, although this lineage has not been proven. The Wrst indication that these glial cells retained immature features came from the Wnding that they are reponsible for generating the neuroblasts of the adult SVZ that migrate into the olfactory bulb and diVerentiate into interneurons (Doetsch et al., 1999). At least some of these cells also have the potential to generate both neurons and glia in culture. Similarly, an astrocyte population that resides in the subgranular region of the hippocampus generates neuroblasts that migrate into the dentate gyrus, where they diVerentiate into granule neurons (reviewed in Fabel et al., 2003). The preceding discussion does not imply that all astrocytes throughout the CNS share neurogenic or stem-like capacities, however, either in vivo or in culture (discussed later).

Astrocytes Also Develop Directly from SVZ Cells The SVZs of the forebrain and cerebellum, established in late gestation and existing in large scale through early post-natal life, constitute a major source of astrocytes and oligodendrocytes. The SVZ contains large numbers of highly proliferative cells, easily labeled by markers of DNA synthesis, 3H-thymidine, or BrdU. Indeed, the classic studies of forebrain gliogenesis from SVZ cells inferred the migration of SVZ cells into white matter and gray matter and their diVerentiation into astrocytes and oligodendrocytes from thymidine pulse labeling (Altman 1963; Fujita, 1965; Paterson et al., 1973). Conclusions reached from replicating DNA labeling can be ambiguous, since there are dividing progenitors outside of the SVZ at this time in development and since the continued division of a progenitor cell will eventually dilute the thymidine or BrdU below the level of detection. In contrast, labeling of dividing cells with replication-deWcient retroviruses provides the incorporation of a heritable marker into a cell’s genome and allows fate tracing, since the progeny of infected cells will continue to express the marker. Retroviral labeling of SVZ cells, by stereotactic injection directly into the SVZ, showed directly that these progenitors emigrate from the SVZ and migrate into white matter and gray matter, where they diVerentiate into astrocytes and oligodendrocytes (Levison and Goldman, 1993; Levison et al., 1993; Luskin and McDermott, 1994). Astrocytes generated from SVZ cells were distributed throughout all cortical layers, up to the pial surface, in white matter, and in the striatum (Figure 12.3).

II. GLIAL CELL DEVELOPMENT

IS ASTROCYTE DEVELOPMENT IDENTICAL

FIGURE 12.3 Migration and glial diVerentiation of SVZ cells in the postnatal rat CNS. The forebrain SVZ of a P1 rat was stereotactically injected with the BAG retrovirus. A. The initial infected population is composed of SVZ cells with simple, unipolar morphologies (cameral lucida drawing). B. Labeled cells migrate from the SVZ to colonize white and gray matter. In white matter, they develop largely into oligodendrocytes (shown are two myelinating oligodendrocytes and a cluster of developing oligodendrocytes), whereas in gray matter, they develop into both oligodendrocytes and astrocytes (subpial astrocyte, with pial based endfeet, and two protoplasmic astrocytes are depicted; cortical oligodendrocytes are not shown in this diagram). (From Goldman, 2001).

Examining early stages of astrocyte development from SVZ cells revealed some of the earliest morphological and molecular changes in this process (Zerlin, et al., 1995; Zerlin and Goldman, 1997). An interaction with blood vessels or the pial surface appears to be one of the Wrst signs of astrocyte diVerentiation (Figure 12.4). These cells still retain the simple, largely unipolar morphology of SVZ and migratory cells. At that time, one sees the expression of the intermediate Wlament proteins, vimentin and nestin, in the progenitors. In rodents, GFAP expression occurs at a later stage. Astrocytes that show more complex branching patterns appear to be already in contact with vessels. This early interaction with blood vessels thus constitutes an early stage in astrocyte development. In fact, the contact with endothelial cells induces astrocyte diVerentiation in astrocyte progenitors cultured from optic nerve (Mi et al., 2001). That is, endothelial cells induce the expression of the astrocyte markers, GFAP and SlOOb, in immature cells. This induction was neutralized with antibodies to leukemia inhibition factor (LIF), a growth factor expressed by endothelia (Mi et al., 2001) (a discussion of molecular signals for astrocyte diVerentiation is presented later). However, the earliest stages of astrocyte diVerentiation from SVZ cells may take place prior to contact with vessels. The early

II. GLIAL CELL DEVELOPMENT

315

316

12. ASTROCYTE LINEAGE

FIGURE 12.4 Early astrocyte-blood vessel interactions. A. beta-gal labeled cells in the rat neocortex 3 days after a P1 injection into the forebrain SVZ show relatively simple morphologies at the times of attachment to blood vessels (camera lucida drawings). B. Early cortical astrocyte enwrapping a blood vessel (asterisk). Typically, the blood vessel contacts exhibit cytoplasmic expansions, similar to the endfoot processes of mature astrocytes. ((A) taken from Goldman, 2001).

FIGURE 12.5 Early expression of zebrinII by astrocytes. Double immunoXuorescence labeling of early astrocytes in the rat neocortex. Two progenitors with early vessel connections (asterisk) and one with a simple form without vascular connection (arrow) label with antibodies to betagalactosidase (A) and zebrin II (B) B. (Taken from Staugaitis et al., 2001).

astrocyte marker, zebrin II, a form of aldolase C, is expressed during the migration of a glial progenitor, before vessel interactions (Staugaitis et al., 2001) (Figure 12.5). In addition, the astrocytic glutamate transporter, GLAST, is expressed by progenitors migrating from the cerebellar SVZ through the white matter toward the cerebellar cortex (Milosovic and Goldman, 2002). Thus, early diVerentiation along an astrocyte lineage may induce molecular changes in progenitors that allow them to bind to young vessels, after which contact further stages of astrocyte diVerentiation ensue.

How Do Astrocyte Progenitors Emigrate from the SVZ and Migrate into Brain Parenchyma? As progenitors emigrate from the SVZ, they usually leave in a radial orientation (Kakita and Goldman, 1999; Suzuki and Goldman, 2003) (Figure 12.6). This suggests that they

II. GLIAL CELL DEVELOPMENT

IS ASTROCYTE DEVELOPMENT IDENTICAL

FIGURE 12.6 Migration paths of progenitors from the SVZ into the adjacent striatum, white matter and cortex. Arrows indicate direction of migration in the SVZ and migration from the SVZ into surrounding structures. Progenitors migrate radially into cortex, but then may move in a tangential direction through the cortex or even back toward the SVZ. Progenitors in white matter tend to migrate in the direction of white matter axons. In the striatum, progenitors at early postnatal times appear to avoid the (unmyelinated) pencil Wbers. (From Kakita and Goldman, 1999)

use radial glia as tracks from which to escape the SVZ. Indeed, the examination of progenitors as they leave the SVZ and as they migrate radially into white matter, and cortex shows that many contact radial glial processes (Suzuki and Goldman, 2003; Zerlin et al., 1995), which the CNS retains into early post-natal life. Furthermore, some SVZ cells migrate laterally along the white matter and then turn radially to enter the cortex, a pattern reminiscent of neuronal migration along radial glia (Bayer et al., 1991; Misson et al., 199la). Glioblasts emigrating from the SVZ migrate radially, but once in white matter can move in a direction parallel to axon pathways or radially into the cortex, where they can either continue or turn to migrate tangentially (Kakita and Goldman, 1999; Suzuki and Goldman, 2003). Glioblasts migrate in a saltatory fashion at an average velocity of about 90 m per hour, but maximal speeds of up to 250 m per hour. As astrocyte progenitors contact vessels or pia, they cease migration, but continue to grow and branch processes extensively, and continue to divide.

Do Different Pathways of Astrocyte Development in the Mammalian Forebrain Give Rise to Different Types of Astrocytes? The studies summarized thus far indicate that there are two (at least) separate pathways through which the CNS provides astrocytes from immature cells: directly from radial glia and from SVZ cells. Is there any reason to think that these pathways generate diVerent astrocyte types? Studies of SVZ cell development with retroviruses suggest that these migrating cells generate astrocytes in gray matter preferentially to white matter, where they largely diVerentiate into oligodendrocytes (Levison and Goldman, 1993). Perhaps, therefore, many white matter astrocytes, at least in rodents, are generated from radial glia. Are the diVerences between gray matter and white matter astrocytes (for example, a higher GFAP content in white matter glia), generated by lineage (radial glia versus SVZ cells) or by local environment (gray matter versus white matter)?

II. GLIAL CELL DEVELOPMENT

317

318

12. ASTROCYTE LINEAGE

Astrocytes constitute a heterogeneous group of cells, which vary in shape and molecular constituents, including growth factor expression, ion channels, levels of GFAP, and types of neurotransmitter transporters. How much of this heterogeneity is determined by lineage and how much is determined by local environmental factors is not yet known. Regardless of the source of a given astrocyte, the CNS must have a mechanism to generate large numbers of astrocytes in the perinatal period. It is in this time of rapid brain growth that the pial surface greatly enlarges, especially in larger mammals with gyriform brains, the vascular tree grows tremendously, and much of synaptogenesis takes place. In the cerebellum, the pial surface enlarges greatly at the time of granule cell migration. Thus, more Bergmann glia are needed to cover the pia and conduct granule cell traYc. Pial, vascular, and neuronal surfaces require investments by astrocyte end feet, and therefore this developmental time requires an increasing number of astrocytes; or an increasing number of astrocyte processes—astrocytes do progress from unipolar or bipolar progenitors to cells with highly branched processes; or astrocytes of an ever increasing size—but as noted previously, there seem to be constraints on astrocyte size. It is likely that the need for astrocytes is satisWed by both the division of radial glia (Schmechel and Rakic, 1979a) and the migration of astrocyte progenitors from the SVZ and the continued division of these progenitors (Levison and Goldman, 1993; Luskin and McDermott, 1994; Zerlin, et al., 1995; Zhang and Goldman, 1996).

Astrocyte Development in the Cerebellum The cerebellum contains Wbrous astrocytes in white matter, velate astrocytes in the internal granule cell layer, and Bergmann glia, whose cell bodies lie within the Purkinje cell layer and send complex branches to the pia (Chan-Palay and Palay, 1972; Palay and ChanPalay, 1974). Bergmann processes guide immature granule cells on their migration from the external to the internal granule cell layer and interact with Purkinje cell-parallel Wber synapses (Grosche et al., 1999). Early studies investigating 3H-thymidine uptake and morphological transformations suggest that Bergmann glia arise from other Bergmann glia or from some sort of immature cells, of unknown nature (Basco et al., 1977; Choi and Lapham, 1980; Moskovkin et al., 1978). Presumably, some Bergmann glia and other astrocytes arise from cerebellar radial glia, but there is no direct evidence for this idea. Fate studies using recombinant retroviruses show that progenitors that originate in the VZ at the base of the cerebellum give rise to Bergmann glia and velate and white matter astrocytes (along with oligodendrocytes) (Miyake et al., 1995). The authors inferred from the fact that the viral-labeled astrocytes were found in clusters that local proliferation continued after migration had ceased. In the post-natal cerebellum, astrocyte progenitors migrate from the base of the cerebellum through the white matter to reach the cortex (Milosevic and Goldman, 2002; Zhang and Goldman, 1996). At least some of the migratory progenitors in the white matter begin to express astrocyte characteristics, such as GLAST, during migration (Milosevic and Goldman, 2002). The details of Bergmann glial growth can in general be inferred from images of Golgi or retroviral labeled cells or immunocytochemical staining in the post-natal Purkinje cell layer (Choi and Lapham, 1980; Yamada et al., 2000; Zhang and Goldman, 1996). The cell bodies stop migration at the Purkinje cell layer and from there extend processes toward the pial surface. Over time, the numbers of processes increase as does the complexity of branches of these processes in dynamic synchrony with Purkinje cell dendritic growth and parallel Wber-Purkinje cell synapse formation. Whether it is the Purkinje cells or the migratory granule cells or other cerebellar components that inXuence astrocyte morphology to develop a Bergmann shape is not known. The inXuence of neurons on astrocyte shape was originally suggested from culture studies, in which astrocytes derived from neonatal cerebellum assumed Bergmann glial-like shapes, with elongated processes, if they were associated with migrating neurons, but a stellate shape when they associate with non-migratory neurons (Hatten et al., 1985, Hatten, 1988). It is unlikely that the astrocyte progenitors of the cerebellum are uniquely able to develop into Bergmann glia, since forebrain SVZ cells, transplanted into cerebellar white matter, give rise to cerebellar-

II. GLIAL CELL DEVELOPMENT

ASTROCYTE DEVELOPMENT IN THE ADULT MAMMALIAN CNS

speciWc astrocyte shapes, such as Bergmann glia and velate astrocytes (A. Milosovic, unpublished observations).

Astrocyte Development in the Spinal Cord A number of immunocytochemical studies suggest that radial glia in the spinal cord, as elsewhere, generate astrocytes (see Hirano and Goldman, 1988, for example), although as noted earlier, it is diYcult to infer lineage pathways from a chronological series of static images. Recent studies have linked the development of motor neurons and interneurons of the cord to oligodendrocyte development, arguing that all of these cell types arise from a single, ventral lineage, the speciWcity of which changes over time in a way correlated with changes in the expression of bHLH transcription factors that regulate patterning and cell fate (see Rowitch et al., 2002, for review). Astrocytes were not included in this lineage. Using Wbroblast growth factor receptor type 3 (Fgfr3) as a marker for astrocytes and their precursors in the developing cord, Pringle et al. (2003) have localized astrocyte precursors to both dorsal and ventral VZ, with the exception of the (ventral) pMN domain, which gives rise to oligodendrocytes and motor neurons. This is consistent with the earlier observation that astrocytes are generated from both dorsal and ventral parts of the neuroepithelium, while oligodendrocytes arise only from ventral parts (Pringle et al., 1998). Thus, the patterning of gliogenesis in the cord as reXected in the domains of the early neuroepithelium is diVerent for astrocytes and oligodendrocytes. It is possible that cells in the pMN domain suppress the ability to diVerentiate into astrocytes. In fact, in mice that are null for the bHLH factors Oligl and Olig2, the pMN domain is converted to an adjacent homeodomain in the VZ, that of p2, and that part of the VZ now generates interneurons followed by astrocytes (Takebayashi et al., 2002; Zhou and Anderson, 2002). Rao and colleagues have isolated a progenitor cell from the embryonic (El3.5) rat spinal cord that is restricted to a glial cell fate, astrocytes and oligodendrocytes (Rao et al., 1998). These precursors are initially A2B5-positive and PSA-NCAM-negative and generate A2B5-negative, Xat astrocytes in serum, A2B5-positive process-bearing astrocytes when exposed to CNTF and bFGF, and oligodendrocytes, but not neurons. Isolated glial restricted progenitors do not express Fgfr3 (at least by immunocytochemistry). A common progenitor for astrocytes and oligodendrocytes is consistent with observations from the forebrain SVZ; it seems inconsistent with Wndings from cord development in vivo. As Pringle et al. have pointed out (2003), the diVerence may reXect suppressive inXuences in vivo that may be abolished once the progenitors are removed and placed in culture. Astrocytes cultured from spinal cord and allowed to proliferate generate clones that display a variety of diVerent shapes (Miller and Szigeti, 1991), suggesting a marked heterogeneity in astrocyte conWgurations. Were these diVerences generated by lineage or by local inXuences? It is not yet clear whether these diVerent morphological forms correspond to speciWc forms in vivo.

ASTROCYTE DEVELOPMENT IN THE ADULT MAMMALIAN CNS Astrocytes continue to be generated in the adult CNS, albeit at an apparently very low rate. Studies that have counted astrocyte numbers or determined how many immature cells, labeled in the adult brain with 3H-thymidine conclude that there is little or no net accumulation of astrocytes with age (Altman, 1963; Hommes and Leblond, 1967; Kaplan and Hinds, 1980; Korr et al., 1973; Ling and Leblond, 1973; McCarthy and Leblond, 1988; Paterson, 1983; Reyners et al., 1986; Vaughan and Peters, 1974). This stands in contrast to the oligodendrocyte population, which continues to increase, slowly, over time (Levison et al., 1999; McCarthy and Leblond, 1988). Nevertheless, the pulse-chase studies indicate that new astrocytes are being generated from somewhere. The possiblities include division of already existing, mature astrocytes, or division and diVerentiation of immature cells, of some as yet undetermined nature. Astrocytes will divide in reponse to

II. GLIAL CELL DEVELOPMENT

319

320

12. ASTROCYTE LINEAGE

pathological conditions (see Norton, 1999 for review), but whether mature astrocytes normally divide is not clear. Cycling cells expressing S-100b or GFAP reside in the adult rat spinal cord, where they constitute a small proportion of the total proliferating cell population (Horner et al., 2000). But a larger proportion of BrdU-labeled cells express GFAPþ four weeks after BrdU injection than 1 hour after injection, suggesting the generation of new GFAPþ cells from some GFAP-population(s). These observations would have in the past been interpreted as reXecting the division of mature astrocytes, but as Horner et al. point out, the expression of GFAP by ‘‘stem-like cells’’ in the adult CNS (discussed later) and the report of S-100b in some oligodendrocytes makes the identiWcation of these cycling cells less clear. A number of recent studies have described cells isolated from the adult CNS that display ‘‘stem-like’’ properties in culture, able to give rise to proliferating cells that generate neurons, astrocytes, and oligodendrocytes (for example, see Reynolds and Weiss, 1992; Richards et al., 1992). Furthermore, at least some of these multipotent cells express astrocyte characteristics, including GFAP (Doetsch et al., 1999; Fabel et al., 2003; Laywell et al., 2000; Seri et al., 2001). While this observation does not imply that all astrocytes have stem cell properties or can potentially become stem cells under the appropriate conditions, it does mean that traditional astrocyte ‘‘markers’’ should be used with caution. The best studied of these astrocytic, stem-like cells reside either in the residual SVZ of the adult brain, where they generate neuroblasts that migrate into the olfactory bulb and become interneurons, or in the subgranular zone of the hippocampus, where they generate dentate granule cells. However, it may well be the case that such astrocytic cells reside throughout the CNS—it has been diYcult to identify them conclusively so far. Another approach to identifying immature glia of the adult CNS involves isolating a population of cycling cells of the adult rat neocortex and subcortical white matter. This turns out to be a heterogeneous population, which contains cells able to generate oligodendrocytes in culture as well as a subset of cells that can generate astrocytes (Gensert and Goldman, 2001). The astrocyte ‘‘progenitor’’ subpopulation appears not to express the early oligodendrocyte marker, O4, but to express the intermediate Wlament, vimentin, a cytoskeletal protein observed early in astrocyte diVerentiation. Thus, immature cells that have the potential to generate astrocytes do exist in the adult CNS, but a clear identiWcation of the nature of those cells and their capacities to diVerentiate into astrocytes in vivo is still lacking.

LINEAGE RELATIONSHIPS BETWEEN ASTROCYTES AND OTHER CNS CELLS Fate mapping studies using retroviruses have argued for a common progenitor for astrocytes and neurons in the retina (Turner and Cepko, 1987), optic tectum (Galileo et al., 1990; Gray and Sanes, 1990, 1992), cortex, and striatum (Halliday and Cepko, 1992; Walsh and Cepko, 1993), although other studies have inferred separate lineages for astrocytes, oligodendrocytes, and neurons (Luskin et al., 1993; Williams and Price, 1995). The recent Wnding that radial glia themselves can give rise to neurons (Malatesta, et al., 2000; Noctor et al., 2001) draws the lineage relationship between these two cell types even closer, suggesting that in some parts of the CNS, radial glia Wrst give rise to neurons and then to astrocytes (see Gray and Sanes, 1991, and the discussion presented earlier). A small proportion of clones generated in vitro by single CNS cells from the embryonic forebrain or post-natal SVZ contain astrocytes, neurons, and oligodendrocytes, arguing that immature cells have the potential to give rise to both glia and neurons (Davis and Temple, 1994; Levison and Goldman, 1997). Interpreting retroviral lineage studies in vivo can be diYcult, depending on how one deWnes a ‘‘clone’’ of cells. Many studies have found that retroviral-labeled cells tend to congregate in homogeneous clusters, with a small proportion of heterogeneous clusters (astrocytes and oligodendrocytes) (Grove et al., 1992, 1993; Levison and Goldman, 1993;

II. GLIAL CELL DEVELOPMENT

WHAT ARE THE MOLECULAR SIGNALS UNDERLYING ASTROCYTE DEVELOPMENT?

Luskin et al., 1988, 1993; Luskin and McDermott, 1994, Parnevales, 1999; Price and Thurlow, 1988). Forebrain gliogenesis has been reexamined using a retroviral ‘‘library’’ (Walsh and Cepko, 1992), so that the proximity of cells to one another becomes irrelevant in judging clonality and two related cells that happen to be separated in space by some distance can be found to come from the same retrovirally infected progenitor. Most clones are indeed homogeneous, but about 15 to 20% are composed of both astrocytes and oligodendrocytes, sometimes appearing in the same cluster of glia (Zerlin and Goldman, submitted). Thus, not all SVZ cells are irrevocably committed to an astrocytic or oligodendrocytic fate before they emigrate from the SVZ. In fact, some do not make a Wnal fate decision until they have stopped migrating, at which time they continue to divide to generate heterogeneous or homogenous clusters. Furthermore, these observations lead to a model in which glial fate is not irrevocably determined within the SVZ. The expression of early astrocyte characteristics (zebrin II, GLAST) by migratory progenitors, however, argues that for some cells, a fate decision takes place prior to the cell’s reaching its Wnal destination. It may well turn out that the astrocyte fate decision is probabilistic, rather than strictly determined at one speciWc place and time.

WHAT ARE THE MOLECULAR SIGNALS UNDERLYING ASTROCYTE DEVELOPMENT? Recent attention has focused on three sets of ligand-receptor interactions that play roles in promoting astrocyte diVerentiation: (1) the IL-6/LIF family of cytokines and the LIF receptor/gpl30 pair; (2) TGFb growth factors, particularly the bone morphogenetic proteins (BMPs) and BMP receptors; and (3) Notch and Notch ligand pairs. Hughes et al. (1988) initially found that ciliary neuronotrophic factor (CNTF), a member of the IL-6 family, would induce astrocyte diVerentiation in glial progenitors (‘‘O-2A’’ cells) isolated from the post-natal optic nerve. CNTF will also promote astrocyte development from cells cultured from the immature CNS (see Bonni et al., 1997; Gard et al., 1995; Johe et al. 1996). This CNTF eVect is generalizable to other members of the IL-6 family, including LIF, cardiotropin 1 (CT1, Ochiai et al., 2001) and oncostatin M (OSM, Yanagisawa et al., 1999), all of which signal through the LIFR/gpl30 receptor complex (Nakashima, 1999a). Binding of ligand to this receptor complex activates the JAK-STAT pathway (Bonni et al., 1997; Kahn et al., 1997), Wrst phosphorylating JAK, and then the transcription factor STAT3, which in turn binds to the CBP/p300 complex to activate the S3BE element in the GFAP promoter (Yanagisawa et al., 2001). Bear in mind that most of these experiments used immature cells cultured from the embryonic CNS, and it is thus diYcult to determine the exact nature of such cells, other than that they proliferate in culture with serum-free medium supplemented with growth factors, usually bFGF, and they do not express markers of mature neurons or glia. Most investigators have used the expression of GFAP as the major assay for astrocyte diVerentiation, a common and convenient marker for astrocytes. Indeed, the GFAP promoter has a consensus STAT sequence (Besnard et al., 1991; Bonni et al., 1997), which in transfection assays mediates the CNTF eVect (Bonni et al., 1997). Obviously, astrocyte diVerentiation involves the induction of many genes, of which GFAP is only one. It will be important in the future to assay for other astrocyte characteristics as well (S100-Jb used in Nakashima et al., 2001). For example, in the rodent CNS, several genes are expressed prior to GFAP: the intermediate Wlament vimentin, zebrin II (Staugaitis et al., 2001), and the glutamate transporter, GLAST (Shibata et al., 1997). Thus, GFAP is likely to be induced after a progenitor has already made an early astrocyte fate decision. This point is also illustrated by the LIF (
/
) mouse, which shows reduced GFAP mRNA levels at El9. CNS cultures that contain cells with the morphology of astrocytes, but lack GFAP (Koblar et al., 1998). Eventually, some GFAPþ cells arise, but only after a long delay. This study shows that LIF does play a role in the induction of GFAP expression, but it is unclear how astrocyte diVerentiation is impaired. Perhaps other LIF-family members can substitute for the molecule.

II. GLIAL CELL DEVELOPMENT

321

322

12. ASTROCYTE LINEAGE

LIF and related cytokines may act synergistically with components of extracellular matrix of mesenchymal cells to promote astrocyte diVerentiation (Lillien and RaV, 1990; Mayer et al., 1994). This may make sense, given that astrocytes interact with basal laminae at blood vessels and at the pial surface, and blood vessel interactions appear to be an early step in astrocyte diVerentiation (Mi et al., 2001; Zerlin and Goldman, 1997). As noted earlier, Mi and colleagues (1999, 2001) reported that LIF expression in cerebral endothelial cells will induce astrocyte characteristics in astrocyte progenitors. BMPs, particularly BMP2 and BMP7, promote astrocyte development from embryonic telencephalic cells (Gross et al., 1996). BMPs through their receptors, activate the Smad family of transcription factors. BMPs apparently act synergistically with LIF family members (Nakashima et al., 1999c), in part by inducing the formation of a transcription factor complex that contains both Smad and STATs (Nakashima et al., 1999b). BMP2 will induce GFAP expression in cultures from the LIF (
/
) mouse, suggesting that the two pathways can work independently. BMP2 may also play a role in inhibiting neuronal diVerentiation from telencephalic precursors as well as promoting astrocytic diVerentiation. BMP2 upregulates the expression of factors that repress neurogenic basic HLH proteins: immature neuroepithelial cells exposed to BMP2 increased their levels of Idl and Id3 as well as Hes-5 (Nakashima et al., 2001). The latter lies downstream in the Notch pathway and inhibits the expression of the pro-neuronal transcription factors, mashl and neurogenin. This Wnding suggests that BMPs actively participate in a fate decision between neuronogenesis and astrocytogenesis, clearly favoring the latter. The fact that E14 telencephalic cells are susceptible to this inXuence suggests additionally that this developmental decision occurs in VZ cells (rather than SVZ cells) and is likely therefore to switch VZ cells to a radial glial and then astrocyte fate. Note that activated Notch expression in VZ cells strongly promotes radial glial and astrocytic fates (discussed below). The family of Notch transmembrane receptors control cell fate decisions by interaction with Notch ligands expressed on the surface of adjacent cells. Recent work in the forebrain and retina indicates that Notch activation can promote the development of progenitors into radial glia and astrocytes. These experiments involve forcing immature cells to express an activated form of Notch encoded by a retroviral vector. When such a vector was introduced into the E9.5 mouse forebrain, it directed the fate of the infected cells toward that of radial glia (Gaiano et al., 2000). In contrast, cells infected by control vectors at this age become either neurons or radial glia. The fates of the (infected) radial glia were traced into post-natal life, where viral infected cells had developed into astrocytes, as expected (noted earlier) and into glial cells that resided in the subependymal region. No oligodendrocytes were observed. Injecting a Notch 1-expressing retrovirus into the PO retina promoted a Mu¨ller glial fate and not a neuronal fate in the infected cells (Furukawa et al., 2000). This fate decision appeared to be at the expense of rod photoreceptors, also born at that time, since infection with a control virus results in a large predominance of rod cells, and only a few Mu¨ller glia. An interplay of these various pathways contributes to astrocyte genesis. In addition, the balance between ‘‘neurogenic’’ and ‘‘gliogenic’’ bHLH factors is likely to regulate the fate decision of an immature cell. Thus, overexpressing the neurogenic factor, neurogenin 1, in embryonic neuroepithelial cells not only promoted neurogenesis, but also decreased the ability of cells to respond to astrocytogenic signals, such as LIF (Sun et al., 2001). Sun et al. speculated that neurogenin 1 binds to the same CBP/p300 complex as do the STATs. Thus, the relative levels of neurogenin 1 and STAT3 may in part determine whether an immature cell becomes a neuron or an astrocyte. Merely overexpressing neurogenins or Mash 1 by retroviral infection in the developing forebrain does not alter dramatically the numbers of neurons versus astrocytes that the infected cells developed into, suggesting that it is not just the levels of one of the bHLH factors that determines cell fates in vivo. Similarly, knocking out both neurogenin 2 and Mash 1 did not produce a dramatic decrease in neurons and a increase in astrocytes, although the cortices of these mice displayed marked disorganization of laminar patterning (Nieto et al., 2001).

II. GLIAL CELL DEVELOPMENT

FUTURE PERSPECTIVES AND UNRESOLVED QUESTIONS

How do these various signaling pathways determine astrocyte development in vivo? The various soluble factors are all present in the developing CNS: CNTF and LIF (Stockli et al., 1991; Yamamori, 1991) and BMPs (see Li et al., 1998; Mabie et al., 1999, for example). The BMP antagonist, Noggin, is expressed in the developing cortex (Li and LoTurco, 2000) and in adult rodents it is found in ependymal cells (Lim et al., 2000), where it may function to counteract a BMP-mediated astrocytic development and promote neurogenesis. In the early post-natal period, such a function is also possible, in that those SVZ cells that are located closest to the ventricle tend to migrate parallel to the ventricle in an caudalrostral dimension, eventually ending up in the olfactory bulb as neurons (Suzuki and Goldman, 2003). In contrast, the glioblasts, which migrate out into the white matter and gray matter, appear concentrated near the periphery of the SVZ, farthest from the ventricle (Marshall et al., 2003; Suzuki and Goldman, 2003), suggesting that there is a gradient of neuronal versus gliogenic signals within the SVZ. With respect to astrocyte diVerentiation, in vitro studies suggest that CNTF/LIF factors operate more eYciently in inducing astrocytic diVerentiation in the presence of extracellular matrix (Lillien and RaV, 1990). LIF is expressed by endothelial cells (Mi and Barres, 1999), again pointing to the vascular system as one potential positive regulator of astrocyte diVerentiation. Notch is expressed on VZ cells, or at least a subset thereof (Yun et al., 2002), and, as noted earlier, expression of constitutively active Notch points VZ cells toward a radial glial fate, from whence they become astrocytes. A further pathway that should be considered is signaling through cyclic-AMP, which increases GFAP transcription in astrocytes (Masood et al., 1993; ShaWt-Zagardo et al., 1988), possibly regulated through the CREB sequence in the GFAP promoter (Besnard et al., 1991; Masood et al., 1993). Whether this pathway is important for astrocyte diVerentiation or only serves to regulate GFAP transcription in astrocytes is not clear, although McManus et al., (1999) provides evidence for astrocyte fate induction by cAMP in precursor cells from the embryonic neocortex.

FUTURE PERSPECTIVES AND UNRESOLVED QUESTIONS 1. Astrocytes vary considerably in form and function from one region of the CNS to another. How and when do these diVerences develop? What local signals instruct astrocyte progenitors to assume forms and functions of astrocytes appropriate to a given part of the CNS? Transplanting progenitors from the neonatal forebrain SVZ into neonatal cerebellar white matter results in their diVerentiation into cerebellar-speciWc astrocyte forms. These include Bergmann glia and velate astrocytes in the internal granule cell layer, forms never observed in the forebrain. This Wnding suggests that some basic quality of ‘‘astrocyte-ness’’ is not region speciWc, but diVerent regions are able to confer on immature astrocytes the forms required to function in a given location. 2. What are the initial stages of astrocyte diVerentiation? How can we recognize these stages? It seems as if at least some astrocytes begin to diVerentiate before they reach their Wnal destinations, as judged by the expression of astrocyte markers in migratory progenitors. If this is indeed the case, what controls this initial expression? How is astrocyte development further promoted by contact of progenitors with basal laminae at blood vessels and pia? Can the BMP and IL-6 pathways account entirely for astrocyte diVerentiation? How are astrocyte genes regulated by these pathways and what are those genes? 3. What are the molecular cues that allow astrocytes to interact with basal laminae and with neurons during development? How does this interaction contribute to the formation of the blood brain barrier and promote or stabilize synaptogenesis? 4. How does the characteristic spacing of astrocytes occur? Since glial progenitors continue to divide after they stop migrating, many sibling astrocytes must begin life next to each other.

II. GLIAL CELL DEVELOPMENT

323

324

12. ASTROCYTE LINEAGE

5. How does cell death shape the distribution and numbers of astrocytes? 6. What are the molecular events and controls that allow radial glia to transform into astrocytes? 7. Why and how do some immature neuroepithelial cells remain immature throughout life, functioning as stem-like precursors in the SVZ and hippocampus? Why do these cells have any astrocyte characteristics at all?

Acknowledgments The author would like to thank present and previous colleagues in my laboratory for their many contributions and critical insights. My studies were supported by National Institutes of Health grant NS17125.

References Abbot, N. J. (2002). Astrocyte-endothelial interactions and blood-brain barrier permeability. J. Anat. 200, 629–638. Altman, J. (1963). Autoradiographic investigation of cell proliferation in the brains of rats and cats. Anat. Rec. 145, 573–591. Andriezen, W. L. (1893). The neuroglia elements in the human brain. Brain J. Medicine 2, 227–230. Basco, E., Hajos, F., and Fulop, Z. (1977). Proliferation of Bergmann-glia in the developing rat cerebellum. Anat. Embryol. 151, 219–222. Bayer, S. A., Altman, J., Russo, R. J., Dai, X., and Simmons, J. A. (1991). Cell migration in the rat embryonic neocortex. J. Compar. Neurol. 307, 499–516. Benjelloun-Touimi , S., Jacque, C. M., Derer, P., De Vitry, F., Maunoury, R., and Dupouey, P. (1985). Evidence that mouse astrocytes may be derived from the radial glia. An immunohistochemical study of the cerebellum in the normal and reeler mouse. J. Neuroimmunol. 9, 87–97. Besnard, F., Brenner, M., Nakatani, Y., Chao, R., Purohit, H. J., and Freese, E. (1991). Multiple interacting sites regulate astrocyte-speciWc transcription of the human gene for glial Wbrillary acidic protein. J. Biol. Chem. 266, 18877–18883. Bonni, A., Sun, Y., Nadal-Vicens, M., Bhatt, A., Frank, D. A., Rozovsky, I., Stahl, N., Yancopovlos, G. D., and Greenberg, M. E. (1997). Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway. Science 278, 477–483. Bushong, E. A., Martone M. E., Jones, Y. Z., and Ellisman, M. H. (2002). Protoplasmic astrocytes in CA1 stratum radiatum occupy separate anatomical domains. J. Neurosci. 22, 183–192. Cameron, R. R. and Rakic, P. (1991). Glial cell lineage in the cerebral cortex: a review and synthesis. Glia 4, 124–137. Campbell, K., and Gotz, M. (2002). Radial glia: Multi-purpose cells for vertebrate brain development. Trends Neurosci. 25(5), 235–238. Cajal, S. R. (1995). ‘‘Histology of the Nervous System of Man and Vertebrates.’’ Vol. 1, pp. 191–208. Oxford University Press, New York. Chan-Ling, T., and Stone, J. (1991). Factors determining the migration of astrocytes into the developing retina: Migration does not depend on intact axons or patent vessels. J. Comp. Neurol. 303, 375–386. Chan-Palay, V., and Palay, S. L. (1972). The form of velate astrocytes in the cerebellar cortex of monkey and rat: High voltage electron microscopy of rapid Golgi preparations. Z. AnatEntwicklungsgesch. 138, 1–19. Choi, B. H., and Lapham, L. W. (1978). Radial glia in the human fetal cerebrum: A combined Golgi, immunoXuorescent and electron microscopic study. Brain. Res. 148, 295–311. Choi, B. H., and Lapham, L. W. (1980). Evolution of Bergmann glia in developing human fetal cerebellum, A Golgi, electron microscopic and immunoXuorescent study. Brain Res. 190, 369–383. Dahl, D., Rueger, D. C., and Bignami, A. (1981). Vimentin, the 57000 molecular weight protein of Wbroblast Wlaments, is the major cytoskeletal component in immature glia. Eur. J. Cell Biol. 24, 191–196. Davis, A. A., and Temple, S. (1994). A self-renewing multipotential stem cell in embryonic rat cerebral cortex. Nature 372, 263–266. Doetsch, .F, Caille, I., Lim, D. A., Garcia-Verdugo, J. M., and Alvarez-Buylla, A. (1999). Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell 97, 730–716. Fabel, K., Todas, H., Fabel, K., and Palmer, T. (2003). Copernican stem cells: Regulatory constellations in adult hippocampal neurogenesis. J. Cell. Biochem. 88, 41–50. Flament-Durand, J., and Brion, JK. P. (1983). Tanycytes: Morphology and functions. Internat. Rev. Cytology 96, 121–155. Fujita, S. (1965). An autoradiographic study on the origin and fate of the sub-pial glioblast in the embryonic chick spinal cord. J. Comp. Neurol. 124, 51–60. Furukawa, T., Mukherjee, S., Bao, Z-Z., Morrow, E. M., and Cepko, C. L. (2000). rax, Hesl, and notchl promote the formation of Muller glia by postnatal retinal progenitor cells. Neuron 26, 383–394.

II. GLIAL CELL DEVELOPMENT

FUTURE PERSPECTIVES AND UNRESOLVED QUESTIONS

Gaiano, N., Nye, J. S., and Fishell, G. (2000). Radial glial identity is promoted by Notchl signaling in the murine forebrain. Neuron 26, 395–404. Galileo, D. S., Gray, G. C., Owens, G. C., Majors, J., and Sanes, J. R. (1990). Neurons and glia arise from a common progenitor in chicken optic tectum: demonstration with two retroviruses and cell-type-speciWc antibodies. Proc. Natl. Acad. Sci. USA 87, 458–462. Gard, A. L., Williams, W. C., and Burrell, M. R. (1995). Oligodendroblasts distinguished from O-2A glial progenitors by surface phenotype (O4þ/GalC-) and response to cytokines using signal transducer LIFR. Dev. Biol. 167, 596–608. Goldman, J. E. (2001). Developmental origins of astrocytes In ‘‘Glial Cell Development, Second Edition.’’ Jessen, K. R. and Richardson, W. D. (eds.). Oxford University Press. Gensert, J. M. and Goldman, J. E. (2001). Heterogeneity of cycling progenitors in the adult mammalian cortex and white matter. J. Neurobiol. 48, 75–86. Gray, G. E., and Sanes, J. R. (1991). Migratory paths and phenotypic choices of clonally related cells in the avian optic tectum. Neuron 6, 211–225. Gray, G. E., and Sanes, J. R. (1992). Lineage of radial glia in the chicken optic tectum. Development 114, 271–283. Grosche, J., Matyash, V., Moller, T., Verkhratsky, A., Reichenbach, A., and Kettenmann, H. (1999). Microdomains for neuron-glia interaction: Parallel Wber signaling to Bergmann glial cells. Nat. Neurosci. 2, 139–143. Gross, R. E., M. F. Mehler, P. C. Mabie, Z. Zang, L. Santschi, and J. A Kessler (1996). Bone morphogenetic proteins promote astroglial lineage commitment by mammalian sub ventricular zone progenitor cells. Neuron 17, 595–606. Grove, E. A., Kirkwood, T. B. L., and Price, J. (1992). Neuronal precursor cells in the rat hippocampal formation contribute to more than one cytoarchitectonic area. Neuron 8, 217–229. Grove, E. A., Williams, B. P., Da-Qing, L., Hajihosseini, M., Friedrich, A., and Price, J. (1993). Multiple restricted lineages in the embryonic rat cerebral cortex. Development 117, 553–561. Halliday, A. L., and Cepko, C. L. (1992). Generation and migration of cells in the developing striatum. Neuron 9, 15–26. Hama, K., Arii, T., and Kosaka, T. (1993). Three dimensional organization of neuronal and glial processes: High voltage electron microscopy. Micr. Res. Tech. 29, 357–367. Hatten, M. E. (1985). Neuronal regulation of astroglial morphology and proliferation in vitro. J. Cell Biol. 100, 384–396. Hatten, M. E., Liem, R. K. H., and Mason, C. A. (1988). Two forms of cerebellar glial cells interact diVerently with neurons in vitro. J. Cell Biol. 98, 193–204. Hirano, M., and Goldman, J. E. (1988). Gliogenesis in rat spinal cord: Evidence for origins of oligodendrocytes and astrocytes from radial precursors. J. Neurosci. Res. 21, 155–167. HockWeld, A., and McKay, R. D. G. (1985). IdentiWcation of major cell classes in the developing mammalian nervous system. J. Neurosci. 5, 3310–3318. Hommes O. R., and Leblond, C. P. (1967). Mitotic division of neuroglia in the normal adult rat. J. Comp. Neurol. 129, 269–278. Horner, P. J., Power, A. E., Kempermann, G., Kuhn, H. G., Palmer, T. D., Winkler, J., Thai, L. J., and Gage, F. H. (2000). Proliferation and diVerentiation of progenitor cells throughout the intact adult rat spinal cord. J. Neurosci. 20, 2218–2228. Hughes, S. M., Lillien, L. E., Raff, M. C., Rohrer, H., Sendtner, M. (1988). Ciliary neurotrophic factor induces type-2 astrocyte differentiation in culture. Nature 335, 70–3. Johe, K. K., Hazel, T. G., Muller, T., et al. (1996). Single factors direct the diVerentiation of stem cells from fetal and adult central nervous system Genes Dev. 10, 3129–3140. Kahn, M. A., Huang, C. J., Caruso, A., et al. (1997). Ciliary neurotrophic factor activates JAK/Stat signal transduction cascade and induces transcriptional expression of glial Wbrillary acidic protein in glial cells. J. Neurochem. 68, 1413–1423. Kakita, A., and Goldman, J. E. (1999). Patterns and dynamics of SVZ cell migration in the postnatal forebrain: Monitoring living progenitors in slice preparations. Neuron 23, 461–472. Kaplan, M. S., and Hinds, J. W. (1980). Gliogenesis of astrocytes and oligodendrocytes in the neocortical grey and white matter of the adult rat: Electron microscopic analysis of light radioautographs. J. Comp. Neurol. 193, 711–727. Khadim, H. J., Gadisseux, J-F, and Evrard, P. (1988). Topographical and cytological evolution of the glial phase during prenatal development of the human brain: histochemical and electronmicroscopic study. J. Neuropathol. Exp. Neurol. 47, 166–188. King, J. S. (1966). A comparative investigation of neuroglia in representative vertebrates. J. Morphol. 119, 435–466. Koblar, S. A., Turnley, A. M., Classon, B. J., Reid, K. L., Ware, C. B., Cheema, S. S., Murphy, M., and Bartlett, P. P. (1998). Neural precursor diVerentiation into astrocytes requires signaling through the leukemia inhibitory factor receptor. Proc. Natl. Acad. Sci. USA 95, 3178–3181. Korr, H. Schultze, B., and Maurer, W. (1973). Autoradiographic investigations of glial proliferation in the brain of adult mice. J. Comp. Neurol. 150, 169–176. Kosaka, T., and Hama, K. (1986). Three-dimensional structure of astrocytes in the rat dentate gyrus. J. Comp. Neurol. 249, 242–260. Laywell, E. D., Rakic, P. Kukekov, V. G., Holland, E., and Steindler, D. A. (2000). IdentiWcation of a multipotent astrocytic stem cell in the immature and adult mouse brain. Proc. Natl. Acad. Sci. USA 97, 13883–13888.

II. GLIAL CELL DEVELOPMENT

325

326

12. ASTROCYTE LINEAGE

au1

LeVine, S. M., and Goldman, J. E. (1988). Embryonic divergence of oligodendrocyte and astrocyte lineages in developing rat cerebrum. J. Neurosci 8, 3992–4006. Levison, S. W., Chuang, C., Abramson, B., and Goldman, J. E. (1993). The migrational patterns and developmental fates of glial precursors in the rat subventricular zone are temporally regulated. Development 119, 611–623. Levison, S. W., and Goldman, J. E. (1993). Both oligodendrocytes and astrocytes develop from progenitors in the subventricular zone of postnatal rat forebrain. Neuron 10, 201–212. Levison, S. W., and Goldman, J. E. (1997). Multipotential progenitors co-exist with lineage restricted precursors in the mammalian postnatal subventricular zone. J. Neurosci. Res. 48, 83–94. Levison, S. W., Young, G. M., and Goldman, J. E. (1999). Cycling cells in the adult rat neocortex produce oligodendrocytes. J. Neurosci. Res. 57, 435–446. Levitt, P., Cooper, M. L., and Rakic, P. (1981). Coexistence of neuronal and glial precursor cells in the cerebral ventricular zone of the fetal monkey: An ultrastructural immunoperoxidase analysis. J. Neurosci 1, 27–39. Li, W., Cogswell, C. A., and LoTurco, J. J. (1998). Neuronal diVerentiation of precursors in the neocortical ventricular zone is triggered by BMP. J. Neurosci. 18, 8853–8862. Li, W., and Lo Turco, J. J. (2000). Noggin is a negative regulator of neuronal diVerentiation in developing neocortex. Dev. Neurosci. 22, 68–73. Lim, D. A., Tramontin, A. D., Trevejo, J. M., Herrera, D. G., Garcia-Verdugo, J. M., and Alvarez-Buylla, A. (2000). Noggin antagonizes BMP signaling to create a niche for adult neurogenesis. Neuron 28, 713–726. Ling, E. A., and Leblond, C. P. (1973). Investigation of glial cells in semithin sections. II. Variation with age in the numbers of the various glial cell types in rat cortex and corpus callosum. J. Comp. Neurol. 149, 73–82. Lillien, L. E., and RaV, M. C. (1990). DiVerentiation signals in the CNS: Type-2 astrocyte development in vitro as a model system. Neuron 5, 111–119. Luskin, M. B., Parnavelas, J. G., and BarWeld, J. A. (1993). Neurons, astrocytes, and oligodendrocytes of the rat cerebral cortex originate from separate progenitor cells: An ultrastructural analysis of clonally related cells. J. Neurosci. 13, 1730–1750. Luskin, M. B., Pearlman, A. L., and Sanes. J. R. (1988). Cell lineage in the cerebral cortex of the mouse studied in vivo and in vitro with a recombinant retrovirus. Neuron 1, 635–647. Luskin, M. B., and McDermott, K. (1994). Divergent lineages for oligodendrocytes and astrocytes originating in the neonatal forebrain subventricular zone. Glia 11, 211–226. Mabie, P. C., Mehler, M. R., and Kessler, J. A. (1999). Multiple roles of bone morphogenetic protein signaling in the regulation of cortical cell number and phenotype. J. Neurosci. 19, 7077–7088. Malatesta, P., Hartfuss, E., and Gotz, M. (2000). Isolation of radial glial cells by Xuorescent-activated cell sorting reveals a neuronal lineage. Development 127, 5253–5263. Marshall, C. A. G., and Goldman, J. E. (2002). Subpallial Dlx2-expressing cells give rise to astrocytes and oligodendrocytes in the cerebral cortex and white matter. J. Neurosci. 22, 9821–9830. Marshall, C. A. G., Suzuki, S., and Goldman, J. E. (2003). Gliogenic and neurogenic progenitors of the subventricular zone; who are they, where did they come from, and where are they going? Glia, 43, 52–61. Masood K., Besnard, F., Su, Y., et al. (1993). Analysis of a segment of the human glial Wbrillary acidic protein gene that directs astrocyte-speciWc transcription. J. Neurochem. 61, 160–166. Mayer, M., Bhakoo, K., and Noble. M. (1994). Ciliary neuronotrophic factor and leukemia inhibitory factor promote the generation, maturation and survival of oligodendrocytes in vitro. Development 120, 143–153. McCarthy, G. F., and Leblond, C. P. (1988). Radioautographic evidence for slow astrocyte turnover and modest oligodendrocyte production in the corpus callosum of adult mice perfused with 3H-thymidine. J. Comp. Neurol. 271, 589–603. McManus, M. F., Chen, L-C, Vallejo, I., and Vallejo, M. (1999). Astroglial diVerentiation of cortical precursor cells triggered by activation of the cAMP-dependent signaling pathway. J. Neurosci. 19, 9004–9015. Mi, H., and Barres, B. A. (1999). PuriWcation and characterization of astrocyte precursor cells in the developing rat optic nerve. J. Neurosci. 19, 1049–1061. Mi, H. M., Haeberle, H., and Barres , B. A. (2001). Induction of astrocyte diVerentiation by endothelial cells. J. Neurosci. 21, 1538–1547. Milosevic, A., and Goldman, J. E. (2002). Progenitors in the postnatal cerebellar white matter are antigenically heterogeneous. J. Comp. Neurol. 452, 192–203. Miller, R. H., and Szigeti, V. (1991). Clonal analysis of astrocyte diversity in neonatal rat spinal cord cultures. Development 113, 353–362. Misson, J.-P., Austin, C. P., Takahashi, T., Cepko, C., and Caviness, V. S., Jr. (1991a). The alignment of migrating neural cells in relation to the murine neopallial radial glial Wber system. Cerebral Cortex 1, 221–229. Misson, J.-P., Edwards, M. A., Yamamoto, M., and Caviness, V. S., Jr. (1988). IdentiWcation of radial glial cells within the developing murine central nervous system: Studies based upon a new immunohistochemical marker. Dev. Brain. Res 44, 95–108. Misson, J.-P., Takahashi, T., and Caviness, V. S., Jr. (1991b). Ontogeny of radial and other astroglial cells in murine cerebral cortex. Glia 4, 138–148. Miyake, T., Fujiwara, T., Fukunaga, T., Takemura, K., and Kitamura, T. (1995). Glial cell lineage in vivo in the mouse cerebellum. Develop. Growth DiVer. 37, 273–285. Moskovkin, G. N., Fulop, Z., and Hajos, F. (1978). Origin and proliferation of astroglia in the immature rat cerebellar cortex. A double label autoradiographic study. Acta Morphol. Acad. Sci. Hung. 26, 101–106.

II. GLIAL CELL DEVELOPMENT

327

FUTURE PERSPECTIVES AND UNRESOLVED QUESTIONS

Mori, K., Ikeda, J., and Hayaishi, O. (1990). Monoclonal antibody R2D5 reveals midsagittal radial glial system in postnatally developing and adult brainstem. Proc. Natl. Acad. Sci. USA 87, 5489–5493. Nakashima, K., Takizawa, T., Ochiai, W., Yanigisawa, M., Hisatsune, T., Hakafuku, M., Miyazono, K., Kishimoto, T., Kageyama, R., and Taga T. (2001). BMP2-mediated alteration in the developmental pathway of fetal mouse brian cells from neurogenesis to astrocytogenesis. Proc. Natl. Acad. Sci. USA 98, 5868–5873. Nakashima, K., Wiese, S., Yanagisawa, M., Arakawa, H., Kimura, N., Hisatsune, T., Yoshida, K., Kikshimoto, T., Sendtner, M., and Taga, T. (1999a). Developmental requirement of gp130 signaling in neuronal survival and astrocyte diVerentiation. J. Neurosci. 19, 5429–5434. Nakashima, K., Yanagisawa, M., Arakawa, H., Kimura, N., Hisatsune, T., Kawabata, M., Miyazono, K., and Taga, T. (1999b). Synergistic signaling in fetal brain by STAT3-Smadl complex bridged by p300. Science 284, 479–482. Nakashima, K., Yanagisawa, M., Arakawa, H., and Taga, T. (1999c). Astrocyte diVerentiation mediated by LIF in cooperation with BMP2. FEES Lett. 457, 43–46. Nieto, M., Schuurmans, C., Britz, O., and Guillemot, F. (2001). Neural bHLH genes control the neuronal versus glial fate decision in cortical progenitors. Neuron 29, 401–413. Noctor, S. C., Flint, A. C., Weissman, T. A. I., Dammerman, R. S., and Kriegstein, A. R. (2001). Neurons derived from radial glial cells establish radial units in neocortex. Nature 409, 714–720. Norton, W. T. (1999). Cell reactions following acute brain injury: A review. Neurochem. Res. 24, 213–218. Ochiai, W., Yanagisawa, M., Takizawa, T., Hakashima, K., and Taga, T. (2001). Astrocyte diVerentiation of fetal neuroepithelial cells involving cardiotrophin-1 -induced activation of STATS. Cytokine 14, 264–271. Palay, S. L., and Chan-Palay, V. (1974). ‘‘Cerebellar cortex, Cytology, and Organization.’’ Springer-Verlag, New York. Parnevales, J. G. (1999). Glial cell lineages in the rat cerebral cortex. Exp. Neurol. 156, 418–429. Paterson, J. A. (1983). Dividing and newly produced cells in the corpus callosum of adult mouse cerebrum as detected by light microscopic radioautography. Anat. Anz., Jena 153, 149–168. Paterson, J. A., Privat, A., Ling, E. A., and Leblond, C. P. (1973). Investigation of glial cells in semithin sections III. Transformation of subependymal cells into glial cells as shown by radioautography afterH-thymidine injection into the lateral ventricle of the brain of young rats. J. Comp. Neurol. 149, 83–102. Peters, A., Palay, S. L., and Webster, H. deF. (1991). ‘‘The Fine Structure of the Nervous System.’’ 3rd ed. Oxford University Press, Oxford. Price, J., and Thurlow, L. (1988). Cell lineage in the rat cerebral cortex: A study using retroviral-mediated gene transfer. Development. 104, 473–482. Ransom, B. R. (1995). Gap junctions. In ‘‘Neuroglia’’ (H. Kettenmann, and B. Ranson, eds.), pp. 299–318. Oxford University Press, London. Pringle, N. P., Guthrie, S., Lumsden, A., and Richardson, W. D. (1998). Dorsal spinal cord neuroepithelium generates astrocytes but not oligodendrocytes. Neuron 20, 883–893. Pringle, N. P., Yu, W-P., Howell, M., Colvin, J., Ornitz, D. M., and Richardson, W. D. (2003). Fgfr3 expression by astrocytes and their precursors: evidence that astrocytes and oligodendrocytes originate in distinct neuroepithelial domains. Development 130, 93–102. Rao, M. S., Noble, M., and Mayer-Proschel, M. (1998). A tripotential glial precursor cells is present in the developing spinal cord. Proc. Natl. Acad. Sci. USA 95, 3996–4001. Reichenbach, A. (1990). Radial glial cells are present in the velum medullare of adult monkeys. J. Hirnforsch. 31, 2690–271. Reyners H., Gianfelici de Reyners E., Regniers L., Maisin J. R. (1986). A glial progenitor cell in the cerebral cortex of the adult rat. J. Neurocytol. 15, 53–61. Reynolds, B. A., and Weiss, S. (1992). Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 255, 1707–1710. Richards, L. J., Kilpatrick, T. J., and Bartlett, P. P. (1992). De novo generation of neuronal cells from the adult mouse brain. Proc. Natl. Acad. Sci. USA 89, 8591–8595. Rowitch, D. H., Lu, R. Q., Kessaris, N., and Richardson, W. D. (2002). An ‘‘oligarchy’’ rules neural development. Trends Neurosci. 25, 417–422. Schmechel, D. E., and Rakic, P. (1979a). Arrested proliferation of radial glial cells during midgestation in rhesus monkey. Nature 277, 303–305. Schmechel , D. E., and Rakic, P. (1979b). A Golgi study of radial glial cells in developing monkey telencephalon: morphogenesis and transformation into astrocytes. Anat. Embryol. (Berl.) 156, 115–152. Seri, B., Garcia-Verdugo, J. M., McEwen, B. S., and Alvarez-Buylla, A. (2001). Astrocytes give rise to new neurons in the adult mammalian hippocampus. J. Neurosci. 21–7153–7161. ShaWt-Zagardo, B., Iwaki, A. K., and Goldman, J. E. (1988). Astrocytes regulate GFAP mRNA levels by cAMP and protein kinase C dependent mechanisms. Glia 1, 346–354. Shibata, T., Yamada, K., Watanabe, M., Ikenaka, K., Wada, K., Tanaka, K., and Inoue, Y. (1997). Glutamate transporter GLAST is expressed in the radial glial-astrocyte lineage of developing mouse spinal cord. J. Neurosci. 17, 9212–9217. Staugaitis, S. M., Zerlin, M., Levine, J. M., Hawkes, R., and Goldman, J. E. (2001). Aldolase C/Zebrin II expression in neonatal rat forebrain reveals cellular heterogeneity within the subventricular zone and early astrocyte diVerentiation. J. Neurosci. 21, 6195–6205.

II. GLIAL CELL DEVELOPMENT

au2

328

12. ASTROCYTE LINEAGE

au3

Stockli, K. A., Lillien, L. E., Haher-Noe, M., Brietfeld, G., Hughes, R. A., RaV, M. C., Thoenen, H., and Sendtner, M. (1991). Regional distribution, developmental changes, and cellular localization of CNTFmRNA and protein in the rat brain. J. Cell Biol. 115, 447.459. Sun, Y., Nadal-Vicens, M., Misono, S., Lin, M. Z., Zubiaga, A., Hua, X., Fan, G., and Greenberg, M. E.(2001). Neurogenin promotes neurogenesis and inhibits glial diVerentiation by independent mechanisms. Cell 104, 365–376. Suzuki, S. O. and Goldman, J. E. (2003) Multiple cell populations in the early postnatal subventricular zone take distinct migratory pathways: a dynamic study of glial and neuronal progenitor migration. J. Neurosci. 23, 4240–4250. Takebayashi, H., Nabeshima, Y., Yoshida, S., Chisaka, O., Ikenanka, K., and Nabeshima, Y. (2002). The basic helix-loop-helix factor olig 2 is essentional for the development of motoneuron and oligodendrocyte lineages. Curr. Biol. 12, 1157–1163. Tohyama, T., Lee, V. M.-Y., Rorke, L. B., Marvin, M., McKay, R. D. G., and Trojanowski, J. Q. (1992). Nestin expression in embyonic human neuroepithelium and in human neuroepithelial tumor cells. Lab. Investig. 66, 303–313. Tout, S., Dreher, Z., Chan-Ling, T., and Stone, J. (1993). Contact spacing between astrocytes is independent of neighboring structures: in vivo and in vitro evidence. J. Comp. Neurol. 332, 433–443. Turner, D., and Cepko, C. (1987). Cell lineage in the rat retina: a common progenitor for neurons and glia persists late in development. Nature, Lond. 328, 131–136. Vaughan, S. W., and Peters, A. (1974). Neuroglial cells in the cerebral cortex of rats from young adulthood to old age: An electron microscopic study. J. Neurocytol. 3, 405–429. Ventura, R., and Harris, K. M. (1999). Three-dimensional relationships between hippocampal synapses and astrocytes. J. Neurosci. 19, 6897–6906. Voigt, T. (1989). Development of glial cells in the cerebral wall of ferrets: Direct tracing of their transformation from radial glia into astrocytes. J. Comp. Neurol. 289, 74–88. Walsh, C., and Cepko, C. L. (1992). Widespread dispersion of neuronal clones across function regions of the cerebral cortex. Science 255, 434–440. Walsh, C., and Cepko, C. L. (1993). Clonal dispersion in proliferating layers of developing cerebral cortex. Nature 362, 632–635. Williams, B. P., and Price, J. (1995). Evidence for multiple precursor cell types in the embryonic rat cerebral cortex. Neuron. 14, 1181–1188. Yamada, K., Fukaya, M., Shibata, T., Kurihara, H., Tanaka, K., Inoue, Y., and Watanabe, M. (2000). Dynamic transformation of Bergmann glial Wbers proceeds in correlation with dendritic outgrowth and synapse formation of cerebellar Purkinje cells. J. Comp. Neurol. 418, 106–120. Yamada, K., Watanabe, M., Shibata, T., Nagashima, M., Tanaka, K., and Inoue, Y. (1998). Glutamate transporter GLT-1 is transiently localized on growing axons in the mouse spinal cord before establishing astrocytic expression. J. Neurosci. 18, 5706–5713. Yamamori, T. (1991). Localization of cholinergic diVerentiation factor/leukemia inhibitory factor mRNA in the rat brain and peripheral tissues. Proc. Natl. Acad. Sci., USA 88, 7298–7302. Yanagisawa, M., Nakashima, K., and Taga, T. (1999). STAT3-mediated astrocyte diVerentiation from mouse fetal neuroepithelial cells by mouse oncostatin M. Neurosci. Lett. 269, 169–172. Yanagisawa, M., Nakashima, K., Takizawa, T., Ochiai , W., Arakawa, H., and Taga, T. (2001). Signaling crosstalk underlying synergistic induction of astrocyte diVerentiation by BMPs and IL-6 family of cytokines. FEBS Lett. 489, 139–143. Yun, K., Fischman, S., Johnson, J., deAngelis, M. H., Weinmaster, G., and Rubenstein, J. L. R. (2002). Modulation of the notch signaling by Mashl and Dlxl/2 regulates sequential speciWcation and diVerentiation of progenitor cell types in the subcortical telencephalon. Development 129, 5029–5040. Zerlin, M., and Goldman, J. E. (1997). Interactions between glial progenitors and blood vessels during early postnatal corticogenesis: Blood vessel contact represents an early stage of astrocyte diVerentiation. J. Comp. Neurol. 387, 537–546. Zerlin, M., Levison, S. W., and Goldman, J. E. (1995). Early patterns of migration, morphogenesis, and intermediate Wlament expression of subventricular zone cells in the postnatal rat forebrain. J. Neurosci. 15, 7238–7249. Zerlin, M.A, Milosevic, A., and Goldman, J.E. (2004). Glial progenitors of the neonatal subventricular zone diVerentiate asychronously, leading to spatial dispersion of glial clones and to the persistence of immature glia in the adult mammalian CNS, (submitted.) Zhang, L., and Goldman, J. E. (1996). Developmental fates and migratory pathways of dividing progenitors in the postnatal rat cerebellum. J. Comp. Neurol. 370, 536–550. Zhou, Q., and Anderson, D. J. (2002). The bHLH transcription factors Olig2 and Oligl couple neuronal and glial subtype speciWcation. Cell 109, 61–73.

II. GLIAL CELL DEVELOPMENT

C H A P T E R

13 Schwann Cell Development K. R. Jessen and R. Mirsky

INTRODUCTION This chapter focuses on glial development in embryonic and early neonatal nerve trunks of rodents. We examine the origin of the Schwann cell lineage and the cellular transformations that lead to the appearance of immature Schwann cells, the predominant glial cell in rodent nerves at birth. The extracellular signaling molecules and matrix receptors that control this system are discussed, and we describe the transcription factors that are known, or suspected, to be important in regulating embryonic Schwann cell development and the onset of myelination. While the myelinating and nonmyelinating Schwann cells form two major categories of peripheral glia, other distinct and important types of peripheral glia exist. These include olfactory ensheathing cells, the teloglia (terminal glia) of somatic motor nerve terminals, satellite cells that envelop neuronal cell bodies in sympathetic, parasympathetic and sensory ganglia, the astrocyte-like enteric glial cells in the autonomic ganglia in the gut wall, and specialized glia found in association with sensory endings such as Pacinian corpuscles (Chuah and West, 2002; Franklin and Barnett, 2000; Gabella, 1981; Gershon, 1998; Jessen and Mirsky, 1983; Landon and Wiseman, 2001; Lubischer and Thompson, 1999; Pannese, 1981, 1994; Raisman, 2001; Ramon-Cueto and Santos-Benito, 2001; Robitaille, 1998; Spencer and Schaumburg, 1973; Zelena, 1994). It is likely that the diVerences between all of these cell classes are acutely dependent on the cellular environment and speciWc location in which they are found, and that the glial cells of the peripheral nervous system (PNS) retain an exceptional degree of plasticity throughout life.

THE SCHWANN CELL LINEAGE: AN OVERVIEW In comparison with many other systems, the lineage of the Schwann cells found in peripheral nerve trunks is simple and well understood in general outline (for reviews, see Jessen and Mirsky, 1999; Lobsiger et al., 2002; Mirsky and Jessen 1996, 1999; Scherer, 1997; Topilko and Meijer, 2001) (Figs. 13.1, 13.2, and 13.3). The origin of these cells is in the neural crest, while the outcome of the developmental process is just two distinct cells, the mature myelin-forming and nonmyelin-forming Schwann cells. Between migrating crest cells and mature Schwann cells lie three main developmental transitions: Wrst, the formation of Schwann cell precursors, second, the formation of immature Schwann cells, and, lastly, the reversible generation of myelinating and nonmyelinating cells. Neural crest cells generate several other cell types besides glia, and immature Schwann cells will either myelinate or not. Therefore, the Wrst and last of these transition points involve fate-choice.

Myelin Biology and Disorders, Volume 1

329

Copyright 2004, Elsevier (USA). All rights reserved.

330

13. SCHWANN CELL DEVELOPMENT

FIGURE 13.1 The Schwann cell lineage in rat and mouse. There are three main transitions in the Schwann cell lineage. I: The formation of Schwann cell precursors from undiVerentiated migrating neural crest cells. This is characterized by the appearance of several diVerentiation markers (a subclass of which is shared by early neurons) and by the assumption of an intimate relationship with axons. II: The precursor-Schwann cell transition. This is characterized by the appearance of a number of diVerentiation markers (some of which are acutely dependent on axonal contact and remain readily reversible) and the establishment of autocrine survival circuits. III: The formation of mature myelinating and nonmyelinating Schwann cells. This involves radical morphological and molecular changes, particularly in myelinating cells; nevertheless, this transition is readily reversible when axonal contact is lost (stippled arrows) and is reestablished in regenerating nerves.

This does not apply to the second step since, apart from programmed cell death, presumably the only fate of Schwann cell precursors in normal development is to become Schwann cells. The extracellular signals that control fate choice in this system have not been established. The mechanisms that enable or direct crest cells to enter the glial lineage are still unclear, and the axon-associated signals that selectively instruct those cells associated with larger diameter axons to myelinate are not known. In contrast to this lack of deWnitive information about signaling at the choice points, both in vitro and in vivo experiments have identiWed two factors, neuregulin-1 and endothelin, that regulate lineage progression, survival and proliferation of Schwann cells in embryonic nerves (Brennan et al., 2000; Garratt et al., 2000a; Jessen and Mirsky, 1999). Two intracellular signals are known to be essential for Schwann cell development. These are the transcription factors Sox-10, which is needed for the establishment of the Schwann cell lineage, and Krox-20, which is needed for myelination (Britsch et al., 2001; Topilko and Meijer, 2001). In addition, the transcription factor Oct-6 has a major role in the timing of myelination (Topilko and Meijer, 2001). Schwann cell precursors and immature Schwann cells are rapidly proliferating cell populations in vivo with DNA incorporation being most rapid in immature Schwann cells (Stewart et al., 1993). The diVerentiation process that transforms the phenotype of crest cells, Wrst to that of Schwann cell precursors and then to that of immature Schwann cells, is therefore compatible with cell division. It is not until the last transition to form myelinating and nonmyelinating cells that developing Schwann cells exit the cell cycle. The ready reversibility of this last step of Schwann cell development is a notable feature of the Schwann cell lineage. The process can be triggered by removing Schwann cells from contact with axons either by injuring nerves in vivo or by dissociating cells from adult nerves and placing them in culture without neurons. Both in vivo and in vitro, it involves the de-diVerentiation or developmental regression of individual Schwann cells and myelin breakdown. As the cells lose the molecular markers and structural features that characterize myelinating and nonmyelinating cells, they reenter the cell cycle and revert back to adopt a phenotype similar, although not identical, to that of immature Schwann cells. Such denervated Schwann cells readily rediVerentiate when they are allowed to reassociate with axons, for example, during nerve regeneration (Fawcett and Keynes, 1990; Scherer and Salzer, 2001). Do other cells in the lineage also show comparable phenotypic plasticity? Reversal of Schwann cell diVerentiation beyond the immature Schwann cell stage and the re-formation of

II. GLIAL CELL DEVELOPMENT

THE SCHWANN CELL LINEAGE: AN OVERVIEW

331

FIGURE 13.2 Cells of the Schwann cell lineage as they appear during peripheral nerve development. Top panel: Four Schwann cell precursors (P1-P4) in the hindlimb nerves of a rat embryo at E15. The nuclei of three of the precursors (P1, P3, and P4) are visible. Schwann cell precursors form close contacts (arrows) with each other and are either embedded between axons inside the nerve, as shown here, or found in close apposition to axons at the surface of the nerve. Because connective tissue spaces and extracellular matrix are essentially absent, these nerves are much more compact than older ones. The axons also have a smaller and more uniform diameter than those seen in mature nerves. Courtesy of Y. Hashimoto. Bar: 3 mm. Center panel: Schwann cells in the sciatic nerve of a newborn mouse (NB). Immature Schwann cells (S**) communally envelop a group of axons of various diameters. Other cells (S) are at a very early stage of myelination, the pro-myelination stage, having formed a 1:1 relationship with large axons (A). S* has progressed slightly farther along the myelin lineage and is starting to form a compact sheath around the axon (A*). Part of a thin myelin sheath can be seen on the extreme right-hand side of the Weld (arrow). Note that in contrast to the compact arrangement seen earlier in development, perinatal nerves contain considerable collagen (C) and extracellular space. Bar: 1.5 mm. Bottom panels :Mature Schwann cells in transverse section of adult rat sciatic nerve (AD). Left: A myelinating Schwann cell forms a compact multilayered sheath (M) around a single large diameter axon (A). Parts of other myelinated Wbers are present. Bar: 1 mm. Right panel: Cross section through the nucleus of a nonmyelinating Schwann cell (N–M) that ensheathes 13 axons (for example, A) each lying in a separate trough in the cell surface. Nonmyelinating Schwann cells can also ensheathe singly, as shown for A*. Myelin sheaths (M) of neighboring axons are visible, and the axon-Schwann cell units are surrounded by collagen-rich extracellular spaces (C). Middle and lower panels courtesy of R. M. King and P. K. Thomas. Bar: 1 mm.

II. GLIAL CELL DEVELOPMENT

332

13. SCHWANN CELL DEVELOPMENT

FIGURE 13.3 DiVerentiation markers in embryonic Schwann cell development. Grey: Markers shared by migrating, undiVerentiated neural crest cells, Schwann cell precursors, and immature Schwann cells. Blue: Markers present on migrating crest cells and precursors but not (or at very low levels) on immature Schwann cells. Green: Markers present on Schwann cell precursors and immature Schwann cells, but absent from migrating undiVerentiated crest cells. Single asterisk*: These proteins also appear on neuroblasts/early neurons. Pink: Markers expressed by immature Schwann cells but not (or at much lower levels) by Schwann cell precursors. Double asterisk**: These markers are acutely dependent on axons for expression. In addition Schwann cells have autocrine survival circuits not present in precursors. For data on survival, see Woodhoo et al., 2003.

Schwann cell precursors has not been described, either because it does not take place or because the number of cells involved is too few for the phenomenon to have been noticed. Although it has been suggested that Schwann cell precursors are likely to be irreversibly committed to the Schwann cell lineage (Morrison et al., 2000), there are several notable examples of speciWed cells, including early developing Schwann cells in the chick, showing unexpected plasticity and ability for respeciWcation (Dupin, et al., 2003; French et al., 2002; Tosh and Slack, 2002). It is therefore prudent to be open-minded about the developmental potential of Schwann cell precursors. Clearly these cells have been speciWed as glial cells, showing a phenotype that diVerentiates them unambiguously from both neural crest and early developing neurons. Furthermore, in their normal signaling environment (i.e., developing nerve trunks), they are likely only to have a Schwann cell fate. Nevertheless, it is possible that exposure to alternative signals could respecify Schwann cell precursors and divert their fate in other directions. None of the three main transitions in Schwann cell development may extend to the entire cell population. Thus, it has been argued that in addition to Schwann cell precursors, E14 rat nerves also harbor a separate population of cells (10 to 15% of total cell number) that shows similarities to neural crest cell phenotype (Morrison et al., 1999, 2000). Some precursor-like cells might remain in perinatal nerves (Bixby et al., 2002) and even mature nerves might hide a population of cells with an early phenotype corresponding to precur-

II. GLIAL CELL DEVELOPMENT

THE ORIGIN OF SCHWANN CELLS IN THE NEURAL CREST

sors or immature Schwann cells (Rizvi et al., 2002). The retention in older, even adult, tissues of a cell population with a phenotype and developmental potential that resemble early cells is seen in many other tissues, including the central nervous system (CNS), enteric nervous system muscle, and the hematopoietic system.

THE ORIGIN OF SCHWANN CELLS IN THE NEURAL CREST The neural crest is a transient group of cells that delaminates from the dorsal part of the neural tube during embryonic development. In the trunk region, cells of the neural crest give rise to glial cells, neurons of sensory, sympathetic, and parasympathetic ganglia, chromaYn cells, and melanocytes. Crest cells in the most anterior part of the trunk, the cardiac crest, also give rise to connective tissue and smooth muscle cells, while crest cells in the head region, the cephalic crest, have the additional potential of generating cartilage and bone cells (Le Douarin and Kalcheim, 1999). Some prospective glial cells in the crest may have already entered the glial lineage at the onset of crest migration, while other cells are likely to start glial development later as they migrate ventrally along the neural tube (Henion and Weston, 1997). It is not clear what cell-extrinsic signals initiate glial development from crest cells or whether the signals responsible for the early and late entry to the glial lineage are the same. As mentioned earlier, two growth factors, neuregulin-1 and endothelin (Brennan et al., 2000; Garratt et al., 2000a), are involved in regulating early Schwann cell development in embryonic nerves. At present it does not seem likely, however, that these factors are part of an instructive signaling mechanism needed for triggering glial development from multipotent crest cells in vivo. The roles of neuregulin-1 is discussed in the sections titled ‘‘The Role of Neuregulin1 in Schwann Cell Development,’’ and endothelin is discussed in the section titled ‘‘The Regulation of the Precurson-Schwann Cell Transition.’’ Cell culture studies indicate that bone morphogenetic proteins (BMPs), which are important for the generation of sympathetic neurons (Schneider et al., 1999; Shah et al., 1996), may act as negative regulators of gliogenesis during crest development (Shah et al., 1996). The possibility that Notch activation promotes gliogenesis has been raised in recent studies (Furukawa et al., 2000; Gaiano et al., 2000; Kubu et al., 2002; Morrison et al., 2000; Wakamatsu et al., 2000). While enforced Notch activation has been shown to promote the emergence of glial cells in the developing CNS in vivo (Furukawa et al., 2000; Gaiano et al., 2000), the evidence for Notch signaling in PNS gliogenesis is less complete. What seems clear is that Notch activation suppresses the generation of neurons from rat and avian neural crest cells both in vivo and in vitro (Kubu et al., 2002; Morrison et al., 2000; Wakamatsu et al., 2000), an observation that is in line with previous Wndings in the CNS (Wakamatsu et al., 1999, and references therein). There is also evidence for asymmetrical segregation of NUMB, a likely Notch antagonist, during division of avian crest cells, and it has been suggested that this would bias cells with higher levels of NUMB toward neuronal development within sensory ganglia. This prediction has received indirect support from the Wnding that without NUMB, dorsal root sensory ganglion (DRG) neurons fail to form in mice (Zilian et al., 2001). It is intriguing that sympathetic neurons appear to be generated normally in these animals. Suppression of neurogenesis from neural crest cells in vitro is a function that Notch activation shares with neuregulin-1 (Shah et al., 1994; also see the section titled ‘‘The Role of Neuregulin-1 in Schwann Cell Development’’). In another parallel with neuregulin, it has proved more problematic to show that Notch activation inductively triggers glial diVerentiation from neural crest cells. In migrating neural crest cells, activation of Notch by application of a soluble form of the Notch ligand delta promotes glial diVerentiation measured as an increase in the number of clones that contain glial Wbrillary acidic protein (GFAP) positive cells in clonal analysis experiments (Kubu et al., 2002; Morrison et al., 2000). In these experiments, glial diVerentiation, albeit at lower levels, takes place without delta application showing that gliogenic signals are available at least to some of the cells without this Notch activation. Although it is possible that Notch signaling directly triggers

II. GLIAL CELL DEVELOPMENT

333

334

13. SCHWANN CELL DEVELOPMENT

glial diVerentiation under these conditions, it is equally possible that the function of Notch activation is to promote the action of other gliogenic signals. Blocking entry to nonglial lineages, an established function of Notch (discussed earlier), provides one possible mechanism since this is likely to prolong the period during which crest cells would be responsive to any instructive gliogenic signals present in the culture system. In conclusion, there is evidence that in the developing neural crest, Notch activation in neighboring cells by early developing neurons prevents excessive neurogenesis without blocking gliogenesis. It remains unclear whether Notch activation acts also as an inductive signal to trigger gliogenesis or whether Notch promotes glial diVerentiation only indirectly. Experiments on cells from E14 rat nerves have raised the additional possibility that Notch signaling promotes the generation of Schwann cells from Schwann cell precursors (see the section titled ‘‘The Developmental Potential of Early PNS Glia’’). The transcription factor Sox-10 appears to be intimately involved in the development of glia from crest cells (Britsch et al., 2001; Kuhlbrodt et al., 1998b; Paratore et al., 2001; Peirano et al., 2000; Sonnenberg-Riethmacher et al., 2001; Southard-Smith et al., 1998; Wegner, 2000). Other members of the Sox family are involved in developmental regulation in a number of diverse systems (Wegner, 1999). At the time of crest formation, expression of Sox-10 mRNA is activated in restricted areas of the closing neural tube corresponding to the sites of origin of migrating crest cells (Kuhlbrodt et al., 1998b). Subsequently Sox-10 mRNA is seen in migrating crest cells (Britsch et al., 2001; Pusch et al., 1998; SouthardSmith et al., 1998) and in vitro studies indicate that all of these cells express Sox-10 protein (Paratore et al., 2001). Sox-10 expression is down-regulated in early developing neurons but persists in developing glia both in ganglia and along nerve trunks (Kuhlbrodt et al., 1998b). By the time of birth, the levels of Sox-10 mRNA in peripheral nerves are, however, much lower than those in neural crest cells (U. Lange and K. R. Jessen, unpublished). The eVects of Sox-10 inactivation have been studied both in the Dominant megacolon (Dom) mice that have spontaneous nonsense or frameshift mutations in the Sox-10 gene, and in mice in which the Sox-10 gene has been inactivated. In these mice, early peripheral glial cells are missing (Britsch et al., 2001). This applies equally to satellite cells within DRGs and Schwann cell precursors in peripheral nerves. These cells were identiWed by expression of brain-speciWc fatty acid binding protein (B-FABP). In the mouse, B-FABP is a useful marker for studying early glial development in the PNS because it distinguishes early glial cells, which express B-FABP protein, from neural crest cells and early neurons, which are B-FABP negative (see the section titled ‘‘DiVerentiation Markers in Embryonic Schwann Cell Development’’). B-FABP positive glia appear in DRGs at embryo day E10–11 in the mouse and in sympathetic ganglia at E12. B-FABP positive cells appear along peripheral nerve trunks at E11. This is consistent with the identiWcation of Schwann cell precursors in mouse nerves at E12, the earliest time point studied, and E13 (Dong et al., 1999). Although B-FABP positive glial precursors are missing in Sox-10 deWcient mice, DRG sensory neurons, identiWed by expression of ß-III tubulin, are initially generated in normal numbers, although they die later (Sonnenberg-Riethmacher et al., 2001). This points to particular functions for Sox-10 in glial development. It is likely that one of these is to maintain the expression of the neuregulin-1 receptor ErbB3. ErbB3 is found in crest cells as they emerge from the neural tube and in early glial cells. In Sox-10 mutant mice, the expression of ErbB3 receptors is initiated normally in crest cells, but the receptors are not maintained as the crest cells migrate away from the dorsal neural tube, commence neuronal or glial diVerentiation, and condense to form DRGs (Britsch et al., 2001). Although neuregulin-1 does not appear to be important for the survival of migrating crest cells in vivo (Britsch et al., 1998), neuregulin-1 signaling and ErbB3 receptors are essential for the survival of Schwann cell precursors in nerve trunks, and neuregulin-1 also acts as a mitogen for these cells (Britsch et al., 1998; Dong et al., 1995, 1999). Therefore the expected eVect of the ErbB3 down-regulation seen in Sox-10 mutants would be increased death and decreased proliferation of those crest cells that start glial diVerentiation. This could lead to selective reduction in the number of glial cells and contribute to the common Wnding in Sox-10 mutants and ErbB3 mutants that the number of Schwann cell precursors is severely reduced in nerve trunks. This mechanism could also

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL PRECURSORS

contribute to the lack of B-FABP positive early glial cells within DRGs in Sox-10 deWcient mice. Non-neuronal cells within condensing ganglia of these animals show increased cell death and decreased proliferation, although cell death in migrating crest cells prior to ganglionic condensation is normal (Britsch et al., 2001; Paratore et al., 2001; SonnenbergRiethmacher et al., 2001). Furthermore, it has been shown directly in vitro that Sox-10 is required for neuregulin-1 mediated survival of p75 positive non-neuronal cells—that is, early glial cells or undiVerentiated crest cells, from mouse E13 DRG (Paratore et al., 2001). The lack of neuregulin-1 signaling cannot, however, be the major reason for the absence of glia in Sox-10 mutant DRGs because DRG glia develop apparently normally in mice in which neuregulin-1 signaling has been inactivated (see the section titled ‘‘The Role of Neuregulin-1 in Schwann Cell Development’’). Rather, the explanation might lie in the interesting possibility that Sox-10 is required for glial speciWcation—that is, the process by which crest cells change to become early glia (Britsch et al., 2001). Early DRGs of Sox-10 mice in which B-FABP positive glial cells are not seen (noted earlier) contain, in addition to neurons, a residual population of cells that appear to retain the phenotype of neural crest cells and some cells of this kind are seen along nerve trunks (Britsch et al., 2001; Sonnenberg-Riethmacher et al., 2001). This is consistent with the idea that without Sox-10 glial speciWcation is blocked although crest cells thrive and can form neurons. This view has received indirect support from cell culture experiments that demonstrate that there can be a link between Sox-10 and lineage decisions in crest-derived cells (Paratore et al., 2001). This study showed that p75 positive non-neuronal cells from E13 mouse DRG of Sox-10 mutants, presumably undiVerentiated neural crest cells (Britsch et al., 2001), tend to leave the neural lineage and take on a smooth muscle-like phenotype, one of the cell types in the crest repertoire, under culture conditions in which cells from wild-type DRG remained in the glial lineage. A further link between Sox-10 and diVerentiation is provided by the Wnding that enforced expression of Sox-10 induces expression of the endogenous P0 gene, a marker of crest cell glial diVerentiation (see the section titled ‘‘DiVerentiation Markers in Embryonic Schwann Cell Development’’), in the neuroblastoma cell line N2A (Peirano et al., 2000). Thus, the eVects of Sox-10 mutations on early glial development can be explained by at least two plausible mechanisms relating on the one hand to neuregulin-1 signaling and, on the other, to glial speciWcation. Mice with mutations in the Sox-10 gene also have deWciencies in pigmentation and in the enteric nervous system, indicating a widespread role for this gene in the development of neural crest derivatives (Wegner, 1999).

SCHWANN CELL PRECURSORS The Wrst clearly deWned stage of glial diVerentiation in peripheral nerve trunks is the Schwann cell precursor. Schwann cell precursors represent the large majority of cells in the limb nerves of E14/15 rats (E12/13 mice). Corresponding cells in dorsal and ventral roots and in early ganglia may not have an identical phenotype and may diVer in the timing of appearance of glial diVerentiation markers (Woodhoo et al., 2003). The precursors are initially found mostly at the edge of nerves but are also seen within more mature nerve branches. They are always in close apposition to axons. They possess extensive sheet-like processes that contact and form junctions with processes from neighboring cells, thereby surrounding large groups of axons, most of which are of similar size at this age (Fig. 13.2). Unlike older nerves, where within a nerve fascicle axon-Schwann cell units are surrounded by abundant Wbrous connective tissue containing blood vessels, E14 rat or E12 mouse nerves contain no signiWcant connective tissue or vessels. From E17/18 onward in the rat (mouse E15/16) nearly all cells in peripheral nerves are Schwann cells. The precursor/Schwann cell transition in the hind limb nerves of rodents therefore centers on the 16th embryonic day, and it is only at this time that substantial numbers of the two cell types, precursors and Schwann cells, can be isolated from limb nerves. The idea of diVerentiating between Schwann cells and Schwann cell precursors came initially from a study of cell survival and expression of the calcium binding protein S100 in embryonic rat nerves. In a comparison of cells from E14 and E18 nerves it was found that

II. GLIAL CELL DEVELOPMENT

335

336

13. SCHWANN CELL DEVELOPMENT

the E14 cells showed very little cytoplasmic S100 immunoreactivity relative to the older cells. Furthermore, the E14 cells died rapidly in vitro without added survival factors, irrespective of cell density, while Schwann cells from E18 and older nerves survived at moderate or higher densities without added factors. This diVerence in survival regulation was later shown to be the presence of autocrine survival circuits in Schwann cells. This mechanism is absent from Schwann cell precursors, which for survival rely instead on paracrine neuregulin-1 signaling from axons (Jessen et al., 1994; Meier et al., 1999). Subsequent work has revealed more clearly the distinct properties of Schwann cell precursors. These cells can be diVerentiated both from the migrating crest cells from which they derive and from the immature Schwann cells that they generate on the basis of a number of markers and other features (Fig. 13.3). Schwann cell precursors are therefore speciWed glial cells. Presumably they are also unifated, namely destined only to become Schwann cells during normal development. This does not necessarily imply that Schwann cell precursors are determined or irreversibly committed to a glial fate (see the section titled ‘‘The Developmental Potential of Early PNS Glia’’). While the essential function of Schwann cell precursors is to generate Schwann cells, an important additional role for these cells has been suggested in studies on mutant mice lacking isoform III of neuregulin-1, the ErbB3 neuregulin receptor, or the transcription factor Sox-10. In all of these mutant strains, Schwann cell precursors are missing or their number severely reduced. In these animals, DRG neurons and motor neurons are initially generated normally but by E14 and E18, respectively, the majority of these cells have died at limb levels of the spinal cord. It has been suggested that this cell death is due to the absence of glial-derived factors that are needed for the survival of sensory and motor neurons (Britsch et al., 2001; Riethmacher et al., 1997). This raises the possibility that a key function of Schwann cell precursors and early Schwann cells is to regulate the survival of discrete pools of embryonic CNS and PNS neurons. In the neuregulin mutants, loss of axonal contact with peripheral targets may also be an important contributor to sensory and motor neuron death (Wolpowitz et al., 2000).

DIFFERENTIATION MARKERS IN EMBRYONIC SCHWANN CELL DEVELOPMENT Much recent analysis of post-natal Schwann cell development and myelination has involved the use of a set of well-deWned molecular markers that characterize distinct stages in the development of myelinating and nonmyelinating cells. There is still a shortage of comparable tools with which to study initial glial diVerentiation in the PNS. Many more markers have been available to monitor the early development of neurons. These include neuron-speciWc tubulin (ß-III tubulin; TUJ1), neuroWlament, peripherin, and early markers of distinct peripheral neuronal lineages such as neurogenins, Phox2, and Mash1 (Anderson, 2000; Memberg and Hall, 1995; Thompson and ZiV, 1989; Tiveron et al., 1996). It has therefore been diYcult to carry out an orderly analysis of glial lineage progression in embryonic nerves. For instance, workers have been forced to use markers that appear at a relatively late stage, in particular GFAP or S100, to monitor early events such as the emergence of the glial lineage from neural crest cells. At the present time, however, a framework of diVerentiation markers and other criteria for monitoring embryonic Schwann cell development in rats and mice is emerging (Fig. 13.3). Using the criterion of stage speciWcity, the markers shown in Fig. 13.3 fall into four categories: (1) markers that are found on embryonic PNS glia but do not diVerentiate between developmental stages, exempliWed by the cell adhesion molecule L1; (2) markers that are found on crest cells and Schwann cell precursors but at very low levels on immature Schwann cells, namely N-cadherin and the transcription factor AP2a; (3) markers that diVerentiate Schwann cell precursors and immature Schwann cells from crest cells; three of these, B-FABP, P0 and desert hedgehog (Dhh), are not found on early developing neurons and therefore distinguish between the neuronal and glial lineage at a very early stage; (4)

II. GLIAL CELL DEVELOPMENT

DIFFERENTIATION MARKERS IN EMBRYONIC SCHWANN CELL DEVELOPMENT

markers that are found on immature Schwann cells and can be used to distinguish between them and Schwann cell precursors or crest cells, exempliWed by S100 or GFAP. Markers in groups three and four are of particular use for monitoring the two main lineage transitions. Since the precursor/Schwann cell transition is relatively well characterized (Jessen and Mirsky, 1999; Mirsky and Jessen, 1999, 2001), the following comments are restricted to the third group, namely the markers that are likely to be of use for studying the fundamental issue of glial speciWcation from the neural crest.

B-FABP This protein is a member of the fatty acid binding protein family. It is highly related to the peripheral myelin protein P2 and the cellular retinoic acid binding protein CRABP. B-FABP is found in many areas of the developing CNS, where it is expressed in particular by radial glial cells (Feng et al., 1994; Kurtz et al., 1994). It is also expressed by satellite cells in DRGs in embryonic and adult mice and by glial cells in embryonic nerve trunks but not by adult Schwann cells (Britsch et al., 2001; Kurtz et al., 1994; Woodhoo et al., 2003). B-FABP is not expressed by migrating crest cells nor developing neurons, and the timing of its appearance in DRGs and nerves, at E10–11, is consistent with the emergence of early glia. It has therefore been suggested that it might mark cells that are common precursors of Schwann and satellite cells (Britsch et al., 2001; Kurtz et al., 1994). Unfortunately, in rat, although expression in the CNS seems comparable to mouse at an equivalent developmental age, expression in both satellite cells and developing nerves at E13/14 is extremely low (K. R. Jessen and R. Mirsky, unpublished), so it is unsuitable as a marker of early glial development in this species.

Protein Zero (P0) Although P0 was initially identiWed as the major protein of Schwann cell myelin, it is now clear that the P0 gene is expressed throughout the embryonic Schwann cell lineage in immature Schwann cells and Schwann cell precursors (Cheng and Mudge, 1996; Lee et al., 1997). Although the P0 gene is unambiguously activated in these cells, P0 expression levels are much lower than the axonally induced P0 expression seen in myelinating cells. In the chick, P0 protein is expressed in a subpopulation of migrating crest cells in stage 19 embryos and subsequently in the earliest glia associated with newly formed nerves (Bhattacharyya et al., 1991). In rats P0 is Wrst activated in migrating crest cells located in a region ventral to condensing DRGs at E11 (Lee et al., 2001). P0 remains expressed in the cells that associate with the earliest axons emanating from the ventral spinal cord and is subsequently seen in the large majority of cells in E14/15 rat nerves (i.e., Schwann cell precursors). P0 is also expressed in most or all immature Schwann cells. At no time is P0 in situ hybridization signal seen in neurons identiWed by the early marker ß-III tubulin (detected by TUJ1 antibodies), indicating that P0 gene expression is excluded from the neuronal lineage from the onset (E. Calle and K. R. Jessen, unpublished; Bhattacharyya et al., 1991; Hagedorn et al., 1999; Lee et al., 2001). These Wndings indicate that P0 gene expression can be used as a very early marker of glial speciWcation in neural crest development (Bhattacharyya et al., 1991; Lee et al., 1997). However, two observations complicate the acceptance of P0 as a marker that speciWcally identiWes those cells that have just entered the glial lineage. First, taking embryonic tissues as a whole, the presence of detectable levels of P0 mRNA is clearly not restricted to PNS glia, since in the rat, the gene is also expressed in the notochord and by cells in nonsensory regions of the ear, neither of which contain, or give rise to, PNS glia (Lee et al., 2001). Second, it is clear that commencement of immunohistochemically detectable P0 expression by early crestderived cells does not signify an irreversible commitment to glial fate, since such cells can, by exposure to diVerentiation signals in vitro, be induced to generate other crest derivatives including, in the chick, melanocytes (Hagedorn et al., 1999; Morrison et al., 1999; Paratore et al., 2001; Dupin, et al., 2003; also see the section titled ‘‘The Developmental Potential of Early PNS Glia’’).

II. GLIAL CELL DEVELOPMENT

337

338

13. SCHWANN CELL DEVELOPMENT

As to the Wrst of these, it remains possible that within the crest sublineage P0 expression exclusively indicates glial fate, although in other tissues the gene is transiently activated in restricted cell populations. The second Wnding touches on an important general issue, which is the relationship between speciWcation and developmental potential. These issues are discussed in the section titled ‘‘The Developmental Potential of Early PNS Glia.’’ BrieXy, the view that P0þ Schwann cell precursors have started glial diVerentiation (i.e., are speciWed glia, but retain at the same time the ability to be respeciWed when removed from their normal environment and challenged by alternative signals in vitro) is in good agreement with the available data and current concepts regarding developmental plasticity and transdiVerentiation. It is more diYcult in the mouse than in the rat to detect P0 mRNA in condensing DRGs and early nerves using DIG-linked in situ hybridization (unpublished). Nevertheless P0 mRNA can clearly be seen in Schwann cell precursors of E12 mouse nerves (Peirano et al., 2000).

Dhh Dhh belongs to the Sonic hedgehog family of signaling molecules. It is expressed by Schwann cells and has an important role in controlling the formation of the perineurium and epineurium around peripheral nerves (Parmantier et al., 1999). Dhh mRNA is present in E12 mouse Schwann cell precursors but is not detectable in migrating crest cells or in developing neurons (Bitgood and McMahon, 1995; and unpublished data).

Oct-6 The transcription factor Oct-6 (SCIP/Tst-1) is important for the timely onset of Schwann cell myelination (Bermingham et al., 1996; Monuki et al., 1990; Topilko and Meijer, 2001). Oct-6 protein is found in the nuclei of essentially all Schwann cells in the perinatal period with particularly high levels in promyelin cells (Arroyo et al., 1998; Scherer et al., 1994a). Oct-6 protein is also present at low levels earlier in development, since it can be seen by immunohistochemistry in the nuclei of most or all late mouse Schwann cell precursors at E13 and in rat nerves at E16 at the time of the transition between precursors and Schwann cells (Blanchard et al., 1996). Oct-6 is not expressed by migrating neural crest cells or by developing peripheral neurons (unlike central neurons; Topilko and Meijer, 2001). Therefore, Oct-6 serves in vivo as a marker for late mouse Schwann cell precursors and in the rat for late precursors/early Schwann cells. It should be noted, however, that Oct-6 depends in vivo on axonal signals (Scherer et al., 1994a) and in vitro on the presence of agents, such as forskolin or cAMP analogues, which activate cAMP-dependent pathways. In vitro, therefore, both Schwann cell precursors and Schwann cells are Oct-6 negative unless cAMP pathways are activated.

Growth Associated Protein (GAP)-43, CD-9, Peripheral Myelin Protein 22 (PMP22), and Proteolipid Protein (PLP) While the Wrst three of these markers are absent from migrating crest, they are found on E14 rat Schwann cell precursors and neurons in E14 DRGs. Therefore they do not distinguish between early neurons and early glia, although they will tell such cells apart from the crest cells from which they originate. GAP-43 (neuromodulin) is a phosphoprotein that is expressed at high levels in growing neurites. It is also found in Schwann cells at the neuromuscular junction and in nerve trunks (Curtis et al., 1992; Scherer et al., 1994b; Sensenbrenner et al., 1997; Stewart et al., 1995). GAP-43 expression is up-regulated in Schwann cells in the distal stump of cut nerves and at denervated synaptic junctions (Plantinga et al., 1993; Woolf et al., 1992). In Schwann cell development, GAP-43 expression can be detected as early as the precursor stage at E14 by antibodies and immunohistochemistry (Jessen et al., 1994). Not all of the currently available antibodies detect GAP-43 at this early stage, however (Morris et al., 1999), while some detect GAP-43 only at low levels or in a subpopulation of precursors

II. GLIAL CELL DEVELOPMENT

THE DEVELOPMENTAL POTENTIAL OF EARLY PNS GLIA

(unpublished observations). It is possible that this is due to early glia expressing a diVerent form of the GAP-43 protein than that found in Schwann cells and neurons. CD-9 is a cell surface protein that has been implicated in Schwann cell migration and described in Schwann cells as early as E18 using immunohistochemistry and tissue sections (Anton et al., 1995; Banerjee and Patterson, 1995; Kaprielian et al., 1995). CD-9 can be clearly seen on E14 Schwann cell precursors and on E14 DRG neurons 2 to 3 hours after plating in vitro using immunohistochemistry, although crest cells migrating from neural tubes are initially CD-9 negative under identical culture conditions (unpublished). In vitro, signiWcant amounts of CD-9 are shed from Schwann cell precursors onto the culture substrate. PMP22 is one of the protein components of the myelin sheath (Nave, 2001). In common with most other myelin proteins, PMP22 is present in some non-neural tissues in embryonic development (Baechner et al., 1995). PMP22 is also expressed by some CNS and PNS neurons (Hagedorn et al., 1999; Parmantier et al., 1995, 1997). While PMP22 cannot be detected in migrating crest cells in E10 rat embryos, PMP22 mRNA is present in the early glia of E12 spinal nerves in the rat and PMP22 mRNA and protein are expressed by E14 Schwann cell precursors (Blanchard et al., 1996; Hagedorn et al., 1999). PLP is the major CNS myelin protein but is also found in low amounts in the myelin sheaths of mature peripheral nerves. Null mutations in PLP can lead to peripheral neuropathy (Garbern et al., 1997). It is present in non-neural tissues such as notochord and otic vesicle during development (Spassky et al., 1998; Timsit et al., 1992). It can be detected in mouse neural crest cells as early as E9.5, in condensing sympathetic, trigeminal, and spinal ganglia, and in peripheral nerves (Spassky et al., 1998). It may also be expressed by neuroblasts or neurons at these early stages of development.

THE DEVELOPMENTAL POTENTIAL OF EARLY PNS GLIA Although there is no reason to expect that during normal development Schwann cell precursors generate other cells than Schwann cells, such as neurons or melanocytes, it is clear that cells from early nerves or DRGs of chick, rat, or mouse can be directed to generate other cell types by exposure to growth factors in cell culture. As will be discussed later, this could be due to two principal reasons. It is possible that the early glia found in these nerves are not yet committed to the glial lineage. Their potential to generate other crest derivatives could be revealed in vitro in a signaling environment that diVers from that normally encountered by these cells in vivo. The other possibility is that, in addition to speciWed glia, early nerves retain a small cell population that has not yet begun to diVerentiate as glia but remains similar to migrating neural crest cells in molecular phenotype and developmental potential. The earliest indication that early nerve trunks could in vitro give rise to cells that are not normally generated by that tissue in vivo came from studies on avian systems. In the chick, nerve segments from E4-E6 embryos, presumed to consist largely of Schwann cells with a minimal amount of Wbroblasts, gave rise to pigmented cells, indicative of melanocyte diVerentiation, when cultured in complex media (Sherman et al., 1993). Nerves from older embryos did not generate melanocytes in these experiments. Recently, however, it has been observed that cells in adult nerves of mice can be diverted to melanocyte diVerentiation. This melanogenesis was triggered by nerve injury and was more extensive in mice heterozygous for neuroWbromin, the defective gene in neuroWbromatosis type 1, than in wild-type mice (Rizvi et al., 2002). Experiments on early quail nerve segments show that melanocyte diVerentiation can arise from cells that are P0 positive and that it depends on the activation of intracellular basic Wbroblast growth factor (bFGF) mediated signaling. The Wndings were interpreted as showing bFGF mediated transdiVerentiation or respeciWcation of Schwann cell precursors that would normally give rise to Schwann cells (Sherman et al., 1993; Stocker et al., 1991). Developing Schwann cells from embryonic chick nerves expressing the glial-specific protein SMP, can also be induced to generate melanocytes by exposure to endothelin (Dupin et al., 2003). Melanocyte diVerentiation can also be obtained in cultures of sensory

II. GLIAL CELL DEVELOPMENT

339

340

13. SCHWANN CELL DEVELOPMENT

ganglia (Cowell and Weston, 1970; Nichols and Weston, 1977). Taken together, these studies indicate that speciWed P0 or SMP positive avian Schwann cell precursors are not irreversibly committed to the Schwann cell lineage and can be diverted to generate at least one other crest derivative (i.e., melanocytes). Observations that are to some extent comparable have been made on early rat nerves (Morrison et al., 1999). In these studies, cells dissociated from E14 rat nerves were divided into subpopulations by Xuorescence activated cell sorting (FACS) prior to culturing, using antibodies to the myelin protein P0 and the p75 low aYnity neurotrophin receptor. When cells that initially sorted as P0þ and p75þ or p75low were clonally analyzed after 2 weeks of culture in complex medium containing chick embyro extract and retinoic acid, it was found that many cells had given rise to neurons or cells that express smooth muscle actin as well as GFAP containing Schwann cells. Development of neurons from many of these cells could be stimulated by BMP2 and neurogenesis could be blocked by neuregulin-1, which is in line with the actions of these signals on migrating crest cells (Morrison et al., 1999). Similar methods were used to examine the developmental potential of non-neuronal cells from E14 rat DRG. These cells co-express the p75 low aYnity neurotrophin receptor NTR, P0, and peripheral myelin protein 22 (PMP22) and are therefore likely to be cells that in normal unperturbed development would give rise to satellite glial cells. They could nevertheless be diverted to generate nonglial crest derivatives in vitro (Hagedorn et al., 1999). Although some uncertainty attends the FACS methodology used in the experiments of Morrison et al., 1999 (discussed later), the results of these investigations are consistent with the idea, derived from studies on avian nerves, that speciWed P0þ early glial cells in the PNS retain a degree of developmental plasticity. As mentioned earlier, embryonic nerves might, in addition to Schwann cell precursors, also contain a distinct population of multipotent crest-like cells that could be diverted to nonglial lineages. It has been suggested that absence of P0 expression distinguishes these cells from Schwann cell precursors and that P0 negative cells of this type account for
15% of the total number of cells dissociated from E14 rat nerves (Morrison et al., 1999, 2000; White et al., 2001). The evidence for this interesting possibility is as yet incomplete. It comes from studies in which cells from E14 rat nerves were fractionated using FACS sorting (noted earlier) (Kubu et al., 2002; Morrison et al., 1999, 2000; White et al., 2001). Cells that sorted as P0, p75þ, which is a molecular phenotype expected of migrating neural crest cells, were found to be able to self-renew and generate neurons, glia or ‘‘myoWbroblasts’’ in response to BMPs, neuregulin-1, and transforming growth factor (TGF) ß. These responses are qualitatively similar to those of migrating crest cells. Subsequent studies showed, however, that the E14 nerve-derived cells were about 10fold less sensitive to the neurogenic signal BMP2 than crest cells derived directly from neural tube cultures. The nerve-derived cells were also shown to be strongly biased toward generating glia, in comparison to tube-derived crest cells, using an in vivo assay (White et al., 2001). Similarly, in vitro the P0, p75þ E14 nerve-derived cells are more sensitive than migrating crest cells to the gliogenic actions of delta-Notch signaling, a change that may be caused by an elevated expression of Notch and reduced expression of the putative Notch inhibitor NUMB in these cells (Kubu et al., 2002). Thus, cells that FACS sort from E14 nerves as P0, p75þ retain more than one developmental option, at least in vitro, but are strongly biased toward gliogenesis, in comparison with migrating crest cells. These cells may represent a distinct population of neural crest cell-like cells resident in embryonic nerves as suggested by Morrison et al. (1999). An alternative interpretation of these results is, however, possible, particularly since the methodology used for obtaining them raises some concern. These cells were FACS sorted as P0, p75þ using the P07 anti-P0 antibody described by Archelos et al. (1993). Two issues raise questions about the suitability of the P07 antibody for sorting cells from embryonic nerves. First, this antibody does not show detectable binding to living, unWxed Schwann cells or their precursors, as judged by immunohistochemistry; although it is generated against an extracellular portion of the P0 protein, signiWcant binding of this antibody to cell surfaces appears to require previous Wxation. Second, another population of cells that sorts from the same tissue as P0þ, p75 forms neither Schwann cells in the presence of neuregulin-1, as would be expected of cells

II. GLIAL CELL DEVELOPMENT

THE ROLE OF NEUREGULIN-1 IN SCHWANN CELL DEVELOPMENT

expressing P0, nor neurons in the presence of BMP2/4 (Morrison et al., 1999). The absence of neuronal and glial diVerentiation in the presence of BMPs and neuregulin, the lack of p75 expression, and population size (
15% of the total cell suspension derived from freshly dissected E14 nerves) are all consistent with the notion that these are non-neural, presumably connective tissue, cells that adhere to dissected E14 nerves, as indicated by Morrison et al. (1999). If this is the case, however, the expression of P0 by these cells would be puzzling since there are no other reports of P0 expression in such cells. Because of these concerns, the identity of the P0, p75þ FACS sorted cells from E14 nerves as a distinct population of crest-like cells remains uncertain. The major alternative possibility is that the FACS sorting is not meaningful in these experiments and that the cells in question represent Schwann cell precursors. The Wnding that these cells can be diverted to more than one lineage in vitro does not clarify the issue since such plasticity may well be a property shared by migrating crest cells and Schwann cell precursors as discussed earlier (see the section titled ‘‘DiVerentiation Markers in Embryonic Schwann Cell Development’’). The possibility that the cells that FACS sorted in these experiments as P0, p75þ represent Schwann cell precursors would be consistent with their strong gliogenic bias. It would also be consistent with the action of Notch activation in these cells, which is to induce an irreversible commitment to the glial lineage (Morrison et al., 1999). Irreversible glial commitment is unlikely to appear as migrating crest cells generate Schwann cell precursors, since these cells show lineage plasticity as discussed earlier. Schwann cells on the other hand show much stronger lineage restriction (Ciment, 1990; Morrison et al., 1999). It is therefore more likely that cells from E14 nerves that become irreversibly restricted to glial diVerentiation by Notch activation represent in fact Schwann cell precursors. This would imply that one of the functions of Notch activation is to promote the generation of Schwann cells from Schwann cell precursors. This idea remains to be tested. In conclusion, there are at least two plausible explanations for the phenotypic plasticity observed in cells from early embryonic nerves and they are not mutually exclusive. First, although Schwann cell precursors are speciWed glial cells they may not yet be determined (i.e., irreversibly committed) to a glial fate. Second, early nerves might well contain a minority population of cells with a phenotype and developmental potential that resembles migrating crest cells. Previously established views on commitment and phenotypic plasticity have been challenged recently by numerous studies on stem cell development, respeciWcation, transdiVerentiation of one cell type into another and other metaplasias (French et al., 2002; Tosh and Slack, 2002). The burden of this work is that the diVerentiated state is less Wxed than previously envisaged. The notion that speciWed precursor cells, such as the Schwann cell precursor, can be reprogrammed using signaling factors in vitro is therefore no longer surprising, but stands in agreement with a large number of experiments. A close parallel is seen, for instance, in the oligodendrocyte lineage (Kondo and RaV, 2000a). This work indicates that extracellular signals can reverse glial speciWcation and convert cells that are unambiguously speciWed as oligodendrocyte precursors into multipotential stem cells. One of the conclusions from this and related work is that during development there may be a signiWcant time lag between speciWcation of a cell to a particular lineage and determination, a state that implies irreversible commitment. It is likely that in Schwann cell development this is reXected in the Wndings that broad developmental potential is more easily demonstrated using rat nerves at E14/15 that contain Schwann cell precursors, than using E17 nerves where most of the cells are Schwann cells (Morrison et al., 1999). Comparable observations have been made using chick DRGs and nerves (Ciment, 1990).

THE ROLE OF NEUREGULIN-1 IN SCHWANN CELL DEVELOPMENT Neuregulin-1 has been implicated as a regulator of various aspects of Schwann cell development and biology at every stage of the lineage from the neural crest to adult cells (Adlkofer and Lai, 2000; Burden and Yarden, 1997; Lemke, 2001). No other factor has

II. GLIAL CELL DEVELOPMENT

341

342

13. SCHWANN CELL DEVELOPMENT

been proposed to be so comprehensively involved in Schwann cell development. Neuregulin-1 has a long history of association with Schwann cells, since one of the Wrst agents found to drive the proliferation of cultured rat Schwann cells was later identiWed as the type II isoform of neuregulin-1 (Gassman and Lemke, 1997; Lemke, 1996; Marchionni et al., 1993). Neurogulin-1 is now considered to be the major axonally-derived Schwann cell mitogen. Three additional members of the neuregulin gene family (neuregulins 2-4) have now been described (Carraway et al., 1997; Harari et al., 1999; Zhang et al., 1997). Neuregulins 2 and 3 are expressed in the nervous system, but it is not known whether they have a role in nerve development (Adlkofer and Lai, 2000). In mice in which neuregulin-1 or the neuregulin receptors ErbB2 or ErbB3 have been genetically inactivated, Schwann cell precursors are essentially absent from embryonic nerve trunks, although, intriguingly, satellite glial cell development within DRGs appears to be normal, as is the initial development of DRG neurons (Garratt et al., 2000a; Meyer and Birchmeier, 1995; Morris et al., 1999; Riethmacher et al., 1997; Woldeyesus et al., 1999). Sympathetic ganglia, however, do not develop properly in these animals (Britsch et al., 1998). What do these striking phenotypes reveal about the role of neuregulin signaling in embryonic development of the PNS and, in particular, the glial lineage? Satellite glia within ganglia, like Schwann cells along nerve trunks, are considered to develop from the neural crest. Therefore the presence of an apparently normal population of DRG satellite cells in neuregulin mutants indicates that neuregulin-1 signaling is not needed for glial diVerentiation from the neural crest in vivo. This is in line with the Wnding that GFAP positive glia develop readily from neural crest cells in culture with or without neuregulin (Shah et al., 1994). Although neuregulin strongly suppresses the generation of neurons from crest cells in culture (Shah et al., 1994), the neuregulin mutants fail to provide any obvious evidence, such as overproduction of neurons, in support of the idea that neuregulin-1 acts as a brake on neurogenesis in vivo. This eVect of neuregulin may be masked in vivo by compensatory mechanisms that are not present in the culture system. A major reason for the striking reduction or absence of Schwann cell precursors from nerve trunks in neuregulin mutants is likely to relate to the function of neuregulin-1 as a survival factor for Schwann cell precursors and early Schwann cells (Dong et al., 1995; Jessen and Mirsky, 1999; Leimeroth et al., 2002). Unlike Schwann cells, Schwann cell precursors cannot support their own survival by autocrine mechanisms (see the sections titled ‘‘The Regulation of the Precursor-Schwann Cell Transition’’ and ‘‘Schwann Cell Survival Signals’’) (Jessen et al., 1994; Meier et al., 1999). They lie in close apposition to axons in embryonic nerves and die if they are removed from axons and plated in vitro without neurons (Dong et al., 1995; Jessen et al., 1994). This has been conWrmed in vivo in studies on chick embryos, which showed that the survival of early glia in embryonic nerves is strongly dependent on the presence of axons (Ciutat et al., 1996; Winseck et al., 2002). All of this indicates that the survival of Schwann cell precursors in developing nerves depends on survival signals from axons. Cell culture studies show that DRG neurons are a potent source of such signals and that they can be either axon associated or secreted. In both cases the signals have been identiWed as neuregulin-1 (Dong et al., 1995). Neuregulin-1 is expressed in vivo in embryonic DRG and motor neurons and accumulates along axonal tracts (Falls et al., 1993; Loeb et al., 1999; Marchionni et al., 1993; Orr-Urtreger et al., 1993). It is therefore present at the right time and place to control the survival of precursors in embryonic spinal nerves. This is further supported by observations on mutants lacking isoform III of neuregulin-1 (Wolpowitz et al., 2000). In these mice, early Schwann cell precursors (at E11) populate nerves, although the number of cells in the nerves is severely depleted by E14, a time when normal nerves contain late precursors that are converting rapidly to Schwann cells. Neuregulin has also been found to support the survival of early glia in embryonic chick nerves (Winseck et al., 2002). All of these Wndings taken together strongly support the notion that a major function of neuregulin-1 in embryonic nerves is to support the survival of Schwann cell precursors (Dong et al., 1995; Jessen and Mirsky, 1999). It is also possible that reduced cell migration contributes to the paucity of Schwann cell precursors in developing nerves of mice deWcient in neuregulin signaling. This is supported

II. GLIAL CELL DEVELOPMENT

343 indirectly by the Wnding that neuregulin-1 promotes the migration of post-natal Schwann cells in vitro and stimulates glial migration from explants of DRGs obtained from newborn mice (Meintanis et al., 2001; Morris et al., 1999). Furthermore, impaired migration of neural crest cells along the ventro-medial pathway is thought to be a major reason why the development of sympathetic ganglia fails in mice lacking ErbB3 receptors (Britsch et al., 1998). Lastly, in cell culture, the migration of glia from E12 DRG of neuregulin mutant mice is impaired (Morris et al., 1999). On the other hand, neuregulin-1 has no eVect on migration of glia from E12 DRG in vitro and neural crest cells migrate normally out from the neural tube and from DRGs in appropriate locations in mice lacking neuregulin-1 or the ErbB3 receptor. While the evidence relating to migration is therefore ambiguous, it is clearly possible that reduced migration contributes to the paucity of precursors in nerves of mice deWcient in neuregulin signaling. Of a large number of neuregulin-1 isoforms, it appears that neuregulin-1, isoform I (NDF, ARIA), isoform II (GGF), and, in particular, isoform III (SMDF, CRD-NRG-1) are of most relevance for PNS and Schwann cell development (Burden and Yarden, 1997; Fischbach and Rosen, 1997; Meyer et al., 1997). This has been concluded from studies on neuregulin mutant mice. Among these mutants, peripheral nerve trunks are most severely depleted by Schwann cell precursors in animals with inactivation of the neuregulin-1 gene and therefore all three isoforms, or of the neuregulin receptors ErbB2 or EbrB3. Mice with selective inactivation of the neuregulin-1 isoform III initially show a milder phenotype, although at later stages Schwann cell precursor numbers are severely depleted and Schwann cells are essentially absent (discussed earlier; also see Wolpowitz et al., 2000). In contrast, mice lacking neuregulin-1 isoforms I and II show normal glial development in peripheral nerves (Meyer et al., 1997). Therefore, of the three main neuregulin-1 isoforms, isoform III is of greatest importance for Schwann cell development. While isoforms I and II bind heparin and may therefore associate with cell surfaces (Adlkofer and Lai, 2000), isoform III, unlike the other isoforms, is mostly expressed as a transmembrane protein (Wang et al., 2001a). This may be critical for function since membrane bound, but not soluble, isoform III induces S100 and Oct-6 expression in P0þ non-neuronal cells from embryonic DRGs (Leimeroth et al., 2002).

SIGNALING PATHWAYS ACTIVATED BY NEUREGULIN-1 IN SCHWANN CELLS The intracellular signaling pathways that transduce neuregulin signals have been extensively studied in other cell types but it is not clear how many of the interacting proteins involved in the neuregulin signaling pathway in these diverse cell types are used by Schwann cells (for reviews see, e.g., Buonanno and Fischbach, 2001; Garratt et al., 2000a; Le et al., 2002; Lemke, 2001; Niemann et al., 2000). In Schwann cells, neuregulin1 induces phosphorylation and dimerization of the neuregulin receptors ErbB2 and ErbB3 (Grinspan et al., 1996; Rahmatullah et al., 1998; Riethmacher et al., 1997; Rosenbaum et al., 1997; Sudhalter et al., 1996; Syroid et al., 1996; Vartanian et al., 1997; Woldeyesus et al., 1999). ErbB2 and ErbB3 are associated with the transmembrane glycoprotein CD44 which enhances neuregulin-induced ErbB2 phosphorylation and ErbB2-ErbB3 dimerization. Furthermore, block of CD44 expression in Schwann cells cultured in low doses of neuregulin-1 induces apoptosis that can be prevented by high concentrations of applied neuregulin-1, suggesting that it facilitates neuregulin signaling (Sherman et al., 2000). In Schwann cells, as in other cell types, neuregulin-1 induces both activation of PI3 kinase and its target Akt, and the mitogen-activated protein kinase (MAPK/ERK) pathway. The PI3 kinase pathway appears to be more important in the proliferative response, since inhibition of this pathway inhibits DNA synthesis in Schwann cells in response to applied neuregulin1 and contact with neurites, whereas inhibition of the MAPK/ERK kinase pathway has no eVect (Kim et al., 1997; Maurel and Salzer, 2000). In breast cancer cells, neuregulin-1 also activates the p38 MAPK pathway since inhibition of this pathway prevents neuregulin-1

II. GLIAL CELL DEVELOPMENT

344

13. SCHWANN CELL DEVELOPMENT

induced proliferation (Neve et al., 2002), and a similar eVect is seen in Schwann cells (D. B. Parkinson, K. R. Jessen and R. Mirsky, unpublished). In Schwann cells, neuregulin-1 activates the Jun-N-terminal kinase (JNK)/c-Jun pathway, and JNK/c-Jun activity is required for neuregulin-1 mediated Schwann cell proliferation. This pathway is also required for Schwann cell death induced by TGFb or serum deprivation (Parkinson et al., 2001), while it is suppressed by the myelin-related transcription factor Krox-20 (discussed later). The JNK/c-Jun pathway is therefore likely to be important in the regulation of Schwann cell development (Parkinson et al., 2003a; D. Parkinson, A. Bhaskaran, R. Mirsky, K. R. Jessen unpublished). Neuregulin-induced Schwann cell survival also requires activation of PI3 kinase and Akt and subsequent phosphorylation and inactivation of the pro-apoptotic protein BAD (Dong et al., 1999; Li et al., 2001; Maurel and Salzer, 2000; Meier et al., 1999). In some, but not all, studies, inhibition of the MAPK/ERK pathway also causes signiWcant cell death (Maurel and Salzer, 2000; Meier et al., 1999; Parkinson et al., 2002b). In Schwann cell precursors inhibition of either the PI3 kinase or the MAPK/ERK pathway completely inhibits survival in neuregulin-1 (Dong et al., 1999). Neuregulin-1 stimulates a delayed rise in intracellular cAMP levels and protein kinase A (PKA) activation that is required for neuregulin-induced Schwann cell proliferation (Kim et al., 1997). This may account for the fact that neuregulin-1, unlike most other growth factors, promotes proliferation in the absence of other agents that stimulate cAMP pathways, although the proliferative response is potentiated by both forskolin and insulin-like growth factor (IGF) (Conlon et al., 2001; Stewart et al., 1996). In Schwann cells, neuregulin-1 also activates PAK65 (a component of the stress-activated signaling pathways), pp70S6 kinase (a downstream eVector of PI3 kinase) (discussed later), p95RSK2 (a CREB kinase activated by MAP kinase), and stimulates CREB by phosphorylation on serine-133. In addition, it stimulates expression of cyclin D and phosphorylation of retinoblastoma protein, both enhanced by addition of forskolin (Rahmatullah et al., 1998; Tabernero et al., 1998). There is also evidence that protein kinase C (PKC), another potential eVector of PI3 kinase, is involved in neuregulin-1 signaling in Schwann cells (Saunders and De Vries, 1988; Yoshimura et al., 1993). Both neuregulin and autocrine-mediated survival of Schwann cells (see the section titled ‘‘Schwann Cell Survival Signals’’) are dependent on the activity of Ets transcription factors, several of which are expressed by Schwann cells (Parkinson et al., 2002b). Interestingly, stimulation of synapse speciWc acetylcholine receptor synthesis by neuregulin-1 is also dependent on an Ets transcription factor binding site (Fromm and Burden, 2001).

THE REGULATION OF THE PRECURSOR-SCHWANN CELL TRANSITION The generation of Schwann cells from Schwann cell precursors involves a coordinated change in a number of disparate and apparently unrelated phenotypic features that relate to antigenic expression (Fig. 13.3), survival regulation, response to mitogens, motility, and cell-cell interactions (see the section titled ‘‘Schwann Cell Precursors’’). This diVerentiation program unfolds remarkably faithfully, both with respect to the range of phenotypic changes and timing, in precursors isolated from neurons and maintained under simple conditions in vitro. Thus if E14 precursors are plated in neuron-free cultures in chemically deWned medium containing neuregulin-1 for 4 days (E14 þ 4 ¼ E18), then dissociated from the coverslips and reexamined for the presence of autocrine survival loops (see the section titled ‘‘Schwann Cell Survival Signals’’), antigenic phenotype and mitogenic responses, over 80% of the cells now show the phenotype of Schwann cells rather than that of precursors (Dong et al., 1995). Therefore neuregulin-1 alone is suYcient not only to support precursor survival, but also lineage progression. Not all precursors convert to Schwann cells, however, under these simpliWed conditions and it is likely that multiple

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL SURVIVAL SIGNALS

signals act together to ensure the appropriate rate of Schwann cell generation in vivo (Brennan et al., 2000; Leimeroth et al., 2002). One of these signals is likely to be endothelin acting through the endothelin B receptor (Brennan et al., 2000). Endothelins and endothelin receptors are present in embryonic nerves, and endothelins 1 to 3 all support the survival of Schwann cell precursors in vitro. Unlike neuregulin, endothelin does not induce DNA synthesis in precursors. Lineage progression is also aVected diVerently by these two factors. Schwann cells are generated very slowly in precursor cultures exposed to endothelin alone and, importantly, Schwann cell generation in the combined presence of endothelin and neuregulin-1 is intermediate between the rates seen in endothelin alone and neuregulin-1 alone. This suggests that neuregulin actively promotes the transition of precursors to Schwann cells, since Schwann cell generation is accelerated by the addition of neuregulin-1 to endothelin. Conversely, since addition of endothelin to neuregulin-1 delays the appearance of Schwann cells compared to the rate seen in neuregulin-1 alone, these observations also indicate that endothelins act as a brake on Schwann cell generation. This has been conWrmed in vivo in experiments on the spotting lethal rat. In these animals, the endothelin B receptors are nonfunctional and Schwann cell generation takes place ahead of schedule. This is the expected result if, in normal nerves, endothelin B receptor activation exerts a tonic negative eVect on the rate of lineage progression (Brennan et al., 2000). FGF2 might be a third regulatory signal in this system, since inclusion of FGF2 accelerates the Schwann cell generation in vitro in comparison with that seen in neuregulin-1 alone (Dong et al., 1999). This remains to be conWrmed in vivo. It is clear that the conversion of precursors to Schwann cells involves orderly changes in the expression of a number of genes, which, in turn, implies the existence of a coordinating gene control mechanism. To date, however, only one transcription factor, AP2a, has been implicated in this process (Stewart et al., 2001). AP2a expression is sharply down-regulated at the precursor-Schwann cell transition in vivo, both in rats and mice, and enforced expression of AP2a in precursors in vitro delays their conversion to Schwann cells (Stewart et al., 2001). This suggests that the rate of AP2a down-regulation is one of the factors that determines the rate of Schwann cell generation in vivo. There is evidence that the time course of oligodendrocyte development may be inXuenced by the Id type of helix-loop-helix (HLH) genes (Kondo and RaV, 2000b; Wang et al., 2001b). Schwann cell precursors and Schwann cells co-express Id 1 to 4. The expression of these factors is not obviously regulated during embryonic nerve development and there is at present no evidence that Id proteins are involved in controlling the speed of the precursor-Schwann cell transition.

SCHWANN CELL SURVIVAL SIGNALS It is likely that two sets of signals play a major role in promoting the survival of developing Schwann cells. These are axon-associated neuregulin-1 (Dong et al., 1995; Grinspan et al., 1996; Syroid et al., 1996; Trachtenberg and Thompson, 1996) and autocrine Schwann cell signals (Cheng et al., 1998; Meier et al., 1999). Transection of neonatal nerves results in increased Schwann cell death within the body of the nerve, although most of the Schwann cells survive, and death of all teloglia at the neuromuscular junction. The cells can be rescued by application of exogenous neuregulin-1 and it is probable that the death is caused by loss of contact with axon-associated neuregulin, as axons degenerate after transection (Grinspan et al., 1996; Syroid et al., 1996; Trachtenburg and Thompson, 1996). After the Wrst post-natal week, there is no comparable Schwann cell death immediately following nerve transection in rats and Schwann cells in the distal stumps of adult animals survive for several months after nerve transection, although their number gradually declines and they become less responsive to extrinsic signals (Grinspan et al., 1996; Li et al., 1998; Sulaiman and Gordon, 2002; Trachtenberg and Thompson, 1996). SigniWcant Schwann cell death is, however, seen in regenerating nerves at three weeks post crush (Ferri and Bisby, 1999). The ability of Schwann cells to survive in the absence of axons is crucial

II. GLIAL CELL DEVELOPMENT

345

346

13. SCHWANN CELL DEVELOPMENT

FIGURE 13.4 Survival regulation in the Schwann cell lineage. Both axon-derived and autocrine signals regulate survival of cells in the Schwann cell lineage. Schwann cell precursors rely exclusively on the axonal signal, neuregulin-1, for survival, but as development proceeds and precursors generate Schwann cells, there is a shift to the establishment of autocrine loops, the main components of which include IGF-2, PDGF-BB and NT3, LIF, and an LPA-like activity.

for nerve regeneration after nerve injury, since Schwann cells provide both trophic factors and adhesive substrates that promote axonal growth. A major reason why Schwann cells survive under these conditions, in which Schwann cell precursors would die, is likely to be the presence of autocrine circuits, which enable Schwann cells to support their own survival (Cheng et al., 1998; Meier et al., 1999) (Fig. 13.4). Autocrine survival circuits are not present in Schwann cell precursors and these cells appear to be wholly dependent on axonal neuregulin signaling for survival (Dong et al., 1999; Meier et al., 1999; Winseck et al., 2002). Because the autocrine Schwann cell survival signal is not mitogenic and does not support precursor survival, it is unlikely to be neuregulin-1 (Cheng et al., 1998; Meier et al., 1999), although low levels of neuregulin-1 can be made by Schwann cells (Carroll et al., 1997; Rosenbaum et al., 1997). IGF-2, neurotrophin-3 (NT3) and platelet-derived growth factor-BB (PDGF-BB) have been identiWed as major components of the autocrine loop. Schwann cells have receptors for these factors and the factors support survival when applied together in very low concentrations. Furthermore, antibodies to these proteins block the Schwann cell survival activity in Schwann cell conditioned medium. IGF has also been shown to act as a potent Schwann cell survival factor using an assay in which Schwann cell death is induced by abrupt serum deprivation. In these experiments, IGF acts via PI3 kinase and Akt to inhibit activation of c-Jun terminal kinases and prevent caspase mediated apoptosis (Campana et al., 1999; Cheng et al., 2000a; Delaney et al., 1999; Syroid et al., 1999). Other potential autocrine Schwann cell survival factors include leukaemia inhibitory factor (LIF) and lysophosphatidic acid (LPA). LIF is secreted by denervated Schwann cells and can also promote Schwann cell survival in the presence of other growth factors (Dowsing et al., 1999). LPA promotes survival in an assay involving abrupt serum deprivation, and Schwann cells secrete an LPA-like activity into the medium. LPA signals through the Gi-protein coupled lpA1 receptor, which in turn activates the PI3 kinase pathway and Akt (Weiner and Chun, 1999; Weiner et al., 2001). In the presence of autocrine signals, longer-term survival is promoted by culture on a laminin substrate, although laminin alone does not support survival (Meier et al., 1999). As mentioned earlier, Schwann cell survival mediated by neuregulin-1 or by Schwann cell conditioned medium requires the activity of Ets transcription factors, whereas LPA induced survival does not (Parkinson et al., 2002b). In summary, Schwann cell precursors do not have autocrine survival circuits and their survival is acutely dependent on axonal neuregulin, while Schwann cells can rescue themselves by secretion of a number of survival factors that are likely to include IGF-2, PDGFBB, NT3, LIF, and LPA. Importantly, Schwann cells in older nerves can therefore survive in the absence of axons, at least in the medium term. Axonal neuregulin-1 is likely to provide additional survival support for Schwann cells in the perinatal period. In addition to positive survival signals, factors that actively promote apoptosis may also play a role in Schwann cell death after injury or infection. There is evidence for the action

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL SURVIVAL SIGNALS

of two such factors, nerve growth factor (NGF) and TGFß, in Schwann cell development (Parkinson et al., 2001; Soilu-Hanninen et al., 1999). NGF, acting via the p75 neurotrophin receptor (p75 NTR), promotes cell death in several systems including Schwann cells, but the eVect is complex since under some conditions NGF promotes Schwann cell survival, underlying the importance of the total cellular and signaling context in the balance between survival and death (Khursigara et al., 2001). In neonatal mice lacking the p75 NTR, Schwann cell death following nerve cut is less than that seen in normal mice. Adult mice lacking p75 NTR also show reduced death three weeks after nerve crush, at a time when axons are regenerating into the distal stump, and there is a requirement to match the number of Schwann cells and axons (Ferri and Bisby, 1999; Syroid et al., 2000). Cultured expanded neonatal Schwann cells from these mice survive better than normal Schwann cells when deprived of serum and growth factors (Frade and Barde, 1999; Soilu-Hanninen et al., 1999; Syroid et al., 2000). Furthermore, antibodies to the p75 NTR block apoptosis in Schwann cells cultured from transected adult rat nerve distal stumps (Hirata et al., 2001). The complexity of the Schwann cell response to p75 NTR activation is underlined by studies on the receptor-interacting protein (RIP)-2 that interacts speciWcally with p75 NTR. In the presence of RIP-2, Schwann cell death in response to NGF is prevented through activation of NFkB. RIP-2, which is a protein kinase and contains a caspase recruitment (CARD) domain, is expressed by freshly isolated Schwann cells, but is lost from Schwann cells on prolonged culture. On loss of RIP-2, Schwann cells become sensitive to NGF induced cell death, and transfection of constitutively active RIP2 confers protection against death. Conversely, expression of dominant-negative RIP-2 in freshly cultured RIP-2 expressing Schwann cells makes them sensitive to NGF-induced cell death (Khursigara et al., 2001). Another factor that associates with p75 NTR is TNF receptor-associated factor (TRAF) 6, which activates the JNK pathway in Schwann cells, and may be responsible for the NGF-induced death seen in the absence of RIP-2 (Khursigara et al., 1999, 2001). In immortalized cell lines expression of other p75 NTR associated proteins has been correlated with cell death, but their importance in Schwann cell death is not clear. These include the zinc Wnger proteins neurotrophin receptor interacting factor (NRIF) -1 and -2, p75 neurotrophin receptor-associated cell death executor (NADE), and neurotrophin receptor-interacting MAGE (NRAGE), which also mediates cell cycle arrest in sympathetic neuroblasts (Benzel et al., 2001; Casademunt et al., 1999; Kendall et al., 2002; Mukai et al., 2000, 2002; Salehi et al., 2000). In addition, two other proteins have been shown to bind to p75 NTR, but these factors do not appear to be involved in cell death. These are the zinc Wnger protein Schwann cell factor-1 (SC-1) which promotes growth arrest in Schwann cells, and RhoA GTPase, which promotes neurite elongation in PC12 cells (Chittka and Chao, 1999). Like NGF, TGFßs can induce apoptotic cell death in neonatal Schwann cells both in vitro and in vivo (Parkinson et al., 2001). Application of TGFß1 kills freshly isolated neonatal Schwann cells in cultured serum-free media by apoptosis and increases cell death when injected into transected, but not normal, neonatal nerves. In vitro, the apoptotic eVect is completely blocked by the combined presence of neuregulin-1 and the autocrine signal combination of IGF-2, PDGF-BB, and NT3. TGFß1 signals via cJun-N-terminal kinase and phosphorylation of c-Jun, since adenoviral expression of dominant-negative c-Jun inhibits TGFß induced apoptosis, while retroviral expression of constitutively active v-Jun promotes cell death. Myelinating Schwann cells from postnatal day 4 nerves are resistant to TGFß induced apoptosis, and this is related to inability of TGFß to phosphorylate c-Jun in myelinating cells (Parkinson et al., 2002a, 2003a). In vitro, enforced expression of the myelin-related transcription factor Krox-20 is suYcient to render Schwann cells insensitive to TGFß-induced death (Parkinson et al., 2002a, 2002b; see the discussion presented later). In the presence of serum, a combination of TGFß and tumor necrosis factor (TNF) a, but neither alone, will induce apoptosis in Schwann cells. This combination is also more potent than TGFß alone in killing freshly isolated cells in serum-free conditions (SkoV et al., 1998; D. Parkinson, R. Mirsky and K. R. Jessen, unpublished). TNFa can also induce death in immortalized Schwann cells in combination with INFg. This occurs via production of nitric oxide and ceramide (Nagano et al., 2001).

II. GLIAL CELL DEVELOPMENT

347

348

13. SCHWANN CELL DEVELOPMENT

In conclusion, Schwann cell survival appears to be regulated by a balance between positive survival signals, principally axon associated neuregulin-1 and a number of autocrine signals, and negative survival signals that include NGF and TGFß.

SCHWANN CELL PROLIFERATION Neuronally derived neuregulin-1 (see the sections titled ‘‘The Role of Neuregulin1 in Schwann Cell Development’’ and ‘‘Signaling Pathways Activated by Neuregulin-1 Schwann Cells’’) is generally considered to be the principal mitogen in Schwann cell development. Schwann cells, however, proliferate in response to a variety of other growth factors present in neurons. Two in vitro models have been used extensively to investigate Schwann cell mitogens and the intracellular pathways by which they work. In the Wrst model, co-cultures of DRG neurons and Schwann cells have been used to study axon-induced proliferation, establishing neuregulin-1 as the primary axonal mitogen for Schwann cells (Bunge et al., 1986; Levi et al., 1995; Morrissey et al., 1995). In this system, axon-induced proliferation and myelination occur sequentially in deWned media with or without serum, and the eVects of antibodies or added growth factors, genetically engineered constructs, or intracellular pathway inhibitors on both proliferation and myelination (see the section titled ‘‘Myelination’’) can also be studied. In the other model, DNA synthesis and proliferative response are measured in puriWed cultures of rat, mouse, or human Schwann cells grown in deWned media with identiWed growth factors or serum, and usually including agents that elevate or mimic cAMP, since in most instances the mitogenic response to growth factors is minimal in the absence of cAMP elevation (for review, see Mirsky and Jessen, 2001). The variety of diVerent protocols that has been used to study proliferation in puriWed Schwann cell cultures makes it diYcult to strictly compare results obtained in diVerent laboratories, particularly when activation of intracellular signaling pathways is investigated. It is important to be aware that cells prepared by diVerent methods, including freshly isolated immunopanned cells, serum puriWed cells, and cells expanded in growth factors and forskolin for short or long periods, may show diVerent responses to applied reagents. Schwann cells in co-cultures with neurons may also show responses that diVer from those seen in puriWed Schwann cell cultures. Using serum puriWed cultures it has been found that, in the presence of cAMP elevation and type 1 IGF receptor stimulation, several growth factors, including FGF-1 and -2, PDGF-BB, TGFßs, and Reg-2, act as Schwann cell mitogens in the presence or absence of serum. IGFs potentiate the eVects of most Schwann cell mitogens including neuregulin-1, acting in Schwann cells and other cells as progression factors (Cheng and Feldman 1997; Cheng et al., 1999; Conlon et al., 2001; Stewart et al., 1996; Stiles et al., 1980). While neuregulin-1 and IGFs both promote cell cycle progression in Schwann cells, IGFs also promote growth in cell size (Conlon et al., 2001; Conlon and RaV, 2003). With the exception of neuregulin-1 and hepatocyte growth factor, none of the potential Schwann cell mitogens stimulate signiWcant Schwann cell proliferation unless cAMP dependent pathways are simultaneously activated by agents such as forskolin or cAMP analogues (Cheng et al., 1999; Krasnoselsky et al., 1994; Livesey et al., 1997; Stewart et al., 1996; for review, see Eccleston, 1992; Mirsky and Jessen, 2001). At present there is no overarching picture of all the pathways involved in the cAMPdependent stimulation of proliferation, but three disparate sets of experiments point to some of the mechanisms that may be involved. First, experiments on thyroid follicular cells and Swiss 3T3 Wbroblasts have implicated the p70 ribosomal S6 kinase (p70s6k). In these cells and in Schwann cells, cAMP dependent proliferation involves activation of p70s6k and inhibition of this kinase with rapamycin inhibits it (Cass and Meinkoth, 1998). Further investigation of this mechanism in thyroid follicular cells reveals that two separate cAMP activated pathways converge on p70s6k to promote cAMP-dependent proliferation. One pathway involves PKA-dependent activation of p70s6k, while the other involves PI3 kinasedependent, PKA-independent, activation of p70s6k (Cass et al., 1999; Cass and Meinkoth,

II. GLIAL CELL DEVELOPMENT

MYELINATION

2000). Second, it has been suggested that a major function of cAMP in Schwann cell proliferation is to up-regulate synthesis of receptors for Schwann cell mitogens such as PDGF (Weinmaster and Lemke, 1990). Third, recent studies, also using PDGF as a mitogen, indicate that the primary action of cAMP elevation is on progression through G1 via sustained elevation of the G1 phase-speciWc protein cyclin D1, which is induced only transiently by PDGF alone. The G1 phase requirement for cAMP can be alleviated by ectopic expression of cyclin D1, while the tumor suppressor gene Nf1 antagonizes accumulation of cAMP and expression of cyclin D1 (Kim et al., 2000, 2001). These studies conWrm the receptor up-regulation seen by Weinmaster and Lemke (1990) but Wnd that it occurs too late to explain the early stimulatory eVect of cAMP on PDGF induced proliferation. Interestingly, retroviral inhibition of cAMP-dependent PKA inhibits mitosis in puriWed Schwann cells exposed to cAMP elevating agents and mitogens, but does not do so in Schwann cell-neuron co-cultures where proliferation is driven by contact with neurites. Neurite-induced proliferation is nevertheless inhibited by PKA inhibitors such as H89 and KT5780, which are less speciWc than the retrovirally mediated vectors. This suggests the involvement of pathways other than PKA in neuron-induced proliferation in development (Howe and McCarthy, 2000, Kim et al., 1997; and see the discussion that follows). Although TGFß induces Schwann cell apoptosis when applied to immunopuriWed cells from nerves of newborn rats (see the section titled ‘‘Schwann Cell Survival Signals’’), it acts to promote Schwann cell DNA synthesis in a variety of other contexts, including with serum or in the presence of cAMP elevating agents or both (Eccleston et al., 1989; Ridley et al., 1989). Yet another response is seen when TGFß is applied to Schwann cells in cocultures with DRG neurons. In this case, TGFß inhibits the Schwann cell proliferation induced by contact with neurites (Einheber et al., 1995; Guenard et al., 1995b). Insofar as neuregulin-1 is believed to be the major axon-associated mitogen, it might be predicted that TGFß would also suppress the DNA synthesis induced by neuregulin-1 in puriWed Schwann cell cultures, but this appears not to have been tested. Clearly, the eVects of TGFß on Schwann cells are strikingly context dependent. It is therefore diYcult to predict what the role of TGFß in nerve development might be. In cyclin D1- or cyclin D2-null mice, Schwann cell proliferation is normal in the early post-natal period. In contrast, little proliferation is seen either after nerve transection in cyclin D1-null mice, or in serum puriWed cultured cyclin D1-null Schwann cells in response to either PDGF plus forskolin or neuregulin-1. This selective involvement of cyclin D1 indicates that diVerent mechanisms regulate the proliferation of Schwann cells in damaged and developing nerves (Atanasoski et al., 2001; Kim et al., 2000). Levels of the cyclin dependent kinases (cdk) 2, 4, and 6, involved in progression through the cell cycle, are high in Schwann cells cultured under proliferating conditions (Mathon et al., 2001). Cdk2 levels are controlled transcriptionally, and are low in growth arrested cells, while levels of the cell cycle inhibitory protein p27 are lower in proliferating cells and higher in arrested cells. In early post-natal nerves cdk2 levels are high while they are downregulated in adult nerves (Tikoo et al., 2000). Schwann cell growth arrest can be induced by constitutive activation of Ras or Raf kinase. This results in constitutive activation of the MAPK/ERK pathway. Cell cycle arrest occurs in the G1 phase of the cell cycle via the induction of p21Cip1 (but not p27Kip1), with subsequent inhibition of cyclin/cdk activity (Lloyd et al., 1997; Ridley et al., 1988). Another MAPK pathway, the JNK/c-Jun pathway, is also involved in control of Schwann cell proliferation. In this case, inhibition of JNK/c-Jun reduces Schwann cell DNA synthesis in response to neuregulin-1 (Parkinson et al., 2003a; D. B. Parkinson, R. Mirsky, and K. R. Jessen, unpublished).

MYELINATION Schwann cell myelination represents an exceptionally striking example of a transforming cell-cell interaction, in which contact with certain axons, the large diameter ones, triggers a radical change in the phenotype of immature Schwann cells, leading to the generation of

II. GLIAL CELL DEVELOPMENT

349

350

13. SCHWANN CELL DEVELOPMENT

one of the most highly specialized cell types in the body, the myelinating Schwann cell (Figs. 13.2 and 13.4). Support for this view comes from numerous experiments showing that myelination and myelin maintenance depend on the presence of axons and on the establishment of close contact between axons and Schwann cells. To date, however, no myelination-inducing signal has yet been isolated from axons, nor have myelin-signal receptors been identiWed in Schwann cells. A clear demonstration at the molecular level of cell-cell signaling leading to myelination is therefore lacking. The myelination process has been widely studied both in vivo and in vitro. The role of extracellular matrix proteins and their receptors in myelination has been established and is discussed brieXy in the section titled ‘‘Schwann Cell Cytoskeleton and Interactions with the Basal Lamina’’ (for reviews see Bunge et al., 1986; Chernousev and Carey, 2000; Previtali et al., 2001). Another area in which signiWcant progress has been made is the analysis of transcriptional control mechanisms. The importance of the transcription factor Sox-10 in early development of the glial lineage has already been discussed, but this factor may also have a role in later lineage events (see the discussion that follows). Two other transcription factors, the POU domain factor Oct-6 (SCIP, Tst-1) and the zinc-Wnger protein Krox-20 (Egr-2), have crucial roles in myelination (Bermingham et al., 1996; Jaegle et al., 1996; Lemke et al., 1991; Monuki et al., 1989; Topilko et al., 1994; for reviews of these and other transcription factors in the Schwann cell lineage, see Mirsky and Jessen, 1999; Topilko and Meijer, 2001). In Oct-6 null mice myelination is severely delayed (Bermingham et al., 1996; Jaegle et al., 1996), while in Krox-20 null mice Schwann cell myelination does not occur (Topilko et al., 1994; Topilko and Meijer, 2001). In these mice, Schwann cells successfully achieve a 1:1 relationship with the larger axons and wrap around them up to one and a half times before myelination is arrested. Krox-20 mutations are found associated with Charcot-Marie-Tooth, Dejerrine-Sottas, and hereditary sensory and motor neuropathies, underlining the crucial importance of Krox-20 in myelination (Bellone et al., 1999; Boerkoel et al., 2001; Pareyson et al., 2000; Timmerman et al., 1999; Warner et al., 1998, 1999; Yoshihara et al., 2001). Krox-20 is highly expressed selectively in Schwann cells that have been signaled to myelinate (in mouse from E16 onward) (Parkinson et al., 2002b; Zorick et al., 1996, 1999;), while Oct-6 is expressed in all Schwann cells in late embryogenesis and the early post-natal period, with highest expression in pro-myelin cells (Arroyo et al., 1998; Blanchard et al., 1996). Schwann cells in peripheral nerves of Oct-6 null mice fail to express high levels of Krox-20 at the appropriate time but do so when they myelinate after developmental delay of about two weeks. There is good evidence that this depends on Brn-2, a transcription factor related to Oct-6 (Jaegle et al., 2003). These factors show a similar expression pattern in developing nerves, and myelination is unperturbed in mice lacking Brn-2 alone. In nerves lacking both Brn-2 and Oct-6, however, myelination is more severely delayed than in nerves lacking Oct-6 alone, and overexpression of Brn-2 partially rescues the Oct-6 phenotype. In the initiation of myelination, Brn-2 can therefore substitute for Oct6. In spite of the obvious requirement of these factors for the proper timing of myelination, a large number of axons eventually myelinate, even in the absence of both Oct-6 and Brn-2. (Ghazvini et al., 2002; Sim et al., 2002; Topilko and Meijer, 2001; Wu et al., 2001). In contrast, Schwann cells in nerves of Krox-20 null mice express Oct-6 on schedule but maintain expression at a higher than normal level in the early post-natal period, suggesting that Krox-20 may take part in down-regulating Oct-6 levels. These Schwann cells also have elevated DNA synthesis and death rates, indicating a role for Krox-20 in controlling these processes (Ghazvini et al., 2002; Topilko and Meijer, 2001; Zorick et al., 1999). Krox-20 levels fall sharply on nerve transection and both Oct-6 and Krox-20 are reexpressed in regenerating nerves and in cultured Schwann cells on elevation of cAMP, indicating that in common with myelin protein genes, they are axonally regulated (Murphy et al., 1996; Zorick et al., 1999). Investigations using adenovirally enforced expression of Krox-20 in cultured Schwann cells in combination with gene array technology have reinforced the importance of this gene in myelination. Under these conditions, mRNA for myelin proteins and myelin lipid synthesis is strongly upregulated together with many other genes of unknown function in the Schwann cell (Nagarajan et al., 2001). These and

II. GLIAL CELL DEVELOPMENT

MYELINATION

other studies (Parkinson et al., 2002a, 2003a,b) have shown that enforced expression of Krox-20 is suYcient to organize a widespread set of phenotypic changes in cultured Schwann cells, all of which are associated with the transition from immature proliferating cells to quiescent myelinating cells. In addition to activation of the myelin proteins periaxin and P0, and down-regulation of L1, TGFß induced cell death is also prevented and proliferation induced by neuregulin-1 blocked, both of which represent changes that are associated with myelination in vivo. There is evidence that the underlying mechanisms hinge on the ability of Krox-20 to inactivate the JNK/c-Jun pathway. Activity in this pathway is required for both Schwann cell proliferation and death and also functions to inhibit myelin differentiation. The ability of Krox-20 to suppress JNK/c-Jun activity therefore gives Krox-20 integrated control over myelination, proliferation and death (Parkinson et al., 2003a; D. B. Parkinson, A. Bahaskaran, R. Mirsky, and K. R. Jessen, unpublished) Surprisingly, Krox-20 induces expression of periaxin and P0 in an unrelated cell type, NIH 3T3 cells. The ability to inhibit proliferation and to induce diVerentiation genes in a cell type in which it is normally not expressed is reminiscent of the action of master regulatory genes such as MyoD, neural bHLH factors, or PPARg (Davis et al., 1987; Lee et al., 1995; Lo et al., 1998; Parkinson et al., 2003a,b; D. B. Parkinson, R. Mirsky and K. R. Jessen, unpublished; Sabourin and Rudnicki, 2000; Tontonoz et al., 1994; Zorick et al., 1999). Analysis of the promoters of Oct-6 and Krox-20 has revealed the presence of Schwann cell speciWc enhancer (SCE) elements, which are controlled by axonally regulated transcription factors. In the case of Oct-6, the SCE, which resides within a 4.3 kb sequence 12 kb downstream of the transcription initiation site, is suYcient to drive spatially and temporally correct expression of the gene both developmentally and during regeneration (Ghazvini et al., 2002; Mandemakers et al., 2000). The Krox-20 promoter contains two separate elements, one of which, the immature Schwann cell element (ISE), which is situated upstream of the transcription start site, is active in immature Schwann cells from E15, presumably in pro-myelinating cells, but not in actively myelinating cells. The other, the myelinating Schwann cell element (MSE), is active from E18 onward in myelinating cells. It is contained within a 1.3 kb sequence situated 35 kb downstream of the Krox-20 transcription initiation site, is dependent on Oct-6 for activation, and contains multiple Oct-6 binding sites. Both enhancers are reexpressed sequentially in regenerating nerves (Ghislain et al., 2002). Besides its function in promoting peripheral gliogenesis in early development (see the section titled ‘‘The Origin of Schwann Cells in the Neural Crest’’), Sox-10 is also likely to regulate later steps in the Schwann cell lineage. In a glioblastoma cell line Sox-10 functions synergistically with Oct-6 to promote gene expression of reporter constructs containing recognition sites for both Sox-10 and Oct-6 and also modulates the activity of Pax-3 positively and Krox-20 negatively (Kuhlbrodt et al., 1998a). Sox-10 positively controls the P0 promoter in vivo and in vitro in the N-2A cell line. It also regulates connexin (Cx) 32 promoter in HeLa cells where it functions synergistically with Krox-20 (Bondurand et al., 2001; Peirano et al., 2000). Mutations in the Cx32 gene and in the Krox-20/Sox-10 binding region of the Cx32 promoter result in X-linked Charcot-Marie-Tooth neuropathies. Furthermore, mutated forms of Sox-10 and Krox-20 identiWed in patients with CharcotMarie-Tooth disease fail to transactivate the Cx32 promoter. All of this argues for important functions for Sox-10 in later stages of peripheral glial development. It should be noted that Sox-10 is also required for terminal diVerentiation of oligodendrocytes, the myelinating cells of the CNS (Stolt et al., 2002). In vitro studies indicate that the transcription factor NFkB has a role in myelination (Nickols et al., 2003). In myelinating co-cultures of neurons and Schwann cells NFkB is expressed before Oct-6 and inhibition of NFkB activity prevents myelination. Myelination failure is also seen if co-cultures are prepared using Schwann cells that are null for the p65 subunit of NFkB. A role for HLH factors in regulating Schwann cell myelination has been suggested but remains to be proved. The B class HLH factor Mash2 has been identiWed in a subset of Krox-20 positive myelinating Schwann cells in adult nerves. Mash2 is down-regulated after

II. GLIAL CELL DEVELOPMENT

351

352

13. SCHWANN CELL DEVELOPMENT

nerve transection and inhibits Schwann cell proliferation in culture. It also regulates the expression of Krox-24, Mob-1, and CXCR4 genes in vitro (Kury et al., 2002). In addition, all four HLH Id genes, which inhibit binding of B class HLH proteins to their A class binding partners such as REB and E12/47, are expressed in the Schwann cell lineage (Stewart et al., 1997; Thatikunta et al., 1999). Id 1 and 3 mRNAs are expressed at higher levels in adult nerves than in 10 day nerves (when peak myelination is occurring). Both Id 1 and 3 strongly repress myelin gene promoter activity and Id 3 mRNA is strongly upregulated after nerve cut (Stewart et al., 1997; Thatikunta et al., 1999). The zinc Wnger transcription factor Krox-24 is closely related to Krox-20 and binds to some of the same DNA sequences (Topilko et al., 1997, and references therein). It is already expressed in mouse Schwann cell precursors at E11/12 and persists in nonmyelinating cells in adult nerves, but is up-regulated after nerve cut (Topilko et al., 1997; but see Kury et al., 2002). The expression of Krox-24 in Schwann cells is controlled by Mash2 (Kury et al., 2002). Myelination and regeneration appear to be normal in peripheral nerves of Krox-24 null mice, but increased cell death is seen after transection of neonatal nerves compared with wild-type nerves (Harris, 2001; Topilko et al., 1998;). When we consider cytoplasmic molecules and pathways that are involved in the changes in the Schwann cell that occur on myelination, the case for an involvement of cAMP related pathways is strong. Retroviral inhibition of PKA prevents myelination in neuronSchwann cell co-cultures (Howe and McCarthy, 2000), and many studies have shown that elevation of intracellular cAMP stimulates the expression of myelin-related molecules such as galactocerebroside, periaxin and P0 protein and mRNA in puriWed Schwann cell cultures (Kamholz et al., 1992; Lemke and Chao, 1988; Mews and Meyer, 1993; Morgan et al., 1991, 1994; Parkinson et al., 2003b; Sobue and Pleasure, 1984). Furthermore, Schwann cells driven to express high levels of P0 by cAMP elevation also down-regulate expression of GFAP, N-CAM and p75 LNTR, all of which are down-regulated on myelinating cells in vivo (Morgan et al., 1991) (Fig. 13.5). As mentioned earlier (see the section titled ‘‘Schwann Cell Proliferation’’), the combined presence of cAMP elevating agents and growth factors induces Schwann cell proliferation. Schwann cell division is not induced, however, if cAMP pathways are stimulated in the absence of growth factors or serum (Morgan et al., 1991). A comparison of cAMP-induced P0 expression in the presence (proliferating conditions) and absence (nonproliferating conditions) of growth factors shows much higher P0 levels in the absence of growth factors (Morgan et al., 1991). Furthermore, in individual cells, cAMP-mediated induction of high P0 protein and mRNA levels is not seen in cells that are undergoing DNA synthesis (Morgan et al., 1991, 1994; Stewart et al., 1993). Therefore, cAMP induced myelin-related diVerentiation in vitro is incompatible with proliferation, as is axonally induced myelination in vivo. As mentioned earlier, IGFs promote cAMP dependent Schwann cell proliferation in the presence of other growth factors (see the section titled ‘‘Schwann Cell Proliferation’’). Interestingly, IGFs also promote cAMP induced P0 expression in the absence of growth factors in nonproliferating Schwann cell cultures (Stewart et al., 1996). Similarly IGFs promote Schwann cell attachment and ensheathment of axons, perhaps by enhancing Schwann cell motility (Cheng et al., 1999, 2000b). They also promote myelination in Schwann cell-DRG neuron co-cultures (Cheng and Feldman, 1997; Cheng et al., 1999; Russell et al., 2000). IGFs therefore have the potential to promote Schwann cell motility, proliferation, and diVerentiation. IGFs stimulate the PI3 kinase and Akt pathway in Schwann cells, and there is evidence that this pathway may be involved in early events of myelination in DRGneuron co-cultures, since pharmacological block of PI3 kinase, but not MAP kinase, reversibly inhibits myelination, but not myelin maintenance, perhaps by interfering with Schwann cell-axon interactions involved in early myelination events (Maurel and Salzer, 2000). Progesterone increases the rate of myelination both in vivo and in vitro, although the evidence suggests that it does not aVect the eventual extent of myelination (Chan et al., 1998; Koenig et al., 1995; Notterpek et al., 1999; for reviews see Magnaghi et al., 2001;

II. GLIAL CELL DEVELOPMENT

MYELINATION

FIGURE 13.5 The main changes in protein and lipid expression that take place as the population of immature Schwann cells diverges to generate myelinating and nonmyelinating cells during development. Myelination involves a combination of down-regulation and up-regulation of key molecules, while the generation of nonmyelinating cells involves many fewer molecular changes. 04 antigen and S100 are expressed by all Schwann cells in peripheral nerves, although 04 is down-regulated in the absence of axons. Hereditary demyelinating neuropathies that result from abnormal expression of some of the myelin-associated proteins are indicated. CMT: Charcot-Marie-Tooth neuropathy; PMD: Pelizaeus-Merzbacher disease.

Schumacher et al., 2001). In addition to availability from blood, both neurons and Schwann cells can synthesize progesterone from cholesterol, and there is evidence that both neurons and Schwann cells possess classical and nonclassical progesterone receptors, with higher expression in neurons (Chan et al., 2000; for review see Magnaghi et al., 2001; Schumacher et al., 2001). Several studies suggest that exposure to progesterone elevates P0, and PMP22 mRNA in Schwann cells and causes transient elevation of Krox-20 mRNA and protein (Desarnaud et al., 1998; Guennoun et al., 2001; Magnaghi et al., 2001; Notterpek et al., 1999; Schumacher et al., 2001). On the other hand, Chan and colleagues, using Schwann cell-DRG neuron co-cultures, provide evidence that progesterone synthesized by Schwann cells acts on progesterone receptors in neurons to activate neuronal target genes, which might in turn regulate Schwann cell myelin genes (Chan et al., 2000). The rate of myelination can also be regulated by the neurotrophins, brain-derived neurotrophic factor (BDNF) and NT3. Reduction of BDNF signaling retards myelination during regeneration of peripheral nerves and in neuron-Schwann cell co-cultures, while increasing BDNF levels enhances the formation of myelin (Chan et al., 2001; Zhang et al., 2000;). NT3 appears to act in the opposite way and inhibit myelination (Chan et al., 2001). It is well established that growth arrested Schwann cells will not myelinate in the absence of axonal signals, indicating that growth arrest is not suYcient to trigger myelination. Conversely, it is possible to inhibit myelin-related diVerentiation without simultaneously stimulating DNA synthesis. In Schwann cell cultures, TGFßs suppress cAMPinduced P0, 04 and galactocerebroside induction, even at concentrations too low to induce

II. GLIAL CELL DEVELOPMENT

353

354

13. SCHWANN CELL DEVELOPMENT

Schwann cell proliferation (Mews and Meyer 1993; Morgan et al., 1994; Stewart et al., 1995). In myelinating neuron-Schwann cell cultures, TGFßs have similar eVects, suppressing DNA synthesis, P0 and galactocerebroside induction, and myelination (Einheber et al., 1995; Guenard et al., 1995a). This shows that TGFßs override the myelin diVerentiation pathways activated either by axon-associated myelination signals or by drug-induced elevation of intracellular cAMP, suggesting that a common intracellular signaling mechanism is involved in both cases. In myelinating co-cultures neuregulin-1 inhibits myelination but also induces demyelination in already myelinating cultures (Zanazzi et al., 2001). On the other hand, in transgenic mice, controlled excision of ErbB2 driven by Krox-20-Cre, which disrupts neuregulin signaling, results in hypomyelination of peripheral nerves, suggesting that in vivo neuregulin signaling promotes rather than inhibits myelination (Garratt et al., 2000a, 2000b; Zanazzi et al., 2001). Another growth factor, glial derived neurotrophic factor (GDNF), promotes Schwann cell proliferation and myelination of small diameter axons that would normally not myelinate when it is administered exogenously to adult rats (Hoke et al., 2003). The direct effect of GDNF is probably to increase the size and axon diameter of the smaller neurons, which in turn promotes proliferation and myelination of Schwann cells.

NEURAL ACTIVITY Recent investigations suggest that neural activity may play a part in regulating glial development in peripheral nerves. In particular, they highlight a role for neuron-derived ATP, acting via P2 type receptors, in regulating the phenotype of immature Schwann cells (Amedee and Despeyroux, 1995; Colomar and Amedee, 2001; Lyons et al., 1994, 1995; Mayer et al., 1997; Stevens et al., 1998; Stevens and Fields, 2000). Electrical stimulation of neurons in co-cultures of DRG neurons and Schwann cells leads to calcium elevation in the soma of individual neurons which is followed by delayed calcium elevation in Schwann cells associated with the neurites. This is caused by nonsynaptic ATP release from the neurons acting via P2 receptors (Stevens and Fields, 2000). In response to ATP, cell proliferation is inhibited at the same time as Schwann cell diVerentiation is delayed, as measured by acquisition of 04 antigen and myelination (Stevens et al., 1998; Stevens and Fields, 2000). Purinergic receptors of the P2Y and P2X(7) subtypes have been reported in primary Schwann cells (Colomar and Amedee, 2001; Lyons et al., 1994), and P2U receptors have been reported in immortalized cells (Berti-Mattera et al., 1996). It is likely that the eVect of ATP on proliferation in the co-culture system is caused by depression of adenylate cyclase activity induced by the action of P2Y receptors. In normal nerve development, cessation of proliferation is associated with the onset of myelination (Brown and Asbury, 1981; Stewart et al., 1993) and takes place well after acquisition of 04 antigen (Dong et al., 1999; Mirsky et al., 1990). Therefore, further studies will be required to clarify how the eVects of ATP signaling seen in these experiments integrate with other signals to generate the co-ordinated pattern of Schwann cell maturation, proliferation, and myelination seen in vivo. Related to these results, in pure neuronal DRG cultures, experiments using a comparable approach have shown that complicated patterns of neural activity lower neuronal expression of adhesion molecules such as N-cadherin and L1 (Itoh et al., 1995, 1997). It has also been reported that in DRG-Schwann cell co-cultures, low frequency stimulation of neurons, which reduces expression of L1 on axons, but not on Schwann cells, reduces myelination in these co-cultures to one third of normal levels, while higher frequency stimulation has no eVect. Decreased axonal L1 also aVects the adhesion of Schwann cells to neurites in the cultures in a short-term assay (Stevens et al., 1998; Stevens and Fields, 2000). A role for L1 prior to myelination is supported by some other studies. Decreased L1 expression increases defasciculation both in vivo and in vitro (Landmesser et al., 1988; Honig et al., 2002; Lin and Goodman, 1994), while antibodies to L1 interfere with axonSchwann cell interactions in a co-culture system and prevent Schwann cell alignment along

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

axons, which leads to a failure to myelinate (Seilheimer and Schachner, 1988; Seilheimer et al., 1989; Wood et al., 1990). Although all of these in vitro studies point to a role for L1 in early axon-Schwann cell interactions and myelination, no major disturbance in development of either myelinated or unmyelinated Wbers has been described in mice in which L1 has been inactivated. In adult life, disturbances in maintenance of axon-Schwann cell relationships in sensory Wbers occur, with subsequent axonal loss (Dahme et al., 1997; Haney et al., 1999).

SCHWANN CELL CYTOSKELETON AND INTERACTIONS WITH THE BASAL LAMINA The Basal Lamina and Laminin Signaling from the extracellular matrix, and in particular the basal lamina, plays an important role in polarization and diVerentiation of diverse cell types, including Schwann cells. Many studies have indicated the importance of the Schwann cell basal lamina in myelin formation (Bunge, 1993), and recent experiments have started to unravel the connections between the major Schwann cell extracellular matrix molecules, their membrane receptors, and the intracellular signaling systems that are linked to them. Laminin, a major component of the basal lamina, exists in many diVerent forms. The major form in the Schwann cell basal lamina is laminin 2 (merosin), consisting of a2, ß1 and g1 chains, although smaller amounts of laminins 4 and 1 and other laminins are also present (Jaakola et al., 1993; Lentz et al., 1997; Nakagawa et al., 2001). The dy/dy mouse has mutations in the a2 chain present in laminins 2 and 4, leading to defective laminin polymer formation. In addition to skeletal muscle atrophy, this mouse shows complex pathology of peripheral motor nerves (for review see Bo¨nnemann et al., 1996; Colognato and Yurchenco, 1999; Matsumura et al., 1997b; Nakagawa et al., 2001). The most severe Schwann cell abnormalities observed in the dy/dy mouse are found in the dorsal and ventral spinal roots where naked axons are surrounded by undiVerentiated Schwann cells that fail to penetrate the axon bundles. More general defects seen not only in the roots but also within peripheral nerve bundles include hypomyelination, excessively wide nodes of Ranvier, overlapping Schwann cells and a patchy basal lamina surrounding individual nerve Wbers (Aguayo and Bray, 1982; Uziyel et al., 2000). The three major receptors for laminin 2 in the Schwann cell membrane are integrins a6ß1 and a6ß4 and a-dystroglycan, a member of the dystrophin complex of proteins (Engvall et al., 1992; Ervasti and Campbell, 1993; Hall et al., 1990; Masaki et al., 2002; Matsumura et al., 1997a).

Integrins Several integrins including a1ß1, a2ß1, a5ß1, a6ß1, a6ß4, a7ß1, ß8 integrin, avß3, and avß1, plus very low levels of a4 and a5, are associated with Schwann cells in vivo or in culture (Einheber et al., 1995; Fernandez-Valle´ et al., 1994; Milner et al., 1997; Stewart et al., 1997; Previtali et al., 2003). It has also been established that axons, Schwann cell precursors, and Schwann cells express combinations of integrin receptors that vary with the stage of peripheral nerve development (reviewed in Previtali et al., 2001). The importance of integrin-ß1 signaling in the late embryonic and subsequent phases of peripheral nerve development has been revealed by conditional disruption of integrin-ß1 speciWcally in Schwann cells, using a P0-Cre transgene to excise ß1 integrin from a Xoxed allele between E13.5-14.5 in mice. Thus Schwann cell precursors can populate the nerves and generate Schwann cells just before the ß1 integrin is excised. The null Schwann cells proliferate and survive normally, but show severely abnormal relationships with axons. In many cases, axons fail to segregate properly, and never achieve the 1:1 relationship with the Schwann cell required for myelination. Although some axons are myelinated in adult mice, even around these axons myelination is developmentally delayed. Some aspects of

II. GLIAL CELL DEVELOPMENT

355

356

13. SCHWANN CELL DEVELOPMENT

the phenotype resemble that seen in the nerve roots of the laminin 2 deWcient dy/dy mouse, and it is suggested that the main partner for the ß1 integrin is likely to be laminin 2, the predominant isoform in Schwann cells, acting via the a6ß1 integrin receptor, which is highly expressed in embryonic Schwann cells (Feltri et al., 2002; Previtali et al., 2002). Absence of ß1 integrin would then lead to a failure to link the laminin in the basal lamina to the Schwann cell cytoskeleton, a process that is required for ensheathment of axons (Feltri et al., 2002). This interpretation is supported by abundant evidence from earlier in vitro studies of the role of the basal lamina in myelination (Bunge, 1993), and on the involvement of ß1 integrins in this process. In myelinating Schwann cell-neuron co-cultures, a6ß1 is expressed by Schwann cells prior to myelination and antibodies to ß1 block myelination whereas antibodies to a1ß1 do not, implying that a6ß1 (the A variant) is the receptor required for myelination to proceed (Fernandez-Valle´ et al., 1994; Hogervost et al., 1993). Integrins are frequently linked to the cytoskeleton and it is relevant to note that drug-induced disruption of the actin cytoskeleton also leads to a block of myelination (Fernandez-Valle´ et al., 1997). Furthermore, in pull-down assays ß1 integrin co-precipitates with focal adhesion kinase and the actin-linked protein paxillin in diVerentiating neuron-Schwann cell cocultures but not in Schwann cell cultures. Tyrosine phosphorylation of both focal adhesion kinase and paxillin also increases in co-cultures as Schwann cells form basal lamina and diVerentiate, suggesting an active role for ß1 mediated intracellular signaling in the diVerentiation process (Chen et al., 2000). Activation of focal adhesion kinase is also induced by IGF-1, acting via a PI3 kinase pathway to activate the small GTPase rac. This acts together with focal adhesion kinase to induce Schwann cell process extension and motility (Cheng et al., 2000b). Whether integrins take part in this process is not known. As mentioned earlier, ß1 interacts with paxillin in diVerentiating Schwann cells. Paxillin, in turn, binds to the tumor suppressor protein Schwannomin/merlin, enabling it to localize at the plasma membrane and associate with ß1 integrin and ErbB2 (FernandezValle´ et al., 2002). Schwannomin, the product of the neuroWbromatosis type 2 tumor suppressor gene, is related to the radixin/ezrin/moesin family of proteins which links membrane proteins to the actin cytoskeleton in epithelia and other cell types (Rouleau et al., 1993; Trofatter et al., 1993). Since mutations in Schwannomin lead to an increased frequency of Schwannomas, and most sporadic Schwannomas and meningiomas carry mutations in Schwannomin, this gene is likely to play an important role in the negative control of Schwann cell proliferation. In myelinating Schwann cells it is co-localized at paranodal membranes with RhoA, a protein associated with actin-based movement in other cell types, and Schwannomin interacts directly with ßII spectrin, present in Schwann cells. Spectrin in turn binds to actin and treatment with antisense nucleotides to Schwannomin causes cells to round up, with concomitant changes to the actin cytoskeleton (Scherer and Gutmann, 1996; Scoles et al., 1998). Schwannomin also binds to the cytoplasmic tail of CD44, a protein implicated in neuregulin-1 signaling (see the section titled ‘‘The Role of Neuregulin-1 in Schwann Cell Development’’ (Sainio et al., 1997). Focal adhesion assembly of paxillin and vinculin via the Rho/p160 Rho-associated kinase (ROCK) pathway, and the induction of N-cadherin/catenin mediated cell-cell contacts can be induced in cultured Schwann cells by the phospholipid signaling molecule LPA (see the section titled ‘‘Schwann Cell Survival Signals’’). LPA also has profound eVects on Schwann cell morphology, inducing bipolar Schwann cells to Xatten and rearrangement of the actin cytoskeleton into wreath-like structures (Weiner et al., 2001). In culture, N-cadherin mediates axon-aligned process growth of Schwann cells and Schwann cell-Schwann cell contacts, a role consistent with its distribution in vivo (Shibuya et al., 1995; Wanner and Wood, 2002). Finally, in mice carrying mutations in the actin-binding protein dystonin, there is abnormal myelination and an autonomous Schwann cell defect, in addition to a neuronal one. Cultured Schwann cells from these mice have a severely disorganized cytoskeleton, fail to attach normally to laminin, and are abnormally polarized (Bernier et al., 1998) Another process likely to involve integrins is Schwann cell migration, which is of importance in embryonic nerves and regeneration. Antibodies to ß1 block Schwann cell

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

migration on laminins 1 and 2 in response to neuregulin and forskolin. Antibodies to a6ß1 block migration on laminin 1 but not laminin 2, while RGD peptides block migration on Wbronectin, suggesting involvement of av integrins (Milner et al., 1997). The main intracellular pathway used in migration on laminin appears to be the MAPK/ERK pathway (Meintanis et al., 2001). Integrins a3, a6, and ß1 also associate with the tetraspan protein CD9, which has been implicated in Schwann cell adhesion, proliferation and migration (Anton et al., 1995; Hadjiargyrou et al., 1996). While a6ß1 is expressed earlier in development, a related integrin, a6ß4, is strongly expressed in diVerentiated Schwann cells including myelinating Schwann cells, but a peripheral nerve phenotype has not been reported in ß4 knockout mice (Einheber et al., 1993; Feltri et al., 1994, 2002; Fernandez-Valle´ et al., 1994; Niessen et al., 1994; van der Neut et al., 1996; see Mirsky and Jessen, 1996, for discussion of this point). Likewise, prominent ß8 immunolabelling has been detected around myelin sheaths in mouse peripheral nerve, suggesting it might be involved in myelination, but this has not been tested (Milner et al., 1997). Possible roles for another integrin, a1ß1, are suggested by its distribution in vivo. It is not strongly expressed by Schwann cell precursors or embryonic or neonatal Schwann cells in developing nerves, but it is expressed by mature nonmyelinating Schwann cells and by all Schwann cells in adult nerve after transection. It might therefore have a role in the interaction of small diameter axons with their surrounding Schwann cells, in the interaction of regenerating axons with Schwann cells or in Schwann cell migration (Stewart et al., 1997). Evidence that integrins have a role in the early embryonic development of the Schwann cell lineage comes from mice lacking a4 or a5 integrins (Haack and Hynes, 2001). The a5 receptor, in combination with integrin b1, binds to Wbronectin, while a4 is more promiscuous, binding not only to Wbronectin but also participating in homophilic and cell-cell interactions. In a4 deWcient embryos, which die at E11.5 of cardiac defects (Yang et al., 1995), initial development of the PNS appears normal, with DRG ganglia and peripheral nerves indistinguishable from wild-type controls. This suggests that absence of a4 does not aVect neural crest migration. In support of this, in cultured explants of the neural crest, lack of a4 does not aVect the number of glial cells generated but is vital for cell survival. When null cells are injected into the neural crest pathway in chick embryos, cell numbers in DRG and peripheral nerves are reduced compared with wild-type injected cells, supporting the idea that a4 is involved in early glial cell survival. a5 null embryos die earlier, during the neural crest migratory phase, but trunk glial development can be followed in vitro, using ventral neural tube explants (Yang et al., 1995). The evidence from explant and transplant studies suggests that a5 is required for proliferation of a cell type (suggested to be a stage prior to the Schwann cell precursor stage) early in the peripheral glial lineage (Haack and Hynes, 2001).

Dystroglycan The role of dystroglycan as an important laminin receptor is increasingly recognized. When laminin binds to dystroglycan, it can assemble a nascent basal lamina, which can then incorporate collagen type IV. In contrast, laminin interaction with ß1 integrin induces a Wbrillar matrix (Tsiper and Yurchenko, 2002). a-dystroglycan binds to the transmembrane protein ß-dystroglycan, which, in turn, binds to dystrophin in the cell interior. Selective deletion of dystroglycan in Schwann cells results in disorganized microvilli at the Node of Ranvier, reduced nodal sodium channel density, and slowed conduction velocity (Saito et al., 2003). The form of dystrophin found in Schwann cells, DRP 116, lacks an actin binding domain, but dystroglycan can bind to Schwann cell utrophin, which has an actin binding domain (Sadoulet-Puccio and Kunkel, 1996; Sherman et al., 2001). Dystroglycan can also bind to the recently described Schwann cell dystrophin isoform DRP2. This forms a complex with a- and ß-dystroglycan and the Schwann cell cytoskeletal protein periaxin (Sherman et al., 2001). Periaxin exists in two alternatively spliced forms, both of which contain an N-terminal PDZ domain. In developing nerves, initial activation

II. GLIAL CELL DEVELOPMENT

357

358

13. SCHWANN CELL DEVELOPMENT

of the periaxin gene is seen in a sub-population of early immature Schwann cells in embryonic nerves where it appears to be an early indicator of myelin diVerentiation. Post-natally periaxin is clearly restricted to myelinating cells (Parkinson et al., 2002b; Scherer et al., 1995). The large form of periaxin is associated with the plasma membrane in myelinating Schwann cells, where it forms clustered patched complexes with DRP2 (Dytrych et al., 1998; Gillespie et al., 1994, 2000; Scherer et al., 1995; Sherman et al., 2001). Mutations of periaxin result in severe hypomyelinating sensory neuropathy in mice and Charcot-Marie-Tooth neuropathy in humans (Gillespie et al., 2000; Sherman et al., 2001; Takashima et al., 2002; Williams and Brophy, 2002). Periaxin null mice myelinate normally but with time develop severe demyelination and loss of sensory function so it is clear that these complexes are essential to the stability of the myelinating Schwann cell although periaxin and DRP2 are not themselves components of the myelin sheath (Gillespie et al., 2000; Sherman et al., 2001).

Laminin and Leprosy An important interaction of Schwann cell-derived laminin 2 is seen in leprosy. This disease, caused by Mycobacterium leprae (M. leprae), is characterized by inWltration and infection of the Schwann cells of sensory nerves of the skin and elsewhere. The phenolic glycolipid1 of the M. leprae cell wall interacts speciWcally with the G domain of the laminin a2 chain of laminin 2. This interferes with the laminin interaction with a-dystroglycan, which normally initiates signaling through the dystroglycan complexes in Schwann cells (Brophy, 2002; Ng et al., 2000; Rambukkana et al., 1997, 1998; Sherman et al., 2001). A recent paper, using Schwann cell-DRG neuron co-cultures, shows that M. leprae interaction induces diVerent responses in myelinating and nonmyelinating Schwann cells. Perhaps diVerent signaling cascades are initiated in the two Schwann cell types because myelinating Schwann cells express the DRP2 dystrophin isoform, in addition to DRP116 and utrophin expressed by both myelinating and nonmyelinating cells (Brophy, 2002; Rambukkana et al., 2002). In nonmyelinating cells the bacteria are taken up, while in myelinating cells, binding of the phenolic glycolipid induces rapid demyelination with little uptake of bacteria (Rambukkana et al., 2002). Whether this is relevant to the pathology of leprosy remains to be determined.

References Adlkofer, K., and Lai, C. (2000). Role of neuregulins in glial cell development. Glia 29, 104–111. Aguayo, A. J., and Bray, G. M. (1982). Developmental disorders of myelination in mouse mutants. In ‘‘Neuronalglial Cell Interrelationships’’ (T. A. Sears, ed.), pp. 57–76. Dahlem Konferenzen, Springer-Verlag, Berlin, Heidelberg, New York. Amedee, T., and Despeyroux, S. (1995). ATP activates cationic and anionic conductances in Schwann cells cultured from dorsal root ganglia of the mouse. Proc. R. Soc. Lond. B Biol. Sci. 259, 277–284. Anderson, D. J. (2000). Genes, lineages and the neural crest: A speculative review. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355, 953–964. Anton, E. S., Hadjiargyrou, M., Patterson, P. H., and Matthew, W. D. (1995). CD9 plays a role in Schwann cell migration in vitro. J. Neurosci. 15, 584–595. Archelos, J. J., Roggenbuck, K., Schneider-Schaulies, J., Linington, C., Toyka, K. V., and Hartung, H. P. (1993). Production and characterization of monoclonal antibodies to the extracellular domain of P0. J. Neurosci. Res. 35, 46–53. Arroyo, E. J., Bermingham, J. R. Jr., Rosenfeld, M. G., and Scherer, S. S. (1998). Promyelinating Schwann cells express Tst-1/SCIP/Oct-6. J. Neurosci. 18, 7891–7902. Atanasoski, S., Shumas, S., Dickson, C., Scherer, S. S., and Suter, U. (2001). DiVerential cyclin D1 requirements of proliferating Schwann cells during development and after injury. Mol. Cell. Neurosci. 18, 581–592. Baechner, D., Liehr, T., Hameister, H., Altenberger, H., Grehl, H., Suter, U., and Rautenstrauss, B. (1995). Widespread expression of the peripheral myelin protein-22 gene (PMP22) in neural and non-neural tissues during murine development. J. Neurosci. Res. 15, 733–741. Banerjee, S. A., and Patterson, P. H. (1995). Schwann cell CD9 expression is regulated by axons. Mol. Cell. Neurosci. 6, 462–473. Bellone, E., Di Maria, E., Soriani, S., Varese, A., Doria, L. L., Ajmar, F., and Mandich, P. (1999). A novel mutation (D305V) in the early growth response 2 gene is associated with severe Charcot-Marie-Tooth type 1 disease. Hum. Mutat. 14, 353–354.

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

Benzel, I., Barde, Y. A., and Casademunt, E. (2001). Strain-speciWc complementation between NRIF1 and NRIF2, two zinc Wnger proteins sharing structural and biochemical properties. Gene 281, 19–30. Bermingham, J. R., Scherer, S. S., O’Connell, S., Arroyo, E., Kalla, K. A., Powell, F. L., and Rosenfeld, M. G. (1996). Tst-1/Oct-6/SCIP regulates a unique step in peripheral myelination and is required for normal respiration. Genes Dev. 10, 1751–1762. Bernier, G., De Repentigny, Y., Mathieu, M., David, S., and Kothary, R. (1998). Dystonin is an essential component of the Schwann cell cytoskeleton at the time of myelination. Development 125, 2135–2148. Berti-Mattera, L. N., Wilkins, P. L., Madhun, Z., and Suchovsky, D. (1996). P2-purigenic receptors regulate phospholipase C and adenylate cyclase activities in immortalized Schwann cells. Biochem. J. 314, 555–561. Bhattacharyya, A., Frank, E., Ratner, N., and Brackenbury, R. (1991). P0 is an early marker of the Schwann cell lineage in chickens. Neuron 7, 831–844. Bitgood, M. J., and McMahon, A. P. (1995). Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev. Biol. 172, 126–138. Bixby, S., Kruger, G. M., Mosher, J. T., Joseph, N. M., and Morrison, S. J. (2002). Cell-intrinsic diVerences between stem cells from diVerent regions of the peripheral nervous system regulate the generation of neural diversity. Neuron 35, 643–656. Blanchard, A. D., Sinanan, A., Parmantier, E., Zwart, R., Broos, L., Meijer, D., Meier, C., Jessen, K. R., and Mirsky, R. (1996). Oct-6 (SCIP/Tst-1) is expressed in Schwann cell precursors, embryonic Schwann cells, and postnatal myelinating Schwann cells: Comparison with Oct-1, Krox-20 and Pax-3. J. Neurosci. Res. 46, 630–640. Boerkoel, C. F., Takashima, H., Bacino, C. A., Daentl, D., and Lupski, J. R. (2001). EGR2 mutation R359W causes a spectrum of Dejerine-Sottas neuropathy. Neurogenetics 3, 153–157. Bondurand, N., Girard, M., Pingault, V., Lemort, N., Dubourg, O., and Goossens, M. (2001). Human Connexin 32, a gap junction protein altered in the X-linked form of Charcot-Marie-Tooth disease, is directly regulated by the transcription factor SOX10. Hum. Mol. Genet. 10, 2783–2795. Bo¨nnemann, C. G., McNally, E. M., and Kunkel, L. M. (1996). Beyond dystrophin: Current progress in the muscular dystrophies. Curr. Opin. Pediatr. 8, 569–582. Brennan, A., Dean, C. H., Zhang, A. L., Cass, D. T., Mirsky, R., and Jessen, K. R. (2000). Endothelins control the timing of Schwann cell generation in vitro and in vivo. Dev. Biol. 227, 545–557. Britsch, S., Goerich, D. E., Riethmacher, D., Peirano, R. I., Rossner, M., Nave, K. A., Birchmeier, C., and Wegner, M. (2001). The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15, 66–78. Britsch, S., Li, L., KirchhoV, S., Theuring, F., Brinkmann, V., Birchmeier, C., and Riethmacher, D. (1998). The ErbB2 and ErbB3 receptors and their ligand, neuregulin-1, are essential for development of the sympathetic nervous system. Genes Dev. 12, 1825–1836. Brophy, P. J. (2002). Microbiology. Subversion of Schwann cells and the leper’s bell. Science 296, 862–863. Brown, M. J., and Asbury, A. K. (1981). Schwann cell proliferation in the postnatal mouse: Timing and topography. Exp. Neurol. 74, 170–186. Bunge, R. P. (1993). Expanding roles for the Schwann cell: Ensheathment, myelination, trophism and regeneration. Curr. Opin. Neurobiol. 3, 805–809. Bunge, R. P., Bunge, M. B., and Eldridge, C. F. (1986). Linkage between axonal ensheathment and basal lamina production by Schwann cells. Annu. Rev. Neurosci. 9, 305–328. Buonanno, A., and Fischbach, G. D. (2001). Neuregulin and ErbB receptor signaling pathways in the nervous system. Curr. Opin. Neurobiol. 11, 287–296. Burden, S., and Yarden, Y. (1997). Neuregulins and their receptors: A versatile signaling module in organogenesis and oncogenesis. Neuron 18, 847–855. Campana, W. M., Darin, S. J., and O’Brien, J. S. (1999). Phosphatidylinositol 3-kinase and Akt protein kinase mediate IGF-I-and prosaptide-induced survival in Schwann cells. J. Neurosci. Res. 57, 332–341. Carraway, K. L. 3rd, Weber, J. L., Unger, M. J., Ledesma, J., Yu, N., Gassmann, M., and Lai, C. (1997). Neuregulin-2, a new ligand of ErbB3/ErbB4-receptor tyrosine kinases. Nature 387, 512–516. Carroll, S. L., Miller, M. L., Frohnert, P. W., Kim, S. S., and Corbett, J. A. (1997). Expression of neuregulins and their putative receptors, ErbB2 and ErbB3, is induced during Wallerian degeneration. J. Neurosci. 17, 1642–1659. Casademunt, E., Carter, B. D., Benzel, I., Frade, J. M., Dechant, G., and Barde, Y. A. (1999). The zinc Wnger protein NRIF interacts with the neurotrophin receptor p75(NTR) and participates in programmed cell death. EMBO J. 18, 6050–6061. Cass, L. A., and Meinkoth, J. L. (1998). DiVerential eVects of cyclic adenosine 3’,5’-monophosphate on p70 ribosomal S6 kinase. Endocrinology 139, 1991–1998. Cass, L. A., and Meinkoth, J. L. (2000). Ras signaling through PI3K confers hormone-independent proliferation that is compatible with diVerentiation. Oncogene 19, 924–932. Cass, L. A., Summers, S. A., Prendergast, G. V., Backer, J. M., Birnbaum, M. J., and Meinkoth, J. L. (1999). Protein kinase A-dependent and -independent signaling pathways contribute to cyclic AMP-stimulated proliferation. Mol. Cell. Biol. 19, 5882–5891. Chan, J. R., Cosgaya, J. M., Wu, Y. J., and Shooter, E. M. (2001). Neurotrophins are key mediators of the myelination program in the peripheral nervous system. Proc. Natl. Acad. Sci. U. S. A. 98, 14661–14668. Chan, J. R., Phillips, L. J. 2nd, and Glaser, M. (1998). Glucocorticoids and progestins signal the initiation and enhance the rate of myelin formation. Proc. Natl. Acad. Sci. U. S. A. 95, 10459–10464.

II. GLIAL CELL DEVELOPMENT

359

360

13. SCHWANN CELL DEVELOPMENT

Chan, J. R., Rodriguez-Waitkus, P. M., Ng, B. K., Liang, P., and Glaser, M. (2000). Progesterone synthesized by Schwann cells during myelin formation regulates neuronal gene expression. Mol. Biol. Cell 11, 2283–2295. Chen, L. M., Bailey, D., and Fernandez-Valle´, C. (2000). Association of beta 1 integrin with focal adhesion kinase and paxillin in diVerentiating Schwann cells. J. Neurosci. 20, 3776–3784. Cheng, H. L., and Feldman, E. L. (1997). Insulin-like growth factor-I (IGF-I) and IGF binding protein-5 in Schwann cell diVerentiation. J. Cell. Physiol. 171, 161–167. Cheng, L., Esch, F. S., Marchionni, M. A., and Mudge, A. W. (1998). Control of Schwann cell survival and proliferation: Autocrine factors and neuregulins. Mol. Cell. Neurosci. 12, 141–156. Cheng, L., and Mudge, A. W. (1996). Cultured Schwann cells constitutively express the myelin protein P0. Neuron 16, 309–319. Cheng, H. L, Russell, J. W., and Feldman, E. L. (1999). IGF-I promotes peripheral nervous system myelination. Ann. N. Y. Acad. Sci. 883, 124–130. Cheng, H. L., Steinway, M., Delaney, C. L., Franke, T. F., and Feldman, E. L. (2000a). IGF-I promotes Schwann cell motility and survival via activation of Akt. Mol. Cell. Endocrinol. 170, 211–215. Cheng, H..L., Steinway, M. L., Russell, J. W., and Feldman, E. L. (2000b). GTPases and phosphatidylinositol 3-kinase are critical for insulin-like growth factor-I-mediated Schwann cell motility. J. Biol. Chem. 275, 27197–27204. Chernousov, M. A., and Carey, D. J. (2000). Schwann cell extracellular matrix molecules and their receptors. Histol. Histopathol. 15, 593–601. Chittka, A., and Chao, M. V. (1999). IdentiWcation of a zinc Wnger protein whose subcellular distribution is regulated by serum and nerve growth factor. Proc. Natl. Acad. Sci. U. S. A. 96, 10705–10710. Chuah, M. I., and West, A. K. (2002). Cellular and molecular biology of ensheathing cells. Microsc. Res. Tech. 58, 216–227. Ciment, G. (1990). The melanocyte Schwann cell progenitor: A bipotent intermediate in the neural crest lineage. Comments Dev. Neurobiol. 1, 207–223. Ciutat, D., Caldero´, J., Oppenheim, R. W., and Esquerda, J. E. (1996). Schwann cell apoptosis during normal development and after axonal degeneration induced by neurotoxins in the chick embryo. J. Neurosci. 16, 3979–3990. Colognato, H., and Yurchenco, P. D. (1999). The laminin alpha2 expressed by dystrophic dy(2J) mice is defective in its ability to form polymers. Curr. Biol. 9, 1327–1330. Colomar, A., and Amedee, T. (2001). ATP stimulation of P2X(7) receptors activates three diVerent ionic conductances on cultured mouse Schwann cells. Eur. J. Neurosci. 14, 927–936. Conlon, I. J., Dunn, G. A., Mudge, A. W., and RaV, M. C. (2001). Extracellular control of cell size. Nature Cell Biol. 3, 918–921. Conlon, I., and Raff, M. (2003). Differences in the way mammalian and yeast cells coordinate cell growth and cellcyle progression. J. Biol. 2, 7. Cowell, L. A., and Weston, J. A. (1970). An analysis of melanogenesis in cultured chick embryo spinal ganglia. Dev. Biol. 22, 670–697. Curtis, R., Stewart, H. J., Hall, S. M., Wilkin, G. P., Mirsky, R., and Jessen, K. R. (1992). GAP-43 is expressed by nonmyelin-forming Schwann cells of the peripheral nervous system. J. Cell Biol. 116, 1455–1464. Dahme, M., Bartsch, U., Martini, R., Anliker, B., Schachner, M., and Mantei, N. (1997). Disruption of the mouse L1 gene leads to malformations of the nervous system. Nat. Genet. 17, 346–349. Davis, R. L., Weintraub, H., and Lassar, A. B. (1987). Expression of a single transfected cDNA converts Wbroblasts to myoblasts. Cell 51, 987–1000. Delaney, C. L., Cheng, H. L., and Feldman, E. L. (1999). Insulin-like growth factor-I prevents caspase-mediated apoptosis in Schwann cells. J. Neurobiol. 41, 540–548. Desarnaud, F., Do Thi, A. N., Brown, A. M., Lemke, G., Suter, U., Baulieu, E. E., and Schumacher, M. (1998). Progesterone stimulates the activity of the promoters of peripheral myelin protein-22 and protein zero genes in Schwann cells. J. Neurochem. 71, 1765–1768. Dong, Z., Brennan, A., Liu, N., Yarden, Y., Lefkowitz, G., Mirsky, R., and Jessen, K. R. (1995). NDF is a neuron-glia signal and regulates survival, proliferation, and maturation of rat Schwann cell precursors. Neuron 15, 585–596. Dong, Z., Sinanan, A., Parkinson, D., Parmantier, E., Mirsky, R., and Jessen, K. R. (1999). Schwann cell development in embryonic mouse nerves. J. Neurosci. Res. 56, 334–348. Dowsing, B. J., Morrison, W. A., Nicola, N. A., Starkey, G. P., Bucci, T., and Kilpatrick, T. J. (1999). Leukemia inhibitory factor is an autocrine survival factor for Schwann cells. J. Neurochem. 73, 96–104. Dupin, E., Real, C., Glavieux-Pardanaud, C., Vaigot, P., and Le Douarin, N. M. (2003). Reversal of developmental restrictions in neural crest lineages: transition from Schwann cells to glial-melanocytic precursors in vitro. Proc. Natl. Acad. Sci. USA 100, 5229–5233. Dytrych, L., Sherman, D. L., Gillespie, C. S., and Brophy, P. J. (1998). Two PDZ domain proteins encoded by the murine periaxin gene are the result of alternative intron retention and are diVerentially targeted in Schwann cells. J. Biol. Chem. 273, 5794–5800. Eccleston, P. A. (1992). Regulation of Schwann cell proliferation: Mechanisms involved in peripheral nerve development. Exp. Cell Res. 199, 1–9. Eccleston, P. A., Jessen, K. R., and Mirsky, R. (1989). Transforming growth factor-beta and gamma-interferon have dual eVects on growth of peripheral glia. J. Neurosci. Res. 24, 524–530.

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

Einheber, S., Hannocks, M.-J., Metz, C. N., Rifkin, D. B., and Salzer, J. L. (1995). Transforming growth factorbeta 1 regulates axon-Schwann cell interactions. J. Cell Biol. 129, 443–458. Einheber, S., Milner, T., Giancotti, F., and Salzer, J. (1993). Axonal regulation of Schwann cell integrin expression suggests a role for alpha6 beta4 in myelination. J. Cell Biol. 123, 1223–1236. Engvall, E., Earwicker, D., Day, A., Muir, D., Manthorpe, M., and Paulsson, M. (1992). Merosin promotes cell attachment and neurite outgrowth and is a component of the neurite-promoting factor of RN22 schwannoma cells. Exp. Cell Res. 198, 115–123. Ervasti, J. M., and Campbell, K. P. (1993). A role for the dystrophin-glycoprotein complex as a transmembrane linker between laminin and actin. J. Cell Biol. 122, 809–823. Falls, D. L., Rosen, K. M., Corfas, G., Lane, W. S., and Fischbach, G. D. (1993). ARIA, a protein that stimulates acetylcholine receptor synthesis, is a member of the neu ligand family. Cell 72, 801–815. Fawcett, J. W., and Keynes, R. J. (1990). Peripheral nerve regeneration. Annu. Rev. Neurosci. 13, 43–60. Feltri, M. L., Graus Porta, D., Previtali, S. C., Nodari, A., Migliavacca, B., Cassetti, A., Littlewood-Evans, A., Reichardt, L. F., Messing, A., Quattrini, A., Mueller, U., and Wrabetz, L. (2002). Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with axons. J. Cell Biol. 156, 199–209. Feltri, M. L., Scherer, S. S., Nemni, R., Kamholz, J., Vogelbacker, H., Oronzi Scott, M., Canal, N., Quaranta, V., and Wrabetz, L. (1994). ß4 integrin expression in myelinating Schwann cells is polarized, developmentally regulated and axonally dependent. Development 120, 1287–1301. Feng, L., Hatten, M. E., and Heintz, N. (1994). Brain lipid-binding protein (BLBP): A novel signaling system in the developing mammalian CNS. Neuron 12, 895–908. Fernandez-Valle´, C., Gorman, D., Gomez, A. M., and Bunge, M. B. (1997). Actin plays a role in both changes in cell shape and gene-expression associated with Schwann cell myelination. J. Neurosci. 17, 241–250. Fernandez-Valle´, C., Gwynn, L., Wood, P. M., Carbonetto, S., and Bunge, M. (1994). Anti-ß1 integrin antibody inhibits Schwann cell myelination. J. Neurobiol. 25, 1207–1226. Fernandez-Valle´, C., Tang, Y., Ricard, J., Rodenas-Ruano, A., Taylor, A., Hackler, E., BiggerstaV, J., and Iacovelli, J. (2002). Paxillin binds schwannomin and regulates its density-dependent localization and eVect on cell morphology. Nat. Genet. 31, 354–362. Ferri, C. C., and Bisby, M. A. (1999). Improved survival of injured sciatic nerve Schwann cells in mice lacking the p75 receptor. Neurosci. Lett. 272, 191–194. Fischbach, G. D., and Rosen, K. M. (1997). ARIA: A neuromuscular junction neuregulin. Annu. Rev. Neurosci. 20, 429–458. Frade, J. M., and Barde, Y. A. (1999). Genetic evidence for cell death mediated by nerve growth factor and the neurotrophin receptor p75 in the developing mouse retina and spinal cord. Development 126, 683–690. Franklin, R. J., and Barnett, S. C. (2000). Olfactory ensheathing cells and CNS regeneration: The sweet smell of success? Neuron 28, 15–18. French, S. W, Hoyer, K. K., Shen, R. R., and Teitell, M. A. (2002). TransdiVerentiation and nuclear reprogramming in hematopoietic development and neoplasia. Immunol. Rev. 187, 22–39. Fromm, L., and Burden, S. J. (2001). Neuregulin-1-stimulated phosphorylation of GABP in skeletal muscle cells. Biochemistry 40, 5306–5312. Furukawa, T., Mukherjee, S., Bao, Z. Z., Morrow, E., and Cepko, C. L. (2000). rax, Hes1, and notch1 promote the formation of Muller glia by postnatal retinal progenitor cells. Neuron 26, 383–394. Gabella, G. (1981). Ultrastructure of the nerve plexuses of the mammalian intestine: The enteric glial cells. Neuroscience 6, 425–436. Gaiano, N., Nye, J. S., and Fishell, G. (2000). Radial glial identity is promoted by Notch1 signaling in the murine forebrain. Neuron 26, 395–404. Garbern, J. Y., Cambi, F., Tang, X. M., Sima, A. A., Vallat, J. M., Bosch, E. P., Lewis, R., Shy, M., Sohi, J., Kraft, G., Chen, K. L., Joshi, I., Leonard, D. G., Johnson, W., Raskind, W., Dlouhy, S. R., Pratt, V., Hodes, M. E., Bird, T., and Kamholz, J. (1997). Proteolipid protein is necessary in peripheral as well as central myelin. Neuron 19, 205–218. Garratt, A. N., Britsch, S., and Birchmeier, C. (2000a). Neuregulin, a factor with many functions in the life of a Schwann cell. BioEssays 22, 987–996. Garratt, A. N., Voiculescu, O., Topilko, P., Charnay, P., and Birchmeier, C. (2000b). A dual role of erbB2 in myelination and in expansion of the schwann cell precursor pool. J. Cell Biol. 148, 1035–1046. Gassman, M., and Lemke, G. (1997). Neuregulins and neuregulin receptors in neural development. Curr. Opin. Neurobiol. 7, 87–92. Gershon, M. D. (1998). V. Genes, lineages, and tissue interactions in the development of the enteric nervous system. Am. J. Physiol. 275, G869–873. Ghazvini, M., Mandemakers, W., Jaegle, M., Piirsoo, M., Driegen, S., Koutsourakis, M., Smit, X., Grosveld, F., and Meijer, D. (2002). A cell type-speciWc allele of the POU gene Oct-6 reveals Schwann cell autonomous function in nerve development and regeneration. EMBO J. 21, 4612–4620. Ghislain, J., Desmarquet-Trin-Dinh, C., Jaegle, M., Meijer, D., Charnay, P., and Frain, M. (2002). Characterisation of cis-acting sequences reveals a biphasic, axon-dependent regulation of Krox20 during Schwann cell development. Development 129, 155–166. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J. (1994). Periaxin, a novel protein of myelinating Schwann cells with a possible role in axonal ensheathment. Neuron 12, 497–508.

II. GLIAL CELL DEVELOPMENT

361

362

13. SCHWANN CELL DEVELOPMENT

Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., Cottrell, D. F., Tait, S., Garry, E. M., Wallace, V. C., Ure, J., GriYths, I. R., Smith, A., and Brophy, P. J. (2000). Peripheral demyelination and neuropathic pain behavior in periaxin-deWcient mice. Neuron 26, 523–531. Grinspan, J. B., Marchionni, M. A., Reeves, M., Coulaloglou, M., and Scherer, S. S. (1996). Axonal interactions regulate Schwann cell apoptosis in developing peripheral nerve: Neuregulin receptors and the role of neuregulins. J. Neurosci. 16, 6107–6118. Guenard, V., Gwynn, L. A., and Wood, P. M. (1995a). Transforming growth factor-beta blocks myelination but not ensheathment of axons by Schwann cells in vitro. J. Neurosci. 15, 419–428. Guenard, V., Rosenbaum, T., Gwynn, L. A., Doetschman, T., Ratner, N., and Wood, P. M. (1995b). EVect of transforming growth factor-beta 1 and -beta 2 on Schwann cell proliferation on neurites. Glia 13, 309–318. Guennoun, R., Benmessahel, Y., Delespierre, B., Gouezou, M., Rajkowski, K. M., Baulieu, E. E., and Schumacher, M. (2001). Progesterone stimulates Krox-20 gene expression in Schwann cells. Brain Res. Mol. Brain Res. 90, 75–82. Haack, H., and Hynes, R. O. (2001). Integrin receptors are required for cell survival and proliferation during development of the peripheral glial lineage. Dev. Biol. 233, 38–55. Hadjiargyrou, M., Kaprielian, Z., Kato, N., and Patterson, P. H. (1996). Association of the tetraspan protein CD9 with integrins on the surface of S-16 Schwann cells. J. Neurochem. 67, 2505–2513. Hagedorn, L., Suter, U., and Sommer, L. (1999). P0 and PMP22 mark a multipotent neural crest-derived cell type that displays community eVects in response to TGF-beta family factors. Development 126, 3781–3794. Hall, D. E., Reichardt, L. F., Crowley, E., Holley, B., Moezzi, H., Sonnenberg, A., and Damsky, C. H. (1990). The 1 1 and 6 1 integrin heterodimers mediate cell attachment to distinct sites on laminin. J. Cell Biol. 110, 2175–2184. Haney, C. A., Sahenk, Z., Li, C., Lemmon, V. P., Roder, J., and Trapp, B. D. (1999). Heterophilic binding of L1 on unmyelinated sensory axons mediates Schwann cell adhesion and is required for axonal survival. J. Cell Biol. 146, 1173–1184. Harari, D., Tzahar, E., Romano, J., Shelly, M., Pierce, J. H., Andrews, G. C., and Yarden, Y. (1999). Neuregulin4: A novel growth factor that acts through the ErbB-4 receptor tyrosine kinase. Oncogene 18, 2681–2689. Harris, B. (2001). Zinc-Wnger transcription factors in the Schwann cell lineage. Ph. D. thesis, University of London. Henion, P. D., and Weston, J. A. (1997). Timing and pattern of cell fate restrictions in the neural crest lineage. Development 124, 4351–4359. Hirata, H., Hibasami, H., Yoshida, T., Ogawa, M., Matsumoto, M., Morita, A., and Uchida, A. (2001). Nerve growth factor signaling of p75 induces diVerentiation and ceramide-mediated apoptosis in Schwann cells cultured from degenerating nerves. Glia 36, 245–258. Hogervorst, F., Admiraal, L. G., Niessen, C., Kuikman, I., Janssen, H., Daams, H., and Sonnenberg, A. (1993). Biochemical characterization and tissue distribution of the A and B variants of the integrin alpha 6 subunit. J. Cell Biol. 121, 179–191. Hoke, A., Ho, T., Crawford, T. O., LeBel, C., Hilt, D., and Griffin, J. W. (2003) Glial cell line-derived neurotrophic factor alters axon schwann cell units and promotes myelination in unmyelinated nerve fibers. J. Neurosci. 23, 561–567. Honig, M. G., Camilli, S. J., and Xue, Q. S. (2002). EVects of L1 blockade on sensory axon outgrowth and pathWnding in the chick hindlimb. Dev. Biol. 243, 137–154. Howe, D. G., and McCarthy, K. D. (2000). Retroviral inhibition of cAMP-dependent protein kinase inhibits myelination but not Schwann cell mitosis stimulated by interaction with neurons. J. Neurosci. 20, 3513–3521. Itoh, K., Ozaki, M., Stevens, B., and Fields, R. D. (1997). Activity-dependent regulation of N-cadherin in DRG neurons: DiVerential regulation of N-cadherin, NCAM, and L1 by distinct patterns of action potentials. J. Neurobiol. 33, 735–748. Itoh, K., Stevens, B., Schachner, M., and Fields, R. D. (1995). Regulated expression of the neural cell adhesion molecule L1 by speciWc patterns of neural impulses. Science 270, 1369–1372. Jaakkola, S., Savunen, O., Halme, T., Uitto, J., and Peltonen, J. (1993). Basement membranes during development of human nerve: Schwann cells and perineurial cells display marked changes in their expression proWles for laminin subunits and beta 1 and beta 4 integrins. J. Neurocytol. 22, 215–230. Jaegle, M., Ghazvini, M., Mandemakers, W., Piirsoo, M., Driegen, S., Levavasseur, F., Raghoenath, S., Grosveld, F., and Meijer, D. (2003). The POU proteins Brn-2 and Oct-6 share important functions in Schwann cell development Genes Dev. 17: 1380–1391. Jaegle, M., Mandemakers, W., Broos, L., Zwart, R., Karis, A., Visser, P., Grosveld, F., and Meijer, D. (1996). The POU factor Oct-6 and Schwann cell diVerentiation. Science 273, 507–510. Jessen, K. R., Brennan, A., Morgan, L., Mirsky, R., Kent, A., Hashimoto, Y., and Gavrilovic, J. (1994). The Schwann cell precursor and its fate: A study of cell death and diVerentiation during gliogenesis in rat embryonic nerves. Neuron 12, 509–527. Jessen, K. R., and Mirsky, R. (1983). Astrocyte-like glia in the peripheral nervous system: An immunohistochemical study of enteric glia. J. Neurosci. 3, 2206–2218. Jessen, K. R., and Mirsky, R. (1999). Schwann cells and their precursors emerge as major regulators of nerve development. Trends Neurosci. 22, 402–410. Kamholz, J., Sessa, M., Scherer, S., Vogelbacker, H., Mokuno, K., Baron, P., Wrabetz, L., Shy, M., and Pleasure, D. (1992). Structure and expression of proteolipid protein in the peripheral nervous system. J. Neurosci. Res. 31, 231–244.

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

Kaprielian, Z., Cho, K. O., Hadjiargyrou, M., and Patterson, P. H. (1995). CD9, a major platelet cell surface glycoprotein, is a ROCA antigen and is expressed in the nervous system. J. Neurosci. 15, 562–573. Kendall, S., Goldhawk, D., Kubu, C., Barker, P., and Verdi, J. (2002). Expression analysis of a novel p75(NTR) signaling protein, which regulates cell cycle progression and apoptosis. Mech. Dev. 117, 187. Khursigara, G., Bertin, J., Yano, H., MoVett, H., DiStefano, P. S., and Chao, M. V. (2001). A prosurvival function for the p75 receptor death domain mediated via the caspase recruitment domain receptor-interacting protein 2. J. Neurosci. 21, 5854–5863. Khursigara, G., Orlinick, J. R., and Chao, M. V. (1999). Association of the p75 neurotrophin receptor with TRAF6. J. Biol. Chem. 274, 2597–2600. Kim, H. A., DeClue, J. E., and Ratner, N. (1997). cAMP-dependent protein kinase A is required for Schwann cell growth: Interactions between the cAMP and neuregulin/tyrosine kinase pathways. J. Neurosci. Res. 49, 236–247. Kim, H. A., Pomeroy, S. L., Whoriskey, W., Pawlitzky, I., Benowitz, L. I., Sicinski, P., Stiles, C. D., and Roberts, T. M. (2000). A developmentally regulated switch directs regenerative growth of Schwann cells through cyclin D1. Neuron 26, 405–416. Kim, H. A., Ratner, N., Roberts, T. M., and Stiles, C. D. (2001). Schwann cell proliferative responses to cAMP and Nf1 are mediated by cyclin D1. J. Neurosci. 21, 1110–1116. Koenig, H. L., Schumacher, M., Ferzaz, B., Thi, A. N., Ressouches, A., Guennoun, R., Jung-Testas, I., Robel, P., Akwa, Y., and Baulieu, E. E. (1995). Progesterone synthesis and myelin formation by Schwann cells. Science 268, 1500–1503. Kondo, T., and RaV, M. (2000a). Oligodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells. Science 289, 1754–1757. Kondo, T., and RaV, M. (2000b). The Id4 HLH protein and the timing of oligodendrocyte diVerentiation. EMBO J. 19, 1998–2007. Krasnoselsky, A., Massay, M. J., DeFrances, M. C., Michalopoulos, G., Zarnegar, R., and Ratner N. (1994). Hepatocyte growth factor is a mitogen for Schwann cells and is present in neuroWbromas. J. Neurosci. 14, 7284–7290. Kubu, C. J., Orimoto, K., Morrison, S. J., Weinmaster, G., Anderson, D. J., and Verdi, J. M. (2002). Developmental changes in Notch1 and numb expression mediated by local cell-cell interactions underlie progressively increasing delta sensitivity in neural crest stem cells. Dev. Biol. 244, 199–214. Kuhlbrodt, K., Herbarth, B., Sock, E., Enderich, J., Hermans-Borgmeyer, I., and Wegner, M. (1998a). Cooperative function of POU proteins and SOX proteins in glial cells. J. Biol. Chem. 273, 16050–16057. Kuhlbrodt, K., Herbarth, B., Sock, E., Hermans-Borgmeyer, I., and Wegner, M. (1998b). Sox10, a novel transcriptional modulator in glial cells. J. Neurosci. 18, 237–250. Kurtz, A., Zimmer, A., Schnutgen, F., Bruning, G., Spener, F., and Muller, T. (1994). The expression pattern of a novel gene encoding brain-fatty acid binding protein correlates with neuronal and glial cell development. Development 120, 2637–2649. Kury, P., Greiner-Petter, R., Cornely, C., Jurgens, T., and Muller, H. W. (2002). Mammalian achaete scute homolog 2 is expressed in the adult sciatic nerve and regulates the expression of Krox24, Mob-1, CXCR4, and p57kip2 in Schwann cells. J. Neurosci. 22, 7586–7595. Landmesser, L., Dahm, L., Schultz, K., and Rutishauser, U. (1988). Distinct roles for adhesion molecules during innervation of embryonic chick muscle. Dev. Biol. 130, 645–670. Landon, D. N, and Wiseman, O. J. (2001). A Pacinian corpuscle in the human bladder lamina propria. J. Neurocytol. 30, 457–464. Le, X. F., Varela, C. R., and Bast, Jr., R. C. (2002). Heregulin-induced apoptosis. Apoptosis 7, 483–491. Le Douarin, N. M., and Kalcheim, C. (1999). ‘‘The Neural Crest.’’ Cambridge University Press, Cambridge. Lee, J. E., Hollenberg, S. M., Snider, L., Turner, D. L., Lipnick, N., and Weintraub, H. (1995). Conversion of Xenopus ectoderm into neurons by NeuroD, a basic helix-loop-helix protein. Science 268, 836–844. Lee, M.-J., Brennan, A., Blanchard, A., Zoidl, G., Dong, Z., Tabernero, A., Zoidl, C., Dent, M. A. R., Jessen, K. R., and Mirsky, R. (1997). P0 is constitutively expressed in the rat neural crest and embryonic nerves and is negatively and positively regulated by axons to generate nonmyelin-forming and myelin-forming Schwann cells, respectively. Mol. Cell. Neurosci. 8, 336–350. Lee, M.-J., Calle, E., Brennan, A., Ahmed, S., Sviderskaya, E., Jessen, K. R., and Mirsky, R. (2001). In early development of the rat mRNA for the major myelin protein P0 is expressed in nonsensory areas of the embryonic inner ear, notochord, enteric nervous system, and olfactory ensheathing cells. Dev. Dyn. 222, 40–51. Leimeroth, R., Lobsiger, C., Lu¨ssi, A., Taylor, V., Suter, U., and Sommer, L. (2002). Membrane-bound neuregulin1 type III actively promotes Schwann cell diVerentiation of multipotent Progenitor cells. Dev. Biol. 246, 245–258. Lemke, G. (1996). Neuregulins in development. Mol. Cell. Neurosci. 7, 247–262. Lemke, G. (2001). Glial control of neuronal development. Annu. Rev. Neurosci. 24, 87–105. Lemke, G., and Chao, M. (1988). Axons regulate Schwann cell expression of the major myelin and NGF receptor genes. Development 102, 499–504. Lemke, G., Kuhn, R., Monuki, E. S., and Weinmaster, G. (1991). Expression and activity of the transcription factor SCIP during glial diVerentiation and myelination. Ann. N. Y. Acad. Sci. 633, 189–195. Lentz, S. I., Miner, J. H., Sanes, J. R, and Snider, W. D. (1997). Distribution of the ten known laminin chains in the pathways and targets of developing sensory axons. J. Comp. Neurol. 378, 547–561.

II. GLIAL CELL DEVELOPMENT

363

364

13. SCHWANN CELL DEVELOPMENT

Levi, A. D., Bunge, R. P., Lofgren, J. A., Meima, L., Hefti, F., Nikolics, K., and Sliwkowski, M. X. (1995). The inXuence of heregulins on human Schwann cell proliferation. J. Neurosci. 15, 1329–1340. Li, H., Wigley, C., and Hall, S. M. (1998). Chronically denervated rat Schwann cells respond to GGF in vitro. Glia 24, 290–303. Li, Y., Tennekoon, G. I., Birnbaum, M., Marchionni, M. A., and Rutkowski, J. L. (2001). Neuregulin signaling through a PI3K/Akt/Bad pathway in Schwann cell survival. Mol. Cell. Neurosci. 17, 761–767. Lin, D. M., and Goodman, C. S. (1994). Ectopic and increased expression of Fasciclin II alters motoneuron growth cone guidance. Neuron 13, 507–523. Livesey, F. J., O’Brien, J. A., Li, M., Smith, A. G., Murphy, L. J., and Hunt, S. P. (1997). A Schwann cell mitogen accompanying regeneration of motor neurons. Nature 390, 614–618. Lloyd, A. C., Obermuller, F., Staddon, S., Barth, C. F., McMahon, M., and Land, H. (1997). Cooperating oncogenes converge to regulate cyclin/cdk complexes. Genes Dev. 11, 663–677. Lo, L., Tiveron, M. C., and Anderson, D. J. (1998). MASH1 activates expression of the paired homeodomain transcription factor Phox2a, and couples pan-neuronal and subtype-speciWc components of autonomic neuronal identity. Development 125, 609–620. Lobsiger, C. S., Taylor, V., and Suter, U. (2002). The early life of a Schwann cell. Biol. Chem. 383, 245–253. Loeb, J. A., Khurana, T. S., Robbins, J. T., Yee, A. G., and Fischbach, G. D. (1999). Expression patterns of transmembrane and released forms of neuregulin during spinal cord and neuromuscular synapse development. Development 126, 781–791. Lubischer, J. L, and Thompson, W. J. (1999). Neonatal partial denervation results in nodal but not terminal sprouting and a decrease in eYcacy of remaining neuromuscular junctions in rat soleus muscle. J. Neurosci. 19, 8931–8944. Lyons, S. A., Morell, P., and McCarthy, K. D. (1994). Schwann cells exhibit P2Y purinergic receptors that regulate intracellular calcium and are up-regulated by cyclic AMP analogues. J. Neurochem. 63, 552–560. Lyons, S. A., Morell, P., and McCarthy, K. D. (1995). Schwann cell ATP-mediated calcium increases in vitro and in situ are dependent on contact with neurons. Glia 13, 27–38. Magnaghi, V., Cavarretta, I., Galbiati, M., Martini, L., and Melcangi, R. C. (2001). Neuroactive steroids and peripheral myelin proteins. Brain Res. Brain Res. Rev. 37, 360–371. Mandemakers, W., Zwart, R., Jaegle, M., Walbeehm, E., Visser, P., Grosveld, F., and Meijer, D. (2000). A distal Schwann cell-speciWc enhancer mediates axonal regulation of the Oct-6 transcription factor during peripheral nerve development and regeneration. EMBO J. 19, 2992–3003. Marchionni, M. A., Goodearl, A. D. J., Chen, M. S., Bermingham-McDonogh, O., Kirk, C., Hendricks, M., Danehy, F., Misumi, D., Sudhalter, J., Kobayashi, K. Wroblewski, D., Lynch, C., Baldassare, M., Hiles, I., Davis, J. B., Hsuan, J. J., Totty, N. F., Otsu, M., McButney, R. N., WaterWeld, M. D., Stroobant, P., and Gwynne, D. (1993). Glial growth factors are alternatively spliced erbB2 ligands expressed in the nervous system. Nature 362, 312–318. Masaki, T., Matsumura, K., Hirata, A., Yamada, H., Hase, A., Arai, K., Shimizu, T., Yorifuji, H., Motoyoshi, K., and Kamakura, K. (2002). Expression of dystroglycan and the laminin-alpha 2 chain in the rat peripheral nerve during development. Exp. Neurol. 174, 109–117. Mathon, N. F., Malcolm, D. S., Harrisingh, M. C., Cheng, L., and Lloyd, A. C. (2001). Lack of replicative senescence in normal rodent glia. Science 291, 872–875. Matsumura, K., Chiba, A., Yamada, H., Fukuta-Ohi, H., Fujita, S., Endo, T., Kobata, A., Anderson, L. V., Kanazawa, I., Campbell, K. P., and Shimizu, T. (1997b). A role of dystroglycan in Schwannoma cell adhesion to laminin. J. Biol. Chem. 272, 13904–13910. Matsumura, K., Yamada, H., Saito, F., Sunada, Y., and Shimizu, T. (1997a). Peripheral nerve involvement in merosin-deWcient congenital muscular dystrophy and dy mouse. Neuromuscul. Disord. 7, 7–12. Maurel, P., and Salzer, J. L. (2000). Axonal regulation of Schwann cell proliferation and survival and the initial events of myelination requires PI 3-kinase activity. J. Neurosci. 20, 4635–4645. Mayer, C., Wachtler, J., Kamleiter, M., and Grafe, P. (1997). Intracellular calcium transients mediated by P2 receptors in the paranodal Schwann cell region of myelinated rat spinal root axons. Neurosci. Lett. 224, 49–52. Meier, C., Parmantier, E., Brennan, A., Mirsky, R., and Jessen, K. R. (1999). Developing Schwann cells acquire the ability to survive without axons by establishing an autocrine circuit involving IGF, NT-3 and PDGF-BB. J. Neurosci. 19, 3847–3859. Meintanis, S., Thomaidou, D., Jessen, K. R., Mirsky, R., and Matsas, R. (2001). The neuron-glia signal betaneuregulin promotes Schwann cell motility via the MAPK pathway. Glia 34, 39–51. Memberg, S. P., and Hall, A. K. (1995). Dividing neuron precursors express neuron-speciWc tubulin. J. Neurobiol. 27, 26–43. Mews, M., and Meyer, M. (1993). Modulation of Schwann cell phenotype by TGF-beta 1: Inhibition of P0 mRNA expression and downregulation of the low aYnity NGF receptor. Glia 8, 208–217. Meyer, D., and Birchmeier, C. (1995). Multiple essential functions of neuregulin in development. Nature 378, 386–390. Meyer, D., Yamaii, T., Garratt, A., Riethmacher-Sonnenberg, E., Kane, D., Theill, L. E., and Birchmeier, C. (1997). Isoform-speciWc expression and function of neuregulin. Development 124, 3575–3586. Milner, R., Wilby, M., Nishimura, S., Boylen, K., Edwards, G., Fawcett, J., Streuli, C., Pytela, R., and VrenchConstant, C. (1997). Division of labor of Schwann cell integrins during migration on peripheral nerve extracellular matrix ligands. Dev. Biol. 185, 215–228.

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

Mirsky, R., Dubois, C., Morgan, L., and Jessen, K. R. (1990). 04 and A007-sulfatide antibodies bind to embryonic Schwann cells prior to the appearance of galactocerebroside; regulation of the antigen by axon-Schwann cell signals and cyclic AMP. Development 109, 105–116. Mirsky, R., and Jessen, K. R. (1996). Schwann cell development, diVerentiation and myelination. Curr. Opin. Neurobiol. 6, 89–96. Mirsky, R., and Jessen, K. R. (1999). The neurobiology of Schwann cells. Brain Pathol. 9, 293–311. Mirsky, R., and Jessen, K. R. (2001). Embryonic and early postnatal development of Schwann cells. In ‘‘Glial Cell Development: Basic Principles and Clinical Relevance’’ (K. R. Jessen and W. D. Richardson, eds.), 2nd ed., pp. 1–20. Oxford University Press, Oxford. Monuki, E. S., Kuhn, R., Weinmaster, G., Trapp, B. D., and Lemke, G. (1990). Expression and activity of the POU transcription factor SCIP. Science 249, 1300–1303. Monuki, E. S., Weinmaster, G., Kuhn, R., and Lemke, G. (1989). SCIP: A glial POU domain gene regulated by cyclic AMP. Neuron 3, 783–793. Morgan, L., Jessen, K. R., and Mirsky, R. (1991). The eVects of cAMP on diVerentiation of cultured Schwann cells: Progression from an early phenotype (04þ) to a myelin phenotype (P0þ, GFAP, N-CAM, NGFReceptor) depends on growth inhibition. J. Cell Biol. 112, 457–467. Morgan, L., Jessen, K. R., and Mirsky, R. (1994). Negative regulation of the P0 gene in Schwann cells: Suppression of P0 MRNA and protein induction in cultured Schwann cells by FGF2 and TGFa1, TGFa2 and TGFa3. Development 120, 1399–1409. Morris, J. K., Lin, W., Hauser, C., Marchuk, Y., Getman, D., and Lee, K.-F. (1999). Rescue of the cardiac defect in ErbB2 mutant mice reveals essential roles of ErbB2 in peripheral nervous system development. Neuron 23, 273–283. Morrison, S. J., Perez, S. E., Qiao, Z., Verdi, J. M., Hicks, C., Weinmaster, G., and Anderson, D. J. (2000). Transient Notch activation initiates an irreversible switch from neurogenesis to gliogenesis by neural crest stem cells. Cell 101, 499–510. Morrison, S. J., White, P. M., Zock, C., and Anderson, D. J. (1999). Prospective identiWcation, isolation by Xow cytometry, and in vivo self-renewal of multipotent mammalian neural crest stem cells. Cell 96, 737–749. Morrissey, T. K., Levi, A. D., Nuijens, A., Sliwkowski, M. X., and Bunge, R. P. (1995). Axon-induced mitogenesis of human Schwann cells involves heregulin and p185erbB2. Proc. Natl. Acad. Sci. U. S. A. 92, 1431–1435. Mukai, J., Hachiya, T., Shoji-Hoshino, S., Kimura, M. T., Nadano, D., Suvanto, P., Hanaoka, T., Li, Y., Irie, S., Greene, L. A., and Sato, T. A. (2000). NADE, a p75NTR-associated cell death executor, is involved in signal transduction mediated by the common neurotrophin receptor p75NTR. J. Biol. Chem. 275, 17566–17570. Mukai, J., Shoji, S., Kimura, M. T., Okubo, S., Sano, H., Suvanto, P., Li, Y., Irie, S., and Sato, T. A. (2002). Structure-function analysis of NADE: IdentiWcation of regions that mediate nerve growth factor-induced apoptosis. J. Biol. Chem. 277, 13973–13982. Murphy, P., Topilko, P., Schneider-Manoury, S., Seitanidou, T., Baron Van Evercooren, A., and Charnay, P. (1996). The regulation of Krox-20 expression reveals important steps in the control of peripheral glial cell development. Development 122, 2847–2857. Nagano, S., Takeda, M., Ma, L., and Soliven, B. (2001). Cytokine-induced cell death in immortalized Schwann cells: Roles of nitric oxide and cyclic AMP. J. Neurochem. 77, 1486–1495. Nagarajan, R., Svaren, J., Le, N., Araki, T., Watson, M., and Milbrandt, J. (2001). EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30, 355–368. Nakagawa, M., Miyagoe-Suzuki, Y., Ikezoe, K., Miyata, Y., Nonaka, I., Harii, K., and Takeda, S. (2001). Schwann cell myelination occurred without basal lamina formation in laminin alpha2 chain-null mutant (dy3K/dy3K) mice. Glia 35, 101–110. Nave, K.-A. (2001). Myelin-speciWc genes and their mutations in the mouse. In ‘‘Glial Cell Development: Basic Principles and Clinical Relevance’’ (K. R. Jessen and W. D. Richardson, eds.), 2nd. ed., pp. 177–208. Oxford University Press, Oxford. Neve, R. M., Holbro, T., and Hynes, N. E. (2002). Distinct roles for phosphoinositide 3-kinase, mitogen-activated protein kinase and p38 MAPK in mediating cell cycle progression of breast cancer cells. Oncogene 21, 4567–4576. Ng, V., Zanazzi, G., Timpl, R., Talts, J. F., Salzer, J. L., Brennan, P. J., and Rambukkana, A. (2000). Role of the cell wall phenolic glycolipid-1 in the peripheral nerve predilection of Mycobacterium leprae. Cell 103, 511–524. Nichols, D. H., and Weston, J. A. (1977). Melanogenesis in cultures of peripheral nervous tissue. I. The origin and prospective fate of cells giving rise to melanocytes. Dev. Biol. 60, 217–225. Nickols, J. C., Valentine, W., Kanwal, S., and Carter, B. D. (2003). Activation of the transcription factor NFkB in Schwann cells is required for peripheral myelin formation. Nat. Neurosci. 6, 161–167. Niemann, C., Brinkmann, V., and Birchmeier, W. (2000). Hepatocyte growth factor and neuregulin in mammary gland cell morphogenesis. Adv. Exp. Med. Biol. 480, 9–18. Niessen, C. M., Cremona, O., Daams, H., Ferraresi, S., Sonnenberg, A., and Marchisio, P. C. (1994). Expression of the integrin a6ß4 in peripheral nerves: Localization in Schwann and perineural cells and diVerent variants of the ß4 subunit. J. Cell Sci. 107, 543–552. Notterpek, L., Snipes, G. J., and Shooter, E. M. (1999). Temporal expression pattern of peripheral myelin protein 22 during in vivo and in vitro myelination. Glia 25, 358–369. Orr-Urtreger, A., Trakhtenbrot, L., Ben-Levy, R., Wen, D., Rechavi, G., Lonai, P., and Yarden, Y. (1993). Neural expression and chromosomal mapping of Neu diVerentiation factor to 8p12-p21. Proc. Natl. Acad. Sci. U. S A. 90, 1867–1871.

II. GLIAL CELL DEVELOPMENT

365

366

13. SCHWANN CELL DEVELOPMENT

Pannese, E. (1981). The satellite cells of the sensory ganglia. Adv. Anat. Embryol. Cell Biol. 65, 1–111. Pannese, E. (1994). ‘‘Neurocytology. Fine Structure of Neurons, Nerve Processes, and Neuroglial Cells.’’ Thieme, Stuttgart, New York. Paratore, C., Goerich, D. E., Suter, U., Wegner, M., and Sommer, L. (2001). Survival and glial fate acquisition of neural crest cells are regulated by an interplay between the transcription factor Sox10 and extrinsic combinatorial signaling. Development 128, 3949–3961. Pareyson, D., Taroni, F., Botti, S., Morbin, M., Baratta, S., Lauria, G., Ciano, C., and Sghirlanzoni, A. (2000). Cranial nerve involvement in CMT disease type 1 due to early growth response 2 gene mutation. Neurology 54, 1696–1698. Parkinson, D. B., Bhaskaran, A., Droggiti, A., Dickinson, S., Mirsky, R., and Jessen, K. R. (2003a). Inhibition of the Jun N-terminal kinase/c-Jun pathway mediates the eVects of Krox-20 on the proliferation and survival of Schwann cells. J. Cell Biol., under revision. Parkinson, D. B., Dickinson, S., Bhaskaran, A., Kinsella, M. T., Brophy, P. J., Sherman, D. L., Sharghi-Namini, S., Duran Alonso, M. B., Jessen, K. R., and Mirsky, R. (2002a). Krox 20 activates a set of complex changes in Schwann cells that characterize myelination. Glia Supplement 1, S69. Parkinson, D. B., Dickinson, S., Bhaskaran, A., Kinsella, M. T., Brophy, P. J., Sherman, D. L., Sharghi-Namini, S., Duran Alonso, M. B., Mirsky, R., and Jessen, K. R. (2003b). Regulation of the myelin gene periaxin provides evidence for Krox-20 independent myelin-related signalling in Schwann cells. Mol. Cell. Neurosci. 23, 13–27. Parkinson, D. B., Dong, Z., Bunting, H., WhitWeld, J., Meier, C., Marie, H., Mirsky, R., and Jessen, K. R. (2001). Transforming growth factor ß (TGFß) mediates Schwann cell death in vitro and in vivo: Examination of c-Jun activation, interactions with survival signals, and the relationship of TGFß mediated death to Schwann cell diVerentiation. J. Neurosci. 21, 8572–8585. Parkinson, D. B., Langner, K., Namini, S. S., Jessen, K. R., and Mirsky, R. (2002b). ß-neuregulin and autocrine mediated survival of Schwann cells requires activity of Ets family transcription factors. Mol. Cell. Neurosci. 20, 154–167. Parmantier, E., Braun, C., Thomas, J. L., Peyron, F., Martinez, S., and Zalc, B. (1997). PMP-22 expression in the central nervous system of the embryonic mouse deWnes potential transverse segments and longitudinal columns. J. Comp. Neurol. 378, 159–172. Parmantier, E., Cabon, F., Braun, C., D’Urso, D., Muller, H. W., and Zalc, B. (1995). Peripheral myelin protein22 is expressed in rat and mouse brain and spinal cord motoneurons. Eur. J. Neurosci. 7, 1080–1088. Parmantier, E., Lynn, B., Lawson, D., Turmaine, M., Sharghi Namini, S., Chakrabarti, L., McMahon, A. P., Jessen, K. R., and Mirsky, R. (1999). Schwann cell-derived Desert Hedgehog controls the development of peripheral nerve sheaths. Neuron 23, 713–724. Peirano, R. I., Goerich, D. E., Riethmacher, D., and Wegner, M. (2000). Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol. Cell. Biol. 20, 3198–3209. Plantinga, L. C., Verhaagen, J., Edwards, P. M., Hol, E. M., Bar, P. R., and Gispen, W. H. (1993). The expression of B-50/GAP-43 in Schwann cells is upregulated in degenerating peripheral nerve stumps following nerve injury. Brain Res. 602, 69–76. Previtali, S. C., Dina, G., Nodari, A., Fasolini, M., Wrabetz, L., Mayer, U., Feltri, M. L., and Quattrini, A. (2003). Schwann cells synthesize a7b1 integrin which is dispensible for peripheral nerve development and myelination. Mol. Cell. Neurosci. 23, 210–218. Previtali, S. C., Feltri, M. L., Archelos, J. J., Quattrini, A., Wrabetz, L., and Hartung, H. (2001). Role of integrins in the peripheral nervous system. Prog. Neurobiol. 64, 35–49. Previtali, S., Nodari, A., Dina, G., Messing, A., Quattrini, A., Wrabetz, L., and Feltri, M. L. (2002). Cell-type speciWc knockouts in the study of glia-basement membrane interactions. Glia supplement I, S7. Pusch, C., Hustert, E., Pfeifer, D., Sudbeck, P., Kist, R., Roe, B., Wang, Z., Balling, R., Blin, N., and Scherer, G. (1998). The SOX10/Sox10 gene from human and mouse: Sequence, expression, and transactivation by the encoded HMG domain transcription factor. Hum. Genet. 103, 115–123. Rahmatullah, M., Schroering, A., Rothblum, K., Stahl, R. C., Urban, B., and Carey, D. J. (1998). Synergistic regulation of Schwann cell proliferation by heregulin and forskolin. Mol. Cell. Biol. 18, 6245–6252. Raisman, G. (2001). Olfactory ensheathing cells—another miracle cure for spinal cord injury? Nature Rev. Neurosci. 2, 369–375. Rambukkana, A., Salzer, J. L., Yurchenco, P. D., and Tuomanen, E. I. (1997). Neural targeting of Mycobacterium leprae mediated by the G domain of the laminin-alpha2 chain. Cell 88, 811–821. Rambukkana, A., Yamada, H., Zanazzi, G., Mathus, T., Salzer, J. L., Yurchenco, P. D., Campbell, K. P., and Fischetti, V. A. (1998). Role of alpha-dystroglycan as a Schwann cell receptor for Mycobacterium leprae. Science 282, 2076–2079. Rambukkana, A., Zanazzi, G., Tapinos, N., and Salzer, J. L. (2002). Contact-dependent demyelination by Mycobacterium leprae in the absence of immune cells. Science 296, 927–931. Ramon-Cueto, A., and Santos-Benito, F. F. (2001). Cell therapy to repair injured spinal cords: Olfactory ensheathing glia transplantation. Restor. Neurol. Neurosci. 19, 149–156. Ridley, A. J., Davis, J. B., Stroobant, P., and Land, H. (1989). Transforming growth factors-beta 1 and beta 2 are mitogens for rat Schwann cells. J. Cell Biol. 109, 3419–3424. Ridley, A. J., Paterson, H. F., Noble, M., and Land, H. (1988). Ras-mediated cell cycle arrest is altered by nuclear oncogenes to induce Schwann cell transformation. EMBO J. 7, 1635–1645.

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

Riethmacher, D., Sonnenberg-Riethmacher, E., Brinkmann, V., Yamaai, T., Lewin, G. R., and Birchmeier, C. (1997). Severe neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature 389, 725–730. Rizvi, T. A., Huang, Y., Sidani, A., Atit, R., Largaespada, D. A., Boissy, R. E., and Ratner, N. (2002). A novel cytokine pathway suppresses glial cell melanogenesis after injury to adult nerve. J. Neurosci. (in press). Robitaille, R. (1998). Modulation of synaptic eYcacy and synaptic depression by glial cells at the frog neuromuscular junction. Neuron 21, 847–855. Rosenbaum, C., Karyala, S., Marchionni, M. A., Kim, H. A., Krasnoselsky, A. L., Happel, B., Isaacs, I., Brackenbury, R., and Ratner, N. (1997). Schwann cells express NDF and SMDF/n-ARIA mRNAs, secrete neuregulin, and show constitutive activation of erbB3 receptors: Evidence for a neuregulin autocrine loop. Exp. Neurol. 148, 604–615. Rouleau, G. A., Merel, P., Lutchman, M., Sanson, M., Zucman, J., Marineau, C., Hoang-Xuan, K., Demczuk, S., Desmaze, C., Plougastel, B., Pulst, S. M., Lenoir, G., Bijlsma, E., Fashold, R., Dumanski, J., de Jong, P., Parry, D., Eldridge, R., Aurias, A., Delattre, O., and Thomas, G. (1993). Alteration in a new gene encoding a putative membrane-organizing protein causes neuro-Wbromatosis type 2. Nature 363, 515–521. Russell, J. W., Cheng, H. L., and Golovoy, D. (2000). Insulin-like growth factor-I promotes myelination of peripheral sensory axons. J. Neuropathol. Exp. Neurol. 59, 575–584. Sabourin, L. A., and Rudnicki, M. A. (2000). The molecular regulation of myogenesis. Clin. Genet. 57, 16–25. Sadoulet-Puccio, H. M., and Kunkel, L. M. (1996). Dystrophin and its isoforms. Brain Pathol. 6, 25–35. Sainio, M., Zhao, F., Heiska, L., Turunen, O., den Bakker, M., ZwarthoV, E., Lutchman, M., Rouleau, G. A., Jaaskelainen, J., Vaheri, A., and Carpen, O. (1997). NeuroWbromatosis 2 tumor suppressor protein colocalizes with ezrin and CD44 and associates with actin-containing cytoskeleton. J. Cell Sci. 110, 2249–2260. Saito, F., Moore, S. A., Barresi, R., Henry, M. D., Messing, A., Ross-Barta, S. E., Cohn, R. D., Williamson, R. A., Sluka, K. A., Sherman, D. L., Brophy, P. J., Schmelzer, J. D., Low, P. A., Wrabetz, L., Feltri, M. L., and Campbell, K. P. (2003). Unique role of dystroglycan in peripheral nerve myelination, nodal structure, and sodium channel distribution. Neuron 38, 747-758. Salehi, A. H., Roux, P. P., Kubu, C. J., Zeindler, C., Bhakar, A., Tannis, L. L., Verdi, J. M., and Barker, P. A. (2000). NRAGE, a novel MAGE protein, interacts with the p75 neurotrophin receptor and facilitates nerve growth factor-dependent apoptosis. Neuron 27, 279–288. Saunders, R. D., and DeVries, G. H. (1988). Schwann cell proliferation is accompanied by enhanced inositol phospholipid metabolism. J. Neurochem. 50, 876–882. Scherer, S. S. (1997). The biology and pathobiology of Schwann cells. Curr. Opin. Neurol. 10, 386–397. Scherer, S. S., and Gutmann, D. H. (1996). Expression of the neuroWbromatosis 2 tumor suppressor gene product, merlin, in Schwann cells. J. Neurosci. Res. 46, 595–605. Scherer, S. S., and Salzer, J. L. (2001). Axon-Schwann cell interactions during peripheral nerve degeneration and regeneration. In ‘‘Glial Cell Development: Basic Principles And Clinical Relevance’’ (K. R. Jessen and W. D. Richardson, eds.), 2nd ed., pp. 299–330. Oxford University Press, Oxford. Scherer, S. S., Wang, D.-Y., Kuhn, R., Lemke, G., Wrabetz, L., and Kamholz, J. (1994a). Axons regulate Schwann cell expression of the POU transcription factor SCIP. J. Neurosci. 14, 1930–1942. Scherer, S. S., Xu, Y. T., Bannerman, P. G., Sherman, D. L., and Brophy, P. J. (1995). Periaxin expression in myelinating Schwann cells: Modulation by axon-glial interactions and polarized localization during development. Development 121, 4265–4273. Scherer, S. S., Xu, Y. T., Roling, D., Wrabetz, L., Feltri, M. L., and Kamholz, J. (1994b). Expression of growthassociated protein-43 kD in Schwann cells is regulated by axon-Schwann cell interactions and cAMP. J. Neurosci. Res. 38, 575–589. Schneider, C., Wicht, H., Enderich, J., Wegner, M., and Rohrer, H. (1999). Bone morphogenetic proteins are required in vivo for the generation of sympathetic neurons. Neuron 24, 861–870. Schumacher, M., Guennoun, R., Mercier, G., Desarnaud, F., Lacor, P., Benavides, J., Ferzaz, B., Robert, F., and Baulieu, E. E. (2001). Progesterone synthesis and myelin formation in peripheral nerves. Brain Res. Brain Res. Rev. 37, 343–359. Scoles, D. R., Huynh, D. P., Morcos, P. A., Coulsell, E. R., Robinson, N. G. G., Tamanoi, F., and Pulst, S. M. (1998). NeuroWbromatosis 2 tumour suppressor schwannomin interacts with ß11-spectrin. Nature Gen. 18, 354–359. Seilheimer, B., Persohn, E., and Schachner, M. (1989). Antibodies to the L1 adhesion molecule inhibit Schwann cell ensheathment of neurons in vitro. J. Cell Biol. 109, 3095–3103. Seilheimer, B., and Schachner, M. (1988). Studies of adhesion molecules mediating interactions between cells of peripheral nervous system indicate a major role for L1 in mediating sensory neuron growth on Schwann cells in culture. J. Cell Biol. 107, 341–351. Sensenbrenner, M., Lucas, M., and Deloulme, J. C. (1997). Expression of two neuronal markers, growthassociated protein 43 and neuron-speciWc enolase, in rat glial cells. J. Mol. Med. 75, 653–663. Shah, N. M., Groves, A. K., and Anderson, D. J. (1996). Alternative neural crest cell fates are instructively promoted by TGFb superfamily members. Cell 85, 331–343. Shah, N. M., Marchionni, M. A., Isaacs, I., Stroobant, P., and Anderson, D. J. (1994). Glial growth factor restricts mammalian neural crest stem cells to a glial fate. Cell 77, 349–360. Sherman, L., Stocker, K. M., Morrison, R., and Ciment, G. (1993). Basic Wbroblast growth factor (bFGF) acts intracellularly to cause the transdiVerentiation of avian neural crest-derived Schwann cell precursors into melanocytes. Development 118, 1313–1326. Sherman, L. S., Rizvi, T. A., Karyala, S., and Ratner, N. (2000). CD44 enhances neuregulin signaling by Schwann cells. J. Cell Biol. 150, 1071–1084.

II. GLIAL CELL DEVELOPMENT

367

368

13. SCHWANN CELL DEVELOPMENT

Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J. (2001). SpeciWc disruption of a schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron 30, 677–687. Shibuya, Y., Mizoguchi, A., Takeichi, M., Shimada, K., and Ide, C. (1995). Localization of N-cadherin in the normal and regenerating nerve Wbers of the chicken peripheral nervous system. Neuroscience 67, 253–261. Sim, F., Zhao, C., Li, W., Lakatos, A., and Franklin, R. (2002). Expression of the POU-domain transcription factors SCIP/Oct-6 and Brn-2 is associated with Schwann cell but not oligodendrocyte remyelination of the CNS. Mol. Cell. Neurosci. 20, 669. SkoV, A. M., Lisak, R. P., Bealmear, B., and Benjamins, J. A. (1998). TNF-alpha and TGF-beta act synergistically to kill Schwann cells. J. Neurosci. Res. 53, 747–756. Sobue, G., and Pleasure, D. (1984). Schwann cell galactocerebroside induced by derivatives of adenosine 3’,5’monophosphate. Science 224, 72–74. Soilu-Hanninen, M., Ekert, P., Bucci, T., Syroid, D., Bartlett, P. F., and Kilpatrick, T. J. (1999). Nerve growth factor signaling through p75 induces apoptosis in Schwann cells via a Bcl-2-independent pathway. J. Neurosci. 19, 4828–4838. Sonnenberg-Riethmacher, E., Miehe, M., Stolt, C. C., Goerich, D. E., Wegner, M., and Riethmacher, D. (2001). Development and degeneration of dorsal root ganglia in the absence of the HMG-domain transcription factor Sox10. Mech. Dev. 109, 253–265. Southard-Smith, E. M., Kos, L., and Pavan, W. J. (1998). Sox 10 mutation disrupts neural crest development in Dom Hirschsprung mouse model. Nature Genet. 18, 60–64. Spassky, N., Goujet-Zalc, C., Parmantier, E., Olivier, C., Martinez, S., Ivanova, A., Ikenaka, K., Macklin, W., Cerruti, I., Zalc, B., and Thomas, J. L. (1998). Multiple restricted origin of oligodendrocytes. J. Neurosci. 18, 8331–8343. Spencer, P. S., and Schaumburg, H. H. (1973). An ultrastructural study of the inner core of the Pacinian corpuscle. J. Neurocytol. 2, 217–235. Stevens, B., and Fields, R. D. (2000). Response of Schwann cells to action potentials in development. Science 287, 2267–2271. Stevens, B., Tanner, S., and Fields, R. D. (1998). Control of myelination by speciWc patterns of neural impulses. J. Neurosci. 18, 9303–9311. Stewart, H. J., Curtis, R., Jessen, K. R., and Mirsky, R. (1995). TGF-beta s and cAMP regulate GAP-43 expression in Schwann cells and reveal the association of this protein with the trans-Golgi network. Eur. J. Neurosci. 7, 1761–1772. Stewart, H. J. S., Bradke, F., Tabernero, A., Morrell, D., Jessen, K. R., and Mirsky, R. (1996). Regulation of rat Schwann cell P0 expression and DNA synthesis by insulin-like growth factors in vitro. Eur. J. Neurosci. 8, 553–564. Stewart, H. J. S., Brennan, A., Rahman, M., Zoidl, G., Mitchell, P. J., Jessen, K. R., and Mirsky, R. (2001). Developmental regulation and overexpression of the transcription factor AP-2, a potential regulator of the timing of Schwann cell generation. Eur. J. Neurosci. 14, 363–372. Stewart, H. J. S., Morgan, L., Jessen, K. R., and Mirsky, R. (1993). Changes in DNA synthesis rate in the Schwann cell lineage in vivo are correlated with the precursor-Schwann cell transition and myelination. Eur. J. Neurosci. 5, 1136–1144. Stewart, H. J. S., Zoidl, G., Rossner, M., Brennan, A., Zoidl, C., Nave, K.-A., Mirsky, R., and Jessen, K. R. (1997). Helix-loop-helix proteins in Schwann cells: A study of regulation and subcellular localization of Ids, REB and E12/47 during embryonic and postnatal development. J. Neurosci. Res. 50, 684–701. Stiles, C. D., Pledger, W. J., Tucker, R. W., Martin, R. G., and Scher, C. D. (1980). Regulation of the Balb/c-3T3 cell cycle-eVects of growth factors. J. Supramol. Struct. 13, 489–499. Stocker, K. M., Sherman, L., Rees, S., and Ciment, G. (1991). Basic FGF and TGF-a1 inXuence commitment to melanogenesis in neural crest-derived cells of avian embryos. Development 111, 635–645. Stolt, C. C., Rehberg, S., Ader, M., Lommes, P., Riethmacher, D., Schachner, M., Bartsch, U., and Wegner M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev. 16, 165–170. Sudhalter, J., Whitehouse, L., Rusche, J. R., Marchionni, M. A., and Mahanthappa, N. K. (1996). Schwann cell heparan sulfate proteoglycans play a critical role in glial growth factor/neuregulin signaling. Glia 17, 28–38. Sulaiman, O. A., and Gordon, T. (2002). Transforming growth factor-beta and forskolin attenuate the adverse eVects of long-term Schwann cell denervation on peripheral nerve regeneration in vivo. Glia 37, 206–218. Syroid, D. E., Maycox, P. J., Soilu-Hanninen, M., Petratos, S., Bucci, T., Burrola, P., Murray, S., Cheema, S., Lee, K. F., Lemke, G., and Kilpatrick, T. J. (2000). Induction of postnatal schwann cell death by the lowaYnity neurotrophin receptor in vitro and after axotomy. J. Neurosci. 20, 5741–5747. Syroid, D. E., Maycox, P. R., Burrola, P. G., Liu, N., Wen, D., Lee, K.-F., Lemke, G., and Kilpatrick, T. J. (1996). Cell death in the Schwann cell lineage and its regulation by neuregulin. Proc. Natl. Acad. Sci. U. S. A. 93, 9229–9234. Syroid, D. E., Zorick, T. S., Arbet-Engels, C., Kilpatrick, T. J., Eckhart, W., and Lemke, G. (1999). A role for insulin-like growth factor-I in the regulation of Schwann cell survival. J. Neurosci. 19, 2059–2068. Tabernero, A., Stewart, H. J. S., Jessen, K. R., and Mirsky, R. (1998). The neuron-glia signal ß neuregulin induces sustained CREB phosphorylation on Ser-133 in cultured rat Schwann cells. Mol. Cell. Neurosci. 10, 309–322. Takashima, H., Boerkoel, C. F., De Jonghe, P., Ceuterick, C., Martin, J. J., Voit, T., Schroder, J. M., Williams, A., Brophy, P. J., Timmerman, V., and Lupski, J. R. (2002). Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51, 709–715.

II. GLIAL CELL DEVELOPMENT

SCHWANN CELL CYTOSKELETON AND INTERACTIONS

Thatikunta, P., Qin, W., Christy, B. A., Tennekoon, G. I., and Rutkowski, J. L. (1999). Reciprocal Id expression and myelin gene regulation in Schwann cells. Mol. Cell. Neurosci. 14, 519–528. Thompson, M. A., and ZiV, E. B. (1989). Structure of the gene encoding peripherin, an NGF-regulated neuronalspeciWc type III intermediate Wlament protein. Neuron 2, 1043–1053. Tikoo, R., Zanazzi, G., ShiVman, D., Salzer, J., and Chao, M. V. (2000). Cell cycle control of Schwann cell proliferation: Role of cyclin-dependent kinase-2. J. Neurosci. 20, 4627–4634. Timmerman, V., De Jonghe, P., Ceuterick, C., De Vriendt, E., Lofgren, A., Nelis, E., Warner, L. E., Lupski, J. R., Martin, J. J., and Van Broeckhoven, C. (1999). Novel missense mutation in the early growth response 2 gene associated with Dejerine-Sottas syndrome phenotype. Neurology 52, 1827–1832. Timsit, S. G., Bally-Cuif, L., Colman, D. R., and Zalc, B. (1992). DM-20 mRNA is expressed during the embryonic development of the nervous system of the mouse. J. Neurochem. 58, 1172–1175. Tiveron, M.-C., Hirsch, M.-R., and Brunet, J.-F. (1996). The expression pattern of the transcription factor Phox2 delineates synaptic pathways of the autonomic nervous system. J. Neurosci. 16, 7649–7660. Tontonoz, P., Hu, E., and Spiegelman, B. M. (1994). Stimulation of adipogenesis in Wbroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 79, 1147–1156. Topilko, P., and Meijer, D. (2001). Transcription factors that control Schwann cell development and myelination. In ‘‘Glial Cell Development: Basic Principles and Clinical Relevance’’ (K. R. Jessen and W. D. Richardson, eds.), 2nd ed., pp. 223–244. Oxford University Press, Oxford. Topilko, P., Levi, G., Merlo, G., Mantero, S., Desmarquet, C., Mancardi, G., and Charnay, P. (1997). DiVerential regulation of the zinc Wnger genes Krox-20 and Krox-24 (Egr-1) suggests antagonistic roles in Schwann cells. J. Neurosci. Res. 50, 702–712. Topilko, P., Schneider-Manoury, S., Levi, G., Baron-Van Evercooren, A., ChennouW, A. B. Y., Seitanidou, T., Babinet, C., and Charnay, P. (1994). Krox-20 controls myelination in the peripheral nervous system. Nature 371, 796–799. Topilko, P., Schneider-Maunoury, S., Levi, G., Trembleau, A., Gourdji, D., Driancourt, M. A., Rao, C. V., and Charnay, P. (1998). Multiple pituitary and ovarian defects in Krox-24 (NGFI-A, Egr-1)-targeted mice. Mol. Endocrinol. 12, 107–122. Tosh, D., and Slack, J. M. (2002). How cells change their phenotype. Nature Rev. Mol. Cell Biol. 3, 187–194. Trachtenberg, J. T., and Thompson, W. J. (1996). Schwann cell apoptosis at developing neuromuscular junctions is regulated by glial growth factor. Nature 379, 174–177. Trofatter, J. A., MacCollin, M. M., Rutter, J. L., Murrell, J. R., Duyao, M. P., Parry, D. M., Eldridge, R., Kley, N., Menon, A. G., Pulaski, K., Haase, V. H., Ambrose, C. M., Munroe, D., Bove, C., Haines, J. L., Martuza, R. L., MacDonald, M. E., Seizinger, B. R., Short, M. P., Buckler, A. J., and Gusella, J. F. (1993). A novel moesin-, ezrin-, radixin-like gene is a candidate for the neuroWbromatosis 2 tumor suppressor. Cell 72, 791–800. Tsiper, M. V., and Yurchenco, P. D. (2002). Laminin assembles into separate basement membrane and Wbrillar matrices in Schwann cells. J. Cell Sci. 115, 1005–1015. Uziyel, Y., Hall, S., and Cohen, J. (2000). InXuence of laminin-2 on Schwann cell-axon interactions. Glia 32, 109–121. van der Neut, R., Krimpenfort, P., Calafat, J., Niessen, C. M., and Sonnenberg, A. (1996). Epithelial detachment due to absence of hemidesmosomes in integrin beta 4 null mice. Nat. Genet. 13, 366–369. Vartanian, T., Goodearl, A., Viehover, A., and Fischbach, G. (1997). Axonal neuregulin signals cells of the oligodendrocyte lineage through activation of HER4 and Schwann cells through HER2 and HER3. J. Cell Biol. 137, 211–220. Wakamatsu, Y., Maynard, T. M., Jones, S. U., and Weston, J. A. (1999). NUMB localizes in the basal cortex of mitotic avian neuroepithelial cells and modulates neuronal diVerentiation by binding to NOTCH-1. Neuron 23, 71–81. Wakamatsu, Y., Maynard, T. M., and Weston, J. A. (2000). Fate determination of neural crest cells by NOTCHmediated lateral inhibition and asymmetrical cell division during gangliogenesis. Development 127, 2811–2821. Wang, J. Y., Miller, S. J., and Falls, D. L. (2001a). The N-terminal region of neuregulin isoforms determines the accumulation of cell surface and released neuregulin ectodomain. J. Biol. Chem. 276, 2841–2851. Wang, S., Sdrulla, A., Johnson, J. E., Yokota, Y., and Barres, B. A. (2001b). A role for the helix-loop-helix protein Id2 in the control of oligodendrocyte development. Neuron 29, 603–614. Wanner, I. B., and Wood, P. M. (2002). N-cadherin mediates axon-aligned process growth and cell-cell interaction in rat Schwann cells. J. Neurosci. 22, 4066–4079. Warner, L. E., Garcia, C. A., and Lupski, J. R. (1999). Hereditary peripheral neuropathies: Clinical forms, genetics, and molecular mechanisms. Annu. Rev. Med. 50, 263–275. Warner, L. E., Mancias, P., Butler, I. J., McDonald, C. M., Keppen, L., Koob, K. G., and Lupski, J. R. (1998). Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nature Genet. 18, 382–384. Wegner, M. (1999). From head to toes: The multiple facets of Sox proteins. Nucleic Acids Res. 27, 1409–1420. Wegner, M. (2000). Transcriptional control in myelinating glia: The basic recipe. Glia 29, 118–123. Weiner, J. A., and Chun, J. (1999). Schwann cell survival mediated by the signaling phospholipid lysophosphatidic acid. Proc. Natl. Acad. Sci. U. S. A. 96, 5233–5238. Weiner, J. A., Fukushima, N., Contos, J. J., Scherer, S. S., and Chun, J. (2001). Regulation of Schwann cell morphology and adhesion by receptor-mediated lysophosphatidic acid signaling. J. Neurosci. 21, 7069–7078.

II. GLIAL CELL DEVELOPMENT

369

370

13. SCHWANN CELL DEVELOPMENT

Weinmaster, G., and Lemke, G. (1990). Cell-speciWc cyclic AMP-mediated induction of the PDGF receptor. EMBO J. 9, 915–920. White, P. M., Morrison, S. J., Orimoto, K., Kubu, C. J., Verdi, J. M., and Anderson, D. J. (2001). Neural crest stem cells undergo cell-intrinsic developmental changes in sensitivity to instructive diVerentiation signals. Neuron 29, 57–71. Williams, A. C., and Brophy, P. J. (2002). The function of the Periaxin gene during nerve repair in a model of CMT4F. J. Anat. 200, 323–330. Winseck, A. K., Caldero´, J., Ciutat, D., Prevette, D., Scott, S. A., Wang, G., Esquerda, J. E., and Oppenheim, R. W. (2002). In vivo analysis of Schwann cell programmed cell death in the embryonic chick: Regulation by axons and glial growth factor. J. Neurosci. 22, 4509–4521. Woldeyesus, M. T., Britsch, S., Riethmacher, D., Xu, L., Sonnenberg-Riethmacher, E., Abou-Rebyeh, F., Harvey, R., Caroni, P., and Birchmeier, C. (1999). Peripheral nervous system defects in erbB2 mutants following genetic rescue of heart development. Genes Dev. 13, 2538–2548. Wolpowitz, D., Mason, T. B., Dietrich, P., Mendelsohn, M., Talmage, D. A., and Role, L. W. (2000). Cysteinerich domain isoforms of the neuregulin-1 gene are required for maintenance of peripheral synapses. Neuron 25, 79–91. Wood, P. M., Schachner, M., and Bunge, R. P. (1990). Inhibition of Schwann cell myelination in vitro by antibody to the L1 adhesion molecule. J. Neurosci. 10, 3635–3645. Woodhoo, A., Dean, C. H., Droggiti, A., Mirsky, R., and Jessen, K. R. (2003). The trunk neural crest and its early glial derivatives: a study of survival responses, developmental schedules and autocrine mechanisms. Mol. Cell. Neurosci., in press. Woolf, C. J., Reynolds, M. L., Chong, M. S., Emson, P., Irwin, N., and Benowitz, L. I. (1992). Denervation of the motor endplate results in the rapid expression by terminal Schwann cells of the growth-associated protein GAP-43. J. Neurosci. 12, 3999–4010. Wu, R., Jurek, M., Sundarababu, S., and Weinstein, D. E. (2001). The POU gene Brn-5 is induced by neuregulin and is restricted to myelinating Schwann cells. Mol. Cell. Neurosci. 17, 683–695. Yang, J. T., Rayburn, H., and Hynes, R. O. (1995). Cell adhesion events mediated by alpha 4 integrins are essential in placental and cardiac development. Development 121, 549–560. Yoshimura, T., Goda, S., Kobayashi, T., and Goto, I. (1993). Involvement of protein kinase C in the proliferation of cultured Schwann cells. Brain Res. 617, 55–60. Yoshihara, T., Kanda, F., Yamamoto, M., Ishihara, H., Misu, K., Hattori, N., Chihara, K., and Sobue G. (2001). A novel missense mutation in the early growth response 2 gene associated with late-onset Charcot-MarieTooth disease type 1. J. Neurol. Sci. 184, 149–153. Zanazzi, G., Einheber, S., Westreich, R., Hannocks, M. J., Bedell-Hogan, D., Marchionni, M. A., and Salzer, J. L. (2001). Glial growth factor/neuregulin inhibits Schwann cell myelination and induces demyelination. J. Cell Biol. 152, 1289–1299. Zelena, J. (1994). ‘‘Nerves and Mechanoreceptors.’’ Chapman and Hall, London. Zhang, J. Y., Luo, X. G., Xian, C. J., Liu, Z. H., and Zhou, X. F. (2000). Endogenous BDNF is required for myelination and regeneration of injured sciatic nerve in rodents. Eur. J. Neurosci. 12, 4171–4180. Zhang, D., Sliwkowski, M. X., Mark, M., Frantz, G., Akita, R., Sun, Y., Hillan, K., Crowley, C., Brush, J., and Godowski, P. J. (1997). Neuregulin-3 (NRG3): A novel neural tissue-enriched protein that binds and activates ErbB4. Proc. Natl. Acad. Sci. U. S. A. 94, 9562–9567. Zilian, O., Saner, C., Hagedorn, L., Lee, H. Y., Sa¨uberli, E., Suter, U., Sommer, L., and Aguet, M. (2001). Multiple roles of mouse Numb in tuning developmental cell fates. Curr. Biol. 11, 494–501. Zorick, T. S., Syroid, D. E., Arroyo, E., Scherer, S. S., and Lemke, G. (1996). The transcription factors SCIP and Krox-20 mark distinct stages and cell fates in Schwann cell diVerentiation. Mol. Cell. Neurosci. 8, 129–145. Zorick, T. S., Syroid, D. E., Brown, A., Gridley, T., and Lemke, G. (1999). Krox-20 controls SCIP expression, cell cycle exit and susceptibility to apoptosis in developing myelinating Schwann cells. Development 126, 1397–1406.

C H A P T E R

14 Olfactory Ensheathing Cells Robin J. M. Franklin and Susan C. Barnett

INTRODUCTION

au1

The inclusion of a cell type that in its natural location is a nonmyelinating cell may, on the face of it, seem strange in a textbook on myelin and its diseases. However, olfactory ensheathing cells (OECs, also referred to as olfactory ensheathing glia), although a distinct cell type, bear a close resemblance to Schwann cells of the PNS and are able to assume a myelinating phenotype when provided with the correct signals. This chapter reviews the biology of this interesting and unusual glial cell and focuses in particular on its myelinating properties and potential for transplant-mediated repair of CNS demyelination (see also Chapter 6). The use of transplanted OECs to promote axon regeneration after traumatic injury of the CNS has received considerable attention recently, but a full discussion of this topic is beyond the scope of this chapter and the interested reader is directed to a number of recent reviews and articles (Bunge, 2001; Franklin and Barnett, 2000; Raisman, 2001; Ramo´n-Cueto and Valverde, 1995; Ramon-Cueto et al., 2000; Takami et al., 2002; Wewetzer et al., 2002).

Anatomy of the Peripheral Olfactory System The peripheral olfactory system, within which OECs reside, comprises a main component and an accessory component called the vomeronasal organ. For the purposes of this chapter, the biology of these two components is suYciently similar that no further distinction between the two will be made. The peripheral olfactory system consists of the olfactory epithelium, the olfactory nerve and the nerve Wber layer of the olfactory bulb (Fig. 14.1). The olfactory epithelium contains the cell bodies of the mature olfactory receptor neurons, as well as these same cells in various stages of maturity, sustentacular cells, horizontal basal cells, and globose basal cells (Calof et al., 2000; Schwob, 2002). The olfactory receptor neurons respond to airborne odorants by way of receptors situated on cilia emanating from their dendritic processes. These project into the mucous layer on the surface of the epithelium. Their axons exit the epithelial layer of the mucosa and traverse the lamina propria where they become loosely gathered together as the olfactory or Wrst cranial nerve. Fiber bundles of the olfactory nerve pass through the cribriform plate at the base of the skull and then form a meshwork of Wbers covering the surface of the distal extent of the olfactory bulb, termed the nerve Wber layer of the olfactory bulb. The axons terminate at synapses on the dendritic arborisations of mitral cells (the second-order neurons) in discrete, clearly deWned and roughly spherical regions of neuropil termed glomeruli. Throughout their passage from the lamina propria to the nerve Wber layer, the axons of the olfactory nerve are ensheathed by OECs. The OECs within the olfactory nerve are sometimes referred to as olfactory nerve ensheathing cells (ONECs), while those in the

Myelin Biology and Disorders, Volume 1

371

Copyright 2004, Elsevier (USA). All rights reserved.

372

14. OLFACTORY ENSHEATHING CELLS

FIGURE 14.1

au6

The peripheral olfactory system, the olfactory bulb, and the anatomical location of OECs. This drawing by Ramo´n y Cajal illustrates the location of the olfactory sensory neurons (A) in the olfactory nasal mucosa (D). The axons of these neurons pass through the cribiform plate (C) and become the nerve Wber layer of the olfactory bulb (B), eventually making synaptic contact (C) with the dendrites of the mitral cells (E) in the glomeruli. The OECs ensheath these axons throughout their length occurring within the olfactory nerve and the olfactory nerve Wber layer. From ‘‘New Ideas on the Structure of the Nervous System in Man and Animals’’ by Santiago Ramo´n y Cajal, 1894, republished in 1990 by MIT Press, Cambridge MA, N. Swanson and L. W. Swanson, eds.

nerve Wber layer are sometimes referred to as olfactory bulb ensheathing cells (OBECs). Whether these are identical populations or manifest subtle diVerences from one another is not clear, and generic terms that do not distinguish between the two are most commonly used. Although OECs broadly resemble the nonmyelinating Schwann cells that associate with small diameter axons in peripheral nerve, the nature of the ensheathment is somewhat diVerent. Whereas nonmyelinating Schwann cells embrace several axons that are separated one from another by folds of Schwann cell cytoplasm, OECs form a much less intimate association where instead thin OEC processes embrace many juxtaposed axons (Doucette, 1990, 1993; Fraher, 1982; Wiley, 1973).

Neurogenesis in the Olfactory System The adult olfactory system is among the few areas of the mammalian CNS that actively engages in neurogenesis throughout life (Graziadei and Monti Graziadei, 1983; Schwob, 2002). Two populations are continuously replaced: the olfactory receptor neurons of the olfactory epithelium and the granule cell interneurons within the olfactory bulb. The latter are generated from precursor cells that originate in the subventricular zone and reach the olfactory bulb by way of the rostral migratory stream (Doetsch and Alvarez-Buylla, 1996; Lois and Alvarez-Buylla, 1994). Since the OECs only ensheath the olfactory receptor neurons, we will limit our discussion to the turnover of neurons in the peripheral olfactory system.

II. GLIAL CELL DEVELOPMENT

INTRODUCTION

By virtue of their projection into the nasal cavity and direct exposure to the external environment, the olfactory receptor neurons are without question the most vulnerable neuronal population in the body, constantly at risk of exposure to potential airborne toxins. For this reason, the peripheral olfactory system of mammals has retained an ability to replenish itself without which an individual would soon lose the important sense of olfaction. Indeed, so eYcient is this system that olfactory receptor neurons are constantly being replenished in response to injury and possibly as part of a normal physiological process (Carr and Farbman, 1992; Farbman, 1990; Graziadei and Monti Graziadei, 1983), although this latter claim has been questioned (Hinds et al., 1984; Mackay-Sim and Kittel, 1991b). Experimentally, widespread death of olfactory receptor neurons can be induced by transection of the olfactory nerve or by irrigating the nasal mucosa with toxins such as zinc sulphate, both of which trigger generation of new, replacement olfactory receptor neurons (Burd, 1993; Chuah et al., 1995; Graziadei and Graziadei, 1979; Harding et al., 1978; Holcomb et al., 1995). The new neurons are thought to derive from the globose basal cells (Caggiano et al., 1994; Mackay-Sim and Kittel, 1991a; Schwartz-Levey et al., 1991; Schwob, 2002), although whether these include a true stem cell population or are exclusively an amplifying progenitor population is unclear (Calof et al., 1996, 2000; Gordon et al., 1995; Mumm et al., 1996; Schwob, 2002). The newly formed olfactory receptor neurons grow axonal processes that extend toward the olfactory bulb, eventually re-establishing the presynaptic element of the synapses with the second-order neurons (mitral cells) within the olfactory bulb glomeruli (Graziadei and Monti Graziadei, 1980; Monti Graziadei et al., 1980). Throughout their growth, the axons are in close contact with OECs that remain within the pathways formerly occupied by the previous population of olfactory receptor neuron axons. The arrangement resembles the regeneration of damaged peripheral nerve axons as they regrow through the bands of Bu¨ngner of the distal stump, and in both cases the glial environment helps guide the axon to its synaptic target (Doucette et al., 1983; Raisman, 1985). Like the Schwann cells within bands of Bu¨ngner, OECs are highly supportive of axonal outgrowth (Goodman et al., 1993; Le Roux and Reh, 1994; Sonigra et al., 1999).

OECs and the CNS/PNS Interface In general there is a clear boundary between the CNS and PNS, a prime example being that which occurs at the dorsal root entry zone (DREZ) where a dorsal root connects to the spinal cord. This boundary is delineated by the glia limitans, an epithelial-like arrangement of astrocyte processes on one side of which occur neuroepithelium-derived CNS glia and on the other neural crest-derived PNS glia. If axons of the dorsal root ganglion (DRG) neurons are cut between the DRG and the DREZ, then the axons will grow through the Schwann cell environment of the dorsal root, but their progress is impeded by contact with the astrocytes at the DREZ (Liuzzi and Lasek, 1987). The axons are therefore unable to pass from a PNS to a CNS environment and make appropriate synaptic contacts. If such a situation pertained in the peripheral olfactory system, then olfactory receptor neuron turnover would be futile. That it does not is thought to be due to the properties of the OECs, which permit growing olfactory receptor neurons to breach the CNS/PNS boundary and synapse with the second-order neurons (Doucette, 1991; Raisman, 1985). A question that has emerged from this property of OECs is where does the boundary between PNS and CNS exist in the olfactory bulb? One view is that the entirety of the nerve Wber layer is peripheral in origin and that it constitutes a cap of PNS tissue on the surface of the olfactory bulb. This would make the boundary where the nerve Wber layer abuts the rest of the bulb, albeit a very indistinct interface that contrasts with the boundary occurring elsewhere in the nervous system (Doucette, 1989; Valverde and Lopez-Mascaraque, 1991). Some authors have described cells present within the nerve Wber layer that they have called astrocytes (Bailey and Shipley, 1993; Doucette, 1984). If these were astrocytes of neuroepithelial origin, then this would mean that the nerve Wber layer contains a mixed population of both PNS and CNS glia and would be unique in this regard (Franklin and Barnett, 1997). However, they may represent a subpopulation of OECs that derive from the same peripheral

II. GLIAL CELL DEVELOPMENT

373

374

14. OLFACTORY ENSHEATHING CELLS

source. This issue remains to be resolved and will require a clear description of the heterogeneity of glial cells within the nerve Wber layer (see the section titled ‘‘OEC Heterogeneity’’), especially the developmental origin of each cell type identiWed.

Development of OECs and the Peripheral Olfactory System The olfactory epithelium and olfactory ensheathing cells originate from the olfactory placodes (Chuah and Au, 1991; De Carlos et al., 1995; Norgren et al., 1992; Valverde et al., 1992). The olfactory placodes are present at the rostral tip of the embryonic head and develop as a localized thickening of the epithelial tissue, forming a pseudostratiWed tissue with an outer layer of non-neural ectoderm and an inner layer of neural ectoderm (Brunjes and Frazier, 1986; De Carlos et al., 1995). This structure invaginates to give rise to the olfactory pit and later on to the olfactory epithelium. During the formation of the olfactory bulb, the axons of the olfactory receptor neurons grow and the OECs migrate toward the telencephalic vesicles forming a structure termed the ‘‘migratory mass’’ (De Carlos et al., 1995; Valverde et al., 1992). The OECs present in this structure are believed to guide the olfactory receptor neuron axons to the telencephelon. The migratory mass also contains gonadotrophin releasing-hormone (GnRH)-expressing neurons that originate in the olfactory placode and are destined for their Wnal location within the hypothalamus. The movement of the migratory mass is disrupted in Kallman’s syndrome, an inherited disorder characterized by anosmia and infertility. Between E15 and E18 the migratory mass extends toward the dorsal and lateral sides of each developing olfactory bulb, giving rise to the nerve Wber layer. To achieve this, the migratory mass becomes interposed between the primordium of the pia mater and the glia limitans delineating the surface of the rostral wall of the cerebral vesicle and the presumptive olfactory bulb (Doucette, 1989; Marin-Padilla and Amieva, 1989). The basal lamina of the glia limitans is broken down where the wall of the cerebral vesicle is contacted by the olfactory axons, which then enter the developing olfactory bulb (Marin-Padilla and Amieva, 1989). Most of the OECs within the migratory mass will remain in the nerve Wber layer (Doucette, 1990), although some authors suggest that some may also penetrate the olfactory bulb (Marin-Padilla and Amieva, 1989).

THE OLFACTORY ENSHEATHING CELL (OEC) The cells that we now call OECs were originally described by the Spanish scientist Blanes at the end of the 19th century (Blanes, 1898). Although initially referred to as a Schwann cells (Barber and Lindsay, 1982; Wiley, 1973), later studies suggested that these cells were more astrocyte-like due to their intense expression of glial Wbrillary acidic protein (GFAP) and their immunoreactivity with an antibody that recognized the GFAP of CNS astrocytes but not that expressed by nonmyelinating Schwann cells (Barber and Dahl, 1987; Barber and Lindsay, 1982; Pixley, 1992). Their astrocyte features also include their ability to form part of the glial limitans of the olfactory bulb (Doucette, 1991). From this emerged a view that OECs represented a class of glia cell sharing properties of both CNS and PNS glia, a view endorsed by the apparent juxtaposition of OECs and astrocytes in the nerve Wber layer of the olfactory bulb (see the section titled ‘‘OECs and the CNS/PNS Interface’’).

OEC Heterogeneity The issue of whether the OEC is more CNS-like or PNS-like has now largely been resolved in favor of the OEC being essentially a unique PNS glial cell but with some features characteristic of CNS glia (Barnett et al., 1993; Ramo´n-Cueto and Valverde, 1995). Thus, the OEC is derived from the olfactory placode, a structure that bears some similarity to neural crest, and in tissue culture most frequently exhibits a phenotype very similar to that of a nonmyelinating Schwann cell, including the expression of the low-aYnity neurotrophin receptor P75NTR and the adhesion molecule L1 (Miragall et al., 1989). Nevertheless, it is

II. GLIAL CELL DEVELOPMENT

THE OLFACTORY ENSHEATHING CELL (OEC)

375

now clear that OECs are an heterogeneous cell population that can exist in more than one distinct phenotype (Doucette, 1984; Pixley, 1992). The best characterized classiWcation of this heterogeneity is the division of OEC into a Schwann-cell like cell that expresses P75NTR but has weak and diVuse expression of GFAP and does not express the polysialylated form (embryonic) of the neural cell adhesion molecule (E-NCAM), and an astrocyte-like cell that is intensely immunoreactive for GFAP, expresses E-NCAM, but does not express P75NTR (Franceschini and Barnett, 1996) (Fig. 14.2). Analysis of a clonal OEC line indicates that these two types of OEC share a common lineage (Franceschini and Barnett, 1996), while studies on primary cells indicate that they may represent two ends of a phenotypic spectrum along which other phenotypes may exist that express features of both cell types to a greater or lesser degree (Alexander et al., 2002). The phenotype of OECs in tissue culture can be manipulated to some extent by the composition of the medium in which they are grown (Alexander et al., 2002; Barnett et al., 2000; Franceschini and Barnett, 1996). The heterogeneity of OEC described in tissue culture studies reXects histological studies on the distribution of OEC markers in the olfactory nerve and nerve Wber layer of the olfactory bulb, where some markers such P75NTR, GFAP, and O4 have distinct distributions (Au et al., 2002; Franceschini and Barnett, 1996; Pixley, 1992). The signiWcance of this heterogeneity is not known at present but may indicate functional subdivisions within the OEC population that might also be harnessed to optimize the repair enhancing capabilities of transplanted OECs. A number of other proteins are expressed by OEC, including as a range of connexins (Barnett et al., 2001), the DM20 isoform of the myelin protein gene proteolipid protein (Dickinson et al., 1997), the signaling molecule desert hedgehog (Smith et al., 2001), and the neuromodulator neuropeptide Y (Ubink et al., 1994). Two antibodies have been described that are claimed to speciWcally identify OECs, but they have not been widely used (Heredia et al., 1998). The functional roles of any of these proteins in OEC biology have not been evaluated.

Growth Factor Responsiveness of OECs A number of growth factors are mitogenic for OECs and are also able to aVect their phenotype. In early studies, medium conditioned by cultured astrocytes was used as a source of mitogens and survival factors (Barnett et al., 1993; Franceschini and Barnett, 1996). Subsequent studies identiWed members of the neuregulin family of signaling molecules, in particular neu diVerentiation factors (NDF) b1,2 and 3, as critical components of astrocyte-conditioned medium that exert these eVects (Pollock et al., 1999). PDGF-BB, FGF-2, NGF, and IGF-I are also found to be mitogenic but with considerably less potency

FIGURE 14.2 Antigenic and morphological appearance of OECs in tissue culture. (A) Schwann cell–like OECs immunolabeled with anti-P75NTR, grown in medium containing FGF-2, heregulin b1 and forskolin. (B) Astrocyte-like OECs immediately after puriWcation, grown in medium containing astrocyteconditioned medium (ACM) and labeled with anti GFAP. (C) Two distinct OEC phenotypes labeled with anti-PSA-NCAM and anti-P75NTR. Parts A and B are of primary neonatal rat OECs and part C is from a rat OEC line (see Franceschini and Barnett, 1996).

II. GLIAL CELL DEVELOPMENT

376

14. OLFACTORY ENSHEATHING CELLS

than neuregulins (Chuah and Teague, 1999; Pollock et al., 1999; Yan et al., 2001). Several mitogens, in particular heregulin (a human recombinant form of neuregulin) together with FGF-2, PDGF-BB or IGF-I, are additive when used in combination (Yan et al., 2001). A particularly potent combination of mitogens consists of heregulin, FGF-2 and forskolin (Alexander et al., 2002). This combination selects for an OEC population that expresses O4 and P75NTR, but with continued exposure the cells also begin to express E-NCAM. Of potential value for future therapeutic applications of OECs, this same combination also allows longer-term expansion of OECs than is achievable with astrocyte conditioned medium alone, making it possible to generate the large numbers of OECs required to make transplant-based strategies feasible. However, not all growth factors that are mitogenic for rodent OECs are also mitogenic for human OECs, which so far have been most successfully expanded using a combination of heregulin b1 and forskolin (Barnett et al., 2000).

MYELINATION BY OECs In their natural location within the peripheral olfactory system, OECs associate with axons that have diameters less than 0.5 micron and are well below the threshold for myelination found in either the PNS or CNS (Doucette, 1990). However, given the OEC’s similarity to the nonmyelinating Schwann cell, it was speculated that the OEC might also be able to myelinate if it encountered an axon of appropriate diameter. This was Wrst demonstrated using an in vitro model in which neurites from DRG neurons were myelinated by co-cultured OECs (Devon and Doucette, 1992) (Fig. 14.3). The myelin sheath formed closely resembled the myelin sheath made by Schwann cells, but unlike Schwann cells did not require L-ascorbic acid to do so (Devon and Doucette, 1995). However, other in vitro data suggest that OEC may be less conducive to assuming a myelinating phenotype (Doucette and Devon, 1994; 1995), and recently it has been shown that OECs are much less eYcient at forming myelin sheaths than are Schwann cells (Plant et al., 2002).

Myelination by Transplanted OECs To further establish the myelinating potential of OECs, an in vivo model was used in which transplanted OECs were brought into contact with large diameter demyelinated axons.

FIGURE 14.3 In vitro myelination of dorsal root ganglion neurites by OECs derived from embryonic rat olfactory bulb nerve Wber layer. Only the larger diameter axons are myelinated. Scale bar ¼ 1 mm (see Doucette and Devon, 1992, 1996).

II. GLIAL CELL DEVELOPMENT

MYELINATION BY OEC

The model used was the X-irradiated ethidium bromide (X-EB) model described in detail in chapter 6. The X-EB lesion has no intrinsic repair potential and thus provides a convenient means of establishing the myelinating potential of cultured glial cells injected directly into the lesion. Given the potential diYculties in determining whether myelin sheaths observed following transplantation of OECs into X-EB lesions were formed by OECs or Schwann cell contaminants, the Wrst transplant experiments used an OEC line created by infecting cultures of OECs with a retrovirus containing the large T oncogene (Franceschini and Barnett, 1996). A temperature-sensitive mutant form of large T was used enabling the eVects of the immortalizing protein to be switched on and oV by altering the ambient temperature. Importantly, the large T protein is rendered inactive at the normal body temperature of a rat meaning that its immortalizing eVects are switched oV following transplantation. It was established by Southern blotting following enzymatic digestion of genomic DNA at a unique restriction site within the retroviral construct that the isolated cells were derived from a single clone and thus constituted a pure OEC preparation free from contaminants (Franceschini and Barnett, 1996). This meant that the myelin sheaths seen following transplantation of this OEC line into X-EB lesions were OEC derived. Using this approach, it was found that the overwhelming majority of the transplanted cells formed myelin sheaths around the demyelinated axons within the lesion (Franklin et al., 1996) (Fig. 14.4). It was subsequently shown that the myelinating properties of OECs were not a peculiarity of the cell line but are also exhibited by primary OECs transplanted X-EB lesions (Imaizumi et al., 1998; Lakatos et al., 2003; Smith et al., 2001) (Fig. 14.5). Moreover, it has been shown that transplanted OECs permit recovery of conduction properties in demyelinated axons that have remyelinated (Imaizumi et al., 1998, 2000). VeriWcation that the myelin sheaths formed by transplantation into X-EB lesions are derived from the transplanted OECs has been achieved in two ways. First, following transplantation of OECs carrying the LacZ reporter gene into X-EB lesions labeled myelinating cells can be

FIGURE 14.4 Remyelination following transplantation of a clonal OEC line into focal areas of demyelination in adult rat spinal cord (X-EB lesion). (A) A myelinating OEC. (B) The myelin sheath is surrounded by a rim of cytoplasm and the cell membrane is covered by a basal lamina (arrow). (C) Small diameter axons are ensheathed but not myelinated. (D) The myelin sheaths express the peripheral myelin speciWc protein periaxin (antibody a gift from Prof. Peter J. Brophy). Scale bar ¼ (A) 1.3 mm, (B) 0.3 mm, (C) ¼ 0.8 mm, (D) ¼ 10 mm.

II. GLIAL CELL DEVELOPMENT

377

378

14. OLFACTORY ENSHEATHING CELLS

FIGURE 14.5 Remyelination by OECs puriWed by immunopanning with P75NTR antibodies following transplantation into focal areas of demyelination in adult rat spinal cord (X-EB lesion). (A) In toluidine blue-stained resin sections, the darkstaining OEC-derived myelin sheaths can be readily identiWed. The adjacent OEC cell body frequently gives rise to a characteristic signet ring appearance in transverse sections. (B) Electronmicrograph of a myelinating OEC that has been labeled with the LacZ marker gene. The gene contains a nuclear localizing signal so the electron dense reaction product is mainly located within the nuclear membrane (unstained section). Scale bar ¼ (A) 9 mm, (B) 1.25 mm

identiWed by electronmicroscopy (Fig. 14.5b). Second, when OECs obtained from adult dog olfactory bulbs are transplanted into X-EB lesions in rats, myelin sheath formation is only seen when the recipients are immunosuppressed (Smith et al., 2002). In the absence of immunosuppression, the dog cells are rejected and no myelin sheaths are seen with the lesion, indicating that the myelin sheaths in the immunosupressed recipients are of donor origin. In addition to rat and dog, myelin sheath formation has also been demonstrated following transplantation of OECs derived from pig (Imaizumi et al., 2000) and, importantly in the context of potential clinical application, from adult human peripheral olfactory tissue (Barnett et al., 2000; Kato et al., 2000) (Fig. 14.6).

Myelination by OECs Resembles Schwann Cell Myelination The myelin formed by OECs is typical of peripheral-type myelin. At the light microscope level, the myelin sheaths are immunoreactive for P0 (Franklin et al., 1996; Smith et al., 2001, 2002) (Fig. 14.6a) and periaxin (unpublished data) (Fig. 14.4d), both of which are proteins found in peripheral but not central myelin (Gillespie et al., 1994; Lemke and Axel, 1985). In toluidine blue-stained resin sections, the myelin sheath has a dark staining hue by which it can be distinguished from myelin formed by myelinating or remyelinating oligodendrocytes but is indistinguishable from myelin formed by Schwann cells. The morphology of the myelinating cell is also very similar to that of a myelinating Schwann cell. Since both cells make a single myelin sheath, the cell body often rests adjacent to the myelinated axon, giving rise to a characteristic ‘‘signet ring’’’ appearance (Fig. 14.5a). Using electronmicroscopy, the similarity of OEC myelination to Schwann cell myelination is even more apparent. In both cases the myelin sheath is surrounded by a distinct rim of cytoplasm (Figs. 14.4, 14.5, and 14.6), and the surface of the myelinating cell is usually covered by a basal lamina (Fig. 14.4b). Very small diameter axons are not myelinated by OECs (Franklin et al., 1996; Plant et al., 2002) (Fig. 14.4c). Indeed it is not currently possible to reliably distinguish myelination by OECs from myelination by Schwann cells. This similarity between the myelinating forms of both cells poses diYculties in distinguishing myelination by one cell type from the other. In the context of their potential use for repairing areas of demyelination in MS, the peripheral-type myelin sheath made by OECs may have the advantage of being resistant to a disease process

II. GLIAL CELL DEVELOPMENT

MYELINATION BY OEC

379

FIGURE 14.6 Remyelination by OECs derived from adult human olfactory bulb following transplantation into focal areas of demyelination in adult rat spinal cord (X-EB lesion). (A) In toluidine blue-stained resin sections, OEC-remyelinated axons can be seen. Myelination by human OEC takes longer than remyelination by rat OECs in this model. (Inset: the myelin sheaths made by the transplanted cells express the peripheral myelin speciWc protein P0.) (B) The ultrastructure of remyelination by human OECs is similar to that achieved by rodent OECs. Scale bar ¼ (A) 8 mm (inset ¼ 7.5 mm), (B) 1.3 mm

au2

directed at antigens expressed in central-type myelin and might therefore create long-lasting repair in the face of continued disease. The transcriptional regulation of Schwann cell myelination has been intensively studied and speciWc roles for Wrst the POU-domain factor SCIP/Oct-6 and then the zinc Wnger factor Krox-20 have been described (Topilko and Meijer, 2001). By in situ hybridization of X-EB lesions into which OECs are transplanted, it has been shown that myelinating OECs also express both of these transcription factors suggesting that both cells, although distinct in their origin and properties, use a similar transcriptional machinery to be come myelinating cells (Smith et al., 2001). OECs, like Schwann cells, also express the secreted signaling factor desert hedgehog (dhh). In the peripheral nervous system this factor has a key role in orchestrating the behavior of perineurial cells, inducing them to adopt arrangements that result in the formation of fascicles (Parmantier et al., 1999). Since fascicles of OEC myelinated axons surrounded by perineurial-like cells are also seen following transplantation of an unpuriWed OEC preparation, it is possible that dhh serves a similar function as that described in the formation of the PNS (Li et al., 1998; Smith et al., 2001).

The Efficiency of Myelination by Transplanted OECs A recent study has addressed how remyelination by transplanted OECs might be optimized (Lakatos et al., 2003). Cell populations isolated from the olfactory nerve Wber layer of the neonatal rat olfactory bulb consists mainly of OECs, but also contain other cell types. Somewhat contrary to expectation it was found that unpuriWed preparation yield more extensive remyelination than puriWed OEC preparation. This result implied that other cells types present within the unpuriWed preparation, although not themselves contributing to remyelination, might assist the remyelination by the transplanted OECs. Subsequent analysis identiWed meningeal as one such cell population since when meningeal cells are co-transplanted with puriWed OECs, there is a signiWcant improvement in the extent of remyelination achieved (Lakatos et al., 2003) (Fig. 14.7). This result provides evidence that the naturally poor remyelinating capacity of OECs can be enhanced by the presence of a

II. GLIAL CELL DEVELOPMENT

380

14. OLFACTORY ENSHEATHING CELLS

80

% Axons remyelinated by OECs

70 60 50 40 30 20 10 0 unpurified

OEC

OEC:M 3:1

M

FIGURE 14.7 Graph showing the percentage of demyelinated axons available for remyelination in an X-EB lesion that have been remyelinated by transplanted OECs. The extent of remyelination achieved by unpuriWed cell preparations for the distal olfactory bulb is greater than that achieved by purifed OECs. The remyelinating capacity of puriWed OECs can be enhanced by co-transplantation of meningeal cells that are themselves unable to mediate remyelination (see Lakatos et al., 2003).

second nonmyelinating cell type, although the basis of this enhancement is not clear at present. By analogy with Schwann cell myelination, it could relate to the increased extracellular matrix generated by mesenchymal cells (Eldridge et al., 1989; Obremski et al., 1993), although a combination of diVerent mechanisms may well be operating. Nevertheless, this experiment indicates that the optimal repair potential of transplanted OECs may not be best served by transplanting pure OEC preparation but may instead depend of the presence of other cell types (Raisman, 2001).

OEC INTERACTIONS WITH ASTROCYTES AND THEIR POTENTIAL CLINICAL APPLICATIONS

au3

The recognition that transplanted OECs can myelinate demyelinated axons raises the possibility that they might be used in transplant-mediated repair of demyelinating disease (see Chapter 6). Since OECs myelinate in a manner that is very similar to that of Schwann cells, it is reasonable to ask why OEC transplantation would have any advantages over Schwann cell transplantation for cellular therapy of demyelinating disease (Franklin and Barnett, 1997; Franklin, 2002). Although transplanted Schwann cells are able to remyelinate demyelinated axons in the CNS, several lines of evidence suggest this is most eVective in circumstances where astrocytes are absent (Franklin and Blakemore, 1993). For example, when Schwann cells are transplanted into EB-induced lesions in the cerebellar white matter of adult rats, the areas of Schwann cell myelination is restricted to areas from which GFAP-expressing astrocytes are absent (Shields et al., 2000). This mutual exclusivity between Schwann cells and astrocytes is not absolute, and there are examples where Schwann cells are able to myelinate in astrocyte containing areas (Baron-Van Evercooren et al., 1996; Duncan et al., 1988; Duncan and HoVman, 1997). Nevertheless, the weight of evidence indicates that environments containing astrocytes are not optimal for Schwann cell remyelination. Since most pathological environments, including both acute and chronic lesions of MS, are characterized by the presence of astrocytes, Schwann cells may not be the cells of choice for transplant-mediated repair of nonremyelinating lesions.

II. GLIAL CELL DEVELOPMENT

OEC INTERACTIONS WITH ASTROCYTES AND THEIR POTENTIAL CLINICAL APPLICATIONS

Several observations indicate that OECs may be more astrocyte-compatible than Schwann cells. For example, when transplanted into areas of CNS trauma, there is minimal activation of host astrocytes leading to better integration of the transplanted OECs with the host environment in a manner less frequently observed with transplanted Schwann cells (Ramo´n-Cueto et al., 1998; Takami et al., 2002). A direct comparison between the interactions of either Schwann cells or astrocytes has been undertaken using a tissue culture approach, developing tissue culture protocols originally devised to show the inhibitory interactions between Schwann cells and astrocytes (Ghirnikar and Eng, 1994; Wilby et al., 1999). Whereas Schwann cells and astrocytes occupy distinct and nonoverlapping domains in co-culture, OECs and astrocytes frequently occupy the same areas (Lakatos et al., 2000). Moreover, whereas migrating Schwann cells are inhibited by contact with astrocytes, inducing in them a reactive hypertrophy, the migration of OECs is unimpeded by astrocytes. These eVects of Schwann on astrocytes in tissue culture are mirrored by the recent Wnding that transplanted Schwann cells may actively induce the increased expression of repair-inhibiting molecules by host astrocytes (Plant et al., 2001). On the basis of these observations one would predict that remyelination following transplantation of OECs will be more extensive than following transplantation of Schwann cells in areas of demyelination where axons are surrounded by astrocytes. Whether this is the case has yet to be established. However, if it were, then OECs would become a strong candidate for use in cell-based therapies for demyelinating disease (Franklin, 2002).

Acknowledgments The authors would like to thank the contributions made by Andras Lakatos, Peter Smith, Clare Alexander, and Mark Dunning for the work described in this chapter originating in their own laboratories.

References Alexander, C. L., FitzGerald, U. F., and Barnett, S. C. (2002). IdentiWcation of growth factors that promote longterm proliferation of olfactory ensheathing cells and modulate their antigenic phenotype. Glia 37, 349–364. Au, W. W., Treloar, H. B., and Greer, C. A. (2002). Sublaminar organisation of the mouse olfactory bulb nerve layer. J. Comp. Neurol. 446, 68–80. Bailey, M. S., and Shipley, M. T. (1993). Astrocyte subtypes in the rat olfactory bulb: Morphological heterogeneity and diVerential laminar distribution. J. Comp. Neurol. 328, 501–526. Barber, P. C., and Dahl, D. (1987). Glial Wbrillary acidic protein (GFAP)-like immunoreactivity in normal and transected rat olfactory nerve. Exp. Brain Res. 65, 681–685. Barber, P. C., and Lindsay, R. M. (1982). Schwann cells of the olfactory nerves contain glial Wbrillary acidic protein and resemble astrocytes. Neurosci. 7, 3077–3099. Barnett, S. C., Alexander, C. L., Iwashita, Y., Gilson, J. M., Crowther, J., Clark, L., Dunn, L. T., Papanastassiou, V., Kennedy, P. G. E., and Franklin, R. J. M. (2000). IdentiWcation of a human olfactory ensheathing cell that can eVect transplant-mediated remyelination of demyelinated CNS axons. Brain 123, 1581–1588. Barnett, S. C., Hutchins, A.-M., and Noble, M. (1993). PuriWcation of olfactory nerve ensheathing cells from the olfactory bulb. Dev. Biol. 155, 337–350. Barnett, S. C., Thompson, R. J., Lakatos, A., and Pitts, J. (2001). Gap junctional communication and connexin expression in cultured olfactory ensheathing cells. J. Neurosci. Res. 65, 520–528. Baron-Van Evercooren, A., Avellana-Adalid, V., Ben Younes-ChennouW, A., Gansmu¨ller, A., Nait-Oumesmar, B., and Vignais, L. (1996). Cell-cell interactions during the migration of myelin-forming cells transplanted in the demyelinated spinal cord. Glia 16, 147–164. Blanes, T. (1898). Sobre algunos puntos dudosos de la estructura del bulbo olfactorio. Rev. Trim. Microg. 3, 99–127. Brunjes, P. C., and Frazier, L. L. (1986). Maturation and plasticity in the olfactory system of vertebrates. Brain Res. 396, 1–45. Bunge, M. B. (2001). Bridging areas of injury in the spinal cord. The Neuroscientist 7, 325–339. Burd, G. D. (1993). Morphological study of the eVects of intranasal zinc sulfate irrigation on the mouse olfactory epithelium and olfactory bulb. Microsc. Res. Tech. 24, 195–213. Caggiano, M., Kauer, J. S., and Hunter, D. D. (1994). Globose basal cells are neuronal progenitors in the olfactory epithelium: A lineage analysis using a replication-incompetent retrovirus. Neuron 13, 339–352. Calof, A. L., Hagiwara, N., Holcomb, J. D., Mumm, J. S., and Shou, J. Y. (1996). Neurogenesis and cell death in olfactory epithelium. J. Neurobiol. 30, 67–81.

II. GLIAL CELL DEVELOPMENT

381

382

14. OLFACTORY ENSHEATHING CELLS

Calof, A. L., Mumm, J. S., Rim, P. C., and Shou, J. (2000). The neuronal stem cell of the olfactory epithelium. J. Neurobiol. 36, 190–205. Carr, V. M., and Farbman, A. I. (1992). Ablation of the olfactory bulb up-regulates the rate of neurogenesis and induces precocious cell death in olfactory epithelium. Exp. Neurol. 115, 55–59. Chuah, M. I., and Au, C. (1991). Olfactory Schwann cells are derived from precursor cells in the olfactory epithelium. J. Neurosci. Res. 29, 172–180. Chuah, M. I., and Teague, R. (1999). Basic Wbroblast growth factor in the primary olfactory pathway: Mitogenic eVect on ensheathing cells. Neurosci. 88, 1043–1050. Chuah, M. I., Tennent, R., and Jacobs, I. (1995). Response of olfactory Schwann cells to intranasal zinc sulfate irrigation. J. Neurosci. Res. 42, 470–478. De Carlos, J. A., Lopez-Mascaraque, L., and Valverde, F. (1995). The telencephalic vesicles are innervated by olfactory placode-derived cells: A possible mechanism to induce neocortical development. Neurosci. 68, 1167–1178. Devon, R., and Doucette, R. (1992). Olfactory ensheathing cells myelinate dorsal root ganglion neurites. Brain Res. 589, 175–179. Devon, R., and Doucette, R. (1995). Olfactory ensheathing cells do not require L-ascorbic acid in vitro to assemble a basal lamina or to myelinate dorsal root ganglion neurites. Brain Res. 688, 223–229. Dickinson, P. J., GriYths, I. R., Barrie, J. M., Kyriakides, E., Pollock, G. F., and Barnett, S. C. (1997). Expression of the dm-20 isoform of the plp gene in olfactory nerve ensheathing cells: Evidence from developmental studies. J Neurocytol. 26, 181–189. Doetsch, F., and Alvarez-Buylla, A. (1996). Network of tangential pathways for neuronal migration in adult mammalian brain. Proc. Natl. Acad. Sci. USA 93, 14895–14900. Doucette, J. R. (1984). The glial cells in the nerve Wber layer of the rat olfactory bulb. Anat. Rec. 210, 385–391. Doucette, J. R., Kiernan, J. A., and Flumerfelt, B. A. (1983). The re-inervation of olfactory glomeruli following transection of primary olfactory axons in the central or peripheral nervous system. J. Anat. 137, 1–19. Doucette, R. (1989). Development of the nerve Wber layer in the olfactory bulb of mouse embryos. J. Comp. Neurol. 285, 514–527. Doucette, R. (1990). Glial inXuences on axonal growth in the primary olfactory system. Glia 3, 433–449. Doucette, R. (1991). PNS-CNS transitional zone of the Wrst cranial nerve. J. Comp. Neurol. 312, 451–466. Doucette, R. (1993). Glial cells in the nerve Wber layer of the main olfactory bulb of embryonic and adult mammals. Microsc. Res. Tech. 24, 113–130. Doucette, R., and Devon, R. (1994). Media that support the growth and diVerentiation of oligodendrocytes do not induce olfactory ensheathing cells to express a myelinating phenotype. Glia 10, 296–310. Doucette, R., and Devon, R. (1995). Elevated intracellular levels of cAMP induce olfactory ensheathing cells to express GAL-C and GFAP but not MBP. Glia 13, 130–140. Duncan, I. D., Hammang, J. P., Jackson, K. F., Wood, P. M., Bunge, R. P., and Langford, L. (1988). Transplantation of oligodendrocytes and Schwann cells into the spinal cord of the myelin-deWcient rat. J. Neurocytol. 17, 351–360. Duncan, I. D., and HoVman, R. L. (1997). Schwann cell invasion of the central nervous system of the myelin mutants. J. Anat. 190, 35–49. Eldridge, C. F., Bunge, M. B., and Bunge, R. P. (1989). DiVerentiation of axon-related Schwann cells in vitro: II. Control of myelin formation by basal lamina. J. Neurosci. 9, 625–638. Farbman, A. I. (1990). Olfactory neurogenesis: Genetic or environmental controls? Trends Neurosci. 13, 362–365. Fraher, J. P. (1982). The ultrastructure of sheath cells in developing rat vomeronasal nerve. J. Anat. 134, 149–168. Franceschini, I. A., and Barnett, S. C. (1996). Low aYnity NGF-receptor and E-N-CAM expression deWne two distinct types of olfactory nerve ensheathing cells that share a common lineage. Dev. Biol. 173, 327–434. Franklin, R. J. M. (2002). Remyelination of the demyelinated CNS: The case for and against transplantation of central, peripheral or olfactory glia. Brain Res. Bull. 57, 827–832. Franklin, R. J. M., and Barnett, S. C. (1997). Do olfactory glia have advantages over Schwann cells for CNS repair? J. Neurosci. Res. 50, 665–672. Franklin, R. J. M., and Barnett, S. C. (2000). Olfactory ensheathing cells and CNS regeneration: The sweet smell of success? Neuron 28, 15–18. Franklin, R. J. M., and Blakemore, W. F. (1993). Requirements for Schwann cell migration within CNS environments: A viewpoint. Int. J. Dev. Neurosci. 11, 641–649. Franklin, R. J. M., Gilson, J. M., Franceschini, I. A., and Barnett, S. C. (1996). Schwann cell-like myelination following transplantation of an olfactory bulb-ensheathing cell line into areas of demyelination in the adult CNS. Glia 17, 217–224. Ghirnikar, R. S., and Eng, L. F. (1994). Astrocyte-Schwann cell interactions in culture. Glia 11, 367–377. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J. (1994). Periaxin, a novel protein of myelinating Schwann cells with a possible role in axonal ensheathment. Neuron 12, 497–508. Goodman, M. N., Silver, J., and Jacobberger, J. W. (1993). Establishment and neurite outgrowth properties of neonatal and adult rat olfactory bulb glial cell lines. Brain Res. 619, 199–213. Gordon, M. K., Mumm, J. S., Davis, R. A., Holcomb, J. D., and Calof, A. L. (1995). Dynamics of MASH1 expression in vitro and in vivo suggest a non-stem cell site of MASH1 action in the olfactory receptor neuron lineage. Mol. Cell. Neurosci. 6, 363–379. Graziadei, G. A., and Graziadei, P. P. (1979). Neurogenesis and neuron regeneration in the olfactory system of mammals. II. Degeneration and reconstitution of the olfactory sensory neurons after axotomy. J Neurocytol. 8, 197–213.

II. GLIAL CELL DEVELOPMENT

OEC INTERACTIONS WITH ASTROCYTES AND THEIR POTENTIAL CLINICAL APPLICATIONS

Graziadei, P. P., and Monti Graziadei, G. A. (1980). Neurogenesis and neuron regeneration in the olfactory system of mammals. III. DeaVerentation and reinnervation of the olfactory bulb following section of the Wla olfactoria in rat. J Neurocytol. 9, 145–162. Graziadei, P. P., and Monti Graziadei, G. A. (1983). Regeneration in the olfactory system of vertebrates. Am. J Otolaryngol. 4, 228–233. Harding, J. W., Getchell, T. V., and Margolis, F. L. (1978). Denervation of the primary olfactory pathway in mice. V. Long-term eVect of intranasal ZnSO4 irrigation on behavior, biochemistry and morphology. Brain Res. 140, 271–285. Heredia, M., Gascuel, J., Ramo´n-Cueto, A., Santacana, M., Avila, J., Masson, C., and Valverde, F. (1998). Two novel monoclonal antibodies (1.9.E and 4.11.C) against olfactory bulb ensheathing glia. Glia 24, 352–364. Hinds, J. W., Hinds, P. L., and McNelly, N. A. (1984). An autoradiographic study of the mouse olfactory epithelium: Evidence for long-lived receptors. Anat. Rec. 210, 375–383. Holcomb, J. D., Mumm, J. S., and Calof, A. L. (1995). Apoptosis in the neuronal lineage of the mouse olfactory epithelium: Regulation in vivo and in vitro. Dev. Biol. 172, 307–323. Imaizumi, T., Lankford, K. L., Burton, W. V., Fodor, W. L., and Kocsis, J. D. (2000). Xenotransplantation of transgenic pig olfactory enseathing cells promotes axonal regeneration in rat spinal cord. Nature Biotech. 18, 949–953. Imaizumi, T., Lankford, K. L., Waxman, S. G., Greer, C. A., and Kocsis, J. D. (1998). Transplanted olfactory ensheathing cells remyelinate and enhance axonal conduction in the demyelinated dorsal columns of the rat spinal cord. J. Neurosci. 18, 6176–6185. Kato, T., Honmou, O., Uede, T., Hashi, K., and Kocsis, J. D. (2000). Transplantation of human olfactory ensheathing cells elicits remyelination of demyelinated rat spinal cord. Glia 30, 209–218. Lakatos, A., Franklin, R. J. M., and Barnett, S. C. (2000). Olfactory ensheathing cells and Schwann cells diVer in their in vitro interactions with astrocytes. Glia 32, 214–225. Lakatos, A., Smith, P. M., Barnett, S. C., and Franklin, R. J. M. (2003). Meningeal cells enhance limited CNS remyelination by transplanted olfactory ensheathing cells. Brain, in press. Le Roux, P. D., and Reh, T. A. (1994). Regional diVerences in glial-derived factors that promote dendritic outgrowth from mouse cortical neurons in vitro. J. Neurosci. 14, 4639–4655. Lemke, G., and Axel, R. (1985). Isolation and sequence of a cDNA encoding the major structural protein of peripheral myelin. Cell 40, 501–508. Li, Y., Field, P. M., and Raisman, G. (1998). Regeneration of adult rat corticospinal axons induced by transplanted olfactory ensheathing cells. J. Neurosci. 18, 10514–10524. Liuzzi, F. J., and Lasek, R. J. (1987). Astrocytes block axonal regeneration in mammals by activating the physiological stop pathway. Science 237, 642–645. Lois, C., and Alvarez-Buylla, A. (1994). Long-distance neuronal migration in the adult mammalian brain. Science 264, 1145–1148. Mackay-Sim, A., and Kittel, P. (1991a). Cell dynamics in the adult mouse olfactory epithelium: A quantitative autoradiographic study. J. Neurosci. 11, 979–984. Mackay-Sim, A., and Kittel, P. W. (1991b). On the life span of olfactory receptor neurons. Eur. J. Neurosci. 3, 209–215. Marin-Padilla, M., and Amieva, M. R. (1989). Early neurogenesis of the mouse olfactory nerve: Golgi and electron microscopic studies. J. Comp. Neurol. 288, 339–352. Miragall, F., Kadman, G., and Schachner, M. (1989). Expression of L1 and N-CAM cell adhesion molecules during development of the mouse olfactory system. Dev. Biol. 135, 272–286. Monti Graziadei, G. A., Karlan, M. S., Bernstein, J. J., and Graziadei, P. P. (1980). Reinnervation of the olfactory bulb after section of the olfactory nerve in monkey (Saimiri sciureus). Brain Res. 189, 343–354. Mumm, J. S., Shou, J., and Calof, A. L. (1996). Colony-forming progenitors from mouse olfactory epithelium: Evidence for feedback regulation of neuron production. Proc. Natl. Acad. Sci. USA 93, 11167–11172. Norgren, R. B., Jr., Ratner, N., and Brackenbury, R. (1992). Development of olfactory nerve glia deWned by a monoclonal antibody speciWc for Schwann cells. Dev. Dyn. 194, 231–238. Obremski, V. J., Johnson, M. I., and Bunge, M. B. (1993). Fibroblasts are required for Schwann cell basal lamina deposition and ensheathment of unmyelinated sympathetic neurites in culture. J. Neurocytol. 22, 102–117. Parmantier, E., Lynn, B., Lawson, D., Turmaine, M., Namini, S. S., Chakrabarti, L., McMahon, A. P., Jessen, K. R., and Mirsky, R. (1999). Schwann cell-derived Desert hedgehog controls the development of peripheral nerve sheaths. Neuron 23, 713–724. Pixley, S. K. (1992). The olfactory nerve contains two populations of glia, identiWed both in vivo and in vitro. Glia 5, 269–284. Plant, G. W., Bates, M. L., and Bunge, M. B. (2001). Inhibitory proteoglycan immunoreactivity is higher at the caudal than the rostral Schwann cell graft-transected spinal cord interface. Mol. Cell. Neurosci. 17, 471–487. Plant, G. W., Currier, P. F., Cuervo, E. P., Bates, M. L., Pressman, Y., Bunge, M. B., and Wood, P. M. (2002). PuriWed adult ensheathing glia fail to myelinate axons under culture conditions that enable Schwann cells to form myelin. J. Neurosci. 22, 6083–6091. Pollock, G. S., Franceschini, I. A., Graham, G., Marchionni, M. A., and Barnett, S. C. (1999). Neuregulin is a mitogen and survival factor for olfactory bulb ensheathing cells and an isoform is produced by astrocytes. Eur. J. Neurosci. 11, 769–780. Raisman, G. (1985). Specialized neuroglial arrangement may explain the capacity of vomeronasal axons to reinnervate central neurons. Neurosci. 14, 237–254.

II. GLIAL CELL DEVELOPMENT

383

384

14. OLFACTORY ENSHEATHING CELLS

Raisman, G. (2001). Olfactory ensheathing cells—another miracle cure for spinal cord injury? Nat. Rev. Neurosci. 2, 369–374. Ramon-Cueto, A., Cordero, M. I., Santos-Benito, F. F., and Avila, J. (2000). Functional recovery of paraplegic rats and motor axon regeneration in their spinal cords by olfactory ensheathing glia. Neuron 25, 425–435. Ramo´n-Cueto, A., Plant, G. W., Avila, J., and Bunge, M. B. (1998). Long distance axonal regeneration in the transected adult rat spinal cord is promoted by olfactory ensheathing glia transplants. J. Neurosci. 18, 3802–3815. Ramo´n-Cueto, A., and Valverde, F. (1995). Olfactory bulb ensheathing glia: A unique cell type with axonal growth-promoting properties. Glia 14, 163–173. Schwartz-Levey, M., Chikaraishi, D. M., and Kauer, J. S. (1991). Characterization of potential precursor populations in the mouse olfactory epithelium using immunocytochemistry and autoradiography. J. Neurosci. 11, 3556–3564. Schwob, J. E. (2002). Neural regeneration and the peripheral olfactory system. Anat. Rec. 269, 33–49. Shields, S. A., Blakemore, W. F., and Franklin, R. J. M. (2000). Schwann cell remyelination is restricted to astrocyte-deWcient areas following transplantation into demyelinated adult rat brain. J. Neurosci. Res. 60, 571–578. Smith, P. M., Lakatos, A., Barnett, S. C., JeVery, N. D., and Franklin, R. J. M. (2002). Cryopreserved cells isolated from the adult canine olfactory bulb are capable of extensive remyelination following transplantation into the adult rat CNS. Exp. Neurol. 176, 402–406. Smith, P. M., Sim, F. J., Barnett, S. C., and Franklin, R. J. M. (2001). SCIP/Oct-6, Krox-20 and desert hedgehog mRNA expression during CNS remyelination by transplanted olfactory ensheathing cells. Glia 36, 342–353. Sonigra, R. J., Brighton, P. C., Jacoby, J., Hall, S., and Wigley, C. B. (1999). Adult rat olfactory nerve ensheathing cells are eVective promoters of adult central nervous system neurite outgrowth in coculture. Glia 25, 256–269. Takami, T., Oudega, M., Bates, M. L., Wood, P. M., Kleitman, N., and Bunge, M. B. (2002). Schwann cell but not olfactory ensheathing glia transplants improve hindlimb locomotor performance in the moderately contused adult rat thoracic spinal cord. J. Neurosci. 22, 6670–6681. Topilko, P., and Meijer, D. (2001). Transcription factors that control Schwann cell development and myelination. In ‘‘Glial Cell Development’’ (K. R. Jessen and W. D. Richardson, eds.), pp. 223–244. Oxford University Press, Oxford. Ubink, R., Halasz, N., Zhang, X., Dagerlind, A., and Hokfelt, T. (1994). Neuropeptide tyrosine is expressed in ensheathing cells around the olfactory nerves in the rat olfactory bulb. Neurosci. 60, 709–726. Valverde, F., and Lopez-Mascaraque, L. (1991). Neuroglial arrangements in the olfactory glomeruli of the hedgehog. J. Comp. Neurol. 307, 658–674. Valverde, F., Santacana, M., and Heredia, M. (1992). Formation of an olfactory glomerulus: Morphological aspects of development and organization. Neurosci. 49, 255–275. Wewetzer, K., Verdu, E., Angelov, D. N., and Navarro, X. (2002). Olfactory ensheathing glia and Schwann cells: Two of a kind? Cell Tissue Res. 309, 337–345. Wilby, M. J., Muir, E. M., Fok-Seang, J., Gour, B. J., Blaschuk, O. W., and Fawcett, J. W. (1999). N-cadherin inhibits Schwann cell migration on astrocytes. Mol. Cell. Neurosci. 14, 66–84. Wiley, T. J. (1973). The ultrastructure of the cat olfactory bulb. J. Comp. Neurol. 152, 211–242. Yan, H., Bunge, M. B., Wood, P. M., and Plant, G. W. (2001). Mitogenic response of adult rat olfactory ensheathing glia to four growth factors. Glia 33, 334–342.

II. GLIAL CELL DEVELOPMENT

C H A P T E R

15 Myelin Basic Protein Gene Anthony T. Campagnoni and Celia W. Campagnoni

INTRODUCTION The major myelin proteins—that is, the myelin basic proteins (MBPs) and the myelin proteolipid proteins (PLPs)—have been the subjects of intense investigation for almost half a century. They have proven to be of interest to a wide variety of biologists including the following: Neurobiologists interested in the function of myelin in the nervous system. Neuropathologists and neurologists interested in the many demyelinating diseases that aZict the human population. Immunologists interested in autoimmune mechanisms because these major myelin proteins induce autoimmune responses in susceptible animals, and they have been used to generate animal models of autoimmune inXammatory diseases, such as multiple sclerosis (MS). Developmental and cellular neurobiologists, because the MBPs and PLPs are markers of end-stage oligodendrocyte (OL) development and are critical for the terminal diVerentiated function of OLs (i.e., the elaboration of the myelin sheath). Molecular neurobiologists, because of the complexity of their gene products, the role the genes may play in neurological diseases, and the multiple functions of these proteins within and without the nervous system. Mutations in the genes for these proteins also account for a number of the dysmyelinating animal mutants that serve as models of human diseases. The myelin basic proteins (MBP) and myelin proteolipid proteins (PLPs) are among the most abundant proteins in the central nervous system. For example, myelin constitutes approximately 25 to 35% of the dry weight of the brain and these major proteins account for at least 80% of the protein found in myelin. So it is not surprising that the structures of the genes encoding these proteins were among the Wrst to be elucidated in the nervous system (Diehl et al., 1986; Milner et al., 1985; Roach et al., 1985; Takahashi et al., 1985;). The myelin basic protein was originally isolated in biochemically pure form from white matter in the late 1960s and the amino acid sequences from several species were determined in the late 1960s and early 1970s (Carnegie, 1971; Carnegie et al., 1974; Eylar, 1970; Kibler et al., 1969). It was then established that the protein existed in multiple isoforms in rodents (Barbarese et al., 1977; Martenson et al., 1971, 1972). Later it was found that these isoforms were produced through the translation of separate mRNAs (Yu and Campagnoni, 1982), which were generated through alternative splicing of the gene (Aruga et al., 1991; de Ferra et al., 1985; Kamholz et al., 1986; Newman et al., 1987; Roth et al., 1987; Takahashi et al., 1985). These initial molecular genetic studies assigned a size of ~35 Kb to the murine MBP gene. It mapped to chromosome 18 in the mouse and the human (Roach et al., 1985; Sparkes et al., 1987).

Myelin Biology and Disorders, Volume 1

387

Copyright 2004, Elsevier (USA). All rights reserved.

388

15. MYELIN BASIC PROTEIN GENE

Because the MBPs are extremely abundant in the CNS (Eng et al., 1968; Norton and Cammer, 1984), it took some time to fully appreciate their heterogeneity because of the dominance of the major isoforms and their mRNAs in the brain. Myelin is produced from OLs in the central nervous system, so expression of the MBP gene is extremely high, particularly during the myelinating period in the brain. During this period, expression of the gene eclipses that of most other genes in the OL and can even mask expression in other cells in the brain by most techniques. IdentiWcation of other products of the MBP gene some years ago suggested a broader range of functions for the MBP gene, and recent molecular, cellular, and biochemical studies have begun to provide stronger evidence for multiple functions for this gene. The literature on the myelin protein genes, their products, expression and genetic mutants is quite large and several reviews have appeared on their structure, expression, and role in disease (Campagnoni and SkoV, 2001; GriYths, 1996; Ikenaka and Kagawa, 1995; Jetten and Suter, 2000; Johns and Bernard, 1999; Kamholz et al., 1999; Mirsky et al., 2001; Staugaitis et al., 1996; Wegner, 2000; Yoshikawa, 2001). This review focuses on the MBP gene, the discovery of newly identiWed gene transcripts of the MBP gene, and alternative functions for the ‘‘new’’ and ‘‘old’’ products of the MBP gene.

THE MYELIN BASIC PROTEIN GENE AND ITS PRODUCTS The MBP gene is ~105 Kb in length in the mouse and ~180 Kb in humans (Campagnoni et al., 1993; Pribyl et al., 1993). The gene contains three transcription start sites, which give rise to three sets of mRNAs that encode two families of proteins. The classic MBP mRNAs are generated primarily from transcription start site 3 (see Fig. 15.1), which is extremely active in myelin-forming cells (i.e., OLs in the CNS and Schwann cells in the PNS). To date, only one mRNA has been identiWed that is derived from the second transcription start site (Kitamura et al., 1990). This mRNA, called the M41 MBP mRNA encodes the classic 14 kDa MBP. It is initiated at transcription start site 2 and it has a longer 5’ untranslated region than its 14 kDa MBP mRNA counterpart generated from transcription start site 3, so it may not be translated as eYciently. The steady-state levels of the M41 MBP mRNA in brain are low, and the possibility that it is translated less eYciently suggests that it does not contribute signiWcantly to the 14 kDa MBP population in the brain. Interestingly, expression of this mRNA has been detected in the murine spleen (C. Campagnoni, K. Kampf, and A. Campagnoni, unpublished observations) by RT-PCR. The expression of low levels of the classic MBPs has been observed in macrophages in spleen and lymph nodes system by Western blot analysis (Liu et al., 2001). The second family of proteins encoded by the MBP gene is the golli (-MBP) proteins. These proteins are produced from the Wrst transcription start site of the gene. In the nervous system the steady-state levels of golli mRNAs are signiWcantly lower than those of the classic MBP mRNAs (Campagnoni et al., 1993). Golli proteins are also expressed throughout the immune system, including the thymus, spleen, and lymph nodes; and the expression levels of their mRNAs and proteins appear to be equivalent to that in the brain. Expression of the golli family of MBP proteins is similar in brain and thymus, but expression of the classic MBPs is substantially higher than golli in the nervous system and substantially lower than golli in the immune system.

Structure and Splicing of the MBP Gene A partial structure of the MBP gene was among the Wrst of the myelin protein genes to be determined (Roach et al., 1985). It was identiWed as the ~32 Kb of the murine gene extending from transcription start site 3 downstream to exon 11 and its exon structure accounted for all the known classic MBPs at that time (see Fig. 15.1). The MBP gene was found to map to human and mouse chromosomes 18 (Roach et al., 1985; Sparkes et al., 1987). In the rat it has been reported to map to chromosome 1 by Koizumi et al. (1991) and

389

THE MYELIN BASIC PROTEIN GENE AND ITS PRODUCTS

Golli-MBPs Golli transcription start site

M41 M41 MBP MBP Classic Classic MBPs MBPs MBP MBP transcription transcription start start sites sites

tss1

tss2 tss2

tss3 tss3

AB

2

3

4

1 2 3

J37 mRNA

AB

5

golli splicing (tss1)

6

7 8 9 10

11

classic classic MBP MBP splicing splicing (tss2 (tss2 & & tss3) tss3)

5

C

A B

BG21 mRNA

C

1 2 3 5 7 8 AB

1 2 3 7 8

4 5 78 9 11

B

11

10

classic classic MBP MBP mRNA mRNA family family

A

M41 mRNA

TP8 mRNA

classic classic MBP MBP ss (133 aa)

BG21 protein

golli domain (133 aa)

J37 protein

TP8 protein

golli domain (47aa) golli domain

(56 aa) MBP domain

14 kDa 17.2 & 17.3 kDa

(116 aa)

18.5 kDa

MBP domain

(22 aa)

20.2 kDa 21.5 kDa

FIGURE 15.1 Diagrammatic representation of the structure and principal products of the murine myelin basic protein gene. The gene contains three transcription start sites (tss1, 2, 3). The Wrst transcription start site generates the golli mRNAs, and the second and third transcription start sites generate the classic MBPs. The golli products all share in common a 47 amino acid N-terminal sequence encoded by exons unique to the golli mRNAs. The two principal golli products are the BG21 and J37 isoforms, which contain varying lengths of classic MBP sequence. The minor golli product, TP8, contains no classic MBP sequence since its splicing patterns is out of frame for classic MBP. The principal classic MBP isoforms are shown, but other minor isoforms have been identiWed. The steady state levels of the M41 MBP mRNA derived from tss2 suggest that it contributes in a minor fashion to the total 14 kDa classic MBP population present in the CNS.

to chromosome 18 by Goldner-Sauve et al. (1991). Elucidation of the complete structure of the MBP gene in mouse and in human (Campagnoni et al., 1993; Pribyl et al., 1993) deWned the origin of the two families of MBP gene transcripts (i.e., golli and ‘‘classic’’ MBPs) and the gene has often been referred to as the golli-MBP to emphasize the two sets of products generated by the gene. The splicing pattern of the MBP gene is somewhat complex with respect to the numbers of products expressed as well as their developmental regulation. In the mouse, the gene contains three transcription start sites (at exons 1, 4, and 5B in Fig. 15.1), each of which becomes active at diVerent times during pre- and post-natal brain development. Transcription start site 1 expresses several golli RNA splice products, which contain exons 1, 2, and 3 of the gene spliced into MBP-encoding exons (see Fig. 15.1). The second start site at exon 4 (Kitamura et al., 1990) and the third transcription start site at exon 5B express MBP mRNAs. The expression of at least six alternatively spliced forms of MBP mRNAs generated from the major (third) transcription start site has been documented (see Aruga et al., 1991). These alternatively spliced MBP mRNAs encode MBP isoforms that range in molecular mass from 14 kDa to 21 kDa, and the onset and proportion of the isoforms is developmentally regulated in the brain. For example, those proteins containing exon 6 (i.e., the 17 and 21.5 kDa isoforms) appear early in myelination, and their proportions relative to the 14 and 18.5 kDa isoforms decline with development in the mouse (Barbarese et al., 1978; Jordan et al, 1989). A pattern of the early appearance of exon 6-containing transcripts prior to the major transcripts has also been observed in viral-induced demyelination

390

15. MYELIN BASIC PROTEIN GENE

(Jordan et al., 1990) and in EAE (Nagasato et al., 1997), and it has been interpreted to indicate endogenous attempts to repair the lesions in the aVected areas. The human gene has been cloned and its organization is similar to that of the mouse gene except that it is much larger, approximately 179 kb (Pribyl et al., 1993). It also expresses alternatively spliced products, some of which are similar but not identical in structure to the mouse products (Pribyl et al., 1996).

Multiple Promoters of the MBP Gene There are three transcription start sites in the MBP gene indicating the presence of three promoters of the gene. The Wrst and third promoters have been studied most extensively, and both have been found to be of great value in targeting transgenes with tissue and cellular speciWcity. The most extensively studied has been the promoter of the classic MBPs (i.e., regulating tss3). This promoter includes exon 5A, which can itself serve as a promoter element to target expression of reporter genes in cell lines (Devine-Beach et al., 1990; Miura et al., 1989) and in transgenic mice where it targets transgenes to OLs with remarkable cellular and developmental speciWcity. Sequences upstream of tss3 containing intron 4 þ exon 5A have been used to drive the expression of many transgenes in OLs in transgenic mice, including oncogenes (Hayes et al., 1992), MHC (major histocompatibility complex) molecules (Turnley et al., 1991), tumor necrosis factor alpha (Taupin et al., 1997), and reporter genes such as lac z (Gow et al., 1992; Miskimins et al., 1992; Vanderluit et al., 2000). Sequences within intron 4 have been reported to modulate expression of the classic MBPs in Schwann cells in vivo (Forghani et al., 2001). Downstream regions of intron 4 þ exon 5A also can be used to drive expression of classic MBP in shiverer mice (which have a deletion of exons 7–11) with a subsequent reduction in the clinical signs of the neurological disorder caused by the mutation (Readhead et al., 1987). Thus, the region upstream of tss3 of the MBP gene, including parts of intron 4 and exon 5A, has been a useful element for targeting transgenes to oligodendrocytes and, in some cases, to Schwann cells with excellent developmental and cellular speciWcity. Transcriptional Regulatory Elements in the Classic MBP Promoters (tss3) IdentiWcation of the cis genetic elements and the trans-acting factors regulating the tissue, cell, and developmental speciWcity of the MBP gene has drawn the attention of several laboratories in recent years. It has generally been felt that the regulation of the gene is complex, involving multiple sites within the region upstream of the MBP transcription start site and encompassing exon 5A of the golli-mbp gene, where trans-acting factors that bind to diVerent parts of the promoter interact to regulate the activity of the MBP transcription unit (Mikoshiba et al., 1991; see Wegner, 2000, for recent review). Accordingly, this region has been isolated, analyzed, and used to deWne putative regulatory elements. The MBP promoter does not contain conventional TATA or CAAT boxes, that is, regulatory sequences common to all promoters and believed to be involved in the binding of general transcription factors (Miura et al., 1989; Takahashi et al, 1985). A number of sequences have been reported to confer either tissue or cellular speciWcity to the expression of the gene (Devine-Beach et al., 1990; Tamura et al., 1989, 1991). Several groups have searched for brain-speciWc factors that regulate the transcription of the classic MBP mRNAs. Khalili and his coworkers have concentrated on a GC rich region between -14 and -50. They have isolated three proteins that bind DNA: Pur alpha, which in conjunction with Sp1, enhances transcription (Tretiakova et al 1999); MyEF-2, which represses transcription (Haas et al 1995); and MyEF-3 (Steplewiski et al 1998). Wrabetz et al. (1993) reported that a NF1 site between -149 and -102 repressed the expression of the classic MBP promoter in other cell types, but not in oligodendroglia. They have isolated MEBA (myelinating glial enriched binding activity), which consists of at least two proteins that bind to the promoter just upstream of, and overlap, the NF1 site and activate the MBP promoter in oligos but not Cos cells (Taveggia et al., 1998). Recently Clark et al. (2002) have reported that the NF1 site is essential for the cAMP induction of the classic MBP promoter.

THE MYELIN BASIC PROTEIN GENE AND ITS PRODUCTS

The regulation of the thyroid response element between -163 and -183 (Farsetti et al. 1991) has been examined by two groups (Farsetti et al., 1997, and Jeanin et al., 1998). Huang et al. (2002) have identiWed a NFkB site between -570 and -562 that is responsible for the stimulation of the MBP promoter by TNFa. Anderson and Miskimins (1994) identiWed a PKC-responsive element at the distal end of a 1.3 Kb MBP promoter fragment, which was responsible for the inhibition of cAMP activation of the promoter. Mutation of an AP-1 like site between -1240 and -1230 results in an increase in reported activity in CG4 cells in diVerentiation media. (Miskimins and Miskimins, 2001). While the 1.3 Kb upstream region of the classic MBP transcription site is able to confer the correct developmental expression of the MBP gene in oligodendrocytes, it does not seem to contain those elements that completely restrict its expression to neural tissue (Asipu et al., 2001; Turnley et al., 1991; Yoshioka et al., 1991). A reporter construct containing 1.9 Kb downstream of tss3 is reported to be expressed solely in oligodendrocytes in transgenic mice (Gow et al., 1992). Schwann cell expression could be conferred by a 0.6 Kb fragment that is 9 Kb upstream from the third transcription initiation site (Forghani et al., 2001). Transcriptional Activity of the Golli Promoter (tss1) and the Second Classic MBP Promoter (tss2) Very little is known about the regulation of the Wrst (golli, tss1) or second (tss2) promoters of the MBP gene. Based on the steady-state levels of the M41-MBP mRNA, which encodes the classic 14 kD MBP, it would appear that the classic MBP promoter at tss2 is substantially less active than the major classic MBP promoter at tss3 in the nervous system. Our unpublished data suggest that the tss2 promoter appears to be active in the spleen as well as the nervous system. A fragment consisting of approximately 1.1 kb upstream þ 0.2 kb downstream of tss1 has been used to drive several transgenes in transgenic mice (Landry et al., 1998; Xie et al., 2002). While not as speciWc as the classic MBP promoter, this fragment appears to target expression to a limited number of neuronal populations, including cortical preplate neurons, cortical subplate neurons, olfactory neurons, and neurons within the dorsal root ganglia. It does not target expression to the immune system or to glia, such as oligodendrocytes. The elements that regulate expression in these cells and tissues have not yet been identiWed. A recent report (Givogri et al., 2001) suggests that in mixed glial cultures and in the OL cell line, N20.1, bFGF, retinoic acid and NT-3 can alter the steady state levels of golli mRNAs, presumably through transcriptional regulation. In these in vitro systems, bFGF and retinoic acid appear to down-regulate golli expression and NT-3 appears to increase the levels of golli transcripts. In a study attempting to explain the low activity of the golli promoter containing ~6.4 kb upstream of tss1 compared to the proximal golli promoter of ~1.3 kb in neural cell lines, Givogri et al. (2000) identiWed a silencer element between
2.7 and
3.0 kb upstream of tss1. This silencer was responsible for a signiWcant inhibition of reporter gene expression in PC12 cells.

The Classic Myelin Basic Proteins The primary cell biological function of the classic MBPs is to maintain the structure of the myelin sheath. There is evidence, however, that a single classic MBP isoform alone is capable of fulWlling this function in the absence of the other isoforms (Katsuki et al, 1988; Kimura et al., 1998). This makes it unclear why the multiple isoforms are needed to maintain myelin structure. It has been observed that the classic MBP isoforms containing the sequence encoded by exon 6 may be transported to the nucleus in vitro and in vivo during early stages of classic MBP expression (Hardy et al., 1996; Pedraza et al., 1997). This translocation appears to occur by active transport (Pedraza et al., 1997). It is not yet clear what role the classic MBPs may play in the nucleus, although it has been speculated that it may have some regulatory function in implementing the myelination program (Pedraza et al., 1997).

391

392

15. MYELIN BASIC PROTEIN GENE

Phosphorylation of the Myelin Basic Proteins The fact that the classic myelin basic proteins are phosphorylated has been known for decades, although the reasons for this phosphorylation are far from clear (Carnegie et al., 1973; 1974; Steck and Appel, 1974). The classic MBPs are excellent substrates for a number of phosphokinases, including protein kinase C, ERK1, and ERK2 (extracellular signalregulated protein kinase 1 and 2), as well as Caþ2/calmodulin-dependent protein kinase 2. The phosphorylation sites for these kinases on the MBP have been identiWed (Martenson et al., 1983, Shoji et al., 1987). Phosphorylation of the classic MBPs has been found to alter their interactions with lipids (Brady et al., 1985; Cheifetz and Moscarello, 1985). This presumably might play a role in altering interactions within the myelin membrane, thereby aVecting its structure. However, neither a given isoform nor a given phosphorylation site is uniformly phosphorylated in the MBPs. There is some evidence for a possible role of phosphorylated classic MBPs in the early stages of myelin formation. During development, classic MBP phosphorylation begins early in the myelination process in vivo (Ulmer and Braun, 1983), and it has been found to be associated with process extension and sheet elaboration of oligodendrocytes in vitro (Vartanian et al., 1986). When puriWed adult oligodendrocytes adhere to a substratum prior to process extension and sheet elaboration, the classic MBPs become phosphorylated through a mechanism involving a substratum-oligodendrocyte induced activation of protein kinase C (Vartanian et al., 1986). A decrease in classic MBP phosphorylation has been found to be associated with a cascade of morphological events induced in cultured oligodendrocytes by treatment with anti-galactocerebroside in vitro (Dyer et al., 1994). This treatment causes a redistribution of membrane surface GalC over internal domains of MBP and the loss of microtubular structures within the sheets (Dyer and Benjamins, 1988, 1989). These authors have presented evidence that classic MBPs may mediate a signaling pathway leading to these events in this system. This would suggest, then, that phosphorylation of classic MBP might play a role in signaling processes related to myelination in the oligodendrocyte. Are the Classic MBPs More Than Just Structural Proteins? Several laboratories have presented evidence of an association of classic MBPs with the cytoskeleton and microtubules, at least in cultured oligodendrocytes (Dyer and Benjamins, 1989; Dyer et al., 1994, 1997; Wilson and Brophy, 1989). Dyer et al. (1994; 1997) have suggested that MBP gene products may mediate extracellular signals that regulate the organization, stability, and cytoskeletal network of oligodendrocytes. In this regard, Harauz and colleagues have examined a number of physical properties and structural motifs in recombinant classic MBPs related to a potential role for MBP in signaling or cytoskeletal rearrangements (Harauz et al., 2000). They have identiWed several structural motifs in the MBP that are similar to MARCKS [myristoylated alanine-rich C kinase substrate]. For example, MBP and MARCKS have similar extended conformations in the absence or presence of bound lipids, they bind to actin, they have pleckstrin homology (i.e., bind phosphatidyl inositides), and MBP has sequence homology to the lipid eVector region of MARCKS. Furthermore, like MARCKS, a number of studies have shown that classic MBP can bind to Caþ2-calmodulin by several methods (Boggs and Rangaraj, 2000; Harauz et al., 2000; Libich and Harauz, 2002a, 2002b). MARCKS is an important downstream target of PKC. In fact, it has been suggested to be the major PKC substrate in OL progenitors (Baron et al., 1999; Bhat, 1995). MARCKS is a member of a family of molecules, which includes GAP43, and can aVect the actin cytoskeleton. It has been implicated in a number of biological processes, including regulation of brain development and cell migration and adhesion (Arbuzova et al., 2002). Phosphorylation of MARCKS by PKC, or by binding of calmodulin, renders it inactive. In its active, unphosphorylated form, MARCKS causes cross-linking of actin Wlaments, which leads to cytoskeletal rearrangements, which presumably promote process extension in neural cells. Establishing a clear link between the classic MBPs and signaling processes or cytoskeletal changes has been suggested but, as yet, unproven. Nonetheless, indirect evidence of

GOLLI PROTEINS ARE LOCALIZED WITHIN FIBERS AND NUCLEI OF CELLS

such a link has continued to accumulate. The classic MBPs can be phosphorylated by PKC and the level of their phosphorylation can change as a function of development and of extracellular signaling events. They share many features in common with MARCKS, an established signaling molecule involved in cytoskeletal rearrangements. The classic MBPs have been shown to be associated with cytoskeletal elements in cultured oligodendrocytes, and cytoskeletal changes are noted in the shiverer mouse in which classic MBP expression is lost.

The Golli-MBPs While the primary, but perhaps not exclusive, biological function of the classic MBPs appears to be that of myelin structural components, the biological function of the golli proteins is only beginning to be understood. Based on their subcellular localization, one role golli proteins do not appear to have is that of structural proteins of myelin. In the CNS, expression of golli proteins is generally developmentally regulated (Campagnoni et al., 1993; Landry et al., 1996). Their expression tends to be higher during embryonic development and declines with age (Landry et al., 1996; 1998). This developmental regulation does not appear to occur in the PNS, in the thymus, or in some brain regions, such as the olfactory system (Landry et al., 1997; Feng et al., 2000). Other than in cells of the nervous system and the immune system, golli immunoreactivity in the periphery appears to result from expression in nerves innervating the tissues (Landry et al., 1996; Feng et al., 2002a). In the nervous system, golli proteins are expressed in oligodendrocyes and in speciWc subsets of neurons, although recently golli expression has been reported in activated microglia, macrophages, and adult oligodendrocyte progenitors around MS lesions (Filipovic et al., 2002). This increased expression in response to an activated immune system has also been noted in macrophages in lymph nodes of mice during the relapsing phase of EAE (MacKenzie-Graham et al., 1997).

GOLLI PROTEINS ARE LOCALIZED WITHIN FIBERS AND NUCLEI OF CELLS Golli is localized within the cytoplasm and process extensions of oligodendrocytes and neurons, and it is often Wrst evident in these cells when they begin to migrate and extend processes (Jacobs et al., 2000; Landry et al., 1997, 1998; Pribyl et al., 1996). For example, within the human lateral funiculus, oligodendrocytes immunostain for golli protein and can be seen along Wber tracts. In these cells, golli protein is distributed throughout the oligodendrocyte cell bodies, nuclei, and the processes connecting the cells to the myelin sheath as they envelope axonal segments. Immunohistochemistry of the human embryonic spinal cord with golli antibody also reveals oligodendrocytes in the earliest stages of elaborating a myelin sheath. The protein localizes to the cell processes and terminal sheet-like structures (Pribyl et al., 1996). Although not appreciated by us at the time, this is reminiscent of more recent in vitro studies localizing golli to the growing tips of OL cell lines induced to elaborate processes by overexpression of golli (S. Reyes and A. Campagnoni, unpublished results). In vitro transfection of golli proteins, or indeed the golli domain alone, into oligodendrocyte cell lines can induce the N19 and N20.1 OL cell lines to elaborate extensive processes and membrane sheets, adopting a morphology similar to oligodendrocytes in culture (Reyes and Campagnoni, 2002). Similarly, transfection of golli into neuronal cell lines results in signiWcant neurite extension, and in PC12 cells it induces neurite outgrowth (Reyes and Campagnoni, 2002). These results complement the known expression of these proteins in vivo in a number of neuronal populations and in oligodendrocyte precursors during process extension and migration. A recent paper by Filipovic et al. (2002) has shown that golli immunoreactivity is high in adult oligodendrocyte progenitors surrounding MS plaques preparatory to remyelination attempts. All these results combine to suggest a role for golli proteins in process extension and myelin elaboration.

393

394

15. MYELIN BASIC PROTEIN GENE

The golli proteins are also localized in the nucleus of many cells, and this localization is developmentally sensitive in a number of neuronal populations. For example, in cerebellar granule cells in the external granule cell layer, golli is localized in cell processes, but after their migration into the inner granule cell layer there is a shift in golli localization to the nucleus (Landry et al., 1996). The site responsible for this translocation is located within the MBP domain of the golli proteins (Reyes and Campagnoni, 2002). In other developing systems, golli protein expression is Wrst seen within the nucleus or cell body of neurons and then later is found only in the Wbers emanating from these neurons. Two examples of this include neurons within the neostriatum (Landry et al., 1996) and cortical subplate (Landry et al., 1998). This behavior is of interest because many cell signaling molecules shuttle between the nucleus and the cytoplasm, raising the possibility that golli might play a role in intracellular signaling.

A Role for Golli Proteins in the Nucleus A molecular partner for golli proteins, called GIP (golli-interacting protein) has been isolated by a yeast two-hybrid approach, which suggests that, in the nucleus, golli may be part of a transcriptional regulatory complex (Fernandes et al., 2000). GIP is a nuclear protein found in all cells in which golli has been found. It can bind to golli, and it can also bind to NLI (nuclear LIM interactor), which is known to be associated with LIMcontaining nuclear complexes that are involved in regulating transcription of some genes (Jurata et al., 1996). GIP appears to be a member of a large family of proteins of unknown function found in species as diverse as yeast and human (Ishikawa, 1997; Marquet et al., 2000; Su et al., 1997). Localization of GIP to the nucleus of neural cells and the demonstration of its interaction with a known nuclear protein represents the Wrst demonstration that this family may have a nuclear function in these organisms. Thus, in the nucleus, golli is likely to be involved in the modulation of genes regulated by LIM family members.

A Role for Golli Proteins in Signaling Pathways A golli KO mouse has been generated, in which golli proteins, but not the classic MBPs, have been selectively ablated (Campagnoni et al., 2000; Feng et al., 2002b; Olmstead et al., 2000, Voskuhl, et al. 2003). In proliferation assays on cells isolated from spleens of golli KO mice, it appears that golli-deWcient splenocytes undergo hyperproliferation compared to controls when treated with anti-CD3, an activator of T-cells (Feng et al., 2002b). These results suggest that a natural role of golli might be to suppress this pathway. Conversely, increased expression of golli BG21 in Jurkat T cells strongly inhibits anti-CD3 induced IL2-luciferase activity, an indicator of T lymphocyte activation. The BG21 and J37 golli isoforms are substrates for PKC (Feng et al., 2002b), and there is an increase in the phosphorylation of golli proteins upon activation of the T-cell line with PMA, presumably a direct result of PKC activation by the PMA. As noted earlier, the classic MBPs are substrates for PKC and MAP kinases (ERKs), so these results with the golli proteins provide further evidence that MBP gene products are natural substrates for these kinases. Golli BG21 inhibits the PMA-induced increase in AP-1 or NF-kB activation, downstream targets of the PKC pathway in T cells, and these results are consistent with golli acting on the PKC pathway (Feng et al., 2002b). Golli BG21 inhibition of the PKC pathway does not appear to be due to a direct action on PKC activation, but occurs in the cascade following PKC activation, since BG21 neither reduces PKC enzyme activity nor blocks the membrane association of PKCy brought on by T lymphocyte activation (Feng et al., 2002b). The inhibitory function of BG21 is independent of its phosphorylation by PKC because a modiWed BG21, in which the PKC sites have been mutated, is as eVective as the wild type BG21 in inhibiting the PMA-induced AP-1 activation. Structure-function assays indicate that BG21 inhibitory activity resides in the golli domain rather than in MBP domain of the molecule (Feng and Campagnoni, unpublished results, submitted). These results are consistent with a role for golli in T-cells, and probably also in neural cells.

SUMMARY AND PERSPECTIVES

A Role for Golli-mbps in Oligodendrocyte Maturation, Process Extension, or Myelination? These results are of interest because there is a growing consensus that during OL development, inhibition of the PKC pathway encourages the diVerentiation of OLs, process elaboration and myelination (Baron et al., 1998; Buttery and Vrench-Constant, 2001; Stariha and Kim, 2001). The observations that golli can inhibit PKC and that golli overexpression leads to process outgrowth in immature OL cell lines is consistent with this notion, and that golli could play a role in process extension in the OL through this mechanism. As indicated earlier, the members of the MARCKS (myristoylated alanine-rich C kinase substrate) family of proteins (which include GAP-43 and CAP23) are major PKC substrates associated with the plasma membrane (Laux et al., 2002). They do not share extensive sequence homology, but they do share many common properties. They bind acidic phospholipids, actin, and calmodulin (CaM). They colocalize in rafts within the plasma membrane. They cause cytoskeletal rearrangements important for process formation in neural cells. They are all PKC substrates. Interestingly, golli proteins share all these properties in common with the MARCKS family members, in part because of the structural similarity to the classic MBPs (Kaur et al., 2003). MARCKS is thought to be the major downstream target of PKC in the OL progenitor/immature OL (Baron et al., 1999; Bhat, 1995) and it is involved in process extension. This similarity in properties is consistent with a possible role for golli in regulating process extension through PKC signaling mechanisms. It is interesting that golli expression has been reported to be upregulated in adult oligodendrocyte precursors during attempts at remyelination in MS lesions (Filipovic et al., 2002).

SUMMARY AND PERSPECTIVES In this short review, we have tried to outline some of the evidence suggesting alternative functions for the myelin basic protein gene. We are, of course, familiar with the role of the classic MBPs in the maintenance of myelin structure. Over the years, however, in the course of studying myelin basic protein metabolism, many aspects of the biochemical and cell biological properties of the proteins have been puzzling and diYcult to reconcile with a role for the classic MBPs in terms of its structural function in myelin. One aspect of the work, not covered in this chapter, is the appearance of at least some of the classic MBPs and golli proteins in human oligodendrocyte (precursors) prior to myelination (Hajihosseini et al., 1996; Pribyl et al., 1996; Tosic et al., 2002). Is this of any signiWcance, or just a sloppy transcriptional mechanism leading to the production of a protein that is easily degraded and not deleterious to the cell? On the other hand, the promoter for the classic MBPs seems to be extraordinarily speciWc, so could this be of greater signiWcance to the oligodendrocyte? Clearly, if the classic MBPs are involved in some aspect of cell signaling, their amounts in the cell need only be a small percentage of the those needed to generate myelin, the most abundant membrane in the brain. To a large extent the huge transcriptional activity required to produce the MBPs for myelin in many ways will mask eVorts to identify additional functions for these proteins, except in systems not complicated by the production of vast quantities of myelin. This is, perhaps, the reason that most of the evidence for another role of the classic MBPs comes from in vitro studies. The possibility that the classic MBPs might be involved in cell signaling mechanisms dovetails well with recent Wndings of their genetically related partners, the golli-mbps. Since the golli-mbps are also expressed in neurons and immune cells as well as oligodendrocytes, their cellular role may be studied without the complications that exist for the classic MBPs. The golli proteins are not myelin structural proteins, so they are not needed by the cell at the levels of their classic MBP counterparts during myelination. The golliMBPs appear to have a role in cell signaling to the extent that they can inhibit the PKC pathway in T-cell lines and in splenocytes and thymocytes. It is likely that they play a

395

396

15. MYELIN BASIC PROTEIN GENE

similar role in neural cells such as neurons and oligodendrocytes. Studies from cell lines and the timing of golli expression in migrating neurons suggest a role for these proteins in process elaboration in neural cells. Golli-mbps appear to be upregulated in lymphoid cells during EAE and in microglia and oligodendrocytes during remyelination attempts in MS. We would like to speculate that the golli proteins, and possibly certain isoforms of the classic MBPs, play a role in oligodendrocyte signaling pathways early in the elaboration of the myelin sheath. Thus, the MBP gene and its products appear to have a broader role in neurobiology and neuroimmunology than we have appreciated up to now. We are only just beginning to understand what these other roles might be, and we expect that these will continue to be clariWed in the next few years.

Acknowledgments This work was supported by NIH grants NS23022 and NS33091, and National Multiple Sclerosis Society grant RG2693.

References Anderson S, Miskimins R (1994). Involvement of Protein Kinase C in cAMP Regulation of myelin basic protein gene expression. J Neurosci Res 37: 604–11. Arbuzova A, Schmitz AAP, Verge`res G (2002). Cross-talk unfolded: MARCKS proteins. Biochem J 362: 1–12. Aruga J, Okano H, Mikoshiba K (1991). IdentiWcation of the new isoforms of mouse myelin basic protein: The existence of exon 5a. J Neurochem 56: 1222–1226. Asipu A, Mellor AL, Blair GE (2001). The speciWcity of the myelin basic protein gene promoter studied in transgenic mice. Biochem Biophys Res Commun 288: 809–18. Barbarese E, Braun PE, Carson JH (1977). IdentiWcation of prelarge and presmall basic proteins in mouse myelin and their structural relationship to large and small basic proteins. Proc Natl Acad Sci USA 74: 3360–4. Barbarese E, Carson JH, Braun PE (1978). Accumulation of the 4 myelin basic proteins in mouse brain during development. J Neurochem 31: 779–782. Baron W, de Jonge JC, de Vries H, Hoekstra D (1998). Regulation of oligodendrocyte diVerentiation: Protein kinase C activation prevents diVerentiation of O2A progenitor cells toward oligodendrocytes. Glia 22: 121–129. Baron W, de Vries DJ, de Vries H, Hoekstra D (1999). Protein kinase C prevents oligodendrocyte diVerentiation: Modulation of actin cytoskelton and cognate polarized membrane traYc. J Neurobiol 41: 385–398. Bhat N (1995). Signal transduction mechanisms in glial cells. Dev Neurosci 17: 267–284. Boggs JM, Rangaraj G (2000). Interaction of lipid-bound myelin basic protein with actin Wlaments and calmodulin. Biochemistry 9: 7799–7806. Brady GW, Fein DB, Wood DD, Moscarello MA (1985). The role of charge microheterogeneity of human myelin basic protein in the formation of phosphatidylglycerol multilayers. Biochem Biophys Res Commun 126: 1161–5. Buttery PC, Vrench-Constant C (2001). Process extension and myelin sheet formation in maturing oligodendrocytes. Prog Brain Res 132: 115–130. Campagnoni AT, Pribyl TM, Campagnoni CW, Kampf K, Amur-Umarjee S, Landry CF, Handley VW, Newman SL, Garbay B, Kitamura K (1993). Structure and developmental regulation of Golli-mbp, a 105-kilobase gene that encompasses the myelin basic protein gene and is expressed in cells in the oligodendrocyte lineage in the brain. J Biol Chem 268: 4930–8. Campagnoni A, Pribyl T Kampf K Jacobs E Bongarzone E Campagnoni C Handley V Skinner E Givogri M and Messing A. (2000). Targeted ablation of the golli products of the myelin basic protein gene: delayed myelination. Society for Neuroscience, Abstracts 26: 74. Campagnoni AT, SkoV RP (2001). The pathobiology of myelin mutants reveals novel biological functions of the MBP and PLP genes. Brain Pathol 11: 74–91. Carnegie PR (1971). Amino acid sequence of the basic protein of human brain myelin. Biochem J 123: 57–67. Carnegie PR, Dunkley PR, Kemp BE, Murray AW (1974). Phosphorylation of selected serine and threonine residues in myelin basic protein by endogenous and exogenous protein kinases. Nature 249: 147–150. Carnegie PR, Kemp BE, Dunkley PR, Murray AW (1973). Phosphorylation of myelin basic protein by an adenosine 3’: 5’-cyclic monophosphate-dependent protein kinase. Biochem J 135: 569–572. Cheifetz S, Moscarello MA (1985). EVect of bovine basic protein charge microheterogeneity on protein-induced aggregation of unilamellar vesicles containing a mixture of acidic and neutral phospholipids. Biochemistry 24: 1909–14. Clark RE, Miskimins WK, Miskimins R (2002). Cyclic AMP inducibility of the myelin basic protein gene promoter requires the NF1 site. Int J Dev Neurosci 20: 103–11. de Ferra F, Engh H, Hudson L, Kamholz J, Puckett C, Molineaux S, Lazzarini RA (1985). Alternative splicing accounts for the four forms of myelin basic protein. Cell 43: 721–7. Devine-Beach K, Lashgari MS, Khalili K (1990). Myelin basic protein gene transcription. IdentiWcation of proximal and distal cis-acting regulatory elements. J Biol Chem 265: 13830–35.

SUMMARY AND PERSPECTIVES

Diehl HJ, Schaich M, Budzinski RM, StoVel W (1986). Individual exons encode the integral membrane domains of human myelin proteolipid protein. Proc Natl Acad Sci USA 83: 9807–11. Dunkley PR, Carnegie PR (1974). Amino acid sequence of the smaller basic protein from rat brain myelin. Biochem J 141: 243–55. Dyer CA, Benjamins JA (1988). Antibody to galactocerebroside alters organization of oligodendroglial membrane sheets in culture. J Neurosci 8: 4307–4311. Dyer CA, Benjamins JA (1989). Organization of oligodendroglial membrane sheets: II. Galactocerebroside:antibody interactions signal changes in cytoskeleton and myelin basic protein. J Neurosci Res 24: 201–211. Dyer CA, Philibotte TM, Wolf MK, Billings-Gagliardi S (1994). Myelin basic protein mediates extracellular signals that regulate microtubule stability in oligodendrocyte membrane sheets. J Neurosci Res 39: 97–107. Dyer CA, Phillbotte T, Wolf MK, Billings-Gagliardi S (1997). Regulation of cytoskeleton by myelin components: studies on shiverer oligodendrocytes carrying an Mbp transgene. Dev Neurosci 19: 395–409. Eng LF, Chao FC, Gerstl B, Pratt D, Tavaststjerna MG (1968). The maturation of human white matter myelin. Fractionation of the myelin membrane proteins. Biochemistry 7: 4455–65. Eylar EH (1970). Amino acid sequence of the basic protein of the myelin membrane. Proc Natl Acad Sci USA 67: 1425–1431. Farsetti A, Lazar J, Phyllaier M, Lippoldt R, Pontecorvi A, Nikodem VM (1997). Active repression by thyroid hormone receptor splicing variant alpha 2 requires speciWc regulatory elements in the context of native triiodothyronine-regulated gene promoters. Endocrinology 138: 4705–12. Farsetti A, Misuhashi T, Desvergne B, Robbins J, Nikodem VM (1991). Molecular basis of thyroid hormone regulation of myelin basic protein gene expression in rodent brain. J Biol Chem 266: 23226–32. Feng J-M, Fernandes AO, Campagnoni AT (2002a). Golli-myelin basic proteins delineate the nerve distribution of lymphoid organs. J. Neuroimmunol. 123: 1–8. Feng JM, Fernandes AO, Pribyl TM, Kampf K, Campagnoni CW, Handley VW, Reyes, S, Messing, A, Campagnoni AT (2002b). Myelin basic protein gene products modulate signal transduction in T cells. FASEB J. 16: A317. Feng JM, Givogri IM, Bongarzone ER, Campagnoni C, Jacobs E, Handley VW, Schonmann V, Campagnoni AT (2000). Thymocytes express the golli products of the myelin basic protein gene and levels of expression are stage-dependent. J. Immunol. 165: 5443–5450. Fernandes AO, Campagnoni CW, Kampf K, Handley VW, Bongarzone, E. R., Schonmann V, Heang C, VonHungen R, Campagnoni AT (2000). Cloning and characterization of a novel protein that interacts with golli-MBP. J Neurochem 74 (Suppl.):S34. Filipovic R, Rakic S, Zecevic N (2002). Expression of Golli proteins in adult human brain and in multiple sclerosis lesions. J. Neuroimmunol. 127: 1–12. Forghani R, Garofalo L, Foran DR, Farhadi HF, Lepage P, Hudson TJ, TretjakoV I, Valera P, Peterson A (2001). A distal upstream enhancer from the myelin basic protein gene regulates expression in myelin-forming Schwann cells. J Neurosci 21: 3780–87. Givogri MI, Bongarzone ER, Campagnoni AT (2000). New insights on the biology of myelin basic protein gene: the neural-immune connection. J Neurosci Res 59: 153–9. Givogri MI, Bongarzone, ER, Schonmann, V, and Campagnoni, AT (2001). Expression and regulation of golli products of the MBP gene during in vitro development of oligodendrocytes. J. Neurosci. Res. 66: 679–690. Goldner-Sauve A, Szpirer C, Szpirer J, Levan G, Gasser DL (1991). Chromosome assignments of the genes for glucocorticoid receptor, myelin basic protein, leukocyte common antigen and TRPM2 in the rat. Biochem Genet 29: 275–286. Gow A, Friedrich VL Jr, Lazzarini RA (1992). Myelin basic protein gene contains separate enhancers for oligodendrocyte and Schwann cell expression. J Cell Biol 119: 605–616. GriYths IR (1996). Myelin mutants: model systems for the study of normal and abnormal myelination. Bioessays 18: 789–97. Haas S, Steplewiski A, Siracusa LD, Amini S Khalili K (1995). IdentiWcation of a sequence-speciWc singlestranded DNA binding protein that suppresses transcription of the mouse myelin basic protein gene. J Biol Chem 270: 12503–10. Hajihosseini M, Tham TN, Dubois-Dalcq M (1996). Origin of oligodendrocytes within the human spinal cord. J Neurosci 16: 7981. Harauz G, Ishiyama N, Bates IR (2000). Analogous structural motifs in myelin basic protein and in MARCKS. Mol Cell Biochem 209: 155–163. Hardy RJ, Lazzarini RA, Colman DR, Friedrich VL Jr (1996). Cytoplasmic and nuclear localization of myelin basic proteins reveals heterogeneity among oligodendrocytes. J Neurosci Res 46: 246–57. Hayes C, Kelly D, Murayama S, Komiyama A, Suzuki K, Popko B (1992). Expression of the neu oncogene under the transcriptional control of the myelin basic protein gene in transgenic mice: Generation of transformed glial cells. J Neurosci Res 31: 175–187. Huang CJ, Nazarian R, Lee J, Zhao PM, Espinosa-JeVrey A de Vellis J (2002). Tumor necrosis factor modulates transcription of myelin basic protein gene through nuclear factor kappa B in a human oligodendroglioma cell line. Int J Dev Neurosci 20: 289–96. Ikenaka K, Kagawa T (1995). Transgenic systems in studying myelin gene expression. Dev Neurosci 17: 127–36. Ishikawa S, Takahashi T, Ogawa M, Nakamura Y (1997). Genomic structure of the human PLCD1 (phospholipase C delta 1) locus on 3p22!p21.3. Cytogenet Cell Genet 78: 58–60.

397

398

15. MYELIN BASIC PROTEIN GENE

Jacobs E, Bongarzone E, Campagnoni C, Kampf K, Campagnoni A. (2000). Expression of the Sr products of the myelin proteolipid protein gene in brain regions undergoing secondary neurogenesis. Soc Neurosci Abstr 26: Abstract No.692.7. Jeanin E, Raobyr D, Desvergne B (1998). Transcriptional regulatory patterns of the myelin basic protein and malic enzyme genes by the thyroid hormone receptors alpha 1 and beta 1. J Biol Chem 273: 24238–48. Jetten AM, Suter U (2000). The peripheral myelin protein 22 and epithelial membrane protein family. Prog Nucleic Acid Res Mol Biol. 64: 97–129. Johns TG, Bernard CC (1999). The structure and function of myelin oligodendrocyte protein. J Neurochem 72: 1–9. Jordan CA, Friedrich VL Jr, de Ferra F, Weismiller DG, Holmes KV, Dubois-Dalcq M (1990). DiVerential exon expression in myelin basic protein transcripts during central nervous system (CNS) remyelination. Cell Mol Neurobiol 10: 3–18. Jordan CA, Friedrich VL Jr, Dubois-Dalcq M (1989). In-situ hybridization analysis of myelin gene transcripts in developing mouse spinal cord. J Neurosci 9: 248–257. Jurata LW, Kenny DA, Gill GN (1996). Nuclear LIM interactor, a rhombotin and LIM homeodomain interacting protein, is expressed early in neuronal development. Proc Natl Acad Sci USA 93: 11693–8. Kamholz J, Awatramani R, Menichella D, Jiang H, Xu W, Shy M. (1999). Regulation of myelin-speciWc gene expression. Relevance to CMT1. Ann NY Acad Sci 883: 91–108. Kamholz J, de Ferra F, Puckett C, Lazzarini R (1986). IdentiWcation of three forms of human myelin basic protein by cDNA cloning. Proc Natl Acad Sci USA 83: 4962–6. Katsuki M, Sato M, Kimura M, Yokoyama M, Kobayashi K, Nomura T (1988). Conversion of normal behavior to shiverer by myelin basic protein antisense cDNA in transgenic mice. Science 241: 593–5. Kaur J, Libach D, Campagnoni C, Wood DD, Moscarello, MA, Campagnoni AT, Harauz G. (2003). Expression and properties of the recombinant murine golli-myelin basic protein (MBP) isoform J37. J Neurosci Res 71: 777–784. Kibler RF, Shapira, R, McKneally S, Jenkins J, Selden P, Chou F (1969). Encephalitogenic protein:structure. Science 164: 577–580. Kimura M, Sato M, Akatsuka A, Saito S, Ando K, Yokoyama M, Katsuki M (1998). Overexpression of a minor component of myelin basic protein isoform (17.2 kDa) can restore myelinogenesis in transgenic shiverer mice. Brain Res 785: 245–52. Kitamura K, Newman SL, Campagnoni CW, Verdi JM, Mohandas T, Handley VW, Campagnoni AT (1990). Expression of a novel transcript of the myelin basic protein gene. J Neurochem 54: 2032–41. Koizumi T, Katsuki M, Kimura M, Hayakawa J (1991). Localization of the gene encoding myelin basic protein to mouse chromosome 18E3—4 and rat chromosome 1p11—p12. Cytogenet Cell Genet 56(3–4): 199–201. Landry CF, Ellison JA, Pribyl, TM, Campagnoni, C, Kampf, K, Campagnoni, AT (1996). Myelin basic protein gene expression in neurons: Developmental and regional changes in protein targeting within neuronal nuclei, cell bodies, and processes. J. Neurosci. 16: 2452–2462. Landry CF, Ellison J, Skinner E, Campagnoni AT (1997). Golli-mbp proteins mark the earliest stages of Wber extension and terminal arboration in the mouse peripheral nervous system. J. Neurosci. Res. 50: 265–271. Landry, CF, Pribyl, TM, Ellison, JA, Givogri, MI, Kampf K, Campagnoni, CW, and Campagnoni, AT (1998). Embryonic expression of the myelin basic protein gene: identiWcation of a promoter region that targets transgene expression to pioneer neurons. J. Neurosci. 18: 7315–7327. Laux T, Fukami K, Thelen M, Golub T, Frey D, Caroni, P (2000). GAP43, MARCKS and CAP23 modulate PI(4,5)P2 at plasmalemmal rafts and regulate cell cortex actin dynamics through a common mechanism. J Cell Biol 149: 1455–1471. Libich DS, Harauz G (2002a). Interactions of the 18.5kDa isoform of myelin basic protein with Ca2þ-calmodulin: in vitro studies using Xuorescence microscopy and spectroscopy. Biochem Cell Biol 80: 395–406. Libich DS, Harauz G (2002b). Interactions of the 18.5 kDa isoform of myelin basic protein with Ca2þ-calmodulin: In vitro studies using gel shift assays. Mol Cell Biochem 241: 45–52. Lui H, MacKenzie-Graham AJ, Palaszynski K, Liva S, Voskuhl RR (2001). ‘‘Classic’’ myelin basic proteins are expressed in lymphoid tissue macrophages. J Neuroimmunol 116: 83–93. MacKenzie-Graham AJ, Pribyl TM, Kim S, Porter VR, Campagnoni, AT, Voskuhl RR (1997). J Immunol 59: 4602–4610. Marquet S, Lepage P, Hudson TJ, Musser JM, Schurr E (2000). Complete nucleotide sequence and genomic structure of the human NRAMP1 gene region on chromosome region 2q35. Mamm Genome 11: 755–62. Martenson RE, Deibler GE, Kies MW (1971). The occurrence of two myelin basic proteins in the central nervous system of rodents in the suborders Myomorpha and Sciuromorpha. J Neurochem 18: 2427–33. Martenson RE, Deibler GE, Kies MW, McKneally SS, Shapira R, Kibler RF (1972). DiVerences between the two myelin basic proteins of the rat central nervous system. A deletion in the smaller protein. Biochim Biophys Acta 263: 193–203. Martenson RE, Law MJ, Deibler GE (1983). IdentiWcation of multiple in vivo phosphorylation sites in rabbit myelin basic protein. J Biol Chem 258: 930–937. Mikoshiba K, Aruga J, Okano H (1991). Structure and function of myelin protein genes. Ann Rev Neurosci 14: 201–217. Milner RJ, Lai C, Nave KA, Lenoir D, Ogata J, SutcliVe JG (1985). Nucleotide sequences of two mRNAs for rat brain myelin proteolipid protein. Cell 42: 931–9.

SUMMARY AND PERSPECTIVES

Mirsky R, Parkinson DB, Dong Z, Meier C, Calle E, Brennan A, Topilko P, Harris BS, Stewart HJ, Jessen KR (2001). Regulation of genes involved in Schwann cell development and diVerentiation. Prog Brain Res 132: 3–11. Miskimins R, Knapp L, Dewey MJ, Zhang X (1992). Cell and tissue-speciWc expression of a heterologous gene under control of the myelin basic protein gene promoter in transgenic mice. Brain Res Dev Brain Res 65: 217–221. Miskimins RA, Miskimins W K (2001). A role for an AP-1-like site in the expression of the myelin basic protein gene during diVerentiation. Int J Devl Neurosci 19: 85–91. Miura M, Tamura T, Aoyama A, Mikoshiba K (1989). The promoter elements of the mouse myelin basic protein gene function eYciently in NG108–15 neuronal/glial cells. Gene 75: 31–8. Nagasato K, Farris RW 2nd, Dubois-Dalcq M, Voskuhl RR (1997). Exon 2 containing myelin basic protein (MBP) transcripts are expressed in lesions of experimental allergic encephalomyelitis (EAE). J Neuroimmunol 72: 21–5. Newman S, Kitamura K, Campagnoni AT (1987). IdentiWcation of a cDNA coding for a Wfth form of myelin basic protein in mouse. Proc Natl Acad Sci USA 84: 886–90. Norton WT and Cammer W (1984). Isolation and characterization of myelin. In ‘‘Myelin’’ (P. Morell, ed), pp 147–195. Plenum Press, New York. Olmstead C, Pribyl T, Kampf K, Jacobs E, Campagnoni C, Handley V, Skinner E, Messing A, LazareV J, Campagnoni A (2000). Targeted ablation of the golli products of the myelin basic protein gene: Learning deWcits. Society for Neuroscience Abstracts 26: 75. Pedraza L, Fidler L, Staugaitis SM, Colman DR (1997). The active transport of myelin basic protein into the nucleus suggests a regulatory role in myelination. Neuron 18: 579–89. Pribyl TM, Campagnoni CW, Kampf K, Ellison JA, Landry CF, Kashima T, McMahon J, and Campagnoni AT (1996). Expression of the myelin basic protein gene locus in neurons and oligodendrocytes in the human fetal central nervous system. J Comp Neurol 374: 342–353. Pribyl TM, Campagnoni CW, Kampf K, Kashima T, Handley VW, McMahon J, Campagnoni AT (1993). The human myelin basic protein gene is included within a 179-kilobase transcription unit: Expression in the immune and central nervous systems. Proc Natl Acad Sci USA 90: 10695–9. Readhead C, Popko B, Takahashi N, Shine HD, Saavedra RA, Sidman RL, Hood L (1987). Expression of a myelin basic protein gene in transgenic shiverer mice: Correction of the dysmyelinating phenotype. Cell 48: 703–712. Reyes SD, Campagnoni AT (2002). Two separate domains in the golli myelin basic proteins are responsible for nuclear targeting and process extension in transfected cells. J. Neurosci. Res 69: 587–596. Roach A, Takahashi N, Pravtcheva D, Ruddle F, Hood L (1985). Chromosomal mapping of mouse myelin basic protein gene and structure and transcription of the partially deleted gene in shiverer mutant mice. Cell 42: 149–55. Roth HJ, Kronquist KE, Kerlero de Rosbo N, Crandall BF, Campagnoni AT (1987). Evidence for the expression of four myelin basic protein variants in the developing human spinal cord through cDNA cloning. J Neurosci Res 17: 321–8. Shoji S, Ohnishi J, Funakoshi T, Fukunaga K, Miyamoto E, Ueki H, Kubota Y (1987). 102: 1113–20. Sparkes RS, Mohandas T, Heinzmann C, Roth HJ, Klisak I, Campagnoni AT (1987). Assignment of the myelin basic protein gene to human chromosome 18q22-qter. Hum Genet 75: 147–50. Stariha RL, Kim SU (2001). Protein kinase C and mitogen-activated protein kinase signalling in oligodendrocytes. Micros Res Tech 52: 680–688. Staugaitis SM, Colman, DR, Pedraza L (1996). Membrane adhesion and other functions for the myelin basic proteins. Bioessays 18: 13–18. Steck AJ, Appel SH (1974). Phosphorylation of the myelin basic protein. J Biol Chem 249: 5416–20. Steplewiski A, Krynska B, Tretiakova A, Haas S, Khalili K Amini S MyEF-3 (1998). A developmentally controlled brain-derived nuclear protein which speciWcally interacts with myelin basic protein proximal regulatory sequences. Biochem Biopohys Res Commun 243: 295–301. Su YA, Lee MM, Hutter CM, Meltzer PS (1997). Characterization of a highly conserved gene (OS4) ampliWed with CDK4 in human sarcomas. Oncogene 15: 1289–94. Takahashi N, Roach A, Teplow DB, Prusiner SB, Hood L (1985). Cloning and characterization of the myelin basic protein gene from mouse: One gene can encode both 14 kd and 18.5 kd MBPs by alternate use of exons. Cell 42: 139–48. Tamura T, Aoyama A, Inoue T, Miura M, Okano H, Mikoshiba K (1989). Tissue-speciWc in vitro transcription from the mouse myelin basic protein promoter. Mol Cell Biol 9: 3122–3126. Tamura T, Sumita K, Mikoshiba K (1991). Sequences involved in brain-speciWc in vitro transcription from the core promoter of the mouse myelin basic protein gene. Biochim Biophys Acta 1129: 83–86. Taupin V, Renno T, Boubonniere L, Peterson AC, Rodriguez M, Owens T (1997). Increased severity of experimental autoimmune encephalomyelitis, chronic macrophage/microglial reactivity, and demyelination in transgenic mice producing tumor necrosis factor-alpha in the central nervous system. Eur J Immunol 27: 905–913. Taveggia C, Pizzagalli A, Feltri ML, Grinspan JB Kamholz J, Wrabetz L (1998). MEBA derepresses the proximal myelin basic protein promoter in oligodendrocytes. J Biol Chem 273: 27741–8. Tosic M, Rakic S, Matthieu JM, Zecevic N (2002). IdentiWcation of golli and myelin basic proteins in human brain during early development. Glia 37: 219–228.

399

400

15. MYELIN BASIC PROTEIN GENE

Tretiakova A., Steplewski A, Johnson EM, Khalili K, Amini S (1999). Regulation of myelin basic protein transcription by Sp1 and Puralpha: Evidence for association of Sp1 and Puralpha in brain. J Cell Physiol. 181: 160–8. Turnley BD, Morahan G, Okano H, Bernard O, Mikoshiba K, Allison J, Barlett PF, Miller JFAP (1991). Dysmyelination in transgenic mice resulting from expression of class I histocompatibility molecules in oligodendrocytes. Nature 353: 566–9. Ulmer JB, Braun PE (1983). In vivo phosphorylation of myelin basic proteins in developing mouse brain: Evidence that phosphorylation is an early event in myelin formation. Dev Neurosci 6: 345–55. Vanderluit JL, Bourque JA, Peterson AC, TetzlaV W (2000). Model for focal demyelination of the spinal dorsal columns of transgenic MBP-LacZ mice by phototargeted ablation of oligodendrocytes. J Neurosci Res 62: 28–39. Vartanian T, Szuchet S, Dawson G, Campagnoni AT (1986). Oligodendrocyte adhesion activates protein kinase C-mediated phosphorylation of myelin basic protein. Science 234: 1395–1398. Voskuhl, R. R., Pribyl, T. M., Kampf, K., Handley, V., Liu, H., Feng, J-M., Campagnoni, C. W., Soldan, S. S., Messing, A., and Campagnoni, A. T. (2003). Experimental autoimmune encephalomyelitis relapses are reduced in heterozygous golli-MBP knock-out mice. J Neuroimmunol 139: 44–50. Wegner M (2000). Transcriptional control in myelinating glia: Flavors and spices. Glia 31: 1–14. Wilson R, Brophy PJ (1989). Role for the oligodendrocyte cytoskeleton in myelination. J Neurosci Res 22: 439–48. Wrabetz L, Shumas S, Grinspan J, Feltri ML, Bozyczko D, McMorris FA, Pleasure D Kamholz J (1993). Analysis of the human MBP promoter in primary cultures of oligodendrocytes; positive and negative cisacting elements in the proximal MBP promoter mediate oligodendrocyte-speciWc expression of MBP. J Neurosci Res 36: 455–71. Xie Y, Skinner E, Landry C, Handley V, Schonmann V, Jacobs E, Fisher R, Campagnoni A (2002). InXuence of the embryonic preplate on the organization of the cerebral cortex: A targeted ablation model. J. Neurosci. 22: 8981–91. Yoshikawa H (2001). Myelin-associated oligodendrocytic basic protein modulates the arrangement of radial growth of the axon and the radial component of myelin. Med Electron Microsc 34: 160–164. Yoshioka TL, Feigenbaum L, Jay G (1991). Transgenic mouse model for central nervous system demyelination. Mol Cell Biol 11: 5479–86. Yu YT, Campagnoni AT (1982). In vitro synthesis of the four mouse myelin basic proteins: Evidence for the lack of a metabolic relationship. J Neurochem. 39: 1559–68.

C H A P T E R

16 Proteolipid Protein Gene Lynn D. Hudson

CONTROL OF PLP GENE EXPRESSION Structure and Alternative Splicing of the PLP Gene The discovery of the myelin proteolipid protein (PLP) by Folch and Lees launched a series of studies where investigators Wrst grappled with an extremely hydrophobic molecule and obtained some precious amino acid sequence information to aid in the eventual isolation of the gene (Folch and Lees, 1951; reviewed in Lees et al., 1979, and Lees and BrostoV, 1984). The PLP locus exists as a single copy gene on the X chromosome (Xq22.2 in human). That only one allele is devoted to the synthesis of one of the most abundant transcripts in oligodendrocytes (estimated at about 1% of the total cellular transcripts) for a protein that constitutes half of the myelin proteins suggests that oligodendrocytes may have evolved special mechansims for maximizing and regulating the output of the PLP gene. Spanning about 17 kb, the PLP locus consists of seven exons (Fig. 16.1) (Diehl et al., 1986; Ikenaka et al., 1988; Macklin et al., 1987; StoVel et al., 1984) and a single promoter region. The Wrst exon encodes the initiator methionine, and the last exon includes the carboxy terminus followed by three polyadenylation sites. The full length gene encodes a 276 amino acid integral membrane protein with a basic isoelectric point. Alternative splicing occurs within an exon, an uncommonly encountered splicing event, to generate PLP and the smaller, functionally distinct DM20 isoform. That the withinexon splicing of exon 3 is a rate limiting step in PLP/DM20 transcript synthesis is suggested by the presence of signiWcant levels of a PLP precursor mRNA containing intron 3 (Vouyiouklis et al., 2000) (Fig. 16.1). The elevated levels of this precursor mRNA during CNS myelination, together with the fact that intron 3 is the most highly conserved intron of PLP/DM20 genes isolated from diVerent species, also points to a step where regulatory controls are exerted. That intron 3 represents a control point in transcript processing was further suggested by experiments introducing a neomycin transgene in the sense orientation into intron 3, an intervention that selectively reduced the levels of PLP mRNA (Uschkureit et al., 2001). Two very closely spaced (about 30 bases apart) transcription initiation sites were originally described in the PLP gene by SutcliVe and colleagues (Milner et al., 1985). That these sites represent transcription initiation, as opposed to premature termination during primer extension or nuclease activity that would obscure the actual cap site, is suggested by two lines of evidence. First, these two transcription initiation sites, located only about 30 bases apart, have been identiWed in the PLP gene of other species (Baumgartner et al., 1999). Second, the two sites are diVerentially used during development, following injury, and in diVerent tissues. Both transcription initiation sites are employed in the CNS, while the downstream site is predominantly used in the PNS

Myelin Biology and Disorders, Volume 1

401

Copyright 2004, Elsevier (USA). All rights reserved.

402

16. PROTEOLIPID PROTEIN GENE

P 1

polyA +

2

3

4 5 6

7

Pre-mRNA PLP mRNA DM20 mRNA

FIGURE 16.1 Structure of the PLP gene. The seven exons are numbered and the coding regions colored black or in the case of the PLP-speciWc segment, gray. The promoter region (P) is indicated by an arrow, as are the three sites of polyadenylation. Transcripts are shown below the PLP locus, including the nuclear precursor transcript containing intron 3 and the two cytoplasmic mRNAs, which diVer by the inclusion or exclusion of the PLP segment of exon 3.

(Kamholz et al., 1992). During the peak of myelination, most transcription initiates from the upstream site (Scherer et al., 1992). Following injury, the reduction in transcription initiation at the PLP locus was more pronounced at the upstream site (Scherer et al., 1992). It is likely that an overlapping constellation of transcription factors act to direct RNA polymerase initiating at the upstream and downstream sites. What is the physiologic relevance of these closely spaced initiation sites in the PLP promoter region? Promoter usage is known to aVect splice site selection. During early oligodendrocyte development, when transcription initiates from the downstream site, DM20 is selectively expressed. The downstream site is also preferentially used in Schwann cells, where DM20 is the predominant spliced isoform. So the prediction is that transcription complexes centered on the downstream site will favor splicing in which exon 3B is subsequently excised to generate DM20, while transcription complexes centered on the upstream site favor PLP spliced products. Additional alternatively spliced products involving a newly described exon 1.1 (Bongarzone et al., 1999) have been detected in the mouse CNS at levels of at least an order of magnitude less than PLP/DM20 transcripts. In the face of the strict conservation of intron/exon junctions in mammalian PLP genes, the absence of a comparable exon in the human argues against the biologic signiWcance of these exon 1.1 splicing events in mice. Moreover, inclusion of exon 1.1 would yield a transcript encoding a 44 amino acid protein that would initiate with the methionine in exon 1, continue through the open reading frame of exon1.1 and terminate out of frame with the PLP/DM20 coding sequence in exon 2. Translation of the predicted soma-restricted PLP (srPLP) and soma-restricted DM20 (srDM20) proteins is unlikely for several reasons: (1) the expected translation at the PLP/DM20 initiation codon in exon 1 would read through the srPLP and srDM20 region out of frame, (2) the lack of evidence for an internal ribosome entry site (IRES) that would enable translation to initiate at the downstream methionine codon, and (3) the extremely unfavorable ribosome binding site context of the putative initiator methionine for srPLP and srDM20 in exon 1.1, in which pyrimidines occupy the critical
3 and þ4 positions.

Regulatory Sites within the PLP Gene PLP gene expression reXects a balance between activators and repressors binding to sites within the PLP gene. At least a subset of these transcription factors are anticipated to recognize additional myelin gene promoters, as one mechanism for coordinately controlling a group of genes is via common DNA sequences (cis elements), which are binding sites for regulatory proteins (trans factors). The PLP promoter does have several binding sites whose sequences are shared with other myelin genes and may represent points of coordinate control for myelination (Berndt et al., 1992). The regions of the PLP gene that participate in transcriptional regulation have been identiWed by a series of biochemical assays (e.g., electrophoretic mobility shift assays, DNA footprinting), inspection of evolu-

CONTROL OF PLP GENE EXPRESSION

tionarily conserved sequences, transfection assays in cultured cells, and expression analysis in transgenic mice (Berndt et al., 1992; Fuss et al., 2000; Janz and StoVel, 1993; Mallon et al., 2002; Nadon et al., 1994; Nave and Lemke, 1991; Spassky et al., 1998; Wight et al., au1 1993, 1997). Transgenic mice expressing a reporter gene from a large fragment of the PLP gene, including 2.4 kb of upstream sequence, exon 1, intron 1 and part of exon 2, exhibit the expected spatial and temporal pattern of PLP/DM20 expression (Fuss et al., 2000; Spassky et al., 1998; Wight et al., 1993). However, a PLP promoter fragment lacking the intron 1 sequence is also capable of directing appropriate oligodendrocyte and Schwann cell–speciWc transgene expression (Nadon et al., 1994; Nadon and West, 1998). These experiments call into question the role of intron 1 in regulating PLP expression in vivo, although elements located within intron 1 of the murine PLP gene do mediate repression of PLP in nonexpressing cultured cells (Li et al., 2002; Wight and Dobretsova, 1997).

Transcriptional Regulators that Bind to the PLP Gene Integral to the regulation of gene expression in oligodendrocytes is the activation of myelin gene loci at the earliest stages of the oligodendrocyte lineage, mouse embryonic day 9 to 10 for the PLP/DM20 locus (Chandross et al., 2001; Mallon et al., 2002; Thomas et al., 2000). This initial wave of transcriptional regulation, which occurs coincident with or shortly following the commitment of neural precursor cells to the oligodendrocyte lineage, must include one or more transcription factors that recognize myelin promoters. One candidate for this task is Myt1, which was originally cloned as a PLP promoter-binding protein expressed by neural precursor and oligodendrocyte progenitor cells (Armstrong et al., 1995; Kim and Hudson, 1992). Myt1 encodes an unusual protein of the zinc Wnger superfamily featuring structurally unique Cys2HisCys zinc Wngers (Kim et al., 1997) that Wgures in neural speciWcation in Xenopus (Bellefroid et al., 1996). Myt1 may represent one of the relatively lineage-speciWc transcription factors needed to push cells towards a myelinating phenotype. Four additional trans-acting factors that recognize sites within the PLP promoter region have been isolated by virtue of their sequence-speciWc binding: MYT2, Yin Yang 1 (YY1), and a nuclear hormone receptor heterodimer (Berndt et al., 2001; Bogazzi et al., 1994; Kim et al., 1998; Peters et al., 2000). The DNA-binding domain of MYT2 is a novel a-helical structure that speciWcally recognizes a TTCCA motif, which in the PLP promoter region is directly adjacent to, and probably overlaps with, the Myt1 binding site (Kim et al., 1998). In the CNS, MYT2 protein is found in oligodendrocyte progenitor cells, subsets of neurons, and cells of the choroid plexus together with ciliated ependymal cells. Most striking for a DNA-binding protein, MYT2 is secreted by cells and is a major constituent of cerebrospinal Xuid (Li et al.,1995). Another unusual feature of the MYT2 transcript is a functional internal ribosome entry sites (IRES) (Kim et al., 1998). The internal initiation of translation, a mechanism infrequently used by cellular messages, avoids the requirement of a methyl cap structure for translation of messenger RNAs and thereby ensures that these transcripts can be translated during mitosis or viral infections or stress. The presence of an IRES site in MYT2, together with the unusual localization of MYT2, suggests that this nucleic acid-binding protein may be in the class of proteins involved in cellular growth control and survival in the nervous system. The transcription factor that binds the PLP promoter at the site essential for maximal transcription of the PLP gene (Berndt et al., 1992) is the ubiquitous transcriptional regulator YY1 (Berndt et al., 2001). YY1 interacts with a number of other transcription factors, components of the histone deacetylase complex and the nuclear matrix. Most of YY1’s target genes are associated with cellular proliferation or are highly expressed following diVerentiation (reviewed in Shi et al., 1997), and PLP Wts into the latter category. A mechanism by which YY1 could act on the PLP gene is to regulate its association with the nuclear matrix, which includes transcriptionally active domains. By tethering the PLP gene to the nuclear matrix of diVerentiating oligodendrocytes and by attracting additional transcription factors, YY1 may promote the formation of domains in which transcription factors required for maximal PLP transcription are clustered. Exploration of this hypothesis will require transient loss-offunction mouse models due to the early lethality of the YY1 knockout mice.

403

404

16. PROTEOLIPID PROTEIN GENE

The direct binding of a nuclear hormone receptor, thyroid hormone receptor (THR), to the PLP promoter merits special attention in view of the impact of thyroid hormone on oligodendrocyte development (Barres et al., 1994). THR heterodimerizes with another nuclear hormone receptor, the peroxisomal proliferator activated receptor b/d (PPARb/ d) and activates PLP transcription in the presence of the thyroid hormone ligand (Bogazzi et al., 1994). PPAR is known to stimulate the expression of a number of genes expressed in peroxisomes, the organelle which is a site of fatty acid metabolism, and the PPARb/d) gene is highly expressed in oligodendrocytes (Granneman et al., 1998). The existence of a common transcription factor (PPAR) that stimulates both the production of the most abundant protein in CNS myelin (PLP) and peroxisomal enzymes pointed to a possibility for coupling the myelin lipid and myelin protein pathways in oligodendrocytes. However, PPARb null mice display unaVected steady state levels of PLP transcripts and myelin loss in only selected tracts (Peters et al., 2000). A null mutant in the other partner that heterodimerize at the PLP promoter, namely THRb, likewise has unexpectedly little impact on CNS myelination (Billon et al., 2001). In each case, compensatory activity by other nuclear hormone isoforms is likely to minimize transcriptional eVects. The transcription factor Sox10 may also bind to the PLP promoter, as gauged by an increase in PLP transcripts upon overexpression of Sox10 in a cell line (Stolt et al., 2002). In the absence of Sox 10, glial development in the PNS is blocked and diVerentiation of oligodendrocytes in the CNS is disrupted (Britsch et al., 2001; Stolt et al., 2002). Given that Sox10 recognizes sites within the promoters of the MBP (Stolt et al., 2002) and P0 (Peirano and Wegner, 2000) genes, the direct contact of Sox10 with the PLP promoter would recommend this factor as one of the general regulators of myelin gene expression.

Post-Transcriptional Controls Exerted on PLP and DM20 Transcripts Messenger stability may be key to regulating PLP expression. Apart from the checkpoint in the nucleus, where the intron 3-containing precursor mRNA accumulates prior to the alternative splicing of exon 3 (Vouyiouklis et al., 2000), PLP mRNA is subjected to further controls prior to translation. By comparing transcription rates with the steady state levels of PLP/DM20 transcripts in developing sciatic nerve, Macklin and coworkers found a constant amount of nuclear run-oV transcripts in the face of a 2.5-fold increase in PLP/ DM20 mRNA (Jiang et al., 2000). This enhanced messenger RNA stability in Schwann cells was restricted to PLP transcripts and was mediated by the 105 nucleotide region of exon 3 that distinguishes PLP from DM20 transcripts. PLP transcript stability is regulated at least in part by axonal contact, as sciatic nerve transection or the withdrawal of forskolin in cultured Schwann cells resulted in a selective loss of PLP transcripts (Jiang et al., 2000). Altered stability of PLP mRNA may also Wgure in transgenic mice containing a neomycin gene cassette within intron 3, as these mice have selectively reduced levels of PLP mRNA yet constant levels of DM20 RNA (Uschkureit et al., 2001). DiVering post-transcriptional controls must be exerted on DM20 transcripts, which are synthesized in both myelinating and nonmyelinating Schwann cells, unlike PLP, which is restricted to myelinating Schwann cells (GriYths et al., 1995). The DM20 transcript, which is the predominant spliced form in the PNS (Ikenaka et al., 1992; Pham-Dinh et al., 1991), is present at relatively constant levels during sciatic nerve development, following transection or crush injury, or in Schwann cell cultures treated with cAMP analogues (Gupta et al., 1991; Jiang et al., 2000; Kamholz et al., 1992).

MOLECULAR EVOLUTION OF THE PLP GENE FAMILY IN VERTEBRATES AND INVERTEBRATES Conservation of PLP/DM20 The ancestral vertebrate gene encoded only the DM20 isoform (referred to as DMa in lower vertebrates) and was coexpressed with the myelin protein P0 in the Wrst myelinated

MOLECULAR EVOLUTION OF THE PLP GENE FAMILY IN VERTEBRATES AND INVERTEBRATES

vertebrates, the cartilaginous Wsh (Kitagawa et al., 1993; Yoshida and Colman, 1996). DMa transcripts are found in lower vertebrates: lobe-Wnned Wshes (lungWsh, coelacanth), ray-Wnned Wshes (fugu, trout, zebraWsh, sturgeon and bichir), and cartilaginous Wsh (sharks and electric rays) (Geltner et al., 1998; Kitagawa et al., 1993; Venkatesh et al., 2001; Yoshida and Colman, 1996; Yoshida et al., 1999). Genomic sequencing of Fugu Wsh established the absence of the 105 bp region in exon 3 that gives rise to PLP in tetrapods (Venkatesh et al., 2001). Of the tetrapods, mammals, birds, and reptiles express both PLP and the smaller DM20 isoform, while only PLP is found in amphibians. This pattern suggests that coincident with or shortly following the acquisition of the PLP-speciWc exon3B by tetrapods, mutations occurred within exon 3 at the DM20 splice donor site in the amphibian genome (Franz et al., 1981; Schliess and StoVel, 1991; Waehneldt et al., 1985). For the species that acquired both the PLP-speciWc exon 3B and the ability to produce the PLP and DM20 alternatively spliced transcripts (i.e., reptiles, birds and mammals), the two protein isoforms evolved to perform functionally distinct roles. In the mouse, DM20 is unable to replace PLP in CNS myelin (Sporkel et al., 2002; Stecca et al., 2000). The primordial PLP locus probably originated 550 million years ago in invertebrates, based on the recent discovery of a similar gene in Drosophila and silkworm (Stecca et al., 2000). The PLP locus has evolved slowly in the past 80 million years, estimated as less than 0.1  10-9 amino acid substitutions per site annually (Kurihara et al., 1997). But more than 300 million years ago, prior to the divergence of reptiles and birds from mammals, the ancestral gene of cartilaginous Wsh underwent a high degree of evolutionary change. In contrast to the diVerent mammalian PLP/DM20 proteins, which display the highest degree of conservation (99.6–100%), Wsh DM20 proteins are strikingly diverse, with as little as 54% conservation present between Polypterus and Protopterus (Geltner et al., 1998) (Fig. 16.2). Remarkably, the DM20 proteins of some Wsh are more closely related to mouse DM20 than to other Wsh species. The early evolutionary experimentation with DM20 has been speculatively coupled with the progressive loss of the myelin protein P0 from CNS myelin (Yoshida and Colman, 1996). As DM20 managed to assume a structural role in CNS myelin, the compacting ability of P0 was no longer required (Geltner et al., 1998). The strict conservation of the mouse, rat, rabbit, dog, pig, bovine, and human PLP/DM20 coding sequence, together with the lack of amino acid polymorphisms, hints at the intense selective pressures that molded the modern mammalian PLP/DM20 proteins and have seemingly restricted further shaping of PLP/DM20. Indeed, any change in the mammalian PLP/DM20 protein is not well tolerated, as witnessed by the large number of missense au2 mutations creating Pelizaeus-Merzbacher disease (see Chapters 37 and 47).

The Lipophilin Family of Tetraspan Proteolipid Proteins That Includes PLP/DM20 DM20 is part of a proteolipid protein gene family, the ‘‘lipophilin’’ family (Gow, 1997) featuring four transmembrane domains with intracellular termini, a large extracellular loop between transmembrane domains 3 and 4 and an intracellular portion between domains 2 and 3 (Fig. 16.3A). The other two members of the DMa/DM20/PLP family, DMb/M6A/EMA and DMg/M6B/Rhombex 29, have been isolated from Wsh, amphibians, and mammals (Baumrind et al., 1992; Kitagawa et al., 1993; Lagenaur et al., 1992; Olinsky et al., 1996; Shimokawa and Miura, 2000; Werner et al., 2001; Yan et al., 1993; Yoshida et al., 1999). The primordial gene of the lipophilin family arose in invertebrates (Stecca et al., 2000). Gow and colleagues described a Drosophila protein with 39 to 48% similarity to the vertebrate lipophilins (aligned in Fig. 16.2) and 83% similarity to another invertebrate lipophilin isolated from the silkworm, Bombyx mori (Stecca et al., 2000). A phylogenetic tree picturing the relationships between the vertebrate and invertebrate lipophilins is shown in Figure 16.4. All family members have a striking degree of amino acid conservation in the four transmembrane domains, while the extramembrane domains display more variety (Fig. 16.2). This feature hints at a close packing or interdigitation of the four a-helical membrane spanning domains, as has been noted in proteins that form membrane channels. Also strictly conserved are the four cysteine residues that in DM20 are sites for

405

406

16. PROTEOLIPID PROTEIN GENE

10

20

30

I

40

M G E C C Q S C M A R

50

60

A T L M C L

F E C C

I

K C L G G

I P Y A S

L

I

A T

I

L L Y A G V A L F

C G C G H E A L S

G T V N

I

L Q T Y F

E

L A R T A G D T

L D V

Frog DMbeta/M6A

M E E N M E E G Q T Q K G C

F E C C

I

K C L G G

I P Y A S

L

I

A T

I

L L Y A G V A L F

C G C G H E A L T G T V N

I

L Q T Y F

E M A R T A G D T

L D V

F C F

L

T V

I M V D Q V

F H L R

-

-

-

-

-

-

-

L E C C A R C L

I G A P F A S

M G L

L E C C A R C L

I G A P F A S

L V A T G L C F

F G V A L F

C G C G H E A L T G T E Q L

I E T Y F

S

K N

-

-

Y Q D Y E Y L

L E C C A R C L V G A P F A S

L V A T G L C F

F G V A L F

C G C G H E A L T G T E K L

I E T Y F

S

K N

-

-

Y Q D Y E Y L

M G L

L E C C A R C L V G A P F A S

L V A T G L C F

F G V A L F

C G C G H E A L T G T E K L

I E T Y F

I E T Y F

S

S

K N

-

-

-

-

Y Q D Y E F

L

M G L

C G C G H E A L T G T E Q L

K N

-

Human PLP Pig PLP

F G V A L F

T M Y R G A S

Chicken PLP

Zebrafinch PLP

L V A T G L C F

I

80

I

M E E N M E E G Q T Q K G C M G L

L G V G

70

I P Y A T L

Mouse M6A

Drosophila DM20/M6

L

Y Q D Y E Y L

Mouse PLP

M G L

L E C C A R C L V G A P F A S

L V A T G L C F

F G V A L F

C G C G H E A L T G T E K L

I E T Y F

S

K N

-

-

Y Q D Y E Y L

Dog PLP

M G L

L E C C A R C L V G A P F A S

L V A T G L C F

F G V A L F

C G C G H E A L T G T E K L

I E T Y F

S

K N

-

-

Y Q D Y E Y L

F G V A L F

C G C E V

S

F G V A L F

Cow PLP Mouse DM20 Trout DMalpha Lungfish DMalpha

M G L

L E C C A R C L V G A P F A S

L V A T G L C F

M G L

L E C C A R C L V G A P F A S

L V A T G L C F

M G W H D G C

I

R C M V G V P F A S

M F P V R H A L L C K A L G C Y D C C

I

R C L G A V P Y P

Frog PLP M G K R K V D S A Q G R F P

P G E R S G L F Y G C V M G C

Shark DMalpha

S

E C C A K C L G G

V

I

A T V

S L V S

T L

I P Y A S M V A T

E C C V R C L G G V P Y A S

L

I

A T

E A L T G T E K L

I E T Y F

C G C G H E A L T G T E K L

I E T Y F

L C F

A G V A L F

C G C G H E A L S

I E T Y F

L C F

T G M A L F

C G C G H E A L A H T E V L V E T Y F

I

L C F

S

I

L C F

V G V A L F

G V A L F

G T E K L

K N

-

-

Y Q D Y E Y L

S

K N

-

Y Q D Y E Y L

S

K N

-

-

Y Q E Y E Y L

V R N

-

-

-

I

Q D Y V

I

L

C G C G H E A L N G T E Y L M E T H F

I

R N

-

-

Y Q H F Q A L

I E L Y F

S

N D

-

-

F M D Y A L

L L

C G C G H E A L T G T E K L

M K P A M E T A A E E N T E Q S Q E R K G C

F E C C

I

K C L G G V P Y A S

L V A T

I

L C F

S

G V A L F

C G C G H V A L A G T V A

L E Q H F S

T N

-

-

T

S D H A L

Frog DMgamma2/M6B

M G C

F E C C

I

K C L G G V P Y A S

L V A T

I

L C F

S

G V A L F

C G C G H V A L T G T V A

I

L E Q H F S

S

N

-

-

P

S D H V

F L

Frog DMgamma1/M6B

M G C

F E C C

I

K C L G G V P Y A S

L V A T

I

L C F

C G V A L F

C G C G H V A L T G T V

T

I

L E Q H F S

S

N

-

-

P

S D H V

L L

R C

L V A T

G V A L F C G C G H E A L T G T E

.

L

I

N

-

-

Y Q D Y

Mouse M6B

-

-

-

-

-

-

-

-

-

-

-

-

-

90

-

-

-

-

-

-

-

- M G

E C C

L

110

G

.

P

A S

120

140

I

- W

I

E A V Q M

I

I

G A G M A A L G F M

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

T M

I

D

I

F K Y V

I Y G

I

A A A F F

V Y G

I

L L M V

E G F

F

T T G A

I K D L Y G D F K

I

T T C G R C V

S A W F

I M L T

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

T M

I

D

I

F K Y V

I Y G

I

A A A F F

V Y G

I

L L M V

E G F

F

T T G A

I K D L Y G D F K

I

T T C G R C V

S A W F

I M L T

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

Zebrafinch PLP

I

D V

I

H G F Q Y F

I Y G T A A F F

F

L Y G A L L L A E G F

Y T T G A V R Q

A T V T G G P

-

K G R G A R G P Q R A H S W Q R V C H C

L G K W

Chicken PLP

I

D V

I

H A F

Q Y V

I Y G T A S

F

F

F

L Y G A L L L A E G F

Y T T G A V R Q

I

F

G D Y R T T

I

C G K G L S

A T V T G G P

-

K G R G A R G P Q R A H S

L Q R V C Q C

L G K W

Human PLP

I

N V

I

H A F

Q Y V

I Y G T A S

F

F

F

L Y G A L L L A E G F

Y T T G A V R Q

I

F

G D Y K T T

I

C G K G L S

A T V T G G Q

-

K G R G S R G Q H Q A H S

L E R V C H C

L G K W

I

H A F

Q Y V

I Y G T A S

F

F

F

L Y G A L L L A E G F

Y T T G A V R Q

I

F

G D Y K T T

I

C G K G L S

K G R G S R G Q H Q A H S

I Y G T A S

F

F

F

L Y G A L L L A E G F

Y T T G A V R Q

I

F

G D Y K T T

I

C G K G L S

A T V T G G Q

-

K G R G S R G Q H Q A H S

L E R V C H C

L G K W

Cow PLP

I Y G T A S

F

F

F

L Y G A L L L A X G F

Y T T G A V R Q

I

F

G D Y K T T

I

C G K G L S

A T V T G G Q

-

K G R G S R G Q H Q A H S

L E R V C H C

L G C W

I Y G T A S

F

G D Y K T T

I

C G K G L S

A T F

I

N V

Frog PLP

I

H V

Trout DMalpha

A S

Mouse DM20

Lungfish DMalpha

S

F

N L

I

Q Y V

N A F

Q F V

K Y F Q Y V

I

D A F

Q Y V

I

A

I

I Y G M A S

F

F

L Y G A L L L A E G F

Y T T G A V R Q

I

F

F F

F

L Y G

I

L L L A E G F

Y T T T A

I

L G E

F K P

F

F

L Y C

I

L L L A E G F

Y T T S

A V K Q T

F G E

F R S T R C G R C

F

F F

F

L Y G V L L L A E G F

Y T S

S

V

I K K V F

G E N K T T L C G R C

I

I

S

V G

-

-

T V T G G P I

I

V

T

F

I

G

-

-

S A T F

I

L S

L T F

S A S

-

-

-

-

-

T R G R Q P V H T

I

E L

I

C R C

L G K W

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

A V K A L F G E

F R T T V C G R C V

E V

I

Q L M Q Y V

I Y G

I

A S

F

F

F

L Y G

I

I

L L A E G F

Y T T S

A V K E L H G E

F K T T A C G R C

I

S G M F V

F L T

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

I

Q L M Q Y V

I Y G

I

A S

F

F

F

L Y G

I

I

L L A E G F

Y T T S

A V K E L H S

E F K T T A C G R C

I

S G M F V

F L T

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

Frog DMgamma1/M6B

T

E V

I

Q V M Q Y V

I Y G

I

A S

F

F

F

L Y G

I

I

L L A E G F

Y T T S

A V K E V H S

E F K T T A C G R C

I

S G M F V

I

V

I

.

I Y G

A S

F F F L Y G

.

L L L A E G F Y T T G A V

A F Q Y V 180

F L Y G V L L L A E G F

III

190

K T T

C G

210

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

240

-

250

-

Q S

-

-

-

-

-

-

-

Y L F M L A W L G V T A F

T S

L P

-

-

-

V Y M Y F

N V W T

I

C R N T T L V

E

-

-

-

-

-

-

G A N L C L D L R Q F G

I V

T

-

-

-

-

-

-

-

Y L F M L A W L G V T A F

T S

L P

-

-

-

V Y M Y F

N L W T

I

C R N A T L

E

-

-

-

-

-

-

E A N F C L D L R Q F G

I V

T V G E E K K L C

Zebrafinch PLP

L G H P D K F V G

I T Y V

L T

I

I W L L V F

S A V P

-

-

-

V Y

N T W T T C Q S

G N P T

-

K T S

Chicken PLP

L G H P D K F V G

I T Y V

L T

I

V W L L A F A C

S A V P

-

-

-

V Y

I Y F

N T W T T C Q S

I

G T L C A D A R M Y G V L P W N A F

Human PLP

L G H P D K F V G

I T Y A L T V V W L L V F

A C

S A V P

-

-

-

V Y

I Y F

N T W T T C Q S

I

A F

P S

-

K T S

A S

I

G S L C A D A R M Y G V L P W N A F

P

-

G K V C G S

N L L

Pig PLP

L G H P D K F V G

I T Y A L T V V W L L V F

A C

S A V P

-

-

-

V Y

I Y F

N T W T T C Q S

I

A F

P S

-

K T S

A S

I

G S L C A D A R M Y G V L P W N A F

P

-

G K V C G S

N L L

I

A C

-

-

-

I Y F

I I

A F

I

P T

-

-

A S

K T T A S

G T L C A D A R M Y G

P N T K L E D M K V C E K Y E I

G E E K K

I L P W N A F

P P

-

I C

T A S S

P

I

E N F

L

S D N F

L

G K V C G S G K V C G S

N L L N L L

I T Y A L T V V W L L V F

A C

S A V P

-

-

-

V Y

I Y F

N T W T T C Q S

A F

P S

-

K T S

G S L C A D A R M Y G V L P W N A F

P

-

G K V C G S

N L L

L G H P D K F V G

I T Y A L T

V W L L V F

A C

S A V P

-

-

-

V Y

I Y F

N T W T T C Q S

I

A F

P S

-

K T S

A S

I

G S L C A D A R M Y G V L P W N A F

P

-

G K V C G S

N L L

Cow PLP

L G H P D K F V G

I T Y A L T V V W L L V F

A C

S A V P

-

-

-

V Y

I Y F

N T W T T C Q S

I

A A P C

-

K T S

A S

I

G T L C A D A R M Y G V L P W N A F

P

-

G K V C G S

N L L

I

N L L

-

-

-

-

-

-

-

-

I T Y A L T V V W L L V F

Mouse DM20

-

Frog PLP

L G H P D K F V G V T Y V -

-

-

-

-

-

-

-

Y

-

-

-

-

-

-

-

-

-

V T Y A L V

-

Y

-

-

-

-

-

-

-

-

-

-

-

-

V

T

-

Lungfish DMalpha Mouse M6B

-

I

Trout DMalpha Shark DMalpha

I

I

L A V L

L W

I

I Y F

N T W T T C Q S

-

-

-

V Y

I Y F

N T W V

I P S

S

S

S

I W H R P A T T S

-

-

L Y

I F H N M W S

V Y

I Y Y T M W S

S A V P

A C

S A V P

A F

T A

I W L A

V

L

I I

F

A L G V T W M G V F

Y V

A C

F

I W L A V F

L G V A W L G V F

F T A L

A F

S A L

P P

G F S A V P

-

-

-

-

-

-

V Y L

V F M F Y N

I W S

I

A F

T C Q S M A F T

S W T

T C K A

P S

A S

G S L C A D A R M Y G V L P W N A F

P

-

E N S G R

-

E N G T G F D D

E L

T

K S

K T T T

I

-

T C Q M V K Y V I

K T S

P G -

T C E V

-

I

E T T P

I

A S

I

C

Dog PLP

Mouse PLP

L G H P D K F V G

L

I D L T Q F H F M F P

-

-

-

-

-

-

-

-

-

-

-

-

-

F L T

230 -

-

-

-

-

-

-

-

S V E H S

-

G

220

I Y T M F W N M C T

-

-

-

F

.

-

-

-

T

L S A T

-

-

-

F L V V V

F G D

-

Frog DMbeta/M6A

-

L C

200

.

-

Mouse M6A

Drosophila DM20/M6

I T Y L L N F V W S

.

F L T

-

P K G R S

E V

F

Y T T S

P A M K G G L

L G K W

S

I

S

I K H

-

T

F Q Y

I Y G T A S

F

Frog DMgamma2/M6B

I

Q V

I Y G

I Y G L A S

I

Mouse M6B

Shark DMalpha

V N V

H A F

I I

F

L E R V C H C

L G K W

Q Y V Q Y V

F

K G R G S R G Q H Q A H S

L E R V C H C

H A F H A F

F

A T V T G G Q

-

Y T T G A V R Q

I I

Q Y V

C G K G L S

A T V T G G Q

L Y G A L L L A E G F

N V N V

H A F

I

C G K G L S

I

I

G D Y K T T

I

I

I

I Y G T A S

G D Y R T T

S C A V L M G

Dog PLP

Mouse PLP

N V

F

I

L 170

F

I

R V G G R

.

160

F

N V

L A T G A T R Y K V Y R A W R S

E T Y F S

150

Frog DMbeta/M6A

I

L F V G F

F

Mouse M6A

Pig PLP

I

L C

130

Drosophila DM20/M6

I F

V

-

II

100

I

S

I

S V S

T L C L D A R M Y G V L P W N A F

N Q H G W

P Q S N G T S

I

I C M D A R Q Y G L

H Q L C V D A R Q Y G

G V E Q

I C V D A R Q Y G I C V D V R Q Y G

A T

I L P W N A S I

P

-

G K V C G S

P

-

G K V C G T S

L P W N A M P

I L P W S I

P W N A F

P

L L

-

G K A C G M T L A

-

G E V C G Y N L Q

P

-

G K

I

C G L

S

P

-

G K

I

C G S

A L E

L A

Frog DMgamma2/M6B

-

-

-

-

-

-

-

-

-

-

-

Y

I

L G V A W L G V F

G F S A

I P

-

-

-

V F M F Y N M W S

S

C E V

I

K A L P

T N L T T

T A D Q

I C V D

I

R Q F G

I

I

P W N A L

P

-

G K A C G Q A L E

Frog DMgamma1/M6B

-

-

-

-

-

-

-

-

-

-

-

Y

I

L G V A W L G V F

G F S A

I P

-

-

-

V F M F Y N M W S

S

C E

F

I

K

L P

T N L T T

T A D Q

I C V D

I

R Q F G

I

I

P W N A L

P

-

G K A C G Q A L E

-

-

-

-

-

-

-

-

-

.

T Y

.

L

-

-

-

V Y

W T T C Q S

I

.

P

-

.

L C

.

L P W N A F P

-

G K V C G

260 Drosophila DM20/M6

K A F C K D G V E N A E V M F

Mouse M6A

R M C E

S T

R M C E

S T

Frog DMbeta/M6A

V

. W L L V

F A

270

S A V 280

L S

T L L V

L L

P

IV S

.

310

D A R

Y G

320

330

L S

A N Y A H

I R D H E K F Q E L Q E

I

Q N L N E L E Y S A T S K D R F

I A M V H Y L M V

L S

A N W A Y V K D A C R M Q K Y E D

I

K S K E E Q E L H D

I

H S

T R S K E R L N A Y T

E L N M T F

H L F

I

V A L A G A G A A V

I A M V H Y L M V

L S

A N W A Y V K D A C R M Q K Y E D

I

K S K E E Q E L H D

I

H S

T R S K E R L N A Y T

E F

Q M T F

H L F

I

A A F

V G A A A T L V S

L V

E F

Q M T F

H L F

I

A A F

V G A A A T L V S

L L T F M

I C K T A E F

Q M T F

H L F

I

A A F

V G A A A T L V S

T F

I

L L T F M

I A T

T Y N F

I A A T Y N F I A A T Y N F

A V L K L M G R G T K F C A V L K L M G R G T K F

S

I C K T A E F

Q M T F

H L F

I

A A F

V G A A A T L V S

L L T F M

I A A T Y N F

A V L K L M G R G T K F

Mouse PLP

S

I C K T A E F

Q M T F

H L F

I

A A F

V G A A A T L V S

L L T F M

I A A T Y N F

A V L K L M G R G T K F

Q M T F

H L F

I

A A F

V G A A A T L V S

L L T F M

I A A T Y N F

A V L K L M G R G T K F

Cow PLP

S

I C K T A E F

Q M T F

H L F

I

A A F

V G A A A T L V S

L V

T F M

I A A T Y N F

A V L K L M G R G T K F

Mouse DM20

S

I C K T A E F

Q M T F

H L F

I

A A F

V G A A A T L V S

L L T F M

I A A T Y N F

A V L K L M G R G T K F

Q M T F

H L F

A A F

V G A A A T L V A L L T Y M V G A S

T Y D L Y

I

A A F

A G A G

Lungfish DMalpha

T V C R T R E F

G M T Y H L F

I

A T

F V G A G A T A V A L L T Y M M S

T T Y N F

Shark DMalpha

A V C N T S

A

I C K T S

S

I C K T K E F

E

I

A T

F A G A A A T V

I A L L T Y M M S

S

Mouse M6B

N

I

V A C A G A G A T V

I A L

L S

E F

E F

I C N T N E F

F

V

L T Y H L F

Y M S

Y H L F

I

I

A L L A L F

I

L Y V V A T

H F L M

Frog DMgamma2/M6B

Q

I C N S N E F

Y M S

Y H L F

I

V A C A G A G A T V

I A L L

I

Frog DMgamma1/M6B

Q

I C N S N E F

Y M S

Y H L F

I

V A C A G A G A T V

I A L

H F L M

.

I C K T

. M T F H L F

I

A A F V G A A A T L V

E F

I

I

Y M M A T

L L T F M

.

N L L

A V L R L M G R G T K

Pig PLP

I C K T A E F

.

340

L V H Y L M C

V A L A G A G A A V

I C K T S

Trout DMalpha

.

L A T

I C K T S

Frog PLP

.

I

S

S

.

I

S

Dog PLP

.

H L F

Chicken PLP

S

.

300

E L N M T F

Zebrafinch PLP Human PLP

I Y F N 290

I

F N Y A V L R V

T G R S

D R S K F

T Y N Y A V L R F L G R K G L R C A

I

V K

I

T Y N Y A V L K F S

L G R E D Y C

T K F

L S

T K F

N W A Y L K D A S

R D D C C

K M Q A Y Q D

T Y N Y A V L K F K S R E D C C

I

L S

A

.

S

N W A Y L K D A S

T Y N

.

A V

L K

.

G R

I

K A K E E Q E L Q D

I

Q S R S K E Q L N S Y T

K A K E E Q E L Q D

I

Q S R S K K Q

T K F

K M Q V Y Q D

I

I

N S Y T

K

FIGURE 16.2 Sequence comparison of PLP and the other lipophilins. ClustalW alignment (MacVector 7.0) is shown with boxes surrounding the areas of amino acid identity or similarity. The four transmembrane domains are numbered and highlighted in yellow. The six cysteine residues that serve as acylation sites in PLP are highlighted in blue and the four cysteine residues that form disulWde bridges are in green. The external glycosylation site (lys217) within the large extracellular loop of PLP is in orange. The three intron/exon junctions that are preserved between the PLP, M6A, M6B genes and their Drosophila counterpart are marked with arrows.

fatty acid attachment (Weimbs and StoVel, 1992) (Fig. 16.2). Although the acylation status of the DMb/M6A/EMA and DMg/M6B/Rhombex 29 family members has not yet been veriWed, this conserved pattern suggests that the fundamental hydrophobic character of these molecules is similarly enhanced by covalently linked lipids. The cysteines that contribute to the structure of the large extracellular loop of the PLP molecule through the formation of disulWde bridges are also remarkably closely aligned (Fig. 16.2), suggesting a common three-dimensional structure for this region of lipophilins. Another feature relating the primordial lipophilin gene discovered in Drosophila with the DMa/DM20/PLP, DMb/ M6A/EMA and DMg/M6B/Rhombex 29 is gene organization, as the intron-exon structure of the Drosophila gene closely mimics that of the vertebrate genes, although the Drosophila locus has only Wve exons (Stecca et al., 2000; Werner et al., 2001) (Fig. 16.2).

MOLECULAR EVOLUTION OF THE PLP GENE FAMILY IN VERTEBRATES AND INVERTEBRATES

A -S SExtracellular -S S-

I

II

III

IV

NH2 PLP-specific

Cytoplasmic

B -S SIntraperiod line -S S-

I

Major dense line

II

III

IV

PLP-specific

FIGURE 16.3 Topology of the lipophilins. (A) Schematic showing the orientation of PLP in the plasma membrane. The other lipophilin proteins (DMa/DM20, DMb/M6A/EMA and DMg/M6B/Rhombex 29) are assumed to have a similar topology, although they would lack the PLP-speciWc segment (dashed line). The four transmembrane domains are numbered and the covalently linked fatty acids are depicted as squiggles within the shaded membrane. The disulWde bridges in the extracellular loop are marked (-S S-), as is the conserved lysine, which can be glycosylated in vivo (open arrow). The M6 proteins also have glycosylated extracellular loops, although complex carbohydrates are attached to these proteins in the RER/Golgi. The site of protease cleavage (amino acids 105-112) deWned by Bizzozero et al. (2000) is shown with a Wlled arrow. (B) The StoVel model of PLP spanning adjacent lipid bilayers in compact myelin (Sporkel et al., 2002). The topology of PLP is the same as in the plasma membrane (panel A), but reversible acylation enables the intracytoplasmic loop of PLP to ‘‘cross’’ the major dense line and associate with the adjacent lipid leaXet via newly attached fatty acids. The major dense line represents the apposition of the cytoplasmic faces of oligodendrocyte plasma membrane, while the intraperiod line is formed by apposition of the extracellular faces.

Distinctive expression patterns are exhibited by the three family members (Tab. 16.1), implying that these proteolipids subserve diVerent cellular functions. The DMa/DM20/ PLP proteins are the major structural component of the myelin sheath in oligodendrocytes. But DM20/PLP expression is not limited to myelin-forming or ensheathing glial cells (Dickinson et al., 1997), as much smaller amounts of these proteins have been detected in immune cells of the thymus and spleen and in cardiac myocytes (Campagnoni et al., 1992; Pribyl et al., 1996). The DMb/M6A/EMA proteins are the neuronal counterpart of the lipophilin family, appearing in presumptive neurons in the gray matter of amphibians (Yoshida et al., 1999) and localized to the surface of axons, dendrites, and neuronal growth cones in mammals (Lagenaur et al., 1992; Baumrind et al., 1992), where M6A/EMA appears at the leading edge of lamellipodia (Sheetz et al., 1990). In the frog, DMb is

407

408

16. PROTEOLIPID PROTEIN GENE

Drosophila DM20/M6 Mouse M6A 0.024 Frog DMbeta/M6A

0.048 0.494 0.015

Human PLP

0.001 0.081 0.12

0.031

Zebrafinch PLP 0.012 Chicken PLP

Pig PLP

0.002 Mouse PLP

0.023 0.022 0.049

Dog PLP

0.083

Cow PLP Mouse DM20

0.26

Frog PLP Trout DMalpha

0.182

Lungfish DMalpha Shark DMalpha

0.084

Mouse M6B

Frog DMgamma2/M6B 0.074 Frog DMgamma1/M6B

FIGURE 16.4 Phylogenetic tree for lipophilins. Phylogeny was reconstructed using the neighbor joining method of ClustalW aligned sequences including a Poissoncorrection. Calculated evolutionary distances show the closeness of the primordial Drosophila gene to the M6A genes.

exclusively expressed in grey matter and in the inner nuclear and ganglion cell layers of the retina (Yoshida et al., 1999). Retinal expression has likewise been noted in the mouse, particularly in the inner plexiform layer and in the endfeet of Muller cells (Mi et al., 1998). Apart from marking CNS neurons, M6A/EMA proteins also appear in transporting epithelial cells such as the proximal tubular cells of the kidney and the choroid plexus (Baumrind et al., 1992; Lagenaur et al., 1992). Very low amounts of M6A have also been isolated from myelin fractions (Klugmann et al., 1997). The DMg/M6B/Rhombex 29 proteins combine the expression patterns of the Wrst two family members, as they are strongly expressed by both oligodendrocytes and neurons (Shimokawa and Miura, 2000; Vouyiouklis et al., 1998; Werner et al., 2001; Yan et al., 1996; Yoshida et al., 1999). In amphibians, DMb and DMg are each expressed in the developing retina, yet their distributions are distinct (Yoshida et al., 1999). DMg is also found within the ventricular zone in tadpoles and mature frogs and in myelinated tracts of the developing spinal cord in the shark (Kitagawa et al., 1993). Analogous to the distribution of the DMb/M6A protein in growth cone membranes, DMg can be detected in the expanded endfeet of radial glial cells in Xenopus (Yoshida et al., 1999). In oligodendrocytes, M6B is found predominantly at the cell surface and intracellular reticular structures, with small quantities of M6B extractable from myelin (Klugmann et al., 1997; Werner et al., 2001). Widespread tissue expression of M6B transcripts, albeit at low levels, has also been described (Werner et al., 2001). Does the lipophilin family of tetraspan proteins form adhesive pore structures, as originally proposed for the DMa/DM20/PLP proteins (Kitagawa et al., 1993)? Phylogenetic inspection of the transmembrane regions certainly argue for a channel or pore structure, as does the similarities of these lipophilin regions to known channel forming

TRANSLATIONAL AND POST-TRANSLATIONAL REGULATION OF PLP AND DM20

TABLE 16.1 Expression Patterns of the Lipophilin Family Lipophilin

Neural cell/subcellular localization

Non-neural tissues

DMa/DM20/PLP DMa Oligodendrocytes/myelin

N.D.

DM20

Oligodendrocytes/myelin Olfactory ensheathing cells/soma Myelinating Schwann cells/soma Nonmyelinating Schwann cells/soma

Spleen, thymus, cardiac myocytes

PLP

Oligodendrocytes/myelin Myelinating Schwann cells/soma

Spleen, thymus

DMb/M6A/EMA DMb

Gray matter, retina

N.D.

M6A/EMA

Neurons/axons, dendrites, growth cones Retina-Muller cells/endfeet Oligodendrocytes/myelin

Kidney proximal tubular cells, choroid plexus

DMg/M6B/Rhombex29 DMg

Cells of ventricular zone, central canal, retina, radial glia (amphibians) White matter (shark)

N.D.

M6B

Cells of ventricular zone, sciatic nerve Neurons/cell surface and intracellular reticular structures Oligodendrocytes/cell surface, myelin, intracellular reticular structures

Liver, muscle, kidney, spleen, lung, heart

Rhombex29

Selected neurons (including ventral medullary surface neurons)

Lung, kidney

N.D. ¼ Not detected.

domains (Kitagawa et al., 1993). The position of DMb/M6A at the leading surface of neuronal growth cones, transporting epithelial cells and at Muller cell endfeet are also consistent with a channel role. Moreover, one family member, Rhombex29, is expressed by the chemosensitive neurons of the ventral medullary surface that sense excess CO2/Hþ in cerebrospinal Xuid (Shimokawa and Miura, 2000). The increased expression of Rhombex29 in response to altered CO2/Hþ levels suggests that Rhombex may transport Hþ, a possibility that invites further electrophysiologic experiments.

TRANSLATIONAL AND POST-TRANSLATIONAL REGULATION OF PLP AND DM20 Synthesis of PLP/DM20 in the Rough Endoplasmic Reticulum The orientation of PLP and DM20 in the endoplasmic reticulum and plasma membrane has been deduced from a variety of experimental approaches, most deWnitively by combined glycosylation scanning, immunocytochemical epitope mapping, and protease domain protection assays (Gow et al., 1997; Wahle and StoVel, 1998). PLP and DM20 were discovered to possess four transmembrane domains, with both amino and carboxy termini exposed to the cytoplasm (Fig. 16.3). This orientation is favored by theoretical considerations (Popot et al., 1991) and is consistent with the model based on the positions of the free, disulWdebonded, and acylated cysteine residues in PLP (Weimbs and StoVel, 1992). While the topology of PLP and DM20 are identical, intriguing diVerences in the traYcking of PLP and DM20 proteins containing various missense mutations indicates that the conformation of these two protein isoforms diVers (Gow et al., 1997). The DM20 protein is largely unaVected by a number of missense mutations, while the same mutations introduced into the PLP protein prevent passage to the cell surface (Gow and Lazzarini, 1996; Gow et al.,

409

410

16. PROTEOLIPID PROTEIN GENE

au4

1994; 1997,1998). Among the mutations that diVerentially act on DM20 and PLP, of particular interest are those located in the large extracellular domain, on the other side of the membrane as the extra cytosolic segment unique to PLP. These mutations hint at a transfer of information across the lipid bilayer mediated by the PLP-speciWc domain behaving as a ‘‘conformation sensor.’’ As proposed by Lazzarini and colleagues, the introduction of the PLP-speciWc segment in tetrapods may have evolved as a mechanism for sensing and responding to extracellular cues during myelin membrane compaction (Gow et al., 1997). Some unresolved issues linger in the bevy of contradictory results on the topology of the DM20 and PLP proteins. Part of the confusion may be attributed to limitations of the experimental approaches, which in addition to glycosylation scanning (Gow et al., 1997; Wahle and StoVel, 1998) included photoaYnity labeling (Kahan and Moscarello, 1985), X-ray diVraction (Inouye and Kirschner, 1989), protease analysis of hyposmotically shocked myelin (StoVel et al., 1984), and immunocytochemical epitope mapping (Greer et al., 1996; Hudson et al., 1989; Konola et al., 1992; Sobel et al., 1994; StoVel et al., 1989). The cotranslational insertion of PLP into the endoplasmic reticulum membrane in vitro led to two diVerent topologies in glycosylation scanning experiments, one with both termini on the cytoplasmic face as described earlier (Fig. 16.3), and another with an opposite orientation (i.e., both termini on the lumenal/extracellular face of the membrane) (Wahle and StoVel, 1998). Whether this phenomenon occurs in myelinating cells is unknown, but could explain some of the disparities between studies. The speculation that PLP/DM20 may be synthesized in both orientations in the RER membrane, yet only molecules with the ‘‘appropriate’’ topology (i.e., both termini on the cytosplasmic face) are transported to the cell surface (Wahle and StoVel, 1998), is an interesting one. Proper folding of PLP/DM20 in the endoplasmic reticulum is essential for subsequent transport. Misfolded PLP triggers an apoptotic response and represents the underlying pathogenic mechanism of missense mutations in the PLP gene (Gow et al., 1994b,1998; Gow and Lazzarini, 1996; Jung et al., 1996; Southwood and Gow, 2001; see Chapters 37 and 42). Multimerization is another processing step that probably occurs within the endoplasmic reticulum. The tendency for PLP molecules to physically interact in vitro is a prominent feature of this protein (reviewed in Lees et al., 1979; Lees and BrostoV, 1984), which suggests that dimer and polymer formation may occur shortly following synthesis.

Trafficking of PLP and DM20

au5

au6

During transit through the Golgi complex, PLP selectively joins cholesterol and galactosylceramide-rich membrane domains in so-called ‘‘rafts,’’ small and dynamic structures formed by self-association of particular lipids (e.g., glycophingolipids and cholesterol) that behave as platforms for the recruitment of a speciWc set of proteins (Simons et al., 2000). This initial stage of myelin assembly can be blocked by either cholesterol depletion or, in the case of the ceramide galactosyl transferase (CGT) knockout mice, the absence of galactosylceramide and sulfatide. PLP still manages to reach the plasma membrane in these situations, as it does when transfected into nonglial cells that lack myelin lipids (Gow et al., 1994a; Gow and Lazzarini et al., 1996; Thomson et al., 1997). When overexpressed, PLP fails to associate with myelin rafts in the Golgi complex (Simons et al., 2002). Instead, PLP accumulates in late endosomes/lysosomes, where the overexpressed protein traps cholesterol and possibly other myelin lipids (Simons et al., 2002). The ensuing disruption of the myelin raft pathway represents a second pathogenetic mechanism for how toxic ‘‘gain-of-function’’ PLP mutations can create Pelizaeus Merzbacher disease (see Chapter 37). In the original mechanism proposed by Gow and Lazzarini (1996), the large amounts of improperly folded mutant PLP and DM20 proteins accumulating in the RER unleashed an apoptotic response. In this case, duplication of the PLP gene leads to overexpression of PLP and DM20 proteins that sequester myelin lipids and thereby perturb myelin assembly. Together, these studies suggest that myelin assembly begins shortly after PLP synthesis and while assembly into raft structures is not essential for transport, the integrity of compact myelin may be adversely aVected by improper ratios of myelin constituents in the Golgi complex.

au3

TRANSLATIONAL AND POST-TRANSLATIONAL REGULATION OF PLP AND DM20

Do PLP and DM20 interact with each other during synthesis, transport, myelin assembly, or compaction? Each protein isoform can be independently transported to the plasma au7 membrane of transfected cells (Gow et al., 1994a; Gow and Lazzarini et al., 1996; Thomson et al., 1997). Nonetheless, facilitated traYcking of mutant PLP proteins by their cognate DM20 proteins is correlated with less severe forms of disease (Gow and Lazzarini, 1996), which supports a report that membrane insertion of PLP is strongly inXuenced by the presence of DM20 (Sinoway et al., 1994). Further investigation by GriYths and colleagues harnessed transgenic technology in which knockout mice were reconstituted with either a PLP or DM20 transgene to demonstrate that an individual isoform can be incorporated into compact myelin (McLaughlin et al., 2002). Although these studies indicate that interactions between PLP and DM20 are not a prerequisite for transport or assembly, they leave open the question of whether the two isoforms normally come in contact with each other during these stages of myelination. Following compaction of the myelin sheath, there is biochemical evidence that PLP and DM20 directly interact. The two protein isoforms PLP and DM20 can be co-immunoprecipitated from isolated myelin, and chromatography yields heteromeric complexes of PLP and DM20 (McLaughlin et al., 2002). In transgenic mice expressing half the normal PLP protein but normal levels of DM20, the altered ratio of PLP to DM20 in myelinating oligodendrocytes has no discernible aVect on myelin compaction or the maintenance of axonal integrity (Uschkureit et al., 2001). Perhaps suYcient amounts of heteromeric complexes are still able to form in these transgenic mice that have an altered composition of PLP and DM20 proteins. In any case, the capacity for PLP and DM20 to form homo- and heteropolymers provides further support for a pore or channel function of these integral membrane proteins.

Post-Translational Modifications of PLP and DM20 The hydrophobic character of PLP/DM20 is further enhanced by the covalent linkage of long chain fatty acids. DM20 is missing two of the six acylation sites present in PLP (Fig. 16.2), a loss that would reduce the hydrophobicity of DM20 compared to PLP. The attachment of fatty acids via thioester linkages to speciWc intracellular cysteine residues of PLP/DM20 is not coupled with protein synthesis and occurs autocatalytically with acylCoA esters as donor molecules (Bizzozero et al., 1987; Ross and Braun, 1988). Deacylation is an active enzymatic process catalyzed by a speciWc myelin-associated fatty acylesterase (Bizzozero et al., 1992). The majority of PLP and DM20 molecules appear to be maximally acylated (six and four molecules, respectively, of fatty acid/molecule protein) when analyzed by matrix-assisted laser desorption ionization time of Xight-mass spectrometry (MALDI-TOF-MS), while chemical analysis yields somewhat lower values (Bizzozero et al., 2002; Weimbs and StoVel, 1992). The predominant fatty acid incorporated into PLP is palmitate. The fatty acid makeup of PLP/DM20 from amphibians, reptiles, birds, and mammals is similar in amount (about 3% w/w) and composition (palmitic, palmitoleic, oleic, and stearic acids), illustrating the constancy of this post-translational modiWcation during evolution (Bizzozero and Lees, 1999). The strict conservation of the four cysteine residues that represent the fatty acid attachment sites (Fig. 16.3) point to the possibility for similar modiWcations in lower organisms. The acyl moieties aVect the adhesive properties of PLP/DM20, as chemical deacylation leads to decompaction of the myelin sheath (Bizzozero et al., 2001). Another post-translational modiWcation of PLP appears to be proteolytic cleavage, a process that targets the intracellular loop and splits the protein into an amino terminal half (amino acids 1-105/112) and a carboxy terminal half (amino acids 113/131–276) (Bizzozero et al., 2002; Lepage et al., 1986; McLaughlin et al., 2002). Minor low molecular weight forms of PLP had been observed before, but they tended to be viewed as possible postmortem artifacts. Demonstration that the equimolar amounts of these two cleavage product do not increase upon post-mortem storage, and that they are diVerentially distributed across the CNS (Bizzozero et al, 2002), together with the report that fragments of PLP can be secreted and exhibit biologic activity (Nakao et al., 1995; Yamada et al., 1999), indicates proteolytic cleavage may be an important step in regulating PLP activity.

411

412

16. PROTEOLIPID PROTEIN GENE

The extracellular domain of PLP is subject to post-translational modiWcation in the form of nonenzymatic glycosylation (Weimbs and StoVel, 1994). The high degree of reactivity of the lysine217 in the extracellular loop of PLP (Fig. 16.3) toward carbonyl compounds led to the discovery of PLP as the most highly nonenzymatically glycosylated brain protein, a modiWcation exacerbated by a hyperglycemic state such as presented by diabetic mice (Weimbs and StoVel, 1994). The M6 proteins are also glycosylated, although in this case complex carbohydrates are enzymatically attached during transit through the endoplasmic reticulum-Golgi compartment, as M6/EMA can be eVectively reduced in size with either neuraminidase (Lagenaur et al., 1992) or endoglycosidase F (Baumrind et al., 1992) treatment.

FUNCTIONS OF PLP AND DM20 Clues to the function of PLP and DM20 have come from analyzing expression patterns, assessing the impact of mutations (discussed in more detail in Chapters 37, 42, and 47), au8 exploring nontraditional roles for membrane proteins, probing biochemical properties, and integrating phylogenetic information. These combined avenues have toppled a longstanding view that the abundantly produced, integral membrane proteins PLP and DM20 are essential for the formation of a myelin sheath and required to Wll out the bulk of the intraperiod line in compact myelin.

Expression Patterns of PLP/DM20: Not Just for Oligodendrocytes Originally considered as one of the deWning characteristics of the oligodendrocyte, PLP/ DM20 expression was extended to the other glial cell types, including myelin-forming and non-myelin-forming Schwann cells, as well as satellite cells of the dorsal root, cranial nerve, and autonomic ganglia (GriYths et al., 1989,1995; Puckett et al., 1987). Non-neural tissues also express PLP/DM20. Both isoforms are present in fetal thymus and spleen (Pribyl et al., 1996), and DM20 is found in the heart (Campagnoni et al., 1992) (see Tab. 16.1). The DM20 pattern does not coincide with myelin formation. The DM20 transcript is selectively expressed in early development (Ikenaka et al, 1992; Timsit et al., 1992), although whether these transcripts are translated is still an open question. Nonetheless, DM20 is the more abundant protein isoform in immature oligodendrocytes (Dickinson et al., 1997; Ikenaka et al., 1992; Schindler et al., 1990; Timsit et al., 1992,1995; Yu et al., 1994). DM20 is also the predominant form in Schwann cells, both myelinating and nonmyelinating (GriYths et al., 1995; Ikenaka et al., 1992; Pham-Dinh et al., 1991), while PLP is restricted to myelinating Schwann cells in the PNS. Another glial cell type that cannot make myelin, the olfactory ensheathing cell, likewise selectively expresses the DM20 isoform (Dickinson et al., 1997). When the PLP/DM20 gene is activated in cells outside of the nervous system, the DM20 isoform usually prevails, as in myocardial cells (Campagnoni et al., 1992). The distinctive developmental and cellular expression pattern of DM20 was the Wrst clue that this protein isoform is functionally distinct from PLP.

Mutational Analysis of PLP: No PLP, No Long-Term Compaction The overarching phenotype of the PLP/DM20 missense mutants, namely oligodendrocyte cell death and its attendant sequellae, makes this class of mutants less informative in addressing the function of PLP. Which phenotypes are a consequence of dysfunctional PLP/DM20 protein, and which are merely due to apoptosis of oligodendrocytes? An exception is the rumpshaker mouse, whose mild phenotype originally showed that oligodendrocyte cell death is not an obligate outcome of PLP missense mutations (Schneider et al., 1992). Supplying a wild-type PLP transgene to rumpshaker mice improved myelin structure yet failed to ameliorate the dysmyelination, an outcome that reinforced the notion that the PLP and DM20 isoforms may perform distinctive roles in myelin structure and oligodendrocyte development (Nadon and West, 1998).

FUNCTIONS OF PLP AND DM20

The loss-of-function mutants provide the best handle on the roles of PLP/DM20

au9 (Chapters 37 and 47). Gene targeting of the PLP locus has generated mice lacking both PLP and DM20 (Boison and StoVel, 1994; Boison et al., 1995; Klugmann et al., 1997), as well as mice selectively deWcient in the PLP isoform (Sporkel et al., 2002; Stecca et al., 2000; Uschkureit et al., 2001). In each transgenic model, oligodendrocyte development is unperturbed. Perhaps more surprisingly, and in deWance of expectations raised by the sheer abundance of PLP in the myelin sheath, oligodendrocytes of the null mice manage to assemble a myelin sheath. These myelin sheaths are not without idiosyncracies. Mice without PLP and DM20 have defects in the intraperiod line, the structure formed by apposition of the extracytoplasmic surfaces of myelin membrane, that translate into reduced conduction velocities and impaired neuromotor coordination (Boison and StoVel, 1994; Rosenbluth et al., 1996). These anomalies in the intraperiod line of PLP/DM20 null mice have been conWrmed (Yool et al., 2002). Still controversial is the status of compact myelin in mice expressing only DM20, in which the periodicity has been reported as expanded (Stecca et al., 2000) or unchanged (Sporkel et al., 2002). An intriguing feature of the transgenic mice expressing solely DM20 is the acquisition of cytoplasmic inclusions within compact myelin (Sporkel et al., 2002), which might be expected to eventually increase the myelin periodicity. This observation lends credence to a topological model in which the intracytoplasmic loop of PLP and DM20 can associate with an adjacent myelin bilayer via its covalently linked fatty acyl chains (the StoVel model, shown in Fig. 16.3B). Three fatty acid attachment sites are present in this loop on the PLP protein (at cys108, cys138, and cys140), while DM20 proteins have only a single fatty acid site (cys108). A solo anchoring site may be insuYcient for stably linking adjacent myelin bilayers, especially in the face of the dynamic processing of factty acyl groups, which are nonenzymatically transferred and enzymatically released (Bizzozero et al., 1987). Thus, myelin composed solely of the DM20 isoform would accumulate cytoplasmic pockets and display increasing myelin periodicity, as observed by the StoVel and Gow groups (Sporkel et al., 2002; Stecca et al., 2000). These results document the individuality of the PLP and DM20 isoforms: DM20 is unable to replace the function(s) of PLP in oligodendrocytes. Moreover, the StoVel model provides a mechanism for PLP-mediated adhesion through reversible acylation (Sporkel et al., 2002). Oligodendrocytes develop normally in both the PLP
and the double PLP
DM20
mice, but age-related degeneration occurs to a certain extent in both models (Boison and StoVel, 1994; Boison et al., 1995; Klugmann et al., 1997; Sporkel et al., 2002; Stecca et al., 2000; Yool et al., 2001). Disturbances in the usual complement of myelinated axons are obvious in the PLP
DM20
knockout, with a selective loss of compact myelin sheaths noted for the small-diameter axons (Yool et al., 2001). Mice that express only DM20 are spared the axonal swellings and defects in the myelination of small diameter axons seen in the doubly deWcient (PLP
DM20
) mice (Boison et al., 1995; GriYths et al., 1998). These results do not obviate a role for PLP and/or DM20 in developing oligodendrocytes, because of the possible redundant action of other lipophilin family members. Oligodendrocytes do synthesize the DM20-like molecules M6B, and at very low levels M6A, whose expression more than doubles in mice lacking PLP/DM20 (Klugmann et al., 1997). These lipophilins may take on the role(s) of DM20 in early development. However, together the M6 proteins are unable to functionally compensate for the loss of DM20/PLP, at least in the myelin sheath (Klugmann et al., 1997). Redundancy may also operate outside of the nervous system. M6B is expressed to a limited extent in most tissues (Werner et al., 2001) and may possibly substitute for PLP/DM20 in nonmyelinating cells, as none of the other type of cells known to express PLP/DM20 are noticeably aVected by mutations at the PLP locus (Boison and StoVel, 1994; Boison et al., 1995; Klugmann et al., 1997; Sporkel et al., 2002; Stecca et al., 2000; Uschkureit et al., 2001).

Secreted Forms of PLP: Another Avenue of Influence? Can an integral membrane protein function in an autocrine or paracrine manner? Support for the possibility that PLP/DM20 molecules or fragments are secreted originated from

413

414

16. PROTEOLIPID PROTEIN GENE

studies in which conditioned media from cells transfected with PLP or DM20 increased oligodendrocyte cell number (Nakao et al., 1995). Enhanced growth extended to jimpy oligodendrocytes (Knapp et al., 1999). The responsible factor was characterized as a carboxy terminal fragment of PLP, whose eVects could be mimicked by picomolar quantities of synthetic peptide (PLP215–232) (Yamada et al., 1999). This region would be included in the proteolytic cleavage product (PLP113/131–276) characterized by Bizzozero and coworkers (2002). PLP may also aVect neuronal growth, as dorsal root ganglia neurons degenerated when cocultured with PLP-expressing cells or media from PLPexpressing cells (Boucher et al., 2002). Since neuronal cell death was associated with PLP-expressing cells, not DM20 expressors, this experiment implicates PLP-speciWc products as responsible for the negative eVects on neuronal survival. However, acidiWcation of the media alone, an event that also occurred with PLP-expressing cells but not DM20expressing cells, was suYcient in these experiments to aVect neuronal viability. In this case, the eVects of secreted PLP products may be secondary to PLP-induced changes in the ionic milieu (Boucher et al., 2002). The in vivo functional relevance of these secreted fragments, along with their mode of secretion, has yet to be established. The phenotypes of the transgenic knockouts in PLP and DM20 include age-related defects in oligodendrocytes and neurons that leave open the possibility of signalling relayed by PLP/DM20 secreted fragments (Anderson et al., 1997; Boison et al., 1995; GriYths et al., 1998). Neuronal pathologies au10 are also apparent in PLP/DM20 overexpressors (Anderson et al., 1997) and in some PLP/ au11 DM20 missense mutants, where altered signalling between oligodendrocytes and neurons produces abnormal neurite outgrowth (Nadon, 2000). But intertwined with any action of secreted PLP products in these transgenic mice are the eVects that membranous PLP and its associated abnormal compact myelin confers on the neuronal cytoskeleton (reviewed by Witt and Brady, 2000). Sorting out indirect and direct eVects will require au12 new transgenic approaches targeted at selectively mutating the protease cleavage site in the PLP gene or interfering with the processing and transport machinery of secreted PLP peptides.

PLP as a Myelin Membrane ‘‘Sensor’’: Are PLP/DM20 Membrane Pore Proteins?

au13

Classical studies in reconstituted membranes have indicated that PLP can behave as a pore protein, allowing the uni- or bidirectional transfer of various ions (e.g., protons, potassium, and sodium) (de Cozar et al., 1987; Helynck et al., 1983; Ting-Beall et al., 1979). This possibility was supported by the decreased pH of media from PLP but not DM20 expressing cells, suggesting that PLP may alter the extracellular ionic environment (Boucher et al., 2002). The resemblance of PLP/DM20 to channel proteins (Kitagawa et al., 1993) and the evolutionarily conserved nature of the transmembrane domains (Fig. 16.2) characteristic of channels whose membrane a helices pack together lend credence to channel functions for lipophilin family members. As channel proteins, PLP and DM20 could sense changes in ion distribution across the myelin sheath. Yet only PLP has the potential for stably linking adjacent myelin bilayers (Fig. 16.3B), given the additional two fatty acid attachment sites in the cytoplasmic loop that are postulated to intercalate into the opposing bilayer leaXet (Sporkel et al., 2002). Analysis of mutations that interfere with PLP/DM20 traYcking also points to this PLPspeciWc segment containing the additional acylation sites as a transmitter of conformational information across the bilayer (Gow et al., 1997). Another property that highlights a ‘‘sensor’’ role of PLP is its interactions with signaling molecules. Unlike DM20, PLP coprecipitates with inositol hexakisphosphate (Yamaguchi et al., 1996). Perhaps PLP not only maintains an ordered myelin structure by pnysically linking adjacent bilayers, but also contributes to the ionic equilibrium of the sheath by forming pores and relaying information to the inositol hexakiphosphate signalling pathway. Since PLP and DM20 proteins interact in myelin (McLaughlin et al., 2002), multimers composed of PLP and DM20 proteins could sense and transmit information.

PROSPECTUS

PROSPECTUS Advances of the past decade have reWned our understanding of how PLP and DM20 evolved to form the structural underpinning of the myelin sheath. The analysis of mutations at the PLP locus has revealed mechanisms by which PLP/DM20 contributes to the structure of myelin sheath, yielded additional insights in the roles PLP/DM20 may play in promoting interactions between axons and oligodendrocytes, and documented the individuality of the PLP and DM20 proteins. But transgenic analysis can only go so far in understanding the molecular details of PLP action. Optimally, the mutation analysis must be coupled with more biochemical and electrophysiologic approaches, particularly to document the channel/pore properties of PLP/DM20. If PLP/DM20 are channels, what ions do they transport in vivo? Does PLP both sense and respond to changes in ion distribution? Regarding the proposed bilayer leaXet-spanning role for PLP, how are the acylation and deacylation steps regulated? Does PLP straddle both the major dense and intraperiod lines, the major dense line through the fatty acids attached to the intracytoplasmic loop and the intraperiod line through homo or heterophilic protein interactions? Given that PLP displaced the homophilic adhesive protein P0 from CNS myelin, the possibility that the extracellular domains of PLP/DM20 interact with one another to stabilize the intraperiod line seems reasonable. The accumulating data of PLP’s pivotal role(s) in oligodendrocytes brings up the need for comprehensive controls on PLP and DM20 synthesis. What forces regulate PLP/DM20 transcription and translation? Many of the trans-acting factors that bind to cis elements in the PLP gene have yet to be identiWed, although expression proWling eVorts should reveal the set of transcription factors that act on the PLP locus in oligodendrocytes and in nonexpressing cells. The splicing factors that favor the production of PLP mRNA and the RNA-binding proteins that selectively stabilize the PLP mRNA are also awaiting discovery. Recent experiments have uncovered the dynamic nature of the myelin sheath and the complex intermingling of the array of myelin lipids and proteins such as PLP. Continued exploration of this essential myelin gene should shed further light on how oliogdendrocytes evolved their spectacular spirals.

References Anderson, T. J., Montague, P., Nadon, N., Nave, K. A., and GriYths, I. R. (1997). ModiWcation of Schwann cell phenotype with Plp transgenes: Evidence that the PLP and DM20 isoproteins are targeted to diVerent cellular domains. J Neurosci Res.50, 13–22. Armstrong, R., Kim, J., and Hudson, L. D. (1995). Expression of Myelin transcription factor 1 (Myt1), a ‘‘ZincFinger’’ DNA binding protein, in developing oligodendrocytes. Glia 14, 303–321. Baumgartner, B. G., Deppe, A., Rettenberger, G., Leeb, T., Hameister, H., and Brenig, B. (1999). Molecular analysis of the porcine proteolipid protein (PLP) gene. Mamm Genome 10(9), 895–899. Baumrind, N. L., Parkinson, D., Wayne, D. B., Heuser, J. E., and Pearlman, A. L. (1992). EMA: A developmentally regulated cell-surface glycoprotein of CNS neurons that is concentrated at leading edge of growth cones. Dev Dyn. 194(4), 311–325. Barres, B. A., Lazar, M. A., and RaV, M. C. (1994). A novel role for thyroid hormone, glucocorticoids and retinoic acid in timing oliogdendrocyte development. Development 120, 1097–1108. Bellefroid, E. J., Bourguignon, C., Hollemann, T., Ma, Q., Anderson, D. J., Kintner, C., and Pieler, T. (1996). X-Myt1, a Xenopus C2HC-Type zinc Wnger protein with a regulatory function in neuronal diVerentiation. Cell 87, 1191–1202. Berndt, J., Kim, J., and Hudson, L. D. (1992). Cis regulatory elements required for expression of the myelin proteolipid protein gene. J. Biol. Chem. 267, 14730–14737. Berndt., J. A., Kim, J. G., Tosic, M., Kim, C., and Hudson, L. D. (2001). The transcriptional regulator Yin Yang 1 activates the myelin PLP gene. J. Neurochem. 77, 935–942. Billon, N., Tokumoto, Y., Forrest, D., and RaV, M. (2001). Role of thyroid hormone receptors in timing oligodendrocyte diVerentiation. Dev Biol. 235(1), 110–120. Bizzozero, O. A., Bixler, H. A., Davis, J. D., Espinosa, A., and Messier, A. M. (2001). Chemical deacylation reduces the adhesive properties of proteolipid protein and leads to decompaction of the myelin sheath. J Neurochem. 76(4),1129–1141. Bizzozero, O. A., and Lees, M. B. (1999). Fatty acid composition of myelin proteolipid protein during vertebrate evolution. Neurochem Res. 24(2), 269–274.

415

416

16. PROTEOLIPID PROTEIN GENE

Bizzozero, O. A., Leyba, J., and Nunez, D. J. (1992). Characterization of proteolipid protein fatty acylesterase from rat brain myelin. J Biol Chem. 267(11), 7886–7894. Bizzozero, O. A., Malkoski, S. P., Mobarak, C., Bixler, H. A., and Evans, J. E. (2002). Mass-spectrometric analysis of myelin proteolipids reveals new features of this family of palmitoylated membrane proteins. J Neurochem. 81(3), 636–645. Bizzozero, O. A., McGarry, J. F., and Lees, M. B. (1987). Autoacylation of myelin proteolipid protein with acyl coenzyme A. Biol Chem. 262(28), 13550–13557. Bogazzi, F., Hudson, L. D., and Nikodem, V. M. (1994). A novel heterodimerization partner for thyroid hormone receptor. J. Biol. Chem. 269, 11683–11686. Boison, D., Bussow, H., D’Urso, D., Muller, H. W., and StoVel, W. (1995). Adhesive properties of proteolipid protein are responsible for the compaction of CNS myelin sheaths. J Neurosci. 15(8), 5502–5513. Boison, D., and StoVel, W. (1994). Disruption of the compacted myelin sheath of axons of the central nervous system in proteolipid protein-deWcient mice. Proc Natl Acad Sci U S A. 91(24), 11709–11713. Bongarzone, E. R., Campagnoni, C. W., Kampf, K., Jacobs, E. C., Handley, V. W., Schonmann, V., and Campagnoni, A. T. (1999). IdentiWcation of a new exon in the myelin proteolipid protein gene encoding novel protein isoformsthat are restricted to the somata of oligodendrocytes and neurons. J Neurosci. 19(19), 8349–8357. Boucher, S. E. M., Cypher, M. A., Carlock, L. R., and SkoV, R. P. (2002). Proteolipid protein gene modulates viability and phenotype of neurons. J. Neurosci 22, 1772–1783. Britsch, S,. Goerich, D. E., Riethmacher, D., Peirano, R. I., Rossner, M., Nave, K. A., Birchmeier, C., and Wegner M. (2001). The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15(1), 66–78. Campagnoni, C. W., Garbay, B., Micevych, P., Pribyl, T., Kampf, K., Handley, V. W., and Campagnoni, A. T. (1992). DM20 mRNA splice product of the myelin proteolipid protein gene is expressed in the murine heart. J Neurosci Res. 33(1), 148–155. Chandross, K. J., Champagne, B., Auble, K., and Hudson, L. D. (2001). Tracking oligodendrocytes during development and regeneration. Microscopy Res. Tech. 52, 766–777. de Cozar, M., Lucas, M., and Monreal, J. (1987). Ionophoric properties of the proteolipid apoprotein from bovine brain myelin. Biochem Int 14, 833–841. Dickinson, P. J., GriYths, I. R., Barrie, J. M., Kyriakides, E., Pollock, G. F., and Barnett, S. C. (1997). Expression of the dm-20 isoform of the plp gene in olfactory nerve ensheathing cells: Evidence from developmental studies. J Neurocytol. 26(3): 181–189. Diehl, H. J., Schaich, M., Budzinski, R. M., and StoVel, W. (1986). Individual exons encode the integral membrane domains of human myelin proteolipid protein. Proc Natl Acad Sci U S A. 83(24), 9807–9811. Folch, J., and Lees, M. B. (1951). Proteolipids, a new type of tissue lipoproteins: Their isolation from brain. J. Biol. Chem. 191, 807–817. Franz, T., Waehneldt, T. V., NeuhoV, V., and Wachtler, K. (1981). Central nervous system myelin proteins and glycoproteins in vertebrates: A phylogenetic study. Brain Res. 226(1–2), 245–258. Fuss, B., Mallon, B., Phan, T., Ohlemeyer, C., KirchhoV, F., Nishiyama, A., and Macklin, W. B. (2000). PuriWcation and analysis of in vivo-diVerentiated oligodendrocytes expressing the green Xuorescent protein. Dev Biol. 218(2), 259–74. Geltner, D., Kitagawa, K., and Yoshida, M. (1998). Remarkable diversity of proteolipid proteins in Wsh. J Neurosci Res. 54(2), 289–295. Gow, A. (1997). RedeWning the lipophilin family of proteolipid proteins. J Neurosci Res. 50(5),659–64. Gow, A., Friedrich, V. L. Jr., and Lazzarini, R. A. (1994a). Intracellular transport and sorting of the oligodendrocyte transmembrane proteolipid protein. J Neurosci Res. 37(5), 563–573. Gow, A., Friedrich, V. L. Jr., and Lazzarini, R. A. (1994b). Many naturally occurring mutations of myelin proteolipid protein impair its intracellular transport. J Neurosci Res. 37(5), 574–583. Gow, A., Gragerov, A., Gard, A., Colman, D. R., and Lazzarini, R. A. (1997). Conservation of topology, but not conformation, of the proteolipid proteins of the myelin sheath. J Neurosci. 17(1), 181–199. Gow, A., and Lazzarini, R. A. (1996). A cellular mechanism governing the severity of Pelizaeus-Merzbacher disease. Nat Genet. 13(4), 422–428. Gow, A., Southwood, C. M., and Lazzarini, R. A. (1998). Disrupted proteolipid protein traYcking results in oligodendrocyte apoptosis in an animal model of Pelizaeus-Merzbacher disease. J Cell Biol. 140(4), 925–934. Granneman, J., SkoV, R., and Yang, X. (1998). Member of the peroxisome proliferator-activated receptor family of transcription factors is diVerentially expressed by oliogdendrocytes. J. Neurosci Res. 51, 563–573. Greer, J. M., Dyer, C. A., Pakaski, M., Symonowicz, C., Lees, M. B. (1996). Orientation of myelin proteolipid protein in the oligodendrocyte cell membrane. Neurochem Res. 21(4), 431–440. GriYths, I. R., Dickinson, P., and Montague, P. (1995). Expression of the proteolipid protein gene in glial cells of the post-natal peripheral nervous system of rodents. Neuropathol Appl Neurobiol. 21(2), 97–110. GriYths, I. R., Klugmann, M., Anderson, T. J., Yool, D., Thomson, C. E., Schwab, M. H., Schneider, A., Zimmermann, F., McCulloch, M. C., Nadon, N. L., and Nave, K.-A. (1998). Axonal swellings and degeneration in mice lacking the major proteolipid of myelin. Science 280, 1610–1613. GriYths, I. R., Mitchell, L. S., McPhilemy, K., Morrison, S., Kyriakides, E., and Barrie, J. A. (1989). Expression of myelin protein genes in Schwann cells. J Neurocytol 18(3), 345–352. Gupta, S. K., Pringle, J., Poduslo, J. F., and Mezei, C. (1991). Levels of proteolipid protein mRNAs in peripheral nerve are not under stringent axonal control. J Neurochem. 56(5), 1754–1762.

417

PROSPECTUS

Helynck, G., Bang, L., Nussbaum, J.-L., Picken, D., Skalidis, G., TriWlieV, E., Van Dorsselaer, A., Seta, P., Sandeaux, R., Gavach, E., Heitz, F., Simon, D., and Spach, G, (1983). Brain proteolipids: Isolation, puriWcation and eVect on ionic permeability of membranes. Eur. J. Biochem 133, 689–695. Hudson, L. D., Friedrich, V., Behar, T., Dubois-Dalcq, M., and Lazzarini, R. A. (1989). The initial events in myelin synthesis: Orientation of proteolipid protein in the plasma membrane of cultured oligodendrocytes. J. Cell Biol. 109, 717–727. Ikenaka, K., Furuichi, T., Iwasaki, Y., Moriguchi, A., Okano, H., Mikoshiba, K. (1988). Myelin proteolipid protein gene structure and its regulaton of expression in normal and jimpy mutant mice. J. Mol Biol. 199, 587–596. Ikenaka, K., Kagawa, T., and Mikoshiba, K. (1992). Selective expression of DM-20, an alternatively spliced myelin proteolipid protein gene product, in developing nervous system and in nonglial cells. J Neurochem. 58(6), 2248–2253. Inouye, H., and Kirschner, D. A. (1989). Orientation of proteolipid protein in myelin: Comparison of models with X-ray diVraction measurements. Dev. Neurosci. 11, 81–89. Janz, R., and StoVel, W. (1993). Characterization of a brain-speciWc Sp1-like activity interacting with an unusual binding site within the myelin proteolipid protein promoter. Biol. Chem. Hoppe Seyler 374, 507–517. Jiang, H., Duchala, C. S., Awatramani, R., Shumas, S., Carlock, L., Kamholz, J., Garbern, J., Scherer, S. S., Shy, M. E., and Macklin, W. B. (2000). Proteolipid protein mRNA stability is regulated by axonal contact in the rodent peripheral nervous system. J. Neurobiol. 44, 7–19. Jung, M., Sommer, I., Schachner, M., Nave, K.-A. (1996). Monoclonal antibody O10 deWnes a conformationally sensitive cell-surface epitope of proteolipid protein (PLP): Evidence that PLP misfolding underlies dysmyelination in transgenic mice. J. Neurosci. 16, 7920–7929. Kahan, I., and Moscarello, M. A. (1985). IdentiWcation of membrane-embedded domains of lipophilin from human myelin. Biochem. 24, 538–454. Kamholz, J., Sessa, M., Scherer, S., Vogelbacker, H., Mokuno, K., Baron, P., Wrabetz, L., Shy, M., and Pleasure, D. (1992). Structure and expression of proteolipid protein in the peripheral nervous system. J Neurosci Res. 31(2), 231–244. Kim, J., and Hudson, L. D. (1992). A novel member of the zinc Wnger superfamily: A C2-HC Finger that recognizes a glial-speciWc gene. Molecular Cell Biol. 12, 5632–5639. Kim, J. G., Armstrong, R. C., Berndt, J. A., Kim, N. W., and Hudson, L. D. (1998). A secreted DNA-binding protein that is translated through an internal ribosome entry site (IRES). Mol. Cell. Neurosci. 12, 119–140. Kim, J. G., Armstrong, R. C., Robinsky, A., Agoston, D. v., Wiese, C., Nagle, J., and Hudson, L. D. (1997). Cloning, mapping and expression of a neuronal zinc Wnger gene closely related to Myelin Transcription Factor 1 (Myt1) of the oligodendrocyte lineage. J. Neurosci. Res. 50, 272–290. Kitagawa, K., Sinoway, M. P., Yang, C., Gould, R. M., and Colman, D. R. (1993). A proteolipid protein gene family: Expression in sharks and rays and possible evolution from anancestral gene encoding a pore-forming polypeptide. Neuron. 11(3), 433–448. Klugmann, M., Schwab, M. H., Puhlhofer, A., Schneider, A, Zimmermann, F., GriYths, I. R., and Nave, K. A. (1997). Assembly of CNS myelin in the absence of proteolipid protein. Neuron. 18(1), 59–70. Knapp, P. E., Bartlett, W. P., Williams, L. A., Yamada, M., Ikenaka, K., and SkoV, R. P. (1999). Programmed cell death without DNA fragmentation in the jimpy mouse: Secreted factors can enhance survival. Cell Death DiVer 6, 136–145. Konola, J. T., Yamamura, T., Tyler, B., and Lees, M. B. (1992). Orientation of the myelin proteolipid protein C-terminus in oligodendroglial membranes. Glia 5, 112–121. Kurihara, T., Sakuma, M., and Gojobori, T. (1997). Molecular evolution of myelin proteolipid protein. Biochem Biophys Res Commun. 237(3),559–61. Lagenaur, C., Kunemund, V., Fischer, G., Fushiki, S., and Schachner, M. (1992). Monoclonal M6 antibody interferes with neurite extension of cultured neurons. J Neurobiol. 23(1), 71–188. Lees, M., and BrostoV, S. L. (1984). Proteins of myelin. In ‘‘Myelin’’ (P. Morell, ed.), 2nd ed., pp. 197–-224. Plenum Press, New York. Lees, M. B., Sakura, J. D., Sapirstein VS, Curatolo W. (1979). Structure and function of proteolipids in myelin and non-myelin membranes. Biochim Biophys Acta. 559(2–3), 209–230. Lepage, P., Helynck, G., Chu, J. Y., Luu, B., Sorokine, O., TriWlieV, E., and Van Dorsselaer, A. (1986). PuriWcation and characterization of minor brain proteolipids: Use of fast atom bombardmentmassspectrometry for peptide sequencing. Biochimie. 68(5), 669–686. Li, A. H. Y., Silverstein, B., Leone, M. G., Giacomelli, S., and Cheng, C. Y. (1995). Cerebrin-50, a human cerebrospinal Xuid protein whose mRNA is present in multiple tissues but predominantly expressed in lymphoblastoid cells and the brain. Biochem. Mol. Biol. 35,135–144. Li, S., Dobretsova, A., Kokorina, N. A., and Wight, P. A. (2002). Repression of myelin proteolipid protein gene expression is mediated through both general and cell type-speciWc negative regulatory elements in nonexpressing cells. J Neurochem. 82(1),159–171. Macklin, W. B., Campagnoni, C. W., Deininger,.PL., and Gardinier, M. V. (1987). Structure and expression of the mouse myelin proteolipid protein gene. J Neurosci Res. 18(3), 383–394. Mallon, B. S., Shick, H. E., Kidd, G. J., and Macklin, W. B. (2002). Proteolipid promoter activity distinguishes two populations of NG2-positive cells throughout neonatal cortical development. J Neurosci. 22(3), 876–885. McLaughlin, M., Hunter, D. J., Thomson, C. E., Yool, D., Kirkham, D., Freer, A. A., and GriYths, I. R. (2002). Evidence for possible interactions between PLP and DM20 within the myelin sheath. Glia. 39(1), 31–36.

au14

418

16. PROTEOLIPID PROTEIN GENE

Mi, Z. P., Weng, W., Hankin, M. H., Narayanan, V., and Lagenaur, C. F. (1998). Maturational changes in cell surface antigen expression in the mouse retina and optic pathway. Brain Res Dev Brain Res. 106(1–2), 145–154. Milner, R. J., Lai, C., Nave, K-A., Lenoir, D., Ogata, J., and SutcliVe, J. G. (1985). Nucleotide sequences of two mRNAs for rat brain myelin proteolipid protein. Cell 42, 931–939. Nadon, N. L. (2000). Aberrant dendritic growth in central nervous system white matter induced by a mutant proteolipid protein transgene. J. Comp. Neurol. 416, 262–268. Nadon, N. L., Arnheiter, H., and Hudson, L. D. (1994). A combination of PLP and DM20 transgenes promotes partial myelination in the jimpy mouse. J Neurochem. 63(3), 822–833. Nadon, N. L., and West, M. (1998). Myelin proteolipid protein: Function in myelin structure is distinct from its role in oligodendrocyte development. Dev Neurosci. 20(6), 533–5339. Nakao, J., Yamada M., Kagawa T., Kim, S. U., Miyao, Y., Shimizu, K., Mikoshiba, K., and Ikenaka, K. (1995). Expression of proteolipid protein gene is directly associated with secretion of a factor inXuencing oligodendrocyte development. J. Neurochem. 64, 2396–2403. Nave, K. A., and Lemke, G. (1991). Induction of the myelin proteolipid protein (PLP) gene in C6 glioblastoma cells: functional analysis of the PLP promoter. J Neurosci. 11(10), 3060–3069. Olinsky, S., Loop, B. T., DeKosky, A., Ripepi, B., Weng, W., Cummins, J., Wenger, S. L., Yan,Y., Lagenaur, C., and Narayanan, V. (1996). Chromosomal mapping of the human M6 genes. Genomics. 33(3), 532–536. Peirano, R. I., and Wegner, M. (2000). The glial transcription factor Sox10 binds to DNA both as monomer and dimer with diVerent functional consequences. Nucleic Acids Res. 28(16), 3047–3055. Peters, J. M, Lee, S. S. T, Li, W., Ward, J. M., Gavrilova, O., Everett, C., Reitman, M. L., Hudson, L. D., and Gonzalez, F. J. (2000). Growth, adipose, brain, and skin alterations resulting from targeted disruption of the mouse peroxisome proliferator-activated receptor b (d). Mol. Cell. Biol. 20, 5119–5128, 2000. Pham-Dinh, D., Birling, M. C., Roussel, G., Dautigny, A., Nussbaum, J. L. (1991). Proteolipid DM-20 predominates over PLP in peripheral nervous system. Neuroreport. 2(2), 89–92. Popot, J. L., Phan Dinh, D., and Dautigny, A., (1991). Major myelin proteolipid: The 4-alpha-helix topology. J. Membr. Biol. 120, 233–246. Pribyl, T. M., Campagnoni, C. W., Kampf, K., Kashima,T., Handley, V. W., McMahon, J., and Campagnoni, A. T. (1996). Expression of the myelin proteolipid protein gene in the human fetal thymus. J Neuroimmunol. 67(2), 125–130. Puckett, C., Hudson, L., Ono, K., Friedrich, V., Benecke, J., Dubois-Dalcq, M., and Lazzarini, R. A. (1987). Myelin-speciWc proteolipid protein is expressed in myelinating Schwann cells but is not incorporated into myelin sheaths. J Neurosci Res. 18(4), 511–518. Rosenbluth, J., StoVel, W., and SchiV, R. (1996). Myelin structure in proteolipid protein (PLP)-null mouse spinal cord. J Comp Neurol. 371(2), 336–344. Ross, N. W., and Braun, P. E. (1988). Acylation in vitro of the myelin proteolipid protein and comparison with acylation in vivo: Acylation of a cysteine occurs nonenzymatically. Neurosci Res. 21(1),35–44. Scherer, S. S., Vogelbacker, H. H., and Kamholz, J. (1992). Axons modulate the expression of proteolipid protein in the CNS. J Neurosci Res. 32(2), 138–148. Schindler, P., Luu, B. Sorokine, O., TriWlieV, E., and Van Dorsselaer, A. (1990). Developmental study of proteolipids in bovine brain: A novel proteolipid and DM20 appear before proteolipid protein (PLP) during myelination. J. Neurochem. 55, 2079–2086. Schliess, F., and StoVel, W. (1991). Evolution of the myelin integral membrane proteins of the central nervous system. Biol Chem Hoppe Seyler. 372(9), 865–874. Schneider, A., Montague, P., GriYths, I., Fanarraga, M., Kennedy, P., Brophy, P., and Nave, K. A. (1992). Uncoupling of hypomyelination and glial cell death by a mutation in the proteolipid protein gene. Nature. 358(6389), 758–761. Sheetz, M. P., Baumrind, N. L., Wayne, D. B., and Pearlman, A. L. (1990). Concentration of membrane antigens by forward transport and trapping in neuronal growth cones. Cell. 61(2), 231–241. Shi, Y. J., Lee, J. S., and Galvin, K. M. (1997). Everything you have ever wanted to know about Yin Yang 1. Biochim. Biophys. Acta 1332, F49-F66. Shimokawa, N., and Miura, M. (2000). Rhombex-29, a novel gene of the PLP/DM20-M6 family cloned from rat medulla oblongata by diVerential display. J Neurosci Res. 62(1), 1–8. Simons, M., Kramer, E. M., Macchi, P., Rathke-Hartlieb, S., Trotter, J., Nave, K.-A., and Schulz, J. B. (2002). Overexpression of the myelin proteolipid protein leads to accumulation of cholesterol and proteolipid protein in endosomes/lysosomes: Implications for Pelizaeus-Merzbacher disease. J Cell Biol. 157, 327–336. Simons, M., Kramer, E. M., Thiele, C., StoVel, W., and Trotter, J. (2000). Assembly of myelin by association of proteolipid protein with cholesterol-andgalactosylceramide-rich membrane domains. J Cell Biol. 151(1), 143–154. Sinoway, M. P., Kitagawa, K., Timsit, S., Hashim, G. A., Colman, D. R. (1994). Proteolipid protein interactions in transfectants: Implications for myelin assembly. J Neurosci Res. 37(5), 551–562. Sobel, R. A., Greer, J. M., Isaac, J., Fondren, G., and Lees, M. B. (1994). Immunolocalization of proteolipid protein peptide 103–116 in myelin. J. Neurosci. Res. 37, 36–43. Southwood, C., and Gow, A. (2001). Molecular pathways of oligodendrocyte apoptosis revealed by mutations in the proteolipid proteingene. Microsc Res Tech. 52(6), 700–708. Spassky, N., Goujet-Zalc, C., Parmantier, E., Olivier, C., Martinez, S., Ivanova, A., Ikenaka, K., Macklin, W., Cerruti, Zalc, B., and Thomas, J. L. (1998). Multiple restricted origin of oligodendrocytes. J Neurosci. 18(20), 8331–8843.

PROSPECTUS

Sporkel, O., Uschkureit, T., Bussow, H., and StoVel, W. (2002). Oligodendrocytes expressing exclusively the DM20 isoform of the proteolipid protein gene: Myelination and development. Glia 37(1), 19–30. Stecca, B., Southwood, C. M., Gragerov, A., Kelley, K. A., Friedrich, V. L. Jr., and Gow, A. (2000). The evolution of lipophilin genes from invertebrates to tetrapods: DM-20 cannot replace proteolipid protein in CNS myelin. J Neurosci. 20(11), 4002–4010. StoVel, W., Hillen, H., and Giersiefen, H. (1984). Structure and molecular arrangement of proteolipid protein of central nervous system myelin. Proc Natl Acad Sci USA 81(16),5012–6. StoVel, W., Subkowski, T., and Jander, S. (1989). Topology of proteolipid protein in the myelin membrane of central nervous system. A study usingantipeptide antibodies. Biol Chem Hoppe Seyler. 370(2), 165–176. Stolt, C. C., Rehberg, S., Ader, M., Lommes, P., Riethmacher, D., Schachner, M., Bartsch, U., and Wegner, M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev. 16(2), 165–170. Thomas, J. L., Spassky, N., Perez Villegas, E. M., Olivier, C., Cobos, I., Goujet-Zalc, C., Martinez, S, and Zalc, B. (2000). Spatiotemporal development of oligodendrocytes in the embryonic brain. J Neurosci Res. 59(4), 47147–6. Thomson, C. E, Montague, P., Jung, M., Nave, K. A., and GriYths, I. R. (1997). Phenotypic severity of murine Plp mutants reXects in vivo and in vitro variations in transport of PLP isoproteins. Glia 20(4), 322–332. Timsit, S. G., Bally-Cuif, L., Colman, D. R., Zalc, B. (1992). DM-20 mRNA is expressed during the embryonic development of the nervous system of the mouse. J Neurochem. 58(3), 1172–1175. Timsit, S,. Martinez, S., Allinquant, B., Peyron, F., Puelles, L., and Zalc, B. (1995). Oligodendrocytes originate in a restricted zone of the embryonic ventral neural tube deWned byDM-20 mRNA expression. J Neurosci. 15(2), 1012–1024. Ting-Beall, H. P., Lees, M. B., Robertson, J. D. (1979). Interactions of Folch-Lees proteolipid apoprotein with planar lipid bilayers. J. Membr. Biol. 51, 33–46. Uschkureit, T., Sporkel, O., Bussow, H., and StoVel, W. (2001). Rumpshaker-Like Proteolipid Protein (PLP) ratio in a mouse model with unperturbed structural and functional integrity of the myelin sheath and axons in the central nervous system. Glia 35, 63–71. Venkatesh, B., Erdmann, M. V., and Brenner, S. (2001). Molecular synapomorphies resolve evolutionary relationships of extant jawed vertebrates. Proc Natl Acad Sci U S A. 98(20), 11382–11387. Vouyiouklis, D. A., Barrie, J. A., GriYths, I. R, and Thomson, C. E. (2000). A proteolipid protein-speciWc premRNA (Ppm-1) contains intron 3 and is up-regulated duringmyelination in the CNS. J Neurochem 74(3), 940–948. Vouyiouklis, D. A., Werner, H, GriYths, I. R., Stewart, G. J., Armin-Nave, K., and Thomson, C. E. (1998). Molecular cloning and transfection studies of M6b-2, a novel splice variant of a member of thePLP-DM20/M6 gene family. J Neurosci Res. 52(6), 633–640. Waehneldt, T. V., Malotka, J., Karin, N. J., and Matthieu, J. M. (1985). Phylogenetic examination of vertebrate central nervous system myelin proteins by electro-immunoblotting. Neurosci Lett. 57(1), 97–102. Wahle, S., and StoVel, W. (1998). Cotranslational integration of myelin proteolipid protein (PLP) into the membrane of endoplasmic reticulum: Analysis of topology by glycosylation scanning and protease domain protection assay. Glia. 24(2), 226–235. Weimbs, T., and StoVel, W. (1992). Proteolipid protein (PLP) of CNS myelin: Positions of free, disulWde-bonded, and fatty acidthioester-linked cysteine residues and implications for the membrane topology of PLP. Biochemistry. 31(49), 12289–12296. Weimbs, T., and StoVel, W. (1994). Topology of CNS myelin proteolipid protein: Evidence for the nonenzymatic glycosylation of extracytoplasmic domains in normal and diabetic animals. Biochemistry. 33(34), 10408–10415. Werner, H., Dimou, L., Klugmann, M., PfeiVer, S., and Nave, K. A. (2001). Multiple splice isoforms of proteolipid M6B in neurons and oligodendrocytes. Mol Cell Neurosci. 18(6), 593–605. Wight, P. A., and Dobretsova, A. (1997). The Wrst intron of the myelin proteolipid protein gene confers cell typespeciWc expression by a transcriptional repression mechanism in non-expressing cell types. Gene. 201(1–2), 111–117. Wight, P. A., Dobretsova, A., and Macklin, W. B. (1997). Regulation of murine myelin proteolipid protein gene expression. J Neurosci Res. 50(6), 917–927. Wight, P. A., Duchala, C. S., Readhead, C., and Macklin, W. B. (1993). A myelin proteolipid protein-LacZ fusion protein is developmentally regulated and targeted to the membrane in transgenic mice. J Cell Biol. 123(2), 443–454. Witt, A., and Brady ST. (2000). Unwrapping new layers of complexity in axon/glial relationships. Glia. 29(2), 112–7. Yamada, M., Ivanova, A., Yamaguchi, Y., Lees, M. B., and Ikenaka, K. (1999). Proteolipid protein gene product can be secreted and exhibit biological activity during earlydevelopment. J Neurosci. 19(6), 2143–2151. Yamaguchi, Y., Ikenaka, K., Niinobe, M., Yamada, H., Mikoshiba, K. (1996). Myelin proteolipid protein (PLP), but not DM-20, is an inositol hexakisphospate-finding protein. J Biol Chem. 271(44), 27838–46. Yan, Y., Lagenaur, C., and Narayanan, V. (1993). Molecular cloning of M6: IdentiWcation of a PLP/DM20 gene family. Neuron. 11(3), 423–431. Yan, Y., Naryanan, V., and Lagenaur, C. (1996). Expression of members of the proteolipid protein gene family in the developing murine central nervous system. J. Comp. Neurol. 370, 465–478. Yool, D., Klugmann, M., Barrie, J. A., McCulloch, M. C., Nave, K. A., and GriYths, I. R. (2002). Observations on the structure of myelin lacking the major proteolipid protein. Neuropathol Appl Neurobiol. 28(1), 75–78.

419

420

16. PROTEOLIPID PROTEIN GENE

Yool, D. A., Klugmann, M., McLaughlin, M., Vouyiouklis, D. A., Dimou, L., Barrie, J. A., McCulloch, M. C., Nave, K. A., and GriYths, I. R. (2001). Myelin proteolipid proteins promote the interaction of oligodendrocytes and axons. J Neurosci Res. 63(2), 151–164. Yoshida, M., and Colman, D. R. (1996). Parallel evolution and coexpression of the proteolipid proteins and protein zero in vertebrate myelin. Neuron. 16(6), 1115–1126. Yoshida, M., Shan, W. S., and Colman, D. R. (1999). Conserved and divergent expression patterns of the proteolipid protein gene family in the amphibiancentral nervous system. J Neurosci Res. 57(1), 13–22. Yu, W. P., Collarini, E. J., Pringle, N. P., and Richardson, W. D. (1994). Embryonic expression of myelin genes: Evidence for a focal source of oligodendrocyte precursors in the ventricular zone of the neural tube. Neuron. 12(6), 1353–1362.

C H A P T E R

17 Myelin-Associated Glycoprotein Gene John Georgiou, Michael B. Tropak, and John C. Roder

MAG INTRODUCTION Specialized glial cells, oligodendrocytes in the central nervous system (CNS), and Schwann cells in the peripheral nervous system (PNS) elaborate cytoplasmic wrappings known as myelin around axons. The myelination process requires a complex series of interactions between the glial cells and axons, which remain poorly understood. Myelin functions to insulate neurons and facilitates the rapid signal conduction required in organisms with complex nervous systems. Myelin-associated glycoprotein (MAG) is a relatively minor constituent of both CNS and PNS myelin that has been implicated in the formation and maintenance of myelin. However, it is also a cell recognition molecule involved in neuronglial interactions, including regulation of axonal outgrowth and nerve regeneration.

Discovery of Myelin-Associated Glycoprotein, MAG Prior to the discovery of MAG in 1973, the major proteins in compact myelin such as myelin basic protein (MBP) and proteolipid protein (PLP) were known. During the early 1970s it became clear that proteins on the surface which mediate adhesion are generally glycosylated. Consequently, Quarles and colleagues used radiolabeled fucose to identify MAG (Quarles et al., 1973), a myelin glycoprotein that might mediate adhesive interactions between glial and neuronal cells that are important for the formation of the myelin sheath. MAG was cloned in 1987 (Arquint et al., 1987), and DNA sequence analysis revealed that the MAG cDNA that was isolated was derived from the same mRNA as clone p1B236, a randomly selected, brain-speciWc, partial cDNA isolated previously in 1983 (SutcliVe et al., 1983). Originally, 1B236 was thought to be a neuronal protein; however, subsequent studies conWrmed that MAG and 1B236 are one and the same (Arquint et al., 1987; Lai et al., 1987a; Noronha et al., 1989). Subsequent Wndings, described here, have yielded insight into the structure and function of MAG, the major CNS myelin glycoprotein.

Nomenclature While antibodies (Ab) consist of domains that belong to the immunoglobulin (Ig) superfamily (IgSF) and are involved in protein-protein interactions, there are several Ig-like proteins that recognize carbohydrates and that are also known as I-type lectins (see the review by Kelm, 2001). Cell surface receptors that bind sialic-acid were distinguished as a subfamily of I-lectins soon after the discovery of sialoadhesin, a macrophage adhesion molecule that recognizes sialylated glycans (Crocker et al., 1994). Sialoadhesin was found to share sequence similarity with the CD22 B-lymphocyte surface receptor and also with

Myelin Biology and Disorders, Volume 1

421

Copyright 2004, Elsevier (USA). All rights reserved.

422

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

MAG. All of these IgSF members bind sialic acid (Freeman et al., 1995; Kelm et al., 1994), and it was subsequently suggested they be referred to as the ‘‘siglec’’ family, derived from sialic acid-binding/immunoglobulin-like/lectin (Crocker et al., 1998). There are now 11 members, which are grouped according to their structural and functional similarities (Fig. 17.1). MAG, which is designated siglec-4a, and the related siglec-4b, which is known as Schwann cell myelin protein (SMP) and found in birds, are two of only three siglecs expressed by cells outside of the haematopoietic system. A third siglec with marked sequence similarity to siglec-10, named siglec-11, was recently cloned and found to be expressed on brain microglia (Angata et al., 2002). However, there is no mouse ortholog of siglec-11, and hence it is believed to have evolved after the split of primate and rodent lineages. Most of the scientiWc community continues to use the term MAG instead of siglec-4a, and hence for convenience we do the same here.

MAG Localization MAG expression is almost exclusively associated with glial cells that will form myelin. Immunolocalization studies at the light and electron microscope levels demonstrate that MAG in the CNS and PNS is found on the Schwann cell and oligodendrocyte membrane in the periaxonal space (Bartsch et al., 1989; Martini and Schachner, 1986; Sternberger et al., 1979; Trapp and Quarles, 1982). The periaxonal space is the 12 to 14 nm interface between the axon and innermost myelin layer. The electron microscopy (EM) images shown in Figure 17.2A reveal that in young post-natal day (P) 10 oligodendrocytes that have just begun to ensheath axons, anti-MAG Abs label the cell surface (cell body and processes). In developing nerves undergoing myelination, MAG is present in the loosely wrapped myelin layers before the compaction process has occurred. Thus, MAG may function to mediate glial-neuron as well as glial-glial interactions. In axons that have been fully myelinated, strong MAG immunoreactivity is localized to the periaxonal region and at the inner mesaxon, as seen in Figures 17.2B and 17.2C. However, no MAG is present within the layers of compact myelin. The light microscopy images in Figure 17.3 show the distribution of MAG in cross sections and longitudinal views of peripheral nerve (Figs. 17.3A and 17.3C, respectively); corresponding schematics appear in Figures 17.3B and 17.3D to help distinguish the distribution pattern of MAG and also to identify various features of myelinated nerve. In the PNS, but not in the CNS, MAG is also found in the paranodal loops, Schmidt-Lantermann incisures (channels within the myelin sheath), and there is some low expression at the external (abaxonal) surface of myelinating Schwann cells (Owens and Bunge, 1989; Trapp and Quarles, 1982; Trapp et al., 1989b). The functional signiWcance of these diVerences is not known, although they may reXect diVerent mechanisms used to target MAG to the myelin membrane (Trapp et al., 1989a). Expression of MAG or even contact with a neuron is not suYcient to induce glial cells to form myelin. Cultured oligodendrocytes are able to elaborate myelin-like membrane whorls in the absence of neurons (Bradel and Prince, 1983; Rome et al., 1986). Schwann cells require neuronal contact in order to initiate expression of the major myelin proteins such as MAG and MBP; however, these proteins can be readily detected in oligodendrocytes cultured in the absence of neurons (Dubois-Dalcq et al., 1986; Owens and Bunge, 1989). Moreover, MAG has been detected on Schwann cells prior to the onset of myelination (Owens and Bunge, 1989; Trapp, 1988), suggesting an additional role before myelination. Evidence is accumulating that transcription factors are involved in controlling whether a glial cell will form myelin, a process that is speculated to be regulated by molecular signals originating from nerves (see reviews by Wegner, 2000; Jessen and Mirsky, 2002). However, the nervederived signals and glial signaling mechanisms involved in myelination remain obscure. Consistent with the notion that MAG also has additional roles besides myelin formation and maintenance is the fact that MAG is expressed in perisynaptic Schwann cells, glia that cover motor nerve terminals at neuromuscular synapses and do not form myelin wrappings (Georgiou and Charlton, 1999). The perisynaptic Schwann cells also express other myelinrelated molecules, including the glycoprotein known as protein zero (P0) over most of their membrane. These glia express MAG at their adaxonal membrane and MAG expression persists

MAG INTRODUCTION

423

FIGURE 17.1 Structure and comparison of siglecs. Schematic showing Ig-like lectins that bind to sialic acid (siglecs). MAG belongs to the siglec-4a group, together with the related Schwann cell myelin protein (SMP, siglec-4b) found in birds. Not shown is the recently cloned human siglec-11, which has no rodent homolog. (A) Siglecs are type-I membrane proteins with an extracellular region containing a homologous V-set Ig-like domain and a varying number of C2-set Ig-like domains at the N-terminus. The cytoplasmic tails of all siglecs apart from sialoadhesin contain tyrosine residues (Y) within potential signaling motifs. Those motifs that Wt the consensus sequence for an immunoreceptor tyrosine-based inhibition motif (ITIM) are shown in pink. The membrane-distal tyrosine-based motifs that are highly conserved in CD33-related siglecs are shown in green. Several siglecs undergo alternative splicing, but only the known full-length forms are illustrated. Potential N-linked glycans are indicated in ball-and-stick form. (B) Alignment of the C-terminal portions of the cytoplasmic tails of CD33-related siglecs reveals two conserved tyrosine-containing motifs. The sequences for siglec-4a and -4b, shown at the top, reveal that the distal motif is similar to the CD33-related siglecs, however, the proximal motif is not conserved. Residues that are identical are boxed in black and residues that are conserved are boxed in gray. The membrane-proximal motif conforms to the consensus ITIM sequence, whereas the distal motif does not. (C) Positions of key residues in sialoadhesin that bind the N-acetylneuraminic acid (Neu5Ac) portion of 3’ sialyllactose, as revealed in a ligand-bound crystal structure of the sialoadhesin N-terminal domain. An essential arginine (Arg97) on the F strand (conserved in the other siglecs) forms a salt bridge with the carboxylate of sialic acid (Neu5Ac) and two tryptophans (Trp2 and Trp106) on the A and G strands form hydrophobic contacts with the N-acetyl and glycerol side groups of Neu5Ac, respectively. (D) Schematic diagram of the V- and C-type domains from immunoglobulins, showing the topology of the b-strands in the two b-sheets. au10 au11 Ig domains belonging to the C2 set have a topology similar to the C-type domain. In domain 1 of sialoadhesin, the inter-sheet disulphide bridge connecting strands B and F is replaced by an intra-sheet dishulphide bridge connecting strands B and E; the C’’ strand is replaced by a coiled structure au12 and G-strand is split in two. Parts A through C have been modiWed with permission from Crocker and Varki, 2001, TRENDS in Immunology 22, 337-342, 2001, Elsevier Science Ltd.

424

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

FIGURE 17.2 Electron microscopy reveals MAG’s periaxonal localization. Detection of MAG by immuno-EM from mouse optic nerve at P10 (A-B) and P14 (C). (A–B) MAG immunoreactivity on the cell surface (A) and processes (B) of an oligodendrocyte prior or during the time of axon (Ax) ensheathment. Preembedding staining procedures were used in association with anti-MAG Abs and visualized by peroxidase-coupled protein A. (C) When compact myelin surrounds the axon, MAG is conWned to the periaxonal region and noncompacted myelin (arrows). Postembedding staining technique is necessary to demonstrate the periaxonal localization of MAG due to presence of compact myelin (M). MAG was detected with secondary Abs absorbed to colloidal gold (gold particles appear as black dots). Scale bars: A ¼ 1 mm; B, C ¼ 0.2 mm. Reprinted with permission from Bartsch, KirchhoV, and Schachner, 1989, Journal of Comparative Neurology 284, 451–462, 1989, Wiley-Liss, Inc.

after denervation including at newly formed glial cell extensions. It is possible that factors present at synapses may prevent myelination, or alternatively, MAG may mediate adhesion between axons and surrounding glia. Regardless, it is clear that expression of myelin-related proteins, including MAG, does not obligate glial cells to form myelin wrappings. Various observations, including MAG adaxonal localization, have suggested that it mediates interactions between the axon and myelin sheath. Expression of MAG on oligodendrocyte tips prior to contact with axons (Bartsch et al., 1989), and the fact that MAG is the Wrst myelin protein exported to the tips, further supported a role for MAG in the initial stages of myelin formation. The myelination process begins when the glial cell process (mesaxon) begins to spiral around the axon, and initial observations in the PNS indicate MAG can be Wrst detected in the periaxonal space after 1.5 turns of the mesaxon around the axon (Martini and Schachner, 1986). This suggests that MAG is involved in some aspect of the wrapping process, and as discussed later in the section titled ‘‘Multiple Gene Knockouts That Include Deletion of MAG,’’ MAG’s involvement in the spiraling of myelinated Schwann cells has also been implicated from studies of double mutant mice lacking P0 and MAG. Finally, MAG has an important role in the control of axonal outgrowth, and this feature is discussed later in the section titled ‘‘Control of Axonal Growth and Regeneration by MAG.’’ Here we introduce that MAG exists primarily as two isoforms, and that the expression of each is regulated temporally and spatially. The isoforms result from alternative splicing to yield a relatively shorter protein known as small MAG (S-MAG) and a longer protein called large MAG (L-MAG). L-MAG predominates during CNS development, including the myelination process, whereas S-MAG accumulates in later stages (Inuzuka et al., 1991; Lai et al., 1987a; Pedraza et al., 1991; Tropak et al., 1988). In the peripheral nervous system, L-MAG is always a minor constituent. MAG isoforms have unique signaling capacities, for instance, L-MAG can activate Fyn kinase, and thus the diVerential expression of MAG isoforms likely has important functional consequences. More information on the unique roles of each MAG isoform will be revealed throughout this chapter.

au13 au14

MAG INTRODUCTION

FIGURE 17.3 MAG distribution in myelinated nerve sections. (A) Confocal image from a transverse section of rat sciatic nerve, double-labeled with a mouse monoclonal Ab against MAG (red, detected by TRITC Xuorescence) and also with rabbit antiserum against b4 integrin (green, detected by FITC Xuorescence). MAG is localized on the inner/ adaxonal membrane, and b4 integrin is localized around the entire circumference of the outer/abaxonal membrane. Compact myelin is not stained and thus appears black. (B) The circumferential organization of a myelinated axon is shown schematically. (C) Image on the left shows MAG immunoXuorescence from a longitudinal section of sciatic nerve. Image on the right shows MAG distribution in teased single Wbers. (D) Schematic showing myelinated axon longitudinal organization and MAG localization. Parts A and B were reprinted with permission from Scherer and Arroyo, 2002, Journal of the Peripheral Nervous System 7, 1–12, 2002 Peripheral Nerve Society, Inc. Part C was modiWed with permission from Altevogt, Kleopa, Postma, Scherer, and Paul, 2002, Journal of Neuroscience 22, 6458–6470, 2002, by the Society for Neuroscience.

MAG Gene Structure MAG Promoter and Gene Regulation Similar to myelin genes such as P0 and P2 basic protein (Peirano et al., 2000), the MAG promoter lacks a TATA box and instead consists of a GC-rich region (Ye et al., 1994). IdentiWcation of consensus sequence sites for SP1 and AP2 transcription factors commonly found at GC-rich promoters is consistent with foot-printing experiments using the MAG promoter transiently expressed in an immortalized glial cell line (Laszkiewicz et al., 1997). The regulatory region of MAG extends to either side of a 152 bp promoter core from
1.6 to þ0.6 kb. A promoter proximal segment that contains sites for strong transcriptional activators was identiWed; however, there is a region downstream containing inhibitory cis-elements (Grubinska et al., 1994). The fact that CpG islands in the MAG regulatory region become

425

426

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

FIGURE 17.4 Schematic of MAG gene locus. The MAG gene contains 13 exons spanning 16 kb and each of its Ig domain is encoded by a single exon (exons 5 to 9). Exons (rectangles), introns (line) and the exons that undergo alternative splicing are indicated (see legend). Exon 12 can be alternatively spliced in, or skipped, to produce S-MAG or L-MAG mRNA transcripts, respectively. There is also alternative splicing of exon 2, which occurs independently of exon 12 splicing (not all combinations are shown); however, it does not contribute to the coding region and its signiWcance is unknown. Alternative polyadenylation cleavage site is depicted as a vertical dashed line in exon 13.

more demethylated as oligodendrocytes progress along the diVerentiation pathway underscores the importance of methylation in regulation of MAG expression. Transcription from the MAG promoter in Schwann cells does not appear to be aVected by glial transcription factors Krox-20 (Topilko et al., 1994) or Sox-10 (Stolt et al., 2002). Although MAG expression is increased in oligodendrocytes in the presence of both transcription factors Oct and Myc, it is not clear whether this is a direct eVect (Jensen et al., 1999). MAG Exon Structure The MAG gene has been localized to human chromosome 19 (Barton et al., 1987), and mouse chromosome 7 (Barton et al., 1987; D’Eustachio et al., 1988), followed by highresolution mapping to 19q13.1 (Spagnol et al., 1989; Trask et al., 1993). Elucidation of the intron/exon structure of rat and mouse MAG genes revealed 13 exons spanning approximately 16 kb where each Ig domain is encoded by a single exon (Lai et al., 1987a; Nakano et al., 1991). Exon 12 can be alternatively spliced in, or skipped, to produce S-MAG or L-MAG, respectively. The schematic shown in Figure 17.4 illustrates the structure and products of the MAG gene. Exon 2 from the 5’ noncoding region can also be spliced (Fujita et al., 1989; Lai et al., 1987a; Tropak et al., 1988). Splicing of exon 2 and exon 12 occur independently. One form of MAG lacking exon 2 is predominant in the PNS, whereas mRNA containing exon 2 predominates in the CNS. In contrast to the known diVerences in signaling capabilities associated with L-MAG versus S-MAG, the signiWcance, if any, of exon 2 splicing is unknown. Alternative Splicing Produces Two Main MAG Isoforms Initially, the S- and L-MAG isoforms detected in the CNS were attributed to diVerential glycosylation (Quarles et al., 1973). Subsequently, the two isoforms were shown to be derived from diVerent mRNAs (Frail and Braun, 1984; Salzer et al., 1987; Tropak et al., 1988) by alternative splicing of exons from a primary RNA transcript (Lai et al., 1987a). The two isoforms of 67 and 72 kDa (sizes of proteins after deglycosylation) share a common region in their cytoplasmic domain but diVer in the length and sequence of the amino acids at the C-terminus. The cytoplasmic domain of MAG is encoded by exons 11, 12, and 13. When exon 12 is included, a premature in-frame stop codon results in the shorter S-MAG isoform with an intracellular sequence of 90 residues that contains a unique 10 amino acid C-terminal end. When exon 12 is excluded, the larger L-MAG is produced, which has an additional 54 amino acids derived from exon 13. DiVerential expression of the two MAG isoforms during development in the PNS and CNS of rat, mouse, and human was veriWed at the RNA (Frail and Braun, 1984; Frail et al.,

MAG BIOCHEMISTRY, STRUCTURE, AND ADHESIVE PROPERTIES

1985; Lai et al., 1987b; Miescher et al., 1997; Tropak et al., 1988) and protein levels (Inuzuka et al., 1991; Pedraza et al., 1991). Developmental regulation may be related to neuronal contact or the local environment because oligodendrocytes in culture express both MAG isoforms (Tropak et al., 1988). Interestingly, when L-MAG is constitutively expressed at high levels in Schwann cells, independent of axonal contact, a greater number of axons are segregated compared to control cells, which express MAG only upon axonal contact (Owens et al., 1990). This observation strongly supports the adhesive role of MAG prior to the start of myelination. However, the biological relevance of these observations is not clear, since only low levels of L-MAG can be detected in control Schwann cells and during PNS development (Pedraza et al., 1991; Tropak et al., 1988). Although myelin formation by oligodendrocytes and Schwann cells appears outwardly similar, the mechanistic details are distinct. However, oligodendrocytes diVer from Schwann cells in several respects. Unlike Schwann cells, which can only myelinate a single segment of an axon, oligodendrocytes are able to simultaneously myelinate multiple segments of diVerent axons. Interestingly, there appears to be a correlation between the relative levels of L-MAG and S-MAG expression and the four morphological types of oligodendrocytes (Butt et al., 1998). Thus, it is possible that L-MAG may function in the early events of myelination related to the ability of oligodendrocytes to myelinate multiple axonal segments. Alternative splicing of MAG is regulated by RNA binding proteins known as QKI. The regulation of MAG isoform expression by QKI was discovered from studies of a naturally occurring mouse mutant known as ‘‘quaking’’ (qk) that has an altered qkI gene regulatory region (noncoding). Reduced QKI levels in qk mice alters the developmental expression of the two MAG isoforms in the CNS and results in S-MAG as the major mRNA throughout development, while L-MAG is scarcely expressed (Frail and Braun, 1985; Fujita et al., 1990). The phenotype of qk mice is described later in the section titled ‘‘Naturally Occurring Mutations AVecting MAG.’’ QKI contains an RNA-binding domain and belongs to the signal transduction and activator of RNA (STAR) family (Ebersole et al., 1996; Vernet and Artzt, 1997). Of the many QKI isoforms that exist, the nuclear localized isoform QKI-5 has been shown to regulate alternative splicing of a MAG minigene as well as the myelin genes PLP and MBP (Wu et al., 2002). In the proposed model, binding of QKI-5 to the QASE consensus sequence downstream of the 5’ splice site may interfere with recognition of the splice site or the downstream intronic enhancer, thereby resulting in skipping of exon 12 and concomitant production of the L-MAG mRNA.

MAG BIOCHEMISTRY, STRUCTURE, AND ADHESIVE PROPERTIES Following translation MAG undergoes several modiWcations including phosphorylation, remodeling of oligosaccharides by sulfotransferases and sialyltransferases, and palmitylation. With the exception of palmitylation, each of these modiWcations will be discussed in greater detail. Palmitylation has been shown to be important in membrane targeting and activity of nonreceptor tyrosine kinases (reviewed in Resh, 1994). Similarly, palmitylation of haemagglutinin is important for the infectivity of inXuenza virus, although the mechanism is unclear (Fischer et al., 1998). Pedraza and colleagues experimentally veriWed the initial prediction that MAG is palmitylated by showing that Cys531 residue in the transmembrane domain is the site of palmitylation (Pedraza et al., 1990). Based on other systems, palmitylation of MAG may be important for some aspect of membrane targeting.

Post-Translational Regulation of MAG MAG Phosphorylation The primary structure of MAG revealed potential Ser, Thr, and Tyr phosphorylation sites within its cytoplasmic domain (Arquint et al., 1987; Salzer et al., 1987). In CNS myelin and oligodendrocytes, L-MAG is the predominant isoform that is phosphorylated, whereas

427

428

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

S-MAG is the major phosphorylated isoform in PNS myelin and Schwann cells (Afar et al., 1990; Agrawal et al., 1990; Bambrick and Braun, 1991; Edwards et al., 1988, 1989; Umemori et al., 1994; Yim et al., 1995). L-MAG is phosphorylated in vivo and in vitro primarily on Ser, but also on Thr and Tyr residues. S-MAG is phosphorylated constitutively only on Ser. In cultured oligodendrocytes, both S- and L-MAG isoforms are phosphorylated on Ser, while in transformed Schwann cells only S-MAG is present and phosphorylated (KirchhoV et al., 1993). Phorbol ester enhances phosphorylation two- to three-fold, suggesting PKC is involved. Recently it was demonstrated that L-MAG is a PKA substrate (Kursula et al., 2000). S-MAG and L-MAG contain potential Tyr phosphorylation sites, and tyrosine kinases can bind and phosphorylate MAG. Phosphorylation of MAG by v-fps and v-src proteintyrosine kinases has been demonstrated in vitro (Afar et al., 1990; Edwards et al., 1988). Fyn kinase, a member of the Src family tyrosine kinases, interacts with MAG and phosphorylates the Tyr at position 620 found in L-MAG (Jaramillo et al., 1994; Umemori et al., 1994). In transfected cells, Fyn only associates with L-MAG, not with S-MAG. In brain lysates however, Fyn was co-immunoprecipitated with both forms of MAG, and conversely, each of L- and S-MAG co-immunoprecipitated with Fyn. The lack of Fyn interaction with S-MAG in the expression system was suggested to be due to the lack of additional molecules that normally mediate this interaction in the brain. Another interpretation is that Fyn does not associate with S-MAG in the brain and that the interaction is seen because S-MAG dimerizes with L-MAG. MAG Glycosylation Approximately 30% of the mass of MAG is due to carbohydrate (Quarles, 1983), which is consistent with the eight N-linked oligosaccharide addition sites identiWed in the predicted amino acid sequence of rat MAG (Arquint et al., 1987; Lai et al., 1987b; Salzer et al., 1987). Mutagenesis experiments of the conserved Asn or Ser residues in the glycosylation consensus sequence have shown that each of the predicted glycosylation sites are utilized when expressed in CHO or COS-1 cell lines (Sgroi et al., 1996; Tropak and Roder, 1997). The majority of oligosaccharides on MAG are of the complex type; about two-thirds are tri- or tetra-antennary, one-third are biantennary, and few or none are the high-mannose type (Noronha et al., 1989). A high proportion of the oligosaccharides are sialylated and sulphated (Matthieu et al., 1975; Quarles et al., 1983). Although some experimental evidence suggested that O-linked oligosaccharides may be present in MAG (Pedraza et al., 1990), these results have been attributed to the presence of N-glycosidases in the batch of O-glycosidase used to test for the presence of O-linked glycans (Salzer, personal communication). The carbohydrate epitope recognized by the L2/human natural killer (HNK)-1 Ab (Noronha et al., 1986) is expressed on MAG (McGarry et al., 1983). It is also found on a number of IgSF member cell adhesion molecules (CAMs) in the nervous system such as the neural cell adhesion molecule (N-CAM), L1, P0 (Bollensen et al., 1988; Kruse et al., 1984), as well as unrelated adhesive molecules such as peripheral myelin-protein-22 (PMP22) and proteoglycans (Snipes et al., 1993). The conservation of the epitope throughout phylogeny (Bajt et al., 1990) is suggestive of the importance of HNK-1 in adhesion. Abs against the L2/HNK-1 epitope have been shown to perturb astrocyte-neuron cell adhesion, neurite outgrowth and attachment of cells to laminin (Hall et al., 1993; Kunemund et al., 1988; Riopelle et al., 1986). Oligosaccharides expressing the HNK-1 epitope have been used to block P0-mediated cell adhesion (GriYth et al., 1992) and cell-cell and cell-substrate interactions (Kunemund et al., 1988). Currently, the gp120 receptor on the AIDS virus is the only protein capable of interacting with the HNK-1 carbohydrate on MAG (van den Berg et al., 1992a). In the case of rat MAG, the HNK-1 epitope has been localized to oligosaccharides in either domain 4 or domain 5 (Pedraza et al., 1995), whereas all oligosaccharides on human MAG have been suggested to express the HNK-1 epitope (Burger et al., 1991). The HNK-1 epitope consists of a 3’ sulphated glucuronyl residue found on glycolipids (Chou et al., 1986), N-linked oligosaccharides (Voshol et al., 1996), and to a limited extent on O-linked oligosaccharides (Ong et al., 2002). Sulphation of the

MAG BIOCHEMISTRY, STRUCTURE, AND ADHESIVE PROPERTIES

glucuronic acid is critical for binding of the Ab. However, sulphation alone is not suYcient for binding of Abs recognizing the HNK-1 epitope, since most oligosaccharides on rat MAG are sulphated, yet the protein is poorly recognized by the Ab (O’Shannessy et al., 1985). The species-dependent expression (O’Shannessy et al., 1985) of the epitope, as well as the fact that not all MAG molecules express the epitope (Burger et al., 1992; Kruse et al., 1984), suggests that HNK-1 may not play a major role in MAG function. Alternatively, one component of the HNK-1 epitope, such as the sulphate group, may be important in MAG function.

MAG Structure On the basis of the amino acid sequence derived from the cDNA for rat MAG, the protein was predicted to be a type I membrane glycoprotein consisting of an N-terminal extracellular domain, a single transmembrane domain and a short cytoplasmic domain (Arquint et al., 1987; Lai et al., 1987b; Salzer et al., 1987). These predictions were later experimentally veriWed (Johnson et al., 1989; Pedraza et al., 1990). The extracellular portion of MAG was predicted to consist of Wve Ig-like domains, based on the presence of amino acid sequences in MAG, which are conserved among all members of the IgSF. Human and mouse MAG have 98% and 95% amino acid sequence identity, respectively, with rat MAG (Fujita et al., 1989; Sato et al., 1989; Spagnol et al., 1989). SMP, which overall shares 45% amino acid sequence identity with rat MAG, is very likely a quail ortholog of MAG (Dulac et al., 1992). IgSF The IgSF consists of a large number of closely related proteins containing one or more domains that are similar in sequence to domains found in Igs (Williams and Barclay, 1988). The proto-typical Ig domain, found in circulating Igs, consists of two anti-parallel b-sheets held together by an inter-sheet disulphide bridge (see Fig. 17.1D and Amzel and Poljak, 1979). Based on the sequences of Ig-like molecules available at that time, the members of the IgSF were divided into three sets based on the size of the linker between the conserved cysteines and the presence of conserved residues characteristic of the variable (V-) domain or constant (C-) domain from Igs (Williams and Barclay, 1988). The variable domain consists of disulphide-linked anti-parallel sheets with four beta strands (referred to as D, E, B, and A), in one face, and Wve strands (referred to as C, C’, C’’, F, and G) in the other (see Figure 17.1D). In contrast, the C-domain consists of two anti-parallel disulphide-linked sheets with four b-strands (referred to as D, E, B, and A) in one face and three strands (referred to as C, F, and G) in the other. Domains with greatest similarity to the V-domain or C-domain were placed in the V- or C1-sets, respectively. Domains which were similar in size to the shorter C-domain and contained conserved residues found in b-strands D and E from the V-domain were placed in the C2-set. Recently, a fourth set, the I-set, has been deWned on the basis of the structure of telokin and its close similarity to other members of the IgSF (Harpaz and Chothia, 1994). Members of the I-set have the same number of b-strands found in the C1- and C2-sets and conserved sequences characteristic of V- and C1-sets. Disulphide Linkage Domain 1 of MAG is most similar to Ig domains in the V-set (Williams and Barclay, 1988), whereas the other four domains are most similar to Ig domains in the C2-set (Arquint et al., 1987; Lai et al., 1987a, 1987b; Salzer et al., 1987). The V-like domain 1 of MAG is unique in that the Cys pair (Cys42 and Cys100) may form an intra-sheet disulphide bridge (Williams and Barclay, 1988) as opposed to the inter-sheet disulphide bridge encountered in most Igs. A similar intra-sheet disulphide bridge has been identiWed in domain 2 of the crystallographically determined structure of CD2, another member of the IgSF (Jones et al., 1992). Domains 1 and 2 of MAG each contain an additional Cys (Cys37 and Cys159) near the N-termini of the domains. Experimental evidence suggests that the additional Cys form an interdomain disulphide bridge between domains 1 and 2 (Pedraza et al., 1990).

429

430

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

Modeling studies suggest that Cys37 and Cys159 are suYciently close to form an interdomain disulphide bridge (Kursula, 2001). MAG Is a Member of the Sialic-Acid Binding Family, Siglec The unique arrangement of conserved Cys residues in domains 1 and 2 in all MAG orthologues is also a distinguishing feature shared by all IgSF members belonging to the siglec family of I-type lectins. The siglec family of proteins including sialoadhesin, CD22, CD33 family all recognize sialylated glycans. Experimental demonstration of MAG’s ability to recognize sialylated glycans on cells was Wrst shown by Kelm and colleagues (1994). In addition to the conserved Cys residues, all functional siglecs contain an invariant Arg residue (Arg118 in MAG, Arg97 in sialoadhesin). The 3D structure of domain 1 from sialoadhesin complexed with sialyllactose has enabled domain modeling studies of MAG and other siglecs (May et al., 1998). Although siglec-1 domain 1 does indeed belong to the V-set of Ig domains, it does, however, diVer in several respects. First, it possesses, as predicted, an intra-sheet disulphide bridge between strands B to E rather than B to F. Second, the spacing between the two sheets is increased from the usual 5.6–7.4A range to 8.6A. Third, the C’’ beta strand, which corresponds to a region of high diversity in the aligned siglec sequences, is replaced with a quasi alpha-helix like coiling strand. Lastly, the G-b strand is split in two and the other region of siglec diversity corresponds to extended BC loop of unknown function. The overall structure of a complete siglec has yet to be determined. However, based on EM pictures of a proteolytically generated soluble form of MAG (dMAG) (Sato et al., 1984), the molecule appears to be folded back on itself with domains 1 and 2 folded back on domains 5 and 4, respectively (Fig. 17.5, and Fahrig et al., 1993). The higher order structure of the domains 1 through 5 is also suggested by biophysical studies using a recombinant soluble derivative of rat MAG (Attia et al., 1993). Furthermore, this type of arrangement, where the IgSF domains are folded back in a hairpin/horseshoe conWguration have been shown in the 3D structures of hemolin and TAG-1/axonin derived by X-ray crystallography (Freigang et al., 2000; Su et al., 1998) and EM pictures of L1 (Schurmann et al., 2001). Although domain 1 alone from sialoadhesin retains the ability to bind sialylated glycans (Nath et al., 1995), the minimum MAG deletion mutant displaying lectin activity consists of domains 1, 2, and 3 (Kelm et al., 1994; Meyer-Franke et al., 1995). The sialic acid binding site on MAG has been indirectly mapped to domain 1 using a MAG mutant in which the invariant Arg in domain 1 found in all siglecs has been mutated to Ala/Asp (Tang et al., 1997a). Deletion mapping studies of the lectin activity and of a conformational epitope recognized by the 513 monoclonal Ab suggest that the three N-terminal domains of MAG are conformationally linked (Meyer-Franke et al., 1995). Similarly, domains 1 and 2 of CD22 appear to be conformationally linked (Nath et al., 1995). Thus, the N-terminal domains in some siglecs may not fold independently, but may require the presence of additional domains to attain the mature conformation. The fact that domains 1, 2, and 3 appear to be conformationally linked is consistent with EM and biophysical studies, which suggest that the domains in MAG adopt a higher-order structure with the domains folded back on each other (Attia et al., 1993; Fahrig et al., 1993). Two of the diVerences that contribute to binding of the 3’ sialyllactose lies to the GFCC’ face of the Ig domain. The conserved residues Trp2 and Trp106 are directed away from the GFCC b-sheet enabling them to hydrophobically interact with the sialyllactose and resulting in the observed increase in inter-sheet distance. Trp2 interacts with the 5’aminoacyl group, whereas Trp106 contacts the glycerol tail of sialic acid. In most Ig domains these two residues are directed between the b-sheets where they contribute to formation of the hydrophobic core. The split G-strand may enable backbone mediated H-bonds with the hydroxyl groups of the sialic acid glycerol tail. The invariant Arg97, which has been shown by mutagenesis to be critically important for sialic acid recognition, forms a salt bridge with the carboxylate of sialic acid. A similar interaction is seen in the 3D structure of an unrelated sialic acid binding lectin (VP1 from polyoma virus) and possibly other bacterial sialic acid binding lectins (May and Jones, 2001).

MAG BIOCHEMISTRY, STRUCTURE, AND ADHESIVE PROPERTIES

FIGURE 17.5 Transmission electron micrographs of dMAG molecules. Transmission EM visualization by rotary-shadowing technique of soluble dMAG (comprising most of the extracellular part of the molecule, 90 kDa molecular weight). (A) Large Weld of MAG molecules shadowed with platinum/carbon. (B–D) Selected images of MAG molecules shadowed with tantalum/tungsten. MAG molecules appear as rod-like structures with a globular domain at one end (shown oriented toward top). Depending on the view or transitional state with varied degree of twisting, the nonglobular part of the molecule appears as a compact structure (B), is divided into two thin parallel arms (C) or with an internal thickening (D). (B’–D’) Schematic representations of the images shown in (B-D). Scale bars: A ¼ 100 nm; B–D ¼ 25 nm (bar shown in D). Reprinted with permission from Fahrig, Probstmeier, Spiess, MeyerFranke, KirchhoV, Drescher, and Schachner, 1993, European Journal of Neuroscience 5, 1118–1126, 1993, European Neuroscience Association.

Structural and Biophysical Features of MAG

au1

All members of the siglec family are type I transmembrane proteins, which have two closely related Ig-like domains belonging to the V and C2-set (Kelm et al., 1994; Powell and Varki, 1995) but diVer in the number of additional Ig-like domains belonging to the C2-set. MAG and SMP consist of Wve Ig-like domains (Dulac et al., 1992). The two isoforms of human and mouse CD22a and b diVer in that CD22b contains seven extracel-

431

432

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

au2

lular Ig domains, whereas CD22a lacks Ig domains 3 and 4 (Engel et al., 1993; Stamenkovic and Seed, 1990; Stamenkovic et al., 1991; Wilson et al., 1991). Sialoadhesin has 17 Ig-like domains (Crocker et al., 1994). However, cDNAs have been isolated, which suggest the existence of soluble isoforms which are missing C-terminal domains and consist of either 3 or 16 Ig domains. Human and mouse CD33 consist solely of the V and C2-like domains, which deWne the I-type sialyl lectins (Simmons and Seed, 1988; Tchilian et al., 1994). Two isoforms of mouse CD33 have been isolated, which diVer in the size and amino acid sequence of the C-terminal tail in the cytoplasmic domain. In addition to the close similarity of the individual members of the I-type lectins, the localization of CD22 (Wilson et al., 1993), CD33 (Peiper et al., 1988), and MAG (Barton et al., 1987) to a syntenic region of human chromosome 19 suggests that the members have evolved from a common ancestor by gene duplication (Mucklow et al., 1995). This prediction is consistent with the fact that the exon boundaries of the CD22, CD33 and MAG (Lai et al., 1987b) genes fall within the same regions of the two N-terminal Ig-like domains, and that codons at the boundaries of the exons are in the same phase. Although sialoadhesion is located on a diVerent chromosome (Mucklow et al., 1995), this does not necessarily exclude the possibility that sialoadhesin arose from an ancestor common to CD22, CD33, and MAG. Each member of the siglec family is expressed in a restricted cell lineage and bind preferentially to certain cell types. SMP, which is most similar to MAG (Dulac et al., 1992), is expressed on quail Schwann cells (Dulac et al., 1988) and oligodendrocytes (Cameron-Curry and Le Douarin, 1995) and is presumed to bind to neurons. CD22, which is expressed on the B-cell lineage (Dorken et al., 1986; Engel et al., 1993; Stamenkovic and Seed, 1990; Stamenkovic et al., 1991), binds to B- and T-lymphocytes and monocytes (Crocker et al., 1995; Engel et al., 1993; Stamenkovic and Seed, 1990; Stamenkovic et al., 1991). In contrast, sialoadhesin is a nonphagocytic receptor, which is expressed on tissue macrophages (Stamenkovic and Seed, 1990; Stamenkovic et al., 1991) and binds selectively to neutrophils, although it will also bind to other cells from the granulocyte lineage such as lymphocytes (Crocker et al., 1995; van den Berg et al., 1992b). CD33 is expressed on the myelomonocytic lineage, such as monocytes and tissue macrophages (Pierelli et al., 1993), and selectively binds to cell lines from the myeloid lineage (Freeman et al., 1995). Sialic acids are a family of 9-carbon carboxylic acids usually added to the terminal positions of oligosaccharides by sialyltransferases in the Golgi apparatus (Schauer, 1982; Varki, 1992). Sialyltransferases form a family of proteins, which transfer sialic acid from the donor, CMP-sialic acid, to a speciWc oligosaccharide acceptor on glycolipids or glycoproteins (CorWeld et al., 1982; Datta and Paulson, 1995; Paulson and Colley, 1989). Each sialyltransferase catalyzes the attachment of sialic acid via an a-ketosidic linkage to diVerent positions on a monosaccharide unit of a speciWc disaccharide acceptor. Individual members of the siglec family preferentially recognize sialylated glycans produced by a speciWc sialyltransferase. CD22 preferentially recognizes sialylated glycans produced by Galb1-4GlcNAc a2,6 sialyltransferase (2,6N ST), which catalyzes the addition of sialic acids linked via an a2,6 linkage to Gal(b1,4)GlcNAc (2,6N), a disaccharide acceptor commonly found on N-linked oligosaccharides (Kelm et al., 1994; Stamenkovic et al., 1991). In contrast MAG (Kelm et al., 1994), sialoadhesin (Crocker et al., 1991; Kelm et al., 1994) and CD33 (Freeman et al., 1995) recognize sialylated glycans produced by either the Ga1b1-3GalNac a2,3 sialyltransferase (2,3-N ST) or Galb1-3(4)GlcNAc a2,3 sialyltransferase (2,3-O ST) which catalyze the addition of sialic acid via an a2,3 linkage to Galb1-3 GalNAc (2,3O) commonly found on O-linked oligosaccharides or Galb1-3(4)GlcNAc (2,3N) commonly found on N-linked oligosaccharides, respectively. MAG preferentially binds sialic acid containing proteins and glycans, especially the NeuAc a3 Gal b3 GalNAc structure, which is commonly found on gangliosides (Collins et al., 1997b; Crocker et al., 1996; Kelm et al., 1994). The ability of MAG and siglecs to bind sialylated glycans on the surface of cells represents not only a mechanism for cell adhesion but also a means for regulating adhesion. Thus, the adhesive functions of MAG (Tropak and Roder, 1997), CD22 (Sgroi et al.,

MAG BIOCHEMISTRY, STRUCTURE, AND ADHESIVE PROPERTIES

FIGURE 17.6 Potential binding partners of siglecs and their potential role in signal transduction. Potential binding partners for siglecs occur on cell surfaces and in the extracellular space. Siglecs can be clustered on cell surfaces by interacting with multivalent binding partners including cell surfaces, extracellular matrix, or soluble glycoproteins. Intracellular signaling cascades can then be activated, such as the tyrosine phosphorylation of MAG and CD22 cytoplasmic domains that occurs upon clustering of these siglecs. Siglecs can interact with glycoconjugates on the opposing cell (trans-interactions) or also with molecules on the same cell (cis-interactions). Binding of siglecs to trans-ligands could also dissociate the siglec molecules from their cis-binding ligands and inXuence the signaling properties of these ligands. Binding partners of siglecs on an opposing cell can be clustered through a sialic acidmediated interaction and subsequently start a signal cascade in the opposing cell. All these events depend on clustering of siglecs, which usually have a greater aYnity for multivalent ligands. Monovalent sialosides can also bind to siglecs and inhibit the interaction with multivalent ligands. Reprinted with permission from Kelm, 2001, Results and Problems in Cell DiVerentiation 33, 153-176, Paul R. Crocker, ed. Mammalian Carbohydrate Recognition Systems, 2001, Springer-Verlag Berlin Heidelberg.

1996), CD33 (Freeman et al., 1995), and sialoadhesin (Barnes et al., 1999) can be masked by high levels of sialylation in the cell expressing the siglec. One interpretation of these Wndings is that siglecs are unable to bind to sialylated trans-ligands on the apposing cell because its lectin binding site is occupied by a cis-ligand present in the cell on which the siglec is expressed (summarized in Fig. 17.6). However, sialoadhesin activity is less susceptible to cis-sialylation, presumably due to its size enabling its binding site to be presented above the sialylated glycocalyx surrounding the cell. In the case of CD22, cis-sialylation may prevent inappropropriate activation of the B-cell receptor (Cyster and Goodnow, 1997). The lectin activity of CD22 is unmasked once B-cells enter the lymph nodes where the B-cell receptor au3 may interact with foreign antigens rather than self-antigens.

MAG Adhesive Functions The adhesive function of MAG was demonstrated experimentally using several diVerent approaches. Poltorak and colleagues Wrst showed that binding of MAG-expressing mouse oligodendrocytes to mouse neurons could be blocked using anti-MAG polyclonal and monoclonal Abs (Poltorak et al., 1987). Subsequent incorporation of puriWed endogenous or recombinant rat MAG into Xuorescently labeled liposomes demonstrated binding to in vitro cultures of rodent dorsal root ganglion (DRG) neurons (Johnson et al., 1989; Poltorak et al., 1987). Figure 17.7 shows Xuorescently labelled liposomes containing recombinant S- or L-MAG, in which L-MAG and S-MAG bound primarily to neurons in spinal cord cultures, whereas only the S-MAG liposomes bound to DRG neurites. MAG liposome binding to neurons was blocked by anti-MAG monoclonal Ab 513 (Figs. 17.7E through 17.7F) and can also be blocked by monoclonal Ab 15 (Meyer-Franke

433

434

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

FIGURE 17.7 Neural Adhesion of Recombinant MAG proteins. One-week-old DRG cultures from newborn mice or 4-week spinal cord cultures from P13 embryonic mice. Liposomes labelled with carboxyXuorescein and containing recombinant S- or L-MAG were incubated for 30 minutes with cultures, washed, then visualized under optics to reveal phase (A, C, E, and G) and also Xuorescence (B, D, F, and H). (A–B) Spinal cord cultures incubated with L-MAG liposomes. (C–D) Spinal cord cultures incubated with S-MAG liposomes. (E–F) Spinal cord cultures were incubated with monoclonal anti-MAG 513 Ab (Fab fragment) and S-MAG liposomes. (G–H) DRG cultures were challenged with S-MAG liposomes. ModiWed with permission from Johnson, Abramow-Newerly, Seilheimer, Sadoul, Tropak, Arquint, Dunn, Schachner, and Roder, 1989, Neuron 3, 377–385, 1989 by Cell Press.

and Barres, 1994). MAb 513 was raised against glycoproteins immunopuriWed from chicken brain membranes using the L2/HNK-1 Ab (Poltorak et al., 1987). In contrast, the IgM rat MAb 15 was raised against immunopuriWed mouse MAG (Meyer-Franke et al., 1995). Initial experiments showed that MAG only bound to DRG neurons, which could be myelinated, but not to cerebellar neurons, which are not myelinated in vivo. Thus, MAG was thought to interact only with a neuronal ligand, which was found on axons destined to be myelinated. Subsequent experiments demonstrated that MAG liposomes could in fact bind to cerebellar neurons, possibly as a result of the diVerent culture conditions used to grow the neurons (Sadoul et al., 1990). Furthermore, it was shown that MAG could mediate the heterophilic aggregation of L cell Wbroblasts (Afar et al., 1991), CHO cells (Attia, 1992), and oligodendrocyte-like cells (Almazan et al., 1992). It is unlikely that these cells expressed a neuronal ligand. These results experimentally veriWed the prediction that MAG could function as a CAM. However, these observations also suggested that MAG could recognize a number of diVerent ligands. As has been demonstrated for most CAMs (see the review by van der Merwe and Barclay, 1994), the interaction of MAG with its ligands appears to be multivalent. Thus, neither an engineered soluble form of MAG nor the proteolytically generated soluble form of dMAG is able to bind to neurons (Attia, 1992; Sadoul et al., 1990). However, dMAG is able to bind with high aYnity (KD ~10-7 M) to various types of collagen and heparin (Fahrig et al., 1987; Probstmeier et al., 1992). The inability of MAG liposomes to bind to these ligands suggests that the manner in which MAG is presented (i.e., soluble versus on the cell surface) may aVect binding to collagen. Although the biological signiWcance of this type of interaction is not clear, it may be indicative of MAG’s ability to bind to diVerent ligands. MAG interacts with several extracellular matrix molecules. MAG binds to speciWc types of collagen and by so doing reduces collagen Wbril formation and also integrin-mediated

MUTATIONS AFFECTING MAG

adhesiveness of neural cells (Bachmann et al., 1995; Probstmeier et al., 1992). The ability of MAG to modify constituents of the extracellular matrix suggests it plays a role in controlling adhesive interactions and recognition between cells.

MUTATIONS AFFECTING MAG Naturally Occurring Mutations Affecting MAG The earliest indication of MAG’s role in mediating interactions between the axon and myelin sheath was provided by studies in the dysmyelinating recessive mouse mutant known as ‘‘quaking’’ (Sidman et al., 1964). The mutation that accounts for the phenotype in the quaking mouse (qk) occurs in the 5’ regulatory (noncoding) region of the qkI gene, which generates three alternatively spliced transcripts encoding the QKI proteins QKI-5, -6, and -7, putative RNA binding proteins suggested to link RNA metabolism with signal transduction (Ebersole et al., 1996; reviewed by Hardy, 1998). Compared to wild-type mice, QKI protein levels are reduced drastically in myelinating cells of qk mice (Hardy et al., 1996). In this mutant, there is an increased molecular weight of both MAG isoforms due to abnormal glycosylation ( Bartoszewicz et al., 1995; Matthieu et al., 1974). However, in qk mice, the expression of L-MAG is greatly reduced both at the RNA (Frail and Braun, 1985; Fujita et al., 1988) and protein levels (Bartoszewicz et al., 1995; Bo et al., 1995; Fujita et al., 1990). Another factor that contributes in reducing L-MAG levels in the qk mice appears to be increased endocytosis of L-MAG from periaxonal membranes (Bo et al., 1995). Regarding S-MAG expression in qk mice, there is an inverse eVect in that the relative levels are increased (Frail and Braun, 1985; Fujita et al., 1990). The modiWed L- and S- MAG levels in qk mice are a result of altered QKI expression. For instance, a recent study found that QKI-5 regulates the alternative splicing of MAG by repressing the inclusion of exon 12, which is skipped in L-MAG and included in S-MAG (Wu et al., 2002). The loss of QKI-5 in qk mice may thus be responsible for the alterations in MAG splicing, although it is not known whether the lack of QKI-6 and -7 also contribute. Myelin sheaths of qk mice are disrupted in regions lacking MAG. Immuno-EM studies revealed that in regions of the periaxonal space where MAG cannot be detected, the characteristic 12 to 14 nm space between the axonal membrane and Schwann cell membrane is altered (Trapp et al., 1984). Furthermore, in the periaxonal region where MAG was absent, the thickness of the periaxonal cytoplasmic collar (PCC) is reduced. Normally, myelinating glia from wild-type mice have a well-developed PCC that spans more than half of the axonal circumference. Although loss of regional expression of MAG is correlated with the previously stated morphological deWcits, the expression of other myelin proteins, such as MBP, is also aVected in qk (Li et al., 2000). Therefore, it is possible that the altered expression of other proteins besides MAG may account for the loss of the periaxonal collar in quaking. B. MAG knockout mice reveal a mixture of minor myelin defects MAG-deWcient mice reveal various subtle defects related to myelin formation and maintenance (also see the review by Schachner and Bartsch, 2000). MAG null mice have been generated by two separate groups using gene targeting methods (Li et al., 1994; Montag et al., 1994). These knockout mice have defects in myelin that are not severe and largely aVect CNS myelin sheaths. The modest changes observed in MAG null mice is in contrast to the observed inability of Schwann cells to segregate large caliber axons and form myelin in vitro, when MAG expression was greatly reduced (more than Wve-fold) using MAG antisense RNA (Owens and Bunge, 1991). Although MAG expression in the Schwann cells was reduced but not eliminated, it is not clear whether this can account for the diVerence between the in vitro and in vivo experiments. The phenotype is probably not due to secondary eVects of using MAG antisense RNA because Schwann cells infected with the MAG sense virus formed normal compact myelin. Mutant mice that express a truncated form of the L-MAG isoform have also been developed and these highlight the disparate importance of L- and S-MAG isoforms in the

435

436

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

FIGURE 17.8 MAG knockout mice have subtle changes in myelin morphology. Myelin morphology in PNS (upper panels) and CNS (lower panels) of MAG wild-type (þ/þ) and knockout (
/
) mice. (A–B) Light micrograph sections of L2 ventral spinal roots do not reveal any striking diVerences in myelin or neurons between MAG þ/þ and
/
mice, respectively. (C) Electron micrograph (EM) of a myelinated axon (Ax) from the L4 ventral root of a MAG
/
mouse, displaying a dilated periaxonal space (arrowheads). (D–E) EMs from L2 ventral root of MAG þ/þ and
/
mice, showing loss of the normal 12 to 14 nm periaxonal space (asterisk) and the Schwann cell cytoplasmic collar (arrow). (F–G) EMs from optic nerve Wber of MAG þ/þ and
/
mice reveal that the normal periaxonal space (see insert, arrowhead) and cytoplasmic collar (insert, arrow) are reduced or missing in oligodendrocytes from a knockout animal (except in the mesaxon region, where the cytoplasmic collar is present, arrowhead). (H) Longitudinal section of a MAG
/
optic nerve Wber showing an oligodendrocyte (ODC) with focal disorganization of the periaxonal cytoplasmic collar (asterisk). (I) Transverse section of MAG
/
optic nerve Wbers showing disrupted compact myelin lamella (asterisks) and redundant compact myelin (arrowheads). Scale bars: A,B ¼ 10 mm; C, F, G, H, I ¼ 0.5 mm; D, E ¼ 0.05 mm. Reprinted with permission from Li, Tropak, Gerlai, ClapoV, Abramow-Newerly, Trapp, Peterson, and Roder, 1994, Nature 369, 747–750, 1994, Macmillan Publishers Ltd.

CNS and PNS, respectively (Fujita et al., 1998). An emerging theme from studies of double knockout mice is that compensation by MAG-related molecules occurs in single mutants, and thus it has been diYcult to fully assess the normal role of MAG in vivo based on analysis of MAG mutant mice alone. Nonetheless, important details on MAG function have emerged. Delayed CNS Myelin Formation in MAG-Deficient Mice Myelin formation is delayed in the CNS, but not PNS, of mice possessing a null mutation in the MAG gene. Compared to wild-type mice, only half as many retinal ganglion cell axons are covered by compact myelin in 10 day old MAG mutants and optic nerves of adult MAG nulls contain more unmyelinated and small-sized axons (Bartsch et al., 1997; Li et al., 1998; Montag et al., 1994). Surprisingly, however, myelination in the PNS of MAG knockouts proceeds normally and essentially normal compact myelin is formed in both the PNS and CNS of young (<3 months) animals (Li et al., 1994; Montag et al., 1994). The changes in morphology of myelin in the PNS and CNS of MAG knockout mice are shown in the EMs of Figure 17.8.

au15 au16 au17

MUTATIONS AFFECTING MAG

Collapse of Periaxonal Cytoplasmic Collar in MAG Null Mice One of the subtle abnormalities observed in MAG-deWcient mice includes collapse of the PCC in the CNS but not everywhere in the PNS (Li et al., 1994; Montag et al., 1994). The EMs in Figures 17.8F and 17.8G show the presence and absence of PCCs from CNS optic nerve Wbers of MAG wild-type and MAG-deWcient mice, respectively. This defect is also shown in the EMs of Figure 17.9, which are from another line of MAG knockout mice. Also, as seen in the EMs of Figures 17.8D and 17.8E, Schwann cells from the L2 ventral roots of MAG null mice, but not wild-type mice, frequently lack PCCs (Li et al., 1994). It should be noted that abnormal PCCs do not occur in all regions of the nervous system since they are not detected in sciatic and femoral nerves (Fruttiger et al., 1995a; Montag et al., 1994). Similar to the diVerence in relative functional importance of MBP in the PNS and CNS, MAG may play a major role in the maintenance of the PCC in the CNS, whereas in the PNS, it seems to play a minor role. The loss of regional expression of MAG is correlated with collapse of the PCC in the PNS of qk mice (Trapp et al., 1984); however, the expression of other myelin proteins is also aVected in qk and thus the loss of additional proteins besides MAG may account for the loss of the PCC. Multiple Myelin Wrappings in MAG-Deficient Mice MAG may regulate CNS myelination via control of glia-axon recognition or adhesion. For instance, MAG-deWcient mice were found to have an increased frequency (2- to 50-fold) of axonal segments ensheathed by two or more individual myelin sheaths in young (<3 months old) and especially in mature (>11 months old) animals (Bartsch et al., 1995b; Li et al., 1994, 1998; Montag et al., 1994). Figures 17.8I and 17.9B show examples of retinal ganglion cell axons from MAG-deWcient mice that are surrounded by two distinct myelin sheaths. In some cases, up to four myelin sheaths have been observed, which do not necessarily spiral in the same direction around the axon. Related abnormalities include oligodendrocyte cytoplasm within compact myelin sheaths, redundant myelin loops, and also myelinated axons and myelin debris, which are grouped by a large myelin sheath (Fig. 17.8I). While myelination does proceed even in the absence of MAG, glia in the CNS fail to establish when, where, and how much myelin to form. It is not yet known whether the multiple myelin wrappings are formed during the initial stages of myelination or whether the original sheaths are later displaced from the axon by additional sheaths. Multiply myelinated structures are generated in the PNS when Schwann cells invade an existing myelin sheath and lay down a new sheath underneath the previous sheath (Kidd and Heath, 1988). In the CNS of MAG deWcient mice, sequential invasion of preexisting myelin sheaths by oligodendrocytic processes may result in multiply myelinated axonal segments. Thus MAG may play a regulatory role in myelination, possibly by recognizing myelination-competent axons. In fact, thin cytoplasmic membranes from several Schwann cells can normally contribute to form a myelin-like sheath, which surrounds the soma of spiral ganglion neurons, and, interestingly, MAG is only expressed in the adjacent myelin that surrounds axons but not in myelin around the soma (Martini, 1994). The lack of MAG may reduce adhesion between glia and axons and allow multiple glial sheaths to surround axonal membranes. Glial Cell Abnormalities in MAG Knockouts Analysis of aged mice lacking MAG unveiled a defect in the long-term maintenance of myelin. While no abnormalities could be detected in the PNS myelin of young (<10 weeks) mice with a null mutation in the MAG gene, degenerating axons and myelin sheaths were observed with increased frequency in the PNS of older (>9 months) MAG knockout mice (Fruttiger et al., 1995a; Weiss et al., 2001; Yin et al., 1998). Morphological analysis of peripheral nerves revealed the presence of onion bulbs, a feature of human peripheral neuropathies, which are formed upon degeneration-induced Schwann cell proliferation and remyelination. The degenerating myelin was often associated with degenerating axons, and it has been suggested that the formation of globular expansions of superXuous myelin, known as myelin tomacula, are responsible for the axonal degeneration.

437

438

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

FIGURE 17.9 Multiple myelin wrappings surround axons of MAG-deWcient mice. EM cross-sections of myelinated retinal ganglion cell axons (marked with an A) from 2-month-old MAG null mice. (A) Normal appearance of compact myelin (M), periaxonal space (arrowheads), and inner mesaxon (large arrowhead). Note the absence of a periaxonal cytoplasmic collar (compare with periaxonal cytoplasmic collar from wild-type axons shown in Fig. 17.8F). (B) Axon surrounded by two distinct myelin sheaths. Cytoplasm at the inner aspect of the inner myelin sheath (one star) is conWned to the inner tongue process (1). The inner myelin sheath terminates in the outer tongue process (2) in the immediate neighbourhood of the inner tongure process (3) of the outer myelin sheath (two stars). Note the separation between the two sheaths. The outer tongue process of the outer myelin sheath (4) appears near the top of the image. Scale bars ¼ 0.2 mm. Reprinted with permission from Montag, Giese, Bartsch, Martini, Lang, Blu¨thmann, Karthigasan, Kirschner, Wintergerst, Nave, Zielasek, Toyka, Lipp and Schachner, 1994, Neuron 13, 229–246, 1998 by Cell Press.

An infrequent CNS abnormality that occurs more often in older MAG null mice involves changes in the structure and function of oligodendrocytes. It is not known how early the changes can be detected; however, oligodendrocytes from 3-month-old MAGdeWcient mice begin to show signs of a disorganized cytoplasm with accumulation of organelles (Fig. 17.8H; also see Li et al., 1994). By 9 months of age, oligodendrocyte processes show a dystrophy that resembles dying-back oligodendrogliopathy (Lassmann et al., 1997; Weiss et al., 2000), commonly found in diseases of toxic or immune-mediated demyelination. Some of the defects that develop in the cytoplasm of oligodendrocyte processes include deposits of vesicular material, multivesicular bodies, mitochondria, and lipofuscin granules. The dying back process is believed to represent a response to injury, which does not involve oligodendrocyte apoptosis. The underlying mechanism is not known, but it could involve a loss of MAG-mediated signaling from myelinated axons to oligodendrocytes. The oligodendrocyte dystrophy in MAG nulls may be related to the similar pathology that occurs in multiple sclerosis (MS) and other demyelinating diseases. It is important to note that the oligodendroglial pathology occurs before any signs of demyelination. Altered Morphology of Axons in MAG Null Mice Axons in MAG knockout mice have a reduced caliber and more densely packed neuroWlaments (Yin et al., 1998). Cross-sections of myelinated axons from MAG null mice reveal that the 10-nm thick dots corresponding to neuroWlaments are spaced closer together, compared to axons from wild-type mice (Fig. 17.10). In unmyelinated Wbers of control mice and MAG-deWcient mice, neuroWlament spacing is identical. Thus, the reduced neuroWlament spacing correlates with the loss of MAG in the adaxonal Schwann cell membrane. Myelin itself is not responsible for regulating neuroWlament spacing because animals that cannot form myelin have increased axon diameters and normal spacing of neuroWlaments (Sanchez et al., 1996). Although it is discussed later on in the section titled ‘‘Autoimmune Neuropathy,’’ we brieXy mention here that neuroWlament spacing is

au18

MUTATIONS AFFECTING MAG

439

FIGURE 17.10

au19

MAG-deWcient axons have reduced neuroWlament spacing and axon caliber. (A) and (B) Electon micrographs showing neuroWlaments, which appear as 10 nm round dots (arrowheads), in cross sections of sciatic nerve from 9 month old control mice and MAG-deWcient mice, respectively. Arrows point to thread-like sidearms that extend from neuroWlaments. Note the wide spacing between neuroWlaments in axons from control mice. Scale bars: A, B ¼ 0.1 mm. (C) Schematic representation of axon-myelin sheath structure from the paranodal region in longitudinal (C1–C2) and transverse (C3–C5) sections. In axons from wild-type mice (C1 and C3), the paranodal region contains neuroWlament, which is spaced apart and correlates with large axonal caliber. Paranodal regions from MAG nulls (C2, C4, and C5) are surrounded by tomaculi, and neuroWlament spacings and axonal calibers are similar to those in nodal regions. Reprinted with permission from Yin, Crawford, GriYn, Tu, Lee, Li, Roder, and Trapp, 1998, Journal of Neuroscience 18, 1953–1962, 1998, by the Society for Neuroscience.

reduced in patients with demyelinating neuropathies associated with serum Ab against MAG (Lunn et al., 2002). Thus, there is adequate evidence to support a role for MAG in modulating the structure and possibly function of axons. Further work will need to establish whether a lack of MAG impedes nerve cell function. Biochemical Changes in MAG-DeWcient Mice MAG null mice have alterations in expression of other proteins, which are summarized in Table 17.1. Levels of the Schwann cell protein CNP (2’, 3’-cyclic nucleotide 3’-phosphodiesterase) are reduced by half in MAG knockout mice compared to control mice (Weiss et al., 2000, 2001). Also, expression of N-CAM was reported to be reduced, although other studies described either no changes or an increase (Li et al., 1994; Montag et al., 1994). There is also decreased expression of certain axonal neuroWlaments, which may contribute to changes in morphology described earlier. It is not clear whether any decrease in protein expression is a direct consequence of MAG absence or due to loss of membranes associated with demyelination. MAG Knockout Mice Have Neurological and Behavioral Defects In addition to morphological and biochemical changes, there are electrophysiological and behavioral abnormalities in MAG knockout mice. Compared to control animals, aged animals lacking MAG have reductions in sciatic nerve conduction velocity (Weiss et al., 2001). However, the changes were not as dramatic as those found in other demylinating neuropathies, and furthermore, there was no signiWcant change in action potential amplitude from which to suggest an axonal neuropathy. The reduction in axonal caliber that occurs in MAG nulls could explain why conduction velocity is impaired. Because MAG is also necessary for long-term maintenance of the myelin sheath, demyelination may also contribute to the electrophysiological abnormality. However, compact myelin is largely unaVected in MAG nulls, which favors the hypothesis that alterations in some aspect of axonal function is responsible.

Mice Lacking L-MAG-isoform Mutant mice lacking L-MAG have a phenotype similar to that of MAG knockout mice. The mice were made by introducing a stop codon into exon 13 of the MAG gene, which normally encodes for L-MAG containing the longer, carboxyl-terminal, cytoplasmic segment (Fujita et al., 1998). The L-MAG mutants produce a truncated L-MAG protein that

440

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

au21

TABLE 17.1 Comparison of MAG Knockout Phenotypes

Genotype

MAG -/-

L-MAG -/-

0

only 1-MAG

PLP -/-

MBP -/shiverer

MAG -/PLP -/-

MBP -/PLP -/-

3.5

0 0

2.7 0 0

MAG -/PLP -/MBP -/-

GaIN AcT -/-

Protein/ganglioside

mRNA

BIOCHEMISTRY MAG PLP MBP N-CAM 140 N-CAM 120 OMgp MAG PLP MBP GD1a/GT1b CNP NF-H, NF-M cdk5 and ERK 1/2

0 1.5 1.6

0

0 0 0

0.5

1.5 0 0.5 S-MAG 0 2 (PNS)

0 0

0 0 2 (PNS)

0 0 0

0 0 0 0

0.5 ~0.5 - 0.7 ~0.6 - 0.7

MORPHOLOGY Myelin

dissociation of outer surfaces, condensed and dissociated IDL

Axon degeneration

late onset (PNS)

late onset (PNS)

Axon Calibre

reduced

Glial cells

multiple myelin reduced PCC, sheaths, multiple myelin redundant sheaths, oligo myelin, dystrophy (CNS), similar to redundant MAG nulls myelin NO myelin and (CNS/PNS) axon axon degeneration of old mice with degeneration in aged mice myelin tomacula (PNS)

dissociation and decompaction of the IDL

hypohypomyelination myelination pseudopseudomyelin (CNS), myelin (CNS), more SLI (PNS)more SLI (PNS) pseudo-IDL

decreased CNS melination and axonal degeneration (CNS and PNS)

more SLI (PNS)

Number of Glia

Neuron

hypomyelination, reduced microtubule expression and density (CNS)

more oligos

early onset (CNS)

more oligos

more oligos

late onset (CNS)

early onset (CNS)

learning (latency)

0.5

0.1

open field

mean swim speed locomotion grooming rearing

0.5 0.3-1.1 0.2

specialized assays

water maze

BEHAVIOR

bar: grooming rotarod (time) horizontal wire horiz. bridge (latency)

0.5

0.2

0.1 0.4 0.1

0.6 0.6

0.6 0.5 0.5

0.1 0.4 0.1

0.4

0.3

0.4

0.1

0.5

0.3

0.1

0.1

tremor by P28

hi-frequency tremor

seizures

>24

>24

7-9

NEUROLOGY tremor / seizures

mild intention tremor

conduction velocity lifespan (months)

only > 12 mo. >24

tremor by P12

>24

3-4

(Continues)

441

MUTATIONS AFFECTING MAG

TABLE 17.1 (Continued )

N-CAM -/Genotype

MAG -/N-CAM -/-

PO -/-

MAG -/PO -/-

CGT -/-

MAG -/CGT -/-

Fyn -/-

MAG -/Fyn -/-

mRNA

MAG PLP MBP N-CAM 140 N-CAM 120 OMgp

Protein/ganglioside

BIOCHEMISTRY

MAG PLP MBP GD1a/GT1b CNP NF-H, NF-M cdk5 and ERK 1/2

0

0 0

0

0

0

0

0

0 0 0

0

MORPHOLOGY

Glial cells

Myelin

degeneration of axons and myelin

degeneration of axons and myelin occurs earlier than in N-CAM null

hypomyelination, abnormal myelin compaction

longer delay in myelin formation, abnormal myelin compaction

multiple myelin sheaths, redundant myelin disoriented paranodal loops, periaxonal splitting (CNS)

no PCC, hypo>50% less myelination, myelin (CNS), multiple myelin myelin sheaths normal sheaths, severely disoriented paranodal loops, periaxonal myelin splitting (CNS)

severe hypomyelination, 80% of optic nerve axons lack myelin, also all of same defects found in MAG nulls

Neuron

Number of Glia Axon degeneration

earlier onset

Axon Calibre learning (latency)

open field

mean swim speed locomotion grooming rearing

specialized assays

water maze

BEHAVIOR

bar: grooming rotarod (time) horizontal wire horiz. bridge (latency)

NEUROLOGY tremor / seizures

tremor by P14

obvious tremor, P12

conduction velocity lifespan (months)

3

<1

Regarding quantitative data, an increase greater than 50% appears in red, a reduction to less than 50% appears in blue, and green signiWes no change. These cut-off values are arbitrarily chosen and do not necessarily correspond with differences by statistical signiWcance. Unless noted otherwise, biochemistry reflects analysis of CNS myelin. Much of the data was obtained from tables found in Uschkureit et al. (2000) and other articles referenced in the chapter. Acronyms: SLI ¼ Schmidt-Lantermann-incisures, IDL ¼ interperiod dense line

442

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

is four amino acids shorter than S-MAG, whereas the endogenous S-MAG isoform is unaVected. CNS axons from L-MAG knockout mice were surrounded by multiple myelin sheaths, there was redundant myelin, and some oligodendrocyte cytoplasm was present in compact myelin. These abnormalities in the L-MAG mutants are similar to and occur at the same frequency as in MAG nulls (13 to 15% of CNS axons). In addition, L-MAG mutants develop a rapid tremor by P12 and tonic seizures in adulthood. Although the total MAG protein level was reduced by half in the L-MAG null mice, the reduced levels are likely not responsible for the observed defects because they do not occur in mice that are heterozygous for the MAG null mutation. Perhaps more interesting is the observation that older L-MAG mutants do not show signs of myelin and axon degeneration in the PNS. This is in contrast to the marked neuropathological abnormalities seen in aged MAG null mice. The simplest interpretation of the data is that L-MAG is not necessary and that S-MAG suYces for PNS myelin formation and maintenance. The extracellular domain of MAG appears to mediate processes involved in maintaining the integrity of the PNS. In contrast, the longer cytoplasmic domain found in L-MAG is critical for CNS myelin formation.

Multiple Gene Knockouts That Include Deletion of MAG There has been a tremendous eVort to Wnd myelin-related molecules that may functionally compensate for the lack of MAG in MAG knockout mice. Some myelin-related molecules can have overlapping function, and thus deletion of a single gene do not produce many defects. Compensation may sometimes be accomplished by up-regulation of related molecules. For instance, in MAG null mice as well as in MAG/PLP double knockouts, MBP expression is up-regulated twofold in peripheral nerves (Li et al., 1994; Uschkureit et al., 2000). There is also a two- or three-fold overexpression of S-MAG in the CNS of MBP nulls and PLP/MBP double mutants. Even when there are no such changes, the potential for overlapping function is still there, and it has become necessary to produce multiple gene knockouts by cross-breeding candidate single mutants to MAG nulls. This search has excluded a compensatory role for L1 because MAG/L1 double mutants do not have more deWcits than what occurs in the MAG knockout alone. Compared to control mice, MAG/ L1 double knockout mice have a similar number of myelinated axons (Haney et al., 1999). However, in several other combinations that are reviewed later in this chapter, there are more complex phenotypes, and these have conWrmed suspected MAG functions and also revealed some novel functions of MAG. Axonal Degeneration in Mice Lacking MAG, PLP, and MBP Studies on double and triple knockouts of MAG in combination with either PLP or MBP have corroborated the belief that MAG has a role in promoting axonal stability (Uschkureit et al., 2000). As an aid to comparing the phenotypes between the various MAG knockouts, a table has been contructed based largely on data from the previously mentioned study (Tab. 17.1). BrieXy, PLP is an integral CNS myelin sheath protein, which plays a role in stabilizing the apposition of the adjacent extracellular membrane surfaces. MBP comprises one-third of the CNS and one-tenth of PNS myelin and has an important role in the compaction of the opposing cytosolic surfaces of the plasma membrane processes. In PLP/MAG double knockouts, severe neurological symptoms develop, which is in contrast to their respective single mutants. A tremor of the hindlimbs develops by P28 and progresses to dragging and jerky-like movement with increasing age. Optic nerves from PLP/MAG double mutants show early onset and rapidly progessing axonal degeneration, which temporally parallels the neurological symptoms. Behavioral tests also revealed more severe motor and learning defects in mutant mice lacking both PLP and MAG (see Tab. 17.1). In triple mutant mice of genotype MAG
/
PLP
/
MBP
/
, a high-frequency tremor develops by 3 months after birth and the life span is reduced to about 8 months, in contrast to the PLP/MBP double-mutants that have a normal life span. There is early onset axonal degeneration in the triple mutant, again suggesting that the deprivation of MAG is detrimental to axons. The Wndings support a role for MAG in maintaining the integrity of CNS axons.

MUTATIONS AFFECTING MAG

MAG/N-CAM Double Knockouts N-CAM partially compensates for MAG functions related to maintenance of axon-myelin integrity. While there is some uncertainty as to whether N-CAM expression is altered in MAG knockouts, N-CAM is still present and could take over functions normally carried out by MAG. To examine the extent of possible compensation by N-CAM in myelin formation and maintenace, MAG/N-CAM double knockouts were generated by crossbreeding single mutants (Carenini et al., 1997). Myelin formation was not aVected in the double mutants; however, degeneration of axons and myelin occurred earlier than in MAG-deWcient single mutants. More pathological abnormalities were detected in the MAG/N-CAM double mutants by 4 to 8 weeks of age, although by 26 weeks both mutants were equally aVected. Therefore, there was no indication that these molecules are necessary for myelin formation, however, both contribute to maintenance of axon and Schwann cell function. A subsequent study determined that N-CAM and MAG can contribute to myelin formation, although these functional roles are only detectable in the absence of P0. Delayed Myelination in MAG/P0 Null Mice The most abundant glycoprotein in PNS myelin is P0, which when absent causes severe dysmelination (Giese et al., 1992). Interestingly, P0 contains an Ig-like extracellular domain, which confers recognition ability. MAG/P0 null mice were generated to analyze their combined roles, and the double mutants had abnormal compaction in myelin sheaths that was similar to that in P0 null mice (Carenini et al., 1999). However, the number of axonSchwann cell units devoid of myelin is greater in 10-day-old MAG/P0 double mutants compared to P0 mutants. This indicates that in mice lacking P0, the formation of turning loops of Schwann cell processes depends on the presence of MAG. The role of MAG in promoting the spiraling of myelinating Schwann cell processes cannot be detected in MAG null mice because P0 seems to compensate. In P0/MAG double-deWcient mutants, the lack of N-CAM also impairs Schwann cell spiraling. However, the eVect is more transient. Thus, MAG has functions in myelin formation that cannot be duplicated fully by other molecules. MAG/Galactolipid-Deficient Mice Show Defects in Adhesion MAG- and galactolipid-deWcient mice have some common deWciencies, and double mutants were generated to test for compensatory activity. Mice in which the gene for the enzyme UDP-galactose:ceramide galactosyltransferase (CGT) is mutated can no longer convert ceramide to the myelin galactolipids galactocerebroside and its sulfated form sulfatide (Bosio et al., 1996; Coetzee et al., 1996). CGT mutant mice have a progressive neuropathological phenotype, which includes tremoring by P14, and they die near P86. However, similar to the MAG mutant, myelin formation is largely unaVected and in the CNS of CGT mutant mice, and there are three times as many unmyelinated processes as well as multiply myelinated axons (Dupree et al., 1998a, 1998b; Marcus et al., 2000). MAG/CGT double mutants have a more severe phenotype than the single CGT mutant with a much more obvious tremor and rapid hindlimb paralysis and reduced life expectancy of 22 days (Marcus et al., 2002). The double knockouts have a noticeably absent PCC whereby eight-fold fewer processes retain a normal periaxonal space and eight times more spinal cord Wbers display periaxonal splitting. There are additional exacerbations, such as in paranode structure, which we will not discuss here. The report concludes that MAG and myelin galactolipids act synergistically at the paranode to promote its structural maintenance. Also, these molecules functionally overlap with one another in maintaining intercellular adhesion along myelinated axons. While there are no recognized axonal ligands for galactolipids, the authors note that each of sulfatide and MAG have been shown to au4 interact with members of the tenascin extracellular matrix protein family, which could modulate adhesion during myelin formation. Hypomyelination in MAG/Fyn Double Mutants The multiple knockout approach has been used to assess whether MAG eVects on myelin involve intracellular signaling via Fyn. Fyn is a nonreceptor-type tyrosine kinase of the Src

443

444

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

family, which when absent leads to signiWcant impairment of myelination (Umemori et al., 1994). Fyn is found in myelin-forming cells and is activated by stimulation through cell receptors including L-MAG, which subsequently leads to transcription of the MBP gene for myelination (Umemori et al., 1999). Compared to wild-type mice, Fyn-deWcient mice have only half as much myelin in their brain and even less in spinal cord. To determine whether MAG and Fyn interact in the myelination process, double knockouts were generated by cross-breeding the single mutants (BiYger et al., 2000). Double knockout of MAG and Fyn revealed hypomyelination that is far more severe than that of single mutants. Half of the axons from optic nerve are unmyelinated in Fyn null mice, but this ratio increases to 80% in MAG/Fyn double knockouts, which is greater than the 20 to 30% found in MAG nulls. If MAG and Fyn operate via the same signaling mechanism, then there should not have been a more severe phenotype in the double mutants. However, it is still possible that MAG and Fyn actions are not independent because compensatory mechanisms may exist in the single mutants. For instance, a MAG-associated kinase activity still persists in Fyn nulls (Umemori et al., 1994), which could partially compensate for the loss of Fyn. MAG/Fyn double mutants have the same morphological defects in their myelin sheaths as in MAG null mice. There is redundant myelin, multiply-myelinated axons, degeneration of myelin sheaths, impairments in the oligodendrocyte cytoplasmic collar, and noncompacted regions of myelin. It is important to note that Fyn single mutants do not have any defects in their myelin sheaths, which suggests that Fyn does not participate in the formation of normal, intact CNS myelin. MAG can therefore activate other downstream signaling molecules to produce morphologically intact myelin. MAG Can Act as an Inhibitor of Axon Regeneration Finally, we will discuss brieXy the inhibitory role of MAG in axonal regeneration, which was demonstrated in crosses of MAG nulls with C57BL/Wlds mutants. After injury to the PNS of C57BL/Wlds mutant mice, axon degeneration occurs slowly, as does removal of myelin due to poor leucocyte invasion (Brown et al., 1991, 1992; Lunn et al., 1989). Together with the observation that Schwann cells with the greatest amount of lesioninduced myelin breakdown are more frequently contacted by regrowing axons (Fruttiger et al., 1995b), this suggested that myelin inhibits axonal regeneration. However, oVspring generated from crosses of C57BL/Wlds and MAG null mice show much improved axonal regeneration in regions along Schwann cells with nondegenerated myelin (Scha¨fer et al., 1996). Still, these mutants have even better axonal growth in regions lacking intact myelin, which has been interpreted to indicate that either additional myelin-associated inhibitors exist or that the persistent myelin sheaths can act as a mechanical barrier for sprouting axons. Nonetheless, the presence of MAG in the PNS of C57BL/Wlds mutant mice partially contributes to inhibit axon growth. It is not known whether this inhibitory activity of MAG has any biological signiWcance, because upon injury to nerves from normal wild-type mice, myelin is removed and Schwann cells down-regulate expression of myelin-related proteins. Thus, MAG is absent in the injured PNS, and its growthinhibitory function may only be important in the normal adult for limiting aberrant growth. The following section focuses on this inhibitory aspect of MAG, together with the receptors involved.

CONTROL OF AXONAL GROWTH AND REGENERATION BY MAG During development of the nervous system, axons send out extensions at a rate and direction that is regulated by factors in their environment. When axons reach myelin, their growth is blocked. MAG and several other proteins found in CNS myelin are inhibitory to axon outgrowth and regeneration. We will discuss the functions that MAG plays in mediating inhibition, including the role of recently identiWed MAG receptors. While the focus will be on axon growth and regeneration, it is important to review these

au5

CONTROL OF AXONAL GROWTH AND REGENERATION BY MAG

events because the mechanisms will almost certainly have some relevancy to the process of myelination.

MAG Promotes Axonal Growth during Early Development Although MAG is a potent axonal growth inhibitor in the adult nervous system, during early development, MAG is not inhibitory. Indeed, most of the founding experiments that studied the role of MAG on neurite outgrowth made use of embryonic and neonatal neurons. MAG increases outgrowth of young cerebellar and spinal cord neurons and also neonatal DRG neurons (Johnson et al., 1989; Matsuda et al., 1996; Mukhopadhyay et al., 1994; Turnley and Bartlett, 1998). However, when post-natal neurons are exposed to MAG, their axonal growth is inhibited. Unfortunately, this developmental switch in the neuronal response to MAG was not realized for quite some time, causing considerable confusion to scientists. It is not clear whether the neurotrophic-like eVect of MAG on young neurons has any biological signiWcance. For instance, axons have stopped growing by the time MAG becomes expressed at detectable levels, and thus MAG is likely not normally involved in the promotion of neurite growth during early development. Nonetheless, a recent investigation has found that the L-MAG isoform is responsible for promoting axonal growth. SpeciWcally, when cerebellar neurons from newborn mice were cultured on BALB/c 3T3 cells transfected with cDNA for either the S- or L-MAG isoform, only the latter had growth promoting activity (Shimizu-Okabe et al., 2001). Because L-MAG is the principle form that is expressed during development, it is speculated to function in enhancing axon outgrowth. It remains to be seen whether low levels of L-MAG exist during development or perhaps could assist axonal recovery following injury or disease. Also, by selecting young neurons that are not inhibited by MAG and do not express receptors that contribute to inhibiton of axonal outgrowth (discussed later), it may be possible to use young neurons as an aid for axon regeneration.

In the Adult, MAG Functions as an Inhibitory Molecule MAG was the Wrst myelin-derived molecule shown to be inhibitory to axonal growth in the adult nervous system. MAG inhibits growth of neurites (including neuronal-like cells) regardless of whether it is presented in soluble forms containing the extracellular domain, is expressed by live cells, or when introduced to axons by way of immobilized membrane prepared from MAG-expressing cells (DeBellard et al., 1996; Li et al., 1996; McKerracher et al., 1994; Mukhopadhyay et al., 1994; Shen et al., 1998; Tang et al., 1997b). Furthermore, Schwann cells that are induced to express MAG also inhibit regeneration, demonstrating that MAG is inhibitory in both the adult CNS and PNS. Several factors that aVect MAG eYcacy were not recognized until recently, and this also hindered progress on accepting it is an inhibitory molecule. For instance, neuronal growth cones were found to collapse upon encountering MAG-coated beads (Li et al., 1996), whereas another study did not Wnd any such eVect using MAG extracted from myelin (Bartsch et al., 1995a). However, the inhibitory role of MAG on neurite outgrowth has also been conWrmed using MAG fused to the Fc portion of human IgG and also dMAG (Tang et al., 1997b). The authors emphasized several important details, which should be considered when studying MAG and which likely explain the seeming discrepancies. For instance, the manner by which MAG is presented can aVect activity, in that MAG is multimeric when coated onto beads, compared to its monomeric, relatively denatured format when extracted from myelin and solubilized. Furthermore, the MAG content in puriWed myelin varies considerably and MAG is also rapidly (within minutes to hours) converted to dMAG, which would be lost upon any washing. MAG-mediated inhibition of axonal growth can also be overcome by the presence of other proteins and also by neurotrophins. Laminin, a component of the Schwann cell basal lamina, can override the growth inhibitory activity of MAG (David et al., 1995). Hence, after puriWcation of MAG from myelin, any residual laminin can nullify MAG inhibition

445

446

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

of neuronal outgrowth. MAG also binds the extracellular matrix glycoprotein tenascin-R which is expressed by oligodendrocytes (Yang et al., 1999). Tenascin-R neutralizes MAGmediated inhibition of neurite outgrowth. Finally, neurotrophins such as BDNF or GDNF have been shown to prevent MAG-mediated inhibition of axonal regeneration (Cai et al., 1999). Upon injury to the adult nervous system, myelin-forming glia will down-regulate MAG. This biological change is of extreme beneWt, given the inhibitory attribute of MAG on axonal outgrowth. Indeed, axons in the PNS can regenerate successfully, especially when aided by the additional removal of any remaining MAG by macrophages and also by the induction of regeneration-associated genes. The inhibitory nature of MAG in the PNS has been demonstrated by the Wnding that axonal regeneration can be improved in C57BL/ Wlds mutant mice (which do not clear myelin and show hardly any axonal regeneration) after crossing them to MAG null mice (as already discussed; also see Scha¨fer et al., 1996). However, axons in the CNS do not grow back, a problem that is of great clinical relevance for injuries and diseases of the nervous system and that has preoccupied researchers for many years. Myelin in the CNS contains two additional proteins that inhibit neuronal growth. Caroni and Schwab identiWed two membrane proteins, NI35 and NI250, and also raised IN-1 Abs, which detected these proteins and could counteract the nonpermissive properties of CNS myelin (Caroni and Schwab, 1988a;1988b). NI250 was next puriWed and shown to inhibit neurite outgrowth (Spillmann et al., 1998). Several labs then used peptide sequences of NI250 (now called Nogo-A) to clone the Nogo gene, named for its ability to block axonal growth (Chen et al., 2000; GrandPre et al., 2000; Prinjha et al., 2000). Nogo codes for three proteins, Nogo-A, -B, and -C, and there is agreement that Nogo-A is inhibitory. All three isoforms share a conserved C-terminal region with two transmembrane domains connected by a 66-amino acid linker sequence known as Nogo-66. The Nogo-66 domain is believed to be on the extracellular side of membranes and is most likely the domain responsible for inhibition of neurite outgrowth. Curiously, the N-terminal region of Nogo-A is also inhibitory, although the mechanism of action is diVerent from Nogo-66 (Chen et al., 2000; Fournier et al., 2001; Prinjha et al., 2000). Also, the aminoNogo domain is suspected to be cytoplasmic and it may only play a role if released following disease or injury. Further work is necessary to sort out the details of Nogo membrane organization and domain diVerences in mediating inhibitory actions. Finally, the last known myelin-derived inhibitor is the oligodendrocyte-myelin glycoprotein (OMgp). OMgp is a glycophosphatidylinositol (GPI)-anchored protein, which remained elusive until very recently when it was identiWed to bind to the same receptor as Nogo and to inhibit neurite outgrowth in vitro (Wang et al., 2002). A receptor for Nogo66, named the Nogo-66 receptor (NgR), was identiWed by expression cloning using a mouse cDNA library (Fournier et al., 2001). NgR was determined to contain a signal sequence followed by eight leucine-rich repeat domains and a cysteine-rich region, followed by a GPI linkage region. The human homologue of the NgR gene has exons (separated by nearly 30 kb) on chromosome 22q11. The NgR cDNA sequence encodes a protein 473 amino acids long with a 85 kDa apparent molecular mass. There is widespread expression in the brain and very little expression in the PNS. In summary, MAG, OMgp, Nogo, and NgR all contribute to the inhibitory activity of CNS myelin.

Identification of Functional Neuronal MAG Receptors MAG Binds the Nogo Receptor, NgR Immediately prior to when this chapter was written, several groups made the important discovery of identifying the NgR as a common functional receptor for MAG and OMgp (Domeniconi et al., 2002; Liu et al., 2002; Wang et al., 2002). Figure 17.11 contains experimental data from the article by Liu and colleagues and elegantly demonstrates that MAG mediates inhibition of axonal outgrowth via NgRs. First, NgR levels in DRG neurons are shown to be minimal during embryonic development and to increase into

CONTROL OF AXONAL GROWTH AND REGENERATION BY MAG

FIGURE 17.11 The Nogo receptor (NgR) controls inhibitory responses induced by MAG. (A) Confocal microscope images of mouse DRG neurons showing increase in NgR levels, as detected by anti-NgR Abs and Xuorescent secondary Abs, from embryonic day 17 (E17) to post-natal day 64 (adult). (B) Comparison of neurite outgrowth under control conditions containing IgG versus Fc-MAG, the extracellular domain of MAG fused to the Fc portion of IgG. Some experiments were carried out in the presence (þ) or absence (
) of PI-PLC (to remove GPI-anchored proteins including NgR) or NgR-Ecto (truncated soluble NgR fragment that antagonizes Nogo-66). The asterisk denotes a statistically signiWcant diVerence between MAG and control treated adult neurons, and there is no diVerence after interfering with NgRs. (C) Embryonic (E7) chick DRG explants were infected using recombinant virus to drive expression of green Xuorescent protein (GFP) or NgR and treated with either GST control (glutathione S-transferase), GST-Nogo-66, IgG control, or Fc-MAG. (D) Quantitation of growth cone (GC) collapse shows that embryonic neurons are normally not responsive to MAG or Nogo-66, unless they are forced to express NgR (asterisks denote statistically signiWcant diVerences). Reprinted with permission from Liu, Fournier, Grandpre´, and Strittmatter, 2002, Science 297, 1190–1193, 2002, American Association for the Advancement of Science.

adulthood (Fig. 17.11A). Second, MAG-mediated neurite outgrowth is eliminated after blockade of NgR operation (Fig. 17.11B). Several methods have been used in this and other studies described earlier to interfere with normal NgR activation, including (1) antibodies against NgRs, (2) soluble NgRs to compete for binding with neuronal receptors, (3) removal of NgRs by cleaving its GPI anchor, and (4) overexpression of truncated NgRs lacking the extreme C-terminal region (e.g., dominant-negative form of NgR), all of which prevent MAG/OMgP/Nogo-mediated inhibition. The last panel of Figure 17.11C shows that embryonic neurons, which are generally unresponsive to growth-inhibition by MAG and Nogo-66 proteins, become responsive when exogenous NgR is introduced into them. The assay involves measurement of growth cone collapse in DRG explants infected with recombinant virus to drive expression of green Xuorescent protein alone or also NgRs. Similar experiments were used by Wang and coworkers to show that introduction of

447

448

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

exogenous NgRs confers OMgP responsiveness to otherwise insensitive neurons (Wang et al., 2002). Thus, the developmental switch in neuronal MAG sensitivity very likely involves an up-regulation of NgRs. Regardless, there is convincing evidence that NgRs are both necessary and suYcient for MAG-mediated growth inhibition in adult axons. Is the interaction between NgRs and their multiple binding partners regulated? Although Nogo-66, OMgp, and MAG do not share any sequence similarity, binding studies show that they all bind to the NgR with very similar aYnities (7, 5, and 8 nM, respectively). Figure 17.12 contains a model that illustrates the interactions of these three myelin-derived inhibitors with neuronal receptors including NgRs, as well as gangliosides and neurotrophin receptors, which will be discussed later. It is interesting that excess Nogo-66 does not prevent MAG-NgR binding (see Liu et al., 2002). Whereas MAG and Nogo-66 have separate binding sites within the same domain of NgR, OMgp and Nogo-66 have an overlapping binding site. If there is no preferential binding partner for NgRs, the particular membrane expression level for each NgR ligand may be an important factor that will determine the extent of axonal inhibition. Also intriguing is the fact that the MAG-NgR interaction is largely unaVected by the introduction of GT1b, a neuronal ganglioside that also interacts with MAG and will be covered in the next section. The MAG-NgR interaction may represent a more speciWc, biologically relevant interaction. Indeed, Nogo and MAG appear to be the main inhibitors in myelin, because when myelin is prepared from mice lacking Nogo-A, there is reduced inhibition of axon growth, and the residual activity is completely blocked by anti-MAG Abs (unpublished data; see Liu et al., 2002). It will be crucial to establish whether MAG, Nogo66, and OMgp bound to NgR and gangliosides all partake to regulate neurite outgrowth in vivo and whether complexes consisting of diVerent combinations of these players can alter function. The competition by multiple ligands for binding to the same receptor is a simple explanation for the redundancy observed when only a single ligand is blocked. The application of IN-1 Ab to block Nogo has only modest eVects on improving axonal regeneration (Schnell and Schwab, 1990). Similarly, axons in MAG null mice do not show any signiWcant improvement in regeneration ( Li et al., 1996; Montag et al., 1994). Although there are no reports published on whether interfering with OMgp can improve axon outgrowth, a lack of signiWcant eVect is predicted. In contrast, a peptide antagonist of NgR (derived from a 40 amino-terminal fragment of Nogo-66, NEP1-40) is an eVective blocker of myelin-mediated inhibition of axonal outgrowth and also signiWcantly improves axon growth and functional recovery following spinal cord injury (GrandPre et al., 2002). Thus, the NgR appears to be a key player in the inhibitory actions of CNS myelin. This would be consistent with a model by which all three inhibitory ligands only exert their physiological inhibitory actions via one receptor, NgR. An obvious next step will be to generate a cocktail containing blockers of Nogo, MAG, OMgp, and also NgR in order to test whether this allows for superior axonal growth. The blockers will need to be designed to act in a transient function, because otherwise there may be excessive outgrowth and possibly an inability to form proper myelin. If this can be overcome, it should yield a potent theurapeutic agent that promotes functional recovery in many diseases and injuries aVecting neurons. Gangliosides Are MAG Receptors The nerve cell surface gangliosides (sialic acid-containing glycosphingolipids) GD1a and GT1b are receptors for MAG (see the commentary by McKerracher, 2002). MAG binds sialic acid containing proteins and glycans, especially the NeuAc a3 Gal b3 GalNAc structure, which is commonly found on gangliosides (Collins et al., 1997b; Crocker et al., 1996; Kelm et al., 1994). Binding studies indicate that GD1a and GT1b are the primary complex gangliosides in the brain that bind MAG (Collins et al., 1997a; Yang et al., 1996). Mice that lack the enzyme GM2/GD2 synthase, which is required for biosynthesis of complex gangliosides including GD1a and GT1b, display pathological features resembling those found in MAG knockouts (Sheikh et al., 1999). There is decreased CNS myelination and axonal degeneration in both CNS and PNS, and demyelination in the PNS. It should be noted that MAG

au6

CONTROL OF AXONAL GROWTH AND REGENERATION BY MAG

FIGURE 17.12 Model of MAG-mediated inhibiton of axon regeneration. Oligodendrocytes in the CNS produce three known growth-inhibitory molecules, MAG (red), OMgp (orange), and Nogo-A (yellow). The exact orientation of Nogo in the myelin membrane is not yet defined, however, both the N-terminal domain and Nogo-66 domain are inhibitory. MAG, OMgp and Nogo-66 all bind to the same neuronal receptor, NgR (Nogo receptor, pink), which results in blockade of neurite outgrowth and growth cone collapse. NgR lacks an intracellular domain and requires an unknown co-receptor to transduce intracellular signals via molecules such as Rho GTPase. MAG also binds to a complex consisting of gangliosides GD1a and GT1b, the latter of which interacts with the neurotrophin receptor p75NTR. A signal for MAG-mediated inhibition of neuronal elongation is transduced through GT1b/p75NTR and activation of Rho. The photomicrographs show DRG neurons grown on a laminin substrate that is permissive for growth (right, GO) or on a CNS myelin substrate that prevents growth (left, NOGO). Reprinted with permission from Woolf and Bloechlinger, 2002, Science 297, 1132–1133, 2002, American Association for the Advancement of Science.

expression is reduced by half in these GalNAcT
/
mice, although MAG heterozygote mice have similar reduction in MAG expression, yet they lack any neuropathology. Neurons from these GalNAcT knockout mice are less sensitive to MAG-mediated inhibition of neurite outgrowth (Vyas et al., 2002). Neurite outgrowth mediated by MAG is reduced upon addition of soluble GD1a and GT1b and also by Abs against GT1b, but not GD1a (Vinson et al., 2001). A subsequent report conWrms the role of ganglioside GT1b in blocking neurite outgrowth and shows that anti-GD1a Ab is also eVective (Vyas et al., 2002). Therefore, gangliosides GT1b and possibly GD1a regulate neurite outgrowth, and are neuronal receptors for MAG. While evidence supports that MAG interacts with gangliosides, there is some debate as to whether gangliosides are necessary for MAG’s inhibition of axonal outgrowth. For instance, removal of sialic acid residues does not aVect binding of MAG to NgRs, receptors that mediate inhibition (Domeniconi et al., 2002). When MAG is mutated at arginine 118 (R118) to abolish sialic acid-dependent binding, MAG can no longer bind to neurons; however, when R118-mutated MAG is expressed on the surface of Schwann cells or CHO

449

450

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

cells, it is still a potent inhibitor of axonal outgrowth (Tang et al., 1997a). These Wndings were interpreted to indicate sialic acid binding is insuYcient to aVect inhibition of axonal regeneration and that a second site on MAG must be responsible for this eVect. Subsequent work suggested that the residual inhibition in the R118 mutant is dependent on sialic acid (Vinson et al., 2001). Therefore, the R118-mutant MAG may have enough sialic acid binding to inhibit neurite outgrowth. However, there is additional evidence that sialic aciddependent binding to neurons is insuYcient to induce inhibition. Sialoadhesin (siglec-1) and a truncated form of MAG, MAG(d1-3)-Fc each recognize the same sialic acid linkage as MAG and also bind neurons in a sialic-dependent manner; however, none are inhibitory. Thus, there are probably multiple unique recognition sites on MAG, at least one of which binds sialoglycoproteins and another that mediates inhibition. MAG may inhibit neurite outgrowth via multivalent clustering of gangliosides on neurons. Ganglioside-mediated inhibition of neurite outgrowth is independent of MAG because such inhibition can be induced in the absence of MAG by Ab-mediated clustering of GT1b (Vinson et al., 2001; Vyas et al., 2002). Compared to bivalent IgG Abs, precomplexed Abs, and decavalent IgM Abs are more eYcient in blocking axonal outgrowth in response to MAG. MAG binding to gangliosides may also function to cluster other signaling molecules, or even to regulate MAG activity, either of which could help potentiate inhibition of axonal growth (see Fig. 17.6). Additional neuronal receptors may exist for MAG. For instance, although they are relatively scarce in brain, the a-series of gangliosides bind MAG with greater potencies than GD1a and GT1b (Collins et al., 1999). However, the roles of a-gangliosides, if any, remain to be elucidated. In addition, De Bellard and Filbin have used MAG-Fc to precipitate several neuronal surface proteins (De Bellard and Filbin, 1999). The two most prominent proteins that bind MAG are of 190 and 250 kDa size, and the former is a sialoglycoprotein. Further work will need to further identify these potential MAG receptors as well as their actions.

SIGNALING EVENTS VIA MAG AND INTERACTING PROTEINS Neuronal Signaling through MAG Receptors Signaling via Neurotrophin Receptor, Rho GTPase, and cAMP NgR lacks a transmembrane domain and is attached to the neuronal extracellular membrane by its GPI anchor. Without any cytoplasmic domains, intracellular signal transduction cannot occur. Thus, MAG (as well as Nogo-66 and OMgp) signaling through NgRs must be mediated via an associated transmembrane receptor. Similarly, gangliosides require signaling partners to transduce any signal across the cell membrane. The only known co-receptor is the low-aYnity nonselective neurotrophin receptor p75 (p75NTR), which binds GT1b (Yamashita et al., 2002). MAG failed to inhibit outgrowth in neurons from mice carrying a mutation in the p75NTR gene. Further experiments on neurons from these mutant mice revealed that activation of the small GTPase Rho was lost. Finally, MAG fusion protein co-localizes with anti-p75NTR staining, suggesting they spatially interact. Hence, MAG inhibits neurite outgrowth via p75NTR-mediated activation of Rho. It is not known whether MAG/NgR-mediated inhibition also operates by way of Rho signaling. The p75NTR-mediated signal was suspected to involve Rho because previous experiments had shown MAG activates Rho. Inactivation of Rho not only blocks growth inhibition by MAG and myelin, it also permits regeneration in vivo (Lehmann et al., 1999). Neurons that cannot transduce Rho signals (expression of dominant-negative Rho) do not respond to MAG. Conversely, blockade of Rho-GTPase activity (using Clostridium botulinum C3 exoenzyme) allows neurons to grow on MAG substrates. An inhibitor of Rho kinase (Y27632), the downstream eVector of Rho, blocks inhibition of neurite outgrowth induced either by myelin, MAG or anti-GT1b Ab ( Dergham et al., 2002; Vinson et al., 2001). Furthermore, the latter study found Y27632 is of theurapeutic value because it promotes

SIGNALING EVENTS VIA MAG AND INTERACTING PROTEINS

axonal regeneration and functional recovery after spinal cord injury. To summarize, MAG binds a receptor complex consisting of ganglioside GT1b and p75NTR, which serves as a signal transducer for neurite outgrowth inhibition via Rho and Rho kinase activation (see Fig. 17.12). It will be important to decipher the molecular participants in signaling events downstream of Rho, including activation of the Ser/Thr kinase known as Rho kinase. Another remaining issue is whether transduction of an inhibitory signal via p75NTR can be modulated according to whether p75NTR interacts with MAG or neurotrophins. Thus, p75NTR may act as a switch that permits outgrowth in environments containing neurotrophins, only to suppress outgrowth upon binding to MAG. Neurotrophins and activation of the cAMP pathway can override MAG-mediated inhibition. Activation of p75NTR by the neurotrophin NGF abolishes Rho activation and enhances neurite elongation (Yamashita et al., 1999). In addition, the neurotrophins BDNF and NT-3, which do not directly bind p75NTR but may interact via association with Trk receptors, can also stimulate neurite outgrowth. Priming of neurons by exposing them to BDNF or GDNF, or also by providing them with a cAMP analog, suspends the inhibitory actions of MAG on axonal outgrowth and regeneration (Cai et al., 1999; Song et al., 1998). It is speculated that MAG binding to neurons activates an inhibitory G-protein (Gi) that functions to block increases in cAMP and prevents growth. Introduction of cAMP analogs into spinal cord neurons allows them to regenerate (Neumann et al., 2002; Qiu et al., 2002). Interestingly, a reduction in endogenous levels of cAMP correlates with the developmental switch in neuronal response to MAG, which begins with promotion of neurite outgrowth in young neurons, and then changes to inhibiton (Cai et al., 2001). Do Rho and cAMP signals interact, and what eVectors control neurite outgrowth? Elevation of cAMP activates protein kinase A (PKA), which subsequently phosphorylates Rho to induce its dissociation from the cell membrane (Lang et al., 1996). This dual regulation of Rho activation and inactivation could act as a switch to control neuronal responses to MAG and neurotrophins. However, there are additional consequences of neurotrophin/cAMP signaling. Just lately, cAMP was shown to up-regulate Arginase I, an enzyme involved in the synthesis of polyamines (Cai et al., 2002). After polyamine synthesis is blocked, BDNF and a cAMP analog can no longer suppress MAG- and myelinmediated inhibition of neurite outgrowth. Either overexpression of Arginase I or priming with polyamines blocks inhibition mediated by MAG or myelin. Interestingly, the endogenous levels of Arginase I in DRG neurons drops during the Wrst post-natal week and appears to coincide with the developmental switch from promotion to inhibition of neurite outgrowth in response to MAG. The authors discuss potential mechanisms by which polyamines can modulate neurite outgrowth and regeneration (also see the review by Skaper et al., 2001). However, as discussed earlier, the developmental increase in NgR receptor also has been suggested to play a role in the post-natal switch in neuronal MAG sensitivity. It is possible that NgR and BDNF receptor signaling pathways interact to control neurite outgrowth. Future work will need to identify the interactions between downstream eVectors of cAMP and Rho kinase. At the receptor level, it will be important to determine whether MAG-mediated signaling events through gangliosides and NgRs have any overlap or interaction. Such research will help in designing strategies that encourage axons to regrow in the CNS. MAG Regulates Neuronal Cytoskeletal Proteins In the section titled ‘‘Altered Morphology of Axons in MAG Null Mice,’’ we noted that axons in MAG knockout mice have an altered morphology involving more densely packed neuroWlaments (also see Fig. 17.10). Compared to wild-type mice, MAG-deWcient mice older than 3 months of age have axonal proWles that are smaller with reduced spacing between neuroWlaments (Yin et al., 1998). The previously stated changes in axon caliber and cytoskeleton are likely a result of reduced neuroWlament phosphorylation, which occurs in MAG knockout mice. This raises the possibility that changes in myelin are secondary to direct eVects on axonal function. Similar alterations in axons and neuroWlaments also occur in Trembler mice, which have a mutation in PMP-22 (de Waegh et al., 1992). In the PMP-22 mutants, however, the alterations only occur in regions of sciatic nerve lacking myelin. MAG

451

452

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

knockouts have axonal defects even though myelin is present, suggesting that some other factor is responsible, such as altered signaling between glia and axons. NeuroWlaments consist of three subunits with low (NF-L), middle (NF-M), and high (NF-H) molecular weights, respectively. MAG-deWcient mice have less NF-L and NF-H in nerve, while total NF-M is unchanged (Yin et al., 1998). Phosphorylation of NF-M and especially NF-H was signiWcantly reduced (30 to 50% reduction depending on assay type). Phosphorylation of NF-H and NF-M increases the total negative charge and repositioning (lateral extension) of the neuroWlament sidearms; this in turn normally functions to increase neuroWlament spacing and axonal caliber. The mechanism of neuroWlament arrangement has been analyzed using experimental data from wild-type and MAG-deWcient mice and the most consistent model involves electrostatic repulsion between the phosphorylated sidearms (Kumar et al., 2002). Thus, under conditions of reduced NF-H and NF-M phosphorylation, the neuroWlament sidearms retract, leading to reduced neuroWlament spacing and smaller axonal calibers. In MAG-deWcient mice, the diameter of the axons and neuroWlament spacing are drastically reduced, particularly in paranodal regions. Whereas the paranodal region in wild-type mice continues to expand between P35 and P90, in MAG null mice the axon shrinks and this results in the formation of paranodal tomaculi. The schematic shown in Figure 17.10C illustrates the appearance of these abnormal paranodal myelin structures and also compares the structure of axons between wild-type and MAG null mice. Finally, it will be important to establish the signaling cascade that MAG initiates to control the neuroWlament subunit composition and phosphorylation. There is evidence that MAG interacts directly with microtubule-associated protein 1B (MAP1B) found on neurons. Although MAP1B is considered to be an intracelluar cytoskeletal component, it was proposed to also be a glycoprotein residing on the cell membrane (Muramoto et al., 1994). Recent analyses by Quarles and colleagues revealed that neurons express MAP1B, and that its carboxy terminus and glycosylation sites are located in the extracellular space, whereas the microtuble binding domains are in the neuronal cytoplasm ( Franzen et al., 2001; Tanner et al., 2000). Although MAP1B is expressed by glial cells, a higher molecular weight isoform is expressed on neurons but not Schwann cells. Coimmunoprecipitation experiments show that MAG speciWcally interacts with MAP1B, and immunohistochemistry reveals they colocalize at axon-glia junctions both in vitro and in vivo. Moreover, when DRG neurons are cocultured with MAG-expressing COS cells, there is a two-fold increase in expression of phosphorylated MAP1B, as well as increases in NF-H and NF-M, which are described in the next paragraph. These MAGmediated cytoskeletal phosphorylation events are important for control of axon caliber, as demonstrated by the axonal changes that occur in MAG nulls when such signaling is missing. Similarly, neuroWlaments in peripheral nerves of Trembler mutants have reduced phosphorylation and increased density, and interestingly there is less MAP1B phosphorylation (Kirkpatrick and Brady, 1994). Shiverer mice have a natural mutation, which yields a truncated form of MBP, and their CNS axon microtubule expression and density is increased (Kirkpatrick et al., 2001). The signaling mechanisms responsible for altering the axonal cytoskeleton in these mutants are not known, but potentially involve changes in MAG expression, which occur in knockouts (see Tab. 17.1). Perhaps most intriguing is the potential for MAG-MAP1B interactions to provide a structural link between axons and glia, and directly contribute to the modulation of myelinated axon structure and stability. The ability of MAG to regulate the axonal cytoskeleton has been demonstrated in cultured neurons and has also been extended to expose the underlying mechanisms. Compared to co-culturing of DRG neurons with control COS cells, co-culturing with MAG-expressing COS cells results in a Wve- and two-fold increase in NF-H and NF-M phosphorylation, respectively (Dashiell et al., 2002). MAG also causes the level of NF-M expression to increase by 50%. Treatment of PC12 neurons with soluble MAG-Fc chimera also eVectively increases NF-M expression and phosphorylation, thus demonstrating that MAG is responsible for the axonal changes, rather than other mechanisms such as cell-cell interactions. MAG also increases MAP1B, phosphorylated MAP1B, MAP2, lowmolecular-weight tau, and growth-associated protein-43 (GAP-43) levels.

SIGNALING EVENTS VIA MAG AND INTERACTING PROTEINS

Whole extracts of sciatic nerves from MAG null mice were also studied to determine whether the reduced neuroWlament phosphorylation in MAG null mice is due to altered kinase activity (Dashiell et al., 2002). Based on an assay that evaluates phosphorylation of preferred kinase substrates, MAG knockout mice had reduced activity of prolinedirected kinases, including cyclin-dependent kinase-5 (cdk5) and extracellular-signal regulated kinases 1 and 2 (ERK1/2). The speciWc activities of the cdk5 and ERK1/2 enzymes were subsequently determined to be reduced by 33 and 38%, respectively. Kinase assays also demonstrate that when these enzymes are immunoprecipitated from PC12 neurons treated with MAG-Fc, their activity is signiWcantly increased. In summary, MAG activates cdk5 and ERK1/2 kinases and this signaling process may be involved in MAG-mediated regulation of the axonal cytoskeleton. Further work will need to determine whether these or other kinases are obligatory, and whether MAG modulates phosphatase activities, transcription, and degradation of cytoskeletal proteins.

Glial Signaling via MAG

au7

au8

The cytoplasmic domains of MAG may be involved in modulating MAG function or signaling through phosphorylation, association with kinases, and also the cytoskeleton. Several interesting signaling events that may be of biological relevance to myelin formation occur within minutes of Ab-induced cross-linking of L-MAG. First, Fyn kinase activity increases, which leads to phosphorylation of L-MAG on Tyr620 (Jaramillo et al., 1994; Umemori et al., 1994). Consequently, this phosphorylation event allows MAG to associate with phospholipase Cg (PLCg). Nonreceptor tyrosine kinases such as Fyn and PLC associate via their SH2 and SH3 domains with speciWc sequences containing a phosphorylated Tyr in the cytoplasmic domain of surface receptors and CAMs. Thus, MAG may serve as a docking protein, which allows interactions between diVerent signaling molecules. Fyn activation by MAG may be of great importance because myelination is partially impaired in fyn-deWcient mice. MAG interaction with neuronal receptors could trigger events such as myelin formation. The regulation of phosphorylation events may be complex, and it cannot be excluded that other signaling events through separate receptors and signaling pathways can feed back onto MAG. We suggested previously that the phosphorylation state of the MAG cytoplasmic domain modulates adhesion and wrapping of the glial cell membrane around the axon during the myelination process (see the review by Attia et al., 1989). The strength of the MAG-neuron interaction could be regulated diVerentially between the leading glial lip and the previous myelin wrap by phosphorylation. In addition to controlling adhesion, the regulation of kinase and phosphatase activities may be responsible for initiating and terminating myelin wrapping. Besides interactions with kinases, MAG binds other cytoplasmic proteins. There is evidence that S100b, a 10 kDa calcium- and zinc-binding protein, can bind to the cytoplasmic domain of L-MAG (Kursula et al., 1999). A S100b synthetic peptide corresponding to the L-MAG interacting cytoplasmic domain (putative basic amphipathic alpha-helix) reproduced this interaction including the Ca2þ-dependency and further revealed a binding stoichiometry of 1:1 with a dissociation constant (KD) of 7 mM. This represents a fairly strong interaction that could also take place in vivo and partially contribute to MAGmediated signaling. Another interesting Wnding is that in the presence of S100b, phosphorylation of the L-MAG cytoplasmic domain by PKA is inhibited (Kursula et al., 2000). The functional consequences of S100b-mediated inhibition of phosphorylation is being investigated. However, S100b interacts with a variety of cytoskeleton-associated proteins, including actin microWlaments, microtubules, and intermediate Wlaments. Thus, MAGmediated control of glial cell morphology and function may be modiWed by interactions with S100b. Finally, there is some progress regarding identiWcation of glial cytoskeletal components that interact with MAG. A potential cytoskeleton binding site, which is similar in sequence to a functional site in the b1 subunit of integrin, is found in both isoforms (Salzer et al.,

453

454

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

1987). Co-localization of MAG and cytoskeletal components such as F-actin and spectrin is consistent with this prediction (Trapp et al., 1989b).

MAG IN INJURY AND DISEASE Human Gene Mutations in MAG? Based on studies of MAG knockout mice, it is probable that human diseases that aVect only MAG will remain elusive. Analyses of MAG single knockout mice have failed to show defects in myelin formation, and instead, it is the long-term maintenance of myelin that is aVected (Fruttiger et al., 1995a; Yin et al., 1998). No electrophysiological defects in peripheral nerve conduction occur in 4- to 8-week-old MAG-deWcient mice (Montag et al., 1994), and most behavioral motor and posture patterns are normal in the open Weld; however, the bar-cross apparatus that tests for Wne motor coordination revealed several abnormalities (Li et al., 1994). MAG nulls have decreased locomotor activity, frequency of exploratory sniYng and grooming, and a mild, transient, trunk tremor on the bars (see Tab. 17-1 for a summary). Similarly, another study found that compared to wild-type mice, only half of the MAG-deWcient mice were able to cross a horizontal bridge and there was an increased latency to reach a platform, whereas open Weld behavior was only mildly aVected (Uschkureit et al., 2000). Thus, there are no gross abnormalities in MAG mutants; however, their Wner motor coordination abilities are impaired. Aging mice do reveal neurological dysfunctions in that there is a signiWcant decrease in conduction velocity, an impairment that also occurs in human demyelinating neuropathies (Weiss et al., 2001). On the other hand, at least seven other action potential parameters were unchanged, again demonstrating that the defects are minor. Given that the neuropathology in MAG knockout mice is not severe, and together with the ability to only detect it easily later in life, this suggests that MAG mutations in humans would be diYcult to diagnose. As far as we are aware, there is no unequivocal evidence for mutations in MAG as contributing to human disease. Although there was some suggestion that genes inXuencing MS reside within or close to human chromosome 19q13, genetic linkage analyses indicate that MAG is not a leading candidate gene (D’Alfonso et al., 2002). Instead, the apolipoprotein E gene, which resides in the same chromosomal region, is more favored to play a role (Schmidt et al., 2002). Still, this locus confers only a part of the genetic susceptibility in MS. However, MAG mutations may play a role in other neuropathies, and further linkage studies will be necessary to determine whether the MAG gene is a candidate. It is also conceivable that alterations in other genes can aVect MAG transcription, such as in the abnormal QKI expression found in qk mice. Thus, future studies will need to consider mutations that are aVecting transcription or post-translational modiWcation of MAG.

Autoimmune Neuropathy Auto-Abs to MAG and also gangliosides are often associated with immune-mediated paraproteinemia and peripheral neuropathies (see reviews by Quarles and Weiss, 1999; Steck et al., 1999). When associated with MAG auto-Ab, the neuropathies involve neuronal or axonal degeneration or demyelination, the condition is chronic, and it aVects both sensory and motor nerves. There are many clinical studies describing the presence of anti-MAG Abs in neuropathies, although they cannot all be referenced here (for example, see Braun et al., 1982; Melmed et al., 1983). For neuropathies associated with monoclonal gammopathy, MAG is the most commonly targeted antigen. In a clinical study of 15 patients with a demyelinating sensory polyneuropathy, most had Abs to MAG (Van den Berg et al., 1996). Larger studies (40 to 75 patients) involving neuropathies associated with IgM gammopathy revealed 56 to 65% had anti-MAG monoclonal Abs (Chassande et al., 1998; Nobile-Orazio et al., 1994;). From other studies, it is

MAG IN INJURY AND DISEASE

known that the MAG auto-Abs commonly react with the HNK-1 epitope, although there are diVerences in speciWcities with regard to binding sulfate and carboxlic acid moieties, and there are also diVerences between patients in the Ab binding strength (Ilyas et al., 1984, 1990; Weiss et al., 1999; Yeung et al., 1991). In demyelinating neuropathies, the Ab titers to MAG appear to correlate with the degree of demyelination observed by histology and electrophysiology (Chassande et al., 1998; Gabriel et al., 1996). The serum anti-MAG titres correlate inversely with the amount of detectable MAG in nerve biopsies from the same patients, even though other myelin-related proteins such as P0, MBP, and periaxin are unaVected. MAG inXuences the structure and biochemistry of axons in patients that have demyelinating neuropathies with MAG auto-Abs present. NeuroWlament phosphorylation and spacing is reduced in MAG null mice (Yin et al., 1998), and similar changes occur in human patients with MAG auto-Ab-related demyelinating neuropathies (Lunn et al., 2002). Electron microscopy reveals that neuroWlaments from diseased human sural nerve axons are more closely spaced in those patients that have anti-MAG autoAbs. Axons from normal subjects and also demyelinated axons from patients with other neuropathies are not aVected, demonstrating that the eVect is speciWc to patients with MAG auto-Abs. There is a nonsigniWcant trend for diseased axons to be smaller in diameter. Axonal constriction may account for the reduced conduction velocity observed in aged MAG null mice. A loss of signaling between Schwann cells and axons followed by decreased neuroWlament phosphorylation is postulated to be responsible for the axonal changes in MAG null mice and in patients with MAG autoAbs. Data support the notion that the auto-Abs are indeed targeting MAG on glial cells. First there is biochemical evidence based on western blotting and also by the ability of auto-Abs to bind MAG peptides. Furthermore, immunocytochemical data reveal that MAG auto-Abs from patient serum binds to myelinating Schwann cells (Takatsu et al., 1985). Prior absorption of serum with MAG abolishes staining. Anti-MAG IgM deposits occur at sites of uncompacted myelin including Schmidt Lanterman incisures and paranodal loops (Gabriel et al., 1998; Steck et al., 1999). Furthermore, Abs bind to the Schwann cell basal lamina of myelinated (but not unmyelinated) Wbers and are co-localized with collagen IV. In summary, auto-Abs to MAG are associated with demyelination, their abundance is related to the severity of the neuropathy, and they bind to myelinating Schwann cells. There is also an indication that in several diseases of the nervous system, anti-ganglioside Abs are involved (see the review by Takamiya et al., 1996). A subset of neuropathies are associated with auto-Abs to GD1a and GT1b, which bind MAG. Patients with demyelinating neuropathy have circulating Abs against GD1a, GT1b, and GM3, the latter of which also binds MAG although with reduced potency ( Mizutani et al., 2001; Vabnick et al., 1997). In a study of 195 patients having an assortment of motor syndromes, high anti-GD1a Ab titers were detected in two patients with IgM monoclonal gammopathy and two patients with Guillain-Barre´ syndrome (Carpo et al., 1996). The latter disease is an acute demyelinating inXammatory polyneuropathy often associated with auto-Abs to gangliosides (including GD1a), and evidence is accumulating to suggest an immune response to an environmental immunogen is involved (see reviews by Quarles and Weiss, 1999; Nachamkin et al., 2002; Schwerer, 2002). In this case, molecular mimicry occurs from Campylobacter jejuni bacterial infection, which presents an antigen similar to human gangliosides. Gangliosides and MAG auto-Abs may also play a role in a variety of other diseases, including MS, optic neuritis, and myasthenia gravis (for example, see Link et al., 1992; Mo¨ller et al., 1987; Sato et al., 1986; Wajgt and Gorny, 1983). Among 25 MS patients, 15 had anti-MAG Abs (12 IgG, 3 IgM) in their CNS Xuid, compared to only 1 out 27 control patients (Baig et al., 1991). A study of 32 MS and 20 optic neuritis patients found 23% and 18%, respectively, had anti-GD1a Abs (Mata et al., 1999). Thus, there is clinical evidence that gangliosides which bind to MAG are involved in neurological disease.

455

456

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

While auto-Abs can induce demyelination, the reasons for their production in neuropathies is unclear. There is suYcient evidence that anti-MAG and anti-GD1a Abs can induce axonal demyelination and degeneration. Injection of cats, rabbits, or chickens with anti-MAG Abs causes demyelination (Hays et al., 1987; Tatum, 1993; Trojaborg et al., 1989; Willison et al., 1988). The ultrastructural pathology in these animal experiments is very similar to that found in human neuropathy, including the widening of myelin lamellae. Mice exposed to anti-GD1a Abs develop axonal degeneration (see the report by Willison et al., 2002). Regarding neuropathologies, the trigger for auto-Ab production by T cells could be antigens that enter the circulation from damaged myelin. Damage to the myelin sheath releases the soluble proteolytic fragment of MAG, dMAG, which inhibits axonal regeneration (Tang et al., 1997b, 2001) and has been suggested to contribute to neurological disease (Johnson et al., 1986). A study of Wve MS patients found that the dMAG form was more abundant than in control subjects and seemed to account for the preferential loss of MAG at MS plaques (Mo¨ller et al., 1987). Another study found that lymphocytes in patients with MS and polyneuropathy responded to MAG and to Wve synthetic MAG peptides to a greater extent relative to healthy subjects (Andersson et al., 2002). However, since none of the peptides appear to be immunodominant, the authors conclude that MAG auto-Abs arise through secondary responses associated with myelin breakdown. However, they did not exclude a possibility that these Abs play a role in the initiation or progression of the diseases. Since there is no evidence for molecular mimicry with a MAG-like epitope, an immune response to endogenous myelin does appear to be a simple reason why MAG is targeted so frequently in a variety of neurological disorders, particularly those involving demyelination. Finally, it is not known whether the pathogenic eVects of MAG Abs are moderated by disruption of normal MAG function or also by cell death through complement-mediated lysis.

SUMMARY AND CLOSING REMARKS MAG is a glycoprotein localized in the periaxonal membranes of glia that form the specialized wrappings known as myelin. Evidence supports multiple functional roles for MAG, including the formation and maintenance of myelin. MAG also aVects axonal viability and cytoskeletal organization, and recent studies show that MAG inhibits axonal sprouting and regeneration. Regarding this latter role, a wealth of information recently became available including the remarkable discovery that MAG, Nogo, and OMgp all bind to NgR. This Wnding explains why removal of only some of the myelin-derived inhibitory molecules is not suYcient to block inhibition mediated by myelin from the adult CNS. Furthermore, the absence of NgR during early development appears to explain why these myelin-derived molecules do not inhibit young neurons. The identiWcation of NgR, together with additional ganglioside receptors for MAG, now makes it possible to focus on and further characterize MAG-mediated molecular interactions and signals. Inhibition of neurite outgrowth by way of MAG/NgR signaling may either overlap or be distinct from ganglioside/p75NTR-mediated intracellular signaling events involving Rho kinase. It is now possible to determine whether these molecular events also apply to processes involving the genesis and stability of myelin. Studies of transgenic mice have revealed important details regarding MAG function. For instance, the L- and S-MAG isoforms have diVerent functions, the former being crucial for formation of CNS myelin and the latter contributing to maintenance of PNS myelin. Mice lacking MAG entirely have mild defects in the formation and maintenance of myelin. Myelination is delayed, superXuous myelin is formed, there is increased degeneration of myelin and axons, and both the ultrastructure and biochemistry of axons are altered. The defects are not prominent, and superWcially it may seem that MAG does not have an essential role. However, there is signiWcant molecular redundancy in the nervous system, and it is diYcult to appreciate the full contribution of endogenous MAG function.

SUMMARY AND CLOSING REMARKS

For instance, the role of MAG in the spiraling process of myelination only becomes apparent in knockout mice lacking both MAG and P0. Behavioral and electrophysiological changes in MAG-deWcient mice present a phenotype similar to that found in human neuropathies and demyelinating diseases. MAG and its ganglioside receptors are a frequent target of auto-Abs in immune-mediated neuropathies. Mice exposed to anti-MAG Abs develop axonal degeneration and demyelination, and the pathology is similar in human patients with MAG auto-Abs. However, there is no human disease that has been mapped to the MAG gene, possibly because the phenotype is not severe in younger individuals and thus is diYcult to detect.

References Afar, D. E., Salzer, J. L., Roder, J., Braun, P. E., and Bell, J. C. (1990). DiVerential phosphorylation of myelinassociated glycoprotein isoforms in cell culture. J Neurochem 55, 1418–1426. Afar, D. E., Marius, R. M., Salzer, J. L., Stanners, C. P., Braun, P. E., and Bell, J. C. (1991). Cell adhesion properties of myelin-associated glycoprotein in L cell Wbroblasts. J Neurosci Res 29, 429–436. Agrawal, H. C., Noronha, A. B., Agrawal, D., and Quarles, R. H. (1990). The myelin-associated glycoprotein is phosphorylated in the peripheral nervous system. Biochem Biophys Res Commun 169, 953–958. Almazan, G., Tropak, M., and Roder, J. (1992). Myelin associated glycoproteins confer heterophilic adhesion properties to an immortalized optic nerve derived cell line. Cell Struct Funct 17, 407–415. Amzel, L. M., and Poljak, R. J. (1979). Three-dimensional structure of immunoglobulins. Annu Rev Biochem 48, 961–997. Andersson, M., Yu, M., So¨derstro¨m, M., Weerth, S., Baig, S., Solders, G., and Link, H. (2002). Multiple MAG peptides are recognized by circulating T and B lymphocytes in polyneuropathy and multiple sclerosis. Eur J Neurol 9, 243–251. Angata, T., Kerr, S. C., Greaves, D. R., Varki, N. M., Crocker, P. R., and Varki, A. (2002). Cloning and characterization of human Siglec-11. A recently evolved signaling molecule that can interact with SHP-1 and SHP-2 and is expressed by tissue macrophages, including brain microglia. J Biol Chem 277, 24466–24474. Arquint, M., Roder, J., Chia, L. S., Down, J., Wilkinson, D., Bayley, H., Braun, P., and Dunn, R. (1987). Molecular cloning and primary structure of myelin-associated glycoprotein. Proc Natl Acad Sci USA 84, 600–604. Attia, J. (1992). Myelin-associated glycoprotein: Production and characterization of a recombinant, soluble form and search for a neuronal ligand. Ph. D. thesis, University of Toronto. 1–179. Attia, J., Hicks, L., Oikawa, K., Kay, C. M., and Dunn, R. J. (1993). Structural properties of the myelinassociated glycoprotein ectodomain. J Neurochem 61, 718–726. Attia, J., Tropak, M., Johnson, P. W., Newerly-Abranow, W., Pawson, T., Roder, J. C., and Dunn, R. J. (1989). Modulated adhesion: A proposal for the role of myelin-associated glycoprotein in myelin wrapping. Clin Chem 35, 717–720. Bachmann, M., Conscience, J. F., Probstmeier, R., Carbonetto, S., and Schachner, M. (1995). Recognition molecules myelin-associated glycoprotein and tenascin-C inhibit integrin-mediated adhesion of neural cells to collagen. J Neurosci Res 40, 458–470. Baig, S., Olsson, T., Yu-Ping, J., Ho¨jeberg, B., Cruz, M., and Link, H. (1991). Multiple sclerosis: Cells secreting antibodies against myelin-associated glycoprotein are present in cerebrospinal Xuid. Scand J Immunol 33, 73–79. Bajt, M. L., Schmitz, B., Schachner, M., and Zipser, B. (1990). Carbohydrate epitopes involved in neural cell recognition are conserved between vertebrates and leech. J Neurosci Res 27, 276–285. Bambrick, L. L., and Braun, P. E. (1991). Phosphorylation of myelin-associated glycoprotein in cultured oligodendrocytes. Dev Neurosci 13, 412–416. Barnes, Y. C., Skelton, T. P., Stamenkovic, I., and Sgroi, D. C. (1999). Sialylation of the sialic acid binding lectin sialoadhesin regulates its ability to mediate cell adhesion. Blood 93, 1245–1252. Barton, D. E., Arquint, M., Roder, J., Dunn, R., and Francke, U. (1987). The myelin-associated glycoprotein gene: Mapping to human chromosome 19 and mouse chromosome 7 and expression in quivering mice. Genomics 1, 107–112. Bartoszewicz, Z. P., Noronha, A. B., Fujita, N., Sato, S., Bo, L., Trapp, B. D., and Quarles, R. H. (1995). Abnormal expression and glycosylation of the large and small isoforms of myelin-associated glycoprotein in dysmyelinating quaking mutants. J Neurosci Res 41, 27–38. Bartsch, S., Montag, D., Schachner, M., and Bartsch, U. (1997). Increased number of unmyelinated axons in optic nerves of adult mice deWcient in the myelin-associated glycoprotein (MAG). Brain Res 762, 231–234. Bartsch, U., Bandtlow, C. E., Schnell, L., Bartsch, S., Spillmann, A. A., Rubin, B. P., Hillenbrand, R., Montag, D., Schwab, M. E., and Schachner, M. (1995a). Lack of evidence that myelin-associated glycoprotein is a major inhibitor of axonal regeneration in the CNS. Neuron 15, 1375–1381. Bartsch, U., KirchhoV, F., and Schachner, M. (1989). Immunohistological localization of the adhesion molecules L1, N-CAM, and MAG in the developing and adult optic nerve of mice. J Comp Neurol 284, 451–462.

457

458

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

Bartsch, U., Montag, D., Bartsch, S., and Schachner, M. (1995b). Multiply myelinated axons in the optic nerve of mice deWcient for the myelin-associated glycoprotein. Glia 14, 115–122. BiYger, K., Bartsch, S., Montag, D., Aguzzi, A., Schachner, M., and Bartsch, U. (2000). Severe hypomyelination of the murine CNS in the absence of myelin-associated glycoprotein and Fyn tyrosine kinase. J Neurosci 20, 7430–7437. Bo, L., Quarles, R. H., Fujita, N., Bartoszewicz, Z., Sato, S., and Trapp, B. D. (1995). Endocytic depletion of L-MAG from CNS myelin in quaking mice. J Cell Biol 131, 1811–1820. Bollensen, E., Steck, A. J., and Schachner, M. (1988). Reactivity with the peripheral myelin glycoprotein P0 in serum from patients with monoclonal IgM gammopathy and polyneuropathy. Neurology 38, 1266–1270. Bosio, A., Binczek, E., and StoVel, W. (1996). Functional breakdown of the lipid bilayer of the myelin membrane in central and peripheral nervous system by disrupted galactocerebroside synthesis. Proc Natl Acad Sci USA 93, 13280–13285. Bradel, E. J., and Prince, F. P. (1983). Cultured neonatal rat oligodendrocytes elaborate myelin membrane in the absence of neurons. J Neurosci Res 9, 381–392. Braun, P. E., Frail, D. E., and Latov, N. (1982). Myelin-associated glycoprotein is the antigen for a monoclonal IgM in polyneuropathy. J Neurochem 39, 1261–1265. Brown, M. C., Lunn, E. R., and Perry, V. H. (1992). Consequences of slow Wallerian degeneration for regenerating motor and sensory axons. J Neurobiol 23, 521–536. Brown, M. C., Perry, V. H., Lunn, E. R., Gordon, S., and Heumann, R. (1991). Macrophage dependence of peripheral sensory nerve regeneration: Possible involvement of nerve growth factor. Neuron 6, 359–370. Burger, D., Perruisseau, G., Simon, M., and Steck, A. J. (1992). Comparison of the N-linked oligosaccharide structures of the two major human myelin glycoproteins MAG and P0: Assessment of the structures bearing the epitope for HNK-1 and human monoclonal immunoglobulin M found in demyelinating neuropathy. J Neurochem 58, 854–861. Burger, D., Perruisseau, G., and Steck, A. J. (1991). Anti-myelin-associated glycoprotein antibodies in patients with a monoclonal IgM gammopathy and polyneuropathy, and a simpliWed method for the preparation of glycolipid antigens. J Immunol Methods 140, 31–36. Butt, A. M., Ibrahim, M., Gregson, N., and Berry, M. (1998). DiVerential expression of the L- and S-isoforms of myelin associated glycoprotein (MAG) in oligodendrocyte unit phenotypes in the adult rat anterior medullary velum. J Neurocytol 27, 271–280. Cai, D., Deng, K., Mellado, W., Lee, J., Ratan, R. R., and Filbin, M. T. (2002). Arginase I and Polyamines Act Downstream from Cyclic AMP in Overcoming Inhibition of Axonal Growth MAG and Myelin in Vitro. Neuron 35, 711–719. Cai, D., Qiu, J., Cao, Z., McAtee, M., Bregman, B. S., and Filbin, M. T. (2001). Neuronal cyclic AMP controls the developmental loss in ability of axons to regenerate. J Neurosci 21, 4731–4739. Cai, D., Shen, Y., De Bellard, M., Tang, S., and Filbin, M. T. (1999). Prior exposure to neurotrophins blocks inhibition of axonal regeneration by MAG and myelin via a cAMP-dependent mechanism. Neuron 22, 89–101. Cameron-Curry, P., and Le Douarin, N. M. (1995). Oligodendrocyte precursors originate from both the dorsal and the ventral parts of the spinal cord. Neuron 15, 1299–1310. Carenini, S., Montag, D., Cremer, H., Schachner, M., and Martini, R. (1997). Absence of the myelin-associated glycoprotein (MAG) and the neural cell adhesion molecule (N-CAM) interferes with the maintenance, but not with the formation of peripheral myelin. Cell Tissue Res 287, 3–9. Carenini, S., Montag, D., Schachner, M., and Martini, R. (1999). Subtle roles of neural cell adhesion molecule and myelin-associated glycoprotein during Schwann cell spiralling in P0-deWcient mice. Glia 27, 203–212. Caroni, P., and Schwab, M. E. (1988a). Antibody against myelin-associated inhibitor of neurite growth neutralizes nonpermissive substrate properties of CNS white matter. Neuron 1, 85–96. Caroni, P., and Schwab, M. E. (1988b). Two membrane protein fractions from rat central myelin with inhibitory properties for neurite growth and Wbroblast spreading. J Cell Biol 106, 1281–1288. Carpo, M., Nobile-Orazio, E., Meucci, N., Gamba, M., Barbieri, S., Allaria, S., and Scarlato, G. (1996). AntiGD1a ganglioside antibodies in peripheral motor syndromes. Ann Neurol 39, 539–543. Chassande, B., Le´ger, J. M., Younes-ChennouW, A. B., Bengoufa, D., Maisonobe, T., Bouche, P., and Baumann, N. (1998). Peripheral neuropathy associated with IgM monoclonal gammopathy: Correlations between M-protein antibody activity and clinical/electrophysiological features in 40 cases. Muscle Nerve 21, 55–62. Chen, M. S., Huber, A. B., van der Haar, M. E., Frank, M., Schnell, L., Spillmann, A. A., Christ, F., and Schwab, M. E. (2000). Nogo-A is a myelin-associated neurite outgrowth inhibitor and an antigen for monoclonal antibody IN-1. Nature 403, 434–439. Chou, D. K., Ilyas, A. A., Evans, J. E., Costello, C., Quarles, R. H., and Jungalwala, F. B. (1986). Structure of sulfated glucuronyl glycolipids in the nervous system reacting with HNK-1 antibody and some IgM paraproteins in neuropathy. J Biol Chem 261, 11717–11725. Coetzee, T., Fujita, N., Dupree, J., Shi, R., Blight, A., Suzuki, K., and Popko, B. (1996). Myelination in the absence of galactocerebroside and sulfatide: Normal structure with abnormal function and regional instability. Cell 86, 209–219. Collins, B. E., Ito, H., Sawada, N., Ishida, H., Kiso, M., and Schnaar, R. L. (1999). Enhanced binding of the neural siglecs, myelin-associated glycoprotein and Schwann cell myelin protein, to Chol-1 (-series) gangliosides and novel sulfated Chol-1 analogs. J Biol Chem 274, 37637–37643.

SUMMARY AND CLOSING REMARKS

Collins, B. E., Kiso, M., Hasegawa, A., Tropak, M. B., Roder, J. C., Crocker, P. R., and Schnaar, R. L. (1997a). Binding speciWcities of the sialoadhesin family of I-type lectins. Sialic acid linkage and substructure requirements for binding of myelin-associated glycoprotein, Schwann cell myelin protein, and sialoadhesin. J Biol Chem 272, 16889–16895. Collins, B. E., Yang, L. J., Mukhopadhyay, G., Filbin, M. T., Kiso, M., Hasegawa, A., and Schnaar, R. L. (1997b). Sialic acid speciWcity of myelin-associated glycoprotein binding. J Biol Chem 272, 1248–1255. CorWeld, A. P., Wember, M., Schauer, R., and Rott, R. (1982). The speciWcity of viral sialidases. The use of oligosaccharide substrates to probe enzymic characteristics and strain-speciWc diVerences. Eur J Biochem 124, 521–525. Crocker, P. R., Clark, E. A., Filbin, M., Gordon, S., Jones, Y., Kehrl, J. H., Kelm, S., Le Douarin, N., Powell, L., Roder, J., Schnaar, R. L., Sgroi, D. C., Stamenkovic, K., Schauer, R., Schachner, M., van den Berg, T. K., van der Merwe, P. A., Watt, S. M., and Varki, A. (1998). Siglecs: A family of sialic-acid binding lectins. Glycobiology 8, v–vi. Crocker, P. R., Freeman, S., Gordon, S., and Kelm, S. (1995). Sialoadhesin binds preferentially to cells of the granulocytic lineage. J Clin Invest 95, 635–643. Crocker, P. R., Kelm, S., Dubois, C., Martin, B., McWilliam, A. S., Shotton, D. M., Paulson, J. C., and Gordon, S. (1991). PuriWcation and properties of sialoadhesin, a sialic acid-binding receptor of murine tissue macrophages. Embo J 10, 1661–1669. Crocker, P. R., Kelm, S., Hartnell, A., Freeman, S., Nath, D., Vinson, M., and Mucklow, S. (1996). Sialoadhesin and related cellular recognition molecules of the immunoglobulin superfamily. Biochem Soc Trans 24, 150–156. Crocker, P. R., Mucklow, S., Bouckson, V., McWilliam, A., Willis, A. C., Gordon, S., Milon, G., Kelm, S., and BradWeld, P. (1994). Sialoadhesin, a macrophage sialic acid binding receptor for haemopoietic cells with 17 immunoglobulin-like domains. Embo J 13, 4490–4503. Cyster, J. G., and Goodnow, C. C. (1997). Tuning antigen receptor signaling by CD22: Integrating cues from antigens and the microenvironment. Immunity 6, 509–517. D’Alfonso, S., Mellai, M., Giordano, M., Pastore, A., Malferrari, G., Naldi, P., Repice, A., Liguori, M., Cannoni, S., Milanese, C., Caputo, D., Savettieri, G., and Momigliano-Richiardi, P. (2002). IdentiWcation of single nucleotide variations in the coding and regulatory regions of the myelin-associated glycoprotein gene and study of their association with multiple sclerosis. J Neuroimmunol 126, 196–204. Dashiell, S. M., Tanner, S. L., Pant, H. C., and Quarles, R. H. (2002). Myelin-associated glycoprotein modulates expression and phosphorylation of neuronal cytoskeletal elements and their associated kinases. J Neurochem 81, 1263–1272. Datta, A. K., and Paulson, J. C. (1995). The sialyltransferase ‘‘sialylmotif’’ participates in binding the donor substrate CMP-NeuAc. J Biol Chem 270, 1497–1500. David, S., Braun, P. E., Jackson, D. L., Kottis, V., and McKerracher, L. (1995). Laminin overrides the inhibitory eVects of peripheral nervous system and central nervous system myelin-derived inhibitors of neurite growth. J Neurosci Res 42, 594–602. De Bellard, M. E., and Filbin, M. T. (1999). Myelin-associated glycoprotein, MAG, selectively binds several neuronal proteins. J Neurosci Res 56, 213–218. DeBellard, M. E., Tang, S., Mukhopadhyay, G., Shen, Y. J., and Filbin, M. T. (1996). Myelin-associated glycoprotein inhibits axonal regeneration from a variety of neurons via interaction with a sialoglycoprotein. Mol Cell Neurosci 7, 89–101. de Waegh, S. M., Lee, V. M., and Brady, S. T. (1992). Local modulation of neuroWlament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68, 451–463. Dergham, P., Ellezam, B., Essagian, C., Avedissian, H., Lubell, W. D., and McKerracher, L. (2002). Rho signaling pathway targeted to promote spinal cord repair. J Neurosci 22, 6570–6577. D’Eustachio, P., Colman, D. R., and Salzer, J. L. (1988). Chromosomal location of the mouse gene that encodes the myelin-associated glycoproteins. J Neurochem 50, 589–593. Domeniconi, M., Cao, Z., Spencer, T., Sivasankaran, R., Wang, K., Nikulina, E., Kimura, N., Cai, H., Deng, K., Gao, Y., He, Z., and Filbin, M. T. (2002). Myelin-associated glycoprotein interacts with the nogo66 receptor to inhibit neurite outgrowth. Neuron 35, 283–290. Dorken, B., Moldenhauer, G., Pezzutto, A., Schwartz, R., Feller, A., Kiesel, S., and Nadler, L. M. (1986). HD39 (B3), a B lineage-restricted antigen whose cell surface expression is limited to resting and activated human B lymphocytes. J Immunol 136, 4470–4479. Dubois-Dalcq, M., Behar, T., Hudson, L., and Lazzarini, R. A. (1986). Emergence of three myelin proteins in oligodendrocytes cultured without neurons. J Cell Biol 102, 384–392. Dulac, C., Cameron-Curry, P., Ziller, C., and Le Douarin, N. M. (1988). A surface protein expressed by avian myelinating and nonmyelinating Schwann cells but not by satellite or enteric glial cells. Neuron 1, 211–220. Dulac, C., Tropak, M. B., Cameron-Curry, P., Rossier, J., Marshak, D. R., Roder, J., and Le Douarin, N. M. (1992). Molecular characterization of the Schwann cell myelin protein, SMP: Structural similarities within the immunoglobulin superfamily. Neuron 8, 323–334. Dupree, J. L., Coetzee, T., Blight, A., Suzuki, K., and Popko, B. (1998a). Myelin galactolipids are essential for proper node of Ranvier formation in the CNS. J Neurosci 18, 1642–1649. Dupree, J. L., Coetzee, T., Suzuki, K., and Popko, B. (1998b). Myelin abnormalities in mice deWcient in galactocerebroside and sulfatide. J Neurocytol 27, 649–659.

459

460

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

Ebersole, T. A., Chen, Q., Justice, M. J., and Artzt, K. (1996). The quaking gene product necessary in embryogenesis and myelination combines features of RNA binding and signal transduction proteins. Nat Genet 12, 260–265. Edwards, A. M., Arquint, M., Braun, P. E., Roder, J. C., Dunn, R. J., Pawson, T., and Bell, J. C. (1988). Myelinassociated glycoprotein, a cell adhesion molecule of oligodendrocytes, is phosphorylated in brain. Mol Cell Biol 8, 2655–2658. Edwards, A. M., Braun, P. E., and Bell, J. C. (1989). Phosphorylation of myelin-associated glycoprotein in vivo and in vitro occurs only in the cytoplasmic domain of the large isoform. J Neurochem 52, 317–320. Engel, P., Nojima, Y., Rothstein, D., Zhou, L. J., Wilson, G. L., Kehrl, J. H., and Tedder, T. F. (1993). The same epitope on CD22 of B lymphocytes mediates the adhesion of erythrocytes, T and B lymphocytes, neutrophils, and monocytes. J Immunol 150, 4719–4732. Fahrig, T., Probstmeier, R., Spiess, E., Meyer-Franke, A., KirchhoV, F., Drescher, B., Schachner, M. (1993). Functional topography of the myelin-associated glycoprotein. I. mapping of domains by electron microscopy. Eur J Neurosci 5, 1118–1126. Fischer, C., Schroth-Diez, B., Herrmann, A., Garten, W., and Klenk, H. D. (1998). Acylation of the inXuenza hemagglutinin modulates fusion activity. Virology 248, 284–294. Fournier, A. E., GrandPre, T., and Strittmatter, S. M. (2001). IdentiWcation of a receptor mediating Nogo-66 inhibition of axonal regeneration. Nature 409, 341–346. Frail, D. E., and Braun, P. E. (1984). Two developmentally regulated messenger RNAs diVering in their coding region may exist for the myelin-associated glycoprotein. J Biol Chem 259, 14857–14862. Frail, D. E., and Braun, P. E. (1985). Abnormal expression of the myelin-associated glycoprotein in the central nervous system of dysmyelinating mutant mice. J Neurochem 45, 1071–1075. Frail, D. E., Webster, H. D., and Braun, P. E. (1985). Developmental expression of the myelin-associated glycoprotein in the peripheral nervous system is diVerent from that in the central nervous system. J Neurochem 45, 1308–1310. Franzen, R., Tanner, S. L., Dashiell, S. M., Rottkamp, C. A., Hammer, J. A., and Quarles, R. H. (2001). Microtubule-associated protein 1B: A neuronal binding partner for myelin-associated glycoprotein. J Cell Biol 155, 893–898. Freeman, S. D., Kelm, S., Barber, E. K., and Crocker, P. R. (1995). Characterization of CD33 as a new member of the sialoadhesin family of cellular interaction molecules. Blood 85, 2005–2012. Freigang, J., Proba, K., Leder, L., Diederichs, K., Sonderegger, P., and Welte, W. (2000). The crystal structure of the ligand binding module of axonin-1/TAG-1 suggests a zipper mechanism for neural cell adhesion. Cell 101, 425–433. Fruttiger, M., Montag, D., Schachner, M., and Martini, R. (1995a). Crucial role for the myelin-associated glycoprotein in the maintenance of axon-myelin integrity. Eur J Neurosci 7, 511–515. Fruttiger, M., Schachner, M., and Martini, R. (1995b). Tenascin-C expression during wallerian degeneration in C57BL/Wlds mice: Possible implications for axonal regeneration. J Neurocytol 24, 1–14. Fujita, N., Kemper, A., Dupree, J., Nakayasu, H., Bartsch, U., Schachner, M., Maeda, N., Suzuki, K., and Popko, B. (1998). The cytoplasmic domain of the large myelin-associated glycoprotein isoform is needed for proper CNS but not peripheral nervous system myelination. J Neurosci 18, 1970–1978. Fujita, N., Sato, S., Ishiguro, H., Inuzuka, T., Baba, H., Kurihara, T., Takahashi, Y., and Miyatake, T. (1990). The large isoform of myelin-associated glycoprotein is scarcely expressed in the quaking mouse brain. J Neurochem 55, 1056–1059. Fujita, N., Sato, S., Kurihara, T., Inuzuka, T., Takahashi, Y., and Miyatake, T. (1988). Developmentally regulated alternative splicing of brain myelin-associated glycoprotein mRNA is lacking in the quaking mouse. FEBS Lett 232, 323–327. Fujita, N., Sato, S., Kurihara, T., Kuwano, R., Sakimura, K., Inuzuka, T., Takahashi, Y., and Miyatake, T. (1989). cDNA cloning of mouse myelin-associated glycoprotein: A novel alternative splicing pattern. Biochem Biophys Res Commun 165, 1162–1169. Gabriel, J. M., Erne, B., Miescher, G. C., Miller, S. L., Vital, A., Vital, C., and Steck, A. J. (1996). Selective loss of myelin-associated glycoprotein from myelin correlates with anti-MAG antibody titre in demyelinating paraproteinaemic polyneuropathy. Brain 119, 775–787. Gabriel, J. M., Erne, B., Bernasconi, L., Tosi, C., Probst, A., Landmann, L., and Steck, A. J. (1998). Confocal microscopic localization of anti-myelin-associated glycoprotein autoantibodies in a patient with peripheral neuropathy initially lacking a detectable IgM gammopathy. Acta Neuropathol (Berl) 95, 540–546. Georgiou, J., and Charlton, M. P. (1999). Non-myelin-forming perisynaptic schwann cells express protein zero and myelin-associated glycoprotein. Glia 27, 101–109. Giese, K. P., Martini, R., Lemke, G., Soriano, P., and Schachner, M. (1992). Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71, 565–576. GrandPre, T., Li, S., and Strittmatter, S. M. (2002). Nogo-66 receptor antagonist peptide promotes axonal regeneration. Nature 417, 547–551. GrandPre, T., Nakamura, F., Vartanian, T., and Strittmatter, S. M. (2000). IdentiWcation of the Nogo inhibitor of axon regeneration as a Reticulon protein. Nature 403, 439–444.

SUMMARY AND CLOSING REMARKS

GriYth, L. S., Schmitz, B., and Schachner, M. (1992). L2/HNK-1 carbohydrate and protein-protein interactions mediate the homophilic binding of the neural adhesion molecule P0. J Neurosci Res 33, 639–648. Grubinska, B., Laszkiewicz, I., Royland, J., Wiggins, R. C., and Konat, G. W. (1994). DiVerentiation-speciWc demethylation of myelin associated glycoprotein gene in cultured oligodendrocytes. J Neurosci Res 39, 233–242. Hall, H., Liu, L., Schachner, M., and Schmitz, B. (1993). The L2/HNK-1 carbohydrate mediates adhesion of neural cells to laminin. Eur J Neurosci 5, 34–42. Haney, C. A., Sahenk, Z., Li, C., Lemmon, V. P., Roder, J., and Trapp, B. D. (1999). Heterophilic binding of L1 on unmyelinated sensory axons mediates Schwann cell adhesion and is required for axonal survival. J Cell Biol 146, 1173–1184. Hardy, R. J. (1998). Molecular defects in the dysmyelinating mutant quaking. J Neurosci Res 51, 417–422. Hardy, R. J., Loushin, C. L., Friedrich, V. L., Jr., Chen, Q., Ebersole, T. A., Lazzarini, R. A., and Artzt, K. (1996). Neural cell type-speciWc expression of QKI proteins is altered in quakingviable mutant mice. J Neurosci 16, 7941–7949. Harpaz, Y., and Chothia, C. (1994). Many of the immunoglobulin superfamily domains in cell adhesion molecules and surface receptors belong to a new structural set which is close to that containing variable domains. J Mol Biol 238, 528–539. Hays, A. P., Latov, N., Takatsu, M., and Sherman, W. H. (1987). Experimental demyelination of nerve induced by serum of patients with neuropathy and an anti-MAG IgM M-protein. Neurology 37, 242–256. Ilyas, A. A., Chou, D. K., Jungalwala, F. B., Costello, C., and Quarles, R. H. (1990). Variability in the structural requirements for binding of human monoclonal anti-myelin-associated glycoprotein immunoglobulin M antibodies and HNK-1 to sphingoglycolipid antigens. J Neurochem 55, 594–601. Ilyas, A. A., Quarles, R. H., MacIntosh, T. D., Dobersen, M. J., Trapp, B. D., Dalakas, M. C., and Brady, R. O. (1984). IgM in a human neuropathy related to paraproteinemia binds to a carbohydrate determinant in the myelin-associated glycoprotein and to a ganglioside. Proc Natl Acad Sci USA 81, 1225–1229. Inuzuka, T., Fujita, N., Sato, S., Baba, H., Nakano, R., Ishiguro, H., and Miyatake, T. (1991). Expression of the large myelin-associated glycoprotein isoform during the development in the mouse peripheral nervous system. Brain Res 562, 173–175. Jaramillo, M. L., Afar, D. E., Almazan, G., and Bell, J. C. (1994). IdentiWcation of tyrosine 620 as the major phosphorylation site of myelin-associated glycoprotein and its implication in interacting with signaling molecules. J Biol Chem 269, 27240–27245. Jensen, N. A., West, M. J., and Celis, J. E. (1999). Oligodendrocyte programmed cell death and central myelination deWciency induced in transgenic mice by synergism between c-Myc and Oct-6. J Biol Chem 274, 29921–29926. Jessen, K. R., and Mirsky, R. (2002). Signals that determine Schwann cell identity. J Anat 200, 367–376. Johnson, D., Sato, S., Quarles, R. H., Inuzuka, T., Brady, R. O., and Tourtellotte, W. W. (1986). Quantitation of the myelin-associated glycoprotein in human nervous tissue from controls and multiple sclerosis patients. J Neurochem 46, 1086–1093. Johnson, P. W., Abramow-Newerly, W., Seilheimer, B., Sadoul, R., Tropak, M. B., Arquint, M., Dunn, R. J., Schachner, M., and Roder, J. C. (1989). Recombinant myelin-associated glycoprotein confers neural adhesion and neurite outgrowth function. Neuron 3, 377–385. Jones, E. Y., Davis, S. J., Williams, A. F., Harlos, K., and Stuart, D. I. (1992). Crystal structure at 2.8 A resolution of a soluble form of the cell adhesion molecule CD2. Nature 360, 232–239. Kelm, S. (2001). Ligands for siglecs. Results Probl Cell DiVer 33, 153–176. Kelm, S., Pelz, A., Schauer, R., Filbin, M. T., Tang, S., de Bellard, M. E., Schnaar, R. L., Mahoney, J. A., Hartnell, A., BradWeld, P., and Crocker, P. R. (1994). Sialoadhesin, myelin-associated glycoprotein and CD22 deWne a new family of sialic acid-dependent adhesion molecules of the immunoglobulin superfamily. Curr Biol 4, 965–972. Kidd, G. J., and Heath, J. W. (1988). Double myelination of axons in the sympathetic nervous system of the mouse. II. Mechanisms of formation. J Neurocytol 17, 263–276. KirchhoV, F., Hofer, H. W., and Schachner, M. (1993). Myelin-associated glycoprotein is phosphorylated by protein kinase C. J Neurosci Res 36, 368–381. Kirkpatrick, L. L., and Brady, S. T. (1994). Modulation of the axonal microtubule cytoskeleton by myelinating Schwann cells. J Neurosci 14, 7440–7450. Kirkpatrick, L. L., Witt, A. S., Payne, H. R., Shine, H. D., and Brady, S. T. (2001). Changes in microtubule stability and density in myelin-deWcient shiverer mouse CNS axons. J Neurosci 21, 2288–2297. Kruse, J., Mailhammer, R., Wernecke, H., Faissner, A., Sommer, I., Goridis, C., and Schachner, M. (1984). Neural cell adhesion molecules and myelin-associated glycoprotein share a common carbohydrate moiety recognized by monoclonal antibodies L2 and HNK-1. Nature 311, 153–155. Kumar, S., Yin, X., Trapp, B. D., Paulaitis, M. E., and Hoh, J. H. (2002). Role of long-range repulsive forces in organizing axonal neuroWlament distributions: Evidence from mice deWcient in myelin-associated glycoprotein. J Neurosci Res 68, 681–690. Kunemund, V., Jungalwala, F. B., Fischer, G., Chou, D. K., Keilhauer, G., and Schachner, M. (1988). The L2/ HNK-1 carbohydrate of neural cell adhesion molecules is involved in cell interactions. J Cell Biol 106, 213–223. Kursula, P. (2001). The current status of structural studies on proteins of the myelin sheath (Review). Int J Mol Med 8, 475–479.

461

462

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

Kursula, P., Lehto, V. P., and Heape, A. M. (2000). S100b inhibits the phosphorylation of the L-MAG cytoplasmic domain by PKA. Brain Res Mol Brain Res 76, 407–410. Kursula, P., Tikkanen, G., Lehto, V. P., Nishikimi, M., and Heape, A. M. (1999). Calcium-dependent interaction between the large myelin-associated glycoprotein and S100b. J Neurochem 73, 1724–1732. Lai, C., Brow, M. A., Nave, K. A., Noronha, A. B., Quarles, R. H., Bloom, F. E., Milner, R. J., and SutcliVe, J. G. (1987a). Two forms of 1B236/myelin-associated glycoprotein, a cell adhesion molecule for postnatal neural development, are produced by alternative splicing. Proc Natl Acad Sci USA 84, 4337–4341. Lai, C., Watson, J. B., Bloom, F. E., SutcliVe, J. G., and Milner, R. J. (1987b). Neural protein 1B236/myelinassociated glycoprotein (MAG) deWnes a subgroup of the immunoglobulin superfamily. Immunol Rev 100, 129–151. Lang, P., Gesbert, F., Delespine-Carmagnat, M., Stancou, R., Pouchelet, M., and Bertoglio, J. (1996). Protein kinase A phosphorylation of RhoA mediates the morphological and functional eVects of cyclic AMP in cytotoxic lymphocytes. Embo J 15, 510–519. Lassmann, H., Bartsch, U., Montag, D., and Schachner, M. (1997). Dying-back oligodendrogliopathy: A late sequel of myelin-associated glycoprotein deWciency. Glia 19, 104–110. Laszkiewicz, I., Grubinska, B., Wiggins, R. C., and Konat, G. W. (1997). Structural characterization of myelinassociated glycoprotein gene core promoter. J Neurosci Res 50, 928–936. Lehmann, M., Fournier, A., Selles-Navarro, I., Dergham, P., Sebok, A., Leclerc, N., Tigyi, G., and McKerracher, L. (1999). Inactivation of Rho signaling pathway promotes CNS axon regeneration. J Neurosci 19, 7537–7547. Li, C., Trapp, B., Ludwin, S., Peterson, A., and Roder, J. (1998). Myelin associated glycoprotein modulates gliaaxon contact in vivo. J Neurosci Res 51, 210–217. Li, C., Tropak, M. B., Gerlai, R., ClapoV, S., Abramow-Newerly, W., Trapp, B., Peterson, A., and Roder, J. (1994). Myelination in the absence of myelin-associated glycoprotein. Nature 369, 747–750. Li, M., Shibata, A., Li, C., Braun, P. E., McKerracher, L., Roder, J., Kater, S. B., and David, S. (1996). Myelinassociated glycoprotein inhibits neurite/axon growth and causes growth cone collapse. J Neurosci Res 46, 404–414. Li, Z., Zhang, Y., Li, D., and Feng, Y. (2000). Destabilization and mislocalization of myelin basic protein mRNAs in quaking dysmyelination lacking the QKI RNA-binding proteins. J Neurosci 20, 4944–4953. Link, H., Sun, J. B., Wang, Z., Xu, Z., Love, A., Fredrikson, S., and Olsson, T. (1992). Virus-reactive and autoreactive T cells are accumulated in cerebrospinal Xuid in multiple sclerosis. J Neuroimmunol 38, 63–73. Liu, B. P., Fournier, A., GrandPre´, T., and Strittmatter, S. M. (2002). Myelin-Associated Glycoprotein as a Functional Ligand for the Nogo-66 Receptor. Science 27, 27. Lunn, E. R., Perry, V. H., Brown, M. C., Rosen, H., and Gordon, S. (1989). Absence of Wallerian Degeneration does not Hinder Regeneration in Peripheral Nerve. Eur J Neurosci 1, 27–33. Lunn, M. P., Crawford, T. O., Hughes, R. A., GriYn, J. W., and Sheikh, K. A. (2002). Anti-myelin-associated glycoprotein antibodies alter neuroWlament spacing. Brain 125, 904–911. Marcus, J., Dupree, J. L., and Popko, B. (2000). EVects of galactolipid elimination on oligodendrocyte development and myelination. Glia 30, 319–328. Marcus, J., Dupree, J. L., and Popko, B. (2002). Myelin-associated glycoprotein and myelin galactolipids stabilize developing axo-glial interactions. J Cell Biol 156, 567–577. Martini, R. (1994). Myelin-associated glycoprotein is not detectable in perikaryal myelin of spiral ganglion neurons of adult mice. Glia 10, 311–314. Martini, R., and Schachner, M. (1986). Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve. J Cell Biol 103, 2439–2448. Mata, S., Lolli, F., Soderstrom, M., Pinto, F., and Link, H. (1999). Multiple sclerosis is associated with enhanced B cell responses to the ganglioside GD1a. Mult Scler 5, 379–388. Matsuda, Y., Okitsu, A., Sato, S., Koito, H., and Yamamoto, H. (1996). Soluble myelin-associated glycoproteinimmunoglobulin G1 chimera protein promotes neurite outgrowth from mouse cerebellar neurons. Neurosci Lett 205, 87–90. Matthieu, J. M., Brady, R. O., and Quarles, R. H. (1974). Anomalies of myelin-associated glycoproteins in ‘Quaking’ mice. J Neurochem 22, 291–296. Matthieu, J. M., Quarles, R. H., Poduslo, J. F., and Brady, R. O. (1975). [35S]sulfate incorporation into myelin glycoproteins. I. Central nervous system. Biochim Biophys Acta 392, 159–166. May, A. P., and Jones, E. Y. (2001). Sialoadhesin structure. Results Probl Cell DiVer 33, 139–151. May, A. P., Robinson, R. C., Vinson, M., Crocker, P. R., and Jones, E. Y. (1998). Crystal structure of the N-terminal domain of sialoadhesin in complex with 3’ sialyllactose at 1.85 A resolution. Mol Cell 1, 719–728. McGarry, R. C., Helfand, S. L., Quarles, R. H., and Roder, J. C. (1983). Recognition of myelin-associated glycoprotein by the monoclonal antibody HNK-1. Nature 306, 376–378. McKerracher, L. (2002). Ganglioside rafts as MAG receptors that mediate blockade of axon growth. Proc Natl Acad Sci USA 99, 7811–7813. McKerracher, L., David, S., Jackson, D. L., Kottis, V., Dunn, R. J., and Braun, P. E. (1994). IdentiWcation of myelin-associated glycoprotein as a major myelin-derived inhibitor of neurite growth. Neuron 13, 805–811. Melmed, C., Frail, D., Duncan, I., Braun, P., DanoV, D., Finlayson, M., and Stewart, J. (1983). Peripheral neuropathy with IgM kappa monoclonal immunoglobulin directed against myelin-associated glycoprotein. Neurology 33, 1397–1405.

SUMMARY AND CLOSING REMARKS

Meyer-Franke, A., and Barres, B. (1994). Axon myelination. Myelination without myelin-associated glycoprotein. Curr Biol 4, 847–850. Meyer-Franke, A., Tropak, M. B., Roder, J. C., Fischer, P., Beyreuther, K., Probstmeier, R., and Schachner, M. (1995). Functional topography of myelin-associated glycoprotein. II. Mapping of domains on molecular fragments. J Neurosci Res 41, 311–323. Miescher, G. C., Lutzelschwab, R., Erne, B., Ferracin, F., Huber, S., and Steck, A. J. (1997). Reciprocal expression of myelin-associated glycoprotein splice variants in the adult human peripheral and central nervous systems. Brain Res Mol Brain Res 52, 299–306. Mizutani, K., Oka, N., Kusunoki, S., Kaji, R., Mezaki, T., Akiguchi, I., and Shibasaki, H. (2001). Sensorimotor demyelinating neuropathy with IgM antibody against gangliosides GD1a, GT1b and GM3. J Neurol Sci 188, 9–11. Mo¨ller, J. R., Yanagisawa, K., Brady, R. O., Tourtellotte, W. W., and Quarles, R. H. (1987). Myelin-associated glycoprotein in multiple sclerosis lesions: A quantitative and qualitative analysis. Ann Neurol 22, 469–474. Montag, D., Giese, K. P., Bartsch, U., Martini, R., Lang, Y., Bluthmann, H., Karthigasan, J., Kirschner, D. A., Wintergerst, E. S., Nave, K. A., Zielasek, J., Toyka, K. V., Lipp, H., and Schachner, M. (1994). Mice deWcient for the myelin-associated glycoprotein show subtle abnormalities in myelin. Neuron 13, 229–246. Mucklow, S., Hartnell, A., Mattei, M. G., Gordon, S., and Crocker, P. R. (1995). Sialoadhesin (Sn) maps to mouse chromosome 2 and human chromosome 20 and is not linked to the other members of the sialoadhesin family, CD22, MAG, and CD33. Genomics 28, 344–346. Mukhopadhyay, G., Doherty, P., Walsh, F. S., Crocker, P. R., and Filbin, M. T. (1994). A novel role for myelinassociated glycoprotein as an inhibitor of axonal regeneration. Neuron 13, 757–767. Muramoto, K., Taniguchi, H., Kawahara, M., Kobayashi, K., Nonomura, Y., and Kuroda, Y. (1994). A substrate of ecto-protein kinase is microtubule-associated protein 1B in cortical cell cultures undergoing synaptogenesis. Biochem Biophys Res Commun 205, 1467–1473. Nachamkin, I., Liu, J., Li, M., Ung, H., Moran, A. P., Prendergast, M. M., and Sheikh, K. (2002). Campylobacter jejuni from Patients with Guillain-Barre Syndrome Preferentially Expresses a GD(1a)-Like Epitope. Infect Immun 70, 5299–5303. Nakano, R., Fujita, N., Sato, S., Inuzuka, T., Sakimura, K., Ishiguro, H., Mishina, M., and Miyatake, T. (1991). Structure of mouse myelin-associated glycoprotein gene. Biochem Biophys Res Commun 178, 282–290. Nath, D., van der Merwe, P. A., Kelm, S., BradWeld, P., and Crocker, P. R. (1995). The amino-terminal immunoglobulin-like domain of sialoadhesin contains the sialic acid binding site. Comparison with CD22. J Biol Chem 270, 26184–26191. Neumann, S., Bradke, F., Tessier-Lavigne, M., and Basbaum, A. I. (2002). Regeneration of sensory axons within the injured spinal cord induced by intraganglionic cAMP elevation. Neuron 34, 885–893. Nobile-Orazio, E., Manfredini, E., Carpo, M., Meucci, N., Monaco, S., Ferrari, S., Bonetti, B., Cavaletti, G., Gemignani, F., Durelli, L., Barbieri, S., Allaria, S., Sgarzi, M., and Scarlato, G. (1994). Frequency and Clinical Correlates of Anti-Neural IgM Antibodies in Neuropathy Associated with IgM Monoclonal Gammopathy. Ann Neurol 36, 416–424. Noronha, A. B., Hammer, J. A., Lai, C., Kiel, M., Milner, R. J., SutcliVe, J. G., and Quarles, R. H. (1989). Myelin-associated glycoprotein (MAG) and rat brain-speciWc 1B236 protein: Mapping of epitopes and demonstration of immunological identity. J Mol Neurosci 1, 159–170. Noronha, A. B., Ilyas, A., Antonicek, H., Schachner, M., and Quarles, R. H. (1986). Molecular speciWcity of L2 monoclonal antibodies that bind to carbohydrate determinants of neural cell adhesion molecules and their resemblance to other monoclonal antibodies recognizing the myelin-associated glycoprotein. Brain Res 385, 237–244. Ong, E., Suzuki, M., Belot, F., Yeh, J. C., Franceschini, I., Angata, K., Hindsgaul, O., and Fukuda, M. (2002). Biosynthesis of HNK-1 glycans on O-linked oligosaccharides attached to the neural cell adhesion molecule (NCAM): The requirement for core 2b1, 6-N-acetylglucosaminyltransferase and the muscle-speciWc domain in NCAM. J Biol Chem 277, 18182–18190. O’Shannessy, D. J., Willison, H. J., Inuzuka, T., Dobersen, M. J., and Quarles, R. H. (1985). The species distribution of nervous system antigens that react with anti-myelin-associated glycoprotein antibodies. J Neuroimmunol 9, 255–268. Owens, G. C., Boyd, C. J., Bunge, R. P., and Salzer, J. L. (1990). Expression of recombinant myelin-associated glycoprotein in primary Schwann cells promotes the initial investment of axons by myelinating Schwann cells. J Cell Biol 111, 1171–1182. Owens, G. C., and Bunge, R. P. (1989). Evidence for an early role for myelin-associated glycoprotein in the process of myelination. Glia 2, 119–128. Owens, G. C., and Bunge, R. P. (1991). Schwann cells infected with a recombinant retrovirus expressing myelinassociated glycoprotein antisense RNA do not form myelin. Neuron 7, 565–575. Paulson, J. C., and Colley, K. J. (1989). Glycosyltransferases. Structure, localization, and control of cell typespeciWc glycosylation. J Biol Chem 264, 17615–17618. Pedraza, L., Frey, A. B., Hempstead, B. L., Colman, D. R., and Salzer, J. L. (1991). DiVerential expression of MAG isoforms during development. J Neurosci Res 29, 141–148. Pedraza, L., Owens, G. C., Green, L. A., and Salzer, J. L. (1990). The myelin-associated glycoproteins: Membrane disposition, evidence of a novel disulWde linkage between immunoglobulin-like domains, and posttranslational palmitylation. J Cell Biol 111, 2651–2661.

463

464

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

Pedraza, L., Spagnol, G., Latov, N., and Salzer, J. L. (1995). Biosynthesis and regulation of expression of the HNK-1 epitope on myelin-associated glycoprotein in a transfected cell model system. J Neurosci Res 40, 716–727. Peiper, S. C., Ashmun, R. A., and Look, A. T. (1988). Molecular cloning, expression, and chromosomal localization of a human gene encoding the CD33 myeloid diVerentiation antigen. Blood 72, 314–321. Peirano, R. I., Goerich, D. E., Riethmacher, D., and Wegner, M. (2000). Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol Cell Biol 20, 3198–3209. Pierelli, L., TeoWli, L., Menichella, G., Rumi, C., Paoloni, A., Iovino, S., Puggioni, P. L., Leone, G., and Bizzi, B. (1993). Further investigations on the expression of HLA-DR, CD33 and CD13 surface antigens in puriWed bone marrow and peripheral blood CD34þ haematopoietic progenitor cells. Br J Haematol 84, 24–30. Poltorak, M., Sadoul, R., Keilhauer, G., Landa, C., Fahrig, T., and Schachner, M. (1987). Myelin-associated glycoprotein, a member of the L2/HNK-1 family of neural cell adhesion molecules, is involved in neuronoligodendrocyte and oligodendrocyte-oligodendrocyte interaction. J Cell Biol 105, 1893–1899. Powell, L. D., and Varki, A. (1995). I-type lectins. J Biol Chem 270, 14243–14246. Prinjha, R., Moore, S. E., Vinson, M., Blake, S., Morrow, R., Christie, G., Michalovich, D., Simmons, D. L., and Walsh, F. S. (2000). Inhibitor of neurite outgrowth in humans. Nature 403, 383–384. Probstmeier, R., Fahrig, T., Spiess, E., and Schachner, M. (1992). Interactions of the neural cell adhesion molecule and the myelin-associated glycoprotein with collagen type I: Involvement in Wbrillogenesis. J Cell Biol 116, 1063–1070. Qiu, J., Cai, D., Dai, H., McAtee, M., HoVman, P. N., Bregman, B. S., and Filbin, M. T. (2002). Spinal axon regeneration induced by elevation of cyclic AMP. Neuron 34, 895–903. Quarles, R. H. (1983). Myelin-associated glycoprotein in development and disease. Dev Neurosci 6, 285–303. Quarles, R. H., Barbarash, G. R., Figlewicz, D. A., and McIntyre, L. J. (1983). PuriWcation and partial characterization of the myelin-associated glycoprotein from adult rat brain. Biochim Biophys Acta 757, 140–143. Quarles, R. H., Everly, J. L., and Brady, R. O. (1973). Myelin-associated glycoprotein: A developmental change. Brain Res 58, 506–509. Quarles, R. H., and Weiss, M. D. (1999). Autoantibodies associated with peripheral neuropathy. Muscle Nerve 22, 800–822. Resh, M. D. (1994). Myristylation and palmitylation of Src family members: The fats of the matter. Cell 76, 411–413. Riopelle, R. J., McGarry, R. C., and Roder, J. C. (1986). Adhesion properties of a neuronal epitope recognized by the monoclonal antibody HNK-1. Brain Res 367, 20–25. Rome, L. H., Bullock, P. N., Chiappelli, F., Cardwell, M., AdinolW, A. M., and Swanson, D. (1986). Synthesis of a myelin-like membrane by oligodendrocytes in culture. J Neurosci Res 15, 49–65. Sadoul, R., Fahrig, T., Bartsch, U., and Schachner, M. (1990). Binding properties of liposomes containing the myelin-associated glycoprotein MAG to neural cell cultures. J Neurosci Res 25, 1–13. Salzer, J. L., Holmes, W. P., and Colman, D. R. (1987). The amino acid sequences of the myelin-associated glycoproteins: Homology to the immunoglobulin gene superfamily. J Cell Biol 104, 957–965. Sanchez, I., Hassinger, L., Paskevich, P. A., Shine, H. D., and Nixon, R. A. (1996). Oligodendroglia regulate the regional expansion of axon caliber and local accumulation of neuroWlaments during development independently of myelin formation. J Neurosci 16, 5095–5105. Sato, S., Baba, H., Inuzuka, T., and Miyatake, T. (1986). Anti-myelin-associated glycoprotein antibody in sera from patients with demyelinating diseases. Acta Neurol Scand 74, 115–120. Sato, S., Fujita, N., Kurihara, T., Kuwano, R., Sakimura, K., Takahashi, Y., and Miyatake, T. (1989). cDNA cloning and amino acid sequence for human myelin-associated glycoprotein. Biochem Biophys Res Commun 163, 1473–1480. Sato, S., Yanagisawa, K., and Miyatake, T. (1984). Conversion of myelin-associated glycoprotein (MAG) to a smaller derivative by calcium activated neutral protease (CANP)-like enzyme in myelin and inhibition by E-64 analogue. Neurochem Res 9, 629–635. Schachner, M., and Bartsch, U. (2000). Multiple functions of the myelin-associated glycoprotein MAG (siglec-4a) in formation and maintenance of myelin. Glia 29, 154–165. Scha¨fer, M., Fruttiger, M., Montag, D., Schachner, M., and Martini, R. (1996). Disruption of the gene for the myelin-associated glycoprotein improves axonal regrowth along myelin in C57BL/Wlds mice. Neuron 16, 1107–1113. Schauer, R. (1982). Chemistry, metabolism, and biological functions of sialic acids. Adv Carbohydr Chem Biochem 40, 131–234. Schmidt, S., Barcellos, L. F., DeSombre, K., Rimmler, J. B., Lincoln, R. R., Bucher, P., Saunders, A. M., Lai, E., Martin, E. R., Vance, J. M., Oksenberg, J. R., Hauser, S. L., Pericak-Vance, M. A., and Haines, J. L. (2002). Association of polymorphisms in the apolipoprotein E region with susceptibility to and progression of multiple sclerosis. Am J Hum Genet 70, 708–717. Schnell, L., and Schwab, M. E. (1990). Axonal regeneration in the rat spinal cord produced by an antibody against myelin-associated neurite growth inhibitors. Nature 343, 269–272. Schurmann, G., Haspel, J., Grumet, M., and Erickson, H. P. (2001). Cell adhesion molecule L1 in folded (horseshoe) and extended conformations. Mol Biol Cell 12, 1765–1773. Schwerer, B. (2002). Antibodies against gangliosides: A link between preceding infection and immunopathogenesis of Guillain-Barre´ syndrome. Microbes Infect 4, 373–384.

SUMMARY AND CLOSING REMARKS

Sgroi, D., Nocks, A., and Stamenkovic, I. (1996). A single N-linked glycosylation site is implicated in the regulation of ligand recognition by the I-type lectins CD22 and CD33. J Biol Chem 271, 18803–18809. Sheikh, K. A., Sun, J., Liu, Y., Kawai, H., Crawford, T. O., Proia, R. L., GriYn, J. W., and Schnaar, R. L. (1999). Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proc Natl Acad Sci USA 96, 7532–7537. Shen, Y. J., DeBellard, M. E., Salzer, J. L., Roder, J., and Filbin, M. T. (1998). Myelin-associated glycoprotein in myelin and expressed by Schwann cells inhibits axonal regeneration and branching. Mol Cell Neurosci 12, 79–91. Shimizu-Okabe, C., Matsuda, Y., Koito, H., and Yoshida, S. (2001). L-isoform but not S-isoform of myelin associated glycoprotein promotes neurite outgrowth of mouse cerebellar neurons. Neurosci Lett 311, 203–205. Sidman, R. L., Dickie, M. M., and Appel, S. H. (1964). Mutant mice (quaking and jimpy) with deWcient myelination in the central nervous system. Science 144, 309–311. Simmons, D., and Seed, B. (1988). Isolation of a cDNA encoding CD33, a diVerentiation antigen of myeloid progenitor cells. J Immunol 141, 2797–2800. Skaper, S. D., Moore, S. E., and Walsh, F. S. (2001). Cell signalling cascades regulating neuronal growthpromoting and inhibitory cues. Prog Neurobiol 65, 593–608. Snipes, G. J., Suter, U., and Shooter, E. M. (1993). Human peripheral myelin protein-22 carries the L2/HNK1 carbohydrate adhesion epitope. J Neurochem 61, 1961–1964. Song, H., Ming, G., He, Z., Lehmann, M., McKerracher, L., Tessier-Lavigne, M., and Poo, M. (1998). Conversion of neuronal growth cone responses from repulsion to attraction by cyclic nucleotides. Science 281, 1515–1518. Spagnol, G., Williams, M., Srinivasan, J., Golier, J., Bauer, D., Lebo, R. V., and Latov, N. (1989). Molecular cloning of human myelin-associated glycoprotein. J Neurosci Res 24, 137–142. Spillmann, A. A., Bandtlow, C. E., Lottspeich, F., Keller, F., and Schwab, M. E. (1998). IdentiWcation and characterization of a bovine neurite growth inhibitor (bNI-220). J Biol Chem 273, 19283–19293. Stamenkovic, I., and Seed, B. (1990). The B-cell antigen CD22 mediates monocyte and erythrocyte adhesion. Nature 345, 74–77. Stamenkovic, I., Sgroi, D., AruVo, A., Sy, M. S., and Anderson, T. (1991). The B lymphocyte adhesion molecule CD22 interacts with leukocyte common antigen CD45RO on T cells and a2–6 sialyltransferase, CD75, on B cells. Cell 66, 1133–1144. Steck, A. J., Erne, B., Gabriel, J. M., and Schaeren-Wiemers, N. (1999). Paraproteinaemic Neuropathies. Brain Pathol 9, 361–368. Sternberger, N. H., Quarles, R. H., Itoyama, Y., and Webster, H. D. (1979). Myelin-associated glycoprotein demonstrated immunocytochemically in myelin and myelin-forming cells of developing rat. Proc Natl Acad Sci USA 76, 1510–1514. Stolt, C. C., Rehberg, S., Ader, M., Lommes, P., Riethmacher, D., Schachner, M., Bartsch, U., and Wegner, M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev 16, 165–170. Su, X. D., Gastinel, L. N., Vaughn, D. E., Faye, I., Poon, P., and Bjorkman, P. J. (1998). Crystal structure of hemolin: A horseshoe shape with implications for homophilic adhesion. Science 281, 991–995. SutcliVe, J. G., Milner, R. J., Shinnick, T. M., and Bloom, F. E. (1983). Identifying the protein products of brainspeciWc genes with antibodies to chemically synthesized peptides. Cell 33, 671–682. Takamiya, K., Yamamoto, A., Furukawa, K., Yamashiro, S., Shin, M., Okada, M., Fukumoto, S., Haraguchi, M., Takeda, N., Fujimura, K., Sakae, M., Kishikawa, M., Shiku, H., and Aizawa, S. (1996). Mice with disrupted GM2/GD2 synthase gene lack complex gangliosides but exhibit only subtle defects in their nervous system. Proc Natl Acad Sci USA 93, 10662–10667. Takatsu, M., Hays, A. P., Latov, N., Abrams, G. M., Nemni, R., Sherman, W. H., Nobile-Orazio, E., Saito, T., and Freddo, L. (1985). ImmunoXuorescence study of patients with neuropathy and IgM M proteins. Ann Neurol 18, 173–181. Tang, S., Qiu, J., Nikulina, E., and Filbin, M. T. (2001). Soluble myelin-associated glycoprotein released from damaged white matter inhibits axonal regeneration. Mol Cell Neurosci 18, 259–269. Tang, S., Shen, Y. J., DeBellard, M. E., Mukhopadhyay, G., Salzer, J. L., Crocker, P. R., and Filbin, M. T. (1997a). Myelin-associated glycoprotein interacts with neurons via a sialic acid binding site at ARG118 and a distinct neurite inhibition site. J Cell Biol 138, 1355–1366. Tang, S., Woodhall, R. W., Shen, Y. J., deBellard, M. E., SaVell, J. L., Doherty, P., Walsh, F. S., and Filbin, M. T. (1997b). Soluble Myelin-Associated Glycoprotein (MAG) Found in Vivo Inhibits Axonal Regeneration. Mol Cell Neurosci 9, 333–346. Tanner, S. L., Franzen, R., JaVe, H., and Quarles, R. H. (2000). Evidence for expression of some microtubuleassociated protein 1B in neurons as a plasma membrane glycoprotein. J Neurochem 75, 553–562. Tatum, A. H. (1993). Experimental paraprotein neuropathy, demyelination by passive transfer of human IgM anti-myelin-associated glycoprotein. Ann Neurol 33, 502–506. Tchilian, E. Z., Beverley, P. C., Young, B. D., and Watt, S. M. (1994). Molecular cloning of two isoforms of the murine homolog of the myeloid CD33 antigen. Blood 83, 3188–3198. Topilko, P., Schneider-Maunoury, S., Levi, G., Baron-Van Evercooren, A., ChennouW, A. B., Seitanidou, T., Babinet, C., and Charnay, P. (1994). Krox-20 controls myelination in the peripheral nervous system. Nature 371, 796–799.

465

466

17. MYELIN-ASSOCIATED GLYCOPROTEIN GENE

Trapp, B. D. (1988). Distribution of the myelin-associated glycoprotein and P0 protein during myelin compaction in quaking mouse peripheral nerve. J Cell Biol 107, 675–685. Trapp, B. D., Andrews, S. B., Cootauco, C., and Quarles, R. (1989a). The myelin-associated glycoprotein is enriched in multivesicular bodies and periaxonal membranes of actively myelinating oligodendrocytes. J Cell Biol 109, 2417–2426. Trapp, B. D., Andrews, S. B., Wong, A., O’Connell, M., and GriYn, J. W. (1989b). Co-localization of the myelinassociated glycoprotein and the microWlament components, F-actin and spectrin, in Schwann cells of myelinated nerve Wbres. J Neurocytol 18, 47–60. Trapp, B. D., and Quarles, R. H. (1982). Presence of the myelin-associated glycoprotein correlates with alterations in the periodicity of peripheral myelin. J Cell Biol 92, 877–882. Trapp, B. D., Quarles, R. H., and Suzuki, K. (1984). Immunocytochemical studies of quaking mice support a role for the myelin-associated glycoprotein in forming and maintaining the periaxonal space and periaxonal cytoplasmic collar of myelinating Schwann cells. J Cell Biol 99, 594–606. Trask, B., Fertitta, A., Christensen, M., Youngblom, J., Bergmann, A., Copeland, A., de Jong, P., Mohrenweiser, H., Olsen, A., Carrano, A., and Tynan, K. (1993). Fluorescence in situ hybridization mapping of human chromosome 19: Cytogenetic band location of 540 cosmids and 70 genes or DNA markers. Genomics 15, 133–145. Trojaborg, W., Galassi, G., Hays, A. P., Lovelace, R. E., Alkaitis, M., and Latov, N. (1989). Electrophysiologic study of experimental demyelination induced by serum of patients with IgM M proteins and neuropathy. Neurology 39, 1581–1586. Tropak, M. B., Johnson, P. W., Dunn, R. J., and Roder, J. C. (1988). DiVerential splicing of MAG transcripts during CNS and PNS development. Brain Res 464, 143–155. Tropak, M. B., and Roder, J. C. (1997). Regulation of myelin-associated glycoprotein binding by sialylated cisligands. J Neurochem 68, 1753–1763. Turnley, A. M., and Bartlett, P. F. (1998). MAG and MOG enhance neurite outgrowth of embryonic mouse spinal cord neurons. Neuroreport 9, 1987–1990. Umemori, H., Kadowaki, Y., Hirosawa, K., Yoshida, Y., Hironaka, K., Okano, H., and Yamamoto, T. (1999). Stimulation of myelin basic protein gene transcription by Fyn tyrosine kinase for myelination. J Neurosci 19, 1393–1397. Umemori, H., Sato, S., Yagi, T., Aizawa, S., and Yamamoto, T. (1994). Initial events of myelination involve Fyn tyrosine kinase signalling. Nature 367, 572–576. Uschkureit, T., Spo¨rkel, O., Stracke, J., Bu¨ssow, H., and StoVel, W. (2000). Early onset of axonal degeneration in double (plp
/
mag
/
) and hypomyelinosis in triple (plp
/
mbp
/
mag
/
) mutant mice. J Neurosci 20, 5225–5233. Vabnick, I., Messing, A., Chiu, S. Y., Levinson, S. R., Schachner, M., Roder, J., Li, C., Novakovic, S., and Shrager, P. (1997). Sodium channel distribution in axons of hypomyelinated and MAG null mutant mice. J Neurosci Res 50, 321–336. Van den Berg, L., Hays, A. P., Nobile-Orazio, E., Kinsella, L. J., Manfredini, E., Corbo, M., Rosoklija, G., Younger, D. S., Lovelace, R. E., Trojaborg, W., Lange, D. E., Goldstein, S., DelWner, J. S., Sadiq, S. A., Sherman, W. H., and Latov, N. (1996). Anti-MAG and anti-SGPG antibodies in neuropathy. Muscle Nerve 19, 637–643. van den Berg, L. H., Sadiq, S. A., Lederman, S., and Latov, N. (1992a). The gp120 glycoprotein of HIV-1 binds to sulfatide and to the myelin associated glycoprotein. J Neurosci Res 33, 513–518. van den Berg, T. K., Breve, J. J., Damoiseaux, J. G., Dopp, E. A., Kelm, S., Crocker, P. R., Dijkstra, C. D., and Kraal, G. (1992b). Sialoadhesin on macrophages: Its identiWcation as a lymphocyte adhesion molecule. J Exp Med 176, 647–655. van der Merwe, P. A., and Barclay, A. N. (1994). Transient intercellular adhesion: The importance of weak protein-protein interactions. Trends Biochem Sci 19, 354–358. Varki, A. (1992). Selectins and other mammalian sialic acid-binding lectins. Curr Opin Cell Biol 4, 257–266. Vernet, C., and Artzt, K. (1997). STAR, a gene family involved in signal transduction and activation of RNA. Trends Genet 13, 479–484. Vinson, M., Strijbos, P. J., Rowles, A., Facci, L., Moore, S. E., Simmons, D. L., and Walsh, F. S. (2001). Myelinassociated glycoprotein interacts with ganglioside GT1b. A mechanism for neurite outgrowth inhibition. J Biol Chem 276, 20280–20285. Voshol, H., van Zuylen, C. W., Orberger, G., Vliegenthart, J. F., and Schachner, M. (1996). Structure of the HNK-1 carbohydrate epitope on bovine peripheral myelin glycoprotein P0. J Biol Chem 271, 22957–22960. Vyas, A. A., Patel, H. V., Fromholt, S. E., HeVer-Lauc, M., Vyas, K. A., Dang, J., Schachner, M., and Schnaar, R. L. (2002). Gangliosides are functional nerve cell ligands for myelin-associated glycoprotein (MAG), an inhibitor of nerve regeneration. Proc Natl Acad Sci USA 99, 8412–8417. Wajgt, A., and Gorny, M. (1983). CSF antibodies to myelin basic protein and to myelin-associated glycoprotein in multiple sclerosis. Evidence of the intrathecal production of antibodies. Acta Neurol Scand 68, 337–343. Wang, K. C., Koprivica, V., Kim, J. A., Sivasankaran, R., Guo, Y., Neve, R. L., and He, Z. (2002). Oligodendrocyte-myelin glycoprotein is a Nogo receptor ligand that inhibits neurite outgrowth. Nature 417, 941–944. Wegner, M. (2000). Transcriptional control in myelinating glia: Flavors and spices. Glia 31, 1–14. Weiss, M. D., Dalakas, M. C., Lauter, C. J., Willison, H. J., and Quarles, R. H. (1999). Variability in the binding of anti-MAG and anti-SGPG antibodies to target antigens in demyelinating neuropathy and IgM paraproteinemia. J Neuroimmunol 95, 174–184.

SUMMARY AND CLOSING REMARKS

Weiss, M. D., Hammer, J., and Quarles, R. H. (2000). Oligodendrocytes in aging mice lacking myelin-associated glycoprotein are dystrophic but not apoptotic. J Neurosci Res 62, 772–780. Weiss, M. D., Luciano, C. A., and Quarles, R. H. (2001). Nerve conduction abnormalities in aging mice deWcient for myelin-associated glycoprotein. Muscle Nerve 24, 1380–1387. Williams, A. F., and Barclay, A. N. (1988). The immunoglobulin superfamily--domains for cell surface recognition. Annu Rev Immunol 6, 381–405. Willison, H., Stoll, G., Toyka, K. V., Berger, T., and Hartung, H. P. (2002). Autoimmunity and inXammation in the peripheral nervous system. Trends Neurosci 25, 127–129. Willison, H. J., Trapp, B. D., Bacher, J. D., Dalakas, M. C., GriYn, J. W., and Quarles, R. H. (1988). Demyelination induced by intraneural injection of human antimyelin-associated glycoprotein antibodies. Muscle Nerve 11, 1169–1176. Wilson, G. L., Fox, C. H., Fauci, A. S., and Kehrl, J. H. (1991). cDNA cloning of the B cell membrane protein CD22: A mediator of B-B cell interactions. J Exp Med 173, 137–146. Wilson, G. L., Najfeld, V., Kozlow, E., Menniger, J., Ward, D., and Kehrl, J. H. (1993). Genomic structure and chromosomal mapping of the human CD22 gene. J Immunol 150, 5013–5024. Wu, J. I., Reed, R. B., Grabowski, P. J., and Artzt, K. (2002). Function of quaking in myelination: Regulation of alternative splicing. Proc Natl Acad Sci USA 99, 4233–4238. Yamashita, T., Higuchi, H., and Tohyama, M. (2002). The p75 receptor transduces the signal from myelinassociated glycoprotein to Rho. J Cell Biol 157, 565–570. Yamashita, T., Tucker, K. L., and Barde, Y. A. (1999). Neurotrophin binding to the p75 receptor modulates Rho activity and axonal outgrowth. Neuron 24, 585–593. Yang, H., Xiao, Z. C., Becker, B., Hillenbrand, R., Rougon, G., and Schachner, M. (1999). Role for myelinassociated glycoprotein as a functional tenascin-R receptor. J Neurosci Res 55, 687–701. Yang, L. J., Zeller, C. B., Shaper, N. L., Kiso, M., Hasegawa, A., Shapiro, R. E., and Schnaar, R. L. (1996). Gangliosides are neuronal ligands for myelin-associated glycoprotein. Proc Natl Acad Sci USA 93, 814–818. Ye, P., Laszkiewicz, I., Wiggins, R. C., and Konat, G. W. (1994). Transcriptional regulation of myelin associated glycoprotein gene expression by cyclic AMP. J Neurosci Res 37, 683–690. Yeung, K. B., Thomas, P. K., King, R. H., Waddy, H., Will, R. G., Hughes, R. A., Gregson, N. A., and Leibowitz, S. (1991). The clinical spectrum of peripheral neuropathies associated with benign monoclonal IgM, IgG and IgA paraproteinaemia. Comparative clinical, immunological and nerve biopsy Wndings. J Neurol 238, 383–391. Yim, S. H., Toda, K., Goda, S., and Quarles, R. H. (1995). Comparison of the phosphorylation of myelinassociated glycoprotein in cultured oligodendrocytes and Schwann cells. J Mol Neurosci 6, 63–74. Yin, X., Crawford, T. O., GriYn, J. W., Tu, P., Lee, V. M., Li, C., Roder, J., and Trapp, B. D. (1998). Myelinassociated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J Neurosci 18, 1953–1962.

467

C H A P T E R

18 Myelin Oligodendrocyte Glycoprotein Gene Danielle Pham-Dinh, Andre Dautigny, and Christopher Linington

INTRODUCTION Myelin oligodendrocyte glycoprotein (MOG) was characterized initially as a target for autoantibody-mediated demyelination in experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis (MS). Studies on EAE date back to the mid1950s, and this model continues to provide new insights into the immunopathogenesis of demyelination in MS. It was recognized in the late 1960s that sera from animals with EAE induced by immunization with central nervous system (CNS) tissue homogenates contained an immunoglobulin fraction that mediated demyelination in vitro. Initially myelin glycolipids were considered the most probable target for this demyelinating autoantibody response, a concept supported by the in vitro demyelinating activity of antibodies to galactosyl ceramide, the major myelin glycolipid (Niedieck, 1969, 1975; au1 Raine et al., 1981; Roth et al., 1985). However, in an elegant series of experiments, Lebar and colleagues demonstrated that the immunodominant target recognized by demyelinating antibodies in animals immunized with CNS tissue homogenates was a previously unknown and quantitatively minor myelin protein that they called M2 (Lebar et al., 1979). This autoantigen was characterized as a CNS myelin/oligodendrocyte glycoprotein that migrated with apparent molecular weights of 27 and 54 kDa when analyzed by SDS-PAGE (Lebar et al., 1986). Its biochemical properties distinguished M2 from the major CNS proteins, proteolipid protein (PLP), myelin basic protein (MBP), Wolfgram protein, and myelin-associated glycoprotein (MAG), which together account for at least 90% of the total proteins of isolated myelin. The biochemical properties of M2 and its tissue distribution did, however, resemble those of myelin oligodendrocyte glycoprotein (MOG), a minor CNS myelin-speciWc glycoprotein, identiWed by Western blotting and immunohistochemistry using a monoclonal antibody (mAb) raised against rat cerebellar glycoproteins (Lebar et al., 1986; Linnington et al., 1984). Studies using M2-speciWc antiserum and this MOG-speciWc mAb, 8-18C5, conWrmed that they recognized the same myelin glycoprotein (Lebar et al., 1986) and initiated a series of studies to investigate the demyelinating activity of mAb 8-18C5 both in vivo (Lassmann et al., 1988; Linington et al., 1988; Schluesener et al., 1987) and in vitro (Kerlero de Rosbo et al., 1990). In vivo, intravenous injection of 8-18C5 into rats with EAE induced by either active immunization with MBP (Schluesener et al., 1987), or the adoptive transfer of MBPspeciWc T cells (Lassmann et al., 1988; Linington et al., 1988) resulted in widespread demyelination and a dramatic increase in disease severity.

Myelin Biology and Disorders, Volume 1

469

Copyright 2004, Elsevier (USA). All rights reserved.

470

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

These early studies relied on the use of MOG-speciWc antibodies to investigate the biochemical and pathophysiological properties of MOG and it soon became apparent that the protein had several features that clearly diVerentiated it from the major components of CNS myelin. Unlike MBP and PLP, MOG was preferentially localized at the outermost surface of the myelin/oligodendrocyte continuum membrane (Brunner et al., 1989). Moreover, in a study using polyclonal MOG/M2-speciWc antisera the protein was shown to be a mammalian specialization that is absent in myelin isolated from Wsh, amphibians, and reptiles (Birling et al., 1993). Once again, this is in contrast to MBP and PLP, which are components of CNS myelin in all vertebrates from Wsh to mammals (Kitagawa et al., 1993; Linington and Waehneldt, 1990; Matthieu et al., 1986; Schliess and StoVel, 1991; Waehneldt et al., 1986). Detailed molecular and pathophysiological studies on MOG, however, had to wait until molecular biological studies provided full length MOG cDNA sequences (Gardinier et al., 1992; Pham-Dinh et al., 1993).

THE MOG GENE Cloning of MOG cDNAs and Gene Two diVerent strategies were used to clone MOG mRNAs. Gardinier and Matthieu used a mixture of MOG mAbs (Linnington et al., 1984; Piddlesden et al., 1991) to screen a rat lgt11 expression library, whereas Pham-Dinh et al. used degenerate primers developed from the N-terminal amino acid sequence of bovine MOG (Birling et al., 1993) to screen bovine, rat, and mouse brain lgt10 phage libraries. Subsequently, the mouse gene was isolated and its organization characterized (Daubas et al., 1994). Similar approaches were used to characterize the human MOG cDNA (Pham-Dinh et al., 1994) and the human gene (Hilton et al., 1995; Pham-Dinh et al., 1995a; Roth et al., 1995b). Sequence analyses indicate that the mature protein is highly conserved between species and consists of 218 amino acids comprising a putative signal peptide, an immunoglobulin (Ig)-like N-terminal domain (residues 1-121), two highly hydrophobic segments (residues 122-150 and 175-200) separated by a short hydrophilic loop, and a short hydrophilic Cterminal domain (residues 201-218) (Fig. 18.1). The presence of two putative membranespanning domains, if veriWed, would make MOG a unique member of the Ig super gene family (IgSF). Moreover, the N-terminal Ig-like domain exhibits a high level of sequence and structural homology with the N-terminal Ig-like domains of two nonmyelin proteins (Gardinier et al., 1992): butyrophilin (BT) (Jack and Mather, 1990), a major component of the milk fat globule membrane that is expressed exclusively in the lactating mammary gland, and the B-G antigens of the chicken major histocompatibility complex (MHC) (Miller et al., 1991) (Figs. 18.2 and 18.3).

Comparison of MOG Gene Organization in Human and Mouse The organization of human and rodent MOG is very similar (Daubas et al., 1994; PhamDinh et al., 1995a; Roth et al., 1995b). Eight exons encode the diVerent domains that make up the major 218 amino acid isoform of MOG. Exon I encodes the putative signal peptide, exon II encodes the N-terminal Ig-like domain, and exon III encodes a hydrophobic sequence assumed to be a transmembrane domain. The following two exons IV and V are very short and together with exons VII and the coding region of exon VIII (VIII A in humans) encode putative cytoplasmic domains of MOG, whereas exon VI (or VI B in humans) encodes a second hydrophobic, potentially membrane spanning, domain. The sizes of exons II to VII, as well as the coding region of exon VIII (VIII A in humans), are identical in mouse and humans (Pham-Dinh et al., 1995a) (Fig. 18.4). The most striking diVerence between mouse and human MOG genes being the size of intron 2 and the presence of an unexpected number of repetitive elements in the human gene (Pham-Dinh et al., 1995a; Roth et al., 1995b).

III. THE MYELIN GENES AND PRODUCTS

471

THE MOG GENE

1 GQFRVIGPRH ---------Q --------G--------G--------GY ---S----A-

PIRALVGDEV ---------------------------A ---------A --Q------I

ELPCRISPGK -----------------------------------------L----

35-55 NATGMEVGWY RPPFSRVVHL ---------- ------------------- ------------------- -S----------------- -S----------------- ----------

60 YRNGKDQDGD ----R----E --------EE --------AE --------AE -----I---E

Human Macaca Bovine Rat Mouse Marsupial

61 QAPEYRGRTE ------------------Q ---------------------------V

LLKDAIGEGK ------------ET-------ES-------ET-S-----EG-----

VTLRIRNVRF -------------------A---Q--------Q---MV-K-QH---

SDEGGFTCFF -----------------------Y--------Y--------Y----

RDHSYQEEAA ----------------------------------------------

120 MELKVEDPFY I-----------------V---------------------------

Human Macaca Bovine Rat Mouse Marsupial

121 Transmembrane domain (HR1) WVSPGVLVLL AVLPVLLLQI TVGLVFLCLQ ----A----- ---------- ----------IN------I ---------- ----------IN----A-I -LV-M----V S------F--------T-I -LV-TI---V S------F--IN--M---I ----I-I--I -I--G--YM-

YRLRGKLRAE ---------R------W-H--------H--------H------Q--

170 IENLHRTFDP ------------------V--------V------------------

Human Macaca Bovine Rat Mouse Marsupial

171 Membrane associated domain (HR2) Human HFLRVPCWKI TLFVIVPVLG PLVALIICYN WLHRRLAGQF Macaca ---------- ---------- ---------- ---------Bovine ---M------ ---------- ---------- ---------Rat ---------- ---------- ---------- ---------Mouse ---------- ---------- ---------- ---------Marsupial ---------- ---------- --I-M----- ----------

218 LEELRNPF --------------------------------KY--

FIGURE 18.1 Phylogenetic comparison of the mature MOG sequence across species. Human, bovine, rat, mouse (Gardinier et al., 1992; Hilton et al., 1995; PhamDinh et al., 1993, 1994), marsupial (Johns and Bernard 1999), and macaque (Delarasse, Pham-Dinh, and Genain, unpublished data). MOG amino acid sequences are aligned. The Ig-like domain comprises the 1-121 N-terminal amino acids. The immunodominant 35-55 peptide is underlined. Both highly hydrophobic regions (HR1, amino acids 122 to 150; and HR2, amino acids 175 to 200) are double underlined. Only the HR1 being a true transmembrane domain (KroepX et al., 1996; della Gaspera et al., 1998).

V motif f - Q - - - - - - - - - - - - - - G - - h Ð f - C - - - - - - - - - - - - - - W h p p - - - - - - - - - - - - - - L - - B-G

- Q I T V V A P S L R V T A I V G Q D V V L R C H L S P C K D V R N S D I R W I Q Q R S S R L V H H Y R N G V D L G - -

MOG

G Q F R V I G P G H P I R A L V G D E V E L P C R I S P G K N A T G M E V G W Y R P P F S R V V H L Y R N G K D Q D E E

BT

A P F D V I G P Q E P I L A V V G E D A E L P C R L S P N V S A K G M E L R W R R E K V S P A V F V S R E G Q E Q E G E

- - - - - - s R h - - - - - - - - - - p h Ð L Ð I - - - - - - a a a Ð Y h C s - - - - - - - - - - - - o G Ð G T p f

B-G

Q M E E Y K G R T E L L R D G L S D G N L D L R I T A V T S S D S G S Y S C A V Q D G D A Y A E A V V N L E V S D P

MOG

Q A P E Y R G R T Q L L K E T I G E G K V T L R I R N V R F S D E G G F T C F F R D H S Y Q E E A A M E L K V E D P

BT

E M A E Y R G R V S L V E D H I A E G S Y A V R I Q E V K A S D D G E Y R C F F R Q D E N Y E E A I V H L K V A A L

FIGURE 18.2 Sequence comparison between Ig-V domain from chicken BG with bovine MOG and BT. Alignments of the extracellular region with the Ig V motif. Lowercase letters identify the amino acids with functional or physical properties as follows: a, acidic (D, E); f, aliphatic (L, I, V); h, hydrophobic (L, I, V, M, Y, F); o, aromatic (Y, F, W); p, polar (K, R, H, D, E, Q, N, T, S); s, small (A, G, S, T, V, N, D). The positions of matches in the three sequences with the V motifs are underlined. Amino acid identities among the three proteins are highlighted.

III. THE MYELIN GENES AND PRODUCTS

472

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

MOG

BG

N

BT

N

N

S

S

CHO

Ig V

S S

CHO CHO

Ig V

Ig V

S

TM

S

TM

HR2

Ig C

S S

TM

C

C

B 30

au14

C

au15

FIGURE 18.3 Structural homology between MOG, BT, and BG. MOG contains a single Ig-V extracellular domain, as well as a transmembrane domain followed by a short cytoplasmic loop, a highly hydrophobic region (HR2) that behaves as semi-embedded in the membrane bilayer on the cytosolic side of the membrane; the C-terminal region ends with a short intracellular tail (KroepX and Gardinier, 1996; Della Gaspera et al., 1998). B-G (uniquely expressed in avian) comprises a single IgV-like extracellular domain, a transmembrane domain, followed by a number of heptad repeat regions of diverse length in the cytoplasmic compartment (according to variation in alternative splicing) (Kaufman et al., 1990; Kaufman and Salomonsen, 1992; Miller et al., 1991). BT (initially designated BTN or BTF) (Jack and Mather, 1990) contains an extracellular domain comprising two Ig-like folds (a distal IgV-like domain and a proximal IgC-like domain) as well as a transmembrane domain and intracellular domain composed of heptads followed by the C-terminal B30.1 globular domain of about 130 amino acids, which is conserved in a family of zinc-Wnger proteins (Henry et al., 1998; Rhodes et al., 2001). Both BT IgV- and IgC-domains bear one glycosylation site, as indicated.

The MOG gene is highly polymorphic. The mouse gene contains a polymorphic microsatellite in intron 2 composed of multiple copies of the dinucleotide 5’-GA-3’ on the sense strand, near the 3’ end of the intron. Analysis of this region by PCR in diVerent mouse strains revealed that the number of GA/CT repeats varied from six copies in SJL/J and 129 DNA to 19 in B10.S. This polymorphism could be used to obtain information about the mouse MOG locus (Daubas et al., 1994). This GA/CT polymorphic repeat is also present in rat MOG, but is not found in the human gene. Three intragenic CA/GT dinucleotide repeats (MOGa, MOGb, and MOGc) are, however, present in the human. MOGa is au16 located in the second intron, whereas MOGb and MOGc are both located upstream of the MOG transcription start site. These repeats were shown to be polymorphic and have provided valuable tools in genetic mapping and population studies (Roth et al., 1995a, 1995b). A second polymorphic repeat used in genetic studies introduces a sequence of either Wve or six leucines into the signal peptide, indicating the presence of a trinucleotidelength polymorphism (Malfroy et al., 1995; Pham-Dinh et al., 1994). However, the most

III. THE MYELIN GENES AND PRODUCTS

THE MOG GENE

FIGURE 18.4 Structure comparison between mouse and human MOG genes. The protein domain structure of the MOG gene correlates with exon structure at the genomic level, which is a characteristic feature of IgSF genes. Exon I encodes a putative signal sequence; exon II the IgV-like domain; exon III the transmembrane region (also designated hydrophobic region 1, HR1, according to the presence of a second highly hydrophobic domain in the MOG protein, HR2 (designated according to Pham-Dinh et al., 1995a). The two short exons IV and V, formed of 21 bases, separate the transmembrane-encoding exon III from exon VI (VI B in humans) encoding HR2; the coding region of the gene ends within exon VIII, using two diVerent potential acceptor sites in the human gene (Ballenthin and Gardinier, 1996; Pham-Dinh et al., 1995a), not found in the mouse gene.

unexpected characteristic of human MOG is the presence of a very high level of Alu repetitive elements in diVerent parts of the gene. As many as fourteen Alu repetitive elements each longer than 270 bases are concentrated within three introns: two are inserted in intron 1, nine in intron 2, and three in intron 5 (Roth et al., 1995b). An additional Alu sequence was also found inserted in the 3’-untranslated region (Pham-Dinh et al., 1994). The presence of the nine Alu elements in intron 2 of human MOG accounts to a large extent for its far larger size than intron 2 in the mouse, 6500 bp as opposed to 2600 (Daubas et al., 1994). The orientation of the Alu sequences relative to the sense of transcription of MOG diVers according to their location in the gene. The majority of repetitive elements located in the 5’ region are diVerently orientated compared to those within the 3’ region of MOG. The Alu repetitive elements in MOG belong to diVerent Alu subfamilies (Jurka and Milosavljevic, 1991; Jurka et al., 1992), suggesting that they were inserted into the gene at diVerent times during primate evolution.

Alternative Splicing Forms The cloning of human MOG cDNA and gene allowed to characterize more than 12 mRNA variants (Ballenthin and Gardinier, 1996; Pham-Dinh et al., 1995a). The intron/exon organization of the human MOG gene supported the hypothesis of a complex alternative splicing pattern arising from processing of a single MOG primary transcript. In addition to the eight exons encoding the major membrane-bound MOG isoform, the human MOG gene also contains several previously unknown alternatively spliced coding exons. One of these new MOG sequences is part of an Alu element, located in intron 2, introducing an inframe premature stop codon in a transcript that would encode a secreted MOG Ig-like domain (Ballenthin and Gardinier, 1996), while predicted proteins encoded by mRNA with exon II spliced out would lack the extracellular Ig-like domain, and even may be constituted of short MOG-unrelated peptides (Delarasse, Genain, and Pham-Dinh, unpublished data). Moreover, alternative use of two acceptor splicing sites in the 3’-terminal exon could encode two diVerent C-termini (Fig. 18.5). To extend these observations, we investigated alternative splicing of the MOG gene in a number of other species by RT-PCR. Analysis of splicing variants in mouse, bovine, marmoset, and macaque MOG revealed at least one splicing variant in mouse, whereas higher primates show the most complex patterns (Delarasse, Pham-Dinh and Genain, unpublished data). This is not a unique feature of the MOG gene, given that most myelin proteins are expressed as families comprising several isoforms generated by alternative

III. THE MYELIN GENES AND PRODUCTS

473

474

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

B

A a1 *

out

a2 *

a3 *

in

a1

b1

a2

a3

*

out

b2 *

b3

in

*

b1

b2

b3

Ig-Alu *

Ig-Alu

FIGURE 18.5 Alternative splicing in human MOG transcripts and model of putative corresponding MOG proteins. (A) Exon composition of 14 alternative variants from the human MOG gene is presented. Two families of RNA are deWned; a and b, based on the use of one of two alternative acceptor splice sites in the 3’-terminal part of the gene.
and þ indicate alternative in-frame stop codons. (B) Schematic topological models of the putative MOG protein isoforms. Note that at present there is no evidence for the translation of any of these alternative mRNA variants. Adapted from Ballenthin and Gardinier, 1996; Pham-Dinh et al., 1995a; and unpublished data from Delarasse, Pham-Dinh, and Genain.

splicing (Baumann and Pham-Dinh, 2001; Campagnoni, 1988). Alternative splicing events are well known to generate protein diversity in the proteome, particularly in the case of surface receptors (Modrek et al., 2001). Alternative splicing of the human MOG primary transcript could generate a number of diVerent products including a soluble MOG Ig-like domain, single transmembrane alpha-helices with variable-length cytoplasmic tails, or even short MOG-unrelated peptides. These additional forms of MOG could be functionally active, in particular in the context of the induction of self-tolerance and autoimmune demyelination, but at present there is no evidence for the translation of any of these alternative MOG mRNA transcripts. However, the low abundance of these alternative transcripts may make it diYcult to detect their protein products (Ballenthin and Gardinier, 1996; Pham-Dinh et al., 1995a).

Regulatory Elements in the MOG Gene The expression of MOG is one of the latest events associated with CNS myelination and several groups have attempted to identify the cis-acting MOG sequences and trans-acting nuclear factors implicated in this phase of myelination. A combination of functional analyses using reporter gene constructs and sequence analysis of consensus regulatory sites allowed a preliminary identiWcation of several potential regulatory elements in the murine gene. The impact of diVerent MOG promoter sequences on the level of reporter gene was determined in the oligodendroglial cell lines, CG4 and C6, as well as in the 3T3

III. THE MYELIN GENES AND PRODUCTS

THE MOG GENE

Wbroblast line (Solly et al., 1997). These experiments deWned the key sites for maximal and tissue-speciWc MOG expression within the MOG promoter sequence 657-bp 5’ of the transcription initiation site. Only this relatively short section of the MOG promoter is necessary to obtain relatively high-level expression of the reporter gene in vitro, but this is nevertheless three-to four-fold less than obtained using the MBP promoter under the same conditions (Solly et al., 1997). Reporter gene activity was considerably greater in the CG4 cells than in C6 and 3T3 cells. This observation is in accordance with the observation that the CG4 cells can diVerentiate and coordinately express in vitro myelin-speciWc proteins, including MOG, indicating they express those transcription factors necessary for the expression of the myelin genes (Louis et al., 1992). The proximal region of the murine MOG promoter, from -188 to -19 bp, which contains a TATA box at
50 bp and a CAAT box at
101, is required for basal and oligodendroglial–speciWc transcription of the reporter gene in CG4 cells. These elements are conserved in rat MOG (Pagany et al., 2003a) and comparison of the genomic sequence of MOG from diVerent rat strains (accession numbers: Brown Norway (BN), AY059598 – AY059605; Lewis (LEW), AY059614 – AY059621; Dark Agouti (DA), AY059606 – AY059613) identiWed multiple strain speciWc polymorphisms. Of particular interest was a BN speciWc small nucleotide polymorphism (SNP) that introduces a new HindII restriction site immediately in front of the CAAT box at -220 to 217, which is compatible with the BN associated RFLP described by Lambracht et al. (1995). When evaluated in the rat glioma cell line C6, this SNP was responsible for a reduction of 35 to 40% in reporter gene expression driven by the rat MOG promoter (
1 to –676). The TATA box is not conserved in human MOG (Roth et al., 1995b) and appears functionally irrelevant in the mouse and may explain at least in part the very low levels of MOG transcription observed in the CNS. The presence of these negative regulatory elements was demonstrated in vitro using MOG reporter gene constructs that extend beyond –657 bp to –1343 upstream of the transcription start site (Solly et al., 1997). These longer constructs inhibited the activity of the minimal MOG promoter, indicating the presence of negative regulatory sites downstream of position
657 bp (Fig. 18.6A). Two clusters of putative binding sites for negative transcriptional factors were identiWed in this region, GTX, a homeobox protein, and MyEF-2, a single-stranded DNA binding protein, which may exert a silencing eVect and be responsible for the lag of expression in vivo. Other potential binding sites for transcriptional regulatory factors can be identiWed from homologies with nuclear binding sites previously identiWed in other myelin-speciWc promoters, such as those for PLP, P0, MBP, MAG, and 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNPase) (Hudson, 1995). In summary, the 188 bp promoter appears suYcient for oligodendrocyte-speciWc expression of MOG and the initial -657 bp of the proximal MOG promoter is suYcient for maximal transcriptional activity in vitro. However, as pointed by Daubas and colleagues (Solly et al., 1997), these sequence homologies and the activity of chimeric promoter constructs in vitro do not necessarily indicate their true functional activity in vivo. To investigate transcriptional regulation of MOG in greater depth, the 657 bp region of the 5’-Xanking sequence of the mouse gene was analyzed by in vitro footprinting and electrophoretic mobility shift assays (Jaquet et al., 1999). DNA footprinting analysis was used to localize the DNA sequences that could interact with trans-acting factors present in nuclear extracts of brain and liver. These experiments identiWed three protected regions, A, B and D (Fig. 18.6B). Protection of regions A and D was identical using both brain and liver extracts, whereas region B was more strongly protected by the brain extract. Region B spans a sequence of 21 nt spanning
93 bp to –73 bp, which is identical in human and mouse MOG and which contains a sequence TGACGTGG, similar to that of the CRE (cAMP responsive element) consensus sequence TGACGTCA (Roth et al., 1995b). Antibody supershift assays using a CREB-1 speciWc mAb, as well as competition experiments with CRE consensus oligonucleotides, caused a marked reduction of nucleoprotein complex B formation, suggesting the involvement of the cAMP transduction pathway in MOG transcription.

III. THE MYELIN GENES AND PRODUCTS

475

476

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

FIGURE 18.6 (A) Deletion analysis of the mouse MOG promoter region. CG4 cells were transiently transfected with
3806 MOG-luciferase constructs and 5’ deletion clones, and luciferase activities were determined. A graphical representation of the MOG-luc plasmids is shown on the left. The luciferase gene is represented in white, and the portion of MOG exon 1 to 85 bp is hatched (from the initiation of transcription, arrow). Two control constructs in which the promoter region is in the reverse orientation are also indicated (rev). For each construct, luciferase activity, normalized to the amount of protein, has been indicated as the percentage of the value obtained with the positive control pMSF-luc, deWned as 100%. The SEM is shown in parentheses. Adapted from Solly et al., 1997. (B) Transcriptional activity of mutated regions in the
657 bp murine MOG promoter sequence. The various constructs were transiently transfected into the OLN-93 cell line. At 24 hours after transfection, luciferase activity was measured. Data are mean þ/
SEM values for at least Wve independent experiments and are expressed relative to the luciferase activity of the plasmid control pluc657. Adapted from Jacquet et al., 1999.

III. THE MYELIN GENES AND PRODUCTS

477

THE MOG GENE

MOG Gene Is Located in the Telomeric Part of the MHC The striking sequence homology of the N-terminal domains of MOG, BT, and the chicken B-G antigen prompted the chromosomal localization of MOG and BT by radio-active in situ hybridization on human and murine metaphase chromosomes. Human MOG was located in p21.3-p22 on chromosome 6 close to human BT (Vernet et al., 1993), while murine MOG was identiWed within band C of mouse chromosome 17 (Pham-Dinh et al., 1993). In both species, these regions correspond to the location of MHC locus, which contains a large cluster of genes with important immunological functions. The location of these genes was further reWned using yeast artiWcial chromosomes (YAC) containing MOG, an approach that demonstrated that human MOG lies 60 kb telomeric to HLA-F (Fig. 18.7 A) in a head-to-head orientation; whereas the mouse gene lies 25 kb telomeric to H2-M5 in a tail-to-head orientation (Jones et al., 1995; Pham-Dinh et al., 1995b). MOG is also located within the telomeric region of the rat genome in the nonclassical class I RT1.M region of the rat MHC (Lambracht et al., 1995).

au2 au3

MOG—A Member of the BT-like Gene Family Sequence analysis of MOG revealed that its N-terminal Ig-like domain shares an amino acid sequence homology of approximately 50% with a structurally similar N-terminal Ig-like domain of the milk protein butyrophilin (BT) as well as the B-G antigens of the chicken MHC locus (Kaufman and Salomonsen, 1992); proWle search analysis indicated also an evolutionary relationship to B-7.1 and B-7.2 costimulary receptors (Linsley et al., 1994). These molecules, expressed on APC, interact with both CTLA4 and CD28, providing the essential second signal to T cells. BT and a cluster of six structurally related BT-like genes (BT2 and BT3 gene subfamilies according to Pontarotti and colleagues’ nomenclature, 1999) are located telomeric of MOG on chromosome 6 in man, in the extended MHC class I region au4 (Henry et al., 1999; Rhodes et al., 2001; Vernet et al., 1993) (Fig. 18.7 A). The general structure of the members of this gene family comprises a signal peptide followed by an Ig variable (V)-like domain, an Ig constant (C)-like domain, a hydrophobic membrane spanning domain, and Wnally a C-terminal cytoplasmic region containing a B30-2 domain. Recently an additional member of this gene family, BTL-II (designated BT4 according to Pontarotti and colleagues’ nomenclature, 1999), was identiWed at the junction of the class II and class III regions of the MHC telomeric of the HLA-DR locus (Stammers et al., 2000). Other BT-like genes have been identiWed as well on other chromosomes of both human and mouse (Shibui et al., 1999; Ye et al., 2000). The protein domain structure of BT-like genes, including MOG, correlates with exon structure at the genomic level, which is a hallmark of IgSF genes. The homology between MOG and the butyrophilin is restricted to the single exon encoding the N-terminal IgV-like domain. There is also signiWcant homology with similar domains of the B-G genes of the chicken blood group system (Miller et al., 1991) (Fig. 18.7B), and the avian genes bg2 (Jarvi et al., 1999) and Zipper (Bikle et al., 1996). This raises the possibility that these homologous au5 N-terminal domains are derived from a common ancestral gene, by duplication and exon shuZing events (Henry et al., 1999), a process resulting in the evolution of new genes with potentially very diVerent functions (Kolkman and Stemmer, 2001). The function of these MOG-related gene products is currently poorly understood. They all appear to be type I transmembrane proteins and the presence of Ig-like domains suggests that they are involved in protein-protein or protein-ligand interactions. BT itself is only expressed in the lactating mammary gland where it is believed to play an important role in stabilizing milk fat globules. BT-like genes located in the murine BTL2 locus are expressed in the gastro-intestinal tract, as is Zipper in chickens. BG genes play an important role in the avian immune system, as potent adjuvant of immunity (Kaufman and Salomonsen, 1992); and in mammals at least one member of the BT gene family is expressed in antigen presenting cells, and on T- and B-cell lines (Greenbaum and Zhuang, 2002; Henry et al., 1997) leading to the speculation that they may be functional members of the extended B7 family; A BT-like gene has recently been found expressed as well in erythroid lineage (Xu et al., 2001). These observations together

III. THE MYELIN GENES AND PRODUCTS

478

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

FIGURE 18.7 (A) Localization of the MOG gene and BT gene families within the human MHC. A schematic representation of the genomic regions containing the class I, class II, and class III genes in the human MHC is depicted. Some HLA (human leukocyte antigen) genes are indicated as representative of the classic MHC genes in class I and II regions. The three regions code for MHC classic (HLA) and nonclassic molecules. The MOG gene is colored in red; the genes of the BT family are in blue/violet. MOG is located 60 kb telomeric to HLA-F, telomeric of the MHC class I nonclassical genes (PhamDinh et al., 1995b); BT (butyrophilin) and the other members of the BT gene family (BT2 and BT3 subfamilies) are located at about 2 Mb in the extended class I MHC region (Rhodes et al., 2001; Ruddy et al., 1997; Vernet et al., 1993). Another related BT-like gene, designated BT4 by Pontarotti and colleagues (1999) (BTLII gene in Stammers et al., 2000), has also been recently located at the border of the MHC class II and III regions in human and mouse (Stammers et al., 2000). (B) Domain organization and phylogenetic relationship between members of the extended B7 family. All members of the family share sequence homology in the IgV-like domain, the rest of the molecule coupled with diVerent domains. IgV domains are in blue, IgC domains in orange. Short C-ter tail of B7-1 in yellow, of MOG in red. Short domains forming heptads are represented by yellow arrowhead-like drawings. The number of heptads varies among the diVerent B-G proteins, with as many as 27 units found within the cytoplasmic tails. Heptad-encoding exons are also found in MOG and BT genes, suggesting an additional possible evolutionary linkage among members of the extended BT gene family in this region of their sequence. Abbreviations: BT, butyrophilin (diVerent abbreviations have been used according to diVerent authors, noted earlier); TM, transmembrane domain; HR2, second hydrophobic region in the membrane bilayer/cytoplasmic compartment of the MOG protein. Adapted from Henry et al., 1999.

III. THE MYELIN GENES AND PRODUCTS

BIOCHEMISTRY AND STRUCTURAL ORGANIZATION OF MOG

with structural similarities between the extracellular domains of the members of the BT gene family and the co-stimulatory molecules CD80 and CD86 (B7.1 and B7.2, respectively) (Linsley et al., 1994) have led to speculation that some BT-like gene products may play a role in the adaptive immune response.

BIOCHEMISTRY AND STRUCTURAL ORGANIZATION OF MOG MOG was identiWed initially as a minor myelin glycoprotein, but the Wne details of its biochemistry only became apparent once the protein had been cloned and its predicted amino acid sequence became available. On SDS-PAGE MOG often migrates as a 26 – 28 kDa doublet together with a higher molecular weight species of approximately 52 – 57 kDa. These components share the same N-terminal sequence, and it appears that the higher molecular weight species is a MOG dimer (Amiguet et al., 1992). In vitro deglycosylation using N-glycanase resolved the 26 – 28 kDa doublet into a single 25 kDa band, indicating that these diVerent molecular forms are due to diVerent degrees of glycosylation. Treatment with speciWc glycosidases conWrmed that MOG was N-glycosylated (Amiguet et al., 1992), an observation in agreement with the presence of only a single N-glycosylation consensus sequence within the predicted amino acid sequence of the mature protein. Human MOG was found later to contain the L2/HNK-1 carbohydrate epitope (Burger et al., 1993), a sulphated glucuronic acid linked to lactosaminyl residues implicated in the functional activity of cell adhesion molecules (GriYth et al., 1992; Quarles, 1997). This glycosylation site is located within the N terminal Ig-like domain of MOG suggesting that this region of the protein should be exposed at the extracellular surface of the myelin membrane. The precise topology of MOG in the membrane was, however, uncertain as hydrophobicity plots identiWed two highly hydrophobic regions (designated as HR1 and HR2, see Fig. 18.1) that could be arranged as a-helical transmembrane domains. If both domains were membrane spanning then MOG would adopt a Ushaped topology across the membrane with both the N- and C-termini of protein exposed at the extracellular surface. (Gardinier et al., 1992; Pham-Dinh et al., 1993, 1995a). This structure is at odds with either the single membrane-spanning domain or glycophospholipid-anchor found in other Ig domain containing membrane proteins (Lindberg et al., 1993; Williams and Barclay, 1988). The presence of three proline residues in the putative HR2 a-helical region argues against this as they would destabilize any interaction within the bilayer (Nilsson et al., 1998). Moreover, the C-terminal of a stretch of basic residues (RLRGK in MOG) is atypical for the extracellular end of a transmembrane segment (Sipos and von Heijne, 1993). The topological organization of MOG was Wnally determined using a combination of immunocytochemical and biochemical approaches. Staining cultures of MOG transfected cell lines and oligodendrocytes with antisera speciWc for MOG1–125, MOG154–169, or MOG198–218 clearly localized the N-terminal Ig-like domain of MOG at the extracellular face of the plasma membrane. However, unlike MOG1–125 speciWc antibodies, which stained the surface of live cells, staining with antisera speciWc for either MOG154–169 or MOG198–218 was only seen after transfected cells were permeabilized allowing antibodies to enter the cytoplasmic compartment (KroepX et al., 1996). These observations suggest that MOG is a type I transmembrane protein in which the second hydrophobic domain interacts with the cytoplasmic face of the membrane. This was conWrmed in a second study, which deWned the structural organization of the second hydrophobic domain of MOG using a combination of transcription and translation-coupled translocation systems within microsomal membranes in vitro (della Gaspera et al., 1998). Protease protection assays and site directed mutagenesis provided deWnitive evidence that while the Wrst hydrophobic region of MOG spans the membrane, the second hydrophobic region is partially buried in the lipid bilayer to generate two short cytoplasmic domains, a short hydrophilic loop and a short C-terminal tail (della Gaspera et al., 1998) (Fig. 18.3). The amino sequence of the second hydrophobic domain and its Xanking hydrophilic residues (amino acids 152 to 218) are nearly 100% identical between human and rodents

III. THE MYELIN GENES AND PRODUCTS

479

480

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

(Fig. 18.1), indicating that conservation of its structure is necessary to maintain the function of this domain. Long-term treatment with antibody directed against the extracellular domain of MOG results in the depolymerization of microtubular veins in the membrane sheets extended by oligodendrocytes in culture (Dyer, 1993; Dyer and Benjamins, 1990; Dyer and Matthieu, 1994), suggesting that signal transduction via the cytoplasmic domain of MOG may inXuence microtubular organization. A second potential function embedded within this domain of MOG may be related to its preferential sequestration in the outermost surface of the myelin membrane (Brunner et al., 1989). The molecular signals necessary to retain MOG at this speciWc location are unknown, but analysis of MDCK cells transfected with MOG or C-terminal truncated mutants identiWed the presence of a peptide sequence that drives MOG to sort exclusively to the MDCK cell basolateral membrane (KroepX and Gardinier, 2001a). Removal of the second hydrophobic domain from MOG resulted in the complete loss of basolateral sorting and speciWc targeting to the apical membrane.

THEORETICAL MODEL OF THE EXTRACELLULAR DOMAIN OF MOG In view of the importance of MOG as a target for autoantibody-mediated demyelination, it is clearly important to determine the structure of its extracellular domain. The sequence of the extracellular domain of MOG deWnes it as a member of the IgSF (Gardinier et al., 1992; Pham-Dinh et al., 1993), but the high resolution crystal structure of the MOG Ig domain (MOGIgd) is not yet available1. Two groups, however, modeled the structure of the extracellular domain of MOG using data obtained from high resolution structural analysis of other IgSF family members (Hjelmstrom et al., 1998; Mesleh et al., 2002). The Wrst approach used data based on the structure of the Wrst domain of CD4, the antibody variable light region of 28B4, and the antibody light region of D1.3, to model the rodent protein (Hjelmstrom et al., 1998). In a second study (Mesleh et al., 2002), a homology model of the human MOG Ig-like domain was built on the analysis of sequence-structure alignments with the high-resolution crystal structure of the myelin P0 Ig-like domain (Shapiro et al., 1996) with which the MOG Ig-like domain shares 35% and 55% of predicted structural homology. Structural organization can only be predicted with any accuracy for those regions involving high levels of sequence/structural homology and this limits the usefulness of the two models. However, both models predict that MOG can adopt a classic IgSF fold forming the so-called Ig ‘‘Greek key’’ motif similar to that seen in IgV domains. Ig domains are highly conserved structurally despite low levels of amino acid sequence homologies between many members of this family of proteins. The structure of the Ig fold is thought to be conserved because it is stable to proteolysis, a characteristic essential for extracellular molecules. The fold contains 70 to 110 amino acid residues in two tightly packed b-sheets, each consisting of four to six antiparallel b-strands containing Wve to ten amino acids juxtaposed to form a sandwich (Harpaz and Chothia, 1994; Williams and Barclay, 1988;). The region between the sheets is hydrophobic, created by the projection of hydrophobic amino acids in this interior region. The hydrophobic amino acids alternate along the b strands with hydrophilic amino acids, which project outward. This structure is stabilized by a disulWde bond between conserved cysteine residues, residues Cys24, and Cys98 in MOG. The accuracy of these models is probably high, as is shown by the localization of the protein’s single glycosylation site (Gardinier et al., 1992) (Fig. 18.8A) and its potential complement-binding domain (Johns and Bernard, 1997) (Fig. 18.8C). As would be anticipated from the accessibility of these domains in experimental studies, the models

1

Note added in proof: The crystal structures of the extracellular Ig domains of rat and mouse MOG (Breithaupt et al., 2003; Clements et al., 2003) and the complex rat protein with the antigen-binding fragment of a demyelinating MOG-specific monoclonal antibody were published while this article was in preparation.

III. THE MYELIN GENES AND PRODUCTS

III. THE MYELIN GENES AND PRODUCTS

N

C⬙

CDR2 C

D

C⬘

N31

E

F

B

CDR3

S−S

W39

D−E

CDR1

N

A⬘

G

A

E107 E108

R52

H49

Y40

Y105 V47

F99

D102

N R101

E36

N

R46

C

Three-dimensional homology model of the Ig-like domain from mouse MOG. The location of the glycosylation site (A) and C1q binding region (C) are indicated in red; the locations of the encephalitogenic peptide, residues 35-55 are colored in dark gray (B). A putative MOG dimer structure is shown (B and C). Adapted from Hjelmstrom et al., 1998; see also Mesleh et al., 2002.

FIGURE 18.8

B

N

A

THEORETICAL MODEL OF THE EXTRACELLULAR DOMAIN OF MOG

481

482

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

locate these regions of MOG at the protein’s surface, rather than being buried within its core (Fig. 18.8). Many reports suggest that MOG dimerizes in vitro (Abo et al., 1993; Amiguet et al., 1992; Birling et al., 1993; Gardinier et al., 1992; Slavin et al., 1997). The models provide some insight into the structural basis for this putative phenomenon suggesting the interface between MOG monomers consists of a region of several hydrophobic and aromatic residues that are conserved in other IgSF proteins (Figs. 18.8B and 18.C).

PATHOPHYSIOLOGICAL PROPERTIES OF MOG MOG is the only CNS myelin autoantigen known to initiate both a demyelinating autoantibody response and an encephalitogenic T cell response in EAE. The ability of MOG to induce a demyelinating autoantibody response reXects its topological organization in the myelin membrane. Unlike the majority of myelin proteins, MOG is sequestered at the outermost surface of the myelin sheath where its single Ig-like domain (MOGIgd) can bind antibodies present in the extracellular milieu (Brunner et al., 1989). Availability of the target antigen at the myelin surface is a prerequisite for an autoantibody response to initiate antibody-mediated demyelination, and this property sets MOG apart from the major structural components of the myelin sheath. Proteins associated with the cytoplasmic surface of myelin such as MBP, buried within compact multilamellar regions of the sheath (e.g., PLP), or sequestered within paranodal myelin/axonal junctions are not normally accessible to antibody. The repertoire of autoantibody responses that can initiate primary demyelination is therefore limited to a relatively small set of myelin proteins, MOG being the only one that has been described to date. In contrast, the number of autoantigens that can induce a CD4þ T cell mediated inXammatory response in the CNS is very large and this is not restricted to components of the myelin sheath. This is because encephalitogenic CD4þ T cells do not recognize their target peptides on the myelin sheath or oligodendrocyte, but bind to MHC class II molecules on the surface of antigen presenting cells. Oligodendrocytes and neurons do not express class II MHC molecules in vivo. In the CNS antigen, presentation to CD4þ T cells is performed by microglia, which takes up and presents peptides derived from myelin/ oligodendrocytes, astrocytes and neurons during the normal turnover of the CNS. Activation of autoreactive T cells following recognition of their target peptide on the surface of microglia initiates a cascade of responses that result in blood brain dysfunction, recruitment of T cells and macrophages into the CNS, and ultimately neurological disease. This response can be initiated by T cells speciWc not only for myelin autoantigens (MBP, PLP, MOBP, MOG, MAG, oligodendrocyte-speciWc glycoprotein [OSP], and CNPase) (Bradl and Linington, 1996; Holz et al., 2000; Maatta et al., 1998; Morris-Downes et al., 2002; Stevens et al., 1999; Weerth et al., 1999), but also for astrocyte-derived autoantigens such as S100ß and GFAP (Bradl and Linington, 1996; Kojima et a., 1994) and the neuronal type I cannabinoid receptor (Proescholdt et al., 2002). Irrespective of the identity of the autoantigen that initiates the inXammatory response in the CNS, demyelination is greatly enhanced if a MOG-speciWc autoantibody response is available to direct immune eVector mechanisms to attack the myelin sheath. Understanding the genetic and structural constraints that determine the pathogenicity of MOGspeciWc autoimmune responses is essential if we are to understand the immunopathological signiWcance of autoimmunity to MOG in MS.

The MOG-Specific Autoantibody Response The demyelinating potential of MOG-speciWc autoantibody responses is now generally accepted, but the precise identity of the epitopes recognized by this pathogenic autoantibody response is still unknown. Analysis of a panel of demyelinating MOGIgd-speciWc mouse mAbs using overlapping linear synthetic MOGIgd peptides in a solid phase ELISA

III. THE MYELIN GENES AND PRODUCTS

PATHOPHYSIOLOGICAL PROPERTIES OF MOG

au17

au6

au7 au18

identiWed two distinct antibody subsets in terms of their epitope speciWcity. Six mAbs only recognize the intact protein indicating that the epitope(s) involved were strictly conformation dependent. The remaining four mAbs not only bound to MOGIgd but also to synthetic peptides spanning amino acids 63 to 100 of the mature protein (Brehm et al., 1999). Regardless of their Wne speciWcity, the MOGIgd epitopes recognized by these mAbs are stable to treatment with SDS as demonstrated by Western blotting experiments, but do not bind to an N-truncated mutant protein, MOG51–120. These observations indicate that the immunodominant epitopes of MOG targeted by the pathogenic autoantibody response in mice are either totally or in part conformation-dependent. A recent study using a combinatorial IgG-Fab library derived from MOG-immunized marmosets indicates that conformation dependent MOG epitopes are also immunodominant in this species (von Budingen et al., 2001). The precise identiWcation of the relevant conformational epitopes will, however, require that we determine the crystal structure of MOG complexed with MOG-speciWc antibodies or their Fab fragments. Intriguingly, DNA vaccination using constructs encoding the full length mature protein results in a MOG-speciWc autoantibody response that is purely conformation-dependent and highly pathogenic in vivo (Bourquin et al., 2000). In contrast, active immunization with recombinant MOGIgd generates a more complex antibody response that recognizes both conformational and linear epitopes in rodents (Adelmann et al., 1995; Brehm et al., 1999; Litzenburger et al., 1998) and primates (Brok et al., 2000; Kerlero de Rosbo et al., 2000). This could be due to the fact that the bacterial product is isolated under denaturing conditions and in addition is not glycosylated. These factors, in addition to the diVerent routes of sensitization, may result in the recombinant protein being processed diVerently resulting in a diVerent spectrum of antibody speciWcities. Currently, whether or not antibody responses to linear MOG peptide epitopes are pathogenic is controversial. MOG-peptide speciWc antibodies cannot bind to native MOGIgd expressed at the surface of MOG transfected target cells (Bourquin et al., 2000; Brehm et al., 1999) indicating that they are unlikely to mediate demyelination in vivo. However, demyelination is reported in some rats with MOG peptide-induced EAE (Ichikawa et al., 1996; Johns et al., 1995; Weissert, personal communication) raising the possibility that part of the MOG peptide-speciWc antibody repertoire is pathogenic. However, it has still to be formally established whether demyelination in this case is mediated by antibodies to the initiating peptides or is due to a spreading of the B cell response to target other, possibly conformation-dependent MOG epitopes.

Encephalitogenic T Cell Responses to MOG The identity of the MOG epitopes that induce an encephalitogenic T cell response in any particular individual is largely determined by genes within the MHC (SteVerl et al., 1999; Weissert et al., 1998), although other non-MHC ‘‘background genes’’ modulate disease penetrance and pathology (Dahlman et al., 1998, 1999; Olsson et al., 2000). Encephalitogenic MOG T cell epitopes are now known for many diVerent rat and mouse strains, as well as certain primate MHC haplotypes (Tab. 18.1). The phenotype of encephalitogenic MOG-speciWc T cell lines is the same as described for MBP-speciWc encephalitogenic T cell lines and clones (Linington et al., 1993). They are a/b T cell receptor, CD4þ, IL-2Rþ, CD8
, CD45R0
, and MHC class II restricted T cells of the Th1 T cell subset that secrete IFNg when stimulated in vitro (Adelmann et al., 1995; Linington et al., 1993; SteVerl et al., 1999). There is, however, an increasing body of evidence indicating that under certain circumstances neuroantigen-speciWc CD8þ MHC class I restricted T cells can also induce EAE in the mouse, including one report of disease induced by the adoptive transfer MOG35-55 speciWc CD8þ T cells in C57/BL6 mice (Sun et al., 2001). The pathogenic potential of MOG speciWc T cells is strain dependent and is not necessarily controlled by the same genetic loci that determine the encephalitogenic potential of MBPspeciWc T cell repertoire. Lewis rats (MHC RT1l) are highly susceptible to EAE induced by the adoptive transfer of MBP-speciWc T cells, but the adoptive transfer of MOG-speciWc T cells results in only a minimal clinical deWcit in this strain (Linington et al., 1993). In contrast,

III. THE MYELIN GENES AND PRODUCTS

483

484

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

TABLE 18.1 Species

MOG T Cell Epitopes in Rodents and Primates

Epitope

Comments

Rat Lewis

1-20 35-55#

Restriction element RT1.Bl, both are epitopes weakly encephalitogenic by T cell transfer. No disease following active immunization (Adelmann et al., 1995).

Dark Agouti

74-93

Restriction element RT1.Bav1, both epitopes are highly encephalitogenic by T cell transfer, but only 93-107 induces EAE following active immunization. CD4þ T cells speciWc for peptide 74-93 cross react with a homologous sequence in bovine butyrophilin (SteVerl et al., 2000).

ACI

93-107

LEW1A Restriction element RT1.Bn, minimal CNS pathology and disease activity after T cell transfer (SteVerl et al., 1999). Peptide 93-107 is a cryptic encephalitognic peptide for LEW.1A.

Brown Norway

21-38

LEW. 1N

(93-107)

Lewis. 1W

1-26 39-58 62-79

Complex polyclonal response, transfer of polyclonal T cell lines recognising all three epitopes induces low grade CNS inXammation, but no clinical disease (unpublished observations).

C57B6/J

35-55

Mendel et al., 1995; Peptide 35-55 induced EAE is highly reproducible in this strain.

Biozzi

35-55

Amor et al., 1994.

PL/J

35-55

Kerlero de Rosbo et al., 1995.

DBA/1

79-96

This strain is highly susceptible to MOG and peptide 79-96 induced EAE Abdul-Majid et al., 2000.

SJL/J

1-20, # 92-106

Amor et al., 1994; Devaux et al., 1997; Tsunoda et al., 2000. Note that only T cells speciWc for peptide 92-106 are pathogenic by adoptive transfer.

Rhesus

4-20, 35-50, 94-116

Dominant epitopes identiWed by proliferation assays on diVerent MHC backgrounds (Kerlero de Rosbo et al., 2000).

Marmoset

14-36

Peptide 14-36 is encephalitogenic in all common marmosets as it is presented by a common monomorphic class II MHC molecule (Brok et al., 2000).

Human

1-22, 34-56 64-96

Three epitopes identiWed as immunodominant in a cohort of MS patients (Kerlero de R). Responses to peptide 63-87 are preferentially enhanced in DR2(15)þ MS patients (Wallstrom et al., 1998). Responses to peptide 63-87 are preferentially enhanced in DR2(15)þ MS patients (Wal 1998).

Mice

Primates

#

Immunodominant epitopes.

484

au19

DA rats (MHC RT1av1) mount a highly encephalitogenic T cell response to both MBP and MOG. The diVerential pathogenicity of the MOG-speciWc T cell responses in these strains is controlled by genes within the MHC, but their identity and their mode of action is unknown. There is no evidence currently available to suggest that similar MHC-mediated eVects inXuence the pathogenic potential of the MOG-speciWc T cell response in other species, although this will be a diYcult question to address in primates as no inbred strains are available. Nevertheless, the Wrst papers describing the T cell response to MOG in marmosets and rhesus monkeys have now been published. A study of MOG-EAE in rhesus monkeys with diVerent MHC backgrounds identiWed three major T cell epitopes localized within amino acid sequences 4-20, 35-50, and 94-116 (Kerlero de Rosbo et al., 2000). The situation in the common marmoset is somewhat simpliWed as this species expresses a common monomorphic class II MHC molecule (Caja-DRB*W1201). This molecule can bind and present MOG14–36 and immunization with MOG14–36 will induce EAE in any individual of this species (Brok et al., 2000). It is notable that many MOG peptide sequences contain determinants that are encephalitogenic in several diVerent species/strains such as MOG35–55

PATHOPHYSIOLOGICAL PROPERTIES OF MOG

485

(Lewis rats, C57B6 mice, and some rhesus monkeys), or MOG90–116 (SJL/J mice, DA rats, rhesus monkeys) (Amor et al., 1994; Brok et al., 2000; Kerlero et al., 2000; Linington et al., 1993; Mendel et al., 1995). However, whether this is due to chance, or reXects some common structural mechanism that restricts the epitope diversity of the MOG-speciWc T cell repertoire to a limited number of ‘‘public’’ determinants is unknown.

MOG-Polymorphisms and Susceptibility to MOG-Induced EAE The cloning of MOG cDNAs resulted in the development of reproducible models of MOG-EAE induced in rats, mice, and primates using recombinant MOGIgd proteins. In these species sensitization with MOGIgd induces variants of EAE that reproduce all major clinical and neuropathological features of MS. The disease follows a chronic relapsing or chronic progressive disease course, and its pathology is characterized by the formation of multifocal CNS speciWc inXammation, MS-like demyelinated plaques and axonal damage (Kornek et al., 2000; Storch et al., 1998; Weissert et al., 1998). The ability of MOG to induce an MS-like disease in a wide range of diVerent species and inbred strains is a direct result of its ability to induce both an encephalitogenic T cell and a demyelinating antibody response. The MOG-speciWc T cell response plays a critical role in the initial stages of MOG-EAE, triggering the recruitment and activation of phagocytes (macrophages and microglia) within the CNS to provide an environment in which immune mediated demyelination can progress more eYciently (Vass et al., 1992). At the same time this local inXammatory response also disrupts the BBB allowing MOG-speciWc antibody and serum complement to enter the CNS compartment. Only then can MOG-speciWc antibodies bind to the surface of the myelin sheath and initiate demyelination by a combination of complement and ADCC (antibody-dependent cell-mediated cytotoxicity) dependent eVector mechanisms. Many diVerent genes are involved in this complex immunopathological response and genetic susceptibility to MOG-EAE is determined by the interplay of multiple genetic loci, including the MHC locus. The role of the MHC in determining genetic susceptibility to EAE is generally discussed in terms the role class II MHC genes play during the selection of the immune repertoire, as well as T cell activation and function. However, there is increasing evidence that other MHC genes also inXuence the clinical course or pathology of EAE. The Wnding that MOG is encoded within the telomeric region of the MHC in rodents and primates prompted speculation that MOG itself may also be a susceptibility gene for EAE and even MS (Hilton et al., 1995; Pham-Dinh et al., 1993; Roth et al., 1995b). As yet, no human MOG polymorphisms have been linked to the human disease, even though several amino acid substitutions are present in human MOG (Gomez-Lira et al., 2002; Ohlenbusch et al., 2002; Rodriguez et al., 1997), but there is evidence that MOG polymorphisms may inXuence the clinical course of MOG-EAE in the rat. In a study of MHC eVects of disease susceptibility in Lewis (RT1l) and Brown Norway (RT1n) rats, early disease onset and increased severity in MOG-EAE segregated with an n haplotype-speciWc MOG RFLP. This eVect is speciWc to MOG-induced EAE suggesting that it involved some modiWcation in the expression of the target antigen. HaplotypespeciWc polymorphisms that modify the expression of MOG may inXuence disease susceptibility in several possible ways. Enhanced expression of MOG outside the CNS, in particular in the thymus may result in tolerance, or inXuence the composition and function of the MOG-speciWc T cell repertoire, as demonstrated for the myelin major proteins, MBP (Harrington et al., 1998; Targoni and Lehmann, 1998) and PLP (Klein et al., 2000; Kuchroo et al., 2002). Alternatively, if a polymorphism increases the availability of antigen in the target organ, it may increase susceptibility to disease induction. Although low levels of MOG mRNAs can be detected in many diVerent organs in the rat, possibly due to the presence of MOG mRNA transcripts in Schwann cells associated with tissue innervation (Pagany et al., 2003b), the absence of detectable levels of MOG in the thymus by immunohistochemistry, Western blotting or immunoprecipitation (unpublished observations) indicates that MOG is not available to mediate central tolerance.

485

486

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

There are, however, signiWcant diVerences in the expression of MOG within the CNS that may account for the early onset and increased severity of MOG-EAE in carrying an n haplotype speciWc RFLP (SteVerl et al., 1999). The presence of this polymorphism is associated with an increase of 25 to 30% in the expression of MOG mRNA transcripts and protein in the CNS as assessed by semi-quantitative PCR and Western blot analysis. The identity of the polymorphism(s) responsible for this enhanced expression of MOG in the CNS is still uncertain. Rat MOG contains multiple SNPs, deletions, and insertions in noncoding regions of the gene, many of which are unique to the n haplotype. Of particular interest was the SNP within the minimal MOG promoter sequence (SteVerl et al., 1999) that introduces a new HindII restriction site into the BN gene adjacent corresponding to the RFLP identiWed by Lambracht et al. (1995). However, in vitro report gene studies failed to demonstrate that this polymorphism alone was responsible for the increased expression of MOG seen in vivo.

MOG—A Relevant Target Autoantigen in MS? The clinical and immunopathological similarities MOG-EAE shares with MS suggest that MOG may also be an important target autoantigen in the human disease. However, it should be stressed that there is increasing evidence that several diVerent CNS autoantigens may be involved in the pathogenesis of MS, and their identity may vary dramatically between patients. Nonetheless, MOG-speciWc autoimmune responses appear to be enhanced in some patients and may directly participate in lesion formation (Genain et al., 1999). Studies using immunopuriWed antigen provided the Wrst indication that the MOG-speciWc T cell response is enhanced in MS and together with the MOG-speciWc B cell response may be preferentially sequestered in the CNS compartment (Sun et al., 1991; Xiao et al., 1991). Bernard and coworkers demonstrated that the peripheral T cell response to MOG is enhanced relative to other candidate autoantigens (MBP, PLP, MAG) in MS patients (Kerlero de Rosbo et al., 1993), and they went on to identify amino acid sequences 1-22, 34-56, and 64-96 as the immunodominant regions of the protein (Kerlero de Rosbo et al., 1997). Subsequent authors used either recombinant MOG proteins or synthetic MOG peptides to analyze the antigen-speciWc response in humans and in general have failed to Wnd diVerent degrees of reactivity to MOG in the circulating T cell population in healthy individuals and in MS patients (Diaz-Villoslada et al., 1999; Lindert et al., 1999; TejadaSimon et al., 2002). However, when MS patients and controls were stratiWed for the genetic risk factor HLA haplotype DR2, an increased number of T cells secreting IFN-g in response to the MOG peptide 63-87 were observed in DR2 positive patients (Wallstrom et al., 1998). Analysis of MOG-speciWc T cell lines, however, suggests that the epitope speciWcity of the MOG-speciWc repertoire in MS patients may be more complex (Lindert et al., 1999). Indeed HLA-DRB1*1501 MS patients have a more focused pattern of T cell reactivity toward MOG peptides as compared with controls and this study identiWed an intracellular MOG epitope (amino acids 146 to 154) as being immunodominant in both MS patients and controls (Weissert et al., 2002). It is clearly crucial to stratify patients and controls with respect to their HLA haplotype when searching for disease related T cell responses, but even when a statistical association with a speciWc T cell response is found, its functional relevance in vivo remains obscure. Cytokine analysis identiWed many MOGIgd-speciWc T cell lines with a Th-0 or Th-2 phenotype suggesting they may be derived from regulatory or memory cells that could provide B cell help, rather than stimulating an inXammatory response in the CNS (Lindert et al., 1999). ‘‘Humanized’’ HLA-DR transgenic mice provide a useful tool to start dissecting the pathogenic potential of human myelin protein speciWc T cell receptors in vivo (Forsthuber et al., 2001). However, this approach is still limited by the complexity of the interactions between the human nervous and immune systems that are not fully replicated in human/mouse chimeras. The evidence for a pathogenic autoantibody response to MOG at least in some patients is, in comparison to the T cell data, compelling. MOG-speciWc antibodies are present in

III. THE MYELIN GENES AND PRODUCTS

487

MOG PUTATIVE PHYSIOLOGICAL FUNCTIONS

MS sera and CSF, but also, albeit at lower frequencies, in patients with other CNS diseases and even normal healthy control donors (Diaz-Villoslada et al., 1999; Karni et al., 1999; Lindert et al., 1999; Reindl et al., 1999). The response in diseases other than MS, however, appears to be transient (Reindl et al., 1999). This lack of disease speciWcity was surprising, as the anti-MOG antibody response in EAE is consistently pathogenic and associated with demyelination. This prompted a study of the epitope speciWcity of MOGIgd-speciWc antibodies isolated from the sera of a panel of MS patients and seropositive controls (Haase et al., 2001). The peptide speciWcity of the MOGIgd-speciWc autoantibody response is heterogeneous, a situation not dissimilar to that described in animals with MOG-induced EAE (Haase et al., 2001). Moreover, it was also possible to demonstrate binding of antibody to the native protein on the surface of MOG-transfected target cells demonstrating that the MOG-speciWc antibody response in MS can satisfy the primary criterion to mediate demyelination in vivo. Surprisingly, antibodies recognizing the native protein were detected in only a minor subset of MS patients (< 5% of the total tested), but this may represent a problem related to the ability of the FACS based assay used in this study to detect low titer antibody responses. Analysis of active MS lesions revealed that not only are antibody and complement activation products deposited on the myelin surface (Lucchinetti et al., 1999; Raine et al., 1999; Storch et al., 1998), but also identiWed MOG-peptide speciWc autoantibodies associated with vesicular myelin debris (Genain et al., 1999; Raine et al., 1999). These observations, together with the presence of C9neo antigen bound to vesicular myelin debris in the cerebrospinal Xuid (Scolding et al., 1989) and the phagocytosis of opsonised myelin by activated macrophages/microglia within the lesions (Prineas and Graham, 1981), provide strong, but circumstantial evidence that humoral eVector mechanisms possibly orchestrated by the MOG-speciWc autoantibody response are involved in the pathogenesis of demyelination.

MOG PUTATIVE PHYSIOLOGICAL FUNCTIONS Despite the importance of MOG as a target for autoimmune attack in EAE and MS, the physiological role of this minor myelin protein is still obscure. The loss of MOG in genetically engineered mice in which MOG gene expression is suppressed using a knockin strategy fails to generate any apparent phenotype. These MOG deWcient mutants develop normally and there are no discernable eVects on myelin morphology or axon structure (Delarasse et al., 2003). However, the absence of any myelin defects or clinical phenotype in these mice oVers the unique opportunity to use them to study the development of immunological self-tolerance to MOG (Litzenburger et al., 2000; Delarasse et al., 2003), as well as in studies of oligodendrogenesis and myelin repair (Picard-Riera et al., 2002).

MOG as an Adhesion Molecule? Although MOG is only a minor component of the intact myelin sheath (0.05 to 0.1 wt%), it is preferentially sequestered at the outermost surface where its eVective concentration will be far higher (Brunner et al., 1989). The surface of the myelin sheath will therefore have a high density of MOG, its Ig-like domain protruding into the extracellular milieu where it could interact with molecules in the extracellular matrix or on adjacent myelinated axons. It is virtually a mirror image of the localization of MAG that is concentrated in the inner tongue and where its Ig-like domains protrude toward the periaxonal space (Quarles, 2002). In this context, it should be noted that MOG bears the HNK-1 glycan epitope (Burger et al., 1993), a marker for many cell adhesion molecules (Schachner, 1989). The HNK-1 epitope is present on several other myelin proteins that are members of the IgSF, such as P0 and MAG. Its function has been particularly well studied in P0, the major PNS myelin protein (Filbin et al., 1990; Filbin and Tennekoon, 1993), where it is involved in maintaining

III. THE MYELIN GENES AND PRODUCTS

au20

488

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

homophilic adhesion between the surfaces of opposing membranes in compact myelin. The observation that an intraperiodic line seems to form at the contact zones between adjacent myelin sheaths in the CNS (Peters et al., 1991) raises the possibility that MOG-MOG interactions may be involved in adhesion between neighboring myelinated Wbers, possibly acting as a glue to maintain bundles of myelinated axons in the CNS (Burger et al., 1993). This could involve homophilic interactions, as well as trans-interactions between opposing cell surfaces or between cell-surface (or myelin sheath surfaces) and the extracellular matrix (Schachner and Martini, 1995). Interestingly, several HNK-1 carbohydrate-binding proteins are present in the CNS including the extracellular matrix glycoproteins laminin (Hall et al., 1997), and the proteoglycan, brevican (Miura et al., 2001). However, there is currently no evidence supporting a direct interaction between MOG and any speciWc components of the extracellular matrix. Immunoelectron microscopy fails to demonstrate any evidence of patching of MOG on the membrane surface at contact zones between adjacent sheaths, and transfection of cell lines with MOG in vitro does not elicit any evidence for homophilic interactions leading to the formation of cell aggregates (Hilton et al., 1995; Pham-Dinh et al., 1994).

MOG as a Receptor? The presence of an extracellular Ig-like domain in the mature protein has lead to speculation that MOG is a cell surface receptor (KroepX et al., 1996; KroepX and Gardinier, 2001a, 2001b) that would transduce signals from the external milieu to the inside of the oligodendrocyte or myelin sheath via its cytoplasmic domains. This concept is supported by the recent demonstration that MOG together with CNPase associates with speciWc lipid rafts within the surface membrane (Kim and PfeiVer, 1999). Lipid rafts are involved in protein sorting or as sites for signaling cascades in the plasma membrane, consistent with the current view that they provide platforms for bringing together and activating components of signal transduction pathways (Lisanti et al., 1995). It is proposed that MOG (and CNPase) may have unique roles among the major myelin proteins in signaling pathways mediated by lipid-protein microdomains formed in myelin membrane loops. Rafts may contribute to mechanisms of both oligodendrocyte/myelin lipid-protein sorting and the mediation of signal transduction (Kim and PfeiVer, 1999). In the absence of a deWned ligand, signal transduction by MOG in oligodendrocytes was assessed in vitro using speciWc antibodies to mimic the action of the putative natural ligand (Dyer, 1993; Dyer and Benjamins, 1990; Dyer and Matthieu, 1994). Short-term exposure of live oligodendrocytes to MOG-speciWc antibody results in a rapid accumulation of antigen/antibody complexes into patches on membrane domains rich in MBP. Subsequently depolymerization of microtubules leads to the loss of the lacy network of microtubules seen in membrane sheets of untreated oligodendrocytes, an eVect that may be mediated by disruption of MBP mediated stabilization of microtubules in the oligodendrocyte cytoskeleton. Microtubule depolarization is also seen in oligodendrocytes treated with a GalC speciWc antibody, but not following treatment with antibody to myelin oligodendrocyte specific protein (MOSP). The mechanism by which two such diverse molecules, one a lipid and the other a protein, mediate the same changes in the cytoskeleton is obscure as the signals transduced by MOG and GalC speciWc antibodies are not identical. Ca2þ plays a crucial role in modulating cytoskeletal organization, but whereas binding to GalC increases intracellular Ca2þ, antibody binding to MOG does not elicit a detectable increase in intracellular Ca2þ (Dyer, 1993; Dyer and Benjamins, 1990).

The Role of the C-terminal Cytoplasmic and Membrane Associated Domains Irrespective of the identity of the putative ligand for the extracellular domain of MOG signal transduction will presumably involve the cytoplasmic domain of the protein. It is striking that the C-terminal regions of MOG, encompassing the second hydrophobic region and the short C-terminal tail, are highly conserved among the six species for

III. THE MYELIN GENES AND PRODUCTS

MOG PUTATIVE PHYSIOLOGICAL FUNCTIONS

FIGURE 18.9 Putative association of MOG with GalC in the membrane bilayer. Given the association of HR2, the second highly hydrophobic region of MOG, to the cytosolic side of the membrane where it could be semi-embedded, the two cysteine residues at position 177 and 198 appear to be ideal targets for post-transductional modiWcation (della Gaspera et al., 1998) such as palmitoylation. Given that ligand-mimicking antibodies to MOG and GalC both activate the same signal transduction pathway (Dyer, 1993), Johns and Bernard (1999) hypothesized that the fatty acid portion of GalC may interact with the lipid-modiWed form of MOG. Adapted from Johns and Bernard, 1999.

which sequence data is available suggesting that this region is of pivotal functional importance (Fig. 18.1). This conserved region may interact with protein or lipid partners in/or close to the cytoplasmic surface of the membrane providing a protein sorting signal during membrane assembly (KroepX et al., 1996; KroepX and Gardinier, 2001a), or as a stabilizing anchor in the mature myelin sheath (della Gaspera et al., 1998; KroepX et al., 1996). The role of this domain may also vary depending on the stage of myelinogenesis. A basolateral membrane targeting signal was identiWed in the cytoplasmic domain of MOG, which is dependent on the integrity of the second hydrophobic region, amino acids 175 to 200 (KroepX and Gardinier, 2001a, 2001b). As discussed earlier, this region of MOG behaves as an integral membrane peptide, which is thought to be embedded in the surface of the lipid bilayer (della Gaspera et al., 1998). This highly hydrophobic segment contains two cysteine residues (C177 and C198), which may be post-transductionaly modiWed by reversible palmitoylation, a feature seen in PLP (Weimbs and StoVel, 1992), as well as in other signaling proteins (Casey, 1995). Although purely speculative, it was proposed that the long-chain fatty acid found in GalC may interact with the lipid-modiWed form of MOG (Johns and Bernard, 1999) (Fig. 18.9). A hypothesis that would explain how some antibodies to either MOG or GalC can activate the same signaling pathway (Dyer, 1993; Dyer and Matthieu, 1994).

Binding of Complement Component C1q to MOG: A Novel Mechanism for Regulating CNS InXammation? The localization of MOG gene in the MHC and its structural homologies with proteins of known immunological functions such as the B-G antigens in the chicken MHC and the B7 family of co-stimulatory molecules led to speculation that MOG has a speciWc immune function within the CNS (Steinman, 1993). Bernard and colleagues identiWed an amino acid motif in the extracellular domain of MOG (see Fig. 18.8C) that was similar to the core C1q-binding sequence previously identiWed in IgG antibodies and demonstrated that MOG can bind the complement component C1q in a dose-dependent and saturating manner (Johns and Bernard, 1997, 1999). In vitro this leads to the activation of complement on the surface of rat oligodendrocytes and ultimately cell lysis. CNS myelin can

III. THE MYELIN GENES AND PRODUCTS

489

490

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

activate the classical complement pathway (Vanguri and Shin, 1986) and MOG may be one of the proteins responsible for this eVect. This direct interaction between a myelin-speciWc protein and C1q has signiWcant implications for CNS inXammation and could be particularly important in demyelinating diseases such as MS (Johns and Bernard, 1997, 1999). In the absence of a suYcient local concentration of complement inhibitors such as CD59, activation of the complement cascade on the myelin surface may not only damage the myelin sheath, but also provide a source of complement derived pro-inXammatory factors that would activate and recruit immune eVector cells into the lesion. In addition, opsonization of myelin debris via antibody independent activation of complement may play an important role in the phagocytosis and clearance of debris from the lesions (Johns and Bernard, 1999).

CONCLUSION MOG is clearly a protein of great neuroimmunological interest, but it is still an orphan in search of a function. Studies in other systems provide hope that we should soon resolve this problem and understand the structural functional relationships of MOG in myelin. Rh blood group proteins were also without a physiological function for many years, other than acting as serological marker that ‘‘decorated’’ the surface of red blood cells and allowed their classiWcation. Only recently were Rh blood group proteins identiWed as ammonium transporters, a function conserved from yeast to mammals (Marini et al., 2000). With that in mind, one should not assume that MOG is a mammalian specialization that simply ‘‘decorates’’ oligodendrocyte and myelin to provide a target for deleterious immune responses in MS. The recent demonstration that MOG is expressed in adult oligodendrocyte progenitors in the rat CNS (Li et al., 2002) also challenges the concept that MOG is only present in the Wnal stages of myelinogenesis (CoVey and MacDermott, 1997; Pham-Dinh et al., 1993; Scolding et al., 1989; Slavin et al., 1997; Solly et al., 1996) and suggests that this molecule has a functional role in development or CNS repair, in addition to its function in the mature myelin sheath.

Acknowledgements The author’s research is supported by the European Leukodystrophy Association, the French Association de Recherche sur la Scle´rose en Plaques and the University of Paris 6 (DPD and AD) and by the Hertie Stiftung and Multiple Sclerosis Society (CL).

References Abo, S., Bernard, C. C., Webb, M., Johns, T. G., Alafaci, A., Ward, L. D., Simpson, R. J., and Kerlero de Rosbo, N. (1993). Preparation of highly puriWed human myelin oligodendrocyte glycoprotein in quantities suYcient for encephalitogenicity and immunogenicity studies. Biochem Mol Biol Int 30, 945–958. Abdul-Majid, K. B., Jirholt, J., Stadelmann, C., Stefferl, A., Kjellen, P., Wallstrom, E., Holmdahl, R., Lassmann, H., Olsson T., and Harris, R. A. (2000). Screening of several H-2 congenic mouse strains identified H-2(q) mice as highly susceptible to MOG-induced EAE with minimal adjuvant requirement. J Neuroimmunol 111, 23–33. Adelmann, M., Wood, J., Benzel, I., Fiori, P., Lassmann, H., Matthieu, J. M., Gardinier, M. V., Dornmair, K., and Linington, C. (1995). The N-terminal domain of the myelin oligodendrocyte glycoprotein (MOG) induces acute demyelinating experimental autoimmune encephalomyelitis in the Lewis rat. J Neuroimmunol 63, 17–27. Amiguet, P., Gardinier, M. V., Zanetta, J. P., and Matthieu, J. M. (1992). PuriWcation and partial structural and functional characterization of mouse myelin/oligodendrocyte glycoprotein. J Neurochem 58, 1676–1682. Amor, S., Groome, N., Linington, C. Morris, M. M., Dommair, K., Gardinier, M. V., Matthieu, J. M., Baker, D. (1994) IdentiWcation of epitopes of myelin oligodendrocyte glycoprotein for the induction of experimental allergic encephalomyelitis in SJL and Biozzi AB/H mice. J Immunol 153, 4349–4356. Ballenthin, P. A., and Gardinier, M. V. (1996). Myelin/oligodendrocyte glycoprotein is alternatively spliced in humans but not mice. J Neurosci Res 46, 271–281. J. Immunol. 153, 4349–4356.

III. THE MYELIN GENES AND PRODUCTS

CONCLUSION

Baumann, N., and Pham-Dinh, D. (2001). Biology of oligodendrocyte and myelin in the mammalian central nervous system. Physiol Rev 81, 871–927. Bikle, D. D., Munson, S., and Komuves, L. (1996). Zipper protein, a B-G protein with the ability to regulate actin/ myosin 1 interactions in the intestinal brush border. J Biol Chem 271, 9075–9083. Birling, M. C., Roussel, G., Nussbaum, F., and Nussbaum, J. L. (1993). Biochemical and immunohistochemical studies with speciWc polyclonal antibodies directed against bovine myelin/oligodendrocyte glycoprotein. Neurochem Res 18, 937–945. Bourquin, C., Iglesias, A., Berger, T., Wekerle, H., and Linington, C. (2000). Myelin oligodendrocyte glycoprotein-DNA vaccination induces antibody-mediated autoaggression in experimental autoimmune encephalomyelitis. Eur J Immunol 30, 3663–3671. Bradl, M., and Linington, C. (1996) Animal models of demyelination. Brain Pathol 6, 303–311. Brehm, U., Piddlesden, S. J., Gardinier, M. V., and Linington, C. (1999). Epitope speciWcity of demyelinating monoclonal autoantibodies directed against the human myelin oligodendrocyte glycoprotein (MOG). J Neuroimmunol 97, 9–15. Breithaupt, C., Schubart, A., Zander, H., Skerra, A., Huber, R., Linington, C., and Jacob U. (2003). Structural insights into the antigenicity of myelin oligodendrocyte glycoprotein. Proc Natl Acad Sci U S A 100, 9446–9451. Brok, H. P., Uccelli, A., Kerlero De Rosbo, N., Bontrop, R. E., Roccatagliata, L., de Groot, N. G., Capello, E., Laman, J. D., Nicolay, K., Mancardi, G. L., Ben-Nun, A., and Hart, B. A. (2000). Myelin/oligodendrocyte glycoprotein-induced autoimmune encephalomyelitis in common marmosets: The encephalitogenic T cell epitope pMOG24–36 is presented by a monomorphic MHC class II molecule. J Immunol 165, 1093–1101. Brunner, C., Lassmann, H., Waehneldt, T. V., Matthieu, J. M., and Linington, C. (1989). DiVerential ultrastructural localization of myelin basic protein, myelin/oligodendroglial glycoprotein, and 2’,3’-cyclic nucleotide 3’-phosphodiesterase in the CNS of adult rats. J Neurochem 52, 296–304. Burger, D., Steck, A. J., Bernard, C. C., and Kerlero de Rosbo, N. (1993). Human myelin/oligodendrocyte glycoprotein: A new member of the L2/HNK-1 family. J Neurochem 61, 1822–1827. Campagnoni, A. T. (1988). Molecular biology of myelin proteins from the central nervous system. J Neurochem 51, 1–14. Casey, P. J. (1995). Protein lipidation in cell signaling. Science 268, 221–225. Clements, C. S., Reid, H. H., Beddoe, T., Tynan, F. E., Perugini, M. A., Johns, T. G., Bernard, C. C., and Rossjohn, J. (2003). The crystal structure of myelin oligodendrocyte glycoprotein, a key autoantigen in multiple sclerosis. Proc Natl Acad Sci U S A 100, 11059–11064. CoVey JC, McDermott KW (1997). The regional distribution of myelin oligodendrocyte glycoprotein (MOG) in the developing rat CNS: An in vivo immunohistochemical study. J Neurocytol 26: 149–161. Dahlman, I., Lorentzen, J. C., de Graaf, K. L., SteVerl, A., Linington, C., Luthman, H., and Olsson, T. (1998). Quantitative trait loci disposing for both experimental arthritis and encephalomyelitis in the DA rat; impact on severity of myelin oligodendrocyte glycoprotein-induced experimental autoimmune encephalomyelitis and antibody isotype pattern. Eur J Immunol 28, 2188–2196. Dahlman, I., Wallstrom, E., Weissert, R., Storch, M., Kornek, B., Jacobsson, L., Linington, C., Luthman, H., Lassmann, H., and Olsson, T. (1999). Linkage analysis of myelin oligodendrocyte glycoprotein-induced experimental autoimmune encephalomyelitis in the rat identiWes a locus controlling demyelination on chromosome 18. Hum Mol Genet 8, 2183–2190. Daubas, P., Pham-Dinh, D., and Dautigny, A. (1994). Structure and polymorphism of the mouse myelin/ oligodendrocyte glycoprotein gene. Genomics 23, 36–41. Delarasse, C., Daubas, P., Mars, L., Vizler, C., Litzenburger, T., Iglesias, A., Bauer, J., Della Gaspera, B., Schubart, A., Decker, L., Roussel, G., Dierich, A., Amor, S., Dautigny, A., Liblau, R., and Pham-Dinh, D. (2003). Myelin/oligodendrocyte glycoprotein (MOG)-deficient mice reveal lack of tolerance to MOG in wildtype mice. J Clin Invest 112, 544–553. della Gaspera, B., Pham-Dinh, D., Roussel, G., Nussbaum, J. L., and Dautigny, A. (1998). Membrane topology of the myelin/oligodendrocyte glycoprotein. Eur J Biochem 258, 478–484. Devaux, B., Enderlin, F., Wallner, B., Smilek, D. E. (1997). Induction of EAE in mice with recombinant human MOG, and treatment of EAE with a MOG peptide. J Neuroimmunol 75, 169–173. Diaz-Villoslada, P., Shih, A., Shao, L., Genain, C. P., and Hauser, S. L. (1999). Autoreactivity to myelin antigens: Myelin/oligodendrocyte glycoprotein is a prevalent autoantigen. J Neuroimmunol 99, 36–43. Dyer, C. A. (1993). Novel oligodendrocyte transmembrane signaling systems. Investigations utilizing antibodies as ligands. Mol Neurobiol 7, 1–22. Dyer, C. A., and Benjamins, J. A. (1990). Glycolipids and transmembrane signaling: Antibodies to galactocerebroside cause an inXux of calcium in oligodendrocytes. J Cell Biol 111, 625–633. Dyer, C. A., and Matthieu, J. M. (1994). Antibodies to myelin/oligodendrocyte-speciWc protein and myelin/ oligodendrocyte glycoprotein signal distinct changes in the organization of cultured oligodendroglial membrane sheets. J Neurochem 62, 777–787. Filbin, M. T., and Tennekoon, G. I. (1993). Homophilic adhesion of the myelin P0 protein requires glycosylation of both molecules in the homophilic pair. J Cell Biol 122, 451–459. Filbin, M. T., Walsh, F. S., Trapp, B. D., Pizzey, J. A., and Tennekoon, G. I. (1990). Role of myelin P0 protein as a homophilic adhesion molecule. Nature 344, 871–872. Forsthuber, T. G., Shive, C. L., Wienhold, W., de Graaf, K., Spack, E. G., Sublett, R., Melms, A., Kort, J., Racke, M. K., and Weissert, R. (2001). T cell epitopes of human myelin oligodendrocyte glycoprotein

III. THE MYELIN GENES AND PRODUCTS

491

492

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

identiWed in HLA-DR4 (DRB1*0401) transgenic mice are encephalitogenic and are presented by human B cells. J Immunol 167, 7119–7125. Gardinier, M. V., Amiguet, P., Linington, C., and Matthieu, J. M. (1992). Myelin/oligodendrocyte glycoprotein is a unique member of the immunoglobulin superfamily. J Neurosci Res 33, 177–187. Genain, C. P., Cannella, B., Hauser, S. L., and Raine, C. S. (1999). IdentiWcation of autoantibodies associated with myelin damage in multiple sclerosis. Nat Med 5, 170–175. Gomez-Lira, M., Moretto, G., Bonamini, D., Benedetti, M. D., Pignatti, P. F., Rizzuto, N., and Salviati, A. (2002). Myelin oligodendrocyte glycoprotein polymorphisms and multiple sclerosis. J Neuroimmunol 133, 241–243. Greenbaum S., and Zhuang Y. (2002). IdentiWcation of E2A target genes in B lymphocyte development by using a gene tagging-based chromatin immunoprecipitation system. Proc Natl Acad Sci U S A 99: 15030–15035. GriYth, L. S., Schmitz, B., and Schachner, M. (1992). L2/HNK-1 carbohydrate and protein-protein interactions mediate the homophilic binding of the neural adhesion molecule P0. J Neurosci Res 33, 639–648. Haase C. G., Guggenmos, J., Brehm, U., Andersson, M., Olsson, T., Reindl, M., Schneidewind, J. M., Zettl, U. K., Heidenreich, F., Berger, T., Wekerle, H., Hohlfeld, R., and Linington, C. (2001). The Wne speciWcity of the myelin oligodendrocyte glycoprotein autoantibody response in patients with multiple sclerosis and normal healthy controls. J Neuroimmunol 114, 220–225. Hall, H., Deutzmann, R., Timpl, R., Vaughan, L., Schmitz, B., and Schachner, M. (1997). HNK-1 carbohydratemediated cell adhesion to laminin-1 is diVerent from heparin-mediated and sulfatide-mediated cell adhesion. Eur J Biochem 246, 233–242. Harpaz, Y., and Chothia, C. (1994). Many of the immunoglobulin superfamily domains in cell adhesion molecules and surface receptors belong to a new structural set which is close to that containing variable domains. J Mol Biol 238, 528–539. Harrington, C. J., Paez, A., Hunkapiller, T., Mannikko, V., Brabb, T., Ahearn, M., Beeson, C., and Goverman, J. (1998). DiVerential tolerance is induced in T cells recognizing distinct epitopes of myelin basic protein. Immunity 8, 571–580. Henry, J., Mather, I. H., McDermott, M. F., Pontarotti, P. (1998). B30.2-like domain proteins: update and new insights into a rapidly expanding family of proteins. Mol Biol Evol 15, 1696–705. Henry, J., Miller, M. M., and Pontarotti, P. (1999). Structure and evolution of the extended B7 family. Immunol Today 20, 285–288. Henry, J., Ribouchon, M. T., OVer, C., and Pontarotti, P. (1997). B30.2-like domain proteins: A growing family. Biochem Biophys Res Commun 235, 162–165. Hilton, A. A., Slavin, A. J., Hilton, D. J., and Bernard, C. C. (1995). Characterization of cDNA and genomic clones encoding human myelin oligodendrocyte glycoprotein. J Neurochem 65, 309–318. Hjelmstrom, P., Penzotti, J. E., Henne, R. M., and Lybrand, T. P. (1998). A molecular model of myelin oligodendrocyte glycoprotein. J Neurochem 71, 1742–1749. Holz, A., Bielekova, B., Martin, R., and Oldstone, M. B. (2000). Myelin-associated oligodendrocytic basic protein: IdentiWcation of an encephalitogenic epitope and association with multiple sclerosis. J Immunol 164, 1103–1119. Hudson, L. D. K., N., and Kim, J. G.. (1995). Control of myelin gene expression. In ‘‘Glial Cell Development: Basic Principle of Clinical Relevance’’ (Richardson, W. D. and Jessen, K. R., ed.), pp. 101–121. Bios ScientiWc, London. Ichikawa, M., Johns, T. G., Adelmann, M., and Bernard, C. C. (1996). Antibody response in Lewis rats injected with myelin oligodendrocyte glycoprotein derived peptides. Int Immunol 8, 1667–1674. Jack, L. J., and Mather, I. H. (1990). Cloning and analysis of cDNA encoding bovine butyrophilin, an apical glycoprotein expressed in mammary tissue and secreted in association with the milk-fat globule membrane during lactation. J Biol Chem 265, 14481–14486. Jaquet, V., Pfend, G., Tosic, M., and Matthieu, J. M. (1999). Analysis of cis-acting sequences from the myelin oligodendrocyte glycoprotein promoter. J Neurochem 73, 120–128. Jarvi, S. I., Goto, R. M., Gee, G. F., Briles, W. E., and Miller, M. M. (1999). IdentiWcation, inheritance, and linkage of B-G-like and MHC class I genes in cranes. J Hered 90, 152–159. Johns, T. G., Kerlero de Rosbo, N., Menon, K. K., Abo, S., Gonzales, M. F., and Bernard, C. C. (1995). Myelin oligodendrocyte glycoprotein induces a demyelinating encephalomyelitis resembling multiple sclerosis. J Immunol 154, 5536–5541. Johns, T. G., and Bernard, C. C. (1997). Binding of complement component Clq to myelin oligodendrocyte glycoprotein: A novel mechanism for regulating CNS inXammation. Mol Immunol 34, 33–38. Johns, T. G., and Bernard, C. C. (1999). The structure and function of myelin oligodendrocyte glycoprotein. J Neurochem 72, 1–9. Jones, E. P., Xiao, H., Schultz, R. A., Flaherty, L., Trachtulec, Z., Vincek, V., Larin, Z., Lehrach, H., and Lindahl, K. F. (1995). MHC class I gene organization in > 1.5-Mb YAC contigs from the H2-M region. Genomics 27, 40–51. Jurka, J., and Milosavljevic, A. (1991). Reconstruction and analysis of human Alu genes. J Mol Evol 32, 105–121. Jurka, J., Walichiewicz, J., and Milosavljevic, A. (1992). Prototypic sequences for human repetitive DNA. J Mol Evol 35, 286–291. Karni, A., Bakimer-Kleiner, R., Abramsky, O., and Ben-Nun, A. (1999). Elevated levels of antibody to myelin oligodendrocyte glycoprotein is not speciWc for patients with multiple sclerosis. Arch Neurol 56, 311–315. Kaufman, J., and Salomonsen, J. (1992). B-G: we know what it is, but what does it do? Immunol, Today 13, 1–3.

III. THE MYELIN GENES AND PRODUCTS

CONCLUSION

Kaufman, J., Salomonsen, J., Skjodt, K., and Thorpe, D. (1990). Size polymorphism of chicken major histocompatibility complex-encoded B-G molecules is due to length variation in the cytoplasmic heptad repeat region. Proc Natl Acad Sci USA 87, 8277–8281. Kerlero de Rosbo, N., Brok, H. P., Bauer, J., Kaye, J. F., t Hart, B. A., and Ben-Nun, A. (2000). Rhesus monkeys are highly susceptible to experimental autoimmune encephalomyelitis induced by myelin oligodendrocyte glycoprotein: Characterisation of immunodominant T- and B-cell epitopes. J Neuroimmunol 110, 83–96. Kerlero de Rosbo, N., HoVman, M., Mendel, I., Yust, I., Kaye, J., Bakimer, R., Flechter, S., Abramsky, O., Milo, R., Karni, A., and Ben-Nun, A. (1997). Predominance of the autoimmune response to myelin oligodendrocyte glycoprotein (MOG) in multiple sclerosis: Reactivity to the extracellular domain of MOG is directed against three main regions. Eur J Immunol 27, 3059–3069. Kerlero de Rosbo, N., Honegger, P., Lassmann, H., and Matthieu, J. M. (1990). Demyelination induced in aggregating brain cell cultures by a monoclonal antibody against myelin/oligodendrocyte glycoprotein. J Neurochem 55, 583–587. Kerlero de Rosbo, N., Mendel, I., Ben-Nun, A. (1995). Chronic relapsing experimental autoimmune encephalomyelitis with a delayed onset and an atypical clinical course, induced in PL/J mice by myelin oligodendrocyte glycoprotein (MOG)-derived peptide: preliminary analysis of MOG T cell epitopes. Eur J Immunol 25, 985–993. Kerlero de Rosbo, N., Milo, R., Lees, M. B., Burger, D., Bernard, C. C., and Ben-Nun, A. (1993). Reactivity to myelin antigens in multiple sclerosis. Peripheral blood lymphocytes respond predominantly to myelin oligodendrocyte glycoprotein. J Clin Invest 92, 2602–2608. Kim, T., and PfeiVer, S. E. (1999). Myelin glycosphingolipid/cholesterol-enriched microdomains selectively sequester the non-compact myelin proteins CNP and MOG. J Neurocytol 28, 281–293. Kitagawa, K., Sinoway, M. P., Yang, C., Gould, R. M., and Colman, D. R. (1993). A proteolipid protein gene family: Expression in sharks and rays and possible evolution from an ancestral gene encoding a pore-forming polypeptide. Neuron 11, 433–448. Klein, L., Klugmann, M., Nave, K. A., Tuohy, V. K., and Kyewski, B. (2000). Shaping of the autoreactive T-cell repertoire by a splice variant of self protein expressed in thymic epithelial cells. Nat Med 6, 56–61. Kojima, K., Berger, T., Lassmann, H., Hinze-Selch, D., Zhang, Y., Gehrmann, J., Reske, K., Wekerle, H., and Linington, C. (1994) Experimental autoimmune panencephalitis and uveoretinitis transferred to the Lewis rat by T lymphocytes speciWc for the S100 beta molecule, a calcium binding protein of astroglia. J Exp Med 180, 817–829. Kolkman, J. A., and Stemmer, W. P. (2001). Directed evolution of proteins by exon shuZing. Nat Biotechnol 19, 423–428. Kornek, B., Storch, M. K., Weissert, R., Wallstroem, E., SteVerl, A., Olsson, T., Linington, C., Schmidbauer, M., and Lassmann, H. (2000). Multiple sclerosis and chronic autoimmune encephalomyelitis: A comparative quantitative study of axonal injury in active, inactive, and remyelinated lesions. Am J Pathol 157, 267–276. KroepX, J. F., and Gardinier, M. V. (2001a). IdentiWcation of a basolateral membrane targeting signal within the cytoplasmic domain of myelin/oligodendrocyte glycoprotein. J Neurochem 77, 1301–1309. KroepX, J. F., and Gardinier, M. V. (2001b). Mutually exclusive apicobasolateral sorting of two oligodendroglial membrane proteins, proteolipid protein and myelin/oligodendrocyte glycoprotein, in Madin-Darby canine kidney cells. J Neurosci Res 66, 1140–1148. KroepX, J. F., Viise, L. R., Charron, A. J., Linington, C., and Gardinier, M. V. (1996). Investigation of myelin/ oligodendrocyte glycoprotein membrane topology. J Neurochem 67, 2219–2222. Kuchroo, V. K., Anderson, A. C., Waldner, H., Munder, M., Bettelli, E., and Nicholson, L. B. (2002). T cell response in experimental autoimmune encephalomyelitis (EAE): Role of self and cross-reactive antigens in shaping, tuning, and regulating the autopathogenic T cell repertoire. Annu Rev Immunol 20, 101–123. Lambracht, D., Prokop, C., Hedrich, H. J., Fischer Lindahl, K., and Wonigeit, K. (1995). Mapping of H2-M homolog and MOG genes in the rat MHC. Immunogenetics 42, 418–421. Lassmann, H., Brunner, C., Bradl, M., and Linington, C. (1988). Experimental allergic encephalomyelitis: The balance between encephalitogenic T lymphocytes and demyelinating antibodies determines size and structure of demyelinated lesions. Acta Neuropathol (Berl) 75, 566–576. Lebar, R., Lubetzki, C., Vincent, C., Lombrail, P., and Boutry, J. M. (1986). The M2 autoantigen of central nervous system myelin, a glycoprotein present in oligodendrocyte membrane. Clin Exp Immunol 66, 423–434. Lebar, R., Vincent, C., and Fischer-le Boubennec, E. (1979). Studies on autoimmune encephalomyelitis in the guinea pig–III. A comparative study of two autoantigens of central nervous system myelin. J Neurochem 32, 1451–1460. Li, G., Crang, A. J., Rundle, J. L., and Blakemore, W. F. (2002). Oligodendrocyte progenitor cells in the adult rat CNS express myelin oligodendrocyte glycoprotein (MOG). Brain Pathol 12, 463–471. Lindberg, F. P., Gresham, H. D., Schwarz, E., and Brown, E. J. (1993). Molecular cloning of integrin-associated protein: An immunoglobulin family member with multiple membrane-spanning domains implicated in alpha v beta 3-dependent ligand binding. J Cell Biol 123, 485–496. Lindert, R. B., Haase, C. G., Brehm, U., Linington, C., Wekerle, H., and Hohlfeld, R. (1999). Multiple sclerosis: B- and T-cell responses to the extracellular domain of the myelin oligodendrocyte glycoprotein. Brain 122 (Pt 11), 2089–2100.

III. THE MYELIN GENES AND PRODUCTS

493

494

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

Linington, C., Berger, T., Perry, L., Weerth, S., Hinze-Selch, D., Zhang, Y., Lu, H. C., Lassmann, H., and Wekerle, H. (1993). T cells speciWc for the myelin oligodendrocyte glycoprotein mediate an unusual autoimmune inXammatory response in the central nervous system. Eur J Immunol 23, 1364–1372. Linington, C., Bradl, M., Lassmann, H., Brunner, C., and Vass, K. (1988). Augmentation of demyelination in rat acute allergic encephalomyelitis by circulating mouse monoclonal antibodies directed against a myelin/oligodendrocyte glycoprotein. Am J Pathol 130, 443–454. Linington, C., and Waehneldt, T. V. (1990). Conservation of the carboxyl terminal epitope of myelin proteolipid protein in the tetrapods and lobe-Wnned Wsh. J Neurochem 54, 1354–1359. Linington, C., Webb, M., and Woodhams, P. L. (1984). A novel myelin-associated glycoprotein deWned by a mouse monoclonal antibody. J Neuroimmunol 6, 387–396. Linsley, P. S., Peach, R., Gladstone, P., and Bajorath, J. (1994). Extending the B7 (CD80) gene family. Protein Sci 3, 1341–1343. Lisanti, M. P., Tang, Z., Scherer, P. E., Kubler, E., Koleske, A. J., and Sargiacomo, M. (1995). Caveolae, transmembrane signalling and cellular transformation. Mol Membr Biol 12, 121–124. Litzenburger, T., Bluthmann, H., Morales, P., Pham-Dinh, D., Dautigny, A., Wekerle, H., and Iglesias, A. (2000). Development of myelin oligodendrocyte glycoprotein autoreactive transgenic B lymphocytes: Receptor editing in vivo after encounter of a self-antigen distinct from myelin oligodendrocyte glycoprotein. J Immunol 165, 5360–5366. Litzenburger, T., Fassler, R., Bauer, J., Lassmann, H., Linington, C., Wekerle, H., and Iglesias, A. (1998). B lymphocytes producing demyelinating autoantibodies: Development and function in gene-targeted transgenic mice. J Exp Med 188, 169–180. Louis, J. C., Magal, E., Muir, D., Manthorpe, M., and Varon, S. (1992). CG-4, a new bipotential glial cell line from rat brain, is capable of diVerentiating in vitro into either mature oligodendrocytes or type-2 astrocytes. J Neurosci Res 31, 193–204. Lucchinetti, C., Bruck, W., Parisi, J., Scheithauer, B., Rodriguez, M., and Lassmann, H. (1999). A quantitative analysis of oligodendrocytes in multiple sclerosis lesions. A study of 113 cases. Brain 122 (Pt 12), 2279–2295. Maatta, J. A., Kaldman, M. S., Sakoda, S., Salmi, A. A., and Hinkkanen, A. E. (1998) Encephalitogenicity of myelin-associated oligodendrocytic basic protein and 2’,3’-cyclic nucleotide 3’-phosphodiesterase for BALB/c and SJL mice. Immunol 95, 383–388. Malfroy, L., Viratelle, C., Coppin, H., Borot, N., and Roth, M. P. (1995). Polymorphic tri- and tetranucleotide repeats in exons 1 and 8 of the myelin oligodendrocyte glycoprotein (MOG) gene. Hum Genet 96, 737–738. Marini, A. M., Matassi, G., Raynal, V., Andre, B., Cartron, J. P., and Cherif-Zahar, B. (2000). The human Rhesus-associated RhAG protein and a kidney homologue promote ammonium transport in yeast. Nat Genet 26, 341–344. Matthieu, J. M., Eschmann, N., Burgisser, P., Malotka, J., and Waehneldt, T. V. (1986). Expression of myelin proteins characteristic of Wsh and tetrapods by Polypterus revitalizes long discredited phylogenetic links. Brain Res 379, 137–142. Mendel, I., Kerlero de Rosbo, N., and Ben-Nun A. (1995) A myelin oligodendrocyte glycoprotein peptide induces typical chronic experimental autoimmune encephalomyelitis in H-2b mice: Fine speciWcity and T cell receptor V beta expression of encephalitogenic T cells. Eur J Immunol. 25, 1951–1959. Mesleh, M. F., Belmar, N., Lu, C. W., Krishnan, V. V., Maxwell, R. S., Genain, C. P., and Cosman, M. (2002). Marmoset Wne B cell and T cell epitope speciWcities mapped onto a homology model of the extracellular domain of human myelin oligodendrocyte glycoprotein. Neurobiol Dis 9, 160–172. Miller, M. M., Goto, R., Young, S., Chirivella, J., Hawke, D., and Miyada, C. G. (1991). Immunoglobulin variable-region-like domains of diverse sequence within the major histocompatibility complex of the chicken. Proc Natl Acad Sci U S A 88, 4377–4381. Miura, R., Ethell, I. M., and Yamaguchi, Y. (2001). Carbohydrate-protein interactions between HNK-1-reactive sulfoglucuronyl glycolipids and the proteoglycan lectin domain mediate neuronal cell adhesion and neurite outgrowth. J Neurochem 76, 413–424. Modrek, B., Resch, A., Grasso, C., and Lee, C. (2001). Genome-wide detection of alternative splicing in expressed sequences of human genes. Nucleic Acids Res 29, 2850–2859. Morris-Downes, M. M., McCormack, K., Baker, D., Sivaprasad, D., Natkunarajah, J., and Amor, S. (2002). Encephalitogenic and immunogenic potential of myelin-associated glycoprotein (MAG), oligodendrocytespeciWc glycoprotein (OSP) and 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNPase) in ABH and SJL mice. J Neuroimmunol 122, 20–33. Niedieck B. (1969). Further studies on the role of lipids and lipid antibodies in experimental allergic encephalomyelitis. Int Arch Allergy & Appl Immunol 36, 328–333. Niedieck, B. (1975). On a glycolipid hapten of myelin. Prog. Allergy. 18, 353–422. Nilsson, I., Saaf, A., Whitley, P., Gafvelin, G., Waller, C., and von Heijne, G. (1998). Proline-induced disruption of a transmembrane alpha-helix in its natural environment. J Mol Biol 284, 1165–1175. Ohlenbusch, A., Pohl, D., and Hanefeld, F. (2002). Myelin oligodendrocyte gene polymorphisms and childhood multiple sclerosis. Pediatr Res 52, 175–179. Olsson, T., Dahlman, I., Wallstrom, E., Weissert, R., and Piehl, F. (2000). Genetics of rat neuroinXammation. J Neuroimmunol 107, 191–200. Pagany, M., Jagodic, M., Bourquin, C., Olsson, T., and Linington, C. (2003). Genetic variation in myelin oligodendrocyte glycoprotein expression and susceptibility to experimental autoimmune encephalomyelitis. J Neuroimmunol 139, 1–8.

III. THE MYELIN GENES AND PRODUCTS

CONCLUSION

Pagany M, Jagodic M, Schubart A, Pham-Dinh D, Bachelin C, Van Evercooren A, Lachapelle F, Olsson T and Linington C. (2003b) Myelin Oligodendrocyte Glycoprotein expression in the peripheral nervous system. Neurosci Lett 350, 165–168. Peters, A., Palay, S., and Webster, H. de F. (1991). ‘‘The Fine Structure of the Nervous System: The Neuron and the Supporting Cells,’’ 3rd ed. Oxford University Press, New York. Pham-Dinh, D., Allinquant, B., Ruberg, M., Della Gaspera, B., Nussbaum, J. L., and Dautigny, A. (1994). Characterization and expression of the cDNA coding for the human myelin/oligodendrocyte glycoprotein. J Neurochem 63, 2353–2356. Pham-Dinh, D., Della Gaspera, B., Kerlero de Rosbo, N., and Dautigny, A. (1995a). Structure of the human myelin/oligodendrocyte glycoprotein gene and multiple alternative spliced isoforms. Genomics 29, 345–352. Pham-Dinh, D., Jones, E. P., Pitiot, G., Della Gaspera, B., Daubas, P., Mallet, J., Le Paslier, D., Fischer Lindahl, K., and Dautigny, A. (1995b). Physical mapping of the human and mouse MOG gene at the distal end of the MHC class Ib region. Immunogenetics 42, 386–391. Pham-Dinh, D., Mattei, M. G., Nussbaum, J. L., Roussel, G., Pontarotti, P., Roeckel, N., Mather, I. H., Artzt, K., Lindahl, K. F., and Dautigny, A. (1993). Myelin/oligodendrocyte glycoprotein is a member of a subset of the immunoglobulin superfamily encoded within the major histocompatibility complex. Proc Natl Acad Sci U S A 90, 7990–7994. Picard-Riera, N., Decker L, Delarasse, C., Goude, K., Nait-Oumesmar, B., Liblau, R., Pham-Dinh, D., Evercooren, A. B. (2002). Experimental autoimmune encephalomyelitis mobilizes neural progenitors from the subventricular zone to undergo oligodendrogenesis in adult mice. Proc Natl Acad Sci U S A 99, 13211–13216. Piddlesden, S., Lassmann, H., LaVaWan, I., Morgan, B. P., and Linington, C. (1991). Antibody-mediated demyelination in experimental allergic encephalomyelitis is independent of complement membrane attack complex formation. Clin Exp Immunol 83, 245–250. Prineas, J. W., and Graham, J. S. (1981). Multiple sclerosis: Capping of surface immunoglobulin G on macrophages engaged in myelin breakdown. Ann Neurol 10, 149–158. Proescholdt, M. G., Quigley, L., Martin, R., and Herkenham, M. (2002). Immunization with a cannabinoid receptor type 1 peptide results in experimental allergic meningocerebellitis in the Lewis rat: A model for cellmediated autoimmune neuropathology. J Neurosci Res 70, 150–160. Quarles, R. H. (1997). Glycoproteins of myelin sheaths. J Mol Neurosci 8, 1–12. Quarles, R. H. (2002). Myelin sheaths: Glycoproteins involved in their formation, maintenance and degeneration. Cell Mol Life Sci 59, 1851–1871. Raine, C. S., Cannella, B., Hauser, S. L., and Genain, C. P. (1999). Demyelination in primate autoimmune encephalomyelitis and acute multiple sclerosis lesions: A case for antigen-speciWc antibody mediation. Ann Neurol 46, 144–160. Raine, C. S., Johnson, A. B., Marcus, D. M.,. Suzuki, A., and Bornstein, M. B. (1981). Demyelination in vitro. Absorption studies demonstrate that galactocerebroside is a major target. J Neurol Sci. 52, 117–131. Reindl, M., Linington, C., Brehm, U., Egg, R., Dilitz, E., Deisenhammer, F., Poewe, W., and Berger, T. (1999). Antibodies against the myelin oligodendrocyte glycoprotein and the myelin basic protein in multiple sclerosis and other neurological diseases: A comparative study. Brain 122 (Pt 11), 2047–2056. Rhodes, D. A., Stammers, M., Malcherek, G., Beck, S., and Trowsdale, J. (2001). The cluster of BT genes in the extended major histocompatibility complex. Genomics 71, 351–362. Rodriguez, D., Della Gaspera, B., Zalc, B., Hauw, J. J., Fontaine, B., Edan, G., Clanet, M., Dautigny, A., and Pham-Dinh, D. (1997). IdentiWcation of a Val I45 Ile substitution in the human myelin oligodendrocyte glycoprotein: Lack of association with multiple sclerosis. The Reseau de Recherche Clinique INSERM sur la Susceptibilite Genetique a la Sclerose en Plaques. Mult Scler 3, 377–381. Roth, G. A., Roytta, M., Yu, R. K., Raine, C. S., and Bornstein, M. B. (1985). Antisera to diVerent glycolipids induce myelin alterations in mouse spinal cord tissue cultures. Brain Res 339, 9–18. Roth, M. P., Dolbois, L., Borot, N., Amadou, C., Clanet, M., Pontarotti, P., and Coppin, H. (1995a). Three highly polymorphic microsatellites at the human myelin oligodendrocyte glycoprotein locus, 100 kb telomeric to HLA-F. Characterization and relation to HLA haplotypes. Hum Immunol 43, 276–282. Roth, M. P., Malfroy, L., OVer, C., Sevin, J., Enault, G., Borot, N., Pontarotti, P., and Coppin, H. (1995b). The human myelin oligodendrocyte glycoprotein (MOG) gene: Complete nucleotide sequence and structural characterization. Genomics 28, 241–250. Ruddy, D. A., Kronmal, G. S., Lee, V. K., Mintier, G. A., Quintana, L., Domingo, R., Jr., Meyer, N. C., Irrinki, A., McClelland, E. E., Fullan, A., Mapa, F. A., Moore, T., Thomas, W., Loeb, D. B., Harmon, C., Tsuchihashi, Z., WolV, R. K., Schatzman, R. C., Feder, J. N. (1997). A 1.1-Mb transcript map of the hereditary hemochromatosis locus. Genome Res 7, 441–456. Schachner, M. (1989). Families of neural adhesion molecules. Ciba Found Symp 145, 156–169, discussion 169–172. Schachner, M., and Martini, R. (1995). Glycans and the modulation of neural-recognition molecule function. Trends Neurosci 18, 183–191. Schliess, F., and StoVel, W. (1991). Evolution of the myelin integral membrane proteins of the central nervous system. Biol Chem Hoppe Seyler 372, 865–874. Schluesener, H. J., Sobel, R. A., Linington, C., and Weiner, H. L. (1987). A monoclonal antibody against a myelin oligodendrocyte glycoprotein induces relapses and demyelination in central nervous system autoimmune disease. J Immunol 139, 4016–4021.

III. THE MYELIN GENES AND PRODUCTS

495

496

18. MYELIN OLIGODENDROCYTE GLYCOPROTEIN GENE

Scolding, N. J., Frith, S., Linington, C., Morgan, B. P., Campbell, A. K., and Compston, D. A. (1989). Myelinoligodendrocyte glycoprotein (MOG) is a surface marker of oligodendrocyte maturation. J Neuroimmunol 22, 169–176. Shapiro, L., Doyle, J. P., Hensley, P., Colman, D. R., and Hendrickson, W. A. (1996). Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17, 435–449. Shibui, A., Tsunoda, T., Seki, N., Suzuki, Y., Sugano, S., and Sugano, K. S. (1999). Cloning, expression analysis, and chromosomal localization of a novel butyrophilin-like receptor. J Hum Genet 44, 249–253. Sipos, L., and von Heijne, G. (1993). Predicting the topology of eukaryotic membrane proteins. Eur J Biochem 213, 1333–1340. Slavin, A. J., Johns, T. G., Orian, J. M., and Bernard, C. C. (1997). Regulation of myelin oligodendrocyte glycoprotein in diVerent species throughout development. Dev Neurosci 19, 69–78. Solly, S. K., Daubas, P., Monge, M., Dautigny, A., and Zalc, B. (1997). Functional analysis of the mouse myelin/ oligodendrocyte glycoprotein gene promoter in the oligodendroglial CG4 cell line. J Neurochem 68, 1705–1711. Solly, S. K., Thomas, J. L., Monge, M., Demerens, C., Lubetzki, C., Gardinier, M. V., Matthieu, J. M., and Zalc, B. (1996). Myelin/oligodendrocyte glycoprotein (MOG) expression is associated with myelin deposition. Glia 18, 39–48. Stammers, M., Rowen, L., Rhodes, D., Trowsdale, J., and Beck, S. (2000). BTL-II: A polymorphic locus with homology to the butyrophilin gene family, located at the border of the major histocompatibility complex class II and class III regions in human and mouse. Immunogenetics 51, 373–382. SteVerl, A., Brehm, U., Storch, M., Lambracht-Washington, D., Bourquin, C., Wonigeit, K., Lassmann, H., and Linington, C. (1999). Myelin oligodendrocyte glycoprotein induces experimental autoimmune encephalomyelitis in the ‘‘resistant’’ Brown Norway rat: Disease susceptibility is determined by MHC and MHC-linked eVects on the B cell response. J Immunol 163, 40–49. Stefferl, A., Schubart, A., Storch, M., Amini, A., Mather, I., Lassmann, H., and Linington, C. (2000). Butyrophilin, a milk protein, modulates the encephalitogenic T cell response to myelin oligodendrocyte glycoprotein in experimental autoimmune encephalomyelitis. J Immunol 165, 2859–2865. Steinman, L. (1993). Connections between the immune system and the nervous system. Proc Natl Acad Sci U S A 90, 7912–7914. Stevens, D. B., Chen, K., Seitz, R. S., Sercarz, E. E., Bronstein, J. M. (1999) Oligodendrocyte-speciWc protein peptides induce experimental autoimmune encephalomyelitis in SJL/J mice. J Immunol 162, 7501–7509. Storch, M. K., SteVerl, A., Brehm, U., Weissert, R., Wallstrom, E., Kerschensteiner, M., Olsson, T., Linington, C., and Lassmann, H. (1998). Autoimmunity to myelin oligodendrocyte glycoprotein in rats mimics the spectrum of multiple sclerosis pathology. Brain Pathol 8, 681–694. Sun, J., Link, H., Olsson, T., Xiao, B. G., Andersson, G., Ekre, H. P., Linington, C., and Diener, P. (1991). T and B cell responses to myelin-oligodendrocyte glycoprotein in multiple sclerosis. J Immunol 146, 1490–1495. Sun, D., Whitaker, J. N., Huang, Z., Liu, D., Coleclough, C., Wekerle, H., and Raine, C. S. (2001). Myelin antigen-speciWc CD8þ T cells are encephalitogenic and produce severe disease in C57BL/6 mice. J Immunol 166, 7579–7587. Targoni, O. S., and Lehmann, P. V. (1998). Endogenous myelin basic protein inactivates the high avidity T cell repertoire. J Exp Med 187, 2055–2063. Tejada-Simon, M. V., Hong, J., Rivera, V. M., and Zhang, J. Z. (2002). Skewed autoantibody reactivity to the extracellular domain of myelin oligodendrocyte glycoprotein in multiple sclerosis. Immunology 107, 403–410. Tsunoda, I., Kuang, L. Q., Theil, D. J., and Fujinami, R. S. (2000). Antibody association with a novel model for primary progressive multiple sclerosis: induction of relapsing-remitting and progressive forms of EAE in H2s mouse strains. Brain Pathol 10, 402–418. Vanguri, P., and Shin, M. L. (1986). Activation of complement by myelin: IdentiWcation of C1-binding proteins of human myelin from central nervous tissue. J Neurochem 46, 1535–1541. Vass, K., Heininger, K., Schafer, B., Linington, C., and Lassmann, H. (1992). Interferon-gamma potentiates antibody-mediated demyelination in vivo. Ann Neurol 32, 198–206. Vernet, C., Boretto, J., Mattei, M. G., Takahashi, M., Jack, L. J., Mather, I. H., Rouquier, S., and Pontarotti, P. (1993). Evolutionary study of multigenic families mapping close to the human MHC class I region. J Mol Evol 37, 600–612. von Budingen, H. C., Tanuma, N., Villoslada, P., Ouallet, J. C., Hauser, S. L., and Genain, C. P. (2001). Immune responses against the myelin/oligodendrocyte glycoprotein in experimental autoimmune demyelination. J Clin Immunol 21, 155–170. Waehneldt, T. V., Matthieu, J. M., and Jeserich, G. (1986). Major central nervous system myelin glycoprotein of the African lungWsh (Protopterus dolloi) cross-reacts with myelin proteolipid protein antibodies, indicating a close phylogenetic relationship with amphibians. J Neurochem 46, 1387–1391. Wallstrom, E., Khademi, M., Andersson, M., Weissert, R., Linington, C., and Olsson, T. (1998). Increased reactivity to myelin oligodendrocyte glycoprotein peptides and epitope mapping in HLA DR2(15)þ multiple sclerosis. Eur J Immunol 28, 3329–3335. Weerth, S., Berger, T., Lassmann, H., and Linington C. (1999). Encephalitogenic and neuritogenic T cell responses to the myelin-associated glycoprotein (MAG) in the Lewis rat. J Neuroimmunol 95, 157–164.

III. THE MYELIN GENES AND PRODUCTS

CONCLUSION

Weimbs, T., and StoVel, W. (1992). Proteolipid protein (PLP) of CNS myelin: Positions of free, disulWde-bonded, and fatty acid thioester-linked cysteine residues and implications for the membrane topology of PLP. Biochemistry 31, 12289–12296. Weissert, R., Kuhle, J., de Graaf, K. L., Wienhold, W., Herrmann, M. M., Muller, C., Forsthuber, T. G., Wiesmuller, K. H., and Melms, A. (2002). High immunogenicity of intracellular myelin oligodendrocyte glycoprotein epitopes. J Immunol 169, 548–556. Weissert, R., Wallstrom, E., Storch, M. K., SteVerl, A., Lorentzen, J., Lassmann, H., Linington, C., and Olsson, T. (1998). MHC haplotype-dependent regulation of MOG-induced EAE in rats. J Clin Invest 102, 1265–1273. Williams, A. F., and Barclay, A. N. (1988). The immunoglobulin superfamily–domains for cell surface recognition. Annu Rev Immunol 6, 381–405. Xiao, B. G., Linington, C., and Link, H. (1991). Antibodies to myelin-oligodendrocyte glycoprotein in cerebrospinal Xuid from patients with multiple sclerosis and controls. J Neuroimmunol 31, 91–96. Xu H., Foltz L., Sha Y., Madlansacay M. R., Cain C., Lindemann G., Vargas J., Nagy D., Harriman B., Mahoney W., Schueler, P.A. (2001). Cloning and characterization of human erythroid membrane-associated protein, human ERMAP. Genomics 76, 2–4. Ye, T. Z., Gordon, C. T., Lai, Y. H., Fujiwara, Y., Peters, L. L., Perkins, A. C., Chui, D. H. (2000). Ermap, a gene coding for a novel erythroid specific adhesion/receptor membrane protein. Gene 242, 337–345.

III. THE MYELIN GENES AND PRODUCTS

497

C H A P T E R

19 2’,3’-Cyclic Nucleotide 3’-Phosphodiesterase: Structure, Biology, and Function Peter E. Braun, John Lee, and Michel Gravel

INTRODUCTION AND HISTORICAL PERSPECTIVE In the four decades since the Wrst report of an enzymatic activity in the brain that hydrolyzes 2’,3’-cyclic nucleotides exclusively to nucleoside 2’-phosphate (Drummond et al., 1962), research progress to understand this unusual enzyme (2’,3’-cyclic nucleotide 3’-phosphodiesterase; CNP; CNPase; EC3.1.4.37) has been slow. The literature describing early studies showing that the CNP enzyme activity was largely associated with myelin, its subsequent puriWcation and characterization, and numerous observations prior to about 1990 have been thoroughly documented and reviewed (Sims and Carnegie, 1978; Sprinkle, 1989; Tsukada and Kurihara, 1992; Vogel and Thompson, 1988). Although many of these early observations relate to the appearance of CNP in oligodendrocytes and Schwann cells at early stages in the developing nervous system, no clear assignments of biological function have been possible, although potential roles have variously been speculated upon. This impasse derives from the fact that the preferred in vitro substrates, 2’,3’-cyclic nucleotides, are not normally observed in metabolite pools of tissues and cells. Several options can then be considered to explain the relevance of this enzyme: (1) the CNP protein serves a function(s) other than that dependent on the observed catalytic in vitro activity; (2) the catalytic site is relevant, but the physiological substrate remains to be discovered; (3) a combination of both of the preceding. Our objective here is to review scholarly advances in CNP research mainly since 1990, but earlier reports may be highlighted where they provide insights and directions for current and future endeavors.

Practical Utility of CNP Detection Following the discovery of an apparent association of CNP with myelin (Kurikara and Tsukada, 1967; Olafson et al., 1969) and the intriguing observations that it comprises 4% of total CNS myelin protein (but only ~ 0.4% in PNS myelin), numerous investigators employed subcellular fractionations, developmental studies, and mutant mice with myelin deWcits to Wrmly establish an avid association of CNP with myelinating cells and with the CNS myelin sheath. This tight binding behavior and speciWcity of expression has made CNP one of the markers of choice for the detection of oligodendrocytes and of myelin elements in tissue sections, cultured cells, and membrane preparations; CNP detection has also proven useful as an indicator of disease activity in neurological patients, where this protein often appears in sera and cerebrospinal Xuids.

Myelin Biology and Disorders, Volume 1

499

Copyright 2004, Elsevier (USA). All rights reserved.

500

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

The most commonly used assay for CNP activity is the spectrophotometric measurement of NADPH, arising from the hydrolytic cleavage of the 3’-phosphodiester bond in 2’,3’-cyclic NADP, followed by the coupling of the 2’-NADP product to the glucose 6-phosphate dehydrogenase reaction to yield NADPH (Sogin 1976); a modiWcation of this procedure to achieve greater sensitivity has also been reported (Stephon et al., 1992). Earlier, less facile assays were reviewed by Sims and Carnegie (1978). The presence of CNP in cells, tissues, and body Xuids is also readily detected by Western blots, immunological assays, or by immunohistochemical procedures that employ monoclonal antibodies; this is currently one of the best ways to detect and identify oligodendrocytes in mixed cell populations. The most useful of these immunoprobes are commercially available from Sternberger Monoclonals, Inc.

Nonmyelin CNP Although the abundance of CNP in myelin and the exclusive location in oligodendrocytes of the CNS caused attention to be focused primarily on these targets, it was always apparent that while CNP is not ubiquitous, some other cell types also express it, albeit at levels at least 10-fold lower than oligodendrocytes. Earlier, CNP was shown to be present in vertebrate erythrocytes (Weissbarth et al., 1981), lymphocytes and platelets (Rastogi and Clausen, 1985; Sheedlo, et al., 1984; Sprinkle et al., 1985), thymus (Bernier et al., 1987), liver (Dreiling et al., 1981a,b), and photoreceptor cells of the retina (Giulian and Moore, 1980; Giulian et al., 1983; Kohsaka et al., 1983; Nishizawa et al., 1982). More recently, there have been reports of CNP in cells of the adrenal medulla (McFerran and Burgoyne, 1997) and in Leydig cells and Sertoli cells of the testis (DavidoV et al., 1997, 2002). Mostly these have been observations on measurable CNP enzyme activity, detection by Western blots, or by immunocytochemical staining, but these studies have not addressed questions of CNP function.

A CNP Homolog in Fish The discovery of a CNP homolog in Wsh optic nerve has been of considerable interest and has helped shape the direction of research on the catalytic function of CNP. Regeneration Induced CNP Homolog (RICH) was Wrst identiWed in axotomized goldWsh retinal ganglion cells in which the process of regeneration was accompanied by a dramatic (~ four-fold to eight-fold) increase in the expression of this protein; two bands of RICH (2 isoforms) were evident by gel electrophoresis, at 68 kDa and 70 kDa. While the protein and its message were most abundant in the regenerating retina, the mRNA was detectable in brain, heart, liver, muscle, testis, and especially in kidney and ovary. The puriWed protein as well as recombinant RICH were found to be enzymatically active, with kinetic parameters similar to CNP, and the amino acid sequence (411 residues; Mr 45,280 Da for gRiCH68 isoform) revealed regions of homology with CNP only in the C-terminal domain (Fig. 19.1) (Ballestero et al., 1995, 1997; Leski and AgranoV, 1994; Wilmot et al., 1993). Subsequently, RICH was also cloned from and described in regenerating zebra Wsh optic nerve (Ballestero et al., 1999). Shared structural features of RICH and CNP are discussed later in this chapter, in the sections titled ‘‘Primary Structural Motifs of CNP and RICH’’ and ‘‘ThreeDimensional Structural Features,’’ in the context of functional implications for this catalytic activity in evolutionarily divergent species.

BIOCHEMICAL AND STRUCTURAL CHARACTERISTICS OF CNP AND RICH Properties of the Isolated Protein CNP can be extracted from brain or myelin in a variety of ways, using combinations of nondenaturing detergents and salts, and it has been puriWed by several procedures

III. THE MYELIN GENES AND PRODUCTS

501

BIOCHEMICAL AND STRUCTURAL CHARACTERISTICS OF CNP AND RICH

Box I Rat Mouse Human Bovine Chicken Bullfrog g-RICH68

MSSSGAKDKPE-LQFPFLQDEDTVATLHECKTLFILRGLPGSGKSTLARLIVEKYHNGTKMVSADAYKIIP LQFPFLQDEDTVATLHECKTLFILRGLPGSGKSTLARLIVEKYHNGTKMVSADAYKIIP MSSSGAKEKPE-LQFPFLQDEDTVATLHECKTLFILRGLPGSGKSTLARLILEKYHDGTKMVSADAYKIIP MSSSGAKDKPE-LQFPFLQDEDTVATLLECKTLFILRGLPGSGKSTLARVIVDKYRDGTKMVSADAYKITP MSSSGAKDKPE-LQFPFLQDEETVATLQECKTLFILRGLPGSGKSTLARFIVDKYRDGTKMVSADSYKITP MSAQAAKERPDSLRFPFLDDEETIATLRESKTFFILRGLPGSGKSTLAQAIQERYRDGCKVIAAENYKITP MSSQASKDQLEGQKPPLLVDDHTVVTVRESKVLVLLRGLPGSGKSTLAKDIELKYKETSRLFSADQYEIKP MDAEQNQEVPE-----AVAETQEVAMKQEEKVESKEVAPSEPEKTPETEHSAGEMPEKEKAMDSKAPPAKP

70 70 70 70 71 71 66

Rat Mouse Human Bovine Chicken Bullfrog g-RICH68

GSRADFSEEYKRLDEDLAGYCRR---------DIRVLVLDDTNHERERLDQLFEMADQYQYQVVLVEPKTA GSRADFSEAYKRLDEDLAGYCRR---------DIRVLVLDDTNHERERLDQLFEMADQYQYQVVLVEPKTA GARGAFSEEYKRLDEDLAAYCRRR--------DIRILVLDDTNHERERLEQLFEMADQYQYQVVLVEPKTA GARGSFSEEYKQLDEDLAACCRR---------DFRVLVLDDTNHERERLEQLFELADQYQYQVVLVEPKTA AVRSGVPEEYGKVDEDLVEYCKR---------DVSVIVLDDTHHERERLDQIFDIADKYRYKVIFAEPKTQ VIRSSSGGDYTKLDDELTTCFERR--------EANLVILDDTHHDHERLDELFDLANKYHFTVVILEPKTP SEPEVAPEKSPEETPAAESSAKPPEPEQKKSEEPPVQVNSEPEKQEEEAVKEAESKPTAVNEAKPEESDKD

132 132 133 132 133 134 137

Rat Mouse Human Bovine Chicken Bullfrog g-RICH68

WRLDCAQLKEKNQWQLS---LDDLKKLKPGLEKDFLPLYFGWFLTKKSSETLRKAGQVFLEELGNHKAFKK WRLDCAQLKEKNQWQLS---ADDLKKLKPGLEKDFLPLYFGWFLTKKSSETLRKAGQVFLEELGNHKAFKK WRLDCAQLKEKNQWQLS---ADDLKKLKPGLEKDFLPLYFGWFLYKKSSETLRKAGQVFLEELGNHKAFKK WRLDCAQLKEKNQWQLS---ADDLKKLKPGLEKDFLPLYFGWFLTKKSSAALWKTGQTFLEELGNHKAFKK WRMDCAQLKDKNQWKLT---AEDLKKMKPSLEKEFLPMYFGWFLSKRSSEILRKAGQVFLDELGSLKAFKK WRLDCAQLKDRNHWKLS---LEELKNLRPSLEKDLLPLYGFWFLAKRDEDSLRKTSHEFLEQLGNLKAFKK EKTKTEGGEEKVQPEADGVKAEPLKETETKQKEPELPLFFGWFLLPEEEERIKCATMDFLKTLDTLEAFKE

200 200 201 200 201 202 208

Rat Mouse Human Bovine Chicken Bullfrog g-RICH68

ELRHFISGDEPKEKLDLVSYFGKRPPGVLHCTTKFCDYGKATGAEEYAQQDVVRRSYGKAFKLSISALFVT ELRHFISGDEPKEKLELVSYFGKRPPGVLHCTTKFCDYGKAAGAEEYAQQEVVKRSYGKAFKLSISALFVT ELRQFVPGDEPREKMDLVTYFGKRPPGVLHCTTKFCDYGKAPGAEEYAQQDVLKKSYSKAFTLTISALFVT ELRHFVSGDEPREKIELVTYFGKRPPGVLHCTTKFCDYGKAAGAEEYAQQDVVKKSYCKAFTLTISALFVT ESKYFTS-EDPKIKIDLTSYFVKRPPGVLHCTTKYTDFGKAAGAEDYAQQEAVRASYGKGFTLSISGLFVT RLQAYGY--EDKHKLDLLKHFAKTP-NILHCTSKFCDYGKAAGSEEYSRQEVVKKSYSKGYTLHITSLFAT HISEFTG--EAEKEVDLEQYF--QNPLQLHCTTKFCDYGKAEGAKEYAELQVVKESLTKSYELSVTALIVT

Rat Mouse Human Bovine Chicken Bullfrog g-RICH68

PKTAGAQVVLNEQELQLWPSDLD-------KPSSSESLPPGSRAHVTLGCAADVQPVQTGLDLLEILQQVK PKTAGAQVVLTDQELQLWPSDLD-------KPSASEGLPPGSRAHVTLGCAADVQPVQTGLDLLDILQQVK PKTTGARVELSEQQLQLWPSDVD-------KLSPTDNLPRGSRAHITLGCAADVEAVQYGLDLLEILRQEK PKTTGARVELSEQQLALWPNDVD-------KLSPSDNLPRGSRAHITLGCAGDVEAVQTGIDLLEIVRQEK TKTVGARVELSEQQLLLWPGDAD-------KLQATDSLPKGSRAHITLGCASGVEAVQTGMDLLEFVKLEK PRTVGARVELTEEQLLLWPLDAER------EVMPKDTFPRGSRAHLTLGTAAEVQDVQTGIDLLEFIKIQQ PRTFGARVALTEAQVKLWPEGADKEGVAPALLPSVEALPAGSRAHVTLGCSAGVETVQTGLDLLEILALQK

Rat Mouse Human Bovine Chicken Bullfrog g-RICH68

GGSQGEEVGELPR---GKLYSLGKGRWMLSLAKKMEVKAIFTGYYGK-GKPVPVHGS-RKGGAMQICTII GGSQGEAVGELPR---GKLYSLGKGRWMLSLTKKMEVKAIFTGYYGK-GKPVPIHGS-RKGGAMQICTII GGSRGEEVGELSR---GKLYSLGNGRWMLTLAKNMEVRAIFTGYYGK-GKPVPTQGS-RKGGALQSCTII GGSRGEEVGELSR---GKLYSLGSGRWMLSLAKKMEVRAIFTGYYGK-GKAVPIRSG-RKGGSFQSCTII AGSKGEEVGEIGG---GKLQYFGNGMWMLTLSKKIDVRAIFSGYYGK-GKLVPTQSTNKRGSVFSSCTIN AGRDGESLGELTGSVAGKVSYFDNGMWMVNLARKIEVKSIFSGYYGKPGVNVPLRGS--KKGLLHQCHIM EGKEGTQVEMDLG----TLTYLSEGRWFLALREPINADTTFTSFSED-KPATSDQGKKNGEKKKKKCTIL

Box II

Box III

Box IV 271 271 272 271 271 270 275

Box V 335 335 336 335 335 335 346

Box VI 400 400 401 400 400 403 411

FIGURE 19.1 Amino acid sequences of CNP1 from bullfrog, chicken, mouse, rat, bovine, and human, and of goldWsh RICH. Boxes I and II: ATPase ‘‘A box’’ and ‘‘B box,’’ respectively. Box III: Adenine recognition motif. Boxes IV and V: The his-containing tetrapeptide motifs essential for catalytic activity. Box VI: Isoprenylation motif; the underlined sequence contains the catalytic domain common to all members of this family.

(reviewed in Sprinkle, 1988; Tsukada and Kurihara, 1992). It’s worth noting that both isoforms, being isoprenylated and palmitoylated, bind tightly to membranes and complete solubilization is diYcult. However, elimination of the C-terminal isoprenyl group by inhibition of the isoprenoid biosynthetic pathway (Braun et al., 1991) or by mutation of the isoprenylated cysteine (DeAngelis and Braun, 1996a,b) allows the protein to be easily solubilized. Equally, when CNP is expressed recombinantly in E. coli (Lee et al., 2001) the protein is readily solubilized and puriWed because it lacks the lipidic C-terminal modiWcation. For many studies, the preparation of recombinant CNP is therefore a convenient alternative to the relatively arduous puriWcation from myelin.

III. THE MYELIN GENES AND PRODUCTS

502

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

The puriWed protein from brain is highly basic, with an isolectric point close to 9. Two isoforms of ~ 46 kDa and 48 kDa are observed in all species studied, but the relative electrophoretic migration pattern by SDS-PAGE diVers slightly among diVerent species (Waehneldt and Malotka, 1980), despite a high degree of sequence homology; the reasons for this are not known. It has been noted that the isolated protein has a propensity to selfassociate under some conditions, leading to observation of dimers on occasion (reviewed in Tsukada and Kurihara, 1992). More extensive aggregates of dimers, trimers, and tetramers have also been observed by cross-linking studies (Braun et al., 1990), but it isn’t clear that this property has biological relevance. The dimers that are sometimes observed by SDS-PAGE may, in fact, arise from nonspeciWc assembly of two monomers during the denaturation of the native protein (Hofmann et al., 2000). Determination of the solution structure by NMR of recombinant rat CNP catalytic fragment (residues 164–378) yielded no evidence for dimerization (Kozlov et al., 2003).

Comparison of Two Isoforms Oligodendrocytes, Schwann cells, and myelin from all species examined manifest two isoforms designated CNP1 (~ 46 kDa) and CNP2 (~ 48 kDa). These have identical amino acid sequences except for a 20-residue extension at the N-terminus that is exclusive to CNP2 (Gravel et al., 1994; Kurihara et al., 1990, 1992). One gene (Douglas et al., 1992; Monoh et al., 1989) consisting of four exons and two promoters gives rise to two RNA transcripts, one of which can generate both proteins (see the section titled ‘‘CNP Gene Structure and Regulation’’). Both polypeptides are synthesized on free ribosomes (Gillespie et al., 1990), both are enzymatically active, and both are post-translationally modiWed by isoprenylation at the C-terminus (DeAngelis and Braun, 1994; 1996a, 1996b; O’Neill and Braun, 2000).

Primary Structural Motifs of CNP and RICH The amino acid sequences of CNP, as deduced from cDNAs for bullfrog, chicken, mouse, rat, bovine, and human, are shown in Figure 19.1; the sequences for RICH from goldWsh is also shown. Although sequence identity or homology is relatively strong among all CNP species, the N-terminal third of the RICH proteins bears no resemblance to CNP; the C-terminal twothirds, however, has substantial homology to CNP, and contains the catalytic domain (see the section title ‘‘Enzymatic Characteristics and Reaction Mechanism’’ ). All CNP species, but not RICH, have strongly conserved consensus motifs for the potential binding of nucleotide phosphoryl groups both in the N-terminal and C-terminal domains (Fig. 19.1). The ATPase ‘‘A box’’ (G/AXXXXGKT/S) and ‘‘B box’’ (XXXXD, where X ¼ hydrophobic residues), along with the adenine recognition motif (YFGKRPPG), are common to many nucleotide binding proteins and enzymes. Interestingly, since they fall outside the essential catalytic domain of CNP (Lee et al., 2001) and are not required for the catalytic activity of RICH, they must serve another nucleotide-related function. ATPase activity has not been detected in puriWed preparations of CNP or in the recombinant protein. Although some preliminary experiments suggest that ATP (and other nucleoside triphosphates) may bind to CNP, deWnitive experiments have not yet been done. By alignments of nucleotide binding motifs with other enzymes, it was noted that CNP shares modest sequence similarities with polynucleotide kinase (Gravel et al., 1994; Koonin et al., 1990), and it was suggested that CNP might also possess this activity. This possibility was subsequently tested with catalytically active recombinant CNP in a polynucleotide kinase assay, with no such activity being evident (Prinos et al., 1995). In another study, Kasama-Yoshida et al. (1997) observed that the catalytic domain of CNP and RICH shared a motif that strongly resembled the active site of b-ketoacyl synthase (bKAS). More recently, Ballestero et al. (1999) tested a possible relationship of CNP/RICH to this enzyme by making site-directed mutants to evaluate the role of cysteines known to be critical for bKAS activity. They found that the cysteines did not, in fact, participate in CNPase activity.

III. THE MYELIN GENES AND PRODUCTS

BIOCHEMICAL AND STRUCTURAL CHARACTERISTICS OF CNP AND RICH

The C-terminal isoprenylation motifs, present in all species of CNP and RICH, have been the focus of some attention. The post-translational modiWcation of a speciWc cysteine at or near the C-terminus by the isoprenoid lipids farnesyl (C-15) or geranylgeranyl (C-20) moieties in thioether linkage occurs in numerous proteins, most of which are members of signal-transducing pathways. CNP and RICH are typical of a large category of proteins bearing a cys-A1,A2-X motif in which A is an aliphatic amino acid and X is any amino acid; these are removed once the cys has become isoprenylated. The terminal carboxyl then may become modiWed by methylation (Sinensky 2000). It was shown that isoprenylation mediates the binding of CNP to membranes, explaining the avid association with myelin in the absence of a protein transmembrane domain (Braun et al., 1991; DeAngelis and Braun, 1994). Further, the C-T-I-I motif common to the mammalian CNPs resulted in modiWcation by either farnesyl or geranylgeranyl in vitro; it remains to be determined which of these (if not both0 is present in native myelin, or for that matter in CNP-expressing cells (DeAngelis and Braun, 1994). Cox et al. (1994) determined that the isoprenylated cysteine in CNP undergoes the additional carboxylmethylation step. The consequences of this for CNP biology was not established, but a body of evidence suggests that there are biological consequences for some proteins so modiWed, and that methylation greatly enhances the association of farnesylated peptides with lipid bilayers (reviewed in Sinensky, 2000). The potential biological signiWcance of CNP isoprenylation was underscored by comparison of transfected ‘‘naive’’ cells (e.g., NIH-3T3; HeLa) with cDNAs encoding either the ‘‘wild-type’’ CNP1 or a CNP1 mutant in which the critical cys (at position 397) was replaced by the nonprenylatable ser. Expression of CNP1 in these nonglial cells dramatically altered their morphology by inducing myriad CNP-enriched Wlopodia and large process extensions, with CNP concentration also occurring at cell margins. On the other hand, transfection of the C397S mutant did not induce these changes, and the nonisoprenylated protein was diffusely distributed throughout the cytoplasm (DeAngelis and Braun, 1994). Implications of these observations for the role of CNP in myelinogenesis are discussed in the section titled ‘‘Functional SigniWcance of CNP.’’ ModiWcation of CNP by fatty acylation (palmitoylation) has also been demonstrated, although the site for palmitoylation (a cysteine residue) has yet to be identified (Agrawal et al., 1990b). Further, acylation occurs independently of isoprenylation since the C397S mutant CNP is just as strongly palmitoylated as the wild-type protein (DeAngelis et al., 1994). The addition of these hydrophobic groups has signiWcance for the targeting to and binding of proteins to speciWc membrane microdomains (‘‘lipid rafts’’) that are rich in glycosphingolipids and cholesterol and that are believed to have important roles in the organization of signaling complexes (Kim and PfeiVer, 1999). The 20 amino acid domain at the N-terminus of CNP2 is highly conserved among all species studied (Fig. 19.2). A recent study focused on the biochemical and biological signiWcance of this domain and provided evidence to show that it serves to target CNP2 to mitochondria (O’Neill, Ph.D. thesis, McGill University, 2002; to be published). Its structure meets the basic requirements for a mitochondrial targeting signal in that it has a net positive charge, some hydroxylated and hydrophobic residues, lacks acidic residues, and is predicted to form an a-helix. Immunocytochemistry and subcellular fractionation experiments with cultured oligodendrocytes and naı¨ve cells (HeLa S3) that had been transfected with a CNP2-expressing vector showed that CNP2, but not CNP1, was localized to mitochondria. This served to explain why CNP had previously been observed 1

Rat Mouse Bovine Human

10

20

MSTSFARKSHTFLPKIFFRKMSSS MNTSFTRKSHTFLPKLFFRKMSSS MSRGFSRKSQTFLPKVFFRKMSSS MNRGFSRKSHTFLPKIFFRKMSSS S9

S22

FIGURE 19.2 Amino acid sequences of the N-terminal domain of CNP2 from rat, mouse, bovine, and human. Putative kinase recognition motifs are underlined and the known phosphorylatable serine residues are indicated.

III. THE MYELIN GENES AND PRODUCTS

503

504

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

associated with mitochondria in adrenal cells (McFerran and Burgogne, 1997). Further, it was established that this domain (referred to as N2 for N-terminal domain of CNP2) was suYcient for mitochondrial localization, in that a fusion protein of N2 with green Xuorescent protein (GFP) was eYciently targeted to mitochondria. Protease digestion of CNP2 in intact mitochondria isolated from transfected cells, as well as in vitro, import assays, and metabolic pulse-chase labeling experiments indicated that CNP2 is not imported into the mitochondria (O’Neill, Ph.D. thesis, McGill University, 2002; to be published). It is more likely that CNP2 is peripherally associated with the outer mitochondrial membrane. The N2 domain is capable of being phosphorylated on serines 9 and 22 (O’Neill and Braun, 2000). In oligodendrocytes phosphorylation of serine 9 occurs after treatment with phorbol ester, a known activator of protein kinase C, whereas serine 22 (putative PKA phosphorylation site) is constitutively phosphorylated. CNP2 may either be phosphorylated by protein kinase C and/or A (Agrawal et al., 1990a, 1994; Bradbury et al., 1984; Bradbury and Thompson, 1984; Vartanian et al., 1988; 1992). This poses the question of a potential role for phosphorylation in regulating the binding of CNP2 to mitochondria.

Enzymatic Characteristics and Reaction Mechanism of CNPase The kinetic parameters of CNPase activity have been amply reviewed previously (Sprinkle, 1989). To summarize brieXy, all 2’,3’-cyclic mono-nucleotides can be hydrolyzed by CNPase to yield 2’-nucleotides exclusively. Linkages of these substrates at the 5’ hydroxyl to other substitutents, such as 2’, 3’-cyclic NADP and oligonucleotides, does not appear to aVect activity. CNP does not act upon 3’, 5’-cyclic mononucleotides (cAMP, cGMP), and no other cyclophosphate rings, either on ribose, hexoses, or inositol, are cleaved by this enzyme. Although physiologically relevant substrates with 2’,3’-cyclic termini have not yet been reported, numerous cyclic phosphate-containing RNAs that are generated as intermediate products of splicing reactions are present in most cells. Although a range of kinetic parameters have been reported, the most rigorous study of CNP to date shows that puriWed recombinant CNP1 hydrolyzes cyclic NADP (in the assay of Sogin, 1976) with a Km of 0.26 mM and a kcat of 836 s
1 at 258C (Lee et al., 2001). In close agreement, a Km of 0.32 mM (for 2’,3’-cyclic AMP) has been reported for recombinant RICH (Ballestero et al., 1999). The catalytic domain resides in the C-terminal region comprising two-thirds of the protein, the region that has signiWcant homology with RICH (Ballestero et al., 1999; Lee et al., 2001). Mapping of the catalytic domain by deletion mutant analysis, by chemical modiWcation studies of speciWc amino acids and by sitedirected mutagenesis have provided new insights. Contrary to earlier reports, cysteine residues have nonessential roles for enzymatic activity. On the other hand, two speciWc histidines (positions 230 and 309 in CNP1) are essential for catalytic activity (Lee et al., 2001); they are part of two tetrapeptide motifs, H-X-T/S-X (where X is often a hydrophobic residue), that are essential for the enzymatic activity of the other three classes of related enzymes (Hofmann et al., 2002) (see the section titled ‘‘Three-Dimensional Structural Features’’). Details of the enzyme mechanism can be deduced for CNP and for RICH by comparison to the mechanism of activity of the plant cyclic phosphodiesterase that hydrolyzes ADP-ribose 1’,2’-cyclic phosphate, an enzyme whose crystal structure has been determined (Hofmann et al., 2000; Kovlov et al., 2003).

Three-Dimensional Structural Features The solution structure of the catalytic core fragment of CNP (residues 163 to 378 of rat CNP1) has been determined by NMR (Kozlov et al., 2002, 2003). Interestingly, the folded structure is remarkably similar to two other proteins with known structures: plant cyclic nucleotide phosphodiesterase (CPD) from Arabidopsis thaliana, (an enzyme involved in the tRNA splicing pathway, known to hydrolyze ADP-ribose 1’,2’-cyclic phosphate) (Hofmann et al., 2000), and most recently, bacterial 2’-5’ RNA ligase from Thermus thermophilus (an enzyme that ligates tRNA half-molecules containing 2’, 3’-cyclic phosphate and 5’-hydroxl termini) (Kato et al., 2003). All three structures show a bilobal

III. THE MYELIN GENES AND PRODUCTS

BIOCHEMICAL AND STRUCTURAL CHARACTERISTICS OF CNP AND RICH

FIGURE 19.3 The structure of the catalytic domain of CNP shows a bilobal arrangement of two modules. The N-terminal and C-terminal lobes are shown in green and blue, respectively. The mobile loop is in red. (A) The secondary structure topology shows that each lobe consists of four beta-strands and two alpha helices. (B) In the topological ribbon representation, the catalytic residues His230, Thr232, His309, and Thr311 are shown explicitly. The mobile loop is proximal to the active site and may act as a Xap upon binding of the substrate.

arrangement of two modules, each consisting of a four-stranded antiparallel b-sheet and two antiparallel a-helices located on the outer part of the modules (Fig. 19.3). The inner region of the modules forms a large water-Wlled cavity, but a more open conformation than that in CPD. The previously identiWed (Lee et al., 2001) catalytically important histidines (His 230; His 309), along with two important threonines (Thr 232; Thr 311), are located in close proximity to each other within the active site. These residues are present as two similarly spaced tetrapeptide motifs (H-X-T/S-X), which, in addition to CNP/RICH, are remarkably found in three other groups of sequence divergent enzymes: 1) fungal/plant RNA ligases 2) bacterial and archaeal RNA ligases (including 2’-5’ RNA ligase from T. thermophilus), and 3) plant/yeast cyclic phosphodiesterases (including CPD from A. thaliana). Collectively, these four groups of enzymes appear to constitute a super-family of proteins, with enzymatic activities not yet well understood in a physiological context, but with an apparent link (inferred for CNP and RICH) to RNA metabolism. This inference for CNP is further strengthened by the observation that the b-sheet surface in the active site has an abundance of positively charged and aromatic residues, a common feature of RNA-binding proteins. Coupled with a large binding cavity, this suggests that RNA could be a substrate for CNP (Kozlov et al., 2003). In preliminary experiments it has been demonstrated that CNP does bind to RNA in vitro, suggesting that CNP may play a role in RNA metabolism (M. Gravel, personal communication). This may explain why CNP is found to become associated with nucleocapsids of some RNA viruses, as described in the section titled ‘‘Functional SigniWcance of CNP.’’

Cytoskeletal Associations and Membrane-Binding Properties of CNP The avid binding of CNP to membranes is not a function of a speciWc polypeptide domain, but rather it depends on the post-translationally added C-terminal polyisoprenoid, as well as one or more palmitoyl groups on cysteines. Kim and PfeiVer (1999) have demonstrated that there are likely to be two ‘‘pools’’ of membrane-bound CNP in myelin; one of them, comprising 40% of the total, being an association with detergent-insoluble glycosphingolipid cholesterol enriched micro-domains (lipid rafts). Not only was nearly half of the CNP of puriWed myelin insoluble in TX100 at 48C (an operational deWnition of lipid rafts), but this binding of CNP to speciWc bilayer microdomains was further established by several other criteria. They showed that the detergent-insoluble complexes could be separated from other myelin membrane complexes by centrifugal Xotation on a density gradient and by dissociation of CNP via TX100 extraction at an elevated temperature (378); or

III. THE MYELIN GENES AND PRODUCTS

505

506

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

CNP could be released by disruption of cholesterol aggregates in the rafts using a cholesterol-binding agent (saponin). Interestingly, it was reported that both isoforms of CNP behaved similarly, suggesting that CNP2, despite its aYnity for mitochondria in the cell, also associates with lipid rafts in membranes of the cytoplasmic compartments of the sheath. Earlier studies (Gillespie et al., 1989; Pereyra et al., 1988; Wilson & Brophy, 1989) on extraction/solubilization properties of proteins, including CNP, from myelin and from oligodendrocytes, using detergents and conditions generally believed to preserve cytoskeletal structures as insoluble aggregates, led to the suggestion that TX100-insoluble (at room temperature) CNP was binding mainly to the actin-based cytoskeleton (de Angelis et al., 1996). This interpretation accorded well with cell biological observations showing intracellular co-localization of CNP with actin-and tubulin-based networks (Dyer and Benjamins, 1988, 1989; Dyer et al., 1994, 1995). It is also consistent with more recent demonstrations of an association of CNP with microtubules (Bifulco et al., 2002; Laezza et al., 1997) and of CNP binding to tubulin (J. Lee, McGill University, to be published). However, none of the above observations preclude the additional participation of CNP in molecular complexes that comprise the lipid bilayer microdomains described by Kim and PfeiVer (1999). Indeed, the data compel consideration of both scenarios. Immunocytochemical visualization of CNP in ODCs and in model cells induced to express CNP shows colocalization of CNP with cytoskeletal elements in the cell soma and in the cell processes, as well as a concentration of CNP at loci in the cell margins. It’s worth noting that the abundant CNP within the whole sheath comprises 4% of total protein and is localized exclusively to the cytoplasm-containing compartments of the sheath (Trapp et al., 1988). Here, the CNP could comprise > 40% of the total protein if these compartments account for 10% or less of the total sheath protein complement; it could be argued that lipid raftassociation cannot likely account for all of the CNP here, and that some of it is associated with other elements in these compartments. In a diVerent approach, Lintner and Dyer (2000) demonstrated that in membrane sheets of cultured rat ODCs cholesterol could be induced to redistribute in the membrane by exposure of the cells to antibodies against either galactocerebroside or to myelin-oligodendrocyte speciWc protein (MOSP). The redistributed cholesterol appeared to co-localize with vein-like structures within the membrane sheet; these also co-stained for CNP and tubulin or actin. Alternatively, disruption of microtubules with colchicine resulted in a redistribution of cholesterol such that it now co-localized with actin microWlaments and, it was suggested, with regions of CNP concentrations. Although insuYcient data exist to attribute these induced perturbations of cholesterol to reorganizations of lipid rafts, it is plausible that cholesterol- and CNP-containing structures participate in some as-yet-to-bedeWned signaling complexes within myelin-associated cytoskeletal structures. Thus far, no clear vision has emerged to explain mechanistically in which signaling cascade CNP might participate.

CNP GENE STRUCTURE AND REGULATION The CNP Gene and Its Expression CNP is found most abundantly in the CNS of vertebrates, especially amphibia and higher forms. Fish have relatively low CNPase activity, while frogs and tadpoles have high levels, possibly even higher than those in mammals. cDNAs of chicken, bullfrog, mouse, rat, bovine, and human CNP have been isolated and the nucleotide sequences determined. Sequence comparisons among species reveals that the variation among diVerent mammals is relatively small. For example, the degree of homology is approximately 85% between mouse and human CNPs. In contrast, sequence variation between bullfrog and chicken compared to mammals in general is much greater. The degree of homology between chicken and human CNPs is 66%, but only 54% between bullfrog and human (KasamaYoshida et al., 1997).

III. THE MYELIN GENES AND PRODUCTS

CNP GENE STRUCTURE AND REGULATION

Structure analysis of cDNAs and genomic regions from rat, mouse, bovine, and human suggest that the overall organization of the gene is very similar in all species (Bernier et al., 1987; Douglas et al., 1992; Gravel et al., 1994; Monoh et al., 1993; Tsukada and Kurihara, 1992). In mouse and human, the CNP gene is about 6 to 9 KB long and comprises four exons. Two genetic loci have been identiWed for the mouse CNP gene from mapping recombinant inbred strains and a limited somatic cell hybrid panel. One of these loci is located on chromosome 11 and is closely linked to the GFAP gene (Bernier et al., 1988). This locus appears to contain all the necessary elements to express both CNP isoforms, and its structure seems to be identical to that of the CNP gene isolated from a genomic library. The second locus maps to chromosome 3, but no functional signiWcance has been ascribed to it. This was initially considered to be a pseudogene, or perhaps an active gene encoding a protein partly homologous to CNP (Bernier et al., 1988), but it has since been reported that it lacks sequences homologous to the 5’-end of the CNP gene, including part of the second exon (Bernier, L., personal communication). Further, no CNPase activity or immunoreactive signals using polyclonal antibody against CNP have been found in brain of CNP null mutant mouse (Lappe-Siefke et al., 2003, see the section titled ‘‘Functional SigniWcance of CNP’’). In this mouse, generated by a knock-in strategy, a large fragment of the CNP gene has been replaced by a mini-gene cassette (containing sequences for the CRE recombinase and the puromycin resistance gene) to destroy the coding potential of the CNP gene. Because of the structure of the targeting vector, only the CNP locus on chromosone 11 can be modiWed by recombination, leaving the second CNP locus on chromosome 3 intact. In humans, however, only one CNP locus has been found on chromosome 17, which provides further evidence of homology between human chromosome 17 and mouse chromosome 11 (Douglas et al., 1992; Sprinkle et al., 1992). Since there is no evidence to the contrary, we suggest that the second CNP locus on the mouse chromosome 3 is not functional. The existence of a single functional gene in both mouse and human supports the idea that both CNP isoforms are encoded by the same gene (Fig. 19.4). The two AUG codons respon-

A P2

P1

Exon 0

Exon 1 AUG

CNP1 mRNA (2.6kb) AUG

AUG

CNP2 mRNA (2.4kb)

5’ untranslated region shared coding region CNP2 unique coding region

B 5’…CUCAUCAUGAGCUUUGCCCGAAAAAGCCACACAUUCCUGCCC AAGAUCUUCUUCAGAAAAAUGUCA…3’

FIGURE 19.4 Representation of the 5’-end of the mouse CNP gene and the resulting primary transcripts. (A) The Wrst two exons of the CNP gene are shown here with the positions of the distal (P2) and proximal (P1) promoters. The larger of the two transcripts, CNP1 mRNA, is generated from P1 whereas the transcription from P2, followed by an additional splicing event within the non-coding region of exon 1 produces the CNP2 mRNA. All coding information for the unique N-terminal extension of CNP2 polypeptide, except for the CNP2 AUG, is contained within exon 1. The remaining part of the coding sequence for isoforms is contained in the last two exons (not shown). (B) Sequence of the bifunctional CNP2 mRNA surrounding the two in-frame initiation sites (italic and underlined). Translational initiation at the Wrst and second AUGs will produce CNP2 and CNP1 polypeptides, respectively.

III. THE MYELIN GENES AND PRODUCTS

507

508

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

sible for the translation of CNP2 and CNP1 polypeptides are, in fact, located in the Wrst and second exons, respectively. Separate promoter regions control the expression of both transcripts. The CNP1 mRNA (2.6 kb) is transcribed from the proximal promoter (P1) located between the Wrst two exons, and contains only the open reading frame for the CNP1 isoform. The distal promoter (P2), upstream of the Wrst exon, regulates the transcription of the CNP2 mRNA (2.4 kb). The primary transcript originating from the distal promoter undergoes an additional splicing event to join sequences encoded by the Wrst exon to sequences located 57 base pairs upstream of the CNP1 start codon in the second exon (Douglas and Thompson, 1993; Monoh et al., 1989), adding 60 base pairs of coding sequence. The resulting CNP2 transcript with its shorter 5’-end contains the two AUG start codons, in frame, from the Wrst and second exons, and therefore has the capability to produce both CNP isoforms (O’Neill et al., 1997). At the primary sequence level of the protein, both CNP isoforms are identical except for the 20-amino acid extension at the N-terminus of CNP2 isoform (Gravel et al., 1994, see the section titled ‘‘Comparison of Two Isoforms’’). The production of two protein isoforms from one gene is not novel. During evolution, diVerent organisms or cell types have devised ways of using similar proteins to perform multiple tasks. As is the case for CNP, the addition of an N-terminal extension is a common strategy for generating two isoforms. This may be done by alternative splicing of a gene, as has been demonstrated for CNP, or by generating one mRNA containing two in-frame translation initiation sites (Acland et al., 1990; Delmas et al., 1992; Descombes and Schibler, 1991; Lock et al., 1991). The latter possibility, however, does not accord with the ‘‘scanning’’ model, a general concept for which the AUG nearest the 5’-end of the mRNA is the unique site for the initiation of translation. According to this model, the 40S ribosomal subunit enters at the 5’-end of the mRNA and migrates linearly until it encounters the Wrst AUG codon. Then the large 60S ribosomal subunit is recruited, and the translational process begins (Kozak, 1989). However, it appears that if the Wrst initiation codon within the mRNA occurs in a suboptimal context, some 40S ribosomal subunits will bypass the Wrst AUG and initiate translation at a downstream site, that is in a more optimal context (Kozak, 1995, 1997). Accordingly, this leaky scanning may result in two independently initiated polypeptides produced from one message (Kozak, 1995). Additionally, other factors that inXuence the decision of the ribosome to bypass an AUG start codon have been reported. These include RNA secondary structure downstream of the codon and proximity of the AUG to the 5’-end of the mRNA. Strong downstream secondary structure increases recognition of an AUG whereas close proximity to the 5’-end decreases recognition (Kozak, 1990, 1991; Sedman et al., 1990). Examples of alternative initiation of translation include a growing number of viral or cellular mRNAs ranging from yeasts to humans. The choice of initiation codon often determines the subcellular fate and the function of the protein. For instance, the use of alternative codons can give rise to a soluble versus a membrane-bound protein, to a nuclear versus secreted protein, or to a mitochondrial versus a cytosolic/nuclear enzyme (Acland et al., 1990; Boguta et al., 1994; Gillman et al., 1991; Tenhunen and Ulmanen, 1993). Previous studies of the expression pattern of the two CNP mRNAs assumed that both CNP1 and CNP2 polypeptides were translated, respectively, from their individual messages (Scherer et al., 1994). This assumption is only partly true. An element of ambiguity became apparent with the observation that when thymus RNA, which possesses only one CNP mRNA corresponding in size to the CNP2 mRNA, is used to program an in vitro translation system, both CNP isoforms are produced (Bernier et al., 1987). To explain this, the authors suspected the presence of another comigrating mRNA species in their preparation (Bernier et al., 1987; Scherer et al., 1994). However, it has recently been demonstrated that the eukaryotic translational machinery is able to produce, in addition to CNP2 protein, the CNP1 isoform from the CNP2 transcript (O’Neill et al., 1997). Both CNP isoforms are produced at the translational level by alternative initiation at the two in-frame start codons. Immunoprecipitation experiments with polyclonal anti-CNP detected both CNP isoforms in thymus and testis, or in cell lines expressing only the CNP2 transcript. Moreover, in vitro translation of synthetic CNP2 mRNA demonstrated that both CNP isoforms are synthesized by initiation at diVerent AUG codons. Furthermore, using expression vectors contain-

III. THE MYELIN GENES AND PRODUCTS

CNP GENE STRUCTURE AND REGULATION

ing the CNP2 cDNA with mutations in and around the second start site, it was demonstrated that both AUGs are functional initiation sites for translation of CNP1 and CNP2. By replacing the second AUG with either CUC or GGG, the synthesis of CNP1 was completely abrogated without aVecting the synthesis of CNP2. Moreover, introducing a stop codon by substitution between the two AUGs completely abolished the synthesis of CNP2. Since in this case the sequence at the second AUG is not changed, this reinforced the conclusion that the CNP1 isoform is not a product of proteolytic cleavage of the larger isoform. Additionally, since both AUGs are in frame, the replacement of the second start site by a stop codon abolished the synthesis of both isoforms. In the past, the possibility of generating both CNP isoforms from the CNP2 mRNA was not considered. We now know that both CNP proteins are present, whether or not tissues express both messages, or just the CNP2 message. This was an important discovery given that previous studies had always presumed that expression of the CNP2 mRNA in a particular tissue or cell type meant that only CNP2 protein was present. Therefore, CNP1 and CNP2 isoforms are more widely distributed than previously thought. However, the precise function of each in diVerent cell types remains to be elucidated.

Developmental Aspects of CNP Expression Studies on the expression of CNP in the developing CNS indicate that the protein appears early in development, preceding myelination, and is maintained at a high level as myelin is produced (Snyder et al., 1983; Sprinkle et al., 1978;). Although high levels of CNP were found to accumulate in diVerentiating and myelinating oligodendrocytes in parallel with the major structural myelin proteins, the onset of CNP expression occurs before that of either MBP or PLP. In rat, CNP expression peaks around 10 days of age at a time when oligodendrocyte precursors enter their terminal diVerentiation stage preceding myelination (Amur-Umarjee et al., 1990; Bansal and PfeiVer, 1985; PfeiVer et al., 1981). Furthermore, it appears that both in culture and in vivo, CNP is expressed as early as GalC, which is considered to be a very early marker for diVerentiating oligodendrocytes (Brenner et al., 1986; Knapp et al., 1988; Reynolds et al., 1987). Northern blot analysis on developing mouse brain (Scherer et al., 1994) showed that the CNP2 mRNA could be detected at low levels at embryonic day 16, but the CNP1 transcript was Wrst detected only at postnatal day 1. In accordance with the protein levels, both mRNAs are induced to much higher levels at the time of oligodendrocyte diVerentiation and increase in parallel with those encoding MBP and PLP. It is now well documented that some of the myelin proteins are expressed, not only in myelinating oligodendrocytes, but also in premyelinating cells. This is the case for an alternatively spliced isoform of PLP known as DM-20, which is expressed at low levels in the embryonic and perinatal brain and spinal cord (Nave et al., 1987; Perron et al., 1997). Some MBP-related mRNAs and proteins are also observed in the developing CNS before the onset of myelination (Sorg et al., 1987; Verity and Campagnoni, 1988). In a set of experiments designed to trace sites of origin and migration routes of oligodendrocyte progenitors cells, it was shown that mRNA encoding CNP was expressed, not only by diVerentiating oligodendrocytes but also, in smaller amount, by proliferating oligodendrocyte progenitors. The CNP2 transcript was present in proliferating oligodendrocyte progenitor cells maintained in culture with growth factors to prevent diVerentiation, while the transcripts for CNP1, PLP, or MBP were totally absent (Scherer et al., 1994; Yu et al., 1994). Removal of growth factors and induction of diVerentiation resulted in both CNP mRNAs being up-regulated as well as MBP and PLP. In situ hybridization was then employed to show that CNP mRNA is present in both brain and spinal cord of rat and mouse embryos. Already at E12, cells located in the ventral ventricular zone of the cord were CNP-positive, coincident in time and space with cells that express the platelet derived growth factor a receptor, another putative marker for oligodendrocytes (Yu et al., 1994). In developing brain, CNP, but not MBP, was detected in the diencephalon and the ventral hypothalamus of E12.5 mouse embryos with a distribution similar to that of DM20positive cells. In later stages, in both brain and spinal cord, intensely CNP-positive cells

III. THE MYELIN GENES AND PRODUCTS

509

510

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

accumulated progressively in developing white matter and nearby gray matter, representative of mature cells that now express all myelin proteins necessary for myelinogenesis (Jordan et al., 1989; Trapp et al., 1988). Since CNP is present in premyelinating cells and also in several non-neural cells, it is quite plausible that CNP may have additional functions that are not directly related to myelination. In the PNS of rodents, myelination begins at birth, peaks at about 2 weeks, and decreases to a basal level by the end of the Wrst month. As in the CNS, at the onset of myelination Schwann cells also express CNP. Whereas in the CNS, CNP expression follows a temporal pattern concordant with the accumulation of myelin, the expression of CNP in PNS does not parallel the myelination process. The CNP message is found at near maximum levels one day after birth, and decreases gradually in older animals (Edwards et al., 1988; Stahl et al., 1990).

Regulation of CNP Gene Expression Oligodendrocyte diVerentiation and maturation are controlled by intrinsic cellular programs as well as by growth factors secreted by neighboring cells (McMorris et al., 1990; PfeiVer et al., 1993). The identiWcation of cis- and trans-acting factors that regulate oligodendrocyte-speciWc gene expression is able to provide important information on the cellular mechanisms involved in the diVerentiation and maturation of the CNS. A number of myelin genes have already been studied with regard to their transcriptional regulation in transgenic mice or in cells in culture by transient transfection, and in vitro by DNA bindings studies (reviewed in Wegner, 2000). As described earlier, two alternative promoters control the transcription of mRNAs that encode the two CNP isoforms. Alternative promoters are commonly used to regulate the expression of a gene at diVerent stages of development or in diVerent cell types (Ayoubi and Van De Ven, 1996). However, the molecular mechanisms regulating the activity of both CNP promoters still remain unknown. Both are diVerentially activated (Scherer et al., 1994), with the distal promoter (P2) being activated quite early during brain development, but the proximal promoter (P1) turned on only in later stages. Further, only the distal promoter is active in several nonmyelinating cell types (O’Neill et al., 1997; Scherer et al., 1994). Taken together, this suggests that the P2 promoter is a constitutive promoter, responsible for CNP expression in a variety of cells, whereas the P1 promoter is responsible for tissue and developmental stage speciWcity of CNP1 expression only in myelinating cells. This diVerential regulation suggests that several mechanisms are involved in regulating CNP gene expression, probably mediated by several transcriptional factors. A recent study of CNP gene regulation focused on the immediate 5’ Xanking region of the mouse CNP gene. A transgenic model was generated with a 4-kb fragment (ending 46 nucleotides before the AUG start codon of the CNP1 isoform) coupled to the bacterial lacZ reporter gene (Gravel et al., 1998). These transgenic mice exhibited a b-galactosidase expression pattern consistent with the complex spatial and temporal expression of endogenous CNP (Fig. 19.5). Transgene expression was detected in immature oligodendrocytes of embryonic brain and spinal cord, and it signiWcantly increased with age. In adult mouse brain, b-galactosidase activity, which appeared to be oligodendrocyte speciWc, was observed mostly in white matter areas of the CNS. Moreover, the transgene was expressed in testis, and thymus, tissues that are known to normally express CNP. Despite the observed similarities in transgene and CNP expression, b-galactosidase activity was detected in the PNS only up to 30 days; Schwann cells of adults had barely detectable levels, contrary to expectation from the known developmental proWle of CNP expression in the PNS. Similar observations were reported for the MBP promoter (Ikenaga and Kagawa, 1995). Therefore, it was suggested that additional regulatory elements are required to maintain Schwann cell expression of the transgene in later stages of development. Taken together, the results provide strong evidence that cis-acting regulatory elements, necessary to direct spatial and temporal expression of CNP in oligodendrocytes, are located within the 4-kb 5’-Xanking sequence of the mouse CNP gene. Similar CNPpromoter elements were used successfully to drive expression of the intact human CNP gene in oligodendrocytes of transgenic mice (Gravel et al., 1996), suggesting that the

III. THE MYELIN GENES AND PRODUCTS

CNP GENE STRUCTURE AND REGULATION

FIGURE 19.5 Visualization of CNP-containing elements in transgenic mice. (A) b-galactosidase activity expressed in a Bluo-gal stained sagittal section of p15 mouse brain. (B) Colocalization of b-galactosidase and CNP expression in the white matter tract of the cerebellum of a 30 d old transgenic mouse (mag  25); brain sections were stained Wrst with X-gal, followed by immunolabeling with anti CNP. (C) An oligodendrocyte and its processes, and CNP-containing elements of myelin as in (B), but at higher magniWcation (from Gravel et al., 1998).

regulatory modules within the CNP promoters are similar in both rodents and humans. Future investigation of the molecular mechanisms underlying the control of cell speciWcity and temporal expression of both CNP transcripts will be necessary to identify cis- and trans-acting elements involved in the regulation of the CNP gene. Subsequently, the CNP promoter has been found to be a valuable tool to target the expression of heterologous genes in the developing CNS. Chandross et al. (1999) used the mouse CNP promoter to generate mice expressing b-galactosidase and neomycin phosphotransferase fusion protein and developed a system to analyze and manipulate puriWed myelin-forming cells at various stages of development. When cells isolated from brain or sciatic nerve were cultured in the presence of G418, 99% of all survivors were oligodendrocytes or Schwann cells, respectively. These transgenic mice proved useful for identifying

III. THE MYELIN GENES AND PRODUCTS

511

512

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

and characterizing oligodendrocyte and Schwann cell progenitors and for selecting viable oligodendrocytes and Schwann cells. Similarly, the CNP promoter, which targets the entire oligodendroglial lineage from embryonic life to adulthood, has been used to generate a CNP-GFP transgenic mouse model to study the in vivo properties of oligodendrocyte precursors cells in normal and damaged CNS (Belachew et al., 2001). To establish the existence and relative abundance of oligodendrocyte progenitors in the adult human white matter, a new strategy has been designed to isolate and enrich native oligodendrocyte precursors from adult brain tissue. To accomplish this, cells from dissociated adult human subcortical white matter were transfected with the GFP reporter gene placed under the control of the human distal CNP promoter (P2) (Roy et al., 1999). Using Xuorescenceactivated cell sorting (FACS), cells expressing GFP were isolated and enriched to near purity. Most of the puriWed cells were oligodendrocytes, which can be used for implantation and cell-based therapy. In the CNS, oligodendrocyte diVerentiation is accompanied by the activation of speciWc transcriptional pathways responsible for the activation of myelin-related genes. One of the signals leading to the activation of those genes is cyclic AMP. Raising the intracellular cAMP levels by treatment with forskolin and other cAMP analogues has been shown to accelerate the rate of diVerentiation of oligodendrocytes (Raible and McMorris, 1989, 1990). In addition, McMorris et al. (1985) have demonstrated that an elevation of cAMP concentration in C6 cells, a rat glioma cell line, or in oligodendrocytes was followed by an increased rate of CNP protein synthesis. Recently, the inXuence of increased intracellular cAMP levels on the transcription of both CNP1 and CNP2 messages in C6 cells has been examined (Gravel et al., 2000). Transcription of CNP1 mRNA was signiWcantly increased in comparison to CNP2 mRNA in cells treated with cAMP analogs, indicating the possibility that the cAMP-mediated pathway is responsible in part for the diVerential regulation of both CNP transcripts during the process of oligodendrocyte diVerentiation. Therefore, deletion-transfection studies with a set of mouse CNP promoter-chloramphenicol acetyltransferase (CAT) reporter gene fusion constructs were used to map speciWc sequences in the 4-kb 5’-Xanking region of the mouse CNP gene that are necessary for the up-regulation of CNP1 expression by cAMP (Gravel et al., 2000). One sequence, located between
126 and
102 nucleotides upstream of the transcriptional start site of CNP1, has been identiWed and characterized. This sequence, 5’-TCTGCACACTCAAGAGAGCCAAGGC-3’ has no apparent similarity to the cAMP-responsive-element consensus sequence, 5’-TGACGTCA (Montminy et al., 1986). However, there is a potential binding site, 5’-GCCAAGGC-3’, for the transcription factor AP-2, which has been shown to mediate cAMP responsiveness alone or in concert with other cAMP-responsive elements (Hyman et al., 1989; Imagawa et al., 1987; Medcalf et al., 1990). Mutational analysis conWrmed the importance of the
126/
102 region in mediating the activation of CNP1 transcription by cAMP. Replacing a block of 10 nucleotides between positions
115 and
106 reduced considerably the level of induction of CAT expression by cAMP. Interestingly, within this mutation, the AP-2 binding site has been partly disrupted, supporting the idea that this element is involved in the activation of CNP1 expression by cAMP. However, it is noteworthy that replacement of the sequence between positions
98 and
89 also reduced signiWcantly the level of induction of CAT expression by cAMP. This suggests that other regulatory elements adjacent to the AP-2 binding site may also be important for the transcriptional activation of the CNP1 promoter. Sequence analysis of the region adjacent to the AP-2 binding site revealed the presence of two putative binding sites for the transcriptional factors AP-4 (5’-GCAGCTGTGG-3’) and NF1 (5’-TGGCTTTGCCCA-3’). Accordingly, it is likely then that the cAMP-induced accumulation of CNP1 mRNA is dependent on the combinatorial action of diVerent nuclear factors to produce maximal responsiveness to cAMP. It appears that the presence of adjacent NF-1 and AP-2 binding sites immediately upstream of the CNP1 transcriptional start site may be a characteristic shared by other glial-speciWc genes, such as JC virus, MBP, PLP, S100b, and GFAP (Amemiya et al., 1992). Moreover, it has been shown that the NF-1 binding site in the MBP promoter, in coordination with upstream regulatory elements mediates the transcriptional response of the MBP gene to elevated cAMP levels

III. THE MYELIN GENES AND PRODUCTS

FUNCTIONAL SIGNIFICANCE OF CNP: CONJECTURE AND PERSPECTIVES

(Clark et al., 2002; Zhang and Miskimmins, 1993). These observations suggest that the two myelin-related genes, CNP and MBP, may be regulated by similar mechanisms, possibly involving the NF-1 binding site. To further support this, it will be necessary to determine whether the AP-4 and NF-1 binding sites, in concert with the AP-2 site, are contributing to the mechanism(s) regulating the expression of CNP1 in response to increased cAMP levels and to determine the identity of the nuclear proteins interacting with the AP-2, AP-4, and NF-1, or other putative regulatory elements of the proximal these sites.

FUNCTIONAL SIGNIFICANCE OF CNP: CONJECTURE AND PERSPECTIVES Immunocytochemical studies, employing both light and election microscopy, have demonstrated the presence of CNP throughout the cytoplasm of oligodendrocytes in brain sections at all stages of myelination (Braun et al., 1988; Trapp et al., 1988). CNP synthesis appears to occur in perinuclear regions, where abundant CNP colocalizes with its mRNA (visualized by in situ hybridization (Trapp et al., 1987, 1988). Further, abundant CNP clearly visualized in the myriad Wne Wlopodia of premyelinating oligodendrocytes in the brain, along with intense concentrations just below the plasma membrane, suggested an association with the cytoskeleton, and an early role in the complex process of myelination (Braun et al., 1988). Once myelination is underway, Wlopodia decrease in abundance and are replaced by well-deWned, CNP-Wlled large processes that terminate on the myelin internodes. Of special interest was the clear demonstration of CNP-Wlled paranodal compartments of the myelin sheath in longtitudinally sectioned spinal cord, and the absence of CNP from lamellar regions, except where there were cytoplasmic inclusions (Trapp et al., 1988). This remarkable concentration of the myelin CNP (comprising ~ 4% of total protein) in specialized compartments has been further borne out by the recent demonstration of CNP enrichment in distinct subcellular domains obtained by biochemical fractionation of myelinated nerves, having characteristics of paranodal membranes associated with the axolemma (S. PfeiVer, personal communication, to be published). All of this implies that CNP is synthesized at an early stage of oligodendrocyte development, is transported to the cell periphery, into Wlopodia, and along major processes that engulf the axon, and that it ultimately accumulates in the cytoplasm-Wlled paranodal compartments that comprise a minor part of the sheath. These must be metabolically active, since they are the site of major protein synthesis and myelin assembly. Numerous possible roles for CNP have been considered, and several strategies are being pursued in a few laboratories. Cell biological approaches and model systems to study CNP have been hampered by lack of good assays for function. To date, observations on morphological changes in cells transfected with CNP cDNA have been one of the principal strategies to determine how CNP might operate in cells. Expression of CNP1 in HeLa cells revealed an intracellular distribution not unlike that of oligodendrocytes, with CNP concentrating more at the cell periphery, near the plasma membrane (Staugitis et al., 1990). A similar approach was employed in studies to determine the role of the C-terminal isoprenoid tail (DeAngelis and Braun, 1994). Here, CNP1 expression was induced by transfection of L cells, NIH 3T3 cells, and HeLa cells (they normally express little or no CNP), and compared to the expression of a CNP1 mutant (C397S) where the substitution of ser for the end-terminal cys precluded isoprenylation. Whereas CNP1 induced dramatic morphological transformation, with extensive Wlopodia production and large process extensions, and CNP distribution similar to that in oligodendrocytes (Braun et al., 1988), the nonprenylated protein failed to elicit these morphological changes, and it was distributed more or less uniformly throughout the cytoplasm of the cell. The larger isoform CNP2 also caused morphological transformation similar to CNP1 in all cell lines examined (293, HeLa S3, COS-7), but it is also targeted to mitochondria in some cell types (oligodendrocytes, HeLa S3) (O’Neill, Ph.D. thesis, McGill University, 2002; to be published). The morphology-altering eVects of CNP are time and dose dependent and appear to be linked

III. THE MYELIN GENES AND PRODUCTS

513

514

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

directly to the association of the protein near the cell membrane and possibly with the cytoskeleton. Although, the mechanism by which CNP induces these cellular changes remains unknown, some aspects of CNP1-binding to cytoskeletal proteins have been explored. Fractionation and detergent partitioning of CNP1-expressing cells, immunocytochemical localization, with and without cytoskeleton-disrupting agents, and coimmunoprecipitation experiments all pointed to an association of CNP1 with actinbased microWlaments (Dyer and Benjamins, 1989; DeAngelis and Braun, 1994, 1996a, 1996b), however, it is unknown whether this is a direct or indirect association. Nevertheless, these observations are not suYciently deWnitive to preclude additional CNP complexes in cells and myelin. In addition to the potential association of CNP with F-actin, several studies point to an association of CNP with microtubules. Initial work by Dyer showed detailed confocal images of CNP and tubulin colocalized in primary cultures of diVerentiated oligodendrocytes (Dyer and Benjamins, 1989), particularly along the microtubular veins in the membraneous sheets and in numerous discrete large ‘‘cuV-like’’ punctate structures that dot the microtubule network. Further biochemical evidence for CNP-tubulin interaction derived from observations that microtubules in FRTL rat thyroid cells became dissociated from the plasma membrane after treatment with lovastatin, a drug that blocks isoprenylation. Since tubulin lacks the motif for isoprenylation, it was reasoned that an isoprenylated linker protein was responsible for membrane attachment of microtubules. A 48 kDa isoprenylated protein was later identiWed to be CNP (Laezza et al., 1997). CNP was found not only to be associated with microtubules in FRTL cells and brain tissue, but it also co-puriWed with microtubules by successive cycles of polymerization and depolymerization, thus identifying CNP as a microtubule-associated protein (MAP). Furthermore, CNP possesses microtubule polymerization activity in vitro (Bifulco et al., 2002). The domain responsible for this activity, but not for tubulin binding, was identiWed in the 13 residue C-terminal end of CNP. Deletion mutants did not promote microtubule polymerization in vitro, and a peptide of the 13 residue region, itself, caused microtubules to polymerize in vitro. Furthermore, preliminary in vitro data suggested that CNP phosphorylation by PKC negatively regulates microtubule polymerization. In parallel, independent studies of CNP binding partners, a predominant 55 kDa polypeptide that co-immunoprecipitated with CNP from various brain tissues, was identiWed as tubulin by mass spectrometry (J. Lee, McGill University, to be published). Both cultured oligodendrocytes and CNP1-transfected COS cells showed microtubule colocalization in the cell body, along the main branches of the processes, and in regions of the processes where further branching occurs. Interestingly, COS cells that expressed high levels of CNP exhibited oligodendrocyte-like morphology (rounding of the cell body, process extension, and extensive arborization of the processes) and extensive alteration and rearrangement of the microtubular network (microtubule breakdown, rearrangement, and bundling), eventually leading to the formation of a microtubule network that contains fewer but thicker microtubule strands. Many of these cells had frayed microtubule ends at the outer cell edge, often pointing in the direction of extending processes. Deletion mutant studies revealed that tubulin binding and cell morphological transformation are mediated by the C-terminal phosphodiesterase catalytic domain, but tublin binding does not occur within or near the active site. NMR studies of tubulin complexed with the CNP catalytic fragment suggested nonactive site binding of tubulin to the positively-charged outer surface of the catalytic fragment, governed by electrostatic interactions (J. Lee, McGill University, to be published). Furthermore, RICH, which shares extensive homology with CNP in the C-terminal domain, can also bind to tubulin and cause similar cytoskeletal transformations in non-neuronal cells (M. Gravel and J. Lee, McGill University, to be published). The fact that the catalytic fragment induces process extension and morphological transformation in non-neuronal cells, and that CNPase activity is not required for this, suggests that binding of CNP to tubulin/microtubules changes microtubule dynamics to facilitate formation and arborization of multiple processes. The molecular mechanism for these events and whether or not CNP-tubulin/microtubule interactions inXuence process extension in oligodendrocytes will need to be determined.

III. THE MYELIN GENES AND PRODUCTS

FUNCTIONAL SIGNIFICANCE OF CNP: CONJECTURE AND PERSPECTIVES

In a molecular genetic approach to CNP function, the expression of CNP in oligodendrocytes was increased by introduction of the human gene into mice (Gravel et al., 1996, 1997). A transgenic line expressing a six-fold increase of CNP protein evinced abnormalities suggestive of a gain of function due to the extra CNP. In adult mice large vacuoles were observed, surrounded by myelin membranes that extended from myelin internodes. Further, in about half of the myelin sheaths cytoplasmic leaXets failed to fuse, resulting in an absence of the major dense lines from compact myelin. Oligodendrocytes cultured from these adult animals manifested a more robust and aggressive regrowth of cellular processes, consistent with other studies described earlier that purported to show a link of CNP to the cytoskeleton and to process extensions. Further observations (Yin et al., 1997) on these CNP over-expressing mice at an earlier developmental stage (d18) revealed aberrant oligodendrocyte and myelin membrane formation during early stages of ODC diVerentiation, consistent with a role in dynamic regulation of the cytoskeleton and membrane expansion. A gain-of-function phenotype was again evident, with the major dense line failing to develop; biochemical data were consistent with the possibility that excess CNP was interfering with the normal deposition of MBP in the developing sheath, but it was not clear if this was due to a competition by CNP for MBP binding sites or interference with mRNA transport, translation, degradation, or on some other interference with the protein synthesis machinery associated with myelination. A long-awaited generation of CNP-null mutant mice (Lappe-Siefke et al., 2003) has thus far provided little insight into CNP function in myelinogenesis. In these animals that lack CNP protein, myelination appears to follow a normal course and is morphologically normal. The adult mice, however, manifest a severe neurodegenerative disorder, with axonal swellings and Wallerian degeneration causing a large microglial response, hydrocephalus and premature death, suggesting that CNP is essential for axonal survival. One can speculate that since CNP appears to be a multifunctional protein, it is possible that the functional activity component responsible for normal ODC development and subsequent myelination is provided by a compensatory protein that becomes up-regulated, although completely lacking the capacity to hydrolyze 2’,3’-cyclic nucleotides. Then, the absence of an additional function, normally provided by CNP (possibly related to CNPase activity) and necessary for the ODC to maintain long-term axonal survival, results in an ‘‘uncoupling’’ of the normal relationship between the ODC and the axon, leading to the observed degenerative responses. Clearly, additional experimentation is required to resolve this scenario. Although deWnitive assignments of CNP function still cannot be made, important insights are emerging that may help resolve this conundrum. The presence of CNP in some nonmyelinating cells such as Leydig and Sertoli cells of the testis (DavidoV et al., 2002) or visual pigment cells of the retina (Giulian and Moore, 1980; Kohsaka et al., 1983) compels a consideration of cellular functions for CNP that are not restricted to myelination. While the focus of investigation directed to understanding the CNP problem has been mainly on myelin and oligodendrocytes, more concerted study of nonmyelinating sources of the enzyme could conceivably yield valuable new answers. Another observation worth noting is that some RNA viruses bind CNP derived from the host cell within which they replicate, with a viral enrichment of CNPase activity exceeding 20-fold that of the cell. This becomes even more intriguing with the Wnding that CNP binds to the viral nucleocapsid (Rosenbergova and Pristasova, 1990, 1991), raising questions about the cell biology of CNP and a possible connection to RNA metabolism, degradation, or regulation. Indeed, a link between CNP and RNA has emerged from another direction. As described in the sections titled ‘‘Enzymatic Characteristics and Reaction Mechanism of CNPase’’ and ‘‘Three-Dimensional Structural Features,’’ the three-dimensional structure of CNP reveals a close resemblance to the structure of a cyclic phosphodiesterase from plants, (an enzyme that hydrolyzes ADP-ribose 1’’, 2’’ cyclic phosphate, a product of the tRNA splicing reaction) and to bacterial 2’-5’ RNA ligase (an enzyme that ligates tRNA half-molecules with a 2’, 3’-cyclic phosphate and 5’-hydroxl termini), despite very low

III. THE MYELIN GENES AND PRODUCTS

515

516

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

sequence similarities. The structural resemblance and conservation of two tetrapeptide motifs places CNP/RICH in a superfamily of proteins, which consists of three other large subfamilies of RNA processing enzymes (Mazumder et al., 2002). Although the physiological signiWcance of some of these enzymes is not clear yet, they all share some relationship to RNA pathways. One can contemplate several potential reasons for CNP-RNA cellular complexes: (1) Since CNP also binds to microtubules, perhaps it is part of an RNA transport complex. (2) CNP could participate in the regulation of RNA function, processing, or degradation in specialized compartments of the cell, such as the paranodal loops of myelin. It is worth mentioning an early study of Starich and Dreiling (1980) purporting to show that partial inhibition of CNPase activity (by 2’-AMP) reduced the amount of labelled leucine incorporated into myelin proteins in vitro. Although this rudimentary study was never conWrmed or followed up with further experiments, the suggestion remains of a CNP association with the protein synthesizing machinery. (3) CNP may bind members of the noncoding family of RNA (e.g., microRNA; intermediates in RNA interference) for a variety of regulatory purposes. This is only a partial list of possibilities, and further speculation is premature. CNP-RNA complexes are in the process of being identiWed and characterized with the anticipation of forthcoming new insights for the role of CNP.

DYSMYELINATING AND DEMYELINATING DISORDERS: IS CNP INVOLVED? Whereas a variety of neurological mutations in mice, causing an assortment of demyelinating and dysmyelinating pathologies, have been valuable model systems to study the function of MBP, PLP, PMP22, and so forth, none of these exist for CNP. To date, CNP has not been linked to any inherited diseases of myelin, and animals with CNP gene mutations have not been reported. In studies of encephalitogenic proteins that could function as self-antigens in inXammatory demyelinating diseases, CNP has been considered as a potential autoantigen. However, CNP was not found not to be encephalitogenic in rodents, although a heat-shock protein related peptide domain of CNP was shown to alter the course of EAE (Birnbaum et al., 1996). Maatta et al. (1998) and Morris-Downes et al. (2002) conWrmed that CNP was nonencephalitogenic in several mouse strains, even though several CNP epitopes could induce T-cell responses in these mice. Several other labs have sought to implicate CNP as an autoantigen in multiple sclerosis. Walsh and Murray (1998) reported that antibodies to CNP were detected in sera of 74% of their patients. These were present as IgM in high titer and also in cerebrospinal Xuid. Further, CNP-containing immune complexes were present in patient brain tissue. Curiously, the antibody response was against CNP1 and not CNP2. However, both isoforms were reported to bind the C3 complement, fueling speculation about a role for CNP in the pathogenesis of MS (Walsh and Murray, 1998). CNP has been considered as a potential antigenic target for a T-cell response in the pathogenesis of demyelinating diseases. Rosener et al. (1997) used a recombinant human CNP to isolate several speciWc T-cell lines from a patient with multiple sclerosis, as well as from a healthy control. In further investigation, this laboratory used human CNP from brain, as well as speciWc CNP peptides to screen for human T-cell responses. Primary proliferative responses were detected in some MS patients as well as healthy controls, and they were directed mainly to the CNP1 polypeptide region 343–373, which also turned out to be one of the main HLA class II-restricted human immunodominant regions of CNP. Despite a detailed analysis of their Wndings, an unambiguous involvement of CNP-speciWc T cells in MS could not be established (Muraro et al., 2002). Taken together with the previously related studies, as well as the curious observations of CNP association with the nucleocapsids of some RNA viruses (see the section titled ‘‘Functional SigniWcance of CNP’’), it is clear that further investigation is warranted.

III. THE MYELIN GENES AND PRODUCTS

DYSMYELINATING AND DEMYELINATING DISORDERS: IS CNP INVOLVED?

References Acland, P., Dixon, M., Peters, G., and Dickson, C. (1990). Subcellular fate of the Int-2 oncoprotein is determined by choice of initiation codon. Nature 343, 662–665. Agrawal, H. C., Sprinkle, T. J., and Agrawal, D. (1990a). 2’,3’-cyclic nucleotide-3’-phosphodiesterase in peripheral nerve myelin is phosphorylated by a phorbol ester-sensitive protein kinase. Biochem. Biophys. Res. Commun. 170, 817–823. Agrawal, H. C., Sprinkle, T. J., and Agrawal, D. (1990b). 2’,3’-cyclic nucleotide-3’-phosphodiesterase in the central nervous system is fatty-acylated by thioester linkage. J. Biol. Chem. 265, 11849–11853. Agrawal, H. C., Sprinkle, T. J., and Agrawal, D. (1994). In vivo phosphorylation of 2’,3’-cyclic nucleotide 3’-phosphohydrolase (CNP): CNP in brain myelin is phosphorylated by forskolin- and phorbol ester-sensitive protein kinases. Neurochem. Res. 19, 721–728. Amemiya, K.,Traub, R., Durham, L., and Major, E. O. (1992). Adjacent nuclear factor-1 and activator protein binding sites in the enhancer of the neurotropic JC virus. A common characteristic of many brain-speciWc genes. J. Biol. Chem. 267, 14204–14211. Amur-Umarjee, S. G., Dasu, R. G., and Campagnoni, A. T. (1990). Temporal expression of myelin-speciWc components in neonatal mouse brain cultures: Evidence that 2’,3’-cyclic nucleotide 3’-phosphodiesterase appears prior to galactocerebroside. Dev. Neurosci. 12, 251–262. Ayoubi, T. A., and Van de Ven, W. J. (1996). Regulation of gene expression by alternative promoters. Faseb J. 10, 453–460. Ballestero, R. P., Wilmot, G. R., AgranoV, B. W., and Uhler, M. D. (1997). gRICH68 and gRICH70 are 2’,3’-cyclic-nucleotide 3’-phosphodiesterases induced during goldWsh optic nerve regeneration. J. Biol. Chem. 272, 11479–11486. Ballestero, R. P., Wilmot, G. R., Leski, M. L., Uhler, M. D., and AgranoV, B. W. (1995). Isolation of cDNA clones encoding RICH: A protein induced during goldWsh optic nerve regeneration with homology to mammalian 2’,3’-cyclic-nucleotide 3’-phosphodiesterases. Proc. Natl. Acad. Sci. U S A. 92, 8621–8625. Ballestero, R. P., Dybowski, J. A., Levy, G., AgranoV, B. W., and Uhler, M. D. (1999). Cloning and characterization of zRICH, a 2’,3’-cyclic-nucleotide 3’-phosphodiesterase induced during zebraWsh optic nerve regeneration. J. Neurochem. 72, 1362–1371. Bansal, R., and PfeiVer, S. E. (1985). Developmental expression of 2’,3’-cyclic nucleotide 3’-phosphohydrolase in dissociated fetal rat brain cultures and rat brain. J. Neurosci. Res. 14, 21–34. Belachew, S., Yuan, X., and Gallo, V. (2001). Unraveling oligodendrocyte origin and function by cell-speciWc transgenesis. Dev. Neurosci. 23, 287–298. Bernier, L., Alvarez, F., Norgard, E. M., Raible, D. W., Mentaberry, A., Schembri, J. G., Sabatini, D. D., and Colman, D. R. (1987). Molecular cloning of a 2’,3’-cyclic nucleotide 3’-phosphodiesterase: mRNAs with diVerent 5’ ends encode the same set of proteins in nervous and lymphoid tissues. J. Neurosci. 7, 2703–2710. Bernier, L., Colman, D. R., and D’Eustachio, P. (1988). Chromosomal locations of genes encoding 2’, 3,-cyclic nucleotide 3’-phosphodiesterase and glial Wbrillary acidic protein in the mouse. J. Neurosci. Res. 20, 497–504. Bifulco, M., Laezza, C., Stingo, S., and WolV, J. (2002). 2’,3’-cyclic nucleotide 3’-phosphodiesterase: A membranebound, microtubule-associated protein and membrane anchor for tubulin. PNAS 99, 1807–1812. Birnbaum, G., Kotilinek, L., Schlievert, P., Clark, H. B., Trotter, J., Horvath, E., Gao, E., Cox. M., and Braun, P. E. (1996). Heat shock proteins and experimental autoimmune encephalomyelitis (EAE): I. Immunization with a peptide of the myelin protein 2’,3’ cyclic nucleotide 3’ phosphodiesterase that is cross-reactive with a heat shock protein alters the course of EAE. J Neurosci Res. 44, 381–396. Boguta, M., Hunter, L. A., Shen, W. C., Gillman, E. C., Martin, N. C., and Hopper, A. K. (1994). Subcellular locations of MOD5 proteins: Mapping of sequences suYcient for targeting to mitochondria and demonstration that mitochondrial and nuclear isoforms commingle in the cytosol. Mol. Cell. Biol. 14, 2298–2306. Bradbury, J. M., Campbell, R. S., and Thompson, R. J. (1984). Endogenous cyclic AMP-stimulated phosphorylation of a Wolfgram protein component in rabbit central-nervous-system myelin. Biochem. J. 221, 351–359. Bradbury, J. M., and Thompson, R. J. (1984). PhotoaYnity labelling of central-nervous-system myelin. Evidence for an endogenous type I cyclic AMP-dependent kinase phosphorylating the larger subunit of 2’,3’-cyclic nucleotide 3’-phosphodiesterase. Biochem. J. 221, 361–368. Braun, P. E., Bambrick, L. L., Edwards, A. M., and Bernier, L. (1990). 2’,3’-Cyclic Nucleotide 3’-Phosphodiesterase (CNP) Has Characteristics of Cytoskeletal Proteins: An Hypothesis for its Function. Annals N. Y. Acad. Sci, 605, 55–65. Braun, P. E., DeAngelis, D., Shtybel, W. W., and Bernier, L. (1991). Isoprenoid modiWcation permits 2’,3’-cyclic nucleotide 3’-phosphodiesterase to bind to membranes. J. Neurosci. Res. 30, 540–544. Braun, P. E., Sandillon, F., Edwards, A., Matthieu, J. M., and Privat, A. (1988). Immunocytochemical localization by electron microscopy of 2’3’-cyclic nucleotide 3’-phosphodiesterase in developing oligodendrocytes of normal and mutant brain. J. Neurosci. 8, 3057–3066. Brenner, T., Lisak, R. P., Rostami, A., McMorris, F. A., Silberberg, D. H. (1986). A monoclonal antibody raised to corpus callosum extract reacts with 2’,3’-cyclic nucleotide 3’-phophoshydrolase. J. Neurochem. 46, 54–60. Chandross, K. J., Cohen, R. I., Paras, P. J., Gravel, M., Braun, P. E., and Hudson, L. D. (1999). IdentiWcation and characterization of early glial progenitors using a transgenic selection strategy. J. Neurosci. 19, 759–774. Clark, Jr, R. E., Miskimins, W. K., and Miskimins, R. (2002). Phosphatidylinositol-3 kinase p85 enhances expression from the myelin basic protein promoter in oligodendrocytes. J. Neurochem. 83, 565–573.

III. THE MYELIN GENES AND PRODUCTS

517

518

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

Cox, M. E., Gao, E. N., and Braun, P. E. (1994). C-terminal CTII motif of 2’,3’-cyclic nucleotide 3’phosphodiesterase undergoes carboxylmethylation. J. Neurosci. Res. 39, 513–518. DavidoV, M. S., MiddendorV, R., Kofuncu, E., Muller, D., Jezek, D., and Holstein, A. F. (2002). Leydig cells of the human testis possess astrocyte and oligodendrocyte marker molecules. Acta Histochem. 104, 39–49. DavidoV, M. S., MiddendorV, R., Muller, D., Kofuncu, E., and Holstein, A. F. (1997). Immunoreactivity for glial cell markers in the human testis. Adv Exp Med Biol. 424, 151–152. DeAngelis, D. A., and Braun, P. E. (1994). Isoprenylation of brain 2’,3’-cyclic nucleotide 3’-phosphodiesterase modulates cell morphology. J. Neurosci. Res. 39, 386–397. DeAngelis, D. A., and Braun, P. E. (1996a). 2’,3’-Cyclic nucleotide 3’-phosphodiesterase binds to actin-based cytoskeletal elements in an isoprenylation-independent manner. J. Neurochem. 67, 943–951. DeAngelis, D. A., and Braun, P. E. (1996b). Binding of 2’,3’-cyclic nucleotide 3’-phosphodiesterase to myelin: An in vitro study. J. Neurochem. 66, 2523–2531. DeAngelis, D. A., Cox, M., Gao, E., and Braun, P. E. (1994). Cellular and molecular characteristics of CNP suggest a regulatory mechanism in myelinogenesis. In ‘‘A Multidisciplinary Approach to Myelin Diseases’’ (S. Salvati, ed.), pp. 49–58. Plenum Press, New York. Delmas, P., Laoide, B. M., Masquilier, D., De Groot, R. P., Foulkes, N. S., and Sassone-Corsi, P. (1992). Alternative usage of initiation codons in mRNA encoding the cAMP-responsive-element modulator generates regulators with opposite functions. Proc. Natl. Acad. Sci. USA 89, 4226–4230. Descombes, P., and Schibler, U. (1991). A liver-enriched transcriptional activator protein, LAP, and a transcriptional inhibitory protein, LIP, are translated from the same mRNA. Cell. 67, 569–579. Douglas, A. J., Fox, M. F., Abbott, C. M., Hinks, L. J., Sharpe, G., Povey, S., and Thompson, R. J. (1992). Structure and chromosomal localization of the human 2’,3’-cyclic nucleotide 3’-phosphodiesterase gene. Ann. Hum. Genet. 56, 243–254. Douglas, A. J., and Thompson, R. J. (1993). Structure of the myelin membrane enzyme 2’,3’-cyclic nucleotide 3’phosphodiesterase: Evidence for two human mRNAs. Biochem. Soc. Trans. 21, 295–297. Dreiling, C. E., Schilling, R. J., and Reitz, R. C. (1981a). 2’,3’-cyclic nucleotide 3’-phosphohydrolase in rat liver mitochondrial membranes. Biochim. Biophys. Acta. 640, 114–120. Dreiling, C. E., Schilling, R. J., and Reitz, R. C. (1981b). EVects of chronic ethanol ingestion on the activity of rat liver mitochondrial 2’,3’-cyclic nucleotide 3’-phosphohydrolase. Biochim. Biophys. Acta. 640, 121–130. Drummond, G. I., Iyer, N. T., and Keith, J. (1962). Hydrolysis of ribonucleoside 2’,3’-cyclic phosphates by a diesterase from brain. J. Biol. Chem., 237, 3535–3539. Dyer, C. A., and Benjamins, J. A. (1988). Antibody to galactocerebroside alters organization of oligodendroglial membrane sheets in culture. J Neurosci. 8, 4307–4318. Dyer, C. A., and Benjamins, J. A. (1989). Organization of oligodendroglial membrane sheets. I: Association of myelin basic protein and 2’,3’-cyclic nucleotide 3’-phosphohydrolase with cytoskeleton. J. Neurosci Res. 24, 201–211. Dyer, C. A., and Matthieu, J. M. (1994). Antibodies to myelin/oligodendrocyte-speciWc protein and myelin/ oligodendrocyte glycoprotein signal distinct changes in the organization of cultured oligodendroglial membrane sheets. J Neurochem. 62, 777–787. Dyer, C. A., Philibotte, T. M., Billings-Gagliardi, S., and Wolf, M. K. (1995). Cytoskeleton in myelin-basicprotein-deWcient shiverer oligodendrocytes. Dev Neurosci. 17, 53–62. Edwards A. M., and Braun, P. E. (1988). Gene expression of the central and peripheral nervous system myelin membrane 2’,3’-cyclic nucleotide 3’-phosphodiesterase in development. Dev. Neurosci. 10, 75–80. Gillespie, C. S., Bernier, L., Brophy, P. J., and Colman, D. R. (1990). Biosynthesis of the myelin 2’,3’-cyclic nucleotide 3’-phosphodiesterases. J Neurochem. 54, 656–661. Gillespie, C. S., Wilson, R., Davidson, A., and Brophy, P. J. (1989). Characterization of a cytoskeletal matrix associated with myelin from rat brain. Biochem J. 260, 689–696. Gillman, E. C., Slusher, L. B., Martin, N. C., and Hopper, A. K. (1991). MOD5 translation initiation sites determine N6-isopentenyladenosine modiWcation of mitochondrial and cytoplasmic tRNA. Mol. Cell. Biol. 11, 2382–2390. Giulian, D., and Moore, S. (1980). IdentiWcation of 2’:3’-cyclic nucleotide 3’-phosphodiesterase in the vertebrate retina. J. Biol Chem. 255, 5993–5995. Gravel, M., DeAngelis, D. A., and Braun, P. E. (1994). Molecular cloning and characterization of 2’,3’-cyclic nucleotide 3’-phosphodiesterase isoform 2. J. Neurosci. Res. 38, 243–247. Gravel, M., Di Polo, A., Valera, P. B., and Braun, P. E. (1998). Four-kilobase sequence of the mouse CNP gene directs spatial and temporal expression of lacZ in transgenic mice. J. Neurosci. Res. 53, 393–404. Gravel, M. Gao, E., Hervouet-Zeiber, C., Parsons, V., and Braun, P. E. (2000). Transcriptional regulation of 2’,3’cyclic nucleotide 3’-phosphodiesterase gene expression by cyclic AMP in C6 cells. J. Neurochem. 75, 1940–1950. Gravel, M., Peterson, J., Yong, V. W., Kottis, V., Trapp, B., and Braun, P. E. (1996). Overexpression of 2’, 3’cyclic nucleotide 3’-phosphodiesterase in transgenic mice alters oligodendrocyte development and produces aberrant myelination. Molec. Cell. Neurosci. 7, 453–466. Gravel, M., Trapp, B., Peterson, J., Braun, P. E. (1997). CNP in myelination: Overexpression alters oligodendrocyte morphogenesis. From ‘‘Cell Biology and Pathology of Myelin.’’ Plenum Press, New York. pp. 75–82.

III. THE MYELIN GENES AND PRODUCTS

DYSMYELINATING AND DEMYELINATING DISORDERS: IS CNP INVOLVED?

Hofmann, A., Grella, M., Botos, I., Filipowicz, W., and Wlodawer, A. (2002). Crystal structures of the semireduced and inhibitor-bound forms of cyclic nucleotide phosphodiesterase from Arabidopsis thaliana. J Biol Chem 277, 1419–1425. Hofmann, A., Zdanov, A., Genschik, P., Ruvinov, S., Fillipowicz, and Wlodawer, A. (2000). Structure and mechanism of activity of the cyclic phosphodiesterase of Appr>p, a product of the tRNA splicing reaction. EMBO 19, 6207–6217. Hyman, S. E., Comb, M., Pearlberg, J., and Goodman, H. M. (1989). An AP-2 element acts synergistically with the cyclic AMP- and phorbol ester-inducible enhancer of the human proenkephalin gene. Mol. Cell. Biol. 9, 321–324. Ikenaka, K., and Kagawa, T. (1995). Transgenic systems in studying myelin gene expression. Dev Neurosci. 17, 127–136.http://highwire. stanford. edu/cgi/medline/pmid;8549424?maxtoshow ¼ & HITS ¼ & hits ¼ & RESULTFORMAT ¼ 1 & author1 ¼ ikenaka & fulltext ¼ MBP & searchid ¼ 1039038349803_1525 & stored_search ¼ & FIRSTINDEX ¼ 0 & sortspec ¼ relevance & fdate ¼ 1/1/1994 & tdate ¼ 12/31/1997 & resourcetype ¼ 1 Imagawa, M., Chiu, R., and Karin, M. (1987). Transcription factor AP-2 mediates induction by two diVerent signal-transduction pathways: Protein kinase C and cAMP. Cell 51, 251–260. Jordan, C., Friedrich, V., Dubois-Dalc, M. (1989). In situ hybridization analysis of myelin gene transcripts in developing mouse spinal cord. J. Neurosci. 9, 248–257. Kasama-Yoshida, H., Tohyama, Y., Kurihara, T., Sakuma, M., Kojima, H., and Tamai, Y. (1997). A comparative study of 2’,3’-cyclic nucleotide 3’-phosphodiesterase in vertebrates: cDNA cloning and amino acid sequences for chicken and bullfrog enzymes. J. Neurochem. 69, 1335–1342. Kim, T., and PfeiVer, S. E. (1999). Myelin glycosphingolip/cholesterol-enriched microdomains selectively sequester the non-compact myelin proteins CNP and MOG. Neurocytology 4, 281–293. Klein, C., Kramer, E. M., Cardine, A. M., Schraven, B., Brandt, R., and Trotter, J. (2002). Process outgrowth of oligodendrocytes is promoted by interaction of fyn kinase with the cytoskeletal protein tau. J. Neurosci 22, 698–707. Knapp, P. E., SkoV, R. P., and Sprinkle, T. J. (1988). DiVerential expression of galactocerebroside, myelin basic protein, and 2’,3’-cyclic nucleotide 3’-phosphodiesterase during development of oligodendrocytes in vitro. J. Neurosci. Res. 21, 249–259. Kohsaka, S., Nishimura, Y., Takamatsu, K., Shimai, K., and Tsukada, Y. (1983). Immunohistochemical localization of 2’,3’-cyclic nucleotide 3’-phosphodiesterase and myelin basic protein in the chick retina. J Neurochem. 41, 434–439. Koonin, E. V., and Gorbalenya, A. E. (1990). Related domains in yeast tRNA ligase, bacteriophage T4 polynucleotide kinase and RNA ligase, and mammalian myelin 2’,3’-cyclic nucleotide phosphohydrolase revealed by amino acid sequence comparison. FEBS Lett. 268, 231–234. Kozak, M. (1989). The scanning model for translation: An update. J. Cell. Biol. 108, 229–241. Kozak, M. (1990). Downstream secondary structure facilitates recognition of initiator codons by eukaryotic ribosomes. Proc. Natl. Acad. Sci. USA. 87, 8301–8305. Kozak, M. (1991). EVects of long 5’ leader sequences on initiation by eukaryotic ribosomes in vitro. Gene expression 1, 117–125. Kozak, M. (1995). Adherence to the Wrst-AUG rule when a second AUG codon follows closely upon the Wrst. Proc. Natl. Acad. Sci. USA. 92, 2662–2666. Kozak, M. (1997). Recognition of AUG and alternative initiator codons is augmented by G in position þ4 but is not generally aVected by the nucleotides in positions þ5 and þ6. EMBO J. 16, 2482–2492. Kozlov, G., Lee, J., Gravel, M., Ekiel, I., Braun, P. E., and Gehring, K. (2002) Letter to the editor: Assignment of the 1H, 13C and 15N resonances of the catalytic domain of the rat 2’,3’-cyclic nucleotide 3’-phosphodiesterase. J. Biomol. NMR 22, 99–100. Kozlov, G., Lee, J., Elias, D., Gravel, M., Gutierrez, P., Ekiel, I., Braun, P. E., and Gehring, K. (2003). Structural evidence that brain cyclic nucleotide phosphodiesterase is a member of the 2H phosphodiesterase superfamily. J. Biol. Chem. 278 (in press). Kra¨mer, E-M., Klein, C., Koch, T., Boytinck, M., and Trotter, J. (1999). Compartmentation of fyn kinase with glycosylphosphatidylinositol-anchored molecules in oligodendrocytes facilitates kinase activation during myelination. J. Biol. Chem. 274, 29042–29049. Kurihara, T., Monoh, K., Sakimura, K., and Takahashi, Y. (1990). Alternative splicing of mouse brain 2’,3’-cyclic nucleotide 3’-phosphodiesterase mRNA. Biochem. Biophys. Res. Commun. 170, 1074–1081. Kurihara, T., Tohyama, Y., Yamamoto, J., Kanamatsu, T., Watanabe, R., and Kitajima, S. (1992). Origin of brain 2’,3’-cyclic-nucleotide 3’-phosphodiesterase doublet. Neurosci Lett. 138, 49–52. Kurihara, T., and Tsukada, Y. (1967). The regional and subcellular distribution of 2’,3’-cyclic nucleotide 3’-phosphohydrolase in the central nervous system. J. Neurochem. 14, 1167–1174. Laezza, C., J. WolV, and M. Bifulco. (1997). IdentiWcation of a 48-kDa prenylated protein that associates with microtubules as 2’,3’-cyclic nucleotide 3’-phosphodiesterase in FRTL-5 cells. FEBS Letters. 413, 260–264. Lappe-Siefke, C., Goebbels, S., Gravel, M., Nicksch, E., Lee, J., Braun, P. E., GriYths, I. R., Nave, K-A (2003). Disruption of the CNP1 gene uncouples oligodendroglial functions in axonal support and myelination. Nature Gen. 33, 366–374. Lee, J., Gravel, M., Gao, E., O’Neill, R. C., and Braun, P. E. (2001). IdentiWcation of Essential Residues in 2’,3’Cyclic Nucleotide 3’-Phosphodiesterase. J. Biol. Chem. 276, 14804–14813.

III. THE MYELIN GENES AND PRODUCTS

519

520

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

Leski, M. L., and AgranoV, B. W. (1994). PuriWcation and characterization of p68/70, regeneration-associated proteins from goldWsh brain. J Neurochem. 62, 1182–1191. Lintner, R. N., and Dyer, C. A. (2000). Redistribution of cholesterol in oligodendrocyte membrane sheets after activation of distinct signal transduction pathways. J. Neurosci. Res. 60, 437–449. Lock, P. Ralph, S., Stanley, E., Boulet, I., Ramsay, R., and Dunn, A. R. (1991). Two isoforms of murine hck, generated by utilization of alternative translational initiation codons, exhibit diVerent patterns of subcellular localization. Mol. Cell. Biol. 11, 4363–4370. Maatta, J. A., Kaldman, M. S., Sakoda, S., Salmi, A. A,, and Hinkkanen, A. E. (1998). Encephalitogenicity of myelin-associated oligodendrocytic basic protein and 2’,3’-cyclic nucleotide 3’-phosphodiesterase for BALB/c and SJL mice. Immunology 95, 383–388. Mazumder, R., Iyer, L. M., Vasudevan, S., and Aravind, L. (2002). Detection of novel members, structurefunction analysis and evolutionary classiWcation of the 2H phosphoesterase superfamily. Nucleic Acids Res. 30, 5229–5243. McFerran, B., and Burgoyne, R. (1997). 2’,3’-Cyclic nucleotide 3’-phosphodiesterase is associated with mitochondria in diverse adrenal cell types. J. Cell Sci. 110, 2979–2985. McMorris, F. A., Smith, T. M., Sprinkle, T. J., and Auszmann, J. M. (1985). Induction of myelin components: Cyclic AMP increases the synthesis rate of 2’,3’-cyclic nucleotide 3’-phosphohydrolase in C6 glioma cells. J. Neurochem. 44, 1242–1251. Medcalf, R. L., Ru¨egg, M., and Schleuning, W. D. (1990). A DNA motif related to the cAMP-responsive element and an exon-located activator protein-2 binding site in the human tissue-type plasminogen activator gene promoter cooperate in basal expression and convey activation by phorbol ester and cAMP. J. Biol. Chem. 265, 14618–14626. Monoh, K., Kurihara, T., Sakimura, K., and Takahashi, Y. (1989). Structure of mouse 2’,3’-cyclic nucleotide 3’-phosphodiesterase gene. Biochem. Biophys. Res. Comm. 165, 1213–1220. Monoh, K., Kurihara, T., Takahashi, T., Ichikawa, Y., Kumanishi, T., Hayashi, S., Minoshima, S., and Shimizu, N. (1993). Structure, expression and chromosomal localization of the gene encoding human 2’,3’-cyclic nucleotide 3’-phosphodiesterase. Gene 129, 297–301. Montminy, M. R., Sevarino, K. A., Wagner, J. A., Mandel, G., and Goodman, R. H. (1986). IdentiWcation of a cyclic-AMP-responsive element within the rat somatostatin gene. Proc. Nat. Acad. Sci. USA 83, 6682–6686. Morris-Downes, M. M., McCormack, K., Baker, D., Sivaprasad, D., Natkunarajah, J., and Amor, S. (2002). Encephalitogenic and immunogenic potential of myelin-associated glycoprotein (MAG), oligodendrocytespeciWc glycoprotein (OSP) and 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNPase) in ABH and SJL mice. J Neuroimmunol 22, 20–33. Muraro, P. A., Kalbus, M., Afshar, G., McFarland, H. F., and Martin, R. (2002). T cell response to 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNNase) in multiple sclerosis patients. Neuroimmunol. 130, 233–242. Nave, K-A, Lai, C., Bloom, F., Milner, R. (1987). Splice site selection in the proteolipid protein gene transcript and primary structure of the DM-20 protein of central nervous system myelin. Proc. Natl. Acad. Sci. USA 84, 5665–5669. Nishizawa, Y., Kurihara, T., and Takahashi, Y. (1982). Immunohistochemical localization of 2’, 3’-cyclic nucleotide 3’-phosphodiesterase in the retina. Brain Res. 251, 384–387. Olafson, R. W., G. I. Drummond, and Lee, J. F. (1969). Studies on 2’,3’-cyclic nucleotide-3’-phosphohydrolase from brain. Can. J. Biochem. 47, 961–966. O’Neill, R. C., and Braun, P. E. (2000). Selective synthesis of 2’,3’-cyclic nucleotide 3’-phosphodiesterase isoform 2 and identiWcation of speciWcally phosphorylated serine residues. J. Neurochem. 74, 540–546. O’Neill, R. C., Minuk, J., Cox, M. E., Braun, P. E., and Gravel, M. (1997). CNP2 mRNA directs synthesis of both CNP1 and CNP2 polypeptides. J. Neurosci. Res. 50, 248–257. Osterhout, D. J., Wolven, A., Wolf, R. M., Resh, M. D., and Chao, M. V. (1999). Morphological diVerentiation of oligodendrocytes requires activation of Fyn tyrosine kinase. J. Cell Biol. 145, 1209–1218. Pereyra, P. M., Horvath, E., and Braun, P. E. (1988). Triton X-100 extractions of central nervous system myelin indicate a possible role for the minor myelin proteins in the stability in lamellae. Neurochem. Res. 13, 583–595. Peyron, F., Timsit, S., Thomas, J-L, Kagawa, K., Ikenaka, K., Zalc, B. (1997). In situ expression of PLP/DM-20, MBP, and CNP during embryonic and postnatal development of the jimpy mutant and of transgenic mice overexpressing PLP. J. Neurosci. Res. 50, 190–201. PfeiVer, S. E., Barbarese, E., and Bhat, S. (1981). Non-coordinate regulation of myelinogenic parameters in primary cultures of dissociated fetal rat brain. J. Neurosci. Res. 6, 369–380. PfeiVer, S. E., Warrington, A. E., and Bansal, R. (1993). The oligodendrocyte and its many cellular processes. Trends Cell Biol 3, 191–197. Prinos, P., Slack, C., and Lasko, D. D. (1995). 5’phosphorylation of DNA in mammalian cells: IdentiWcation of a polymin P-precipitable polynucleotide kinase. J Cell Biochem 58, 115–131. Raible, D. W., and McMorris, F. A. (1989). Cyclic AMP regulates the rate of diVerentiation of oligodendrocytes without changing the lineage commitment of their progenitors. Dev. Biol. 133, 437–446. Raible, D. W., and McMorris, F. A. (1990). Induction of oligodendrocyte diVerentiation by activators of adenylate cyclase. J. Neurosci. Res. 27, 43–46. Rastogi, S. C., and Clausen, J. (1985). Membrane-bound 2’,3’,-cyclic nucleotide 3’-phosphohydrolase activity of lymphocytes, granulocytes and erythrocytes in multiple sclerosis. Acta Neurol Scand. 71, 303–308.

III. THE MYELIN GENES AND PRODUCTS

DYSMYELINATING AND DEMYELINATING DISORDERS: IS CNP INVOLVED?

Reynolds, R., Carey, E. M., and Herschkowitz, N. (1989). Immunohistochemical localization of myelin basic protein and 2’,3’-cyclic nucleotide 3’-phosphohydrolase in Xattened membrane expansions produced by cultured oligodendrocytes. Neuroscience 28, 181–188. Rosenbergova, M., and Pristasova, S. (1990). Localization of 2’,3’-decycling phosphodiesterases in the Newcastle disease virus virion. Acta. Virol. 34, 508–516. Rosenbergova, M., and Pristasova, S. (1991). Relationship of 2’,3’-cyclic nucleotide 3’-phosphohydrolase activity of large enveloped RNA viruses to host cell activity. Acta. Virol. 35, 401–407. Rosener, M., Muraro, P. A., Riethmuller, A., Kalbus, M., Sappler, G., Thompson, R. J., Lichtenfels, R., Sommer, N., McFarland, H. F, and Martin, R. (1997). 2’,3’-cyclic nucleotide 3’-phosphodiesterase: A novel candidate autoantigen in demyelinating diseases. J Neuroimmunol 75, 28–34. Roy, N. S., Wang, S., Harrison-Restelli, C., Benraiss, A., Fraser, R. A. R., Gravel, M., Braun, P. E., and Goldman, S. A. (1999). IdentiWcation, isolation, and promoter-deWned separation of mitotic oligodendrocyte progenitor cells from the adult human subcortical white matter. J. Neurosci. 19, 9986–9995. Scherer, S. S., Braun, P. E., Grinspan, J., Collarini, E., Wang, D. Y., and Kamholz, J. (1994). DiVerential regulation of the 2’,3’-cyclic nucleotide 3’-phosphodiesterase gene during oligodendrocyte development. Neuron 12, 1363–1375. Sedman, S. A., Gelembiuk, G. W., and Mertz, J. E. (1990). Translation initiation at a downstream AUG occurs with increased eYciency when the upstream AUG is located very close to the 5’ cap. J. Virol. 64, 453–457. Sheedlo, H. J., Doran, J. E., and Sprinkle, T. J. (1984). An investigation of 2’:3’-cyclic nucleotide 3’-phosphodiesterase (EC 3.1.4.37, CNP) in peripheral blood elements and CNS myelin. Life Sci. 34, 1731–1737. Sims, N. R., and Carnegie, P. R. (1978). 2’,3’-cyclic-nucleotide 3’-phosphodiesterase. In ‘‘Advances in Neurochemistry’’ (B. W. AgranoV and M. H. Aprison, eds.) Chapter 1, pp. 1–41. Plenum Press, New York. Sinensky, M. (2000). Recent advances in the study of prenylated proteins. Biochim Biophys Acta 1484, 93–106 Snyder, D. S., Zimmerman, T. R., Jr., Farooq, M., Norton, W. T., and Cammer, W. (1983). Carbonic anhydrase, 5’-nucleotidase, and 2’,3’-cyclic nucleotide 3’-phosphodiesterase activities in oligodendrocytes, astrocytes and neurons isolated from the brains of developing rats. J. Neurochem. 40, 120–127. Sogin, D. C. (1976). 2’,3’-Cyclic NADP as a substrate for 2’,3’-cyclic nucleotide 3’-phosphohydrolase. J. Neurochem. 27, 1333–1337. Sorg, B. A., Smith, M. M., and Campagnoni, A. T. (1987). Developmental expression of the myelin proteolipid protein and basic protein mRNAs in normal and dysmyelinating mutant mice. J. Neurochem. 49, 1146–1154. Sprinkle, T. J. (1989). 2’,3’-cyclic nucleotide 3’-phosphodiesterase, an oligodendrocyte-Schwann cell and myelin associated enzyme of the nervous system. CRC Crit. Rev. Neurobiol. 4, 235–300. Sprinkle, T. J., Lanclos, K. D., and Lapp, D. F. (1992). Assignment of the human 2’, 3’-cyclic nucleotide 3’phosphodiesterase gene to chromosome 17. Genomics 13, 877–880. Sprinkle, T. J., McMorris, F. A., Yoshino, J., and DeVries, G. H. (1985). DiVerential expression of 2’:3’-cyclic nucleotide 3’-phosphodiesterase in cultured central, peripheral, and extraneural cells. Neurochem Res. 10, 919–931. Sprinkle, T. J., Zaruba, M. E., and McKhann, G. M. (1978). Activity of 2’,3’-cylic nucleotide 3’-phosphodiesterase in regions of rat brain during development: Quantitative relationship to myelin basic protein. J. Neurochem. 30, 309–314. Stahl, N., Harry, J., and Popko, B. (1990). Quantitative analysis of myelin protein gene expression during development in the rat sciatic nerve. Mol. Brain Res. 8, 209–212. Starich, G. H., and Dreiling, C. E. (1980). Simultaneous inhibition of guinea pig brain 2’,3’-cyclic nucleotide 3’phosphohydrolase and myelin protein synthesis by 2’-adenosine monophosphate. Life Sci 18, 567–572. Staugaitis, S. M., Bernier, L., Smith, P. R., and Colman, D. R. (1990). Expression of the oligodendrocyte marker 2’3’-cyclic nucleotide 3’-phosphodiesterase in non-glial cells. J. Neurosci. Res. 25, 556–560. Stephon, R. L., Niedbala, R. S., Schray, K. J., and Heindel, N. D. (1992). An enzymatic cycling procedure for beta-NADPþ generated by 3’-phosphodiesterase, 2’:3’-cyclic nucleotide. Anal Biochem. 202, 6–9. Tenhunen, J., and Ulmanen, I. (1993). Production of rat soluble and membrane-bound catechol O-methyltransferase forms from bifunctional mRNAs. Biochem. J. 296, 595–600. Trapp, B. D., Bernier, L., Andrews, S. B., Colman, D. R. (1988). Cellular and subcellar distribution of 2’, 3’-cyclic nucleotide 3’-phosphodiesterase and its mRNA in the rat nervous system. J. Neurochem. 51, 859–868. Tsukada, Y., and Kurihara, T. (1992). 2’,3’-cyclic nucleotide 3’-phosphodiesterase: Molecular characterization and possible functional signiWcance. In ‘‘Myelin: Biology and Chemistry’’’ (R. E. Martenson, ed.), pp. 449–480. CRC Press. Vartanian, T., Sprinkle, T. J., Dawson, G., and Szuchet, S. (1988). Oligodendrocyte substratum adhesion modulates expression of adenylate cyclase-linked receptors. Proc. Natl. Acad. Sci. U S A. 85, 939–943. Vartanian, T., Szuchet, S., and Dawson, G. (1992). Oligodendrocyte-substratum adhesion activates the synthesis of speciWc lipid species involved in cell signalling. J. Neurosci. Res. 32, 69–78. Verity, A. N., and Campagnoni, A. T. (1988). Regional expression of myelin protein genes in the developing mouse brain: In situ hybridization studies. J. Neurosci. Res. 21, 238–248. Vogel, U. S., and Thompson, R. J. (1988). Molecular structure, localization, and possible functions of the myelinassociated enzyme 2’,3’-cyclic nucleotide 3’-phosphodiesterase. J. Neurochem. 50, 1667–1677. Waehneldt, T. V., and Malotka, J. (1980). Comparative electrophoretic study of the Wolfgram proteins in myelin from several mammalia. Brain Res. 189, 582–587.

III. THE MYELIN GENES AND PRODUCTS

521

522

19. 2’,3’-CYCLIC NUCLEOTIDE 3’-PHOSPHODIESTERASE

Walsh, M. J., and Murray, J. M. (1998). Dual implication of 2’,3’-cyclic nucleotide 3’ phosphodiesterase as major autoantigen and C3 complement-binding protein in the pathogenesis of multiple sclerosis. J Clin Invest. 101, 1923–1931. Wegner, M. (2000). Transcriptional control in myelinating glia: The basic recipe. Glia 29, 118–123. Weissbarth, S., Maker, H. S., Raes, I., Brannan, T. S., Lapin, E. P., and Lehrer, G. M. (1981). The activity of 2’,3’-cyclic nucleotide 3’-phosphodiesterase in rat tissues. J. Neurochem. 37, 677–680. Wilmot, G. R., Raymond, P. A., and AgranoV, B. W. (1993). The expression of the protein p68/70 within the goldWsh visual system suggests a role in both regeneration and neurogenesis. J Neurosci. 13, 387–401. Wilson, R., and Brophy, P. J. (1989). Role for the oligodendrocyte cytoskeleton in myelination. J. Neurosci. Res. 22, 439–448. Yin, X., Peterson, J., Gravel, M., Braun, P. E., and Trapp, B. D. (1997). CNP overexpression induces aberrant oligodendrocyte membranes and inhibits MBP accumulation and myelin compaction. J. Neurosci. Res. 50, 238–247. Yu, W. P., Collarini, E. J., Pringle, N. P., and Richardson, W. D. (1994). Embryonic expression of myelin genes: Evidence for a focal source of oligodendrocyte precursors in the ventricular zone of the neural tube. Neuron 12, 1353–1362. Zhang, X., and Miskimins, R. (1993). Binding at an NFI site is modulated by cyclic AMP-dependent activation of myelin basic protein gene expression. J. Neurochem. 60, 2010–2017.

III. THE MYELIN GENES AND PRODUCTS

C H A P T E R

20 The P0 Gene Daniel A. Kirschner, Lawrence Wrabetz, and Maria Laura Feltri

INTRODUCTION Protein zero (P0) is the major protein in myelin of the peripheral nervous system (PNS). Discovered and characterized nearly 30 years ago, P0 was found to be glycosylated and to constitute 50 to 60% of the myelin protein (Everly et al., 1973; GreenWeld et al., 1973). Its solubility properties were consistent with it being an integral membrane protein, perhaps analogous in function to proteolipid protein (PLP) of the central nervous system (CNS). As the major protein remaining in the well-myelinated, MBP-free sheaths of the shiverer mutant mouse PNS, P0 was proposed as playing the role of ‘‘structural cement’’ (Kirschner and Ganser, 1980)—that is, it was responsible for adhesion at both the extracellular and cytoplasmic surfaces of the myelin membranes. Subsequent studies of P0 have substantiated this idea and have moreover implicated this protein in certain demyelinating peripheral neuropathies. Thus, what has driven research on P0 has been its importance as a major PNS myelin protein, its structural role in membrane adhesion, and its involvment in peripheral neuropathy. The principal features of human P0 are summarized in Table 20.1. Essentially, this ~30 kDa (Mr) glycosylated, integral membrane protein is one of the simplest members of the immunoglobulin gene superfamily, having a single disulWde-stabilized VH-like domain, a single transmembrane domain, one glycosylation site, and relatively few other post-translational modiWcations (a single acylation site, one or more phosphorylation sites, and sulfation of the carbohydrate moiety). The N-linked oligosaccharide is of the complex type, and shows considerable heterogeneity pointing to additional roles for the protein (Gallego et al, 2001; Uyemura et al., 1992). The theoretical isoelectric point of the full protein is basic (~9–10), owing to the substantial net positive charge of the cytoplasmic domain (pI~11). P0 is a major constituent of peripheral myelin across a wide phylogenetic range (Fig. 20.1) that includes mammals, reptiles, birds, and amphibia; and in Wsh, including both teleosts and elasmobranchs, P0 or P0-like proteins are major components in both PNS and CNS myelin (Kirschner et al., 1989; Lanwert and Jeserich, 2001; Stratmann and Jeserich, 1995; Waehneldt et al., 1986). This chapter focuses on summarizing what is now known about P0 and updating what has been presented in some earlier and recent reviews (Eichberg, 2002; Filbin et al., 1997; Kirschner et al., 1996a; Quarles, 1997; Spiryda, 1998; Uyemura et al., 1992).

MOLECULAR GENETICS P0 Gene MPZ/Mpz P0 has been cloned from human (Hayasaka et al., 1991), rat (Lemke and Axel, 1985), mouse (Lemke and Axel, 1985; Lemke et al., 1988; You et al., 1991), chicken (Barbu, 1990), shark

Myelin Biology and Disorders, Volume 1

523

Copyright 2004, Elsevier (USA). All rights reserved.

524

20. THE P0 GENE

TABLE 20.1

Summary of Human P0 Characteristicsa

Feature

Description

Name

Myelin P0 protein (precursor)

Aliases

P0 glycoprotein, Myelin protein zero, myelin peripheral protein, MPP; myelin membrane adhesion molecule

Gene symbol

MPZ

Aliases

HMSN1B; CMT1; CMT1B

Chromosome mapping

1q21.3-q23

Gene structure

~7 kb, Exons I-VI

Accession numbers

P25189 (Swiss-Prot), 159440 (OMIM), 1NEU (PDB)

Subcellular location

Type 1 membrane protein

Tissue speciWcity

Schwann cells of peripheral nervous system

Structural information Unprocessed protein

248 amino acids, MW ¼ 27,555 Da

Signal sequence

residues 1-29b; Exons I & II 219 amino acids, MW ¼ 24,763 Da; /þ, 23/34; pI ¼ 9.57

Processed protein Extracellular

residues 30-153; MW ¼ 14,217 Da; Exons II & III; /þ, 17/13; pI ¼ 5.47; crystal structurec, 1NEU (PDB)

Transmembrane

residues 154-179; Exon IV

Cytoplasmic

residues 180-248; MW ¼ 7923 Da; Exons IV-VI; /þ, 6/21; pI ¼ 10.88

Domains

one immunoglobulin-like V-type domainc

DisulWde

between Cys50/21 and Cys127/98

c

d

Carbohydrate

N-linked (Asn122/93), complex

Post-translational modiWcations

acylation (Cys182/153); phosphorylation (Ser210/181, Ser233/204, Ser243/214; Tyr220/191 );

Allelic variants

f

Charcot-Marie-Tooth Neuropathy, Type 1B (CMT-1B); De´je´rine-Sottas Syndrome (DSS); Congenital Hypomyelination (CH)

a

Sources of data: SWISS-PROT: http://us.expasy.org/cgi-bin/niceprot.pl?P/25189; OMIM: www3.ncbi.nlm.nih.gov/htbin-post/Omim; PDB: http://www.rcsb.org/pdb/cgi/explore.cgi?pid ¼ 57471043166228&page ¼ 0&pdbId ¼ 1NEU, plus references footnoted below; Molecular mass (MW; includes neither the carbohydrate moiety, nor other post-translational modiWcations), charges (expressed as [Asp þ Glu]/[Arg þ Lys]), and pI calculated using ProtParam tool of ExPASy. b Numbering of residues in the literature may or may not include the signal sequence. In this table, when speciWc residues are cited, both numbers are given as preprotein/protein. c Lemke et al., 1988; Shapiro et al., 1996; Uyemura et al., 1987. d Gallego et al., 2001; Uyemura et al., 1992; Voshol et al., 1996. e Bizzozero et al., 1994. f Iyer et al., 2000.

(Saavedra et al., 1989), trout (Stratmann et al., 1995), and zebraWsh (Bro¨samle and Halpern, 2002; Schweitzer et al., 2003) (Fig. 20.1). In Wsh and amphibian, P0 or P0-like proteins are expressed both in central and peripheral nervous system myelin (Jeserich and Waehneldt, 1986). The cDNA for P0, which was Wrst isolated from the rat (Lemke and Axel, 1985), is 1.85 kb in length and has a coding sequence of 744 bp and a long 3’ untranslated region. It encodes a single RNA species of 1.9 kb. The predicted amino acid sequence, subsequently conWrmed by direct amino acid sequencing of bovine P0 (Sakamoto et al., 1987), consists of a 29-residue signal peptide, a 124-residue extracellular domain, a 26-residue transmembrane domain, and a 69-residue intracellular domain (Fig. 20.1). The extracellular domain contains an immunoglobulin-like fold, is stabilized by a disulWde bond between Cys21 and Cys98, and is N-glycosylated on Asn93. Owing to the large number of lysine and arginine residues and paucity of acidic residues, the intracellular domain is highly basic, having a theoretical pI
11 (Tab. 20.1). Three alternatively spliced isoforms that cause skipping of exon 3 have been reported in nerve and white blood cells (Besancon et al, 1999). These isoforms all cause a premature stop in exon 4 that should code for a protein truncated at the transmembrane domain and is thus probably not inserted in the membrane. The myelin protein zero gene (Mpz), Wrst isolated from rat and mouse, is relatively small (7 kb) and consists of 6 exons (Lemke et al, 1988; You et al., 1991). Exon 1 encodes for a 5’untranslated mRNA region, the signal peptide and Wrst 7 residues; exons 2 and 3 encode for

MOLECULAR GENETICS

525

FIGURE 20.1 Comparison of P0 sequences for diVerent species. Amino acids are numbered (according to the human sequence) both for the preprotein (residues 1 to 248, which includes the signal sequence, 1 to 29) and for the processed protein (1 to 219). The sequence boundaries corresponding to exons I-to-VI of the human P0 gene MPZ are indicated. The extracellular, transmembrane, and cytoplasmic domains are indicated by the horizontal bands of green, yellow, and pink, respectively. The multiple sequence alignment (Corpet et al., 1988) used here for the nine diVerent species reveals the extent of homology: 100%, white letters on red Weld; >50%, red letters. The sequence consensus is summarized below the species: capital letters, 100% consensus; lower case, >50%; !, Ile or Val; #, Asn, Asp, Gln, Glu; %, Phe, Tyr. References for species: human (Hayasaka et al., 1991), rat (Lemke and Axel, 1985), mouse (You et al., 1991), bovine (Sakamoto et al., 1987), chicken (Barbu, 1990), Xenopus (Kirschner et al., unpublished), horn shark (Saavedra et al, 1989), zebraWsh (Bro¨samle and Halpern, 2002), and trout (Stratmann and Jeserich, 1995).

the extracellular or ectodomain, exon 4 for the transmembrane domain, and exons 5 and 6 for the cytoplasmic domain (Fig. 20.1). Exon 6 also codes for the 3’ untranslated region. Human MPZ has a genomic organization similar to that of the mouse and rat genes, with six exons and conservation between exonic and promoter sequences; however, the sequence and length diVers between human and rodent (Pham-Dinh et al., 1993). The mouse Mpz and human MPZ are both located on chromosome 1 (Hayasaka et al,. 1993; Kuhn et al., 1990; You et al., 1991), whereas the rat gene is on chromosome 13 (Liehr et al., 1995). Mpz Homologs Recently, two genes with homology to Mpz have been reported: PZR and EVA. Human PZR, for ‘‘protein zero related,’’ is a widely expressed binding protein and possible substrate of tyrosine phosphatase SHP-2. Like P0, PZR is a single pass transmembrane protein with

526

20. THE P0 GENE

one immunoglobulin fold. Its extracellular domain shares 46% identity with P0, whereas its intracellular and transmembrane domains share no similarity with P0 (Zhao and Zhao, 1998). The gene for PZR is also referred to as myelin protein zero-like 1 (MPZL1). No data on the expression of PZR in peripheral nerve are available, and no mutations have been found in screening Charcot-Marie-Tooth families (Tang et al., 2000). Epithelial V-like antigen (EVA), a P0 homolog identiWed in thymocytes, has 33% amino acid identity with P0, is a single pass transmembrane protein with an immunoglobulin-like fold, and has homophilic adhesive properties when expressed in heterologous cells (Guttinger et al., 1998). It is thought to play a role as a homotypic adhesion molecule in thymus development. In addition to thymus, it is expressed in other embryonic and adult epithelia, such as gut, kidney, and liver. Interestingly, EVA mRNA is expressed in postnatal rat sciatic nerve and regulated during myelination (Feltri and Wrabetz, unpublished).

Gene Expression and Transcriptional Regulation Mpz expression has been studied extensively both in vivo and in vitro. Because Mpz is expressed almost exclusively in Schwann cells, and at extraordinary levels during myelinogenesis (P0 mRNA accounts for 8% of total mRNA and P0 glycoprotein more than 50% of total protein in nerve), unraveling the regulation of Mpz might have revealed mechanisms that deWne terminal diVerentiation of myelin-forming Schwann cells. Unfortunately, progress has been frustratingly slow. Mpz is expressed primarily in the Schwann cell lineage. In addition, Mpz mRNA is detected in otic placode and vesicle, notocord, enteric neural crest, and olfactory ensheathing cells (Lee et al., 2001). Before birth, basal levels of P0 mRNA Wrst appear in a subset of neural crest cells and remain in Schwann cell precursors and embryonic Schwann cells (Baron et al., 1994; Bhattacharyya et al., 1991; Lee et al., 1997; Zhang et al., 1995). P0 protein has been detected in the neural crest of chicken (Bhattacharyya et al., 1991) and on the surface of freshly cultured cells from E14.5 rat nerve (Lee et al., 1997). However, the levels of P0 mRNA and protein are remarkably lower in embryonic as compared to P1 nerves (Lee et al., 1997). No role has been assigned to P0 in other tissues, or in embryonic nerve, although P0/PMP22 co-expression marks a subset of pluripotent neural crest cells that respond diVerentially to TGF-ß in culture depending on their number (Hagedorn et al., 1999). In Mpz-null animals, cursory examination of other tissues or neonatal nerve has revealed no defects that could be attributed to loss of embyronic expression (R. Martini, personal communication). After birth, P0 mRNA levels distinguish the diVerentiation of non myelin-forming from myelin-forming Schwann cells; levels fall to undetectable in the former, and are induced remarkably in the latter (Lee et al., 1997). In both cases, axons determine the diVerentiated phenotype of the Schwann cell and its state of Mpz expression. Various agents repress (in some cases indirectly) Mpz expression in Schwann cells, including SV40 T-antigen with c-Jun (Bharucha et al., 1994), serum, GGF, TGFßs, bFGF2 (Cheng and Mudge, 1996; Fernandez-Valle et al., 1993; Mews and Meyer, 1993; Morgan et al., 1991; 1994), and neurotrophin-3 (NT3) via the TrkC receptor (Cosgaya et al., 2002). Some of these may maintain low-level Mpz expression in premyelinating Schwann cells, or further downregulate Mpz expression during diVerentiation of non myelin-forming Schwann cells. Myelin-forming Schwann cells require contact with axons for both the induction (Lemke and Chao, 1988; Scherer et al., 1994) and maintenance of high-level Mpz expression (Gupta et al., 1988; Trapp et al., 1988). Candidate mediators of inductive axonal signals include increased intracellular levels of cAMP (Lemke and Chao, 1988; Morgan et al., 1991) and brain-derived neurotrophic factor (BDNF) via p75neurotrophin receptor (Cosgaya et al., 2002). Axonally induced Mpz expression is also modulated in Schwann cells by signals originating from alterations in basal lamina/cytoskeletal linkage (Fernandez-Valle et al., 1993, 1997) and by progesterone (Koenig et al., 1995; Melcangi et al., 1998). Because its mRNA appears just before P0 glycoprotein in development, and its mRNA level is markedly upregulated during myelination, Mpz expression is thought to be primarily transcriptionally regulated (Lee et al., 1997; Lemke et al., 1988; Stahl et al., 1990). IdentiWca-

MOLECULAR GENETICS

tion of the genomic structure of Mpz and mapping of the 5’ end of exon 1 predicted a promoter region without a canonical TATAA box. However, 1.1 kilobases of this 5’ Xanking region activated expression of the CAT reporter gene speciWcally in transfected, cultured Schwann cells, but not other cell types (Lemke et al., 1988). Finally, transcriptional run-oV analysis of cultured Schwann cells treated with forskolin, a model of axonal induction of Mpz expression, showed directly that a change in the rate of transcription accounts at least in part for the induction of Mpz expression (Awatramani et al., 2002). In keeping with these data, the relatively small 1.1 kilobase P0 promoter fragment successfully activated expression of toxins, hormones, oncogenes, transcription factors, and enzymes speciWcally in Schwann cells in transgenic mice (Akagi et al., 1997; Giovannini et al., 1999; Messing et al., 1992; Messing et al., 1994; Nikitin et al., 1996; Weinstein et al., 1995; Yamauchi et al., 1999). However, transgene expression with the 1.1 kilobase promoter has been inconsistent. Thus, phenotypes suggesting transgene expression have been obtained with potent molecules (e.g., diphtheria toxin or Cre recombinase), although transgenic mRNA expression was low and was not measured as a direct proportion of endogenous P0 mRNA (Messing et al., 1992). Also, more consistent expression was obtained when the transgene contained both exons and natural introns from heterologous genes, as for growth hormone, SV40 T antigen, and connexin 32 (Bone et al., 1997; Messing et al., 1992, 1994). Instead, many attempts to express transgenes containing P0 promoter fragments fused to cDNAs or reporter genes, even with heterologous introns (e.g., rabbit b-globin) included, have had only limited success, sometimes requiring many independent transgenic lines in order to obtain even low-level expression (Maycox et al., 1997, and Wrabetz & Feltri unpublished observation). Thus, 1.1 kilobases of proximal Mpz promoter are suYcient to activate appropriate cell-speciWc and developmentally regulated expression, but it is unclear whether this region contains all regulatory elements for full amplitude of Mpz expression. In contrast, various insertions into exon 1 of an Mpz-based transgene, containing 6 kilobases of 5’Xanking region and 400 nucleotides of 3’ region produced not only appropriately regulated topographical and temporal expression of lacZ reporter (Feltri et al., 1999b), and Cre recombinase (Feltri et al., 1999a, 2002), but also levels of mRNA and protein expression comparable to endogenous Mpz (Previtali et al., 2000; Wrabetz et al., 2000; Yin et al., 2000). There have been no attempts in vivo to identify the subregions necessary for regulation in this transgene, and it is not even clear that regulatory elements outside the 1.1 kilobase proximal promoter need be invoked to explain the more consistent and higher level expression. First, the larger size of the whole Mpz may simply isolate the transgene better from regulatory eVects of surrounding chromatin. Second, it is important to note that improved expression from the whole Mpz gene as compared to the 1.1 kilobase promoter may not result from altered rates of transcription. Mechanisms that regulate message stability or translation may be relevant, since all exons, including the 3’ untranslated region are included in the Mpz construct and in the resulting transgenic fusion transcript. Given its suYcience for Schwann cell-speciWc and developmentally regulated expression of heterologous genes in vivo, the cis-acting elements within the 1.1 kilobase Mpz promoter have been thoroughly dissected in studies of cultured Schwann cells (Brown and Lemke, 1997). Several DNA sequences in the proximal 350 nucleotides are bound by various transcription factors, including Sp1 and NF-Y, but the basis for cell-speciWc activation remains unclear. The major diYculty for these studies may be the cell model. Schwann cells cultured in isolation from axons, with growth factors and forskolin to support their survival and proliferation, probably model premyelinating Schwann cells, where expression of Mpz is basal. It is possible, therefore, that the trans-acting factors that cooperate to markedly induce Mpz transcription in myelin-forming Schwann cells are not recruited to the Mpz promoter under those circumstances. Finally, genetic alterations that produce dysmyelination in mouse and human (Table 20.2), together with expression analysis, suggest that several other transcription factors are upstream regulators of Mpz in developing nerve, including Pou3f1(Oct6/SCIP/ TST1), Krox20, and Sox10 (see chapters 39 and 48 in volume 2 for details). Whether

527

528

20. THE P0 GENE

regulation is direct remains unclear for all three. Oct6 strongly represses the Mpz promoter in vitro (Monuki et al., 1989), but is likely an activator in vivo based on genetic evidence (Bermingham et al., 1996; Jaegle et al., 1996; and see Topilko and Meijer, 2001, for discussion). Krox20 is absolutely required in vivo for terminal diVerentiation of myelinforming Schwann cells (Topilko et al., 1994), as Krox20-null mice produce none of the normal 100-fold up-regulation of Mpz expression in post-natal nerve. Krox20 also induces Mpz expression 59-fold within 24 hours of infection in Schwann cells transduced with a Krox20 adenovirus vector (Nagarajan et al., 2001), whereas Krox20 only activates the P0 promoter two to four-fold in co-transfection studies (Zorick et al., 1999). Finally, the strongest evidence for direct regulation is for Sox10, which is required for formation of Schwann cells (Britsch et al., 2001), is expressed at all stages of Schwann cell life and upregulates both endogenous Mpz expression, as well as the Mpz promoter 10- to 20-fold when ectopically expressed in N2A neuroblastoma cells (Peirano et al., 2000). Both proximal and distal portions of the 1.1kilobase Mpz promoter contain probable Sox10 binding sites, but chromatin cross-linking followed by immunoprecipitation (ChiP) experiments will be required to conWrm Sox10 binding at these sites in Schwann cells actively transcribing Mpz.

FUNCTION AND CELL BIOLOGY Synthesis, Post-Translational Modification, and Targeting During development, Mpz expression and P0 mRNA amounts increase dramatically, and in the adult, they fall to steady-state levels (reviewed by Gould et al., 1992). Corresponding to this temporal proWle, the level of P0 in the PNS increases during development and correlates with myelin formation. Like the proteins of other plasma membranes, P0 synthesis is localized to the endoplasmic reticulum, is processed through the Golgi network, and Wnally targeted (perhaps via a YAML motif in its cytoplasmic domain) to the plasma membrane of Schwann cells (reviewed by Eichberg, 2002). Post-Golgi sorting of P0 is presumably based on signals inherent to the cytoplasmic or extracellular domains of the polypeptide, and transport to the internodal site of myelin formation is mediated by microtubules. As summarized in Table 20.1 and depicted in Figure 20.2, post-translational modiWcations of P0 include in the extracellular domain: disulWde bond formation between Cys21 and Cys98, which stabilizes the pair of apposed b-sheets of the Ig fold (Shapiro et al., 1996); glycosylation of Asn93 with a complex carbohydrate (Gallego et al., 2001; Poduslo, 1990; Uyemura et al., 1992); and sulfation of the N-acetylglucosamine (GlcNAc) residues of the carbohydrate (reviewed by Uyemura et al., 1992; Gallego et al., 2001). In the cytoplasmic domain, the post-translational modiWcations include acylation of Cys153 with palmitic acid (Bizzozero et al., 1994); and phosphorylation of Tyr191, Ser181, and Ser204 (reviewed by Eichberg, 2002).

Cell Adhesion The extensive intermembrane adhesion in myelin is particularly evident from electron micrographs of chemically Wxed, internodal compact sheaths, which show the membranes to be closely apposed at their cytoplasmic (‘‘major dense line’’) and extracellular surfaces (‘‘double intraperiod line’’). X-ray diVraction studies on unWxed myelinated nerves show that the intermembrane spaces are, indeed, very narrow—for example, in human PNS (sciatic nerve) myelin, the cytoplasmic space is 3.5 nm and the extracellular space is 5.3 nm (reviewed by Kirschner and Blaurock, 1992; these spaces are deWned by the center-tocenter distance between lipid headgroup layers that border the space separating neighboring membranes); and across phylogeny, from amphibians to humans, these values range from 3.3 to 3.8 nm and 4.5 to 5.4 nm for the cytoplasmic and extracellular spaces, respectively (Tab. 20.3). Studies on the reversible eVects of ionic strength and pH on myelin membrane packing emphasize the stability of the cytoplasmic apposition and the lability of

FUNCTION AND CELL BIOLOGY

529

FIGURE 20.2 Structural aspects and mutations for human P0. The primary sequence of the processed protein is illustrated (residues 1 to 219; see Fig. 20.1). Acidic, basic, and His residues are denoted by backgrounds of blue, pink, and light pink, respectively. The extracellular, transmembrane, and cytoplasmic domains are indicated by the letter colors dark-blue, yellow, and red, respectively. Residues having post-translational modiWcations—disulWde between Cys21 and Cys98, N-linked glycosylation at Asn93, acylation at Cys153 (Bizzozero et al., 1994), and phosphorylation at Tyr191 (Iyer et al., 2000) and putatively at Ser181, Ser204, Ser214—are highlighted by a larger font. Structural features of the extracellular domain of P0 that are based on crystallographic analysis (Shapiro et al., 1996) and depicted here include the positions of the 10 b-strands (horizontal bars over the residues, labeled A, A’, B, C, C’, C’’, D-G), and the intermolecular (crystal) contacts between the protomers, where v is the interface between each head-to-tail pair of like-directed (parallel) molecules within each cyclic tetramer, is the twofold adhesion interface between the sides of molecules from laterally arranged, oppositely directed (antiparallel) tetramers, and is the head-to-head contact between monomers of the packed, apposed tetramers in the crystal. Below the primary sequence are indicated the known mutations in MPZ, which include amino acid substitutions (moderate to severe phenotypes shown in red), deletions (D), frameshifts (fs), polymorphisms (o), and stops (X), and the replacement of four residues at 57 to 60 by three and the insertion of two residues after D89 are shown. See also Table 20.2.

the extracellular apposition—that is, while the former can change in width by at most 0.5 nm, the latter changes by 8 to 20 nm or more (reviewed by Kirschner et al., 1984). In general, two factors determine the distance between membrane surfaces: (1) molecular contacts between apposed constituents that protrude into the spaces between the surfaces, which can either promote or hinder close approach, and (2) a balance of attractive and repulsive ‘‘nonspeciWc’’ forces (van der Waals, electrostatic, and hydration forces). Whereas a simple balance of nonspeciWc forces accounts for much of the change in intermembrane distances, some striking interactions cannot be explained—for example, the coexistence of swollen and compact membrane packings, discontinuous changes from one type of packing to another, and the virtually invariant cytoplamic spacing. Clearly, not only are there speciWc factors regulating membrane packing, but also there are diVerent mechanisms of adhesion at the two membrane surfaces. What are the molecular elements underlying these mechanisms? Because the intermembrane spacing values cited here are signiWcantly larger than the 1.5 to 3 nm separations for diVerent protein-free, phospholipid-cholesterol lipid multibilayers in water, then it must be the myelin protein that accounts for the diVerence; both freezefracture and x-ray diVraction studies support this conclusion (reviewed in Kirschner et al., 1984; see also Inouye and Kirschner, 1988). That it is speciWcally the P0 protein that accounts for spacing and adhesion at both membrane surfaces derives from application of these same techniques to studying myelin membrane packing in sciatic nerves from normal and MBP-deWcient shiverer mice, where the nerves have been treated so as to produce two diVerent membrane domains—one containing intramembranous particles and the other domain particle-free. The former one maintains the native period, while the latter one assumes

530

20. THE P0 GENE

a period that corresponds to twice the thickness of a protein-free lipid bilayer. As P0 is the predominant protein remaining in shiverer PNS myelin, then it must be P0 that corresponds to the intramembranous particles and that maintains the normal membrane packing at both cytoplasmic and extracellular appositions. This role of P0 in myelin membrane adhesion based on structural analyses is consistent with subsequent molecular genetics experiments, which proposed that this protein, as a primitive member of the Ig-superfamily, likely functions in cell-cell recognition or adhesion (Lemke et al., 1988). TABLE 20.2 MPZ Mutationsa #Pre

#Pro

Orig

Change

30

1

I

M

32

3

V

F, severe

34

5

T

I

42

13

V

D , dm with A221/192T, severe

b

44

15

S

F

54

25

S

C, severe; P

58

29

V

F, moderate

61

32

D

Gc

63

34

S

D; C; F; Ld, mild or moderate

64

35

F

D

65

36

T

Il

ED

68

39

Y

C, severe

71

42

E

X , mild

75

46

D

frameshift; Vf, mild.

78

49

S

L , severe

81

52

H

R, severe; Yh, dm with V113/84F, mild

e

g

82

53

Y

C

86–89

57–60

QPYI

HLF (8 kb!5 kb; severe)i

90

61

D

E

92

63

V

V, polymorphism

93

64

G

E

96

67

K

E

98

69

R

C, severe; H; P; S

99

70

I

T, moderate

101

72

W

C

102

73

103

74

G

Ej, severe

frameshift

112

83

I

T, severe

113

84

V

Fh, dm with H81/52Y, mild; Ik

114

85

I

Tl, tm with N116/87H and D128/99N, severe

116

87

N

Hl, tm with I114/85T and D128/99N, severe

118

89

D

DFY (insertion)

119

90

Y

C , (CMT2)

c

g

122

93

N

S , moderate to severe

124

95

T

D; Mf, mild.

127

98

C

Y, severe

128

99

D

E; N , tm with I114/85T and N116/87H, severe

l

130

101

K

R

131

102

N

Km, mild

132

103

P

L, moderate (Continues)

531

FUNCTION AND CELL BIOLOGY

TABLE 20.2 (continued) #Pre

#Pro

Orig

Change

134

105

D

E; N

135

106

I

L; T

137

108

G

S

141 143

112 114

Q T

X (CMT2) M

154

125

Y

X

163

134

G

R

167

138

G

A, severe; R, severe; frameshift

170

141

L

Rk

n

172

143

frameshift

174

145

frameshift

181

152

Y

185

156 171

204

175

206

177

G

Xo, mild

215

186

Q

X, severe

192

224

195

228

199

233

204

244

215

TM

X

200

221

ED

frameshift G

G, polymorphism frameshift

A

CD

b

X; T , dm with V42/13
, severe frameshift

S

S, polymorphism frameshift, severe

R

L

a

Key, abbreviations, sources: #Pre: amino acid numbering of preprotein includes signal sequence; #Pro: amino acid numbering of processed protein; Orig: residue in normal protein; Change: result of genetic mutation in MPZ (including amino acid substitutions, truncations (X), frameshifts, polymorphisms), with severity of disease indicated (if apparent from the data sources); dm: double mutation; tm: triple mutation. Sources of data: SWISSPROT: http://us.expasy.org/cgi-bin/niceprot.pl?P25189; OMIM: http://www3.ncbi.nlm.nih.gov/htbin-post/Omim; Nelis et al. (1999); and footnoted references. The extents of the extracellular (ED), transmembrane (TM), and cytoplasmic (CD) domains are indicated by the vertical arrows (solid line indicating at the boundary, dashed line indicating not at boundary). b Plante´-Bordeneuve et al., 2001. c Senderek et al., 2000. d Fabrizi et al., 2000. e Lagueny et al., 2001. f Misu et al., 2000. g Sindou et al., 1999; J-M. Vallat, personal communication. h Bienfait et al., 2002. i Silander et al., 1996; K. Silander, personal communication. j Fabrizi et al., 2001. k Numakura et al., 2002. l Warner et al., 1997. m Plante´-Bordeneuve et al., 1999. n Young et al., 2001. o Senderek et al., 2001.

Formation of Double Intraperiod Line Electron micrographs of developing myelin sheaths show the spiral infolding of Schwann cell processes. During this dynamic process of myelination captured by chemical Wxation, one may be fortunate enough to observe a regular, narrowed apposition of extracellular membrane surfaces and a substantial but thin and irregular layer of cytoplasm still separating the inner membrane surfaces (e.g., see Figure 1.7 in Landon and Hall, 1976, and Figure 6-7 in Peters et al., 1991). This morphology suggests that the end-to-end interaction of P0 molecules across the extracellular space might be the Wrst homophilic interaction between molecules on opposing membranes, occurring before the ensuing process leads to further narrowing at both appositions—which result in forming the double intraperiod line and the major dense line. The widened but regular extracellular space

532

20. THE P0 GENE

corresponds to the incompletely swollen state of myelin that is observed during the discontinuous swelling of membranes from their native separation when the ionic strength is lowered to 0.12–0.14 or less from 0.15 (Inouye and Kirschner, 1988). Further, this particular homophilic interaction could correspond to the head-to-head interaction between P0 monomers of the crystal-packed tetramers (Shapiro et al., 1996; discussed later). In vitro studies of various transfected cell lines or myelinating cultures, and in vivo studies on transgenic mice provide additional evidence that the homophilic interaction of P0 may be involved in myelin compaction. Cells transfected with P0 acquire the ability to adhere to each other (D’Urso et al., 1990; Filbin et al., 1990; Schneider-Schaulies et al., 1990). The adhesive plane between adjacent cells is narrow (though wider than in myelin), contains P0, and often forms desmosomes; and anti-P0 antibodies inhibit the adhesion (D’Urso et al., 1990; Filbin et al., 1990; Schneider-Schaulies et al., 1990). The cytoplasmic domain of P0 is required for extracellular adhesion to occur, and truncation of the cytoplasmic domain inhibits adhesion of full-length P0 by a dominant-negative mechanism (Wong and Filbin, 1994, 1996). In addition, chimeric Wsh/rat P0 in this system is strongly adhesive only when rat P0 cytoplasmic domain is present (Lanwert and Jeserich, 2001). The same in vitro paradigm using transfected cell lines also identiWed the possible role of post-translational modiWcations to P0’s adhesive function. Glycosylation at Asn93 with complex carbohydrate (Filbin and Tennekoon, 1991, 1993), disulWde bond formation (Zhang and Filbin, 1994), and possibly acylation (Gao et al., 2000) are all apparently required for adhesive function. Antibodies against the L2/HNK1 chain partially inhibit P0 homophilic interaction, also supporting a role for glycosylation (GriYth et al., 1992). In vitro and in vivo myelination experiments further conWrm P0’s role in adhesive homophilic interaction and formation of the intraperiod line. First, Schwann cells infected with a retrovirus coding for P0 antisense RNA show diminished levels of P0 protein and are either unable to myelinate, or form myelin with uncompacted lamellae when co-cultured with DRG neurons (Owens and Boyd, 1991). Similarly, mice deWcient in P0 show severe hypomyelination, with absent or poorly compacted sheaths, and supernumerary Schwann cell processes (Giese et al., 1992). Intraperiod lines are diVusely absent, conWrming a role in homophilic adhesion. Surprisingly, some myelin sheaths show compaction, particularly at the major dense line, suggesting that other molecules compensate for the loss of P0. Indeed, up-regulation of L1, NCAM, MAG, PLP, and tenascin is found in P0 null myelin (Giese et al., 1992); however, NCAM/P0 doubly deWcient mice have a phenotype identical to P0 null mice (Carenini et al., 1999), suggesting that NCAM does not compensate for P0 in the function of compaction. Formation of Major Dense Line The unique nature of the cytoplasmic apposition between Schwann cell plasma membranes in compact myelin is evident from electron microscopy and membrane diVraction studies. First of all, one must realize that the appearance of a ‘‘fused’’ apposition is an artifact of processing prior to thin-sectioning. X-ray diVraction measurements carried out on dissected sciatic nerve after each step in the chemical Wxation, dehydration, and embedding procedure demonstrate that what was originally a double line bordering the cytoplasmic apposition— where each line is centered at the lipid headgroup layers—becomes, after osmication and during the dehydration steps, a single electron dense peak centered at the boundary between membranes (Kirschner and Hollingshead, 1980). By contrast, the intraperiod line retains its ‘‘double’’ appearance, pointing to a major chemical diVerence between the two surfaces of the membrane. Contrary to its appearance by electron microscopy as an occluded space, then, the cytoplasmic apposition has a Wnite width (3.3 to 3.8 nm; noted earlier). Moreover, diVraction experiments establish that the cytoplasmic space is readily accessible to ions and sucrose (Blaurock, 1971), and to water (Kirschner et al., 1975). In fact, the diVusion time for water in intact myelin sheaths is about 15 minutes, which is comparable to that in cholesterol-lecithin multilayers (Franks and Lieb, 1980). Although the cytoplasmic space is aqueous and accessible, its width is surprisingly stable to changes in pH and ionic strength (described earlier). The strength of the intermembrane interaction at the major dense line is further emphasized if one considers that the vigorous shearing forces that are carried out under hypotonic conditions during myelin isolation are not suYcient to separate myelin

FUNCTION AND CELL BIOLOGY

lamellae at this apposition, which moreover retains its invariant spacing with changes in pH and ionic strength (Inouye et al., 1989). It is commonly written that electrostatic interactions between the highly basic cytoplasmic domain of P0 and opposing acidic phospholipids underlie compaction of the cytoplasmic myelin leaXets and formation of the major dense line—indeed, a peptide containing 65 of the 69 residues of the intracellular domain of P0 binds to and aggregates artiWcial phospholipid vescicles (Ding and Brunden, 1994). This hypothesis is strongly questioned, however, by the evidence summarized in the foregoing paragraph, namely that the water- and ion-accessible cytoplasmic space remains virtually unaltered across wide variations of pH and ionic strength. It has been proposed (Inouye et al., 1999) that this very stable apposition may instead depend on lipid-anchoring and additional hydrophobic ˚ -long hydrocarbon interactions between apposed P0 cytoplasmic domains. The ~20 A chains of the palmitic acid that acylates Cys153 (Bizzozero et al., 1994) near the membrane surface may be directed away from the surface to interact with an apposing hydrocarbon chain. Amphiphilic a-helical sequences and hydrophobic side chains of b-strands, which are predicted secondary structure elements in the cytoplasmic domain of P0, may also stabilize a hydrophobic pocket at the cytoplasmic apposition. Finally, mouse genetics experiments indicate that P0 and MBP cooperate in the formation of the major dense line in peripheral myelin. Some myelin sheaths in P0 null mice still contain major dense lines, but myelin sheaths in mice deWcient for both P0 and MBP are completely uncompacted and have no major dense lines (Martini et al., 1995b). Heterophilic Adhesive Function? P0 mRNA and protein are expressed at low levels before birth in a subpopulation of neural crest cells and subsequently in Schwann cell precursors (Baron, 1994; Bhattacharyya, 1991; Lee, 1997), and at least P0 mRNA is expressed in otic placode, enteric nervous system and olfactory ensheathing cells (Lee et al., 2001). Thus, P0 may also have additional functions outside the myelin sheath. One possibility is that P0 may mediate heterophilic adhesion to neurites. Cells expressing ectopic P0 not only adhere to each other, but also promote neurite outgrowth, suggesting that P0 may mediate adhesion to neurites during development (Schneider-Schaulies, 1990). Altered prenatal development in P0-null mice has not yet been reported. In addition, a series of observations predict a possible role for P0 in the maintenance of axons. First, both P0 heterozygous and homozygous null mice develop axonal degeneration (Frei et al., 1999; Giese et al., 1992; Martini et al., 1995a). Second, several P0 mutations in patients cause the axonal form of Charcot-Marie-Tooth, (CMT-2, see Chapter 39 by Wrabetz et al.), with minimal myelin involvement, pupillary signs, and deafness (De Jonghe et al., 1999; Marrosu et al., 1998). Finally P0 may function ed1 in axonal regeneration. P0 is expressed by olfactory-ensheathing glia, cells in contact with axons that regenerate throughout life, and that support axonal regeneration in the central nervous system (reviewed in Franklin and Barnett, 2000). Also, P0 is re-expressed during regeneration, when it switches from a complex-type oligosaccharide form to the high-mannose carbohydrate form (Poduslo et al., 1985); this transition could support a change from homophilic (Filbin and Tennekoon, 1991) to heterophilic adhesive function.

Signaling Based on extensive morphological studies of mice having double mutations in genes that encode for diVerent myelin proteins, Billings-Gagliardi et al. (1987) proposed an hypothesis of multiple primary gene functions: a gene could code for a protein that not only has a structural role in myelin but also a ‘‘cytogogic’’ role—that is, directing some cellular function. While this hypothesis was derived from observations on mice with mutations in the genes for proteolipid protein and myelin basic protein, the principle may extend to P0, which like other adhesion receptors could also have signal transduction roles, either during cellular diVerention (outside-in signaling) or in regulation of its ectodomain adhesiveness (inside-out signaling). In the Wrst case, a signal may be transduced by P0 to

533

534

20. THE P0 GENE

regulate cellular functions. In the second case, the cytoplasmic domain of P0 may signal to the extracellular domain, regulating its own adhesiveness. This would resemble the inside-out signaling mechanism well described for other adhesion receptors such as integrins. A signal transduction role for P0 is supported by the Wnding of phosphorylation of serine and tyrosine residues in the cytoplasmic domain (Brunden and Poduslo, 1987; Hilmi et al.; 1995, Iyer et al., 1996; Suzuki et al., 1990). Outside-in Signaling There is evidence suggesting that P0 controls aspects of Schwann cell diVerentiation, such as polarization and gene expression. First, when P0 is introduced into carcinoma cell lines it is able to induce the expression and redistribution of adherens junction proteins and the formation of adherens junctions, desmosomes (Doyle et al., 1995), and physiologically operative tight junctions (Spiryda and Colman, 1998a). Furthermore, expression of P0 is able to revert transformation of these cells (Spiryda and Colman, 1998b). Finally, in P0null mice, myelin gene expression is altered, MAG, E-cadherin, and b-catenin are mislocalized, and adherens junctions are absent (Menichella et al., 2001). These data together suggest that P0 may signal to regulate Schwann cell polarization, junction formation, and gene expression during myelination. Inside-out Signaling Recent evidence suggests that serine phosphorylation of P0 may be important for regulated adhesion. Serine residues in the cytoplasmic domain of P0 are phosphorylated by PKC, mostly during active myelination (reviewed, in Eichberg and Iyer, 1996). Futhermore, P0 binds PKCa and the PKC-binding protein RACK1. Mutations of the RSTK motif cause loss of P0 adhesiveness in a cell system and cause CMT1B neuropathy in humans, suggesting that phosphorylation at this site is required for regulated adhesion of the extracellular domain (Xu et al., 2001). Whereas we reviewed indications that serine phosphorylation may be important in inside-out signaling (discussed earlier), the role of tyrosine phosphorylation remains to be determined. Tyrosine phosphorylation is higher during active myelinogenesis (Iyer et al., 2000), occurs mainly at residue Tyr191 (Iyer et al., 2000; Xu et al., 2000), and causes association with other phosphoproteins (Xu et al., 2000) suggesting a role during myelination. However, mutations at this site do not inXuence P0 adhesiveness (Xu et al., 2001). It is therefore possible that tyrosine phosphorylation of P0 serves a diVerent role, possibly in regulation of its own expression (discussed in Xu et al., 2000) or in other aspects of diVerentiation.

Structure As described earlier, P0 has three domains: extracellular, transmembrane, and cytoplasmic. To date, only the recombinant extracellular domain corresponding to the sequence for rat has been crystallized (Shapiro et al., 1996); however, small-angle x-ray scattering has been carried out on the full (glycosylated) polypeptide isolated from bovine PNS tissue (Inouye et al., 1999). The crystal data indicate the atomic structure of the extracellular domain and suggest what homotypic interactions might be relevant to P0 in the native myelin membrane, while the scattering data indicate the oligomeric state of the protein in a membrane-like environment. The aqueous extracellular domain of P0 was crystallized at slightly alkaline pH, conditions where the myelin period is native. The atomic structure of this domain shows that the asymmetric unit is a single molecule, or protomer, having an immunoglobulin variable-like fold, and forming cyclic tetramers (Fig. 20.3A). In the standard immunoglobulin designation for the disulWde-bonded b-sandwich, one b-sheet is constituted of strands D, E, B, and A, while the other consists of strands A’, G, F, C, C’, and C’’. The three diVerent sites of interactions that the protomer exhibits with neighboring P0 molecules, which are based on the crystal contacts, may indicate some of the homotypic interactions of P0 in the native membrane. Figure 20.2 indicates which residues are involved in these interactions. The Wrst

535

FUNCTION AND CELL BIOLOGY

FIGURE 20.3 Molecular structure of P0 extracellular domain and its tetrameric arrangement. The molecule is represented by its Ca backbone; and the disulWde bond in each molecule is indicated in yellow, and the N- and C-termini are shown. (A) The tetramer of protomers is viewed looking down onto the membrane surface. The loop postulated to be part of the adhesive interface (panel C) is in pink, and side chains for Trp28 and Trp78 are indicated. (B) The tetramer is viewed perpendicular to that in (A), parallel to the membrane surface. The C-termini point down toward the membrane surface, and the Trp28 side chains at the apices of the molecules are directed toward the apposing membrane surface. (C) Lateral view (parallel to the membrane surfaces) of two protomers forming the postulated adhesive interface. The left molecule emanates from the upper membrane, and the right molecule comes from the lower membrane. The C-termini are truncated by Wve residues, as these are too disoriented in the crystal for atomic resolution. The Trp28 side chains at the apices may be intercalated into the apposing membrane. (D) The head-to-head interface, which might correspond to earliest events in myelin formation when Schwann cell plasma membranes make initial contact with one another. Reproduced with permission from Shapiro et al., 1996.

packing interface is between the tilted, head-to-tail protomers within the tetramer, and shows relatively sparse interactions (Figs. 20.3A and B). The second packing interface is between laterally arranged, antiparallel protomers belonging to apposed tetramers, and is proposed to be the adhesive interface (Fig. 20.3C). The third interface is a head-to-head contact between the apices of apposed P0 extracellular domains (Fig. 20.3D). The Wrst two ˚ -wide space similar molecular interfaces in the crystal show contacts within a narrow, ~46 A in width to the extracellular separation in native myelin (Inouye & Kirschner, 1988). The third interface likely corresponds to the incompletely swollen state of myelin that precedes the compaction of the spirally infolded multilayer that occurs during myelination (discussed earlier). Supporting evidence that the tetrameric assembly of P0 protomers may be physiological are the Wndings that such an oligomerization of the extracellular domains

au2

536

20. THE P0 GENE

(recombinant rat sequence) is energetically favorable in solution as shown by analytical ultracentrifugation (Shapiro et al., 1996), and that a similar tetrameric assembly of full sequence P0 molecules (isolated from bovine) is detected by x-ray scattering of the protein in dilute sodium dodecylsulfate, which mimics the membrane lipid environment (Inouye et al., 1999). Tetrameric as well as dimeric P0 are detected in peripheral myelin protein of Xenopus, and when native gels are used, the major stable species is the dimer (Thompson et al, 2002), which is consistent with the more extensive protomer interactions at the proposed adhesive interface (Shapiro et al., 1996). A major constituent of P0 is the carbohydrate N-linked at Asn93. This moiety accounts for nearly 20% of the molecular mass of the extracellular domain of the glycoprotein. Using DEAE-Sephadex chromatography, Uyemura et al. (1992) originally identiWed Wve glycopeptides, the major one of which contained a nonasaccharide. Using the recently developed and far more sensitive technique of magic angle spinning proton NMR spectroscopy in a nanoprobe and in combination with mass spectrometry, Gallego et al. (2001) report a far more complex and heterogeneous structure. While the core structure is

Man 1 GlcNAc1 Man 1

6 4 Man 1-4GlcNAc1-4(Fuc1-6)GlcNAc-Asn933

as previously determined, the nonreducing termini af the Man-a1,6- and Mana1,3-branches show great variety, and include an HNK-1 epitope, (6-O-sulfo) HNK-1, 6-O-sulfo sialyl Lewis X, disialo, and galactosyl structures. Based on its estimated size and localization near the membrane-proximal base of the molecule (Wells et al., 1993), this carbohydrate may have a role in orienting the folded polypeptide correctly at the membrane surface. Additionally, the complex epitope library inherent in this structural diversity suggests cytogogic involvment in signal transduction, diVerentiation, and cell adhesion (Gallego et al., 2001). Neither full sequence P0 (containing the transmembrane domain) nor its cytoplasmic domain have yet been crystallized. Across a wide phylogenetic range the cytoplasmic appos˚ -wide space (Inouye & Kirschner, 1988; ition in myelin is typically a very narrow, 33 A Kirschner et al., 1989). This apposition is stabilized by very strong adhesion, as shown by ‘‘electrostatic stressing’’ across widely varying pH and ionic strength (discussed earlier). As indicated previously, these results argue against the idea that ionic interaction between basic sidechains of P0 and acidic head groups of lipids mediate the cytoplasmic adhesion. Sequence analysis of the cytoplasmic domain of P0 does not indicate homology to any other protein, but does suggest that there is an a-b-a folding pattern (Inouye & Kirschner, 1991).

GENOTYPE/PHENOTYPE COMPARISONS AND HEREDITARY NEUROPATHIES In terms of myelin structure, what are the phenotypic ramiWcations of genetic diVerences in MPZ/Mpz? That is, how do alterations in the primary structure of P0 aVect its role in membrane adhesion? Understanding this relationship could provide insight into the mechanism of dys- or demyelination in certain hereditary neuropathies where the diVerences in MPZ arise from mutations. Because P0 mediates myelin adhesion at both membrane appositions (discussed earlier), the way to approach this question must be by methods that have the spatial resolution to distinguish between the extracellular and cytoplasmic surfaces within the myelin sheath. Electron microscopy of thin-sectioned material and xray (or neutron) diVraction on unWxed nerves are the only methods that are capable of ˚ or more detecting such diVerences. While both can detect changes on the order of ~30 A (such as the approximate diVerence in period between CNS and PNS myelins), only the ˚ ngstroms. Thus, electron microscopy is well latter can resolve changes at the level of A suited to scrutinizing material where there may be relatively large changes in myelin period

537

GENOTYPE/PHENOTYPE COMPARISONS AND HEREDITARY NEUROPATHIES

TABLE 20.3 Comparison of Myelin Periods, and Cytoplasmic and Extracellular Spaces among Species Specimena

˚) Appositionc (A

˚) Myelin Periodb (A Cytoplasmic

Extracellular

Primate

180

37

47

Rodent

176–178

34–35

~47

Bovine

182

39

46–48

Avian

182

38

48

Amphibian

170–174

33–34

45–48

Teleost

158–161

24–26

36–47

Elasmobranch

177–178

36–37

37

a Data for the diVerent species, which are indicated as a range of values, are from native, unWxed samples of peripheral nerve, as summarized from Kirschner et al. (1989), except where noted: primate – monkey; rodent – hamster, guinea pig, rat, mouse; avian – chicken (Blaurock, 1986); amphibian – bullfrog, frog, mudpuppy; teleost – ‘‘Jack Dempsey’’, goldWsh, trout; elasmobranch – dogWsh shark, torpedo. b Measured directly from x-ray diVraction patterns. c Determined from membrane proWles that were calculated from the periods and scattered intensities of the xray patterns. The widths of the appositions are measured between the middles of the lipid headgroup layers across the interbilayer spaces.

TABLE 20.4

Sequence Identities among P0 from Different Species (%)a Human

Rat Bovine Chicken

Full

Cytoplasmic

Extracellular

94 93 78

91 93 79

97 93 81 Rat

Xenopus Trout Shark

Full

Cytoplasmic

Extracellular

65 44 54

72 14b 53

71 59 57

a

Percentages calculated using ClustalW (Combet et al., 2000; Thompson et al., 1994). Trout with 30 residues and rat with 69 have only 10 identical residues.

b

owing to expansion of the spaces between membranes, but it is not appropriate for detecting subtle but real changes in membrane-membrane packing. Examination of the sequence homologies among a phylogenetic range of vertebrates (Fig. 20.1) shows the extent of absolute conservation; and while electron microscopy shows compact myelin with about the same periodicity for all of these species, x-ray diVraction shows some signiWcant diVerences (Kirschner et al., 1989). Table 20.3 summarizes the period and widths of the intermembrane spaces for most of the species included in Figure 20.1. Despite extensive sequence identity among primate, rodent, bovine, chicken, and amphibian P0s (Tab. 20.4), there are small diVerences in myelin period and membrane packing, particularly at the cytoplasmic apposition. Whether this is due to the cytoplasmic domain of P0 or to other myelin proteins (such as P2, Blaurock, 1986; or MBP, or PMP22) is not known. The most striking packing diVerences are among teleost, elasmobranch, and the higher verte˚ narrower cytoplasmic apposition in teleost compared to the brates. In particular, the ~10 A others is most likely due to the truncated cytoplasmic domain of P0, which is shorter by nearly ˚ narrower extracellular 40 amino acid residues (see Fig. 20.1). Another diVerence is the ~10 A space in elasmobranch compared to the higher vertebrates, and this might be caused by an additional kink close to the membrane surface due to a second proline insertion immediately

538

20. THE P0 GENE

following Pro122 and subsequent reorientation of the extracellular domain. Detailed correlations between such myelin packing data and the sequence diVerences would require knowing the three dimensional structures of the various P0 extracellular domains from homology modeling or preferably from crystallographic analyses. In the foregoing genotype/phenotype comparisons, intact nerves from the diVerent species are easily accessible for diVraction analysis of native myelin membrane packing, and the P0 cDNA and protein sequences can be determined using standard procedures. There is a considerably greater challenge in relating MPZ mutations to myelin phenotypes in human diseases as diVraction analysis is impractical and electron microscopy of nerve biopsy is no longer routine for diagnostic purposes. Initial attempts to link myelin packing alterations with amino acid changes in human P0 were based on comparing the sites of these side chain in a molecular model of the extracellular domain of P0 with the structural data provided by electron microscopy of biopsy samples (Kirschner et al., 1996b). More recently, addition of the myc epitope tag to the mature amino terminus of P0 has been shown to produce a CMT1B-like neuropathy in transgenic mice. Ultrastructural analysis shows widening of the intraperiod line and myelin packing alteration in abnormal nerves, and immuno-electron microscopic analysis directly reveals that the mutant P0 was inserted in these abnormal myelin subregions (Previtali, et al., 2000). While some of the correlations are tantalizing in terms of explaining irregular or dys-compaction due to nonconservative substitutions resulting in altered interactions, this approach is of limited usefulness owing to the paucity of data. Currently, ~40 amino acids in the extracellular domain, four in the transmembrane domain, and seven in the cytoplasmic domain have been identiWed as sites of substitutions (Fig. 20.2). (Please refer to the Inherited Peripheral Neuropathies Database for the most up-to-date information: http://molgen-www.uia.ac.be/CMTMutations/.) The mutation sites are fairly well distributed within the domains. Taken together, the naturally occurring isomorphisms among species and the pathological forms in human P0 should provide a rich source of variants for testing hypotheses about the roles of speciWc amino acid side chains in P0 function and structure.

CONCLUSION AND FUTURE PERSPECTIVES Ample evidence demonstrates that P0 is a structural protein on which PNS myelin organizational integrity and function depend, and there is mounting evidence that P0 also has physiological roles. What are some of the research directions we can expect, and questions to be addressed in the coming years? Related to its structural role in myelin membrane adhesion, what is the mechanism whereby P0 homophilic interactions are initiated when the plasma membranes of the Schwann cell begin to elaborate? How does the end-to-end interaction of apposed P0 extracellular domains segue to lateral interactions? How do the speciWc amino acid substitutions in P0 result in the altered compaction evident from the observed pathology? What roles can be ascribed to P0’s carbohydrate moiety and its considerable heterogeneity (Gallego et al., 2001), to the palmitoylation of Cys153, and to the phosphorylation of tyrosine (Iyer et al., 2000; Xu et al., 2000)? Consideration of these and related questions will depend crucially on developing a model system in which alterations in P0 (e.g., deletions, substitutions, post-translational modiWcations) can be engineered and introduced transgenically into an animal, and in which ramiWcations across the hierarchy of myelin development, nerve function, and myelin membrane structure and interactions can be determined. While use of in vitro systems has provided some clues, and considerable progress has been made using transgenic mice, it can be anticipated that exploitation of the zebraWsh vertebrate model for myelination (Bro¨samle and Halpern, 2002; Schweitzer et al., 2003) may also lead to additional rapid progress. Beyond structure, we can anticipate clarifying the possible role of P0 in cell-cell adhesion and signal transduction in the PNS, where it is expressed well before myelination (Hagedorn et al., 1999), and possibly also in the CNS, where P0 mRNA or protein are detected in regions outside the peripheral nervous system during early development of the

CONCLUSION AND FUTURE PERSPECTIVES

FIGURE 20.4 Molecular model of P0 extracellular domain (PDB #1NEU; Shapiro et al., 1996) showing sites of amino acid substitutions that Wgure in human peripheral demyelinating neuropathies. The N- and C-termini are indicated, as is the apical Trp28. The domain is oriented with its C-terminus, which is lacking residues 120–124 (dotted), directed down toward the membrane surface. Residues 103–106 (part of the F-G loop) are disordered so their backbone trace is omitted (dotted lines). The disulfide is between the two b-sheets of the Ig fold (arrowhead). Color key for the original residues that are sites of substitutions: yellow – G, A, V, L, I, M, P, F, W; brown—S, T, N, Q, C, Y; purple—K, R, H, D, E; green, Asn93 (site of glycosylation). The Wgure was prepared using RASMOL (Sayle and Milner-White, 1995).

rat (Lee et al., 2001) and in spinal cord of young rats (Sato et al., 1999). Demonstration of P0 as a receptor for Theiler’s murine encephalomyelitis virus (TMEV; Libbey et al., 2001) may further substantiate the idea that P0 has additional ‘‘cytogogic’’ or nonstructural roles. The evidence that P0 might serve as an autoantigen in certain autoimmune diseases—for example, chronic inXammatory demyelinating polyradiculoneuropathy (Yan et al., 2001) and auditory dysfunction (Matsuoka et al., 1999; Boulassel et al., 2001)—is controversial in point of fact, but highlights that our notion of P0 has evolved from that of a simple structural protein of myelin to that of a protein having a more complex, integral role in the nervous system.

Acknowledgments D. A. K. is grateful to his colleague Dr. Hideyo Inouye for numerous insightful discussions and collaboration over the past 20 years, and to Dr. Deepak Sharma and Mr. Xiao (Tony)

539

540

20. THE P0 GENE

Luo for assistance with some of the P0 sequence data analyses. D. A. K. acknowledges research funding from NIH (NINDS #NS39650), institutional support from Boston College, and a Fulbright Senior Research Scholar Award from the Binational U.S.-Italian Fulbright Scholar Program for his sabbatical visit to the Feltri-Wrabetz lab in Milano. L. W. and M. L. F. acknowledge research support from the NIH (NS41319 and NS45630); Telethon, Italy; Great Britain Multiple Sclerosis Society; and the European Community.

References Akagi, K., Sandig, V., Vooijs, M., Van der Valk, M., Giovannini, M., Strauss, M., and Berns, A. (1997). Cremediated somatic site-speciWc recombination in mice. Nucleic Acids Res. 25, 1766–1773. Awatramani, R., Shumas, S., Kamholz, J., and Scherer, S. S. (2002). TGFbeta1 modulates the phenotype of Schwann cells at the transcriptional level. Mol. Cell Neurosci. 19, 307–319. Barbu, M. (1990). Molecular cloning of cDNAs that encode the chicken P0 protein: Evidence for early expression in avians. J. Neurosci. Res. 25, 143–151. Baron, P., Shy, M., Honda, H., Sessa, M., Kamholz, J., and Pleasure, D. (1994). Developmental expression of Po mRNA and Po protein in the sciatic nerve and the spinal nerve roots of the rat. J. Neurocytol. 23, 249–257. Bermingham, J. R., Jr., Scherer, S. S., O’Connell, S., Arroyo, E., Kalla, K. A., Powell, F. L., and Rosenfeld, M. G. (1996). Tst-1/Oct-6/SCIP regulates a unique step in peripheral myelination and is required for normal respiration. Genes Dev. 10, 1751–1762. Besancon, R., Prost, A. L., Konecny, L., Latour, P., Petiot, P., Boutrand, L., Kopp, N., Mularoni, A., Chamba, G., Vandenberghe, A. (1999). Alternative exon 3 splicing of the human major protein zero gene in white blood cells and peripheral nerve tissue. FEBS Lett. 457, 339–342. Bharucha, V. A., Peden, K. W., and Tennekoon, G. I. (1994). SV40 large T antigen with c-Jun down-regulates myelin P0 gene expression: A mechanism for papovaviral T antigen-mediated demyelination. Neuron, 12, 627–637. Bhattacharyya, A., Frank, E., Ratner, N., and Brackenbury, R. (1991). P0 is an early marker of the Schwann cell lineage in chicken. Neuron 7, 831–844. Bienfait, H. M. E., Baas, F., Gabree¨ls-Festen, A. A. W. M., Koelman, J. H. T. M., Langerhorst, C. T., De Visser, M. (2002). Two amino-acid substitutions in the myelin protien zero gene of a case of Charcot-Marie-Tooth disease associated with light-near dissociation. Neuromusc. Disord. 12, 281–285. Billings-Gagliardi, S., Kerner, A.-L., Kirschner, D. A., and Wolf, M. K. (1987). Shiverer*Jimpy double mutant mice. III. Comparison of shimld*jpmsd and shi*jp phenotypes demonstrates dissimilar interactions of allelic mutations. Mol. Brain Res. 2, 199–214. Bizzozero, O. A., Fridal, K., and Pastuszyn, A. (1994). IdentiWcation of the palmitoylation site in rat myelin P0 glycoprotein. J. Neurochem. 62, 1163–1171. Blaurock, A. E. (1971). Structure of the nerve myelin membrane: Proof of the low-resolution proWle. J. Mol. Biol. 56, 35–52. Blaurock, A. E. (1986). X-ray and neutron diVraction by membranes: How great is the potential for deWning the moleuclear interactions? In ‘‘Progress in Protein-Lipid Interactions’’ (A. Watts and J. J. H. H. M. DePont, eds.), Vol. 2, pp. 1. Elsevier, Amsterdam, New York. Bone, L. J., Balice-Gordon, R. J., Abel, A., Messing, A., Fischbeck, K., and Scherer, S. S. (1997). Transgenic mice expressing a connexin 32 mutation develop demyelinating peripheral neuropathy. Soc. Neurosci. Abstr. 23, 586. Boulassel, M. R., Guerit, J. M., Denison, S., de TourtchaninoV, M., Wenderickx, L., Botterman, N., Deggouj, N., GersdorV, M., Tomasi, J. P. (2001). No evidence of auditory dysfunction in guinea pigs immunized with myelin P0 protein. Hearing Res. 152, 10–16. Britsch, S., Goerich, D. E., Riethmacher, D., Peirano, R. I., Rossner, M., Nave, K. A., Birchmeier, C., and Wegner, M. (2001). The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15, 66–78. Bro¨samle, C., and Halpern, M. E. (2002). Characterization of myelination in the developing zebraWsh. Glia 39, 47–57. Brown, A. M., and Lemke, G. (1997). Multiple regulatory elements control transcription of the peripheral myelin protein zero gene. J. Biol. Chem. 272, 28939–28947. Brunden, K. R., and Poduslo, J. F. (1987). A phorbol ester-sensitive kinase catalyzes the phosphorylation of P0 glycoprotein in myelin. J. Neurochem. 49, 1863–1872. Carenini, S., Montag, D., Schachner, M., and Martini, R. (1999). Subtle roles of neural cell adhesion molecule and myelin-associated glycoprotein during Schwann cell spiralling in P0-deWcient mice. Glia 27, 203–212. Cheng, L., and Mudge, A. W. (1996). Cultured Schwann cells constitutively express the myelin protein P0. Neuron 16, 309–319. Combet, C., Blanchet, C., Geourjon, C., and Deleage, G. (2000). NPS@: Network protein sequence analysis. Trends Biochem. Sci. 25, 147–150. Corpet, F. (1988). Multiple sequence alignment with hierarchical clustering. Nucl. Acids Res. 16, 10881–10890.

CONCLUSION AND FUTURE PERSPECTIVES

Cosgaya, J. M., Chan, J. R., and Shooter, E. M. (2002). The neurotrophin receptor p75NTR as a positive modulator of myelination. Science 298, 1245–1248. DeJonghe, P., Timmerman, V., Ceuterick, C., Nelis, E., DeVriendt, E., Lofgren, A., Vercruyssen, A., Verellen, C., VanMaldergem, L., Martin, J-J., and VanBroeckhoven, C. (1999). The Thr124Met mutation in the peripheral myelin protein zero (MPZ) gene is associated with a clinically distinct Charcot-Marie-Tooth phenotype. Brain. 122, 281–290. Ding, Y., and Brunden, K. R. (1994). The cytoplasmic domain of myelin glycoprotein P0 interacts with negatively charged phospholipid bilayers. J. Biol. Chem. 269, 10764–10770. Doyle, J. P., Stempak, J. G., Cowin, P., Colman, D. R., and D’Urso, D. (1995). Protein zero, a nervous system adhesion molecule, triggers epithelial reversion in host carcinoma cells. J. Cell Biol. 131, 465–482. D’Urso, D., Brophy, P. J., Staugaitus, S. M., Gillespie, C. S., Frey, A. B., Stempack, J. G., and Colman, D. R. (1990). Protein zero of peripheral nerve myelin: Biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron 2, 449–460. Eichberg, J. (2002). Myelin P0: New knowledge and new roles. Nerochem. Res. 27, 1331–1340. Eichberg, J., and Iyer, S. (1996). Phosphorylation of myelin protein: Recent advances. Neurochem. Res. 21, 527–35. Everly, J. L, Brady, R. O., and Quarles, R. H. (1973). Evidence that the major protein in rat sciatic nerve myelin is a glycoprotein. J. Neurochem. 21, 329–334. Fabrizi, G. M., Ferrarini, M., Cavallaro, T., Jarre, L., Polo, A., and Rizzuto, N. (2001). A somatic and germline mosaic mutation in MPZ/P0 mimics recessive inheritance of CMT1B. Neurol. 57, 101–105. Fabrizi, G. M., Taioli, F., Cavallaro, T., Rigatelli, F, Simonati, A., Mariani, G., Perrone, P., and Rizzuto, N. (2000). Focally folded myelin in Charcot-Marie-Tooth neuropathy type 1B with Ser49Leu in the myelin protein zero. Acta Neuropathol. 100, 299–304. Feltri, M. L., D’Antonio, M., Previtali, S., Fasolini, M., Messing, A., and Wrabetz, L. (1999a). P0-Cre transgenic mice for inactivation of adhesion molecules in Schwann cells. Ann. N. Y. Acad. Sci. 883, 116–123. Feltri, M. L., D’Antonio, M., Quattrini, A., Numerato, R., Arona, M., Previtali, S., Chiu, S. Y., Messing, A., and Wrabetz, L. (1999b). A novel P0 glycoprotein transgene activates expression of lacZ in myelin-forming Schwann cells. Eur. J. Neurosci. 11, 1577–1586. Feltri, M. L., Graus Porta, D., Previtali, S. C., Nodari, A., Migliavacca, B., Cassetti, A., Littlewood-Evans, A., Reichardt, L. F., Messing, A., Quattrini, A., Mueller, U., and Wrabetz, L. (2002). Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with axons. J. Cell Biol. 156, 199–209. Fernandez-Valle, C., Fregien, N., Wood, P. M., and Bunge, M. B. (1993). Expression of the protein zero myelin gene in axon-related Schwann cells is linked to basal lamina formation. Development 119, 867–880. Fernandez-Valle, C., Gorman, D., Gomez, A. M., and Bunge, M. B. (1997). Actin plays a role in both changes in cell shape and gene-expression associated with Schwann cell myelination. J. Neurosci. 17, 241–250. Filbin, M. T., and Tennekoon, G. I. (1991). The role of complex carbohydrates in adhesion of the myelin protein, P0. Neuron 7, 845–855. Filbin, M. T., and Tennekoon, G. I. (1993). Homophilic adhesion of the myelin P0 protein requires glycosylation of both molecules in the homophilic pair. J. Cell Biol. 122, 451–459. Filbin, M. T., Walsh, F. S., Trapp, B. D., Pizzey, J. A., and Tennekoon, G. I. (1990). Role of myelin P0 protein as a homophilic adhesion molecule. Nature 344, 871–872. Filbin, M. T., Wong, M. H., Zhang, K., and Li, W. H. (1997). The role of myelin P0 protein. In ‘‘Cell Biology and Pathology of Myelin’’ (B. H. J. Juurlink, R. M. Devon, J. R. Doucette, A. J. Nazarali, D. J. Schreyer, and V. M. K. Verge, eds.), pp. 103–114. Plenum Press, New York. Franklin, R. J., and Barnett, S. C. (2000). Olfactory ensheathing cells and CNS regeneration: The sweet smell of success? Neuron 28, 15–8. Franks, N. P., and Lieb, W. R. (1980). Rapid movement of molecules across membranes: Measuerment of the permeability coeYcient of water using neutron diVraction. J. Mol. Biol. 141, 43–61. Frei, R., Motzing, S., Kinkelin, I., Schachner, M., Koltzenburg, M., and Martini, R. (1999). Loss of distal axons and sensory Merkel cells and features indicative of muscle denervation in hindlimbs of P0-deWcient mice. J. Neurosci. 19, 6058–6067. Gallego, R. G., Blanco, J. L. J., Thijssen-van Zuylen, C. W. E. M., Gotfredsen C. H., Voshol, H., Duus, J. Ø., Schachner, M., and Vliegenthart, J. F. G. (2001). Epitope diversity of N-glycans from bovine peripheral myelin glycoprotein P0 revealed by mass spectrometry and nano probe magic angle spinning 1H NMR spectroscopy. J. Biol. Chem. 276, 30834–30844. Gao, Y., Li, W., and Filbin, M. T. (2000). Acylation of myelin P0 protein is required for adhesion. J. Neurosci. Res. 60, 704–713. Giese, K. P., Martini, R., Lemke, G., Soriano, P., and Schachner, M. (1992). Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71, 565–576. Giovannini, M., Robanus-Maandag, E., Niwa-Kawakita, M., van der Valk, M., WoodruV, J. M., Goutebroze, L., Merel, P., Berns, A., Thomas, G. (1999). Schwann cell hyperplasia and tumors in transgenic mice expressing a naturally occurring mutant NF2 protein. Genes Dev. 13, 978–86. Gould, R. M., Jessen, K. R., Mirsky, R., Tennekoon, G. (1992). The cell of Schwann: An update. In ‘‘Myelin: Biology and Chemistry’’ (R. E. Martenson, Ed.), pp. 123–171. CRC Press, Boca Raton, FL. GreenWeld, S., BrostoV, S., Eylar, E. H., and Morell, P. (1973). Protein composition of myelin of the peripheral nervous system. J. Neurochem. 20, 1207–1216.

541

542

20. THE P0 GENE

GriYth, L. S., Schmitz, B., (1992). L2/HNK-1 carbohydrate and protein-protein interactions mediate the homophilic binding of the neural adhesion molecule P0. J. Neurosci. Res. 33, 639–648. Gupta, S. K., Poduslo, J. F., and Mezei, C. (1988). Temporal changes in P0 and MBP gene expression after crushinjury of the adult peripheral nerve. Mol. Brain Res. 4, 133–141. Guttinger, M., Sutti, F., Panigada, M., Porcellini, S,. Merati, B., Mariani, M., Teesalu, T., Consalez, G. G., Grassi, F. (1998). Epithelial V-like antigen (EVA), a novel member of the immunoglobulin superfamily, expressed in embryonic epithelia with a potential role as homotypic adhesion molecule in thymus histogenesis. J. Cell Biol. 141, 1061–1071. Hagedorn, L., Suter, U., and Sommer, L. (1999). P0 and PMP22 mark a multipotent neural crest-derived cell type that displays community eVects in response to TGF-beta family factors. Development 126, 3781–3794. Hayasaka, K., Himoro, M., Wang, Y., Takata, M., Minoshima, S., Shimizu, HN., Miura, M., Uyemura, K., and Takada, G. (1993). Structure and chromosomal localization of the gene encoding the human myelin protein zero (MPZ) Genomics 17, 755–758. Hayasaka, K., Nanao,K., Tahara,M., Sato,W., Takada,G., Miura,M., and Uyemura,K. (1991). Isolation and sequence determination of cDNA encoding the major structural protein of human peripheral myelin. Biochem. Biophys. Res Commun. 180, 515–518. Hilmi, S., Fournier, M., Valeins, H., Gandar, G. C., and Bonnet, J. (1995). Myelin P0 glycoprotein: IdentiWcation of the site phosphorylated in vitro and in vivo by endogenous protein kinases. J. Neurochem. 64, 902–907. Inouye, H., Karthigasan, J., and Kirschner, D. A. (1989). Membrane structure in isolated and intact myelins. Biophys. J. 56, 129–137. Inouye, H., and Kirschner, D. A. (1988). Membrane interactions in nerve myelin: I. Determination of surface charge from eVects of pH and ionic strength on period. Biophys. J. 53, 235–246. Inouye, H., and Kirschner, D. A. (1991). Folding and function of the myelin proteins from primary sequence data. J. Neurosci. Res. 28, 1–17. Inouye, H., Tsuruta, H., Sedzik, J., Uyemura, K., and Kirschner, D. A. (1999). Tetrameric assembly of full sequence P0 myelin glycoprotein by synchrotron x-ray scattering. Biophys. J. 76, 423–437. Iyer, S., Bianchi, R., and Eichberg, J. (2000). Tyrosine phosphorylation of PNS myelin P0 occurs in the cytoplasmic domain and is maximal during early development. J. Neurochem. 75, 347–354. Iyer, S., Rowe-Rendleman, C. L., Bianchi, R., and Eichberg, J. (1996). Tyrosine phosphorylation of myelin protein P0. J. Neurosci. Res. 46, 531–539. Jaegle, M., Mandemakers, W., Broos, L., Zwart, R., Karis, A., Visser, P., Grosveld, F., and Meijer, D. (1996). The POU factor Oct-6 and Schwann cell diVerentiation. Science 273, 507–510. Jeserich, G., and Waehneldt, T. V. (1986). Bony Wsh myelin: Evidence for common major structural glycoproteins in central and peripheral myelin of trout. J. Neurochem. 46, 525–533. Kirschner, D. A., and Blaurock, A. E. (1992). Organization, phylogenetic variations and dynamic transitions of myelin structure. In ‘‘Myelin: Biology and Chemistry’’ (R. E. Martenson, ed.), pp. 3–78. CRC Press, Boca Raton, FL. Kirschner, D. A., Caspar, D. L. D., Schoenborn, B. P., and Nunes, A. C. (1975). Neutron diVraction of nerve myelin. In ‘‘Neutron Scattering for the Analysis of Biological Structure’’ (B. P. Schoenborn, ed.), Brookhaven Symposium in Biology, No. 27, III, 68–76. Brookhaven National Laboratory, Upton, NY. Kirschner, D. A., and Ganser, A. L. (1980). Compact myelin exists in the absence of basic protein in the shiverer mutant mouse. Nature 283, 207–210. Kirschner, D. A., Ganser, A. L., and Caspar, D. L. D. (1984). DiVraction studies of molecular organization and interactions in myelin membranes. In ‘‘Myelin’’ (P. Morell, ed.), 2nd ed., pp. 51–95. Plenum Press, New York. Kirschner, D. A., and Hollingshead, C. J. (1980). Processing for electron microscopy alters membrane structure and packing in myelin. J. Ultrastruc. Res. 73, 211–232. Kirschner, D. A., Inouye, H., Ganser, A. L., Mann, V. (1989). Myelin membrane structure and composition correlated: A phylogenetic study. J. Neurochem. 53, 1599–1609. Kirschner, D. A., Inouye, H., and Saavedra, R. A. (1996a). Membrane adhesion in peripheral myelin: Good and bad wraps with protein P0. Structure 4, 1239–1244. Kirschner, D. A., Szumowski, K., Gabree¨ls-Festen, A. A. W. M., Hoogendijk, J. E., and Bolhuis, P. A. (1996b). Inherited demyelinating peripheral neuropathies: Relating myelin packing abnormalities to P0 molecular defects. J. Neurosci. Res. 46, 502–508. Koenig, H. L., Schumacher, M., Ferzaz, B., Thi, A. N., Ressouches, A., Guennoun, R., Jung-Testas, I., Robel, P., Akwa, Y., and Baulieu, E. E. (1995). Progesterone synthesis and myelin formation by Schwann cells. Science 268, 1500–1503. Kuhn, R., Pravtcheva, D., Ruddle, F., and Lemke, G. (1990). The gene encoding peripheral myelin protein zero is located on mouse chromosome 1. J. Neurosci. 10, 205–209. Lagueny, A., Latour, P., Vital, A. Le Masson G., Rouanet, M., Ferrer, X., Vital, C., and Vandenberghe, A. (2001). Mild recurrent neuropathy in CMT1B with a novel nonsense mutation in the extracellular domain of the MPZ gene. J. Neurol Neurosurg. Psychiatry 70, 232–235. Landon, D. N., and Hall, S. (1976). The myelinated nerve Wber. In ‘‘The Peripheral Nerve’’ (D. N. Landon, ed.), pp. 1–105. Chapman and Hall, London. Lanwert, C., and Jeserich, G. (2001). Structure, heterologous expression, and adhesive properties of the P(0)-like myelin glycoprotein IP1 of trout CNS. Micros. Res. Tech. 52, 637–644. Lee, M., Brennan, A., Blanchard, A., Zoidl, G., Dong, Z., Tabernero, A., Zoidl, C., Dent, M. A., Jessen, K. R., and Mirsky, R. (1997). P0 is constitutively expressed in the rat neural crest and embryonic nerves and is

CONCLUSION AND FUTURE PERSPECTIVES

negatively and positively regulated by axons to generate non-myelin-forming and myelin-forming Schwann cells, respectively. Mol. Cell Neurosci. 8, 336–350. Lee, M-J., Calle, E., Brennan, A., Ahmed, S., Sviderskaya, E., Jessen, K. R., and Mirsky, R. (2001). In early development of the rat mRNA for the major myelin protein P0 is expressed in nonsensory areas of the embryonic inner ear, notochord, enteric nervous system, and olfactory ensheathing cells. Develop. Dynamics 222, 40–51. Lemke, G., and Axel, R. (1985). Isolation and sequence of a cDNA encoding the major structural protein of peripheral myelin. Cell 40, 501–508. Lemke, G., and Chao, M. (1988). Axons regulate Schwann cell expression of the major myelin and NGF receptor genes. Develop. 102, 499–504. Lemke, G., Lamar, E., and Patterson, J. (1988). Isolation and analysis of the gene encoding peripheral myelin protein zero. Neuron 1, 73–83. Libbey, J. E., McCright Ingeborg, J., Tsunoda, I., Wada, Y., and Fujinami, R. S. (2001). Perpheral nerve protein, P0, as a otential receptor for Theiler’s muroine encephalomyelitis virus. J. Neurovirol. 7, 97–104. Liehr, T., Ekici, A., Hillenbrand, R., and Rautenstrauss, B. (1995). Regional localization of rat myelin protein zero Mpz gene to chromosome 13q24–25 by means of FISH. Mamm. Genome. 6, 824–825. Marrosu, M. G., Vaccargiu, S., Marrosu, G., Vannelli, A., Cianchetti, C., and Muntoni, F. (1998). CharcotMarie-Tooth disease type 2 associated with mutation of the myelin protein zero gene. Neurology 50, 1397–1401. Martini, R., Mohajeri, M. H., Kasper, S., Giese, K. P., and Schachner, M. (1995b). Mice double deWcient in the genes for P0 and myelin basic protein show that both proteins contribute to the formation of the major dense line in peripheral nerve myelin. J. Neurosci. 15, 4488–4495. Martini, R., Zielasek, J., Toyka, K. V., Giese, KP., and Schachner, M. (1995a). Protein zero (P0)-deWcient mice show myelin degeneration in peripheral nerves characteristic of inherited human neuropathies. Nat. Genet. 11, 281–286. Matsuoka, H., Cheng, K-C., Krug, M. S., Yazawa, Y., and Yoo, T-J. (1999). Murine model of autoimmune hearing loss induced by myelin protein P0. Ann. Otol. Rhinol. Laryngol. 108, 255–264. Maycox, P., Ortuno, D., Burrola, P., Kuhn, R., Bieri, P., Arrezo, J., and Lemke, G. (1997). A transgenic mouse model for human hereditary neuropathy with liability to pressure palsies. Mol. Cell. Neurosci. 8, 405–416. Melcangi, R. C., Magnaghi, V., Cavarretta, I., Martini, L., and Piva, F. (1998). Age-induced decrease of glycoprotein Po and myelin basic protein gene expression in the rat sciatic nerve. Repair by steroid derivatives. Neuroscience 85, 569–578. Menichella, D. M., Arroyo, E. J., Awatramani, R., Xu, T., Baron, P., Vallat, J-M., Balsamo, J., Lilien, J, Scarlato, G., Kamholz, J., Scherer, S. S., and Shy, M. E. (2001). Protein zero is necessary for E-cadherin-mediated adherens junction formation in Schwann cells. Mol. Cell Neurosci. 18, 606–618. Messing, A., Behringer, R. R., Hammang, J. P., Palmiter, R. D., Brinster, R. L., and Lemke, G. (1992). P0 promoter directs expression of reporter and toxin genes to Schwann cells of transgenic mice. Neuron 8, 507–520. Messing, A., Behringer, R. R., Wrabetz, L., Hammang, J. P., Lemke, G., Palmiter, R. D., and Brinster, R. L. (1994). Hypomyelinating peripheral neuropathies and schwannomas in transgenic mice expressing SV40 T-antigen. J. Neurosci. 14, 3533–3539. Mews, M., and Meyer, M. (1993). Modulation of Schwann cell phenotype by TGF-beta 1: Inhibition of P0 mRNA expression and downregulation of the low aYnity NGF receptor. Glia 8, 208–217. Misu, K., Yoshihara, T., Shikama, T., Awaki, E., Yamamoto, M., Hattori, N., Hirayama, M., Takegami, T., Nakashima, K., and Sobue, G. (2000). An axonal form of Charcot-Marie-Tooth disease showing distinctive features in asociation with mutations in the peripheral myelin protein zero gene (Thr124Met or Asp75Val). J. Neurol. Neurosurg. Psychiatry 69, 806–811. Monuki, E. S., Weinmaster, G., Kuhn, R., and Lemke, G. (1989). SCIP: A glial POU domain gene regulated by cyclic AMP. Neuron 3, 783–793. Morgan, L., Jessen, K. R., and Mirsky, R. (1991). The eVects of cAMP on diVerentiation of cultured Schwann cells: Progression from an early phenotype (04þ) to a myelin phenotype (P0þ, GFAP-, N-CAM-, NGFreceptor-) depends on growth inhibition. J. Cell Biol. 112, 457–467. Morgan, L., Jessen, K. R., and Mirsky, R. (1994). Negative regulation of the P0 gene in Schwann cells: Suppression of P0 mRNA and protein induction in cultured Schwann cells by FGF2 and TGF beta 1, TGF beta 2 and TGF beta 3. Development 120, 1399–1409. Nagarajan, R., Svaren, J., Le, N., Araki, T., Watson, M., and Milbrandt, J. (2001). EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30, 355–368. Nelis, E., Haites, N., and Van Broeckhoven, C. (1999). Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum. Mutat. 13, 11–28. Nikitin, A. Y., Jin, J. J., Papewalis, J., Prokopenko, S. N., Pozharisski, K. M., Winterhager, E., Flesken-Nikitin, A., and Rajewsky, M. F. (1996). Wild type neu transgene counteracts mutant homologue in malignant transformation of rat Schwann cells. Oncogene 12, 1309–1317. Numakura, C., Ln, C., Ikegami, T., Guldberg, P., and Hayasaka, K. (2002). Molecular analysis in Janpanese patients with Charcot-Marie-Tooth: DGGE analysis for PMP22, MPZ, and Cx32/GJB1 mutations. Hum. Mutat. 20, 392–398. Owens, G. C., and Boyd, C. J. (1991). Expressing antisense P0 RNA in Schwann cells perturbs myelination. Development 112, 639–649.

543

544

20. THE P0 GENE

Peirano, R. I., Goerich, D. E., Riethmacher, D., and Wegner, M. (2000). Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol. Cell Biol. 20, 3198–3209. Peters, A., Palay, S. L., and Webster, H. deF. (1991). ‘‘The Fine Structure of the Nervous System: Neurons and Their Supporting Cells.’’ 3rd ed., p. 231. Oxford University Press, New York. Pham-Dinh, D., Fourbil, Y., Blanquet, F., Mattei, M-G., Roeckel, N., Latour, P., Chazot, G., Vandenberghe, A., and Dautigny, A. (1993). The major peripheral myelin protein zero gene: Structuree and localization in the cluster of Fcg receptor genes on human chromosome 1q21.3-q.23. Hum. Molec. Genet. 2, 2051–2054. Plante´-Bordeneuve, V., Guiochon-Mantel, AS., Lacroix, C., Lapresle, J., and Said, G. (1999). The Roussy-Le´vy family: From the original description to the gene. Ann. Neurol. 46, 770–773. Plante´-Bordeneuve, V., Parman, Y., Guiochon-Mantel, A., Alj, Y., Deymeer, F., Serdaroglu, P., Eraksoy, M., and Said, G. (2001). The range of chronic demyelinating neuropathy of infancy: A clinico-pathological and genetic study of 15 unrelated cases. J. Neurol. 248, 795–803. Poduslo, J. F. (1990). Golgi sulfation of the oligosaccharide chain of P0 occurs in the presence of myelin assembly but not in its absence. J. Biol. Chem. 265, 3719–3725. Poduslo, J. F., Dyck, P. J., and Berg, C. T. (1985). Regulation of myelination: Schwann cell transition from a myelin-maintaining state to a quiescent state after permanent nerve transection. J. Neurochem. 44, 388–400. Previtali, S. C., Quattrini, A., Fasolini, M., Panzeri, M. C., Villa, A., Filbin, M. T., Li, W., Chiu, S. Y., Messing, A., Wrabetz, L., and Feltri, M. L. (2000). Epitope-tagged P(0) glycoprotein causes Charcot-Marie-Tooth-like neuropathy in transgenic mice. J. Cell Biol. 151, 1035–1046. Quarles, R. H. (1997). Glycoproteins of myelin sheaths. J. Moec. Neurosci. 7, 1–12. Saavedra, R. A., Fors,L., Aebersold, R. H., Arden, B., Horvath, S., Sanders, J., and Hood, L. (1989). The myelin proteins of the shark brain are similar to the myelin proteins of the mammalian peripheral nervous system. J. Mol. Evol. 29, 149–156. Sakamoto, Y., K. Kitamura, Yoshimura,K., Nishijima,T., and Uyemura, K. (1987). Complete amino acid sequence of P0 protein in bovine peripheral nerve myelin. J. Biol. Chem. 262, 4208–4214. Sato, Y., Kimura, M., Yasuda, Ch., Nakano, Y., Tomita, M., Kobata, A., and Endo, T. (1999). Evidence for the presence of major peripheral myelin glycoprotein P0 in mammalian spinal cord and a change of its glycosylation state during aging. Glycobiol. 9, 655–660. Sayle, R., and Milner-White, E. J. (1995). RasMol: Biomolecular graphics for all. Trends Biochem. Sci. 20, 374. Scherer, S. S., Wang, D. Y., Kuhn, R., Lemke, G., Wrabetz, L., and Kamholz, J. (1994). Axons regulate Schwann cell expression of the POU transcription factor SCIP. J. Neurosci. 14, 1930–1942. Schneider-Schaulies, J., von Brunn, A., and Schachner, M. (1990). Recombinant peripheral myelin protein P0 confers both adhesion and neurite outgrowth-promoting properties. J. Neurosci. Res. 27, 286–297. Schweitzer, J., Becker, T., Becker, C. G., Schachner, M. (2003). Expression of protein zero is increased in lesioned axon pathways in the central nervous system of adult zebraWsh. Glia 41, 301–317. Senderek, J., Hermanns, B., Lehmann, U., Bergmann, C., Marx, G., Kabus, C., Timmerman, V., StoltenburgDidinger, G., and Schroeder, M. J. (2000). Charcot-Marie-Tooth neuropathy type 2 and P0 point mutations: Two novel amino acid substitutions (Asp61Gly; Tyr119Cys) and a possible ‘‘hotspot’’ on Thr124Met. Brain Pathol. 10, 235–248. Senderek, J., Ramaekers, V. T., Zerres, K., Rudnik-Schoeneborn, S., Schroeder, J. M., and Bergmann, C. (2001). Phenotypic variation of a novel nonsense mutation in the P0 intracellular domain. J. Neurol. Sci. 192, 49–51. Shapiro, L., Doyle, J. P., Hensley, P., Colman, D. R., and Hendrickson, W. A. (1996). Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17, 435–449. Silander, K., P. Meretoja, Nelis, E., Timmerman, V., Van Broeckhoven, C., Aula, P., and Savontaus, M-L. (1996). A de novo duplication in 17p11.2 and a novel mutation in the P0 gene in two De´je´rine-Sottas syndrome patients. Hum. Mutat. 8, 304–310. Sindou, P., Vallat, J-M., Chapon, F., Archelos, J-J., Tabaraud, F., Anani, T., Braund, K. G., Maisonobe, T., Hauw, J-J., and Vandenberghe, A. (1999). Ultrastructural protein zero expression in Charcot-Marie-Tooth type 1B disease. Muscle Nerve 22, 99–104. Spiryda, L. (1998). Myelin protein zero and membrane adhesion. J. Neurosci. Res. 54, 137–146. Spiryda, L. B., and Colman, D. R. (1998a). Protein zero, a myelin IgCAM, induces physiologically operative tight junctions in nonadhesive carcinoma cells. J. Neurosci. Res. 54, 282–8. Spiryda, L. B., and Colman, D. R. (1998b). Suppression of tumorigenicity in an aggressive cervical carcinoma induced by protein zero, a nervous system IgCAM. J. Cell Sci. 111, 3253–60. Stahl, N., Harry, J., and Popko, B. (1990). Quantitative analysis of myelin protein gene expression during development in the rat sciatic nerve. Mol. Brain Res. 8, 209–212. Stratmann, A., and Jeserich, G. (1995). Molecular cloning and tissue expression of a cDNA encoding IP1 – a P0like glycoprotein of trout CNS myelin. J. Neurochem. 64, 2427–2436. Suzuki, M., Sakamoto, Y., Kitamura, K., Fukunaga, K., Yamamoto, H., Miyamoto, E., and Uyemura, K. (1990). Phosphorylation of P0 glycoprotein in peripheral nerve myelin. J. Neurochem. 55, 1966–1971. Tang, D. S., Yu, K. P., Tang, X., Zhang, H., Pan, Q., Dai, H., and Xia, J. (2000). In silicon [sic] cloning of human myelin protein zero-like genes by bioinformatics strategy. Amer. J. Hum. Genet. 67(Suppl.), 177. Thompson, A. J., Cronin, M. S., and Kirschner, D. A. (2002). Myelin protein zero (P0) exists as dimers and tetramers in native membranes of Xenopus laevis peripheral nerve. J. Neurosci. Res. 67, 766–771. Thompson, J. D., Higgins, D. G, Gibson, T. J. (1994). Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-speciWc gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673–4680.

CONCLUSION AND FUTURE PERSPECTIVES

Topilko, P., and Meijer, D. (2001). Transcription factors that control Schwann cell development and myelination. In ’’Glial Cell Development: Basic Principles and Clinical Relevance’’ (K. R. Jessen and W. R. Richardson, eds.), pp. 223–244. Oxford University Press, Oxford. Topilko, P., Schneider-Maunoury, S., Levi, G., Baron-Van Evercooren, A., ChennouW, A. B., Seitanidou, T., Babinet, C., and Charnay, P. (1994). Krox-20 controls myelination in the peripheral nervous system. Nature 371, 796–799. Trapp, B. D., Hauer, P., and Lemke, G. (1988). Axonal regulation of myelin protein mRNA levels in actively myelinating Schwann cells. J. Neurosci. 8, 3515–3521. Uyemura, K., Kitamure, K., and Miura, M. (1992). Structure and molecular biology of P0 protein. In ‘‘Myelin: Biology and Chemistry’’ (R. E. Martenson, ed.), pp. 481–508. CRC Press, Boca Raton, FL. Uyemura, K., Suzuki, M., Sakamoto, Y., and Tanaka, S. (1987). Structure of P0 protein: Homology to immunoglobulin superfamily. Biomed. Res. 8, 353–357. Voshol, H., van Zuylen, C. W. E. M., Orberger, G., Vliegenthart, J. F. G., and Schachner, M. (1996). Structure of the HNK-1 carbohydrate epitope on bovine peripheral myelin glycoprotein P0. J. Biol. Chem. 271, 22957–22960. Waehneldt, T. V., Matthieu, J-M., Jeserich, G. (1986). Appearance eof myelin proteins during vertebrate evolution. Neurochem. Int. 9, 463–474. Warner, L. E., Shohat, M., Shorer, Z, and Lupski, J. R. (1997). Multiple de novo MPZ (P0) point mutations in a sporadic Dejerine-Sottas case. Hum. Mutat. 10, 21–24. Weinstein, D. E., Burrola, P. G., and Lemke, G. (1995). Premature Schwann cell diVerentiation and hypermyelination in mice expressing a targeted antagonist of the POU transcription factor SCIP. Mol. Cell. Neurosci. 6, 212–229. Wells, C. A., Saavedra, R. A., Inouye, H., and Kirschner, D. A. (1993). Myelin P0-glycoprotein: Predicted structure and interactions of extracellular domain. J. Neurochem. 61, 1987–1995. Wong, M. H., and Filbin, M. T. (1994). The cytoplasmic domain of the myelin P0 protein inXuences the adhesive interactions of its extracellular domain. J. Cell Biol. 126, 1089–1097. Wong, M. H., and Filbin, M. T. (1996). Dominant-negative eVect on adhesion by myelin Po protein truncated in its cytoplasmic domain. J. Cell Biol. 134, 1531–1541. Wood, J. G., and Dawson, R. M. C. (1973). A major myelin glycoprotein of sciatic nerve. J. Neurochem. 21, 717–719. Wrabetz, L., Feltri, M., Quattrini, A., Inperiale, D., Previtali, S., D’Antonio, M., Martini, R., Yin, X., Trapp, B., Zhou, L., Chiu, S.-Y., and Messing, A. (2000). P0 overexpression causes congenital hypomyelination of peripheral nerve. J. Cell Biol. 148, 1021–1033. Xu, M-J., Zhao, R., Sui, X., Xu, F., and Zhao, Z. J. (2000). Tyrosine phosphyorylation of myelin P0 and its implication in signal transduction. Biochem. Biophys. Res. Commun. 267, 820–825. Xu, W., Shy, M., Kamholz, J., Elferink, L., Xu, G., Lilien, J., and Balsamo, J. (2001). Mutations in the cytoplasmic domain of P0 reveal a role for PKC-mediated phosphorylation in adhesion and myelination. J. Cell Biol. 155, 439–446. Yamauchi, Y., Abe, K., Mantani, A., Hitoshi, Y., Suzuki, M., Osuzu, F., Kuratani, S., and Yamamura, K. (1999). A novel transgenic technique that allows speciWc marking of the neural crest cell lineage in mice. Dev. Biol. 212, 191–203. Yan W. X., Archelos, J. J., Hartung, H-P., and Pollard, J. D. (2001). P0 protein is a target antigen in chronic inXammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50, 282–292. Yin, A., Kidd, G., Wrabetz, L., Feltri, M., Messing, A., and Trapp, B. (2000). Schwann cell myelination requires timely and precise targeting of P0 protein. J. Cell Biol. 148, 1009–1020. You, K-H., Hsieh, C-L., Hayes, C., Stahl, N., Francke, U., and Popko, B. (1991). DNA sequence, genomic organization, and chromosomal localization of the mouse peripheral myelin protein zero gene: IdentiWcation of polymorphic alleles. Genomics 9, 751–757. Young, P., Grote, K., Kuhlenba¨umer, G., Debus, O., Kurlemann, H., Halfter, H., Funke, H., Ringelstein, E. B., and Sto¨gbauer, F. (2001). Mutation analysis in Charcot-Marie-Tooth disease type 1: Point mutations in the MPZ gene and GJB1 gene cause comparable phenotypic heterogeneity. J. Neurol. 248, 410–415. Zhang, K., and Filbin. M. T. (1994). Formation of a disulWde bond in the immunoglobulin domain of the myelin P0 protein is essential for its adhesion. J. Neurochem. 63, 367–370. Zhang, S. M., Marsh, R., Ratner, N., and Brackenbury, R. (1995). Myelin glycoprotein P0 is expressed at early stages of chicken and rat embryogenesis. J. Neurosci. Res. 40, 241–250. Zhao, Z. J., Zhao, R. (1998). PuriWcation and cloning of PZR, a binding protein and putative physiological substrate of tyrosine phosphatase SHP-2. J. Biol. Chem. 273, 29367–29372. Zorick, T. S., Syroid, D. E., Brown, A., Gridley, T., and Lemke, G. (1999). Krox-20 controls SCIP expression, cell cycle exit and susceptibility to apoptosis in developing myelinating Schwann cells. Development 126, 1397–1406.

545

C H A P T E R

21 PMP 22 Gene Ueli Suter

INTRODUCTION The Pperipheral myelin protein 22 (PMP22) is a minor component of the myelin sheath of peripheral nerves. It belongs to a family of membrane proteins that is characterized by four hydrophobic domains and conserved amino acid motifs (Jetten and Suter, 2000). These small vertebrate integral membrane glycoproteins are evolutionary related. The group includes epithelial membrane protein-1 (EMP-1, also known as CL20, TMP, B4B, or PAP); epithelial membrane protein-2 (EMP-2, also referred to as XMP); epithelial membrane protein-3 (EMP-3, also named YMP or HNMP-1), and, more distantly related, PERP, a downstream eVector of p53-dependent apoptosis (Attardi et al., 2000), and the eye lens speciWc membrane protein 20 (MP20 or MP19). The proteins of this family are about 160 to 180 amino acid residues in size and sequence comparisons suggest that the group may also include the claudins, components of tight junctions (www.sanger. ac.uk/cgi-bin/Pfam/getacc?PF00822). PMP22 can be regarded as the prototypic member of the family. It has been characterized most extensively and is the topic of this review. Besides the known function of the claudins (Tsukita et al., 2001), the precise roles of the other members of the family have not yet been Wrmly established. They are likely to have important functions in the regulation of cell proliferation, diVerentiation and apoptosis in various tissues (Attardi et al., 2000; Ben-Porath et al., 1999; Jetten and Suter, 2000; Wang et al., 2001; Wilson et al., 2002). PMP22 plays a crucial functional role in peripheral nerves based on the observation that genetic alterations in the PMP22 gene lead to various forms of myelination deWciencies in humans and rodents (Naef and Suter, 1998; Suter and Snipes, 1995). This chapter will review what is known about the structure, regulation and biological function of this protein in myelination, myelin maintenance, and disease.

CLONING AND EXPRESSION OF PMP22 Cloning of PMP22 cDNA has been Wrst described in the mouse as the result of a screening eVort aimed at the elucidation of the genetic program that regulates cellular growth arrest in Wbroblasts. PMP22 was isolated as a transcript that was strongly upregulated in quiescent NIH3T3 Wbroblasts and was called gas-3 for growth-arrest speciWc mRNA number 3 (Ciccarelli et al., 1990; ManWoletti et al., 1990; Schneider et al., 1988). In 1991, the rat PMP22 cDNA was cloned, initially termed SR13 or CD25, based on two other diVerential screenings of cDNA libraries generated from injured versus noninjured sciatic nerves (De Leon et al., 1991; Spreyer et al., 1991; Welcher et al., 1991). Welcher and colleagues (1991) also realized that the corresponding bovine PMP22 protein had been

Myelin Biology and Disorders, Volume 1

547

Copyright 2004, Elsevier (USA). All rights reserved.

548

21. PMP 22 GENE

described previously, together with Protein zero (P0) as one of two peripheral myelinspeciWc glycoproteins, PASII and PASI, respectively, that stained with periodate-SchiV ’s reagent (Kitamura et al., 1976). Subsequently, the human PMP22 cDNA was identiWed (Patel et al., 1992). More recently, the cloning of the cDNA of a PMP22 orthologue in the zebraWsh (Wulf et al., 1999) and a PMP22-related gene in C. elegans (Agostoni et al., 1999) have been reported. PMP22 is widely expressed in neural and non-neural tissues during embryonic development and in the adult (Baechner et al., 1995; De Leon et al., 1994; Kuhn et al., 1993; Lobsiger et al., 1996; Spreyer et al., 1991; Taylor et al., 1995; Welcher et al., 1991). In early embryonic PNS development of the rat, PMP22 is found in peripheral nerves and in dorsal root ganglia (DRG) from E12 onwards (Hagedorn et al., 1999). In contrast, expression in the early developing mouse DRG is very weak or absent (Paratore et al., 2002). Thus, PMP22 appears to be a valuable marker for PNS progenitors in rat embryos but not for early mouse neural crest derivatives. DRG sensory neurons and satellite cells continue to express PMP22 also into adulthood (De Leon et al., 1994). In zebraWsh, PMP22 expression is found in embryonic sclerotome cells, in neural crest cells, and in migratory derivatives of both populations (Wulf et al., 1999). PMP22 is most highly expressed by myelinating Schwann cells and is strongly upregulated in parallel with the initiation of myelination (Notterpek et al., 1999b; Snipes et al., 1992). Detailed immunohistochemical studies have localized PMP22 to the plasma membrane of nonmyelinating and myelinating Schwann cells as well as to the compact portion of myelin (Haney et al., 1996; Snipes et al., 1992). Following sciatic nerve injury, PMP22 expression is rapidly down-regulated in the degenerating nerve segments distal to the site of injury but recovers with nerve regeneration (De Leon et al., 1991; Kuhn et al., 1993; Snipes et al., 1992; Welcher et al., 1991). These data suggest that axons are required for high induction of PMP22 expression (Maier et al., 2002b). This hypothesis is supported by the Wnding that Schwann cells express low levels of PMP22 in the absence of neurons in culture. Only if myelin is formed, PMP22 expression increases strongly (Pareek et al., 1997). The majority of newly synthesized PMP22 in Schwann cells is rapidly degraded in the endoplasmic reticulum (Pareek et al., 1993). Only a minor portion of the synthesized PMP22 is complex glycosylated and accumulates in the Golgi apparatus. These proteins are translocated to the Schwann cell membrane in detectable amounts only when axonal contact and myelination occur. The rapid turnover of PMP22 in Schwann cells, however, is not altered by myelination. PMP22 is also found in the central nervous system (CNS) but at much lower levels than in peripheral nerves. In young mice, nuclei of the oculomotor and trochlear nerves express PMP22 but not the nucleus of the abducens nerve. PMP22 is also present in the motor nuclei of the trigeminal, facial, ambigus, vagus, hypoglossal and accessory spinal nerves, and motoneurons of the ventral horn of the spinal cord (De Leon et al., 1994; Parmantier et al., 1995). During mouse CNS development, PMP22 expression starts at E11.5 in restricted longitudinal and transverse domains, in the ventricular zone of the spinal cord, the rhombencephalon, mesencephalon and prosencephalon. PMP22 continues to be expressed in a patterned manner and diVerences in the level of PMP22 expression in p2, p3, and p4 morphologically deWne the p2/p3 and p3/p4 neuromeric boundaries (Parmantier et al., 1997). In the adult, PMP22 mRNA levels are approx. 10-fold higher in sciatic nerve compared to the lung and intestine and about 50 to 100-fold higher than in brain. Even lower levels are observed in testis, muscle and liver (Lobsiger et al., 1996; Spreyer et al., 1991; Suter et al., 1994; Taylor et al., 1995; Welcher et al., 1991). In the intestinal tract, signiWcant levels of PMP22 are found in the colon and cecum, with weaker expression in ileum and jejunum. In the stomach, the highest expression is found in the fundus and corpus gastricum. No PMP22 expression is found in preimplantation embryos (Fleming et al., 1997), but PMP22 expression is widespread in several ectodermal, endodermal, and mesodermal tissues during mouse development (Baechner et al., 1995).

III. THE MYELIN GENES AND PRODUCTS

STRUCTURE OF PMP22

STRUCTURE OF PMP22 PMP22 encodes a hydrophobic integral membrane protein of 160 amino acids with a predicted molecular weight of approx. 18 kD (Pareek et al., 1993; Patel et al., 1992; Sedzik et al., 1998; Spreyer et al., 1991; Suter et al., 1992b; Welcher et al., 1991). Computer-aided analysis suggests that the protein consists of four hydrophobic domains that fulWll the requirement for potential transmembrane domains. In the absence of direct structural information, which might be soon forthcoming due to recent progress in the puriWcation of PMP22 (Sedzik et al., 1998; Sedzik and Tsukihara, 2000), the topology of PMP22 has been examined using chimeric proteins consisting of diVerent PMP22 domains fused to reporter proteins and internally tagged molecules. Based on a series of such experiments, it has been proposed that PMP22 may adopt both tetraspan and nontetraspan topologies (D’Urso and Muller, 1997; Taylor et al., 2000). This is an intriguing hypothesis considering recent results demonstrating that the once attained protein topology of polytopic membrane proteins can be changed in a reversible manner in response to alterations in phospholipid composition and may be subject to post-assembly proofreading to correct misfolded structures (Bogdanov et al., 2002). With regard to the topology of PMP22, this might relate to the fast turnover of newly synthesized, presumably improperly folded or aberrantly membraneinserted PMP22 (Pareek et al., 1993, 1997) and the unique lipid composition of myelin (StoVel and Bosio, 1997). In general, the PMP22 amino acid sequence is highly conserved between species, although not identical (Patel et al., 1992; Wulf et al., 1999). The putative intracellular domains of PMP22 are small and it appears unlikely that they are involved in speciWc interactions with intracellular proteins. Intracellular signaling, however, may still be regulated by PMP22-associated proteins given the combined Wndings that PMP22 appears to regulate cell spreading (Brancolini et al., 1999) and the recently described interactions of integrins with polytopic membrane proteins including the CNS myelin component OSP/claudin 11, a distant relative of PMP22 (Tiwari-WoodruV et al., 2001). The extracellular loops of PMP22 may directly interact with other molecules, but so far, no such extracellular interaction has conclusively shown, although direct association of PMP22 with P0 has been reported (D’Urso et al., 1999). Of particular relevance in this regard might be that PMP22 is glycosylated (Fabbretti et al., 1995; ManWoletti et al., 1990; Pareek et al., 1993; Welcher et al., 1991). Intracellularly, PMP22 associates in a glycosylation-dependent manner with the chaperon calnexin (Dickson et al., 2002). PMP22 runs as a 22 kD protein on denaturing SDS polyacrylamide gels but shifts in the presence of the N-glycosylation inhibitor tunicamycin or treatment with N-glycosidase F to 18 kD. This is consistent with a single N-linked glycosylation chain attached to aparagine 41 as suggested by the appropriate consensus sequence in the primary PMP22 polypeptide. Human and cat PMP22 carry the HNK-1 carbohydrate epitope (Hammer et al., 1993; Snipes et al., 1993). This epitope has been identiWed previously as a sulfated glucoronic acid on other cell surface glycoproteins, including P0 (Schachner and Martini, 1995), and Kitamura and colleagues demonstrated directly that also bovine PMP22 contains this peculiar sugar structure (Kitamura et al., 2000). Many of the proteins carrying the HNK-1 epitope function in cell-cell and cell-extracellular matrix adhesion and thus, PMP22 might also be involved in adhesive processes although strong homophilic, protein-based PMP22PMP22 interactions have been excluded (Takeda et al., 2001). PMP22 and P0 are similarly regulated during development and peripheral nerve injury and colocalize in compact myelin. Structural work has revealed that the extracellular domain of P0 forms a tetramer with four molecules arranged around a central cavity (Shapiro et al., 1996). It has been proposed that PMP22 may be associated with this tetrameric complex (D’Urso and Muller, 1997; D’Urso et al., 1999). Such an association Wts well with the fact that mutations aVecting PMP22 or P0 are both associated with inherited demyelinating neuropathies (Berger et al., 2002; Muller, 2000) since the highly ordered structure and speciWc function of myelin implies a requirement for speciWc interactions between various myelin components (Scherer, 1997). Mutations or changes in the stoichiometry of these components are likely to aVect the proper structure and function of

III. THE MYELIN GENES AND PRODUCTS

549

550

21. PMP 22 GENE

such protein complexes and oVer also an attractive hypothesis how changes in PMP22 gene dosage may cause neuropathies (Snipes and Suter, 1995a). Recently, biochemical analysis revealed that PMP22 forms homodimers and also larger complexes (Tobler et al., 1999, 2002), a process that might, in part, be stabilized by PMP22 glycosylation (Ryan et al., 2000). The exact functional implications of PMP22 oligomerization remain unknown, but these processes are likely to play a role in disease processes in Charcot-Marie-Tooth type 1 A (CMT1A), Dejerine-Sottas Syndrome (DSS) or congenital hypomylination (CH) caused by PMP22 point mutations (Tobler et al., 1999; Tobler et al., 2002).

THE PMP22 GENE The genomic structure of the human PMP22 gene has been determined (Patel et al., 1992; Suter et al., 1994). It spans approx. 40 kb and consists of six exons. Exons 1A and 1B are alternatively transcribed resulting in two diVerent mRNAs (Suter et al., 1994). Both transcripts encode the same PMP22 polypeptide but diVer in the sequence of their 5’ untranslated regions. Consequently, the expression of the two mRNAs is regulated by two diVerent promoters P1 and P2. A possible additional promoter has been described, but its relevance in normal tissue remains to be determined (Huehne and Rautenstrauss, 2001). Exon 2 encodes the N-terminus consisting of the Wrst hydrophobic domain of PMP22. Exon 3 encodes the Wrst extracellular loop including the glycosylation site. Exon 4 encodes the second and half of the third hydrophobic domains, and exon 5 covers the remaining of the third, the second putative extracellular loop, the fourth hydrophobic domain, as well as the 3’-untranslated region. The exon-intron structure of PMP22 is highly conserved to those of the EMP-1 and EMP–3 genes (Bolin et al., 1997; Chen et al., 1997; Lobsiger et al., 1996). The number of coding exons and the positions of introns are completely conserved supporting the hypothesis that these genes belong to the same family and are likely derived from duplications of a common ancestral gene. In contrast, the genomic structure of the MP20 gene is not conserved as expected for a more distantly related gene family member (Church and Wang, 1993). The human PMP22 gene has been mapped to human chromosome 17p11.2-p12 (Patel et al., 1992), the mouse PMP22 gene to chromosome 11 (Suter et al., 1992a), and the rat PMP22 gene to 10q22 (Liehr and Rautenstrauss, 1995).

REGULATION OF PMP22 The expression of PMP22 is controlled by a combination of transcriptional and posttranscriptional mechanisms. Furthermore, there is an additional level of regulation aVecting PMP22 protein stability and traYcking (Pareek et al., 1997), and the association of PMP22 with speciWc lipid rafts may be yet another regulatory aspect to be considered (Erne et al., 2002; Hasse et al., 2002). In particular, it has been suggested that certain myelin proteins might be transported in the same raft intracellularly, providing a platform for correct delivery to myelin. If PMP22 and P0 would be indeed associated in such common rafts, it may reinforce the concept of a strict requirement for correct stoichiometry of these two proteins. On the transcriptional level, two diVerent promoters, P1 and P2, have been described (Suter et al., 1994). P1 appears to regulate myelinating Schwann cell–speciWc expression, while P2 is more ubiquitously active (Suter et al., 1994). In line with these data obtained by measuring steady-state PMP22 mRNA levels in diVerent tissues, during nerve development, and in peripheral nerve regeneration, sequence characterization of the promoter regions reveals that P1 contains a TATA-box-like element at the appropriate distance from the transcription initiation site. In contrast, no TATA-box-like sequence could be found in the P2 promoter. The immediate upstream sequence of the P2 promoter has a high GC content and resembles the promoter of a house-keeping gene.

III. THE MYELIN GENES AND PRODUCTS

PMP22 IN DISEASE

First attempts have been made to identify transcriptional regulators of PMP22. Progesterone activates speciWcally the PMP22 promoter 1 in Schwann cell transfection assays (Desarnaud et al., 1998) and glucocorticosteroids stimulate both PMP22 promoters in the same experimental paradigm (Desarnaud et al., 2000). The Wndings are consistent with complementary studies demonstrating positive eVects of sex steroid hormones on gene expression of PMP22 (Melcangi et al., 2000). These results may also be related to the tantalizing data that progesterone is beneWcial for myelination in peripheral nerve regeneration (Koenig et al., 1995). Further transfection studies suggest that 300 bp upstream of the transcription initiation site on exon 1A contain the elements required for Schwann cell–speciWc expression (Saberan-Djoneidi et al., 2000). This minimal promoter activity appears to be under the control of a silencer element sensitive to cAMP, located between
0.3 kb and
3. 5 kb from the start of transcription. Computer analysis of 2 kb of the promoter predicts several transcription factor binding sites, including CREB (potentially involved in the response of PMP22 expression to cAMP stimulation) and steroid receptors. The CREB binding element might be involved in silencing the PMP22 promoter activity. Serial deletions of P1 revealed a positive regulatory element just in front of promoter 1 and a prominent sequence-dependent DNA-protein complex was detected in electrophoretic mobility shift assays using nuclear extracts from the Schwann cell line RT4-D6P2T (Hai et al., 2001b). Site-directed mutagenesis of the binding region identiWed nucleotides at positions
46 to
43 as the crucial elements for the formation of the complex. Nucleotides at positions
46 and
45 were essential for transactivation. Such studies may provide the basis for competitive binding of triplex-forming oligonucleotides to regulate PMP22 expression in vivo (Hai et al., 2001a), a potential gene therapy approach to normalize PMP22 expression in Charcot-Marie-Tooth disease type 1A (CMT1A) due to PMP22 overexpression (Vallat et al., 1996). Although transfection studies in Schwann cells are informative, there are limitations since Schwann cells do not myelinate in vitro without co-culturing with neurons. Transgenic mice provide an ideal system to examine the cis-regulatory elements within the PMP22 gene that are controlled by the intense axon-glia interactions in PNS development and during regeneration. Ten kb upstream of the PMP22 translation start codon have been demonstrated to direct temporal and spatial expression in development and regeneration of peripheral nerves (Maier et al., 2002a). Post-transcriptional regulation of PMP22 has also been suggested, but the responsible elements, which include the two diVerent 5’ and the 3’-untranslated regions, remain to be determined (Bosse et al., 1999).

PMP22 IN DISEASE Most of what we know about the crucial function of PMP22 in proper development and maintenance of the nervous system has been learned from genetics since PMP22 is the culprit gene in the most common form of hereditary motor and sensory neuropathies in human and rodents (Naef and Suter, 1998). This critical role of PMP22 in peripheral nerves became clear soon after the discovery that PMP22 was strongly expressed by myelinating Schwann cells. The Wnding that the mouse mutants Trembler (Tr) and Tr-J carry point mutations in hydrophobic regions of the PMP22 protein sparked the Weld initially (Suter et al., 1992a, 1992b, 1993). These natural mouse mutants had been long suggested to be potential animal models for severe forms of congenital hereditary motor and sensory neuropathies (HMSN) called Dejerine-Sottas Syndrome (DSS) (Ayers and Anderson, 1973; Dejerine and Sottas, 1893; Gabreels-Festen, 2002; Mcleod and Low, 1979). Indeed, the same mutations have been found in humans as described in Tr-J (Valentijn et al., 1992a) and Tr (Ionasescu et al., 1997a) mice before. Fine mapping of the mouse PMP22 gene on chromosome 11 revealed that the PMP22 locus lies on a chromosomal segment potentially syngenic to human chromosome 17p11.2 (Suter et al., 1992a, 1992b). This region had been linked to the most common form of HMSN (Raeymaekers et al., 1989; Vance et al., 1989), 70% of all cases (Nelis et al., 1996;

III. THE MYELIN GENES AND PRODUCTS

551

552

21. PMP 22 GENE

Wise et al., 1993) called CMT1A. CMT1A belongs to a heterogeneous group of neurological disorders (Berger et al., 2002; Maier et al., 2002b; Suter and Snipes, 1995; Suter, 1998) and is dominantly inherited. With a prevalence of 1/2500, these diseases are the most common monogenetic disorders of the nervous system (see Chapter 48). The usual onset of CMT1A is in the second decade of life, manifested by distally pronounced progressive muscle weakness of the legs and hands associated with variable degrees of sensory loss (Birouk et al., 1997; Thomas et al., 1997). The clinical phenotypes vary, even in identical twins, suggesting the involvement other modulatory factors (Garcia et al., 1995). Sural CMT1A nerve biopsies show demyelination and remyelination, accompanied by Schwann cell proliferation and the formation of Schwann cell onion bulb structures (GabreelsFesten et al., 1995). Reduced nerve conduction velocities (NCV) are characteristic of CMT1A (Birouk et al., 1997; Thomas et al., 1997). Most cases of CMT1A are associated with an 1.4 megabase (Mb) intrachromosomal duplication (Inoue et al., 2001; Lupski et al., 1991; Raeymaekers et al., 1991) that contains the intact PMP22 gene (Matsunami et al., 1992; Patel et al., 1992; Timmerman et al., 1992; Valentijn et al., 1992b). The CMT1A duplication segment consists of a tandem repeat which arises from an unequal crossing over due to misalignment of repetitive sequences (termed CMT1A-REP) during meiosis (Pentao et al., 1992). The CMT1A-REP are Xanking the normal CMT1A monomer and are present in an additional copy on the CMT1A-duplicated chromosome (Suter and Patel, 1994). CMT1A-REP contains an insect-derived, functionally defective, mariner transposon-like element (MITE) near a recombination hotspot (Reiter et al., 1996). This may facilitate the recombination event and potentially explain the high frequency of de novo recombination events observed in isolated CMT1A (Hoogendijk et al., 1992). The entire DNA sequence of the rearranged chromosomal region is available and has been analyzed in detail (Inoue et al., 2001). The reciprocal intrachromosomal deletion to the CMT1A duplication is associated with the dominant motor and sensory neuropathy, hereditary neuropathy with liability to pressure palsies (HNPP) (Chance et al., 1993, 1994). HNPP is a recurrent neuropathy that is precipitated by minor trauma to peripheral nerves. With age, the disease may become chronic resembling demyelinating CMT (Cruz-Martinez et al., 1997; Windebank, 1993). Sausage-like myelin structures (called tomacula ¼ sausages) on teased nerve Wber preparations are typical for HNPP (Snipes and Suter, 1995b). In line with the PMP22 being the disease-causing gene heterozygous PMP22 frame-shift mutations that are likely to be null alleles are associated with HNPP (Lenssen et al., 1998; Nicholson et al., 1994; Young et al., 1997) (Fig. 21.1). PMP22 expression studies on biopsies from CMT1A duplication and HNPP deletion patients showed that the altered gene dosage is also reXected at the PMP22 mRNA and protein level (Gabriel et al., 1997; Vallat et al., 1996; Hanemann et al., 1994, 1995; Kamholz et al., 1994; Schenone et al., 1997a, 1997b; Yoshikawa et al., 1994). PMP22transgenic animals provided the Wnal proof that altered PMP22 gene dosage is suYcient to cause HMSN.

POINT MUTATIONS IN PMP22: STRUCTURE/FUNCTION ANALYSIS Appoximately 40 mutations aVecting the PMP22 gene, besides the much more frequent duplications and deletions, have been reported (for a comprehensive list, see http://molgenwww.uia.ac.be/CMTMutations/) (Fig. 21.1). These mutations provide a valuable source for the understanding of PMP22 since they represent a potential in vivo structure/function analysis of the protein. However, most PMP22 mutations alter the traYcking of the PMP22 protein in Schwann cells, and this diVerent traYcking is likely the underlying mechanism of the disease (D’Urso et al., 1998; Naef et al., 1997; Naef and Suter, 1998, 1999; Tobler et al., 1999). Thus, the phenotypes of PMP22 mutations have to be interpreted with some caution since it may often represent the propensity of the mutated protein to aggregate or the degree of interactions of the mutant protein with the wild-

III. THE MYELIN GENES AND PRODUCTS

553

POINT MUTATIONS IN PMP22: STRUCTURE/FUNCTION ANALYSIS

S W L

Q N

S

T C

G

D

T A

H C

S

S

S

V

s

G L

Extracellular

s S

V

I L M

L

L

V

L L

T

F

S

L

V

A V

I

T

L S L

H I

F

S L F

L

L

S

Q

L F

A

I

M

F

L

F

Q L

CMT1

Dominant DSS / CHN V S F

L

Recessive DSS In-Frame Deletion / CMT1 In-Frame Deletion / DSS / Cranial

I

F T

K

G G

A

S

G

T

Cytoplasmic

Y G F

C

I

V

V

L

A

L

Y

A

s

S

A

W

F

P

L

L

V

L A A G

I

F

Y

I

V I

R

HNPP / Stop Frameshift (HNPP)

A

Y

I

Y

I

G

C T

S

P

V

I

I

W

T

M

I

L

H

Y

R

V

Q

A

E

W

D

L

H

N

Q

S

W

Q

S

V

V

I

S

N

P

E

G V

S

H

N

H

N

s Splice Mutations

S

F

L

R

K

R

E

Frameshift (DSS) Frameshift (CMT1 / HNPP)

HNPP / V30M

FIGURE 21.1 Disease-associated PMP22 mutations. Summary of the PMP22 mutations that have been identiWed in human patients with hereditary motor and sensory neuropathies. For details and updates, see http://molgen-www.uia.ac.be/CMTMutations/. Please note that the structure of PMP22 has not been determined and several potential conformations have been suggested (Taylor et al., 2000).

type protein (Dickson et al., 2002; Ryan et al., 2002; Tobler et al., 1999, 2002). Frameshifting mutation in the PMP22 gene tend to be associated with HNPP, consistent with PMP22 null alleles Leu7fs (Nicholson et al., 1994), Gly94fs (Lenssen et al., 1998; Young et al., 1997; but see also Ionasescu et al., 1997b), and Pro122fs (Bissar-Tadmouri et al., 2000). The same correlation is observed with mutations of splice sites in the Wrst part of the PMP22 gene (Bort et al., 1997; Meuleman et al., 2001; but see also Ekici et al., 2000; Nelis et al., 1994) as well as with a nonsense mutation generating an early stop codon (Pareyson and Taroni, 1996). The eVects of other point mutations vary from CMT1, to the more severe DSS, to the most severe phenotype CH, in which Wbers are virtually devoid of myelin (for details on mutations and associated phenotypes, see http://molgen-www.uia.ac.be/CMTMutations/). In general, there is a tendency that PMP22 point mutations show a more severe phenotype than PMP22 duplications (Gabreels-Festen et al., 1995; Tyson et al., 1997). Almost all of these mutations are disease-causing in a heterozygous state. They are either inherited in a dominant fashion or represent de novo mutations. However, there are exceptions to this rule. The signiWcance of a potentially recessive Thr118Met mutation is debated (Naef and Suter, 1998; Nelis et al., 1997; Roa et al., 1993a; Seeman et al., 1999; Young et al., 2000), but a family with proven recessive DSS carrying a Arg157Trp mutation has been described (Parman et al., 1999). A similar but not the same mutation was also found in a hemizygous state (Arg157Gly allele combined with a PMP22 deletion allele), leading to the CMT1 phenotype (Numakura et al., 2000; see also comments in Beckmann and Schroder, 2000; Lupski, 2000). Some speciWc mutants show unusual features that are not seen with other PMP22 mutations. The mutations W28R (Boerkoel et al., 2002), Ala67Pro (Kovach et al., 1999), Ser72Leu (Ionasescu et al., 1996), and Ser72Ile (Tyson et al., 1997) are associated with severe sensorineural deafness. The reasons for this phenotype are not clear. However,

III. THE MYELIN GENES AND PRODUCTS

554

21. PMP 22 GENE

PMP22 is expressed in cranial nerves during development and in early migratory neural crest cells as well as other tissues (Baechner et al., 1995; Hagedorn et al., 1999; Maier et al., 2002a; Paratore et al., 2002; Parmantier et al., 1995, 1997). These observations suggest that these PMP22 mutations might cause sensorineural deafness by aVecting the eighth cranial nerve and the inner ear (a neural crest derivative) during development or by demyelination. Interestingly in this context, a patient with a deletion of Phe84 has bee described with DSS including multiple cranial nerve alterations (Yener et al., 2001). Furthermore, one of the mutations associated with deafness, Ala67Pro, was found in a large family and showed the phenomenon of anticipation as determined by careful clinical evaluation (Kovach et al., 2002). The reason for this Wnding remains unclear. Fabrizi and colleagues identiWed a particularly interesting mutation, Asp37Val, that is classiWed as CMT1 and shows a very unusual prominent uncompaction of the myelin sheath (Fabrizi et al., 1999). This bears some similarity to CMT1B caused by P0 mutations. Interestingly, this PMP22 alteration is one of the few disease-causing PMP22 mutations that are not located in hydrophobic domains of the protein (see the discussion that follows and Fig. 21.1). Thus, it is tantalizing to speculate that this particular mutant protein may not be mainly trapped within the Schwann cells as are many other PMP22 mutants (Naef and Suter, 1999). The mutation may rather aVect and reveal an important function of PMP22 in the compaction of the myelin sheath, potentially by interacting with P0 (D’Urso et al., 1999). If the frequency of mutations at diVerent locations in the PMP22 polypeptide is analyzed, a mutational hotspot at Ser72 becomes notable. The mutation Ser72Leu has been found in multiple families and de novo cases (Bissar-Tadmouri et al., 2000; CeuterickDe Groote et al., 2001; Ionasescu et al., 1996; Marques et al., 1998; Mostacciuolo et al., 2001; Roa et al., 1993b; Simonati et al., 1999). Interestingly, some patients have been classiWed with CMT1, others with DSS, and a third group with CH emphasizing the diYculties in generating genotype-phenotype correlations in PMP22-based diseases. In addition, Ser72Pro, Ser72Trp, and Ser76Ile mutations have been found in DSS (Ekici et al., 2001; Tyson et al., 1997). These diVerent mutations at the same amino acid residue identify Ser72 as being critical in the proper function of PMP22 in the normal organism. The Val30Met mutant described by Sahenk and colleagues (Sahenk et al., 1998) is the only known PMP22 missense mutation that leads to a comparable phenotype to the PMP22 null allele in HNPP. All other missense mutations lead to CMT1, DSS, or CH. Thus, these mutants represent not null but gain-of-function alleles. A series of transfection studies and in situ analysis in Tr and Tr-J, suggested that the pathology results from a combination of mistraYcking-based loss of function, further augmented by the ability of some mutants to disrupt normal traYcking of the product of the wild-type protein (Naef et al., 1997; Tobler et al., 1999). Wild-type PMP22 protein appears to heterodimerize with traYcking-incompetent mutants that would reduce the amount of PMP22 transported to the myelin sheath (Sanders et al., 2001). Indeed, mutated Tr and Tr-J PMP22 protein and complexes containing this aberrant forms of the protein cannot reach the myelin sheath in detectable amounts (Colby et al., 2000). To explore this problem further, Tobler and colleagues (Tobler et al., 2002) have compared the aggregation of wild-type PMP22 with those of Tr and Tr-J mutant protein in transfected cells. All three proteins form homodimers and heterodimers with each other and build up also larger high molecular weight complexes. Both mutant proteins also sequester the same amount of wild-type PMP22 in heterodimers and heterooligomers. However, Tr PMP22 appears to be more prone to aggregation. Thus, the authors suggest that the diVerences in the phenotypes of heterozygous Tr and TrJ mice (Tr is more aVected than TrJ; see Henry et al., 1983; Notterpek et al., 1997) may depend more on the ability of the mutant protein to aggregate than on the dominant-negative eVect of the mutant PMP22 on wild-type PMP22 traYcking. Whether the fact that Tr PMP22 is mainly retained in the endoplasmic reticulum (ER) (Naef et al., 1997) while the Tr-J protein reaches the intermediate compartment between the ER and the Golgi apparatus is of functional signiWcance remains to be elucidated. Aggregation in special intracellular structures called ‘‘aggresomes,’’ and regulation of protein degradation by the proteasome pathway appears to be an important issue in the function of PMP22 in

III. THE MYELIN GENES AND PRODUCTS

MICE LACKING PMP22

health and disease (Notterpek et al., 1999a; Ryan et al., 2002). However, the detailed mechanisms have not been worked out. In particular, it is not clear how PMP22 traYcking alterations and the observed aggregations are related with each other and how this applies to other PMP22 mutations. Furthermore, comparative analysis of Tr/Tr, Tr/þ and compound heterozygous mice that carried a null allele and a Tr allele revealed that the mutated PMP22 protein can act by a true gain-of-function mechanism having some deleterious eVects on its own without wild-type protein (Adlkofer et al., 1997b). Recent data suggest that calnexin is the major folding partner of PMP22 in myelinating Schwann cells (Dickson et al., 2002), and the same study suggests that sequestration of calnexin by mutant PMP22 into myelin-like intracellular structures might contribute to the disease phenotype. Such astonishing structures of multiple membrane lamellae, stacked onto each other and reminiscent of myelin, were found in PMP22 transfected non-neural HeLa and 293A cells, suggesting a role of PMP22 in the organization of membranes (Dickson et al., 2002). As mentioned earlier, almost all mutations that are associated with disease are located within hydrophobic domains of PMP22. Exceptions are The HNPP-linked Val30Met mutant; the Asp37Val mutant that shows myelin uncompaction; a Gly93Arg (Ohnishi et al., 1995), which is located either intracellular or extracellular depending on the PMP22 model applied (Taylor et al., 2000) and does not display obvious peculiarities; the recessive DSS mutation Arg157Trp (Parman et al., 1999); as well as the hemizygous Arg157Gly (Numakura et al., 2000) mutant, which is located near the intracellular carboxyterminus of PMP22 (D’Urso and Muller, 1997) (Fig. 21.1). One might hypothesize that mutations aVecting the hydrophobic domains of PMP22 may invariably have aggregation or traYcking problems that also aVect the wild-type protein (Naef and Suter, 1999), while the other mutants may act by diVerent mechanisms (Beckmann and Schroder, 2000; Lupski, 2000).

MICE LACKING PMP22 Despite all the knowledge obtained from the genetics of PMP22 in disease, the basic function of PMP22 is not yet clear (Naef and Suter, 1998). To address this question in vivo, mice carrying an inactivated PMP22 gene were generated using homologous recombination in embryonic stem cells (Adlkofer et al., 1995, 1997a; Sancho et al., 1999, 2001). Mice that completely lack PMP22 are viable but develop walking diYculties due to progressive weakness of the hind limbs after approximately 2 weeks of age. Morphological analysis of peripheral nerves during development revealed a mildly delayed onset of myelination in young animals. This indicated some important function of PMP22 in the initial steps of myelination (Carenini et al., 1999). Later, characteristic paranodal but also internodal tomacula are formed, which degenerate with progressing age and remodel the Schwann cell and axonal protein composition by demyelination (Neuberg et al., 1999). Later in life, Schwann cell onion bulbs, associated with very slow NCV, become predominant although some tomacula can still be found (Adlkofer et al., 1995; Naef and Suter, 1998; Sancho et al., 1999, 2001). Schwann cell onion bulb formation is accompanied by aberrant proliferation and cell death (Sancho et al., 2001). Furthermore, axonal atrophy and axonal loss are also observed and are likely to contribute to the clinical phenotype (Sancho et al., 1999). Finally, muscular atrophy develops as characterized by extensive type grouping of muscle Wbers and ultraterminal axonal sprouting (Maier et al., 2002b). Other tissues than the peripheral nerves are not aVected despite the widespread expression of PMP22 in development and in adults. Possibly, more subtle abnormalities have yet to be discovered. Potential compensatory eVects by other members of the PMP22/EMP/MP20 family are also conceivable due to their co-expression in various tissues. Heterozygous ‘‘knock-out’’ PMP22 (þ/0) mice have retained only one PMP22 gene copy and mimic genetically, behaviorally, and morphologically HNPP (Adlkofer et al., 1997a). Mice and rats with increased PMP22 gene dosage have also been generated as animal models for CMT1A (Huxley et al., 1996, 1998; Magyar et al., 1996; Niemann et al., 2000; Norreel et al., 2001; Perea et al., 2001; Robertson et al., 2002; Sancho et al., 1999, 2001; Sereda et al., 1996). These animals will be further discussed in Section V.

III. THE MYELIN GENES AND PRODUCTS

555

556

21. PMP 22 GENE

CELLULAR APPROACHES TOWARD THE FUNCTION OF PMP22 PMP22 has been implicated in proliferation control and apoptosis. It is induced by growth arrest of NIH3T3 cells and embryonic Wbroblasts after serum starvation (or at conXuence), while its expression becomes down-regulated after serum addition (Ciccarelli et al., 1990; ManWoletti et al., 1990; Suter et al., 1994). High levels of PMP22 are also found in postmitotic adipocytes (Shugart et al., 1995), and PMP22 is up-regulated in diVerentiated Schwann cells (Welcher et al., 1991; Zoidl et al., 1995). PMP22 is also induced in rat pheochromocytoma PC12 cells during neuronal diVerentiation after nerve growth factor (NGF) treatment (De Leon et al., 1994). However, PMP22 is not always regulated by growth arrest. In C6 glial cells, NGF-induced growth arrest has no eVect on PMP22 expression (De Leon et al., 1994), and in L6 myoblasts, PMP22 mRNA increases only slightly when cells are grown to conXuency and become quiescent, but is strongly downregulated when cells are induced to terminally diVerentiate into myotubules (Cowled et al., 1994). Functional experiments have shown that overexpression of PMP22 in NIH3T3 cells induces apoptosis (Fabbretti et al., 1995). The extent of cell death appears to correlate with the level of PMP22 expression. N-acetyl cysteine and ascorbic acid prevent this induction, suggesting that generation of active oxygen intermediates may be part of the apoptosisinducing mechanism (Fabbretti et al., 1995). When the apoptotic response triggered by PMP22 was prevented by Bcl-2 coexpression, altered cell spreading was observed (Brancolini et al., 1999). RhoA counteracts the PMP22-dependent morphological response but cannot neutralize the apoptotic response. Thus, regulation of Schwann cell shape and spreading regulated by PMP22 through the Rho GTPase might play a role during Schwann cells diVerentiation and myelination. Mutant PMP22 containing point mutations found in CMT1A (within the hydrophobic domains of PMP22) showed signiWcantly reduced ability to induce apoptosis in NIH3T3 cells (Fabbretti et al., 1995). Consistent with these Wndings, Brancolini and colleagues (Brancolini et al., 2000) demonstrated that cell surface expression of PMP22 is required to regulate both cell death and cell spreading. Since most mutant PMP22 proteins do not reach the plasmamebrane and surface expression of wild-type PMP22 protein is also compromised by oligomerization, reduced exposure of PMP22 at the cell surface may contribute to the disease phenotype. It is also interesting to note in this context that loss of Schwann cells in biopsies from CMT1A and HNPP patients has been reported due to apoptosis (Erdem et al., 1998). In contrast, only alterations in Schwann cell diVerentiation were observed in other experiments (Hanemann et al., 1997). Overexpression of PMP22 in cultured Schwann cells also markedly reduced the rate of proliferation (Hanemann et al., 1998; Zoidl et al., 1995). Mechanistically, PMP22 delayed the entry of cells from G0/G1 into the S phase or growth arrested cells in G0. The signiWcance of these in vitro studies in vivo is unclear. Both overexpression and reduced PMP22 expression lead to Schwann cell hypertrophy (Adlkofer et al., 1995; Magyar et al., 1996; Sereda et al., 1996), and no diVerences in the proliferation rates in early post-natal development of mutant sciatic nerves has been observed (Sancho et al., 2001). However, an association of PMP22 (and the EMP proteins; Jetten and Suter, 2000) with the C-terminal domain of the ATP receptor P2X7 has been found (Wilson et al., 2002). Binding of extracellular ATP to the P2X7 receptor opens an integral cation permeable channel and leads to membrane blebbing and eventual cell death. Since overexpression of PMP22 leads to a similar phenotype, the interaction with P2X7 may mediate some aspects of the downstream signaling. Interestingly, ATP released by DRG neurons in culture inhibits Schwann cells proliferation and arrests development in response to impulse activity, at least in vitro (Fields and Stevens, 2000). Whether this mechanism may be related to PMP22 function in peripheral nerves and how this might potentially even relate to disease mechanisms in PMP22-based disorders remains open. It had been noted, based on primary amino acid sequence comparisons, that PMP22 shares some similarities with the crucial tight junction components of the claudin family.

III. THE MYELIN GENES AND PRODUCTS

CELLULAR APPROACHES TOWARD THE FUNCTION OF PMP22

Interestingly, PMP22-like immunoreactivity was found in adult rat liver and intestine, and cultured epithelial cells, associated with markers of the tight junctional complex, including zonula occludens 1 (ZO-1) and occludin (Notterpek et al., 2001). Exogenous myc-tagged PMP22 was targeted to apical cell junctions in polarized epithelia and to anti-ZO-1 antibody immunoreactive cell contacts of Wbroblasts. Although it is not formally clear whether PMP22 is suYcient to induce proper tight junction strands, these studies support a role for PMP22 at intercellular junctions of epithelia. PMP22 appears not to be necessary for tight junction formations though, since epithelia are not aVected in mice without PMP22. However, the members of the EMP family of proteins are expressed in these tissues and may compensate for the loss of PMP22.

Acknowledgments This work was supported by the Swiss National Science Foundation, the Swiss Muscle Disease Foundation, the Wolfermann-Na¨geli Stiftung, the National Center of Competence in Research ‘‘Neural Plasticity and Repair,’’ and the Swiss Bundesamt for Science related to the Commission of the European Communities, speciWc RTD program ‘‘Quality of Life and Management of Living Resources,’’ QLK6-CT-2000-00179.

References Adlkofer, K., Martini, R., Aguzzi, A., Zielasek, J., Toyka, K. V., and Suter, U. (1995). Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deWcient mice, Nat Genet 11, 274–280. Adlkofer, K., Frei, R., Neuberg, D. H., Zielasek, J., Toyka, K. V., and Suter, U. (1997a). Heterozygous peripheral myelin protein 22-deWcient mice are aVected by a progressive demyelinating tomaculous neuropathy, J Neurosci 17, 4662–4671. Adlkofer, K., Naef, R., and Suter, U. (1997b). Analysis of compound heterozygous mice reveals that the Trembler mutation can behave as a gain-of-function allele, J Neurosci Res 49, 671–680. Agostoni, E., Gobessi, S., Brancolini, C., and Schneider, C. (1999). IdentiWcation and characterization of a new member of the gas3/PMP22 gene family in C. elegans, Gene 234, 267–274. Attardi, L. D., Reczek, E. E., Cosmas, C., Demicco, E. G., McCurrach, M. E., Lowe, S. W., and Jacks, T. (2000). PERP, an apoptosis-associated target of p53, is a novel member of the PMP-22/gas3 family, Genes Dev 14, 704–718. Ayers, M. M., and Anderson, R. M. (1973). Onion bulb neuropathy in the Trembler mouse: A model of hypertrophic interstitial neuropathy (Dejerine-Sottas) in man, Acta Neuropathol 25, 54–70. Baechner, D., Liehr, T., Hameister, H., Altenberger, H., Grehl, H., Suter, U., and Rautenstrauss, B. (1995). Widespread expression of the peripheral myelin protein-22 gene (PMP22) in neural and non-neural tissues during murine development, J Neurosci Res 42, 733–741. Beckmann, A., and Schroder, J. M. (2000). Recessive PMP22 mutations: Quantitative and qualitative aspects, Ann Neurol 48, 131–132. Ben-Porath, I., Yanuka, O., and Benvenisty, N. (1999). The tmp gene, encoding a membrane protein, is a c-Myc target with a tumorigenic activity, Mol Cell Biol 19, 3529–3539. Berger, P., Young, P., and Suter, U. (2002). Molecular cell biology of Charcot-Marie-Tooth disease, Neurogenetics 4, 1–15. Birouk, N., Gouider, R., Le Guern, E., Gugenheim, M., Tardieu, S., Maisonobe, T., Le Forestier, N., Agid, Y., Brice, A., and Bouche, P. (1997). Charcot-Marie-Tooth disease type 1A with 17p11.2 duplication. Clinical and electrophysiological phenotype study and factors inXuencing disease severity in 119 cases, Brain 120, 813–823. Bissar-Tadmouri, N., Parman, Y., Boutrand, L., Deymeer, F., Serdaroglu, P., Vandenberghe, A., and Battaloglu, E. (2000). Mutational analysis and genotype/phenotype correlation in Turkish Charcot-Marie-Tooth Type 1 and HNPP patients, Clin Genet 58, 396–402. Boerkoel, C. F., Takashima, H., Garcia, C. A., Olney, R. K., Johnson, J., Berry, K., Russo, P., Kennedy, S., Teebi, A. S., Scavina, M., et al. (2002). Charcot-Marie-Tooth disease and related neuropathies: Mutation distribution and genotype-phenotype correlation, Ann Neurol 51, 190–201. Bogdanov, M., Heacock, P. N., and Dowhan, W. (2002). A polytopic membrane protein displays a reversible topology dependent on membrane lipid composition, Embo J 21, 2107–2116. Bolin, L. M., McNeil, T., Lucian, L. A., DeVaux, B., Franz-Bacon, K., Gorman, D. M., Zurawski, S., Murray, R., and McClanahan, T. K. (1997). HNMP-1: A novel hematopoietic and neural membrane protein diVerentially regulated in neural development and injury, J Neurosci 17, 5493–5502. Bort, S., Nelis, E., Timmerman, V., Sevilla, T., Cruz-Martinez, A., Martinez, F., Millan, J. M., Arpa, J., Vilchez, J. J., Prieto, F., et al. (1997). Mutational analysis of the MPZ, PMP22 and Cx32 genes in patients of Spanish ancestry with Charcot-Marie-Tooth disease and hereditary neuropathy with liability to pressure palsies, Hum Genet 99, 746–754.

III. THE MYELIN GENES AND PRODUCTS

557

558

21. PMP 22 GENE

Bosse, F., Brodbeck, J., and Muller, H. W. (1999). Post-transcriptional regulation of the peripheral myelin protein gene PMP22/gas3, J Neurosci Res 55, 164–177. Brancolini, C., Edomi, P., Marzinotto, S., and Schneider, C. (2000). Exposure at the cell surface is required for gas3/PMP22 to regulate both cell death and cell spreading: Implication for the Charcot-Marie-Tooth type 1A and Dejerine-Sottas diseases, Mol Biol Cell 11, 2901–2914. Brancolini, C., Marzinotto, S., Edomi, P., Agostoni, E., Fiorentini, C., Muller, H. W., and Schneider, C. (1999). Rho-dependent regulation of cell spreading by the tetraspan membrane protein Gas3/PMP22, Mol Biol Cell 10, 2441–2459. Carenini, S., Neuberg, D., Schachner, M., Suter, U., and Martini, R. (1999). Localization and functional roles of PMP22 in peripheral nerves of P0-deWcient mice, Glia 28, 256–264. Ceuterick-de Groote, C., De Jonghe, P., Timmerman, V., Van Goethem, G., Lofgren, A., Ceulemans, B., Van Broeckhoven, C., and Martin, J. J. (2001). Infantile demyelinating neuropathy associated with a de novo point mutation on Ser72 in PMP22 and basal lamina onion bulbs in skin biopsy, Pathol Res Pract 197, 193–198. Chance, P. F., Abbas, N., Lensch, M. W., Pentao, L., Roa, B. B., Patel, P. I., and Lupski, J. R. (1994). Two autosomal dominant neuropathies result from reciprocal DNA duplication/deletion of a region on chromosome 17, Hum Mol Genet 3, 223–228. Chance, P. F., Alderson, M. K., Leppig, K. A., Lensch, M. W., Matsunami, N., Smith, B., Swanson, P. D., Odelberg, S. J., Disteche, C. M., and Bird, T. D. (1993). DNA deletion associated with hereditary neuropathy with liability to pressure palsies, Cell 72, 143–151. Chen, Y., Medvedev, A., Ruzanov, P., Marvin, K. W., and Jetten, A. M. (1997). cDNA cloning, genomic structure, and chromosome mapping of the human epithelial membrane protein CL-20 gene (EMP1), a member of the PMP22 family, Genomics 41, 40–48. Church, R. L., and Wang, J. H. (1993). The human lens Wber-cell intrinsic membrane protein MP19 gene: Isolation and sequence analysis, Curr Eye Res 12, 1057–1065. Ciccarelli, C., Philipson, L., and Sorrentino, V. (1990). Regulation of expression of growth arrest-speciWc genes in mouse Wbroblasts, Mol Cell Biol 10, 1525–1529. Colby, J., Nicholson, R., Dickson, K. M., Orfali, W., Naef, R., Suter, U., and Snipes, G. J. (2000). PMP22 carrying the trembler or trembler-J mutation is intracellularly retained in myelinating Schwann cells, Neurobiol Dis 7, 561–573. Cowled, P. A., Ciccarelli, C., Coccia, E., Philipson, L., and Sorrentino, V. (1994). Expression of growth arrestspeciWc (gas) genes in senescent murine cells, Exp Cell Res 211, 197–202. Cruz-Martinez, A., Bort, S., Arpa, J., Duarte, J., and Palau, F. (1997). Clinical, genetic and electrophysiological correlation in hereditary neuropathy with liability to pressure palsies with involvement of the PMP22 gene at chromosome 17, Eur J Neurol 4, 274–286. D’Urso, D., Ehrhardt, P., and Muller, H. W. (1999). Peripheral myelin protein 22 and protein zero: A novel association in peripheral nervous system myelin, J Neurosci 19, 3396–3403. D’Urso, D., and Muller, H. W. (1997). Ins and outs of peripheral myelin protein-22: Mapping transmembrane topology and intracellular sorting, J Neurosci Res 49, 551–562. D’Urso, D., Prior, R., Greiner-Petter, R., Gabreels-Festen, A. A., and Muller, H. W. (1998). Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutations of the peripheral myelin protein PMP22, J Neurosci 18, 731–740. De Leon, M., Nahin, R. L., Mendoza, M. E., and Ruda, M. A. (1994). SR13/PMP-22 expression in rat nervous system, in PC12 cells, and C6 glial cell lines, J Neurosci Res 38, 167–181. De Leon, M., Welcher, A. A., Suter, U., and Shooter, E. M. (1991). IdentiWcation of transcriptionally regulated genes after sciatic nerve injury, J Neurosci Res 29, 437–448. Dejerine, J., and Sottas, J. (1893). Sur la ne´vrite interstitielle, hypertrophique et progressive de l’enfance, CR Soc Biol 45, 63–96. Desarnaud, F., Bidichandani, S., Patel, P. I., Baulieu, E. E., and Schumacher, M. (2000). Glucocorticosteroids stimulate the activity of the promoters of peripheral myelin protein-22 and protein zero genes in Schwann cells, Brain Res 865, 12–16. Desarnaud, F., Do Thi, A. N., Brown, A. M., Lemke, G., Suter, U., Baulieu, E. E., and Schumacher, M. (1998). Progesterone stimulates the activity of the promoters of peripheral myelin protein-22 and protein zero genes in Schwann cells, J Neurochem 71, 1765–1768. Dickson, K. M., Bergeron, J. J., Shames, I., Colby, J., Nguyen, D. T., Chevet, E., Thomas, D. Y., and Snipes, G. J. (2002). Association of calnexin with mutant peripheral myelin protein-22 ex vivo: A basis for ‘‘gain-offunction’’ ER diseases, Proc Natl Acad Sci USA 99, 9852–9857. Ekici, A. B., Park, O., Korinthenberg, R., Grehl, H., and Rautenstrauss, B. (2001). T>C transition in codon 72 (TCG!CCG), S72P, a putative hotspot in PMP22, Hum Mutat 17, 81. Ekici, A. B., Schweitzer, D., Park, O., Lorek, D., Rautenstrauss, B., Kruger, G., Friedl, W., Uhlhaas, S., Bathke, K., Heuss, D., et al. (2000). Charcot-Marie-Tooth disease and related peripheral neuropathies: Novel mutations in the peripheral myelin genes connexin 32 (Cx32), peripheral myelin protein 22 (PMP22), and peripheral myelin protein zero (MPZ), Neurogenetics 3, 49–50. Erdem, S., Mendell, J. R., and Sahenk, Z. (1998). Fate of Schwann cells in CMT1A and HNPP: Evidence for apoptosis, J Neuropathol Exp Neurol 57, 635–642. Erne, B., Sansano, S., Frank, M., and Schaeren-Wiemers, N. (2002). Rafts in adult peripheral nerve myelin contain major structural myelin proteins and myelin and lymphocyte protein (MAL) and CD59 as speciWc markers, J Neurochem 82, 550–562.

III. THE MYELIN GENES AND PRODUCTS

CELLULAR APPROACHES TOWARD THE FUNCTION OF PMP22

Fabbretti, E., Edomi, P., Brancolini, C., and Schneider, C. (1995). Apoptotic phenotype induced by overexpression of wild-type gas3/PMP22: Its relation to the demyelinating peripheral neuropathy CMT1A, Genes Dev 9, 1846–1856. Fabrizi, G. M., Cavallaro, T., Taioli, F., Orrico, D., Morbin, M., Simonati, A., and Rizzuto, N. (1999). Myelin uncompaction in Charcot-Marie-Tooth neuropathy type 1A with a point mutation of peripheral myelin protein-22, Neurology 53, 846–851. Fields, R. D., and Stevens, B. (2000). ATP: An extracellular signaling molecule between neurons and glia, Trends Neurosci 23, 625–633. Fleming, J. V., Fontanier, N., Harries, D. N., and Rees, W. D. (1997). The growth arrest genes gas5, gas6, and CHOP-10 (gadd153) are expressed in the mouse preimplantation embryo, Mol Reprod Dev 48, 310–316. Gabreels-Festen, A. (2002). Dejerine-Sottas syndrome grown to maturity: Overview of genetic and morphological heterogeneity and follow-up of 25 patients, J Anat 200, 341–356. Gabreels-Festen, A. A., Bolhuis, P. A., Hoogendijk, J. E., Valentijn, L. J., Eshuis, E. J., and Gabreels, F. J. (1995). Charcot-Marie-Tooth disease type 1A: Morphological phenotype of the 17p duplication versus PMP22 point mutations, Acta Neuropathol 90, 645–649. Gabriel, J. M., Erne, B., Pareyson, D., Sghirlanzoni, A., Taroni, F., and Steck, A. J. (1997). Gene dosage eVects in hereditary peripheral neuropathy. Expression of peripheral myelin protein 22 in Charcot-Marie-Tooth disease type 1A and hereditary neuropathy with liability to pressure palsies nerve biopsies, Neurology 49, 1635–1640. Garcia, C. A., Malamut, R. E., England, J. D., Parry, G. S., Liu, P., and Lupski, J. R. (1995). Clinical variability in identical twins with Charcot-Marie-Tooth disease type 1A duplication, Neurology 45, 2090–2093. Hagedorn, L., Suter, U., and Sommer, L. (1999). P0 and PMP22 mark a multipotent neural crest-derived cell type that displays community eVects in response to TGF-beta family factors, Development 126, 3781–3794. Hai, M., Bidichandani, S. I., Hogan, M. E., and Patel, P. I. (2001a). Competitive binding of triplex-forming oligonucleotides in the two alternate promoters of the PMP22 gene, Antisense Nucleic Acid Drug Dev 11, 233–246. Hai, M., Bidichandani, S. I., and Patel, P. I. (2001b). IdentiWcation of a positive regulatory element in the myelinspeciWc promoter of the PMP22 gene, J Neurosci Res 65, 508–519. Hammer, J. A., O’Shannessy, D. J., De Leon, M., Gould, R., Zand, D., Daune, G., and Quarles, R. H. (1993). Immunoreactivity of PMP-22, P0, and other 19 to 28 kDa glycoproteins in peripheral nerve myelin of mammals and Wsh with HNK1 and related antibodies, J Neurochem 35, 546–558. Hanemann, C. O., Gabreels-Festen, A. A., Stoll, G., and Muller, H. W. (1997). Schwann cell diVerentiation in Charcot-Marie-Tooth disease type 1A (CMT1A): Normal number of myelinating Schwann cells in young CMT1A patients and neural cell adhesion molecule expression in onion bulbs, Acta Neuropathol (Berl) 94, 310–315. Hanemann, C. O., Rosenbaum, C., Kupfer, S., Wosch, S., Stoegbauer, F., and Muller, H. W. (1998). Improved culture methods to expand Schwann cells with altered growth behaviour from CMT1A patients In Process Citation, Glia 22, 89–98. Hanemann, C. O., Stoll, G., D’Urso, D., Fricke, W., Martin, J. J., Van Broeckhoven, C., Mancardi, G. L., Bartke, I., and Muller, H. W. (1994). Peripheral myelin protein-22 expression in Charcot-Marie-Tooth disease type 1a sural nerve biopsies, J Neurosci Res 37, 654–659. Hanemann, C. O., Stoll, G., and Muller, H. W. (1995). PMP22 expression in CMT1a neuropathy, Ann Neurol 37, 136. Haney, C., Snipes, G. J., Shooter, E. M., Suter, U., Garcia, C., GriYn, J. W., and Trapp, B. D. (1996). Ultrastructural distribution of PMP22 in Charcot-Marie-Tooth disease type 1A, J Neuropathol Exp Neurol 55, 290–299. Hasse, B., Bosse, F., and Muller, H. W. (2002). Proteins of peripheral myelin are associated with glycosphingolipid/cholesterol-enriched membranes, J Neurosci Res 69, 227–232. Henry, E. W., Cowen, J. S., and Sidman, R. L. (1983). Comparison of trembler and trembler-J phenotypes: Varying severity of peripheral hypomyelination, J Neuropathol Exp Neurol 42, 688–706. Hoogendijk, J. E., Hensels, G. W., Gabreels, F. A., Gabreels, F. J., Janssen, E. A., de, J. P., Martin, J. J., van, B. C., Valentijn, L. J., Baas, F., et al. (1992). De-novo mutation in hereditary motor and sensory neuropathy type I, Lancet 339, 1081–1082. Huehne, K., and Rautenstrauss, B. (2001). Transcriptional startpoints and methylation patterns in the PMP22 promoters of peripheral nerve, leukocytes and tumor cell lines, Int J Mol Med 7, 669–675. Huxley, C., Passage, E., Manson, A., Putzu, G., Figarella-Branger, D., Pellissier, J. F., and Fontes, M. (1996). Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear injection of human YAC DNA, Hum Mol Genet 5, 563–569. Huxley, C., Passage, E., Robertson, A. M., Youl, B., Huston, S., Manson, A., Saberan-Djoniedi, D., FigarellaBranger, D., Pellissier, J. F., Thomas, P. K., and Fontes, M. (1998). Correlation between varying levels of PMP22 expression and the degree of demyelination and reduction in nerve conduction velocity in transgenic mice, Hum Mol Genet 7, 449–458. Inoue, K., Dewar, K., Katsanis, N., Reiter, L. T., Lander, E. S., Devon, K. L., Wyman, D. W., Lupski, J. R., and Birren, B. (2001). The 1.4-Mb CMT1A duplication/HNPP deletion genomic region reveals unique genome architectural features and provides insights into the recent evolution of new genes, Genome Res 11, 1018–1033. Ionasescu, V. V., Searby, C., and Greenberg, S. A. (1996). Dejerine-Sottas disease with sensorineural hearing loss, nystagmus, and peripheral facial nerve weakness: De novo dominant point mutation of the PMP22 gene, J Med Genet 33, 1048–1049.

III. THE MYELIN GENES AND PRODUCTS

559

560

21. PMP 22 GENE

Ionasescu, V. V., Searby, C. C., Ionasescu, R., Chatkupt, S., Patel, N., and Koenigsberger, R. (1997a). DejerineSottas neuropathy in mother and son with same point mutation of PMP22 gene, Muscle Nerve 20, 97–99. Ionasescu, V. V., Searby, C. C., Ionasescu, R., Reisin, R., Ruggieri, V., and Arberas, C. (1997b). Severe CharcotMarie-Tooth neuropathy type 1A with 1-base pair deletion and frameshift mutation in the peripheral myelin protein 22 gene, Muscle Nerve 20, 1308–1310. Jetten, A. M., and Suter, U. (2000). The peripheral myelin protein 22 and epithelial membrane protein family, Prog Nucleic Acid Res Mol Biol 64, 97–129. Kamholz, J., Shy, M., and Scherer, S. (1994). Elevated expression of messenger RNA for peripheral myelin protein 22 in biopsied peripheral nerves of patients with Charcot-Marie-Tooth disease type 1A [letter; comment], Ann Neurol 36, 451–452. Kitamura, K., Suzuki, M., and Uyemura, K. (1976). PuriWcation and partial characterization of two glycoproteins in bovine peripheral nerve myelin membrane, Biochim Biophys Acta 455, 806–816. Kitamura, K., Uyemura, K., Shibuya, K., Sakamoto, Y., Yoshimura, K., and Nomura, M. (2000). Structure of a major oligosaccharide of PASII/PMP22 glycoprotein in bovine peripheral nerve myelin, J Neurochem 75, 853–860. Koenig, H. L., Schumacher, M., Ferzaz, B., Thi, A. N., Ressouches, A., Guennoun, R., Jung-Testas, I., Robel, P., Akwa, Y., and Baulieu, E. E. (1995). Progesterone synthesis and myelin formation by Schwann cells, Science 268, 1500–1503. Kovach, M. J., Campbell, K. C., Herman, K., Waggoner, B., Gelber, D., Hughes, L. F., and Kimonis, V. E. (2002). Anticipation in a unique family with Charcot-Marie-Tooth syndrome and deafness: Delineation of the clinical features and review of the literature, Am J Med Genet 108, 295–303. Kovach, M. J., Lin, J. P., Boyadjiev, S., Campbell, K., Mazzeo, L., Herman, K., Rimer, L. A., Frank, W., Llewellyn, B., Wang Jabs, E., et al. (1999). A unique point mutation in the PMP22 gene is associated with Charcot-Marie-Tooth disease and deafness, Am J Hum Genet 64, 1580–1593. Kuhn, G., Lie, A., Wilms, S., and Muller, H. W. (1993). Coexpression of PMP22 gene with MBP and P0 during de novo myelination and nerve repair, Glia 8, 256–264. Lenssen, P. P., Gabreels-Festen, A. A., Valentijn, L. J., Jongen, P. J., van Beersum, S. E., van Engelen, B. G., van Wensen, P. J., Bolhuis, P. A., Gabreels, F. J., and Mariman, E. C. (1998). Hereditary neuropathy with liability to pressure palsies. Phenotypic diVerences between patients with the common deletion and a PMP22 frame shift mutation, Brain 121, 1451–1458. Liehr, T., and Rautenstrauss, B. (1995). Regional localization of rat peripheral myelin protein 22 (Pmp22) gene to chromosome 10q22 by FISH, Mamm Genome 6, 489. Lobsiger, C. S., Magyar, J. P., Taylor, V., Wulf, P., Welcher, A. A., Program, A. E., and Suter, U. (1996). IdentiWcation and characterization of a cDNA and the structural gene encoding the mouse epithelial membrane protein-1, Genomics 36, 379–387. Lupski, J. R. (2000). Recessive Charcot-Marie-tooth disease, Ann Neurol 47, 6–8. Lupski, J. R., de Oca Luna, R. M., Slaugenhaupt, S., Pentao, L., Guzzetta, V., Trask, B. J., Saucedo-Cardenas, O., Barker, D. F., Killian, J. M., Garcia, C. A., et al. (1991). DNA duplication associated with Charcot-MarieTooth disease type 1A, Cell 66, 219–232. Magyar, J. P., Martini, R., Ruelicke, T., Aguzzi, A., Adlkofer, K., Dembic, Z., Zielasek, J., Toyka, K. V., and Suter, U. (1996). Impaired diVerentiation of Schwann cells in transgenic mice with increased PMP22 gene dosage, J Neurosci 16, 5351–5360. Maier, M., Berger, P., Nave, K. A., and Suter, U. (2002a). IdentiWcation of the regulatory region of the peripheral myelin protein 22 (PMP22) gene that directs temporal and spatial expression in development and regeneration of peripheral nerves, Mol Cell Neurosci 20, 93–109. Maier, M., Berger, P., and Suter, U. (2002b). Understanding Schwann cell-neurone interactions: The key to Charcot-Marie-Tooth disease?, J Anat 200, 357–366. ManWoletti, G., Ruaro, M. E., Del Sal, G., Philipson, L., and Schneider, C. (1990). A growth arrest-speciWc (gas) gene codes for a membrane protein, Mol Cell Biol 10, 2924–2930. Marques, W., Jr., Thomas, P. K., Sweeney, M. G., Carr, L., and Wood, N. W. (1998). Dejerine-Sottas neuropathy and PMP22 point mutations: A new base pair substitution and a possible ‘‘hot spot’’ on Ser72, Ann Neurol 43, 680–683. Matsunami, N., Smith, B., Ballard, L., Lensch, M. W., Robertson, M., Albertsen, H., Hanemann, C. O., Muller, H. W., Bird, T. D., White, R., and Chance, P. F. (1992). Peripheral myelin protein-22 gene maps in the duplication in chromosome 17p11.2 associated with Charcot-Marie-Tooth 1A., Nat Genet 1, 176–179. McLeod, J. G., and Low, P. A. (1979). Animal models of peripheral neuropathy: The Trembler mouse, Muscle Nerve 2, 495–502. Melcangi, R. C., Magnaghi, V., Galbiati, M., Ghelarducci, B., Sebastiani, L., and Martini, L. (2000). The action of steroid hormones on peripheral myelin proteins: A possible new tool for the rebuilding of myelin?, J Neurocytol 29, 327–339. Meuleman, J., Pou-Serradell, A., Lofgren, A., Ceuterick, C., Martin, J. J., Timmerman, V., Van Broeckhoven, C., and De Jonghe, P. (2001). A novel 3’-splice site mutation in peripheral myelin protein 22 causing hereditary neuropathy with liability to pressure palsies, Neuromuscul Disord 11, 400–403. Mostacciuolo, M. L., Righetti, E., Zortea, M., Bosello, V., Schiavon, F., Vallo, L., Merlini, L., Siciliano, G., Fabrizi, G. M., Rizzuto, N., et al. (2001). Charcot-Marie-Tooth disease type I and related demyelinating neuropathies: Mutation analysis in a large cohort of Italian families, Hum Mutat 18, 32–41.

III. THE MYELIN GENES AND PRODUCTS

561

CELLULAR APPROACHES TOWARD THE FUNCTION OF PMP22

Muller, H. W. (2000). Tetraspan myelin protein PMP22 and demyelinating peripheral neuropathies: New facts and hypotheses, Glia 29, 182–185. Naef, R., Adlkofer, K., Lescher, B., and Suter, U. (1997). Aberrant protein traYcking in Trembler suggests a disease mechanism for hereditary human peripheral neuropathies, Mol Cell Neurosci 9, 13–25. Naef, R., and Suter, U. (1998). Many facets of the peripheral myelin protein PMP22 in myelination and disease, Microsc Res Tech 41, 359–371. Naef, R., and Suter, U. (1999). Impaired intracellular traYcking is a common disease mechanism of PMP22 point mutations in peripheral neuropathies, Neurobiol Dis 6, 1–14. Nelis, E., Holmberg, B., Adolfsson, R., Holmgren, G., and van Broeckhoven, C. (1997). PMP22 Thr(118)Met: Recessive CMT1 mutation or polymorphism?, Nat Genet 15, 13–14. Nelis, E., Timmerman, V., De Jonghe, P., and Van Broeckhoven, C. (1994). IdentiWcation of a 5’ splice site mutation in the PMP-22 gene in autosomal dominant Charcot-Marie-Tooth disease type 1, Hum Mol Genet 3, 515–516. Nelis, E., Van Broeckhoven, C., and Consortium, C. (1996). Estimation of the mutation frequencies in CharcotMarie-Tooth disease type 1 and hereditary neuropathy with liability to pressure palsies: A european collaborative study, Eur J Hum Genet 4, 25–33. Neuberg, D. H., Sancho, S., and Suter, U. (1999). Altered molecular architecture of peripheral nerves in mice lacking the peripheral myelin protein 22 or connexin32, J Neurosci Res 58, 612–623. Nicholson, G. A., Valentijn, L. J., Cherryson, A. K., Kennerson, M. L., Bragg, T. L., DeKroon, R. M., Ross, D. A., Pollard, J. D., McLeod, J. G., Bolhuis, P. A., et al. (1994). A frame shift mutation in the PMP22 gene in hereditary neuropathy with liability to pressure palsies, Nat Genet 6, 263–266. Niemann, S., Sereda, M. W., Suter, U., GriYths, I. R., and Nave, K. A. (2000). Uncoupling of myelin assembly and schwann cell diVerentiation by transgenic overexpression of peripheral myelin protein 22, J Neurosci 20, 4120–4128. Norreel, J. C., Jamon, M., Riviere, G., Passage, E., Fontes, M., and Clarac, F. (2001). Behavioural proWling of a murine Charcot-Marie-Tooth disease type 1A model, Eur J Neurosci 13, 1625–1634. Notterpek, L., Ryan, M. C., Tobler, A. R., and Shooter, E. M. (1999a). PMP22 accumulation in aggresomes: Implications for CMT1A pathology, Neurobiol Dis 6, 450–460. Notterpek, L., Shooter, E. M., and Snipes, G. J. (1997). Upregulation of the endosomal-lysosomal pathway in the trembler-J neuropathy, J Neurosci 17, 4190–4200. Notterpek, L., Snipes, G. J., and Shooter, E. M. (1999b). Temporal expression pattern of peripheral myelin protein 22 during in vivo and in vitro myelination, Glia 25, 358–369. Notterpek, L., Roux, K. J., Amici, S. A., Yazdanpour, A., Rahner, C., and Fletcher, B. S. (2001). Peripheral myelin protein 22 is a constituent of intercellular junctions in epithelia, Proc Natl Acad Sci USA 98, 14404–14409. Numakura, C., Lin, C., Oka, N., Akiguchi, I., and Hayasaka, K. (2000). Hemizygous mutation of the peripheral myelin protein 22 gene associated with Charcot-Marie-Tooth disease type 1, Ann Neurol 47, 101–103. Ohnishi, A., Yoshimura, T., Kanehisa, Y., and Fukushima, Y. (1995). A case of hereditary motor and sensory neuropathy type I with a new type of peripheral myelin protein (PMP)-22 mutation, Rinsho Shinkeigaku 35, 788–792. Paratore, P., Hagedorn, L., Floris, J., Hari, L., Kleber, M., Suter, U., and Sommer, L. (2002). Cell-intrinsic and cell-extrinsic cues regulating lineage decisions in multipotent neural crest-derived progenitor cells, Int J Dev Biol 46, 193–200. Pareek, S., Notterpek, L., Snipes, G. J., Naef, R., Sossin, W., Laliberte, J., Iacampo, S., Suter, U., Shooter, E. M., and Murphy, R. A. (1997). Neurons promote the translocation of peripheral myelin protein 22 into myelin, J Neurosci 17, 7754–7762. Pareek, S., Suter, U., Snipes, G. J., Welcher, A. A., Shooter, E. M., and Murphy, R. A. (1993). Detection and processing of peripheral myelin protein PMP22 in cultured Schwann cells, J Biol Chem 268, 10372–10379. Pareyson, D., and Taroni, F. (1996). Deletion of the PMP22 gene and hereditary neuropathy with liability to pressure palsies, Curr Opin Neurol 9, 348–354. Parman, Y., Plante-Bordeneuve, V., Guiochon-Mantel, A., Eraksoy, M., and Said, G. (1999). Recessive inheritance of a new point mutation of the PMP22 gene in Dejerine-Sottas disease, Ann Neurol 45, 518–522. Parmantier, E., Braun, C., Thomas, J. L., Peyron, F., Martinez, S., and Zalc, B. (1997). PMP-22 expression in the central nervous system of the embryonic mouse deWnes potential transverse segments and longitudinal columns, J Comp Neurol 378, 159–172. Parmantier, E., Cabon, F., Braun, C., D’Urso, D., Muller, H. W., and Zalc, B. (1995). Peripheral myelin protein22 is expressed in rat and mouse brain and spinal cord motoneurons, Eur J Neurosci 7, 1080–1088. Patel, P. I., Roa, B. B., Welcher, A. A., Schoener-Scott, R., Trask, B. J., Pentao, L., Snipes, G. J., Garcia, C. A., Francke, U., Shooter, E. M., et al. (1992). The gene for the peripheral myelin protein PMP-22 is a candidate for Charcot-Marie-Tooth disease type 1A, Nat Genet 1, 159–165. Pentao, L., Wise, C. A., Chinault, A. C., Patel, P. I., and Lupski, J. R. (1992). Charcot-Marie-Tooth type 1A duplication appears to arise from recombination at repeat sequences Xanking the 1.5 Mb monomer unit, Nat Genet 2, 292–300. Perea, J., Robertson, A., Tolmachova, T., Muddle, J., King, R. H., Ponsford, S., Thomas, P. K., and Huxley, C. (2001). Induced myelination and demyelination in a conditional mouse model of Charcot-Marie-Tooth disease type 1A, Hum Mol Genet 10, 1007–1018.

III. THE MYELIN GENES AND PRODUCTS

au1

562

21. PMP 22 GENE

Raeymaekers, P., Timmerman, V., De Jonghe, P., Swerts, L., Gheuens, J., Martin, J. J., Muylle, L., De Winter, G., Vandenberghe, A., and Van Broeckhoven, C. (1989). Localization of the mutation in an extended family with Charcot-Marie-Tooth neuropathy (HMSN I), Am J Hum Genet 45, 953–958. Raeymaekers, P., Timmerman, V., Nelis, E., De Jonghe, P., Hoogendijk, J. E., Baas, F., Barker, D. F., Martin, J. J., De Visser, M., Bolhuis, P. A., and Van Broeckhoven, C. (1991). Duplication in chromosome 17p11.2 in Charcot-Marie-Tooth neuropathy type 1a (CMT 1a). The HMSN Collaborative Research Group, Neuromuscul Disord 1, 93–97. Reiter, L. T., Murakami, T., Koeuth, T., Pentao, L., Muzny, D. M., Gibbs, R. A., and Lupski, J. R. (1996). A recombination hotspot responsible for two inherited peripheral neuropathies is located near a mariner transposon-like element, Nat Genet 12, 288–297. Roa, B. B., Garcia, C. A., Pentao, L., Killian, J. M., Trask, B. J., Suter, U., Snipes, G. J., Ortiz-Lopez, R., Shooter, E. M., Patel, P. I., et al. (1993a). Evidence for a recessive PMP22 point mutation in Charcot-MarieTooth disease type 1A, Nat Genet 5, 189–194. Roa, B. B., Garcia, C. A., Suter, U., Kulpa, D. A., Wise, C. A., Mueller, J., Welcher, A. A., Snipes, G. J., Shooter, E. M., Patel, P. I., et al. (1993b). Charcot-Marie-Tooth disease type 1A. Association with a spontaneous point mutation in the PMP22 gene, N Engl J Med 329, 96–101. Robertson, A., Perea, J., Tolmachova, T., Thomas, P. K., and Huxley, C. (2002). EVects of mouse strain, position of integration and tetracycline analogue on the tetracycline conditional system in transgenic mice, Gene 282, 65–74. Ryan, M. C., Notterpek, L., Tobler, A. R., Liu, N., and Shooter, E. M. (2000). Role of the peripheral myelin protein 22 N-linked glycan in oligomer stability, J Neurochem 75, 1465–1474. Ryan, M. C., Shooter, E. M., and Notterpek, L. (2002). Aggresome formation in neuropathy models based on peripheral myelin protein 22 mutations, Neurobiol Dis 10, 109–118. Saberan-Djoneidi, D., Sanguedolce, V., Assouline, Z., Levy, N., Passage, E., and Fontes, M. (2000). Molecular dissection of the Schwann cell speciWc promoter of the PMP22 gene, Gene 248, 223–231. Sahenk, Z., Chen, L., and Freimer, M. (1998). A novel PMP22 point mutation causing HNPP phenotype: Studies on nerve xenografts, Neurology 51, 702–707. Sancho, S., Magyar, J. P., Aguzzi, A., and Suter, U. (1999). Distal axonopathy in peripheral nerves of PMP22mutant mice, Brain 122, 1563–1577. Sancho, S., Young, P., and Suter, U. (2001). Regulation of Schwann cell proliferation and apoptosis in PMP22deWcient mice and mouse models of Charcot-Marie-Tooth disease type 1A, Brain 124, 2177–2187. Sanders, C. R., Ismail-Beigi, F., and McEnery, M. W. (2001). Mutations of peripheral myelin protein 22 result in defective traYcking through mechanisms which may be common to diseases involving tetraspan membrane proteins, Biochemistry 40, 9453–9459. Schachner, M., and Martini, R. (1995). Glycans and the modulation of neural-recognition molecule function, Trends Neurosci. 18, 183–191. Schenone, A., Nobbio, L., Caponnetto, C., Abbruzzese, M., Mandich, P., Bellone, E., Ajmar, F., Gherardi, G., Windebank, A. J., and Mancardi, G. (1997a). Correlation between PMP-22 messenger RNA expression and phenotype in hereditary neuropathy with liability to pressure palsies, Ann Neurol 42, 866–872. Schenone, A., Nobbio, L., Mandich, P., Bellone, E., Abbruzzese, M., Aymar, F., Mancardi, G. L., and Windebank, A. J. (1997b). Underexpression of messenger RNA for peripheral myelin protein 22 in hereditary neuropathy with liability to pressure palsies, Neurology 48, 445–449. Scherer, S. S. (1997). The biology and pathobiology of Schwann cells, Curr Opin Neurol 10, 386–397. Schneider, C., King, R. M., and Philipson, L. (1988). Genes speciWcally expressed at growth arrest of mammalian cells, Cell 54, 787–793. Sedzik, J., Kotake, Y., and Uyemura, K. (1998). PuriWcation of PASII/PMP22–an extremely hydrophobic glycoprotein of PNS myelin membrane, Neuroreport 9, 1595–1600. Sedzik, J., and Tsukihara, T. (2000). Solubilization of PNS myelin membrane proteins by detergents, Neuroreport 11, 2559–2563. Seeman, P., Mazanec, R., Marikova, T., and Rautenstrauss, B. (1999). Charcot-Marie-Tooth 1A: Heterozygous T118M mutation over a CMT1A duplication has no inXuence on the phenotype, Ann NY Acad Sci 883, 485–489. Sereda, M., GriYths, I., Puhlhofer, A., Stewart, H., Rossner, M. J., Zimmerman, F., Magyar, J. P., Schneider, A., Hund, E., Meinck, H. M., et al. (1996). A transgenic rat model of Charcot-Marie-Tooth disease, Neuron 16, 1049–1060. Shapiro, L., Doyle, J. P., Hensley, P., Colman, D. R., and Hendrickson, W. A. (1996). Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin, Neuron 17, 435–449. Shugart, E. C., Levenson, A. S., Constance, C. M., and Umek, R. M. (1995). DiVerential expression of gas and gadd genes at distinct growth arrest points during adipocyte development, Cell Growth DiVer 6, 1541–1547. Simonati, A., Fabrizi, G. M., Pasquinelli, A., Taioli, F., Cavallaro, T., Morbin, M., Marcon, G., Papini, M., and Rizzuto, N. (1999). Congenital hypomyelination neuropathy with Ser72Leu substitution in PMP22, Neuromuscul Disord 9, 257–261. Snipes, G. J., and Suter, U. (1995a). Molecular anatomy and genetics of myelin proteins in the peripheral nervous system, J Anat 186, 483–494. Snipes, G. J., and Suter, U. (1995b). Molecular basis of common hereditary motor and sensory neuropathies in humans and in mouse models, Brain Pathol 5, 233–247.

III. THE MYELIN GENES AND PRODUCTS

CELLULAR APPROACHES TOWARD THE FUNCTION OF PMP22

Snipes, G. J., Suter, U., and Shooter, E. M. (1993). Human peripheral myelin protein-22 carries the L2/HNK-1 carbohydrate adhesion epitope, J Neurochem 61, 1961–1964. Snipes, G. J., Suter, U., Welcher, A. A., and Shooter, E. M. (1992). Characterization of a novel peripheral nervous system myelin protein (PMP-22/SR13), J Cell Biol 117, 225–238. Spreyer, P., Kuhn, G., Hanemann, C. O., Gillen, C., Schaal, H., Kuhn, R., Lemke, G., and Muller, H. W. (1991). Axon-regulated expression of a Schwann cell transcript that is homologous to a ‘growth arrest-speciWc’ gene, Embo J 10, 3661–3668. StoVel, W., and Bosio, A. (1997). Myelin glycolipids and their functions, Curr Opin Neurobiol 7, 654–661. Suter, U. (1998). Charcot-Marie-Tooth disease. Pathology, genetics and animal models. In ‘‘Mouse Models of Human Neurological Diseases’’ (B. Popko, ed.). Plenum Press, New York. Suter, U., Moskow, J. J., Welcher, A. A., Snipes, G. J., Kosaras, B., Sidman, R. L., Buchberg, A. M., and Shooter, E. M. (1992a). A leucine-to-proline mutation in the putative Wrst transmembrane domain of the 22-kDa peripheral myelin protein in the trembler-J mouse, Proc Natl Acad Sci USA 89, 4382–4386. Suter, U., and Patel, P. I. (1994). Genetic basis of inherited peripheral neuropathies, Hum Mutat 3, 95–102. Suter, U., and Snipes, G. J. (1995). Biology and genetics of hereditary motor and sensory neuropathies., Ann Rev Neurosci 18, 45–75. Suter, U., Snipes, G. J., Schoener-Scott, R., Welcher, A. A., Pareek, S., Lupski, J. R., Murphy, R. A., Shooter, E. M., and Patel, P. I. (1994). Regulation of tissue-speciWc expression of alternative peripheral myelin protein-22 (PMP22) gene transcripts by two promoters, J Biol Chem 269, 25795–25808. Suter, U., Welcher, A. A., Ozcelik, T., Snipes, G. J., Kosaras, B., Francke, U., Billings-Gagliardi, S., Sidman, R. L., and Shooter, E. M. (1992b). Trembler mouse carries a point mutation in a myelin gene, Nature 356, 241–244. Suter, U., Welcher, A. A., and Snipes, G. J. (1993). Progress in the molecular understanding of hereditary peripheral neuropathies reveals new insights into the biology of the peripheral nervous system, Trends Neurosci 16, 50–56. Takeda, Y., Notsu, T., Kitamura, K., and Uyemura, K. (2001). Functional analysis for peripheral myelin protein PASII/PMP22: Is it a member of claudin superfamily?, Neurochem Res 26, 599–607. Taylor, V., Welcher, A. A., Program, A. E., and Suter, U. (1995). Epithelial membrane protein-1, peripheral myelin protein 22, and lens membrane protein 20 deWne a novel gene family, J Biol Chem 270, 28824–28833. Taylor, V., Zgraggen, C., Naef, R., and Suter, U. (2000). Membrane topology of peripheral myelin protein 22, J Neurosci Res 62, 15–27. Thomas, P. K., Marques, W., Jr., Davis, M. B., Sweeney, M. G., King, R. H., Bradley, J. L., Muddle, J. R., Tyson, J., Malcolm, S., and Harding, A. E. (1997). The phenotypic manifestations of chromosome 17p11.2 duplication, Brain 120, 465–478. Timmerman, V., Nelis, E., Van Hul, W., Nieuwenhuijsen, B. W., Chen, K. L., Wang, S., Ben Othman, K., Cullen, B., Leach, R. J., Hanemann, C. O., et al. (1992). The peripheral myelin protein gene PMP-22 is contained within the Charcot-Marie-Tooth disease type 1A duplication, Nat Genet 1, 171–175. Tiwari-WoodruV, S. K., Buznikov, A. G., Vu, T. Q., Micevych, P. E., Chen, K., Kornblum, H. I., and Bronstein, J. M. (2001). OSP/claudin-11 forms a complex with a novel member of the tetraspanin super family and beta1 integrin and regulates proliferation and migration of oligodendrocytes, J Cell Biol 153, 295–305. Tobler, A. R., Liu, N., Mueller, L., and Shooter, E. M. (2002). DiVerential aggregation of the Trembler and Trembler J mutants of peripheral myelin protein 22, Proc Natl Acad Sci USA 99, 483–488. Tobler, A. R., Notterpek, L., Naef, R., Taylor, V., Suter, U., and Shooter, E. M. (1999). Transport of Trembler-J mutant peripheral myelin protein 22 is blocked in the intermediate compartment and aVects the transport of the wild-type protein by direct interaction, J Neurosci 19, 2027–2036. Tsukita, S., Furuse, M., and Itoh, M. (2001). Multifunctional strands in tight junctions, Nat Rev Mol Cell Biol 2, 285–293. Tyson, J., Ellis, D., Fairbrother, U., King, R. H., Muntoni, F., Jacobs, J., Malcolm, S., Harding, A. E., and Thomas, P. K. (1997). Hereditary demyelinating neuropathy of infancy. A genetically complex syndrome, Brain 120, 47–63. Valentijn, L. J., Baas, F., Wolterman, R. A., Hoogendijk, J. E., van den Bosch, N. H., Zorn, I., Gabreels-Festen, A. W., de Visser, M., and Bolhuis, P. A. (1992a). Identical point mutations of PMP-22 in Trembler-J mouse and Charcot-Marie-Tooth disease type 1A, Nat Genet 2, 288–291. Valentijn, L. J., Bolhuis, P. A., Zorn, I., Hoogendijk, J. E., van den Bosch, N., Hensels, G. W., Stanton Jr., V. P., Housman, D. E., Fischbeck, K. H., Ross, D. A., et al. (1992b). The peripheral myelin gene PMP-22/GAS-3 is duplicated in Charcot-Marie-Tooth disease type 1A, Nat Genet 1, 166–170. Vallat, J. M., Sindou, P., Preux, P. M., Tabaraud, F., Milor, A. M., Couratier, P., LeGuern, E., and Brice, A. (1996). Ultrastructural PMP22 expression in inherited demyelinating neuropathies, Ann. Neurol. 39, 813–817. Vance, J. M., Nicholson, G. A., Yamaoka, L. H., Stajich, J., Stewart, C. S., Speer, M. C., Hung, W. Y., Roses, A. D., Barker, D., and Pericak-Vance, M. A. (1989). Linkage of Charcot-Marie-Tooth neuropathy type 1a to chromosome 17, Exp Neurol 104, 186–189. Wang, C. X., Wadehra, M., Fisk, B. C., Goodglick, L., and Braun, J. (2001). Epithelial membrane protein 2, a 4-transmembrane protein that suppresses B-cell lymphoma tumorigenicity, Blood 97, 3890–3895. Welcher, A. A., Suter, U., De Leon, M., Snipes, G. J., and Shooter, E. M. (1991). A myelin protein is encoded by the homologue of a growth arrest-speciWc gene, Proc Natl Acad Sci USA. 88, 7195–7199. Wilson, H. L., Wilson, S. A., Surprenant, A., and North, R. A. (2002). Epithelial membrane proteins induce membrane blebbing and interact with the P2X7 receptor C-terminus, J Biol Chem 277, 34017–34023.

III. THE MYELIN GENES AND PRODUCTS

563

564

21. PMP 22 GENE

Windebank, A. J. (1993). Inherited recurrent neuropathy. In ‘‘Peripheral Neuropathy’’ (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low, and J. F. Poduslo, eds.), pp. 1094–1136. WB Saunders, Philadelphia. Wise, C. A., Garcia, C. A., Davis, S. N., Heju, Z., Pentao, L., Patel, P. I., and Lupski, J. R. (1993). Molecular analysis of unrelated Charcot-Marie-Tooth (CMT) disease patients suggest a high frequency of the CMT1A duplication, Am J Hum Genet 53, 853–863. Wulf, P., Bernhardt, R. R., and Suter, U. (1999). Characterization of peripheral myelin protein 22 in zebraWsh (zPMP22) suggests an early role in the development of the peripheral nervous system, J Neurosci Res 57, 467–478. Yener, G. G., Guiochon-Mantel, A., Obuz, F., Baklan, B., Ozturk, V., Kovanlikaya, I., Cakmur, R., and Genc, A. (2001). Phe 84 deletion of the PMP22 gene associated with hereditary motor and sensory neuropathy HMSN III with multiple cranial neuropathy: Clinical, neurophysiological and magnetic resonance imaging Wndings, J Neurol 248, 193–196. Yoshikawa, H., Nishimura, T., Nakatsuji, Y., Fujimura, H., Himoro, M., Hayasaka, K., Sakoda, S., and Yanagihara, T. (1994). Elevated expression of messenger RNA for peripheral myelin protein 22 in biopsied peripheral nerves of patients with Charcot-Marie-Tooth disease type 1A, Ann Neurol 35, 445–450. Young, P., Stogbauer, F., Eller, B., de Jonghe, P., Lofgren, A., Timmerman, V., Rautenstrauss, B., Oexle, K., Grehl, H., Kuhlenbaumer, G., et al. (2000). PMP22 Thr118Met is not a clinically relevant CMT1 marker, J Neurol 247, 696–700. Young, P., Wiebusch, H., Stogbauer, F., Ringelstein, B., Assmann, G., and Funke, H. (1997). A novel frameshift mutation in PMP22 accounts for hereditary neuropathy with liability to pressure palsies, Neurology 48, 450–452. Zoidl, G., Blass-Kampmann, S., D’Urso, D., Schmalenbach, C., and Muller, H. W. (1995). Retroviral-mediated gene transfer of the peripheral myelin protein PMP22 in Schwann cells: Modulation of cell growth, Embo J 14, 1122–1128.

III. THE MYELIN GENES AND PRODUCTS

C H A P T E R

22 The Claudin 11 Gene Alexander Gow

INTRODUCTION Despite a relatively simple protein composition of the central nervous system (CNS) myelin sheath, where two proteins comprise as much as 80% of total protein (Lees and BrostoV, 1984), identiWcation of the cellular and molecular mechanisms governing synthesis, assembly, maintenance, and destruction of this membrane has proven to be a complex and protracted process. In large part, the diYculties associated with dissecting the form and function of the myelin sheath stem from inadequate biophysical and biochemical techniques to deal with the compact multilamellar structure of myelin membrane. Even the recent application of powerful genetic approaches such as transgenic and homologous recombination technologies have frequently yielded mere glimpses of myelin biology and have highlighted a number of surprisingly complex features of myelin. For example, ablation of the Plp1 gene, from which roughly 50% of myelin protein is derived, confers almost imperceptible behavioral and morphological phenotypes in mice until adulthood when the subtle consequences of long-term myelin instability and defective axon-oligodendrocyte communication become apparent (Boison and StoVel, 1994; Klugmann et al., 1997; Stecca et al., 2000). Indeed, nearly a decade after generation of the Wrst of three Plp1-null mouse strains in diVerent laboratories, the functions of PLP1 gene products remain as obscure as the underlying pathogenesis in the null mutants (Garbern et al., 2002; GriYths et al., 1998). Perhaps equally perplexing is the function of myelin basic protein (MBP). The gene encoding this protein is functionally null in two naturally occurring mouse mutants, shi and mld, and it is clear from the behavioral and CNS morphological phenotypes that myelin membrane compaction is critically dependent on MBP (reviewed by Mikoshiba et al., 1992). However, molecular details of the interactions between MBP and other lipid and/or protein components at the major dense line are completely unknown. The best indication from in vitro systems of MBP function is that this protein precipitates on the negatively charged cytoplasmic surfaces of the myelin bilayer and glues the membrane surfaces together (Boggs et al., 2000). Ablation of a third gene encoding the myelinoligodendrocyte glycoprotein (MOG) has yielded little or no insight into function, in similar fashion to over-expression of this gene in oligodendrocytes or ectopically in many diVerent tissues (unpublished observations, Jaquet et al., 1996). A fourth gene, for which ablation in mice has to date yielded only glimpses of the normal function in CNS myelin, encodes the oligodendrocyte speciWc protein (OSP). Importantly, the function of this protein in the inner ear and testis is quite clear (see the section titled ‘‘Functions of Claudin 11 Tight Junctions in Non-CNS Tissue’’). Initial assessment of function from amino acid similarities with Gas-3/PMP-22 suggested that OSP might regulate oligodendrocyte proliferation and diVerentiation (Bronstein et al., 1996). Moreover, biochemical immunoprecipitation experiments indicate that OSP interacts with the tetra-

Myelin Biology and Disorders, Volume 1

565

Copyright 2004, Elsevier (USA). All rights reserved.

566

22. THE CLAUDIN 11 GENE

spanin protein, OAP-1, and b1-integrin. These data suggest a role for OSP in oligodendrocyte migration and are further supported by cell culture experiments in which migration on a Wbronectin substrate is attenuated for primary oligodendrocyte progenitors from Osp-null mice compared to controls (Bronstein et al., 2000b; Tiwari-WoodruV et al., 2001). Finally, recent amino acid alignment analyses indicate that OSP is a member of the claudin family of integral membrane tight junction proteins, designated claudin 11, which suggests a potentially important contribution to the electrophysiological properties of the myelin sheath.

PROTEINS OF TIGHT JUNCTIONS In most polarized epithelia, the molecular architecture of tight junctions is that of an extensive complex that includes the physical paracellular barrier of transmembrane proteins and a cytoplasmic scaVold or plaque that is frequently attached to the cytoskeleton. Evidence to date suggests that the PDZ-domain zonula occludens proteins, ZO-1, ZO-2, and ZO-3, may constitute the principal components of the cytoplasmic scaVold because of their abundance and modular design that accommodates interactions with the integral membrane tight junction proteins, occludin and claudin family members, as well as binding to actin and myosin Wlaments. Nonetheless, more than a dozen components of the cytoplasmic plaque have been identiWed, including cingulin and VAP-33, but little is known about the functions of most of these proteins at the tight junction (reviewed by Lapierre et al., 1999; Mitic and Anderson, 1998).

THE CLAUDIN FAMILY OF TIGHT JUNCTION PROTEINS The Wrst transmembrane component of tight junctions to be identiWed was occludin, which is a constituent of the junctions in most cell types, can assemble into intramembranous Wbrils in transfected Wbroblasts, and is phosphorylated under conditions that regulate tight junction assembly (Furuse et al., 1993; Wong, 1997). However, doubts about the likelihood of a single transmembrane protein accounting for the diverse morphological and electrophysiological properties observed for tight junctions from diVerent tissues were conWrmed by ablation of the mouse Occludin gene by homologous recombination in embryonic stem cells (Saitou et al., 1998). Occludin-null mice exhibit phenotypes associated with most polarized cell types, but tight junctions in these animals are, in general, functional and morphologically normal (Saitou et al., 2000). An important exception is the Sertoli cell junctions in the seminiferous tubules of the testis. Null mutants are sterile and exhibit similar pathophysiology as Claudin 11-null males (Gow et al., 1999). A reevaluation of tight junction-enriched subcellular fractions from liver turned up another transmembrane protein, called claudin 1, that shared limited homology with a second protein and both of these components were shown to assemble into intramembranous Wbrils in transfected Wbroblasts (Furuse et al., 1998). Subsequent database homology searches revealed a large family of claudins with varying tissue distributions and antibodies raised against many of these proteins demonstrate their localization to tight junctions in vitro and in vivo (Gow et al., 1999; Morita et al., 1999a; Simon et al., 1999; Wilcox et al., 2001). At present, the size of the CLAUDIN family stands at 20 genes in mice and humans and the chromosomal localization of most of these genes has been determined for one or both species (Tab. 22.1). The CLAUDIN family can be divided into two groups on the basis of the presence or absence of introns and a Clustal analysis of the encoded proteins shows that intronless genes are more closely related to each other than to intron-containing genes (Southwood and Gow, 2001). According to this analysis, the archetypal claudin gene appears to have contained introns and subsequent gene duplication gave rise to at least six additional genes before the emergence and expansion of the intronless gene group. Evidence for these evolutionary relationships is found on human chromosomes 3, 7, 16, and 18 where gene clusters comprise only intron-containing or intronless CLAUDIN genes.

567

CLAUDIN 11 TIGHT JUNCTIONS

TABLE 22.1 Genomic Structure data of mouse and human CLAUDIN genes. Claudin gene

Species

Number of nucleotide triplets per coding exon

Number of Exons 1st

1a

Hu

4

74

2

Hu/Ms

1

230

3d

Hu/Ms

1

220

4d

Hu/Ms

1

5

1/3

1/3

1/3

3q28-29

209/210

AC023010

7q11.23

AC000071

1

218

1

220

7

1/3

53

5th

Xq22.3-23

Hu/Ms

28

4th

AC009520

Hu/Ms

55

3rd

Human chromosome

AL158821

6

b

2nd

Human gene accession #

7q11.23

AC004643 55

28

1/3

53

1/3

AC003688

22q11.2 16p13.3 17

Hu

5

74

c

8

Hu/Ms

1

225

AP001846

18q22

9b

Hu/Ms

1

217

AC004643

16p13.3

10

Hu

5

73

1/3

54

27

1/3

11

Hu/Ms

3

75

1/3

55

76

2/3

12

Hu/Ms

1

244

13d

Ms

1

211

14

Hu/Ms

1

239

15

Hu

5

70

1/3

a

Hu

5

108

17

c

Hu/Ms

1

224

18e

Hu

5

73

1/3 1/3

16

19

Hu

4

74

20

Hu

1

219

371/3

36

AL139376

13

AC008041

3q26.2-26.3

AC006153

7q2

AC000005

21

AC006329

7

7q11.23? 55 34 55 55

27 1/3

1/3

55 39 1/3 28

1/3

39

34 1/3

64

2/3

43

56 1/3

37 53

1/3

AC009520

3q28-29

AP001846

18q22

AC016252

3

AL136383

1

AL139101

6

a,b,c

Several human genes are located in close proximity to each other. Five mouse claudin genes, including Claudins 3, 4, 13, and two novel claudins, are in close proximity to each other and map to a single contig from mouse chromosome 5 (Genebank Acc# AC079938). Currently, it is not known if the genomic organization is similar at human chromosome 7q11.23, where CLAUDIN 3 and CLAUDIN 4 lie in close proximity; however, mouse chromosome 5 and human chromosome 7q11 are largely syntenic. e An open reading frame containing 73 1/3 codons and a canonical splice donor site lies upstream of the CLAUDIN 18 mRNA published in Genbank, suggesting that this gene has two promoters. However, ESTs containing the upstream exon have not been found. Hu, human; Ms, mouse. d

CLAUDIN 11 TIGHT JUNCTIONS Claudin 11 was initially isolated as an oligodendrocyte-speciWc protein (OSP) in a diVerential display of dorsal versus ventral spinal cord from mice that was designed to identify genes involved in myelinogenesis (Bronstein et al., 1996). Claudin 11 is targeted to oligodendrocyte processes shortly after they contact axons and persists in mature myelin sheaths as a relatively abundant protein comprising approximately 7% of total protein in rodents (Bronstein et al., 1997). However, claudin 11 expression is not limited to the central nervous system and is expressed broadly during development as well as in several adult tissues (Gow et al., 1999; Morita et al., 1999b). During embryonic development in rodents, claudin 11 is detected in the neural tube as early as E8.5 and becomes more widely expressed by E14.5 when it is detected in foregut, epaxial muscles, mesonephric ducts, urogenital tract, bone primordia, nasal epithelium, hair follicles of the vibrissae, perioptic mesenchyme, lacrimal glands, leptomeninges around the brain and neural tube and membranous labyrinth of the inner ear (Bronstein et al., 2000a; Gow et al., 1999). Postnatally, claudin 11 is expressed in testis, leptomeninges, choroid plexis, CNS white matter, and inner ear (Bronstein et al., 1997; Bronstein et al., 1996; Gow et al., 1999; Morita et al., 1999b; Wolburg et al., 2001). A satisfying aspect of claudin 11 tight junction biology, at least for three tissues that we have examined (Gow et al., 1999; B. Kachar, unpublished data), is revealed by freeze

568

22. THE CLAUDIN 11 GENE

au1

fracture electron microscopy where intramembranous particles of similar morphology are found in diVerent tissues. The parallel, rarely anastomosing morphology of claudin 11 intramembranous Wbrils in the Stria vascularis of the inner ear (Fig. 22.1A), Sertoli cells in the testis (Fig. 22.1B) and myelin sheaths in the CNS (Fig. 22.1C) are clearly visible (arrowheads) and may be interspersed with gap junctions (black arrows, Fig. 22.1A). These data generate the expectation that tight junctions in other tissues that exhibit similar morphology will be largely comprised of claudin 11.

FIGURE 22.1 Freeze fracture morphology of claudin 11 intramembranous Wbrils The morphology of claudin 11 in the bilayer is that of parallel, rarely anastomosing intramembranous Wbrils and is similar in diVerent tissues. a. Claudin 11 tight junctions between basal cells of the Stria vascularis in the lateral wall of the cochlear duct. Arrowheads show several parallel tight junction strands that are interspersed with connexons (black arrows). These gap junctions are likely comprised of connexin 26 and represent part of the potassium recycling machinery in the cochlea. b. Claudin 11 tight junctions between Sertoli cells in the seminiferous tubules of the testis. The organization of these junctions is quite similar to those in the cochlea (arrowheads). c. A myelinated axon (lower, lower left to upper right) showing claudin 11 tight junctions coursing obliquely to the long axis of the fiber (arrowheads). At high magnification, these fibrils are resolved into parallel rows of intramembranous particles (inset). Cross-section of the axon in the center of the myelinated fiber is apparent, as is the multilamellar structure of the myelin sheath (white arrows).

CLAUDIN 11 TIGHT JUNCTIONS

Functions of Claudin 11 Tight Junctions in Non-CNS Tissue Despite the initial morphological characterization of CNS myelin tight junctions more than three decades ago, relatively little is known about the function of these structures. In contrast, claudin 11 tight junctions in the cochlea and testis have been characterized in considerable detail and consideration of data from these studies reveals important characteristics that may pertain to CNS myelin tight junction function. Claudin 11 Junctions Are Required for the Generation of Endocochlear Potential Claudin 11 is widely expressed in both the vestibular and auditory portions of the inner ear and is delineated by wholemount X-gal staining of tissue from Claudin 11-null mice in which the lacZ coding region substitutes for the Claudin 11 gene (Gow et al., 1999). In Figure 22.2A, X-gal staining is present in the epithelium that lines the bony labyrinth of all three semicircular canals. In addition, X-gal also stains the lateral wall of the spiral cochlear duct. Transverse sections through the cochlea reveal that b-galactosidase expression is limited to a narrow epithelial layer in the lateral wall that runs the length of the cochlear duct known as the Stria vascularis (Fig. 22.2B). The Stria vascularis is composed of two polarized cell layers, known as the marginal and basal cell layers, that are separated by a mesodermal cell layer, the intermediate cells, which are derived from melanocytes (Fig. 22.2C). Although the auditory apparatus is formed during embryogenesis, many mammals are born deaf. In rodents, hearing begins in the second postnatal week and coincides with the establishment of the claudin 11 tight junction network in the Stria vascularis (Souter and Forge, 1998). A major function of this epithelium is secretion into the cochlear duct of endolymph that baths the Organ of Corti, which is the mechanosensory apparatus of the cochlea. Hair cells within the Organ of Corti transduce sound vibrations into focal ion currents that depolarize neurons along the central axis of the cochlea for signal transmission to the brain stem. The ion composition of the endolymph is unusual for an extracellular Xuid in that it contains 150 mM Kþ and 1-2 mM Naþ (Salt and Konishi, 1986). In contrast, most extracellular Xuids have ion compositions that are similar to that in blood with 4 mM Kþ and close to 150 mM Naþ. Importantly, vibrational deXection of the basilar membrane of the Organ of Corti distorts the hair cells and opens mechanosensory Kþ channels in the apical surface. The Kþ in the endolymph Xows into the hair cells down a concentration gradient and through channels in the basolateral surface of these cells where it depolarizes local nerve terminals. A second critical function of the Stria vascularis is maintenance of an 80 to 100 mV endocochlear potential in the endolymph that is generated as a consequence of Kþ recycling via gap junction networks in the support cells that surround the Organ of Corti and the Stria vascularis (Kikuchi et al., 2000; Salt et al., 1987). Thus, extracellular Kþ in the vicinity of the basolateral surfaces of hair cells is sequestered by surrounding cells and Xows along a gap junction network toward the spiral ligament in the lateral wall of the cochlear duct. The basal and mesodermal intermediate cells of the Stria vascularis are the Wnal components of this gap junction network, and Kþ exits into the extracellular Xuid of the intermediate compartment in the vicinity of the basolateral surfaces of the marginal cells. As shown in the schematic in Figure 22.2C, these cells form extensive infoldings in their basolateral surfaces which serve as a large surface area for rapid uptake of Kþ through Naþ/Kþ pumps in these membranes. Furthermore, these infoldings contain a high density of mitochondria, which presumably generate the large amount of ATP required to operate the ion pumps. Thereafter, the Kþ Xows down its concentration gradient to exit marginal cells through the apical surface into the endolymph. Early electrophysiological studies in the cochlea used voltage- and Kþ-sensitive electrodes to map the origin of endocochlear potential across the Stria vascularis (Salt et al., 1987). These elegant experiments clearly demonstrate an endocochlear potential within the intermediate compartment that must be generated by basal and intermediate cells. Other studies have used small tracer molecules injected into cells that line the cochlear duct to demonstrate gap junction networks between cells in the Organ of Corti and the Stria

569

570

22. THE CLAUDIN 11 GENE

FIGURE 22.2 Expression of claudin 11 in the inner ear a. The Claudin 11 gene was ablated in mice by replacing the coding region in exon 1 with the open reading frame of bacterial b-galactosidase (Gow et al., 1999). X-gal histochemistry shows the expression of this enzyme in the inner ear throughout the semicircular canals and a wide stripe in the spiral ligament on the lateral wall of the cochlear duct. The narrow spiral stripe of staining in the cochlea reflects b-galactosidase expression by cells adjacent to Reissner’s membrane (lower), b-galactosidase staining of oligodendrocytes in the CNS component of the VIIIth cranial nerve. b. Cross-section of the inner ear shows the organization of the cochlear duct with the lateral wall to the left and the spiral ganglion out of view to the right. The basal epithelial layer of the Stria vascularis is stained by the X-gal reaction product. c. The schematic depicts a cross-section through the Stria vascularis and shows the relationships between cells comprising the three layers of this organ. The positions of tight junctions comprised of either claudin 1 or claudin 11 are shown. The surfaces of basal and marginal cells are in close proximity which may serve in the ecient recycling of potassium ions between the two cell types. The infoldings of the marginal cell basolateral membranes are densely packed with mitochondria (lower) which likely supplies ATP required to drive sodium/potassium ATPase pumps in this region of the cell.

vascularis (Takeuchi and Ando, 1998). Thus, together these data provide strong evidence for a model of Kþ recycling into the endolymph, which predicts that basal cell tight junctions form the principal permeability barrier necessary for generating an endocochlear potential (Kikuchi et al., 2000). Indeed, electrophysiological data from Claudin 11-null mice conWrms a major prediction from the Wve compartment model of the Stria vascularis,

CLAUDIN 11 TIGHT JUNCTIONS

that basal cell tight junctions are essential for generating an endocochlear potential. Auditory brain stem response and DPOAE measurements demonstrate that null mutants are deaf from an early age and that the endocochlear potential generated by these animals is less than 50% of littermate controls. Claudin 11 Junctions Constitute the Blood-Testis Barrier In many mammalian species, postnatal formation of male sex organs occurs in temporally regulated waves of proliferation, migration, and diVerentiation. For example, cells within the sex cords of mice at birth are primarily composed of two cell types: proliferating Sertoli cells in contact with the basement membrane and a small population of centrally located dormant germ cells. In the perinatal period, Sertoli cell proliferation subsides and germ cell proliferation begins. These cells migrate radially to take up positions along the basement membrane between Sertoli cells and constitute the stem cell population that gives rise to sperm cells throughout the life of the animal. By the third postnatal week, the initial wave of germ cell proliferation and migration subsides and a second wave of proliferation ensues, which is coupled with diVerentiation and centripetal migration. At this stage, Sertoli cells elongate and establish two compartments within the seminiferous tubules delimited by a claudin 11 tight junction network: a basal compartment that includes the stem cell population and is in contact with blood-derived factors, as well as a luminal compartment for spermatocyte diVerentiation with a solute composition that is distinct from that of blood and is tightly regulated by the Sertoli cells. The claudin 11 tight junction network is the major component of the blood-testes barrier (Dym and Fawcett, 1970; Gilula et al., 1976; Pelletier and Byers, 1992). After germ cells migrate through the Sertoli cell tight junctions and take up residence in the luminal compartment, they undergo meiosis, chromatin condensation and diVerentiation to spermatids (Leblond and Clermont, 1952). Disruption of the claudin 11 tight junctions in the testis leads to widespread cell death of the germ cell lineage in the lumenal compartment and has been observed in seasonal breeders, Claudin 11-null and Occludin-null mice, and after treatment with drugs that destroy the blood-testis barrier (Gow et al., 1999; Pelletier and Byers, 1992; Pelletier and Friend, 1986; Saitou et al., 2000). On the other hand, stem cells in the basal compartment appear relatively unaVected. Data from several studies indicate that the unique solute composition of the luminal compartment is paramount for germ cell diVerentiation and one of the unusual constituents of this Xuid is 50 mM Kþ (Waites and Gladwell, 1982). In view of a primary function of claudin 11 junctions in the Stria vascularis, which is to separate two compartments of vastly diVerent Kþ concentrations, Sertoli cell junctions can be considered to play a similar role in separating the basal and luminal compartments of the seminiferous tubule where the Kþ concentration gradient across Sertoli cells exceeds tenfold.

Claudin 11 Tight Junctions in CNS Myelin A distinctive feature of tight junctions in CNS myelin compared to most other polarized epithelia examined to date is the apparent absence of all major structural components beyond claudin 11. For example, the modular proteins ZO
1,
2, and
3, which are known to play crucial scaVolding roles for cytoskeleton attachment as well as for anchoring claudin family members and occludin to the tight junction plaque via PDZ domain interactions, are not detected in CNS white matter tracts. In contrast, the Schwann cells that myelinate peripheral nervous system axons make tight junctions containing ZO-1 and ZO-2 (Poliak et al., 2002). The distribution of claudin 11 within the myelin sheath diVers from that of most other structural proteins that have been characterized in this membrane. In the relatively small myelinated Wbers of rodents, claudin 11 is usually localized to continuous narrow channels Wlled with cytoplasm that track as loose spirals from end to end along the myelin sheath; in contrast, other structural myelin proteins such as PLP1 and MBP are relatively evenly distributed throughout the sheath. On the other hand, claudin 11 exhibits much more complex localization in the large diameter myelin sheaths of primates. For example, in monkey myelin, immunoXuorescence staining using anti-claudin 11 antibodies reveals tight

571

572

22. THE CLAUDIN 11 GENE

spiral channels at each end of myelin sheaths in paranodal regions. A loose spiral channel runs along the sheath between the two paranodes that is reminiscent of the rodent distribution and, in addition, short tight spirals emanate from this channel at regular intervals and reXect the presence of cytoplasmic channels coursing through the compact myelin, called Schmidt-Lanterman incisures (Gow et al., 1999). Schematics of the positions and organization of these tight junctions are illustrated in Figure 22.3. The immunocytochemical localization of claudin 11 in CNS myelin corresponds closely to extensive cytoplasmic channels observed by investigators under the light microscope in the early part of the 20th century (del Rio-Hortega, 1928) as well as to the distribution of zonulae occludentes and the radial component observed in freeze fracture replicas and transmission electron micrographs of white matter (Mugnaini and Schnapp, 1974; Schnapp and Mugnaini, 1978; Tabira et al., 1978). Indeed, recent morphological data show that myelin zonulae occludentes and the radial component are absent in CNS sheaths from Claudin 11-null mice, which provides strong evidence that claudin 11 is the principal transmembrane component of intramembranous Wbrils in this tissue (Gow et al., 1999). An unexpected Wnding in the CNS of Claudin 11-null mice is the absence of immune activation toward protein components of the myelin sheath. Early hypotheses regarding the functions of tight junctions in this membrane contend that these structures provide an immune privileged compartment for myelin membrane proteins that are not expressed until long after immune self-recognition is established in the perinatal period (Mugnaini and Schnapp, 1974). Such notions were modeled on data suggesting that autoimmunemediated infertility in males results from breakdown of the blood-testis barrier. However, there is no evidence of T-cell inWltration into the CNS of Claudin 11-null mutants or demyelination of white matter tracts out more than 12 months of age (Gow et al., 1999). Another role of CNS myelin tight junctions proposed from early studies is that these barriers serve to stabilize the multilamellar organization of compact myelin. Indeed, these morphological studies showed that multilamellar regions of puriWed compact myelin membrane treated with hypoosmotic solutions tended to split along the major dense line only to points of intersection with the radial component (Nagara and Suzuki, 1982; Tabira et al., 1978). However, claudin 11 tight junctions do not appear to be the major or lone supramolecular assembly to impart such structural stability on compact myelin because myelin ultrastructure and period are unaVected in Claudin 11-null mice (Gow et al., 1999). In contrast, the absence of other myelin components—including MBP, MAG, galactocerebrosides and Plp1 gene products or replacement of PLP1 with DM-20—in myelin sheaths does have ultrastructural consequences in the short term in the form of an altered myelin period (Coetzee et al., 1996; Kimura et al., 1989; Klugmann et al., 1997; Lassmann et al., 1997; Stecca et al., 2000). The consequences to the behavior of mice in the absence of claudin 11 in CNS myelin are surprisingly minor and there is no detectable myelin loss in aging animals at the ultrastructural, transcriptional or posttranslational levels (Gow et al., 1999). However, a

FIGURE 22.3 Organization of claudin 11 tight junctions in CNS myelin sheaths. The schematic shows portions of two CNS myelin sheaths separated by a node of Ranvier (Node). The sheath to the left is cut away at the paranode and shows the radial organization of tight junctions that are continuous between the inner and out loops of the compact myelin internode. In addition, tight junctions in the paranode seal the extracellular surfaces between the lateral myelin loops. The sheath to the right is rendered transparent to reveal the spiral organization of tight junctions at the paranode and the Schmidt-Lanterman incisures (*). The loose spirals that course obliquely to the long axis of the myelinated fiber (arrows) represent tight junctions under the outer loop of the myelin membrane; for simplicity, the analogous tight junction strands running along the inner loop are not depicted.

TOWARD THE FUNCTION OF CLAUDIN 11 TIGHT JUNCTIONS IN MYELIN

relatively Wne shivering behavior is revealed when these animals are held by the tail. This phenotype becomes apparent in the third postnatal week, lasts until approximately two months of age and is most obvious at earlier ages. The mice also exhibit an apparent hind limb weakness that persists into adulthood and is detectable when balancing on a horizontal bar or during motor coordination tests such as the rotarod. Currently, the underlying etiology of these phenotypes is unknown but, conceivably, may stem from neurological abnormalities associated with aberrant myelin sheath function (see the section titled ‘‘Toward the Function of Claudin 11 Tight Junctions in Myelin’’). Conduction velocities along CNS myelinated Wbers may also be perturbed in Claudin 11-null mice. The latency period of strobe Xash-evoked potentials recorded from scalp electrodes over the visual cortex is increased by 15 to 20% in mutants compared to control littermates (Gow et al., 1999). A simple interpretation of these data is that conduction velocity along the optic nerve and elsewhere in the visual system is decreased by 15 to 20% in the absence of myelin tight junctions. This decrease is relatively modest and suggests that the tight junctions may serve to optimize conduction velocity. In other studies, soundevoked responses in the auditory system that are associated with nerve conduction from the cochlea to the brain stem (the so-called peak V latency) are slowed approximately 20% in Claudin 11-null mice. These data also suggest a reduction in conduction velocity in the mutants that is consistent with the visual evoked response data (unpublished observations). On the other hand, other explanations could, conceivably, account for increased latencies in the null mutants. For example, synaptic connections between retinal ganglion cells and other neurons may function more slowly in the null mutants compared to controls which could increase latency times.

TOWARD THE FUNCTION OF CLAUDIN 11 TIGHT JUNCTIONS IN MYELIN In view of the general function of tight junctions in polarized epithelia from many tissues and the electrophysiological abnormalities observed in visual and auditory evoked responses in Claudin 11-null mice, testable hypotheses concerning the function of the radial component of myelin are at hand. Thus, tight junctions are known to occlude the paracellular space to the passive diVusion of large and many small molecules, and recent evidence indicates that at least some claudin family members act as ion-speciWc channels between epithelial cells (Colegio et al., 2002; Simon et al., 1999). On the other hand, claudin 11 tight junctions form highly impermeable junctions in at least some tissues, which, in the case of Sertoli cell junctions in the testis, are among the most impermeable barriers known (Claude and Goodenough, 1973). In this light, claudin 11 tight junctions appear to form impermeable barriers to separate compartments of vastly diVerent concentrations of Kþ as suggested for the cochlea and seminiferous tubules. Claudin 11 intramembranous Wbrils are continuous around the perimeter of the myelin sheath (Mugnaini and Schnapp, 1974), which is most clearly seen when the sheath is represented as an unraveled Xat sheet of membrane (Fig. 22.4A). This barrier is ideally placed to prevent passive diVusion of solutes into the sheath that might otherwise enter at the outer and inner edges of the myelin along the internode (Fig. 22.4B) or at the paranodes (Fig. 22.4C). In eVect, these tight junctions could serve to seal this multilamellar structure to generate a thick, impermeable lipid insulator with the high-resistance and lowcapacitance electrophysiological properties favorable for saltatory conduction (Ritchie, 1984). The time interval between sodium channel activation at successive Nodes determines conduction velocity and is critically dependent on the properties of the myelin sheath. The more eVective the insulation provided by this membrane, the lower the current loss from myelin sheaths and the more rapid the depolarization at successive nodes of Ranvier. Thus, current leakage is a major consequence of damage or perturbation to the structure or composition of myelin, and available electrophysiological data indicate that the paranode

573

574

22. THE CLAUDIN 11 GENE

FIGURE 22.4 Claudin 11 tight junctions may serve to seal the edges of the myelin sheath a. When depicted as a flat membrane sheet, the organization of claudin 11 tight junctions is apparent. Tight junction strands completely surround the edges of the membrane at the outer, inner and lateral loops. Furthermore, the radial tight junctions form linear stripes along the internode with the distance between each strip equal to the circumference of the fiber. b. This schematic depicts a cross-section through the internodes of myelin sheaths (for simplicity, only two myelin lamellae are shown) from wild type (upper) or Claudin 11-null mice (lower). The radial tight junctions from the outer to the inner loops in wild type myelin are ideally positioned to occlude the entry of solutes between the lamellae into the interior of the sheath both from the parenchyma as well as the periaxonal space (arrows). In contrast, the absence of tight junctions in myelin theoretically provides an unobstructed path from the periaxonal space to the parenchyma. In practical terms, it is possible that solutes could penetrate a short distance between the myelin lamellae from the parenchyma and the periaxonal space. In this regard, perhaps the more important region of the sheath is from the periaxonal space. c. In analogous fashion to solute penetration in the internode, tight junctions at the paranodes occlude the space between the lateral myelin loops. During saltatory conduction, relatively large numbers of ions move across the node of Ranvier and some of these ions could, conceivably, be driven into the CNS myelin sheaths of Claudin 11-null mice.

is the most vulnerable region of the sheath leading to reduced conduction velocity or intermittent conduction (Rasband et al., 1998). Indeed, mouse mutants that lack any of several axoglial junction components including Caspr, contactin, or galactocerebroside exhibit severe ataxia and conduction velocity abnormalities because myelin paranodal loops fail to adhere to the axolemma (Bhat et al., 2001; Boyle et al., 2001; Coetzee et al., 1996).

TOWARD THE FUNCTION OF CLAUDIN 11 TIGHT JUNCTIONS IN MYELIN

Consideration of the current Xow at and around a depolarizing node of Ranvier during saltatory conduction in CNS myelinated axons provides insights into the mechanism by which claudin 11 tight junctions may serve to maximize conduction velocity and may explain abnormalities in visual and auditory evoked responses that have been observed in Claudin 11-null mice (unpublished observations, Gow et al., 1999). In this regard, Figure 22.5 shows a schematic of a myelinated axon that has been sectioned through the long axis.

FIGURE 22.5 Role of claudin 11 tight junctions in saltatory conduction of CNS myelinated Wbers. a. A longitudinal section through a myelinated axon to show a myelin sheath and flanking Nodes of Ranvier. For clarity, the periaxonal and intramyelinic spaces are drawn disproportionately large, and only the two innermost myelin lamellae are represented. The paranodal myelin loops (hatched) in wild type mice (top) are interconnected by tight junctions and adhere to the axon via axoglial junctions. Sodium ions moving through ion channels at a depolarizing Node of Ranvier (left) displace Kþ in the region to generate a capacitive shunt current at the adjacent Node (right). The Naþ ions at the external surface of this Node are displaced from the membrane and complete a circuit with the depolarizing Node along the outside of the myelin sheath. However, in the absence of myelin tight junctions (bottom), some of the displaced Kþ at the depolarizing Node activate internodal Kþ-channels in the axonal membrane and Kþ can flow into the periaxonal and intramyelinic spaces to generate a resistive shunt current which spirals through the myelin and completes the circuit with the depolarizing Node. Consequently, less current is available at the adjacent Node which increases the time taken to drive those Naþ-channels to threshold and depolarize the Node i.e. conduction velocity is slowed. b. Cross-section through the internode of a myelinated axon to show more clearly the spiral path taken by the resistive shunt current (dashed spiral arrow) through the connected periaxonal and intramyelinic spaces in Claudin 11-null mice (left). Only the first turn of the multilamellar membrane is represented (thick spiral line). In wild type mice (right), these shunt currents are not normally generated because the continuity of the tight junctions around the inner, outer and paranodal myelin loops electrically-seal the intramyelinic space and ensure that the myelin sheath acts as a capacitor of high resistance.

575

576

22. THE CLAUDIN 11 GENE

Nodes of Ranvier are shown on either side of a myelin sheath and the node to the left is depolarized. The direction of saltatory conduction is left to right. As voltage-gated sodium channels at the left-hand node of Ranvier open and sodium ions Xow down their concentration gradient into the axon, positive charge increases in the vicinity of the axolemma. This charge buildup exerts a bidirectional electrostatic force on ions along the axon. Cations in the vicinity of the axolemma at the node of Ranvier to the right are repelled by this charge buildup and move toward the membrane, thereby exerting a repulsive force on ions across the bilayer (a capacitative current). The resulting movement of these extracellular ions away from the axolemma allows a buildup of ions at the cytoplasmic surface that depolarizes this node suYciently to open voltage-gated sodium channels. Such capacitative shunt currents at nodes are minimized in the myelin internode by the narrow periaxonal space. Importantly, the tight junctions that seal the inner myelin loop along its length are a key component of the periaxonal space that limits the size of this compartment. Thus, cations in the periaxonal space cannot enter the myelin sheath as charge builds up on the opposite side of the membrane because movement away from the axolemma is physically constrained. In similar fashion, capacitative current leakage from the axon into the myelin sheath at the paranodes is also prevented by claudin 11 tight junctions between paranodal loops. Indeed, if such a capacitative shunt were to occur, it may be suYcient to open the normally inactive voltage-gated potassium channels that are known to be located in the juxtaparanode (Rasband et al., 1998). In summary, the role of claudin 11 tight junctions in CNS myelin sheaths remains uncertain, although recent data suggest a barrier function that is similar to that observed for tight junctions in many polarized epithelia. Although it is clear from mouse mutant studies that the integrity of the adhesive axoglial junctions at the paranode is paramount to normal conduction along myelinated Wbers, the hypothesis developed herein makes room for claudin 11 to make an important contribution to saltatory conduction. Indeed, experiments are currently underway to test this hypothesis.

Acknowledgments I thank Ms. Cherie Southwood and Mr. Ramaswamy Sharma (Center for Molecular Medicine and Genetics, Wayne State University) for their critical evaluation of this work. as well as the National Institutes of Health (RO1 NS43783), National Multiple Sclerosis Society (RG 2891-B-2). and Children’s Research Center of Michigan for their generous Wnancial support.

References Bhat, M. A., Rios, J. C., Lu, Y., Garcia-Fresco, G. P., Ching, W., Martin, M. S., Li, J., Einheber, S., Chesler, M., Rosenbluth, J., Salzer, J. L., and Bellen, H. J. (2001). Axon-glia interactions and the domain organization of myelinated axons requires neurexin iv/caspr/paranodin. Neuron 30, 369–383. Boggs, J. M., Rangaraj, G., Koshy, K. M., and Mueller, J. P. (2000). Adhesion of acidic lipid vesicles by 21.5 kDa (recombinant) and 18.5 kDa isoforms of myelin basic protein. Biochim Biophys Acta 1463, 81–7. Boison, D., and StoVel, W. (1994). Disruption of the compacted myelin sheath of axons of the central nervous system in proteolipid protein-deWcient mice. Proc. Natl. Acad. Sci. USA 91, 11709–11713. Boyle, M. E., Berglund, E. O., Murai, K. K., Weber, L., Peles, E., and Ranscht, B. (2001). Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30, 385–397. Bronstein, J. M., Chen, K., Tiwari-WoodruV, S., and Kornblum, H. I. (2000a). Developmental expression of OSP/claudin-11. J. Neurosci. Res. 60, 284–290. Bronstein, J. M., Micevych, P. E., and Chen, K. (1997). Oligodendrocyte-speciWc protein (OSP) is a major component of CNS myelin. J. Neurosci. Res. 50, 713–720. Bronstein, J. M., Popper, P., Micevych, P. E., and Farber, D. B. (1996). Isolation and characterization of a novel oligodendrocyte-speciWc protein. Neurology 47, 772–778. Bronstein, J. M., Tiwari-WoodruV, S., Buznikov, A. G., and Stevens, D. B. (2000b). Involvement of OSP/claudin11 in oligodendrocyte membrane interactions: Role in biology and disease. J. Neurosci. Res. 59, 706–711. Claude, P., and Goodenough, D. A. (1973). Fracture faces of zonulae occludentes from ‘‘tight’’ and ‘‘leaky’’ epithelia. J. Cell Biol. 58, 390–400. Coetzee, T., Fujita, N., Dupree, J., Shi, R., Blight, A., Suzuki, K., and Popko, B. (1996). Myelination in the absence of galactocerebroside and sulfatide: Normal structure with abnormal function and regional instability. Cell 86, 209–219.

au2

TOWARD THE FUNCTION OF CLAUDIN 11 TIGHT JUNCTIONS IN MYELIN

au3

au4

Colegio, O. R., Van Itallie, C. M., McCrea, H. J., Rahner, C., and Anderson, J. M. (2002). Claudins create charge-selective channels in the paracellular pathway between epithelial cells. Am J Physiol Cell Physiol 283, C142–7. del Rio-Hortega, P. (1928). Tercera aportacion al conocimiento morfologico e interpretacion funcional de la oligodendroglia. Mem. Real. Soc. Espan. Hist. Nat. 14, 5–119. Dym, M., and Fawcett, D. W. (1970). The blood-testis barrier in the rat and the physiological compartmentation of the seminiferous epithelium. Biol. Reprod. 3, 308–326. Furuse, M., Fujita, K., Hiiragi, T., Fujimoto, K., and Tsukita, S. (1998). Claudin-1 and -2: Novel integral membrane proteins localizing at tight junctions with no sequence similarity to occludin. J. Cell Biol. 141, 1539–1550. Furuse, M., Hirase, T., Itoh, M., Nagafuchi, A., Yonemura, S., Tsukita, S., and Tsukita, S. (1993). Occludin: A novel integral membrane protein localizing at tight junctions. J. Cell Biol. 123, 1777–1788. Garbern, J. Y., Yool, D. A., Moore, G. J., Wilds, I. B., Faulk, M. W., Klugmann, M., Nave, K. A., Sistermans, E. A., van der Knaap, M. S., Bird, T. D., Shy, M. E., Kamhaz, J. A., and Griffiths, I. R. (2002). Patients lacking the major CNS myelin protein, proteolipid protein 1, develop length-dependent axonal degeneration in the absence of demyelination and inXammation. Brain 125, 551–61. Gilula, N. B., Fawcett, D. W., and Aoki, A. (1976). The Sertoli cell occluding junctions and gap junctions in mature and developing mammalian testis. Dev. Biol. 50, 142–168. Gow, A., Southwood, C. M., Li, J. S., Pariali, M., Riordan, G. P., Brodie, S. E., Danias, J., Bronstein, J. M., Kachar, B., and Lazzarini, R. A. (1999). CNS myelin and sertoli cell tight junction strands are absent in Osp/ Claudin 11-null mice. Cell 99, 649–659. GriYths, I., Klugmann, M., Anderson, T., Yool, D., Thomson, C., Schwab, M. H., Schneider, A., Zimmermann, F., McCulloch, M., Nadon, N., and Nave, K. A. (1998). Axonal swellings and degeneration in mice lacking the major proteolipid of myelin. Science 280, 1610–1613. Jaquet, V., Gow, A., Tosic, M., Suchanek, G., Breitschopf, H., Lassmann, H., Lazzarini, R. A., and Matthieu, J. M. (1996). An antisense transgenic strategy to inhibit the myelin oligodendrocyte glycoprotein synthesis [In Process Citation]. Brain Res Mol Brain Res 43, 333–7. Kikuchi, T., Kimura, R. S., Paul, D. L., Takasaka, T., and Adams, J. C. (2000). Gap junction systems in the mammalian cochlea. Brain Res. Brain Res. Rev. 32, 163–166. Kimura, M., Sato, M., Akatsuka, A., Nozawakimura, S., Takahashi, R., Yokoyama, M., Nomura, T., and Katsuki, M. (1989). Restoration of myelin formation by a single type of myelin basic protein in transgenic shiverer mice. Proc. natl. Acad. Sci. U. S. A. 86, 5661–5665. Klugmann, M., Schwab, M. H., Puhlhofer, A., Schneider, A., Zimmermann, F., GriYths, I. R., and Nave, K.-A. (1997). Assembly of CNS myelin in the absence of proteolipid protein. Neuron 18, 59–70. Lapierre, L. A., Tuma, P. L., Navarre, J., Goldenring, J. R., and Anderson, J. M. (1999). VAP-33 localizes to both an intracellular vesicle population and with occludin at the tight junction. J. Cell Sci. 112, 3723–3732. Lassmann, H., Bartsch, U., Montag, D., and Schachner, M. (1997). Dying-back oligodendrogliopathy: A late sequel of myelin-associated glycoprotein deWciency. Glia 19, 104–10. Leblond, C. P., and Clermont, Y. (1952). DeWnition of the stages of the cycle of the seminiferous epithelium in the rat. N.Y. Acad. Sci. 55, 548–573. Lees, M. B., and BrostoV, S. W. (1984). Proteins in myelin. In ‘‘Myelin’’ (P. Morell, ed.), pp. 197–224. Plenum, New York and London. Mikoshiba, K., Aruga, J., Ikenaka, K., and Okano, H. (1992). Shiverer and allelic mutant MLD mice. In ‘‘Myelin: Biology and Chemistry’’ (R. E. Martenson, ed.), pp. 723–744. CRC Press, Boca Raton, FL. Mitic, L. L., and Anderson, J. M. (1998). Molecular architecture of tight junctions. Annu. Rev. Physiol. 60, 121–142. Morita, K., Furuse, M., Fujimoto, K., and Tsukita, S. (1999a). Claudin multigene family encoding fourtransmembrane domain protein components of tight junction strands. Proc. Natl. Acad. Sci. U.S.A. 96, 511–516. Morita, K., Sasaki, H., Fujimoto, K., Furuse, M., and Tsukita, S. (1999b). Claudin-11/OSP-based tight junctions of myelin sheaths in brain and Sertoli cells in testis. J. Cell Biol. 145, 579–588. Mugnaini, E., and Schnapp, B. (1974). Possible role of zonula occludens of the myelin sheath in demyelinating conditions. Nature 251, 725–727. Nagara, H., and Suzuki, K. (1982). Radial component of the central myelin in neurologic mutant mice. Lab. Invest., 47–51. Pelletier, R. M., and Byers, S. W. (1992). The blood-testis barrier and Sertoli cell junctions: Structural considerations. Microsc. Res. Tech. 20, 3–33. Pelletier, R. M., and Friend, D. S. (1986). Sertoli cell junctional complexes in gossypol-treated neonatal and adult guinea pigs. J Androl 7, 127–39. Poliak, S., Matlis, S., Ullmer, C., Scherer, S. S., and Peles, E. (2002). Distinct claudins and associated PDZ proteins form diVerent autotypic tight junctions in myelinating Schwann cells. J Cell Biol 159, 361–372. Rasband, M. N., Trimmer, J. S., Schwarz, T. L., Levinson, S. R., Ellisman, M. H., Schachner, M., and Shrager, P. (1998). Potassium channel distribution, clustering, and function in remyelinating rat axons. J. Neurosci. 18, 36–47. Ritchie, J. M. (1984). Physiological basis of conduction in myelinated nerve Wbers. In ‘‘Myelin’’ (P. Morell, ed.), pp. 117–145. Plenum, New York.

577

578

22. THE CLAUDIN 11 GENE

Saitou, M., Fujimoto, K., Doi, Y., Itoh, M., Fujimoto, T., Furuse, M., Takano, H., Noda, T., and Tsukita, S. (1998). Occludin-deWcient embryonic stem cells can diVerentiate into polarized epithelial cells bearing tight junctions. J. Cell Biol. 141, 397–408. Saitou, M., Furuse, M., Sasaki, H., Schulzke, J. D., Fromm, M., Takano, H., Noda, T., and Tsukita, S. (2000). Complex phenotype of mice lacking occludin, a component of tight junction strands. Mol Biol Cell 11, 4131–4142. Salt, A. N., and Konishi, T. (1986). The cochlear Xuids: Perilymph and endolymph. In ‘‘Neurobiology of Hearing: The Cochlea’’ (R. A. Altschuler, D. W. HoVman, and R. P. Bobbin, eds.), pp. 109–122. Raven, New York. Salt, A. N., Melichar, I., and Thalmann, R. (1987). Mechanisms of endocochlear potential generation by stria vascularis. Laryngoscope 97, 984–991. Schnapp, B., and Mugnaini, E. (1978). Membrane architecture of myelinated Wbers as seen by freeze-fracture. In ‘‘Physiology and Pathobiology of Axons’’ (S. G. Waxman, eds.), pp. 83–123. Raven, New York. Simon, D. B., Lu, Y., Choate, K. A., Velazquez, H., Al-Sabban, E., Praga, M., Casari, G., Bettinelli, A., Colussi, G., Rodriguez-Soriano, J., McCredie, D., Milford, D., Sanjad, S., and Lifton, R. P. (1999). Paracellin-1, a renal tight junction protein required for paracellular Mg2þ resorption. Science 285, 103–106. Souter, M., and Forge, A. (1998). Intercellular junctional maturation in the stria vascularis: Possible association with onset and rise of endocochlear potential. Hear. Res. 119, 81–95. Southwood, C. M., and Gow, A. (2001). Functions of OSP/claudin 11-containing parallel tight junctions: Implications from the knockout mouse. In ‘‘Tight Junctions’’ (J. M., anderson and M. Cereijido, eds.), pp. 719–741. CRC Press, New York. Stecca, B., Southwood, C. M., Gragerov, A., Kelley, K. A., Friedrich, V. L. J., and Gow, A. (2000). The evolution of lipophilin genes from invertebrates to tetrapods: DM-20 cannot replace PLP in CNS myelin. J. Neurosci. 20, 4002–4010. Tabira, T., Cullen, M. J., Reier, P. J., and Webster, H. (1978). An experimental analysis of interlamellar tight junctions in amphibian and mammalian C. N. S. myelin. J. Neurocytol. 7, 489–503. Takeuchi, S., and Ando, M. (1998). Dye-coupling of melanocytes with endothelial cells and pericytes in the cochlea of gerbils. Cell Tissue Res. 293, 271–275. Tiwari-WoodruV, S. K., Buznikov, A. G., Vu, T. Q., Micevych, P. E., Chen, K., Kornblum, H. I., and Bronstein, J. M. (2001). OSP/claudin-11 forms a complex with a novel member of the tetraspanin super family and beta1 integrin and regulates proliferation and migration of oligodendrocytes. J Cell Biol 153, 295–305. Waites, G. M., and Gladwell, R. T. (1982). Physiological signiWcance of Xuid secretion in the testis and bloodtestis barrier. Physiol. Rev. 62, 624–671. Wilcox, E. R., Burton, Q. L., Naz, S., Riazuddin, S., Smith, T. N., Ploplis, B., Belyantseva, I., Ben-Yosef, T., Liburd, N. A., Morell, R. J., Kachar, B., Wu, D. K., GriYth, A. J., and Friedman, T. B. (2001). Mutations in the gene encoding tight junction claudin-14 cause autosomal recessive deafness DFNB29. Cell 104, 165–72. Wolburg, H., Wolburg-Buchholz, K., Liebner, S., and Engelhardt, B. (2001). Claudin-1, claudin-2 and claudin-11 are present in tight junctions of choroid plexus epithelium of the mouse. Neurosci Lett 307, 77–80. Wong, V. (1997). Phosphorylation of occludin correlates with occludin localization and function at the tight junction. Am. J. Physiol. 273, C1859-C1867.

au5

au6

C H A P T E R

23 The Neurexin and NCP Gene Families Manzoor A. Bhat

INTRODUCTION Neuronal development and function involves intricate molecular mechanisms, which are currently being unraveled using a combination of genetic and molecular methods. The generation of neuronal speciWcity and the precision with which each individual neuron performs its function relies on the expression and presence of speciWc protein components. Some of these proteins play dual functions, as they may be involved in establishing speciWc neuronal connections and also act as signaling molecules in signal transduction pathways. These signaling pathways allow the nervous system to coordinate, interpret, and respond to various stimuli within and from its immediate environment. One class of such molecules, the neurexins, was originally identiWed as a polymorphic family of neuronal-speciWc type-I cell surface membrane proteins that were proposed to be involved in specifying synaptic speciWcity and synaptic vesicle docking at the synaptic active zone. Most of the proposed functions are based on the biochemical interactions between neurexins and a component of the black widow spider venom, a-latrotoxin. In addition, neurexins have been found to interact with a number of unrelated proteins that include synaptotagmin, neurexophilins, neuroligins, and CASK to suggest a role in synaptic function. Some of the conclusions on the functions of the neurexins remain an area of immense interest due to their proposed synaptic functions and their presence in invertebrates (e.g., C. elegans and Drosophila, two important genetic model systems). Application of genetic and molecular techniques particularly gene ablation mutagenesis and homologous recombination in embryonic stem cells has clearly demonstrated the critical role of some of the neurexin family members in neuronal function. The chapter oVers a brief review of the emerging functions of neurexins and identiWes additional members of the neurexin superfamily.

MEMBERS OF THE NEUREXIN SUPERFAMILY During the past decade, a number of novel proteins were identiWed that displayed domain homology similar to those of neurexins, with some modiWcations. It has become clear with the completion of the sequencing of the human, mouse, Drosophila, and C. elegans genomes that neurexins and neurexin-related proteins form a superfamily that evolved from a common ancestral gene. Based on the sequence and domain homology, and functional analysis, the invertebrate and vertebrate members of the neurexin superfamily fall in two related groups: the neurexin family and the NCP family (Neurexin IV/Caspr/ Paranodin; Bellen et al., 1998). Detailed information about the members of the neurexin superfamily identiWed thus far is given in Table 23.1. These families, with their individual members for which in vivo functional data are available, are discussed next.

Myelin Biology and Disorders, Volume 1

579

Copyright 2004, Elsevier (USA). All rights reserved.

580

23. THE NEUREXIN AND NCP GENE FAMILIES

TABLE 23.1 Gene Structure of Neurexin Superfamily Members Gene

Species

Number of exons

Number of amino acids

Accession number

Chromosome location

1477

BAA87821

2p21

1712

BAA94075

Neurexins Neurexin I

Human

24

Neurein I

Mouse

24

Neurexin II

Human

23

Neurexin II

Mouse

23

Neurexin III

Human

24

Neurexin III

Mouse

24

17E5 11q13 19A 1061

NP_004787

14q31 12D2-3

Neurexin

Drosophila

13

1837

NP_524449

3(94B)

Neurexin

C. elegans

27

1560

NP_505767

V

Axotactin

Drosophila

15

1685

AAD29408

3(64B)

NCPs NCP1/Caspr

Human

24

1384

AAB48481

17q21

NCP1/Caspr

Mouse

24

1385

AAB96760

11

NCP2/Caspr2

Human

25

1331

Q9UHC6

7q36

NCP2/Caspr2

Mouse

NCP3/Caspr3

Human

NCP3/Caspr3

Mouse

6 21

1288

AAG52889

9q12

AAL68839

16q22

NCP4/Caspr4

Human

24

1311

NCP4/Caspr4

Mouse

24

1310

NCP5/Caspr5

Human

24

1306

NP_570129

2q14

NCP5/Caspr5

Mouse 13

1283

CAA60383

3(68F)

1193

NP_502312

NCP/NRX IV

Drosophila

NCP

C. elegans

8

Neurexin Family Neurexins were identiWed in a biochemical screen for putative a-latrotoxin binding proteins or receptors from synaptic membranes using immobilized a-latrotoxin in aYnity chromatography (Ushkaryov et al., 1992). a-Latrotoxin is a very potent neurotoxin from black widow spider venom, which triggers exocytosis of synaptic vesicles in a calcium-independent manner to release neurotransmitters at the synapse. Many other studies have also reported a-latrotoxin-binding proteins using aYnity chromatography, which include proteins other than neurexins (Petrenko et al., 1990, 1993; Scheer and Meldolesi, 1985; Scheer et al., 1986). Sequence analysis of one of these a-latrotoxinbinding proteins showed that it constitutes a highly polymorphic family of neuron-speciWc cell-surface proteins, which were named neurexins (Ushkaryov et al., 1992). Further molecular characterization identiWed two additional members that displayed identical domain structure. Thus vertebrates contain at least three neurexin genes, now referred to as neurexins I, II, and III. As shown in Figure 23.1, the vertebrate neurexin genes encode a large a-isoform (1507 to 1578 amino acids, 160 to 220 kDa) and a small b-isoform (437 to 471 amino acids). The primary structure of neurexins contains a large extracellular domain with O-linked sugar attachment sites in proximity to a single transmembrane domain, and a short cytoplasmic segment (40-55 amino acids), therefore resembling cell surface receptors. The domain homology search revealed that the extracellular sequences of a-neurexins are composed of three unique sets, each of which consists of a central epidermal growth factor (EGF)-like sequence Xanked on each side by distantly related LNS domains. The LNS domain refers to its homology to the G domain of laminin A, its presence in neurexins, and sex-hormone binding globulin (Fig. 23.1). These unique laminin G type domains—also found in agrin, slit, protein S, and perlecan—constitute protein-protein

MEMBERS OF THE NEUREXIN SUPERFAMILY

581

FIGURE 23.1 Members of the neurexin superfamily. (A) The domain structure of three representative neurexin family members from human, Drosophila and C. elegans are shown. The human a-neurexins are shown as H.s. NRXIa and b-neurexins are shown as H.s. NRX1b. The invertebrate neurexin genes from Drosophila (D.m. NRX1) and C. elegans (C.e. NRX1) seem to encode only the long a isoforms; however, the presence of the b-neurexin isoforms cannot be ruled out. (B) The domain structure of the representative NCP family members from human and Drosophila. Most of the NCP members display almost identical domain structure with a carbohydrate-binding discoidin domain replacing an LNS and EGF repeat of the neurexin family. There is no indication that NCP members generate any short b-isoforms. (C) Evolutionary analysis results obtained by comparing the protein sequences of all the neurexin superfamily members given in Table 23.1, using the Grow Tree Software of GCG. The neurexins and NCPs clearly fall in two diVerent groups with ancestors in C. elegans and Drosophila and might have originated from a common ancestor. The Drosophila axotactin does not seem to be a close member of the neurexin or NCP family and seems distantly related to neurexins.

interaction domains that in neurexins speciWcally bind to a-latrotoxin in the presence of Caþþ. A detailed sequence analysis showed that neurexin gene transcription involves two independent promoters, an external promoter, which directs the transcription of the larger a-neurexins, and an internal intronic promoter that directs the transcription of the shorter b-neurexins (Ushkaryov et al., 1992, 1994; Ushkaryov and Siidhof, 1993). Consequently, b-neurexins are truncated forms of a-neurexins with a short unique sequence at the N terminus, thereby making b-neurexins diVerent from a-neurexins due to their internal transcription. The primary transcripts of neurexins are subject to extensive alternative splicing, theoretically resulting in hundreds or even thousands of isoforms, which display a diVerential distribution in various areas of the nervous system. There are Wve canonical sites of alternative splicing in a-neurexin transcripts, the last two of which are shared with b-neurexin transcripts (discussed later). Neurexin III, but not neurexin I or II, generates secreted forms, which are produced from alternative splicing caused by in-frame stop codons before the transmembrane region (Ushkaryov et al., 1993). Neurexins are mostly expressed in the brain and in situ hybridization and northern blot analysis revealed diVerential but overlapping expression of various neurexin isoforms in diVerent regions of the brain, and during diVerent stages of embryonic development (Puschel and Betz, 1995; Patzke and Ernsberger, 2000; Ullrich et al., 1995; Ushkaryov et al., 1992;). The transcripts of all three neurexin genes have been detected in the spinal cord of mice as early as embryonic day 10, approximately one day before synaptogenesis is initiated (Puschel and Betz, 1995). It is not clear whether this pre-synaptogenesis expression is critical for synapse formation.

582

23. THE NEUREXIN AND NCP GENE FAMILIES

In addition to binding with a-latrotoxin, neurexins were also found to interact with synaptotagmin, a synaptic vesicle calcium-binding protein (Hata et al., 1993; Petrenko et al., 1991). The interaction of multiple synaptotagmins with neurexins via mirror image motifs in the C-terminus of synaptotagmins was shown by in vitro biochemical binding, suggesting that this interaction may mediate docking of synaptic vesicles or modulation of neurotransmitter release (Perin, 1994, 1996). The binding of neurexins to synaptotagmin provided an excellent hypothesis and novel clues about the mechanism of a-latrotoxin function in eliciting the synaptic response. This interaction with synaptotagmin also suggested that neurexins could play a key role in synaptogenesis in the formation of the active zones by binding to postsynaptic receptors or extracellular matrix components at the synapses and, subsequently aid in docking the synaptic vesicles at these sites. Immunolocalization using anti-neurexin antibodies suggested that neurexins were selectively enriched at the synapses (Ushkaryov et al., 1992). Based on these studies, it was proposed that alternative splicing resulted in the generation of thousands of isoforms to create diversity and thus allow neurexins to specify synaptic targeting and hence synaptic plasticity. However, some of these proposed functions of neurexins have remained inconclusive due to the lack of a Wne subcellular immunohistochemical or ultrastructural localization to the synapses. Additional molecular studies have also raised some doubts about neurexins being the functional a-latrotoxin receptors as their interaction is Caþþ-dependent. a-Latrotoxin exerts its eVects in a calcium independent manner, and surprisingly, a-latrotoxin has also been shown to cause exocytosis in mutant mice lacking synaptotagmin I (Geppert et al., 1994), raising additional questions about whether a neurexinsynaptotagmin interaction has a physiological signiWcance at the synapses. More recently, a Discopyge homolog of vertebrate neurexins has been identiWed, and immunoXuorescence studies showed that neurexins are expressed along the myelinated axonal membranes and not at the synapses, arguing against a universal role for neurexins as nerve terminal speciWc proteins, but rather required for axon-glial interactions (Russell and Carlson, 1997). Recent genomic sequence analyses for human and mouse neurexin genes have provided a detailed description of the genomic structure and various alternative splice variants (Rowen et al., 2002; Tabuchi and Siidhof, 2002). These studies have highlighted the fact that the neurexin genes in human and mouse genomes display identical exon/intron structures and that these genes evolved as a result of relatively recent gene duplications. The neurexin genes have an extremely large size, especially neurexin I and III with ~1.1 and 1.6 Mb, respectively. On the other hand, neurexin II is relatively small spanning a total of~110 kb). It has been suggested that large genes which need a long time for transcription may undergo cotranscriptional splicing and thus generate a smaller transcript than the length of the locus, as is the case for the dystrophin gene (2.25Mb). It is plausible that neurexin genes I and III also use this cotranscriptional splicing mechanism to generate smaller transcripts. A detailed exon-intron structure for vertebrate neurexins was recently described in Rowen et al. (2002) and Tabuchi and Siidhof (2002) and is shown in Figure 23.2. The homologs of vertebrate neurexins have been identiWed in Drosophila and C. elegans based on the genomic sequences in the respective databases, suggesting an evolutionary conservation (Fig. 23.2). Both Drosophila and C. elegans contain a single gene against three vertebrate neurexin genes, and comparison of the primary sequences showed ~30% sequence identity between Drosophila and vertebrate neurexins and ~23% identity between C. elegans and vertebrate neurexins. The two invertebrate neurexins show a relatively lower sequence identity of ~22%. In addition to sequence and domain homology, all neurexin family members have a PDZ domain-binding motif that potentially interacts with a PDZ domain containing protein (Bhat et al., 1999; Hata et al., 1996;). In vertebrates, CASK has been shown to interact with the cytoplasmic domain of neurexin-I (Hata et al., 1996) and interestingly, Drosophila and C. elegans also contain homologs of CASK known as CamGUK and Lin-2, respectively (Dimitratos et al., 1997). Preliminary genetic and molecular studies indicate that the Drosophila neurexin-I is expressed exclusively in the nervous system and may undergo alternative splicing (Li and Bhat, unpublished data). Whether Drosophila and C. elegans neurexin loci can encode homologs of vertebrate b-neurexins remains to be determined.

MEMBERS OF THE NEUREXIN SUPERFAMILY

583

FIGURE 23.2 Genomic structure of neurexin genes. The C. elegans, and Drosophila neurexin genes show variable exon-intron structures and based on sequence information, it is not clear whether the invertebrate neurexins encode b-neurexin isoforms. The vertebrate neurexin genes I-III encode a and b isoforms from an external a promoter and an internal intronic b promoter. Neurexin genes I and III are among the largest genes spanning ~1.7b mb, whereas neurexin gene II is much smaller in size spanning a total of 120 kb. However, the number of exons present in the vertebrate neurexins is remarkably similar irrespective of their large size diVerences (for more details see Rowen et al., 2002; Tabuchi and Su¨dhof, 2002).

The Drosophila genome also contains another neurexin-like gene, axotactin. The primary structure of the axotactin protein contains many additional domains that are not present in neurexins and lacks the transmembrane and cytoplasmic region present in neurexins. This gene is expressed by glial cells and after translation the protein is transported to axonal tracts. Genetic and molecular studies of axotactin mutants indicated that loss of axotactin causes temperature sensitive paralysis and a corresponding blockade of axonal conduction. These studies suggested that axotactin protein is a component of glial neuronal signaling mechanism that determines the 8 membrane electrical properties of target axons (Yuan and Ganetzky, 1999). The overall domain homology between verte-

584

23. THE NEUREXIN AND NCP GENE FAMILIES

brate and invertebrate neurexins is shown in Figure 23.1A, and the evolutionary relationship between the neurexin family members is given in Figure 23.1C. Extracellular Ligands of Neurexins: Neurexophilins, Neuroligins, and Dystroglycan The extracellular domains of neurexins contain laminin G and EGF repeats, which are now known to serve as protein-protein interaction domains. During aYnity chromatography on immobilized a-latrotoxin, a novel small secreted glycoprotein was identiWed which co-puriWed with neurexin la and named neurexophilin (Petrenko et al., 1996). Further molecular analysis revealed the presence of four neurexophilin genes expressed in mammalian brains with the exception of neurexophilin 2, which was not detected in rodents. The molecular structure of neurexophilins includes a signal peptide, a variable N terminal region, a highly conserved central domain, a short linker region and a conserved cysteine rich C-terminal domain (Missler and Sudhof, 1998b). It was also shown that only neurexophilin 1 and 3 bind to a-neurexins through the second LNS domain, but not to b-neurexins (Missler et al., 1998a). Neurexophilins are preferentially expressed in the brain, and the high level expression is enriched in a restricted subset of neurons, namely inhibitory interneurons (Missler et al., 1998b; Petrenko et al., 1996). In vitro cell culture expression experiments in diVerent types of cells showed that proteolytic processing of neurexophilins was only observed in neuron-like cells (Missler et al., 1998b; Petrenko et al., 1996). Given these biochemical interactions, so far no in vivo evidence exists on the possible signaling function of these neuropeptide-like molecules and whether they behave as ligands for cell surface neurexin receptors. Another family of proteins that showed interaction with neurexins is neuroligin. The neuroligins interact with a splice site-speciWc variant of b-neurexins in the presence of Ca++ (Ichtchenko et al., 1995). There are at least three neuroligins in vertebrates and their extracellular domains interact with all three b-neurexin splice variants via their extracellular regions, but not to a-neurexins, only when b-neurexins lack an insert in splice site 4 (Ichtchenko et al., 1995, 1996). The primary structure of neuroligins is composed of a large enzymatically inactive esterase homology domain, a transmembrane region and a short cytoplasmic tail with a PDZ-binding motif. The intracellular tail of neuroligins binds to the third PDZ domain of PSD-95 that is thought to be involved in the assembly and organization of a signal transduction complex at postsynaptic densities at the synapses (Irie et al., 1997). The Wrst two PDZ domains of PSD-95 have been shown to bind N-methyl-D-aspartate (NMDA) receptor and Kþ channels. Cell aggregation assays using cell lines expressing neuroligin I and neurexin Ib, and subcellular localization of neuroligin to postsynaptic densities in excitatory synapses suggest that the neurexin Ibneuroligin interaction may somehow trigger synapse assembly (Nguyen and Siidhof, 1997; Song et al. 1999). Using an in vitro system, it was recently shown that non-neuronal cells that exogenously express neuroligins induce morphological and functional presynaptic diVerentiation in contacting axons, suggesting that the neuroligin-neurexin interaction may be involved in synaptogenesis (ScheiVele et al., 2000). Further experiments need to be carried out to establish whether neuroligins play an instructive role in synapse formation and that this function involves neurexins. Recent aYnity chromatography experiments using immobilized neurexin la have identiWed the neuronal form of dystroglycan as another neurexin-interacting protein. Dystroglycan is composed of two subunits: a highly glycosylated, extracellular a-dystroglycan and a b-dystroglycan containing a short extracellular sequence, a transmembrane region, and a cytoplasmic tail, a- and b-dystroglycan exist as heterodimers and are expressed ubiquitously on the cell surface. The extracellular domain of a-dystroglycan binds to extracellular matrix proteins, laminin, agrin, and perlecan,. Intracellularly, dystrophin provides a link between b-dystroglycan with the actin cytoskeleton. It has been shown that dystroglycan binds tightly to both a- and b-neurexins via a single LNS domain in a manner regulated by alternative splicing (Sugita et al., 2001). Like binding to other extracellular partners, glycosylation of a-dystroglycan is required for binding to neurexins. Further more, new insights into the function of neurexins have come from recent studies on the hypoglycosylation condition in congenital muscular dystrophies, which disrupts the interaction of

au1

585

MEMBERS OF THE NEUREXIN SUPERFAMILY

a-dystroglycan with the extracellular neurexin, laminin, and agrin. This could be the underlying cause of pathology in cases involving muscular dystrophy with brain defects. The phenotype of brain-selective deletion of dystroglycan mice suggests dystroglycanligands that include neurexins are involved in neuronal migration, learning and memory, and other neural functions (Michele et al., 2002; Moore et al., 2002). Neurexins and Synapse Formation The initial observation that neurexins were synapse speciWc cell adhesion molecules has not been conWrmed by subsequent studies and still remains an issue of much controversy (Bellen et al., 1998; Russell and Carlson, 1997; Ushkaryov et al., 1992). Indirect evidence based on biochemical interactions with neuroligins and neurexophilins suggests that neurexins may localize to synapses (Ichtchenko et al., 1995, 1996; Missler and Sudhof, 1998a; Missler et al., 1998b; Petrenko et al., 1996). In addition, this localization by association has raised issues whether this interaction is physiologically relevant. It has been also proposed that the binding of neurexins to neuroligins forms an intercellular junction, which suggests that these proteins function as cell adhesion molecules (Nguyen and Sudhof, 1997). Recently, expression of neuroligins in heterologous cells has been shown to induce the generation of some sort of presynaptic specializations in neurons, indicating a function for neuroligins in synapse formation (ScheiVele et al., 2000). Based on these studies, it has been concluded that at least b-neurexins may have some role to play in synapse formation. The other neurexin-interacting proteins, neurexophilins are secreted proteins and only bind to a-neurexins irrespective of their splice variations (Missler and Sudhof, 1998b; Missler et al., 1998b). Other indirect evidence gathered over the years suggests that neurexins are neuronally expressed proteins and may be either enriched at synaptic junctions or may be part of junctions formed between axons and glial cells (Butz et al., 1999; Hsueh et al., 1998; Song et al., 1999; Sugita et al., 1999). In spite of all the biochemical interactions that have been reported between neurexins and their various putative ligands/receptors over the past decade, the issue of precise subcellular localization of neurexins remains to be addressed. Unless this is clearly established, much of the biochemical studies may not be physiologically relevant. Crystal Structure of an LNS Domain of Neurexins The three-dimensional structure of the last LNS domain of neurexins, which is conserved between a-and b-neurexins, and also found in many other proteins like laminin, agrin, and slit, has been established (Rudenko et al., 1999). This structure reveals two seven-strand b-sheets forming a jellyroll fold with close structural similarities to the carbohydratebinding pentraxins and other lectins. The LNS domains of neurexins seem to have a preferential ligand-interaction site, which is distinct from the carbohydrate-binding sites found in most lectins, thereby allowing LNS domains to accommodate not only sugar moieties but also steroids and other proteins (Rudenko et al., 2001). The splicing sites of neurexins, which have been proposed to tightly regulate the binding of neurexins to a-latrotoxin, neuroligins, and dystroglycan, are located in the loops at the edge of the jellyroll. How the speciWcity of the binding of various ligands to LNS domains in diVerent proteins is achieved remains to be addressed.

NCP (Neurexin IV/Caspr/Paranodin) Family The family derives its name from the identiWcation of several overlapping human expressed sequence tags (ESTs) that were isolated, sequenced, and mapped to 17q21 in a search to identify the breast cancer (BRCA1) gene (Brody et al., 1995; Friedman et al., 1995). The human ESTs were referred to as neurexin-like sequences (Brody et al., 1995). The second member of this family was identiWed by degenerate polymerase chain reaction to identify Drosophila homologs of neurexins and named neurexin IV (Baumgartner et al., 1996). Subsequent work in many laboratories identiWed the homologs of neurexin IV in rat and human, named contactin-associated protein, Caspr (Peles et al., 1997). The same protein was independently isolated as a major rat brain glycoprotein that bound to speciWc lectins. The

au2

586

23. THE NEUREXIN AND NCP GENE FAMILIES

protein was named paranodin as it is localized to the nodes of Ranvier, in the paranodal space (Menegoz et al., 1997). Recently the murine homolog of neurexin IV was identiWed and characterized (Bhat et al., 2001). This family of proteins was named NCP family for Neurexin IV/Caspr/Paranodin (Bellen et al., 1998). The proteins of the NCP family typically contain two or three repeats of a laminin G domain-epidermal growth factor (EGF)-laminin G domain, also named LNS motifs (as described in the section titled ‘‘Extracellular Ligands of Neurexins’’), followed by a transmembrane domain and a short cytoplasmic tail. The NH2-terminal domains within and between the neurexin and NCP families are the most variable domains, and the major diVerence between the NCP members and the original neurexins is that they contain an NH2-terminal discoidin domain that is not found in neurexins I, II, and III (compare Fig. 23.1A with Fig. 23.1B). Although the overall domain structure of these proteins is quite similar, the overall identity between neurexins I, II, and III and the NCP members is only 21 to 29% throughout their entire length. Recently, four additional members of the NCP family have been identiWed in vertebrates. These members have been named Caspr2 (Poliak et al., 1999), Caspr3 and Caspr4 (Spiegel et al., 2002),and Caspr5 (found in the Genebank under accession number NP_570129). Neurexin IV and Septate Junctions in Drosophila

au3

Invertebrate septate junctions (S Js) have been proposed to play a role in cellular cohesion/ cell adhesion (Lane, 1991), blood-brain barrier formation (Carlson et al., 1997, 2000; Juang and Carlson, 1992, 1994; Lane, 1991), and intercellular communication (Woods and Bryant, 1991). The Wrst clue about the function of Drosophila NCP/Neurexin IV came from its subcellular localization to speciWc junctions in the epithelial and glial cells. The Drosophila protein localized to SJs in all cells that were previously shown to contain pleated SJs by electron microscopy (Baumgartner et al., 1996; Tepass and Hartenstein, 1994). Neurexin IV expression occurred just prior to the morphological appearance of the SJs and colocalization studies with other proteins that were previously shown to localize to SJs—for example, Coracle and Discs large (DLG) conWrmed that neurexin IV is a speciWc marker for pleated SJs. The pleated SJs are more prominent in perineurial glial cells that are the outermost glial cells of the CNS and PNS (Juang and Carlson, 1992, 1994; Lane, 1991; Carlson et al., 1997; see the chapter 8 by Bellen and Schulze on Invertebrate Glia for more details). These perineurial glial cells express neurexin IV in the PNS of embryos and third instar larvae (Fig. 23.3). The localization of septate junctions in the insulating glial cells that wrap around axon bundles is shown in Figure 23.4. The SJs have previously been proposed to form the blood-nerve barrier in other insects and also to slow down paracellular transport. Based on these proposed functions, they are therefore considered equivalent to vascular endothelium of higher organisms (Carlson et al., 1997; Juang and Carlson, 1992; Lane, 1991). The Drosophila nrx IVmutants are paralyzed and ultrastructural analysis revealed that the morphology of SJs is aVected (Baumgartner et al., 1996). The characteristic morphology of normal SJs as ladder-like septa had been transformed into what are referred to as smooth SJs (Baumgartner et al., 1996; Tepass and Hartenstein, 1994). Electrophysiological analyses of these mutants clearly indicated a defect in barrier formation, because by varying the extracellular Kþ concentration, the proWle of spontaneous neural activity in mutants was severely aVected. Such changes in Kþ concentration had no eVect in wild-type embryos (Baumgartner et al., 1996). Hence, the key function of neurexin IV in Drosophila embryos is to form SJs, which serve as barriers to prevent ions and other small molecules from breaching into neighboring cells. This function could very well be established by proper localization of other macromolecules at the SJs that are required in controlled ion and small molecular transport (Lane, 1991). These in vivo studies demonstrated that Drosophila neurexin IV plays a critical role in establishing the septate junctions between glial cells to preserve the ionic microenvironment of the neuronal cells and axons, which is required for action potential propagation. NCP1 and Septate Junctions in Vertebrates As discussed above, the presence of vertebrate homologs of Drosophila neurexin IV became evident during a search for candidate breast cancer genes that mapped to human chromo-

au4

MEMBERS OF THE NEUREXIN SUPERFAMILY

FIGURE 23.3 Drosophila neurexin IV is expressed in epithelial and perineurial glial cells. (A) A stage 15 wild-type embryo doublelabeled with anti-neurexin IV antibody (green) and sensory neuronal membrane speciWc mAb 22CI0 (red). Drosophila neurexin IV is expressed in ectodermal epithelial cells and the perineurial glial cells of the PNS and CNS. In the PNS, the protein is expressed in the perineurial glia (small green arrows) insulating the axons. Neurexin IV is also expressed in midline glial cells (large green arrow). The sensory neurons in the PNS are stained with 22C10 (large red arrow). (B) A stage 15 wild type embryo double-labeled with anti-neurexin IV antibody (green) and motor neuron speciWc anti-fasciclin II antibody (red). The anti-fasciclin II antibody stains the motor axons in the periphery (small red arrow) and in the longitudinal axon tracts in the CNS (large red arrow). The small and large green arrows highlight the perineurial glial cells and the midline glial cells respectively. (C) The whole third instar larval CNS with brain lobes (BL) and ventral nerve cord (VNC). Immunolocalization of neurexin IV in the larval VNC and BLs shows that the protein is expressed in midline glia as well as at the edges of the interdigitating perineurial glial cells required for the maintenance of the blood–brain barrier in third instar larvae (white arrows). Perineurial glia are very large cells that ensheath the whole CNS. In the ventral nerve cord, only eight cells per segment are required to circumference the nerve cord. *corresponds to a single perineurial glial cells surrounding the BLs.

FIGURE 23.4 Schematic diagram of a Drosophila abdominal nerve in cross-section. The invertebrate nerve consists of bundles of axons (Ax) surrounded by a sheath formed by a single glial cell (G) similar to oligodendrocytes in the vertebrate CNS. This cell is surrounded by multiple perineurial glial cells (P) that form septate junctions shown at a higher magniWcation as white ladder-like structures. Drosophila neurexin IV localizes to these septate junctions, which are required to form the blood–nerve barrier preventing hemolymph from the hemocoel (HC) to enter the nervous system. The whole nerve Wber is surrounded by neuralemma (NL).

587

588

23. THE NEUREXIN AND NCP GENE FAMILIES

some 17q21 (Brody et al., 1995; Friedman et al., 1995). Since this new human neurexin-like gene contained a discoidin domain and vertebrate neurexins I to III lacked a discoidin domain, it was proposed that this novel gene was a homolog ofDrosophila neurexin IV and was therefore referred to as hNRXIV (Baumgartner et al., 1996). The complete cDNA sequence for this neurexin-like gene was independently established by several groups and named as contactin-associated protein (caspr) (Peles et al., 1997), as Paranodin (Menegoz et al., 1997), as mouse homolog of neurexin IV (accession number AF039833; Bellen et al., 1998) and recently as NCP1 (Bhat et al., 2001; Fig. 23.5). The human and rat proteins were identiWed as a 190 kDa proteins that co-precipitated with contactin bound to carbonic anhydrase (CAH) domain of receptor phosphotyrosine phosphatase-p and proposed to play a role in contactin-mediated signal transduction (Peles et al., 1997). Simultaneously, Menegoz et al., (1997) puriWed a 180-kDa neuronal glycoprotein from rat brain using concanavalin A (a lectin) aYnity chromatography. Immunohistochemical localization showed that the protein was expressed throughout the neuropil in the cerebellum and highly enriched in the paranodal area of the nodes of Ranvier in both the CNS and PNS, hence the name paranodin. This protein was initially expressed at the onset of myelination and became localized progressively to the paranodal sites during neuronal maturation (Menegoz et al., 1997). The protein levels are downregulated during myelination and this down-regulation and axonal localization is paired with a dramatic redistribution to the paranodal space during the formation of the nodes. The paranodal localization of this protein was subsequently conWrmed and it was established that this protein was speciWcally expressed by neurons and not by Schwann cells or oligodendrocytes (Einheber et al., 1997). Immunoelectron microscopy studies further demonstrated that the protein is an axonal component localized to the septate-like junctions of the paranodal region that are formed by the noncompact myelin loops and the paranodal axonal axolemma (Einheber et al., 1997). These studies demonstrated that Drosophila neurexin IV and vertebrate NCP1 are functional homologs and both localize to septate junctions. Additional Members of the NCP Family: Caspr2, Caspr3, Caspr4, and CasprS Whole genome sequence information in human and mouse allowed the identiWcation of additional members of the NCP family. Recently, the second member of the NCP family named Caspr2 has been localized to a region next to the paranodal area and distal to the node that is referred to as juxtaparanodal domain. This domain is covered by the compact myelin sheath and is also considered as a specialized portion of the internode (Poliak et al., 1999). In addition to Caspr2, the Shaker-type delayed-rectiWer Kþ channels Kvl.l, Kvl.2, and their cytoplasmic subunit KV02 are also highly enriched in the juxtaparanodal region and have been suggested to promote membrane repolarization and to maintain the inter-

FIGURE 23.5 Localization of NCP1 in mouse sciatic nerve Wbers. The teased sciatic nerves from wild-type mice were stained with antibodies to NCP1 (blue, blue arrows pointing to paranodal domains), Naþ channels (green, green arrows pointing to the nodal domain), and Kþ channels (red, red arrows pointing to the juxtaparanodal domains). The paranodal staining of NCP1 ensures the demarcation and separation of Naþ and Kþ channels at the node of Ranvier (Bhat et al., 2001).

ROLE OF NCP1 AT PARANODAL JUNCTIONS AND DOMAIN ORGANIZATION AT THE NODE OF RANVIER

nodal resting potential (Rasband et al., 1998; Vabnick et al., 1999; Wang et al., 1993). Based on their colocalization at the juxtaparanodes, immunoprecipitation experiments showed that Caspr2, Kvl.2, and Kvb2 associated into a molecular complex, but Caspr2 did not associate with the Kv2.1 subunit (Poliak et al., 1999). It would be interesting to know the consequences of loss of Caspr2 on the juxtaparanodal domain organization and clustering and localization of Kþ channels. Two more members of the NCP family were recently identiWed: Caspr3, and Caspr4, which are expressed in the nervous system (Spiegel et al., 2002). Caspr3 is expressed along axons in the corpus callosum, spinal cord, basket cells in the cerebellum, and in peripheral nerves, as well as in oligodendrocytes. Caspr4 shows a more restricted expression and is present in speciWc neuronal subpopulations in the olfactory bulb, hippocampus, deep cerebellar nuclei, and the substantia nigra. The cytoplasmic domains of Caspr3 and Caspr4 interact diVerentially with PDZ domain-containing proteins of the CASK/Lin2Veli/Lin7-Mintl/LinlO complex (Spiegel et al., 2002). The molecular function of these additional NCP members and what role they play in cell-cell interaction or signal transduction in the nervous system remains to be elucidated. A Wfth member of the NCP family, Caspr5, has been identiWed. It is expressed in the nervous system, but no functional data are available for this member of the NCP family.

ROLE OF NCP1 AT PARANODAL JUNCTIONS AND DOMAIN ORGANIZATION AT THE NODE OF RANVIER The myelinated Wbers are organized into anatomically distinct domains. Recent identiWcation of molecular markers have conWrmed these anatomical domains, which can now be identiWed as molecular domains. These domains are the node of Ranvier, the paranodal and juxtaparanodal regions, and the intemode (Arroyo and Scherrer, 2000; Peles and Salzer, 2000; Salzer, 1997). The molecular organization at these domains is thought to play a very critical role in fast conduction of nerve impulses via saltatory conduction. The highly complex molecular interactions between the myelinating glial cells (Schwann cells in the PNS and oligodendrocytes in the CNS) and the underlying axonal axolemma are still poorly understood. Recent studies have signiWcantly contributed to a better understanding of the molecular composition of these domains, establishing the presence of several protein complexes that include voltage gated Naþ channels that interact with ankyrina and also cell adhesion molecules like Neurofascin 186 at the node, NCPl/Caspr/paranodin and contactin at the paranodes and delayed rectiWer Kþ channels of the Drosophila Shaker family and Caspr2/NCP2 at the juxtaparanodal region. The contributions of NCP family members (NCP1 and NCP2) at the paranodes and juxtaparanodes, respectively, in the domain organization of myelinated axons is currently being investigated. Flanking on either side of the node of Ranvier, the paranodal space is created by noncompact myelin lamellae or loops that spiral around the axon in close proximity to the axonal axolemma. These noncompact cytoplasm-Wlled myelin loops, in association with the axonal surface, generate a series of septate like junctions, which are also referred to as paranodal junctions (Figs. 23.5 and 23.6). These paranodal junctions have been proposed to perform several functions at this axo-glial interface, including (1) anchoring of the glial myelin loops to the axon, (2) creating an ionic diVusion barrier into the periaxonal space, and (3) serving as a fence to maintain the axonal domains and preventing lateral diVusion of the various membrane protein complexes. The close apposition of the axonglial interface at the paranodes could serve as a potential site for bidirectional signaling between axons and glial cells. The septate junctions formed between the myelinating glial cells and the axonal surface begin to form during early postnatal development, but their development is complete by the third postnatal week (Garcia-Fresco et al., unpublished data). The formation or the appearance of the junctions begins at the proximal side of the node with outer most myelin loop connecting Wrst and gradually continuing into the distal end as additional loops make

589

590

23. THE NEUREXIN AND NCP GENE FAMILIES

FIGURE 23.6 Paranodal morphology in wild-type and NCP1 mutant mice. Electron micrographs through the paranodal region of the spinal cords in the wild-type (þ/þ) and NCP1 mutant mice (
/
) show the presence of the characteristic transverse septa at the wild-type paranodes shown by red arrowheads in the upper panel. The NCP1 mutant spinal cords show loss of transverse septa and display an evertion of paranodal loops away from the axonal Axolemma shown by red arrowheads in the lower panel (see Bhat et al., 2001).

connections with the axonal surface (Pedraza et al., 2001; TaoCheng and Rosenbluth, 1983). As the myelin sheath wraps around, at each turn it generates contiguous rings around the axon with septate junctions, which can be seen as indentations in the freeze fracture analysis (Tao Chang and Rosenbluth, 1983). The molecules that have been localized to paranodal septate junctions include NCPl/Caspr/Paranodin (Bhat et al., 2001; Einheber et al., 1997, Menegoz et al., 1997) and NCP 1-interacting protein, Contactin, which is anchored into the membrane through a glycosylphosphatidylinositol (GPI) moiety (Bhat et al., 2001; Poliak et al., 2001; Rios et al., 2000) and band 4. IB (Ohara etal, 2000). The components of paranodal junctions on the glial side, which interact with the NCPl/contactin complex at the septate junctions, remain largely unknown. Since the axonal proteins NCP 1/Caspr/Paranodin and contactin form a cis complex in the axonal axolemma at the paranodal axon-glial interface, a putative glial protein must interact with this complex at these junctions. A glial protein, neurofascin 155 (155kDa isoform) is expressed at the paranodal loops and may be the putative glial receptor for the NCP 1/Caspr-contactin complex (Tait et at., 2000). It was recently reported that the extracellular domain of neurofascin 155 binds speciWcally to transfected cells expressing the NCPlcontactin complex at the cell surface and to the same complex from brain lysates. These studies suggest that NF155 may be the glial ligand of NCPl/cntactin complex to establish septate junctions at the paranodes (Charles et al., 2002). The Wrst molecular evidence in support of the role of paranodal septate junctions in domain organization emerged from studies of ceramide galactosyl transferase (CGT) mutant mice (Bosio et al., 1998; Coetzee et al., 1996). The CGT mutant mice are unable to synthesize galactocerebroside and sulfatide, two glycolipids that are very abundant in myelin. These mice display a variety of abnormalities at the paranodal and nodal regions, including absence of the transverse bands—the hallmark of the paranodal junctions (Dupree et al., 1999). It is not clear how the loss of myelin galactolipids causes the underlying abnormalities at the paranodes (Dupree and Popko, 1999). The major Wnding

au5

ROLE OF NCP1 AT PARANODAL JUNCTIONS AND DOMAIN ORGANIZATION AT THE NODE OF RANVIER

591

from these studies was that Kþ channels were mislocalized from the juxtaparanodes into paranodes in these CGT mutant mice (Dupree et al., 1999), thus highlighting that the paranodal junctions play a key role in the delineation of channel domains at and around the node. Recently, NCP1 mutant mice were generated that displayed phenotypes similar to that of CGT mice (Bhat et al., 2001). Since NCP1 is a paranodal septate junction speciWc protein, loss of NCP 1 in mice results in loss of the transverse septa at the paranodes, severe neurologic defects and aberrant organization, and axo-glial interactions in the paranodal region (Fig. 23.6). In addition, the strict separation seen between Naþ channels at the node and Kþ channels at the juxtaparanode is abolished and nerve conduction velocity is substantially reduced in the absence of NCP1 (Fig. 23.7). Hence, NCP1 plays a key role in the formation of the paranodal junctions and in establishing and maintenance of molecular domains at the node of Ranvier required for normal saltatory conduction (Bhat et al., 2001; Fig, 23.7). It was recently shown that the export of NCPl/Caspr from the endoplasmic reticulum to the plasma membrane in transfected CHO or neuroblastoma cells depends on its interaction with contactin. In the absence of contactin, NCP1 failed to reach to the surface of the transfected cells and from soma to axon in the neurons in contactin mutant animals (Boyle et al., 2001; Faivre-Sarrailh et al., 2000). The paranodal junctional complex is completely destabilized in the absence of NCP1, as both contactin and Neurofascin 155 get mislocalized or degraded at the paranodes (Bhat et al., 2001, Boyle et al., 2001, Poliak et al., 2001), further highlighting the role of NCP1 in the organization of the paranodal septate junctions and domains in the myelinated axons. These studies showed that NCP1 and contactin are essential for the generation of the paranodal junctions, and their absence results in the loss of the transverse bands or septa that are the hallmark of the paranodal axo-glial contact, resulting in a lower conduction velocity (Bhat et al., 2001, Boyle et al., 2001). The loss of CGT, contactin, and NCP1 phenocopied each other in the peripheral nerve Wbers, where they all displayed similar phenotypes—that is, loss of paranodal septate

FIGURE 23.7 Voltage-gated channel distribution is perturbed in NCP1 mutant mice. (A) Teased sciatic nerves from wild type mice were stained with antibodies to ion channels. In the wild-type sciatic nerves, the Naþ channels (green) and Kþ channels (red) are completely separated by NCP1 (blue) in the paranodes. (B) In the NCP1 mutant nerves, Kþ channels are aberrantly localized to the paranodes and overlap slightly with Naþ channels (regions of overlap are yellow). Thus, the molecular fence function provided by paranodal septate junctions is lost in the absence of the NCP1 protein (Bhat et al., 2001).

592

23. THE NEUREXIN AND NCP GENE FAMILIES

junctions and mislocalization of Kþ channels to the paranodal area from the juxtaparanodal space (Bhat et al., 2001; Boyle et al., 2001; Dupree et al., 1999). These mutant phenotypes suggest that paranodal junctions may perform distinct functions and serve as barriers to restrict the Xow of ions at the paranodes from entering the periaxonal space and act as fences by restricting the diVusion or transport of proteins from crossing into other domains. This barrier and fence function of paranodal junctions is similar to that proposed for epithelial tight junctions, which also restrict paracellular transport and movement of proteins into apical domains. Thus the gene ablation studies in mice demonstrated that NCP1 at septate junctions in the paranodes plays a critical role in performing a fence role to maintain the molecular separation of ion channels at the node.

NEUREXINS LINK SEPTATE JUNCTIONS WITH THE CYTOSKELETON THROUGH BAND 4.1 PROTEINS All neurexin superfamily members including neurexins I to III and NCP 1–5 of vertebrates and neurexin I and IV of Drosophila, contain a conserved Band 4.1-binding domain (Marfatia et al., 1995). This binding site is in close proximity and located a few amino acids away from the C terminus of the transmembrane domain (Littleton et al., 1997). Immunocolocalization of neurexin IV and Coracle (Band 4.1 homolog) in Drosophila embryos showed that this domain plays an important role in linking septate junctions to the actin cytoskeleton. A Drosophila neurexin IV allele that made essentially all the extracellular domain but truncated close to the transmembrane domain resulted in mislocalization of the septate junction protein, Coracle (Baumgartner, et al., 1996). In these mutants Coracle remained diVused in the cytoplasm and plasma membrane and did not localize to the septate junctions. The coracle mutants also displayed phenotypes similar to that of neurexin IV mutants (i.e., loss of septate junctions and dorsal closure defects). In addition, neurexin IV localization was not restricted to septate junctions but showed more basolateral localization in coracle mutants (Ward et al., 1998). Thus a molecular interaction between the septate junction speciWc protein, neurexin IV and an actin cytoskeletal protein, coracle, links these junctions with the cellular cytoskeleton (Fig. 8). Vertebrates contain four diVerent Band 4.1 proteins that show expression in the nervous system with varying subcellular localizations (Ohara et al., 1999; Parra et al., 2000; Walensky et al., 1999). One of these 4.1 proteins, 4.1B, speciWcally localizes to paranodal and juxtaparanodal regions in the myelinated Wbers (Ohara et al., 2000). Menegoz et al., (1997) demonstrated that partially puriWed Band 4.1 from erythrocytes and rat brain was able to bind rat NCP 1/caspr/Paranodin, thus showing a direct interaction between Band 4.1 and NCP 1. Based on these observations it seems likely that most members of the neurexin superfamily, which contain this consensus sequence will bind to Band 4.1 (Littleton et al., 1997). As in Drosophila, the localization of paranodal junction speciWc Band 4.1 protein, 4. IB, is aVected in NCP1 mutants. The protein is diVusely present at the paranodes, however, the total protein levels do not seem to change in NCP1 mutants (Garcia-Fresco and Bhat, unpublished data). Similarly, protein 4.IB was abnormally distributed along peripheral myelinated axons of CGT deWcient mice, which display mislocalization of NCPl/contactin complex at their paranodes (Poliak et al., 2001). Members of the Band 4.1 family contain the FERM domain (F for 4.1 protein, E for ezrin, R for radixin and M for moesin). This name reXects the initial identiWcation of this domain in a family of peripheral membrane proteins that function as membrane-cytoskeleton linkers (Chishti et al., 1998). These proteins act as linkers between the plasma membrane and the cytoskeleton by interacting with F-actin and with each other in vertebrate cells (Vaheri et al., 1997). Mutations in the gene encoding the merlin protein are involved in tumor formation, including schwannomas and meningiomas, but the process by which these tumors arise is poorly understood (Gusella et al., 1996). It remains to be seen whether the disruption of the molecular link between septate junctions and cytoskeleton at the paranodes results in axonal cytoskeletal defects and neuronal pathology in NCP1 or CGT deWcient mice.

au6

FUNCTIONAL CONSERVATION AND EVOLUTIONARY SIGNIFICANCE

FUNCTIONAL CONSERVATION AND EVOLUTIONARY SIGNIFICANCE IdentiWcation of vertebrate neurexins and their invertebrate counterparts show that the members of the neurexin superfamily constitute a family of highly conserved neuronal cell adhesion molecules. The neurexin family thus includes the three vertebrate neurexins and single neurexin homologs in Drosophila and C. elegans (Fig. 23.1A). These members display identical domain structure in their extracellular domain, transmembrane regions, and cytoplasmic C-terminal domains. Vertebrate studies have shown that neurexin genes I to III display overlapping expression patterns in diVerent classes of neuronal populations (Ullrich et al., 1995). Biochemical studies in vertebrates suggested that neurexins interact with other proteins in an isoform speciWc manner, for example, the long extracellular domain containing a-neurexins interacts with neurexophilin and short extracellular domain containing b-neurexins interact with neuroligins. The biochemical interactions between these proteins have not made signiWcant impact on the in vivo function of the neurexins. Mutational analysis of the vertebrate neurexin genes has been hampered due to redundant expression patterns and absence of any obvious phenotype in NRX la mutants. Mutational analysis of the Drosophila neurexin 1 homolog is currently underway that will provide novel insights into the in vivo function of this neurexins in neural development and function (Li and Bhat, unpublished data). The members of the NCP family include Wve vertebrate proteins (NCP1–5), Drosophila neurexin IV and a C. elegans NCP like gene (Fig. 23.1B). Genetic and molecular characterization of the only Drosophila member of the NCP family (i.e., neurexin IV) uncovered its novel functions in septate junction and blood-nerve barrier formation. The septate junctions form between the ensheathing glial cells that surround the nerve bundles in Drosophila and are essential for establishment of the blood-nerve barrier (Bellen et al., 1998). It was previously shown that the Drosophila NRX IV protein localizes to and is required for the formation of the ladder-like septae characteristic of these junctions (Baumgartner et al., 1996). Recent mutational analysis of the vertebrate homolog of neurexin IV—that is, NCPl/ Caspr/Paranodin demonstrated that the septate and paranodal junctions serve conserved functions in maintaining the axonal milieu required for action potential propagation. These mutants lacked paranodal septate junctions as seen in Drosophila neurexin IV mutants, and further loss of NCP 1 aVected the domain organization at the node of Ranvier. However, the topology and localization of the proteins is clearly diVerent. In Drosophila, neurexin IV is expressed by and localized between glial cells where, as in mice and other vertebrates, NCP1 is expressed by neurons and localized between the axon and glial cells. Hence, even though there is functional conservation, signiWcant changes in expression occurred during evolution. These new studies have provided novel insights into the function of the neurexin family in organizing cellular junctions either in epithelial and glial cells in Drosophila or at axo-glial contact sites in the vertebrates (see Fig. 23.8). Essential questions still remain to be answered concerning the function of vertebrate neurexins. A precise subcellular localization of the known vertebrate neurexins is required to determine if they may also be involved in axonalglial interactions, similar to that reported for the NRXIV and NCPl/Caspr/Paranodin. What role do neurexins/NCPs play in synaptogenesis and ion channel clustering? What are the functional ligands of neurexins/NCPs in vertebrates and invertebrates? Are these molecules part of a signal transduction pathway and if so, what are the down stream eVectors of these signaling pathways? Using a combined approach of genetics and cell biology, a functional dissection of the role of neurexins in cell junction formation or synaptogenesis can now be undertaken in various model systems.

Acknowledgments I would like to thank my colleagues German P. Garcia-Fresco, Jingjun Li, Afshan Ismat, and Swati Banerjee for their comments on this article and many thoughtful discussions; and German P. Garcia-Fresco for making the schematic Wgures. Work in my laboratory is

593

594

23. THE NEUREXIN AND NCP GENE FAMILIES

FIGURE 23.8 Schematic representation of the molecular domains at the node of Ranvier. The nodal complex is established by inter- and intra-molecular interactions between ankyrin, spectrin, sodium channels, and other cell adhesion molecules, and this macromolecular complex remains at the node. The sodium channels clustered at the node play a critical role in saltatory conduction. The paranodal complex that is the hallmark of the paranodal septate junctions is established by NCP1, contactin, 4.1B, and other yet unidentiWed proteins on the axonal side and possibly NF155 and other yet unidentiWed proteins on the glial side. How these junctions are established at the axo-glial contact sites is still not clear? The juxtaparanodal complex contains NCP2, potassium channels, and cell adhesion molecules. The molecular events that underlie the formation of juxtaparanodal region are also not well established. At both the paranodal and juxtaparanodal domains, NCP1 and 2, respectively, link these domains with the actin cytoskeleton via band 4.1B protein, which is expressed at the para- and juxta-paranodes.

supported by grants from the National Institute of General Medical Sciences and the National Cancer Institute at the National Institutes of Health, the Hirschl Foundation, and Cardiovascular Research Institute of the Mount Sinai School of Medicine.

References Arroyo, E. J., and Scherer, S. S. (2000). On the molecular architecture of myelinated Wbers. Histochem. Cell Biol. 3, 1–18. Baumgartner, S., Littleton, J. T., Broadie, K., Bhat, M. A., Harbecke, R., Lengyel, J. A., Chiquet-Ehrismann, R., Prokop, A., and Bellen, H. J. (1996). A Drosophila neurexin is required for septate junction and blood-nerve barrier formation and function. Cell 87, 1059–1068. Bellen, H. J., Lu, Y., Beckstead, R., and Bhat, M. A. (1998). Neurexin IV, caspr and paranodin-novel members of the neurexin family: Encounters of axons and glia. Trends Neurosci. 21, 444–449. Bhat, M. A., Izadoost, S., Lu, Y., Cho, K., Choi, K., and Bellen, J. B. (1999). Discs Lost, a Novel Multi-PDZ Domain Protein, Establishes and Maintains Epithelial Polarity. Cell 96, 833–845. Bhat, M. A., Rios, J. C., Lu, Y., Garcia-Fresco, G. P., Ching, W., St. Martin, M., Li, J., Einheber, S., Chesler, M., Rosenbluth, J., Salzer, J. L., and Bellen H. J. (2001). Axon-glia interactions and the domain organization of myelinated axons requires neurexin IV/Caspr/Paranodin. Neuron 30, 369–383. Bosio, A., Bussow, H., Adam, J., and StoVel, W. (1998). Galactosphingolipids and axon-glial interaction in myelin of the central nervous system. Cell Tissue Res. 292, 199–210.

au7

595

FUNCTIONAL CONSERVATION AND EVOLUTIONARY SIGNIFICANCE

Boyle, M. E., Berglund, E. O., Murai, K. K., Weber, L., Peles, E., and Ranscht, B. (2001). Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30, 385–397. Brody, L. C., Abel, K. J., Castilla, L. H., Couch, F. J., McKinley, D. R., Yin, G., Ho, P., Merajver, S., Chandrasekharappa, S. C., Xu, J., Cole, J. L., Struewing, J. P., Vaides, J. M., Collins, F. S., and Weber, B. L. (1995). Construction of a transcription map surrounding the BRCA1 locus of human chromosome 17. Genomics 25, 238–247. Butz, S., Fernandez-Chacon, R., Schmitz, F., Jahn, R., and Siidhof, T. C. (1999). The subcellular localizations of atypical synaptotagmins: Synaptotagmin III is enriched in synapses and synaptic plasma membranes but not in synaptic vesicles. J. Biol. Chetn. 274, 18290–18296. Carlson, S. D., Hilgers, S. L., and Juang, J. L. (1997). infrastructure and blood-nerve barrier of chordotonal organs in the Drosophila embryo./. Neurocytol. 6, 377–388. Carlson, S. D., Juang, J. L., Hilgers, S. L., and Garment, M. B. (2000) Blood barriers of the insect. Annu. Rev. Entomol. 45, 151–174. Charles, P., Tait, S., Faivre-Sarrailh, C., Barbin, G., Gunn-Moore, F., Denisenko-Nehrbass, N., Guennoc, A. M., Girault, J. A., Brophy, P. J., and Lubetzki, C. (2002). Neurofascin is a glial receptor for the paranodin/Casprcontactin axonal complex at the axoglial junction. Curr. Biol. 12, 217–220. Chishti, A. H., Kim, A. C., Marfatia, S. M., Lutchman, M., Hanspal, M., Jindal, H., Liu, S. C., Low, P. S., Rouleau, G. A., Mohandas, N., Chasis, J. A., Conboy, J. G., Gascard, P., Takakuwa, Y., Huang, S. C., Benz, E. J. Jr., Bretscher, A., Fehon, R. G., Gusella, J. F., Ramesh, V., Solomon, F., Marchesi, V. T., Tsukita, S., Tsukita, S., Hoover, K. B., et al. (1998). The PERM domain: A unique module involved in the linkage of cytoplasmic proteins to the membrane. Trends Biochem. Sci. 23, 281–282. Coetzee, T., Fujita, N., Dupree, J., Shi, R., Blight, A., Suzuki, K., and Popko, B. (1996). Myelination in the absence of galactocerebroside and sulfatide: Normal structure with abnormal function and regional instability. Cell 86, 209–219. Dimitratos, S. D., Woods, D. P., and Bryant, P. J. (1997). Camguk, Lin-2, and CASK: Novel membraneassociated guanylate kinase homologs that also contain CaM kinase domains. Mech. Dev. 63, 127–130. Dupree, J. L., Girault, J. A., and Popko, B. (1999). Axo-glial interactions regulate the localization of axonal paranodal proteins. J. Cell Biol. 147, 1145–1152. Dupree, J. L., and Popko, B. (1999). Genetic dissection of myelin galactolipid function. J. Neurocytol. 28, 271–279. Einheber, S., Zanazzi, G., Ching, W., Scherer, S., Milner, T. A., Peles, E., and Salzer, J. L. (1997). The axonal membrane protein Caspr, a homologue of neurexin IV, is a component of the septate-like paranodal junctions that assemble during myelination. J. Cell Biol. 139, 1495–1506. Faivre-Sarrailh, C., Gauthier, F., Denisenko-Nehrbass, N., Le Bivic, A., Rougon, G., and Girault, J. A. (2000). The glycosylphosphatidyl inositol-anchored adhesion molecule F3/Contactin is required for surface transport of Paranodin/Contactin-associated protein (caspr). J. Cell Biol. 149, 491–502. Friedman, L. S., Ostermeyer, E. A., Lynch, E. D., Welsch, P., Szabo, C. I., Meza, J. E., Anderson, L. A., Dowd, P., Lee, M. K., Rowell, S. E., Ellison, J., Boyd, J., and King, M. C. (1995). 22 genes from chromosome 17q21: Cloning, sequencing, and characterization of mutations in breast cancer families and tumors. Genomics 25, 256–263. Geppert, M., Goda, Y., Hammer, R. E., Li, C., Rosahl, T. W., et al. (1994). Synaptotagmin I: A major Caþþ sensor for transmitter release at a central synapse. Cell 79, 717–727. Gusella, J. F., Ramesh, V., MacCollin, M., and Jacoby, L. B. (1996). NeuroWbromatosis 2: Loss of merlin’s protective spell. Curr. Opin. Genet. Dev. 6, 87–92. Hata, Y., Davletov, B., Petrenko, A. G., Jahn, R., and Sudhof, T. C. (1993). Interaction of synaptotagmin with the cytoplasmic domains of neurexins. Neuron 10, 307–315. Hata, Y., Butz, S., and Sudhof, T. C. (1996). CASK: A novel dlg/PSD95 homolog with an N-terminal calmodulindependent protein kinase domain identiWed by interaction with neurexins. J. Neurosci. 16, 2488–2494. Hsueh, Y. P., Yang, F. C., Kharazia, V., Naisbitt, S., Cohen, A. R., et al. (1998). Direct interaction of CASK/ LIN-2 and syndecan heparan sulfate proteoglycan and their overlapping distribution in neuronal synapses./. Cell Biol. 142, 139–151. Ichtchenko, K., Hata, Y., Nguyen, T., Ullrich, B., Missler, M., Moomaw, C., and Sudhof, T. C. (1995). Neuroligin 1: A splice site-speciWc ligand for p-Neurexins. Cell 81, 435–443. Ichtchenko, K., Nguyen, T., and Sudhof, T. C. (1996). Structures, alternative splicing, and neurexin binding of multiple neuroligins. J. Biol. Chem. 271, 2676–2682. Me, M., Hata, Y., Takeuchi, K., Ichtchenko, K., Toyoda, A., Hirao, K., Takai, Y., Rosahl, T. W., and Sudhof, T. C. (1997). Binding of neuroligins to PSD-95. Science 277, 1511–1515. Juang, J. L., and Carlson, S. D. (1992). Fine structure and blood-brain barrier properties of the central nervous system of a dipteran larva. J. Comp. Neurol. 324, 343–352. Juang, J. L., and Carlson, S. D. (1994). Analog of vertebrate anionic sites in blood-brain interface of larval Drosophila. Cell Tissue Res. 277, 87–95. Lane, N. J. (1991). Morphology of glial blood-brain barriers. Ann. N. Y. Acad. Sci. 633, 348–362. Littleton, J. T., Bhat, M. A., and Bellen, H. J. (1997). Deciphering the function of neurexins at cellular junctions./. CellBiol. 137, 793–796. Marfatia, S. M., Leu, R. A., Branton, D., and Chishti, A. H. (1995). IdentiWcation of the protein 4.1 binding interface on glycophorin C and p55, a homologue of the Drosophila discs-large tumor suppressor protein. J. Biol. Chem. 270, 715–719.

au8

au9 au10

au11

596

23. THE NEUREXIN AND NCP GENE FAMILIES

Menegoz, M., Caspar, P., Le Bert, M., Galvez, T., Burgaya, F., Palfrey, C., Ezan, P., Arnos, F., and Girault, J. A. (1997). Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron 19, 319–331. Michele, D. E., Barresi, R., Kanagawa, M., Saito, F., Cohn, R. D., Satz, J. S., Dollar, J., Nishino, L, Kelley, R. I., Somer, H., Straub, V., Mathews, K. D., Moore, S. A., and Campbell, K. P. (2002). Post-translational disruption of dystroglycan-ligand interactions in congenital muscular dystrophies. Nature 418, 417–22 Missler, M., Femandez-Chacon, R., and Siidhof, T. C. (1998a). The making of neurexins./. Neurochem. 71, 1339–1347. Missler, M., Hammer, R. E., and Siidhof, T. C. (1998b). Neurexophilin binding to a-neurexins. A single LNS domain functions as an independently folding ligand-binding unit. J. Biol Chem. 273, 34716–34723. Missler, M., and Siidhof, T. C. (1998a). Neurexins: Three genes and 1001 products. Trends Genet. 14, 20–25. Missler, M., and Siidhof, T. C. (1998b). Neurexophilins form a conserved family of neuropeptide-like glycoproteins. J. Neurosci. 18, 3630–3638. Moore, S. A., Saito, F., Chen, J., Michele, D. E., Henry, M. D., Messing, A., Cohn, R. D., Ross-Barta, S. E., Westra, S., Williamson, R. A., Hoshi, T., and Campbell KP. (2002). Deletion of brain dystroglycan recapitulates aspects of congenital muscular dystrophy. Nature 418, 422–425. Nguyen, T., and Siidhof, T. C. (1997). Binding properties of neuroligin 1 and neurexin ip reveal function as heterophilic cell adhesion molecules. J. Biol. Chem. 272, 26032–26039. Ohara, R., Yamakawa, H., Nakayama, M., and Ohara, O. (2000). Type II brain 4.1 (4.1B/KIAA0987), a member of the protein 4.1 family, is localized to neuronal paranodes. Brain Res. Mol. Brain Res. 85, 41–52. Ohara, R., Yamakawa, H., Nakayama, M., Yuasa, S., and Ohara, O. (1999). Cellular and subcellular localization of a newly identiWed member of the protein 4.1 family, brain 4.1 in the cerebellum of adult and postnatally developing rats. Brain Res. Dev. Brain Res. 117, 127–138. Parra, M., Gascard, P., Walensky, L. D., Gimm, J. A., Blackshaw, S., Chan, N., Takakuwa, Y., Berger, T., Lee, G., Chasis, J. A., Snyder, S. H., Mohandas, N., and Conboy, J. G. (2000). Molecular and functional characterization of protein 4. IB, a novel member of the protein 4.1 family with high level, focal expression in brain. J. Biol. Chem. 275, 3247–3255. Patzke, H., and Ernsberger, U. (2000). Expression of neurexin I alpha splice variants in sympathetic neurons: Selective changes during diVerentiation and in response to neurotrophins. Mol. Cell Neurosci. 15, 561–572. Pedraza, L., Huang, J. K., and Colman, D. R. (2001). Organizing principles of the axoglial apparatus. Neuron 30, 335–344. Peles, E., Nativ, M., Lustig, M., Grumet, M., Schilling, J., Martinez, R., Plowman, G. D., and Schlessinger, J. (1997). IdentiWcation of a novel contactin-associated transmembrane receptor with multiple domains implicated in protein-protein interactions. EMBO J. 16, 978–988. Peles, E., and Salzer, J. L. (2000). Molecular domains of myelinated axons. Curr. Opin. Neurobiol. 10, 558–865. Perin, M. S. (1994). The COOH terminus of synaptotagmin mediates interaction with the Neurexins. J. Biol. Chem. 269, 8576–8581. Perin, M. S. (1996). Mirror image motifs mediate the interaction of the COOH terminus of multiple synaptotagmins with the neurexins and calmodulin. Biochemistry 29, 13808–13816. Petrenko, A. G., Kovalenko, V. A., Shamotienko, O. G., Surkova, I. N., Tarasyuk, T. A., Ushkaryov, Y. A., and Grishin, E. V. (1990). Isolation and properties of the a-latrotoxin receptor. EMBOJ. 9, 2023–2027. Petrenko, A. G., Lazaryeva, V. D., Geppert, M., Tarasyuk, T. A., Moomaw, C, Khokhlatchev, A. V., Ushkaryov, Y. A., Slaughter, C., Nasimov, I. V., and Siidhof, T. C. (1993). Polypeptide composition of the a-latrotoxin receptor. J. Biol. Chem. 268, 1860–1867. Petrenko, A. G., Perin, M. S., Bazbek, A., Davletov, B. A., Ushkaryov, Y. A., Geppert, M., and Siidhof, T. C. (1991). Binding of synaptotagmin to the a-latrotoxin receptor implicates both in synaptic vesicle exocytosis. Nature 353, 65–68. Petrenko, A. G., Ullrich, B., Missler, M., Krasnoperov, V., Rosahl, T. W., and Siidhof, T. C. (1996). Structure and evolution of neurexophilin. J. Neurosci. 16, 4360–4369. Poliak, S., Gollan, L., Martinez, R., Custer, A., Einheber, S., Salzer, J. L., Trimmer, J. S., Shrager, P., and Peles, E. (1999). Caspr2, a new member of the neurexin superfamily, is localized at the juxtaparanodes of myelinated axons and associates with Kþ channels. Neuron 24, 1037–1047. Puschel, A. W., and Betz, H. (1995). Neurexins are diVerentially expressed in the embryonic nervous system of mice./. Neurosci. 15, 2849–2856. Rasband, M. N., Trimmer, J. S., Schwarz, T. L., Levinson, S. R., Ellisman, M. H., Schachner, M., and Shrager, P. (1998). Potassium channel distribution, clustering, and function in remyelinating rat axons./. Neurosci. 8, 36–47. Rios, J. C., Melendez-Vasquez, C. V., Einheber, S., Lustig, M., Grumet, M., Hemperly, J., Peles, E., and Salzer, J. L. (2000). Contactin-associated protein (Caspr) and contactin form a complex that is targeted to the paranodal junctions during myelination. J. Neurosci. 20, 8354–8364. Rowen, L., Young, J., Birditt, B., Kaur, A., Madan, A., Philipps, D. L., Qin, S., Minx, P., Wilson, R. K., Hood, L., and Graveley, B. R. (2002). Analysis of the human neurexin genes: Alternative splicing and the generation of protein diversity. Genomics 79, 587–597. Rudenko, G., Hohenester, E., and Muller, Y. A. (2001). LG/LNS domains: Multiple functions–one business end? Trend. Biochem. Sci. 26, 363–368. Rudenko, G., Nguyen, T., Chelliah, Y., Sudhof, T. C., and Deisenhofer, J. (1999). The structure of the ligand-binding domain of neurexin 16: Regulation of LNS domain function by alternative splicing. Cell 99, 93–101.

597

FUNCTIONAL CONSERVATION AND EVOLUTIONARY SIGNIFICANCE

Russell, A. B., and Carlson, S. S. (1997). Neurexin is expressed on nerves, but not at nerve terminals, in the electric organ. J. Neurosci. 17, 4734–4743. Salzer, J. L. (1997). Clustering sodium channels at the node of Ranvier: Close encounters of the axon-glia kind. Neuron 18, 843–846. Scheer, H., and Meldolesi, J. (1985). PuriWcation of the putative a-latrotoxin receptor from bovine synaptosomal membranes in an active binding form. EMBO J. 4, 323–327. Scheer, H., Prestipino, G., and Meldolesi, J. (1986). Reconstitution of the puriWed a-latrotoxin receptor in liposomes and planar lipid membranes. Clues to the mechanism of toxin action. EMBO J. 5, 2643–2648. ScheiVele, P., Fan, J., Choih, J., Fetter, R. D., and SeraWni, T. (2000). Neuroligin expressed in nonneuronal cells triggers presynaptic development in contacting axons. Cell 101, 657–669. Song, J., Ichtchenko, K., Siidhof, T. C., and Brose, N. (1999). Neuroligin 1 is a postsynaptic cell-adhesion molecule of excitatory synapses. PNAS 96, 1100–1105. Spiegel, I., Salomon, D., Erne, B., Schaeren-Wiemers, N., and Peles, E. (2002). CasprS and caspr4, two novel members of the caspr family are expressed in the nervous system and interact with PDZ domains. Mol. Cell. Neurosci. 20, 283–297. Sugita, S., Khvochtev, M., and Siidhof, T. C. (1999). Neurexins are functional a-latrotoxin receptors. Neuron 22, 489–496. Sugita, S., Saito, F., Tang, J., Satz, J., Campbell, K., and Siidhof, T. C. (2001). A stoichiometric complex of neurexins and dystroglycan in brain. J. Cell Biol. 154, 435–445. Tabuchi, K., and Siidhof, T. C. (2002). Structure and evolution of neurexin genes: Insight into the mechanism of alternative splicing. Genomics 79, 849–859. Tait, S., Gunn-Moore, F., Collinson, J. M., Huang, J., Lubetzki, C., Pedraza, L., Sherman, D. L., Colman, D. R., and Brophy, P. J. (2000). An oligodendrocyte cell adhesion molecule at the site of assembly of the paranodal axo-glial junction. J. Cell Biol. 150, 657–666. Tao-Cheng, J. H., and Rosenbluth, J. (1983). Axolemmal diVerentiation in myelinated Wbers of rat peripheral nerves. Brain Res. 285, 251–263. Tepass, U., and Hartenstein, V. (1994). The development of cellular junctions in the Drosophila embryo. Dev. Biol 161, 563–596. Ullrich, B., Ushkaryov, Y. A., and Siidhof, T. C. (1995). Cartography of neurexins: More than 1000 isoforms generated by alternative splicing and expressed in distinct subsets of neurons. Neuron 14, 497–507. Ushkaryov, Y. A., Hata, Y., Ichtchenko, K., Moomaw, C., Afendis, S., Slaughter, C. A., and Sudhof, T. C. (1994). Conserved domain structure of p-neurexins. J. Biol. Chem. 269, 11987–11992. Ushkaryov, Y. A., Petrenko, A. G., Geppert, M., and Siidhof, T. C. (1992). Neurexins: Synaptic cell surface proteins related to the a-latrotoxin receptor and laminin. Science 257, 50–55. Ushkaryov, Y. A., and Sudhof, T. C. (1993). Neurexin Ilia: Extensive alternative splicing generates membranebound and soluble forms. Proc. Natl. Acad. Sci. USA 90, 6410–6414. Vabnick, I., Trimmer, J. S., Schwarz, T. L., Levinson, S. R., Risal, and D., Shrager, P. (1999).Dynamic potassium channel distributions during axonal development prevent aberrant Wring patterns. J Neurosci;19(2):747–758. Vaheri, A., Carpen, O., Heiska, L., Helander, T. S., Jaaskelainen, J., Majander-Nordenswan, P., Sainio, M., Timonen, T., and Turunen, O. (1997). The ezrin protein family: Membrane-cytoskeleton interactions and disease associations. Curr. Opin. Cell Biol 9, 659–666. Van Renterghem, C., Iborra, C., Martin-Moutot, N., Lelianova, V., Ushkaryov, Y., and Seagar, M. (2000). -latrotoxin forms calcium-permeable membrane pores via interactions with latrophilin or neurexin. Eur. J. Neurosci. 12, 3953–3962. Walensky, L. D., Blackshaw, S., Liao, D., Watkins, C. C., Weier, H. U., Parra, M., Huganir, R. L., Conboy, J. G., Mohandas, N., and Snyder, S. H. (1999) A novel neuron-enriched homolog of the erythrocyte membrane cytoskeletal protein 4.1. J Neurosci 19:6457–6467. Wang, H., Kunkel, D. D., Martin, T. M., Schwartzkroin, P. A., and Tempel, B. L. (1993). Heteromultimeric Kþ channels in terminal and juxtaparanodal regions of neurons. Nature 365, 75–79. Ward, R. E. 4th, Lamb, R. S., and Fehon, R. G. (1998). A conserved functional domain of Drosophila coracle is required for localization at the septate junction and has membrane-organizing activity. J. CellBiol. 140, 1463–1473. Woods, D. F., and Bryant, P. J. (1991). The discs-large tumor suppressor gene of Drosophila encodes a guanylate kinase homolog localized at septate junctions. Cell 66, 451–464. Yuan, L, L., and Ganetzky, B. (1999). A glial-neuronal signaling pathway revealed by mutations in a neurexinrelated protein. Science 283, 1343–1345.

au12

C H A P T E R

24 The Connexin32 and Connexin29 Genes Steven S. Scherer and David L. Paul

CONNEXIN GENES The connexins are a family of highly related genes encoding a group of channel-forming proteins (Bruzzone et al., 1996; White and Paul, 1999). Connexin genes have been identiWed in a variety of chordates including mammals, birds, amphibians, bony Wshes, cartilaginous Wshes, and tunicates. Arthropods and nematodes do not have connexin genes. Instead, their gap junctions contain innexins, which have almost no primary sequence relationship to connexins. In the most commonly used nomenclature, each connexin gene is named according to the organism from which it is derived and the predicted molecular mass (in kDa) of the protein that it encodes (Fig. 24.1). The mouse and human genomes contain 19 and 20 connexin genes, respectively. However, one mouse connexin (connexin33; mCx33) does not have an ortholog in the human genome database, while two human connexins (hCx25 and hCx59) have no obvious mouse orthologs. These discrepancies could mean that genomic databases are not yet complete or that some connexin genes are species-speciWc. Thus, there are potentially 21 diVerent mammalian connexin genes (Willecke et al., 2002). All connexin genes encode proteins with a similar four transmembrane domain topology (Bruzzone et al., 1996; Unger et al., 1999; White and Paul, 1999). The short cytoplasmic amino terminus, four transmembrane domains, and two extracellular loops are highly conserved phylogenetically (Fig. 24.2). The extracellular loops contain three invariant cysteines and are joined together via disulWde bonds, stabilizing their structure. These domains are critical for regulating interactions between connexins in the plasma membranes of apposing cells. The major cytoplasmic domains, consisting of the intracellular region between the second third and third transmembrane segments and the carboxy terminus, are highly variable in sequence and size. DiVerences in these regions account for the diVerent molecular masses of the connexins and are the likely basis for the many connexin-speciWc diVerences in channel gating (Bruzzone et al., 1996; Unger et al., 1999; White and Paul, 1999). The transmembrane domains are generally modeled as alpha helixes, and most studies suggest that the third transmembrane domain forms an amphipathic helix whose polar residues line the pore of the channel.

CONNEXINS FORM GAP JUNCTIONS Gap junctions are aggregations of intercellular channels, which provide a direct pathway for the diVusion of small molecules, usually between the cytoplasms of adjacent cells. They are observed in all metazoan organisms and are found in most types of cells and tissues

Myelin Biology and Disorders, Volume 1

599

Copyright 2004, Elsevier (USA). All rights reserved.

600

24. THE CONNEXIN32 AND CONNEXIN29 GENES

FIGURE 24.1 The phylogenetic relationships of connexins based on CLUSTAL algorithm. In this dendrogram, the connexin genes are named according to the predicted molecular mass of the encoded protein (in kDa). From Altevogt et al. (2002), with permission of the Society for Neuroscience.

au1

(Bruzzone et al., 1996; White and Paul, 1999). Gap junctions mediate diverse behaviors, such as propagation of electrical excitation (KirchhoV et al., 1998; Simon et al., 1998) and the control of cell proliferation and diVerentiation (Simon et al., 1997; White, 2002). Intercellular channels are composed of two apposed hemichannels (or connexons) that can form a contiguous pathway between the adjacent cells. Each hemichannel is composed of six connexins arranged to form a central pore. The permselectivity of the pore is connexin dependent but is typically large enough to allow the diVusion of ions, amino acids, and nucleotides, although it is too small to allow the diVusion of proteins or nucleic acids. Most connexins are expressed in multiple cell types and in most cell types express more than one connexin. Thus, many connexins interact in many diVerent combinations. Such interactions are potentially complex, as hemichannels may be composed of a single connexin (homomeric) or more than one connexin (heteromeric). Further, hemichannels in one cell may be composed of a diVerent set of connexins than the hemichannel of the apposing cell. When apposed cells express diVerent connexins, the intercellular channel is described as heterotypic. However, not all combinations of connexins are functional; only some result in functional intercellular channels. The potential interactions between diVerent connexins may relate to the pathogenesis of X-linked Charcot-Marie-Tooth disease (CMTX), which is caused by mutations in the Cx32 gene (see Chapter 39 by Wrabetz, et al.).

THE CX32 GENE Cx32 (also described as b1 connexin) was the Wrst connexin to be cloned (Kumar and Gilula, 1986; Paul, 1986). It is highly conserved across all mammalian species; for example, the amino acid sequence of human Cx32 is 98% identical to the mouse Cx32 (Fig. 24.2A). Like most connexins, the coding region of Cx32 is contained within a single exon (Fig. 24.3). However, the Cx32 gene contains three alternative promoters (Hennemann et al., 1992; So¨hl et al., 2001b). The human, rat, and bovine Cx32 genes have similar structures (Duga et al., 1999; Miller et al., 1988; Neuhaus et al., 1995, 1996; So¨hl et al., 1996). Cx32 transcripts in peripheral nerve and brain are mainly initiated at the promoter termed P2— the one nearest the second exon (Neuhaus et al., 1996; So¨hl et al., 1996). Transcripts in the liver, embryonic stem cells, and pancreas are initiated at the P1 promoter, while the P3 promoter is active only in oocytes (Neuhaus et al., 1996; So¨hl et al., 2001b). In transgenic mice, a 1.8 kB region containing the rat P2 promoter directed expression of a luciferase reporter speciWcally in peripheral nerve and the CNS, indicating that this minimal pro-

THE CX32 GENE

FIGURE 24.2 Schematic representation of hCx32. (A) hCx31.3. (B) The amino acids of human genes; amino acid diVerences in their mouse orthologs are indicated in gray. The positions of the transmembrane domains are based on the model of Yeager and Nicholson (1996).

moter contains the cis-acting elements that are required for expression in myelinating glia (Neuhaus et al., 1995). Transcripts initiated at the diVerent promoters have partially divergent 5’ untranslated regions (UTR), while the portion of the 5’ UTR close to the start codon, the coding region itself, and 3’ UTR are always identical. As shown in Figure 24.3, the human GJB1/Cx32 promoter contains a TATAA box, as well as putative binding sites for the transcription factors SOX10 and EGR2 (Bondurand et al., 2001). Myelinating Schwann cells express SOX10 and EGR2; oligodendrocytes

601

602

24. THE CONNEXIN32 AND CONNEXIN29 GENES

FIGURE 24.2 (Continued)

express SOX10 but probably not EGR2 (Kuhlbrodt et al., 1998). The idea that SOX10 and EGR2 directly regulate the expression of the Cx32 gene is supported by the Wndings that dominant SOX10 and EGR2 mutations cause inherited dysmyelinating diseases: EGR2 mutations aVect only PNS myelination (Kleopa and Scherer, 2002; Lupski and Garcia, 2001; Wrabetz et al., 2001), whereas SOX10 mutations aVect both CNS and PNS myelination (Inoue et al., 1999, 2003; Kuhlbrodt et al., 1998; Pingault et al., 2000; Stolt et al., 2002). Musso et al. (2001) found that a mutant EGR2 showed reduced aYnity for the most 3’ EGR2 binding site shown in Figure 24.3. The phenotype of the patient who had this mutation was more severe than is typical for CMTX, so that reduced expression of Cx32 alone does not account for the entire clinical picture. EGR2 appears to regulate the expression of several myelin-related genes, including GJB1/Cx32 (Nagarajan et al., 2001).

THE CX32 GENE

FIGURE 24.3 The structure of the mouse and human Cx32 genes. (A) The mouse gene has three alternative promoters, P1-P3, which give rise to three transcripts that diVer only in part of their 5’ untranslated region (UTR); these are shown schematically. Note that all three transcripts have an identical open reading frame (ORF). (B) The human gene has two alternative promoters that correspond to P1 and P2 in mice; whether the P3 promoter exists in humans is unknown. The nucleotide sequence of the P2 promoter, exon 1B, and the most 5’ aspect of exon 2. The locations of potential SOX10 and EGR2 binding sites, a TATAA box (asterisk), a promoter mutation (at 528), the start site of transcription (arrow), exon 1A (capitalized letters), a mutation in the 5’ UTR (at 458), a 355 bp intron, and exon 2 (capitalized letters). Bases are named according to their position from the ATG initiation codon.

Some CMTX kindreds do not have mutations in the GJB1/Cx32 open reading frame (ORF) (BergoVen et al., 1993; Nelis et al., 1999). In these families, mutations might aVect the promoter, splice sites, or the untranslated portions of the mRNA. Ionasescu et al. (1996) reported noncoding region mutations in two CMTX kindreds that lacked mutations in the ORF. One mutation (-528T>G) was just proximal to the start site of transcription and alters a putative SOX10 binding site (Bondurand et al., 2001), and in transient cotransfection assays, this mutation decreases the expression of the Cx32 gene (Bondurand et al., 2001). The other mutation (-458C>T) was in the 5’ UTR and appears to abolish an

603

604

24. THE CONNEXIN32 AND CONNEXIN29 GENES

FIGURE 24.4 Expression of Cx29 and Cx32 mRNA in the CNS and PNS. (A) Northern blot analysis of developing mouse cerebellum. The post-natal ages (in days) are indicated for each sample. Note that the levels of Cx29 and Cx32 mRNA, as well as that of proteolipid protein (PLP), increase in parallel. (B) Northern blot analysis of Cx29 and Cx32 in developing rat sciatic nerve. The age of the rats (in post-natal days; P) is indicated. Note that Cx29 mRNA is expressed prior to that of Cx32, and its expression falls to relatively lower levels in adult nerves; the expression of P0 mRNA is shown for comparison. (C) Northern blot analysis of Cx29 and Cx32 in injured adult rat sciatic nerve. RNA was isolated from distal nerve-stumps at the indicated times (in days; d) post-transection or post-crush. Transections were done in such a way to prevent axonal regeneration; axons regenerate and are remyelinated after nerve-crush. The distal nerve-stumps of crush-injured nerves were divided into proximal (D1) and distal (D2) segments. Note that Cx29 and Cx32 are expressed in parallel in crush-injured nerves; the expression of P0 mRNA is shown for comparison. Adapted from Altevogt et al. (2002), with permission of the Society for Neuroscience.

au2

internal ribosome entry site (IRES) that is essential for the translation of Cx32 mRNA (Hudder and Werner, 2000). The analysis of another 5’ UTR mutation indicates that the mutant mRNA is unstable, as Cx32 mRNA from sural nerve biopsies of two heterozygous carriers of this mutation revealed only wild-type Cx32 mRNA (Flagiello et al., 1998). A third putative promoter mutation ( 713G>A; (Wang et al., 2000) appears to be a polymorphism (Bergmann et al., 2001), and it does not decrease expression in transient co-transfection assays (Bondurand et al., 2001). Although Cx32 is most abundant in liver, it is also expressed by many tissues, including kidney, intestine, lung, spleen, stomach, pancreas, uterus, testes, brain (by oligodendrocytes and perhaps some kinds of neurons), and myelinating Schwann cells (Chandross et al., 1996; Ressot and Bruzzone, 2000; Scherer et al., 1995; So¨hl et al., 1996). In the PNS and CNS, the expression of Cx32 mRNA and protein parallels that of other myelin-related genes (Fig. 24.4). Furthermore, in the PNS, the expression of Cx32 requires the integrity of axon-Schwann cell interactions (Fig. 24.4C). Axotomy results in a rapid inhibition of Cx32 expression distal to the lesion (Scherer et al., 1995; So¨hl et al., 1996)—the pattern observed for many other myelin-related genes (Scherer and Salzer, 2001). This

THE CX29 GENE

FIGURE 24.5 The structure of the mCx29 and hCx31.3 genes. (A) Based on the sequences of ESTs, the mCx29 has two exons that are separated by a 4.8 kb intron. The Wrst exon contains only 5’ UTR. (B) The hCx31.3 gene has two exons that are separated by 5.2 kb. It remains to be determined whether the hCx31.3 gene has a 5’ exon corresponding to the one found in the mouse.

decreased expression of Cx32 likely reXects decreased transcription from the P2 promoter, as the levels of P2-initiated transcripts fall after axotomy (So¨hl et al., 1996). As axons regenerate and are remyelinated, the Cx32 expression increases (Scherer et al., 1995). In the CNS, Cx32 mRNA levels are profoundly reduced in jimpy mice and myelin-deWcient rats, both of which have Plp mutations resulting in oligodendrocyte cell death and profound dysmyelination (Scherer et al., 1995).

THE CX29 GENE A novel murine connexin, mCx29 (AF503616), was identiWed by using degenerate primers to amplify mouse CNS cDNA (Altevogt et al., 2000). Independently, So¨hl et al. (2001a) retrieved the same sequence from the HUSAR/EMBL/Heidelberg database. Several mCx29 ESTs (BB625925, BB644041, BB646257, BE945479) contain 5’ UTR sequences that are located 4.8 kB upstream of the ORF, indicating that the mCx29 gene contains a 5’ exon encoding only 5’ UTR (Fig. 24.5). As noted by Willecke et al. (2002), the splice acceptor site and start codon for mCx29 are separated by only 4 bp. Human genomic sequences corresponding to mCx29 were identiWed by TblastN searches of the Genbank HTGS database (Altevogt et al., 2002). Two entries (AC004977 and AC011904) predicted identical connexins with a molecular mass of 30.3 kDa, while the connexin-related sequence in the third entry (AC004522) exhibited multiple stops and was likely a nonfunctional gene. The putative 30.3 kDa connexin was also described by So¨hl et al. (2001a), who designated it hCx30.2 (the designation of 30.2 was selected since a human Cx30.3 gene had already been described). However, EST database searches revealed a human EST (BI860607) where the ultimate C-terminal diverged from the genomic ORF,

605

606

24. THE CONNEXIN32 AND CONNEXIN29 GENES

predicting a 31.3 kDa protein. Altevogt et al. (2002) conWrmed the EST by RT-PCR of human brain cDNA and in addition found no evidence of transcripts corresponding to the shorter form (hCx30.2). The transcribed C-terminal coding sequence was found 5.2 kb downstream of the original ORF in human genomic DNA. Together, these data indicate that the human, but not mouse, gene contains an intron interrupting the coding region. Thus, the human connexin has been redesignated hCx31.3 (AF503615). Alignment of mCx29 and hCx31.3 ORFs revealed 61% identity of the amino acid sequence (Altevogt et al., 2002). As indicated by CLUSTAL analysis (Fig. 24.1), mCx29 was more closely related to hCx31.3 than to any other murine connexin. This high level of relatedness suggests that mCx29 and hCx31.3 are orthologs. The CLUSTAL algorithm also demonstrates that mCx29 and hCx31.3 were among the most divergent members of the connexin family. Although computer-aided analyses of connexin sequences suggest the grouping of connexins into subclasses, consistent structural rules by which to deWne these subclasses have not emerged (Willecke et al., 2002). Because the signiWcance of these subclasses is unclear, mCx29 and hCx31.3 were not assigned to a subclass, previously deWned or new. Unlike Cx32, the expression of mCx29 appears to be restricted to myelinating glia (Altevogt et al., 2002; Li et al., 2002; So¨hl et al., 2001a). In the CNS, oligodendrocytes appear to be the sole source of Cx29 mRNA by in situ hybridization, and myelin-deWcient rats have dramatically reduced levels of Cx29 mRNA (Altevogt et al., 2002). In the developing CNS, Cx29 mRNA is expressed in concert with those of Cx32 and other myelin-related genes (Fig. 24.4). Cx29 is localized in the myelin sheaths of small myelinated axons in the CNS, particularly in the juxtaparanodal region (Altevogt et al., 2002). Cx32, in contrast, is mainly localized to the somata and proximal processed of oligodendrocytes (Altevogt et al., 2002; Li et al., 1997; Scherer et al., 1995). Moreover, double-labeling for Cx29 and Cx32 demonstrates that they are expressed by a distinct subsets of oligodendrocytes (Altevogt et al., 2002). These data provide further evidence of molecular heterogeneity of oligodendrocytes and indicate that this may be related to the caliber of axons that they myelinate, thereby extending the anatomical descriptions of del Rio Hortega (1928) to the molecular level. In the PNS, Cx29 mRNA and protein are expressed by myelinating Schwann cells. Cx29 mRNA and protein are detected prior to those of Cx32, and the level of Cx29 mRNA in adult nerve is relatively low compared to that of Cx32 (Fig. 24.4B). Similarly, the level of Cx29 mRNA increases as Schwann cells remyelinate regenerating axons (Fig. 24.4C), indicating that Cx29 mRNA expression is regulated by axon-Schwann cell interactions (Altevogt et al., 2002). Cx29 is localized to the paranodes, incisures, inner mesaxon, and especially the juxtaparanodal region of myelinating Schwann cells (Altevogt et al., 2002).

SUMMARY

au3

In spite of their diVerent embryological origins, myelinating Schwann cells and oligodendrocytes in rodents express the same two connexins, Cx29 and Cx32. Both are expressed in parallel with other myelin genes in the CNS and PNS, but in the PNS, the expression of Cx29 appears to begin early and decrease more substantially than that of Cx32. The gene structure and ORF of Cx32 are highly conserved in mammals, including the alterative promoter that is used by myelinating glial cells. The misregulation of Cx32 expression may contribute to the pathogenesis of the dysmyelination that is seen in patients with certain SOX10 and EGR2 mutations. The structure of the mCx29/hCx31.3 genes are surprising divergent—as the hCx31.3 gene has an additional exon that is not found in the mouse. The mCx29/hCx31.3 promoters have not yet been elucidated.

Acknowledgments Our work is supported by the National Institutes of Health.

SUMMARY

References Altevogt B. M., Kleopa K. A., Postma F. R., Scherer S. S., and Paul D. L. (2002). Cx29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J Neurosci 22, 6458–6470. Altevogt B. M., Paul D. L., and Goodenough D. A. (2000). Cloning and characterization of a novel central nervous system speciWc connexin, mouse connexin29. Mol Cell Biol S11, 330a. Bergmann C., Schro¨der J. M., Rudnik-Schneborn S., Zerres K., and Senderek J. (2001). A point mutation in the human connexin32 promoter P2 does not correlate with X-linked dominant Charcot-Marie-Tooth neuropathy in Germany. Mol Brain Res 88, 183–185. BergoVen J., Scherer S. S., Wang S., Oronzi-Scott M., Bone L., Paul D. L., Chen K., Lensch M. W., Chance P., and Fischbeck K. (1993). Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262, 2039–2042. Bondurand N., Girard M., Pingault V., Lemort N., Dubourg O., and Goossens M. (2001). Human connexin 32, a gap junction protein altered in the X-linked form of Charcot-Marie-Tooth disease, is directly regulated by the transcription factor SOX10. Hum Mol Genet 10, 2783–2795. Bruzzone R., White T. W., and Paul D. L. (1996). Connections with connexins: The molecular basis of direct intercellular signaling. Eur J Biochem 238, 1–27. Chandross K. J., Kessler J. A., Cohen R. I., Simburger E., Spray D. C., Bieri P., and Dermietzel R. (1996). Altered connexin expression after peripheral nerve injury. Mol Cell Neurosci 7, 501–518. Del Rio-Hortega P. (1928). Tercera aportacion al conocimiento morgologico e interpretcion functioal de la oldigodendoroglia. Mem Real Soc Exsp d Hist Nat 14, 5–122. Duga S., Asselta, R., DelGiacco L., Malcovati M., Ronchi S., Tenchini M. L., and Simonic T. (1999). A new exon in the 5 ‘ untranslated region of the connexin32 gene. Eur J Biochem 259, 188–196. Flagiello L., Cirigliano V., Strazzullo M., Cappa V., Ciccodicola A., D’Esposito M., Torrente I., Werner R., De Iorio G., Rinaldi M., Dallapiccola B., Forabosco A., Venruto V., and D’Urso M. (1998). Mutation in the nerve-speciWc 5’ non-coding region of Cx32 gene and absence of speciWc mRNA in a CMTX1 Italian family. Hum Mutat 12, 361–363. Hennemann H., Kozjek G., Dahl E., Nicholson B., and Willecke K. (1992). Molecular cloning of mouse connexins26 and -32: Similar genomic organization but distinct promoter sequence of two gap junction genes. Eur J Cell Biol 158:81–89. Hudder A, and Werner R (2000). Analysis of a CMTX mutation reveals an essential IRES element in the connexin-32 gene. J Biol Chem 275, 34586–34591. Inoue K., Shilo K., Boerkoel C. F., Crowe C., Sawady J., Lupski J. R., and Agamanolis D. P. (2003). Congenital hypomyelinating neuropathy, central dysmyelination, and Waardenburg-Hirschsprung disease: Phenotypes linked by SOX10 mutation. Ann Neurol 52, 836–842. Inoue K., Tanabe Y., and Lupski J. R. (1999). Myelin deWciencies in both the central and the peripheral nervous systems associated with a SOX10 mutation. Ann Neurol 46, 313–318. Ionasescu V. V., Searby C., Ionasescu R., Neuhaus I. M., and Werner R. (1996). Mutations of noncoding region of the connexin32 gene in X-linked dominant Charcot-Marie-Tooth neuropathy. Neurology 47, 541–544. KirchhoV S., Nelles E., HagendorV A., Kruger O., Traub O., and Willecke K. (1998). Reduced cardiac conduction velocity and predisposition to arrhythmias in connexin40-deWcient mice. Curr Biol 8, 299–302. Kleopa K. A., and Scherer S. S. (2002). Inherited Neuropathies. Neurol Clin N Am 20, 679–709. Kuhlbrodt K., Herbarth B., Sock E., Hermans-Borgmeyer I., and Wegner M. (1998). Sox10, a novel transcriptional modulator in glial cells. J Neurosci 18, 237–250. Kumar N. M., and Gilula N. B. (1986). Cloning and characterization of human and rat liver cDNAs coding for a gap junction protein. J Cell Biol 103, 767–776. Li X., Lynn B. D., Olson C., Meier C., Davidson K. G. V., Yasumura T., Rash J. E., and Nagy J. L. (2002). Connexin29 expression, immunocytochemistry and freeze-fracture replica immunogold labelling (FRIL) in sciatic nerve. Eur J Neurosci 16, 795–806. Li J., Hertzberg E. L., and Nagy J. I. (1997). Connexin32 in oligodendrocytes and association with myelinated Wbers in mouse and rat brain. J Comp Neurol 379, 571–591. Lupski J. R., and Garcia C. A. (2001). Charcot-Marie-Tooth peripheral neuropathies and related disorders. In ‘‘The Metabolic & Molecular Basis of Inherited Disease’’ (C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, eds.), pp. 5759–5788. McGraw-Hill, New York. Miller R., Dahl G., and Werner R. (1988). Structure of a gap junction gene: Rat connexin-32. Biosci Reports 8, 455–464. Musso M., Balestra P., Bellone E., Cassandrini D., Di Maria E., Doria L. L., Grandis M., Mancardi G., Schenone A., Levi G., Ajmar F., and Mandar P. (2001). The D355V mutation decreases EGR2 binding to an element within the Cx32 promoter. Neurobiol Dis 8, 700–706. Nagarajan R., Svaren J., Le N., Araki T., Watson M., and Milbrandt J. (2001). EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30, 355–368. Nelis E., Haites N., and Van Broeckhoven C. (1999). Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum Mutat 13, 11–28. Neuhaus I. M., Bone L., Wang S., Ionasescu V., and Werner R. (1996). The human connexin32 gene is transcribed from two tissue-speciWc promoters. Biosci Reports 16, 239–248. Neuhaus I. M., Dahl G., and Werner R. (1995). Use of alternative promoters for tissue-speciWc expression of the gene coding for connexin32. Gene 158, 257–262.

607

608

24. THE CONNEXIN32 AND CONNEXIN29 GENES

Paul D. L. (1986). Molecular cloning of cDNA for rat liver gap junction protein. J Cell Biol 103, 123–134. Pingault V., Guiochon-Mantel A., Bondurand N., Faure C., Lacroix C., Lyonnet S., Goosens M., and Landrieu P. (2000). Peripheral neuropathy with hypomyelination, chronic intestinal pseudo-obstruction and deafness: A developmental neural crest syndrome related to a SOX10 mutation. Ann Neurol 48, 671–676. Ressot C., and Bruzzone R. (2000). Connexin channels in Schwann cells and the development of the X-linked form of Charcot-Marie-Tooth disease. Brain Res Rev 32, 192–202. Scherer S. S., Descheˆnes S. M., Xu Y.-T., Grinspan J. B., Fischbeck K. H., and Paul D. L. (1995). Connexin32 is a myelin-related protein in the PNS and CNS. J Neurosci 15, 8281–8294. Scherer S. S., and Salzer J. (2001). Axon-Schwann cell interactions in peripheral nerve degeneration and regeneration. In ‘‘Glial Cell Development’’ (K. R. Jessen, and W. D. Richardson WD, eds.), pp. 299–330. Oxford University Press, Oxford. Simon A. M., Goodenough D. A., Li E., and Paul D. L. (1997). Female infertility in mice lacking connexin 37. Nature 385, 525–529. Simon A. M., Goodenough D. A., and Paul D. L. (1998). Mice lacking connexin40 have cardiac conduction abnormalities characteristic of atrioventricular block and bundle branch block. Curr Biol 8, 295–298. So¨hl G., Eiberger J., Jung Y. T., Kozak C. A., and Willecke K. (2001a). The mouse gap junction gene connexin29 is highly expressed in sciatic nerve and regulated during brain development. Biol Chem 382, 973–978. So¨hl G., Gillen C., Bosse F., Gleichmann M., Muller H. W., and Willecke K. (1996). A second alternative transcript of the gap junction gene connexin32 is expressed in murine Schwann cells and modulated in injured sciatic nerve. Eur J Cell Biol 69, 267–275. So¨hl G., Theis M., Hallas G., Brambach S., Dahl E., Kidder G., and Willecke K. (2001b). A new alternatively spliced transcript of the mouse connexin32 gene is expressed in embryonic stem cells, oocytes, and liver. Exp Cell Res 266, 177–186. Stolt C. C., Rehberg S., Ader M., Lommes P., Riethmacher D., Schachner M., Bartsch U., and Wegner M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev 16, 165–170. Unger V. M., Kumar N. M., Gilula N. B., and Yeager M. (1999). Three-dimensional structure of a recombinant gap junction membrane channel. Science 283, 1176–1180. Wang H. L., Wu T., Chang W. T,. Li A. H., Chen M. S., Wu C. Y., and Fang W. (2000). Point mutation associated with X-linked dominant Charcot-Marie-Tooth disease impairs the P2 promoter activity of human connexin-32 gene. Mol Brain Res 78, 146–153. White T. W. (2002). Unique and redundant connexin contributions to lens development. Science 295, 319–320. White T. W., and Paul D. L. (1999). Genetic diseases and gene knockouts reveal diverse connexin functions. Annu Rev Physiol 61, 283–310. Willecke K., Eiberger J., Degen J., Eckardt D., Romualdi A., Guldenagel M., Deutsch U., and So¨hl G. (2002). Structural and functional diversity of connexin genes in the mouse and human genome. Biol Chem 383, 725–737. Wrabetz L., Feltri M. L., Hanemann C. O., and Muller H. W. (2001). The molecular genetics of hereditary demyelinating neuropathies. In ‘‘Glial Cell Development’’ (K. R. Jessen, and W. D. Richardson WD, eds.), pp. 331–354. Oxford University Press, Oxford. Yeager M., and Nicholson B. J. (1996). Structure of gap junction intercellular channels. Curr Opin Struct Biol 6, 183–192.

C H A P T E R

25 Integrins Charles ffrench-Constant

INTRODUCTION The correct formation of a series of myelin sheaths along the length of every axon requires a number of distinctive developmental processes. These include (1) the production of oligodendrocyte or Schwann cell precursor cells (in the PNS, these are generated from neural crest cells that originate within the dorsal neural tube; in the CNS, oligodendrocyte precursor cells arise within the ventral neural tube in response to signaling by Shh and are speciWed, at least in part by expression of the olig transcription factors); (2) migration of the precursors from their origins within the neural tube into the PNS (for neural crest cells) or throughout most regions of the central nervous system (for oligodendrocyte precursor cells); (3) proliferation within the region of the axons to be myelinated; (4) diVerentiation with establishment of axonal contact, expression of myelin genes, and formation and wrapping of the multilamellar sheath; and (5) apoptosis to remove excess cells and ensure that correct Wnal number of myelin forming glia for the available axons (Barres and RaV, 1999; Bronner-Fraser, 1993; Butt and Berry, 2000; Compston et al., 1997; Le Douarin et al., 1991; Miller, 1996; Mirsky and Jessen, 1996; Richardson et al., 1997; Rowitch et al., 2002; Tsai and Miller, 2002). The basic processes of migration, proliferation, and survival are shared with many other cell types during development. However, the precision required of myelinating glia to ensure appropriate axo-glial interactions and the uniquely spectacular structure of the sheath itself makes myelination of axons a particularly fascinating problem for cell biologists. It is also one of considerable clinical importance in light of the relevance to studies on multiple sclerosis (MS) and other demyelinating diseases. It is already clear that multiple signaling molecules and their glial cell surface receptors coordinate to provide the short- and long-range cues required for myelination. This chapter focuses on one such family of receptors, the integrins. The major goals are to review the expression and function of integrins in Schwann cells and oligodendrocytes, the myelin-forming cells of the peripheral and central nervous system (PNS and CNS), respectively, and to describe how integrin signaling might interact with other signaling pathways such as growth factors. In addition, this chapter examines the contributions that studies on integrins have made to our understanding of the mechanisms that might underlie the regulation of axon-glial interaction and the failure of repair following demyelination in MS. This chapter compliments a recent, excellent review on integrins in the PNS (Previtali et al., 2001), while the role of integrins in CNS myelination has not previously been systematically reviewed.

Myelin Biology and Disorders, Volume 1

609

Copyright 2004, Elsevier (USA). All rights reserved.

610

25. INTEGRINS

INTEGRIN STRUCTURE AND SIGNALING This is a large and rapidly expanding area of research, and a full discussion is beyond the scope of a review focused on myelination. This discussion summarizes the essential details and highlights reviews that provide a more in-depth analysis of some areas. An individual integrin receptor comprises two distinct transmembrane molecules, termed a and b subunits, each encoded by separate genes (Fig. 25.1). Each subunit has a single transmembrane domain and a short cytoplasmic domain. The only exception to this is the b4 subunit, which has a very large cytoplasmic domain of greater than 1000 amino acids. All integrin cytoplasmic domains lack intrinsic kinase or other enzymatic properties. A ligand binding site is created by the combination of the two large extracellular chains (Hynes, 1992). Currently, 18 a and 8 b subunits have been characterized, with 24 heterodimeric combinations detected (Hynes and Zhao, 2000). Each heterodimer has a distinct speciWcity for a single ligand or range of ligands, although two broad groups comprising those binding molecules such as Wbronectin or vitronectin containing RGD or RGD-like amino acid sequences or those binding laminins can be distinguished. These ligands for integrins are usually extracellular matrix molecules (Ruoslahti, 1996), but a number of cell surface ligands including immunoglobulin superfamily molecules such as Thy-1, V-CAM, and L1 as well as molecules containing disintegrin-like domains of the ADAM family have also been described (Evans, 2001; Humphries et al., 1995; Leyton et al., 2001; Montgomery et al., 1996; Wolfsberg et al., 1995). Integrins transmit signals following ligand binding by the assembly of a signaling complex in association with the cytoplasmic domains (Clark and Brugge, 1995; Dedhar, 2000; Giancotti and Ruoslahti, 1999). The components of the complex include kinases, such as FAK, Src family kinases, and ILK as well as adaptor proteins such as paxillin. In addition, integrins are linked directly to the cytoskeleton with molecules such as talin and vinculin also present in the complex. While receptor occupancy by monovalent ligands such as peptides can induce the association of some of these molecules with the cytoplasmic domain of single integrins, binding to multivalent extracellular ligands such as those present in the extracellular matrix is required for clustering of integrins within the membrane. This clustering initiates full signaling complex assembly and cytoskeletal association (Cukierman et al., 2001; Miyamoto et al., 1995a, 1995b) in a structure often referred to in cultured cells as a focal adhesion. The molecules of the complex then interact directly, or via the recruitment of further molecules such as the adaptor protein Crk, with components of the MAP kinase (MAPK), PI3 kinase (PI3K), and Jun kinase (JNK) signaling pathways. These pathways are well recognized regulators of cell behavior, linking integrins with deWned eVector molecules. These include cyclin D, regulating cell cycle progression by activating cyclin dependent kinases, the pro-apoptotic molecule BAD, and the Rho-family GTPases that regulate cell shape (Assoian and Schwartz, 2001; Datta et al., 1999; Etienne-Manneville and Hall, 2002). In addition to molecules linked to the cytoplasmic domains, integrins can also signal via an association with other membrane-associated molecules. The receptors for the growth factors PDGF, EGF, and VEGF have all been shown to interact directly with integrins (Borges et al., 2000; Miyamoto et al., 1996; Moro et al., 1998), providing a molecular basis for integrin-growth factor interactions that will be discussed in more detail in a later section of this chapter. The integral membrane protein caveolin-1 has been shown to recruit the adapter protein Shc to the signaling complex, which in turn can bind the Grb2 and so activate the MAPK pathway (Wary et al., 1996). Another potential pathway is provided by the tetraspanin family of proteins (so called because of their four membrane-spanning domain structure), which include members of the PLP and PMP 22 gene families as well as CD9, a marker of mature myelin. These can interact with the extracellular domain of the integrin a subunit (Gudz et al., 2002) and have been directly implicated in the control of cell motility (Berditchevski, 2001). Together these signaling mechanisms contribute to ‘‘outside-in’’ signaling, by which extracellular cues modulate intracellular signaling pathways. A second direction of integrin signaling, ‘‘inside-out’’ signaling, plays an important role in the regulation of integrin

INTEGRIN FUNCTIONS IN NEURAL DEVELOPMENT AND REPAIR

A matrix

b a

b a ba b a

b a integrin

Signalling complex

cytoskeletal assembly

migration survival

proliferation

B laminin fibronectin vitronectin

α9 α8

α10

β3 β5

α7 β4

α6

αv

β1

β6

α1

α5 α4

α3

α2

β8

β7

FIGURE 25.1 (A) A schematic diagram showing integrin structure. Note the presence of two chains (3a and b), which together form the ligand-binding site. Ligand binding is associated with clustering, which initiates the formation of a signaling complex as discussed in the text. (B) Heterodimerization patterns of integrin subunits—note that many diVerent a subunits associate with b1, while a number of the other diVerent b subunits will associate with aV. The ligand speciWcity of those expressed on myelinating glia and their precursors is discussed in the text. Note that b2 integrins, present on many cells involved in immune function, are not shown here.

ligand binding (Hynes, 1992). In normal circumstances, only a small fraction of cell-surface integrins bind their extracellular ligands. Ligand binding is enhanced by a process termed activation, which can result either from a change in shape of the integrin extracellular domains increasing ligand aYnity or by clustering of integrins increasing ligand avidity (Bazzoni and Hemler, 1998). The conformational changes in the integrin extracellular domains associated with increased ligand aYnity have recently been clariWed by structural studies (Takagi et al., 2002; Vinogradova et al., 2002). These reveal that inactive integrins may have a bent conformation with the ligand binding site next to the membrane, while activated integrins straighten and so increase exposure of the site to extracellular and other molecules (Hynes, 2002; Liddington and Ginsberg, 2002). The mechanisms by which this change can be regulated by intracellular signaling pathways remain poorly understood. However, it is clear that this form of signaling provides an important physiological mechanism for the control of integrin function in many cell types.

INTEGRIN FUNCTIONS IN NEURAL DEVELOPMENT AND REPAIR Given the diversity of signaling pathways activated by integrins, it is not surprising that these receptors have been shown to regulate cell behavior in a wide variety of systems. The

611

612

25. INTEGRINS

nervous system is no exception, with some important experiments highlighting integrin function in neural development, maintenance, and repair. Mice in which all b1 integrins have been ablated from neural precursor cells by the use of cre-lox technology show abnormalities of radial glial endfeet within the cortical marginal zone and alterations in cortical lamina formation as neuronal precursors migrate too far (Graus-Porta et al., 2001). Migration abnormalities are also seen in mice lacking either the a3 or a6 subunits, both of which heterodimerize with b1 to generate laminin receptors. The patterns, however, are diVerent. In a6 deWcient mice, the neurones migrate too far, producing nodules on the surface of the cortex (Georges-Labouesse et al., 1998), while in a3-deWcient mice migrating neurones appeared to leave the radial glial Wbres along which they normally migrate prematurely resulting in a partial inversion of the normal patterning of the cortex (Anton et al., 1999). aV integrins are expressed in radial glial cells, and antibody blocking studies suggest a role for these integrins in the interaction with migrating neurones (Anton et al., 1999). The abnormalities of cerebral vasculature development and consequent haemorrhage seen in aV knockout mice (McCarty et al., 2002) have so far prevented a genetic analysis to conWrm this role of aV in neuronal migration. Cell culture experiments have shown that integrins can promote neurite outgrowth on a number of diVerent extracellular matrix substrates including Wbronectin and laminin (Reichardt and Tomaselli, 1991). These studies suggest that integrins will contribute to axon outgrowth in vivo during development and repair. In keeping with this result, mice lacking the a7 subunit (part of the a7b1 laminin receptor) have reduced motor neurone regeneration following a facial nerve injury (Werner et al., 2000). Regenerating PNS sensory axons express a4 integrin, which binds to a region of Wbronectin whose expression is increased following nerve injury as a consequence of altered alternative splicing (Vogelezang et al., 2001). Moreover, it is possible to enhance neurite outgrowth from adult DRG neurones (which normally lose their ability to grow neurites well in culture) on diVerent extracellular matrix substrates by increasing the expression levels of appropriate integrins in these cells (Condic, 2001). While the extent to which neurite outgrowth in vitro mirrors axon regeneration is unclear, these results do point to a role for changes in integrin expression in deWning the ability of the neurones to regrow after injury. Integrins may also be essential for the higher order functions of the CNS, as other work implicates integrins in the formation of memory. RGD peptides, which can block the interaction of many integrins with many nonlaminin extracellular matrix ligands such as Wbronectin and vitronectin, reduce the stabilization of long-term potentiation in hippocampal slice cultures (Bahr et al., 1997). Long-term potentiation represents a well-studied model of learning and memory, and more recent work has shown that maturation of hippocampal synapses is associated with changes in glutamate release and NMDA receptor composition. These changes are blocked by antibodies against aVb3, an RGD-binding integrin expressed at the synapses (Chavis and Westbrook, 2001). Genetic studies have also shown that integrins can contribute to memory function in drosophila. Mutations in an a subunit gene velado cause loss of short-term memory (Grotewiel et al., 1998) associated with abnormalities in synaptic structure and conduction (Rohrbough et al., 2000). Given the role of synapse morphology in memory formation, and the ability of integrins to interact with signaling pathways regulating cell shape, these results implicating integrins in memory may reXect an essential role in regulating the shape changes of the synapse that underlie some aspects of learning and memory. Integrins are also expressed on astrocytes in the CNS (Tawil et al., 1994). Cell culture studies using cells from b5-deWcient transgenic mice that will lack the Wbronectin/vitronectin binding integrin aVb5 (Milner et al., 1999), or cells in which aV integrin expression has been manipulated by overexpression of diVerent b subunits that heterodimerise with aV (Milner et al., 2001), reveal roles for aV integrins in cell migration. However, there are as yet no studies examining integrin function in astrocytes in vivo and the relevance of these studies to the intact CNS will also be unclear until well-deWned extracellular matrix ligands for av integrins are characterized in CNS tissue. Equally, the likely presence of many diVerent classes of astrocytes, which cannot be readily distinguished by currently available markers, will make an analysis of astrocyte integrin expression and function a challenging task.

INTEGRIN EXPRESSION IN MYELIN FORMING GLIAL CELLS

INTEGRIN EXPRESSION IN MYELIN FORMING GLIAL CELLS In contrast to astrocytes, a number of diVerent stages in the development of myelinforming cells can be identiWed in both the Schwann cell and oligodendrocyte lineage. The ability to grow many of these diVerent stages in cell culture in suYcient quantities to allow biochemical analysis, combined with the use of markers to identify cells in vivo, has allowed the characterization of integrin expression patterns during myelination.

Schwann Cells

au1

Schwann cells arise from a population of migrating cells, neural crest cells, that originate from the dorsal neural tube. In vivo studies on these cells have shown expression in avian embryos of the laminin receptors a1b1 and (for a subpopulation of neural crest cells) a6b1 and a7b1 and of Wbronectin receptors containing the a4 and aV subunits (Bronner-Fraser et al., 1992; Duband et al., 1992; Kil and Bronner-Fraser, 1996; Kil et al., 1998). A more complete repertoire of the expression of integrins has been determined on neural crest cells grown in culture, and four laminin receptors (a1b1, a3b1, a6b1and a7b1), seven Wbronectin receptors (a3b1, which also binds laminins, as well as a4b1, a5b1, a8b1, aVb1, aVb3, and a b8 integrin), and three vitronectin receptors (avb1,avb3 and avb5) have been detected (Testaz et al., 1999). Neural crest cells in turn give rise to committed Schwann cell precursors within early embryonic nerves. Integrin expression on these precursor cells has not been systematically addressed, although in the rat they do not express the a1b1 collagen/laminin binding integrin present on their parent neural crest cells and on nonmyelinating diVerentiated Schwann cells (Stewart et al., 1997). Integrin expression on diVerentiated Schwann cells has been examined both in cell culture and in vivo. Many integrins have been described, with Schwann cells shown in diVerent studies to express a1b1, a2b1, a6b1, a6b4, aVb3, and aVb8, with low levels of a4b1 and a5b1 (Einheber et al., 1993; Feltri et al., 1994; Hsiao et al., 1991; Milner et al., 1997c; Niessen et al., 1994). A Schwann cell line, MSC80, also expresses a1b1, a5b1, and a6b1 as well as aV integrins (Detrait et al., 1999). One study has reported a3b1 on human Schwann cells grown in cell culture (Hsiao et al., 1991). Three of these integrins, a1b1, a6b1, and a6b4, show interesting patterns of developmental regulation. Once axonal contact is established during normal development, Schwann cells diVerentiate into myelinating or nonmyelinating cells. a1b1 is normally up-regulated only in the nonmyelinating population, but increases also in the myelin-forming population following loss of axons associated with peripheral nerve injury. a1b1 does not, however, increase in Schwann cell precursors removed from axonal contact by being grown in cell culture (Stewart et al., 1997). a6 switches b partners from b1 to b4 at the onset of myelination, even though b1 integrin subunit expression remains unchanged in the cell. a6b4 is then expressed on the outside (abaxonal surface) of the Schwann cell, adjacent to the overlying basal lamina that surrounds the cell. As with a1b1, axonal contact appears to be important in regulating this switch. Schwann cells up-regulate b4 when co-cultured with dorsal root ganglion (DRG) axons and b4 expression falls after axon transection in these cultures or after peripheral nerve lesions in vivo. Axonal regeneration is associated with a reappearance of b4, although interestingly b4 also returns at lower levels in the presence of permanent nerve transection, indicating that expression is not entirely dependent on axons (Einheber et al., 1993; Feltri et al., 1994; Niessen et al., 1994).

Oligodendrocytes The intrinsic mechanisms within oligodendrocyte precursor cells that results in their diVerentiation in culture with a time scale that mirrors that seen in vivo allows a developmental analysis of integrin expression using immunoprecipitation techniques. This revealed a rather more restricted pattern of integrin expression in that seen in the Schwann

613

614

25. INTEGRINS

au2

au3

cell lineage. a6b1 was expressed throughout oligodendrocyte diVerentiation, while aV integrins showed developmental regulation with sequential expression of aVb1, aVb3, and aVb5. A fourth aV integrin, aVb8, was expressed throughout diVerentiation. a8b1 had also been reported in oligodendrocytes but was not seen in our cell culture conditions (Milner and Vrench-Constant, 1994; Milner et al., 1997b). To address the eVects of axonal contact on oligodendrocyte integrin expression, we used a xeno-culture system in which DRG neurones from rat were co-cultured with mouse oligodendroglial cells and species speciWc anti-integrin antibodies then used to distinguish rat and mouse integrins. These experiments showed that oligodendrocytes, unlike Schwann cells, do not express a6b4 at the time of myelin formation (Shaw et al., 1996). This result was expected, given the lack of b4 expression within white matter tracts in vivo in the CNS (Sonnenberg et al., 1990). We also found only modest changes in the levels of aV integrins in the co-cultures, with the developmental switching of b subunits preserved (Milner et al., 1997b). In situ hybridization studies of white matter tracts have shown diVuse labeling for a6 and aV mRNAs, consistent with our in vitro observations. b3 was not detected in the CNS, an apparent contradiction to our observation that aVb3 is present during oligodendrocyte (PinkstaV et al., 1999) diVerentiation. However, more recent studies in hippocampus have shown aVb3 to be expressed at synapses, as discussed earlier (Chavis and Westbrook, 2001). These observations illustrate an important point to be considered when comparing the in vitro and in vivo studies. Integrin expression levels may be very low in vivo, making them very diYcult to detect by in situ hybridization, but still suYcient to exert an important biological eVect. The evidence that blocking aVb3 modulates hippocampal synaptic transmission emphasizes this point, as does our observation (described later) that a6-deWcient mice show abnormalities in oligodendrocyte survival while previous immunocytochemical studies did not reveal a6 protein in white matter tracts (Sonnenberg et al., 1990). When grown in cell culture, cells may up-regulate these integrin subunits substantially, facilitating detection but creating the possibility that some of the observed expression does not accurately reXect the pattern seen in vivo.

INTEGRIN FUNCTION IN MYELINATING GLIAL CELLS Schwann Cells Migration Neural crest cell migration has been examined extensively in vitro, using explant culture systems in which the cells migrate across extracellular matrix components such as Wbronectin or laminin. A number of diVerent anti-integrin antibodies have been shown to inhibit migration in this system. DiVerent results are seen on diVerent extracellular matrix molecules. Inhibition of a4b1 and avb3 is able to block migration on Wbronectin substrates, on vitronectin avb3 and avb5 appear to play the major role in migration and on laminin substrates a1b1 is responsible for the majority of migration (Delannet et al., 1994; Desban and Duband, 1997; Testaz et al., 1999). These results provide valuable insights into the mechanisms of neural crest cell migration, but are less informative as to which are the critical integrins in vivo, when many diVerent extracellular matrix molecules including Wbronectin and laminins are present together in the migratory pathway (Perris, 1997). To examine integrin function in vivo, three main approaches have been used: injection of blocking antibodies, antisense oligonucleotides, and analysis of transgenic ‘‘knockout’’ animals. Both of the former approaches can perturb neural crest cell migration (BronnerFraser and Lallier, 1988; Kil et al., 1996), but questions about whether observed eVects are due to loss of integrin function in the neural crest cells themselves or their cellular neighbors limit the power of these experiments. Transgenic animals also have this disadvantage (unless a cell-type speciWc ablation is performed using a cre/lox strategy) but, crucially, are informative if neural crest cell migration is unaVected despite the loss of an

INTEGRIN FUNCTION IN MYELINATING GLIAL CELLS

integrin subunit. An example of this is provided by the a4 integrin knockout. Using explant and grafting techniques to allow analysis of neural crest cell migration at later stages than feasible in intact embryos due to the lethality of the knock out, it was found that a4-deWcient neural crest cells migrated normally but showed increased apoptosis (Haack and Hynes, 2001). While the latter eVect was also seen in a later antibody blocking study (Testaz and Duband, 2001), the normal migration apparently contradicts the in vitro work and also emphasizes the diYculty of interpreting experiments where the normal cellular environment is being perturbed. Schwann cells also migrate during development and repair of peripheral nerves. This migration is essential for normal development, allowing interactions between Schwann cells and the growing axons. While the role of integrins in vivo is unknown, cell culture studies have shown that, like neural crest cells, Schwann cells can use multiple extra au4 cellular matrix ligands and integrins for migration (Milner et al., 1997c). Proliferation While much attention has focused on the role of growth factors in Schwann cell proliferation, two experiments point to an important role for integrins. First, antibodies against the tetraspan molecule CD9 promote proliferation (Hadjiargyrou and Patterson, 1995). CD9 is associated with a3b1 and a6b1, implicating integrin signaling in proliferation (Hadjiargyrou et al., 1996). More direct evidence comes from the observation that a5deWcient Schwann cells proliferate more slowly than do wild-type cells (Haack and Hynes, 2001). Differentiation The pioneering work of the Bunge laboratory using DRG neurone-Schwann cell cultures established the need for Schwann cells to assemble a basal lamina prior to the formation of a myelin sheath (Moya et al., 1980). An important observation was that the addition of ascorbate to the culture medium allowed collagen biosynthesis and the assembly of the basal lamina scaVold into which laminins and other extracellular matrix molecules are incorporated (Eldridge et al., 1987). More recent work has shown that the critical component of the basal lamina is laminin, as laminin deposition on the Schwann cell surface is suYcient for myelination in the absence of basal lamina formation (Podratz et al., 2001). These results implicate integrin laminin receptors in Schwann cell myelination and, consistent with this result, antibodies against b1 integrin inhibit myelination in the cocultures (Fernandez-Valle et al., 1994). These experiments cannot, however, distinguish between the eVects of antibody blockade on the neuronal and glial integrins present in the co-cultures. This problem has been overcome in an elegant genetic study using cre/lox technology to ablate b1 speciWcally in Schwann cells (Feltri et al., 2002). This study used mice in which both copies of the b1 gene have been modiWed so as to contain two lox sites on either side of the Wrst coding exon. These were crossed with mice that express cre recombinase driven from the Po promoter and are also heterozygous for a b1 null allele. This cross generates progeny with the genotype b1 Xox/null//Po.cre, which are therefore heterozygous for the b1 null allele and develop normally until Schwann cell myelination commences. Then, activation of the Po promoter in Schwann cells results in recombination and ablation of the remaining b1 allele, but only in the Schwann cells. Consequently, the role of b1 can be assessed in the context of a normal cellular background. These mice show a peripheral neuropathy, and histological examination of the nerve shows abnormal Schwann cell precursors that often fail to establish contact with individual axons. Even if contact is established, the cells appear to fail to maintain contact and thus retract their processes. However, some normal myelination is observed in the mice, although this occurs later than in wild-type animals. These abnormalities of Schwann cell myelination did not reXect changes in Schwann cell numbers due to abnormalities of proliferation or migration of precursor cells, as Schwann cell numbers and proliferation rates are normal in the mutant nerves. The loss of the b1 integrin therefore seems to cause abnormalities in the dramatic changes in cell shape associated with myelination. In contrast, and perhaps surprisingly given the clear switch in a6 partners at the time of myelination, loss of b4

615

616

25. INTEGRINS

integrin does not appear to aVect myelin formation (Frei et al., 1999). It may, however, have an important role in maintenance of the myelin sheath, and the perinatal lethality of the b4-null mouse makes this harder to address in vivo without using the cre/lox strategy described earlier. Survival The conditional knockout study described earlier shows how b1 integrins are not required for the survival for the Schwann cells themselves. However, the timing of gene excision in that study meant that Schwann cell precursors were not examined, as these do not express the cre recombinase. As the survival response to growth factors diVers between Schwann cell precursors and diVerentiating Schwann cells (with the former dependent on axonal signals probably mediated by neuregulin; Dong et al., 1995; Garratt et al., 2000), it is possible that Schwann cell precursors do require integrins for survival. This appears to be the case for neural crest cells, as described earlier, and further studies of the role of integrins in the axon-dependent survival of Schwann cell precursors is required given the evidence to be described later in this chapter that integrins are involved in axon-mediated oligodendrocyte survival.

Oligodendrocytes Migration Oligodendrocyte precursors migrate from distinct sites in the germinal regions of the developing CNS (Miller, 1996; Thomas et al., 2000). This migration appears to be essential for the generation of oligodendrocytes. Dorsal regions of embryonic spinal cord separated from the normal ventral source of oligodendrocyte precursors do not generate oligodendrocytes when cultured as explants under conditions that promote oligodendrocyte precursor diVerentiation (Warf et al., 1991). Although other studies have shown that mitogenic growth factors can induce oligodendrocyte formation from cells within the dorsal cord (Chandran et al., 1998), it is not clear to what extent this represents reprogramming of a population of cells within the dorsal cord that would not normally produce oligodendrocytes. Within the eye, the lack of myelination in the retinal nerve Wber layer results from a barrier to oligodendrocyte precursor migration at the retina-optic nerve junction (the lamina cribrosa) (Vrench-Constant et al., 1988; Perry and Lund, 1990). Antibodies that block b1 integrins will inhibit oligodendrocyte precursor migration in cell culture (Milner et al., 1996; Tiwari-WoodruV et al., 2001), demonstrating a role for these integrins. avb1 has been suggested to be the most signiWcant b1 integrin regulating migration (Milner et al., 1996), but the lack of speciWc antibodies against this heterodimer, and the large number of potential partners for both the av and b1 subunits has made it impossible to conWrm this directly at present. However, a role of avb1 would be interesting, as this integrin is down-regulated as oligodendrocyte precursors diVerentiate, suggesting a model in which switching of the av integrin b partners might regulate, at least in part, the loss of the migratory phenotype (Milner and Vrench-Constant, 1994). It should be noted that a molecule on the migrating precursors themselves cannot act as a guidance cue, only a motor for migration. These cues may be provided by extracellular factors such as netrins, semaphorins, and chemokines, some of which have well-established roles in axonal guidance (Spassky et al., 2002; Sugimoto et al., 2001; Tsai et al., 2002). Proliferation Oligodendrocyte precursors continue to proliferate once migration has ceased, so amplifying the numbers of cells available for diVerentiation into myelin-forming oligodendrocytes. As will be discussed later, most attention has focused on growth factors as regulators of proliferation. Integrins are also involved, as over-expression of avb3 promotes proliferation of the oligodendrocyte precursor cell line CG-4, while blocking antibodies and overexpression of dominant-negative inhibitors of this integrin will inhibit proliferation in cell culture (Baron et al., 2002; Blaschuk et al., 2000).

617

INTEGRIN FUNCTION IN MYELINATING GLIAL CELLS

Differentiation Both av and a6 integrins appear to be involved in diVerentiation. Expression of the myelinspeciWc gene myelin basic protein (MBP) can be blocked by the addition of antibodies against avb5 to oligodendrocytes diVerentiating in cell culture (Blaschuk et al., 2000). Given the sequential expression pattern reported for avb1, avb3, and avb5, and their roles in migration, proliferation, and diVerentiation, respectively in cell culture, we have proposed a model in which the timing of these diVerent stages of oligodendrocyte precursor development is regulated by the switching of av-associated b subunits (Blaschuk et al., 2000). There are, however, a number of important points of caution to be considered. First, there is at present no support for this model from genetic studies in intact animals. The transgenic mice lacking b3 and b5 appear fertile and viable (Hodivala-Dilke et al., 1999; Huang et al., 2000), although myelination has not been examined. However, as shown by the tenascin-CdeWcient mice as discussed later, signiWcant abnormalities of oligodendrocyte development may be present in apparently normal mice. aV knockout mice die of cerebral hemorrhage before myelination can be analyzed (McCarty et al., 2002), and cell-type speciWc knockouts of this integrin have not yet been reported. Second, the model predicts that the diVerent integrins will be expressed sequentially during myelination in vivo, but this has yet to be conWrmed and may be diYcult given the potentially low expression levels as discussed earlier. Third, our original biochemical studies on oligodendrocytes were performed using the detergent Triton X-100 at 48C to extract membrane proteins. Recent work has characterized membrane microdomains enriched in speciWc signaling molecules termed lipid rafts (Simons and Toomre, 2000), which are not soluble at 48C in this detergent but require the higher temperature of 378C. Oligodendrocytes contain more than one type of lipid raft (Simons et al., 2000). Our original studies on oligodendrocyte integrins would not have identiWed raft proteins, and we and others have shown that lipid rafts contain integrins (Baron et al., 2003; Claas et al., 2001; Green et al., 1999; Krauss and Altevogt, 1999; Leitinger and Hogg, 2002; Thorne et al., 2000). It is possible therefore that the pattern of developmental regulation is more complicated than originally reported. Transplantation of precursor cells expressing a dominant-negative b1 integrin chimera into a experimentally created demyelinated lesion shows that this integrin is required for myelination, as these cells will not remyelinate the axons eVectively while control transplants repair eYciently (Relvas et al., 2001) (Fig. 25.2). The dominant-negative integrin was made by fusing the cytoplasmic domain of b1 with the transmembrane and extracellular domain of the interleukin-2 receptor (LaFlamme et al., 1994). This strategy generates single b1 cytoplasmic domains in the membrane, able to associate with normal signaling partners but unable to heterodimerize with an a subunit and so bind ligand to initiate the clustering required for signaling. The only b1 integrin expressed at this stage of oligodendrocyte development is a6b1 (Milner and Vrench-Constant, 1994), and cell culture experiments have shown that a6 integrins appear to contribute to the changes in cell shape associated with diVerentiation. Cells plated on substrates of laminin-2 (also known as merosin), recognized by the a6b1 receptor, elaborate extensive myelin-like membrane sheets. In contrast, cells plated on av ligands such as Wbronectin extend processes just as well as do cells on laminin-2 but form fewer sheets (Buttery and Vrench-Constant, 1999). This appears to be mediated by signaling from the b1 rather than the a6 subunit, as cells engineered to express the a5 subunit, so generating an a5b1 Wbronectin receptor, now show myelin-like membrane formation on Wbronectin substrates (Relvas et al., 2001). This ability to switch or manipulate the substrate speciWcity of oligodendrocyte diVerentiation may have therapeutic signiWcance if, as discussed later, the altered extracellular matrix in the damaged brain is an important contributor to the pathogenesis of diseases such as multiple sclerosis. Survival The strongest evidence for a role of integrins in oligodendrocyte development comes from studies of apoptosis. During normal development, as many as 50% of newly formed oligodendrocytes will die (Barres et al., 1992a). This represents target-dependent cell

au5

618

25. INTEGRINS

FIGURE 25.2 Oligodendrocyte precursor cells expressing a dominant-negative b1 integrin construct do not remyelinate eYciently. Note the lack of thin myelin sheaths (as are normally seen in remyelination) in the lesion injected with cells expressing the dominant-negative construct (panels on right;the upper shows a light micrograph, while the lower is an electron micrograph), while the control cells expressing only the retroviral vector used remyelinate eYciently, as shown in the panels on the left.

death; those oligodendrocytes that fail to establish axonal contact can be seen to undergo apoptosis (Trapp et al., 1997), and genetic manipulation of the number of axons in optic nerve also increases the number of surviving oligodendrocytes (Burne et al., 1996). Cell culture experiments point to a role for a6 integrins. Blocking antibodies inhibit the survival promoting eVect on oligodendrocytes either of neurones (when grown in rat/mouse coculture to allow direct cell-cell contact and speciWc inhibition of oligodendrocyte integrins) or astrocytes (Corley et al., 2001; Frost et al., 1999). More compelling, however, are experiments examining oligodendrocytes survival in a6-deWcient mice (Colognato et al., 2002). These mice show increased levels of oligodendrocyte apoptosis in the developing brain stem, a tract in which myelination appears before birth, so allowing analysis of the phenotype before the perinatal lethality resulting from a blistering skin defect (GeorgesLabouesse et al., 1996) prevents further analysis of the mice.

INTERACTION OF INTEGRINS WITH OTHER SIGNALING MOLECULES IN MYELINATING GLIA A number of cell-surface and extracellular signaling molecules have been implicated in the regulation of oligodendrocyte development and the formation of a myelin sheath including cell-surface adhesion molecules (CAMs) (Ben-Hur et al., 1998; Bhat and Silberberg, 1988; Charles et al., 2000; Decker et al., 2000; Diers-Fenger et al., 2001; Gard et al., 1996; Hughson et al., 1998; Kramer et al., 1999; Martini and Schachner, 1997; Ono et al., 1997; Payne et al., 1996; Pedraza et al., 2001; Schnadelbach et al., 2000, 2001; Tait et al., 2000;

au11

INTERACTION OF INTEGRINS WITH OTHER SIGNALING MOLECULES IN MYELINATING GLIA

au9

Tiwari-WoodruV et al., 2001; Wang et al., 1996), matrix metalloproteases (Oh et al., 1999), extracellular matrix (ECM) molecules (Bartsch et al., 1994; Cardwell and Rome, 1988; Fuss et al., 1993; Garcion et al., 2001; Kiernan et al., 1996; Pesheva et al., 1997; Schirmer et al., 1994), neurotransmitters (Bergles et al., 2000; Ghiani et al., 1999; Ghiani and Gallo, 2001), ion channels (Knutson et al., 1997; Neusch et al., 2001), Notch (Wang et al., 1998), and chemokine receptors (Wu et al., 2000). Two approaches can be taken to establish the mechanisms by which integrins interact with some of these signaling pathways. First is the identiWcation of other molecules that are associated directly with integrins. In oligodendrocytes, these include two tetraspanin molecules, CD9 and Tspan-2, as well as the tetraspanin-like molecule and well-characterized myelin protein PLP. The former have been shown to interact with b1 integrins while PLP interacts with avb5 (Birling et al., 1999; Gudz et al., 2002; Terada et al., 2002). Both observations are extremely interesting, as tetraspanins have been implicated in the regulation of integrin-mediated migration in other cell types (Berditchevski, 2001). They therefore represent an important class of potential regulators of myelination requiring further study. In Schwann cells, b1 is associated with the kinase FAK, the adaptor protein paxillin, and schwannomin, a tumor suppressor molecule encoded by the neuroWbromatosis type 2 gene (discussed in more detail at the end of the chapter) (Chen et al., 2000; Fernandez-Valle et al., 2002). These molecules can potentially interact with a number of diVerent intracellular signaling pathways including those downstream of growth factor receptors. The observation that disruption of the actin cytoskeleton with cytochalasin inhibits myelin gene expression (Fernandez-Valle et al., 1997) suggests that the well-recognized role of integrins in cytoskeletal reorganization may also be important for myelination. A second approach to the question of integrin interactions during myelination is to study the eVect of integrin ligation and inhibition on individual signaling pathways. Of these, most attention has focused on the role of growth factors in the CNS. Oligodendrocyte precursor migration in vitro is driven by PDGF and FGF-2 (Armstrong et al., 1990; Milner et al., 1997a; Noble et al., 1988). Transplanted precursors expressing a au6 dominant-negative form of the FGF receptor fail to migrate (Osterhout et al., 1997), supporting a role for FGF signaling in migration in vivo. Proliferation is stimulated by PDGF, FGFs, neurotrophin-3, neuregulins (NRGs) and insulin-like growth factors (IGFs) (Barres et al., 1994; Bogler et al., 1990; Canoll et al., 1996; Jiang et al., 2001; McMorris and Dubois-Dalcq, 1988; Richardson et al., 1988). Changes in proliferation in vivo in mice with altered levels of PDGF conWrm an essential role for this growth factor (Calver et al., 1998; Fruttiger et al., 1999), and studies examining cell cycle times in response to PDGF in vitro and in vivo suggest that the physiological concentration of PDGF is less then 1 ng/ml (van Heyningen et al., 2001). The timing of diVerentiation is regulated by a combination of an intrinsic clock mechanism and the availability of growth factors, with exit from the cell cycle preceding terminal diVerentiation (Barres and RaV, 1994; van Heyningen et al., 2001). The PDGFa receptor is down-regulated once terminal diVerentiation has been initiated (Butt et al., 1997; Hart et al., 1989a), although au7 responsiveness to PDGF decreases prior to receptor loss (Hart et al., 1989a). Growth au8 factors can also inXuence cell fate choices by the precursors. FGFs prevent or reverse expression of myelin markers (Bansal and PfeiVer, 1997; Cohen and Chandross, 2000; Goddard et al., 1999). BMPs can induce astrocyte diVerentiation (Grinspan et al., 2000; Mabie et al., 1997) and reversion of OPs to an FGF-responsive neural stem cell phenotype when used on PDGF-expanded precursor cells (Kondo and RaV, 2000). Multiple growth factors can mediate survival of newly formed oligodendrocytes in vitro including PDGF, IGFs, NRGs, and CNTF (Barres et al., 1993; Flores et al., 2000; Louis et al., 1993). In vivo, however, evidence exists only for PDGF and NRG as survival factors during development (Barres et al., 1992a; Colognato et al., 2002; Fernandez et al., 2000). The role of growth factors in myelin membrane formation is not well understood, but mice lacking IGF1 show hypomyelination (Beck et al., 1995), while overexpression or administration of exogenous IGF-1 promotes myelination in vivo (Butt and Berry, 2000; Carson et al., 1993; Ye et al., 1995). Some of the intracellular signaling molecules activated by these growth factors and other extracellular cues have been identiWed. Precursor proliferation involves

619

620

25. INTEGRINS

au10

PI3K, PKC, MAPK, and pp70 S6 kinase signaling (Baron et al., 2000, 2002; Ebner et al., 2000). PI3K and MAPK are also components of survival signaling in newly formed oligodendrocytes (Colognato et al., 2002; Flores et al., 2000; Vemuri and McMorris, 1996), and the survival-promoting growth factors PDGF and CNTF also activate the JAK/STAT signaling pathway (Dell’Albani et al., 1998). Given these well-proven instructive roles for growth factors in oligodendrocyte diVerentiation, ampliWcation of growth factor signaling pathways represents an attractive candidate mechanism by which integrins might regulate oligodendrocyte development. We have demonstrated such ampliWcation for two integrins, avb3 and a6b1. As described earlier, avb3 stimulates proliferation when the integrin is over-expressed in an oligodendrocyte precursor cell line (Blaschuk et al., 2000), and over-expression of a dominantnegative b3 integrin reduces the proliferative eVect of PDGF in primary oligodendrocyte precursors (Baron et al., 2002). Experiments using diVerent extracellular matrix substrates to bind either av integrins (Wbronectin, vitronectin) or a6 integrins (laminins) show that integrin signaling reduces the concentration of PDGF required for proliferation or survival when cells are grown in minimal culture conditions. The level required for proliferation drops from 10 to 0.1–1ng /ml, an eVect that can be blocked with antibodies against avb3 conWrming the role of the integrin in this ampliWcation (Baron et al., 2002). Equally, a6b1 substrates reduce the concentration of PDGF or NRG required to promote oligodendrocyte survival, while those binding to av integrins have no eVect on survival (Colognato et al., 2002; Frost et al., 1999). The signiWcance of these results becomes clear when one considers the likely physiological concentrations of these growth factors in vivo. As cited earlier, studies comparing cell cycle lengths of proliferating embryonic oligodendrocyte precursors in vivo with those grown in vitro under diVerent PDGF concentrations suggest a physiological concentration of this mitogen in between 0.1 and 1 ng/ml (van Heyningen et al., 2001). Concentrations of PDGF and NRG in vivo during the stage at which newly diVerentiated oligodendrocytes undergo extensive apoptosis are not established, but two sets of experiments show that both are present at limiting concentrations. Cell lines expressing PDGF will, when transplanted into the developing CNS, increase the number of surviving oligodendrocytes (Barres et al., 1992). Soluble forms of the erbB NRG receptors that can act as inhibitors of NRG interactions with normal cell surface erbB receptors will increase oligodendrocyte apoptosis in vivo (Fernandez et al., 2000). The results on avb3 emphasise that integrin signaling will enhance oligodendrocyte precursor proliferation in vivo. Indeed, the integrin appears to be required as proliferation at physiological concentrations of PGDF in the cell culture experiments was only seen on avb3 ligands. If avb3 integrin ligands are expressed on the axonal surface, as is known to be the case for the immunoglobulin superfamily molecules Thy-1 (Leyton et al., 2001; Morris, 1985; Xue et al., 1991) and L1 (Montgomery et al., 1996; Persohn and Schachner, 1987; Yip et al., 1998), then integrin signaling provides a mechanism to ensure that proliferation in response to PDGF is restricted to those locations such as axon tracts where it is appropriate. However, other ligands such as extracellular matrix molecules expressed in developing white matter tracts may also play a signiWcant part. Mice lacking the extracellular matrix molecule tenascin-C, which binds a number of integrins including avb3, show decreased rates of oligodendrocyte precursor proliferation in the developing CNS (Garcion et al., 2001). Cell culture experiments using cells from these mice show, as would be predicted from the results stated earlier, that tenascin-C null oligodendrocyte precursors show reduced sensitivity to the proliferative eVects of PDGF, and the addition of exogenous tenascin-C to the experiment rescues this eVect. It is likely, therefore, that a combination of axonal signals and cues deriving from other cells, such as astrocytes, within white matter tracts will provide integrin ligands. This may ensure that the proliferative response to long-range cues such as PDGF is spatially controlled by the much more restricted distribution of integrin ligands. This concept of integrins acting to restrict the response to soluble growth factors to biological appropriate locations is also very clearly demonstrated by studies on oligodendrocyte survival. As discussed earlier, this is an example of ligand-dependent survival

MECHANISMS AND CONSEQUENCES OF INTEGRIN GROWTH FACTOR INTERACTION IN OLIGODENDROCYTES

with axonal contact preventing apoptosis. The identity of the axonal signals has been elusive. A recent study showed that NRGs, of which some isoforms are expressed on the axonal surface, promotes survival (Fernandez et al., 2000). This class of growth factors therefore provides one such signal for those isoforms attached to the cell surface. Another axonal survival signal, however, appears to be the extracellular matrix molecule laminina2. Perhaps surprisingly for a member of a family of extracellular matrix molecules often associated with basal lamina, laminin-a2 is found on the surface of axons at the time of myelination in cerebellum and brain stem (Colognato et al., 2002; Powell et al., 1998). As discussed earlier, laminin-a2 enhances survival signaling in response to both PDGF and NRG, and a6 deWcient mice show increased oligodendrocyte apoptosis in the brain stem (Colognato et al., 2002). As discussed earlier, laminins containing the a2 chain enhance the sensitivity of the cells contacting the axon to PDGF and NRG, so enabling that cell to compete more eVectively for the limiting concentrations of available growth factor and thereby survive.

MECHANISMS AND CONSEQUENCES OF INTEGRIN GROWTH FACTOR INTERACTION IN OLIGODENDROCYTES Integrin/growth factor signaling interactions have been studied in many diVerent cell types (Assoian and Schwartz, 2001; Boudreau and Bissell, 1998; Schwartz and Baron, 1999), and these interactions have been described at the level of shared downstream signaling molecules as well as directly between the two cell surface receptors. In particular, one PDGF receptor, the PDGFb receptor, has been shown to interact with avb3 (Schneller et al., 1997). This suggests that a direct interaction between the integrin and the PDGF receptor present in oligodendrocytes, the PDGFa receptor, might mediate the observed interaction in the regulation of proliferation. We conWrmed that such an interaction does occur in the presence of both PDGF and an integrin ligand, and also showed that one eVect of growth factor signaling was to activate the integrin (Baron et al., 2002). As described earlier, activation involves a change in shape of the extracellular domain resulting in an increase in aYnity of the integrin for ligand. This shape change can result from binding to the cytoplasmic domain of intracellular signaling molecules, a process termed ‘‘inside-out’’ signaling (Hynes, 2002). In oligodendrocyte precursors, we showed using an engineered monovalent antibody that recognises activated av integrins (Pampori et al., 1999) that activation can be driven by PDGF and occurs in the absence of any integrin ligand (Baron et al., 2002). However, this in itself does not result in increased proliferation. The recognition of a ligand by the activated integrin does now stimulate proliferation and, critically, activation of the integrins and integrin ligand binding is suYcient for proliferation; under these conditions the growth factor PDGF is not required for proliferation. A similar mechanism has been proposed for the response of endothelial cells to VEGF (Byzova et al., 2000), and it is of interest for two reasons. First, it explains the requirement for an extracellular matrix ligand in the cell culture experiments using physiological concentrations of growth factors, as activated integrins cannot signal without such a ligand. Higher, nonphysiological, concentrations of growth factors are presumably able to activate alternative signaling pathways in the cell that do not require integrins. Second, a model in which integrins are downstream of the mitogenic growth factor receptor rather than operating in parallel allows them to act as integrators of many diVerent signaling pathways that can aVect their activation. One obvious example is provided by the eph/ephrins, shown to regulate integrin activation and levels of cell adhesion in other systems (Davy and Robbins, 2000; Huai and Drescher, 2001; Huynh-Do et al., 1999). We have previously suggested a role in migration for another of the av integrins, avb1 (Milner et al., 1996). Future work will therefore determine whether migration, like proliferation, is regulated at least in part by the control of integrin activation and also whether the survival-promoting a6b1 integrin interacts with growth factor receptors in a similar manner. The observation that growth factor signaling may require cooperation with integrins to exert a biological eVect also has the important implication that the identity of the integrin

621

622

25. INTEGRINS

associated with a growth factor receptor at a particular developmental stage may be instructive in determining the eVect of the growth factor. This might be particularly important both for PDGF, which has been shown (as described earlier) to regulate migration, proliferation, and survival, and for NRG that regulates both proliferation and survival. The model would propose that changes in integrin association of a given growth factor receptor during development will switch the downstream pathways activated and so allow the growth factor to have distinct eVects at diVerent stages of development— in other words, explain how one growth factor can do two things at diVerent times. This hypothesis is supported by two independent lines of work examining the transition from proliferation to survival signaling in oligodendrocytes. First, studies of the composition of the lipid rafts previously described in oligodendrocytes (Simons et al., 2000) reveal that the PDGFa receptor becomes enriched in the rafts as oligodendrocyte diVerentiation commences (Baron et al., 2003). Additionally, the two major classes of integrins are present in diVerent compartments of the membrane; av integrins are present in the nonraft compartment while a6 integrins are highly enriched in the rafts. Consequently, a shift of the PDGFa receptor from the nonraft to the raft compartment will alter the potential partner from av to a6 integrins. ImmunoXuorescence studies of individual cells and immunoprecipitation experiments of the lipid raft fraction from oligodendrocyte cultures do show that the PDGFa receptor is associated with a6b1 integrin when the cells are exposed to laminin-2 substrates (Baron et al., 2003). However, similar immunoprecipitation studies earlier in oligodendroglial development reveal an association of the PDGFa receptor and avb3 when the cells are exposed to av ligands (Baron et al., 2002). As av and a6 integrins regulate proliferation and survival respectively, the eVect of this switch in integrin partner will be a change in PDGF signaling from proliferation to survival without any necessary changes in the expression level of the growth factor receptor itself (Fig. 25.3). A second example of how integrins control growth factor signaling comes from work on NRG survival signaling. As discussed earlier, the axonal laminin-a2 provides an axonal survival signal. Studies using pharmacological inhibitors and antibodies that distinguish the phosphorylation patterns of the pro-apoptotic molecule BAD in response to either PI3K or MAPK signaling show that NRG signaling is switched from PI3K to a MAPK pathway when a6b1 binds laminin-2. Critically, the presence of this switch can be conWrmed in vivo by the observation that the a6-deWcient mice show reduced levels of BAD phosphorylated by MAPK (Colognato et al., 2002). As PI3K potentiates survival but inhibits diVerentiation, while MAPK promotes survival and at the same time facilitates diVerentiation, the consequence will be a switch such that NRG signaling now promotes diVerentiation while maintaining survival at precisely the time that diVerentiation is appropriate (i.e., when axonal contact has been established). This result provides an explanation for the puzzling previous observations that NRG appears to inhibit diVerentiation (Canoll et al., 1996, 1999). As these experiments were performed on non-laminin substrates, the growth factor will activate PI3K rather than the physiological MAPK, and PI3K signaling will prevent diVerentiation.

INTEGRINS AND MYELIN DISEASES The recognition that integrin/extracellular matrix interactions are regulators of oligodendrocyte and Schwann cell behavior provides insight into the pathogenesis of a number of well-recognized diseases of myelin. The genetic disorder congenital muscular dystrophy can be caused by a deWciency of laminin-2/merosin and in these cases can be associated with hypomyelination in the CNS (as judged by MRI abnormalities) (Jones et al., 2001; Miyagoe-Suzuki et al., 2000). While the muscle dystrophy is most likely to reXect a lack of laminin-2 in the basal lamina of muscle and peripheral nerve (Matsumura et al., 1997), the CNS phenotype has been harder to explain. The evidence that laminin-a2 provides an axonal survival and diVerentiation signal for oligodendrocytes (Colognato et al., 2002) suggests that laminin-2 deWciency would be associated with increased oligodendrocyte

623

INTEGRINS AND MYELIN DISEASES

Lipid raft

αvβ3

PDGFαR

α6β1

proliferation

survival

FIGURE 25.3 A hypothesis for the regulation of growth factor eVect by integrin association. In oligodendrocyte precursors (left), the PDGF receptor is associated with, and activates, avb3 and so stimulates cell proliferation. In newly diVerentiated oligodendrocytes (right), the PDGF receptor switches to the raft compartment (as illustrated by the curved arrow) and becomes associated with the a6b1 integrin present in this compartment. As a result, the eVect of the growth factor is now to promote survival.

apoptosis and delayed diVerentiation. This would in turn delay myelination (a phenotype also seen in the a6 knockout mice) and so explain the MRI abnormalities seen during development in these children. However, laminin-2 can also promote myelin membrane formation (Buttery and Vrench-Constant, 1999), so other deWcits in oligodendrocyte development could also contribute to the radiological abnormalities observed. Perturbations in integrin signaling might also explain an entirely diVerent genetic disorder, neuroWbromatosis type 2. This disease is associated with Schwann cell tumors in vestibular and other nerves, reXecting a loss of normal Schwann cell growth control (Evans et al., 2000). This results from abnormalities in schwannomin (also known as merlin) that result from mutations in the encoding gene. Fernandez-Valle and colleagues have recently shown that schwannomin is associated with b1 integrin and paxillin in Schwann cells (Fernandez-Valle et al., 2002). Some mutations that cause neuroWbromatosis type 2 result in amino acid changes in the domains of schwannomin that interact directly with paxillin. They show that schwannomin can promote the internalization of paxillin away from the integrin complex. This might limit the cell motility and proliferation that results from the recruitment by paxillin of the PXL-PIX-PaK complex, and the loss of regulated activity of integrin-associated complexes may therefore be important in allowing the Schwann cell overgrowth seen in the disease. PLP mutations cause myelin diseases in a wide range of species including mice (the jimpy mutation) and Pelizaeus-Merzbacher disease and spastic paraparesis in humans (Koeppen and Robitaille, 2002; SchiVmann and BoespXug-Tanguy, 2001). The important observation from the Macklin laboratory that PLP interacts with integrins in oligodendrocytes (Gudz et al., 2002) raises the possibility that Pelizaeus-Merzbacher disease is, at least in part, caused by alterations in intracellular pathways to which integrins may make a signiWcant contribution. The three genetic diseases described earlier are all comparatively rare. A much more common disorder of myelin, aVecting as many as 1 in 800 of Northern Europeans and descendants, is multiple sclerosis. The aetiology and pathology of this disease are subjects of intense study, but an important point that has already emerged is that failure of repair is a signiWcant contributor to the chronic demyelination and axonal loss that characterizes the disease (Compston and Coles, 2002). A key observation has been the description of cells with the phenotype of oligodendrocyte precursors within chronic MS lesions (Chang

624

25. INTEGRINS

et al., 2000; Scolding et al., 1998; Wolswijk, 1998). The problem, therefore, is one of stalled diVerentiation rather than depletion of cells with repair capacity. Why does diVerentiation fail in these lesions? It is not simply loss of axons, as cells contacting axons but unable to diVerentiate are seen in some lesions (Chang et al., 2002). The work on integrins described earlier suggests one mechanism. If growth factor signaling requires the presence of integrin ligands, as suggested from studies of proliferation and survival during development, then the ‘‘correct’’ extracellular matrix within the MS lesion will be required for repair. However, the extracellular matrix of the lesion will have been altered extensively by the loss of the blood brain barrier (with the entry of many diVerent serum extracellular matrix components) and the invasion of the tissue by T-cells and macrophages. Changes in extracellular matrix proteins such as tenascins and proteoglycans have already been described (Gutowski et al., 1999; Sobel and Ahmed, 2001), and it therefore seems certain that the extracellular matrix of an MS lesion will be very diVerent to that present during development. We suggest that this radically diVerent extracellular matrix will inhibit eVective oligodendrocyte precursor diVerentiation and repair by preventing appropriate growth factor responses. This hypothesis needs to be tested by a complete analysis of the extracellular matrix in multiple sclerosis, with the diYculty that mRNA expression studies will not be suYcient as they will not detect the many serum components that have leaked into the tissue. Two important consequences of the hypothesis are, Wrst, that growth factor therapy alone will not be suYcient to promote repair and, second, that strategies designed to activate integrin signaling pathways that lie downstream of both integrin and growth factor receptors represent attractive therapeutic approaches for enhancing repair in MS.

Acknowledgments Work in the author’s laboratory cited in this chapter has been funded by the Wellcome Trust, the Multiple Sclerosis Society of Great Britain and Northern Ireland, the Medical Research Council, and Action Research. I am grateful to Holly Colognato, Laurence Decker, Jon Moore, and Mie Olsen for helpful comments on the manuscript.

References Anton, E. S., Kreidberg, J. A., and Rakic, P. (1999). Distinct functions of alpha3 and alpha(v) integrin receptors in neuronal migration and laminar organization of the cerebral cortex. Neuron 22, 277–289. Armstrong, R. C., Harvath, L., and Dubois-Dalcq, M. E. (1990). Type 1 astrocytes and oligodendrocyte-type 2 astrocyte glial progenitors migrate toward distinct molecules. J Neurosci Res 27, 400–407. Assoian, R. K., and Schwartz, M. A. (2001). Coordinate signaling by integrins and receptor tyrosine kinases in the regulation of G1 phase cell-cycle progression. Curr Opin Genet Dev 11, 48–53. Bahr, B. A., Staubli, U., Xiao, P., Chun, D., Ji, Z. X., Esteban, E. T., and Lynch, G. (1997). Arg-Gly-Asp-Serselective adhesion and the stabilization of long-term potentiation: Pharmacological studies and the characterization of a candidate matrix receptor. J Neurosci 17, 1320–1329. Bansal, R., and PfeiVer, S. E. (1997). FGF-2 converts mature oligodendrocytes to a novel phenotype. J Neurosci Res 50, 215–228. Baron, W., Decker, L., Colognato, H., and Vrench-Constant, C. (2003). Regulation of integrin growth factor interactions in oligodendrocytes by lipid raft microdomains. Curr. Biol., 13, 151–155. Baron, W., Metz, B., Bansal, R., Hoekstra, D., and de Vries, H. (2000). PDGF and FGF-2 signaling in oligodendrocyte progenitor cells: Regulation of proliferation and diVerentiation by multiple intracellular signaling pathways. Mol Cell Neurosci 15, 314–29. Baron, W., Shattil, S. J., and Vrench-Constant, C. (2002). The oligodendrocyte precursor mitogen PDGF stimulates proliferation by activation of alpha(v)beta3 integrins. Embo J 21, 1957–66. Barres, B. A., Hart, I. K., Coles, H. S., Burne, J. F., Voyvodic, J. T., Richardson, W. D., and RaV, M. C. (1992a). Cell death and control of cell survival in the oligodendrocyte lineage. Cell 70, 31–46. Barres, B. A., Hart, I. K., Coles, H. S., Burne, J. F., Voyvodic, J. T., Richardson, W. D., and RaV, M. C. (1992b). Cell death in the oligodendrocyte lineage. J Neurobiol 23, 1221–1230. Barres, B. A., and RaV, M. C. (1994). Control of oligodendrocyte number in the developing rat optic nerve. Neuron 12, 935–492. Barres, B. A., and RaV, M. C. (1999). Axonal control of oligodendrocyte development. J Cell Biol 147, 1123–1128. Barres, B. A., RaV, M. C., Gaese, F., Bartke, I., Dechant, G., and Barde, Y. A. (1994). A crucial role for neurotrophin-3 in oligodendrocyte development. Nature 367, 371–375.

INTEGRINS AND MYELIN DISEASES

Barres, B. A., Schmid, R., Sendnter, M., and RaV, M. C. (1993). Multiple extracellular signals are required for long-term oligodendrocyte survival. Development 118, 283–295. Bartsch, U., Faissner, A., Trotter, J., Dorries, U., Bartsch, S., Mohajeri, H., and Schachner, M. (1994). Tenascin demarcates the boundary between the myelinated and nonmyelinated part of retinal ganglion cell axons in the developing and adult mouse. J Neurosci 14, 4756–4768. Bazzoni, G., and Hemler, M. E. (1998). Are changes in integrin aYnity and conformation overemphasized? Trends Biochem Sci 23, 30–34. Beck, K. D., Powell-Braxton, L., Widmer, H. R., Valverde, J., and Hefti, F. (1995). Igf1 gene disruption results in reduced brain size, CNS hypomyelination, and loss of hippocampal granule and striatal parvalbumincontaining neurons. Neuron 14, 717–730. Ben-Hur, T., Rogister, B., Murray, K., Rougon, G., and Dubois-Dalcq, M. (1998). Growth and fate of PSANCAMþ precursors of the postnatal brain. J Neurosci 18, 5777–5788. Berditchevski, F. (2001). Complexes of tetraspanins with integrins: More than meets the eye. J Cell Sci 114, 4143–4151. Bergles, D. E., Roberts, J. D., Somogyi, P., and Jahr, C. E. (2000). Glutamatergic synapses on oligodendrocyte precursor cells in the hippocampus. Nature 405, 187–191. Bhat, S., and Silberberg, D. H. (1988). Developmental expression of neural cell adhesion molecules of oligodendrocytes in vivo and in culture. J Neurochem 50, 1830–1838. Birling, M. C., Tait, S., Hardy, R. J., and Brophy, P. J. (1999). A novel rat tetraspan protein in cells of the oligodendrocyte lineage. J Neurochem 73, 2600–8. Blaschuk, K. L., Frost, E. E., and Vrench-Constant, C. (2000). The regulation of proliferation and diVerentiation in oligodendrocyte progenitor cells by alphaV integrins. Development 127, 1961–1969. Bogler, O., Wren, D., Barnett, S. C., Land, H., and Noble, M. (1990). Cooperation between two growth factors promotes extended self-renewal and inhibits diVerentiation of oligodendrocyte-type-2 astrocyte (O-2A) progenitor cells. Proc Natl Acad Sci U S A 87, 6368–6372. Borges, E., Jan, Y., and Ruoslahti, E. (2000). Platelet-derived growth factor receptor beta and vascular endothelial growth factor receptor 2 bind to the beta 3 integrin through its extracellular domain. J Biol Chem 275, 39867–39873. Boudreau, N., and Bissell, M. J. (1998). Extracellular matrix signaling: Integration of form and function in normal and malignant cells. Curr Opin Cell Biol 10, 640–646. Bronner-Fraser, M. (1993). Mechanisms of neural crest cell migration. Bioessays 15, 221–230. Bronner-Fraser, M., Artinger, M., Muschler, J., and Horwitz, A. F. (1992). Developmentally regulated expression of alpha 6 integrin in avian embryos. Development 115, 197–211. Bronner-Fraser, M., and Lallier, T. (1988). A monoclonal antibody against a laminin-heparan sulfate proteoglycan complex perturbs cranial neural crest migration in vivo. J Cell Biol 106, 1321–1329. Burne, J. F., Staple, J. K., and RaV, M. C. (1996). Glial cells are increased proportionally in transgenic optic nerves with increased numbers of axons. J Neurosci 16, 2064–2073. Butt, A. M., and Berry, M. (2000). Oligodendrocytes and the control of myelination in vivo: New insights from the rat anterior medullary velum. J Neurosci Res 59, 477–4. Butt, A. M., Hornby, M. F., Ibrahim, M., Kirvell, S., Graham, A., and Berry, M. (1997). PDGF-alpha receptor and myelin basic protein mRNAs are not coexpressed by oligodendrocytes in vivo: A double in situ hybridization study in the anterior medullary velum of the neonatal rat. Mol Cell Neurosci 8, 311–322. Buttery, P. C., and Vrench-Constant, C. (1999). Laminin-2/integrin interactions enhance myelin membrane formation by oligodendrocytes. Mol Cell Neurosci 14, 199–212. Byzova, T. V., Goldman, C. K., Pampori, N., Thomas, K. A., Bett, A., Shattil, S. J., and Plow, E. F. (2000). A mechanism for modulation of cellular responses to VEGF: Activation of the integrins. Mol Cell 6, 851–860. Calver, A. R., Hall, A. C., Yu, W. P., Walsh, F. S., Heath, J. K., Betsholtz, C., and Richardson, W. D. (1998). Oligodendrocyte population dynamics and the role of PDGF in vivo. Neuron 20, 869–882. Canoll, P. D., Kraemer, R., Teng, K. K., Marchionni, M. A., and Salzer, J. L. (1999). GGF/neuregulin induces a phenotypic reversion of oligodendrocytes. Mol Cell Neurosci 13, 79–94. Canoll, P. D., Musacchio, J. M., Hardy, R., Reynolds, R., Marchionni, M. A., and Salzer, J. L. (1996). GGF/ neuregulin is a neuronal signal that promotes the proliferation and survival and inhibits the diVerentiation of oligodendrocyte progenitors. Neuron 17, 229–243. Cardwell, M. C., and Rome, L. H. (1988). RGD-containing peptides inhibit the synthesis of myelin-like membrane by cultured oligodendrocytes. J Cell Biol 107, 1551–1559. Carson, M. J., Behringer, R. R., Brinster, R. L., and McMorris, F. A. (1993). Insulin-like growth factor I increases brain growth and central nervous system myelination in transgenic mice. Neuron 10, 729–740. Chandran, S., Svendsen, C., Compston, A., and Scolding, N. (1998). Regional potential for oligodendrocyte generation in the rodent embryonic spinal cord following exposure to EGF and FGF-2. Glia 24, 382–329. Chang, A., Nishiyama, A., Peterson, J., Prineas, J., and Trapp, B. D. (2000). NG2-positive oligodendrocyte progenitor cells in adult human brain and multiple sclerosis lesions. J Neurosci 20, 6404–6412. Chang, A., Tourtellotte, W. W., Rudick, R., and Trapp, B. D. (2002). Premyelinating oligodendrocytes in chronic lesions of multiple sclerosis. N Engl J Med 346, 165–173. Charles, P., Hernandez, M. P., StankoV, B., Aigrot, M. S., Colin, C., Rougon, G., Zalc, B., and Lubetzki, C. (2000). Negative regulation of central nervous system myelination by polysialylated-neural cell adhesion molecule. Proc Natl Acad Sci U S A 97, 7585–7590.

625

626

25. INTEGRINS

au12

Chavis, P., and Westbrook, G. (2001). Integrins mediate functional pre-and postsynaptic maturation at a hippocampal synapse. Nature 411, 317–321. Chen, L. M., Bailey, D., and Fernandez-Valle, C. (2000). Association of beta 1 integrin with focal adhesion kinase and paxillin in diVerentiating Schwann cells. J Neurosci 20, 3776–3784. Claas, C., Stipp, C. S., and Hemler, M. E. (2001). Evaluation of prototype transmembrane 4 superfamily protein complexes and their relation to lipid rafts. J Biol Chem 276, 7974–8794. Clark, E. A., and Brugge, J. S. (1995). Integrins and signal transduction pathways: The road taken. Science 268, 233–239. Cohen, R. I., and Chandross, K. J. (2000). Fibroblast growth factor-9 modulates the expression of myelin related proteins and multiple Wbroblast growth factor receptors in developing oligodendrocytes. J Neurosci Res 61, 273–287. Colognato, C., Baron, W., Avellana-Adalid, V., Relvas, J., Baron-van Evercooren, A., Georges-Labouesse, E., and Vrench-Constant, C. (2002). CNS integrins switch growth factor signalling to promote target dependent survival. Nature Cell Biol. 4, 833–841. Compston, A., and Coles, A. (2002). Multiple sclerosis. Lancet 359, 1221–1231. Compston, A., Zajicek, J., Sussman, J., Webb, A., Hall, G., Muir, D., Shaw, C., Wood, A., and Scolding, N. (1997). Glial lineages and myelination in the central nervous system. J Anat 190, 161–200. Condic, M. L. (2001). Adult neuronal regeneration induced by transgenic integrin expression. J Neurosci 21, 4782–4788. Corley, S. M., Ladiwala, U., Besson, A., and Yong, V. W. (2001). Astrocytes attenuate oligodendrocyte death in vitro through an alpha(6) integrin-laminin-dependent mechanism. Glia 36, 281–294. Cukierman, E., Pankov, R., Stevens, D. R., and Yamada, K. M. (2001). Taking cell-matrix adhesions to the third dimension. Science 294, 1708–1712. Datta, S. R., Brunet, A., and Greenberg, M. E. (1999). Cellular survival: A play in three Akts. Genes Dev 13, 2905–2927. Davy, A., and Robbins, S. M. (2000). Ephrin-A5 modulates cell adhesion and morphology in an integrindependent manner. Embo J 19, 5396–5405. Decker, L., Avellana-Adalid, V., Nait-Oumesmar, B., Durbec, P., and Baron-Van Evercooren, A. (2000). Oligodendrocyte precursor migration and diVerentiation: Combined eVects of PSA residues, growth factors, and substrates. Mol Cell Neurosci 16, 422–439. Dedhar, S. (2000). Cell-substrate interactions and signaling through ILK. Curr Opin Cell Biol 12, 250–256. Delannet, M., Martin, F., Bossy, B., Cheresh, D. A., Reichardt, L. F., and Duband, J. L. (1994). SpeciWc roles of the alpha V beta 1, alpha V beta 3 and alpha V beta 5 integrins in avian neural crest cell adhesion and migration on vitronectin. Development 120, 2687–2702. Dell’Albani, P., Kahn, M. A., Cole, R., Condorelli, D. F., GiuVrida-Stella, A. M., and de Vellis, J. (1998). Oligodendroglial survival factors, PDGF-AA and CNTF, activate similar JAK/STAT signaling pathways. J Neurosci Res 54, 191–205. Desban, N., and Duband, J. L. (1997). Avian neural crest cell migration on laminin: Interaction of the alpha1beta1 integrin with distinct laminin-1 domains mediates diVerent adhesive responses. J Cell Sci 110, 2729–2744. Detrait, E., Laduron, S., Meremans, V., Baron-Van Evercooren, A., van den Bosch de Aguilar, P., and Knoops, B. (1999). Expression of integrins by murine MSC80 Schwann cell line: Relationship to cell adhesion and migration. Neurosci Lett 267, 49–52. Diers-Fenger, M., KirchhoV, F., Kettenmann, H., Levine, J. M., and Trotter, J. (2001). AN2/NG2 proteinexpressing glial progenitor cells in the murine CNS: Isolation, diVerentiation, and association with radial glia. Glia 34, 213–28. Dong, Z., Brennan, A., Liu, N., Yarden, Y., Lefkowitz, G., Mirsky, R., and Jessen, K. R. (1995). Neu diVerentiation factor is a neuron-glia signal and regulates survival, proliferation, and maturation of rat Schwann cell precursors. Neuron 15, 585–596. Duband, J. L., Belkin, A. M., Syfrig, J., Thiery, J. P., and Koteliansky, V. E. (1992). Expression of alpha 1 integrin, a laminin-collagen receptor, during myogenesis and neurogenesis in the avian embryo. Development 116, 585–600. Ebner, S., Dunbar, M., and McKinnon, R. D. (2000). Distinct roles for PI3K in proliferation and survival of oligodendrocyte progenitor cells. J Neurosci Res 62, 336–345. Einheber, S., Milner, T. A., Giancotti, F., and Salzer, J. L. (1993). Axonal regulation of Schwann cell integrin expression suggests a role for alpha 6 beta 4 in myelination. J Cell Biol 123, 1223–36. Eldridge, C. F., Bunge, M. B., Bunge, R. P., and Wood, P. M. (1987). DiVerentiation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. J Cell Biol 105, 1023–34. Etienne-Manneville, S., and Hall, A. (2002). Rho GTPases in cell biology. Nature 420, 629–635. Evans, D. G., Sainio, M., and Baser, M. E. (2000). NeuroWbromatosis type 2. J Med Genet 37, 897–904. Evans, J. P. (2001). Fertilin beta and other ADAMs as integrin ligands: Insights into cell adhesion and fertilization. Bioessays 23, 628–639. Feltri, M. L., Graus Porta, D., Previtali, S. C., Nodari, A., Migliavacca, B., Cassetti, A., Littlewood-Evans, A., Reichardt, L. F., Messing, A., Quattrini, A., Mueller, U., and Wrabetz, L. (2002). Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with axons. J Cell Biol 156, 199–209.

INTEGRINS AND MYELIN DISEASES

Feltri, M. L., Scherer, S. S., Nemni, R., Kamholz, J., Vogelbacker, H., Scott, M. O., Canal, N., Quaranta, V., and Wrabetz, L. (1994). Beta 4 integrin expression in myelinating Schwann cells is polarized, developmentally regulated and axonally dependent. Development 120, 1287–1301. Fernandez, P. A., Tang, D. G., Cheng, L., Prochiantz, A., Mudge, A. W., and RaV, M. C. (2000). Evidence that axon-derived neuregulin promotes oligodendrocyte survival in the developing rat optic nerve. Neuron 28, 81–90. Fernandez-Valle, C., Gorman, D., Gomez, A. M., and Bunge, M. B. (1997). Actin plays a role in both changes in cell shape and gene-expression associated with Schwann cell myelination. J Neurosci 17, 241–250. Fernandez-Valle, C., Gwynn, L., Wood, P. M., Carbonetto, S., and Bunge, M. B. (1994). Anti-beta 1 integrin antibody inhibits Schwann cell myelination. J Neurobiol 25, 1207–1226. Fernandez-Valle, C., Tang, Y., Ricard, J., Rodenas-Ruano, A., Taylor, A., Hackler, E., BiggerstaV, J., and Iacovelli, J. (2002). Paxillin binds schwannomin and regulates its density-dependent localization and eVect on cell morphology. Nat Genet 31, 354–362. Vrench-Constant, C., Miller, R. H., Burne, J. F., and RaV, M. C. (1988). Evidence that migratory oligodendrocyte-type-2 astrocyte (O-2A) progenitor cells are kept out of the rat retina by a barrier at the eye-end of the optic nerve. J Neurocytol 17, 13–25. Flores, A. I., Mallon, B. S., Matsui, T., Ogawa, W., Rosenzweig, A., Okamoto, T., and Macklin, W. B. (2000). Akt-mediated survival of oligodendrocytes induced by neuregulins. J Neurosci 20, 7622–7630. Frei, R., Dowling, J., Carenini, S., Fuchs, E., and Martini, R. (1999). Myelin formation by Schwann cells in the absence of beta4 integrin. Glia 27, 269–274. Frost, E. E., Buttery, P. C., Milner, R., and Vrench-Constant, C. (1999). Integrins mediate a neuronal survival signal for oligodendrocytes. Curr Biol 9, 1251–4. Fruttiger, M., Karlsson, L., Hall, A. C., Abramsson, A., Calver, A. R., Bostrom, H., Willetts, K., Bertold, C. H., Heath, J. K., Betsholtz, C., and Richardson, W. D. (1999). Defective oligodendrocyte development and severe hypomyelination in PDGF-A knockout mice. Development 126, 457–467. Fuss, B., Wintergerst, E. S., Bartsch, U., and Schachner, M. (1993). Molecular characterization and in situ mRNA localization of the neural recognition molecule J1–160/180: A modular structure similar to tenascin. J Cell Biol 120, 1237–1249. Garcion, E., Faissner, A., and Vrench-Constant, C. (2001). Knockout mice reveal a contribution of the extracellular matrix molecule tenascin-C to neural precursor proliferation and migration. Development 128, 2485–96. Gard, A. L., Maughon, R. H., and Schachner, M. (1996). In vitro oligodendrogliotrophic properties of cell adhesion molecules in the immunoglobulin superfamily: Myelin-associated glycoprotein and N-CAM. J Neurosci Res 46, 415–426. Garratt, A. N., Britsch, S., and Birchmeier, C. (2000). Neuregulin, a factor with many functions in the life of a schwann cell. Bioessays 22, 987–996. Georges-Labouesse, E., Mark, M., Messaddeq, N., and Gansmuller, A. (1998). Essential role of alpha 6 integrins in cortical and retinal lamination. Curr Biol 8, 983–986. Georges-Labouesse, E., Messaddeq, N., Yehia, G., Cadalbert, L., Dierich, A., and Le Meur, M. (1996). Absence of integrin alpha 6 leads to epidermolysis bullosa and neonatal death in mice. Nat Genet 13, 370–373. Ghiani, C., and Gallo, V. (2001). Inhibition of cyclin E-cyclin-dependent kinase 2 complex formation and activity is associated with cell cycle arrest and withdrawal in oligodendrocyte progenitor cells. J Neurosci 21, 1274–1282. Ghiani, C. A., Yuan, X., Eisen, A. M., Knutson, P. L., DePinho, R. A., McBain, C. J., and Gallo, V. (1999). Voltage-activated Kþ channels and membrane depolarization regulate accumulation of the cyclin-dependent kinase inhibitors p27(Kip1) and p21(CIP1) in glial progenitor cells. J Neurosci 19, 5380–5392. Giancotti, F. G., and Ruoslahti, E. (1999). Integrin signaling. Science 285, 1028–1032. Goddard, D. R., Berry, M., and Butt, A. M. (1999). In vivo actions of Wbroblast growth factor-2 and insulin-like growth factor-I on oligodendrocyte development and myelination in the central nervous system. J Neurosci Res 57, 74–85. Graus-Porta, D., Blaess, S., Senften, M., Littlewood-Evans, A., Damsky, C., Huang, Z., Orban, P., Klein, R., Schittny, J. C., and Muller, U. (2001). Beta1-class integrins regulate the development of laminae and folia in the cerebral and cerebellar cortex. Neuron 31, 367–379. Green, J. M., Zhelesnyak, A., Chung, J., Lindberg, F. P., Sarfati, M., Frazier, W. A., and Brown, E. J. (1999). Role of cholesterol in formation and function of a signaling complex involving alphavbeta3, integrinassociated protein (CD47), and heterotrimeric G proteins. J Cell Biol 146, 673–682. Grinspan, J. B., Edell, E., Carpio, D. F., Beesley, J. S., Lavy, L., Pleasure, D., and Golden, J. A. (2000). StagespeciWc eVects of bone morphogenetic proteins on the oligodendrocyte lineage. J Neurobiol 43, 1–17. Grotewiel, M. S., Beck, C. D., Wu, K. H., Zhu, X. R., and Davis, R. L. (1998). Integrin-mediated short-term memory in Drosophila. Nature 391, 455–460. Gudz, T. I., Schneider, T. E., Haas, T. A., and Macklin, W. B. (2002). Myelin proteolipid protein forms a complex with integrins and may participate in integrin receptor signaling in oligodendrocytes. J Neurosci 22, 7398–7407. Gutowski, N. J., Newcombe, J., and Cuzner, M. L. (1999). Tenascin-R and C in multiple sclerosis lesions: Relevance to extracellular matrix remodelling. Neuropathol Appl Neurobiol 25, 207–214. Haack, H., and Hynes, R. O. (2001). Integrin receptors are required for cell survival and proliferation during development of the peripheral glial lineage. Dev Biol 233, 38–55.

627

628

25. INTEGRINS

Hadjiargyrou, M., Kaprielian, Z., Kato, N., and Patterson, P. H. (1996). Association of the tetraspan protein CD9 with integrins on the surface of S-16 Schwann cells. J Neurochem 67, 2505–2513. Hadjiargyrou, M., and Patterson, P. H. (1995). An anti-CD9 monoclonal antibody promotes adhesion and induces proliferation of Schwann cells in vitro. J Neurosci 15, 574–583. Hart, I. K., Richardson, W. D., Bolsover, S. R., and RaV, M. C. (1989a). PDGF and intracellular signaling in the timing of oligodendrocyte diVerentiation. J Cell Biol 109, 3411–3417. Hart, I. K., Richardson, W. D., Heldin, C. H., Westermark, B., and RaV, M. C. (1989b). PDGF receptors on cells of the oligodendrocyte-type-2 astrocyte (O-2A) cell lineage. Development 105, 595–603. Hodivala-Dilke, K. M., McHugh, K. P., Tsakiris, D. A., Rayburn, H., Crowley, D., Ullman-Cullere, M., Ross, F. P., Coller, B. S., Teitelbaum, S., and Hynes, R. O. (1999). Beta3-integrin-deWcient mice are a model for Glanzmann thrombasthenia showing placental defects and reduced survival. J Clin Invest 103, 229–238. Hsiao, L. L., Peltonen, J., Jaakkola, S., Gralnick, H., and Uitto, J. (1991). Plasticity of integrin expression by nerve-derived connective tissue cells. Human Schwann cells, perineurial cells, and Wbroblasts express markedly diVerent patterns of beta 1 integrins during nerve development, neoplasia, and in vitro. J Clin Invest 87, 811–820. Huai, J., and Drescher, U. (2001). An ephrin-A-dependent signaling pathway controls integrin function and is linked to the tyrosine phosphorylation of a 120-kDa protein. J Biol Chem 276, 6689–6694. Huang, X., GriYths, M., Wu, J., Farese, R. V., Jr., and Sheppard, D. (2000). Normal development, wound healing, and adenovirus susceptibility in beta5-deWcient mice. Mol Cell Biol 20, 755–759. Hughson, E., Dowler, S., Geall, K., Johnson, G., and Rumsby, M. (1998). Rat oligodendrocyte O-2A precursor cells and the CG-4 oligodendrocyte precursor cell line express cadherins, beta-catenin and the neural cell adhesion molecule, NCAM. Neurosci Lett 251, 157–60. Humphries, M. J., Sheridan, J., Mould, A. P., and Newham, P. (1995). Mechanisms of VCAM-1 and Wbronectin binding to integrin alpha 4 beta 1: Implications for integrin function and rational drug design. Ciba Found Symp 189, 177–191. Huynh-Do, U., Stein, E., Lane, A. A., Liu, H., Cerretti, D. P., and Daniel, T. O. (1999). Surface densities of ephrin-B1 determine EphB1-coupled activation of cell attachment through alphavbeta3 and alpha5beta1 integrins. Embo J 18, 2165–73. Hynes, R. O. (1992). Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69, 11–25. Hynes, R. O. (2002). Integrins: Bidirectional, allosteric signaling machines. Cell 110, 673–687. Hynes, R. O., and Zhao, Q. (2000). The evolution of cell adhesion. J Cell Biol 150, F89-F96. Jiang, F., Frederick, T. J., and Wood, T. L. (2001). IGF-I synergizes with FGF-2 to stimulate oligodendrocyte progenitor entry into the cell cycle. Dev Biol 232, 414–423. Jones, K. J., Morgan, G., Johnston, H., Tobias, V., Ouvrier, R. A., Wilkinson, I., and North, K. N. (2001). The expanding phenotype of laminin alpha2 chain (merosin) abnormalities: Case series and review. J Med Genet 38, 649–657. Kiernan, B. W., Gotz, B., Faissner, A., and Vrench-Constant, C. (1996). Tenascin-C inhibits oligodendrocyte precursor cell migration by both adhesion-dependent and adhesion-independent mechanisms. Mol Cell Neurosci 7, 322–335. Kil, S. H., and Bronner-Fraser, M. (1996). Expression of the avian alpha 7-integrin in developing nervous system and myotome. Int J Dev Neurosci 14, 181–90. Kil, S. H., Krull, C. E., Cann, G., Clegg, D., and Bronner-Fraser, M. (1998). The alpha4 subunit of integrin is important for neural crest cell migration. Dev Biol 202, 29–42. Kil, S. H., Lallier, T., and Bronner-Fraser, M. (1996). Inhibition of cranial neural crest adhesion in vitro and migration in vivo using integrin antisense oligonucleotides. Dev Biol 179, 91–101. Knutson, P., Ghiani, C. A., Zhou, J. M., Gallo, V., and McBain, C. J. (1997). Kþ channel expression and cell proliferation are regulated by intracellular sodium and membrane depolarization in oligodendrocyte progenitor cells. J Neurosci 17, 2669–2682. Koeppen, A. H., and Robitaille, Y. (2002). Pelizaeus-Merzbacher disease. J Neuropathol Exp Neurol 61, 747–59. Kondo, T., and RaV, M. (2000). Oligodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells. Science 289, 1754–1757. Kramer, E. M., Klein, C., Koch, T., Boytinck, M., and Trotter, J. (1999). Compartmentation of Fyn kinase with glycosylphosphatidylinositol-anchored molecules in oligodendrocytes facilitates kinase activation during myelination. J Biol Chem 274, 29042–29049. Krauss, K., and Altevogt, P. (1999). Integrin leukocyte function-associated antigen-1-mediated cell binding can be activated by clustering of membrane rafts. J Biol Chem 274, 36921–36927. LaFlamme, S. E., Thomas, L. A., Yamada, S. S., and Yamada, K. M. (1994). Single subunit chimeric integrins as mimics and inhibitors of endogenous integrin functions in receptor localization, cell spreading and migration, and matrix assembly. J Cell Biol 126, 1287–1298. Le Douarin, N., Dulac, C., Dupin, E., and Cameron-Curry, P. (1991). Glial cell lineages in the neural crest. Glia 4, 175–184. Leitinger, B., and Hogg, N. (2002). The involvement of lipid rafts in the regulation of integrin function. J Cell Sci 115, 963–972. Leyton, L., Schneider, P., Labra, C. V., Ruegg, C., Hetz, C. A., Quest, A. F., and Bron, C. (2001). Thy-1 binds to integrin beta(3) on astrocytes and triggers formation of focal contact sites. Curr Biol 11, 1028–38. Liddington, R. C., and Ginsberg, M. H. (2002). Integrin activation takes shape. J Cell Biol 158, 833–839.

INTEGRINS AND MYELIN DISEASES

Louis, J. C., Magal, E., Takayama, S., and Varon, S. (1993). CNTF protection of oligodendrocytes against natural and tumor necrosis factor-induced death. Science 259, 689–692. Mabie, P. C., Mehler, M. F., Marmur, R., Papavasiliou, A., Song, Q., and Kessler, J. A. (1997). Bone morphogenetic proteins induce astroglial diVerentiation of oligodendroglial-astroglial progenitor cells. J Neurosci 17, 4112–4120. Martini, R., and Schachner, M. (1997). Molecular bases of myelin formation as revealed by investigations on mice deWcient in glial cell surface molecules. Glia 19, 298–310. Matsumura, K., Yamada, H., Saito, F., Sunada, Y., and Shimizu, T. (1997). Peripheral nerve involvement in merosin-deWcient congenital muscular dystrophy and dy mouse. Neuromuscul Disord 7, 7–12. McCarty, J. H., Monahan-Earley, R. A., Brown, L. F., Keller, M., Gerhardt, H., Rubin, K., Shani, M., Dvorak, H. F., Wolburg, H., Bader, B. L., Dvorak, A. M., and Hynes, R. O. (2002). Defective associations between blood vessels and brain parenchyma lead to cerebral hemorrhage in mice lacking alphav integrins. Mol Cell Biol 22, 7667–7677. McMorris, F. A., and Dubois-Dalcq, M. (1988). Insulin-like growth factor I promotes cell proliferation and oligodendroglial commitment in rat glial progenitor cells developing in vitro. J Neurosci Res 21, 199–209. Miller, R. H. (1996). Oligodendrocyte origins. Trends Neurosci 19, 92–96. Milner, R., Anderson, H. J., Rippon, R. F., McKay, J. S., Franklin, R. J., Marchionni, M. A., Reynolds, R., and Vrench-Constant, C. (1997a). Contrasting eVects of mitogenic growth factors on oligodendrocyte precursor cell migration. Glia 19, 85–90. Milner, R., Edwards, G., Streuli, C., and Vrench-Constant, C. (1996). A role in migration for the alpha V beta 1 integrin expressed on oligodendrocyte precursors. J Neurosci 16, 7240–7252. Milner, R., and Vrench-Constant, C. (1994). A developmental analysis of oligodendroglial integrins in primary cells: Changes in alpha v-associated beta subunits during diVerentiation. Development 120, 3497–506. Milner, R., Frost, E., Nishimura, S., Delcommenne, M., Streuli, C., Pytela, R., and Vrench-Constant, C. (1997b). Expression of alpha vbeta3 and alpha vbeta8 integrins during oligodendrocyte precursor diVerentiation in the presence and absence of axons. Glia 21, 350–360. Milner, R., Huang, X., Wu, J., Nishimura, S., Pytela, R., Sheppard, D., and Vrench-Constant, C. (1999). Distinct roles for astrocyte alphavbeta5 and alphavbeta8 integrins in adhesion and migration. J Cell Sci 112, 4271–4279. Milner, R., Relvas, J. B., Fawcett, J., and Vrench-Constant, C. (2001). Developmental regulation of alphav integrins produces functional changes in astrocyte behavior. Mol Cell Neurosci 18, 108–118. Milner, R., Wilby, M., Nishimura, S., Boylen, K., Edwards, G., Fawcett, J., Streuli, C., Pytela, R., and VrenchConstant, C. (1997c). Division of labor of Schwann cell integrins during migration on peripheral nerve extracellular matrix ligands. Dev Biol 185, 215–228. Mirsky, R., and Jessen, K. R. (1996). Schwann cell development, diVerentiation and myelination. Curr Opin Neurobiol 6, 89–96. Miyagoe-Suzuki, Y., Nakagawa, M., and Takeda, S. (2000). Merosin and congenital muscular dystrophy. Microsc Res Tech 48, 181–91. Miyamoto, S., Akiyama, S. K., and Yamada, K. M. (1995). Synergistic roles for receptor occupancy and aggregation in integrin transmembrane function. Science 267, 883–5. Miyamoto, S., Teramoto, H., Coso, O. A., Gutkind, J. S., Burbelo, P. D., Akiyama, S. K., and Yamada, K. M. (1995). Integrin function: Molecular hierarchies of cytoskeletal and signaling molecules. J Cell Biol 131, 791–805. Miyamoto, S., Teramoto, H., Gutkind, J. S., and Yamada, K. M. (1996). Integrins can collaborate with growth factors for phosphorylation of receptor tyrosine kinases and MAP kinase activation: Roles of integrin aggregation and occupancy of receptors. J Cell Biol 135, 1633–1642. Montgomery, A. M., Becker, J. C., Siu, C. H., Lemmon, V. P., Cheresh, D. A., Pancook, J. D., Zhao, X., and Reisfeld, R. A. (1996). Human neural cell adhesion molecule L1 and rat homologue NILE are ligands for integrin alpha v beta 3. J Cell Biol 132, 475–485. Moro, L., Venturino, M., Bozzo, C., Silengo, L., Altruda, F., Beguinot, L., Tarone, G., and DeWlippi, P. (1998). Integrins induce activation of EGF receptor: Role in MAP kinase induction and adhesion-dependent cell survival. Embo J 17, 6622–6632. Morris, R. (1985). Thy-1 in developing nervous tissue. Dev Neurosci 7, 133–160. Moya, F., Bunge, M. B., and Bunge, R. P. (1980). Schwann cells proliferate but fail to diVerentiate in deWned medium. Proc Natl Acad Sci U S A 77, 6902–6906. Neusch, C., Rozengurt, N., Jacobs, R. E., Lester, H. A., and Kofuji, P. (2001). Kir4.1 potassium channel subunit is crucial for oligodendrocyte development and in vivo myelination. J Neurosci 21, 5429–5438. Niessen, C. M., Cremona, O., Daams, H., Ferraresi, S., Sonnenberg, A., and Marchisio, P. C. (1994). Expression of the integrin alpha 6 beta 4 in peripheral nerves: Localization in Schwann and perineural cells and diVerent variants of the beta 4 subunit. J Cell Sci 107, 543–552. Noble, M., Murray, K., Stroobant, P., WaterWeld, M. D., and Riddle, P. (1988). Platelet-derived growth factor promotes division and motility and inhibits premature diVerentiation of the oligodendrocyte/type-2 astrocyte progenitor cell. Nature 333, 560–562. Oh, L. Y., Larsen, P. H., Krekoski, C. A., Edwards, D. R., Donovan, F., Werb, Z., and Yong, V. W. (1999). Matrix metalloproteinase-9/gelatinase B is required for process outgrowth by oligodendrocytes. J Neurosci 19, 8464–8475.

629

630

25. INTEGRINS

Ono, K., Yasui, Y., Rutishauser, U., and Miller, R. H. (1997). Focal ventricular origin and migration of oligodendrocyte precursors into the chick optic nerve. Neuron 19, 283–292. Osterhout, D. J., Ebner, S., Xu, J., Ornitz, D. M., Zazanis, G. A., and McKinnon, R. D. (1997). Transplanted oligodendrocyte progenitor cells expressing a dominant-negative FGF receptor transgene fail to migrate in vivo. J Neurosci 17, 9122–9132. Pampori, N., Hato, T., Stupack, D. G., Aidoudi, S., Cheresh, D. A., Nemerow, G. R., and Shattil, S. J. (1999). Mechanisms and consequences of aYnity modulation of integrin alpha(V)beta(3) detected with a novel patchengineered monovalent ligand. J Biol Chem 274, 21609–21616. Payne, H. R., Hemperly, J. J., and Lemmon, V. (1996). N-cadherin expression and function in cultured oligodendrocytes. Brain Res Dev Brain Res 97, 9–15. Pedraza, L., Huang, J. K., and Colman, D. R. (2001). Organizing principles of the axoglial apparatus. Neuron 30, 335–344. Perris, R. (1997). The extracellular matrix in neural crest-cell migration. Trends Neurosci 20, 23–31. Perry, V. H., and Lund, R. D. (1990). Evidence that the lamina cribrosa prevents intraretinal myelination of retinal ganglion cell axons. J Neurocytol 19, 265–272. Persohn, E., and Schachner, M. (1987). Immunoelectron microscopic localization of the neural cell adhesion molecules L1 and N-CAM during postnatal development of the mouse cerebellum. J Cell Biol 105, 569–76. Pesheva, P., Gloor, S., Schachner, M., and Probstmeier, R. (1997). Tenascin-R is an intrinsic autocrine factor for oligodendrocyte diVerentiation and promotes cell adhesion by a sulfatide-mediated mechanism. J Neurosci 17, 4642–51. PinkstaV, J. K., Detterich, J., Lynch, G., and Gall, C. (1999). Integrin subunit gene expression is regionally diVerentiated in adult brain. J Neurosci 19, 1541–1556. Podratz, J. L., Rodriguez, E., and Windebank, A. J. (2001). Role of the extracellular matrix in myelination of peripheral nerve. Glia 35, 35–40. Powell, S. K., Williams, C. C., Nomizu, M., Yamada, Y., and Kleinman, H. K. (1998). Laminin-like proteins are diVerentially regulated during cerebellar development and stimulate granule cell neurite outgrowth in vitro. J Neurosci Res 54, 233–247. Previtali, S. C., Feltri, M. L., Archelos, J. J., Quattrini, A., Wrabetz, L., and Hartung, H. (2001). Role of integrins in the peripheral nervous system. Prog Neurobiol 64, 35–49. Reichardt, L. F., and Tomaselli, K. J. (1991). Extracellular matrix molecules and their receptors: Functions in neural development. Annu Rev Neurosci 14, 531–570. Relvas, J. B., Setzu, A., Baron, W., Buttery, P. C., LaFlamme, S. E., Franklin, R. J., and Vrench-Constant, C. (2001). Expression of dominant-negative and chimeric subunits reveals an essential role for beta1 integrin during myelination. Curr Biol 11, 1039–1043. Richardson, W. D., Pringle, N., Mosley, M. J., Westermark, B., and Dubois-Dalcq, M. (1988). A role for plateletderived growth factor in normal gliogenesis in the central nervous system. Cell 53, 309–319. Richardson, W. D., Pringle, N. P., Yu, W. P., and Hall, A. C. (1997). Origins of spinal cord oligodendrocytes: Possible developmental and evolutionary relationships with motor neurons. Dev Neurosci 19, 58–68. Rohrbough, J., Grotewiel, M. S., Davis, R. L., and Broadie, K. (2000). Integrin-mediated regulation of synaptic morphology, transmission, and plasticity. J Neurosci 20, 6868–6878. Rowitch, D. H., Lu, Q. R., Kessaris, N., and Richardson, W. D. (2002). An ‘oligarchy’ rules neural development. Trends Neurosci 25, 417–422. Ruoslahti, E. (1996). RGD and other recognition sequences for integrins. Annu Rev Cell Dev Biol 12, 697–715. SchiVmann, R., and BoespXug-Tanguy, O. (2001). An update on the leukodsytrophies. Curr Opin Neurol 14, 789–794. Schirmer, E. C., Farooqui, J., Polak, P. E., and Szuchet, S. (1994). GRASP: A novel heparin-binding serum glycoprotein that mediates oligodendrocyte-substratum adhesion. J Neurosci Res 39, 457–473. Schnadelbach, O., Blaschuk, O. W., Symonds, M., Gour, B. J., Doherty, P., and Fawcett, J. W. (2000). N-cadherin inXuences migration of oligodendrocytes on astrocyte monolayers. Mol Cell Neurosci 15, 288–302. Schnadelbach, O., Ozen, I., Blaschuk, O. W., Meyer, R. L., and Fawcett, J. W. (2001). N-cadherin is involved in axon-oligodendrocyte contact and myelination. Mol Cell Neurosci 17, 1084–1093. Schneller, M., Vuori, K., and Ruoslahti, E. (1997). Alphavbeta3 integrin associates with activated insulin and PDGFbeta receptors and potentiates the biological activity of PDGF. Embo J 16, 5600–5607. Schwartz, M. A., and Baron, V. (1999). Interactions between mitogenic stimuli, or, a thousand and one connections. Curr Opin Cell Biol 11, 197–202. Scolding, N., Franklin, R., Stevens, S., Heldin, C. H., Compston, A., and Newcombe, J. (1998). Oligodendrocyte progenitors are present in the normal adult human CNS and in the lesions of multiple sclerosis. Brain 121, 2221–2228. Shaw, C. E., Milner, R., Compston, A. S., and Vrench-Constant, C. (1996). Analysis of integrin expression on oligodendrocytes during axo-glial interaction by using rat-mouse xenocultures. J Neurosci 16, 1163–1172. Simons, K., and Toomre, D. (2000). Lipid rafts and signal transduction. Nat Rev Mol Cell Biol 1, 31–39. Simons, M., Kramer, E. M., Thiele, C., StoVel, W., and Trotter, J. (2000). Assembly of myelin by association of proteolipid protein with cholesterol-and galactosylceramide-rich membrane domains. J Cell Biol 151, 143–154. Sobel, R. A., and Ahmed, A. S. (2001). White matter extracellular matrix chondroitin sulfate/dermatan sulfate proteoglycans in multiple sclerosis. J Neuropathol Exp Neurol 60, 1198–1207. Sonnenberg, A., Linders, C. J., Daams, J. H., and Kennel, S. J. (1990). The alpha 6 beta 1 (VLA-6) and alpha 6 beta 4 protein complexes: Tissue distribution and biochemical properties. J Cell Sci 96, 207–217.

INTEGRINS AND MYELIN DISEASES

Spassky, N., de Castro, F., Le Bras, B., Heydon, K., Queraud-LeSaux, F., Bloch-Gallego, E., Chedotal, A., Zalc, B., and Thomas, J. L. (2002). Directional guidance of oligodendroglial migration by class 3 semaphorins and netrin-1. J Neurosci 22, 5992–6004. Stewart, H. J., Turner, D., Jessen, K. R., and Mirsky, R. (1997). Expression and regulation of alpha1beta1 integrin in Schwann cells. J Neurobiol 33, 914–928. Sugimoto, Y., Taniguchi, M., Yagi, T., Akagi, Y., Nojyo, Y., and Tamamaki, N. (2001). Guidance of glial precursor cell migration by secreted cues in the developing optic nerve. Development 128, 3321–3330. Tait, S., Gunn-Moore, F., Collinson, J. M., Huang, J., Lubetzki, C., Pedraza, L., Sherman, D. L., Colman, D. R., and Brophy, P. J. (2000). An oligodendrocyte cell adhesion molecule at the site of assembly of the paranodal axo-glial junction. J Cell Biol 150, 657–666. Takagi, J., Petre, B. M., Walz, T., and Springer, T. A. (2002). Global conformational rearrangements in integrin extracellular domains in outside-in and inside-out signaling. Cell 110, 599–611. Tawil, N. J., Wilson, P., and Carbonetto, S. (1994). Expression and distribution of functional integrins in rat CNS glia. J Neurosci Res 39, 436–447. Terada, N., Baracskay, K., Kinter, M., Melrose, S., Brophy, P. J., Boucheix, C., Bjartmar, C., Kidd, G., and Trapp, B. D. (2002). The tetraspanin protein, CD9, is expressed by progenitor cells committed to oligodendrogenesis and is linked to beta1 integrin, CD81, and Tspan-2. Glia 40, 350–359. Testaz, S., Delannet, M., and Duband, J. (1999). Adhesion and migration of avian neural crest cells on Wbronectin require the cooperating activities of multiple integrins of the (beta)1 and (beta)3 families. J Cell Sci 112, 4715–4728. Testaz, S., and Duband, J. L. (2001). Central role of the alpha4beta1 integrin in the coordination of avian truncal neural crest cell adhesion, migration, and survival. Dev Dyn 222, 127–140. Thomas, J. L., Spassky, N., Perez Villegas, E. M., Olivier, C., Cobos, I., Goujet-Zalc, C., Martinez, S., and Zalc, B. (2000). Spatiotemporal development of oligodendrocytes in the embryonic brain. J Neurosci Res 59, 471–476. Thorne, R. F., Marshall, J. F., Shafren, D. R., Gibson, P. G., Hart, I. R., and Burns, G. F. (2000). The integrins alpha3beta1 and alpha6beta1 physically and functionally associate with CD36 in human melanoma cells. Requirement for the extracellular domain OF CD36. J Biol Chem 275, 35264–35275. Tiwari-WoodruV, S. K., Buznikov, A. G., Vu, T. Q., Micevych, P. E., Chen, K., Kornblum, H. I., and Bronstein, J. M. (2001). OSP/claudin-11 forms a complex with a novel member of the tetraspanin super family and beta1 integrin and regulates proliferation and migration of oligodendrocytes. J Cell Biol 153, 295–305. Trapp, B. D., Nishiyama, A., Cheng, D., and Macklin, W. (1997). DiVerentiation and death of premyelinating oligodendrocytes in developing rodent brain. J Cell Biol 137, 459–468. Tsai, H. H., Frost, E., To, V., Robinson, S., Vrench-Constant, C., Geertman, R., RansohoV, R. M., and Miller, R. H. (2002). The chemokine receptor CXCR2 controls positioning of oligodendrocyte precursors in developing spinal cord by arresting their migration. Cell 110, 373–383. Tsai, H. H., and Miller, R. H. (2002). Glial cell migration directed by axon guidance cues. Trends Neurosci 25, 173–5; discussion 175–176. van Heyningen, P., Calver, A. R., and Richardson, W. D. (2001). Control of progenitor cell number by mitogen supply and demand. Curr Biol 11, 232–241. Vemuri, G. S., and McMorris, F. A. (1996). Oligodendrocytes and their precursors require phosphatidylinositol 3kinase signaling for survival. Development 122, 2529–2537. Vinogradova, O., Velyvis, A., Velyviene, A., Hu, B., Haas, T., Plow, E., and Qin, J. (2002). A structural mechanism of integrin alpha(IIb)beta(3) ‘‘inside-out’’ activation as regulated by its cytoplasmic face. Cell 110, 587–597. Vogelezang, M. G., Liu, Z., Relvas, J. B., Raivich, G., Scherer, S. S., and Vrench-Constant, C. (2001). Alpha4 integrin is expressed during peripheral nerve regeneration and enhances neurite outgrowth. J Neurosci 21, 6732–4674. Wang, C., Pralong, W. F., Schulz, M. F., Rougon, G., Aubry, J. M., Pagliusi, S., Robert, A., and Kiss, J. Z. (1996). Functional N-methyl-D-aspartate receptors in O-2A glial precursor cells: A critical role in regulating polysialic acid-neural cell adhesion molecule expression and cell migration. J Cell Biol 135, 1565–1581. Wang, S., Sdrulla, A. D., diSibio, G., Bush, G., Nofziger, D., Hicks, C., Weinmaster, G., and Barres, B. A. (1998). Notch receptor activation inhibits oligodendrocyte diVerentiation. Neuron 21, 63–75. Warf, B. C., Fok-Seang, J., and Miller, R. H. (1991). Evidence for the ventral origin of oligodendrocyte precursors in the rat spinal cord. J Neurosci 11, 2477–2488. Wary, K. K., Mainiero, F., IsakoV, S. J., Marcantonio, E. E., and Giancotti, F. G. (1996). The adaptor protein Shc couples a class of integrins to the control of cell cycle progression. Cell 87, 733–743. Werner, A., Willem, M., Jones, L. L., Kreutzberg, G. W., Mayer, U., and Raivich, G. (2000). Impaired axonal regeneration in alpha7 integrin-deWcient mice. J Neurosci 20, 1822–1830. Wolfsberg, T. G., PrimakoV, P., Myles, D. G., and White, J. M. (1995). ADAM, a novel family of membrane proteins containing A Disintegrin And Metalloprotease domain: Multipotential functions in cell-cell and cellmatrix interactions. J Cell Biol 131, 275–278. Wolswijk, G. (1998). Chronic stage multiple sclerosis lesions contain a relatively quiescent population of oligodendrocyte precursor cells. J Neurosci 18, 601–609. Wu, Q., Miller, R. H., RansohoV, R. M., Robinson, S., Bu, J., and Nishiyama, A. (2000). Elevated levels of the chemokine GRO-1 correlate with elevated oligodendrocyte progenitor proliferation in the jimpy mutant. J Neurosci 20, 2609–2617.

631

632

25. INTEGRINS

Xue, G. P., Rivero, B. P., and Morris, R. J. (1991). The surface glycoprotein Thy-1 is excluded from growing axons during development: A study of the expression of Thy-1 during axogenesis in hippocampus and hindbrain. Development 112, 161–176. Ye, P., Carson, J., and D’Ercole, A. J. (1995). In vivo actions of insulin-like growth factor-I (IGF-I) on brain myelination: Studies of IGF-I and IGF binding protein-1 (IGFBP-1) transgenic mice. J Neurosci 15, 7344–7356. Yip, P. M., Zhao, X., Montgomery, A. M., and Siu, C. H. (1998). The Arg-Gly-Asp motif in the cell adhesion molecule L1 promotes neurite outgrowth via interaction with the alphavbeta3 integrin. Mol Biol Cell 9, 277–290.

C H A P T E R

26 The Periaxin Gene Diane L. Sherman and Peter J. Brophy

INTRODUCTION In recent years, much has been learned about the embryological origin of Schwann cells and their developmental program (Mirsky and Jessen, 1996). There is also a growing knowledge of the trophic polypeptides and transcription factors that regulate cell number and fate in the developing PNS (Jessen and Mirsky, 1999; Mirsky et al., 2001). In contrast to these new insights into the development and speciWcation of Schwann cells, it is much less clear how internodal length and the Wnal stages of myelination, including the determination of sheath thickness, are regulated; nor have many clues emerged about the factors that stabilize the mature Schwann cell–axon unit. The growing realization that myelin-forming glia in general, and Schwann cells in particular, play a critical role in promoting neuron survival has added urgency to understanding how the Schwann cell–axon unit is stabilized in the mature nerve (Riethmacher et al., 1997; Trapp et al., 1998). Hence, proteins that play a critical role in these later stages of nerve maturation are of great interest. Mice that lack a functional Periaxin (Prx) gene produce an apparently normal myelin sheath, which subsequently becomes unstable in the mature animal. This has lent credence to the view that by studying this gene and its protein products, we may gain some insight into the factors that determine the stable association between glia and neurons and those that undermine it in demyelinating disease (Gillespie et al., 2000).

THE PRX GENE AND THE PERIAXINS Gene Structure L-periaxin is one of two proteins encoded by the Prx gene and was Wrst identiWed in a screen for cytoskeleton-associated proteins in Schwann cells (Gillespie et al., 2000). The murine Prx gene is primarily expressed in myelinating Schwann cells (Gillespie et al., 1994; Scherer et al., 1995). It has seven exons and encodes two mRNAs of 4.6 and 5.2 kb, respectively, the Wrst of which comprises exons 1 to 7, whereas the second includes a retained intron between exons 6 and 7 (Fig. 26.1). The smaller mRNA encodes L-periaxin (147 kD), whereas the larger mRNA encodes S-periaxin (16 kD), a truncated protein, in which the unique C-terminus of the protein and a stop codon are supplied by the retained intron (Dytrych et al., 1998). The inclusion of the intron is probably favored by the presence of suboptimal Xanking 5’- and 3’-splice sites together with a downstream putative exonic splicing enhancer whose main distinguishing feature is a high purine content (Dirksen, Sun, and Rottman, 1995).

Myelin Biology and Disorders, Volume 1

633

Copyright 2004, Elsevier (USA). All rights reserved.

634

26. THE PERIAXIN GENE

TAA

PDZ

5.2 kb

S-periaxin 16 kD TAA

Basic domain

4.6 kb L-periaxin 147 kD

FIGURE 26.1 Structure of the murine Prx gene and the proteins it encodes. Two mRNAs are produced by alternative splicing. The larger mRNA encodes S-periaxin owing to the retention of an intron that includes a stop codon; note that in addition to truncating the protein, it provides S-periaxin with a unique C-terminus. Exon 7 is translated in the smaller mRNA and encodes the rest of the basic domain (white), the repeat region (yellow), and an acidic domain (dark gray).

The human PRX gene, like the murine gene, comprises seven exons, and exon VII is also by far the largest (Boerkoel et al., 2001; Guilbot et al., 2001). As with the murine gene, the translational start site is located in exon IV. The human gene encodes a 1461 amino acid L-periaxin, 70 amino acids longer than the mouse protein, and most of this extra sequence can be accounted for within the central repeat region of the protein (discussed later). The human and murine L-periaxins share 78% amino acid identity. In human peripheral nerve, it has been reported that the 5.1-kb PRX mRNA is more abundant than the 5.6-kb species, suggesting that L-periaxin is the dominant isoform in human PNS myelin (Boerkoel et al., 2001).

Structure of L- and S-periaxin L- and S-periaxin are relatively abundant proteins in myelinating Schwann cells. Perhaps the most remarkable feature of the primary sequence of both proteins is the presence of a PDZ domain at the N-terminus between residues 13 and 100 (Fig. 26.1). This proteininteracting domain was originally named after the three proteins in which it was Wrst identiWed, namely post-synaptic density protein PSD-95, Drosophila discs large (dlg) tumor suppressor gene and the tight junction-associated protein ZO-1. PDZ domains are believed to share a common three-dimensional structure that interacts with a sequence found at the carboxy-terminus of certain plasma membrane proteins. Binding occurs through a b-sheet-antiparallel interaction where the carboxy-terminal amino acid of the transmembrane protein Wts into a hydrophobic pocket in the folded PDZ module at the carboxylate binding loop. Although the complementary binding site for some PDZ domains is the simple peptide sequence (S/T)XV, others recognize somewhat diVerent sequences and can even bind to other PDZ domains (Hillier et al., 1999). It appears that the PDZ domain of L-periaxin falls into this latter category in that it dimerizes to form L-periaxin homodimers. Although both L- and S-periaxin contain the N-terminal PDZ domain (Fig. 26.1), only L-periaxin is localized to the Schwann cell plasma membrane. In contrast, S-periaxin, although approximately equal in abundance to L-periaxin in the mouse, is distributed uniformly throughout the cytoplasm of the cell (Dytrych et al., 1998). This suggests that L-periaxin might participate in a cortical signaling complex from which S-periaxin is excluded. Overall, murine L-periaxin is slightly basic (pI, 8.6), but this obscures the presence of a highly basic domain near the N-terminus between amino acids 118 and 196 and an acidic region encompassed by residues 1036 to 1163 toward the C-terminus (Fig. 26.1). The central domain of the protein is characterized by the presence of a series of repeats

III. THE MYELIN GENES AND PRODUCTS

EMBRYONIC EXPRESSION AND LOCALIZATION OF L-PERIAXIN

throughout a rather long region comprising amino acids 430 to 730 (Fig. 26.1). The core repeat is a pentamer in which the variable amino acid is always either alanine or glutamine or one of the basic amino acids as follows: aliphatic nonpolar: pro, glu, or asp; aliphatic nonpolar: variable Higher-order repeats are also apparent, and this region appears to be less Xexible than the rest of the polypeptide so that it may extend the protein, thus keeping the PDZ and basic domains well separated from the acidic region in the C-terminus.

EMBRYONIC EXPRESSION AND LOCALIZATION OF L-PERIAXIN Like the major integral membrane protein of PNS myelin, P0, L-periaxin is detectable in Schwann cells at an early stage of embryonic development (Sherman and Brophy, 2000). In contrast to P0, which is expressed in neural crest cells, the Prx gene is not transcribed in the Schwann cell lineage until these cells diVerentiate from their precursors in the mouse embryo, at around E14. At this time, L-periaxin is localized to the nuclei of Schwann cells where it remains for an additional 3 days of Schwann cell development (Sherman and Brophy, 2000). A nuclear localization signal (NLS) of an unusual tripartite type in the N-terminal basic domain (see the earlier section titled ‘‘Structure of L- and S-periaxin’’) appears to be responsible for translocating L-periaxin into the nucleus of Schwann cells (Sherman and Brophy, 2000). The classical NLS is typiWed by that of the SV40 large T antigen (PKKKRKV) or the bipartite NLS of nucleoplasmin (KRPAAIKKAGQAKKKK). These signals interact with the importin/karyopherin complex, which is subsequently translocated into the nucleus in an energy-dependent fashion by a mechanism that requires, among other things, GTP and the GTPase, Ran. Recent data suggest that the nucleocytoplasmic distribution of several proteins that undergo active nuclear uptake is aVected by cell-cell contact. This also appears to be the case for L-periaxin, at least in transfected cells. Although it is not yet clear what the nuclear function of L-periaxin might be, it joins a growing list of proteins that redistribute between the nucleus and cortical signaling/adhesion complexes. The best characterized example of these is the Wnt-signaling pathway in which b-catenin shuttles between adherens junctions and the nucleus by means of a transcription factor complex (Gumbiner, 1998). Proteins of L-periaxin’s size (147 kD) would be expected to require active transport through the nuclear pore complex rather than diVuse passively into the nucleus. Given the requirement for energy and the eVect of cell-cell contact, it seems likely that nuclear translocation of L-periaxin does take place by the classical active route even though it does not utilize a classical NLS. Thus far, little is known of how a tripartite signal might be recognized by the import machinery. However, it is known that more than one copy of a basic NLS can increase the eYciency of nuclear uptake. Does the developmentally regulated nucleocytoplasmic redistribution of L-periaxin in embryonic Schwann cells have a function? It is possible that nuclear targeting of L-periaxin in embryonic Schwann cells may sequester the PDZ domain from inappropriate interactions in the cytoplasm until the correct ligand becomes available at the cell cortex of the maturing myelin-forming Schwann cell. The appearance of an appropriate binding partner at the cell surface of the Schwann cell at the early stages of myelination may then act as a stimulus for the translocation of L-periaxin from the nucleus to myelinating processes as they ensheath the axon. Such a protein partner would not be predicted to be expressed in the myelin-forming glia of the CNS, since no such redistribution from the nucleus occurs when L-periaxin is ectopically expressed in the myelinating oligodendrocytes of transgenic mice (Sherman and Brophy, 2000). A binding partner for L-periaxin with these attributes has been identiWed (see the following section). Alternatively, and by analogy with the Wntsignaling pathway and b-catenin, L-periaxin may have a cofactor role in regulating embryonic gene transcription in the Schwann cell. Once myelination begins in mouse sciatic nerve around E19, L-periaxin disappears from the nucleus and concentrates initially at the adaxonal membrane (apposing the axon) and

III. THE MYELIN GENES AND PRODUCTS

635

636

26. THE PERIAXIN GENE

Basal lamina α β

Laminin

βδ ε

SG

Utrophin

α α β dystroglycan β βδε

βδ ε

DRP2

DRP2

F-actin

α β βδ ε

Dp116

L-periaxin

FIGURE 26.2 Structure of the L-periaxin-DRP2-dystroglycan (PDG) complex in the Schwann cell plasma membrane. L-periaxin dimerizes at its N-terminal PDZ domain, which clusters the PDG complex. The other two dystroglycan complexes utilize utrophin and a truncated form of dystrophin (Dp116). These probably interact with the PDG complex. SG: sarcoglycans.

at the abaxonal membrane (apposing the basal lamina) (Gillespie et al., 1994). As myelin sheaths mature, L-periaxin becomes concentrated in the abaxonal plasma membrane (Scherer et al., 1995). This shift in localization led to the suggestion that L-periaxin might participate in the membrane-protein interactions required to stabilize the sheath.

COMPOSITION AND FUNCTION OF THE L-PERIAXIN COMPLEX Composition of the L-periaxin Complex One way of gaining clues about the function of a protein is to identify the proteins with which it interacts and this approach is likely to be particularly informative for members of macromolecular signaling complexes. L-periaxin was suspected to be such a protein since, although it is neither a transmembrane protein nor a constituent of compact myelin, the protein is closely associated with the Schwann cell plasma membrane in the mature cell (Gillespie et al., 1994; Scherer et al., 1995). Yeast two-hybrid screens of rat peripheral nerve cDNA libraries using the N-terminus of L-periaxin as a bait identiWed dystrophin-related protein 2 (DRP2) as an interacting protein. DRP2, as its name suggests, is a member of a family of dystrophin-like proteins, the other members of which, dystrophin, utrophin (dystrophin-related protein 1), and dystrobrevin, are all components of complexes with the plasma membrane protein dystroglycan (Fig. 26.2). The dystroglycan complexes (DGCs) have been well characterized in skeletal muscle where they link extracellular proteins, such as the laminins of the basal lamina, to dystrophin and the cortical actin cytoskeleton by virtue of interactions with a- and bdystroglycan, respectively. Mutations in components of the DGC disrupt the complex and cause several types of muscular dystrophy, possibly due to a lack of mechanical buVering against the shearing forces of contraction. It is also likely that such mutations disrupt the signaling function of DGCs, a role that is increasingly recognized due to the identiWcation of the syntrophins as PDZ-domain-containing proteins that interact with dystrophin (Kachinsky, Froehner, and Milgram, 1999). DGCs are also widely distributed in the vertebrate nervous system (Love et al., 1989; Matsumura et al., 1993). In the peripheral nervous system (PNS), Schwann cells express dystroglycan, utrophin, and a truncated form of dystrophin, Dp116 (Fig. 26.2). DGCs in the PNS have been implicated in mediating the responsiveness of Schwann cells to extracellular signals from laminin-2 since contact with the basal lamina modulates the expression of genes required for myelination (Bunge et al., 1990) and mutations in the

III. THE MYELIN GENES AND PRODUCTS

COMPOSITION AND FUNCTION OF THE L-PERIAXIN COMPLEX

FIGURE 26.3 Ramon y Cajal’s silver stain revealed the domains enriched in the L-periaxin-DRP2 complex. A teased sciatic nerve stained with silver nitrate by Santiago Ramon y Cajal reveals the unstained patches of plasma membrane (arrowhead) embedded in a longitudinal and transverse meshwork of protoplasmic trabeculae (reproduced from ‘‘Histology’’ by S. Ramon y Cajal (1933) with permission, Elsevier Press). These patches or spheroids are also revealed by immunoXuorescence staining with antibodies to DRP2 (green) and S100 (red), which delineates the Schwann cell cytoplasm (arrowhead). Note that the plasma membrane of the Schwann cell soma is deWcient in such complexes; the position of the nucleus is shown with an arrow.

laminin-2 gene, LAMA2, cause peripheral demyelination as well as muscular dystrophy (Xu et al., 1994). Although DRP2 was Wrst identiWed as a transcript in the CNS (Roberts et al., 1996), the protein appears to be particularly abundant in the plasma membrane of myelinating Schwann cells (Sherman et al., 2001). Immunocytochemistry reveals that the L-periaxinDRP2-dystroglycan complex is clustered in distinctive patches at the Schwann cell plasma membrane (Sherman et al., 2001), and it seems likely that this clustering is promoted or stablilized by dimerization of L-periaxin at its PDZ domain (Gillespie et al., 2000). In contrast, the Dp116 and utrophin DGCs appear to be evenly distributed around the circumference of the Schwann cell plasma membrane. In spite of these distinct localizations, when biochemically isolated by immunoaYnity chromatography, the L-periaxinDRP2-dystroglycan complex also contains some Dp116 and utrophin, indicating that a subset of their DGC complexes coclusters with L-periaxin and DRP2 (Sherman et al., 2001). The distribution of the spheroidal clusters in the plasma membrane of the Schwann cell does not appear to be random. Figure 26.3 shows that they tend to localize to regions of the plasma membrane that are closest to the abaxonal (i.e., furthest away from the axon) surface of the myelin sheath. These plasma membrane domains would be expected to overlie regions of the Schwann cell less rich in cytoplasm than other areas, a point alluded to previously (Arroyo et al., 2001). As for many other aspects of nervous system structure, these domains were Wrst described by the great Spanish neuroanatomist Santiago Ramon y Cajal in the late 19th century (Ramon-Cajal, 1933). Remarkably, the L-periaxin-DRP2dystroglycan complex is clustered in the Schwann cell plasma membrane in a pattern that accurately delineates the unstained domains embedded in a meshwork of protoplasmic trabecula, which Ramon y Cajal described (Fig. 26.4). The discovery of the L-periaxinDRP2-dystroglycan complex is the Wrst insight into the composition and function of these plasma membrane domains. Whether or not these regions of the Schwann cell represent specialized signaling domains remains to be seen.

The L-periaxin-DRP2 Complex Is Essential for Stable Axon-Glia Interaction The presence of three diVerent DGCs in the Schwann cell plasma membrane, each utilizing a diVerent member of the dystrophin family, suggests that these complexes may have distinct functions. To explore this issue further Prx / , mice have been generated by gene targeting in ES cells (Gillespie et al., 2000). In periaxin-null mice, the DRP2-DGC complex is mislocalized and DRP2 is unstable, which supports the view that the ability of L-periaxin to homodimerize at its PDZ domain probably accounts for the clustering of the complex (Sherman et al., 2001). What eVect

III. THE MYELIN GENES AND PRODUCTS

637

638

26. THE PERIAXIN GENE

FIGURE 26.4 The L-periaxin-DRP2 domains are concentrated where the abaxonal surface of the myelin sheath is closest to the plasma membrane of the Schwann cell. The micrograph shows the superimposition of a cross-section of a trochlear nerve immunostained for DRP2 (green) on a transmitted light image, and the diagram demonstrates that the patchy staining of the Schwann cell plasma membrane corresponds to regions that are close to the myelin sheath.

does this have on myelination? Prx / mice are able to assemble compact myelin, but by 6 weeks of age the sciatic nerves of periaxin-null mice contain some focal thickenings (tomacula) and infoldings of internodal myelin (Gillespie et al., 2000). The levels of the major myelin proteins, myelin-associated glycoprotein (MAG), P0, and myelin basic protein (MBP) are normal in the sciatic nerves from mutant animals, which conWrms that at this age there has not been extensive demyelination. A variety of spinal, cranial, and autonomic nerves also display some limited abnormalities at this age. However, by 6 months, sensory, motor, and autonomic (vagus) nerves are extensively demyelinated, and most internodes in the sciatic nerves of periaxin-null mice contain focal thickenings or infoldings of the sheath (Fig. 26.5). Profound disruption of axonal ensheathment and segmental demyelination are apparent in teased sciatic nerve Wbers at this age together with clear evidence of supernumerary Schwann cells (Gillespie et al., 2000) (Fig. 26.5). Schmidt-Lanterman incisures, which are normally visible as cytoplasm-Wlled structures along the length of the internodes, are also deranged. Electron microscopy demonstrates that saphenous nerves, which are predominantly sensory, are extensively hypermyelinated but that unmyelinated C-Wber bundles appear to be morphologically normal, which is consistent with the absence of the periaxins in nonmyelin-forming Schwann cells (Scherer et al., 1995). Hypermyelination results in sheath infolding and axonal compression, and naked or thinly myelinated axons are common in sciatic nerve Wbers by 8 months, often surrounded by redundant basal lamina and supernumerary Schwann cells, which form onion bulb structures, diagnostic of attempts to remyelinate demyelinated Wbers (Gillespie et al., 2000) (Fig. 26.4). There is no evidence that demyelination is the result of a macrophage-mediated inXammatory process, and the damage seems to be conWned to the myelin sheath at 6 weeks in that there is no evidence of axonal degeneration. Furthermore, neuronal cell bodies in spinal gray matter and spinal ganglia show no evidence of degenerative changes at this age. However, in older animals extensive demyelination is presumably responsible for the concomitant loss of both sensory neurons and axons.

Reduced Nerve Conduction Velocity and Neuropathic Pain in the Prx

/

Mouse /

At 6 to 8 months, the extensive pathology noted in peripheral nerves from Prx mice is reXected in reduced conduction velocities (Gillespie et al., 2000). Remarkably, even at 6 weeks after birth, when there is very little demyelination and no loss of sensory neurons, Prx / mice exhibit marked diVerences from wild-type animals in reXex behavioral tests which are reliable indicators of sensory abnormalities such as mechanical allodynia and

III. THE MYELIN GENES AND PRODUCTS

au1

639

COMPOSITION AND FUNCTION OF THE L-PERIAXIN COMPLEX

FIGURE 26.5 Demyelination in the sciatic nerves of periaxin-null mice. Teased sciatic nerve Wbers from wild-type and periaxinnull (Prx / ) mice were immunostained for myelin P0 (green), axonal neuroWlaments (red), and nuclei (blue). Note the segmental demyelination, hypermyelination (tomacula), and supernumerary cells (primarily Schwann cells) in the mutant Wbers.

Filament Indentation Pressure (mN/mm2) Ascending tracts to brain for perception

+/+

460

40

-/-

274

23

DRG Aβ

Sensory nerve

Motor nerve

Von Frey filament force

FIGURE 26.6 Mechanical allodynia in Prx / mice. In this reXex behavioral test, graded Von Frey Wlaments are brushed against the paws and the applied pressure threshold at which paw-withdrawal occurrences are measured. The threshold is markedly reduced in mutant mice, indicating the development of pain behavior as a result of demyelination. This diagram is reproduced with permission of Susan Fleetwood-Walker.

thermal hyperalgesia (Gillespie et al., 2000). Mechanical allodynia is the perception of a normally innocuous mechanical stimulus, such as gentle stroking of the skin with a Wne hair, as painful and the way in which this is conveniently measured in mice using Von Frey Wlaments is shown in Figure 26.6. What might cause the central change that leads to pain? NMDA receptor-dependent central sensitization characteristically underpins the mechanical allodynia and thermal hyperalgesia seen in a variety of other models of chronic pain states (Chaplan et al., 1994; Woolf and Costigan, 1999). Hence, the fact that NMDA receptor-dependent events at a central site are essential for the phenomenon of mechanical allodynia and thermal hyperalgesia behavior to become manifest in Prx / mice indicates that some form of central change, probably sensitization, is a feature of the phenotype (Gillespie et al., 2000). How might peripheral demyelination initiate such changes in the spinal cord? Peripheral ectopic repetitive Wring has been proposed to be the origin of abnormal sensation in diseases characterized by segmental demyelination, particularly when there is no evidence for axonopathic changes. And demyelinated segments can certainly serve as foci for

III. THE MYELIN GENES AND PRODUCTS

au2

640

26. THE PERIAXIN GENE

spontaneous multiple spikes or for those evoked by mechanical stimuli (Baker and Bostock, 1992; Rasminsky, 1981). A common feature of the sensory saphenous nerves of mutant animals is the presence of a spontaneous low-frequency discharge (1 to 3 Hz), in marked contrast to normal littermates in which no spontaneous or background activity is detected. It seems likely that this spontaneous discharge may be linked to the fact that even modest demyelination can lead to marked neuropathic pain behavior in the Prx / mouse (Gillespie et al., 2000). By analogy with other more drastic kinds of peripheral nerve injury, it seems likely that demyelination causes a change in the expression of diVerent sodium channel genes in DRG neurons, which in turn can lead to abnormal Wring. Indeed, an acquired channelopathy has also been proposed to occur in response to CNS demyelination (Waxman, 2001).

Reinnervation and Remyelination in the Periaxin-Null Mouse A major question in peripheral nerve demyelination and the consequent axonopathy is the extent to which reinnervation and remyelination might be feasible. In so far as the Schwann cells of Prx / mice appear to produce relatively normal compact myelin, it might be expected that they would be able to repair damaged Wbers quite eYciently. This is of relevance to human disease, since persisting crush injury can cause pressure palsies of long duration in demyelinating Charcot-Marie-Tooth disease. The extent to which reinnervation and repair recapitulate the normal processes of development in the PNS of mutant mice has been determined after crush lesion in young (6 weeks) and older (4 months) mice. In younger animals, although the number of myelinated axons returns to normal after crush, the diameters of the reinnervating axons in the periaxin-null mice are smaller than in the contra-lateral uncrushed nerve (Williams and Brophy, 2002). At this age, control periaxin-null mice have more hypermyelinated axons than their wild-type counterparts, and this hypermyelination is recapitulated during regeneration after nerve crush. In fact, although periaxin-null mice can undergo peripheral nerve remyelination, the dysregulation of peripheral myelin thickness is even more severe. This derangement is even more pronounced in 4-month-old Prx / mice and results in many incompletely myelinated axons (Gillespie et al., 2000). Hence, it would appear that there is an age-dependent loss in the ability of Prx / Schwann cells to regenerate the myelin sheath after crush injury. This may be a consequence of the extensive accumulation of redundant basal laminas in the nerve, or it may reXect a change in the susceptibility of the axon to Schwann cell ensheathment. Alternatively, it may be that Schwann cells distal to the site of crush injury do not dediVerentiate fully to the type of premyelinating cells analogous to their embryonic counterparts.

CMT4F DISEASE The murine Prx gene maps to a region of mouse chromosome 7 syntenic to human chromosome 19, and the localization of the human gene has been conWrmed (Boerkoel et al., 2001; Gillespie et al., 1997; Guilbot et al., 2001; Takashimi et al., 2002). An autosomal recessive form of Charcot-Marie-Tooth disease, CMT4F, maps to the same chromosomal localization at 19q13, indicating that mutations in the PRX gene are responsible for some of the rarer forms of peripheral demyelinating neuropathy (Boerkoel et al., 2001; Guilbot et al., 2001). Thus far, eight mutations in seven diVerent families have been identiWed, with two families being compound heterozygotes in that they each carry two distinct mutant PRX alleles (Boerkoel et al., 2001; Guilbot et al., 2001; Takashimi et al., 2002). The eVects of these frame shift and nonsense mutations on the amino acid sequence of L-periaxin are indicated in Figure 26.7. Mutations that introduce stop codons usually cause mRNAs to be unstable due to nonsense-mediated RNA decay. However, two of the mutant proteins (C715X R1070X) have been detected in the residual myelinated Wbers present in sural nerve biopsies from

III. THE MYELIN GENES AND PRODUCTS

641

CMT4F DISEASE

PDZ Basic/NLS Repeats V763fs X774

Acidic

S929 fs X957 R82 fs X96 R196 X

13-98 118-196

R368 X

C715 X

R1070 X

R953 X

429-788

1098-1235

1461

FIGURE 26.7 Mutations in CMT4F. The location and type of mutations that have been identiWed so far in CMT4F are shown in the L-periaxin polypeptide. Only the frameshift mutation at the N-terminus (R82fsX96) is expected to also aVect S-periaxin.

these patients (Takashimi et al., 2002, and Parman et al., submitted). Both proteins lack the C-terminus of L-periaxin to a greater or lesser extent, but their N-termini, including the PDZ domain and the DRP2-binding region, are predicted to be intact. This strongly suggests that in addition to these N-terminal protein binding motifs there are other important functional domains toward the C-terminus of the protein. Determination of the proteins involved in these interactions is likely to be extremely informative. Recent unpublished data (Vasiliou et al., in preparation) suggests that one of these may be important in regulating ubiquitination of L-periaxin. It is often diYcult to relate the phenotype of CMT patients to their genotype, and such correlations must be particularly tentative for CMT4F given the small number of patients that has been characterized thus far and the uncertainty about gain versus loss of function eVects where mutant proteins are expressed. In general, most patients show an early age of onset, and in most cases the disease is quite severe. A distinctive feature of PRX mutations is that they seem to cause a prominent sensory neuropathy. This marked sensory involvement has resonance in the pain behavioral phenotype displayed by Prx / mice (section IV.C). Since severe sensory involvement is rare among patients with mutations in other genes linked to CMT, it may be a distinctive clinical feature in many cases of CMT4F neuropathy. Although their function is still enigmatic, it is clear that the clusters of the L-periaxinDRP2-dystroglycan complex in the Schwann cell plasma membrane are essential for the stabilization of the Schwann cell–axon unit. Furthermore, the spheroids represent a striking example of the type of domains that are believed to be a widespread feature of plasma membranes but that are rarely so accessible to study. Since the integrity of these clustered L-periaxin-DRP2-dystroglycan complexes is so important to peripheral nerve myelination, it seems likely that the characterization of other components of this novel dystroglycan complex will identify candidate genes for other autosomal recessive forms of CMT.

References Arroyo, E. J., Xu, T., Poliak, S., Watson, M., Peles, E., and Scherer, S. S. (2001). Internodal specializations of myelinated axons in the central nervous system. Cell Tissue Res 305, 53–66. Baker, M., and Bostock, H. (1992). Ectopic activity in demyelinated spinal root axons of the rat. J. Physiol. (London) 451, 539–552. Boerkoel, C. F., Takashima, H., Stankiewicz, P., Garcia, C. A., Leber, S. M., Rhee-Morris, L., and Lupski, J. R. (2001). Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68, 325–333. Bunge, M. B., Clark, M. B., Dean, A. C., Eldridge, C. F., and Bunge, R. P. (1990). Schwann-Cell Function Depends Upon Axonal Signals and Basal Lamina Components. Ann. NY Acad. Sci. 580, 281–287.

III. THE MYELIN GENES AND PRODUCTS

au3

au6

642

26. THE PERIAXIN GENE

Chaplan, S. R., Bach, F. W., Pogrel, J. W., Chung, J. M., and Yaksh, T. L. (1994). QuantiWed assessment of tactile allodynia in rat paw. J. Neurosci. Meths. 50, 91–93. Dirksen, W. P., Sun, Q., and Rottman, F. M. (1995). Multiple splicing signals control alternative intron retention of Bovine growth hormone pre-mRNA. J. Biol. Chem. 270, 5346–5352. Dytrych, L., Sherman, D. L., Gillespie, C. S., and Brophy, P. J. (1998). Two PDZ-domain proteins encoded by the murine Periaxin gene are the result of alternative intron retention and are diVerentially targeted in Schwann cells. J. Biol. Chem. 273, 5794–5800. Gillespie, C. S., Lee, M., Fantes, J. F., and Brophy, P. J. (1997). The gene encoding the Schwann cell protein periaxin localises on mouse chromosome 7 (Prx). Genomics 41, 297–298. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J. (1994). Periaxin, a novel protein of myelinating Schwann cells with a possible role In axonal ensheathment. Neuron 12, 497–508. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., Cottrell, D. F., Tait, S., Garry, E. M., Wallace, V. C., Ure, J., GriYths, I. R., Smith, A., and Brophy, P. J. (2000). Peripheral demyelination and neuropathic pain behavior in periaxin-deWcient mice. Neuron 26, 523–531. Guilbot, A., Williams, A., Ravise, N., Verny, C., Brice, A., Sherman, D. L., Brophy, P. J., LeGuern, E., Delague, V., Bareil, C., Megarbane, A., and Claustres, M. (2001). A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10, 415–421. Gumbiner, B. M. (1998). Propagation and localization of Wnt signaling. Curr. Opin Genet. Dev. 8, 430–435. Hillier, B. J., Christopherson, K. S., Prehoda, K. E., Bredt, D. S., and Lim, W. A. (1999). Unexpected modes of PDZ domain scaVolding revealed by structure of nNOS-syntrophin complex. Science 284, 812–815. Jessen, K. R., and Mirsky, R. (1999). Schwann cells and their precursors emerge as major regulators of nerve development. Trends Neurosci 22, 402–10. Kachinsky, A. M., Froehner, S. C., and Milgram, S. L. (1999). A PDZ-containing scaVold related to the dystrophin complex at the basolateral membrane of epithelial cells. J. Cell Biol. 145, 391–402. Love, D. R., Hill, D. F., Dickson, G., Spurr, N. K., Byth, B. C., Marsden, R. F., Walsh, F. S., Edwards, Y. H., and Davies, K. E. (1989). An autosomal transcript in skeletal muscle with homology to dystrophin. Nature 339, 55–58. Matsumura, K., Yamada, H., Shimizu, T., and Campbell, K. P. (1993). DiVerential expression of dystrophin, utrophin and dystrophin-associated proteins in peripheral nerve. FEBS Lett. 334, 281–285. Mirsky, R., and Jessen, K. R. (1996). Schwann cell development, diVerentiation and myelination. Curr. Opin. Neurobiol. 6, 89–96. Mirsky, R., Parkinson, D. B., Dong, Z., Meier, C., Calle, E., Brennan, A., Topilko, P., Harris, B. S., Stewart, H. J., and Jessen, K. R. (2001). Regulation of genes involved in Schwann cell development and diVerentiation. Prog Brain Res 132, 3–11. Ramon-Cajal, S. (1933). In ‘‘Histology,’’ pp. 312–313. Bailliere, Tindall & Cox, London. Rasminsky, M. (1981). Hyperexcitability of pathologically myelinated axons and positive symptoms in multiple sclerosis. In ‘‘Demyelinating Diseases: Basic and Clinical Electrophysiology’’ (S. G. Waxman, and J. M. Ritchie, eds.), pp. 289–297. Raven Press, New York. Riethmacher, D., Sonnenberg-Riethmacher, E., Brinkmann, V., Yamaai, T., Lewin, G. R., and Birchmeier, C. (1997). Severe neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature 389, 725–730. Roberts, R. G., Freeman, T. C., Kendall, E., Vetrie, D. L., Dixon, A. K., Shaw-Smith, C., Bone, Q., and Bobrow, M. (1996). Characterization of DRP2, a novel human dystrophin homologue. Nat. Genet. 13, 223–226. Scherer, S. S., Xu, Y. T., Bannerman, P. G. C., Sherman, D. L., and Brophy, P. J. (1995). Periaxin expression in myelinating Schwann cells: Modulation by axon-glial interactions and polarized localization during development. Development 121, 4265–4273. Sherman, D. L., and Brophy, P. J. (2000). A tripartite nuclear localization signal in the PDZ-domain protein L-periaxin. J. Biol. Chem. 275, 4537–4540. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J. (2001). SpeciWc disruption of a Schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron 30, 677–687. Takashimi, H., Boerkoel, C. F., De Jonghe, P., Ceuterick, C., Williams, A., Brophy, P. J., Timmerman, V., and Lupski, J. R. (2002). Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51, 709–715. Trapp, B. D., Peterson, J., RansohoV, R. M., Rudick, R., Mork, S., and Bo, L. (1998). Axonal transection in the lesions of multiple sclerosis. N. Engl. J. Med. 338, 278–285. Waxman, S. G. (2001). Acquired channelopathies in nerve injury and MS. Neurology 56, 1621–1627. Williams, A. C., and Brophy, P. J. (2002). The function of the Periaxin gene during nerve repair in a model of CMT4F. J Anat 200, 323–330. Woolf, C. J., and Costigan, M. (1999). Transcriptional and post-translational plasticity and the generation of inXammatory pain. Proc. Natl. Acad. Sci. (USA) 96, 7723–7773. Xu, H., Wu, X. R., Wewer, U. M., and Engvall, E. (1994). Murine muscular dystrophy caused by a mutation in the laminin alpha 2 (Lama2) gene. Nat. Genet. 8, 297–302.

III. THE MYELIN GENES AND PRODUCTS

C H A P T E R

27 The QKI Gene Rebecca J. Hardy

INTRODUCTION Although we are now beginning to understand the complex intracellular processes within myelin-forming forming cells that lead to the assembly of myelin proteins and lipids into the myelin sheath, much can still be learned. Dysmyelinating mutants such as the shiverer and jimpy have yielded information about the function of structural myelin proteins, and this chapter describes how the investigation of another dysmyelinating mutant, quaking, has led to the discovery and characterization of a novel family of proteins, QKI. QKI proteins play a critical role in myelinogenesis, and the available evidence indicates that QKI proteins have multiple regulatory roles within myelin-forming cells.

THE QUAKING MUTANT Morphology and Biochemistry in qkv Mice The quaking mutant is a spontaneous autosomal recessive mutation found on mouse chromosome 17 (Ebersole et al., 1992), which results in a severe dysmyelination. Quaking, also known as quakingviable (qk v) to distinguish it from other quaking alleles (see the sections titled ‘‘The qk v Gene’’ and ‘‘Expression of QKI outside the Nervous System’’), was Wrst described by Sidman and colleagues (1964). The Wrst abnormalities manifest at around 10 to 12 days post-natal and appear as axial tremor. Mice live a normal life span, but are prone to tonic seizures. Observation of qkv CNS tissue with myelin stains reveals that the amount of myelin is severely reduced. Ultrastructural studies show that qkv mice have reduced numbers of myelin lamellae together with poor compaction of the myelin sheath, increased frequency of cytoplasmic loops, and atypical Schmidt-Lanterman incisures (Berger, 1971; BillingsGagliardi et al., 1980; Samorajski et al., 1970; Suzuki and Zagoren, 1976; Wisniewski and Morell, 1971). The quantity of myelin is more severely reduced in rostral areas and varies between tracts according to their phylogenetic age (Friedrich, 1974). Interestingly, the more severely aVected rostral tracts contain normal numbers of oligodendrocytes, but oligodendrocyte hyperplasia is observed in less severely aVected caudal tracts (Friedrich, 1975). In the PNS, the qkv phenotype is less pronounced, but myelin is also abnormally thin, with occasional lamellar inclusions (Samorajski et al., 1970; Watanabe and Bingle, 1972). Irregular nodal termination and abnormal Schmidt-Lanterman incisures are also seen (Suzuki and Zagoren, 1977). Extensive biochemical analysis of qkv myelin has been undertaken and has been thoroughly reviewed (Hogan and GreenWeld, 1984). BrieXy, qkv mice have quantitative

Myelin Biology and Disorders, Volume 1

643

Copyright 2004, Elsevier (USA). All rights reserved.

644

27. THE QKI GENE

deWciencies in myelin-associated lipids, an ‘‘immature’’ fatty acid composition characterized by a reduction in long-chain fatty acids of the myelin-associated sphingoglycolipids, a deWciency in the myelin proteins PLP and MBP (especially the 14kd and 17kd isoforms), and a reduction in biosynthetic capacity for myelin-associated galactolipids, sulfolipids, long-chain fatty acids, and sterols, without apparent abnormalities in the relevant enzymes.

Molecular Abnormalities in qkv Mice The qkv mutant exhibits many molecular abnormalities including impaired expression of a panel of myelin structural genes (reviewed in Campagnoni and Macklin, 1988). For example, the amount of MBP is reduced by as much as 5 to 20% in qkv brain (Delassalle et al., 1981; Jacque et al., 1983), with the 14kd isoform the most severely aVected (Carnow et al., 1984). Sensitive mRNA studies have revealed that the levels of all major MBP mRNAs are signiWcantly reduced in qkv mice (Li et al., 2000). Furthermore, individual isoforms are aVected diVerently in that the 14kd MBP isoform mRNA is the most severely reduced and the 21.5kd MBP isoform mRNA is the less severely aVected. Because MBP transcription appears normal, these reductions appear to arise from post-transcriptional destablization that occurs in the oligodendrocyte cytoplasm, probably from mislocalization of MBP mRNA (Li et al., 2000; also see the section titled ‘‘QKV Oligodendrocytes Fail to EYciently Localize MBP mRNAs to the Myelin Sheath’’). Levels of 2’,3’-cyclic nucleotide 3’-phosphohydrolase (CNPase) enzyme activity are also reduced in qkv mice (Kurihara et al., 1970). However, in contrast with MBP, only a slight reduction in CNP mRNA is found compared to heterozygous littermates (Zhang and Feng, 2001). The reduction in CNPase activity is not likely to be due to a failure of CNP translation, since CNP mRNAs display normal association with translating polyribosomes. It is also unlikely to be a primary defect of qkv, as QKI proteins do not bind CNP mRNA in wildtype mice (Zhang and Feng, 2001). It has therefore been suggested that accelerated protein degradation is responsible for reduced CNP levels in qkv mice, possibly due to a failure to correctly localize it within oligodendrocytes (Zhang and Feng, 2001). Myelin-associated glycoprotein (MAG) is also aVected in qkv mice, most probably through defects in alternative splicing. Relative levels of L-MAG and S-MAG mRNAs are abnormal in that levels of L-MAG mRNA are severely reduced, whereas S-MAG mRNAs levels are increased. However, S-MAG protein levels are actually reduced, but not as dramatically as those of L-MAG (Bartoszewicz et al., 1995; Braun et al., 1990; Frail and Braun, 1985; Fujita et al., 1988, 1990). Furthermore, both MAG isoforms appear to be abnormally glycosylated (Bartoszewicz et al., 1995; Matthieu et al., 1974). Although reduced levels of L-MAG may be partly due to its increased endocytosis from the cell membrane (Bo et al., 1995), defects in alternative splicing are probably responsible for the MAG abnormalities seen, and these are likely a direct result of misregulation of the qkI gene (see the section titled ‘‘Regulation of Alternative Splicing’’). au4

The qkv Gene Until the mid-1990s, it was not clear which of the many reported abnormalities, if any, were the primary defect in qkv mice, and which were secondary eVects arising from feedback loops in myelinogenesis; the underlying molecular defect of qkv remained elusive for decades after its Wrst description in 1964. Molecular studies had shown that the quaking defect is a large deletion on chromosome 17 (Ebersole et al., 1992), but because none of the myelin genes cloned to that date were localized to the deleted region, the cause of dysmyelination in quaking remained obscure. Several other alleles of quaking have been described that are N-ethyl-nitrosourea (ENU)-induced mutations, which are embryonic lethal in homogygotes (Justice and Bode, 1988; also see the section titled ‘‘Expression of QKI outside the Nervous System’’). qkv complements these alleles for embryonic lethality but not defective myelination. As the ENU mutations are presumably point mutations, this suggested to investigators that a single quaking gene functions in both embryogenesis and myelination (Ebersole et al., 1996). Because the ENU mutations are recessive lethals, it was not thought likely that

III. THE MYELIN GENES AND PRODUCTS

THE QKI GENE

qkv was a deletion of the coding region of this gene, as this would not be viable. Instead, it was suggested that the qkv deletion aVected a regulatory element necessary for myelination but not embryogenesis. If this were true, Artzt and colleagues reasoned that some 5’ regulatory sequences of that gene might lie within the deletion on chromosome 17. They therefore sequenced chromosome 17 distal to the 3’ breakpoint of the qkv deletion and in doing so discovered the candidate qkv gene, qkI (Ebersole et al., 1996). Cloning of the qkI gene in wild-type and qkv mice conWrmed that the qkI coding region is intact in qkv mutants, and that the 3’ breakpoint of the qkv deletion lies some 913bp upstream of the qkI transcription start site in exon 1 (Fig. 27.1; Ebersole et al., 1996; Kondo et al., 1999). Analysis of the qkI gene in one of the homozygous embryonic lethal ENU-induced mutants, qkkt4, revealed that its point mutation lies within the qkI coding region (Ebersole et al., 1996; also see the section titled ‘‘Expression of QKI outside the Nervous System’’). As qkkt4 fails to complement qkv for its dysmyelinating phenotype, the qkI gene must therefore be aVected by the qkv deletion. Furthermore, the fact that qkv homozygous mice exhibit a less severe phenotype than qkkt4 homozygotes (Justice and Bode, 1988), is further evidence that qkv is not a null allele of qkI (see Bedell et al., 1996).

THE QKI GENE The qkI gene has been cloned and described in detail and consists of nine exons spanning around 65kb of DNA (Cox et al., 1999; Kondo et al., 1999). The transcription start site lies almost 1kb downstream of the qkv deletion, and the translation start site lies 21bp downstream of the beginning of exon 2 (Fig. 27.1; Kondo et al., 1999). The main structural feature of the qkI gene is the KH domain located within exons 3, 4, and 5 (Fig. 27.1). The KH domain was originally identiWed as a repeated sequence in the heterogeneous nuclear ribonucleoprotein particle (hnRNP) K and is an RNA-binding domain (Siomi et al., 1993, 1994). Flanking the KH domain, in exons 2/3 and exon 5, are the QUA1 and QUA2 domains, also known as the GSG domain (Fig. 27.1; Ebersole et al., 1996; Jones and Schedl, 1995). The qkI gene is transcribed into a family of alternatively spliced mRNAs, all but one of which consist of a common core sequence with distinct COOH terminal tails. The common sequences are found in the Wrst six exons and downstream exons are alternatively spliced to generate at least four diVerent short COOH terminal sequences and variable 3’UTRs (Fig. 27.1).

Alternative Splicing of the QKI Gene qkI transcripts theoretically generate at least Wve QKI polypeptides, the best characterized of which are QKI5, QKI6, and QKI7, which possess COOH terminal tails of 30, 8, and 14 amino acids, respectively (Fig. 27.1; Ebersole et al., 1996). The COOH terminal tails and 3’UTRs of QKI5, QKI6, and QKI7 are encoded by exons 6 to 9 in a rather complex manner. Exon 7 encodes both the COOH terminal tails and 3’UTRs of QKI6 and QKI7 and part of the COOH terminal tail of QKI5 by way of internal splice acceptor sites (exons 7a, 7b, and 7c, Fig. 27.1). The remainder of QKI5’s COOH terminal tail and 3’UTR is encoded for by exons 8 and 9. As well as the major 3 QKI isoforms, two additional transcripts have been described. The QKIG transcript is generated by failing to splice out the intron between exon 6 and 7, thus encoding a novel COOH terminal tail of eight amino acids (Fig. 27.1; Cox et al., 1999; Hardy, unpublished). A further qkI transcript contains only exons 1 to 4 and fails to utilize the splice site at the 3’ end of exon 4, which results in the addition of three amino acids encoded by sequences within the intron between exons 4 and 5 (Fig. 27.1; Kondo et al., 1999). Because this abbreviated form truncates the KH domain, it has been named QKI
KH. QKIG and QKI
KH have been detected by RT PCR in brain RNA, but their existence at the protein level has yet to be conWrmed (Cox et al., 1999; Kondo et al., 1999). The remainder of this chapter focuses on the three best characterized isoforms, QKI5, QKI6, and QKI7.

III. THE MYELIN GENES AND PRODUCTS

645

646

27. THE QKI GENE

transcription start site ATG qkv deletion breakpoint 1

626bp

2

3

4

5

161bp 144bp 116bp

217bp

6

7

315bp

5294bp

8

9

14bp

2969p

GSG domain

913bp

QUA1

KH domain QUA2

7c

QKI5 76bp 7b

QKI6 24bp 7a

QKI7 45bp 6a

QKIG 24bp 4a

QKI ∆KH

QKI5 GAVATKVRRHDMRVHPYQRIVTADRAATGN QKI6 GMAFPTKG QKI7 EWIMPVMPDASAH QKIG GKYDSCTM QKI∆KH VSR

FIGURE 27.1 The qkI gene is made up of nine exons, which are alternatively spliced to generate at least Wve diVerent isoforms: QKI5, QKI6, QKI7, QKIG, and QKI
KH. Positions of the qkv deletion breakpoint and transcription and translation start sites are shown, as well as exon sizes (Kondo et al., 1999). Approximate positions of the KH domain and GSG (QUA1 and QUA2) domains are indicated by shading. COOH terminal tails unique to each isoform are encoded by exons 7c (QKI5), 7b (QKI6), 7c (QKI7), and 6a (QKIG). The three additional amino acids present in QKI
KH follow on from exon 4. Sequences unique to each isoform are shown in the boxed inset. Underlined sequences refer to amino acids important for nuclear localization (QKI5) or an SH3 binding consensus sequence (QKI7).

Homologues of QKI Homologues of qkI have been described in several other species, including human, Xenopus, zebraWsh, Drosophila, and chicken (Baehrecke, 1997; Lo and Frasch, 1997; Mezquita et al., 1998; Tanaka et al., 1997; ZaVran et al., 1997; Zorn et al., 1997). Sequence alignment of these homologues reveals that their primary sequences are well conserved. As in mice, the qkI genes of other species are alternatively spliced to generate diVerent size transcripts, most of which diVer only in the COOH termini of their coding regions and in their 3’UTRs. Many of these transcripts are not well characterized, but several of their COOH terminal tails show similarity with those of murine QKI (e.g., Zorn and Krieg, 1997). The function of QKI homologues in other species is diverse. For example, who/how/ struthio in Drosophila plays a role in muscle development, and Xenopus Xqua is involved in notochord development. However, though they act in diVerent systems, the cellular functions of QKI homologues maybe conserved in diVerent species.

EXPRESSION OF THE QKI GENE QKI Expression in the Nervous System As would be expected from the gene responsible for the qkv phenotype, qkI is expressed in the adult mouse brain, where all three main qkI isoforms can be detected (Ebersole et al., 1996). A developmental proWle of the expression of qkI5, qkI6, and qkI7 transcripts in total

III. THE MYELIN GENES AND PRODUCTS

au5

647

EXPRESSION OF THE QKI GENE

brain reveals that expression of individual isoforms is developmentally regulated (Hardy et al., 1996). During post-natal development, qkI5 is highly expressed from birth but expression levels decline at around 2 to 3 weeks post-natal to signiWcantly lower levels at 1 month and into adulthood. By contrast, qkI6 and qkI7 messages, while expressed at birth, show increased levels of expression throughout the Wrst 3 post-natal weeks and which are maintained into adulthood. This pattern of expression is consistent with a role in myelinogenesis, which occurs chieXy during the Wrst month of life in rodents. The cellular localization of the QKI proteins in the nervous system has been determined using isoform-speciWc antisera raised against the COOH terminal tails speciWc to each QKI isoform (Hardy et al., 1996). Immunocytochemical analyses of wild-type mouse brain reveals that QKI5, QKI6, and QKI7 are strongly expressed in oligodendrocytes. Examination of sciatic nerve tissue demonstrates that all three isoforms are also expressed in myelin-forming Schwann cells. Thus, QKI proteins are expressed in both central and peripheral nervous system myelinating cells. Although all three isoforms are present in the same population of cells, there is a striking diVerence in the intracellular distribution of individual isoforms. In both oligodendrocytes and Schwann cells, QKI5 is always restricted to the nucleus, whereas QKI6 and QKI7, while present in the nucleus, are concentrated in the perikaryal cytoplasm (Hardy et al., 1996). This compartmentalization of distinct QKI isoforms is also observed in cultured oligodendrocytes, where QKI5 is restricted to the nucleus and QKI6 and QKI7 are found in the cell body and major process networks (Wu et al., 2001). In addition to their expression in myelin-forming cells, QKI proteins are also present in other glial subtypes (Hardy et al., 1996). They are expressed in Bergman glia of the Purkinje cell layer in the cerebellum, and at lower levels in astrocytes throughout the CNS. Interestingly, unlike that in oligodendrocytes and Schwann cells, astrocytic expression of QKI5 is not restricted to the nucleus. All three QKI isoforms are also present at low levels in nonmyelin-forming Schwann cells (Hardy et al., 1996). None of the QKI isoforms studied to date have been found in neurons (with one exception; see the section titled ‘‘Expression of QKIs during Nervous System Development’’).

Developmental Expression of QKI Proteins in Oligodendrocytes As mentioned earlier, the expression of qkI transcripts is developmentally regulated during the early post-natal period in mice. During this period, oligodendrocytes diVerentiate from their progenitor cells into premyelinating cells with multiple branching processes, but no apparent contact to axons (Hardy and Friedrich, 1996b). Oligodendrocytes at this stage are thought to be searching for axons to myelinate. Once they Wnd a suitable axon and begin to elaborate myelin, various changes occur within the cell including expansion of cytoplasmic volume, pruning of the process network, and up-regulation of various myelin proteins (Hardy and Friedrich, 1996b; Wu et al., 2001). Using QKI isoform-speciWc antisera, it has been shown that these changes are accompanied by the regulation of QKI expression in oligodendrocytes (Wu et al., 2001). Premyelinating oligodendrocytes express the myelin proteins CNP and MBP as well as QKI proteins, but as oligodendrocytes contact axons and begin myelination, they dramatically up-regulate their expression of the cytoplasmic QKI proteins, QKI6 and QKI7 (Wu et al., 2001). The majority of QKI6 and QKI7 is found in perikaryal cytoplasm, but some is also found in connecting processes to myelin. Interestingly, the up-regulation of cytoplasmic QKI proteins coincides with the dramatic upregulation of MBP in the oligodendrocyte cell body. As oligodendrocytes mature further, the majority of MBP is lost from the cell body, presumably as its mRNA is translocated into processes for translation at the myelin sheath (see the section titled ‘‘Translocation of mRNAs’’). This loss of MBP from the cell body in mature myelinating oligodendrocytes is accompanied by a down-regulation of cytoplasmic QKI expression (Wu et al., 2001). The mechanisms that lead to QKI up-regulation are unknown, but they may be a result of intracellular signaling following oligodendrocyte process contact with axons. Irrespective of the mechanism, these observations show that the regulation of cytoplasmic QKI expression is closely associated with the initial stages of myelination.

III. THE MYELIN GENES AND PRODUCTS

au6

648

27. THE QKI GENE

DEFECTS IN QKI EXPRESSION IN THE qkv MUTANT Although upstream regions of the qkI gene are deleted in qkv mice, this does not prevent its transcription; the three major qkI transcripts are easily detected in qkv brain (Ebersole et al., 1996). However, consistent with the suggestion that some aspect of qkI regulation is impaired in qkv mutants, immunocytochemical analysis of qkv brains using QKI isoformspeciWc antisera shows that QKI protein levels are severely reduced in these mice (Hardy et al., 1996). As would be expected, levels are signiWcantly reduced in oligodendrocytes and Schwann cells, but interestingly, expression of QKI proteins in astrocytes appears normal. This suggests that the regulatory sequences that control QKI expression in myelin-forming cells are aVected by the qkv deletion, but that those controlling astrocytic expression are not. Thus, diVerent neural cell types use distinct regulatory sequences to control expression of QKI proteins. Even within myelin forming cells, however, individual QKI isoforms are not equally aVected as QKI5 is present in a subset of oligodendrocytes (Hardy et al., 1996). Detailed analysis of qkv mice in the 1970s showed clear diVerences in severity of dysmyelination between diVerent CNS regions (Friedrich, 1974). Areas such as the hindbrain, cerebellum, and optic nerve display less severe dysmyelination whereas more caudal regions such as the anterior commissure are more severely dysmyelinated. Hyperplasia of oligodendrocytes is pronounced in regions with more myelin, and it has been proposed that the increased number of cells is responsible for the less severe phenotype in these areas (Friedrich, 1975). The expression of the nuclear QKI5 isoform is also similarly regionally aVected such that in more severely dysmyelinated regions, oligodendrocytes lack all three QKI isoforms, but in less severely aVected tracts, oligodendrocytes express QKI5, but not QKI6 or QKI7 (Hardy et al., 1996). Thus, the molecular defect in qkv mice leads to the absence of QKI proteins in myelinating cells. This deWciency is likely responsible for dysmyelinating qkv phenotype, indicating that QKI proteins play a critical role in myelinogenesis. Analysis of qkI and its gene products in wild-type and qkv mice will help to elucidate this role and further our understanding of the mechanisms involved in the assembly of myelin.

CELL BIOLOGY OF QKI PROTEINS RNA Binding The structure of the QKI proteins gives us some clues as to their function. Most signiWcantly, the presence of the RNA-binding KH domain suggests that QKI proteins play a role in some aspect of RNA regulation. The QKI proteins belong to a subset of KH domain containing proteins that have a single KH domain. In members of this family, the KH domain is Xanked by a larger sequence of approximately 170 amino acids known as the GSG domain (Fig. 27.1), initially identiWed by aligning the KH domains of the Wrst three members of the family, GRP33, Sam68, and GLD-1 (Jones and Schedl, 1995). The GSG domain has also been referred to as QUA1 and QUA2 (Ebersole et al., 1996; Kondo et al., 1999) and members of this family as STAR proteins, for signal transduction and RNA binding (Vernet and Artzt, 1997). The QKI and related proteins are unusual in that most KH domain-containing proteins described to date possess multiple copies of the KH motif. Cooperation between individual KH domains has been shown to be required for optimal RNA binding (Siomi et al., 1994). Therefore, the functional signiWcance of the single KH domain in QKI proteins is not immediately clear. One possibility is that QKI proteins form multimers, either by selfassociation or by interaction with related proteins, thereby providing two or more proximate KH domains available for RNA binding. Several investigators have demonstrated the ability of QKI proteins to bind RNA (Chen et al., 1997; Chen and Richard, 1998, Li et au7 al., 2001; Saccomanno et al., 1999; Zorn and Krieg, 1997), illustrating that the presence of a single KH domain is suYcient to confer RNA-binding ability. However, structure/function

III. THE MYELIN GENES AND PRODUCTS

CELL BIOLOGY OF QKI PROTEINS

studies have revealed that domains of the protein other than the KH domain are required for RNA binding (Chen and Richard, 1998; Zorn and Krieg, 1997). Initial studies were performed on the Xenopus homologue of QKI, Xqua (Zorn and Krieg, 1997). Xqua shares 94% homology with murine QKI proteins and possesses a COOH terminal tail identical to that of QKI5 (Zorn et al., 1997). Recombinant Xqua binds total embryonic Xenopus RNA, and deletion of the KH domain abolishes this ability (Zorn and Krieg, 1997). However, truncations of Xqua that include portions of the GSG domain downstream of the KH domain (QUA2) abolish its RNA-binding ability, demonstrating that the KH domain is necessary but not suYcient for RNA binding. Studies on mouse QKI7 gave slightly diVerent results and indicate that the N terminal portion of the GSG (QUA1), as well as the KH domain, is vital for RNA binding (Chen and Richard, 1998). These discrepancies are likely due to diVerent stringencies of RNA-binding assays, or to intrinsic diVerences between Xqua and QKI7. It is clear that RNA binding is dependent on the GSG domain, and further evidence has shown that this region is responsible for self-association of QKI proteins. Xqua can homodimerize (Zorn and Krieg, 1997), as can QKI7 (Chen et al., 1997; Chen and Richard, 1998). Chen and Richard (1998) determined that in mouse QKI7, the portion of the protein critical for self-association lay between amino acids 18 to 57, which are found in the GSG domain upstream of the KH domain (QUA1). A point mutation in this domain resulting in a nonconservative amino acid change at position 48 (glutamic acid to glycine) is found in the ENU mutant qkkt4 (Ebersole et al., 1996), and this change abolishes QKI7 selfassociation. A computer prediction of secondary structure indicates that the QUA1 region forms coiled coils, and that these would be disrupted by such an amino acid change at position 48 (Chen and Richard, 1998). This suggests that QKI proteins form dimers through coiled coil interactions mediated by the GSG domain. A coiled coil is also predicted at the 3’ end of the KH domain in QUA2. Studies on mouse QKI7 suggest that this region is not suYcient for self-association but it is probably involved in other protein: protein interactions (Chen and Richard, 1998). As all three QKI isoforms have identical GSG and KH domains, it is reasonable to assume that not only can they self-associate, but that they can also associate with each other. The formation of QKI5:QKI6 or QKI5:QKI7 dimers may be a mechanism by which intracellular localization of isoforms may be regulated (see the following section) or by which function of individual isoforms can be modiWed (see the section titled ‘‘Induction of Apoptosis’’).

Function of QKI COOH Terminal Tails QKI5: Nuclear Localization As QKI proteins are identical save for their COOH terminal tails, it seems reasonable to suppose that these tails are required for some aspect of QKI function. An early observation using QKI isoform-speciWc antisera was that QKI isoforms have diVerent intracellular distributions: QKI5 is restricted to the nucleus, whereas QKI6 and QKI7 are found primarily in the cytoplasm, although they are also present at lower levels in the nucleus (see the section titled ‘‘QKI Expression in the Nervous System’’). Because the QKI5 COOH terminal tail is the only portion of the protein that distinguishes it from the other isoforms, it presumably confers the protein’s distinct localization within the cell. This was originally determined using the Xenopus homologue Xqua (Zorn and Krieg, 1997) and was subsequently conWrmed for mouse QKI5 in a study in which cultured Hela cells were transfected with a construct encoding a protein with the QKI5 COOH terminal tail fused to green Xuorescent protein (GFP; Wu et al., 1999). This protein localizes exclusively to the nuclear compartment, indicating that the QKI5 COOH terminal tail is suYcient to target heterologus proteins to the nucleus. This 30-amino-acid sequence does not contain a classic bipartite or basic residue nuclear localization signal, but the investigators determined that a 7-amino-acid sequence within the COOH terminal tail, RVHPYQR, is responsible for its nuclear localization (Fig. 27.1). Furthermore, the outer Xanking arginines of this motif are critical for correct transport to the nucleus. Additional experiments showed that QKI5

III. THE MYELIN GENES AND PRODUCTS

649

650

27. THE QKI GENE

self-association is not required, but that both the ability to bind RNA and active transcription are necessary for nuclear localization (Wu et al., 1999). Although in vivo QKI5 is mostly found in the nucleus, it is also present in the cytoplasm of cells at some stages of development (e.g., glial progenitor cells; see the section titled ‘‘Expression of QKIs during Nervous System Development’’). This raises the possibility that QKI5 shuttles between the nuclear and cytoplasmic compartments, and the shuttling ability of QKI5 has been demonstrated in in vitro systems (Wu et al., 1999). Because of its ability to heterodimerize with other QKI isoforms, QKI5 may be responsible for ‘‘piggybacking’’ QKI6 and QKI7 into the nucleus. Indeed, in transfected Hela cells, cytoplasmic QKI7 is translocated to the nucleus in the presence of QKI5 (Pilotte et al., 2001). As yet we do not know the mechanism regulating nuclear/cytoplasmic shuttling of QKI5, but it likely has important consequences for QKI function, particularly during development of glial cells (see the section titled ‘‘Expression of QKIs during Nervous System Development’’). QKI7: SH3-Binding and Apoptosis As the COOH terminal tail of QKI5 is responsible for that isoform’s cellular location, it seems likely that the COOH terminal tails of the other QKI isoforms are involved in their localization as well. In fact, there is evidence that the COOH terminal tails of QKI6 and QKI7 are involved in association with various cytoskeletal elements (see the section titled ‘‘QKI Associates with the Cytoskeleton’’). The COOH terminal tail of QKI7 contains the SH3-binding consensus sequence PXXP (Fig. 27.1), a protein:protein interaction domain often utilized in interactions with the cytoskeleton (Morton and Campbell, 1994). However, one group of investigators has shown that in vitro, QKI7 acts as an inducer of apoptosis, and they have termed the 14-amino-acid QKI7 COOH terminal tail a ‘‘killer sequence’’ (Pilotte et al., 2001). This sequence confers the ability to induce apoptosis when fused to heterologous proteins (see the section titled ‘‘Induction of Apoptosis’’). The amino acid sequence of the QKI7 COOH terminal tail does not resemble known apoptotic signals and is not present in any other proteins described to date.

FUNCTION OF QKI PROTEINS IN MYELIN FORMING CELLS Due to the presence of the KH domain, and their ability to bind RNA, it is likely that QKI proteins function as regulators of some aspect of RNA metabolism. The evidence available to date indicates that QKI isoforms have multiple roles within myelin-forming cells, including various aspects of RNA regulation.

Translocation of mRNAs Levels of MBPs are dramatically reduced in qkv myelin (Delassalle et al., 1981; Jacque et al., 1983), and myelin lamellae of qkv mice fail to properly compact. This suggests that qkv mice fail to correctly incorporate MBP into developing myelin, possibly contributing to a general defect in the assembly of myelin components (Hogan and GreenWeld, 1984). Myelin-forming cells target MBPs to the myelin membrane by translocation of MBP mRNAs along their processes and into the myelin membrane, where its local translation enables newly synthesized MBPs to be directly inserted into myelin. It is therefore possible that the failure of qkv mice to correctly incorporate MBPs into myelin is due to an inability to eYciently localize MBP mRNAs to the myelin sheath. This suggests a role for QKI proteins as facilitators of MBP mRNA translocation, and possibly that of other myelin mRNAs, within myelin-forming cells. The available evidence supports a role for QKI in MBP mRNA translocation and is discussed next. qkv Oligodendrocytes Fail to Efficiently Transport MBP mRNAs to Peripheral Processes in Culture Cultured rodent oligodendrocytes have been shown to eYciently transport MBP mRNAs away from the cell soma and into their extensive process networks (Carson et al., 1998).

III. THE MYELIN GENES AND PRODUCTS

au8

651

FUNCTION OF QKI PROTEINS IN MYELIN FORMING CELLS

MBP mRNAs are present in suYcient quantities in peripheral processes to allow detection by in situ hybridization, and in normal cells, up to 80% of MBP mRNA is localized to processes. However, in oligodendrocytes cultured from qkv mice, only 23% of MBP mRNAs are localized in process networks by in situ hybridization (Barbarese, 1991). The transport of other mRNAs appears unaVected. Thus, in culture, qkv oligodendrocytes, which lack QKI6 and QKI7 (see the section titled ‘‘Defects in QKI Expression in the qkv Mutant’’), display a defect in MBP mRNA transport. qkv Oligodendrocytes Fail to Efficiently Localize MBP mRNAs to the Myelin Sheath Following subcellular fractionation of normal rodent brain, MBP and some other myelin mRNAs (e.g., myelin-associated oligodendrocyte basic protein, MOBP) can be found in the myelin fraction as a result of their transport to the myelin sheath in vivo (Colman et al., 1982; Gould et al., 1999). In qkv mice, however, the proportion of MBP mRNA found in the myelin fraction is greatly reduced compared to phenotypically normal littermates (Li et al., 2000). This demonstrates that, as in culture, qkv oligodendrocytes fail to eYciently transport MBP mRNA to myelin in the intact brain. It seems likely that the instability of MBP mRNA in qkv mice (see the section titled ‘‘Molecular Abnormalities in qkv Mice’’) is a result of its mislocalization; degradation of improperly localized mRNAs may be a mechanism by which cells ensure that proteins are translated at the preferred location. QKI7 Binds to MBP mRNA If QKI proteins are involved in mRNA localization within myelin forming cells, such a role would presumably be mediated by the binding of mRNA to QKI proteins via their KH domain. QKI proteins are capable of RNA binding (see the section titled ‘‘RNA Binding’’) and, in fact, in vitro translated QKI7 has been demonstrated to bind directly to MBP mRNA (Li et al., 2000). It is not known whether other QKI isoforms are capable of binding MBP mRNA, but it would seem likely given that the region of the proteins responsible for RNA binding are common to all isoforms. In addition, removal of the 3’UTR from the MBP mRNA signiWcantly reduces binding activity, indicating that this region is important for QKI binding. This is interesting in the light of the fact that sequences present in the 3’UTR of MBP mRNA are critical for its correct localization away from the oligodendrocyte perikaryon (Carson et al., 1998). QKI Associates with the Cytoskeleton Consistent with a role in the localization of mRNAs to the myelin sheath, QKI proteins have been shown to localize with cytoskeletal elements. The cytoskeleton is critical in the traYcking of mRNAs to myelin, providing a molecular highway upon which the mRNAs are transported. For example, the translocation of MBP mRNA requires an intact microtubule network and is dependent on the microtubule motor kinesin (Carson et al., 1997). Studies on cultured oligodendrocytes show that the correct localization of the cytoplasmic QKI isoforms, QKI6 and QKI7, is also dependent upon an intact cytoskeleton (Wu et al., 2001). Immuncytochemical analyses and treatment with cytoskeletal disrupting drugs reveals that QKI6 associates with microtubules and actin Wlaments, whereas QKI7 associates with microtubules but not actin Wlaments. To date it is unclear whether QKI proteins bind directly to cytoskeletal proteins or rather associate with the cytoskeleton through intermediary proteins. In either case, it seems likely that the ability of QKI6 to interact with microtubules or actin Wlaments is mediated through distinct domains of the protein. As QKI6 and QKI7 diVer only in their short COOH terminal tails, the QKI6 speciWc tail is probably responsible for association with actin Wlaments, whereas association with microtubules would be conferred by sequences common to both QKI6 and QKI7. Of course, these sequences would also be present in QKI5, but this isoform is presumably excluded from cytoskeletal interactions in the cytoplasm as it is targeted to the nucleus by its own speciWc COOH terminal tail (Wu et al., 1999; also see the section titled ‘‘QKI7: SH3Binding and Apoptosis’’).

III. THE MYELIN GENES AND PRODUCTS

au9

652

27. THE QKI GENE

QKI Proteins Facilitate Translocation of MBP mRNAs Taken together, these lines of evidence strongly suggest a role for cytoplasmic QKI isoforms in the transport of MBP mRNA in myelin-forming cells. In the qkv mutant, oligodendrocytes, which lack QKI, fail to eYciently transport MBP mRNA to peripheral processes in vitro and to the myelin sheath in vivo. QKI7 binds MBP mRNA directly via its 3’UTR, a region known to be critical for mRNA translocation. In addition, cytoplasmic QKI isoforms associate with the cytoskeleton. It therefore seems reasonable to hypothesize that the cytoplasmic QKI isoforms mediate interactions between mRNAs and the cytoskeleton during mRNA translocation in oligodendrocytes. Furthermore, distinct QKI isoforms facilitate mRNA traYcking on diVerent cytoskeletal highways: MBP mRNAs leave the oligodendrocyte cell body and move along major processes via the microtubule network (Carson et al., 1997), which may be mediated by either QKI6 or QKI7, but local delivery of mRNAs to the periphery of the cell or the myelin sheath would be mediated by the association of QKI6 with actin Wlaments. The Function of QKI Proteins during mRNA Translocation Are QKI proteins therefore mere scaVolding molecules that anchor mRNAs to the cytoskeleton during translocation? There is some circumstantial evidence that suggests they have a more active role: the suppression of translation of transported transcripts. This came to light as a result of similarities between QKI proteins and the C. elegans protein GLD-1. GLD-1 inhibits translation of another C. elegans gene, tra2, by binding to two elements in its 3’UTR, which have been termed TGEs (for tra2 and GLI [the human gene, GLI-1] elements; Jan et al., 1997, 1999). The fact that QKI proteins and GLD-1 are both members of the STAR family of KH domain containing proteins (Vernet and Artzt, 1997) suggests that QKI proteins might also be TGE-dependent translational repressors. Experiments have now shown that not only do QKI proteins bind speciWcally to TGEs, but that in doing so, QKI6 represses translation of TGE-containing reporter constructs (Saccomanno et al., 1999). Thus, at least in in vitro systems, QKI proteins can repress translation by binding to TGEs. Does TGE-binding and repression of translation have any relevance to the transport of mRNAs to the myelin sheath? This is not clear to date, and the MBP mRNA 3’UTR does not contain any strong TGE consensus sequences. However, there is a putative TGE consensus sequence present in exon8B of MOBP (Fig. 27.2). MOBP is alternatively spliced to generate at least six isoforms (McCallion et al., 1999). Unlike MBP isoforms, all of whose mRNAs are transported to myelin, only the mRNAs of four MOBP isoforms are translocated to the myelin sheath, whereas the rest remain in the cell body (Gould et al., 1999). Those MOBP isoforms whose mRNAs are transported to myelin diVer from their nontransported counterparts in that they contain exon 8B as part of their 3’UTR (Gould et al., 1999; McCallion et al., 1999). Thus, exon 8B probably contains sequences important for mRNA transport, and this may include its TGE. QKI proteins may bind to the TGE consensus and repress translation of MOBP mRNA during its transport. Suppression of mRNA translation during translocation is presumably important for ensuring correct targeting of protein to its intended destination, in this case the myelin sheath.

Regulation of Alternative Splicing While the evidence strongly suggests that cytoplasmic QKI proteins are involved in mRNA transport, this is probably not the case for QKI5, given is predominantly nuclear localization (Hardy et al., 1996; Wu et al., 1999, 2001). Evidence exists, in fact, that QKI5 regulates a diVerent aspect of RNA metabolism: alternative splicing. The alternative splicing of several genes is abnormal in qkv mutant mice (Hardy, 1998a), suggesting a role for QKI proteins in the alternative splicing of more than one gene. One such gene is MAG. In normal mice, MAG is alternatively spliced into two isoforms, S- au10 MAG and L-MAG. These are generated by the inclusion (S-MAG) or exclusion (L-MAG) of exon 12 (Fig. 27.3), and the proportion of each is developmentally regulated: L-MAG is

III. THE MYELIN GENES AND PRODUCTS

653

FUNCTION OF QKI PROTEINS IN MYELIN FORMING CELLS

(n)

CC C UUUC U AA U

Consensus

CUCA

(n)

UUUCU

MOBP exon 8B

CUCAGCCAGGAAUUAUUUCUUAUUAUACUUUUAUUUCU GACAGGACAAUUUCU

Human GLI-1

CUCAUCCAUCACAGAUCGCAUUUCCUAAGGGGUUUCU

FIGURE 27.2 Consensus sequence for TGE (Jan et al., 1997) and putative TGE in exon 8B of MOBP. The TGE sequence in human GLI-1 is included for comparison.

stop codon

11

12

13

S-MAG

11

13

L-MAG

FIGURE 27.3 Alternative splicing of the MAG gene generates the two isoforms: S-MAG and L-MAG. S-MAG utilizes exons 11, 12, and 13, but the coding region terminates at a stop codon within exon 12 (arrow). L-MAG splices out exon 12, thus avoiding the stop codon and resulting in a longer COOH terminal tail.

the predominant isoform during early myelination, whereas S-MAG is predominant in the adult. In qkv mice, levels of L-MAG mRNA are dramatically reduced, whereas levels of SMAG mRNA are increased. At the protein level, amounts of S-MAG are reduced but not as dramatically as those of L-MAG (Bartoszewicz et al., 1995; Braun et al., 1990; Frail and Braun, 1985; Fujita et al., 1988, 1990). Therefore, abnormal splicing of MAG transcripts is clearly a feature of qkv pathology. It has now been demonstrated that the regulation of MAG mRNA alternative splicing is regulated by QKI5, at least in vitro (Wu et al., 2002). When QKI5 and a MAG mini-gene are co-expressed in COS7 cells, QKI5 represses the inclusion of exon12, and therefore the generation of S-MAG. This indicates that QKI5 is involved in the developmentally regulated alternative splicing of MAG and acts to promote the generation of L-MAG over S-MAG. Thus, in qkv mice, where QKI5 is absent in most myelin-forming cells, S-MAG mRNA is in abundance at the expense of L-MAG. It has been determined that QKI5 binds directly to MAG pre-mRNA at a 53nt intronic sequence termed the QASE, where it may interfere with splice site recognition (Wu et al., 2002). It is likely, however, that MAG is but one of the targets of QKI5. Qkv mice have defects in alternative splicing of several myelin genes, the most notable of which is MBP; qkv mice contain abnormal ratios of MBP isoforms (Carnow et al., 1984; Li et al., 2000). QKI5 therefore may be involved in the regulation of alternative splicing of a panel of myelin genes. QKI5 may also regulate the alternative splicing of nonmyelin genes in other tissues (see the section titled ‘‘Expression of QKI outside the Nervous System’’).

Induction of Apoptosis As mentioned earlier (see the section titled ‘‘Induction of Apoptosis’’), QKI7 can act as an inducer of apoptosis, at least in vitro. Apoptotic activity of QKI7 was originally described in NIH 3T3 cells, in which transfection of a GPF:QKI7 fusion protein resulted in the death

III. THE MYELIN GENES AND PRODUCTS

au16

654

27. THE QKI GENE

by apoptosis of 90% of cells after 48 hours (Chen et al., 1997). Similar expression of GFP:QKI5 does not induce apoptotic cell death (Pilotte et al., 2001). Subsequent experiments showed that over-expression of QKI7 in cultured Wbroblasts or primary rat oligodendrocytes also induces apoptosis (Pilotte et al., 2000). QKI7 appears to utilize classical apoptotic/survival pathways because its activity is caspase dependent and is suppressed by over-expression of the survival protein Bcl-2 (Pilotte et al., 2001). Structure/function analysis of QKI7 indicates that the GSG domain, and therefore selfassociation and RNA binding, is not required for apoptotic activity (Pilotte et al., 2001). However, deletion of the COOH terminal 14 amino acids, which distinguish QKI7 from the other QKI isoforms, is suYcient to prevent cell death. Furthermore, these same 14 amino acids, when fused with the heterologous nonapoptotic proteins GFP or GLD-1, conferred the ability to induce apoptosis. These results show that the QKI7 COOH terminal tail is necessary and suYcient for apoptotic activity. Hence, this ‘‘killer sequence’’ functions with the apoptotic machinery independently of other functional QKI domains. Intracellular localization of QKI7 is important in regulating its function as an inducer of apoptosis. Directing QKI7 to the nucleus by fusion with a nuclear localization signal inactivates apoptotic activity (Pilotte et al., 2001), indicating that the QKI7 ‘‘killer sequence’’ is functional only in the cytoplasm. Sequestration of QKI7 to the nucleus seems to be achieved in vivo by heterodimerization with other QKI isoforms. For example, cooverexpression of QKI7 with either QKI5 or QKI6 results in its translocation to the nucleus, accompanied by inactivation of apoptotic activity. Mutations of the QKI proteins that destroy their ability to heterodimerize result in a failure to translocate QKI7 to the nucleus and a failure to suppress cell death. Thus, both heterodimerization and nuclear localization are required for suppression of apoptotic activity. These experiments suggest that the balance between expression of QKI7 and the other QKI isoforms is critical for regulation of QKI7 induced apoptosis, and therefore cell survival. Over-expression of QKI7 will induce cell death, whereas over-expression of QKI5 or QKI6 will promote cell survival. This is interesting in the light of the observed hyperplasia of oligodendrocytes in less severely aVected tracts in the qkv mutant (Friedrich, 1975), which may arise due to disregulation of cell number as a result of the absence of QKI proteins. This method of cell number control may also be important during the generation of glial progenitors (see the section titled ‘‘Expression of QKIs during Nervous System Development’’) and in organogenesis in other tissues (see the section titled ‘‘Expression of QKI outside the Nervous System’’). However, it is not yet clear how the role of QKI7 as an inducer of apoptosis can be reconciled with its interaction with the cytoskeleton (Wu et al., 2001) and ability to bind MBP mRNA (Li et al., 2001). It must also be considered that these experiments involve over-expression of QKI7 in in vitro systems and their relevance to physiological expression of QKI proteins in the whole animal is yet to be determined.

OTHER ROLES FOR QKI PROTEINS Expression of QKIs during Nervous System Development As well as their role in myelinogenesis, there is evidence that QKI proteins may also function during earlier stages of glial development. Embryonic expression of qkI transcripts (Ebersole et al., 1996) suggested that QKI proteins might be expressed in the developing embryonic brain and possibly in precursor cells of oligodendrocytes. Immunocytochemical analysis of embryonic mouse tissue has shown that all three QKI isoforms are expressed in the neural progenitors of the proliferative ventricular zone (vz) throughout the CNS (Hardy, 1998b). Intracellular localization of QKIs in these cells mirrors that in oligodendrocytes: QKI5 is restricted to the nucleus, whereas QKI6 and QKI7 are found predominantly in the cytoplasm. Double immunolabeling of QKIþ cells with early neuronal markers shows that newly formed neurons have signiWcantly reduced levels of QKI proteins compared to undiVer-

III. THE MYELIN GENES AND PRODUCTS

au11

655

OTHER ROLES FOR QKI PROTEINS

entiated neural progenitors. Therefore, neural progenitors down-regulate QKI expression as they acquire a neuronal fate and migrate away from the vz (Fig. 27.4). One exception to this observation is in the ventral spinal cord where motor neurons express QKI5 in their nuclei. They do not express other QKI isoforms. Although all vz cells express QKIs, those in strictly delineated subregions of the vz express dramatically higher levels of QKI5 (Hardy, 1998b). This expression pattern is transient and is followed by the appearance of QKIþ cells apparently emerging from this subdomain and migrating into the surrounding parenchyma. Interestingly, QKI5 is not restricted to the nucleus in these cells, but rather is found in the cytoplasm of the cell body and leading processes. This is interesting in the light of the fact that in in vitro systems, QKI5 can shuttle between the nucleus and cytoplasm (Wu et al., 1999). Movement of QKI5 out of the nucleus may be a way of regulating its function, or a mechanism by which other QKI isoforms are transported via ‘‘piggyback’’ into the cytoplasm. It is postulated that the QKIþ cells that emerge from subdomains of the vz characterized by elevated QKI5 expression are glial progenitors. They arise after the main period of neurogenesis is complete, they express the glial progenitor Wlament protein nestin, they are proliferative, and they emerge from regions of the vz shown to give rise to oligodendrocyte progenitors and astrocytes (Hardy, 1998b). Also, in early post-natal development, QKI proteins are expressed in the subventricular zone, a region known to be the source of oligodendrocyte progenitor cells. In vitro experiments using the P19 embryonal carcinoma cell line have also shown that neural progenitors generated by aggregation and retinoic acid treatment of P19 cells down-regulate QKI expression as they acquire a neuronal phenotype but instead maintain expression of QKI as they generate diVerentiated glia (Hardy, 1998b). Therefore, it appears that neural progenitors regulate their expression of QKI proteins as they acquire a neuronal or glial fate (Fig. 27.4). Neural progenitors generating neurons down-regulate QKI expression, whereas those acquiring a glial fate maintain QKI expression. Whether or not QKI proteins play a critical role in this process has yet to be demonstrated. It seems reasonable to suppose that QKI proteins are involved in the regulation of mRNA localization, translation, or splicing in diVerentiating neural progenitors and immature glia, just as they are in diVerentiated myelin-forming cells.

Expression of QKI outside the Nervous System Early genetic analysis of quaking indicated a role for the quaking gene in embryogenesis. Four ENU-induced alleles of quaking, collectively referred to as qke, cause embryonic death at midgestation in homozygotes (Justice and Bode, 1988; Shedlovsky et al., 1988). qkkt4 contains point mutation that results in a glutamic acid to glycine change at position 48, within the GSG domain (Ebersole et al., 1996). As mentioned earlier (see the section titled ‘‘RNA Binding’’), this substitution abolishes dimerization and likely obliterates QKI function. Certainly, constructs with this mutation induce apoptosis in vitro (Chen et al., 1997). The qkl-1 allele has a point mutation that results in the loss of a splice site necessary for generation of the qkI5 transcript (Cox et al., 1999), indicating that the loss of the QKI5 nuclear isoform alone is suYcient to cause embryonic lethality. The qkk2 allele contains a nonconservative amino acid change in the KH domain, presumably aVecting RNAbinding properties of the protein (Cox et al., 1999). The defect in the qkkt1 allele does not lie in the QKI coding region and is yet to be identiWed. qkI messages have been detected in early embryonic brain and heart (Ebersole et al., 1996), suggesting that failure of QKI function in these organs might be responsible for the embryonic death of the qke alleles. Their demise occurs generally around E10 to E12.5 (Justice and Bode, 1988), some 2 to 4 days before the onset of myelination (Hardy and Friedrich, 1996a). The only allele to be studied in detail is qkk2, and in these mice, neurological and cardiac development appears normal prior to their death (Noveroske et al., 2002). However, qkk2 homozygote embryos do have a severe defect in yolk sac vascular remodeling, together with abnormal vessels in the embryo. In wild-type embryos, QKI5 is expressed in the yolk sack endoderm, adjacent to the vascular cells (Noveroske et al.,

III. THE MYELIN GENES AND PRODUCTS

au12

656

27. THE QKI GENE

neuronal progenitor

neuron

QKI – QKI5 + + +

QKI + QKI + vz

oligodendrocyte

glial progenitor QKI + astrocyte

FIGURE 27.4 Schematic representation of QKI expression during neurogenesis and gliogenesis. QKIþ glial progenitors emerge from subdomains of the ventricular zone (vz) that express elevated levels of QKI5. QKI expression is maintained in these cells as they migrate away from the vz and diVerentiate into mature glia. By contrast, neurons emerging from the vz down-regulate QKI expression.

2002). It is postulated that absence of functional QKI5 in these cells leads to failure of proper vascularization and therefore leads to embryonic lethality. Analysis of the qkk2 allele has revealed a novel function for QKI proteins and thorough characterization of remaining qke and other quaking alleles may reveal additional QKI functions in other tissues. Thus, it is likely that the cellular roles of QKI proteins in regulation of mRNA translocation, alternative splicing, and suppression of translation investigated in myelinating cells also operate in other cell types. Of particular relevance is the proposed role of QKI7 as an inducer of apoptosis (see the section titled ‘‘Induction of Apoptosis’’), which indicates that QKI proteins are involved in the regulation of cell number during organogenesis.

Regulation of HIV Viral Gene Expression QKI proteins are also potential mediators of post-transcriptional regulation of human immunodeWciency virus (HIV) replication. The HIV Rev protein facilitates the nuclear export of unspliced or singly spliced viral mRNA (Hope, 1999). It contains an RNAbinding domain that interacts with a target sequence named the RRE (Rev response element). The STAR family protein, Sam68 binds to the RRE and can replicate Rev function, as well as synergize with Rev in RRE-mediated gene expression and virus replication (Reddy et al., 1999). Due to the sequence homology between Sam68 and QKI proteins, it might be expected that the latter is also able to substitute for Rev in regulating HIV replication. In fact, QKI5, QKI6, and QKI7 fail to bind to RRE-containing RNA and fail to transactivate RRE-directed reporter gene expression directly (Reddy et al., 2002). They are, however, able to complex with Rev and enhance Rev transactivation on the RRE. These preliminary studies suggest that QKI proteins may play a role in the post-transcriptional regulation of HIV.

Acknowledgments I would like to thank Matt Flaherty, without whose cooperation, understanding, and support this chapter could not have been written. Thanks also go to Richard Reynolds for helpful comments on the manuscript and practical and moral support, as always. This chapter is dedicated to Rose Charlotte Flaherty.

References Baehrecke, E. H. (1997). who encodes a KH RNA binding protein that functions in muscle development. Development 124, 1323–1332.

III. THE MYELIN GENES AND PRODUCTS

au13

657

OTHER ROLES FOR QKI PROTEINS

Barbarese, E. (1991). Spatial distribution of myelin basic protein mRNA and polypeptide in quaking oligodendrocytes in culture. J. Neurosci. Res. 29, 271–281. Bartoszewicz, Z. P., Noronha, A. B., Fujita, N., Sato, S., Bo, L., Trapp, B. D., and Quarles, R. H. (1995). Abnormal expression and glycosylation of the large and small isoforms of myelin-associated glycoprotein in dysmyelinating quaking mutants. J. Neurosci Res. 41, 27–38. Bedell, M. A., Jenkins, N. A., and Copeland, N. G. (1996). Good genes in bad neighbourhoods. Nat. Genet. 12, 229–232. Berger, B. (1971). Quelques aspects ultrastructuraux de la substance blanche chez la souris quaking. Brain Res. 25, 35–53. Bo, L., Quarles, R. H., Fujita, N., Bartoszewicz, Z., Sato, S., and Trapp, B. D. (1995). Endocytic depletion of L-MAG from CNS myelin in quaking mice. J Cell Biol. 131, 1811–1820. Braun, P. E., Horvath, E., and Edwards, A. M. (1990). Two isoforms of myelin-associated glycoprotein accumulate in quaking mice: only the large polypeptide is phosphorylated. Dev. Neurosci. 12, 286–292. Billings-Gagliardi, S., Adcock, L. H., and Wolf, M. K. (1980). Hypomyelinated mutant mice: description of jpmsd and comparison with jp and qk on their present genetic backgrounds. Brain Res. 194, 325–338. Campagnoni, A. T., and Macklin, W. B. (1988). Cellular and molecular aspects of myelin protein gene expression. Mol. Neurobiol. 2, 41–89. Carnow, T. B., Carson, J. H., BrostoV, S. W., and Hogan, E. L. (1984). Myelin basic protein gene expression in quaking, jimpy, and myelin synthesis-deWcient mice. Dev. Biol. 106, 38–44. Carson, J. H., Worboys, K., Ainger, K., and Barbarese, E. (1997). Translocation of myelin basic protein mRNA in oligodendrocytes requires microtubules and kinesin. Cell. Motil. Cytoskeleton 38, 318–328. Carson, J. H., Kwon, S., and Barbarese, E. (1998). RNA traYcking in myelinating cells. Curr. Opin. Neurobiol. 8, 607–612. Chen, T., Damaj, B. B., Herrera, C., Lasko, P., and Richard, S. (1997). Self-association of the single-KH-domain family members Sam 68, GRP 33, GLD-1, and QkI: Role of the KH domain. Mol. Cell. Biol. 17, 5707–5718. Chen, T., and Richard, S. (1998). Structure-function analysis of QkI: A lethal point mutation in mouse quaking prevents homodimerization. Mol. Cell Biol. 18, 4863–4871. Colman, D. R., Kreibich, G., Frey, A. B., and Sabatini, D. D. (1982). Synthesis and incorporation of myelin polypeptides into CNS myelin. J. Cell Biol. 95, 598–608. Cox, R. D., Hugill, A., Shedlovsky, A., Noveroske, J. K., Best, S., Justice, M. J., Lehrach, H., and Dove, W. F. (1999). Contrasting eVects of ENU induced embryonic lethal mutations of the quaking gene. Genomics 57, 333–341. Delassalle, A., Zalc, B., Lachapelle, F., Raoul, M., Collier, P., and Jacque, C. (1981). Regional distribution of myelin basic protein in the central nervous system of quaking, jimpy, and normal mice during development and aging. J. Neurosci. Res. 6, 303–313. Ebersole, T. A., Chen, Q., Justice, M. J., and Artzt, K. (1996). The quaking gene product necessary in embryogenesis and myelination combines features of RNA binding and signal transduction proteins. Nat. Genet. 12, 260–265. Ebersole, T., Rho, O., and Artzt, K. (1992). The proximal end of mouse chromosome 17: new molecular markers identify a deletion associated with quakingviable. Genetics 131, 183–190. Frail, D. E., and Braun, P. E. (1985). Abnormal expression of the myelin-associated glycoprotein in the central nervous system of dysmyelinating mutant mice. J. Neurochem. 45, 1071–1075. Friedrich, V. L., Jr. (1974). The myelin deWcit in quaking mice. Brain Res. 82, 168–172. Friedrich, V. L., Jr. (1975). Hyperplasia of oligodendrocytes in quaking mice. Anat. Embryol. (Berl). 147, 259–271. Fujita, N., Sato, S., Ishiguro, H., Inuzuka, T., Baba, H., Kurihara, T., Takahashi, Y., and Miyatake, T. (1990). The large isoform of myelin-associated glycoprotein is scarcely expressed in the quaking mouse brain. J. Neurochem. 55, 1056–1059. Fujita, N., Sato, S., Kurihara, T., Inuzuka, T., Takahashi, Y., and Miyatake, T. (1988). Developmentally regulated alternative splicing of brain myelin-associated glycoprotein mRNA is lacking in the quaking mouse. FEBS Lett. 232, 323–327. Fyrberg, C., Becker, J., Barthmaier, P., MahaVey, J., and Fyrberg, E. (1998). A family of Drosophila genes encoding quaking-related maxi-KH domains. Biochem. Genet. 36, 51–64. Gould, R. M., Freund, C. M., and Barbarese, E. (1999). Myelin-associated oligodendrocytic basic protein mRNAs reside at diVerent subcellular locations. J. Neurochem. 73, 1913–1924. Hardy, R. J. (1998a). Molecular defects in the dysmyelinating mutant quaking. J. Neurosci. Res. 51, 417–422. Hardy, R. J. (1998b). QKI expression is regulated during neuron-glial cell fate decisions. J. Neurosci. Res. 54, 46–57. Hardy, R. J., and Friedrich, V. L., Jr. (1996a). Oligodendrocyte progenitors are generated throughout the embryonic mouse brain, but diVerentiate in restricted foci. Development 122, 2059–69. Hardy, R. J., and Friedrich, V. L., Jr. (1996b). Progressive remodeling of the oligodendrocyte process arbor during myelinogenesis. Dev. Neurosci. 18, 243–54. Hardy, R. J., Loushin, C. L., Friedrich, V. L., Jr, Chen, Q., Ebersole, T. A., Lazzarini, R. A., and Artzt, K. (1996). Neural cell type-speciWc expression of QKI proteins is altered in quakingviable mutant mice. J. Neurosci. 16, 7941–7949. Hogan, E. L., and GreenWeld, S. (1984). Animal models of genetic disorders of myelin. In ‘‘Myelin’’ (P. Morrell, ed.), pp. 489–534. Plenum Press, New York.

III. THE MYELIN GENES AND PRODUCTS

au14 au15

658

27. THE QKI GENE

Hope, T. J. (1999). The ins and outs of HIV Rev. Arch. Biochem. Biophys. 365, 186–191. Jacque, C., Delassalle, A., Raoul, M., and Baumann, N. (1983). Myelin basic protein deposition in the optic and sciatic nerves of dysmyelinating mutants quaking, jimpy, Trembler, mld, and shiverer during development. J. Neurochem. 41, 1335–1340. Jan, E., Motzny, C. K., Graves, L. E., and Goodwin, E. B. (1999). The STAR protein, GLD-1, is a translational regulator of sexual identity in Caenorhabditis elegans. EMBO J. 18, 258–269. Jan, E., Yoon, J. W., Walterhouse, D., Iannaccone, P., and Goodwin, E. B. (1997). Conservation of the C. elegans tra-2 3’UTR translational control. EMBO J. 16, 6301–6313. Jones, A. R., and Schedl, T. (1995). Mutations in gld-1, a female germ cell-speciWc tumor suppressor gene in Caenorhabditis elegans, aVect a conserved domain also found in Src-associated protein Sam68. Genes Dev. 9, 1491–1504. Justice, M. J., and Bode, V. C. (1988). Three ENU-induced alleles of the murine quaking locus are recessive embryonic lethal mutations. Genet. Res. 51, 95–102. Kondo, T., Furuta, T., Mitsunaga, K., Ebersole, T. A., Shichiri, M., Wu, J., Artzt, K., Yamamura, K., and Abe, K. (1999). Genomic organization and expression analysis of the mouse qkI locus. Mamm. Genome. 10, 662–669. Kurihara, T., Nussbaum, J. L., and Mandel, P. (1970). 2’,3’-cyclic nucleotide 3’-phosphohydrolase in brains of mutant mice with deWcient myelination. J. Neurochem. 17, 993–997. Li,Z., Zhang,Y., Li, D., and Feng, Y. (2000). Destabilization and mislocalization of myelin basic protein mRNAs in quaking dysmyelination lacking the QKI RNA-binding proteins. J. Neurosci. 20, 4944–4953. Lo, P. C., and Frasch, M. (1997). A novel KH-domain protein mediates cell adhesion processes in Drosophila. Dev. Biol. 190, 241–256. Matthieu, J. M., Daniel, A., Quarles, R. H., and Brady, R. O. (1974). Interactions of concanavalin A and other lectins with CNS myelin. Brain Res. 81, 348–353. McCallion, A. S., Stewart, G. J., Montague, P., GriYths, I. R., and Davies, R. W. (1999). Splicing pattern, transcript start distribution, and DNA sequence of the mouse gene (Mobp) encoding myelin-associated oligodendrocytic basic protein. Mol. Cell Neurosci. 13, 229–236. Mezquita, J., Pau, M., and Mezquita, C. (1998). Four isoforms of the signal-transduction and RNA-binding protein QKI expressed during chicken spermatogenesis. Mol. Reprod. Dev. 50, 70–78. Morton, C. J., and Campbell, I. D. (1994). SH3 domains. Molecular ‘Velcro’. Curr. Biol. 4, 615–617. Noveroske, J. K, Lai, L., Gaussin, V., Northrop, J. L., Nakamura, H., Hirschi, K. K., and Justice, M. J. (2002). Quaking is essential for blood vessel development. Genesis 32, 218–230. Pilotte, J., Larocque, D., and Richard, S. (2001). Nuclear translocation controlled by alternatively spliced isoforms inactivates the QUAKING apoptotic inducer. Genes Dev. 15, 845–858. Reddy, T. R., Suhasini,M., Xu, W., Yeh, L. Y., Yang, J. P., Wu, J., Artzt, K., and Wong-Staal, F. (2002). A role for KH domain proteins (Sam68-like mammalian proteins and quaking proteins) in the post-transcriptional regulation of HIV replication. J. Biol. Chem. 277, 5778–5784. Reddy, T. R., Xu, W., Mau, J. K., Goodwin, C. D., Suhasini, M., Tang, H., Frimpong, K., Rose, D. W., and Wong-Staal, F. (1999). Inhibition of HIV replication by dominant negative mutants of Sam 68, a functional homolog of HIV-1 Rev. Nat. Med. 5, 635–642. Saccomanno, L., Loushin, C., Jan, E., Punkay, E., Artzt, K., and Goodwin, E. B. (1999). The STAR protein QKI6 is a translational repressor. Proc. Natl. Acad. Sci. U. S. A. 96, 12605–12610. Samorajski, T., Friede, R. L., and Reimer, P. R. (1970). Hypomyelination in the quaking mouse. A model for the analysis of disturbed myelin formation. J. Neuropathol. Exp. Neurol. 29, 507–523. Sidman, R. L., Dickie, M. M., and Appel, S. H. (1964). Mutant mice (quaking and jimpy) with deWcient myelination in the central nervous system. Science 144, 309–312. Shedlovsky, A., King, T. R., and Dove, W. F. (1988). Saturation germ line mutagenesis of the murine t region including a lethal allele at the quaking locus. Proc. Natl. Acad. Sci. U. S. A. 85, 180–184. Siomi, H., Choi, M., Siomi, M. C., Nussbaum, R. L., and Dreyfuss, G. (1994). Essential role for KH domains in RNA binding: impaired RNA binding by a mutation in the KH domain of FMR1 that causes fragile X syndrome. Cell 77, 33–39. Siomi, H., Matunis, M. J., Michael, W. M., and Dreyfuss, G. (1993). The pre-mRNA binding K protein contains a novel evolutionarily conserved motif. Nucleic Acids Res. 21, 1193–1198. Suzuki, K., and Zagoren, J. C. (1976). Variations of Schmidt-Lanterman incisures in Quaking mouse. Brain Res. 106, 146–151. Suzuki, K., and Zagoren, J. C. (1977). Quaking mouse: An ultrastructural study of the peripheral nerves. J. Neurocytol. 6, 71–84. Tanaka, H., Abe, K., and Kim, C. H. (1997). Cloning and expression of the quaking gene in the zebraWsh embryo. Mech. Dev. 69, 209–213. Vernet, C., and Artzt, K. (1997). STAR, a gene family involved in signal transduction and activation of RNA. Trends Genet. 13, 479–484. Watanabe, I., and Bingle, G. J. (1972). Dysmyelination in ‘‘quaking’’ mouse. Electron microscopic study. J. Neuropathol. Exp. Neurol. 31, 352–369. Wisniewski, H., and Morell, P. (1971). Quaking mouse: ultrastructural evidence for arrest of myelinogenesis. Brain Res. 29, 63–73. Wu, H. Y., Dawson, M. R., Reynolds, R., and Hardy, R. J. (2001) Expression of QKI proteins and MAP1B identiWes actively myelinating oligodendrocytes in adult rat brain. Mol. Cell Neurosci. 17, 292–302.

III. THE MYELIN GENES AND PRODUCTS

OTHER ROLES FOR QKI PROTEINS

Wu, J. I., Reed R. B., Grabowski, P. J., and Artzt K. (2002). Function of quaking in myelination: regulation of alternative splicing. Proc. Natl. Acad. Sci. U. S. A 99, 4233–4238. Wu, J., Zhou, L., Tonissen, K., Tee, R., and Artzt, K. (1999). The quaking I-5 protein (QKI-5) has a novel nuclear localization signal and shuttles between the nucleus and the cytoplasm. J. Biol. Chem. 274, 29202–29210. ZaVron, S., Astier, M., Gratecos, D., and Semeriva, M. (1997). held out wings (how) Drosophila gene encodes a putative RNA-binding protein involved in the control of muscular and cardiac activity. Development 124, 2087–2098. Zhang, Y., and Feng, Y. (2001). Distinct molecular mechanisms lead to diminished myelin basic protein and 2’,3’cyclic nucleotide 3’-phosphodiesterase in qk(v) dysmyelination. J. Neurochem. 77, 165–172. Zorn, A. M., Grow, M., Patterson, K. D., Ebersole, T. A., Chen, Q., Artzt, K., and Krieg, P. A. (1997). Remarkable sequence conservation of transcripts encoding amphibian and mammalian homologues of quaking, a KH domain RNA-binding protein. Gene 188, 199–206. Zorn, A. M., and Krieg, P. A. (1997). The KH domain protein encoded by quaking functions as a dimer and is essential for notochord development in Xenopus embryos. Genes Dev. 11, 2176–2190.

III. THE MYELIN GENES AND PRODUCTS

659

C H A P T E R

28 The Leukodystrophies: Overview and Classification James M. Powers

DISORDERS OF CENTRAL WHITE MATTER: LEUKOENCEPHALOPATHIES, LEUKODYSTROPHIES DEMYELINATION, DYSMYELINATION ‘‘When I use a word,’’ Humpty Dumpty said in a rather scornful tone, ‘‘It means just what I choose it to mean. Neither more nor less.’’ Through the Looking Glass Lewis Carroll Myelin, the lipid-protein insulating coat of axons, is largely restricted to white matter of the central nervous system (CNS) and to larger axons of the peripheral nervous system (PNS). Therefore, diseases that speciWcally aVect myelin, often referred to as primary diseases of myelin (Figures 28.1A, B), by necessity aVect either CNS white matter, myelinated axons in the PNS, or both. The converse is not always true. That is, there are several disease processes that aVect CNS white matter, or less commonly myelinated PNS Wbers, but do not speciWcally or primarily aVect myelin or myelinating cells. These latter entities are in reality secondary or coincidental disorders of myelin, and their etiologies are diverse. One of the clearest examples of this secondary type of myelin loss, usually referred to as secondary demyelination, is due to a primary loss of axons from the death of its distant parent neuron (Wallerian or Wallerian-like degeneration) (Fig. 28.1C). Another is vascular disease, such as subcortical arteriosclerotic leukoencephalopathy (often referred to as ‘‘Binswanger disease’’), which is attributed to the small vessel lesions of chronic hypertension (Fig. 28.1D) (Babikian and Ropper, 1987). In ‘‘Binswanger disease,’’ there is a predominant destruction of cerebral white matter (Fig. 28.2) that varies from lacunar (small cystic) infarcts to concomitant and equivalent losses of myelin and axons to a preferential loss of myelin, in association with markedly hyalinized regional blood vessels (De Reuck et al., 1980). Another cerebral leukoencephalopathy presumed to be due to ischemia is its rare occurrence in amyloid angiopathy (Gray et al., 1985); here, however, the vascular lesions reside in amyloidotic leptomeningeal arteries/arterioles. Vascular lesions displaying the preferential loss of myelin share the major neuropathologic characteristic of primary diseases of myelin: loss of myelin and oligodendrocytes with relative sparing of axons; but the primary cause of this myelin loss is ischemia (vascular insuYciency) in which oligodendrocytes and their myelin sheaths are more aVected than their neighbors. The vulnerability of oligodendrocytes to oxygen deprivation, particularly in the human postnatal period when they are heavily involved in myelination, and their destruction by

Myelin Biology and Disorders, Volume 2

663

Copyright 2004, Elsevier (USA). All rights reserved.

664

28. THE LEUKODYSTROPHIES

White Matter Lesions Normal

A

oligodendrocyte blood vessel perivascular cell

neuron axon

myelin

Primary demyelination

B

x

C

Secondary demyelination

xx

D

Leukoencephalopathy

variable loss of myelin, axons and oligodendrocytes

x x FIGURE 28.1

Schematic diagrams of primary and secondary lesions of myelin. X identiWes the initial or primary lesion in B, C, and D.

DISORDERS OF CENTRAL WHITE MATTER

FIGURE 28.2 DiVuse atrophy with focal gray discoloration and cavitation of parietal white matter in ‘‘Binswanger disease.’’ (B) Hyalinized blood vessels within myelin-depleted white matter of ‘‘Binswanger disease.’’ Hematoxylin-eosin.

FIGURE 28.3 Bilateral symmetrical necrosis of frontal white matter with sparing of cortical gray in severe hypoxia-ischemia, such as CO poisoning.

hypoxic-ischemic insults are well documented (reviewed in Ludwin, 1997). The clinical presentation of such an hypoxic or ischemic destruction of white matter may occur cataclysmically, as in the delayed leukoencephalopathy of carbon monoxide (CO) poisoning (Fig. 28.3) (Grinker’s myelinopathy) (Plum et al., 1962), or more surreptitiously as in ‘‘Binswanger disease.’’ The latter’s chronic and progressive clinical course mimics that of primary diseases of myelin, in particular the major focus of this chapter: the leukodystrophies. CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy) also may exhibit a chronic and progressive clinical course, often dementing, due primarily to the destruction of cerebral white matter; it is familial and is associated with a defect in the Notch 3 gene (Joutel et al., 1996; Tournier-Lasserve et al., 1993). In view of its genetic causation, its progressive clinical course, and its predominant white matter lesion, CADASIL has been included by some under the rubric of ‘‘leukodystrophies.’’ In my opinion, this is inappropriate because CADASIL lacks the most important pathogenetic criterion of the leukodystrophies: the primary involvement of myelin sheaths/myelinating cells. Available evidence for the destruction of cerebral white matter in CADASIL indicates that it too is fundamentally ischemic and due to arterial abnormalities (Fig. 28.4) obvious at both light (granular medial myocytic degeneration) and electron (granular osmiophilic material, GOM) microscopic levels (Baudrimont et al., 1993). Cer-

665

666

28. THE LEUKODYSTROPHIES

FIGURE 28.4 Granular degeneration of medial myocytes in small arteries of CADASIL. Hematoxylin-eosin.

tain chemotherapeutic agents and illicit drugs or toxins, such as solvent vapor exposure, also can lead to conXuent losses of cerebral myelin (Kornfeld et al., 1994). ‘‘It was the best of times, it was the worst of times . . . it was the season of Light, it was the season of Darkness,’’ wrote Charles Dickens concerning England and France in 1775 (Dickens, 1894). In respect to our molecular understanding of genetic diseases, including those of myelin, the present is the best of times and a season of Light; one only has to witness the transformation of the single volume of Stanbury et al.’s The Metabolic Basis of Inherited Disease of 1983 to its four-volume heir, Scriver et al.’s The Metabolic and Molecular Bases of Inherited Disease of 2001. Many of the pathogenetic advances that have expanded this tome were derived from mouse mutants, both natural and genetically engineered, and by other powerful molecular methodologies. At the same time, it would proWt us to interpret these modern data within the context of a large repository of clinical and basic neuroscience, such as classical neuropathologic studies of the leukodystrophies (reviewed in Powers, 1996). While darkness no longer prevails, there still are many persistent glimmers in our current scientiWc world that are aggravated by the imprecise use of language. Sometimes a linguistic imprecision reXects a limited understanding of a scientiWc process, such as the diYculty in determining neuropathologically whether a developmental delay or a complete arrest in myelination is responsible for some cases of ‘‘hypomyelination,’’ but too often it is due to an ignorance of, or unwillingness to acknowledge, welldocumented scientiWc precedent. Enlightened scientiWc discourse can be facilitated by the persistent use of traditional terms and concepts, rather than their distortion or the fabrication of neologisms. As cases in point, consider the current designation of Alzheimer’s disease by some as an inXammatory disease, the same historical etiologic category as bacterial meningitis; or the term ‘‘intracellular amyloid’’ for a substance (amyloid) deWned in part for over a century by its extracellular localization; or the inclusion of CADASIL within the leukodystrophies. The latter decision ignores both the pathologic data mentioned above and its neuropathologic appellation: leukoencephalopathy. A more relevant, and pervasive, linguistic impropriety has occurred with the terms ‘‘demyelination’’ and ‘‘dysmyelination, which has led to much confusion. Some use ‘‘demyelination’’ to describe any loss of myelin staining in CNS white matter. Such linguistic promiscuity has been rightly condemned by Raine, one of the world’s experts in myelin diseases (discussed in Raine, 1997, p. 627, and the earlier editions 1991 and 1985); he also objects to the term ‘‘secondary demyelination’’ as used earlier. Moreover, both he and Prineas (Prineas et al., 2002, and the previous edition 1997), another leader in this Weld, equate ‘‘demyelinating’’ diseases with acquired autoimmune and suspected autoimmune diseases usually accompanied by perivascular demyelination/inXammation, in keeping with the historical precedents set by Adams and Kubik (1952) for demyelinating diseases and Poser (1962, 1968). Poser divided the primary diseases of myelin into two major types: myelinoclastic (currently referred to as demyelinating or demyelinative) and

au3

DISORDERS OF CENTRAL WHITE MATTER

au4

au5 au6

dysmyelinating (Poser, 1968). These two major types were established primarily on the basis of their diVering pathogeneses: acquired immune and inXammatory for demyelinating disorders versus an heredofamilial metabolic abnormality in myelin without inXammation for dysmyelinating disorders. To quote Poser: ‘‘The Wrst, the ‘myelinoclastic’ type, constitutes the true demyelinating diseases. . . . The other type, exempliWed by the leukodystrophies . . . groups together the ‘dysmyelinating’ diseases.’’ While Adams’s and Poser’s classiWcations were not universally accepted and had some Xaws (as do all classiWcation schemes), most found their concepts of demyelinating and dysmyelinating diseases to be rational and useful. To quote Raine: ‘‘ ‘Demyelination’ is a term carefully avoided by most authors in reference to the hereditary dysmyelinating diseases (metachromatic leukodystrophy, Krabbe’s disease, adrenoleukodystrophy, etc.). . . . Thus, there is currently consensus among neuropathologists that ‘demyelination’ should be a term reserved for the multiple sclerosis group of conditions.’’ (Raine, 1997, 1991, 1985). Would that life were that simple! Recently, two other good friends and colleagues, as preeminent in the neuroradiological elucidation of pediatric myelin disorders as Adams and Poser were in the neuropathologic elucidation of adult and pediatric myelin disorders, respectively, re-examined these concepts through a semantic eye; this approach led to dramatically diVerent deWnitions. Demyelinating disorders are . . . metachromatic leukodystrophy, multiple sclerosis. . . . Dysmyelination is, as the literal translation of the name implies, reserved for conditions in which the process of myelination is disturbed, leading to abnormal, patchy, irregular myelination, sometimes but not necessarily combined with signs of myelin breakdown. Examples: some amino acidopathies, damaged structure of unmyelinated white matter after perinatal hypoxia or encephalitis’’ (van der Knaap and Valk, 1995).

au7

au8

au9

au10

These latter deWnitions Xy in the face of the neuropathologic precedents mentioned previously and in particular those concepts proposed by Poser, whom they quote: ‘‘The dysmyelinating disorders comprise those disorders in which ‘myelin is not formed properly, or in which myelin formation is delayed or arrested, or in which the maintenance of already formed myelin is disturbed.’ Examples are metachromatic leukodystrophy and adrenoleukodystrophy.’’ Semantic considerations do have some merit, particularly when a conceptual framework is being established or Wne tuned. When they threaten or disassemble an established framework, however, they cause confusion and distract us from more important issues. For example, a reviewer of this chapter argued similarly that ‘‘dysmyelination is deWned as a process of defective myelin formation during development, while demyelination describes a process where myelin is formed correctly during development but is destroyed later. In this sense, ALD and MLD, for instance, should be demyelinating diseases.’’ The easiest way to respond to this latter criticism is again to cite the historical scientiWc precedents described earlier and perhaps speciWcally foreseen by Poser when he deWned dysmyelinating diseases: ‘‘or in which the maintenance of already formed myelin is disturbed.’’ Note that he does not say ‘‘attacked’’ or ‘‘destroyed,’’ hallmarks of the acquired immune ‘‘demyelinating’’ diseases. The words ‘‘maintenance’’ and ‘‘disturbed’’ reXect the metabolic imbalance in myelin that he believed was operative in these diseases (Poser, 1962, 1968). One could also play the semantic card and deWne ‘‘myelination’’ as fundamentally a metabolic (molecular, biochemical) process that is developmentally regulated; hence, ‘‘dysmyelination’’ refers more to a biochemical aberration in myelin (as Poser viewed it) than a developmental error. Putting this aside, I would argue that while the bulk of myelination in the human cerebrum is completed by 2 years postnatally, the morphologic data derived from Weigert (myelin) stained brain sections by Flechsig and Kaes ‘‘indicate that throughout the Wrst decade of life, myelination continues in the non-speciWc thalamic projections and in the association areas, and into adult life in the reticular formation and intracortical neuropil’’ (E. P. Richardson, Jr., 1982) and ‘‘Histologically, myelination reaches completion in early adulthood’’ (van der Knaap and Valk, 1995). The usual clinical onset of classical MLD, the late infantile form, is between the latter part of the Wrst through the second year of life, precisely the developmental period

667

668

28. THE LEUKODYSTROPHIES

au11

when cerebral myelination is in full swing; the myelin abnormality precedes the clinical onset. Moreover, the majority of the leukodystrophies have their onsets in infancy to childhood, but most also have juvenile- and adult-onset forms (e.g., MLD, GLD, ALD). Are these later-onset forms to be considered a diVerent class of myelin disease? Finally, myelin is not an inert substance and its biochemical components undergo a regular pattern of turnover throughout life. Is not this remodeling part of the biological process of myelination? When exactly does the ‘‘developmental’’ period of human myelination end? Who can say, and does it really matter? Such queries border more on the philosophical than the scientiWc. In ALD I believe that the myelin only becomes unstable when a suYcient amount of its abnormal fatty acid becomes incorporated into the myelin lipids and proteolipid protein. In most ALD males, this occurs at around 5 to 9 years; perhaps in the later onsets this is a slower process. In the Wnal analysis, I personally Wnd the logic and classiWcation schemes for human demyelinating and dysmyelinating diseases initiated by Adams, Kubik and Poser, while imperfect, to be prescient and far more cogent than the others. Thus, I fully endorse Poser’s deWnition, oVered in 1957, and quoted earlier: Dysmyelinating diseases are those heredofamilial disorders in which myelin is not formed properly, or in which myelin formation is delayed or arrested, or in which the maintenance of already formed myelin is disturbed. These deWnitional categories can be expanded or reWned to include new diseases and more scientiWcally valid data. For example, the pathogenetic logic behind Poser’s ‘‘dysmyelination’’ emanated from the scientiWc wisdom of his day: anabolic enzyme defects due to inborn errors of metabolism a` la Archibald Garrod. Today, we recognize the shortcomings of his pathogenetic formulation, but this does not diminish the value of the classiWcation. The term leukodystrophy (leuko-white; dystroph—defective nutrition) (Bielschowsky and Henneberg, 1928), on the other hand, has traditionally been used for genetically determined, hence usually familial, and clinically progressive disorders that primarily aVect CNS myelin. While all leukodystrophies are ‘‘dysmyelinating’’ diseases, the converse is not always true. Out of respect for historical precedent and for the sake of this present discussion leukodystrophies should have (1) a known or presumptive genetic causation, (2) a progressive clinical course, (3) a predominant and usually conXuent involvement of CNS white matter, and (4) a primary lesion of myelin or myelinating cells. The latter may be manifested by either a loss or failed development of CNS white matter due to a biochemical abnormality in myelin or a molecular abnormality in myelinating cells. Those other white matter lesions that lack at least one of these diagnostic attributes can be referred to as leukoencephalopathies. This deWnition of a leukodystrophy may seem too restrictive to some, but it has historical precedent (brieXy reviewed in van der Knaap and Valk, 1995, pp. 14–15) and the diagnosis of ‘‘leukodystrophy’’ has genetic and prognostic implications. Such a restricted use of ‘‘dystrophy’’ is not limited to diseases of CNS white matter. An analogous situation has existed in diseases of skeletal muscle with the terms ‘‘muscular dystrophy’’ and myopathy. In both dystrophic (i.e., myelin and muscle) situations, a primary genetic causation and somewhat predictable, often severe, clinical progression are characteristic. The remainder of this chapter, therefore, will be conWned to primary diseases of myelin and, in particular, to the major leukodystrophies (Table 28.1). It should be emphasized that Table 28.1, based on neuropathologic data, is neither all inclusive nor written in stone, but rather is an evolving process; some white matter diseases, at least presently, do not ‘‘Wt’’ well or have debatable placements (e.g., cerebrotendinous xanthomatosis). Not all primary diseases of myelin are leukodystrophies, and some are referred to as leukoencephalopathies. For example, progressive multifocal leukoencephalopathy does not Wt well into the demyelinative category and fulWlls several criteria of a leukodystrophy. The JC papova virus directly infects and lyses oligodendrocytes resulting usually in a multifocal loss of CNS myelin (Fig. 28.5A) with relative sparing of axons, but usually without inXammation; it also is a progressive illness and may be conXuent, such as in AIDS (Fig. 28.5B). However, it lacks the genetic element of a leukodystrophy. Some diseases also referred to as leukoencephalopathies (e.g., vacuolating megalencephalic leukoencephalopathy with subcortical cysts) have been assigned provisionally to the dysmyelinative

DISORDERS OF CENTRAL WHITE MATTER

TABLE 28.1

669

Primary Diseases of Myelin

I. DYSMYELINATING DISEASES (LEUKODYSTROPHIES) Classical Dysmyelinative Adrenoleukodystrophy (MIM 300100) Metachromatic leukodystrophy (MIM 250100) Globoid cell leukodystrophy (Krabbe’s disease, MIM 245200) Sudanophilic (orthochromatic) leukodystrophies (MIM 272100) Simple Type Pigmentary Type With meningeal angiomatosis Polycystic lipomembranous osteodysplasia with sclerosing leukodystrophy (membranous lipodystrophy, Nasu-Hakola, MIM 221770) Neuroaxonal leukodystrophy, hereditary diffuse leukoencephalopathy with spheroids; autosomal dominant diffuse leukoencephalopathy with neuroaxonal spheroids Sjogren-Larsson (MIM 270200) Others Hypomyelinative Pelizaeus-Merzbacher disease (MIM 312080) Alexander disease (MIM 203450) Vanishing white matter disease/childhood ataxia with diffuse cerebral hypomyelination (MIM 603896) Aicardi-Goutie`res syndrome (MIM 225750) Cockayne syndrome (MIM 216400) Spongiform Spongy degeneration of central nervous system (Canavan or VanBogaert-Bertrand disease, MIM 271900) Adult-onset spongiform leukodystrophy (MIM 169500) Vacuolating megalencephalic leukoencephalopathy with subcortical cysts (Infantile-onset spongiform leukoencephalopathy, MIM 604004)

II. MYELINOLYTIC DISEASES (SPONGY MYELINOPATHIES) Central pontine myelinolysis Aminoacidurias, Organic acidurias Mitochondrial disorders: Kearns-Sayre syndrome (MIM 530000) Vitamin B12 (folate) deficiency and HIV vacuolar myelopathy Exogenous toxins: Heroin vapor, hexachlorophene and triethyl tin

III. DEMYELINATING DISEASES (CNS) Multiple (disseminated) sclerosis (MIM 126200) 1. Chronic a. Classical (Charcot) b. Diffuse cerebral variant (Schilder) 2. Acute variants a. Disseminated (Marburg) b. Concentric sclerosis (Balo) b. Neuromyelitis optica (Devic) Acute disseminated encephalomyelitis 1. Classical a. Postinfectious encephalomyelitis b. Postvaccinal encephalomyelitis 2. Hyperacute a. Acute hemorrhagic leukoencephalitis (Hurst). Focal inflammatory demyelinating lesions with mass effect (Pseudotumor) Infectious 1. Progressive multifocal leukoencephalopathy 2. Subacute sclerosing panencephalitis

670

28. THE LEUKODYSTROPHIES

FIGURE 28.5 (A) Multifocal loss of myelin in PML. Heidenhain myelin. (B) ConXuent to cavitary loss of frontal white matter in PML of AIDS patient.

disease (leukodystrophy) category, because they have shown an apparent primary lesion of myelin (van der Knaap et al., 1996), in addition to the other criteria mentioned earlier. Several other familial and progressive white matter disorders have been placed provisionally in the hypomyelinative group in spite of their controversial or unknown pathogeneses: vanishing white matter disease (VWM) (van der Knaap, et al. 1998), Aicardi-Goutie`res syndrome (Razavi et al., 1988), and Cockayne syndrome (reviewed in Nance and Berry, 1992), because of their similarity to Alexander disease and to the prototypic hypomyelinative leukodystrophy, Pelizaeus-Merzbacher disease (PMD), respectively. The latter calcifying disorders mimic the hypomyelinative lesions of PMD, while the cavitating quality of VWM approximates Alexander disease. Reviewing the morphology of myelin degradation and comparing and contrasting the neuropathologic features of its major types to those of other primary diseases of myelin and to each other can provide us with a neuropathologic overview of the leukodystrophies.

MYELIN DEGRADATION Myelin breakdown is a dynamic morphologic process in which speciWc cells participate in characteristic patterns for a particular disease, and the biochemical degradation of myelin can be appreciated with traditional carbohydrate and lipid stains (Adams, 1965). The galactolipids, cerebroside and sulfatide, and cholesterol are major lipid components of myelin sheaths that are liberated during myelin breakdown. Cerebroside and sulfatide

PRIMARY DISEASES OF MYELIN

contain 1, 2-glycol groups and hence are periodic acid-SchiV (PAS) positive. Sulfatide (cerebroside sulfate) has the additional property of metachromasia due to its anionic sulfate groups. A substance is metachromatic when it can produce a spectral shift, usually toward longer wavelengths such as blue to red, when stained with some basic dye (e.g., toluidine blue or cresyl violet). When an individual has normal lysosomal galactocerebrosidase and arylsulfatase A activities, the respective degradative enzymes for these substrates, the catabolic reactions are rapid and their morphologic correlates are of short duration. However, when either of these degradative enzymes is deWcient, such as in globoid cell leukodystrophy (Krabbe disease; GLD) or metachromatic leukodystrophy (MLD), the PAS positivity of the nondegraded galactocerebroside or sulfatide and the metachromasia of sulfatide persist. Cholesterol is esteriWed primarily by macrophages that become vacuolated and have been referred to historically as gitter cells, compound granular corpuscles, or lipid-laden macrophages (lipophages). Normally the intracellular esteriWed cholesterol persists much longer than the galactolipids or their degradation products. Therefore, the major degradative end point of myelin in an individual with biochemically normal myelin and normal lysosomal enzymes is cholesterol ester, the neurochemist’s ‘‘Xoating fraction’’ (Norton et al., 1966). It is important to note that the histochemical or biochemical detection of these degradative products and particularly cholesterol esters in white matter is not restricted to primary diseases of myelin and may be seen in a variety of lesions, such as an infarct. Cholesterol esters also are detectable when myelin is being laid down by oligodendrocytes (Ramsay and Davison, 1974). Cholesterol esters within macrophages can be demonstrated with ‘‘Sudan’’ and other neutral lipid dyes, such as Oil Red O, in frozen sections. Hence, the end point of normal myelin degradation is referred to as sudanophilic; it is also said to be orthochromatic, because cholesterol ester does not possess metachromatic properties. Sudanophilia is typical of the major human demyelinative disease, multiple sclerosis. However, sudanophilic myelin debris also occurs in adrenoleukodystrophy (ALD), where biochemical abnormalities in both myelin lipids and proteolipid protein (PLP) have been clearly demonstrated. Comparable, but currently unrecognized, biochemical abnormalities in myelin probably underlie at least some of the ‘‘sudanophilic’’ (orthochromatic) leukodystrophies (SLD), the category in which ALD had been placed until the identiWcation of its biochemical abnormality. While these histochemical reactions (Wolman, 1970) are unfolding in the classical dysmyelinative (leukodystrophies) and demyelinative diseases, myelin sheaths are generally undergoing some loss of stainability, vacuolation, and fragmentation prior to and in concert with a variable macrophage inWltration and astrocytic hypertrophy/hyperplasia eventuating in severe myelin loss, relative axonal sparing, decreased numbers of oligodendrocytes, few remaining lipophages, and chronic Wbrillary astrogliosis.

PRIMARY DISEASES OF MYELIN As discussed earlier, primary diseases of central myelin have been divided into two major types: myelinoclastic (demyelinating/demyelinative) and dysmyelinating/dysmyelinative (Poser, 1962, 1968), and they are both characterized by a primary loss of myelin with a relative sparing of axons. Neuropathologically, this is conWrmed by staining two serial sections of a white matter lesion for myelin or axons and demonstrating a loss of myelin in one and a sparing of axons in the corresponding area of the other (Fig. 28.6). A third category was added by the author, in part because those diseases shared a common light microscopic feature of spongy to vacuolated myelin (Fig. 28.7); we designated them as myelinolytic diseases or spongy myelinopathies (Powers and Horoupian, 1995) in keeping with earlier reports. The vacuoles are due to the accumulation of Xuid within myelin sheaths and astrocytes; in the former they usually originate at the intraperiod line that is continuous potentially with the extracellular space (Jellinger and Seitelberger, 1970). It is diYcult to determine the exact site of splits in myelin sheaths in human biopsy or autopsy specimens due to poor tissue preservation. Intramyelinic edema is considered a subtype of

671

672

28. THE LEUKODYSTROPHIES

FIGURE 28.6 (A) Primary demyelination. Complete loss of myelin, reactive astrocytosis, and perivascular inXammation. Luxol fast blue—PAS myelin. (B) Slight to moderate loss of axons in the identical Weld of a serial section. Bodian axon.

FIGURE 28.7 Spongy myelinopathy of cerebral white matter with good preservation of oligodendrocytes in Kearns-Sayre syndrome. Hematoxylin-eosin.

cytotoxic (intracellular) edema (Klatzo, 1967). Myelinolytic diseases may be either genetically transmitted or acquired; they are etiologically diverse but most often toxic or metabolic (e.g., vitamin B12 deWciency). In many myelinolytic lesions, axonal and oligodendroglial sparing is characteristic, at least of the early stages, and inXammatory cells including macrophages do not participate to any appreciable degree; myelin debris is also

PRIMARY DISEASES OF MYELIN

sparse but is sudanophilic. Astrocytosis varies, but usually occurs. As a result and most important, the spongy myelinopathies diVer from other primary diseases of myelin, particularly leukodystrophies, in that some are reversible, such as central pontine myelinolysis (Wakui et al., 1991) and hexachlorophene toxicity (Kimbrough and Gaines, 1971). In some myelinolytic disorders, soluble toxins may be acting directly on myelin sheaths; in others, osmotic factors may be responsible (osmotic myelinolysis); yet in others, spongy myelin may reXect a potentially correctable metabolic dysfunction of the supporting oligodendrocytes/astrocytes. The myelin appears to be biochemically normal on the basis of its sudanophilia and biochemical analyses (Cammer et al., 1975). PNS lesions are usually absent in myelinolytic diseases. One could include Canavan disease in this category because of its vacuolated myelin, a meager myelin debris-macrophage response, and the preferential involvement of subcortical arcuate Wbers (Adachi et al., 1973); however, its genetic defect, infantile onset, progressive clinical course, and conXuent white matter lesions align it more closely to the leukodystrophies. Some aminoacidurias, organic acidurias, heritable mitochondrial disorders, and other inherited vacuolar leukoencephalopathies could fulWll the diagnostic criteria of leukodystrophies but have not yet been accorded this position. In demyelinative diseases, such as multiple sclerosis, inXammatory cells are prominent (Fig. 28.8) and destroy biochemically normal myelin or myelinating cells (oligodendrocytes/Schwann cells). Demyelinative diseases are typically acquired, even though genetic factors also may play a role in some; their clinical progression tends to be more variable than that of the leukodystrophies. Considerable axonal sparing in the early phase of the chronic types has traditionally distinguished demyelinative diseases from the leukodystrophies (Figures 28.9A, B), but the recent demonstration of early and progressive loss of axons in the demyelinative plaques of multiple sclerosis vitiates this distinction (Trapp et al., 1998). Myelin breakdown products are sudanophilic and orthochromatic. Demyelinative lesions usually do not respect the subcortical arcuate or ‘‘U’’ Wbers, have sharp edges (Fig. 28.9A) and are usually asymmetric when bilateral. Fibrillary astrogliosis with markedly diminished myelin and oligodendrocytes (Fig. 28.9C) is the ultimate outcome of the demyelinative plaque. The prominent participation of inXammatory cells (T and B lymphocytes, macrophages, and plasma cells) and their products (cytokines and chemokines) provide convincing evidence of an immune destruction of myelin that appears to be directed against antigenic myelin proteins. Some of these proteins reside in either the CNS or PNS (e.g., the CNS PLP), and therefore the involvement of either central or peripheral myelin is typical of most demyelinative diseases. As mentioned previously, leukodystrophies are genetically transmitted and progressive diseases in which either biochemically abnormal myelin or some molecular abnormality in

FIGURE 28.8 Prominent lymphocytic perivascular cuV in demyelinated white matter exhibiting a loss of oligodendrocytes and reactive astrocytosis in multiple sclerosis. Hematoxylin-eosin.

673

674

28. THE LEUKODYSTROPHIES

FIGURE 28.9 (A) Demyelinative plaque with sharp border in multiple sclerosis. Heidenhain myelin. (B) Axonal sparing in demyelinative plaque. Serial section of same lesion as in Figure 28.9a. Bodian axon. (C) Isomorphic Wbrillary astrogliosis of chronic demyelinative plaque in multiple sclerosis. Hematoxylin-eosin.

myelinating cells has been identiWed. They have been classiWed as dysmyelinative to set them apart from demyelinative diseases. Such classical leukodystrophies include ALD, MLD, GLD, and SLD. Their biochemical/molecular defects involve myelin lipids, which are qualitatively similar in CNS and PNS; hence, the involvement of both central and peripheral myelin is typical of most classical leukodystrophies. ALD, in spite of its profound lymphocytic element that is typical of demyelinative diseases, has a Wrm nosological position within the dysmyelinative diseases due to the incorporation of abnormal very long chain saturated fatty acids (VLCFA) into several myelin components (reviewed in Moser, 1997). The second subgroup of dysmyelinative diseases, hypomyelinative, diVers

LEUKODYSTROPHIES: CLINICAL SIMILARITIES AND DISSIMILARITIES

from the classical leukodystrophies in that their primary defect appears to relate more to inadequate myelinogenesis rather than myelin breakdown. At present classical/connatal X-linked PMD is the most legitimate member of this divergent group, and GLD also could call this its home. In PMD and its allelic cousin, spastic paraplegia type 2 (SP2), the molecular defect involves the PLP gene and the variable absence of myelin PLP; hence, the lesions are restricted to the CNS for all practical purposes, as is true for the others. The third subgroup, spongiform, displays the same spongy to vacuolated change in myelin that epitomizes the myelinolytic diseases (Fig. 28.7), but they are genetically determined and their primary locus is cerebral white matter that displays conXuent abnormalities. PNS lesions also are not typical of spongiform leukodystrophies. One disease that doesn’t Wt well into this classiWcation scheme is VWM, because it exhibits a mixed picture of hypomyelination, classical dysmyelination with some myelin breakdown and marked axonal loss, as well as spongiform changes.

LEUKODYSTROPHIES: CLINICAL SIMILARITIES AND DISSIMILARITIES

au12

Most leukodystrophies have similar clinical Wndings that reXect damage to central white matter: abnormal motor function, vision, hearing, and cognition. The speciWc manifestations of these system abnormalities can vary somewhat, depending on the age of the patient at the time of onset. Nevertheless, developmental delay or regression, spasticity and hypertonia or hypotonia, quadriparesis to quadriplegia, decerebrate posturing, visual and auditory agnosia, decreased visual or auditory acuity to blindness and deafness, ataxia, nystagmus, other abnormal movements, and mental retardation to dementia can be detected in patients with a leukodystrophy. Genetic transmission patterns can be autosomal recessive (e.g., MLD, 22q; GLD, 14q; Canavan disease, 17p; VWM, 3q), autosomal dominant (adult spongiform leukodystrophy, 5q) and X-linked (ALD, PMD); some are more typically sporadic (Alexander disease, SLD). The typical age of onset is also variable: infancy (GLD, Canavan disease, Alexander disease), late infancy (MLD, PMD), childhood (ALD, VWM), and adult (AMN, SLD, SP2).

LEUKODYSTROPHIES: NEUROPATHOLOGIC SIMILARITIES AND DISSIMILARITIES Leukodystrophies generally exhibit similar gross neuropathologic features: reduced brain size, except for the megalencephaly of Canavan and Alexander diseases, optic atrophy, ventriculomegaly, atrophy of the corpus callosum, and bilaterally symmetrical, diVuse to conXuent, loss or lack of cerebral and cerebellar white matter with their replacement by Wrm gray to beige Wbrillary astrogliosis (sclerosis) (Fig. 28.10). A patchy or tigroid myelinopathy, due to the perivascular presence of myelin, typiWes PMD (Fig. 28.11) (Seitelberger, 1995); but also Cockayne syndrome (Leech et al., 1985) and adult-onset spongiform leukodystrophy (Fig. 28.12) (Eldridge et al., 1984). Brainstem lesions in Alexander disease often simulate the demyelinative plaques of multiple sclerosis. Preservation of subcortical ‘‘U’’ Wbers is characteristic, except in Canavan disease (Fig. 28.13) and PMD, but severe or protracted courses can lead to their loss in the others (Fig. 28.14). Asymmetry is seen in the advancing edges of ALD, usually frontal (Schaumburg et al., 1975). Cavitation of white matter due to massive axonal loss and an inadequate astrocytic reparative response rarely occurs in the classical leukodystrophies, but is characteristic of Alexander disease (Borrett and Becker, 1985) and VWM (van der Knaap et al., 1997) (Fig. 28.15). The parieto-occipital lobes in ALD and late-onset GLD, the frontal lobes in Alexander disease and several late-onset leukodystrophies such as MLD, and the arcuate Wbers in Canavan disease are favored early sites. Gray matter is usually not involved, except when there has been superimposed hypoxic-ischemic events

675

676

28. THE LEUKODYSTROPHIES

FIGURE 28.10 ConXuent loss of myelin with sparing of arcuate Wbers and replacement by gray-tan astrocytic tissue in frontal white matter of ALD.

FIGURE 28.11 Absence of myelin staining, including arcuate Wbers, except for small circular foci around blood vessels in PMD. Weil myelin.

or when cavitation of white matter leads to direct or transynaptic atrophy of its associated gray matter. These same patterns of gross neuropathologic lesions now can be appreciated in living patients with modern imaging techniques, such as magnetic resonance imaging (MRI) (van der Knaap and Valk, 1995). T2-weighted images are particularly eVective in demonstrating myelin abnormalities (Fig. 28.16). An MRI-pattern recognition approach has been developed and applied to unclassiWed white matter diseases in children with great success (van der Knaap et al., 1999). MR spectroscopy (MRS), in which speciWc biochemical moieties within white matter lesions can be measured, also can contribute to a clinical diagnosis. Perhaps more important, MRS along with the recent modiWcations of magnetization transfer and diVusion anisotropy, especially when utilized in longitudinal studies of living patients and correlated with postmortem imaging and neuropathologic analysis,

LEUKODYSTROPHIES: NEUROPATHOLOGIC SIMILARITIES AND DISSIMILARITIES

FIGURE 28.12 ConXuent, but patchy, loss of myelin in frontal white matter of adult-onset spongiform leukodystrophy.

FIGURE 28.13 Mucoid appearance of arcuate Wbers in swollen frontal white matter of Canavan disease.

should provide powerful pathogenetic insights (e.g., in ‘‘hypomyelination’’). (The reader is referred to van der Knaap, 2001, and van der Knaap and Valk, 1995, for more speciWc details.) Likewise, with the light microscope most leukodystrophies display common features: reduced myelin staining, loss of oligodendrocytes, considerable axonal loss but still a

677

678

28. THE LEUKODYSTROPHIES

FIGURE 28.14 ConXuent loss of myelin, including arcuate Wbers, in occipital white matter of same ALD patient depicted in Figure 28.10.

FIGURE 28.15 Cavitation of frontal white matter in Alexander disease.

relative sparing of axons and reactive astrocytosis in the early stages (Fig. 28.17A) to Wbrillary astrogliosis that may be either isomorphic or anisomorphic in later stages (Fig. 28.17B). Macrophages with myelin debris are typical of the classical types but tend to be sparse in the hypomyelinative and spongiform subgroups, such as PMD and Canavan disease. Axonal loss is greater in the leukodystrophies than in the demyelinative diseases

LEUKODYSTROPHIES: NEUROPATHOLOGIC SIMILARITIES AND DISSIMILARITIES

FIGURE 28.16 ConXuent, bilaterally symmetrical high signal abnormalities of parieto-occipital white matter in ALD. T2-MRI.

FIGURE 28.17 (A) Marked loss of myelin, axons and oligodendrocytes with reactive astrocytosis, typical of most classical leukodystrophies. Such perivascular lymphocytes are essentially restricted to ALD. Hematoxylin-eosin. (B) Fibrillary astrogliosis of demyelinated cerebral white matter in ALD. Hematoxylin-eosin.

679

680

28. THE LEUKODYSTROPHIES

and is particularly prominent in the cavitating lesions of VWM and Alexander disease, as well as the posterior limb of the internal capsule in ALD and GLD, characteristically leading to secondary or Wallerian-like corticospinal tract degeneration (Fig. 28.18). Other than macrophages, traditional inXammatory cells (e.g., lymphocytes) are usually inconspicuous, except in ALD (Fig. 28.19), to a mild degree in GLD, and in some early lesions of Alexander disease. In spite of the similarities, signiWcant diVerences between the leukodystrophies can be seen with both the light and electron microscopes. For example, even with a routine hematoxylin-eosin (H-E) stain, the macrophages are often distinctive because the nature and amount of myelin breakdown products are characteristic of each leukodystrophy. In the classical leukodystrophies, myelin debris is substantial during the active phase. Lipophages tend to be vacuolated in SLD, except in the pigmentary form of SLD where they are yellow-brown and granular, vacuolated and striated in ALD, coarsely granular in MLD, and epithelioid to multinucleated (globoid) in GLD (Fig. 28.20). The myelin debris is predominantly sudanophilic and orthochromatic in ALD and SLD, acid fast/PASþ, autoXuorescent and variably iron positive in pigmentary SLD, metachromatic and PASþ in MLD or PASþ, and orthochromatic in GLD (Fig. 28.21). Macrophages are diastase resistant, PASþ; so the macrophages in ALD and SLD also can be PASþ (Fig. 28.22), but the more striated the cytoplasm in ALD the less sudanophilic. The myelin debris also displays a highly characteristic and often diagnostic ultrastructural appearance:

FIGURE 28.18. Bilateral, secondary corticospinal tract degeneration in basis pontis of ALD. Luxol fast blue—eosin.

LEUKODYSTROPHIES: PATHOGENETIC SIMILARITIES AND DISSIMILARITIES

FIGURE 28.19 Prominent lymphocytic inWltrates characteristic of ALD and comparable to those of MS. Hematoxylin-eosin.

multangular crystalloids in GLD, prismatic and tuVstone bodies in MLD and lamellae and lamellar-lipid proWles in ALD. Abnormal mitochondria in Alzheimer type II astrocytes are seen in Canavan disease (Fig. 28.23), and proteinaceous Rosenthal Wbers with their granular precursors in astrocytes are the hallmark of Alexander disease (Fig. 28.24). An apparent increase in the number of oligodendrocytes in the early lesions and ‘‘foamy’’ oligodendrocytes in later stages have been reported in VWM. Small foci of mineralization of the abnormal white matter have been noted in several leukodystrophies (Fig. 28.25) but are prominent and more widespread in Cockayne and Aicardi-Goutiere`s syndromes. While the involvement of CNS white matter is constant, the association of other pathologic lesions is characteristic of speciWc leukodystrophies. The PNS is commonly involved in infantile GLD, late infantile MLD, and the adult ALD variant, adrenomyeloneuropathy-AMN. Concomitant neuronal storage (neurolipidosis) in subcortical sites is seen only in MLD and its variant, multiple sulfatase deWciency (mucosulfatidosis) (Fig. 28.26). Neuronal loss of the dentate nuclei, brainstem, and thalamus seems characteristic of GLD, while neuronal loss in cerebellar cortex may be conspicuous in MLD. The mineralizations mentioned earlier in Cockayne and Aicardi-Goutiere`s syndromes also involve gray matter, particularly the deep gray matter. Extraneural lesions are typical of ALD (adrenocortical and Leydig cells) (Fig. 28.27) and MLD (renal tubules and other epithelial cells) (Fig. 28.28). (See Harding and Surtees, 2002, Powers and DeVivo, 2002, and Suzuki and Suzuki, 2002, for further details.) It is noteworthy that adult variants of some leukodystrophies are characterized by bilaterally symmetrical tract degenerations, such as of the optic radiations, implying a primary axonal problem rather than the dysmyelination of the pediatric patients. This is well documented in AMN where several supraspinal tracts (e.g., optic radiations and medial lemnisci) have been shown to have equivalent losses of axons and myelin, while a dominant dying-back axonopathy of the gracile, corticospinal, and spinocerebellar tracts (Fig. 28.29) is associated with atrophic dorsal root ganglion neurons containing lipidic mitochondrial inclusions (Powers et al., 2000; Powers et al., 2001; Schaumburg et al., 1977). Similar tract lesions can be seen in juvenile-adult Alexander disease (Fig. 28.30) (personal observation), adult-onset GLD (Choi et al., 1991), and perhaps SP2.

LEUKODYSTROPHIES: PATHOGENETIC SIMILARITIES AND DISSIMILARITIES In contrast to the common elements of their gross and microscopic neuropathologic lesions, but comparable to their distinctive morphologic features, each leukodystrophy is unique in

681

682

28. THE LEUKODYSTROPHIES

FIGURE 28.20 Macrophages. (A) Smooth to vacuolated macrophages in ALD; (B) brown macrophages in pigmentary form of SLD; (C) large striated macrophage adjacent to blood vessel in ALD; (D) coarsely granular macrophages in MLD; (E) clusters of globoid cells with single and multiple nuclei, often around blood vessels, in GLD. Hematoxylin-eosin.

its biochemical/molecular defect and, presumably, pathogenesis. In most leukodystrophies, only glimpses of their pathogenetic mechanisms exist. Genotype-phenotype correlations among the leukodystrophies are highly variable: from excellent in PMD (Hudson, 2001), to good in MLD and GLD with certain common homozygous mutations (D. Wenger, personal communication), to poor in ALD/AMN (Smith et al., 1999). In spite of this season of ‘‘Light,’’ more needs to be learned about most leukodystrophies than is known. For example, some consider the primary pathogenetic problem in VWM as glial, while others

LEUKODYSTROPHIES: PATHOGENETIC SIMILARITIES AND DISSIMILARITIES

FIGURE 28.21 Myelin debris. (A) Red (sudanophilic) macrophages in ALD. frozen section, oil red 0; (B) deep red-violet macrophages in pigmentary SLD, acid fast; (C) brown-yellow metachromasia of sulfatide within macrophages of MLD, frozen section, acid-cresyl violet; (D) magenta globoid cells of GLD, PAS.

FIGURE 28.22 PAS positive macrophages around blood vessel in ALD. LFB-PAS.

see it as axonal (van der Knaap et al., 1998). How the novel and revolutionary discovery of a defective translation initiation factor gene relates to this exclusively CNS white matter disorder also is a mystery at present (Leegwater et al., 2001). Speaking for myself, I have been trying to understand the pathogenesis of ALD and AMN for over three decades and am still in the dark, even after the identiWcation of the genetic defect on the X chromosome.

683

684

28. THE LEUKODYSTROPHIES

FIGURE 28.23 Ultrastructure of myelin debris. (A) Crystalloids of GLD; (B) prismatic structures of MLD; (C) lamellae and lamellar-lipid proWles of ALD; (D) abnormal mitochondria of Canavan disease.

FIGURE 28.24 Countless eosinophilic Rosenthal Wbers in white matter of infantile Alexander disease. Hematoxylin-eosin.

Usually the primary molecular abnormality in a leukodystrophy has been demonstrated, or assumed to be, in the oligodendrocyte, except in Alexander disease with its mutations of the astrocytic glial Wbrillary acidic protein gene (Brenner et al., 2001). An astrocytic and perhaps mitochondrial participation, secondary at least, may also occur in the pathogenesis of Canavan disease (Jellinger and Seitelberger, 1970) that is caused by aspartoacylase deWciency (Matalon et al., 1988). Both in MLD and GLD, the enzymatic

LEUKODYSTROPHIES: PATHOGENETIC SIMILARITIES AND DISSIMILARITIES

FIGURE 28.25 Round to elliptical blue mineralizations in gliotic white matter devoid of myelin and oligodendrocytes in ALD. Hematoxylin-eosin.

FIGURE 28.26 Neuronal storage in dentate nucleus devoid of myelin in MLD. Hematoxylin-eosin.

FIGURE 28.27 Pale striated and ballooned adrenocortical cells (left side of Weld) adjacent to normal eosinophilic counterparts in ALD.

685

686

28. THE LEUKODYSTROPHIES

FIGURE 28.28 Brown sulfatide in biliary epithelium of MLD. Frozen section, Acid-cresyl violet.

FIGURE 28.29 Comparable loss of axons (NF) and myelin (PLP) in gracile and corticospinal tracts of cervical spinal cord in AMN. Immunostains to neuroWlament (NF) and proteolipid (PLP) proteins.

defect resides in the lysosome, arylsulfatase A in MLD and galactosylceramidase (galactocerebrosidase) in GLD. There is evidence that the excessive sulfatide in myelin destabilizes the sheath and leads to its breakdown in MLD (Ginsberg and Gershfeld, 1991), but a toxic eVect on oligodendrocytes by lysosulfatide also has been proposed. A comparable toxic pathogenetic mechanism has considerable support in GLD, where galactosylsphingosine (psychosine) has been implicated in the death of oligodendrocytes and abortive myelination (Miyatake and Suzuki, 1972). The myelin instability model also has been suggested in the later onset and milder cases of PMD and its allelic SP2, whereas a toxic eVect of mutant PLP or its alternatively spliced isoform DM20 trapped in dilated cisterns

LEUKODYSTROPHIES: PATHOGENETIC SIMILARITIES AND DISSIMILARITIES

FIGURE 28.30 Similar degeneration of medial gracile and corticospinal tracts in late-onset Alexander disease evident with both (A) myelin and (B) axon stains. (A) Luxol fast blue—PAS myelin, (B) Bodian axon.

of rough endoplasmic reticulum has been proposed to cause the apoptotic death of oligodendrocytes with resultant hypomyelination in connatal and classical PMD (reviewed in Hudson, 2001). The pathogeneses of SLDs remain unknown, but the abundance of ceroid-lipofuscin and the presence of iron in the glia of its pigmentary form (Gray et al., 1987) may indicate an oxidative insult. The peroxisomal (and perhaps mitochondrial) disease ALD/AMN, on the other hand, has been considered to have a more complicated two-stage pathogenetic mechanism: dysmyelination followed by inXammatory demyelination. The dysmyelinative process was originally considered to result from a toxic eVect of VLCFA on oligodendrocytes, but currently a myelin instability caused by the incorporation of VLCFA into myelin lipids, particularly phosphatidylcholine and gangliosides, and into the major myelin protein PLP is favored (Ho et al., 1995; Powers et al., 1992; Powers et al., 2000). Subsequently, a profound and rapid demyelinative (immune) destruction of myelin supervenes, in which T cells (particularly CD8 cytotoxic T cells), reactive astrocytes, macrophages, nitric oxide, TNF-a, and CD1-mediated lipid antigen presentation participate (Ito et al., 2001; Powers et al., 1992). (See speciWc chapters in this volume and the relevant chapters in Scriver et al., 2001, for further details.)

Acknowledgments

ed13

The author thanks Tina Blazey for her usual outstanding secretarial assistance, both Jenny Smith and Nancy Dimmick for their artistic and photographic expertise, and Professor Marjo van der Knaap for Figure 28.16.

687

688

28. THE LEUKODYSTROPHIES

References Adachi, M., Schneck, L., Cara, J., and Volk, B. W. (1973). Spongy degeneration of the central nervous system (Van Bogaert and Bertrand type: Canavan disease). A review. Hum. Pathol. 4, 331–347. Adams, C. W. M. (1965). ‘‘Neurohistochemistry.’’ pp. 437–517, Elsevier, Amsterdam. Adams, R. D., and Kubik, C. S. (1952). The morbid anatomy of the demyelinative diseases. Am. J. Medicine 12, 510–546. Babikian, V., and Ropper, A. H. (1987). Binswanger’s disease. A review. Stroke 18, 2–12. Baudrimont, M., Dubas, F., Joutel, A., Tournier-Lasserve, E., and Bousser, M. G. (1993). Autosomal dominant leukoencephalopathy and subcortical ischemic stroke. Stroke 24, 122–125. ¨ ber familiare diVuse Sklerose (Leukodystrophia cerebri progresBielschowsky, M., and Henneberg, R. (1928). U siva hereditaria). J. Psychol. Neurol. (Lpz.) 36, 131–181. Borrett, D., and Becker, L. E. (1985). Alexander’s disease: A disease of astrocytes. Brain 108, 367–385. Brenner, M., Johnson, A. B., BoespXug-Tanguy, O., Rodriguez, D., Goldman, J. E., and Messing, A. (2001). Mutations in GFAP, encoding glial Wbrillary acidic protein, are associated with Alexander disease. Nature Genet. 27, 117–120. Cammer, W., Rose, A. L., and Norton, W. T. (1975). Biochemical and pathological studies of myelin in hexachlorophene intoxication. Brain Res. 98, 547–559. Choi, K. G., Sung, J. H., Clark, H. B., and Krivit, W. (1991). Pathology of adult-onset globoid cell leukodystrophy (GLD) (Abstract). J. Neuropathol. Exp. Neurol. 50, 336. De Reuck, J., Crevits, L., De Coster, W., Sieben, G., and vander Eecken, H. (1980). Pathogenesis of Binswanger chronic progressive subcortical encephalopathy. Neurology 30, 920–928. Dickens, C. (1894). ‘‘A Tale of Two Cities.’’ Eldridge, R., Anayiotos, C. P., Schlesinger, S., Cowen, D., Bever, C., Patronas, N., and McFarland, N. (1984). Hereditary adult-onset leukodystrophy simulating chronic progressive multiple sclerosis. N. Engl. J. Med. 311, 948–953. Ginsberg, L., and Gershfeld, N. L. (1991). Membrane bilayer instability and the pathogenesis of disorders of myelin. Neurosci. Lett. 130, 133–136. Gray, F., Destee, A., Bourre, J.-M., Gherardi, R., Krivosic, I., Warot, P., and Poirier, J. (1987). Pigmentary type of orthochromatic leukodystrophy (OLD): A new case with ultrastructural and biochemical study. J. Neuropathol. Exp. Neurol. 46, 585–596. Gray, F., Dubas, F., Roullet, E., and Escourolle, R. (1985). Leukoencephalopathy in diVuse hemorrhagic cerebral amyloid angiopathy. Ann. Neurol. 18, 54–59. Harding, B. H., and Surtees, R. (2002). Metabolic and neurodegenerative diseases of childhood. In ‘‘GreenWeld’s Neuropathology’’ (D. I. Graham and P. L. Lantos, Eds.), 7th ed., Vol. 1, pp. 485–517. Arnold, London. Ho, J. K., Moser, H., Kishimoto, Y., and Hamilton, J. A. (1995). Interactions of a very long chain fatty acid with model membranes and serum albumin. Implications for the pathogenesis of adrenoleukodystrophy. J. Clin. Invest. 96, 1455–1463. Hudson, L. D. (2001). Pelizaeus-Merzbacher disease and the allelic disorder X-linked spastic paraplegia type 2. In ‘‘The Metabolic & Molecular Bases of Inherited Disease’’ (C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, Eds.), 8th ed., Vol. IV, pp. 5789–5805. McGraw-Hill, New York. Ito, M., Blumberg, B. M., Mock, D. J., Goodman, A. D., Moser, A. B., Moser, H. W., Smith, K. D., and Powers, J. M. (2001). Potential environmental and host participants in the early white matter lesion of adrenoleukodystrophy: Morphologic evidence of CD8 cytotoxic T cells, cytolysis of oligodendrocytes, and CD1mediated lipid antigen presentation. J. Neuropathol. Exp. Neurol. 60, 1004–1019. Jellinger, K., and Seitelberger, F. (1970). Spongy degeneration of the central nervous system in infancy. Curr. Topics Pathol. 53, 90–159. Joutel, A., Corpechot, C., Ducros, A., Vahedi, K., Chabriat, H., Mouton, P., Alamowitch, S., Domenga, V., Cecillion, M., Marechal, E., Maciazek, J., Vayssiere, C., Cruaud, C., Cabanis, E. A., Ruchoux, M. W., Weissenbach, J., Bach, J. F., Bousser, M. G., and Tournier-Lasserve, E. (1996). Notch3 mutations in CADASIL, a hereditary late-onset condition causing stroke and dementia. Nature 383, 707–710. Kimbrough, R. D., and Gaines, T. B. (1971). Hexachlorophene eVects on the rat brain: Study of high doses by light and electron microscopy. Arch. Environm. Hlth. 23, 114–118. Klatzo, I. (1967). Neuropathological aspects of brain edema. J. Neuropathol. Exp. Neurol. 26, 1–14. Kornfeld, M., Moser, A., Moser, H., Kleinschmidt-DeMasters, B., Nolte, K., and Phelps, A. (1994). Solvent vapor abuse leukoencephalopathy. Comparison to adrenoleukodystrophy. J. Neuropathol. Exp. Neurol. 43, 389–398. Leech, R. W., Brumback, R. A., Miller, R. H., Otsuka, F., Tarone, R. E., and Robbins, J. H. (1985). Cockayne syndrome: Clinicopathologic and tissue culture studies of aVected siblings. J. Neuropathol. Exp. Neurol. 44, 507–519. Leegwater, P. A., Vermeulen, G., Konst, A. A., Naidu, S., Mulders, J., Visser, A., Kersbergen, P., Mobach, D., Fonds, D., van Berkel, C. G., Lemmers, R. J., Frants, R. R., Oudejans, C. B., Schutgens, R. B., Pronk, J. C., and van der Knaap, M. S. (2001). Subunits of the translation initiation factor eIF2B are mutant in leukoencephalopathy with vanishing white matter. Nat. Genet. 29, 383–388. Ludwin, S. K. (1997). The pathobiology of the oligodendrocyte. J. Neuropathol. Exp. Neurol. 56, 111–124. Matalon, R., Michals, K., Sebesta, D., Deanching, M., GashkoV, P., and Casanova, J. (1988). Aspartoacylase deWciency and N-acetylaspartic aciduria in patients with Canavan disease. Am. J. Med. Genet. 25, 463–471.

LEUKODYSTROPHIES: PATHOGENETIC SIMILARITIES AND DISSIMILARITIES

au14

Miyatake, T., and Suzuki, K. (1972). Globoid cell leukodystrophy. Additional deWciency of psychosine galactosidase. Biochem. Biophys. Res. Commun. 48, 538–543. Moser, H. W. (1997). Adrenoleukodystrophy: phenotype, genetics, pathogenesis, and therapy. Brain 120, 1485–1508. Nance, M. A., and Berry, S. A. (1992). Cockayne syndrome: review of 140 cases. Am. J. Med. Genet. 42, 68–84. Norton, W. T., Poduslo, S. E., and Suzuki, K. (1966). Subacute sclerosing leukoencephalitis. II. Chemical studies including abnormal myelin and an abnormal ganglioside pattern. J. Neuropathol. Exp. Neurol. 25, 582–597. Plum, F., Posner, J. B., and Hain, R. F. (1962). Delayed neurological deterioration after anoxia. Arch. Intern. Med. 110, 56–63. Poser, C. M. (1962). Concepts of dysmyelination. In ‘‘Cerebral Sphingolipidoses’’ (S. M. Aronson and B. W. Volk, eds.), pp. 141–164. Academic Press, New York. Poser, C. M. (1968). Diseases of the myelin sheath. In ‘‘Pathology of the Nervous System’’ (J. Minckler, ed.), Vol. I, pp. 767–820. McGraw-Hill, New York. Powers, J. M. (1996). A neuropathologic overview of the neurodystrophies and neurolipidoses. Handbook of Clin. Neurol. 66 (revised series 22), 1–32. Powers, J. M., DeCiero, D. P., Cox, C., RichWeld, E. K., Ito, M., Moser, A. B., and Moser, H. B. (2001). The dorsal root ganglia in adrenomyeloneuropathy: Neuronal atrophy and abnormal mitochondria. J. Neuropathol. Exp. Neurol. 60, 493–501. Powers, J. M., DeCiero, D. P., Ito, M., Moser, A. B., and Moser, H. W. (2000). Adrenomyeloneuropathy: A neuropathologic review featuring its noninXammatory myelopathy. J. Neuropathol. Exp. Neurol. 59, 89–102. Powers, J. M., and De Vivo, D. C. (2002). Peroxisomal and mitochondrial disorders. In ‘‘GreenWeld’s Neuropathology’’ (D. I. Graham and P. L. Lantos, eds.), 7th ed., Vol. 1, pp. 737–797. Arnold, London. Powers, J. M., and Horoupian, D. S. (1995). Central nervous system. In ‘‘Anderson’s Pathology’’ (I. Damjanov and J. Linder, eds.), 10th ed., pp. 2777–2782. Mosby, Philadelphia. Powers, J. W., Liu, Y., Moser, A. B., and Moser, H. W. (1992). The inXammatory myelinopathy of adrenoleukodystrophy. J. Neuropathol. Exp. Neurol. 51, 630–643. Prineas, J. W., McDonald, I., and Franklin, R. J. M. (2002). Demyelinating diseases. In ‘‘GreenWeld’s Neuropathology’’ (D. I. Graham and P. L. Lantos, eds.), 7th ed., Vol. 1, pp. 471–550. Arnold, London. Raine, C. S. (1997). Demyelinating diseases. In ‘‘Textbook of Neuropathology’’ (R. L. Davis, D. M. Robertson, eds.), 3rd ed., pp. 627–714. Williams & Wilkins, Baltimore. Ramsey, R. B., and Davison, A. N. (1974). Steryl esters and their relationship to normal and diseased human central nervous system. J. Lipid Res. 15, 249–255. Razavi, E. F., Larroche, J. C., and Gaillard, D. (1988). Infantile familial encephalopathy with cerebral calciWcations and leukodystrophy. Neuropediatrics 19, 72–79. Richardson, E. P., Jr., (1982). Myelination in the human central nervous system. In ‘‘Histology and Histopathology of the Nervous System’’ (W. Haymaker, R. D. Adams, eds.), pp. 146–173. Charles C. Thomas, SpringWeld, IL. Schaumburg, H. H., Powers, J. M., Raine, C. S., Spencer, P. S., GriYn, J. W., Prineas, J. W., and Boehme, D. M. (1977). Adrenomyeloneuropathy: A probable variant of adrenoleukodystrophy. II. General pathologic, neuropathologic, and biochemical aspects. Neurology 27, 1114–1119. Schaumburg, H. H., Powers, J. M., Raine, C. S., Suzuki, K., and Richardson, Jr., E. P. (1975). Adrenoleukodystrophy. A clinical and pathological study of 17 cases. Arch. Neurol. 33, 577–591. Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (2001). ‘‘The Metabolic & Molecular Bases of Inherited Disease,’’ 8th ed. McGraw-Hill, New York. Seitelberger, F. (1995). Neuropathology and genetics of Pelizaeus-Merzbacher disease. Brain Pathol. 5, 267–274. Smith, K. D., Kemp, S., Braiterman, L. T., Lu, J-F, Wei, H-M, Geraghty, M., Stetten, G., Bergin, J. S., Peysner, J., and Watkins, P. A. (1999). X-linked adrenoleukodystrophy: Genes, mutations, and phenotypes. Neurochem. Res. 24, 521–535. Suzuki, K., and Suzuki, K. (2002). Lysosomal diseases. In ‘‘GreenWeld’s Neuropathology’’ (D. I. Graham and P. L. Lantos, eds.), 7th ed., Vol. 1, pp. 653–735. Arnold, London. Tournier-Lasserve, E., Joutel, A., Meilki, J., Weissenbach, J., Lathrop, G. M., Chabriat, H., Mas, J. L., Cabanis, E. A., Baudrimont, M., and Maciazek, J. (1993). Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy maps to chromosome 19q12. Nat. Genet. 3, 256–259. Trapp, B. D., Peterson, J., RansohoV, R. M., Rudick, R., Mork, S., and Bo, L. (1998). Axonal transection in the lesions of multiple sclerosis. N. Engl. J. Med. 338, 278–285. van der Knaap, M. S. (2001). Magnetic resonance in childhood white-matter disorders. Dev. Med. Child Neurol. 43, 705–712. van der Knaap, M. S., Barth, P. G., Gabreels, F. J., Franzoni, E., Begeer, J. H., Stroink, H., Rotteveel, J. J., and Valk, J. (1997). A new leukoencephalopathy with vanishing white matter. Neurology 48, 845–855. van der Knaap, M. S., Barth, P. G., Vrensen, G. F. J. M., and Valk, J. (1996). Histopathology of an infantileonset spongiform leukoencephalopathy with a discrepantly mild clinical course. Acta Neuropathol. 92, 206–212. van der Knaap, M. S., Breiter, S. N., Naidu, S., Hart, A. A. M., and Valk, J. (1999). DeWning and categorizing leukoencephalopathies of unknown origin: MR imaging approach. Radiology 213, 121–133. van der Knaap, M. S., Kamphorst, W., Barth, P. G., Kraaijeveld, C. L., Gut, E., and Valk, J. (1998). Phenotypic variation in leukoencephalopathy with vanishing white matter. Neurology 51, 540–547.

689

690

28. THE LEUKODYSTROPHIES

van der Knaap, M. S., and Valk, J. (1995). ‘‘Magnetic resonance of myelin, myelination, and myelin disorders,’’ 2nd ed. Springer, Berlin. Wakui, H., Nishimura, S., Watahiki, Y., Endo, Y., Nakamoto, Y, and Miura, A. B. (1991). Dramatic recovery from neurological deWcits in a patient with central pontine myelinolysis following severe hyponatremia. Jpn. J. Med. 30, 281–284. Wolman, M. (1970). Histochemistry of myelination and demyelination. In ‘‘Handbook of Clinical Neurology’’ (P. J. Vinken and W. G. Bruyn, eds.), Vol. 9, pp. 24–44. North-Holland Publishing, Amsterdam.

C H A P T E R

29 Multiple Sclerosis Classification and Overview Fred D. Lublin

INTRODUCTION Multiple sclerosis (MS) has been a recognized clinical entity since the latter part of the 19th century, following the clinical description by Charcot. Several pathologic descriptions preceded this one, but Charcot is credited with providing the synthesis of the clinical and pathologic pictures. Anecdotes that describe rather typical cases can be found as far back as the Middle Ages. MS is the commonest of the demyelinating diseases and the commonest cause of neurologic disability in young adults. The prevalence of MS in North America is about 100 per 100,000 and incidence of about 6 per 100,000, increasing with latitude. Approximately 350,000 persons in the United States have MS, and this number may be an underestimate. The average age at onset is 32, and patients tend to live in excess of 35 years from time of diagnosis. Therefore, although the actual number of individuals is not large compared to some other diseases, the longevity and the potential for serious disability produce considerable economic consequences. The cost of MS in the United States is $9.6 billion per year (in 1994 dollars), around $34,000 per year for each patient, exclusive of the costs of disease-modifying agents. The signs and symptoms of MS are the consequence of the underlying neuropathologic changes that occur in patients. The primary mechanism of injury is by inXammatory demyelination and, to a variable degree, axonal damage. Either mechanism may produce clinical features. The role of axonal damage is clear cut, disrupting conduction completely. Demyelination may result in either slowing of conduction or complete failure of transmission. The former will produce symptoms when the slowing becomes critical. As the pathologic damage may involve any area of the central nervous system (CNS), the location of the lesion(s) also plays a role in symptom production. MS can produce any symptom or sign that might occur with damage to the CNS, especially white matter tracks. The most common Wndings include optic neuritis, weakness, sensory loss, ataxia, nystagmus, bladder dysfunction, and cognitive impairment, but the full list is quite long. The progressive impairment or disability that occurs over time with MS results from one of two mechanisms. There is either stepwise worsening due to accumulated deWcits from residua of exacerbations or gradual, inexorable progressive disease, independent of the exacerbations (see the fuller deWnition of an exacerbation presented later in the chapter). The relative role of exacerbations and progressive disease in the accumulation of deWcits has been debated, but the data are clear that both impact the long-term course of the illness. Recent data from a meta-analysis of several clinical trials in MS demonstrate that

Myelin Biology and Disorders, Volume 2

691

Copyright 2004, Elsevier (USA). All rights reserved.

692

29. MULTIPLE SCLEROSIS CLASSIFICATION AND OVERVIEW

residual deWcit from exacerbations occurs after at least 50% of attacks. Later in the course of MS, progressive disease seems to contribute more strongly to the disability (Confavreux et al., 2000).

DIAGNOSING MS The diagnosis of MS is based on Wnding clinical evidence of lesions of the CNS, disseminated in time and space. For this reason, some have referred to the illness as disseminated sclerosis. Dissemination in time implies that there is more than one episode of CNS dysfunction. Dissemination in space implies involvement of more than one area of the CNS. This is accomplished through a careful medical history and a detailed neurologic examination. Although all MS begins with a Wrst attack, the essence of the clinical disease is the multiplicity of attacks. Several diagnostic criteria have been proposed over the past several decades, all aYrming the need for dissemination, primarily white matter involvement, young age at onset (20 to 50), and an important caveat: that there be no better diagnosis. Twenty years ago, a committee organized by the National Multiple Sclerosis Society (NMSS), developed the penultimate MS diagnostic criteria, incorporating the clinical aspects of MS, paraclinical evidence (MRI, urodynamics, evoked potentials) and cerebrospinal Xuid (CSF) immunoglobulin abnormalities (Poser et al., 1983). This diagnostic schema segregated patients into deWnite, probable, or possible MS categories and was designed primarily for clinical research protocols. The role of MRI, which was in its infancy as a diagnostic tool, was not deWned. In 2000, another committee of the NMSS convened to update and revise the MS diagnostic criteria for MS, with the intention of increasing the role of MRI in the diagnostic schema. Over the past 20 years, there has been considerable knowledge of the importance of MRI in providing an in vivo view of the neuropathology of MS. Other objectives for this revision of the diagnostic criteria were to simplify the categories and produce guidelines that would be useful to practicing clinicians. Outside experts then reviewed the deliberations of that group. The resultant manuscript was submitted for peer review and published in the Annals of Neurology in July 2001 (McDonald et al., 2001). The need for this most current revision is underscored by the development of MS disease-modifying agents (DMAs) over the past decade. Since the advent of DMAs, the need for early, accurate therapy has become extremely important as the accumulated data suggest that the earlier treatment is started, the less there is risk of accumulating impairment/disability. The general conclusions from the new criteria are that the diagnosis of MS requires objective evidence of lesions disseminated in time and space; MRI Wndings may contribute to determination of dissemination in time or space; other supportive investigations include CSF and visual evoked potential (VEP); and the diagnostic categories are possible MS, MS, or not MS (the category of ‘‘probable MS,’’ used in Poser, had little practical value for clinicians, other than suggesting more certainty than ‘‘possible,’’ and was of no value in clinical trials). The new guidelines reaYrmed the classical approach to diagnosing MS by clinical means only: the Wnding of evidence of lesions of the CNS, disseminated in time and space, based on a detailed history and a complete neurological examination. This schema allows for diagnosing MS in regions of the world where there is limited access to newer technologies. Most MS starts with an attack/relapse/exacerbation—that is, an acute episode of CNS dysfunction lasting at least 24 hours, occurring in the absence of fever or metabolic derangement. All events occurring within 30 days of such an event are considered, by convention, to be part of a single event, even though multiple areas of the CNS may be involved. The commonest clinical course of MS follows a relapsing-remitting course of multiple attacks, as will be described. The role of MRI in the diagnosis of MS has been expanded considerably in this latest MS diagnostic guideline. After careful consideration of the various MRI studies of patients with MS, the committee determined that the criteria of Barkhof (Barkhof et al., 1997), as

DIAGNOSING MS

amended by Tintore (Tintore et al., 2000), provided the best combination of speciWcity and sensitivity, with emphasis on accuracy, as appropriate for a diagnostic guideline. For dissemination in space (Tab. 29.1), these criteria require three of the following four elements: (1) at least one gadolinium enhancing lesion or nine T2 hyperintense lesions, (2) at least one infratentorial lesion, (3) at least one juxtacortical lesion, and (4) at least three periventricular lesions. A spinal cord lesion can substitute for any of these brain lesions. If there are immunoglobulin abnormalities in the CSF, then the MRI criteria are relaxed to only two T2 lesions typical of MS. The MRI also can be used to conWrm dissemination in time (Tab. 29.2). If an MRI scan of the brain performed at 3 or more months after an initial clinical event demonstrates a new gadolinium-enhancing lesion, this would indicate a new CNS inXammatory event, as the duration of gadolinium enhancement in MS is usually less than 6 weeks. If there are no gadolinium-enhancing lesions but a new T2 lesion (presuming an MRI at the time of the initial event), then a repeat MRI scan after another 3 months is needed with demonstration of a new T2 lesion or gadolinium-enhancing lesion. The reason for the second scan to establish that a new T2 lesion has occurred relates to the inclusion of all events within 30 days of an exacerbation as part of the initial exacerbation. A new T2 lesion developing within the 30 days following an exacerbation would not count as a new event, although it would show up as new on the Wrst 3-month scan, thus necessitating a new T2 lesion on the second 3month scan. The 3-month interval was a consensus decision and is oVered as a guideline. Spinal Xuid analysis is also useful for diagnosing MS, but not diagnostic by itself (as other conditions can produce similar abnormalities). The presence of immunoglobulin abnormalities in the CSF indicates the production of immunoglobulin within the CNS. This is best determined by the Wnding of oligoclonal bands of IgG present in the CSF, but not in serum. This is best determined by using the isoelectric focusing technique. The Wnding of an elevated IgG index (a ratio of the IgG to protein in the serum and CSF) is equally helpful. The total protein in the CSF is almost always less than 100 in MS and the Wnding of more than 50 WBCs in the CSF is quite uncommon. Evoked responses can also be used to provide evidence for dissemination in space. The most valuable is the visual evoked potential (VEP). The Wnding of a prolonged VEP in an individual without clinical evidence of an optic nerve lesion indicates subclinical involvement of the optic nerve. The other evoked responses were not found to be discriminative or sensitive enough to provide useful diagnostic information (Gronseth and Ashman, 2000). Putting these diagnostic guidelines into practice is rather straightforward. If the patient has had two or more attacks with objective evidence of involvement of two or more areas of the CNS, and there is no better diagnosis, then the patient has met the criteria for dissemination in time and space. In this circumstance, additional studies are not necessary, an important issue in regions where access to MRI might be limited or nonexistent. If

TABLE 29.1 New Diagnostic Criteria for MRI Determination of Dissemination in Space (after Barkhof et al. and Tintore et al. (Tintore et al., 2000) ) Three out of four of the following: .

One Gdþ lesion or 9 T2 hyperintense lesions

.

One infratentorial lesion

.

One juxtacortical lesion

.

Three periventricular lesions (One spinal cord lesion ¼ one brain lesion)

TABLE 29.2 .

.

New Diagnostic Criteria for MRI Determination of Dissemination in Time

Gadolinium-enhancing lesion demonstrated in a scan done at least 3 months following onset of a clinical attack at a site diVerent from attack/ In the absence of gadolinium-enhancing lesions at the 3 month scan, follow-up scan after an additional 3 months showing a gadolinium-enhancing lesion or new T2 lesion.

693

694

29. MULTIPLE SCLEROSIS CLASSIFICATION AND OVERVIEW

additional studies are undertaken, such as MRI or CSF analysis, they should yield consistent results or the diagnosis should be questioned. If there has been one attack and evidence of two lesions in the CNS, then one needs to obtain evidence for dissemination in time. This can be accomplished by either a second attack or by evidence of dissemination in time on MRI scanning (see Tab. 29.2) If there are two or more attacks but evidence on exam of involvement of only one area of the CNS, then one needs to conWrm dissemination in space. A second attack, involving a new area of the CNS or changes on subsequent MRI scan (see Tab. 29.1), will fulWll this criterion. If there has been one attack and evidence of only one area of CNS involvement (the clinically isolated syndrome), then one need conWrm dissemination in both time and space. This can be accomplished clinically or by fulWlling the MRI criteria for dissemination for each (Tabs. 29.1 and 29.2). In the case of insidious onset and progression of neurologic dysfunction suggestive of MS (usually primary progressive MS), one needs to Wnd positive CSF and evidence of dissemination in time by MRI or continued progression for at least 1 year and dissemination in space by MRI (see Tab. 29.1) or two or more spinal MRI abnormalities or four to eight cerebral MRI lesions and one spinal cord lesion or an abnormal VEP and four to eight cerebral lesions on MRI or an abnormal VEP and less than four cerebral lesions plus one spinal cord lesion. These criteria were adapted from those previously published by Thompson et al. (Thompson et al., 2000). As opposed to the other guidelines, it was felt that CSF was necessary to diagnose primary progressive MS, in that the risk of alternative diagnoses was greater than for other forms of MS. There is not an abundance of demographic data on pure primary progressive MS, as currently deWned. In the past, the practice of including patients with relapses, often remote, has confounded this category. Analysis of recent clinical trials of primary progressive MS should provide important additional data that can be used to support or modify the diagnostic guidelines. These new diagnostic criteria are relatively easy to employ in clinical practice and in clinical trials. Figures 29.1 and 29.2 outline the major features of the new criteria and are available on a laminated card from the National Multiple Sclerosis Society (USA). Initial studies show that the new criteria are highly sensitive and speciWc, leading to earlier diagnosis. One study shows more than twice the rate of conversion from a clinically isolated syndrome to MS using the new criteria as compared to the older Poser et al. criteria (Dalton et al., 2002). Similar results are seen in therapeutic trials of clinically isolated syndromes and early MS.

CLINICAL SUBTYPES OF MS The clinical course of MS, although quite variable in temporal sequence, tends to follow one of several speciWc courses characterized by either a relapsing pattern or a progressive course. In relapsing forms of MS, there occur multiple acute exacerbations of neurologic dysfunction lasting days to months, with a variable degree of recovery and then stability until the next exacerbation, which can occur weeks to decades later. There are at present no reliable biologic makers (either MRI or clinical laboratory) that distinguish the disease course patterns, so they were decided by consensus, based on a survey of the international MS clinical research community published in 1996 (Lublin and Reingold, 1996). The clinical course patterns can be divided into four subtypes: relapsing-remitting, primary progressive, secondary progressive, and progressive-relapsing, as outlined next. Relapsing-remitting. Relapsing-remitting (RR) MS (Figs. 29.3A and B) is the commonest form at presentation and is characterized by clearly deWned disease relapses with full recovery or with sequelae and residual deWcit upon recovery. Periods between disease relapses are characterized by a lack of disease progression (Figs. 29.5A and B). The deWning elements of RR MS are episodes of acute worsening of neurologic function followed by a variable degree of recovery, with a stable course between attacks. Approximately 85 to 90% of patients with MS start with an RR course.

CLINICAL SUBTYPES OF MS

New Multiple Sclerosis Diagnostic Criteria CLINICAL (ATTACKS)

OBJECTIVE LESIONS

2 or more

2 or more

2 or more

1

1

2 or more

1 monosymptomatic

1

• Dissemination in space by MRI or positive CSF and 2 or more MRI lesions consistentwith MS AND • Dissemination in time by MRI or second clinicalattack

1

• Positive CSF AND • Dissemination in space by MRI evidence of 9 or more T2 brain lesions or 2 ormore cord lesions or 4-8 brain and 1 cord lesion or positive VEP with 4-8 MRIlesions or positive VEP withless than4 brainlesions plus1 cordlesion AND • Dissemination in time by MRI or continued progressionfor 1 year

0 (progression from onset)

ADDITIONAL REQUIREMENTSTOMAKEDIAGNOSIS • None;clinical evidence willsuffice (additional evidencedesirable but mustbe consistentwith MS) • Dissemination in space by MRI or positive CSF and 2 or more MRI lesions consistentwith MS or further clinicalattack involving different site

• Dissemination in time by MRI or second clinicalattack

FIGURE 29.1

ParaclinicalEvidence in MS Diagnosis What is a Positive MRI? 3 out of 4of the following: 1 Gd-enhancing lesion or 9 T2 hyperintense lesions if no Gd-enhancing lesion 1 or more infratentorial lesions 1 or more juxtacortical lesions 3 or more periventricular lesions Note: 1 cord lesion can substitute for 1 brain lesion

WhatProvides MRI Evidence of Dissemination in Time? A Gd-enhancinglesion demonstrated in a scan done at least 3 months following onset of clinicalattack at a site different from attack, or In absence of Gd-enhancing lesions at 3 monthscan, follow-up scan after an additional 3 months showing Gd-lesion ornew T2lesion.

What is Positive CSF? Oligoclonal IgG bandsin CSF (and not serum) or elevated IgG index

What is Positive VEP? Delayed but well-preserved wave form

2001 The National Multiple Sclerosis Society

FIGURE 29.2

Primary progressive. Primary progressive (PP) MS (Figs. 29.4A and B) is characterized by disease progression from onset with occasional plateaus and temporary minor improvements allowed. Approximately 10% of patients have this form of MS. The essential element

695

696

29. MULTIPLE SCLEROSIS CLASSIFICATION AND OVERVIEW

Increasing Disability

A

TIME

B

TIME

Increasing Disability

FIGURE 29.3 Relapsing-remitting (RR) MS is characterized by clearly deWned acute attacks with full recovery (A) or with sequelae and residual deWcit upon recover (B). Periods between disease relapses are characterized by lack of disease progression.

Increasing Disability

A

TIME

Increasing Disability

B

TIME

FIGURE 29.4 Primary-progressive (PP) MS is characterized by disease showing progression of disability from onset, without plateaus or remissions (A) or with occasional plateaus and temporary minor improvements (B).

CLINICAL SUBTYPES OF MS

in PP MS is a gradual, nearly continuously worsening baseline with minor Xuctuations, but no distinct relapses. While near continuous progression is required in this deWnition, it was recognized that progression at a constant rate throughout disease (Fig. 29.4A) was unlikely and that accommodation must be made for variations in the rate of progression over time (Fig. 29.4B). PP MS is quite distinct from RR MS (especially the absence of any exacerbations), causing some to suggest that it may represent a diVerent disease. However, current evidence suggests that PP is a subtype of typical MS. Secondary progressive. Secondary progressive (SP) MS (Figs. 29.5A and B) is characterized by an initial relapsing-remitting disease course followed by progression with or without occasional relapses, minor remissions, and plateaus. SP MS may be seen as a long-term outcome of RR MS, in that almost all SP patients initially begin with RR disease as deWned here. However, once the baseline between relapses begins to progressively worsen, the patient has switched from RR MS to SP MS. This transition from RR to SP occurs in up to 50% of RR MS patients, although it can take many years and is unpredictable. Progressive-relapsing. In the progressive-relapsing (PR) form of MS (Figs. 29.6A and B), there is progressive disease from onset, with clear acute relapses, with or without recovery, with periods between relapses characterized by continuing progression. Approximately 5 to 6% of patients have this form of MS, but there are data now accruing that PP MS patients may convert to PR at a rate of almost 1% per year. This will be better understood once a large clinical trial in PP MS has completed. The term ‘‘chronic progressive MS,’’ used frequently in the past has been discarded in favor of one of the more descriptive progressive forms just described. MS can also be categorized by outcome. At the extremes, MS can be designated as either benign or malignant. Benign MS has been deWned as disease that allows patients to remain fully functional in all neurologic systems 15 years after disease onset. Although this form may comprise 10 to 15% of patients, diagnosis, and thus prognosis, is diYcult and by deWnition requires 15 years. Even then, relapses or progression can occur, sometimes as late as 25 years later.

Increasing Disability

A

TIME

B

TIME

Increasing Disability

FIGURE 29.5 Secondary progressive (SP) MS begins with an initial RR course, followed by progression of variable rate (A) or may also include occasional and minor remissions (B).

697

698

29. MULTIPLE SCLEROSIS CLASSIFICATION AND OVERVIEW

Increasing Disability

A

TIME

Increasing Disability

B

TIME

FIGURE 29.6 Progressive-relapsing (PR) MS shows progression from onset, but with clear acute relapses with (A) or without (B) full recovery.

Malignant MS is deWned as disease with a rapid, progressive course, leading to signiWcant disability in multiple neurologic systems or death in a relatively short time after disease onset. This is, fortunately, quite rare. It is hoped that in the near future, we will have biologically based characterizations of the diVerent MS disease courses, likely using advanced MRI metrics and possibly immunologic markers. These could improve our prognostic abilities and allow for more rational use of the various therapies available for treating MS, and perhaps provide increased insight into the complex underlying pathophysiologic aberration that is MS.

References Barkhof, F., Filippi, M., Miller, D. H., Scheltens, P., Campi, A., Polman, C. H., Comi, G., Ader, H. J , LosseV, N., and Valk, J. (1997). Comparison of MRI criteria at Wrst presentation to predict conversion to clinically deWnite multiple sclerosis. Review. Brain 120, 2059–2069. Confavreux, C., Vukusic, S., Moreau, T., and Adeleine, P. (2000). Relapses and progression of disability in multiple sclerosis. N. Engl. J. Med. 343, 1430–1438. Dalton, C. M., Brex, P. A., Miszkiel, K. A., Hickman, S. J., MacManus, D. G., Plant, G. T., Thompson, A. J., and Miller, D. H. (2002). Application of the new McDonald criteria to patients with clinically isolated syndromes suggestive of multiple sclerosis. Ann. Neurol. 52, 47–53. Gronseth, G. S., and Ashman E. J. (2000). Practice parameter: The usefulness of evoked potentials in identifying clinically silent lesions in patients with suspected multiple sclerosis (an evidence-based review): Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 54, 1720–1725. Lublin, F. D., and Reingold S. C. (1996). DeWning the clinical course of multiple sclerosis: Results of an international survey. National Multiple Sclerosis Society (USA) Advisory Committee on Clinical Trials of New Agents in Multiple Sclerosis. Neurology 46, 907–911 McDonald, W. I., Compston, A., Edan, G., Goodkin, D., Hartung, H. P., Lublin, F. D., McFarland, H. F., Paty, D. W., Polman, C. H., Reingold, S. C., Sandberg-Wollheim, M., Sibley, W., Thompson, A., van Den, N. S., Weinshenker, B. Y., and Wolinsky, J. S. (2001). Recommended diagnostic criteria for multiple sclerosis: Guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann. Neurol. 50, 121–127. Poser C. M, Paty, D. W., Scheinberg, L., McDonald W. I., Davis, F. A., Ebers, G. C., Johnson, K. P., Sibley, W. A., Silberberg, D. H., and Tourtellotte, W. W. (1983). New diagnostic criteria for multiple sclerosis: Guidelines for research protocols. Ann. Neurol. 13, 227–231.

au1

CLINICAL SUBTYPES OF MS

Thompson, A. J., Montalban, X., Barkhof, F., Brochet, B., Filippi, M., Miller, D. H., Polman, C.H., Stevenson, V. L., and McDonald, W. I. (2000), Diagnostic criteria for primary progressive multiple sclerosis: A position paper. Ann. Neurol. 47, 831–835. Tintore, M., Rovira, A., Martinez, M. J., Rio, J., Diaz-Villoslada, P., Brieva, L., Borras, C., Grive, E., Capellades, J., and Montalban, X .(2000). Isolated demyelinating syndromes: Comparison of diVerent MR imaging criteria to predict conversion to clinically deWnite multiple sclerosis. AJNR Am. J. Neuroradiol. 21, 702–706.

AUTHOR QUERIES [au1] As meant? [au2] Spell out here? [au3] Figure captions to come, or is the text on the Wgure manuscript to serve as the caption text?

699

C H A P T E R

30 Genetic Susceptibility and Epidemiology Alastair Compston

THE METHODOLOGY OF EPIDEMIOLOGICAL STUDIES IN MULTIPLE SCLEROSIS Epidemiology remains one of the most active areas of multiple sclerosis research. The emerging statistics serve several purposes: generating etiological hypotheses, establishing health care needs in the community, deWning the natural history of multiple sclerosis as the basis for understanding the evolving clinical expression of tissue injury, and providing a yardstick against which the results of therapies can be compared. Because the many studies have been performed over the past 100 years in diVerent places and at diVerent times, comparisons—which aim to see the big epidemiological picture—are notoriously diYcult. It follows that attempts to formulate reliable hypotheses using the epidemiological evidence are potentially vulnerable. Most sensitive to artifact have been the temporal and geographical trends emerging from comparisons of prevalence between regions and the serial study of individual locations. An everyday word such as frequency is useful in conveying a general impression of statistics for multiple sclerosis but it lacks precision. Cumulative frequency or lifetime risk is the maximum chance that the disease will occur during the lifetime of an at-risk individual; it is around 1:400 for northern European Caucasians. Risk factors alter this rate. Their contribution to the underlying pathogenesis can be expressed as the relative risk (the product of the proportions of cases and controls with and without the risk factor) or the odds ratio (the ratio of incidence rates for individuals who have and have not been exposed to the risk factor). Relative risk is a collective descriptor and the contribution made by any one factor is the attributable risk. Incidence describes new events (the numerator) in a deWned group (the denominator) over a given period. Each is liable to ascertainment error, but the impact on statistics is greater when mistakes occur in estimating the numerator. Prevalence describes the number of aVected individuals in a population at risk on a given occasion. Ascertainment will vary inversely with size and accessibility of the at-risk population, and security of the diagnosis, and it tends to increase with repeated survey as awareness and vigilance improve among participants. The at-risk population should be demographically based. Mortality describes the number of individuals dying with or as a result of multiple sclerosis among the at-risk population over a given period. With the decline in autopsy rates and the trend for death certiWcation to reXect administrative needs rather than pathological veriWcation, mortality is a poor statistic for evaluating the epidemiology of multiple sclerosis. Incidence, prevalence, and mortality have a close relationship. In a population not experiencing recent demographic change and where mortality returns are accurate and complete, incidence will equal mortality, and prevalence will be the product of either

Myelin Biology and Disorders, Volume 2

701

Copyright 2004, Elsevier (USA). All rights reserved.

702

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

statistic and disease duration. In practice, changes in frequency usually arise from predictable cycles in these statistics (regression to the mean) and structure of the population being surveyed, rather than alteration in etiological and biological factors causing the disease. In a recent population-based study, survival was >50% at 40 years from onset with an excess of deaths from suicide and neoplasia; complications of multiple sclerosis accounted for 70% of deaths (Sumelahti et al., 2002). The diYculty of describing statistics that incorporate lifetime risk in children and young adults is addressed by quoting age- and sex-speciWc rates for morbidity; these relate numerators to a denominator conWned to that proportion of the at-risk population, which has the same age and gender structure. One further reWnement is to relate statistics to a single representative or virtual population and derive a standardized prevalence ratio. The 95% conWdence limits (upper and lower with attention to whether or not these straddle unity) provide a statement on the likely reproducibility of a given epidemiological Wnding. The easy route to answering an epidemiological problem in multiple sclerosis is to retrieve cases from an existing register, usually hospital or clinic based, of validated cases. The inclusion or exclusion of marginal cases will vary with the purpose of the survey. In seeking to identify biological features, the error should be toward inclusion of individuals who probably have the disease process even if this is not yet clinically deWnite. In other contexts, it is advisable to restrict the register to those who meet strict criteria (McDonald et al., 2001). Most investigators segregate cases of diVerent racial origin since sociohistoric factors may create variations in risk status across even quite small regions. However, some epidemiological questions can only be answered by comparing speciWcally diVerent groups or locations. Choosing a population with a low prevalence of multiple sclerosis for the study of a rare event, such as twinning, guarantees frustration and a less than deWnitive result since the numerator will be low. An important genetic or biological feature may not diVer signiWcantly between groups in places where multiple sclerosis is frequent and risk factors are over-represented in the at-risk population. Paradoxically, the chance of identifying factors that are common in the at-risk population and make a major contribution to the pathogenesis is improved by surveying regions of low prevalence. Conversely, those risk factors for multiple sclerosis that are not over-represented in the normal population will be identiWed more easily in high prevalence regions.

THE DISTRIBUTION OF MULTIPLE SCLEROSIS By the beginning of the 20th century, multiple sclerosis—a disease that merited individual case reports 25 years previously—had become one of the most common reasons for admission to a neurological ward. The period 1900–1950 saw a gradual maturation of methods for accurate deWnition of population-based statistics. Thereafter, surveys from many parts of the world established the geography of multiple sclerosis and allowed speculation on the reasons for this pattern. Kurtzke (1975) Wrst systematically collated published surveys of prevalence and suggested that the distribution fell into zones of low, medium, and high prevalence. The high risk band (>30/105) extended throughout northern Europe, the northern United States, Canada, southern Australia, and New Zealand. Areas of medium risk (5-25/105) were southern Europe, the southern United States, and northern Australia. Low risk (<5/105) extended to Asia and South America and was assumed in many uncharted regions. Improved methodology, alterations in classiWcation, and enhanced investigator vigilance have all impacted on serial updating of these Wgures showing a rise in the absolute number of cases in almost all parts of the globe (Bauer, 1987; Lauer and Firnhaber, 1994; Kurtzke, 1993). Many of the claims for latitudinal gradients now seem less secure but multiple sclerosis does seem to show genuine variation in its distribution over quite small distances. The explanation for these patterns has been much debated in the reasonable belief that they provide important clues to the cause of the disease. The raw statistics for prevalence are shown in Figure 30.1; these are rates per 105 of the population, often rounded up or down to provide a best guess for regions subjected to

703

THE DISTRIBUTION OF MULTIPLE SCLEROSIS

A

55

120

110

68

94

165

110 65

70 22

B

110

73

70

66 93 75 120

280

96

200

55 105 120 60 70 140 76 100 85 40 51 90 120 25 98 80 110 26 60 80 50 40 50 120 65 30 30 60 35 10 32 144 10 17 90 53 50120

230

13

13

C

11 18 21 29 29

33

24 38 72

74

77

FIGURE 30.1 Prevalence rates for multiple sclerosis (per 105) in Europe (B), the United States and Canada (A), and Australia & New Zealand (C). Numbers are approximations based on best available morbidity statistics. Estimates published since 1998 are shown in large font.

704

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

serial or multiple simultaneous survey; conWdence intervals have not been calculated or shown for those studies in which these statistics were originally included. To summarize the situation, multiple sclerosis remains a relatively common disorder of young adults in northern Europe, continental North America, and Australasia. The indigenous peoples of the Orient, Arabian peninsula, Africa, continental South America, and India are less susceptible, although recent studies are starting to erode these diVerentials. Cases are now identiWed among small population isolates, previously considered to be absolutely protected; these include examples of multiple sclerosis in Saami, Hutterites, and African blacks, among others. In northern Europe, incidence and prevalence are higher in southern Scandinavia, northern Germany, the United Kingdom, and parts of Italy than in northern Scandinavia, France, Spain, and the eastern Mediterranean countries. Multiple sclerosis is more prevalent in northeast Scotland and the Orkney and Shetland islands than in other parts of the United Kingdom (Robertson and Compston, 1995). DiVerences in prevalence exist between Mediterranean regions and islands that are geographically close but diVer in their genetic and cultural histories (Dean et al., 2002; Rosati 1994). A diagonal gradient in frequency is seen in North America with highest rates in the Midwest and lowest in the Mississippi delta (Bulman and Ebers, 1992). A latitudinal gradient is apparent for the white Australian population with higher rates in the south than north based on deWnitive survey of four regions using comparable methods and working to a common prevalence date (Hammond et al., 1988).

MULTIPLE SCLEROSIS IN MIGRANTS Populations are not stable, geographically or socially, and migrations involving relatively large numbers of people have aVected the distribution of multiple sclerosis. The message from these studies of migration has been to regard multiple sclerosis as an acquired disorder triggered by environmental conditions. The contrary view, derived more from the study of stable indigenous populations, is that the distribution depends on endogenous (i.e., genetic) susceptibility easily seen at the population level and increasingly apparent from the study of individuals. But the race versus place, nature versus nurture debate is increasingly sterile given the concession that however low the threshold for susceptibility, disease processes must still be triggered; for most, there is room both for genes and the environment in considering the etiology and pathogenesis of all complex traits, including multiple sclerosis. Comparisons of morbidity statistics for multiple sclerosis between racial groups in a single geographical setting Wrst achieved prominence with the studies of GeoVrey Dean in South Africa from the 1950s. Dean made two inXuential observations. First, higher agecorrected incidence and prevalence rates were seen in immigrants from Europe compared to South African English, Afrikaners, and Cape Coloreds (with mixed African and European ancestry)—in that order—and with no cases identiWed in African blacks. Second, within the English-speaking white population, those moving from northern Europe to southern Africa as adults took with them the high frequency of their country of origin, whereas those migrating below the age of 15 years showed the lower rates characteristic of native-born white South Africans (Dean, 1967). Subsequently, GeoVrey Dean and colleagues looked at the corollary to his pioneering studies from South Africa. A survey of immigrants showed prevalence rates for multiple sclerosis in the United Kingdom–born children of West Indians, Africans, and Asians approximating to those in equivalent Caucasian age groups (Elian and Dean, 1987; Elian et al., 1990). Various methodological problems arise in studying ethnic minorities within a large metropolis. It is easy to miss rare events and so underestimate the numerator whereas reliable census data are available on which to base the denominator; conversely, members of ethnic minorities aVected by an illness such as multiple sclerosis might tend to cluster in areas of maximum population density so as to beneWt from family and other community support, thereby inXating the apparent concentration of cases. But the trawl of immigrant cases has been widespread and it is more likely that the number of patients with multiple

MULTIPLE SCLEROSIS IN MIGRANTS

sclerosis, born in the United Kingdom of parents who were migrants from the West Indies, Africa, or Asia, has actually been underascertained. That said, it is wise not to rush to judgment by extrapolating from studies involving a small numerator living in an unusual environment. For example, three cases were identiWed among a cohort of around 3400 persons born of Vietnamese mothers who came to France under the age of 20 years providing, by 1975, an age-speciWc prevalence of 169/105 (95% CI 94-I35) for the third decade (Kurtzke and Bui, 1980). SuperWcially, this suggests a change from the inherent low risk of multiple sclerosis in Vietnamese, but the study is confounded by the fact that having a French-born father was a requirement for immigration; genetic admixture therefore introduced a major potential bias into the study. Another illustrative series of epidemiological studies compared the frequency of multiple sclerosis among Japanese living in Japan, Hawaii, and the West Coast of the United States. The prevalence among Japanese (7/105) in Hawaii was only one-quarter that seen in Caucasian immigrants (Alter et al., 1971; Detels et al., 1972; Lauer, 1994) and virtually identical to Japanese (and Caucasians) living in California (Detels et al., 1977). At that time, rates quoted for native Japanese were around 2/105 (Kuroiwa et al., 1983), and the evidence favored a strong protective eVect for Japanese irrespective of environment. But as in so many other parts of the world, subsequent surveys suggest that the baseline rates were underestimates and recent prevalence rates in Japan are higher both in the northern and in southern parts of the country. However, before reaching a Wnal position on the frequency of multiple sclerosis in Japanese, it is important to recall that deWnitions have changed. The classic view was that multiple sclerosis in Japan is characterized by an opticospinal phenotype with very few cases conforming to the pattern seen in Caucasians; now, the opticospinal pattern is seen less frequently, whereas the so-called Western features are on the increase (Fukazawa et al., 2000). Thus, alterations in deWnition and recognition may have contributed to the changing clinical and epidemiological pattern of multiple sclerosis in Japanese. A third location where migration shows age-related diVerences in prevalence of multiple sclerosis is Israel. Starting from the higher prevalence in migrants from northern Europe (Ashkenazis) than Asia and Africa (Sephardis), an age-at-migration eVect was apparent in Ashkenazi Jews with those arriving in Israel before adolescence underrepresented, compared to Sephardis (Alter, et al., 1962; Alter, et al., 1978). Age-adjusted prevalence in the Israeli-born children both of Ashkenazi and Sephardic Jews was the same as seen in the United States. Updated Wgures conWrm these observations—although, as expected, the numbers have changed (Kahana et al., 1994): after adjusting to the 1960 Israeli population and classifying on place of paternal birth, prevalences (per 105) were 32 (fathers born in Israel), 38 (fathers born in Europe or North America), and 29 (fathers born in Africa or Asia) compared with 14 in immigrants. The possibility of sampling error is raised by the evidence for regional variation in these diVerences—with higher rates overall and less diVerential in Jerusalem (where ascertainment was optimized) compared to more rural areas. Taken together, the evidence from the study of migrants supports the macroview that racially determined diVerences in risk for multiple sclerosis are modiWed by environment. Geographically more restricted movements support this hypothesis. Mortality with multiple sclerosis is higher for southern born United States citizens dying in the north (0.68/ 105/year) compared with those remaining in the south (0.46/105/year). For United States army veterans born in northern states but entering service from the middle and southern parts of North America, mortality ratios fall from 1.48 to 1.27 and 0.74, respectively. Taking those born in the middle states, mortality ratio decreases with entry into military service in the south and increases in the north (1.4 and 0.73, compared to baseline, respectively: Kurtzke et al., 1971). Attention needs to be directed at the Australian epidemiological studies, since these describe a large geographical area with a relatively homogenous population mix. The distribution of migrants from the United Kingdom and Ireland shows the same latitudinal gradient as for the entire population but with a bias introduced by the high prevalence for multiple sclerosis in the cohort from Hobart (Tasmania); that location apart, the rates are

705

706

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

lower than for contemporary studies from the United Kingdom but with no diVerences dependent on age at migration taking 15 years as the point of stratiWcation (Hammond et al., 2000). These studies are the epidemiological platform for suggesting a contribution from environmental factors in the etiology of multiple sclerosis, but they do not establish at what age these inXuences occur. The studies from South Africa and the United States argue for a critical period in childhood although it is diYcult to be more precise and the Australian survey somewhat undermines this tight window of microbial opportunity.

EPIDEMICS OF MULTIPLE SCLEROSIS John Kurtzke has been the main protagonist for the occurrence of point source epidemics of multiple sclerosis (Kurtzke, 1993). Skeptics prefer the position that these merely follow cycles of increased ascertainment due to local enthusiasm rather than genuine changes in incidence arising from the introduction of transmissible etiological factors into virgin populations. The fact that all the better-documented examples have occurred in small island communities in the north Atlantic can be seen to support either view. In the Wrst survey of Iceland, 168 cases of multiple sclerosis were identiWed with onset between 1900 and 1975 (Kurtzke et al., 1982). Incidence seemed retrospectively to have risen from about 1922, then stabilized until again increasing for around 10 years from 1945 and declining thereafter. Local observers point out the correlation between incidence and changes in health care provision with the arrival of specialists in neurology and serial reexamination of the morbidity statistics. According to this analysis, among 252 of 323 patients with onset of symptoms attributed to multiple sclerosis after January 1, 1900, who remained alive in December 1989, incidence rates were <1/105/year until the 1930s and 2.5/ 105/year thereafter, coinciding with the arrival of two neurologists; with waning enthusiasm, there then followed a lull until 1945–1954 when incidence increased to 3.3/105/year following the Wrst systematic survey of the disease; with nine neurologists in practice from 1975, there was a further increase to a peak incidence of 4.1/105/year, which has been maintained. Just as John Kurtzke takes support for a post–World War II epidemic from the lower age-at-onset between 1945 and 1954, John Benedicz holds up the long interval between onset and diagnosis, the disproportionate lack of benign cases (more easy to overlook) and overrepresentation of patients with severe multiple sclerosis before 1950 alongside the reduction in interval between onset and diagnosis after 1940 as evidence for the impact of neurological expertise on early diagnosis and improved case ascertainment making the Icelandic saga a virtual rather than a real multiple sclerosis epidemic (Benedikz et al., 1994). The epidemiology of multiple sclerosis in the Faroe Islands has assumed special signiWcance in considering the etiology of multiple sclerosis. The story begins with the identiWcation of fewer cases than expected from comparisons with neighboring Orkney and Shetland. Subsequently, despite diligent searches, Kurtzke was unable to identify any patient with an estimated date of onset earlier than July 1943. Among 83 individuals with multiple sclerosis resident in the Faroes during the 20th century, 55 had spent their entire life in the Faroes, or only short periods overseas, whereas 28 were either foreign born or had lived elsewhere for prolonged periods. Based on year of onset, these cases are considered to cluster into four epidemics involving 21, 10, 10, and 13 cases, and resulting from exposure between 1941–1944, 1945–1957, 1958–1970, and 1971–1983, respectively. Kurtzke has proposed that the factor determining multiple sclerosis in the Faroes was introduced by British troops between 1940 and1943, and the temporal and spatial distribution of the disease mapped nicely onto the pattern of villages where troops were billeted. The story is summarized by Kurtzke and Heltberg (2001). The Faroes hypothesis has not escaped criticism and counter-criticism with exchange of salvoes relating to validity of the diagnoses, exclusions, case ascertainment, the deWnition of epidemics, and the role of British occupation in triggering the putative epidemic.

EXOGENOUS FACTORS PREDICTING THE CLINICAL COURSE

Morbidity statistics for multiple sclerosis have never been higher than the rates for incidence and prevalence seen in the Orkney and Shetland Islands. Estimates of prevalence (per 105) carried out on four occasions between 1954 and 1974, showed an increase from 111 to 309 in Orkney and 134 to 184 in Shetland (Cook et al., 1985; Robertson and Compston, 1995). Although the overall frequency of the disease is without precedent, the evidence suggests that the incremental change in incidence and mortality are attributable to alterations in classiWcation and ascertainment: during these decades, systematic depopulation in Orkney and Shetland left an aging population no longer at risk of multiple sclerosis; the rise in prevalence relates to increased survival (by approximately one decade) and better case recognition. Perhaps, the more recent updating of morbidity statistics does suggest signiWcantly less multiple sclerosis in Shetland, but not Orkney, since 1965. Outside the north Atlantic, an epidemic has been claimed for Key West (south west Florida) where 37 patients with peak onset in and around 1977–1979 were identiWed in 1984 (prevalence 140/100,00), suggesting an exceptional frequency compared to neighboring parts of the United States not attributable to increased clinical vigilance or diVerential migration of symptomatic individuals to a more favorable climate (Sheremata et al., 1985).

EXOGENOUS FACTORS PREDICTING THE CLINICAL COURSE Multiple sclerosis is almost always found to be more common in females than males in population based surveys, with a ratio of at least 2F:M in all ethnic groups. This distortion is even more marked in children with multiple sclerosis (Duquette et al., 1987) but reversed in the context of late onset disease which is more common in males who show overrepresentation of the primary progressive phenotype. Their prognosis is less favorable that younger age-at-onset cases, females with multiple sclerosis, and cases with a phenotype characterized by infrequent sensory episodes that recover fully. There have been many attempts to categorize the course and prognosis of multiple sclerosis; this reductionist approach is tricky in the context of a disease that is notoriously Wckle in its behavior, but the pattern of primary progressive multiple sclerosis in 20%, relapses with full remission in the remaining 80% of cases at onset, with conversion over time to persistent deWcits and then secondary progression in 60% of this group, is a reasonable caricature. Against this background of the natural history, with all its uncertainties, various extrinsic factors have been considered that might alter the natural history either by exposing latent disease, precipitating disease activity in those with established multiple sclerosis or changing the later course of the illness. There is no epidemiological evidence to suggest that multiple sclerosis is contagious. After correction for age and disease duration, and anticipation of the diagnosis in the second aVected through enhanced awareness, Robertson et al. (1997) showed no clustering for year at onset in conjugal pairs, suggesting that multiple sclerosis had not been passed from one spouse to the other. A recent study purporting to provide evidence that the disease is sexually transmitted and attributable to abuse when onset is in childhood (Hawkes, 2002) is speculative, lacks scientiWc rigor, and is deeply oVensive to many aVected individuals, especially those forming part of multiplex families and their relatives. But what is the environmental trigger? It is a rare moment in the history of research in multiple sclerosis for there not to be a current hot microbial favorite for the cause of multiple sclerosis based either on population serology, identiWcation of particles in one tissue or another, or recovery of deWned organisms from body Xuids. Few have stood the test of time and most have disappeared from interest often due to the subsequent recognition of technical artifacts and as each is displaced by new and more compelling candidates. Although individual reports continue to describe an excess of antibodies to chlamydia pneumonia in the cerebrospinal Xuid of patients with multiple sclerosis, perhaps the most telling study is a comparison between three laboratories of polymerase chain reaction detection rates for chlamydia pneumonia in the same samples; one (the laboratory reporting the original Wndings: Sriram et al., 1999) found the organism in the majority of patients but

707

708

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

few controls; the other two were unable to detect chlamydia in any samples (Kaufman et al., 2002). HHV-6 remains a candidate (Ablashi et al., 1998). Friedman et al. (1999) reported an excess of HHV-6 structural protein and antibody to HHV-6 in serum and brain tissue in multiple sclerosis compared to controls; Soldan et al. (2000) have shown increased peripheral blood lymphocyte reactivity of the neurotrophic HHV-6A variant. The issue of vaccination has recently been resolved: three studies have shown no increase in demyelinating disorders or activity of preexisting disease in individuals undergoing hepatitis B vaccination (Ascherio et al., 2001; Confavreux et al., 2001; Zipp et al., 1999). The United States study showed a relative risk for the onset of multiple sclerosis of 0.7 (95% CI 0.3-1.8) in the 2 years after hepatitids B vaccination (Ascherio et al., 2001); the French study (Confavreux et al., 2001) showed a relative risk of 0.71 (95% CI 0.40-1.26) for relapse in the 1-, 2-, or 3-month period following vaccination (mainly for hepatitis B but also tetanus and inXuenza). Trauma as a trigger of disease activity in multiple sclerosis has been much considered and still features in the courts where plaintiVs may claim that an accident has provoked the Wrst appearance of multiple sclerosis or altered the course of preexisting manifestations. Sibley et al. (1991) prospectively studied disease activity by questionnaire and physical examination for 8 years: taking either the 3- or 6-month period following each event as atrisk, only electrical trauma showed an association with new episodes; in fact, all other forms of trauma were negatively correlated both with clinical exacerbations and disease progression. Siva et al. (1993), using the Mayo Clinic cohort, also concluded that disease exacerbations are no more frequent in the 6 months after limb fracture than at other times. The situation with respect to the risk from anesthesia is unresolved in that no epidemiologically based studies have been performed and the evidence is entirely anecdotal; some neurologists advise patients to avoid elective interventions while remaining sensible about treatments or procedures that justify the small risk of increasing disease activity—if it actually exists (Baskett and Armstrong, 1970; Siemkowicz, 1976). Several authors have demonstrated prospectively that new episodes of demyelination increase after (presumed) viral exposure, but no single agent has been implicated (Anderson et al., 1991; Sibley et al., 1985): 9% of presumed infections are followed by relapse, and 27% of new episodes are related to infection; the relative risk for relapse in the 4-week period after upper respiratory (especially adenovirus) or gastrointestinal infections is 1.3. Most recently, Buljevac et al. (2002) prospectively studied 73 patients with relapsing multiple sclerosis experiencing 167 infections during 6466 patient-weeks (1.7 years) and observed a twofold increased relapse rate, and greater risk of lasting clinical deWcits, during the at-risk period (
2 to þ5 weeks), associated with markers of immune activation but no change in permeability of the blood brain barrier. Rather softer, but nevertheless consistent, is the evidence that new episodes of demyelination—especially optic neuritis—cluster in the au1 spring, at least in northern Europe (Jin et al., 2000). Another attempt to link the epidemiological patterns to speciWc infections has recently been revisited—and with broadly similar conclusions. Compston et al. (1986) retrospectively compared historical and laboratory evidence for previous virus exposure in patients with multiple sclerosis and optic neuritis with HLA-DR-matched controls. Patients with demyelinating disease had a later age at infection by measles, mumps, and rubella. Martyn et al. (1993) later reanalyzed this series to show an even more marked eVect of EBV infection; those who reported having infectious mononucleosis before the age of 18 years had a relative risk for multiple sclerosis of 7.9 (95% CI 1.7-37.9). Meanwhile, Lindberg et al. (1991), looking at registries for infectious mononucleosis and multiple sclerosis in Sweden, found a slight excess of demyelinating disease following EBV infection compared with controls. Haahr et al. (1995) used records from the Danish State Serum Institute register of EBV infections and the Danish Multiple Sclerosis Registry to identify 16 individuals who developed multiple sclerosis among 6853 experiencing EBV infection in the decade 1968–1978 (relative risk, 2.8); the median age at infection was 17 years, which did not diVer from the age at infection in heterophile antibody-positive individuals who did not develop multiple sclerosis. Vaughan et al. (1996) used more discriminating laboratory markers and also showed an enhanced immune response to EBV, involving epitopes also

ANALYZING THE EPIDEMIOLOGICAL PATTERN IN MULTIPLE SCLEROSIS

expressed by other pathogens one of which appeared to cross-react with (murine) glia cells, in patients with multiple sclerosis, especially during relapse. Now, Hernan et al. (2001) have again shown an increased of multiple sclerosis following infectious mononucleosis (odds ratio 2.1: 95% CI 1.5-2.9) and older age-at infection by mumps and measles. Anecdotal evidence on whether pregnancy aVects the immediate or long-term course of multiple sclerosis has been supplemented by prospective surveys. A major confounder in the detailed interpretation of these studies is the decision by women with severe disability not to embark on pregnancy and the corresponding preparedness of those with mild disease to start or extend their families. Onset of multiple sclerosis does not cluster around pregnancy, and having children does not alter the long-term course of the disease, but there is an increase in relapse rate during the puerperium. The studies indicate approximately a threefold higher risk in the 3 to 6 months after term than during pregnancy and suggest that the attacks may be more severe; but since there is evidence for reduction in relapse rate during the pregnancy itself, the high puerperial rate makes for no overall increase in disease activity over the entire pregnancy year. Runmarker and Andersen (1995) studied an inception cohort in Goteborg, Sweden, and disposed of the hypothesis that the onset of multiple sclerosis is inXuenced by pregnancy; in fact, there was a conspicuous absence of onset bouts during pregnancy compared with nonpregnant epochs including the puerperal 8 months. Fecundity was reduced in women with multiple sclerosis, presumably by choice especially in the context of signiWcant disability and this is the probable explanation for the conclusion that pregnancy after onset is associated with a lower risk of progression. A prospective survey (PRIMS: PRegnancy In Multiple Sclerosis: Confavreux et al., 1998) recently surveyed 222 completed European pregnancies; annualized relapse rates in the four quarters before pregnancy were 0.63, 0.74, 0.70, and 0.81, respectively, falling to 0.47, 0.58, and 0.22 during the three trimesters of pregnancy, and rising to 1.23 and 0.88 in the Wrst 6 months of the puerperium. Confavreux et al. (1998) suggested that the perturbation in disease activity around the time of parturition relates to T-helper (Th)-2-like cytokine performance of the placenta.

ANALYZING THE EPIDEMIOLOGICAL PATTERN IN MULTIPLE SCLEROSIS The interplay of nature and nurture is reXected in the distribution of many diseases but factors determining the geography of complex traits remain to be deWned. As the global pattern of multiple sclerosis has emerged, ideas have also evolved on the interplay of factors that shape the inXuence of race and place. At Wrst, the accumulated evidence from serial surveys of large populations deWned gradients in frequency within the high prevalence regions: northern Europe, North America, and Australia. These patterns suggested a dominant environmental eVect on the etiology of multiple sclerosis. This interpretation was supported by the surveys of multiple sclerosis in migrants from northern Europe showing, with few exceptions, lower prevalence rates than the country of origin—and by the putative epidemics. But the search for an environmental cause of multiple sclerosis remains stubbornly unproductive. Systematic screening for candidate infectious agents using population serology and sophisticated methods for virus detection has not revealed the microbial cause for multiple sclerosis. And, as a result, increasingly complex alternative theories have been advanced involving the role of climate, diet, geomagnetism, sunlight, air pollutants, radioactive rocks, and toxins to account for the global pattern of the disease. Failure to deWne the environmental cause therefore led others to interpret the distribution of multiple sclerosis as a function of genetic susceptibility. The geography of multiple sclerosis matches the distribution of northern Europeans. Racial groups thought to be at low risk of the disease begin to show increased frequencies as their peoples move and mix with northern Europeans (Chakraborty et al., 1992). Coming at the issue of ancestry from various directions (physiognomy, social history, surname—among other indirect clues), commentators have argued for nearly 100 years that the distribution arises because

709

710

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

au2

multiple sclerosis is a Nordic-Celtic disease: credit should go to Davenport (1922) for the original observation, to Sutherland (1956) for the Wrst systematic demonstration of diVerential rates depending on ancestry in the same geographical region, to Poser (1994) for tying this hypothesis into the rich history of the Viking sagas, and to Skegg et al. (1987) for the amusing trick of mapping multiple sclerosis to the density of surnames with the preWx ‘‘Mc..’’ in the telephone directories of New Zealand. Further support for a geo-genetic solution to the distribution of multiple sclerosis is provided by maps showing a reasonable correlation between global or regional prevalence and genetic clines for markers of susceptibility in at-risk populations (Cavalli-Svorza et al., 1994). At a more micropopulation genetics level, isolates have arisen as a consequence of sociohistorical events and population migrations, leading to disproportionate concentration of susceptibility genes in small communities (with others excluded) thus adding relic groups to the global pattern that exists for many polymorphic genetic markers. Based on the putative epidemic of multiple sclerosis in the Faroes, John Kurtzke (Kurtzke, 1993) concluded that multiple sclerosis originated in Scandinavia (central Norway or the south-central lake district of Sweden) in the early 18th century and diVused across the Baltic states and northern Europe including the British Isles over the next 100 years. From there, it was exported to North America and Australasia, to South Africa and Italy. It is an attractive hypothesis, but whereas for Kurtzke the factors being distributed are germs, for others they are genes.

FAMILIAL MULTIPLE SCLEROSIS The concept of genetic susceptibility to multiple sclerosis began with observations on diVerences in racial susceptibility and gained momentum with descriptions of familial clustering. Many interesting pedigrees were reported, but at Wrst their collection did not follow sound epidemiological principles and the studies were therefore subject to ascertainment bias. From the 1980s, the frequency and pattern of familial multiple sclerosis has been documented in population-based samples. These large cohorts have been used systematically to document the frequency of familial clustering, thereby consolidating the case for genetic susceptibility in multiple sclerosis and illuminating many aspects of its complexity. Multiplex families can be used to assess the extent to which genetic factors determine clinical features of the disease making the distinction between susceptibility and inXuences on the course. Access to family pedigrees provides the opportunity for laboratory studies aimed at mapping, identifying, and describing the function of the several genes that are assumed to make a contribution based on the pattern of familial clustering, individually and through epistatic interactions. The applications of that knowledge are several: to provide recurrence risks for individual categories of relatives, to prevent the disease by intervention in high risk individuals, to understand the mechanisms of tissue injury, to predict the long term course of the illness, and to select individuals more or less likely to beneWt from treatments. Of these, the most immediate and realistic application is improved understanding of the pathogenesis and the design of novel therapies. In using multiplex families to explore genetic aspects of multiple sclerosis, diagnostic criteria for inclusion may deliberately be adjusted, estimating recurrence risks and performing analyses under conditions of varying clinical stringency. For example, since the majority of patients with isolated optic neuritis eventually convert to multiple sclerosis, these may reasonably be assigned aVected status in genetic studies whereas clinical phenotypes with a lower recurrence risk and a less clear relationship to multiple sclerosis should be treated diVerently. Multiple sclerosis has a familial recurrence rate of approximately 15%. A summary of recurrence risks for various categories of family member is shown in Figure 30.2. Although it is usually assumed that familial aggregation is due to coinheritance of susceptibility factors, the case has to be argued and the alternative is that this results from common exposure to environmental factors in childhood. The initial study of recurrence (from

711

FAMILIAL MULTIPLE SCLEROSIS

Recurrence risks for multiple sclerosis Genetic sharing

Relationship

MZ twin

100%

Sibling, 2 aff.parents Sibling, 1 aff. parent

DZ twin

50%

Sibling Parent Child Half sibling 25%

Aunt / uncle Nephew / niece Cousin

12.5%

Adoptee

0%

General population

0

5

10

15 20 Lifetime risk (%)

25

30

35

FIGURE 30.2 Age-adjusted recurrence risks for diVerent relatives of probands with multiple sclerosis. Pooled data from population-based surveys. Estimated 95% conWdence intervals are shown. Figures on left show the degree of genetic sharing between relative and proband. Kindly prepared by Dr Simon Broadley. Reproduced from Compston, D. A. S., and Coles, A. J. (2002). Multiple sclerosis: Seminar. Lancet 359, 1221–1231

Canada) took as its baseline a lifetime risk of 0.2% for the entire population, and showed an increase to 3% in other Wrst degree relatives (relative risk 20) and 1% in second degree relatives (relative risk 5.5: Sadovnick et al., 1988). The United Kingdom survey conWrms that the highest age-adjusted recurrence rate is for sisters (4%) and brothers (3%), compared with parents (2%) and oVspring (2%). Overall, the reduction in risk changes from 3% (relative risk 9) in Wrst-degree relatives to 1% (relative risk 3.4) and 1% (relative risk 2.9) in second- and third-degree relatives, respectively, compared with a background age-adjusted risk in this population of 0.3% (Robertson et al., 1996b). In the remaining large survey from Belgium, recurrence risks were also 10-12-fold for Wrst degree and threefold for second-degree relatives (Carton et al., 1996). With some variations in methodology, three contemporary studies have described the concordance rates for multiple sclerosis in twins identiWed from a population base. Surveys from Canada (Sadovnick et al., 1993) and the United Kingdom (Mumford et al., 1994) show a higher clinical concordance rate in monozygotic (approximately 25%) than dizygotic twin pairs (about 3%). The outlier, from France (French Research Group on Multiple Sclerosis (1992), did not identify a diVerence between monozygotic and dizygotic twins although this result was within the conWdence limits of the two other surveys. Taken together, the relative risk for multiple sclerosis in the monozygotic twin partner of an aVected proband is around 190. Adopted individuals who subsequently develop multiple sclerosis, and probands who have themselves adopted children, provide a rare but informative resource for studying the relative contribution of genes and the environment in causing multiple sclerosis. Considering individuals with multiple sclerosis adopted in infancy, and those having adopted siblings or children, the frequency of multiple sclerosis in these social

712

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

relatives matches that seen in the general at-risk population and not the recurrence rates expected for the biological Wrst-degree relatives of index cases (Ebers et al., 1995). Halfsiblings share a proportion of parental genes and stratify into those reared together and apart during the period considered critical for the development of multiple sclerosis. Their age-adjusted risk of multiple sclerosis is signiWcantly lower than for full siblings, and there is no diVerence in risk for half-siblings reared together and apart (Sadovnick et al., 1996). Two studies of conjugal pairs with multiple sclerosis who have children provide a special opportunity for assessing the eVect of genetic loading on susceptibility. Five of 86 oVspring from 45 conjugal pairs living in the United Kingdom also had multiple sclerosis, and characteristic imaging abnormalities or clinical symptoms consistent with demyelination but not meeting criteria for clinically deWnite disease were seen in Wve others (Robertson et al., 1997). Ebers et al. (2000b) identiWed 23 conjugal cases among the spouses of 13,550 probands; 6 of their 49 oVspring were aVected providing an age-corrected recurrence risk of 30%. Although subject to ascertainment bias, the lack of an increased rate of multiple sclerosis among spouses—compared to the at-risk population—and recurrence in oVspring approaching that seen in monozygotic twins emphasizes that the familial risk is genetically determined and not due to a shared environment. Although these studies favor early ascertainment, the young age at onset suggests that genetic loading does count; inheritance of factors from both parents provides evidence for epistatic eVects or homogeneity in the contribution made by genetic factors.

THE ANALYSIS OF COMPLEX TRAITS A remarkable change in the ability to characterize the genetic basis of complex traits such as multiple sclerosis has occurred over the past two decades. In the early 1980s, activities were limited to comparing the phenotypes of those few genetic systems that were identiWed and known to be polymorphic—blood groups, the major histocompatibility complex, immunoglobulin isotypes, and a few others. In the mid-1980s, restriction fragment length polymorphisms opened up the opportunity to characterize genotypes at sites where genes were also known to be located. Thereafter, the cataloging and mapping of microsatellite markers widely distributed across the genome, but usually in the intervals between coding regions and made up of repeats involving two or more nucleotides increased the range and diversity of genotype deWnition. Widely distributed single nucleotide polymorphisms now provide a suYcient density of markers in and around genes to oVer a deWnitive approach for indirect screens. Ultimately, coding sequence can be used to provide a direct screen for each and every gene making it no longer necessary to rely on markers, however closely placed. But the availability of reagents brings new challenges for the development of technologies needed to derive these genotypes and statistical methods for their analysis. Two methods—linkage and association—underpin the analysis of complex traits (Risch, 1990, 2000; Risch and Merikangas,1996). Their principles are illustrated in Figure 30.3. Strategies for detecting susceptibility genes have moved from case-control studies seeking associations (relatively easy to organize but vulnerable to mismatch between cases and controls) to family-based association studies comparing the distribution of alleles between aVected and nonaVected Wrst-degree relatives (usually family trios or quartets made up of cases, parents, or siblings; harder to collect in suYcient numbers and of low power given the overmatching of susceptibility factors but not vulnerable to erroneous selection of controls) to family-based linkage analysis. The choice of markers has either been driven by a priori guesses on the possible nature of susceptibility (candidate genes), systematic screening of whole chromosomes or the entire genome, or identiWcation of candidates known to map within regions of chromosomal interest (positional candidates). Markers may be analyzed individually (single point analysis); alternatively, the probability that a marker is linked or associated with multiple sclerosis is corrected based on information available from juxtaposed markers (multipoint analysis). Finally, the cumulative probability that a given marker or region of interest is linked or associated with multiple sclerosis in

713

THE ANALYSIS OF COMPLEX TRAITS

A

B

C

Distant marker Linked not associated

Small sample Associated not linked

Large sample Linked and associated

FIGURE 30.3 (A) Linkage is present when two or more aVected relatives share parts of the genome encoding susceptibility genes, identiWed by adjacent markers, more often than expected by chance. Linkage has low statistical power but is able to track the passage of extended haplotypes. (B) Association is present when people with multiple sclerosis have a higher frequency of alleles, and markers in linkage disequilibrium, increasing susceptibility than unaVected individuals. This has higher statistical power but depends on linkage disequilibrium, which decays with genetic time (meioses) and mutation of the (microsatellite) markers. (C) Markers distant from the susceptibility gene may be linked but not associated, being outside the range of linkage disequilibrium. Markers close to the susceptibility gene may be associated but not linked if the sample size of multiplex families is small. Close markers, and disease susceptibility alleles themselves, will be linked and associated.

several studies can be formally tested by meta-analysis. Despite these advances in methodology and range of strategies, progress in identifying genes or even regions conferring susceptibility to multiple sclerosis has been slow. The lesson learned thus far is that no one gene makes a major contribution to susceptibility although collectively they determine a relative risk (for siblings) of around 20. The classical Mendelian and biometric views of inheritance distinguish diseases attributable to the eVect of single mutant alleles from those inXuenced by a large number of traits, each exerting only a minimal eVect. Mutations are genetic variations in coding sequence occurring in <1% and polymorphisms are the alleles occurring in >1% of the population. Polymorphisms may therefore be rare or common in the population, with either category exerting trivial or substantial biological eVects on disease mechanisms. It follows that disease frequency will depend on the interplay between frequency of susceptibility alleles and

714

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

au3

magnitude of their biological eVects. When a small family population expands numerically in relative isolation, genetic drift can inXuence the frequency of mutant and polymorphic alleles. As a consequence, some genetically determined diseases may occur at higher than expected frequency. Isolated populations such as Finns, Sardinians, Icelanders, and Tasmanians could in theory provide special opportunities for resolving the issue of genetic susceptibility to complex traits such as multiple sclerosis. Linkage is seen when two or more aVected relatives share parts of the genome encoding susceptibility genes and nearby markers for a complex trait more often than expected by chance: It has low power but operates over a wide genetic distance. Association is present when people with the disorder have a higher frequency of alleles and nearby markers of factors increasing susceptibility than unaVected individuals: association has high statistical power but depends on the presence of linkage disequilibrium and therefore operates only over short genetic distances. Linkage analysis and association mapping either select candidate genes based on a priori considerations of disease pathogenesis or perform systematic screens of the whole genome. During meiosis, juxtaposed segments of paired chromosomes recombine such that the gametal chromosome is a mixture of the two parental copies. Linkage exploits the fact that susceptibility genes and juxtaposed markers will be carried in one or more of these recombined segments or hapotypes. The breakpoints vary with each meiosis and the size of blocks carrying the susceptibility genes will also therefore vary. Markers placed close to each susceptibility gene remain within the blocks inherited by each aVected individual and so they are linked. Linkage deteriorates with distance of the marker from the susceptibility gene until it is no longer present in that family. But because the passage of genes and linked markers is being studied across very few meioses, the topographical eYciency of linkage is high and relatively few markers (i.e., relatively wide separation between those selected for a linkage screen) are needed to stand a good chance of detecting linkage. Even though the marker allele present in each family may diVer, the locus remains linked to the gene inXuencing susceptibility in these families. However, it turns out that linkage performs poorly for genes of small biological eVect and the number of families needed to show that a particular marker (allele or locus) is being inherited in multiplex families more often than expected by chance is often prohibitively large. The main failure of linkage analysis, despite economy of genotyping in the laboratory, is therefore small sample size. Where extensive genome screens fail to provide evidence for linkage, rare genes with large eVects probably do not exist for that disease. The power of a linkage genome screen is critically dependent on the frequency of susceptibility alleles in the population studied and is thus expected to vary between studies. Negative studies are not necessarily deWnitive and linkage analysis may still be worth pursuing in locations or populations where the frequency of susceptibility alleles is more favorable for historical reasons. Any allele actually encoding a product that causes a disease process (as in mutations of genes for Mendelian disorders) must be linked and associated with that disorder or component of the disease mechanism. But the usefulness of association lies in the fact that, for reasons of population history, susceptibility genes making even a small contribution to the disease process can also be identiWed by screening nearby anonymous markers making it no longer necessary to hit the disease locus itself. Association depends on the phenomenon of linkage disequilibrium. Because association and linkage refer to diVerent aspects of genetic analysis—each providing entirely separate but complementary categories of information—it is unfortunate that the linkage word has crept into both deWnitions; this causes much confusion. When polymorphisms that increase susceptibility to disease arise in a founder group, these will necessarily be located within a large group of linked genes. This block is subject to recombination during subsequent meioses and is gradually whittled down. Markers close to the susceptibility gene are the last to become separated. It follows that the progeny of this founder continue to share segments of DNA identically by descent for many generations. Their genes are said to be in linkage disequilibrium. In time, the extent of linkage disequilibrium degrades until it is nonexistent and it is then no longer possible to detect susceptibility genes by association mapping because the population is now in linkage equilibrium.

THE ANALYSIS OF COMPLEX TRAITS

This is the rule for older populations such the ‘‘out of Africa’’ races. But for younger populations that went through and survived population bottlenecks, the residual degree of linkage disequilibrium may be considerable and suYcient to use markers remote from the disease promoting polymorphism to track the genetic basis for susceptibility. However, mutation of the marker for a disease associated susceptibility allele also decays linkage disequilibrium and so reduces the ability to detect associations between that marker and the disease. Microsatellites mutate once every 1000 generations; those that mutate soon after appearance of the disease susceptibility allele will signiWcantly reduce the number of chromosomes carrying both the marker and the susceptibility gene in the expanded present-day population, whereas more recent mutations will be much less disruptive. Estimates for the extent of linkage disequilibrium are a moving target, but crucial in assessing the density of markers needed to explore a given region for association. At Wrst, crude theoretical modeling of human population history suggested that variants which are common in the population as a whole will generally be very old and therefore accompanied by rather little linkage disequilibrium (Krugylak, 1999). However, empirical evidence in Europeans started to show that linkage disequilibrium is in fact much more extensive than these predictions. Initial average distances for the extent of linkage disequilibrium ranged from 60 kb to 385 kb (Collins et al., 2001; Reich et al., 2001) for SNPs and 0.5 Mb to 1 Mb for microsatellites (Eaves et al., 2000; Kendler et al., 1999). But it soon became clear that even though many parts of the genome remain to be explored, linkage disequilibrium is highly variable and therefore not uniformly accessible to association screening (Dunning et al., 2000; Martin et al., 2000; Taillon-Miller et al., 2000). Subsequently, a diVerent extent and structure of linkage disequilibrium in Europeans has been shown from that predicted. Recombination is not uniformly distributed, as assumed, but rather is concentrated in hot spots separating regions of marked linkage disequilibrium (Daly et al., 2001; JeVreys et al., 2001). These blocks of extensive linkage disequilibrium with sharp boundaries vary in size but average around 20 kb (Daly et al., 2001). It follows that the limited haplotype diversity within blocks favors testing these regions (but not the adjacent boundaries) for association with appropriately informative markers (Johnson et al., 2001). A Wrst-generation linkage disequilibrium map of human chromosome 22 in Centre d’Etude du Polymorphisme Humain (CEPH) families shows extremely variable linkage disequilibrium with blocks up 804 kb interspersed with regions of little or no detectable linkage disequilibrium. In these long blocks, 76% of human diversity could be deWned using just three tagging SNPs to deWne the Wve common haplotypes; elsewhere, much more extensive typing would be needed to screen short haplotypes (Dawson et al., 2002). Association has much greater power to detect genes of small biological eVect than linkage, and it is logistically easier to collect cases and controls than multiplex families. The downside of association screening is that the range of linkage disequilibrium is much narrower than for linkage; it degrades faster and the density of markers needed to cover a region of interest is necessarily much greater. For high-density screening, genotyping eVort can be reduced using equal amounts of DNA from a large number of cases, controls, or parents pooled into a single sample and screened against a panel of markers (Barcellos et al., 1997). Pooling introduces several additional potential sources of error in the analysis of complex traits (miscalling of alleles due to the presence of stutter bands; and the artifact of length-dependent ampliWcation, whereby short alleles are preferentially ampliWed in the polymerase chain reaction skewing the apparent distribution of alleles seen with individual typing and either favoring or inhibiting the demonstration of associations resulting from short and long alleles, respectively); but allowing for these technical limitations, pooling can detect allelic frequency diVerences of around 5%. So association is easy on the Weldworkers but tougher on the laboratory staV. Linkage strains clinical resources for collection but is lighter on laboratory eVort; oV-the-shelf statistical programs are now developed for analyzing either, but their further reWnements require continued biostatistical wizardry. Associations can be demonstrated using population or family-based studies. Ethnic mismatching has proved a major limiting factor for population studies especially in large

715

716

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

outbred populations. Family-based methods aim to show that a marker allele is transmitted to aVected individuals more often than expected by chance. Transmission disequilibrium testing (TDT) uses trios consisting of single aVecteds and both parents, who are usually (but need not be) unaVected. Each parental allele has a 1:2 chance of being transmitted. The test provides evidence for allelic association by demonstrating excess transmission and identiWes that allele which is responsible for disequilibrium in the sample. Although not every family will be informative for this marker, those in linkage disequilibrium with susceptibility genes for multiple sclerosis will be adequately represented. One further factor limiting the extent to which a sample of family trios may not prove fully informative for a particular polymorphic locus is parental homozygosity since the transmitted allele cannot then be identiWed. Segregation distortion arises when a particular allele confers a survival advantage and therefore appears to be disproportionately transmitted; in fact, the allele is preferentially transmitted to all surviving oVspring whether or not they have the disease in question. This issue can be resolved by studying aVected and unaVected children. But a new problem then arises, which does not confuse the aVected family member approach, and that is diYculty in assigning unaVected status to a young adult in a disease that may not manifest clinically until late in life, if at all, given the prevalence of pathologically veriWed but clinically silent disease in autopsy series. Selecting probands from among families in which there are multiple cases increases the power and reduces the scale of the study especially in the context of rare alleles. Associations may be with a deWned allele or locus. Those that are reproducible between populations carry conviction with respect to their causal relevance, but population speciWcity does not exclude causality. Where there is reasonable but variable linkage disequilibrium and several markers are found to be associated in separate populations, cross-cultural studies (exploiting diVerent degrees of linkage disequilibrium) may narrow the region of interest and select the closest marker; for example, Caucasians show an association between narcolepsy and both DR and DQ alleles, but only the DQ association is found in Africans (Rogers et al., 1997). Linkage and association provide diVerent categories of information. Assuming a reasonable sample size, any marker placed well distant from a susceptibility gene may be linked but not associated since it will lie outside the region of linkage disequilibrium. A marker that is close to a susceptibility gene may not be linked if the sample is underpowered but it may nevertheless be associated. Only the combination of a close marker tested in an adequate sample will show both linkage and association. If these conditions are met in independent tests, the candidature of a given chromosomal region is much enhanced. Many genes are mapped but have no assigned function. This is the obvious next step once a susceptibility locus has been identiWed (Eisenberg et al., 2000). The function of genes can be assessed by computational methods using similarities and dissimilarities between the phylogenetic expression of two proteins in diVerent organisms with fully sequenced genomes. The knowledge that two gene products are fused proteins in another species implies a generic sharing of function. Genes that are linked on the same chromosome in more than one species tend to have related functions. Those showing the same pattern of increased and decreased expression in diVerential displays are likely to share function providing guilt by association. The increasing availability of reagents for distinguishing short sections of DNA, accurate mapping of these microsatellites and single nucleotide polymorphisms (SNPs) across the genome, access to extended pedigrees in which aVected status can reliably be determined, the collection of large cohorts of cases and controls, and the deployment of appropriate statistical methods for assessing statistical signiWcance across large data sets collectively have made it possible to tackle the genetics of complex traits. Meta-analysis of linkage and association screens has been explored in the expectation that this will reduce the evidence for false positive results and strengthen the candidature of those that are genuine in order to select regions for more detailed studies of positional candidates. Combining data sets from ethnically diverse groups necessarily selects for genes conferring susceptibility between populations and obscures the identiWcation of others restricted to particular populations. For a disease that is notoriously variable in its clinical pheno-

CANDIDATE GENES IN MULTIPLE SCLEROSIS

type and in which a case can be made for heterogeneity in the pathogenesis, lumping all cases under the rubric of ‘‘multiple sclerosis’’ reduces the power to identify any one susceptibility factor and restricts the analysis to susceptibility genes ignoring those that may focus the pathological process on deWned pathways, shape the clinical course, and (perhaps) determine the response to treatment. Since these distinguishing features do not segregate within families, linkage is essentially restricted to the study of susceptibility genes. Association mapping can in theory be used to reveal either susceptibility genes or those inXuencing the clinical course if appropriate stratiWcations are employed in designing the study.

CANDIDATE GENES IN MULTIPLE SCLEROSIS

au4

Much eVort has gone into the assessment of candidate susceptibility genes chosen on the basis of prevailing ideas concerning the pathogenesis of multiple sclerosis and therefore providing a logical basis for their role in shaping the pathogenesis. It is hard to argue that progress in advancing knowledge has not been painfully slow since the burst of activity in the early 1970s, which established the range of associations with alleles of the major histocompatability complex—remarkable for its extensive linkage disequilibrium. As expected, it proved much more diYcult formally to demonstrate that the region is also linked to multiple sclerosis in family studies, as of course it must be if the association is genuine. To summarize the present situation, population studies comparing unrelated cases and controls show an association between the class II major histocompatibility complex alleles DR15 and DQ6 and their underlying genotypes (DRB1*1501, DRB5*0101 and DQA1*0102, DQB2*0602) (Olerup and Hillert, 1991). This is seen in almost all populations (Caucasian, Oriental, Arab, Hispanic, Finnish, Russian, and Jewish), although the strengths of the association diVer. Even those ethnic groups in which the frequency of multiple sclerosis is low, or the phenotype distinct from that usually observed in northern Europe, are now acknowledged to be DR15 associated with one regional exception. In Sardinians, there is a speciWcally diVerent association with DR4 (DRB1*0405-DQA1*0301-DQB1*0302: Marrosu et al., 1992). In the Canaries, the primary association is with DR15 and DQ6, but a secondary association also exists with DR4 (Coraddu et al., 1998). In Turkey, there is an allelic association with both DR2 (DBR1*1501, DQA1*0102, DQB1*0602) and DR4 (DRB1*04, DQA1*03, DQB1*0302: Saruhan-Direskeneli et al., 1997). Thus, there is a DR4 Xavor to the association in some southern Mediterranean populations. But since this Mediterranean DR4 haplotype is not identical, opportunities exist to narrow the region. That said, most investigators assume that—based on the genetics and obvious candidature through its role in restricting the immune response—DR (or DQ) is itself the susceptibility gene encoded at 6p21. Outside the major histocompatibility complex, the list of immunologically related intelligent or positional candidates now screened includes >50 adhesion molecules, immune receptors, accessory molecules, cytokines, and chemokines and their receptors or antagonists (selected references include He et al., 1998; Jacobsen et al., 2000; Kuhlman et al., 2002; McDonnell et al., 2000; Modin et al., 2001; and Reboul et al., 2000). In the isolated population of Finns, association and linkage of susceptibility to multiple sclerosis exist for the myelin basic protein gene encoded on chromosome 18 (Tienari et al., 1992). Even in this cohort from Vaasa, the eVect can be traced to a subset of families with common ancestry and does not hold up in the larger cohort (Pihlaja et al., 2003). But opportunities now exist for tracing the genealogy of this isolate, linking it to other regions through population history and so seeing more clearly the genetic path explaining one particular route to increased disease susceptibility. Studies of structural genes of myelin have otherwise been uninformative (Seboun et al., 1999), as have eVorts directed at implicating growth factors determining oligodendrocyte development and their receptors in susceptibility (Mertens et al., 1998). More than 35 growth factors or genes encoding structural components of the nervous system have been assessed as candidates. The current list of candidates is shown in Figure 30.4; those for which results are entirely negative, and therefore excluded within the limits of sample size, are distinguished from the smaller number of candidates for which a role in susceptibility or inXuences on the course

717

718

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

Candidate gene studies in multiple sclerosis Susceptibility

Course

HLA-DR/DQ

(Severe)

haplotype

CTLA4 IgVH IL-1Ra IL-4/R ICAM-1 MCP-3 IFNa VitD R GABA A3R D2 R

Resistance

HLA-DR1/7 IL-12p40 IFN?CA12 HLA-DR15 DR15-TCRß-DR15 FasL ApoE2 IL-1Ra IL-2 IL-10 GSTM3

Phenotype HLA-DR15 (female, young at onset), ?CCR5 (late onset) HLA-DR4, CTLA4 (primary prog)

Mt DNA (Harding’s disease) Osteopontin, oestrogen receptor (Japanese western) HLA-DP5 (Japanese opticospinal)

Isolates

CNTF

ApoE4 HLA-DR4 (Med)

(Benign)

MBP (Finns) PRK CA (Finns) DPB1*0501 (Japan) HLA-DR8 (Japan)

IL-6 MMP-9

Negative IL-1aß, IL-7R, IL-9, TCRa, CD40L, CD45, B7-1, GelatinaseB, PECAM,TIMP3 PRKAR1A, IFNß, G-proteinB3, HSP70, p53, Bax, Bcl 2, Bcl X, a2 macroglobulin, lipoprotein R related protein, myeloperoxidase, C3/4/6/7, NOS, NOS2A, ? 2-adrenergic R, ApoI, Rhesus blood group, plasminogenactivator inhibitor, CNPase, MOG, MAG, OMGP, PLP, Trk-C, NF2, SCA2/3/6/8/11, CYP2D6CYP2D6, Growth factors, ErbB4, Notch3 FIGURE 30.4

ed15

A list of candidate genes studied in multiple sclerosis. Most are neither linked nor associated (light blue), but provisional evidence is available for an eVect on susceptibility (black), resistance (green), or disease progression or severity (red), population isolates (orange), and clinical phenotype (dark blue). The central panel correlates the clinical course—relapses with recovery; relapses with persistent disability and secondary progression—with the immunological interactions that disrupt normal axon-glial interactions subserving salutatory conduction and indicates potential targets for the eVect of these gene products. Reproduced from Compston, D. A. S., and Coles, A. J. (2002). Multiple sclerosis: Seminar. Lancet 359, 1221–1231.

of multiple sclerosis remain a possibility. The number increases weekly but with rather a poor dividend for increased knowledge. Many of the associations are negative, yet the size of the studies often leaves open the issue of whether these are type 2 errors. Others claim an association with susceptibility but the Wndings have proved hard to reproduce.

LINKAGE GENOME SCREENS

au5

If the dividend from attempting to fast-track the solution to susceptibility in multiple sclerosis has been small, the problem is also not solved by the nine whole genome linkage analyses using variable numbers of families from the United States (Haines et al., 1996), Canada (Ebers et al., 1996), the United Kingdom (Chataway et al., 1998; Sawcer et al., 1996), Finland (Kuokkanen et al., 1997), Sardinia (Coraddu et al., 2001), Italy (Broadley et al., 2001), Turkey (Eraksoy et al., 2003), Scandinavia (Akesson et al., 2002), and Australia (Ban et al., 2002). In the three original large and ethnically diverse populations (United States, Canada, United Kingdom), genotyping was completed on cohorts of between 21 and 225 families, together involving in excess of 1000 individuals, for each of 257 to 443 microsatellite markers. These were chosen to have an average spacing of around 10 centiMorgans giving enough power to identify regions encoding a major susceptibility gene; and the markers were suYciently polymorphic to make a high proportion of the

WHOLE GENOME ASSOCIATION SCREENING

available families fully informative. Although several new genomic regions of interest were revealed, many are likely to be false positives. Failure to identify a major susceptibility factor may indicate that no such gene exists, that it was missed by each group, or that genetic heterogeneity obscured the picture. Other than the major histocompatability complex, several of the other new putative susceptibility loci were clearly unique to each screen. The regions of interest emerging from the United Kingdom genome screen were 1cen, 5 cen, 6p, 7p, 14q, 17q, 19q, Xp (Chataway et al., 1998; Sawcer et al., 1996). They were 2p, 3p, 5p, 11q, and Xp in the Canadian series (Ebers et al., 1996). The United States/ French consortium identiWed 6p, 7q, 11p, 12q, and 19q (Haines et al., 1996). There were no statistically signiWcant regions of interest in the Finnish screen, although positive lod scores were obtained for 6p21 (MHC), 5p14-p12 and 17q22-q24 (Kuokkanen et al., 1997). These studies were followed by smaller screens in families from Italy (Broadley et al., 2001), Sardinia (Coraddu et al., 2001), and Turkey (Eraksoy et al., 2003), and by surveys of comparable size from Scandinavia (Akesson et al., 2002) and Australia (Ban et al., 2002). Although the Italian, Sardinian, Scandinavian, and Australian linkage screens failed to provide statistically unequivocal linkages, more regions of potential interest were identiWed than expected on the basis of chance alone. The main positive result from the Scandinavian, Italian, and Sardinian studies, backed by an updated screen of United Kingdom families, is provisional linkage to chromosome 10p (Akesson et al., 2003). Others have looked at these whole genome screens and explored regions of interest in more detail hoping to consolidate their status based on mapping but without picking out positional candidates. As a result, the linkage peak on chromosome 17q, originally identiWed in the United Kingdom screen, is now supported by additional positional screens from Denmark (Larsen et al., 2000), Canada (Dyment et al., 2001), and Finland (Saarela et al., 2002). Support is also provided for chromsome 5p, 7p, and 12q based on direct evidence and synteny with genes determining susceptibility to experimental forms of demyelination (Oturai et al., 1999; Xu et al., 1999; 2001). Given the lack of overlap between screens and the provisional implication of diVerent regions of interest, meta-analysis has been deployed in the expectation that this will reduce the evidence for false positive peaks and strengthen the candidature of those that are genuine, providing the best guide to shared regions of interest as the map is serially updated. By-eye analysis (Becker et al., 1998) provided a start, but making the separate screens compatible has proved no trivial task. Two meta-analyses are currently available: the original United States, Canadian, and United Kingdom screen (The Transatlantic Multiple Sclerosis Genetics Cooperative, 2001), and all the published families (GAMES and The Transatlantic Multiple Sclerosis Genetics Cooperative, 2003). SuperWcially, these show the same peaks and troughs, but the comparison is confounded by the fact that many of the same families feature in each and the total number is dominated by the same large contributors (Fig. 30.5). The failure of linkage genome screens reliably to identify regions of interest conferring disease susceptibility both in mixed and isolated European populations, each having a high prevalence of multiple sclerosis, has obvious implications for future genetic analyses of complex traits. Linkage data from new populations are being systematically added with the aim of conWrming and excluding regions of interest; studies in progress include new cohorts from the United States and Canada. Meanwhile, continued scrutiny of the human genome map for positional candidates may shorten the search for genes that determine susceptibility and inXuence the course of multiple sclerosis. However, power calculations indicate that tests based on association will generally be more powerful in multiple sclerosis than those dependant on linkage, and this has led to a further switch in strategy (Risch, 1996).

WHOLE GENOME ASSOCIATION SCREENING Until recently, whole genome linkage disequilibrium mapping was considered impractical and dependent on chance colocalization of susceptibility genes and markers applied

719

720

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

Linkage and association in multiple sclerosis: the present position: 5.0

NPL

NPL

NPL

4.0

2.0

NPL

2.0

2.0

3.0

1.0

2.0

1.0

1.0

1.0

1 0.0

16

0.0

0.0 NPL

MS1

NPL

11

6 0.0

NPL

2.0

2.0

NPL

2.0

2.0

1.0

2.0

1.0

12 0.0

2 0.0

17

NPL

NPL

2.0

MS3

2.0

NPL

NPL 2.0

2.0

1.0

1.0

1.0

1.0

13

18

0.0

3

1.0

0.0

7 0.0

0.0

NPL

1.0

1.0

2.0

NPL

1.0

0.0

8 0.0

21

0.0

NPL

NPL

22

0.0

2.0

2.0 NPL

NPL

2.0

2.0

1.0

1.0

2.0

1.0

NPL

1.0

19 14

0.0

1.0

0.0

9

4

0.0

0.0

0.0

NPL

NPL

NPL

2.0

2.0

2.0

2.0

1.0

1.0

1.0

1.0

5 0.0

10 0.0

15 0.0

X

MS2

NPL

20 0.0

FIGURE 30.5 Whole genome linkage screening based on c600 European families is shown for each chromosome. The vertical axis shows the nonparametric likelihood (NPL) score indicating the statistical probability that a gene conferring susceptibility to multiple sclerosis is encoded at that site. Based on these results and whole genome linkage screening, a provisional list of multiple sclerosis susceptibility genes is MS1 (6p), 2 (19q), and 3 (17q).

randomly and distributed at low density. This situation changed with the increased availability of widely distributed microsatellite markers and is set further to increase with the identiWcation and mapping of single nucleotide polymorphisms. Both have required high throughput methods for genotyping helped by economies arising, for example, from the use of pooled DNA (Barcellos et al., 1997). A Wrst pass at screening the genome for association in multiple sclerosis has now been completed based on a 0.5 cM map of microsatellite markers and using DNA pools derived from cases with multiple sclerosis, unrelated controls and trio families (aVected individuals and their parents). The initial sample came from the United Kingdom (Sawcer et al., 2002), but this was the Wrst of 19 carried out more or less simultaneously and together representing the Genetic Analysis of Multiple sclerosis in EuropeanS (GAMES). The study was designed to extend the number of families studied by linkage analysis and to place on this provisional map— assumed to show true and false positives—the location of associated markers, seeking the double hit of linkage (albeit provisional) and association. The use of pooled DNA from cases of multiple sclerosis not subject to stratiWcation on the basis of disease course or phenotype (apart from the exclusion of primary progressive cases) restricted GAMES to the identiWcation of susceptibility genes. The 19 populations were expected to show individual associations not necessarily replicated across the diVerent screens. Although these domestic associations were likely to be of both local and general interest, special signiWcance was attached to those that did survive testing in multiple populations. A highly conservative approach was adopted in deWning associations emerging from testing pools for these 6000 markers. Those showing greatest deviation from the expected distribution of results were ranked. The 19 separate studies were then considered together, searching for the best ranked markers and recognizing that local associations, even if genuine, might be lost by meta-analysis. A deWning and validating subset of populations was used based on typing in each of the two host laboratories. Markers showing the same allelic association in this metaanalysis and those where diVerent alleles of the same polymorphic marker were each listed.

GENETIC INFLUENCES ON THE CLINICAL COURSE

au6

Finally, as a Wrst trawl, individual typing was performed on selected markers. These oVered the most likely prospect of identifying associations even though others lower down the ranking might, in time, also be shown to mark for susceptibility genes. The aim was to Wnd regions outside the major histocompatability complex showing both linkage and association so as both to conWrm the status and narrow the region of interest restricting the identiWcation of candidate genes encoded within that region of linkage disequilibrium. The 10 markers showing the best evidence for association with multiple sclerosis to emerge from analysis of the United Kingdom screen included three from the HLA region on chromosome 6p (D6S1615, D6S2444 and TNFa), providing a positive control for the method, four from regions previously identiWed by linkage analysis in United Kingdom multiplex families (two mapping to chromosome 17q: GCT6E11 and D17S1535; one to chromosome 1p: GGAA30B06; and one to 19q: D19S585), and three from novel sites with respect to linkage analysis (D1S1590 at 1q; D2S2739 at 2p; and D4S416 at 4q). These results provisionally strengthened the candidature of 6p, 17q, 19q, and 1p as regions most likely to encode susceptibility genes for multiple sclerosis. As the remaining 18 components of GAMES were completed and analyzed, many new markers were shown provisionally to be associated. All of these now need to be confirmed using individual typing in data set replications. Figure 30.4 shows the present position with respect to linked and associated markers and provides a best current estimate for the location of multiple sclerosis susceptibility genes. The number of microsatellite markers used necessarily makes GAMES a lowdensity screen and in many chromosomal regions quite possibly insuYcient for the stochastic nature of linkage disequilibrium present in European populations. Even though it may only have covered 10% of the genome in detail and another 20% in part leaving much yet to be explored, the aim was to Wnd some new regions outside the major histocompatibility complex showing linkage and association with multiple sclerosis not each and every such locus—and to exclude those parts of the genome not meriting further search. Once regions of interest are mapped, the next aim is to move from whole genome screening to the identiWcation of functional polymorphisms, which condition one component or another of the disease process and determine variations in the clinical course and features. How to reach that position is less clear and several parallel strategies have been suggested. One is to add incrementally to the number of multiplex families available for linkage until thresholds are reached for the identiWcation of secure loci using statistical criteria for genome wide signiWcance. An alternative is to accept that the combination of linkage and association now available is suYcient to concentrate the search for positional candidates within regions of interest. Each provisional site already oVers several interesting possibilities although the number of genes encoding components of the nervous, immune and signaling systems is such as to make practically any region suggestive with respect to sensible candidates. Finally, there is the option of narrowing the region of interest—still large even if the spacing of GAMES markers has reliably identiWed one of the larger blocks of linkage disequilibrium—by identifying the haplotypes associated with multiple sclerosis. Sequencing one or a few genomes is a remarkable feat, but it tells us little about the rest of humankind. In time, it is expected that the whole genome will be characterized for the size, distribution, and diversity of blocks containing a restricted number of haplotypes. If the preliminary evidence holds up, it will be possible to tag the common variants within each block in populations (such as Europeans) retaining signiWcant linkage disequilibrium and screen individuals for the susceptibility haplotype with relative economy. To date, the hapmap is incomplete but a start is being made on applying such topographical information as does exist to the problem of complex traits in general and multiple sclerosis in particular. Its future use is depicted in Figure 30.6.

GENETIC INFLUENCES ON THE CLINICAL COURSE Clinical analysis of the course and concordance for other features of the disease within pedigrees can be used to gauge the inXuence of genetic factors in determining the clinical

721

722

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

Genetic analysis of multiple sclerosis: 1972-2002 1972

3.0 NPL 2.0 1.0 0.0

3.0 NPL

3.0

2.0

2.0

3.0

NPL

I I II

II IIII III

II IIIIIIIIIIIIII II

I I

I I I I II II

I II

I

IIIIII

II

I

II II III I I

0.0 II I II II I

3.0

NPL

1

I III II

I

I I

I I I I I IIIIII I I

0.0 II II

II I

III

I

5

II IIIIIIIIIIIIII III

I I

I I I I II II

III

I

0.0

IIIIII

3.0

NPL

1.0 III I II II

IIIIIIII

IIII II

III II

I

1.0

I I I

I I I

III I III II III IIIII IIII I

0.0

6

NPL

2.0

2.0

1.0

1.0 III II

3.0

NPL

2.0

2.0

1.0

1.0 III I I I

I

6

1996 0.0

III

17

0.0

II I I I I II IIIIII

19

II

I I

I

IIII II III I I I

I II II I I

X

5.0 NPL

2003

4.0

2.0

2.0 NPL

1.0

2.0

1.0

2.0 1.0

1.0 2

0.0

5

NPL

2.0 1.0

1.0

6

6

10

Candidate genes

0.0

0.0

11

16

NPL

2.0 NPL

1.0 16

0.0

0.0

2.0

NPL

1.0

10

0.0

1

2.0

NPL

5

0.0

2004

NPL

3.0

NPL 2.0

1.0

0.0

0.0

17

NPL

22

0.0

17

2.0

1.0

19

22

X

Haplotype mapping

150Kb

50Kb

FIGURE 30.6 Association (but not linkage) between multiple sclerosis and alleles of the major histocompatibility complex was shown in 1972 and subsequently conWrmed on many occasions and in many populations. Provisional linkage to novel chromosomal regions of interest was provided in 1996 by whole genome screens. Some of these regions are conWrmed and others newly identiWed by whole genome (low-density) association mapping and updated linkage analysis. Future strategies either involve testing positional candidates suggested by reference to the human genome map or narrowing the region of interest by identifying the associated haplotype using tagged single nucleotide (SNP) markers.

au7

au8

au9

phenotype of multiple sclerosis in the individual. In a large population-based sample, Ebers et al. (2000a) showed no diVerences in time to reach the later stages of disability among 206 familial cases, irrespective of relatedness to the proband, compared to sporadic multiple sclerosis although onset was earlier and the usual female predominance no longer seen in the families with most cases—providing no evidence for genetic eVects on the clinical features or course. In the United Kingdom survey of conjugal pairs, there was no evidence for clinical concordance, clustering at year of onset or distortion of the expected pattern of age at onset in the second aVected spouse from 33 pairs in whom these comparisons could be made (Robertson et al., 1997). Statistical analysis of sibling pairs is prone to sampling bias; concordance for year at onset is much inXuenced by the tendency for earlier recognition of symptoms in the second individual aVected in a pair through heightened awareness of the possibility of multiple sclerosis. After appropriate correction, Robertson et al. (1996a) found no correlation with age or year at onset, or with mode of presentation and disability, although there was a correlation with disease course, and aVected siblings tended to be same-sex. The most recent assessment of concordance in coaVected siblings and parent-child pairs supports a role for genetic factors in determining eventual disability and handicap in multiple sclerosis but not the initial presentation (Chataway et al., 2001). Concordant parent-child pairs show no distortion in the random distribution of male-female pairings, and neither sex nor line of inheritance inXuence

GENETIC INFLUENCES ON THE CLINICAL COURSE

au10

au11

au12

disability, age at onset or course; however, disability is highest in the male oVspring of aVected fathers who more commonly follow a primary progressive course (Hupperts et al., 2001). It could be argued that comparing year on age at onset and clinical features described over many years is too imprecise a method for assessing the relative contribution of genes and environmental factors in determining the development and course of multiple sclerosis in individuals; but with large numbers of siblings, artefactual correlations are reduced and useful statistical trends emerge. Turning to speciWc associations, the majority of surveys investigating HLA class 2 alleles and disease course fail to conWrm speciWcally diVerent associations with any one of these phenotypes; the presence of DR15 is associated with younger age at diagnosis and female gender but does not distinguish features relating to disease course, outcome, speciWc clinical features, or paraclinical investigations (Celius et al., 2000; Hensiek et al., 2002 Masterman et al., 2000). This suggests that DR15 exerts an eVect on susceptibility rather than modifying the course of sporadic multiple sclerosis. Most recently, these observations have been extended to familial multiple sclerosis, with additional conWrmation of heterogeneity since DR linkage (and concordance for early onset) are seen only in families where DR2(15) is present (The Multiple Sclerosis Genetics Group, 2002). Some genetic studies have been conWned to eVects on the clinical course or prognostic factors using stratiWed populations (for example, Evangelou et al., 1999; Giess et al., 2002; Hogh et al., 2000; for review, see Kantarci et al., 2002); yet the stratiWcations are necessarily arbitrary and of doubtful provenance. Appropriately, the majority of these studies are conWned to discrete populations and positive Wndings cannot necessarily be generalized to other groups. The example of CTLA4 serves well to illustrate the diYculty of achieving a secure position: Ligers et al. (1999) reported population- and family-based associations with homozygosity of the G49 allele of the CTLA4 gene in relapsing-remitting multiple sclerosis; this was conWrmed by Harbo et al. (2000); homozygosity of the AA allele of exon 1 was then primarily associated with increased severity in Japanese patients with relapsing-remitting or secondary progressive multiple sclerosis (Fukazawa et al., 1999); but Dyment et al. (2002) were unable to demonstrate linkage or association to susceptibility or the clinical course with a microsatellite marking exon 1 of the CTLA-4 gene in Canadian patients; Maurer et al. (2002) and Masterman et al. (2002), working from the groups that originally claimed an association to susceptibility in relapsing-remitting multiple sclerosis, then claimed a relationship with primary progressive disease; Alizadeh et al. (2003) identiWed a single nucleotide polymorphism strongly associated with susceptibility to relapsing-remitting multiple sclerosis in French and Iberian patients; Wnally, Roxburgh et al. (2003) failed to show any association in family-based studies with susceptibility or disease course. The risk of autoimmunity is increased in the relatives of probands with multiple sclerosis. Heinzlef et al. (1999) studied 1971 relatives of 357 probands and identiWed recurrence of multiple sclerosis in 15% with another autoimmune disease in a further 8% and several disorders (Graves’s disease, rheumatoid arthritis, and diabetes) occurring in 5% of pedigrees. Broadley et al (2000) assessed the frequency of multiple sclerosis, Wve deWnite and three putative autoimmuine disorders, and control conditions in the 1315 Wrst degree relatives of 571 probands, compared with 375 control families. Whereas patients themselves did not have an excess of autoimmune diseases, these were over-represented in family members of individuals with multiple sclerosis. The main eVect was attributed to an excess of autoimmune thyroid disease. Recurrence risk was especially high in multiplex families. Henderson et al. (2000) identiWed an odds ratio of 2.2 for recurrence of autoimmune disease in 722 Wrst degree relatives of 117 Australian patients with multiple sclerosis. Several other disorders have been considered more frequent than expected in patients with multiple sclerosis. None of these is entirely secure, but there may be comorbidity between neuroWbromatosis 1 and primary progressive multiple sclerosis; if so, this is not explained by mutation of the oligodendrocyte myelin glycoprotein gene, which maps to the same locus (Johnson et al., 2000).

723

724

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

HETEROGENEITY IN MULTIPLE SCLEROSIS The existing genetic analyses are predicated on the assumption that multiple sclerosis is one disease. A major part of future genetic studies will be to resolve the question of disease heterogeneity. Even now, it is possible to dissect the disease phenotype and suggest features that may have arisen through a distinct etiology and mechanism. Furthermore, preliminary dissections of the provisional genetic descriptions suggest discrete clusters of regions forming epistatic groups. A minority of patients who meet the clinical criteria for deWnite multiple sclerosis and in whom there are associated magnetic resonance imaging abnormalities and cerebrospinal Xuid oligoclonal bands have an illness in which there is disproportionate involvement of the anterior visual pathway. These are commonly women with male relatives already known to be aVected by Leber’s hereditary optic neuropathy and they have pathological mutations of mitochondrial (mt) DNA (Harding et al., 1992; Kellar Wood et al., 1994). Visual involvement in multiple sclerosis may generally be associated with the haplotype J of mtDNA (Reynier et al., 1999); but severe optic neuropathy occurring in the context of the Devic phenotype is not associated with mutations of mtDNA (Kalman and Mandler, 2002). Harding’s disease therefore represents true heterogeneity in that a speciWc genotype determines a characteristic phenotype. What remains unresolved is whether the mutation of mtDNA directs the process of brain inXammation onto a particular site—constituting selective tissue vulnerability—or merely represents the chance occurrence of relatively mild multiple sclerosis and Leber’s hereditary optic neuropathy. The latter seems unlikely given the number of cases of Harding’s disease and the rarity of isolated Leber’s hereditary optic neuropathy in women. The distinct clinical features of demyelinating disease seen in Orientals and Africans provide another probable example of heterogeneity. The suggestion is that, in Japan, multiple sclerosis shows either a Western phenotype in which a number of sites are involved, or an opticospinal pattern in which the clinical picture is dominated by involvement of visual and spinal cord pathways having more neurodegenerative and less inXammatory activity and with a speciWcally diVerent genetic background (HLA-DP*1501 rather than DRB1*1501 seen with the Western phenotype: Yamasaki et al., 1999). In its extreme form, this mimics Devic’s disease—a disorder that also occurs in Europeans, albeit rarely. However, as with many aspects of multiple sclerosis, the early descriptions have not always proved stable, and recent reports of multiple sclerosis in Japanese highlight the previously underreported extent of the so-called Western phenotype. ConWning their analysis to patients with this Western phenotype, Kikuchi et al. (2003) segregated cases into two clinically indistinguishable groups deWned by DR15, oligoclonal bands, and high magnetic resonance lesion load compared with patients positive for DR4 but without oligoclonal bands and having fewer magnetic resonance lesions. Demyelinating disease is considered to be extremely rare in Africans, but a number of cases are described and the phenotype is typically a severe illness dominated by one or more episodes usually aVecting the anterior visual pathway and spinal cord—again combining the anatomical features of Devic’s disease with the clinical course of moderately severe relapsing remitting multiple sclerosis (Dean et al., 1994). Unlike Harding’s disease, the evidence that this represents disease heterogeneity arising from speciWc genetic or environmental modiWcation of a core pathological process remains circumstantial. The concept of heterogeneity is further developed in the recent pathological studies using biopsy and autopsy material in which four distinct types of disease process are proposed (Luchinetti et al., 2000). Allelic heterogeneity provides evidence for the functional relevance of a given locus since the probability of several diVerent associated polymorphisms occurring at the same site in unrelated individuals would be low if this locus was altogether irrelevant. Allelic heterogeneity is suggested by detailed analysis of the most secure Wnding in multiple sclerosis—association with alleles of the major histocompatibility complex. In addition to the increase in DR15/DQ6 phenotype in Northern Europeans, the subsidiary association with DR3 (DR17)-DQ2 (and its associated DRB1*0301-DRB5*0101-DQA1*0501-DQB1*0201 genotype) and speciWcally diVerent as-

CONCLUSIONS

au13

sociations in Mediterranean populations cannot be explained by site speciWc similarities in sequence of the crucial peptide binding elements. The implication must be that either these alleles are in linkage disequilibrium with another (shared) susceptibility gene encoded within the major histocompatibility complex (which may or may not have a primary immunological function), or that the environmental trigger which initiates the disease process in multiple sclerosis varies and so selects speciWcally diVerent at-risk populations. The evidence for genetic heterogeneity is also addressed by the analysis of primary progressive multiple sclerosis. The claim has been made that DR4 confers susceptibility and DQ7 resistance to the primary progressive form of multiple sclerosis in Scandinavian patients (Hillert et al., 1992). Although there is support from a more recent study of primary progressive multiple sclerosis in Spain (de la Concha et al., 1997), consensus has not been reached (McDonnell et al., 1999; Weinshenker et al., 1998) and deWnitive studies are needed. Furthermore, the opticospinal but not the Western type of multiple sclerosis is associated with DPB1*0501 in northern and southern Japan, whereas the western type is DRB1*1501 associated (Ma et al., 1998; Yamasaki et al., 1999). Conditioning the United Kingdom genome screen for DR15 (or an extended DR15 linked haplotype also encoding alleles of TNF and the DQ locus) showed that the regions of interest on 1p, 17p, 17q, and X clustered in families that are identical by state for DR15, whereas the nonsharing group was associated with 1cen, 3p, 5cen, 7p, 14q, and 22q; in addition new regions of interest were found at 5q and 13p (DR15 sharing) and 16p and 20p (DR15 nonsharing: Chataway et al., 1998; Coraddu et al., 1999). As knowledge accumulates, conditioned analyses may routinely be needed in order to suggest or exclude new regions of interest or positional candidate susceptibility genes. The presence of phenocopies is a major concern in the analysis of complex traits where diagnosis depends on pattern recognition of symptoms, signs, and laboratory investigations occurring in the absence of a test for the disease. Reassuringly, the cohort of cases included in the United Kingdom linkage genome screen was not contaminated by cases of CADASIL, spinocerebellar degenerations, or adrenoleukodystrophy in male-male pairs (Chataway et al., 1998; Sawcer et al., 1996). However, this search did identify a potentially important aspect of genetic heterogeneity that may have implications for selective tissue vulnerability. Although there were no individuals having an excess of triplet repeats for SCA2, the 22kb allele occurred at a higher frequency in cases than controls reported in the literature. This result prompted an assessment of transmission disequilibrium in family trios, which supported an association between multiple sclerosis and the 22kb allele (Chataway et al., 1999). If conWrmed, one interpretation of this Wnding would be that, in individuals who have a tendency for autoimmunity as a result of the interplay between genetic susceptibility and environmental factors, the inXammatory process may be targeted onto a particular system or pathway within the brain and spinal cord in those who have genetic polymorphisms exposing that pathway to tissue injury. In the case of SCA2, individuals with the 22kb allele polymorphism may have disproportionate inXammatory demyelination of the spinocerebellar pathways—similar by analogy to involvement of the anterior visual pathway in Harding’s disease.

CONCLUSIONS To risk a summary of data that are distinctly ambiguous, genes showing some reproducibility for an eVect on susceptibility are HLA DR/DQ, TNF-A, IL-1Ra, IL-4, and CTLA 4. Those apparently associated with disease protection are FAS-670 and IL-12p40. Genes that may inXuence the course or phenotype of multiple sclerosis include IL-10, CNTF, and ApoE4. Mutations of mitochondrial DNA may determine the phenotype. Eventually, six main categories of susceptibility gene can be predicted: genes that determine susceptibility to the process of inXammation across a range of disorders—the autoimmune genes; those that determine the speciWcity of that process for the development of multiple sclerosis—the ubiquitous genes; those that are relevant for the pathogenesis in isolated populations—the

725

726

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

au14

au16

domestic genes; those that determine particular phenotypes—the pleotropic genes; those that determine variations in the clinical course—the modifying genes; and those that cluster to provide speciWcally diVerent (heterogenous) contributions to the pathogenesis—the epistatic genes. Although perhaps overly speculative, some examples can be oVered. Taken with the increased recurrence risk of autoimmunity in relatives of probands with multiple sclerosis (Broadley et al., 2000; Heinzlef et al., 1999, Henderson et al., 2000), 10p (diabetes and multiple sclerosis: Broadley et al., 2001; Coraddu et al., 2001; Dyment et al., 2001; Reed et al., 1997),18q (diabetes, rheumatoid arthritis and multiple sclerosis: Merriman et al., 2001), and 6p21 (major histocompatibility locus) may each encode genes that aVect the general process of autoimmunity—autoimmune genes. 19q and 17q are provisionally linked and associated in several populations (Dyment et al., 2001; Ebers et al., 1996; Kuokkenen et al., 1996; Larsen et al., 2000; Saarla et al., 2002; Sawcer et al., 1996; Sawcer et al., 2002)—ubiquitous multiple sclerosis genes. Myelin basic protein is linked and associated in Finns and Turks (who share their population history) with multiple sclerosis (Eraksoy et al., 2003; Tiennari et al., 1992)—a domestic gene. Mutations of mitochondrial DNA are associated with a deWned phenotype, Harding’s disease, in multiple sclerosis and the SCA2 association may implicate spinocerebellar pathways (Chataway et al., 1999; Harding et al., 1992; Kellar Wood et al., 1994)—pleotropic genes. DR15 may be associated with earlier onset and female gender but not more severe clinical course in multiple sclerosis (Celius et al., 2000; Hensiek et al., 2002; Masterman et al., 2000)—a modifying gene. Further, some regions of interest cluster within families stratiWed for the presence or not of DR15—epistatic groups. Clearly, much work lies ahead to consolidate these provisional classiWcations of gene eVects in multiple sclerosis. A major part of future studies will be to resolve the question of disease heterogeneity in multiple sclerosis. When eventually in place, the potential of this genetic knowledge for improved understanding of the pathogenesis of multiple sclerosis and designing novel treatments is considerable. Resolving the issues of complexity and heterogeneity in multiple sclerosis, and other complex traits, has practical dividends. Without knowledge linking etiology to pathogenesis and phenotype, putative new treatments—selected on the basis of an imperfect understanding of the pathogenesis—will continue to be screened in cohorts who may or may not have an appropriate pathological substrate for that particular intervention.

References Ablashi, D. V., Lapps, W., Kaplan, M., et al. (1998). Human herpesvirus-6 (HHV-6) infection in multiple sclerosis: A preliminary report. Multiple Sclerosis 4, 490–496. ˚ kesson, E., Coraddu. F., Marrosu, M., et al. (2003). Fine mapping of a candidate region on chromosome 10 in A 449 sib-pairs with multiple sclerosis. J. Neuroimmunology (in press). ˚ kesson, E., Oturai, A., Berg, J., et al. (2002). A genome-wide screen for linkage in Nordic sib-pairs with multiple A sclerosis. Genes and Immunity 3, 279–285, Andersen, O., Lygner, P.-E., Berstrom, T., Andersson, M., and Vahlne, A. (1991). Viral infections trigger multiple sclerosis relapses: A prospective seroepidemiological study. J. Neurology 240, 417–422. Alizadeh M., Babron M. C., Birebent, B., et al. (2003). Involvement of CTLA-4 gene promoter SNP in MS genetics: A European collaborative family study Ann. Neurol. (in press). Alter, M., Halpern, L., Kurland, L. T., Bornstein, V., Tikva, P., Leibowitz, U., and Silberstein, J. (1962). Multiple sclerosis in Israel: Prevalence among immigrants and native inhabitants. Arch. Neurology 7, 253–263. Alter, M., Kahana, E., and Loewenson, R. (1978). Migration and risk of multiple sclerosis. Neurology 28, 1089–1093. Alter M., Okihiro, M., Rowley, W., and Morris, T. (1971). Multiple sclerosis among Orientals and Caucasians in Hawaii. Neurology 2, 122–130. Ascherio, A., Zhang, S. M., Hernan, M. A., et al. (2001). Hepatitis B vaccination and the risk of multiple sclerosis. New England J. Medicine, 344, 327–332. Ban, M., Stewart, G., Bennetts, B., Heard, R., et al. (2002) A genome screen for linkage in Australian sibling-pairs with multiple sclerosis. Genes and Immunity 3, 464–469. Barcellos, L. F., Klitz ,W., Field, L. L., et al. (1997). Association mapping of disease loci using a pooled DNA genomic screen. Am. J. Human Genetics 61, 734–747. Baskett, P. J. F., and Armstrong, R. (1970). Anaesthetic problems in multiple sclerosis. Anaesthesia 25, 397–401. Bauer, H. J. (1987). Multiple sclerosis in Europe. Symposium Report. J. Neurology 234, 195–206.

CONCLUSIONS

Becker, K. G., Simon R. M., and Bailey-Wilson, J. E. (1998). Clustering of non-major histocompatibility complex susceptibility candidate loci in human autoimmune disease. Proceedings of the National Academy of Sciences of the USA 95, 9979–9984. Benedikz, J. G., Magnusson, H., and Gudmundsson, G. (1994). Multiple sclerosis in Iceland, with observations on the alleged epidemic in the Faroe Islands. Ann. Neurology 36 (suppl 2), S175–S179. Broadley, S., Deans, J., Sawcer, S. J., et al. (2000). Autoimmune disease in Wrst degree relatives of patients with multiple sclerosis in the United Kingdom. Brain 123, 1102–1111. Broadley, S., Sawcer, S., D’Alfonso, S., et al. (2001). A genome screen for multiple sclerosis in Italian families. Genes and Immunity 2, 205–210. Bulman, D., and Ebers, G. C. (1992). The geography of multiple sclerosis reXects genetic susceptibility. J. Tropical and Geographical Neurology 2, 66–72. Buljevac, D., Flach, H. Z., Hop, W. C. J., et al. (2002). Prospective study on the relationship between infections and multiple sclerosis exacerbations. Brain 125, 952–960. Carton, H., Vlietinck, R., Debruyne, J., et al. (1996). Recurrence risks of multiple sclerosis in relatives of patients in Flanders, Belgium. J. Neurology Neurosurgery and Psychiatry 62, 329–333. Cavalli-Sforza, L. L., Menozzi, P., and Piazza, A. (1994). ‘‘The History and Geography of Human Genes.’’ Princeton University Press, Princeton, NJ. Celius, E. G., Harbo, H. F., Egeland, T. et al. (2000). Sex and age at diagnosis are correlated with the HLA-DR2, DQ6 haplotype in multiple sclerosis. J. Neurological Sciences 178, 132–135. Chakraborty, R., Kamboh, M. I., Nwankwo, M., and Ferrell, R. E. (1992). Caucasian genes in American Blacks: New data. Am. J. Human Genetics 50, 145–155. Chataway, J., Feakes, R., Coraddu, F. et al. (1998). The genetics of multiple sclerosis: Principles, background and updated results of the United Kingdom systematic genome screen. Brain 121, 1869–1887. Chataway, J., Sawcer, S., Coraddu, F. et al. (1999). Allelic variants of the spinocerebellar ataxia genes contribute to multiple sclerosis susceptibility. Neurogenetics 2, 91–96. Chataway, S. J. S., Mander, A., Robertson, N., et al. (2001). Multiple sclerosis in sibling pairs: An analysis of 250 families. J. Neurology Neurosurgery and Psychiatry 71, 757–761. Collins, A., Ennis, S., Taillon-Miller, P. et al. (2001). Allelic association with SNPs: Metrics, populations, and the linkage disequilibrium map. Human Mutation 17, 255–262. Compston, D. A. S., Vakarelis, B.N., P. E., McDonald, W. I., Batchelor, J. R., and Mims, C. A. (1986). Viral infection in patients with multiple sclerosis and HLA-DR matched controls. Brain 109, 325–344. Confavreux, C., Hutchinson, M., Hours, M., et al. (1998). Rate of pregnancy-related relapse in multiple sclerosis. The New England J. Medicine 339, 285–291. Confavreux, C., Suissa, S., Saddier, P., et al. (2001). Vaccinations and the risk of relapse in multiple sclerosis. The New England J. Medicine 344, 319–326. Cook, S. D., Cromarty, M. B., Tapp, W., Poskanzer, D., Walker, J. D., and Dowling, P. C. (1985). Declining incidence of multiple sclerosis in the Orkney Islands. Neurology 35, 545–551. Coraddu, F., Reyes-Yanez, M. P., Aladro, Y., et al. (1998). HLA associations with multiple sclerosis in the Canary Islands. J. Neuroimmunology 87, 130–135. Coraddu, F., Sawcer, S., D’Alfonso, S., et al. (2001). A genome screen for multiple sclerosis in Sardinian multiplex families. European J. Human Genetics 9, 621–626. Coraddu, F., Sawcer, S., Feakes, R., et al. (1999). HLA typing in the United Kingdom multiple sclerosis genome screen. Neurogenetics 2, 24–33. Daly, M. J., Rioux, J. D., SchaVner, S. F., et al. (2001). High-resolution haplotype structure in the human genome. Nature Genetics 29, 229–332. Davenport, C.B. (1922). Multiple sclerosis from the standpoint of geographic distribution and race. Arch. Neurology 8, 51–58. Dawson, E., Abecasis, G. R., Bumpstead, S. et al. (2002) A Wrst-generation linkage disequilibrium map of human chromosome 22. Nature 418, 544–548. de la Concha, E. G., Arroyo, R., Crusius, J. B., et al. (1997). Combined eVect of HLA-DRB1*1501 and interleukin-1 receptor antagonist gene allele 2 in susceptibility to relapsing/remitting multiple sclerosis. J. Neuroimmunology 80, 172–8. Dean, G. (1967). Annual incidence, prevalence and mortality of MS in white South African-born and in white immigrants to South Africa. British Med. J.l 2, 724–730. Dean, G., Bhighee, A. I. G., Bill, P. L. A., et al. (1994). Multiple sclerosis is black South Africans and Zimbabweans. J. Neurology Neurosurgery and Psychiatry 57, 1064–1069. Dean, G., Elian, M., de Bono, A. G., et al. (2002). Multiple sclerosis in Malta in 1998: An update. J. Neurology Neurosurgery and Psychiatry 73, 256–260. Detels, R., Brody, J. F., Edgar, A. H. (1972). Multiple sclerosis among American, Japanese and Chinese migrants to California and Washington. J. Chronic Diseases 25, 3–10. Detels, R., Visscher, B., Malmgrem, R. M., Coulson, A. H., Lucia, M. V., and Dudley, J. P. (1977). Evidence for lower susceptibility to multiple sclerosis in Japanese-Americans. Am. J. Epidemiology 105, 303–310. Duquette, P., Murray T. J., Pleines, J., Ebers, G. C., Sadovnick, D., Weldon, P., Warren, S., Paty, D. W., Upton, A., Hader, W., Nelson, R., Auty, A., Neufeld, B., and Meltzer, C. (1987). Multiple sclerosis in childhood: Clinical proWle in 125 patients. J. Pediatrics 3, 359–363. Dunning, A. M., Durocher, F., Healey, C. S., et al. (2000). The extent of linkage disequilibrium in four populations with distinct demographic histories. Am. J. Human Genetics 67, 1544–1554.

727

728

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

au17

Dyment, D. A., Steckley, J. L., Willer, C. J., et al. (2002). No evidence to support CTLA-4 as a susceptibility gene in MS families: The Canadian Collaborative Study. J. Neuroimmunology 123, 193–198. Dyment, D. A., Willer, C. J., Scott, B., et al. (2001). Genetic susceptibility to MS: A second stage analysis in Canadian MS families. Neurogenetics 3, 145–51. Eaves, I. A., Merriman T. R., Barber, R. A., et al. (2000). The genetically isolated populations of Finland and Sardinia may not be a panacea for linkage disequilibrium mapping of common disease genes. Nature Genetics 25, 320–323. Ebers, G. C., Koopman, W. J., Hader, W., et al. (2000a). The natural history of multiple sclerosis: A geographically based study. Brain 123, 641–649. Ebers, G.C., Kukay, K., Bulma,n D.,et al. (1996). A full genome search in multiple sclerosis. Nature Genetics 13, 472–476. Ebers, G. C., Sadovnick, A. D., and Risch NJ. (1995). A genetic basis for familial aggregation in multiple sclerosis. Nature 377, 150–151. Ebers, G. C., Yee, I. M., Sadovnick, A. D., and Duquette, P. (2000b). Conjugal multiple sclerosis: Population-based prevalence and recurrence risks in oVspring. Canadian Collaborative Study Group. Ann. Neurology 48, 927–931. Eisenberg, D., Marcotte, E. M., Xenaris, I., and Yeats, T. O. (2000). Protein function in the post-genome era. Nature 405, 823–826. Elian, M., and Dean, G. (1987). Multiple sclerosis among United Kingdom born children of immigrants from the West Indies. J. Neurology, Neurosurgery and Psychiatry 50, 327–332. Elian, M., Nightingale, S., and Dean, G. (1990). Multiple sclerosis among United Kingdom-born children of immigrants from the Indian subcontinent, Africa and the West Indies. J. Neurology Neurosurgery and Psychiatry 53, 906–911. Eraksoy, M., Hensiek, A., Ku¨rtu¨ncu¨, M., et al. (2003). A genome screen for linkage disequilibrium in Turkish multiple sclerosis. J. Neuroimmunology (in press). Evangelou,, N., Jackson, M., Beeson, D., and Palace, J. (1999). Association of the APOE e4 allele with diseas activity in multiple sclerosis. J. Neurology, Neurosurgery and Psychiatry 67, 203–205. Friedman, J. E., Lyons, M. J., Cu, G., et al. (1999). The association of the human herpesvirus-6 and MS. Multiple Sclerosis 5, 355–362. French Research Group on Multiple Sclerosis. (1992). Multiple Sclerosis in 54 twinships: Concordance rate is independent of zygosity. Ann. Neurology 32, 724–727. Fukazawa, T., Kikuchi, S., Sasaki, H., et al. (2000) Genomic HLA proWles of MS in Hokkaido, Japan: Important role of DPB1*0501 allele. J. Neurology 247, 175–178. Fukazawa, T., Yanagawa, T., Kikuchi, S., et al. (1999). CTLA-4 gene polymorphism may modulate disease in Japanese multiple sclerosis patients. J. Neurological Sciences 171, 49–55. GAMES and The Transatlantic Multiple Sclerosis Genetics Cooperative. (2001). A meta-analysis of genome screens in multiple sclerosis. J. Multiple Sclerosis 7, 3–11. Giess, R., Maurer, M., and Pohl, D. (2002). A null mutation in the CNTF gene is associated with early onset of multiple sclerosis. Archives of Neurology, 59 407–409. Haahr, S., Koch-Henriksen, N., Moller-Larsen, A., et al. (1995). Increased risk of multiple sclerosis after late Epstein-Barr virus infection: A historical prospective study. Multiple Sclerosis 1, 73–77. Haines, J. L., Ter-Minassian, M., Bazyk, A., et al. (1996). A complete genomic screen for multiple sclerosis underscores a role for the major histocompatibility complex. Nature Genetics 13, 469–471. Hammond, S. R., English, D. R, and McLeod, J. G. (2000). The age-range of risk of developing multiple sclerosis. Evidence from a migrant population in Australia. Brain 123, 968–974. Hammond, S. R., McLeod, J. G., Millingen, K. S., Stewart-Wynne, E. G., English, D., Holland, J. T., and McCall, M. G. (1988). The epidemiology of multiple sclerosis in 3 Australian cities: Perth, Newcastle and Hobart. Brain 111, 1–25. Harbo, H. F., Celius, E. G., Vardtal, F., and Spurkland, A. (2000). CTLA4 promoter and exon 1 dimorphisms in multiple sclerosis. Tissue Antigens 53, 106–110. Harding, A. E., Sweeney, M. G., Brockington, M., et al. (1992). Occurrence of a multiple sclerosis-like illness in women who have a Leber’s hereditary optic neuropathy mitochondrial DNA mutation. Brain 115, 989–989. Hawkes, C. H. (2002). Is multiple sclerosis a sexually transmitted infection? J. Neurology Neurosurgery and Psychiatry 73, 439–443. He, B., Xu, C., Yang, B., et al. (1998). Linkage and association analysis of genes encoding cytokines and myelin proteins in multiple sclerosis. J. Neuroimmunology 86, 13–19. Heinzlef, O., Alamowitch, S., Sazdovitch, V., et al. (1999). Autoimmune disease in families of French patients with multiple sclerosis. Acta Neurologica Scandinavica 100, 1–5. Henderson, R. D., Bain, C. J., and Pender, M. P. (2000). The occurrence of autoimmune diseases in patients with multiple sclerosis and their families. J. Clinical Neuroscience 7(5), 434–437. Hensiek, A. E., Sawcer, S. J., Feakes, R., et al. (2002) HLA-DR 15 is associated with female gender and younger age at diagnosis in multiple sclerosis. J. Neurology Neurosurgery and Psychiatry 72, 184–187. Hernan, M. A., Zhang, S. M., Lipworth, L., et al. (2001). Multiple sclerosis and age at infection with common viruses. Epidemiology 12, 301–06. Hillert, J., Gronning, M., Hyland, H., Link, H., and Olerup, O. (1992). Immunogenetic heterogeneity in multiple sclerosis. J. Neurology Neurosurgery and Psychiatry 55, 887–890. Hogh, P., Oturai, A., Schreiber, K., et al. (2000). Apoliprotein E and multiple sclerosis: Impact of the epsilon-4 allele on susceptibility, clinical type and progression rate. Mutiple Sclerosis 6, 226–230.

CONCLUSIONS

Hupperts, R., Broadley, S., Mander, A., et al. (2001). Patterns of disease in concordant parent-child pairs with multiple sclerosis. Neurology 57, 290–295. Jacobsen, M., Schweer, D., Ziegler, A., et al. (2000). A point mutation in PTPRC is associated with the development of multiple sclerosis. Nature Genetics 26, 495–499. JeVreys, A. J,. Kauppi, L., Neumann, R. (2001). Intensely punctate meiotic recombination in the class II region of the major histocompatibility complex. Nature Genetics 29, 217–222. Jin, Y.-P., de Pedro-Cuesta, J., Soderstrom, M., et al. (2000). Seasonal patterns in optic neuritis and multiple sclerosis: A meta-analysis. J. Neurol. Sci. 181, 56–64. Johnson, G. C., Esposito, L., Barratt B. J., et al. (2001). Haplotype tagging for the identiWcation of common disease genes. Nature Genetics 29, 233–237. Johnson, M. R., Ferner, R. E., Bobrow, M., et al. (2000). Detailed analysis of the oligodendrocyte myelin glycoprotein gene in four patients with neuroWbromatosis 1 and primary progressive multiple sclerosis. J. Neurology, Neurosurgery and Psychiatry 68(5), 643–6. Kahana, E., Zilber, N., Abramson, J. H., Biton, Y., Leibowitz, Y., and Abramsky, O. (1994). Multiple sclerosis: Genetic versus environmental aetioogy: Epidemiology in Israel updated. J. Neurology 241, 341–346. Kalman, B., and Mandler, R. N. (2002). Studies of mitochondrial DNA in Devic’s disease revealed no pathogenic mutations, but polymorphisms also found in association with multiple sclerosis. Ann. Neurology 51 5, 661. Kantarci, O. H., de Andrade, M., Weinshenker, B. G. (2002) Identifying disease modifying genes in multiple sclerosis. J. Neuroimmunology 123, 144–159. Kaufman, M., Gaydos, C. A., Sriram, S., et al. (2002). Is Chlamydia pneumoniae found in spinal Xuid samples from multiple sclerosis patients? ConXicting results. Multiple Sclerosis 8, 289–294. Kellar Wood, H., Robertson, N., Govan, G. G., et al. (1994). Leber’s hereditary optic neuropathy mitochondrial DNA mutations in multiple sclerosis. Ann. Neurology 36, 109–112. Kendler, K. S., MacLean, C. J,. Ma, Y., et al. (1999). Marker-to-marker linkage disequilibrium on chromosomes 5q, 6p and 8p in Irish high-density achizophreniua pedigrees. Am. J. Human Genetics 88, 29–33. Kikuchi, S., Fukazawa, T., Niino, M., et al. (2003). HLA-related subpopulations of MS in Japanese with and without oligoclonal bands. Neurology (in press). Kruglyak, L. (1999). Prospects for whole-genome linkage disequilibrium mapping of common disease genes. Nature Genetics 22, 139–144. Kuhlmann, T., Glas, M., zum Bruch, C., et al. (2002). Investigation of bax, bcl-2, bcl-x and p53 gene polymorphisms in multiple sclerosis. J. Neuroimmunology 129, 154–160. Kuokkanen, S, Gschwend, M., Rioux, J. D., et al. (1997). Genomewide scan of multiple sclerosis in Finnish multiplex families. Am. J. Human Genetics 61, 1379–1387. Kuroiwa, Y., Shibasaki, H,. and Ikeda, M. (1983). Prevalence of multiple sclerosis and its north-south gradient in Japan. Neuroepidemiology 2, 62–69. Kurtzke, J. F. (1975). A reassessment of the distribution of multiple sclerosis. Acta Neurologica Scandinavica 51, 110–136; 137–157. Kurtzke, J. F. (1993). Epidemiologic evidence for multiple sclerosis as an infection. Clinical Microbiology Reviews 6, 382–427. Kurtzke, J. F., and Bui, Q. H. (1980). Multiple sclerosis in a migrant population to half Orientals immigrating in childhood. Ann. Neurology 8, 256–260. Kurtzke, J. F., Gudmundsson, K. R., and Bergmann, S. (1982). Multiple sclerosis in Iceland. 1. Evidence of a post-war epidemic. Neurology 32, 143–150. Kurtzke, J. F., and Heltberg, A. (2001). Multiple sclerosis in the Faroe Islands: An epitome. J. Clinical Epidemiology 54, 1–22. Kurtzke, J. F., Kurland, L. T., and Goldberg, I. D. (1971). Mortality and migration in multiple sclerosis. Neurology 21, 1186–1197. Larsen, F., Oturai, A., Ryder, L. P., et al. (2000). A. linkage analysis of a candidate region in Scandinavian sib pairs with multiple sclerosis reveals linkage to chromosome 17q. Genes and Immunity 1, 456–459. Lauer, K. (1994). The risk of multiple sclerosis in the USA in relation to sociogeographic features: a factoranalytic study. J. Clinical Epidemiology 47, 43–48. Lauer, K., and Firnhaber, W. (1994). Multiple sclerosis in Europe: An epidemiological update. LeuchturmVerlag/LTV Press, Darmstadt. Ligers, A., Xu, C., Saarinen, S., et al. (1999). The CTLA-4 gene is associated with multiple sclerosis. J. Neuroimmunology 97, 182–190. Lindberg, C., Andersen, O., Vahlne, A., et al. (1991). Epidemiological investigation of the association between infectious mononucleosis and multiple sclerosis. Neuroepidemiology 10, 62–65. Lucchinetti, C., Bruck, W., Parisi, J., et al. (2000). Heterogeneity for multiple sclerosis lesions: Implications for the pathogenesis of demyelination. Annals of Neurology 47, 707–717. Ma, J. J., Nishimura, M., Mine, H., et al. (1998). HLA-DRB1 and tumor necrosis factor gene polymorphisms in Japanese patients with multiple sclerosis. J. Neuroimmunology 92, 109–112. Masterman, T., Ligers, A., Olsson, T., et al. (2000). HLA-DR15 is associated with lower age at onset in multiple sclerosis. Ann. Neurology 48, 211–219. Masterman, T., Ligers, A., Zhang, Z., et al. (2002). CTLA4 polymorphisms inXuence disease course in multiple sclerosis. J. Neuroimmunal. 131, 208–212. Maurer, M., Ponath, A., Kruse, N., et al. (2002) CTLA-4 wxon 1 dimorphism is associated with primary progressive multiple sclerosis J. Neuroimmunal. 131, 213–215.

729

730

30. GENETIC SUSCEPTIBILITY AND EPIDEMIOLOGY

McDonald, W. I., Compston, D. A. S., Edan. G., et al. (2001). International panel on the diagnosis of multiple sclerosis: New diagnostic criteria for multiple sclerosis. Ann. Neurology 50, 121–127. McDonnell, G. V., Kirk, C. W., Hawkins, S. A., and Graham, C. A. (2000). An evaluation of interleukin genes as susceptibility loci for multiple sclerosis. J. the Neurological Sciences, 176, 4–12. McDonnell, G. V., Mawhinney, H., Graham, C. A., et al. (1999). A study of the HLA-DR region in clinical subgroups of multiple sclerosis and its inXuence on prognosis. J. Neurological Science 165(1), 77–83. Marrosu, M. G., Muntoni, F., Murru, M. R., et al. (1992). HLA-DQB1 genotype in Sardinian multiple sclerosis: Evidence for a key role of DQB1. 0201 and DQB1.0302 alleles. Neurology 42, 883–886. Martin, E. R., Lai, E. H., Gilbert, J. R., et al. (2000). SNPing away at complex diseases: Analysis of singlenucleotide polymorphisms around APOE in Alzheimer disease. Am. J. Human Genetics 67, 383–394. Martyn, C. N., Cruddas, M., and Compston, D. A. S. (1993). Symptomatic Epstein-Barr virus infection and multiple sclerosis. J. Neurology Neurosurgery and Psychiatry 56, 167–168. Merriman, A., Cordell, H. J., Eaves, I. A., et al. (2001). Suggestive evidence for association of human chromosome 18q12-q21 and its orthologue on rat and mouse chromosome 18 with several autoimmune diseases. Diabetes 50, 184–194. Mertens, C., Brassat, D., Reboul, J., et al. (1998). A systematic study of oligodendrocyte growth factors as candidates for genetic susceptibility to multiple sclerosis. Neurology 51, 748–753. Modin, H., Dai, Y., Masterman T., et al. (2001). No linkage or association of the nitric oxide synthase genes to multiple sclerosis. J. Neuroimmunology 119, 95–100. The Multiple Sclerosis Genetics Group, Barcellos, L. F., Oksenberg, J. R., et al. (2002). Genetic basis for clinical expression in multiple sclerosis.. Brain 125, 150–158. Mumford, C. J., Wood, N. W., Kellar-Wood, H. F., et al. (1994). The British Isles survey of multiple sclerosis in twins. Neurology 44, 11–15. Olerup, O., and Hillert, J. (1991). HLA class II-associated genetic susceptibility in multiple sclerosis: A critical evaluation. Tissue Antigens 38, 1–15. Oturai, A., Larsen, F., Ryder, L. P., et al. (1999). Linkage and association analysis of susceptibility regions on chromosomes 5 and 6 in 106 Scandinavian sibling pair families with multiple sclerosis. Annals of Neurology 46, 612–616. Pihlaja, H., Rentamaki, T., Wikstrom, J., et al. (2003). Linkage disequilibrium between the MBP tetranucleotide repeat and multiple sclerosis is restricted to a geographically deWned subpopulation in Finland. Genes and Immunity 4, 138–146. Poser, C. M. (1994). The dissemination of multiple sclerosis: A Viking saga? A historical essay. Ann.Neurology 36 (suppl 2), S231–S243. Reboul, J., Mertens, C., Levillayer, F., et al. (2000). Cytokines in genetic susceptibility to multiple sclerosis: A candidate gene approach. J. Neuroimmunology 102, 107–12. Reed, P., Cucca, F., Jenkins, S., et al. (1997). Evidence for a type 1 diabetes susceptibility locus (IDDM10) on human chromosome 10p11-q11. Human Molecular Genetics 6, 1011–6. Reich, D. E., Cargill, M., Bolk, S., et al. (2001). Linkage disequilibrium in the human genome. Nature 411, 199–204. Reynier, P., Pennisson-Besnier, I., Moreau, C., et al. (1999). MtDNA haplopgroup J: A contributing factor of optic neuritis. European J. Human Genetics Apr. 7(3), 404–6. Risch, N. (1990). Linkage strategies for genetically complex traits. Am. J. Human Genetics 46, 222–253. Risch, N., and Merikangas, K. (1996). The future of genetic studies of complex human diseases. Science 23, 1516–1517. Risch, N. J. (2000). Searching for genetic determinants in the new millennium. Nature 405, 847–856. Robertson, N. P., Compston, D. A. S. (1995). Surveying multiple sclerosis in the United Kingdom. J. Neurology Neurosurgery and Psychiatry 58, 2–6. Robertson, N. P., Clayton, D., Fraser, M. B., et al. (1996a). Clinical concordance in sibling pairs with multiple sclerosis. Neurology 47, 347–352. Robertson, N. P., Fraser, M., Deans, J., et al. (1996b). Age adjusted recurrence risks for relatives of patients with multiple sclerosis. Brain 119, 449–455. Robertson, N. P., O’Riordan, J. I., Chataway, J., et al. (1997). Clinical characteristics and oVspring recurrence rates of conjugal multiple sclerosis. Lancet 349, 1587–1590. Rogers, A. E., Meehan, J., Guilleminault, C., et al. (1997). HLA DR15 (DR2) and DQB1*0602 typing studies in 188 narcoleptic patients with cataplexy. Neurology 48, 1550–1556. Rosati, G. (1994). Descriptive epidemiology of multiple sclerosis in Europe in the 1980s: A critical overview. Annals of Neurology 36 (suppl 2), S164–S174. Roxburgh, R., Sawcer, S., Deans, J., et al. (2003). Familial recurrence and genetic risk factors for autoimmune thyroid disease in patients with multiple sclerosis (submitted) Runmarker, B., and Andersen, O. (1995). Pregnancy is associated with a lower risk of onset and a better prognosis in multiple sclerosis. Brain 118, 253–261. Saarela, J., Schoenberg Fejzo, M., Chen, D., et al. (2002). Fine mapping of a multiple sclerosis locus to 2.5 Mb on chromosome 17q22-q24. Human Molecular Genetics 11, 2257–2267. Sadovnick, A. D., Armstrong, H., Rice, G. P. A., et al. (1993). A population-based study of multiple sclerosis in twins: Update. Annals of Neurology 33, 281–285. Sadovnick, A. D., Baird, P. A., Ward, R. H. (1988). Multiple sclerosis: Updated risks for relatives. Am. J. Medical Genetics 29, 533–541.

CONCLUSIONS

Sadovnick, A. D., Ebers, G. C., Dyment, D. A., et al. (1996). Evidence for genetic basis of multiple sclerosis. Lancet 347, 1728–1730. Saruhan-Direskeneli, G., Esin, S., Baykan-Kurt, B., et al. (1997). HLA-DR and -DQ associations with multiple sclerosis in Turkey. Human Immunology 55, 59–65. Sawcer, S., Jones, H. B., Feakes R., et al. (1996). A genome screen in multiple sclerosis reveals susceptibility loci on chromosome 6p21 and 17q22. Nature Genetics 13, 464–468. Sawcer, S., Meranian, M., Setakis, E., et al. (2002). A whole genome screen for linkage disequilibrium in multiple sclerosis conWrms disease associations with regions previously linked to susceptibility. Brain, 125, 1337–1347. Seboun, E., Oksenberg, J. R., Rombos, A., et al. (1999). Linkage analysis of candidate myelin genes in familial multiple sclerosis. Neurogenetics Sep; 2(3), 155–62. Sheremata, W. A., Poskanzer, D. C., Withum, D. G., MacLeod, C. L., and Whiteside, M. E. (1985). Unusual occurrence on a tropical island of multiple sclerosis. Lancet (letter) 2, 618. Skegg, D. C. G., Corwin, P. A., Craven, R. S., Malloch, J. A., and Pollock, M. (1987). Occurrence of multiple sclerosis at the north and south of New Zealand. J. Neurology Neurosurgery and Psychiatry 50, 134–139. Sibley, W. A., Bamford, C. R., and Clark, K. (1985). Clinical viral infections and multiple sclerosis. Lancet i, 1313–1315. Sibley, W. A., Bamford, C. R., Clark, K., Smith, M. S., and Laguna, J. F. (1991). A prospective study of physical trauma and multiple sclerosis. J. Neurology Neurosurgery and Psychiatry 54, 584–589. Siemkowicz, E. (1976). Multiple sclerosis and surgery. Anaesthesia 31, 1211–1216. Siva, A., Radhakrishnan, K., Kurland, L. T., O’Brien, P. C., Swanson, J. W., and Rodriguez, M. (1993). Trauma and multiple sclerosis: A population based cohort study from Olmsted County, Minnesota. Neurology 43, 1878–1882. Soldan, S. S., Leist, T. P., Juhng, K. N., et al. (2000). Increased lymphoproliferative response to human herpesvirus type 6A variant in multiple sclerosis patients. Annals of Neurology 47, 306–313. Sriram, S., Stratton, C. W., Yao, S., et al. (1999). Chlamydia pneumoniae infection of the central nervous system in multiple sclerosis. Annals of Neurology 46, 6–14. Sumelahti, M.-L., Tienari, P. J., Wikstrom, J., et al. (2002). Survival of multiple sclerosis in Finland between 1964–1993. Multiple Sclerosis 8, 350–355. Sutherland, J. M. (1956). Observations on the prevalence of multiple sclerosis in Northern Scotland. Brain 79, 635–654. Taillon-Miller, P., Bauer-Sardina, I., Saccone, N. L., et al. (2000). Juxtaposed regions of extensive and minimal linkage disequilibrium in human Xq25 and Xq28. Nature Genetics 25, 324–328. Tienari, P., Wikstrom, J., Sajantila, A., Palo, J., and Peltonen L. (1992). Genetic susceptibility to multiple sclerosis linked to myelin basic protein gene. Lancet 340, 987–991. The Transatlantic Multiple Sclerosis Genetics Cooperative. (2003). A meta-analysis of genome screens in multiple sclerosis. J. Neuroimmunology (in press). Vaughan, J. H., Riise, T., Rhodes, G. H., et al. (1996). An Epstein Barr virus-related cross reactive autoimmune response in multiple sclerosis in Norway. J. Neuroimmunology 69, 95–102. Weinshenker, B. G., Santrach, P., Bissonet, A. S., et al. (1998). Major histocompatibility complex class II alleles and the course and outcome of MS: A population-based study. Neurology 51, 742–7. Xu, C., Dai, Y., Fredrickson, S., et al. (1999). Association and linkage analysis of candidate chromosomal regions in multiple sclerosis: Indication of disease genes in 12q23 and 7ptr-15. European J. Human Genetics 7, 110–116. Xu, C., Dai, Y., and Lorentzen, J. C., (2001). Linkage analysis in multiple sclerosis of chromosomal regions syntenic to experimental autoimmune disease loci. European J. Human Genetics 9, 458–463. Yamasaki, K., Horiuchi, I., Minohara, M., et al. (1999). HLA-DPB1*0501-associated opticospinal multiple sclerosis: Clinical, neuroimaging and immunogenetic studies. Brain 122, 1689–1696. Zipp, F., Weil, J. G., and Einhaupl, K. M. (1999). No increase in demyelinating diseases after hepatitis B vaccination. Nature Medicine 5, 964–965.

731

C H A P T E R

31 Cellular Damage and Repair in Multiple Sclerosis Hans Lassmann

The basic features of the pathology of multiple sclerosis (MS) were deWned at the end of the 19th century (Charcot, 1868; Marburg, 1906). It is a chronic inXammatory demyelinating disease. This means that an inXammatory process of the central nervous system is associated with the formation of large lesions of primary demyelination with partial axonal preservation and reactive glial scar formation. The lesions are basically formed around small veins and venules, which show focal perivascular inXammation (Dawson, 1916; RindXeisch, 1963). Perivenous lesions, when they enlarge, can fuse with other adjacent demyelinating areas and form large demyelinating plaques, which may reach a diameter of several centimeters. In addition, lesions can grow by radial expansion. In this case, active demyelination occurs at the periphery of the lesion, resulting in a gradual expansion of the plaque into the surrounding normal white matter. Thus, the pathology of multiple sclerosis is deWned by the presence of demyelinated plaques (Carswell, 1838; Cruveilhier, 1841, Figure 31.1). Such demyelinating lesions may occur both in the white as well as the gray matter of the central nervous system, in the latter they are, however, diYcult to identify due to the low density of myelin (Brownell and Hughes, 1962; Lumsden, 1970). Although any brain or spinal cord region can be aVected by the disease process, certain predilection sites are aVected more frequently than others. These are the periventricular white matter, in particular the lateral angles of the lateral ventricles, the subcortical white matter, the optic nerves and chiasm, the cerebellar peduncles with the adjacent cerebellar white matter, and the spinal cord (Fog, 1950; Lumsden, 1970; Steiner, 1931). This uneven distribution of lesions in the central nervous system is due in part to their formation around small to medium-sized veins and venules. This implies that in areas with high density of drainage veins and venules, the probability for initiating an MS lesion is higher than in areas with low venous density (Lassmann, 1983). Multiple sclerosis plaques can be easily identiWed by the unaided eye in the freshly dissected brain (Lumsden, 1970). They appear as sharply demarcated areas of gray discoloration within the white matter. By touching the surface of the cut brain tissue, they reveal a much higher consistency compared to the adjacent tissue. Thus, they represent multiple islets of hard ‘‘sclerotic’’ scar tissue within the brain and this feature is reXected by the name of the disease ‘‘multiple sclerosis.’’ These gray, hard, and sharply demarcated lesions are old chronic plaques. Fresh lesions, which are still in the process of myelin destruction, are in contrast yellow to brown and of soft consistency. In addition, in cases with very severe disease more extensive tissue destruction may lead to cystic loss of tissue, in particular in the center of the lesions. Finally, in some multiple sclerosis patients with an acute, rapidly progressive disease course, very large lesions may appear in the hemispheric

Myelin Biology and Disorders, Volume 2

733

Copyright 2004, Elsevier (USA). All rights reserved.

au1

734

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

FIGURE 31.1 Demyelinated plaques in the central nervous system of MS patients. (A) acute MS (Marburg’s type) with massive brain stem involvement. Demyelinated plaques are seen in the pons, the cerebellar peduncles and the cerebellar hemisphere. Two types of lesions can be diVerentiated: large conXuent demyelinated plaques as well as small perivascular lesions, which may form Wnger-like extensions of the large plaques. Weigert myelin stain;
1. (B) chronic MS with large periventricular plaque surrounding the occipital horn of the lateral ventricle (V) as well as some smaller perivenous lesions in the depth of the white matter. Weigert myelin stain,
0.5.

white matter, which show a concentric layering of myelinated and demyelinated tissue. This variant of multiple sclerosis is called Balo’s concentric sclerosis (Balo, 1928). The hallmark of multiple sclerosis is the plaque with primary demyelination and reactive glial scar formation. Primary demyelination means that myelin sheaths are completely destroyed, while axons remain relatively well preserved in relation to the complete myelin loss. However, as stressed already in the earliest descriptions of multiple sclerosis pathology (for review, see Kornek and Lassmann, 1999), emphasis has to be laid on the term ‘‘relative.’’ Every MS plaque has some degree of axonal injury and loss; its extent varies from plaque to plaque in a given patient but even more between plaques of diVerent patients. The reduction of axonal density in diVerent plaques may range from as few as

THE NATURE OF THE INFLAMMATORY REACTION IN MULTIPLE SCLEROSIS

10% to as much as 90%. As noted already, in the early years of the twentieth century, the presence of lesions with a high extent of axonal loss was generally associated with poor recovery from clinical deWcit in remission periods (Kornek and Lassmann, 1999). Besides these classical plaques, the multiple sclerosis brain frequently shows areas of decreased myelin density. Such areas of partial myelin loss may be due to diVerent underlying pathologies. They may either represent the so-called shadow plaques (Markschattenherde, Schlesinger, 1909). These shadow plaques are very similar to classical MS plaques, being sharply demarcated discolored areas in the white matter with hard sclerotic consistency and located in MS-typical distribution in the central nervous system (CNS). The major diVerence to classical MS plaques is that myelin is not completely lost, but present in reduced density throughout the whole lesion area. In addition, individual myelin sheaths appear unusually thin and the internode length is shortened (Prineas, 1985). As will be discussed in detail, it is now clear that such shadow plaques are areas of remyelination. Besides the classical shadow plaques, other areas of reduced myelin density can be found, which are not sharply demarcated. They are mainly present in patients with multiple lesions at the late stage of the disease. In the brain hemispheres their location shape and distribution is irregular, but in areas of deWned tract systems, such as for instance the spinal cord or the corpus callosum, these lesions follow the anatomical course of preexisting tracts. In histology, such lesions show a profound reduction in the density of myelinated axons, but the thickness of myelin in the preserved Wbers is normal. These lesions represent Wallerian tract degeneration, due to axonal destruction in adjacent demyelinated plaques (Jellinger, 1969; Lumsden, 1970). Finally, in patients with long-standing severe disease diVuse cerebral and spinal cord, atrophy is common. In this situation, the white matter volume is extensively reduced and the ventricles are enlarged. This may also be associated with atrophy of the gray matter, reXected by a narrow cortical ribbon and increased subarachnoide space (Jellinger, 1969). All these tissue alterations occur on the background of a chronic inXammatory process (Charcot, 1868; Marburg, 1906), and it is most likely, but not deWnitely proven, that this inXammatory process is the primary cause of all further destructive events in the central nervous system. For this reason, anti-inXammatory, immunosuppressive, or immunomodulatory treatments are currently in the center of interest in multiple sclerosis therapy.

THE NATURE OF THE INFLAMMATORY REACTION IN MULTIPLE SCLEROSIS Despite the focal nature of the demyelination in multiple sclerosis, inXammation is a diVuse process, which aVects the brain and spinal cord of multiple sclerosis patients as whole. The diVerence in inXammation between plaques and the so-called normal white matter in multiple sclerosis is quantitative, but not qualitative (Allen and McKeown, 1979; Lassmann, 1998). Thus, actively demyelinating lesions in general show more profound inXammation than inactive plaques or the ‘‘unaVected’’ white matter. This is not only the case for inXammatory leucocytic inWltrates in the tissue, but also for the expression of immune associated molecules, such major histocompatibility antigens, adhesions molecules or cytokines (Bo et al., 1994; Cannella and Raine, 1995; Woodroofe and Cuzner, 1993). Thus, there is a diVuse process of inXammation throughout the whole CNS tissue, which is, however, accentuated within the focal lesions. The inXammatory response consists of perivascular inXammatory inWltrates and a diVuse inWltration of the CNS tissue by leukocytes (Prineas and Wright, 1978). The inXammatory inWltrates are essentially composed of mononuclear leukocytes (Fraenkel and Jakob, 1913), which mainly include lymphocytes and macrophages (Traugott et al., 1983; Figure 31.2). In addition, this inXammatory process is associated with a general activation of the local microglia cell population, which may even precede the bulk inWltration

735

736

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

FIGURE 31.2 InXammation in multiple sclerosis lesions. (A) Perivascular inWltrate with mononuclear cells, which mainly consist of T-cells (brown) and macrophages. Immunocytochemistry for CD3;
400. (B) DiVuse inWltrate of T-cells (brown) in the CNS parenchyma within an actively demyelinating lesion. Immunocytochemistry for CD8;
400. (C) Perivascular inWltrate with numerous plasma cells (brown). Immunocytochemistry for IgG;
400. (D) DiVuse inWltrate in an actively demyelinating lesion with T-cells, expressing granzyme B (brown) in their cytotoxic granules. Immunocytochemistry for granzyme B;
400.

of the tissue by hematogenous cells (Gay et al., 1997; Trebst et al., 2001). Activated macrophages and microglia cells are associated with active myelin destruction (Babinski, 1885), and their activation state, as well as their phagocytosis of myelin fragments can be used for exact staging of demyelinating activity of the lesions (Bru¨ck et al., 1995). Within the lymphocyte population, T-cells outnumber B-cells (Esiri, 1980, Esiri et al., 1989). In acute MS and early stages of chronic MS, T-cells are much more abundant than B-cells, generally outnumbering them by 50 to 100 times. This is diVerent in at least a subpopulation of cases with chronic MS, in which the number of B-cells and antibody producing plasma cells may become much more prominent (Ozawa et al., 1994). However, even in these cases T-cells still remain to represent the majority of the lymphocyte population.

T-Lymphocytes It was controversial for many years: What T-cell subpopulation dominates within MS lesions. Some authors claimed that, similar as in autoimmune encephalomyelitis, Class II MHC restricted CD4þ T-lymphocytes prevailed, in particular during the stage of active demyelination (Traugott et al., 1983). This was contradicted by others, who described a profound dominance of CD8þ Class I MHC restricted T-cells in all lesions, irrespective

THE NATURE OF THE INFLAMMATORY REACTION IN MULTIPLE SCLEROSIS

their type or stage of development (Booss et al., 1983). Such a discrepancy is surprising, but may be explained by several technical problems. When these studies were performed, all these markers had to be used on native frozen tissue. Such studies are inevitably restricted to small tissue samples of few patients. In addition, quantitative evaluation of CD4þ T-cells is complicated by the expression of this antigen in activated microglia. Finally, some of these studies were heavily inXuenced by the immunological concept, that MS is, like EAE, an autoimmune disease mediated by Class II restricted T-cells. More detailed recent studies show, that CD8þ T-cells represent 60 to 80% of the total T-cell population within MS plaques (Babbe et al., 2000; Gay et al., 1997). This is the case in active as well as inactive lesions of acute as well as chronic MS. Furthermore, CD8þ T-cells are particularly prominent in the diVuse inWltrates within the CNS tissue, while their relative number is less in the meninges and the perivascular space. It thus appears that Class I MHC restricted T-cells are concentrated within the areas where tissue damage occurs. The mere presence of a given cell population in a lesion alone does not allow conclusions about its pathogenic role or importance in the disease process. Indirect evidence for a possible pathogenic involvement of T-cells can be obtained by determination of clonal expansion of T-cell populations. When T-cells are activated by their speciWc antigenic peptide together with MHC on the surface of an antigen presenting cell, they start to proliferate. If this occurs within a site of inXammation, certain T-cell clones will develop within the tissue. Thus, the degree of clonality of a T-cell population may provide an indirect evidence that they have seen a speciWc antigen and were stimulated to proliferate. By immunological methods, clonal expansion of T-cells within MS lesions has been shown for T-cells with a/b (Oksenberg et al., 1993) and g/d T-cell receptors (Wucherpfennig et al., 1992b), although this has not been reproduced for a/b T-cells in other studies (Birnbaum and Van Ness, 1992; Wucherpfennig et al., 1992a). Since these studies performed T-cell receptor analysis in whole tissue extracts, the clonal expansion of individual cell populations could not be determined. This was overcome recently by applying the method of single cell PCR to analyze T-cell clonality in individual T-cells picked from deWned areas of MS lesions (Babbe et al., 2000). This study clearly showed that it is mainly the CD8þ cell population that is present within the active and inactive lesions, which show clonal expansion, while CD4þ T-cells and T-cells, which are present in the perivascular connective tissue, reveal only a low incidence of clonality. These studies suggest that Class I restricted T-cells not only numerically dominate the inXammatory inWltrate, but possibly recognize their speciWc antigen at the sites of CNS damage in MS. Class I MHC restricted T-cells can either be cytotoxic T-cells, which may directly destroy antigen-containing targets, or they may be regulatory (‘‘suppressor’’) cells. The question thus arises as to whether these CD8þ cells are cytotoxic and involved in the pathogenic process of tissue destruction or whether they down-regulate the inXammatory response. In vitro studies show that these cells can directly recognize their antigen on the surface of diVerent cells of the CNS, such as astrocytes, oligodendrocytes and neurons (Neumann et al., 2002). In addition, they may induce speciWc cytotxicity, mediated either through the content of their cytotoxic granules or through activation of death receptors. They thus have the full potential to destroy target cells, such as oligodendrocytes, neurons, or axons, within multiple sclerosis lesions. In multiple sclerosis lesions, up-regulation of death-associated molecules, involved in T-cell cytotoxicity, such as Fas, Fas-ligand (D’Souza et al., 1996; Dowling et al., 1996), TNF-a, lymphotoxin (Selmaj et al., 1991), and the respective TNF receptors (Bonetti and Raine, 1997), have been described. We have recently analyzed the interaction of cytotoxic, Class I MHC restricted T-cells with local tissue elements in multiple sclerosis. We found, in particular in the lesions of acute MS, a massive expression of granzyme B in the cytotoxic granules of inWltrating T-cells and direct attachment of these granzyme B positive T-cells to oligodendrocytes and axons in actively demyelinating plaques. These data strongly suggest that CD8þ T-cells play an important role in the pathogenesis of the lesions, they do, however, not rule out an additional important pro-inXammatory role of Class II MHC restricted CD4 positive T-cells in the process of inXammation in multiple sclerosis.

737

738

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

B-lymphocytes and Plasma Cells As mentioned earlier, B-lymphocytes and plasma cells are present within MS lesions, although they are much less abundant than T-cells (Esiri, 1977; Prineas and Wright, 1978). They are mainly located in perivascular and meningeal inWltrates, while T-cells are much more dispersed within the CNS tissue in the lesions. Plasma cells in multiple sclerosis brains mainly produce immunoglobulin G; the number of IgA and IgM positive B-cells and plasma cells is substantially lower in comparison to the IgG containing cells (Mussini et al., 1977). Such B-cells and plasma cells are diagnostically important, since they are responsible for the intrathecal production of immunoglobulin, which is a characteristic, but nonspeciWc paraclinical marker used in multiple sclerosis diagnosis. The function of B-cells and plasma cells within MS lesions is so far not clear (Archelos and Hartung, 2000). They may be functionally irrelevant cells, which are secondarily recruited into the lesions in the course of the inXammatory process. This view is supported by the fact that most of the immunoglobulins, which are intrathecally produced in MS patients, are low aYnity antibodies against many diVerent epitopes of viruses and bacteria. B-cells may, however, play a role as antigen presenting cells for T-lymphocytes, which is independent from the speciWcity of the antibodies they produce. Finally, some of the intrathecal B-cells and plasma cells may produce auto-antibodies, directed against myelin proteins, such as myelin basic protein (Gerritse et al., 1994) and demyelinating antibodies may be involved in the pathogenesis of demyelination, at least in a subset of MS patients (Genain et al., 1999; Lucchinetti et al., 2000).

Macrophages and Microglia The vast majority of cells within actively demyelinating lesions in multiple sclerosis are macrophages (Adams and Poston, 1990; Esiri and Reading, 1987). They in general outnumber lymphocytes by at least 10 to 20 times. The show a phenotype of phagocytic macrophages and contain within their cytoplasm the remnants of myelin, which have been released in the demyelinating process (Babinski, 1885, Figure 31.3). These myelin degradation products within macrophages are a useful tool to classify the age of a given demyelinating lesions (Bru¨ck et al., 1995). When macrophages take up myelin fragments in the process of demyelination, these degradation products still contain all myelin proteins. With a few days, however, the minor myelin proteins, such as myelin oligodendrocyte glycoprotein or myelin associated glycoprotein, are degraded, while major myelin proteins like proteolipid protein are still preserved. These major myelin proteins are digested within the macrophages within one to two weeks. Thus, the presence of diVerent myelin proteins within the digestion chambers of macrophages in MS lesions allows a fairly accurate determination of demyelinating activity and lesion age. Macrophages are apparently instrumental in the process of myelin and tissue destruction in multiple sclerosis lesions. During active demyelination, they are closely attached to degenerating myelin sheaths and take up fragments of the dissolved myelin (Babinski, 1885). As discussed in detail in the chapter on EAE, they produce a variety of toxic products, which can destroy myelin in vitro and the can interact with antibodies and complement components, bound to myelin sheaths, by their Fc- and complement receptors (Ulvestad et al., 1994). Macrophages may either be derived from circulating monocytes or from the pool of resident microglia. The extent of macrophage recruitment versus microglia activation contributing to the lesion macrophages in MS is so far not determined, since no reliable markers are available, which allows the distinction of the monocyte or microglia origin of a macrophage. A detailed analysis of the dynamics of plaque formation (Woodroofe et al., 1986; Gay et al., 1997) and of the expression patterns of chemokine receptors within the lesions at diVerent stages of their development (Trebst et al., 2001) suggest that the majority of macrophages are derived from microglia. Although this seems to be the case in the majority of cases, there are some cases of acute MS, where hematogenous monocytes apparently play a much more important role. There, in general, granulocytes are found in addition to

THE NATURE OF THE INFLAMMATORY REACTION IN MULTIPLE SCLEROSIS

FIGURE 31.3 Active demyelination in MS. (A) Active MS plaque with complete loss of myelin and sharp demarcation from the surrounding periplaque white matter. At the left side of the plaque are perivenous areas of demyelination, which in part are in continuity with the plaque. Immunocytochemistry for MOG;
20. (B) Edge of the plaque, stained for macrophages; numerous macrophages are present within the demyelinated lesions and there is a sharp border towards the normal periplaque white matter (WM); immunocytochemistry for CD 68;
80. (C) Plaque edge, stained with Luxol fast blue for myelin. Only few myelin sheaths are preserved and show irregular contours (stage of myelin dissolution). In between there are macrophages, which contain small cytoplasmic granules with the same staining properties as myelin sheaths (early myelin degradation products); Luxol fast blue myelin stain,
600.

macrophages and a large population of cells is present, which expresses an antigen that is predominately present in circulating leukocytes (MRP 14; Bru¨ck et al., 1995).

Expression of Histocompatibility Antigens T-lymphocytes do not recognize their speciWc antigenic peptide in a soluble form. For T-cell activation, antigen recognition has to occur together with MHC-molecules on the surface of antigen presenting cells. Thus, the local expression of MHC molecules is an absolute requirement for antigen recognition by T-cells. The normal CNS tissue diVers from other organs by the low to absent expression of MHC antigens, which is part of the immunoprivilege of the CNS (Hayes et al., 1987). Cells that contain Class I or Class II MHC antigens in the normal brain and spinal cord are perivascular and meningeal macrophages. In addition, cerebral endothelial cells express Class I MHC antigen under normal conditions. Both. Class I and Class II MHC antigens can be induced in cells after stimulation with proinXammatory cytokines, such as gamma-interferon or tumor necrosis factor alpha. Class I antigens can be present widely in all diVerent cells of the CNS, while the expression of Class II MHC antigen is more restricted. The latter is found in the inXamed CNS tissue mainly on microglia and sometimes on astrocytes and ependymal cells.

739

740

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

In MS lesions the expression of MHC antigens is highly regulated by the activity of the inXammatory process. They were mainly described on microglia cells, invading leucocytes and in rare instances on astrocytes. In addition, endothelial cells consitutively express MHC Class I molecules and have also been reported in some studies to be Class II positive (Hayashi et al., 1988; RansohoV and Estes, 1991; Traugott, 1987). In principle, however, also neurons, axons, and oligodendrocytes can express Class I MHC antigens (Neumann, 2002), and such an expression has been found in other inXammatory brain diseases (Bien et al., 2002). To what extent this is also the case in multiple sclerosis has to be determined. MHC Class I and Class II antigens are responsible for the presentation of peptide antigens. The presentation of glycolipid and carbohydrate antigens, however, occurs through CD1 molecules. These molecules are similar to MHC molecules, practically absent in the normal CNS tissue. In MS lesions, however, a massive up-regulation of CD1 is found, mainly on astrocytes but sometimes also on macrophages or microglia cells (Battistini et al., 1996).

Adhesion Molecules, Chemokines and Chemokine Receptors Migration of leukocytes from the circulation into the tissue is the prerequisite for an inXammatory response. This process is regulated by two sets of molecules, the adhesion molecules and the chemokines. Adhesion molecules are cell surface proteins, which mediate binding of cells to other cells or components of the extracellular matrix and can also be involved in intercellular signaling. In inXammatory conditions, adhesion molecules are expressed on the endothelial surface of vessels and interact with their speciWc binding partners on leukocytes. This interaction leads to the attachment of circulating inXammatory cells on the endothelial surface and initiates transmigration of leukocytes through the vessels wall (Springer, 1994). DiVerent molecular classes of adhesion molecules exist, including selectins, integrins, and adhesion molecules of the immunoglobulin gene superfamily. They diVer in their binding properties as well as in their ability to mediate, in addition to adhesion, intracellular signaling upon their ligation and activation. In addition to adhesion molecules, chemokines, interacting with their speciWc receptors, are instrumental in leukocyte migration from the circulation to the tissue (Luster, 1998). Chemokines may act simply as chemoattractants, inducing migration of cells along their concentration gradient. In addition, they may, through intracellular signaling, enhance the binding properties of adhesions molecules and may through this mechanisms facilitate the transmigration of leukocytes through the vessel wall. There are many diVerent chemokines, which act through a substantial number of diVerent chemokine receptors. Importantly, diVerent leukocyte subsets express diVerent chemokine receptor proWles. Thus, it is the chemokines that by and large determine what types of inXammatory cells are recruited into a lesion (Luster, 1998). Both adhesion molecules and chemokines are very attractive targets for anti-inXammatory treatment, since their blockade essentially prevents the entrance of leukocytes into an inXammatory locus. Cerebral endothelial cells in the normal brain express only very few adhesion molecules, in particular intercellular adhesion molecule 2 (ICAM-2). In contrast, in multiple sclerosis lesions, many diVerent adhesion molecules (including selectins, integrins, and immunoglobulin family members) appear on the luminal surface of cerebral endothelial cells and their expression is related to the inXammatory activity in the lesion (Cannella and Raine, 1995; Dore DuVy et al., 1993; Sobel et al., 1990; Washington et al., 1994). This expression is, however, not speciWc for MS, since very similar expression patterns are found in other inXammatory diseases of the central nervous system, being particularly similar in diseases such as virus encephalitis, which too show a T-cell and macrophage dominated inXammatory response. A quite similar situation is found with chemokines. These molecules too, with a few exceptions, are practically absent within the normal mature CNS tissue, but they are heavily up-regulated within the inXammatory foci of multiple sclerosis. Certain members of the chemokine family, in particular those involved in the recruitment of T-cells and macrophages, are prominently expressed in the lesions, both by invading inXammatory

THE NATURE OF THE INFLAMMATORY REACTION IN MULTIPLE SCLEROSIS

cells as well as by astrocytes (McManus et al., 1998; Simpson et al., 1998; Simpson, 2000a, b; Sorensen et al., 1999, 2002). In parallel with these chemokines, the respective receptors are present in MS lesions on invading lymphocytes and monocytes as well as on local microglia (Balashov et al., 1999; Sorensen et al., 1999, 2002; Trebst et al., 2001). The expression of both, the chemokines and their receptors, is tightly associated with the activity of the lesions (both the inXammatory and the demyelinating activity), being most prominently expressed in fresh lesions.

Cytokines The regulation of an inXammatory process within the tissue is accomplished by cytokines, produced either by the inXammatory cells themselves or by local cells within the target tissue. Due to the similarities of the inXammatory response between multiple sclerosis and experimental autoimmune encephalomyelitis and the treatment eVects of certain immunomodulatory compounds, multiple sclerosis is generally believed to be a disease, driven by cytokines of the Th-1 type. This would imply that in initial stages, when new lesions are formed, Th-1-related cytokines such as interferon-gamma, interleukin 12, interleukin 2, and tumor necrosis factor alpha dominate, while during remission Th-2-related or purely immunoregulatory cytokines such as interleukin 10 or transforming growth factor beta prevail. Studies on cytokine expression in MS lesions by immunocytochemistry or in situ hybridization, so far, however, were rather inconclusive. They showed, not unexpectedly, a general up-regulation of cytokine expression in MS lesions (Cannella and Raine, 1995; Merrill, 1992; Woodroofe and Cuzner, 1993). However, both pro-as well as anti-inXammatory cytokines were expressed in parallel and there was no clear-cut patterning of cytokine expression in relation to lesional activity. The source of the cytokines within the lesions are invading leukocytes (Woodroofe and Cuzner, 1993), but also local microglia (Aloisi 2001) and astrocytes (Dong and Benveniste 2001).

Other Markers for Immune Activation In addition, some other immune-associated molecules are expressed in MS lesions, which may play an important role in the initiation or propagation of the lesions. Antigen-speciWc stimulation of T-lymphocytes can be enhanced or modulated by the simultaneous signaling through co-stimulatory molecules. This may also occur directly in the lesions, since such co-stimulatory molecules, such as CD-40 and its ligand or the molecules of the B-7 family, are expressed on macrophages and lymphocytes, respectively (Gerritse et al., 1996; Windhagen et al., 1995). Interaction of CD40 with its ligand is a crucial step in the activation and Th1 polarization of T-cells and, thus, blockade of this pathway may be an interesting approach in anti-inXammatory therapy. The activation of CTLA-4 through B-7 molecules may be involved in down-regulation of the T-cell response. Interestingly, a genetic polymorphism in the CTLA-4 gene has been reported to be associated with multiple sclerosis susceptibility (Ligers et al., 1999).

Neurotrophins InXammation has for long been regarded as being deleterious for the nervous system. Recently, however, experimental studies have shown that traumatic or ischemic injury of the central nervous system can be counteracted by the presence of a moderate inXammatory response, induced by autoimmune T-cells (Moalem et al., 1999). The reason for this paradox observation seems to be that inXammatory cells can be a potent source of neurotrophic factors (Moalem et al., 2000). In fact, human leukocytes produce considerable amounts of bioactive BDNF when stimulated in vitro (Kerschensteiner et al., 1999). BDNF is synthesized by many diVerent inXammatory cells, such as T- and B-lymphocytes, macrophages, and granulocytes. Within inXammatory brain lesions of multiple sclerosis patients, massive local expression of BDNF was found in leukocytes by immunocytochemistry, the intensity of the reaction being much more prominent than in neurons

741

742

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

(Kerschensteiner et al., 1999). In addition, the trk-B neurotrophin-(BDNF-) receptor was present on astrocytes as well as on neurons and axons (Stadelmann et al., 2002). These data suggest that brain inXammation in multiple sclerosis is not always bad, but that it may also have a neuroprotective function or may play a role in the induction or glial scar formation, remyelination, and possibly also axonal sprouting.

THE REACTION OF CEREBRAL VESSELS TO INFLAMMATION IN MULTIPLE SCLEROSIS The normal brain is protected from the entrance of circulating toxic molecules by the blood brain barrier. A chronic inXammatory process in the central nervous system, like that in multiple sclerosis, has several major consequences on blood brain barrier integrity. During the phase of acute inXammation, when leukocytes migrate through the vessel wall, endothelial cells are partly damaged (Minagar et al., 2001). This is associated with a massive disruption of blood brain barrier integrity, mediated either through a disruption of interendothelial junctions (Plumb et al., 2002) or an increased vesicular trans-endothelial transport (Hawkins et al., 1990). Endothelial damage in active MS lesions is suggested by early tracer studies, which show leakage of the blood brain barrier even in post mortem tissue (Broman, 1964). Such a leakage cannot be due to an increased active and energy dependent trans-endothelial transport, but has to reXect opening of the inter-endothelial junctions or endothelial damage. In pathology, this stage is associated with massive penetration of serum proteins into the tissue (Tavolato, 1975), associated with profound extracellular accumulation of Xuid (extracellular edema). When the acute inXammatory stage passes by, the blood brain barrier function becomes partially repaired. However, it has to be emphasized that in old inactive MS plaques there is never a complete restoration of the blood brain barrier (Kwon and Prineas, 1994), but that there is in every old lesion some degree of serum protein leakage. In these vessels endothelial junctions seem to be unaVected, but the endothelial cells show increased numbers of vesicles, indicating enhanced active trans-endothelial transport (Brown, 1978). Structurally, the vessels in old inactive MS plaques are aVected by profound vascular Wbrosis. This means that the vessel wall is detached from the astroglial limiting membrane (Rafalowska et al., 1992) and separated from it by several layers of connective tissue cells and Wbers. This is sometimes associated with the formation of small perivascular channels, which by ultrastructural analysis resemble lymphatic vessels (Prineas, 1979), although they do not express markers for lymphatic endothelial cells. Whether they function as lymphatic vessels has still to be determined. In addition to these changes in structure and function, cerebral vessels in MS plaques also increase in density. This is not only a relative increase, due to loss of tissue elements, such as myelin and axons, but a signiWcant increase in the total number of vessels within the tissue. This is also reXected in fresh lesions by morphological signs of angiogenesis and the expression of markers for angiogenesis, such as VEGF-receptors (Ludwin et al., 2001). This angiogenesis may either be stimulated by cytokines in the inXammatory environment or by hypoxia.

SYNOPSIS OF BRAIN INFLAMMATION IN MULTIPLE SCLEROSIS All the data, discussed so far in relation to inXammation in multiple sclerosis, are consistent with the concept that the disease is driven by a T-lymphocyte mediated inXammatory process. All molecules that are known to be important for recruitment and local activation of T-cells and macrophages are expressed in multiple sclerosis lesions and their degree and type of expression correlates well with the activity of the inXammatory process. Thus, it seems likely that in the initial stage T-cells, which are activated in the peripheral immune system, enter the CNS, where they Wnd their speciWc (possibly cross reacting) antigen. This

DEMYELINATION AND OLIGODENDROCYTE INJURY

leads to reactivation of the T-cells and the production of pro-inXammatory cytokines, such as g-IFN or TNF-a. This in turn stimulates the production of adhesion molecules and chemokines at the site of initial inXammation, thus recruiting other T-cells as well as monocytes and activating the local microglia population. As will be discussed in the following sections in detail, tissue damage may then occur through direct T-cell cytotoxicity, through toxic products of activated macrophages and microglia cells or through additional antibodies, recognizing proteins or glycolipids on the surface of CNS cells. T-cells are locally destroyed by apoptosis, and remission of disease episodes as well as clearance of inXammation from the CNS occurs, when the rate of T-cell apoptosis in the nervous tissue exceeds the generation of new disease propagating T-cells in the peripheral lymphatic tissue and their migration into the lesions. Although these basic principles of inXammation in multiple sclerosis are well established, there are still many open questions. The type of inXammation in multiple sclerosis is very similar to that found in other inXammatory diseases of the central nervous system, such as in particular in diVerent forms of virus encephalomyelitis (Bauer et al., 2001). In fact, so far no feature of inXammation has been reported, which is speciWc for MS and not found in these other diseases. Yet, in MS this inXammatory process is associated with widespread demyelination, which is largely absent in other diseases of virus-induced encephalomyelitis. Thus, there must be some pathogenetic mechanisms, which in multiple sclerosis drive the inXammatory process speciWcally against the myelin/oligodendroglia complex. Another aspect, which needs further attention is the fact that nearly any immune associated molecule that has been looked for so far has been identiWed within MS plaques. This could mean that the immune response in a chronic inXammatory lesion is so redundant and overlapping that a therapy which targets speciWc immunological mechanisms is unlikely to succeed. Alternatively, there may be quantitative diVerences in the involvement of diVerent immunological eVector pathways, some being more and others less important in the formation of the lesions. Such possible quantitative diVerences have, however, so far not been identiWed and their possible existence has so far even not been addressed in proper studies. Finally, it has to be considered that inXammation in multiple sclerosis is a two-edged sword. On the one hand, it induces and propagates disease and tissue damage, but on the other hand, it also may play an important role in limiting tissue destruction and in stimulating remyelination and repair. This leads to an obvious dilemma with regard to anti-inXammatory therapy.

DEMYELINATION AND OLIGODENDROCYTE INJURY As mentioned earlier, primary demyelination is the hallmark of multiple sclerosis pathology (Fig. 31.1). This means that myelin sheaths are completely lost, while axons are still present in high numbers within the lesions and are embedded in a dense astroglial scar tissue. This dissociation between complete myelin loss and partial or even complete preservation of axons is unique for demyelinating diseases and is the most important criterion to distinguish MS lesions from those arising in the course of other diseases. Demyelination in multiple sclerosis is a segmental process (Babinski, 1885). Thus, at the edge of established multiple sclerosis plaques, the myelin sheaths terminate with a node of Ranvier, from where the denuded axons can be traced into astrocytic scar tissue. Active demyelination is associated with inXammation and myelin destruction occurs in close contact between degenerating myelin sheaths with activated macrophages or microglia cells (Babinski, 1885). The myelin sheaths may either be stripped from the axons by processes of phagocytic cells, which penetrate the space between myelin lamellae or the periaxonal space (Lassmann, 1983). Small fragments of myelin may also be removed from the myelinated Wbers by macrophages, forming coated pits and vesicles (Prineas et al., 1984). This type of myelin phagocytosis is reminiscent of a receptor mediated process, possibly involving speciWc anti-myelin antibodies and Fc-receptors or complement receptors on phagocytic cells. In cases with severe and rapid myelin destruction the whole myelin sheaths may be

743

744

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

transformed into small vesicles, a process called ‘‘vesicular disruption of myelin’’ (Guo and Gao, 1983; Kirk, 1979). Again in other cases initial demyelination is accompanied by degenerative changes in the most distal cell processes, which is also reXected by a primary loss of those myelin proteins, which are predominantly located in these areas, such as myelin associated glycoprotein and cyclic nucleotide phosphodiesterase (CNPase). Such a pattern of demyelination is suggestive for a ‘‘distal’’ or ‘‘dying back’’ oligodendrogliopathy (Itoyama et al., 1980, Rodriguez et al., 1993). These data suggest, that the mechanisms of demyelination are diVerent in plaques of diVerent multiple sclerosis patients. Myelin sheaths are formed by oligodendrocytes and it can therefore be expected that in a disease, where myelin is destroyed these cells are aVected too. Indeed, when old chronic completely demyelinated MS plaques are analyzed, there are only very few oligodendrocytes left within the tissue (Lumsden, 1970, Prineas, 1985). As will be discussed below, there are, however, many lesions, which show partial or even complete remyelination, which implies that either oligodendrocytes have survived or that new oligodendrocytes have been recruited from progenitor cells. The fate of oligodendrocytes in acutely demyelinating lesions has been controversial for many years. In some studies very few oligodendrocytes were found in active plaques, suggesting that these cells were lost in parallel or even prior to the myelin sheaths (Lumsden, 1970). However, many other studies showed that in freshly demyelinated plaques, the density of oligodendrocytes may be very high, sometimes even exceeding that in the periplaque white matter (Bru¨ck et al., 1994; Prineas et al., 1984; Raine et al., 1981). This controversy has recently been resolved in a systematic study on a very large sample of cases and lesions. This study showed a profound intra-individual heterogeneity with respect to oligodendrocyte destruction and loss in multiple sclerosis plaques (Lucchinetti et al., 1999). In some patients there was a massive loss of oligodendrocytes already at the earliest stages of demyelination, and there was also no hint for a recruitment of new oligodendrocytes from the pool of progenitor cells. Such patients only had completely demyelinated lesions without remyelination. In most of the patients, however, there was, as described before by Prineas (1985), an initial loss of oligodendrocytes of variable extent in acute lesions, while new oligodendrocytes, apparently derived from progenitor cells, appeared in more advanced plaques. In other patients and lesions, only a minor loss of oligodendrocytes was found in the initial stages of demyelination, which was followed by a progressive loss of these cells at later stages of lesion development (Fig. 31.4). The mode of cell death of oligodendrocytes within actively demyelinating lesions has been described to resemble either necrosis or apoptosis. Again, when a large sample of lesions from diVerent MS patients was analyzed, it became clear that there are major interindividual diVerences in the patterns of oligodendrocyte destruction (Lucchinetti et al., 1999). Within the demyelinated plaque, two essentially diVerent populations of oligodendrocytes can be found (Wolswijk, 2000). One, which is present at the time when myelin falls apart and then gradually disappears, is a small round cell with little or no cell processes, which expresses myelin oligodendrocyte glycoprotein on its surface. It is negative for most other myelin and oligodendrocyte antigens and does not or only in a limited degree contain mRNA for major myelin proteins. It is suggested that these cells represent mature oligodendrocytes, which have survived the demyelinating attack. These cells disappear at later stages of plaque formation, but it is not clear, whether they are removed by programmed cell death or whether they may dediVerentiate and reenter the pool of remyelinating cells. The other cell type is a large and activated cell with elaborate cell processes, which contains abundant mRNA for myelin proteins. It expresses diVerent myelin proteins, including MBP, CNPase, and MAG in its cytoplasm, but shows a low and variable expression of myelin oligodendrocyte glycoprotein on its surface. These cells apparently are immature myelinating cells. It is likely, but so far not deWnitely proven, that these cells are derived from progenitor cells and are the prime candidates, mediating remyelination in the lesions. Summarizing the patterns of myelin and oligodendrocyte damage in multiple sclerosis, it becomes clear that there is a major heterogeneity in the patterns of demyelination between diVerent MS lesions, in particular, when lesions of diVerent patients are analyzed (Lucchinetti et al., 1996). This strongly argues in favor for an inter-individual variability in

OLIGODENDROCYTE PROGENITORS AND REMYELINATION IN MULTIPLE SCLEROSIS LESIONS

FIGURE 31.4 Oligodendrocytes in active multiple sclerosis lesions. (A) Actively demyelinating lesion in chronic MS with oligodendrocytes (black cells) in the periplaque white matter. These cells are absent in the zone of active demyelination (A), but reappear within the demyelinated areas (DM) of the plaque. (B) Actively demyelinating chronic MS lesion with extensive loss of oligodendrocytes (black cells). These cells can be seen in the periplaque white matter, but are nearly completely lost at the actively demyelinating plaque edge (A) and within the demyelinated center of the plaque (DM). In situ hybridization for PLP mRNA (black cells) and immunocytochemistry for PLP protein (red);
200.

the mechanisms of demyelination and tissue destruction in this disease. The data further show that at least in the early stages of disease and lesion formation, there are high numbers of oligodendrocytes within the tissue, and this is associated with extensive remyelination. However, with increasing chronicity of lesions and disease, this remyelinating potential is gradually lost.

OLIGODENDROCYTE PROGENITORS AND REMYELINATION IN MULTIPLE SCLEROSIS LESIONS In chronic multiple sclerosis in patients with long-standing disease the typical lesion in the central nervous system is the completely demyelinated sclerotic plaque. There

745

746

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

remyelination is sparse and restricted to a very small rim at the border to the adjacent periplaque white matter. It has therefore been believed for long time that remyelination and reparation of tissue damage does not occur in multiple sclerosis. The situation, however, is diVerent when cases of acute or rapidly progressive MS are analyzed. Paradoxically, in this most severe clinical subtype of multiple sclerosis, remyelination is extensive and may lead to complete repair of myelin in plaques of several centimeters in diameter (Lassmann, 1983). Such lesions are called ‘‘Markschattenherde’’ or shadow plaques (Fig. 31.5). They were Wrst described by Schlesinger (1912) as sharply demarcated areas of reduced myelin density and reactive gliosis. Schlesinger interpreted these lesions as areas or incomplete demyelination, although possible remyelination was already suggested earlier by Marburg (1906). With the use of ultrastructural techniques and by the analysis of myelin protein and mRNA synthesis, such lesions now can clearly be classiWed as remyelinated plaques. Remyelination is, however, not restricted to patients with acute MS, but also occurs in patients with classical chronic disease (Prineas et al., 1993a, 1993b). Such lesions contain nerve Wbers with unusually thin myelin sheaths and shortened myelin internodes (Prineas, 1985). In addition, the oligodendrocytes within these plaques are increased in size, show signs of metabolic activation, and contain high amounts of mRNAs for diverse myelin proteins (Ozawa et al., 1994). At the peak of remyelination even the myelin proteins, such as myelin basic protein or myelin associated glycoprotein, are present within the perinuclear cytoplasm of these cells (Prineas et al., 1984, 1989). Remyelination may be quite extensive in actively demyelinating, but not in inactive plaques even in patients with long-standing chronic disease. As with demyelination and oligodendrocyte damage, however, the presence and extent of remyelination is highly variable in lesions of diVerent MS patients (Lucchinetti et al., 1999; Ozawa et al., 1994). Not unexpectedly, oligodendrocyte progenitor cells have been identiWed in large numbers in remyelinating MS lesions (Wolswijk, 2000). However, such cells, are even present in chronic demyelinated plaques, which do not show any signs of ongoing myelin repair. They are present in highest numbers in lesions, which still contain macrophages and decrease in number in parallel with the extent of macrophage inWltration of the plaques (Wolswijk, 2002) and with the age of the lesion (Chang et al., 2002; Wolswijk, 2002). However, even in most chronic plaques, some progenitor cells are present, in spite of complete absence of remyelination (Chang et al., 2002). Thus, in summary there is a high potential for remyelination in multiple sclerosis lesions, but it fails to be eVective in a large proportion of lesions and patients. Overall remyelination is much more eVective in early stages of the disease and in actively demyelinating lesions than at later stages or in inactive plaques, although there are additional major diVerences in the remyelinating capacity between diVerent MS patients (Lucchinetti et al., 1999). There are several diVerent explanations for this complex situation. First, the extent of remyelination may depend on the mechanism of demyelination, which vary between diVerent patients (Lucchinetti et al., 2000). Apparently in many patients the destructive process involves mature myelin and oligodendrocytes, leaving progenitor cells unaVected. In such a situation, new remyelinating oligodendrocytes can be recruited and remyelination can be accomplished (Prineas, 1993b). In some patients, however, progenitor cells seem to be aVected together with mature oligodendrocytes and myelin. Certain macrophage toxins—such as, for instance, excitotoxins—aVect progenitor cells more eVectively than mature oligodendrocytes (Fern and Moller, 2000). Thus, when demyelination is mediated through such toxins, both the mature and immature oligodendrocytes will be lost. In addition, some patients develop antibodies, which are directed against the AN-2/ NG-2 epitope (Niehaus et al., 2000). Such antibodies may selectively destroy progenitor cells within an inXammatory demyelinating lesion and thus reduce the number of cells capable for remyelination of the lesions. Another mechanism may involve demyelinating antibodies. When they are present in high titers, they may directly inhibit remyelination (Seil et al., 1975) or destroy remyelinating oligodendrocytes at the moment when they express their respective target antigen, such as, for instance, myelin oligodendrocyte glycoprotein. In such lesions the remyelination

OLIGODENDROCYTE PROGENITORS AND REMYELINATION IN MULTIPLE SCLEROSIS LESIONS

FIGURE 31.5 Remyelinated, shadow plaques in multiple sclerosis. (A) Sharply demarcated plaque with reduced myelin staining intensity, oriented around a central vein (V). Luxol fast blue myelin stain;
20. (B) Sharply demarcated remyelinated shadow plaque (S) and a demyelinated plaque (D) without remyelination located side by side. Within the remyelinated area and the periplaque white matter there are numerous oligodendrocytes. In the demyelinated plaque these cells are only present at the plaque margin; Co: Cortex; in situ hybridization for PLP mRNA (black) and PLP protein (red);
10.

program is started, but new myelin and oligodendrocytes are quickly removed during maturation (Storch et al., unpublished). Remyelination may also be impaired due to multiple demyelinating episodes, occurring in the same brain area. Remyelinated MS plaques may frequently become the target of new demyelinating attacks (Prineas et al., 1993b; Figure 31.6), possibly due to the disturbed integrity of the blood brain barrier within these areas. In experimental models, it has been shown that repeated demyelinating episodes in the same area may decrease the capacity for remyelination (Linington et al., 1992; Ludwin, 1980). It is thus feasible that in MS demyelination occurs in a given area, which is followed by remyelination. During one of the next bouts of the disease, the same area becomes aVected with a new demyelinating episode, which again is followed by remyelination (Prineas et al., 1993b). When this happens several times, the pool of progenitor cells may be depleted and remyelination may be severely compromised. Finally, it has to be taken into account that in some chronic, persistently demyelinated lesions oligodendrocyte progenitor cells are present, which, however, fail to start the remyelinating process. It has been suggested that in such lesions, axons may be defect in a way that they are unable to initiate the remyelinating program (Chang et al., 2002). Alternatively, progenitor cells may require stimulation by trophic factors, which are produced by the inXammatory cells and in particular by macrophages. Indeed, depletion of macrophages impairs remyelination in experimental models (Kotter et al., 2001). In both, acute as well as chronic MS lesions remyelination is most prominent during and immediately following the stage of acute demyelination and the extent of remyelination correlates with the degree of macrophage inWltration in the lesions. Thus, lack of remyelination in chronic inactive MS plaques may in part be due to the lack of ‘‘beneWcial’’ inXammation.

747

748

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

FIGURE 31.6 Repeated demyelination in the same area of the CNS. One demyelinated plaque without remyelination (D) overlaps with a remyelinated shadow plaque (S). This suggests that an old remyelinated plaque partially became demyelinated again during the formation of the new demyelinated plaque. Axonal staining in Figures B and D shows reduction of axonal density in both plaques, being more pronounced in the area of overlap. (A, C) Immunocytochemistry for MOG. (B, D) Bielschowsky silver impregnation; C and D are higher magniWcation images of the areas labeled by * in Figures A and B; A and B:
20; Cc and D:
180.

THE NEURODEGENERATIVE COMPONENT IN MULTIPLE SCLEROSIS: NEURONAL AND AXONAL INJURY IN INFLAMMATORY DEMYELINATING LESIONS Primary demyelination is the key feature of multiple slcerosis pathology. This means that in comparison to the complete destruction and loss of myelin sheaths axons are relatively spared within the plaques (Charcot, 1868, for review see Kornek and Lassmann, 1999). The term ‘‘relative,’’ however, implies that axons are not completely unaVected and that in every MS lesion there is some degree of axonal injury and loss (Marburg, 1906). Its extent is variable between diVerent plaques in the same MS patient and even more variable between plaques of diVerent patients (Fig. 31.7). In addition, MS lesions are not restricted to the white matter but may also aVect gray matter areas, such as for instance the cortex. Gray matter lesions too are deWned by demyelination, but neuronal and axonal injury is profound (Peterson et al., 2001). Axonal injury in plaques is associated with secondary Wallerian degeneration, aVecting tracts that pass through the demyelinated areas (Evangeou et al., 2000a). Furthermore, it may result in anterograde or retrograde neuronal degeneration, which may further contribute to neuronal and axonal loss (Evangelou et al., 2001). As a result of all these neurodegenerative events, multiple sclerosis in patients with long-standing and severe disease is generally associated with brain atrophy, aVecting both the gray and white matter (Lumsden, 1970). All these neurodegenerative features have been noticed and carefully described in the earliest reports of multiple sclerosis pathology (Kornek and Lassmann, 1999), but have not been appreciated very much during past decades. Only recently magnetic resonance imaging studies brought these feature back into the center of interest, since these studies suggest that it is this neurodegenerative component, which plays an important role in the establishment of permanent functional deWcit and irreversible clinical disease (see the chapter on MRI).

THE NEURODEGENERATIVE COMPONENT IN MULTIPLE SCLEROSIS

FIGURE 31.7 Axonal loss in multiple sclerosis lesions. Chronic active multiple sclerosis with multiple plaques in the subcortical white matter. Axonal staining shows profound diVerences in axonal density within diVerent lesions. In some lesions, reduction of staining intensity is moderate (*), while in others there is nearly complete loss of tissue integrity (**). Bielschowsky silver impregnation;
10.

Axonal Injury in Demyelinated Plaques In comparison to the adjacent normal periplaque, white matter the density of axons is always reduced in demyelinated plaques of MS patients (Lassmann, 1998). This reduction of axonal density is highly variable, ranging from less than 10% to up to 90%. A decreased axonal density not necessarily has to reXect true axonal loss, since it may also occur when axons are separated from each other by pronounced tissue edema and the inWltrating leukocytes. Thus, in acute plaques it is in general more pronounced in very early stages of demyelination than in later ones, where inXammation and edema have partially been cleared. Quantitative assessment of true axonal loss is diYcult to perform in the human central nervous system and is therefore restricted so far to deWned tract areas, such as for instance the spinal cord, where total numbers of axons of a given tract can be counted (Bjatmar et al., 2000). This study showed an average of true axonal loss between 60 and 70%, which is quite comparable to the data obtained by evaluating relative axonal density in brain lesions (Mews et al., 1998). In addition, clinico-pathological correlation suggests that overt clinical deWcit requires a profound loss of axons in a given tract system, which is in the range of 40 to 50%. This means that there is a very high reserve capacity in the human brain and spinal cord, which allows functional compensation of a loss of nearly half of the axons in a functionally important system such as the pyramidal tract. Axonal loss in the spinal cord is not restricted to the demyelinated plaques, but occurs in quite similar extent also in white matter areas, devoid of demyelinated plaques (Bjartmar et al., 2001; Ganter et al., 1999; Lovas et al., 2000). Since demyelinated plaques in MS patients may aVect all levels of the cord, the most reasonable explanation for this observation is that axonal loss in myelinated areas of the spinal cord reXects tract degeneration due to demyelinating lesions in upper or lower spinal cord segments. Such a mechanism may also explain the rather diVuse atrophy of the cord, seen in MRI images of MS patients, which is apparently not related to the location of individual demyelinated lesions. Such tract degeneration has also convincingly been shown to occur in the corpus callosum in relation to deep white matter lesions (Evangelou et al., 2000b). Acute axonal injury results in a disturbance of axonal transport, which is reXected by focal axonal swelling and the accumulation of amyloid precursor protein (Fig. 31.8). Both features can be used to identify acutely injured axons in multiple sclerosis lesions (Ferguson et al., 1997; Trapp et al., 1998). Acute axonal injury is most pronounced in

749

750

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

FIGURE 31.8 Axonal pathology in actively demyelinating plaque in multiple sclerosis. Reduced axonal density and multiple axons with focal enlargements of their diameter (axonal spheroids), suggesting acute disturbance of axonal transport. Immuncytochemistry for phosphorylated neuroWlament;
300.

actively demyelinating MS lesions (Ferguson et al., 1997; Kornek et al., 2000; Trapp et al., 1998). In addition, however, there is a low density of acutely injured axons also in chronic established demyelinated plaques, which do not show any signs of ongoing destruction of myelin (Kornek et al., 2000). This slow burning axonal degeneration in old lesions is not seen in remyelinated shadow plaques. Thus, in MS axonal destruction seems to occur in two diVerent steps. Many axons are injured during a limited period of time, spanning the phase of active demyelination. In addition, there seems to be a slow process of axonal injury, which occurs in inactive plaques and may progress for months or years. Little is known so far about the mechanisms of axonal damage in multiple sclerosis. Within the lesions as well as in the ‘‘normal’’ white matter, small Wbers are more severely aVected than thick axons (Evangelou et al., 2000a, 2001). The extent of acute axonal injury correlates with the number of activated macrophages (Ferguson et al., 1997) and—much less signiWcantly—with the density of cytotoxic T-cells in the lesions (Bitsch et al., 2000). This is not only the case in acute lesions, but also in inactive demyelinated ‘‘silent’’ plaques (Kornek et al., 2000). It is thus likely that toxic mediators, produced by activated macrophages and microglia cells, play a key role in the induction of axonal death. Macrophage produced proteases (Anthony et al., 1998) and nitric oxide radicals (Smith and Lassmann, 2001) are Wrst line candidates for such an eVect. Further axonal disintegration may be due to calcium inXux into the axolemma and subsequent activation of calcium dependent proteases (Kornek et al., 2001).

Gray Matter Lesions/Cortical Plaques Multiple sclerosis is generally regarded as a disease aVecting the white matter. This is due to the fact that white matter lesions in this disease can easily be seen by pathologists and are distinctly visible in magnetic resonance images. However, gray matter areas are frequently aVected in this disease, as for instance in cases with brain stem and spinal cord involvement. In addition, cortical plaques are abundant in patients with chronic multiple sclerosis. In some patients up to 60% of all hemispheric lesions involve the cortex (Brownell and Hughes, 1962; Lumsden, 1970). As in white matter lesions, those in the gray matter and in particular cortical plaques are deWned by the demyelinating process (Lumsden, 1970). Thus, myelin sheaths are completely lost, while other structures of the cortex are relatively well preserved. According to their shape and distribution, several types of cortical lesions can be distinguished (Kidd et al., 1999; Peterson et al., 2001). One type is deWned as a subcortical white matter plaque, which extends into the cortex. Other lesions are sharply demarcated perivascular intracortical lesions, which, however, do not reach the subcortical white matter. The third type is

MECHANISMS OF DEMYELINATION AND TISSUE DAMAGE IN MULTIPLE SCLEROSIS LESIONS

oriented toward the outer surface of the cortex. It is generally large, aVecting one or more gyri as a whole and extending from the outer surface into the three to four outer layers of the cortex. Thus, in such lesions there is demyelination from the molecular to the inner granular layer of the cortex, while myelin in the deeper cortical layers remains intact (Peterson et al., 2001). Since the density of myelin sheaths in the normal cortex is low, such cortical plaques are diYcult to identify and generally require for their detection very sensitive myelin stains. Gray mater lesions and in particular cortical plaques diVer in several aspects from those arising in the white matter (Peterson et al., 2001). Although also in cortical lesions inXammation by lymphocytes and macrophages as well as microglia activation are present, the density of macrophages and microglia cells even in actively demyelinating lesions is about 10 times lower than in white matter plaques. In addition, large foamy macrophages with myelin degradation products, the hallmark of active white matter lesions are hardly detectable in active cortical plaques, remnants of myelin sheaths can mainly be found in small ramiWed activated microglia cells. Similar as in white matter plaques there is profound axonal injury in cortical lesions but in addition, neuronal damage can also be seen. Some neurons within the lesions show DNA fragmentation and sometimes the neuronal expression of activated caspase 3 as a marker for apoptotic cells death can be seen. Oligodendrocyte injury in cortical plaques is reXected by accumulation of the complement component C4d in their cytoplasm (Schwab and McGeer, 2002). Thus, cortical lesion can be abundant in MS patients and may in such patients contribute to a major extent to the clinical disease burden. These lesions are mainly found in patients with progressive MS at late disease stages, but are rare in the early stages of the disease and nearly absent in patients with Marburg’s type of acute MS (Lassmann, unpublished). In comparison to white matter plaques, the extent of the inXammatory response is less, but in addition to demyelination, neurodegeneration is prominent. All these neurodegenerative changes within the central nervous system of MS patients Wnally lead to extensive brain changes and atrophy, denominated ‘‘MS-encephalopathy’’ (Jellinger, 1969). Brain atrophy aVects both, the white and the gray matter, although in general the white matter is more severely damaged (Prineas and McDonald, 1997). White matter atrophy is diVuse and aVects all portions of the hemispheres, some areas, such as the corpus callosum, however, are more severely aVected than others. It is characteristically associated with dilatation of the cerebral ventricles. Gray matter, and in particular the cortex too can show some degree of atrophy, reXected by a widening of cerebral sulci and a thinning of the cortical ribbon. It is, however, much less extensive than that in the white matter and is found, when at all, mainly in patients with very severe chronic disease of long duration. Atrophy cannot only be seen in pathology, but is quantiWable in magnetic resonance images and thus became a useful quantitative indicator for progression of brain damage in multiple sclerosis patients (Fisher et al., 2000).

MECHANISMS OF DEMYELINATION AND TISSUE DAMAGE IN MULTIPLE SCLEROSIS LESIONS There is little doubt that multiple sclerosis lesions develop on the background of a T-lymphocyte driven inXammatory response in the central nervous system (Noseworthy et al., 2000). Some data suggest that the inXammatory response is due to an autoimmune reaction against central nervous system antigens (Martino and Hartung, 1999). Arguments in favor of the autoimmune concept are that all attempts to transmit the disease to animals so far failed, that immune suppression has a beneWcial eVect on disease development, and that autoimmune reactions, both on the level of T-cells as well as of antibodies, have been found in multiple sclerosis patients. Proof for a pathogenic autoimmunity in MS patients is indirect; current evidence does not rule out the involvement of an infectious agent at some time points in disease development (Noseworthy et al., 2000). Indeed, epidemiological data argue in favor of an exogenous agents being involved in the induction of the disease.

751

752

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

The induction and development of disease in MS is, however, to a large extent determined by the host’s genes. Although the genetic contribution to disease development is high, there is not a single gene responsible for this eVect. In fact, multiple genes are involved in the control of this disease and the genetic risk factors appear to diVer between patient subgroups. MS is thus a chronic disease with complex pathogenetic background, determined by multiple genes as well as by environmental factors (Kalman and Lublin, 1999; Noseworthy et al., 2000). The pathology of multiple sclerosis apparently reXects a common basic mechanism of inXammatory tissue damage, which is principally mediated by T-lymphocytes and activated macrophages or microglia cells. This basic pathology can then in diVerent patients be modiWed by additional immune reactions, which may amplify disease and tissue damage. In addition, the genetic susceptibility of the target tissue may to a major extent inXuence the extent and type of tissue damage within the lesions (Lassmann et al., 2001). Based on this concept of pathogenesis, diVerent patterns of active demyelination, occurring in diVerent subgroups of MS patients, have recently been deWned (Lucchinetti et al., 2000).

The Basic Pattern of T-cell Mediated Inflammatory Tissue Damage The basic inXammatory reaction within the CNS is similar in all MS patients, irrespective of the severity or the clinical course of the disease. It consists of an inWltration, dominated by cytotoxic Class I MHC restricted T-lymphocytes, which is associated with recruitment of hematogenous macrophages and a massive activation of local microglia (Babbe et al., 2000; Gay et al., 1997). Demyelination and tissue destruction in such lesions apparently is mediated by the cytotoxic T-cells themselves, as well as by activated macrophages. Several diVerent eVector pathways may be involved. For cytotoxic T-cells, it may be either the direct action of cytotoxic granules, destroying the target tissue through perforin and granzymes, or the activation of so-called death receptors, such as the Fas- or the TNF receptors (Neumann et al., 2002). In addition to T-cells, however, activated macrophages and microglia cells seem to play a major role. These cells can destroy local target cells through various toxins (see the chapter on experimental autoimmune encephalomyelitis). Many cytotoxic molecules have been identiWed in multiple sclerosis lesions, mainly produced by activated macrophages and microglia cells or some of them also by astrocytes, and their possible function in MS lesions has been extensively reviewed. Proteolytic Enzymes Proteases, and in particular matrix metallproeinases and their inhibitors are highly expressed in actively demyelinating multiple sclerosis lesions (Cuzner et al., 1996; Maeda and Sobel, 1996), but are absent in normal human brain. They are mainly present in activated macrophages and in part in lymphocytes. They may play a major role in the inXammatory cascade by promoting leukocyte migration into the lesions (Cuzner and Opdenakker, 1999). When released in the lesions in high concentration, they may also induce myelin and axonal damage, although, due to the poor selectivity of tissue damage induced by proteases, they are unlikely to be responsible for the large areas of primary demyelination in MS (Anthony et al., 1998) Tumor Necrosis Factor Members of the TNF family are present predominantly in active MS lesions, but to a lower extent they are also present in inactive plaques. Lymphotoxin-alpha is mainly found in lymphocytes and microglia, while TNF-a dominates in macrophages, microglia and some astrocytes. (Hofman et al., 1989; Selmaj et al., 1991). Oligodendrocytes and microglia cells express the respective TNF-receptors (TNF-R1 and TNF-R2; Bonetti and Raine, 1997). TNF, is a pro-inXammatory cytokine, which appears to be involved in the recruitment of leukocytes into the lesions. In addition, oligodendrocytes are particularly vulnerable for the induction of TNF-mediated cell death in vitro and in vivo (Probert et al., 2000).

MECHANISMS OF DEMYELINATION AND TISSUE DAMAGE IN MULTIPLE SCLEROSIS LESIONS

Reactive Oxygen and Nitrogen Specie (ROS and RNI) Direct visualization of ROS and RNI within tissue sections is so far not possible. However, the enzymes involved in their synthesis as well as downstream footprints of oxidative and nitrative damage have been identiWed in actively demyelinating multiple sclerosis lesions. The cellular source within the lesions are macrophages, microglia cells, and possibly astrocytes (Smith et al., 1999; Smith and Lassmann 2002). Tissue damage induced by ROS and RNIs is mostly accomplished in cooperation—for instance, by the formation of peroxynitrite. They may play diVerent roles in the pathogenesis of multiple sclerosis lesions. They may increase blood brain barrier permeability and are cytotoxic for oligodendrocytes and neurons. In addition, they can block axonal function and induce axonal degeneration (for details see Smith and Lassmann, 2002). Excitotoxins Macrophages are a potent source of excitotoxins (Lipton, 1998). Since oligodendrocytes as well as neurons and axons are highly susceptible to excitotoxic injury, such a mechanism may in part be responsible for tissue damage in an inXammatory demyelinating disease (Pitt et al., 2000; Smith et al., 2000). An imbalanced glutamate homeostasis has recently been suggested to occur in multiple sclerosis lesions, reXected by increased glutaminase expression in macrophages or microglia and reduced expression of glutamate dehydrogenase in oligodendrocytes (Werner et al., 2001). Complement Components Activated complement, in particular activation of the terminal lytic complement complex, is present in a subset of MS cases, located on myelin at sites of active demyelination (Compston et al.; 1989; Prineas et al.; 2001; Storch et al., 1998). Its activation on myelin may either occur nonspeciWcally or through speciWc antibodies, bound to the myelin surface. Myelin and oligodendrocytes are particularly vulnerable to the toxic eVects of complement activation (Benn et al., 2001). General Significance of Macrophage Mediated Tissue Injury in MS Lesions The previously discussed data provide compelling evidence that macrophage toxins may play a crucial role in demyelination and tissue damage in multiple sclerosis. However, macrophages and microglia cells are nonspeciWc cells of the defense system and do not recognize speciWc antigenic targets. It is thus surprising that an inXammatory reaction, which induces tissue damage through activated macrophages, is able to mediate selective primary demyelination. The reason for this appears to reside in a diVerential vulnerability of diVerent target cells and structures for macrophage mediated damage. Overall, in vitro studies show that there is a diVerential vulnerability of diVerent brain cells to most macrophage toxins, oligodendrocytes and myelin being the most susceptible, followed by neurons and axons, astrocytes, and, as the least susceptible cell, the microglia or macrophage itself (Benn et al., 2001). Thus, direct application of macrophage toxins on CNS tissue in vitro or in vivo leads at low concentration to selective primary demyelination, while higher concentration increasingly also destroy other cell types. However, there are many virus-induced diseases of the nervous system that show similar inXammation and macrophage activation in the lesions, but lack primary demyelination. Thus, there has to be some speciWc mechanism that drives the process in MS so selectively against myelin sheaths and oligodendrocytes. Whether this is accomplished by speciWc cytotoxic T-cells or autoantibodies has to be determined. A question of major importance for the design of new therapeutic strategies is, what macrophage or microglia toxin is the most relevant in the pathogenesis of multiple sclerosis lesions? Studies in the model of EAE show that in principle all diVerent molecules are involved in the destructive process, their relative importance varies between diVerent models (see the chapter on experimental autoimmune encephalomyelitis). In MS lesions, all these molecules are expressed and detailed quantitative studies on their level of expression in diVerent MS patients are not available. Thus, in a situation like this, a prediction regarding

753

754

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

the best therapeutic target is diYcult. Two key players as cytotoxic eVector molecules in inXammatory tissue damage are either tumor necrosis factor alpha or nitrogen and oxygen radicals. Both of them seem to be involved at several steps of the injury cascade in inXammatory demyelinating lesions. Yet experimental studies aimed to therapeutically block these molecules in EAE revealed controversial results. Indeed, in many studies disease was not ameliorated but rather augmented. The reason for this dilemma is that such macrophage toxins are not only involved in inXammatory tissue damage, but are also essential for the elimination of pathogenic T-lymphocytes and thus for the termination of the inXammatory response (Probert et al., 2000; Smith and Lassmann, 2002; Willenborg et al., 1999).

Modulation of T-Cell Mediated Inflammation by Antibodies In a subgroup of MS patients, speciWc antibodies seem to be involved in demyelination and tissue injury (Lucchinetti et al., 2000). Active lesions in such cases show profound deposition of immunoglogulin and activated complement on the surface of myelin sheaths and oligodendrocytes at the sites where myelin is falling apart (Prineas and Graham, 1981; Storch et al., 1998; Figure 31.9). This may be associated with the interaction of macrophages with myelin fragments through coated pits and vesicles, suggestive of a receptor mediated phagocytosis (Prineas et al., 1984). Whether the antibodies that are bound to degenerating myelin are really directed against myelin oligodendrocyte glycoprotein, as suggested in one study (Genain et al., 1999), has to await conWrmation. Possible other targets for demyelinating antibodies are, in particular, glycolipids, such as galactocerebroside (Dubois-Dalq et al., 1970). Besides demyelination, speciWc antibodies could also play other roles in the pathogenesis of MS lesions. In a subgroup of MS patients, circulating antibodies against oligodendrocyte progenitor cells have been detected, which could, when they reach the lesion, destroy these cells and impair remyelination (Niehaus et al., 2000). In other patients, massive immunoglobulin and complement deposition is found in vessel walls, which may indicate additional antibody mediated vasculitis as a potentiation factor of inXammation (Lucchinetti et al., 2002). Within the spectrum of multiple sclerosis patients, the most intense antibody involvement can be found in cases with Devic’s type of neuromyelitis optica (Luchinetti et al., 2002). In these patients, the massive complement activation is also associated with an inXammatory inWltrate, which contains signiWcant numbers of granulocytes, including eosinophilic granulocytes. This granulocyte component, which is also observed in the cerebrospinal Xuid of the respective patients, can be diagnostically useful. It may further help in the therapeutic decision to concentrate on immunosuppressive strategies, which in particular ameliorate B-lymphocyte and antibody responses.

Modification of MS Lesions by Hypoxia-Like Tissue Damage In some MS patients, active myelin destruction follows a structural pattern, which is closely reXected in acute lesions of white matter stroke (Lassmann et al., 2001). This type of demyelination is in its initial phases characterized by a loss of myelin associated glycoprotein and cyclic nucleotide phosphodiesterase, while other myelin proteins, such as myelin basic protein, proteolipid protein, or myelin oligodendrocytes glycoprotein, initially remain preserved (Itoyama et al., 1980). This process is associated with nuclear condensation and fragmentation of oligodendrocytes, structurally resembling apoptotic cell death (Lucchinetti et al., 2000). When the oligodendrocytes are subsequently destroyed, myelin sheaths completely fall apart and are taken up by macrophages or microglia cells. Active lesions, which are demyelinated by this type of injury, diVer in several aspects from the previously described MS plaques. They are ill demarcated from the surrounding periplaque white matter, and in some instances they may show a concentric, rim-like pattern of demyelination. In addition, inXamed vessels in the center of the lesions are frequently surrounded by a rim of preserved myelin (Fig. 31.10). These changes of myelin are closely similar to those occurring in initial stages of ischemic myelin damage in white matter stroke lesions. In addition, this pattern of

MECHANISMS OF DEMYELINATION AND TISSUE DAMAGE IN MULTIPLE SCLEROSIS LESIONS

FIGURE 31.9 Active demyelination in a multiple sclerosis patient with antibody mediated demyelination. Massive deposition of activated complement (C9neo-antigen) on damaged myelin sheaths and in myelin degradation products within macrophages in the areas of demyelinating activity (A); very little deposition of activated complement in the surrounding normal white matter (WM). Immunocytochemistry for C9neo-antigen;
900.

FIGURE 31.10 Actively demyelinating lesion in a patient with acute multiple sclerosis and a pattern of demyelination, resembling hypoxia/ischemia. In contrast to other MS lesions, these plaques are not sharply demarcated; MOG (Fig. A) is partially preserved, while other myelin components are lost (Fig. B, stained with Luxol fast blue); the central vessel in the plaque (*) shows a perivascular cuV of inXammatory cells, but is surrounded by a rim of preserved myelin. Within the lesion there is massive perivascular inXammation (Fig. C), but the diVuse tissue inWltration with macrophages or activated microglia cells is less profound compared to MS lesions, following other types of demyelination (see Figure 31.3). (A) Immunocytochemistry for MOG;
50. (B) Luxol fast blue myelin stain;
50. (C) Immunocytochemistry for CD68;
100.

demyelination, being present in ischemia or in multiple sclerosis lesions, is always associated with increased expression and nuclear translocation of hypoxia inducible factor 1 alpha, which is a sensitive and speciWc sensor for decreased oxygen tension in the tissue (Lassmann, unpublished). Hypoxia-like tissue damage can be induced in inXammatory brain lesions by a variety of diVerent mechanisms. It may just occur as a consequence of edema in tissues, which are constrained in swelling by connective tissue or bone barriers, as they exist in the spinal cord or the optic nerve (Prineas and McDonald, 1997). Such an edema-induced increase in tissue pressure may result in disturbance of microcirculation with subsequent ischemia. Hypoxia and ischemia could also be induced through vasculitic reactions, which may lead to thrombotic occlusion of cerebral vessels in MS plaques (WakeWeld et al., 1994). Vasculitis in brain inXammation can either be mediated by cytotoxic T-cells or by speciWc antibodies, both recognizing their speciWc antigens within the vessel wall. Finally, hypoxia-like tissue injury can also be induced by toxic inXammatory mediators. Toxicity of nitric oxide radicals is at least in part mediated by a blockade of mitochondrial function within the respective target cells (Bolanos et al., 1997). Mitochondrial failure will lead to a hypoxia-like metabolic tissue injury even under normoxic conditions.

755

756

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

Hypoxia-like tissue injury is frequently observed in patients with Marburg’s type of acute multiple sclerosis, but is rare in the lesions of patients with classical chronic MS. Interestingly, all MS lesions with concentric demyelinating lesions of Balo’s disease type follow in their acute stage the pattern of hypoxia-like demyelination.

Differential Susceptibility of the Target Tissue for Inflammatory Demyelination in Multiple Sclerosis Patients In addition to the quality and quantity of the immune response, the genetic susceptibility of the target tissue may inXuence the outcome of inXammatory demyelinating lesions. This has been shown in the model of autoimmune encephalomyelitis, in which, for instance, a deWciency of ciliary neurotrophic factor in the CNS may augment demyelination and oligodendrocyte destruction (Linker et al., 2002). In multiple sclerosis too, the genetic properties of the target tissue may modify the pathology of the lesions. In some patients, demyelination and oligodendrocyte destruction is signiWcantly more severe compared to other patients with identical quality and quantity of the inXammatory response (Lucchinetti et al., 2000). Similarly, the extent of axonal destruction and neurodegeneration is highly variable between patients, in spite of very similar patterns of inXammation (Lassmann, 1998). Genetic studies in large cohorts of multiple sclerosis patients revealed various diVerent candidate genes, which may modify the course of multiple sclerosis. They have no major inXuence on the global disease incidence, but they are associated with more severe disease or with speciWc atypical features of the disease. As examples, deWciency of CNTF has been described to be associated with earlier onset and more severe disease (Giess et al., 2002). Other gene defects or polymorphisms that are associated with certain atypical features of the disease are related to genes for apolipoprotein E (Fazekas et al., 2000) or mitochondrial DNA (Mojon et al., 1999). The speciWc pathological substrate of this modiWcation of multiple sclerosis is so far not known, but it is expected that it may be related to the degree of demyelination, oligodendrocyte destruction, and axonal injury, but may also aVect the topography of the lesions.

CONCLUSIONS Numerous very elaborate studies of the central nervous system in multiple sclerosis patients have deWned an inXammatory demyelinating disease process as the basic pathological process. InXammation is apparently driven by a T-lymphocyte mediated immune response, which may be directed against one or more autoantigens of the nervous system, although the additional contribution of unknown infectious agents cannot be excluded. Demyelination is associated with a variable extent of axonal and neuronal injury, which may lead to secondary degeneration of respective tracts or projection areas. The mechanisms of demyelination and tissue destruction is complex and heterogeneous between patients. This heterogeneity is most likely controlled by genes of the patients, inXuencing qualitatively and quantitatively the type of the immune response that is responsible for tissue injury. In addition, genetic factors appear to control the susceptibility of the target tissue for immune-mediated injury and to determine the capacity of the tissue for remyelination and repair. MS is thus not unlike other chronic immunological diseases, in which an exogenous trigger initiates a cascade of events, which in their quantity and quality are determined by the genetic background of the patient. The complexity of such a disease, however, explains, why curative therapies are so diYcult to establish. Since the immune system is armed with many diVerent eVector mechanisms, which in many situations of immune defense or autoimmunity work in parallel, treatments focusing on a single mechanism are unlikely to succeed. In addition, the heterogeneity in the mechanisms of demyelination between diVerent MS patients implies that diVerent patient subgroups will require diVerent therapeutic strategies.

CONCLUSIONS

Apart from these complications, there are other implications for therapy from the data reviewed here. Obviously in an inXammatory demyelinating disease, inXammation is one of the prime targets for intervention. In this case, however, it has to be considered that inXammation in the CNS is not only bad, but can also through the production of neurotrophins limit damage and stimulate remyelination or tissue repair. Thus, a modulatory therapy, changing the quality of the inXammatory reaction, may be the better choice than global immune suppression. In addition, other therapies should be developed, aimed to stimulate remyelination and to prevent axonal and neuronal degeneration. Remyelination in MS plaques seems to be instrumental to prevent further axonal degeneration in established lesions. Finally, axonal protection appears to be crucial. Functional deWcit, induced by inXammation and demyelination, is potentially reversible, while that due to axonal loss is irreversible. Since the Wnal pathway of axonal injury and destruction seems to be similar in inXammatory conditions like MS and other pathologies such as stroke or trauma, it can be expected that neuroprotective therapies developed in the latter conditions will also be beneWcial in multiple sclerosis patients.

References Adams, C. W., and Poston, R. N. (1990). Macrophage histology in paraYn-embedded multiple sclerosis plaques is demonstrated by the monoclonal pan-macrophage marker HAM-56: Correlation with chronicity of the lesion. Acta. Neuropathol. (Berl) 80, 208–211. Allen, I. V., and McKeown, S. R. (1979). A histological, histochemical and biochemical study of the macroscopically normal white matter in multiple sclerosis. J. Neurol. Sci. 41, 81–91. Aloisi, F. (2001). Immune function of microglia. Glia 36, 165–179. Anthony, D. C., Miller, K. M., Fearn, S., Townsend, M. J., Opdenakker, G., Wells, G. M., Clements, J. M., Chandler, S., Gearing, A. J., and Perry, V. H. (1998). Matrix metalloproteinase expression in an experimentally-induced DTH model of multiple sclerosis in the rat CNS. J. Neuroimmunol. 87, 62–72. Archelos, J. J., and Hartung, H. P. (2000). Pathogenetic role of autoantibodies in neurological disease. Trends Neurosci. 23, 317–327. Babbe, H., Roers, A., Waisman, A., Lassmann, H., Goebels, N., Hohlfeld, R., Friese, M., Schro¨der, R., Deckert, M., Schmidt, S., Ravid, R., and Rajewsky, K. (2000). Clonal expansion of CD8þ T cells dominate the T cell inWltrate in active multiple sclerosis lesions as shown by micromanipulation and single cell polymerase chain reaction. J. Exp. Med. 192, 393–404. Babinski, J. (1885). Recherches sur l’anatomie pathologique de la sclerose en plaque et etude comparative des diverses varietes de la scleroses de la moelle. Arch. Physiol. (Paris) 5–6, 186–207. Balashov, K. E., Rottman, J. B., Weiner, H. L., and Hancock, W. W. (1999). CCR5(þ) and CXCR3(þ) T cells are increased in multiple sclerosis and their ligands MIP-1alpha and IP-10 are expressed in demyelinating brain lesions. Proc. Natl. Acad. Sci. U.S.A. 96, 6873–6878. Balo, J. (1928). Encephalitis periaxialis concentrica. Arch. Neurol. 19, 242–264. Battistini, L., Fischer, F. R., Raine, C. S., and Brosnan, C. F. (1996). CD1b is expressed in multiple sclerosis lesions. J. Neuroimmunol. 67, 145–151. Bauer, J., Rauschka, H., and Lassmann, H. (2001). InXammation in the nervous system: The human perspective. Glia 36, 235–243. Benn, T., Halfpenny, C., and Scolding, N. (2001). Glial cells as targets for cytotoxic immune mediators. Glia 36, 200–211. Bien, C. G., Bauer, J., Deckwerth, T. L., Wiendl, H., Deckert, M., Wiestler, O. D., Schramm, J., Elger, C. E., and Lassmann, H. (2002). Destruction of neurons by cytotoxic T cells: A new pathogenic mechanism in Rasmussen’s encephalitis. Ann. Neurol. 51, 311–318. Birnbaum, G., and van Ness, B. (1992) Quantitation of T-cell receptor V beta chain expression on lymphocytes from blood, brain, and spinal Xuid in patients with multiple sclerosis and other neurological diseases. Ann. Neurol. 32, 24–30. Bitsch, A., Schuchardt, J., Bunkowski, S., Kuhlmann, T., and Bruck, W. (2000). Acute axonal injury in multiple sclerosis. Correlation with demyelination and inXammation. Brain 2000 123, 1174–83. Bjartmar, C., Kidd, G., Mork, S., Rudick, R., and Trapp, B. D. (2000). Neurological disability correlates with spinal cord axonal loss and reduced N-acetyl aspartate in chronic multiple sclerosis patients. Ann. Neurol. 48, 893–901. Bjartmar, C., Kinkel, R. P., Kidd, G., Rudick, R. A., and Trapp, B. D. (2001). Axonal loss in normal-appearing white matter in a patient with acute MS. Neurology 57, 1248–52. Bo, L., Mork, S., Nyland, H., Pardo, C. A., and Trapp, B. D. (1994). Detection of MHC class II antigens on macrophages and microglia, but not on astrocytes and endothelia in active multiple sclerosis lesions. J. Neuroimmunol. 51, 135–146. Bolanos, J. P., Almeida, A., Stewart, V., Peuchen, S., Land, J. M., Clark, J. B., and Heales, S. J. (1997). Nitric oxide-mediated mitochondrial damage in the brain: Mechanisms and implications for neurodegenerative diseases. J. Neurochem. 68, 2227–40.

757

758

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

Bonetti, B., and Raine, C. S. (1997). Multiple sclerosis: Oligodendrocytes display cell death-related molecules in situ but do not undergo apoptosis. Ann. Neurol. 42, 74–84. Booss, J., Esiri, M. M., Tourtellotte, W. W., and Mason, D. Y. (1983). Immunohistological analysis of T lymphocyte subsets in the central nervous system in chronic progressive multiple sclerosis. J. Neurol. Sci. 62, 219–232. Broman, T. (1964). Blood brain barrier damage in multiple sclerosis. Supra-vital test observations. Acta. Neurol. Scand. 10, 21–24. Brown, W. J. (1978). The capillaries in acute and subacute multiple sclerosis plaques: A morphometric analysis. Neurology 28, 84–92. Brownell, B., and Hughes, J. T. (1962). The distribution of plaques in the cerebrum in multiple scleosis. J. Neurol. Neurosurg. Psychiat. 25, 315–320. Bru¨ck, W., Porada, P., Poser, S., Riechmann, P., Hanefeld, F., Kretschmar, H-A., and Lassmann, H. (1995). Monocyte/macrophage diVerentiation in early multiple sclerosis lesions. Ann. Neurol. 38, 788–796. Bru¨ck, W., Schmied, M., Suchanek, G., Bru¨ck, Y., Breitschopf, H., Poser, S., Piddlesden, S., and Lassmann, H. (1994). Oligodendrocytes in the early course of multiple sclerosis. Ann. Neurol. 35, 65–73. Cannella, B., and Raine, C. S. (1995). The adhesion molecule and cytokine proWle of multiple sclerosis lesions. Ann. Neurol. 37, 424–435. Carswell, R. (1838). ‘‘Pathological Anatomy: Illustrations on Elementary Forms of Disease.’’ Longman, London. Chang, A., Tourtellotte, W. W., Rudick, R., and Trapp, B. D. (2002). Premyelinating oligodendrocytes in chronic lesions of multiple sclerosis. New Engl. J. Med. 346, 165–173. Charcot, J. M. (1868). Histologie de la sclerose en plaque. Gaz Hopital (Paris) 41, 554–566. Compston, D. A., Morgan, B. P., Campbell, A. K., Wilkins, P., Cole, G., Thomas, N. D., and Jasani, B. (1989). Immunocytochemical localization of the terminal complement complex in multiple sclerosis. Neuropathol. Appl. Neurobiol. 15, 307–16. Cruveilhier, J. (1829–1842). ‘‘Anatomie pathologique du corps humain.’’ JB Bailliere, Paris. Cuzner, M. L., Gveric, D., Strand, C., Loughlin, A. J., Paemen, L., Opdenakker, G., and Newcombe, J. (1996). The expression of tissue-type plasminogen-activator, matrix metalloproteinases and endogenous inhibitors in the central nervous system in multiple sclerosis: Comparison of stages of lesion evolution. J. Neuropathol. Exp. Neurol. 55, 1194–1204. Cuzner, M. L., and Opdenakker, G. (1999). Plasminogen activators and matrix metalloproteinases, mediators of extracellular proteolysis in inXammatory demyelination of the central nervous system. J. Neuroimmunol. 94, 1–14 D’Souza, S. D., Bonetti, B., Balasingam, V., Cashman. N. R., Barker, B. A., Troutt, A. B., Raine, C. S., and Antel, J. P. (1996). Multiple sclerosis: Fas signaling in oligodendrocyte death. J. Experimental Medicine 184, 2361–2370. Dawson, J. W. (1916). The histology of disseminated sclerosis. Trans. R. Soc. 50, 517–540. Dong, Y., and Benveniste, E. N. (2001). Immune function of astrocytes. Glia 36, 180–190. Dore-DuVy, P., Washington, R., and Dragovic, L. (1993). Expression of endothelial cell activation antigens in microvessels from patients with multiple sclerosis. Adv. Exp. Med. Biol. 331, 243–248. Dowling, P., Shang, G., Raval, S., Menona, J., Cook, S., and Husar, W. (1996). Involvement of the CD95 (APO1/ Fas) receptor/ligand system in multiple sclerosis brain. J. Exp. Med. 184, 1513–1518. Dubois-Dalcq, M., Niedieck, B., and Buyse, M. (1970). Action of anti-cerebroside sera on myelinated nervous tissue cultures. Pathologica Europea 5, 331–347. Esiri, M. M. (1977). Immunoglobulin-containing cells in multiple sclerosis plaques. Lancet 2, 478–480. Esiri, M. M., and Reading, M. C. (1987). Macrophage populations associated with multiple sclerosis plaques. Neuropathol. Appl. Neurobiol. 13, 451–465. Esiri, M. M., Reading, M. C., Squier, M. V., and Hughes, J. T. (1989). Immunocytochemical characterization of the macrophage and lymphocyte inWltrate in the brain in six cases of human encephalitis. Neuropathol. Appl. Neurobiol. 15, 289–305. Esiri, M. M. (1980). Multiple sclerosis: A quantitative and qualitative study of immunoglobulin-containing cells in the central nervous system. Neuropathol. Appl. Neurobiol. 6, 9–21. Evangelou, N., Esiri, M. M., Smith, S., Palace, J., and Matthews, P. M. (2000a). Quantitative pathological evidence for axonal loss in normal appearing white matter in multiple sclerosis. Ann. Neurol. 47, 391–5. Evangelou, N., Konz, D., Esiri, M. M., Smith, S., Palace, J., and Matthews, P. M. (2000b). Regional axonal loss in the corpus callosum correlates with cerebral white matter lesion volume and distribution in multiple sclerosis. Brain 123, 1845–9. Evangelou, N., Konz, D., Esiri, M. M., Smith, S., Palace, J., and Matthews, P. M. (2001). Size-selective neuronal changes in the anterior optic pathways suggest a diVerential susceptibility to injury in multiple sclerosis. Brain 124, 1813–20. Fazekas, F., Strasser Fuchs, S., Schmidt, H., Enzinger, C., Ropele, S., Lechner, A., Flooh, E., Schmidt, R., and Hartung, H. P. (2000). Apolipoprotein E genotype related diVerences in brain lesions of multiple sclerosis. J. Neurol. Neurosurg. Psychiat. 69, 25–28. Ferguson, B., Matyszak, M. K., Esiri, M. M., and Perry, V. H. (1997). Axonal damage in acute multiple sclerosis lesions. Brain 120, 393–9. Fern, R., and Moller, T. (2000). Rapid ischemic cell death in immature oligodendrocytes: A fatal glutamate release feedback loop. J. Neurosci. 20, 34–42.

CONCLUSIONS

Fisher, E., Rudick, R. A., Cutter, G., Baier, M., Miller, D., Weinstock-Guttman, B., Mass, M. K., Dougherty, D. S., and Simonian, N. A. (2000). Relationship between brain atrophy and disability: An 8-year follow-up study of multiple sclerosis patients. Mult. Scler. 6, 373–7. Fog, T. (1950). Topographic distribution of plagues in the spinal cord in multiple sclerosis. A.M.A. Arch Neurol Psychiat 63, 382–414. Fraenkel, M., and Jakob, A. (1913). Zur Pathologie der multiplen Sklerose mit besonderer Beru¨cksichtigung der akuten Formen. Z. Neurol. 14, 565–603. Ganter, P., Prince, C., and Esiri, M. M. (1999). Spinal cord axonal loss in multiple sclerosis: A post-mortem study. Neuropathol. Appl. Neurobiol. 25, 459–67. Gay, F. W., Drye, G. W., Dick, G. W. A., and Esiri, M. M. (1997). The application of multifactorial cluster analysis in the staging of plaques in early multiple sclerosis: IdentiWcation and characterization of the primary demyelinating lesion. Brain 120, 1461–1483. Genain, C. P., Cannella, B., Hauser, S. L., and Raine, C. S. (1999). Autoantibodies to MOG mediate myelin damage in MS. Nat. Med. 5, 170–175. Gerritse, K., Deen, C., Fasbender, M., Ravid, R., Boersma, W., and Claassen, E. (1994). The involvement of speciWc anti myelin basic protein antibody-forming cells in multiple sclerosis immunopathology. J. Neuroimmunol 49, 153–159. Gerritse, K., Laman, J. D., Noelle, R. J., AruVo, A., Ledbetter, J. A., Boersma, W. J., and Claassen,-E. (1996). CD40-CD40 ligand interactions in experimental allergic encephalomyelitis and multiple sclerosis. Proc. Natl. Acad. Sci. U.S.A. 93, 2499–504. Giess, R., Maurer, M., Linker, R., Gold, R., Warmuth-Metz, M., Toyka, K. V., Sendtner, M., and Rieckmann, P. (2002). Association of a null mutation in the CNTF gene with early onset of multiple sclerosis. Arch. Neurol. 59, 407–9. Guo, Y. P., and Gao, S. F. (1983). Concentric sclerosis. In ‘‘Clinical and Experimental Neurology. Proc Australian Association of Neurologists,’’ (J. H. Tyrer and M. J. Eadie, eds.), 19; pp. 67–76. Adis Health Science Press Sydney. Hawkins, C. P., Munro, P. M., MacKenzie, F., Kesselring, J., Tofts, P. S., du-Boulay, E. P., Landon, D. N., and McDonald, W. I. (1990). Duration and selectivity of blood-brain barrier breakdown in chronic relapsing experimental allergic encephalomyelitis studied by gadolinium-DTPA and protein markers. Brain 113, 365–78. Hayashi, T., Morimoto, C., Burks, J. S., Kerr, C., and Hauser, S. L. (1988). Dual-label immunocytochemistry of the active multiple sclerosis lesion: Major histocompatibility complex and activation antigens. Ann. Neurol. 24, 523–31. Hayes, G. M., Woodroofe, M. N., and Cuzner, M. L. (1987). Microglia are the major cell type expressing MHC class II in human white matter. J. Neurol. Sci. 80, 25–37. Hofman, F. M., Hinton, D. R., Johnson, K., and Merrill, J. E. (1989). Tumor necrosis factor identiWed in multiple sclerosis brain. J. Exp. Med. 170, 607–612. Itoyama, Y., Sternberger, N. H., Webster, H. deF., Quarles, R. H., Cohen, S. R., and Richardson, E. P., Jr. (1980). Immunocytochemical observation on the distribution of myelin-associated glycoprotein and myelin basic protein in multiple sclerosis lesions. Ann. Neurol. 7, 167–177. Jellinger, K. (1969). Einige morphologische Aspekte der Multiplen Sklerose. Wien Z Nervenheilk (Suppl) II,12–37. Kalman, B., and Lublin, F. D. (1999). The genetics of multiple sclerosis. A review. Biomed. Pharmacother. 53, 358–70. Kerschensteiner, M., Gallmeier, E., Behrens, L., Leal, V. V., Misgeld, T., Klinkert, W. E., Kolbeck, R., Hoppe, E., Oropeza-Wekerle, R. L., Bartke, I., Stadelmann, C., Lassmann, H., Wekerle, H., and Hohlfeld, R. (1999). Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor in vitro and in inXammatory brain lesions: A neuroprotective role of inXammation? J. Exp. Med. 189, 865–70., Kidd, T., Barkhof, F., McConnell, R., Algra, P. R., Allen, I. V., and Revesz, T. (1999). Cortical lesions in multiple sclerosis. Brain 122, 17–26. Kirk, J. (1979). The Wne structure of the CNS in multiple sclerosis. Vesicular demyelination in an acute case. Neuropathol. Appl. Neurobiol. 5, 289–294. Kornek, B., and Lassmann, H. (1999). Axonal pathology in multiple sclerosis: A historical note. Brain Pathol. 9, 651.656. Kornek, B., Storch, M. K., Bauer, J., Djamshidian, A., Weissert, R., Wallstrom, E., SteVerl, A., Zimprich F., Olsson, T., Linington, C., Schmidbauer, M., and Lassmann, H. (2001). Distribution of calcium channel subunit in dystrophic axons in multiple sclerosis and experiemntal autoimmune encephalomyelitis. Brain 124, 1114–1124. Kornek, B., Storch, M., Weissert, R., Wallstroem, E., SteVerl, A., Olsson, T., Linington C., Schmidbauer, M., and Lassmann, H. (2000). Multiple sclerosis and chronic autoimmune encephalomyelitis: A comparative quantitative study of axonal injury in active, inactive and remyelinated lesions. Amer. J. Pathol. 157, 267–276. Kotter, M. R., Setzu, A., Sim, F. J., Van-Rooijen, N., and Franklin, R. J. (2001). Macrophage depletion impairs oligodendrocyte remyelination following lysolecithin-induced demyelination. Glia. 35, 204–12. Kwon, E. E., and Prineas, J. W. (1994). Blood brain barrier abnormalities in longstanding multiple sclerosis lesions. An immunohistochemical study. J. Neuropath. Exp. Neurol. 53, 625–636. Lassmann, H. (1983). Comparative neuropathology of chronic experimental allergic encephalomyelitis and multiple sclerosis. Springer Schriftenr. Neurol. 25:1–135. Lassmann, H. (1998). Pathology of multiple sclerosis. In ‘‘McAlpine’s Multiple Sclerosis’’ (A. Compston, ed.), pp323–358. Churchill Liningstone, London. Lassmann, H., Bru¨ck, W., and Lucchinetti, C. (2001). Heterogeneity of multiple sclerosis pathogenesis: Implications for diagnosis and therapy. Trends Mol. Med. 7, 115–121.

759

760

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

Ligers, A., Xu, C., Saarinen, S., Hillert, J., and Olerup, O. (1999). The CTLA-4 gene is associated with multiple sclerosis. J. Neuroimmunol. 97, 182–190. Linington, C., Engelhardt, B., Kapocs, G., and Lassmann, H. (1992). Induction of persistently demyelinated lesions in the rat following the repeated adoptive transfer of encephalitogenic T cells and demyelinating antibody. J. Neuroimmunol. 40, 219–24. Linker, R. A., Maurer, M., Gaupp, S., Martini, R., Holtmann, B., Giess, R., Rieckmann, P., Lassmann, H., Toyka, K. V., Sendtner, M., and Gold, R. (2002). CNTF is a major protective factor in demyelinating CNS disease: A neurotrophic cytokine as modulator in neuroinXammation. Nat. Med. 8, 620–4. Lipton, S. A. (1998). Neuronal injury associated with HIV-1: Approaches and treatment. Annu. Rev. Pharmacol. Toxicol. 38, 159–177. Lovas, G., Szilagyi, N., Majtenyi, K., Palkovits, M., and Komoly, S. (2000). Axonal changes in chronic demyelinated cervical spinal cord plaques. Brain 123, 308–17. Lucchinetti, C., Bru¨ck, W., Parisi, J., Scheithauer, B., Rodriguez, M., and Lassmann, H. (2000). Heterogeneity of multiple sclerosis lesions: Implications for the pathogenesis of demyelination. Ann. Neurol. 47, 707–717. Lucchinetti, C., Bru¨ck, W., Parisi, J., Scheithauer, B., Rodriguez, M., and Lassmann, H. (1999). A quantitative analysis of oligodendrocytes in multiple sclerosis lesions. A study of 117 cases. Brain 122, 2279–2295. Lucchinetti, C. F., Bru¨ck, W., Rodriguez, M., and Lassmann, H. (1996). Distinct patterns of multiple sclerosis pathology indicates heterogeneity in pathogenesis. Brain Pathol. 6, 259–274. Lucchinetti, C. F., Mandler, R., McGavern, D., Bru¨ck, W., Gleich, G., RansohoV, R. M., Trebst, C., Weinshenker, B., Wingerchuck, D., Parisi, J., and Lassmann, H. (2002). A role for humoral mechanisms in he pathogenesis of Devic’s neuromyelitis optica. Brain (in press). Ludwin, S. K. (1980). Chronic demyelination inhibits remyelination in the central nervous system. An analysis of contributing factors. Lab. Invest. 43, 382–387. Ludwin, S. K., Henry, J. M., McFarland, H. (2001). Vascular proliferation and angiogenesis in multiple sclerosis: Clinical and pathogenetic implications. J. Neuropathol. Exp. Neurol. 60, 505. Lumsden, C. E. (1970) The neuropathology of multiple sclerosis. In ‘‘Handbook of clinical Neurology’’ (Vinken, P. I., Bruyn, and G. W., eds.), Vol. 9, pp. 217–309. Elsevier, New York. Luster, A. D. (1998). Chemokines — chemotactic cytokines that mediate inXammation. N. Engl. J. Med. 338, 436–445. Maeda, A., and Sobel, R. A. (1996). Matrix metalloproteinases in the normal human central nervous system, microglia nodules and multiple sclerosis lesions. J. Neuropathol. Exp. Neurol. 55, 300–309. Marburg, O. (1906). Die sogenannte ‘‘akute Multiple Sklerose’’ Jahrb Psychiatrie 27, 211–312. Martino, G., and Hartung, H. P. (1999). Immunopathogenesis of multiple sclerosis: The role of T cells. Curr. Opin. Neurol. 12, 309–321. McManus, C., Berman, J. W., Brett, F. M., Staunton, H., Farrell, M., and Brosnan, C. F. (1998). MCP-1, MCP-2 and MCP-3 expression in multiple sclerosis lesions: An immunohistochemical and in situ hybridization study. J. Neuroimmunol. 86, 20–9. Merrill, J. E. (1992). ProinXammatory and antiinXammatory cytokines in multiple sclerosis and central nervous system acquired immunodeWciency syndrome. J Immunother. 12, 167–170. Mews, I., Bergmann, M., Bunkowski, S., Gullotta, F., and Bru¨ck, W. (1998). Oligodendrocyte and axon pathology in clinically silent multiple sclerosis lesions. Multiple Sclerosis 4, 55–62. Minagar, A., Jy, W., Jimenez, J. J., Sheremata, W. A., Mauro, L. M., Mao, W. W., Horstman, L L., and Ahn, Y. S. (2001). Elevated plasma endothelial microparticles in multiple sclerosis. Neurology. 56, 1319–24. Moalem, G., Gdalyahu, A., Shani, Y., Otten, U., Lazarovici, P., Cohen, I. R., and Schwartz, M. (2000). Production of neurotrophins by activated T cells: Implications for neuroprotective autoimmunity. J. Autoimmun. 15, 331–45. Moalem, G., Leibowitz-Amit, R., Yoles, E., Mor, F., Cohen, I. R., and Schwartz, M. (1999). Autoimmune T cells protect neurons from secondary degeneration after central nervous system axotomy. Nat. Med. 5, 49–55. Mojon, D., Fujihara, K., Hirano, M., Miller, C., LincoV, N., Jacobs, D., and Greenberg, S. (1999). Leber’s hereditary optic neuropathy mitochondrial DNA mutations in familial multiple sclerosis. Graefes Arch. Clin. Exp. Ophtalmol. 237, 348–350. Mussini, J. M., Hauw, J. J., and Escourolle, R. (1977). ImmunoXuorescence studies of intra cytoplasmatic immunoglobulin binding lymphoid cells in the central nervous system. Report of 32 cases including, 19 multiple sclerosis. Acta. Neuropathol. (Berl.) 40, 227–232. Neumann, H., Medana, I., Bauer, J., and Lassmann, H. (2002). Cytotoxic T lymphocytes in autoimmune and degenerative CNS diseases. Trend Neurosci. 25, 313–319. Niehaus, A., Shi, J., Grzenkowski, M., Diers-Fenger, M., Hartung, H. P,. Toyka, K., Bruck, W., and Trotter, J. (2000). Patients with active relapsing-remitting multiple sclerosis synthesize antibodies recognizing oligodendrocyte progenitor cell surface protein: Implications for remyelination. Ann. Neurol. 48, 362–71. Noseworthy, J. H., Lucchinetti, C., Rodriguez, M., and Weinshenker, B. G. (2000). Multiple sclerosis. New Engl. J. Med. 343, 938–952. Oksenberg, J. R., Panzara, M. A., Begovich, A. B., Mitchell, D., Ehrlich, H. A., Murray, R. S., Shimonkevitz, R., Sherritt, M., Rothbard, J., and Bernard, C. C. (1993). Selection for T-cell receptor V beta-D beta-J beta gene rearrangements with speciWcity for a myelin basic protein peptide in brain lesions of multiple sclerosis. Nature 362, 68–70. Ozawa, K., Suchanek, G., Breitschopf, H., Bru¨ck, W., Budka, H., Jellinger, K., and Lassmann, H. (1994). Patterns of oligodendroglia pathology in multiple sclerosis. Brain 117, 1311–1322.

CONCLUSIONS

Peterson, J. W., Bo, L., Mork, S., Chang, A., and Trapp, B. D. (2001). Transected neurites, apoptotic neurons and reduced inXammation in cortical multiple sclerosis lesions. Ann. Neurol. 50, 389–400. Pitt, D., Werner, P., Raine, C. S. (2000). Glutamate excitotoxicity in a model of multiple sclerosis. Nature Med. 6, 67–70. Plumb, J., McQuaid, S., Mirakhur, M., and Kirk, J. (2002). Abnormal endothelial tight junctions in active lesions and normal appearing white matter in multiple sclerosis. Brain Pathol. 12,, 199–211. Prineas, J. W. (1979). Multiple sclerosis: Presence of lymphatic capillaries and lymphoid tissue in the brain and spinal cord. Science 203, 1123–1125. Prineas, J. W. (1985). The neuropathology of multiple sclerosis. In ‘‘Handbook of Clinical Neurology’’ (J. C. Koetsier, ed.), Vol. 47, pp. 337–395. Elsevier, New York. Prineas, J. W., Barnard, R. O., Kwon, E. E, Sharer, L. R., and Cho, E. S. (1993a). Multiple sclerosis: Remyelination of nascent lesions. Ann. Neurol. 33, 137–151. Prineas, J. W., Barnard, R. O., Revesz, T., Kwon, E. E., Sharer, L., and Cho, E. S. (1993b). Multiple sclerosis: Pathology of recurrent lesions. Brain 116, 681–693. Prineas, J. W., and Graham, J. S. (1981). Multiple sclerosis: Capping of surface immunoglobulin G on macrophages engaged in myelin breakdown. Ann. Neurol. 10, 149–158. Prineas, J. W., Kwon, E. E., Cho, E. S., and Sharer, L. R. (1984). Continual breakdown and regeneration of myelin in progressive multiple sclerosis plaques. Ann. NY Acad. Sci. 436, 11–32. Prineas, J. W., Kwon, E. E., Cho, E. S., Sharer, L. R., Barnett, M. H., Oleszak, E. L., HoVman, B., and Morgan, B. P. (2001). Immunopathology of secondary-progressive multiple sclerosis. Ann. Neurol. 50, 646–657. Prineas, J. W., Kwon, E. E., Goldenberg, P. Z., Ilyas, A. A., Quarles, R. H., Benjamins, J. A., and Sprinkle, T. J. (1989). Multiple sclerosis: Oligodendrocyte proliferation and diVerentiation in fresh lesions. Lab. Invest. 61, 489–503. Prineas, J. W., and McDonald, I. W. (1997). Demyelinating diseases. In ‘‘GreenWeld’s Neuropathology’’ (D. I. Graham and P. L. Lantos, eds.), 6th ed., pp. 813–896. Arnold, London, Sidney, Auckland. Prineas, J. W., and Wright, R. G. (1978). Macrophages, lymphocytes, and plasma cells in the perivascular compartment in chronic multiple sclerosis. Lab. Invest. 38, 409–421. Probert, L., Eugster, H. P., Akassoglou, K., Bauer, J., Frei, K., Lassmann, H., and Fontana, A. (2000). TNFR1 signalling is critical for the development of demyelination and the limitation of T-cell responses during immune-mediated CNS disease. Brain 123, 2005–19. Rafalowska, J., Krajewski, S., Dolinska, E., and Dziewulska, D. (1992). Does damage to perivascular astrocytes in multiple sclerosis participate in blood brain barrier permeability. Neuropathol. Pol. 30, 73–80. Raine, C. S., Scheinberg, L., and Waltz, J. M. (1981). Multiple sclerosis: Oligodendrocyte survival and proliferation in an active established lesion. Lab. Invest. 45, 534–546. RansohoV, R. M., and Estes, M. L. (1991). Astrocyte expression of major histocompatibility complex gene products in multiple sclerosis brain tissue obtained by stereotactic biopsy. Arch. Neurol. 48, 1244–6 RindXeisch, E. (1863). Histologisches Detail zur grauen Degeneration von Gehirn und Ru¨ckenmark. Arch. Pathol. Anat. Physiol. Klin. Med. (Virchow) 26, 474–483. Rodriguez, M., Scheithauer, B. W., Forbes, G., and Kelly, P. J. (1993) Oligodendrocyte injury is an early event in lesions of multiple sclerosis. Mayo Clin. Proc. 68, 627–636. Schlesinger, H. (1909). Zur Frage der akuten multiplen Sklerose und der encephalomyelitis disseminata im Kindesalter. Arb. Neurol. Inst. (Wien) 17, 410–432. Schwab, C., and McGeer, P. L. (2002). Complement activated C4d immunoreactive oligodendrocytes delineate small cortical plaques in multiple sclerosis. Exp. Neurol. 174, 81–88. Seil, F. J., Smith, M. E., Leiman, A. L., and Kelly, J. M. (1975). Myelination inhibiting neuroelectric blocking factors in experimental allergic encephalomyelitis. Science 187, 951–953. Selmaj, K., Raine, C. S., Cannella, B., and Brosnan, C. F. (1991). IdentiWcation of lymphotoxin and tumor necrosis factor in multiple sclerosis lesions. J. Clin. Invest. 87, 949–954. Simpson, J., Rezaie, P., Newcombe, J., Cuzner, M. L., Male, D., and Woodroofe, M. N. (2000b). Expression of the beta-chemokine receptors CCR2, CCR3 and CCR5 in multiple sclerosis central nervous system tissue. J. Neuroimmunol. 108, 192–200. Simpson, J. E., Newcombe, J., Cuzner, M. L., and Woodroofe, M. N. (1998). Expression of monocyte chemoattractant protein-1 and other beta-chemokines by resident glia and inXammatory cells in multiple sclerosis lesions. J. Neuroimmunol. 84, 238–49. Simpson, J. E., Newcombe, J., Cuzner, M. L., and Woodroofe, M. N. (2000a). Expression of the interferongamma-inducible chemokines IP-10 and Mig and their receptor, CXCR3, in multiple sclerosis lesions. Neuropathol. Appl. Neurobiol. 26, 133–42. Smith, K. J., and Lassmann, H. (2002). The role of nitric oxide in multiple sclerosis. Lancet Neurology 1, 232–241. Smith, K. J., Kapoor, R., and Felts, P. A. (1999). Demyelination: The role of reactive oxygen and nitrogen species. Brain Pathol. 9, 69–92. Smith, T., Groom, A., Zhu, B., and Turski, L. (2000). Autoimmune encephalomyelitis ameliorated by AMPA antagonists. Nature Med. 6, 62–66. Sobel, R. A., Mitchell, M. E., and Fondren, G. (1990). Intercellular adhesion molecule-1 (ICAM-1) in cellular immune reactions in the human central nervous system. Am. J. Pathol. 136, 1309–1316. Sorensen,T. L., Tani, M., Jensen, J., Pierce, V., Lucchinetti, C., Folcik, V. A., Qin, S., Rottman, J., Sellebjerg, F., Strieter, R. M., Frederiksen, J. L., and RansohoV, R. M. (1999). Expression of speciWc chemokines and chemokine receptors in the central nervous system of multiple sclerosis patients. J. Clin. Invest. 103, 807–15.

761

762

31. CELLULAR DAMAGE AND REPAIR IN MULTIPLE SCLEROSIS

Sorensen, T. L., Trebst, C., Kivisakk, P., Klaege, K. L., Majmudar, A., Ravid, R, Lassmann, H., Olsen, D. B., Strieter, R. M., RansohoV, R. M., and Sellebjerg, F. (2002). Multiple sclerosis: A study of CXCL10 and CXCR3 co-localization in the inXamed central nervous system. J. Neuroimmunol. 127, 59–68 Springer, T. A. (1994). TraYc signals for lymphocyte recirculation and leucocyte emigration: The multistep paradigm. Cell 76, 301–314. Stadelmann, C., Kerschensteiner, M., Misgeld, T, Bru¨ck, W., Hohlfeld, R., and Lassmann, H. (2002). BDNF and gp145trkB in multiple sclerosis brain lesions: Neuroprotective interactions between immune cells and neuronal cells? Brain 125, 75–85. Steiner, G. (1931). Regionale Verteilung der Entmarkungsherde in ihrer Bedeutung fu¨r die Pathogenese der multiplen Sklerose. In ‘‘Krankheitserreger und Gewebsbefund bei multipler Sklerose,’’ pp. 108–120. Springer, Berlin. Storch, M. K., Piddlesden, S., Haltia, M., Iivanainen, M., Morgan, P., and Lassmann, H. (1998). Multiple sclerosis: In situ evidence for antibody and complement mediated demyelination. Ann. Neurol. 43, 465–471. Tavolato, B. (1975). Immunoglobulin G distribution in multiple sclerosis brain. An immunoXuorescence study. J. Neurol. Sci. 24, 1–11. Trapp, B. D., Peterson, J., RansohoV, R. M., Rudick, R., Mork, S., and Bo, L. (1998). Axonal transection in the lesions of multiple sclerosis. N. Engl. J. Med. 338, 278–85. Traugott, U. (1987). Multiple sclerosis: Relevance of class I and class II MHC-expressing cells to lesion development. J. Neuroimmunol. 16, 283–302. Traugott, U., Reinherz, E. L., and Raine, C. S. (1983). Multiple sclerosis: Distribution of T cells, T cell subsets and Ia-positive macrophages in lesions of diVerent ages. J. Neuroimmunol. 4, 201–221. Trebst, C., Sorensen, T. L., Kivisakk, P., Cathcart, M. K., Hesselgesser, J., Horuk, R., Sellebjerg, F., Lassmann, H., and RansohoV, R. M. (2001). CCR1þ/CCR5þ mononuclear phagocytes accumulate in the central nervous system of patients with multiple sclerosis. Am. J. Pathol. 159, 1701–10. Ulvestad, E., Williams, K., Vedeler, C., Antel, J., Nyland, H., Mork, S., and Matre, R. (1994). Reactive microglia in multiple sclerosis lesions have an increased expression of receptors for the Fc part of IgG. J. Neurol. Sci. 121, 125–131. WakeWeld, A. J., More, L. J., DiVord, J., and McLaughlin, J. E. (1994). Immunohistochemical study of vascular injury in acute multiple sclerosis. J. Clin. Pathol. 47, 129–133. Washington, R., Burton, J., Todd, R. F., 3rd, Newman, W., Dragovic, L., and Dore-DuVy, P. (1994). Expression of immunologically relevant endothelial cell activation antigens on isolated central nervous system microvessels from patients with multiple sclerosis. Ann. Neurol. 35, 89–97. Werner, P., Pitt, P., and Raine, C. S. (2001). Multiple sclerosis: Altered glutamate homeostasis in lesions correlates with oligodendrocyte and axonal damage. Ann. Neurol. 50, 169–180. Willenborg, D. O., Staykova, M. A., and Cowden, W. B. (1999). Our shifting understanding of the role of nitric oxide in autoimmune encephalomyelitis: A review. J. Neuroimmunol. 100, 21–35. Windhagen, A., Newcombe, J., Dangond, F., Strand, C., Woodroofe, M. N., Cuzner, M. L., and HaXer, D. A. (1995). Expression of costimulatory molecules B7–1 (CD80), B7–2 (CD86), and interleukin 12 cytokine in multiple sclerosis lesions. J. Exp. Med. 182, 1985–96. Wolswijk, G. (2000) Oligodendrocyte survival, loss and birth in lesions of chronic-stage multiple sclerosis. Brain 123, 105–115. Wolswijk, G. (2002). Oligodendrocyte precursor cells in the demyelinated multiple sclerosis spinal cord. Brain 125, 338–349. Woodroofe, M. N., Bellamy, A. S., Feldmann, M., Davison, A. N., and Cuzner, M. L. (1986). Immunocytochemical characterisation of the immune reaction in the central nervous system in multiple sclerosis. Possible role for microglia in lesion growth. J. Neurol. Sci. 74, 135–52. Woodroofe, M. N., and Cuzner, M. L. (1993). Cytokine mRNA expression in inXammatory multiple sclerosis lesions: Detection by nonradioactive in situ hybridization. Cytokine 5, 583–588. Wucherpfennig, K. W., Newcombe, J., Li, H., Keddy, C., Cuzner, M. L., and HaXer, D. A. (1992a). Gamma delta T-cell receptor repertoire in acute multiple sclerosis lesions. Proc. Natl. Acad. Sci. (USA) 89, 4588–4592. Wucherpfennig, K. W., Newcombe, J., Li, H., Keddy, C., Cuzner, M. L., and HaXer, D. A. (1992b). T-cell receptor V alpha-V beta repertoire and cytokine gene expression in active multiple sclerosis lesions. J. Exp. Med. 175, 993–1002.

C H A P T E R

32 MRI Visualization of Multiple Sclerosis Alberto Cifelli, D. L. Arnold, and P. M. Matthews

ed1

INTRODUCTION Magnetic resonance imaging (MRI) and related techniques make possible the noninvasive visualization of structural, biochemical, and functional changes in the brain (Gadian, 1995; Hashemi and Bradley, 1997; Jezzard et al., 2001). This visualization of pathological changes in vivo has had a profound impact on the clinical management of patients with multiple sclerosis (MS) and on the understanding of the disease and its treatment (Miller et al., 1997). MRI has substantially changed the way the diagnosis of MS is made (Arnold and Matthews, 2002). Diagnosis with high conWdence now is possible very early in the course of disease (McDonald et al., 2001). More accurate information regarding prognosis is available, e.g., for patients presenting with clinically isolated syndromes, MRI data help to estimate the risk of progression to MS and for patients early in the course of MS it provides information relevant to predicting disease severity (Brex et al., 2002). MRI and related techniques also have been able to shed considerable light on the natural history of the pathology of MS (Miller et al., 1998) and its responses to treatments (Arnold and Matthews, 2002; Rovaris and Filippi, 1999). This has improved the eYciency of trial designs, particularly for the earlier stages of evaluation of new drugs. MRI provides a range of techniques for visualizing the central nervous system (CNS) in patients with MS, each of which is sensitive to diVerent aspects of pathological change (Arnold and Matthews, 2002). The application of these MRI techniques to MS thus can be viewed in a similar way to the application of diVerent staining techniques to post-mortem material by the pathologist. Each MR-based technique allows visualization of particular characteristics of the pathology (Barkhof and van Walderveen, 1999). By integration of data from multiple MR techniques, a usefully comprehensive view of the disease emerges (Rovaris and Filippi, 2000b). With greater understanding of the genesis of the pathological changes, neurobiologically driven hypotheses regarding disease evolution or the eVects of interventions now can be tested by observations during life. The relationship between changes in MS and in animal models of CNS inXammatory disease also can be evaluated directly, enhancing the information available from the models (Hawkins et al., 1990). This chapter outlines the basic principles of MRI and related techniques. MRI techniques deWne pathology on the basis of changes in biophysical parameters of tissue. With an understanding of these biophysical parameters, it can be appreciated that they have variably speciWc relations with pathological features as deWned using traditional methods. The chapter then reviews several contributions of MRI and related techniques to the understanding of MS and to pathology. MR and related techniques have been important particularly in emphasizing that much pathological change is subclinical, the diVuse nature of pathology, the prominence of gray matter involvement, and the role of neurodegeneration, even in the early stages of disease evolution. Together these observations provide a

Myelin Biology and Disorders, Volume 2

763

Copyright 2004, Elsevier (USA). All rights reserved.

764

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

remarkably comprehensive view of key features essential for understanding the way MS progresses.

BASIC PRINCIPLES OF MRI Conventional Magnetic Resonance Imaging Techniques MRI and related techniques are based on the weak interaction between the proton nuclear magnetic dipole and a strong, homogeneous applied magnetic Weld. The conventional MRI techniques used to study MS patients produce images that reXect primarily the physicochemical state of protons in brain water. Contrast in such images is derived from tissuespeciWc diVerences in (1) water content, (2) spin-spin relaxation time (i.e., the time constant for the decay of magnetization in the plane perpendicular to the magnetic Weld, referred to as T2), and (3) spin-lattice relaxation time (i.e., the time constant for the recovery of magnetization along the axis of the magnetic Weld, referred to as T1) (for reviews of MRI theory and applications see, for example, Gadian, 1995; Hashemi and Bradley, 1997; Jezzard et al., 2001). Conventional MRI techniques include (1) T2-weighted imaging, (2) proton-density-weighted imaging, (3) Xuid-attenuated inversion-recovery (FLAIR) imaging, (4) standard T1-weighted imaging, and (5) gadolinium-enhanced T1-weighted imaging (Fig. 32.1). T2-Weighted Imaging MR images are T2-weighted by allowing more time for T2 relaxation to occur during a relatively-long echo time (i.e., the time between radio-frequency excitation and subsequent acquisition of the signal from a ‘‘spin-echo’’, referred to as TE). Signals from water protons located in tissues associated with longer T2 values decay less during a long TE; because of this, such tissues appear hyperintense on T2-weighted images relative to tissues that have shorter T2 relaxation times. The T2 relaxation rate (the reciprocal of the relaxation time) is determined both by local static magnetic interactions and by interactions that Xuctuate at the resonance (Larmor) frequency. The resonance frequency is determined by an unchanging physical property of the nuclear magnetic moment of the proton (the gyromagnetic ratio) and the strength of

A

B

C

D

FIGURE 32.1 Use of diVerent MRI pulse sequences allows contrast to be generated on the basis of diVerent biophysical parameters. These have diVerent sensitivities to histopathological change. (A) A T1-weighted image deWnes areas of substantial demyelination, axonal loss, and matrix destruction as hypointense volumes. (B) A T2-weighted image more sensitively identiWes pathology associated with multiple sclerosis, but shows T2-hyperintense signal change nonspeciWcally for edema, demyelination, and gliosis. (C) A proton density-weighted image has a similar relative lack of pathological speciWcity, but can be useful particularly for identifying the extent of lesions in the periventricular regions because CSF is hypointense, in contrast to the hyperintensity of MS lesions. (D) A magnetization transfer ratio (MTR) image shows decreased intensity in areas of demyelination and axonal loss. This technique thus deWnes lesion distribution similarly to T2-weighted imaging, although the latter is more sensitive to edema.

BASIC PRINCIPLES OF MRI

the applied magnetic Weld. At 1.5T, this is in the radiofrequency range at 62 MHz. Because the T2 relaxation rate depends on both static and Larmor frequency interactions, it is very sensitive to even small changes in the nature and extent of the association between water and local macromolecules in tissue. T2 therefore is highly sensitive to changes in tissue water content and to the nature and concentration of tissue macromolecules. A practical problem with T2-weighted brain images for the quantitative characterization of MS pathology is that both CSF and focal lesions are hyperintense, so the precise extent of periventricular lesions often cannot be deWned clearly. The FLAIR sequence is a T2-weighted imaging sequence modiWed to ‘‘null’’ signal from the CSF compartment: it is often used with (or in preference to) a simple T2-weighting. T1-Weighted Imaging T1-weighted images are produced by shortening the time between repetitions of a pulse sequence (TR), allowing less time for water to regain its equilibrium magnetization. Water protons in tissues with a relatively short T1 relaxation time recover more quickly and generate more signal with a shorter TR than do those in tissues with a longer T1. Protons in bulk water (e.g., in CSF or in tissue that is associated with edema or a loss of structural integrity) have a long T1 and appear hypointense on T1-weighted sequences. The T1 relaxation rate is determined by interactions only at the resonance frequency, giving T1-weighted imaging changes less sensitivity but greater speciWcity to pathological phenomena. T1-weighted images are less sensitive to variations in either the matrix water- or myelin content than are T2-weighted images, for example. Proton Density-Weighted Imaging Contrast in conventional MRI arises from diVerences in water content, as well as intrinsic relaxation characteristics between diVerent tissues or between normal and pathological tissue. Signal intensity in images acquired with a very long inter-pulse interval (TR) and a very short delay before data acquisition (TE) is related directly to the concentration of tissue water protons. This type of image would be a true proton density (PD) image. In practice, so-called PD-weighted images are obtained with relatively long TR and relatively short TE, such that CSF is hypointense and lesions are hyperintense. Both PD- and T2-weighted MRI provide qualitatively similar pathological information in MS. In practice, PD-weighted images (and FLAIR images) are more useful for identiWcation of lesions adjacent to the ventricles, because the CSF in their proximity is hypointense (i.e., darker than lesions) (Fig. 32.1). Gadolinium-Enhanced T1-Weighted Imaging As reviewed by Rovaris and Filippi (Rovaris and Filippi, 2000a), the sensitivity of lesion detection in T1-weighted imaging can be increased selectively with the injection of gadolinium-diethylenetriaminepentacetate (Gd-DTPA), a chelated form of gadolinium (Gd) (Fig. 32.2). Gd is one of the lanthanide elements and has strong paramagnetic properties, so that in its nontoxic chelate with DTPA it changes the local rate of water MR relaxation. Gd-DTPA facilitates both longitudinal (T1) and transverse (T2) water proton relaxation in proportion to its concentration in local tissue water (Tofts and Kermode, 1991). As water proton T1 is 5-10 times longer than T2 in most tissues, the relaxation time shortening eVects of Gd-DTPA are more pronounced on T1 than on T2. Normally, Gd-DTPA in serum does not cross the blood-brain barrier (BBB) and would enhance only signal in blood vessels. However, when BBB integrity is compromised, Gd-DTPA can enter the CNS and, with shortening of T1, increases signal intensity on T1-weighted images selectively at the sites of BBB breakdown.

‘‘Nonconventional’’ MR Imaging Techniques Several ‘‘nonconventional’’ MR imaging measures can be deWned arbitrarily on the basis of their lack of integration into routine clinical practice at present. Each generates image contrast based on diVerent biophysical parameters. In combination with conventional

765

766

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

A

1

B

3

FIGURE 32.2 Gd-enhancement can identify focal breakdown of the blood brain barrier (BBB). (A) This T2-weighted MRI scan from a patient with multiple sclerosis demonstrates multiple T2-hyperintense lesions. Acute and chronic lesions cannot be distinguished on the basis of the T2-weighted MRI signal alone. (B) This gadolinium-enhanced T1-weighted image shows focal hyperintense signal (arrowhead) in one of the lesions identiWed as T2hyperintense in (A). This lesion is an acute or subacute lesion, distinguishing it from the other T2-hyperintense lesions as shown in (A). Together, images in (A) and (B) provide evidence for dissemination of lesions in both space and time.

measures, they can enhance the speciWcity and allow more precise quantiWcation of pathological changes. Magnetization Transfer Imaging Magnetization transfer (MT) images are created by applying a selective radiofrequency irradiation that saturates the signal from protons in macromolecules not normally visualized by conventional MRI (Grossman, 1999; McGowan, 1999) (Fig. 32.3). When this saturation is subsequently transferred to the bulk water pool that is normally visualized by MRI, the bulk water MR signal decreases by an amount that depends on (1) the density of the macromolecules at a given location and (2) the nature of their interactions with bulk water. The extent of the eVect can be quantiWed conveniently as the magnetization transfer ratio (MTR), which is deWned as the percent decrease in signal between the image with the saturating irradiation of macromolecule protons and that without their saturation. A greater magnetization transfer between macromolecules and associated water (giving a higher MTR) implies a greater or stronger interaction between water and macromolecular protons. Loss of macromolecular density or structure (as occurs with demyelination) will reduce the MTR. As conventionally applied, the MTR does not speciWcally probe any particular class of macromolecules. However, because the relative abundance of myelin proteins in the brain is so high, changes in myelin content dominate MTR changes in MS. Even aside from technical factors related to variations in the eYciency of macromolecule saturation, MTR is not completely speciWc for changes in magnetization transfer, as it is also inXuenced by T1 changes associated with severe edema. This limitation may be circumvented by quantifying magnetization transfer, rather than relying simply on the ratio. Such quantiWcation is nontrivial, but is being pursued in a number of laboratories since it may improve the pathological speciWcity of the technique (Brass et al., 2002a, 200b; Dalton et al., 2002; Levesque et al., 2002; Ramani et al., 2001; Ropele et al., 2002; Santos et al., 2002).

767

BASIC PRINCIPLES OF MRI

A

B

C

FIGURE 32.3 Magnetization transfer imaging generates contrast on the basis of relaxation interactions between tissue water and macromolecules. (A) Image acquired without oV-resonance irradiation to saturate macromolecular protons. (B) An image from the same level with oV-resonance irradiation to fully saturate the macromolecule protons. The intensity is diminished in (B) relative to (A) because of transfer of saturation from macromolecule protons to tissue water. (C) A magnetization transfer ratio (MTR) image in which the contrast expresses the signal in (B) relative to (A). The white matter is bright because of the strong cross-relaxation eVects of myelin proteins on tissue water.

Diffusion-Weighted Imaging Application of high magnetic Weld gradients during the imaging sequence can make it sensitive to the diVusion of water (Cercignani and HorsWeld, 2001; Cercignani et al., 1999). DiVusion properties of water in tissue can be expressed without regard to direction, as an apparent diVusion coeYcient (ADC), or with respect to the relative anisotropy, either as a simple measure, such as the fractional anisotropy (FA), or as a full diVusion tensor, deWning the direction of anisotropy explicitly in three-dimensional space. Water diVusion in healthy white matter is highly anisotropic because of the oriented axons and myelin, along the axes of which water diVusion distances are greatest. Although the relative contributions of water in myelin and axons to diVusion anisotropy measurements are not well deWned, it is clear that factors associated with demyelination and an increase in the extracellular space markedly decrease diVusion anisotropy. Magnetic Resonance Spectroscopy and Spectroscopic Imaging Proton magnetic resonance spectroscopy (MRS) is fundamentally diVerent from the waterproton-based MRI techniques described earlier in that it records signals that arise from protons in CNS metabolites (Rudkin and Arnold, 1999). Because the concentration of such tissue metabolites is approximately one-thousandth that of tissue water, the signalto-noise ratio and image resolution of these metabolite-based images is much lower than in water-based images. Nevertheless, the resulting metabolite images can provide chemicopathological speciWcity that is not possible with conventional water-based images. As shown in Figure 32.4, the water-suppressed, localized proton MRS spectrum of the normal human brain that is recorded at relatively long echo times (i.e., TE of 136 or 272 ms) reveals three major resonance peaks [the locations of which are expressed as the diVerence in parts per million (ppm) between the resonance frequency of the compound of interest and that of a standard compound (i.e., tetramethylsilane)]. The three major peaks are ascribed to choline (Cho), creatine (Cr), and N-acetylaspartate (NAA). The Cho peak at 3.2 ppm arises from tetramethyl amines, which are mainly found in cholinecontaining phospholipids participating in membrane synthesis and degradation. The Cr peak, which resonates at 3.0 ppm, arises from both creatine and phosphocreatine (Cr). The

768

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

Normal

MS NA

[1] NWM NA/Cr = 3.83 Cho/Cr = 1.30

[1] NAWM NA/Cr = 2.86 NA Cho/Cr = 1.84 Cho

Cho

Cr

[2] NWM NA/Cr = 3.04 Cho/Cr = 1.14

Cr

NA

[2] Lesion NA/Cr = 1.91 Cho/Cr = 1.12 NA

Cho

Cr

Cho

Cr

FIGURE 32.4 Proton-density-weighted magnetic resonance images through the centrum semiovale and the results of proton MRSI in a patient with multiple sclerosis and in a normal control subject. The superimposed grid in each image represents individual 1H-MRSI voxels, and the large white box represents the entire proton MRSI volume of interest for that individual. The smaller, numbered boxes represent voxels of normal-appearing white matter (NAWM) and lesional brain tissue in the patient and normal white matter (NWM) in the normal control subject. The 1H-MRSI spectra from within each of these voxels are shown to the right of each image. The areas under the NA and Cho peaks (normalized to Cr) are shown above each spectrum. The spectra have been scaled so that the Cr peak in each of them has the same height.

N-acetylaspartate resonance at 2.0 ppm originates from N-acetyl groups (NA), which, in the brain, are found primarily in the neuronally localized compound N-acetylaspartate (NAA). A fourth peak, which is attributed to the methyl resonance of lactate (LA) and resonates at 1.3 ppm, is normally only barely visible above the baseline noise; in certain pathological conditions, it may increase many-fold and a methyl resonance from lipids may also be detected in the same region of the spectrum. The resonance intensity that is ascribed to NAA is, arguably, the most important proton MRS signal in the characterization of MS pathology, because NAA is localized within neurons and neuronal processes such as axons and dendrites (Bitsch et al., 1999; Bjartmar et al., 2000; MoVett et al., 1991; Simmons et al., 1991; Wujek et al., 2002). MRS can acquire biochemical data from a single volume of tissue (single-voxel spectroscopy) or from multiple voxels simultaneously, thereby generating an image based on the distribution of a selected proton resonance (magnetic resonance spectroscopic imaging or MRSI) (Salibi and Brown, 1998). Functional Magnetic Resonance Imaging FMRI is a magnetic resonance technique based on monitoring blood oxygen level dependent (BOLD) contrast (Jezzard et al., 2001). Neuronal activity is associated with an increased local synaptic activity. The increased metabolism is associated with an increase of blood Xow and oxygen consumption. The relative increment in blood Xow is greater than the increase in oxygen consumption. This leads to a boost of the oxyhaemoglobin/ deoxyhaemoglobin ratio. Contrast changes arise because the eVects of (diamagnetic) oxyhaemoglobin and (paramagnetic) deoxyhaemoglobin on local water relaxation times are diVerent. Increased oxyhaemoglobin levels lead to a relatively longer T2 relaxation time for water in and around blood vessels, resulting in an enhancement of the MR signal (0.5–1.5% at 1.5 T). This signal change can be visualized by subtracting images acquired during a period of greater neural activation (e.g., during a movement) from images acquired during a relative resting state (Fig. 32.5). FMRI measures neural activation only indirectly by sampling the relative changes in microvascular (mainly venular) blood oxygenation. Albeit a young technique (Ogawa et al., 1993), fMRI already has been found application in a vast array of research and clinical settings.

MRI VISUALIZATION OF MULTIFOCAL LESIONS FOR EARLY DIAGNOSIS OF MS

769

MRI VISUALIZATION OF MULTIFOCAL LESIONS FOR EARLY DIAGNOSIS OF MS MRI has substantially increased the sensitivity for detection of pathological changes associated with MS and allows the diagnosis of MS to be made much earlier than in the past, when it was based to a greater extent on clinical manifestations of the disease (McDonald et al., 2001). Although CT changes can be seen with MS (Hershey et al., 1979), most lesions are not highlighted by contrast and abnormalities on unenhanced scans are typically apparent only with very large acute lesions or in the later stages of the disease. Even then, CT abnormalities have limited speciWcity. In contrast, characteristic patterns of lesions are found on MRI in approximately 95% of MS patients (Paty et al., 1988) and in the vast majority of patients with clinically isolated syndromes (initial attacks) suggestive of MS, who will go on to develop additional episodes or MRI lesions that would be suYcient to make a diagnosis (Brex et al., 2002). A number of studies have deWned criteria with diVerent sensitivity and speciWcity for interpretation of MRI in the context of suspected MS (Barkhof et al., 1997; Fazekas et al., 1988; Gean-Marton et al., 1991; Horowitz et al., 1989; Paty et al., 1988; Tas et al., 1995; Tintore et al., 2000) (Tab. 32.1). Further speciWcity arises from gadolinium enhancement of lesions deWning acute to subacute blood brain barrier (BBB) breakdown due to inXammation. The currently recommended criteria for use of MRI in the diagnosis of MS (McDonald et al., 2001) are shown in Table 32.2. They are relatively stringent in order to increase the speciWcity of the MRI Wndings (Barkhof et al., 1997; Tintore et al., 2000). Although other pathologies also can give rise to multiple T2 hyperintense lesions, these have rather diVerent characteristics, often allowing them to be discriminated on the basis of imaging alone (Arnold and Matthews, 2002; Barkhof and Scheltens, 2002). Additional information can come from spinal cord imaging (Simon, 2000). However, spinal cord MRI is much less sensitive to the pathological changes of MS than is brain (Bergers et al., 2002).

FIGURE 32.5 Functional magnetic resonance imaging (fMRI) can deWne altered patterns of brain activation associated with sensory, motor, or cognitive tasks in patients with multiple sclerosis. (A) The pattern of activation in the superior aspect of the brain from a healthy control during simple hand Xexionextension movements is shown. The contralateral primary sensorimotor area (large arrow), the midline supplementary motor area and a small region in the ipsilateral premotor cortex show activation. (B) In a patient with multiple sclerosis, activation in the contralateral primary sensorimotor area may be diminished with signiWcantly increased activation in the ipsilateral motor cortex (black borders). This relative recruitment of ipsilateral motor cortex for a simple action is potentially adaptive, limiting disability in the patients.

770

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

TABLE 32.1

Interpretation of MRI in the Diagnosis of Multiple Sclerosis

Reference

DeWnition

Study design

Sensitivity (%)

SpeciWcity (%)

Fazekas et al., 1988


3 lesions with two of 1. An infratentorial lesion 2. A periventricular lesion 3. A lesion >6 mm

Retrospective

88

100

Paty et al., 1988


four lesions, or three lesions of which one is periventricular

Prospective (following from Wrst presentation)

94

57

Tas et al., 1995


1 Gda-enhancing lesion and
one nonenhancing lesion

Prospective

59

80

At least 1 Gd-enhancing lesion or 9 T2-hyperintense lesions including 1. At least one juxtacortical lesion 2. At least three periventricular lesions 3. At least one infratentorial lesion

Prospective

82

78

b,c

Barkhof et al., 1997

a Gd ¼ gadolinium-DTPA. b According to Tintore’ et al (2000), if three out of four criteria are fulWlled, the highest accuracy and best compromise between sensitivity and speciWcity are achieved. c

McDonald et al. (2001) allow the substitution of one spinal cord lesion for one brain lesion. Adapted from Barkhof et al., 1997.

TABLE 32.2 Clinical (attacks)

McDonald (McDonald et al., 2001) Criteria for Diagnosis of Multiple Sclerosis

Objective lesions

Additional requirements to make diagnosis

Two or more

Two or more

Two or more

One

None: Clinical evidence will suYce (additional evidence is desirable but must be consistent with MS) Dissemination in space by MRI or CSFþa and two or more MRI lesions consistent with MS or further clinical attack involving diVerent site

One

Two or more

Dissemination in time by MRI or second clinical attack

One (monosymptomatic)

One

Dissemination in space by MRI or CSFþ and two or more MRI lesions consistent with MS And dissemination in time by MRI or second clinical attack

Zero (progression from onset)

One

CSFþ And dissemination in space with MRI evidence of nine or more brain lesions or two or more cord lesions or four to eight brain and one cord lesion Or positive VEPb with four to eight MRI lesions Or positive VEP with less than four brain lesions plus one cord lesion And dissemination in time by MRI or continued progression for one year

a CSFþ ¼ Presence in the cerebrospinal Xuid of oligocloclonal bands diVerent from any such bands in serum or of a raised IgG index. b

VEP ¼ Visual evoked potentials.

When patients present with their Wrst symptoms suggestive of MS, it is important to be able to predict the risk of developing another attack (which would qualify the patient for a diagnosis of MS) and also the risk of signiWcant future disability. MRI data cannot yet be used to predict disability reliably. However, the presence of asymptomatic lesions in the brain at the time of initial presentation is a strong predictor that the patient will eventually develop clinically deWnite MS (Brex et al., 2002). An increased proportion of T1 hypointense changes within chronic lesions is associated with more severe disease, although speciWc criteria for this have not been evaluated yet (Fazekas et al., 2000).

MRI VISUALIZATION OF THE PROGRESSION OF MS LESIONS While progression of lesions in MS has long been inferred on the basis of variations in pathology at post-mortem, longitudinal MRI studies have provided a direct view of the dynamics of evolution of individual lesions (Grossman et al., 1988; Kermode et al., 1990a; Thompson et al., 1992). A basic model has developed in which the earlier stages of evolution are associated with blood brain barrier (BBB) breakdown, followed by later inXammatory changes leading to demyelination, axonal loss, and gliosis. With more severe

771

MRI VISUALIZATION OF THE PROGRESSION OF MS LESIONS

TABLE 32.3

Pathological Correlations in MS for MR-based Imaging Changes

Technique

Findings

Conventional T2-weighted imaging

Hyperintensity

Conventional T1-weighted images

Acute hypointensity Chronic hypointensity a

Pathological correlates InXammation, oedema, demyelination, gliosis, remyelination, axonal loss Oedema Demyelination, axonal loss, and matrix destruction

Gd -enhancement

Blood brain barrier disruption

Cerebral (or spinal cord) atrophy

Demyelination, axonal loss and gliosis (relative contributions uncertain)

Conventional PDb -weighted images

Hyperintensity

As for T2-weighted imaging, but better contrast between lesions and CSFc

Magnetic resonance spectroscopy (MRS) or spectoscopic imaging (MRSI)

N-acetyl-aspartate (NAA) decrease

Atrophy, metabolic dysfunction, or loss of axons or neurons Early myelin damage Myelin breakdown and inXammatory cell inWltration Acute inXammation Glial changes InXammatory or glial response

Increased macromolecule (lipid) resonances Increased choline resonances Increased lactate resonance Decreased creatine resonances Increased myo-inositol resonance Magnetisation transfer (MT) imaging

Reduced magnetisation transfer ratio (MTR)

Demyelination and axonal loss

DiVusion weighted imaging (DWI)

Increased water diVusivity and decreased diVusion anisotropy

Oedema, demyelination, and axonal loss

Functional MRI

Altered patterns of cerebral activation during sensory, motor, or cognitive tasks

Systems-level, potentially adaptive functional reorganization

d

a Gd ¼ Gadolinium-DTPA. b

PD ¼ Proton density.

c CSF ¼ Cerebrospinal Xuid. d

No pathological evidence from human studies available.

inXammation, substantial matrix destruction and local axonal damage can occur. However, as will be discussed subsequently, important reWnements to this model continue to be made.

Visualization of Blood Brain Barrier Breakdown with Contrast-Enhanced T1-Weighted MRI BBB disruption is an early event in the pathogenesis of an MS lesion. The BBB refers to several mechanisms that restrict free exchange of non-lipid soluble molecules between blood and the CSF space. The BBB includes both active (energy-requiring) processes (e.g., amino-acid transporters) and passive mechanisms (e.g., endothelial tight junctions) that regulate the physiological environment in the central nervous system (CNS). Anatomical barriers contributing to the BBB include the capillary endothelium, with its tight junctions, adjacent glia, and components of the extracellular matrix. Intravenous injection of an exogenous contrast agent that is normally excluded from the CNS space, but can enter with damage to the integrity of the BBB, allows MRI to visualize breakdown of the BBB (Fig. 32.2). The contrast agent in most common use is Gd-DTPA. Local Gd-enhancement is associated with active lesions in MS (Table 32.3) Movement of Gd-DTPA across the BBB can be shown to be driven by altered endothelial pinocytotic transport in animal models, as well as possibly by loss of the integrity of tight junctions between endothelial cells, as suggested by a study of post-mortem MS brains (Plumb et al., 2002). Rare observations have provided additional, direct histopathological correlations of Gd-enhancement in MS as well. These were Wrst described in a study of a secondary progressive (SP) MS patient scanned Wrst at 4 weeks and then at 10 days prior to death (Katz et al., 1990, 1993). Enhancing lesions were extensively demyelinated and contained abundant perivascular cuVs with lymphocytes, macrophages, and plasma cells. Within

772

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

areas of enhancement were multiple small areas of demyelination centered around a perivascular cuV and bordered by macrophages Wlled with undigested myelin debris. Active lesions are highly likely to show Gd-enhancement. In a study based on both biopsy and autopsy material (Nesbit et al., 1991), all of the histologically active lesions showed enhancement whereas none of the inactive lesions did. Not surprisingly, Gdenhancement is associated with clinical relapses (Miller et al., 1988). A majority of patients in relapse with focal neurological symptoms will show Gd-enhancement in lesions anatomically localized to account for new symptoms (Miller et al., 1988). However, most enhancing lesions are asymptomatic. This likely reXects the variable severity of the associated conduction block or axonal injury, as well as the fact that many areas of brain do not eloquently express deWcits from smaller lesions. Gd-enhancement of lesions is 5 to 10 times higher than the number of clinical exacerbations. Activity can be high even in the early RR phase of the disease (Harris et al., 1991). Most lesions enhance for less than 1 month. Enhancement for greater than 6 months is rare (McFarland et al., 1992). Gd-enhancement thus is a useful marker of acute to subacute inXammatory activity. Dynamic contrast studies follow the time course of signal changes with a rapidly acquired series of images after Gd-DTPA injection. In a dynamic study of RR MS (Kermode et al., 1990b), images acquired 2 to 4 minutes after injection of Gd-DTPA showed a variable extent of enhancement, which was typically smaller than the corresponding volumes on unenhanced scans and frequently had the appearance of conXuent rings. This ring enhancement likely reXects acute inXammation at the border of chronic active demyelinating lesions. By 16 to 20 minutes after injection, most lesions enhanced homogeneously. Over several hours after the injection the initially hypointense center of many of the ring lesions then became brighter than the periphery. By 5 hours postinjection, the enhancing volumes enlarged to the full size of the corresponding T2 hyperintense lesions. The mean time to the peak enhancement of lesions was just under 30 minutes, but newer lesions tended to show peak enhancement earlier than older ones, suggesting that time course may provide a marker of the timing of inXammation. Dynamic contrast studies are not performed routinely, however. As a practical guide, the optimal timing to maximize the sensitivity to enhancement of active lesions after contrast injection is between 10 and 30 minutes with conventional techniques (Kermode et al., 1990b; Silver et al., 1997). Most clinical Gd-enhanced scans are begun 5 minutes after the injection, in the interest of time eYciency. The extent of contrast enhancement for a speciWc lesion is determined by the dose of contrast agent injected, as well as the time delay after administration before imaging. Use of triple dose gadolinium can increase the frequency of detection of enhancing lesions by 66 to 75% (Filippi et al., 1996b; Silver et al., 1997). The primary application of Gd-enhancement is for assessment of lesion activity. In conjunction with conventional T2-weighted or PD-weighted imaging, it can increase overall lesion detection, but the beneWts are modest. The gain in sensitivity is greatest for lesions at cortical-subcortical junctions (Miller et al., 1993). Gd-enhancement is highly variable between patients. However, in a fairly large group of relapsing-remitting patients studied monthly three consecutive times 78% had evidence of BBB breakdown on at least one MRI (Stone et al., 1995). There is a rather cyclical trend to the variation in frequency of Gd-enhancement in individual patients (McFarland et al., 1992). Gd-enhancing lesion number or volume is predictive of relapse frequency in RR MS (Smith et al., 1993). The frequency of Gd-enhancing lesions also may be predictive of subsequent development of disability (Khoury et al., 1994; Koudriavtseva et al., 1997; LosseV et al., 1996a; Smith et al., 1993). Although there is a strong relationship between histopathologically deWned active lesions and Gd-enhancement, it is not clear that all lesions visible by T2-weighted MRI necessarily evolve through an early, Gd-enhancing phase (Bruck et al., 1997). There are discrepancies between patterns of Gd-enhancement and T2-hyperintense lesions, for example. The rate of Gd-enhancing lesion appearance may decrease in later stages of the disease, without an associated decrease in T2-hyperintense lesion accumulation rates (Filippi et al., 1997). The rate of enhancing lesion formation in SP MS patients can be

MRI VISUALIZATION OF THE PROGRESSION OF MS LESIONS

signiWcantly lower than in RR patients, despite comparable increments in the rates of increase of the unenhanced lesion load and in disability. A possible explanation for this was provided by Lee et al. (1999), who demonstrated that the spatial distributions of T2-hyperintense and Gd-enhancing lesions were diVerent across cerebral white matter in a population of patients with established MS. There was a much higher probability for T2-hyperintense lesions to be periventricular than for Gd-enhancing lesions, which tended to be more peripheral in the white matter. This implies that the periventricularly localized component of the T2 hyperintense lesion burden less frequently involves an early stage of BBB breakdown that could be detected using Gd-enhancement. T2-signal changes in the periventricular area may, at least in part, reXect gliosis secondary to the Wallerian degeneration of descending Wbers transected in more peripheral lesions. Proton Density and T2-Weighted MRI Allows Visualization of Inflammation, Demyelination, and Gliosis T2- and PD-weighted images allow discrimination of MS lesions from the surrounding normal appearing white matter because of lesional changes in the water proton T2 relaxation time and content. These techniques are very sensitive to pathology; even direct inspection of unWxed brains at post-mortem examination does not reveal all the lesions seen on T2-weighted images of the same specimens (Newcombe et al., 1991), for example. The sensitivity of the proton density or T2-weighted image for MS lesions is important for diagnosis, but as a routine follow-up implement it is pathologically nonspeciWc. Histopathological studies of biopsy material show that areas of hyperintense signal on T2-weighted MRI deWne the whole spectrum of MS lesion evolution from pathologically early active, through late active to chronic inactive and including remyelinating lesions (Bruck et al., 1997; van Waesberghe et al., 1999). While both acute and chronic lesions show contrast changes with PD- or T2-weighted imaging, the underlying pathological correlations may vary (Table 32.3). Consider the pathological changes potentially associated with increased signal on PD-weighted images, for example. Acute lesions show increased water content with breakdown of the BBB and associated transudation of soluble serum proteins. Chronic lesions have increased water content with reduced myelin lipids and changes in cellularity with a greater glial component. These diVerent mechanisms and the diVerences in the extent of associated increases in water content lead to subtle, time-dependent changes in lesion appearance. Lesions that are less than approximately 30 days old, for example, have a pattern of central hyperintensity in the PD-weighted image. Around 2 to 3 months of age, approximately 60% of lesions acquire a ring-like appearance with a darker central core and brighter periphery. This may correspond with more extensive demyelination and other chronic changes centrally, and with active inXammation at the lesion rim (Guttmann et al., 1995). Similarly, quantitatively diVerent changes in T2 relaxation might deWne diVerences in pathology more speciWcally. Early active lesions, for example, have a border of decreased intensity contrasting with a brighter center (Bruck et al., 1997). More recent studies with direct measurement of T2 relaxation times eventually might allow this speciWcity to be applied to discrimination of diVerent lesions types (Santos et al., 2002), although at present the spatial resolution for these T2 relaxometric techniques is substantially lower than for the structural images, limiting their practical application. Serially acquired T2-weighted images emphasize that the pathology of MS is evolving continuously (Fig. 32.6); clinical relapses identify only a minority of new lesions. In general, there may be as much as a 10-fold greater activity deWned by T2-weighted MRI than from clinical course (Miller et al., 1997). Asymptomatic lesions may either be in noneloquent parts of the CNS, represent less destructive pathology, or develop slowly enough that functional adaptation may occur to prevent their expression.

T1-Weighted MRI Allows Visualization of Matrix Destruction in Lesions As noted earlier, T2- or PD-weighted hyperintense lesions are pathologically nonspeciWc. An attractive approach to improving speciWcity is to combine deWnitions of pathology

773

774

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

0

2 months

4 months

8 months

FIGURE 32.6 MRI demonstrates dynamic changes in T2 hyperintense lesion size and distribution. These serially acquired T2-weighted MRI scans from the same level of the brain of a patient with multiple sclerosis demonstrate both lesion growth (large arrow) and lesion shrinkage (small arrow) over time.

based on T2-weighted imaging with a diVerent range of pathological sensitivity acquired using diVerent pulse sequences. One of the most practical eVorts toward this integration is the study of T1 hypointensity, arising from increases in water T1 or an increased proportion of free water in lesions. Uhlenbrock and Sehlen (1989) described focal T1-hypointensities or ‘‘black holes’’ in some T2-hyperintense lesions in brains of patients with MS. They postulated that they represented regions of axonal loss and gliosis (Table 32.3). This was conWrmed for chronic lesions by post-mortem studies (Bruck et al., 1997; van Walderveen et al., 1998; van Waesberghe et al., 1999). T1 hypointensity in acute lesions may be less pathologically speciWc as it is signiWcantly inXuenced by the acute edema. Axonal loss is the major determinant of chronic disability (Matthews et al., 1998). Not surprisingly, therefore, increases in the volume of chronic T1-hypointensity are related to progression of disability. In a pioneering study (Truyen et al., 1996), T1 lesion load was found to be more strongly correlated with disability than the less pathologically speciWc T2 hyperintense lesion load. Objective deWnition of focal T1 lesions using voxel-by-voxel T1-mapping has conWrmed that the volume of the most abnormally prolonged T1 is correlated strongly with disability (Parry et al., 2002a). These observations have led to increasingly widespread use of the chronic T1-hypointense lesion volume as a measure of the progression of pathology relevant to disability.

Visualization of Focal Demyelination Myelin is a primary target of tissue damage in MS. Understanding the dynamics of myelin loss and its repair by remyelination therefore is an important goal for in vivo pathological studies of MS. The lack of speciWcity of T1 and T2 relaxation time changes limits their usefulness as MR-based indices of myelin damage. MT imaging deWnes changes in biophysical parameters that are altered by myelin loss more selectively than are water proton T1 or T2 (McGowan, 1999) (Table 32.3). The most direct evidence for this has come from correlative pathological and imaging studies of a primate model (Brochet and Dousset, 1999). In this EAE model, there is a close correlation between MTR changes prior to sacriWce and histopathological changes of demyelination post-mortem. An alternative and promising newer technique for myelin visualization is selective imaging of the component of water with a very short T2 relaxation time (about 20 msec), which includes water trapped between the layers of myelin (Fig. 32.7) (Laule et al., 2002; Webb et al., 2002). This short T2 water may be more speciWc for myelin than water associated with macromolecules in general and correlates extremely well with myelin content on histological examination (Gareau et al., 2000). MacKay and his coworkers (MacKay et al., 1994) have spatially mapped the amount of this ‘‘myelin water’’ in the brains of patients with MS. Postmortem studies conWrmed an association between loss of the short T2 component and

MRI VISUALIZATION OF THE PROGRESSION OF MS LESIONS

775

FIGURE 32.7 Mapping the distribution of the short T2-relaxation time component of the tissue water relaxation decay curve for the brain may help to identify myelin distribution. Water associated with myelin may have the shortest T2 relaxation time of any brain water compartment. In a normal brain (A) the water short T2 relaxation time component distribution is relatively evenly distributed over the central white matter (B). In contrast, in a patient with multiple sclerosis and focal white matter lesions (C), the short T2 component is reduced particularly in areas of focal lesions, suggesting demyelination. (Images courtesy of C. Laule and A. McKay, University of British Columbia).

demyelination (Moore et al., 2000). The myelin-associated water content may be more than 50% lower in lesions than in the surrounding white matter (Laule et al., 2002). A limitation to the technique at present is the long time necessary for acquisition of the full relaxation time dataset, as well as the hardware and analytical demands for measuring such short relaxation time components of the total water relaxation accurately. There is a general problem with the interpretation of changes in MT or myelin-associated water exclusively in terms of demyelination, as myelin and axonal loss usually occur concomitantly (Arnold et al., 1992; van Waesberghe et al., 1999). Thus, while MT is sensitive to demyelination, it is diYcult in practice to quantitatively assess the extent of demyelination independent of axonal loss using MT alone (Filippi, 1999). Quantitative evaluation of myelin integrity also suggests that not all the pathology in MS is conWned to focal lesions. In the extra-lesional white matter there are in fact changes

776

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

relative to healthy white matter, although these are more modest than in focal lesions. T2 compartmentation studies, for example, suggest an increase in total water content of about 2% and a reduction in myelin-associated water content of approximately 15% in the socalled normal-appearing white matter (NAWM) (Laule et al., 2002).

MRI VISUALIZATION OF DIFFUSE WHITE MATTER INVOLVEMENT IN MS Pathologists traditionally have focused attention on the plaques of MS. An important general contribution of MRI techniques has been to emphasize the importance of pathology outside of the lesions. It is becoming increasingly clear that in established MS all white matter shows some evidence of pathological change if techniques with appropriate sensitivity are used. A compelling reason to be interested in these diVuse changes is that the focal lesion load in a typical MS brain constitutes at most a few percent of the total white matter volume: the bulk of changes occurs diVusely.

Magnetic Resonance Spectroscopy for Assessment of Diffuse Axonal Injury and Loss None of the MRI techniques that measure signals from water are able to directly detect and quantify the dysfunction of neurons and their axonal processes. However, this information can be obtained from MRS or MRSI (Rudkin and Arnold, 1999). In one of the earliest MRS reports (Arnold et al., 1992), N-acetylaspartate (NAA) was shown to be decreased in MS patients when a large central brain volume of interest was used for acquisition of the proton spectrum. N-acetylaspartate is a speciWc marker of axonal integrity in the adult central nervous system (Clark, 1998; Matthews and Arnold, 2001; Simmons et al., 1991; Trapp et al., 1998; Tsai and Coyle, 1995) (Table 32.3). As the volume of lesions within the large spectroscopic volume was small, the bulk of changes must have occurred diVusely in the normal-appearing tissue, which was predominantly NAWM. A diVuse decrease of white matter NAA has been subsequently observed in many studies (Cucurella et al., 2000; De Stefano et al., 2001; Fu et al., 1998; Husted et al., 1994; Leary et al., 1999; Sarchielli et al., 1999; Tedeschi et al., 2002). The extent of the diVuse reduction of NAA measured by MRS is at least approximately consistent with the relative loss of axons measured directly in white matter projection volumes using histopathological methods. The relative contribution of extra-lesional NAA decrease is diVerent in patients with RR and SP disease and accounts for the axonal injury and loss that correlates best with the progression of disability (Matthews et al., 1996). Later studies directly contrasted the relative concentrations of NAA within lesions and in surrounding white matter. Decreases are most marked in lesions, but smaller reductions also occur outside of plaques. It remains unclear how much of the diVuse loss is due to secondary consequences of axonal transection in focal lesions (anterograde or retrograde degeneration), diVuse inXammation, or a more primary neurodegenerative process. The extent of this NAA reduction decreases with the distance from the core of a lesion (Arnold et al., 1992), consistent with the notion that the diVuse changes are at least in part related to dying back of axons transected (Trapp et al., 1998) within plaques. There is also a correlation between the extent of diVuse axon loss and local lesion load suggested both by spectroscopic imaging (Matthews et al., 1996) and direct histopathological observations (Evangelou et al., 2000). Although retrograde or anterograde changes resulting from focal lesions contribute to the diVuse abnormalities, other factors may be involved. DiVusible toxins such as proteolytic enzymes, cytokines and nitric oxide, and other free radicals may damage axons and glial cells outside, and sometimes at a considerable distance from, the focal lesions. Immunoglobulins directed against both neurons and oligodendroylial cells may cause damage or dysfunction (Bauer et al., 2001; Rieckmann and Smith, 2001).

MRI VISUALIZATION OF DIFFUSE WHITE MATTER INVOLVEMENT IN MS

Brain and Spinal Cord Atrophy: Measures of Neuronal and Glial Changes DiVuse damage also is demonstrated by atrophy of brain and spinal cord. Many measures of atrophy based on imaging have been proposed. Although these vary in sensitivity to change and to some extent in their regional speciWcity, they commonly demonstrate enhanced rates of loss of CNS parenchyma in MS. Rates of volume change are in general well correlated with the progression of disability in later stages of MS (LosseV et al., 1996b). A striking Wnding from recent longitudinal MRI studies has been that brain and spinal cord atrophy begins early in the disease. RR patients with mild disability may have substantially increased rates of brain substance loss, both in the white and the gray matter (Chard et al., 2002). SigniWcant cerebral atrophy reXected in lateral ventricular enlargement can occur even in the interval between the Wrst clinical presentation and clinical diagnosis of MS (Brex et al., 2000). The speciWc tissue changes that contribute to the genesis of this atrophy remain uncertain (Miller et al., 2002). These likely include axonal loss and demyelination in the white matter, as well as glial changes in chronic lesions. In gray matter changes in myelin content and axonal loss are also found, as well as atrophy of dendritic arborisations and loss of neurons. It is likely, but not well established, that the relative contributions of atrophy in diVerent tissue compartments may change during the course of the disease. The primary utility of atrophy as a marker of disease progression lies in the extent to which its magnitude and rate of increase reXect irreversible nervous system injury and are correlated with disability and its worsening (Edwards et al., 1999; LosseV et al., 1996b; Nijeholt et al., 1998). As atrophy can be measured from serially acquired T1-weighted images entirely automatically (Stevenson et al., 2002), it also provides a measure that is reasonably sensitive to change and not very demanding of special hardware or analysis capabilities. There is evidence that the rate of atrophy may be related to inXammatory activity, at least in the RR stage of MS. The number of Gd-enhancing lesions at baseline in the placebo arm of an interferon-beta trial predicted the relative extent of atrophy over the subsequent 2-year period (Simon et al., 1998), a Wnding supported by cross-sectional (Lin and Blumhardt, 2001) and longitudinal (Leist et al., 2001) studies. However, this relationship may either be variable or potentially confounded by other factors (Paolillo et al., 2000; Rudick et al., 1999; Saindane et al., 2000). One factor likely to contribute is time after injury, perhaps because demyelinated axons are chronically deprived of the trophic support of myelin. A recent study has shown that optic nerve atrophy continues for more than two years after an episode of optic neuritis (Hickman et al., 2002). However, whether some degree of ongoing, subclinical inXammatory activity might contribute to this apparent progressive axonal degeneration is at present unresolved.

Other Quantitative Techniques That Show Diffuse White Matter Abnormalities Other quantitative measures provide further evidence for diVuse pathology. The MTR is diVusely low in white matter of patients with MS (Catalaa et al., 2000; Cercignani et al., 2001; Ge et al., 2002; Guo et al., 2001; Siger-Zajdel and Selmaj, 2001). Water proton relaxation time measurements also show diVuse changes in the NAWM (Miller et al., 1989; Ormerod et al., 1986). Recent data from a novel, rapid T1 mapping technique have deWned changes in the T1 relaxation time histogram for extra-lesional white matter that are strongly correlated with disability (Parry et al., 2002a). T2-relaxation times may be similarly diVusely prolonged in white matter of MS patients (Barbosa et al., 1994). Finally, diVusion anisotropy measurements also show signiWcant increases in white matter that appears normal on conventional imaging (Bammer et al., 2000; Cercignani et al., 2000; Christiansen et al., 1993; Ciccarelli et al., 2001; Filippi et al., 2000, 2001; Guo et al., 2001; HorsWeld et al., 1996; Rocca et al., 2000; Werring et al., 1999, 2001). Recent histopathological studies have conWrmed the substantial diVuse damage in the white matter suggested by imaging studies (Allen et al., 2001). Ferguson and coworkers (1997) noted that the expression of amyloid precursor protein (APP), a marker of axonal

777

778

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

injury, was abnormally elevated around active chronic lesions. Trapp et al. (1998) have reported abnormal hypophosphorylated neuroWlaments in axons outside of lesions. More recent work has directly measured the diVuse axonal loss and matrix changes distant from plaques. Ganter and coworkers (1999) noted that the density of axons in cervical thoracic spinal cord outside of lesions was reduced up to 42% relative to controls. Further work by Evangelou et al. (Evangelou et al., 2000) showed that in the corpus collosum, a brain region ideal for axon quantiWcation because of the highly oriented structure, axonal density was reduced by approximately 35% outside of lesions. An important observation in this study was that both axonal density and the cross-sectional area of the corpus collosum decrease in MS patients, suggesting that independent measures of either brain volume or axonal density changes underestimate the total injury in the diVuse white matter.

MRI VISUALIZATION OF GRAY MATTER PATHOLOGY IN MS Lesions have long been described in gray matter (Brownell and Hughes, 1962; Lumsden, 1970). The relative lack of previous appreciation for the importance of gray matter abnormalities arose because traditional pathological approaches (Peterson et al., 2001) are relatively insensitive to cortical lesions. Conventional T2-weighted imaging also is insensitive to these gray matter changes (Kidd et al., 1999; Miller et al., 1998). For example, in a post-mortem imaging study of unWxed brains, out of 54 gray matter lesions identiWed histologically, only 2 were detected by T2-weighted MRI (Newcombe et al., 1991). FLAIR imaging increases the sensitivity for cortical lesions, particularly those that are juxtacortical (Bakshi et al., 2001; Boggild et al., 1996; Filippi et al., 1996a; Gawne-Cain et al., 1997), but is still relatively insensitive to the majority of lesions, which are intracortical. The small sizes of the discrete cortical lesions, and diVerences in the nature of the inXammatory changes in gray matter lesions and in the structure of gray relative to white matter, likely account for the insensitivity of conventional MRI to gray matter lesions (Peterson et al., 2001). It is also possible that imaging characteristics of gray matter lesions may be fundamentally diVerent from plaques in the white matter. Bakshi et al. (2002) have reported that T2-hypointense lesions can be deWned in most gray matter regions in patients with established MS. The hypointensity may be related to T2-shortening with deposition of paramagnetic iron in the lesions, a Wnding associated nonspeciWcally with neurodegeneration in other contexts. As more sensitive techniques begin to be applied, it is likely that the contribution of gray matter to the total brain pathology in MS will be shown to be substantial (Chard et al., 2002). Peterson et al. (2001) have emphasized that large, conXuent volumes of hypomyelination in neocortex are common. While these have not yet been visualized using MRI, imaging suggests that gray matter atrophy may be substantial. Chard et al. (2002) demonstrated that as much as 50% of total brain atrophy could be ascribed to neocortical atrophy in RR MS. Because gray matter constitutes well over 50% of brain volume, gray matter atrophy contributes profoundly to total brain volume changes. However, these changes are diYcult to measure, because the neocortex is so thin. Recent work by Chen et al. (2002) has attempted to more precisely deWne the loss of neocortex in diVerent stages of multiple sclerosis. Substantial loss was shown both in early RR and later SP stages. Although the sample was limited, the data suggested that the relative contribution of the neocortical volume loss to total brain atrophy was substantially greater in RR than in the SP patients. In fact, loss of neocortex accounted for most of the total brain atrophy in patients in the earlier stages of the disease. MRS measurements have shown signiWcant decreases in NAA in the neocortex of MS patients, consistent with neuronal or axonal injury (Kapeller et al., 2001; Presciutti et al., 2000). This NAA reduction, however, could also be explained by a retraction of dendritic arborisation in the gray matter (Tseng and Hu, 1996). Interpretation of these Wndings is confounded by the concomitant development of atrophy and partial volume eVects.

779

VISUALIZATION OF ‘‘PRELESIONAL’’ CHANGES

This problem was addressed by Cifelli and coworkers (2002), who studied neurodegeneration in the thalamus using either MRS or histopathological methods for similar SP MS cohorts. Because the thalamus does not include sulcal CSF spaces, its MRS investigation can be performed without the confound of partial volume eVects. Using a specially tailored MRI sequence that deWned the borders of the thalamus well, Cifelli et al. (2002) demonstrated a mean 17% loss of thalamic volume in patients with SP MS (Fig. 32.8). This was associated with a 19% decrease in the relative NAA concentration (a measure of the loss of neuronal or axonal density) to suggest a total neuronal loss of about 30%. In a parallel histopathological study with comparable post-mortem specimens, they measured a similar volume loss in the mediodorsal nucleus of the thalamus directly and showed that this was associated with a 22% loss of neuronal density, a 21% loss of volume and thus a total loss of neurons of about 35%, which is comparable to the decrease deWned in vivo. Just as in the earlier studies of axonal loss in the corpus callosum, both neuronal density loss and tissue volume loss contributed to estimation of the total change. More recently, this work was extended to patients with RR MS, who showed similar changes, the magnitude of which was related to the duration of disease (Wylezinska et al., 2003). Together, these MRS and atrophy studies emphasize that gray matter pathology contributes a substantial proportion of the load of disease in MS. DiVerences in gray matter pathology could account for the apparent lack of a consistent association between measures of focal white matter disease and disability across diVerent clinical subtypes of MS. The diVerences in sensitivity of MR measures to gray and white matter abnormalities emphasize the need to use multiple MR-based techniques simultaneously in order to describe the full range of pathology in this disease.

au2

VISUALIZATION OF ‘‘PRELESIONAL’’ CHANGES Although all focal lesion changes may not be initiated by a phase of increased BBB permeability, until recently Gd-enhancement was the earliest focal change that could be

A

B

FIGURE 32.8 The acquisition parameters of these MRI images have been optimized by means of simulations in order to achieve high contrast between central gray matter and surrounding white matter. Moderate enlargement of both lateral and third ventricles can be noted in the MS patient (B). Volume loss of the thalamic gray matter can also be observed in the patient as compared with the normal control (A).

780

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

detected in the evolution of new lesions. However, other, quantitative MR techniques reveal focal changes in NAWM that precede Gd-enhancement and the appearance of T2weighted hyperintensity. Spectroscopy acquisitions at short TE can reveal signals from mobile macromolecules (which arise mainly from lipids): they become MRS-visible due to increases in mobility associated with demyelination. In their longitudinal study, Narayana and his colleagues (1998) found examples of the focal appearance of lipid peaks in regions that later developed new T2-hyperintense lesions. De Stefano and colleagues (2001) found a focal increase in Cho preceding the development of new T2 lesions. This suggests that low-grade, focal myelin pathology may antedate the development of acute, severe inXammation. Focal MTR changes also can occur prior to the appearance of T2-hyperintense lesions. Filippi’s group (1998) serially studied RR patients over 3 months and outlined the contours of new enhancing lesions. These regions then were mapped onto coregistered MT images acquired at previous time points. Even before lesion development, MTR was focally reduced in these volumes. Changes were progressive and proportionally higher in the month preceding enhancement. Similar Wndings have been described by others (Goodkin et al., 1998). A more recent study (Pike et al., 2000) based on less frequent scanning over a much longer period demonstrated subtle MTR changes predating the development of T2 lesions by years. The rates of change of MTR were remarkably consistent before and after lesion appearance, suggesting that the pathology associated with MTR decline is continuous and accelerates only transiently during the acute inXammation associated with Gd-enhancement and new T2 lesion formation.

ADAPTIVE REORGANIZATION CAN BE VISUALIZED USING FUNCTIONAL MAGNETIC RESONANCE IMAGING (FMRI) Forms of Adaptive Reorganization after Brain Injury in MS Recovery from the brain injury of MS involves several mechanisms that can be visualized by MR-based techniques. Resolution of the primary inXammation and repair of myelin must have a role, which can be demonstrated as reduction of Gd-enhancement and resolution of T2-hyperintensity, respectively. Also, impairment (and partial restoration) of axonal and neuronal metabolic function takes place and can be monitored by changes in NAA. However, an increasingly strong case can be made for the importance of adaptive cerebral plasticity, which can occur at a number of levels: 1. Axonal, with expression of new sodium channels (Waxman, 2001) 2. Neuronal, with enhanced dendritic arborisation from surviving neurons (Jones and Schallert, 1992) 3. Synaptic, with changes in synapse number or distribution with respect to the soma 4. Systems organization, with altered recruitment of parallel processing pathways or other ‘‘latent connections’’ (Jacobs and Donoghue, 1991) FMRI is proving useful in deWning the systems-level changes directly. Animal studies have shown a direct correlation between behavior and electrophysiologically deWned changes in the cortical representations for movement or sensation in primary motor cortex around focal lesions. For example, after an ischaemic lesion of the hand area in the motor cortex of an adult squirrel monkey, the hand movement representation changed over time after infarction (Nudo et al., 1996). However, while with injury alone the hand representation decreased by at least 25%, if aggressive physiotherapy was used the representation could increase by 10%. The increased representation in some areas occurs at the expense of neighboring regions and is correlated with improved function. Cortical reorganization also can occur at a distance from a focal lesion. Dendritic remodeling is stimulated in homotopic neocortex contralateral to a focal injury. Immobilization of the paretic limb prevents the dendritic growth and impairs functional recovery (Kozlowski et al., 1996).

ADAPTIVE REORGANIZATION CAN BE VISUALIZED USING FMRI

Functional Magnetic Resonance Imaging (FMRI): Imaging Functional Reorganization of the Human Brain after Injury from MS FMRI allows patterns of brain activation associated with sensory, motor or cognitive tasks to be mapped with greater sensitivity than was possible before (Matthews, 2001). FMRI demonstrates a widely distributed network of regions involved in the control of even a simple hand movement, for example. DiVerences in the patterns of activity in such a network can be deWned between patients and healthy controls (Rocca et al., 2002a, 2002b). One of the most consistent Wndings has been relatively increased ipsilateral motor cortex activation (Lee et al., 2000; Pantano et al., 2002). Just as has been demonstrated with white matter ischaemic disease (Reddy et al., 2002a), in MS there is a strong correlation between increasing disease burden and the extent of the functional changes (Lee et al., 2000; Pantano et al., 2002; Rocca et al., 2002a). The functional changes identiWed also appear to be dynamic, just like the pathology of MS. In a case report based on the study of a relapsing-remitting MS patient with a very large demyelinating lesion of the left hemisphere and resolving right hemiplegia, Reddy et al. (2000b) correlated serially acquired measures of clinical evolution, lesion size assessed from conventional MRI, biochemical pathology deWned with magnetic resonance spectroscopic imaging (MRSI), and relative cortical activation during a Wnger-thumb opposition paradigm with fMRI over the 6 months following presentation of the lesion. The NAA concentrations in the corticospinal tract increased in parallel with recovery from functional impairment. Abnormal patterns of fMRI activation were found throughout the period of study. Although the extent of the abnormality was greatest when the lesion was largest and NAA in the corticospinal tracts was lower fMRI activation remained abnormal even after apparent clinical recovery of motor function. Together, these observations are consistent with the hypothesis that the changes are functionally adaptive, although often only incompletely so. In some cases, cortical areas related to polymodal or higher levels of processing may become more involved with injury to primary pathways in sensory system. Werring et al. (2000) made the intriguing observation that patients who had suVered from optic neuritis showed reduced primary visual cortical activation with photic stimulation, but increased activation of extra-striate visual areas, including polymodal sensory areas such as the claustrum. The apparent relationship to primary tissue injury was demonstrated by a correlation between delay in the P100 visual evoked potential peak and extent of extrastriate activations.

Adaptive Change May Limit the Clinical Expression of Deficits in MS A common concern in the interpretation of imaging studies of brain plasticity is that the activation changes observed might not be due to adaptive phenomena, but instead could simply result from diVerences in performance between the patients and healthy subjects. Performance cannot be easily controlled for in studies of movement. Even if behavior is similar, the relative diYculty or ‘‘eVortfulness’’ may not be well matched (Lee et al., 2000). One partial solution to this conundrum is the study of patients without any clinical deWcits for the movement under examination. Reddy and colleagues (Reddy et al., 2000a), for example, studied MS patients without any clinically evident motor or sensory impairment in the upper limbs with a Wnger Xexion-extension paradigm. Even in these patients (whose behavior was well matched with the healthy controls), there was a strong correlation between the extent of ipsilateral sensorimotor cortex activation and concentration of NAA in voxels localized to the descending corticospinal tract. An alternative approach is to use a task that reports on elements of relevance to motor control but that is intrinsically well matched between patients and healthy controls. Because of the rich, reciprocal innervations between motor and sensory cortex, passive movement of the hand activates cortical regions that would be active with a similar, active movement (Reddy et al., 2001, 2002c). Comparison of fMRI activation patterns associated with active and passive hand movements in patients and healthy controls conWrmed that

781

782

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

diVerences in patterns of activation related to disease burden can be found even when performance and its diYculty is matched (Reddy et al., 2002b). However, disease burden is just one factor that may contribute to determining the extent of functional reorganization with MS. A recent study attempted to test whether altered patterns of limb use have a distinct and direct inXuence on patterns of brain organization. Three groups of MS patients with diVerent degrees of disability and white matter disease burden as assessed from brain atrophy and NAA decreases were studied (Reddy et al., 2002b). One group had no evidence of either substantial injury or functional impairment, a second group had signiWcant white matter injury but no upper limb impairment, and the Wnal group had a similar burden of brain injury but showed substantial upper limb impairment. Contrast of patients with no hand impairment, but with reduced or normal white matter NAA, showed signiWcant activation increases in the ipsilateral premotor cortex and the supplementary motor area bilaterally. To assess whether disability itself can alter patterns of cortical activation associated with hand weakness, a contrast was made between patients with decreased NAA and either impaired or unimpaired hand function. This contrast demonstrated greater bilateral primary and secondary somatosensory cortex activation with greater limb disability. The authors concluded that the pattern of cerebral activity with Wnger movements changes independently both with increasing injury and with increasing disability. It was hypothesized that the changes related to disability may be caused by altered patterns of use. Potentially adaptive functional changes also may occur with purely cognitive processes. In a study of MS patients with mild neuropsychological deWcits, StaVen et al. demonstrated that patients show abnormal, increased recruitment of prefrontal cortex (corresponding approximately to Brodman areas 6, 8, and 9) during a visual serial addition task. Parry et al. (2002b) identiWed abnormally increased activity in a similar region of frontopolar cortex in MS patients performing the Stroop task, a test of executive function, and demonstrated that the extent of recruitment of this region is increased with greater disease burden (Fig. 32.9). Functional imaging has the potential to identify regions that may be critical for the genesis of symptoms diYcult to localize using conventional strategies. Filippi’s group (Filippi et al., 2002) applied a simple hand movement paradigm to distinguish brain activity changes associated with MS-related fatigue. MS patients with fatigue showed relative increases in brain activity in several brain regions, including the thalamus, intraparietal sulcus, and rolandic operculum. The extent of the increases in these regions were correlated with fatigue scores, suggesting that activity in these areas may contribute to the genesis of symptoms.

CONCLUSIONS MRI and related MR-based techniques oVer an increasingly comprehensive view of the pathology of MS. Because these methods are noninvasive and well tolerated, longitudinal studies have allowed the dynamics of pathological changes to be deWned directly. Such studies have emphasized that clinical expression has a complex relation to the underlying dynamics of the pathological change, suggesting that maybe the latter should be targeted directly in neurobiologically driven, rational treatment strategies. While inXammatory demyelination may be the most obvious histological feature of MS, imaging studies have led to a shift in focus toward associated pathologies. It has become clear that axonal and neuronal loss, rather than the damage to myelin, is responsible for the irreversible progression of disability. One important consequence of this concept is that it links strategies for limiting disability in MS to those for control of progression in the broad range of primary neurodegenerative diseases. With the advent of functional imaging methods, it may be expected that better characterization of previously poorly understood symptoms such as fatigue, attentional, and memory impairments will be possible. In general, by allowing clear relationships to be

CONCLUSIONS

783

FIGURE 32.9 Functional magnetic resonance imaging can deWne altered patterns of brain activation with cognitive tasks in patients with multiple sclerosis. The Stroop paradigm is a test of executive function demanding inhibition of a preferred response. The counting Stroop task, for example, demands subjects to report the number of words presented on a screen. The words may be either neutral words (e.g., cat, dog) or number words (e.g., three, four). Response times for the latter are prolonged relative to the former, because of the need to suppress answers based on the words’ meaning rather than their number. In healthy controls, this produces activation in multiple areas (red and yellow). Patients with multiple sclerosis activate similar areas (yellow), but fail to show signiWcant activation in the right inferior frontal cortex (red). Unlike controls, they show signiWcant activation in predominantly left frontopolar areas (blue). Activation in the frontopolar areas relative to the right inferior frontal area is related directly to disease burden.

established between structural, functional, and behavioral changes in individual patients, imaging is ushering in an exciting new era of MRI-based pathology that should play a critical role in relieving suVering from this disease.

Acknowledgments AC and PMM thank the MS Society of Great Britain and Northern Ireland and the Medical Research Council, and DLA thanks the MS Society of Canada and the Canadian Institutes of Health Research for support. All of us acknowledge the considerable assistance we have received over the years from our many collaborators, particularly those within the Brain Imaging Laboratory Linkage.

References Allen, I. V., McQuaid, S., Mirakhur, M., and Nevin, G. (2001). Pathological abnormalities in the normalappearing white matter in multiple sclerosis. Neurol. Sci. 22, 141–144. Arnold, D. L., and Matthews, P. M. (2002). MRI in the diagnosis and management of multiple sclerosis. Neurology 58, S23-S31. Arnold, D. L., Matthews, P. M., Francis, G. S., O’Connor, J., and Antel, J. P. (1992). Proton magnetic resonance spectroscopic imaging for metabolic characterization of demyelinating plaques. Ann. Neurol. 31, 235–241. Bakshi, R., Ariyaratana, S., Benedict, R. H., and Jacobs, L. (2001). Fluid-attenuated inversion recovery magnetic resonance imaging detects cortical and juxtacortical multiple sclerosis lesions. Arch. Neurol. 58, 742–748. Bakshi, R., Benedict, R. H., Bermel, R. A., Caruthers, S. D., Puli, S. R., Tjoa, C. W., Fabiano, A. J., and Jacobs, L. (2002). T2 hypointensity in the deep gray matter of patients with multiple sclerosis: A quantitative magnetic resonance imaging study. Arch. Neurol. 59, 62–68. Bammer, R., Augustin, M., Strasser-Fuchs, S., Seifert, T., Kapeller, P., Stollberger, R., Ebner, F., Hartung, H. P., and Fazekas, F. (2000). Magnetic resonance diVusion tensor imaging for characterizing diVuse and focal white matter abnormalities in multiple sclerosis. Magn Reson. Med. 44, 583–591. Barbosa, S., Blumhardt, L. D., Roberts, N., Lock, T., and Edwards, R. H. (1994). Magnetic resonance relaxation time mapping in multiple sclerosis: normal appearing white matter and the ‘‘invisible’’ lesion load. Magn. Reson. Imaging. 12, 33–42.

784

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

Barkhof, F., Filippi, M., Miller, D. H., Scheltens, P., Campi, A., Polman, C. H., Comi, G., Ader, H. J., LosseV, N., and Valk, J. (1997). Comparison of MRI criteria at Wrst presentation to predict conversion to clinically deWnite multiple sclerosis. Brain 120 (Pt. 11), 2059–2069. Barkhof, F., and van Walderveen, M. (1999). Characterization of tissue damage in multiple sclerosis by nuclear magnetic resonance. Philos. Trans. R. Soc. Lond B Biol. Sci. 354, 1675–1686. Barkhof, F., and Scheltens, P. (2002). Imaging of white matter lesions. Cerebrovasc. Dis. 13 Suppl 2, 21–30. Bauer, J., Rauschka, H., and Lassmann, H. (2001). InXammation in the nervous system: The human perspective. Glia 36, 235–243. Bergers, E., Bot, J. C., van, d., V, Castelijns, J. A., Nijeholt, G. J., Kamphorst, W., Polman, C. H., Blezer, E. L., Nicolay, K., Ravid, R., and Barkhof, F. (2002). DiVuse signal abnormalities in the spinal cord in multiple sclerosis: direct postmortem in situ magnetic resonance imaging correlated with in vitro high-resolution magnetic resonance imaging and histopathology. Ann. Neurol. 51, 652–656. Bitsch, A., Bruhn, H., Vougioukas, V., Stringaris, A., Lassmann, H., Frahm, J., and Bruck, W. (1999). InXammatory CNS demyelination: histopathologic correlation with in vivo quantitative proton MR spectroscopy. AJNR Am. J. Neuroradiol. 20, 1619–1627. Bjartmar, C., Kidd, G., Mork, S., Rudick, R., and Trapp, B. D. (2000). Neurological disability correlates with spinal cord axonal loss and reduced N-acetyl aspartate in chronic multiple sclerosis patients. Ann. Neurol. 48, 893–901. Boggild, M. D., Williams, R., Haq, N., and Hawkins, C. P. (1996). Cortical plaques visualised by Xuid-attenuated inversion recovery imaging in relapsing multiple sclerosis. Neuroradiology 38 Suppl 1, S10-S13. Brass, S. D., Santos, A. C., Francis, S. J., Caramanos, Z., Parrilla, G., Lapierre, Y., Levesque, I., Narayanan, S., Sled, J. G., Pike, G. B., and Arnold, D. L. (2002a). Regional variations of quantitative magnetisation transfer imaging parameters in normal-appearing gray matter and normal-appearing white matter in multiple sclerosis. Proc. Int. Soc. Magn. Reson. Med. 10, 1178. Brass, S. D., Santos, A. C., Francis, Caramanos, Z., Parrilla, G., Lapierre, Y., Levesque, I., Narayanan, S., Sled, J. G., Pike, G. B., and Arnold, D. L. (2002b). Quantitative magnetization transfer imaging parameters in white matter, grey matter, and lesions in patients with multiple sclerosis compared to normal controls. Proc. Int. Soc. Magn. Reson. Med. 10, 1179. Brex, P. A., Ciccarelli, O., O’Riordan, J. I., Sailer, M., Thompson, A. J., and Miller, D. H. (2002). A longitudinal study of abnormalities on MRI and disability from multiple sclerosis. N. Engl. J. Med. 346, 158–164. Brex, P. A., Jenkins, R., Fox, N. C., Crum, W. R., O’Riordan, J. I., Plant, G. T., and Miller, D. H. (2000). Detection of ventricular enlargement in patients at the earliest clinical stage of MS. Neurology 54, 1689–1691. Brochet, B., and Dousset, V. (1999). Pathological correlates of magnetization transfer imaging abnormalities in animal models and humans with multiple sclerosis. Neurology 53, S12-S17. Brownell, B., and Hughes J. T. (1962). The distribution of plaques in the cerebrum in multiple sclerosis. J. Neurol. Neurosurg. Psychiat. 25, 315–320. Bruck, W., Bitsch, A., Kolenda, H., Bruck, Y., Stiefel, M., and Lassmann, H. (1997). InXammatory central nervous system demyelination: correlation of magnetic resonance imaging Wndings with lesion pathology. Ann. Neurol. 42, 783–793. Catalaa, I., Grossman, R. I., Kolson, D. L., Udupa, J. K., Nyul, L. G., Wei, L., Zhang, X., Polansky, M., Mannon, L. J., and McGowan, J. C. (2000). Multiple sclerosis: Magnetization transfer histogram analysis of segmented normal-appearing white matter. Radiology 216, 351–355. Cercignani, M., Bozzali, M., Iannucci, G., Comi, G., and Filippi, M. (2001). Magnetisation transfer ratio and mean diVusivity of normal appearing white and grey matter from patients with multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 70, 311–317. Cercignani, M., and HorsWeld, M. A. (2001). The physical basis of diVusion-weighted MRI. J. Neurol. Sci. 186 Suppl 1, S11-S14. Cercignani, M., Iannucci, G., and Filippi, M. (1999). DiVusion-weighted imaging in multiple sclerosis. Ital. J. Neurol. Sci. 20, S246-S249. Cercignani, M., Iannucci, G., Rocca, M. A., Comi, G., HorsWeld, M. A., and Filippi, M. (2000). Pathologic damage in MS assessed by diVusion-weighted and magnetization transfer MRI. Neurology 54, 1139–1144. Chard, D. T., GriYn, C. M., Parker, G. J., Kapoor, R., Thompson, A. J., and Miller, D. H. (2002). Brain atrophy in clinically early relapsing-remitting multiple sclerosis. Brain 125, 327–337. Chen, J. T., Smith, S. M., Arnold, D. L., and Matthews, P. M. (2002). QuantiWcation of change in cortical grey matter thickness in multiple sclerosis. Proc. Int. Soc. Magn. Reson. Med. 10, 351. Christiansen, P., Gideon, P., Thomsen, C., Stubgaard, M., Henriksen, O., and Larsson, H. B. (1993). Increased water self-diVusion in chronic plaques and in apparently normal white matter in patients with multiple sclerosis. Acta Neurol. Scand. 87, 195–199. Ciccarelli, O., Werring, D. J., Wheeler-Kingshott, C. A., Barker, G. J., Parker, G. J., Thompson, A. J., and Miller, D. H. (2001). Investigation of MS normal-appearing brain using diVusion tensor MRI with clinical correlations. Neurology 56, 926–933. Cifelli, A., Arridge, M., Jezzard, P., Esiri, M. M., Palace, J., and Matthews, P. M. (2002). Thalamic neurodegeneration in multiple sclerosis. Ann. Neurol. (in press). Clark, J. B. (1998). N-acetyl aspartate: a marker for neuronal loss or mitochondrial dysfunction. Dev. Neurosci. 20, 271–276.

CONCLUSIONS

Cucurella, M. G., Rovira, A., Rio, J., Pedraza, S., Tintore, M. M., Montalban, X., and Alonso, J. (2000). Proton magnetic resonance spectroscopy in primary and secondary progressive multiple sclerosis. NMR Biomed. 13, 57–63. Dalton, C. M., Ramani, A., Wheeler-Kingshott, C., Barker, G. J., Miller, D. H., and Tofts, P. S. (2002). Bound water magnetization transfer measurements in patients with multiple sclerosis: a pilot study. Proc. Int. Soc. Magn. Reson. Med. 10, 1180. De Stefano, N., Narayanan, S., Francis, G. S., Arnaoutelis, R., Tartaglia, M. C., Antel, J. P., Matthews, P. M., and Arnold, D. L. (2001). Evidence of axonal damage in the early stages of multiple sclerosis and its relevance to disability. Arch. Neurol. 58, 65–70. Edwards, S. G., Gong, Q. Y., Liu, C., Zvartau, M. E., Jaspan, T., Roberts, N., and Blumhardt, L. D. (1999). Infratentorial atrophy on magnetic resonance imaging and disability in multiple sclerosis. Brain 122 (Pt. 2), 291–301. Evangelou, N., Konz, D., Esiri, M. M., Smith, S., Palace, J., and Matthews, P. M. (2000). Regional axonal loss in the corpus callosum correlates with cerebral white matter lesion volume and distribution in multiple sclerosis. Brain 123 (Pt. 9), 1845–1849. Fazekas, F., OVenbacher, H., Fuchs, S., Schmidt, R., Niederkorn, K., Horner, S., and Lechner, H. (1988). Criteria for an increased speciWcity of MRI interpretation in elderly subjects with suspected multiple sclerosis. Neurology 38, 1822–1825. Fazekas, F., Strasser-Fuchs, S., Schmidt, H., Enzinger, C., Ropele, S., Lechner, A., Flooh, E., Schmidt, R., and Hartung, H. P. (2000). Apolipoprotein E genotype related diVerences in brain lesions of multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 69, 25–28. Ferguson, B., Matyszak, M. K., Esiri, M. M., and Perry, V. H. (1997). Axonal damage in acute multiple sclerosis lesions. Brain 120 (Pt 3), 393–399. Filippi, M. (1999). Magnetization transfer imaging to monitor the evolution of individual multiple sclerosis lesions. Neurology 53, S18–S22. Filippi, M., Cercignani, M., Inglese, M., HorsWeld, M. A., and Comi, G. (2001). DiVusion tensor magnetic resonance imaging in multiple sclerosis. Neurology 56, 304–311. Filippi, M., Iannucci, G., Cercignani, M., Assunta, R. M., Pratesi, A., and Comi, G. (2000) A quantitative study of water diVusion in multiple sclerosis lesions and normal-appearing white matter using echo-planar imaging. Arch. Neurol. 57, 1017–1021. Filippi, M., Rocca, M. A., Martino, G., HorsWeld, M. A., and Comi, G. (1998). Magnetization transfer changes in the normal appearing white matter precede the appearance of enhancing lesions in patients with multiple sclerosis. Ann. Neurol. 43, 809–814. Filippi, M., Rocca, M. A., Martino, G., HorsWeld, M. A., and Comi, G. (1998). Magnetization transfer changes in the normal appearing white matter precede the appearance of enhancing lesions in patients with multiple sclerosis. Ann. Neurol. 43, 809–814. Filippi, M., Rossi, P., Campi, A., Colombo, B., Pereira, C., and Comi, G. (1997). Serial contrast-enhanced MR in patients with multiple sclerosis and varying levels of disability. AJNR Am. J. Neuroradiol. 18, 1549–1556. Filippi, M., Yousry, T., Baratti, C., HorsWeld, M. A., Mammi, S., Becker, C., Voltz, R., Spuler, S., Campi, A., Reiser, M. F., and Comi, G. (1996a). Quantitative assessment of MRI lesion load in multiple sclerosis. A comparison of conventional spin-echo with fast Xuid-attenuated inversion recovery. Brain 119 (Pt. 4), 1349–1355. Filippi, M., Yousry, T., Campi, A., Kandziora, C., Colombo, B., Voltz, R., Martinelli, V., Spuler, S., Bressi, S., Scotti, G., and Comi, G. (1996b). Comparison of triple dose versus standard dose gadolinium-DTPA for detection of MRI enhancing lesions in patients with MS. Neurology 46, 379–384. Filippi, M., Rocca, M. A., Colombo, B., Falini, A., Codella, M., Scotti, G., and Comi, G. (2002). Functional magnetic resonance imaging correlates of fatigue in multiple sclerosis. Neuroimage. 15, 559–567. Fu, L., Matthews, P. M., De Stefano, N., Worsley, K. J., Narayanan, S., Francis, G. S., Antel, J. P., Wolfson, C., and Arnold, D. L. (1998). Imaging axonal damage of normal-appearing white matter in multiple sclerosis. Brain 121 (Pt. 1), 103–113. Gadian, D. G. (1995). ‘‘NMR and Its Applications to Living Systems.’’ Oxford University Press, Oxford, UK. Ganter, P., Prince, C., and Esiri, M. M. (1999). Spinal cord axonal loss in multiple sclerosis: a post-mortem study. Neuropathol. Appl. Neurobiol. 25, 459–467. Gareau, P. J., Rutt, B. K., Karlik, S. J., and Mitchell, J. R. (2000). Magnetization transfer and multicomponent T2 relaxation measurements with histopathologic correlation in an experimental model of MS. J. Magn Reson. Imaging 11, 586–595. Gawne-Cain, M. L., O’Riordan, J. I., Thompson, A. J., Moseley, I. F., and Miller, D. H. (1997). Multiple sclerosis lesion detection in the brain: a comparison of fast Xuid-attenuated inversion recovery and conventional T2-weighted dual spin echo. Neurology 49, 364–370. Ge, Y., Grossman, R. I., Udupa, J. K., Babb, J. S., Mannon, L. J., and McGowan, J. C. (2002). Magnetization transfer ratio histogram analysis of normal-appearing gray matter and normal-appearing white matter in multiple sclerosis. J. Comput. Assist. Tomogr. 26, 62–68. Gean-Marton, A. D., Vezina, L. G., Marton, K. I., Stimac, G. K., Peyster, R. G., Taveras, J. M., and Davis, K. R. (1991). Abnormal corpus callosum: a sensitive and speciWc indicator of multiple sclerosis. Radiology 180, 215–221.

785

786

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

Goodkin, D. E., Rooney, W. D., Sloan, R., Bacchetti, P., Gee, L., Vermathen, M., Waubant, E., Abundo, M., Majumdar, S., Nelson, S., and Weiner, M. W. (1998). A serial study of new MS lesions and the white matter from which they arise. Neurology 51, 1689–1697. Grossman, R. I. (1999). Application of magnetization transfer imaging to multiple sclerosis. Neurology 53, S8–11. Grossman, R. I., BraVman, B. H., Brorson, J. R., Goldberg, H. I., Silberberg, D. H., and Gonzalez-Scarano, F. (1988). Multiple sclerosis: serial study of gadolinium-enhanced MR imaging. Radiology 169, 117–122. Guo, A. C., Jewells, V. L., and Provenzale, J. M. (2001). Analysis of normal-appearing white matter in multiple sclerosis: comparison of diVusion tensor MR imaging and magnetization transfer imaging. AJNR Am. J. Neuroradiol. 22, 1893–1900. Guttmann, C. R., Ahn, S. S., Hsu, L., Kikinis, R., and Jolesz, F. A. (1995). The evolution of multiple sclerosis lesions on serial MR. AJNR Am. J. Neuroradiol. 16, 1481–1491. Harris, J. O., Frank, J. A., Patronas, N., McFarlin, D. E., and McFarland, H. F. (1991). Serial gadoliniumenhanced magnetic resonance imaging scans in patients with early, relapsing-remitting multiple sclerosis: implications for clinical trials and natural history. Ann. Neurol. 29, 548–555. Hashemi, R. H., and Bradley, W. G. (1997). ‘‘MRI: The Basics.’’ Lippincott, Williams & Wilkins, Baltimore. Hawkins, C. P., Munro, P. M., Mackenzie, F., Kesselring, J., Tofts, P. S., Du Boulay, E. P., Landon, D. N., and McDonald, W. I. (1990). Duration and selectivity of blood-brain barrier breakdown in chronic relapsing experimental allergic encephalomyelitis studied by gadolinium-DTPA and protein markers. Brain 113 (Pt .2), 365–378. Hershey, L. A., Gado, M. H., and Trotter, J. L. (1979). Computerized tomography in the diagnostic evaluation of multiple sclerosis. Ann. Neurol. 5, 32–39. Hickman, S. J., Brierley, C. M., Brex, P. A., MacManus, D. G., Scolding, N. J., Compston, D. A., and Miller, D. H. (2002). Continuing optic nerve atrophy following optic neuritis: a serial MRI study. Mult. Scler. 8, 339–342. Horowitz, A. L., Kaplan, R. D., Grewe, G., White, R. T., and Salberg, L. M. (1989). The ovoid lesion: A new MR observation in patients with multiple sclerosis. AJNR Am. J. Neuroradiol. 10, 303–305. HorsWeld, M. A., Lai, M., Webb, S. L., Barker, G. J., Tofts, P. S., Turner, R., Rudge, P., and Miller, D. H. (1996). Apparent diVusion coeYcients in benign and secondary progressive multiple sclerosis by nuclear magnetic resonance. Magn Reson. Med. 36, 393–400. Husted, C. A., Goodin, D. S., Hugg, J. W., Maudsley, A. A., Tsuruda, J. S., de Bie, S. H., Fein, G., Matson, G. B., and Weiner, M. W. (1994). Biochemical alterations in multiple sclerosis lesions and normal-appearing white matter detected by in vivo 31P and 1H spectroscopic imaging. Ann. Neurol. 36, 157–165. Jacobs, K. M., and Donoghue, J. P. (1991). Reshaping the cortical motor map by unmasking latent intracortical connections. Science 251, 944–947. Jezzard, P., Matthews, P. M., and Smith, S. (Eds.). (2001). ‘‘Functional Magnetic Resonance Imaging: An Introduction to Methods.’’ Oxford: Oxford University Press. Jones, T. A., and Schallert, T. (1992). Overgrowth and pruning of dendrites in adult rats recovering from neocortical damage. Brain Res. 581, 156–160. Kapeller, P., McLean, M. A., GriYn, C. M., Chard, D., Parker, G. J., Barker, G. J., Thompson, A. J., and Miller, D. H. (2001). Preliminary evidence for neuronal damage in cortical grey matter and normal appearing white matter in short duration relapsing-remitting multiple sclerosis: A quantitative MR spectroscopic imaging study. J. Neurol. 248, 131–138. Katz, D., Taubenberger, J., Raine, C., McFarlin, D., and McFarland, H. (1990). Gadolinium-enhancing lesions on magnetic resonance imaging: Neuropathological Wndings. Ann. Neurol. 28, 243. Katz, D., Taubenberger, J. K., Cannella, B., McFarlin, D. E., Raine, C. S., and McFarland, H. F. (1993). Correlation between magnetic resonance imaging Wndings and lesion development in chronic, active multiple sclerosis. Ann. Neurol. 34, 661–669. Kermode, A. G., Thompson, A. J., Tofts, P., MacManus, D. G., Kendall, B. E., Kingsley, D. P., Moseley, I. F., Rudge, P., and McDonald, W. I. (1990a). Breakdown of the blood-brain barrier precedes symptoms and other MRI signs of new lesions in multiple sclerosis. Pathogenetic and clinical implications. Brain 113 (Pt. 5), 1477–1489. Kermode, A. G., Tofts, P. S., Thompson, A. J., MacManus, D. G., Rudge, P., Kendall, B. E., Kingsley, D. P., Moseley, I. F., Du Boulay, E. P., and McDonald, W. I. (1990b). Heterogeneity of blood-brain barrier changes in multiple sclerosis: An MRI study with gadolinium-DTPA enhancement. Neurology 40, 229–235. Khoury, S. J., Guttmann, C. R., Orav, E. J., Hohol, M. J., Ahn, S. S., Hsu, L., Kikinis, R., Mackin, G. A., Jolesz, F. A., and Weiner, H. L. (1994). Longitudinal MRI in multiple sclerosis: Correlation between disability and lesion burden. Neurology 44, 2120–2124. Kidd, D., Barkhof, F., McConnell, R., Algra, P. R., Allen, I. V., and Revesz, T. (1999). Cortical lesions in multiple sclerosis. Brain 122 (Pt. 1), 17–26. Koudriavtseva, T., Thompson, A. J., Fiorelli, M., Gasperini, C., Bastianello, S., Bozzao, A., Paolillo, A., Pisani, A., Galgani, S., and Pozzilli, C. (1997). Gadolinium enhanced MRI predicts clinical and MRI disease activity in relapsing-remitting multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 62, 285–287. Kozlowski, D. A., James, D. C., and Schallert, T. (1996) Use-dependent exaggeration of neuronal injury after unilateral sensorimotor cortex lesions. J. Neurosci. 16, 4776–4786. Laule, C., Vavasour, I. M., Oger, J., Paty, D. W., Li, D. K. B., and MacKay, A. L. (2002). T2 relaxation measurements of in-vivo water content and myelin water content in normal appearing white matter and lesions in multiple sclerosis. Proc. Intl. Soc. Mag. Reson. Med. 10, 185.

au3

CONCLUSIONS

au4

Leary, S. M., Davie, C. A., Parker, G. J., Stevenson, V. L., Wang, L., Barker, G. J., Miller, D. H., and Thompson, A. J. (1999). 1H magnetic resonance spectroscopy of normal appearing white matter in primary progressive multiple sclerosis. J. Neurol. 246, 1023–1026. Lee, M., Reddy, H., Johansen-Berg, H., Pendlebury, S., Jenkinson, M., Smith, S., Palace, J., and Matthews, P. M. (2000). The motor cortex shows adaptive functional changes to brain injury from multiple sclerosis. Ann. Neurol. 47, 606–613. Lee, M. A., Smith, S., Palace, J., Narayanan, S., Silver, N., Minicucci, L., Filippi, M., Miller, D. H., Arnold, D. L., and Matthews, P. M. (1999). Spatial mapping of T2 and gadolinium-enhancing T1 lesion volumes in multiple sclerosis: evidence for distinct mechanisms of lesion genesis? Brain 122 (Pt. 7), 1261–1270. Leist, T. P., Gobbini, M. I., Frank, J. A., and McFarland, H. F. (2001). Enhancing magnetic resonance imaging lesions and cerebral atrophy in patients with relapsing multiple sclerosis. Arch. Neurol. 58, 57–60. Levesque, I., Sled J. G., Brass S. D., Santos, A. C., Narayanan, S., Francis, S. J., Arnold, D. L., and Pike, G. B. (2002). Comparison of MTR and qMT imaging of multiple sclerosis lesions. Proc. Int. Soc. Magn. Reson. Med. 10, 1183 Lin, X., and Blumhardt, L. D. (2001). InXammation and atrophy in multiple sclerosis: MRI associations with disease course. J. Neurol. Sci. 189, 99–104. LosseV, N. A., Kingsley, D. P., McDonald, W. I., Miller, D. H., and Thompson, A. J. (1996a). Clinical and magnetic resonance imaging predictors of disability in primary and secondary progressive multiple sclerosis. Mult. Scler. 1, 218–222. LosseV, N. A., Wang, L., Lai, H. M., Yoo, D. S., Gawne-Cain, M. L., McDonald, W. I., Miller, D. H., and Thompson, A. J. (1996b). Progressive cerebral atrophy in multiple sclerosis. A serial MRI study. Brain 119 (Pt. 6), 2009–2019. Lumsden, C. E. (1970). The neuropathology of multiple sclerosis. In ‘‘Multiple Sclerosis and Other Demyelinating Diseases. Handbook of Clinical Neurology’’ (P. J. Vinken and G. W. Bruyn, eds.), Vol. 9, pp. 217–309. NorthHolland, Amsterdam. MacKay, A., Whittall, K., Adler, J., Li, D., Paty, D., and Graeb, D. (1994). In vivo visualization of myelin water in brain by magnetic resonance. Magn Reson. Med. 31, 673–677. Matthews, P. M. (2001). An introduction to functional magnetic resonance imaging. In ‘‘Functional Magnetic Resonance Imaging: An Introduction to Methods’’ (P., Jezzard, P. M. Matthews, and S. Smith, Eds.). Oxford: Oxford University Press. Matthews, P. M., and Arnold, D. L. (2001). Magnetic resonance imaging of multiple sclerosis: new insights linking pathology to clinical evolution. Curr. Opin. Neurol. 14, 279–287. Matthews, P. M., Pioro, E., Narayanan, S., De Stefano, N., Fu, L., Francis, G., Antel, J., Wolfson, C., and Arnold, D. L. (1996). Assessment of lesion pathology in multiple sclerosis using quantitative MRI morphometry and magnetic resonance spectroscopy. Brain 119 (Pt. 3), 715–722. Matthews, P. M., De Stefano, N., Narayanan, S., Francis, G. S., Wolinsky, J. S., Antel, J. P., and Arnold, D. L. (1998). Putting magnetic resonance spectroscopy studies in context: Axonal damage and disability in multiple sclerosis. Semin. Neurol. 18, 327–336. McDonald, W. I., Compston, A., Edan, G., Goodkin, D., Hartung, H. P., Lublin, F. D., McFarland, H. F., Paty, D. W., Polman, C. H., Reingold, S. C., Sandberg-Wollheim, M., Sibley, W., Thompson, A., van den, N. S., Weinshenker, B. Y., and Wolinsky, J. S. (2001). Recommended diagnostic criteria for multiple sclerosis: Guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann. Neurol. 50, 121–127. McFarland, H. F., Frank, J. A., Albert, P. S., Smith, M. E., Martin, R., Harris, J. O., Patronas, N., Maloni, H., and McFarlin, D. E. (1992). Using gadolinium-enhanced magnetic resonance imaging lesions to monitor disease activity in multiple sclerosis. Ann. Neurol. 32, 758–766. McGowan, J. C. (1999). The physical basis of magnetization transfer imaging. Neurology 53, S3-S7. Miller, D. H., Barkhof, F., Frank, J. A., Parker, G. J., and Thompson, A. J. (2002). Measurement of atrophy in multiple sclerosis: Pathological basis, methodological aspects and clinical relevance. Brain 125, 1676–1695. Miller, D. H., Barkhof, F., and Nauta, J. J. (1993). Gadolinium enhancement increases the sensitivity of MRI in detecting disease activity in multiple sclerosis. Brain 116 (Pt. 5), 1077–1094. Miller, D. H., Grossman, R. I., Reingold, S. C., and McFarland, H. F. (1998). The role of magnetic resonance techniques in understanding and managing multiple sclerosis. Brain 121 (Pt. 1), 3–24. Miller, D. H., Johnson, G., Tofts, P. S., MacManus, D., and McDonald, W. I. (1989). Precise relaxation time measurements of normal-appearing white matter in inXammatory central nervous system disease. Magn Reson. Med. 11, 331–336. Miller, D. H., Kesselring, J., McDonald, W. I., Paty, D. W., and Thompson, A. J. (1997). ‘‘Magnetic Resonance in Multiple Sclerosis.’’ Cambridge University Press, Cambridge, United Kingdom. Miller, D. H., Rudge, P., Johnson, G., Kendall, B. E., MacManus, D. G., Moseley, I. F., Barnes, D., and McDonald, W. I. (1988) Serial gadolinium enhanced magnetic resonance imaging in multiple sclerosis. Brain 111 (Pt. 4), 927–939. MoVett, J. R., Namboodiri, M. A., Cangro, C. B., and Neale, J. H. (1991). Immunohistochemical localization of N-acetylaspartate in rat brain. Neuroreport 2, 131–134. Moore, G. R., Leung, E., MacKay, A. L., Vavasour, I. M., Whittall, K. P., Cover, K. S., Li, D. K., Hashimoto, S. A., Oger, J., Sprinkle, T. J., and Paty, D. W. (2000). A pathology-MRI study of the short-T2 component in formalin-Wxed multiple sclerosis brain. Neurology 55, 1506–1510.

787

788

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

au5

Narayana, P. A., Doyle, T. J., Lai, D., and Wolinsky, J. S. (1998). Serial proton magnetic resonance spectroscopic imaging, contrast-enhanced magnetic resonance imaging, and quantitative lesion volumetry in multiple sclerosis. Ann. Neurol. 43, 56–71. Nesbit, G. M., Forbes, G. S., Scheithauer, B. W., Okazaki, H., and Rodriguez, M. (1991). Multiple sclerosis: Histopathologic and MR and/or CT correlation in 37 cases at biopsy and three cases at autopsy. Radiology 180, 467–474. Newcombe, J., Hawkins, C. P., Henderson, C. L., Patel, H. A., Woodroofe, M. N., Hayes, G. M., Cuzner, M. L., MacManus, D., Du Boulay, E. P., and McDonald, W. I. (1991). Histopathology of multiple sclerosis lesions detected by magnetic resonance imaging in unWxed postmortem central nervous system tissue. Brain 114 (Pt. 2), 1013–1023. Nijeholt, G. J., van Walderveen, M. A., Castelijns, J. A., van Waesberghe, J. H., Polman, C., Scheltens, P., Rosier, P. F., Jongen, P. J., and Barkhof, F. (1998). Brain and spinal cord abnormalities in multiple sclerosis. Correlation between MRI parameters, clinical subtypes and symptoms. Brain 121 (Pt. 4), 687–697. Nudo, R. J., Wise, B. M., SiFuentes, F., and Milliken, G. W. (1996). Neural substrates for the eVects of rehabilitative training on motor recovery after ischemic infarct. Science 272, 1791–1794. Ogawa, S., Menon, R. S., Tank, D. W., Kim, S. G., Merkle, H., Ellermann, J. M., and Ugurbil, K. (1993). Functional brain mapping by blood oxygenation level-dependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model. Biophys. J. 64, 803–812. Ormerod, I. E., Johnson, G., MacManus, D., Du Boulay, E. P., and McDonald, W. I. (1986). Relaxation times of apparently normal cerebral white matter in multiple sclerosis. Acta Radiol. Suppl 369, 382–384. Pantano, P., Iannetti, G. D., Caramia, F., Mainero, C., Di Legge, S., Bozzao, L., Pozzilli, C., and Lenzi, G. L. (2002). Cortical motor reorganization after a single clinical attack of multiple sclerosis. Brain 125, 1607–1615. Paolillo, A., Pozzilli, C., Gasperini, C., Giugni, E., Mainero, C., Giuliani, S., Tomassini, V., MilleWorini, E., and Bastianello, S. (2000). Brain atrophy in relapsing-remitting multiple sclerosis: relationship with ‘black holes’, disease duration and clinical disability. J. Neurol. Sci. 174, 85–91. Parry, A., Clare, S., Jenkinson, M., Smith, S., Palace, J., and Matthews, P. M. (2002a). White matter and lesion T1 relaxation times increase in parallel and correlate with disability in multiple sclerosis. J. Neurol. (in press). Parry, A. M. M., Scott, R. B., Palace, J., Smith, S. M., and Matthews, P. M. (2002b). Potentially adaptive functional changes for cognitive processing in multiple sclerosis. Submitted for publication. Paty, D. W., Oger, J. J., KastrukoV, L. F., Hashimoto, S. A., Hooge, J. P., Eisen, A. A., Eisen, K. A., Purves, S. J., Low, M. D., Brandejs, V., Robertson, W. D. and Li, D. K. B. (1988). MRI in the diagnosis of MS: A prospective study with comparison of clinical evaluation, evoked potentials, oligoclonal banding, and CT. Neurology 38, 180–185. Peterson, J. W., Bo, L., Mork, S., Chang, A., and Trapp, B. D. (2001). Transected neurites, apoptotic neurons, and reduced inXammation in cortical multiple sclerosis lesions. Ann. Neurol. 50, 389–400. Pike, G. B., De Stefano, N., Narayanan, S., Worsley, K. J., Pelletier, D., Francis, G. S., Antel, J. P., and Arnold, D. L. (2000). Multiple sclerosis: Magnetization transfer MR imaging of white matter before lesion appearance on T2-weighted images. Radiology 215, 824–830. Plumb, J., McQuaid, S., Mirakhur, M., and Kirk, J. (2002). Abnormal endothelial tight junctions in active lesions and normal-appearing white matter in multiple sclerosis. Brain Pathol. 12, 154–169. Presciutti, O., Sarchielli, P., Gobbi, G., Tarducci, R., Pelliccioli, G. P., Chiarini, P., Gentile, E., and Gallai, V. (2000). 1H MRS study in occipital gray matter of multiple sclerosis patients. Proc. Int. Soc. Magn. Reson. Med. 8, 627. Ramani, A., Barker, G. J., and Tofts, P. S. (2001). Fast measurement of quantitative MT parameters in Wxed multiple sclerosis brain. Proc. Int. Soc. Magn. Reson. Med. 9, 259. Reddy, H., Bendahan, D., Lee, M. A., Johansen-Berg, H., Donaghy, M., Hilton-Jones, D., and Matthews, P. M. (2002c). An expanded cortical representation for hand movement after peripheral motor denervation. J. Neurol. Neurosurg. Psychiatry. 72, 203–210. Reddy, H., De Stefano, N., Mortilla, M., Federico, A., and Matthews, P. M. (2002a). Functional reorganization of motor cortex increases with greater axonal injury from CADASIL. Stroke 33, 502–508. Reddy, H., Floyer, A., Donaghy, M., and Matthews, P. M. (2001). Altered cortical activation with Wnger movement after peripheral denervation: comparison of active and passive tasks. Exp. Brain Res. 138, 484–491. Reddy, H., Narayanan, S., Arnoutelis, R., Jenkinson, M., Antel, J., Matthews, P. M., and Arnold, D. L. (2000a). Evidence for adaptive functional changes in the cerebral cortex with axonal injury from multiple sclerosis. Brain 123 (Pt. 11), 2314–2320. Reddy, H., Narayanan, S., Matthews, P. M., Hoge, R. D., Pike, G. B., Duquette, P., Antel, J., and Arnold, D. L. (2000b). Relating axonal injury to functional recovery in MS. Neurology 54, 236–239. Reddy, H., Narayanan, S., Woolrich, M., Mitsumori, T., Lapierre, Y., Arnold, D. L., and Matthews, P. M. (2002b). Functional brain reorganisation for hand movements in patients with multiple sclerosis: deWning distinct eVects of injury and disability. Brain (in press). Rieckmann, P., and Smith, K. J. (2001). Multiple sclerosis: More than inXammation and demyelination. Trends Neurosci. 24, 435–437. Rocca, M. A., Cercignani, M., Iannucci, G., Comi, G., and Filippi, M. (2000). Weekly diVusion-weighted imaging of normal-appearing white matter in MS. Neurology 55, 882–884. Rocca, M. A., Falini, A., Colombo, B., Scotti, G., Comi, G., and Filippi, M. (2002a) Adaptive functional changes in the cerebral cortex of patients with nondisabling multiple sclerosis correlate with the extent of brain structural damage. Ann. Neurol. 51, 330–339.

CONCLUSIONS

au6

Rocca, M. A., Matthews, P. M., Caputo, D., Ghezzi, A., Falini, A., Scotti, G., Comi, G., and Filippi, M. (2002b). Evidence for widespread movement-associated functional MRI changes in patients with PPMS. Neurology 58, 866–872. Ropele, S., Strasser-Fuchs, S., Seifert, T., Enzinger, C., and Fazekas, F. (2002). Development of active multiple sclerosis lesions: A quantitative MT study. Proc. Int. Soc. Magn. Reson. Med. 10, 180. Rovaris, M., and Filippi, M. (1999). Magnetic resonance techniques to monitor disease evolution and treatment trial outcomes in multiple sclerosis. Curr. Opin. Neurol. 12, 337–344. Rovaris, M., and Filippi, M. (2000a). Contrast enhancement and the acute lesion in multiple sclerosis. Neuroimaging Clin. N. Am. 10, 705–716. Rovaris, M., and Filippi, M. (2000b). The value of new magnetic resonance techniques in multiple sclerosis. Curr. Opin. Neurol. 13, 249–254. Rudick, R. A., Fisher, E., Lee, J. C., Simon, J., and Jacobs, L. (1999) Use of the brain parenchymal fraction to measure whole brain atrophy in relapsing-remitting MS. Multiple Sclerosis Collaborative Research Group. Neurology 53, 1698–1704. Rudkin, T. M., and Arnold, D. L. (1999). Proton magnetic resonance spectroscopy for the diagnosis and management of cerebral disorders. Arch. Neurol. 56, 919–926. Saindane, A. M., Ge, Y., Udupa, J. K., Babb, J. S., Mannon, L. J., and Grossman, R. I. (2000). The eVect of gadolinium-enhancing lesions on whole brain atrophy in relapsing-remitting MS. Neurology 55, 61–65. Salibi, N., and Brown, M. A. (1998). ‘‘Clinical MR Spectroscopy. First Principles.’’ Wiley-Liss: New York. Santos, A. C., Sled, J. G., Narayanan, S., Francis, S. J., Brass, S. D., Levesque, I., Caramanos, Z., Pike, G. B., and Arnold, D. L. (2002). Quantitative magnetisation transfer for assessment of tissue damage in individual multiple sclerosis lesions. Proc. Int. Soc. Magn. Reson. Med. 10, 1177. Sarchielli, P., Presciutti, O., Pelliccioli, G. P., Tarducci, R., Gobbi, G., Chiarini, P., Alberti, A., Vicinanza, F., and Gallai, V. (1999). Absolute quantiWcation of brain metabolites by proton magnetic resonance spectroscopy in normal-appearing white matter of multiple sclerosis patients. Brain 122 (Pt. 3), 513–521. Siger-Zajdel, M., and Selmaj, K. (2001). Magnetisation transfer ratio analysis of normal appearing white matter in patients with familial and sporadic multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 71, 752–756. Silver, N. C., Good, C. D., Barker, G. J., MacManus, D. G., Thompson, A. J., Moseley, I. F., McDonald, W. I., and Miller, D. H. (1997). Sensitivity of contrast enhanced MRI in multiple sclerosis. EVects of gadolinium dose, magnetization transfer contrast and delayed imaging. Brain 120 (Pt. 7), 1149–1161. Simmons, M. L., Frondoza, C. G., and Coyle, J. T. (1991). Immunocytochemical localization of N-acetylaspartate with monoclonal antibodies. Neuroscience 45, 37–45. Simon, J. H. (2000). The contribution of spinal cord MRI to the diagnosis and diVerential diagnosis of multiple sclerosis. J. Neurol. Sci. 172, Suppl 1, S32–S35. Simon, J. H., Jacobs, L. D., Campion, M., Wende, K., Simonian, N., Cookfair, D. L., Rudick, R. A., Herndon, R. M., Richert, J. R., Salazar, A. M., Alam, J. J., Fischer, J. S., Goodkin, D. E., Granger, C. V., Lajaunie, M., Martens-Davidson, A. L., Meyer, M., Sheeder, J., Choi, K., Scherzinger, A. L., Bartoszak, D. M., Bourdette, D. N., Braiman, J., Brownscheidle, C. M., and Whitham, R. H. (1998). Magnetic resonance studies of intramuscular interferon beta-1a for relapsing multiple sclerosis. The Multiple Sclerosis Collaborative Research Group. Ann. Neurol. 43, 79–87. Smith, M. E., Stone, L. A., Albert, P. S., Frank, J. A., Martin, R., Armstrong, M., Maloni, H., McFarlin, D. E., and McFarland, H. F. (1993). Clinical worsening in multiple sclerosis is associated with increased frequency and area of gadopentetate dimeglumine-enhancing magnetic resonance imaging lesions. Ann. Neurol. 33, 480–489. Stanisz, G. J., and Webb, S. (2002). Understanding the MT, T1 and T2 changes during demyelination and inXammation. Proc. Int. Soc. Magn. Reson. Med. 10, 183. Stevenson, V. L., Smith, S. M., Matthews, P. M., Miller, D. H., and Thompson, A. J. (2002). Monitoring disease activity and progression in primary progressive multiple sclerosis using MRI: sub-voxel registration to identify lesion changes and to detect cerebral atrophy. J. Neurol. 249, 171–177. Stone, L. A., Smith, M. E., Albert, P. S., Bash, C. N., Maloni, H., Frank, J. A., and McFarland, H. F. (1995). Blood-brain barrier disruption on contrast-enhanced MRI in patients with mild relapsing-remitting multiple sclerosis: Relationship to course, gender, and age. Neurology 45, 1122–1126. Tas, M. W., Barkhol, F., van Walderveen, M. A., Polman, C. H., Hommes, O. R., and Valk, J. (1995). The eVect of gadolinium on the sensitivity and speciWcity of MR in the initial diagnosis of multiple sclerosis. AJNR Am. J. Neuroradiol. 16, 259–264. Tedeschi, G., Bonavita, S., McFarland, H. F., Richert, N., Duyn, J. H., and Frank, J. A. (2002). Proton MR spectroscopic imaging in multiple sclerosis. Neuroradiology 44, 37–42. Thompson, A. J., Miller, D., Youl, B., MacManus, D., Moore, S., Kingsley, D., Kendall, B., Feinstein, A., and McDonald, W. I. (1992). Serial gadolinium-enhanced MRI in relapsing/remitting multiple sclerosis of varying disease duration. Neurology 42, 60–63. Tintore, M., Rovira, A., Martinez, M. J., Rio, J., Diaz-Villoslada, P., Brieva, L., Borras, C., Grive, E., Capellades, J., and Montalban, X. (2000). Isolated demyelinating syndromes: Comparison of diVerent MR imaging criteria to predict conversion to clinically deWnite multiple sclerosis. AJNR Am. J. Neuroradiol. 21, 702–706. Tofts, P. S., and Kermode, A. G. (1991). Measurement of the blood-brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson. Med. 17, 357–367. Trapp, B. D., Peterson, J., RansohoV, R. M., Rudick, R., Mork, S., and Bo, L. (1998). Axonal transection in the lesions of multiple sclerosis. N. Engl. J. Med. 338, 278–285.

789

790

32. MRI VISUALIZATION OF MULTIPLE SCLEROSIS

Truyen, L., van Waesberghe, J. H., van Walderveen, M. A., van Oosten, B. W., Polman, C. H., Hommes, O. R., Ader, H. J., and Barkhof, F. (1996). Accumulation of hypointense lesions (‘‘black holes’’) on T1 spin-echo MRI correlates with disease progression in multiple sclerosis. Neurology 47, 1469–1476. Tsai, G., and Coyle, J. T. (1995). N-acetylaspartate in neuropsychiatric disorders. Prog. Neurobiol. 46, 531–540. Tseng, G. F., and Hu, M. E. (1996). Axotomy induces retraction of the dendritic arbor of adult rat rubrospinal neurons. Acta Anat.(Basel) 155, 184–193. Uhlenbrock, D., and Sehlen, S. (1989). The value of T1-weighted images in the diVerentiation between MS, white matter lesions, and subcortical arteriosclerotic encephalopathy (SAE). Neuroradiology 31, 203–212. van Waesberghe, J. H., Kamphorst, W., De Groot, C. J., van Walderveen, M. A., Castelijns, J. A., Ravid, R., Nijeholt, G. J., van, d., V, Polman, C. H., Thompson, A. J., and Barkhof, F. (1999). Axonal loss in multiple sclerosis lesions: Magnetic resonance imaging insights into substrates of disability. Ann. Neurol. 46, 747–754. van Walderveen, M. A., Kamphorst, W., Scheltens, P., van Waesberghe, J. H., Ravid, R., Valk, J., Polman, C. H., and Barkhof, F. (1998). Histopathologic correlate of hypointense lesions on T1-weighted spin-echo MRI in multiple sclerosis. Neurology 50, 1282–1288. Waxman, S. G. (2001). Acquired channelopathies in nerve injury and MS. Neurology 56, 1621–1627. Webb, S., Munro, C. A., Midha, R., and Stanisz, G. J. (2002). Is quantitative T2 a good measure of myelin content in white matter pathologies? Proc. Int. Soc. Magn. Reson. Med. 10, 184. Werring, D. J., Bullmore, E. T., Toosy, A. T., Miller, D. H., Barker, G. J., MacManus, D. G., Brammer, M. J., Giampietro, V. P., Brusa, A., Brex, P. A., Moseley, I. F., Plant, G. T., McDonald, W. I., and Thompson, A. J. (2000). Recovery from optic neuritis is associated with a change in the distribution of cerebral response to visual stimulation: a functional magnetic resonance imaging study. J. Neurol. Neurosurg. Psychiatry 68, 441–449. Werring, D. J., Clark, C. A., Barker, G. J., Thompson, A. J., and Miller, D. H. (1999). DiVusion tensor imaging of lesions and normal-appearing white matter in multiple sclerosis. Neurology 52, 1626–1632. Werring, D. J., Clark, C. A., Droogan, A. G., Barker, G. J., Miller, D. H., and Thompson, A. J. (2001). Water diVusion is elevated in widespread regions of normal-appearing white matter in multiple sclerosis and correlates with diVusion in focal lesions. Mult. Scler. 7, 83–89. Wujek, J. R., Bjartmar, C., Richer, E., RansohoV, R. M., Yu, M., Tuohy, V. K., and Trapp, B. D. (2002). Axon loss in the spinal cord determines permanent neurological disability in an animal model of multiple sclerosis. J. Neuropathol. Exp. Neurol. 61, 23–32. Wylezinska, M., Cifelli, A., Jezzard, P., Palace, J., Alecci, M., and Matthews, P. M. (2002a). Thalamic neurodegeneration in relapsing-remitting multiple sclerosis Neurology 60, 1949–1959. Wylezinska, M., Cifelli, A., Matthews, P., Palace, J., and Jezzard, P. (2002b). Neuronal damage in thalamic grey matter in relapsing-remitting multiple sclerosis. Proc. Int. Soc. Magn. Reson. Med. 10, 592. Zorzon, M., De Masi, R., Nasuelli, D., Ukmar, M., Mucelli, R. P., Cazzato, G., Bratina, A., and Zivadinov, R. (2001). Depression and anxiety in multiple sclerosis. A clinical and MRI study in 95 subjects. J. Neurol. 248, 416–421.

C H A P T E R

33 Multiple Sclerosis: Therapy au1

Jack Antel and Amit Bar-Or

INTRODUCTION

au2

au3

The clinical and pathologic features of multiple sclerosis (MS) began to be described in the mid-1800s and by the 1870s had been synthesized into a recognizable entity by Charcot and colleagues. Charcot emphasized the loss of the myelin sheath with relative, but not absolute, preservation of axons. He referred to the observation by ReinXeisch in 1863 of inXammation around a vessel in the center of MS plaques; this can be viewed as the beginning of the continuing debate of the relative contributions of immune mediated versus ‘‘neurodegenerative’’ processes as a basis for the disease pathology (Hoeber, 1922; Murray, 2000). As reviewed in other chapters, since that time there have been major advances in our understanding of the clinical, pathologic and imaging aspects of the disease. Charcot had stated in his lectures that the time had not yet come to consider therapy for the disorder (Charcot, 1877). Over the ensuing years, multiple therapeutic interventions were attempted, often based on the ideas of the time regarding disease pathogenesis or on availability of therapies for other diseases. The National Multiple Sclerosis Society, through the various editions of its publication Therapeutic Claims in Multiple Sclerosis, has compiled a comprehensive list of such therapies (National Multiple Sclerosis Society, 1982; Therapeutic Claims in Multiple Sclerosis, 1992). Expert committees were used to evaluate therapeutic claims in the era that preceded the controlled clinical trial supported by magnetic resonance (MR) based imaging. Examples of therapies listed as in use prior to 1935 include arsenic, belladonna, fever therapy, and hypnotism; between 1935 and 1950, dicoumaral, histamine, and vitamins B12, D, E, and K were among those mentioned. Theories of disease pathogenesis that generated associated therapies between 1950 and 1965 included those directed at nutritional status (including vitamin supplementation), metabolic status (with carbohydrate and fat supplements), vascular abnormalities (using either anticoagulants or vasodilators), and infectious etiologies (including antispirochete therapies). None of these interventions were judged to have demonstrated eYcacy. Candidate infectious agents such as chlamydia, Epstein-Barr virus, and human herpes virus type 6 continue to be evaluated (Kaufman et al., 2002; Moore et al., 2002; Talbot et al., 2001). The postulate that immune-mediated mechanisms underlie the pathogenesis of MS has been raised since the initial descriptions of MS and has dominated thinking in the past several decades, as discussed in other chapters. As they became available, an array au4 of anti-inXammatory (e.g., corticosteroids) and immune-suppressive agents (e.g., cyclophosphamide) were tested on patients with MS (reviewed in Antel et al., 2003, and Smith et al., 1998). The results of modern clinical/MRI trials indicate the limitations of the study designs used in these early studies—for example, length of study, number of patients,

Myelin Biology and Disorders, Volume 2

791

Copyright 2004, Elsevier (USA). All rights reserved.

792

33. MULTIPLE SCLEROSIS: THERAPY

phase of disease, placebo control, primary end point. A number of the immune modulators for which there were uncertain conclusions in the 1982 edition of MS Therapeutic Claims, such as azathioprine and cyclophosphamide, have been and are being reexamined in light of results with current therapies and insights regarding the immunobiology of MS. Some therapies have been abandoned for toxicity reasons (lymphoid irradiation), although greater risks may now be acceptable in speciWc patient groups (e.g., those who fail currently approved therapies.) The major themes considered in this chapter relate to deWning the relationship between the clinical phenotypes of MS and their underlying immunobiologic and neurobiologic substrates and explaining results of or predicting outcomes of past, current, and future clinical trials in the context of the underlying disease process. Although many of the clinical features of MS were described in the early writings, recent natural history and large-scale epidemiologic studies have provided a more solid source of information, as reviewed in separate chapters in this book. The insights into immunobiology have been derived from pathologic analyses of MS tissues, in vivo imaging, largely using MR techniques, and animal models. All of these are also reviewed in separate chapters. The results of clinical trials conducted on well-deWned subtypes of MS provide further insights and identify new challenges regarding the biologic substrates that underlie the various phases of the MS disease process.

IMMUNOBIOLOGY OF THE CLINICAL FEATURES OF MS

au5

As described in chapter 29, the clinical phenotypes can be summarized into four major categories although overlaps exist. These include relapsing-remitting (RR), secondary progressive (SP), primary progressive (PP), and progressive-relapsing.(PR). Natural history studies indicate that more than 50% of RR cases will evolve into the SP phenotype over about 15 years with such evolution being almost inevitable once Wxed motor deWcits become evident (reviewed in (Paty and Ebers, 1997) ). A meaningful number, albeit a minority of MS patients do, however, follow a relatively benign course, with autopsydeWned cases being reported in which no clinical manifestations were ever evident ( (McAlpine, 1961; Paty and Ebers, 1997) ).

Clinically Isolated Syndrome (CIS) Most cases of MS begin with a discrete neurologic event consistent with demyelination within the CNS. Common examples would include optic neuritis, brain stem dysfunction, or transverse myelitis. If the individual is seen at this time, the condition is referred to as a clinically isolated syndrome (CIS). CIS may occur with or with or without multifocal lesions on MRI. The MRI Wndings are now shown to be very signiWcant predictors of future disease course. The risk of recurrent disease in cases of CIS with multifocal MRI abnormalities approaches 80% over the subsequent 10 years but is only 5 to 10% in cases with normal MRIs (Brex et al., 2002). New deWnitions of MS attempt to incorporate recurrent MR-deWned lesions as being suYcient to accept the diagnosis of the disease (McDonald et al., 2001). An unresolved question is whether CIS with and without multifocal lesions have a common underlying pathogenesis. The clinical disorder acute disseminated encephalomyelitis (ADEM) provides a prototype of an immune mediated demyelinating disorder that aVects the CNS. The disorder is characterized by a uniphasic course followed by a variable degree of recovery. MRI data conWrm the multifocal involvement of the CNS. Pathologic analysis shows multiple perivascular demyelinating lesions, associated with inXammation. All are of the same age. This disorder was Wrst described after Pasteur introduced a neural tissue containing vaccine as a therapy for rabies. It should be noted that, unlike MS, peripheral nervous system involvement is described in >50% of cases (Swamy et al., 1984). The frequency of

IMMUNOBIOLOGY OF THE CLINICAL FEATURES OF MS

post-rabies vaccine–associated ADEM has declined with introduction of vaccines prepared in the absence of neural tissues. ADEM is also recognized to occur after exposure to an array of viral infections. Viral infection is also an established risk factor for disease exacerbation in relapsing MS. Recurrent forms of ADEM have been described in children (an age in which typical MS is unusual); the relationship of this disorder to MS remains to be established (Gusev et al., 2002; Murthy et al., 2002; Tourbah et al., 1999). The Pasteur vaccine complication was initially reproduced in animals by immunization with neural tissue, resulting in an inXammatory demyelinating disorder termed experimental autoimmune encephalomyelitis (EAE). The disorder can now also be induced by means of adoptive transfer of myelin-reactive pro-inXammatory CD4þ T cells into the systemic circulation. Of note, direct injection of such autoreactive T cells into the CNS does not produce EAE. Distinct phenotypes of EAE can be induced that have characteristic regional involvement such as spinal cord or optic nerve. Relapsing and chronic disease forms of EAE can be obtained by means of selection of precise myelin antigen, strain of inbred animal, and immunization regimen. The overlap of acute and chronic/relapsing syndromes in the EAE model raise the issue of overlapping pathogenic mechanisms of uniphasic, recurrent, and chronic phenotypes in the human. Initiation of the MS disease process, by analogy with EAE, has been attributed to CNSdirected autoreactive T cells. Such autoreactive T cells can, however, be derived from the circulation of normal individuals, as well as MS patients. There is an apparent increase in the frequency in the latter, particularly if one considers only cells that show evidence of previous activation by antigen (Bieganowska et al., 1997). Whether this increased frequency in MS reXects a unique exposure to a speciWc antigen or a defect in immune regulation remains speculative. One possible explanation for antigen exposure in MS patients invokes the concept of molecular mimicry. This refers to shared antigenicity between autoantigens (such as myelin constituents) and exogenous antigens (infectious agents) (Wucherpfennig, 2002). The recognition that cross reactivity among peptide antigens is determined by the very limited number of amino acids that make the crucial contact with the receptor on speciWc T cells predicts that the possibilities for cross reactivity are extremely high (Lang et al., 2002). Results of studies in which myelin reactive T cell clones have been exposed to combinatorial libraries of peptides demonstrate that optimal reactivity of such clones is more likely to be induced by a nonmyelin peptide (Sung et al., 2002). In the EAE model, almost always performed in inbred animal strains, the actual autoantigen and the actual encephalitogenic peptide portion thereof varies from strain to strain. Additional models of immune mediated CNS demyelination involve those in which a persistent viral infection in the CNS results in generation of autoreactive T cells. Generation of such cells presumably develops in response to myelin antigens released as a consequence of viral induced neural cell injury (Miller et al., 2001). T cell sensitization could occur either within the CNS or in regional draining lymph nodes. Support for the former is derived from observations that lymphocyte traYcking is ongoing in the CNS even under physiologic conditions and that perivascular and parenchymal microglia have the capacity to serve as competent antigen presenting cells (reviewed in Becher et al., 2000). The documentation that CNS-released antigens are transported back to regional lymph nodes also implicates this site for generation of the putative disease relevant immune response. Entry of autoreactive T cells into the CNS requires a number of molecular events to occur at the level of the blood brain barrier (BBB) and within the CNS parenchyma (Prat et al., 2001). These include the processes of immune cell—endothelial cell adhesion, chemoattraction, and proteolytic digestion of the basement membrane that comprises the BBB and of the extracellular matrix (see other chapters for detailed discussion). Using a Boyden chamber model in which the dual compartments are divided by a porous Wbronectin-coated membrane on which are grown human brain endothelial cells (HBECs), we have found that lymphocyte migration rates are increased in patients during times of relapses compared to times of clinically stable disease (Prat et al., 2002).

793

794

33. MULTIPLE SCLEROSIS: THERAPY

Once antigen reactive T cells enter a tissue, their persistence is dependent on being presented with antigen by competent antigen presenting cells. Animal model studies show that T cells reactive with either neural (MBP reactive) or non-neural (ovalbumin) antigens can transmigrate into the CNS but that only the former will persist (Owens et al., 1998). As mentioned, perivascular and parenchymal microglia are competent APCs. An issue to consider both in the experimental model and in MS is what is the source of antigen that is being presented to the T cells in the CNS of a previously healthy individual. Only now are the techniques becoming available to determine what, if any, peptides are sitting within the MHC class II groove of microglia under ‘‘physiologic conditions’’ (Santambrogio et al., 2001). One speculates whether there is suYcient turnover of myelin under even normal conditions so that its processed peptides are expressed in APCs in the CNS. If tissue injury has occurred, greater amounts of peptide would be expected to be available. Perhaps the initial wave of activated cells that enters the CNS induces suYcient injury via release of pro-inXammatory molecules to result in antigen release. The initial autoreactive T cells interacting with the resident neural cells could initiate a cascade of events that would lead to the tissue injury characteristic of MS (see other chapters).

Remission: Basis for Recovery The basis of clinical recovery following immune mediated tissue injury with resultant neurologic dysfunction appears to involve cessation of the injury process and functional recovery by the injured tissue. The former could result from a relatively passive process in which the inWltrating disease relevant immune cells can no longer sustain their activation state. The process may also be impacted by active immune regulatory mechanisms that can involve functionally distinct cell subsets acting via cell-cell contact or release of soluble anti-inXammatory mediators (Anderton et al., 1999; Kohm et al., 2002). Recovery of function of injured tissue could reXect restoration of impaired axonal function or by remyelination. Restoration of axonal function may occur as a result of redistribution of sodium channels on demyelinated nerve segments (Waxman, 2001). Remyelination could arise from previously myelinating mature oligodendrocytes or, as suggested by most experimental studies, from progenitor cells that have diVerentiated into oligodendrocytes, as discussed later (Chari et al., 2002). Functional recovery, as most clearly shown by functional MRI activation studies, can also reXect the recruitment of additional brain regions to mediate the functions previously associated with the actual sites of injury (Reddy et al., 2000).

Recurrence of Neurologic Episodes Relapses in MS have been empirically deWned as the appearance or reappearance of one or more neurologic abnormalities persisting for at least 24 to 48 hours and occurring 30 days or more after any previous relapse. The requirement that the symptoms or signs persist for at least 1 to 2 days is used to distinguish new pathologic events from transient physiological dysfunction that often occurs in previously damaged tissue. The interval of a month is an arbitrary attempt to deWne whether repeated events belong to one ongoing relapse or to diVerent relapses. As discussed in detail in chapter 32, the MRI correlate of the relapse is a new T2 deWned lesion generally associated with gadolinium enhancement on T1, indicating a local breakdown of the blood brain barrier. The rate of new MRI lesion formation is approximately tenfold greater than the clinical relapse rate. New MRI lesion formation without clinical features is common and as mentioned has been incorporated into the McDonald diagnostic criteria. The best pathologic correlate of the acute relapse and new MRI lesion is perivascular inXammation with extension into the parenchyma. Analysis of cells recovered from the CNS of animals with recurrent forms of EAE suggests that such recurrences can reXect expansion of the immune response over time with development of T cell populations that are reactive with CNS antigens other than those used to initially induce the disease (Tuohy et al., 2000). This is referred to as epitope spreading. Although this phenomenon has been demonstrated to occur in MS patients

au6

THERAPY AND MS

using functional in vitro measures of frequency of circulating myelin reactive T cells (limiting dilution, ELISPOT) (Pelfrey et al., 2000), stronger conWrmation is to be expected with advances in methodology (e.g., MHC class II tetramers). Such methodology should also help address the heterogeneity of autoantigens involved initially in any individual cases. Reactivation or expansion of a disease relevant immune response could occur either in the systemic compartment or within the CNS.

Secondary Progression As stated, approximately 50 to 60% of RR patients will evolve into a secondary progressive (SP) phase of disease after 10 to 15 years. Features of the SP phase of the disease include dysfunction at all levels of the neuraxis including cognitive impairment. SigniWcant diYculty walking (EDSS >4) correlates most closely with spinal cord involvement. In approximately 10 to 15% of cases the disease is progressive from onset with or without intermixed relapses—primary progressive (PP) and progressive relapsing (PR). Whether the basis for progression in these clinical phenotypes is identical to that which accounts for the more common SP form of disease remains speculative. As will be noted, there is an apparent signiWcant lack of eYcacy of systemic immunomodulatory and immunosuppressive therapies in the progressive forms or phases of MS compared to the relapsing forms. Progressive forms of disease can also be induced in the EAE model dependent on animal strain and immunization regimen. In both the progressive MS case and the animal model, the pathology is characterized by extensive tissue destruction within the lesion with loss of oligodendrocytes and axons. At the lesion edge there is accumulation of activated microglia/macrophage with evidence of continued myelin breakdown. Lymphocyte accumulation is less evident. There is little ongoing remyelination. These features are consistent with MRI Wndings of expanded total lesion burden with reduction in the number of gadolinium enhancing lesions. The basis for the evolution of the lesion in MS over time remains to be deWned. The activated microglia/macrophages are sources of potential eVector molecules that could induce tissue injury. Furthermore, uptake of tissue debris serves as means of activation of these cells. Removal of CNS tissue debris by invading macrophages may be necessary for optimal tissue repair. We have proposed that the end-stage loss of oligodendrocytes in the later phases of MS could reXect the increased vulnerability of initial sublethally injured oligodendrocytes to eVector molecules derived from inXammatory cells present in the chronic inXammatory lesion (Wosik et al., 2001). We showed that human oligodendrocytes experimentally injured by transfection with sublethal levels of p53 up-regulated expression of death receptors on the cells making them vulnerable to programmed cell death mediated by death receptor ligands including Fas-ligand and TRAIL. One further speculates whether the lack of remyelination reXects failure of progenitor cells due to such cells themselves being targets of immune mediated injury (Niehaus et al., 2000) or that these cells or that these cells exhaust their capacity to replace destroyed oligodendrocytes/myelin. As regards axonal/nerve cell injury, initial transection of axons would result in extensive Wallerian degeneration of distal segments and loss of cell bodies if the initial injury were close to the cell body. In parallel with observations made at the neuromuscular junction in cases of post-polio syndrome, failure of compensatory synapses over time may also be expected (Cashman et al., 1995).

THERAPY AND MS One might consider this issue in terms of those therapies directed at the immune system and those that are directed at the nervous system. The former therapies are aimed at halting disease development or progression; therapies targeting the nervous system are aimed at protecting tissue from injury, promoting recovery, or enhancing function. A consideration of therapeutic claims in MS needs to take into account the actual design of the clinical studies from which relevant data were generated and how widely the results

795

796

33. MULTIPLE SCLEROSIS: THERAPY

can be extrapolated across the diVerent clinical phenotypes and phases of the disease. Trials in which stringent clinical trial methodology has been applied are the easiest to evaluate. A major dilemma is how to relate results of short-term studies to the long-term natural history of MS. The pivotal phase 3 studies leading to drug approval have been powered to assess eVects on relapse rate and short-term progression. Most current and proposed immune-directed phase 2 studies are powered to generate MRI rather than clinical evidence of potential treatment eYcacy.

IMMUNOMODULATORY AND IMMUNE SUPPRESSIVE TREATMENT Trials with these agents are based on the immune mediated hypothesis of MS pathogenesis; conversely, results of these trials provide support for the hypothesis. These therapies can be divided into those aimed at limiting the severity of relapses and those that prevent them.

TREATMENT OF RELAPSES

au7

Glucocorticoid therapy has become an accepted means to reduce the severity and shorten the duration of clinical attacks of MS, although the lack large-scale clinical trials makes it diYcult to quantitate the magnitude of the eVect (reviewed in Goodin et al., 2002). Existing data do suggest the superiority of a short course of high dose intravenous therapy compared to relatively low and more protracted doses of oral therapy. A single study implicating a deleterious short-term eVect of low dose therapy has not yet been conWrmed (Beck et al., 1993). DeWnitive data on high dose oral therapy are awaited. There seems to be little long-term beneWcial impact on the course of the illness. The apparent limited magnitude of beneWt of IV Ig and plasma exchange therapy for relapses of MS, coupled with the complexity or expense of administering these therapies, limit their use for the usual clinical situation. An open label study suggested that patients who have a very severe attack of fulminant demyelinating disease (not only MS) and who are unresponsive to high-dose glucocorticoids may beneWt from a course of plasma exchange (Weinshenker et al., 1999). Phase 2 clinical trials using an anti-VLA-4 adhesion molecule antibody (Natalizumab, Antegren) indicated that this therapy produced signiWcant results with regard to reducing new MRI lesion frequency and clinical relapse rate (Miller et al., 2003). A phase 3 clinical trial with this agent is currently in progress.

TREATMENTS DIRECTED AT MODIFYING THE COURSE OF MS—RRMS

au8 Immunomodulatory Treatment of RRMS Two families of molecules, interferon b (IFNb) and glatiramer acetate (GA), have undergone large-scale clinical trials in patients with established RRMS. These trials have used both clinical and MR-based measures to demonstrate that the therapies reduced disease activity or development of further neurologic disability. Although the trials share many principles of design, it remains hazardous to do comparative analyses in the absence of head-to-head trials. Interferon b —Interferon beta-1b (IFNb-1b, Betaseron, Betaferon) The results of a large, multicenter, placebo-controlled trial using IFNb-1b were reported in 1993 (The IFNB Multiple Sclerosis Study Group, 1993) and demonstrated that compared to treatment with placebo, treatment with 8 million international units (MIU) of IFNb-1b injected subcutaneously (sc) every other day (qod) reduced the annual clinical attack rate (- 34%; p ¼ 0.0001), rapidly reduced MRI activity as assessed by the number of newly forming gadolinium enhancing lesions (- 59%; p ¼ 0.0089), and decreased the volume of developing white matter disease seen on MRI (- 20%; p ¼ 0.001). This trial also showed a trend

TREATMENTS DIRECTED AT MODIFYING THE COURSE OF MS—RRMS

toward reduction in conWrmed 1-point EDSS progression rate (- 29%; p ¼ 0.16). Treatment with 1.6 mIU sc qod of IFNb-1b (Betaseron) was also better than placebo on several outcome measures but was, in general, not as beneWcial as the higher dose. The approved dose of Betaseron is 8 MIU, sc, qod. IFNb-1a (Avonex) The placebo-controlled clinical trial, published in 1996 (Jacobs et al., 1996), involved administration of 30 mg (6 mIU) of IFNb-1a intramuscularly weekly for up to 2 years to relatively mildly aVected patients (EDSS 1
3.5). The treatment produced a reduction in the clinical attack rate (- 18%; p ¼ 0.04), MRI gadolinium activity (- 52%; p ¼ 0.05) and conWrmed 1-point EDSS progression rate (- 37%; p ¼ 0.02). The total volume of white matter disease seen on MRI may also have been reduced in the treated group (- 6.7%; p ¼ 0.36). A subsequent clinical trial comparing 30 and 60 mg concluded there was no signiWcant dose related eVect (Clanet et al., 2001). 30 mg remains the recommended dose; this would be considered as the lowest dose of currently used IFNbs. IFNb-1a (Rebif) The PRISMS trial, published in 1998 (PRISMS Study Group, 1998), demonstrated that compared with placebo, sc injection of 44 mg (12 mIU) of IFNb-1a (Rebif) three times a week was associated with a reduction in clinical attack rate (- 33%; p < 0.005), the conWrmed 1-point EDSS progression rate (- 30%; p ¼ 0.01), the number of newly forming gadolinium enhancing lesions on MRI (- 78%; p < 0.0001), and the volume of white matter disease seen on MRI (-14.7%; p < 0.0001). Treatment with 22 mg sc three times a week was also eVective on each of these outcome measures. The high-dose IFNb-1a (Rebif) appeared to do better than the lower dose on each of the four main outcome measures, though these diVerences were not statistically signiWcant; 44 mg sc, three times a week is the only approved dose in Europe. The neuroimaging and pathological studies that underscore that damage occurs early in the CNS of patients have sparked an interest in the potential of immunomodulators to change the disease course more eVectively if administered at the time of the initial clinical event. The CHAMPS study (Jacobs et al., 2000) reported that IFNb-1a (Avonex), used in patients with clinically isolated syndromes (CIS) and multifocal MRI abnormalities signiWcantly delayed the progression to clinically deWnite MS compared with placebo. The ETOMS study (Comi et al., 2000) reported similar Wndings when IFNb-1a (Rebif) was used to treat selected patients with CIS. These observations further support the use of IFNb therapies early in the course of disease. The bulk of evidence suggests that there is a dose response eVect of IFNb on the rate of new lesion formation in the MS disease process, as measured by short-term studies of relapse rate (EVIIDENCE) (Panitch H et al., 2002). However, even at the highest IFNb doses used currently, the therapeutic beneWts seem to approach a plateau. As a result, it seems unlikely that clinical relapse rate will ever be reduced by more than 30 to 40% using IFNb alone. IFNb therapies are thought to reduce new lesion formation by acting at several levels of the putative immune cascade that underlies development of such lesions (see Fig. 33.1). These include suppressing T cell activation, inhibiting pro-inXammatory cytokine production, inducing IL-l0 production, and inhibiting the expression of adhesion molecules (Fig. 33.1, step 2) and tissue breaking enzymes (Fig. 33.1, step 4). Because of the latter, interferons would be expected to have an important impact at the level of the blood brain barrier, which is indeed reXected by the consistent demonstration of signiWcant and early suppression of gadolinium enhancement across all IFNb trials. How signiWcant an impact the initial reduction on lesion formation has on subsequent development of long-term neurologic disability remains to be established. A complicating factor regarding long-term eYcacy of IFNb therapy relates to generation of anti-IFNb antibodies during the course of treatment (Bertolotto et al., 2002; Coles, 2002; Pozzilli et al., 2002). Such antibodies may interfere with the beneWcial action of the medication in some patients. Generation of antibodies may reXect the relative dose or preparation (and hence the immunogenicity) of the IFNb used.

797

798

33. MULTIPLE SCLEROSIS: THERAPY

FIGURE 33.1 Potential sites of action of immunomodulatory therapies in MS. In MS, immune cells activated in the periphery (step 1) are thought to migrate across the blood brain barrier (BBB) through a series of molecular interactions involving adhesion (step 2), chemoattraction, (step 3) and transendothelial invasion (step 4). Once in the central nervous system (CNS), tissue directed immune cells may become reactivated (step 5) and participate in the local disease process. Available, and experimental, therapies for MS are likely to mediate their eVects, in part, by acting on these potential targets.

Little data exist to date comparing the impact of current IFNb therapies on quality-oflife measures. Although in all of the preceding trials, a high proportion of patients were able to meet the study end point, the array of side eVects associated with IFNb therapy, particularly the systemic ‘‘Xu-like symptoms’’ and injection site reactions, do impact on patient decisions to accept or continue with therapy in the clinical situation. The rate of discontinuation of IFNb therapy in clinical practice approaches 35 to 40% over 1 to 2 years). Glatiramer Acetate (Copaxone, GA) is a copolymer comprised of four randomly assembled amino acids. In a multicenter, placebo-controlled trial published in 1995 (54), daily 20 mg sc injections of GA were associated with a reduction in the clinical attack rate over a 2-year period (- 29%; p ¼ 0.007). The conWrmed 1-point EDSS progression rate also appeared to be slightly reduced (- 13%; ns). In a subsequent short-duration trial, speciWcally directed toward MRI outcome measures, both the number of newly forming gadolinium enhancing lesions and the volume of white matter disease accumulation seen on MRI, were reduced in the group receiving GA, compared to placebo (Comi et al., 2001). The eVects on MRI reached statistical signiWcance between 6 to 7 months after initiation of therapy. These MRI results contrast with the rapid suppression of new gadolinium enhancing lesions seen with IFNb therapies. These results would be consistent with current data indicating that GA acts via inducing a deviation of the immune response in a Th2 direction, rather than by blocking lymphocyte traYcking across the blood brain barrier (Neuhaus et al., 2001). Thus, the presence or absence of gadolinium enhancement per se does not directly inform us about the pathologic state of the CNS tissue within that lesion. GA therapy can be associated with localized skin reactions, hives, and occasional episodes of immediate post-injection reactions. Flu-like symptoms and fatigue are not an expected

TREATMENTS DIRECTED AT MODIFYING THE COURSE OF MS—RRMS

side eVect. The conclusion from a recent clinical trial with an oral form of Copaxone was that no eYcacy was apparent at the dose being studied. Glucocorticoids Chronic oral steroid therapy has been abandoned as a treatment strategy. A recent trial of regular 3 monthly pulses of high dose I-V corticosteroids concluded that such therapy reduced the development of long-term disability and inhibited development of brain atrophy (Zivadinov et al., 2001). These data raise the issue of the corticosteroids acting via a neuroprotective mechanism. IVIg Several studies have reported that IVIg therapy reduces the frequency of clinical relapses (summarized in 39). To date, however, the eVects of IVIg on disability or on the accumulation of MRI lesion burden have not been Wrmly established. Campath-1 This lympholytic anti-T cell antibody has been shown to virtually eliminate relapse rate and MRI new lesion formation but without an apparent eVect on disease progression (Paolillo et al., 1999). These data again raise the issue as to whether some of the progressive component of MS reXects pathophysiologic mechanisms that are distinct from those underlying the relapsing-remitting elements of the disease. Also of concern is identifying when the progressive process begins. When Campath was initially used, transient worsening of patients was observed that could be prevented by corticosteroid therapy. This eVect was shown to be due to release of cytokines and nitric oxide (NO) and was reproduced in vitro by showing conduction block with NO. Treatments That Aggravate Disease Activity (reviewed in Owens et al., 2001, and Wiendl et al., 2002) Systemic IFN g therapy was reported to increase symptoms in MS patients with RR disease. This study was conducted in the pre-MRI era leaving open the question whether the exacerbation of neurologic deWcits reXected actual new inXammatory lesion formation or physiologic dysfunction of previously demyelinated axons, in parallel with results seen with initial use of Campath-1. Systemic therapies aimed at reducing TNF levels, namely anti-TNF antibody and soluble TNF receptor increased MRI and clinically deWned disease activity. The mechanisms accounting for these results remains to be explained. The deleterious eVects of systemic IFNg and anti-TNF directed therapies in MS were not predicted by results obtained in the EAE model. Future Directions (see Fig. 33.1) Based on growing insights into the immune pathophysiology of MS and the presumed mechanisms of action of the currently approved therapies, a variety of new immunomodulators are being introduced into early phase clinical trials. The process of T cell stimulation (Fig. 33.1, step 1) is being targeted with the development of molecules such as CTLA4-Ig and anti-CD40 antibody that block costimulatory signals required for T cell activation (Laman et al., 1998; Racke et al., 2000). Other molecules including an altered peptide ligand (APL) of myelin basic protein (MBP) have been designed to shift responses of activated T cells from pro-inXammatory to anti-inXammatory response proWles (Bielekova et al., 2000; Kappos et al., 2000). Administration of mixtures of myelin antigens or myelin encoding DNA vaccines are being tested as means to eliminate autoreactive T cells recognizing these antigens via high dose tolerance (Robinson et al., 2002). Studies continue with T cell vaccines using whole T cells, autoreactive T cells, or peptides encoded by T cell receptor V-beta genes whose products are implicated as recognizing disease relevant autoantigens (Vandenbark et al., 1996; Zhang et al., 2002). The anti-VLA-4 antibody (Antegren) noted earlier targets immune cell adhesion to endothelial cells (Fig. 33.1, step 2) and thus prevents entry of inXammatory cells into the CNS. Additional molecules are being designed to limit immune cell inWltration by targeting chemokine-chemokine receptor interactions (Fig. 33.1, step 3) and by suppressing the ability of immune cells to release tissue-breaking enzymes (Fig. 33.1, step 4).

799

800

33. MULTIPLE SCLEROSIS: THERAPY

Immunomodulatory Treatment of SPMS The European trial with IFNb-1b (Betaferon) for SPMS reported beneWt of treatment over placebo in signiWcantly delaying time to sustained progression of disability, with additional beneWts on relapse rate and MRI variables (European Study Group on interferon beta-1b in secondary progressive MS, 1998). The North American study of IFNb-1b (Betaseron) in SPMS and the recently completed trial of IFNb-1a (Rebif) in SPMS both failed to demonstrate a signiWcant reduction in the conWrmed 1-point EDSS progression (the predeWned primary endpoint of both trials) (randomized controlled trial of interferonbeta-1a in secondary progressive MS: Clinical results, 2001; Goodkin et al., 2000). Both reported signiWcant reductions in clinical attack rate, MRI gadolinium activity, and MRI accumulation of white matter disease. The apparent discrepancy between these two trials and the earlier IFNb-1b (Betaferon) trial would appear to relate to the variable inclusion in these trials of progressive patients with concomitant relapses. The use of a 1-point EDSS change to deWne disease progression, and diVerences in the entry EDSS between trial populations, may have also contributed to the disparate results. A trial with Avonex reported a beneWcial eVect on the newly developed MSFC scale but not on the EDSS (Li et al., 2001). While IFNb therapies may therefore have a role in the treatment of patients with progressive MS who also have relapses, there is currently no Wrm evidence to support the use of this family of immunomodulators in patients with SPMS without relapses. IVIg also has no proven eYcacy in the progressive forms of MS.

Immunosuppressive Treatments In the past several decades, there has been considerable interest in the potential use of immunosuppressive agents for patients with MS, particularly those who fail the currently approved therapies described earlier. With the exception of mitoxantrone (recently approved in Europe and by the FDA for use in selected patients with MS), these therapies remain either ‘‘oV label’’ in MS (to be considered only in select circumstances with careful review of toxicity proWles) or strictly experimental. Multiple agents have been tested including methotrexate, cyclosporine A, and cladribine (Goodkin et al., 1995; Rice et al., 2000; Zhao et al., 1997). Described next are those agents that remain in active clinical use or that provide particular insights into the basis of disease development. Azathioprine Meta-analysis of published trials (mainly SPMS) suggests that this agent is marginally eVective (Fernandez et al., 2002; Palace et al., 1997). It is generally administered at a total daily dose of 2 to 3 mg per kg with the goal of lowering the white blood cell count to between 3500 and 4000 cells/mL. Cyclophosphamide This alkylating agent has potent cytotoxic and immunosuppressive eVects. Short-term side eVects include alopecia and hemorrhagic cystitis, and longer-term toxicities include infertility and a potential increase in the risk of bladder cancer. The Northeast Cooperative Treatment Group reported a beneWt of ‘‘pulse’’ therapy on clinical disease stabilization at 24 months though this eVect was no longer seen at 36 months (Hauser et al., 1983; Smith et al., 1998). The Canadian Cooperative MS Study did not show a beneWt of cyclophosphamide therapy on disease progression in an older cohort with more advanced disease (Noseworthy et al., 1991). These apparently contrasting results can be viewed as supporting the concept that early lesion formation in MS reXects an immune-mediated process that is responsive to systemic immune directed therapies, whereas the later progressive phase is more dependent on neurobiologic variables that are resistant to systemic immune therapies. A current focuses of immunosuppressive therapy in MS is on patients with frequent relapses or those who are transitioning into a progressive course despite the use of approved immunomodulating therapies. Treatment protocols involving immunosuppressive induction regimes at disease outset are also being investigated.

NEURAL DIRECTED THERAPY

Mitoxantrone (Novantrone) This anti-neoplastic agent intercalates with DNA and potently suppresses cellular and humoral immune responses. An initial clinical trial involved intravenous (IV) infusion of doses of 12 mg per m2 or 5 mg per m2 every 3 months for 2 years to both RRMS and SPMS patients (Edan et al., 1997; van de Wyngaert et al., 2001). The high-dose mitoxantrone resulted in signiWcant reductions in the clinical attack rate (P ¼ 0.0002), the MRI development of new gadolinium enhanced lesions (P < 0.05), and the total MRI lesion accumulation (P < 0.05), as well as a beneWt in time to reach a 1-point EDSS change (P ¼ 0.04). It has been approved by the FDA for the treatment of SPMS patients and may also be considered in severe, refractory relapses. Because of concerns regarding cardiac toxicity, the recommended total lifetime dose of mitoxantrone is limited, and the proposed regimen is generally oVered for only 2 years. Such a limitation will be problematic for patients expected to require treatment over many years Chronic corticosteroid use, including repeated pulse therapy, in SPMS has been evaluated in several studies that, for the most part, have not demonstrated sustained beneWts. Total lymphoid irradiation is considered too risky for the marginal beneWt it may provide. Plasmapheresis in SPMS has not been shown to be eVective. Intense Immunoablation Followed by Autologous Stem Cell Rescue Early unblinded studies in advanced patients have had mixed results, with ‘‘ ‘remissions’’ reported in some but not all patients and a small but real risk of mortality. A Canadian trial is under way in relatively early aggressive patients who are refractory to conventional therapies. The rational is that such patients may have more to gain and be at lower risk of serious complications. Overall, immunomodulatory and immunosuppressive therapies appear to provide their greatest beneWt early in the disease, during a period in which inXammatory responses are major contributors to tissue injury. As progressive disease sets in and irreversible axonal injury becomes the major process underlying clinical deterioration, the relative contribution of inXammation to ongoing injury may diminish and with it the role of anti-inXammatory therapies. This underscores the importance of initiating therapy early in the disease.

NEURAL DIRECTED THERAPY These can be considered in terms of therapies that mediate neuroportection or repair and those that improve function of the damaged nervous system (i.e., symptomatic therapy). Neurprotection/Regeneration To date there are no signiWcant deWnitive trials or approved therapies related to this category. IFNb itself is shown to increase NAA (a magnetic resonance spectroscopy marker of axonal integrity) after a 6-month delay (Narayanan et al., 2001). Short-term systemic GA therapy is reported to accelerate recovery from experimental acute CNS traumatic lesions, an eVect attributed to pro-inXammatory GA reactive cells that access the injury site. Neuroprotective therapies such as those used in ALS eg the gltutamate inhibitor Riluzole are being explored (Miller et al., 2002). IV Ig therapy failed to promote functional recovery in MS patients (Noseworthy et al., 2001). Initial clinical proof-ofprincipal trials using myelin forming cell transplants and gene transfer of neurotrophic agents are under way.

SYMPTOMATIC TREATMENT IN MS MS produces an array of symptoms that, if eVectively treated, would have major impact on the aVected individual’s quality of life. The symptoms considered next represent those that presumably arise directly consequent to CNS demyelination and axonal injury.

801

802

33. MULTIPLE SCLEROSIS: THERAPY

Focal Weakness fMRI studies conWrm the positive eVects on neural plasticity and functional reorganization achieved by rehabilitation programs. Fatigue Fatigue remains a very common and often very debilitating symptom for patients with MS. Potential biologic correlates of this symptom include critical lesion sites such as brain stem, release of cytokines, and ineVective neural transmission consequent to demyelination. Pharmacologic strategies to combat the fatigue of MS include pemoline, amantadine, and modaWnil. All have demonstrated signiWcant albeit small beneWts in clinical trials. The potassium channel blockers (e.g., 4-aminopyridine, 10 to 40 mg per day; and 3,4-diaminopyridine, 40 to 80 mg per day) may help some MS symptoms (especially heat-sensitive symptoms). These drugs presumably work by prolonging the duration of the nerve action potential. This would facilitate conduction through demyelinated Wbers. At high doses, they may also cause seizures for similar reasons. Spasticity This symptom reXects interruption of transmission along corticospinal tracts. Forty percent of patients rate their spasticity as moderate to severe. Nonpharmacologic approaches to the management of spasticity include physical therapy and exercise. Pharmacologic agents for reducing both spasticity and relalted spasms include baclofen (Lioresal), 20 to 120 mg per day; diazepam (Valium), 2 to 40 mg per day; and tizanidine (ZanaXex), 8 to 32 mg per day. Clonazepam (Klonopin, Rivotril) can be useful, particularly at bedtime, to decrease spasticity and improve sleep quality in some patients. Other medications less well established to provide beneWt for patients with spasticity include carbamazepine, phenytoin, gabapentin, tetrahydrocannabinol, barbiturates, and alcohol. When the spasticity is particularly severe and the patient already has limited use of the lower extremities, a surgically implanted baclofen pump can often provide substantial relief. Destructive procedures such as selective rhyzotomy, tenotomy, myotomy, and phenol injections are reserved for only the most extreme situations. Paroxysmal Symptoms Several diVerent paroxysmal syndromes occur in MS. These syndromes are distinguished by brief duration (30 seconds to 2 minutes); high frequency of occurrence (5 to 40 paroxysms per day; lack of any alteration of consciousness or change in background electroencephalogram during the events); and a self-limited nature (generally lasting only months and then subsiding). They may be precipitated by hyperventilation or movement. These syndromes include the familiar Lhermitte’s sign (electric shock-like sensations induced by neck Xexion), tonic seizures, paroxysmal dysarthria/ataxia, paroxysmal sensory disturbances, and several other less well characterized syndromes. These syndromes are also distinguished by their marked responsiveness to very low dosages of anticonvulsant medications such as carbamazepine (Tegretol), 50 to 400 mg per day; phenytoin (Dilantin), 50 to 300 mg per day; or acetazolamide (Diamox), 200 to 600 mg per day. Pain Pain is an under-appreciated symptom in MS. More than half of patients with MS complain of pain and, in a substantial fraction, the pain is described as severe, at least at times. An improved understanding of the mechanisms that produce pain of central origin has produced several successful approaches to its management, including the anticonvulsant drugs such as carbamazepine, 100 to 1000 mg per day, phenytoin, 300 to 600 mg per day, or gabapentin (Neurontin), 300 to 4800 mg per day; the antidepressant drugs such as amitriptyline, 25 to 150 mg per day, nortriptyline, 25 to 150 mg per day, desipramine, 100 to 300 mg per day, or venlafaxine, 75 to 225 mg per day; or the antiarrhythmic drugs such as mexiletine (Mexitil), 300 to 900 mg per day.

SYMPTOMATIC TREATMENT IN MS

Ataxia or Tremor Ataxia or tremor is a relatively common and often intractable symptom in MS that is diYcult to treat eVectively. Some medications are occasionally helpful including clonazepam (Klonopin, Rivotril), 1.5 to 20 mg per day; primidone (Mysoline), 50 to 250 mg per day; propranolol (Inderal), 40 to 200 mg per day; and on dansetron (Zofran), 8 to 16 mg per day. For the most part, however, the success of such therapy is limited. Recently, there has been interest in the use thalamotomy or the placement of deep brain stimulators to control tremor. However, the response to this intervention, even when performed by a highly qualiWed surgeon, is often partial, the response rates are limited (< 50%), and the duration of any therapeutic beneWt is unknown. Moreover, the surgical procedure itself carries risk.

References

au9 au10

Anderton, S., Burkhart, C., Metzler, B., and Wraith, D. Mechanisms of central and peripheral T-cell tolerance: lessons from experimental models of multiple sclerosis. Immunol. Rev. 169, 123–137. 1999. Antel, J. P. and Bar-Or, A. Multiple sclerosis. (R. E. Rakel and E. T. Bope, eds.). ‘‘Conn’s Current Therapy.’’ 55th. 1-1-2003. W. B. Saunders Company. Becher, B., Prat, A., and Antel, J. P. Brain-immune connection: Immuno-regulatory properties of CNS-resident cells. Glia 29(4), 293–304. 2-15-2000. Beck, R. W., Cleary, P. A., Trobe, J. D., Kaufman, D. I., Kupersmith, M. J., Paty, D. W., and Brown, C. H. The eVect of corticosteroids for acute optic neuritis on the subsequent development of multiple sclerosis. The Optic Neuritis Study Group. N. Engl. J. Med. 329(24), 1764–1769. 12-9-1993. Bertolotto, A., Malucchi, S., Sala, A., OreWce, G., Carrieri, P. B., Capobianco, M., Milano, E., Melis, F., and Giordana, M. T. DiVerential eVects of three interferon betas on neutralising antibodies in patients with multiple sclerosis: a follow up study in an independent laboratory. J. Neurol. Neurosurg. Psychiatry 73(2), 148–153. 2002. Bieganowska, K. D., Ausubel, L. J., Modabber, Y., Slovik, E., Messersmith, W., and HaXer, D. A. Direct ex vivo analysis of activated, Fas-sensitive autoreactive T cells in human autoimmune disease. J. Exp. Med. 185(9), 1585–1594. 5-5-1997. Bielekova, B., Goodwin, B., Richert, N., Cortese, I., Kondo, T., Afshar, G., Gran, B., Eaton, J., Antel, J., Frank, J. A., McFarland, H. F., and Martin, R. Encephalitogenic potential of the myelin basic protein peptide (amino acids 83–99) in multiple sclerosis: Results of a phase II clinical trial with an altered peptide ligand. Nat. Med. 6(10), 1167–1175. 2000. Brex, P. A., Ciccarelli, O., O’Riordan, J. I., Sailer, M., Thompson, A. J., and Miller, D. H. A longitudinal study of abnormalities on MRI and disability from multiple sclerosis. N. Engl. J. Med. 346(3), 158–164. 1-17-2002. Cashman, N. R. and Trojan, D. A. Correlation of electrophysiology with pathology, pathogenesis, and anticholinesterase therapy in post-polio syndrome. Ann. N. Y. Acad. Sci. 753, 138–150. 5-25-1995. Charcot J. M. Lectures on diseases of the nervous system. translated by G. Sigerson. 158–222. 1877. The New Sydenham Society, London. Chari, D. M., and Blakemore, W. F. New insights into remyelination failure in multiple sclerosis: Implications for glial cell transplantation. Mult. Scler. 8(4), 271–277. 2002. Clanet, M., Raduc, E. W., Kappos, L., Hartung, H. P., Hohlfeld, R., Sandberg-Wollheim, M., KooijmansCoutinho, M. F., Tsao, E. C., Sandrock, A. W. (2002). A randomized, double blind, dose-comparison study of weekly interferon beta-1a in relapsing MS. J. Neurol. 59 (10), 1507–1517. Coles, A. J. Neutralising antibodies to the beta interferons. J. Neurol. Neurosurg. Psychiatry 73(2), 110–111. 2002. Comi, G., Filippi, M., Barkhof, F., Durelli, L., Edan, G., Fernandez, O., Hartung, H., Seeldrayers, P., Sorensen, P. S., Rovaris, M., Martinelli, V., Hommes, O. R. (2001). Effect of early interferon treatment on conversion to definite multiple sclerosis: a randomised study. Lancet 357 1576–1582. Comi, G., Filippi, M., and Wolinsky, J. S. European/Canadian multicenter, double-blind, randomized, placebocontrolled study of the eVects of glatiramer acetate on magnetic resonance imaging—measured disease activity and burden in patients with relapsing multiple sclerosis. European/Canadian Glatiramer Acetate Study Group. Ann. Neurol. 49(3), 290–297. 2001. Edan, G., Miller, D., Clanet, M., Confavreux, C., Lyon-Caen, O., Lubetzki, C., Brochet, B., Berry, I., Rolland, Y., Froment, J. C., Cabanis, E., Iba-Zizen, M. T., Gandon, J. M., Lai, H. M., Moseley, I., and Sabouraud, O. Therapeutic eVect of mitoxantrone combined with methylprednisolone in multiple sclerosis: A randomised multicentre study of active disease using MRI and clinical criteria. J. Neurol. Neurosurg. Psychiatry 62(2), 112–118. 1997. European Study Group on Interferon Beta-1b in Secondary Progressive MS. Placebo-controlled multicentre randomised trial of interferon beta-1b in treatment of secondary progressive multiple sclerosis. Lancet 352(9139), 1491–1497. 11-7-1998.

803

804

33. MULTIPLE SCLEROSIS: THERAPY

au11

Fernandez, O., Guerrero, M., Mayorga, C., Munoz, L., Lean, A., Luque, G., Hervas, M., Fernandez, V., Capdevila, A., and De Ramon, E. Combination therapy with interferon Beta-1b and azathioprine in secondary progressive multiple sclerosis: A two-year pilot study. J. Neurol. 249(8), 1058–1062. 2002. Goodin, D. S., Frohman, E. M., Garmany, G. P., Jr., Halper, J., Likosky, W. H., Lublin, F. D., Silberberg, D. H., Stuart, W. H., and van den, Noort S. Disease modifying therapies in multiple sclerosis: Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology and the MS Council for Clinical Practice Guidelines. Neurology 58(2), 169–178. 1-22-2002. Goodkin, D., and the North American Study Group on Interferon beta-1b in Secondary Progressive MS. Interferon beta-1b in secondary progressive MS: clinical and MRI results of a 3-year randomized controlled trial. Neurology 54, 2352. 2000. Goodkin, D. E., Rudick, R. A., VanderBrug, Medendorp S., Daughtry, M. M., Schwetz, K. M., Fischer, J., and Van Dyke, C. Low-dose (7.5 mg) oral methotrexate reduces the rate of progression in chronic progressive multiple sclerosis. Ann. Neurol. 37(1), 30–40. 1995. Gusev, E., Boiko, A., Bikova, O., Maslova, O., Guseva, M., Boiko, S., Vorobeichik, G., and Paty, D. The natural history of early onset multiple sclerosis: comparison of data from Moscow and Vancouver. Clin. Neurol. Neurosurg. 104(3), 203–207. 2002. Hauser, S. L., Dawson, D. M., Lehrich, J. R., Beal, M. F., Kevy, S. V., Propper, R. D., Mills, J. A., and Weiner, H. L. Intensive immunosuppression in progressive multiple sclerosis. A randomized, three-arm study of high-dose intravenous cyclophosphamide, plasma exchange, and ACTH. N. Engl. J. Med. 308(4), 173–180. 1-27-1983. Hoeber, Paul B. Association for Research in Nervous and Mental Disease: Multiple Sclerosis (disseminated sclerosis). [1). 1922. New York. Interferon Beta-1b Is EVective in Relapsing-Remitting Multiple Sclerosis. I. Clinical results of a multicenter, randomized, double-blind, placebo-controlled trial. The IFNB Multiple Sclerosis Study Group. Neurology 43(4), 655–661. 1993. Jacobs, L. D., Beck, R. W., Simon, J. H., Kinkel, R. P., Brownscheidle, C. M., Murray, T. J., Simonian, N. A., Slasor, P. J., and Sandrock, A. W. Intramuscular interferon beta-1a therapy initiated during a Wrst demyelinating event in multiple sclerosis. CHAMPS Study Group. N. Engl. J. Med. 343(13), 898–904. 9-28-2000. Jacobs, L. D., Cookfair, D. L., Rudick, R. A., Herndon, R. M., Richert, J. R., Salazar, A. M., Fischer, J. S., Goodkin, D. E., Granger, C. V., Simon, J. H., Alam, J. J., Bartoszak, D. M., Bourdette, D. N., Braiman, J., Brownscheidle, C. M., Coats, M. E., Cohan, S. L., Dougherty, D. S., Kinkel, R. P., Mass, M. K., Munschauer, F. E., III, Priore, R. L., Pullicino, P. M., Scherokman, B. J., Whitham, R. H., and Intramuscular interferon beta-1a for disease progression in relapsing multiple sclerosis. The Multiple Sclerosis Collaborative Research Group (MSCRG). Ann. Neurol. 39(3), 285–294. 1996. Kappos, L., Comi, G., Panitch, H., Oger, J., Antel, J., Conlon, P., and Steinman, L. Induction of a nonencephalitogenic type 2 T helper-cell autoimmune response in multiple sclerosis after administration of an altered peptide ligand in a placebo-controlled, randomized phase II trial. The Altered Peptide Ligand in Relapsing MS Study Group. Nat. Med. 6(10), 1176–1182. 2000. Kaufman, M., Gaydos, C. A., Sriram, S., Boman, J., Tondella, M. L., and Norton, H. J. Is Chiamydia pneumoniae found in spinal Xuid samples from multiple sclerosis patients? ConXicting results. Mult. Scler. 8(4), 289–294. 2002. Kohm, A. P., Carpentier, P. A., Anger, H. A., and Miller, S. D. Cutting Edge: CD4(þ)CD25(þ) Regulatory T Cells Suppress Antigen-SpeciWc Autoreactive Immune Responses and Central Nervous System InXammation During Active Experimental Autoimmune Encephalomyelitis. J. Immunol. 169(9), 4712–4716. 11-1-2002. Laman, J. D., Maassen, C. B., Schellekens, M. M., Visser, L., Kap, M., de Jong, E., van Puijenbroek, M., van Stipdonk, M. J., van Meurs, M., Schwarzler, C., and Gunthert, U. Therapy with antibodies against CD40L (CD154) and CD44-variant isoforms reduces experimental autoimmune encephalomyelitis induced by a proteolipid protein peptide. Mult. Scler. 4(3), 147–153. 1998. Lang, H. L., Jacobsen, H., Ikemizu, S., Andersson, C., Harlos, K., Madsen, L., Hjorth, P., Sondergaard, L., Svejgaard, A., Wucherpfennig, K., Stuart, D. I., Bell, J. I., Jones, E. Y., and Fugger, L. A functional and structural basis for TCR cross-reactivity in multiple sclerosis. Nat. Immunol. 3(10), 940–943. 2002. Li, D. K., Zhao, G. J., and Paty, D. W. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: MRI results. Neurology 56(11), 1505–1513. 6-12-2001. McAlpine D. The benign form of multiple sclerosis. a study based on 241 cases seen within 3 years of onset and followed up until the 10th year or more of the disease. Brain 84, 186–203. 1961. McDonald, W. I., Compston, A., Edan, G., Goodkin, D., Hartung, H. P., Lublin, F. D., McFarland, H. F., Paty, D. W., Polman, C. H., Reingold, S. C., Sandberg-Wollheim, M., Sibley, W., Thompson, A., van den Noort S., Weinshenker, B. Y., and Wolinsky, J. S. Recommended diagnostic criteria for multiple sclerosis: Guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann. Neurol. 50(1), 121–127. 2001. Miller, D. H., Khan, O. A., Sheremata, W. A., Blumbardt, L. D., Rice, G. P., Libonati, M. A., Willmer-Hulme, A. J. Dalton, C. M., Miszkiel, K. A., O’Connor, P. W. (2003). A controlled trial of natalizumab for relapsing multiple sclerosis. N. Engl. J. Med. 348, 15–23. Miller, R. G., Mitchell, J. D., Lyon, M., and Moore, D. H. Riluzole for amyotrophic lateral sclerosis (ALS)/motor neuron disease (MND). Cochrane. Database. Syst. Rev. (2), CD001447. 2002. Miller, S. D., Katz-Levy, Y., Neville, K. L., and Vanderlugt, C. L. Virus-induced autoimmunity: Epitope spreading to myelin autoepitopes in Theiler’s virus infection of the central nervous system. Adv. Virus Res. 56, 199–217. 2001.

805

SYMPTOMATIC TREATMENT IN MS

au13

Moore, F. G., and Wolfson, C. Human herpes virus 6 and multiple sclerosis. Acta Neurol. Scand. 106(2), 63–83. 2002. Murray T. J., Burks, J. S., and Johnson, K. P. ‘‘Multiple Sclerosis: Diagnosis, Medical Management, and Rehabilitation,’’ (1), 1–34. 2000. Demos Medical Publishing, New York. Murthy, S. N., Faden, H. S., Cohen, M. E., and Bakshi, R. Acute disseminated encephalomyelitis in children. Pediatrics 110 (2 Pt 1), e21. 2002. Narayanan, S., De Stefano, N., Francis, G. S., Arnaoutelis, R., Caramanos, Z., Collins, D. L., Pelletier, D., Arnason, B. G. W., Antel, J. P., and Arnold, D. L. Axonal metabolic recovery in multiple sclerosis patients treated with interferon beta-1b. J. Neurol. 248(11), 979–986. 2001. National Multiple Sclerosis Society. ‘‘Therapeutic Claims in Multiple Sclerosis,’’ 1st. ed. 1982. New York. Neuhaus, O., Farina, C., Wekerle, H., and Hohlfeld, R. Mechanisms of action of glatiramer acetate in multiple sclerosis. Neurology 56(6), 702–708. 3-27-2001. Niehaus, A., Shi, J., Grzenkowski, M., Diers-Fenger, M., Archelos, J., Hartung, H. P., Toyka, K., Bruck, W., and Trotter, J. Patients with active relapsing-remitting multiple sclerosis synthesize antibodies recognizing oligodendrocyte progenitor cell surface protein: implications for remyelination. Ann. Neurol. 48(3), 362–371. 2000. Noseworthy, J. H., O’Brien, P. C., Petterson, T. M., Weis, J., Stevens, L., Peterson, W. K., Sneve, D., Cross, S. A., Leavitt, J. A., Auger, R. G., Weinshenker, B. G., Dodick, D. W., Wingerchuk, D. M., and Rodriguez, M. A randomized trial of intravenous immunoglobulin in inXammatory demyelinating optic neuritis. Neurology 56(11), 1514–1522. 6-12-2001. Noseworthy, J. H., Vandervoort, M. K., Penman, M., Ebers, G., Shumak, K., Seland, T. P., Roberts, R., Yetisir, E., Gent, M., and Taylor, D. W. Cyclophosphamide and plasma exchange in multiple sclerosis. Lancet 337(8756), 1540–1541. 6-22-1991. Owens, T., Tran, E., Hassan-Zahraee, M., and Krakowski, M. Immune cell entry to the CNS–a focus for immunoregulation of EAE. Res. Immunol. 149(9), 781–789. 1998. Owens, T., Wekerle, H., and Antel, J. Genetic models for CNS inXammation. Nat. Med. 7(2), 161–166. 2001. Palace, J., and Rothwell, P. New treatments and azathioprine in multiple sclerosis. Lancet 350(9073), 261. 7-261997. Panitch, H., Goodin, D. S., Francis, G., Chang, P., Coyle, P. K., O’Connor, P., Monaghan, E., Li, D., Weinsberkar, B. (2002). Randomized, comparative study of interferon beta-1a treatment regimens in MS: the evidence trail. Neurology 59(10), 1496–1506. Paolillo, A., Coles, A. J., Molyneux, P. D., Gawne-Cain, M., MacManus, D., Barker, G. J., Compston, D. A., and Miller, D. H. Quantitative MRI in patients with secondary progressive MS treated with monoclonal antibody Campath 1H. Neurology 53(4), 751–757. 9-11-1999. Paty, D., and Ebers, G. C. ‘‘Multiple Sclerosis.’’ 1997.F. A. Davis Company, Philadelphia. Pelfrey, C. M., Rudick, R. A., Cotleur, A. C., Lee, J. C., Tary-Lehmann, M., and Lehmann, P. V. QuantiWcation of self-recognition in multiple sclerosis by single-cell analysis of cytokine production. J. Immunol. 165(3), 1641–1651. 8-1-2000. Pozzilli, C., Antonini, G., Bagnato, F., Mainero, C., Tomassini, V., Onesti, E., Fantozzi, R., Galgani, S., Pasqualetti, P., MilleWorini, E., Spadaro, M., Dahlke, F., and Gasperini, C. Monthly corticosteroids decrease neutralizing antibodies to IFNbeta1 b: A randomized trial in multiple sclerosis. J. Neurol. 249(1), 50–56. 2002. Prat, A., Biernacki, K., Lavoie, J. F., Poirier, J., Duquette, P., and Antel, J. P. Migration of multiple sclerosis lymphocytes through brain endothelium. Arch. Neurol. 59(3), 391–397. 2002. Prat, A., Biernacki, K., Wosik, K., and Antel, J. P. Glial cell inXuence on the human blood-brain barrier. Glia 36(2), 145–155. 2001. PRISMS (Prevention of Relapses and Disability by Interferon beta-1a Subcutaneously in Multiple Sclerosis) Study Group. Randomised double-blind placebo-controlled study of interferon beta-1a in relapsing/remitting multiple sclerosis. Lancet 352(9139), 1498–1504. 11-7-1998. Racke, M. K., Ratts, R. B., Arredondo, L., Perrin, P. J., and Lovett-Racke, A. The role of costimulation in autoimmune demyelination. J. Neuroimmunol. 107(2), 205–215. 7-24-2000. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: Clinical results. Neurology 56(11), 1496–1504. 6-12-2001. Reddy, H., Narayanan, S., Arnoutelis, R., Jenkinson, M., Antel, J., Matthews, P. M., and Arnold, D. L. Evidence for adaptive functional changes in the cerebral cortex with axonal injury from multiple sclerosis. Brain 123 (Pt. 11), 2314–2320. 2000. Rice, G. P., Filippi, M., and Comi, G. Cladribine and progressive MS: Clinical and MRI outcomes of a multicenter controlled trial. Cladribine MRI Study Group. Neurology 54(5), 1145–1155. 3-14-2000. Robinson, W. H., Garren, H., Utz, P. J., and Steinman, L. Millennium Award. Proteomics for the development of DNA tolerizing vaccines to treat autoimmune disease. Clin. Immunol. 103(1), 7–12. 2002. Santambrogio, L., Belyanskaya, S. L., Fischer, F. R., Cipriani, B., Brosnan, C. F., Ricciardi-Castagnoli, P., Stern, L. J., Strominger, J. L., and Riese, R. Developmental plasticity of CNS microglia. Proc. Natl. Acad. Sci. USA 98(11), 6295–6300. 5-22-2001. Smith, D. R., Olek, M. J., Balashov, K. E., Khoury, S. J., HaXer, D. A., and Weiner, H. L. Principles of immunotherapy. ‘‘Clinical Neuroimmunology’’ (J. P. Antel, G. Birnbaum, and H-P Hartung), (7), 92–104. 1998. Blackwell Science, Oxford. Sung, M. H., Zhao, Y., Martin, R., and Simon, R. T-cell epitope prediction with combinatorial peptide libraries. J. Comput. Biol. 9(3), 527–539. 2002.

au12

806

33. MULTIPLE SCLEROSIS: THERAPY

Swamy, H. S., Shankar, S. K., Chandra, P. S., Aroor, S. R., Krishna, A. S., and Perumal, V. G. Neurological complications due to beta-propiolactone (BPL)-inactivated antirabies vaccination. Clinical, electrophysiological and therapeutic aspects. J. Neurol. Sci. 63(1), 111–128. 1984. Talbot, P. J., Arnold, D., and Antel, J. P. Virus-induced autoimmune reactions in the CNS. Curr. Top. Microbiol. Immunol. 253, 247–271. 2001. ‘‘Therapeutic Claims in Multiple Sclerosis.’’ 3rd. ed. Sibley, W. 1992. Demo Publications, New York. Tourbah, A., Gout, O., Liblau, R., Lyon-Caen, O., Bougniot, C., Iba-Zizen, M. T., and Cabanis, E. A. Encephalitis after hepatitis B vaccination: recurrent disseminated encephalitis or MS? Neurology 53(2), 396–401. 7-22-1999. Tubridy, N., Behan, P. O., Capildeo, R., Chaudhuri, A., Forbes, R., Hawkins, C. P., Hughes, R. A., Palace, J., Sharrack, B., Swingler, R., Young, C., Moseley, I. F., MacManus, D. G., Donoghue, S., and Miller, D. H. The eVect of anti-alpha4 integrin antibody on brain lesion activity in MS. The UK Antegren Study Group. Neurology 53(3), 466–472. 8-11-1999. Tuohy, V. K., and Kinkel, R. P. Epitope spreading: a mechanism for progression of autoimmune disease. Arch. Immunol. Ther. Exp. (Warsz). 48(5), 347–351. 2000. van de Wyngaert, F. A., Beguin, C., D’Hooghe, M. B., Dooms, G., Lissoir, F., Carton, H., and Sindic, C. J. A double-blind clinical trial of mitoxantrone versus methylprednisolone in relapsing, secondary progressive multiple sclerosis. Acta Neurol. Belg. 101(4), 210–216. 2001. Vandenbark, A. A., Chou, Y. K., Whitham, R., Mass, M., Buenafe, A., Liefeld, D., Kavanagh, D., Cooper, S., Hashim, G. A., and OVner, H. Treatment of multiple sclerosis with T-cell receptor peptides: Results of a double-blind pilot trial. Nat. Med. 2(10), 1109–1115. 1996. Waxman, S. G. Acquired channelopathies in nerve injury and MS. Neurology 56(12), 1621–1627. 6-26-2001. Weinshenker, B. G., O’Brien, P. C., Petterson, T. M., Noseworthy, J. H., Lucchinetti, C. F., Dodick, D. W., Pineda, A. A., Stevens, L. N., and Rodriguez, M. A randomized trial of plasma exchange in acute central nervous system inXammatory demyelinating disease. Ann. Neurol. 46(6), 878–886. 1999. Wiendl, H., and Hohlfeld, R. Therapeutic approaches in multiple sclerosis: Lessons from failed and interrupted treatment trials. BioDrugs. 16(3), 183–200. 2002. Wosik, K., Antel, J., Kuhlmann, T., Bruck, W., Massie, B., and Nalbantoglu, J. (2003). Oligodendrocyte injury in multiple sclerosis: a role for p53. J. Neurochem. 85(3), 635–644. Wucherpfennig, K. W. Infectious triggers for inXammatory neurological diseases. Nat. Med. 8(5), 455–457. 2002. Zhang, J. Z., Rivera, V. M., Tejada-Simon, M. V., Yang, D., Hong, J., Li, S., Haykal, H., Killian, J., and Zang, Y. C. T cell vaccination in multiple sclerosis: Results of a preliminary study. J. Neurol. 249(2), 212–218. 2002. Zhao, G. J., Li, D. K., Wolinsky, J. S., Koopmans, R. A., Mietlowski, W., Redekop, W. K., Riddehough, A., Cover, K., and Paty, D. W. Clinical and magnetic resonance imaging changes correlate in a clinical trial monitoring cyclosporine therapy for multiple sclerosis. The MS Study Group. J. Neuroimaging 7(1), 1–7. 1997. Zivadinov, R., Rudick, R. A., De Masi, R., Nasuelli, D., Ukmar, M., Pozzi-Mucelli, R. S., Grop, A., Cazzato, G., and Zorzon, M. EVects of IV methylprednisolone on brain atrophy in relapsing-remitting MS. Neurology 57(7), 1239–1247. 10-9-2001.

C H A P T E R

34 Adrenoleukodystrophies Hugo W. Moser

HISTORY The disorder now referred to as X-linked adrenoleukodystrophy (X-ALD) was Wrst described by Siemerling and Creutzfeldt in 1923 (Siemerling and Creutzfeldt, 1923). They reported a 7-year-old boy who had been well until age 3 or 4, when he was Wrst noted to be hyperpigmented. At 61/2 years he became disturbed, and his speech and gait deteriorated. He became spastic, unable to walk or swallow, and died at 7 years. Postmortem examination showed adrenal atrophy and extensive demyelination combined with perivascular accumulation of lymphocytes and plasma cells in the central nervous system. They referred to this condition as Bronzekrankheit und Skelosierendc Encephalomyelitis, and noted the resemblance of the neuropathological features to the encephalitis periaxialis diVusa described by Paul Schilder in 1912 and in 1913. In retrospect it is now clear that the case reported by Haberfeld and Spieler in 1910, and studied neuropathologically by Schilder (Schilder, 1924), had the same condition, in view of the fact that the clinical history and neuropathological Wndings were similar and the note in the case history that the patient had become hyperpigmented, even though adrenal pathology was not commented on. Additional cases were reported by PWster (1936), Hampel (1937), Adams and Kubic (Adams and Kubic, 1952), Gagnon (1959), Lichenstein and Rosenbluth (1959), Brun and Voigt (1960), Hoefnagel and Van Den Noort (1962), Fanconi et al. (1963), and Blaw et al. (1964). The cases were referred to variously as ‘‘DiVuse Hirnsklerose’’ (PWster, 1936), ‘‘Morbus Addisonii und Skleroriesiende Erkrankung’’ (Hampel, 1937), ‘‘Sclerose Cerebrale diVuse avec Melanoderme et atrophie surrenale’’ (Gagnon, 1959), ‘‘Entzundlische cerebrale Sklerose mit NebenniereninsuYzienz’’ (Brun and Voigt, 1960), and ‘‘Addison’s disease and diVuse sclerosi’’ (Hoefnagel and Van Den Noort, 1962). It has also been referred to as Addison-Schilder disease, and less speciWcally as Schilder’s disease. The designation adrenoleukodystrophy was introduced by Blaw in 1970 (Blaw, 1970) and is now used worldwide. On the basis of pedigree analysis Fanconi et al. proposed in 1963 (Fanconi et al., 1963) that the disorder had an X-linked recessive mode of inheritance, and this has been conWrmed. In 1976 and in 1977, Budka et al. and GriYn et al. independently described X-ALD adult patients with progressive paraparesis and primary adrenal insuYciency (Budka et al., 1976; GriYn et al., 1977). This entity is now referred to as adrenomyeloneuropathy (AMN). AMN has now been shown to have the same biochemical and genetic basis as the childhood forms of X-ALD and they often co-occur in the same family. We recommend that the designation X-linked adrenoleukodystrophy (X-ALD) be applied to encompass all of the phenotypic variants listed in Table 34.1. X-ALD must be distinguished sharply from the entity referred to as connatal adrenoleukodystrophy (Ulrich et al., 1978) or neonatal adrenoleukodystrophy (NALD) (Kelley et al., 1986). Although NALD shares certain biochemical and pathological features

Myelin Biology and Disorders, Volume 2

807

Copyright 2003, Elsevier (USA). All rights reserved.

au1

808

34. ADRENOLEUKODYSTROPHIES

TABLE 34.1 Phenotypes in Males and Females Phenotypes in Males Phenotype

Description

Estimated relative frequency

Childhood cerebral (CCER)

Onset at 3–10 years of age. Progressive behavioral, cognitive and neurologic deWcit, often leading to total disability within 3 years. InXammatory brain demyelination

31–35%

Adolescent

Like childhood cerebral. Onset age 11–21 years. Somewhat slower progression.

4–7%

Adrenomyeloneuropathy (AMN)

Onset 28 + 9 years, progressive over decades. Involves spinal cord mainly, distal axonopathy inXammatory response mild or absent. Approximately 40% have or develop cerebral involvement with varying degrees of inXammatory response and more rapid progression.

40–46%

Adult cerebral

Dementia, behavioral disturbances. Sometimes focal deWcits, without preceding AMN. White matter inXammatory response present. Progression parallels that of childhood cerebral form.

2–5%

Olivo-ponto-cerebellar

Mainly cerebellar and brainstem involvement in adolescence or adulthood.

1–2%

Aaddison-only@

Primary adrenal insuYciency without apparent neurologic involvement. Onset common before 7.5 years. Most eventually develop AMN.

Varies with age. Up to 50% in childhood.

Asymptomatic

Biochemical and gene abnormality without demonstrable adrenal or neurologic deWcit. Detailed studies often show adrenal hypofunction or subtle signs of AMN.

Diminishes with age. Common < 4 years. Very rare > 40 years

Phenotypes in Female X-ALD Carriers Phenotype

Description

Estimated relative frequency

Asymptomatic

No evidence of adrenal or neurologic involvement

Diminishes with age. Most women < 30 years neurologically uninvolved.

Mild myelopathy

Increased deep tendon reXexes and distal sensory changes in lower extremities with absent or mild disability.

Increases with age. Approximately 50% > 40 years.

Moderate to severe myeloneuropathy

Symptoms and pathology resemble AMN, but milder and later onset.

Increases with age. Approximately 15% > 40 years.

Cerebral involvement

Rarely seen in childhood and slightly more common in middle age and later.

Approximately 2%.

Clinically evident adrenal insuYciency

Rare at any age.

Approximately 1%.

with X-ALD, it is fundamentally distinct. It is a disorder of peroxisome biogenesis with an autosomal recessive mode of inheritance (Gould et al., 2001). The discussion in this chapter will be conWned to X-ALD. A key advance about the understanding of X-ALD was achieved at the Albert Einstein College of Medicine in New York between 1973 and 1976 when Powers and Schaumburg (1973) demonstrated characteristic inclusions in the adrenal cortical cells and brain macrophages and Igarashi et al. (1976b) showed that these inclusions contained large amounts of cholesterol esteriWed with saturated unbranched very long chain fatty acids (VLCFA). These VLCFA consisted mainly of tetracosanoic (C24:0) and hexacosanoic (C26:0) acids. This led to the recognition that X-ALD is a lipid storage disease. In 1980 and 1981, Moser et al. showed that VLCFA levels were also increased in cultured skin Wbroblasts and plasma. The plasma VLCFA assay is the most widely used diagnostic assay and has led to the identiWcation of thousands of patients (Moser et al., 1999). In 1984 Singh et al. (1984b) showed that white blood cells and cultured skin Wbroblasts had an impaired capacity to degrade VLCFA. This led to the conclusion that the basic biochemical defect in X-ALD is the impaired capacity to degrade VLCFA. Two laboratories (Lazo et al., 1988; Wanders et al., 1988) reported that the biochemical defect involved the impaired capacity to form the coenzyme derivative of VLCFA, a reaction catalyzed by the perox-

IV. DISEASES OF MYELIN

CLINICAL FEATURES

isomal enzyme VLCFA synthetase (VLCS). This conclusion had been generally accepted, but has been thrown into question recently by the studies of McGuinness (McGuinness et al., 2003), which indicate that the earlier report of Tsuji et al. (1981a) that VLCFA synthesis is increased in patients with X-ALD must also be considered as a factor in the pathogenesis of the VLCFA accumulation. The gene that is defective in X-ALD was mapped to Xq28 in 1981 (Migeon et al., 1981). Mosser et al. identiWed the defective gene by positional cloning (Mosser et al., 1993). It is now referred to as ABCD1. The gene product has no homology to VLCS, but codes for a peroxisomal membrane protein (ALDP) that is a member of the ATP binding cassette (ABC) transporter protein family (Higgins, 1992). The mechanism through which the ALDP deWciency leads to the accumulation of VLCFA or to the adrenal and brain pathology is not yet understood. Mouse models of X-ALD by targeted inactivation of ABCD1 were developed in 1997 (Forss-Petter et al., 1997; Kobayashi et al., 1997; Lu et al., 1997).

CLINICAL FEATURES X-ALD has been observed in all ethnic groups and aVects approximately 1:20,000 males (Bezman et al., 2001). The pattern of inheritance is X-linked recessive. The clinical manifestations of X-ALD vary widely and for reasons that are still unknown the various phenotypes often co-occur within the same family. Table 34.1 shows the subgroupings that have been established and their relative frequency. Figure 34.1 shows a survival analysis for the cerebral phenotypes.

Childhood Cerebral X-ALD The designation of childhood cerebral X-ALD (CCER) is applied to boys who develop evidence of neurologic involvement before 10 years of age. Prenatal, perinatal, and early postnatal development are entirely normal. In a 1987 survey (Moser et al., 1987), we found the mean age of onset of neurologic symptoms to be 7.2 + 1.0 years with a range of 2.75 to 10 years, but recently we became aware of a patient who became symptomatic at 21 months of age. Table 34.2 lists the initial symptoms in a series of 180 patients. Most commonly, the Wrst neurologic symptoms are in the behavioral sphere and are often mistaken for hyperactivity attention deWcit disorder. The child may also exhibit emotional lability, withdrawn behavior, or school failure or combination of the three. The behavioral disturbances vary with the location of the demyelinating lesions on MRI. In approximately 80%, the initial lesions involve the parieto-occipital white matter (Kumar et al., 1987) and manifest as disturbances of visuospatial or auditory perception and spatial orientation. DiYculty in understanding speech in a noisy room or over the telephone are common early symptoms and reXect impaired auditory discrimination, often with retention in normal pure tone perception. Visual impairment is an early symptom in approximately one-third of the patients and often is due to a combination of optic nerve, optic tract, and occipital lobe involvement. Visual Weld cuts and impaired visual acuity are common. In approximately 15% of patients, the initial lesion involves the frontal lobes. These patients often present with disinhibited behavior and other behavioral disturbances without the focal neuropsychological deWcits that occur in patients with the parieto-occipital lesions. Shapiro and collaborators have described a series of neuropsychologic tests that are particularly relevant to the evaluation of children with leukodystrophies (Shapiro et al., 1995, 2000; Shapiro and Klein, 1994). These tests assess the Wve major domains of language, visual perception, visual/constructional function, memory, and executive function. They are of value in assessing disease progression and prognosis and for the selection of patients who are candidates for bone marrow transplantation (Shapiro et al., 2000). Apart from the neuropsychological, auditory and visual disturbances, early neurologic symptoms may include impaired sport performance, unsteady gait, poor handwriting,

IV. DISEASES OF MYELIN

809

810

34. ADRENOLEUKODYSTROPHIES

FIGURE 34.1 Time is age in years. See Table 34.1 for deWnition of phenotypes. Note that the rapid downward slopes for each of the cerebral phenotypes are approximately equal. As noted in the sections on pathology and pathogenesis, the rapid progression of X-ALD is due the inXammatory cerebral demyelination. Most commonly, this response manifests in childhood. Less commonly, it manifests in adolescence or adulthood, but once it commences the rate of progression is approximately the same, irrespective of age. The approximately equal downward slope of the survival curves of each of the cerebral phenotypes demonstrates this.

TABLE 34.2

Initial Symptoms in 160 Patients with the Childhood Cerebral Form of Adrenoleukodystrophy

Symptom

Percentage

School diYculty

16

Behavioral disturbances

13

Impaired vision

11

Impaired hearing

8

Poor coordination

8

Dementia

7

Seizure

7

Hyperactivity

6

Squint, double vision

5

DiYculty walking

4

Speech diYculty

4

Limb weakness

3

Poor handwriting

2

Headaches

2

Loss of athletic ability

1

Urinary incontinence

1

Tics

1

Fecal incontinence

0.3

Increased intracranial pressure

0.3

DiYculty swallowing

0.3

Coma

0.3

IV. DISEASES OF MYELIN

CLINICAL FEATURES

strabismus, and seizures. A seizure was the Wrst neurologic manifestation in 7% of the patients (Table 34.2). Once the neurologic symptoms become manifest, progression often is rapid. In the 1987 Kennedy Krieger Institute series (Moser et al., 1987) that involved 167 patients with CCER, the mean interval between Wrst neurologic symptoms and an apparently vegetative state was 1.9 + 2 years (range 0.5 to 10.5 years). In this state, the child is bedridden, is unable to speak or see, and is fed via nasogastric tube or gastrostomy. Ability to interact may be retained to a variable extent that may be diYcult to assess. The child or adolescent may remain in this apparently vegetative state for several years, in some instances more than 5 years. The mean age at death in the childhood form was 9.4 + 2 years. Recent follow-up studies at the Kennedy Krieger Institute have shown that information of prognostic signiWcance can be obtained by subdividing the patients in accordance with age and degree of MRI abnormality (Moser et al., 2000). Eighteen subgroups were established. The purpose of the study was to determine whether the presence or absence of brain MRI abnormality at a given age correlates with outcome. The MRI involvement was assessed with the 34-point scale developed speciWcally for X-ALD by Loes et al. (1994). As demonstrated in Figures 34.2A and 34.2B, the correlation in some of the age groups was striking. In the 7-to-10-year age group, all of the 22 patients with normal MRI (Loes score < 1) survived and remained neurologically intact, whereas half of the patients with MRI score > 3 died during the follow-up period. Table 34.3 summarizes the results.

Adolescent Cerebral ALD In one series (Moser et al., 1987) of 837 patients, there were 42 patients in whom Wrst symptoms occurred between age 11 and 21. Symptoms resemble those in CCER.

Adrenomyeloneuropathy Adrenomyeloneuropathy (AMN) is a disorder that involves the spinal cord mainly. It is now subdivided into ‘‘pure AMN,’’ where neuropathological changes are conWned to the spinal cord and peripheral nerves (Powers et al., 2000), and ‘‘AMN-cerebral,’’ in which there is also diVuse involvement of cerebral white matter (Schaumburg et al., 1975). As discussed in the section on pathogenesis, the distinction between AMN and the cerebral forms of the disease is fundamental. Kumar et al. (1995) found that the brain MRI is normal in 54% of AMN patients. Some degree of cerebral involvement coexists in 46%. In a 10-year follow-up study, van Geel et al. found that, 19% of pure AMN patients developed cerebral involvement during that period (van Geel et al., 2001). Typically, a man with pure AMN had been neurologically normal, often with good athletic skills, until his twenties when he noted stiVness or clumsiness in his legs. Adrenocortical insuYciency is demonstrable in at least 50% of AMN patients (Brennemann et al., 1996). The neurologic disability is slowly progressive, so that within the next to 15 years, the gait disturbance becomes severe and requires the use of a cane or a wheelchair. Urinary disturbances and sexual dysfunction are noted in the twenties or thirties. The somatosensory and brainstem auditory evoked responses are nearly always abnormal. Visual evoked and peripheral nerve abnormalities occur less frequently. Neuropsychologic function in patients with pure AMN is normal or only minimally impaired (Edwin et al., 1996).

AMN-Cerebral Phenotype GriYn et al. noted already in 1977 (GriYn et al., 1977) that some pure AMN patients develop diVuse cerebral involvement. In a cross sectional study, Kumar et al. (1995) found that 46% of AMN patients had some degree of cerebral involvement. The cerebral involvement may already be present when AMN is Wrst diagnosed, or it may develop later in patients with pure AMN. The inXammatory cerebral pathology may not begin until the Wfth decade (van Geel et al., 2001), but once it manifests it may progress as rapidly as in patients with CCER.

IV. DISEASES OF MYELIN

811

812

34. ADRENOLEUKODYSTROPHIES

FIGURES 34.2A AND 34.2B MRI score is based on the 34-point scale developed by Loes et al. (1994) speciWcally for X-ALD. A score of < 1 is classiWed as within normal limits. The Wgures show that for the 7 to 10 years age group, the MRI score is highly predictive. If a 7- to 10-year-old boy has an MRI score that is greater than three, he has a 50% chance of dying in the next 5 years. None of the boys in this age group who had a normal MRI died during that period. However, they will most likely develop the milder AMN in adulthood.

IV. DISEASES OF MYELIN

CLINICAL FEATURES

TABLE 34.3 X-ALD Prognosis as a Function of Age and MRI Score at First Encounter MRI score > 3: 70–80% worsen irrespective of age; 8/80 remain neurological stable MRI score 1–3: 60% worsen irrespective of age; Longer survival MRI score < 1: Age 3–7: 30% develop cerebral involvement; Ages 7–10: 10% develop cerebral involvement Age > 10: Cerebral involvement rare

Adult Cerebral X-ALD The term ‘‘adult cerebral X-ALD’’ is applied to patients with the biochemical defect of XALD who develop cerebral symptoms after 21 years of age, but who do not have signs of spinal cord involvement; that is, they do not have evidence of AMN. This form is relatively rare. Twenty-three cases have been reported (Moser et al., 2000). Age of onset ranged from the early twenties to the Wfties. Symptoms resembled schizophrenia with dementia or there may be a focal neurological deWcit that led to suspicion of a brain tumor (Bresnan and Richardson, 1979). The most common initial psychiatric manifestations are signs of hypomania including disinhibition, impulsivity, increased spending, hypersexuality, loudness, and perserveration (Garside et al., 1999). Adult cerebral X-ALD has a serious prognosis. The mean interval between Wrst neurologic symptom and an apparently vegetative state or death was 3 to 4 years.

Presentation AS Olivocerebellar Atrophy There are eight reports of patients who presented with clinical manifestations of olivopontocerebellar atrophy (Kurihara et al., 1993). Age at presentation ranged from 5 years to adulthood. Cerebellar and pontine atrophy was present in all patients in whom imaging studies were performed and may be the only demonstrable abnormality initially. The illness was progressive and cerebral white matter abnormalities became evident later in all instances.

Addison Disease Only X-ALD patients are assigned to the ‘‘Addison only’’ phenotype category if they have adrenal insuYciency without demonstrable evidence of nervous system involvement. X-ALD is one of the causes of Addison disease. Lauretti et al. (1996) reported it as the cause of adrenal insuYciency in 35% of male patients who had previously been diagnosed as having primary idiopathic adrenocortical insuYciency. They noted that none of the X-ALD patients had adrenocortical antibodies. Analogous results were obtained by Jorge et al. (1994), who demonstrated X-ALD as the cause of Addison disease in 5 of 24 patients. Statistical analysis indicated that the likelihood of X-ALD as the cause is age dependent. It is most likely when Addison manifests before the age of 7.5 years. Most patients with the Addison-only phenotype later develop neurologic involvement. Van Geel et al. (2001) found that during a 10-year follow-up period this occurred in 50%.

Asymptomatic Male X-ALD Patients While many young X-ALD patients are asymptomatic, most develop either neurologic or adrenal involvement or both at some time and we are not aware of any male patients who remained asymptomatic after age 60.

IV. DISEASES OF MYELIN

813

814

34. ADRENOLEUKODYSTROPHIES

Manifesting Heterozygotes

au2

Some degree of neurologic involvement occurs frequently in women who are heterozygous for X-ALD. In a series of 104 women who attended the annual meeting of the United Leukodystrophy Foundation, 61% had some neurologic abnormalities that resembled pure AMN (Moser et al., 1991). It is likely that this group was reasonably representative of the overall X-ALD heterozygote population because most of the women attended the conference because of concern about their sons rather than their own health. Their mean age was 32 years. Disability was severe in 14%. Twenty percent had slight to moderate symptoms, while another 22% had not complained of symptoms but were found to have hyperreXexia and impaired vibration sense in the lower extremities. The mean age of onset of neurologic symptoms was 37.8 + 14.8 years. Restuccia et al. (1999) demonstrated abnormalities of motor and sensory evoked responses in 12 of, 191 heterozygotes. The pure AMN-like neurologic symptoms are progressive and are present in more than 50% heterozygous women in late middle age or later. Less than 1% of heterozygous women develop Addison disease or cerebral involvement. Naidu et al. (1997) have proposed that these rare occurrences are associated with skewed X-inactivation patterns in which the normal allele is not expressed. Maier et al. (2002) reported recently that neurologically symptomatic women are more likely to have skewed X-inactivation patterns. Hershkovitz (Hershkovitz et al., 2002) have reported a 9-year-old girl with cerebral involvement as severe as in boys with CCER. She had a pathogenic X-ALD mutation on the maternally derived X-chromosome combined with a de-novo deletion of Xq28 on the paternally derived X, and thus was totally deWcient in the X-ALD gene.

PATHOLOGY

au3 It is of key importance to note that the pathology of the cerebral forms of X-ALD diVers fundamentally from that of pure AMN. The cerebral forms are associated with an inXammatory response in the cerebral white matter. Pure AMN is mainly a distal axonopathy (Powers et al., 2000), and the inXammatory response is absent or mild. Pathology of the Cerebral Forms The gray matter is usually intact, but the centrum semiovale is consistently Wrm (sclerotic) and replaced by large areas of brown to gray translucent tissue. The loss of myelin is conXuent, usually asymmetric, and most prominent in the parieto-occipital regions with caudorostral progression. Several less common patterns has been reported, including the forms in which the demyelinative process starts in the frontal region or in the cerebellum and pons (Kurihara et al., 1993). Cavitation and calciWcation of white mater may be seen in severe cases. Arcuate Wbers are relatively spared. In the most common form with posterior presentation, the posterior cingulum, corpus callosum, fornix, hippocampas commissure, posterior limb of the internal capsule, and optic systems are typically involved. The cerebellar white matter usually exhibits a similar but milder, conXuent loss of myelin and sclerosis. Secondary corticospinal tract degeneration extending down through the peduncles, basis pontis, medullary pyramids, and spinal cord is characteristic. The spinal cord is spared in the CCER phenotype except for the descending tract degeneration. Histopathologically there is a marked loss of myelinated axons (Myelin > axons) and oligodendrocytes in association with hypertrophic reactive astrocytosis. The advancing active edges of myelin loss are sites of intense perivascular inXammation and lipid-laden macrophages. DiVuse inWltration of macrophages and large perivascular collections of mononuclear cells, particularly lymphocytes, are highly characteristic of areas of early myelin breakdown. Recent studies have revealed that most of the lymphocytes are CD8 cytotoxic cells (Ito et al., 2001). Pathology of Pure AMN The spinal cord bears the brunt of the disease process in men with pure AMN and also in neurologically symptomatic heterozygous women. Loss of myelinated axons and a milder

IV. DISEASES OF MYELIN

BIOCHEMICAL ABNORMALITIES

loss of oligodendrocytes is observed in the long ascending and descending tracts of the spinal cord, especially the fasciculus gracilis and the lateral corticospinal tracts. The pattern of Wber loss is consistent with a distal axonopathy in that the greatest losses are observed in the distal segments—that is, in the cervical region for the ascending fasciculus gracilis and the descending corticospinal tract. Recent pathological studies of the dorsal root ganglia in AMN patients (Powers et al., 2001) provide further support for this formulation. Powers et al. point out that a dying-back pattern of axonal degeneration such as seen in Friedreich’s ataxia, and AMN could be due either to neuronal death or to axonal damage. The former is the case in Friedreich’s ataxia, where there is neuronal loss and nodules of Nageoti are prominent (Hughes et al., 1968). In contrast, in the AMN patients the number of dorsal root ganglion cells was not reduced, there was no evidence of necrosis or apoptosis, and nodules of Nageoti were not observed. This points toward axonal pathology as the primary event. Two other Wndings were of interest: even though the total number of neurons was not reduced, there was a reduction of the number of large neurons and ultrastructural studies of the neurons revealed electron-dense lipidic inclusions. The implications of these Wndings will be discussed in the section on pathogenesis. Peripheral Nerves Peripheral nerve lesions in AMN are variable and mild compared to the myelopathy. Sural and peroneal nerves have displayed loss of large and small diameter myelinated Wbers, endoneurial Wbrosis, and thin myelin sheaths (Julien et al., 1981; Martin et al., 1980). Chaudhry et al. (1996) studied 13 variables of peripheral nerve function in 99 men with AMN and 37 heterozygous women. At least one variable was abnormal in 87% of the men and 67% of the women. Abnormalities were more common in the men than the women. They concluded that the abnormalities represented a mixture of axonal loss and multifocal demyelination. Van Geel et al. (1996) studied 18 men with AMN and Wve neurologically symptomatic heterozygotes. Sixty-Wve percent of the patients had a polyneuropathy with predominantly axonal sensorimotor features, and they concluded that primary axonal degeneration is the principal abnormality. Only two (9%) of the patients fulWlled the electrodiagnostic criteria for primary demyelination. Adrenal Cortex and Testis Adrenocortical cells, particularly those of the inner fasciculata-reticularis, become ballooned and striated due to accumulations of lamellae, lamellar lipid-laden proWles, and Wne lipid clefts (Fig. 34.3). The striated material, which contains cholesterol esters esteriWed with VLCFA, appears to lead to cell dysfunction, atrophy, and death (Powers and Schaumburg, 1974). InXammatory cells are only rarely observed. Ultimately primary atrophy of the adrenal cortex ensues. In fetuses aVected by X-ALD, the fetal adrenal zone is already severely involved (Powers et al., 1982). In the testes, lamellae and lamellarlipid proWles are present in the interstitial cells of Leydig and their precursors. Degenerative changes in the seminiferous tubules and Sertoli cells are observed in AMN (Powers and Schaumburg, 1981), and may eventually lead to azoospermia (Aversa et al., 1998). It should be noted, however, that many X-ALD patients have fathered children. The Kennedy Krieger Institute records include 964 children fathered by 336 X-ALD patients.

BIOCHEMICAL ABNORMALITIES Accumulation of Very Long Chain Fatty Acids Very long chain fatty acids (VLCFA) are deWned as saturated and unsaturated fatty acids with carbon chain lengths longer than 22 atoms. Rezanka (1989) reviewed the literature about VLCFA up to 1989 and points out that they are almost omnipresent in the animal and plant kingdoms, varying from 0343.1% to 10% of total fatty acids. Normally saturated VLCFA occur in highest concentration in myelin lipids and red blood cell sphingomyelin. They occur in much lower concentration in other tissues. The accumulation of very long

IV. DISEASES OF MYELIN

815

816

34. ADRENOLEUKODYSTROPHIES

FIGURE 34.3 This adrenocortical cell contains both unilamellate and multilammete inclusions, which are both free in the cytoplasm and are attached to various organelles (arrows). (MagniWcation X 23,625.) Electron micrograph was taken from uranyl acetate-lead citrate-stained thin sections. From CRC critical reviews of Neurobiology 3, 29–88 1987, with permission

chain fatty acids is the principal biochemical abnormality in X-ALD. This was demonstrated Wrst in post-mortem brain tissue and adrenal gland by Igarashi et al. (1976b), where these fatty acids accounted for 11 to 40% of fatty acids in the cholesterol ester fraction, whereas they were virtually absent from similar fractions in control tissues. The excess of VLCFA in postmortem brain tissue has been conWrmed in numerous subsequent studies (Brown et al., 1983; Menkes and Corbo, 1977; Molzer et al., 1981; Ramsey et al., 1979; Reinecke et al., 1985; Taketomi et al., 1987; Theda et al., 1992; Wilson and Sargent, 1993). The VLCFA that accumulate in X-ALD are saturated and unbranched, and involve mainly those with a carbon chain length of 26 (hexacosanoic acid, C26:0) or 24 (tetracosanoic acid, C24:0). The accumulation of VLCFA in X-ALD brain has a speciWc pattern. While some degree of excess is present in most tissues, the most striking increases are found in the cholesterol ester fractions of the brain and the adrenal glands. The VLCFA excess in the brain cholesterol ester fraction correlates with histopathology. The greatest enrichment (approximately 40%) is found in actively demyelinating areas, whereas it was similar to control (6%) in the cholesterol ester fraction of histologically normal brain regions, and two and a half times normal (15%) in a gliotic region (Theda et al., 1992). VLCFA levels are also increased in the proteolipid fraction (Bizzozero et al., 1991) and in gangliosides. VLCFA of brain gangliosides in X-ALD patients account for 25 to 40% of their total fatty acids, compared to less than 1% in controls. Brown et al. (1983) examined the lipid

IV. DISEASES OF MYELIN

au6

817

GENE DEFECT

composition of puriWed myelin from three X-ALD patients. Unlike the controls, the lipids of this preparation included 8.9% cholesterol ester. This may have been due to contamination with a low-density fraction from regions where myelin had been destroyed. While this cholesterol ester fraction contained tenfold C26:0 increase, the C26:0 increase in the cerebroside, sulfatide, and sphingomyelin fractions was less than twofold. VLCFA levels are increased also in the total lipid fractions of plasma (Moser et al., 1981, 1999) and in the erythrocyte membrane sphingomyelin fraction (Antoku et al., 1987; Tsuji et al., 1981c). Measurement of VLCFA levels in plasma (Moser et al., 1981, 1999) or red blood cells (Antoku et al., 1987; Tsuji et al., 1981c) are the most commonly used diagnostic assays. The mechanisms responsible for the VLCFA accumulation will be discussed in the section on pathogenesis.

GENE DEFECT ABCD1, the gene that is defective in X-ALD, was mapped to Xq28 in 1981 (Migeon et al., 1981) and isolated and cloned by positional cloning in 1993 (Mosser et al., 1993). It occupies approximately 26 kb of genomic DNA. It is composed of 10 exons and encodes an mRNA of 4.3 kb and a predicted protein (ALDP) of 745 amino acids (Sarde et al., 1994). ALDP, contrary to earlier expectations (see section on pathogenesis), has no sequence homology to any of the VLCS, but instead was found to be a member of the ATP binding cassette (ABC) transmembrane transporter superfamily that transport a wide variety of substrates, including ions, sugars, amino acids, proteins, and lipids (Higgins, 1992). Forty-eight mammalian ABC transporters are estimated to exist (Dean et al., 2001). Mammalian ABC transporter proteins typically consist of two hydrophobic transmembrane domains and two hydrophobic nucleotide-binding folds encoded by a single gene. Peroxisomal ABC transporters have been designated as subgroup (D), which at this time includes four members that are listed in Table 34.4. ABCD1 is the gene that is defective in X-ALD. ABCD2 codes for ALDR (adrenoleukodystrophy related protein). ALDR maps to 12q11 has 66 homology to ALDP (Lombard-Platet et al., 1996) and its exon structure is similar to that of ALDP (Broccardo et al., 1998). The ABCD2 expression patterns diVers from that of ABCD1. ABCD1 is expressed strongly in glia and the adrenal cortex, and ABCD2 in neurons and the adrenal medulla (Dubois-Dalq et al., 1999). ABCD3 codes for PMP70 (Kamijo et al., 1990) and ABCD4 for PMP70/69R (Holzinger et al., 1997; Shani et al., 1997). They are all encoded as half-transporters with a single transporter domain and a single binding fold (Dean et al., 2001). They function as dimers, either as homodimers or as heterodimers with other members of the ABCD group (Liu et al., 1999; Smith et al., 1999).

Mutation Analyses Figure 34.4 shows a topographic model of ALDP. The Kennedy Krieger Institute and the Laboratory for Genetic Metabolic Disease at the Academic Medical Center in Amsterdam have established a website (www.x-ald.nl), which lists and updates the mutations that have been identiWed worldwide. Four-hundred-six mutations had been deWned by 2001 (Kemp et al., 2001). Their nature and location in respect to exons and domains is shown in Figure 34.5. Two-hundred-thirty-four (57.6%) of the mutations are nonrecurrent and unique to a kindred. Of all the mutations, 227 (55.9%) are missense, 110 (27.1%) are nonsense, 16 (3.9%) are small in-frame amino acid insertions or deletions, and 16 (3.9%) are large deletions of one or more exons. Figure 34.4 shows that disease-causing mutations are distributed throughout the gene, but the distribution is not even. There is clustering of mutations in the transmembrane domain (40%), in the ATP binding domain (30%) and in exon 5 (14%). A hotspot has been identiWed on exon 5 (Kemp et al., 1994; Kok et al., 1995). This mutation involves the deletion of two nucleotides (AG) at cDNA position 1415–1416. This results in a frame shift at amino acid residue Glu 471 (fs E471) and a premature stop codon at position 554. The predicted ALDP lacks the ALDP-binding

IV. DISEASES OF MYELIN

au4

818

34. ADRENOLEUKODYSTROPHIES

TABLE 34.4 Peroxisomal ABC Half-Transporters Protein

Percentage identity to ALDP

Chromosomal location

100

Xq28

ALDR

66

12q11

PMP70

38

1p21

P70R (PMP69)

27

14q24

ALD

FIGURE 34.4

IV. DISEASES OF MYELIN

819

GENE DEFECT

FIGURE 34.5 Graphic presentations of mutations in the X-ALD gene identiWed worldwide up to 1998. Schematic presentation of the open reading frame. The putative transmembrane and ATP-binding domain are shown in A. B. Distribution of nonrecurrent mutations in the X-ALD gene. Each vertical bar represents the location of a ‘‘private’’ mutation. Mutations are grouped by type. The density of missense mutations is greatest in the transmembrane and ATP-binding domains. The arrows in the splice defect columns indicate whether the defect aVects the splice donor or spice acceptor site. C. Presentation of all mutations identiWed in the ALD gene other than the 15 large chromosomal deletions that had been identiWed by 1998. The common dinucleotide deletion (415delAG) that was observed in 24 families is not to scale. From Archives of Neurology 56, 273–275 1999, with permission.

domain and is both unstable and inactive. This deletion has been found in 10.3% (42/ 406) of X-ALD kindreds. Haplotype analysis has excluded a founder eVect (Kemp et al., 1994). Combined mutation and ALDP expression data (Feigenbaum et al., 1996; Watkins et al., 1995) were available in 216 X-ALD cell lines as of June 2001 (Kemp et al., 2001). ALDP was not immunologically detectable in 178 (82.4%). The 216 informative cell lines included 78 non-recurrent mutations. Sixty of these (79.5%) result in the absence of detectable protein levels. Normal levels of ALDP expression were found in 38 (17.6%). In these 38 cell lines, 16 diVerent missense mutations were identiWed. All ABCD1 mutations other than missense mutations resulted in unstable and therefore non-detectable levels of protein. Expression data was available on 137 missense mutations, which included 52 that were unique. Only 31% of the nonrecurrent mutations failed to aVect expression of ALDP. Of these, 11% resulted in strongly reduced expression, and 58% in unstable or absent levels. Missense mutations located around the ATP-binding cassette are more likely to aVect protein stability. Liu et al. (1999) demonstrated that amino acids near the carboxyl site, which includes this region, are important for the dimerization of ALDP with itself, with ALDRP, or with PMP70. The disease-causing eVect of some these missense mutations thus may be due to the inhibition of dimer formation.

Genotype-Phenotype Correlations It has not been possible to establish genotype-phenotype correlations so far. Mild phenotypes have been observed in patients with large deletions (Mosser et al., 1993) and in a large number of mutations that result in absence of protein, and conversely severe phenotypes have been associated with missense mutations in which ALDP is expressed. The mutation 1415 del AG has been identiWed in patients with all possible X-ALD phenotypes

IV. DISEASES OF MYELIN

au7

820

34. ADRENOLEUKODYSTROPHIES

(Kemp et al., 1994). Another striking demonstration of the complexity of genotypephenotype correlation is the presence of Wve diVerent phenotypes in six members of a family with a destabilizing missense mutation P 484R (Berger et al., 1994). However, a recent intriguing observation by O’Neill et al. suggests that genotype-phenotype correlation may exist in some families (O’Neill et al., 2001). They studied a family in which the phenotype pattern is highly unusual and possibly unique in that all nine aVected males and aVected heterozygous women were concordant for the pure AMN phenotype. They demonstrated a so far unique mutation that aVects an ABCD1 translation-initiation codon, which results in an N-terminal truncated ALDP missing the Wrst 65 amino acids. Possibly this mutation exerts a dominant eVect. The frequent co-occurrence of diverse phenotypes suggest the action of a modiWer gene. Genetic segregation support the hypothesis that at least one autosomal gene plays a role (Maestri and Beaty, 1992; Moser et al., 1992; Smith et al., 1991).

PATHOGENESIS Pathogenesis of VLCFA Accumulation Impaired Peroxisomal Beta Oxidation in X-ALD Fibroblasts, White Blood Cells, and Amniocytes: Recent Controversy In 1984, Singh et al. (1984b) demonstrated that the capacity to oxidize C24:0 was reduced to 20 to 30% of control in cultured skin Wbroblasts and in white blood cells of patients with X-ALD. This Wnding has been conWrmed in three other laboratories (Kemp et al., 1998; Poulos et al., 1986; Rizzo et al., 1984; Wanders et al., 1987). The defect has also been demonstrated in cultured skin Wbroblasts of the X-ALD mouse model (Kemp et al., 1998; Lu et al., 1997) and in cultured human X-ALD amniocytes and in one instance in an adrenal biopsy sample from an X-ALD patient (Singh et al., 1981). The oxidation of VLCFA takes place in the peroxisome (Singh et al., 1984a) by a series of biochemical reactions that diVer from those used for fatty acid oxidation in the mitochondrion (Hashimoto, 1996). These results had led to the generally accepted conclusion that the accumulation of VLCFA is due to defective VLFA oxidation in the peroxisome. However, this conclusion was recently challenged by McGuinness et al. (2003). While their studies in the X-ALD mouse model conWrmed that VLCFA oxidation in Wbroblasts was impaired, they reported the surprising Wnding that VLCFA oxidation the brain, adrenal, heart, liver, and kidney did not diVer from control even though VLCFA levels were increased in all of these tissues. There is no explanation at this time for the diVerence in VLCFA oxidation in Wbroblasts in mice and humans and in human white blood cells and amniocytes versus those in X-ALD mouse tissues. Comparable studies in human X-ALD tissues are not available, except for the earlier study of Singh et al. of the adrenal biopsy sample of one X-ALD patient (Singh et al., 1981), in which VLCFA beta-oxidation was reduced. McGuinness et al. carried out an extensive investigation of the mechanism of VLCFA accumulation, which also included studies of the eVects of pharmacological agents such as 4-phenylbutyrate and studies of VLCFA oxidation in cell lines of patients with various mitochondrial disorders. They present evidence that the gene defect in X-ALD aVects the interaction between peroxisomes and mitochondria, with the primary defect involving the oxidation of long chain fatty acids in mitochondria. They found that increased levels of saturated long chain fatty acids (C16:0) impair peroxisomal VLCFA oxidation in both normal and in X-ALD Wbroblasts. They propose that an increase in long chain fatty acid levels is a primary event and that increased VLCFA levels are secondary. They postulate that in X-ALD mouse tissues there is an imbalance between the rate of incorporation VLCFA into complex lipids and the rate of complex lipid degradation, with a shift toward synthesis in X-ALD. Consistent with this Wnding are the earlier reports of Tsuji et al. (1981a) that the rate of C26:0 synthesis is increased in X-ALD Wbroblasts. The discrepancy between these recent Wndings of McGuinness et al. in X-ALD mouse tissues and the previously generally accepted conclusion that the VLCFA accumulation in

IV. DISEASES OF MYELIN

PATHOGENESIS

Wbroblasts and white blood cells is due to impaired peroxisomal VLCFA oxidation is unresolved at this time. Intuitively it seems unlikely that diVerent mechanisms would be the cause of VLCFA in Wbroblasts and the tissues. However, as noted in the next section, the VLCFA CoA synthetases, enzymes involved in peroxisomal beta oxidation, have tissue speciWcity, and this could contribute to diVerences in the mechanism of VLCFA accumulation in various tissues. In any case, the recent data of McGuinness et al. indicate that the previous conclusion that the VLCFA excess is fully accounted for by a defect in VLCFA oxidation must be reexamined. VLCFA CoA Synthetases Involved in Peroxisomal Beta Oxidation In 1984, Singh et al. (1984a) demonstrated that C24:0 is oxidized mainly in the peroxisome. Peroxisomal beta oxidation shortens VLCFA by a series of cycles each of which shortens the fatty acid by two carbons. Each cycle involves four enzymatic reactions (Hashimoto, 1996). The Wrst of these involves the formation of the thioester with Coenzyme A. Three groups have reported that this reaction is deWcient in X-ALD (Hashmi et al., 1986; Lazo et al., 1988; Wanders et al., 1988). The reaction is catalyzed by VLCFA CoA synthetase (VLCS). The family of proteins that have VLCS activity includes six members (Watkins et al., 1999) that diVer in regard to substrate speciWcity, species and tissue distribution. One is involved mainly in bile acid metabolism (Mihalik et al., 2002; Steinberg et al., 2000a). Most are localized to the peroxisomes and microsomes. The VLCS (VLCS 1) that has been studied in greatest detail in respect to ALD was the Wrst to be puriWed (Heinzer et al., 2002) and cloned (Uchiyama et al., 1996). The human and murine enzymes have 83% identity contain 620 amino acids and contain both peroxisomal and microsomal targeting sequences with most activity found in the peroxisome. All of the VLCS activate also long chain fatty acids such as C16:0 in addition to VLCFA, with the activation of long chain fatty acids being greater than that of VLCFA. They have two highly conserved motifs. Motif 1 is an AMP-binding domain that is also present in other acyl-CoA synthetases, suggesting that the reaction involves the hydrolysis of ATP to AMP and pyrophosphate. The second motif, the function of which is still unknown, was unique to the VLCS that had been identiWed by, 1997 (Black et al., 1997). VLCS1 is expressed in human and mouse Wbroblasts. In the mouse VLCS1 RNA is most abundant in the liver and kidney. It is also present in the brain and the adrenal, but only at approximately one Wfth the level of that in the liver (Heinzer et al., 2002). Recently, a newly identiWed VLCS has attracted interest in regard to X-ALD. This enzyme was Wrst identiWed in a Drosophila Melanogaster neurodegenerative mutant that accumulates neuronal lipids that under the microscope look like bubblegum. Brain VLCFA levels were increased (Min and Benzer, 1999). The enzyme that is deWcient in this mutant was found to have VLCS activity and is referred to as bubblegum. Human bubblegum has been cloned (Steinberg et al., 2000b). While it has VLCS activity, it does not contain the Motifs associated with all of the other mammalian VLCS that has been identiWed up to 1999 (Watkins et al., 1999). It also diVers from the other VLCS in that it is localized to the cytosol and plasma membrane rather than the peroxisome. However, it is of interest in relation to X-ALD because, unlike the other VLCS, it is expressed primarily in the brain (Steinberg et al., 2000b) and also in the adrenal gland and testis (Watkins, P., personal communication), the tissues that are involved in X-ALD. VLCS Has No Direct Role in X-ALD When deWcient VLCS activity was demonstrated in X-ALD Wbroblasts in 1984, it was considered likely that the gene defect in X-ALD would be found to involve this enzyme. However, when the defective gene was cloned in 1993, it was found to code for a peroxisomal membrane protein (ALDP) that is a member of the ATP binding cassette (ABC) transporter superfamily (Higgins, 1992) and the gene product has no homology to VLCS. VLCS activity and localization in the mouse model of X-ALD is normal (Heinzer et al., 2002) and a VLCS1 mouse knockout model does not show accumulation of VLCFA or any other features of X-ALD (Heinzer, A., unpublished observation).

IV. DISEASES OF MYELIN

821

822

34. ADRENOLEUKODYSTROPHIES

It is thus clear that VLCS deWciency is not the primary biochemical defect in X-ALD. Nevertheless, some interactions between ALDP, the gene product that is deWcient in XALD, and VLCS have been described and will be discussed in the section on the physiological role of ALDP.einze VLCFA Synthesis and Its Relation to X-ALD Synthesis of fatty acids with chain length greater than 16 carbons is carried out by the fatty acid elongating system. The system occurs in both mitochondria and microsomes. The microsomal system appears to be more active and to have greater physiological signiWcance (Murad and Kishimoto, 1978). The stoichiometry of the reaction is as follows: Palmitoyl CoA þ Malonyl CoA þ NADPH þ Hjstearoyl CoA þ CO2 þ NADPþ CoA þ H2O With this reaction, a C16:0 fatty acid (palmitic) is elongated to C18:0 (stearic acid). VLCFA synthesis is achieved by repeated additions of malonyl CoA, so that two carbon units are added until the desired chain length is achieved. Fatty acid synthesis and elongation are complex and highly regulated processes involving multiple enzymes and acyl carrier proteins (Volpe and Vagelos, 1976). The maturational changes in activity of the brain microsomal fatty acid elongating system correlate with the deposition of myelin (Murad and Kishimoto, 1978). Formation of saturated VLCFA was Wrst demonstrated in rat sciatic nerve (Cassagne et al., 1978). The factors that control the rate of elongation are still poorly understood. The chain length of the substrate is an important factor. In a study utilizing swine cerebral microsomes, it was found that elongation of C20:0 CoA yielded C22:0 and 24:0 concomitantly, whereas elongation of C22:0 CoA yielded only negligible amounts of C24:0. Kinetic studies in this system suggested that elongation of C20:0 CoA and of C22:0 CoA are carried out through two separate pathways with that for the C20:0 substrate more active (Yoshida and Takeshita, 1987). Bourre et al. concluded that a single enzyme is responsible for the elongation of behenic acid (2w23:0) and its monounsaturated counterpart erucic acid (22:1) (Bourre et al., 1976). This presumably is the basis for the striking lowering of the levels of saturated VLCFA when patients with X-ALD are treated with a 4.1 mixture of glyceryl trioleate and trierucate (Lorenzo’s Oil), the active component of which is erucic acid (Rizzo et al., 1989) (see section on therapy). Further study of the elongating system in X-ALD is indicated for three reasons: (1) studies in which deuterated water was administered orally to a patient with X-ALD demonstrated substantial incorporation of deuterium into VLCFS, indicating that there was substantial endogenous synthesis (Moser et al., 1983); (2) administration of a 4:1 glyceryl trioleate and glyceryl trieuricate mixture, which is thought to act by inhibiting VLCFA synthesis, leads to a striking reduction in plasma VLCFA levels (Rizzo et al., 1989); )and 3) kinetic studies in cultured skin Wbroblasts of X-ALD patients conducted by Tsuji and associates indicate that the activity of the fatty acid elongating system that forms C26:0 is increased (Koike et al., 1991; Tanaka, 1988; Tsuji et al., 1981b). Further discussion of this topic is presented in the next section. Physiological Role of ALDP The physiological role of ALDP is not yet understood. ALDP has been localized to the peroxisomal membrane in human Wbroblasts (Mosser et al., 1994) with the hydrophylic carboxyl-terminal domain oriented toward the cytoplasm (Watkins et al., 1995). That ALDP does have a role in VLCFA metabolism is indicated by the demonstration in several laboratories that overexpression of ALDP increases VLCFA oxidation in cultured Wbroblasts of X-ALD patients (Braiterman et al., 1998; Flavigny et al., 1999; Kemp et al., 1998; Netik et al., 1999). See Figure 34.7. Note that addition of the other peroxisomal ABC transporter proteins (ALDRP and PMP 70) also had this stimulatory eVect. The mechanism of this eVect is not known. ALDP does not have VLCS activity (Steinberg, S. J., and Watkins, P. A., unpublished data, 1998). In view of the transport function of the ABC

IV. DISEASES OF MYELIN

PATHOGENESIS

proteins, it has been postulated that ALDP is involved in the anchoring or transport of VLCS into the peroxisomal membrane (Contreras et al., 1994; Mosser et al., 1993). The subsequent studies of Steinberg et al. (1999) make this unlikely. They cloned VLCS and showed by immunocytochemical studies that VLCS in X-ALD Wbroblasts is localized in the peroxisome in an amount indistinguishable from that in controls. They also conducted topographical studies, which showed that VLCS in both X-ALD and control cells faces the peroxisomal surface of the membrane. If ALDP is required for VLCS translocation into the peroxisome, then VLCS would have been expected to be found on the exterior of the organelle in the X-ALD cells. They also studied the eVect of overexpression of either ALDP or VLCS in SV 40 transformed human X-ALD Wbroblasts. As noted previously, ALDP overexpression improved VLCFA oxidation, but it did not increase VLCS activity. VLCS overexpression did not alter VLCFA oxidation. However, combined overexpression of ALDP and VLCS showed a synergistic eVect that was statistically signiWcant. The mechanism of this eVect is not known. Braiterman et al. (1999) proposed that ALDP plays a role in the traYcking of VLCFA between microsomes and peroxisomes. This aspect has not been examined experimentally, but deserves study in view of the earlier reports by Tsuji et al. (Tsuji et al., 1981a) that VLCFA synthesis is increased in Wbroblasts of X-ALD patients (Koike et al., 1991). The possibility exists that in the absence of ALDP, VLCFA do not enter the peroxisome at the normal rate and that the microsomal elongation pathway is favored. McGuinness et al. have proposed recently that the primary eVect of ALDP is on mitochondrial metabolism (McGuinness et al., 2003). They propose that ALDP facilitates the interaction between peroxisomes and mitochondria, resulting in increased VLCFA levels when ALDP is deWcient.

Pathogenesis of the Nervous System Lesions In this section we consider separately the pathogenesis of the noninXammatory myelopathy of adrenomyeloneuropathy (AMN) and the inXammatory white matter response in the cerebral forms of the disease (CCER, adolescent and adult cerebral ALD, and AMNcerebral). The severe inXammatory white matter demyelinative sets X-ALD apart from the other leukodystrophies. It is the cause of rapid progression and rapidly fatal outcome. It is most common in childhood. However, approximately 40% of male patients and 99% of heterozygous women escape the inXammatory form of the disease. We hypothesize that all aVected males and 50% of heterozygous women would develop the noninXammatory AMN syndrome in adulthood, but that approximately 40% of males die in childhood or adolescence due to the inXammatory brain disease before AMN manifests. It should be noted, however, that while the clinical and pathological diVerences between childhood cerebral X-ALD and pure AMN are striking, the diVerence is not absolute. Detailed neuropathological studies may show mild inXammatory response in some pure AMN patients. Pathogenesis of Pure Adrenomyeloneuropathy The previously cited studies of Powers et al. (2000 and 2001) provide strong evidence that the primary disease process in AMN is a distal axonopathy. It is hypothesized that this is caused by impaired membrane stability and function secondary to the accumulation of VLCFA based upon the following observations: 1. VLCFA accumulation is the principal biochemical abnormality in X-ALD (Bizzozero et al., 1991; Brown et al., 1983; Igarashi et al., 1976b; Menkes and Corbo, 1977; Molzer et al., 1981; Ramsey et al., 1979; Reinecke et al., 1985; Taketomi et al., 1987; Theda et al., 1992; Wilson and Sargent, 1993). 2. VLCFA accumulation alters the biophysical properties and function of cell membranes. a. The desorption rate constant of saturated fatty acids from phospholipid membranes of VLCFA is much lower than that of shorter length fatty acids. The constant for C26:0 is 10,000 times lower than that for C16:0. b. While albumin has six or more high or low aYnity binding sites for fatty acids with 12 to 18 carbon chain length (Hamilton et al., 1991), it has only a single low aYnity

IV. DISEASES OF MYELIN

823

824

34. ADRENOLEUKODYSTROPHIES

binding site for C26:0 (Ho et al., 1995). Recent structural studies have shown that C26:0 cannot be accommodated in the binding groove of the high aYnity binding sites (Ho et al., 2002). c. The viscosity of red cell membranes in X-ALD and AMN patients is increased (Knazek et al., 1983). Normalization of VLCFA levels with ‘‘Lorenzo’s oil’’ (see section on therapy) also normalizes red blood cell membrane viscosity (HW Moser, unpublished observation). d. Microcalometric studies have shown that the inclusion of C26:0 in a model membrane disrupts membrane structure (Ho et al., 1995). e. VLCFA excess impairs function in cultured human adrenal cells. Whitcomb et al. (1988) assessed ACTH-stimulated cortisol release in cultured human adrenocortical cells. The addition of C26:0 or C24:0 to the culture medium in concentrations equivalent to those in X-ALD plasma increased microviscosity of adrenocortical cell membranes and decreased ACTH-stimulated cortisol secretion. It has also been hypothesized that gangliosides that contain excess VLCFA play a role in the axonal dysfunction in AMN (Powers et al., 2000). In gangliosides isolated from X-ALD brain fatty acids with chain lengths greater than C22:0 account for 28 to 50% of total fatty acids compared to 2.5% in controls (Igarashi et al., 1976a). Gangliosides are present in high concentration in plasma membranes and play a role in cell function (Zeller and Marchase, 1992). It is of interest that gangliosides containing ganglioside GT1b appear to be restricted to the axolemma (Sheikh et al., 1999). Excess VLCFA content may alter ganglioside function. For instance, gangliosides that contain VLCFA have less immunosuppressive activity than those with fatty acids. It is this ganglioside fraction that has the highest proportion of VLCFA (Igarashi et al., 1976a). At this time the evidence that the axonopathy in AMN is a consequence of VLCFA in excess is still indirect and conjectural and other pathogenetic mechanisms must be considered. As noted, McGuinness et al. (2003) and Ito et al. (2001) have demonstrated the existence of mitochondrial abnormalities in X-ALD and energy deWciencies or accumulation of oxygen radicals could contribute to axon dysfunction. Furthermore, since the function of ALDP is not yet understood, there may be disease mechanisms that have not yet been considered. Pathogenesis of the Inflammatory Response The most generally accepted concept for the pathogenesis of the inXammatory responses is that the accumulation of VLCFA has an adverse impact on myelin, and oligodendrocyte stability and function, which renders them vulnerable to various other adverse events (a ‘‘second hit’’), which then initiate a destruction cascade that results in the death of oligodendrocytes and rapid breakdown of myelin (Dubois-Dalcq et al., 1999; Feigenbaum et al., 2000; Ito et al., 2001; Powers et al., 1992). In the absence of a ‘‘second hit,’’ the nervous system pathology is mainly the distal axonopathy of AMN. Cytokines and immune mechanisms have been postulated to play a role in the initiation of the destructive cascade. DeWnitive conclusions are hampered by the lack of an animal model of the X-ALD inXammatory demyelination. Several investigators have implicated the role of cytokines in the inXammatory cascade associated with the cerebral forms of X-ALD. Powers et al. (1992) emphasized the presence of tumor necrosis factor alpha (TNF) in the astrocytes at the active edge of the lesion. TNF produced by stimulated astrocytes has been shown to be toxic to oligodendrocytes (Robbins et al., 1987). Immunocytochemical studies demonstrated increased interleukin1 and ICAM-1 expression in astrocytes and microvessels at the edge and within the lesion. However, a study in which cytokine gene expression patterns in X-ALD were considered by reverse transcriptase polymerase chain reaction showed only small increases and less than those in multiple sclerosis (McGuinness et al., 1997). Feigenbaum et al. (2000) reported apoptosis in the oligodendrocytes of actively demyelinating lesions of X-ALD patients and postulated that this was triggered or enhanced by inXammatory cells.

IV. DISEASES OF MYELIN

825

ANIMAL MODELS OF X-ALD

Ito et al. (2001) reexamined the pathogenesis of the inXammatory with the aid of recently developed antibodies. They demonstrated that most of the perivascular lymphocytes in the acute demyelinating lesions were CD8 cytotoxic T-cells, many of which showed strong immunoreactivity for granzyme B and CD44 and often showed intimate topographical association with oligodendrocytes. There was severe loss of oligodendrocytes. However, contrary to the previously cited Wndings of Feigenbaum et al. (2000), these authors concluded that in most instances the mechanisms of oligodendrocyte death was lytic/ cytolytic rather than apoptotic. An intriguing new Wnding is that there was strong CD1 immunoreactivity (particularly CD1b and CD1c) in astrocytes and microglia. CD1 molecules are antigen presenting surface glycoproteins that are encoded on chromosome 1, and unlike the MHC complex proteins can present self-lipid antigens to T-cells (Moody et al., 1999). This Wnding is of particular interest in respect to X-ALD, since these lipid antigens may contain VLCFA such as those associated with the mycobacterium tuberculosis. CD1d has a molecular conformation in which there is a hydrophobic groove that optimally accommodates lipid of approximately 32 carbon chain length. The authors postulate that CD1-lipid presentation may play a key role in the destructive cascade in cerebral X-ALD. As before, they consider that the primary event is biochemical membrane instability due to the VLCFA excess associated with the gene defect, and that this results in predisposition to some degree of spontaneous breakdown. They state that ‘‘prior to the second stage of fulminant inXammatory demyelination at least 3 additional events occur: the cytolytic killing of oligodendrocytes by CD8 cytotoxic T-cells, the MCH class 11-restricted presentation of peptide antigens by microglia, and the CD1-restricted presentation of lipid antigens’’ (Ito et al., 2001). These additional events then lead to the rapidly progressive demyelinative cascade. The CD1-lipid presentation may play a key role in this process. The presence of CD1b, c, and d in X-ALD suggests that several VLCFA-containing lipid classes are being presented. It is also of interest that gangliosides that contain VLCFA fatty acids bind eight times more avidly to anti-gangliosidase antibodies than those that contain 18 to 20 chain length fatty acids (Tagawa et al., 2002).

ANIMAL MODELS OF X-ALD Mice with targeted inactivation of the X-ALD gene have been produced independently in three laboratories (Forss-Petter et al., 1997; Kobayashi et al., 1997; Lu et al., 1997). All showed the same features. VLCFA levels were increased in the same tissue distribution, but somewhat less markedly than in the human disease. Excess was greatest in the brain and adrenal, with smaller increases in other tissues. Unlike the human disease, plasma VLCFA levels are normal in the X-ALD mouse (A. Moser, unpublished observation). Beta oxidation of VLCFA is decreased to the same extent as in the human disease, and the animal shows the characteristic needle-like inclusions in the adrenal cortex, testis and ovaries. During the Wrst year the growth, motor function, behavior and brain structure of the mouse model is entirely normal. However, at 15 months and later, the animals show impaired rotarod performance, moderately impaired nerve conduction velocity, and histological abnormalities of myelin and axons in the spinal cord and sciatic nerve, but not the brain (Pujol et al., 2002). It is concluded that the mouse presents a relatively mild model of ‘‘pure adrenomyeloneuropathy.’’ There has been no clinical evidence of adrenal insuYciency. The severe inXammatory brain disease has never been observed in the mouse model in spite of repeated attempts to induce it. This limits the value of the X-ALD mouse to serve as a model to test the eVectiveness of therapeutic interventions.

DIAGNOSIS OF X-ALD The diagnosis of X-ALD will be discussed in four diVerent settings: (1) symptomatic patients, (2) screening of at-risk family members, (3) prenatal diagnosis, and (4) mass

IV. DISEASES OF MYELIN

au5

826

34. ADRENOLEUKODYSTROPHIES

neonatal screening. Measurements of VLCFA is a key diagnostic technique in all four settings. Mutation analysis, when available, is of great value in settings 1 to 3, and of key importance for the identiWcation of heterozygotes. Neuroimaging studies are important in the Wrst setting. The plasma assay of VLCFA is the most commonly used diagnostic technique (Moser et al., 1999). It is reliable for the identiWcation of aVected males, irrespective of age, but false negative or borderline results are obtained in approximately 20% of heterozygotes. Mutation analysis is the most accurate method for the identiWcation of heterozygotes (Boehm et al., 1999). As noted previously, more than 400 diVerent mutations have been identiWed and often are unique to one kindred. Mutation analysis is now available on a service basis. The initial characterization of the mutation in a kindred requires approximately 4 to 6 weeks and is relatively expensive. It can be performed on blood samples of aVected males or obligate heterozygotes. Once the family mutation has been identiWed, at-risk family members can be screened for the presence of this mutation with a shorter turn around time and at lesser cost. Prenatal identiWcation of aVected male fetuses can be accomplished by measuring VLCFA levels in cultured amniocytes or chorion villus cells. A series of 255 studies performed at the Kennedy Krieger Institute up to 1998 identiWed 63 aVected male fetuses without known false negatives, subject to the caution that long-term follow-up was not available for some of the samples with normal results (Moser and Moser, 1999). However, the literature includes two reports of false negative results in male fetuses later found to have been aVected (Carey et al., 1994; Gray et al., 1995). When the family mutation has been deWned, the risk of false negatives can be minimized by performing mutation analysis in the fetal cells, and this procedure is recommended.

Diagnosis of Symptomatic Patients The diagnosis of X-ALD must be considered in boys with various neurological syndromes that manifest after 3 years of age. These include acquired attention deWcit disorders, or hyperactivity, progressive behavioral deWcits and progressive school failure, seizure disorders, disorders of vision or hearing, and progressive incoordination. An abnormal brain MRI study is often the initial reason for suspecting the diagnosis, with the MRI not infrequently having been performed for another reason, such as ruling out brain damage following an injury. The MRI abnormality precedes clinical manifestations, and abnormalities can be extensive even when clinical manifestations are mild. Approximately 80% of patients with the childhood cerebral form of X-ALD have the classical posterior patterns (Fig. 34.6) with symmetrical involvement of the parieto-occipital lobes and the splenium of the corpus callosum and contrast enhancement at the margins of the lesion (Kumar et al. 1987; Loes et al. in press). In 15% of the patients, the initial involvement is in the frontal lobes. Less common patterns are those with initial cerebellopontione involvement. Cerebral lesions may be unilateral and when this occurs have been mistaken for brain tumor. Even when the radiological pattern is classical, it is not speciWc for X-ALD and the diagnosis must be conWrmed by biochemical assay. A more complete discussion of the diVerential diagnosis is provided in a recent review article (Moser et al., 2000). A progressive myelopathy with paraparesis and sphincter disturbances in adults is the other common clinical presentation of X-ALD. When this presentation is combined with adrenal insuYciency, adrenomyeloneuropathy (AMN) is by far the most common cause and readily conWrmed by biochemical assay. However, in approximately 30% of patients with AMN, adrenal function is not demonstrably abnormal or is only mild. Under these circumstances, the diagnosis is often not made and patients may be misdiagnosed as having progressive multiple sclerosis. We recommend that plasma levels of VLCFA be measured in all patients with progressive myelopathy. The diagnosis of the AMN-like syndrome in women heterozygous for X-ALD is a challenge. Less than 1% have adrenal insuYciency, it presents most commonly in middle age or later, and it must be distinguished from many other causes of myelopathy. Furthermore, the biochemical abnormality in plasma is not as marked as in aVected males. False negative tests occur in approximately 20% of heterozygotes. Most of the women with this syndrome have been identiWed because they were

IV. DISEASES OF MYELIN

DIAGNOSIS OF X-ALD

FIGURE 34.6 ‘‘Classical’’ MRI pattern in a patient with the childhood cerebral form of X-ALD. T-1 weighted image obtained following intravenous injection of pentaacetic acid. The symmetric regions of decreased signal intensity in the parieto-occipital regions are indicative of loss of myelin and gliosis. The garland of increased signal density that surrounds these regions is the zone in which the gadolinium contrast material has accumulated due to the breakdown of the blood-brain barrier associated with the inXammatory response. From Radiology 165, 496, 1987, with permission

relatives of aVected male patients. However, as awareness about this syndrome has increased, more women without known aVected relatives are being diagnosed. This is important because it increases the opportunity for genetic counseling. Primary adrenal insuYciency without neurological involvement, the ‘‘Addison-only’’ phenotype is the third mode of clinical presentation of X-ALD. It cannot be distinguished clinically from other forms of primary adrenocortical insuYciency. It has been estimated that X-ALD is the cause of adrenal insuYciency in 35% patients with idiopathic Addison disease that manifested before 7.5 years of age) (Jorge et al., 1993; Laureti et al., 1996). It is recommended that plasma VLCFA levels be measured in all males with adrenal insuYciency of unknown cause.

Extended Family Screening Extended family screening programs have identiWed many asymptomatic males with X-ALD and many heterozygous women. This is clinically important, since many of the asymptomatic males have unrecognized adrenal insuYciency and can be started on appropriate hormone replacement therapy before they develop clinical symptoms. Furthermore, as discussed in the next section, current therapy for the neurological manifestations (bone marrow transplantation) is most eVective when it is initiated when involvement is still mild, and dietary therapy in young asymptomatic patients appears to reduce the risk of developing the childhood cerebral disease. IdentiWcation of heterozygotes is of key importance for genetic counseling. Bezman et al. (2001) used the plasma VLCFA assay to screen 4169 atrisk members of the extended families of known X-ALD patients and identiWed 594 aVected males, 250 of whom were asymptomatic, and 1270 heterozygous women.

Prenatal Diagnosis Prenatal identiWcation can be achieved by VLCFA analysis in cultured amniocytes or chorion villus cells (Moser and Moser, 1999). The risk of false negatives can be reduced by

IV. DISEASES OF MYELIN

827

828

34. ADRENOLEUKODYSTROPHIES

immunoXuorescence assay (Ruiz et al., 1997). Mutation analysis is the preferred procedure when the mutation in an aVected member of the family has been deWned.

Mass Neonatal Screening Plasma VLCFA levels in aVected males are increased already on the day of birth (Moser et al.). Studied to determine the feasibility of screening all newborn males with tandem mass spectrometry techniques are now in progress.

TREATMENT The design of therapy of the neurological manifestation X-ALD is hampered by the incomplete understanding of its pathogenesis: The function of the defective gene product, ALDP, is still not understood, nor is the pathogenesis of the inXammatory demyelination that causes the rapid disease progression. The development of an animal model that displays the inXammatory demyelination would enhance greatly the capacity to design and evaluate therapeutic intervention. This section will begin with an appraisal of the therapies that are in current use and cite brieXy to therapies that have been proposed for the future. It should be noted that the development of mass neonatal screening, which may become available during the next 5 years, would have a profound eVect on therapeutic strategies: nervous system function in X-ALD patients does not become abnormal until 2 to 3 years of age, and often considerably later. Neonatal screening thus has the potential of detecting all aVected males years before they develop nervous system dysfunction and therapeutic approaches could then focus on the prevention of nervous system damage.

Adrenal Steroid Replacement Therapy Adrenal steroid replacement therapy is mandatory and life saving. Even though relatively simple, this form of therapy is often not properly implemented, and preventable deaths due to adrenal crisis continues to occur. Glucocorticoid dose requirements are generally those used for other forms of adrenal insuYciency. Most patients do not require mineralocorticoid replacement. While the adrenal steroid replacement therapy can improve strength and well being, it generally does not appear to alter progression of neurological disability. However, neurologic improvement coincident with glucocorticoid replacement has been reported in some cases (Peckham et al., 1982; Zhang and Moser, 2003).

Bone Marrow Transplantation-BeneWt in Patients with Early Cerebral Involvement It has been shown that allogeneic bone marrow transplantation (BMT) can prevent further progression and occasionally reverse deWcits in children and adolescents with cerebral X-ALD who received the transplant when brain involvement was still relatively mild (Aubourg et al., 1990; Malm et al., 1997; Shapiro et al., 2000). Some of the patients have now been followed for 10 years and have normal cognitive function and the brain MRI abnormality has remained stable, and in one patient it disappeared (Aubourg et al., 1990; Shapiro et al., 2000). However, when the cerebral involvement is already advanced, the eVect of BMT is not favorable. The deWcit may increase during the days and weeks following the transplant, and this may lead to fatal outcome. When long-term stabilization occurs in patients with advanced disability, the quality of life often is impaired to such an extent that the ethical justiWcation for the procedure is questionable. It is not known whether BMT has a favorable eVect in adult patients with AMN, the noninXammatory form of X-ALD. It has not been performed in pure AMN patients because the risk of graft versus host disease (GVH) is higher than in children and adolescents, and in view of the slow progression of AMN, the risk/beneWt ratio is not favorable. A point of key interest will be whether the children who had been transplanted successfully for cerebral ALD will

IV. DISEASES OF MYELIN

TREATMENT

develop AMN in adulthood. While there is no doubt that BMT is the most eVective therapy currently available for children and adolescents, and possibly young adults with cerebral X-ALD, patients must be selected with great care so that advantage is taken of the relatively small ‘‘window of opportunity.’’ This window can be identiWed by serial MRI and neuropsychological studies. The most favorable patients are those who show evidence of relatively mild but progressive MRI abnormalities and still have a performance intelligence quotient (PIQ) of 80 or above (Shapiro et al., 2000). These conditions are encountered most often in the follow-up of asymptomatic patients identiWed by family screening. Unfortunately, many children and adolescents who are identiWed because of clinically identiWed abnormalities already have such advanced disabilities that they are no longer considered to be candidates for BMT. The mechanism of the favorable eVect of BMT is not yet clear. The BMT-derived cells have the capacity to metabolize VLCFA (Moser et al., 1984), and the plasma VLCFA levels are reduced, although not normalized (Shapiro et al., 2000). Unlike the lysosomal disorders, where the normal gene product is excreted by donor cells and can be taken up by host cells, this is not applicable to X-ALD because the transfer of a peroxisomal membrane protein from donor to host cells is not possible. The favorable eVect in X-ALD brain may result from the donor-derived microglia. Microglia are bone marrow derived at least in part (Hickey and Kimura, 1988), and BMT-derived cells do enter the nervous system (Unger et al., 1993). The donor-derived microglia may have a favorable eVect on local brain metabolism. Microglia have a slow turnover rate (Hickey et al., 1992). This could account for the Wndings that BMT beneWcial eVects are not observed until 6 to 12 months after BMT and may continue to increase thereafter (Aubourg et al., 1990). An alternative or additive mechanism for the favorable eVect of BMT may be due to the immunosuppression that forms part of the preparative regimen. So far, the beneWcial eVects of BMT have been documented only in patients with inXammatory demyelination. BMT has been shown to abolish the accumulation of contrast material in MRI studies, which is an index of the inXammatory response (Charnas, L., unpublished observation). Were this to be the case, then an immunosuppressive regimen alone could be beneWcial. Against this hypothesis is that immunosuppressive regimens tested so far have not been beneWcial (discussed later), and that no improvement was observed in a patient who received a BMT who had been immunosuppressed but failed to engraft (Nowaczyk et al., 1997).

Reduction of VLCFA Levels by Dietary Therapy Oral administration of a 4:1 mixture of glyceryl-trioleate, and glyceryl-trierucate, also referred to as ‘‘Lorenzo’s oil,’’ when combined with reduction of fat intake, normalizes plasma VLCFA levels in X-ALD patients within 4 weeks (Odone and Odone, 1989; Rizzo et al., 1989), probably by competitive inhibition with the microsomal elongating system for saturated long chain fatty acids (Bourre et al., 1976). In spite of this striking biochemical eVect, it does not appear to alter disease progression in patients who already are neurologically symptomatic (Aubourg et al., 1993; Kaplan et al., 1993; Uziel et al., 1991; van Geel et al., 1999). The lack of clinical beneWt has been attributed to the apparently low rate of entry of erucic acid, the active component of Lorenzo’s oil’’ into the brain (Poulos et al., 1994; Rasmussen et al., 1994). This clinical experience in several clinical centers combined with the observation that about 30% of treated patients developed moderate thrombocytopenia (Zinkham et al., 1993) and a lesser incidence of other side eVects (van Geel et al., 1999) led to the consensus that this therapy is not indicated in patients who are already symptomatic. An international multicenter study to evaluate the preventive eVect of Lorenzo’s oil has been conducted. Enacted since 1989, the study involves Wve centers in the United States and two in Europe and included a total of 104 boys who all had proven X-ALD, were less than 6 years old, and had normal neurological examination and MRI. They received Lorenzo’s oil in accordance with a previously described protocol. Lorenzo’s oil therapy combined with reduced fat intake from other sources, was initiated between 1 and 6 years of age using a regimen that has been described previously (Moser and Borel, 1995). Outcome measurements were the time of development of neurologic and MRI

IV. DISEASES OF MYELIN

829

830

34. ADRENOLEUKODYSTROPHIES

abnormalities, which were evaluated separately independently by standardized criteria. An open study rather than a randomized placebo-controlled study design was selected after ethical advice because of the devastating nature of childhood cerebral X-ALD combined with the possibility that clinical beneWt might result from reduction of plasma VLCFA levels. Two criteria were used to evaluate preventive eVect: (1) comparison of age of onset of neurologic symptoms in the treated group (1989–1999) with that in historical controls; and (2) correlation between neurologic outcome and MRI progression and the degree of reduction of plasma VLCFA levels, a measure of compliance. Analysis of these data is complex because of the nonrandomized study design and incomplete information about historical controls. Preliminary appraisal of the data by Hugo Moser and Ann Moser presented at a meeting on peroxisomal diseases in 2002 (Moser and Moser, 2003) indicated that the oil diminished the subsequent risk of neurological involvement. However, additional analysis of the data is required for deWnitive conclusion. It was recommended that this therapy be oVered to boys in this category, subject to the caution that the beneWt may represent a delay in disease onset rather than prevention and that some patients did develop neurological involvement in spite of good control of plasma levels. It was emphasized that the therapy be supervised by a multidisciplinary team to help assure normal growth and development and prevent complications such as reduction of platelet count. Adrenal function must be monitored and deWciency treated by adrenal steroid therapy. Levels of essential and polyunsaturated fatty acids, including docosahexaenoic acid, must be monitored and appropriate supplementation provided. Finally, it is of crucial importance that brain MRI be monitored so that those patients who can beneWt from bone marrow transplant be identiWed in a timely manner.

Phenylbutyrate—Favorable Biochemical Effect in X-ALD Mouse Model Kemp et al. (1998) reported that 4-phenylbutyrate (4PB) normalizes VLCFA levels and restores VLCFA oxidation in cultured skin Wbroblasts of patients with X-ALD and of X-ALD mice. They also made the important observation that oral administration to X-ALD mice reduced substantially levels of VLCFA in the brain and adrenal glands of these animals. The mechanism of this eVect is not yet fully understood. The authors made the intriguing observation that 4PB increased the expression of ALDR (ABCD2), a homologue of ABCD1 (the gene that is defective in X-ALD). ALDR can substitute at least in part for the eVects of ALDP on VLCFA metabolism in X-ALD cells (Fig. 34.7) (Kemp et al., 1998; Netik et al., 1999). In their initial publication, these authors suggested that 4PB therapy here could be an example of pharmacological gene therapy made possible by the existence of gene redundancy. However, later studies by the same group led them to conclude that this is not the mechanism (McGuinness et al., 2001), based on their observation that the eVect on VLCFA levels preceded the increased expression of ALDR. The authors found that 4PB has a rapid but as yet not fully understood eVects on both peroxisomal and mitochondrial fatty acid oxidation and that the reduction of VLCFA levels in X-ALD cells appears to be a consequence of this eVect. 4PB therapy has not been evaluated systematically in X-ALD patients. Moser et al. (unpublished observation) conducted a phase I study in six men with AMN who received oral 4PB at a dose of 20 grams per day for two 6-week periods. While platelet VLCFA levels were reduced by approximately 50%, levels of VLCFA in plasma and red cells were unchanged. VLCFA oxidation in white blood cells and the expression of ALDR were not increased in white blood cells. There were no adverse eVects. Clinical Wndings were not changed, but this was not expected over such a short time in patients with the slowly progressive pure AMN. Systematic long-term clinical trials of 4PB therapy in X-ALD have not been conducted.

Immunosuppression The aim of immunosuppression therapy is to abolish the inXammatory response in cerebral X-ALD and thus convert this rapidly progressive form into the milder pure AMN pheno-

IV. DISEASES OF MYELIN

TREATMENT

FIGURE 34.7 Peroxisomal ABC half-transporter complementation of C24:0 ß-oxidation. Human X-ALD Wbroblasts transformed with SV40T antigen were transfected with recombinant expression vector (pCDNA3) alone or with vectorcontaining cDNA for PMP70, ALDP, or ALDRP (hatched bars). The rates of C24:0 ß-oxidation in the transfected cells were corrected for the fraction of cells expressing the transgene, as determined by immunoXuorescence staining of the transgene. The adjusted rates were compared with the rates of C24:0 ß-oxidation determined in transformed Wbroblasts from normal individuals (Wlled bar). The indicated values are the mean and standard deviation for pCDNA3 (n ¼ 6); ALDP (n ¼ 5); ALDRP (n ¼ 4); and normal m (n ¼ 5). PMP70, n ¼ 1. From Nature Medicine 4, 1261, 1998.

type. Agents tested so far are cyclophophamide (Naidu et al., 1988; Stumpf et al., 1981), beta interferon (Korenke et al., 1997), and Rolipram (Netik et al., 1999). They have not been successful. Other recently developed approaches that have shown promise in the treatment of multiple sclerosis, such as the alpha 4 integrin antagonist Natalizumab (Miller et al., 2003), have not been tested. Better understanding of the pathogenesis of the inXammatory response in X-ALD may lead to more eVective immunosuppressive therapies, and in view of the devastating eVect of the inXammatory response, this represents a high priority.

Lovastatin Singh and associates have proposed lovastatin as a possible therapeutic agent. They have shown that lovastatin increases the capacity of cultured X-ALD cells to metabolize tetracosanoic acid (C24:0) and that it normalized the levels of VLCFA in these cells (Singh et al., 1998). Oral administration of lovastatin in a dosage of 40 mg per day lowered the levels of VLCFA in the plasma of X-ALD patients (Pai et al., 2000; Singh et al., 1998), but the lowering is not as consistent as that achieved with Lorenzo’s oil (Moser, A. B., unpublished observation). Lovastatin administration in X-ALD mice did not reduce VLCFA levels in their brain (Cartier et al., 2000; Yamada et al., 2000). Weinhofer et al. have shown that lowering cholesterol leads to activation of sterol regulatory element binding protein and increased expression of the ABCD2 gene and reduced VLCFA accumulation in cultured

IV. DISEASES OF MYELIN

831

832

34. ADRENOLEUKODYSTROPHIES

ABCD1 cells. This may account for the VLCFA lowering action of lovastatin (Weinhofer et al., 2002). Lovastatin has another action, which may be relevant to X-ALD. It has been shown to reduce the inXammatory demyelination in experimental allergic encephalitis in Lewis rats (Stanislaus et al., 1999), at least in part due to its induction of nitric oxide synthase (Pahan et al., 1997). Nitric oxide synthase induction appears to play a role in the pathogenesis of the inXammatory response in X-ALD (Gilg et al., 2000). Lovastatin did not appear to alter neurologic progression in brief noncontrolled clinical trials (Pai, et al., 2000), but longer-term controlled trials have not been conducted.

Gene Replacement Therapy Retroviral transfer of the X-ALD gene into C34þ cells of X-ALD patients has been achieved (DoerXinger et al., 1998). A higher and more persistent transduction has been achieved with lenti-virus (Aubourg, P., personal communication) and warrants consideration of ex vivo therapy with bone marrow derived cells. However, such trials must be deferred until the safety of this vector has been examined in greater detail.

PREVENTION OF X-ALD In spite of emerging therapies, X-ALD is a devastating experience to the patients and families. The importance of disease prevention through genetic counseling cannot be overemphasized.

References Adams R. D., and Kubik C. S. (1952). The morbid anatomy of the demyelinative diseases. Am J Med 12, 510–546. Antoku Y., Sakai T., Tsukamoto K., Imanishi K., Ohtsuka Y., Iwashita H., and Goto I. (1987). A simple diagnostic method of adrenoleukodystrophy: Total fatty acid analysis of erythrocyte membranes. Clin Chim Acta 169, 121–125. Aubourg P., Adamsbaum C., Lavallard-Rousseau M. C., Rocchiccioli F., Cartier N., Jambaque I., Jakobezak C., Lemaitre A., Boureau F., Wolf C., and Bougneres, P. F. (1993). A two-year trial of oleic and erucic acids (‘‘Lorenzo’s oil’’) as treatment for adrenomyeloneuropathy. N Engl J Med 329, 745–752. Aubourg P., Blanche S., Jambaque I., Rocchiccioli F., Kalifa G., Naud-Saudreau C., Rolland M. O., Debre M., Chaussain J. L., Griscelli C., Fischer, A., and Bougneres, P. F. (1990). Reversal of early neurologic and neuroradiologic manifestations of X-linked adrenoleukodystrophy by bone marrow transplantation. N Engl J Med 322, 1860–1866. Aversa A., Palleschi S., Cruccu G., Silvestroni L., Isidori A., and Fabbri A. (1998). Rapid decline of fertility in a case of adrenoleukodystrophy. Hum Reprod 13, 2474–2479. Berger J., Molzer B., Fae I., and Bernheimer H. (1994). X-linked adrenoleukodystrophy (ALD).: A novel mutation of the ALD gene in 6 members of a family presenting with 5 diVerent phenotypes. Biochem Biophys Res Commun 205, 1638–1643. Bezman L., Moser A. B., Raymond G. V., Rinaldo P., Watkins P. A., Smith K. D., Kass N. E., and Moser H. W. (2001). Adrenoleukodystrophy: Incidence, new mutation rate, and results of extended family screening. Ann Neurol 49, 512–517. Bizzozero O. A., Zuniga G., and Lees M. B. (1991). Fatty acid composition of human myelin proteolipid protein in peroxisomal disorders. J Neurochem 56, 872–878. Black P. N., Zhang Q., Weimar J. D., and DiRusso C. C. (1997). Mutational analysis of a fatty acyl-coenzyme A synthetase signature motif identiWes seven amino acid residues that modulate fatty acid substrate speciWcity. J Biol Chem 272, 4896–4903. Blaw M. E. (1970). Melanodermic type leucodystrophy (adrenoleukodystrophy). In ‘‘Leucodystrophies and Poliodystrophies’’ (Vinken P. J., and Bruyn G. W. E., eds.), pp. 128–133. North Holland Publishing Company, Amsterdam. Blaw M. E., Osterberger K., Kozak P., and Nelson E. (1964). Sudanophilic leukodystrophy and adrenal cortical atrophy. Arch Neurol 11, 626–631. Boehm C. D., Cutting G. R., Lachtermacher M. B., Moser H. W., and Chong S. S. (1999). Accurate DNA-based diagnostic and carrier testing for X-linked adrenoleukodystrophy. Mol Genet Metab 66, 128–136.

IV. DISEASES OF MYELIN

au8 au9

au10

833

PREVENTION OF X-ALD

Bourre J. M., Daudu O., and Baumann N. (1976). Nervonic acid biosynthesis by erucyl-CoA elongation in normal and quaking mouse brain microsomes. Elongation of other unsaturated fatty acyl-CoAs (mono and poly-unsaturated). Biochim Biophys Acta 424, 1–7. Braiterman L. T., Watkins P. A., Moser A. B., and Smith K. D. (1999). Peroxisomal very long chain fatty acid beta-oxidation activity is determined by the level of adrenodeukodystrophy protein (ALDP). expression. Mol Genet Metab 66, 91–99. Braiterman L. T., Zheng S., Watkins P. A., Geraghty M. T., Johnson G., McGuinness M. C., Moser A. B., and Smith K. D. (1998). Suppression of peroxisomal membrane protein defects by peroxisomal ATP binding cassette (ABC). proteins. Hum Mol Genet 7, 239–247. Brennemann W., Kohler W., Zierz S., and Klingmuller D. (1996). Occurrence of adrenocortical insuYciency in adrenomyeloneuropathy. Neurology 47, 605. Bresnan M. J., and Richardson E. P. (1979). Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 18–1979. N Engl J Med 300, 1037–1045. Broccardo C., TroVer-Charlier N., Savary S., Mandel J. L., and Chimini G. (1998). Exon organisation of the mouse gene encoding the Adrenoleukodystrophy related protein (ALDRP). Eur J Hum Genet 6, 638–641. Brown F. R. 3rd, Chen W. W., Kirschner D. A., Frayer K. L., Powers J. M., Moser A. B., and Moser H. W. (1983). Myelin membrane from adrenoleukodystrophy brain white matter–biochemical properties. J Neurochem 41, 341–348. Brun A., and Voigt G. E. (1960). Entzundliche cerebrale sklerose mit nebenniereninsuYzienz. Deutsche Zellschrift F. Nervenheilkunde 180, 654–664. Budka H., Sluga E., and Heiss W. D. (1976). Spastic paraplegia associated with Addison’s disease: Adult variant of adreno-leukodystrophy. J Neurol 213, 237–250. Carey W. F., Poulos A., Sharp P., Nelson P. V., Robertson E. F., Hughes J. L., and Gill A. (1994). Pitfalls in the prenatal diagnosis of peroxisomal beta-oxidation defects by chorionic villus sampling. Prenat Diagn 14, 813–819. Cartier N., Guidoux S., Rocchiccioli F., and Aubourg P. (2000). Simvastatin does not normalize very long chain fatty acids in adrenoleukodystrophy mice. FEBS Lett 478, 205–208. Cassagne C., Darriet D., and Bourre J. M. (1978). Biosynthesis of very long chain fatty acids by the sciatic nerve of the rabbit. FEBS Lett 90, 336–340. Chaudhry V., Moser H. W., and Cornblath D. R. (1996). Nerve conduction studies in adrenomyeloneuropathy. J Neurol Neurosurg Psychiatry 61, 181–185. Cox R. D. (1972). Regression models and life tables. J Royal Statistics Society 34, 187. Dean M., Hamon Y., and Chimini G. (2001). The human ATP-binding cassette (ABC). transporter superfamily. J Lipid Res 42, 1007–1017. DoerXinger N., Miclea J. M., Lopez J., Chomienne C., Bougneres P., Aubourg P., and Cartier N. (1998). Retroviral transfer and long-term expression of the adrenoleukodystrophy gene in human CD34þ cells. Hum Gene Ther 9, 1025–1036. Dubois-Dalcq M., Feigenbaum V., and Aubourg P. (1999). The neurobiology of X-linked adrenoleukodystrophy, a demyelinating peroxisomal disorder. Trends Neurosci 22, 4–12. Edwin D., Speedie L. J., Kohler W., Naidu S., Kruse B., and Moser H. W. (1996). Cognitive and brain magnetic resonance imaging Wndings in adrenomyeloneuropathy. Ann Neurol 40, 675–678. Fanconi V. A., Prader A., Isler W., Luthy F., and Siebenmann R. (1963). Morbus Addison mit hirnsklerose im kindesalter–Ein hereditares syndrom mit X-chromosomaler vererbung? Helv Paediatr Acta 18, 480–501. Feigenbaum V., Gelot A., Casanova P., Daumas-Duport C., Aubourg P., and Dubois-Dalcq M. (2000). Apoptosis in the central nervous system of cerebral adrenoleukodystrophy patients. Neurobiol Dis 7, 600–612. Feigenbaum V., Lombard-Platet G., Guidoux S., Sarde C. O., Mandel J. L., and Aubourg P. (1996). Mutational and protein analysis of patients and heterozygous women with X-linked adrenoleukodystrophy. Am J Hum Genet 58, 1135–1144. Flavigny E., Sanhaj A., Aubourg P., and Cartier N. (1999). Retroviral-mediated adrenoleukodystrophy-related gene transfer corrects very long chain fatty acid metabolism in adrenoleukodystrophy Wbroblasts: Implications for therapy. FEBS Lett 448, 261–264. Forss-Petter S., Werner H., Berger J., Lassmann H., Molzer B., Schwab M. H., Bernheimer H., Zimmermann F., and Nave K. A. (1997). Targeted inactivation of the X-linked adrenoleukodystrophy gene in mice. J Neurosci Res 50, 829–843. Gagnon J. L. R. (1959). Sclerose cerebral diVuse avec melanodermie et atrophie surrenale. Union Med Canada 88, 392–418. Garside S., Rosebush P. I., Levinson A. J., and Mazurek M. F. (1999). Late-onset adrenoleukodystrophy associated with long-standing psychiatric symptoms. J Clin Psychiatry 60, 460–468. Gilg A. G., Singh A. K., and Singh I. (2000). Inducible nitric oxide synthase in the central nervous system of patients with X-adrenoleukodystrophy. J Neuropathol Exp Neurol 59, 1063–1069. Gould S. J., Raymond G. V., and Valle D. (2001). The Peroxisome Biogenesis Disorders. In ‘‘The Metabolic and Molecular Bases of Inherited Diseases’’ (Scriver C. R., Beaudet A. L., Sly W. S., and Valle D., eds.), pp. 3181–3217. McGraw Hill, New York. Gray R. G., Green A., Cole T., Davidson V., Giles M., Schutgens R. B., and Wanders R. J. (1995). A misdiagnosis of X-linked adrenoleukodystrophy in cultured chorionic villus cells by the measurement of very long chain fatty acids. Prenat Diagn 15, 486–490.

IV. DISEASES OF MYELIN

au11

834

34. ADRENOLEUKODYSTROPHIES

GriYn J. W., Goren E., Schaumburg H., Engel W. K., and Loriaux L. (1977). Adrenomyeloneuropathy: A probable variant of adrenoleukodystrophy. I. Clinical and endocrinologic aspects. Neurology 27, 1107–1113. Hamilton J. A., Era S., Bhamidipati S. P., and Reed R. G. (1991). Locations of the three primary binding sites for long-chain fatty acids on bovine serum albumin. Proc Natl Acad Sci U S A 88, 2051–2054. Hampel E. (1937). Morbus Addisonii und sklerosierende erkrankung des hemispharenmarks. Beitrag zu den hirnveranderungen beim morbus Addisonii und Zum kapitel der diVusen sklerosen. Dtsch Z Nervenheilk 142, 186–208. Hashimoto T. (1996). Peroxisomal beta-oxidation: Enzymology and molecular biology. Ann N Y Acad Sci 804, 86–98. Hashmi M., Stanley W., and Singh I. (1986). Lignoceroyl-CoASH ligase: Enzyme defect in fatty acid betaoxidation system in X-linked childhood adrenoleukodystrophy. FEBS Lett 196, 247–250. Heinzer A. K., Kemp S., Lu J. F., Watkins P. A., and Smith K. D. (2002). Mouse very long-chain acyl-CoA synthetase in X-linked adrenoleukodystrophy. J Biol Chem 277, 28765–28773. Hershkovitz E., Narkis G., Shorer Z., Moser A. B., Watkins P. A., Moser H. W., and Manor E. (2002). Cerebral X-linked adrenoleukodystrophy in a girl with Xq27-ter deletion. Ann Neurol 52, 234–237. Hickey W. F., and Kimura H. (1988). Perivascular microglial cells of the CNS are bone marrow-derived and present antigen in vivo. Science 239, 290–292. Hickey W. F., Vass K., and Lassmann H. (1992). Bone marrow-derived elements in the central nervous system: An immunohistochemical and ultrastructural survey of rat chimeras. J Neuropathol Exp Neurol 51, 246–256. Higgins C. F. (1992). ABC transporters: From microorganisms to man. Annu Rev Cell Biol 8, 67–113. Ho J. K., Duclos R. I. Jr and Hamilton J. A. (2002). Interactions of acyl carnitines with model membranes: A (13).C-NMR study. J Lipid Res 43, 1429–1439. Ho J. K., Moser H., Kishimoto Y., and Hamilton J. A. (1995). Interactions of a very long chain fatty acid with model membranes and serum albumin. Implications for the pathogenesis of adrenoleukodystrophy. J Clin Invest 96, 1455–1463. Hoefnagel D., and Van Den Noort S. (1962). DiVuse cerebral sclerosis with endocrine abnormalities in young males. Brain 85, 553–568. Holzinger A., Kammerer S., and Roscher A. A. (1997). Primary structure of human PMP69, a putative peroxisomal ABC-transporter. Biochem Biophys Res Commun 237, 152–157. Hughes J. T., Brownell B., and Hewer R. L. (1968). The peripheral sensory pathway in friedreich’s ataxia. An examination by light and electron microscopy of the posterior nerve roots, posterior root ganglia, and peripheral sensory nerves in cases of friedreich’s ataxia. Brain 91, 803–18. Igarashi M., Belchis D., and Suzuki K. (1976a). Brain gangliosides in adrenoleukodystrophy. J Neurochem 27, 327–328. Igarashi M., Schaumburg H. H., Powers J., Kishmoto Y., Kolodny E., and Suzuki K. (1976b). Fatty acid abnormality in adrenoleukodystrophy. J Neurochem 26, 851–860. Ito M., Blumberg B. M., Mock D. J., Goodman A. D., Moser A. B., Moser H. W., Smith K. D., and Powers J. M. (2001). Potential environmental and host participants in the early white matter lesion of adreno-leukodystrophy: Morphologic evidence for CD8 cytotoxic T cells, cytolysis of oligodendrocytes, and CD1-mediated lipid antigen presentation. J Neuropathol Exp Neurol 60, 1004–1019. Jorge P., Quelhas D., and Nogueira A. (1993). Characterization of X-linked adrenoleukodystrophy in diVerent biological specimens from ten Portuguese families. J Inherit Metab Dis 16, 55–62. Jorge P., Quelhas D., Oliveira P., Pinto R., and Nogueira A. (1994). X-linked adrenoleukodystrophy in patients with idiopathic Addison disease. Eur J Pediatr 153, 594–597. Julien J. J., Vallat J. M., Vital C., Lagueny A., Ferrer X., and Darriet D. (1981). Adrenomyeloneuropathy: Demonstration of inclusions at the level of the peripheral nerve. Eur Neurol 20, 367–373. Kamijo K., Taketani S., Yokota S., Osumi T., and Hashimoto T. (1990). The 70-kDa peroxisomal membrane protein is a member of the Mdr (P-glycoprotein).-related ATP-binding protein superfamily. J Biol Chem 265, 4534–4540. Kaplan P. W., Tusa R. J., ShankroV J., Heller J., and Moser H. W. (1993). Visual evoked potentials in adrenoleukodystrophy: A trial with glycerol trioleate and Lorenzo oil. Ann Neurol 34, 169–174. Kelley R. I., Datta N. S., Dobyns W. B., Hajra A. K., Moser A. B., Noetzel M. J., Zackai E. H., and Moser H. W. (1986). Neonatal adrenoleukodystrophy: New cases, biochemical studies, and diVerentiation from Zellweger and related peroxisomal polydystrophy syndromes. Am J Med Genet 23, 869–901. Kemp S., Ligtenberg M. J., van Geel B. M., Barth P. G., Wolterman R. A., Schoute F., Sarde C. O., Mandel J. L., van Oost B. A., and Bolhuis P. A. (1994). IdentiWcation of a two base pair deletion in Wve unrelated families with adrenoleukodystrophy: A possible hot spot for mutations. Biochem Biophys Res Commun 202, 647–653. Kemp S., Pujol A., Waterham H. R., van Geel B. M., Boehm C. D., Raymond G. V., Cutting G. R., Wanders R. J., and Moser H. W. (2001). ABCD1 mutations and the X-linked adrenoleukodystrophy mutation database: Role in diagnosis and clinical correlations. Hum Mutat 18, 499–515. Kemp S., Wei H. M., Lu J. F., Braiterman L. T., McGuinness M. C., Moser A. B., Watkins P. A., and Smith K. D. (1998). Gene redundancy and pharmacological gene therapy: Implications for X-linked adrenoleukodystrophy. Nat Med 4, 1261–1268. Knazek R. A., Rizzo W. B., Schulman J. D., and Dave J. R. (1983). Membrane microviscosity is increased in the erythrocytes of patients with adrenoleukodystrophy and adrenomyeloneuropathy. J Clin Invest 72, 245–248. Kobayashi T., Shinnoh N., Kondo A., and Yamada T. (1997). Adrenoleukodystrophy protein-deWcient mice represent abnormality of very long chain fatty acid metabolism. Biochem Biophys Res Commun 232, 631–636.

IV. DISEASES OF MYELIN

PREVENTION OF X-ALD

au12

au13

Koike R., Tsuji S., Ohno T., Suzuki Y., Orii T., and Miyatake T. (1991). Physiological signiWcance of fatty acid elongation system in adrenoleukodystrophy. J Neurol Sci 103, 188–194. Kok F., Neumann S., Sarde C. O., Zheng S., Wu K. H., Wei H. M., Bergin J., Watkins P. A., Gould S., Sack G., Moser, H., Mandel, J-L., and Smith, K. D. (1995). Mutational analysis of patients with X-linked adrenoleukodystrophy. Hum Mutat 6, 104–115. Korenke G. C., Christen H. J., Kruse B., Hunneman D. H., and Hanefeld F. (1997). Progression of X-linked adrenoleukodystrophy under interferon-beta therapy. J Inherit Metab Dis 20, 59–66. Kumar A. J., Kohler W., Kruse B., Naidu S., Bergin A., Edwin D., and Moser H. W. (1995). MR Wndings in adult-onset adrenoleukodystrophy. AJNR Am J Neuroradiol 16, 1227–1237. Kumar A. J., Rosenbaum A. E., Naidu S., Wener L., Citrin C. M., Lindenberg R., Kim W. S., Zinreich S. J., Molliver M. E., Mayberg H. S., and Moser, H. W. (1987). Adrenoleukodystrophy: Correlating MR imaging with CT. Radiology 165, 497–504. Kurihara M., Kumagai K., Yagishita S., Imai M., Watanabe M., Suzuki Y., and Orii T. (1993). Adrenoleukomyeloneuropathy presenting as cerebellar ataxia in a young child: A probable variant of adrenoleukodystrophy. Brain Dev 15, 377–380. Laureti S., Casucci G., Santeusanio F., Angeletti G., Aubourg P., and Brunetti P. (1996). X-linked adrenoleukodystrophy is a frequent cause of idiopathic Addison’s disease in young adult male patients. J Clin Endocrinol Metab 81, 470–474. Lazo O., Contreras M., Hashmi M., Stanley W., Irazu C., and Singh I. (1988). Peroxisomal lignoceroyl-CoA ligase deWciency in childhood adrenoleukodystrophy and adrenomyeloneuropathy. Proc Natl Acad Sci U S A 85, 7647–7651. Lichtenstein B. W., and Rosenbluth P. R. (1959). Schilder’s disease with melanoderma. J Neuropathol Exp Neurol 18, 384–396. Liu L. X., Janvier K., Berteaux-Lecellier V., Cartier N., Benarous R., and Aubourg P. (1999). Homo- and heterodimerization of peroxisomal ATP-binding cassette half-transporters. J Biol Chem 274, 32738–32743. Loes D. J., Fatemi A., Melhem E. R., Gupte N., Bezman L., Moser H. W., and Raymond G. V. (in press). Brain MRI patterns and their predictive role in cerebral X-linked adrenoleukodystrophy. Neurology. Loes D. J., Hite S., Moser H., Stillman A. E., Shapiro E., Lockman L., Latchaw R. E., and Krivit W. (1994). Adrenoleukodystrophy: A scoring method for brain MR observations. AJNR Am J Neuroradiol 15, 1761–1766. Lombard-Platet G., Savary S., Sarde C. O., Mandel J. L., and Chimini G. (1996). A close relative of the adrenoleukodystrophy (ALD). gene codes for a peroxisomal protein with a speciWc expression pattern. Proc Natl Acad Sci U S A 93, 1265–1269. Lu J. F., Lawler A. M., Watkins P. A., Powers J. M., Moser A. B., Moser H. W., and Smith K. D. (1997). A mouse model for X-linked adrenoleukodystrophy. Proc Natl Acad Sci U S A 94, 9366–9371. Maestri N. E., and Beaty T. H. (1992). Predictions of a 2-locus model for disease heterogeneity: Application to adrenoleukodystrophy. Am J Med Genet 44, 576–582. Maier E. M., Kammerer S., Muntau A. C., Wichers M., Braun A., and Roscher A. A. (2002). Symptoms in carriers of adrenoleukodystrophy relate to skewed X inactivation. Ann Neurol 52, 683–688. Malm G., Ringden O., Anvret M., von Dobeln U., Hagenfeldt L., Isberg B., Knuutila S., Nennesmo I., Winiarski J., and Marcus C. (1997). Treatment of adrenoleukodystrophy with bone marrow transplantation. Acta Paediatr 86, 484–492. Martin J. J., Ceuterick C., and Libert J. (1980). Skin and conjunctival nerve biopsies in adrenoleukodystrophy and its variants. Ann Neurol 8, 291–5. McGuinness M. C., Lu J. F., Zhang H. P., Dong G. X., Heinzer A. K., Watkins P. A., Powers J., and Smith K. D. (2003). Role of ALDP (ABCD1)., and Mitochondria in X-Linked Adrenoleukodystrophy. Mol Cell Biol 23, 744–753. McGuinness M. C., Powers J. M., Bias W. B., Schmeckpeper B. J., Segal A. H., Gowda V. C., Wesselingh S. L., Berger J., GriYn D. E., and Smith K. D. (1997). Human leukocyte antigens and cytokine expression in cerebral inXammatory demyelinative lesions of X-linked adrenoleukodystrophy and multiple sclerosis. J Neuroimmunol 75, 174–182. McGuinness M. C., Zhang H. P., and Smith K. D. (2001). Evaluation of pharmacological induction of fatty acid beta-oxidation in X-linked adrenoleukodystrophy. Mol Genet Metab 74, 256–263. Menkes J. H., and Corbo L. M. (1977). Adrenoleukodystrophy. Accumulation of cholesterol esters with very long chain fatty acids. Neurology 27, 928–932. Migeon B. R., Moser H. W., Moser A. B., Axelman J., Sillence D., and Norum R. A. (1981). Adrenoleukodystrophy: Evidence for X linkage, inactivation, and selection favoring the mutant allele in heterozygous cells. Proc Natl Acad Sci U S A 78, 5066–5070. Mihalik S. J., Steinberg S. J., Pei Z., Park J., Kim do G., Heinzer A. K., Dacremont G., Wanders R. J., Cuebas D. A., Smith K. D., and Watkins P. A. (2002). Participation of two members of the very long-chain acyl-CoA synthetase family in bile acid synthesis and recycling. J Biol Chem 277, 24771–24779. Miller D. H., Khan O. A., Sheremata W. A., Blumhardt L. D., Rice G. P., Libonati M. A., Willmer-Hulme A. J., Dalton C. M., Miszkiel K. A., and O’Connor P. W. (2003). A controlled trial of natalizumab for relapsing multiple sclerosis. N Engl J Med 348, 15–23. Min K. T., and Benzer S. (1999). Preventing neurodegeneration in the Drosophila mutant bubblegum. Science 284, 1985–1988.

IV. DISEASES OF MYELIN

835

836

34. ADRENOLEUKODYSTROPHIES

au14

Molzer B., Bernheimer H., Budka H., Pilz P., and ToiX K. (1981). Accumulation of very long chain fatty acids is common to 3 variants of adrenoleukodystrophy (ALD). ‘‘Classical’’ ALD, atypical ALD (female patient)., and adrenomyeloneuropathy. J Neurol Sci 51, 301–310. Moody D. B., Besra G. S., Wilson I. A., and Porcelli S. A. (1999). The molecular basis of CD1-mediated presentation of lipid antigens. Immunol Rev 172, 285–296. Moser A. B., Kreiter N., Bezman L., Lu S., Raymond G. V., Naidu S., and Moser H. W. (1999). Plasma very long chain fatty acids in 3,000 peroxisome disease patients and 29,000 controls. Ann Neurol 45, 100–110. Moser A. B., and Moser H. W. (1999). The prenatal diagnosis of X-linked adrenoleukodystrophy. Prenat Diagn 19, 46–48. Moser H. W., and Borel J. (1995). Dietary management of X-linked adrenoleukodystrophy. Annu Rev Nutr 15, 379–397. Moser H. W., Loes D. J., Melhem E. R., Raymond G. V., Bezman L., Cox C. S., and Lu S. E. (2000). X-Linked adrenoleukodystrophy: Overview and prognosis as a function of age and brain magnetic resonance imaging abnormality. A study involving 372 patients. Neuropediatrics 31, 227–239. Moser H. W., Moser A. B. (2003). Evaluation of the preventive eVect of glyceryl trioleate-trierucate (Lorenzo’s oil). therapy in asymptomatic X-linked adrenoleukodystrophy: Results of two concurrent studies. Peroxisomal Disorders and Regulation of Genes. Frank Roels (Ed). Kluwer Academic, Plenum Publishers, London, in press. Moser H. W., Moser A. B., Frayer K. K., Chen W., Schulman J. D., O’Neill B. P., and Kishimoto Y. (1981). Adrenoleukodystrophy: Increased plasma content of saturated very long chain fatty acids. Neurology 31, 1241–1249. Moser H. W., Moser A. B., Naidu S., and Bergin A. (1991). Clinical aspects of adrenoleukodystrophy and adrenomyeloneuropathy. Dev Neurosci 13, 254–261. Moser H. W., Moser A. B., Smith K. D., Bergin A., Borel J., ShankroV J., Stine O. C., Merette C., Ott J., Krivit W., and Shapiro, E. (1992). Adrenoleukodystrophy: Phenotypic variability and implications for therapy. J Inherit Metab Dis 15, 645–664. Moser H. W., Naidu S., Kumar A. J., and Rosenbaum A. E. (1987). The adrenoleukodystrophies. Crit Rev Neurobiol 3, 29–88. Moser H. W., Pallante S. L., Moser A. B., Rizzo W. B., Schulman J. D., and Fenselau C. (1983). Adrenoleukodystrophy: Origin of very long chain fatty acids and therapy. Pediatr Res 17, 293A. Moser H. W., Smith K. D., Watkins P. A., Powers J., and Moser A. B. (2000). X-linked adrenoleukodystrophy. In ‘‘The Metabolic and Molecular Bases of Inherited Disease’’ (Scriver C. R., Beaudet A. L., Sly W. S., and Valle D., eds.), pp. 3257–3301. McGraw Hill, New York. Moser H. W., Tutschka P. J.; Brown F. R. III; Moser A. B.; Yeager A. M.; Singh I.; Mark S. A.; Kumar A. A. J.; McDonnell J. M.; White C. L. III; Maumenee I. H.; Green W. R.; Powers J. M.; and Santos G. W. (1984). Bone marrow transplant in adrenoleukodystrophy. Neurology 34, 1410–1417. Mosser J., Douar A. M., Sarde C. O., Kioschis P., Feil R., Moser H., Poustka A. M., Mandel J. L., and Aubourg P. (1993). Putative X-linked adrenoleukodystrophy gene shares unexpected homology with ABC transporters. Nature 361, 726–730. Mosser J., Lutz Y., Stoeckel M. E., Sarde C. O., Kretz C., Douar A. M., Lopez J., Aubourg P., and Mandel J. L. (1994). The gene responsible for adrenoleukodystrophy encodes a peroxisomal membrane protein. Hum Mol Genet 3, 265–271. Murad S., and Kishimoto Y. (1978). Chain elongation of fatty acid in brain: A comparison of mitochondrial and microsomal enzyme activities. Arch Biochem Biophys 185, 300–306. Naidu S., Bresnan M. J., GriYn D., O’Toole S., and Moser H. W. (1988). Childhood adrenoleukodystrophy. Failure of intensive immunosuppression to arrest neurologic progression. Arch Neurol 45, 846–848. Naidu S., Washington C., Thirumalai S., Smith K. D., Moser H. W., and Watkins P. A. (1997). X-chromosome inactivation in symptomatic heterozygotes of X-linked adrenoleukodystrophy. Ann Neurol 42, 498a. Netik A., Forss-Petter S., Holzinger A., Molzer B., Unterrainer G., and Berger J. (1999). Adrenoleukodystrophyrelated protein can compensate functionally for adrenoleukodystrophy protein deWciency (X-ALD): Implications for therapy. Hum Mol Genet 8, 907–913. Nowaczyk M. J., Saunders E. F., Tein I., Blaser S. I., and Clarke J. T. (1997). Immunoablation does not delay the neurologic progression of X-linked adrenoleukodystrophy. J Pediatr 131, 453–455. O’Neill G. N., Aoki M., and Brown R. H. Jr (2001). ABCD1 translation-initiator mutation demonstrates genotype-phenotype correlation for AMN. Neurology 57, 1956–1962. Odone A., and Odone M. (1989). Lorenzo’s oil. J Pediatr Neurosci 5, 55. Pahan K., Sheikh F. G., Namboodiri A. M., and Singh I. (1997). Lovastatin and phenylacetate inhibit the induction of nitric oxide synthase and cytokines in rat primary astrocytes, microglia, and macrophages. J Clin Invest 100, 2671–2679. Pai G. S., Khan M., Barbosa E., Key L. L., Craver J. R., Cure J. K., Betros R., and Singh I. (2000). Lovastatin therapy for X-linked adrenoleukodystrophy: Clinical and biochemical observations on 12 patients. Mol Genet Metab 69, 312–322. Peckham R. S., Marshall M. C. Jr, Rosman P. M., Farag A., Kabadi U., and Wallace E. Z. (1982). A variant of adrenomyeloneuropathy with hypothalamic-pituitary dysfunction and neurologic remission after glucocorticoid replacement therapy. Am J Med 72, 173–176. PWster R. (1936). Beitrag sur kenntnis der diVusen kirnsklerose. Archiv fur Psychiatrie 105, 1–16. Poulos A., Gibson R., Sharp P., Beckman K., and Grattan-Smith P. (1994). Very long chain fatty acids in X-linked adrenoleukodystrophy brain after treatment with Lorenzo’s oil. Ann Neurol 36, 741–746.

IV. DISEASES OF MYELIN

PREVENTION OF X-ALD

Poulos A., Singh H., Paton B., Sharp P., and Derwas N. (1986). Accumulation and defective beta-oxidation of very long chain fatty acids in Zellweger’s syndrome, adrenoleukodystrophy and Refsum’s disease variants. Clin Genet 29, 397–408. Powers J. M., DeCiero D. P., Cox C., RichWeld E. K., Ito M., Moser A. B., and Moser H. W. (2001). The dorsal root ganglia in adrenomyeloneuropathy: Neuronal atrophy and abnormal mitochondria. J Neuropathol Exp Neurol 60, 493–501. Powers J. M., DeCiero D. P., Ito M., Moser A. B., and Moser H. W. (2000). Adrenomyeloneuropathy: A neuropathologic review featuring its noninXammatory myelopathy. J Neuropathol Exp Neurol 59, 89–102. Powers J. M., Liu Y., Moser A. B., and Moser H. W. (1992). The inXammatory myelinopathy of adrenoleukodystrophy: Cells, eVector molecules, and pathogenetic implications. J Neuropathol Exp Neurol 51, 630–643. Powers J. M., Moser H. W., Moser A. B., and Schaumburg H. H. (1982). Fetal adrenoleukodystrophy: The signiWcance of pathologic lesions in adrenal gland and testis. Hum Pathol 13, 1013–1019. Powers J. M., and Schaumburg H. H. (1973). The adrenal cortex in adreno-leukodystrophy. Arch Pathol 96, 305–310. Powers J. M., and Schaumburg H. H. (1974). Adreno-leukodystrophy (sex-linked Schilder’s disease). A pathogenetic hypothesis based on ultrastructural lesions in adrenal cortex, peripheral nerve and testis. Am J Pathol 76, 481–491. Powers J. M., and Schaumburg H. H. (1981). The testis in adreno-leukodystrophy. Am J Pathol 102, 90–98. Pujol A., Hindelang C., Callizot N., Bartsch U., Schachner M., and Mandel J. L. (2002). Late onset neurological phenotype of the X-ALD gene inactivation in mice: A mouse model for adrenomyeloneuropathy. Hum Mol Genet 11, 499–505. Ramsey R. B., Banik N. L., and Davison A. N. (1979). Adrenoleukodystrophy brain cholesteryl esters and other neutral lipids. J Neurol Sci 40, 189–196. Rasmussen M., Moser A. B., Borel J., Khangoora S., and Moser H. W. (1994). Brain, liver, and adipose tissue erucic and very long chain fatty acid levels in adrenoleukodystrophy patients treated with glyceryl trierucate and trioleate oils (Lorenzo’s oil). Neurochem Res 19, 1073–1082. Reinecke C. J., Knoll D. P., Pretorius P. J., Steyn H. S., and Simpson R. H. (1985). The correlation between biochemical and histopathological Wndings in adrenoleukodystrophy. J Neurol Sci 70, 21–38. Restuccia D., Di Lazzaro V., Valeriani M., Oliviero A., Le Pera D., Barba C., Cappa M., Bertini E., Di Capua M., and Tonali P. (1999). Neurophysiologic follow-up of long-term dietary treatment in adult-onset adrenoleukodystrophy. Neurology 52, 810–816. Rezanka T. (1989). Very-long-chain fatty acids from the animal and plant kingdoms. Prog Lipid Res 28, 147–187. Rizzo W. B., Avigan J., Chemke J., and Schulman J. D. (1984). Adrenoleukodystrophy: Very long-chain fatty acid metabolism in Wbroblasts. Neurology 34, 163–169. Rizzo W. B., Leshner R. T., Odone A., Dammann A. L., Craft D. A., Jensen M. E., Jennings S. S., Davis S., Jaitly R., and Sgro J. A. (1989). Dietary erucic acid therapy for X-linked adrenoleukodystrophy. Neurology 39, 1415–1422. Robbins D. S., Shirazi Y., Drysdale B. E., Lieberman A., Shin H. S., and Shin M. L. (1987). Production of cytotoxic factor for oligodendrocytes by stimulated astrocytes. J Immunol 139, 2593–2597. Ruiz M., Coll M. J., Pampols T., and Giros M. (1997). ALDP expression in fetal cells and its application in prenatal diagnosis of X-linked adrenoleukodystrophy. Prenat Diagn 17, 651–656. Sarde C. O., Mosser J., Kioschis P., Kretz C., Vicaire S., Aubourg P., Poustka A., and Mandel J. L. (1994). Genomic organization of the adrenoleukodystrophy gene. Genomics 22, 13–20. Schaumburg H. H., Powers J. M., Raine C. S., Suzuki K., and Richardson E. P. Jr (1975). Adrenoleukodystrophy. A clinical and pathological study of 17 cases. Arch Neurol 32, 577–591. Schilder P. (1924). Die encephalitis periaxilis. Difusa. Arch Psychiatr Nervenkr 71, 327–356. Shani N., Jimenez-Sanchez G., Steel G., Dean M., and Valle D. (1997). IdentiWcation of a fourth half ABC transporter in the human peroxisomal membrane. Hum Mol Genet 6, 1925–1931. Shapiro E., Krivit W., Lockman L., Jambaque I., Peters C., Cowan M., Harris R., Blanche S., Bordigoni P., Loes D., Ziegler R., Crittenden M., Ris D., Berg B., Cox C., Moser H., Fischer A., and Aubourg P. (2000). Longterm eVect of bone-marrow transplantation for childhood-onset cerebral X-linked adrenoleukodystrophy. Lancet 356, 713–718. Shapiro E. G., and Klein K. A. (1994). Dementiq in childhood: Issues in neuropsychological assessment with application to the natural history and treatment of degenerative storage diseases. In ‘‘Advances in Child Neuropsychology’’ (Tramontana M. G., and Hooer S. R., eds.), pp. 119–171. Springer Verlag, New York. Shapiro E. G., Lockman L. A., Balthazor M., and Krivit W. (1995). Neuropsychological outcomes of several storage diseases with and without bone marrow transplantation. J Inherit Metab Dis 18, 413–429. Sheikh K. A., Deerinck T. J., Ellisman M. H., and GriYn J. W. (1999). The distribution of ganglioside-like moieties in peripheral nerves. Brain 122 (Pt 3), 449–60. Siemerling E., and Creutzfeldt H. G. (1923). Bronzekrankheit und Sklerosierende Encephalomyelitis. Arch Psychiatr Nervenkr 68, 217–244. Singh I., Khan M., Key L., and Pai S. (1998). Lovastatin for X-linked adrenoleukodystrophy. N Engl J Med 339, 702–703. Singh I., Moser A. E., GoldWscher S., and Moser H. W. (1984a). Lignoceric acid is oxidized in the peroxisome: Implications for the Zellweger cerebro-hepato-renal syndrome and adrenoleukodystrophy. Proc Natl Acad Sci U S A 81, 4203–4207.

IV. DISEASES OF MYELIN

837

838

34. ADRENOLEUKODYSTROPHIES

Singh I., Moser A. E., Moser H. W., and Kishimoto Y. (1984b). Adrenoleukodystrophy: Impaired oxidation of very long chain fatty acids in white blood cells, cultured skin Wbroblasts, and amniocytes. Pediatr Res 18, 286–290. Singh I., Moser H. W., Moser A. B., and Kishimoto Y. (1981). Adrenoleukodystrophy: Impaired oxidation of long chain fatty acids in cultured skin Wbroblasts an adrenal cortex. Biochem Biophys Res Commun 102, 1223–1229. Singh I., Pahan K., and Khan M. (1998). Lovastatin and sodium phenylacetate normalize the levels of very long chain fatty acids in skin Wbroblasts of X-adrenoleukodystrophy. FEBS Lett 426, 342–346. Smith K. D., Kemp S., Braiterman L. T., Lu J. F., Wei H. M., Geraghty M., Stetten G., Bergin J. S., Pevsner J., and Watkins P. A. (1999). X-linked adrenoleukodystrophy: Genes, mutations, and phenotypes. Neurochem Res 24, 521–535. Smith K. D., Sack G., Beaty T., Bergin A., Naidu S., Moser A., and Moser H. (1991). A genetic basis for multiple phenotypes of X-ALD. Am J Hum Genet 49(Suppl)., 865. Stanislaus R., Pahan K., Singh A. K., and Singh I. (1999). Amelioration of experimental allergic encephalomyelitis in Lewis rats by lovastatin. Neurosci Lett 269, 71–74. Steinberg S. J., Kemp S., Braiterman L. T., and Watkins P. A. (1999). Role of very-long-chain acyl-coenzyme A synthetase in X-linked adrenoleukodystrophy. Ann Neurol 46, 409–412. Steinberg S. J., Mihalik S. J., Kim D. G., Cuebas D. A., and Watkins P. A. (2000a). The human liver-speciWc homolog of very long-chain acyl-CoA synthetase is cholate:CoA ligase. J Biol Chem 275, 15605–15608. Steinberg S. J., Morgenthaler J., Heinzer A. K., Smith K. D., and Watkins P. A. (2000b). Very long-chain acylCoA synthetases. Human ‘‘bubblegum’’ represents a new family of proteins capable of activating very longchain fatty acids. J Biol Chem 275, 35162–35169. Stumpf D. A., Hayward A., Haas R., Frost M., and Schaumburg H. H. (1981). Adrenoleukodystrophy. Failure of immunosuppression to prevent neurological progression. Arch Neurol 38, 48–49. Tagawa Y., Laroy W., Nimrichter L., Fromholt S. E., Moser A. B., Moser H. W., and Schnaar R. L. (2002). Antiganglioside antibodies bind with enhanced aYnity to gangliosides containing very long chain fatty acids. Neurochem Res 27, 847–855. Taketomi T., Hara A., Kitazawa N., Takada K., and Nakamura H. (1987). An adult case of adrenoleukodystrophy with features of olivo-ponto-cerebellar atrophy: II. Lipid biochemical studies. Jpn J Exp Med 57, 59–70. Tanaka Y. A. S. T. S. M. T. (1988). Enhanced synthesis of hexacosanoic acid in the cultured Wbroblasts from patients with adrenoleukodystrophy. Biomed Res 9, 451–456. Theda C., Moser A. B., Powers J. M., and Moser H. W. (1992). Phospholipids in X-linked adrenoleukodystrophy white matter: Fatty acid abnormalities before the onset of demyelination. J Neurol Sci 110, 195–204. Tsuji S., Sano T., Ariga T., and Miyatake T. (1981a). Increased synthesis of hexacosanoic acid (C23:0). by cultured skin Wbroblasts from patients with adrenoleukodystrophy (ALD)., and adrenomyeloneuropathy (AMN). J Biochem (Tokyo). 90, 1233–1236. Tsuji S., Sano T., Ariga T., and Miyatake T. (1981b). Increased synthesis of hexacosanoic acid (C26:0). by cultured skin Wbroblasts from patients with adrenoleukodystrophy (ALD)., and adrenomyeloneuropathy (AMN). Biochem J (Tokyo). 90, 1233–1236. Tsuji S., Suzuki M., Ariga T., Sekine M., Kuriyama M., and Miyatake T. (1981c). Abnormality of long-chain fatty acids in erythrocyte membrane sphingomyelin from patients with adrenoleukodystrophy. J Neurochem 36, 1046–1049. Uchiyama A., Aoyama T., Kamijo K., Uchida Y., Kondo N., Orii T., and Hashimoto T. (1996). Molecular cloning of cDNA encoding rat very long-chain acyl-CoA synthetase. J Biol Chem 271, 30360–30365. Ulrich J., Herschkowitz N., Heitz P., Sigrist T., and Baerlocher P. (1978). Adrenoleukodystrophy. Preliminary report of a connatal case. Light- and electron microscopical, immunohistochemical and biochemical Wndings. Acta Neuropathol (Berl). 43, 77–83. Unger E. R., Sung J. H., Manivel J. C., Chenggis M. L., Blazar B. R., and Krivit W. (1993). Male donor-derived cells in the brains of female sex-mismatched bone marrow transplant recipients: A Y-chromosome speciWc in situ hybridization study. J Neuropathol Exp Neurol 52, 460–470. Uziel G., Bertini E., Bardelli P., Rimoldi M., and Gambetti M. (1991). Experience on therapy of adrenoleukodystrophy and adrenomyeloneuropathy. Dev Neurosci 13, 274–279. van Geel B. M., Assies J., Haverkort E. B., Koelman J. H., Verbeeten B. Jr, Wanders R. J., and Barth P. G. (1999). Progression of abnormalities in adrenomyeloneuropathy and neurologically asymptomatic X-linked adrenoleukodystrophy despite treatment with ‘‘Lorenzo’s oil’’. J Neurol Neurosurg Psychiatry 67, 290–299. van Geel B. M., Bezman L., Loes D. J., Moser H. W., and Raymond G. V. (2001). Evolution of phenotypes in adult male patients with X-linked adrenoleukodystrophy. Ann Neurol 49, 186–194. van Geel B. M., Koelman J. H., Barth P. G., and Ongerboer de Visser B. W. (1996). Peripheral nerve abnormalities in adrenomyeloneuropathy: A clinical and electrodiagnostic study. Neurology 46, 112–118. Volpe J. J., and Vagelos P. R. (1976). Mechanisms and regulation of biosynthesis of saturated fatty acids. Physiol Rev 56, 339–417. Wanders R. J., van Roermund C. W., van Wijland M. J., Heikoop J., Schutgens R. B., Schram A. W., Tager J. M., van den Bosch H., Poll-The B. T., Saudubray J. M., et al. (1987). Peroxisomal very long-chain fatty acid beta-oxidation in human skin Wbroblasts: Activity in Zellweger syndrome and other peroxisomal disorders. Clin Chim Acta 166, 255–263.

IV. DISEASES OF MYELIN

au15

839

PREVENTION OF X-ALD

Wanders R. J., van Roermund C. W., van Wijland M. J., Schutgens R. B., van den Bosch H., Schram A. W., and Tager J. M. (1988). Direct demonstration that the deWcient oxidation of very long chain fatty acids in X-linked adrenoleukodystrophy is due to an impaired ability of peroxisomes to activate very long chain fatty acids. Biochem Biophys Res Commun 153, 618–624. Watkins P. A., Gould S. J., Smith M. A., Braiterman L. T., Wei H. M., Kok F., Moser A. B., Moser H. W., and Smith K. D. (1995). Altered expression of ALDP in X-linked adrenoleukodystrophy. Am J Hum Genet 57, 292–301. Watkins P. A., Pevsner J., and Steinberg S. J. (1999). Human very long-chain acyl-CoA synthetase and two human homologs: Initial characterization and relationship to fatty acid transport protein. Prostaglandins Leukot Essent Fatty Acids 60, 323–328. Weinhofer I., Forss-Petter S., Zigman M., Berger J. (2002). Cholesterol regulates ABCD2 expression: Implications for the therapy of X-linked adrenoleukodystrophy. Hum Mol Genet 11, 2701–2708. Whitcomb R. W., Linehan W. M., and Knazek R. A. (1988). EVects of long-chain, saturated fatty acids on membrane microviscosity and adrenocorticotropin responsiveness of human adrenocortical cells in vitro. J Clin Invest 81, 185–188. Wilson R., and Sargent J. R. (1993). Lipid and fatty acid composition of brain tissue from adrenoleukodystrophy patients. J Neurochem 61, 290–297. Yamada T., Shinnoh N., Taniwaki T., Ohyagi Y., Asahara H., Horiuchi and Kira J. (2000). Lovastatin does not correct the accumulation of very long-chain fatty acids in tissues of adrenoleukodystrophy protein-deWcient mice. J Inherit Metab Dis 23, 607–614. Yoshida S., and Takeshita M. (1987). Analysis of the condensation step in elongation of very-long-chain saturated and tetraenoic fatty acyl-CoAs in swine cerebral microsomes. Arch Biochem Biophys 254, 170–179. Zeller C. B., and Marchase R. B. (1992). Gangliosides as modulators of cell function. Am J Physiol 262, C1341–55. Zhang and Moser H. W. (2003). Arch Neurol in press. Zinkham W. H., Kickler T., Borel J., and Moser H. W. (1993). Lorenzo’s oil and thrombocytopenia in patients with adrenoleukodystrophy. N Engl J Med 328, 1126.

IV. DISEASES OF MYELIN

au16

C H A P T E R

35 Krabbe Disease Kunihiko Suzuki

HISTORY In 1916, a Danish physician, Knud Krabbe, described clinical and pathological Wndings in Wve infants from two families who died of an ‘‘acute infantile familial diVuse brain sclerosis’’ (Krabbe, 1916). These infants developed episodes of violent crying and irritability beginning at age 4 to 6 months, followed by progressive muscular rigidity, tonic spasms evoked by such stimuli as noise, light, and touching. Death occurred between 11 months to 1.5 years. He provided a detailed description of the globoid cells, the histologic hallmark of the disease. Thus, the Wrst description of the disease is credited to Krabbe. Retrospectively, however, similar abnormal cells in the white matter had been described in neuropathological literature. Collier and GreenWeld (1924) Wrst used the term ‘‘globoid cells’’ to describe the PAS-positive macrophages unique in this disease. In 1970, deWciency in the activity of a lysosomal enzyme, galactosylceramidase, was identiWed as the underlying genetic cause (Suzuki and Suzuki, 1970) that made noninvasive antemortem diagnosis possible (Suzuki and Suzuki, 1971). Prenatal diagnosis of an aVected fetus was Wrst accomplished in 1971 (Suzuki et al., 1971). Toxic eVect of a related metabolite, galactosylsphingosine (psychosine), was Wrst proposed in 1972 as critical in the biochemical pathogenesis (Miyatake and Suzuki, 1972). The psychosine hypothesis has since been generally substantiated both in the human disease and in animal models (Suzuki, 1998). In 1990, the gene encoding galactosylceramidase was mapped to human chromosome 14 (Zlotogora et al., 1990). Human galactosylceramidase cDNA was cloned in 1993–1994 (Chen et al., 1993; Sakai et al., 1994), and more than 60 disease causing mutations have been identiWed (Wenger et al., 2001).

INCIDENCE, GENETICS, AND CLINICAL FORMS Globoid cell leukodystrophy is inherited as an autosomal recessive trait with a wide geographical distribution. Until enzymatic diagnosis became feasible, the diagnosis was made on the basis of the characteristic neuropathology. Since the gene was cloned, recently described cases have enzymatic as well as molecular diagnosis. Typical infantile patients develop Wrst clinical signs and symptoms at 3 to 6 months after birth, but there are cases of very early or late onset with atypical clinical manifestations. The incidence of the typical infantile form is estimated as 1 in 100,000 births in the United States, 2 in 100,000 births in Sweden, and 0.5 to 1 in 100,000 births in Japan. There are pockets of populations where the incidence is unusually high; for example, the Druze community in Israel has an incidence of 6 in 1000 births. In contrast, no Jewish patients have ever been reported. Late-onset and adult forms of the disease are even rarer but are being reported in

Myelin Biology and Disorders, Volume 2

841

Copyright 2004, Elsevier (USA). All rights reserved.

842

35. KRABBE DISEASE

increasing frequency. Since deWnitions of ‘‘late-onset’’ and ‘‘adult’’ forms are not necessarily standardized, precise incidence of the later-onset forms is diYculty to assess. In contrast to the infantile form, which appears to have somewhat higher incidence in Nordic countries, patients with the late-onset form appear to be more common in Southern Europe (Barone et al., 1996; Fiumara et al., 1990). The frequency of late-onset cases was estimated to be approximately 10% of the nearly 350 GLD patients diagnosed by Wenger (Wenger et al., 2001).

CLINICAL MANIFESTATIONS Infantile GLD Clinical phenotype of the classical infantile Krabbe disease is relatively stereotypic. Hagberg et al. (Hagberg et al., 1963) divided the steady and rapidly progressive clinical course into three stages. The general clinical picture is that of a progressive white matter disorder. Stage I is characterized by generalized hyperirritability, hyperesthesia, episodic fever of unknown origin, and some stiVness of the limbs. The child, apparently normal for the Wrst few months after birth, becomes hypersensitive to auditory, tactile, or visual stimuli and begins to cry frequently without apparent cause. Slight retardation or regression of psychomotor development, vomiting with feeding diYculty, and convulsive seizures may occur as initial clinical symptoms. The cerebrospinal Xuid protein level is already highly increased. In stage II, rapid and severe motor and mental deterioration develops. There is marked hypertonicity, with extended and crossed legs, Xexed arms, and the backward-bent head. Tendon reXexes are hyperactive. Minor tonic or clonic seizures occur. Optic atrophy and sluggish pupillary reactions to light are common. Stage III is the ‘‘burnt-out’’ stage, sometimes reached within a few weeks or months. The infant is decerebrate and blind and has no contact with the surroundings. Deafness may appear. Patients rarely survive for more than 2 years. Clinical examination may not always reveal neuropathy, especially in the early stages, because symptoms and signs of central nervous system involvement are overwhelming. Krabbe (1916), however, noted in his original Wve patients that kneejerks could not be elicited and that stiVness passed into a Xaccid state toward the end of the disease. The typical pathology is always present in the peripheral nerves, and careful clinical examination combined with appropriate electrophysiological studies should reveal presence of PNS involvement. Spinal Xuid protein is invariably highly elevated in patients with infantile GLD. The symptoms and signs are, for all practical purposes, conWned to the nervous system. No visceromegaly is present.

Late-Onset GLD Earlier, patients with late-onset globoid cell leukodystrophy were often misdiagnosed during life, and the deWnite diagnosis could be established only by histological examination. Since the advent of the enzymatic diagnosis, late-onset GLD has been reported in increasing frequency. Most patients develop initial clinical signs and symptoms by 10 years of age, but some may develop neurological signs after 40 years of age. Late-onset GLD is often divided into two types; late infantile (or early childhood) and juvenile (late childhood). In the late infantile group (onset 6 months to 3 years), irritability, psychomotor regression, stiVness, ataxia, and loss of vision are frequent initial symptoms. The course is progressive resulting in death approximately in 2 to 3 years after the onset. In the juvenile group (onset 3 to 8 years), patients commonly develop loss of vision, together with hemiparesis, ataxia, and psychomotor regression. Most patients with the juvenile form show rapid deterioration initially, followed by a more gradual progression possibly lasting for years. The number of reports on adult cases is also increasing steadily. Adult patients may develop slowly progressive spastic paraparesis or slow, unsteady stiV- and wide-based gait during life. Progressive and generalized neurological deterioration may not be

IV. DISEASES OF MYELIN

au1

PATHOLOGY

observed until 40 years of age. Some adult patients may have a normal life span. The cerebrospinal Xuid protein is normal or only mildly elevated in the juvenile or adult type patients. Peripheral nerve conduction velocity is generally reduced in late infantile patients but may be normal in juvenile patients with some exceptions.

PATHOLOGY Pathology is, for all practical purposes, limited to the nervous system (Suzuki and Suzuki, 2002). In the most common infantile type, the brain is atrophic with Wrm rubbery gliotic white matter. At the terminal stage, loss of myelin is nearly complete with possible exception of the subcortical intergyral arcuate U-Wbers. Microscopically, marked paucity of myelin with some axonal degeneration is present throughout the brain. Extensive Wbrillary gliosis and inWltration of numerous macrophages, often multinucleated (‘‘globoid cells’’), are the unique features. The globoid cells are abundant in the region of active demyelination and often clustered around blood vessels. Oligodendrocytes are markedly reduced. Correlative MRI and neuropathological studies showed that the areas of marked hyper-intensity on the T2-weighted MRI images corresponded to the areas of demyelination with globoid cell inWltration (Percy et al., 1994). Globoid cells contain PAS-positive storage materials. On the ultrastructural level, the globoid cells contain tubular and Wlamentous structures with polygonal cross sections that are structurally identical with chemically pure galactosylceramide (Yunis and Lee, 1970). In long surviving cases, the white matter may be totally gliotic and devoid of macrophages. The optic nerves are usually atrophic but, in some cases, they are markedly enlarged with extensive gliosis. The peripheral nerves are often grossly enlarged and Wrm with marked endoneurial Wbrosis, segmental demyelination, and evidence of remyelination process with onion bulb formation. Quantitative analyses demonstrated a severe loss of large myelinated Wbers without loss of unmyelinated Wbers. Endoneurial macrophages and also Schwann cells contain tubular inclusions similar to those in the globoid cells in the cerebral white matter. Neuropathological reports of late onset cases, however, are limited. In a meeting abstract, Choi et al. described the neuropathology of the adult onset GLD in 18-year-old twins. Their clinical symptoms developed 12 and 7 months prior to their death, respectively. Both died of severe graft-versus-host disease 2 months after allogeneic bone marrow transplantation. The brains showed degeneration of the optic radiation and frontoparietal white matter with corticospinal tract degeneration. Multiple necrotic foci with calcium deposits were found within the lesion. Globoid cell inWltration was present in actively degenerating white matter. In the peripheral nerves, loss of myelinated Wbers, disproportionately thin myelin sheaths and inclusions in Schwann cells were described.

ANALYTICAL BIOCHEMISTRY The genetic cause of all so far known human patients with Krabbe disease is deWcient activity of galactosylceramidase (Fig. 35.1). Galactosylceramidase is a degradative enzyme with an acid ph optimum localized in the lysosome. Thus, the disease conceptually belongs to the category of genetic disorders, the lysosomal disease as originally deWned by Hers (Hers, 1966). Essentially all lysosomal diseases are ‘‘storage diseases,’’ in which subtrates of the genetically defective enzymes accumulate to abnormally high levels. The enzyme is fairly speciWc for glycolipids with a terminal galactose moiety in a b anomeric conWguration. Quantitatively, by far the major natural substrate is galactosylceramide, which is highly localized in the myelin sheath. Other known natural substrates are psychosine (galactosylsphingosine), monogalactosyldiglyceride, and the precursor of seminolipid (1-alkyl, 2-acyl-, 3-galactosyl glycerol). In vivo degradation of these substrates requires, in addition to the enzyme, galactosylceramidase, an activator protein, saposin A. In addition, the two lysosomal ß-galactosidases, galactosylceramide and GM1-ganglioside ß-galactosidase,

IV. DISEASES OF MYELIN

843

844

35. KRABBE DISEASE

lactosylceramide

[other glycosphingolipids/gangliosides]

Saposin B,C

sphingomyelin

glucosylceramide

digalactosylceramide galactose

Saposin C

GM4 CGT

ceramide

x

Saposin B

galactosylceramidase Saposin A

FA Saposin D?

sphingosine

galactose

sulfatide

galactosylceramide

CGT

x

SO4

psychosine

galactosylceramidase

FIGURE 35.1 Metabolic pathways pertinent to galactosylceramide and related compounds. In the synthetic pathway, sphingosine is Wrst acylated to ceramide, which in turn is galactosylated by UDP-galactose:ceramide galactosyltransferase (CGT) to form galactosylceramide. The same enzyme can galactosylate sphingosine directly to generate psychosine. Both galactosylceramide and psychosine are degraded by galactosylceramidase, which is genetically deWcient in Krabbe disease. In vivo degradation of galactosylceramide requires, in addition to the enzyme, a sphingolipid activator protein, saposin A. Galactosylceramide is further sulfated to form sulfatide. Both galactosylceramide and sulfatide are characteristic myelin glycolipids.

share lactosylceramide as their common substrate. The unique biochemical characteristic of Krabbe disease is lack of abnormal accumulation of galactosylceramide in the brain, contrary to what is expected from the enzymatic defect (Svennerholm et al., 1980; Vanier and Svennerholm, 1975). This paradoxical phenomenon results from the unique localization of galactosylceramide in the myelin sheath and very rapid and early disappearance of the myelinating cells in the process of the disease. Since the myelinating cells disappear at a very early stage of myelination and since no further synthesis of galactosylceramide occurs, it does not accumulate beyond the level attained at the early stage of myelination. Instead, however, a related toxic metabolite, psychosine (galactosylsphingosine), does accumulate abnormally and is considered the key compound in the pathogenesis of the disease (discussed later) (Miyatake and Suzuki, 1972; Vanier and Svennerholm, 1976). Although there is no abnormal accumulation of galactosylceramide in the brain tissue as a whole, there is clear evidence to indicate that localized accumulation of galactosylceramide does occur within the characteristic globoid cells. Biochemical analysis of a fraction enriched with the characteristic globoid cells contained a relatively large amounts of galactosyceramide (Austin, 1963). Galactosylceramide has unique capacity to elicit inWltration of globoid cells when it is implanted into the brain (Austin and Lehfeldt, 1965) and such experimentally induced globoid cells appear morphologically identical to those seen in patients with Krabbe disease (Andrews and Menkes, 1970; Suzuki, 1970). Biochemical abnormalities are essentially limited to the nervous system, at least in human patients (see Chapter 45, ‘‘Models of Krabbe Disease’’. au2

PATHOPHYSIOLOGY Metabolism of Myelin and Its Galactolipids Galactosylceramide has a uniquely restricted tissue distribution. It is mostly, but not exclusively, localized in the myelin sheath and thus almost exclusively synthesized within the oligodendrocytes and the Schwann cells. Its sulfate ester, sulfatide, is synthesized only by sulfation of galactosylceramide (Fig. 35.1). Thus, both galactosylceramide and sulfatide

IV. DISEASES OF MYELIN

PATHOPHYSIOLOGY

are characteristically the glycolipids of the myelin sheath and are virtually absent in the brain before myelination and are present at abnormally low concentrations in any pathologic conditions where severe loss of myelin occurs. The amount of total brain galactosylceramide correlates precisely with the amount of myelin that can be isolated from the brain, whereas amounts of other lipids do not (Norton and Poduslo, 1973). It is practically absent in systemic organs except in the kidney, which normally contains appreciable amounts of galactosylceramide, although much less than in the nervous system. Myelin of adult mammalian brain generally contains galactosylceramide at a concentration of 15 to 18% of total lipid. The sum of galactosylceramide and sulfatide makes up to 20% of the dry weight of myelin. The content of galactosylceramide in myelin from the peripheral nerve is somewhat less than that of CNS myelin. In view of the unusually high concentrations of galactosylceramide and sulfatide in the myelin sheath, metabolic diseases involving these lipids (Krabbe disease and metachromatic leukodystrophy) would be expected to manifest themselves primarily as disorders of white matter and peripheral nerves. The most signiWcant metabolic features of CNS myelin are its high rate of formation and turnover during a relatively short period of brain development and its slow turnover in the adult brain. The period of most active myelination in humans probably extends from the perinatal period to about age 18 months. Myelination does not stop after this period, and in the human brain, it may not be complete until age 20 years. The amount of galactosylceramide in immature brain is very low. Activity of galactosylceramide synthase, UDP-galactose:ceramide galactosyltransferase (CGT), peaks sharply at 20 to 25 days after birth in rodent brains well correlating with the most active period of myelination (Costantino-Ceccarini and Morell, 1972). The recent cloning of the rat and mouse CGT conWrmed that the peak levels of the corresponding mRNA occur during the period of most active myelination and that relatively high mRNA levels are found in the brain and kidney (Stahl et al., 1994). Synthesis and turnover of galactosylceramide occurs at much lower rates in the adult brain.

Pathogenesis Some aspects of the chemistry and metabolism of galactosylceramide should be kept in mind when the pathogenetic mechanism of GLD is considered. (1) Galactosylceramide consists of sphingosine, fatty acid, and galactose. (2) Galactosylceramide is the precursor of sulfatide. (3) Both galactosylceramide and sulfatide are highly concentrated in the myelin sheath. (4) Galactosylceramidase degrades galactosylceramide to ceramide and galactose. (5) A few related galactolipids also serve as substrates for galactosylceramidase, including psychosine, monogalactosyldiglyceride, seminolipid precursor, and lactosylceramide. (6) Biosynthesis of galactosylceramide reaches a peak, coincident with the maximum period of myelination (during the Wrst year and a half in humans), when myelin also turns over relatively rapidly. (7) Once formed, adult myelin is relatively stable metabolically, although there is some turnover. (8) Galactosylceramide is uniquely capable of inducing inWltration of globoid cells when implanted into the brain but does not appear toxic, while another normally insigniWcant substrate, psychosine, is highly cytotoxic. Two Separate but Related Pathogenetic Mechanisms? Three of the most characteristic pathological features of Krabbe disease are (1) the inWltration of macrophages that are often multinucleated and contain strongly PAS-positive materials (‘‘globoid cells’’), (2) the rapid and almost complete disappearance of the oligodendrocytes, and (3) lack of abnormal tissue accumulation of the primary substrate of the defective enzyme, galactosylceramide, contrary to what is expected in a ‘‘storage disease’’ due to genetic defect in degradative enzymes. These phenotypic characteristics must be explained as consequences of the underlying genetic defect. Defective degradation of two substrates, galactosylceramide and psychosine (galactosylsphingosine), appears to play critical roles in the pathogenesis. While these mechanisms are fundamentally distinct from each other, they are closely intertwined to result in the unique phenotype of the disease.

IV. DISEASES OF MYELIN

845

846

35. KRABBE DISEASE

Globoid cells The genetic defect in degradation of galactosylceramide clearly is a major factor for the unique pathological feature of the disease, the globoid cells. It has long been known that free galactosylceramide has a speciWc capacity to elicit inWltration of macrophages into the brain (Austin and Lehfeldt, 1965; Suzuki et al., 1976). Once in the brain, they phagocytize galactosylceramide and are transformed to multinucleated globoid cells. The characteristic inclusions in the globoid cells have morphological appearance identical to galactosylceramide itself (Yunis and Lee, 1970). No other agent is known to have a similar capacity in vivo. The globoid cell reaction can be reconstructed in the following way. Once the active period of myelination begins, turnover of already formed myelin also begins. In patients’ brains, however, galactosylceramide cannot be degraded due to the underlying galactosylceramidase deWciency. Free galactosylceramide thus generated elicits inWltration of macrophophages, which become the characteristic PAS-positive, often multinucleated globoid cells. Psychosine hypothesis On the other hand, the devastating early destruction of the myelin-forming cells is diYcult to explain on the basis of undegradable galactosylceramide because galactosylceramide implanted in the brain does not exhibit any functionally detrimental capacity other than eliciting the globoid cell reaction. There is no experimental evidence that galactosylceramide is a metabolic toxin. On the other hand, a closely related metabolite, psychosine (galactosylsphingosine), is highly cytotoxic (Taketomi and Nishimura, 1964) and causes fatal hemorrhagic infarct when implanted into the brain (Miyatake and Suzuki, 1972). At least in mammalian tissues, psychosine can be generated only by galactosylation of sphingosine by galactosylceramide synthase, UDP-galactose:ceramide galactosyltransferase (CGT), but not by de-acylation of galactosylceramide. Since CGT is nearly exclusively localized in the myelin-forming cells, synthesis of psychosine should also occur only in the oligodendrocytes and Schwann cells. Psychosine is detectable in normal brain with highly sensitive analytical methods but its concentration is minuscule (less than 10 picomoles/mg protein). It appears to be a dead-end product, which is normally degraded immediately. However, psychosine is degraded also by galactosylceramidase. Therefore, patients with Krabbe disease cannot degrade psychosine. A hypothesis, known as the psychosine hypothesis, was Wrst proposed on the basis of this enzymological consideration (Miyatake and Suzuki, 1972), and then its abnormal accumulation was analytically demonstrated in the brain of patients (Svennerholm et al., 1980; Vanier and Svennerholm, 1976) and in canine and murine models (Igisu and Suzuki, 1984). The psychosine hypothesis postulates that, in globoid cell leukodystrophy, not only the primary substrate of the defective enzyme, galactosylceramide, but also the toxic metabolite, galactosylsphingosine (psychosine), cannot be degraded and the consequent abnormal accumulation of psychosine causes the uniquely rapid destruction of the myelin-forming cells. The hypothesis initially met considerable skepticism but has survived the intervening 30 years (Suzuki, 1998). In fact, the basic premise of the hypothesis has been extended to other sphingolipidoses (Hannun and Bell, 1987). For varieties of reasons, however, its plausibility for other disorders is not as Wrm as it is for GLD, with possible exceptions of neuronopathic form of Gaucher disease and Niemann-Pick type A disease. Overall pathogenesis Close interactions of these two pathogenetic mechanisms can explain the most important aspects of the characteristic phenotype of Krabbe disease (Fig. 35.2). The fundamental cause of the disease is the genetic defect in galactosyceramidase activity. Since galactosylceramide synthesis is limited to actively myelinating cells, the disease process does not begin until the active myelination period. Once myelination begins, its metabolic turnover also starts. This generates free galactosylceramide in the brain of patients because of the inability to degrade galactosylceramide, which in turn elicits the characteristic globoid cell reaction. Galactosylceramide synthase also synthesizes psychosine within the actively myelinating cells. Normally, it is immediately degraded and never reaches beyond a barely detectable levels. In Krabbe disease, however, an abnormal accumulation of psychosine occurs to the level toxic to cellular metabolism. This causes the other characteristic feature of the disease, a rapid and almost complete disappearance of

IV. DISEASES OF MYELIN

MOLECULAR GENETICS

FIGURE 35.2 Pathogenetic mechanisms operating in Krabbe disease. See text for explanation.

the oligodendrocytes. Psychosine is as potent an apoptosis inducer as C6 ceramide (Tohyama et al., 2001). The cellular death results in further destruction of already formed myelin, which contributes more free galactosylceramide that in turn further elicits the globoid cell inWltration. On the other hand, myelination ceases at a very early stage due to the near-complete loss of the oligodendrocytes. This explains the paradoxical characteristics of the disease that the primary substrate of the defective enzyme, galactosylceramide, does not accumulate abnormally.

MOLECULAR GENETICS Gene Structure Human galactosylceramidase gene, galc (GenBank database Accession No. 119970), was localized to the region of 14q24.3-q32.1 by linkage analysis (Oehlmann et al., 1993) and later further narrowed to 14q31 by in situ hybridization (Cannizzaro et al., 1994). Using Nterminal amino acid sequence information, the human cDNA was cloned in 1993–1994 (Chen et al., 1993; Sakai et al., 1994). The full-length cDNA consists of 3795 bp, including 2007 bp of the coding region, 47 base pairs of 5’ untranslated sequence and 1741 bp of 3’ untranslated sequence. The base and amino acid sequences have no similarities and thus no suggestion of evolutionary relationship with the other b-galactosidases or any other known genes. The encoded protein consists of 669 amino acids with six potential glycosylation sites. The Wrst 26-amino acids have the characteristic of a leader sequence. The precursor protein is approximately 80-85 kD, which is processed to 50–52 kD and 30 kD subunits. The coding sequence for the 50–52 kD subunit is at the 5’ end and the 30 kD subunit is at the 3’ end of the coding region. The organization of the human gene was characterized in 1995 (Luzi et al., 1995). It consists of 17 exons spread over about 56 kb. Other than exons 1 and 17, the other exons range in size from 39 to 181 nucleotides. The 200 nucleotides preceding the initiation codon and the 5’ end of intron 1 are GC-rich, including 13 GGC trinucleotides. The 5’ Xanking region includes a YYI element and one potential SP1 binding site. No consensus TATA box or CAAT box is present among the 800 nucleotides preceding the initiation codon. A construct containing nucleotides
176 to
24 had the strongest promoter activity. However, evidence for inhibitory sequences was found just upstream of the promoter region and also at the 5’ end of intron 1. There is another potential initiation codon located 48 nucleotides upstream; however, there is no evidence that it is utilized or if it might play a role in tissue speciWc expression.

Disease-Causing Mutations More than 60 disease-causing missense, nonsense mutations, deletions, and insertions have been identiWed in the human galactosylceramidase gene (Wenger et al., 2001). A major deletion of 30-kb from the middle of intron 10 to beyond the end of the gene that always

IV. DISEASES OF MYELIN

847

848

35. KRABBE DISEASE

occurs on a 502T polymorphic background (502T/del) is common among patients from Northern Europe and the United States, including those with Mexican ancestry. The 30 kb deletion eliminates all of the coding region for the 30 kD subunit and about 15% of the coding region for the 50–52 kD subunit. A survey conducted among patients within the Dutch population and from other parts of Europe conWrmed that 502T/del mutation makes up about 50% of the total mutant alleles. In infantile Swedish patients, this mutation makes up 75% of the mutant alleles. This mutation probably initially occurred in Sweden and was transmitted from there throughout Europe, Near Asia, and the United States. It has not been found among Japanese patients. Two other mutations (C1538T and A1652C) make up an additional 10 to 15% of the mutant alleles in infantile patients with European ancestry. Three mutations (635delþins, A198G, and T1853C) have been found in multiple unrelated Japanese patients. In Israel there are two populations with an extremely high carrier rate for Krabbe disease, and they have diVerent mutations. All infantile patients in the Druze population in Northern Israel are homozygous for the T!G transversion at nucleotide 1748 (I583S), and patients from a Moslem village near Jerusalem are homozygous for the G!A transition at nucleotide 1582 (D528N).

Polymorphisms There is a relatively broad range of galactosylceramidase activities in the ‘‘normal’’ population and among the obligate heterozygotes. This makes enzyme-based carrier testing in the general population nearly impossible. Also, there are normal individuals, including obligate heterozygotes, who have galactosylceramidase activity suYciently low for diagnosis for Krabbe disease but who are clinically normal. These phenomena can be explained at least partially by the presence of polymorphisms in the galactosylceramidase gene that result in amino acid substitutions. The C502T polymorphism is widespread but the A865G polymorphism has been reported only among Japanese. These polymorphisms generate galactosylceramidase proteins that are less active than the most common type. It has been observed by several groups that these polymorphisms occur on the same alleles as disease-causing mutations at a higher than expected frequency. Some ‘‘disease-causing’’ mutations may in fact be deleterious only when the polymorphism is present on the same allele. Polymorphisms may also play a role in the development of clinical disease when inherited either in multiple copies, on the same allele with another mutation, or together with a known disease-causing mutation on the other chromosome.

TREATMENT Only supportive care is available for patients with the classical infantile form of the disease, who are diagnosed too late for hematopoietic stem cell transplantation. For patients with either late-onset, slowly progressive disease or infantile disease prior to the onset of neurological manifestations, clinical improvements can occur by bone marrow transplantation (Krivit et al., 1998).

ANIMAL MODELS Genetic galactosylceramidase deWciency (Krabbe disease) occurs naturally in the mouse (twitcher), sheep, dogs (West Highland white terriers and Cairn terriers; blue-tick hound and beagles), and Rhesus monkeys. Clinical and pathological features of these models are similar to those of the human disease. Galactosylceramidase cDNA was cloned and disease-causing mutations have been identiWed in the mouse (Sakai et al., 1996), West Highland and Cairn terriers (Victoria et al., 1996), and the Rhesus monkey (Luzi et al., 1997). More details about animal models are found elsewhere in this volume (see Section V, au3 ‘‘Animal Models of Human Disease’’).

IV. DISEASES OF MYELIN

ANIMAL MODELS

References Andrews, J. M., and Menkes, J. H. (1970). Ultrastructure of experimentally produced globoid cells in the rat. Exp. Neurol. 29, 483–493. Austin J. H. (1963). Studies in globoid (Krabbe) leukodystrophy II. Controlled thin-layer chromatographic stuides of globoid body fractions in seven patients. J. Neurochem. 10, 921–930. Austin, J. H., and Lehfeldt, D. (1965). Studies in globoid (Krabbe) leukodystrophy. III. SigniWcance of experimentally-produced globoid-like elements in rat white matter and spleen. J. Neuropathol. Exp. Neurol. 24, 265–289. Barone, R., Bru¨hl, K., Stoeter, P., Fiumara, A., Pavone, L., and Beck, M. (1996). Clinical and neuroradiological Wndings in classic infantile and late-onset globoid-cell leukodystrophy (Krabbe disease). Am. J. Med. Genet. 63, 209–217. Cannizzaro, L. A., Chen, Y. Q., RaW, M. A., and Wenger, D. A. (1994). Regional mapping of the human galactocerebrosidase gene (GALC) to 14q31 by in situ hybridization. Cytogenet. Cell Genet. 66, 244–245. Chen, Y. Q., RaW, M. A., De Gala, G., and Wenger, D. A. (1993). Cloning and expression of cDNA encoding human galactocerebrosidase, the enzyme deWcient in globoid cell leukodystrophy. Hum. Mol. Genet. 2, 1841–1845. Collier, J., and GreenWeld, J. G. (1924). The encephalitis periaxialis of Schilder: A clinical and pathological study with an account of two cases, one of which as diagnosed during life. Brain 47, 489–519. Costantino-Ceccarini, E., and Morell, P. (1972). Biosynthesis of brain sphingolipids and myelin accumulation in the mouse. Lipids 7, 656–659. Fiumara, A., Pavone, L., Siciliano, L., Tine, A., Parano, E., and Innico, G. (1990). Late-onset globoid cell leukodystrophy. Report on seven new patients. Childs. Nerv. Syst. 6, 194–197. Hagberg, B., Sourander, P., and Svennerholm, L. (1963). Diagnosis of Krabbe’s infantile leukodystrophy. J. Neurosurg. Psychiat. 26, 195–204. Hannun, Y. A., and Bell, R. M. (1987). Lysosphingolipids inhibit protein kinase C: Implications for the sphingolipidoses. Science 235, 670–674. Hers, H. G. (1966). Inborn lysosomal disease. Gastroenterology 48, 625–633. Igisu, H., and Suzuki, K. (1984). Progressive accumulation of toxic metabolite in a genetic leukodystrophy. Science 224, 753–755. Krabbe, K. (1916). A new familial, infantile form of diVuse brain sclerosis. Brain 39, 74–114. Krivit, W., Shapiro, E. G., Peters, C., Wagner, J. E., Cornu, G., Kurtzberg, J., Wenger, D. A., Kolodny, E. H., Vanier, M. T., Loes, D. J., Dusenbery, K., and Lockman, L. A. (1998). Hematopoietic stem-cell transplantation in globoid-cell leukodystrophy. N. Engl. J. Med. 338, 1119–1126. Luzi, P., RaW, M. A., Victoria, T., Baskin, G. B., and Wenger, D. A. (1997). Characterization of the rhesus monkey galactocerebrosidase (GALC) cDNA and gene and identiWcation of the mutation causing globoid cell leukodystrophy (Krabbe disease) in this primate. Genomics 42, 319–324. Luzi, P., RaW, M. A., and Wenger, D. A. (1995). Structure and organization of the human galactocerebrosidase (GALC) gene. Genomics 26, 407–409. Miyatake, T., and Suzuki, K. (1972). Globoid cell leukodystrophy: Additional deWciency of psychosine galactosidase. Biochem. Biophys. Res. Commun. 48, 538–543. Norton, W. T., and Poduslo, S. E. (1973). Myelination in rat brain: changes in myelin composition during brain maturation. J. Neurochem. 21, 759–773. Oehlmann, R., Zlotogora, J., Wenger, D. A., and Knowlton, R. G. (1993). Localization of the Krabbe disease gene (GALC) on chromosome 14 by multipoint linkage analysis. Am. J. Hum. Genet. 53, 1250–1255. Percy, A. K., Odrezin, G. T., Knowles, P. D., Rouah, E., and Armstrong, D. D. (1994). Globoid cell leukodystrophy: Comparison of neuropathology with magnetic resonance imaging. Acta Neuropathol. (Berl) 88, 26–32. Sakai, N., Inui, K., Fujii, N., Fukushima, H., Nishimoto, J., Yanagihara, I., Isegawa, Y., Iwamatsu, A., and Okada, S. (1994). Krabbe disease: Isolation and characterization of a full-length cDNA for human galactocerebrosidase. Biochem. Biophys. Res. Commun. 198, 485–491. Sakai, N., Inui, K., Tatsumi, N., Fukushima, H., Nishigaki, T., Taniike ,M., Nishimoto, J., Tsukamoto, H., Yanagihara, I., Ozone, K., and Okada, S. (1996). Molecular cloning and expression of cDNA for murine galactocerebrosidase and mutation analysis of the twitcher mouse, a model of Krabbe’s disease. J. Neurochem. 66, 1118–1124. Stahl, N., Jurevics, H., Morell, P., Suzuki, K., and Popko, B. (1994). Isolation, characterization, and expression of cDNA clones that encode rat UDP-galactose:ceramide galactosyltransferase. J. Neurosci. Res. 38, 234–242. Suzuki, K. (1970). Ultrastructural study of experimental globoid cells. Lab. Invest. 23, 612–619. Suzuki, K. (1998). Twenty Wve years of the psychosine hypothesis: A personal perspective of its history and present status. Neurochem. Res. 23, 251–259. Suzuki, K., Schneider, E. L., and Epstein, C. J. (1971). In utero diagnosis of globoid cell leukodystrophy. Biochem. Biophys. Res. Commun. 45, 1363–1366. Suzuki, K., and Suzuki, K. (2002). Lysosomal disease. In ‘‘GreenWeld’s Neuropathology’’ (Graham, D. I., and Lantos, P. L., eds.), pp. 653–735. Edward Arnold, London. Suzuki, K., and Suzuki, Y. (1970). Globoid cell leucodystrophy (Krabbe’s disease): DeWciency of galactocerebroside ß-galactosidase. Proc. Natl. Acad. Sci., USA 66, 302–309.

IV. DISEASES OF MYELIN

849

850

35. KRABBE DISEASE

Suzuki, K., Tanaka, H., and Suzuki, K. (1976). Studies on the pathogenesis of Krabbe’s leukodystrophy: Cellular reaction of the brain to exogenous galactosylsphingosine, monogalactosyl diglyceride and lactosylceramide. In ‘‘Current Trends in Sphingolipidoses and Allied Disorders’’ (Volk, B. W., and Schneck, L., eds.), pp. 99–113. Plenum Press, New York. Suzuki, Y., and Suzuki, K. (1971). Krabbe’s globoid cell leukodystrophy: DeWciency of galactocerebrosidase in serum, leukocytes, and Wbroblasts. Science 171, 73–75. Svennerholm, L., Vanier, M.-T., and Ma˚nsson, J.-E. (1980). Krabbe disease: A galactosylsphingosine (psychosine) lipidosis. J. Lipid Res. 21, 53–64. Taketomi, T., and Nishimura, K. (1964). Physiological activity of psychosine. Jap. J. Exp. Med. 34, 255–265. Tohyama, J., Matsuda, J., and Suzuki, K. (2001). Psychosine is as potent an inducer of cell death as C6-ceramide in cultured Wbroblasts and in MOCH-1 cells. Neurochem. Res. 26, 667–671. Vanier, M.-T., and Svennerholm, L. (1975). Chemical pathology of Krabbe’s disease. III. Ceramide hexosides and gangliosides of brain. Acta Paediat. Scand. 64, 641–648. Vanier, M.-T., and Svennerholm, L. (1976). Chemical pathology of Krabbe disease: The occurrence of psychosine and other neutral sphingoglycolipids. In ‘‘Current Trends in Sphingolipidoses and Allied Disorders’’ (Volk, B. W., and Schneck, L., eds.), pp. 115–126. Plenum Press, New York. Victoria, T., RaW, M. A., and Wenger, D. A. (1996). Cloning of the canine GALC cDNA and identiWcation of the mutation causing globoid cell leukodystrophy in west highland white and cairn terriers. Genomics 33, 457–462. Wenger, D. A., Suzuki, K., Suzuki, Y., and Suzuki, K. (2001). Galactosylceramide lipidosis: Globoid cell leukodystrophy (Krabbe disease). In ‘‘The Metabolic and Molecular Basis of Inherited Disease’’ (Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D., eds.), pp. 3669–3694. McGraw-Hill, New York. Yunis, E. J., and Lee, R. E. (1970). Tubules of globoid cell leukodystrophy: A right-handed helix. Science 169, 64–66. Zlotogora, J., Chakraborty, S., Knowlton, R. G., and Wenger, D. A. (1990). Krabbe disease locus mapped to chromosome 14 by genetic linkage. Am. J. Hum. Genet. 47, 37–44.

IV. DISEASES OF MYELIN

C H A P T E R

36 Alexander Disease Albee Messing and James E. Goldman

EARLY HISTORY AND CLINICAL PRESENTATIONS In 1949, W. S. Alexander described a boy who died at 16 months of age with a history of megalencephaly, hydrocephalus, and psychomotor retardation (Alexander, 1949). A striking feature of the neuropathology in this child was the accumulation of Rosenthal Wbers within astrocytes, and an associated degeneration or failure of myelination. During the ensuing 15 years, several similar patients were reported and given such descriptive diagnoses as Wbrinoid degeneration of astrocytes, dysmyelogenic leukodystrophy, leukodystrophy with megalobarencephaly, Wbrinoid leukodystrophy, and megalencephaly with hyaline pan-neuropathy. When Friede described the sixth case in 1964, he also provided relief from the growing chaos in nomenclature by coining the eponym Alexander disease to refer to this fascinating but mysterious childhood disease (Friede, 1964). Subsequently, other patients were described who, despite sharing the common neuropathologic feature of prominent accumulation of Rosenthal Wbers, diVered markedly in age of onset, clinical presentation, and distribution of lesions (Borrett and Becker, 1985; Johnson, 1996; Pridmore et al., 1993; Russo et al., 1976). Hence, presently three main forms of Alexander disease are recognized: infantile, juvenile, and adult (Russo et al., 1976). There has been considerable debate about whether these are diVerent manifestations of the same disease or fundamentally diVerent disorders (Herndon, 1999), though recent genetic studies now shed light on this issue (discussed later). There is no sex predilection, and the disease occurs in diverse ethnic groups. The infantile form, with onset between birth and about 2 years of age, is the most common (Arend et al., 1991; Deprez et al., 1999; Klein and Anzil, 1994; Neal et al., 1992; Townsend et al., 1985; Wohlwill et al., 1959). It is usually accompanied by megalencephaly, but this is not invariantly present (Rodriguez et al., 2001). Seizures and developmental delay or regression are commonly found. Motor function gradually deteriorates to quadriparesis and spasticity. Hydrocephalus is often present, and though found occasionally along with stenosis of the cerebral aqueduct (Ni et al., 2002; Sherwin and Berthrong, 1970), there is no direct evidence that the stenosis results from Rosenthal Wber accumulation. There is often profound mental retardation, but other patients have been described with only late, minimal, or even no cognitive diYculties. Survival varies from only a few weeks to several years, with some patients surviving into their early teens. Nearly all cases are sporadic. A few putative sib ships have been described (discussed later), and two examples are known of monozygotic twins where both were aVected (Brenner et al., unpublished observations; Meins et al., 2002). Some investigators have argued for a distinct early-onset

Myelin Biology and Disorders, Volume 2

851

Copyright 2004, Elsevier (USA). All rights reserved.

852

36. ALEXANDER DISEASE

or neonatal form that is rapidly fatal and where seizures and elevated intracranial pressure are the predominant signs without spasticity (Springer et al., 2000). The juvenile form shows a later onset and initial signs may not be seen until the midteens (Deprez et al., 1999; Neal et al., 1992). Especially prominent are bulbar signs, with diYculties in swallowing or speech and vomiting, often accompanied by lower limb spasticity and incoordination. There may be some slow loss of intellectual function. The juvenile form generally progresses more slowly than the infantile, but the brain stem involvement can be life threatening. Some patients presented with signs resembling those of brain stem tumors, with the confusion resolved only by biopsy (Duckett et al., 1992). Adult-onset Alexander disease is the most variable and the least common form (Honnorat et al., 1993; Howard et al., 1993; Martidis et al., 1999; Okamoto et al., 2002; Rizzuto et al., 1980; Schwankhaus et al., 1995; Seil et al., 1968). Sometimes it mimics the juvenile form, but with a later onset and slower progression. Other times it may simulate multiple sclerosis. Some cases exhibit palatal myoclonus and there may be abnormal eye movements (Martidis et al., 1999). Three reports exist of apparent autosomal dominant inheritance. These individuals developed symptoms late and lived well into reproductive age (Howard et al., 1993; Okamoto et al., 2002; Schwankhaus et al., 1995). Okanoto, et al. (2002) describe a heterozygous GFAP mutation in parent and children (discussed later).

PATHOLOGY Alexander disease is usually grouped among the leukodystrophies because of the pronounced white matter deWciency seen in children with this disorder. Infants, who typically have rapid clinical courses, do not myelinate appropriately and manifest widespread destruction of white matter, even to the point of cavitation (Klein and Anzil, 1994; Schochet et al., 1968). Young patients typically show megalencephaly. Older children, who have a longer clinical course, show less white matter degeneration, although in longstanding cases glio-vascular scars with little myelin are seen in the deep white matter. In addition, older children show prominent involvement of the brain stem. Myelination in arcuate Wbers is relatively spared, as is the case in many leukodystrophies. Patients with adult forms of the disease may show only patchy zones of myelin pallor or cavitation (Honnorat et al., 1993; Schwankhaus et al., 1995; Spalke and Mennel, 1982) or more widespread myelin loss (Walls et al., 1984). Presumably, infants with early forms of the disorder never myelinate properly to begin with, while children and adults with later onset forms may myelinate and then demyelinate focally as the disease progresses. Rare reports of defects in PNS myelin are not convincing (Terao et al., 1983). EVects on neurons have also been observed, but have not been investigated thoroughly. A loss of axons has been reported as variable, ranging from none to severely diminished, particularly in the more gliotic regions (Borrett and Becker, 1985; Schochet et al., 1968). It is not clear when axonal degeneration begins during the evolution of the Alexander pathology, nor what causes axonal loss. Even when axonal degeneration is present, the myelin loss is far more severe (Walls et al., 1984). Similarly, neuronal somatal pathology has occasionally been reported. For example, Russo et al. (1976) described neuronal loss and chromatolytic changes in the brain stem. A number of authors have noted a paucity of oligodendrocytes in aVected areas, but oligodendrocyte death has not been carefully studied. The most characteristic Wnding in the Alexander brain is the presence of enormous numbers of Rosenthal Wbers. These astrocytic inclusions appear by routine staining as eosinophilic, refractile, often rod-shaped bodies, varying in size from less than 1 micron to dozens of microns in length (Fig. 36.1). Although Rosenthal Wbers are distributed in all regions of the CNS, they are particularly concentrated in astrocytic processes in the deep white matter, at the glial limitans and periventricular zones, in spinal white matter and intracranial regions of the optic nerve, and in the brain stem, particularly at the medullary level, where they accumulate in tegmental and ventral regions. Thalamic and basal ganglionic astrocytes can also accumulate large numbers of Rosenthal Wbers (TowWghi et al., 1983).

PATHOLOGY

FIGURE 36.1 Rosenthal Wbers concentrated in the astrocytic endfeet surrounding a blood vessel in the brain stem of a 1-year-old child with Alexander disease. Hematoxylin and eosin stain, paraYn section. Reprinted with permission from Elsevier (Lancet Neurology, 2003, 2, 75).

FIGURE 36.2 Rosenthal Wbers in an astrocyte cell body from a 17-month-old child with Alexander disease, viewed by electron microscopy. Reprinted from Eng et al., 1998, with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.

Although astrocytes in cortical gray matter do not seem as prone to develop Rosenthal Wbers, small inclusions are often found by careful observation. Examination of autopsies and biopsies of infants, or of children in the early stages of the disease, reveals small Rosenthal Wbers in cell bodies (Borrett and Becker, 1985; Townsend et al., 1985). During the course of the disease, the Rosenthal Wbers enlarge and eventually come to reside in astrocyte processes and endfeet, hence the pronounced subpial and perivascular localization typically seen. Ultrastructural examination of Rosenthal Wbers reveals intracellular osmiophilic deposits in intimate contact with bundles of intermediate Wlaments (Fig. 36.2) (Herndon et al.,

853

854

36. ALEXANDER DISEASE

1970; Seil et al., 1968). Rosenthal Wbers are not enclosed by membranes and show no lamellar features. The Wlaments contain GFAP (Johnson and Bettica, 1989) and vimentin (Tomokane et al., 1991), both normally expressed by astrocytes. The osmiophilic matrix contains GFAP as well as two members of the small heat shock protein (hsp) family, aBcrystallin (Iwaki et al., 1989) and hsp 27 (Iwaki et al., 1993) and likely other components yet to be deWned. A fraction of the aB-crystallin is ubiquitinated (Goldman and Corbin, 1991). Other post-translational modiWcations reportedly associated with Rosenthal Wbers include lipid peroxidation adducts (Castellani et al., 1998) and advanced glycation end products (Castellani et al., 1997). The amount of inXammation in the Alexander brain is variable. Some cases show appreciable lymphocytic accumulation around blood vessels, particularly in the brain stem (Russo et al., 1976; TowWghi et al., 1983), while others show little. It is not known what produces an inXammatory response or if this response contributes to the pathology. Several patients who have not manifested any neurological signs or symptoms have been found at autopsy to have widespread accumulation of Rosenthal Wbers in their brains (Mastri and Sung, 1973; Riggs et al., 1988). This accumulation was not accompanied by demyelination, however. All of these patients suVered such systemic illnesses as lymphoma, ovarian carcinoma, cardiac and respiratory insuYciency, diabetes, and myocardial infarction. Whether these individuals should be considered to have a form of Alexander disease is dubious, and whether there is a causal relationship between their illnesses and the deposition of Rosenthal Wbers, is not clear (Herndon, 1999). Finally, we note that while most patients with Alexander disease exhibit widespread pathology in the CNS, there are individuals with focal lesions, particularly in bulbar regions (Duckett et al., 1992; Goebel et al., 1981; Russo et al., 1976; SoVer and Horoupian, 1979). Localized brain stem pathology can be confused with pilocytic astrocytomas. Indeed, astrocytes in the Alexander brain can show a moderate degree of nuclear and cytoplasmic pleomorphism, thus compounding this rare diagnostic dilemma. Another consideration is that Rosenthal Wbers are not speciWc to Alexander disease; for example, sporadic Rosenthal Wbers can be found in the context of old glial scars, in pilocytic astrocytomas, or in the walls of syrinx cavities where they were Wrst described (Rosenthal, 1898).

CLINICAL DIAGNOSIS Until recently there has been no deWnitive laboratory diagnostic test for Alexander disease (but see the discussion of a new genetic analysis, presented later in the chapter). It is important to rule out other more common leukodystrophies and to diVerentiate Alexander disease from other causes of megalencephaly (Matalon et al., 1996). In particular, Canavan disease is indicated by a positive test for urinary N-acetylaspartic acid, very low activity of aspartoacylase in skin Wbroblasts, or detection of known mutations in the aspartoacylase gene (Kaul et al., 1993). One infantile patient presented with clinical signs and laboratory results indicative of Leigh’s encephalopathy, with elevations in serum pyruvate and lactate and CSF pyruvate, but was subsequently found on autopsy to have Alexander disease (Gingold et al., 1999). Brain lactate was also elevated in a patient with biopsyproven Alexander disease as determined by MR spectroscopy (Kang et al., 2001). Some have proposed evaluating CSF levels of HSP27 or aB-crystallin as indicators of Rosenthal Wber accumulation, but these Wndings are likely to be nonspeciWc (Takanashi et al., 1998). Radiology, especially MR imaging, are the current standard tools in the diagnosis of Alexander disease for both the infantile and juvenile forms (Hess et al., 1990; Takanashi et al., 1998; van der Knaap and Valk, 1995; van der Knaap et al., 2001). Typical cases show bilateral frontal predominance of white matter changes with relative sparing of the posterior and parietal lobes (Fig. 36.3). They also show abnormalities of the basal ganglia, especially the caudate and sometimes the thalami, and a periventricular rim of abnormal signal

855

GFAP MUTATIONS

A

B

C

D

E

F

FIGURE 36.3 MR images of a 10-year-old boy with Alexander disease. The axial T2-weighted images (A–C) show extensive abnormalities of the cerebral white matter with frontal predominance. The parieto-occipital white matter is partially spared (B, C). There is a thin periventricular rim of low signal intensity (B, C). The basal ganglia and thalamus show some signal changes and are mildly atrophic (B). The cerebellar white matter is abnormal, and there is a lesion in the dorsal medulla (A). After contrast administration, enhancement is seen of the lesion in the medulla (D), the dentate nucleus (E), the cerebellar surface (D, E), and parts of the ependymal lining of the lateral ventricles (F). Generously provided by Dr. Marjo van der Knaap.

intensity, particularly around the frontal horns. Sometimes the medulla and brain stem show changes, and less commonly the cerebellum. The white matter changes show decreased signal in T1-weighted images, but increased signal on T2-weighted images. Often there is swelling of the periventricular rim, with abnormal contrast enhancement. Other advanced imaging techniques such as PET (Bobele et al., 1990; Sawaishi et al., 1999) and NMR spectroscopy (Kang et al., 2001; Shiroma et al., 2001) have been used occasionally, showing reduced metabolism in white matter and low NAA/creatinine ratios, but have not signiWcantly aided in the diagnosis. Using the criteria of van der Knaap et al. (2001), diagnostic accuracy approaches 90% based on comparison with subsequent pathological analyses of the same patients. In atypical cases, however, pathological analysis of biopsy or autopsy samples is still essential for deWnitive diagnosis of Alexander disease.

GFAP MUTATIONS Although a genetic basis for Alexander disease had been speculated for some time, the rarity and sporadic nature of the disorder precluded traditional linkage analysis and few

856

36. ALEXANDER DISEASE

candidate genes were considered. As described earlier, the primary constituents of Rosenthal Wbers include GFAP, aB-crystallin, and hsp 27. A reasonable supposition is that an abnormality in one of these components might cause Alexander disease. Indeed, aB-crystallin was sequenced to investigate this possibility for two patients, but with negative results (Iwaki et al., 1992a). GFAP would be an even more attractive candidate, as it is expressed almost exclusively in astrocytes, the apparent focus for the disease. In fact, Becker and Teixiera (1988) suggested GFAP as a candidate gene more than a decade ago. However, it was the accidental discovery that transgenic mice engineered to constitutively overexpress GFAP developed a fatal encephalopathy with formation of bona Wde Rosenthal Wbers that gave new impetus to this idea (Messing et al., 1998). This discovery provided strong evidence that a primary alteration in the expression of GFAP could lead to the hallmark feature of Alexander disease. Prompted by these Wndings, Brenner et al. (2001) evaluated the GFAP coding region and proximal promoter in DNA from 13 patients who had died of biopsy-or autopsyproven Alexander disease. Nonconservative, heterozygous point mutations were found in 11 of 12 infantile cases and in the single older patient examined (who had onset at 10 years with survival to 48 years). These mutations were found to alter seven diVerent nucleotides, predicting changes in Wve diVerent amino acids (all arginines). Following this initial report, a number of other studies have now conWrmed and extended these Wndings (Aoki et al., 2001; 2002; Gorospe et al., 2002; Meins et al., 2002; Okamoto et al., Rodriguez, 2001; Sawaishi et al., 2002; Shiihara et al., 2002; Shiroma et al., 2001). A diagram showing the location of all published mutations in relation to the protein domains of GFAP and variant of Alexander disease is shown in Fig. 36.4, and detailed discussion of these mutations and other polymorphisms in GFAP can be found in the review by Li et al. (2002). Websites that will continue to provide current updates on GFAP mutations can be found at the University of Wisconsin-Madison (www.waisman.wisc.edu/alexander) and the Human Intermediate Filament Mutation Database (www.interWl.org). As of fall 2002, the GFAP gene has been evaluated in more than 60 Alexander disease patients. Mutations were present in nearly all (~95%) cases of infantile Alexander disease and in at least a certain proportion of juvenile and adult cases. Remarkably, mutations at only two amino acids, Arg79 or Arg239, account for nearly half of all cases. All of the mutations are heterozygous, presumably acting in an autosomal dominant fashion. For all of the infantile and juvenile cases where parents were available for testing, the parents were normal (i.e., did not have the mutation present in their child), conWrming that the mutations occurred de novo. The one example of a GFAP mutation being inherited occurred in a family of adult-onset cases where a mother and two adult children were aVected and carried the same mutation (Okamoto et al., 2002). The penetrance also approaches 100%, the only exceptions being two children whose initial evaluation for other problems led to MRI diagnoses of leukodystrophy, with subsequent genetic analysis revealing GFAP mutations (Gorospe et al., 2002), and one of these children is now showing signs. The Wnding of GFAP mutations in nearly all cases of Alexander disease, and of diverse types, supports a uniform underlying mechanism for the disease. However, the same data do not provide a strong case for genotype-phenotype correlations. For instance, two children with the juvenile form (onset at 4 and 10 years of age) have the same R416W mutation previously found in two infantile patients (Brenner et al., 2001; Gorospe et al., 2002). Rodriquez et al. have argued that the R79 mutations may cause a relatively mild phenotype compared to the R239 mutations (Rodriguez et al., 2001). The mutation associated with the single adult onset family, V87G (Okamoto et al., 2002), has not yet been found in any other patients, but more adult-onset cases need to be tested. Many of the GFAP mutations occur within an amino acid sequence that is highly conserved among intermediate Wlament proteins (Quinlan et al., 1995). Mutations at the homologous sites of other intermediate Wlament proteins are associated with human diseases involving skin blistering, cataracts, cardiomyopathies, and muscular dystrophies (reviewed by Quinlan, 2001). However, although homologous sites are aVected, most of these other mutations lead to a dominant loss of function, whereas the GFAP mutations appear to produce a dominant gain of function. For example, a loss of function is indicated

GFAP MUTATIONS

FIGURE 36.4 Location of Alexander’s disease-associated mutations in GFAP mutations in relation to protein domain structure of intermediate Wlaments. Like other intermediate Wlaments, the structure of GFAP consists of randomly coiled N-terminal and C-terminal regions Xanking four segments of an a-helical rod (shown by boxes) that are interconnected by nonhelical linkers. Mutations that are homologous to disease-causing mutations in other intermediate Wlaments are shown in black, whereas those that currently appear unique to GFAP are shown in red. Multiple independent cases with a given mutation are indicated by the number of symbols shown to the right (a single symbol is used for the set of cases from a single family). ClassiWcation of each case by age of onset is indicated by the color of the symbol (infantile ¼ teal; juvenile ¼ orange; adult ¼ black; asymptomatic ¼ open). N ¼ N-terminal, C ¼ C terminal. Adapted and updated from Figure 36.1 of Li et al., 2002.

for keratin mutations because both heterozygous mutations in humans and null mutations of the homologous gene in mice disrupt the keratin Wlament network and produce a similar blistering disease (reviewed by Fuchs and Cleveland, 1998). In contrast, GFAP Wlaments are present in Alexander disease patients, and GFAP null mice are fully viable and their pathology does not resemble Alexander disease (Gomi et al., 1995). Thus, the GFAP mutations do not appear to be acting by reducing or eliminating normal GFAP function, but rather by producing a new, deleterious, activity. The GFAP mutations could arise either in the developing embryo or in the germ cells of one of the parents. If the mutation arose early in development, all astrocytes might be aVected. However, if the mutation arose after the Wrst several cellular divisions of the embryo, it is possible that not all cells or CNS areas would be involved, and that the disease could be more or less severe depending on when in development the mutation occurred. Focal lesions, such as those mimicking brain stem gliomas, might be explained by such a mechanism. Finally, it is important to keep in mind that GFAP mutations have not been found in all cases of pathologically proven Alexander disease, suggesting that there could be other genetic or nongenetic causes for this disease. Since Rosenthal Wbers appear to form in response to chronic overexpression of GFAP, another potential cause is increased synthesis due to a mutation in the transcriptional control region or to gene duplication. Gene duplication has been found to cause other human neurological disorders (Lupski et al.,

857

858

36. ALEXANDER DISEASE

1991; Sistermans et al., 1998). However, we have found no evidence yet for duplication of GFAP in any patients (Hagemann, unpublished observations). There are a number of methodological and conceptual reasons that could explain these negative Wndings, and for the purpose of genetic counseling one must stress that this absence of information is simply that, and is unfortunately noninformative. GFAP mutations may also eventually be found associated with disorders other than Alexander disease—at present there is no hint as to what those disorders might be. The immediate clinical beneWt derived from the close association of GFAP mutations with Alexander disease is that invasive diagnostic procedures such as brain biopsy have now been replaced by rapid and accurate DNA analysis. In addition, parents now have the option of fetal testing for subsequent pregnancies.

FAMILIAL CASES For many years, Alexander disease was speculated to be a genetic disorder and likely autosomal recessive given the apparent normalcy of the parents (for instance, see discussion in Pridmore et al., 1993). In part, this speculation was fueled by occasional reports of families with more than one aVected sibling. Given the signiWcance attributed to these sib ships, and the understandable concern of parents regarding the risk of having other aVected children and the need for guidance in genetic counseling, a careful review of these reports is warranted. The Wrst was by Wohlwill et al. (1959), who described a family of nine children in which one sister and three brothers died between the ages of 3 and 6 years with macrocephaly and hydrocephalus. An autopsy of the last child revealed Alexander disease (though not so named at the time), but autopsies were not performed on any of the other children. Two subsequent reports appeared of infantile sibling pairs produced from consanguineous marriages, strongly implying a recessive genetic trait. However, all four of these patients were diagnosed as having Alexander disease based on clinical signs, progression, and radiology, without any pathological conWrmation (Barbieri et al., 1980; Springer et al., 2000). Three other reports exist of sibling pairs with later onset or very slow course of Alexander disease. In the Wrst and best-documented pair, Duckett et al. (1992) describe a brother and sister who developed symptoms at ages 11 and 26 years, respectively. Although the initial considerations pointed toward gliomas of the posterior fossa, biopsies of the brain stem (brother) and cerebellum (sister) instead revealed Rosenthal Wbers without any evidence of a tumor. A second pair, reported as a personal communication to V. McKusick (McKusick, 2001), presented at the ages of 11 and 13 years with nonprogressive macrocephaly, developmental delay, and radiological evidence of a leukodystrophy. Both of these patients are still alive in their twenties (P. Pearl, personal communication). The third pair (Klein et al., 1988) is very unusual in that both siblings had early onsets (birth to 6 months) but are known to have lived into their mid-twenties, with one still alive at the age of 26. The latter two pairs lack pathological conWrmation of the diagnosis. These sibships lack absolute certainty that all are aVected with Alexander disease. Nevertheless, it is worth considering how such familial cases could occur given the Wndings described earlier emphasizing heterozygous de novo mutations in GFAP. If the mutation arose in a parental germinal stem cell instead of in the patient, multiple gametes could carry the defect. Although highly unusual as a pattern of inheritance (Vogel and Motulsky, 1997), such gonadal mosaicism could account for the rare sibships described here. Thus, it will be of considerable interest to determine if aVected sibs have GFAP mutations and, if so, if the same mutation is found in each aVected family member. In addition, although all known GFAP mutations appear to act in the heterozygous state, and thus Wt with mechanisms known from other intermediate Wlament disease, it remains possible on theoretical grounds that some GFAP mutations could act in an autosomal recessive fashion to cause Alexander disease.

OTHER CANDIDATE GENES

Finally, three reports describe adult-onset syndromes in families in which more than one generation was aVected. All noted similar clinical signs of palatal myoclonus, ataxia, and paresis or paraplegia. These are the only examples of Alexander disease in which any of the parents have been aVected. In the Wrst family, a father developed signs in his early thirties and died after a 15-year course, whereupon autopsy revealed numerous Rosenthal Wbers (Seil et al., 1968). Subsequently, Schwankhaus et al. (1995) reported that three of his Wve children developed similar neurological signs beginning in their twenties to forties, and one was autopsy-conWrmed as Alexander disease. In a second family, Howard et al. (1993) reported that three of ten adult sibs developed similar syndromes in their twenties, with one having a biopsy showing Rosenthal Wbers. Their mother, maternal aunt, and two of four children of the aunt were reported by family members to have had similar neurological problems. Although the authors of this report distinguished their patients from what had been described as adult Alexander disease, the striking similarities between this family and that of Schwankhaus et al. suggest that these two families suVered from the same disease. Most recently, Okamoto et al. (2002) described a family of three individuals in which a mother and one adult child developed palatal myoclonus, spastic paraparesis, and atrophy of the caudal brain stem and spinal cord in their Wfties and thirties, respectively, with another adult child showing subtle neurological signs in his early thirties. Subsequent genetic analysis of this family indicated that all three shared the same mutation, thus supporting the concept that infantile, juvenile, and adult forms of adult Alexander disease can share a common etiology. The adult-onset families are consistent with the autosomal dominant pattern of inheritance predicted from the heterozygous nature of the GFAP mutations observed in Alexander disease.

OTHER CANDIDATE GENES As previously noted, a small subset of Alexander disease patients do not have GFAP mutations. Alexander disease might theoretically arise from alterations in proteins that interact with GFAP. There is precedent for the involvement of such genes in several other disorders associated with intermediate Wlament mutations. These include epidermolysis bullosa simplex, which can result from either mutations in keratins or the interacting plectin protein (Fuchs and Cleveland, 1998); Emery-Dreifuss muscular dystrophy, which can be produced by mutation in either nuclear lamins or the interacting emerin protein (RaVaele Di Barletta et al., 2000); and desmin-related myopathy, which can be produced either by mutations in desmin (Goldfarb et al., 1998) or the associated aB-crystallin (Vicart et al., 1998). aB-crystallin interacts with GFAP as well as with desmin (Nicholl and Quinlan, 1994; Wisniewski and Goldman, 1998) and thus remains a candidate gene for patients in whom a GFAP mutation is not present (the two patients whose aB-crystallin gene was sequenced were among those subsequently found to have GFAP mutations). Although none of the Alexander disease patients has been reported to have either the cataracts or myopathies observed in the one aB-crystallin mutation family documented (Vicart et al., 1998), diVerent mutations in the protein might lead to diVerent phenotypes. In the normal mammalian brain, aB-crystallin is expressed at low levels in astrocytes and oligodendrocytes (Iwaki et al., 1990), but accumulates in glia and neurons in a variety of neurological disorders (Iwaki et al., 1992b). In most of these examples, Rosenthal Wbers do not form, suggesting that its elevation per se does not trigger Rosenthal Wber formation. On the contrary, aB-crystallin appears to decrease Wlament aggregation, probably by inhibiting interactions between Wlaments (Koyama and Goldman, 1999; Perng et al., 1999). Thus, Rosenthal Wbers may form because conditions prevent aB-crystallin from organizing a normal intermediate Wlament network. Such conditions could include the presence of excess or mutant GFAP protein or a defect in the aB-crystallin itself. Another candidate gene for Alexander disease is NDUFV1, which encodes a component of mitochondrial complex I. In 1999, Schuelke and colleagues (Schuelke et al., 1999)

859

860

36. ALEXANDER DISEASE

described a young girl with mitochondrial complex I deWciency and clinical signs resembling Alexander disease. Sequencing of the NDUFVI open reading frame revealed homozygosity for a nonconservative mutation in the coding region. Both parents were conWrmed as heterozygotes, supporting a recessive mode of transmission. In the absence of pathological conWrmation, it cannot be certain that the diagnosis of Alexander disease in this child is accurate. If correct, however, this case would suggest a link to mitochondrial dysfunction as a pathway of injury in Alexander disease (Castellani et al., 1998), as also suggested by the patient initially thought to have Leigh’s encephalopathy mentioned earlier (Gingold et al., 1999).

POSSIBLE DISEASE MECHANISMS GFAP missense mutations have been found in a high proportion of cases of Alexander disease. A strong argument can be made that these mutations are responsible for the disease based on their homology to known disease-causing mutations in other intermediate Wlament genes, and the Wnding that they arose de novo in all instances that parental DNAs were analyzed (the probability of this occurring at random is miniscule). Formal proof that the GFAP mutations cause Alexander disease must await demonstration that their introduction into mice reproduces the salient features of the disease, or direct demonstration of a biological eVect of mutant protein. Mice are being engineered with point mutations of GFAP that correspond to the most common mutations found in Alexander disease patients, and preliminary analysis indicates that these mutations are indeed suYcient to induce formation of Rosenthal Wbers in brain (Hagemann, unpublished observations). How might GFAP mutations lead to Alexander disease? One approach to this question is to look for a common feature among instances in which Rosenthal Wbers form; for example, in response to chronic gliosis, in the GFAP transgenic mice, and as a result of GFAP mutations. In the Wrst two conditions there is a sustained elevation of GFAP. Elevated levels of GFAP may also be instrumental in Alexander disease, as Rosenthal Wbers are typically found in patients at sites where GFAP is normally highly expressed, such as the glial limitans, white matter, and the subependymal zone (which might also explain why pathology is not seen in nonastrocytic cells that are known to express GFAP, albeit at much lower levels, such as nonmyelinating Schwann cells or lens epithelium). Thus, one possible mechanism by which the GFAP mutations might cause the disease is by raising levels of the protein; for example, by increasing its stability. However, the mutations might instead cause accumulation of a particular form of GFAP that compromises astrocyte function, rather than by increasing the amount of total GFAP. For example, the mutant protein could assume a conformation, or receive a modiWcation, that occurs less frequently for the wild type protein and that leads to an abnormal association with other cellular constituents (possibly itself). Such a scenario has been reported for a-synuclein, a protein found in the Lewy bodies of Parkinson’s disease, and which is mutated in several familial cases of this disease (Dawson, 2000). Aggregation of a-synuclein is promoted either by increasing its level of expression or by the presence of the Parkinson’s disease–associated mutations (Conway et al., 1998; Dawson, 2000; Giasson et al., 1999; Narhi et al., 1999). Another way that the mutations could lead to accumulation of a toxic form of GFAP is by interfering with its usual polymerization into Wlaments. It was noted earlier that homologous mutations in other intermediate Wlament proteins disrupt their incorporation into normal Wlament networks. Although typical intermediate Wlaments are observed in Alexander disease, it is possible that they form with reduced eYciency, resulting in the accumulation of GFAP oligomers that react to form toxic products. Such side reactions might also be promoted by GFAP accumulation resulting from overexpression. It is unclear whether the Rosenthal Wber itself compromises astrocyte function, or whether it instead is formed as a protective mechanism to sequester aberrant GFAP-

POSSIBLE DISEASE MECHANISMS

containing complexes. In addition, the distribution of Rosenthal Wbers in the brain does not always coincide with the location of the most severe myelin defects (perhaps implying a pathogenic role for toxic soluble forms of GFAP, as discussed earlier). Recent evidence suggests that the protein aggregates associated with several other diseases may indeed be benign or protective, including the Lewy bodies of Parkinson’s disease (Mizuno et al., 1999), the inclusions found in Huntington’s disease (Kim et al., 1999), and the Mallory bodies present in liver cirrhosis (Zatloukal et al., 2000). It should also be borne in mind that there is no evidence for destruction of astrocytes in Alexander disease; instead, a primary clinical feature is hypomyelination or demyelination. Apparently the GFAP mutations lead to aberrant interactions between astrocytes and oligodendrocytes. However, Alexander disease astrocytes do display characteristics of physiological stress, as evidenced by the elevation of the small hsps aB-crystallin and hsp 27 (Head et al., 1993). Interestingly, these same stress proteins are increased in the GFAP overexpressing mice (Messing et al., 1998). Additional evidence for stress, and a suggestion that it may involve oxidative damage, is the association of lipid peroxidation products with Rosenthal Wbers (Castellani et al., 1998). On the other hand, the components of Rosenthal Wbers are readily dissociated by SDS or urea (Goldman and Corbin, 1988), and so do not display the extensive cross-linking that is found in many other disease-associated protein aggregates and that has been attributed to oxidative damage (Giasson et al., 2000). What consequences the stress response of Alexander disease astrocytes have for the functions of these cells and their interactions with other CNS cells remains to be determined. Whatever the triggering mechanism, once the disease process begins, it would likely lead to a catastrophic positive feedback loop. Almost any insult to astrocytes prompts a reactive response, which includes a strong upregulation of GFAP synthesis. Robustly reactive astrocytes are abundant in the GFAP overexpressing mice, and if GFAP mutations provoke a similar reactive response in humans, levels of the mutant protein would be further increased, reactivity further stimulated, and so on in a noxious spiral. Another positive feedback loop might operate through eVects on protein degradation. It has been recently suggested that the intermediate Wlament network may play a role in organizing degradative complexes that remove aberrant protein (Garcia-Mata et al., 1999; Johnston et al., 1998), and if this function were compromised by the mutant GFAP, another spiral of increasing levels of mutant GFAP would ensue. Finally, although the presence of GFAP mutations in Alexander disease suggests that this is a primary disorder of astrocytes, the clinical and pathological features suggest that there must also be signiWcant dysfunction of oligodendrocytes and perhaps neurons as well. For instance, patients with Alexander disease display either lack of proper myelin development (as in infants) or loss of myelination (as in older patients) at many levels of the CNS, including optic nerves, subcortical white matter, cerebellum, and spinal cord. We are therefore faced with a situation in which the expression of a mutated gene in one cell type (the astrocyte) has deleterious eVects on the functions of another cell type (the oligodendrocytes). What mechanism might underlie these eVects? One possibility is that astrocytes expressing a mutant GFAP undergo a stress response, possibly due to the accumulation of intermediate Wlaments, and as part of that response, secrete factors that are toxic to oligodendrocytes. These compounds could include TNF-a, a known stress protein that at least under some conditions induced oligodendrocyte cell death (Selmaj and Raine, 1988). A second possibility based on astrocyte toxicity is a secondary loss of axons, which would in turn prevent or cause loss of myelination. Axonal loss has not been demonstrated in Alexander disease, although clearly it takes place in severe cavitating lesions or in children with long-term clinical courses. Whether the broader features of the Alexander phenotype truly reXect secondary eVects of astrocyte dysfunction alone remains an open question. Clearly there is abundant evidence for how astrocytes might regulate the properties of both oligodendrocytes and neurons, but one should also consider the possibility that GFAP expression (and hence eVects of mutant GFAPs) might not strictly be limited to astrocytes. Brenner (1994) reviewed the transcriptional regulation of GFAP, and there are several isoforms about which much less is known than the major GFAPa species of mRNA. A recently identiWed

861

862

36. ALEXANDER DISEASE

isoform, termed GFAPe, arises by alternative splicing and produces a protein that interacts with presenilin, although the precise cell in which GFAPe is expressed is not yet clear (Nielsen et al., 2002). Transcriptional inWdelity might also lead to markedly abnormal forms of GFAP, some of which may be expressed in neurons (Hol et al., 2001; van Leeuwen et al., 1998). Finally, considerable evidence now suggests that either GFAP is expressed at low levels in a multipotential stem cell (Doetsch et al., 1999; Johansson et al., 1999; Laywell et al., 2000; Zhuo et al., 2001), or that radial glia (which in primates do express GFAP) regularly give rise to neurons as a normal pathway of diVerentiation (Campbell and Gotz, 2002; Malatesta et al., 2000; Noctor et al., 2001). If mutant GFAPs were to exert their eVects at the level of such stem cells, the defects could then manifest in a diverse population of cellular progeny.

SUMMARY Prior to Wnding that GFAP mutations underlie many cases of Alexander disease, it was unclear whether the disease originated in astrocytes or if the formation of Rosenthal Wbers was a response to an external insult. It was also unclear whether the etiology of the disease was environmental or genetic. For many cases of Alexander disease, these questions have now been answered. An immediate clinical beneWt of this discovery is the possibility of diagnosing most cases of Alexander disease through analysis of patient DNA samples, rather than resorting to brain biopsy. In addition, fetal testing is now an option for parents who have had an Alexander disease child with an identiWed mutation and who wish to have additional children. For the future, these mutations should provide a unique window for illuminating the mechanism of the disease, for further understanding the role of astrocytes and GFAP in myelination, and eventually for suggesting means of treatment.

Acknowledgments We thank our collaborators, especially Michael Brenner, all the Alexander patients, and their families who have participated in this research, and Marjo van der Knaap for contributing MRI images. This work was supported by grants from the NIH (NS-22475, NS-41803, and NS17125). This chapter is adapted and updated from a review published in the Journal of Neuropathology & Experimental Neurology (60:563, 2001).

References Alexander, W. S. (1949). Progressive Wbrinoid degeneration of Wbrillary astrocytes associated with mental retardation in a hydrocephalic infant. Brain 72, 373–381. Aoki, Y., Haginoya, K., Munakata, M., Yokoyama, H., Nishio, T., Togashi, N., Ito, T., Suzuki, Y., Kure, S., Iinuma, K., Brenner, M., and Matsubara, Y. (2001). A novel mutation in glial Wbrillary acidic protein gene in a patient with Alexander disease. Neurosci. Lett. 312, 71–74. Arend, A. O., Leary, P. M., and Rutherfoord, G. S. (1991). Alexander’s disease: A case report with brain biopsy, ultrasound, CT scan and MRI Wndings. Clinical Neuropathology 10, 122–126. Barbieri, F., Filla, A., De Falco, F. A., and Buscaino, G. A. (1980). Alexander’s disease. A clinical study with computerized tomographic scans of the Wrst two Italian cases. Acta Neurologica 2, 1–9. Becker, L. E., and Teixeira, F. (1988). Alexander’s disease. In ‘‘The Biochemical Pathology of Astrocytes’’ (M. D. Norenberg, L. Hertz, and A. Schousboe, eds.), pp. 179–190. Alan R. Liss, New York. Bobele, G. B., Garnica, A., Schaefer, G. B., Leonard, J. C., Wilson, D., Marks, W. A., Leech, R. W., and Brumback, R. A. (1990). Neuroimaging Wndings in Alexander’s disease. Journal of Child Neurology 5, 253–258. Borrett, D., and Becker, L. E. (1985). Alexander’s disease. A disease of astrocytes. Brain 108, 367–385. Brenner, M. (1994). Structure and transcriptional regulation of the GFAP gene. Brain Pathol. 4, 245–257. Brenner, M., Johnson, A. B., BoespXug-Tanguy, O., Rodriguez, D., Goldman, J. E., and Messing, A. (2001). Mutations in GFAP, encoding glial Wbrillary acidic protein, are associated with Alexander disease. Nature Genet. 27, 117–120. Campbell, K., and Gotz, M. (2002). Radial glia: Multi-purpose cells for vertebrate brain development. Trends Neurosci. 25, 235–238.

SUMMARY

Castellani, R. J., Perry, G., Harris, P. L. R., Cohen, M. L., Sayre, L. M., Salomon, R. G., and Smith, M. A. (1998). Advanced lipid peroxidation end-products in Alexander’s disease. Brain Res. 787, 15–18. Castellani, R. J., Perry, G., Harris, P. L. R., Monnier, V. M., Cohen, M. L., and Smith, M. A. (1997). Advanced glycation modiWcation of Rosenthal Wbers in patients with Alexander disease. Neurosci. Lett. 231, 79–82. Conway, K. A., Harper, J. D., and Lansbury, P. T. (1998). Accelerated in vitro Wbril formation by a mutant alphasynuclein linked to early-onset Parkinson disease. Nature Med. 4, 1318–1320. Dawson, T. M. (2000). New animal models for Parkinson’s disease. Cell 101, 115–118. Deprez, M., D’Hooghe, M., Misson, J. P., de Leval, L., Ceuterick, C., Reznik, M., and Martin, J. J. (1999). Infantile and juvenile presentations of Alexander’s disease: A report of two cases. Acta Neurologica Scandinavica 99, 158–165. Doetsch, F., Caille, I., Lim, D. A., Garcia-Verdugo, J. M., and Alvarez-Buylla, A. (1999). Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell 97, 703–716. Duckett, S., Schwartzman, R. J., Osterholm, J., Rorke, L. B., Friedman, D., and McLellan, T. L. (1992). Biopsy diagnosis of familial Alexander’s disease. Pediatr. Neurosurg. 18, 134–138. Eng, L. F., Lee, Y. L., Kwan, H., Brenner, M., and Messing, A. (1998). Astrocytes cultured from transgenic mice carrying the added human glial Wbrillary acidic protein gene contain Rosenthal Wbers. J. Neurosci. Res. 53, 353–360. Friede, R. L. (1964). Alexander’s disease. Arch. Neurol. 11, 414–422. Fuchs, E., and Cleveland, D. W. (1998). A structural scaVolding of intermediate Wlaments in health and disease. Science 279, 514–519. Garcia-Mata, R., Bebok, Z., Sorscher, E. J., and Sztul, E. S. (1999). Characterization and dynamics of aggresome formation by a cytosolic GFP-chimera. J. Cell Biol. 146, 1239–1254. Giasson, B. I., Duda, J. E., Murray, I. V. J., Chen, Q. P., Souza, J. M., Hurtig, H. I., Ischiropoulos, H., Trojanowski, J. Q., and Lee, V. M. Y. (2000). Oxidative damage linked to neurodegeneration by selective alpha-synuclein nitration in synucleinopathy lesions. Science 290, 985–989. Giasson, B. I., Uryu, K., Trojanowski, J. Q., and Lee, V. M. (1999). Mutant and wild type human alphasynucleins assemble into elongated Wlaments with distinct morphologies in vitro. J. Biol. Chem. 274, 7619–7622. Gingold, M. K., Bodensteiner, J. B., Schochet, S. S., and Jaynes, M. (1999). Alexander’s disease: Unique presentation. Journal of Child Neurology 14, 325–329. Goebel, H. H., Bode, G., Caesar, R., and Kohlschutter, A. (1981). Bulbar palsy with Rosenthal Wber formation in the medulla of a 15-year-old girl. Localized form of Alexander’s disease? Neuropediatrics 12, 382–391. Goldfarb, L. G., Park, K. Y., Cervenakova, L., Gorokhova, S., Lee, H. S., Vasconcelos, O., Nagle, J. W., Seminomora, C., Sivakumar, K., and Dalakas, M. C. (1998). Missense mutations in desmin associated with familial cardiac and skeletal myopathy. Nature Genet. 19, 402–403. Goldman, J. E., and Corbin, E. (1988). Isolation of a major protein component of Rosenthal Wbers. Am. J. Pathol. 130, 569–578. Goldman, J. E., and Corbin, E. (1991). Rosenthal Wbers contain ubiquitinated alpha B-crystallin. Am. J. Pathol. 139, 933–938. Gomi, H., Yokoyama, T., Fujimoto, K., Ideka, T., Katoh, A., Itoh, T., and Itohara, S. (1995). Mice devoid of the glial Wbrillary acidic protein develop normally and are susceptible to scrapie prions. Neuron 14, 29–41. Gorospe, J. R., Naidu, S., Johnson, A. B., Puri, V., Raymond, G. V., Jenkins, S. D., Pedersen, R. C., Lewis, D., Knowles, P., Fernandez, R., De Vivo, D., van der Knaap, M. S., Messing, A., Brenner, M., and HoVman, E. P. (2002). Molecular Wndings in symptomatic and pre-symptomatic Alexander disease patients. Neurology 58, 1494–1500. Head, M. W., Corbin, E., and Goldman, J. E. (1993). Overexpression and abnormal modiWcation of the stress proteins alpha B-crystallin and HSP27 in Alexander disease. Am. J. Pathol. 143, 1743–1753. Herndon, R. M. (1999). Is Alexander’s disease a nosologic entity or a common pathologic pattern of diverse etiology? Journal of Child Neurology 14, 275–276. Herndon, R. M., Rubinstein, L. J., Freeman, J. M., and Mathieson, G. (1970). Light and electron microscopic observations on Rosenthal Wbers in Alexander’s disease and in multiple sclerosis. J. Neuropathol. Exp. Neurol. 29, 524–551. Hess, D. C., Fischer, A. Q., Yaghmai, F., Figueroa, R., and Akamatsu, Y. (1990). Comparative neuroimaging with pathologic correlates in Alexander’s disease. Journal of Child Neurology 5, 248–252. Hol, E. M., Roelofs, R. F., Moraal, M., Sonnemans, M. A., Sluijs, J. A., van Tijn, P., Proper, E. A., de Graan, P. N., and van Leeuwen, F. W. (2001). Expression of novel splice variants of GFAP in human brain and neuronal accumulation of mutant GFAP protein in Alzheimer patients. Society for Neuroscience 429.10. Abstract. Honnorat, J., Flocard, F., Ribot, C., Saint-Pierre, G., Pineau, D., Peysson, P., and Kopp, N. (1993). [Alexander’s disease in adults and diVuse cerebral gliomatosis in 2 members of the same family]. [French]. Revue Neurologique 149, 781–787. Howard, R. S., Greenwood, R., Gawler, J., Scaravilli, F., Marsden, C. D., and Harding, A. E. (1993). A familial disorder associated with palatal myoclonus, other brainstem signs, tetraparesis, ataxia and Rosenthal Wbre formation. J. Neurol. Neurosurg. Psychiatry 56, 977–981. Iwaki, A., Iwaki, T., Goldman, J. E., Ogomori, K., Tateishi, J., and Sakaki, Y. (1992a). Accumulation of alpha Bcrystallin in brains of patients with Alexander’s disease is not due to an abnormality of the 5’-Xanking and coding sequence of the genomic DNA. Neurosci. Lett. 140, 89–92.

863

864

36. ALEXANDER DISEASE

Iwaki, T., Iwaki, A., Tateishi, J., Sakaki, Y., and Goldman, J. E. (1993). Alpha B-crystallin and 27-kd heat shock protein are regulated by stress conditions in the central nervous system and accumulate in Rosenthal Wbers. Am. J. Pathol. 143, 487–495. Iwaki, T., Kume-Iwaki, A., and Goldman, J. E. (1990). Cellular distribution of alpha B-crystallin in non-lenticular tissues. Journal of Histochemistry & Cytochemistry 38, 31–39. Iwaki, T., Kume-Iwaki, A., Liem, R. K. H., and Goldman, J. E. (1989). aB-Crystallin is expressed in nonlenticular tissues and accumulates in Alexander’s disease brain. Cell 57, 71–78. Iwaki, T., Wisniewski, T., Iwaki, A., Corbin, E., Tomokane, N., Tateishi, J., and Goldman, J. E. (1992b). Accumulation of alpha B-crystallin in central nervous system glia and neurons in pathologic conditions. Am. J. Pathol. 140, 345–356. Johansson, C. B., Momma, S., Clarke, D. L., Risling, M., Lendahl, U., and Frise´n, J. (1999). IdentiWcation of a neural stem cell in the adult mammalian central nervous system. Cell 96, 25–34. Johnson, A. B. (1996). Alexander disease. In ‘‘Handbook of Clinical Neurology’’ (H. G. Moser, ed.), Vol. 22 (66): Neurodystrophies and Neurolipidoses, pp. 701–710. Elsevier, Amsterdam. Johnson, A. B., and Bettica, A. (1989). On-grid immunogold labeling of glial intermediate Wlaments in epoxyembedded tissue. Am. J. Anat. 185, 335–341. Johnston, J. A., Ward, C. L., and Kopito, R. R. (1998). Aggresomes: A cellular response to misfolded proteins. J. Cell Biol. 143, 1883–1898. Kang, P. B., Hunter, J. V., and Kaye, E. M. (2001). Lactic acid elevation in extramitochondrial childhood neurodegenerative diseases. Journal of Child Neurology 16, 657–660. Kaul, R., Ping Gao, G., Balamurugan, K., and Matalon, R. (1993). Cloning of the human aspartoacylase cDNA and a common missense mutation in Canavan disease. Nature Genet. 5, 118–123. Kim, M., Lee, H. S., LaForet, G., McIntyre, C., Martin, E. J., Chang, P., Kim, T. W., Williams, M., Reddy, P. H., Tagle, D., Boyce, F. M., Won, L., Heller, A., Aronin, N., and DiFiglia, M. (1999). Mutant huntingtin expression in clonal striatal cells: Dissociation of inclusion formation and neuronal survival by caspase inhibition. J. Neurosci. 19, 964–973. Klein, E. A., and Anzil, A. P. (1994). Prominent white matter cavitation in an infant with Alexander’s disease. Clin. Neuropathol. 13, 31–38. Klein, S. K., Goulden, K. J., and Rapin, I. (1988). Evidence for autosomal recessive transmission of Alexander’s disease. Annals of Neurology 24, 302. Abstract. Koyama, Y., and Goldman, J. E. (1999). Formation of GFAP cytoplasmic inclusions in astrocytes and their disaggregation by aB-crystallin. Am. J. Pathol. 154, 1563–1572. Laywell, E. D., Rakic, P., Kukekov, V. G., Holland, E. C., and Steindler, D. A. (2000). IdentiWcation of a multipotent astrocytic stem cell in the immature and adult mouse brain. Proc. Natl. Acad. Sci. USA 97, 13883–13888. Li, R., Messing, A., Goldman, J. E., and Brenner, M. (2002). GFAP mutations in Alexander disease. Int. J. Dev. Neurosci. 20, 259-268. Lupski, J. R., Montes de Oca-Luna, R., Slaugenhaupt, S., Pentao, L., Guzzetta, V., Trask, B. J., SaucedoCardenas, O., Barker, D. F., Killian, J. M., Garcia, C. A., Chakravarti, A., and Patel, P. I. (1991). DNA duplication associated with Charcot-Marie-Tooth disease type 1A. Cell 66, 219–232. Malatesta, P., Hartfuss, E., and Go¨tz, M. (2000). Isolation of radial glial cells by Xuorescent-activated cell sorting reveals a neuronal lineage. Development 127, 5253–5263. Martidis, A., Yee, R. D., Azzarelli, B., and Biller, J. (1999). Neuro-ophthalmic, radiographic, and pathologic manifestations of adult-onset Alexander disease. Arch. Ophthalmol. 117, 265–267. Mastri, A. R., and Sung, J. H. (1973). DiVuse Rosenthal Wber formation in the adult: A report of four cases. J. Neuropathol. Exp. Neurol. 32, 424–436. Matalon, R., Michals, K., and Kaul, R. (1996). Canavan disease. In ‘‘Handbook of Clinical Neurology,’’ (H. G. Moser, ed.), Vol. 22 (66): Neurodystrophies and Neurolipidoses, pp. 661–669. Elsevier, Amsterdam. McKusick, V. A. Alexander disease. OMIM. 2001. Electronic Citation. Meins, M., Brockmann, K., Yadav, S., Haupt, M., Sperner, J., Stephani, U., and Hanefeld, F. (2002). Infantile Alexander disease: A GFAP mutation in monozygotic twins and novel mutations in two other patients. Neuropediatrics. In press. Messing, A., Head, M. W., Galles, K., Galbreath, E. J., Goldman, J. E., and Brenner, M. (1998). Fatal encephalopathy with astrocyte inclusions in GFAP transgenic mice. Am. J. Pathol. 152, 391–398. Mizuno, Y., Hattori, N., and Mori, H. (1999). Genetics of Parkinson’s disease. Biomedicine & Pharmacotherapy 53, 109–116. Narhi, L., Wood, S. J., Steavenson, S., Jiang, Y., Wu, G. M., AnaW, D., Kaufman, S. A., Martin, F., Sitney, K., Denis, P., Louis, J. C., Wypych, J., Biere, A. L., and Citron, M. (1999). Both familial Parkinson’s disease mutations accelerate alpha-synuclein aggregation [published erratum appears in J Biol Chem 1999 May 7;274(19):13728]. J. Biol. Chem. 274, 9843–9846. Neal, J. W., Cave, E. M., Singhrao, S. K., Cole, G., and Wallace, S. J. (1992). Alexander’s disease in infancy and childhood: A report of two cases. Acta Neuropathol. (Berl.) 84, 322–327. Ni, Q., Johns, G. S., Manepalli, A., Martin, D. S., and Geller, T. J. (2002). Infantile Alexander’s disease: Serial neuroradiologic Wndings. J. Child Neurol. 17, 463–466. Nicholl, I. D., and Quinlan, R. A. (1994). Chaperone activity of a-crystallins modulates intermediate Wlament assembly. EMBO J. 13, 945–953.

SUMMARY

Nielsen, A. L., Holm, I. E., Johansen, M., Bonven, B., Jørgensen, P., and Jørgensen, A. L. (2002). A new splice variant of glial Wbrillary acidic protein, GFAPe, interacts with the presenilin proteins. J. Biol. Chem. 277, 29983–29991. Noctor, S. C., Flint, A. C., Weissman, T. A., Dammerman, R. S., and Kriegstein, A. R. (2001). Neurons derived from radial glial cells establish radial units in neocortex. Nature 409, 714–720. Okamoto, Y., Mitsuyama, H., Jonosono, M., Hirata, K., Arimura, K., Osame, M., and Nakagawa, M. (2002). Autosomal dominant palatal myoclonus and spinal cord atrophy. J. Neurol. Sci. 195, 71–76. Perng, M. D., Cairns, L., van den IJssel, P., Prescott, A., Hutcheson, A. M., and Quinlan, R. A. (1999). Intermediate Wlament interactions can be altered by HSP27 and aB-crystallin. J. Cell Sci. 112, 2099–2112. Pridmore, C. L., Baraitser, M., Harding, B., Boyd, S. G., Kendall, B., and Brett, E. M. (1993). Alexander’s disease: Clues to diagnosis. J. Child. Neurol. 8, 134–144. Quinlan, R. (2001). Cytoskeletal catastrophe causes brain degeneration. Nature Genet. 27, 10–11. Quinlan, R., Hutchison, C., and Lane, B. (1995). Intermediate Wlament proteins. Protein ProWle 2, 801–952. RaVaele Di Barletta, M., Ricci, E., Galluzzi, G., Tonali, P., Mora, M., Morandi, L., Romorini, A., Voit, T., Orstavik, K. H., Merlini, L., Trevisan, C., Biancalana, V., Housmanowa-Petrusewicz, I., Bione, S., Ricotti, R., Schwartz, K., Bonne, G., and Toniolo, D. (2000). DiVerent mutations in the LMNA gene cause autosomal dominant and autosomal recessive Emery-Dreifuss muscular dystrophy. Am. J. Hum. Genet. 66, 1407–1412. Riggs, J. E., Schochet, S. S. J., and Nelson, J. (1988). Asymptomatic adult Alexander disease: Entity or nosological misconception? Neurology 38, 152–154. Rizzuto, N., Ferrari, G., and Piscioli, A. (1980). DiVuse Rosenthal Wber formation in adults. A case report. Acta Neuropathol. (Berl.) 50, 237–240. Rodriguez, D. (2001). Infantile Alexander disease: Spectrum of GFAP mutations and genotype-phenotype correlation (vol 69, pg 1134, 2001). Am. J. Hum. Genet. 69, 1413. Rodriguez, D., Gauthier, F., Bertini, E., Bugiani, M., Brenner, M., N’guyen, S., Goizet, C., Gelot, A., Surtees, R., Pedespan, J.-M., Hernandorena, X., Troncoso, M., Uziel, G., Messing, A., Ponsot, P., Pham-Dinh, D., Dautigny, A., and BoespXug-Tanguy, O. (2001). Infantile Alexander disease: Spectrum of GFAP mutations and genotype-phenotype correlation. Am. J. Hum. Genet. 69, 1134–1140. ¨ ber eine eigenthu¨mliche, mit syringomyelie complicirte geschwulst des ru¨ckenmarks. Rosenthal, W. (1898). U Bietr. Pathol. Anat. 23, 111–143. Russo, L. S., Jr., Aron, A., and Anderson, P. J. (1976). Alexander’s disease: A report and reappraisal. Neurology 26, 607–614. Sawaishi, Y., Hatazawa, J., Ochi, N., Hirono, H., Yano, T., Watanabe, Y., Okudera, and Takada, G. (1999). Positron emission tomography in juvenile Alexander disease. J. Neurol. Sci. 165, 116–120. Sawaishi, Y., Yano, T., Takaku, I., and Takada, G. (2002). Juvenile Alexander’s disease with a novel mutation in glial Wbrillary acidic protein gene. Neurology 58, 1541–1543. Schochet, S. S., Lampert, P. W., and Earle, K. M. (1968). Alexander’s disease: A case report with electron microscopic observations. Neurology 18, 543–549. Schuelke, M., Smeitink, J., Mariman, E., LoeVen, J., Plecko, B., Trijbels, F., Stockler-Ipsiroglu, S., and van den Heuvel, L. (1999). Mutant NDUFV1 subunit of mitochondrial complex I causes leukodystrophy and myoclonic epilepsy. Nature Genet. 21, 260–261. Schwankhaus, J. D., Parisi, J. E., Gulledge, W. R., Chin, L., and Currier, R. D. (1995). Hereditary adult-onset Alexander’s disease with palatal myoclonus, spastic paraparesis, and cerebellar ataxia. Neurology 45, 2266–2271. Seil, F. J., Schochet, S. S. J., and Earle, K. M. (1968). Alexander disease in an adult. Report of a case. Arch. Neurol. 19, 494–502. Selmaj, K. W., and Raine, C. S. (1988). Tumor necrosis factor mediates myelin and oligodendrocyte damage in vitro. Ann. Neurol. 23, 339–346. Sherwin, R. M., and Berthrong, M. (1970). Alexander disease with sudanophilic leukodystrophy. Archives of Pathology & Laboratory Medicine 89, 321–328. Shiihara, T., Kato, M., Honma, T., Ohtaki, S., Sawaishi, Y., and Hayasaka, K. (2002). Fluctuation of computed tomographic Wndings in white matter in Alexander’s disease. Journal of Child Neurology 17, 227–230. Shiroma, N., Kanazawa, N., Izumi, M., Sugai, K., Fukumizu, M., Sasaki, M., Hanaoka, S., Kaga, M., and Tsujino, S. (2001). Diagnosis of Alexander disease in a Japanese patient by molecular genetic analysis. Journal of Human Genetics 46, 579–582. Sistermans, E. A., Decoo, R. F. M., Dewijs, I. J., and Vanoost, B. A. (1998). Duplication of the proteolipid protein gene is the major cause of Pelizaeus-Merzbacher disease. Neurology 50, 1749–1754. SoVer, D., and Horoupian, D. S. (1979). Rosenthal Wbers formation in the central nervous system. Its relation to Alexander’s disease. Acta Neuropathol. (Berl.) 47, 81–84. Spalke, G., and Mennel, H. D. (1982). Alexander’s disease in an adult: Clinicopathologic study of a case and review of the literature. Clinical Neuropathology 1, 106–112. Springer, S., Erlewein, R., Naegele, T., Becker, I., Auer, D., Grodd, W., and Krageloh-Mann, I. (2000). Alexander disease - ClassiWcation revisited and isolation of a neonatal form. Neuropediatrics 31, 86–92. Takanashi, J., Sugita, K., Tanabe, Y., and Niimi, H. (1998). Adolescent case of Alexander disease: MR imaging and MR spectroscopy. Pediatr. Neurol. 18, 67–70. Terao, Y., Ishi, M., Hamada, T., and Murata, R. (1983). [Ultrastructural changes of skin nerve in Alexander’s disease]. [Japanese]. Nippon Hifuka Gakkai Zasshi - Japanese Journal of Dermatology 93, 1533–1535.

865

866

36. ALEXANDER DISEASE

Tomokane, N., Iwaki, T., Tateishi, J., Iwaki, A., and Goldman, J. E. (1991). Rosenthal Wbers share epitopes with alpha B-crystallin, glial Wbrillary acidic protein, and ubiquitin, but not with vimentin. Immunoelectron microscopy with colloidal gold. Am. J. Pathol. 138, 875–885. TowWghi, J., Young, R., Sassani, J., Ramer, J., and Horoupian, D. S. (1983). Alexander’s disease: Further light-, and electron-microscopic observations. Acta Neuropathol. (Berl.) 61, 36–42. Townsend, J. J., Wilson, J. F., Harris, T., Coulter, D., and Fife, R. (1985). Alexander’s disease. Acta Neuropathol. (Berl.) 67, 163–166. van der Knaap, M. S., Naidu, S., Breiter, S. N., Blaser, S., Stroink, H., Springer, S., Begeer, J. C., Van Coster, R., Barth, P. G., Thomas, N. H., Valk, J., and Powers, J. M. (2001). Alexander disease: Diagnosis with MR imaging. Am. J. Neuroradiol. 22, 541–552. van der Knaap, M. S., and Valk, J. (1995). Magnetic Resonance of Myelin, Myelination, and Myelin Disorders. In Magnetic Resonance of Myelin, Myelination, and Myelin Disorders (Berlin: Springer-Verlag), pp. 259–264. van Leeuwen, F. W., Burbach, J. P., and Hol, E. M. (1998). Mutations in RNA: A Wrst example of molecular misreading in Alzheimer’s disease. [Review] [48 refs]. Trends Neurosci. 21, 331–335. Vicart, P., Caron, A., Guicheney, P., Li, Z. L., Prevost, M. C., Faure, A., Chateau, D., Chapon, F., Tome, F., Dupret, J. M., Paulin, D., and Fardeau, M. (1998). A missense mutation in the alphaB-crystallin chaperone gene causes a desmin-related myopathy. Nature Genet. 20, 92–95. Vogel, F., and Motulsky, A. G. (1997). Human Genetics: Problems and Approaches. (Berlin: Springer), pp. pg. 409–410. Walls, T. J., Jones, R. A., Cartlidge, N., and Saunders, M. (1984). Alexander’s disease with Rosenthal Wbre formation in an adult. J. Neurol. Neurosurg. Psychiatry 47, 399–403. Wisniewski, T., and Goldman, J. E. (1998). aB-crystallin is associated with intermediate Wlaments in astrocytoma cells. Neurochem. Res. 23, 385–392. Wohlwill, F. J., Bernstein, J., and Yakovlev, P. I. (1959). Dysmyelinogenic leukodystrophy: Report of a case of a new, presumably familial type of leukodystrophy with megalobarencephaly. J. Neuropathol. Exp. Neurol. 18, 359–383. Zatloukal, K., Stumptner, C., Lehner, M., Denk, H., Baribault, H., Eshkind, L. G., and Franke, W. W. (2000). Cytokeratin 8 protects from hepatotoxicity, and its ratio to cytokeratin 18 determines the ability of hepatocytes to form mallory bodies. Am. J. Pathol. 156, 1263–1274. Zhuo, L., Theis, M., Willecke, K., Brenner, M., and Messing, A. (2001). hGFAP-cre transgenic mice for manipulation of glial and neuronal function in vivo. genesis 31, 85–94.

C H A P T E R

37 Pelizaeus-Merzbacher Disease Lynn D. Hudson, James Y. Garbern, and John A. Kamholz

A HISTORICAL PERSPECTIVE OF PELIZAEUS-MERZBACHER DISEASE (PMD) In 1885, Friedrich Pelizaeus described a family with an unusual inherited disease that now bears his name. In this description he also noted ‘‘that the disease is passed on by the mother but does not hurt her’’ (Pelizaeus, 1885) consistent with an X-linked mode of inheritance (Boulloche and Aicardi, 1986; Seitelberger, 1970). Twenty-Wve years later, Ludwig Merzbacher reinvestigated 12 aVected individuals from the same family and performed a detailed pathological analysis of the brain of one of its members. In this analysis he identiWed the widespread loss of myelin in the cortical white matter (Merzbacher, 1910). In his exhaustive clinical description, Merzbacher noted that the disease began in early neonatal life with aimless, wandering eye movements, followed by nystagmus. Infants failed to develop normal head control and displayed tremors or shaking movements of the head. The disease was slowly progressive, with additional signs including bradylalia, scanning speech, ataxia and intention tremor of the upper limbs, spastic contractions of the lower limbs, athetotic movements, and cognitive impairment (Merzbacher, 1910; Pelizaeus, 1885). A disorder with pathology similar to that described by Merzbacher was reported by Franz Seitelberger in 1954 (Seitelberger, 1954). In this condition, however, there was nearly complete absence of myelin sheaths and a profound loss of myelin-forming oligodendrocytes. Seitelberger suggested that this disease, which he called the connatal form of PMD, was similar to that described by Pelizaeus and Merzbacher, which he designated the classical form of PMD. In addition, he noted that in both disorders the absence of myelin was the primary biochemical defect, suggesting that both were leukodystrophies. Zeman and coworkers subsequently suggested that the defect in PMD resided in the proteolipid protein (PLP; also known as lipophilin or Folch-Lees protein), the major protein component of myelin (Zeman et al., 1964), a hypothesis veriWed by the sequencing of mutations in the PLP gene of several patients with the disease a quarter of a century later (Gencic et al., 1989; Hudson et al., 1989; Trofatter et al., 1989). In the early 1990s, BoespXug-Tanguy and collaborators found genetic linkage of spastic paraparesis type 2 to the Xq22 region where PMD also mapped, suggesting that PLP1 mutations could produce this syndrome. Then these researchers identiWed several patients with PLP1 mutations who presented with spastic paraplegia without the other signs of PMD (Saugier-Veber et al., 1994; reviewed in Nave and BoespXug-Tanguy, 1996). Evaluation of additional patients with X-linked spastic paraplegia and mutations in the PLP1 gene showed that this syndrome could exist as either a ‘‘complicated’’ or a milder ‘‘pure’’ form in which the clinical phenotype is conWned to lower limb spasticity. The detection of families in which PMD and SPG2 coexist emphasize the broad clinical continuum of these disorders (Tab. 37.1), all of which

Myelin Biology and Disorders, Volume 2

867

Copyright 2004, Elsevier (USA). All rights reserved.

TABLE 37.1 Phenotype Connatal PMD

Neurologic Wndings

Age of onset Neonatal period

Spectrum of PLP-Related Disorders

*

Ambulation

Speech

Lifespan

Nystagmus at birth Pharyngeal weakness * Stridor * Hypotonia * Severe spasticity * + Seizures * Cognitive impairment

Never achieved

Absent

Death in childhood to third decade

*

Classic PMD

First year

*

Nystagmus in Wrst two months * Initial hypotonia * Spastic quadriparesis * Ataxia titubation * + Dystonia, athetosis * Cognitive impairment

With assistance if achieved; lost in childhood/adolescence

Usually present

Death in 3rd to 7th decade

PLP1 null syndrome

First 1–5 years

*

Present

Present; usually worsens after adolescence

Death in 5th to 7th decade

+ Nystagmus Ataxia * Autonomic dysfunction (spastic urinary bladder) * Spastic gait * Little or no cognitive impairment

þ

Present

Normal

*

þ

Present

Normal

No nystagmus

*

Mild spastic quadriparesis Ataxia * Peripheral neuropathy * Mild to moderate cognitive impairment *

Complicated spastic paraplegia (SPG2)

First 1–5 years

Pure spastic paraplegia (SPG2)

First 1–5 years

*

Autonomic dysfunction (spastic urinary bladder) Spastic gait * Normal cognition *

GENETICS OF PMD/SPG2

share a phenotype of spasticity and hypomyelination (Cambi, F et al., 1996; Bond et al., 1997; Kobayashi et al., 1994; Osaka et al., 1995).

GENETICS OF PMD/SPG2 The Proteolipid Protein (PLP) Gene Is Mutated in PMD and SPG2 Proteolipid protein (PLP), the predominant protein of CNS myelin (reviewed in Chapter 16) is one of nature’s most hydrophobic proteins. What gives this integral membrane protein extra hydrophobic character is an unusual degree of fatty acid acylation. Six fatty acid chains are covalently linked to a PLP molecule (Weimbs and StoVel, 1992), and those fatty acids attached to the intracellular loop of PLP have been proposed to mediate the association of PLP with the adjacent lipid leaXet in compact myelin (Sporkel et al., 2002, see Fig. 3B in Chapter 16). Acylation occurs autocatalytically at a stage following translation of PLP mRNA (Bizzozero et al., 1987; Ross and Braun, 1988), a temporal pattern consistent with a role for this post-translational modiWcation in the stabilization or compaction of myelin. PLP is synthesized in the rough endoplasmic reticulum as a tetraspan intrinsic membrane protein with both termini on the cytoplasmic face (Gow et al., 1997; Wahle and StoVel, 1998) and subsequently transported through the Golgi complex, where other myelin lipid constituents such as cholesterol and galactocerebroside associate with PLP in ‘‘rafts’’ (Simons et al., 2000). Raft formation is one of the initial stages of myelin assembly and is followed by the vesicular transport of PLP to the myelin membrane. The exceptional nature of PLP as a protein is echoed in the gene, which is extremely well conserved. PLP gene structure is preserved among tetrapods and readily discernible in the primordial gene of the lipophilin family present in invertebrates (Stecca et al., 2000). Mammals share a nearly identical coding capacity for PLP (reviewed in Chapter 16). Moreover, no amino acid polymorphisms have been detected in the thousands of coding regions sequenced in the human PLP gene. PLP is encoded by a single gene, composed of seven exons located on the X chromosome (Xq22.2) (Diehl et al., 1986; Ikenaka et al., 1988; Macklin et al., 1987; StoVel et al., 1984) (see Fig. 1, Chapter 16). The Wrst exon ends one base after the initiator methionine, which is cleaved oV the nascent protein. An additional exon with the potential of encoding an alternate amino terminus was reported in mouse (Bongarzone et al., 1999) but is not present in the human gene. The third exon contains an internal donor splice site, which is used to generate transcripts encoding the smaller (20 kDa) DM20 isoform. While identical to PLP in topology, DM20 is missing part of the intracellular loop that contains two acylation sites, an absence which may account for the altered conformation and physical properties observed for DM20 (Gow et al., 1997; Helynck et al., 1983; Skalidis et al., 1986). Like PLP, DM20 is abundantly produced, estimably at 60% of the level of PLP (Schindler et al., 1990), and the two proteins can form heteromers (McLaughlin et al., 2002). PLP and DM20 perform distinct roles in the maintenance of myelin structure, as DM20 cannot fully compensate for a loss of PLP in myelin (Sporkel et al., 2002; Stecca et al., 2000). Other functions have been proposed for PLP/DM20 based on features of these lipophilin family members resembling channel proteins, the detection of secreted fragments of PLP/DM20, and the expression in other glial cells as well as outside of the nervous system (reviewed in Chapter 16). The intriguing hypothesis that PLP acts as a sensor in transmitting information across the lipid bilayer (Gow and Lazzarini, 1996) was validated by the discovery that PLP, but not DM20, interacts with av-integrin as part of a signaling complex (Gudz et al., 2002). Apart from oligodendrocytes, the PLP gene is transcriptionally active in the nervous system in olfactory ensheathing cells (Dickinson et al., 1997), satellite cells (GriYths et al., 1995), and Schwann cells (Garbern et al., 1997; GriYths et al., 1989; Puckett et al., 1987), where the predominant isoform expressed is DM20 (PhamDinh et al., 1991). Schwann cell expression of PLP/DM20 is an order of magnitude lower than that observed in oligodendrocytes, and most of the proteins produced are not

869

870

37. PELIZAEUS-MERZBACHER DISEASE

normally incorporated into the myelin sheath (Anderson et al., 1997; Garbern et al., 1997). A low level of PLP/DM20 expression also occurs outside of the nervous system, in the heart (Campagnoni et al., 1992), fetal thymus, spleen (Pribyl et al., 1996), thyroid, testes, and skin (SkoV, unpublished). In general, cells other than myelinating oligodendrocytes tend to favor the synthesis of DM20 over PLP. Even in oligodendrocytes, the DM20 expression proWle is not always coincident with myelination, as immature oligodendrocytes selectively express DM20 (Ikenaka et al., 1992; Schindler et al., 1990; Timsit et al., 1992; Timsit et al., 1995; Yu et al., 1994).

Mutations Result in Overexpression, Loss-of-Function, or Gain-of-Function of PLP In keeping with the conserved nature of the PLP gene, all types of mutations at the PLP locus have discernible eVects in humans. Most frequently encountered are duplications of the PLP gene, which have been estimated to account for 60 to 70% of cases (Inoue et al., 1996; Mimault et al., 1999; Sistermans et al., 1998; Wang et al., 1997), as depicted in Figure 37.1. PLP duplications are typically tandem in nature, involving a large genomic segment that includes neighboring genes (Inoue et al., 1996; Inoue et al., 1999a; Woodward et al., 1998). Striking variation in the position of the breakpoints occurs in diVerent PMD families (Inoue et al., 1999a; Woodward et al., 1998), unlike the situation with other inherited duplications such as Charcot-Marie-Tooth disease type 1A (CMT1A) (reviewed in Inoue and Lupski, 2002). The duplicated segment can be as large as three megabases (Mb), more than 150 times the size of the PLP locus (Inoue et al., 1999a). Therefore, not only will PLP be overexpressed in these patients, a number of other X-linked genes will be inappropriately expressed. Only a fraction of genes are sensitive to dosage eVects, and in the segments of the X chromosome duplicated in the PMD patients, PLP is apparently the sole gene for which changes in copy number spawn phenotypic aberrations. Not that the families with duplications display a uniform phenotype (Inoue et al., 1999a; Sistermans et al., 1998; Woodward et al., 1998), as the position of the breakpoints can disrupt other X-linked genes, which if haplo-insuYcient could contribute to the overall phenotype. That PLP is a gene subject to dosage control is reinforced by the description of possibly three copies of the PLP gene in patients with a more severe form of PMD (Harding et al., 1995; Woodward et al., 1998). Unequal sister chromatid exchange in male meiosis is the major mechanism leading to duplication of the PLP gene (Inoue et al., 1999a; Mimault et al., 1999). Additional mechanisms of genomic rearrangements operate at the PLP locus, as indicated by several families in which the duplicated copy invades another spot on the X chromosome (Hodes et al., 2000). Despite the large number of duplications arising from sister chromatid exchange, the expected reciprocal recombination event, namely deletion of the PLP locus, rarely occurs (Inoue and Lupski, 2002; Raskind et al., 1991). Deletion of the PLP gene encompasses a much smaller segment of the X chromosome, with only two neighboring genes (Inoue and Lupski, 2002). Probably the deletion of larger sections of the X chromosome, which would comprise the majority of reciprocal recombination events arising from duplications of PLP, would cause lethality or infertility. By examining the deletion breakpoints in the three identiWed families, Lupski and coworkers discovered several diVerent modes of genome rearrangement (Inoue et al., 2002). This study reinforces the complexities of recombination involving the PLP locus that were initially observed with the PLP duplications and suggests that nonhomologous joining of ends causes PLP deletions (Inoue et al., 2002). In addition to the loss-of-function mutations arising from deletion events, two point mutations in the PLP coding region at the initiation codon (Sistermans et al., 1996) or the second codon (Garbern et al., 1997) are null for PLP expression. Unlike the PLP deletions characterized to date, these null point mutations allow for a direct examination of PLP loss without complicating considerations from deletion of those genes neighboring PLP, namely the RAS superfamily member RAB9L and the thymosin b family member TMSNB (Inoue et al., 2002). About 20% of PMD patients have point mutations (single base changes or small deletions or insertions) at the PLP locus that alter the amino acid sequence of the PLP/

871

Number of Patients

GENETICS OF PMD

Connatal PMD

Transitional PMD

Classical PMD

Complicated spastic paraparesis Pure SPG

More severe

Clinical severity

Less severe Deletion / null mutations Duplication / triplication Point mutation

FIGURE 37.1 The clinical spectrum and relative frequency of PLP1 mutations. The range of clinical syndromes caused by PLP1 mutations is a set of overlapping distributions rather than a linear gradation from very severe to mild disease. The areas under the curves approximate the observed frequencies of clinical subtypes.

DM20 proteins. These include missense, nonsense, frameshift, and splicing mutations (Fig. 37.2), all of which produce abnormal PLP/DM20 proteins. Of the abnormal PLP/ DM20 proteins, a majority result in severely aVected patients through a toxic gain-offunction (discussed later), while the remainder create a milder form of the disease that may be categorized as loss-of-function. The loss-of-function class of abnormal proteins cannot fully perform the roles of PLP or DM20, but they do not take on new roles in oligodendrocytes. Approximately 100 distinct mutations have been discovered to date (for an up-todate accounting of the various point mutations, refer to www.med.wayne.edu/Neurology/ plp.html). An extensive collection of missense mutations (those mutations that result in amino acid substitution) in the PLP/DM20 gene exist. Certain amino acid codons have a particularly rich array of changes that oVer an opportunity to investigate the consequences of a speciWc protein alteration on myelination and the clinical manifestations of PMD/ SPG2 (Cailloux et al., 2000; Hodes et al., 1999). Three codons were mutated in two missense versions (V165E, V165G; L223I, L223P; Q233Z, Q233P), and one codon was subjected to Wve diVerent missense mutations: the aspartate at position 202 was changed to an asparagine, histidine, valine, glycine, or glutamate residue in diVerent PMD patients. Codon 202, located in the large external loop of PLP/DM20 (Fig. 37.2), represents a mutational ‘‘hot spot.’’ Indeed, the entire external loop has an excessive number of mutations. While mutations are distributed throughout the PLP/DM20 coding sequence, appearing in both the transmembrane and extra-membrane domains, half of the missense mutations occur within the large external loop. The susceptibility of this region hints at conformational cues that may be important in maintaining the intraperiod line in compact myelin. A signiWcant number of mutations are also available in the intracellular domain that is speciWc for PLP, including a nonsense mutation in exon3B, that enable a comparison of the roles of DM20 and PLP in the myelin sheath of man (Tab. 37.1/Fig. 37.2). A number of splice site mutations have been uncovered in PMD patients. Of most interest are the splicing mutations that are not located at the strictly conserved positions in the donor and acceptor splice sites, including a deletion of 19 bp within intron 3 and 26 bp in intron 5 (Cailloux et al., 2000; Hobson et al., 2000). Although the spliced products have not been characterized in these families, splicing mutations would most likely result in

872

37. PELIZAEUS-MERZBACHER DISEASE

FIGURE 37.2 Point mutations of the PLP1 gene. The orientation of the tetraspan PLP and DM20 proteins in the plasma membrane and myelin sheath is depicted together with the acylation sites (cysteine residues #5,6,8,108,138, and 140) and the disulWde bridges in the large extracellular loop. Citations for the mutations described in this Wgure can be found at www.med.wayne.edu/Neurology/ClinicalPrograms/PelizaeusMerzbacher/plp.html. When more than one mutation occurs at a single position, the most severe phenotype is indicated. For completeness, some mutations are presented that have been described without clinical information. When limited clinical information was available, the disease severity was assigned as follows: very severe (equivalent to ‘‘connatal’’) syndrome was ascribed if the patient was explicitly described as connatal, neurologic signs were present at birth or death occurred before 20 years of age; severe: neurologic signs were present in Wrst few months of life, patient was described explicitly as intermediate or classical PMD or death occurred after 20 years of age; complicated SPG: patient was able to walk eVectively for at least a few years and had CNS signs in addition to spastic paraparesis; SPG2: ‘‘pure’’ spastic paraparesis not associated with other CNS signs.

skipping of an exon, an event that would create an internally deleted and possibly a frameshifted abnormal PLP protein. However, mutations within intron 3 that eliminate the donor splice site have the potential of leaving the DM20 transcript and protein unscathed. The atypical splicing mutations at the PLP locus (Cailloux et al., 2000; Hobson et al., 2000) suggest that even more splicing mutations may be found in PMD/SPG2 patients, mutations that have eluded detection because sequencing eVorts usually concentrate on coding regions and intron/exon junctions.

DIAGNOSIS, PRESENTATION AND CLINICAL COURSE OF PMD/SPG2

Another category of point mutations is the regulatory mutations that alter the expression of the PLP gene without aVecting the protein sequence. A putative promoter mutation has been reported in a PMD family at
34 of the PLP1 gene (Kawanishi et al., 1996). Whether this is the causative change that alters PLP/DM20 expression in the reported family is not known. Additional changes may occur within regulatory elements of the PLP gene, which are not yet fully deWned, or splice sites that aVect PLP gene expression. Nonetheless, the C to T transition at
34 is of interest as it is within the area bound by the RNA polymerase complex prior to the commencement of transcription at the upstream initiation site (reviewed in Chapter 16). A second class of regulatory mutations that create a PMD-like disorder may arise in genes encoding transcription factors that recognize the PLP promoter. The transcription factors that directly bind to the PLP promoter are candidates (reviewed in Chapter 16), as are factors known to aVect PLP expression, such as the homeodomain protein Nkx2.2 (Fu et al., 2002; Qi et al., 2001) or the high-mobilitygroup regulator Sox10 (Stolt et al., 2002). One Sox10 mutation has been described that combines features of PMD, Charcot-Marie-Tooth disease type 1, and WaardenburgHirschsprung syndrome (Inoue et al., 1999b). The occurrence of such mutations is rare, as no additional Sox10 mutations were found when screening 56 patients with CharcotMarie-Tooth disease or 88 leukodystrophies, all patients previously sequenced for the usual candidate genes (Pingault et al., 2002). Nonethess, this cohort of patients may be aVected in Nkx2.2 (Fu et al., 2002; Qi et al., 2001) or one of the other transcription factors that inXuence PLP expression. Apart from the broader phenotypes expected from the mutation of a transcription factor that acts on multiple target genes, the absence of Xlinkage in disorders caused by mutated transcription factors would genotypically distinguish them from PMD patients with mutations in the PLP1 gene.

DIAGNOSIS, PRESENTATION AND CLINICAL COURSE OF PMD/SPG2 The nomenclature of clinical syndromes arising from PLP1 mutations has generated confusion and some controversy, and stems from the variability of syndromes caused by diVerent mutations and from the variable expressivity of an individual mutation among family members (see Fig. 37.1 and Tab. 37.1). The most consistent features of PMD include spasticity, a lack of evidence of male-to-male transmission in the family, and generalized leukodystrophy on magnetic resonance imaging scans. However, even these relaxed criteria, applied too strictly, might exclude some patients who have only ataxia and tremor, or those infants with severe mutations who have hypotonia at onset. The diagnosis of Pelizaeus-Merzbacher disease (PMD) is thus suggested in aVected individuals by the presence of a characteristic set of neurological signs and symptoms, including nystagmus, spastic paraparesis, and limb ataxia, a family history of disease consistent with an X-linked recessive pattern of inheritance, and an MRI scan demonstrating diVuse central nervous system abnormalities of myelination. The diagnosis can be unequivocally established in about 80% of patients by molecular genetic testing to identify a PLP1 gene duplication or PLP1 gene mutation, as discussed in the previous section on genetic mechanisms and in the discussion that follows. Although the clinical presentation and course of PMD varies depending upon the nature of the PLP1 mutation, the disease usually presents in one of three typical patterns. The most severe form of disease, connatal PMD, begins during the Wrst weeks of life, and is associated with hypotonia, respiratory distress, stridor, nystagmus, and sometimes seizures. Because of the prominence of hypotonia and respiratory symptoms, connatal PMD can be confused with motor neuron disease or spinal muscular atrophy (Kaye et al., 1994). Individuals with connatal PMD go on to develop severe spasticity with little voluntary movement, and never ambulate. In addition, they have very poor head control and cannot sit unsupported. Growth is poor, and they develop very limited language skills. These individuals usually die before the third decade of life. The most common form of disease, in contrast, classic PMD, begins during the neonatal period, usually within the Wrst

873

874

37. PELIZAEUS-MERZBACHER DISEASE

year of life and is associated with nystagmus, lower extremity weakness, and head titubation. Respiration is normal. Muscle tone is often reduced during infancy, but progresses to spasticity later during childhood. Motor milestones are also usually delayed in classic PMD, and most individuals never walk independently. Patients with this disease go on to develop a spastic quadraparesis, worse in the lower than in the upper extremities. Ataxia of trunk and limb movements is also a prominent feature of classic PMD, and dystonic posturing and movements also occur. Most individuals acquire some degree of language skill, which may even approach normal levels, but the speech is dysarthric and the speed of language output is usually slow. In addition, patients with the classic form of PMD also have some degree of cognitive disability. These patients can survive until the sixth decade of life. The mildest form of PMD merges clinically with syndromes of X-linked spastic paraparesis (SPG2). This disease begins during childhood, usually within the Wrst 5 years of life and is associated with a mild to severe spastic paraplegia, although there may also be limb and gait ataxia. Patients with SPG2 may have nystagmus as an early sign (Bonneau et al., 1993; Saugier-Veber et al., 1994) but others may not develop this, or it may occur as a late sign and manifest as end-gaze rather than primary position nystagmus (Garbern et al., 1997; Johnston and McKusick, 1962). Motor milestones are usually delayed, but most individuals learn to walk independently during childhood, although this ability may be lost later in life. Language skills and intelligence can be normal, although there also may be mild cognitive impairment. Individuals with this form of PMD usually have a normal life span. MRI scans are usually abnormal, but the Wndings can be regional or very subtle in comparison to those of patients with the more severe forms of PMD (see Fig. 37.3). Table 37.1 summarizes these three typical patterns of PMD, based on our own clinical observations as well as those of Boulloche and Aicardi (1986), Hodes et al. (1993), and Cailloux et al. (2000). Magnetic resonance imaging (MRI) analysis of the brain is essential in the evaluation of individuals with clinical signs and symptoms of PMD/SPG2 (Ono et al., 1994; Nezu et al., 1998). Virtually all patients with PMD eventually have MRI Wndings consistent with a leukodystrophy, including diVusely increased signal intensity within the central white matter of the cerebral hemispheres, cerebellum, and brain stem, best seen on either T2-weighted or Xuid attenuated inversion recovery (FLAIR) sequences, as shown in Figure 37.3. Although the relative white matter volume can be reduced and the corpus callosum thinned, brain structure is otherwise normal, including the ventricular system, basal ganglia, and cortical surface. Because myelination is an ongoing process during postnatal development, the signal intensity of myelinated tracks is not constant during this time. For this reason, the T2-weighted or FLAIR images may not be unequivocally abnormal in PMD until a child is older than 2 years of age. Patients with the more mild spastic paraplegia phenotype have similar MRI changes, but these may be less pronounced or more patchy in nature (Cambi et al., 1995; Hodes et al., 1999) (Fig. 37.3). Other diagnostic considerations for individuals with the clinical features of PMD include metachromatic leukodystrophy, adrenoleukodystrophy, Krabbe disease, Cockayne disease, and Canavan disease. None of these diseases, however, are associated with nystagmus, which is common in PMD, and their diagnosis can usually be made by analysis of the appropriate lysosomal enzyme. In addition, the white matter abnormalities in these conditions are often regional rather than diVuse: The occipital white matter is most aVected in adrenoleukodystrophy, while the frontal white matter is most aVected in metachromatic leukodystrophy. Infants with merosin deWciency can also have dramatically increased T2 signal in the cerebral white matter, but the presence of severe weakness and hypotonia and the absence of nystagmus should direct the clinician toward consideration of myopathy. A fatal X-linked syndrome of ataxia, blindness, deafness, and mental retardation has been described and is linked to Xq21-24, but the MRI does not show a pattern of leukodystrophy, and mutations in the PLP coding regions have been excluded. Finally, mutations in the cell adhesion molecule gene L1CAM cause X-linked spastic paraplegia type 1 (SPG1), a disorder associated with mental retardation and adducted thumbs, which is allelic to the MASA syndrome (mental retardation, aphasia, shuZing gait, adducted thumbs) and X-linked hydrocephalus. MRIs of these disorders may show

DIAGNOSIS, PRESENTATION AND CLINICAL COURSE OF PMD/SPG2

875

FIGURE 37.3 MRI of PLP1 mutations with a spectrum of severity. All images are T2 weighted scans of aVected males, with the exception of the control. Ages of the patients are control, 17 year old female; PLP1 null, 17 years; duplication, 12 years; Pro14Leu, 20 years; Ile186Thr, 45 years; and IVS3 deletion, 9 years. The Pro14Leu mutation represents the severely aVected, connatal form of the disease, the duplication corresponds to the intermediate classical form, and the remaining images represent the mild SPG2 syndrome (PLP null, Ile186Thr and the splicing defect marked 19bp del IVS 3). The arrowhead points to areas of frontal lobe white matter and illustrates the diVerence in signal abnormality among this group of patients (on T2 weighted scans signal brightness increases with myelin abnormality). Note that the SPG2 patients have normal to subtle signal abnormalities compared to the duplication and connatal PMD patients.

enlarged ventricles or agenesis of the corpus callosum, but not the diVuse abnormalities of white matter consistent with a leukodystrophy. Women with a PLP1 gene mutation may have neurological signs and symptoms, but are not usually index cases. Several investigators have observed that, in families with severely aVected males, the heterozygous women are unlikely to have clinical manifestations of PMD/SPG2, whereas in families with mildly aVected males, the heterozygous women are more likely to have symptoms (Bond et al., 1997; Sivakumar et al., 1999). In a family with a particularly mild syndrome characterized by ataxia and mild spastic paraplegia, all three heterozygous females had neurologic signs (Hodes et al., 1997). Animals with mutations of PLP1 also exist in which severe alleles may cause transient neurologic signs in heterozygotes (Cuddon et al., 1998) and mild alleles cause persistent neurologic signs in heterozygotes (Fanarraga et al., 1991). These interesting phenomena will be discussed further in a subsequent section, and in Chapter 47.

876

37. PELIZAEUS-MERZBACHER DISEASE

Peripheral Neuropathy in PMD Some patients with PMD have a demyelinating peripheral neuropathy (Garbern et al., 1997). The neuropathy in these patients is mild, however, and not usually clinically signiWcant. Electrophysiological studies demonstrate areas of modestly slowed nerve conduction velocities distributed nonuniformly along the nerve (Garbern et al., 1999). Genetic Testing in PMD Approximately 80% of patients with clinical, genetic, and MRI features consistent with PMD have been found to have PLP1 mutations. The genetic etiology for the other 20% of patients is not known but may be due either to mutations in areas of the PLP1 gene not routinely analyzed, such as introns and regulatory regions, or the presence of an additional autosomal or X-linked mutation that can cause the same phenotype (Osaka et al., 1999). Duplication of the region surrounding the PLP1 gene at Xq22 accounts for the majority of mutations in patients with PMD (Mimault et al., 1999), perhaps up to 70%, while point mutations, small deletions, or insertions make up the rest. Deletion of the entire PLP1 gene has been identiWed in a small number of patients with PMD (BoespXug-Tanguy et al., 1999; Raskind et al., 1991), and interstitial duplications or more complex rearrangements of the X chromosome visible on routine cytogenetic studies have been found in several others (Carrozzo et al., 1997; Cremers et al., 1987; Zackai et al., 1997). Because duplication of the PLP1 region is the most common cause of PMD, identiWcation of a PLP1 gene duplication is the most eYcient initial genetic screening test for diagnosing PMD. The duplications are of variable size, but they are usually found within an 800-kb region of the X chromosome including the PLP1 gene (Inoue et al., 1996; Inoue et al., 1999a; Sistermans et al., 1998; Woodward et al., 1998). The duplicated region can also be found, however, at some distance from Xq22. One PLP1 duplication has been identiWed at Xp22 and a second at Xq28 (Hodes et al., 2000; Woodward et al., 1998). Both interphase Xuorescent in situ hybridization (FISH) and quantitative polymerase chain reaction (QPCR) have been used to detect PLP1 duplications (Hobson et al., 2000; Inoue et al., 1996; Inoue et al., 1999a; ShaVer et al., 1997). Duplications smaller than 50 kb, however, may not be resolved by FISH, while QPCR does not provide important cytogenetic information on the location of the duplication (ShaVer et al., 1997). For these reasons, both methods, interphase FISH and QPCR, should be routinely employed for the molecular genetic diagnosis of PMD. If neither interphase FISH nor QPCR demonstrates a PLP1 duplication, direct sequence analysis of the PLP1 gene should be performed. The PLP1 gene encodes a relatively small protein of 277 amino acids (831 bp of DNA) and the coding sequences are contained within only seven exons. Using automated sequencing methods, therefore, it is cost-eVective as well as technically straightforward to obtain the DNA sequence of the PLP1 exons and portions of their surrounding introns. When a small mutation is found, it is sometimes possible to design an allele-speciWc oligonucleotide hybridization test or a simple PCR/restriction digestion assay to detect the mutation (Hobson et al., 2000), which can be particularly helpful for conWrmation of carrier status in females. Prenatal DNA diagnosis of PMD in aVected males at risk for the disease has been accomplished by several groups of investigators but is not routinely available. A PLP1 duplication in an at-risk male fetus has been identiWed using both interphase FISH (Inoue et al., 2001a) and QPCR (Regis et al., 2001). A PLP1 point mutation in an at-risk male fetus, however, has not been reported, although prenatal testing has successfully excluded such mutations (Maenpaa et al., 1990; Strautnieks et al., 1992). Preimplantation genetic diagnosis for PMD is possible but has not been reported. IdentiWcation of a PLP1 duplication in single cells is also possible, although technical diYculties currently preclude its use as a diagnostic tool in patients. Genotype-Phenotype Correlations No simple correlation has been found between a particular PLP1 mutation or genotype, and the clinical manifestation of the disease or phenotype. Although most patients with

MOLECULAR PATHOGENESIS OF PMD

duplications have the classic form of PMD (Inoue et al., 1999a), some have the more severe connatal form (Ellis and Malcolm, 1994), while others have a milder spastic paraparesis. Inoue and coworkers recently analyzed the duplication size and structure in 20 families with PMD and suggested that the size of the PLP1 duplication correlated with the clinical phenotype (Inoue et al., 1999a), so that patients with larger duplications, had more severe disease. In a similar study of 16 families with PLP duplications, however, Hobson and coworkers did not conWrm this Wnding (Hobson, unpublished), suggesting that other structural features of the duplication, such as the location, breakpoint, or orientation, may also play a role. Callioux and colleagues recently compared the clinical phenotype and genotype in 33 families with PLP1 point mutations (Cailloux et al., 2000). They found that single amino acid changes within evolutionarily conserved regions of the protein produced the most severe disease, while substitutions of less conserved amino acids, protein truncations, null mutations, and mutations within the PLP1-speciWc region (amino acids 116–150) produced a milder form of disease. Although exceptions to this rule occur, more severe forms of disease are likely to be associated with missense mutations within highly conserved regions of the protein. Garbern and coworkers analyzed several families with PLP1 mutations in which no protein product was produced, so-called null mutations (Garbern et al., 1997; Garbern et al., 2002). These patients all had a relatively mild spastic paraparesis, which progressed during adolescence as well as an associated demyelinating peripheral neuropathy identiWed during electrophysiological testing. The neuropathy was not correlated with disease severity and was not found in patients with either duplications or point mutations in which protein was produced. Taken together, these data suggest that individuals with a relatively mild form of disease and peripheral neuropathy are likely to have a null mutation.

MOLECULAR PATHOGENESIS OF PMD Gow and Lazzarini proposed that diVerences in clinical severity in patients with PLP1 coding region mutations can be accounted for by a diVerential eVect of the speciWc mutation on the folding and intracellular traYcking of the protein (reviewed in Southwood and Gow, 2001). Mutations that aVect the folding and transport to the cell surface of both PLP and DM20 are associated with the most severe PMD phenotypes and also cause increased oligodendrocyte cell death, while mutations that impair transport of PLP but not DM20 produce a less severe PMD phenotype that is not associated with oligodendrocyte cell death (Gow and Lazzarini, 1996; Gow et al., 1998). Since mutations in which no mutant protein is synthesized cause the mildest disease, the predominant eVect of PLP1 coding region mutations is probably due to misfolded PLP1 gene products. The cellular and molecular eVects of the accumulation of misfolded PLP1 and DM20 in the RER of oligodendrocytes, rather than the absence of these proteins in the myelin sheath, are thus the cause of the clinical signs and symptoms of PMD. Not only does the protein-misfolding hypothesis explain the diVerential clinical eVect of PLP mutations, it also explains the eVect of these mutations on female carriers. Female dogs that are heterozygous for a severe mutation in the canine PLP, for example, have neurological abnormalities early in life, but by adulthood are clinically normal, have normal numbers of oligodendrocytes, and express very little mutant PLP messenger RNA (Cuddon et al., 1998). Female PMD carriers are also usually clinically unaVected, although some may have transient neurological abnormalities as children (Hodes et al., 1995). In some PMD families, however, female heterozygotes are clinically aVected, as in the family described by Pelizaeus and Merzbacher (Pelizaeus, 1885). Because of random inactivation of the X chromosome on which the PLP gene is located, females who are heterozygous for PLP mutations should express the abnormal protein in approximately 50% of their oligodendrocytes. Oligodendrocytes expressing a more severe PLP mutation, however, in which both PLP and DM20 are aVected, undergo increased cell death and are eliminated during

877

878

37. PELIZAEUS-MERZBACHER DISEASE

myelination and replaced by normal oligodendrocytes. In contrast, oligodendrocytes expressing a less severe PLP mutation, which does not cause cell death, are not eliminated, thus producing abnormal myelin and neurological dysfunction. Paradoxically then, females who are heterozygous for the less severe PLP mutations are more likely to experience neurological diYculties as adults than are females who are heterozygous for the more severe PLP alleles (Hodes et al., 1995; Inoue et al., 2001b; Sivakumar et al., 1999; Sambuughin et al., 1998). Similar observations have been made with experimental and naturally occurring murine PLP mutations (see Chapter 47). How does accumulation of misfolded DM-20/PLP1 in the ER of oligodendrocytes aVect their function? Several lines of evidence point to the involvement of the unfolded protein response (UPR), a network of genes that are induced in response to unfolded proteins and that act to regulate expression of molecular chaperones, transcription factors, caspases, and other genes (reviewed by Kaufman et al., 2002). Two bZip transcription factors, CHOP (CEBPb-homologous protein) and ATF3, are induced during the UPR and have been shown to cause apoptosis when overexpressed in transfected cells. Gow and coworkers recently found that both CHOP and ATF3 expression as well as several other RER-resident molecular chaperones are similarly induced in oligodendrocytes in response to the synthesis of mutant PLP1 gene products, implicating the UPR in the pathogenesis of oligodendrocyte cell death in PMD (Southwood et al., 2002). These investigators also discovered that rumpshaker(rsh) mice without a functional Chop gene (rsh/chop-null double mutants) have a more severe disease than rsh mice, directly implicating CHOP expression in the pathogenesis of PMD. It is thus likely that the set of genes induced during the UPR plays a role in PMD pathogenesis by protecting oligodendrocytes from the toxic eVects of misfolded DM-20 and PLP. Protein misfolding has been implicated as a pathogenic mechanism in several other neurodegenerative diseases, including Alzheimer’s, Parkinson’s, and Huntington’s diseasse (Aridor and Balch, 1999; Kopito and Ron, 2000; Taylor et al., 2002). The morphological features associated with protein accumulation in these diseases include amorphous aggregates in the RER, cytoplasm or nucleus, and intermediate Wlament-containing aggresomes in the cytoplasm (Johnston et al., 1998). Perinuclear inclusions are also observed in a variety of cell types, particularly in cultured cells treated with proteasome complex inhibitors, and are thought to form when the RER-to-cytoplasm delivery of unfolded proteins exceeds degradation by the proteasome complex. Aggresome-like inclusions are rarely found in PMD, however, because myelinating oligodendrocytes do not normally synthesize intermediate Wlaments. Proliferating oligodendrocyte precursor cells express vimentin and nestin in culture, but the expression of these genes is switched oV as the cells diVerentiate (Almazan et al., 2001). Although protein misfolding has been implicated in all of these diseases, the molecular mechanisms of oligodendrocyte cell death in PMD may thus be diVerent than those in the more classic neurodegenerative diseases. A second pathogenic mechanism in PMD is associated with the overexpression of PLP in patients with duplications of the PLP gene. Excessive amounts of normal PLP proteins have been shown to accumulate in the late endosome and lysosomal compartments of rodent cells overexpressing PLP (Simons et al., 2002). Since PLP typically associates with cholesterol and other lipids to form myelin ‘‘rafts’’ as it traYcks through the Golgi compartment (Simons et al., 2000), the shunting of excess PLP into the endosomal/lysosomal compartment eVectively drains myelin lipids from the Golgi (Simons et al., 2002). Presumably the transport and assembly of myelin constituents is altered in cells overexpressing PLP. Thus, while abnormal PLP proteins trigger a protein misfolding response in the rough endoplasmic reticulum, excessive PLP proteins create an imbalance in myelin constituents that adversely aVects the subsequent stage of nascent myelin assembly in the Golgi network. Occasionally females with a duplication of the PLP gene manifest on earlyonset neurological phenotype (Inoue et al., 2001b). Like some carriers of PLP point mutations, these patients with mild PMD or spastic paraplegia show sustained clinical improvement. The recovery of these heterozygous females is probably attributable to the same mechanism favored for the point mutations, namely compensatory myelin production by the normal oligodendrocytes that contain only a single copy of the PLP gene.

MOLECULAR PATHOGENESIS OF PMD

A third mechanism of molecular pathogenesis in PMD/SPG2 occurs by loss-of-function, in patients with a deletion of the PLP gene (BoespXug-Tanguy et al., 1999; Inoue et al., 2002; Raskind et al., 1991) or with point mutations at the beginning of the coding region that preclude translation (Garbern et al., 1997; Sistermans et al., 1996). These patients have less severe forms of the disease, with the PLP deletions giving rise to either a complicated form of SPG2 or a mild form of PMD (Inoue et al., 2002). In mice lacking PLP, oligodendrocytes develop normally and manage to assemble a myelin sheath, yet defects in the intraperiod line of these sheaths translate into reduced conduction velocities and impaired motor coordination (Boison and StoVel, 1994; Boison et al., 1995; Klugmann et al., 1997; Rosenbluth et al., 1996; Yool et al., 2002, reviewed in Chapter 47 by Nave). In addition, null mutations also produce axonal pathology (discussed later). These pathological changes suggest there is an absolute requirement for PLP both to maintain the structure of compact myelin and to maintain axonal integrity and function. Thus, the absence of PLP would neither trigger the unfolded protein response nor derail myelin assembly, but would instead negatively aVect maintenance of the myelin sheath. PLP Mutations Cause Axonal Damage Evidence of axonal damage has been recently found in both PMD and its animal models, a Wnding that is important for future understanding of the pathogenesis of demyelinating disease and its treatment. In his original description of the neuropathological features of PMD, Merzbacher noted: ‘‘Es stellt sich na¨mlich heraus: da dort, wo die Markscheiden fehlen, auch keine Achsencylinder nachweisbar sind’’ (It is evident that there are no axons demonstrated where the myelin sheaths are absent) (Merzbacher, 1910). Unconvinced that axonal damage existed in PMD, Merzbacher nonetheless concluded that axons were much thinner and did not stain well with axonal stains. DeWnitive evidence for axonal damage in PMD was subsequently provided in several rodent models, including those caused by PLP point mutations (Rosenfeld and Freidrich, 1983), increased PLP gene dosage (Anderson et al., 1998), and PLP null mutation (Garbern et al., 2002; GriYths et al., 1998). Consistent with this interpretation, Garbern and coworkers have found evidence for axonal damage in both mice and patients with a PLP1 null mutation by a combination of direct pathological examination of brain tissue and magnetic resonance spectroscopy (Garbern et al., 2002). The axonal injury is not due to demyelination, since myelin is intact in both patients and experimental animals, or oligodendrocyte cell death, since these cells appear healthy and ensheathe axons. The extent of axonal injury increases with age and probably accounts for the progression of neurological signs and symptoms. In addition, the axonal degeneration is length dependent, suggesting that impaired axonal transport is a cause. These data suggest that progressive axonal damage is not only a common feature of the pathogenesis of PMD, it is also clinically relevant. Because axonal degeneration occurs without signiWcant demyelination, it probably arises from the absence or perturbation of PLP-mediated oligodendrocyteaxonal interactions. Consistent with this notion, Scherer and collaborators have shown that axoglial junctions at the paranodal region are disrupted in md rats (Arroyo et al., 2002), an animal model with a plp point mutation, and that these changes are probably involved in disease pathogenesis. Interestingly, the axonal abnormalities in PMD are very similar to those described by Trapp and coworkers in multiple sclerosis (MS) (Trapp et al., 1998), suggesting that the axonal abnormalities in MS, like those in PMD, may likewise result from disruption of oligodendrocyte-axonal interactions. Axonal degeneration is clinically relevant in MS, since the N-acetyl aspartate/creatine ratio is decreased in the brains of patients with MS, even in regions outside of MS lesions, and correlates well with clinical disability (De Stefano et al., 1997; Fu et al., 1998; Grossman et al., 1992). Also, axonal damage in MS may underlie the secondary progressive phase of the disease, which does not respond signiWcantly to immune modulation. Further understanding of the mechanisms of axonal degeneration in PMD will thus also be important in MS and may lead to the development of new treatment strategies for both diseases.

879

880

37. PELIZAEUS-MERZBACHER DISEASE

Management and Future Prospects Currently, there is no speciWc therapy for patients with PMD. The observation that most patients with PMD have a gene duplication and thus overexpress PLP or have a point mutation causing a gain-of-function precludes simple replacement gene therapy, even if appropriate delivery vehicles were to become available. In fact, for most patients, the more appropriate goal might be the reduction of PLP expression, such as through antisense gene therapy, since absence of PLP results in a less severe syndrome. The Wnding that axonal degeneration is clinically relevant in the pathogenesis of PMD also raises the possibility that therapy directed at maintaining the integrity of axons might be eVective in this disorder. Cellular therapy, such as transplantation of oligodendrocyte precursors into the CNS, has shown potential in animal models of PMD (Brustle et al., 1999; Duncan et al., 1988; Duncan and Milward, 1995; Duncan et al., 1997; Lachapelle et al., 1990) and might therefore be eVective in patients. Cellular therapy has not yet reversed the clinical deWcits in animal models, however, and for maximum eVectiveness, this therapy may need to be initiated either in utero or shortly after birth.

References Almazan, G., Vela, J. M., Molina-Holgado, E., and Guaza, C. (2001). Re-evaluation of nestin as a marker of oligodendrocyte lineage cells. Microsc. Res. Tech. 52(6), 753–765. Anderson, T. J., Montague, P., Nadon, N., Nave, K. A., and GriYths, I. R. (1997). ModiWcation of Schwann cell phenotype with Plp transgenes: Evidence that the PLP and DM20 isoproteins are targeted to diVerent cellular domains. J. Neurosci. Res. 50(1), 13–22. Anderson, T. J., Schneider, A., Barrie, J. A., Klugmann, M., McCulloch, M. C., Kirkham, D., Kyriakides, E., Nave, K.-A., and GriYths, I. R. (1998). Late-Onset Neurodegeneration in Mice with Increased Dosage of the Proteolipid Protein Gene. J. Comp. Neurol. 394, 506–519. Aridor, M., and Balch, W. E. (1999). Integration of endoplasmic reticulum signaling in health and disease. Nat. Med. 5, 745–751. Arroyo, E. J., Xu, T., Grinspan, J., Lambert, S., Levinson, S. R., Brophy, P. J., Peles, E., and Scherer, S. S. (2002). Genetic dysmyelination alters the molecular architecture of the nodal region. J. Neurosci. 22(5), 1726–1737. Bizzozero, O. A., McGarry, J. F., and Lees, M. B. (1987). Autoacylation of myelin proteolipid protein with acyl coenzyme A. Biol Chem. 262(28), 13550–13557. BoespXug-Tanguy, O., Giraud, G., Mimault, C., Isabelle, V., and Dinh, D. P. (1999). Heterogeneous rearrangements of the PLP genomic region in Pelizaeus-Merzbacher Disease: Genotype-phenotype correlation in 41 patients. Am. J. Hum. Genet. 65, Program Nr: 1597. Boison, D., Bussow, H., D’Urso, D., Muller, H. W., and StoVel, W. (1995). Adhesive properties of proteolipid protein are responsible for the compaction of CNS myelin sheaths. J. Neurosci. 15(8), 5502–5513. Boison, D., and StoVel, W. (1994). Disruption of the compacted myelin sheath of axons of the central nervous system in proteolipid protein-deWcient mice. Proc. Natl. Acad. Sci. U. S. A. 91(24), 11709–11713. Bond, C., Si, X., Crisp, M., Wong, P., Paulson, G. W., Boesel, C. P., Dlouhy, S. R., and Hodes, M. E. (1997). Family with Pelizaeus-Merzbacher disease/X-linked spastic paraplegia and a nonsense mutation in exon 6 of the proteolipid protein gene. Am. J. Med. Genet. 71, 357–360. Bongarzone, E. R., Campagnoni, C. W., Kampf, K., Jacobs, E. C., Handley, V. W., Schonmann, V., and Campagnoni, A. T. (1999). IdentiWcation of a New Exon in the Myelin Proteolipid Protein Gene Encoding Novel Protein Isoforms That Are Restricted to the Somata of Oligodendrocytes and Neurons. J. Neurosci. 19(19), 8349–8357. Bonneau, D., Rozet, J. M., Bulteau, C., Berthier, M., Mettey, R., Gil, R., Munnich, A., and Le Merrer, M. (1993). X linked spastic paraplegia (SPG2): Clinical heterogeneity at a single gene locus. J. Med. Genet. 30(5), 381–384. Boulloche, J., and Aicardi, J. (1986). Pelizaeus-Merzbacher disease: Clinical and nosological study. J. Child Neurol. 1(3), 233–239. Brustle, O., Jones, K. N., Learish, R. D., Karram, K., Choudhary, K., Wiestler, O. D., Duncan, I. D., and McKay, R. D. (1999). Embryonic stem cell-derived glial precursors: A source of myelinating transplants Science 285(5428), 754–756. Cailloux, F., Gauthier-Barichard, F., Mimault, C., Isabelle, V., Courtois, V., Giraud, G., Dastugue, B., and BoespXug-Tanguy, O. (2000). Genotype-phenotype correlation in inherited brain myelination defects due to proteolipid protein gene mutations Eur. J. Hum. Genet. 8(11), 837–845. Cambi, F., Tang, X. M., Cordray, P., Fain, P. R., Keppen, L. D., and Barker, D. F. (1996). ReWned genetic mapping and proteolipid protein mutation analysis in X-linked pure hereditary spastic paraplegia. Neurology 46(4), 1112–1117. Cambi, F., Tartaglino, L., Lublin, F., and McCarren, D. (1995). X-Linked Pure Familial Spastic Paraparesis: Characterization Of a Large Kindred With Magnetic-Resonance-Imaging Studies. Arch. Neurol. 52(7), 665–669.

MOLECULAR PATHOGENESIS OF PMD

Campagnoni, C. W., Garbay, B., Micevych, P., Pribyl, T., Kampf, K., Handley, V. W., and Campagnoni, A. T. (1992). DM20 mRNA splice product of the myelin proteolipid protein gene is expressed in the murine heart. J. Neurosci. Res. 33(1), 148–155. Carrozzo, R., Arrigo, G., Rossi, E., Bardoni, B., Cammarata, M., Gandullia, P., Gatti, R., and ZuVardi, O. (1997). Multiple congenital anomalies, brain hypomyelination, and ocular albinism in a female with dup(X) (pter!q24::q21.32!qter) and random X inactivation. Am. J. Med. Genet. 72(3), 329–334. Cremers, F. P., PfeiVer, R. A., van de Pol, T. J., Hofker, M. H., Kruse, T. A., Wieringa, B., and Ropers, H. H. (1987). An interstitial duplication of the X chromosome in a male allows physical Wne mapping of probes from the Xq13-q22 region. Hum. Genet. 77(1), 23–27. Cuddon, P. A., Lipsitz, D., and Duncan, I. D. (1998). Myelin mosaicism and brain plasticity in heterozygous females of a canine X-linked trait Ann. Neurol. 44(5), 771–779. De Stefano, N., Matthews, P. M., Narayanan, S., Francis, G. S., Antel, J. P., and Arnold, D. L. (1997). Axonal dysfunction and disability in a relapse of multiple sclerosis: Longitudinal study of a patient. Neurology 49(4), 1138–1141. Dickinson, P. J., GriYths, I. R., Barrie, J. M., Kyriakides, E., Pollock, G. F., and Barnett, S. C. (1997). Expression of the DM-20 isoform of the PLP gene in olfactory nerve ensheathing cells: Evidence from developmental studies. J. Neurocytol. 26(3), 181–189. Diehl, H. J., Schaich, M., Budzinski, R. M., and StoVel, W. (1986). Individual exons encode the integral membrane domains of human myelin proteolipid protein [published erratum appears in Proc Natl Acad Sci U S A 1991 Apr;86(6), 617–8]. Proc. Natl. Acad. Sci. U. S. A. 83(24), 9807–9811. Duncan, I. D., Grever, W. E., and Zhang, S. C. (1997). Repair of myelin disease: Strategies and progress in animal models. Mol. Med. Today 3(12), 554–561. Duncan, I. D., Hammang, J. P., Jackson, K. F., Wood, P. M., Bunge, R. P., and Langford, L. (1988). Transplantation of oligodendrocytes and Schwann cells into the spinal cord of the myelin-deWcient rat. J. Neurocytol. 17(3), 351–360. Duncan, I. D., and Milward, E. A. (1995). Glial cell transplants: Experimental therapies of myelin diseases.. Brain Pathol. 5(3), 301–310. Ellis, D., and Malcolm, S. (1994). Proteolipid protein gene dosage eVect in Pelizaeus-Merzbacher disease Nat. Genet. 6(4), 333–334. Fanarraga, M. L., GriYths, I. R., McCulloch, M. C., Barrie, J. A., Cattanach, B. M., Brophy, P. J., and Kennedy, P. G. (1991). Rumpshaker: An X-linked mutation aVecting CNS myelination. A study of the female heterozygote. Neuropathol. Appl. Neuro. Biol. 17(4), 323–334. Fu, H., Qi, Y., Tan, M., Cai, J., Takebayashi, H., Nakafuku, M., Richardson, W., and Qiu, M. (2002). Dual origin of spinal oligodendrocyte progenitors and evidence for the cooperative role of Olig2 and Nkx2.2 in the control of oligodendrocyte diVerentiation. Development 129(3), 681–693. Fu, L., Matthews, P. M., De Stefano, N., Worsley, K. J., Narayanan, S., Francis, G. S., Antel, J. P., Wolfson, C., and Arnold, D. L. (1998). Imaging axonal damage of normal-appearing white matter in multiple sclerosis. Brain 121(Pt 1), 103–113. Garbern, J. Y., Cambi, F., Lewis, R., Shy, M., Sima, A., Kraft, G., Vallat, J. M., Bosch, E. P., Hodes, M. E., Dlouhy, S., Raskind, W., Bird, T., Macklin, W., and Kamholz, J. (1999). Peripheral neuropathy caused by proteolipid protein gene mutations. Ann. N.Y. Acad. Sci 883, 351–365. Garbern, J. Y., Cambi, F., Tang, X. M., Sima, A. A., Vallat, J. M., Bosch, E. P., Lewis, R., Shy, M., Sohi, J., Kraft, G., Chen, K. L., Joshi, I., Leonard, D. G., Johnson, W., Raskind, W., Dlouhy, S. R., Pratt, V., Hodes, M. E., Bird, T., and Kamholz, J. (1997). Proteolipid protein is necessary in peripheral as well as central myelin. Neuron 19(1), 205–218. Garbern, J. Y., Yool, D. A., Moore, G. J., Wilds, I. B., Faulk, M. W., Klugmann, M., Nave, K. A., Sistermans, E. A., van der Knaap, M. S., Bird, T. D., Shy, M. E., Kamholz, J. A., and GriYths, I. R. (2002). Patients lacking the major CNS myelin protein, proteolipid protein 1, develop length-dependent axonal degeneration in the absence of demyelination and inXammation. Brain 125(Pt 3), 551–561. Gencic, S., Abuelo, D., Ambler, M., and Hudson, L. D. (1989). Pelizaeus-Merzbacher disease: An X-linked neurologic disorder of myelin metabolism with a novel mutation in the gene encoding proteolipid protein. Am.. J. Hum. Genet. 45(3), 435–442. Gow, A., Gragerov, A., Gard, A., Colman, D. R., and Lazzarini, R. A. (1997). Conservation of topology, but not conformation, of the proteolipid proteins of the myelin sheath. J. Neurosci. 17(1), 181–189. Gow, A., and Lazzarini, R. A. (1996). A cellular mechanism governing the severity of Pelizaeus-Merzbacher Disease. Nat. Genet. 13(4), 422–428. Gow, A., Southwood, C. M., and Lazzarini, R. A. (1998). Disrupted proteolipid protein traYcking results in oligodendrocyte apoptosis in an animal model of Pelizaeus-Merzbacher disease. J. Cell Biol. 140(4), 925–934. GriYths, I., Klugmann, M., Anderson, T., Yool, D., Thomson, C., Schwab, M. H., Schneider, A., Zimmermann, F., McCulloch, M., Nadon, N., and Nave, K.-A. (1998). Axonal swellings and degeneration in mice lacking the major proteolipid of myelin. Science 280, 1610–1613. GriYths, I. R., Dickinson, P., and Montague, P. (1995). Expression of the proteolipid protein gene in glial cells of the post-natal peripheral nervous system of rodents. Neuropathol. Appl. Neurobiol. 21(2), 97–110. GriYths, I. R., Mitchell, L. S., McPhilemy, K., Morrison, S., Kyriakides, E., and Barrie, J. A. (1989). Expression of myelin protein genes in Schwann cells. J. Neurocytol. 18(3), 345–352. Grossman, R. I., Lenkinski, R. E., Ramer, K. N., Gonzalez-Scarano, F., and Cohen, J. A. (1992). MR proton spectroscopy in multiple sclerosis. AJNR Am. J. Neuroradiol. 13(6), 1535–1543.

881

882

37. PELIZAEUS-MERZBACHER DISEASE

Gudz, T. I., Schneider, T. E., Haas, T. A., and Macklin, W. B. (2002). Myelin proteolipid protein forms a complex with integrins and may participate in integrin receptor signaling in oligodendrocytes. J. Neurosci 22(17), 7398–7407. Harding, B., Ellis, D., and Malcolm, S. (1995). A case of Pelizaeus-Merzbacher disease showing increased dosage of the proteolipid protein gene. Neuropathol. Appl. Neurobiol. 21(2), 111–115. Helynck, G., Luu, B., Nussbaum, J. L., Picken, D., Skalidis, G., TriWlieV, E., Van Dorsselaer, A., Seta, P., Sandeaux, R., Gavach, C., Heitz, F., Simon, D., and Spach, G. (1983). Brain proteolipids. Isolation, puriWcation and eVect on ionic permeability of membranes. Eur. J. Biochem. 133(3), 689–695. Hobson, G. M., Davis, A. P., Stowell, N. C., Kolodny, E. H., Sistermans, E. A., de Coo, I. F. M., Funanage, V. L., and Marks, H. G. (2000). Mutations in noncoding regions of the proteolipid protein gene in PelizaeusMerzbacher disease. Neurology 55, 1089–1096. Hodes, M. E., Blank, C. A., Pratt, V. M., Morales, J., Napier, J., and Dlouhy, S. R. (1997). Nonsense mutation in exon 3 of the proteolipid protein gene (PLP) in a family with an unusual form of Pelizaeus-Merzbacher disease. Am. J. Med. Genet. 69(2), 121–125. Hodes, M. E., DeMyer, W. E., Pratt, V. M., Edwards, M. K., and Dlouhy, S. R. (1995). Girl with signs of Pelizaeus-Merzbacher disease heterozygous for a mutation in exon 2 of the proteolipid protein gene. Am. J. Med. Genet. 55(4), 397–401. Hodes, M. E., Pratt, V. M., and Dlouhy, S. R. (1993). Genetics of Pelizaeus-Merzbacher disease. Dev. Neurosci. 15(6), 383–394. Hodes, M. E., Woodward, K., Spinner, N. B., Emanuel, B. S., Enrico-Simon, A., Kamholz, J., Stambolian, D., Zackai, E. H., Pratt, V. M., Thomas, I. T., Crandall, K., Dlouhy, S. R., and Malcolm, S. (2000). Additional copies of the proteolipid protein gene causing Pelizaeus-Merzbacher disease arise by separate integration into the X chromosome. Am. J. Hum. Genet. 67(1), 14–22. Hodes, M. E., Zimmerman, A. W., Aydanian, A., Naidu, S., Miller, N. R., Oller, J. L. G., Barker, B., Aleck, K. A., Hurley, T. D., and Dlouhy, S. R. (1999). DiVerent mutations in the same codon of the proteolipid protein gene, PLP, may help in correlating genotype with phenotype in Pelizaeus-Merzbacher disease/X-linked spastic paraplegia (PMD/SPG2). Am. J. Med. Genet. 82(2), 132–139. Hudson, L. D., Puckett, C., Berndt, J., Chan, J., and Gencic, S. (1989). Mutation of the proteolipid protein gene PLP in a human X chromosome-linked myelin disorder. Proc. Natl. Acad. Sci. USA 86(20), 8128–8131. Ikenaka, K., Furuichi, T., Iwasaki, Y., Moriguchi, A., Okano, H., and Mikoshiba, K. (1988). Myelin proteolipid protein gene structure and its regulation of expression in normal and jimpy mutant mice. J. Mol. Biol. 199(4), 587–596. Ikenaka, K., Kagawa, T., and Mikoshiba, K. (1992). Selective expression of DM-20, an alternatively spliced myelin proteolipid protein gene product, in developing nervous system and in nonglial cells. J. Neurochem. 58(6), 2248–2253. Inoue, K., Kanai, M., Tanabe, Y., Kubota, T., Kashork, C. D., Wakui, K., Fukushima, Y., Lupski, J. R., and ShaVer, L. G. (2001a). Prenatal interphase FISH diagnosis of PLP1 duplication associated with PelizaeusMerzbacher disease. Prenat. Diagn. 21(13), 1133–1136. Inoue, K., and Lupski, J. R. (2002). Molecular mechanisms for genomic disorders. Annu. Rev. Genomics Hum. Genet. 3, 199–242. Inoue, K., Osaka, H., Imaizumi, K., Nezu, A., Takanashi, J., Arii, J., Murayama, K., Ono, J., Kikawa, Y., Mito, T., ShaVer, L. G., and Lupski, J. R. (1999a). Proteolipid protein gene duplications causing Pelizaeus-Merzbacher disease: Molecular mechanism and phenotypic manifestations. Ann. Neurol. 45(5), 624–632. Inoue, K., Osaka, H., Sugiyama, N., Kawanishi, C., Onishi, H., Nezu, A., Kimura, K., Kimura, S., Yamada, Y., and Kosaka, K. (1996). A duplicated PLP gene causing Pelizaeus-Merzbacher disease detected by comparative multiplex PCR. Am. J. Hum. Genet. 59, 32–39. Inoue, K., Osaka, H., Thurston, V. C., Clarke, J. T., Yoneyama, A., Rosenbarker, L., Bird, T. D., Hodes, M. E., ShaVer, L. G., and Lupski, J. R. (2002). Genomic Rearrangements Resulting in PLP1 Deletion Occur by Nonhomologous End Joining and Cause DiVerent Dysmyelinating Phenotypes in Males and Females. Am. J. Hum. Genet. 71(4), 838–853. Inoue, K., Tanabe, Y., and Lupski, J. R. (1999b). Myelin deWciencies in both the central and the peripheral nervous systems associated with a SOX10 mutation. Ann. Neurol. 46(3), 313–318. Inoue, K., Tanaka, H., Scaglia, F., Araki, A., ShaVer, L. G., and Lupski, J. R. (2001b). Compensating for central nervous system dysmyelination: Females with a proteolipid protein gene duplication and sustained clinical improvement. Ann. Neurol. 50(6), 747–754. Johnston, A. W., and McKusick, V. A. (1962). A sex-linked recessive form of spastic paraplegia. Am. J. Hum. Genet. 14, 83–94. Johnston, J. A., Ward, C. L., and Kopito, R. R. (1998). Aggresomes: A cellular response to misfolded proteins. J. Cell Biol. 143(7), 1883–1898. Kaufman, R. J., Scheuner, D., Schroder, M., Shen, X., Lee, K., Liu, C. Y., and Arnold, S. M. (2002). The unfolded protein response in nutrient sensing and diVerentiation. Nat. Rev. Mol. Cell Biol. 3(6), 411–421. Kawanishi, C., Sugiyama, N., Osaka, H., Inoue, K., Suzuki, K., Onishi, H., Yamada, Y., Nezu, A., Kimura, S., and Kosaka, K. (1996). Pelizaeus-Merzbacher disease: A novel mutation in the 5’-untranslated region of the proteolipid protein gene. Hum. Mutat. 7(4), 355–357. Kaye, E. M., Doll, R. F., Natowicz, M. R., and Smith, F. I. (1994). Pelizaeus-Merzbacher disease presenting as spinal muscular atrophy: Clinical and molecular studies. Ann. Neurol. 36(6), 916–919.

MOLECULAR PATHOGENESIS OF PMD

Klugmann, M., Schwab, M. H., Pu¨hlhofer, A., Schneider, A., Zimmermann, F., GriYths, I. R., and Nave, K.-A. (1997). Assembly of CNS myelin in the absence of proteolipid protein. Neuron 18(1), 59–70. Kobayashi, H., HoVman, E. P., and Marks, H. G. (1994). The rumpshaker mutation in spastic paraplegia Nat. Genet. 7(3), 351–352. Kopito, R. R., and Ron, D. (2000). Conformational disease. Nat. Cell Biol. 2(11), E207–209. Lachapelle, F., Lapie, P., Gansmuller, A., Villarroya, H., Baumann, N., and Gumpel, M. (1990). What have we learned about the jimpy phenotype expression by intracerebral transplantations? Ann. N.Y. Acad. Sci. 605, 332–345. Macklin, W. B., Campagnoni, C. W., Deininger, P. L., and Gardinier, M. V. (1987). Structure and expression of the mouse myelin proteolipid protein gene. J. Neurosci. Res. 18(3), 383–394. Maenpaa, J., Lindahl, E., Aula, P., and Savontaus, M. L. (1990). Prenatal diagnosis in Pelizaeus-Merzbacher disease using RFLP analysis. Clin. Genet. 37(2), 141–147. McLaughlin, M., Hunter, D. J., Thomson, C. E., Yool, D., Kirkham, D., Freer, A. A., and GriYths, I. R. (2002). Evidence for possible interactions between PLP and DM20 within the myelin sheath. Glia 39(1), 31–36. Merzbacher, L. (1910). Eine eigenarige familia¨r-hereditare Erkrankungsform (Aplasia axialis extra-corticalis congenita). Z ges. Neurol. Psych. 3, 1–138. Mimault, C., Giraud, G., Courtois, V., Cailloux, F., Boire, J. Y., Dastugue, B., and BoespXug-Tanguy, O. (1999). Proteolipoprotein gene analysis in 82 patients with sporadic Pelizaeus-Merzbacher Disease: Duplications, the major cause of the disease, originate more frequently in male germ cells, but point mutations do not. The Clinical European Network on Brain Dysmyelinating Disease. Am. J. Hum. Genet. 65(2), 360–369. Nave, K.-A., and BoespXug-Tanguy, O. (1996). X-linked developmental defects of myelination: From mouse mutants to human genetic diseases. Neuroscientist 2(1), 33–43. Nezu, A., Kimura, S., Takeshita, S., Osaka, H., Kimura, K., and Inoue, K. (1998). An MRI and MRS study of Pelizaeus-Merzbacher disease. Pediatr. Neurol. 18(4), 334–337. Ono, J., Harada, K., Sakurai, K., Kodaka, R., Shimidzu, N., Tanaka, J., Nagai, T., and Okada, S. (1994). MR diVusion imaging in Pelizaeus-Merzbacher disease. Brain & Development 16(3), 219–223. Osaka, H., Kawanishi, C., Inoue, K., Onishi, H., Kobayashi, T., Sugiyama, N., Kosaka, K., Nezu, A., Fujii, K., Sugita, K., Kodama, K., Murayama, K., Murayama, S., Kanazawa, I., and Kimura, S. (1999). PelizaeusMerzbacher disease: Three novel mutations and implication for locus heterogeneity. Ann. Neurol. 45(1), 59–64. Osaka, H., Kawanishi, C., Inoue, K., Uesugi, H., Hiroshi, K., Nishiyama, K., Yamada, Y., Suzuki, K., Kimura, S., and Kosaka, K. (1995). Novel nonsense proteolipid protein gene mutation as a cause of X-linked spastic paraplegia in twin males. Biochem. Biophys. Res. Commun. 215(3), 835–841. ¨ ber eine eigenthu¨mliche Form Spastischer La¨hmung mit Cerebralerschinungen auf heredPelizaeus, F. (1885). U ita¨rer Grundlage (Multiple Sklerose). Arch. Psychiatr. Nervenkr. 16, 698–710. Pham-Dinh, D., Birling, M. C., Roussel, G., Dautigny, A., and Nussbaum, J. L. (1991). Proteolipid DM-20 predominates over PLP in peripheral nervous system. Neuroreport 2(2), 89–92. Pingault, V., Girard, M., Bondurand, N., Dorkins, H., Van Maldergem, L., Mowat, D., Shimotake, T., Verma, I., Baumann, C., and Goossens, M. (2002). SOX10 mutations in chronic intestinal pseudo-obstruction suggest a complex physiopathological mechanism. Hum. Genet. 111(2), 198–206. Pribyl, T. M., Campagnoni, C., Kampf, K., Handley, V. W., and Campagnoni, A. T. (1996). The major myelin protein genes are expressed in the human thymus. J. Neurosci. Res. 45(6), 812–819. Puckett, C., Hudson, L., Ono, K., Friedrich, V., Benecke, J., Dubois-Dalcq, M., and Lazzarini, R. A. (1987). Myelin-speciWc proteolipid protein is expressed in myelinating Schwann cells but is not incorporated into myelin sheaths. J. Neurosci. Res. 18(4), 511–518. Qi, Y., Cai, J., Wu, Y., Wu, R., Lee, J., Fu, H., Rao, M., Sussel, L., Rubenstein, J., and Qiu, M. (2001). Control of oligodendrocyte diVerentiation by the Nkx2.2 homeodomain transcription factor. Development 128(14), 2723–2733. Raskind, W. H., Williams, C. A., Hudson, L. D., and Bird, T. D. (1991). Complete deletion of the proteolipid protein gene (PLP) in a family with X-linked Pelizaeus-Merzbacher disease. Am. J. Hum. Genet. 49(6), 1355–1360. Regis, S., Filocamo, M., Mazzotti, R., Cusano, R., Corsolini, F., Bonuccelli, G., Stroppiano, M., and Gatti, R. (2001). Prenatal diagnosis of Pelizaeus-Merzbacher disease: Detection of proteolipid protein gene duplication by quantitative Xuorescent multiplex PCR. Prenat. Diagn. 21(8), 668–671. Rosenbluth, J., StoVel, W., and SchiV, R. (1996). Myelin structure in proteolipid protein (PLP)-null mouse spinal cord. J. Comp. Neurol. 371(2), 336–344. Rosenfeld, J., and Freidrich, V. L., Jr. (1983). Axonal swellings in jimpy mice:Does lack of myelin cause neuronal abnormalities? Neuroscience 10(3), 959–966. Ross, N. W., and Braun, P. E. (1988). Acylation in vitro of the myelin proteolipid protein and comparison with acylation in vivo: Acylation of a cysteine occurs nonenzymatically. Neurosci Res. 21(1), 35–44. Sambuughin, N., Sivakumar, K., Selenge, B., Baasanjav, D., and Goldfarb, L. G. (1998). New mutation in exon 3B of proteolipid protein gene in Mongolian family with a benign variant of X linked spastic paraplegia. Am. J. Hum. Genet. 63(4), A383. Saugier-Veber, P., Munnich, A., Bonneau, D., Rozet, J. M., Le Merrer, M., Gil, R., and BoespXug-Tanguy, O. (1994). X-linked spastic paraplegia and Pelizaeus-Merzbacher disease are allelic disorders at the proteolipid protein locus. Nat. Genet. 6(3), 257–262. Schindler, P., Luu, B., Sorokine, O., TriWlieV, E., and Van Dorsselaer, A. (1990). Developmental study of proteolipids in bovine brain: A novel proteolipid and DM-20 appear before proteolipid protein (PLP) during myelination. J. Neurochem. 55(6), 2079–2085.

883

884

37. PELIZAEUS-MERZBACHER DISEASE

Seitelberger, F. (1954). Die Pelizaeus-Merzbachersche Krankheit. Klinischanatomische Untersuchungen zum Problem ihrer Stellung unter den diVusen Sklerosen. Wien Z Nervenheilk 9, 228–289. Seitelberger, F. (1970). Pelizaeus-Merzbacher disease. In ‘‘Handbook of Clinical Neurology’’ (P. J. Vinken and G. W. Bruyn, G., eds.), 10, 150–220. Amsterdam, North Holland Publishing. ShaVer, L. G., Kennedy, G. M., Spikes, A. S., and Lupski, J. R. (1997). Diagnosis of CMT1A duplications and HNPP deletions by interphase FISH: Implications for testing in the cytogenetics laboratory. Am. J. Med. Genet. 69(3), 325–331. Simons, M., Kramer, E. M., Macchi, P., Rathke-Hartlieb, S., Trotter, J., Nave, K. A., and Schulz, J. B. (2002). Overexpression of the myelin proteolipid protein leads to accumulation of cholesterol and proteolipid protein in endosomes/lysosomes: Implications for Pelizaeus-Merzbacher disease. J. Cell Biol. 157(2), 327–336. Simons, M., Kramer, E. M., Thiele, C., StoVel, W., and Trotter, J. (2000). Assembly of Myelin by Association of Proteolipid Protein with Cholesterol- and Galactosylceramide-rich Membrane Domains. J. Cell Biol. 151(1), 143–154. Sistermans, E. A., de Coo, R. F., de Wijs, I. J., and van Oost, B. A. (1998). Duplication of the proteolipid protein gene is the major cause of Pelizaeus-Merzbacher disease. Neurology 50(6), 1749–1754. Sistermans, E. A., de Wijs, I. J., de Coo, R. F. M., Smit, L. M. E., Menko, F. H., and van Oost, B. A. (1996). A (G-to-A) mutation in the initiation codon of the proteolipid protein gene causing a relatively mild form of Pelizaeus-Merzbacher disease in a Dutch family. Hum. Genet. 97(3), 337–339. Sivakumar, K., Sambuughin, N., Selenge, B., Nagle, J. W., Baasanjav, D., Hudson, L. D., and Goldfarb, L. G. (1999). Novel exon 3B proteolipid protein gene mutation causing late-onset spastic paraplegia type 2 with variable penetrance in female family members. Ann. Neurol. 45(5), 680–683. Skalidis, G., TriWlieV, E., and Luu, B. (1986). Selective extraction of the DM-20 brain proteolipid. J. Neurochem. 46(1), 297–299. Southwood, C., and Gow, A. (2001). Molecular pathways of oligodendrocyte apoptosis revealed by mutations in the proteolipid protein gene. Microsc. Res. Tech. 52(6), 700–708. Southwood, C. M., Garbern, J., Jiang, W., Gow A. (2002). The unfolded protein response modulates diseases severity in Pelizaeus-Merzbacher disease. Neuron. 36(4), 585–96. Sporkel, O., Uschkureit, T., Bussow, H., and StoVel, W. (2002). Oligodendrocytes expressing exclusively the DM20 isoform of the proteolipid protein gene: Myelination and development. Glia 37(1), 19–30. Stecca, B., Southwood, C. M., Gragerov, A., Kelley, K. A., Friedrich, V. L., and Gow, A. (2000). The evolution of lipophilin genes from invertebrates to tetrapods: DM-20 cannot replace proteolipid protein in CNS myelin. J. Neurosci. 20(11), 4002–4010. StoVel, W., Hillen, H., and Giersiefen, H. (1984). Structure and molecular arrangement of proteolipid protein of central nervous system myelin. Proc. Natl. Acad. Sci. U. S. A. 81(16), 5012–5016. Stolt, C. C., Rehberg, S., Ader, M., Lommes, P., Riethmacher, D., Schachner, M., Bartsch, U., and Wegner, M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev. 16(2), 165–170. Strautnieks, S., Rutland, P., Winter, R. M., Baraitser, M., and Malcolm, S. (1992). Pelizaeus-Merzbacher disease: Detection of mutations Thr181—Pro and Leu223—Pro in the proteolipid protein gene, and prenatal diagnosis. Am. J. Hum. Genet. 51(4), 871–878. Taylor, J. P., Hardy, J., and Fischbeck, K. H. (2002). Toxic proteins in neurodegenerative disease. Science 296(5575), 1991–1995. Timsit, S., Martinez, S., Allinquant, B., Peyron, F., Puelles, L., and Zalc, B. (1995). Oligodendrocytes originate in a restricted zone of the embryonic ventral neural tube deWned by DM-20 mRNA expression. J. Neurosci. 15(2), 1012–1024. Timsit, S. G., Bally-Cuif, L., Colman, D. R., and Zalc, B. (1992). DM-20 mRNA is expressed during the embryonic development of the nervous system of the mouse. J. Neurochem. 58(3), 1172–1175. Trapp, B. D., Peterson, J., RansohoV, R. M., Rudick, R., Mork, S., and Bo, L. (1998). Axonal transection in the lesions of multiple sclerosis N. Engl. J. Med. 338(5), 278–285. Trofatter, J. A., Dlouhy, S. R., DeMyer, W., Conneally, P. M., and Hodes, M. E. (1989). Pelizaeus-Merzbacher disease: Tight linkage to proteolipid protein gene exon variant. Proc. Natl. Acad. Sci. USA 86(23), 9427–9430. Wahle, S., and StoVel, W. (1998). Cotranslational integration of myelin proteolipid protein (PLP) into the membrane of endoplasmic reticulum: Analysis of topology by glycosylation scanning and protease domain protection assay. Glia 24(2), 226–235. Wang, P. J., Hwu, W. L., Lee, W. T., Wang, T. R., and Shen, Y. Z. (1997). Duplication of proteolipid protein gene: A possible major cause of Pelizaeus-Merzbacher disease. Pediatr. Neurol. 17(2), 125–128. Weimbs, T., and StoVel, W. (1992). Proteolipid protein (PLP) of CNS myelin: Positions of free, disulWde-bonded, and fatty acid thioester-linked cysteine residues and implications for the membrane topology of PLP. Biochemistry 31(49), 12289–12296. Woodward, K., Kendall, E., Vetrie, D., and Malcolm, S. (1998). Pelizaeus-Merzbacher disease: IdentiWcation of Xq22 proteolipid-protein duplications and characterization of breakpoints by interphase FISH. Am. J. Hum. Genet. 63(1), 207–217. Yool, D., Klugmann, M., Barrie, J. A., McCulloch, M. C., Nave, K. A., and GriYths, I. R. (2002). Observations on the structure of myelin lacking the major proteolipid protein. Neuropathol. Appl. NeuroBiol. 28(1), 75–78. Yu, W. P., Collarini, E. J., Pringle, N. P., and Richardson, W. D. (1994). Embryonic expression of myelin genes: Evidence for a focal source of oligodendrocyte precursors in the ventricular zone of the neural tube. Neuron 12(6), 1353–1362.

MOLECULAR PATHOGENESIS OF PMD

Zackai, E. H., Stambolian, D., Enrico, A., McDonald-McGinn, D. M., Kamholz, J., Emanuel, B. S., and Spinner, N. B. (1997). Familial Pelizaeus Merzbacher disease with a pericentric inversion of the X chromosome [inv(X) (p11.4q22.1)] resulting in PLP gene duplication. Am. J. Hum. Genet. 61, A144. Zeman, W., DeMyer, W., and Falls, H. F. (1964). Pelizaeus-Merzbacher disease: A study in nosology. J. Neuropath. Exp. Neurol. 23, 334–354. Zhang, S. C., and Duncan, I. D.(2000). Remyelination and restoration of axonal function by glial cell transplantation. Prog. Brain Res. 127, 515–33.

885

C H A P T E R

38 Guillain-Barre Syndrome John W. Griffin and Kazim Sheikh

INTRODUCTION With the near eradication of polio by vaccination, Guillain-Barre syndrome (GBS) has emerged as the most frequent cause of acute Xaccid paralysis worldwide. Its most frequent form, acute inXammatory demyelinating polyneuropathy (AIDP), is the prototypic acquired demyelinating disease of the peripheral nervous system. The importance of GBS in this text lies both in its own prominence as a major cause of neurologic morbidity and in the similarities and contrasts with acquired demyelinating disorders of the central nervous system.

HISTORY The recognition that the peripheral nervous system could be the site of involvement for paralytic diseases of the nervous system came surprisingly late. It may seem intuitive that PNS disease could lead to sensory changes and paralysis, but the publications of the work of Sir Charles Bell and Francois Magendie on the diVerent roles of the ventral and dorsal roots emerged in the early 1820s, the deWnitive experiments of Johannes Muller on this topic were not published until 1831, and the consequences of nerve section awaited Augustus Waller’s description in 1852 (Sanders, 1948). In any event, Robert Graves, in 1843, reviewed an epidemic outbreak of a painful paralytic disorder that occurred in Paris in 1828 and concluded that, because of the lack of documented involvement of the CNS, the disorder most likely occurred within the peripheral nervous system (Graves, 1884). The Wrst clear description of Guillain Barre syndrome was made by the marvelously intuitive French physician Jean Baptiste Octave Landry de Thezillat in 1859 (Landry, 1859). Landry described Wve of his own cases and Wve others from the literature, including a detailed description of a woman, initially considered hysterical by her physician, who died of respiratory insuYciency. He emphasized the ascending sequence of involvement and the rapid progression. Paralysis developed within 15 days in his cases and could evolve over 2 to 3 days or ‘‘occasionally only a few hours.’’ Of his initial 10 cases, two died of respiratory insuYciency and others had well-documented bulbar paralysis. In the late 1890s, the degeneration of Wbers in the peripheral nervous system was clear from occasional pathologic studies of acute Xaccid paralysis, and one report had identiWed inXammatory cells around vessels and in the endoneurial space, although this feature generated little comment (Eichhorst, 1877). In 1892, Sir William Osler, the professor of medicine at Johns Hopkins Hospital, described six cases of ‘‘acute febrile polyneuritis’’ (Osler, 1892). These cases clinically correspond well with what we now recognize as the Guillain-Barre syndrome, although they had fever at the peak of their paralysis, a Wnding

Myelin Biology and Disorders, Volume 2

887

Copyright 2004, Elsevier (USA). All rights reserved.

888

38. GUILLAIN-BARRE SYNDROME

that remains unexplained. In 1893, Bury and Ross published an important and little known review of peripheral nerve diseases and documented more than 60 cases of acute polyneuropathies to that time (Ross and Bury, 1893). The diVerential diagnosis of acute Xaccid paralysis 100 years ago was much longer than it is at present and included infectious and toxic disorders rarely encountered today. While it is intriguing to speculate on the diagnosis of each individual summarized by Bury and Ross, in the aggregate most of them are certainly Guillain-Barre syndrome. Guillain-Barre syndrome was deWnitively described in 1916 by two French neurologists who were friends and colleagues, Georges Guillain and Jean-Alexander Barre, along with their little recognized coauthor, Andre´ Strohl (Guillain, Barre, et al., 1916). Their report detailed the stories of two French soldiers in World War I who developed acute Xaccid paresis. An early symptom in one was inability to stand wearing his pack and to rise after falling. Ironically, also in 1916, the Wrst cases of epidemic polio were described from New York City, perhaps in part accounting for the fact that Guillain-Barre syndrome has been occasionally referred to as ‘‘French polio.’’ The report by Guillain, Barre, and Strohl was a model of clinical lucidity, pointing out the depressed tendon reXexes, the rapid recovery, and the laboratory Wnding of elevated protein throughout cells in the spinal Xuid. This Wnal feature utilized the technique—new at that time—of lumbar puncture for diagnostic purposes and set ‘‘their’’ syndrome apart from poliomyelitis. By 1948, Kernohan and Haymaker had studied 50 cases pathologically and recognized demyelination could occur in GBS (Haymaker and Kernohan, 1949). Parenthetically, the length of time required to identify the prominence of demyelination in GBS reXected the diYculty in ‘‘seeing’’ demyelination of individual Wbers using standard paraYn sections. The best technique is teasing of nerve Wbers, in which individual Wbers can be followed longitudinally over multiple internodes. Teasing was utilized by Gombault in his experimental studies, in the 1880s, but was little used in subsequent pathologic studies until the 1960s. In 1969, Asbury, Arnason, and Adams wrote their classic pathologic review of 19 Boston cases and pointed out the prominence as well as the variability of lymphocytic inXammation in these cases (Asbury, Arnason, et al., 1969). Prineas in 1976 produced his classic electron micrographs of macrophage-mediated demyelination and myelin stripping (Prineas, 1981). Perhaps because of the similarities of the images in these last two reports with those of experimental allergic neuritis induced by immunization with myelin, GBS was frequently equated with EAN, and pathologic analysis of GBS largely languished for the next decade. In 1986, Feasby and colleagues reported a patient with acute paralysis in whom physiologic and pathologic data suggested axonal degeneration without demyelination (Feasby, Gilbert, et al., 1986; Feasby, Hahn, et al., 1993). The possibility of ‘‘axonal GBS’’ was subsequently conWrmed and extended by a series of studies from Northern China, conducted as collaborative studies involving Second Teaching Hospital in Shijiazhuang, Hebei Province, the Beijing Children’s Hospital, the University of Pennsylvania, and the Johns Hopkins University (GriYn, Li, et al., 1995; GriYn, Li, et al., 1996; Siebert and Larrick, 1992). The investigators identiWed epidemic paralytic disease that occurred annually in the summer and had many features that were identical to AIDP in the West. However, most of the summertime cases diVered from AIDP in the prominence of physiologic and pathologic features of axonal degeneration. Initially termed the ‘‘Chinese Paralytic Syndrome’’ to reXect its uncertain relationship to AIDP (McKhann, Cornblath, et al., 1991), similar cases with prominent motor axonal involvement as a distinguishing feature were identiWed in many countries. For this reason, this syndrome is now termed acute motor axonal neuropathy (AMAN) (GriYn, Li, Ho, Xue, Macko, Cornblath, Gao, Yang, Tian, Mishu, McKhann, and Asbury, 1995; GriYn, Li, Macko, Ho, Hsieh, Xue, Wang, Cornblath, McKhann, and Asbury, 1996; Hafer-Macko, Hsieh, et al., 1996; McKhann, Cornblath, et al., 1993). In such cases, the motor system is involved exclusively or nearly exclusively (Lu, Sheikh, et al., 2000). Other cases, such as those described initially by Feasby and colleagues, in which both motor involvement and sensory involvement are prominent, are designated acute motor sensory axonal neuropathy (AMSAN).

IV. DISEASES OF MYELIN

CLINICAL MANIFESTATIONS

CLINICAL MANIFESTATIONS AIDP constitutes over 90% of GBS cases in the United States and Western Europe (Emilia-Romagna Study Group on Clinical and Epidemiological Problems in Neurology, 1998; Hadden, Cornblath, et al., 1998). The dominant manifestation is usually weakness, often leading to paralysis (Asbury, Arnason, and Adams, 1969; Hughes, 1990; Ropper, Wijdicks, et al., 1991). However, the initial manifestations are frequently paresthesias. In addition, there may be pain of lesser or greater severity, often vaguely localized to the back. The paresthesias may begin in the toes and ‘‘ascend’’ up the leg and to the hands. They occasionally involve the face or the trunk. The initial manifestations of weakness may be in the feet and ankles, but they usually come to involve proximal muscles of the arms and legs, and most ominously, the respiratory muscles. Bulbar nerves may be aVected, as most often evidenced by facial weakness and diYculty with eye closure, and by dysarthria and diYculty swallowing. Hearing, vision, and smell are not aVected. The involvement of extraocular movements and pupillary responses suggests elements of the Fisher syndrome described later. The abrupt onset of weakness in GBS can cause understandable alarm to patients, but at the outset the Wndings on examination may be surprisingly scanty. For this reason, dating back to the original cases of Landry, hysteria has been considered a part of the initial diVerential diagnosis. An important clue to the neurologic nature of the underlying disease is the presence of depressed or absent tendon reXexes early. As also noted by Landry, progression can be frighteningly abrupt. In exceptional cases, patients may walk in and be in respiratory diYculties within hours. Progression may continue up to 4 weeks, with about half reaching their nadir in strength by 14 days after neurologic onset (Asbury, Arnason, et al., 1978; Asbury and Cornblath, 1990; Ravn, 1967). A longer phase of initial progression suggests that the patient may have a subacute or chronic inXammatory polyneuropathy rather than GBS. For the most part, laboratory tests served to exclude other disorders; there are no laboratory tests diagnostic of GBS. Even the ‘‘albuminocytologic dissociation’’ emphasized by Guillain, Barre, and Strohl (Guillain, Barre´, and Strohl, 1916) is of limited diagnostic value, since the elevated spinal Xuid protein often develops 2 or more weeks after onset of the disease, at a time when the diagnosis has already been established. Thus, while a markedly elevated protein at the start of the syndrome can be supportive of the diagnosis, its absence does not refute the diagnosis. Diseases that can mimic GBS necessitate, as appropriate, thorough inspection of the scalp for ticks that might cause tick paralysis, inquiry for possible exposure to botulinum toxin (for example, home canned foods), and historical inquiries for features that might suggest in intermittent porphyrias. Historical and, when relevant, serologic investigation for associated HIV or Lyme disease may be indicated. The most valuable laboratory study is electrodiagnosis. Electrodiagnostic studies are abnormal in some fashion in most patients with GBS (Brown and Feasby, 1984; Cornblath, Mellits, et al., 1988; van der Meche, Meulstee, et al., 1988; van der Meche´, Schmitz et al., 1992b). In the uncommon instances where they are normal on the Wrst examination, they can be expected to become abnormal on subsequent reexamination. In addition, electrodiagnostic data, as noted later, give prognostic information and help establish the class of GBS, AIDP, AMAN, or AMSAN. The Wndings suggestive of AIDP are those consistent with demyelination, including prolongation of F waves, prolongation of distal latencies, and reduction in motor or sensory nerve conduction velocities at the time when the action potential amplitudes are relatively preserved. Conversely, the AMAN syndrome suggested by markedly reduced compound motor action potential amplitudes with relatively normal motor conduction velocities, normal sensory nerve action potential amplitudes and velocities (McKhann, Cornblath, GriYn, Ho, Li, Jiang, Wu, Zhaori, Liu, Jou, Liu, Gao, Mao, Blaser, Mishu, and Asbury, 1993). Even this combination, however, cannot eliminate the possibility of restricted motor nerve terminal demyelination, as has been clearly demonstrated in physiologic-pathologic correlation studies in which extensive

IV. DISEASES OF MYELIN

889

890

38. GUILLAIN-BARRE SYNDROME

demyelination of motor nerve terminals has been demonstrated (Reisin, Cersosimo, et al., 93 A.D.; Reisin, Pociecha, et al.). The prognosis of AIDP is inXuenced by age (the older the patient, the slower the recovery), the severity at nadir, and whether or not eVective immunomodulatory therapy (intravenous immunoglobulin or plasmapheresis) were used early in the treatment (Cornblath, Mellits, GriYn, McKhann, Albers, Miller, Feasby, Quaskey, and Guillain-Barre Study Group, 1988; van der Meche, Meulstee, Vermeulen, and Kievit, 1988; van der Meche´, Schmitz, Meulstee, and Oomes, 1992b). In addition, the electrodiagnostic studies provide important prognostic data. Absent evoked compound motor action potentials predict a longer period to recovery (Cornblath, Mellits, GriYn, McKhann, Albers, Miller, Feasby, Quaskey, and Guillain-Barre Study Group, 1988; van der Meche, Meulstee, Vermeulen, and Kievit, 1988; van der Meche´, Schmitz, Meulstee, and Oomes, 1992b). Although throughout his life Guillain emphasized the favorable outcome of patients with GBS, and the mortality in the best centers is 1 to 2.5%, 15 to 20% of patients have substantial residual weakness. Both plasmapheresis and infusion of human immunoglobulin (HIG) improve the outcome of AIDP. Plasmapheresis, the Wrst therapy demonstrated to beneWt GBS (Consensus Conference, 1986; French Cooperative Group on Plasma Exchange in GuillainBarre Syndrome, 1987; Guillain-Barre Study Group, 1985), has the disadvantage of the necessity for line placement in many patients and therefore is less frequently used than infusion of HIG (van der Meche´, Schmitz, et al., 1992a). A recent trial comparing plasmapheresis and HIG found no advantage of one over the other and no advantage to using both together (Hadden, Cornblath, Hughes, Zielasek, Hartung, Toyka, and Swan, 1998). Corticosteroids alone are of no beneWt in treatment (Hughes and Swan, 1995). Ongoing trials are assessing the combination of HIG and corticosteroids.

PATHOLOGY OF AIDP Lymphocytic inWltration is characteristic of AIDP, but its severity can vary markedly. Typically, lymphocytes are present around endoneurial vessels, and individual lymphocytes are scattered throughout the endoneurial space (Fig. 38.1). The extensive perivascular and endoneurial cuVs of lymphocytes and the ‘‘plaque-like’’ demyelination seen in experimental allergic neuritis is exceptional in GBS but can occur (Fig. 38.1 A, B). Demyelination can occur anywhere in the peripheral nervous system, from the ventral roots to the nerve terminals, but typically it clusters in the spinal roots and just distal to the spinal foramina, in the mixed spinal roots and plexuses. There can be substantial endoneurial and subperineurial edema in these sites, suYcient to lead some investigators to question if, in the unyielding space of the spinal foramina, ischemic or compressive injury to nerve could contribute to the pathogenesis of Guillain-Barre (Berciano and Garcia, 2002). In any event, extensive proximal demyelination can be associated with Wallerianlike degeneration of Wbers more distally, so that many patients with absent evoked compound of motor action potential amplitudes or sensory nerve action potentials have few surviving nerve Wbers in the distal regions (Haymaker and Kernohan, 1949; Honavar, Tharakan, et al., 1991), yet have evidence of extensive demyelination more proximally. In other individuals, nerve terminal demyelination may be prominent (Hall, Hughes, et al., 1992; Massaro, Rodriguez, et al., 1998; Reisin, Cersosimo, Garcia Alvarez, Massaro, and Fejerman, 93 A.D.). Staining for immunoglobulins in GBS nerve often shows evidence of a break in the blood nerve barrier, so that all plasma constituents, including IGM, can be found within the endoneurial space. This makes it diYcult to tell with certainty whether there is binding of immunoglobulin to individual nerve Wbers. In some cases, binding of complement constituents on nerve Wbers has been demonstrated, and this can occur before demyelination is advanced. In three fatal cases of GBS in children, Hafer-Macko and colleagues (Hafer-Macko, Sheikh, et al., 1996) found that C3d and the complement membrane attack

IV. DISEASES OF MYELIN

PATHOLOGY OF AIDP

FIGURE 38.1 Pathology of AIDP. (A) This ventral rootlet from a fatal case of AIDP shows an admixture of normal myelinated nerve Wbers and fully demyelinated axons, both scattered throughout the root and in the central large plaque-like zone. In the inset, a demyelinated axon is identiWed at higher magniWcation by the arrow. One micron plastic section scale bar ¼ 50 microns. (B) A similar region in longitudinal section shows a central zone Wlled with demyelinated axons, Xanked above and below by myelinated nerve Wbers. One micron plastic section. Scale bar ¼ 20 microns. (C) Scattered mononuclear cells in the endoneurial space in this H&E stained paraYn section of a ventral root. Scale bar ¼ 50 microns. (D) Around the vessel to the lower left, there is extensive endoneurial edema and numerous mononuclear cells. The boxed region is seen at higher power in E. Scale bar ¼ 50 microns. (E) A myelinated internode from the boxed region in D is seen to end abruptly at the heminode. To the right, the axon is entirely demyelinated. A large nucleus, probably of a macrophage, sits near the node. One micron plastic section. Scale bar ¼ 20 microns.

complex neoantigen, C5b-9, aYxed to the outside of nerve Wbers, surrounding the abaxonal Schwann cell plasmalemma. Strikingly, these complement activation markers were not found on myelin per se. By doing light microscopic-electronmicroscopic comparisons on the same nerve Wbers, such Wbers were found to have early myelin vacuolization involving the outermost myelin lamellae. This vacuolization went on in other Wbers to involve extensive vacuolar degeneration of myelin, and clearance of the myelin by macrophages. This sequence raises the possibility that immunoglobulin binding and complement activation lead to the formation of sublytic complement pores in the abaxonal Schwann cell cytoplasm, with entry of calcium, activation of phospholipase A and other calciumsensitive enzymes, and consequent demyelination.

IV. DISEASES OF MYELIN

891

892

38. GUILLAIN-BARRE SYNDROME

This immunopathology is not universal in AIDP; in many cases it is not possible to demonstrate complement activation markers on the surface, and macrophages invade entirely normal appearing myelin sheaths, engaging in stripping of myelin. Such cases may have diVerent immunopathogenesis.

ANTECEDENTS OF GBS AIDP may follow antecedent illnesses, including infectious diseases in about two-thirds of cases and, less frequently, surgery, parturition, and other life events. Post-infectious GBS typically follows a bacterial or viral infection by 10 to 14 days. The antecedent infections that have clear links to GBS include infection with the gram-negative bacteria Campylobacter-jejuni (C. jejuni) (Jacobs, Van Doorn, et al., 1996; Kaldor and Speed, 1984; Rees, Gregson, et al., 1995; Walsh, Cronin, et al., 1991) and the herpes virus, Cytomegalovirus (Dowling and Cook, 1981; Visser, van der Meche´, et al., 1996). Mycoplasma pneumoniae infection can also precede GBS (Jacobs, Rothbarth, et al., 1998). A variety of disorders that alter immune function have been associated with GBS, including HIV infection (Cornblath, McArthur, et al., 1987), Hodgkins disease (Lisak, Mitchell, et al., 1977), and pharmacologic immunosuppression (Drachman, Patterson, et al., 1970). Other antecedents include pregnancy and delivery, surgery, and an extraordinary variety of other infections. It can be associated with Lyme disease. A few vaccinations have had a suggested relationship to GBS. The Semple Rabies vaccine can clearly produce GBS, due almost certainly to the inclusion of myelin constituents in the vaccine (Hemachudha, Phanuphak, et al., 1987). The 1976 inXuenza vaccine, the swine Xu, produced a modest increase in case rate within the Wrst weeks after immunization (Kaplan, Schonberger, et al., 1983; Lasky, Terracciano, et al., 1998; Schonberger, Bregman, et al., 1979). When an antecedent infection such as Campylobacter enteritis precedes GBS, the acute infectious manifestations and fever have generally abated before onset of the neurologic disorder. In the case of Campylobacter, the organism usually has been cleared by the time of neurologic presentation. Antecedent infection can be suggested by serologic studies, although Campylobacter serology is a specialized test and better suited to comparison of populations than individual diagnosis. In GBS, the seroprevalence of Campylobacter ranges from 15 to 75% in various parts of the world (Ho, Mishu, et al., 1995; Kaldor and Speed, 1984; Mishu, Ilyas, et al., 1993; McKhann, Cornblath, GriYn, Ho, Li, Jiang, Wu, Zhaori, Liu, Jou, Liu, Gao, Mao, Blaser, Mishu, and Asbury, 1993; Rees, Soudain, et al., 1995; Speed, Kaldor, et al., 1984). Enteric infection with Campylobacter is the most frequent bacterial cause of diarrhea worldwide, with an estimated 2.4 million cases per year in the United States alone. Yet only about 2500 individuals develop GBS in North America annually, and only a portion of these can be ascribed to antecedent Campylobacter infection (Buzby, Allos, et al., 1997; McCarthy, Andersson, et al., 1999; Mishu and Blaser, 1993; Tauxe, 1992). Some Campylobacter strains are more likely to be associated with GBS than others. HS (Penner) serotyping, which detects capsular polysaccharides distinct from the lipopolysaccharide (LPS), is an important epidemiological tool in studying C. jejuni-associated GBS (Penner and Hennessy, 1980; Penner, Hennessy, et al., 1983). Certain HS serotypes are overrepresented in GBS patients in diVerent parts of the world. For example, HS:19, an uncommon serotype in patients with diarrhea, is isolated in up to 90% of Japanese Campylobacter associated GBS (Fujimoto, Yuki, et al., 1992; Kuroki, Saida, et al., 1993; Yuki, Takahashi, et al., 1997). This serostrain is also over-represented in Chinese (Sheikh, Nachamkin, et al., 1998) and Mexican (Irving Nachamkin, unpublished observations) patients with GBS, whereas HS:41 serotype is overrepresented in GBS patients from South Africa (Lastovica, Goddard, et al., 1997; Prendergast, Lastovica, et al., 1998). In the United Kingdom (UK), neither serotype is overrepresented (Rees, Soudain, Gregson, and Hughes, 1995). Even though the risk of GBS following Campylobacter is low, understanding the neuritogenic properties of the organisms has important implications for development of a Campylobacter jejuni vaccine, currently a priority for the military.

IV. DISEASES OF MYELIN

PATHOGENETIC MECHANISMS

In general post-Campylobacter cases are more likely to be severe, to have prominent axonal involvement, and to have disproportionate motor involvement (Jacobs, Van Doorn, Schmitz, Tio-Gillen, Herbrink.P, Visser, Hooijkaas, and van der Meche, 1996; Rees, Gregson, and Hughes, 1995; Visser, van der Meche´, et al., 1995). In contrast, postCMV cases appear are more likely to have prominent or predominant sensory involvement (Visser, van der Meche´, Meulstee, Rothbarth, Jacobs, Schmitz, Van Doorn, and Dutch Guillain-Barre´ Study Group, 1996).

PATHOGENETIC MECHANISMS One of the ironies of GBS at the beginning of the 21st century is that the pathogeneses of the less frequent ‘‘variants’’ of GBS are better understood than that of AIDP. The model of molecular mimicry is attractive in AIDP, but the pathogenetic reconstruction of AIDP remains incomplete. The target antigens in AIDP are usually unknown and are likely to diVer among diVerent cases. The extent to which T cell-and antibody-mediation are involved is unresolved, and again may diVer. These issues contrast with AMAN and with the Fisher syndrome. These variants have provided one of the most attractive examples of ‘‘molecular mimicry,’’ in which immune attack is directed toward an antigen of an infectious agent that is similar to an antigen that is present and ‘‘seen’’ by the immune system on nerve Wbers. These disorders are regularly associated with speciWc antiganglioside antibodies, and ganglioside-like moieties to be present on organisms isolated from these patients. The best-documented example is Campylobacter jejuni, which can have relevant ganglioside-like antigens in its lipooligo-saccharide. Among patients with Campylobacter infection, only those who go on to GBS have high titers of anti-ganglioside antibodies (Oomes, Jacobs, et al., 1995; Rees, Gregson, and Hughes, 1995; Sheikh, Nachamkin, Ho, Willison, Veitch, Ung, Nicholson, Li, Wu, Shen, Cornblath, Asbury, McKhann, and GriYn, 1998). The recent development of successful animal models based on sensitization to these gangliosides and to the relevant Campylobacter antigens have substantially ‘‘closed the loop’’ in understanding the role of molecular mimicry in the AMAN syndrome, as described here. The GBS variant that may bridge the way to comparable studies of AIDP is the Fisher syndrome. Described in 1956 on the basis of the triad of internal and external ophthalmoplegia, ataxia, and areXexia with little weakness, the Fisher syndrome was proposed to be a variant of GBS (Fisher, 1956). That suggestion was initially controversial, but it has subsequently been observed that many cases that present as Fisher syndrome evolve into typical AIDP. These patients may require respiratory support, ventilator assistance, and have all the other electrodiagnostic manifestations of typical AIDP. The pathology of ‘‘pure’’ Fisher syndrome, without more widespread weakness, is unknown because it is not a fatal disorder. However, the few cases that had gone on to death after developing a paralysis have had evidence of inXammatory demyelination. Serologic studies established that the 90% of Fisher syndrome or Fisher-AIDP overlap cases have acute-phase serum antibodies against GQ1b gangliosides; these antibodies disappear with clinical recovery (Chiba, Kusunoki, et al., 1992; Willison, Veitch, et al., 1993; Yuki, Sato, et al., 1993b). Fisher syndrome patients may also have antibodies that react with structurally related gangliosides containing disialosyl moieties, including GT1a, GD1b, and GT1b (Willison, Almemar, et al., 1994). GQ1b is enriched in oculomotor nerves (Chiba, Kusunoki, et al., 1993; Chiba, Kusunoki, et al., 1997), the principal motor site aVected in Fisher syndrome. Fisher syndrome occasionally follows such antecedent infections with Campylobacter. Campylobacter jejuni lipopolysaccharides can contain ganglioside-like moieties, and several C. jejuni isolates from Fisher patients contain the structurally similar GQ1b-, GT1a-, and GD3-like moieties (Aspinall, McDonald, et al., 1994; Jacobs, Endtz, et al., 1995; Salloway, Mermel, et al., 1996; Yuki, Taki, et al., 1994). Rabbits immunized with C. jejuni LPS from patients with Fisher syndrome can produce cross-reactive antibodies recognizing

IV. DISEASES OF MYELIN

893

894

38. GUILLAIN-BARRE SYNDROME

GQ1b (Goodyear, O’Hanlon, et al., 1999). Willison and colleagues have undertaken a detailed analysis of the pathogenetic role of anti-GQ1b antibodies in experimental settings. They showed that anti-GQ1b antibodies stain the terminal axon at the neuromuscular junction in rat phrenic nerve-diaphragm preparations, and that anti-GQ1b antibodies could produce complement-dependent motor nerve terminal degeneration (Goodyear, O’Hanlon, Plomp, Wagner, Morrison, Veitch, Cochrane, Bullens, Molenaar, Conner, and Willison, 1999; Plomp, Molenaar, et al., 1999a; Roberts, Willison, et al., 1994). This degeneration was heralded by massive release of quanta from the motor nerve terminal (Goodyear, O’Hanlon, Plomp, Wagner, Morrison, Veitch, Cochrane, Bullens, Molenaar, Conner, and Willison, 1999). This eVect reXected calcium entry into the terminal and preceded swelling and destruction of the terminal (Plomp, Molenaar, et al., 1999b; Plomp, Molenaar, O’Hanlon, Jacobs, Veitch, Daha, Van Doorn, van der Meche, Vincent, Morgan, and Willison, 1999a). Buchwald and Toyka have found evidence of a complement-independent eVect of anti-GQ1b antibodies on motor nerve terminals (Buchwald, Weishaupt, et al., 1998). Thus, an attractive reconstruction of the Fisher syndrome is that GQ1b is enriched in oculomotor nerves and that the generation of anti-GQ1b antibodies by an antecedent infection produces ophthalmoparesis because of immune-mediated injury of the oculomotor nerve terminals. BickerstaV’s brain stem encephalitis has ocular features similar to the Fisher syndrome, but associated with evidence of central nervous system involvement and T2 brightness in the brain stem on MRI studies. Such cases can also follow Campylobacter infection and are associated with anti-GQ1b antibodies (Yuki, Sato, et al., 1993a). This suggests that antiganglioside antibodies might also be capable of producing CNS involvement when they have access. The association of the Fisher syndrome and the axonal forms of GBS with speciWc antiganglioside immune responses raises the possibility that a similar association underlies AIDP. Some cases of AIDP have antiganglioside antibodies, and there have been associations of speciWc antiganglioside antibody patterns with speciWc clinical or prognostic patterns. For example, cases of AIDP with predominant motor involvement and with a poorer prognosis have been associated with IgG anti-GM1 antibodies (Jacobs, Van Doorn, Schmitz, Tio-Gillen, Herbrink.P, Visser, Hooijkaas, and van der Meche, 1996; Rees, Gregson, and Hughes, 1995; Visser, van der Meche´, Van Doorn, and, et al., 1995). This raises the possibility that in these cases, the poorer prognosis is associated with an increased ‘‘axonal’’ component, related to the anti-GM1 antibodies. One possible exception is antibody against the major PNS myelin ganglioside, LM1 (sialsylneolacto tetrasylceromide). In one thoroughly reported AIDP patient, antibody against LM1, a peripheral nerve ganglioside enriched in myelin was present in high titers at the onset of disease and fell over time (Ilyas, Willison, et al., 1988). Two recent reports have identiWed anti-LM1 antibodies in 5 to 25% of AIDP patients (Harukawa, Utsumi, et al., 2002; Yako, Kusunoki, et al., 1999); in these reports they were rarely associated with axonal cases, although one other found anti-LM1 antibodies in 29% of axonal GBS. Other candidate antigens for AIDP, including GD1b, asialo-GM1, Gal (b1-3)GalNAc epitope, GM2 and GT1b (Fredman, Vedeler, et al., 1991; Gregson, Koblar, et al., 1993; Ho, Mishu, Li, Gao, Cornblath, GriYn, Asbury, Blaser, and McKhann, 1995; Ilyas, Mithen, et al., 1992; Ilyas, Willison, Quarles, Jungawala, Cornblath, Trapp, GriYn, GriYn, and McKhann, 1988; Rees, Gregson, and Hughes, 1995;), are infrequently identiWed. It is notable that anti-GM1 antibodies can be present in both AMAN and AIDP. The Wne speciWcities of these antibodies or diVerential antibody aYnity for gangliosides of neuronal and glial origin need to be explored to explain this apparent paradox. As noted, anti-galactocerebroside and anti-GM2 antibodies have been identiWed after antecedent infections with Mycoplasma and CMV, respectively. Prominent bulbar and facial involvement have been associated with anti-GT1a antibodies (Kashihara, Shiro, et al., 1998; Koga, Yuki, et al., 1998) and a sensory ataxic neuropathy with anti-GD1b antibodies (Miyazaki, Kusunoki, et al., 2001). The latter pattern has been reproduced in a rabbit model by immunization with GD1b (Kusunoki, Hitoshi, et al., 1999), in this model the target appears to be sensory ganglion cells rather than myelin.

IV. DISEASES OF MYELIN

PATHOGENETIC MECHANISMS

On balance, it seems likely that relatively few cases of AIDP represent immune responses to gangliosides. The frequency of antiganglioside antibodies in AIDP, in comparison to the axonal and Fisher syndrome of patients, is low. These issues are complicated by the wellpublicized diYculties in reproducibility and reliability of antiganglioside antibody assays and by the diYculties in localization in gangliosides on nerve Wbers. The problem of localization of antigen with gangliosides has been improved by the availability of high titer monospeciWc antiganglioside antibodies, generated by immunization of genetically engineered mice deWcient in gangliosides of interest. The most widely used such mice are the GD3/GD2 synthesis knockout mice, which make GD3, but no subsequent complex gangliosides (Sheikh, Sun, et al., 1999). Immunization of these animals with such gangliosides as GD1a, GM1, GD1b, and GT1b have produced the high titer monospeciWc antibodies that can either be complement Wxing (mouse IgG2a or 2b) or noncomplement Wxing (mouse IgG1 or 3). Lessons from the use of these high titer monospeciWc antibodies to assess ganglioside localization include the recognition that the Wxation and preparation of the tissue aVects the apparent localization (Gong, Tagawa, et al., 2002; Lunn, Johnson, et al., 2000; Sheikh, Deerinck, et al., 1999). It is diYcult to ‘‘see’’ myelin gangliosides with immunocytochemistry, except in paranodal regions, if the method of preparation does not open the myelin sheath. Thus, they may be poorly seen on Wxed teased Wber preparations and yet relatively abundant in transverse fresh frozen cryostat sections. Similarly, the biochemical ‘‘surround’’ of gangliosides within membrane may produce crypticity of gangliosides in some settings. The abaxonal surface of Schwann cells is frequently stained by anti-ganglioside antibodies. This observation implies that if anti-glycolipid antibodies are targets in demyelination, the surface expression of glycolipid antigens on glial cells may be suYcient for antibody binding, complement activation, and calcium entry that can lead to demyelination, without a direct immune attack on myelin. Finally, there are important Wne speciWcities of anti-gangliosides antibodies, so that some antibodies selectively stain certain neuronal and nerve Wber populations. These diVerences in staining patterns are not explained by ganglioside content or antibody binding to extracted gangliosides from these diVerentially stained neurons or nerve Wbers (Fig. 38.2). Compared to the axonal forms, AIDP has greater T-cell-mediated inWltration component (Asbury, Arnason, and Adams, 1969), the presence of T-cell activation markers in the serum and CSF of AIDP patients (Bansil, Mithen, et al., 1991; Sharief, McLean, et al., 1993; Sharief, Ingram, et al., 1997; Sivieri, Ferrarini, et al., 1997), and on the pathologic similarities to the animal model EAN (Arnason and Soliven, 1993; Hartung, Pollard, et al., 1995a; Hartung, Pollard, et al., 1995b). Although the trigger for T-cell activation is not clear, T cells could contribute to pathogenesis of AIDP in several ways. Activated T cells may play an important role in breakdown of the blood-nerve-barrier in recruitment of macrophages. They may also contribute to Schwann cell and myelin injury, either by direct cytotoxic mechanisms or indirectly through proinXammatory cytokines. There is a growing interest in the role of antibody-mediated demyelination in AIDP. As noted, in some AIDP cases complement activation markers are found on the outermost Schwann cell surface and can be associated with vesicular demyelination (Hafer-Macko, Sheikh, Li, Ho, Cornblath, McKhann, Asbury, and GriYn, 1996). This pattern closely resembles the experimental nerve Wber demyelination induced by anti-galactocerebroside (GalC). GalC is a glycosphingolipid enriched in myelin (Saida, Saida, et al., 1979). It is possible that in these cases the antibody and complement is more directly involved in targeting the Schwann cell and myelin, and the role of T cells may be to open the blood nerve barrier (Pollard, Westland, et al., 1995; Spies, Pollard, et al., 1995; Spies, Westland, et al., 1995). Several clinical and experimental observations support a role for antibodymediated mechanisms, including the response to plasmapheresis (French Cooperative Group on Plasma Exchange in Guillain-Barre Syndrome, 1987; Guillain-Barre Study Group, 1985), the presence of anti-myelin (Koski, Chou, et al., 1989; Koski, Humphrey, et al., 1985) and anti-glycoconjugate antibodies, and the ability of AIDP sera to induce demyelination after intraneural injection (Saida, Saida, et al., 1982) or in vitro incubation (Birchem, Mithen, et al., 1987; Koski, Chou, and Jungalwala, 1989; Mithen, Ilyas, et al., 1992; Sawant-Mane, Clark, et al., 1991; Sawant-Mane, Estep, et al., 1994).

IV. DISEASES OF MYELIN

895

896

38. GUILLAIN-BARRE SYNDROME

FIGURE 38.2 (A) Teased Wber preparation showing cholera toxin staining (ganglioside GM1) a node of Ranvier and paranodal Schwann cell. Scale bar ¼ 20 microns. (B) A mixed spinal root section stained for ganglioside GD1a showing preferential staining of motor (M) compared with sensory (S) root. Scale bar ¼ 50 microns.

Molecular Mimicry and the AMAN Syndrome The molecular basis for ganglioside-like mimicry is beginning to come to light following the sequencing of the Campylobacter genome (Parkhill, Wren, et al., 2000). There are sialyltransfereases and sialic acid synthetases involved in ganglioside-like epitope expression (Gilbert, Brisson, et al., 2000; Linton, Gilbert, et al., 2000; Linton, Karlyshev, et al., 2000). Ganglioside-like moieties have regularly been found in the GBS isolates. For example, C. jejuni isolates from AMAN patients with anti-GM1 antibodies patients contain GM1-like moieties (Yuki, Handa, et al., 1992; Yuki, Taki, et al., 1993). The antiganglioside antibodies in AMAN are IgG species, and typically directed against GM1, GD1a, GalNac-GD1a, and occasionally GM1b (Gregson, Koblar, and Hughes, 1993; Ho, Mishu, Li, Gao, Cornblath, GriYn, Asbury, Blaser, and McKhann, 1995; Kornberg, Pestronk, et al., 1994; Kuwabara, Yuki, et al., 1998; Ogino, Orazio, et al., 1995; Willison and Veitch, 1994; Yuki, Takahashi, Tagawa, Kashiwase, Tadokoro, and Saito, 1997; Yuki, Yoshino, Sato, and Miyatake, 1990; Yuki, Yoshino, et al., 1990). Immunopathologic studies have found few lymphocytes, even in fatal cases (McKhann, Cornblath, GriYn, Ho, Li, Jiang, Wu, Zhaori, Liu, Jou, Liu, Gao, Mao, Blaser, Mishu, and Asbury, 1993), and identiWed binding of complement to nodes of Ranvier and the internodal axolemma (Hafer-Macko, Hsieh, Li, Ho, Sheikh, Cornblath, McKhann, Asbury, and GriYn, 1996). A characteristic feature of the axonal cases is recruitment of macrophages to the nodes and into the periaxonal space surrounding the nodes of Ranvier (GriYn, Li, Macko, Ho, Hsieh, Xue, Wang, Cornblath, McKhann, and Asbury, 1996). In the most severe cases, motor axons undergo Wallerian-like degeneration that extends from the ventral root exit zone to the motor nerve terminal (GriYn, Li, Ho, Xue, Macko, Cornblath, Gao, Yang, Tian, Mishu, McKhann, and Asbury, 1995; GriYn, Li, Macko, Ho, Hsieh, Xue, Wang, Cornblath, McKhann, and Asbury, 1996). In other cases, motor axons appear to degenerate only in their terminal regions, where the motor nerve terminal is outside the blood nerve barrier (Ho, Hsieh, et al., 1997). Such cases can recover surprisingly promptly.

IV. DISEASES OF MYELIN

RELEVANCE OF GBS TO ACQUIRED DEMYELINATING DISORDERS OF THE CNS

In northern China, the speciWcity of antiganglioside antibodies for the AMAN syndrome is greatest for antibodies against GD1a and the minor gangliosides, GalNAc-GD1a and GM1b. Anti-GD1a antibodies are signiWcantly elevated in patients with AMAN (60%) compared to AIDP (4%) (Ho, Willison, et al., 1999). In this group anti-GM1, anti-GM1b, and anti-GalNAc-GD1a were also more frequent in AMAN than AIDP (Yuki, Ho, et al., 1999). Elevated titers of anti-GM1b and anti-GalNAc-GD1a were more commonly associated with motor-predominant variants in Japanese and Dutch patients (Ang, Yuki, et al., 1999; Hao, Kaida, Kusunoki, et al., 2000; Saida, et al., 1999; Yuki, Ang, et al., 2000). The presence of antibodies against these gangliosides is strongly related to preceding C. jejuni infection (Ang, Yuki, Jacobs, Koga, Van Doorn, Schmitz, and van der Meche, 1999; Chiba, et al., 1994; Hao, Saida, Yoshino, Kuroki, Nukina, and Saida, 1999; Ho, Willison, Nachamkin, Li, Veitch, Ung, Wang, Liu, Cornblath, Asbury, GriYn, and McKhann, 1999; Kaida, Kusunoki, Kamakura, Motoyoshi, and Kanazawa, 2000; Kusunoki, Iwamori, et al., 1996; Kusunoki, Yuki, Ho, Tagawa, Koga, Li, Hirata, and GriYn, 1999; Yuki, Ang, Koga, Jacobs, Van Doorn, Hirata, and van der Meche, 2000; Yuki, Yoshino, et al., 1992), and the LPSs of C. jejuni carry GD1a- and GM1b-like moieties. The correlation between AMAN, anti-GM1 antibodies and C. jejuni infection is not found in all patient populations (Enders, Karch, et al., 1993; Vriesendorp, Mishu, et al., 1993). Biochemical data and localization studies indicate that GM1-like moieties are present in both axons and myelin (O’Hanlon, Paterson, et al., 1996) perhaps explaining the fact that IgG anti-GM1 antibodies are also seen in some cases of AIDP. The basis for the motor-predominant symptoms in AMAN remains a target of investigation. The diVerential display of speciWc gangliosides on motor rather than sensory Wbers is a potential explanation. GD1a localization supports this concept. GD1a has been localized by taking advantage of the ability to generate high titer monospeciWc antiganglioside antibodies, using mice in which the enzyme GM2/GD2 synthase was genetically targeted. These mice respond to immunization with gangliosides by raising much higher titers of monospeciWc antibody than seen in wild-type animals (Lunn, Johnson, Fromholt, Itonori, Huang, Vyas, Hildreth, GriYn, Schnaar, and Sheikh, 2000). The resulting antiGD1a antibodies bind to motor axons but only a subpopulation of small sensory axons (Gong, Tagawa, Lunn, Laroy, HeVer-Lauc, Li, GriYn, Schnaar, and Sheikh, 2002). GalNAc-GD1a, a minor ganglioside in peripheral nerves, has also been associated with AMAN, and is reported to be present in human spinal motor neurons and motor nerves but not sensory nerves (Hao, Saida, Yoshino, Kuroki, Nukina, and Saida, 1999; Yoshino, 1997). Recently Yuki et al. have developed a model of AMAN by immunization of rabbits with GM1 and complete Freund’s adjuvant for several months (Yuki, Yamada, et al., 2001). They produced Wallerian-like degeneration of motor axons without inXammation, and with evidence of IgG antibody binding to axolemma and the distinctive periaxonal macrophages seen in human AMAN. Sheikh and colleagues (unpublished) have shown that high titer monoclonal IgG anti-GD1a antibodies administered to mice produce noninXammatory axonal degeneration. Li et al. (Li, Xue, et al., 1996) developed an animal model of AMAN by feeding a C. jejuni isolate from a patient with AMAN to chickens. The pathological changes in the nerves of aVected chickens were similar to those seen in human cases. Unfortunately, this model has not been reproduced in other laboratories, perhaps reXecting diVerences in the chicken strains. Taken together, these in vivo models conWrm that antiganglioside antibodies can produce noninXammatory axonal injury.

RELEVANCE OF GBS TO ACQUIRED DEMYELINATING DISORDERS OF THE CNS The Guillain Barre syndrome (GBS) and multiple sclerosis (MS) represent the major immune-mediated disorders of the PNS and the CNS, respectively. Several common themes have recently emerged. There are similarities in the immune organization of

IV. DISEASES OF MYELIN

897

898

38. GUILLAIN-BARRE SYNDROME

the PNS and CNS. Several recent studies have underlined the heterogeneity in the pathology and immunopathology of MS (Bruck, Lucchinetti, et al., 2002; Lassmann, 1998; Lucchinetti, Bruck, et al., 2000), as well as GBS. This heterogeneity includes variation in the severity of lymphocytic inXammation, in the extent to which antibody and complement appear to play roles in tissue injury, in the extent of axonal degeneration (Trapp, Bo, et al., 1999). Several conclusions from GBS might provide questions for the future in MS. First, the data for GBS suggest that the immunologic mechanism can involve molecular mimicry, at least in some GBS variants. Through this mechanism, infectious agents as diverse as CMV, EBV, Mycoplasma, and Campylobacter jejuni have been linked to GBS syndromes that appeared similar until detailed electrophysiology, pathology, and immunology were applied. Could such etiologic heterogeneity apply to MS? Second, the GBS experience suggests that the pathogenetic eVector mechanisms can diVer, with some disorders largely antibody-dependent, whereas the role of antibody is uncertain in others. The immunopathologic data suggest this may also apply to MS (Bruck, Lucchinetti, and Lassmann, 2002; Lassmann, 1998; Lucchinetti, Bruck, Parisi, Scheithauer, Rodriguez, and Lassmann, 2000). Third, as the GBS experience has shown, the target antigens need not be intrinsic myelin proteins, and indeed need not be proteins. Glycolipid serology is a notoriously diYcult Weld. Might glycolipids be antigenic in some cases of MS, or become antigenic if antigen spreading develops? The experience with BickerstaV’s brain stem encephalitis focuses this question. As noted earlier, BickerstaV’s encephalitis is an acute monophasic CNS disorder associated with eye movement abnormalities. A small number of cases have recently demonstrated that BickerstaV’s can follow Campylobacter jejuni infection, and can be associated with antiGQ1b antibodies, and can overlap with the Fisher syndrome (Yuki, Sato, Tsuji, Hozumi, and Miyatake, 1993a). Finally, in both GBS and MS it is likely that multiple mechanisms render the axon vulnerable. These mechanisms include damage as a bystander to inXammatory disease, as a consequence of the intimate cell-cell interactions between the myelin-forming cell and axon, and possibly as the target of the immune attack.

Acknowledgments We wish to thank and acknowledge our long-term collaborators Arthur K. Asbury, David R. Cornblath, Tony W. Ho, Chung Yun Li, Guy M. McKhann, Irving Nachamkin, and Hugh J. Willison.

References Ang, C. W., Yuki, N., Jacobs, B. C., Koga, M., Van Doorn, P. A., Schmitz, P. I., and van der Meche, F. G. (1999). Rapidly progressive, predominantly motor Guillain-Barre syndrome with anti-GalNAc-GD1a antibodies. Neurology 53, 2122–2127. Arnason, B. G. W., and Soliven, B. (1993). Acute inXammatory demyelinating polyradiculopathy. In ‘‘Peripheral Neuropathy’’ (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low, and J. F. Poduslo, eds.), pp. 1437–1497. W. B. Saunders, Philadelphia. Asbury, A. K., Arnason, B. G., and Adams, R. D. (1969). The inXammatory lesion in idiopathic polyneuritis. Medicine 48, 173–215. Asbury, A. K., Arnason, B. G., Karp, H. R., and McFarlin, D. E. (1978). Criteria for diagnosis of Guillain-Barre syndrome. Annals of Neurology 3, 565–566. Asbury, A. K., and Cornblath, D. R. (1990). Assessment of current diagnostic criteria for Guillain-Barre syndrome. Annals of Neurology 27 (Supp), S21–S24. Aspinall, G. O., McDonald, A. G., Pang, H., Kurjanczyk, L. A., and Penner, J. L. (1994). Lipopolysaccharides of Campylobacter jejuni serotype O:19: Structures of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-Barre´ syndrome. Biochemistry 33, 241–249. Bansil, S., Mithen, F. A., Cook, S. D., SheVet, A., and Rohowsky-Kochan, C. (1991). Clinical correlation with serum-soluble interleukin-2 receptor levels in Guillain-Barre´ syndrome. Neurology 41, 1302–1305. Berciano, J., and Garcia, A. (2002). Nerve ischemia in Guillain-Barre syndrome: An alternative mechanism for early conduction failure. Rev. Neurol (Paris) 158, 364–365.

IV. DISEASES OF MYELIN

RELEVANCE OF GBS TO ACQUIRED DEMYELINATING DISORDERS OF THE CNS

Birchem, R., Mithen, F. A., L’Empereur, K. M., and Wessels, M. M. (1987). Ultrastructural eVects of GuillainBarre´ serum in cultures containing only rat Schwann cells and dorsal root ganglion neurons. Brain Research 421, 173–185. Brown, W. F., and Feasby, T. E. (1984). Conduction block and denervation in Guillain-Barre polyneuropathy. Brain 107, 219–239. Bruck, W., Lucchinetti, C., and Lassmann, H. (2002). The pathology of primary progressive multiple sclerosis. Mult. Scler. 8, 93–97. Buchwald, B., Weishaupt, A., Toyka, K. V., and Dudel, J. (1998). Pre and postsynaptic blockade of neuromuscular transmission by MillerFishersyndrome IgG at mouse motor nerve terminals. European Journal of Neuroscience 10, 281–290. Buzby, J. C., Allos, B. M., and Roberts, T. (1997). The economic burden of Campylobacter-associated GuillainBarre syndrome. Journal of Infectious Disease 176 Suppl 2, S192-S197. Chiba, A., Kusunoki, S., Obata, H., Machinami, R., and Kanazawa, I. (1993). Serum anti-GQ1b antibody is associated with ophthalmoplegia in Miller Fisher syndrome and Guillain-Barre syndrome: Clinical and immunohistochemical studies. Neurology 43, 1911–1917. Chiba, A., Kusunoki, S., Obata, H., Machinami, R., and Kanazawa, I. (1997). Ganglioside composition of the human cranial nerves, with special reference to pathophysiology of Miller Fisher syndrome. Brain Research 745, 32–36. Chiba, A., Kusunoki, S., Shimizu, T., and Kanazawa, I. (1992). Serum IgG antibody to ganglioside GQ1b is a possible marker of Miller Fisher syndrome. Annals of Neurology 31, 677–679. Consensus Conference (1986). Consensus Conference: The utility of therapeutic plasmapheresis for neurological disorders. Journal of the American Medical Association 256, 1333–1337. Cornblath, D. R., McArthur, J. C., Kennedy, P. G. E., Witte, A. S., and GriYn, J. W. (1987). InXammatory demyelinating peripheral neuropathies associated with human T-cell lymphotropic virus type III infection. Annals of Neurology 21, 32–40. Cornblath, D. R., Mellits, E. D., GriYn, J. W., McKhann, G. M., Albers, J. W., Miller, R. G., Feasby, T. E., Quaskey, S. A., and Guillain-Barre Study Group (1988). Motor conduction studies in the Guillain-Barre syndrome: Description and prognostic value. Annals of Neurology 23, 354–359. Dowling, P. C., and Cook, S. D. (1981). Role of infection in Guillain-Barre syndrome: Laboratory conWrmation of Herpes viruses in 41 cases. Annals of Neurology 9(Suppl), 44–55. Drachman, D. A., Patterson, P. Y., Berlin, B., and Roguska, J. (1970). Immunosuppression in the Guillain-Barre syndrome. Archives of Neurology 23, 385–393. Eichhorst, H. (1877). Neuritis acuta progressiva. Virchow’s Archiv. 268. Emilia-Romagna Study Group on Clinical and Epidemiological Problems in Neurology. (1998). Guillain-Barre´ syndrome variants in Emilia-Romagna, Italy, 1992–3: Incidence, clinical features, and prognosis. Journal of Neurology, Neurosurgery, and Psychiatry 65, 218–224. Enders, U., Karch, H., Toyka, K. V., Michels, M., Zielasek, J., Pette, M., Heesemann, J., and Hartung, H.-P. (1993). The spectrum of immune responses to Campylobacter jejuni and glycoconjugates in Guillain-Barre syndrome and in other neuroimmunological disorders. Annals of Neurology 34, 136–144. Feasby, T. E., Gilbert, J. J., Brown, W. F., et al. (1986). An acute axonal form of Guillain-Barre polyneuropathy. Brain 109, 1115–1126. Feasby, T. E., Hahn, A. F., Brown, W. F., Bolton, C. F., Gilbert, J. J., and Koopman, W. J. (1993). Severe axonal degeneration in acute Guillain-Barre syndrome: Evidence of two diVerent mechanisms? Journal of Neurological Science 116, 185–192. Fisher, M. (1956). An unusual variant of acute idiopathic polyneuritis (syndrome of ophthalmoplegia ataxia and areXexia). New England Journal of Medicine 255, 57–65. Fredman, P., Vedeler, C. A., Nyland, H., Aarli, J. A., and Svennerholm, L. (1991). Antibodies in sera from patients with inXammatory demyelinating polyradiculoneuropathy react with ganglioside LM1 and sulphatide of peripheral nerve myelin. Journal of Neurology 238, 75–79. French Cooperative Group on Plasma Exchange in Guillain-Barre Syndrome (1987). EYcacy of plasma exchange in Guillain-Barre syndrome: Role of replacement Xuids. Annals of Neurology 22, 753–761. Fujimoto, S., Yuki, N., Itoh, T., and Amako, K. (1992). SpeciWc serotype of Campylobacter jejuni associated with Guillain-Barre syndrome. Journal of Infectious Disease 165, 183 (letter). Gilbert, M., Brisson, J. R., Karwaski, M. F., Michniewicz, J., Cunningham, A. M., Wu, Y., Young, N. M., and Wakarchuk, W. W. (2000). Biosynthesis of ganglioside mimics in Campylobacter jejuni OH4384. IdentiWcation of the glycosyltransferase genes, enzymatic synthesis of model compounds, and characterization of nanomole amounts by 600-mhz (1)h and (13)c NMR analysis. Journal of Biological Chemistry 275, 3896–3906. Gong, Y., Tagawa, Y., Lunn, M. P., Laroy, W., HeVer-Lauc, M., Li, C. Y., GriYn, J. W., Schnaar, R. L., and Sheikh, K. A. (2002). Localization of major gangliosides in the PNS: Implications for immune neuropathies. Brain 125, 2491–2506. Goodyear, C. S., O’Hanlon, G. M., Plomp, J. J., Wagner, E. R., Morrison, I., Veitch, J., Cochrane, L., Bullens, R. W., Molenaar, P. C., Conner, J., and Willison, H. J. (1999). Monoclonal antibodies raised against GuillainBarre syndrome-associated Campylobacter jejuni lipopolysaccharides react with neuronal gangliosides and paralyze muscle-nerve preparations [published erratum appears in J Clin Invest 1999 Dec; 104 (12): 1771]. J. Clin. Invest 104, 697–708. Graves, R. J. (1884). Clinical Lectures on the Practice of Medicine. New Syd. Soc. (reprinted from the Wrst edition, 1843)1, 578.

IV. DISEASES OF MYELIN

899

900

38. GUILLAIN-BARRE SYNDROME

Gregson, N. A., Koblar, S., and Hughes, R. A. (1993). Antibodies to gangliosides in Guillain-Barre syndrome: SpeciWcity and relationship to clinical features [see comments]. Q. J. Med. 86, 111–117. GriYn, J. W., Li, C. Y., Ho, T. W., Xue, P., Macko, C., Cornblath, D. R., Gao, C. Y., Yang, C., Tian, M., Mishu, B., McKhann, G. M., and Asbury, A. K. (1995). Guillain-Barre syndrome in northern China: The spectrum of neuropathologic changes in clinically deWned cases. Brain 118, 577–595. GriYn, J. W., Li, C. Y., Macko, C., Ho, T. W., Hsieh, S.-T., Xue, P., Wang, F. A., Cornblath, D. R., McKhann, G. M., and Asbury, A. K. (1996). Early nodal changes in the acute motor axonal neuropathy pattern of the Guillain-Barre syndrome. Journal of Neurocytology 25, 33–51. Guillain-Barre Study Group (1985). Plasmapheresis and acute Guillain-Barre syndrome. Neurology 35, 1096–1104. Guillain, G., Barre´, J. A., and Strohl, A. (1916). Sur un syndrome de radiculonebrite avec hyperalbuminose du liquide cephalo-rachidien sans reaction cellulaire. Remarques sur les caracteres cliniques et graphiques des reXexes tendineux. Bulletin de Societe des Medicines Hopitals de Paris 40, 1462. Hadden, R. D., Cornblath, D. R., Hughes, R. A., Zielasek, J., Hartung, H. P., Toyka, K. V., and Swan, A. V. (1998). Electrophysiological classiWcation of Guillain-Barre syndrome: Clinical associations and outcome. Plasma Exchange/Sandoglobulin Guillain-Barre Syndrome Trial Group. Annals of Neurology 44, 780–788. Hafer-Macko, C., Hsieh, S.-T., Li, C. Y., Ho, T. W., Sheikh, K. A., Cornblath, D. R., McKhann, G. M., Asbury, A. K., and GriYn, J. W. (1996). Acute motor axonal neuropathy: An antibody-mediated attack on axolemma. Annals of Neurology 40, 635–644. Hafer-Macko, C., Sheikh, K. A., Li, C. Y., Ho, T. W., Cornblath, D. R., McKhann, G. M., Asbury, A. K., and GriYn, J. W. (1996). Immune attack on the Schwann cell surface in acute inXammatory demyelinating polyneuropathy. Annals of Neurology 39, 625–635. Hall, S. M., Hughes, R. A., Atkinson, P. F., McColl, I., and Gale, A. (1992). Motor nerve biopsy in severe Guillain-Barre syndrome. Annals of Neurology 31, 441–444. Hao, Q., Saida, T., Yoshino, H., Kuroki, S., Nukina, M., and Saida, K. (1999). Anti-GalNAc-GD1a antibodyassociated Guillain-Barre syndrome with a predominantly distal weakness without cranial nerve impairment and sensory disturbance. Annals of Neurology 45, 758–768. Hartung, H.-P., Pollard, J. D., Harvey, G. K., and Toyka, K. V. (1995a). Immunopathogenesis and treatment of the Guillain-Barre syndrome–Part I. Muscle & Nerve 18, 137–153. Hartung, H.-P., Pollard, J. D., Harvey, G. K., and Toyka, K. V. (1995b). Immunopathogenesis and treatment of the Guillain-Barre syndrome–Part II. Muscle & Nerve 18, 154–164. Harukawa, H., Utsumi, H., Asano, A., and Yoshino, H. (2002). Anti-LM1 antibodies in the sera of patients with Guillain-Barre syndrome, Miller Fisher syndrome, and motor neuron disease. J Peripher. Nerv. Syst. 7, 54–58. Haymaker, W., and Kernohan, J. W. (1949). The Landry-Guillain-Barre syndrome. A clinicopathologic report of Wfty fatal cases and a critique of the literature. Medicine 28, 59–141. Hemachudha, T., Phanuphak, P., Johnson, R. T., GriYn, D. E., Ratanavongsiri, J., and Siriprasomsup, W. (1987). Neurologic complications of Semple-type rabis vaccine: Clinical and immunologic studies. Neurology 37, 550–556. Ho, T., Hsieh, S., Nachamkin, I., Willison, H., Sheikh, K., Kiehlbauch, J., Flanigan, K., McArthur, J., Cornblath, D., McKhann, G., and GriYn, J. (1997). Motor nerve terminal degeneration provides a potential mechanism for rapid recovery in acute motor axonal neuropathy after Campylobacter infection. Neurology 48, 717–724. Ho, T. W., Mishu, B., Li, C. Y., Gao, C. Y., Cornblath, D. R., GriYn, J. W., Asbury, A. K., Blaser, M. J., and McKhann, G. M. (1995). Guillain-Barre syndrome in northern China: Relationship to Campylobacter jejuni infection and anti-glycolipid antibodies. Brain 118, 597–605. Ho, T. W., Willison, H. J., Nachamkin, I., Li, C. Y., Veitch, J., Ung, H., Wang, G. R., Liu, R. C., Cornblath, D. R., Asbury, A. K., GriYn, J. W., and McKhann, G. M. (1999). AntiGD1a antibody is associated with axonal but not demyelinating forms of Guillain-Barre´ syndrome. Annals of Neurology 45, 168–173. Honavar, M., Tharakan, J. K. J., Hughes, R. A. C., Leibowitz, S., and Winer, J. B. (1991). A clinicopathological study of the Guillain-Barre syndrome. Brain 114, 1245–1269. Hughes, R., and Swan, A. (1995). Treatment of Guillain-Barre syndrome with intravenous methylprednisolone [letter; comment]. Annals of Neurology 37, 683–684. Hughes, R. A. C. (1990). ‘‘Guillain-Barre Syndrome.’’ Springer-Verlag, New York. Ilyas, A. A., Mithen, F. A., Dalakas, M. C., Chen, Z.-W., and Cook, S. D. (1992). Antibodies to acidic glycolipids in Guillain-Barre´ syndrome and chronic inXammatory demyelinating polyneuropathy. Journal of Neurological Science 107, 111–121. Ilyas, A. A., Willison, H. J., Quarles, R. H., Jungawala, F. B., Cornblath, D. R., Trapp, B. D., GriYn, D. E., GriYn, J. W., and McKhann, G. M. (1988). Serum antibodies to gangliosides in Guillain-Barre syndrome. Annals of Neurology 23, 440–447. Jacobs, B., Van Doorn, P. A., Schmitz, P. I., Tio-Gillen, A. P., Herbrink. P, Visser, L. H., Hooijkaas, H., and van der Meche, F. G. (1996). Campylobacter jejuni Infections and Anti-GM1 Antibodies in Guillain-Barre Syndrome. Annals of Neurology 40, 181–187. Jacobs, B. C., Endtz, H., van der Meche´, F. G. A., Hazenberg, M. P., Achtereekte, H. A., and Van Doorn, P. A. (1995). Serum anti-GQ1b IgG antibodies recognize surface epitopes on Campylobacter jejuni from patients with Miller Fisher syndrome. Annals of Neurology 37, 260–264.

IV. DISEASES OF MYELIN

RELEVANCE OF GBS TO ACQUIRED DEMYELINATING DISORDERS OF THE CNS

Jacobs, B. C., Rothbarth, P. H., van der Meche, F. G. A., Herbrink. P, Schmitz, P. I. M., de Klerk, M. A., and van Doorn, PA. (1998). The spectrum of antecedent infections in Guillain-Barre´ syndrome. Neurology 51, 1110–1115. Kaida, K., Kusunoki, S., Kamakura, K., Motoyoshi, K., and Kanazawa, I. (2000). Guillain-Barre syndrome with antibody to a ganglioside, N-acetylgalactosaminyl GD1a. Brain 123 (Pt 1), 116–124. Kaldor, J., and Speed, B. R. (1984). Guillain-Barre syndrome and Campylobacter jejuni: A serological study. British Medical Journal 288, 1867–1870. Kaplan, J. E., Schonberger, L. B., Hurwitz, E. S., and Katona, P. (1983). Guillain-Barre syndrome in the United States, 1978–1981: Additional observations from the national surveillance system (Letter). Neurology 33, 633–636. Kashihara, K., Shiro, Y., Koga, M., and Yuki, N. (1998). IgG Anti-GT1a antibodies which do not cross react with GQ1b ganglioside in a pharyngeal-cervical-brachial variant of Guillain-Barre´ syndrome. Journal of Neurology, Neurosurgery, and Psychiatry 65, 799. Koga, M., Yuki, N., Ariga, T., Morimatsu, M., and Hirata, K. (1998). Is IgG anti-GT1a antibody associated with pharyngeal-cervical-brachial weakness or oropharyngeal palsy in Guillain-Barre´ syndrome? Journal of Neuroimmunology 86, 74–79. Kornberg, A. J., Pestronk, A., Bieser, K., Ho, T. W., McKhann, G. M., Wu, H. S., and Jiang, Z. (1994). The clinical correlates of high-titer IgG anti-GM1 antibodies. Annals of Neurology 35, 234–237. Koski, C. L., Chou, D. K. H., and Jungalwala, F. B. (1989). Anti-peripheral nerve myelin antibodies in GuillainBarre syndrome bind a neutral glycolipid of peripheral myelin and cross-react with Forssman antigen. Journal of Clinical Investigation 84, 280–287. Koski, C. L., Humphrey, R., and Shin, M. L. (1985). Anti-peripheral myelin antibodies in patients with demyelinating neuropathy: Quantitative and kinetic determination of serum antibody by complement component 1 Wxation. Proceedings of the National Academy of Science USA 82, 905–909. Kuroki, S., Saida, T., Nukina, M., Haruta, T., Yoshioka, M., Kobayashi, Y., and Nakanishi, H. (1993). Campylobacter jejuni strains from patients with Guillain-Barre syndrome belong mostly to Penner serogroup 19 and contain beta-N-acetylglucosamine residues. Annals of Neurology 33, 243–247. Kusunoki, S., Chiba, A., Kon, K., Ando, S., Arisawa, K., Tate, A., and Kanazawa, I. (1994). N-acetylgalactosaminyl GD1a is a target molecule for serum antibody in Guillain-Barre syndrome. Annals of Neurology 35, 570–576. Kusunoki, S., Hitoshi, S., Kaida, K., Arita, M., and Kanazawa, I. (1999). MonospeciWc anti-GD1b IgG is required to induce rabbit ataxic neuropathy. Ann. Neurol 45, 400–403. Kusunoki, S., Iwamori, M., Chiba, A., et al. (1996). GM1b is a new member of antigen for serum antibody in Guillain-Barre´ syndrome. Neurology 47, 237–242. Kuwabara, S., Yuki, N., Koga, M., Hattori, T., Matsuura, D., Miyake, M., and Noda, M. (1998). IgG anti-GM1 antibody is associated with reversible conduction failure and axonal degeneration in Guillain-Barre syndrome. Annals of Neurology 44, 202–208. Landry, O. (1859). Note sur paralysie ascendante aigue. Gaz. Hebdom 472–486. Lasky, T., Terracciano, G. J., Magder, L., Koski, C. L., Ballesteros M., Nash, D., Claek, S., Haber, P., Stolley, P. D., Schonberger, L. B., and Chen, R. T. (1998). The Guillain-Barre syndrome and the 1992–1993 and 1993–1994 inXuenza vaccines. New England Journal of Medicine 339, 1797–1802. Lassmann, H. (1998). Neuropathology in multiple sclerosis: New concepts. Mult. Scler. 4, 93–98. Lastovica, A. J., Goddard, E. A., and Argent, A. C. (1997). Guillain-Barre syndrome in South Africa associated with Campylobacter jejuni O:41 strains. Journal of Infectious Disease 176 Suppl 2, S139-S143. Li, C. Y., Xue, P., Gao, C. Y., Tian, W. Q., Liu, R. C., and Yang, C. (1996). Experimental Campylobacter jejuni infection in the chicken: An animal model of axonal Guillain-Barre´ syndrome. Journal of Neurology,Neurosurgery,and Psychiatry 61, 279–284. Linton, D., Gilbert, M., Hitchen, P. G., Dell, A., Morris, H. R., Wakarchuk, W. W., Gregson, N. A., and Wren, B. W. (2000). Phase variation of a beta-1,3 galactosyltransferase involved in generation of the ganglioside GM1-like lipo-oligosaccharide of campylobacter jejuni [In Process Citation]. Molecular Microbiology 37, 501–514. Linton, D., Karlyshev, A. V., Hitchen, P. G., Morris, H. R., Dell, A., Gregson, N. A., and Wren, B. W. (2000). Multiple N-acetyl neuraminic acid synthetase (neuB) genes in Campylobacter jejuni: IdentiWcation and characterization of the gene involved in sialylation of lipo-oligosaccharide. Molecular Microbiology 35, 1120–1134. Lisak, R. P., Mitchell, M., Zweiman, B., et al. (1977). Guillain-Barre syndrome and Hodgkin’s disease: Three cases with immunological studies. Annals of Neurology 1, 72–78. Lu, J. L., Sheikh, K. A., Wu, H. S., Zhang, J., Jiang, Z. F., Cornblath, D. R., McKhann, G. M., Asbury, A. K., GriYn, J. W., and Ho, T. W. (2000). Physiological-pathological correlation in Guillain-Barre syndrome. Neurol. 54, 33–39. Lucchinetti, C., Bruck, W., Parisi, J., Scheithauer, B., Rodriguez, M., and Lassmann, H. (2000). Heterogeneity of multiple sclerosis lesions: Implications for the pathogenesis of demyelination. Ann. Neurol 47, 707–717. Lunn, M. P., Johnson, L. A., Fromholt, S. E., Itonori, S., Huang, J., Vyas, A. A., Hildreth, J. E., GriYn, J. W., Schnaar, R. L., and Sheikh, K. A. (2000). High-aYnity anti-ganglioside IgG antibodies raised in complex ganglioside knockout mice: Reexamination of GD1a immunolocalization [In Process Citation]. Journal of Neurochemistry 75, 404–412.

IV. DISEASES OF MYELIN

901

902

38. GUILLAIN-BARRE SYNDROME

Massaro, M. E., Rodriguez, E. C., Pociecha, J., Arroyo, H. A., Sacolitti, M., Taratuto, A. L., Fejerman, N., and Reisin, R. C. (1998). Nerve biopsy in children with severe Guillain-Barre syndrome and inexcitable motor nerves. Neurology 51, 394–398. McCarthy, N., Andersson, Y., Jormanainen, V., Gustavsson, O., and Giesecke, J. (1999). The risk of GuillainBarre syndrome following infection with Campylobacter jejuni. Epidemiol. Infect. 122, 15–17. McKhann, G. M., Cornblath, D. R., GriYn, J. W., Ho, T. W., Li, C. Y., Jiang, Z., Wu, H. S., Zhaori, G., Liu, Y., Jou, L. P., Liu, T. C., Gao, C. Y., Mao, J. Y., Blaser, M. J., Mishu, B., and Asbury, A. K. (1993). Acute motor axonal neuropathy: A frequent cause of acute Xaccid paralysis in China. Annals of Neurology 33, 333–342. McKhann, G. M., Cornblath, D. R., Ho, T. W., Li, C. Y., Bai, A. Y., Wu, H. S., Yei, Q. F., Zhang, W. C., Zhaori, Z., Jiang, Z., GriYn, J. W., and Asbury, A. K. (1991). Clinical and electrophysiological aspects of acute paralytic disease of children and young adults in northern China. The Lancet 338, 593–597. Mishu, B., and Blaser, M. J. (1993). The role of Campylobacter jejuni infection in the initiation of Guillain-Barre syndrome. Clinical Infectious Disease 17, 104–108. Mishu, B., Ilyas, A. A., Koski, C. L., Vriesendorp, F., Cook, S. A., Mithen, F., and Blaser, M. J. (1993). Serologic evidence of previous Campylobacter jejuni infection in patients with the Guillain-Barre syndrome. Annals of Internal Medicine 118, 947–953. Mithen, F. A., Ilyas, A. A., Birchem, R., and Cook, S. D. (1992). EVects of Guillain-Barre´ sera containing antibodies against glycolipids in cultures of rat Schwann cells and sensory neurons. Journal of Neurological Science 112, 223–232. Miyazaki, T., Kusunoki, S., Kaida, K., Shiina, M., and Kanazawa, I. (2001). Guillain-Barre syndrome associated with IgG monospeciWc to ganglioside GD1b. Neurology 56, 1227–1229. O’Hanlon, G. M., Paterson, G. J., Wilson, G., et al. (1996). Anti-GM1 ganglioside antibodies cloned from autoimmune neuropathy patients show diverse binding patterns in the rodent nervous system. J. Neuropath. Exp. Neurol. 55, 184–195. Ogino, M., Orazio, N., and Latov, N. (1995). IgG anti-GM1 antibodies from patients with acute motor neuropathy are predominantly of the IgG1 and IgG3 subclasses. Journal of Neuroimmunology 58, 77–80. Oomes, P. G., Jacobs, B. C., Hazenberg, M. P. H., BanVer, J. R. J., and van der Meche´, F. G. A. (1995). AntiGM1 IgG antibodies and Campylobacter jejuni bacteria in Guillain-Barre syndrome: Evidence of molecular mimicry. Annals of Neurology 38, 170–175. Osler, W. (1892). ‘‘The Principles and Practice of Medicine.’’ D. Appleton and Company, New York. Parkhill, J., Wren, B. W., Mungall, K., Ketley, J. M., Churcher, C., Basham, D., Chillingworth, T., Davies, R. M., Feltwell, T., Holroyd, S., Jagels, K., Karlyshev, A. V., Moule, S., Pallen, M. J., Penn, C. W., Quail, M. A., Rajandream, M. A., Rutherford, K. M., van Vliet, A. H., Whitehead, S., and Barrell, B. G. (2000). The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403, 665–668. Penner, J. L., and Hennessy, J. N. (1980). Passive hemagglutination techniques for serotyping Campylobacter fetus subsp. jejuni on the basis of soluble heat-stable antigens. Journal of Clinincal Microbiology 12, 732–737. Penner, J. L., Hennessy, J. N., and Congi, A. R. V. (1983). Serotyping of Campylobacter jejuni and Campylobacter coli on the basis of thermostable antigens. European Journal of Clinical Microbiology 2, 378–383. Plomp, J. J., Molenaar, P. C., O’Hanlon, G. M., Jacobs, B. C., Veitch, J., Daha, M. R., Van Doorn, P. A., van der Meche, F. G., Vincent, A., Morgan, B. P., and Willison, H. J. (1999a). Miller Fisher anti-GQ1b antibodies: Alpha-latrotoxin-like eVects on motor end plates [published erratum appears in Ann. Neurol. 1999;45:823]. Annals of Neurology 45, 189–199. Plomp, J. J., Molenaar, P. C., O’Hanlon, G. M., Jacobs, B. C., Veitch, J., Daha, M. R., vanDoorn, P. A., van der Meche, F. G. A., Vincent, A., Morgan, B. P., and Willison, H. J. (1999b). Miller Fisher Anti-GQ1b Antibodies: a-Latrotoxin-Like EVects on Motor End Plates. Annals of Neurology 45, 189–199. Pollard, J. D., Westland, K. W., Harvey, G. K., et al. (1995). Activated T cells of nonneural speciWcity open the blood-nerve barrier to circulating antibody. Annals of Neurology 37, 467–475. Prendergast, M. M., Lastovica, A. J., and Moran, A. P. (1998). Lipopolysaccharides from Campylobacter jejuni O:41 strains associated with Guillain-Barre syndrome exhibit mimicry of GM1 ganglioside. Infection and Immunology 66, 3649–3655. Prineas, J. W. (1981). Pathology of the Guillain-Barre syndrome. Annals of Neurology 9(Suppl), 6–19. Ravn, H. (1967). The Landry-Guillain-Barre syndrome. Acta Neurologica Scandinavica 30, 8–64. Rees, J. H., Gregson, N. A., and Hughes, R. A. C. (1995). Anti-ganglioside GM1 antibodies in Guillain-Barre syndrome and their relationship to Campylobacter jejuni infection. Annals of Neurology 38, 809–816. Rees, J. H., Soudain, S. E., Gregson, N. A., and Hughes, R. A. (1995). Campylobacter jejuni infection and Guillain-Barre´ syndrome. New England Journal of Medicine 333, 1374–1379. Reisin, R. C., Cersosimo, R., Garcia Alvarez, M., Massaro, M. E., and Fejerman, N. (93 A. D.). Acute ‘‘axonal’’ Guillain-barre syndrome in childhood. Muscle & Nerve 16, 1310–1316. Reisin, R. C., Pociecha, J., Rodriguez, E., Massaro, M. E., Arroyo, H. A., and Fejerman, N. Severe GuillainBarre syndrome in childhood treated with human immune globulin. Pediatric. Neurology 1996.May. 14, 308–312. Roberts, M., Willison, H., Vincent, A., and Newsom-Davis, J. (1994). Serum factor in Miller-Fisher variant of Guillain-Barre syndrome and neurotransmitter release. The Lancet 343, 454–455. Ropper, A. H., Wijdicks, E. F. M., and Truax, B. T. (1991). ‘‘Guillain-Barre Syndrome.’’ F. A. Davis Company, Philadelphia. Ross, J., and Bury, J. S. (1893). ‘‘A Treatise on Peripheral Neuritis.’’ Charles GriYn and Company, London.

IV. DISEASES OF MYELIN

RELEVANCE OF GBS TO ACQUIRED DEMYELINATING DISORDERS OF THE CNS

Saida, K., Saida, T., Brown, M. J., and Silberberg, D. H. (1979). In vivo demyelination induced by intraneural injection of antigalactocerebroside serum. American Journal of Pathology 95, 99–116. Saida, T., Saida, K., Lisak, R. P., Brown, M. J., Silberberg, D. H., and Asbury, A. K. (1982). In vivo demyelinating activity of sera from patients with Guillain-Barre syndrome. Annals of Neurology 11, 69–75. Salloway, S., Mermel, L. A., Seamans, M., Aspinall, G. O., Nam Shin, J. E., Kurjanczyk, L. A., and Penner, J. L. (1996). Miller-Fisher syndrome associated with Campylobacter jejuni bearing lipopolysaccharide molecules that mimic human ganglioside GD3. Infection and Immunology 64, 2945–2949. Sanders, F. K. (1948). The thickness of myelin sheaths of normal and regenerating peripheral nerve Wbers. Proceedings of the National Academy of Science,London 135, 323–357. Sawant-Mane, S., Clark, M. B., and Koski, C. L. (1991). In vitro demyelination by serum antibody from patients with Guillain-Barre syndrome requires terminal complement complexes. Annals of Neurology 29, 397–404. Sawant-Mane, S., Estep, A. 3., and Koski, C. L. (1994). Antibody of patients with Guillain-Barre syndrome mediates complement-dependent cytolysis of rat Schwann cells: Susceptibility to cytolysis reXects Schwann cell phenotype. Journal of Neuroimmunology 49, 145–152. Schonberger, L. B., Bregman, D. J., Sullivan-Bolyai, J. Z., Keenlyside, R. A., Ziegler, D. W., Retailliau, H. F., Eddins, D. L., and Bryan, J. A. (1979). Guillain-Barre syndrome following vaccination in the national inXuenza immunization program, United States, 1976–1977. American Journal of Epidemiology 110, 105–123. Sharief, M. K., Ingram, D. A., and Swash, M. (1997). Circulating tumor necrosis factor-alpha correlates with electrodiagnostic abnormalities in Guillain-Barre syndrome. Annals of Neurology 42, 68–73. Sharief, M. K., McLean, B., and Thompson, E. J. (1993). Elevated serum levels of tumor necrosis factor-alpha in Guillain-Barre syndrome. Annals of Neurology 33, 591–596. Sheikh, K. A., Deerinck, T. J., Ellisman, M. H., and GriYn, J. W. (1999). The distribution of ganglioside-like moieties in peripheral nerves. Brain 122 (Pt 3), 449–460. Sheikh, K. A., Nachamkin, I., Ho, T. W., Willison, H. J., Veitch, J., Ung, H., Nicholson, M., Li, C. Y., Wu, H. S., Shen, B. Q., Cornblath, D. R., Asbury, A. K., McKhann, G. M., and GriYn, J. W. (1998). Campylobacter jejuni lipopolysaccharides in Guillain-Barre syndrome: Molecular mimicry and host susceptibility. Neurology 51, 371–378. Sheikh, K. A., Sun, J. L. Y., Kawai, H., Crawford, T. O., Proia, R. L. G. J. W., and Schnaar, R. L. (1999). Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proceedings of the National Academy of Science USA 96, 7532–7537. Siebert, P. D., and Larrick, J. W. (1992). Competitive PCR. Nature 359, 557–558. Sivieri, S., Ferrarini, A. M., Lolli, F., Mata, S., Pinto, F., Tavolato, B., and Gallo, P. (1997). Cytokine pattern in the cerebrospinal Xuid from patients with GBS and CIDP. Journal of Neurological Science 147, 93–95. Speed, B., Kaldor, J., and Cavanagh, P. (1984). Guillain-Barre syndrome associated with Campylobacter jejuni enteritis. Journal of Infectious Disease 8, 85–86. Spies, J. M., Pollard, J. D., Bonner, J. G., Westland, K. W., and McLeod, J. G. (1995). Synergy between antibody and P2-reactive T cells in experimental allergic neuritis. Journal of Neuroimmunology 57, 77–84. Spies, J. M., Westland, K. W., Bonner, J. G., and Pollard, J. D. (1995). Intraneural activated T cells cause focal breakdown of the blood-nerve barrier. Brain 118, 857–868. Tauxe, R. V. (1992). Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations. In ‘‘Campylobacter jejuni: Current Status and Future Trends’’ (I. Nachamkin, M. J. Blaser, and L. S. Tompkins, Eds.), pp. 9–19. American Society for Microbiology, Washington, D. C. Trapp, B. D., Bo, L., Mork, S., and Chang, A. (1999). Pathogenesis of tissue injury in MS lesions. Journal of Neuroimmunology 98, 49–56. van der Meche, F. G. A., Meulstee, J., Vermeulen, M., and Kievit, A. (1988). Patterns of conduction failure in the Guillain-Barre syndrome. Brain 111, 405–416. van der Meche´, F. G. A., Schmitz, P. I. M., and Dutch Guillain-Barre Study Group. (1992a). A randomized trial comparing intravenous immune globulin and plasma exchange in Guillain-Barre syndrome. New England Journal of Medicine 326, 1123–1129. van der Meche´, F. G. A., Schmitz, P. I. M., Meulstee, J., and Oomes, P. G. (1992b). Prognostic factors in the Dutch Guillain-Barre study. Journal of Neurology 239 (suppl 2), S52. Abstract. Visser, L. H., van der Meche´, F. G. A., Meulstee, J., Rothbarth, P. Ph., Jacobs, B. C., Schmitz, P. I. M., Van Doorn, P. A., and Dutch Guillain-Barre´ Study Group (1996). Cytomegalovirus infection and Guillain-Barre´ syndrome: The clinical, electrophysiologic, and prognostic features. Neurology 47, 668–673. Visser, L. H., van der Meche´, F. G. A., Van Doorn, P. A., et al. (1995). Guillain-Barre´ syndrome without sensory loss (acute motor neuropathy): A subgroup with speciWc clinical, electrodiagnostic and laboratory features. Brain 118, 841–847. Vriesendorp, F. J., Mishu, B., Blaser, M., and Koski, C. L. (1993). Serum antibodies to GM1, peripheral nerve myelin, and Campylobacter jejuni in patients with Guillain-Barre syndrome and controls: Correlation and prognosis. Annals of Neurology 34, 130–135. Walsh, F. S., Cronin, M., Koblar, S., Doherty, P., Winer, J., Leon, A., and Hughes, R. A. C. (1991). Association between glycoconjugate antibodies and Campylobacter infection in patients with Guillain-Barre syndrome. Journal of Neuroimmunology 34, 43–51. Willison, H. J., Almemar, A., Veitch, J., and Thrush, D. (1994). Acute ataxic neuropathy with cross-reactive antibodies to GD1b and GD3 gangliosides. Neurology 44, 2395–2397. Willison, H. J., and Veitch, J. (1994). Immunoglobulin subclass distribution and binding characteristics of antiGQ1b antibodies in Miller Fisher syndrome. Journal of Neuroimmunology 50, 159–165.

IV. DISEASES OF MYELIN

903

904

38. GUILLAIN-BARRE SYNDROME

Willison, H. J., Veitch, J., Patterson, G., and Kennedy, P. G. E. (1993). Miller Fisher syndrome is associated with serum antibodies to GQ1b ganglioside. Journal of Neurology,Neurosurgery,and Psychiatry 56, 204–206. Yako, K., Kusunoki, S., and Kanazawa, I. (1999). Serum antibody against a peripheral nerve myelin ganglioside, LM1, in Guillain-Barre syndrome. J Neurol Sci. 168, 85–89. Yoshino, H. (1997). Distribution of gangliosides in the nervous tissues recognized by axonal form of GuillainBarre syndrome (in Japanese). Neuroimmunology 5, 174–175. Yuki, N., Ang, C. W., Koga, M., Jacobs, B. C., Van Doorn, P. A., Hirata, K., and van der Meche, F. G. (2000). Clinical features and response to treatment in Guillain-Barre syndrome associated with antibodies to GM1b ganglioside. Annals of Neurology 47, 314–321. Yuki, N., Handa, S., Taki, T., Kasama, T., Takahashi, M., and Saito, K. (1992). Cross-reactive antigen between nervous tissue and a bacterium elicits Guillain-Barre syndrome: Molecular mimicry between gangliocide GM1 and lipopolysaccharide from Penner’s serotype 19 of Campylobacter jejuni. Biomedical Research 13, 451–453. Yuki, N., Ho, T. W., Tagawa, Y., Koga, M., Li, C. Y., Hirata, K., and GriYn, J. W. (1999). Autoantibodies to GM1b and GalNAc-GD1a: Relationship to Campylobacter jejuni infection and acute motor axonal neuropathy in China. Journal of Neurological Science 164, 134–138. Yuki, N., Sato, S., Tsuji, S., Hozumi, I., and Miyatake, T. (1993a). An immunologic abnormality common to BickerstaV’s brain stem encephalitis and Fisher’s syndrome. Journal of Neurological Science 118, 83–87. Yuki, N., Sato, S., Tsuji, S., Ohsawa, T., and Miyatake, T. (1993b). Frequent presence of anti-GQ1b antibody in Fisher’s syndrome. Neurology 43, 414–417. Yuki, N., Takahashi, M., Tagawa, Y., Kashiwase, K., Tadokoro, K., and Saito, K. (1997). Association of Campylobacter jejuni Serotype with Antiganglioside Antibody in Guillain-Barre Syndrome and Fisher’s Syndrome. Annals of Neurology 42, 28–33. Yuki, N., Taki, T., Inagaki, F., et al. (1993). A bacterium lipopolysaccharide that elicits Guillain-Barre syndrome has a GM1 ganglioside-like structure. Journal of Experimental Medicine 178, 1771–1775. Yuki, N., Taki, T., Takahashi, M., Saito, K., Yoshino, H., Tai, T., Handa, S., and Miyatake, T. (1994). Molecular mimicry between GQ1b ganglioside and lipopolysaccharides of Campylobacter jejuni isolated from patients with Fisher’s syndrome. Annals of Neurology 36, 791–793. Yuki, N., Yamada, M., Koga, M., Odaka, M., Susuki, K., Tagawa, Y., Ueda, S., Kasama, T., Ohnishi, A., Hayashi, S., Takahashi, H., Kamijo, M., and Hirata, K. (2001). Animal model of axonal Guillain-Barre syndrome induced by sensitization with GM1 ganglioside. Annals of Neurology 49, 712–720. Yuki, N., Yoshino, H., Sato, S., and Miyatake, T. (1990). Acute axonal polyneuropathy associated with anti-GM1 antibodies following Campylobacter jejuni enteritis. Neurology 40, 1900–1902. Yuki, N., Yoshino, H., Sato, S., Shinozawa, K., and Miyatake, T. (1992). Severe acute axonal form of GuillainBarre syndrome associated with IgG anti-GD1a antibodies. Muscle & Nerve 15, 899–903.

IV. DISEASES OF MYELIN

C H A P T E R

39 Inherited Neuropathies: Clinical, Genetic, and Biological Features Lawrence Wrabetz, M. Laura Feltri, Kleopas A. Kleopa, and Steven S. Scherer

INTRODUCTION Neuropathy is a frequent component of numerous inherited syndromes. When it occurs in isolation, it is usually called Charcot-Marie-Tooth disease (CMT). The biology of axons and myelinating Schwann cells makes them vulnerable to the eVects of mutations in a large number of genes. We emphasize the varieties of inherited demyelinating neuropathies, their clinical phenotypes, the mutations that cause these phenotypes, and update their pathogenesis gene-by-gene (for reviews, see Berger et al., 2002b; Dyck et al., 1993a; Harding, 1995; Kleopa and Scherer, 2002; Lupski and Garcia, 2001; Wrabetz et al., 2001).

THE CLINICAL CLASSIFICATION OF INHERITED NEUROPATHIES Inherited neuropathies have been recognized since the late 1800s, when various forms were described by Charcot, Marie, Tooth, Herringham, De´je´rine, and Sottas (Dyck et al., 1993a). The dominantly inherited forms have come to be known as CMT, although an alternative designation, hereditary motor and sensory neuropathy (HMSN), has been widely used, too. CMT/HMSN was subdivided into demyelinating (CMT1) and axonal (CMT2) forms according to clinical, electrophysiological, and histological features. CMT1/HMSN I is more common and is characterized by an earlier age of onset (Wrst or second decade of life), nerve conduction velocities (NCVs) less than 38 m/s in upper limb nerves, and segmental demyelination, remyelination, and onion bulb formations in nerve biopsies. CMT2/HMSN II has a later onset, NCVs greater than 38 m/s, and biopsies mainly show loss of myelinated axons (Dyck et al., 1993a). The terms De´je´rine-Sottas syndrome (DSS), HMSN III, and CMT3 denote children who have a severe neuropathy (Dyck et al., 1993a; Gabreels-Festen, 2002; Ouvrier, 1996; Ouvrier et al., 1990; Plante-Bordeneuve and Said, 2002). Motor development is delayed before 3 years of age, sometimes extending to infancy. Motor abilities typically improve during the Wrst decade, but this may be followed by progressive weakness to the point that many aVected individuals use wheelchairs. Ventilatory failure (presumably caused by phrenic nerve involvement) can occur, even during infancy or childhood. Kyphoscoliosis, short stature, and foot deformities are common in older children. Sensory loss is profound, especially for modalities subserved by myelinated axons, to the point that some children have a severe sensory ataxia. Tendon reXexes are absent. Occasional patients have cranial nerve involvement—miosis, reduced pupillary responses to light, ptosis, facial weakness, nystagmus, and hearing loss. CSF protein may be elevated, and nerve roots may enhance by MRI. NCVs are very slow (<10 m/s), with marked temporal dispersion but no

Myelin Biology and Disorders, Volume 2

905

Copyright 2004, Elsevier (USA). All rights reserved.

906

39. INHERITED NEUROPATHIES

conduction block (Benstead et al., 1990). Nerves are often enlarged, and biopsies reveal a complete absence of Wbers containing normal/thick myelin sheaths and prominent ‘‘onion bulbs’’; edema may be present. Axons have inappropriately thin myelin sheaths for the axonal caliber and/or are segmentally demyelinated (Fig. 39.1). Historically, DSS/CMT3/ HMSN III was thought to be recessively inherited. Now it is clear that new dominant mutations in Myelin Protein Zero (MPZ), Peripheral Myelin Protein 22 kDa (PMP22), and Early Growth Response Gene 2 (EGR2) are the commonest causes (Nelis et al., 1999); homozygous recessive mutations of MPZ, PMP22, EGR2, or other genes are rarer causes (Tab. 39.1). Thus, while DSS/CMT3/HMSN III are meaningful labels of a clinical phenotype, their lack of speciWcity makes them diYcult to use in genetic classiWcations. The rare recessive forms have been placed in a separate group, ‘‘CMT4’’ (Tab. 39.1).

a

a

a a

a a

a

a

a

a

A: normal a

tom

a

a

a

a

B: HNPP a

a a

a

a

n

a n’

*

a n

C: CMT1A

a

n n

D: DSS

n

*

n

n

n’

n a n

*

n

FIGURE 39.1 The pathological features of HNPP, CMT1, and DSS. These are photomicrographs of semi-thin sections of sural nerve biopsies stained with methylene blue-azure II-basic fuchsin from a 27-year-old patient with normal Wndings (A), a 32-year-old with HNPP (B), a 46-year-old with CMT1A (C), and a 16-year-old with dominantly inherited DSS (D), previously reported (Lynch et al., 1997). Myelinated axons (a); demyelinated axons (*) and their associated Schwann cell nuclei (n); remyelinated axons (arrowheads). In B, note the tomaculum (tom). In C and D, note the supernumerary Schwann cell processes and their nuclei (n’) that form ‘‘onion bulbs’’ around demyelinated and remyelinated axons. Scale bar 10 mm.

907

THE CLINICAL CLASSIFICATION OF INHERITED NEUROPATHIES

TABLE 39.1 Genetic Classification of the Non-syndromic Inherited Neuropathies The table lists non-syndromic inherited neuropathies as they are classified by MIM (http://www.ncbi.nlm.nih.gov/ Omim/). Unless otherwise noted, the references are listed in the website (http://molgen-www.uia.ac.be/CMTMutations/DataSource/MutByGene. cfm). For syndromic neuropathies, see the website (http://www.neuro.wustl.edu/neuromuscular/time/hmsn.html). Disease (MIM)

Linkage

Affected gene

References

CMT1 (autosomal or X-linked dominant demyelinating) HNPP (MIM 162500)

17p11

PMP22

See text

CMT1A (MIM 118220)

17p11

PMP22 (MIM 601097)

See text

CMT1B (MIM 118200)

1q22-23

MPZ (MIM 159440)

See text

CMT1C (MIM 601098)

16p13.1.-12.3

LITAF/SIMPLE

See text

CMT1

10q21

EGR2 (MIM 129010)

See text

CMT1X (MIM 302800)

Xq13.1

GJB1 (MIM 304040)

See text

Intermediate CMT (I-CMT; autosomal dominant) I-CMT1

10q24.1-25.1

I-CMT2

19p12-13.2

Outside EGR2 and HSMN-R loci

(Verhoeven et al., 2001; Villanova et al., 1998) (Kennerson et al., 2001)

CMT2 (autosomal dominant axonal/neuronal) CMT2A (MIM 118210)

1p35-36

KIF1Bb (MIM 605995)

CMT2B (MIM 600882)

3q13-22

RAB7

CMT2B-not linked to 3q13-22 (no MIM)

See text (Verhoeven et al., 2003) (Auer-Grumbach et al., 2000b)

CMT2C (with vocal cord paresis) (MIM 606071)

12q23-24

Klein et al., 2003

CMT2D (MIM 601472) May be allelic to distal SMA/ HMN-V (MIM 600794)

7p14

GARS

Antonellis et al., 2003

CMT2E (MIM 162280)

8p21

NEFL

See text

CMT2F (no MIM)

7q11-21

(Ismailov et al., 2001)

CMT2G (MIM 604484) (proximal, 3q13.1 Okinawa type, HSMNP or HSMNO) CMT2-P0 (MIM 118200)

(Takashima et al., 1997; Takashima et al., 1999)

1q22-23

MPZ

See text

Severe demyelinating, autosomal dominant or recessive phenotypes (‘‘CMT3’’) Dejerine-Sottas Disease/HMSN-III 17p11 (MIM 145900) 1p22-23 Xq13.1 10q21 19q13.1 8p21 Congenital Hypomyelinating Neuropathy (CHN) (MIM 605253); see also CMT4E

10q21 17p11 1p22-23 11q22

PMP22 MPZ GJB1 (F235C) EGR2 PRX NEFL

See text See text See text See text See text See text

EGR2, PMP22 MPZ MTMR2

See text See text See text See text

Autosomal recessive demyelinating neuropathy (‘‘CMT4’’) CMT4A (MIM 214400)

8q13-q21.1

GDAP1

See text

CMT4B-1 (MIM 601382)

11q22

MTMR2 (MIM 603557)

See text

CMT4B-2 (MIM 604563)

11p15

MTMR13

CMT4C (MIM 601596)

5q32

See text (Guilbot et al., 1999; Le Guern et al., 1996) (Continues)

908

39. INHERITED NEUROPATHIES

TABLE 39.1

(Continued)

Disease (MIM)

Linkage

Affected gene

References

CMT4D (Lom) (MIM 601455)

8q24

NDRG1 (MIM 605626)

See text

CMT4F (MIM 605260)

19q13

See text

HMSN-R (Russe type) (MIM 605285)

10q23.2

PRX (periaxin) (MIM 605725) (EGR2 excluded)

CMT4E; same as AR CHN

See text

(Rogers et al., 2000)

Autosomal recessive axonal neuropathy (‘‘AR-CMT2’’) AR-CMT2A (CMT2B1) (MIM 605588)

1q21.2-1q21.3

AR-CMT2B (CMT2B2) (MIM 605589)

19q13.3

LMNA

(Leal et al., 2001) (Gabreels-Festen et al., 1991; Ouvrier et al., 1981)

Early onset AR-CMT2 Congenital AR axonal neuropathy

(Chaouch et al., 2003; De Sandre-Giovannoli et al., 2002)

5q deletion (SMA area)

(Korinthenberg et al., 1997) (Vedanarayanan et al., 1998; Wilmshurst et al., 2001)

Lethal neonatal AR-axonal neuropathy (MIM 604431) X-linked recessive CMT CMTX2 (MIM 302801)

Xp22.2

(Ionasescu et al., 1991, 1992)

CMTX3 (MIM 302802)

Xq26

(Ionasescu et al., 1991, 1992)

Cowchock syndrome (MIM 310490)

Xq13

(Cowchock et al., 1985; Fischbeck et al., 1986)

Hereditary sensory (and autonomic) neuropathies (HSN or HSAN) HSN-1 (MIM 162400)

9q22.1-22.3 (dominant)

SPTLC1

(Bejaoui et al., 2001; Dawkins et al., 2001) (Ohta et al., 1973)

HSN-2 HSN 3 (Riley-Day) (MIM 223900)

9q31 (recessive)

IKBKAP

(Anderson et al., 2001; Slaugenhaupt et al., 2001)

HSN 4 (CIPA) (MIM 256800)

1q21-q22, recessive

TRKA/NGF receptor

(Indo, 2001)

Hereditary motor neuropathies (HMN or ‘‘distal SMA’’) HMN 1 (MIM 182960) (early adulthood onset, AD or AR)

(Nelson and Amick, 1966)

HMN 2 (MIM 158590) (also SMA-IV)

12q24

HMN 5 (MIM 600794)

7p

HMN 7 (MIM 158580) (distal HMN with vocal paralysis-DHMNVP)

2q14

Distal infantile spinal muscular atrophy with diaphragm paralysis (SMARD1) (MIM 604320)

11q13.2-4

Distal hereditary motor neuropathy, Jerash type (HMNJ) (MIM 605726)

9p21.1-p12

(Irobi et al., 2000, 2002) GARS

Antonellis, et al., 2003 (McEntagart et al., 2001)

IGHMBP2 (MIM 600502) (Grohmann et al., 1999, 2001)

(Christodoulou et al., 2000)

Please note the following: HNPP is included with CMT1; dominant EGR2 mutations that cause a CMT1 phenotype are not given as separate name (such as CMT1D); recessive EGR2 mutations are not listed under CMT4; CMT4E is the same as autosomal recessive CHN. For abbreviations see text.

A CELLULAR BASIS OF CLASSIFYING INHERITED NEUROPATHIES

While these clinical, electrophysiological, and pathological features distinguish DSS from CMT1 (Gabreels-Festen et al., 1995; Ouvrier et al., 1987), the literature contains many examples of patients who are said to have CMT (often ‘‘severe CMT’’) who could have just as readily been labeled DSS (see Fig. 39.2). This confusion underscores that idea that the severity of clinical Wndings in inherited neuropathies is a continuum; discrete subtypes are the exception, not the rule. Congenital hypomyelinating neuropathy (CHN) is a term that should be reserved for infants who have hypotonic weakness at birth, caused by a severe neuropathy (GabreelsFesten, 2002; Guzzetta et al., 1982; Ouvrier, 1996; Ouvrier et al., 1990; Phillips et al., 1999; Plante-Bordeneuve and Said, 2002). Some may even have arthrogryposis caused by a prenatal onset, and swallowing or respiratory diYculties may result in death during infancy or later. NCVs are usually severely reduced (<5 m/s). Nerve biopsies reveal features that overlap with DSS, but with diVerent predominance—greater paucity of myelin and a failure of myelinating Schwann cells to progress beyond the promyelinating stage, less frequent onion bulbs consisting of Schwann cell processes, and more redundant basal lamina (‘‘basal lamina onion bulbs’’). As in DSS, new dominant mutations, as well as recessive mutations, of MPZ, PMP22, and EGR2, cause CHN (Tab. 39.1). Because of their overlapping clinical and pathological features, CHN has been considered to be a more severe subtype of DSS (Guzzetta et al., 1982; Tachi et al., 1984). However, it is likely that the most severe cases of CHN, and those with unique pathological Wndings, have diVerent genetic causes, as yet undiscovered. The traditional approach to evaluating a neuropathy is to determine whether it is a primary ‘‘axonal’’ or ‘‘demyelinating’’ disorder. If the axon (or even the neuron) suVers the primary injury, then it is an ‘‘axonal’’ neuropathy; the subsequent degeneration of myelin sheaths is Wallerian degeneration and not demyelination (GriYn and HoVman, 1993; Scherer and Salzer, 2001). Axonal neuropathies may even result in ‘‘secondary demyelination,’’ often associated with axonal atrophy (Dyck et al., 1993b). If the myelinating Schwann cells are aVected Wrst, then it is a ‘‘demyelinating’’ neuropathy, even though axons may degenerate subsequent to demyelination for reasons that are not well understood (Frei et al., 1999; Sancho et al., 1999). For peripheral neuropathies in general, it may be diYcult or even impossible, to determine whether a given neuropathy is a primary demyelinating or a primary axonal neuropathy. This is also true for inherited neuropathies, as the most widely used clinical criterion—forearm motor conduction velocities (greater or less than 38 m/s)—has turned out to be inadequate. This idea was Wrst promulgated before the era of molecular diagnosis (Harding and Thomas, 1980) and was a heuristic criterion, as it facilitated the separation of CMT1 from CMT2. It was based on a large number of genetically heterogeneous CMT patients, whose median motor NCVs ranged from normal to very low. The distribution appeared bimodal, with one peak around 20 m/s (corresponding to CMT1), and another between 40 and 60 m/s (corresponding to CMT2). Although the two groups overlapped, they appeared to intersect at 38 m/s. This imperfect distinction between CMT1 and CMT2 has been useful, but an uncritical adherence to this criterion has contributed to the misconception that some demyelinating neuropathies are ‘‘axonal,’’ exempliWed by CMT1X. Even prior to 1980, a group of patients with ‘‘intermediate’’ slowing had been identiWed (Bradley et al., 1977; Davis et al., 1977; Madrid et al., 1977); Figure 39 in Madrid et al. (1977) looks like a typical CMT1X biopsy (Sander et al., 1998).

A CELLULAR BASIS OF CLASSIFYING INHERITED NEUROPATHIES Whereas much has been learned from comparing the genotypes and phenotypes of CMT patients, cell biological approaches have provided the key concepts of their pathogenesis. The cellular expression of the mutant protein, and the analysis of genetically authentic animal models, indicate that inherited demyelinating neuropathies are caused by the eVect

909

910

39. INHERITED NEUROPATHIES G V Y T E P Q Y RW NH2 I T F 70 F GG V V S N R 30 G I D C Y M A D F C T I V D S I F 60 S D V S I R Y F V F 80 L E H W V Y E Y H A S C R G K S P C A 40 G VF W Q V H Q G F 50 S L P S C H L T V R F Y E S F K Y I I L G 130 S 0 N T D E V K D N C 90 P T L P E F D V I Y N R T S V E G E G M G N T N S K E F D H N T K V S C S P S E C 120 140 M Y Q R T H D Y F V I C F T Q L T F Y L E 100 W V N T N V H F G I D E V K V P RW P T K D GS I 150 110 R CHN/ CMT/ G Y V DSS DSS CMT HNPP ÒCMT2Ó V L Z Z Z Z Z Missense 160 A G V R Z Z Z Z I Nonsense Mutation G G R Z Z Deletion Mutation L V G V V Insertion L L 170 Z Z Deletion&Insertion L L L L Frameshift F Y V Possible Polymorphism or Recessive Mutation Z V R K G P K H L K G K E M A S L R R Q L A A Q R R L W C Y 180 D A S K 210 200 190 R L COOH G RQ T P V L Y A M L D H S R S T K A V S E K K A K G L G E S R K D K K

A

B T

A H G N

P (MPZ)

D L W Q V

V

N

C

S T S

S

S

G

V N 50

HH C F

M

G

V

30

V

40

S

S

Q

T

I W

P

I

P

W

L

I

F

L L 20 P V L V A P V L H V I 10 I Q S L L L L L M NH 2

C

P

P

S

Q

L

W

V K A Q T M I L 70 S I V I F S I L R L S E L F 80 F F C L Q F L T T 90 K G

S

E

S S P N

60

A I A S C L G L Q I I G T 100 Y F R G

M

H

P

R 120 V T Y M A V

110

A F

R R

I

R

PMP22 (PMP22)

E

W H L N

S

D Y 130 S Y G F A V I Y L A V W140 A F L P R A C D L S L G R V 150 I Y V I L L R W K G R E COOH 160

Z

Z

220

230

240

C R

F

C D V V K Y 170 S R L L P C R Q P G C R Y Y L P V 60 N T M N N S D A C T I V S P P I C Y 160 V S F E A H G D Y F L H C Y S P C D S 50 P R N L R S A S 180 F Q Y Y S L C K L V F M R F F L S E L C T V 70 P R S C H D Y A S I F P R L G S V L T W F H Y M E K V S L Extracellular S M 150 K V E S 40 E K V P F R A T I V WQ P A L V V A F A M W A I G S E W V T 190 V H 80 L V F Q V F L Q L F V I K M M F R V L M L M C I R 140 L F T V N L L F A L M S C RI F A G V V A Lipid bilayer F F R S S T SV G I T I P M N 30 P V I P I F C A C I V 200 L R V L Y N L I I S A R T N W L L T L W S 90 AV N N L V VW LW M V F I H A V R P Q G M T F V S R H A 130 G P E K V Y G S E D P C Intracellular A Q S A M S I N V 210 V P W G I V H V A 20 M Y Q H L I I R A C A R R T A Q 220 V S H Q L L K H 100 H N P Y H R H C C H R I Q W R C I R S E N K G R H G F G S G K R S P P N L 230 K K L M Q V S 240 120 V K G E P S T M S E G E 250 E L L 10 Y K Q N E I N K L L S E Q D P G L R L W S W L H 110 T P L E L Y C G K 260 P missense H D G L G P S R R L I D T G Z nonsense G A 270 T Z deletion G G S R W L Z frameshift COOH N A C S A C R D S K E insertion NH2 M 280 F

H

R

R

S

QW

I

D

C H

W R C RS R S R S 410 Y NH 2 L S T N S S 380 D A D D V I P F E F F H E G R T R 350 L N L K R N 270 L P R R R T K T D F P W D R M G T R R T P 260 C H H C C H G K L P I I 360 T I E Y Zn Zn K Zn R P R G V A R D F I G R T 390 P C H I C C H 420 H P 250 P L A Q L P T T 400 C 340 S T G F G F R Y A G H K P 370 P L D EK P P Q

Krox20 (EGR2)

COOH

Cx32 (GJB1)

FIGURE 39.2 Schematic summary of how MPZ, PMP22, EGR2, and GJB1 mutations aVect P0 (A), PMP22 (B), EGR2 (C), and Cx32 (D), respectively. The positions of the amino acids aVected by the mutations, as well as their phenotypes, are indicated in the legend. N-linked glycosylation sites are shown for P0 (amino acid 122) and PMP22 (amino acid 41). Mutations that aVect the 5’ untranslated region (GJB1), promoter (GJB1), or splice sites (MPZ and PMP22) are not depicted. The patients were classiWed according to the published information given in the references on the website (http:// molgen-www.uia.ac.be/CMTMutations/DataSource/MutByGene.cfm), but in many cases this information is insuYcient to make a deWnitive diagnosis. ModiWed from (Kleopa and Scherer, 2002), with permission of Elsevier Science.

CLINICAL AND MOLECULAR GENETICS OF INHERITED NEUROPATHIES

of mutations in genes that are expressed by myelinating Schwann cells and are Wrst manifested in the myelinating Schwann cells themselves (Chapter 48 by Wrabetz et al.; Martini, 1997; Nave, 2001). In other words, mutations have cell autonomous eVects, at least at the onset of the neuropathy. While these concepts are consistent with much of the data, there may be exceptions. For example, the clinical phenotypes of certain mutations of ‘‘myelin-related genes’’ indicate that their initial manifestations may be on axons (discussed later). Further, some genes that cause inherited demyelinating neuropathies may be expressed by both myelinating Schwann cells and neurons (MTMR2, GDAP1) and may have cell autonomous eVects in both cell types. The analysis of animal models of these apparent exceptions should prove to be informative (Chapter 48 by Wrabetz et al.). Such data will likely provide new insights into how myelinating Schwann cells and axons interact to form a fundamental vertebrate adaptation—the myelinated axon. The structure and function of myelinated axons provide many clues to understanding how mutations cause inherited demyelinating neuropathies. Because these topics are considered in detail elsewhere in this book, only a few points will be emphasized here. The myelin sheath itself can be divided into two domains, compact and noncompact myelin; each contains a nonoverlapping set of proteins. Compact myelin forms the bulk of the myelin sheath. It is largely composed of lipids, mainly cholesterol and sphingolipids, including galactocerebroside and sulfatide, and three proteins: P0, PMP22, and myelin basic protein (MBP). Noncompact myelin is found in the paranodes and incisures, and it contains connexin32 (Cx32), Cx29, and myelin-associated glycoprotein (MAG). In keeping with their uniquely specialized structure, myelinating Schwann cells express high levels of myelin-related proteins and their cognate mRNAs, including P0, PMP22, Cx32, Cx29, MAG, MBP, and periaxin (Scherer and Salzer, 2001). The maintenance of a myelinating phenotype depends on the integrity of axon-Schwann cell interactions. Axonal degeneration leads to Wallerian degeneration, in which myelin sheaths are phagocytosed and previously myelinating Schwann cells dediVerentiate, expressing markers that are more characteristic of nonmyelinating Schwann cells, such as the neural cell adhesion molecule (NCAM) and the low-aYnity nerve growth factor/ neurotrophin receptor, p75NTR (GriYn and HoVman, 1993; Scherer and Salzer, 2001). The expression of myelin-related mRNAs is down-regulated, but if axons regenerate, Schwann cells will re-ensheathe them in a manner that is highly reminiscent of development, re-express high levels of myelin-related proteins and mRNAs, and cease expressing p75NTR and NCAM. Nearly all myelin-related proteins and mRNAs are aVected in this manner (Scherer and Salzer, 2001). It is widely believed that axons determine whether a Schwann cell will diVerentiate into a myelinating or nonmyelinating phenotype (Chapter 13 by Jessen and Mirsky). While the nature of the axonal signal remains to be determined, two unrelated transcription factors, POU3f1 (also known as tst-1, SCIP, and Oct-6; a POU domain family) and Egr2 (also known as Krox20, a zinc Wnger protein), are essential for the normal development of myelinating Schwann cells (Chapter 13 by Jessen and Mirsky). Both Pou3f1- and Egr2-null mice have drastically reduced numbers of myelinated axons, presumably because myelinating Schwann cells fail to diVerentiate properly (Bermingham et al., 1996; Jaegle et al., 1996; Topilko et al., 1994). There is a surfeit of promyelinating Schwann cells—Schwann cells that associate with axons in a 1:1 manner, but lack a myelin sheath. These pathological Wndings resemble those found in some cases of DSS and CHN.

CLINICAL AND MOLECULAR GENETICS OF INHERITED NEUROPATHIES The following sections have been organized on a gene-by-gene basis. For each gene, the relevant aspects of the gene, its normal gene product, and its role in myelinating Schwann cells or neurons are mentioned to frame a discussion of how mutations aVect the protein

911

912

39. INHERITED NEUROPATHIES

and result in neuropathy. More detailed information of each gene, the biology of its normal gene product, and relevant animal models are found in other chapters. For the original references of the mutations, see the website (http://molgen-www.uia.ac.be/ CMTMutations/DataSource/MutByGene.cfm). We have focused on dysmyelinating (myelin sheaths do not form properly) and demyelinating (myelin sheaths form properly, but then breakdown) neuropathies.

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN) This is a clinically heterogenous group of disorders, spanning a wide range of disease severity, and of genetically diverse causes (Tab. 39.1).

PMP22 Deletions and Duplications Cause HNPP and CMT1A Genetic Aspects A reciprocal genetic mechanism causes the two most common inherited neuropathies, CMT1A and HNPP (Lupski and Garcia, 2001). Prior to the discovery of their genetic origins, a relationship between these two diseases had not been suspected because they are so clinically distinct. After CMT1A had been mapped to chromosome 17, it was subsequently associated with a small (1.4 megabase) internal duplication (Lupski et al., 1991; Raeymaekers et al., 1991). The observation that Pmp22 mutations cause Trembler and TremblerJ (Suter et al., 1992a; Suter et al., 1992b)—well known mouse models of dominantly inherited demyelinating neuropathies—quickly led to the discovery that PMP22 mapped to the duplicated segment of chromosome 17, as the region of synteny on mouse chromosome 11 containing Pmp22 was conserved in the region of human chromosome 17 containing the duplication (Matsunami et al., 1992; Patel et al., 1992; Timmerman et al., 1992; Valentijn et al., 1992b). HNPP was subsequently mapped to the same region of chromosome 17, and the reciprocal deletion was determined to be associated with the disease (Chance et al., 1993). Two homologous DNA sequences Xanking the PMP22 gene are the molecular basis for its deletion/duplication (Lupski and Garcia, 2001). Their high degree of homology promotes unequal crossing over during meiosis, which simultaneously generates a duplicated and a deleted allele. These homologous regions are found in humans and chimpanzees, so that deletions/duplications have arisen many times during human evolution. When de novo deletions and duplications occur, they originate in the father. Because most cases of HNPP and CMT1A are inherited from an aVected parent, their high prevalence in various ethnic groups indicates that they do not impact reproductive Wtness. For example, the prevalence of HNPP in a Finnish population was 16/100,000 (Meretoja et al., 1997), likely an underestimate, because many aVected individuals are asymptomatic. The prevalence of CMT in various populations ranges from 4.7-41/100,000 (Dyck et al., 1993a; Morocutti et al., 2002; Skre, 1974), and these estimates do not include HNPP. Clinical Features of HNPP Episodic mononeuropathies at typical sites of nerve compression are the hallmark of HNPP (Windebank, 1993). In order of frequency, they occur in the peroneal nerve at the Wbular head, the ulnar nerve at the elbow, the brachial plexus, the radial nerve at the spiral groove, and the median nerve at the wrist (Gouider et al., 1995). Other nerves may be aVected, and atypical presentations have been described (Mouton et al., 1999; Pareyson et al., 1996; Tyson et al., 1996). More than half of the patients recover completely, usually within days to months, but deWcits may persist. Ankle jerks may be absent in over onethird of the patients, and 12.5% are areXexic (Gouider et al., 1995). Some patients with HNPP develop a progressive generalized sensory-motor neuropathy, with or without episodes of pressure palsies.

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN)

During the acute episodes of pressure palsies, electrophysiological studies may show conduction block (Sellman and Mayer, 1987). In addition to focal changes at common sites of nerve entrapment (Li et al., 2002a), genetically aVected individuals have a mild, sensorymotor polyneuropathy (Mouton et al., 1999; Pareyson et al., 1996). Sensory velocities and distal latencies are diVusely slowed, especially in the upper extremities. Motor NCVs are minimally slowed, but distal motor latencies are consistently prolonged, especially at sites prone to entrapment (Li et al., 2002a). Biopsies of unpalsied nerves show focal thickenings (tomacula) caused by folding of the myelin sheath (Fig. 39.1), typically in the paranodal region, as well as segmental demyelination and remyelination (Windebank, 1993). Regions of uncompacted myelin have been noted (Jacobs and Gregory, 1991; Yoshikawa and Dyck, 1991), but this is not a speciWc Wnding (Fabrizi et al., 1999; Mendell et al., 1985). While it is plausible that severe focal demyelination (causing conduction block) followed by remyelination (restoring conduction) are the cellular alterations associated with the focal neuropathies, this remains to be directly demonstrated. Clinical Features of CMT1A Caused by PMP22 Duplications CMT1A, by deWnition, is CMT1 linked to chromosome 17, and is almost always caused by the inheritance of the PMP22 duplication (Lupski and Garcia, 2001). In large series, the age at onset ranges from 2 to 76 years, with most patients aVected by end of the second decade (Birouk et al., 1997; Harding and Thomas, 1980; Hoogendijk et al., 1994; Thomas et al., 1997). Neuropathy can be detected in CMT1A patients by age 5 (Berciano et al., 2003), and slow nerve conduction velocities are evident even earlier (discussed later). Weakness, atrophy, and sensory loss in the lower limbs, foot deformities and areXexia are invariably present in aVected patients. At a mean age of 40 years, functional disability is mild or absent in 35% of the patients, 25% are asymptomatic, 61% have diYculty walking or running but are still independent, and only one patient (less than 1%) requiring a wheelchair (Birouk et al., 1997). There is considerable variability in the degree of neurological deWcits within families, and even between identical twins, indicating that stochastic and environmental factors modulate disease severity (Garcia et al., 1995). In CMT1A patients with duplications, the NCVs are abnormally slowed, ranging from 5 to 35 m/s in forearm motor nerves, but most average around 20 m/s (Birouk et al., 1997; Kaku et al., 1993a; Lewis et al., 2000; Thomas et al., 1997). The lack of conduction block or temporal dispersion and the high correlation between the motor NCVs in diVerent nerves are all hallmarks. NCVs are slow in children, even well before the clinical onset of disease. In individual patients, the motor NCVs do not change signiWcantly even over 2 decades, whereas the motor amplitudes decrease, albeit slowly (Killian et al., 1996). Conduction velocities, like clinical disability, vary widely within families (Birouk et al., 1997; Kaku et al., 1993b), but an early age at onset and greatly reduced median NCV are associated with a more severe course (Birouk et al., 1997; Hoogendijk et al., 1994). Nevertheless, motor amplitudes and the number of motor units correlate with clinical disability, indicating that axonal loss and not conduction velocity per se, cause weakness (Berciano et al., 2000; Hoogendijk et al., 1994; Krajewski et al., 2000). Nerve biopsies from CMT1A patients with duplications evolve during the disease (Fabrizi et al., 1998; Gabreels-Festen et al., 1992; Thomas et al., 1997). Demyelination is more prevalent in children, and ‘‘hypomyelinated’’ axons (remyelinated axons with myelin sheaths that are inappropriately thin for the axonal caliber) become relatively more numerous with age (Fig. 39.1). Although some of the data were generated before genetic testing was possible, age-related axonal loss is a feature of CMT1 (Dyck et al., 1989; Dyck et al., 1974). PMP22 Mutations Cause CMT1A, HNPP, and DSS Besides gene duplications and deletions, more than 40 diVerent PMP22 mutations have been identiWed, the majority of which lie in membrane-associated domains. These mutations are predicted to cause amino acid substitutions (missense mutations), premature stops (nonsense mutations), or frameshifts, as shown schematically in Figure 39.2, along with their associated phenotypes. Of the 38 mutations depicted, 7 cause HNPP, 6 cause

913

914

39. INHERITED NEUROPATHIES

CMT1, 4 cause ‘‘severe CMT’’ (some of these could have been equally well called DSS), 17 cause DSS, and 4 are recessive or possible polymorphisms. Except for those with the HNPP phenotype, most patients with PMP22 point mutations, including those with a CMT1 phenotype, are more severely aVected than those harboring the duplication. They typically have an earlier age of onset, a more severe phenotype at any given age, more slowing of NCVs, and distinctive pathological changes in biopsies (Lupski and Garcia, 2001; Roa et al., 1993a; Valentijn et al., 1992a). Individuals who are heterozygous for one of three PMP22 mutations, Thr118Met, Arg157Trp, and Arg157Gly, have no discernible phenotype. Further, 1% of Europeans are heterozygous for the Thr118Met allele, which might be taken as evidence that it is a benign polymorphism (Nelis et al., 1997; Young et al., 2000). Homozygosity of Arg157Trp, however, causes DSS (Parman et al., 1999). Further, individuals who are compound heterozygotes—a PMP22 deletion and either Thr118Met or Arg157Gly— have a CMT1 phenotype (Numakura et al., 2000; Roa et al., 1993b). Thus, demyelinating neuropathy develops only when these mutant proteins are not paired with wild-type PMP22, but when it does occur, it is more severe than that caused by a null PMP22 allele paired with wild-type PMP22 (HNPP). This confusing situation has been termed a ‘‘recessive gain of function’’ (Lupski, 2000). How Do Loss of Function PMP22 Mutations Cause HNPP? The observation that a 1.4 megabase deletion in chromosome 17 is the common cause of HNPP immediately suggested that HNPP results from haplotype insuYciency of PMP22, although the potential contributions of other genes could not be excluded (Chance et al., 1993). This idea was soon supported by the Wnding that frameshift mutations (presumed to be loss of function mutations) within PMP22 also cause HNPP (Nicholson et al., 1994; Young et al., 1997), as well as the analysis of a genetically authentic animal model—Pmp22 heterozygous null (Pmp22þ/
) mice (Adlkofer et al., 1997a; Adlkofer et al., 1995; Adlkofer et al., 1997b; see also Chapter 48 by Wrabetz et al.). Finally, the level of PMP22 mRNA is relatively reduced in nerve biopsies from patients with HNPP, and the amount of PMP22 protein in compact myelin also appears to be reduced compared to normal nerves (Gabriel et al., 1997; Schenone et al., 1997; Vallat et al., 1996). PMP22 is a 160 amino acid protein with four putative transmembrane domains (Fig. 39.2). It is most abundant in peripheral nerve, mainly expressed by myelinating Schwann cells, and constitutes 2 to 5% of protein in compact myelin (Snipes et al., 1992; Spreyer et al., 1991; Welcher et al., 1991). Although it is also expressed in a number of cell types (Baechner et al., 1995; Parmantier et al., 1995; Welcher et al., 1991), there is no evidence that this expression plays a role in the pathogenesis of HNPP or CMT1A (Robertson et al., 2002). PMP22 is synthesized in the endoplasmic reticulum (ER), where it is glycosylated, transported to the Golgi, and then to the cell membrane (Pareek et al., 1993). PMP22 forms dimers, likely in the ER (D’Urso et al., 1999; Tobler et al., 1999). Given that reduced PMP22 expression causes HNPP, it seems paradoxical that much of it is rapidly degraded in proteasomes in normal myelinating Schwann cells (Pareek et al., 1997). The function of PMP22 is still unknown. In addition to its role in compact myelin, it was also identiWed as a gene (gas3) expressed in growth-arrested murine Wbroblasts (ManWoletti et al., 1990). Studies in vitro and in patient material further suggest that PMP22 could regulate cell proliferation (D’Urso et al., 1997; Hanemann and Mu¨ller, 1998; Zoidl et al., 1995), cell death (Brancolini et al., 1999; Fabbretti et al., 1995; Zoidl et al., 1997), and diVerentiation (Brancolini et al., 2000; Brancolini et al., 1999; Hanemann et al., 1996; Hanemann et al., 1997). These Wndings have led to alternative hypotheses regarding the role of PMP22 overexpression in the pathogenesis of CMT1A, as described later. How does a relative reduction in the amount of PMP22 produce the various features of HNPP: tomacula, uncompacted myelin, and a propensity for demyelination? One idea is that the components of compact myelin have a set stoichiometry, so that underexpression or overexpression of one component destabilizes the myelin sheath (Scherer, 1997). The evidence for this idea is still indirect. PMP22 and P0 are coordinately expressed in

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN)

developing and regenerating nerve (De Leon et al., 1991; Notterpek et al., 1999b; Snipes et al., 1992; Suter et al., 1994). In the interval of P0 overexpression that permits normal myelination, the level of PMP22 increases proportionally to that of P0, whereas those of other myelin proteins fall (Wrabetz et al., 2000). An altered PMP22/P0 ratio is associated with myelin pathology in CMT1A with duplicaton and HNPP (Vallat et al., 1996), as well as in Pmp22þ/
(Adlkofer et al., 1995; Gabriel et al., 1997) and Mpzþ/
mice (Martini et al., 1995b). PMP22 and P0 may directly interact within compact myelin itself (D’Urso et al., 1999; compare with Tobler et al., 2002), and possibly in the sphingolipid/cholesterol rafts that deliver them to the myelin sheath (Erne et al., 2002; Hasse et al., 2002). It is conceivable that reduced PMP22/P0 ratios produce a mechanical fragility in compact myelin, rendering them abnormally susceptible to compression, resulting in demyelination. Tomacula may form independently of compression, but aVect paranodes, which are crucial for saltatory conduction, resulting in a mild slowing of conduction (Chapter 4 by Scherer et al.). What Is the Gain of Function Caused by PMP22 Duplication? Increased gene dosage of PMP22 is the genetic basis for most patients with CMT1A; aVected individuals have one extra copy of PMP22 (three copies instead of two). Further, patients who are homozygous for the PMP22 duplication (four copies of PMP22) sometimes manifest a more severe phenotype (Kaku et al., 1993a; Le Guern et al., 1997). The idea that gene dosage is the essential abnormality is supported by the Wnding that the PMP22 mRNA levels are increased in CMT1A biopsies (Hanemann et al., 1994; Yoshikawa et al., 1994; compare with, Kamholz et al., 1994). The most direct evidence to date is that PMP22-immunoreactivity in nerve biopsies are higher in young or mildly aVected patients, and return to normal in parallel with axonal loss (Gabriel et al., 1997; Vallat et al., 1996). It seems unlikely that the expression of PMP22 in the Schwann cells that form ‘‘onion bulbs’’ in CMT1A patients is related to gene dosage (Haney et al., 1996; Nishimura et al., 1996), as onion bulbs are PMP22-positive in nerves from patients with missense mutations (Hanemann et al., 2000). One explanation of how PMP22 overexpression results in demyelination is that PMP22 acts as a growth-arrest gene (Hanemann and Mu¨ller, 1998). In keeping with this idea, overexpressing PMP22 in cultured rat Schwann cells decreases their proliferation (Zoidl et al., 1995). In addition, human Schwann cells cultured from sural nerve biopsies of CMT1A patients have decreased proliferation (Hanemann et al., 1998). In other studies, the overexpression of PMP22 causes Schwann cell apoptosis (Fabbretti et al., 1995; Zoidl et al., 1995; Zoidl et al., 1997), a feature that has been noted in inherited demyelinating neuropathies (Erdem et al., 1998; Fidzianska et al., 2002; Sancho et al., 2001). These observations alone, however, are not deWnitive, as Schwann cell proliferation and even apoptosis probably occur in all neuropathies (Berciano et al., 1999; Bradley and Asbury, 1972; Perkins et al., 1981; Weinberg and Spencer, 1978), perhaps serving to match the number of Schwann cells and axons (Grinspan et al., 1996; Trachtenberg and Thompson, 1996). Furthermore, there is no evidence for increased Schwann cell proliferation or apoptosis early in the pathogenesis in PMP22 overexpressing mice or Pmp22-null (Pmp22
/
) mice (Sancho et al., 2001). PMP22 overexpression may perturb Schwann cell diVerentiation. In CMT1A biopsies, myelinating Schwann cells may express markers of nonmyelinating/denervated Schwann cells, such as p75NTR and NCAM, even in young patients, when most myelin sheaths look normal (Hanemann et al., 1996; Hanemann et al., 1997). However, this may not be a speciWc response to PMP22 overexpression, since these molecular changes have also been described in other myelin mutants (Giese et al., 1992; Magyar et al., 1996; Montag et al., 1994). The PMP22 overexpressing rat (’’CMT rat’’) provides an interesting twist on this theme (Niemann et al., 2000; Sereda et al., 1996). Homozygous rats have a dysmyelinating neuropathy that is much more severe than that in typical CMT1A. Nevertheless, myelinrelated mRNAs are robustly expressed, but steady state levels of their cognate proteins are markedly reduced, although both PMP22 and P0 reach the Golgi. As in CMT1A nerves, myelinating Schwann cells co-express p75NTR. Thus, in the ‘‘CMT rat,’’ myelin gene transcription appears to be uncoupled from myelin assembly, and the normal, reciprocal

915

916

39. INHERITED NEUROPATHIES

expression of markers of myelinating versus nonmyelinating Schwann cells is altered as described in CMT1A. Even the extreme overexpression of PMP22, however, does not prevent Schwann cells from ensheathing axons and forming at least some myelin sheaths in vitro (D’Urso et al., 1997; Nobbio et al., 2001) and in vivo (Robertson et al., 2002; Sereda et al., 1996; Chapter 48 by Wrabetz et al.). Another plausible idea is that excess PMP22 arrives to and destabilizes the myelin sheath, resulting in demyelination in CMT1A (see above) (Gabriel et al., 1997; Vallat et al., 1996). This explanation, however, does not readily account for why even higher levels of PMP22 cause more severe dysmyelination/demyelination than in CMT1A, and that Schwann cells have altered phenotypes. Yet these more dramatic models of dysmyelination may not be relevant to the pathogenesis of CMT1A, in which the degree of overexpression is subtle. Most PMP22 Point Mutations Cause a Gain of Abnormal Function Most PMP22 point mutations cause phenotypes that are more severe than in duplicationrelated CMT1A or HNPP (Fig. 39.2). Since a simple loss of function of one PMP22 allele should result in HNPP, these mutations must produce a gain of abnormal function. This idea has been directly conWrmed in mice: the neuropathy in Trembler mice is more severe than in TremblerJ mice; the neuropathy in both of these mutants is much more severe than in Pmp22 heterozygous null (Pmp22þ/
) mice (Martini, 1997). The gain of abnormal function could be either a dominant negative eVect on the other wild-type PMP22 allele, or it could be a ‘‘toxic’’ eVect. The Wrst clue to diVerentiate these possibilities is the intracellular location of mutant PMP22 in Schwann cells. In cultured cells, dominant PMP22 mutant proteins are retained intracellularly, in the ER or the intermediate compartment (Brancolini et al., 2000; Brancolini et al., 1999; Colby et al., 2000; D’Urso et al., 1998; Dickson et al., 2002; Naef et al., 1997; Naef and Suter, 1999; Notterpek et al., 1997; Tobler et al., 1999). Thr118Met, encoded by one of the ‘‘recessive’’ mutations/polymorphisms (discussed earlier), in contrast, reaches the cell membrane, although not as eYciently as wild-type PMP22 (Naef and Suter, 1999), proving that this simple assay is potentially informative. In addition, the mutant proteins encoded by the Trembler (G150D) and TremblerJ alleles (L16P), aggregate abnormally in transfected cells (Tobler et al., 2002). Heterologous cells, however, may be diVerent than myelinating Schwann cells, which are highly specialized and require axonal interactions for eYcient insertion of PMP22 into their cell membrane (Notterpek et al., 1999b; Pareek et al., 1997). Thus, it is important to conWrm these Wndings in animal models and in patient material. To date, intracellular PMP22 has been found in nerves from both Pmp22Tr/þ and Pmp22TrJ/þ mice (Naef and Suter, 1999), and in rats that have been transfected with adenovirus expressing epitope-tagged versions of G150D or L16P (Colby et al., 2000). Immunostaining a nerve biopsy from a ‘‘severe CMT1’’ patient with the same mutation that is found in TremblerJ mice (L16P) showed intracellular accumulation of PMP22 (wild-type and mutant protein cannot be distinguished in patients) that colocalized with an ER marker, BiP (Hanemann et al., 2000). Biopsies from patients who had milder phenotypes showed diVerent results—L105R did not show intracellular accumulation of PMP22, and G107V showed accumulation of PMP22 in supernumerary Schwann cells of onion bulbs, but not colocalized with BiP. How can this type of intracellular retention produce gain of abnormal function? PMP22 forms dimers, so that mutant PMP22 retained in the ER or intermediate compartment could block a proportion of wild-type PMP22 from arriving to the cell membrane (a dominant-negative eVect) (Naef and Suter, 1999; Tobler et al., 2002; Tobler et al., 1999). Such an eVect could theoretically produce a phenotype somewhere between that of a heterozygous and a homozygous PMP22-null individual. A trans-dominant eVect of PMP22 mutants on the traYcking of P0 is yet another possibility. However, intracellularly retained mutant PMP22 did not alter the localization of P0 in Trembler (G150D) and TremblerJ (L16P) nerves (Naef et al., 1997), or in transfected, cultured Schwann cells (Tobler et al., 1999). Nor can dominant-negative eVects be the whole story, as Pmp22Tr/mice have a worse phenotype than Pmp22
/
mice (Adlkofer et al., 1997b). This conWrms

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN)

that Pmp22Tr is more deleterious than the null Pmp22 allele, and more important, it shows that Pmp22Tr can have its eVect without the presence of wild-type PMP22. Taken together, these results suggests some PMP22/Pmp22 alleles have a toxic gain of function deriving from an intracellular location. One possibility is that accumulation of mutant PMP22 triggers the unfolded protein response (UPR) in the ER. The UPR increases chaperone expression, attenuates protein synthesis, and increases degradation of proteins in order to eliminate abnormal proteins, which can produce cell death (reviewed in Chapter 42 by Gow; also see Mori, 2000). In Pmp22þ/þ, Pmp22Tr/þ, and Pmp22TrJ/þ nerves, however, PMP22 does not interact with BiP, thought to be a prerequisite for the UPR. Further, markers of the UPR, BiP, and CHOP (a transcription factor whose expression is typically induced in UPR) are not altered in mutant nerves (Dickson et al., 2002). Instead, wild-type, L16P (TremblerJ) and G150D (Trembler) PMP22 complex with another ER chaperone, calnexin, both in nerves and in transfected cells, suggesting that calnexin sequestration may be deleterious (Dickson et al., 2002). The second kind of toxic gain of function could result from the rapid degradation of PMP22 from the ER (Pareek et al., 1997) via the ubiquitin/proteasome pathway (Tsai et al., 2002). When proteasome function is ineYcient, aggresomes containing wild-type, L16P (TremblerJ) and G150D (Trembler) PMP22 form in transfected Schwann cells (Ryan et al., 2002), and even in Schwann cells cultured from TremblerJ mice (Notterpek et al., 1999a; Ryan et al., 2002). L16P (TremblerJ) and G150D (Trembler) mutants aggregate more in vitro than does wild-type PMP22 (Tobler et al., 2002). Thus, the aggregation or retention of PMP22 mutants may lead to ineYcient proteasome function, which could be toxic to the cell (Bence et al., 2001). However, the half-life of PMP22 mutants are only slightly increased, indicating proteasome function is only mildly impaired (Ryan et al., 2002). Moreover, such toxicity usually produces increased cell death; this is not a feature of Trembler nerves at the onset of myelination (Sancho et al., 2001). Finally, PMP22 accumulates in lysosomes in myelinating Schwann cells of Trembler and TremblerJ mice, raising the possibility that the endosomal/lysosomal degradation pathway plays a role in the pathogenesis of demyelination (Notterpek et al., 1999a; Notterpek et al., 1997; Ryan et al., 2002). Intracellular retention of mutant PMP22 (and, if wild-type PMP22 dimerizes with mutant PMP22, it could be retained too) could produce unstable myelin, its subsequent destruction and upregulation of the lysosomal pathway (Colby et al., 2000). Alternatively, unfolded luminal protein in the ER could be delivered to lysosomes more directly, without ever passing through the myelin sheath (Dickson et al., 2002; Notterpek et al., 1999a; Ryan et al., 2002). In TremblerJ nerves, there is increased degradation not only of PMP22, but also of P0, which is also synthesized in the ER, and of MBP (Notterpek et al., 1997), which is synthesized on free ribosomes and inserted locally into forming myelin sheaths (Colman et al., 1982). This degradation of myelin components, however, could be an eVect rather than a cause of demyelination.

MPZ Mutations Cause CMT1B, DSS, CHN, and a CMT2-Like Phenotype Genetic Aspects CMT1B was the Wrst CMT to be mapped (Bird et al., 1982), and when MPZ was later mapped near the CMT1B locus on chromosome 1 (Hayasaka et al., 1993c), it was the obvious candidate gene. MPZ mutations were soon found to cause CMT1B, DSS, and CHN (Hayasaka et al., 1993a, 1993b, 1993d; Warner et al., 1996). MPZ mutations account for as much as 10% of CMT not due to chromosome 17 duplication (Tab. 39.2), and as ascertainment of the range of phenotypes associated with MPZ mutations improves, this percentage may rise. Many diVerent MPZ mutations have been identiWed, including missense, nonsense, and frameshift mutations. Of the 84 mutations depicted in Figure 39.2, 38 cause CMT1, 22 cause ‘‘severe CMT,’’ 13 cause DSS or CHN, 8 have been reported to cause a CMT2-like phenotype, 2 are recessive or possible polymorphisms, and 1 causes an HNPP-like phenotype. Moreover, the phenotypes of patients labeled

917

918

39. INHERITED NEUROPATHIES

TABLE 39.2

The Proportion of Different CMT Mutations

PMP22 duplic.

PMP22 (other)

MPZ

GJB1

Other

None

Comments

52% (79)

7% (5)

3% (5)

8% (11)

2% (3)

33% (50)

1 EGR2 1 PRX 0 MTMR2 1 NEFL

44 CMT1 2 CMT2 1 DSS 1 CHN 1 HNPP

(Boerkoel et al., 2002) 153 unrelated patients with CMT1, CMT2, CMT1X, DSS, or CHN

31% (40)

5% (6)

9% (12)

11% (14)

N.D.

44% (56) (phenotypes not stated)

(Numakura et al., 2002) 128 Japanese CMT patients (relatedness not stated) 74 CMT1 40 CMT2 47 unclassiWed

58% (98)

1% (2)

2% (4)

7% (12)

0% 0 EGR2

38% (54) (phenotypes not stated)

(Mostacciuolo et al., 2001) 170 unrelated Italian patients with CMT or DSS

34% (59)

1% (2)

3% (6)

7% (12)

N.D.

55% (95) (phenotypes not stated)

(Mersiyanova et al., 2000a) 174 unrelated Russian patients with CMT 108 CMT1 32 CMT2 34 unclassified

55% (64)

1% (1)

3% (3)

10% (12)

N.D.

31% (36) (phenotypes not stated)

(Silander et al., 1998) 116 unrelated Finnish patients with CMT1 or DSS

71% (579)

(3/98)

(14/107)

(10/36)

N.D.

60% (38 families)

N.D.

2% (1 family)

16% (10 families)

N.D.

22% (14 CMT2 families)

(Ionasescu et al., 1993) 63 families (356 patients), 38 (193 patients) CMT1A, 1 (5 patients) CMT1B, 10 (110 patients) CMT1X, 14 (48 patients) CMT2

57% (43)

N.D.

N.D.

N.D.

N.D.

27% (20)

(Wise et al., 1993) 75 unrelated patients

(Nelis et al., 1996) 819 unrelated CMT1 patients; PMP22, MPZ, GJB1 mutations were looked for in some nonduplicated patients

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN)

CMT1 ranges from mildly aVected individuals to patients who could have equally well been called DSS. Genotype-Phenotype Correlations How can one make sense of so many mutations and the diversity of phenotypes they cause? The lesson from PMP22 mutations is to compare the eVects of MPZ mutations to a null allele. Deletion of one MPZ allele in humans has not been described, but the comparable mutation in mice causes a late onset, demyelinating neuropathy (Martini et al., 1995b; Shy et al., 1997; Zielasek et al., 1996). The V102 frameshift mutation, however, may be comparable to an MPZ deletion, as the protein would not function as a cell adhesion molecule because it lacks its transmembrane domain. It was discovered in two siblings with DSS who were both homozygous for this mutation, but their heterozygous parents and grandparents were asymptomatic (Sghirlanzoni et al., 1992). Further, these heterozygous individuals had normal sensory and motor amplitudes, and their NCVs were only mildly slowed (for example, the median motor NCVs ranged from 35-42m/s). Thus, given the mild phenotypes of MPZ mutations with loss of function, mutations that cause DSS, severe CMT, or even CMT, must produce a gain of function, although perhaps in conjunction with a loss of function, too. There are other mildly aVected kindreds. Four siblings in their Wfties were found to have the I99T mutation only after one of them appeared to develop chronic inXammatory demyelinating polyneuropathy (Donaghy et al., 2000). The other three siblings had mildly reduced or even normal NCVs; one had a nerve biopsy that showed some thinly myelinated axons as well as clusters of regenerated axons. Kindreds with the D35Y, S51F, F64del, or E71x also appear to have milder neuropathies than typical CMT1B kindreds; the patient with the E71x mutation has an HNPP-like phenotype (Lagueny et al., 2001). A few MPZ mutations cause a CMT2-like phenotype. The best example is the S44F mutation, as clinically aVected individuals have little or no slowing of NCVs despite clinically signiWcant axonal loss (Hanemann et al., 2001; Marrosu et al., 1998). In contrast, clinically aVected individuals with the T124M mutation have median/ulnar motor NCVs that range from ‘‘intermediate’’ slowing (25 to 40 m/s) to normal, whereas those in presymptomatic individuals tend to be normal (Chapon et al., 1999; De Jonghe et al., 1999; Misu et al., 2000; Senderek et al., 2000). Nerve biopsies from clinically aVected patients show axonal loss, clusters of regenerated axons, and some thinly myelinated axons (Hanemann et al., 2001). In spite of a late onset, many patients progress relatively rapidly to the point of using a wheelchair. In addition, abnormally reactive pupils (often the Wrst clinical Wnding), dysphagia, positive sensory phenomena (including painful lacinations), and hearing loss are common. One family with the E97V mutation has a comparable phenotype, with a late onset progressive neuropathy associated with hearing loss and pupillary abnormalities (Seeman et al., 2001). Two clinically aVected individuals have intermediate slowing of their median motor NCVs (29 to 30 m/s), whereas a clinically unaVected boy has a nearly normal NCV (45 m/s; Seeman, personal communication). Patients with the D75V mutation have a late onset neuropathy associated with hearing loss and pupillary abnormalities, and relatively normal motor NCVs; it was not stated whether their weakness progressed rapidly (Misu et al., 2000). Patients with D61G or Y119C mutations have an adult onset neuropathy (but without hearing loss or pupillary abnormalities); their NCVs were mildly slowed (median motor NCVs 39 to 43 m/s), and their biopsies showed decreased numbers of myelinated axons and clusters of regenerated axons (Senderek et al., 2000). Finally, a Q141x mutation was found in two children (11 and 16 years old) who had a mild CMT phenotype and little if any slowing of motor NCVs (Young et al., 2001). In sum, a few MPZ mutations cause a mild demyelinating neuropathy (D35Y, S51F, F64del, E71x, V102frameshift), others (D61G, I99T, Y119C) cause a late onset neuropathy that does not progress, and others (D75V, E97V, T124M) cause a complicated phenotype (rapid progression, hearing loss, or pupillary abnormalities). S44F and Q141x are the only mutations for which there is scant evidence for demyelination. The range in clinical severity and the large number of MPZ mutations gives the appearance of a continuum, so that comparisons are tenuous. In addition, intrafamilial

919

920

39. INHERITED NEUROPATHIES

variability is notable for certain mutations (e.g., G206x; (Senderek et al., 2001), but not for others (e.g., T124M; see above, or D90E; discussed later). At best, the clinical Wndings can be correlated with the clinical electrophysiology and nerve biopsy Wndings. DSS patients (described above) cannot be separated from DSS caused by mutations in other genes. Patients with ‘‘severe CMT1’’ overlap those with DSS, in terms of their clinical onset (often before age 3), clinical electrophysiology (forearm motor responses 8 to 15 m/s; becoming unobtainable with advanced distal muscle atrophy; absent sensory responses), and biopsy Wndings (severe loss of myelinated Wbers, hypertrophic changes, onion bulbs, and proliferation of basal lamina). Most MPZ mutations appear to cause a CMT1 phenotype that is more like the one caused by PMP22 missense mutations than by PMP22 duplications: in the former, the age of onset is younger, atrophy appears earlier and extends to proximal muscles, and NCVs are slower (typically 10 to 20 m/s). Nerve biopsies have revealed some possible clues. Uncompacted myelin has been reported with the S63del, E71x, R98C, R98H, C127Y mutations; other mutations have abnormally folded myelin sheaths (V32F, D61N, I62F, S78L, K96E, D109N, K130R, N131K, I135L). All of these mutations are localized in the extracellular domain, which mediates adhesion, but the phenotypes associated with both uncompacted myelin and abnormally folded myelin sheaths range from mild CMT to DSS (Gabreels-Festen et al., 1996; Tachi et al., 1997; Thomas et al., 1994). The report of Bird and colleagues (Bird et al., 1997) of a large family with a D90E mutation is particularly instructive. The youngest members were all aVected before age 10, several before age 3. Delayed walking and clumsy running were common early symptoms; absent tendon reXexes, followed by distal weakness and atrophy were early clinical Wndings. Proximal weakness and atrophy developed over time, and motor impairment outweighed sensory Wndings in most patients. The disease was insidiously progressive but did not aVect longevity. The median and ulnar motor NCVs ranged from 6 to 15 m/s, changing little over time, except that distal motor responses became unobtainable, presumably due to the loss of motor axons. An autopsy of a 92-year-old aVected woman revealed degeneration of fasciculus gracilis, which is consistent with severe axonal/neuronal loss caused by demyelination. In this family, an example of what is called ‘‘severe CMT1’’ in Figure 39.2, there was some variability in age of onset and disease severity at any given age, but the variability seems less than what has been described for CMT1A caused by PMP22 duplications. Loss and Gain of Function Effects of MPZ Mutations P0 is the most abundant protein in peripheral nerve myelin (GreenWeld et al., 1973). It is a single pass type I transmembrane protein with an immunoglobulin (Ig)-like fold in the extracellular domain, predicting a role in adhesion (Lemke and Axel, 1985). It has a basic/ positively charged, cytoplasmic domain that likely interacts with the negatively charged phospholipids of the membrane, thereby augmenting the role of MBP in holding the apposed cytoplasmic surfaces of the membranes together (Ding and Brunden, 1995; Martini et al., 1995a; Martini et al., 1995b). Crystallographic analysis of the extracellular domain suggests that four P0 molecules form a tetramer in cis, that interacts homophilically in trans with tetramers in apposing membrane (Shapiro et al., 1996) to form the intraperiod line. The analysis of mice that lack P0 (Mpz
/
mice) conWrms the importance of P0, as Schwann cells form a multilamellar spiral of membrane around axons, but the myelin does not compact properly (Giese et al., 1992). Mpzþ/
mice develop a late-onset demyelinating neuropathy, indicating that halving the amount of P0 destabilizes compact myelin (Martini et al., 1995b; Shy et al., 1997; Zielasek et al., 1996). Conversely, overexpressing P0 also causes dysmyelination, the severity of which is related to the level of transgene mRNA expression (Wrabetz et al., 2000; Yin et al., 2000). Martini (1997) has drawn a parallel between Mpzþ/
mice and mild CMT and between Mpz
/
mice and DSS. This predicts that the range of severity in these human cases could represent the interval between loss of function of one and two MPZ alleles, consistent with a dominant-negative gain of function. That is, the mutant allele would lose adhesive function and provoke varying loss of adhesive function in the partner wild-type allele.

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN)

Aggregation assays of transfected cells show that the extracellular domain of P0 can mediate homophilic intermolecular interactions (D’Urso et al., 1990; Filbin et al., 1990; Schneider-Schaulies et al., 1990). The possibility that P0 mutants have diminished adhesion has been evaluated in similar assays. Mutations in the extracellular domain that disrupt the disulWde bond (Zhang and Filbin, 1998) or N-linked glycosylation (Filbin and Tennekoon, 1993), and the I62F mutant (Matsuyama et al., 2002), as well as cytoplasmic mutants that alter either an acylation site (C182A) or a PKC phosphorylation site (RSTK, residues 227 to 230) (Gao et al., 2000; Wong and Filbin, 1996; Xu et al., 2001) all reduced adhesion, sometimes even of a co-expressed wild-type P0 (a dominant-negative eVect). Whereas the loss of adhesion alone might be expected to produce uncompacted myelin and even demyelination, as in Mpz
/
mice, what accounts for the wide range of morphological phenotypes observed in patients with MPZ mutations? These Wndings include focally-folded myelin sheaths, tomacula, axonal alterations with minimal myelin changes, and even the lack of formation of myelin. There are likely to be additional gain of function eVects, stemming either from defects of the myelin sheath itself, or from another intracellular location. Comparing diverse missense mutations of the same residue raises some possibilities. For example, the S54P, S63del, S63F, R98P, R98H, and R98S mutations result in a milder phenotype (CMT1B) than the S54C, S63C, and R98C mutations, which cause severe CMT/DSS. A substitution of a cysteine generates a thiol group, which could form aberrant disulWde bonds, improperly fold, and aVect traYcking. The location or type of mutation in the P0 intracellular domain is not obviously correlated with phenotype. For example, how Q215x causes CHN (Mandich et al., 1999; Warner et al., 1996), whereas an even longer deletion (Y181x) or a frameshift beginning after S226 causes a milder phenotype (CMT1B), remains to be determined. R228S disrupts a PKC phosphorylation site (an RSTK domain, residues 227 to 230; Brunden and Poduslo, 1987; Hilmi et al., 1995; Suzuki et al., 1990), and causes CMT1B (Xu et al., 2001). Intracellular PKC activity modulates P0 extracellular adhesion, and P0 intracellular domain has been proposed to initiate signals that regulate terminal diVerentiation of Schwann cells (Xu et al., 2001). Thus, intracellular MPZ mutations may provoke variable combinations of loss of extra- and intracellular adhesion, (loss of intraperiod line and major dense line compaction, respectively), together with loss of signaling function. These results further support that MPZ mutations act through various pathogenetic mechanisms, some mutation speciWc. Unlike PMP22, there are no naturally occurring Mpz mutant mice, in which to analyze the eVects of point mutations on traYcking. We (LW, MLF) have introduced randomly inserted copies of Mpz—either wild-type or mutated in the extracellular domain—that produce CHN or CMT1B-like neuropathies in transgenic mice. The nerve pathology is mutation-speciWc and includes either lack of myelin formation, uncompaction of myelin and tomacula, or demyelination with Schwann cell onion bulbs (Previtali et al., 2000; Wrabetz et al., 2000, 2002; Yin et al., 2000). Since these mice also contain two normal copies of Mpz, the eVect of the mutant alleles must include gain of function. Of note, some of the mutant proteins arrive to the myelin sheath (Previtali et al., 2000), as predicted by analysis of the intraperiod line in CMT1B biopsy material (Kirschner et al., 1996), whereas another (S63del) is retained intracellularly (Wrabetz et al., 2002), in concordance with a study of transfected cells (Gao et al., 2000). Thus, various gain of abnormal function mechanisms, originating from diverse intracellular locations are suYcient to phenocopy the variety seen in MPZ-related neuropathies.

GJB1/Cx32 Mutations Cause CMT1X Genetic Aspects Shortly after the initial reports of autosomal dominant kindreds of CMT, Herringham (1889) reported a family in which only the males had peripheral neuropathy, many years before Morgan’s description of X-linked inheritance. Although long under-recognized, CMT1X is the second most common form of CMT1, between 7 and 16% of CMT families (Tab. 39.2).

921

922

39. INHERITED NEUROPATHIES

Clinical Features of CMT1X CMT1X is considered to be an X-linked dominant trait because it aVects female carriers; their clinical involvement varies, likely owing to the proportion of myelinating Schwann cells that inactivate the X chromosome carrying the mutant GJB1 allele (Scherer et al., 1998). AVected women usually have a later onset (after the end of second decade) and a milder version of the same phenotype. In aVected males, the clinical onset is between 5 and 20 years (Birouk et al., 1998; Nicholson and Nash, 1993). Initial symptoms include diYculty running and frequent ankle sprains; foot drop and sensory loss in the legs develop later. Depending on the tempo of the disease, the distal weakness may progress to involve the gastrocnemius and soleus muscles, even to the point where assistive devices are required for ambulation. Weakness, atrophy, and sensory loss also develop in the hands, particularly in thenar muscles. These clinical manifestations are the result of a chronic, length-dependent neuropathy, and are nearly indistinguishable from those seen in patients with CMT1A or CMT1B, although atrophy, particularly of intrinsic hand muscles, positive sensory phenomena, and sensory loss may be more prominent in CMT1X patients (Hahn et al., 1990). Besides the pattern of inheritance, the degree of NCV slowing helps to distinguish CMT1A/B from CMT1X (Nicholson and Nash, 1993). In aVected men, motor NCVs in the arms show ‘‘intermediate’’ slowing (25 to 40 m/s)—faster than in typical CMT1A (10 to 30 m/s) but slower than normal (>50 m/sec). Slowing is even evident in presymptomatic male children, prior to axonal loss (Kuntzer et al., 2003). Some degree of slowing is usually found in female carriers, even asymptomatic ones. As compared to CMT1A, CMT1X patients have more heterogeneous NCVs and more distal axonal loss (Gutierrez et al., 2000; Tabaraud et al., 1999). The overlap in motor NCVs in CMT1X and CMT2 (discussed later) has led to the erroneous conclusion that GJB1 mutations cause CMT1X and CMT2 (Birouk et al., 1998; Boerkoel et al., 2002; Silander et al., 1998; Timmerman et al., 1996). In all of these cases, however, CMT1X should have been considered, as there was no male-to-male transmission, and the motor NCVs of aVected male patients showed intermediate slowing. In contrast to biopsies of CMT1A/B patients, there is less evidence of demyelination and remyelination and more evidence of axonal degeneration/regeneration in CMT1X (Hahn et al., 2001; Sander et al., 1998; Vital et al., 2001). Genotype-Phenotype Correlations CMT1X is caused by mutations in GJB1, the gene that encodes Cx32 (BergoVen et al., 1993). An amazing number (more than 240) and variety of mutations have been identiWed, aVecting every domain of the Cx32 protein (Fig. 39.2), the promoter and 3’ untranslated region (Chapter 24 by Scherer and Paul), as well as deletions of the entire gene. Missense mutations alone aVect 122 out of 283 amino acids, emphasizing the importance of each amino acid. Yet in spite of the large number of diVerent mutations, the clinical severity caused by GJB1 mutations appears to be relatively uniform in aVected men, including those with a deleted gene (Dubourg et al., 2001; Hahn et al., 2000; Nakagawa et al., 2001); PMP22, MPZ, and EGR2 mutations clearly cause more diverse phenotypes. Only two mutations, the 265-273 deletion (Ionasescu et al., 1996a) and F235C (Lin et al., 1999), appear to cause ‘‘severe CMT’’ and/or DSS.; several others (e.g., R22x, R22Q, V163I, A147frameshift) appear to cause a more severe form of CMT1X. N175D and T191A mutations are associated with a mild phenotype. It remains to be determined whether there is a genotype-phenotype correlation in CMT1X, as most of the kindreds that have been associated with either mild or severe phenotypes are small. Genotype-phenotype correlations are being addressed in a ‘‘CMT database’’ led by Dr. Michael Shy at Wayne State University in collaboration with Indiana University. Other associated manifestations of CMT1X have been reported. Hearing loss has been associated with the R142Q, T55R, and T191frameshift mutations (Lee et al., 2002; Stojkovic et al., 1999). This is intriguing because mutations in the genes encoding Cx43, Cx31, Cx30, and especially Cx26 are common causes of inherited hearing loss (http:// www.crg.es/deafness); why these various connexins all cause hearing loss remains to be

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN)

determined, but interactions between them are likely. Many mutations likely have subclinical electrophysiological evidence of slowed CNS conduction (Ba¨hr et al., 1999; Nicholson et al., 1998). A few mutations (W24C, M34V, A39V, T55R, T55I, M93V, E09x, R164Q, R183S, T191frameshift; (Kleopa et al., 2002; Lee et al., 2002; Marques et al., 1999; Panas et al., 1998) have been reported to have clinical Wndings (such as extensor plantar responses; abnormal MRIs) that indicate CNS involvement. Two mutations (R142W and C168Y) may predispose CMT1X patients to an acute, transient CNS-related syndrome after visiting high altitudes (Paulson et al., 2002). GJB1/Cx32 Mutations Act Primarily through Variable Loss of Function Cx32 is one of about 20 connexins in mammals (Chapter 24 by Scherer and Paul). All connexins are highly homologous, with the overall structure shown in Figure 12.2—four transmembrane domain proteins, two extracellular loops, an intracellular loop, and cytoplasmic N- and C-termini. Six connexins oligomerize to form a hemichannel, or connexon; two hemichannels from apposing membranes interact to form a gap junction, a pathway that permits diVusion of ions and small molecules across membranes (Bruzzone et al., 1996; White and Paul, 1999). Cx32 is localized to incisures and paranodal loops of myelinating Schwann cells and likely forms gap junctions between adjacent layers of the myelin sheath (Chapter 1 by Trapp and Kidd; Chapter 4 by Scherer et al.). Such a radial pathway—directly across the layers of the myelin sheath—would be advantageous as it provides a much shorter pathway (up to 1000-fold) than a circumferential route. Disruption of this radial pathway may be the reason that GJB1/Gjb1 mutations cause demyelination in humans and in mice (Anzini et al., 1997; Scherer et al., 1998). However, the pathway and the rate of 5,6-carboxyXuorescein diVusion in Gjb1
/
mice did not appear to be diVerent than in wild type mice (Balice-Gordon et al., 1998), implying that another connexin also forms functional gap junctions in PNS myelin sheaths. This connexin is probably Cx29 (its human homologue is Cx31.3), which has been recently localized to the incisures (Altevogt et al., 2002; Li et al., 2002b). Whether another connexin, such as Cx31 (Lopez-Bigas et al., 2001), is also present in incisures, remains to be determined. The similar clinical phenotypes of most CMT1X patients suggests that most GJB1 mutations cause a loss of function. This issue has been analyzed by expressing the mutants in heterologous cells (reviewed by Abrams et al., 2000). Many mutants do not form functional channels in Xenopus oocytes (G12S, R15W, R22G, R22P, L90H, V95M; P172S, E208L; Y211x) or mammalian cells (C60F, V139M, R215W). Other mutants form functional channels with altered biophysical characteristics (S26L, I30N, M34T, V35M, V38M, L56F, V63I, P87A, E102G, del111-116, R220x); two of these (S26L; M34T) maintain electrical coupling, but have reduced pore diameter such that may prevent the diVusion of second messengers like IP3, cAMP, and Ca2þ (Oh et al., 1997). These studies also demonstrate that the nature of the mutation may be important, as the R15W and H94Q mutants form normal functional channels, whereas R15Q and H94Y do not (Abrams et al., 2001). On the other hand, because several disease-related mutants (R15Q, H94Q, C217x, R238H, C280G, S281x) form fully functional channels (Abrams et al., 2001; Castro et al., 1999), these kinds of studies are limited in relation to pathogenesis of CMT1X. Other studies have used mammalian cells to study traYcking of Cx32 mutants (reviewed in Yum et al., 2002). Only one mutant is not detectably expressed in mammalian cells (e.g., 175 frameshift), but many mutants do not appear to reach the cell membrane; they appear to be retained in the ER or Golgi (G12S, R22Q, M34T, M34K, V38M, A39P, A39V, C53S, T55I, R75W, R75Q, R75P, R142W, R164Q, R164W, P172R, R183C, E186K, N205I, E208K, Y211x, C217x). Other mutants reach the cell membrane, although many of these appear to accumulate in the Golgi more than wild-type Cx32 (W3S, V13L, R15Q, S26L, M34I, M34V, V35M, V37M, C60F, V63I, M93V, del 112-117, V139M, R183H, R183S, N205S, I213V, A216W, R219C, R219H, R220x, R200G, R230C, R230L, F235C, R238H, L239I, C280G, S281x). The actual traYcking of mutants, however, may be more complex than indicated by these apparent patterns of retention (VanSlyke et al., 2000; VanSlyke and Musil, 2002).

923

924

39. INHERITED NEUROPATHIES

The preceding work leads to a few generalizations. Although diVerent mutants of the same amino acid (R75P, R75Q, R75W) exhibit similar patterns of intracellular retention, this does not appear to be generally the case (Yum et al., 2002). In contrast to the rest of the protein, missense mutants of intracellular carboxy terminus reach the cell membrane. Mutants that reach the cell membrane of mammalian cells usually form functional gap junctions in oocytes. Conversely, mutants that do not form functional gap junctions in oocytes do not reach the cell membrane in transfected mammalian cells. A few mutants have been expressed in cultured Schwann cells, and their localization agrees with that of transfected cell lines (Kleopa et al., 2002; Yum et al., 2002), but only two have been examined in myelinating Schwann cells. The 175 frameshift does not cause any discernible eVects, and, as in transfected cells, no mutant protein can be detected (Abel et al., 1999). The R142W mutation appears to be retained in the Golgi, and causes the retention of wildtype Cx32; it results in a mild demyelinating neuropathy (Scherer et al., 1999). While much has been learned by expressing Cx32 mutants in vitro, the eVects of these mutations are manifested in myelinating Schwann cells. Given their overall structural similarity, it is a paradox why Cx32 mutants that accumulate in the ER do not cause as much demyelination as PMP22 mutants, which also accumulate in the ER (discussed earlier). Unlike PMP22, inhibiting proteosomes does not cause Cx32 to accumulate in aggresomes (Kleopa et al., 2002; Notterpek et al., 1999a; VanSlyke et al., 2000). Perhaps the association between PMP22 and calnexin (Dickson et al., 2002) is important in this regard. Finally, the Wnding that PLP mutants that cause devastating demyelination in the CNS have no known eVects in myelination in the PNS may also be related to this paradox. Thus, despite their structural similarities, PMP22, Cx32, and PLP mutants have speciWc eVects in myelinating Schwann cells. CMT1X patients who are more severely aVected, or have other manifestations such as CNS Wndings or hearing loss, raise the possibility that some Cx32 mutants have an additional toxic gain of function. Because connexin monomers interact, one reasonable possibility is that some Cx32 mutants have dominant-negative aVects; this possibility has been shown experimentally (Bruzzone et al., 1994; Omori et al., 1996; Scherer et al., 1999). However, Cx32 mutants cannot normally interact with themselves, as GJB1 is subjected to X inactivation (Scherer et al., 1998). Nevertheless, Cx32 mutants could have dominantnegative eVects on other connexins expressed by Schwann cells or oligodendrocytes. Bruzzone et al. (1994) demonstrated that R142W had a dominant-negative eVect on Cx26 in oocytes, but this connexin is probably not expressed by myelinating Schwann cells (Nagaoka et al., 1999). In addition, channels formed by some Cx32 mutants could have a gain of function (Abrams et al., 2002), although the mutant with this eVect (S85C) is not associated with an unusual phenotype. Finally, expressing Cx32 ‘‘CNS mutants’’ (A39V, T55I, M93V, R164Q, and R183H) did not alter the distribution of Cx45 in stably transfected HeLa cells (Kleopa et al., 2002). It now appears that oligodendrocytes do not express Cx45 (Kruger et al., 2000); the relevant connexin may be Cx47 (Menichella et al., 2003; Odermatt, 2003).

EGR2 Mutations Cause CMT1, DSS, and CHN Egr2/Krox20 is a member of the early growth response gene family and encodes the zincWnger transcription factor Krox20. As predicted by segment-speciWc pattern of expression of Krox20 in developing hindbrain, Krox20/Egr2-null mice lack rhombomeres 3 and 5 (Schneider-Maunoury et al., 1993; Swiatek and Gridley, 1993). Unexpectedly, these mice were also found to have a severe dysmyelinating neuropathy (Topilko et al., 1994), making EGR2 an excellent candidate gene (Warner et al., 1998), even though no CMT locus had been linked to its locus (10q21-22). To date, one recessive and six dominant EGR2 mutations have been detected, associated with CMT1, DSS, or CHN (Fig. 39.2). How does Krox20 function in myelination? In Krox20/Egr2-null mice, the development of myelinating Schwann cells appears to be ‘‘arrested’’ at the promyelinating stage, in which presumptive myelinating Schwann cells have formed 1:1 relationships with axons, but not yet made a myelin sheath (Topilko et al., 1994). Like promyelinating Schwann cells

DOMINANT DYSMYELINATING AND DEMYELINATING NEUROPATHIES (CMT1, HNPP, DSS, CHN)

in normal nerves, these mutant Schwann cells express early markers such as MAG, but have not up-regulated their expression of myelin-related genes. These Wndings suggest that axons regulate the expression of Krox20; this has been directly demonstrated (Ghazvini et al., 2002; Murphy et al., 1996; Zorick et al., 1996, 1999). Thus, Krox20 may mediate the massive induction of myelin-related mRNAs that accompanies myelination (Stahl et al., 1990), a process that has been shown to require axon-Schwann cell interactions (Scherer et al., 1994). In keeping with this idea, the forced expression of EGR2 in rat Schwann cells strongly increases the expression of many myelin-related genes, including those of P0, PMP22, periaxin, MAG, Cx32, MBP, choline kinase, and stearyl CoA desaturase (Nagarajan et al., 2001). Whether these are direct or indirect eVects remains to be determined, but Krox20 can activate the rat Mpz promoter 4-fold in cotransfected, cultured rat Schwann cells (Zorick et al., 1999). In addition, the GJB1/Cx32 promoter has three putative EGR2 binding sites (see Fig. 3 in Chapter 24 by Scherer and Paul), and the D355V EGR2 mutant has reduced aYnity for the most 3’ EGR2 binding site (Musso et al., 2001). The one recessive mutation (I268N) falls in the R1 repressor domain (Svaren et al., 1996; Svaren et al., 1998); it prevents interactions with NAB co-repressors and can activate a synthetic Krox20/Egr2 target promoter in vitro 15-fold (Warner et al., 1999). Thus, this recessive mutation may produce a gain of function by altering gene expression of myelinrelated genes! In keeping with this idea, unlike Krox20/Egr2-null mice (Levi et al., 1996; Topilko et al., 1994), the three siblings who are homozygous for I268N have no brain stem or bone abnormalities and have a few myelinated axons in a sural nerve biopsy (Harati and Butler, 1985). Of the target genes that are likely to be perturbed by EGR2 overexpression in Schwann cells (Nagarajan et al., 2001), PMP22 and MPZ are the most likely to produce dysmyelination (Magyar et al., 1996; Sereda et al., 1996; Wrabetz et al., 2000). Six dominant EGR2 mutations aVect the highly conserved zinc Wnger DNA binding domains (Fig. 39.2). They likely act though gain of function, as heterozygous Krox20/ Egr2-null mice appear normal. Three (R359W, S382R/D383Y, R409W) reduce DNA binding and transactivation by in vitro assays (Warner et al., 1999), but how they function as dominants is unclear. The S382R/D383Y mutant inhibits the activation of endogenous myelin genes by wild-type EGR2 in co-transduced rat Schwann cells (Nagarajan et al., 2001), but a dominant-negative eVect was not found in cotransfection assays with a synthetic target promoter-reporter (Nagarajan et al., 2001; Warner et al., 1999). Because EGR2 is not thought to dimerize, this eVect may result from a dominant eVect on a partner of EGR2, such as a co-activator or adaptor. The sequestration of an EGR2 partner could explain the lack of a direct correlation between loss of DNA binding in vitro and the severity of the associated phenotype (Warner et al., 1999). Thus, as for PMP22 and MPZ, the pathogenesis of EGR2 mutations includes components of both loss of function (reduced transactivation) and gain of abnormal function.

Mutations in LITAF Cause CMT1C Two families with dominantly inherited, typical CMT1 were mapped to 16p13.1-12.3 (Chance et al., 1992b; Street et al., 2002). These families, along with a new one, were found to have mutations in LITAF (Street et al., 2003). LITAF (lipopolysaccharideinduced TNF-alpha factor) encodes a widely expressed transcript that is translated into a 161 amino acid protein. LITAF mRNA is expressed in sciatic nerve, but its level of expression is not altered in response to nerve injury, in contrast to other genes known to cause CMT1. The LITAF gene was originally cloned as a putative nuclear transcription factor involved in TNFa gene regulation, but subsequent studies indicated that the LITAF gene encodes a lysosomal protein (Moriwaki et al., 2001). Cytoplasmic LITAF expression is found in the epidermis basal cell layer in human skin (Street et al., 2003). The mutations associated with CMT1C may cluster, deWning a putative domain of the LITAF protein having a critical role in peripheral nerve function. Western blot analysis suggested that the T115N and W116G mutations do not alter the level of LITAF protein in peripheral blood lymphocytes (Street et al., 2003).

925

926

39. INHERITED NEUROPATHIES

RECESSIVELY INHERITED DEMYELINATING NEUROPATHIES (CMT4) This is a heterogenous group of disorders that cause demyelinating neuropathy, although one is syndromic (Tab. 39.1). The neuropathies are usually severe, meriting the clinical diagnosis of DSS or ‘‘severe CMT’’ (although using the term CMT for any recessively inherited neuropathy is an expansion of its original, narrower meaning—dominantly inherited neuropathies). Each kind is rare, and tends to be more common in certain inbred populations.

GDAP1 Mutations Are Associated with CMT4A and AR-CMT2-Like Neuropathy CMT4A was found in Tunisian families and mapped to the 8q13 locus (Ben Othmane et al., 1993a; Ben Othmane et al., 1995). Patients have a severe neuropathy that starts in the Wrst 2 years of life, causing delayed motor development, and progressing to involve proximal muscles by the end of the Wrst decade, to the point that many patients become wheelchair-dependent. Motor NCVs average 29 m/s in the arms, and nerve pathology includes hypomyelination, basal lamina onion bulbs, and loss of large myelinated axons. These individuals could be considered to have ‘‘severe CMT’’ or even DSS. These Wndings have been substantiated in three additional kindreds (Nelis et al., 2002b). Mutations have been identiWed in ganglioside-induced diVerentiation-associated protein-1 (GDAP1) in the Tunisian families (Baxter et al., 2002). Surprisingly, recessive GDAP1 mutations were found in three Spanish families manifesting a ‘‘severe CMT2’’ phenotype (AR-CMT2; Tab. 39.1). These patients also had a childhood onset, with distal weakness and atrophy, motor NCV of 41 m/s, and sural nerve biopsies showing loss of myelinated Wbers, axonal degeneration, and no signs of demyelination (Cuesta et al., 2002). Three mutations (two nonsense and one missense) were found in the Tunesian families; aVected individuals were homozygous. Three alleles (one frameshift leading to premature stop and two nonsense mutations) were found in the Spanish families; aVected individuals were compound heterozygous or homozygous. The aVected patients described by Nelis et al. (2002) were homozygous for two novel mutations and one mutation previously described in compound heterozygosity. All of these mutations would be predicted to cause partial or complete loss of function. As yet, the speciWc mutation does not predict demyelinating CMT4A versus axonal AR-CMT2 phenotype. For example, homozygous S194x produced CMT4A in one family (Baxter et al., 2002), whereas it caused AR-CMT2 in another (Nelis et al., 2002b). GDAP1 was identiWed as one of several genes whose expression was induced by ganglioside synthesis in Neuro2a neuroblastoma cells (Liu et al., 1999). It encodes a protein predicted to have two transmembrane domains and a glutathione S transferase domain, suggesting a role in antioxidant pathways or detoxiWcation. It is highly expressed in brain and at lower levels in nerve and other tissues, so that it could have cell-autonomous roles in both neurons and Schwann cells (Cuesta et al., 2002). Thus, speciWc mutations may cause cell-speciWc eVects, producing phenotypes that suggest a neuronal (AR-CMT2) or a Schwann cell (CMT4) defect. Alternatively, a partial loss of function could have a less severe phenotype, akin to the Spanish patients, rather than a complete loss of function, akin to the Tunisian patients. Finally, GDAP1 could be involved in Schwann cell/axonal interactions, such that loss of function produces abnormalities in both (Cuesta et al., 2002).

MTMR2 Mutations Cause CMT4B1 CMT4B has been described in Italian and Saudi Arabian families and causes a similar phenotype. Weakness was recognized between 2 and 4 years of age, progressing to wheelchair dependency by adulthhood (Houlden et al., 2001; Quattrone et al., 1996). Motor NCVs are uniformly slow (10 to 20 m/s in the arms), and nerve biopsies have a

RECESSIVELY INHERITED DEMYELINATING NEUROPATHIES (CMT4)

distinctive feature—irregular folding and redundant loops of myelin. CMT4B is caused by mutations in the myotubularin-related protein-2 gene, MTMR2, which belongs to the family of myotubularin-related dual speciWc phosphatases (Bolino et al., 2000). When two Tunisian families with the same phenotype, including focally folded myelin sheaths, were linked to a diVerent locus (11p15) named CMT4B2 (Ben Othmane et al., 1999), CMT4B caused by MTMR2 mutations was designated CMT4B1. MTMR2 is one of 13 members of the highly conserved, ‘myotubularin-related’ family in humans (Laporte et al., 2001). It has at least four functional domains: a GRAM domain (glucosyltransferase, Rab-like GTPase activator and myotubularin), a phosphatase domain, a SID (SET-interacting) domain, and a PDZ binding site (Laporte et al., 2001). Like other members of this family, MTMR2 is widely expressed in many tissues, but is enriched in both neurons and Schwann cells from embryonic development onward (Berger et al., 2002a; Bolino et al., 2002). The membrane phospholipid, phosphatidylinositol-3phosphate, PI(3)P, is a substrate of both MTMR2 and MTM1 in vitro (Kim et al., 2002; Laporte et al., 2002), although PI3,5P2 may be the preferred substrate of MTMR2 (Berger et al., 2002a). MTM1 and MTMR2 mutations produce muscle- and nerve-speciWc syndromes, respectively; tissue-enriched expression, diverse substrate speciWcity, or diVerential subcellular localization could explain the lack of redundant function between family members. The MTMR2 mutations identiWed thus far predict loss of function. Five homozygous loss of function mutations in unrelated CMT4B patients, consisting of three nonsense mutations in exons 9, 11, and 12; one splice site mutation leading to the skipping of exon 13 (in-frame deletion), and one complex mutation of a 8-bp deletion combined with a 2 bp insertion, in exon 14 (Bolino et al., 2000). To evaluate the frequency of MTMR2 mutations, 183 unrelated patients with a broad spectrum of CMT and related neuropathies, were screened (Bolino et al., 2000), revealing two additional heterozygous mutations in MTMR2 associated with CHN in two unrelated patients. Heterozygous relatives were normal suggesting that these patients harbored an undetected mutation in the other MTMR2 allele. Mutations have been subsequently reported by others in two unrelated families with CMT4B1, consisting of one base pair deletion mutation in exon 4 predicting a truncated protein, and missense mutations in exons 4 and 9 (Houlden et al., 2001; Nelis et al., 2002a). The majority of mutations fall in predicted functional domains. The normal role of MTMR2 in peripheral nerve is not known; the loss of MTMR2 may have cell autonomous eVects in neurons and Schwann cells. Phosphoinositides regulate intracellular membrane traYcking (Simonsen et al., 2001). PI(3)P, in particular, may regulate autophagy—a transport pathway that targets cytosolic proteins and organelles to lysosomes for hydrolase-mediated degradation (Petiot et al., 2000). A cell autonomous eVect in myelinating Schwann cells would suggest a role for MTMR2 in the regulation of autophagy of myelin-related lipids and proteins, although why oligodendrocytes are not similarly aVected is unexplained. That altered lysosomal degradation of myelin glycolipids causes PNS and CNS demyelination (the leukodystrophies) is a precedent for this idea. A cell autonomous eVect in neurons would suggest that membrane recycling in axons is crucial, but a neuronal aVect alone would not account for the pronounced demyelinating phenotype in these patients. Perhaps MTMR2 mutations perturb axon-Schwann cell interactions (Berger et al., 2002a). Cell-speciWc disruption of Mtmr2 in mice is underway to test these ideas.

MTMR13/SBF2 Mutations Cause CMT4B2 with or without Glaucoma In two large consanguineous families from Tunisia and Morocco, recessive demyelinating neuropathy associated with early onset glaucoma was mapped to 11p15, overlapping the locus for CMT4B2. Clinically, the mean age of onset was 8 years, with distal motor deWcit in all four limbs, motor NCV<20m/s and myelin outfoldings on nerve biopsy. All aVected family members exhibited congenital glaucoma progressing to loss of vision. An additional consanguineous Turkish family with four aVected members manifested a similar neuropathy, but without glaucoma.

927

928

39. INHERITED NEUROPATHIES

Two homozygous nonsense mutations in the Tunisia/Morocco families and an in frame deletion in the Turkish family were identiWed in MTMR13/SBF2, a pseudophosphatase homologue of MTMR2, predicting loss of function (Azzedine et al., 2003; Senderek et al., 2003). MTMR13/SBF2 is expressed in multiple tissues, especially in brain, spinal cord and sciatic nerve. Its function is unknown, but by analogy to MTM1 and the pseudophosphatase MTMR9, that may form a functional pair to regulate levels of PI(3)-phosphates (Wishart and Dixon, 2002), mutations in MTMR13 may impair cooperative function with MTMR2, producing a CMT4B2 demyelinating phenotype similar to CMT4B1. The presence or absence of glaucoma in CMT4B2 may be mutation-speciWc.

NDRG1 Mutations Cause CMT4D CMT4D is a syndromic neuropathy, with hearing loss and dysmorphic features. It was initially described in Gypsies from Lom, Bulgaria, but has been subsequently found in Gypsies in other countries (Kalaydjieva et al., 1996). It has an onset in childhood, with foot deformities and abnormal gait, and progresses to severe disability by the Wfth decade. Motor NCVs are severely reduced and become unobtainable after the age of 15, indicating severe axonal loss. Biopsies show demyelination, onion bulb formation, and cytoplasmic inclusions in Schwann cells (Kalaydjieva et al., 1998). CMT4D is caused by a nonsense mutation that should truncate the product of N-myc downstream-regulated gene-1 (NDRG1). Its function is unknown. Many cell types, including Schwann cells, express NDRG1 (Kalaydjieva et al., 2000; Lachat et al., 2002). It has been suggested to be associated with adherens junctions (Lachat et al., 2002), which are present in myelinating Schwann cells (Fannon et al., 1995; Young et al., 2002). Structural prediction algorithms place NDRG1 in the a/b hydrolase superfamily, but it lacks crucial residues thought to be necessary for enzymatic activity (Shaw et al., 2002). One potential clue is that its expression can be induced as part of the endoplasmic reticulum stress response (Agarwala et al., 2000) perhaps CMT4D is caused by impaired endoplasmic reticulum-associated protein degradation.

PRX Mutations Cause CMT4F A large Lebanese family originally deWned CMT4F (Delague et al., 2000). AVected members have a DSS phenotype, with delayed motor milestones, progressing to severe distal weakness, areXexia, sensory loss, and even sensory ataxia. NCVs are severely slowed (< 5 m/s) or undetectable, and biopsies show thinly myelinated axons, onion bulbs, and severe axonal loss. Mutations in the periaxin gene (PRX) were identiWed in this family (Guilbot et al., 2001) and in three unrelated cases (Boerkoel et al., 2000). The identiWcation of other mutations has broadened the spectrum of associated phenotypes, including patients who might have been considered to have more typical CMT1 (Takashima et al., 2002). Periaxin is a membrane-associated protein with a PDZ domain that is solely expressed in myelinating Schwann cells (Gillespie et al., 1994). Its localization changes during myelination—from the adaxonal membrane (nearest the axon) to the abaxonal membrane (nearest the basal lamina) in mature myelin sheaths (Scherer et al., 1995). Abaxonal periaxin interacts with the dystroglycan complex through dystrophin related protein -2 (DRP-2), thereby linking laminin-2 in the basal lamina to the actin cytoskeleton in the cytoplasm (Sherman et al., 2001). This complex may also activate signal transduction from laminin to the Schwann cell (Wrabetz and Feltri, 2001). Homozygous or compound heterozygous nonsense and frameshift mutations in PRX cause CMT4F. All mutations would be expected to cause loss of function; this was conWrmed in one patient whose sural nerve biopsy had no periaxin (Takashima et al., 2002). Myelination begins normally in prx-null mice; myelin outfolding develops after 6 weeks, followed by extensive demyelination by 6 months of age. In contrast to other CMT animal models, they develop a prominent allodynia and hyperalgesia (Gillespie et al., 2000). Such sensory abnormalities are reminiscent of CMT4F patients.

DEMYELINATING NEUROPATHIES AS PARTS OF OTHER SYNDROMES

DEMYELINATING NEUROPATHIES AS PARTS OF OTHER SYNDROMES Dominant demyelinating neuropathies associated with syndromes are uncommon. Patients who have large duplications of chromosome 17, including PMP22, are one example (Chance et al., 1992a). Loss of function mutations in PLP cause Pelizeaus-Merzbacher disease and a mild demyelinating neuropathy (Garbern et al., 1997). Dominant SOX10 mutations cause Waardenburg-Shah/Waardenburg type IV syndrome and other ‘‘neurocristopathies.’’ Moderate (CMT1-like) and severe (CHN-like) demyelinating neuropathies have been described, as well as CNS dysmyelination (Inoue et al., 1999, 2002; Pingault et al., 2000; Touraine et al., 2000). Nevertheless, a screen for SOX10 mutations in uncomplicated CMT neuropathies was unrevealing (Pingault et al., 2001). SOX10 is expressed by oligodendrocytes, neural crest, and by all stages in the Schwann cell lineage (Stolt et al., 2002). In keeping with its pattern of expression, SOX10 likely plays multiple roles in cellular diVerentiation, especially in the development of cells from the neural crest, including glial cells (Paratore et al., 2001; Southard-Smith et al., 1998). SOX10 may directly regulate the expression of MPZ and GJB1 (Bondurand et al., 2001; Peirano et al., 2000). One CMT1X mutation in the GJB1 promoter (Chapter 24 by Scherer and Paul) alters a putative SOX10 binding site and decreases the expression of the Cx32 gene in transient co-transfection assays (Bondurand et al., 2001). Demyelinating neuropathies are part of several recessive neurological syndromes, but are frequently overshadowed by other manifestations. The ‘‘classic form’’ of Refsum disease has a childhood onset (<20 yrs); the major clinical features are ataxia, retinitis pigmentosa, cardiomyopathy, deafness, ichthyosis, and a demyelinating neuropathy. It is caused by mutations in phytanic acid oxidase (Jansen et al., 1997; Mihalik et al., 1997) and is treated with a diet low in phytanic acid. Another form of this disease is caused by mutations in PEX7, which encodes a receptor required for importing proteins into the peroxisome (van den Brink et al., 2003). A more severe variant with infantile onset is caused by mutations of Peroxin-1, located on chromosome 7q21-22 (Reuber et al., 1997). Peroxin is required for importing peroxisomal matrix proteins. This variant is allelic to neonatal adrenoleukodystrophy and Zellweger’s and their clinical features are overlapping. They include mental retardation associated with leukodystrophy, deafness, hepatomegaly, and neuropathy, although it is unclear if the neuropathy is demyelinating. Both phytanic acid and very long chain fatty acids accumulate in these patients. Two leukodystrophies cause a recessively inherited demyelinating neuropathy. Metachromatic leukodystrophy (MIM 250100) is caused by mutations in Arylsulfatase A (von Figura et al., 2001). Globoid cell leukodystrophy (also known as Krabbe’s disease; MIM 245200) is caused by mutations in Galactosylceramide b-galactosidase (Wenger et al., 2001). Both of these are lysosomal enzymes that are required for sequential steps in the degradation of sulfatide > galactocerebroside > galactose þ ceramide. These glycolipids are highly enriched in oligodendrocytes and myelinating Schwann cells. In both of these disorders, the symptoms and signs referable to the CNS usually overshadow those of the PNS. At least for MLD, the severity of peripheral neuropathy seems to be related to the duration of disease, and is therefore generally inversely correlated to the severity of the CNS presentation (von Figura et al., 2001). Rarely, some patients may present with a demyelinating peripheral neuropathy with little or no evident CNS Wndings (Comabella et al., 2001; Marks et al., 1997). In both disorders, peripheral nerve alterations include segmental demyelination, thin myelin sheaths, and Schwann cell inclusions.

DOMINANT AXONAL NEUROPATHIES (CMT2) Patients with CMT2 usually present in the second decade, but one-third of patients present up to decades later (Dyck et al., 1993a). Physical Wndings are typically conWned to the legs and progress slowly, with areXexia, distal weakness, and atrophy of both anterior and

929

930

39. INHERITED NEUROPATHIES

posterior calf muscles; sensory Wndings may be relatively mild. Sural nerve biopsies show reduced numbers of large myelinated axons, clusters of regenerated axons, and few onion bulb-like structures. CMT2 is genetically heterogeneous; kindreds have been linked to seven diVerent loci (Tab. 39.1). The genetic abnormalities have been found in three, CMT2A, CMT2B, and CMT2E; these are discussed next.

KIF1B Mutations Cause CMT2A CMT2A patients have a typical phenotype (Ben Othmane et al., 1993b; Saito et al., 1997; Timmerman et al., 1996). A mutation in the gene encoding kinesin1B isoforms alpha and beta (KIF1B), molecular motors for transporting synaptic vesicles, has been identiWed in one CMT2A kindred. A Q98L mutation aVecting the highly conserved ATP binding region results in a loss of function (Zhao et al., 2001). Haplotype insuYciency can account for the dominant phenotype in humans, as mice that are heterozygous for a null Kif1B allele develop neuropathy. Kinesins are microtubule-activated ATPases—molecular ‘‘motors’’ that use microtubules as tracks. They form a large gene family, and each kinesin transports a speciWc organelle (Hirokawa, 1998). For example, in heterozygous kifb-null mice the levels of synaptic vesicle proteins are selectively decreased in peripheral axons. Which kinesin 1B isoform is important in the pathogenesis of CMT2A is unclear. As Mok et al. (2002) noted, the mutant ATP binding region is contained in both (the mutant mouse is also deWcient in both by design), and cell transfection studies aimed to rescue the speciWc loss of a or b isoform function may be Xawed. This distinction is important because the cargo is isoform speciWc; via its carboxyl PDZ domain, KIF1B a may interact not only with synaptic proteins, but also Kv1.4 potassium channels, NR2 subunits of NMDA receptors, and neuronal nitric oxide synthase. Isoform-speciWc disruption of KIF1B in transgenic mice will address this question. In any case, the functions of kinesin1B isoforms explains why CMT2A is length dependent—a relative reduction in this molecular motor leads to axonal impairment starting at the most remote part of the neuron. Because axons are the longest cells in the body, they may be particularly vulnerable to defective axonal transport.

RAB7 Mutations Cause CMT2B The CMT2B patients described in three reports appear to be clinically similar (AuerGrumbach et al., 2000a; De Jonghe et al., 1997; Elliot et al., 1997). In addition to length-dependent weakness and severe sensory loss, distal ulcerations in the feet are common, often leading to toe amputations. A similar, ulceromutilating neuropathy has been reported in an Austrian family, which did not link to the CMT2B locus (AuerGrumbach et al., 2000b), indicating further genetic heterogeneity in this phenotype. In three CMT2B families and three other unrelated CMT2B patients, two misense mutations have been described in RAB7, encoding a member of the Rab family of Ras-related GTPases, that are essential for proper intracellular membrane traYcking (Verhoeven et al., 2003). The pathogenesis of RAB7 mutations is unknown. RAB7 is widely expressed, including in motor and sensory neurons, and may function in vesicle transport between late endosomes and lysosomes, or in Golgi targeting of glycosphingolipids.

NEFL Mutations Cause CMT2E Mutations in the gene that encodes the neuroWlament light subunit (NEFL) cause CMT2E (De Jonghe et al., 2001; Mersiyanova et al., 2000b). In a large Russian kindred, clinical manifestations of CMT2E become apparent in the second or third decades, with diYculty walking, distal weakness, diminished sensation (including pain sensation), and absent reXexes in the legs, followed by slow progression (Mersiyanova et al., 2000b). In a smaller kindred, distal weakness progressed to paresis by the fourth decade (De Jonghe et al., 2001) In the Russian kindred, the median motor NCVs were mildly reduced or normal (38 to 52 m/s), whereas they were slowed well into the ‘‘demyelinating range’’ (median motor NCV

RECESSIVELY INHERITED AXONAL NEUROPATHIES (AR-CMT2)

25 to 39 m/s; ulnar motor NCV 30 to 42 m/s) in the other kindred (De Jonghe et al., 2001). Three novel mutations in Japanese patients have been reported (Yoshihara et al., 2002), but the clinical description of these patients was limited to the observation that the motor NCVs were slowed (22 to 36 m/s). Finally, Jordanova et al. (2003) found six NEFL mutations (Wve missense and one three nucleotide deletion) in 323 unrelated patients with CMT neuropathies, some with a more severe (DSS) phenotype. A sural nerve biopsy from one patient showed both myelin outfolding, loss of myelinated Wbers, and onion bulbs, as well as clusters of regenerating axons. NeuroWlaments are the main neuronal intermediate Wlaments and are composed of three subunits, termed heavy, medium, and light. In animal models, NeX mutations have profound eVects on axonal caliber (Lee et al., 1994; Ohara et al., 1993), potentially accounting for the slowed NCVs seen in CMT2E patients. CMT2E is autosomal dominant, suggesting that mutations produce either haploinsuYciency or gain of function. NeXnull mice, however, do not appear to develop a CMT2-like neuropathy (Zhu et al., 1997), whereas the L394P mutation, which is predicted to aVect the coil 2B domain and thus disrupt neuroWlament assembly, causes a severe peripheral neuropathy/neuronopathy in transgenic mice (Lee et al., 1994). Thus, missense mutants that disorganize assembly could result in neuropathy. Consistent with this idea, several missense mutations have been reported thus far, and two of these mutant genes encode proteins that disrupt assembly and axonal transport of neuroWlaments in transfected DRG neurons (Brownlees et al., 2002; Perez-Olle et al., 2002). In addition to axons, Schwann cells also express NEFL (Fabrizi et al., 1997); whether this relates to the phenotype of CMT2E remains to be determined.

RECESSIVELY INHERITED AXONAL NEUROPATHIES (AR-CMT2) Axonal neuropathy is found in many recessively inherited neurological syndromes, but is rarely recognized as an isolated inherited disease. Sporadic axonal neuropathies that are presumed to be recessively inherited typically present from birth to 5 years, and progress to severe disability by the second decade or even to death during infancy (Gabreels-Festen et al., 1991; Ouvrier et al., 1981). NCVs and pathology are consistent with axonal degeneration, thereby distinguishing these cases from inherited demyelinating neuropathies and CMT2. We have used the term ‘‘autosomal recessive CMT2’’ (AR-CMT2); ‘‘CMT2B’’ is an alternative designation used by OMIM. Both of these terms are confusing, if not misleading, designations, as the term ‘‘CMT’’ was originally used to designate dominantly inherited neuropathies. The X-linked recessive CMT types listed on Table 39.1 are probably syndromic.

MUTATIONS IN LMNA CAUSE AR-CMT2A/CMT2B1 In a large Moroccan family linked to 1q21.2-q21.3, the onset of symptoms in nine aVected siblings was in the second decade, with severe weakness and atrophy of the distal limb muscles; proximal muscles were involved in some patients. Motor NCVs were normal or slightly slowed and biopsies showed diminished numbers of myelinated axons and clusters of regenerated axons (Bouhouche et al., 1999). Two inbred Algerian families of similar phenotype were mapped to chromosome 1q21.2–21.3 and a homozygous missense mutation in LMNA was found to segregate with disease in both families, as well as an independent third family (Chaouch et al., 2003; De Sandre-Giovannoli et al., 2002). LMNA encodes four isoforms of lamins A/C, all components of the nuclear envelope, and amino acid substitutions of which have already been described in limb-girdle muscular dystrophy 1B, autosomal dominant Emery-Dreifuss muscular dystrophy, dilated cardiomyopathy type 1A, autosomal dominant partial lipodystrophy, and Hutchinson–Gilford

931

932

39. INHERITED NEUROPATHIES

progeria syndrome. Lamins A/C are expressed in most cells. The R298C substitution in Lamin A/C aVects all four isoforms and probably produces AR-CMT2A via loss of function, as Lmna-null mice develop a neuropathy similar to that of CMT2A patients (De Sandre-Giovannoli et al., 2002; Sullivan et al., 1999). In contrast, most of the other LMNA associated diseases occur in heterozygosity, suggesting that these diverse phenotypes arise via diVerent gain of function eVects.

HEREDITARY SENSORY (AND AUTONOMIC) NEUROPATHIES (HSN OR HSAN) They are grouped together for their shared characteristics—the loss of sensory (especially of small Wbers) and autonomic Wbers, resulting in sensory loss to the point of mutilating deformities of the hands and feet. They proved to be genetically heterogeneous (Tab. 39.1). HSN-1 is an autosomal dominant trait and manifests in adolescence with small Wber sensory loss, burning pain (distal>proximal and legs>arms), pedal deformity, acromutilation, and distal weakness. It is caused by mutations in the gene encoding serine palmitoyl transferase, long-chain base subunit 1 (SPTLC1) (Bejaoui et al., 2001; Dawkins et al., 2001). The mutations that cause HSN-1 (C133Y, C133W, V144D) reside in a conserved region, and the corresponding mutations in the yeast enzyme act as dominants because the enzyme is part of a heterodimer (Gable et al., 2002). HSN-2, HSN-3, and HSN-4 are autosomal recessive. HSN-2 has its onset in early childhood with similar phenotype to HSN-1 (Dyck, 1993). HSN-3, also known as the Riley-Day syndrome or familial dysautonomia with congenital indiVerence to pain, is caused by mutations of IKBKAP, the inhibitor of k light polypeptide gene enhancer in B cells, kinase complexassociated protein (Anderson et al., 2001; Slaugenhaupt et al., 2001). Onset is in infancy with absent fungiform papillae of the tongue and mainly small Wber involvement, including autonomic crises with postural hypotension and tachycardia. HSN-4 is characterized by congenital insensitivity to pain with anhydrosis (hence the alternative name, CIPA syndrome), with the associated features of small Wber sensory loss, autonomic failure, mental retardation, and acromutilation. It is caused by mutations in TRKA, which encodes a receptor tyrosine kinase for nerve growth factor (Indo, 2001).

HEREDITARY MOTOR NEUROPATHIES (HMN OR ‘‘DISTAL SMA’’) In these diseases, motor neurons are preferentially aVected; they are also known as distal spinal muscular atrophy, ‘‘distal SMA’’ (Harding and Thomas, 1980). One presumes that these diseases are caused by cell autonomous eVects of mutations in motor neurons, but their genetic basis is largely unknown (Tab. 39.1).

THE PATHOGENESIS OF INHERITED AXONAL NEUROPATHIES In many neuropathies, the clinical features tend to have a distal predilection, both in terms of Wrst appearance and in ultimate severity. This suggests that axonal length is a factor in determining which neural elements are at risk. But distal distribution does not mean that the defect necessarily lies in the axon; it could just as well represent a primary neuron cell body abnormality. For instance, large doses of pyridoxine promptly kill large primary sensory neurons, whereas smaller doses cause only subtle shrinkage of these neurons and indolent, distal axonal degeneration (Xu et al., 1989). Thus, a modest neuronal abnormality may result in distal axonopathy, but a more severe insult of the same type may cause the neuron itself to degenerate as the primary event. The inclusion of HSN-3 and HSN-4 as inherited ‘‘axonal neuropathies’’ is even more problematic, as they appear to result from the developmental degeneration of neurons.

MAKING A MOLECULAR DIAGNOSIS

In nonsyndromic neuropathies, mutations appear to act on a cell-autonomous basis, aVecting either neurons or myelinating Schwann cells. The selective vulnerability of PNS neurons that leads to neuropathy may be the axons themselves—the longest cells in the body. Their prominent cytoskeleton contains intermediate Wlaments and microtubules, and dominant mutations in NEFL, the gene encoding the light subunit, cause neuropathy. Further, recessive mutations in the gigaxonin gene, which encodes a protein that likely interacts with cytoskeleton proteins, cause giant axonal neuropathy (Bomont et al., 2000; Ding et al., 2002; Kuhlenbaumer et al., 2002). The reason for the selective vulnerability of Schwann cells (and oligodendrocytes) is less clear. Two mutations that cause dys/demyelination aVect genes that are largely if not exclusively expressed in myelinating Schwann cells (MPZ, PRX); these mutations need not aVect other cell types. However, more mutations are expressed by multiple cell types (PMP22, GJB1, EGR2, LITAF, NDRG1, MTMR2); why these mutations mainly aVect myelinating Schwann cells remains largely a mystery. Disability in inherited neuropathies, even demyelinating ones, correlates with axonal loss. This has been well documented in animal models (Chapter 48 by Wrabetz et al.) and in CMT patients (Berciano et al., 2000; Dyck et al., 1974, 1989; Krajewski et al., 2000). The lack of a myelin sheath has pronounced eVects on axonal caliber, axonal transport, and the phosphorylation and packing of neuroWlaments (Chapter 1 by Trapp and Kidd). How does a myelinating Schwann cell communicate with its axon, and how is this altered by demyelination? Possibilities include trophic support from myelinating Schwann cells, the altered myelin sheath itself (Hanemann and Gabreels-Festen, 2002), signals emanating from the adaxonal Schwann cell membrane and/or cytoplasm, especially in the paranodal region, where axons and Schwann cells are intimately joined (Chapter 4 by Scherer, et al.). The evidence for these possibilities is provided by Mag-null mice, which have axonal changes that mimic those found in demyelinated axons (Yin et al., 1998). Other contributing factors to demyelination or axonal loss include inXammatory changes initiated by demyelination (Maurer et al., 2002) and remodeling of the extracellular matrix (Misko et al., 2002; Palumbo et al., 2002). In some CMT1B mutations, particularly MPZT124M, axons appear to be disproportionately aVected, and in a few CMT1B mutations, axonal loss is detected before altered myelin sheaths (discussed earlier). While these examples are provocative, rigorous documentation that axonal alterations preceed demyelination is still lacking.

MAKING A MOLECULAR DIAGNOSIS Determining the phenotypes of patients—their age of onset, historical progression, as well as their current physical Wndings—will usually distinguish between the CHN, DSS, CMT, and HNPP phenotypes. Other distinctive features discussed earlier, such as X-linked inheritance, may suggest a speciWc type of CMT, but these may not be present. Electrophysiological studies, including the median and ulnar motor NCVs, should be performed, as the degree of slowing and its uniformity should conWrm or refute the clinical diagnosis and even suggest a speciWc subtype. In our opinion, a nerve biopsy should not be part of the initial diagnostic evaluation except in problematic cases or for research purposes. In cases of suspected HNPP, testing for the PMP22 deletion should be performed, as this is by far the major cause; if this is negative, then PMP22 exons should be sequenced. In CMT patients with uniform slowing of motor NCVs between 10 and 35 m/s, testing for the PMP22 duplication should be performed, as this is the major cause of CMT1. If the patient does not have the duplication, then PMP22, MPZ, GJB1, EGR2, and NEFL exons should be sequenced. In DSS with appropriately slowed NCVs, testing for the duplication will usually be negative, and sequencing of the PMP22, MPZ, GJB1, EGR2, PRX, and NEFL genes should be included in the initial evaluation. In cases where CMT1X is the most likely diagnosis (intermediate slowing of NCVs, no male-to-male transmission), sequencing GJB1 alone is the appropriate initial test; if negative, consider sequencing MPZ and

933

934

39. INHERITED NEUROPATHIES

NEFL. In cases of suspected CMT2, it is probably premature to test for MPZ and NEFL mutations alone (the only commercially available tests), but a more comprehensive battery should be available in the future as more genetic causes are found (Fig. 39.3). Discovering these causes depends on the participation of referring physicians and aVected families.

IDENTIFICATION OF OTHER CMT GENES As summarized in Table 39.2, there are many reports regarding the proportion of diVerent mutations that cause CMT. As noted in the table, these reports diVer in the kinds of patients who were analyzed and, depending on the year of publication, whether other genes were sequenced. If one considers CMT overall, PMP22 duplications accord for about onehalf of all kindreds; the proportion goes up markedly if only CMT1 patients are considered, and even more if one only considers families with dominant inheritance. Mutations of GJB1, MPZ, and PMP22 account for 10 to 25% of the other cases of CMT, mutations of EGR2, PRX, MTMR2, and NEFL are even rarer. A surprising Wnding is that no mutations have yet been identiWed for one-third of CMT kindreds. This issue was most elegantly illustrated in the analysis of Boerkoel et al. (2002), who did not identify mutations in many CMT1 kindreds. How will we Wnd these other CMT genes? AVected families that have no known cause of their CMT and that are large enough for linkage analysis are increasingly less common. Owing to automated sequencing, it is now possible to screen ‘‘candidate genes’’ as causes of CMT. To date, most of these candidates are derived from advances in the genetics and biology of myelinated axons, as described elsewhere in this book. Transcriptome analysis (Nagarajan et al., 2002; Nagarajan et al., 2001) and proteomics will provide even more candidates. Their sequence immediately ties them to a physical map that can be compared to mapped CMT loci.

WORK UP FOR INHERITED NEUROPATHIES CMT

1. Clinical evaluation

2. Forearm motor NCV slowing

3. Genetic testing

CMT1

CMT1X

CMT2

HNPP

severe (<10 m/s)

moderate (10-35 m/s)

intermediate (25-40 m/s)

mild (>40m/s) /normal

focal and/or mild

sequence: PMP22 MPZ GJB1/Cx32 EGR2 PRX NEFL MTMR2* MTMR13*

PMP22 duplication

sequence: GJB1/Cx32

CMT2 sequence: NEFL MPZ KIFIB RAB7*

PMP22 deletion

DSS/CHN

−

−

*not commercially available

sequence: PMP22

FIGURE 39.3 How to order the appropriate test in inherited neuropathies (updated information on the availability of new genetic testing can be found on the website www.genetests.org). ModiWed from (Kleopa and Scherer, 2002), with permission of Elsevier Science.

SUMMARY

SUMMARY Inherited neuropathies are common. They are usually caused by mutations in genes that are expressed by myelinating Schwann cells or neurons, which is the biological basis for the long-standing distinction between primary ‘‘demyelinating’’ and ‘‘axonal’’ neuropathies. Neuropathies can be isolated, the primary manifestation of a more complex syndrome, or overshadowed by other aspects of the inherited disease. Increasing knowledge of the molecular genetic causes of inherited neuropathies facilitates a faster and more accurate diagnosis, setting the stage for the development of speciWc therapeutic interventions.

References Abel, A., Bone, L. J., Messing, A., Scherer, S. S., and Fischbeck, K. F. (1999). Studies in transgenic mice indicate a loss of connexin32 function in X-linked Charcot-Marie-Tooth disease. J. Neuropathol. Exp. Neurol. 58, 702–710. Abrams, C. K., Bennett, M. V. L., Verselis, V. K., and Bargiello, T. A. (2002). Voltage opens unopposed gap junction hemichannels formed by a connexin 32 mutant associated with X-linked Charcot-Marie-Tooth disease. Proc. Natl. Acad. Sci. USA 99, 3980–3984. Abrams, C. K., Freidin, M. M., Verselis, V. K., Bennett, M. V. L., and Bargiello, T. A. (2001). Functional alterations in gap junction channels formed by mutant forms of connexin 32: Evidence for loss of function as a pathogenic mechanism in the X-linked form of Charcot-Marie-Tooth disease. Brain Res. 900, 9–25. Abrams, C. K., Oh, S., Ri, Y., and Bargiello, T. A. (2000). Mutations in connexin 32: The molecular and biophysical bases for the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Rev. 32, 203–214. Adlkofer, K., Frei, R., Neuberg, D. H.-H., Zielasek, J., Toyka, K. V., and Suter, U. (1997a). Heterozygous peripheral myelin protein 22-deWcient mice are aVected by a progressive demyelinating peripheral neuropathy. J. Neurosci. 17, 4662–4671. Adlkofer, K., Martini, R., Aguzzi, A., Zielasek, J., Toyka, K. V., and Suter, U. (1995). Hypermyelination and demyelinating peripheral neuropathy in pmp22-deWcient mice. Nat. Genet. 11, 274–280. Adlkofer, K., Naef, R., and Suter, U. (1997b). Analysis of compound heterozygous mice reveals that the Trembler mutation can behave as a gain-of-function allele. J. Neurosci. Res. 49, 671–680. Agarwala, K. L., Kokame, K., Kato, H., and Miyata, T. (2000). Phophorylation of RTP, an ER stress-response cytoplasmic protein. Biochem. Biophys. Res. Comm. 272, 641–647. Altevogt, B. M., Kleopa, K. A., Postma, F. R., Scherer, S. S., and Paul, D. L. (2002). Cx29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J. Neurosci. 22, 6458–6470. Anderson, S. L., Coli, R., Daly, I. W., Kichula, E. A., Rork, M. J., Volpi, S. A., Ekstein, J., and Rubin, B. Y. (2001). Familial dysautonomia is caused by mutations of the IKAP gene. Amer. J. Hum. Genet. 68, 753–758. Antonellis, A., Ellsworth, R. E., Sambuughin, N., Puls, I., Abel, A., Lee-Lin, S. Q., Jordanova, A., Kremensky, I., Christodoulou, K., Middleton, L. T., Sivakumar, K., Ionasescu, V., Funalot, B., Vance, J. M., Goldfarb, L. G., Fischbeck, K. H., and Green, E. D. (2003) Glycyl tRNA synthetase mutations in Charcot-Marie-Tooth disease type 2D and distal spinal muscular atrophy type V. Amer, J. Hum. Genet. 72, 1293–1299. Anzini, P., Neuberg, D. H.-H., Schachner, M., Nelles, E., Willecke, K., Zielasek, J., Toyka, K., Suter, U., and Martini, R. (1997). Structural abnormalities and deWcient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin32. J. Neurosci. 17, 4545–4561. Auer-Grumbach, M., De Jonghe, P., Wagner, K., Verhoeven, K., Hartung, H.-P., and Timmerman, V. (2000a). Phenotype-genotype correlations in a CMT2B family ith reWned 3q13-q22 locus. Neurology 55, 1552–1557. Auer-Grumbach, M., Wagner, K., Timmerman, V., De Jonghe, P., and Hartung, H.-P. (2000b). Ulceromutilating neuropathy in an Austrian kinship without linkage to hereditary motor and sensory neuropathy IIB and hereditary sensory neuropathy I loci. Neurology 54, 45–52. Azzedine, H., Bolino, A., Taieb, T., Birouk, N., Di Duca, M., Bouhouche, A., Benamou, S., Mrabet, A., Hammadouche, T., Chkili, T., Gouider, R., Ravazzolo, R., Brice, A., Laporte, J., and Le Guern, E. (2003). Mutations in MTMR13, a new pseudo-phosphatase homologue of MTMR2 and Sbf1, are responsible for an autosomal recessive demyelinating form of Charcot-Marie-Tooth disease associated with early-onset glaucoma. Amer. J. Hum. Genet. 72, 1143–1153. Baechner, D., Liehr, T., Hameister, H., Altenberger, H., Grehl, H., Suter, U., and Rautenstrauss, B. (1995). Widespread expression of the peripheral myelin protein-22 gene (PMP22) in neural and non-neural tissues during murine development. J. Neurosci. Res. 42, 733–741. Ba¨hr, M., Andres, F., Timmerman, V., Nelis, E., Van Broeckhoven, C., and Dichgans, J. (1999). Central visual, acoustic, and motor pathway involvement in a Charcot-Marie-Tooth family with an Asn205Ser mutation in the connexin32 gene. J. Neurol. Neurosurg. Psychiat. 66, 202–206. Balice-Gordon, R. J., Bone, L. J., and Scherer, S. S. (1998). Functional gap junctions in the Schwann cell myelin sheath. J. Cell Biol. 142, 1095–1104. Baxter, R. V., Ben Othmane, K., Rochelle, J. M., Stajich, J. E., Hulette, C., Dew-Knight, S., Hentati, F., Ben Hamida, M., Bel, S., Stenger, J. E., Gilbert, J. R., Pericak-Vance, M., and Vance, J. M. (2002). Ganglioside-

935

936

39. INHERITED NEUROPATHIES

induced diVerentiation-associated protein-1 is mutant in Charcot-Marie-Tooth disease type 4A/8q21. Nat. Genet. 30, 21–22. Bejaoui, K., Wu, C. Y., ShefXer, M. D., Haan, G., Ashby, P., Wu, L. C., De Jonghe, P., and Brown, R. H. (2001). SPTLC1 is mutated in hereditary sensory neuropathy, type 1. Nat. Genet. 27, 261–262. Ben Othmane, K., Hentati, F., Lennon, F., Hamida, C. B., Biel, S., Roses, A. D., Pericak-Vance, M. A., Hamida, M. B., and Vance, J. M. (1993a). Linkage of a locus (CMT4A) for autosomal recessive Charcot-Marie-Tooth disease to chromosome 8q. Hum. Mol. Genet. 2, 1625–1628. Ben Othmane, K., Johnson, E., Menold, M., Graham, F. L., BenHamida, M., Hasegawa, O., Rogala, A. D., Ohnishi, A., PericakVance, M., Hentati, F., and Vance, J. M. (1999). IdentiWcation of a new locus for autosomal recessive Charcot-Marie-Tooth disease with focally folded myelin on chromosome 11p15. Genomics 62, 344–349. Ben Othmane, K., Loeb, D., Hayworth-Hodgte, R., Hentati, F., Rao, N., Roses, A. D., Hamida, K. B., Pericak, M. A., and Vance, J. M. (1995). Physical and genetic mapping of the CMT4A locus and exclusion of PMP-22 as a defect in CMT4A. Genomics 28, 286–290. Ben Othmane, K., Middleton, L. T., Lopest, L. J., Wilkinson, K. M., Lennon, F., Rozear, M. P., Stajich, J. M., Gaskell, P. C., Roses, A. D., Pericak-Vance, M. A., and Vance, J. M. (1993b). Localization of a gene (CMT2A) for autosomal dominant Charcot-Marie-Tooth disease type 2 to chromosome 1p and evidence of genetic heterogeneity. Genomics 17, 370–375. Bence, N. F., Sampat, R. M., and Kopito, R. R. (2001). Impairment of the ubiquitin-proteasome system by protein aggregation. Science 292, 1552–1555. Benstead, T., Kuntz, N., Miller, R., and Daube, J. (1990). The elecrophysiologic proWle of De´je´rine-Sottas disease (HMSN III). Muscle Nerve 13, 586–592. Berciano, J., Garcia, A., Calleja, J., and Combarros, O. (2000). Clinico-electrophysiological correlation of extensor digitorum brevis muscle atrophy in children with Charcot-Marie-Tooth disease 1A duplication. Neuromuscular Disord. 10, 419–424. Berciano, J., Garcia, A., and Combarros, O. (2003). Initial semeiology in childrin with Charcot-Marie-Tooth disease 1A duplication. Muscle Nerve 27, 34–39. Berciano, M. T., Fernandez, R., Pena, E., Calle, E., Villagra, N. T., and Lafarga, M. (1999). Necrosis of Schwann cells during tellurium-induced primary demyelination: DNA fragmentation, reorganization of splicing machinery, and formation of intranuclear rods of actin. J. Neuropathol. Exp. Neurol. 58, 1234–1243. Berger, P., Bonneick, S., Willi, S., Wymann, M., and Suter, U. (2002a). Loss of phosphatase activity in myotubularin-related protein 2 is associated with Charcot-Marie-Tooth disease type 4B1. Hum. Mol. Genet. 11, 1569–1579. Berger, P., Young, P., and Suter, U. (2002b). Molecular cell biology of Charcot-Marie-Tooth disease. Neurogenetics 4, 1–15. BergoVen, J., Scherer, S. S., Wang, S., Oronzi-Scott, M., Bone, L., Paul, D. L., Chen, K., Lensch, M. W., Chance, P., and Fischbeck, K. (1993). Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262, 2039–2042. Bermingham, J. R., Jr.,, Scherer, S. S., O’Connell, S., Arroyo, E., Kalla, K., Powell, F. R., and Rosenfeld, M. G. (1996). tst-1/SCIP/Oct-6 regulates a unique step in peripheral myelination and is required for normal respiration. Genes Dev. 10, 1751–1762. Bird, T. D., Kraft, G. H., Lipe, H. P., Kenney, K. L., and Sumi, S. M. (1997). Clinical and pathological phenotype of the original family with Charcot-Marie-Tooth type 1B: A 20-year study. Ann. Neurol. 41, 463–469. Bird, T. D., Ott, J., and Giblett, E. R. (1982). Evidence for linkage of Charcot-Marie-Tooth neuropathy to the DuVy locus on chromosome 1. Am. J. Genet. 34, 388–394. Birouk, N., Gouider, R., Le Guern, E., Gugenheim, M., Tardieu, S., Maisonobe, T., Le Forestier, N., Agid, Y., Brice, A., and Bouche, P. (1997). Charcot-Marie-Tooth disease type 1A with 17p11.2 duplication—Clinical and electrophysiological phenotype study and factors inXuencing disease severity in 119 cases. Brain 120, 813–823. Birouk, N., Le Guern, E., Maisonobe, T., Rouger, H., Gouider, R., Gugenheim, M., Tardieu, S., Gugenheim, M., Routon, M. C., Leger, J. M., Agid, Y., Brice, A., and Bouche, P. (1998). X-linked Charcot-Marie-Tooth disease with connexin 32 mutations—clinical and electrophysiological study. Neurology 50, 1074–1082. Boerkoel, C. F., Takashima, H., Garcia, C. A., Olney, R. K., Johnson, J., Berry, K., Russo, P., Kennedy, S., Teebi, A. S., Scavina, M., Williams, L. L., Mancias, P., Butler, I. J., Krajewski, K., Shy, M., and Lupski, J. R. (2002). Charcot-Marie-Tooth disease and related neuropathies: Mutation distribution and genotype-phenotype correlation. Ann. Neurol. 51, 190–201. Boerkoel, C. F., Takashima, H., Stankiexicz, P., Garcia, C. A., Leber, S. M., Rhee-Morris, L., and Lupski, J. R. (2000). Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Amer. J. Hum. Genet. 68, 325–333. Bolino, A., Marigo, V., Ferrera, F., Loader, J., Romio, L., Leoni, A., Di Duca, M., Cinti, R., Cecchi, C., Feltri, M. L., Wrabetz, L., Ravazzolo, R., and Monaco, A. P. (2002). Molecular characterization and expression analysis of Mtmr2, a mouse homologue of MTMR2, the myotubularin-related-2 gene mutated in CMT4B. Gene 283, 17–26. Bolino, A., Muglia, M., Conforti, F. L., LeGuern, E., Salih, M. A. M., Georgiou, D. M., Christodoulou, K., Hausmanowa-Petrusewicz, I., Mandich, P., Schenone, A., Gambardella, A., Bono, F., Quattrone, A., Devoto, M., and Monaco, A. P. (2000). Charcot-Marie-Tooth type 4B is caused by mutations in the gene encoding myotubularin-related protein-2. Nat. Genet. 25, 17–19.

SUMMARY

Bomont, P., Cavalier, L., Blondeau, F., Hamida, C. B., Belal, S., Tazir, M., Demir, E., Topaloglu, H., Korinthenberg, R., Tuysuz, B., Landrieu, P., Hentati, F., and Koenig, M. (2000). The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat Genet 26, 370–374. Bondurand, N., Girard, M., Pingault, V., Lemort, N., Dubourg, O., and Goossens, M. (2001). Human connexin 32, a gap junction protein altered in the X-linked form of Charcot-Marie-Tooth disease, is directly regulated by the transcription factor SOX10. Hum. Mol. Genet. 10, 2783–2795. Bouhouche, A., Benomar, A., Birouk, N., Mularoni, A., Meggouh, F., Tassin, J., Grid, D., Vandenberghe, A., Yahyaoui, M., Chkili, T., Brice, A., and LeGuern, E. (1999). A locus for an axonal form of autosomal recessive Charcot-Marie-Tooth disease maps to chromosome 1q21.2q21.3. Amer. J. Hum. Genet. 65, 722–727. Bradley, W. G., and Asbury, A. K. (1972). Radioautographic studies of Schwann cell behavior: I. Acrylamide neuropathy in the mouse. J. Neuropathol. Exp. Neurol. 29, 500–506. Bradley, W. G., Madrid, R., and Davis, D. J. F. (1977). The peroneal muscular atrophy syndrome—clinical, genetic, electrophysiological and nerve biopsy studies. 3. Clinical, electrophysiological and pathological correlations. J. Neurol. Sci. 32, 123–136. Brancolini, C., Edomi, P., Marzinotto, S., and Schneider, C. (2000). Exposure at the cell surface is required for gas3/PMP22 to regulate both cell death and cell spreading: Implication for the Charcot-Marie-Tooth type 1A and Dejerine-Sottas diseases. Mol. Biol. Cell. 11, 2901–2914. Brancolini, C., Marzinotto, S., Edomi, P., Agostoni, E., Fiorentini, C., Mu¨ller, H. W., and Schneider, C. (1999). Rho-dependent regulation of cell spreading by the tetraspan membrane protein Gas3/PMP22. Mol. Biol. Cell. 10, 2441–2459. Brownlees, J., Ackerley, S., Grierson, A. J., Jacobsen, N. J., Shea, K., Anderton, B. H., Leigh, P. N., Shaw, C. E., and Miller, C. C. (2002). Charcot-Marie-Tooth disease neuroWlament mutations disrupt neuroWlament assembly and axonal transport. Hum. Mol. Genet. 11, 2837–2844. Brunden, K. R., and Poduslo, J. F. (1987). A phorbol ester-sensitive kinase catalyzes the phosphorylation of P0 glycoprotein in myelin. J. Neurochem. 49, 1863–1872. Bruzzone, R., T. W. White, S. S. Scherer, Fischbeck, K. H., and Paul, D. L. (1994). Null mutations of connexin32 in patients with X-linked Charcot-Marie-Tooth disease. Neuron 13, 1253–1260. Bruzzone, R., White, T. W., and Paul, D. L. (1996). Connections with connexins: The molecular basis of direct intercellular signaling. Eur. J. Biochem. 238, 1–27. Castro, C., Gomez-Hernandez, J. M., Silander, K., and Barrio, L. C. (1999). Altered formation of hemichannels and gap junction channels caused by C-terminal connexin-32 mutations. J. Neurosci. 19, 3752–3760. Chance, P. F., Alderson, M. K., Leppig, K. A., Lensch, M. W., Matsunami, N., Smith, B., Swanson, P. D., Odelberg, S. J., Disteche, C. M., and Bird, T. D. (1993). DNA deletion associated with hereditary neuropathy with liability to pressure palsies. Cell 72, 143–151. Chance, P. F., Bird, T. D., Matsunami, N., Lensch, M. W., Brothman, A. R., and Feldman, G. M. (1992a). Trisomy-17p associated with Charcot-Marie-Tooth neuropathy Type-1A phenotype—evidence for gene dosage as a mechanism in CMT1A. Neurology 42, 2295–2299. Chance, P. F., Matsunami, N., Lensch, W., Smith, B., and Bird, T. D. (1992b). Analysis of the DNA duplication 17p11.2 in Charcot-Marie-Tooth neuropathy type-1 pedigrees—additional evidence for a third autosomal CMT1 locus. Neurology 42, 2037–2041. Chaouch, M., Allal, Y., De Sandre-Giovannoli, A., Vallat, J. M., Amerel-Khedoud, A., Kassouri, N., Chaouch, A., Sindou, P., Hammadouche, T., Tazir, M., Levy, N., and Grid, D. (2003). The phenotypic manifestations of autosomal recessive axonal Charcot-Marie-Tooth due to a mutation in Lamin A/C gene. Neuromuscular Disord. 13, 60–67. Chapon, F., Latour, P., Diraison, P., SchaeVer, S., and Vandenberghe, A. (1999). Axonal phenotype of CharcotMarie-Tooth disease associated with a mutation in the myelin protein zero gene. J. Neurol. Neurosurg. Psychiat. 66, 779–782. Christodoulou, K., Zamba, E., Tsingis, M., Mubaidin, A., Horani, K., AbuSheik, S., El Khateeb, M., Kyriacou, K., Kyriakides, T., Al Qudah, A. K., and Middleton, L. (2000). A novel form of distal hereditary motor neuronopathy maps to chromosome 9p21.1-p12. Ann. Neurol. 48, 877–884. Colby, J., Nicholson, R., Dickson, K. M., Orfali, W., Naef, R., Suter, U., and Snipes, G. J. (2000). PMP22 carrying the Trembler or Trembler-J mutation is intracellularly retained in myelinating Schwann cells. Neurobiol. Dis. 7, 561–573. Colman, D. R., Kreibich, G., Frey, A. B., and Sabatini, D. D. (1982). Synthesis and incorporation of myelin polypeptides into CNS myelin. J. Cell Biol. 95, 598–608. Comabella, M., Waye, J. S., Raguer, N., Eng, B., Dominguez, C., Navarro, C., Borras, C., Krivit, W., and Montalban, X. (2001). Late-onset metachromatic leukodystrophy clinically presenting as isolated peripheral neuropathy: Compound heterozygosity for the IVS2þ1G -> A mutation and a newly identiWed missense mutation (Thr408Ile) in a Spanish family. Ann. Neurol. 50, 108–112. Cowchock, R. S., Duckett, S. W., Streletz, L. J., Graziani, L. J., and Jackson, L. G. (1985). X-linked sensory-motor neuropathy type II with deafness and mental retardation: A new disorder. Am. J. Hum. Genet. 20, 307–315. Cuesta, A., Pedrola, L., Sevilla, T., Garcia-Planells, J., Chumillas, M. J., Mayordomo, F., Le Guern, E., Marin, I., Vilchez, J. J., and Palau, F. (2002). The gene encoding ganglioside-induced diVerentiation-associated protein 1 is mutated in axonal Charcot-Marie-Tooth type 4A disease. Nat. Genet. 30, 22–25. D’Urso, D., Brophy, P. J., Staugaitis, S. M., Gillespie, C. S., Frey, A. B., Stempak, J. G., and Colman, D. R. (1990). Protein zero of peripheral nerve myelin: Biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron 4, 449–460.

937

938

39. INHERITED NEUROPATHIES

D’Urso, D., Ehrhardt, P., and Mu¨ller, H. W. (1999). Peripheral myelin protein 22 and protein zero: A novel association in peripheral nervous system myelin. J. Neurosci. 19, 3396–3403. D’Urso, D., Prior, R., Greiner-Petter, R., Gabreels-Festen, A. A. W. M., and Mu¨ller, H. W. (1998). Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutations of the peripheral myelin protein PMP22. J. Neurosci. 18, 731–740. D’Urso, D., Schmalenbach, C., Zoidl, G., Prior, R., and Mu¨ller, H. W. (1997). Studies on the eVects of altered PMP22 expression during myelination in vitro. J. Neurosci. Res. 48, 31–42. Davis, D. J. F., Bradley, W. G., and Madrid, R. (1977). The peroneal muscular atrophy syndrome—clinical, genetic, electrophysiological and nerve biopsy studies. 1. Clinical, genetic, and electrophysiological Wndings and classiWcation. J. Ge´ne´t. Hum. 26, 311–349. Dawkins, J. L., Hulme, D. J., Brahmbhatt, S. B., Auer-Grumbach, M., and Nicholson, G. A. (2001). Mutations in SPTLC1, encoding serine palmitoyltransferase, long chain base subunit-1, cause hereditary sensory neuropathy type I. Nat. Genet. 27, 309–312. De Jonghe, P., Mersivanova, I., Nelis, E., Del Favero, J., Martin, J. J., Van Broeckhoven, C., Evgrafov, O. C., and Timmerman, V. (2001). Further evidence that neuroWlament light chain gene mutations can cause Charcot-Marie-Tooth disease type 2E. Ann. Neurol. 49, 245–249. De Jonghe, P., Timmerman, V., Ceuterick, C., Nelis, E., De Vriendt, E., Lofgren, A., Vercruyssen, A., Verellen, C., Van Maldergem, L., Martin, J. J., and Van Broeckhoven, C. (1999). The Thr124Met mutation in the peripheral myelin protein zero (MPZ) gene is associated with a clinically distinct Charcot-Marie-Tooth phenotype. Brain 122, 281–290. De Jonghe, P., Timmerman, V., Nelis, E., Martin, J.-J., and Van Broeckhoven, C. (1997). Charcot-Marie-Tooth disease and related peripheral neuropathies. J. Peripher. Nerv. Syst. 2, 370–387. De Leon, M., Welcher, A. A., Suter, U., and Shooter, E. M. (1991). IdentiWcation of transcriptionally regulated genes after sciatic nerve injury. J. Neurosci. Res. 29, 437–448. De Sandre-Giovannoli, A., Chaouch, M., Kozlov, S., Vallat, J. M., Tazir, M., Kassouri, N., Szepetowski, P., Hammadouche, T., Vandenberghe, A., Stewart, C. L., Grid, D., and Levy, N. (2002). Homozygous defects in LMNA, encoding lamin A/C nuclear-envelope proteins, cause autosomal recessive axonal neuropathy in human (Charcot-Marie-Tooth disorder type 2) and mouse. Amer. J. Hum. Genet. 70, 726–736. Delague, V., Bareil, C., TuVery, S., Bouvagnet, P., Chouery, E., Koussa, S., Maisonobe, T., Loiselet, J., Megarbane, A., and Claustres, M. (2000). Mapping of a new locus for autosomal recessive demyelinating Charcot-Marie-Tooth disease to 19q13.1–13.3 in a large consanguineous Lebanese family: Exclusion of MAG as a candidate gene. Amer. J. Hum. Genet. 67, 236–243. Dickson, K. M., Bergeron, J. J. M., Shames, I., Colby, J., Nguyen, D. T., Chevet, E., Thomas, D. Y., and Snipes, G. J. (2002). Association of calnexin with mutant peripheral myelin protein-22 ex vivo: A basis for gain-offunction ER diseases. Proc. Natl. Acad. Sci. USA 99, 9852–9857. Ding, J. Q., Liu, J. J., Kowal, A. S., Nardine, T., Bhattacharya, P., Lee, A., and Yang, Y. M. (2002). Microtubuleassociated protein 1B: A neuronal binding partner for gigaxonin. J. Cell Biol. 158, 427–433. Ding, Y., and Brunden, K. R. (1995). The cytoplasmic domain of myelin glycoprotein P0 interacts with negatively charged phospholipid bilayers. J. Biol. Chem. 269, 10764–10770. Donaghy, M., Sisodiya, S. M., Kennett, R., McDonald, B., Haites, N., and Bell, C. (2000). Steroid responsive polyneuropathy in a family with a novel myelin protein zero mutation. J. Neurol. Neurosurg. Psychiat. 69, 799–805. Dubourg, O., Tardieu, S., Birouk, N., Gouider, R., Leger, J. M., Maisonobe, T., Brice, A., Bouche, P., and LeGuern, E. (2001). Clinical, electrophysiological and molecular genetic characteristics of 93 patients with X-linked Charcot-Marie-Tooth disease. Brain 124, 1958–1967. Dyck, P. J. (1993). Neuronal atrophy and degeneration predominantly aVecting peripheral sensory and autonomic neurons. In ‘‘Peripheral Neuropathy’’ 3rd ed., vol.2 (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low and J. F. Poduslo, eds.), 1065–1093. W. B. Saunders, Philadelphia. Dyck, P. J., Chance, P., Lebo, R., and Carney, J. A. (1993a). Hereditary motor and sensory neuropathies. In ‘‘Peripheral Neuropathy’’ 3rd ed., vol.2 (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low, and J. F. Poduslo, eds.), 1094–1136. W. B. Saunders,, Philadelphia. Dyck, P. J., Giannini, D., and Lais, A. (1993b). Pathologic alterations of nerves. In ‘‘Peripheral Neuropathy ’’ 3rd ed., vol.1 (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low, and J. F. Poduslo, eds.),3rd ed., 514–595. W. B. Saunders, Philadelphia. Dyck, P. J., Karnes, J. L., and Lambert, E. H. (1989). Longitudinal study of neuropathic deWcits and nerve conduction abnormalities in hereditary motor and sensory neuropathy type 1. Neurology 39, 1302–1308. Dyck, P. J., Lais, A. C., and OVord, K. P. (1974). The nature of myelinated nerve Wber degeneration in dominately inherited hypertrophic neuropathy. Mayo Clinic Proc. 49, 34–39. Elliot, J. L., Kwon, J. M., Goodfellow, P. J., and Yee, W.-C. (1997). Hereditary motor and sensory neuropathy IIB: Clinical and electrophysiological characteristics. Neurology 48, 23–28. Erdem, S., Mendell, J. R., and Sahenk, Z. (1998). Fate of Schwann cell in CMT1A and HNPP: Evidence for apoptosis. J. Neuropathol. Exp. Neurol. 57, 635–642. Erne, B., Sansano, S., Frank, M., and Schaeren-Wiemers, N. (2002). Rafts in adult peripheral nerve myelin contain major structural myelin proteins and myelin and lymphocyte protein (MAL) and CD59 as speciWc markers. J. Neurochem. 82, 550–562.

SUMMARY

Fabbretti, E., Edomi, P., Brancolini, C., and Schneider, C. (1995). Apoptotic phenotype induced by overexpression of wild-type gas3/PMP22: Its relation to the demyelinating peripheral neuropathy CMT1A. Genes Dev. 9, 1846–1856. Fabrizi, C., Kelly, B. M., Gillespie, C. S., Schlaepfer, W. W., Scherer, S. S., and Brophy, P. J. (1997). Transient expression of the neuroWlament proteins NF-L and NF-M by Schwann cells is regulated by axonal contact. J. Neurosci. Res. 50, 291–299. Fabrizi, G. M., Cavallaro, T., Taioli, F., Orrico, D., Morbin, M., Simonati, A., and Rizzuto, N. (1999). Myelin uncompaction in Charcot-Marie-Tooth neuropathy type 1A with a point mutation of peripheral myelin protein-22. Neurology 53, 846–851. Fabrizi, G. M., Simonati, A., Morbin, M., Cavallaro, T., Taioli, F., Benedetti, M. D., Edomi, P., and Rizzuto, N. (1998). Clinical and pathological correlations in Charcot-Marie-Tooth neuropathy type 1A with the 17p11.2p12 duplication: A cross-sectional morphometric and immunohistochemical study in twenty cases. Muscle Nerve 21, 869–877. Fannon, A. M., Sherman, D. L., Ilyina-Gragerova, G., Brophy, P. J., Friedrich, V. L., and Colman, D. R. (1995). Novel E-cadherin mediated adhesion in peripheral nerve: Schwann cell architecture is stabilized by autotypic adherens junctions. J. Cell Biol. 129, 189–202. Fidzianska, A., Drac, H., and Rafalowska, J. (2002). Phenomenon of Schwann cell apoptosis in a case of congenital hypomyelinating neuropathy with basal lamina onion bulb formation. Brain Develop. 24, 727–731. Filbin, M. T., and Tennekoon, G. I. (1993). Homophilic adhesion of the myelin P0 protein requires glycosylation of both molecules in the homophilic pair. J. Cell Biol. 122, 451–459. Filbin, M. T., Walsh, F. S., Trapp, B. D., Pizzey, J. A., and Tennekoon, G. I. (1990). Role of P0 protein as a homophilic adhesion molecule. Nature 344, 871–872. Fischbeck, K. H., ar-Rushdi, N., Pericak-Vance, M., Rozear, M., Roses, A. D., and Fryns, J. P. (1986). X-linked neuropathy: Gene localization with DNA probes. Ann. Neurol. 20, 527–532. Frei, R., Motzing, S., Kinkelin, I., Schachner, M., Koltzenburg, M., and Martini, R. (1999). Loss of distal axons and sensory Merkel cells and features indicative of muscle denervation in hindlimbs of P0-deWcient mice. J. Neurosci. 19, 6058–6067. Gable, K., Han, G., Monaghan, E., Bacikova, D., Natarajan, M., Williams, R., and Dunn, T. M. (2002). Mutations in the yeast LCB1 and LCB2 genes, including those corresponding to the hereditary sensory neuropathy type I mutations, dominantly inactivate serine palmitoyltransferase. J. Biol. Chem. 277, 10194–10200. Gabreels-Festen, A. (2002). Dejerine-Sottas syndrome grown to maturity: Overview of genetic and morphological heterogeneity and follow-up of 25 patients. J. Anat. 200, 341–356. Gabreels-Festen, A. A. W. M., Bolhuis, P. A., Hoogendijk, J. E., Valentijn, L. J., Eshuis, E. J. H. M., and Gabreels, F. J. M. (1995). Charcot-Marie-Tooth disease type 1A: Morphological phenotype of the 17p duplication versus PMP22 point mutations. Acta Neuropathol. 90, 645–649. Gabreels-Festen, A. A. W. M., Hoogendijk, J. E., Meijerink, P. H. S., Gabreels, F. J. M., Bolhuis, P. A., Vanbeersum, S., Kulkens, T., Nelis, E., Jennekens, F. G. I., Devisser, M., Vanengelen, B. G. M., Van Broeckhoven, C., and Mariman, E. C. M. (1996). Two divergent types of nerve pathology in patients with diVerent P0 mutations in Charcot-Marie-Tooth disease. Neurology 47, 761–765. Gabreels-Festen, A. A. W. M., Joosten, E. M. G., Gabreels, F. J. M., Jennekens, F. G. I., Gooskens, R. H. J. M., and Stegeman, D. F. (1991). Hereditary motor and sensory neuropathy of neuronal type with onset in early childhood. Brain 114, 1855–1870. Gabreels-Festen, A. A. W. M., Joosten, E. M. G., Gabreels, F. J. M., Jennekens, F. G. I., and Kempen, T. W. J. (1992). Early morphological features in dominantly inherited demyelinating motor and sensory neuropathy (HMSN Type-I). J. Neurol. Sci. 107, 145–154. Gabriel, J. M., Erne, B., Pareyson, D., Sghirlanzoni, A., Taroni, F., and Steck, A. J. (1997). Gene dosage eVects in hereditary peripheral neuropathy—Expression of peripheral myelin protein 22 in Charcot-Marie-Tooth disease type 1A and hereditary neuropathy with liability to pressure palsies nerve biopsies. Neurology 49, 1635–1640. Gao, Y., Li, W., and Filbin, M. T. (2000). Acylation of myelin P0 protein is required for adhesion. J. Neurosci. Res. 60, 704–713. Garbern, J. Y., Cambi, F., Tang, X. M., Sima, A. A. F., Vallat, J. M., Bosch, E. P., Lewis, R., Shy, M., Sohi, J., Kraft, G., Chen, K. L., Joshi, I., Leonard, D. G. B., Johnson, W., Raskind, W., Dlouhy, S. R., Pratt, V., Hodes, M. E., Bird, T., and Kamholz, J. (1997). Proteolipid protein is necessary in peripheral as well as central myelin. Neuron 19, 205–218. Garcia, C. A., Malamut, R. E., Englnad, J. D., Parry, G. S., Liu, P., and Lupski, J. R. (1995). Clinical variability in two pairs of identical twins with the Charcot-Marie-Tooth disease type 1A duplication. Neurology 45, 2090–2093. Ghazvini, M., Mandemakers, W., Jaegle, M., Piirsoo, M., Driegen, S., Koutsourakis, M., Smit, X., Grosveld, F., and Meijer, D. (2002). A cell type-speciWc allele of the POU gene Oct-6 reveals Schwann cell autonomous function in nerve development and regeneration. EMBO J. 21, 4612–4620. Giese, K. P., Martini, R., Lemke, G., Soriano, P., and Schachner, M. (1992). Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71, 565–576. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J. (1994). Periaxin, a novel protein of myelinating Schwann cells with a possible role in axonal ensheathment. Neuron 12, 497–508.

939

940

39. INHERITED NEUROPATHIES

Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., Cottrell, D. F., Tait, S., Garry, E. M., Wallace, V. C. J., Ure, J., GriYths, I. R., Smith, A., and Brophy, P. J. (2000). Peripheral demyelination and neuropathic pain behavior in periaxin-deWcient mice. Neuron 26, 523–531. Gouider, R., Le Guern, E., Gugenheim, M., Tardieu, S., Maisonobe, T., Leger, J. M., Vallat, J. M., Agid, Y., Bouche, P., and Brice, A. (1995). Clinical, electrophysiologic, and molecular correlations in 13 families with hereditary neuropathy with liability to pressure palsies and a chromosome 17p11.2 deletion. Neurology 45, 2018–2023. GreenWeld, S., BrostoV, S., Eylar, E. H., and Morell, P. (1973). Protein composition of myelin of the peripheral nervous system. J. Neurochem. 20, 1207–1216. GriYn, J. W., and HoVman, P. N. (1993). Degeneration and regeneration in the peripheral nervous system. In ‘‘Peripheral Neuropathy’’ 3rd ed., vol.1 (P. J. Dyck, P. K. Thomas, P. A. Low, and J. F. Poduslo, eds.), 361–376. W. B. Saunders, Philadelphia. Grinspan, J. B., Marchionni, M., Reeves, M., Coulaloglou, M., and Scherer, S. S. (1996). Axonal interactions regulate Schwann cell apoptosis in developing peripheral nerve: Neuregulin receptors and the role of neuregulins. J. Neurosci. 16, 6107–6118. Grohmann, K., Schuelke, M., Diers, A., HoVmann, K., Lucke, B., Adams, C., Bertini, E., Leonhardt-Horti, H., Muntoni, F., Ouvrier, R., Pfeufer, A., Rossi, R. N., Van Maldergem, L., Wilmshurst, J. M., Wienker, T. F., Sendtner, M., Rudnik-Scho¨neborn, S., Zerres, K., and Hu¨bner, C. (2001). Mutations in the gene encoding immunoglobulin mu-binding protein 2 cause spinal muscular atrophy with respiratory distress type 1. Nat. Genet. 29, 75–77. Grohmann, K., Wienker, T. F., Saar, K., Rudnik-Scho¨neborn, S., Stoltenburg-Didinger, G., Rossi, R. N., Nu¨rnberg, G., Pfeufer, A., Wirth, B., Reis, A., Zerres, K., and Hu¨bner, C. (1999). Diaphragmatic spinal muscular atrophy with respiratory distress is heterogeneous, and one form is linked to chromosome 11q13q21. Amer. J. Hum. Genet. 65, 1459–1462. Guilbot, A., Kessali, M., Ravise, N., Hammadouche, T., Bouhouche, A., Maisonobe, T., Grid, D., Brice, A., and Guern, E. L. (1999). The autosomal recessive form of CMT disease linked to 5q31-q33. Ann. N. Y. Acad. Sci. 883, 56–59. Guilbot, A., Williams, A., Ravise, N., Verny, C., Brice, A., Sherman, D. L., Brophy, P. J., Le Guern, E., Delague, V., Bareil, C., Megarbane, A., and Claustres, M. (2001). A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10, 415–421. Gutierrez, A., England, J. D., Sumner, A. J., Ferer, S., Warner, L. E., Lupski, J. R., and Garcia, C. A. (2000). Unusual electrophysiological Wndings in X-linked dominant Charcot-Marie-Tooth disease. Muscle Nerve 23, 182–188. Guzzetta, F., Ferriere, G., and Lyon, G. (1982). Congenital demyelinating neuropathy. Pathological Wndings compared with polyneuropathies starting later in life. Brain 105, 395–416. Hahn, A., Ainsworth, P. J., Bolton, C. F., Bilbao, J. M., and Vallat, J.-M. (2001). Pathological Wndings in the X-linked form of Charcot-Marie-Tooth disease: A morphometric and ultrastructural analysis. Acta Neuropathol. 101, 129–139. Hahn, A. F., Ainsworth, P. J., Naus, C. C. G., Mao, J., and Bolton, C. F. (2000). Clinical and pathological observations in men lacking the gap junction protein connexin 32. Muscle Nerve, S39-S48. Hahn, A. F., Brown, W. F., Koopman, W. J., and Feasby, T. E. (1990). X-linked dominant hereditary motor and sensory neuropathy. Brain 113, 1511–1525. Hanemann, C. O., D’Urso, D., Gabreels-Festen, A. A. W. M., and Mu¨ller, H. W. (2000). Mutation-dependent alteration in cellular distribution of peripheral myelin protein 22 in nerve biopsies from Charcot-Marie-Tooth type 1A. Brain 123, 1001–1006. Hanemann, C. O., and Gabreels-Festen, A. A. (2002). Secondary axon atrophy and neurological dysfunction in demyelinating neuropathies. Curr. Opin. Neurol. 15, 611–615. Hanemann, C. O., Gabreels-Festen, A. A. W. M., and De Jonghe, P. (2001). Axon damage in CMT due to mutation in myelin protein P0. Neuromuscular Disord. 11, 753–756. Hanemann, C. O., Gabreels-Festen, A. A. W. M., Mu¨ller, H. W., and Stoll, G. (1996). Low aYnity NGF receptor expression in CMT1A nerve biopsies of diVerent disease stages. Brain 119, 1461–1469. Hanemann, C. O., Gabreels-Festen, A. A. W. M., Stoll, G., and Mu¨ller, H. W. (1997). Schwann cell diVerentiation in Charcot-Marie-Tooth disease type 1A (CMT1A): Normal number of myelinating Schwann cells in young CMT1A patients and neural cell adhesion molecule expression in onion bulbs. Acta Neuropathol. 94, 310–315. Hanemann, C. O., and Mu¨ller, H. W. (1998). Pathogenesis of Charcot-Marie-Tooth IA (CMTIA) neuropathy. Trends Neurosci. 21, 282–286. Hanemann, C. O., Rosenbaum, C., Kupfer, S., Wosch, S., Stoegbauer, F., and Mu¨ller, H. W. (1998). Improved culture methods to expand Schwann cells with altered growth behaviour from CMT1A patients. Glia 23, 89–98. Hanemann, C. O., Stoll, G., D’Urso, D., Fricke, W., Martin, J. J., Van Broeckhoven, C., Mancardi, G. L., Bratke, I., and Mu¨ller, H. W. (1994). Peripheral myelin protein-22 expression in Charcot-Marie-Tooth disease type Ia sural nerve biopsies. J. Neurosci. Res. 37, 654–659. Haney, C., Snipes, G. J., Shooter, E. M., Suter, U., Garcia, C., GriYn, J. W., and Trapp, B. D. (1996). Ultrastructural distribution of PMP22 in Charcot-Marie-Tooth disease type 1A. J. Neuropathol. Exp. Neurol. 55, 290–299. Harati, Y., and Butler, I. J. (1985). Congenital hypomyelinating neuropathy. J. Neurol. Neurosurg. Psychiat. 48, 1269–1276.

SUMMARY

Harding, A. E. (1995). From the syndrome of Charcot, Marie and Tooth to disorders of peripheral myelin proteins. Brain 118, 809–818. Harding, A. E., and Thomas, P. K. (1980). The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103, 259–280. Hasse, B., Bosse, F., and Mu¨ller, H. W. (2002). Proteins of peripheral myelin are associated with glycosphingolipid/cholesterol-enriched membranes. J. Neurosci. Res. 69, 227–232. Hayasaka, K., Himoro, M., Sato, W., Takada, G., Uyemura, K., Shimizu, N., Bird, T. D., Conneally, P. M., and Chance, P. F. (1993a). Charcot-Marie-Tooth neuropathy type 1B is associated with mutations of the myelin P0 gene. Nat. Genet. 5, 31–34. Hayasaka, K., Himoro, M., Sawaishi, Y., Nanao, K., Takahashi, T., Takada, G., Nicholson, G. A., Ouvrier, R. A., and Tachi, N. (1993b). De novo mutation of the myelin Po gene in Dejerine-Sottas disease (hereditary motor and sensory neuropathy type III). Nat. Genet. 5, 266–268. Hayasaka, K., Himoro, M., Wang, Y. M., Takata, M., Minoshima, S., Shimizu, N., Miura, M., Uyemura, K., and Takada, G. (1993c). Structure and chromosomal localization of the gene encoding the human myelin protein zero (MPZ). Genomics 17, 755–758. Hayasaka, K., Takada, G., and Ionasescu, V. V. (1993d). Mutation of the myelin P0 gene in Charcot-Marie-Tooth neuropathy type-1B. Hum. Mol. Genet. 2, 1369–1372. Herringham, W. P. (1889). Muscular atrophy of the peroneal type aVecting many members of a family. Brain 11, 230–236. Hilmi, S., Fournier, M., Valeins, H., Gandar, J. C., and Bonnet, J. (1995). Myelin P0 glycoprotein: IdentiWcation of the site phosphorylated in vitro and in vivo by endogenous protein kinases. J. Neurochem. 64, 902–907. Hirokawa, N. (1998). Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science 279, 519–526. Hoogendijk, J. E., de Visser, M., Bolhuis, P. A., Hart, A. A. M., and de Visser, B. W. O. (1994). Hereditary motor and sensory neuropathy type I—clinical and neurographical features of the 17p duplication subtype. Muscle Nerve 17, 85–90. Houlden, H., King, R. H. M., Wood, N. W., Thomas, P. K., and Reilly, M. M. (2001). Mutations in the 5’ region of the myotubularin-related protein 2 (MTMR2) gene in autosomal recessive hereditary neuropathy with focally folded myelin. Brain 124, 907–915. Indo, Y. (2001). Molecular basis of congenital insensitivity to pain with anhidrosis (CIPA): Mutations and polymorphisms in TRKA (NTRK1) gene encoding the receptor tyrosine kinase for nerve growth factor. Hum. Mutat. 18, 462–471. Inoue, K., Shilo, K., Boerkoel, C. F., Crowe, C., Sawady, J., Lupski, J. R., and Agamanolis, D. P. (2002). Congenital hypomyelinating neuropathy, central dysmyelination, and Waardenburg-Hirschsprung disease: Phenotypes linked by SOX10 mutation. Ann Neurol 52, 836–842. Inoue, K., Tanabe, Y., and Lupski, J. R. (1999). Myelin deWciencies in both the central and the peripheral nervous systems associated with a SOX10 mutation. Ann. Neurol. 46, 313–318. Ionasescu, V., Ionasescu, R., and Searby, C. (1996a). Correlation between connexin 32 gene mutations and clinical phenotype in X-linked dominant Charcot-Marie-Tooth neuropathy. Am. J. Med. Genet. 63, 486–491. Ionasescu, V. V., Ionasescu, R., and Searby, C. (1993). Screening of dominantly inherited Charcot-Marie-Tooth neuropathies. Muscle Nerve 16, 1232–1238. Ionasescu, V. V., Trofatter, J., and Haines, J. L. (1991). Heterogeneity in X-linked recessive Charcot-Marie-Tooth neuropathy. Amer. J. Hum. Genet. 48, 1075–1083. Ionasescu, V. V., Trofatter, J., Haines, J. L., Summers, A. M., Ionasescu, R., and Searby, C. (1992). X-linked recessive Charcot-Marie-Tooth neuropathy—clinical and genetic study. Muscle Nerve 15, 368–373. Irobi, J., Nelis, E., Verhoeven, K., DeVriendt, E., Dierick, I., DeJonghe, P., VanBroeckhoven, C., and Timmerman, V. (2002). Mutation analysis of 12 candidate genes for distal hereditary motor neuropathy type II (Distal HMN II) linked to 12q24.3. J. Peripher. Nerv. Syst. 7, 87–95. Irobi, J., Tissir, F., De Jonghe, P., De Vriendt, E., Van Broeckhoven, C., Timmerman, V., and Beuten, J. (2000). A clone contig of 12q24.3 encompassing the distal hereditary motor neuropathy type II gene. Genomics 65, 34–43. Ismailov, S. M., Fedotov, V. P., Dadali, E. L., Polyakov, A. V., Van Broeckhoven, C., Ivanov, V. I., De Jonghe, P., Timmerman, V., and Evgrafov, O. V. (2001). A new locus for autosomal dominant Charcot-Marie-Tooth disease type 2 (CMT2F) maps to chromosome 7q11-q21. Eur. J. Human Genet. 9, 646–650. Jacobs, J. M., and Gregory, R. (1991). Uncompacted lamellae as a feature of tomaculous neuropathy. Acta Neuropathol. 83, 87–91. Jaegle, M., Mandemakers, W., Broos, L., Zwart, R., Karis, A., Visser, P., Grosveld, F., and Meijer, D. (1996). The POU factor Oct-6 and Schwann cell diVerentiation. Science 273, 507–510. Jansen, G. A., Hogenhout, E. M., Ferdinandusse, S., Waterham, H. R., Ofman, R., Jakobs, C., Skjeldal, O. H., and Wanders, R. J. A. (1997). Human phytanoyl-CoA hydroxylase: Resolution of the gene structure and the molecular basis of Refsum’s disease. Nat. Genet. 17, 190–193. Jordanova, A., De Jonghe, P., Boerkoel, C. F., Takashima, H., De Vriendt, E., Ceuterick, C., Martin, J.-J., Butler, I. J., Mancias, P., Papasozomenos, S. C., Terespolsky, D., Potochi, L., Brown, C. W., Shy, M., Rita, D. A., Tournev, I., Kremensky, I., Lupski, J. R., and Timmerman, V. (2003). Mutations in the neuroWlament light chain geen (NEFL) cause early onset severe Charcot-Marie-Tooh disease. Brain 126, 590–597.

941

942

39. INHERITED NEUROPATHIES

Kaku, D. A., Parry, G. J., Malamut, R., Lupski, J. R., and Garcia, C. A. (1993a). Nerve conduction studies in Charcot-Marie-Tooth polyneuropathy associated with a segmental duplication of chromosome 17. Neurology 43, 1806–1808. Kaku, D. A., Parry, G. J., Malamut, R., Lupski, J. R., and Garcia, C. A. (1993b). Uniform slowing of conduction velocities in Charcot-Marie-Tooth disease polyneuropathy type 1. Neurology 43, 2664–2667. Kalaydjieva, L., Gresham, D., Gooding, R., Heather, L., Baas, F., de Jonge, R., Blechschmidt, K., Angelicheva, D., Chandler, D., Worsley, P., Rosenthal, A., King, R. H. M., and Thomas, P. K. (2000). N-myc downstreamregulated gene 1 is mutated in hereditary motor and sensory neuropathy-Lom. Amer. J. Hum. Genet. 67, 47–58. Kalaydjieva, L., Hallmayer, J., Chandler, D., Savov, A., Nikolova, A., Angelicheva, D., King, R. H. H., Ishpekova, B., Honeyman, K., Calafell, F., Shmarov, A., Petrova, J., Turnev, I., Hristova, A., Moskov, M., Stancheva, S., Petkova, I., Bittles, A. H., Georgieva, V., Middleton, L., and Thomas, P. K. (1996). Gene mapping in Gypsies identiWes a novel demyelinating neuropathy on chromosome 8q24. Nat. Genet. 14, 214–217. Kalaydjieva, L., Nikolova, A., Turnev, I., Petrova, J., Hristova, A., Ishpekova, B., Petkova, I., Shmarov, A., Stancheva, S., Middleton, L., Merlini, L., Trogu, A., Muddle, J. R., King, R. H. M., and Thomas, P. K. (1998). Hereditary motor and sensory neuropathy—LOM, a novel demyelinating neuropathy associated with deafness in gypsies—Clinical, electrophysiological and nerve biopsy Wndings. Brain 121, 399–408. Kamholz, J., Shy, M., and Scherer, S. S. (1994). Elevated expression of messenger RNA for peripheral myelin protein 22 in biopsied peripheral nerves of patients with Charcot-Marie-Tooth disease type IA. Ann. Neurol. 36, 451–452. Kennerson, M. L., Zhu, D., Gardner, R. J. M., Storey, E., Merory, J., Robertson, S. P., and Nicholson, G. A. (2001). Dominant intermediate Charcot-Marie-Tooth neuropathy maps to chromosome 19p12-p13.2. Amer. J. Hum. Genet. 69, 883–888. Killian, J. M., Tiwari, P. S., Jacobson, S., Jackson, R. D., and Lupski, J. R. (1996). Longitudinal studies of the duplication form of Charcot-Marie-Tooth polyneuropathy. Muscle Nerve 19, 74–78. Kim, S. A., Taylor, G. S., Torgersen, K. M., and Dixon, J. E. (2002). Myotubularin and MTMR2, phosphatidylinositol 3-phosphatases mutated in myotubular myopathy and type 4B Charcot-Marie-Tooth disease. J. Biol. Chem. 277, 4526–4531. Kirschner, D. A., Szumowski, K., Gabreels-Festen, A. A. W. M., Hoogendijk, J. E., and Bolhuis, P. A. (1996). Inherited demyelinating peripheral neuropathies: Relating myelin packing abnormalities to P0 molecular defects. J. Neurosci. Res. 46, 502–508. Klein, C. J., Cunningham, I. M., Atkinson, E. J., Schaid, D. J., Hebbring, S. J., Anderson, S. A., Klein, D. M., Dyck, P. J. B., Litchy, W. J., Thibodeau, S. N. (2003). The gene for HMSN2C maps to 12q23-24 - A region of neuromuscular disorders. Neurology 60, 1151–1156. Kleopa, K. A., and Scherer, S. S. (2002). Inherited Neuropathies. Neurol. Clin. N. Am. 20, 679–709. Kleopa, K. A., Yum, S. W., and Scherer, S. S. (2002). Cellular mechanisms of connexin32 mutations associated with CNS manifestations. J. Neurosci. Res. 68, 522–534. Korinthenberg, R., Sauer, M., Ketelsen, U. P., Hanemann, C. O., Stoll, G., Graf, M., Baborie, A., Volk, B., Wirth, B., Rudnik-Schoneborn, S., and Zerres, K. (1997). Congenital axonal neuropathy caused by deletions in the spinal muscular atrophy region. Ann. Neurol. 42, 364–368. Krajewski, K. M., Lewis, R. A., Fuerst, D. R., Turansky, C., Hinderer, S. R., Garbern, J., Kamholz, J., and Shy, M. E. (2000). Neurological dysfunction and axonal degeneration in Charcot-Marie-Tooth disease. Brain 123, 1516–1527. Kruger, O., Plum, A., Kim, J. S., Winterhager, E., Maxeiner, S., Hallas, G., KirchhoV, S., Traub, O., Lamers, W. H., and Willecke, K. (2000). Defective vascular development in connexin 45-deWcient mice. Development 127, 4179–4193. Kuhlenbaumer, G., Young, P., Oberwittler, C., Hunermund, G., Schirmacher, A., Domschke, K., Ringelstein, B., and Stogbauer, F. (2002). Giant axonal neuropathy (GAN): Case report and two novel mutations in the gigaxonin gene. Neurology 58, 1273–1276. Kuntzer, T., Dunarnd, M., Schorderet, D. F., Vallat, J.-M., Hahn, A. F., and Bogousslavsky, J. (2003). Phenotypic expresssion of a Pro 87 to Leu mutatin in the connexin 32 gene in a large Swiss family with Charcot-Marie-Tooth neuropathy. J. Neurol. Sci., 87, 77–86. Lachat, P., Shaw, P., Gebhard, S., Van Belzen, N., Chaubert, P., and Bosman, F. T. (2002). Expression of NDRG1, a diVerentiation-related gene, in human tissues. Histochem. Cell. Biol. 118, 399–408. Lagueny, A., Latour, P., Vital, G., LeMasson, G., Rouanet, M., Ferrer, X., Vital, C., and Vandenberghe, A. (2001). Mild recurrent neuropathy in CMT1B with a novel nonsense mutation in the extracellular domain of the MPZ gene. J Neurol Neurosurg Psychiat 70, 232–235. Laporte, J., Blondeau, F., Buj-Bello, A., and Mandel, J. L. (2001). The myotubularin family: From genetic disease to phosphoinositide metabolism. Trends Genet 17, 221–228. Laporte, J., Liaubet, L., Blondeau, F., Tronchere, H., Mandel, J. L., and Payrastre, B. (2002). Functional redundancy in the myotubularin family. Biochem Biophys Res Commun 291, 305–312. Le Guern, E., Gouider, R., Mabin, D., Tardieu, S., Birouk, N., Parent, P., Bouche, P., and Brice, A. (1997). Patients homozygous for the 17p11.2 duplication in Charcot-Marie-Tooth type 1A disease. Ann. Neurol. 41, 104–108. Le Guern, E., Guilbot, A., Kessali, M., Ravise, N., Tassin, J., Maisonobe, T., Grid, D., and Brice, A. (1996). Homozygosity mapping of an autosomal recessive form of demyelinating Charcot-Marie-Tooth disease to chromosome 5q23-q33. Hum. Mol. Genet. 5, 1685–1688.

SUMMARY

Leal, A., Morera, B., DelValle, G., Heuss, D., Kayser, C., BerghoV, M., Villegas, R., Hernandez, E., Mendez, M., Hennies, H. C., Neundorfer, B., Barrantes, R., Reis, A., and Rautenstrauss, B. (2001). A second locus for an axonal form of autosomal recessive Charcot-Marie-Tooth disease maps to chromosome 19q13.3. Amer. J. Hum. Genet. 68, 269–274. Lee, M. J., Nelson, I., Houlden, H., Sweeney, M. G., Hilton-Jones, D., Blake, J., Wood, N. W., and Reilly, M. M. (2002). Six novel connexin32 (GJB1) mutations in X-linked Charcot-Marie-Tooth disease. J. Neurol. Neurosurg. Psychiat. 73, 304–306. Lee, M. K., Marzalek, J. R., and Cleveland, D. W. (1994). A mutant neuroWlament subunit causes massive, selective motor neuron death: Implications for the pathogenesis of human motor neuron disease. Neuron 13, 975–988. Lemke, G., and Axel, R. (1985). Isolation and sequence of a cDNA encoding the major structural protein of peripheral myelin. Cell 40, 501–508. Levi, G., Topilko, P., Schneider-Maunoury, S., Lasagna, M., Mantero, S., Cancedda, R., and Charnay, P. (1996). Defective bone formation in Krox-20 mutant mice. Development 122, 113–120. Lewis, R. A., Sumner, A. J., and Shy, M. E. (2000). Electrophysiological features of inherited demyelinating neuropathies: A reappraisal in the era of molecular diagnosis. Muscle Nerve 23, 1472–1487. Li, J., Krajewski, K., Shy, M. E., and Lewis, R. A. (2002a). Hereditary neuropathy with liability to pressure palsy—The electrophysiology Wts the name. Neurology 58, 1769–1773. Li, X., Lynn, B. D., Olson, C., Meier, C., Davidson, K. G. V., Yasumura, T., Rash, J. E., and Nagy, J. L. (2002b). Connexin29 expression, immunocytochemistry and freeze-fracture replica immunogold labelling (FRIL) in sciatic nerve. Eur. J. Neurosci. 16, 795–806. Lin, G. S., Glass, J. D., Shumas, S., Scherer, S. S., and Fischbeck, K. H. (1999). A unique mutation in connexin32 associated with severe, early onset CMTX in a heterozygous female. Ann. N. Y. Acad. Sci. 883, 481–485. Liu, H., Nakagawa, T., Kanematsu, T., Uchida, T., and Tsuji, S. (1999). Isolation of 10 diVerentially expressed cDNAs in diVerentiated Neuro2a cells induced through controlled expression of the GD3 synthase gene. J. Neurochem. 72, 1781–1790. Lopez-Bigas, N., Olive, M., Rabionet, R., Ben-David, O., Martinez-Matos, J. A., Bravo, O., Bachs, I., Volpini, V., Gasparini, P., Avraham, K. B., Ferrer, I., Arbones, M. L., and Estivill, X. (2001). Connexin 31 (GJB3) is expressed in the peripheral and auditory nerves and causes neuropathy and hearing impairment. Hum. Mol. Genet. 10, 947–952. Lupski, J. R. (2000). Recessive Charcot-Marie-Tooth disease. Ann. Neurol. 47, 6–8. Lupski, J. R., and Garcia, C. A. (2001). Charcot-Marie-Tooth peripheral neuropathies and related disorders. In ‘‘The Metabolic & Molecular Basis of Inherited Disease’’ vol.4 (C. R. Scriver, A. L. Beaudet, W. S. Sly, D. Valle, B. Childs, and K. W. Kinzler, eds.), 5759–5788. McGraw-Hill, New York. Lupski, J. R., Montes de Oca-Luna, R., Slaugenhaupt, S., Pentao, L., Guzzetta, V., Trask, B. J., SaucedoCardenas, O., Barker, D. F., Chakravarti, A., and Patel, P. I. (1991). DNA duplication associated with Charcot-Marie-Tooth disease type IA. Cell 66, 219–232. Lynch, D. R., Hara, H., Yum, S., Chance, P. J., Scherer, S. S., Bird, S. J., and Fischbeck, K. H. (1997). Autosomal dominant transmission of Dejerine-Sottas disease (HMSN III). Neurology 49, 601–603. Madrid, R., Bradley, W. G., and Davis, D. J. F. (1977). The peroneal muscular atrophy syndrome—clinical, genetic, electrophysiological and nerve biopsy studies. 2. Observations on pathological changes in sural nerve biopsies. J. Neurol. Sci. 32, 91–122. Magyar, J. P., Martini, R., Ruelicke, T., Aguzzi, A., Adlkofer, K., Dembic, Z., Zielasek, J., Toyka, K. V., and Suter, U. (1996). Impaired diVerentiation of Schwann cells in transgenic mice with increased PMP22 gene dosage. J. Neurosci. 16, 5351–5360. Mandich, P., Mancardi, G. L., Varese, A., Soriani, S., DiMaria, E., Bellone, E., Bado, M., Gross, L., Windebank, A. J., Ajmar, F., and Schenone, A. (1999). Congenital hypomyelination due to myelin protein zero Q215X mutation. Ann. Neurol. 45, 676–678. ManWoletti, G., Ruaro, M. E., Del Sal, G., Philipson, L., and Schneider, C. (1990). A growth arrest-speciWc (gas) gene codes for a membrane protein. Mol. Cell. Biol. 10, 2914–2930. Marks, H. G., Scavina, M. T., Kolodny, E. H., Palmieri, M., and Childs, J. (1997). Krabbe’s disease presenting as a peripheral neuropathy. Muscle Nerve 20, 1024–1028. Marques, W., Sweeney, M. G., Wood, N. W., and Wroe, S. J. (1999). Central nervous system involvement in a novel connexin 32 mutation aVecting identical twins. J. Neurol. Neurosurg. Psychiat. 66, 803–804. Marrosu, M. G., Vaccargiu, S., Marrosu, G., Vannelli, A., Cianchetti, C., and Muntoni, F. (1998). CharcotMarie-Tooth disease type 2 associated with mutation of the myelin protein zero gene. Neurology 50, 1397–1401. Martini, R. (1997). Animal models for inherited peripheral neuropathies. J. Anat. 191, 321–336. Martini, R., Mohajeri, M. H., Kasper, S., Giese, K. P., and Schachner, M. (1995a). Mice doubly deWcient in the genes for P0 and myelin basic protein show that both proteins contribute to the formation of the major dense line in peripheral nerve myelin. J. Neurosci. 15, 4488–4495. Martini, R., Zielasek, J., Toyka, K. V., Giese, K. P., and Schachner, M. (1995b). Protein zero (P0)-deWcient mice show myelin degeneration in peripheral nerves characteristic of inherited human neuropathies. Nat. Genet. 11, 281–285. Matsunami, N., Smith, B., Ballard, L., Lensch, M. W., Robertson, M., Albertsen, H., Hanemann, C. O., Muller, H. W., Bird, T. D., White, R., and Chance, P. F. (1992). Peripheral myelin protein-22 gene maps in the duplication in chromosome-17p11.2 associated with Charcot-Marie-Tooth-1A. Nat. Genet. 1, 176–179.

943

944

39. INHERITED NEUROPATHIES

Matsuyama, W., Nakagawa, M., Takashima, H., and Osame, M. (2002). Altered traYcking and adhesion function of MPZ mutations and phenotypes of Charcot-Marie-Tooth disease 1B. Acta Neuropathol. 103, 501–508. Maurer, M., Kobsar, I., BerghoV, M., Schmid, C. D., Carenini, S., and Martini, R. (2002). Role of immune cells in animal models for inherited neuropathies: Facts and visions. J. Anat. 200, 405–414. McEntagart, M., Norton, N., Williams, H., Teare, M. D., Dunstan, M., Baker, P., Houlden, H., Reilly, M., Wood, N., Harper, P. S., Futreal, P. A., Williams, N., and Rahman, N. (2001). Localization of the gene for distal hereditary motor neuronopathy VII (DHMN-VII) to chromosome 2q14. Amer. J. Hum. Genet. 68, 1270–1276. Mendell, J. R., Sahenk, Z., Whitaker, J. N., Trapp, B. D., Yates, A. J., Griggs, R. C., and Quarles, R. H. (1985). Polyneuropathy and IgM monoclonal gammopathy: Studies on the pathogenetic role of anti-myelin-associated glycoprotein antibody. Ann. Neurol. 17, 243–254. Menichella, D. M., Goodenough, D. A., Sirkowski, E., Scherer, S. S., Paul, D. L., (2003). Connexins are critical for normal myelination in the central nervous system. J. Neurosci. 23, 5963–5973. Meretoja, P., Silander, K., Kamino, H., Aula, P., Meretoja, A., and Savontaus, M.-L. (1997). Epidemiology of herediatry neuropathy with liability to pressure palises (HNPP) in southwestern Finland. Neuromusc. Disord. 7, 529–532. Mersiyanova, I. V., Ismailov, S. M., Polyakov, A. V., Dadali, E. L., Fedotov, V. P., Nelis, E., Lofgren, A., Timmerman, V., Van Broeckhoven, C., and Evgrafov, O. V. (2000a). Screening for mutations in the peripheral myelin genes PMP22, MPZ and Cx32 (GJB1) in Russian Charcot-Marie-Tooth neuropathy patients. Hum. Mutat. 15, 340–347. Mersiyanova, I. V., Perepelov, A. V., Polyakov, A. V., Sitnikov, V. F., Dadali, E. L., Oparin, R. B., Petrin, A. N., and Evgrafov, O. V. (2000b). A new variant of Charcot-Marie-Tooth disease type 2 is probably the result of a mutation in the neuroWlament-light gene. Amer. J. Hum. Genet. 67, 37–46. Mihalik, S. J., Morrell, J. C. D., Sacksteder, K. A., Watkins, P. A., and Gould, S. J. (1997). IdentiWcation of PAHX, a Refsum disease gene. Nat. Genet. 17, 185–189. Misko, A., Ferguson, T., and Notterpek, L. (2002). Matrix metalloproteinase mediated degradation of basement membrane proteins in Trembler J neuropathy nerves. J. Neurochem. 83, 885–894. Misu, K., Yoshihara, T., Shikama, Y., Awaki, E., Yamamoto, M., Hattori, N., Hirayama, M., Takegami, T., Nakashima, K., and Sobue, G. (2000). An axonal form of Charcot-Marie-Tooth disease showing distinctive features in association with mutations in the peripheral myelin protein zero gene (Thr124Met or Asp75Val). J. Neurol. Neurosurg. Psychiat. 69, 806–811. Mok, H., Shin, H., Kim, S., Lee, J. R., Yoon, J., and Kim, E. (2002). Association of the kinesin superfamily motor protein KIF1Balpha with postsynaptic density-95 (PSD-95), synapse-associated protein-97, and synaptic scaVolding molecule PSD-95/discs large/zona occludens-1 proteins. J. Neurosci. 22, 5253–5258. Montag, D., Giese, K. P., Bartsch, U., Martini, R., Lang, Y., Bluthmann, H., Karthigasan, J., Kirschner, D. A., Wintergerst, E. S., Nave, K.-A., Zielasek, J., Toyka, K. V., Lipp, H.-P., and Schachner, M. (1994). Mice deWcient for the myelin-associated glycoprotein show subtle abnormalities in myelin. Neuron 13, 229–246. Mori, K. (2000). Tripartite management of unfolded proteins in the endoplasmic reticulum. Cell 101, 451–454. Moriwaki, Y., Begum, N. A., Kobayashi, M., Matsumoto, M., Toyoshima, K., and Seya, T. (2001). Mycobacterium bovis bacillus Calmette-Guerin and its cell wall complex induce a novel lysosomal membrane protein, SIMPLE, that bridges the missing link between lipopolysaccharide and p53-inducible gene, LITAF (PIG7) and estrogen-inducible gene, EET-1. J. Biol. Chem. 276, 23065–23076. Morocutti, C., Colazza, G. B., Soldati, G., DAlessio, C., Damiano, M., Casali, C., and Pierelli, F. (2002). Charcot-Marie-Tooth disease in Molise, a central-southern region of Italy: An epidemiological study. Neuroepidemiology 21, 241–245. Mostacciuolo, M. L., Righetti, E., Zortea, M., Bosello, V., Schiavon, F., Vallo, L., Merlini, L., Siciliano, G., Fabrizi, G. M., Rizzuto, N., Milani, M., Baratta, S., and Taroni, F. (2001). Charcot-Marie-Tooth disease type I and related demyelinating neuropathies: Mutation analysis in a large cohort of Italian families. Hum. Mutat. 18, 32–41. Mouton, P., Tardieu, S., Gouider, R., Birouk, N., Maisonobe, T., Dubourg, O., Brice, A., Le Guern, E., and Bouche, P. (1999). Spectrum of clinical and electrophysiological features in HNPP patients with the 17p11.2 deletion. Neurology 52, 1440–1446. Murphy, P., Topilko, P., Schneider-Maunoury, S., Seitanidou, T., Baron-Van Evercooren, A., and Charnay, P. (1996). The regulation of Krox-20 expression reveals important steps in the control of peripheral glial cell development. Development 122, 2847–2857. Musso, M., Balestra, P., Bellone, E., Cassandrini, D., Di Maria, E., Doria, L. L., Grandis, M., Mancardi, G., Schenone, A., Levi, G., Ajmar, F., and Mandar, P. (2001). The D355V mutation decreases EGR2 binding to an element within the Cx32 promoter. Neurobiol. Dis. 8, 700–706. Naef, R., Adlkofer, K., Lescher, B., and Suter, U. (1997). Aberrant protein traYcking in Trembler suggests a disease mechanism for hereditary human peripheral neuropathies. Mol. Cell. Neurosci. 9, 13–25. Naef, R., and Suter, U. (1999). Impaired intracellular traYcking is a common disease mechanism of PMP22 point mutations in peripheral neuropathies. Neurobiol. Dis. 6, 1–14. Nagaoka, T., Oyamada, M., Okajima, S., and Takamatsu, T. (1999). DiVerential expression of gap junction proteins connexin26, 32, and 43 in normal and crush-injured rat sciatic nerves: Close relationship between connexin43 and occludin in the perineurium. J. Histochem. Cytochem. 47, 937–948.

SUMMARY

Nagarajan, R., Le, N., Mahoney, H., Araki, T., and Milbrandt, J. (2002). Deciphering peripheral nerve myelination by using Schwann cell expression proWling. Proc Nat Acad Sci USA 99, 8998–9003. Nagarajan, R., Svaren, J., Le, N., Araki, T., Watson, M., and Milbrandt, J. (2001). EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30, 355–368. Nakagawa, M., Takashima, H., Umehara, F., Arimura, K., Miyashita, F., Takenouchi, N., Matsuyama, W., and Osame, M. (2001). Clinical phentoype in X-linked Charcot-Marie-Tooth disease with an entire deletion of the connexin 32 coding sequence. J. Neurol. Sci. 185, 31–36. Nave, K.-A. (2001). Myelin-speciWc genes and their mutations in the mouse. In ‘‘Glial Cell Development’’ (K. R. Jessen and W. D. Richardson, eds.), 177–208. Oxford University Press, Oxford. Nelis, E., Erdem, S., Tan, E., Lofgren, A., Ceuterick, C., De Jonghe, P., Van Broeckhoven, C., Timmerman, V., and Topaloglu, H. (2002a). A novel homozygous missense mutation in the myotubularin-related protein 2 gene associated with recessive Charcot-Marie-Tooth disease with irregularly folded myelin sheaths. Neuromuscul Disord 12, 869–873. Nelis, E., Erdem, S., Van den Bergh, P. Y. K., Belpaire-Dethiou, M.-C., Ceuterick, C., Van Gerwen, V., Cuesta, A., Pedrola, L., Palau, F., Gabreels-Festen, A. A. W. M., Verellen, C., Tan, E., Demirci, M., Van Broeckhoven, C., De Jonghe, P., Topaloglu, H., and Timmerman, V. (2002b). Mutations in GDAP1. Autosomal recessive CMT with demyelination and axonopathy. Neurology 59, 1865–1872. Nelis, E., Haites, N., and Van Broeckhoven, C. (1999). Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum. Mutat. 13, 11–28. Nelis, E., Holmberg, B., Adolfsson, R., Holmgren, G., and VanBroeckhoven, C. (1997). PMP22 Thr(118)Met: Recessive CMT1 mutation or polymorphism? Nat. Genet. 15, 13–14. Nelis, E., Van Broeckhoven, C., and co-authors (1996). Estimation of the mutation frequencies in Charcot-MarieTooth disease type 1 and hereditary neuropathy with liability to pressure palsies: A European collaborative study. Eur. J. Human Genet. 4, 25–33. Nelson, J. W., and Amick, L. D. (1966). Heredofamilial progressive spinal muscular atrophy: A clinical and electromyographic study of a kinship. Neurology 16, 306. Nicholson, G., and Nash, J. (1993). Intermediate nerve conduction velocities deWne X-linked Charcot-MarieTooth neuropathy families. Neurology 43, 2558–2564. Nicholson, G., Valentijn, L. J., Cherryson, A. K., Kennerson, M. L., Bragg, T. L., DeKroon, R. M., Ross, D. A., Pollard, J. D., McLeod, J. G., Bolhuis, P. A., and Baas, F. (1994). A frame shift mutation in the PMP22 gene in hereditary neuropathy with liability to pressure palsies. Nat. Genet. 6, 263–266. Nicholson, G. A., Yeung, L., and Corbett, A. (1998). EYcient neurophysiological selection of X-linked CharcotMarie-Tooth families. Neurology 51, 1412–1416. Niemann, S., Sereda, M. W., Suter, U., GriYths, I. R., and Nave, K. A. (2000). Uncoupling of myelin assembly and Schwann cell diVerentiation by transgenic overexpression of peripheral myelin protein 22. J. Neurosci. 20, 4120–4128. Nishimura, T., Yoshikawa, H., Fujimura, H., Sakoda, S., and Yanagihara, T. (1996). Accumulation of peripheral myelin protein 22 in onion bulbs and Schwann cells of biopsied nerves from patients with Charcot-MarieTooth disease type 1A. Acta Neuropathol. 92, 454–460. Nobbio, L., Mancardi, G., Grandis, M., Levi, G., Surer, U., Nave, K. A., Windebank, A. J., Abbruzzese, M., and Schenone, A. (2001). PMP22 transgenic dorsal root ganglia cultures show myelin abnormalities similar to those of human CMT1A. Ann. Neurol. 50, 47–55. Notterpek, L., Ryan, M. C., Tobler, A. R., and Shooter, E. M. (1999a). PMP22 accumulation in aggresomes: Implications for CMT1A pathology. Neurobiol. Dis. 6, 450–460. Notterpek, L., Shooter, E. M., and Snipes, G. J. (1997). Upregulation of the endosomal-lysosomal pathway in the Trembler-J neuropathy. J. Neurosci. 17, 4190–4200. Notterpek, L., Snipes, G. J., and Shooter, E. M. (1999b). Temporal expression pattern of peripheral myelin protein 22 during in vivo and in vitro myelination. Glia 25, 358–369. Numakura, C., Lin, C. Q., Ikegami, T., Guldberg, P., and Hayasaka, K. (2002). Molecular analysis in Japanese patients with Charcot-Marie-Tooth disease: DGGE analysis for PMP22, MPZ, and Cx32/GJB1 mutations. Hum. Mut. 20, 392–398. Numakura, C., Lin, C. Q., Oka, N., Akiguchi, I., and Hayasaka, K. (2000). Hemizygous mutation of the peripheral myelin protein 22 gene associated with Charcot-Marie-Tooth disease type 1. Ann. Neurol. 47, 101–103. Odermatt, B., Wellershaus, K., Wallraff, A., Seifert, G., Degen, G., Euwens, C., Fuss, B., Bussow, H., Schilling, K., Stenhauser, C., Willecke, K. (2003). Connexin 47 (Cx47)-deWcient mice with enhanced green fluorescent protein reporter gene reveal predominant oligodendrocytic expression of Cx47 and display vacuolized myelin in the CNS. J. Neurosci. 23, 4549–4559. Oh, S., Ri, Y., Bennett, M. V. L., Trexler, E. B., Verselis, V. K., and Bargiello, T. A. (1997). Changes in permeability caused by connexin 32 mutations underlie X-linked Charcot-Marie-Tooth disease. Neuron 19, 927–938. Ohara, O., Gahara, Y., Miyake, T., Teraoka, H., and Kitamura, T. (1993). NeuroWlament deWciency in quail caused by nonsense mutation in neuroWlament-L gene. J. Cell Biol. 121, 387–395. Ohta, M., Ellefson, R. D., Lambert, E. H., and Dyck, P. J. (1973). Hereditary sensory neuropathy, type II. Clinical, electrophysiologic, histologic, and biochemical studies of a Quebec kinship. Arch. Neurol. 29, 23–37. Omori, Y., Mesnil, M., and Yamasaki, H. (1996). Connexin 32 mutations from X-linked Charcot-Marie-Tooth disease patients: Functional defects and dominant negative eVects. Mol. Biol. Cell 7, 907–916.

945

946

39. INHERITED NEUROPATHIES

Ouvrier, R. (1996). Correlation between the histopathologic, genotypic, and phenotypic features of hereditary peripheral neuropathies in childhood. J. Child Neurol. 11, 133–146. Ouvrier, R. A., McLeod, J. G., and Conchin, T. E. (1987). The hypertrophic forms of hereditary motor and sensory neuropathy. A study of hypertrophic Charcot-Marie-Tooth disease (HMSN type I) and DejerineSottas disease (HMSN type III) in childhood. Brain 110, 121–148. Ouvrier, R. A., McLeod, J. G., Morgan, G. J., Wise, G. A., and Conchin, T. E. (1981). Hereditary motor and sensory neuropathy of neuronal type with onset in early childhood. J. Neurol. Sci. 51, 181–197. Ouvrier, R. A., McLeod, J. G., and Pollard, J. D. (1990). ‘‘Peripheral Neuropathy in Childhood.’’ Raven Press, New York. Palumbo, C., Massa, R., Panico, M. B., DiMuzio, A., Sinibaldi, P., Bernardi, G., and Modesti, A. (2002). Peripheral nerve extracellular matrix remodeling in Charcot-Marie-Tooth type 1 disease. Acta Neuropathol. 104, 287–296. Panas, M., Karadimas, C., Avramopoulos, D., and Vassilopoulos, D. (1998). Central nervous system involvement in four patients with Charcot-Marie-Tooth disease with connexin 32 extracellular mutations. J. Neurol. Neurosurg. Psychiat. 65, 947–948. Paratore, C., Goerich, D. E., Suter, U., Wegner, M., and Sommer, L. (2001). Survival and glial fate acquisition of neural crest cells are regulated by an interplay between the transcription factor Sox10 and extrinsic combinatorial signaling. Development 128, 3949–3961. Pareek, S., Notterpek, L., Snipes, G. J., Naef, R., Sossin, W., Laliberte, J., Iacampo, S., Suter, U., Shooter, E. M., and Murphy, R. A. (1997). Neurons promote the translocation of peripheral myelin protein 22 into myelin. J. Neurosci. 17, 7754–7762. Pareek, S., Suter, U., Snipes, G. J., Welcher, A. A., Shooter, E. M., and Murphy, R. A. (1993). Detection and processing of peripheral myelin protein PMP22 in cultured Schwann cells. J. Biol. Chem. 268, 10372–10379. Pareyson, D., Scaioli, V., Taroni, F., Botti, S., Lorenzetti, D., Solari, A., Ciano, C., and Sghirlanzoni, A. (1996). Phenotypic heterogeneity in hereditary neuropathy with liability to pressure palsies associated with chromosome 17p11.2–12 deletion. Neurology 46, 1133–1137. Parman, Y., Plante-Bordeneuve, V., Guiochon-Mantel, A., Eraksoy, M., and Said, G. (1999). Recessive inheritance of a new point mutation of the PMP22 gene in Dejerine-Sottas disease. Ann. Neurol. 45, 518–522. Parmantier, E., Cabon, F., Braun, C., Durso, D., Mu¨ller, H. W., and Zalc, B. (1995). Peripheral myelin protein-22 is expressed in rat and mouse brain and spinal cord motoneurons. Eur. J. Neurosci. 7, 1080–1088. Patel, P. I., Roa, B. B., Welcher, A. A., Schoenerscott, R., Trask, B. J., Pentao, L., Snipes, G. J., Garcia, C. A., Francke, U., Shooter, E. M., Lupski, J. R., and Suter, U. (1992). The gene for the peripheral myelin proteinPMP-22 is a candidate for Charcot-Marie-Tooth disease type-1A. Nat. Genet. 1, 159–165. Paulson, H., Garbern, J. Y., Hoban, T. F., Krajewski, K. M., Lewis, R. A., Fischbeck, K. H., Grossman, R. I., Lenkiski, R., Kamholz, J. A., and Shy, M. E. (2002). Transient CNS white matter abnormality in X-linked Charcot-Marie-Tooth disease. Ann. Neurol. 52, 429–434. Peirano, R. I., Goerich, D. E., Riethmacher, D., and Wegner, M. (2000). Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol. Cell. Biol. 20, 3198–3209. Perez-Olle, R., Leung, C. L., and Liem, R. K. H. (2002). EVects of Charcot-Marie-Tooth-linked mutations of the neuroWlament light subunit on intermediate Wlament formation. J. Cell Sci. 115, 4937–4946. Perkins, C. S., Aguayo, A. J., and Bray, G. M. (1981). Schwann cell multiplication in trembler mice. Neuropathol. Appl. Neurobiol. 7, 115–126. Petiot, A., Ogier-Denis, E., Blommaart, E. F., Meijer, A. J., and Codogno, P. (2000). Distinct classes of phosphatidylinositol 3’-kinases are involved in signaling pathways that control macroautophagy in HT-29 cells. J. Biol. Chem. 275, 992–998. Phillips, J. P., Warner, L. E., Lupski, J. R., and Garg, B. P. (1999). Congenital hypomyelinating neuropathy: Two patients with long-term follow-up. Pediat. Neurol. 20, 226–232. Pingault, V., Bondurand, N., LeCaignec, C., Tardieu, S., Lemort, N., Dubourg, O., Le Guern, E., Goossens, M., and BoespXug-Tanguy, O. (2001). The SOX10 transcription factor: Evaluation as a candidate gene for central and peripheral hereditary myelin disorders. J. Neurol. 248, 496–499. Pingault, V., Guiochon-Mantel, A., Bondurand, N., Faure, C., Lacroix, C., Lyonnet, S., Goosens, M., and Landrieu, P. (2000). Peripheral neuropathy with hypomyelination, chronic intestinal pseudo-obstruction and deafness: A developmental neural crest syndrome related to a SOX10 mutation. Ann. Neurol. 48, 671–676. Plante-Bordeneuve, V., and Said, G. (2002). Dejerine-Sottas disease and hereditary demyelinating neuropathy of infancy. Muscle Nerve 26, 608–621. Previtali, S. C., Quattrini, A., Fasolini, M., Panzeri, M. C., Villa, A., Filbin, M. T., Li, W. H., Chiu, S. Y., Messing, A., Wrabetz, L., and Feltri, M. L. (2000). Epitope-tagged P0 glycoprotein causes Charcot-MarieTooth-like neuropathy in transgenic mice. J. Cell Biol. 151, 1035–1045. Quattrone, A., Gambardella, A., Bono, F., Aguglia, U., Bolino, A., Bruni, A. C., Montesi, M. P., Oliveri, R. L., Sabatelli, M., Tamburrini, O., Valentino, P., Van Broeckhoven, C., and Zappia, M. (1996). Autosomal recessive hereditary motor and sensory neuropathy with focally folded myelin sheaths: Clinical, electrophysiologic, and genetic aspects of a large family. Neurology 46, 1318–1324. Raeymaekers, P., Timmerman, V., Nelis, E., De Jonghe, P., Hoogendijk, J., Baas, F., Barker, D. F., Martin, J. J., De Visser, M., Bolhius, P. A., and Van Broeckhoven, C. (1991). Duplication in chromosome 17p11.2 in Charcot-Marie-Tooth neuropathy type 1a (CMT1a). Neuromusc. Disord. 1, 93–97.

SUMMARY

Reuber, B. E., Germain-Lee, E., Collins, C. S., Morrell, J. C., Ameritunga, R., Moser, H. W., Valle, D., and Gould, S. J. (1997). Mutations in PEX1 are the most common cause of peroxisome biogenesis disorders. Nat. Genet. 17, 445–448. Roa, B. B., Dyck, P. J., Marks, H. G., Chance, P. F., and Lupski, J. R. (1993a). Dejerine-Sottas syndrome associated with point mutation in the peripheral myelin protein 22 (PMP22) gene. Nat. Genet. 5, 269–273. Roa, B. B., Garcia, C. A., Pentai, L., Killian, J. M., Trask, B. J., Suter, U., Snipes, G. J., Oritz-Lopez, R., Shooter, E. M., Patel, P. I., and Lupski, J. R. (1993b). Evidence for a recessive PMP22 point mutation in Charcot-Marie-Tooth disease type IA. Nat. Genet. 5, 189–194. Robertson, A. M., Perea, J., McGuigan, A., King, R. H. M., Muddle, J. R., Gabreels-Festen, A. A., Thomas, P. K., and Huxley, C. (2002). Comparison of a new pmp22 transgenic mouse line with other mouse models and human patients with CMT1A. J. Anat. 200, 377–390. Rogers, T., Chandler, D., Angelicheva, D., Thomas, P. K., Youl, B., Tournev, I., Gergelcheva, V., and Kalaydjieva, L. (2000). A novel locus for autosomal recessive peripheral neuropathy in the EGR2 region on 10q23. Amer J Hum Genet 67, 664–671. Ryan, M. C., Shooter, E. M., and Notterpek, L. (2002). Aggresome formation in neuropathy models based on peripheral myelin protein 22 mutations. Neurobiol. Dis. 10, 109–118. Saito, M., Hayashi, Y., Suzuki, T., Tanaka, H., Hozumi, I., and Tsuji, S. (1997). Linkage mapping of the gene for Charcot-Marie-Tooth disease type 2 to chromosome 1p (CMT2A) and the clinical features of CMT2A. Neurology 49, 1630–1635. Sancho, S., Magyar, J. P., Aguzzi, A., and Suter, U. (1999). Distal axonopathy in peripheral nerves of PMP22 mutant mice. Brain 122, 1563–1577. Sancho, S., Young, P., and Suter, U. (2001). Regulation of Schwann cell proliferation and apoptosis in PMP22deWcient mice and mouse models of Charcot-Marie-Tooth disease type 1A. Brain 124, 2177–2187. Sander, S., Nicholson, G. A., Ouvrier, R. A., McLeod, J. G., and Pollard, J. D. (1998). Charcot-Marie-Tooth disease: Histopathological features of the peripheral myelin protein (PMP22) duplication (CMT1A) and connexin32 mutations (CMTX1). Muscle Nerve 21, 217–225. Schenone, A., Nobbio, L., Caponnetto, C., Abbruzzese, M., Mandich, P., Bellone, E., Ajmar, F., Gherardi, G., Windebank, A. J., and Mancardi, G. (1997). Correlation between PMP-22 messenger RNA expression and phenotype in hereditary neuropathy with liability to pressure palsies. Ann. Neurol. 42, 866–872. Scherer, S. S. (1997). Molecular genetics of demyelination: New wrinkles on an old membrane. Neuron 18, 13–16. Scherer, S. S., Bone, L. J., Abel, A., Descheˆnes, S. M., Balice-Gordon, R., and Fischbeck, K. (1999). The role of the gap junction protein connexin32 in the pathogenesis of X-linked Charcot-Marie-Tooth disease. In ‘‘Gap Junction-Mediated Intercellular Signalling in Health and Disease’’ (G. Cardew, ed., 175–185. John Wiley & Sons, New York. Scherer, S. S., and Salzer, J. (2001). Axon-Schwann cell interactions in peripheral nerve degeneration and regeneration. In ‘‘Glial Cell Development,’’ (K. R. Jessen and W. D. Richardson, eds.), 299–330. Oxford University Press, Oxford. Scherer, S. S., Wang, D.-y., Kuhn, R., Lemke, G., Wrabetz, L., and Kamholz, J. (1994). Axons regulate Schwann cell expression of the POU transcription factor SCIP. J. Neurosci. 14, 1930–1942. Scherer, S. S., Xu, Y.-T., Bannerman, P., Sherman, D. L., and Brophy, P. J. (1995). Periaxin expression in myelinating Schwann cells: Modulation by axon-glial interactions and polarized localization during development. Development 121, 4265–4273. Scherer, S. S., Xu, Y.-T., Nelles, E., Fischbeck, K., Willecke, K., and Bone, L. J. (1998). Connexin32-null mice develop a demyelinating peripheral neuropathy. Glia 24, 8–20. Schneider-Maunoury, S., Topilko, P., Switanidou, T., Levi, G., Cohen-Tannoudji, M., Pournin, S., Babinet, C., and Charnay, P. (1993). Disruption of Krox-20 results in alteration of rhombomeres 3 and 5 in the developing hindbrain. Cell 75, 1199–1214. Schneider-Schaulies, J., von Brunn, A., and Schachner, M. (1990). Recombinant peripheral myelin protein Po confers both adhesion and neurite outgrowth-promoting properties. J. Neurosci. Res. 27, 286–297. Seeman, P., Huehne, K., Rautenstrauss, B., Mazanec, R., Suslvkova, P., and Keller, O. (2001). An axonal form of Charcot-Marie-Tooth disease beginning with deafness and abnormal pupillar reaction: New mutation in the myelin protein zero gene (MPZ) in three generations of a Czech family. Nervenheilkunde 20, S28. Sellman, M. S., and Mayer, R. F. (1987). Conduction block in hereditary neuropathies with susceptability to pressure palsies. Muscle Nerve 10, 621–625. Senderek, J., Bergmann, C., Weber, S., Ketelsen, U. P., Schorle, H., Rudnik-Schoneborn, S., Buttner, R., Buchheim, E., and Zerres, K. (2003). Mutation of the SBF2 gene, encoding a novel member of the myotubularin family, in Charcot-Marie-Tooth neuropathy type 4B2/11p15. Hum. Mol. Genet. 12, 349–356. Senderek, J., Hermanns, B., Lehmann, U., Bergmann, C., Marx, G., Kabus, C., Timmerman, V., StoltenburgDidinger, G., and Schro¨der, J. M. (2000). Charcot-Marie-Tooth neuropathy type 2 and P0 point mutations: Two novel amino acid substitutions (Asp61Gly; Tyr119Cys) and a possible hotspot on Thr124Met. Brain Pathol. 10, 235–248. Senderek, J., Ramaekers, V. T., Zerres, K., Rudnik-Schoneborn, S., Schro¨der, J. M., and Bergmann, C. (2001). Phenotypic variation of a novel nonsense mutation in the P0 intracellular domain. J. Neurol. Sci. 192, 49–51. Sereda, M., GriYths, I., Puhlhofer, A., Stewart, H., Rossner, M. J., Zimmermann, F., Magyar, J. P., Schneider, A., Hund, E., Meinck, H. M., Suter, U., and Nave, K. A. (1996). A transgenic rat model of Charcot-MarieTooth disease. Neuron 16, 1049–1060.

947

948

39. INHERITED NEUROPATHIES

Sghirlanzoni, A., Pareyson, D., Balestrini, M. R., Bellone, E., Berta, E., Ciano, C., Mandich, P., and Marazzi, R. (1992). HMNS III phenotype due to homozygous expression of a dominant HMNS II gene. Neurology 42, 2201–2203. Shapiro, L., Doyle, J. P., Hensley, P., Colman, D. R., and Hendrickson, W. A. (1996). Crystal stucture of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17, 435–449. Shaw, E., McCue, L. A., Lawrence, C. E., and Dordick, J. S. (2002). IdentiWcation of a novel class in the alpha/ beta hydrolase fold superfamily: the N-myc diVerentiation-related proteins. Proteins 47, 163–168. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J. (2001). SpeciWc disruption of a Schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron 30, 677–687. Shy, M. E., Arroyo, E., Sladky, J., Menichella, D., Jiang, H., Kamholz, J., and Scherer, S. S. (1997). Heterozygous P0 knockout mice develop a peripheral neuropathy that resembles chronic inXammatory demyelinating polyneuropathy (CIDP). J. Neuropathol. Exp. Neurol. 56, 811–821. Silander, K., Meretoja, P., Juvonen, V., Ignatius, J., Pihko, H., Saarinen, A., Wallden, T., Herrgard, E., Aula, P., and Savontaus, M. L. (1998). Spectrum of mutations in Finnish patients with Charcot-Marie-Tooth disease and related neuropathies. Hum. Mutat. 12, 59–68. Simonsen, A., Wurmser, A. E., Emr, S. D., and Stenmark, H. (2001). The role of phosphoinositides in membrane transport. Curr. Opin. Cell Biol. 13, 485–492. Skre, H. (1974). Genetic and clinical aspects of Charcot-Marie-Tooth’s disease. Clin. Genet. 6, 98–118. Slaugenhaupt, S. A., Blumenfeld, A., Gill, S. P., Leyne, M., Mull, J., Cuajungco, M. P., Liebert, C. B., Chadwick, B., Idelson, M., Reznik, L., Robbins, C. M., Makalowska, I., Brownstein, M. J., Krappmann, D., Scheidereit, C., Maayan, C., Axelrod, F. B., and Gusella, J. F. (2001). Tissue-speciWc expressin of a splicing mutatin in the IKBKAP gene causes familial dysautonomia. Amer. J. Hum. Genet. 68, 598–605. Snipes, G. J., Suter, U., Welcher, A. A., and Shooter, E. M. (1992). Characterization of a novel peripheral nervous system myelin protein (PMP-22/SR13). J. Cell Biol. 117, 225–238. Southard-Smith, E. M., Kos, L., and Pavan, W. J. (1998). Sox10 mutation disrupts neural crest development in DOM Hirschsprung mouse model. Nat. Genet. 18, 60–64. Spreyer, P., Kuhn, G., Hanemann, C. O., Gillen, C., Schaal, H., Kuhn, R., Lemke, G., and Mu¨ller, H. W. (1991). Axon-regulated expression of a Schwann cell transcript that is homologous to a ‘‘growth-arrest-speciWc’’ gene. EMBO J. 10, 3661–3668. Stahl, N., Harry, J., and Popko, B. (1990). Quantitative analysis of myelin protein gene expression during development in the rat sciatic nerve. Mol. Brain Res. 8, 209–212. Stojkovic, T., Latour, P., Vandenberghe, A., Hurtevent, J. F., and Vermersch, P. (1999). Sensorineural deafness in X-linked Charcot-Marie-Tooth disease with connexin 32 mutations (R142Q). Neurology 52, 1010–1014. Stolt, C. C., Rehberg, S., Ader, M., Lommes, P., Riethmacher, D., Schachner, M., Bartsch, U., and Wegner, M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev. 16, 165–170. Street, V. A., Bennett, C. L., Goldy, J. D., Shirk, A. J., Kleopa, K. A., Tempel, B. L., Lipe, H. P., Scherer, S. S., Bird, T. D., and Chance, P. F. (2003). Mutations in LITAF (SIMPLE) associated with Charcot-Marie-Tooth disease Type 1C. Neurology 60, 22–26. Street, V. A., Goldy, J. D., Golden, A. S., Tempel, B. L., Bird, T. D., and Chance, P. F. (2002). Mapping of Charcot-Marie-Tooth disease type 1C to chromosome 16p identiWes a novel locus for demyelinating neuropathies. Amer. J. Hum. Genet. 70, 244–250. Sullivan, T., Escalante-Alcalde, D., Bhatt, H., Anver, M., Bhat, N., Nagashima, K., Stewart, C. L., and Burke, B. (1999). Loss of A-type lamin expression compromises nuclear envelope integrity leading to muscular dystrophy. J Cell Biol 147, 913–920. Suter, U., Moskow, J. J., Welcher, A. A., Snipes, G. J., Kosaras, B., Sidman, R. L., Buchberg, A. M., and Shooter, E. M. (1992a). A leucine-to-proline mutation in the putative 1st transmembrane domain of the 22kDa peripheral myelin protein in the Trembler-J mouse. Proc. Natl. Acad. Sci. USA 89, 4382–4386. Suter, U., Snipes, G. J., Schoener-Scott, R., Welcher, A. A., Pareek, S., Lupski, J. R., Murphy, R. A., Shooter, E., and Patel, P. I. (1994). Regulation of tissue-speciWc expression of alternative peripheral myelin protein-22 (PMP22) gene transcripts by two promoters. J. Biol. Chem. 269, 25795–25808. Suter, U., Welcher, A. A., Ozcelik, T., Snipes, G. J., Kosaras, B., Francke, U., Billings-Gagliardi, S., Sidman, R. L., and Shooter, E. M. (1992b). Trembler mouse carries a point mutation in a myelin gene. Nature 356, 241–244. Suzuki, M., Sakamoto, Y., Kitamura, K., Fukunaga, K., Yamamoto, H., Miyamoto, E., and Uyemara, K. (1990). Phosphorylation of P0 glycoprotein in peripheral nerve myelin. J. Neurochem. 55, 1966–1971. Svaren, J., Sevetson, B. R., Apel, E., Zimonjic, D. B., Popescu, N. C., and Milbrandt, J. (1996). NAB2, a corepressor of NGFI-A (Erg-1) and Krox20, is induced by proliferative and diVerentiative stimuli. Mol. Cell. Biol. 16, 3545–3553. Svaren, J., Sevetson, B. R., Stanton, J. J., SwirnoV, A. H., and Milbrandt, J. (1998). Novel mutants of NAB corepressors enhance activation by Erg transactivators. EMBO J. 17, 6010–6019. Swiatek, P. J., and Gridley, T. (1993). Perinatal lethality and defects in hindbrain development in mice homozygous for a targeted mutation of the zinc Wnger gene Krox20. Genes Dev. 7, 2071–2084. Tabaraud, F., Lagrange, E., Sindou, P., Vandenberghe, A., Levy, N., and Vallat, J. M. (1999). Demyelinating X-linked Charcot-Marie-Tooth disease: Unusual electrophysiological Wndings. Muscle Nerve 22, 1442–1447. Tachi, N., Ishikawa, Y., and Minami, R. (1984). Two cases of congenital hypomyelinating neuropathy. Brain Dev. 6, 560–565.

SUMMARY

Tachi, N., Kozuka, N., Ohya, K., Chiba, S., and Sasaki, K. (1997). Tomaculous neuropathy in Charcot-MarieTooth disease with myelin protein zero gene mutation. J. Neurol. Sci. 153, 106–109. Takashima, H., Boerkoel, D. F., De Jonghe, P., Ceuterick, C., Martin, J.-J., Voit, T., Schroder, J.-M., Williams, A., Brophy, P. J., Timmerman, V., and Lupski, J. R. (2002). Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51, 709–715. Takashima, H., Nakagawa, M., Nakahara, K., Suehara, M., Matsuzaki, T., Higuchi, I., Higa, H., Arimura, K., Iwamasa, T., Izumo, S., and Osame, M. (1997). A new type of hereditary motor and sensory neuropathy linked to chromosome 3. Ann. Neurol. 41, 771–780. Takashima, H., Nakagawa, M., Suehara, M., Saito, M., Saito, A., Kanzato, N., Matsuzaki, T., Hirata, K., Terwilliger, J. D., and Osame, M. (1999). Gene for hereditary motor and sensory neuropathy (Proximal dominant form) mapped to 3q13.1. Neuromuscular Disord. 9, 368–371. Thomas, F. P., Lebo, R. V., Rosoklija, G., Ding, X. S., Lovelace, R. E., Latov, N., and Hays, A. P. (1994). Tomaculous neuropathy in chromosome 1 Charcot-Marie-Tooth syndrome. Acta Neuropathol. 87, 91–97. Thomas, P. K., Marques, W., Davis, M. B., Sweeney, M. G., King, R. H. M., Bradley, J. L., Muddle, J. R., Tyson, J., Malcolm, S., and Harding, A. E. (1997). The phenotypic manifestations of chromosome 17p11.2 duplication. Brain 120, 465–478. Timmerman, V., De Jonghe, P., Spoelders, P., Simokovic, S., Lofgren, A., Nelis, E., Vance, J., Martin, J.-J., and Van Broeckhoven, C. (1996). Linkage and mutation analysis of Charcot-Marie-Tooth neuropathy type 2 families with chromosomes 1p35-p36 and Xq13. Neurology 46, 1311–1318. Timmerman, V., Nelis, E., Vanhul, W., Nieuwenhuijsen, B. W., Chen, K. L., Wang, S., Othman, K. B., Cullen, B., Leach, R. J., Hanemann, C. O., De Jonghe, P., Raeymaekers, P., van Ommen, G.-J. B., Martin, J.-J., Mu¨ller, H. W., Vance, J. M., Fischbeck, K. H., and Van Broeckhoven, C. (1992). The peripheral myelin protein gene PMP-22 is contained within the Charcot-Marie-Tooth disease Type-1A duplication. Nat. Genet. 1, 171–175. Tobler, A. R., Liu, N., Mueller, L., and Shooter, E. M. (2002). DiVerential aggregation of the Trembler and TremblerJ mutants of peripheral myelin protein 22. Proc. Natl. Acad Sci. USA 99, 483–488. Tobler, A. R., Notterpek, L., Naef, R., Taylor, V., Suter, U., and Shooter, E. M. (1999). Transport of Trembler-J mutant peripheral myelin protein 22 is blocked in the intermediate compartment and aVects the transport of the wild-type protein by direct interaction. J. Neurosci. 19, 2027–2036. Topilko, P., Schneider-Maunoury, S., Levi, G., Baron-Van Evercooren, A., Ben Younes ChennouW, A., Seitanidou, T., Babinet, C., and Charnay, P. (1994). Krox-20 controls myelination in the peripheral nervous system. Nature 371, 796–799. Touraine, R. L., Attie-Bitach, T., Manceau, E., Korsch, E., Sarda, P., Pingault, V., EnchaRazavi, F., Pelet, A., Auge, J., Nivelon-Chevallier, A., Holschneider, A. M., Munnes, M., DoerXer, W., Goossens, M., Munnich, A., Vekemans, M., and Lyonnet, S. (2000). Neurological phenotype in Waardenburg syndrome type 4 correlates with novel SOX10 truncating mutations and expression in developing brain. Amer. J. Hum. Genet. 66, 1496–1503. Trachtenberg, J. T., and Thompson, W. J. (1996). Schwann cell apoptosis and its regulation at the developing neuromuscular junction. Nature 379, 174–177. Tsai, B., Ye, Y., and Rapoport, T. A. (2002). Retro-translocation of proteins from the endoplasmic reticulum into the cytosol. Nat. Rev. Mol. Cell. Biol. 3, 246–255. Tyson, J., Malcolm, S., Thomas, P. K., and Harding, A. E. (1996). Deletions of chromosome 17p11.2 in multifocal neuropathies. Ann. Neurol. 39, 180–186. Valentijn, L. J., Baas, F., Wolterman, R. A., Hoogendijk, J. E., van den Bosch, N. H. A., Zorn, I., GabreelsFesten, A. W. M., de Visser, M., and Bolhuis, P. A. (1992a). Identical point mutations of PMP-22 in TremblerJ mouse and Charcot-Marie-Tooth disease Type-1A. Nat. Genet. 2, 288–291. Valentijn, L. J., Bolhuis, P. A., Zorn, I., Hoogendijk, J. E., Vandenbosch, N., Hensels, G. W., Stanton, V. P., Housman, D. E., Fischbeck, K. H., Ross, D. A., Nicholson, G. A., Meershoek, E. J., Dauwerse, H. G., Vanommen, G. J. B., and Baas, F. (1992b). The peripheral myelin gene PMP-22/GAS-3 is duplicated in Charcot-Marie-Tooth disease type-1A. Nat. Genet. 1, 166–170. Vallat, J. M., Sindou, P., Preux, P. M., Tabaraud, F., Milor, A. M., Couratier, P., Le Guern, E., and Brice, A. (1996). Ultrastructural PMP22 expression in inherited demyelinating neuropathies. Ann. Neurol. 39, 813–817. van den Brink, D. M., Brites, P., Haasjes, J., Wierzbicki, A. S., Mitchell, J., Lambert-Hamill, M., de Belleroche, J., Jansen, G. A., Waterham, H. R., and Wanders, R. J. A. (2003). IdentiWcation of PEX7 as the second gene involved in Refsum disease. Amer. J. Hum. Genet. 72, 471–477. VanSlyke, J. K., Descheˆnes, S. M., and Musil, L. S. (2000). Intracellular transport, assembly, and degradation of wild-type and disease-linked mutant gap junction proteins. Mol. Biol. Cell. 11, 1933–1946. VanSlyke, J. K., and Musil, L. S. (2002). Dislocation and degradation from the ER are regulated by cytosolic stress. J. Cell Biol. 157, 381–394. Vedanarayanan, V. V., Smith, S., Subramony, S. H., Bock, G. O., and Evans, O. B. (1998). Lethal neonatal autosomal recessive axonal sensorimotor polyneuropathy. Muscle Nerve 21, 1473–1477. Verhoeven, K., De Jonghe, P., Coen, K., Verpoorten, N., Auer-Grumbach, M., Kwon, J. M., FitzPatrick, D., Schmedding, E., De Vriendt, E., Jacobs, A., Van Gerwen, V., Wagner, K., Hartung, H.-P., and Timmerman, V. (2003). Mutations in the small GTP-ase late endosomal protein RAB7 cause Charcot-Marie-Tooth disease type 2B neuropathy. Amer. J. Hum. Genet. 72,. Verhoeven, K., Villanova, M., Rossi, A., Malandrini, A., DeJonghe, P., and Timmerman, V. (2001). Localization of the gene for the intermediate form of Charcot-Marie-Tooth to chromosome 10q24.1-q25.1. Amer J Hum Genet 69, 889–894.

949

950

39. INHERITED NEUROPATHIES

Villanova, M., Timmerman, V., De Jonghe, P., Malandrini, A., Rizzuto, N., Van Broeckhoven, C., Guazzi, G. C., and Rossi, A. (1998). Charcot-Marie-Tooth disease: An intermediate form. Neuromuscular Disord. 8, 392–393. Vital, A., Ferrer, X., Lagueny, A., Vandenberghe, A., Latour, P., Goizet, C., Canron, M. H., Louiset, P., Petry, K. G., and Vital, C. (2001). Histopathological features of X-linked Charcot-Marie-Tooth disease in 8 patients from 6 families with diVerent connexin32 mutations. J. Peripher. Nerv. Syst. 6, 79–84. von Figura, K., Gieselmann, V., and Jaeken, J. (2001). Metachromatic leukodystrophy. In ‘‘The Metabolic & Molecular Basis of Inherited Disease,’’ vol.3 (C. R. Scriver, A. L. Beaudet, W. S. Sly, D. Valle, B. Childs, and K. W. Kinzler, eds.), 3695–3724. McGraw-Hill, New York. Warner, L. E., Hilz, M. J., Appel, S. H., Killian, J. M., Watters, G. V., Wheeler, C., Witt, D., Bodell, A., Nelis, E., Van Broeckhoven, C., and Lupski, J. R. (1996). Clinical phenotypes of diVerent MPZ (P0) mutations may include Charcot-Marie-Tooth type 1B, Dejerine-Sottas, and congenital hypomyelination. Neuron 17, 451–460. Warner, L. E., Mancias, P., Butler, I. J., McDonald, C. M., Keppen, L., Koob, K. G., and Lupski, J. R. (1998). Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18, 382–384. Warner, L. E., Svaren, J., Milbrandt, J., and Lupski, J. R. (1999). Functional consequences of mutations in the early growth response 2 gene (EGR2) correlate with severity of human myelinopathies. Hum. Mol. Genet. 8, 1245–1251. Weinberg, H. J., and Spencer, P. S. (1978). The fate of Schwann cells isolated from axonal contact. J. Neurocytol. 7, 555–569. Welcher, A. A., Suter, U., DeLeon, M., Snipes, G. J., and Shooter, E. M. (1991). A myelin protein is encoded by the homologue of a growth arrest-speciWc gene. Proc. Natl. Acad. Sci. USA 88, 7195–7199. Wenger, D. A., Suzuki, K., Suzuki, Y., and Suzuki, K. (2001). Galactosyl ceramide lipidosis: Globoid cell leukodystrophy (Krabbe disease). In ‘‘The Metabolic & Molecular Basis of Inherited Disease’’ vol.3 (C. R. Scriver, A. L. Beaudet, W. S. Sly, D. Valle, B. Childs, and K. W. Kinzler, eds.), 3669–3694. McGraw-Hill, New York. White, T. W., and Paul, D. L. (1999). Genetic diseases and gene knockouts reveal diverse connexin functions. Annu. Rev. Physiol. 61, 283–310. Wilmshurst, J. M., Bye, A., Rittey, C., Adams, C., Hahn, A. F., Ramsay, D., Pamphlett, R., Pollard, J. D., and Ouvrier, R. (2001). Severe infantile axonal neuropathy with respiratory failure. Muscle Nerve 24, 760–768. Windebank, T. (1993). Inherited recurrent focal neuropathies. In ‘‘Peripheral Neuropathy’’ vol.2 (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low, and J. F. Poduslo, eds.), 1137–1148. W. B. Saunders, Philadelphia. Wise, C. A., Garcia, C. A., Davis, S. N., Zhang, H. J., Liu, P. T., Patel, P. I., and Lupski, J. R. (1993). Molecular analyses of unrelated Charcot-Marie-Tooth (CMT) disease patients suggest a high frequency of the CMTIA duplication. Amer. J. Hum. Genet. 53, 853–863. Wishart, M. J., and Dixon, J. E. (2002). PTEN and myotubularin phosphatases: From 3-phosphoinositide dephosphorylation to disease. Phosphatase and tensin homolog deleted on chromosome ten. Trends Cell Biol. 12, 579–585. Wong, M. H., and Filbin, M. T. (1996). Dominant-negative eVect on adhesion by myelin P0 protein truncated in its cytoplasmic domain. J. Cell Biol. 134, 1531–1541. Wrabetz, L., D’Antonio, M., Dati, G., Fratta, P., Previtali, S., Imperiale, D., Zielasek, J., Toyka, K., Messing, A., Feltri, M.-L., and Quattrini, A. (2002). Transgenic mice expressing the CMT1B mutant Mpz(delser63) develop demyelinating neuropathy. 2002 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience (Online), Program No. 523.527. Wrabetz, L., and Feltri, M. L. (2001). Do Schwann cells stop, Dr(o)P2, and roll? Neuron 30, 642–644. Wrabetz, L., Feltri, M. L., Hanemann, C. O., and Mu¨ller, H. W. (2001). The molecular genetics of hereditary demyelinating neuropathies. In ‘‘Glial Cell Development’’ (K. R. Jessen and W. D. Richardson, eds.), 331–354. Oxford University Press, Oxford. Wrabetz, L., Feltri, M. L., Quattrini, A., Imperiale, D., Previtali, S., DAntonio, M., Martini, R., Yin, X. H., Trapp, B. D., Zhou, L., Chiu, S. Y., and Messing, A. (2000). P0 glycoprotein overexpression causes congenital hypomyelination of peripheral nerves. J Cell Biol 148, 1021–1033. Xu, W. B., Shy, M., Kamholz, J., Elferink, L., Xu, G., Lilien, J., and Balsamo, J. (2001). Mutations in the cytoplasmic domain of P0 reveal a role for PKC-mediated phosphorylation in adhesion and myelination. J Cell Biol 155, 439–445. Xu, Y., Sladky, J. T., and Brown, M. J. (1989). Dose-dependent expression of neuronopathy after experimental pyridoxine intoxication. Neurology 39, 1077–1083. Yin, X., Kidd, G. J., Wrabetz, L., Feltri, M. L., Messing, A., and Trapp, B. D. (2000). Schwann cell myelination requires timely and precise targeting of P0 protein. J. Cell Biol. 148, 1009–1020. Yin, X. H., Crawford, T. O., GriYn, J. W., Tu, P. H., Lee, V. M. Y., Li, C. M., Roder, J., and Trapp, B. D. (1998). Myelin-associated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J. Neurosci. 18, 1953–1962. Yoshihara, T., Yamamoto, M., Hattori, N., Misu, K., Mori, K., Koike, H., and Sobue, G. (2002). IdentiWcation of novel sequence variants in the neuroWlament-light gene in a Japanese population: Analysis of CharcotMarie-Tooth disease patients and normal individuals. J. Peripher. Nerv. Syst. 7, 221–224. Yoshikawa, H., and Dyck, P. J. (1991). Uncompacted inner myelin lamellae in inherited tendency to pressure palsy. J. Neuropathol. Exp. Neurol. 50, 649–657.

SUMMARY

Yoshikawa, H., Nishimura, T., Nakatsuji, Y., Fujimura, H., Himoro, M., Hayasaka, K., Kakoda, S., and Yanagihara, T. (1994). Elevated expression of messenger RNA for peripheral myelin protein 22 in biopsied peripheral nerves of patients with Charcot-Marie-Tooth disease type 1A. Ann. Neurol. 35, 445–450. Young, P., Boussadia, O., Berger, P., Leone, D. P., Charnay, P., Kemler, R., and Suter, U. (2002). E-cadherin is required for the correct formation of autotypic adherens junctions of the outer mesaxon but not for the integrity of myelinated Wbers of peripheral nerves. Mol. Cell. Neurosci. 21, 341–351. Young, P., Grote, K., Kuhlenbaumer, G., Debus, O., Kurlemann, H., Halfter, H., Funke, H., Ringelstein, E. B., and Stogbauer, E. (2001). Mutation analysis in Charcot-Marie Tooth disease type 1: Point mutations in the MPZ gene and the GJB1 gene cause comparable phenotypic heterogeneity. J. Neurol. 248, 410–415. Young, P., Stogbauer, F., Eller, B., deJonghe, P., Lofgren, A., Timmerman, V., Rautenstrauss, B., Oexle, K., Grehl, H., Kuhlenbaumer, G., VanBroeckhoven, C., Ringelstein, E. B., and Funke, H. (2000). PMP22 Thr118Met is not a clinically relevant CMT1 marker. J. Neurol. 247, 696–700. Young, P., Wiebusch, H., Stogbauer, F., Ringelstein, B., Assmann, G., and Funke, H. (1997). A novel frameshift mutation in PMP22 accounts for hereditary neuropathy with liability to pressure palsies. Neurology 48, 450–452. Yum, S. W., Kleopa, K. A., Shumas, S., and Scherer, S. S. (2002). Diverse traYcking abnormalities for connexin32 mutants causing CMTX. Neurobiol. Dis. 11, 43–52. Zhang, K., and Filbin, M. T. (1998). Myelin P0 protein mutated at Cys21 has a dominant-negative eVect on adhesion of wild type P0. J. Neurosci. Res. 53, 1–6. Zhao, C., Takita, J., Tanaka, Y., Setou, M., Nakagawa, T., Takeda, S., Yang, H. W., Terada, S., Nakata, T., Takei, Y., Saito, M., Tsuji, S., Hayashi, Y., and Hirokawa, N. (2001). Charcot-Marie-Tooth disease type 2A caused by mutation in a microtubule motor KIF1Bb. Cell 105, 587–597. Zhu, Q., Couillard-Despres, S., and Julien, J. P. (1997). Delayed maturation of regenerating myelinated axons in mice lacking neuroWlaments. Exp. Neurol. 148, 299–316. Zielasek, J., Martini, R., and Toyka, K. V. (1996). Functional abnormalities in P0-deWcient mice resemble human hereditary neuropathies linked to P0 gene mutations. Muscle Nerve 19, 946–952. Zoidl, G., Blass-Kampmann, S., D’Urso, D., Schmalenback, C., and Mu¨ller, H. W. (1995). Retroviral-mediated gene transfer of the peripheral myelin protein PMP22 in Schwann cells: Modulation of cell growth. EMBO J. 14, 1122–1128. Zoidl, G., D’Urso, D., Blass-Kampmann, S., Schmalenbach, C., Kuhn, R., and Muller, H. W. (1997). InXuence of elevated expression of rat wild-type PMP22 and its mutant PMP22Trembler on cell growth of NIH3T3 Wbroblasts. Cell Tiss. Res. 287, 459–470. Zorick, T. S., Syroid, D. E., Arroyo, E., Scherer, S. S., and Lemke, G. (1996). The transcription factors SCIP and Krox-20 mark distinct stages and cell fates in Schwann cell diVerentiation. Mol. Cell. Neurosci. 8, 129–145. Zorick, T. S., Syroid, D. E., Brown, A., Gridley, T., and Lemke, G. (1999). Krox-20 controls SCIP expression, cell cycle exit and susceptibility to apoptosis in developing myelinating Schwann cells. Development 126, 1397–1406.

951

C H A P T E R

40 Infectious Demyelinating Diseases Richard T. Johnson and Eugene O. Major

INTRODUCTION Viral infections can lead to acute or chronic demyelinating diseases of the central or peripheral nervous systems in animals and humans. These diseases can be acute or chronic with progressive or relapsing-remitting courses. The loss of myelin with relative sparing of axons may or may not be associated with inXammation or gliosis, and oligodendrocytes or Schwann cells may or may not be altered. The mechanisms of demyelination are varied. Natural infections of rodents (mouse hepatitis and Theiler’s viruses), canines (canine distemper virus), and ruminants (visna of sheep and caprine arthritis-encephalitis viruses) have been associated with central nervous system (CNS) demyelination; see Chapter 44. This chapter focuses on three human CNS demyelinating diseases: acute disseminated encephalomyelitis (ADEM), also known as post-infectious encephalomyelitis, progressive multifocal leukoencephalopathy (PML), and multiple sclerosis (MS). These three diseases have very diVerent clinical courses and distinctive pathological features, although all share the essential element of demyelination. ADEM and PML have antithetic modes of pathogenesis; the former is a predominantly extraneural infection resulting in a virus-induced host autoimmune response, and the latter is a direct lytic infection of oligodendrocytes in an immunocompromised host. The role of infections in MS is unclear; epidemiological studies implicate an early life exposure in the genesis of the disease that could represent an infection. Exacerbations of disease more often follow virallike illnesses and patients with MS have abnormal immune responses to viruses, including the intrathecal generation of antibodies to measles and a variety of other agents.

PROPOSED MECHANISMS OF VIRUS-INDUCED DEMYELINATION Several mechanisms have been proposed to explain the demyelination seen with viral infections, and in MS when an infectious etiology has been postulated (Tab. 40.1). The most straightforward mechanism to explain CNS myelin loss is the selective destruction of the myelin maintaining cells, oligodendrocytes. This mechanism is responsible for the loss of myelin in PML, a disease characterized by infection and lysis of oligodendrocytes (discussed later). In the immunocompetent host, immune responses are often proposed to explain myelin destruction. Most agents associated with CNS demyelination are enveloped viruses. Viral envelopes are formed by insertion of virus-coded proteins into the lipid bilayer of the cell membrane. The core or nucleocapsid of the virus then acquires the envelope by budding through the modiWed membrane. Hypothetically, immune responses against viral proteins within the membrane could caused myelin membrane damage in situ. Alternatively, virus

Myelin Biology and Disorders, Volume 2

953

Copyright 2004, Elsevier (USA). All rights reserved.

954

40. INFECTIOUS DEMYELINATING DISEASES

TABLE 40.1 Proposed Mechanisms of Virus-Induced CNS Demyelination CNS infection Infection of oligodendrocytes Direct destruction Pathogenic immune response to viral antigens on cell membranes Introduction of cell membranes into systematic circulation Infection of other CNS cells Release of cytokines or viral proteins toxic to myelin supporting cells or myelin membranes Extraneural infection Molecular mimicry (virus proteins and myelin proteins) Disruption of immune responses

replicating in oligodendrocytes could transport sequestered myelin antigens into the systemic circulation. Infection of macrophages, microglia or astrocytes may release soluble cytokines, chemokines or viral proteins, which are toxic to other uninfected CNS cells. Visna virus infections of sheep have a long incubation period followed by either a progressive or remitting and relapsing course, accompanied by patchy demyelinated lesions simulating MS. However, infection is limited to cells of macrophage origin. In the brain, macrophages and microglia appear to release a cytokine or similar soluble substance that results in demyelination (Kennedy et al., 1985). A related lentivirus, human immunodeWciency virus (HIV), has subsequently been shown to infect the same restricted cell population in the human brain. Viral proteins as well as cytokines released from infected macrophages and microglia have also been implicated in the pathogenesis of HIV encephalopathy (Power and Johnson, 2001). Demyelination can also result from immune responses against myelin in the absence of nervous system infection. Molecular mimicry, in which an immune response to an environmental agent cross-reacts with a host antigen, has been a popular postulated mechanism. A similar amino acid sequence contained in both viral protein and myelin proteins might allow systemic virus replication to induce an immune response against an epitope on the CNS myelin. Searching sequence databases for commonality of encephalitogenic sequences of myelin basic protein and viral proteins turned up a sequence within the P protein of hepatitis B virus. Inoculation of this synthesized viral sequence into rabbits resulted in an inXammatory response in the brain (Fujinami and Oldstone, 1985). In natural infections, Camphylobactor infections followed by the axonal form of GuillianBarre syndrome are probably due to similarities between bacterial and axonal proteins (Moran and Prendergast, 2001), and cellular damage in HTLV-1-associated myelopathy (tropical spastic paraparesis) appears related, in part, to homology between nuclear ribonuclear neuronal protein and the tax protein of HTLV-1 (Levin et al., 2002). Finally, infection of lymphoid cells may disrupt normal cellular immune responses. Activated T cells normally traYc through the CNS, but only those recognizing an antigen remain (Irani and GriYn, 1996). Thus, in systemic infections causing lymphocyte activation, CNS traYc of T cells is increased. If normally suppressed responses to self-antigens are disrupted by the infection, autoimmune disease can occur such as ADEM associated with measles virus infections (discussed later). For over half a century, the focus on experimental autoimmune (allergic) encephalomyelitis (EAE) as the prototype autoimmune disease biased thought on virus-induced demyelination. In recent years, studies of humoral immune responses in the pathogenesis of Guillian-Barre syndrome and focus on toxic eVects of cytokines and viral proteins in studies of neurological complications of HIV infections have provided a more balanced perspective.

955

ACUTE DISSEMINATED ENCEPHALOMYELITIS

ACUTE DISSEMINATED ENCEPHALOMYELITIS Definition ADEM is an acute, inXammatory, demyelinating disease of the brain and spinal cord. In most patients it has an abrupt onset days to several weeks after a viral exanthem or viruslike illness. But the disease is not speciWc to viruses and has been reported after several bacterial illnesses, immunizations, and drug and serum administration. The nosology is confusing, since the disease has been described under a remarkable variety of names. Post-infectious, parainfectious, post-exanthematous, post-vaccinal, postmeasles, and post-inXuenzal encephalomyelitis have been applied to describe the clinical settings. Acute disseminated encephalomyelitis, perivascular myelinoclasis, perivenous encephalitis, and acute demyelinating encephalomyelitis have been coined to describe the pathological features. Allergic encephalomyelitis, immune-mediated encephalomyelitis, hyperergic encephalomyelitis, and disseminated vasculomyelinopathy imply knowledge of pathogenetic mechanisms (Johnson et al., 1985). Since the essential features for diagnosis are the neuropathological changes, ADEM will be used here except in speciWc cases such as post-measles and post-vaccinal encephalomyelitis. Acute hemorrhagic necrotizing leukoencephalits is generally regarded as a more intense, ‘‘hyperacute’’ form of ADEM. However, this rare acute demyelinating disease has distinct clinical and pathological features and is associated with a diVerent spectrum of antecedent infections.

Epidemiology ADEM was a common disease in the mid-20th century, representing about one-third of all cases of encephalitis. The most common cause of ADEM was measles, which, along with ADEM cases following rubella and mumps, has been largely eliminated in regions of the world that have adequately protected children with the measles-mumps-rubella (MMR) vaccine. The second major cause of ADEM was a vaccine, vaccinia virus, which was discontinued after the worldwide eradication of natural smallpox in 1977. More recently, the varicella vaccine has further reduced the risk of ADEM. Now in countries with active childhood vaccination programs, ADEM makes up less than 10% of the cases of encephalitis, and the most common antecedent illnesses are nonspeciWc respiratory infections (Johnson, 1998). The incidence of ADEM after clinically distinctive virus illnesses is highly variable (Tab. 40.2); the incidence after Epstein-Barr (EB) virus, Mycoplasma pneumoniae, inXuenza, and nonspeciWc upper respiratory infections are uncertain. The clinical Wndings after speciWc infections show some distinctive features, and the mortality and morbidity rates are quite diVerent. Despite the common pathology, the pathogenesis may vary. Most data on pathogenesis relate to post-measles encephalomyelitis.

Pathology In acute fatal cases the brain may be congested and swollen. On gross sections, vessels are prominent in white matter with discoloration along the veins. TABLE 40.2 Postinfectious Encephalomyelitis with Perivenular Demyelination Associated with Exanthematous Viral Infections Case rate

Fatality rate

Vaccinia

1:63 to 1:250,000

10%

Rare

Measles

1:1000

25%

Frequent

a

Varicella

1:10,000

5%

Rubellaa

1:20,000

20%

a

Sequelae rate

10% Very rare

Estimates diYcult to determine because of frequency of toxic encephalopathy or Reye’s syndrome (diVerent pathology) and acute cerebellar ataxia (unknown pathology) and the rare documentation of perivenular demyelinating disease.

956

40. INFECTIOUS DEMYELINATING DISEASES

On microscopic examination mononuclear cells are prominent along the small veins. In intense acute cases, polymorphonuclear cells may also be present. Pallor and loss of myelin staining is seen along the vessels; its perivenular localization often produces Xame shaped lesions. In the spinal cord, this causes a characteristic radial pattern (Fig. 40.1), a pattern distinctive from the plaques of demyelination seen in PML or MS. Patients dying later in disease show even more sharply demarcated lesions and lipid-laden macrophages (Johnson et al., 1985). Immunocytochemical staining of myelin proteins also distinguishes ADEM from PML and MS. In ADEM, as in EAE, areas demonstrating loss of myelin basic protein and myelinassociated glycoprotein are concordant (Gendelman et al., 1984). In PML, the area of myelin-associated glycoprotein loss is distinctly larger than the area of myelin basic protein loss in demyelinated lesions. Presumably this reXects a direct attack on the myelin membranes in ADEM and EAE; whereas in PML, where disease results from oligodendrocyte infection, the myelin-associated glycoprotein, concentrated in the periaxonal, most distal extensions of the myelin membrane, is lost Wrst. Thus, in PML lesions, areas of decreased myelin-associated glycoprotein staining are two to three times larger than areas of myelin basic protein loss. Demyelinated plaques in MS show a mixture of patterns, suggesting diVerent pathogenic mechanisms accounting for myelin loss (Gendelman et al., 1985). In acute hemorrhagic necrotizing leukoencephalitis, the brain is usually strikingly swollen with evidence of herniation. Gross hemorrhages are evident. Microscopically both veins and arterioles show Wbrinoid necrosis. This vascular necrosis is associated with transudates of Wbrin into the tissue, extravasation of red blood cells, and tissue

FIGURE 40.1 ADEM following varicella. This 12-year-old girl developed paraparesis abruptly 2 weeks after the onset of uncomplicated chickenpox. Over the next 3 days the disease evolved with arm weakness, blindness, respiratory distress, seizures, and death. Section of spinal cord shows the characteristic pattern of perivenular demyelination.

957

ACUTE DISSEMINATED ENCEPHALOMYELITIS

necrosis. The inXammatory inWltrate is predominantly polymorphonuclear. These Wndings are most intense in the white matter; and despite the large necrotic areas, there are regions where myelin loss with relative sparing of axons is evident.

Pathogenesis The pathological changes in ADEM resemble those seen in the ‘‘neuroparalytic accidents’’ reported after post-exposure vaccination for rabies using killed virus prepared in animal brain and spinal cord. Indeed, the similarity between the demyelinating encephalomyelitis after vaccination with vaccinia virus to prevent smallpox and the complications of rabies vaccine led Rivers and Schwentker (1935) to studies in monkeys. Animals repeatedly inoculated with emulsions of normal brain developed neurological signs and had perivenular demyelination. This was the discovery of EAE. Subsequently Kabat and colleagues (1947) found that in some species, a single injection of brain could induce the disease if brain inoculum was emulsiWed in adjuvant. Others showed that speciWc sequences of myelin basic protein and proteolipid protein could cause disease, and that disease could be passively transferred with sensitized lymphocytes (Paterson, 1960) (see Chapter 43). EAE mimics ADEM and post-rabies vaccine encephalitis (Tab. 40.3). Lymphocytes from patients with post-rabies vaccine encephalomyelitis (Hemachudha et al., 1988), postmeasles encephalitis, post-varicella cerebellar ataxia, and encephalomyelitis following respiratory infections have been shown to proliferate when cultured in the presence of myelin basic protein. The gap in these parallels is that patients with ADEM have not been injected with myelin proteins. Studies of the pathogenesis of ADEM have focused primarily on measles because the clinical diagnosis is easy, the incidence of ADEM is high compared to other infections, and the neurological complications of measles are homogeneous. Generalization of these studies to ADEM and other post-infectious complications of other viruses is cautioned, since the systemic pathogenesis and eVects on the immune system by the other agents are variable. TABLE 40.3

Comparisons of Experimental Allergic Encephalomyelitis with Encephalomyelitis after Rabies Vaccine and Viral Infections Experimental allergic encephalomyelitis

Post-rabies vaccine encephalomyelitis

Postinfectious encephalomyelitis

Inducing event

Inoculation with CNS tissue or myelin basic Protein

Inoculation with CNS tissue

Infection with enveloped viruses

Latent period

10–21 days

7–42 days

10–40 daysa

Acute onset

þ

þ

þ

Monophasic disease

þ

þ

þ

Occasional chronic or relapsing

þ

þ

þ

þ

þ

þ

þ

þ

þ

In vitro demyelination by lymphocytes

þ

?

þ

Anti-myelin protein antibodies

þ

þ




Clinical forms

forms Pathologic Wndings Perivenular lymphocytes Perivenular demyelination Immunological studies Lymphocytes stimulated in vitro by myelin basic protein

a

From beginning of incubation periods. From Johnson (1998).

958

40. INFECTIOUS DEMYELINATING DISEASES

Measles was the Wrst virus shown to cause immunosuppression. von Pirquet (1908) demonstrated that children had a conversion of positive tuberculin reactions during measles, and for weeks thereafter. Subsequent studies showed inhibition of lymphocyte responses to mitogens for up to 4 weeks after uncomplicated measles virus infection. The magnitude of inhibition of lymphocyte proliferation was the same in children with uncomplicated measles, in those with pneumonitis related to immunodeWciency, and in those with ADEM thought to be due to an autoimmune response (Hirsch et al., 1984). In contrast, spontaneous proliferation of CD4, CD8, and B lymphocytes was found, as well as signs of immune activation, which included lymphoproliferation of lymphocytes in the presence of myelin basic protein in 15% of cases of uncomplicated measles and in 47% of those with ADEM (Johnson et al., 1984). The mechanisms underlying the profound suppression of cell-mediated immunity accompanying measles is still not fully understood. In vitro infection of human monocytes speciWcally down-regulates IL-12 production, a cytokine critical in cell-mediated immunity (Karp et al., 1996), however disruption at other sites in the complex cytokine network is likely. In a subsequent study of ADEM after varied infections, T-cell lines were established from patients. The frequency of cell lines reactive to myelin basic protein was ten times higher in patients with ADEM than patients with encephalitis or controls. IL-4 was the prominent cytokine secreted by T-cell lines from patients with ADEM during the recovery phase (Pohl-Koppe et al., 1998). This Wnding supports a more general relevance of T-cell responses to myelin proteins in the pathogenesis of ADEM.

Clinical Features ADEM is best deWned by its unique pathology, because many of the causative agents are associated with multiple post-infectious syndromes and many of the clinical syndromes and imaging studies overlap with other disease processes. The two most clearly deWned cases of ADEM are those following vaccinia and measles virus infections. Each follows a clinically distinct exanthem and presents a similar clinical course and consistent pathology. In contrast, neurological syndromes associated with rubella, varicella, mumps, and inXuenza may include direct encephalitis, Reye’s syndrome, and acute cerebellar ataxia— all of which may have diVerent mechanisms of pathogenesis not characterized by demyelination (Tab. 40.2). The common clinical features are a lag of 3 days to 3 weeks after the exanthem or respiratory disease, an abrupt onset of headache, fever, and impaired consciousness, and the Wnding of focal neurological signs. The disease reaches a nadir within days, and recovery is variable, depending, in large part, on the causative agent. The spinal Xuid usually shows a modest pleocytosis and mild protein elevation, but can also appear normal. The myelin basic protein content may be elevated, particularly early in disease. In some cases, magnetic resonance imaging has proved an eVective method of diVerentiation from acute encephalitis or Reye’s syndrome. Multifocal white matter lesions of similar age are found, which may or may not enhance. When enhancement is seen in all lesions simultaneously (Fig. 40.2), the image is characteristic of ADEM. Acute ADEM not associated with viral exanthems can be diYcult to diVerentiate from acute viral encephalitis or, in some cases, the initial attack of MS. A viral prodrome, high lesion load on magnetic resonance imaging involving deep gray matter, and the absence of oligoclonal bands in the spinal Xuid favor a diagnosis of ADEM (Hynson et al., 2001). Followup studies of patients diagnosed with ADEM have shown a subsequent diagnosis of MS in some patients, raising the question of whether ADEM might be a part of an ‘‘MS spectrum’’ (Hartung and Grossman, 2001). MS is not, however, an outcome of classical ADEM following measles or vaccinia infections, and ADEM has a very distinctive neuropathology. Measles Measles probably remains the most common cause of ADEM worldwide. Although indigenous measles has been eliminated from the Western Hemisphere and Western

ACUTE DISSEMINATED ENCEPHALOMYELITIS

FIGURE 40.2 ADEM after primary Epstein-Barr virus infection. This college student had classical monospot positive infectious mononucleosis. Two weeks after onset, she developed multifocal neurological signs and coma. The enhanced MRI at that time shows widespread white mater lesions with intense enhancement. She subsequently recovered and returned to school with minimal sequelae.

Europe, it still causes over 1 million childhood deaths each year, largely in developing countries. Measles continues to rank as the third most common infectious cause of childhood death worldwide, following diarrheal illnesses and malaria. The most frequent fatal complications of measles, pneumonitis, gastroenteritis, and secondary bacterial infections, result from a depression of cell-mediated immune responses extending for 1 to 4 weeks after the rash. Prior to the emergence of HIV, measles virus was the most important and lethal causes of a virus-induced immunodeWciency syndrome. However, the major neurological complication of measles infection, ADEM, was assumed to be an allergic or autoimmune disease. Although this initially appeared to be a paradox, studies of measles and HIV have shown that immune activation can suppress immune responses as well as release autoimmune responses. The incidence of post-measles encephalitis is reasonably constant, at about 1 per 1000 cases, and is age dependent, since children under 2 years of age seldom develop encephalitis. It does not appear to be nutritionally dependent, as are some opportunistic infections. Thus, populations such as those of West Africa tend to suVer high rates of measles mortality from opportunistic infections and infant deaths. On the other hand, more aVluent countries without adequate immunization programs have greater mortality and morbidity from post-measles encephalitis, because older children become infected. Although polyneuritis, toxic encephalopathy and acute hemiplegia of possible vascular etiology have been reported with measles, over 95% of the neurological illnesses are represented by ADEM.

959

960

40. INFECTIOUS DEMYELINATING DISEASES

The incubation period of measles is 10 to 14 days. The prodrome and period of infectivity is marked by coryza, conjunctivitis, cough, and the pathognomonic Koplik spots on the buccal mucosa. Coincident with the antibody response and the end of infectivity, a maculopapular rash develops on the face and trunk, later spreading to the extremities. Virus can be recovered from the rash, and for up to 5 days after the appearance of the rash, viral antigen or RNA can be found in epithelioid cells of the lung, gut, bladder, and lymphoid organs (Moench et al., 1988). On rare occasions, viral antigen and RNA can also be detected in cerebrovascular endothelial cells (Esolen et al., 1995). Although there is no evidence of infection of neural cells, approximately 50% of children have an abnormal electroencephalogram during the rash (Gibbs et al., 1959), and approximately 30% have a pleocytosis (Ojala, 1947). Post-measles encephalitis typically develops 4 to 5 days after the onset of the rash, but may precede the rash or be delayed until 3 weeks after. Typically the child is afebrile, the rash is fading and the child is returning to normal activities, when fever returns with headache. Obtundation is frequent and may progress to coma. Generalized or focal seizures occur in about half of the children. Multifocal neurological signs may include cranial nerve abnormalities, pyramidal tract signs, abnormal movements, and ataxia. Sensory deWcits are infrequent. The spinal Xuid may show a modest mononuclear cell pleocytosis but is acellular in about one-third of the patients. Protein elevations are variable, but many have high levels of myelin basic protein in spinal Xuid, particularly early in the course of the encephalitis. Elevated IgG synthesis and oligoclonal bands are usually absent (Johnson et al., 1984). Mortality and morbidity are high. Between 10 and 40% mortality are reported, and neurological sequellae are found in the majority of survivors. Prognosis has been linked to length of stupor or coma (Tyler, 1957), but remarkable recoveries can be seen after prolonged coma (Johnson et al., 1984). The measles vaccine is an attenuated live virus, which has led to the question of whether or not the vaccine may, on rare occasions, cause ADEM. The vaccine virus can produce fever, but in early tests of the vaccine, no abnormalities were found on electroencephalograms that are common with wild-type virus infections. Post-liscencing studies reported one case of encephalitis per million children following vaccination, a number lower than the observed background of two per million cases of encephalitis or encephalopathy per month, suggesting a coincidental relationship between vaccine and disease. Further analysis, however, showed some clustering of cases during the 6 to 15 days after immunization, suggesting a possible relationship (Landrigan and Witte, 1973). Since no histopathological studies have been reported in post-vaccine illnesses, whether or not a few of these cases represent rare cases of ADEM remains unknown. Vaccinia Until recently, the neurological complications of smallpox and vaccinia were of only historical interest; the last case of natural smallpox was observed in 1977, and within a few years vaccinia inoculation had been abandoned worldwide. Recent threats of bioterrorism have led to consideration of renewed vaccinia virus inoculation, and the risks of ADEM must be reconsidered. In retrospect, ADEM accompanied smallpox but was hidden under the devastating systemic disease (Marsden and Hurst, 1932). In the 1920s, ADEM became appreciated as a complication of immunization with vaccinia virus. The incidence of complications after vaccination is extraordinarily variable; an incidence of 1 in 63 is cited for one Dutch vaccine program (but a variety of types of illness were included) (DeVries, 1960). During World War II, an incidence of post-vaccinal encephalomyelitis in England was estimated 1 per 175,000 (Miller, 1953), a subsequent retrospective analysis in the United States estimated 1 per 200,000; during the sidewalk vaccination of 5 million people in New York in 1947 a similar incidence of about 1 per 100,000 was estimated, and during the more recent mass vaccination during the smallpox outbreak in the United Kingdom in 1962, passive reporting estimated neurological complications in 1 per 20,000 (Spillane and Wells, 1964). The variation in incidence may be due to diVerent ethnic populations, diVerent patches of virus, and certainly variable and often poor data collection. ADEM

ACUTE DISSEMINATED ENCEPHALOMYELITIS

is clearly more common with primary immunization, and some data suggest that incidence increases with age. At the time of maximal cutaneous reaction or shortly thereafter, patients with ADEM develop fever, nuchal rigidity and obtundation, Movement disorders including tremor, ataxia, trismus, and myoclonus are speciWcally mentioned in several reports (Miller and Stanton, 1954; Spillane and Wells, 1964). Fatality rates are as varied as incidence rates, but are generally lower than those for post-measles encephalomyelitis, and sequellae are less frequent. The pathogenesis is assumed to resemble that of measles, but vaccinia virus usually replicates only in the dermis. In some patients a viremia is found, and this is more prolonged in patients with ADEM. Virus has been recovered from brains and spinal Xuids of patients with ADEM (Brooks, 1979). Which hematogenous cells are infected and the eVect on immune responses has not been studied. In considering the resumption of vaccination some have recommended (1) vaccination of those likely to encounter victims (Wrst-responders, family health care providers, and clinic and emergency room personnel), (2) vaccination of all who request vaccine, or (3) mandatory universal vaccination. In any of these scenarios, most of those receiving vaccine would be over age 2 and a majority would be receiving primary vaccinations, two factors that presumably increase the risk of ADEM. If the incidence of ADEM were 1 per 20,000 vaccinees (some would consider that Wgure high; others low) in the United States, we might anticipate 10,000 cases of ADEM with 1000 deaths if universal immunization were the option chosen. Varicella Of the neurological complications accompanying chickenpox, fully 50% are acute cerebellar ataxia, which complicates 1 in 4000 chickenpox cases. Typical ADEM is quite rare but does occur (Fig. 40.1). The most dreaded complication is Reye’s syndrome, in which fatty degeneration of the liver is accompanied by life threatening brain edema, but in this disease both inXammation and demyelination are absent. Post-varicella cerebellar ataxia has a good prognosis and the absence of fatalities leave the histopathology undeWned. In several cases of acute ataxia with chickenpox, lymphoproliferative responses in the presence of myelin basic protein have been reported, suggesting a pathogenesis similar to post-measles encephalomyelitis (Johnson et al., 1984; Lisak et al., 1977). Antibodies reacting with sections of cerebrum and cerebellum were reported in 3 of 8 children with post-varicella cerebellar ataxia, suggesting that humoral immune responses may also be involved (Adams et al., 2000). The typical ADEM resembles post-measles ADEM; disease onset is abrupt at 4 to 15 days after the onset of rash. Fever, headache, obtundation, seizures, and focal neurological signs are common (Gollomp and Fahn, 1987). Neuropathological studies are similar. Rubella Rubella is the fourth exanthem classically associated with ADEM. It is estimated that an incidence of 1 per 20,000 cases develop ADEM. The onset is typical; during the week after the rash, fever, headache, and obtundation develop often with seizures. Focal neurological signs are usually not found, but a pleocytosis is. Fatalities do occur, but most have failed to show the typical hallmarks of ADEM; a minority has shown characteristic perivenular inXammation and demyelination suggesting varied types of post-rubella encephalopathies. In a single child, a lymphoproliferative response to myelin basic protein was documented (Johnson et al., 1985). Human Immunodeficiency Virus HIV has been associated with a remarkable spectrum of neurological diseases. Acute meningitis and acute demyelinating polyneuritis (Guillian-Barre syndrome) have often been seen about the time of initial seroconversion. These are assumed to be autoimmune disorders associated with the initial activation of CD4 cells. A small number of newly

961

962

40. INFECTIOUS DEMYELINATING DISEASES

infected individuals have developed a fatal encephalomyelitis, and pathological studies have shown ADEM (Narciso et al., 2001; Silver et al., 1997). Years later with onset of AIDS diVerent central and peripheral nervous system diseases are prominent associated with intense immunodeWciency. HIV dementia, which develops in 20 to 40% of AIDS patients, can be characterized by ‘‘diVuse myelin pallor,’’ visualized on magnetic resonance imaging as a hypodense lesion of white matter, primarily in the cerebral hemispheres, and in pathological sections by a pallor of myelin staining. Initially thought to represent demyelination, immunocytochemical staining for myelin proteins has failed to show myelin loss. The abnormal signal on imaging and pallor of staining seem to reXect a breakdown of the blood-brain barrier (Power et al., 1993). In addition to the prominent diVuse pallor of myelin, small Xame-shaped areas of demyelination without inXammation are occasionally seen along vessels; these may represent the residua of minor or subclinical demyelinating encephalitis that occurred at the time of seroconversion. Other Agents Several viruses have been associated with ADEM that are also associated with apparent acute encephalitis or meningitis. In neurological complications of mumps virus, EB virus, and inXuenza A and B virus infections, virus has been recovered from brain or spinal Xuid and pathological Wndings have been varied; some suggest direct eVects of virus replication in neural cells and some histopathologically are ADEM (Hart and Earle, 1975). Mumps was the single most common viral cause of viral meningitis, but over 90% of these illnesses are now prevented in countries with adequate MMR vaccine administration. Indeed, mumps may be the most neuroinvasive virus, since examination of spinal Xuid in patients with uncomplicated parotitis showed that fully 50% had a pleocytosis. Fortunately, mumps is not highly neurovirulent, and the common neurological complication is benign meningitis. In patients with mumps meningitis, virus can readily be isolated from spinal Xuid, and viral nucleocapsids can be visualized by electron microscopy within ependymal cells found in spinal Xuid (Herndon et al., 1974). This suggests a similar pathogenesis of CNS invasion in humans as seen in hamsters, where mumps virus selectively infects ependymal cells (Johnson, 1968). Serious CNS complications of mumps virus infections are rare. Even prior to immunization, when mumps virus infection was universal, only four to Wve deaths from mumps encephalitis were reported to the Centers for Disease Control each year. Of those cases, about half showed the histological Wndings of perivenular demyelination (Schwarz et al., 1968). The neurological complications of EB virus infections pose an even more confusing spectrum of disease. Typically these disorders arise 1 to 2 weeks after the onset of clinical infectious mononucleosis. About 1% of patients have neurological complications, but these vary from Guillian-Barre syndrome and acute cerebellar ataxia (usually associated with autoimmune responses) to meningitis, encephalitis, and myelitis typical of direct infections (Gautier-Smith, 1965). The Wnding of viral DNA by PCR is not meaningful, since EB virus is latent in B lymphocytes and a single B cell traversing through cerebral circulation or drifting into spinal Xuid could give a positive signal. As with mumps virus infections, intrathecal antibody synthesis and oligoclonal bands of IgG in the spinal Xuid may be found; this is in contrast to post-measles encephalomyelitis but does not exclude an immunopathological mechanism. Again, the diagnosis of ADEM is pathology-based. In the rare fatal cases of encephalitis, a necrotizing polioencephalitis is found; but in others perivenular demyelination has been reported (Paskavitz et al., 1995). Both inXuenza A and B have been, on rare occasions, related to Reye’s syndrome, acute transverse myelitis, and Guillian-Barre syndrome. The association with pathologically veriWed ADEM is very rare (Hoult and Flewett, 1960). Mycoplasma pneumonia is also associated with a variety of neurological complications including the occasional case of ADEM (Riedel et al., 2001). A diagnosis of ADEM is often entertained clinically, but autopsy studies of fatal cases have generally shown brain edema or perivascular inXammation without convincing demyelination. Recent anecdotal reports of probable ADEM associated with hepatitis C virus (Sacconi et al., 2001), attenuated polio vaccine virus (Ozawa et al., 2000), and acute herpetic gingivostomatitis (Ito et al., 2000) may be coincidental.

PROGRESSIVE MULTIFOCAL LEUKOENCEPHALOPATHY

Acute Hemorrhagic Necrotizing Leukoencephalopathy This acute hemorrhagic disease has been regarded as a more intense form of ADEM and as a distinct entity. The clinical setting is similar. Usually 1 to 20 days following a virus-like illness, the disease develops with fever, obtundation, seizures, and focal neurological signs. Although reported in individual cases after measles and chickenpox, most follow a nonspeciWc upper respiratory infection. In Asia, a number of cases have been associated with inXuenza virus infections (Voudris et al., 2001); recently several cases were reported with Mycoplasma pneumoniae infections (Pfausler et al., 2002). In contrast to ADEM, the illness is fulminant with signs suggesting an expanding mass lesion, and the majority of patients die within 5 days. The spinal Xuid shows polymorphonuclear cells and red cells; in addition, a peripheral leukocytosis and proteinuria are usually found. The arguments linking acute hemorrhagic necrotizing leukoencephalitis to ADEM are (1) the similar clinical setting despite a distinct spectrum of precipitating illnesses, (2) an apparent continuum of pathology with cases clearly showing features of both diseases, and (3) the animal model of hyperacute EAE that resembles hemorrhagic necrotizing leukoencephalitis (Levine and Wenk, 1965). In a single case, a proliferative response of the patient’s lymphocytes was demonstrated when cultured with myelin basic protein (Behan et al., 1968).

Prevention and Treatment Few arenas of medicine have celebrated the extraordinary success in disease prevention at as an astonishing cost-beneWt ratio as in the prevention of infections with vaccines. The measles vaccine alone prevents between 2 million and 3 million deaths each year, reduces the burden of permanently neurologically impaired by even greater numbers, has reduced childhood deafness by 10%, and has virtually eliminated subacute sclerosing panencephalitis from countries with sustained vaccine programs. The cessation of vaccination to prevent smallpox and the introduction of vaccine programs to prevent mumps, rubella, and chickenpox have decreased the incidence of ADEM dramatically. Treatment is less eVective. Although the literature abounds with anecdotal claims of the beneWts of corticosteroids, no randomized placebo-controlled study supports their value. Several studies of sequential patients with measles encephalitis who did or did not receive steroids or ACTH showed no diVerence in mortality or morbidity (Ziegra, 1961). These studies may be applicable only to measles, and empiric treatment with steroids remains common in ADEM, particularly when there is any evidence of increased intracranial pressure. One retrospective study of post-infectious encephalomyelitis showed higher mortality and morbidity rates among those who had received corticosteroids. On the assumption that the more seriously ill would tend to be treated, the data were reanalyzed to include only patients admitted in coma. Even in this reevaluation the mortality was higher in treated patients (Boe et al., 1965). Supportive treatment is important since children can have remarkable recoveries after prolonged coma. Therefore lowering of fever, management of seizures, careful management of Xuids, reducing increased intracranial pressure, and prevention of urinary and respiratory tract and skin infections are paramount. Mechanical ventilation is often necessary.

PROGRESSIVE MULTIFOCAL LEUKOENCEPHALOPATHY Definition PML In the 1950s, a rapidly progressing neurologic syndrome was observed in a patient suVering from chronic lymphatic leukemia (CLL). Although the cause was ruled to be a result of leukemic cells that had inWltrated the brain, post-mortem examination revealed bizarre, enlarged nuclei and demyelination, which was not consistent with the diagnosis. A few

963

964

40. INFECTIOUS DEMYELINATING DISEASES

years later, another patient with CLL developed similar neurologic symptoms. Again, autopsy revealed demyelinated plaques populated by cells with enlarged, dense nuclei, which were hypothesized to be oligodendrocytes with an unusual, undescribed pathology. The results were published in 1958 as the Wrst clinical and neuropathological description of a new disease, progressive multifocal leukoencephalopathy (PML), also known as Richardson’s disease after E. P. Richarson, who was responsible for the initial description (Astrom et al., 1958). There were early suspicions that PML was caused by a virus; however, the etiologic agent was not identiWed until 1971 as the human polyomavirus, JC Virus. Like other polyomaviruses, JCV was named after the initials of the patient from whom virus was Wrst isolated (Padgett et al., 1971) Initially the disease was mainly associated with patients suVering from lymphoproliferative and myeloproliferative diseases such as CLL, Hodgkin’s disease, and sarcoidosis (Richardson, 1974). Following the description of PML, retrospective review of the literature uncovered several similar accounts in patients with dementia suVering from a variety of underlying immunosuppressive disorders. The accounts dated back as far as 1930 with pathologic features consistent with PML (Hallervorden, 1930). PML is almost exclusively associated with an underlying cellular immunodeWciency. Prior to the AIDS epidemic, the association was most frequently observed in patients with Hodgkin’s disease and CLL. In the decades following, the exponential rise in the incidence of AIDS resulted in a much larger and rapidly growing population of immunosuppressed individuals. Presently, PML has become an increasingly common complication in AIDS patients, and has been the AIDS deWning illness in approximately 1% of the cases. It is expected that approximately 4 to 5% of all AIDS patients will eventually develop this disease (Berger and Major, 1999). JC Virus JCV is a small, noneneveloped virus, approximately 45 nm in diameter. The viral capsid is icosahedral in symmetry, and made up of the 3 viral capsid proteins encoded by the late region of the JCV genome. After translation and transcription in the cytoplasm, nuclear localization signals located on the amino terminal region transport the proteins to the nucleus, where the virion particles are assembled (Moreland and Garcea, 1991). Capsid assembly is governed by the major structural protein, Vp1, which accounts for more than 70% of the entire viral protein content. Vp1 also contains the antigenic epitopes to which a speciWc antibody response is mounted, and it is responsible for the ability of JCV to agglutinate human type O erythrocytes (Shah et al., 1977; Wang et al., 1999). Located inside the viral capsid are the JCV minchromosomes, each of which is a single molecule of viral DNA complexed with cellular and nuclear histones. The complete genome is a closed, circular supercoiled structure, approximately 5.13 kb in length (Frisque et al., 1984). Transcription occurs in both directions starting from the highly conserved origin of replication, marked by a single EcoR1 restriction site. Extensive sequencing data has revealed that the genome can be functionally divided into three regions: the early region, encoding two nonstructural proteins; the late region, encoding three capsid proteins; and the regulatory region. The early region, located on the proximal side of the origin, is transcribed and expressed early after viral entry. Counterclockwise transcription of this region starting from the EcoR1 site, followed by diVerential splicing, produces two major mRNA species encoding the Large T and small t antigens. It is known that Large T is a nonstructural, DNA binding protein with multiple functions (Fanning, 1992), one of which is the initiation of JC viral DNA replication by the unwinding and separation of the two strands of DNA so that the polymerases can function. Furthermore, it autoregulates to prevent transcription of the early genes during the later stages of infection, when the structural genes are being produced. Large T also plays a part in the malignant transformation of cells by binding to cell cycle regulatory proteins as well as tumor suppressors such as p53 and pRB, resulting in cellular malignant transformation. The late region, located on the distal side of the origin, is encoded in the strand of DNA complementary to the strand that encodes the early genes. Clockwise transcription from the origin of replication yields the capsid proteins VP1, VP2, and VP3, as well as the Agnoprotein. VP1, VP2, and VP3, as discussed earlier, are the structural proteins that make up the

PROGRESSIVE MULTIFOCAL LEUKOENCEPHALOPATHY

viral capsid. Agnoprotein in related viruses has been implicated in DNA binding and localization of VP1 to the nucleus (Cole, 1996, Carswell and Alwine, 1986). A recent report describing the presence of JCV agnoprotein in the cytoplasm of infected cells, suggests that protein shuttles freely between the cytoplasm and the nucleus and is important for JCV proliferation (Okada et al., 2001). Early gene expression must take place before viral DNA replication can proceed, since the early products, Large T and small t, are necessary for the expression of the late gene products. Thus, eYcient expression of those late structural proteins occurs only after viral transcription and translation have taken place. The noncoding sequences located between the early and late genes contain the origin of replication, the JCV promoter, and enhancer sequences. Collectively, this area is also known as the viral regulatory region (RR), thought to control host range for lytic infection. Studies of JCV infection in various cell types have demonstrated the extremely narrow host range of this virus. It has been shown that in cell culture systems, human glial cells are the most susceptible to infection as well as the most conducive to virion production (Wroblewska et al., 1980). It was originally hypothesized that the Large T viral protein deWned the neurotropism of JCV, since the virus was shown to be able to replicate in nonglial cell types only in the presence of JCV or SV40 Large T protein, both of which share signiWcant homology (Feigenbaum et al., 1987). Subsequent studies interchanged the JCV promoter with other viral promoters and used the constructs to infect various cell types. The results have suggested that the cell type speciWcity of JCV may also be a function of the viral promoter (Feigenbaum et al., 1992). The restricted growth in cell types from extra neural tissue such as kidney and tonsil has led to questions regarding the nature of JCV susceptibility. Is JCV host cell restriction at the level of binding and entry or is it at the molecular level? The recent identiWcation of the speciWc cell surface receptor for JCV has provided some answers to these questions (Liu et al., 1998). The receptor, an a 2-6 linked sialic acid, is a commonly expressed ganglioside found on the surface of various cells. Indeed, binding studies of JCV have shown that the virus can bind to and enter numerous cell types via this receptor, both susceptible and nonsusceptible. Thus, unlike other viruses, viral susceptibility to JCV is not a function of speciWc receptor binding and entry. The focus has now shifted to intracellular factors that may contribute to the speciWcity of this virus. The regulatory region of JCV has several binding sites for the NF1 family of transcription factors. A comparison of relative NF1 expression between highly susceptible cells from the human fetal brain (HFB) and a nonsusceptible epithelial cell line (HeLa), demonstrated that one class in particular, NF1-D, was elevated in the HFB cells (Sumner et al., 1996). Further studies have shown that overexpression of NF1-D in a normally nonpermissive cell line can render the cells susceptible to JCV infection, as determined by early and late gene production (Monaco et al., 2001). The regulatory region of this virus has been of much interest, because it is becoming increasingly clear that despite eVective virus binding or the presence of viral DNA in the cells, productive infection only occurs in the cell types that express the appropriate transcriptional factors. The very narrow host range and the cell type speciWcity of JCV are unique properties of this virus.

Epidemiology It would be expected that the incidence of PML would be higher in the geographic areas coinciding with a large HIV infected population. However, since the technology required to make a deWnitive diagnosis of PML may not readily be available in such areas, the actual incidence of this disease may be masked to some extent. As such, the entire worldwide incidence of PML has not yet been determined. However, serological studies of human polyomaviruses in geographically isolated regions have yielded much information on the natural distribution of JCV in human populations. The Wrst study published in 1973 described a random sampling of 406 individuals from the state of Wisconsin, screened for the presence of JCV antibodies. The results showed that approximately 70% of assayed individuals had signiWcant antibody titers to JCV in their blood. However, the percentage dropped dramatically in very young children, suggesting that initial infection

965

966

40. INFECTIOUS DEMYELINATING DISEASES

and seroconversion most likely occurs during early childhood (Padgett and Walker, 1973). Much larger studies were conducted shortly thereafter to study the worldwide distribution of human polyomaviruses. Again, JCV infection was prevalent worldwide, as demonstrated by the presence of antibodies in infected individuals. The majority of adolescents and a much higher percentage of adults (80%) showed elevated levels of JCV-speciWc antibodies (Brown et al., 1975,Walker and Padgett, 1983). A more recent epidemiological study of the incidence of PML in the United States was conducted from 1979 to 1994. The survey reported the incidence of PML among HIV infected individuals as 1.6%. However, since the data only included PML cases that were diagnosed ante mortem, and not new cases discovered at autopsy, the authors felt that the numbers were an under-representation of the actual prevalence of the disease. It was also reported that 89% of the PML cases were attributed to HIV infection during this time period (Holman et al., 1991). A recently concluded seroepidemiological study examining the circulation of human polyomaviruses in the population determined that 80% of the population exhibited antibodies to JCV (Maher, D. et al., in press).

Pathogenesis The exact route of JCV transmission, initial infection and pathogenesis continues to be investigated. Seroepidemiologial studies have proven that the majority of the healthy, human population has antibodies against the virus, although the percentage drops in very young children. Thus, it is postulated that initial infection and seroconversion occurs within the Wrst 6 years of life with no associated clinical symptoms. The initial route of infection remains elusive, although one study has reported JCV replication in human tonsillar tissue, implicating a site of latency and also a primary infection route via inhalation (Monaco et al., 1998). TraYcking B lymphocytes (Tornatore et al., 1992) may then carry the virus from the tonsillar and stromal tissue to other latent sites, including the bone marrow, tonsils, colon epithelial cells, and kidney (Jensen and Major, 1999; Laghi et al., 1999) (Fig. 40.3). It is well documented that the virus is excreted in the urine of healthy individuals as well as patients with PML (Agostini et al., 1996). Transmission via this route is unlikely, however, because the predominant genotype found in urine isolates has not be shown to be able to cause a robust infection in any type of human cell (Ault, 1997). Systemic circulation of the virus to these diVerent compartments is thought to be via a hematogenous route, most likely involving infected B lymphocytes. Lymphocytes may shuttle the virus across the blood brain barrier during a period of severe immune deWciency, which sets the stage for the onset of PML when JCV infection is passed from the lymphocytes to the highly susceptible glial cell population.

Pathology The initial descriptions of PML were focused on the histopathological features associated with the disease, such as the enlarged oligodendroglial nuclei and bizarre, giant astrocytes. The swollen nuclei of infected oligodendrocytes usually display a change in chromatin pattern. They often appear more homogenous or will have the chromatin concentrated at the periphery, near the nuclear membrane. The nuclei have also been described as having ‘‘ground glass’’ appearance, which is due to the presence of numerous inclusion bodies. Electron microscopy has revealed that these inclusion bodies are actually a dense, crystalline, or Wlamentous array of JC virion particles (Aksamit, 1995). Upon gross examination, demyelinated plaques and lesions can be visible to the naked eye. The foci of demyelination are initially few and randomly distributed. They can range in size from millimeters to centimeters in diameter, occasionally coalescing to form even larger lesions. The lesions have been identiWed throughout the white matter of the brain, including the cerebral hemispheres, cerebellum, medulla, and even spinal cord. However, lesions are most commonly located in the subcortical white matter, near the gray-white matter junction, which is an area of increased cerebral blood Xow. This lends support to the hypothesis of viral entry into the CNS via hematogenous dissemination. However,

967

PROGRESSIVE MULTIFOCAL LEUKOENCEPHALOPATHY

Suggested primary routes of infection

Respiratory inhalation

Gastrointestinal tract

Initial Infection Asymptomatic

Viral dissemination

B

lung

colon

tonsil †

liver Bone marrow*

spleen

Viral latency and activation Periphery

Kidney*

†

B

Blood brain barrier

Brain

astrocytes

oligos

Progressive Multifocal Leukoencephalopathy FIGURE 40.3 Pathogenesis of progressive multifocal leukoencephalopathy. Primary infection is followed by an extended period of latency in several anatomical compartments. Current data implicate the kidneys and bone marrow as potential sites for JCV latency (*). Active JCV replication has been demonstrated in tonsillar tissue as well as in kidney ({). Following viral activation, infected B lymphocytes may traYc the virus across the blood brain barrier and into the parenchyma of the brain, where the infection is then passed to the highly susceptible glial cell population, ultimately resulting in the pathological and clinical symptoms associated with PML.

lesions do not follow the cerebral vasculature of the brain. Microscopic examination of the plaques shows lytic destruction of oligodendrocytes, myelin degradation, but sparing of the associated axons. Active viral replication and capsid formation in infected oligodendrocytes is followed by cytolysis and viral release to surrounding cells. Susceptible cell types will follow the same pattern, resulting in a lesion where infected oligodendrocytes are concentrated at the periphery, encroaching outwards as the lesion grows. Thus, JCV is disseminated by cell-to-cell contact. Oligodendrocytes can be found throughout the lesion, often with enlarged, basophilic nuclei. Astrocytes found in this area may be hypertrophied and severely enlarged, while neurons are typically spared. In some cases, lipid-laden macrophages have been identiWed in the center of the lesions as well, evidence of active myelin degradation. The mechanism behind JCV induced demyelination is primarily through the lytic infection of oligodendrocytes. However, there is some evidence that JCV large T protein

968

40. INFECTIOUS DEMYELINATING DISEASES

may interfere with the production of myelin. InXammatory inWltrates are rare in PML, although there is a pronounced active astrocytosis in the lesions.

Radiographic Findings Besides the onset of clinical symptoms, one of the Wrst indicators of PML is a lesion visualized by CT or MRI. Lesions are often diYcult to attribute to a speciWc disease because, radiologically, they can appear very similar. As a general rule, PML lesions are multifocal and noncontrast enhancing. The location and extent of lesioning can vary, however. As such, radiographic imaging cannot be used alone to diagnose PML. Neuroimaging has remained one of the most important tools in studying the PML patient (Thurnher et al., 1997). The noninvasive visualization of lesions is helpful for a preliminary diagnosis, especially in cases where suspected lesions are in areas where the risk of biopsy is too high. By computerized tomography (CT) scan, the demyelinated lesions appear as asymmetrically distributed subcortical areas of decreased signaling intensity. Contrast enhancement is very rarely seen, indicating an intact blood brain barrier. CT scans are limited in detecting the extent of lesions occurring deeper in the brain, such as in the cerebellum or brainstem. Magnetic resonance imaging (MRI) is extremely sensitive in determining not only the number of lesions, but the extent as well. In fact, MRI has been shown to be more sensitive than CT in both aspects, often revealing lesions where the CT scan previously appeared normal. As such, MRI is the preferred diagnostic technique for the evaluation of potential PML cases. In contrast to the hypodense lesions seen in CT scans, demyelinated lesions appear as patchy areas of increased signal intensity by T2 weighted MRI. T1 weighted images exhibit a decrease in signal intensity, as opposed to the hyperintense T2 weighted images. T1 related decrease in signal intensity is consistent with demyelination. Metabolic assays generally have not been of much use in studying PML because lesions tend to occur either in the myelinated white matter or at the gray white matter junction.

Clinical Features Viral reactivation and lytic JCV infection of oligodendrocytes can cause demyelination in any white matter tract located throughout the brain. The clinical symptoms seen in PML patients are consistent with the extent and location of subcortical white matter destruction, with no inXammatory changes in the CSF. Furthermore, the spectrum of symptoms in HIV-associated PML patients is nearly identical to that of PML associated with other underlying immune deWciencies. The most common presenting symptoms are known collectively as the ‘‘triad,’’ which include a progressive deterioration of visual, motor, and cognitive functions. In HIV associated PML cases, the onset of neurological signs and symptoms may actually precede a diagnosis of AIDS in HIV infected patients and has been added to the list of AIDS deWning illnesses. The most common presenting ailments are motor abnormalities. By the time of diagnosis, the majority of patients will exhibit signs of moderate to severe weakness, which typically aVects limbs on one side of the body (hemiparesis). Gait abnormalities (Berger and Major, 1999) or diYculty performing routine motor tasks are often accompanied by complaints of lethargy or impaired movement of the arms and legs. Visual deWcits also account for a signiWcant percentage of presenting symptoms. The severity of the symptoms often correlate with the extent of lesioning. Hemianopsia, or blindness in one-half the visual Weld in each eye is common. Predictors of longer survival time with PML include lack of clinical progression during the Wrst 2 months of treatment (De Luca et al., 2001), contrast enhancement and mass eVect (Berger, 2000) in HIV patients, concomitant treatment with HAART and importantly, low JCV virus levels in the CSF (Yiannoutsos et al., 1999).

PROGRESSIVE MULTIFOCAL LEUKOENCEPHALOPATHY

Treatment and Immune Deficiency There is no established therapy, as yet, for the eVective treatment of PML. Although there have been isolated reports in individual patients, large-scale studies have been diYcult to conduct, due to the limited number of possible test subjects. Attempts to treat the disease have traditionally been aimed at curing the underlying immune deWcit, thereby alleviating symptoms of opportunistic infections. Therapies aimed speciWcally against JCV have also been investigated, but with mixed results. Such antiviral therapies are diYcult due to the fact that the drugs not only interfere with the virus but also with the normal functioning of the host cell. Cytarabine, or Ara-C, is a nucleoside, well established as a chemotherapeutic agent to treat various malignant disorders, but has also been reported to be beneWcial in the treatment of PML patients. Positive case reports, in addition to preliminary results suggesting that the drug may have some activity against JCV in primary human cells in vitro (Hou and Major, 1998), prompted the undertaking of ACTG 243, the largest clinical trial, as yet, for any opportunistic infection associated with AIDS. HIV infected, biopsy proven PML patients were administered Ara-C, either intravenously or intrathecally, with or without concomitant antiretroviral therapy. Patients were monitored for clinical signs and viral load during treatment. The results from this trial showed that the administration of Ara-C resulted in no statistical diVerence between the outcome of treated patients to untreated controls (Hall et al., 1998). However, this and other trials investigating Ara-C resulted in data showing the prognostic value of JC viral load in the CSF. Patients with a reduction in the levels of virus present in the CSF had a signiWcant increase in survival time (De Luca et al., 1999; Yiannoutsos et al., 1999). Cidofovir is an antiviral agent that is active against a broad spectrum of human DNA viruses, best documented in the treatment of cytomegalovirus (CMV) retinitis, a signiWcant complication in AIDS patients, or genital herpes. There have been several reports of success in treating AIDS-related PML with the drug, particulary in European countries. Although failures have been reported, the addition of cidofovir to preexisting antiretroviral therapy in AIDS patients seemed to have improved the symptoms of PML, even in cases where the course of the disease worsened despite highly active antiretroviral therapy (Portilla et al., 2000). SigniWcant radiological improvements, marked by decreases in the extent of lesions, have also been reported in response to cidofovir therapy (Cardenas et al., 2001). In contrast, the results from the only clinical trial speciWcally using cidofovir in relation to PML, ACTG protocol 363, were far less encouraging. Reported at the 8th Conference on Retroviruses and Opportunistic Infections, the preliminary statistical analysis revealed no diVerence in prognosis between patients receiving cidofovir and those without (Marra et al., 2002). Since the beginning of the AIDS pandemic, the most common underlying immune deWcit resulting in PML has been HIV infection. Combination retroviral therapy also known as highly active anti-retroviral therapy or HAART has been frequently reported to increase the immune status, clinical and radiological symptoms, and survival times in some PML patients (Inui et al., 1999). The presence of JCV in the CSF or the initiation of a JCV-speciWc humoral intrathecal response have been used as markers for viral replication and immune status, both of which may possess prognostic value for the progression of PML. This is particularly true with the advent of HAART therapy, which in multicenter analyses has been successfully correlated with longer survival. However, despite prolonged survival, HAART did not always halt the rapid neurological deterioration of PML, suggesting that early intervention with a drug speciWc for JCV will be necessary to stabilize the destruction of the white matter (Gasnault et al., 1999). Similar case reports have also shown that patients can develop PML while on HAART therapy, and that a decrease in HIV viral load and increase in CD4þ cell counts does not always correlate with resolution of neurological symptoms, again emphasizing the need for a deWnitive antiviral treatment (Tantisiriwat et al., 1999). It also appears that the initiation of HAART therapy in patients with a high JC viral load at the time of diagnosis does not have a signiWcant eVect on the surival of the patients following treatment, whereas

969

970

40. INFECTIOUS DEMYELINATING DISEASES

patients with median or low viral loads at diagnosis do have prolonged survival on HAART (TaouWk et al., 2000). One recently completed study retrospectively analyzed a series of PML patients being treated with HAART in three northern Italian neurological clinics. The trial revealed that although approximately 50 percent of the patients did show disease stabilization and longer survival in response to HAART, the rapid immune reconstitution could speed the progression of PML as a result of the circulation of lymphocytes that are actively infected by JCV (Cinque et al., 2001).

Conclusions PML is the only human demyelinating disease with a known direct viral infection of oligodendrocytes. It remains a challenge to understand how a virus with such a widespread presence in the normal human population can be targeted so speciWcally to oligodendrocytic destruction in the brain of immune compromised individuals. Once considered exceedingly rare, PML has been given much attention recently due to the continual rise in the incidence of AIDS and other immunosuppressive disorders due to chemotherapy in cancer patients and preventative measures against graft rejection in transplant recipients. Since the majority of the population has already been infected with JCV, and a signiWcant percentage of AIDS patients will develop the disease, it is critically important to establish an eVective treatment regimen and develop methods to screen severely immunocompromised patients for the potential risk of developing PML.

MULTIPLE SCLEROSIS Definition Although the plaques of MS on gross brain and spinal cord specimens had been previously described, the clinical complex of multifocal remitting and relapsing disease and pathological correlation with plaques of sclerosis was made by Charcot in 1868. In keeping with his times, he postulated that exposure to cold, physical injury or emotional stress caused the disease. Over the next 15 years, Koch and Pasteur laid the foundations of microbiology and immunology, so it is not surprising that in 1884 Pierre Marie, a student and successor of Charcot as Professor of Neurology in Paris, proposed a microbial cause of MS. Indeed, in the afterglow of Pasteur’s discovery of a post-exposure vaccine to protect against rabies, Marie prematurely predicted that a vaccine would soon be available for MS. Speculation concerning an infectious agent as a cause of MS has recurred over the past century but has gained greater credence over the past 50 years because of three areas of investigation: (1) epidemiologic studies have suggested that MS results, in part, from a childhood environmental exposure followed by a long latency; (2) studies of viral diseases of animals and humans have documented that infections can have long incubation periods, give rise to relapsing and remitting courses, and cause demyelination; (3) studies of patients with MS have consistently shown abnormal immune responses to viral antigens, particularly the intrathecal synthesis of antibodies to viral antigens. MS is deWned as a demyelinating disease of the CNS with lesions separated in space and time and in which other diagnoses have been ruled out. Before the recovery of Borellia burgdorferi or the human T-cell lymphotropic virus, some cases of Lyme disease and tropical spastic paraparesis fell within the deWnition of MS. We may view these observations from two perspectives: (1) in the past we made erroneous diagnoses now corrected by greater knowledge or (2) what we deWne as MS may have multiple causes and manifestations, and with more speciWc diagnoses the syndrome will become better and more narrowly deWned. The epidemiology, pathology, and clinical features of MS are comprehensively covered in other chapters. Discussion in this chapter will focus on features suggesting possible infectious causes.

MULTIPLE SCLEROSIS

Epidemiology Geographic and Ethnic Distribution The prevalence of MS is highly variable, dependent on geography and ethnicity. The prevalence of over 200 per 100,000 in the Shetland and Orkney Islands of the North Atlantic contrasts with prevalences approaching 1 per 100,000 in areas of Africa. This variance is not explained by clinical sophistication or quality of health care as once believed. Even within Europe and North America a north/south gradient or zones exist with higher incidence related to higher latitude. This appears to be reXected in the Southern Hemisphere, where the incidence of MS is higher in southern than northern areas of Australia and higher on the south than the north island of New Zealand (Kurtzke, 1993). In Northern Europe, Canada, and the Northern United States, prevalence rates are high and range from 30 to 80 cases per 100,000. In Southern Europe and the Southern United States, moderate rates of 6 to 29 are usual. Low prevalence rates are deWned as those below 5 per 100,000 and prevail in Northern South America and all known areas of Africa and Asia. Remarkable exceptions to the general correlation of latitude to prevalence are exempliWed by the absence of MS in Eskimos, and a prevalence in excess of 150 per 100,000 in Sardinia (Montomoli et al., 2002). The incidence of MS is clearly tied to geography, but whether this reXects immigration routes of Northern Europeans carrying susceptibility genes, or whether it reXects regional exposure still provokes controversy. MS is more common in women than men in all regions, which supports the postulated immune-mediated pathogenesis. Distribution among racial groups is also unequal; it is more common in whites than Asians, and more common in Asians than African Americans. This supports genetic factors in causation, but it is diYcult to accredit this solely to genetic factors since the prevalence in non-whites in the United States increases with increasing latitude, implicating the importance of environmental agents (Lowis, 1988). MS prevalence also correlates to a lesser extent with higher socioeconomic class and urbanization. Little data are available to determine whether the prevalence of MS has increased over time. The increasing identiWcation of occasional cases in those ethnic group previously thought to be spared from the disease (native-born Andeans residents, black South Africans, Eastern European gypsies, etc.) may represent better medical care and availability of imaging or may represent a genuine spread of MS to previously unaVected populations. In Rochester, Minnesota, a stable rate was noted for many years and a more recent survey showed a striking rise (Wynn et al., 1990). Prevalence data are hard to compare, since the earlier diagnosis and longer survival increase prevalence rates and incidence rates are less accessible Familial Aggregation and Genetics The risk of MS is signiWcantly increased in the siblings and progeny of MS patients. This is striking in twins where the Canadian study has shown monozygotic twins had a concordance rate of 30.8% and dizygotic sex-alike twins had a concordance rate of 4.7% (Sadovnick et al., 1993). This strengthens the long recognized genetic component of causation; however, 70% of the monozygotic twins were discordant suggesting important nongenetic factors. The Canadian study subsequently evaluated adopted, nonbiological relatives and could Wnd little eVect of shared environment (Ebers et al., 1995). Migration Studies Studies of populations migrating from high-risk regions to low-risk regions and studies of patterns of disease on North Atlantic islands have provided the strongest support for the role of a childhood exposure followed by a long incubation period. These investigations began after World War II when Dean (1970) observed that the majority of patients with MS in South Africa were immigrants from the United Kingdom or Northern Europe, even though they made up less that 10% of the population. He determined a prevalence of 50 per 100,000 in this population, compared with 11 per 100,000 in native born English-speaking South Africans, 3 per 100,000 in white Africaans speaking natives, and absence of MS in

971

972

40. INFECTIOUS DEMYELINATING DISEASES

black, native-born South Africans. Further analysis showed that migration prior to age 15 years led to a risk similar to that of native born English South Africans; and migration as an adult resulted in a risk similar to that of the country of origin (Dean and Kurtzke, 1971). Similar risk of early life exposure for migrants have been reported from Israel, Australia, Hawaii, and the Antilles. Recent studies of migrants from the United Kingdom and Ireland to varied latitudes in Australia showed prevalence in migrants was considerably less than their country of origin, but did not show a shift at age 15, suggesting that environmental factors may operate over a period of many years and not only in childhood (Hammond et al., 2000). Studies of migrants from low risk regions to high risk regions suggest a similar phenomenon but are less complete. Similar evidence of early life exposure is provided by studies of World War II veterans. In the United States, residence at birth and military induction showed a sharp north to south diVerential in risk for MS. Residence after induction and at the onset of disease showed no geographic correlation indicating that risk was acquired prior to conscription (Beebe et al., 1967). Apparent Epidemics Continental incidence rates appear rather stable, but studies of island populations of Norse ancestry in the North Atlantic have shown Xuctuations suggesting epidemics. Prevalence rates on the Shetland and Orkney Islands repeatedly showed the highest rates worldwide, but the breakdown of incidences of new cases per year show an abrupt upsurge in the late 1930s, and an equally sudden decline in new cases in 1971, suggesting an epidemic during the intervening 3 decades (Kurtzke, 2000) The data from the Faeroe Islands are even more dramatic. No cases of MS were documented prior to 1943; then 16 patients had onsets between 1943 and 1949 with three subsequent waves of cases. The onset of the initial outbreak coincided with the British occupation of the islands from 1940 to 1945, and detailed analysis showed a spatial relationship between villages where MS patients lived and where British troops had been quartered. The studies in the Faeroes have concluded that MS is not only an acquired disease but that it is a transmissible disease (Kurtzke, 2000). MS has long been recognized in Iceland, but a reexamination of their rates show a rise in incidence in 1922, which plateaued until 1945 when a rise in incidence occurred over the subsequent decade. Again this followed the occupation by American, Canadian, and British military forces. Epidemiological Evidence of a Specific Virus Although late childhood or adolescent infection with a virus, followed by a long incubation has been suggested by epidemiological investigations, very little data implicate a speciWc agent. A number of reports conWrm that measles and other common childhood infections occur at a more advanced age in MS patients than in controls (Bachmann and Kesselring, 1998). A history of infectious mononucleosis is more frequent in MS patients, indicating later infection, since infections with the ubiquitous EB virus are generally asymptomatic in children and manifest with mononucleosis only when infections is delayed until adolescence and young adult life. These Wndings favoring late acquisition of common infections simply may reXect that persons who develop MS may have had more sheltered childhoods. An unexplained north to south gradient of varicella-zoster virus infection occurs with chickenpox being an almost universal disease of elementary school children in temperate zones and an infrequent disease of children in the tropics. In the tropics more adult cases, and even adult epidemics, occur (Ross, 1998). An excess of spring births of persons who develop MS has been cited to implicate an infectious etiology, but this correlation would implicate maternal or neonatal infection, which is not consistent with the migration studies. Animal exposures have been investigated extensively. Initially, the Faeroe Island outbreak was postulated to be related to canine distemper virus, a viral infection thought to have been imported by the British oYcers’ dogs. Subsequently a cluster of MS cases in

MULTIPLE SCLEROSIS

Sitka, Alaska, was noted to have followed a canine distemper outbreak 4 to 5 years previously (Cook and Dowling, 1982). Studies of MS patients have not conWrmed antibody responses to canine distemper speciWc polypeptides. Viral Infections and Exacerbations Patients often relate exacerbations of MS to psychological stress, physical trauma or physical fatigue, but prospective studies quite consistently show a relationship only with symptoms of respiratory infections (Casetta and Granieri, 2000; Marrie et al., 2000). A recent study extended the Wndings to show that exacerbations in the contest of a systemic infections lead to more sustained damage (Buljevac et al., 2002). A number of studies have also examined activation of human herpesviruses with exacerbations, particularly EB virus and human herpesvirus 6 (HHV6), but also active replication of herpes simplex type 1 has been associated with exacerbations (Ferrante et al., 2000). The obvious dilemma in these studies is the precise timing of the onset of the exacerbation and the activation of the latent virus—that is, determining which came Wrst.

Pathology Electron microscopic studies in 1964 identiWed papovavirus particles in inclusion bodies of PML and in 1965 identiWed structures resembling morbillivirus nucleocapsids in the inclusions in subacute sclerosing panencephalitis. Amid the exuberance over Wnding viruses by electron microscopy of poorly Wxed autopsy and biopsy tissues, a number of studies reported ‘‘viruslike’’ particles in MS brains. The ovoid membrane bodies of 30 to 200 nm in diameter are now thought to represent myelin breakdown products, the dense intracytoplasmic granules of 60 to 80 nm diameter probably represent nonspeciWc changes in reactive astrocytes, and the intranuclear structures in inXammatory cells originally identiWed as myxovirus nucleocapsids are now believed to be nonspeciWc alterations in nuclear chromatin. None of these or other structures have been deWnitively identiWed as viral in nature (Johnson and Herndon, 1974). Although the neuropathology of MS in covered in Chapter 31, one perspective of the pathology that may be relevant to causation is the heterogeneity of lesions in MS. Lucchinetti, et al. (2000) analyzed acute demyelinating lesions in 51 biopsies and 32 autopsies from MS patients. All cases contained at least one active lesion with inXammation, macrophages with myelin debris, and demyelination. CD3 T cells, macrophages, and occasional plasma cells characterized all patterns, but in some cases the demyelination was perivenular and in others not. Some had plaques with sharp margins, others were illdeWned. In some, oligodendrocytes persisted in demyelinated foci and remyelination was evident, in others apoptosis of oligodendrocytes was associated with wider loss of myelin associated glycoprotein. Despite this heterogeneity there was homogeneity among active lesions in the same patient. They divided the cases into four patterns, but the parallels of some to ADEM and others to PML are very suggestive of diVerent modes of pathogenesis. In viral diseases, consistency of a clinical-pathological syndrome does not imply a single causative agent. When lymphocytic choriomeningitis virus was recovered from spinal Xuid it was regarded as the cause of benign aseptic meningitis, yet subsequent studies have implicated more than 100 diVerent viruses in that syndrome. Conversely a single virus can evoke varied clinical-pathological responses such a varicella-zoster virus which can cause ADEM, Reye’s disease, acute myelitis, vasculitis, and postherpetic neuralgia. Among clinical laboratory tests, the intrathecal synthesis of IgG and the presence of spinal Xuid oligoclonal bands are hallmarks of MS. Other diseases that consistently show these abnormalities are infectious processes—neurosyphilis, neuroborreliosis, subacute sclerosing panencephalitis, HTLV-1 associated myelitis, and other chronic infections. In these diseases, intrathecal antibodies can be shown to react with viral antigens; the antigen(s) in MS are unknown but the analysis of IgG heavy chain sequences in MS brains suggest an antigen driven response rather than a nonspeciWc B-cell activation (SmithJensen et al., 2000).

973

974

40. INFECTIOUS DEMYELINATING DISEASES

Virological Studies Studies of patients with MS have implicated poxviruses, herpesviruses, rhabdoviruses, orthomyxoviruses, paramyxoviruses, coronaviruses, Xaviviruses, picornaviruses, retroviruses, and a variety of unclassiWed or mythical agents. Several parasites and bacteria have also been implicated. Assays of antibodies in serum and in spinal Xuid have consistently shown higher titers of various antibodies in MS patients than in controls, inoculations of patient Xuids or tissues into cell cultures or laboratory animals have shown changes interpreted as evidence of virus replication, or Wnally the change in quantity or topography of an agent known to persist in humans has been interpreted as suggesting a causal relationship to MS. Serological Studies In 1962, the Wrst report appeared that titers of antibodies to measles virus were higher in the serum of MS patients than controls. The same antibodies were detectable in the spinal Xuids of 75% of MS patients and not in spinal Xuids of controls (Adams and Imagawa, 1962). Initially, these Wndings were regarded with skepticism, but study after study conWrmed these odd Wndings; more than 30 conWrmations have been published. Subsequently other studies showed higher titers and intrathecal synthesis of antibodies to a variety of other viruses, but never with the magnitude or consistency of measles antibodies (Tab. 40.4). In one study, 23% of patients with MS had disproportionately high antibodies to 2 or more viruses in the spinal Xuid (Norrby et al., 1974); in another study one patient was reported who had evidence of intrathecal synthesis of antibody to 11 diVerent viruses (Salmi et al., 1983). In general, twin studies have shown higher levels in antibody in the serum or spinal Xuid of the aVected twin than of the healthy twin (Kinnunen et al., 1990). The range of responses has suggested a nonspeciWc activation of B cells, but responses against one protein of a virus and not another (Nath and Wolinsky, 1990), and studies of the antibody variable regions suggest a more speciWc response. The Wnding of higher serum antibody titers to measles is not speciWc to MS. Similar higher titers have been reported in systemic lupus erythematosus, chronic hepatitis, and Reiter syndrome. In MS, as in these other diseases, individual levels of antimeasles antibody are not remarkable, as they are in subacute sclerosing panencephalitis, but the mean titer of large group of patients is consistently higher than the mean of a group of matched controls. Furthermore, measles infection is not a prerequisite to MS, since cases of MS have been observed, presenting prior to the acquisition of measles. TABLE 40.4

Higher Antiviral Antibodies in Multiple Sclerosis Than in Controls

Serum

CSF

Measles ParainXuenza 3

Measles ParainXuenza 1, 2, 3

InXuenza C

InXuenza A, B

Varicella

Varicella

Herpes simplex

Herpes simplex

Human herpes virus—6

Human herpes virus 6

Epstein-Barr

Epstein-Barr

Rubella

Rubella Mumps Respiratory syncytial Coronaviruses Adenoviruses

Borna disease virus

Borna disease virus

HTLV-I (gag)

HTLV-I (gag)

HTLV-II

Simian Virus-5

ModiWed from Johnson (1998).

975

MULTIPLE SCLEROSIS

Isolation Reports Recovery of agents from MS has a colorful history. During the Wrst half of the 20th century extensive interest was given a putative spirochete. In 1917, the agent was claimed to have been recovered from the spinal Xuid of patients with MS inoculated into guinea pigs and rabbits. In 1952, direct staining of the agent in brain and spinal cord led to its naming as Spirochaeta myelophthora (Steiner, 1952). Interest was rekindled in 1957 with further claims of cultivation of spirochetes from spinal Xuids (Ichelson, 1957), claims that later reports failed to conWrm. This controversy was Wnally settled by extensive negative results using the precise methods recommended for cultivation but substituting autoclaved water (Kurtzke et al., 1962). In the 1930s in England, an organism tentatively named Spherula insularis, possibly a Mycoplasma, was reported to have been isolated from spinal Xuid of 176 of 189 patients with MS (Chevassut, 1930). A vaccine was made, and more than 100 patients were given the vaccine prior to an abrupt retraction (Purves-Stewart, 1931). In 1956, Toxoplasma gondii was alleged to have been isolated from spinal Xuid and blood of MS patients, but that too went unconWrmed. A number of claims of transmission to primates and other animals were made, but agents were not characterized and results remained unconWrmed (Johnson, 1985). The Wrst recovery of a virus that evoked serious consideration was the 1946 Soviet claim of recovery of a virus in mice inoculated with spinal Xuid and brain tissue of two patients with MS (Margulis et al., 1946). The virus was shown independently to be rabies virus; whether the patient diagnosis was wrong or the agents were laboratory contaminants was unknown. No laboratories conWrmed the isolations, although the original laboratory reported subsequent similar isolations. In 1964, herpes simplex virus was recovered from the spinal Xuid of a patient with MS (Gudnadottir et al., 1964). This virus subsequently was shown to be a type 2 herpes simplex strain; the type that has subsequently be isolated frequently from spinal Xuid in recurrent meningitis. This may also represent the Wrst isolation of ‘‘normal Xora’’ from MS specimens and the beginning of the persistent question of cause or eVect. Some of the viruses on Table 40.5 have subsequently been retracted as laboratory contaminants (scrapie and measles), some are thought to represent animal viruses recovered from inoculated laboratory animals (chimpanzee cytomegalovirus and corona-

TABLE 40.5 Viruses Recovered from Patients with Multiple Sclerosis (MS) Rabies virus

1946 1964

Herpes simplex virus, type 2

1964

Scrapie agent

1965

MS-associated agent

1972

ParainXuenza virus 1

1972

Measles virus

1972

Simian virus 5

1978

Chimpanzee cytomegalovirus

1979

Coronavirus

1980

SMON-like virus

1982

Tick-borne encephalitis Xavivirus

1982

HTLV-I

1985

LM7 (retrovirus)

1989

Herpes simplex virus, type 1

1989

Human herpesvirus 6

1994

Endogenous retroviruses

1998

HTLV, human T-cell lymphotrophic virus, SMON, subacute myelo-optico-neuropathy.

976

40. INFECTIOUS DEMYELINATING DISEASES

virus), and some probably do not represent viruses but only laboratory observations interpreted as representing viral activity not veriWed in independent laboratories (MSassociated agent and SMON virus). A critique of these reported isolations previously has been published (Johnson, 1998). During the National Institutes of Health studies of slow infections, tissues from MS patients were inoculated into chimpanzees. No evidence of transmission has been observed over the subsequent 30 years. This is not deWnitive evidence against a viral cause, however. During those studies tissue from PML and subacute sclerosing panencephalitis, diseases now known to be caused by viruses, were similarly inoculated into chimpanzees with negative long-term observations. Agents of Current Interest Over the past 5 years the literature on infectious agents and MS has been dominated by studies of the herpesviruses, EB and human herpesvirus 6 (HHV6), endogenous retroviruses, and a bacterium, Chlamydia pneumoniae. In contrast to prior attempts to recover a unique MS virus, these all represent ubiquitous agents that persist in humans, and studies have focused on quantitation and cellular sites of infection. In each case the diYcult question is whether changes are related to causation or whether replication and host cells changes are secondary to the immunological changes in MS. Epstein Barr virus Interest continues in the long postulated role of EB virus in MS. As mentioned earlier, the age of acquisition determines disease in this infection. Early life infections, as occur in tropical climes and impoverished communities, lead to immunity but no clinical illness; delayed infections in adolescence and young adult life often lead to the syndrome of infectious mononucleosis. Furthermore, even in case-controlled studies MS patients report a greater frequency of preceding infectious mononucleosis. Prevalence of antibodies to EB virus in MS patients is greater than in controls, and in most studies 100% have antibodies against EB, an extent of seropositivity unique to EB virus. A number of authors have suggested that EB infection is a prerequisite to development of MS (Ascherio and Munch, 2000; Munch et al., 1998; Myhr et al., 1998) Longitudinal studies have also found an association between EB virus activation and disease activity in MS patients (Wandinger et al., 2000) ). There have been several reports of patients with neurological complications of primary EB virus infections who went on to develop progressive or relapsing disease subsequently diagnosed as MS (Bray et al., 1992; Shaw and Alvord, 1987). One 6-year-old had 11 episodes of relapsing disease with high titers of EB antibodies. At death, the neuropathological diagnosis was typical MS, and PCR of the brain showed EB virus sequences (Pedneault et al., 1992). EB virus maintains latency in B cells, which are not present, or at least very rare, in normal nervous tissue. B cells are a feature of the inXammatory response in MS and other inXammatory and infectious diseases. Therefore, the presence of EB DNA determined by PCR may only reXect the presence of B cells; detection of viral proteins in cells or infectious virus in brain or spinal Xuid are evidence of active infection, but again this could represent nonspeciWc activation during the attack of MS. Human herpesvirus 6 HHV6 is a recently recovered human herpesvirus. The virus is ubiquitous with a 70 to 100% seroprevalence in adult populations worldwide. Two variants have been distinguished, and the B variant is the predominant cause of exanthem subitum in childhood. Encephalitis has long been a recognized complication of exanthem subitum. After primary infection, HHV6 remains latent primarily in T cells, but the virus is pleiotropic, with latency in B cells and CNS glial cells having been reported (Soldan et al., 2001). Similar with many other viruses, higher levels of serum antibodies and presence of spinal Xuid antibodies to HHV6 were reported in many MS patients. In addition, HHV6 DNA was detected in spinal Xuid by PCR (Wilborn et al., 1994). The report by Challoner and colleagues in 1995 made HHV6 a serious candidate as the cause of MS. They found

MULTIPLE SCLEROSIS

HHV6 DNA in the majority of MS and control brains, but protein expression was primarily in MS brains. Furthermore, immunocytochemical staining showed positive meningeal cells in both groups, but in MS lesions there was staining of cells adjacent to the plaques thought to be neurons and oligodendrocytes. Many conXicting publications have followed. HHV6 IgM in serum and spinal Xuid, higher titers of antibodies in serum, and higher frequency of antibody in spinal Xuid have all been reported in MS patients compared to controls; and all of these Wndings have been refuted in other studies. More frequent detection of HHV6 DNA in peripheral blood mononuclear cells, serum, spinal Xuid, and brain of MS patients have been reported; and again others have failed to conWrm the claims. In one study of lymphoproliferative responses to HHV6 antigens by MS patients, more frequent responses of patients was found to the A variant, although most data has implicated the B variant (Soldan et al., 2000). Several reported results raise a need for conWrmation or extension. One group reported immunocytochemical staining in 90% of sections showing active demyelinating lesions and only 13% in tissue sections free of active disease (Knox et al., 2000); this level of sensitivity and speciWcity is unique among reports. Recently elevated serum and spinal Xuid levels of membrane cofactor protein CD46 were reported in MS patients; this is important because CD46 is a receptor for HHV6 (as well as measles, the traditionally most implicated virus) and is a regulator of the complement cascade involved in antibody mediated immunopathology (Soldan et al., 2001). This provocative Wnding may lead to new mechanisms by which viruses might evoke in immune-mediated diseases. Endogenous retroviruses Human endogenous retroviruses (HERVs) are DNA sequences present within human chromosomes and make up about 2% of the human genome. The characteristic presence of long terminal repeats followed by gag, pol, and env genes identify their retroviral origins; they are thought to represent ancestral infections in which integrated DNA is now passed on in Mendelian fashion. Comparisons to ERVs of apes and old world monkeys suggest that some entered our genome 25 million years ago (Voisset et al., 1999). HERVs are defective in that they do not code infectious particles or transmit horizontally. Some do encode functional proteins, and complementation may result in virion formation. No endogenous retroviruses have been convincingly associated with human disease. The potential to enhance downstream cellular genes has led to speculation that they might be involved in autoimmune disease. They have been proposed as factors in the pathogenesis of MS, systemic lupus erythematosus, Sjogren’s disease, and type 1 diabetes (Perron and Seigneurin, 1999). In 1991 a retrovirus was reported budding from a cell line of meningeal cells established from the spinal Xuid of a patient with MS (Perron et al., 1991). Subsequently, C-type retrovirus particles were found in peripheral blood mononuclear cells cultured from several patients with MS, but this proved to be a diVerent HERV (Christensen et al., 1998). Several recent studies suggest that increased expression of these viruses is a consequence of immune activity rather than the cause. Johnston et al. (2001) showed that levels of HERV RNA increased in cultured macrophages nonspeciWcally stimulated in vitro; several HERVs were also expressed in brain tissue of patients with human immunodeWciency virus infections and with MS, correlating with tumor necrosis factor expression and macrophage activation. In another study (Dolei et al., 2002), analysis of blood for HERV reported detection in all MS patients, most patients with other inXammatory neurological disease and rarely in healthy donors. The role of these viruses as cofactors rather than simply secondary responders remains unclear. Chlamydia pneumoniae C. pneumoniae is an obligate intracellular Gram-negative bacterium. It is a common respiratory pathogen causing pharyngitis, bronchitis and atypical pneumonia. Seroprevalence is 40 to 70% in adults, with most seroconversions occurring during adolescence. Numerous attempts to relate C. pneumoniae to chronic disease have been made, most notably coronary artery disease and atherosclerosis. Because

977

978

40. INFECTIOUS DEMYELINATING DISEASES

the bacteria persist in human macrophages, PCR studies of tissues with macrophage inWltrates often are positive. After observing a single patient with apparent acute MS from whom C. pneumonia was recovered from spinal Xuid and who improved with antibiotic therapy, the Vanderbilt group undertook an extensive study. From spinal Xuid of MS patients, they cultured C. pneumoniae from 64% but recovered bacteria from only 11% of spinal Xuids from patients with other diseases. PCR was positive in 97% compared to 18% in control patients. Spinal Xuid IgG directed against the bacterium was found in 97%, compared to 18% with other neurological diseases (Sriram et al., 1999). Subsequently they reported that oligoclonal bands in spinal Xuids of patients with MS not only reacted with C. pneumoniae antigens (116 of 17 patients) but could be partially or completely adsorbed by antigens (Yao et al., 2001). A large number of contradictory reports have followed. Some have been conWrmatory to some facets, but none with the high percentages of the initial report; the majority have been negative A recent report, for example, using PCR detected C. pneumoniae in spinal Xuid of 21% of MS patients, in 43% of patients with other neurological diseases, and in no healthy controls (GieVers et al., 2001). This would suggest that inXammatory responses that recruit macrophages into the spinal Xuid, where under normal circumstances they are not found, can carry in C. pneumonia. Better standardization of both cultivation and PCR methods are needed before fair comparisons of studies can be made.

SUMMARY It is clear that a diverse array of viruses can infect the human central nervous system. The resulting viral infections can result in a wide variety of clinical and pathological symptoms. While some viruses may cause widespread inXammation and neurodegeneration, other viruses may remain latent in the CNS and only produce pathological changes during reactivation. Viral induced demyelination can occur both from direct infection of the myelin-producing oligodendrocytes, as in PML, or by indirect mechanisms that have yet to be determined. The Wnal outcome of a viral CNS infection will depend not only on viral characteristics, but also the interaction between virus and host cell. Understanding factors such as immune modulation and host cell regulation of viral gene expression could improve current methods of diagnosis or even lead to innovative methods for therapeutic intervention.

References Adams C, Diadori P, et al. (2000). Autoantibodies in childhood post-varicella acute cerebellar ataxia. Can J Neurol Sci 27, 316–320. Adams JM, Imagawa DT (1962). Measles antibodies in multiple sclerosis. Proc Soc Exp Biol Med 111. Agostini HT, Ryschkewitsch CF, et al. (1996). Genotype proWle of human polyomavirus JC excreted in urine of immunocompetent individuals. J Clin Microbiol 34, 159–164. Aksamit AJ, Jr. (1995). Progressive multifocal leukoencephalopathy: A review of the pathology and pathogenesis. Microsc Res Tech 32, 302–311. Ascherio A, Munch M (2000). Epstein-Barr virus and multiple sclerosis. Epidemiology 11, 220–224. Astrom KE, Mancall EL, et al. (1958). Progressive Multifocal Leukoencephalopathy. Brain 81, 93–127. Ault GS (1997). Activity of JC virus archetype and PML-type regulatory regions in glial cells. J Gen Virol 78 (Pt 1), 163–169. Bachmann S, Kesselring J (1998). Multiple sclerosis and infectious childhood diseases. Neuroepidemiology 17, 154–160. Beebe GW, Kurtzke JF, et al. (1967). Studies on the natural history of multiple sclerosis. 3. Epidemiologic analysis of the army experience in World War II. Neurology 17, 1–17. Behan PO, Geschwind N, et al. (1968). Delayed hypersensitivity to encephalitogenic protein in disseminated encephalomyelitis. Lancet 2, 1009–1012. Berger JR (2000). Progressive Multifocal Leukoencephalopathy. Curr Treat Options Neurol 2, 361–368. Berger JR, Major EO (1999). Progressive Multifocal Leukoencephalopathy. In ‘‘Textbook of AIDS Medicine’’ (T.C. Merigan, J. G. Bartlett, and D. Bolognesi, eds.), pp. 403–419. Williams and Wilkins, Baltimore. Boe J, Solberg CO, et al. (1965). Corticosteroid treatment of acute meningoencephalitis: A retrospective study of 346 cases. Br Med J 1, 1094–1095.

SUMMARY

Bray PF, Culp KW, et al. (1992). Demyelinating disease after neurologically complicated primary Epstein-Barr virus infection. Neurology 42, 278–282. Brooks BR (1979). Virus-associated neurological syndromes. In ‘‘Diagnosis of Viral Infections’’ (D. Lennette, S. Spector, and K. Thompson, eds.), pp. 183–203. University Park Press. Baltimore, Maryland. Brown P, Tsai T, et al. (1975). Seroepidemiology of human papovaviruses. Discovery of virgin populations and some unusual patterns of antibody prevalence among remote peoples of the world. Am J Epidemiol 102, 331–340. Buljevac, D., Flach, H. Z., Hop, W. C., Hijdra, D., Laman, J. D., Savelkoul, H. F., van Der Meche, F. G., van Doorn, P. A., Hintzen, R. O. (2002). Prospective study on the relationship between infections and multiple sclerosis exacerbations. Brain 125, 952–960. Cardenas RL, Cheng KH, et al. (2001). The eVects of cidofovir on progressive multifocal leukoencephalopathy: An MRI case study. Neuroradiology 43, 379–382. Carswell S, Alwine JC (1986). Simian virus 40 agnoprotein facilitates perinuclear-nuclear localization of VP1, the major capsid protein. J Virol 60, 1055–1061. Casetta I, Granieri E (2000). Clinical infections and multiple sclerosis: Contribution from analytical epidemiology. J Neurovirol 6 Suppl 2, S147–151. Chevassut K (1930). The etiology of disseminated sclerosis. Lancet 1, 552–560. Christensen T, Dissing Sorensen P, et al. (1998). Expression of sequence variants of endogenous retrovirus RGH in particle form in multiple sclerosis. Lancet 352, 1033. Cinque P, Pierotti C, et al. (2001). The good and evil of HAART in HIV-related progressive multifocal leukoencephalopathy. J Neurovirol 7, 358–363. Cole CN (1996). Polyomavirinae: The viruses and their replication. In ‘‘Fields Virolog’’ (B. N. Fields, D. M. Knipe, and P. M. Howley, eds.), pp 1997–2025. Lippincott-Raven, Philadelphia. Cook SD, Dowling PC (1982). Distemper and multiple sclerosis in Sitka, Alaska. Ann Neurol 11, 192–194. De Luca A, Giancola ML, et al. (1999). Clinical and virological monitoring during treatment with intrathecal cytarabine in patients with AIDS-associated progressive multifocal leukoencephalopathy. Clin Infect Dis 28, 624–628. De Luca A, Giancola ML, et al. (2001). Potent anti-retroviral therapy with or without cidofovir for AIDSassociated progressive multifocal leukoencephalopathy: Extended follow-up of an observational study. J Neurovirol 7, 364–368. Dean G (1970). The multiple sclerosis problem. Sci Am 223, 40–46. Dean G, Kurtzke JF (1971). On the risk of multiple sclerosis according to age at immigration to South Africa. Br Med J 3, 725–729. DeVries E (1960). ‘‘Postvaccinal Perivenous Encephalitis.’’ Elsevier Science, Amsterdam. Dolei A, Serra C, et al. (2002). Multiple sclerosis-associated retrovirus (MSRV) in Sardinian MS patients. Neurology 58, 471–473. Ebers GC, Sadovnick AD, et al. (1995). A genetic basis for familial aggregation in multiple sclerosis. Canadian Collaborative Study Group. Nature 377, 150–151. Esolen LM, Takahashi K, et al. (1995). Brain endothelial cell infection in children with acute fatal measles. J Clin Invest 96, 2478–2481. Fanning E (1992). Simian virus 40 large T antigen: The puzzle, the pieces, and the emerging picture. J Virol 66, 1289–1293. Feigenbaum L, Hinrichs SH, et al. (1992). JC virus and simian virus 40 enhancers and transforming proteins: Role in determining tissue speciWcity and pathogenicity in transgenic mice. J Virol 66, 1176–1182. Feigenbaum L, Khalili K, et al. (1987). Regulation of the host range of human papovavirus JCV. Proc Natl Acad Sci U S A 84, 3695–3698. Ferrante P, Mancuso R, et al. (2000). Molecular evidences for a role of HSV-1 in multiple sclerosis clinical acute attack. J Neurovirol 6 Suppl 2, S109–114. Frisque RJ, Bream GL, et al. (1984). Human polyomavirus JC virus genome. J Virol 51, 458–469. Fujinami RS, Oldstone MB (1985). Amino acid homology between the encephalitogenic site of myelin basic protein and virus: Mechanism for autoimmunity. Science 230, 1043–1045. Gasnault J, TaouWk Y, et al. (1999). Prolonged survival without neurological improvement in patients with AIDSrelated progressive multifocal leukoencephalopathy on potent combined antiretroviral therapy. J Neurovirol 5, 421–429. Gautier-Smith PC (1965). Neurological complications of glandular fever (infectious mononucleosis). Brain 88, 323–334. Gendelman HE, Pezeshkpour GH, et al. (1985). A quantitation of myelin-associated glycoprotein and myelin basic protein loss in diVerent demyelinating diseases. Ann Neurol 18, 324–328. Gendelman HE, Wolinsky JS, et al. (1984). Measles encephalomyelitis: Lack of evidence of viral invasion of the central nervous system and quantitative study of the nature of demyelination. Ann Neurol 15, 353–360. Gibbs FA, Gibbs EL, et al. (1959). Electroencephalographic abnormality i ‘‘uncomplicated’’ childhood diseases. JAMA 171, 1050–1055. GieVers J, Pohl D, et al. (2001). Presence of Chlamydia pneumoniae DNA in the cerebral spinal Xuid is a common phenomenon in a variety of neurological diseases and not restricted to multiple sclerosis. Ann Neurol 49, 585–589. Gollomp SM, Fahn S (1987). Transient dystonia as a complication of varicella. J Neurol Neurosurg Psychiatry 50, 1228–1229.

979

980

40. INFECTIOUS DEMYELINATING DISEASES

Gudnadottir M, Helgadottir H, et al. (1964). Virus isolated from the brain of a patient with multiple sclerosis. Exp Neurol 9, 85–95. Hall CD, Dafni U, et al. (1998). Failure of cytarabine in progressive multifocal leukoencephalopathy associated with human immunodeWciency virus infection. AIDS Clinical Trials Group 243 Team. N Engl J Med 338, 1345–1351. Hallervorden J (1930). Eigennartige und nicht rubriziebare Prozesse. In ‘‘Hadbuch der Geiteskranheiten’’ (O. Bumke, ed.), pp 1063–1107. Springer, Berlin. Hammond SR, English DR, et al. (2000). The age-range of risk of developing multiple sclerosis: Evidence from a migrant population in Australia. Brain 123 (Pt 5), 968–974. Hart MN, Earle KM (1975). Haemorrhagic and perivenous encephalitis: A clinical-pathological review of 38 cases. J Neurol Neurosurg Psychiatry 38, 585–591. Hartung HP, Grossman RI (2001). ADEM: Distinct disease or part of the MS spectrum? Neurology 56, 1257–1260. Hemachudha T, GriYn DE, et al. (1988). Immunologic studies of patients with chronic encephalitis induced by post-exposure Semple rabies vaccine. Neurology 38, 42–44. Herndon RM, Johnson RT, et al. (1974). Ependymitis in mumps virus meningitis. Electron microscopical studies of cerebrospinal Xuid. Arch Neurol 30, 475–479. Hirsch RL, GriYn DE, et al. (1984). Cellular immune responses during complicated and uncomplicated measles virus infections of man. Clin Immunol Immunopathol 31, 1–12. Holman RC, Janssen RS, et al. (1991). Epidemiology of progressive multifocal leukoencephalopathy in the United States: Analysis of national mortality and AIDS surveillance data. Neurology 41, 1733–1736. Hou J, Major EO (1998). The eYcacy of nucleoside analogs against JC virus multiplication in a persistently infected human fetal brain cell line. J Neurovirol 4, 451–456. Hoult JG, Flewett TH (1960). InXuenza encephalopathy and post-inXuenzal encephalitis: Histological and other observations. BMG 1, 1847–1850. Hynson JL, Kornberg AJ, et al. (2001). Clinical and neuroradiologic features of acute disseminated encephalomyelitis in children. Neurology 56, 1308–1312. Ichelson RR (1957). Cultivation of spirochaetes from spinal Xuids of multiple sclerosis cases and engative controls. Proc Soc Exp Biol Med 95, 57–58. Inui K, Miyagawa H, et al. (1999). Remission of progressive multifocal leukoencephalopathy following highly active antiretroviral therapy in a patient with HIV infection. Brain Dev 21, 416–419. Irani DN, GriYn DE (1996). Regulation of lymphocyte homing into the brain during viral encephalitis at various stages of infection. J Immunol 156, 3850–3857. Ito T, Watanabe A, et al. (2000). Acute disseminated encephalomyelitis developed after acute herpetic gingivostomatitis. Tohoku J Exp Med 192, 151–155. Jensen PN, Major EO (1999). Viral variant nucleotide sequences help expose leukocytic positioning in the JC virus pathway to the CNS. J Leukoc Biol 65, 428–438. Johnson RT (1968). Mumps virus encephalitis in the hamster. Studies of the inXammatory response and noncytopathic infection of neurons. J Neuropathol Exp Neurol 27, 80–95. Johnson RT (1985). Viral aspects of multiple sclerosis. In ‘‘Handbook of Clinical Neurology’’ (P. J. Vinken, G. W. Bruyn, and H. L. Klawans, eds.), pp 319–336. Elsevier Science, Amsterdam. Johnson RT (1998). Postinfectious demyleinating diseases. In ‘‘Viral Infections of the Nervous System’’ (R. T. Johnson , ed.), pp 181–210. Lippincott-Raven, Philadelphia. Johnson RT, GriYn DE, et al. (1984). Measles encephalomyelitis–clinical and immunologic studies. N Engl J Med 310, 137–141. Johnson RT, GriYn DE, et al. (1985). Postinfectious encephalomyelitis. Semin Neurol 5, 180–190. Johnson RT, Herndon RM (1974). Virologic studies of multiple sclerosis and other chronic and relapsing neurological diseases. Prog Med Virol 18, 214–228. Johnston JB, Silva C, et al. (2001). Monocyte activation and diVerentiation augment human endogenous retrovirus expression: Implications for inXammatory brain diseases. Ann Neurol 50, 434–442. Kabat EA, Wolfe A, et al. (1947). The rapid production of acute disseminated encephalomyelitis in rhesus monkey by injection of heterologous and homologous brain tissue with adjuvants. J Exp Med 85, 117–130. Karp CL, Wysocka M, et al. (1996). Mechanism of suppression of cell-mediated immunity by measles virus. Science 273, 228–231. Kennedy PG, Narayan O, et al. (1985). Persistent expression of Ia antigen and viral genome in visna-maedi virusinduced inXammatory cells. Possible role of lentivirus-induced interferon. J Exp Med 162, 1970–1982. Kinnunen E, Valle M, et al. (1990). Viral antibodies in multiple sclerosis. A nationwide co-twin study. Arch Neurol 47, 743–746. Knox KK, Brewer JH, et al. (2000). Human herpesvirus 6 and multiple sclerosis: Systemic active infections in patients with early disease. Clin Infect Dis 31, 894–903. Kurtzke JF (1993). Epidemiologic evidence for multiple sclerosis as an infection. Clin Microbiol Rev 6, 382–427. Kurtzke JF (2000). Multiple sclerosis in time and space-geographic clues to cause. J Neurovirol 6 Suppl 2, S134–140. Kurtzke JF, Martin A, et al. (1962). Microbiology in multiple sclerosis-evaluation of Ichelson’s organism. neurology 12, 915–922. Laghi L, Randolph AE, et al. (1999). JC virus DNA is present in the mucosa of the human colon and in colorectal cancers. Proc Natl Acad Sci U S A 96, 7484–7489.

SUMMARY

Landrigan PJ, Witte JJ (1973). Neurologic disorders following live measles-virus vaccination. Jama 223, 1459–1462. Levin MC, Lee SM, et al. (2002). Autoimmunity due to molecular mimicry as a cause of neurological disease. Nat Med 8, 509–513. Levine S, Wenk EJ (1965). A hyperacute form of allergic encephalomyelitis. Am J Pathol 47, 61–88. Lisak RP, Mitchell M, et al. (1977). Guillain-Barre syndrome and Hodgkin’s disease: Three cases with immunological studies. Ann Neurol 1, 72–78. Liu CK, Wei G, et al. (1998). Infection of glial cells by the human polyomavirus JC is mediated by an N-linked glycoprotein containing terminal alpha(2–6)-linked sialic acids. J Virol 72, 4643–4649. Lowis GW (1988). Ethnic factors in multiple sclerosis: A review and critique of the epidemiological literature. Int J Epidemiol 17, 14–20. Lucchinetti C, Bruck W, et al. (2000). Heterogeneity of multiple sclerosis lesions: Implications for the pathogenesis of demyelination. Ann Neurol 47, 707–717. Margulis MS, Soloviev VD, et al. (1946). Aetiology and pathogenesis of acute sporadic disseminated encephalomyelitis and multiple sclerosis. J Neurol Neurosurg Psychiatry. Marra CM, Rajicic N, et al. (2002). A pilot study of cidofovir for progressive multifocal leukoencephalopathy in AIDS. Aids 16, 1791–1797. Marrie RA, Wolfson C, et al. (2000). Multiple sclerosis and antecedent infections: A case-control study. Neurology 54, 2307–2310. Marsden JP, Hurst EW (1932). Acute perivascular myelinoclasis (‘‘acute disssemination encephalopmyelitis’’) in smallpox. Brain 55, 181–225. Miller HG (1953). Prognosis of neurologic illness following vaccination against smallpox. Arch Neurol 69, 695–706. Miller HG, Stanton JB (1954). Neurological sequelae of prophylactic inoculation. Q J Med 23, 1–27. Moench TR, GriYn DE, et al. (1988). Acute measles in patients with and without neurological involvement: Distribution of measles virus antigen and RNA. J Infect Dis 158, 433–442. Monaco MC, Jensen PN, et al. (1998). Detection of JC virus DNA in human tonsil tissue: Evidence for site of initial viral infection. J Virol 72, 9918–9923. Monaco MC, Sabath BF, et al. (2001). JC virus multiplication in human hematopoietic progenitor cells requires the NF-1 class D transcription factor. J Virol 75, 9687–9695. Montomoli C, Allemani C, et al. (2002). An ecologic study of geographical variation in multiple sclerosis risk in central Sardinia, Italy. Neuroepidemiology 21, 187–193. Moran AP, Prendergast MM (2001). Molecular mimicry in Campylobacter jejuni and Helicobacter pylori lipopolysaccharides: Contribution of gastrointestinal infections to autoimmunity. J Autoimmun 16, 241–256. Moreland RB, Garcea RL (1991). Characterization of a nuclear localization sequence in the polyomavirus capsid protein VP1. Virology 185, 513–518. Munch M, Riisom K, et al. (1998). The signiWcance of Epstein-Barr virus seropositivity in multiple sclerosis patients? Acta Neurol Scand 97, 171–174. Myhr KM, Riise T, et al. (1998). Altered antibody pattern to Epstein-Barr virus but not to other herpesviruses in multiple sclerosis: A population based case-control study from western Norway. J Neurol Neurosurg Psychiatry 64, 539–542. Narciso P, Galgani S, et al. (2001). Acute disseminated encephalomyelitis as manifestation of primary HIV infection. Neurology 57, 1493–1496. Nath A, Wolinsky JS (1990). Antibody response to rubella virus structural proteins in multiple sclerosis. Ann Neurol 27, 533–536. Norrby E, Link H, et al. (1974). Comparison of antibodies against diVerent viruses in cerebrospinal Xuid and serum samples from patients with multiple sclerosis. Infect Immun 10, 688–694. Ojala A (1947). On changes in the cerebrospinal Xuid during measles. Ann Med Intern Fenn 36, 321–331. Okada Y, Endo S, et al. (2001). Distribution and function of JCV agnoprotein. J Neurovirol 7, 302–306. Ozawa H, Noma S, et al. (2000). Acute disseminated encephalomyelitis associated with poliomyelitis vaccine. Pediatr Neurol 23, 177–179. Padgett BL, Walker DL (1973). Prevalence of antibodies in human sera against JC virus, an isolate from a case of progressive multifocal leukoencephalopathy. J Infect Dis 127, 467–470. Padgett BL, Walker DL, et al. (1971). Cultivation of papova-like virus from human brain with progressive multifocal leucoencephalopathy. Lancet 1, 1257–1260. Paskavitz JF, Anderson CA, et al. (1995). Acute arcuate Wber demyelinating encephalopathy following EpsteinBarr virus infection. Ann Neurol 38, 127–131. Paterson PY (1960). Transfer of allergic encephalomyelitis in rats by means of lymph node cells. J Exp Med 111, 119–136. Pedneault L, Katz BZ, et al. (1992). Detection of Epstein-Barr virus in the brain by the polymerase chain reaction. Ann Neurol 32, 184–192. Perron H, Lalande B, et al. (1991). Isolation of retrovirus from patients with multiple sclerosis. Lancet 337, 862–863. Perron H, Seigneurin JM (1999). Human retroviral sequences associated with extracellular particles in autoimmune diseases: Epiphenomenon or possible role in aetiopathogenesis? Microbes Infect 1, 309–322. Pfausler B, Engelhardt K, et al. (2002). Post-infectious central and peripheral nervous system diseases complicating Mycoplasma pneumoniae infection. Report of three cases and review of the literature. Eur J Neurol 9, 93–96.

981

982

40. INFECTIOUS DEMYELINATING DISEASES

Pohl-Koppe A, Burchett SK, et al. (1998). Myelin basic protein reactive Th2 T cells are found in acute disseminated encephalomyelitis. J Neuroimmunol 91, 19–27. Portilla J, Boix V, et al. (2000). Progressive multifocal leukoencephalopathy treated with cidofovir in HIVinfected patients receiving highly active anti-retroviral therapy. J Infect 41, 182–184. Power C, Johnson RT (2001). Neuroimmune and neurovirological aspects of human immunodeWciency virus infection. Adv Virus Res 56, 389–433. Power C, Kong PA, et al. (1993). Cerebral white matter changes in acquired immunodeWciency syndrome dementia: Alterations of the blood-brain barrier. Ann Neurol 34, 339–350. Purves-Stewart J (1931). Disseminated sclerosis-experimental vaccine treatment. Lancet 1, 440–441. Richardson EP, Jr. (1974). Our evolving understanding of progressive multifocal leukoencephalopathy. Ann N Y Acad Sci 230, 358–364. Riedel K, Kempf VA, et al. (2001). Acute disseminated encephalomyelitis (ADEM) due to Mycoplasma pneumoniae infection in an adolescent. Infection 29, 240–242. Rivers TM, Schwentker FF (1935). Encephalomyelitis accompanied by myelin destruction experimentally produced in monkeys. J Exp Med 61, 689–702. Ross RT (1998). The varicella-zoster virus and multiple sclerosis. J Clin Epidemiol 51, 533–535. Sacconi S, Salviati L, et al. (2001). Acute disseminated encephalomyelitis associated with hepatitis C virus infection. Arch Neurol 58, 1679–1681. Sadovnick AD, Armstrong H, et al. (1993). A population-based study of multiple sclerosis in twins: Update. Ann Neurol 33, 281–285. Salmi A, Reunanen M, et al. (1983). Intrathecal antibody synthesis to virus antigens in multiple sclerosis. Clin Exp Immunol 52, 241–249. Schwarz GA, Yang DC, et al. (1968). Meningoencephalomyelitis with epidemic parotitis. Arch Neurol 11, 453–462. Shah KV, Ozer HL, et al. (1977). Common structural antigen of papovaviruses of the simian virus 40-polyoma subgroup. J Virol 21, 179–186. Shaw CM, Alvord EC, Jr. (1987). Multiple sclerosis beginning in infancy. J Child Neurol 2, 252–256. Silver B, McAvoy K, et al. (1997). Fulminating encephalopathy with perivenular demyelination and vacuolar myelopathy as the initial presentation of human immunodeWciency virus infection. Arch Neurol 54, 647–650. Smith-Jensen T, Burgoon MP, et al. (2000). Comparison of immunoglobulin G heavy-chain sequences in MS and SSPE brains reveals an antigen-driven response. Neurology 54, 1227–1232. Soldan SS, Fogdell-Hahn A, et al. (2001). Elevated serum and cerebrospinal Xuid levels of soluble human herpesvirus type 6 cellular receptor, membrane cofactor protein, in patients with multiple sclerosis. Ann Neurol 50, 486–493. Soldan SS, Leist TP, et al. (2000). Increased lymphoproliferative response to human herpesvirus type 6A variant in multiple sclerosis patients. Ann Neurol 47, 306–313. Spillane JD, Wells CEC (1964). The neurology of Jennerian vaccination: A clinical account of the neurological complications which occurred during the smallpox epidemic in South Wales in 1962. Brain 87, 1–44. Sriram S, Stratton CW, et al. (1999). Chlamydia pneumoniae infection of the central nervous system in multiple sclerosis. Ann Neurol 46, 6–14. Steiner G (1952). Acute plaques in multiple sclerosis, their pathogenetic signiWcance and role of spirochaetes as etiological factor. J Neuropathol Exp Neurol 11, 343–372. Sumner C, Shinohara T, et al. (1996). Expression of multiple classes of the nuclear factor-1 family in the developing human brain: DiVerential expression of two classes of NF-1 genes. J Neurovirol 2, 87–100. Tantisiriwat W, Tebas P, et al. (1999). Progressive multifocal leukoencephalopathy in patients with AIDS receiving highly active antiretroviral therapy. Clin Infect Dis 28, 1152–1154. TaouWk Y, Delfraissy JF, et al. (2000). Highly active antiretroviral therapy does not improve survival of patients with high JC virus load in the cerebrospinal Xuid at progressive multifocal leukoencephalopathy diagnosis. Aids 14, 758–759. Thurnher MM, Thurnher SA, et al. (1997). Progressive multifocal leukoencephalopathy in AIDS: Initial and follow-up CT and MRI. Neuroradiology 39, 611–618. Tornatore C, Berger JR, et al. (1992). Detection of JC virus DNA in peripheral lymphocytes from patients with and without progressive multifocal leukoencephalopathy. Ann Neurol 31, 454–462. Tyler HR (1957). Nuerological complications of rubeola (measles). Medicine (Baltimore) 36, 147–167. Voisset C, Blancher A, et al. (1999). Phylogeny of a novel family of human endogenous retrovirus sequences, HERV-W, in humans and other primates. AIDS Res Hum Retroviruses 15, 1529–1533. von Pirquet C (1908). Das Verhalten der kutanen Tuberkulin-reaktion wahrend der Masern. Dtsch Med Wochenschr 34, 1297–1300. Voudris KA, Skaardoutsou A, et al. (2001). Brain MRI Wndings in inXuenza A-associated acute necrotizing encephalopathy of childhood. Eur J Paediatr Neurol 5, 199–202. Walker DL, Padgett BL (1983). The epidemiology of human polyomaviruses. Prog Clin Biol Res 105, 99–106. Wandinger K, Jabs W, et al. (2000). Association between clinical disease activity and Epstein-Barr virus reactivation in MS. Neurology 55, 178–184. Wang M, Tzeng TY, et al. (1999). Human anti-JC virus serum reacts with native but not denatured JC virus major capsid protein VP1. J Virol Methods 78, 171–176. Wilborn F, Schmidt CA, et al. (1994). A potential role for human herpesvirus type 6 in nervous system disease. J Neuroimmunol 49, 213–214.

SUMMARY

Wroblewska Z, Wellish M, et al. (1980). Growth of JC virus in adult human brain cell cultures. Arch Virol 65, 141–148. Wynn DR, Rodriguez M, et al. (1990). A reappraisal of the epidemiology of multiple sclerosis in Olmsted County, Minnesota. Neurology 40, 780–786. Yao SY, Stratton CW, et al. (2001). CSF oligoclonal bands in MS include antibodies against Chlamydophila antigens. Neurology 56, 1168–1176. Yiannoutsos CT, Major EO, et al. (1999). Relation of JC virus DNA in the cerebrospinal Xuid to survival in acquired immunodeWciency syndrome patients with biopsy-proven progressive multifocal leukoencephalopathy. Ann Neurol 45, 816–821. Ziegra SR (1961). Corticosteroid treatment for measles encephalitis. J Pediatr 59, 322–323.

983

C H A P T E R

41 Ischemic White Matter Damage Peter K. Stys and Stephen G. Waxman

Central nervous system axons play the critical role of transmitting electrical impulses within the central nervous system (CNS) with high Wdelity and reliability. The unique architecture of myelinated axons allows these structures to conduct action potentials rapidly in an energy eYcient manner. A coordinated organization of axonal ion channels is required to eVect this, in addition to a highly specialized myelin sheath. Together, the axonal ion channels conduct the necessary ionic currents, while the myelin sheath provides a capacitative shield that greatly reduces stray trans-axolemmal current leaks, thereby ensuring rapid, reliable, and eYcient saltatory conduction (Waxman et al., 1995). The highly specialized architecture of myelinated Wbers renders them prone to functional disruption when any of the critical components are deranged. A variety of axonal disorders is characterized by irreversible compromise of conduction through central tracts, resulting in varying degrees of clinical disability, which depends on the severity of damage and location of the aVected pathways. Common examples include acute stroke, hypoxic/ischemic white matter injury that can result in periventricular leukomalacia and cause cerebral palsy, and more chronic states such as vascular dementia from long-standing microangiopathic pathology, trauma (for example, in closed head and spinal cord injuries), and a variety of demyelinating disorders such as multiple sclerosis. Together, these disorders represent a huge personal and socioeconomic burden on Western populations. We believe that a key to devising successful therapies for these disorders lies in a thorough understanding of the fundamental mechanisms of ischemic CNS injury. While white matter injury mechanisms share a number of common steps with those seen in gray matter (for an excellent review, see Lipton, 1999), there are also unique features. This chapter summarizes our current knowledge of the deleterious events triggered to induce anoxic/ischemic damage in mammalian white matter.

THE SCOPE OF WHITE MATTER INJURY Cerebrovascular disease is a leading cause of death and disability, with close to 600,000 new cases each year in Canada and the United States alone. While much work has focused on gray matter injury mechanisms, white matter tracts are also damaged in the vast majority of strokes. Up to one-quarter of all strokes are lacunar in nature, involving predominantly white matter tracts in the brain (Fisher, 1982). Indeed, ~ 50% of the adult human CNS is composed of white matter (Zhang and Sejnowski, 2000), yet we know far less about this tissue than gray matter regions. Traumatic spinal cord injury is another example of a devastating condition aVecting mainly young adults, with more than 10,000 new cases per year in the United States (Gibson, 1992); the yearly cost of care for the estimated 250,000 individuals living with spinal cord injuries exceeds $10 billion per year. In the United States alone, it is estimated that nearly 2 million people suVered a head injury

Myelin Biology and Disorders, Volume 2

985

Copyright 2004, Elsevier (USA). All rights reserved.

986

41. ISCHEMIC WHITE MATTER DAMAGE

in 1990 (Collins, 1990). Recent data suggest that the majority of these (~ 1.5 million) suVered some form of brain injury as a result (Sosin et al., 1996). DiVuse axonal injury is a central feature of brain trauma. Microscopically, this type of injury is characterized by the appearance of numerous swollen axonal ‘‘retraction balls’’ separated from the distal Wber, mainly due to ‘‘secondary axotomy,’’ whereby axons are biochemically (rather than mechanically) transected in the Wrst hours after mild to moderate injury (Povlishock and Christman, 1995). The mechanisms of this axotomy are not well understood but seem to be accompanied by mitochondrial swelling, nodal ‘‘blebs,’’ focal loosening of the myelin sheath, disorganization of neuroWlaments, localized disruption of axoplasmic transport, and Wnally axonal disconnection (Maxwell et al., 1997). A recent study suggests focal accumulation of Ca with activation of calpain may be responsible (Buki et al., 1998), and we have recently shown that ‘‘controlled’’ Ca inXux (rather than nonspeciWc Ca overload through disrupted axolemma) via reverse Na-Ca exchange and voltage-gated Ca channels plays a key role (Wolf et al., 2001). Finally, lessons learned about ischemic injury to white matter may be very relevant to multiple sclerosis, aZicting an estimated 350,000 people in Europe and North America alone (Weinshenker, 1996). Recent work has made it clear that this disease causes irreversible disability partly, if not mainly, because of axonal degeneration, rather than demyelination (Bjartmar and Trapp, 2001; Ferguson et al., 1997; Kornek and Lassmann, 1999; Rieckmann and Smith, 2001; Trapp et al., 1998; Waxman, 2000). Although the mechanisms responsible for this axonal loss are not yet fully understood, there are indications that some mechanisms that have been elucidated from studies of ischemic white matter injury may play a role, and this may open up therapeutic opportunities.

CNS ENERGY METABOLISM Cells of the mammalian CNS require a continuous supply of oxygen and glucose to support normal metabolism and uninterrupted signaling. Indeed, it is estimated that over 50% of resting ATP consumption is accounted for by ion pumping, with the Na-K-ATPase being by far the greatest consumer (Erecinska and Dagani, 1990; Erecinska and Silver, 1989), operating to maintain normal Na and K gradients across cell membranes. While glucose consumption is two to three times greater in gray than in white matter (Clarke and SokoloV, 1994), white matter is nevertheless heavily dependent on a continuous supply of energy: optic nerves maintained in vitro quickly depolarize within minutes of the onset of anoxia, with only a small fraction of resting membrane potential supported by glycolysis (Leppanen and Stys, 1997). Conversely, blocking glycolysis either pharmacologically—for example, using iodoacetate, an irreversible blocker of glyceraldehyde 3-phosphate dehydrogenase (Devlin, 1992, p. 308) or by removing glucose—also causes a marked depolarization of central axons; however, the onset of membrane potential failure is delayed compared to anoxia, with the compound action and resting membrane potentials being well preserved for 20 minutes or more (Brown et al., 2001a; Stys, 1998; Stys et al., 1998) (Fig. 41.1). In the case of pharmacological inhibition of glycolysis, this delay is thought to be due residual consumption of energy substrates (e.g., amino acids) by the Krebs cycle and ATP generation by oxidative metabolism (Leppanen and Stys, 1997; Stryer, 1988). Removing glucose also has a delayed eVect on membrane potential and the disappearance of the compound action potential, which begin to decay after only 40 minutes of aglycemia (Brown et al., 2001a; Wender et al., 2000). In this case, with the glycolytic pathways unblocked, it is likely that once residual substrates that fuel the Krebs cycle are depleted, energy is supplied to axons by catabolism of glycogen (Wender et al., 2000). Because astrocytes contain the only substantial source of glycogen in the CNS (Koizumi, 1974; Sorg and Magistretti, 1991; Swanson et al., 1989), it is thought that these cells generate lactate through glycolysis, which is exported to, then taken up by, axons via speciWc monocarboxylate transporters (Wender et al., 2000). Although there are suggestions that neurons use lactate preferentially, especially during activation (Tsacopoulos and Magistretti, 1996), the issue of whether neurons and axons utilize glucose directly is still

EARLY CONSEQUENCES OF ENERGY DEPRIVATION

987

FIGURE 41.1 (A, B) Representative compound action and resting membrane potentials from rat optic nerve exposed to in vitro anoxia (beginning at time 0). Compound action potentials are abolished within minutes of anoxia, paralleling the quick depolarization of resting potential. Blocking the Na-KATPase with ouabain (B) reveals that a small component of resting membrane potential is supported by glycolytic ATP even after 1 hr of anoxia. (C, D) Similar records from optic nerves this time exposed to in vitro glycolytic inhibition using iodoacetate (a paradigm that is independent of residual traces of glucose or glycogen availability). In contrast to anoxia, during glycolytic block excitability and resting membrane potential are maintained for at least 20 minutes before rapidly collapsing, probably at the time of exhaustion of alternate substrates being metabolized by the Krebs cycle (see text). Reproduced from Stys, 1998, with permission.

unsettled. Recent evidence suggests that neuronal elements will use both substrates if available (Chih et al., 2001). Figure 41.2 summarizes the presumed modes of energy substrate utilization in white matter of the mammalian CNS.

EARLY CONSEQUENCES OF ENERGY DEPRIVATION The heavy reliance of white matter on a continuous supply of energy substrates implies that transmembrane ionic gradients and membrane polarization will be rapidly compromised when ATP levels fall. Using ion-sensitive microelectrodes, Ransom and colleagues demonstrated a quick rise of [K]o in optic nerve within 5 minutes of anoxia onset, from a baseline of 3 mM to ~15 mM, coupled with an acidiWcation of the extracellular space by 0.3 pH units (Ransom et al., 1992). The rise in [K]o closely parallels the rapid anoxic depolarization of optic nerves and loss of excitability (which may be reversible depending on the duration of anoxia, see below). Using Ca-sensitive microelectrodes, Brown et al. (1998) observed a fall in [Ca]o, presumably reXecting accumulation into an intracellular compartment. What these studies do not reveal is which cellular elements

988

41. ISCHEMIC WHITE MATTER DAMAGE

FIGURE 41.2 SimpliWed model depicting utilization of energy substrates in white matter. Astrocytes are known to take up glucose and metabolize it to lactate through glycolytic pathways. Under substrate-limited conditions, glial-derived lactate can support axonal function. Although not Wrmly established, under energy-replete conditions it is likely that both glia and axons can take up and metabolize glucose directly. ModiWed from Wender et al., 2000, with permission, and B. R. Ransom, personal communication.

(e.g., axons or glia) are sourcing or accumulating ions. It is well known that astrocytes are able to survive on glycolysis alone for prolonged periods of time (Callahan et al., 1990; Yu et al., 1989). Hypoxia alone causes little change in astrocytic [Na] and only a modest increase of free [Ca]i from ~150 to ~280 nM (Rose et al., 1998; Silver et al., 1997). Even ischemic astrocytes do not accumulate substantial Na acutely, but cytosolic [Ca] rises more than with hypoxia alone, to ~600 nM (Rose et al., 1998; Silver et al., 1997). Central axons however behave more like neuronal elements, being much more sensitive even to anoxia alone. Figure 41.3 shows a plot of axoplasmic Na, K and Ca changes in large optic nerve axons subjected to in vitro anoxia. Axoplasmic [K] falls from a resting level of ~150 mM to less than 10% after 1h of anoxia, while [Na] rises from ~20 to ~100 mM. Because Ca is largely bound in the cytosol, the concentrations of this ion are shown in mmol/kg dry weight because the electron probe microanalysis technique used to measure these ions (more accurately, elements) reports total (free þ bound) amounts. Total axonal Ca rises by approximately Wve-fold during 60 minutes of in vitro anoxia, but the ionized [Ca], most relevant to biological (including pathological) processes, may increase much more. Using guinea pig spinal cord slices subjected to focal compression injury in vitro, LoPachin and colleagues (LoPachin et al., 1999) observed elemental changes similar to those seen during anoxia. Thus, a rise in axoplasmic Na and Ca, and a loss of K seem to be a general feature of the response of injured nerve Wbers. We have recent evidence of a very substantial rise in axoplasmic free [Ca] using confocal microscopy and Ca-sensitive dyes. During in vitro ischemia, we observed a substantial Xuorescence increase, indicating a rise in free [Ca] into the micromolar range, possibly exceeding 10 mM or more, though accurate calibrations in myelinated Wbers are diYcult to perform (Nikolaeva and Stys, 2002; Ren et al., 2000). Figure 41.4 shows confocal images of optic nerve subjected to in vitro ischemia in perfusate containing normal [Ca] (2 mM) and in zero-Ca/EGTA bath. Surprisingly, there was a substantial increase in ionized axoplasmic Ca in the absence of bath Ca, implying release of this ion from an intracellular pool. Recent experiments from our lab indicate that endoplasmic reticulum Ca stores, controlled by ryanodine and IP3 receptors, play a major role in injury to certain white matter tracts such

INJURY MECHANISMS TRIGGERED BY ENERGY FAILURE

FIGURE 41.3 Graph of changes in axoplasmic [Na], [K] and [Ca] in large optic nerve axons subjected to in vitro anoxia (beginning at time 0). [Na] and [K] are shown as free, ionized fractions in mM (see Stys et al., 1997, for details), whereas Ca, which is almost totally bound in cells, is reported in mmol/kg dry weight. During 1 hour of anoxic exposure, axoplasmic Na increases about Wve-fold from a baseline of ~ 20 mM. There is a parallel severe loss of K. Total Ca content (including axoplasm and ER, but excluding mitochondria) increases gradually about Wve-fold over baseline. Note that normal baseline Ca content in axons approaches 1 mM as an equivalent concentration, indicating substantial intracellular sources of Ca that likely contribute to white matter injury when released during ischemia (see text).

as dorsal columns of the spinal cord. While dorsal columns respond to anoxia in a manner that is similar to optic nerves (degree of functional injury, rescue by zero-Ca perfusate) (Imaizumi et al., 1997, 1999; Li et al., 1999, 2000), ischemia (oxygen and glucose deprivation) causes far greater injury that cannot be prevented by removal of Ca from the bath (Ouardouz et al., 2003). More detailed investigation revealed that L-type Ca channel blockers are robustly protective in ischemic dorsal columns, but only in Ca-free perfusate. This apparent paradox can be explained by the fact that dihydropyridine Ca channel blockers inhibit the voltage sensor on L-type Ca channels (Rios and Brum, 1987), which is crucial for depolarization-mediated activation of ryanodine receptors and Ca release from the endoplasmic reticulum, a mechanism identical to ‘‘excitation-contraction coupling’’ in skeletal muscle. Taken together, it is very likely that perhaps the most proximal source of deleterious Ca increase during ischemia is release of this ion from ‘‘axoplasmic reticulum’’ (see Fig. 41.5), both via ryanodine receptor activation by axonal depolarization sensed by L-type Ca channels, and by activation of IP3 receptors by second messengers (Ouardouz et al., 2003; Thorell et al., 2002). In addition to release of internal Ca stores into the cytosol, there is also a net accumulation of Ca originating from the extracellular space inferred by [Ca]o measurements (Brown et al., 1998) and demonstrated directly by electron probe microanalysis (Fig. 41.3) (LoPachin and Stys, 1995; Stys and LoPachin, 1998).

INJURY MECHANISMS TRIGGERED BY ENERGY FAILURE As mentioned in the previous section, excitability of white matter tracts is abolished within minutes of the onset of anoxia or ischemia, but this does not imply an irreversible loss of

989

990

41. ISCHEMIC WHITE MATTER DAMAGE

FIGURE 41.4 Confocal images of rat optic nerve axons (arrows) loaded with the Ca-sensitive dye Oregon Green 488 BAPTA1 dextran. The Ca-sensitive green signal is weak in control nerves (A) because of the normally low resting [Ca] in healthy Wbers. In vitro ischemia in normal CSF (containing 2 mM Ca) causes axoplasmic [Ca] to rise (B). Panels C and D are the same Welds as in A and B but in pseudocolor to better illustrate the changes in [Ca]. A diVerent experiment shows control nerve (E) and a substantial Ca increase in axons bathed in Ca-free perfusate during ischemia (F), indicating release of Ca from intracellular stores. Quantitative changes in Ca-sensitive Xuorescence are shown for optic nerves subjected to in vitro ischemia in normal Ca-replete CSF (G) and Ca-free bath (H).

INJURY MECHANISMS TRIGGERED BY ENERGY FAILURE

991

FIGURE 41.5 Electron micrographs of rat dorsal column axons reveal endoplasmic reticulum proWles in the cortical as well as the central axoplasm. Circular, elongated, or irregular cisternae frequently abutted the axolemma. It is hypothesized that this ‘‘axoplasmic reticulum’’ represents internal Ca storage compartments that can be released by depolarization-induced activation of ryanodine receptors (analogous to excitation-contraction coupling in skeletal muscle) or chemically by receptor-controlled synthesis of IP3. MY: myelin; AX: axoplasm; AL: axolemma. Scale bars 200 nm. Modified from Ouardouz et al., 2003 with permission.

function. For instance in adult optic nerve, if anoxia is maintained for only 10 to 15 minutes in vitro, although electrogenesis is completely abolished, re-oxygenation allows complete functional recovery; the longer anoxia is applied, the worse the post-anoxic recovery, so that after 90 minutes of anoxic exposure at 378C, no return of function is observed (Fern et al., 1998). Predictably, in vitro ischemia produces greater injury compared to anoxia alone (Ouardouz et al., 2003; Stys and Jiang, 2002; Stys and Ouardouz, 2002; Tekkok and Goldberg, 2001). The key role of cellular Ca overload in anoxic white matter damage was demonstrated in 1990 by Stys and colleagues: removal of bath Ca (with the addition of the Ca chelator EGTA) allowed complete functional (Stys et al., 1990) and structural (Waxman et al., 1993) recovery, despite a more rapid acute loss of excitability (Stys, Ransom, and Waxman, unpublished; Tekkok and Goldberg, 2001). Indeed, Ca depletion allowed substantial functional recovery of optic nerve function even after 3 hours of continuous in vitro ischemia, whereas with Ca-replete perfusate, such an insult causes liquefaction and complete destruction of the tissue (Sim and Stys, unpublished). The central role of Ca in white matter injury was subsequently conWrmed in a number of other tracts such as spinal cord dorsal columns and corpus callosum, using various paradigms including anoxia, simulated ischemia, and traumatic injury (Agrawal and Fehlings, 1996; Brown et al., 2001a; Imaizumi et al., 1997; Li et al., 2000; Tekkok and Goldberg, 2001). Thus, it appears that excess accumulation of Ca ions in the cytosol is a central event. The key question then is, what are the mechanisms promoting Ca overload in axons and glia during injury? In a series of experiments in 1992 on anoxic optic nerve, we showed that functional injury (and presumably Ca overload, a prediction which was later conWrmed directly by electron probe x-ray microanalysis; see Stys and LoPachin, 1998) was almost completely dependent on Na inXux through Na channels which could be blocked by micromolar TTX (Stys et al., 1992a). This Na- and depolarization-dependent Ca overload was found be mediated in large part by reverse Na-Ca exchange. The Na-Ca exchanger is a key player in maintaining Ca homeostasis in all excitable cells, utilizing the transmembrane Na gradient to export one Ca ion from the cytosol for three or four Na ions Xowing in (Blaustein and Lederer, 1999; Dong et al., 2002) (although recent work indicates that a 4Na-1Ca:1K exchanger exists in the brain as well; see Dong et al., 2001; Kiedrowski et al., 2002). The electrogenic nature and reversibility of this transport system implies that collapse of the Na

992

41. ISCHEMIC WHITE MATTER DAMAGE

FIGURE 41.6 SimpliWed diagram illustrating sequence of interrelated events leading to anoxic injury of a central myelinated axon. Interruption of energy supply leads to failure of ATP-dependent pumps such as axolemmal Na-K-ATPase (1). Reduction of Na pumping across the axolemma leads to accumulation of axoplasmic Na mainly through noninactivating Na channels (2). The rise in [Na]i, coupled with depolarization caused by K eZux through a variety of K channels (3) stimulates the Na-Ca exchanger to operate in the Ca import mode, overloading the axon with damaging amounts of this cation (4).

gradient or membrane depolarization (two invariable consequences of compromised white matter tracts) will bias the exchanger toward the Ca import mode, potentially culminating in Ca overload and irreversible cellular injury. From these studies emerged a simple model of white matter anoxic injury summarized in Figure 41.6. Subsequent experiments conWrmed that this mechanism operates in many other axonal tracts as well, in response to anoxia, ischemia and various traumatic models (Imaizumi et al., 1997; Li et al., 2000; Kapoor et al., 2003; Tekkok et al., 2000; Wolf et al., 2001). Careful inspection of the time course of axonal Na accumulation (see Fig. 41.3) reveals a continuous rise in [Na] over tens of minutes, suggesting that a substantial axolemmal Na conductance persists throughout most of the anoxic period, even at a time when axons are strongly depolarized (compare with Fig. 41.1B). Classical Hodgkin-Huxley kinetics imply inactivation of Na channels in depolarized axonal membranes, yet the preceding results suggest that a Wnite Na permeability persists even in depolarized axons. Such ‘‘noninactivating’’ Na conductance has been demonstrated in many types of neurons (for a review, see Taylor, 1993) and was conWrmed directly in optic nerve (Stys et al., 1993) axons as well. This route of pathological Na Xux in injured axons may have important implications for the design of therapeutic agents, with several classes of Na channel blockers having preferential selectivity at open, noninactivating Na channels. More recent studies indicate that reverse Na-Ca exchange may not be the only route of extracellular Ca inXux. Several groups found that a variety of voltage-gated channel subtypes contribute to white matter injury. Antagonists of L-type Ca channels were found to be partially protective against optic nerve anoxic injury, with the Cav1.2 and 1.3 isoforms demonstrated immunohistochemically in axons and astrocytes (Brown et al., 2001b; Fern et al., 1995a). Curiously, the dihydropyridines nifedipine and nimodipine failed to reduce accumulation of total axoplasmic Ca induced by anoxia (Stys and LoPachin, 1998). The reason for this discrepancy is not clear, but in light of recent data (Ouardouz et al., 2003), perhaps these Ca channel blockers are protective because they reduce depolarization-induced release from ryanodine-sensitive stores (see previous section), which would explain the improved physiological outcome with no diVerence in total axonal Ca accumulation. In addition to L-type, N-type Ca channels have also been implicated in anoxic and traumatic white matter injury (Agrawal et al., 2000; Fern et al., 1995a; Wolf et al., 2001).

INJURY MECHANISMS TRIGGERED BY ENERGY FAILURE

Interestingly, while the mechanisms of axonal injury in MS are poorly understood, its pathophysiology may include some features in common with ischemia and trauma, in that an important role of Ca is suggested by several lines of evidence (for a review see Rieckmann and Maurer, 2002): glutamate, released by activated macrophages/microglia, has deleterious eVects on CNS white matter, in part through activation of glial AMPA/ kainate receptors (Li and Stys, 2000; Pitt et al., 2000; Stys and Li, 2000). We have evidence that group I metabotropic glutamate receptors may be involved in the release of toxic amounts of Ca from IP3-dependent stores in white matter (Stys and Ouardouz, 2002). Moreover, a very interesting recent study by Goldberg and colleagues suggests that AMPA/kainate receptor-mediated damage to oligodendrocytes in vitro induces release of reactive oxygen species that secondarily injure axons (Underhill and Goldberg, 2002). From these studies, it is very likely that the toxic amounts of glutamate known to be released in inXammatory areas of MS plaques may promote Ca-dependent damage not only to glia but to axons as well, ultimately causing axonal degeneration. Nitric oxide (NO) is another factor that is produced in relatively high concentrations by macrophages (and other cells) at sites of inXammation (Bo et al., 1994; Cross et al., 1998). NO causes axonal dysfunction probably because of an interaction with mitochondrial cytochrome c oxidase, leading to a competitive inhibition of O2 utilization (Bolanos et al., 1997; Cooper, 2002; Garthwaite et al., 2002; Smith et al., 2001); thus, it is possible that tissue in the vicinity of an inXammatory lesion remains ‘‘chemically hypoxic.’’ Such a state could in turn invoke the Ca-dependent injury processes that have been described in anoxic/ischemic white matter. Such mechanistic overlap might allow us to capitalize on knowledge gained from work on white matter anoxia/ischemia to devise successful protective strategies for inXammatory white matter disorders. Complicating the issue of white matter damage in inXammatory disorders further, recent studies indicate that, in addition to demyelination and axon degeneration, dysregulated expression of ion channels (an ‘‘acquired channelopathy’’) occurs in experimental allergic encephalomyelitis (EAE) and MS (Black et al., 2000; Waxman, 2001). An interesting recent report indicates that Cav2.2, the pore-forming subunit of N-type Ca channels, not only accumulates within axonal spheroids of actively demyelinating lesions in MS and EAE but is also inserted into the axolemma, suggesting that injured axons may be preferentially susceptible to Ca-mediated injury by ion entering through ectopically targeted channels (Kornek et al., 2001). As in many other cell types (Orrenius and Nicotera, 1996; Schanne et al., 1979), Ca is thought to represent the ‘‘Wnal common pathway’’ of cell injury in white matter as well. It is highly likely that once cytosolic [Ca] rises excessively, a large number of Ca-dependent pathways are stimulated culminating in irreversible tissue destruction. Many crucial biochemical pathways are modulated by Ca ions including calpains, phospholipases, endonucleases, NO synthase, mitochondrial free radical generation, protein kinase C, and others. Experiments in both traumatic, anoxic/ischemic and immune-mediated demyelinating paradigms clearly indicate the important role of calpain overactivation in producing breakdown of key structural proteins including spectrin, neuroWlament, and myelin basic protein (Buki et al., 1999; Jiang and Stys, 2000; McCracken et al., 2001; Schumacher et al., 2000, 2001; Stys and Jiang, 2002). Calpain substrates are not limited to just these proteins, with additional relevant targets for possible proteolysis including the Ca-ATPase (Salamino et al., 1994); tau, tubulin, ankyrin (Rami et al., 1997); the myelin proteins MAG and PLP (Deshpande et al., 1995; Schaecher et al., 2001; Shields et al., 1999); calmodulin-binding proteins (e.g., G-proteins), protein kinase C, calcineurin, phospholipase C (KampX et al., 1997); NMDA and AMPA receptors (Bi et al., 1998a, 1998b, 1998c; Gellerman et al., 1997); L-type Ca channels (Hell et al., 1996); NCAM and N-cadherin (Covault et al., 1991); and calpain also appears to play an important role in the induction of apoptosis and the mitochondrial permeability transition (Lipton, 1999) (discussed later). In all likelihood, many if not most of the other Ca-dependent pathways are also overdriven in injured white matter, conspiring to cause irreversible damage to this tissue. The corollary is that blocking only one (e.g., calpain) or even a few of the Ca-stimulated

993

994

41. ISCHEMIC WHITE MATTER DAMAGE

pathways would not improve functional outcome, as has been shown in optic nerve: predictably, direct pharmacological calpain inhibition was very eVective at reducing calpain-dependent spectrin degradation (and by inference calpain activation in general), but did not at all improve physiological outcome (Jiang and Stys, 2000; Stys and Jiang, 2002). Although white matter is structurally and functionally simpler than gray matter, recent data indicate that many more signaling pathways may be involved in the genesis of white matter ischemic injury than was previously appreciated (e.g., Fig. 41.6). A thorough discussion of each is beyond the scope of this chapter, and the reader is referred to the cited references for more details. In addition to ‘‘direct’’ Ca sourcing pathways such as NaCa exchange, voltage-gated Ca channels and intracellular Ca release channels, other pathways modulate the degree of injury, including somewhat surprisingly, several neurotransmitters. For instance, both endogenously released GABA and adenosine are protective against anoxia in the optic nerve: application of exogenous GABA or adenosine, or blocking reuptake of GABA or adenosine with nipecotic acid or propentofylline, respectively, are protective, whereas blocking GABA-B (phaclofen) or adenosine receptors (theophylline) worsens outcome (Fern et al., 1994, 1995b, 1996b). Release of both adenosine and GABA in white matter regions was conWrmed in an in vivo cat model of global cerebral ischemia (Dohmen et al., 2001), lending further support to the ‘‘autoprotective’’ role of these substances proposed by Fern et al. (1996a). These two neurotransmitter receptors appear to modulate a convergent pathway involving a G-protein/protein kinase C cascade, although which molecular complex is then targeted and which cellular elements are protected by this mechanism (i.e., axon cylinder versus glia) is unknown. Most investigations on white matter injury have been focused on aberrant cation Xux, but Malek and Stys (2002) have data indicating that anion transporters may also contribute to damage. Evidence from the anoxic optic nerve model suggests that Cl channels attempt to normalize a Cl dysequilibrium mediated by the K-Cl co-transporter, itself driven to accumulate abnormal amounts of Cl because of the collapse of the K gradient. Thus, blocking certain Cl channels with niXumic acid worsens injury, but blocking the K-Cl co-transporter with furosemide is partially protective. Analysis of waveshapes suggests that these eVects may be mediated in large part at the myelin sheath, so that pathological Cl-dependent volume changes might be disrupting axo-glial architecture leading to slowing and failure of action potential propagation (Malek et al., 2002). Reperfusion injury has been implicated in many tissues including brain, heart, and even in spinal cord white matter directly (Darley et al., 1991; Jalc et al., 1995; White et al., 2000). As if starvation of the CNS of energy substrates is not bad enough, ironically, restoring oxygen and glucose supply can be additionally injurious. Free radical and NO production are both stimulated during the reperfusion period, and mitochondrial matrix Ca rises to an extreme degree, exceeding its baseline level by ~30 to 100-fold; this is true for organelles in both hippocampal neurons (Taylor et al., 1999) and white matter axons (LoPachin and Stys, 1995; Stys and LoPachin, 1996). Such a large Ca accumulation will not only damage mitochondria, as these quantities of ion precipitate out of the aqueous phase, but the accumulating electrical charge carried by a rapid inXux of Ca ions will also severely hamper attempts by a rejuvenated electron transport chain to restore the negative matrix potential; this potential is essential for the resumption of ATP synthesis, rather than consumption of precious glycolytically-derived ATP, which occurs when mitochondria are depolarized (Nicholls and Budd, 2000). Moreover, high mitochondrial Ca loads will also promote opening of the mitochondrial permeability transition, further accelerating the rundown of any remaining electrochemical gradients across the inner membrane. Therefore, paradoxically, these events may induce a permanent state of ‘‘chemical hypoxia’’ and a worsening energy deWcit at a time when cellular energy metabolism should be restarting. A recent report demonstrated an apoptotic mode of cell death, particularly involving oligodentrocytes in traumatically injured spinal cord white matter (Casha et al., 2001). It is interesting to speculate on the link between the mitochondrial permeability transition (promoted by excessive Ca levels), release of cytochrome c, and the induction of apoptosis by activation of caspases (Lipton, 1999; MacManus and Buchan, 2000), perhaps reXecting an additional,

WHITE MATTER EXCITOTOXICITY

more delayed, deleterious eVect of mitochondrial Ca overload that occurs during, and especially after, the termination of ischemia and other insults.

WHITE MATTER EXCITOTOXICITY White matter is a tissue devoid of chemical synapses in the traditional sense, but this does not preclude the possibility of physiological and pathological signaling between elements involving traditional neurotransmitters. Indeed, the previous section describes experiments clearly demonstrating a role of endogenously released GABA and adenosine in white matter anoxic injury. While gray matter ‘‘excitotoxicity’’ has been the subject of intense study for several decades, the eVect of glutamate on white matter has received attention only in recent years, with this excitotoxin implicated in all modes of white matter injury examined so far. Numerous studies have shown that both oligodendrocytes and astrocytes possess glutamate receptors of the AMPA and kainate (but not NMDA) subtypes (Agrawal and Fehlings, 1997b; Garcia-Barcina and Matute, 1996; Jensen and Chiu, 1993; Matute et al., 1997; for reviews see Matute et al., 2002; Steinhauser and Gallo, 1996). White matter astrocytes express all AMPA and kainate receptor subunits except GluR4. In contrast, oligodendrocytes express only GluR3 and GluR4 AMPA receptor subunits (notably lacking GluR2), as well as all kainate subunits except GluR5 (Garcia-Barcina and Matute, 1996, 1998; Matute et al., 2002). The absence of GluR2 in oligodendrocytes may render them particularly susceptible to Ca Xux through these Ca-permeable receptors. Persistent activation of these non-NMDA ionotropic receptors causes injury to oligodendrocytes both in cell culture and in vivo (Matute et al., 1997, 1998; McDonald et al., 1998; Yoshioka et al., 1995, 1996). Astrocytes are more resistant to excitotoxicity, but can be severely injured when exposed to AMPA receptor agonists particularly when desensitization is blocked (David et al., 1996). As an extension to these Wndings, a number of investigators have shown that glutamate receptor antagonism is protective against a number of insults aVecting white matter. Spinal cord injury examined using both an in vitro clip compression model (Agrawal and Fehlings, 1997b) and an in vivo contusion model (Rosenberg et al., 1999a; Wrathall et al., 1997) was ameliorated in the presence of 2,3-dihydro-6-nitro-7sulfamoyl-benzo(f)quinoxaline (NBQX), a selective antagonist of AMPA/kainate receptors. More recent work has conWrmed that AMPA/kainate receptors play an important role in hypoxic/ischemic white matter injury as shown by in vitro models, for example, corpus callosum (Tekkok and Goldberg, 2001) and spinal dorsal column (Li et al., 1999). This mechanism was conWrmed in vivo both in the adult brain (McCracken et al., 2002) and in the immature animal, with oligodendrocytes being particularly vulnerable (Follett et al., 2000). Predictably, NMDA receptor antagonists are ineVective at protecting white matter from ischemic injury (Yam et al., 2000), in line with the known lack of eVects of NMDA receptor activation (or inhibition) in this tissue (Agrawal and Fehlings, 1997b; Li and Stys, 2000). Interestingly, two recent studies using a model of EAE reported a beneWcial eVect of systemically administered NBQX, with reduction of oligodendroglial and axonal damage, along with a parallel improvement in clinical scores (Pitt et al., 2000; Smith et al., 2000). Finally, a role for metabotropic glutamate receptors in ischemic and traumatic white matter injury has been proposed, which may actually couple back to the potentially very important mechanism of release of internal Ca stores, via a phospholipase C-dependent mechanism acting on IP3 receptors (Agrawal et al., 1998; Stys and Ouardouz, 2002). Taken together, these data strongly suggest that as in gray matter, a variety of disorders damage white matter via excitotoxic mechanisms. Important questions that arise include which white matter elements are injured by excessive glutamate release? And how is this excitotoxin released in a tissue without synaptic machinery? Using an isolated spinal white matter preparation, Li and Stys (2000) showed that activation of AMPA receptors caused signiWcant functional injury to dorsal columns measured electrophysiologically. Immunohistochemistry revealed that oligodendrocytes and astrocytes were damaged, revealed by an increase in calpain-mediated breakdown of the cytoskeletal protein spectrin (Fig. 41.7). In addition, using a marker for degenerated myelin basic protein (Matsuo et al., 1998), it

995

996

41. ISCHEMIC WHITE MATTER DAMAGE

FIGURE 41.7 Confocal microscopic images of dorsal columns stained with standard markers (neuroWlament, CNPase, GFAP to identify axon cylinders, oligodendrocytes, and astrocytes, respectively) and markers of cellular injury after 3 hour in normal CSF or CSF containing 1 mM glutamate. Panels A, B, and C show tissue double stained for neuroWlament (green) and degenerated myelin basic protein (red). Control images show no myelin damage, whereas exposure to glutamate caused marked injury to myelin (arrowheads). Panels D and E identify oligodendrocytes using CNPase (green), showing cytoskeletal damage revealed by spectrin breakdown (‘‘SBP,’’ red) in glutamate-treated versus control tissue. GFAP (green in panels F and G) identiWes astrocytes that also sustained cytoskeletal damage. In contrast, axon cylinders showed no appreciable increase in spectrin breakdown after an equivalent treatment (not shown). Scale bars 10 mm. ModiWed from Li and Stys, 2000, with permission.

was shown that the myelin sheath itself is damaged by AMPA receptor stimulation (Figs. 41.7B and 41.7C) (Li and Stys, 2000). While it is not yet clear whether myelin damage results primarily from overactivation of Ca-permeable AMPA receptors, or whether the sheath degenerated secondarily as a result of injury to the cell body of the oligodendrocyte, immunohistochemistry suggests that GluR4 (but not GluR2) subunits may be located on the myelin sheath itself (Li and Stys, 2000), raising the possibility that this structure may be directly vulnerable to elevated ambient glutamate levels. Whether or not axons per se are vulnerable to glutamate exposure is less certain. In contrast to glia, axons did not exhibit

WHITE MATTER EXCITOTOXICITY

any discernible breakdown of spectrin in response to an excitotoxic insult (Li and Stys, 2000). On the other hand, Tekko¨k and Goldberg (2001) presented evidence of axonal protection by NBQX in an in vitro model of central white matter ischemic injury. However, recent evidence suggests that axonal degeneration in response to an excitotoxic challenge occurs secondarily to oligodendroglial injury, mediated by generation of free radicals (Underhill and Goldberg, 2002). How glutamate eVects injury is poorly understood, but several mechanisms are possible. The Wrst is simple Ca overload mediated by Ca-permeable AMPA receptors. During periods of energy depletion such as hypoxia/ischemia, what might normally be a manageable Ca load could now become a toxic accumulation leading to overactivation of the many Ca-dependent systems discussed earlier. A second mechanism that is perhaps important during lower intensity but more chronic exposures to glutamate does not involve not a receptor-dependent pathway, but the uptake of glutamate by oligodendrocytes via a glutamate-cystine exchange mechanism. This in turn depletes cystine and thereby glutathione, rendering the cells vulnerable to oxidative stress (Oka et al., 1993). A similar mechanism has also been proposed for astrocytes (Chen et al., 2000). This mechanism may represent a component of the free radical-mediated axonal damage originating from oligodendrocytes challenged with glutamate (Underhill and Goldberg, 2002). There is now little doubt that many white matter tracts are vulnerable to excitotoxins, whether applied exogenously or released endogenously in the setting of an insult such as ischemia, trauma, or immune-mediated demyelination. In general, glutamate can be released (1) in a vesicular manner (i.e., Ca-dependent synaptic release), (2) in a nonvesicular manner by reversal of the Na-dependent glutamate transporters (Attwell et al., 1993), (3) through volume-sensitive anion channels (Rutledge et al., 1998), or (4) by exocytosis from astrocytes (Pasti et al., 2001). As noted in previous sections, anoxia, ischemia, and trauma all result in marked ionic deregulation mainly in axons, causing depolarization, loss of K and a rise in Na. A major mechanism for terminating excitatory synaptic transmission is uptake of glutamate by Na-dependent glutamate transporters. This transport system couples the movement of glutamate or aspartate with Na and H in exchange for K in an electrogenic manner, with a typical stoichiometry of 3 Na, 1 H, 1 glutamate: 1 K (Levy et al., 1998; Zerangue and Kavanaugh, 1996). This ratio of ion coupling implies electrogenic transport, indicating that collapse of Na and K gradients, together with depolarization, will drive this transporter in the reverse, glutamate export mode, in a manner similar to that proposed for physiological release of glutamate in immature optic nerve (Kriegler and Chiu, 1993). Cytoplasm, including axoplasm, is known to contain millimolar concentrations of glutamate that far exceed the low micromolar levels in brain extracellular space (Attwell et al., 1993; Fonnum, 1984). Calculations suggest that extracellular glutamate concentrations may reach hundreds of micromolar in response to an anoxic/ischemic insult in white matter (Li et al., 1999), supported by direct measurements of elevated extracellular glutamate levels using microdialysis in vivo (Graf et al., 1998). Further evidence for a major role of reverse Na-dependent glutamate transport was provided by the signiWcant protection against both anoxia and trauma in spinal cord white matter, aVorded by dihydrokainate and L-trans-pyrrolidine-2,4-dicarboxylic acid (Li et al., 1999), speciWc inhibitors of Na-dependent glutamate transport (Arriza et al., 1994; GriYths et al., 1994). Using a semiquantitative immunohistochemical technique, these authors went on to show that it is mainly axon cylinders, and to a lesser extent oligodendrocytes, that source glutamate during an anoxic challenge (Li et al., 1999), in keeping with the proposed order of ischemic vulnerability of these elements in adult white matter (axons > oligodendrocytes >> astrocytes). During ischemia (versus anoxia as in the above study), however, immature oligodendrocytes release substantial quantities of glutamate that in turn activates receptors on these same cells to cause death, in eVect creating a ‘‘fatal glutamate release feedback loop’’ (Fern and Moller, 2000). Other modes of glutamate release have not been investigated in injured white matter, but may also include anion channel mediated release, or exocytotic release during more severe injury where astrocytic [Ca] would be expected to rise.

997

998

41. ISCHEMIC WHITE MATTER DAMAGE

RATIONAL NEUROPROTECTIVE STRATEGIES In the past 10 years, our knowledge of the basic mechanisms of white matter injury has progressed signiWcantly from the relatively simple three-step model shown in Figure 41.6 to a much more comprehensive and complex picture summarized in Figure 41.8. The most eVective therapeutic strategies can now be rationally proposed by carefully observing the interdependencies of the various steps in the injury cascade. One can see that, for example, a noninactivating voltage gated Na conductance (probably due, at least in part in mature axons, to Nav1.6, which displays a prominent persistent conductance; see Smith et al., 1998) occupies a central location in the injury process for the following reasons: (1) during Na pump failure, this conductance will provide a pathway for inXux of Na into the axoplasm, in turn promoting K eZux through a variety of K channels, resulting in axonal depolarization and axoplasmic Na accumulation. (2) The depolarization will in turn gate Cav1.2, which will activate ryanodine receptors and release Ca from intracellular stores.

9

AMPA/KA-R

AMPA/KA-R

Extracellular Space 12

K

3

1

Cl

4

O2 + glucose

mGluR1 3Na

2K

KCC

MY

Na-K ATPase

Na

3Na

1Ca

GluR4

NaCaX

GABA-B ADENOSINE

K

?

Cav1.2

G

ATP K 10

PKC

mGluR1 AMPA-R glut Na

8

2

Na

RyR1

K

PLC

9

Ca IP3R

“AR”

? 11

“AR”

Ca-ATPase Ca

6

spectrin, NF... calpain phospholipase C phospholipase A2 protein kinase C nitric oxide synthase

5

Axoplasm

Ca-ATPase Ca

7

NO

FIGURE 41.8 Diagram illustrating interrelated events leading to injury of a central myelinated axon. Interruption of energy supply leads to failure of ATPdependent pumps (1). Perhaps the Wrst source of raised axonal [Ca] is from internal stores, which may be released by an ‘‘excitation-contraction coupling’’-like mechanism and by generation of IP3 from activation of mGluR1 (2). Reduction of Na pumping across the axolemma leads to accumulation of axoplasmic Na via a noninactivating Na conductance (3). The rise in [Na]i, coupled with depolarization caused by K eZux through a variety of K channels (10) stimulates the Na-Ca exchanger to operate in the Ca import mode (4). The accumulation of axonal Ca in turn leads to mitochondrial injury, especially during reoxygenation (5), and to activation of a number of Ca-dependent enzyme systems that damage the Wber (6). One of these is nitric oxide synthase, which may generate suYcient quantities of NO to further inhibit mitochondrial respiration (7). Some Na inXux may occur through Na/K permeable inward rectiWer channels (8) (Eng et al., 1990; Stys et al., 1998). Glial injury, especially oligodendrocytes and myelin (MY) damage, is exacerbated by excess glutamate release by the Na-K-glutamate transporter, which releases this transmitter under conditions of axoplasmic Na loading and depolarization. AMPA receptors on astrocytes, oligodendrocytes, and even myelin mediate direct damage to these structures (9). Endogenously released transmitters such as GABA and adenosine appear to play an ‘‘autoprotective’’ role (11). Recent evidence also suggest anion transporters such as the K-Cl co-transporter participate in volume dysregulation in glia and the myelin sheath, contributing to conduction abnormalities (12). The locations of the various channels and transporters are drawn for convenience and do not necessarily reXect their real distributions in axons.

999

RATIONAL NEUROPROTECTIVE STRATEGIES

(3) Depolarization and increased [Na]i will together drive the reversal of the Na-Ca exchanger leading to axonal Ca overload and reverse Na-dependent glutamate transport, which will cause release of toxic amounts of glutamate, leading to activation of AMPA/ kainate and metabotropic glutamate receptors. For these reasons, targeting Na channels is an attractive option. A major problem with such a strategy is that Na channels also play a key physiological role: the genesis of action potentials. While TTX, a speciWc, state-independent blocker of voltage-gated Na channels (Catterall, 1980), is very protective in a variety of white matter injury paradigms both in vitro and in vivo (Imaizumi et al., 1997; Rosenberg et al., 1999b; Stys et al., 1992a; Tekkok and Goldberg, 2001; Teng and Wrathall, 1997), it is not a viable therapeutic option in the clinical setting because of the potency of the Na channel blocking eVect. Certain agents from the local anesthetic and anti-arrhythmic classes are preferentially active at the open conformation of the voltage-gated Na channel (Khodorov, 1991; Wang et al., 1987; Yeh and Tanguy, 1985) and should therefore be relatively selective for the noninactivating Na channel subtype implicated in axonal injury. These ‘‘use-dependent’’ or ‘‘phasic’’ Na channel blockers have the potential of allowing normal signaling to proceed unhindered along axons (because the time a rapidly inactivating Na channel spends in the open conformation during an action potential is very brief, therefore the agent will have little opportunity to access its binding site on the channel protein), yet eVectively block a persistently open state that occurs during pathological conditions with prolonged depolarization. This hypothesis was tested using analogs of local anesthetics and anti-arrhythmics known to preferentially block open, noninactivation Na channels, in the in vitro anoxic optic nerve model (Stys et al., 1992b; Stys, 1995). Examples are shown in Figure 41.9. While the prototypical local anesthetic lidocaine is an eVective neuroprotective agent, it exerts its actions at the expense of severe depression of preinjury action potentials. In contrast, the permanently charged quaternary lidocaine analog QX-314 was very eVective

pre-anoxia / in drug post-anoxia / wash 100

CAP area recovery (%)

80

60

40

20

0 none

lidocaine 1m

QX-314 0.3m

prajmaline 10µ

tocainide 1m

FIGURE 41.9 Bar graph showing examples of eVects of various Na channel blockers on pre-anoxic and post-anoxic optic nerve compound action potential (CAP). Gray bars show the degree of depression of excitability in optic nerve before anoxia is applied. Black bars show CAP recovery after anoxia and wash of drug. Control recovery without blockers is about 20 to 30% of pre-anoxic CAP area. Lidocaine (1 mM) is a very eVective neuroprotectant, but at the expense of severe depression of electrogenesis. In contrast, ‘‘use-dependent’’ Na channel blockers, particularly the permanently charged analogs such as QX-314 and prajmaline that are thought to be more selective for the open conformation of the Na channel, are highly neuroprotective with minimal anesthetic eVect. Data from Stys, 1995.

1000

41. ISCHEMIC WHITE MATTER DAMAGE

at a concentration that showed little inhibition of normal electrogenesis. Charged molecules such as QX-314 are thought to be more selective for the open conformation of Na channels (Khodorov, 1991; Wang et al., 1987; Yeh and Tanguy, 1985), which is likely the main reason for its favorable proWle. A serious practical problem with charged compounds is their poor penetration across the blood-brain barrier. QX-314, for instance, although it is a very eVective neuroprotectant in vitro, fails to enter into the rat CNS to any measurable degree after systemic administration (Stys, unpublished observations). We are then faced with the paradoxical requirement of a charged species for maximal open Na channel blockade versus neutrality to allow CNS penetration. A compromise can be reached by selecting an ionizable compound with a pKa near physiological pH that will exist in both neutral and charged forms. Studies with the antiarrhythmic mexiletine, a primary amine with a pKa of 8.4, showed that this drug is not only eVective in vitro against optic nerve anoxia but also penetrates into the CNS reaching neuroprotective concentrations after intraperitoneal injection (Hewitt et al., 2001; Stys and Lesiuk, 1996). Moreover, the intracellular acidosis that occurs during anoxia/ischemia will increase the proportion of the protonated form, trapping more drug in the cytosol, further raising the concentration of charged drug where it is most needed. This rationale was conWrmed with other Na channel blocking agents including anticonvulsants and a number of other antiarrhythmics and local anesthetics (Fern et al., 1993; Stys et al., 1992b; Stys, 1995). The in vivo situation is likely far more complex (discussed later), which may be why these agents confer only modest protection in the whole animal subjected, for example, to an in vivo spinal cord injury (Agrawal and Fehlings, 1997a). Moreover, in the clinical setting, systemic side eVects of drugs need to be considered. Na channel blockers may have potent adverse eVects on peripheral organs such as the cardiovascular system. Nevertheless, proof of principle in an in vivo model of axonal degeneration is provided by the observation by Lo et al. (2002) of robust protection of white matter axons in EAE with phenytoin at serum levels within the range used in the clinical setting (Fig. 41.10) and by Bechtold and colleagues (2002) using the antiarrhythmic Xecainide at doses that had very few adverse clinical eVects. Recent advances in our molecular understanding of Na channel subtype distributions in various organs and diVerent parts of the CNS raises the potential for the design of agents that will speciWcally target a certain Na channel isoform and even a speciWc conformational state using knowledge of molecular pharmacology. In the case of white matter, mature myelinated Wbers are endowed mainly with Nav1.6 at nodes of Ranvier (Caldwell et al., 2000), so targeting these channels with selective agents may minimize peripheral cardiovascular side eVects while maximizing the desired blocking action. Anoxic in vitro models of white matter injury have been very useful for dissecting out injury mechanisms, but may display only a subset of the steps unleashed during more severe (and perhaps more clinically realistic) injury in vivo. For instance, as eVective as TTX is at protecting white matter against anoxia in vitro (Imaizumi et al., 1997; Stys et al., 1992a), this agent is ineVective during more severe insults including more prolonged simulated in vitro ischemia using oxygen-glucose deprivation (Stys and Ouardouz, 2002), where previously dormant pathways may now be recruited. In ischemic (as opposed to anoxic) dorsal columns, protection is only apparent if Ca is removed from the bath and Na channels are blocked (Ouardouz et al., 2003), suggesting that during ischemia internal Ca stores are unloaded independently of Ca entry across cell membranes, with both sources needing to be blocked for protection. If this observation is conWrmed, it would suggest that neuroprotective strategies designed to control Ca inXux across plasma membranes are necessary but not suYcient to fully protect tissue after more severe insults; optimal protection against ischemia will likely need to simultaneously address the release of internal Ca stores. Another target apparent from Figure 41.8 is the glutamate receptor, particularly the AMPA/kainate class. This receptor appears to be primarily involved in damage to glia and myelin. A number of studies have conWrmed the eYcacy of AMPA receptor antagonists such as NBQX and GYKI52466 in in vitro anoxic, ischemic and traumatic white matter injury models (Agrawal and Fehlings, 1997b; Li et al., 1999; Stys and Ouardouz, 2002).

1001

RATIONAL NEUROPROTECTIVE STRATEGIES

50,000

*

* Total optic nerve axons

40,000

30,000

20,000

10,000

0 CONT+Phen

EAE

EAE+Phen

FIGURE 41.10 Total number of axons within optic nerves of control (no EAE, treated with phenytoin), EAE, and phenytoin treated EAE mice. There is a signiWcant decrease in number of optic nerve axons in EAE. Phenytoin has a protective eVect and produces a signiWcant increase in the number of axons, compared to untreated EAE. Bars, means + SD, * p < 0.001 compared to EAE. Reproduced from Lo et al., 2002, with permission.

Equally important is the recent observation of white matter protection by AMPA receptor blockers in several in vivo models of ischemia, spinal cord injury, and auto-immune demyelination (Kanellopoulos et al., 2000; McCracken et al., 2002; Pitt et al., 2000; Rosenberg et al., 1999a; Smith et al., 2000; Wrathall et al., 1994, 1997). Coupled with the known beneWcial eVects of these agents against gray matter injury (Akins and Atkinson, 2002), this class of drug represents an attractive option and deserves further study. While much of our understanding of white matter injury mechanisms has been obtained from in vitro models of anoxia or ischemia, emerging evidence suggests an unexpected degree of overlap with other white matter disorders, including trauma and immunemediated demyelination. In the latter case, NO produced in inXammatory lesions may play a key role. Kapoor et al. (2003) have hypothesized that NO acts by impeding mitochondrial respiration. Drawing parallels from work on white matter anoxic injury, they have shown that partial blockade of Na channels with low doses of Xecainide or lidocaine, or of the Na-Ca exchanger with bepridil, have a protective eVect on axons exposed to NO. Extending these predictions to in vivo models of demyelination, two groups have recently demonstrated the utility of state-dependent Na channel blockers in ameliorating histological damage and clinical disability in an animal model of MS, experimental allergic encephalomyelitis (EAE). Bechtold et al. (2002) showed that Xecainide, a class I anti-arrhythmic that blocks open Na channels preferentially (Ragsdale et al., 1996), and which reduces NO-mediated axonal injury (discussed earlier), was signiWcantly neuroprotective in a model of chronic relapsing EAE. Using a diVerent class of Na channel blocking agent, Lo et al. (2002) have demonstrated that the anticonvulsant phenytoin (which is partially protective in an anoxic white matter model; see Fern et al., 1993), can ameliorate clinical progression and prevent axonal degeneration in a similar model. Finally, with the recent advent of reperfusion strategies in stroke using intravenous or intra-arterial thrombolysis (Meschia et al., 2002) we are now faced with the potential of

1002

41. ISCHEMIC WHITE MATTER DAMAGE

inducing reperfusion injury when ischemic regions have their blood Xow restored. Mitochondria appear particularly vulnerable to reperfusion injury as described previously. Therefore strategies aimed at limiting mitochondrial Ca overload, perhaps by timed inhibition of the Ca uniporter, or the mitochondrial permeability transition or introduction of free radical scavengers, may reduce injury during the reperfusion phase. This has not been tested in white matter injury models.

CONCLUSION Given the key role that white matter tracts play in the overall operation of the CNS, it is essential that any protective or restorative strategies consider this tissue on a footing equal to its gray matter counterpart. Over the past decade, the many neuroprotective trials in stroke and traumatic brain injury were designed mainly if not exclusively with the goal of rescuing the neuron and the synapse, exhibiting a naı¨ve inattention to the importance of axonal connections. We would argue that one of the main reasons for the failure of so many (indeed all!) acute neuroprotective trials (DeGraba and Pettigrew, 2000; Lees, 2000; Maas, 2001) is the fact that white matter pathophysiology was neglected in the design of these strategies. Over the past 10 years we have learned a great deal about how this tissue is damaged by ischemia, trauma, and immune attack, which positions the research community to now rationally devise the ‘‘next phase’’ of neuroprotective studies on the CNS, tailored to the protection of both gray and white matter regions.

Acknowledgments Work in the laboratory of PKS supported in part by grants from the Heart and Stroke Foundation of Ontario, National Institute of Neurological Disorders and Stroke, Canadian Institutes of Health Research, Ontario Neurotrauma Foundation, Canadian Stroke Network, Premier’s Research Excellence Award from the Province of Ontario, Canadian MS Society, and the generosity of private donors. Work in the laboratory of SGW supported in part by grants from the Rehabilitation Research Service and Medical Research Service, VA, and the National Multiple Sclerosis Society, and by gifts from the PVA and EPVA.

References Agrawal S. K., Fehlings M. G. (1996). Mechanisms of secondary injury to spinal cord axons in vitro–role of Naþ, Naþ-Kþ-ATPase, the Naþ-Hþ exchanger, and the Naþ-Ca2þ exchanger. J. Neurosci. 16:545–552. Agrawal S. K., Fehlings M. G. (1997a). The eVect of the sodium channel blocker QX-314 on recovery after acute spinal cord injury. J Neurotrauma 14:81–88. Agrawal S. K., Fehlings M. G. (1997b). Role of NMDA and non-NMDA ionotropic glutamate receptors in traumatic spinal cord axonal injury. J. Neurosci. 17:1055–1063. Agrawal S. K., Theriault E., Fehlings M. G. (1998). Role of group I metabotropic glutamate receptors in traumatic spinal cord white matter injury. J Neurotrauma 15:929–941. Agrawal S. K., Nashmi R., Fehlings M. G. (2000). Role of L- and N-type calcium channels in the pathophysiology of traumatic spinal cord white matter injury. Neuroscience 99:179–188. Akins P. T., Atkinson R. P. (2002). Glutamate AMPA receptor antagonist treatment for ischaemic stroke. Curr Med Res Opin 18 Suppl 2:s9–13. Arriza J. L., Fairman W. A., Wadiche J. I., Murdoch G. H., Kavanaugh M. P., Amara S. G. (1994). Functional comparisons of three glutamate transporter subtypes cloned from human motor cortex. J. Neurosci. 14:5559–5569. Attwell D., Barbour B., Szatkowski M. (1993). Nonvesicular release of neurotransmitter. Neuron 11:401–407. Bechtold D. A., Kapoor R., Smith K. J. (2002). Axonal protection in experimental inXammatory demyelinating disease achieved by Xecainide therapy. Soc. Neurosci. Abstr. 93.7. Bi R., Bi X., Baudry M. (1998a). Phosphorylation regulates calpain-mediated truncation of glutamate ionotropic receptors. Brain Res 797:154–158. Bi X., Chen J., Baudry M. (1998b). Calpain-mediated proteolysis of GluR1 subunits in organotypic hippocampal cultures following kainic acid treatment. Brain Res 781:355–357.

CONCLUSION

Bi X., Rong Y., Chen J., Dang S., Wang Z., Baudry M. (1998c). Calpain-mediated regulation of NMDA receptor structure and function. Brain Res 790:245–253. Bjartmar C., Trapp B. D. (2001). Axonal and neuronal degeneration in multiple sclerosis: Mechanisms and functional consequences. Curr. Opin. Neurol. 14:271–278. Black J. A., Dib-Hajj S., Baker D., Newcombe J., Cuzner M. L., Waxman S. G. (2000). Sensory neuron -speciWc sodium channel SNS is abnormally expressed in the brains of mice with experimental allergic encephalomyelitis and humans with multiple sclerosis. Proc. Natl. Acad. Sci. U.S.A. 97:11598–11602. Blaustein M. P., Lederer W. J. (1999). Sodium/calcium exchange: Its physiological implications. Physiol. Rev. 79:763–854. Bo L., Dawson T. M., Wesselingh S., Mork S., Choi S., Kong P. A., Hanley D., Trapp B. D. (1994). Induction of nitric oxide synthase in demyelinating regions of multiple sclerosis brains. Ann Neurol 36:778–786. Bolanos J. P., Almeida A., Stewart V., Peuchen S., Land J. M., Clark J. B., Heales S. J. (1997). Nitric oxide mediated mitochondrial damage in the brain: Mechanisms and implications for neurodegenerative diseases. J Neurochem 68:2227–2240. Brown A. M., Fern R., Jarvinen J. P., Kaila K., Ransom B. R. (1998). Changes in [Ca2þ]o during anoxia in CNS white matter. Neuroreport 9:1997–2000. Brown A. M., Wender R., Ransom B. R. (2001a). Ionic mechanisms of aglycemic axon injury in mammalian central white matter. J. Cereb. Blood Flow Metab. 21:385–395. Brown A. M., Westenbroek R. E., Catterall W. A., Ransom B. R. (2001b). Axonal L-type Ca(2þ) channels and anoxic injury in rat CNS white matter. J. Neurophysiol. 85:900–911. Buki A., Povlishock J. T., Siman R., Christman C. W. (1998). the role of calpain-mediated spectrin proteolysis (CMSP) in traumatically induced axonal injury (AI). Soc. Neurosci. Abstr. 24:252. Buki A., Siman R., Trojanowski J. Q., Povlishock J. T. (1999). The role of calpain-mediated spectrin proteolysis in traumatically induced axonal injury. J. Neuropathol. Exp. Neurol. 58:365–375. Caldwell J. H., Schaller K. L., Lasher R. S., Peles E., Levinson S. R. (2000). Sodium channel Nav1.6 is localized at nodes of Ranvier, dendrites, and synapses. Proc. Natl. Acad. Sci. U.S.A. Callahan D. J., Engle M. J., Volpe J. J. (1990). Hypoxic injury to developing glial cells: Protective eVect of high glucose. Pediatr Res 27:186–190. Casha S., Yu W. R., Fehlings M. G. (2001). Oligodendroglial apoptosis occurs along degenerating axons and is associated with FAS and p75 expression following spinal cord injury in the rat. Neuroscience 103:203–218. Catterall W. A. (1980). Neurotoxins that act on voltage-sensitive sodium channels in excitable membranes. Annu. Rev. Pharmacol. Toxicol. 20:15–43. Chen C. J., Liao S. L., Kuo J. S. (2000). Gliotoxic action of glutamate on cultured astrocytes. J Neurochem 75:1557–1565. Chih C. P., Lipton P., Roberts E. L., Jr. (2001). Do active cerebral neurons really use lactate rather than glucose? Trends Neurosci. 24:573–578. Clarke D. D., SokoloV L. (1994). Circulation and energy metabolism of the brain. In ‘‘Basic Neurochemistry’’ (G. J. Siegel, B. W. AgranoV, R. W., Albers, and P. B. MolinoV, eds.), 5th ed., pp. 645–680. Raven Press, New York. Collins J. G. (1990). Types of injuries by selected characteristics. Vital Health Stat.:1–68. Cooper C. E. (2002). Nitric oxide and cytochrome oxidase: Substrate, inhibitor or eVector? Trends Biochem. Sci. 27:33–39. Covault J., Liu Q. Y., el-Deeb S. (1991). Calcium-activated proteolysis of intracellular domains in the cell adhesion molecules NCAM and N-cadherin. Brain Res Mol Brain Res 11:11–16. Cross A. H., Manning P. T., Keeling R. M., Schmidt R. E., Misko T. P. (1998). Peroxynitrite formation within the central nervous system in active multiple sclerosis. J Neuroimmunol 88:45–56. Darley U. V., Stone D., Smith D., Martin J. F. (1991). Mitochondria, oxygen and reperfusion damage. Ann. Med. 23:583–588. David J. C., Yamada K. A., Bagwe M. R., Goldberg M. P. (1996). AMPA receptor activation is rapidly toxic to cortical astrocytes when desensitization is blocked. J. Neurosci. 16:200–209. DeGraba T. J., Pettigrew L. C. (2000). Why do neuroprotective drugs work in animals but not humans? Neurol Clin 18:475–493. Deshpande R. V., Goust J. M., Hogan E. L., Banik N. L. (1995). Calpain secreted by activated human lymphoid cells degrades myelin. J. Neurosci. Res. 42:259–265. Devlin T. M. (Ed.). (1992). ‘‘Textbook of Biochemistry,’’ 3rd ed. Wiley-Liss, New York. Dohmen C., Kumura E., Rosner G., Heiss W. D., Graf R. (2001). Adenosine in relation to calcium homeostasis: Comparison between gray and white matter ischemia. J Cereb Blood Flow Metab 21:503–510. Dong H., Dunn J., Lytton J. (2002). Stoichiometry of the Cardiac Naþ/Ca2þ exchanger NCX1.1 measured in transfected HEK cells. Biophys J 82:1943–1952. Dong H., Light P. E., French R. J., Lytton J. (2001). Electrophysiological characterization and ionic stoichiometry of the rat brain K(þ)-dependent NA(þ)/CA(2þ) exchanger, NCKX2. J Biol Chem 276:25919–25928. Eng D. L., Gordon T. R., Kocsis J. D., Waxman S. G. (1990). Current-clamp analysis of a time-dependent rectiWcation in rat optic nerve. J. Physiol. (Lond.) 421:185–202. Erecinska M., Dagani F. (1990). Relationships between the neuronal sodium/potassium pump and energy metabolism. EVects of Kþ, Naþ, and adenosine triphosphate in isolated brain synaptosomes. J. Gen. Physiol. 95:591–616. Erecinska M., Silver I. A. (1989). ATP and brain function. J. Cereb. Blood Flow Metab. 9:2–19.

1003

1004

41. ISCHEMIC WHITE MATTER DAMAGE

Ferguson B., Matyszak M. K., Esiri M. M., Perry V. H. (1997). Axonal damage in acute multiple sclerosis lesions. Brain 120:393–399. Fern R., Davis P., Waxman S. G., Ransom B. R. (1998). Axon conduction and survival in CNS white matter during energy deprivation: A developmental study. J. Neurophysiol. 79:95–105. Fern R., Moller T. (2000). Rapid ischemic cell death in immature oligodendrocytes: A fatal glutamate release feedback loop. J Neurosci 20:34–42. Fern R., Ransom B. R., Waxman S. G. (1995a). Voltage-gated calcium channels in CNS white matter: Role in anoxic injury. J. Neurophysiol. 74:369–377. Fern R., Ransom B. R., Waxman S. G. (1996a). Autoprotective mechanisms in the CNS: Some new lessons from white matter. Mol. Chem. Neuropathol. 27:107–129. Fern R., Ransom B. R., Waxman S. G. (1996b). White matter stroke: Autoprotective mechanisms with therapeutic implications. Cerebrovasc. Dis. 6:59–65. Fern R., Ransom B., Stys P. K., Waxman S. G. (1993). Pharmacological protection of CNS white matter during anoxia: Actions of phenytoin, carbamazepine and diazepam. J. Pharmacol. Exp. Ther. 266:1549–1555. Fern R., Waxman S. G., Ransom B. R. (1994). Modulation of anoxic injury in CNS white matter by adenosine and interaction between adenosine and GABA. J. Neurophysiol. 72:2609–2616. Fern R., Waxman S. G., Ransom B. R. (1995b). Endogenous GABA attenuates CNS white matter dysfunction following anoxia. J. Neurosci. 15:699–708. Fisher C. M. (1982). Lacunar strokes and infarcts: A review. Neurology 32:871–876. Follett P. L., Rosenberg P. A., Volpe J. J., Jensen F. E. (2000). NBQX attenuates excitotoxic injury in developing white matter. J. Neurosci. 20:9235–9241. Fonnum F. (1984). Glutamate: A neurotransmitter in mammalian brain. J. Neurochem. 42:1–11. Garcia-Barcina J. M., Matute C. (1996). Expression of kainate-selective glutamate receptor subunits in glial cells of the adult bovine white matter. Eur. J. Neurosci. 8:2379–2387. Garcia-Barcina J. M., Matute C. (1998). AMPA-selective glutamate receptor subunits in glial cells of the adult bovine white matter. Brain Res. Mol. Brain Res. 53:270–276. Garthwaite G., Goodwin D. A., Batchelor A. M., Leeming K., Garthwaite J. (2002). Nitric oxide toxicity in CNS white matter: An in vitro study using rat optic nerve. Neuroscience 109:145–155. Gellerman D. M., Bi X., Baudry M. (1997). NMDA receptor-mediated regulation of AMPA receptor properties in organotypic hippocampal slice cultures. J Neurochem 69:131–136. Gibson C. J. (1992). Overview of spinal cord injury. Phys. Med. Rehab. Clin. North Am. 3:699–709. Graf R., Dohmen C., Kumura E., Rosner G., Heiss W.-D. (1998). Extracellular shifts of GABA, adenosine and ion activities in cortical gray and subcortical white matter during global ischemia in cats. Soc. Neurosci. Abstr. 24:983. GriYths R., Dunlop J., Gorman A., Senior J., Grieve A. (1994). L-trans-pyrrolidine-2,4-dicarboxylate and cis-1aminocyclobutane-1,3-dicarboxylate behave as transportable, competitive inhibitors of the high-aYnity glutamate transporters. Biochem. Pharmacol. 47:267–274. Hell J. W., Westenbroek R. E., Breeze L. J., Wang K. K., Chavkin C., Catterall W. A. (1996). N-methyl-Daspartate receptor-induced proteolytic conversion of postsynaptic class C L-type calcium channels in hippocampal neurons. Proc Natl Acad Sci U S A 93:3362–3367. Hewitt K. E., Stys P. K., Lesiuk H. J. (2001). The use-dependent sodium channel blocker mexiletine is neuroprotective against global ischemic injury. Brain Res. 898:281–287. Imaizumi T., Kocsis J. D., Waxman S. G. (1997). Anoxic injury in the rat spinal cord-pharmacological evidence for multiple steps in Ca2þ-dependent injury of the dorsal columns. J. Neurotrauma 14:299–311. Imaizumi T., Kocsis J. D., Waxman S. G. (1999). The role of voltage-gated Ca2þ channels in anoxic injury of spinal cord white matter. Brain Res. 817:84–92. Jalc P., Marsala J., Jalcova H. (1995). Postischemic reperfusion causes a massive calcium overload in the myelinated spinal cord Wbers. Mol. Chem. Neuropathol. 25:143–153. Jensen A. M., Chiu S. Y. (1993). Expression of glutamate receptor genes in white matter: Developing and adult rat optic nerve. J. Neurosci. 13:1664–1675. Jiang Q., Stys P. K. (2000). Calpain inhibitors confer biochemical, but not electrophysiological, protection against anoxia in rat optic nerves. J. Neurochem. 74:2101–2107. KampX A., Posmantur R. M., Zhao X., Schmutzhard E., Clifton G. L., Hayes R. L. (1997). Mechanisms of calpain proteolysis following traumatic brain injury: Implications for pathology and therapy. A review and update. J. Neurotrauma 14:121–134. Kanellopoulos G. K., Xu X. M., Hsu C. Y., Lu X., Sundt T. M., Kouchoukos N. T. (2000). White matter injury in spinal cord ischemia: Protection by AMPA/kainate glutamate receptor antagonism. Stroke 31:1945–1952. Kapoor R., Davies M., Blaker P. A., et al (2003). Protection of axons from degeneration caused by nitric oxide can be achieved using blockers of the sodium channel and sodium-calcium exchanger. Ann. Neurol. in press. Khodorov B. I. (1991). Role of inactivation in local anesthetic action. Ann. N. Y. Acad. Sci. 625:224–248. Kiedrowski L., Czyz A., Li X. F., Lytton J. (2002). Preferential expression of plasmalemmal K-dependent Naþ/ Ca2þ exchangers in neurons versus astrocytes. Neuroreport 13:1529–1532. Koizumi J. (1974). Glycogen in the central nervous system. Prog. Histochem. Cytochem. 6:1–37. Kornek B., Lassmann H. (1999). Axonal pathology in multiple sclerosis. A historical note. Brain Pathol. 9:651–656.

CONCLUSION

Kornek B., Storch M. K., Bauer J., Djamshidian A., Weissert R., Wallstroem E., SteVerl A., Zimprich F., Olsson T., Linington C., Schmidbauer M., Lassmann H. (2001). Distribution of a calcium channel subunit in dystrophic axons in multiple sclerosis and experimental autoimmune encephalomyelitis. Brain 124:1114–1124. Kriegler S., Chiu S. Y. (1993). Calcium signaling of glial cells along mammalian axons. J. Neurosci. 13:4229–4245. Lees K. R. (2000). Neuroprotection. Br Med Bull 56:401–412. Leppanen L. L., Stys P. K. (1997). Ion transport and membrane potential in CNS myelinated axons. II: EVects of metabolic inhibition. J. Neurophysiol. 78:2095–2107. Levy L. M., Warr O., Attwell D. (1998). Stoichiometry of the glial glutamate transporter GLT-1 expressed inducibly in a chinese hamster ovary cell line selected for low endogenous Naþ-dependent glutamate uptake. J. Neurosci. 18:9620–9628. Li S., Jiang Q., Stys P. K. (2000). Important role of reverse Na(þ)-Ca(2þ) exchange in spinal cord white matter injury at physiological temperature. J. Neurophysiol. 84:1116–1119. Li S., Mealing G. A., Morley P., Stys P. K. (1999). Novel injury mechanism in anoxia and trauma of spinal cord white matter: Glutamate release via reverse Naþ-dependent glutamate transport. J. Neurosci. 19:RC16. Li S., Stys P. K. (2000). Mechanisms of ionotropic glutamate receptor-mediated excitotoxicity in isolated spinal cord white matter. J. Neurosci. 20:1190–1198. Lipton P. (1999). Ischemic cell death in brain neurons. Physiol. Rev. 79:1431–1568. Lo A. C., Black J. A., Waxman S. G. (2002). Neuroprotection of axons with phenytoin in experimental allergic encephalomyelitis. Neuroreport 13:1909–1912. LoPachin R. M., Gaughan C. L., Lehning E. J., Kaneko Y., Kelly T. M., Blight A. (1999). Experimental spinal cord injury: Spatiotemporal characterization of elemental concentrations and water contents in axons and neuroglia. J. Neurophysiol. 82:2143–2153. LoPachin R. M., Stys P. K. (1995). Elemental composition and water content of rat optic nerve myelinated axons and glial cells: EVects of in vitro anoxia and reoxygenation. J. Neurosci. 15:6735–6746. Maas A. I. (2001). Neuroprotective agents in traumatic brain injury. Expert Opin Investig Drugs 10:753–767. MacManus J. P., Buchan A. M. (2000). Apoptosis after experimental stroke: Fact or fashion? J Neurotrauma 17:899–914. Malek S., Coderre E., Stys P. K. (2002). K-Cl cotransporter mediates injurious Cl Xux in anoxic/ischemic CNS white matter. J. Neurosci.:in prep. Malek S., Stys P. K. (2002). EVects of Na and Cl Xuxes during anoxia on the compound action potential of rat optic nerve. Soc. Neurosci. Abstr. 95.9. Matsuo A., Akiguchi I., Lee G. C., McGeer E. G., McGeer P. L., Kimura J. (1998). Myelin degeneration in multiple system atrophy detected by unique antibodies. Am J Pathol 153:735–744. Matute C. (1998). Characteristics of acute and chronic kainate excitotoxic damage to the optic nerve. Proc. Natl. Acad. Sci. U.S.A. 95:10229–10234. Matute C., Alberdi E., Ibarretxe G., Sanchez-Gomez M. V. (2002). Excitotoxicity in glial cells. Eur J Pharmacol 447:239–246. Matute C., Sanchez-Gomez M. V., Martinez-Millan L., Miledi R. (1997). Glutamate receptor-mediated toxicity in optic nerve oligodendrocytes. Proc. Natl. Acad. Sci. U.S.A. 94:8830–8835. Maxwell W. L., Povlishock J. T., Graham D. L. (1997). A mechanistic analysis of nondisruptive axonal injury: A review. J. Neurotrauma 14:419–440. McCracken E., Dewar D., Hunter A. J. (2001). White matter damage following systemic injection of the mitochondrial inhibitor 3-nitropropionic acid in rat. Brain Res 892:329–335. McCracken E., Fowler J. H., Dewar D., Morrison S., McCulloch J. (2002). Grey matter and white matter ischemic damage is reduced by the competitive AMPA receptor antagonist, SPD 502. J Cereb Blood Flow Metab 22:1090–1097. McDonald J. W., Althomsons S. P., Hyrc K. L., Choi D. W., Goldberg M. P. (1998). Oligodendrocytes from forebrain are highly vulnerable to AMPA/kainate receptor-mediated excitotoxicity. Nat. Med. 4:291–297. Meschia J. F., Miller D. A., Brott T. G. (2002). Thrombolytic treatment of acute ischemic stroke. Mayo Clin Proc 77:542–551. Nicholls D. G., Budd S. L. (2000). Mitochondria and neuronal survival. Physiol. Rev. 80:315–360. Nikolaeva M. A., Stys P. K. (2002). Calcium changes in ischemic rat optic nerve axons studied by confocal microscopy. Soc. Neurosci. Abstr. 299.6. Oka A., Belliveau M. J., Rosenberg P. A., Volpe J. J. (1993). Vulnerability of oligodendroglia to glutamate: Pharmacology, mechanisms, and prevention. J. Neurosci. 13:1441–1453. Orrenius S., Nicotera P. (1996). Mechanisms of calcium-related cell death. In ‘‘Cellular and Molecular Mechanisms of Ischemic Brain Damage’’ (B. K. Siesjo and T. Wieloch, eds.), pp. 137–152. Raven Press, New York. Ouardouz, M., Nikolaeva, M., Coderre, E., Zamponi, G. W., McRory, J. E., Trapp, B. D., Yin, X., Wang, W., Woulfe, J., Stys, P. K. (2003). Depolarization-induced Ca2þrelease in ischemic spinal cord white matter involves L-type Ca2þ channel activation of ryanodine receptors. Neuron 40, 53–63. Pasti L., Zonta M., Pozzan T., Vicini S., Carmignoto G. (2001). Cytosolic calcium oscillations in astrocytes may regulate exocytotic release of glutamate. J. Neurosci. 21:477–484. Pitt D., Werner P., Raine C. S. (2000). Glutamate excitotoxicity in a model of multiple sclerosis. Nat. Med. 6:67–70. Povlishock J. T., Christman C. W. (1995). DiVuse axonal injury. In ‘‘The Axon: Structure, Function and Pathophysiology’’ (S. G. Waxman, J. D. Kocsis, and P. K. Stys, eds.), pp. 504–529. Oxford University Press, New York.

1005

1006

41. ISCHEMIC WHITE MATTER DAMAGE

Ragsdale D. S., McPhee J. C., Scheuer T., Catterall W. A. (1996). Common molecular determinants of local anesthetic, antiarrhythmic, and anticonvulsant block of voltage-gated Naþ channels. Proc Natl Acad Sci U S A 93:9270–9275. Rami A., Ferger D., Krieglstein J. (1997). Blockade of calpain proteolytic activity rescues neurons from glutamate excitotoxicity. Neurosci. Res. 27:93–97. Ransom B. R., Walz W., Davis P. K., Carlini W. G. (1992). Anoxia-induced changes in extracellular Kþ and pH in mammalian central white matter. J. Cereb. Blood Flow Metab. 12:593–602. Ren Y., Ridsdale A., Coderre E., Stys P. K. (2000). Calcium imaging in live rat optic nerve myelinated axons in vitro using confocal laser microscopy. J. Neurosci. Meth. 102:165–176. Rieckmann P., Maurer M. (2002). Anti-inXammatory strategies to prevent axonal injury in multiple sclerosis. Curr Opin Neurol 15:361–370. Rieckmann P., Smith K. J. (2001). Multiple sclerosis: More than inXammation and demyelination. Trends Neurosci 24:435–437. Rios E., Brum G. (1987). Involvement of dihydropyridine receptors in excitation-contraction coupling in skeletal muscle. Nature 325:717–720. Rose C. R., Waxman S. G., Ransom B. R. (1998). EVects of glucose deprivation, chemical hypoxia, and simulated ischemia on Naþ homeostasis in rat spinal cord astrocytes. J. Neurosci. 18:3554–3562. Rosenberg L. J., Teng Y. D., Wrathall J. R. (1999a). 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo(f)quinoxaline reduces glial loss and acute white matter pathology after experimental spinal cord contusion. J. Neurosci. 19:464–475. Rosenberg L. J., Teng Y. D., Wrathall J. R. (1999b). EVects of the sodium channel blocker tetrodotoxin on acute white matter pathology after experimental contusive spinal cord injury. J. Neurosci. 19:6122–6133. Rutledge E. M., Aschner M., Kimelberg H. K. (1998). Pharmacological characterization of swelling-induced D[3H]aspartate release from primary astrocyte cultures. Am. J. Physiol. 274:C1511–1520. Salamino F., Sparatore B., Melloni E., Michetti M., Viotti P. L., Pontremoli S., Carafoli E. (1994). The plasma membrane calcium pump is the preferred calpain substrate within the erythrocyte. Cell Calcium 15:28–35. Schaecher K. E., Shields D. C., Banik N. L. (2001). Mechanism of myelin breakdown in experimental demyelination: A putative role for calpain. Neurochem Res 26:731–737. Schanne F. A., Kane A. B., Young E. E., Farber J. L. (1979). Calcium-dependence of toxic cell death: A Wnal common pathway. Science 206:700–702. Schumacher P. A., Siman R. G., Fehlings M. G. (2000). Pretreatment with calpain inhibitor CEP-4143 inhibits calpain I activation and cytoskeletal degradation, improves neurological function, and enhances axonal survival after traumatic spinal cord injury. J. Neurochem. 74:1646–1655. Shields D. C., Schaecher K. E., Saido T. C., Banik N. L. (1999). A putative mechanism of demyelination in multiple sclerosis by a proteolytic enzyme, calpain. Proc. Natl. Acad. Sci. U.S.A. 96:11486–11491. Silver I. A., Deas J., Erecinska M. (1997). Ion homeostasis in brain cells: DiVerences in intracellular ion responses to energy limitation between cultured neurons and glial cells. Neuroscience 78:589–601. Smith K. J., Kapoor R., Hall S. M., Davies M. (2001). Electrically active axons degenerate when exposed to nitric oxide. Ann. Neurol. 49:470–476. Smith M. R., Smith R. D., Plummer N. W., Meisler M. H., Goldin A. L. (1998). Functional analysis of the mouse Scn8a sodium channel. J. Neurosci. 18:6093–6102. Smith T., Groom A., Zhu B., Turski L. (2000). Autoimmune encephalomyelitis ameliorated by AMPA antagonists. Nat Med 6:62–66. Sorg O., Magistretti P. J. (1991). Characterization of the glycogenolysis elicited by vasoactive intestinal peptide, noradrenaline and adenosine in primary cultures of mouse cerebral cortical astrocytes. Brain Res 563:227–233. Sosin D. M., Sniezek J. E., Thurman D. J. (1996). Incidence of mild and moderate brain injury in the United States, 1991. Brain Inj. 10:47–54. Steinhauser C., Gallo V. (1996). News on glutamate receptors in glial cells. Trends Neurosci 19:339–345. Stryer L. (1988). ‘‘Biochemistry.’’ W. H. Freeman, New York. Stys P. K. (1995). Protective eVects of antiarrhythmic agents against anoxic injury in CNS white matter. J. Cereb. Blood Flow Metab. 15:425–432. Stys P. K. (1998). Anoxic and ischemic injury of myelinated axons in CNS white matter: From mechanistic concepts to therapeutics. J. Cereb. Blood Flow Metab. 18:2–25. Stys P. K., Hubatsch D. A., Leppanen L. L. (1998). EVects of Kþ channel blockers on the anoxic response of central myelinated axons. NeuroReport 9:447–453. Stys P. K., Jiang Q. (2002). Ca2þ-mediated calpain-dependent neuroWlament breakdown in anoxic and ischemic central white matter axons. Neurosci. Lett. 328:150–154. Stys P. K., Lehning E. J., Sauberman A. J., LoPachin R. M. (1997). Intracellular concentrations of major ions in rat myelinated axons and glia: Calculations based on electron probe X-ray microanalyses. J. Neurochem. 68:1920–1928. Stys P. K., Lesiuk H. (1996). Correlation between electrophysiological eVects of mexiletine and ischemic Stys P. K., Li S. (2000). Glutamate-induced white matter injury: Excitotoxicity without synapses. The Neuroscientist 6:230–233. Stys P. K., LoPachin R. M. (1996). Elemental composition and water content of rat optic nerve myelinated axons during in vitro post-anoxia reoxygenation. Neuroscience 73:1081–1090. protection of CNS white matter. Neuroscience 71:27–36.

CONCLUSION

Stys P. K., LoPachin R. M. (1998). Mechanisms of ion Xux in anoxic myelinated CNS axons. Neuroscience 82:21–32. Stys P. K., Ouardouz M. (2002). Role of glutamate receptors in spinal cord dorsal column ischemia. Soc. Neurosci. Abstr.:298.298. Stys P. K., Ransom B. R., Waxman S. G. (1992b). Tertiary and quaternary local anesthetics protect CNS white matter from anoxic injury at concentrations that do not block excitability. J. Neurophysiol. 67:236–240. Stys P. K., Ransom B. R., Waxman S. G., Davis P. K. (1990). Role of extracellular calcium in anoxic injury of mammalian central white matter. Proc. Natl. Acad. Sci. USA. 87:4212–4216. Stys P. K., Sontheimer H., Ransom B. R., Waxman S. G. (1993). Non-inactivating, TTX-sensitive Naþ conductance in rat optic nerve axons. Proc. Natl. Acad. Sci. USA. 90:6976–6980. Stys P. K., Waxman S. G., Ransom B. R. (1992a). Ionic mechanisms of anoxic injury in mammalian CNS white matter: Role of Naþ channels and Naþ -Ca2þ exchanger. J. Neurosci. 12:430–439. Swanson R. A., Yu A. C., Sharp F. R., Chan P. H. (1989). Regulation of glycogen content in primary astrocyte culture: EVects of glucose analogues, phenobarbital, and methionine sulfoximine. J. Neurochem. 52:1359–1365. Taylor C. P. (1993). Naþ currents that fail to inactivate. Trends Neurosci. 16:455–460. Taylor C. P., Weber M. L., Gaughan C. L., Lehning E. J., LoPachin R. M. (1999). Oxygen/glucose deprivation in hippocampal slices: Altered intraneuronal elemental composition predicts structural and functional damage. J. Neurosci. 19:619–629. Tekkok S. B., Goldberg M. P. (2001). AMPA/kainate receptor activation mediates hypoxic oligodendrocyte death and axonal injury in cerebral white matter. J. Neurosci. 21:4237–4248. Tekkok S. B., Hyrc K. L., Underhill S. M., Goldberg M. P. (2000). Naþ/Ca2þ exchange blocker KB-R7943 protects axons and oligodendrocytes during oxygen glucose deprivation. Soc. Neurosci. Abstr. 26:2065. Teng Y. D., Wrathall J. R. (1997). Local blockade of sodium channels by tetrodotoxin ameliorates tissue loss and long -term functional deWcits resulting from experimental spinal cord injury. J. Neurosci. 17:4359–4366. Thorell W. E., Leibrock L. G., Agrawal S. K. (2002). Role of RyRs and IP3 receptors after traumatic injury to spinal cord white matter. J. Neurotrauma 19:335–342. Trapp B. D., Peterson J., RansohoV R. M., Rudick R., Mork S., Bo L. (1998). Axonal transection in the lesions of multiple sclerosis. N. Engl. J. Med. 338:278–285. Tsacopoulos M., Magistretti P. J. (1996). Metabolic coupling between glia and neurons. J. Neurosci. 16:877–885. Underhill S. M., Goldberg M. P. (2002). Axons potentiate oligodendrocyte vulernability to excitotoxicity. Soc. Neurosci. Abstr. 299.7. Wang G. K., Brodwick M. S., Eaton D. C., Strichartz G. R. (1987). Inhibition of sodium currents by local anesthetics in chloramine-T-treated squid axons. The role of channel activation. J. Gen. Physiol. 89:645–667. Waxman S. G. (2000). Multiple sclerosis as a neuronal disease. Arch Neurol 57:22–24. Waxman S. G. (2001). Transcriptional channelopathies: an emerging class of disorders. Nat. Rev. Neurosci. 2:652–659. Waxman S. G., Black J. A., Ransom B. R., Stys P. K. (1993). Protection of the axonal cytoskeleton in anoxic optic nerve by decreased extracellular calcium. Brain Res. 614:137–145. Waxman S. G., Kocsis J. D., Stys P. K., (Eds.). (1995). ‘‘The Axon: Structure, Function and Pathophysiology.’’ Oxford University Press, New York. Weinshenker B. G. (1996). Epidemiology of multiple sclerosis. Neurol Clin 14:291–308. Wender R., Brown A. M., Fern R., Swanson R. A., Farrell K., Ransom B. R. (2000). Astrocytic glycogen inXuences axon function and survival during glucose deprivation in central white matter. J Neurosci 20:6804–6810. White B. C., Sullivan J. M., DeGracia D. J., O’Neil B. J., Neumar R. W., Grossman L. I., Rafols J. A., Krause G. S. (2000). Brain ischemia and reperfusion: Molecular mechanisms of neuronal injury. J Neurol Sci 179:1–33. Wolf J. A., Stys P. K., Lusardi T., Meaney D., Smith D. H. (2001). Traumatic axonal injury induces calcium inXux modulated by tetrodotoxin-sensitive sodium channels. J. Neurosci. 21:1923–1930. Wrathall J. R., Choiniere D., Teng Y. D. (1994). Dose-dependent reduction of tissue loss and functional impairment after spinal cord trauma with the AMPA/kainate antagonist NBQX. J. Neurosci. 14:6598–6607. Wrathall J. R., Teng Y. D., Marriott R. (1997). Delayed antagonism of AMPA/kainate receptors reduces longterm functional deWcits resulting from spinal cord trauma. Exp. Neurol. 145:565–573. Yam P. S., Dunn L. T., Graham D. I., Dewar D., McCulloch J. (2000). NMDA receptor blockade fails to alter axonal injury in focal cerebral ischemia. J Cereb Blood Flow Metab 20:772–779. Yeh J. Z., Tanguy J. (1985). Na channel activation gate modulates slow recovery from use-dependent block by local anesthetics in squid giant axons. Biophys. J. 47:685–694. Yoshioka A., Bacskai B., Pleasure D. (1996). Pathophysiology of oligodendroglial excitotoxicity. J. Neurosci. Res. 46:427–437. Yoshioka A., Hardy M., Younkin D. P., Grinspan J. B., Stern J. L., Pleasure D. (1995). Alpha-amino-3 -hydroxy5-methyl-4-isoxazolepropionate (AMPA) receptors mediate excitotoxicity in the oligodendroglial lineage. J. Neurochem. 64:2442–2448. Yu, A. C. H., Gregory G. A., Chan P. H. (1989). Hypoxia-induced dysfunctions and injury of astrocytes in primary cell cultures. J. Cereb. Blood Flow Metab. 9:20–28. Zerangue N., Kavanaugh M. P. (1996). Flux coupling in a neuronal glutamate transporter. Nature 383:634–637. Zhang K., Sejnowski T. J. (2000). A universal scaling law between gray matter and white matter of cerebral cortex. Proc Natl Acad Sci U S A 97:5621–5626.

1007

C H A P T E R

42 Protein Misfolding as a Disease Determinant Alexander Gow

INTRODUCTION ‘‘Conformational diseases’’ is the term used increasingly to refer to a large eclectic group of degenerative disorders with far greater similarities in pathophysiology than is apparent clinically. As this umbrella term suggests, conformational diseases arise at least in part from perturbations to three-dimensional structure in one or more cellular proteins stemming from genetic missense or nonsense mutations or from unclear etiology. At least some of these diseases can be loosely regarded as evolutionarily conserved in the sense that molecular mechanisms underlying pathogenesis in mammals are known to operate in and modify the behavior of yeast and fungi and, perhaps, organisms in many phyla (Prusiner, 1998; Uptain and Lindquist, 2002). In humans, the deposition of abnormally folded proteins have been found in most organs including CNS, kidney, liver, heart, bones, and joints. Curiously, deposits typically occur in CNS but not other organs or vice versa, but the etiology of this apparent arbitrary division is currently unknown (Pepys, 2001). The amyloidosis group of conformational diseases comprises systemic degenerative disorders that mainly involve the peripheral organs. As the name indicates, these diseases arise from the deposition of amyloid in tissue that is extracellular and is composed of any one of approximately 20 proteins having little in common except for a capacity to generate insoluble Wbrillar deposits with a stable b-sheet structure. This common secondary structure motif, which involves interactions between the amino acid main chain atoms, is a generic property of polypeptides that can be generated using puriWed proteins in vitro (reviewed by Dobson, 2001). Nevertheless, speciWc diseases with distinct clinical phenotypes result from the accumulation of particular proteins, which suggests that unique properties of the misfolded polypeptide chains drive the pathophysiology. An important aspect of pathogenesis in systemic amyloidoses is the dependence on abnormally high concentrations of major components of the amyloid Wbrils. For example, chronic infections and inXammation cause sustained overproduction of acute phase proteins such as serum amyloid A protein, which is the apolipoprotein component of high density lipoprotein. Processed forms of this protein are most commonly deposited in kidney leading to renal failure. Ab2M amyloidosis is a complication of long-term hemodialysis, for which the b2-microglobulin component of HLA class I molecules accumulates in the bones and joints. b2-microglobulin is normally exclusively removed from blood by the proximal tubules of the kidney; however, this clearance is ineYcient after kidney failure and cannot be eVected by hemodialysis, which leads to chronic elevation of b2microglobulin levels in plasma.

Myelin Biology and Disorders, Volume 2

1009

Copyright 2004, Elsevier (USA). All rights reserved.

1010

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

The major focus of this review is the group of conformational diseases that stem from aberrant protein folding in the central nervous system (CNS). The number of disorders included in this group has been growing rapidly and includes several well-known diseases—Alzheimer’s disease, Parkinson’s disease, prion protein diseases, Huntington’s disease and frontotemporal dementias—for which important Wndings have been reported in recent years.

PROTEIN INCLUSION DISEASES A major focus of a large number of studies appearing in the literature in recent years has been to identify cellular conditions leading to the generation of altered protein conformations as well as changes in protein processing and aggregation. Such pathology is commonplace in neurodegenerative diseases and stems from increased b-sheet secondary structure and aggregation of one or more of several proteins that lead to formation of the amyloid-like deposits and inclusion bodies observed in Alzheimer’s disease, Parkinson’s disease, advanced Down’s syndrome, and frontotemporal dementia and Parkinsonism linked to chromosome 17 (FTDP-17). The proteins involved in this pathology include amyloid precursor protein-derived peptides (e.g., Ab), a-synuclein and the microtubuleassociated protein, tau (Trojanowski, 2002). However, amyloid deposits and inclusions are also found in other diseases including prion diseases, Huntington’s disease, most spinocerebellar ataxias, and serpinopathies (reviewed in Cummings and Zoghbi, 2000; Lomas and Carrell, 2002). The degree to which amyloid deposits and inclusions contribute to neurodegenerative pathophysiology is currently unclear and, in some if not most cases, their presence may reXect secondary processes stemming from upstream events (Cummings and Zoghbi, 2000; Trojanowski, 2002). Despite the persistence of these protein aggregates in the parenchyma as prominent features or hallmarks of speciWc disorders, deposits and inclusions are increasingly thought to be nontoxic. Indeed, for several spinocerebellar ataxias inclusions are not detected in neuronal populations that are primarily aVected but rather are observed in nontarget neurons from brain regions exhibiting little to no other pathology or clinical involvement (Cummings and Zoghbi, 2000). On the other hand, proteins that comprise the protein aggregates are directly implicated in pathogenesis of several diseases. For example, amyloid Wbrils composed of the Ab peptide are prominent features of sporadic Alzheimer’s disease and mutations in the amyloid precursor protein lead to increased deposition of Ab amyloid in familial Alzheimer’s disease. In similar fashion, sporadic and familial forms of Parkinson’s disease involve a-synuclein in Lewy bodies while FTDP-17 involves tau in neuroWbrillary tangles and prion disease involves PrP protein in inclusions. Thus, the genes that are mutated in familial forms of these diseases encode proteins, which are major components of the inclusions observed in sporadic forms of the same diseases. A relatively simple explanation to accommodate apparently opposing views about the relevance of amyloid and inclusions in neurodegenerative disease is that the monomeric or oligomeric forms of the proteins are the toxic species and cells attempt to neutralize this toxicity by sequestering the protein into large aggregates. Indeed, recent in vivo data indicate as much for neurodegenerative changes associated with alternatively folded forms of the prion protein in scrapie and Creutzfeldt-Jakob disease (CJD) and b-amyloid protein in Alzheimer’s disease (Ma and Lindquist, 2002; Ma et al., 2002a; Walsh et al., 2002). Accordingly, several factors—the level of expression of amyloidogenic or inclusionforming protein, a cell’s capacity to form nontoxic deposits, and the sequence of events leading to altered protein conformation or oligomerization (genetic mutation or upstream pathologic processes)—are thought to inXuence the susceptibility of any given cell to alternatively folded forms of proteins that give rise to a particular disease. In the discussion that follows, we consider data for a number of degenerative diseases in the context of protein misfolding.

PROTEIN INCLUSION DISEASES

Prion Diseases Prion-related disorders in mammals are perhaps the archival class of conformational diseases and have been identiWed in various mammalian species including scrapie in sheep and goats, bovine spongioform encephalitis, chronic wasting disease in deer and elk, as well as Creutzfeldt-Jakob disease (CJD) and Kuru in humans (Prusiner, 1998). These diseases involve a cellular prion protein, PrPC, which is encoded by the evolutionarily conserved PRNP gene. The function of PrPC is poorly deWned and can adopt an abnormal higher-ordered conformation under certain conditions to generate the PrPSc isoform (Wechselberger et al., 2002). This alternative conformation exhibits unusual properties because it propagates by converting other PrPC molecules to the PrPSc form. An important aspect of pathogenesis, and ultimately therapy, is that continued expression of PRNP is necessary for development of disease, which has been demonstrated by the complete resistance of Prnp-null mice to high titer infection with infective prion particles (Bueler et al., 1993). Thus, prions are infectious particles and, when transferred between organisms, elicit degenerative disease through subversion of normal protein folding. Details of the mechanism underlying conversion of PrPC to PrPSc or PrPSc–like structural isoforms have come to light and can be divided into two major stages (Hegde et al., 1998, 1999; Ma and Lindquist, 2002; Ma et al., 2002a). The Wrst stage involves a byproduct of PrPC synthesis and folding in the endoplasmic reticulum (ER) prior to traYcking through the secretory pathway to the cell surface. A signiWcant proportion of nascent PrP polypeptide chains either fail to adopt appropriate higher-ordered structures or adopt alternate transmembrane topologies (CtmPrP), and these unstable intermediates are retrotranslocated from the ER to the cytosol for rapid degradation by the ubiquitin-proteasome complex. Missense mutations in PrP identiWed from CJD patients generate these intermediates at higher frequencies than those observed for either wild-type PrP or for amino acid changes that are not associated with CJD, resulting in retrotranslocation of a greater proportion of nascent PrP polypeptides. The second stage of PrP conversion is thought to occur stochastically at low frequency and involves adoption of PrPSc-like conformations by retrotranslocated polypeptides in the cytosol. Such structural changes serve as molecular templates for the structural conversion of additional PrPC molecules, which homopolymerize and are extremely resistant to degradation by the cell. Furthermore, polymer formation is dependent on monomer concentration in the cytosol; thus, the higher likelihood for misfolding of familial CJD variant PrP accounts for the greater eYciency of PrPSc conversion and self-association. In this light, the mechanism through which prions contribute to pathogenesis appears to involve a toxic gain-of-function of misfolded PrP rather than a loss of function for the cell. In strong support of this notion, Prnp-null mice exhibit no observable neurologic phenotype (Bueler et al., 1993).

Serpinopathies One of the clearest examples of diseases that stem from a toxic gain-of-function associated with altered protein conformation involves mutations in a1-ANTITRYPSIN (a1-AT) and related genes that encode serine protease inhibitors (reviewed in Carrell and Lomas, 2002; Lomas and Carrell, 2002). a1-AT is secreted into the bloodstream by the liver and functions primarily to protect connective tissue by irreversibly inhibiting elastase that is secreted by leukocytes in the lungs. Patients harboring mutations in the coding region of a1-AT exhibit reduced serpin activity in the blood and are at risk for developing emphysema in adulthood. Patients that are homozygous for mutations causing the most severe form of disease, so-called PiZZ alleles, show virtually no serpin activity or protein in the blood. Rather, nascent polypeptides homopolymerize to form linear chains that are visible using electron microscopy and accumulate in the ER of hepatocytes. Approximately 10 to 15% of PiZZ patients develop neonatal cirrhosis. Because a1-AT has no known function in liver, this additional hepatic phenotype likely represents an acquired toxic property of the mutant protein related to its accumulation in the ER.

1011

1012

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

Major factors contributing to liver damage involve those that cause inXammatory responses that raise body temperature and induce the expression of acute phase genes such as a1-AT (Crowther, 2002). Furthermore, those patients that develop severe liver pathology exhibit other abnormalities in skin Wbroblasts. SpeciWcally, the Wbroblasts exhibit reduced capacity to degrade misfolded polypeptides through the ubiquitin-proteasome pathway following retrotranslocated from the ER (Wu et al., 1994). Currently, the relevance to disease severity of these lower-than-normal rates of ER-associated protein degradation (ERAD) is unknown. However, it is conceivable that hepatocytes in most PiZZ patients are resistant to damage because the ERAD pathway is suYciently active to minimize the accumulation of polymerized mutant a1-AT that would otherwise lead to cell dysfunction or death. More recently, one form of autosomal-dominant familial encephalopathy with inclusion bodies (FEN1B) has been linked to mutations in the PI12 gene, which encodes the neurally restricted neuroserpin protein. A hallmark of this neurodegenerative disease is the presence of eosinophilic Collins bodies in neurons of the substantia nigra and deeper layers of cerebral cortex, which bear striking resemblance to the periodic-acid-SchiV positive inclusions in hepatocytes of PiZZ patients. Indeed, in one pedigree with a particularly severe form of this dementia, life span is reduced to 2 decades and inclusions develop in essentially all neurons (Davis et al., 1999a, 1999b). Currently, the aggregation state of the mutant serpins that mediate the toxic eVects in neurons and hepatocytes—monomers, small aggregates, or large polymers—has not been determined; however, these mutant proteins adopt stable higher-ordered structures within the ER and, at least for a1-AT, do not induce expression of the molecular chaperone, BiP (Graham et al., 1990). Nevertheless, a substantial proportion of these serpin polymers are stably associated with the ER-localized calcium binding-protein, calnexin (Le et al., 1994).

Cytoplasmic Inclusion Diseases A conformational disease for which defective ERAD in neurons of the substantia nigra plays a critical role in pathogenesis is Parkinson’s disease (reviewed in Crowther, 2002; Kruger et al., 2002). To date, mutations in two genes, PARKIN and a-SYNUCLEIN, have been clearly linked to familial disease. PARKIN encodes an E3 ubiquitin ligase, which normally polyubiquitinates proteins that are destined for proteasomal degradation (Shimura et al., 2000). Furthermore, considerable evidence links another locus containing the UCH-L1 gene to familial Parkinson’s disease, the gene product of which encodes a ubiquitin hydrolase that serves to recycle ubiquitin monomers during ubiquitinated substrate degradation by the proteasome. The fact that an O-glycosylated form of a-synuclein is a major component of cytoplasmic inclusions, called Lewy bodies, in nigral neurons from most Parkinson’s patients provides a compelling link between the rare familial forms and sporadic forms of disease. In addition, a-synuclein that is present in Lewy bodies is typically ubiquitinated (by the PARKIN gene product), which suggests that perturbed or defective proteasome function is a major mechanism underlying Lewy body formation and pathogenesis in most patients. However, Lewy bodies are rarely observed in patients harboring PARKIN mutations, which indicates that the pathophysiology of Parkinson’s disease is independent of these inclusions; thus, Lewy bodies probably are not toxic to neurons and may be secondary aspects of the disease phenotype. Rather, oligomers of at least some of the components of inclusions are currently thought to be the active species for cell dysfunction or death. Parkinson’s disease is a member of a diverse group of cytoplasmic inclusion disorders known collectively as tauopathies (reviewed in Lee et al., 2001). Other well-known members include Alzheimer’s disease, amyotrophic lateral sclerosis, Creutzfeldt-Jakob disease, Down’s syndrome, FTDP-17, and Pick’s disease. The hallmark pathological features of tauopathies are the appearance of Wlaments in the parenchyma, called neuropil threads, and Wlamentous inclusions composed of paired-helical and straight Wlaments. These protein aggregates are labeled by antibodies raised against phosphorylated tau protein. Other pathology may also accompany the tau-positive Wlament tangles, including: b-amyloid deposits observed in Alzheimer’s disease and Lewy bodies found in Parkinson’s disease.

PROTEIN INCLUSION DISEASES

Clear evidence that tau plays an important role in the pathophysiology of tauopathies comes from mapping of mutations in many FTDP-17 kindreds to the TAU gene. These mutations appear to alter splicing of TAU primary transcripts or induce hyperphosphorylation of tau and reduce its aYnity for microtubules, which leads to tau aggregation and neuroWbrillary tangle formation. Furthermore, several sporadic disorders known as frontotemporal dementias cause clinical symptoms that are similar to FTDP-17 and appear to perturb TAU gene expression in several ways, including modifying the splicing of primary transcripts, increasing gene expression, or increasing tau phosphorylation. On the other hand, pathogenesis in sporadic Alzheimer disease is widely thought to involve an amyloid cascade with Ab peptide as the building block and tau pathology as a secondary event (Hardy and Allsop, 1991).

Expanded Trinucleotide-Repeat Diseases The spinocerebellar ataxia (SCA) group of neurological diseases are autosomal-dominant neurodegenerative disorders characterized by progressive motor symptoms, with a particularly severe impact on hypoglossal nuclei, in the absence of cognitive impairment (Orr and Zoghbi, 2001). Trinucleotide expansions in some of these diseases occur in the untranslated regions of unrelated genes and appear to cause disease through disparate mechanisms. On the other hand, expanded repeats also occur in coding regions to form large polyglutamine tracts (encoded by the trinucleotide, CAG) beyond a threshold size of 20 to 40 residues. Additional diseases known to be caused by trinucleotide repeat expansions in protein coding regions include dentatorubropallidoluysian atrophy, Huntington and Kennedy diseases (reviewed in Cummings and Zoghbi, 2000). Studies in transgenic mice have revealed a number of important principles governing pathogenesis in the trinucleotide expansion diseases, some of which are known to be relevant to other conformational diseases. Working with pathogenic forms of SCA1 containing arrays of 77 to 82 CAG repeats (corresponding to a glutamine tract, 77 to 82Q), which encode ataxin-1 and aggregate to form nuclear inclusions, Orr and colleagues Wrst determined that nuclear localization of this protein is required for aggregation and subsequent neurodegeneration (Klement et al., 1998). Second, ataxin-1 self-association/ aggregation and inclusion body formation are not required for pathogenesis, which suggests that nuclear-localized, monomeric ataxin-1 is the primary toxic species once the glutamine domain is suYciently large (Klement et al., 1998). Third, the ubiquitinproteasome pathway directly modulates disease severity by serving to ubiquitinate pathogenic forms of ataxin-1 and promote incorporation into inclusion bodies. Indeed, this pathway normally serves to ameliorate disease as demonstrated by an exacerbation of the phenotype of ataxin-1(82Q) transgenic mice that harbor two null alleles of the E3 ubiquitin ligase gene, Ube3a (Cummings et al., 1999). Finally, the presence of small chaperones in ataxin-1 inclusion bodies, and the mitigating eVects of overexpressing inducible hsp70 in ataxin-1(82Q) transgenic mice, indicate that inappropriate or incomplete protein folding may be an important mediator of pathophysiology (Cummings et al., 1998, 2001). Notably, overexpression of hsp70 does not inXuence formation of inclusions but may reduce the toxicity of ataxin-1 monomers by sequestration to form nontoxic complexes. An important aspect of pathology in trinucleotide repeat diseases is reXected by the observation that the polyglutamine tract may be suYcient for pathogenesis. For example, widespread expression of a polyglutamine tract under transcriptional control of the human cytomegalovirus immediate-early promoter in transgenic mice conferred progressive behavioral abnormalities but only localized neuronal loss (Reddy et al., 1998). Furthermore, addition of 146 CAG repeats into the coding region of the Hprt gene, which encodes a ubiquitously expressed cytoplasmic protein that does not normally contain a glutaminerich domain, causes Huntington’s disease–like symptoms and premature death of mice (Ordway et al., 1997). However, the absence of cell death in these animals indicates that the neuronal loss observed in Huntington’s disease may arise from a loss of function of the huntingtin protein. Thus, these data indicate that although polyglutamine tracts are likely to be intrinsically toxic, disease-speciWc phenotypes, and the loss of particular neuronal

1013

1014

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

subtypes are likely to stem from loss- or gain-of-function associated with the mutant protein harboring the polyglutamine domain.

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY In contrast to the diseases discussed previously, a number of familial degenerative disorders are not associated with protein deposits or inclusion bodies despite the fact that changes in protein conformation are known to arise from the associated genetic lesions. Presumably, either these proteins are toxic at far lower concentrations than proteins causing inclusion diseases or the primary structures of these proteins do not comprise amino acid domains that are amyloidogenic if improperly folded. One of the best characterized of these noninclusion diseases is the leukodystrophy, Pelizaeus-Merzbacher disease (PMD). Clinical symptoms associated with mutations causing PMD are apparent at birth or within the Wrst year of life, which is consistent with the notion that the mutant gene products are extremely toxic to cells and cause apoptosis long before they accumulate in amounts necessary to form inclusion bodies. In comparison, ages of onset of even juvenile forms of inclusion body diseases are rarely less than the Wrst decade of life. A major focus of studies aimed at characterizing the molecular mechanisms underlying pathogenesis of PMD has been to identify signaling pathways downstream of the synthesis of the mutant proteins that lead to cell dysfunction or apoptosis. These pathways are discussed in detail in this chapter because of their growing relevance to the pathogenesis of inclusion body diseases.

The PLP1 gene and Pelizaeus-Merzbacher disease The leukodystrophies are a diverse group of rare central nervous system (CNS) disorders that arise from a variety of neurodegenerative processes in the white matter. Genetic lesions and biochemical defects underlying pathogenesis of several of these diseases have been described over the past 15 years (Berger et al., 2001; SchiVmann and BoespXugTanguy, 2001). Pelizaeus-Merzbacher disease (PMD) is one example for which mutations were identiWed in the proteolipid protein 1 gene (PLP1) in the late 1980s (Gencic et al., 1989; Hudson et al., 1989) but for which the molecular mechanisms are only now coming to light. Thus, a novel mechanism involving a recently elucidated signaling cascade that is activated by the accumulation of mutant proteins in the secretory pathway appears to play a direct role in pathogenesis. This mechanism could account for 20 to 25% of PMD cases and may be causative for many degenerative diseases involving protein misfolding. PMD is an X-chromosome linked hypomyelinating disease caused by mutations in the PLP1 gene, which encodes a major CNS myelin protein. The PLP1 gene is comprised of 7 exons, all of which contribute to the open reading frame, and exon 3 contains a central cryptic splice-site that is utilized to generate two protein isoforms, PLP1 and DM-20. Thus, the smaller DM-20 protein lacks a 35 amino acid polypeptide encoded by exon 3B, which, in PLP1, contributes to a hydrophilic domain in the middle of the protein. This domain is exposed to the cytoplasm and is known as the PLP1-speciWc peptide. Absence of the PLP1speciWc peptide has no eVect on overall membrane topology, and both protein isoforms span the bilayer four times with the amino- and carboxyl-termini exposed to the cytoplasm (Gow et al., 1997). However, topography of the second extracellular loop of DM-20 may diVer from that of PLP1, which may serve to regulate myelin membrane compaction and lamellar spacing in a manner slightly diVerent to that of PLP1 (Stecca et al., 2000).

Three Genetic Lesions Confer Pelizaeus-Merzbacher Disease The clinical spectrum of PMD is very broad and ranges from severe forms associated with a phenotype at birth, quadriparesis and dramatically reduced life span, to milder forms

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

associated only with spastic paraparesis. Three types of genetic lesions associated with PMD account for 80 to 90% of diagnoses and are characterized by gene deletions, gene duplications and coding region mutations (Garbern et al., 1999). PLP1 Gene Null Alleles PLP1-null mutations have been identiWed in several patients, including a complete deletion of the gene and an in-frame stop codon after the Wrst amino acid (Garbern et al., 1997), and all exhibit a similar mild clinical phenotype characterized by moderate spastic quadriparesis, mild cognitive delay, ataxia, and a demyelinating peripheral neuropathy (Garbern et al., 1997; Nave and BoespXug-Tanguy, 1996; Raskind et al., 1991; Sistermans et al., 1996). In contrast to this Schwann cell pathology, demyelination in the CNS is virtually absent although neurons undergo progressive length-dependent axonal degeneration principally in cervical and thoracic regions of the ascending and descending corticospinal tracts (Garbern et al., 2002). These data demonstrate the importance of PLP1 gene products for long term maintenance of oligodendrocyte-axon interactions even in the PNS where this gene is expressed at very low levels by Schwann cells. Supernumerary Copies of the PLP1 gene The second and most common cause of PMD is the duplication within an approximately 800 kb region of the X-chromosome Xanking the PLP1 gene. Studies in animal models show that overexpression of PLP1 and/or DM-20 is suYcient to cause CNS hypomyelination in transgenic mice (Gow et al., 1998; Kagawa et al., 1994; Macklin et al., 1995; Mastronardi et al., 1993; Readhead et al., 1994); thus, gene duplications in humans likely cause PMD in similar fashion. The molecular mechanism responsible for generating this genetic lesion is currently unknown, but may involve extensive DNA repeat regions that predispose the Xchromosome to nonreciprocal homologous recombination as is known for duplications in Charcot-Marie-Tooth 1A (Reiter et al., 1996) and Smith-Magenis syndrome (Chen et al., 1997). Clinical phenotypes associated with duplications are variable and have not been characterized in detail; however, data from transgenic mice suggest a direct relationship between the level of overexpression and disease severity (Inoue et al., 1996). Missense and Nonsense Mutations in the PLP1 Gene The third major genetic defect in PMD involves coding region mutations in the PLP1 gene, which arise from missense, nonsense, and splice-site mutations that either preserve the reading frame or cause frame-shifted and truncated proteins. More than 50 distinct mutations have been described worldwide and the clinical phenotypes associated with these mutations span the severity spectrum. Most coding region mutations identiWed to date aVect the primary structure of both DM-20 and PLP1, and the majority of these mutations confer severe disease. On the other hand, perhaps the only region of the open reading frame where diVerent missense mutations confer milder phenotypes encodes the PLP1-speciWc peptide.

Animal Models of Pelizaeus-Merzbacher Disease An important tool set that has contributed signiWcantly to the current level of understanding about the pathogenesis of PMD is the large number of spontaneously occurring animal models from several species that are available for study. The best characterized models in older literature are the canine shaking pup (shp), myelin-deWcient rat (md) and jimpy mouse (jp). Mouse mutants have received considerable attention in recent years because of the identiWcation of patients with the same mutations (Kobayashi et al., 1994; Yamamoto et al., 1998), including the rumpshaker mouse (rsh) and the myelin synthesis-deWcient mouse (msd), and perhaps more importantly because of the application of genetic manipulations to generate wild-type Plp1 overexpressing transgenic mice—strains designated 4e-Plp, #72 and #66—and homologous recombinant null alleles (Boison and StoVel, 1994; Hudson and Nadon, 1992; Kagawa et al., 1994; Klugmann et al., 1997; Readhead et al., 1994; Stecca et al., 2000).

1015

1016

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

A useful aspect of pathophysiology associated with Plp1 coding region mutations in the mouse models is a relatively reliable association between disease severity and life span. Mutations causing severe disease such as jp, msd and the recently described 4J mutation (Pearsall et al., 1997), shorten life span to less than 4 weeks while mild phenotypes are generally associated with normal life spans. This correlation is not clear-cut in PMD patients, and mutations causing severe disease do so through greater functional disability. Together, the mouse mutants comprise the full spectrum of genetic lesions and disease severities found in patients, including mild and severe forms of disease arising from overexpression and coding region mutations. Moreover, a number of the clinical symptoms as well as hypomyelination and oligodendrocyte cell death observed in patients are recapitulated in these animals.

Evidence for Three Distinct Pathogenic Mechanisms for Pelizaeus-Merzbacher Disease Although PMD is found to be a monogenic disease in the majority of patients, considerable evidence indicates that each of the three types of genetic lesions in the PLP1 gene causes disease by distinct mechanisms. Early data from Plp1 overexpressing transgenic mice show that morphological abnormalities in oligodendrocyte cell bodies from these animals are conWned to Golgi proWles, which become distended during myelinogenesis (Kagawa et al., 1994; Readhead et al., 1994). Immunocytochemical evidence indicates that oligodendrocytes in these animals appear to function relatively normally at early ages; however, the processes from these cells eventually lose contact with axons leading to demyelination (Gow et al., 1998). Recent in vitro studies have now shed light on a cellular mechanism that underlies pathophysiology, which revolves around inappropriate traYcking and accumulation of cholesterol and lipid raft components into lysosomes (Simons et al., 2002). There are good reasons to suspect that disease associated with coding region mutations arises from gain-of-function eVects rather than the loss of function observed for gene deletions. First, coding region mutations cause a broad spectrum of disease severities, while null alleles consistently yield mild disease (Garbern et al., 1999). Second, peripheral neuropathy is associated with null alleles but not with the vast majority of coding region mutations. Third, the degree of myelination in null patients and transgenic mice is close to normal whereas most coding region mutations are associated with a significant loss to virtual absence of myelin. Finally, gene expression proWles between null alleles and coding region mutations are distinct (see Figures 42.3B and 42.4B).

Molecular Pathogenesis of Missense Mutations Causing Pelizaeus-Merzbacher Disease The availability of naturally occurring, noncomplimenting myelin mutant rodents such as the jp and msd mouse strains as well as mutations in other species enabled detailed analyses of the biochemical and cellular phenotypes associated with mutations in the PLP1 gene long before identiWcation of the genetic lesions. Early Observations Characterization of the molecular mechanism underlying coding region mutations in PMD has been greatly assisted by the availability of animal models; biopsy procedures in PMD patients were discontinued in the 1950s and this disease is rare enough so that autopsy material is not forthcoming. Early studies in animals such as the jp mouse, md rat and canine shp predate identiWcation of the genetic lesions in these animals; however, X-chromosome linkage and similarities in phenotype, morphology and biochemical analyses of white matter suggested that the aVected loci in these animals might be allelic (for reviews see Hudson and Nadon, 1992; SkoV and Knapp, 1992). In these early studies, several observations have been key to the development of current hypotheses to account for pathogenesis in PMD (Gow et al., 1998). First, immunoelectron microscopy studies

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

looking at Plp1 gene products in wild-type and jp mice revealed striking diVerences in the intracellular localization of these proteins in the secretory pathway of oligodendrocytes (Nussbaum and Roussel, 1983; Roussel et al., 1987; Schwob et al., 1985). Furthermore, transmission electron microscopy of oligodendrocytes revealed amorphous precipitates in the ER of some Plp1 mutant animals, including shp and jp mice (reviewed in Duncan, 1990). Such precipitates are relatively common in a number of cell types in which mutant proteins accumulate in the ER (Rutishauser and Spiess, 2002). Another series of watershed morphometric observations in jp mice laid out the consequences for oligodendrocytes expressing many mutant forms of Plp1 gene products (Ghandour and SkoV, 1988; Knapp et al., 1986). Although not fully appreciated at the time, jp oligodendrocytes exhibit the canonical morphological features of apoptosis (Gow et al., 1998; Grinspan et al., 1998; Lipsitz et al., 1998) but decreases in the size of the oligodendrocyte population in the Wrst 14 days after birth is not a feature of disease (Ghandour and SkoV, 1988). Presumably, the rate of diVerentiation of oligodendrocyte progenitors in these mutants at early ages is roughly equivalent to cell death. However, whether this population receives environmental signals at later times to shut down proliferation and diVerentiation, or whether the progenitor population becomes depleted, remains an open question. A common interpretation of the phenotypes of Plp1 mutant animals at one time was that the oligodendrocytes failed to diVerentiate to the point that they would express high levels of Plp1 gene products. Findings from embryos showing Plp1 gene expression in the CNS from mid-gestation further fueled this view (Timsit et al., 1992); however, ablation of the Plp1 gene in mice by homologous recombination reveal no apparent consequence for oligodendrocyte diVerentiation and has eVectively put such a notion to rest (Boison and StoVel, 1994; Klugmann et al., 1997; Stecca et al., 2000). A more likely interpretation of pathology in the mouse mutants is that as oligodendrocyte progenitors diVerentiate and express Plp1 gene products, they undergo apoptosis, which eVectively eliminates the diVerentiated cell pool. The fact that progenitors continue to divide and replenish this cell population provides a plausible explanation for the apparent diVerentiation defect. An eVective demonstration of the presence of diVerentiated oligodendrocytes in msd brain is shown in Figure 42.1. Myelin basic protein (MBP) immunoXuorescence staining of cerebellum from a P18 msd mouse shows an abundance of this protein in the white matter, which is similar to littermate controls (Fig. 42.1A). Such staining would not be detected in the absence of diVerentiated, myelin membranesynthesizing oligodendrocytes, and previous studies have reached similar conclusions (Ghandour and SkoV, 1988). Indeed, staining of cortex in P18 msd mice using antibodies against MBP and PLP1/DM-20 (top row, Fig. 42.1B) clearly demonstrates the presence of oligodendrocyte membrane around nearby axons with a morphological appearance similar to myelin sheaths in wild-type and rsh mice (Gow et al., 1998). In contrast, PLP1/DM-20 is conWned to oligodendrocyte cell bodies in msd and rsh mice (bottom row, Fig. 42.1B). Together, these data lend strong support to the view that the oligodendrocyte cell lineage diVerentiates normally before undergoing apoptosis in response to the expression of mutant Plp1 gene products. Recent Data and the Protein Misfolding Hypothesis The search for molecular mechanisms underlying pathogenesis of PMD led to a series of in vitro experiments by several groups where wild-type and mutant forms of PLP1 and DM20 were expressed in transfected or transduced cells in culture (Gow et al., 1994a, 1997; Gow and Lazzarini, 1996; Jung et al., 1996; Kalwy et al., 1997; Sinoway et al., 1994; Thomson et al., 1997; Yamada et al., 1999, 2001). Together, these data demonstrate that normal intracellular traYcking of PLP1 gene products through the secretory pathway to the cell surface and into the endocytic pathway to lysosomes is perturbed by single amino acid changes throughout the coding region. An essentially invariant consequence of missense mutations is defective traYcking of PLP1. However, mutations fall into two classes based on whether they block traYcking of PLP1 and DM-20 (class I mutations) or PLP1 but not DM-20 (class II mutations). Interestingly, a strong correlation is apparent between

1017

1018

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

FIGURE 42.1 MBP and PLP1/DM-20 immunocytochemistry in msd and rsh mouse brains. (A) The intensities of MBP staining in wild-type and msd cerebella are similar. (B) MBP staining in the cortex shows many myelin sheath proWles in wild-type as well as msd and rsh mice. Although PLP1/DM-20 colocalizes with MBP in wild-type tissue, this protein is largely conWned to the cell bodies of oligodendrocytes in the mutant animals.

class II mutations and mild forms of disease in patients (Tab. 42.1) and a simple interpretation of these data is that disease severity correlates with the amount of protein that is retained in the perinuclear regions of the cells (Gow and Lazzarini, 1996). A hallmark of interrupted traYcking of these proteins is their accumulation in the ER, which is often seen as perinuclear localization in tubulovesicular proWles that colocalize with ER markers such as the molecular chaperone, BiP, as well as their absence on the cell surface and in lysosomes (Gow et al., 1994b). Importantly, observations from in vitro paradigms using transfected Wbroblasts are similar to those observed for mutant forms of PLP1/DM-20 in vivo (Gow et al., 1998; Roussel et al., 1987) and suggests that disruption to PLP1/DM-20 traYcking is not an artifact or a consequence of cell type-speciWc processes but rather is a general phenomenon in cell biology. Indeed, interrupted traYcking and ER accumulation of dozens of secretory pathway proteins with missense mutations have been reported and a number have been shown to be stably associated with molecular chaperones such as BiP. Unfortunately, the propensity for PLP1 and DM-20 to aggregate in solution renders similar experiments technically challenging. The intracellular traYcking defects that are observed for most mutant forms of PLP1 and DM-20 suggest that these proteins do not adopt stable higher-ordered structures after

1019

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

TABLE 42.1 Intracellular Trafficking of PLP1 and DM-20 for Mutations Identified in PMD Patients and Animal Models Intracellular Accumulation in Transfected COS-7 Cells Class

Disease Severity

PLP1

Missense Mutation

Endoplasmic Reticulum Wild type

-

-

I

Severe

L14P H36P (shp) A38S (4J) T74P (md) W162R T181P L223P A242V (msd)

x x x x x x x x

II

Mild

H36Q (pt) T155I I186T (rsh) D202H D202E P215S V218F V218I L223I

x x x x x x x x x

DM-20 Cell Surface

Endoplasmic Reticulum

x

synthesis in the ER. In such cases, proteins generally remain in this compartment until they are hydrolyzed via the ubiquitin/proteasome pathway. This pathway is known as ERassociated degradation (ERAD), which operates constitutively in most cell types and involves: retrotranslocation of defective or excess wild-type polypeptides through the ER membrane into the cytoplasm, polyubiquitination, and proteolysis in the proteasome (reviewed in Brodsky and McCracken, 1999). In light of the facts that ERAD is evolutionarily conserved in eukaryotes and that secretory pathway proteins are constantly Xowing through this pathway, it is reasonable to expect that cells should cope eYciently with mutant proteins in the ER. However, conceivably, the rates of synthesis of some mutant proteins could be so great as to overwhelm the cells capacity to degrade them. In such cases, misfolded proteins would accumulate in the ER and eventually threaten cell viability (reXecting a loss of homeostasis), and the cell might be expected to react to this metabolic stress by suppressing protein synthesis and activating genetic programs to alleviate the toxic eVects of protein accumulation. One such signaling pathway that serves to reestablish homeostasis in cells in response to misfolded protein accumulation in the ER is called the unfolded protein response (UPR).

The UPR Signaling Cascade Early awareness of the UPR at the cellular level dates back to the 1970s and 1980s as a result of examining the intracellular traYcking of wild-type and mutant forms of transmembrane proteins through the secretory pathway (reviewed by GaroV, 1985; Kozutsumi et al., 1988). Retention of mutants in the ER, which could also be eVected by drugs such as tunicamycin to block N-glycosylation of glycoproteins, led to the notion that the ER serves as a quality control organelle in which individual proteins must adopt stable three-dimensional structures before gaining entry into the remaining membrane compartments of the cell (Bonifacino and Lippincott-Schwartz, 1991; Gething and Sambrook, 1992; Hammond and Helenius, 1995; Klausner and Sitia, 1990; Rose and Doms, 1988). A number of ‘‘UPReVector’’ proteins, such as molecular chaperones like BiP that belong to the hsp70 family, had been identiWed by the late 1980s, but the bulk of the characterization and integration

Cell Surface x

x x x x x x x x x x x x x x x x x

1020

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

of UPR components into a cohesive signaling pathway has been achieved over the past 10 years. Components of the UPR Although our knowledge about the complexity and function of the UPR is still evolving, a number of central players have been deWned (for recent reviews, Harding et al., 2002; Kaufman et al., 2002; Ma and Hendershot, 2001). The UPR can be broadly divided into two components; one that modulates homeostasis at the level of transcription while the other component modulates homeostasis through translational repression. Table 42.2 lists well characterized upstream components of the UPR in mammals as well as their major functions and targets and Figure 42.2 shows the Xow of information in this signaling cascade from activation in the ER to the nucleus. Three proteins that have been characterized to date, which we refer to as ‘‘UPRinitiator’’ proteins, serve as direct monitors of protein Xux through the ER. A fourth protein, ATF4, which may be activated by signals outside of the UPR, regulates the expression of important UPR target genes. These initiators activate a number of ‘‘UPReVector’’ proteins that assist cells in adapting to imposed stress to establish homeostasis, as well as to participate negative feedback loops that deactivate UPR signaling. Thus, Ire1 (Fig. 42.2A) and PERK (Fig. 42.2B) are ER-resident receptor kinases, each with a luminal domain that associates with the molecular chaperone, BiP. In the event of an increase in the level of unstable nascent polypeptide intermediates in the ER, BiP dissociates from BiP-Ire1 and/or BiP-PERK complexes, which frees up these initiator proteins for dimerization, autotransphosphorylation, and signaling to the nucleus (Bertolotti et al., 2000). Activated Ire1 and PERK also exhibit additional properties—endoribonuclease activity for Ire1 and an eIF2a kinase activity for PERK—and eVect transcriptional modulation and global translational repression, respectively (Harding et al., 2002). Ire1 catalyzes processing of an mRNA that encodes the XBP-1 transcription factor, which in turn induces expression of molecular chaperone genes (reviewed in Ma and Hendershot, 2001). On the other hand, PERK-mediated phosphorylation of the alpha subunit of eIF2 (eIF2a) inhibits GDP/GTP exchange on this factor by the guanine nucleotide exchange factor (GEF), eIF2B, and blocks binding of Met-tRNAfmet to eIF2. Consequently, blocking formation of the eIF2.GTP.Met-tRNAfmet ternary complex ensures that

TABLE 42.2 Summary of the Functions and Targets of UPR Associated Proteins Protein Ire1

Group Initiator

Function Kinase/ endonuclease

Targets Xbp-1 mRNA splicing

Reference (Mori et al., 1993) (Cox and Walter, 1996) (Wang et al., 1998a)

PERK

eIF2a-kinase

eIF2

(Harding et al., 1999)

ATF4

Transcription factor

CHOP gene

(Fawcett et al., 1999)

ATF6

Transcription factor

XBP-1 gene

(Yoshida et al., 2000)

Transcription factor

Molecular Chaperone expression

(Yoshida et al., 2001) (Calfon et al., 2002)

CHOP

Transcription factor

ATF3, apoptosis-related genes

(Ubeda et al., 1996)

ATF3

Transcription factor

CHOP gene repressor

(Wolfgang et al., 1997) (Bole et al., 1986)

BiP

Molecular chaperone

Unfolded protein binding

(Kozutsumi et al., 1988)

Xbp-1

Effector

ERp59

Disulfide bond formation

Nascent ER polypeptides

(Ferrari and Soling, 1999)

ERp72

Disulfide bond formation

Nascent ER polypeptides

(Ferrari and Soling, 1999)

eIF2

Ribosome assembly

Met-tRNAi

(Harding et al., 2000)

Gadd34

Phosphatase binding

PP1

(Novoa et al., 2001)

1021

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

A. IRE1 activation

B. PERK/ATF4 activation UPR Mutant Protein

BiP

UPR Mutant Protein

ER

ER BiP

BiP

BiP

AA

A AA

XBP-1

P

A

P

Ire1

P

PERK

P

P eIF2α

XBP pre-mRNA

PP1 GADD34

Global Translation

Cyt

eIF2α

Nuc

A

ATF4

AA

AA A

A

AA A

AA

Molecular Chaperones

C. ATF6 activation UPR

Golgi

Trafficking

CHOP GADD34 Molecular Chaperones

Mutant Protein

ER

S1P S2P

BiP

BiP

ATF6

Cyt Nuc XBP BIP ERp59 ERp72

FIGURE 42.2 UPR signaling is initiated through three proteins and regulates transcription and translation.(A) The Ire1 receptor is localized to the ER and modulates transcription through XBP-1. (B) The PERK receptor is localized to the ER and represses global translation through phosphorylation of eIF2a. ATF4 is encoded by a cytoplasmic mRNA that is translated in response to changes in the rate of protein synthesis and regulates transcription. (C) ATF6 is a membrane-tethered protein that is localized in the ER but is released to the Golgi during a UPR and regulates transcription after cleavage by proteases.

recruitment of initiator tRNA to the 40S ribosome preinitiation complex cannot occur (reviewed in Kimball, 1999). Furthermore, phosphorylated eIF2 competitively inhibits the GEF activity of eIF2B and leads to rapid repression of all translation in the cell. Under cellular conditions where the eIF2 is partially phosphorylated and global protein synthesis is only partially inhibited, the translation of speciWc classes of mRNAs yield proteins that serve to propagate UPR signaling into the nucleus. One such class of mRNAs

1022

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

is constitutively present in most cell types and a representative example of this class is an mRNA that encodes the transcription factor, ATF4. Importantly, the 5’ region of this mRNA includes a short open reading frame (called a 5’ ORF) that, under normal conditions, is translated to yield an inactive protein. However, when the available pool of eIF2 becomes partially phosphorylated, translation initiation on this mRNA switches to a downstream open reading frame (ORF) that encodes the active transcription factor. Apparently, as ribosomal scanning positions the 40S ribosomal subunit at the initiation codon of the 5’ ORF under normal cellular conditions, the abundance of ternary complexes ensures rapid binding and prompt 80S ribosome assembly. However, eIF2a phosphorylation drives down the level of ternary complexes, which delays formation of the preinitiation complex and increases the probability that the small ribosomal subunit will move beyond the AUG of the 5’ UTR. This process is known as bypass scanning. In such cases, the 40S ribosome continues along the mRNA to Wnd the downstream ORF where ternary complex binding enables ribosome assembly and translation to generate ATF4 (DeGracia et al., 2002; Harding et al., 2000). One of the consequences of bypass scanning is induction of ATF4 target genes that encode: several molecular chaperones including BiP, which help to alleviate ER stress and negatively regulate the initiation of UPR signaling (Bertolotti et al., 2000); GADD34 protein, which negatively regulates translational repression by recruiting the protein phosphatase PP1 to dephosphorylate eIF2 and reestablish ribosome assembly; and the transcription factor CHOP, which plays a pivotal role in UPR signaling and has been shown to induce apoptosis in several cell types (Harding et al., 2000). Similar to ATF4, CHOP mRNA contains a 5’ ORF (Jousse et al., 2001), which suggests that CHOP function is likely to be important under conditions where eIF2 is phosphorylated. Furthermore, accumulation of GADD34 in the cytoplasm likely ensures that translational repression triggered by PERK is temporary and that CHOP is only transiently translated (Novoa et al., 2001). A further requirement for CHOP to function as a transcription factor is phosphorylation by p38 kinase, which imparts additional constraints on CHOP function (Wang and Ron, 1996). ATF6 is a third UPR-initiator protein and is an ER membrane-tethered transcription factor that is released into the cytosol by S1P/S2P proteases in response to the accumulation of unfolded proteins. These proteases are unlikely to be speciWc activators for the UPR signaling pathway because they also process another well-characterized membranetethered transcription factor, SREBP, which plays a central role in cholesterol biosynthesis (Lee et al., 2002; Ye et al., 2000). Interestingly, tethered-ATF6 is retained in the ER in a complex with BiP under normal conditions but is released for traYcking to the Golgi apparatus in response to ER stress where it is proteolytically cleaved (Shen et al., 2002). Targets of ATF6 include XBP-1, genes encoding molecular chaperones and the protein disulWde isomerases, ERp59, and ERp72 (Okada et al., 2002). Many UPR-Effector Proteins Are Constitutively Expressed During the course of normal development and maintenance, individual cells continually synthesize membrane components in the ER for delivery to other compartments, including the Golgi apparatus, cell surface, endosomes, and lysosomes. In addition to general transcriptional regulation of individual components by cell-type speciWc factors, membrane biogenesis through the secretory pathway is regulated at the level of bulk Xow, or Xux, by several components of the UPR. For example, molecular chaperones such as BiP are normally present in the ER of most cells under physiological conditions, which suggests that XBP-1 and ATF6 are continually processed at low levels. Transcription factors such as CHOP and ATF3 are also expressed at low levels, thereby implying PERK activation and ATF4 translation. However, the limited studies of UPR-eVector genes carried using normal cells at cellular homeostasis suggests that such low level expression reXects the activity of UPR signaling-independent mechanisms (Brewer et al., 1997). Nevertheless, the fact that UPR-eVectors are present in cells under normal conditions and are strongly induced by the UPR suggests that these proteins, and by direct extension the UPR, operate to protect cells against metabolic stress involving the ER and to preserve or re-establish

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

cellular homeostasis. The consequence for cells in which homeostasis cannot be regained is often apoptosis by activation of caspase cascades. These cascades may involve the initiator-caspase 12 in mice but not in humans (Fischer et al., 2002) where, presumably, another caspase family member is activated. In light of the constitutive activity of UPR-eVectors, how is the normal expression of genes encoding these proteins distinguished from cell stress-related expression? One possibility may be through the utilization of conserved cis-elements in the proximal promoters of UPReVector genes to induce transcription (Lee, 2001; Ma et al., 2002b; Yoshida et al., 1998). Indeed, two such elements have been well-characterized in mammalian cells, namely the unfolded protein response element (UPRE) and the endoplasmic reticulum stress element (ERSE), which are found in the promoters of many molecular chaperones. A third element, called the C/EBP-ATF composite element, has recently been identiWed in the CHOP promoter and is critical for induction of this gene by UPR signaling. A number of transcription factors activated by the UPR are known to bind to the UPRE, ERSE and C/EBP-ATF motifs, including XBP-1, ATF3, ATF4, ATF6, C/EBP, NF-Y and YY1, and TFII-I.

Involvement of the UPR in Pelizaeus-Merzbacher Disease and Animal Models The Wrst indication that activation of the UPR is a feature of disease in Plp1 mutant animals came from a demonstration that the Bip gene is induced in the brains of jp and msd mice although, curiously, Bip gene induction is not observed in md rats or the canine shp model (Hudson and Nadon, 1992). More recent experiments focused exclusively on the mouse mutants demonstrate induction of a number of molecular chaperones, but the magnitudes of these increases are relatively modest (Southwood and Gow, 2001). Nevertheless, these observations have prompted a detailed examination of UPR activation in Plp1 mutant mice (Southwood et al., 2002). The UPR Is Activated in Pelizaeus-Merzbacher Disease Northern blots in Figure 42.3A reveal induced expression of six UPR-eVector genes in the spinal cords of msd, jp and rsh mice. The Chop gene is genetically downstream of PERK and ATF4 and Atf3 is regulated by CHOP (Chen et al., 1996). On the other hand, Bip, ERp59 and ERp72 are downstream of Ire1 and ATF6. Importantly, northern blots in Figure 42.3B demonstrate that the same set of UPR-eVector proteins are induced in subcortical white matter compared with occipital gray matter from a PMD patient (Dexon6). This patient harbors a splice-site mutation at the splice donor site of exon 6 that causes skipping of this exon (Hobson et al., 2000; Southwood et al., 2002). These data contrast with the results obtained from the nonplaque region of a patient with multiple sclerosis (MS). In this case, the BIP gene is the only gene for which expression is signiWcantly diVerent between white and gray matter, and is actually increased in gray matter. The signiWcance of these data for MS is presently unclear and is under further investigation. Importantly, characterization of UPR-eVector genes inside active plaques and adjacent gray and white matter regions may reveal important aspects of the pathophysiology as has been alluded to in previous studies. Induced expression of UPR-eVector genes in msd and rsh mice has been conWrmed with immunocytochemical data, which show not only that CHOP is localized to the nuclei of Plp1-positive oligodendrocytes, but also ATF3. Both of these bZip transcription factors are thought to remain cytoplasmic until their activation through phosphorylation by MAP kinases p38 and JNK, respectively (Wang and Ron, 1996). It is not currently known if CHOP and ATF3 are present in the same oligodendrocytes at the same time or whether there is a temporal expression pattern; however, developmental northern analyses in spinal cords of msd mice indicate that Chop induction precedes that of Atf3, possibly by as much as 5 days (Southwood et al., 2002). Studies of liver regeneration in rat indicate that these two transcription factors are not normally colocalized in the same cells and may antagonize each others expression (Chen et al., 1996; Wolfgang et al., 1997). A signiWcant distinction between CHOP and ATF3 expression in msd mice is the expression of ATF3 by microglia in addition to oligodendrocytes, while CHOP

1023

1024

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

A

B P15 wt msd

P18 wt

jp

P21 wt rsh

Plp1

MS WM

GM

Æexon6 WM GM

PLP

Chop CHOP Atf3 ATF3

18S

wt

P16 msd

P18 wt

jp

wt

P17 rsh

BIP

BiP ERp59 ERp59 ERp72 ERp72 18S

18S

FIGURE 42.3 Northern blotting shows the induction of six genes encoding UPR-eVector proteins. (A) Induction of genes in msd, jp, and rsh mice. (B) Induction of genes in PMD but not an MS patient.

is expressed only in oligodendrocytes. Although morphology, proliferation, and gene expression of microglia are altered in response to mutant Plp1 expression (reviewed by Vela et al., 1998), the signals that induce these changes remain unclear. However, there is no evidence for Plp1 expression and activation of the UPR in this cell type, which indicates that Atf3 expression is induced by a broader range of signals than is Chop and that ATF3 may have additional functions outside of the UPR. The UPR Modulates Disease Severity in Animal Models of Pelizaeus-Merzbacher Disease Considering changes in cell behavior at the level of transcription in response to UPR signaling, the greatest attention has been focused on the transcription factors that eVect these changes and to gain insight into the compensatory mechanisms that cells employ to establish or maintain homeostasis. CHOP has been particularly well studied in this regard, partly because it was one of the Wrst UPR-induced transcription factors identiWed and because of the important role this protein plays in initiating apoptosis in at least some cell types (Ubeda et al., 1996; Wang et al., 1996; Zinszner et al., 1998). From these studies, CHOP is widely believed to be a pro-apoptotic transcription factor (reviewed in Harding et al., 2002). The central role proposed for CHOP in the UPR, as well as its strong induction in oligodendrocytes of Plp1 mutant mice, suggests that this transcription factor might be a critical modulator of disease severity in PMD patients. Indeed, previous studies in animal models have suggested that severe disease correlates with abnormally high levels of oligodendrocyte death (Gow et al., 1998; Schneider et al., 1992). To explore the potential involvement of CHOP in modulating disease severity, rsh mice were crossed with Chop-null mice (Zinszner et al., 1998) and examined using a Kaplan-Meier analysis (Southwood et al., 2002). Changes in life span for rsh mice in the absence of CHOP is dramatic. While rsh mice live approximately 2 years, Chop-null/rsh double-mutant mice die as early as 5 weeks of age and have an average life span of less than 10 weeks. Furthermore, these animals exhibit seizures not unlike that of jp and msd mice, which is a more severe form of disease than is usually observed for rsh mice, and indicates that CHOP serves as an anti-apoptotic transcription factor in PMD. Although the mechanism underlying the functional switch

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

between pro-apoptotic and anti-apoptotic properties is unknown, it is likely that the fate of cells after UPR activation is determined by the downstream target genes of CHOP as well as other UPR-induced transcription factors. In this regard, CHOP appears to regulate overlapping but distinct targets in oligodendrocytes compared to other cell types (Southwood et al., 2002; Wang et al., 1998b). A similar heterogeneity in CHOP target genes has also been observed in transfected COS-7 cells expressing a mutant form of the mitochondrial protein, OTC, that misfolds in the mitochondrial matrix. Under these conditions, expression of Chop and four nuclearencoded mitochondrial molecular chaperones is induced but not the expression of Wve ERlocalized molecular chaperones. Furthermore, promoters regulating expression of the mitochondrial chaperones contain cis-elements that bind the CHOP/C/EBPb heterodimer in vitro. Conversely, induction of a UPR using tunicamycin or thapsigargin induces expression of ER-localized molecular chaperones but not those localized to mitochondria (Zhao et al., 2002). A simple interpretation of these data is that CHOP is expressed in response to the accumulation of unfolded proteins in at least two organelles and CHOP target gene selection is regulated in conjunction with unknown organelle-speciWc factors. Activation of the UPR Is Specific for Coding Region Mutations in PLP1 In the light of data from early immunoelectron microscopy studies in jp mice, which showed that mutant PLP1/DM-20 were largely localized to the ER of oligodendrocytes (Roussel et al., 1987), the demonstration that the UPR is activated in PMD tissue and Plp1 mutant animals is a consistent and satisfying outcome. But is the UPR activated when the PLP1 gene is overexpressed in duplication patients or transgenic mice? Immunoelectron microscopy and immunocytochemical studies show that most wild-type PLP1/DM-20 accumulates in the myelin sheaths. Nonetheless, these proteins are detectable in oligodendrocyte cell bodies and are localized largely to the Golgi apparatus (Nussbaum et al., 1985; Schwob et al., 1985). Interestingly, morphological evidence from Plp1 overexpressor transgenic mice suggests that the products from this gene also accumulate in the Golgi apparatus and may perturb the function of this organelle. If so, the levels of PLP1/DM-20 in the ER should be relatively normal and the UPR would not be expected to be induced. Indeed, the data in Figure 42.4 suggest that this is the case. Immunocytochemical staining of wild-type cerebellum shows strong anti-PLP1/DM-20 reactivity in white matter tracts (Fig. 42.4Aa). CHOP is not expressed (Fig. 42.4Ab) by any of the cells in this Weld, the positions of which are revealed by the nuclear DAPI stain (Fig. 42.4Ac). White matter tracts in the cerebellum are weakly labeled with anti-PLP1/DM-20 antibodies in transgenic 4e-Plp overexpressor mice and several oligodendrocyte cell bodies are heavily labeled (arrowheads, Fig. 42.4Ad). Nevertheless, CHOP is not induced in these cells. Staining of cerebellum with antibodies against ATF3 also fails to label oligodendrocytes (data not shown). Consistent with these Wndings, northern blots of autopsy material from a PMD duplication patient (Fig. 42.4B) show that neither CHOP nor ATF3 are induced in white matter compared to gray matter. Together, these data support the contention that the UPR is an important feature of the pathophysiology in PMD caused by coding region mutations.

Other Noninclusion Diseases Associated with Activation of the UPR With a great deal of attention focused on the delineation of UPR signaling pathways in the basic science and clinical arenas, a growing number of diseases that stem from mutations in proteins that may activate a UPR are coming to light. For example, patients with WolcottRallison syndrome disease harbor mutations in the PERK gene and clinical symptoms such as neonatal type I diabetes, mental retardation, cardiovascular abnormalities, hepatic and renal dysfunction and other multisystemic manifestations (Delepine et al., 2000). The most likely cause of disease in these patients is the failure of cells to repress protein translation in response to metabolic or heat-induced stress, which places homeostasis at risk and leads to cell death. Currently, it is unclear why many of these diseases do not appear to be associated with intracellular inclusion bodies.

1025

1026

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

A

B Duplication

PLP

WM

GM

PLP

CHOP a

d

CHOP

ATF3

e

c

f

DAPI

b

18S

FIGURE 42.4 UPR induction is not a feature of disease in PLP1 overexpressors. (A) CHOP is not expressed by wild-type oligodendrocytes or oligodendrocytes overexpressing a wild-type Plp1 transgene. (B) Northern blotting shows that neither CHOP nor ATF3 are induced in white matter of a PMD patient with a duplicated wild-type PLP1 gene.

Missense mutations in several genes cause the group of peripheral neuropathies known collectively as Charcot-Marie-Tooth disease and involve the PMP22, MPZ, and CONNEXIN 32 genes. Heterologous expression, in transfected Wbroblasts, of mutant forms of PMP-22 identiWed from patients have been examined in greatest detail. In similar fashion to PLP1, missense mutations disrupt the normal intracellular traYcking of PMP-22 through the secretory pathway to the cell surface and the mutant proteins accumulate in the ER (D’Urso et al., 1998; Naef et al., 1997; Sanders et al., 2001; Tobler et al., 1999). Such accumulation has been demonstrated in PNS-myelinating Schwann cells from patient biopsy specimens and is associated with increased expression of molecular chaperones. These data suggest that the UPR is activated in CMT; however, widespread apoptosis is not a feature of even the most severe forms of CMT, which may mean that misfolded PNS myelin-speciWc proteins are more easily degraded than those in the CNS or that Schwann cells and oligodendrocytes diVer in their sensitivity to UPR inducing metabolic stress. Cystic Wbrosis stems from genetic mutations that perturb normal function of the chloride transporter protein, CFTR (recently reviewed by Gelman and Kopito, 2002). Although this disease aVects several cell types in patients, the most prominent pathology is observed in lung. Extensive analysis of CFTR mutations from the 1200 or so mutations and polymorphisms that have been described for this gene indicate that four or more distinct mechanisms may account for pathophysiology: class I mutations are null or functionally null; class II mutations are associated with interrupted intracellular traYcking in the ER; class III mutations arise from abnormal plasma membrane recycling, endocytosis, or lysosomal transport; and class IV mutations stem from defects in channel activation and regulation by interacting proteins. By far the most common mutation in the CFTR gene (> 90% of patients of Northern European descent) involves an in-frame deletion of three nucleotides that encode phenylalanine at position 508 (DF508) of this

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

1480 amino acid protein. The DF508 mutation is commonly designated as a class II mutation from both in vitro and in vivo studies (Denning et al., 1992; Kartner et al., 1992), although the extent of the intracellular traYcking defect exhibits cell type dependence (Kalin et al., 1999). The DF508 mutation is also noteworthy from the perspective of pharmacological strategies for ameliorating disease. Thus, this mutation can form functional chloride channels in cells if it reaches the plasma membrane, which can be accomplished by reducing the growth temperature or by treating cells with protein stabilizing agents such as glycerol and sorbitol (Colledge et al., 1995; Gelman and Kopito, 2002; Sato et al., 1996). Despite the fact that mutant CFTR does not form inclusions in patients, in vitro paradigms can be manipulated to yield perinuclear inclusions of polyubiquitylated protein (Johnston et al., 1998). Intermolecular folding of wild-type CFTR does not appear to proceed eYciently in the ER because >75% of nascent chains are degraded by the proteasome complex under normal physiological conditions (Jensen et al., 1995). Inhibition of this ERAD pathway leads to accumulation of wild-type or mutant CFTR in the cytosol in the form of large perinuclear aggregates called aggresomes (Johnston et al., 1998). In Wbroblasts, aggresomes are bordered by intermediate Wlaments such as vimentin and it is possible that these structures serve to sequester the mutant protein aggregates until such times as they can be degraded by the proteasome complex. Aggresome formation is not restricted to CTFR accumulation; presenilin I also forms these complexes (Johnston et al., 1998), and we have also observed these structures in transfected cells overexpressing wildtype PLP1 in the presence or absence of the proteasome inhibitor MG132 (unpublished observations, and Gow et al., 1994a). The morphological and biochemical characterization of aggresomes by Kopito and colleagues, as well as the demonstration that misfolded protein aggregates impair the function of the ERAD pathway (Bence et al., 2001; Johnston et al., 1998; Kopito and Sitia, 2000) provide an important interpretation of inclusion pathology that is observed in many neurodegenerative diseases. Another example of a disease that may involve the UPR is a cell-type speciWc disease, called leukoencephalopathy with vanishing white matter (VWM). VWM is associated with mutations in any of the subunits of the ubiquitously expressed GEF protein, eukaryotic initiation factor 2B (eIF2B). eIF2B catalyzes the GDP-GTP exchange on eIF2 to enable binding of Met-tRNAfmet and its subsequent recruitment to the ribosome preinitiation complex. Furthermore, eIF2 is the central regulator of global protein translation in the cell and, in particular, the phosphorylation status of the eIF2a subunit determines whether-ornot eIF2B GEF activity is induced or inhibited by eIF2.GDP (Kimball, 1999). Mutations that modify the interaction between eIF2B and eIF2 or alter the activities of these initiation factors, as is likely to be the case in VWM, diminish the ability of cells to regulate protein synthesis in response to cell stress and even under normal physiological conditions. For unknown reasons in VWM disease, oligodendrocytes and neurons appear to be selectively sensitive compared to other cell types following transient core temperature elevations such as those experienced during infections (Leegwater et al., 2001; van der Knaap et al., 1997; van der Knaap et al., 2002). Evidence that is now beginning to emerge from several groups implicates the UPR in familial forms of Parkinson’s disease, in particular those associated with mutations in the PARKIN gene. Although Parkinson’s disease is considered here and elsewhere as an inclusion disorder, patients with PARKIN mutations rarely exhibit inclusion pathology. One of the substrates of parkin E3 ligase is the parkin-associated endothelin receptor-like receptor, Pael-R, which has been shown to accumulate with a-synuclein in Lewy bodies from sporadic Parkinson’s patients (Imai et al., 2001). Pael-R is a transmembrane protein that is expressed by oligodendrocyte lineage cells and dopamine neurons and is most abundant in corpus callosum and substantia nigra. A substantial portion of this receptor is normally conjugated to ubiquitin, suggesting either that the protein is synthesized in excess and residual polypeptide chains are subsequently degraded by the ERAD pathway, or that the protein does not readily adopt a stable higher-ordered conformation and many of the nascent chains are targeted to the ERAD pathway. When overexpressed in transfected cells, Pael-R is also polyubiquitinated and induces robust apoptosis that is enhanced

1027

1028

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

by concomitant inhibition of proteasome function and diminished by coexpression of wildtype parkin. These data indicate that Pael-R expression induces an unfolded protein response, which leads to apoptosis of these cells (Imai et al., 2001). Consistent with this notion, analysis of autopsied frontal lobes from autosomal-recessive juvenile Parkinson’s patients (AR-JP, associated with mutations in the PARKIN gene) reveals substantial accumulation of detergent-insoluble Pael-R compared to control tissue. Importantly, the absence of Lewy bodies in the parenchyma of AR-JP patients indicates that Pael-R may accumulate in the ER and may kill the nigral neurons through activation of the unfolded protein response pathway in similar fashion to that observed in cell culture experiments (Imai et al., 2001). Unexpectedly, the pathophysiology of several experimental models of Parkinson’s disease also involves activation of the UPR, at least for in vitro systems (Ryu et al., 2002). As determined by SAGE analysis and northern blotting, PC12 cell-derived dopamine neurons grown in the presence of 6-hydroxydopamine, MPPþ or rotenone induce a number of UPR target genes encoding: heat shock proteins and molecular chaperones; CHOP and one of its DNA-binding partners, C/EBPb ubiquitin and ubiquitin-conjugating enzymes and proteasome subunits. Furthermore, western blotting reveals that two of the UPR-initiator proteins, PERK and Ire1, are phosphorylated, that eIF2a is phosphorylated and that the mRNA encoding ATF4 is translated in the presence of these Parkinson’s mimetic agents. Several recent studies also suggest that familial Alzheimer’s disease may involve activation of the UPR although some of these data are currently conXicting. While investigating the proteolytic cleavage and nuclear localization of the cytoplasmic tail of Ire1 in response to UPR activation, two groups have demonstrated that the polytopic presenilin (PS) protein and Ire1 physically interact and that proteolytic processing of Ire1 is defective in cells derived from PS I-null mice or transfected neurons expressing mutant forms of PS I identiWed in patients (Katayama et al., 1999; Niwa et al., 1999); two related PRESENILIN genes are known in mammals and their encoded products are believed to exhibit g-secretase activity or are closely associated with g-secretase. Moreover, these groups Wnd that tunicamycin-induced signal transduction from the ER to the nucleus is attenuated in cells from PS I-null mice, which results in blunted induction of the molecular chaperone, BiP, in response to tunicamycin treatment. On the other hand, a third group (Sato et al., 2000) has demonstrated in PS I/PS II-double null cells that Ire1 and PERK activation and signaling are not compromised after indution of the UPR. Furthermore, induction of Bip and Chop expression occur normally in the absence of PS or the presence of mutant forms of PS I identiWed in familial Alzheimer’s patients or in autopsied brains from sporadic Alzheimer’s patients.

Toward a Set of Criteria to Identify Novel UPR-Associated Diseases To recognize additional degenerative diseases that may involve mutations in secretory pathway proteins that activate the UPR, it is important to develop a set of criteria that may be used to screen out diseases for which other mechanisms underlie pathogenesis. Toward this end, three factors are worthy of consideration. Null Alleles Complete gene deletions or nonsense mutations at the 5’ end of the open reading frame represent the strongest cases for loss-of-function disease and particular care should be taken to characterize the phenotype from these genetic lesions for comparison with phenotypes arising from other mechanisms. Although these null alleles could yield recessive or semi-dominant phenotypes, they would be expected to confer disease that is signiWcantly more mild than, or clinically distinct from, at least a portion of allelic mutations. With this criterion, it is possible to eliminate diseases for which mutations simply abolish protein function even in the event that those mutations disrupt protein traYcking. For example, mutations in the immunoglobulin superfamily neural cell adhesion molecule, L1, underlie a variety of clinical symptoms collectively referred to as CRASH

DISEASES FOR WHICH DEPOSITS OR INCLUSIONS ARE NOT PROMINENT FEATURES OF PATHOLOGY

(Fransen et al., 1995). Heterologous expression of missense mutations in transfected Wbroblasts reveals that most fail to traverse the secretory pathway but rather are retained in the ER (De Angelis et al., 2002; Moulding et al., 2000). However, available evidence suggests that the clinical symptoms are probably not associated with activation of the UPR. In particular, expression of mutant L1 under transcriptional control of the L1 promoter in transgenic mice yields a phenotype that is indistinguishable from the null allele using several criteria (Runker et al., 2003). These data indicate that mutant forms of L1 are nonfunctional and that the disease symptoms are consistent with a loss-of-function phenotype. A second example of a disease where missense mutations cause protein accumulation in the ER, but where clinical symptoms more likely arise from loss of function is autosomal dominant retinitis pigmentosa (ADRP). In this case, Rhodopsin-null mice exhibit very similar retinal degeneration as do transgenic mice expressing mutant forms of rhodopsin identiWed in ADRP patients (Humphries et al., 1997; Olsson et al., 1992). In contrast, PLP1-null alleles identiWed in PMD patients consistently yield more mild forms of disease than most missense mutations. In addition, null patients exhibit peripheral neuropathy, which is a feature of disease distinct from that observed in patients with coding region mutations (Shy et al., 2003). Dominant and Semi-Dominant Alleles Accumulation of mutant proteins in the ER frequently occurs at the level of nascent polypeptide synthesis and prior to the integration of these protein monomers into multimeric complexes in the ER. Accordingly, a single mutant copy of a gene encoding an abundant secretory pathway protein may be suYcient to induce the clinical symptoms of a UPR-associated disease. Moreover, addition of wild-type copies of this gene (e.g., by gene therapy) would not be expected to ameliorate the phenotype because toxicity of the mutant protein would not be associated with a loss of function. In fact, such a strategy may exacerbate disease by increasing Xux through the secretory pathway and placing heavier demands on the ER. In this regard, the introduction of supernumerary copies of a wildtype Plp1 gene into mice harboring the Plp1jp allele fails to ameliorate the severe disease conferred by the jp allele even though transgene-derived wild-type PLP1 and DM-20 in these animals can traverse the secretory pathway to the myelin sheath (Schneider et al., 1995). Clearly, disease in this case is not a consequence of the absence of Plp1 gene products in the myelin but rather is a consequence of mutant protein accumulation in the secretory pathway (Gow et al., 1998; Nussbaum and Roussel, 1983; Roussel et al., 1987). Recessive Alleles In general, it is likely to be diYcult to identify UPR-associated diseases exhibiting recessive phenotypes without the aid of biopsy or autopsy material to demonstrate perinuclear accumulation of mutant proteins or without using in vitro approaches to examine intracellular traYcking of the mutant proteins in transfected cells in culture. For genes expressed at relatively low levels, it is reasonable to expect that a coding region mutation in one allele might not yield a suYcient amount of protein accumulation to overwhelm the UPR and cause disease. However, such mutations may do so if both gene copies are mutated. This case has clearly been shown by several groups in mice (Power et al., 1996; Turnley et al., 1991; Yoshioka et al., 1991). For example, transgene-derived expression of the H-2Kb protein in oligodendrocytes using the cell type-speciWc MBP promoter/enhancer causes a dysmyelinating phenotype in some hemizygous lines while others are asymptomatic. Although this MHC molecule has a wild-type sequence, it normally resides in the ER until forming a ternary complex with b2-microglobulin and a polypeptide supplied by a TAP tansporter. Because oligodendrocytes usually express neither b2-microglobulin nor the transporter, H-2Kb accumulates in the ER of these cells. Importantly, the disease phenotype is dependent on transgene expression above a threshold, which can be breached in high expressing hemizygous lines, by breeding an asymptomatic line to homozygosity or by increasing gene dosage through the breeding of two asymptomatic hemizygous lines.

1029

1030

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

SUMMARY Although involvement of the UPR is suspected in a number of common degenerative disorders, until recently too little has been known or understood about this signaling cascade for any meaningful exploration of human disease. However, rapid progress over the past 2 years by several prominent laboratories has yielded an important framework of information with which to assess the role of this signaling pathway in many degenerative diseases. In this regard, coding region mutations in the PLP1 gene are associated with induction of the UPR in vitro, in the CNS of animal models of PMD, and in a PMD patient. The activation of this signaling cascade occurs in animals with severe and mild forms of disease and does not determine disease severity per se. However, UPR activation is directly involved in modulating disease severity in the sense that interfering with the function of this pathway negatively aVects cell survival. Thus, PMD is the Wrst example of a degenerative disease in which the UPR is not only activated but also modulates disease severity.

Acknowledgments My warm thanks to Cherie Southwood for her critical review of this work and comments. This work is supported by grants from NINDS (RO1 NS43783), the National Multiple Sclerosis Society (RG 2891-B-2), and the Childrens Research Center of Michigan.

References Bence, N. F., Sampat, R. M., and Kopito, R. R. (2001). Impairment of the ubiquitin-proteasome system by protein aggregation. Science 292, 1552–15525. Berger, J., Moser, H. W., and Forss-Petter, S. (2001). Leukodystrophies: Recent developments in genetics, molecular biology, pathogenesis and treatment. Curr Opin Neurol 14, 305–312. Bertolotti, A., Zhang, Y., Hendershot, L. M., Harding, H. P., and Ron, D. (2000). Dynamic interaction of BiP and ER stress transducers in the unfolded-protein response. Nat. Cell Biol. 2, 326–332. Boison, D., and StoVel, W. (1994). Disruption of the compacted myelin sheath of axons of the central nervous system in proteolipid protein-deWcient mice. Proc. Natl. Acad. Sci. USA 91, 11709–11713. Bole, D., Hendershot, L. M., and Kearney, J. (1986). Posttranslational association of immunoglobulin heavy chain binding protein with nascent heavy chains in nonsecreting and secreting hybridomas. J. Cell Biol. 102, 1558–1566. Bonifacino, J. S., and Lippincott-Schwartz, J. (1991). Degradation of proteins within the endoplasmic reticulum. Curr. Opin. Cell Biol. 3, 592–600. Brewer, J. W., Cleveland, J. L., and Hendershot, L. M. (1997). A pathway distinct from the mammalian unfolded protein response regulates expression of endoplasmic reticulum chaperones in non-stressed cells. EMBO J. 16, 7207–7216. Brodsky, J. L., and McCracken, A. A. (1999). ER protein quality control and proteasome-mediated protein degradation. Semin. Cell Dev. Biol. 10, 507–513. Bueler, H., Aguzzi, A., Sailer, A., Greiner, R. A., Autenried, P., Aguet, M., and Weissmann, C. (1993). Mice devoid of PrP are resistant to scrapie. Cell 73, 1339–1347. Calfon, M., Zeng, H., Urano, F., Till, J. H., Hubbard, S. R., Harding, H. P., Clark, S. G., and Ron, D. (2002). IRE1 couples endoplasmic reticulum load to secretory capacity by processing the XBP-1 mRNA. Nature 415, 92–96. Carrell, R. W., and Lomas, D. A. (2002). Alpha1-antitrypsin deWciency—a model for conformational diseases. N Engl J Med 346, 45–53. Chen, B. P., Wolfgang, C. D., and Hai, T. (1996). Analysis of ATF3, a transcription factor induced by physiological stresses and modulated by gadd153/Chop10. Mol. Cell. Biol. 16, 1157–1168. Chen, K. S., Manian, P., Koeuth, T., Potocki, L., Zhao, Q., Chinault, A. C., Lee, C. C., and Lupski, J. R. (1997). Homologous recombination of a Xanking repeat gene cluster is a mechanism for a common contiguous gene deletion syndrome. Nat. Genet. 17, 154–163. Colledge, W. H., Abella, B. S., Southern, K. W., RatcliV, R., Jiang, C., Cheng, S. H., MacVinish, L. J., Anderson, J. R., Cuthbert, A. W., and Evans, M. J. (1995). Generation and characterization of a delta F508 cystic Wbrosis mouse model. Nat Genet 10, 445–452. Cox, J. S., and Walter, P. (1996). A novel mechanism for regulating activity of a transcription factor that controls the unfolded protein response. Cell 87, 391–404. Crowther, D. C. (2002). Familial conformational diseases and dementias. Hum Mutat 20, 1–14.

SUMMARY

Cummings, C. J., Mancini, M. A., AntalVy, B., DeFranco, D. B., Orr, H. T., and Zoghbi, H. Y. (1998). Chaperone suppression of aggregation and altered subcellular proteasome localization imply protein misfolding in SCA1. Nat Genet 19, 148–154. Cummings, C. J., Reinstein, E., Sun, Y., AntalVy, B., Jiang, Y., Ciechanover, A., Orr, H. T., Beaudet, A. L., and Zoghbi, H. Y. (1999). Mutation of the E6-AP ubiquitin ligase reduces nuclear inclusion frequency while accelerating polyglutamine-induced pathology in SCA1 mice. Neuron 24, 879–692. Cummings, C. J., Sun, Y., Opal, P., AntalVy, B., Mestril, R., Orr, H. T., Dillmann, W. H., and Zoghbi, H. Y. (2001). Over-expression of inducible HSP70 chaperone suppresses neuropathology and improves motor function in SCA1 mice. Hum Mol Genet 10, 1511–1518. Cummings, C. J., and Zoghbi, H. Y. (2000). Trinucleotide repeats: Mechanisms and pathophysiology. Annu Rev Genomics Hum Genet 1, 281–328. D’Urso, D., Prior, R., Greiner-Petter, R., Gabreels-Festen, A. A., and Muller, H. W. (1998). Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutations of the peripheral myelin protein PMP22. J. Neurosci. 18, 731–740. Davis, R. L., Holohan, P. D., Shrimpton, A. E., Tatum, A. H., Daucher, J., Collins, G. H., Todd, R., Bradshaw, C., Kent, P., Feiglin, D., Rosenbaum, A., Yerby, M. S., Shaw, C. M., Lacbawan, F., and Lawrence, D. A. (1999a). Familial encephalopathy with neuroserpin inclusion bodies. Am J Pathol 155, 1901–1913. Davis, R. L., Shrimpton, A. E., Holohan, P. D., Bradshaw, C., Feiglin, D., Collins, G. H., Sonderegger, P., Kinter, J., Becker, L. M., Lacbawan, F., Krasnewich, D., Maenke, M., Lawrence, D. A., Yerby, M. S., Shaw, C. M., Gooptu, B., Elliot, P. R., Finch, J. T., Carrell, R. W., and Lomas, D. A. (1999b). Familial dementia caused by polymerization of mutant neuroserpin. Nature 401, 376–379. De Angelis, E., Watkins, A., Schafer, M., Brummendorf, T., and Kenwrick, S. (2002). Disease-associated mutations in L1 CAM interfere with ligand interactions and cell-surface expression. Hum Mol Genet 11, 1–12. DeGracia, D. J., Kumar, R., Owen, C. R., Krause, G. S., and White, B. C. (2002). Molecular pathways of protein synthesis inhibition during brain reperfusion: Implications for neuronal survival or death. J Cereb Blood Flow Metab 22, 127–141. Delepine, M., Nicolino, M., Barrett, T., Golamaully, M., Lathrop, G. M., and Julier, C. (2000). EIF2AK3, encoding translation initiation factor 2-alpha kinase 3, is mutated in patients with Wolcott-Rallison syndrome. Nat Genet 25, 406–409. Denning, G. M., Anderson, M. P., Amara, J. F., Marshall, J., Smith, A. E., and Welsh, M. J. (1992). Processing of mutant cystic Wbrosis transmembrane conductance regulator is temperature-sensitive. Nature 358, 761–764. Dobson, C. M. (2001). The structural basis of protein folding and its links with human disease. Philos Trans R Soc Lond B Biol Sci 356, 133–145. Duncan, I. D. (1990). Dissection of the phenotype and genotype of the X-linked myelin mutants. In ‘‘Myelination and Dysmyelination’’ (I. D. Duncan, R. P. SkoV, and D. R. Colman, eds.), Vol. 605 , pp. 110–121. New York Academy of Sciences, New York. Fawcett, T. W., Martindale, J. L., Guyton, K. Z., Hai, T., and Holbrook, N. J. (1999). Complexes containing activating transcription factor (ATF)/cAMP-responsive-element-binding protein (CREB) interact with the CCAAT/enhancer-binding protein (C/EBP)-ATF composite site to regulate Gadd153 expression during the stress response. Biochem J 339, 135–141. Ferrari, D. M., and Soling, H. D. (1999). The protein disulphide-isomerase family: Unravelling a string of folds. Biochem J 339, 1–10. Fischer, H., Koenig, U., Eckhart, L., and Tschachler, E. (2002). Human caspase 12 has acquired deleterious mutations. Biochem Biophys Res Commun 293, 722–726. Fransen, E., Lemmon, V., Van Camp, G., Vits, L., Coucke, P., and Willems, P. J. (1995). CRASH syndrome: Clinical spectrum of corpus callosum hypoplasia, retardation, adducted thumbs, spastic paraparesis and hydrocephalus due to mutations in one single gene, L1. Eur J Hum Genet 3, 273–284. Garbern, J., Cambi, F., Shy, M., and Kamholz, J. (1999). The molecular pathogenesis of Pelizaeus-Merzbacher disease. Arch. Neurol. 56, 1210–1214. Garbern, J. Y., Cambi, F., Tang, X. M., Sima, A. A., Vallat, J. M., Bosch, E. P., Lewis, R., Shy, M., Sohi, J., Kraft, G., Chen, K. L., Joshi, I., Leonard, D. G., Johnson, W., Raskind, W., Dlouhy, S.R., Pratt, V., Hodes, M.E., Bird, T., and Kamholz, J. (1997). Proteolipid protein is necessary in peripheral as well as central myelin. Neuron 19, 205–218. Garbern, J. Y., Yool, D. A., Moore, G. J., Wilds, I. B., Faulk, M. W., Klugmann, M., Nave, K. A., Sistermans, E. A., van der Knaap, M. S., Bird, T. D., Shy, M. E., Kamholz, J. A., and Griffiths, I. R. (2002). Patients lacking the major CNS myelin protein, proteolipid protein 1, develop length-dependent axonal degeneration in the absence of demyelination and inXammation. Brain 125, 551–561. GaroV, H. (1985). Using recombinant DNA techniques to study protein targeting in the eukaryotic cell. Annu. Rev. Cell Biol. 1, 403–445. Gelman, M. S., and Kopito, R. R. (2002). Rescuing protein conformation: Prospects for pharmacological therapy in cystic Wbrosis. J Clin Invest 110, 1591–1597. Gencic, S., Abuelo, D., Ambler, M., and Hudson, L. D. (1989). Pelizaeus-Merzbacher disease: An X-linked neurologic disorder of myelin metabolism with a novel mutation in the gene encoding proteolipid protein. Am. J. Hum. Genet. 45, 435–442. Gething, M.-J., and Sambrook, J. (1992). Protein folding in the cell. Nature 355, 33–45. Ghandour, M., and SkoV, R. (1988). Expression of galactocerebroside in developing normal and jimpy oligodendrocytes in situ. J. Neurocytol. 17, 485–498.

1031

1032

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

Gow, A., Friedrich, V. L., and Lazzarini, R. A. (1994a). Intracellular transport and sorting of the oligodendrocyte transmembrane proteolipid protein. J. Neurosci. Res. 37, 563–573. Gow, A., Friedrich, V. L., and Lazzarini, R. A. (1994b). Many naturally occurring mutations of myelin proteolipid protein impair its intracellular transport. J. Neurosci. Res. 37, 574–583. Gow, A., Gragerov, A., Gard, A., Colman, D. R., and Lazzarini, R. A. (1997). Conservation of topology, but not conformation, of the proteolipid proteins of the myelin sheath. J. Neurosci. 17, 181–189. Gow, A., and Lazzarini, R. A. (1996). A cellular mechanism governing the severity of Pelizaeus-Merzbacher disease. Nat. Genet. 13, 422–428. Gow, A., Southwood, C. M., and Lazzarini, R. A. (1998). Disrupted proteolipid protein traYcking results in oligodendrocyte apoptosis in an animal model of Pelizaeus-Merzbacher disease. J. Cell Biol. 140, 925–934. Graham, K., Le, A., and Sifers, R. (1990). Accumulation of the insoluble PiZ variant of human a1-antitrypsin within the hepatic endoplasmic reticulum does not elevate the steady-state level of grp78/BiP. J. Biol. Chem. 265, 20463–20468. Grinspan, J. B., Coulalaglou, M., Beesley, J. S., Carpio, D. F., and Scherer, S. S. (1998). Maturation-dependent apoptotic cell death of oligodendrocytes in myelin-deWcient rats. J. Neurosci. Res. 54, 623–634. Hammond, C., and Helenius, A. (1995). Quality control in the secretory pathway. Curr. Op. Cell Biol. 7, 523–529. Harding, H. P., Calfon, M., Urano, F., Novoa, I., and Ron, D. (2002). Transcriptional and Translational Control in the Mammalian Unfolded Protein Response. Annu Rev Cell Dev Biol 18, 18. Harding, H. P., Novoa, I., Zhang, Y., Zeng, H., Wek, R., Schapira, M., and Ron, D. (2000). Regulated translation initiation controls stress-induced gene expression in mammalian cells. Mol. Cell 6, 1099–1108. Harding, H. P., Zhang, Y., and Ron, D. (1999). Protein translation and folding are coupled by an endoplasmic reticulum-resident kinase. Nature 397, 271–274. Hardy, J., and Allsop, D. (1991). Amyloid deposition as the central event in the aetiology of Alzheimer’s disease. Trends Pharmacol Sci 12, 383–388. Hegde, R. S., Mastrianni, J. A., Scott, M. R., DeFea, K. A., Tremblay, P., Torchia, M., DeArmond, S. J., Prusiner, S. B., and Lingappa, V. R. (1998). A transmembrane form of the prion protein in neurodegenerative disease. Science 279, 827–834. Hegde, R. S., Tremblay, P., Groth, D., DeArmond, S. J., Prusiner, S. B., and Lingappa, V. R. (1999). Transmissible and genetic prion diseases share a common pathway of neurodegeneration. Nature 402, 822–826. Hobson, G. M., Davis, A. P., Stowell, N. C., Kolodny, E. H., Sistermans, E. A., de Coo, I. F., Funanage, V. L., and Marks, H. G. (2000). Mutations in noncoding regions of the proteolipid protein gene in PelizaeusMerzbacher disease. Neurology 55, 1089–96. Hudson, L. D., and Nadon, N. L. (1992). Amino acid substitutions in proteolipid protein that cause dysmyelination. In ‘‘Myelin: Biology and Chemistry’’ (R. E. Martenson, ed.), pp. 677–702. CRC Press, Boca Raton, FL. Hudson, L. D., Puckett, C., Berndt, J., Chan, J., and Gencic, S. (1989). Mutation of the proteolipid protein gene PLP in a human X chromosome-linked myelin disorder. Proc. Natl. Acad. Sci. U. S. A. 86, 8128–8131. Humphries, M. M., Rancourt, D., Farrar, G. J., Kenna, P., Hazel, M., Bush, R. A., Sieving, P. A., Sheils, D. M., McNally, N., Creighton, P., Erven, A., Boros, A., Gulya, K., Capecchi, M. R., and Humphries, P. (1997). Retinopathy induced in mice by targeted disruption of the rhodopsin gene. Nat. Genet. 15. Imai, Y., Soda, M., Inoue, H., Hattori, N., Mizuno, Y., and Takahashi, R. (2001). An unfolded putative transmembrane polypeptide, which can lead to endoplasmic reticulum stress, is a substrate of Parkin. Cell 105, 891–902. Inoue, Y., Kagawa, T., Matsumura, Y., Ikenaka, K., and Mikoshiba, K. (1996). Cell death of oligodendrocytes or demyelination induced by overexpression of proteolipid protein depending on expressed gene dosage. Neurosci. Res. 25, 161–172. Jensen, T. J., Loo, M. A., Pind, S., Williams, D. B., Goldberg, A. L., and Riordan, J. R. (1995). Multiple proteolytic systems, including the proteasome, contribute to CFTR processing. Cell 83, 129–135. Johnston, J. A., Ward, C. L., and Kopito, R. R. (1998). Aggresomes: A cellular response to misfolded proteins. J. Cell Biol. 143, 1883–1898. Jousse, C., Bruhat, A., Carraro, V., Urano, F., Ferrara, M., Ron, D., and Fafournoux, P. (2001). Inhibition of CHOP translation by a peptide encoded by an open reading frame localized in the chop 5’UTR. Nucleic Acids Res 29, 4341–4351. Jung, M., Sommer, I., Schachner, M., and Nave, K. A. (1996). Monoclonal antibody O10 deWnes a conformationally sensitive cell-surface epitope of proteolipid protein (PLP): Evidence that PLP misfolding underlies dysmyelination in mutant mice. J. Neurosci. 16, 7920–7929. Kagawa, T., Ikenaka, K., Inoue, Y., Kuriyama, S., Tsujii, T., Nakao, J., Nakajima, K., Aruga, J., Okano, H., and Mikoshiba, K. (1994). Glial cell degeneration and hypomyelination caused by overexpression of myelin proteolipid protein gene. Neuron 13, 427–442. Kalin, N., Claass, A., Sommer, M., Puchelle, E., and Tummler, B. (1999). DeltaF508 CFTR protein expression in tissues from patients with cystic Wbrosis. J Clin Invest 103, 1379–1389. Kalwy, S. A., Smith, R., and Kidd, G. J. (1997). Myelin proteolipid protein expressed in COS-1 cells is targeted to actin-associated surfaces. J Neurosci Res 48, 201–211. Kartner, N., Augustinas, O., Jensen, T. J., Naismith, A. L., and Riordan, J. R. (1992). Mislocalization of delta F508 CFTR in cystic Wbrosis sweat gland. Nat Genet 1, 321–327. Katayama, T., Imaizumi, K., Sato, N., Miyoshi, K., Kudo, T., Hitomi, J., Morihara, T., Yoneda, T., Gomi, F., Mori, Y., Nakane, Y., Takeda, J., Tsuda, T., Itoyama, Y., Murayama, O., Takashima, A., St. George-Hyslop, P., Takeda, M., and Tohyama, M. (1999). Presenilin-1 mutations downregulate the signalling pathway of the unfolded-protein response. Nat Cell Biol 1, 479–485.

SUMMARY

Kaufman, R. J., Scheuner, D., Schroder, M., Shen, X., Lee, K., Liu, C. Y., and Arnold, S. M. (2002). The unfolded protein response in nutrient sensing and diVerentiation. Nat Rev Mol Cell Biol 3, 411–421. Kimball, S. R. (1999). Eukaryotic initiation factor eIF2. Int. J. Biochem. Cell Biol. 31, 25–29. Klausner, R. D., and Sitia, R. (1990). Protein degradation in the endoplasmic reticulum. Cell 62, 611–614. Klement, I. A., Skinner, P. J., Kaytor, M. D., Yi, H., Hersch, S. M., Clark, H. B., Zoghbi, H. Y., and Orr, H. T. (1998). Ataxin-1 nuclear localization and aggregation: Role in polyglutamine-induced disease in SCA1 transgenic mice. Cell 95, 41–53. Klugmann, M., Schwab, M. H., Puhlhofer, A., Schneider, A., Zimmermann, F., GriYths, I. R., and Nave, K.-A. (1997). Assembly of CNS myelin in the absence of proteolipid protein. Neuron 18, 59–70. Knapp, P., SkoV, R., and Redstone, D. (1986). Oligodendroglial cell death in jimpy mice: An explanation for the myelin deWcit. J. Neurosci. 6, 2813–2822. Kobayashi, H., HoVman, E. P., and Marks, H. G. (1994). The rumpshaker mutation in spastic paraplegia. Nat. Genet. 7, 351–352. Kopito, R. R., and Sitia, R. (2000). Aggresomes and Russell bodies. Symptoms of cellular indigestion? EMBO Rep 1, 225–231. Kozutsumi, Y., Segal, M., Normington, K., Gething, M. J., and Sambrook, J. (1988). The presence of malfolded proteins in the endoplasmic reticulum signals the induction of glucose-regulated proteins. Nature 332, 462–464. Kruger, R., Eberhardt, O., Riess, O., and Schulz, J. B. (2002). Parkinson’s disease: One biochemical pathway to Wt all genes? Trends Mol Med 8, 236–240. Le, A., Steiners, J. L., Ferrell, G. A., Shaker, J. C., and Sifers, R. N. (1994). Association between calnexin and a secretion-incompetent variant of human a1-antitrypsin. J. Biol. Chem. 269, 7514–7519. Lee, A. S. (2001). The glucose-regulated proteins: Stress induction and clinical applications. Trends Biochem Sci 26, 504–510. Lee, K., Tirasophon, W., Shen, X., Michalak, M., Prywes, R., Okada, T., Yoshida, H., Mori, K., and Kaufman, R. J. (2002). IRE1-mediated unconventional mRNA splicing and S2P-mediated ATF6 cleavage merge to regulate XBP1 in signaling the unfolded protein response. Genes Dev 16, 452–466. Lee, V. M., Goedert, M., and Trojanowski, J. Q. (2001). Neurodegenerative tauopathies. Annu Rev Neurosci 24, 1121–1159. Leegwater, P. A., Vermeulen, G., Konst, A. A., Naidu, S., Mulders, J., Visser, A., Kersbergen, P., Mobach, D., Fonds, D., van Berkel, C. G., Lemmers, R. J., Frants, R. R., Oudejans, C. B., Schutgens, R. B., Prank, J. C., and vander Knapp, M. S. (2001). Subunits of the translation initiation factor eIF2B are mutant in leukoencephalopathy with vanishing white matter. Nat Genet 29, 383–388. Lipsitz, D., Goetz, B. D., and Duncan, I. D. (1998). Apoptotic glial cell death and kinetics in the spinal cord of the myelin-deWcient rat. J. Neurosci. Res. 51, 497–507. Lomas, D. A., and Carrell, R. W. (2002). Serpinopathies and the conformational dementias. Nat Rev Genet 3, 759–768. Ma, J., and Lindquist, S. (2002). Conversion of PrP to a Self-Perpetuating PrPSc-like Conformation in the Cytosol. Science 298, 1785–1788. Ma, J., Wollmann, R., and Lindquist, S. (2002a). Neurotoxicity and Neurodegeneration When PrP Accumulates in the Cytosol. Science 298, 1781–1785. Ma, Y., Brewer, J. W., Diehl, J. A., and Hendershot, L. M. (2002b). Two distinct stress signaling pathways converge upon the CHOP promoter during the mammalian unfolded protein response. J Mol Biol 318, 1351–1365. Ma, Y., and Hendershot, L. M. (2001). The unfolding tale of the unfolded protein response. Cell 107, 827–30. Macklin, W. B., Trapp, B. D., Kagawa, T., and Ikenaka, K. (1995). Hypomyelination in transgenic mice with multiple copies of the myelin proteolipid protein gene. J. Neurochem. 64, S86. Mastronardi, F. G., Ackerley, C. A., Arsenault, L., Roots, B. I., and Moscarello, M. A. (1993). Demyelination in a transgenic mouse: A model for multiple sclerosis. J. Neurosci. Res. 36, 315–324. Mori, K., Ma, W., Gething, M.-J., and Sambrook, J. (1993). A transmembrane protein with a cdc2þ/CDC28related kinase activity is required for signalling from the ER to the nucleus. Cell 74, 743–756. Moulding, H. D., Martuza, R. L., and Rabkin, S. D. (2000). Clinical mutations in the L1 neural cell adhesion molecule aVect cell-surface expression. J Neurosci 20, 5696–5702. Naef, R., Adlkofer, K., Lescher, B., and Suter, U. (1997). Aberrant protein traYcking in Trembler suggests a disease mechanism for hereditary human peripheral neuropathies. Mol. Cell. Neurosci. 9, 13–25. Nave, K.-A., and BoespXug-Tanguy, O. (1996). X-linked developmental defects of myelination: From mouse mutants to human genetic diseases. The Neuroscientist 2, 33–43. Niwa, M., Sidrauski, C., Kaufman, R. J., and Walter, P. (1999). A role for presenilin-1 in nuclear accumulation of Ire1 fragments and induction of the mammalian unfolded protein response. Cell 99, 691–702. Novoa, I., Zeng, H., Harding, H. P., and Ron, D. (2001). Feedback inhibition of the unfolded protein response by GADD34-mediated dephosphorylation of eIF2alpha. J Cell Biol 153, 1011–1022. Nussbaum, J. L., and Roussel, G. (1983). Immunocytochemical demonstration of the transport of myelin proteins through the Golgi apparatus. Cell Tissue Res. 234, 547–559. Nussbaum, J.-L., Roussel, G., Wunsch, E., and Jolles, P. (1985). Site-speciWc antibodies to rat myelin proteolipids directed against the C-terminal hexapeptide. J. Neurol. Sci. 68, 89–100. Okada, T., Yoshida, H., Akazawa, R., Negishi, M., and Mori, K. (2002). Distinct Roles of ATF6 and PERK in Transcription during the Mammalian Unfolded Protein Response. Biochem J 16.

1033

1034

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

Olsson, J. E., Gordon, J. W., Pawlyk, B. S., Roof, D., Hayes, A., Molday, R. S., Mukai, S., Cowley, G. S., Berson, E. L., and Dryja, T. P. (1992). Transgenic mice with a rhodopsin mutation (pro23his): A mouse model of autosomal dominant retinitis pigmentosa. Neuron 9, 815–830. Ordway, J. M., Tallaksen-Greene, S., Gutekunst, C. A., Bernstein, E. M., Cearley, J. A., Wiener, H. W., Dure, L. S. t., Lindsey, R., Hersch, S. M., Jope, R. S., Albin, R. L., and Detloff, P. J. (1997). Ectopically expressed CAG repeats cause intranuclear inclusions and a progressive late onset neurological phenotype in the mouse. Cell 91, 753–763. Orr, H. T., and Zoghbi, H. Y. (2001). SCA1 molecular genetics: A history of a 13 year collaboration against glutamines. Hum Mol Genet 10, 2307–2311. Pearsall, G. B., Nadon, N. L., Wolf, M. K., and Billings-Gagliardi, S. (1997). Jimpy-4J mouse has a missense mutation in exon 2 of the Plp gene. Dev. Neurosci. 19, 337–341. Pepys, M. B. (2001). Pathogenesis, diagnosis and treatment of systemic amyloidosis. Philos Trans R Soc Lond B Biol Sci 356, 203–10; discussion 210–211. Power, C., Kong, P., and Trapp, B. (1996). Major histocompatibility class I expression in oligodendrocytes induces hypomyelination in transgenic mice. J. Neurosci. Res. 44, 165–173. Prusiner, S. B. (1998). Prions. Proc Natl Acad Sci U S A 95, 13363–13383. Raskind, W. H., Williams, C. A., Hudson, L. D., and Bird, T. D. (1991). Complete deletion of the proteolipid protein gene (PLP) in a family with X-linked Pelizaeus-Merzbacher disease. Am. J. Hum. Genet. 49, 1355–1360. Readhead, C., Schneider, A., GriYths, I. R., and Nave, K.-A. (1994). Premature arrest of myelin formation in transgenic mice with increased proteolipid protein gene dosage. Neuron 12, 583–595. Reddy, P. H., Williams, M., Charles, V., Garrett, L., Pike-Buchanan, L., Whetsell, W. O., Jr., Miller, G., and Tagle, D. A. (1998). Behavioural abnormalities and selective neuronal loss in HD transgenic mice expressing mutated full-length HD cDNA. Nat Genet 20, 198–202. Reiter, L. T., Murakami, T., Koeuth, T., Pentao, L., Muzny, D. M., Gibbs, R. A., and Lupski, J. R. (1996). A recombination hotspot responsible for two inherited peripheral neuropathies is located near a mariner transposon-like element. Nat. Genet. 12, 288–297. Rose, J. K., and Doms, R. W. (1988). Regulation of protein export from the endoplasmic reticulum. Annu. Rev. Cell Biol. 4, 257–288. Roussel, G., Meskovic, N. M., TriWlieV, E., Artault, J.-C., and Nussbaum, J.-L. (1987). Arrest of proteolipid transport through the Golgi apparatus in Jimpy brain. J. Neurocytol. 16, 195–204. Runker, A. E., Bartsch, U., Nave, K. A., and Schachner, M. (2003). The C264Y missense mutation in the extracellular domain of L1 impairs protein traYcking in vitro and in vivo. J Neurosci 23, 277–286. Rutishauser, J., and Spiess, M. (2002). Endoplasmic reticulum storage diseases. Swiss Med Wkly 132, 211–222. Ryu, E. J., Harding, H. P., Angelastro, J. M., Vitolo, O. V., Ron, D., and Greene, L. A. (2002). Endoplasmic reticulum stress and the unfolded protein response in cellular models of Parkinson’s disease. J Neurosci 22, 10690–10698. Sanders, C. R., Ismail-Beigi, F., and McEnery, M. W. (2001). Mutations of peripheral myelin protein 22 result in defective traYcking through mechanisms which may be common to diseases involving tetraspan membrane proteins. Biochemistry 40, 9453–9459. Sato, N., Urano, F., Yoon Leem, J., Kim, S. H., Li, M., Donoviel, D., Bernstein, A., Lee, A. S., Ron, D., Veselits, M. L., Sisoda, S. S., and Thinakaran, G. (2000). Upregulation of BiP and CHOP by the unfolded-protein response is independent of presenilin expression. Nat. Cell Biol. 2, 863–870. Sato, S., Ward, C. L., Krouse, M. E., Wine, J. J., and Kopito, R. R. (1996). Glycerol reverses the misfolding phenotype of the most common cystic Wbrosis mutation. J Biol Chem 271, 635–638. SchiVmann, R., and BoespXug-Tanguy, O. (2001). An update on the leukodsytrophies. Curr Opin Neurol 14, 789–794. Schneider, A., GriYths, I., Readhead, C., and Nave, K.-A. (1995). Dominant-negative action of the jimpy mutation in mice complemented with an autosomal transgene for myelin proteolipid protein. Proc. Natl. Acad. Sci. 92, 4447–4451. Schneider, A., Montague, P., GriYths, I., Fanarraga, M., Kennedy, P., Brophy, P., and Nave, K.-A. (1992). Uncoupling of hypomyelination and glial cell death by a mutation in the proteolipid protein gene. Nature 358, 758–761. Schwob, V. S., Clark, H. B., Agrawal, D., and Agrawal, H. C. (1985). Electron microscopic immunocytochemical localization of myelin proteolipid protein and myelin basic protein to oligodendrocytes in rat brain during myelination. J. Neurochem. 45, 559–571. Shen, J., Chen, X., Hendershot, L., and Prywes, R. (2002). ER stress regulation of ATF6 localization by dissociation of BiP/GRP78 binding and unmasking of Golgi localization signals. Dev Cell 3, 99–111. Shimura, H., Hattori, N., Kubo, S., Mizuno, Y., Asakawa, S., Minoshima, S., Shimizu, N., Iwai, K., Chiba, T., Tanaka, K., and Suzuki, T. (2000). Familial Parkinson disease gene product, parkin, is a ubiquitin-protein ligase. Nat Genet 25, 302–305. Shy, M., Hobson, G., BoespXug-Tanguy, O., Garbern, J., Sperle, K., Jain, M., Li, W., Gow, A., Rodriguez, D., Bertini, E., Mancias, P., Krajewski, K., Lewis, R. A., Kamholz, J., and members of the ENEDD (2003). Schwann cell expression of PLP1 but not its alternatively spliced isoform DM20 is necessary to prevent demyelinating peripheral neuropathy. Ann Neurol (in press).

SUMMARY

Simons, M., Kramer, E. M., Macchi, P., Rathke-Hartlieb, S., Trotter, J., Nave, K. A., and Schulz, J. B. (2002). Overexpression of the myelin proteolipid protein leads to accumulation of cholesterol and proteolipid protein in endosomes/lysosomes: Implications for Pelizaeus-Merzbacher disease. J Cell Biol 157, 327–336. Sinoway, M. P., Kitagawa, K., Timsit, S., Hashim, G. A., and Colman, D. R. (1994). Proteolipid protein interactions in transfectants: Implications for myelin assembly. J. Neurosci. Res. 37, 551–562. Sistermans, E. A., de Wijs, I. J., de Coo, R. F. M., Smit, L. M. E., Menko, F. H., and van Oost, B. A. (1996). A (G-to-A) mutation in the initiation codon of the proteolipid protein gene causing a relatively mild form of Pelizaeus-Merzbacher disease in a Dutch family. Hum. Genet. 97, 337–339. SkoV, R. P., and Knapp, P. E. (1992). Phenotypic expression of X-linked genetic defects aVecting myelination. In ‘‘Myelin: Biology and Chemistry’’ (R. E. Martenson, ed.), pp. 653–676. CRC Press, Boca Raton, FL. Southwood, C. M., Garbern, J., Jiang, W., and Gow, A. (2002). The unfolded protein response modulates disease severity in Pelizaeus-Merzbacher disease. Neuron 36, 585–596. Southwood, C. M., and Gow, A. (2001). Molecular mechanisms of disease stemming from mutations in the proteolipid protein gene. Micros. Res. Tech. 52, 700–708. Stecca, B., Southwood, C. M., Gragerov, A., Kelley, K. A., Friedrich, V. L. J., and Gow, A. (2000). The evolution of lipophilin genes from invertebrates to tetrapods: DM-20 cannot replace PLP in CNS myelin. J. Neurosci. 20, 4002–4010. Thomson, C. E., Montague, P., Jung, M., Nave, K.-A., and GriYths, I. R. (1997). Phenotypic severity of murine Plp mutants reXects in vivo and in vitro variations in transport of PLP isoproteins. Glia 20, 322–332. Timsit, S. G., Bally-Cuif, L., Colman, D. R., and Zalc, B. (1992). DM-20 mRNA is expressed during the embryonic development of the nervous system of the mouse. J. Neurochem. 58, 1172–1175. Tobler, A. R., Notterpek, L., Naef, R., Taylor, V., Suter, U., and Shooter, E. M. (1999). Transport of Trembler-J mutant peripheral myelin protein 22 is blocked in the intermediate compartment and aVects the transport of the wild-type protein by direct interaction. J. Neurosci. 19, 2027–2036. Trojanowski, J. Q. (2002). Tauists, Baptists, Syners, Apostates, and new data. Ann Neurol 52, 263–265. Turnley, A. M., Morahan, G., Okano, H., Bernard, O., Mikoshiba, K., Allison, J., Bartlett, P. F., and Miller, J. A. F. P. (1991). Dysmyelination in transgenic mice resulting from expression of class I histocompatibility molecules in oligodendrocytes. Nature 353, 566–569. Ubeda, M., Wang, X.-Z., Zinszner, H., Wu, I., Habener, J., and Ron, D. (1996). Stress-induced binding of transcription factor CHOP to a novel DNA-control element. Mol. Cell. Biol. 16, 1479–1489. Uptain, S. M., and Lindquist, S. (2002). Prions as protein-based genetic elements. Annu Rev Microbiol 56, 703–41. van der Knaap, M. S., Barth, P. G., Gabreels, F. J., Franzoni, E., Begeer, J. H., Stroink, H., Rotteveel, J. J., and Valk, J. (1997). A new leukoencephalopathy with vanishing white matter. Neurology 48, 845–855. van der Knaap, M. S., Leegwater, P. A., Konst, A. A., Visser, A., Naidu, S., Oudejans, C. B., Schutgens, R. B., and Pronk, J. C. (2002). Mutations in each of the Wve subunits of translation initiation factor eIF2B can cause leukoencephalopathy with vanishing white matter. Ann Neurol 51, 264–270. Vela, J. M., Gonzalez, B., and Castellano, B. (1998). Understanding glial abnormalities associated with myelin deWciency in the jimpy mutant mouse. Brain Res Brain Res Rev 26, 29–42. Walsh, D. M., Klyubin, I., Fadeeva, J. V., Cullen, W. K., Anwyl, R., Wolfe, M. S., Rowan, M. J., and Selkoe, D. J. (2002). Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature 416, 535–539. Wang, X.-Z., Harding, H. P., Zhang, Y., Jolicoeur, E. M., Kuroda, M., and Ron, D. (1998a). Cloning of mammalian Ire1 reveals diversity in the ER stress responses. EMBO J. 17, 5708–5717. Wang, X.-Z., Kuroda, M., Sok, J., Batchvarova, N., Kimmel, R., Chung, P., Zinszner, H., and Ron, D. (1998b). IdentiWcation of novel stress-induced genes downstream of chop. EMBO J. 17, 3619–3630. Wang, X.-Z., Lawson, B., Brewer, J. W., Zinszner, H., Sanjay, A., Mi, L.-J., Boorstein, R., Kreibich, G., Hendershot, L. M., and Ron, D. (1996). Signals from the stressed endoplasmic reticulum induce C/EBPhomologous protein (CHOP/GADD153). Mol. Cell. Biol. 16, 4273–4280. Wang, X.-Z., and Ron, D. (1996). Stress-induced phosphorylation and activation of the transcription factor CHOP (GADD153) by p38 MAP Kinase. Science 272, 1347–1349. Wechselberger, C., Wurm, S., Pfarr, W., and Hoglinger, O. (2002). The physiological functions of prion protein. Exp Cell Res 281, 1–8. Wolfgang, C. D., Chen, B. P., Martindale, J. L., Holbrook, N. J., and Hai, T. (1997). gadd153/Chop10, a potential target gene of the transcriptional repressor ATF3. Mol. Cell. Biol. 17, 6700–6707. Wu, Y., Whitman, I., Molmenti, E., Moore, K., Hippenmeyer, P., and Perlmutter, D. H. (1994). A lag in intracellular degradation of mutant a1-antitrypsin correlates with the liver disease phenotype in homozygous PiZZ a1-antitrypsin deWciency. Proc. Natl. Acad. Sci. USA 91, 9014–9018. Yamada, M., Ivanova, A., Yamaguchi, Y., Lees, M. B., and Ikenaka, K. (1999). Proteolipid protein gene product can be secreted and exhibit biological activity during early development. J. Neurosci. 19, 2143–2151. Yamada, M., Jung, M., Tetsushi, K., Ivanova, A., Nave, K. A., and Ikenaka, K. (2001). Mutant Plp/DM20 cannot be processed to secrete PLP-related oligodendrocyte diVerentiation/survival factor. Neurochem Res 26, 639–645. Yamamoto, T., Nanba, E., Zhang, H., Sasaki, M., Komaki, H., and Takeshita, K. (1998). Jimpy(msd) mouse mutation and connatal Pelizaeus-Merzbacher disease. Am. J. Med. Genet. 75, 439–440. Ye, J., Rawson, R. B., Komuro, R., Chen, X., Dave, U. P., Prywes, R., Brown, M. S., and Goldstein, J. L. (2000). ER stress induces cleavage of membrane-bound ATF6 by the same proteases that process SREBPs. Mol Cell 6, 1355–1364.

1035

1036

42. PROTEIN MISFOLDING AS A DISEASE DETERMINANT

Yoshida, H., Haze, K., Yanagi, H., Yura, T., and Mori, K. (1998). IdentiWcation of the cis-acting endoplasmic reticulum stress response element responsible for transcriptional induction of mammalian glucose-regulated proteins. Involvement of basic leucine zipper transcription factors. J. Biol. Chem. 273, 33741–33749. Yoshida, H., Matsui, T., Yamamoto, A., Okada, T., and Mori, K. (2001). XBP1 mRNA Is Induced by ATF6 and Spliced by IRE1 in Response to ER Stress to Produce a Highly Active Transcription Factor. Cell 107, 881–891. Yoshida, H., Okada, T., Haze, K., Yanagi, H., Yura, T., Negishi, M., and Mori, K. (2000). ATF6 activated by proteolysis binds in the presence of NF-Y (CBF) directly to the cis-acting element responsible for the mammalian unfolded protein response. Mol Cell Biol 20, 6755–6767. Yoshioka, T., Feigenbaum, L., and Jay, G. (1991). Transgenic mouse model for central nervous system demyelination. Mol. Cell. Biol. 11, 5479–5486. Zhao, Q., Wang, J., Levichkin, I. V., Stasinopoulos, S., Ryan, M. T., and Hoogenraad, N. J. (2002). A mitochondrial speciWc stress response in mammalian cells. Embo J 21, 4411–4419. Zinszner, H., Kuroda, M., Wang, X., Batchvarova, N., Lightfoot, R. T., Remotti, H., Stevens, J. L., and Ron, D. (1998). CHOP is implicated in programmed cell death in response to impaired function of the endoplasmic reticulum. Genes Dev. 12, 982–995.

C H A P T E R

43 Experimental Autoimmune Encephalomyelitis Hans Lassmann

INTRODUCTION The concept that brain inXammation can be induced by an autoimmune reaction against the central nervous system was born from clinical observations made at the turn of the 20th century in the course of testing the safety and eYcacy of diVerent vaccinations (Remlinger, 1928). Immunization with certain vaccines, in particular with rabies vaccines, which were produced by the propagation of the infectious agents in nervous tissue, in rare instances were associated with an acute inXammatory central nervous system (CNS) disease—postvaccinal leucoencephalomyelitis. Based on these observations, Rivers et al. (1933) immunized monkeys by repeated intramuscular injections of sterile extracts of normal rabbit brain tissue. They observed after 46 to 85 such injections the development of a chronic inXammatory demyelinating diseases of the central nervous system, which in its essential features resembled the lesions present in multiple sclerosis. Although these studies established the concept of autoimmunity as a mechanism to induce inXammatory demyelinating brain disease, the practical use of this experimental model was limited due to species restrictions and poor reproducibility. This problem was overcome by the introduction of potent immune stimulating adjuvants, such as Freund’s complete adjuvans (Freund et al., 1950). With these tools the disease could be induced in nearly all vertebrate species in a highly reproducible manner. These models, which were then denominated ‘‘experimental autoimmune encephalomyelitis (EAE),’’ became the most important tools to study brain inXammation in general or the immunological mechanisms of organ-speciWc autoimmunity. They further were instrumental to unravel the mechanisms, leading to immune-mediated tissue damage in the central nervous system. The pathology in the central nervous system in autoimmune encephalomyelitis reXects an inXammatory demyelinating disease, which in many respects is similar to its human counterparts, such as acute disseminated leucoencephalomyelitis, acute haemorhagic leucoencephalomyelitis, and multiple sclerosis (Alvord, 1970; Lassmann, 1983; Raine et al., 1974). The inXammatory reaction is dominated by a mononuclear cell inWltrate, mainly consisting of T lymphocytes and macrophages accompanied by microglia activation. Pathological alterations are concentrated in the white matter (leucoencephalitis), although gray matter areas can be involved too. The essential structural lesions in the CNS are conXuent areas (plaques) of primary demyelination with partial axonal preservation. However, axonal injury and destruction is regularly found in all areas of demyelination. The lesions are further characterized by the formation of a dense astrocytic scar, called reactive gliosis. Parallel to demyelination a variable degree of remyelination can be found,

Myelin Biology and Disorders, Volume 2

1039

Copyright 2004, Elsevier (USA). All rights reserved.

1040

43. MYELIN AND ITS DISEASES

which can lead to complete restoration of the demyelinated tissue. A detailed comparison of such EAE lesions with those found in acute disseminated leucoencephalitis, and multiple sclerosis revealed that most essential pathological features of human inXammatory demyelinating disease are closely reXected in these EAE models (Lassmann, 1983).

BASIC MECHANISMS OF AUTOIMMUNE-MEDIATED INFLAMMATORY BRAIN DISEASE Since sensitization with nervous system tissue leads to an autoimmune disease, which is restricted to the central or the peripheral nervous system, it became clear already at an early stage that speciWc immune reactions, either mediated by T lymphocytes or antibodies, have to be involved in its pathogenesis. Classical work by Philip Paterson (1960) provided Wrst clear evidence that the disease is basically mediated by T lymphocytes. T cells, isolated from actively sensitized animals with autoimmune encephalomyelitis, induced a similar disease, when passively transferred to normal naı¨ve recipients, provided the donor cells and recipient animals are matched by their histocompatibility antigens. In contrast, intravenous transfer of serum or puriWed antibodies, derived from sensitized and diseased animals, did not induce disease (Chase, 1959). A major breakthrough in the study of T-cell immunology in EAE was the development of a technique, which allowed to isolate and propagate monospeciWc T-cell lines or clones, which were able to transfer inXammatory brain disease to naı¨ve recipients, which was Wrst established in rats (Ben Nun et al., 1981) and later in mice (Zamvil et al., 1985). This allowed for the Wrst time an in depth analysis of the antigen-speciWcity and the immunological properties of T cells, responsible for the initiation and propagation of autoimmune CNS inXammation. There is, however, a major diVerence in the pathology of the CNS lesions between EAE, induced by passive transfer of T lymphocytes and that present in animals, which have been actively sensitized with whole CNS tissue or myelin. While in the latter the inXammatory process is associated with widespread primary demyelination or tissue damage, passive transfer of autoreactive T cells in general leads to an inXammatory brain disease with little or absent demyelination (Fig. 43.1). Thus, in actively induced EAE additional immune responses seem to be involved which modify the basic T-cell-mediated inXammatory process. A major component, which is involved in this modiWcation, are antibodies, directed against antigens, which are exposed in the extracellular space and are thus accessible for antibodies in vivo. Active sensitization of animals with whole CNS tissue or myelin may give rise to antibodies, which are able to induce demyelination in vitro in organotypic myelinating tissue culture (Bornstein and Appel, 1961). Intravenous injections of such antibodies fails to induce disease, because their entry into the CNS is restricted by the intact blood brain barrier. However, when directly injected into the cerebrospinal Xuid or the CNS tissue with either complement or cytokines, which activate macrophages, they induce primary demyelination (Lassmann et al.; 1981; Vass et al., 1992). Furthermore, when such antibodies are injected intravenously into animals with EAE, induced by transfer of auto-reactive T lymphocytes, they may eVectively transform the pure T-cellmediated inXammation into an inXammatory demyelinating type of pathology, which is similar to that observed after active sensitization of the animals with brain tissue (Linington et al., 1988; Schlu¨sener et al., 1987). Thus, on this basis three diVerent disease models are available, which can be used to address fundamentally diVerent questions regarding the pathogenesis of inXammatory demyelinating diseases.

1. EAE Induced by Passive Transfer of Autoreactive T-Lymphocytes (PT-EAE) In this model memory T lymphocytes, directed against antigens of the CNS, are isolated from sensitized donors and are transferred into naı¨ve histocompatible recipients (Ben Nun

BASIC MECHANISMS OF AUTOIMMUNE-MEDIATED INFLAMMATORY BRAIN DISEASE

FIGURE 43.1 Brain inXammation induced by passive transfer of MBP-reactive Th1 cells; perivascular inXammatory inWltrates in the spinal cord (A) with minimal or absent demyelination or tissue injury (B). The inWltrates are composed of T lymphocytes (C) and macrophages (D). (A) immunocytochemistry for ED1 (macrophages);
: 100. (B) Luxol fast blue myelin stain;
: 100. (C) immunocytochemistry for W 3/13 (T cells);
500. (D) Immunocytochemistry for ED1 (macrophages);
500.

et al., 1981). Thus, this model is induced by T cells, which are already immunologically primed and it is not complicated by the immunological processes, which occur during the sensitization in the peripheral immune system. It is thus an ideal model to study the mechanisms, how T cells start the inXammatory process in the CNS, and to analyze the eVects on the target tissue, which they exert either alone or in cooperation with activated macrophages or microglia cells. SpeciWc questions that have been addressed in these models include those that involve the mechanisms of immune surveillance of the CNS, the interaction of T cells and macrophages with the blood brain barrier and the analysis, how inXammatory cells—and in particular T cells—are cleared from established inXammatory brain lesions. When monospeciWc T-cell lines or clones are used, the disease of PT-EAE follows a very reproducible course. Disease and brain inXammation starts between 3 and 5 days after transfer, reaches its clinical peak between days 5 to 8, and declines thereafter (Ben Nun et al., 1981; Wekerle et al., 1986). In general, the inXammatory process in the CNS is cleared after 15 days, leaving the CNS tissue either unaVected or in the case of severe encephalitis with some degree of perivascular astrocytic scaring. In rare instances, in particular in mouse models, a single passive transfer of autoreactive T cells may lead to a chronic disease (Mokhtarian et al., 1984; Zamvil et al., 1985), which is associated with more widespread demyelination and tissue damage (Fig. 43.2).

2. Co-transfer EAE (CoT-EAE) In this model disease is induced, as in classical PT-EAE, by the passive transfer of encephalitogenic T lymphocytes (Linington et al., 1988). At the onset of brain inXammation and disease (3 to 4 days after T-cell transfer) antibodies, soluble inXammatory

1041

1042

43. MYELIN AND ITS DISEASES

FIGURE 43.2 Chronic T-cell-mediated EAE in the mouse following sensitization with MOG peptide 35–55. The inXammatory process is associated with demyelination and tissue damage. (A) Luxol fast blue myelin stain, sowing areas of myelin loss in the ventral portions of the white matter;
: 50. (B) Adjacent section stained with the macrophage marker Mac 3 reveals massive macrophage inWltration in the areas of demyelination;
: 50.

mediators or other cells of the immune system are transferred intravenously. This allows to study, whether inXammatory mediators or immune cells, which by themselves are unable to induce brain inXammation, may modulate disease and pathology of T-cell-mediated disease. Co-transfer EAE has been extensively analyzed to deWne the pathogenic role of demyelinating antibodies in vivo (Linington et al., 1988). When such antibodies are transferred at the onset of T-cell-mediated brain inXammation the severity of clinical disease is massively augmented and the pathological alterations in the CNS are transformed from a pure inXammatory to an inXammatory demyelinating phenotype. The pathological outcome is determined by the balance between the encephalitogenic T-cell response and the concentration of circulating demyelinating antibodies (Lassmann et al., 1988). This model allows to evaluate the pathogenic potential of CNS directed autoantibodies (Weerth et al., 1999) and to study the immunological mechanisms of antibody-mediated brain tissue damage in vivo (Litzenburger et al., 1998). Furthermore, it is also useful to address more general questions on the cell biology and functional consequences of de- and remyelination, in particular, when an exactly timed model of de and remyelination is required (Di Bello et al., 1999). Co-transfer EAE is, however, not restricted to the study of antibodies. It further allows to test immunomodulatory eVects of cytokines or the functional role of regulatory T-cell populations in acute T-cell-mediated brain inXammation.

3. EAE, Induced by Active Sensitization (A-EAE) This is the most complex model of EAE, since it covers not only the eVector stage of brain inXammation, as it is done in PT-EAE and CoT-EAE, but also includes the induction phase of the immune response, such as the sensitization of naı¨ve T cells and B cells as well as their propagation and polarization into diVerent eVector and regulatory T-cell populations. Active immunization of susceptible animals in general induces an acute neurological disease, which starts between 10 and 20 days after sensitization (Alvord, 1970). Depending on the mode of sensitization, the nature of the sensitizing antigen, and the genetic background of the animals, a chronic relapsing or chronic progressive disease may develop (Lassmann, 1983, Raine, 1985). The pathology of the acute phase of the disease (10 to 20 days after sensitization) is dominated by the inXammatory reaction and mainly mediated through the T-cell response. Chronic disease is generally associated with severe demyelination or tissue damage and involves in addition to T cells other immunological mechanisms, such as for instance pathogenic antibodies (Fig. 43.3). In comparison to models of transfer EAE, A-EAE has several advantages. It can easily be induced even in laboratories, where the complex T-cell line technology is not established and, if properly handled, chronic inXammatory demyelinating disease can be induced for studies of basic cell biological

TARGET ANTIGENS IN THE CENTRAL NERVOUS SYSTEM FOR AUTOIMMUNITY

FIGURE 43.3 InXammatory demyelinating disease induced in rats by active sensitization with MOG. In this case, the lesions are induced by a cooperation of encephalitogenic T cells and demyelinating antibodies. Large plaques of demyelination are present in the cerebellar white matter (A), the optic nerve (B) and the spinal cord (C). The axonal density within the lesion is reduced, in part by edema and tissue swelling and in part by axonal injury (D). (A–C) Luxol fast blue myelin stain. (D) Bielschowsky silver impregnation for axons. (A)
15. (B) 100. (C, D)
40.

aspects of demyelination, remyelination, and tissue injury. It furthermore allows the study of basic mechanisms of the induction of autoimmunity and of immune regulation. Its disadvantage lies in its complexity, which makes it diYcult to clearly spot the individual immunopathogenetic mechanisms involved in its pathogenesis.

TARGET ANTIGENS IN THE CENTRAL NERVOUS SYSTEM FOR AUTOIMMUNITY T lymphocytes and antibodies recognize their antigen in diVerent ways (for review, see Wekerle, 1998). T cells recognize small peptides, bound to molecules of the major histocompatibility complex (MHC) on the surface of antigen presenting cells. CD8þ T cells use class I MHC and CD4þ T cells class II MHC molecules for antigen recognition. In the normal central nervous system, the expression of both classes of MHC molecules is restricted to some perivascular macrophages and microglia cells, class I molecules being additionally present on endothelial cells (Traugott et al., 1983, Matsumoto and Fujiwara, 1986). While class I MHC can be expressed on all cell types of the nervous system under inXammatory conditions, the expression of class II molecules is more restricted and predominantly found on leukocytes and microglia cells (Vass and Lassmann, 1990). For T-cell antigen recognition, proteins have to be cleaved into small peptides in intracellular compartments of the antigen presenting cells; they have then to be attached to respective MHC molecule and transported to the surface of the antigen presenting cell, where they are recognized by the T cell through their speciWc T-cell receptor. This antigen

1043

1044

43. MYELIN AND ITS DISEASES

recognition requires additional co-ligation between T cells and antigen presenting cells through adhesion molecules and co-stimulatory molecules. The selection of the proper peptide not only depends on the T-cell receptor, but also on its binding abilities to the MHC molecule and its proteolytic cleavage sites. In an outbred population or in diVerent inbred strains, which have diVerent MHC haplotypes, the antigenic peptides will be diVerent. Thus, the antigenic peptide, which drives an autoimmune disease, may be the same in diVerent animals from the same inbred strain, but in general will diVer between diVerent individuals of an outbred population. It has, however, to be noted that both extracellular as well as intracellular proteins may be presented in this way and may thus be targets for the cell-mediated immunity. In contrast, antibodies recognize their antigen within the intact protein and do not require speciWc antigen presentation. Thus, a pathogenic antibody will recognize its target across histocompatibility barriers, provided the antigen has the respective interspecies homology in protein sequence and structure. The limitation of antibodies to be pathogenic is its accessibility of the target epitope in vivo. This implies that the antigen (epitope) has to be either located on the cell surface or in the extracellular space and the epitope recognized by the antibody must not be hidden within the tertiary structure of the protein. Due to these major diVerences in epitope recognition between T cells and antibodies, their roles in CNS autoimmunity have to be discussed separately.

T-Cell-mediated Autoimmunity For a long time myelin basic protein, one of the major myelin proteins, was believed to be the only brain antigen responsible for the induction of EAE. It was the Wrst that had been identiWed and chemically characterized (Kies et al., 1960). The reason for this was that it is relatively easy to purify and due to its hydrophilic nature it is readily available for use in immunological test systems. Myelin basic protein is highly encephalitogenic, which means that it is able to induce EAE even in minute concentrations (Alvord, 1970). MBP induces a strong encephalitogenic T-cell response in all vertebrate species tested so far. However, the peptide epitopes of the MBP molecule, which are recognized by encephalitogenic T cells, diVer between diVerent species and between diVerent strains of the same species (Alvord, 1984; Swanborg, 2001). Furthermore, even within a given inbred strain frequently more than one peptide epitopes of the MBP molecule can drive a T-cell response, which each may diVer in their encephalitogenic potential. This complex situation makes it diYcult or even impossible to prove the pathogenetic potential of a T-cell response against MBP in humans. This problem is further complicated by the fact that MBP is only one candidate antigen for an encephalitogenic T-cell response. Another antigen, which can drive an encephalitogenic T-cell response, is proteolipid protein (PLP; Sobel and Kuchroo, 1992). Similar to the situation with MBP the peptide epitopes within the PLP molecule, which are responsible for the induction of T-cell-mediated encephalitis, vary between animal species and strains. In recent years encephalitogenic T-cell responses were identiWed, directed against a variety of other myelin and oligodendrocyte proteins, such as myelin oligodendrocyte glycoprotein (Linington et al., 1993), myelin associated glycoprotein (Morris-Downes et al., 2002; Weerth et al., 1999), cyclic nucleotide phosphodieserase (Morris-Downes et al., 2002; Rosener et al., 1997), and oligodendrocytic basic protein (Maatta et al., 1998; Morris-Downes et al., 2002), as well as against proteins expressed in other glial cells, such as for instance S-100 protein (Kojima et al., 1994). These data suggest that any CNS protein, which is able to stimulate a T-cell response (and this probably applies to most of them), is a possible candidate for a T-cell-mediated encephalomyelitis. The spectrum of possible target auto-antigens for an encephalitogenic T-cell response in humans is thus very broad. In addition, brain inXammation was also induced by T cells, which react through molecular mimicry with brain proteins as well as proteins from infectious agents (Bhardwaj et al., 1993; Moktharian et al., 1999; Wucherpfennig and Strominger, 1995). This does not

TARGET ANTIGENS IN THE CENTRAL NERVOUS SYSTEM FOR AUTOIMMUNITY

require complete sequence identity between the respective peptides, but just a peptide structure, which in its motive is similar in the essential anchor sequences that mediate the binding to the MHC molecule and the T-cell receptor (Wucherpfennig and Strominger, 1995). This further increases the possible number of peptide sequences, which may give rise for an encephalitogenic T-cell response. T-cell autoimmunity can also be induced against other molecules than peptides, such as for instance glycolipids and complex carbohydrates. In this situation antigen presentation is diVerent, since these molecules are not presented by classical MHC molecules but by molecules of the CD1 complex (Moody et al., 1999; Moody and Besra, 2001). CD1 molecules are nonpolymorphic and thus do not diVer between diVerent outbred members of the same species. Some of the CD1 molecules have a deep hydrophobic grove, which allows the binding of glycolipid molecules. The speciWc T-cell response is directed against the complex carbohydrate structure of the respective glycolpipds. SpeciWc T-cell reactions against a variety of glycolipids in the central nervous system, such as gangliosides or sulfatides have been identiWed in humans, in particular in multiple sclerosis patients and such molecules are, thus, potential targets for T-cell-mediated autoimmunity (Shamshiev et al., 1999, 2002). Due to species diVerences in the structure and function of CD1 molecules and the anti-lipid T-cell response, experimental evidence for the encephalitogenic potential of such T cells so far is indirect.

Target Antigens for Antibody-Mediated Demyelination and Tissue Destruction As mentioned earlier, the basic requirement for an antibody to be pathogenic is the accessibility of the target antigen in the intact tissue. Thus, the target antigen has either to be present in the extracellular space or expressed on the extracellular surface of cells. This to a large extent reduces the number of potentially pathogenic target antigens, since it excludes all those epitopes which are exclusively present in the cell cytoplasm or in other compartments, which are not accessible from the extracellular space. The pathogenic potential of antibodies can be tested when they are directly applied together with either activated macrophages or complement to organotypic cultures of CNS tissue (Bornstein and Appel, 1961). Alternatively, such antibodies can be directly injected into the cerebrospinal Xuid or the CNS tissue (Jankovic et al., 1965, Lassmann et al., 1981). Furthermore, their pathogenic potential in vivo can be tested in models of co-transfer EAE (Linington et al., 1988). This basic principle has been extensively studies for antibodies against myelin components. Thus, antibodies against MBP, which is located at the cytoplasmic face of the cell membrane in compact myelin sheaths, do not induce demyelination in vitro or in vivo. In contrast, MOG, which is located on the extracellular surface of oligodendrocytes and myelin sheaths, is a target for antibody-mediated demyelination (Linington et al., 1988). The situation is more complex for antigens, such as PLP or MAG, which at least contain epitopes, which are located on the extracellular face of the myelin membrane. Nevertheless, antibodies against these antigens did not induce demyelination in vitro (Seil et al., 1980,, 1981). This could be explained by the complex architecture of the myelin sheath. The extracellular compartment within the compact myelin sheath (interperiod line) is secluded from the brain extracellular space by tight junctions, which prohibit the entry of serum proteins into the compact myelin. Similarly, diVusion of extracellular proteins is restricted into the periaxonal space where MAG is located. In addition, the epitope recognized by the antibody has to be accessible in vivo within the complex tertiary structure of the molecule. As an example, most antibodies, which were directed against linear peptide epitopes of MOG, even against those which are contained in the extracellular domain of the molecule, did not bind to native MOG on the surface of transfected cells (Haase et al., 2001). Thus, so far only a small number of molecules has been identiWed as targets for demyelinating antibodies. They include MOG as well as certain glycolipids, which are exposed on the myelin surface, such as galactocerebroside (Dubois-Dalcq et al., 1970). In principle, however, also antibodies against other cellular targets of the CNS, such as

1045

1046

43. MYELIN AND ITS DISEASES

astrocytes or neurons, could be pathogenic and could modulate the pathology of T-cellmediated encephalitis. Interestingly, in some multiple sclerosis patients, antibodies have been detected, which are directed against progenitor cells of oligodendrocytes (Niehaus et al., 2000). It has to be determined in the future whether such antibodies can impair remyelination in inXammatory demyelinating lesions of the CNS.

WHAT T-CELL POPULATIONS ARE RESPONSIBLE FOR THE INDUCTION OF BRAIN INFLAMMATION? The immune system is controlled by diVerent T lymphocyte populations. Helper T cells (Th-cells), by their production of cytokines may either recruit and activate eVector cells, such as macrophages or granulocytes or may stimulate B-cell diVerentiation and antibody production. Cytotoxic T lymphocytes (Tc-cells) destroy in an antigen-speciWc manner speciWc target cells through their cytotoxic machinery. Regulatory T cells (Tr-cells), again through cytokine production, modulate or down-regulate immune reactions. Furthermore, antigen recognition diVers between T-cell subsets. CD4þ T cells recognize antigenic peptides in the context of class II MHC molecules, CD8þ T cells use class I MHC molecules and T cells directed against complex carbohydrates or glycolipis require CD1 molecules for antigen presentation. Finally, the Wrst antigen contact of naı¨ve T cells may polarize T cells toward the production of Th/Tc1 cytokines such as interferongamma (IFN-g) or tumor necrosis factor alpha (TNF-a), of Th/Tc2 cytokines, such as interleukin 4, 5, and 10 or of immunoregulatory cytokines, such as transforming growth factor beta (TGF-b). Thus, many diVerent T-cell populations with fundamentally diVerent function may be involved in the initiation and regulation of T-cell-mediated brain inXammation.

Th-1 Cells in EAE EAE has for long been regarded as the paradigmatic model of an autoimmune disease mediated by Th-1 cells (for review see Lafaille, 1998; Owens et al., 2001). This assumption is mainly based on studies of EAE models, induced by sensitization with MBP. The disease can be transferred by intravenous injection of monospeciWc autoreactive Th-1 polarized Tcell lines and clones. The initial stage of brain inXammation is dominated by local production of Th-1 cytokines, while during the recovery phase of the disease Th-2 cytokines become more prominent in the lesions (Issazadeh et al., 1995a, 1995b). The disease, induced by active sensitization can be inhibited by deletion of CD4þ T cells and by blockade of class II MHC molecules and inhibition or genetic deletion of Th-1-related cytokines, in particular interleukin 12 may ameliorate disease. In addition, disease susceptibility of diVerent animal strains is to a large extent controlled by the MHC class II complex (Olsson et al., 2000). These data clearly document the ability and importance of Th-1 cells for disease induction in EAE, but they do not rule out possible additional pathogenetic roles of other T-cell populations.

Th-2 Cells in EAE Th-2 cells are generally believed to be protective in EAE (Lafaille, 1998). They dominate in lesions of animals, which recover from disease (Issazadeh et al., 1995a, 1995b). Sensitization strategies, which favor a Th-2 response, are in general protective against brain inXammation in this model (Wekerle, 1998). Thus, immune deviation from Th-1 to Th-2 responses is proposed as a valid therapeutic option for T-cell-mediated brain diseases, such as multiple sclerosis and this notion is supported by a large number of experimental studies, showing such eVects in MBP-induced EAE (Hohlfeld, 1997). There are, however, a variety of new data that indicate that the situation is more complicated. It was shown that Th-2 polarized T cells against myelin basic

WHAT T-CELL POPULATIONS ARE RESPONSIBLE FOR THE INDUCTION OF BRAIN INFLAMMATION?

protein can induce brain inXammation following passive transfer, although profound immunosuppression of recipient animals was required (LaVaille et al., 1997). The pathology of the brain lesions in this model was diVerent from that induced by Th-1 cells, the lesions being associated with profound tissue destruction, and the inXammatory inWltrates dominated by granulocytes. Thus, in principle, Th-2 cells are able to directly induce brain inXammation and cerebral tissue damage. A possible role of Th-2 immune responses is also suggested from studies in EAE, induced by sensitization with MOG. With this antigen the disease is driven by a synergy of pathogenic T-cell and antibody responses. Furthermore, in the chronic stage of the disease its severity is determined rather by the antibody than the T-cell response (SteVerl et al., 1999). Thus, in such a situation any stimulation of the antibody response may be detrimental. In this model, sensitization procedures or therapeutic strategies, which stimulate Th-2 responses may be associated with more severe disease as well as more demyelination and tissue damage compared to that in a classical Th-1-mediated autoimmune encephalomyelitis (Genain et al., 1996). This may also be associated with a Th-2 driven antibody response (Tsunoda et al., 2000) and an inXammatory pathology, dominated by the inWltration of granulocytes and eosinophils (Storch et al., 1998).

Class I MHC restricted Cytotoxic T cells Similar to that of Th-2 cells, the role of class I restricted CD8þ T lymphocytes has for long been regarded as inhibitory or regulatory. This again is mainly based on data, obtained in EAE models, induced with myelin basic protein. Blockade of antigen presentation by the class II pathway, either through antibodies or by genetic deletion of CD4 or MHC class II molecules in general prevents EAE. In contrast, blockade of CD8þ lymphocytes through antibodies or genetic deletion resulted in increased disease severity and, in chronic models, in increased numbers of relapses (Jiang et al., 1992; Koh et al., 1992). In addition, cytotoxic class I MHC restricted T cells were shown to destroy encephalitogenic Th-1 cells in an antigen speciWc manner through recognition of peptides of the respective T-cell receptor (Sun et al., 1988). Thus, these data taken together suggest that class I restricted T cells in EAE may function as regulatory T cells instrumental in suppression of acute disease and relapses. This concept was challenged by the observation that passive transfer of a T-cell population, which was enriched with MOG reactive, class I restricted CD8þ cells can induce brain inXammation in normal recipient mice. In addition, these T cells induced more severe tissue damage than Th-1 cells alone (Sun et al., 2001). In another experiment, myelin basic protein reactive CD8þ T cells were raised from animals, which were genetically deWcient for MBP. Such T cells, when passively transferred into recipients, which did express MBP in the CNS, induced brain inXammation (Huseby et al., 2001). Disease, however, was only seen when the recipients were immunosuppressed and treated with Il-2 before transfer. The lesions in the brain showed T-cell and macrophage inWltrates, like in classical PT-EAE; brain tissue damage, however, when present resembled the pathology of brain ischemia. Whether this was due to a vasculitic reaction, induced by class I restricted T cells or by other mechanisms is so far unclear. To avoid the problem of tolerance for class I–restricted T cells against autoantigens, their possible pathogenicity has also been tested in transgenic models. The strategy in these experiments was to express a foreign antigen under a CNS speciWc promotor. Upon peripheral immunization with the respective antigen, the animals develop an inXammatory disease of the central nervous system, associated with destruction of the cells, expressing the transgene, which was mainly mediated by class I restricted T cells (Evans et al., 1996). This was also achieved following passive transfer of cytotoxic T cells (Oldstone et al., 1993). We recently used a similar model (Cornet et al., 2001), in which inXuenza haemagglutinin is expressed under the GFAP promotor. We then transferred class I restricted cytotoxic T cells, derived from a T-cell receptor transgenic animal, in which the T-cell repertoire entirely consists of HA reactive class I MHC restricted T cells (Lassmann et al. unpublished). These studies showed that such T cells can induce brain inXammation in normal transgenic recipients. Similar as in PT-EAE with Th-1 cells, antigen speciWc

1047

1048

43. MYELIN AND ITS DISEASES

activation of the T cells is required to allow them to home in the CNS. The pathological alterations in the CNS in this model were characterized by inXammatory inWltrates, mainly composed by activated T lymphocyte. The lesions showed little recruitment of haematogenous macrophages but were associated with profound microglia activation, selective loss, and apoptotic destruction of antigen-containing astrocytes without bystander damage of other CNS elements, such as myelin, oligodendrocytes, neurons, or axons (Fig. 43.4). Thus, the pathology of these lesions was much more selective compared to that found after transfer of Th-1 cells. These data together with recent in vitro studies suggest a major pathogenic role of cytotoxic, class I MHC restricted T cells in inXammatory brain diseases (Neumann et al., 2002). These studies clearly show that not only Th-1 cells, but also class I restricted cytotoxic T cells and, under certain conditions also Th-2 cells, can induce or contribute to T-cellmediated autoimmune inXammation in the CNS. The question is why this was overlooked during many years of EAE research. The reason for this may be twofold. First, most studies on the immunology of EAE concentrated on the T-cell response against myelin basic protein. This antigen induces a disease which is purely T-cell mediated. Antibodies against MBP do not recognize their target antigen in vivo and thus are not pathogenic. When, however, like in MOG-induced EAE antibodies play a major role in the pathogenesis of brain disease, a stimulation of Th-2 responses may play a role in augmentation and

FIGURE 43.4 Brain inXammation induced by class I MHC restricted cytotoxic T cells; passive transfer of haemaglutinin reactive CD8þ T cells in recipient animals, which express haemaglutinin under the GFAP promotor in astrocytes. Perivascular inXammation with vacuolization of the tissue (A) is associated with a difuse inWltration of T lymphocytes in the tissue (B) and an activation of local microglia cells (C). This process leads to degenerative changes and loss of astrocytes in the inXamed areas (D). (A) Haematoxylin/eosin stain. (B) Immunocytochemistry for CD3. (C) Immunocytochemistry for Mac-3. (D) Immunocytochemistry for GFAP. All Wgures
: 300.

IMMUNE SURVEILLANCE OF THE CNS AND THE INDUCTION OF INFLAMMATION

modiWcation of the disease. Second, thymic deletion and peripheral tolerization may be more eYcient for class I restricted cytotoxic T cells than for Th-1 cells. This may account for many classical EAE antigens, such as MBP or PLP, since they are expressed not only in the nervous system, but also in peripheral tissues, including the thymus (Voskuhl, 1998). A strong encephalitogenic class I restricted T-cell response has, thus, so far only been achieved by sensitization with either a foreign antigen or with an autoantigen, which is not expressed at all in peripheral tissues, such as MOG (Sun et al., 2001) or MBP in MBPdeWcient animals (Huseby et al., 2001). This, however, does not rule out that a pathogenic class I restricted autoimmune response could be induced by molecular mimicry between infectious agents and autoantigens in the CNS (Evans et al., 1996; Wucherpfennig and Strominger, 1995).

IMMUNE SURVEILLANCE OF THE CNS AND THE INDUCTION OF INFLAMMATION The models of PT-EAE and in particular the introduction of monospeciWc autoreactive T-cell lines and clones (Ben Nun et al., 1981) paved the way for detailed studies on the mechanisms of immune surveillance of the brain and on the molecular requirements for the induction of brain inXammation. As mentioned earlier, intravenous injection of activated autoreactive Th-1 cells leads to an inXammatory brain disease (Fig. 43.5), which starts 3 to 4 days after transfer. InXammation reaches its peak around day 5 to 7 and then declines. The time of disease onset in part depends upon the number of T cells, which are transferred, with lower numbers leading to a delay of a few days. However, even with transfer of

FIGURE 43.5 InXammatory inWltrate in T-cell-mediated EAE; massive perivascular accumulation of Lymphocytes and macrophages. Toluidine blue stained plastic section;
2000.

1049

1050

43. MYELIN AND ITS DISEASES

very high cell numbers, an incubation period of at least 72 hours remains (Flu¨gel et al., 2001). Detailed quantitative studies on T-cell inWltration in the CNS revealed that within the Wrst hours after intravenous transfer some T cells enter the CNS. This early wave of T cells is antigen independent, being similar after transfer of autoreactive MBP-speciWc cells or T lymphocytes directed against an irrelevant antigen, which is not present in the CNS (Hickey et al., 1991; Wekerle et al., 1986). However, only CNS-speciWc T cells induce an inXammatory CNS disease. During the initial phase of CNS inXammation, 3 to 4 days after transfer, antigen speciWc T cells dominate in the lesions (Bauer et al., 1998; Flu¨gel et al., 2001). Thereafter, additional T cells and macrophages from the host are recruited into the lesion, which at the peak of clinical disease and brain inXammation dominate the inXammatory inWltrates (Bauer et al., 1998). This is associated with local up-regulation of class I and class II MHC molecules, both on inXammatory cells and local glia cells and by profound activation of local microglia. Both the antigen speciWc as well as the secondarily recruited T cells are rapidly cleared from the lesions by apoptotic destruction (Bauer et al., 1998). These data suggest that the normal blood brain barrier does restrict the entry of most circulating leukocytes into the CNS. This, however, is not the case for T lymphocytes, which have been activated by their speciWc antigen in the periphery. A peripheral activation of these cells, for instance, in the course of an infection, will start the process of immune surveillance of the brain. In the absence of the speciWc antigen, the T cells will either be locally destroyed by apoptosis or possibly also leave the brain through lymphatic drainage. In case they encounter their speciWc antigen in the CNS, they will be reactivated, start to produce pro-inXammatory cytokines, such as gamma-interferon or TNF-a, and initiate the local inXammatory response. It is, however, important to note that freshly in vitro activated T cells have only a limited capability to reach the CNS and their direct injection into the cerebrospinal Xuid does not induce inXammation and disease. Instead these cells have after their transfer to migrate in a predictable way through the peripheral lymphatic system, where they obtain speciWc phenotypic and functional characteristics that allow them to migrate to and home in the CNS and mediate the inXammatory response (Flu¨gel et al., 2001).

Leukocyte Entry into the CNS Is Controlled by Adhesion Molecules and Chemokines The process of leukocyte migration through vascular barriers is controlled by adhesion molecules and chemokines and occurs in several consecutive steps (Luster, 1998; Springer, 1994). The Wrst step is characterized by loose attachment of the cells to the endothelial surface. It leads to rolling of leukocytes along the luminal surface of the vessels and is mainly mediated by the interaction between selectins and carbohydrates. Then the leukocytes Wrmly attach to the endothelial surface by the interaction of adhesion molecules of the intergrin and immunoglobulin family. This Wrm attachment initiates transmigration, which either may occur in a trans-endothelial route through endothelial channels or in a paracellular route. Migration of inXammatory cells through the vessel wall requires the activation of proteases, which dissolve the inter-endothelial junctions and/or the basement membrane. This process of trans-endothelial migration of leukocytes is further regulated by the interaction between chemokines and their receptors (reviewed by Luster, 1998). Chemokines are a family of small proteins which are produced and secreted both, by leukocytes and by cells of the target tissue. They diVuse toward the vessel wall and may become attached to the luminal surface of endothelial cells. There they are recognized by leukocytes, which express the respective receptors. In addition, leukocytes are attracted by soluble chemokines and migrate along their concentration gradient. Chemokines may either activate leukocyte adhesion molecules and thus support the Wrm attachment of the cells to the endothelial surface. In addition, through diVerent members of the chemokine family and their interaction with speciWc receptors, which are expressed on diVerent leukocyte subsets, they determine which leukocytes (T cells, B cells, macrophages, or

IMMUNE SURVEILLANCE OF THE CNS AND THE INDUCTION OF INFLAMMATION

granulocytes) are attracted to the inXammatory focus. Thus, chemokines determine the quality of the inXammatory reaction. Adhesion molecules and chemokines play a major role in controlling brain inXammation (for review see Lee and Benveniste, 1999; Huang et al., 2000; RansohoV, 1999). Endothelial cells of the normal blood brain barrier show only a low basal level of adhesion molecule expression (Lee and Benveniste, 1999), and there is only a minor synthesis of chemokines in the normal brain (Huang et al., 2000). This may explain, why it is permissive for transmigration only of activated T cells with high levels of adhesion molecule and chemokine receptor expression (Flu¨gel et al., 2001). When inXammation, however, is initiated cerebral endothelial cells become activated and express high levels of adhesion molecules. Furthermore local synthesis of chemokines in perivascular macrophages and astrocytes create a pro-inXammatory environment, which attracts more T cells and macrophages into the lesions. In line with this concept a variety of adhesion molecules have been detected on brain endothelia within EAE lesions and this is associated with increased binding of leukocytes to the inXamed vessels in in vitro assays (Engelhardt et al., 1994). Vascular cell adhesion molecule (VCAM) seems to be particularly important, since blockade of the interaction with its binding partner on leukocytes consistently reduces the severity of brain inXammation in EAE in vivo (Yednock et al., 1992). The spectrum of chemokines, present in classical EAE lesions is broad and includes mainly chemokines, which are associated with the recruitment of T lymphocytes and monocytes/macrophages (Godiska et al., 1995) and it appears to diVer depending upon the stage of the disease. Thus, in the acute phase of EAE diVerent chemokines may be involved than in chronic lesions (Kennedy et al., 1998) Their respective receptors are expressed on leukocytes within the lesions (Fife et al., 2001; Izikson et al., 2000; Jee et al., 2002; Matsui et al., 2002,). Blockade of certain chemokine receptors, such as for instance CCR2 through antibodies or through genetic deletion is a potent anti-inXammatory therapy in EAE (Fife et al., 2000; Izikson et al., 2000). To what extent other chemokines are present in atypical EAE lesions, such as those with profound granulocyte inWltration, is so far undetermined. Besides classical chemokines there are other molecules with similar pro-inXammatory function. One of those are components of the complement system, such as C3 or C5, which interact with the respective complement receptors on leukocytes. Thus, complement inactivation or blockade of complement receptors may not only inhibit tissue damage but in addition has a profound anti-inXammatory eVect (Huitinga et al., 1993). Chemokine function may however also be mediated through certain neuropeptides (Numao and Agrawal, 1992; RuV et al., 1985), and in EAE the neuropeptide secretoneurin appears to attract macrophages into sites of brain inXammation (Storch et al., 1996).

MHC-Expression and T-cell Activation in the CNS As discussed earlier in detail, T lymphocytes recognize their antigen only, when it is presented in the context of MHC molecules. Thus, MHC expression in the CNS is the prerequisite that T cells in the inXammatory focus get activated through speciWc recognition of their antigen, liberated within the tissue. MHC expression in the normal CNS tissue is strictly controlled (Matsumoto and Fujiwara, 1986) and only present on perivascular and meningeal macrophages and few microglia cells. In addition, MHC class I molecules are present on the luminal surface of cerebral endothelial cells. The suppression of MHC antigen expression is an active process and depends on the electrical activity of the neurons (Neumann et al., 1996, 1997). Blockade of electrical activity by tetrodotoxin in brain slices leads to rapid up-regulation of MHC expression not only in the neurons themselves but also in glia cells, which are located in their vicinity. The active suppression of MHC expression by electrically active neurons is at least in part mediated by neurotrophic factors, signaled through the p75 neurotrophin receptor (Neumann et al., 1998). In EAE, irrespective of the mode of disease induction, MHC molecules are up-regulated within the inXammatory CNS lesions (Hickey et al., 1985; Matsumoto and Fujiwara, 1986;

1051

1052

43. MYELIN AND ITS DISEASES

Sobel and Colvin, 1985). The extent of up-regulation reXects the severity of the inXammatory process. Most extensive expression is found on invading leukocytes and on microglia. In case of intense inXammation, class I molecules can also be found on other cells, such as astrocytes and some oligodendrocytes or neurons. In contrast, class II molecules can be found on microglia and only rarely on some astrocytes and ependymal cells. This graded response of MHC molecule up-regulation in the CNS is also seen after direct injection of gamma-interferon into the cerebrospinal Xuid or the CNS tissue (Vass and Lassmann, 1990). Also in this situation microglia respond most prominently, while expression on other glial cells is only found after application of rather high concentrations of this cytokine. These data suggest that antigen recognition by T cells in the CNS occurs in diVerent steps. In the normal CNS tissue in the course of immune surveillance, antigen presentation occurs on meningeal and perivascular macrophages. This implies that an autoantigen, which may be recognized by T cells in the normal brain, has to be liberated into the brain’s extracellular space and has to diVuse to the perivascular or menigeal compartment, where it is presented by cells, constitutively expressing MHC molecules. When inXammation is started, MHC antigens are up-regulated in a graded manner, Wrst on microglia and only in conditions with intense inXammation also on other cells of the CNS. Thus, during the development of inXammatory brain lesions and in particular in chronic inXammatory processes more and additional antigen presenting cells are recruited, which may amplify the T-cell-mediated inXammatory reaction. This concept is supported also by data, obtained in the model of EAE induced in radiation bone marrow chimeras. In this model, bone marrow transplantation is performed, using cells with a partially diVerent MHC haplotype. These new bone marrow cells, at least in this particular rat model, replace the population of meningeal and perivascular macrophages, but not the parenchymal cells in the CNS tissue such as neurons, oligodendrocytes, astrocytes, and microglia (Hickey et al., 1992). PT-EAE induced by T cells with an MHC haplotype, corresponding to that of the transplanted bone marrow cells, results in brain inXammation, similar to that present in fully histocompatible animals (Hickey and Kimura, 1988). These data show that antigen presentation by the population of meningeal and perivascular macrophages alone is suYcient to induce the basic inXammatory reaction of autoimmune encephalomyelitis. It is, however, expected, but so far not formally proven that this basic inXammatory reaction may be ampliWed by antigen presentation through additional cells of the CNS tissue. Direct antigen presentation by cells of the CNS is required for cytotoxicity mediated by class I MHC restricted T cells. Thus, when EAE is induced by transfer of class I restricted cytotoxic T cells the apoptotic destruction of the speciWc target cells is associated with profound class I expression in the respective cells (Lassmann unpublished). Not surprisingly no disease and CNS lesions are found when such T cells are transferred into ß-2 microglobulin deWcient mice (Sun et al., 2001).

CLEARANCE OF BRAIN INFLAMMATION IN AUTOIMMUNE ENCEPHALOMYELITIS The CNS is diVerent from other organs, since it lacks classical lymphatic vessels. Yet in acute autoimmune encephalomyelitis the inXammatory lesions subside within Wve to ten days after the peak of inXammation and clinical disease. Thus, there must be ways and mechanisms, how inXammatory cells, which have entered the CNS in the course of immune surveillance or inXammation, are cleared from the tissue. An important observation was reported by Pender (Pender et al., 1991), who draw the attention to the fact that numerous cells within EAE lesions are destroyed by apoptosis. A more detailed analysis of the cell types undergoing apoptosis in the lesions revealed that they were predominantly T lymphocytes. Furthermore, since programmed cell death is a rapid procedure, the quantitative data suggested that within a 24-hour time period more

DETERMINANT-SPREADING AS A MECHANISM TO PROPAGATE AUTOIMMUNITY

than twice the number of cells present in the lesions are removed by apoptosis (Schmied et al., 1993). Thus, T-cell-mediated inXammation in acute EAE can only be maintained as long as there is a continuous supply of new T cells entering the CNS from the peripheral circulation. Using prelabeled T lymphocytes for PT-EAE it became clear that not only those T cells that recognize the speciWc antigen in the CNS, but also all other secondarily recruited T cells are eliminated from the lesions through programmed cell death (Bauer et al., 1998). Besides T cells, B lymphocytes also have been shown to undergo apoptosis in inXammatory CNS lesions (White et al., 2000), while plasma cells seem to be resistant. Detailed quantitative data on the elimination of B cells are, however, so far not available. The immunological mechanisms that are responsible for this eYcient destruction of T cells in the CNS are so far not clear and several diVerent pathways may be involved (Gold et al., 1997; Pender and Rist, 2001). Antigen speciWc T cells may be driven into apoptosis by contact with excess of soluble antigen, liberated in the CNS in the course of tissue damage, and TNF signaling through the TNF-receptor I seems to be involved in this process (Bachmann et al., 1999). In addition, astrocytes express Fas-ligand, which may induce cell death through activation of the Fas-receptor, which is expressed on activated T lymphocytes (Bechmann et al., 2000). This may explain why in the CNS apoptotic destruction of T cells is mainly found within the CNS parenchyme and not in the meningeal and perivascular space. These data, however, do not exclude that other, so far unknown CNS-speciWc mechanisms play and additional role. This eYcient local destruction of lymphocytes within inXamed brain tissue has several important consequences. It suggests that apoptotic destruction of the speciWc cells of the immune system is the major mechanism to terminate T-cell-mediated inXammation in the CNS. It may further explain why after acute EAE the original encephalitogenic T-cell population may be clonaly deleted from the immune repertoire. Finally, an impairment of this clearance process may be involved in the pathogenesis of chronic inXammatory diseases of the CNS. The lack of apoptotic destruction of fully diVerentiated plasma cells may explain the long lasting intrathecal immunoglobulin synthesis following cerebral inXammation. While it is clear now that the vast majority of lymphocytes do not leave the CNS but are locally destroyed within the tissue, the situation seems to be diVerent for macrophages. Although some of these cells within the lesions can be found, which show morphological changes of apoptosis, their number is very low (White et al., 1998). Whether macrophages leave the CNS through lymphatic drainage pathways has to be determined.

DETERMINANT-SPREADING AS A MECHANISM TO PROPAGATE AUTOIMMUNITY In some animal species and strains, EAE induced by a single antigenic peptide may develop into a chronic inXammatory disease of the nervous system. When the T-cell immune response in such animals is followed over time, a broadening of the immune response to diVerent peptide epitopes of the same protein or even to peptides of diVerent proteins can be seen (Mc Rae et al., 1995; Yu et al., 1996). This process has been deWned as determinantor epitope-spreading and it is believed to be a major mechanism, responsible for chroniWcation of an autoimmune disease. This cannot only be observed in primary autoimmune diseases, such as EAE, but may also happen as a consequence of virus-induced inXammation in the brain and may be one possible mechanism, how virus infections can trigger a neurological autoimmune disease (Miller and Eagar, 2001). The basis for the induction of antigen-spreading is the liberation of brain antigens in an inXammatory environment and the genetic susceptibility of the organism to mount a respective autoimmune response. The question of where the autoimmune response is triggered is still unresolved. This is not trivial, since the induction of a new T-cell response requires the stimulation of naı¨ve, unprimed T cells by professional antigen presenting cells, such as dendritic cells. The normal brain does not contain dendritic cells and they are also

1053

1054

43. MYELIN AND ITS DISEASES

not induced in or recruited into the CNS by simple tissue injury. However, in chronic inXammatory lesions, in particular also in chronic EAE lesions, cells can be found, which phenotypically resemble dendritic cells and express the costimulatory molecules, necessary to stimulate naı¨ve T cells (SeraWni et al., 2000). The source of these dendritic like cells is still controversial. They may either be recruited from the circulation by speciWc chemokines. Alternatively, some data suggest that microglia are undiVerentiated haematopoetic progenitor cells, which have the potential to diVerentiate either into tissue macrophages or dendritic cells, depending on the cytokine milieu in the lesions (Fischer and Reichmann, 2001). It is thus feasible that naı¨ve T cells, which are nonspeciWcally recruited into the inXammatory focus, are then locally primed with CNS antigen, presented by local dendritic cells and locally diVerentiate into autoimmune memory cells (Miller et al., 1995; Vanderlugt et al., 1998). The alternative mechanism for the induction of antigen-spreading is that brain antigens are either de novo expressed in or reach the lymphatic tissues, after being liberated in the lesions (Voskuhl, 1998).

MECHANISMS OF TISSUE DAMAGE IN INFLAMMATORY LESIONS OF THE CNS Brain inXammation in EAE is associated with damage and destruction of CNS tissue. DiVerent forms of tissue damage may occur in the white matter. Primary demyelination is deWned as a process, which leads to destruction of myelin sheaths with relative preservation of axons. Although all primary demyelinating lesions show some degree of axonal injury and loss, the characteristic feature of such lesions is that within the areas of complete loss of myelin preserved axons are still present in high density. Such primary demyelinating lesions have to be distinguished from lesions with unselective tissue damage In this situation, axons are aVected to a similar extent compared to myelin and the degeneration of myelin occurs in parallel or secondary to axonal loss (secondary demyelination). In both situations, astrocytes built up a dense scar tissue (Fig. 7). Finally, in most severe lesions, all tissue elements can be destroyed, leading to cytstic necrosis. All three diVerent forms of tissue injury can be found in EAE lesions (Lassmann, 1983), the quality and quantity of tissue damage is highly variable between diVerent models and depends on the genetic background of the animals, the mode of sensitization, the nature of the sensitizing antigen, and the chronicity of the disease. Overall in acute EAE the pathology is dominated by inXammation and tissue damage or demyelination is sparse. In contrast, chronic models of EAE are generally associated with extensive demyelination or tissue damage. In addition, EAE models that are exclusively mediated by Th1 cells show less pronounced tissue injury than those in which the immune response is directed against multiple antigens and inXammation is mediated by multiple diVerent immune mechanisms (Lassmann, 1983). In immunological terms tissue destruction can occur by speciWc immune reactions, either through antigen recognition by cytotoxic T cells or by speciWc antibodies. This is generally associated with a high degree of selectivity of the cellular injury. In contrast, tissue may be destroyed by activated eVector cells, such as macrophages, microglia, or granulocytes, independently of speciWc antigen recognition on the target cells. This type of damage is called ‘‘bystander damage’’ (Wisniweski and Bloom, 1975). Such bystander damage may show a certain degree of speciWcity, since for instance myelin sheaths or oligodendrocytes are in general more susceptible to toxic macrophage products than neurons or astrocytes. In more severe lesions, bystander damage, however, is always associated with profound unselective tissue injury. A variety of diVerent immunological mechanisms appear to be involved in the induction of tissue injury in EAE and their relative importance for disease outcome varies between species and strains of animals. These mechanisms include direct T-cell-mediated cytotoxicity, tissue injury induced by toxic products of activated eVector cells, such as macrophages, microglia, or granulocytes, speciWc antibodies that can damage tissue either

MECHANISMS OF TISSUE DAMAGE IN INFLAMMATORY LESIONS OF THE CNS

FIGURE 43.7 Glial scar formation in chronic EAE lesions; numerous reactive astrocytes are present within the lesion, forming the scar with a dense network of cell processes. Toluidine blue-stained plastic section;
2000.

through complement activation or the interaction with activated macrophages, or, in very severe lesions, hypoxic or ischemic metabolic disturbances.

Direct T-cell-mediated Cytotoxicity In principle, both class I and class II–restricted T lymphocytes have a cytotoxic potential. CD8þ T cells, however, are more eVective in tissue destruction, since class I MHC expression is less restricted in comparison to class II. Thus, in vitro, all cells of the nervous system, including neurons, axons, oligodendrocytes, astrocytes, or microglia can be destroyed by cytotoxic T cells in an antigen-speciWc class I MHC restricted way (Neumann et al., 2002). This is diVerent for CD4þT lymphocytes, which can only lyse antigen presenting macrophages or microglia and astrocytes (Sun and Wekerle, 1986). As discussed earlier, passive transfer of autoreactive CD8þ T lymphocytes can lead to brain inXammation with very selective destruction of the respective antigen-containing target cells. The immunological requirements for such a direct T-cell-mediated killing is the expression of MHC class I molecules on the surface of the target cells, allowing speciWc antigen presentation and the interaction with T cells via an ‘‘immunological synapse’’ (Grakoui et al., 1999). T cells can kill their targets either through their cytotoxic granules, which mainly contain perforin and granzymes. Perforin impairs the membrane stability of the target cells, while granzymes, getting access to the cell cytoplasm, start the apoptotic cascade. This is a rapid process, leading to cell destruction within less than 1 hour. In addition, T cells can destroy their targets also through the activation of death receptors, belonging to the Fas- or TNF receptor family. This death pathway is more protracted, the target cells being destroyed within several hours after the insult (for review, see Neumann et al., 2002).

1055

1056

43. MYELIN AND ITS DISEASES

EAE is generally regarded as a disease mediated by class II restricted Th2 cells. Most of the immunological data in this respect, however, were obtained from EAE models induced by sensitization with myelin basic protein. Since MBP is at least in part expressed in peripheral lymphatic tissue (Voskuhl, 1998), immune control of class I restricted T-cell reactions against this antigen may be under strict regulation. In line with this concept, a signiWcant immune response of CD8þ T cells can be induced only in MBP-deWcient animals (Huseby et al., 2001). However, sensitization with MOG, which is not expressed in peripheral tissues of mice, leads to a profound class I MHC restricted immune response, which was shown to be encephalitogenic after passive transfer of the respective T cells. Thus, these data suggest that cytotoxic T lymphocytes may play a major pathogenic role in chronic MOG-induced EAE in mice, while in other EAE models their role may be limited.

Antibody-Mediated Tissue Injury As mentioned earlier, antibodies directed against target antigens, which are accessible from the extracellular space in vivo, may modulate clinical disease and pathology in T-cellmediated EAE. This process has most intensively been investigated for demyelinating antibodies, which have been recognized already in early studies of EAE (Bornstein and Appel, 1961). Antibody-mediated tissue injury in the course of T-cell-mediated EAE leads to a very selective destruction of the target structures and plays a key role in the pathogenesis of the extensive and widespread demyelination, present in MOG induced EAE in rats (Storch et al., 1998) or primates (Genain et al., 1999) and in guinea pigs after sensitization with whole CNS tissue or myelin (Lebar et al., 1986, Linington and Lassmann, 1987, Figs. 43.3 and 43.6). Although mice with chronic EAE may too develop a potential demyelinating antibody response, its role in the pathogenesis of the lesions is less evident. This is best exempliWed in anti-MOG B-cell transgenic animals (Litzenburger et al., 1998). These animals develop a massive antibody response against MOG, which recognizes a surface determinant of the molecule and is thus potentially pathogenic. In spite of this, these animals only show a moderate increase in demyelination in comparison to wild-type animals with the same T-cell-mediated EAE. The reason for this rather poor antibody-mediated demyelinating response may reside in the relative ineYcacy of most mouse strains in the antibodymediated activation of the complement system. Tissue damage through antibodies may be accomplished by two diVerent ways. Complement is activated by the immune complexes, leading to accumulation of the terminal lytic complement complex on the surface of the target cells. This will lead to disturbance of membrane homeostasis and lysis of the respective cells. Thus, blockade of the complement system is an eVective way to prevent antibody-mediated demyelination in EAE (Piddlesden et al., 1994). In addition, macrophages or microglia express Fc- and complement-receptors, which allow them to attach to antibody targeted myelin and initiate its destruction (Brosnan et al., 1977).

Tissue Injury through Activated Macrophages or Microglia Cells Macrophages or activated microglia cells play a central role in EAE. The severity of clinical disease as well as the amount of tissue damage within the CNS correlates signiWcantly with the extent of macrophage, but not of T-cell inWltration in the lesions (Berger et al., 1997). In addition, deletion of macrophages or blockade of macrophage function ameliorate disease and tissue damage (Huitinga et al., 1990). As shown by studies performed in vitro and in vivo, activated macrophages produce a variety of diVerent toxic mediators, which are not only instrumental in the defense against foreign pathogens but can also destroy local tissue in the course of an inXammatory reaction. Proteases Therapeutic intervention in EAE with protease inhibitors ameliorates clinical disease and tissue damage in the CNS (Brosnan et al., 1980). Based on these Wndings the role of

MECHANISMS OF TISSUE DAMAGE IN INFLAMMATORY LESIONS OF THE CNS

FIGURE 43.6 Primary demyelination in EAE, induced by a cooperation of encephalitogenic T cells and demyelinating antibodies in a guinea pig, sensitized with whole CNS tissue. Numerous demyelinated axons are embedded in a dense glial scar tissue. Toluidine blue stained plastic section;
2000.

proteases in the pathogenesis of this disease and as possible targets for therapy has been extensively studied (reviewed in Cuzner and Opdenakker, 1999). Various diVerent proteases are produced and expressed within the lesions, predominantly within activated inXammatory cells such as T cells, macrophages or, when present, granulocytes (Clements et al., 1997; Teesalu et al., 2001), as well as in astrocytes (Teesalu et al., 2001), and can also be detected in the cerebrospinal Xuid (Gijbels et al., 1993). Their biological activity in the lesions was shown by in situ zymography (Teesalu et al., 2001). They are in general secreted as inactive precursor and have to be cleaved to become biologically active (Cuzner and Opdenakker, 1999). Furthermore, their activity within the lesions is strictly controlled by speciWc inhibitors, such as tissue inhibitors of metalloproteases (TIMPs; Cuzner and Opdenakker, 1999), which are also expressed within the lesions, in particular in astrocytes (Teesalu et al., 2001). Thus, their biological activity in the lesions depends on the balance between the active enzyme and its inhibitor. Proteases are involved in several essentially diVerent steps in the development of EAE lesions. First, they are required for the migration of inXammatory cells through the blood brain barrier. Second, they may cleave extracellularely liberated CNS proteins, such as for instance MBP, and thus increase the amount of peptides, which may be presented to T cells (Fabry et al., 1994; Opdenakker and Van Damme, 1994). Both eVects will potentiate the inXammatory reaction and a blockade of proteases will have an anti-inXammatory eVect. Third, proteases liberated in the lesions may directly be involved in tissue destruction.

1057

1058

43. MYELIN AND ITS DISEASES

Direct intracerebral injection of proteases in concentrations, similar to those present in inXammatory lesions, induce myelin and axonal damage (Anthony et al., 1998). Protease inhibition may thus be a valuable target for anti-inXammatory therapy. This has been established in diVerent EAE models in vivo, in which treatment with various protease inhibitors has a beneWcial eVect on clinical disease and tissue damage (Clements et al., 1997; Gijbels et al., 1994). Interestingly, some of the eVects of b-interferon on brain inXammation may also be mediated through its inhibition of matrix metalloprotease activity (Leppert et al., 1996). Tumor Necrosis Factor-Alpha (TNF-a) TNF-a is a key cytokine, regulating inXammation and tissue damage in brain inXammation (for review, see Probert et al., 2000). It is highly expressed in the lesions of EAE, in particular in the early stages of their evolution (Issazadeh et al., 1995a). The cellular sources of TNF-a in inXammatory brain lesions are lymphocytes and macrophages as well as activated microglia (Frei et al., 1987) and astrocytes (Liebermann et al., 1989). TNF-a exerts serveral pro-inXammatory actions, such as the induction of adhesion molecules and chemokines (Butcher and Picker, 1996), and may induce macrophage activation. In addition, TNF-a and lymphotoxin-alpha (LT-a) may directly induce cell damage by the activation of the death domain of the TNF-receptor-1. This may occur in oligodendrocytes in vitro (D’Souza et al., 1995; Selmaj and Raine, 1988) and may lead to primary demyelination (Selmaj and Raine, 1988). Over-expression of TNF-a in vivo in transgenic animals leads to demyelination and apoptosis of oligodendrocytes (Akassoglou et al., 1998) or when expressed only at low levels in the CNS may, although not being pathogenic by itself, aggravate demyelination in EAE (Taupin et al., 1997). For these reasons, TNF-a has long been regarded as a prime candidate for therapeutic intervention in inXammatory demyelinating diseases. However, treatment of EAE animals with antibodies against TNF-a or TNF-R1 blocking agents revealed conXicting results (for review, see Probert et al., 2000). While most studies showed a beneWcial eVect some studies revealed the opposite (Willenborg et al., 1995). The situation may be in part explained by results obtained in transgenic models with a deletion of TNF/LT or the TNF-R1. When EAE is induced in these animals, demyelination and tissue destruction are reduced by 80% in comparison to that in wild-type controls. On the other hand, the degree of brain inXammation is signiWcantly increased (Eugster et al., 1999), and local destruction of T cells in the lesions by apoptosis is reduced (Bachmann et al., 1999). Thus, as recently reviewed by Probert et al. (2000), signaling through the TNFR1 pathway in EAE may have a dual function. It may thus be pro-inXammatory and may induce selective tissue damage, but simultaneously it may have an anti-inXammatory eVect, possibly being involved in the control of T-cell autoimmunity. Importantly, these TNFR1-mediated eVects on EAE diVer between animal strains (Kassiotis et al., 1999; Ko¨rner et al., 1997; Liu et al., 1998), underlining the genetic inXuence on the mechanisms of inXammation and tissue damage in EAE. Reactive Oxygen and Nitrogen Species Reactive oxygen and nitrogen species (ROS and RNI) are important toxic mediators produced mainly by activated leukocytes. They may play a major role in the pathogenesis of inXammatory demyelinating diseases, both in humans and experimental models (for review, see Smith et al., 1999; Smith and Lassmann, 2002; Willenborg et al., 1999). ROS may directly destroy target cells through direct membrane damage by lipid peroxidation and the induction of apoptosis. Oligodendrocytes and myelin are particularly vulnerable to their action, possible due to the elaborate cell processes built by these cells (Kim and Kim, 1991; Noble et al., 1994). Thus, therapies that limit superoxide production show consistent beneWcial eVects in EAE (Smith et al., 1999). The situation with reactive nitrogen intermediates is more complex. They may either alone or in combination with ROS interfere with the pathogenesis of EAE lesions at a variety of diVerent levels. Similar as ROS nitric oxide radicals induce oligodendrocyte death in vitro, these cells being much more sensitive compared to astrocytes and microglia

MECHANISMS OF TISSUE DAMAGE IN INFLAMMATORY LESIONS OF THE CNS

(Merrill et al., 1993). In addition RNIs can impair axonal function. At low concentration, they can functionally block axon conduction and may thus be involved in the initiation of clinical deWcit. This eVect is much more pronounced in demyelinated compared to intact axons (Kapoor et al., 1999). At higher concentrations and in particular when the axons are electrically active, NO induces permanent degeneration of axons (Smith et al., 2001). The mechanisms of RNI action are complex and may involve direct interference with ion channels (Bielefeldt et al., 1999), they may induce DNA strand breaks (Zhang et al., 1994), impair mitochondrial function (Bolanos et al., 1997), or exert an indirect toxicity through activation of matrix metalloproteinases (Maeda et al., 1998). Besides their possible involvement in demyelination and axonal injury, RNIs may also aVect the gray matter. When produced in inXammatory lesions, they may directly impair synaptic transmission and may induce neuronal damage and destruction (Smith and Lassmann, 2002). Thus, it is likely that RNIs play a major role in several diVerent aspects of brain inXammation, leading to both structural damage and functional deWcit and are therefore very attractive candidates for therapeutic intervention. Unfortunately, studies on iNOS blockade in EAE animals revealed very conXicting results and many studies unexpectedly showed disease aggravation instead of amelioration (Smith and Lassmann, 2002; Willenborg et al., 1999). The reason for this appears to be that RNIs have additional functions in controlling the inXammatory process. Several studies suggest that T-cell activation (Albina and Henry, 1991) and the expression of endothelial adhesion molecules (Kubes et al., 1991) are inhibited by NO derivatives and that these compounds are involved in the apoptotic elimination of T cells from the lesions (Okuda et al., 1997; Zettl et al., 1997). Thus, as discussed earlier for TNF, RNIs also exert pro- as well as anti-inXammatory actions besides their direct cytotoxic role in the destruction of the target tissue. A simple therapy by just blocking nitric oxide production is thus unlikely to succeed. Excitotoxins: Tissue damage in EAE lesions is in part mediated also by excitotoxins. Early studies showed a beneWcial eVect on EAE by treatment with memantine, an inhibitor of NMDA receptors (Wallstrom et al., 1996). More recently, a profound eVect of AMPA/Kainate receptor blockade on the clinical course of EAE was described (Pitt et al., 2000; Smith et al., 2000). It is well established that nerve cells can be destroyed by over-excitation through glutamate receptor agonists. During the last years, however, it became clear that oligodendrocytes too are highly vulnerable for excitotoxic injury in vitro and in vivo. Some data also indicate that glutamate receptors may be expressed on axons, which raises the possibility that their activation may also be involved in axonal injury. In line with these data the clinical eVects of AMPA/Kainate receptor blockade in EAE was associated with signiWcant reduction in the extent of demyelination as well as axonal and neuronal injury (Pitt et al., 2000; Smith et al., 2000). The source of excitotoxins in EAE lesions is not entirely clear. In part, they may be released by injured neurons and axons. In addition, however, activated macrophages and microglia cells have been shown to be a potent source of exctitotoxic molecules, such as for instance quinolinic acid (Lipton, 1998).

Other Mechanisms of Tissue Injury Depending on the model of EAE, other mechanisms can contribute to tissue injury and destruction. Severe lesions of acute and chronic EAE, in particular after immunization with MOG, may contain a prominent number of granulocytes and also eosinophils in the inXammatory inWltrates (Storch et al., 1996). These cells can produce a variety of very potent cytotoxic mediators, which in the case of eosinophils may be particularly toxic for the CNS tissue. In other models, hypoxic or ischemic tissue damage may occur. Massive blood brain barrier damage and edema is an important component of brain inXammation. If the swelling of edematous tissue is limited by bony or meningeal constrains, cerebral vessels may become compressed, resulting in disturbance of microcirculation (Prineas and

1059

1060

43. MYELIN AND ITS DISEASES

McDonald, 1997). In addition, antigen, which is liberated in the CNS, may diVuse to the vessel wall and may there be recognized by speciWc antibodies or cytotoxic T cells. This may result in an inXammatory response in the vessel wall itself, which may impair microcirculation. Finally, mitochondrial function can be disturbed by RNIs (Bolanos et al., 1997), locally produced by activated macrophages. Mitchondrial dysfunction may result in a hypoxia-like tissue damage.

REMYELINATION IN EAE LESIONS In contrast to the situation in multiple sclerosis, spontaneous remyelination in EAE is extensive and inactive completely demyelinated lesions are rare in EAE models (Lassmann, 1983) (Fig. 43.8). Although in EAE, induced by active immunization against MOG, extensive conXuent demyelination occurs, already at very early stages of lesion formation oligodendrocytes reappear and start to form new thin myelin sheaths (Storch et al., 1996). This happens even in lesions, which still contain abundant macrophages with myelin degradation products. This indicates that remyelination starts within a few days following the demyelinating insult. Remyelination may be accomplished by either oligodendrocytes or Schwann cells. The latter is mainly found in spinal cord lesions and depends on profound astroglial injury during the acute phase of demyelination (Lassmann, 1983). As long as the subpial astroglial limiting membrane remains intact, peripheral remyelination is absent. The source of remyelinating cells in EAE lesions in vivo is still unresolved. In vitro data as well as results from toxic models of demyelination suggest that remyelination can only be accomplished by undiVerentiated progenitor cells, which still have the capacity to proliferate (Blakemore and Keirstead, 1999). In line with this concept, numerous progenitor cells are found in acutely demyelinated EAE lesions (DiBello et al., 1999). Whether mature oligodendrocytes that have survived the demyelinating attack are also able to dediVerentiate and form new remyelinating oligodendrocytes is still unresolved. Irrespective of their source, remyelinating oligodendrocytes are abundant in EAE lesions and are present even in the earliest stages of lesion formation (Fig. 43.9). Remyelination in EAE in vivo can be stimulated further by growth factors (Canella et al., 1998; McMorris and McKinnon, 1996; RuYni et al., 2001) or immunoglobulins (Rodriguez and Miller, 1994). The interpretation of such experiments, however, is diYcult for several reasons. First, spontaneous remyelination in EAE models is very eVective and it may be hard to detect any additional remyelination induced by the therapy. Second, as will be discussed later, neurotrophins have additional functions on the immune system, which

FIGURE 43.8 Remyelination in EAE, shown in a lesion from a guinea pig with chronic disease, following sensitization with whole CNS tissue. In comparison to the normal myelin, present in the left half of the Wgure, remyelinated Wbers are characterized by abnormally thin myelin sheaths (A). In higher magniWcation the nonproportionally thin myelin sheaths can be appreciated (B). Toluidine blue-stained plastic sections. (A)
1200. (B)
3000.

THE GOOD SIDE OF INFLAMMATION: PROTECTIVE AUTOIMMUNITY

FIGURE 43.9 Oligodendrocytes in an actively demyelinating EAE lesions stained by in situ hybridization for PLP mRNA (black cells). In comparison to the normal white matter, the density of oligodendrocytes is reduced in the lesions. However, even during the process of active demyelination, there are oligodendrocytes present, which contain myelin protein mRNAs and in some areas they form clusters of high density. In situ hybridization for PLP mRNA (black) and immunocytochemistry for PLP protein (red);
60.

may interfere with the development of the disease. Thus, an eVect of the treatment on clinical disease or lesion volume in the CNS does not necessarily imply that this is due to remyelination. There are only few instances where remyelination is impaired in EAE lesions. Obviously in very destructive EAE lesions, remyelination is limited by axonal loss. In addition, repeated demyelinating episodes within the same area of the CNS reduces the remyelinating capacity of the tissue. This is best exempliWed in the model of repeated co-transfer EAE with encephalitogenic T cells and demyelinating anti-MOG antibodies. While after the Wrst co-transfer remyelination occurs rapidly and completely, after each subsequent co-transfer of T cells and antibodies, the remyelinating capacity of the tissue declines (Linington et al., 1992). Although not formally proven, these data suggest that the pool of progenitor cells that may accomplish remyelination is limited and gradually becomes used up in the course of repeated demyelinating episodes. Another model that shows impaired remyelination is MOG-induced EAE in the Brown Norway (BN) rat. These rats diVer from other rat strains in two ways. First, immunization with MOG in this strain results in a vigorous demyelinating antibody response, which is more severe compared to other rat strains (SteVerl et al., 1999). Second, these animals express MOG at an earlier stage of myelination and remyelination in comparison to other rat strains. The consequence of these two factors is that MOG is expressed in remyelinating lesions already at a time when the blood brain barrier is still impaired and anti-MOG antibodies can reach the lesions in high concentration. These remyelinating oligodendrocytes are then rapidly destroyed by antibodies and complement (Storch et al., unpublished). Although these data give some insights as to how remyelination can be impaired in inXammatory demyelinating lesions, there must be additional other mechanisms that are responsible for the low degree of remyelination in chronic MS plaques, which so far are not reproduced in EAE.

THE GOOD SIDE OF INFLAMMATION: PROTECTIVE AUTOIMMUNITY For many years, inXammation has been considered to inevitably be harmful for the central nervous system. On the other hand, it was diYcult to explain why nature permitted brain

1061

1062

43. MYELIN AND ITS DISEASES

autoimmunity to occur with relative ease. In other words, liberation of antigen in the central nervous system can induce autoimmune reactions on the T-cell as well as antibody level. The reason for this apparent discrepancy may be found in recent immunological studies, which show that inXammatory cells can produce a variety of cytokines or growth factors that help to repair damaged tissue (Schwartz, 2001). Cells of the immune system, including T cells, B cells, and macrophages, in particular when they are activated, can produce neurotrophic factors (Besser and Wank, 1999; Kerschensteiner et al., 1999; Moalem et al., 2000). These neurotrophic factors may exert their action not only on the nervous tissue, but also on the immune system. Within the nervous system, activated leukocytes within an inXammatory focus may be the major source of neurotrophins, being more important than resident neurons or glia cells, and leukocyte-derived neurotrophins are biologically active and can rescue neurons in vitro (Kerschensteiner et al., 1999). In addition, neurotrophins may directly act on cells of the immune system, mainly having anti-inXammatory functions. As an example, nerve growth factor can modulate EAE by its direct action on macrophages through the p75 NT-receptor (Flu¨gel et al., 2001). It inhibits MHC-expression and impairs immigration of inXammatory cells into the brain, possibly by down-regulating adhesion molecule expression. Recent studies suggest that inXammatory reactions in the brain may stimulate regeneration and repair of CNS lesions. In experimental models of brain or spinal cord trauma, clinical deWcit is less pronounced in animals with mild T-cell-mediated autoimmune inXammation (Hamarberg et al., 2000; Moalem et al., 1999). This is associated with signiWcantly less tissue damage and attempts of axonal regeneration and sprouting. The mechanisms behind these eVects are apparently mediated by neurotrophins, produced by inXammatory cells in these conditions. Functional blockade of neurotrophins with sepciWc antibodies abolishes the beneWcial eVect of inXammation. InXammation apparently has a major eVect on remyelination in demyelinating lesions. Remyelination is stimulated by macrophages in vitro (Diemel et al., 1998), and macrophage depletion blocks remyelination in a model of toxic demyelination (Kotter et al., 2001). Interestingly, this eVect may in part be due to the action of TNF-a, which may promote the proliferation of oligodendrocyte progenitors as well as remyelination (Arnett et al., 2001). Thus, the same cytokine may be cytotoxic in mature oligodendrocyte but stimulate the recruitment of new oligodendrocytes from progenitors. Although the concept that inXammation may promote tissue repair is not new and has been recognized for long time as a mechanism of wound healing, its importance in inXammatory CNS diseases has only recently received attention. This has major consequences. In a disease like multiple sclerosis, remyelination is extensive in acute and in active lesions of chronic disease, but is sparse in chronic inactive demyelinated plaques. Active remyelination thus occurs on a background of profound inXammation. This is similar in EAE models, where remyelination is present and extensive in lesions that contain profound inXammatory inWltrates. In the light of these Wndings, it may turn out that complete blockade of inXammation in multiple sclerosis patients by extensive immunosuppression may be counterproductive. Thus, changing the quality of inXammation by immune modulation may be the better therapeutic strategy than complete immunosuppression.

EAE IS AN ENVIRONMENTALLY INDUCED DISEASE WITH POLYGENIC BACKGROUND Although T-cell or antibody-mediated autoimmune reactions have been described to occur following trauma or viral infections of the CNS, so far no spontaneous form of autoimmune encephalomyelitis has been described in any animal species. Active immunization with nervous system antigens is therefore essential for the induction of autoimmune encephalomyelitis. In addition, however, the incidence, severity and phenotypic manifestation of disease is highly inXuenced by the genetic background of the animal. The genetic

CONCLUSIONS

regulation of EAE is very complex and its complete review would go beyond the scope of this chapter. Thus, only some principal aspects will be discussed (for more detailed review, see Olsson et al., 2000, and Sundvall et al., 1995). The gene region with the most prominent eVect on disease susceptibility in EAE is the MHC region (Encinas et al., 1996; Olsson et al., 2000). Within the MHC complex, class II genes have the strongest eVect and MHC-controlled high disease susceptibility is associated with increased production of pro-inXammatory Th1 cytokines by encephalitogenic T cells. Thus, a major inXuence of the MHC region is due to its qualitative and quantitative control of T-cell antigen recognition (Weissert et al., 1998). A detailed analysis of intraMHC recombinant rat strains, however, revealed additional eVects, mediated by genes in the class I and class III region (Weissert et al., 1998). Whether these eVects reXect a pathogenic contribution of class I MHC restricted T cells is so far undetermined, since the MHC complex contains numerous other genes, which are involved in the control of immune function. Besides the control by the MHC, complex EAE is regulated by many other, non-MHC genes (Dahlmann et al., 1999). Genome wide screening of EAE animals revealed more than 10 additional susceptibility loci with inXuence on disease incidence, chronicity, or severity. Interestingly, many of these gene regions are not speciWc for EAE, but are additionally associated with other autoimmune diseases, such as autoimmune neurits or uveitis, adjuvant or collagen-induced arthritis, or insulin-dependent diabetes melitus, and synthenic regions appeared in the genomic screen of multiple sclerosis patients (Olsson et al., 2000). The individual genes and their possible function are so far undetermined but may be involved in general immune regulation, in the mechanisms of immune-mediated tissue damage, or may determine the susceptibility of the nervous tissue for inXammationinduced damage. There are several possible examples for such eVects. The major diVerences in the extent of antibody-mediated demyelination in MOG-EAE between diVerent species and strains may be due to the genetically determined ability of antibodies to activate complement and by polymorphisms in the MOG gene itself (SteVerl et al., 1999). In addition, the type and activation stage of eVector cells within EAE lesions seems to be genetically controlled. While classical EAE in rats is associated with massive recruitment of haematogenous macrophages, in Wistar rats lesions are characterized by massive T-cell inWltration, microglia activation but nearly complete absence of macrophage recruitment (Storch et al., 2002). Detailed genetic mapping revealed that this eVect is associated with genes in the class II and the nonclassical class I region of the MHC complex. This diVerence in eVector cells in the lesions has a major inXuence on the type of immune-mediated tissue damage. In spite of a similar extent of demyelination, axonal injury is much more severe in animals with macrophage-dominated lesions, compared to those where microglia prevail (Storch et al., 2002). Susceptibility of the target tissue may also be controlled by neurotrophic factors. As an example, the extent of oligodendrocyte damage and demyelination in EAE is much more severe in animals with a deWciency for ciliary neurotrophic factor compared to intact controls (Linker et al., 2002).

CONCLUSIONS EAE research was started with the naı¨ve approach to develop a simple model for multiple sclerosis that allows to unravel its pathogenesis and evaluate preclinically new therapeutic strategies. After nearly 70 years of EAE research, we have learned a lot about basic immunology, the mechanisms of autoimmunity, and the pathogenesis of brain inXammation. It, however, also became clear that EAE is a very complex disease with a broad spectrum of diVerent immunological mechanisms, their pathogenic contribution being determined by the genetic background of the animal and the mode of disease induction. Thus, there is no single EAE. Instead there are multiple diVerent models, each of which

1063

1064

43. MYELIN AND ITS DISEASES

allows us to study only certain speciWc aspects of brain inXammation or inXammationinduced tissue damage. This is extremely useful for elucidating basic pathogenetic mechanisms involved in inXammatory brain diseases and can also be utilized to test the principal feasibility of therapeutic approaches. However, since human inXammatory brain diseases and in particular multiple sclerosis are similarly complex, its use for screening of new MS therapies deserves a critical and cautious approach.

References Akassoglou, K., Bauer, J., Kassiotis, G., Pasparakis, M., Lassmann, H., Kollias, G., and Probert, L. (1998). Oligodendrocyte apoptosis and primary demyelination induced by local TNF/p55TNF receptor signaling in the central nervous system of transgenic mice: Models for multiple sclerosis with primary oligodendrogliopathy. Amer. J. Pathol. 153, 801–813. Albina, J. E., Henry, W. L Jr. (1991). Suppression of lymphocyte proliferation through the nitric oxide synthesizing pathway. J Surg Res. 50, 403–409. Alvord, E. C. Jr. (1970). Acute disseminated encephalomyelitis and ‘‘allergic’’ encephalopathies. In ‘‘Handbook of Clinical Neurology’’ (P. I. Vinken and G. W. Bruyn, eds.), Vol. 9, pp. 500–571. Elsevier, New York. Alvord, E. C. Jr. (1984). Species restricted encephalitogenic determinants. In ‘‘Experomental Allergic Encephalomyelitis: A Useful Model for Multiple Sclerosis’’ (E. C. Alvord, M. W. Kies, A. J. Suckling, eds.), pp. 523–537, Alan R. Liss, New York. Anthony, D. C., Miller, K. M., Fearn, S., Townsend, M. J., Opdenakker, G., Wells, G. M., Clements, J. M., Chandler, S., Gearing, A. J., and Perry, V. H. (1998). Matrix metalloproteinase expression in an experimentally-induced DTH model of multiple sclerosis in the rat CNS. J Neuroimmunol. 87, 62–72. Arnett, H. A., Mason, J., Marino, M., Suzuki, K., Matsushima, G. K., and Ting, J. P. (2001). TNF alpha promotes proliferation of oligodendrocyte progenitors and remyelination. Nat Neurosci. 4, 1116–1122. Bachmann, R., Eugster, H. P., Frei, K., Fontana, A., and Lassmann, H. (1999). Impaiment of TNF-receptor 1 signalling but not Fas-signalling diminishes T-cell apoptosis in MOG-peptide induced chronic demyelinating autoimmune encephalomyelitis in mice. Amer J Pathol 154, 1417–1422. Bauer, J., Bradl, M., Hickey, W. F., Forss-Petter, S., Breitschopf, H., Linington, C., Wekerle, H., and Lassmann, H. (1998). T cell apoptosis in inXammatory brain lesions. Destruction of T cells does not depend on antigen recognition. Am J Pathol 153, 715–724. Bechmann, I.; Lossau, S., Steiner, B., Mor, G., Gimsa, U., and Nitsch, R. (2000). Reactive astrocytes upregulate Fas (CD95) and Fas ligand (CD95L) expression but do not undergo programmed cell death during the course of anterograde degeneration. Glia. 32, 25–41. Ben Nun, A., Wekele H., and Cohen, I. R. (1981). The rapis isolation of clonable antigen-speciWc T lymphocyte lines capable of mediating autoimmune encephalomyelitis. Eur. J. Immunol. 11, 195–199. Berger, T., Weerth, S., Kojima, K., Linington, C., Wekerle, H., and Lassmann, H. (1997). Experimental autoimmune encephalomyelitis: The antigen speciWcity of T-lymphocytes determines the topography of lesions in the central and peripheral nervous system. Lab Invest 76, 355–364. Besser, M., and Wank, R. (1999). Cutting edge: Clonally restricted production of the neurotrophins brain-derived neurotrophic factor and neurotrophin-3 mRNA by human immune cells and Th1/Th2-polarized expression of their receptors. J Immunol. 162, 6303–6306. Bhardwaj, V., Kumar, V., Geysen, H. M., and Sercarz, E. E. (1993). Degenerate recognition of a dissimilar antigenic peptide by myelin basic protein-reactive T cells. Implications for thymic education and autoimmunity. J Immunol. 151, 5000–5010. Bielefeldt, K., Whiteis, C. A., Chapleau, M. W., and Abboud, F. M. (1999). Nitric oxide enhances slow inactivation of voltage-dependent sodium currents in rat nodose neurons. Neurosci Lett. 271, 159–162. Blakemore, W. F., and Keirstead, H. S. (1999). The origin of remyelinating cells in the central nervous system. J Neuroimmunol. 98, 69–76. Bolanos, J. P., Almeida, A., Stewart, V., Peuchen, S., Land, J. M., Clark, J. B., and Heales, S. J. (1997). Nitric oxide-mediated mitochondrial damage in the brain: Mechanisms and implications for neurodegenerative diseases. J Neurochem. 68, 2227–2240. Bornstein, M. B., and Appel S. H. (1961). The application of tissue culture to the study of experimental ‘‘allergic’’ encephalomyelitis. I. Patterns of demyelination. J. Neuropathol. Exp. Neurol. 20, 141–147. Brosnan, C. F., Cammer, W., Norton, W. T., and Bloom, B. R. (1980). Proteinase inhibitors suppress the development of experimental allergic encephalomyelitis. Nature. 285, 235–237. Brosnan, C. F., Stoner, G. L., Bloom, B. R., and Wisniewski, H. M. (1977). Studies on demyelination by activated lymphocytes in the rabbit eye. II. Antibody dependent cell mediated demyelination. J. Immunol. 118, 2103–2111. Butcher, E. C., and Picker, L. J. (1996). Lymphocyte homing and homeostasis. Science 272, 60–66. Cannella, B., Hoban, C. J., Gao, Y. L., Garcia-Arenas, R., Lawson, D., Marchionni, M., Gwynne, D., and Raine, C. S. (1998). The neuregulin, glial growth factor 2, diminishes autoimmune demyelination and enhances remyelination in a chronic relapsing model for multiple sclerosis. Proc Natl Acad Sci US 95, 10100–10105. Chase, W. M. A. (1959). A critique of attempts at passive transfer of sensitivity to nervous tissue. In ‘‘Allergic Encephalomyelitis’’ (M. vW. Kies and E. vC. Alvord, eds.), pp. 348–374. Thomas, SpringWeld, IL.

CONCLUSIONS

Clements, J. M., Cossins, J. A., Wells, G. M., Corkill, D. J., Helfrich, K., Wood, L. M., Pigott, R., Stabler, G., Ward, G. A., Gearing, A. J., and Miller, K. M. (1997). Matrix metalloproteinase expression during experimental autoimmune encephalomyelitis and eVects of a combined matrix metalloproteinase and tumour necrosis factor-alpha inhibitor. J Neuroimmunol. 74, 85–94. Cornet, A., Savidge, T. C., Cabarrocas, J., Dend, W. L., Colombel, J. F., Lassmann, H., Desreumaux, P., and Liblau, R. S. (2001). Enterocolitis induced by autoimmune targeting of enteric glia cells: A possible mechanism in Crohn’s disease? Proc Natl Acad Sci (USA) 98, 13306–13311. Cuzner, M. L., and Opdenakker, G. (1999). Plasminogen activators and matrix metalloproteinases, mediators of extracellular proteolysis in inXammatory demyelination of the central nervous system. J. Neuroimmunol. 94, 1–14. Dahlmann, I., Wallstrom E., Weissert, R., Storch, M., Kornek, B., Jacobsson, L., Linington, C., Luthman, H., Lassmann, H., and Olsson, T. (1999). Linkage analysis of myelin oligodendrocyte glycoprotein-induced experimental autoimmune encephalomyelitis in the rat identiWes a locus controlling demyelination on chromosome 18. Hum Mol Genet 8, 2183–2190. DiBello, I. C., Dawson, M. R., Levine, J. M., and Reynolds, R. (1999). Generation of oligodendroglial progenitors in acute inXammatory demyelinating lesions of the rat brain stem is associated with demyelination rather than inXammation. J. Neurocytol. 28, 365–381. Diemel, L. T., Copelman, C. A., and Cuzner, M. L. (1998). Macrophages in CNS remyelination: Friend or foe? Neurochem-Res. 23, 341–347. D’Souza, S., Alinauskas, K., McCrea, E., Goodyer, C., and Antel, J. P. (1995). DiVerential susceptibility of human CNS-derived cell populations to TNF-dependent and independent immune-mediated injury. J Neurosci. 15, 7293–7300. Dubois-Dalcq, M., Niedieck, B., and Buyse, M. (1970)/ Action of anti-cerebroside sera on myelinated nervous tissue cultures. Pathologica Europea 5, 331–347. Encinas, J. A., Weiner, H. L., and Kuchroo, V. K. (1996). Inheritance of susceptibility to experimental autoimmune encephalomyelitis. J Neurosci Res. 45, 655–669. Engelhardt, B., Conley, F. K., and Butcher, E. C. (1994). Cell adhesion molecules on vessels during inXammation in the mouse central nervous system. J Neuroimmunol. 51, 199–208. Eugster, H. P., Frei, K., Bachmann, R., Bluethmann, H., Lassmann, H., and Fontana, A. (1999). Severity of symptoms and demyelination in MOG-induced EAE depends on TNFR1. Eur. J. Immunol. 29, 626–632. Evans, C. F., Horwitz, M. S., Hobbs, M. V., and Oldstone, M. B. (1996). Viral infection of transgenic mice expressing a viral protein in oligodendrocytes leads to chronic central nervous system autoimmune disease. J Exp Med. 184, 2371–2384. Fabry, Z., Raine, C. S., and Hart, M. N. (1994). Nervous tissue as an immune compartment: The dialect of the immune response in the CNS. Immunol Today. 15, 218–224. Fife, B. T., HuVnagle, G. B., Kuziel, W. A., and Karpus, W. J. (2000). CC chemokine receptor 2 is critical for induction of experiemntal autoimmune encephalomyelitis. J. Exp. Med. 192, 899–905. Fife, B. T., Paniagua, M. C., Lukcs, N. W., Kunkel, S. L., and Karpus, W. J. (2001). Selective CC chemokine receptor expression by central nervous system-inWltrating encephalitogenic T cells during experimental autoimmune encephalomyelitis. J. Neurosci. Res. 66, 705–714. Fischer, H. G., and Reichmann, G. (2001). Brain dendritic cells and macrophages/microglia in central nervous system inXammation. J Immunol 166, 2717–2726. Flugel, A., Berkowicz, T., Ritter, T., Labeur, M., Jenne, D. E., Li, Z., Ellwart, J. W., Willem, M., Lassmann, H., and Wekerle, H (2001). Migratory activity and functional changes of green Xuorescent eVector cells before and during experiemntal autoimmune encephalomyelitis. Immunity 14, 547–560. Flugel, A., Matsumoro, K., Neumann, H., Klinkert, W. E., Birnbacher, R., Lassmann, H., Otten, U., and Wekerle, H (2001). Anti inXammatory activity of nerve growth factor in experimental autoimmune encephalomyelitis: Inhibition of monocyte transendeothelial migration. Eur J Immunol 13, 11–22. Frei, K., Siepl, C., Groscurth, P., Bodmer, S., Schwerdel, C., and Fontana, A. (1987). Antigen presentation and tumor cytotoxicity by interferon-gamma-treated microglial cells. Eur J Immunol. 17, 1271–1278. Freund, J., Lipton, M. M., and Morrison, L. R. (1950). Demyelination in the guinea pig chronic allergic encephalomyelitis produced by injecting guinea pig brain in oil emulsion containing a variant of mycobacterium butyricum. Arch. Pathol. 50, 108–121. Genain, C. P., Abel, K., Belmar, N., Villinger, F., Rosenberg, D. P., Linington, C., Raine, C. S., and Hauser, S. L. (1996). Late complications of immune deviation therapy in a non-human primate. Science 274, 2054–2057. Genain, C. P., Cannella, B., Hauser, S. L., and Raine, C. S. (1999). Autoantibodies to MOG mediate myelin damage in MS. Nat. Med. 5, 170–175. Gijbels, K., Galardy, R. E., and Steinman, L. (1994). Reversal of experimental autoimmune encephalomyelitis with a hydroxamate inhibitor of matrix metalloproteases. J Clin Invest. 94, 2177–2182. Gijbels, K., Proost, P., Masure, S., Carton, H., Billiau, A., and Opdenakker, G. (1993). Gelatinase B is present in the cerebrospinal Xuid during experimental autoimmune encephalomyelitis and cleaves myelin basic protein. J Neurosci Res. 36, 432–440. Godiska, R., Chantry, D., Dietsch, G. N., and Gray, P. W. (1995) Chemokine expression in murine experimental allergic encephalomyelitis. J. Neuroimmunol. 58, 167–176. Gold, R., Hartung, H. P., and Lassmann, H. (1997). T-cell apoptosis in autoimmune diseases: Termination of inXammation in the nervous system and other sites with specialized immune-defense mechanisms. Trends Neurosci 20, 399–404.

1065

1066

43. MYELIN AND ITS DISEASES

Grakoui, A., Bromley, S. K., Sumen, C., Davis, M. M., Shaw, A. S., Allen, P. M., and Dustin, M. L. (1999). The immunological synapse: A molecular machine controlling T cell activation. Science 285, 221–227. Haase, C. G., Guggenmos, -J., Brehm, U., Andersson, M., Olsson, T., Reindl, M., Schneidewind, J. M., Zettl, U. K., Heidenreich, F., Berger, T., Wekerle, H., Hohlfeld, R., and Linington, C. (2001). The Wne speciWcity of the myelin oligodendrocyte glycoprotein autoantibody response in patients with multiple sclerosis and normal healthy controls. J Neuroimmunol. 114, 220–225. Hammarberg, H., Lidman, O., Lundberg, C., Eltayeb, S. Y., Gielen, A. W., Muhallab, S., Svenningsson, A., Linda, H., van-Der-Meide, P. H., Cullheim, S., Olsson, T., and Piehl, F. (2000). Neuroprotection by encephalomyelitis: Rescue of mechanically injured neurons and neurotrophin production by CNS-inWltrating T and natural killer cells. J Neurosci. 20, 5283–5291. Hickey, W. F., Hsu, B. L., and Kimura, H. (1991). T lymphocyte entry into the central nervous system. J Neurosci Res 28, 254–260. Hickey, W. F., and Kimura, H. (1988). Perivascular microglia cells of the CNS are bone marrow derived and present antigen in vivo. Science 239, 290–293. Hickey, W. F., Osborn, J. F., and Kirby, W. W. (1985) Expression of Ia molecules by astrocytes during acute experiemntal allergic encephalomyelitis. Cell. Immunol. 91, 528–535. Hickey, W. F., Vass, K., and Lassmann, H. (1992). Bone marrow derived elements in the central nervous system: An immunohistochemical and ultrastructural survey of rat chimeras. J Neuropath Exp Neurol 51, 246–256. Hohlfeld, R. (1997). Biotechnological agents for the immunotherapy of multiple sclerosis. Principles, problems and persectives, Brain 120, 865–916. Huang, D., Han, Y., Rani, M. R., Glabinski, A., Trebst, C., Sorensen, T., Tani, M., Wang, J., Chien, P., O’Bryan, S., Bielecki, B., Zhou, Z. L., Majumder, S., and RansohoV, R. M. (2000). Chemokines and chemokine receptors in inXammation of the nervous system: Manifold roles and exquisite regulation. Immunol Rev. 177, 52–67. Huitinga, I., Damoiseaux, J. G., Dopp, E. A., and Dijkstra, C. D. (1993). Treatment with anti-CR3 antibodies ED7 and ED8 suppresses experimental allergic encephalomyelitis in Lewis rats. Eur J Immunol. 23, 709–715. Huitinga. I., Rooijen, N., deGroot, C. J. A., Uitdehaag, B. M. J., and Dijkstra, C. D. (1990). Suppression of experiemntal allergic encephalomyelitis after elimination of macrophages. J. Exp. Med.172, 1025–1033. Huseby, E. S., Liggitt, D., Brabb, T., Schnabel, B., Ohlen, C., and Goverman, J. (2001). A pathogenic role for myelin-speciWc CD8 (þ) T cells in a model for multiple sclerosis. J Exp Med 194, 669–676. Issazadeh, S., Ljungdahl, A., Ho¨jeberg, B., Mustafa, M., and Olsson, T. (1995a). Cytokine production in the central nervous system of Lewis rats with experimental autoimmune encephalomyelitis: Dynamics of mRNA expression for interleukin 10, interleukin 12, cytolysin, tumor necrosis factor a and tumor necrosis factor ß. J Neuroimmunol. 61, 205–212. Issazadeh, S., Mustafa, M., Ljungdahl, A., Hojeberg, B., Dagerlind, A., Elde, R., and Olsson, T. (1995b). Interferon gamma, interleukin 4 and transforming growth factor beta in experimental autoimmune encephalomyelitis in Lewis rats: Dynamics of cellular mRNA expression in the central nervous system and lymphoid cells. J Neurosci Res. 1995 40, 579–590. Izikson, L., Klein, R. S., Charo, I. F., Weiner, H. L., and luster, A. D. (2000). Resistance to experimental autoimmune encephalomyelitis in mice lacking the CC chemokine receptor (CCR) 2. J. Exp. Med. 192, 1075–1080.. Jankovic, B., Draskoci, M., and Janjic, M. (1965). Passive transfer of ‘‘allergic’’ encephalomyelitis with anti-brain serum, injected into the lateral ventricle of the brain. Nature 207, 428–429. Jee, Y., Yoon, W. K., Okura, Y., Tanuma, N., and Matsumoto, Y. (2002). Upregulation of monocyte chemotactic protein-1 and CC chemokine receptor 2 in the central nervous system is closely associated with relapse of autoimmune encephalomyelitis in Lewis rats. J. Neuroimmunol. 128, 49–57. Jiang, H., Zhang, S. L., and Pernis, B. (1992). Role of CD8þ T cells in murine experimental allergic encephalomyelitis. Science 256, 1213–1215. Kapoor, R., Davies, M., and Smith, K. J. (1999). Temporary axonal conduction block and axonal loss in inXammatory neurological disease. A potential role for nitric oxide? Ann N Y Acad Sci. 893, 304–308. Kassiotis, G., Pasparakis, M., Kollias, G., and Probert, L. (1999). TNF accelerates the onset but does not alter the incidence and severity of myelin basic protein-induced experimental autoimmune encephalomyelitis. Eur J Immunol. 29, 774–780. Kennedy, K. J., Strieter, R. M., Kunkel, S. L., Lucas, N. W., and Karpus, W. J. (1998). Acute and relapsing experimental autoimmune encephalomyelitis are regulated by diVerential expression of the CC chemokines macrophage inXammatory protein-1alpha and monocyte chemotactic protein-1. J. Neuroimmunol. 92, 98–108. Kerschensteiner, M., Gallmeier, E., Behrens, L., Leal, V. V., Misgeld, T., Klinkert, W. E., Kolbeck, R., Hoppe, E., Oropeza-Wekerle, R. L., Bartke, I., Stadelmann, C., Lassmann, H., Wekerle, H., and Hohlfeld, R. (1999). Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor in vitro and in inXammatory brain lesions: A neuroprotective role of inXammation? J Exp Med. 189, 865–870. Kies, M. W., Murphy, J. B., and Alvord, E. C. (1960). Fractionation of guinea pig brain proteins with encephalitogenic activity. Fed. Proc. 19, 207. Kim, Y. S., and Kim, S. U. (1991). Oligodendroglial cell death induced by oxygen radicals and its protection by catalase. J Neurosci Res. 29, 100–106. Koh, D. R., Fung-Leung, W. P., Ho, A., Gray, D., Acha-Orbea, H., and Mak, T. W. (1992). Less mortality but more relapses in experimental allergic encephalomyelitis in CD8 / mice. Science 256, 1210–1213.

CONCLUSIONS

Kojima, K., Berger, Th., Lassmann, H., Hinze-Selch, D., Zhang, Y., Germann, J., Reske, K., Wekerle, H., and Linington, C. (1994). Experimental autoimmune panencephalitis and uveoretinitis transfered to the Lewis rat by T-lymphocytes speciWc for the S100ß molecule, a calcium binding protein of astroglia. J Exp Med 180, 817–829. Korner, H., Lemckert, F. A., Chaudhri, G., Etteldorf, S., and Sedgwick, J. D. (1997). Tumor necrosis factor blockade in actively induced experimental autoimmune encephalomyelitis prevents clinical disease despite activated T cell inWltration to the central nervous system. Eur J Immunol. 27, 1973–1981. Kotter, M. R., Setzu, A., Sim, F. J., Van-Rooijen, N., and Franklin, R. J. (2001). Macrophage depletion impairs oligodendrocyte remyelination following lysolecithin-induced demyelination. Glia. 35, 204–212. Kubes, P., Suzuki, M., and Granger, D. N. (1991). Nitric oxide: An endogenous modulator of leukocyte adhesion. Proc Natl Acad Sci USA. 88, 4651–4655. Lafaille, J. (1998). The role of Helper T cell subsets in autoimmune disease. Cytokine & Growth Factor Reviews 9, 139–151. Lafaille, J. J., Keere, F. V., Hsu, A. L., Baron, J. L., Haas, W., Raine, C. S., and Tonegawa, S. (1997). Myelin basic protein-speciWc T helper 2 (Th2) cells cause experimental autoimmune encephalomyelitis in immunodeWcient hosts rather than protect them from disease. J. Exp. Med. 186, 307–312. Lassmann, H. (1983). ‘‘Comparative Neuropathology of Chronic Experimental Allergic Encephalomyelitis and Multiple Sclerosis.’’ Springer, Berlin, Heidelberg, New York. Lassmann, H., Brunner, C., Bradl, M., and Linington, C. (1988). Experimental allergic encephalomyelitis: The balance between encephalitogenic T-lymphocytes and demyelinating antibodies determines size ans structure of demyelinated lesions. Acta Neuropathol. 75, 566–576. Lassmann, H., Kitz, K., and Wisniewski, H. M. (1981). In vivo eVect of sera from animals with chronic relapsing experimental allergic encephalomyelitis on central and peripheral myelin. Acta. Neuropathol. 55, 297–306. Lebar, R., Lubetzki, C., Vincent, C., Lombrail, P., and Poutry, J. M. (1986). The MS autoantigen of central nervous system myelin, a glycoprotein present in oligodendrocyte membranes. Experimental Immunol. 66, 423–443. Lee, S. J., and Benveniste, E. N. (1999). Adhesion molecule expression and regulation on cells of the central nervous system. J. Neuroimmunol. 98, 77–88. Leppert, D., Waubant, E., Burk, M. R., Oksenberg, J. R., and Hauser, S L. (1996). Interferon beta-1b inhibits gelatinase secretion and in vitro migration of human T cells: A possible mechanism for treatment eYcacy in multiple sclerosis. Ann Neurol. 40, 846–852. Lieberman, A. P., Pitha, P. M., Shin, H. S., Shin, M. L. (1989). Production of tumor necrosis factor and other cytokines by astrocytes stimulated with lipopolysaccharide or a neurotropic virus. Proc Natl Acad Sci USA 86, 6348–6352. Linington, C., Berger, Th., Perry, L., Weerth, S., Hinze-Selch, D., Zhang, Y., Lu, H., Lassmann, H., and Wekerle, H. (1993). T cells speciWc for the myelin oligodendroglia glycoprotein mediate an unusual autoimmune inXammatory response in the central nervous system. Eur J Immunol 23, 1364–1372. Linington, C., Bradl, M., Lassmann, H., Brunner, C., and Vass, K. (1988). Augmentation of demyelination in rat acute allergic encephalomyelitis by circulating antibodies against a myelin oligodendrocyte glycoprotein. Amer. J. Pathol. 130,443–454. Linington, C., Engelhardt, B., Kapocs, G., and Lassmann, H. (1992). Induction of persistently demyelinated lesions in the rat following the repeated adoptive transfer of encephalitogenic T cells and demyelinating antibody. J Neuroimmunol. 40, 219–224. Linington, C., and Lassmann, H. (1987). Antibody responses in chronic relapsing experimental allergic encephalomyelitis: Correlation of serum demyelinating activity with the antibody titre to the myelin/oligodendrocyte glycoprotein (MOG). J Neuroimmunol 17, 61–69. Linker, R. A., Maurer, M., Gaupp, S., Martini, R., Holtmann, B., Giess, R., Rieckmann, P., Lassmann, H., Toyka, K. V., Sendtner, M., and Gold, R. (2002). CNTF is a major protective factor in demyelinating CNS disease: A neurotrophic cytokine as modulator in neuroinXammation. Nat Med. 8, 620–624. Lipton, S. A. (1998). Neuronal injury associated with HIV-1: Approaches and treatment. Annu Rev Pharmacol Toxicol 38, 159–177. Litzenburger, T., Fa¨ssler, R., Bauer, J., Lassmann, H., Linington, C., Wekerle, H., and Iglesias, A. (1998). B lymphocytes producing demyelinating antibodies: Development and function in gene-trageted transgenic mice. J Exp Med 188, 169–180. Liu, J., Marino, M. W., Wong, G., Grail, D., Dunn, A., Bettadapura, J., Slavin, A. J., Old, L., and Bernard, C. C. (1998). TNF is a potent anti-inXammatory cytokine in autoimmune-mediated demyelination. Nat Med. 4, 78–83. Luster, A. D. (1998). Chemokines–chemotactic cytokines that mediate inXammation. N. Engl. J. Med. 338, 436–445. Maatta, J. A., Kaldman, M. D., Sakoda, S., Salmi, A. A., and Hinkkanen, A. E. (1998). Encephalitogenicity of myelin associated oligodendrocytic basic protein and 2’,3’-cyclic nucleotide 3’-phosphodieserase for BALB/c and SJL mice. Immunology 95, 383–388. Maeda, H., Okamoto, T., and Akaike, T. (1998). Human matrix metalloprotease activation by insults of bacterial infection involving proteases and free radicals. Biol Chem. 379, 193–200. Matsui, M., Weaver, J., Proudfoot, A. E. I., Wujek, J. R., Wei, T., Richter, E., Trapp, B. D., Rao, A., and RansohoV, R. M. (2002). Treatment of experiemntal autoimmune encephalomyelitis with the chemokine receptor antagonist Met-RANTES. J. Neuroimmunol. 128, 16–22.

1067

1068

43. MYELIN AND ITS DISEASES

Matsumoto, Y., and Fujiwara, M. (1986). In situ detection of class I and II major histocompatibility complex antigens in the rat central nervous system during experimental allergic encephalomyelitis. An immunohistochemical study. J Neuroimmunol. 12, 265–277. McMorris, F. A., and McKinnon, R. D. (1996). Regulation of oligodendrocyte development and CNS myelination by growth factors: Prospects for therapy of demyelinating disease. Brain Pathol. 6, 313–329. McRae, B. L., Vanderlugt, C. L., Dal-Canto, M. C., and Miller, S. D. (1995). Functional evidence for epitope spreading in the relapsing pathology of experimental autoimmune encephalomyelitis. J Exp Med. 182, 75–85. Merrill, J. E., Ignarro, L. J., Sherman, M. P., Melinek, J., and Lane, T. E. (1993). Microglial cell cytotoxicity of oligodendrocytes is mediated through nitric oxide. J Immunol. 151, 2132–2141. Miller, S. D., and Eagar, T. N. (2001). Functional role of epitope spreading in the chronic pathogenesis of autoimmune and virus-induced demyelinating diseases. Adv Exp Med Biol. 490, 99–107. Miller, S. D., Vanderlugt, C. L., Lenschow, D. J., Pope, J. G., Karandikar, N. J., Dal-Canto, M. C., and Bluestone, J. A. (1995). Blockade of CD28/B7–1 interaction prevents epitope spreading and clinical relapses of murine EAE. Immunity. 3, 739–745. Moalem, G., Gdalyahu, A., Shani, Y., Otten, U., Lazarovici, P., Cohen, I. R., and Schwartz, M. (2000). Production of neurotrophins by activated T cells: Implications for neuroprotective autoimmunity. J Autoimmun. 15, 331–345. Moalem, G., Leibowitz-Amit, R., Yoles, E., Mor, F., Cohen, I. R., and Schwartz, M. (1999). Autoimmune T cells protect neurons from secondary degeneration after central nervous system axotomy. Nat Med. 5, 49–55. Mokhtarian, F., McFarlin, D. E., and Raine, C. S. (1984). Adoptive transfer of myelin basic protein-sensitized T cells produce chronic relapsing demyelinating disease in mice. Nature 309, 356–358. Mokhtarian, F., Zhang, Z., Shi, Y., Gonzales, E., and Sobel, R. A. (1999). Molecular mimicry between a viral peptide and a myelin oligodendrocyte glycoprotein peptide induces autoimmune demyelinating disease in mice. J Neuroimmunol. 95, 43–54. Moody, D. B., and Besra, G. S. (2001). Glycolipid targets of CD1-mediated T-cell responses. Immunology 104, 243–251. Moody, D. B., Besra, G. S., Wilson, I. A., and Porcelli, S. A. (1999). The molecular basis of CD1-mediated presentation of lipid antigens. Immunol. Reviews 172, 285–296. Morris-Downes, M. M., McCormack, K., Baker, D., Sivaprasad, D., Natkunarajah, J., and Amor, S. (2002). Encephalitogenic and immunogenic potential of myelin associated glycoprotein (MAG), oligodendrocytespeciWc glycoprotein (OSP) and 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNPase) in ABH and SJL mice. J Neuroimmunol. 122, 20–33. Neumann, H., Boucraut, J., Hahnel, C., Misgeld, T., and Wekerle,-H. (1996). Neuronal control of MHC class II inducibility in rat astrocytes and microglia. Eur J Neurosci. 8, 2582–2590. Neumann, H., Medana, I. M., Bauer, J., and Lassmann, H. (2002). Cytotoxic T lymphocytes in autoimmune and degenerative CNS disease. Trends Neurosci. 25, 313–319. Neumann, H., Misgeld, T., Matsumuro, K., and Wekerle, H. (1998). Neurotrophins inhibit major histocompatibility class II inducibility of microglia: Involvement of the p75 neurotrophin receptor. Proc Natl Acad Sci US 95, 5779–5784. Neumann, H., Schmidt, H., Cavalie, A., Jenne, D., and Wekerle, H. (1997). Major histocompatibility complex (MHC) class I gene expression in single neurons of the central nervous system: DiVerential regulation by interferon (IFN)-gamma and tumor necrosis factor (TNF)-alpha. J Exp Med. 185, 305–316. Niehaus, A., Shi, J., Grzenkowski, M., Diers-Fenger, M., Hartung, H. P,. Toyka, K., Bruck, W., and Trotter, J (2000). Patients with active relapsing-remitting multiple sclerosis synthesize antibodies recognizing oligodendrocyte progenitor cell surface protein: Implications for remyelination. Ann Neurol. 48, 362–371. Noble, P. G., Antel, J. P., and Yong, V. W. (1994). Astrocytes and catalase prevent the toxicity of catecholamines to oligodendrocytes. Brain Res. 633, 83–90. Numao, T., and Agrawal, D. K. (1992). Neuropeptides modulate human eosinophil chemotaxis. J. Immunol. 149, 3309–3315. Okuda, Y., Sakoda, S., Fujimura, H., and Yanagihara, T. (1997). Nitric oxide via an inducible isoform of nitric oxide synthase is a possible factor to eliminate inXammatory cells from the central nervous system of mice with experimental allergic encephalomyelitis. J Neuroimmunol. 73, 107–116. Oldstone, M. B., and Southern, P. J. (1993). TraYcking of activated cytotoxic T lymphocytes into the central nervous system: Use of a transgenic model. J Neuroimmunol. 46, 25–31. Olsson, T., Dahlman, I., Wallstrom, E., Weissert, R., and Piehl, F. (2000). Genetics of rat neuroinXammation. J. Neuroimmunol. 107, 191–200. Opdenakker, G., Van-Damme, J. (1994). Cytokine-regulated proteases in autoimmune diseases. Immunol Today. 15 103–107. Owens, T., Wekerle, H., and Antel, J. (2001) Genetic models for CNS inXammation. Nature Med. 7, 161–166. Paterson, P. Y. (1960). Transfer of allergic encephalomyelitis in rats by means of lymph node cells. J. Exp. Med. 111, 119–135. Pender, M. P., Nguyen, K. B., McCombe, P. A., and Kerr, J. F. (1991). Apoptosis in the nervous system in experimental allergic encephalomyelitis. J Neurol Sci. 104, 81–87. Pender, M. P and Rist, M. J. (2001). Apoptosis of inXammatory cells in immune control of the nervous system: Role of glia. Glia. 36, 137–144.

CONCLUSIONS

Piddlesden, S. J., Storch, M., Hibbs, M., Freeman, A. M., Lassmann, H., and Morgan, B. P. (1994). Soluble recombinant complement receptor 1 inhibits inXammation and demyelination in antibody-mediated demyelinating experimental allergic encephalomyelitis. J Immunol 152, 5477–5484. Pitt, D., Werner, P., Raine, C. S. (2000). Glutamate excitotoxicity in a model of multiple sclerosis. Nature Med 6, 67–70. Prineas, J. W., and McDonald, I. W. (1997). Demyelinating diseases. In ‘‘GreenWeld’s Neuropathology’’ (D. I. Graham and P. L. Lantos, eds.), 6th ed., pp. 813–896. Arnold, London, Sidney, Auckland. Probert, L., Eugster, H. P., Akassoglou, K., Bauer, J., Frei, K., Lassmann, H., and Fontana, A. (2000). TNFR1 signalling is critical for the development of demyelination and the limitation of T-cell responses during immune-mediated CNS disease. Brain. 123, 2005–2019. Raine, C. S. (1985). Experimental allergic encephalomyelitis and experimental allergic neuritis. In ‘‘Handbook of Clinical Neurology’’ (P. I. Vinken and G. W. Bruyn, eds.), Vol. 47, pp. 429–466. Elsevier, New York. Raine, C. S., Snyder, D. H., Valsamis, M. D., and Stone, S. H. (1974). Chronic experimental allergic encephalomyelitis in inbred guinea pigs–an ultrastructural study. Lab. Invest. 31, 369–380. RansohoV, R. M. (1999). Mechanisms of inXammation in MS tissue: Adhesion molecules and chemokines. J Neuroimmunol 98, 57–68. Remlinger, P. (1928). Les paralysies due traitment antirebique. Annals of the Insitute Pasteur 55 (Suppl), 35–68. Rivers. T. M., Sprunt, D. H., and Berry, G. P. (1933). Observations on attempts to produce acute disseminated encephalomyelitis in monkeys. J. Exp. Med. 58, 39–53. Rodriguez, M., and Miller, D. J. (1994). Immune promotion of central nervous system remyelination. Prog-BrainRes. 103, 343–355. Rosener, M. Muraro, P. A., Riethmuller, A., Kalbus, M., Sappler, G., Thompson, R. J., Lichtenfels, R., Sommer, N., McFarland, H. F., and Martin, R. (1997). 2’,3’-cyclic nucleotide 3’-phosphodiesterase: A novel candidate autoantigen in demyelinating diseases. J Neuroimmunol 75, 28–34. RuV, M., Wahl, S. M., and Pert, C. B. (1985). Substance P receptor mediated chemotaxis of human monocytes. Peptide 6, 107–111. RuYni, F., Furlan, R., Poliani, P. L., Brambilla, E., Marconi, P. C., Bergami, A., Desina, G., Glorioso, J. C., Comi, G., and Martino, G. (2001). Fibroblast growth factor-II gene therapy reverts the clinical course and the pathological signs of chronic experimental autoimmune encephalomyelitis in C57BL/6 mice. Gene-Ther. 8, 1207–1213. Schlu¨sener, H., Sobel, R. A., Linington, C., and Weiner, H. L. (1987). A monoclonal antibody against a myelin oligodendrocyte glycoprotein induces relapses and demyelination in central nervous system autoimmune disease. J. Immunol. 139, 4016–4021. Schmied, M., Breitschopf, H., Gold, R., Zischler, H., Rothe, G., Wekerle, H., and Lassmann, H. (1993). Apoptosis of T lymphocytes in experimental autoimmune encephalomyelitis: Evidence for prgrammed cell death as a mechanism to control inXammation in the brain. Amer J Pathol 143, 446–452. Schwartz, M. (2001). Harnessing the immune system for neuroprotection: Therapeutic vaccines for acute and chronic neurodegenerative disorders. Cell Mol Neurobiol. 21, 617–27. Seil, A. J., Quarles, R. H., Johnson, D., and Brady, R. O. (1981). Immunisation with puriWed myelin associated glycoprotein does not evoke myelination inhibiting or demyelinating antibodies. Brain Research 209, 470–475. Seil, F. J., and Agrawal, H. C. (1980). Myelin proteolipid protein does not induce demyelinating or myelin inhibiting antibodies. Brain Research 194, 273–277. Selmaj, K. W., and Raine, C. S. (1988). Tumor necrosis factor mediates myelin and oligodendrocyte damage in vitro. Ann Neurol. 23, 339–346. SeraWni, B., Columba-Cabezas, S., Di Rosa F., and Aloisi, F. (2000). Intracerebral recruitment and maturation of dendritic cells in the onset and progression of experimental autoimmune encephalomyelitis. Amer J Pathol 157, 1991–2002. Shamshiev, A., Donda, A., Carena, I., Mori, L., Kappos, L., and DeLibero, G. (1999). Self glycolipids as T-cell autoantigens. Eur. J. Immunol. 29, 1667–1675. Shamshiev, A., Gober, H. J., Donda, A., Mazorra, Z., Mori, L., and DeLibero, G. (2002). Presentation of the same glycolipid by diVerent CD1-molecules. J. Exp. Med. 195, 1013–1021. Smith, T., Groom, A., Zhu, B., and Turski, L. (2000). Autoimmune encephalomyelitis ameliorated by AMPA antagonists. Nature Med. 6, 62–66. Smith, K. J., and Lassmann, H. (2002). The role of nitric oxide in multiple sclerosis. Lancet Neurology 1, 232–241. Smith, K. J., Kapoor, R., and Felts, P. A. (1999). Demyelination: The role of reactive oxygen and nitrogen species. Brain Pathol. 9, 69–92. Smith, K. J., Kapoor, R., Hall, S. M., and Davies, M. (2001). Electrically active axons degenerate when exposed to nitric oxide. Ann Neurol. 49, 470–476. Sobel, R. A., and Colvin, R. B. (1985). The immunopathology of experiemntal allergic encephalomyelitis (EAE). III. DiVerential in situ expression of strain 13 Ia on endothelial and inXammatory cells of (strain 2
strain 13) F1 guinea pigs with EAE. J. Immunol. 91, 2333–2337. Sobel, R. A., and Kuchroo, V. K. (1992). The immunopathology of acute experimental allergic encephalomyelitis induced with myelin proteolipid protein. T cell receptors in inXammatory lesions. J. Immunol. 149, 1444–1451. Springer, T. A. (1994). TraYc signals for lymphocyte recirculation and leucocyte emigration: The multistep paradigm. Cell 76, 301–314. SteVerl, A., Brehm, U., Storch, M., Lambracht-Washington, D., Bourquin, C., Wonigeit, K. Lassmann, H., and Linington, C. (1999). Myelin oligodendrocyte glycoprotein induces experimental autoimmune encephalomy-

1069

1070

43. MYELIN AND ITS DISEASES

elitis in the resistant Brown Norway rat: Disease susceptibility is determined by MHC and MHC-linked eVects on the B-cell response. J Immunol 163, 40–49. Storch, M. K., Fischer-Colbrie, R., Smith, T., Rinner, W. A., Hickey, W. F., Cuzner, M. L., Winkler, H., and Lassmann, H. (1996). Co-localization of secretoneurin immunoreactivity and macrophage inWltration in the lesions of experiemntal autoimmune encephalomyelitis. Neurosci. 71, 885–893. Storch, M. K., SteVerl, A., Brehm, U., Weissert, R., Wallstro¨m, E., Kerschensteiner, M., Olsson, T., Linington, C., and Lassmann, H (1998). Autoimmunity to myelin oligodendrocyte glycoprotein in rats mimics the spectrum of multiple sclerosis pathology. Brain Pathol 8, 681–694. Storch, M. K., Weissert, R., SteVerl, A., Birnbacher, R., Wallstrom, E., Dahlman, I., Ostensson, C. G., Linington, C., Olsson, T and Lassmann, H. (2002). MHC gene related eVects on microglia and macrophages in experiemntal autoimmune encephalomyelitis determine the extent of axonal injury. Brain Pathol. 12, 287–299. Sun, D., Qin, Y., Chluba, J.,, Epplen J. T., and Wekerle, H. (1988). Suppression of experimentally induced autoimmune encephalomyelitis by cytolytic T-T-cell interactions. Nature 332, 843–845. Sun, D., and Wekerle, H. (1986). Ia-restricted encephalitogenic T lymphocytes mediating EAE lyse autoantigenpresenting astrocytes. Nature. 320, 70–72. Sun, D., Whitaker, J. N., Huang, Z., Liu, D., Coleclough, C., Wekerle, H., and Raine, C. S. (2001). Myelin antigen speciWc CD8þ T cells are encephalitogenic and produce severe disease in C57BL/6 mice. J Immunol 166, 7579–7587. Sundvall, M., Jirholt, J., Yang, H. T., Jansson, L., Engsro¨m, A., Pettersson, U., and Holmdahl, R. (1995). IdentiWcation of murine loci associated with susceptibility to chronic experimental autoimmune encephalomyelitis. Nature Genetics 10, 313–317. Swanborg, R. H. (2001). Experimental autoimmune encephalomyelitis in the rat: Lessons in T-cell immunology and autoreactivity. Immunol. Reviews 184, 129–135. Taupin, V., Renno, T., Bourbonniere, L., Peterson, A. C., Rodriguez, M., and Owens, T. (1997). Increased severity of experimental autoimmune encephalomyelitis, chronic macrophage/microglial reactivity, and demyelination in transgenic mice producing tumor necrosis factor-alpha in the central nervous system. Eur J Immunol. 27, 905–913 Teesalu, T., Hinkkanen, A. E., and Vaheri, A. (2001). Coordinated induction of extracellular proteolysis systems during experimental autoimmune encephalomyelitis in mice. Am J Pathol. 159, 2227–2237. Traugott. U., Reinherz, E. L., and Raine, C. S. (1983). Multiple sclerosis: Distribution of T cell subsets within active chronic lesions. Science 219, 308–310. Tsunoda, I., Kuang, L. Q., Theil, D. J., and Fujinami, R. S. (2000). Antibody association with a novel model for primary progressive multiple sclerosis: Induction of relapsing-remitting and progressive forms of EAE in H2s mouse strains. Brain Pathol. 10, 402–418. Vanderlugt, C. L., Begolka, W. S., Neville, K. L., Katz-Levy, Y. Howard, L.,M; Eagar, T. N., Bluestone, J. A., and Miller, S. D. (1998). The functional signiWcance of epitope spreading and its regulation by co-stimulatory molecules. Immunol Rev. 164, 63–72. Vass, K., Heininger, K., Scha¨fer, B., Linington, C., and Lassmann, H. (1992). Interferon-gamma potentiates antibody-mediated demyelination in vivo. Ann. Neurol. 32, 189–206. Vass, K., and Lassmann, H. (1990). Intrathecal application of interferon-gamma: Progressive appearance of MHC antigens within the rat nervous system. Amer. J. Pathol. 137, 789–800. Voskuhl, R. R. (1998). Myelin protein expression in lymphoid tissues: Implications for peripheral tolerance. Immunol Rev. 164, 81–92. Wallstrom, E., Diener, P., Ljungdahl, A., Khademi, M., Nilsson, C. G., and Olsson, T. (1996). Memantine abrogates neurological deWcits, but not CNS inXammation, in Lewis rat experimental autoimmune encephalomyelitis. J Neurol Sci. 137, 89–96. Weerth, S., Berger, T., Lassmann, H., and Linington, C. (1999). Encephalitogenic and neuritogenic T cell responses ti the myelin associated glycoprotein (MAG) in the Lewis rat. J Neuroimmunol. 95, 157–164. Weissert, R., Wallstrom, E., Storch, M. K., SteVerl, A., Lorentzen, J. Lassmann, H., Linington, C., and Olsson, T. (1998). MHC haplotype-dependent regulation of MOG-induced EAE in rats. J-Clin-Invest. 102, 1265–1273. Wekerle, H. (1998) Immunology of multiple sclerosis. In ‘‘McAlpine’s Multiple Sclerosis’’ (A. Compston, ed.), pp. 379–407. Churchill Livingstone, London. Wekerle, H., Linington, C., Lassmann, H., and Meyermann, R. (1986). Cellular immune reactivity within the CNS. Trends Neurosci. 9, 271–277. White, C. A., McCombe, P. A., and Pender, M. P. (1998). Microglia are more susceptible than macrophages to apoptosis in the central nervous system in experimental autoimmune encephalomyelitis through a mechanism not involving Fas (CD95). Int Immunol. 10, 935–941. White, C. A., Nguyen, K. B., and Pender, M. P. (2000). B cell apoptosis in the central nervous system in experimental autoimmune encephalomyelitis: Roles of B cell CD95, CD95L and Bcl-2 expression. J Autoimmun. 14, 195–204. Willenborg, D. O., Fordham, S. A., Cowden, W. B., and Ramshaw, I. A. (1995). Cytokines and murine autoimmune encephalomyelitis: Inhibition or enhancement of disease with antibodies to select cytokines, or by delivery of exogenous cytokines using a recombinant vaccinia virus system. Scand-J-Immunol. 41, 31–41. Willenborg, D. O., Staykova, M. A., and Cowden, W. B. (1999). Our shifting understanding of the role of nitric oxide in autoimmune encephalomyelitis: A review. J Neuroimmunol. 100, 21–35. Wisniewski, H. M., and Bloom, B. R. (1975). Primary demyelination as a nonspeciWc consequence of a cell mediated immune reaction. J. Exp. Med. 141, 346–459.

CONCLUSIONS

Wucherpfennig, K. W., and Strominger, J. L. (1995). Molecular mimicry in T cell-mediated autoimmunity: Viral peptides activate human T cell clones speciWc for myelin basic protein. Cell. 80, 695–705. Yednock, T. A., Cannon, C., Fritz, L. C., Sanchez-Madrid, F., and Steinman, L. (1992). Prevention of experimental autoimmune encephalomyelitis by antibodies against a4ß1 integrin. Nature 356, 63–66. Yu, M., Johnson, J. M., and Tuohy, V. K. (1996). A predictable sequential determinant spreading cascade invariably accompanies progression of experimental autoimmune encephalomyelitis: A basis for peptidespeciWc therapy after onset of clinical disease. J Exp Med. 183, 1777–88. Zamvil, S., Nelson, P., Trotter, J., Mitchell, D., Knobler, R., Fritz, R., and Steinman, L. (1985). T-cell clones speciWc for myelin basic protein induce chronic relapsing paralysis and demyelination. Nature 317, 355–358. Zettl, U. K., Mix, E., Zielasek, J., Stangel, M., Hartung, H. P., and Gold, R. (1997). Apoptosis of myelin-reactive T cells induced by reactive oxygen and nitrogen intermediates in vitro. Cell Immunol. 178, 1–8. Zhang, J., Dawson, V. L., Dawson, T. M., and Snyder, S. H. (1994). Nitric oxide activation of poly(ADP-ribose) synthetase in neurotoxicity. Science. 263, 687–689.

1071

C H A P T E R

44 Experimental Models of Virus-Induced Demyelination A. J. Bieber and M. Rodriguez

MEDICAL RELEVANCE OF ANIMAL MODELS OF DEMYELINATING DISEASE Multiple sclerosis (MS) is an inXammatory demyelinating disease of the human central nervous system (CNS) and is the most common cause of acquired nontraumatic neurologic disability in young adults. The pathologic hallmark of the disease is damage to oligodendrocytes and CNS myelin, which results in demyelinated white matter lesions, followed by axonal loss and glial scarring (Lucchinetti et al., 1997; Noseworthy et al., 2000; Trapp et al., 1998). Areas of active disease display a prominent inXammatory response with tissue inWltration by mononuclear cells, primarily T cells and macrophages. Despite decades of research, the causes of myelin damage and neurologic dysfunction in MS remain largely unknown. Chapters 29 through 33 in Section IV of these volumes provide an overview of the classiWcation, pathology, genetics, and potential pathogenic mechanisms of multiple sclerosis. There is no cure for MS, and attempts to develop eVective therapies have met with only limited success (Noseworthy, 1998). Immunosuppression with corticosteroids is commonly used as a short-term therapy to control MS relapses. Long-term therapy generally consists of treatment with immunomodulatory drugs such as glatiramer acetate (Copolymer-1/ Copaxone), interferon-b-1a (Avonex), or interferon-b-1b (Betaseron). Each of these drugs has been shown to reduce relapses of the disease, but whether they alter the long-term clinical endpoint is still being debated. Our continued lack of understanding about the causes of MS and the mechanisms of disease progression, and the lack of truly eVective therapies for the disease, underscore the need for good animal models of MS on which to conduct research. In this chapter we review two of the most widely studied animal models of virus-induced demyelinating disease: Theiler’s murine encephalomyelitis virus and murine hepatitis virus. Both viruses produce acute inXammatory encephalitis that is followed by chronic CNS demyelinating disease. The clinical and pathologic correlates of virus-induced demyelination are largely immune mediated. Several pathologic mechanisms have been proposed to explain the development of myelin damage and neurologic deWcits, and each of the proposed mechanisms may play a role in disease progression depending on the genetic constitution of the infected animal. The induction of demyelinating disease by virus may be directly relevant to human MS. Several viruses are known to cause demyelination in humans, and viral infection is an epidemiologic factor that is consistently associated with clinical exacerbation of MS (Sarchielli et al., 1993; Sibley et al., 1985). It has been suggested that viral infection may be a cause of MS, although no speciWc virus has been identiWed as a causative agent.

Myelin Biology and Disorders, Volume 2

1073

Copyright 2004, Elsevier (USA). All rights reserved.

au1

1074

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

The clinical and pathologic presentation of virus-induced demyelination, the immunemediated nature of the disease mechanisms, and the variability of the disease presentation depending on genetic background are all very similar to what is observed in MS. The many similarities between human MS and the demyelinating diseases that are induced by infection with TMEV and MHV make these animal models very attractive for the study of multiple sclerosis.

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV) Theiler’s murine encephalomyelitis Virus (TMEV) belongs to the cardiovirus genus of the picornavirus family (Nitayaphan et al., 1986; Pevear et al., 1987). The original isolation and characterization of the virus was reported by Theiler in 1937 when the eVects of TMEV infection of the CNS were observed in a mouse that spontaneously developed Xaccid paralysis. The virus was subsequently transmitted to other mice by intracerebral inoculation with a suspension made from the brain and spinal cord of the infected mouse. TMEV is a ubiquitous enteric pathogen that usually causes asymptomatic intestinal infections in mice. Occasionally, the virus will spread beyond the intestinal tract and enter the CNS, resulting in both acute and chronic CNS infections. Chronic CNS infection results in extensive demyelination with accumulating neurologic deWcits and has provided a valuable experimental model for human demyelinating disease.

Virus Biology and Life Cycle TMEV Subgroups

au2

Since its initial discovery, isolates of TMEV have been recovered in several laboratories. The diVerent TMEV strains have been divided into two subgroups depending on the disease that they induce after infection of the CNS (Lorch et al., 1981). The GDVII and FA strains are highly virulent with an LD50 as low as 1 to 10 plaque forming units (PFU). CNS infection causes acute encephalitis that is characterized by the destruction of a large number of CNS neurons, especially in the cortex, hippocampus, thalamus, brain stem, and in the anterior horns of the spinal cord. The encephalitis is usually fatal, and the highly virulent strains do not result in persistent infection of the CNS. A second TMEV subgroup is far less virulent (LD50 > 106 PFU) and results in a distinctly diVerent disease presentation. This subgroup, which includes the TO4, DA, BeAn 3886, Yale, and WW strains, induces a biphasic CNS disease (Daniels et al., 1952; Lipton, 1978; Wroblewska et al., 1977). Following intracerebral infection, these strains replicate in neurons in the brain resulting in encephalitis similar to that observed with the GDVII strain but nonlethal. The virus is cleared from the brain by the host immune response but persists in the spinal cord, eventually resulting in the development of chronic demyelinating disease (Lehrich et al., 1976; Lipton, 1975; Njenga et al., 1997; Rodriguez et al., 1987). The disease is characterized by viral persistence in oligodendrocytes (Rodriguez et al., 1983) and macrophages (Lipton et al., 1995), with chronic demyelination and progressive loss of motor function (McGavern et al., 1999). The pathology is largely immune mediated with animals demonstrating a range of disease phenotypes depending on their genetic background. In the SJL strain, demyelination is evident within 30 days after infection. By 90 days, infected animals begin to develop spasticity and gait abnormalities, and weakness of the lower extremities, with paralysis eventually occurring by 6 to 9 months (Lipton and Dal Canto, 1976a). Capsid Structure The X-ray crystallographic structures of DA, BeAn and GDVII viruses have been determined to about 3 angstrom resolution (Grant et al., 1992; Luo et al., 1992, 1996; Toth et al., 1993). A schematic representation of the TMEV capsid, based on the crystallograghic data, is presented in Figure 44.1. Each virus consists of a protein shell that is composed of 60

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV)

FIGURE 44.1 TMEV capsid and genome structure. The upper panel shows a schematic representation of the TMEV capsid, based on the X-ray crystallograghic data. Lower panel shows a map of the TMEV DA strain genome. The 5’ and 3’ untranslated (UTR) regions are indicated. The 5’ end of the RNA is bound by the viral VPg protein (ball) and the 3’ end is polyadenylated (AAAAAn). The open bar indicates the open reading frame from nucleotides 1066 to 7968. The heavy lines dividing the bar indicate the positions of the initial proteolytic cleavages after the leader peptide (L) and between the P1, P2, and P3 precursor peptides. The lighter lines dividing the bar indicate the arrangement of the Wnal gene products. The identities of the Wnal gene products are discussed in the text. The TMEV capsid structure is reproduced with permission from Webster and GranoV (1995).

protomers, arranged as 12 pentamers with icosahedral symmetry. Each protomer contains one of each of the four capsid peptides (VP1, VP2, VP3, VP4). The VP1, VP2, and VP3 capsid proteins are each composed of an eight-stranded antiparallel b-barrel. The loops that connect the strands of the b-barrel form the surface structure of the viral capsid and may play important roles as sites of binding for neutralizing antibody and in conformational determinants of virus persistence. A second surface feature, the ‘‘pit’’ formed at the contact region between VP1 and VP3, is the likely binding site for the cellular receptor (Zhou, et al., 2000; Jnaoui, et al., 2002). Viral Genome The TMEV genome consists of a positive sense, single stranded RNA molecule. Genomes for both the GDVII and TO subgroups have been sequenced and found to be very similar, with about 90% identity at the nucleotide level and 96% at the amino acid level (Pevear et al., 1988). The genome of the DA strain is 8093 nucleotides long (Ohara et al., 1988) and contains a single 6903 nucleotide open reading frame that starts at nucleotide 1066 and encodes a 2301 amino acid polyprotein. The genome terminates with a poly(A) tail. The 12 mature TMEV gene products are generated from the polyprotein by post-translational proteolytic cleavage. The gene products include a leader peptide, four capsid polypeptides (VP1-4), two viral proteases, a polymerase/helicase, the RNA-dependent RNA polymerase that is involved in genome replication, and a small basic protein which becomes covalently linked to the 5’ end of viral RNAs. Two other gene products have unknown functions. Virus Life Cycle The TMEV infection cycle is typical of the picornavirus family (Rueckert, 1996). Attachment of the virion to speciWc cell surface receptors serves to position the virus close to the cell membrane. Attachment also induces a conformational change in the virion that results in the loss of the VP4 capsid protein and the translocation of the RNA genome across the cell membrane and into the cytoplasm. Co-opting the translational machinery of the cell,

1075

1076

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

the RNA genome directs the synthesis of a single large polyprotein. While nascent on the ribosome, the polyprotein is cleaved by virus encoded protease activity, into three large precursor proteins: P1, P2, and P3. The precursor proteins are further cleaved to produce the viral structural proteins and proteins necessary for completion of the infection cycle (Fig. 44.1). Precursor P3 is processed to produce a viral proteinase (designated 3C), a protein involved in initiating replication of the viral genome (3B or VPg) and the viral RNAdependent RNA polymerase (3D). The viral polymerase copies the incoming viral genome to produce a negative-strand RNA that then serves as the template for replication of the positive-strand viral genome. The newly formed positive-strand RNA molecules may be copied to form more negative-strand templates or be packaged into virions as virus assembly proceeds. The P1 precursor protein is cleaved to produce the viral capsid proteins VP0 (1AþB), VP1 (1D), and VP3 (1C). As the concentration of capsid proteins increases in the cell, they begin to aggregate into protomers composed of one copy of each of the capsid proteins. These protomers then assemble into pentamers, and 12 pentamers assemble with a positivestranded, VPg bound viral RNA, to produce a noninfectious provirion. The Wnal step in maturation to infectious virus requires a maturation cleavage in which the VP0 protein is cleaved to give the Wnal VP2 (1B) and VP4 (1A) capsid proteins that are found in the mature virus. In neonates and in some cells in culture, mature virus may accumulate to very high levels in infected cells and can often be observed as paracrystalline arrays in electron micrograghs (Fig. 44.2). Virus is generally released from the cell by infection-mediated cytolysis. The time that is required to complete the infection cycle is generally 7 to 12 hours. TMEV Receptor In culture, TMEV is able to infect a wide variety of cell types from several diVerent species. Although a unique protein receptor for the virus has not been identiWed, several lines of evidence suggest that cell surface carbohydrate residues may play a central role in the binding and entry of virus. Removal of sialic acid from the cell surface by treatment with sialidase signiWcantly reduces the infectivity of BeAn virus in BHK cells (Fotiadis et al., 1991). Similar decreases are observed when binding to sialic acid is blocked with wheat germ agglutinin, which binds to sialic acid on the cell surface (Fotiadis et al., 1991).

FIGURE 44.2 Crystalline virus in neonatal brain. A paracrystalline viral array in the cytoplasm of a spinal cord neuron from a neonatal SJL mouse infected with the DA strain of TMEV. An intact mitochondrion is visible (m), but the numerous cytoplasmic vacuoles (v) indicate the cytopathic eVects of viral infection.

1077

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV)

Sialyllactose comprises the terminal three sugar residues of cell surface oligosaccarides, and inclusion of sialyllactose in cell culture medium inhibits viral binding and infectivity by blocking receptor binding sites on the virus (Zhou et al., 1997). Together these data demonstrate an important role for sialic acid in the recognition of the cell surface by the TO subgroup TMEV strains. Sialyloligosaccaride addition is a common post-translation modiWcation of cell surface proteins, which is consistent with the observation that TMEV can infect a variety of cell types. Which speciWc sialic acid containing proteins might be required for virus binding and entry has not yet been established. The neurovirulent strains of TMEV (GDVII and FA) do not require sialic acid for cell surface binding (Fotiadis et al., 1991) but instead use cell surface heparin sulfate proteoglycan. Pretreatment of virus with heparin sulfate, treatment of host cells with heparin sulfate cleaving enzymes, or treatment with heparin sulfate speciWc antibodies all inhibit GDVII binding and infection (Reddi and Lipton, 2002). Proteoglycan deWcient CHO cells are resistant to infection, further supporting a role for proteoglycans in GDVII binding. The interaction of proteins with heparin sulfate is usually mediated by heparin-binding domains (HBD) and a heparin-binding domain consensus sequence has been described (Cardin and Weintraub, 1989). A putative HBD is present in the amino acid sequence of the VP1 capsid protein of both the GDVII and TO subgroups. This sequence may play a role in the binding of GDVII to heparin sulfate but if so, conformational diVerences must exist between VP1 peptides of the TMEV subgroups because the TO subgroup viruses do not bind heparin sulfate. Proteoglycans are known to serve as co-receptors or attachment factors for a variety of diVerent viruses. For several of these viruses, heparin sulfate acts primarily as an attachment factor and is not fully suYcient to mediate virus entry. The extent to which additional protein entry receptors are needed for GDVII infection is not yet clear (Reddi and Lipton, 2002).

Disease Pathology The encephalitic phase of TMEV-induced disease develops rapidly after intracerebral infection. Virus infects and replicates in neurons of the brain and spinal cord, with titers of infectious virus reaching their peak by 5 to 7 days. Immunohistochemical staining demonstrates the highest viral antigen loads in the thalamus, hypothalamus, midbrain, brain stem, and spinal cord gray matter. Neuronal infection and lysis is particularly pronounced with the GDVII subgroup viruses, and infection usually results in death soon after inoculation (Patick et al., 1990). Acute disease following infection with the TO subgroup viruses is also characterized by intense mononuclear cell inWltration in the brain and spinal cord. The combined eVects of direct viral lysis of neurons and damage from inWltrating cells may result in transient Xaccid paralysis within 2 weeks after infection, but in many cases the early phase of the disease is largely asymptomatic. The acute phase of the disease is followed by clearance of virus from the gray matter by the host immune response (Njenga et al., 1997). Resistant strains of mice achieve virtually complete viral clearance. In susceptible strains, virus is cleared from the gray matter but persists in the white matter of the spinal cord for the lifetime of the animal. During persistent infection, the amount of infectious virus, and the number of cells with detectable virus antigen in the spinal cord is usually quite low. This is in contrast to the high level of viral genome that is present throughout the disease (Trottier et al., 2001). Infected cells most often appear in the lateral or anterior columns of the thoracic spinal cord and the presence of infected cells is often accompanied by an inXammatory parenchymal inWltrate and perivascular cuYng. The diVuse inXammatory inWltrate consists of large numbers of CD4þ and CD8þ T lymphocytes, activated macrophages and microglia, B cells, and reactive astrocytes (Fig. 44.3; Lindsley et al., 1989). The presence of inXammatory cells correlates closely with the development of myelin damage, and electron microscopy reveals inWltrating mononuclear cells stripping myelin from axons. Macrophages containing engulfed myelin debris are numerous in demyelinating lesions. The speciWc roles that these diVerent cell types play in the development of demyelination and neurologic deWcits will be discussed in subsequent sections.

au3

au4

1078

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

FIGURE 44.3 Immunochemical characterization of a chronic demyelinated lesion. Adjacent sections from the spinal cord of an SJL mouse, 8 months post-infection, were stained for CD4þ T cells with the Ly-2 antibody (A), for CD8þ T cells with the GK1.5 antibody (B) and for macrophages with the M1/70 antibody against Mac-1 (C).

Demyelination is apparent by as early as 22 days after infection and in many mouse strains, as demyelination progresses, neurologic deWcits begin to appear. Demyelination pathology is illustrated in Figure 44.4. The development of neurologic deWcits in

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV)

FIGURE 44.4 Demyelinated and remyelination in the TMEV-infected spinal cord. (A) Normal spinal cord white matter stained for myelin sheaths with para-phenylene-diamine. (B) A demyelinated lesion from an FVB mouse, 3 months postinfection with the DA strain of TMEV. Almost complete myelin destruction is apparent. (C) Extensive remyelination is apparent by 8 months post-infection. Schwann cell remyelination is evident as heavily stained myelin Wgures in the upper right corner of the panel, while the lightly staining myelin in the center of the panel is oligodendrocyte remyelination.

1079

1080

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

infected mice involves not just the disruption of neuronal function but neuronal and axonal loss as well (McGavern et al., 2000). Quantitative assessment of CNS injury during TMEV-induced disease shows that signiWcant atrophy of the lateral and anterior white matter regions of the spinal cord can be detected by 45 days post-infection. Atrophy is most pronounced in the posterior cervical and anterior thoracic regions of the spinal cord. While levels of demyelination reach a peak by 100 days post-infection, spinal cord atrophy continues to increase between 100 and 220 days. A 25% reduction was observed in the area of the anterior and lateral white matter at the C8-T11 level of the spinal cord by 195 to 229 days after infection. In this study, atrophy was speciWcally assessed in areas of normal appearing white matter and not in lesion areas, to avoid complications due to area changes resulting from inXammation and edema. This marked spinal cord atrophy is the result of a 30% loss of medium to large axonal Wbers by 195 to 229 days post-infection. The progressive loss of axons correlates well with electrophysiologic changes in the spinal cord and with changes in motor ability. These observations demonstrate that the neurologic deWcits that develop during chronic TMEV-induced demyelinating disease are primarily due to axonal damage and loss rather than to demyelination alone.

Viral Persistence Intracerebral inoculation with the highly virulent GDVII subgroup viruses results in fatal encephalitis, while infection with the less virulent TO viruses results in milder encephalitis, followed by clearance of virus from the brain and virus persistence in the spinal cord. There is still some debate concerning whether the diVerence in disease that is induced by these viruses is due simply to diVerences in virulence. Would the GDVII strains persist and cause demyelination if animals survived the initial encephalitis or are determinants for persistence and demyelination a speciWc genetic property of the TO strains? Recombinant chimeric viruses, resulting from the exchange of sequences between the GDVII and BeAn (or DA) virus genomes, have been useful in beginning to map the determinants of virus neurovirulence and persistence. Initial studies localized the majority of neurovirulence to a segment of the GDVII genome from the middle of 1B to the middle of 2C, although additional upstream sequences had a more minor eVect (Fig. 44.1; Fu et al., 1990a). Infection with recombinant DA virus that contains these GDVII regions results in death at a much reduced LD50 compared to wildtype DA virus. Subsequent studies narrowed the critical segment to a region of P1 from 1B to 1D, the capsid coding region (Zhang et al., 1993). Several studies have since conWrmed that major determinants of viral persistence lie in the region of the viral genome that encodes the capsid proteins and have identiWed regions in two of these proteins, VP1 and VP2, that appear to be necessary for persistence (Adami et al., 1998; Lin et al., 1998). Molecular modeling suggests that these sequences may be positioned on the surface of the virus in a manner consistent with a role in the binding of the virus to its cellular receptor (Zhou et al., 1997). DiVerences in receptor binding between the TMEV subgroups may therefore play an important role in determining the balance between neurovirulence and the ability to produce persistent infection. Several mutated strains of the GDVII virus have been characterized in which neurovirulence is greatly attenuated, and while these strains are viable, replicate in the brain, and may even induce encephalitis, they do not persist or induce chronic demyelinating disease (Lipton et al., 1991, 1998). These experiments suggest that ability of a TMEV strain to produce persistent infection and demyelination is not solely a function of neurovirulence but may be a speciWc genetic property of the strain. Other investigators have reported that insertion into the GDVII backbone, of DA strain segments from across the length of the DA genome, attenuates virulence and results in virus persistence and demyelination (Fu et al., 1990b; Rodriguez and Roos, 1992; see also Jakob and Roos, 1996). Further studies will be needed to determine whether there are speciWc genetic determinants in the TO subtype that control persistence and to map these determinants to speciWc viral proteins. An additional interesting diVerence between the GDVII and TO subgroups involves the expression of the leader peptide region (L) of the TMEV genome at the 5’ end of the open

1081

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV)

reading frame (see Fig. 44.1). The TO strains use an alternative initiation codon to produce L*, a protein whose coding sequence is out of frame with the polyprotein reading frame. The GDVII strains do not encode the L* protein. Mutations in L* dramatically decrease virus-induced demyelination indicating that it plays a role in the development of disease (Chen et al., 1995). The L* protein is essential for replication of the DA strain in a macrophage-like cell line in culture, while the GDVII strain does not replicate in these cells (Takata et al., 1998). L* appears to function by preventing apoptosis of the DAinfected cells (Ghadge, et al., 1998). The production of the L* protein therefore plays an essential role in the ability of TO subtype viruses to persist in speciWc cell types in the CNS. Analysis of the cytotoxic T lymphocyte (CTL) response in the CNS of mice infected with L* mutant virus, suggests that the inhibition of apoptosis by L* may prevent development of virus-speciWc cytotoxicity, permitting persistent virus infection and demyelinating disease (Lin et al., 1999). TMEV persistence in susceptible mice is associated with continuous virus replication and infectious virus and viral RNA can be isolated from the CNS for many months after initial infection. There is no single obvious cellular site of virus persistence in the spinal cord. Both viral antigen and RNA have been reported in oligodendrocytes, macrophages, and astrocytes (Aubert et al., 1987; Brahic et al., 1981; Lipton et al., 1995; Rodriguez et al., 1983). The typical observation that only very small numbers of infectious virus can be recovered from chronically infected spinal cord initially led to the notion that the virus persists at low levels in the tissue. However, recent quantitative studies on the level of viral RNA in the CNS have demonstrated the presence of large amounts of full-length viral RNA, about 109 copies/spinal cord, throughout the course of the chronic disease (Trottier et al., 2001). In contrast, infectious virus reaches a peak of >106 PFU/spinal cord by 15 days after intracerebral inoculation but then falls signiWcantly to 102 to 104 PFU/spinal cord after 28 days. This diVerence between viral genome load and infectious virus may reXect the rapid neutralization of virus by virus-speciWc antibodies. These data indicate that active viral replication continues throughout persistent viral infection. The identiWcation of a primary cellular site and speciWc mechanism of TMEV persistence has been diYcult. Based on immunostaining experiments, the predominant virus antigen burden in infected spinal cord has been reported to reside in either oligodendrocytes or macrophages (Lipton et al., 1995; Rodriguez et al., 1983). More recently, fractionation of macrophages and oligodendrocytes from the chronically infected CNS revealed large amounts of viral genome in both cell types (Trottier et al., 2001). Restriction of virus replication has been suggested as a mechanism for virus persistence and such restriction has been reported in macrophages isolated from TMEV infected CNS (Clatch et al., 1990; Levy et al., 1992). However, the ratio of plus- to minus-strand TMEV RNA appears to be similar in macrophages and oligodendrocytes from infected spinal cord, suggesting that neither cell type is generally restricted for virus replication (Trottier et al., 2001). As a potential site for TMEV persistence, macrophages have been the most widely studied. As mentioned earlier, restriction of virus replication has been reported in macrophages isolated from the TMEV infected CNS, and such restriction has been suggested as a possible mechanism for persistence. Depletion of blood borne macrophages early during the course of disease (7 to 21 days post-infection) decreases the amount of viral RNA in the CNS, supporting a role for macrophages in viral persistence (Rossi et al., 1997). Studies in macrophage cell lines have demonstrated a block at the point of virus assembly rather than replication and it has been demonstrated that the diVerentiation state of the macrophage may be critical for the ability of the cell to support persistence (Jelachich et al., 1995; Shaw-Jackson and Michiels, 1997). Similar experiments, to look at potential viral persistence in the precursors of other glial types, have not yet been reported. Other investigators have suggested a role in virus persistence for other CNS cells, such as astrocytes and vascular endothelial cells (Sapatino et al., 1995; Zheng et al., 2001). Whether there is a primary cellular site and a unique mechanism for TMEV persistence in the chronically infected CNS, is still an unresolved matter.

au5

1082

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

Host Determinants of Resistance and Susceptibility Inbred strains of mice can be either resistant or susceptible to persistent TMEV infection and demyelinating disease. Upon intracerebral infection with virus, resistant strains mount an anti-viral immune response that controls virus replication and eventually clears virus from the animal. Upon clearance, infectious virus, virus antigen, and viral RNA can no longer be detected. Susceptible strains also mount a strong immune response against virus but virus is not cleared. A persistent viral infection is established along with the development of demyelinating disease, as described earlier. Since virus clearance is mediated by the immune response, mutations that disrupt major functional elements of the immune system generally result in some level of susceptibility, even in an animal with an otherwise resistant genetic background. The eVects of mutations of this type will be discussed in subsequent sections. Here we review naturally occurring genetic loci that act as modiWers of the response to TMEV infection and that have been identiWed using standard techniques of genetic analysis in mice. Early studies with diVerent strains of mice demonstrated a wide range of susceptibility to TMEV-induced disease (Lipton and Del Canto, 1979). F1 and F2 backcrossing experiments between resistant and susceptible strains revealed that several loci were important for virus clearance and that one of these loci was linked to the H-2 major histocompatau6 ibility complex on chromosome 17 (Lipton and Melvold, 1984). Further genetic analysis of this region showed that mice with certain H-2 haplotypes (H-2f, p, q, r, s, v ) are susceptible to persistent virus infection, while others (H-2d,b,k) conferred resistance (Rodriguez and au7 David, 1985). Resistance was a dominant eVect (Patick et al., 1990) and mapped to the D region of the H-2 complex, suggesting that a class-I mediated element of the immune response was essential for virus resistance (Clatch et al., 1985; Rodriguez et al., 1986b). au8 Direct demonstration that an H-2 linked class-I gene was responsible for virus resistance was provided when the H-2Dd class-I gene was introduced into a susceptible mouse strain and these transgenic animals were shown to be resistant (Azoulay et al., 1994). A key role for class-I restricted CD8þ T cells as eVectors of resistance has since been conWrmed with targeted mutations in CD8 and b2-microglobulin (which lacks MHC class-I function). Both of these mutations result in TMEV susceptibility in animals with otherwise resistant genetic backgrounds (Fiette et al., 1993; Murray et al., 1998; Pullen et al., 1993; Rodriguez et al., 1993). The H-2Db gene exerts its eVect on the stimulation of virus-speciWc CTL through the presentation of an immunodominant peptide from the VP2 capsid protein: VP2121–130 (Dethlefs et al., 1997; Johnson et al., 1999). There are other class-I molecules whose peptide binding sites are identical to those of the D locus and therefore have the potential to present TMEV peptide. These molecules, however, have not been shown to play a role in establishing resistance, possibly due to diVerential patterns of expression or to structural diVerences outside of the peptide binding domain that are important for antigen presentation (Altintas et al., 1993; Lin et al., 1997a). Non-MHC loci also have an inXuence on virus resistance/susceptibility. The T-cell receptor genes lie on chromosome 6 (the Tcrb locus) in the mouse, and genetic mapping studies have indicated that a gene aVecting resistance maps close to the Tcrb locus. Several commonly used strains of susceptible mice have large deletions of the Vb T-cell receptor genes, further suggesting a link between T-cell function and virus resistance (Rodriguez et al., 1992). Genetic analysis of congenic mice with speciWc Vb deletions indicates that in susceptible genetic backgrounds, Vb deletions enhance susceptibility but that have little eVect on resistance. These data generally support a role for T-cell receptor genes in determining virus resistance (Rodriguez et al., 1994). Mapping experiments using the susceptible SJL strain and the resistant B10.S strain, two strains that both have the same H-2S haplotype, have identiWed loci on chromosomes 10 and 18 (Bihl et al., 1999; Bureau et al., 1993). There are two loci close to the gene for interferon-g (Ifn-g) on chromosome 10. While the possibility of a resistance locus at the site of the Ifn-g gene was of obvious interest, Wne mapping experiments have excluded Ifn-g as a candidate (Bihl et al., 1999). A second locus is near the gene for myelin basic protein (MBP) on chromosome 18. The exact identity of the gene involved has not yet been

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV)

au9

established but the possibility that TMEV resistance might be inXuenced by one of the major structural genes for myelin is intriguing. Studies on TMEV infection of MBP deWcient shiverer mice demonstrate enhanced susceptibility in the absence of MBP, supporting the idea that MBP might play a role in determining resistance (Bihl et al., 1997). Loci that aVect TMEV resistance and persistence have also been identiWed on chromosome 3 near the carbonic anhydrase-2 (Car-2) locus (Melvold et al., 1990), on chromosome 11 (Aubagnac et al., 1999), and on chromosome 14 (Bureau et al., 1998). The identity of the genes that are involved at these loci has not yet been established. Gender has also been reported to inXuence the susceptibility of certain mouse strains to TMEV-induced demyelinating disease (Kappel et al., 1990).

Mechanisms of Demyelination, Axonal Damage and the Development of Neurologic Deficits When considering TMEV-induced disease as a model for human demyelinating disease, questions regarding the mechanisms of demyelination and axonal damage are of central importance. Several diVerent mechanisms have been proposed to explain the pathology and neurologic deWcits that usually accompany persistent TMEV infection, but there is still disagreement in the Weld about the relative importance of the various potential mechanisms for the development of demyelination and axonal damage. Whatever the exact mechanisms of disease resistance and susceptibility, it seems certain that a balance between persistent virus infection and immune cell activation determines whether demyelinating disease will develop in immunocompetent mice. Mice with severe combined immunodeWciency (SCID) normally die from acute encephalitis when infected with TMEV. Adoptive transfer of normal spleen cells can result in survival and the development of demyelinating disease, but the results of adoptive transfer depend critically on the number of cells transferred (Rodriguez et al., 1996). Transfer of too few cells fails to ameliorate the SCID phenotype and most mice die. Transfer of too many cells results in complete virus clearance and the mice survive but do not develop disease. Only when an intermediate number of cells is transferred do the surviving mice develop persistent virus infection and demyelination. These results demonstrate the balance between immune activation and virus persistence that determines the quality and extent of demyelinating disease. Mechanisms of Demyelination Proposed mechanisms of TMEV-induced demyelination include (1) demyelination resulting from direct virus lysis of infected oligodendrocytes or as a ‘‘dying back’’ response of oligodendrocytes following infection; (2) TMEV-speciWc, immune-mediated destruction of infected oligodendrocytes; (3) ‘‘bystander’’ damage to glia by toxic mediators from activated macrophages; (4) ‘‘epitope spreading,’’ an autoimmune attack against myelin antigens resulting from the de novo processing and immune recognition of myelin antigens that follows the initial events of virus-induced myelin damage; and (5) ‘‘molecular mimicry,’’ an autoimmune attack against myelin antigens based on the activity of virusactivated T cells that are cross-reactive with myelin antigens. While several distinct mechanisms of demyelination have been proposed and supported by experimental evidence, this does not preclude the likelihood that diVerent combinations of these mechanisms may be active at various times during the course of TMEV-induced disease and in animals with varying genetic composition. Direct virus effects on oligodendrocytes While it is generally agreed that immune activation plays a major role in demyelination, several lines of evidence also support a role for direct eVects of viral infection on oligodendrocytes. In cultures of mixed neonatal brain cells, TMEV infection of oligodendrocytes is primarily lytic (Graves et al., 1986; Rodriguez et al., 1988). Two days after infection, many cultured oligodendrocytes display cytopathic ultrastructural changes characteristic of picornavirus infection, such as numer-

1083

1084

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

au10

ous cytoplasmic vacuoles and disruption of the endoplasmic reticulum, with preservation of mitochondria (see Fig. 44.2). The number of oligodendrocytes in these cultures begins to diminish within 48 hours after infection and by 7 days has decreased to 50% of the value in uninfected control cultures (Rodriguez et al., 1988). Since TMEV infects oligodendrocytes in vivo (Graves et al., 1986; Ohara et al., 1990; Rodriguez et al., 1988) and kills these cells in vitro, demyelination as a direct cellular consequence of the infection of oligodendrocytes seems plausible. A direct demonstration that infection can produce demyelination without the need for an accompanying activation of the immune response comes from studies of demyelination in nude mice which lack circulating T cells and cell mediated immunity (Roos and Wollmann, 1984; Rosenthal et al., 1986). In these studies, infection of BALB/c nu/nu mice with the DA strain of TMEV resulted in the development of signiWcant neurologic deWcits by 2 to 3 weeks and death within 4 to 8 weeks. Demyelinated foci were observed in the spinal cord by 14 days post-infection even though inXammatory inWltrates were minimal. Electron microscopy revealed numerous infected and degenerating glial cells, some that were identiWed as oligodendrocytes. Damaged myelin was observed around intact axons as well as completely demyelinated axons. In general, the demyelinated lesions in these mice were small and inXammation is minimal compared to what might be expected in most susceptible strains of mice, suggesting that other mechanisms of demyelination may normally play a more prominent role. However, these observations do indicate that TMEV-induced glial damage can occur in the absence of a T-cell mediated immune response. In immunocompetent mice, the cytoplasm of most oligodendrocytes from infected animals appears normal on electron microscopic examination (Dal Canto and Lipton, 1975; Rodriguez et al., 1983), even though viral RNA and antigen have been detected in oligodendrocytes throughout the course of persistent infection (Brahic et al., 1981; Rodriguez et al., 1983; Trottier et al., 2001). Rodriguez (Rodriguez et al., 1983; Rodriguez, 1985) reported ultrastructural abnormalities of the innermost extent of the myelin sheath and has proposed a ‘‘dying back’’ mechanism for demyelination. In persistently infected SJL mice, the inner tongue of the myelin sheath, which is the most distal extent of the myelin lamellae, often exhibits swelling and abnormalities in the normal cytoachitecture. Abnormalities included vacuolization, degeneration of mitochondria, increased numbers of microWlaments, and the appearance of electron-dense bodies and debris. These indications of damage to the innermost layer of the myelin sheath were observed in the absence of any signs of damage to the external surface of the sheath, indicating that damage to the oligodendrocyte may occur before immune-mediated damage to the myelin surface. Immunoperoxidase studies revealed that viral antigen was present on both the inner and outer extents of the myelin lamellae, even though the ultrastructural abnormalities were restricted to the inner loops. The observed degeneration appears to begin in the inner loop of the oligodendrocyte lamellae and then proceeds proximally toward the cytoplasm with the ultimate destruction of the myelin sheath. The ‘‘dying back’’ concept was initially proposed as a mechanism underlying peripheral neuropathies in which the cell body is unable to support the structure and metabolism of a distant cellular process (Spencer and Schaumburg, 1976). It seems particularly relevant that Ludwin and Johnson (1981) initially reported the ‘‘dying back’’ of oligodendrocytes in a very diVerent model for CNS demyelination in which demyelination is induced by ingestion of the toxic, copper-chelating agent, cuprizone. Dying back oligodendrogliopathy may therefore represent a general mechanism for demyelination in the CNS. TMEV-specific immune damage Dying back oligodendrogliopathy could be the direct result of viral infection on the cellular physiology of oligodendrocytes. However, it has been demonstrated that the demyelination observed following persistent TMEV infection in susceptible mouse strains is, at least in part, immune mediated (Lipton and Dal Canto, 1976b). Dying back might also be the result of damage due to TMEV-speciWc immune reactions directed against infected oligodendrocytes.

1085

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV)

In cultures of mixed neonatal brain cells, immunoperoxidase staining visualized with electron microscopy demonstrated large amounts of viral antigen on the oligodendrocyte cell membrane (Rodriguez et al., 1988). Similarly, staining in vivo can detect virus antigen in persistently infected oligodendrocytes by 28 days post-infection (Rodriguez et al., 1983). Antigen is present in the cell cytoplasm as well as in the myelin lamellae. Infected animals typically produce high titers of virus-speciWc antibody and the recognition of virus antigen on the cell surface by these immunoglobulins may result in oligodendrocyte injury through several antibody dependent mechanisms including complement activation, antibody-dependent cell-mediated cytotoxicity (ADCC), or receptor mediated phagocytosis by macrophages. Alternatively, viral antigen might also be presented on the cell surface in the context of the MHC class I or class II proteins. MHC class I, in particular, is known to be upregulated in oligodendrocytes following TMEV infection of the CNS (Altintas et al., 1993; Lindsley et al., 1992). In susceptible mouse strains, class-I MHC is up-regulated in the CNS by one day after infection and remains high in the spinal cord white matter throughout chronic infection and demyelination. A novel antigen such as viral peptide, presented on the cell surface by MHC class-I, might generate a class-I restricted cytotoxic T lymphocyte (CTL) response, which might cause primary demyelination as the result of chronic nonlethal injury to TMEV-infected oligodendrocytes. However, mice with a disrupted b2microglobulin (b2-m) gene do not express signiWcant levels of MHC class-I, do not have functional CD8þ T cells, and therefore cannot mount a CTL response. When b2-m knockout mice from a resistant haplotype are infected with TMEV, resistance is abrogated and the virus establishes persistent CNS infection. However, signiWcant demyelination develops in these mice (Fiette et al., 1993; Pullen et al., 1993; Rodriguez et al., 1993), demonstrating that CD8þ class I-restricted cytotoxic T cells are probably not the primary mediators of CNS demyelination. In general agreement with these observations, b2-m knockout mice from a susceptible haplotype display somewhat enhanced levels of demyelination (Begolka et al., 2001). Similar results have been obtained with mice carrying a genetic deletion of the gene encoding CD8, which display little alteration in the disease phenotype in either resistant or susceptible strains (Murray et al., 1998). Bystander damage In contrast to observations with MHC class-I and CD8, treatment of susceptible mice with neutralizing antibodies against MHC class-II (anti-Ia) and antiCD4 has been shown to reduce demyelination (Gerety et al., 1994; Rodriguez et al., 1986a; Welsh et al., 1987). These observations suggest that class-II restricted CD4þ T lymphocytes might play a role in disease pathogenesis. Clatch and coworkers (1986) attempted to correlate the development of clinical disease with several pathophysiologic parameters including virus titer, titer of anti-TMEV speciWc antibody, delayedtype hypersensitivity (DTH), and T-cell proliferative responses. The extent of TMEVinduced demyelinating disease correlated most closely with the presence of high levels of TMEV-speciWc DTH. The development of TMEV speciWc delayed-type hypersensitivity would require an initial presentation of virus antigen to TH cells by MHC class-II, resulting in activation and clonal expansion. Generally, the activated T cells are of the CD4þ TH1 subtype and are often referred to as TDTH cells. Upon secondary contact with antigen, TDTH cells secrete cytokines that recruit and activate macrophages and other nonspeciWc inXammatory cells. Development of the DTH response may take 48 to 72 hours and by the time the reaction is fully developed, the vast majority of the activated cells at the site of inXammation are nonspeciWc immune cells and macrophages which act as the primary eVector cells of the DTH response. These cells release a variety of cytotoxic mediators which results in nonspeciWc ‘‘bystander’’ damage to cells in the area of the response. In the CNS of an animal with persistent TMEV infection, any cell that presents virus antigen with MHC class-II might therefore serve as a nidus for the development of a DTH response. The concept of DTH as a mechanism of demyelination is supported by adoptive transfer experiments using the sTV1 T-cell line or using TMEV sensitized lymph node T cells. sTV1 is a DTH-mediating CD4þ TH1 line that is speciWc for the immunodominant

au11

1086

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

VP274–86 peptide of TMEV. Transfer of 5
106 peptide-stimulated sTV1 cells into SJL mice that had been infected with a suboptimal dose of virus, signiWcantly potentiated clinical disease in the recipient mice. Animals that received cells but did not receive virus to provide secondary cellular stimulation did not develop disease. An experimental scheme using lymph node T cells from mice primed with UV-inactivated virus instead of sTV1 cells resulted in similar potentiation of disease but only when virus was present for secondary stimulation (Gerety et al., 1994). Further supporting the notion of DTH mediated demyelination, induction of peripheral immune tolerance to TMEV signiWcantly reduces the development of DTH, general inXammation, virus-speciWc immune responses, and demyelination, following infection with the virus (Karpus et al., 1995).

au12

Epitope spreading While the precise mechanisms may not yet be clear, it is well accepted that TMEV persistence in the CNS stimulates a virus-speciWc immune response that initiates damage to CNS cells and myelin. Recent studies have also detected myelinreactive T cells in chronically infected mice. The appearance of T cells that are reactive to myelin autoepitopes is presumably due to the development of an immune response to myelin antigens that are released during the initial stages of TMEV-induced demyelination. It has been proposed that these autoreactive T cells are pathogenic and that they may be a major source of pathology during the chronic stages of demyelinating disease. The generation of a myelin-speciWc pathologic immune response, following an initial event of virus-induced demyelination, is called ‘‘Epitope spreading’’ (Lehmann et al., 1993; McRae et al., 1995). Earlier attempts to detect myelin reactive T cells from TMEV infected mice had been unsuccessful (Miller et al., 1990). However, Miller and coworkers (1997) have now reported that 3 to 4 weeks after initiation of demyelinating disease, T-cell responses to a series of myelin autoepitopes begin to appear in an ordered progression. T-cell proliferation and DTH responses to myelin peptides were Wrst demonstrated at about 50 days post-infection. By 164 days post-infection, reactivity was shown for multiple peptide epitopes from proteolipid protein (PLP), myelin basic protein (MBP), myelin oligodendrocyte glycoprotein (MOG), and mouse spinal cord homogenate. T-cell hybridoma clones speciWc for the immunodominant peptides of TMEV, and lymph node T cells from TMEV primed mice, did not show reactivity to myelin peptides. Conversely, lymph node T cells from myelin primed mice did not show reactivity to TMEV peptides. These results suggest that the reactivity to myelin epitopes was speciWc and not due to cross-reactivity with TMEV. Additionally, sequence comparison of the reactive myelin peptides to TMEV showed no strong homologies that might indicate a likely source of cross-reactivity. An epitope spreading mechanism for TMEV-induced demyelinating disease would be very similar to the pathologic mechanism of EAE. Adoptive transfer of myelin reactive immune cells from animals with EAE eVectively transfers the disease to naı¨ve animals. However, attempts to transfer TMEV-induced demyelinating disease by the transfer of T cells from infected animals have been unsuccessful (Barbano et al., 1984). Initial attempts to suppress TMEV-induced disease by inducing neuroantigen-speciWc tolerance to a heterogeneous mixture of CNS antigens were unsuccessful (Miller et al., 1990). More recently, induction of myelin-speciWc tolerance with a fusion protein containing MBP and PLP sequences has been reported to reduce demyelination and CNS inXammation in mice with TMEV-induced demyelinating disease (Neville et al., 2002). Studies of TMEV-induced disease in CD4 knockout mice are in potential disagreement with the idea of CD4-dependent mechanisms of demyelination such as bystander damage (DTH) or epitope spreading. Murray and coworkers (1998) examined the eVect of CD4 deletion in both resistant (C57BL/6) and susceptible (SJL and PLJ) genetic backgrounds. In a resistant background, the absence of CD4 resulted in the establishment of persistent viral infection in the CNS and with the development of demyelination by 90 days postinfection. In both susceptible strains, demyelination was dramatically increased. It is not immediately obvious how best to address and reconcile the various observations described here. It might be noted, however, that CD4 plays an important role in immune control of viral infection and that TMEV titers in CD4 knockout mice were generally

THEILER’S MURINE ENCEPHALOMYELITIS VIRUS (TMEV)

elevated during the disease. The development of TMEV-induced pathology likely reXects a delicate balance between viral titer and activity and a dynamic immune response to the virus. As experimental protocols alter various aspects of this balance, such as virus titer and the quality and quantity of the participating immune cells, the mechanisms of pathology may also change in ways that make the interpretation of the data less straightforward than expected. In addition to the complexity of the system, diVerent strains of TMEV are currently being used by the most active research groups. The BeAn 8386 and Daniel’s (DA) strains are both widely used and although both establish robust, persistent infections of the spinal cord, the BeAn strain is generally less virulent that the DA strain. The precise signiWcance of the use of these two strains to the interpretation and comparison of data is not yet clear. Molecular mimicry Like the epitope spreading model, molecular mimicry involves pathogenesis due to myelin reactive autoimmunity. However, in molecular mimicry, the autoimmune response results from structural homologies between TMEV and myelin proteins. Immune recognition of the virus therefore also gives rise to an anti-myelin response, and this autoimmunity is pathogenic. While molecular mimicry is probably an important mechanism for other CNS diseases (Rouse and Deshpande, 2002; Yuki, 2001), several observations suggest that it may not play a prominent role in TMEV-induced disease. TMEV infection is followed by a rapid and robust immune response that includes the development of both TMEV-speciWc antibodies and T cells by 7 days post-infection. Demyelinating disease is Wrst observed by about 21 days post-infection and does not reach a peak until 90 to 100 days. The temporal diVerence between the development of the TMEV-speciWc immune response and the appearance of demyelinating disease argues against molecular mimicry as an important component of the demyelination mechanism. Furthermore, the initial immune response to TMEV does not cross-react with several myelin antigens including MBP, PLP, and MOG (Miller et al., 1997). No strong homologies have been reported between TMEV protein sequences and those of the major myelin proteins. Mechanisms of Axonal Damage and the Development of Neurologic Deficits In the past, demyelination alone has often been considered to be suYcient for the development of axonal damage and neurologic deWcits. However, both the timing and mechanisms of demyelination and axonal damage can be distinct and diVerent in the TMEV model. Quantitative assessment of demyelination and neurologic ability in two susceptible mouse strains shows that small demyelinated lesions, accompanied by statistically signiWcant decreases in motor ability, can occur as early as 24 days post-infection (McGavern et al., 1999). However, axonal damage and neurologic deWcits do not necessarily follow demyelination. Several reports have documented that deWciency of MHC class-I (b2-microglobulin mutants) or MHC class-II (Abo mutants) function, will abrogate resistance to TMEV-induced disease in otherwise resistant strains of mice (Fiette et al., 1993; Njenga et al., 1996; Pullen et al., 1993; Rodriguez et al., 1993). Resistant mice that are homozygous for either of these mutations develop extensive demyelination. However, class-II deWcient mice develop severe neurologic deWcits as indicated by signiWcant reductions in spontaneous activity, while class-I deWcient mice show no clinical signs. Electrophysiologic measurement of the conduction velocity of motor-evoked potentials shows that decreased conduction velocities accompany deWcits in class-II deWcient mice with chronic demyelination. Conduction velocities in class-I deWcient mice are no diVerent that those in uninfected controls (Rivera-Quinones et al., 1998). These Wndings demonstrate that demyelination and neurologic deWcits are genetically and functionally separable, and they implicate a role for MHC class-I in the development of neurologic deWcits following demyelination. A role for MHC class-I in the development of neurologic deWcits implicates CD8þ T cells as the pathologic eVector cell for the induction of neurologic disease. Granule exocytosis (perforin release) and Fas ligand expression are two of the common cytopathic eVector functions of CD8 cells, and mice deWcient for perforin, Fas, and Fas ligand were tested to determine whether these molecules play a role in the development of

1087

1088

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

demyelination and clinical disease (Murray et al., 1998). Lpr (Fas mutation) and Gld (Fas ligand) mutant mice, on a TMEV-resistant genetic background, maintained resistance to viral persistence and demyelinating disease. Perforin-deWcient mice developed persistent viral infection and chronic demyelination of spinal cord white matter. Despite demyelination and virus persistence, these mice showed only minimal neurologic deWcits. These studies indicate that perforin release is the pathologic eVector for the induction of neurologic disease by CD8þ T cells. The experiments described earlier for the determination of the roles played by class-I and class-II related mechanisms for the development of axonal damage were all performed in mice with the C57BL/6 or C57BL/6
129 genetic background. C57BL/6 mice mount a vigorous CTL response to TMEV. This response is predominantly directed against a single immunodominant TMEV peptide that represents amino acids 121 to 130 of the VP2 capsid protein, which is presented in the context of the H-2Db MHC class-I molecule (Borson et al., 1997). At 7 days post-infection, staining of CD8þ brain inWltrating lymphocytes with Db:VP2121–130 tetramers shows that 50 to 63% of the inWltrating CD8þ cells are virus speciWc (Johnson et al., 1999). Mice with a targeted disruption of the Ifn-g receptor gene respond to virus infection with rapidly progressing disease and severe clinical deWcits. These mice also develop a class-I restricted anti-viral response that is dominanted by the VP2121–130 epitope. In these mice, intravenous injection of the VP2 peptide one day before virus infection completely eliminates the immunodominant T-cell response (Johnson et al., 2001). Elimination of the response does not cause a signiWcant increase in viral titer. There is no change in the overall pattern of infection: by 45 days virus is primarily localized to the white matter of the spinal cord, there is no signiWcant increase in inXammation, and demyelination is similar in extent to control animals. Elimination of the response did, however, signiWcantly preserve motor function in infected animals. At 45 days after infection, VP2 peptide treated animals performed signiWcantly better than control animals in tests of motor ability, although not as well as uninfected animals. These results suggest that the CD8þ T cells that mediate neurologic dysfunction are most likely the TMEVspeciWc cells that inWltrate the CNS following infection. In contrast to the work done in mice with the TMEV-resistant C57BL/6 genetic background, susceptible SJL mice that lack CD8þ eVector function (b2-microglobulin knockouts) are not rescued from TMEV-induced neurologic deWcits (Begolka et al., 2001). Instead, these mice show an earlier onset of neurologic disease, with enhanced CNS demyelination and macrophage inWltration at 50 days post-infection. These results indicate that in these mice, CD8þ cells are not required for the initiation or progression of demyelinating disease and may illustrate the strong inXuence that genetic background can have on the development of this disease.

Remyelination Following TMEV-Induced Demyelination Myelin repair, or remyelination, is a normal physiologic response to myelin damage. However, the ability of mice to repair the damage caused by TMEV-induced demyelinating disease is strongly dependent on the genetic constitution of the animal. b2-microglobulin deWcient mice lack MHC class-I function. On a C57BL/6
129 genetic background, these mice develop extensive demyelination following TMEV infection (Rivera-Quinones et al., 1998). Despite the presence of persistent viral infection, these animals show extensive remyelination by 6 months post-infection (Miller et al., 1995). This spontaneous remyelination of CNS axons involves both oligodendrocytes and Schwann cells. In contrast, MHC class-II deWcient mice on the same genetic background also develop demyelinating disease but subsequent remyelination is largely absent (Njenga et al., 1999). Class-II deWcient mice fail to produce TMEV-speciWc IgG and eventually developed high viral titers. However, the animals do not die during the acute phase of the disease but survive well into the chronic demyelinating phase, even though most die by 120 days postinfection. Failure to observe remyelination may simply reXect insuYcient time for myelin repair to occur before death. SigniWcant axonal damage is also apparent in these mice and this damage may also preclude remyelination.

MOUSE HEPATITIS VIRUS (MHV)

Like MHC class-II deWcient mice, the SJL strain is susceptible to TMEV infection, develops extensive demyelinating disease, and does not repair myelin damage. In contrast to class-II deWcient mice, SJL mice control viral infection eVectively and can survive for over a year with persistent infection and chronic demyelinating disease. The disease in SJL mice progresses from demyelination, through mild neurologic deWcits, to profound deWcits and paralysis. Two general hypotheses have been proposed to explain the absence of remyelination in these mice (Miller et al., 1996). First, an inhibitory local environment may be present, which prevents spontaneous repair. Local inXammation, reactive astrocytes, microglia, or damaged axons might play a role. Second, an absence of growth factors or oligodendrocyte precursor cells may prevent eYcient remyelination. Which of these mechanisms is the most important remains unclear. The absence of spontaneous remyelination in SJL mice, however, has provided an excellent experimental system for the study of ways to enhance the remyelination process, and several potential therapeutic approaches have been studied and developed using this system (Asakura et al., 1998; Drescher et al., 1998; Njenga et al., 2000; Ure and Rodriguez, 2002; Warrington et al., 2000). Whatever the mechanisms of remyelination, it seems clear that remyelination is associated with improved neurologic function. Like the SJL strain, PL/J mice develop chronic demyelinating disease but do not remyelinate. PL/J mice that are deWcient for either CD8þ or CD4þ T cells show a signiWcant increase in the severity of disease pathogenesis (Murray et al., 1998). The disease is particularly severe in PL/J CD4 mutant mice. While early mortality is common in these animals, many survive up to 6 months post-infection, and many of the mice that survive to the chronic phase of the disease have nearly complete spontaneous remyelination. Importantly, all of the surviving mice show partial recovery of motor function that coincides with myelin repair, and the extent of recovery correlates strongly with the percentage of lesioned white matter area that has remyelinated (Murray et al., 2001). These observations clearly demonstrate that functional recovery is possible despite previous demyelination and ongoing persistent virus infection, and that remyelination correlates with this recovery.

MOUSE HEPATITIS VIRUS (MHV) Murine hepatitis virus (MHV) is a member of the Coronaviridae, a family of large, enveloped, plus-strand RNA viruses. MHV is a common enteric pathogen of mice and rats that occasionally disseminates to the CNS. Like TMEV, the original isolation and characterization of one of the most commonly studied strains, MHV-JHM, occurred with the spontaneous appearance of two paralyzed mice, and the subsequent passage of the virus in mouse brain were it continued to produce paralytic disease (Cheever et al., 1948; Pappenheimer, 1958). Later passages of the virus resulted primarily in encephalitic disease with accompanying demyelination. The demyelinating phase of MHV infection has been used as a model for human demyelinating disease

Virus Biology and Life Cycle The MHV genome is a 31 kb message-strand RNA that is capped and polyadenylated. The genomic RNA is complexed with the viral nucleocapsid phosphoprotein (N) to form a long, helical nucleocapsid. The nucleocapsid lies within a lipoprotein envelop that consists largely of intracellular host cell membrane and three viral capsid proteins: membrane protein (M), spike protein (S), and hemagglutinin-esterase protein (HE). MHV virions bind to the plasma membrane of the host cell by the interaction of the spike proteins with the host receptor. Biliary glycoprotein 1a (Bgp1a), a member of the immunoglobulin superfamily, is a known MHV receptor (Williams et al., 1991). The main determinant of cellular tropism is the ability of the spike protein to react with a suitable receptor on the cell surface and infected cell populations in the animal closely match those

1089

1090

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

with demonstrated expression of the Bgp1a receptor. However, very little Bgp1a is expressed on CNS glia and neurons, and the mechanism of virus entry into these cells is still uncertain. The MHV virion also contains the HE protein. Binding of HE to 9-O-acetylated neuraminic acid residues on the plasma membrane may serve as an important pre-receptor function that facilitates viral entry. After receptor binding, the viral envelope fuses with the plasma membrane in an Sprotein mediated event. Once the RNA genome has entered the cytoplasm of the cell, it is translated into a polyprotein that is co- and post-translationally processed into multiple proteins. These proteins, which include the capsid proteins (N, M, S, and HE), the virusspeciWc RNA-dependent RNA polymerase, and virus encoded proteases, play various roles in the replication and development of infectious virus. The RNA-dependent RNA polymerase uses the positive-strand genomic RNA as a template to produce full-length negative-strand RNA, which in turn is used to make more positive-strand genomic RNA. Overlapping sets of 3’ co-terminal subgenomic positive-strand RNAs are also produced and translated into viral gene products. The newly synthesized genomic RNA assembles with nucleocapsid (N) protein to form the helical nucleocapsid. The S, HE, and M glycoproteins are all translated on membrane bound polysomes and are co-translationally inserted into intracellular membranes. Virions begin to form when nucleocapsids bind to intracellular membranes that contain the viral M protein and virus then buds from these specialized membranes. S and HE proteins are incorporated into the virions during budding while host proteins are excluded. After budding, mature virions accumulate in large, smooth-walled vesicles that release virus when they fuse with the plasma membrane.

Disease Pathology The outcome of MHV infection depends on several variables such as virus strain, virus dose and route of infection, age, and genetic composition of the host. Two diVerent MHV strains, MHV-A59 and MHV-JHM, have been used in most studies of MHV-induced demyelination. As the name suggests, hepatitis is a common outcome of MHV infection but while infection with the MHV-A59 strain results in prominent hepatitis, MHV-JHM infected animals show little or no evidence of hepatitis. However, intracerebral or intranasal infection with either strain results in acute, sometimes fatal, encephalitis, followed by the development of chronic demyelination in animals that survive. To decrease mortality and thereby facilitate the study of MHV pathogenesis, several MHV-JHM variants have been developed that have limited neurotropism related lethality but that continue to produce an early encephalitis followed by acute primary demyelination (Erlich et al., 1987; Fleming et al., 1986; Haspel et al., 1978). MHV infects a variety of CNS cell types including astrocytes, oligodendrocytes, microglia and neurons. As virus replicates to high titers in the brain and spinal cord, it is controlled by the development of an anti-viral immune response that is largely restricted to the CNS. Control of infection correlates well with the appearance of an inXammatory inWltrate by 4 to 7 days post-infection. This acute inXammatory response is characterized by the inXux of many types of immune eVector cells including CD4þ and CD8þ T cells, B cells, macrophages, and natural killer cells (Williamson et al., 1991). Virus neutralizing antibodies appear in the serum by 7 to 9 days post-infection but probably do not play a critical role in initial virus clearance. Infectious virus is cleared from the CNS by 10 to 14 days but complete clearance is not achieved and virus persists in the CNS. Infectious virus is rarely recovered from spinal cord tissue after the acute phase of infection. Virus antigen can be detected for several months after initial clearance and appears to persist in astrocytes and oligodendrocytes (Sun et al., 1995). Viral RNA can be detected in the CNS for at least a year after infection (Lavi et al., 1984). Demyelination with axonal sparing can be detected as early as 5 days post-infection. MHV-induced demyelination is illustrated in Figure 44.5. Relatively few inXammatory cells but large numbers of debris containing macrophages characterize demyelinated lesions, and their appearance correlates well with the development of lesions (Lane et al., 2000). Electron microscopy demonstrates macrophage processes extending between the

au13

MOUSE HEPATITIS VIRUS (MHV)

FIGURE 44.5 Pathology of the MHV-infected spinal cord. (A) A section of spinal cord tissue from an SJL mouse infected with MHV-A59 and then stained for myelin with para-phenylene-diamine shows multiple white matter lesions (arrows). (B) Normal white matter lies adjacent to a demyelinated lesion in the ventral spinal cord tracts. Myelin destruction in the lesioned area is apparent.

layers of myelin sheaths, strongly suggesting an active role for macrophages in the demyelination process (Powell and Lampert, 1975). Demyelination continues to develop until about 21 days after infection and then begins to slow. Spontaneous remyelination is common following the initial phase encephalitis and demyelination. After the initial phase of encephalitis, demyelination, and viral clearance, virus persistence results in a second phase that probably continues for the life of the animal and that is associated with the recurrent episodes of demyelination.

Virus Resistance Infected adult mice generate a potent immune response that clears the virus in 10 to 14 days. CD4þ and CD8þ T cells are the most critical elements of this response. Viral infection results in the rapid recruitment of large numbers of CD8þ T cells to the CNS parenchyma, while CD4þ cells are primarily localized to the perivasculature and meninges. CD8þ T cells are the primary immune eVectors of virus clearance. MHC class-I tetramer staining shows that over 50% of the CD8þ T cells in the CNS after infection are virus speciWc (Bergmann et al., 1999). As virus is cleared, the total number of CD8þ cells in the

1091

1092

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

CNS decreases but the percentage of cells that are virus-speciWc remains high. Initially the inWltrating CD8þ cells display virus-speciWc cytolytic activity. As clearance progresses and persistence is established, these cells lose their cytolytic function while retaining their ability to secrete Ifn-g, indicating diVerential regulation of CD8þ eVector functions as the course of infection progresses. Consistent with a role for CD8þ cells in virus clearance, CD8 deWcient mice show increases in viral titer, increased mortality, and an inability to eYciently clear virus after infection (Lane et al., 2000; Williamson et al., 1990). CD8þ cells probably use two diVerent eVector mechanisms to clear virus from diVerent subsets of infected cells. Perforin-mediated cytolysis may play an essential role in the control of virus infection of astrocytes and microglia. Perforin deWcient mice are able to clear MHV from the CNS but the clearance is delayed by several days (Lin et al., 1997b), indicating that perforin-dependent CTL plays a role in viral clearance from the CNS. Histologic analysis has shown that virus-speciWc CTL are more eYcient at the clearance of virus from astrocytes and macrophages than from oligodendrocytes (Stohlman et al., 1995). Ifn-g is the antiviral mechanism primarily responsible for controlling the infection of oligodendrocytes (Parra et al., 1999). In Ifn-g deWcient mice, MHV infection results in increased neurologic disability and mortality that is associated with increased viral titers and delayed viral clearance. The anti-viral antibody response and the CTL response are not aVected. Increases of viral antigen in Ifn-g deWcient mice are located almost exclusively in oligodendrocytes and are accompanied by increased numbers of CD8þ cells in CNS white matter. CD8þ cells are therefore major participants in MHV clearance and employ at least two diVerent eVectors mechanisms to clear virus from diVerent types of infected cells. Some studies have also suggested that the Fas-Fas ligand system may play a role (Parra et al., 2000). Although the ability to control MHV replication is predominantly controlled by CD8þ cells, this response is highly dependent on the CD4þ T-cell response (Stohlman et al., 1998). Adoptive transfer of activated CTL into recipients that have been depleted of CD4þ cells shows that although the CD8þ cells can eVectively traYc to sites of CNS infection in the absence of CD4þ cells, once they arrive they rapidly undergo apoptosis. These observations indicate that the survival of the CD8þ cells that are primarily responsible for virus clearance may depend critically on support from CD4þ cells. Consistent with these observations, studies on CD4 deWcient mice show increases in viral titer, increased mortality, and an inability to eYciently clear virus, a pathologic phenotype very similar to that observed for CD8 deWcient mice (Lane et al., 2000). CD4 deWcient mice also exhibit a signiWcant reduction in the number of activated macrophages and microglia in the CNS of infected animals (Lane et al., 2000). This may result from a decrease in the expression of the chemokine RANTES that is secreted by CD4þ cells and acts to recruit inWltrating mononuclear cells to the CNS. In strong support of this idea, in vivo depletion of RANTES with RANTES-speciWc antiserum mimics the phenotype seen in CD4 deWcient mice. Despite the increases in viral titer, increased mortality, and inability to clear virus, CD4 mutant mice that survive infection show a general decrease in inXammation and demyelination. It has been proposed that these decreases may result directly from a reduction in the inWltration of macrophages that are involved in myelin destruction. However, depletion of blood-borne macrophages with toxic liposomes prior to infection has little eVect on demyelination (Xue et al., 1999). Large numbers of activated macrophages/microglia are still present in the CNS after liposome treatment suggesting that the activity of resident microglia may play a central role in demyelination. Macrophage depletion prior to infection does result in increased mortality, suggesting a role for macrophages in the control of viral titers prior to the development of the T-cell responses (Wijburg et al., 1997; Xue et al., 1999). The role of antibody in viral clearance and the establishment of persistence has been studied in mice that lack a normal humoral immune response (Lin et al., 1999). Mice homozygous for a disruption in the immunoglobulin mu gene lack B cells. In these mice, virus is cleared normally during acute infection but then virus titers increase dramatically after about 11 days post-infection, when the titer of infectious virus in wild-type animals

MOUSE HEPATITIS VIRUS (MHV)

has fallen to undetectable levels. Passive transfer of MHV immune serum following initial virus clearance prevents this secondary rise in virus titers. These data demonstrate that initial clearance of virus is largely antibody independent but that antibody may play a key role in controlling the reemergence of virus in the CNS during persistent infection.

Mechanisms of Demyelination and Axonal Damage The mechanisms of demyelination during MHV infection are not yet entirely clear. A priori, the potential mechanisms involved in MHV-induced demyelination might be similar to those involved in TMEV-induced demyelinating disease—that is, direct virus lysis of infected oligodendrocytes, virus-speciWc immune-mediated destruction of oligodendrocytes and myelin, macrophage-mediated ‘‘bystander’’ damage to oligodendrocytes and myelin, and myelin damage by ‘‘epitope spreading’’ or ‘‘molecular mimicry.’’ Several studies have shown that virus actively replicates in oligodendrocytes and that morphologic and biochemical changes accompany oligodendrocyte infection in vivo (Barac-Latas et al., 1997; Powell et al., 1975). Down-regulation of oligodendrocyte gene expression along with oligodendrocyte destruction by necrosis and apoptosis have been reported for both the early and late stages of demyelinating disease (Barac-Latas et al., 1997). These types of observations have supported the suggestion that direct viral cytolysis might play a role in the development of demyelination. However, it seems clear that an immune component is essential for the full development of demyelinating disease. JHM-MHV infection of severe combined immunodeWciency (SCID) mice, which lack T lymphocytes, does not result in either demyelination or paralysis (Houtman and Fleming, 1996). Similarly, mice deWcient in recombinase-activating gene (RAG) activity lack both Band T-cell function and do not develop demyelination after MHV infection (Wu and Perlman, 1999). Infected RAG mice, whose immune system has been reconstituted by the adoptive transfer of spleen cells from genetically identical immunocompetent mice that have been immunized for MHV, developed demyelination within 7 to 9 days after transfer. Demyelination rarely occurs if donor mice have not been preimmunized against MHV, demonstrating that MHV-speciWc T cells are critical for the development of demyelination. The demyelination that follows reconstitution of immune function is characterized by the extensive recruitment of activated macrophages and microglia to sites of demyelinating lesions, consistent with a potential role for these cells in the development of myelin damage. These observations are also consistent with the idea that RANTES, secreted by CD4þ cells, plays a key role in the recruitment of macrophages to lesions and that activated macrophages and microglia may be primary mediators of myelin destruction (Lane et al., 2000). A mechanism of CD4þ-dependent, nonspeciWc damage mediated by inWltrating macrophages bears strong resemblance to the ‘‘bystander’’ damage induced by the development of a DTH response in the TMEV model. Although these data suggest a signiWcant role for the immune system in the development of demyelination, a recent report on the infection of RAG mice with MHV-A59 demonstrated typical demyelinated lesions (Matthews et al., 2002). T cells may therefore play a central role in the development of demyelination but other mechanisms may also be important, and the relative contributions of speciWc mechanisms may depend on variables such as virus strain and the host genetic background. The question of axonal damage after MHV infection is separate from the question of demyelination. MHV infection is generally characterized by demyelination with axonal sparing, but axonal injury at sites of demyelination has been described (Stohlman and Weiner, 1981). In contrast to the TMEV model, where neurologic deWcits tend to occur well after primary demyelination is established, most studies have shown that MHV induced paralysis and demyelination occur almost simultaneously, indicating that axonal damage occurs early in disease. In the experiments described earlier, animals that lack T cells, such as RAG and SCID mice, do not develop demyelination, but reconstitution of these animals by the adoptive transfer of spleen cells results in demyelination by 7 days post-transfer. Axonal damage appears with exactly the same pattern following MHV

1093

1094

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

infection in these mice (Dandekar et al., 2001). Using the appearance of nonphosphorylated neuroWlament as a marker for axonal damage, the onset of axonal damage was shown to be coincident with the appearance of demyelination. Damaged axons were detected not only at sites of active demyelination but also in areas with heavy macrophage inWltration but no obvious demyelination. Axonal damage therefore occurs very early in the disease process and nonspeciWc inXammation may play a key role. In the TMEV model, the secondary development of a pathogenic immune response directed against myelin antigens or ‘‘epitope spreading’’ has been proposed as an important mechanism for the exacerbation and maintenance of demyelination. JHM-MHV infection of either Brown Norway or Lewis rats results in a subacute demyelinating disease. Infection of Lewis rats also results in lymphocytes that proliferate in response to myelin basic protein and that produce mild encephalitis upon adoptive transfer into naı¨ve recipients (Watanabe et al., 1983). Lymphocytes from infected Brown Norway rats did not develop MBP sensitivity, demonstrating a strong host strain dependence for this response (Watanabe et al., 1987). The production of pathogenic anti-myelin antibodies demonstrates potential for an epitope spreading mechanism of MHV-induced disease, but the extent to which this mechanism contributes to the development of the chronic demyelinating disease is an unresolved issue. Antibodies, either autoreactive or anti-viral, might directly mediate demyelination by mechanisms such as antibody-dependent or complement-mediated cytotoxicity. In general, little correlation has been found between anti-MHV antibody titers and the severity of demyelination. The role of B cells and antibody has been studied directly in mice with mutations in the immunoglobulin mu gene; these mice lack B cells and make no antibody. In B-cell deWcient mice, demyelination following MHV infection is signiWcantly more severe than in controls at 30 to 60 days post-infection (Matthews et al., 2002). This increase may be due to the inability of the animals to control virus after the acute phase of the disease. Animals that can produce antibody but lack antibody receptors (Fc) or components of the complement pathway are able to clear virus but still develop demyelination (Matthews et al., 2002). These data indicate that antibody and antibody-dependent eVector mechanisms are probably not central to the demyelination process. Demyelination after MHV infection is clearly a complex event and probably results from some combination of direct viral destruction and immune-mediated damage. There does not, however, appear to be any single lymphocyte or monocyte function that is primarily responsible for the immune component of the damage. Both MHV-speciWc and nonspeciWc immune responses are probably involved.

References

au14

Adami, C., Pritchard, A. E., Knauf, T., Luo, M., and Lipton, H. L. (1998). A determinant for central nervous system persistence localized in the capsid of Theiler’s murine encephalomyelitis virus by using recombinant viruses. J. Virol. 72, 1662–1665. Altintas, A., Cai, Z., Pease, L. R., and Rodriguez, M. (1993). DiVerential expression of H-2K and H-2D in the central nervous system of mice infected with Theiler’s virus. J. Immunol. 151, 2803–2812. Asakura, K., Miller, D. J., Pease, L. R., and Rodriguez, M. (1998). Targeting of IgM kappa antibodies to oligodendrocytes promotes CNS remyelination. J. Neurosci. 18, 7700–7708. Aubagnac, S., Brahic, M., and Bureau, J. F. (1999). Viral load and a locus on chromosome 11 aVect the late clinical disease caused by Theiler’s virus. J. Virol. 73, 7965–7971. Aubert, C., Chamorro, M., and Brahic, M. (1987). IdentiWcation of Theiler’s virus infected cells in the central nervous system of the mouse during demyelinating disease. Microb Pathog. 3, 319–326. Azoulay, A., Brahic, M., and Bureau, J. F. (1994). FVB mice transgenic for the H-2Db gene become resistant to persistent infection by Theiler’s virus. J. Virol. 68, 4049–4052. Barac-Latas, V., Suchanek, G., Breitschopf, H., Stuehler, A., Wege, H., and Lassmann, H. (1997). Patterns of oligodendrocyte pathology in coronavirus-induced subacute demyelinating encephalomyelitis in the Lewis rat. Glia 19, 1–12. Barbano, R. L., and Dal Canto, M. C. (1984). Serum and cells from Theiler’s virus-infected mice fail to injure myelinating cultures or to produce in vivo transfer of disease. The pathogenesis of Theiler’s virus-induced demyelination appears to diVer from that of EAE. J. Neurol. Sci. 66, 283–293.

MOUSE HEPATITIS VIRUS (MHV)

au15

au16

Begolka, W. S., Haynes, L. M., Olson, J. K., Padilla, J., Neville, K. L., Dal Canto, M., Palma, J., Kim, B. S., and Miller, S. D. (2001). CD8-deWcient SJL mice display enhanced susceptibility to Theiler’s virus infection and increased demyelinating pathology. J. Neurovirol. 7, 409–420. Bergmann, C. C., Altman, J. D., Hinton, D., and Stohlman, S. A. (1999). Inverted immunodominance and impaired cytolytic function of CD8þ T cells during viral persistence in the central nervous system. J. Immunol. 163, 3379–3387. Bihl, F., Brahic, M., and Bureau, J. F. (1999). Two loci, Tmevp2 and Tmevp3, located on the telomeric region of chromosome 10, control the persistence of Theiler’s virus in the central nervous system of mice. Genetics 152, 385–392. Bihl, F., Pena-Rossi, C., Guenet, J. L., Brahic, M., and Bureau, J. F. (1997). The shiverer mutation aVects the persistence of Theiler’s virus in the central nervous system. J. Virol. 71, 5025–5030. Borson, N. D., Paul, C., Lin, X., Nevala, W. K., Strausbauch, M. A., Rodriguez, M., and Wettstein, P. J. (1997). Brain-inWltrating cytolytic T lymphocytes speciWc for Theiler’s virus recognize H-2Db molecules complexed with a viral VP2 peptide lacking a consensus anchor residue. J. Virol. 71, 5244–5250. Brahic, M., Stroop, W. G., and Baringer, J. R. (1981). Theiler’s virus persists in glial cells during demyelinating disease. Cell 26, 123–128. Bureau, J. F., Drescher, K. M., Pease, L. R., Vikoren, T., Delcroix, M., Zoecklein, L., Brahic, M., and Rodriguez, M. (1998). Chromosome 14 contains determinants that regulate susceptibility to Theiler’s virus-induced demyelination in the mouse. Genetics 148, 1941–1949. Bureau, J. F., Montagutelli, X., Bihl, F., Lefebvre, S., Guenet, J. L., and Brahic, M. (1993). Mapping loci inXuencing the persistence of Theiler’s virus in the murine central nervous system. Nat. Genet. 5, 87–91. Cardin, A. D., and Weintraub, H. J. (1989). Molecular modelling of protein-glycosaminoglycan interactions. Arteriosclerosis 9, 21–32. Cheever, F. S., Daniels, J. B., Pappenheimer, A. M., and Bailey, O. T. (1948). A murine virus (JHM) causing disseminated encephalomyelitis with extensive destruction of myelin. J. Exp. Med. 90, 181–194. Chen, H. H., Kong, W. P., Zhang, L., Ward, P. L., and Roos, R. P. (1995). A picornaviral protein synthesized out of frame with the polyprotein plays a key role in a virus-induced immune-mediated demyelinating disease. Nat. Med. 1, 927–931. Clatch, R. J., Lipton, H. L., and Miller, S. D. (1986). Characterization of Theiler’s murine encephalomyelitis virus (TMEV)-speciWc delayed-type hypersensitivity responses in TMEV-induced demyelinating disease: Correlation with clinical signs. J. Immunol. 136, 920–927. Clatch, R. J., Melvold, R. W., Miller, S. D., and Lipton, H. L. (1985). Theiler’s murine encephalomyelitis virus (TMEV)-induced demyelinating disease in mice is inXuenced by the H-2D region: Correlation with TEMVspeciWc delayed-type hypersensitivity. J Immunol. 135, 1408–1414. Clatch, R. J., Miller, S. D., Metzner, R., Dal Canto, M. C., and Lipton, H. L. (1990). Monocytes/macrophages isolated from the mouse central nervous system contain infectious Theiler’s murine encephalomyelitis virus (TMEV). Virology 176, 244–254. Dal Canto, M. C., and Lipton, H. L. (1975). Primary demyelination in Theiler’s virus infection. An ultrastructural study. Lab Invest. 33, 626–637. Dandekar, A. A., Wu, G. F., Pewe, L., and Perlman, S. (2001). Axonal damage is T cell mediated and occurs concomitantly with demyelination in mice infected with a neurotropic coronavirus. J. Virol. 75, 6115–6120. Daniels, J. B., Pappenheimer, A. M., and Richardson, S. (1952). Observations on Encephalitis of Mice. J. Exp. Med. 96, 517. Dethlefs, S., Escriou, N., Brahic, M., van der Werf, S., and Larsson-Sciard, E. L. (1997). Theiler’s virus and Mengo virus induce cross-reactive cytotoxic T lymphocytes restricted to the same immunodominant VP2 epitope in C57BL/6 mice. J. Virol. 71, 5361–5365. Drescher, K. M., Rivera-Quinones, C., Lucchinetti, C. F., and Rodriguez, M. (1998). Failure of treatment with Linomide or oral myelin tolerization to ameliorate demyelination in a viral model of multiple sclerosis. J. Neuroimmunol. 88, 111–119. Erlich, S. S., Fleming, J. O., Stohlman, S. A., and Weiner, L. P. (1987). Experimental neuropathology of chronic demyelination induced by a JHM virus variant (DS). Arch. Neurol. 44, 839–842. Fiette, L., Aubert, C., Brahic, M., and Rossi, C. P. (1993). Theiler’s virus infection of beta 2-microglobulindeWcient mice. J. Virol. 67, 589–592. Fleming, J. O., Trousdale, M. D., el-Zaatari, F. A., Stohlman, S. A., and Weiner, L. P. (1986). Pathogenicity of antigenic variants of murine coronavirus JHM selected with monoclonal antibodies. J. Virol. 58, 869–875. Fotiadis, C., Kilpatrick, D. R., and Lipton, H. L. (1991). Comparison of the binding characteristics to BHK-21 cells of viruses representing the two Theiler’s virus neurovirulence groups. Virology 182, 365–70. Fu, J., Rodriguez, M., and Roos, R. P. (1990b). Strains from both Theiler’s virus subgroups encode a determinant for demyelination. J. Virol. 64, 6345–6348. Fu, J. L., Stein, S., Rosenstein, L., Bodwell, T., Routbort, M., Semler, B. L., and Roos, R. P. (1990a). Neurovirulence determinants of genetically engineered Theiler viruses. Proc. Natl. Acad. Sci. USA 87, 4125–4129. Ghadge, G. D., Ma, L., Sato, S., Kim, J., and Roos, R. P. (1998). A protein critical for a Theiler’s virus-induced immune system-mediated demyelinating disease has a cell type-speciWc antiapoptotic eVect and a key role in virus persistence. J. Virol. 72, 8605–8612. Gerety, S. J., Rundell, M. K., Dal Canto, M. C., and Miller, S. D. (1994). Class II-restricted T cell responses in Theiler’s murine encephalomyelitis virus-induced demyelinating disease. VI. Potentiation of demyelination with and characterization of an immunopathologic CD4þ T cell line speciWc for an immunodominant VP2 epitope. J. Immunol. 152, 919–929.

1095

1096

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

Grant, R. A., Filman, D. J., Fujinami, R. S., Icenogle, J. P., and Hogle, J. M. (1992). Three-dimensional structure of Theiler virus. Proc. Natl. Acad. Sci. USA 89, 2061–2065. Graves, M. C., Bologa, L., Siegel, L., and Londe, H. (1986). Theiler’s virus in brain cell cultures: Lysis of neurons and oligodendrocytes and persistence in astrocytes and macrophages. J. Neurosci. Res. 15, 491–501. Haspel, M. V., Lampert, P. W., and Oldstone, M. B. (1978). Temperature-sensitive mutants of mouse hepatitis virus produce a high incidence of demyelination. Proc. Natl. Acad. Sci. USA 75, 4033–6. Hertzler, S., Luo, M., and Lipton, H. L. (2000). Mutation of predicted virion pit residues alters binding of Theiler’s murine encephalomyelitis virus to BHK-21 cells. J. Virol. 74, 1994–2004. Houtman, J. J., and Fleming, J. O. (1996). Dissociation of demyelination and viral clearance in congenitally immunodeWcient mice infected with murine coronavirus JHM. J. Neurovirol. 2, 101–110. Jakob, J., and Roos, R. P. (1996). Molecular determinants of Theiler’s murine encephalomyelitis-induced disease. J. Neurovirol. 2, 70–77. Jelachich, M. L., Bandyopadhyay, P., Blum, K., and Lipton, H. L. (1995). Theiler’s virus growth in murine macrophage cell lines depends on the state of diVerentiation. Virology 209, 437–444. Jnaoui, K., Minet, M., and Michiels, T. (2002). Mutations that aVect the tropism of DA and GDVII strains of Theiler’s virus in vitro inXuence sialic acid binding and pathogenicity. J. Virol. 76, 8138–8147. Johnson, A. J., Njenga, M. K., Hansen, M. J., Kuhns, S. T., Chen, L., Rodriguez, M., and Pease, L. R. (1999). Prevalent class I-restricted T-cell response to the Theiler’s virus epitope Db:VP2121–130 in the absence of endogenous CD4 help, tumor necrosis factor alpha, gamma interferon, perforin, or costimulation through CD28. J. Virol. 73, 3702–3708. Johnson, A. J., Upshaw, J., Pavelko, K. D., Rodriguez, M., and Pease, L. R. (2001). Preservation of motor function by inhibition of CD8þ virus peptide-speciWc T cells in Theiler’s virus infection. FASEB J. 15, 2760–2762. Kappel, C. A., Melvold, R. W., and Kim, B. S. (1990). InXuence of sex on susceptibility in the Theiler’s murine encephalomyelitis virus model for multiple sclerosis. J. Neuroimmunol. 29, 15–19. Karpus, W. J., Pope, J. G., Peterson, J. D., Dal Canto, M. C., and Miller, S. D. (1995). Inhibition of Theiler’s virus-mediated demyelination by peripheral immune tolerance induction. J. Immunol. 155, 947–957. Lane, T. E., Liu, M. T., Chen, B. P., Asensio, V. C., Samawi, R. M., Paoletti, A. D., Campbell, I. L., Kunkel, S. L., Fox, H. S., and Buchmeier, M. J. (2000). A central role for CD4(þ) T cells and RANTES in virus-induced central nervous system inXammation and demyelination. J. Virol. 74, 1415–1424. Lavi, E., Gilden, D. H., Highkin, M. K., and Weiss, S. R. (1984). Persistence of mouse hepatitis virus A59 RNA in a slow virus demyelinating infection in mice as detected by in situ hybridization. J. Virol. 51, 563–566. Lehmann, P. V., Sercarz, E. E., Forsthuber, T., Dayan, C. M., Gammon, G. (1993). Determinant spreading and the dynamics of the autoimmune T-cell repertoire. Immunol. Today 14, 203–208. Lehrich, J. R., Arnason, B. G., and Hochberg, F. H. (1976). Demyelinative myelopathy in mice induced by the DA virus. J. Neurol. Sci. 29, 149–160. Levy, M., Aubert, C., and Brahic, M. (1992). Theiler’s virus replication in brain macrophages cultured in vitro. J. Virol. 66, 3188–3193. Lin, M. T., Hinton, D. R., Marten, N. W., Bergmann, C. C., and Stohlman, S. A. (1999). Antibody prevents virus reactivation within the central nervous system. J Immunol. 162, 7358–7368. Lin, M. T., Stohlman, S. A., and Hinton, D. R. (1997b). Mouse hepatitis virus is cleared from the central nervous systems of mice lacking perforin-mediated cytolysis. J. Virol. 71, 383–391. Lin, X., Pease, L. R., and Rodriguez, M. (1997a). DiVerential generation of class I H-2D-versus H-2K-restricted cytotoxicity against a demyelinating virus following central nervous system infection. Eur. J. Immunol. 27, 963–970. Lin, X., Sato, S., Patick, A. K., Pease, L. R., Roos, R. P., and Rodriguez, M. (1998). Molecular characterization of a nondemyelinating variant of Daniel’s strain of Theiler’s virus isolated from a persistently infected glioma cell line. J. Virol. 72, 1262–1269. Lindsley, M. D., Patick, A. K., Prayoonwiwat, N., and Rodriguez, M. (1992). Coexpression of class I major histocompatibility antigen and viral RNA in central nervous system of mice infected with Theiler’s virus: A model for multiple sclerosis. Mayo Clin. Proc. 67, 829–838. Lindsley, M. D., and Rodriguez, M. (1989). Characterization of the inXammatory response in the central nervous system of mice susceptible or resistant to demyelination by Theiler’s virus. J. Immunol. 142, 2677–2682. Lipton, H. L. (1975). Theiler’s virus infection in mice: An unusual biphasic disease process leading to demyelination. Infect. Immun. 11, 1147–1155. Lipton, H. L. (1978). Characterization of the TO strains of Theiler’s mouse encephalomyelitis viruses. Infect. Immun. 20, 869–872. Lipton, H. L., CalenoV, M., Bandyopadhyay, P., Miller, S. D., Dal Canto, M. C., Gerety, S., and Jensen, K. (1991). The 5’ noncoding sequences from a less virulent Theiler’s virus dramatically attenuate GDVII neurovirulence. J. Virol. 65, 4370–4377. Lipton, H. L., and Dal Canto, M. C. (1976a). Chronic neurologic disease in Theiler’s virus infection of SJL/J mice. J. Neurol. Sci. 30, 201–207. Lipton, H. L., and Dal Canto, M. C. (1976b). Theiler’s virus-induced demyelination: Prevention by immunosuppression. Science 192, 62–64. Lipton, H. L., and Dal Canto, M. C. (1979). Susceptibility of inbred mice to chronic central nervous system infection by Theiler’s murine encephalomyelitis virus. Infect. Immun. 26, 369–374.

au17

1097

MOUSE HEPATITIS VIRUS (MHV)

Lipton, H. L., and Melvold, R. (1984). Genetic analysis of susceptibility to Theiler’s virus-induced demyelinating disease in mice. J. Immunol. 132, 1821–1825. Lipton, H. L., Pritchard, A. E., and CalenoV, M. A. (1998). Attenuation of neurovirulence of Theiler’s murine encephalomyelitis virus strain GDVII is not suYcient to establish persistence in the central nervous system. J. Gen. Virol. 79, 1001–1004. Lipton, H. L., Twaddle, G., and Jelachich, M. L. (1995). The predominant virus antigen burden is present in macrophages in Theiler’s murine encephalomyelitis virus-induced demyelinating disease. J. Virol. 69, 2525–2533. Lorch, Y., Friedmann, A., Lipton, H. L., and Kotler, M. (1981). Theiler’s murine encephalomyelitis virus group includes two distinct genetic subgroups that diVer pathologically and biologically. J. Virol. 40, 560–567. Lucchinetti, C. F., and Rodriguez, M. (1997). The controversy surrounding the pathogenesis of the multiple sclerosis lesion. Mayo Clin. Proc. 72, 665–678. Ludwin, S. K., and Johnson, E. S. (1981). Evidence for a ‘‘dying-back’’ gliopathy in demyelinating disease. Ann. Neurol. 9, 301–305. Luo, M., He, C., Toth, K. S., Zhang, C. X., and Lipton, H. L. (1992). Three-dimensional structure of Theiler murine encephalomyelitis virus (BeAn strain). Proc. Natl. Acad. Sci. USA 89, 2409–2413. Luo, M., Toth, K. S., Zhou, L., Pritchard, A., and Lipton, H. L. (1996). The structure of a highly virulent Theiler’s murine encephalomyelitis virus (GDVII) and implications for determinants of viral persistence. Virology 220, 246–250. Matthews, A. E., Lavi, E., Weiss, S. R., and Paterson, Y. (2002). Neither B cells nor T cells are required for CNS demyelination in mice persistently infected with MHV-A59. J. Neurovirol. 8, 257–264. McGavern, D. B., Murray, P. D., Rivera-Quinones, C., Schmelzer, J. D., Low, P. A., and Rodriguez, M. (2000). Axonal loss results in spinal cord atrophy, electrophysiological abnormalities and neurological deWcits following demyelination in a chronic inXammatory model of multiple sclerosis. Brain 123, 519–531. McGavern, D. B., Zoecklein, L., Drescher, K. M., and Rodriguez, M. (1999). Quantitative assessment of neurologic deWcits in a chronic progressive murine model of CNS demyelination. Exp. Neurol. 158, 171–181. McRae, B. L., Vanderlugt, C. L., Dal Canto, M. C., and Miller S. D. (1995). Functional evidence for epitope spreading in the relapsing pathology of experimental autoimmune encephalomyelitis. J. Exp. Med. 182, 75–85. Melvold, R. W., Jokinen, D. M., Miller, S. D., Dal Canto, M. C., and Lipton, H. L. (1990). IdentiWcation of a locus on mouse chromosome 3 involved in diVerential susceptibility to Theiler’s murine encephalomyelitis virus-induced demyelinating disease. J. Virol. 64, 686–690. Miller, D. J., Asakura, K., and Rodriguez, M. (1996). Central nervous system remyelination clinical application of basic neuroscience principles. Brain Pathol. 6, 331–344. Miller, D. J., Rivera-Quinones, C., Njenga, M. K., Leibowitz, J., and Rodriguez, M. (1995). Spontaneous CNS remyelination in beta 2 microglobulin-deWcient mice following virus-induced demyelination. J. Neurosci. 15, 8345–8352. Miller, S. D., Gerety, S. J., Kennedy, M. K., Peterson, J. D., Trotter, J. L., Tuohy, V. K., Waltenbaugh, C., Dal Canto, M. C., and Lipton, H. L. (1990). Class II-restricted T cell responses in Theiler’s murine encephalomyelitis virus (TMEV)-induced demyelinating disease. III. Failure of neuroantigen-speciWc immune tolerance to aVect the clinical course of demyelination. J. Neuroimmunol. 26, 9–23. Miller, S. D., Vanderlugt, C. L., Begolka, W. S., Pao, W., Yauch, R. L., Neville, K. L., Katz-Levy, Y., Carrizosa, A., and Kim, B. S. (1997). Persistent infection with Theiler’s virus leads to CNS autoimmunity via epitope spreading. Nat. Med. 3, 1133–1136. Murray, P. D., McGavern, D. B., Sathornsumetee, S., and Rodriguez, M. (2001). Spontaneous remyelination following extensive demyelination is associated with improved neurological function in a viral model of multiple sclerosis. Brain 124, 1403–1416. Murray, P. D., Pavelko, K. D., Leibowitz, J., Lin, X., and Rodriguez, M. (1998). CD4(þ) and CD8(þ) T cells make discrete contributions to demyelination and neurologic disease in a viral model of multiple sclerosis. J. Virol. 72, 7320–7329. Neville, K. L., Padilla, J., and Miller, S. D. (2002). Myelin-speciWc tolerance attenuates the progression of a virusinduced demyelinating disease: Implications for the treatment of MS. J. Neuroimmunol. 123, 18–29. Nitayaphan, S., Omilianowski, D., Toth, M. M., Parks, G. D., Rueckert, R. R., Palmenberg, A. C., and Roos, R. P. (1986). Relationship of Theiler’s murine encephalomyelitis viruses to the cardiovirus genus of picornaviruses. Intervirology 26, 140–148. Njenga, M. K., Asakura, K., Hunter, S. F., Wettstein, P., Pease, L. R., and Rodriguez, M. (1997). The immune system preferentially clears Theiler’s virus from the gray matter of the central nervous system. J. Virol. 71, 8592–8601. Njenga, M. K., Coenen, M. J., DeCuir, N., Yeh, H. Y., and Rodriguez, M. (2000). Short-term treatment with interferon-alpha/beta promotes remyelination, whereas long-term treatment aggravates demyelination in a murine model of multiple sclerosis. J. Neurosci. Res. 59, 661–670. Njenga, M. K., Murray, P. D., McGavern, D., Lin, X., Drescher, K. M., and Rodriguez, M. (1999). Absence of spontaneous central nervous system remyelination in class II-deWcient mice infected with Theiler’s virus. J. Neuropathol. Exp. Neurol. 58, 78–91. Njenga, M. K., Pavelko, K. D., Baisch, J., Lin, X., David, C., Leibowitz, J., and Rodriguez, M. (1996). Theiler’s virus persistence and demyelination in major histocompatibility complex class II-deWcient mice. J. Virol. 70, 1729–1737.

au18 au19

1098

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

au20 au21

au22

Noseworthy, J. H. (1998). Multiple sclerosis clinical trials: Old and new challenges. Semin. Neurol. 18, 377–388. Noseworthy, J. H., Lucchinetti, C., Rodriguez, M., and Weinshenker, B. G. (2000). Multiple sclerosis. N. Engl. J. Med. 343, 938–952. Ohara, Y., Konno, H., Iwasaki, Y., Yamamoto, T., Terunuma, H., and Suzuki, H. (1990). Cytotropism of Theiler’s murine encephalomyelitis viruses in oligodendrocyte-enriched cultures. Arch. Virol. 114, 293–298. Ohara, Y., Stein, S., Fu, J. L., Stillman, L., Klaman, L., and Roos, R. P. (1988). Molecular cloning and sequence determination of the DA strain of Theiler’s murine encephalomyelitis viruses. Virology 164, 245–255. Pappenheimer, A. M. (1958). Pathology of infection with the JHM virus. J. Natl. Cancer Inst. 20, 879–891. Parra, B., Hinton, D. R., Marten, N. W., Bergmann, C. C., Lin, M. T., Yang, C. S., and Stohlman, S. A. (1999). IFN-gamma is required for viral clearance from central nervous system oligodendroglia. J. Immunol. 162, 1641–1647. Parra, B., Lin, M. T., Stohlman, S. A., Bergmann, C. C., Atkinson, R., and Hinton, D. R. (2000). Contributions of Fas-Fas ligand interactions to the pathogenesis of mouse hepatitis virus in the central nervous system. J. Virol. 74, 2447–2450. Patick, A. K., Lindsley, M. D., and Rodriguez, M. (1990). DiVerential pathogenesis between mouse strains resistant and susceptible to Theiler’s virus-induced demyelination. Sem. Virol. 1, 281–288. Patick, A. K., Pease, L. R., David, C. S., and Rodriguez, M. (1990). Major histocompatibility complex-conferred resistance to Theiler’s virus-induced demyelinating disease is inherited as a dominant trait in B10 congenic mice. J. Virol. 64, 5570–5576. Pevear, D. C., Borkowski, J., CalenoV, M., Oh, C. K., Ostrowski, B., and Lipton, H. L. (1988). Insights into Theiler’s virus neurovirulence based on a genomic comparison of the neurovirulent GDVII and less virulent BeAn strains. Virology 165, 1–12. Pevear, D. C., CalenoV, M., Rozhon, E., and Lipton, H. L. (1987). Analysis of the complete nucleotide sequence of the picornavirus Theiler’s murine encephalomyelitis virus indicates that it is closely related to cardioviruses. J. Virol. 61, 1507–1516. Pevear, D. C., Luo, M., and Lipton, H. L. (1988). Three-dimensional model of the capsid proteins of two biologically diVerent Theiler virus strains: Clustering of amino acid diVerence identiWes possible locations of immunogenic sites on the virion. Proc. Natl. Acad. Sci. USA 85, 4496–4500. Powell, H. C., and Lampert, P. W. (1975). Oligodendrocytes and their myelin-plasma membrane connections in JHM mouse hepatitis virus encephalomyelitis. Lab Invest. 33(4):440–445. Pullen, L. C., Miller, S. D., Dal Canto, M. C., and Kim, B. S. (1993). Class I-deWcient resistant mice intracerebrally inoculated with Theiler’s virus show an increased T cell response to viral antigens and susceptibility to demyelination. Eur. J. Immunol. 23, 2287–2293. Reddi, H. V., and Lipton, H. L. (2002). Heparan sulfate mediates infection of high-neurovirulence Theiler’s viruses. J. Virol. 76, 8400–8407. Rivera-Quinones, C., McGavern, D., Schmelzer, J. D., Hunter, S. F., Low, P. A., and Rodriguez, M. (1998). Absence of neurological deWcits following extensive demyelination in a class I-deWcient murine model of multiple sclerosis. Nat. Med. 4, 187–193. Rodriguez, M. (1985). Virus-induced demyelination in mice: ‘‘dying back’’ of oligodendrocytes. Mayo Clin. Proc. 60, 433–438. Rodriguez, M., and David, C. S. (1985). Demyelination induced by Theiler’s virus: InXuence of the H-2 haplotype. J. Immunol. 135, 2145–2148. Rodriguez, M., Dunkel, A. J., Thiemann, R. L., Leibowitz, J., Zijlstra, M., and Jaenisch, R. (1993). Abrogation of resistance to Theiler’s virus-induced demyelination in H-2b mice deWcient in beta 2-microglobulin. J. Immunol. 151, 266–276. Rodriguez, M., Lafuse, W. P., Leibowitz, J., and David, C. S. (1986a). Partial suppression of Theiler’s virusinduced demyelination in vivo by administration of monoclonal antibodies to immune-response gene products (Ia antigens). Neurology 36, 964–970. Rodriguez, M., Leibowitz, J., and David, C. S. (1986b). Susceptibility to Theiler’s virus-induced demyelination. Mapping of the gene within the H-2D region. J. Exp. Med. 163, 620–631. Rodriguez, M., Leibowitz, J. L., and Lampert, P. W. (1983). Persistent infection of oligodendrocytes in Theiler’s virus-induced encephalomyelitis. Ann. Neurol. 13, 426–433. Rodriguez, M., Nabozny, G. H., Thiemann, R. L., and David, C. S. (1994). InXuence of deletion of T cell receptor V beta genes on the Theiler’s virus model of multiple sclerosis. Autoimmunity 19, 221–230. Rodriguez M, Oleszak E., and Leibowitz J. (1987). Theiler’s murine encephalomyelitis: A model of demyelination and persistence of virus. Crit Rev Immunol. 7, 325–365. Rodriguez, M., Patick, A. K., Pease, L. R., and David, C. S. (1992). Role of T cell receptor V beta genes in Theiler’s virus-induced demyelination of mice. J. Immunol. 148, 921–927. Rodriguez, M., Pavelko, K. D., Njenga, M. K., Logan, W. C., and Wettstein, P. J. (1996). The balance between persistent virus infection and immune cells determines demyelination. J. Immunol. 157, 5699–5709. Rodriguez, M., and Roos, R. P. (1992). Pathogenesis of early and late disease in mice infected with Theiler’s virus, using intratypic recombinant GDVII/DA viruses. J. Virol. 66, 217–225. Rodriguez, M., Siegel, L. M., Hovanec-Burns, D., Bologa, L., and Graves, M. C. (1988). Theiler’s virusassociated antigens on the surfaces of cultured glial cells. Virology 166, 463–474. Roos, R. P., and Wollmann, R. (1984). DA strain of Theiler’s murine encephalomyelitis virus induces demyelination in nude mice. Ann. Neurol. 15, 494–499.

1099

MOUSE HEPATITIS VIRUS (MHV)

Rosenthal, A., Fujinami, R. S., and Lampert, P. W. (1986). Mechanism of Theiler’s virus-induced demyelination in nude mice. Lab Invest. 54, 515–522. Rossi, C. P., Delcroix, M., Huitinga, I., McAllister, A., van Rooijen, N., Claassen, E., and Brahic, M. (1997). Role of macrophages during Theiler’s virus infection. J. Virol. 71, 3336–3340. Rouse, B. T., and Deshpande, S. (2002). Viruses and autoimmunity: An aVair but not a marriage contract. Rev. Med. Virol. 12, 107–113. Rueckert, R. R. (1996). Picornaviridae: The Viruses and Their Replication. In ‘‘Fields Virology’’ (Fields, Knipe, Howley, Chanock, Melnick, Monath, Roizman, and Strauss, eds.), 3rd ed., pp. 609–654. Lippincott-Raven, Philadelphia. Sapatino, B. V., Petrescu, A. D., Rosenbaum, B. A., Smith, R., 3rd, Piedrahita, J. A., and Welsh, C. J. Characteristics of cloned cerebrovascular endothelial cells following infection with Theiler’s virus. II. Persistent infection. J. Neuroimmunol. 62, 127–135. Sarchielli, P., Trequattrini, A., Usai, F., Murasecco, D., and Gallai, V. (1993). Role of viruses in the etiopathogenesis of multiple sclerosis. Acta Neurol. (Napoli) 15, 363–381. Shaw-Jackson, C., and Michiels, T. (1997). Infection of macrophages by Theiler’s murine encephalomyelitis virus is highly dependent on their activation or diVerentiation state. J. Virol. 71, 8864–8867. Sibley, W. A., Bamford, C. R., and Clark, K. (1985). Clinical viral infections and multiple sclerosis. Lancet 1, 1313–1315. Spencer, P. S., and Schaumburg, H. H. (1976). Central-peripheral distal axonopathy: The pathology of dyingback polyneuropathies. Prog. Neuropathol. 3, 253–295. Stohlman, S. A., and Weiner, L. P. (1981). Chronic central nervous system demyelination in mice after JHM virus infection. Neurology 31, 38–44. Stohlman, S. A., Bergmann, C. C., Lin, M. T., Cua, D. J., and Hinton, D. R. (1998). CTL eVector function within the central nervous system requires CD4þ T cells. J. Immunol. 160, 2896–2904. Stohlman, S. A., Bergmann, C. C., van der Veen, R. C., and Hinton, D. R. (1995). Mouse hepatitis virus-speciWc cytotoxic T lymphocytes protect from lethal infection without eliminating virus from the central nervous system. J. Virol. 69, 684–694. Sun, N., Grzybicki, D., Castro, R. F., Murphy, S., and Perlman, S. (1995). Activation of astrocytes in the spinal cord of mice chronically infected with a neurotropic coronavirus. Virology 213, 482–493. Takata, H., Obuchi, M., Yamamoto, J., Odagiri, T., Roos, R. P., Iizuka, H., and Ohara, Y. (1998). L* protein of the DA strain of Theiler’s murine encephalomyelitis virus is important for virus growth in a murine macrophage-like cell line. J. Virol. 72, 4950–4955. Toth, K. S., Zhang, C. X., Lipton, H. L., and Luo, M. (1993). Crystallization and preliminary X-ray diVraction studies of Theiler’s virus (GDVII strain). J. Mol. Biol. 231, 1126–1129. Trapp, B. D., Peterson, J., RansohoV, R. M., Rudick, R., Mork, S., and Bo, L. (1998). Axonal transection in the lesions of multiple sclerosis. N. Engl. J. Med. 338, 278–285. Trottier, M., Kallio, P., Wang, W., and Lipton, H. L. (2001). High numbers of viral RNA copies in the central nervous system of mice during persistent infection with Theiler’s virus. J. Virol. 75, 7420–7428. Ure, D. R., and Rodriguez, M. (2002). Polyreactive antibodies to glatiramer acetate promote myelin repair in murine model of demyelinating disease. FASEB J. 16, 1260–1262. Warrington, A. E., Asakura, K., Bieber, A. J., Ciric, B., Van Keulen, V., Kaveri, S. V., Kyle, R. A., Pease, L. R., and Rodriguez, M. (2000). Human monoclonal antibodies reactive to oligodendrocytes promote remyelination in a model of multiple sclerosis. Proc. Natl. Acad. Sci. USA 97, 6820–6825. Watanabe, R., Wege, H., and ter Meulen, V. (1983). Adoptive transfer of EAE-like lesions from rats with coronavirus-induced demyelinating encephalomyelitis. Nature 305, 150–153. Watanabe, R., Wege, H., and ter Meulen, V. (1987). Comparative analysis of coronavirus JHM-induced demyelinating encephalomyelitis in Lewis and Brown Norway rats. Lab Invest. 57, 375–384. Webster, R. G., and GranoV, A. (Eds.). (1995). ‘‘Encyclopedia of Virology.’’ Academic Press, San Diego, CA. Welsh, C. J., Tonks, P., Nash, A. A., and Blakemore, W. F. (1987). The eVect of L3T4 T cell depletion on the pathogenesis of Theiler’s murine encephalomyelitis virus infection in CBA mice. J. Gen. Virol. 68, 1659–1667. Wijburg, O. L., Heemskerk, M. H., Boog, C. J., and Van Rooijen, N. (1997). Role of spleen macrophages in innate and acquired immune responses against mouse hepatitis virus strain A59. Immunology 92, 252–258. Williams, R. K., Jiang, G. S., and Holmes, K. V. (1991). Receptor for mouse hepatitis virus is a member of the carcinoembryonic antigen family of glycoproteins. Proc. Natl. Acad. Sci. USA 88, 5533–5536. Williamson, J. S., and Stohlman, S. A. (1990). EVective clearance of mouse hepatitis virus from the central nervous system requires both CD4þ and CD8þ T cells. J. Virol. 64, 4589–4592. Williamson, J. S., Sykes, K. C., and Stohlman, S. A. (1991). Characterization of brain-inWltrating mononuclear cells during infection with mouse hepatitis virus strain JHM. J. Neuroimmunol. 32, 199–207. Wroblewska, Z., Gilden, D. H., Wellish, M., Rorke, L. B., Warren, K. G., and Wolinsky, J. S. (1977). Virusspecific intracytoplasmic inclusions in mouse brain produced by a newly isolated strain of Theiler virus. I. Virologic and morphologic studies. Lab. Invest. 37, 395–602. Wu, G. F., and Perlman, S. (1999). Macrophage inWltration, but not apoptosis, is correlated with immunemediated demyelination following murine infection with a neurotropic coronavirus. J. Virol. 73, 8771–8780. Xue, S., Sun, N., Van Rooijen, N., and Perlman, S. (1999). Depletion of blood-borne macrophages does not reduce demyelination in mice infected with a neurotropic coronavirus. J. Virol. 73, 6327–6334. Yuki, N. (2001). Infectious origins of, and molecular mimicry in, Guillain-Barre and Fisher syndromes. Lancet Infect. Dis. 1, 29–37.

au23

au24

1100

44. EXPERIMENTAL MODELS OF VIRUS-INDUCED DEMYELINATION

Zhang, L., Senkowski, A., Shim, B., and Roos, R. P. (1993). Chimeric cDNA studies of Theiler’s murine encephalomyelitis virus neurovirulence. J. Virol. 67, 4404–4408. Zheng, L., CalenoV, M. A., and Dal Canto, M. C. (2001). Astrocytes, not microglia, are the main cells responsible for viral persistence in Theiler’s murine encephalomyelitis virus infection leading to demyelination. J. Neuroimmunol. 118, 256–267. Zhou, L., Lin, X., Green, T. J., Lipton, H. L., and Luo, M. (1997). Role of sialyloligosaccharide binding in Theiler’s virus persistence. J. Virol. 71, 9701–9712. Zhou, L., Luo, Y., Wu, Y., Tsao, J., and Luo, M. (2000). Sialylation of the host receptor may modulate entry of demyelinating persistent Theiler’s virus. J. Virol. 74, 1477–1485.

C H A P T E R

45 Models of Krabbe Disease Kunihiko Suzuki

INTRODUCTION The classical form of human Krabbe disease (globoid cell leukodystrophy, GLD) is caused by genetic deWciency of galactosylceramidase activity (Wenger et al., 2001; also see Chapter 35, in this book). The genetic leukodystrophy in West Highland and Cairn terrier dogs had been described earlier as globoid cell leukodystrophy on the basis of the characteristic neuropathology (Fankha¨user et al., 1963). As soon as the genetic enzymatic defect in the human disease was identiWed (Suzuki and Suzuki, 1970), the dog GLD was recognized as genetically equivalent to human disease (Suzuki et al., 1970). Several years later, a spontaneous mouse mutant was discovered at the Jackson Laboratory that showed characteristic neuropathology of GLD and was named twitcher (Duchen et al., 1980). Its enzymatic basis was also identiWed as genetic galactosylceramidase deWciency (Kobayashi et al., 1980). Bridging the gap between the human and other mammalian species is a mutant that occurs in the Rhesus monkey, discovered much later (Baskin et al., 1998). Galactosylceramidase cDNA was cloned, and disease-causing mutations have been identiWed in the mouse (Sakai et al., 1996), West Highland and Cairn terriers (Victoria et al., 1996), and the Rhesus monkey (Luzi et al., 1997). These models have been extensively utilized for studies of the pathogenesis and treatment of GLD. They provide varying advantages and disadvantages as animal models because of their diVerent sizes and life spans. From time to time, globoid cell leukodystrophy identiWed by its unique neuropathology has also been described in other mammalian species and some other strain of dogs, but these conclusions were generally sporadic and are unavailable for follow-up studies. Thus, utility of these sporadic models, even if they might be enzymatically authentic, is limited. In addition to the naturally occurring animal models, Luzi et al. recently generated an experimental mouse model in which a mutation, H168C, was introduced by homologous recombination (Luzi et al., 2001). A genetically distinct form of globoid cell leukodystrophy has also been experimentally generated in the mouse by inactivating one of the sphingolipid activator proteins, saposin A (Matsuda et al., 2001b). This resulted in a late-onset, chronic form of GLD. All aspects of its clinical, pathological, and biochemical phenotype were consistent with impaired degradation of galactosylceramidase substrates. This model established that saposin A is indispensable for normal physiological function of galactosylceramidase in vivo. This chapter emphasizes the mouse models and refers to other models from time to time as appropriate. For more details on models in other species, refer to (Wenger et al., 2001).

Myelin Biology and Disorders, Volume 2

1101

Copyright 2004, Elsevier (USA). All rights reserved.

au1 au2

au3

1102

45. MODELS OF KRABBE DISEASE

MOUSE MODELS Twitcher Mutant The twitcher mutant was originally discovered as an autosomal recessive myelin mutant in the Jackson Laboratory (Bar Harbor, Maine) in l976. Since the Wrst clinico-pathological documentation (Duchen et al., 1980) and the identiWcation of the genetic defect in a lysosomal hydrolase, galactosylceramidase (Kobayashi et al., 1980), it has been widely used as a convenient and useful experimental tool for investigations of the pathogenesis of and therapeutic approaches to GLD. Despite the large number of known neurological mouse mutants, the twitcher remains the only naturally occurring genetically authentic mouse model of human sphingolipidoses. Although the lack of authentic models of human sphingolipidoses among small laboratory animals is no longer a problem with the advent of the homologous recombination and transgenic technology, which allows experimental generation of basically any mouse models once the gene is isolated, the twitcher mutant remains the most thoroughly characterized and most extensively utilized because of its more than 20-year history. Successful cloning of the galactosylceramidase gene (Chen et al., 1993; Sakai et al., 1994) has led to identiWcation of the mutation in the twitcher mutant (Sakai et al., 1996). Clinical Phenotype Clinical manifestations are exclusively neurological, and their onset coincides with the onset of active myelination. This is understandable because the primary substrate of the defective enzyme, galalctosylceramide, is a nearly exclusive constituent of the myelin sheath and is not even synthesized until myelin formation begins. Thus, homozygous twitcher mice are indistinguishable at birth from their unaVected littermates. AVected mice stop gaining weight and become generally less active than their littermates by 15 to 20 days. Generalized tremulousness sets in before 20 days and progressive muscle weakness and eventual paralysis develop, particularly conspicuous in the hind limbs and the neck muscles. At the terminal stage around 40 to 45 days, their body weight is about one-third of normal mice (Igisu et al., 1983a). Motor functions generally deteriorate rapidly after 20 days. Genetic background appears to inXuence the course of the disease signiWcantly. Currently available twitcher mice are on the genetic background of C57/BL, and their life span rarely extends beyond 45 days. However, Duchen et al. observed in their Wrst description of the mutant on a diVerent genetic background that its life span was ‘‘up to 3 months,’’ much longer than twitcher mice on the C57/BL background (Duchen et al., 1980). As in most autosomal recessive disorders, heterozygous carriers are essentially normal. However, interestingly, subtle developmental delays in neurological and neurobehavioral parameters have been described during early development in male but not in female heterozygous carriers, although they eventually caught up with wild-type littermates (Olmstead 1987). Pathology Neuropathology is generally well correlated with progression of the clinical manifestations and is also almost entirely limited to the nervous system. Prior to clinical onset around 15 days, no neuropathological changes are apparent on the light microscopic level, although inclusions in oligodendrocytes may be detected on the ultrastructural level as early as 5 days. Macrophages inWltrate concomitantly with myelin degeneration into the spinal cord, cerebellar white matter, and brain stem after 20 days, then into the cerebral white matter after 25 days, and into cerebellar and cerebral gray matter after 30 days. Myelin degeneration begins at 10 to 20 days after commencement of myelination in any of the given nerve Wber tracts (Taniike and Suzuki, 1994). The brain of aVected mice is slightly smaller than normal at the terminal stage, but no other gross abnormalities are present. The characteristic pathology in the central nervous system is demyelination, degeneration of oligodendrocytes, reactive astrocytosis, and inWltration of macrophages that contain periodic acid-SchiV (PAS)-positive materials (‘‘globoid cells’’). The macrophages are often clustered

V. ANIMAL MODELS OF HUMAN DISEASE

MOUSE MODELS

around blood vessels. Crystalloid or slender tubular inclusions, typical in human GLD, are present most abundantly in the macrophages but also in the oligodendrocytes. These pathological features are essentially identical with those in human Krabbe disease. The cerebral cortex and other regions of gray matter are relatively unaVected pathologically, except for occasional perivascular or perineuronal microglial cells containing similar inclusions. GFAP-immunoreactive astrocytes are abundant throughout the central nervous system, including gray matter (Kobayashi et al., 1986; Mikoshiba et al., 1985). This increase of GFAP-immunoreactive cells is accompanied by an increase in GFAP mRNA, which is detectable as early as 10 days when demyelination is not yet obvious. Contrary to the central nervous system, peripheral nerves are grossly abnormal. They are abnormally thick, translucent, and Wrm and can easily be distinguished from normal nerves by visual inspection. There is segmental demyelination with evidence of remyelination associated with inWltration of macrophages containing the typical GLD inclusions and marked endoneurial edema that widely separates individual nerve Wbers (Powell et al., 1983). The Schwann cells of myelinated Wbers frequently contain the characteristic GLD inclusions, and the Schwann cells of unmyelinated Wbers are attenuated with unusually elongated or branching processes (Duchen et al., 1980; Kobayashi et al., 1988). GFAP immunoreactivity increases in Schwann cells as demyelination progresses (Kobayashi et al., 1986). Microglia/macrophages in both CNS and PNS react positively with anti-Mac-1 and F4/80 antibodies. Although the pathology is largely conWned to the nervous system, the kidney is an exception. Abnormal inclusions are frequently present in the renal epithelial cells. This renal pathology is distinctly diVerent from the human disease, in which the kidney is essentially normal. This diVerence also reXects a distinct diVerence in biochemical abnormalities in the kidney among the human, canine, and murine diseases (discussed later). Biochemistry The similarities of the clinical and pathological Wndings of the twitcher mouse to those in human Krabbe disease strongly suggested the same biochemical basis underlying GLD in the two species. A profound deWciency in galactosylceramidase activity in the brain and liver of the twitcher mice established it as an authentic mouse model of the human disease (Kobayashi et al., 1980). Clinically and pathologically normal heterozygous carrier mice showed intermediate activities. Thus, the genetic cause in the twitcher mouse is the same as that in human Krabbe disease (Suzuki and Suzuki, 1970). Enzymatic assays using tips of clipped tails of newborn mice allows identiWcation of the genetic status and thus makes experiments before the onset of clinical symptoms feasible (Kobayashi et al., 1982). These enzymatic diagnoses have been largely superceded by molecular diagnosis since the mutation underlying twitcher mouse was identiWed. Twitcher mouse shares the same unusual and apparently paradoxical analytical biochemistry with the human disease. Despite the genetic defect in the lysosomal hydrolase, galactosylceramidase, its primary physiological substrate, galactosylceramide, does not accumulate abnormally in the brain. The composition of galactolipids in the brain, spinal cord, and the sciatic nerve of twitcher mice is surprisingly normal during development (Igisu et al., 1983a). Galactosylceramide and sulfatide are signiWcantly decreased in later stages in twitcher brains, but the degree of decrease is far less than that in human GLD brains. In the twitcher spinal cord, the proportions of both galactosylceramide and sulfatide in the total lipids are normal at 16 and 25 days, but at 42 days, galactosylceramide is slightly less than the control (Igisu and Suzuki, 1984a). However, the total lipid was markedly decreased in the twitcher spinal cord at this age. Therefore, there is a much greater loss of galactosylceramide and sulfatide in the spinal cord than in the brain. In contrast to the brain, there is a massive accumulation of galactosylceramide in the kidney (Ida et al., 1982; Igisu and Suzuki, 1984a; Igisu et al., 1983b). Both HFA- and NFAgalactosylceramides are increased even at 16 days. They increase rapidly with progression of the disease. The parallel with the human disease breaks down here in that no increase in galactosylceramide was detected in the kidneys of human GLD patients (Suzuki, 1971).

V. ANIMAL MODELS OF HUMAN DISEASE

1103

1104

45. MODELS OF KRABBE DISEASE

au4

au5

The kidney in the dog model is intermediate between the human and murine disease in that there is only a slight increase in galactosylceramide associated with occasional inclusion bodies. In addition, moderate increases of galactosylceramide occur in the liver and lung of aVected mice (Igisu and Suzuki, 1984a). In contrast to galactosylceramide, another metabolically related substrate of the defective enzyme, galactosylsphingosine (psychosine), is consistently and progressively increased in CNS of twitcher mice up to 20 times normal (Igisu and Suzuki, 1984b). The degree of the psychosine accumulation generally correlates well with severity of pathology. Psychosine similarly accumulates also in human GLD brains (see Chapter 35 in this volume). Two other minor galactolipids are also natural substrates of galactosylceramidase: monogalactosyldiglyceride and the seminolipid precursor (1-alkyl, 2-acyl, galactosylglycerol). These compounds also accumulate moderately in tissues where they are normal constituents, the seminolipid precursor almost exclusively in the testis and monogalactosyldiglyceride in the brain and kidney (Matsuda et al., 2001b). Functional signiWcance of their accumulation is not clear. A synthetic enzyme, UDP-galactose:ceramide galactosyltransferase (CGT), catalyzes the last step of galactosylceramide synthesis. Its activity in the normal brain is known to change dramatically during CNS development in parallel to the period of active myelination (Costantino-Ceccarini and Morell, 1972). The activity is very low during the premyelination period and then increases sharply at the onset of myelination. It then declines nearly as sharply to a lower steady-state level. The activity of the enzyme in the twitcher spinal cord was normal during the early stage of myelination, but the activity decreased after 20 days (Kodama et al., 1982). At 25 and 33 days, galactosyltransferase activity was drastically reduced compared to controls. Recently this study has been corroborated by quantitative estimates of CGT mRNA (Taniike et al., 1998), made possible by recent cloning of the cDNA coding for rat CGT (Schulte and StoVel, 1993; Stahl et al., 1994). There was an early reduction of CGT mRNA, disproportionate to later decreases in mRNA of myelin protein genes. This pattern of decrease in enzyme activity is consistent with the morphological evidence of normal early myelination followed by destruction of myelin and oligodendrocytes after 20 days. Abnormal accumulation of a toxic metabolite, psychosine, is considered the underlying cause of this rapid degeneration of the oligodendrocytes (psychosine hypothesis) (Suzuki, 1998) (regarding detailed discussions of the psychosine hypothesis, refer to Chapter 35 in this volume). Galactosylceramidase had for many years eluded the best eVorts of several laboratories to purify. However, Wenger and colleagues recently succeeded to purify and clone the human cDNA (Chen et al., 1993). Later Sakai et al. also cloned the cDNA independently (Sakai et al., 1994). Then, the mutation underlying the twitcher mutant was identiWed (Sakai et al., 1996). It is a nonsense mutation near the middle of the coding sequence that results in a mutation of codon 339 to a stop codon (TGG!TGA).

Experimental Mouse Models of GLD In addition to the naturally occurring twitcher mutant, a new experimental model of GLD due to genetic galactosylceramidase deWciency and an entirely new, genetically distinct mouse model of GLD have recently been generated. Particularly, the latter, a genetic deWciency of a sphingolipid activator protein, saposin A, introduced a new gene, defect of which can also cause impaired degradation of galactosylceramidase substrates and thus globoid cell leukodystrophy. While GLD due to saposin A deWciency is not known in humans, an equivalent human disease can be anticipated. Point Mutation in Galactosylceramidase Gene There are several base changes in human galactosylceramidase gene, which are observed in the general population and which by themselves do not cause globoid cell leukodystrophy. They are therefore by deWnition polymorphisms. Some of them, however, result in galactosylceramidase protein with reduced catalytic activity, although not to the extent to cause

V. ANIMAL MODELS OF HUMAN DISEASE

MOUSE MODELS

the disease. Wenger noted that some individuals who simultaneously have three or more of these ‘‘polymorphisms’’ often exhibit late-onset spasticity and other white matter signs. Tested in an in vitro system, one of the polymorphisms, H168C, in human gene gave 80% of normal activity in COS-1 cells, but the same base change in the mouse gene gave only 10 to 20% of normal activity. Based on this observation, H168C was introduced into the mouse gene by homologous recombination technology. Mice homogygous for H168C exhibited a phenotype of GLD with slightly later onset and milder course (Luzi et al., 2001). Clinical onset was about 25 to 30 days (20 days in twitcher mouse), and the average life span was about 50 days with large variations. Galactosylceramidase activities in various tissues were so low that it could not be reliably compared with that in twitcher mouse. Pathology was similar to that of the twitcher mouse. Psychosine accumulation in the brain was approximately 80% of the level observed in the twitcher mouse. Thus, although not by itself pathogenic in humans, H168C is a disease-causing mutation in the mouse. These characterizations were done relatively soon after the generation of the mutant when the genetic background was still mixed. In view of the longerliving twitcher mice with diVerent genetic background (Duchen et al., 1980), it is of interest to follow up this initial observation when the genetic background is made more homogeneous. Saposin A Deficiency Sphingolipid activator proteins (saposins A, B, C, D) are small heat-stable glycoproteins required for in vivo degradation of some sphingolipids with short carbohydrate chains (SandhoV et al., 2001). They are derived from a common precursor protein (prosaposin), which is encoded by a single gene and proteolytically processed to saposins A, B, C, and D. These four saposins are all homologous to each other, having six conserved cysteines and one common glycosylation site. In spite of these structural similarities, their activator functions are relatively speciWc, with some overlaps, for individual sphingolipid hydrolases. Human patients with mutations in the saposin B and C domains are known and they show phenotypes of metachromatic leukodystrophy and Gaucher disease, indicating that they primarily activate arylsulfatase A and glucosylceramide in vivo, respectively (Christomanou et al., 1986; Shapiro et al., 1979). Two mutations are known in humans that result in complete inactivation of all four saposins and prosaposin (Hulkova´ et al., 2001; Schnabel et al., 1992). A mouse model of human total saposin deWciency closely mimics the human disease (Fujita et al., 1996). There have been circumstantial evidence that saposin A is a galactosylceramidase activator (Harzer et al., 1997; Morimoto et al., 1989). A saposin AdeWcient mouse line has been generated experimentally by introducing an amino acid substitution (C106F) into the saposin A domain by the Cre/loxP system (Matsuda et al., 2001b) that eliminated one of the three conserved disulWde bonds considered essential for the functional properties of saposins. In humans, an equivalent mutation in the 4th cysteine to phenylalanine in saposin C causes speciWc saposin C deWciency, and a mutation of the 5th cysteine to serine in saposin B causes speciWc saposin B deWciency (SandhoV et al., 2001). Saposin A
/
mice developed slowly progressive hind leg paralysis with the clinical onset around 2.5 months and survival up to 5 months. Tremors and shaking prominent in other myelin mutants were not conspicuous until the terminal stage. Both males and females of saposin A
/
are fertile, and females are able to raise pups up to three times. In every respect, pathology (Wrm and thick peripheral nerves, demyelination, reactive astrogliosis, and inWltration of ‘‘globoid cells’’ that contain the characteristic inclusion bodies) and analytical biochemistry (consequences of impaired degradation of all of the galactosylceramidase substrates in the brain, kidney, and testis) were qualitatively identical with but generally much milder than those seen in the twitcher mouse. Accumulation of psychosine in the brain was only twice normal compared to the 10- to 20-fold increase in twitcher mice. The saposin A
/
mouse clearly established that saposin A is indispensable for galactosylceramidase activation in vivo in the sense that normal cellular functions cannot be maintained in its absence. It should now be recognized that, in addition to galactosylceramidase deWciency, genetic saposin A deWciency can also cause chronic globoid cell leukodystrophy. Genetic saposin A deWciency should be suspected in

V. ANIMAL MODELS OF HUMAN DISEASE

1105

1106

45. MODELS OF KRABBE DISEASE

human patients with undiagnosed late-onset chronic leukodystrophy without known enzymatic deWciencies.

Utilities of Mouse Models Human Krabbe disease is a rare disorder. It is always diYcult to design well-controlled studies with human patients because of the relatively long life and genetic complexity. This problem is compounded when the disease is rare. Furthermore, any studies with human individuals pose serious ethical constraints. Appropriate animal models equivalent to human disease can alleviate many of these limitations of human studies. Thus, they are being utilized with increasing frequency and the GLD animal models are no exception. Pathogenesis and Other Basic Studies Although the fundamental genetic basis of GLD is well understood and the globoid cell reaction due to impaired galactosylceramide degradation and disappearance of myelinforming cells due to accumulation of cytotoxic psychosine appear to be the major pathogenetic mechanisms underlying the disease, many details are still uncertain. The mouse models are well suited for exploration of these questions. Some Wndings related to this topic are discussed in other sections of this chapter. Involvement of apoptotic process While it is well accepted now that cytotoxicity of psychosine is likely to be responsible for the characteristic rapid disappearance of the oligodendrocytes, the exact mechanism is not well characterized. Involvement of the cellular apoptotic processes has been explored in this regard, and an increasing number of oligodendrocytes in twitcher mouse brains were found to be dying from apoptotic processes (Taniike and Suzuki 1995; Taniike et al., 1999). This observation is consistent with an in vitro study, which indicated that psychosine is as potent an inducer of apoptosis as C-6 ceramide (Tohyama et al., 2001). Involvement of immune and inflammatory processes One of the issues related to the pathogenesis of GLD is the extent and nature of immune and inXammatory system involvement. The role of the major histocompatibility complex in the pathogenesis of twitcher mice was evaluated. With progression of demyelination, GFAP-, Mac-1- and F4/80-positive cells increased both in the twitcher CNS and PNS. Some Mac-1-positive cells also expressed the major histocompatibility complex (MHC) class II (Ia). Emergence of Ia-expressing cells was largely coincident with the onset of demyelination. Ia-immunoreactive cells gradually increased in areas of demyelination, reached a plateau between 30 to 40 days of age in the cerebrum, and then rapidly decreased despite continuous demyelination. In the spinal cord, however, Ia-immunoreactive cells did not decline even at 50 days (Ohno et al., 1993). These results may indicate either a speciWc involvement of immunological factors in the pathogenesis of this genetic demyelinating disease or a nonspeciWc reaction to degenerating tissue components, such as myelin. Cross-breeding the twitcher mice with a MHC class II knockout mouse line generated a twitcher mouse line, which is also deWcient with MHC class II molecules (Grusby et al., 1991). In these mice, clinical symptoms and histopathology of the cerebrum and the brain stem–cerebellar region were milder than those in twitcher mice with the MHC-II positive background but there was no noticeable improvement in the pathology of the spinal cord (Matsushima et al., 1994). Preliminary analysis of psychosine level also showed less accumulation than in the MHCII-positive twitcher mice, consistent with the lesser degree of pathology in these mice. Interleukins, cytokines, and their receptors have also been examined in the twitcher mice to evaluate participation of inXammatory processes in the pathogenesis of GLD (LeVine et al., 1994; LeVine and Brown 1997; Matsuda et al., 2001a; Pedchenko and LeVine 1999; Podchenko et al., 2000; Wu et al., 2001). Interactions between the two lysosomal ß-galactosidases Two ß-galactosidases are present in the lysosome. One is galactosylceramidase responsible for human Krabbe

V. ANIMAL MODELS OF HUMAN DISEASE

MOUSE MODELS

disease and several animal models. The other is acid ß-galactosidase, which primarily hydrolyzes the terminal galactose from GM1-ganglioside and its asialo-derivative, GA1, and is responsible for GM1-gangliosidosis/Morquio B disease. Cross-breeding of twitcher mice and acid ß-galactosidase knockout mice indicated that the acid ß-galactosidase gene dosage exerts unexpected and paradoxical inXuence on the twitcher phenotype (Tohyama et al., 2000). Twitcher mice with additional complete deWciency of acid ß-galactosidase showed the mildest phenotype with the longest life span and nearly rescued CNS pathology. In contrast, twitcher mice with a single functional acid ß-galactosidase gene exhibited the most severe disease with the shortest life span. A signiWcant proportion of these galc
/
, bgal þ/
mice clinically developed additional extreme hyperreactivity and generalized seizures not seen in any other genotypes. Widespread neuronal degeneration was present in the galc
/
, bgal þ/
mice, most prominently in the CA3 region of the hippocampus. The double knockout mice showed a massive accumulation of lactosylceramide in all tissues as had been expected from earlier enzymological studies (Tanaka and Suzuki, 1977). The brains inexplicably contained only a half-normal amount of galactosylceramide, which may account for the mild clinical and pathological phenotype. On the other hand, galc
/
, bgal þ/
mice showed a signiWcantly higher level of brain psychosine than other genotypes. The reduced galactosylceramide in the brain of the double knockout mice, and the signiWcantly higher psychosine in the brain of the galc
/
, bgal þ/
mice cannot readily be explained from the genotypes of these mice. These observations are contrary to the expected outcome of Mendelian autosomal recessive single gene disorders. Acid ß-galactosidase gene may function as a modiWer gene for the phenotypic expression of genetic galactosylceramidase deWciency. Therapeutic Trials Nerve graft experiment The twitcher mutant has been used extensively for therapeutic trials for GLD since the identiWcation of its underlying genetic defect as galactosylceramidase deWciency. In earlier studies, sciatic nerve of aVected mice was grafted to the sciatic nerve of either normal or trembler mutant mice (Scaravilli and Jacobs, 1982; Scaravilli and Suzuki, 1983). When twitcher sciatic nerves were grafted to the sciatic nerve of normal littermates, endoneurial edema became pronounced and features of demyelination appeared 2 months after the graft operation. However, at 6 months, there were signs of improvement in pathology. Many nerve Wbers were myelinated, although myelin was far thinner for the size of the axons. Galactosylceramidase activities of the grafted twitcher sciatic nerves were indistinguishable from those in the host nerves at 4, 5, and 9 months. By using trembler mice as the host, migration of host Schwann cells into the grafted twitcher nerve could be excluded as the source of the high galactosylceramidase activity in the graft. Bone marrow transplantation The bone marrow transplantation experiments prolonged life span of twitcher mice with a 36% probability of survival at 100 days in contrast to untreated mice that do not survive beyond 35 to 40 days (Hoogerbrugge et al., 1988a; Yeager et al., 1984). Hind limb weakness was signiWcantly improved but tremulousness still remained. Nevertheless, body weight of the treated mice remained much below that of normal mice throughout their life span. After bone marrow transplantation, galactosylceramidase activity in the spleen and bone marrow increased to the donor level. Also the enzyme activity was signiWcantly increased in the liver, lung and heart. In the CNS, the enzyme activity gradually increased to 15% of the control level in 100-day old treated mice (Hoogerbrugge et al., 1988a). On the other hand, another group reported restoration of normal level of galactosylceramidase activity in the twitcher mouse brain by 90 days after transplantation (Ichioka et al., 1987). Psychosine accumulation decreased in the CNS following bone marrow transplantation and remained stable (Hoogerbrugge et al., 1988a; Ichioka et al., 1987). Motor nerve conduction velocity improved in the sciatic nerve (Toyoshima et al., 1986). In the late stage, morphological evidence of remyelination was present in the CNS (Suzuki et al., 1988). Many nerve Wbers in the cerebellum, brain stem, and spinal cord were myelinated, and myelin degeneration was rare in these regions, although thickness of the myelin sheaths was far less than expected for the size of the

V. ANIMAL MODELS OF HUMAN DISEASE

1107

1108

45. MODELS OF KRABBE DISEASE

axons, indicating hypomyelination or remyelination. With allogeneic transplantation using C3H strain of mice [H-2Kk allele of the major histocompatibility complex (MHC)] as the bone marrow donor to twitcher mice that are of C57BL strain [H-2Kb allele of MHC complex], donor origin of inWltrating macrophages was demonstrated with immunocytochemical techniques (Hoogerbrugge et al., 1988b; Huppes et al., 1992). Other studies also showed that hematogenous cells can inWltrate into the central nervous system (Wu et al., 2000a, 2000b). Remyelinating oligodendrocytes appeared normal at the light microscopic level with abundant cytoplasm. However, examination on the ultrastructural level indicated presence of GLD inclusions in their perikarya as well as in the inner or outer tongue processes (Suzuki et al., 1988). Apparent recovery from demyelination in the regions where foamy cells were conspicuously present suggested that these cells were responsible for the improvement of the CNS pathology by somehow providing the normal enzyme. Cellular transplantation Transplantation of isolated and cultured glial cells has been used to remyelinate demyelinated or dysmyelinated white matter in experimental animals (Gumpel et al., 1983). Transplanted glial cells migrate along the nerve Wbers, blood vessels, and the subependymal region. They appear to migrate preferentially toward the site of demyelination. Intracerebrally transplanted fetal neuroglial cells were found to spread widely within the host twitcher brain (Huppes et al., 1992). Intracerebal transplantation of neural stem cells is being actively pursued, and it is hoped that the results would be published in due course. Transgenic treatment Introduction of normal gene as a transgene into the genome is not and will not be in the foreseeable future a viable means for clinical treatment of genetic disease. Nevertheless, the procedure is of interest as a basic study. Matsumoto et al. introduced normal human galactosylceramidase cDNA as a transgene into twitcher mice under control of the promoter of a myelin-speciWc gene, myelin basic protein (Matsumoto et al., 1997), in an attempt to inXuence the natural course of the disease. Expression of the transgene was unexpectedly low and could be detected only by RT-PCR. When determined at 35 days, galactosylceramidase activities in the brain and kidney showed a general trend of increasing activities from the untreated twitcher mice, twitcher mice carrying a single transgene, to twitcher mice carrying a double dose of the transgene. However, increment in the enzymatic activities in mice carrying the transgene was very minor. Nevertheless, while twitcher mice without the transgene developed the disease early, lost weight, and died by 35 + days, those carrying one transgene developed the disease more slowly, gained weight longer, and survived generally to 50 + days. Those with double dose of the transgene developed the disease even later, gained weight much longer, and survived up to 66 days. Pathology at 35 days suggested corresponding improvements in transgene-carrying twitcher mice. Most important, psychosine levels in the brain were reduced to nearly half of those without the transgene. This ‘‘failed’’ experiment with respect to ‘‘curing’’ the disease by introduction of the transgene nevertheless provided encouraging evidence to support the hypothesis that a relatively small increment in the enzymatic activity may be suYcient to eVect ‘‘cure’’ of many lysosomal disorders. Gene therapy Twitcher mice are being used to explore possible avenues for gene therapy. Highly promising results have been obtained in cell culture systems (CostantinoCeccarini et al., 1999; RaW et al., 1996). The most commonly employed viral vector is the retroviruses for their well-characterized properties and the eYcient and relatively nondiscriminatory capacity to infect various cellular types. If a retrovirus is appropriately modiWed with an inserted exogenous gene, it can infect the target cells and produce the gene product lacking in the host cell. For stable expression, the retrovirus must be integrated into the host genome, and the integration requires mitosis of the host cells. This property seriously limits the usefulness of retroviruses as the vector for gene therapy of genetic diseases of the central nervous system, in which most cells are post-mitotic. Use of either adenovirus or replication defective Herpes simplex 1 virus can overcome this disadvantage, since they do not need to be integrated into the host cell genome for its

V. ANIMAL MODELS OF HUMAN DISEASE

MOUSE MODELS

expression. Thus, the exogenous gene engineered into adenovirus or Herpes simplex 1 virus can be expressed in post-mitotic cells and actively replicating cells with similar eYciency. This ability to replicate and express its own genome independent of the host cell replication further provides an advantage in that there is much less risk of inadvertently activating oncogenes or disrupting important endogenous genes by random integration into the host genome. The disadvantage is that the introduced genes are ‘‘diluted’’ and eventually lost as the host cells undergo series of mitosis. More recently, adeno-associated virus (AAV) is gaining popularity as a vector for overcoming many of the shortcomings of other vectors. Some of the advantages of the AAV are (1) AAV is nonpathogenic, since the integrating DNA vector has no remaining viral genes and the co-transfected helper adenovirus is completely eliminated in time; (2) integrated gene carried by the AAV vector is stable at least in tissue culture for more than 100 passages; (3) AAV integrates itself into a small predictable region of the host genome (chromosome 19 in humans) and thus has little danger of disrupting essential host genes; (4) AAV can be introduced into most cells types; (5) the replication-defective helper adenovirus can be produced at high titers; and (6) AAV is the only known human virus that has targeting capability. AAV has also been explored as the viral vehicle for gene therapy of twitcher mice (Chen et al., 1998, 1999). Pregnancy dramatically alleviates saposin A deficient phenotype During routine breeding process, it was noted that aVected saposin A
/
females that were continually pregnant showed greatly improved neurological symptoms compared to aVected females that do not experience pregnancy or aVected males (Matsuda et al., 2001a). The pathological hallmark of globoid cell leukodystrophy, demyelination with inWltration of globoid cells, largely disappeared. The immune-related gene expression (MCP-1, TNF-a) was signiWcantly down-regulated in the brain of pregnant saposin A
/
mice. In addition, intense expression of estrogen receptors (ER-a and ER-b) on the globoid cells, activated astrocytes and microglia in the demyelinating area of saposin A
/
mice was observed. When saposin A
/
mice were subcutaneously implanted with time-release 17b-estradiol (E2) pellets from 30 to 90 days, the pathology was vastly improved. These Wndings suggest that higher level of estrogen during pregnancy is one of the important factors in the protective eVect of pregnancy. While we should be cautious in extrapolating these observations in the mouse to human disease, the phenomenon is spectacularly dramatic and estrogen administration might be worth a consideration as a supplementary treatment for some chronic genetic leukodystrophies. It must be emphasized that the degree of the clinical and pathological improvement by pregnancy is much greater than those seen in any other therapeutic trials so far attempted for this disease. Substrate reduction All lysosomal diseases result from failure to degrade certain cellular constituents due to genetic defects in the degradative pathways. Radin originally proposed a concept of substrate reduction as a means to alleviate the detrimental outcome of lysosomal disease by limiting synthesis of the substrates and thus reducing burden on the degradative system (Radin, 1996, 1999; Vunnam and Radin, 1980). Radin could not test his concept because no suitable animal models were available when he proposed the idea in the 1970s. With the advent of gene targeting technology, the concept can now be tested and has in fact become fashionable. While most of the experimental tests have been done with animal models, in which glucosylceramide-based sphingolipids are stored, using inhibitors of glucosylceramide synthesis, this approach is not feasible for GLD because the disease does not involve impaired degradation of glucosylceramide-based sphingolipids. Instead, use of cycloserine, which inhibits all sphingolipid synthesis, was suggested to treat twitcher mice in 1985 (Sundaram and Lev, 1984), and recent actual trials found it to be slightly eVective (Biswas and LeVine, 2002; LeVine et al., 2000). A logical weakness of all experiments with exogenous administration of inhibitors of substrate synthesis is that one can only know what the compounds do or do not do with respect to the functions tested but never know what else they might do. Thus, it is extremely diYcult to prove a causal relationship between substrate inhibition and outcome on the course of the disease. A biologically ‘‘cleaner’’ experiment was done by

V. ANIMAL MODELS OF HUMAN DISEASE

1109

1110

45. MODELS OF KRABBE DISEASE

cross-breeding twitcher mice with galactosylceramide synthase (UDP-galactose:ceramide galactosyltransferase, CGT) knockout mice (Ezoe et al., 2000a, 2000b). The prediction that the phenotype of the doubly deWcient mice should be the same as the cgt
/
mice, since the degrading enzyme should not be necessary if the substrate is not synthesized, proved to be only partially correct. In early stages of the disease, the doubly deWcient mice (galc
/
, cgt
/
) were essentially indistinguishable from the cgt
/
mice. However, the doubly deWcient mice had a much shorter life span than cgt
/
mice. Both galactosylceramide and galactosylsphingosine (psychosine) were undetectable in the brain of the cgt
/
and the doubly deWcient mice. The characteristic twitcher pathology was never seen in the galc
/
, cgt
/
mice. However, after 43 days, neuronal pathology was observed in the brain stem and spinal cord. This late neuronal pathology has not been seen in the CGT knockout mice. These observations indicate that the functional relationship between galactosylceramidase and galactosylceramide synthase is complex. This crossbreeding also generates hybrids with a genotype of galc
/
, cgtþ/
, in addition to doubly deWcient mice. They are ideally suited to evaluate the potential usefulness of limiting synthesis of the substrate as a treatment of genetic disorders due to degradative enzyme defects. The rate of accretion of galactosylceramide in the brain of CGT knockout carrier mice (cgtþ/
) is approximately two-thirds of the normal rate, suggesting a gene-level compensation for the reduced gene dosage. Phenotype of twitcher mice with a single dose of normal cgt gene was indeed milder with statistical signiWcance, albeit only slightly. Neuropathologists were able to distinguish blindly galc
/
, cgtþ/
mice from galc
/
, cgtþ/þ mice. The brain psychosine level in galc
/
, cgtþ/
mice was also approximately two-thirds of the galc
/
, cgtþ/þ mice. These observations indicate that reduction of galactosylceramide synthesis to two-thirds of the normal level results in minor but clearly detectable phenotypic improvements. This approach by itself is unlikely to be useful as the sole treatment but may be helpful as a supplement to other therapies.

References Baskin G. B., Ratterree M., Davison B. B., Falkenstein K. P., Clarke M. R., England J. D., Vanier M. T., Luzi P., RaW M. A., and Wenger D. A. (1998). Genetic galactocerebrosidase deWciency (globoid cell leukodystrophy, Krabbe disease) in Rhesus monkeys (Macaca mulatta). Lab. Anim. Sci. 48, 476–482. Biswas S., and LeVine S. M. (2002). Substrate-reduction therapy enhances the beneWts of bone marrow transplantation in young mice with globoid cell leukodystrophy. Pediatr. Res. 51, 40–47. Chen H., McCarty D. M., Bruce A. T., Suzuki K., and Suzuki K. (1998). Gene transfer and expression in oligodendrocytes under the control of myelin basic protein transcriptional control region mediated by adenoassociated virus. Gene Therapy 5, 50–58. Chen H., McCarty D. M., Bruce A. T., Suzuki K., and Suzuki K. (1999). Oligodendrocyte-speciWc gene expression in mouse brain: Use of a myelin-forming cell type-speciWc promoter in an adeno-associated virus. J. Neurosci. Res. 55, 504–513. Chen Y. Q., RaW M. A., De Gala G., and Wenger D. A. (1993). Cloning and expression of cDNA encoding human galactocerebrosidase, the enzyme deWcient in globoid cell leukodystrophy. Hum. Mol. Genet. 2, 1841–1845. Christomanou H., Aignesberger A., and Link R. P. (1986). Immunochemical characterization of two activator proteins stimulating enzymic sphingomyelin degradation in vitro: Absence of one of them in a human Gaucher disease variant. Biol. Chem. Hoppe-Seyler 367, 879–890. Costantino-Ceccarini E., Luddi A., Volterrani M., Strazza M., RaW M. A., and Wenger D. A. (1999). Transduction of cultured oligodendrocytes from normal and twitcher mice by a retroviral vector containing human galactocerebrosidase (GALC). cDNA. Neurochem. Res. 24, 287–293. Costantino-Ceccarini E., and Morell P. (1972). Biosynthesis of brain sphingolipids and myelin accumulation in the mouse. Lipids 7, 656–659. Duchen L. W., Eicher E. M., Jacobs J. M., Scaravilli F., and Teixeira F. (1980). Hereditary leukodystrophy in the mouse – the new mutant twitcher. Brain 103, 695–710. Ezoe T., Vanier M. T., Oya Y., Popko B., Tohyama J., Matsuda J., Suzuki K., and Suzuki K. (2000a). Biochemistry and neuropathology of mice doubly deWcient in synthesis and degradation of galactosylceramide. J. Neurosci. Res. 59, 170–178. Ezoe T., Vanier M. T., Oya Y., Popko B., Tohyama J., Matsuda J., Suzuki K., and Suzuki K. (2000b). Twitcher mice with only a single active galactosylceramide synthase gene exhibit clearly detectable but therapeutically minor phenotypic improvements. J. Neurosci. Res. 59, 179–187. Fankha¨user R., Luginbu¨hl H., and Hartley W. J. (1963). Leukodystrophie vom Typus Krabbe beim Hund. Schweiz. Arch. Tierheilkd. 105, 198–207.

V. ANIMAL MODELS OF HUMAN DISEASE

MOUSE MODELS

Fujita N., Suzuki K., Vanier M. T., Popko B., Maeda N., Klein A., Henseler M., SandhoV K., Nakayasu H., and Suzuki K. (1996). Targeted disruption of the mouse sphingolipid activator protein gene: A complex phenotype, including severe leukodystrophy and wide-spread storage of multiple sphingolipids. Hum. Mol. Genet. 5, 711–725. Grusby M. L., Johnson R. S., Papaloannou V. E., and Glimcher L. H. (1991). Depletion of CD4þ T cells in major histocompatibility compex class II-deWcient mice. Science 253, 670–674. Gumpel M., Baumann N., Raoul M., and Jacque C. (1983). Survival and diVerentiation of oligodendrocytes from neural tissue transplanted into new-born mouse brain. Neurosci. Lett. 37, 307–312. Harzer K., Paton B. C., Christomanou H., Chatelut M., Levade T., Hiraiwa M., and O’Brien J. S. (1997). Saposins (sap). A and C activate the degradation of galactosylceramide in living cells. FEBS Lett. 417, 270–274. Hoogerbrugge P. M., Poorthuis B. J. H. M., Romme A. E., van de Kamp J. J. P., Wagemaker G., and van Bekkum D. W. (1988a). EVect of bone marrow transplantation on enzyme levels and clinical course in the neurologically aVected twitcher mouse. J. Clin. Invest. 81, 1790–1794. Hoogerbrugge P. M., Suzuki K., Suzuki K., Poorthuis B. J. H. M., Kobayashi T., Wagemaker G., and van Bekkum D. W. (1988b). Donor derived cells in the central nervous system of twitcher mice after bone marrow transplantation. Science 239, 1035–1038. Hulkova´ H., Cervenkova´ M., Ledvinova´ J., Tochackova´ M., Hrebı´cek M., Poupetova´ H., Befekadu A., Berna´ L., Paton B. C., Harzer K., Bo¨o¨r A., Smı´d F., and Elleder M. (2001). A novel mutation in the coding region of the prosaposin gene leads to a complete deWciency of prosaposin and saposins, and is associated with a complex sphingolipidosis dominated by lactosylceramide accumulation. Hum. Mol. Genet. 10, 927–940. Huppes W., De Groot C. J. A., Ostendorf R. H., Bauman J. G. J., Gossen J. A., Smit V., Vijg J., and Dijkstra C. D. (1992). Detection of migrated allogeneic oligodendrocytes throughout the central nervous system of the galactocerebrosidase-deWcient twitcher mouse. J Neurocytol 21, 129–136. Ichioka T., Kishimoto Y., Brennan S., Santos G. W., and Yeager A. M. (1987). Hematopoietic cell transplantation in murine globoid cell leukodystrophy (the twticher mouse): EVects on levels of galactosylceramidase, psychosine, and galactocerebrosides. Proc. Natl. Acad. Sci. USA 84, 4259–4263. Ida H., Umezawa F., Kasai E., Eto Y., and Maekawa K. (1982). An accumulation of galactocerebroside in kidney from mouse globoid cell leukodystrophy (twitcher). Biochem. Biophys. Res. Commun. 109, 634–638. Igisu H., Shimomura K., Kishimoto Y., and Suzuki K. (1983a). Lipids of developing brain of twitcher mouse— An authentic murine model of human Krabbe disease. Brain 106, 405–417. Igisu H., and Suzuki K. (1984a). Glycolipids of the spinal cord, sciatic nerve and systemic organs of the twitcher mouse. J. Neuropath. Exper. Neurol. 43, 22–36. Igisu H., and Suzuki K. (1984b). Progressive accumulation of toxic metabolite in a genetic leukodystrophy. Science 224, 753–755. Igisu H., Takahashi H., Suzuki K., and Suzuki K. (1983b). Abnormal accumulation of galactosylceramide in the kidney of twitcher mouse. Biochem. Biophys. Res. Commun. 110, 940–944. Kobayashi S., Chiu F.-C., Katayama M., Sacchi R. S., Suzuki K., and Suzuki K. (1986). Expression of glial Wbrillary acidic protein in the CNS and PNS of murine globoid cell leukodystrophy, the twitcher. Am. J. Pathol. 125, 227–243. Kobayashi S., Katayama M., Satoh J., Suzuki Ku., and Suzuki K. (1988). The twitcher mouse: An alteration of the unmyelinated Wbers in the PNS. Am. J. Pathol. 131, 308–319. Kobayashi T., Nagara H., Suzuki K., and Suzuki K. (1982). The twitcher mouse: Determination of genetic status by galactosylceramidase assays on clipped tail. Biochemical Medicine 27, 8–14. Kobayashi T., Yamanaka T., Jacobs J. M., Teixeira F., and Suzuki K. (1980). The twitcher mouse: An enzymatically authentic model of human globoid cell leukodystrophy (Krabbe disease). Brain Res. 202, 479–483. Kodama S., Igisu H., Siegel D. A., and Suzuki K. (1982). Glycosylceramide synthesis in the developing spinal cord and kidney of the twitcher mouse, an enzymatically authentic model of human Krabbe disease. J. Neurochem. 39, 1314–1318. LeVine S. M., and Brown D. C. (1997). IL-6 and TNFa expression in brains of twitcher, quaking and normal mice. J. Neuroimmunol. 73, 47–56. LeVine S. M., Pedchenko T. V., Bronshteyn I. G., and Pinson D. M. (2000). L-cycloserine slows the clinical and pathological course in mice with globoid cell leukodystrophy (twitcher mice). J. Neurosci. Res. 60, 231–236. LeVine S. M., Wetzel D. L., and Eiler A. J. (1994). Neuropathology of twitcher mice: Examination by histochemistry, immunohistochemistry, lectin histochemistry and Fourier transform infrared microspectroscopy. In. J. Develop. Neurosci. 12, 275–288. Luzi P., RaW M. A., Victoria T., Baskin G. B., and Wenger D. A. (1997). Characterization of the rhesus monkey galactocerebrosidase (GALC). cDNA and gene and identiWcation of the mutation causing globoid cell leukodystrophy (Krabbe disease). in this primate. Genomics 42, 319–324. Luzi P., RaW M. A., Zaka M., Curtis M., Vanier M. T., and Wenger D. A. (2001). Generation of a mouse with low galactocerebrosidase activity by gene targeting: A new model of globoid cell leukodystrophy (Krabbe disease). Molecular Genetics and Metabolism 73, 211–223. Matsuda J., Vanier M. T., Saito Y., Suzuki K., and Suzuki K. (2001a). Dramatic phenotypic improvement during pregnancy in a genetic leukodystrophy: Estrogen appears to be a critical factor. Hum. Mol. Genet. 10, 2709–2715.

V. ANIMAL MODELS OF HUMAN DISEASE

1111

1112

45. MODELS OF KRABBE DISEASE

Matsuda J., Vanier M. T., Saito Y., Tohyama J., Suzuki K., and Suzuki K. (2001b). A mutation in the saposin A domain of the sphingolipid activator protein (prosaposin). gene causes a late-onset, slowly progressive form of globoid cell leukodystrophy in the mouse. Hum. Mol. Genet. 10, 1191–1199. Matsumoto A., Vanier M. T., Oya Y., Kelly D., Popko B., Wenger D. A., Suzuki K., and Suzuki K. (1997). Transgenic introduction of human galactosylceramidase into twitcher mouse: SigniWcant phenotype improvement with a minimal expression. Develop. Brain Dysfunction 10, 142–154. Matsushima G. K., Taniike M., Glimcher L. H., Grusby M. J., Frelinger J. A., Suzuki K., and Ting J. P. Y. (1994). Absence of MHC class II molecules reduces CNS demyelination, microglial; macrophage inWltration, and twitching in murine globoid cell leukodystrophy. Cell 78, 645–656. Mikoshiba K., Fujishiro M., Kohsaka S., Okano H., Takamatsu K., and and Tsukada Y. (1985). Disorders in myelination in the twitcher mutant: Immunohistochemical and biochemical studies. Neurochem. Res. 10, 1129–1141. Morimoto S., Martin B. M., Yamamoto Y., Kretz K. A., O’Brien J. S., and Kishimoto Y. (1989). Saposin-A – 2nd Cerebrosidase Activator Protein. Proc. Natl. Acad. Sci., USA 86, 3389–3393. Ohno M., Komiyama A., Martin P. M., and Suzuki K. (1993). MHC class II antigen expression and T-cell inWltration in the demyelinating CNS and PNS of the twitcher mouse. Brain Res. 625, 186–196. Olmstead C. E. (1987). Neurological and neurobehavioral development of the mutant ‘twitcher’ mouse. Behav. Brain Res. 25, 143–152. Podchenko T. V., Broshteyn I. G., and LeVine S. M. (2000). TNF-receptor I deWciency fails to alter the clinical and pathological course in mice with globoid cell leukodystrophy (twitcher mice). but aVords protection following LPS challenge. J. Neuroimmunol. 110, 186–194. Pedchenko T. V., and LeVine S. M. (1999). IL-6 deWciency causes enhanced pathology in twitcher (globoid cell leukodystrophy). mice. Exp. Neurol. 158, 459–468. Powell H. C., Knobler R. L., and Myers R. R. (1983). Peripheral neuropathy in the twitcher mutant: A new experimental model of endoneurial edema. Lab. Invest. 49, 19–25. Radin N. S. (1996). Treatment of Gaucher disease with an enzyme inhibitor. Glycoconjugate J. 13, 153–157. Radin N. S. (1999). Chemotherapy by slowing glucosphingolipid synthesis. Biochem. Pharmacol. 57, 589–595. RaW M. A., Fugaro J., Amini S., Luzi P., De Gala G., Victoria T., Dubell C., Shahinfar M., and Wenger D. A. (1996). Retroviral vector-mediated transfer of the galactocerebrosidase (GALC). cDNA leads to overexpression and transfer of GALC activity to neighboring cells. Biochem. Molec. Med. 58, 142–150. Sakai N., Inui K., Fujii N., Fukushima H., Nishimoto J., Yanagihara I., Isegawa Y., Iwamatsu A., and Okada S. (1994). Krabbe disease: Isolation and characterization of a full-length cDNA for human galactocerebrosidase. Biochem. Biophys. Res. Commun. 198, 485–491. Sakai N., Inui K., Tatsumi N., Fukushima H., Nishigaki T., Taniike M., Nishimoto J., Tsukamoto H., Yanagihara I., Ozone K., and Okada S. (1996). Molecular cloning and expression of cDNA for murine galactocerebrosidase and mutation analysis of the twitcher mouse, a model of Krabbe disease. J. Neurochem. 66, 1118–1124. SandhoV K., Kolter T., and Harzer K. (2001). Sphingolipid activator proteins. In ‘‘The Metabolic and Molecular Basis of Inherited Disease’’ (C. R. Scriver A. L. Beaudet, W. S. Sly, and D. Valle, eds.), pp. 3371–3388. McGraw-Hill, New York. Scaravilli F., and Jacobs J. M. (1982). Improved myelination in nerve grafts from the leukodystrophic twitcher into trembler mice – Evidence for enzyme replacement. Brain Res. 237, 163–172. Scaravilli F., and Suzuki K. (1983). Enzyme replacement in grafted nerve of twitcher mouse. Nature 305, 713–715. Schnabel D., Schro¨der M., Fu¨rst W., Klein A., Hurwitz R., Zenk T., Weber J., Harzer K., Paton B. C., Poulos A., Suzuki K., and SandhoV K. (1992). Simultaneous deWciency of sphingolipid activator proteins 1 and 2 is caused by a mutation in the initiation codon of their common gene. J. Biol. Chem. 267, 3312–3315. Schulte S., and StoVel W. (1993). Ceramide UDPgalactosyltransferase from myelinating rat brain: PuriWcation, cloning, and expression. Proc. Natl. Acad. Sci., USA 90, 10265–10269. Shapiro L. J., Aleck K. A., Kaback M. M., Itabashi H., Desnick R. J., Brand N., Stevens R. L., Fluharty A. L., and Kihara H. (1979). Metachromatic leukodystrophy without arylsulfatase A deWciency. Pediat. Res. 13, 1179–1181. Stahl N., Jurevics H., Morell P., Suzuki K., and Popko B. (1994). Isolation, characterization, and expression of cDNA clones that encode rat UDP-galactose:ceramide galactosyltransferase. J. Neurosci. Res. 38, 234–242. Sundaram K. S., and Lev M. (1984). Inhibition of sphingolipid synthesis by cycloserine in vitro and in vivo. J. Neurochem. 42, 577–581. Suzuki K. (1971). Renal cerebrosides in globoid cell leukodystrophy (Krabbe disease). Lipids 6, 433–436. Suzuki K. (1998). Twenty Wve years of the psychosine hypothesis: A personal perspective of its history and present status. Neurochem. Res. 23, 251–259. Suzuki K., Hoogerbrugge P. M., Poorthuis B. J. H. M., Romme A. E., van de Kamp J. J. P., Wagemaker G., van Bekkum D. W., and Suzuki K. (1988). The twitcher mouse: Central nervous system pathology after bone marrow transplantation (BMT). Lab. Invest. 58, 302–309. Suzuki K., and Suzuki Y. (1970). Globoid cell leucodystrophy (Krabbe disease): DeWciency of galactocerebroside b-galactosidase. Proc. Natl. Acad. Sci. USA 66, 302–309. Suzuki Y., Austin J., Armstrong D., Suzuki K., Schlenker J., and Fletcher T. (1970). Studies in globoid leukodystrophy: Enzymatic and lipid Wndings in the canine form. Exp. Neurol. 29, 65–75. Tanaka H., and Suzuki K. (1977). Sustrate speciWcities of the two genetically distinct human brain b-galactosidases. Brain Res. 122, 325–335.

V. ANIMAL MODELS OF HUMAN DISEASE

MOUSE MODELS

Taniike M., Marcus J. R., Nishigaki T., Fujita N., Popko B., and Suzuki K. (1998). Suppressed UDP-galactose:ceramide galactosyltransferase and myelin protein mRNA in twitches mouse brain. J. Neurosci. Res. 51, 536–540. Taniike M., Mohri I., Eguchi N., Irikura D., Urade Y., Okada S., and Suzuki K. (1999). An apoptotic depletion of oligodendrocytes in the twitcher, a murine model of globoid cell leukodystrophy. J. Neuropathol. Exp. Neurol. 58, 644–653. Taniike M., and Suzuki K. (1994). Spacio-temporal progression of demyelination in twitcher mouse: With clinicopathological correlation. Acta Neuropathol. (Berl). 88, 228–236. Taniike M., and Suzuki K. (1995). Proliferative capacity of oligodendrocytes in the demyelinating twitcher spinal cord. J Neurosci Res 40, 325–332. Tohyama J., Matsuda J., and Suzuki K. (2001). Psychosine is as potent an inducer of cell death as C6-ceramide in cultured Wbroblasts and in MOCH-1 cells. Neurochem. Res. 26, 667–671. Tohyama J., Vanier M. T., Suzuki K., Ezoe T., Matsuda J., and Suzuki K. (2000). Paradoxical inXuence of acid b-galactosidase gene dosage on phenotype of the twitcher mouse (genetic galactosylceramidase deWciency). Hum. Mol. Genet. 9, 1699–1707. Toyoshima E., Yeager A. M., Brennan S., Santos G. W., Moser H., W., and Mayer R. F. (1986). Nerve conduction studies in the twitcher mouse (murine globoid cell leukodystrophy). J. Neurol. Sci. 74, 307–318 Victoria T., RaW M. A., and Wenger D. A. (1996). Cloning of the canine GALC cDNA and identiWcation of the mutation causing globoid cell leukodystrophy in west highland white and cairn terriers. Genomics 33, 457–462. Vunnam R. R., and Radin N. S. (1980). Analogs of ceramide that inhibit glucocerebroside synthetase in mouse brain. Chem. Phys. Lipids 26, 265–278. Wenger D. A., Suzuki K., Suzuki Y., and Suzuki K. (2001). Galactosylceramide lipidosis: Globoid cell leukodystrophy (Krabbe disease). In ‘‘The Metabolic and Molecular Basis of Inherited Disease’’ (C. R. Scriver A. L. Beaudet, W. S. Sly, and D. Valle, eds.), pp. 3669–3694. McGraw-Hill, New York. Wu Y. P., Matsuda J., Kubota A., Suzuki K., and Suzuki K. (2000a). InWltration of hematogenous lineage cells into the demyelinating central nervous system of twitcher mice. J. Neuropathol. Exp. Neurol. 59, 628–639. Wu Y. P., McMahon E., Kraine M. R., Tisch R., Meyers A., Frelinger J., Matsushima G. K., and Suzuki K. (2000b). Distribution and characterization of GFPþ donor hematogenous cells in twitcher mice after bone marrow transplantation. American Journal of Pathology 156, 1849–1854. Wu Y. P., McMahon E. J., Matsuda J., Suzuki K., Matsushima G. K., and Suzuki K. (2001). Expression of immune-related molecules is downregulated in Twitcher mice following bone marrow transplantation. J. Neuropathol. Exp. Neurol. 60, 1062–1074. Yeager A. M., Brennan S., TiVany C., Moser H. W., and Santos G. W. (1984). Prolonged survival and remyelination after hematopoietic cell transplantation in the twitcher mouse. Science 225, 1053–1054.

V. ANIMAL MODELS OF HUMAN DISEASE

1113

C H A P T E R

46 Models of Alexander Disease Albee Messing and Michael Brenner

ALEXANDER DISEASE Alexander disease, Wrst described in 1949 (Alexander, 1949), is an uncommon but fatal CNS disorder usually aVecting children (for a review, see chapter 36 by Messing and Goldman in this volume). In the most common form (‘‘infantile’’), children present before the age of 2 years with white matter loss especially in the frontal lobes. Seizures and spasticity that are diYcult to control are prominent symptoms, as is hydrocephalus and psychomotor developmental delay. These children suVer progressive deterioration with death usually before the age of 10. The key diagnostic feature of the neuropathology in Alexander disease is the widespread deposition of Rosenthal Wbers in subpial, periventricular, and white matter astrocytes throughout the CNS. Morphologically, Rosenthal Wbers consist of two components: bundles of intermediate Wlaments surrounding irregular deposits of dense material (Herndon et al., 1970). Biochemically, Rosenthal Wbers are composed of a complex ubiquitinated mixture of the intermediate Wlament GFAP in association with other constituents, especially the small stress proteins aB-crystallin and HSP25 (Head and Goldman, 2000; Iwaki et al., 1989). Rosenthal Wbers are also found sporadically in several other settings, such as some astrocytomas and chronic gliosis (their original description was in a young adult with syringomyelia; see Rosenthal, 1898), but never to the extent seen in Alexander disease. SigniWcant accumulations of Rosenthal Wbers also occur in individuals with much later onset and diVerent signs, which led to a broader classiWcation of Alexander disease that now includes juvenile and adult forms. In contrast to the infantile form, where the leukodystrophy predominantly aVects the frontal lobes and there is mental retardation, the older patients tend to have less cognitive impairment and the most severe pathology in the hindbrain, with corresponding prominence of bulbar signs. The extent to which infantile, juvenile, and adult forms of Alexander disease are all fundamentally the same disease is unclear (though at least some have a common genetic etiology as described later) (Russo et al., 1976; see also Herndon, 1999; Li et al., 2002). Why Rosenthal Wbers form, and what relation they have to the disease process, is not known. Nevertheless, the localization of Rosenthal Wbers in astrocytes led to the suggestion many years ago that Alexander disease was essentially a primary disorder of astrocytes (as originally proposed by Alexander), with the drastic consequences for other cell types reXecting a secondary eVect of astrocyte dysfunction (Borrett and Becker, 1985). The hydrocephalus that is often associated with infantile cases is sometimes attributed to Rosenthal Wbers causing stenosis of the cerebral aqueduct (Ni et al., 2002; Sherwin and Berthrong, 1970), but no direct proof of this mechanism exists and instead it could conceivably reXect primary astrocyte dysfunction.

Myelin Biology and Disorders, Volume 2

1115

Copyright 2004, Elsevier (USA). All rights reserved.

1116

46. MODELS OF ALEXANDER DISEASE

SPONTANEOUS ANIMAL MODELS Based strictly on the morphological criterion of diVuse Rosenthal Wber deposition, Alexander-like diseases have been reported occasionally in dogs (Cox et al., 1986; McGrath, 1980; Richardson et al., 1991; Weissenbo¨ck et al., 1996) and sheep (Fankhauser et al., 1980), and there is an anecdotal report of a case in a captive deer. However, these cases were not extensively studied and the degree to which they truly resembled the human disease is diYcult to judge. The canine cases typically presented as young adults with histories of progressive ataxia and hind-limb weakness, and on autopsy showed diVuse deposition of Rosenthal Wbers in association with mild leukodystrophy. There was no frontal predominance or seizures. It is interesting that the earliest of these reports concerned two Labrador retriever littermates (McGrath, 1980), which is not the typical pattern of inheritance seen in human Alexander disease (nearly all of which have been sporadic).

GENETIC BASIS OF ALEXANDER DISEASE In 1998 we reported the unexpected Wnding that transgenic mice designed to constitutively over-express wild-type GFAP developed a lethal phenotype associated with formation of Rosenthal Wbers throughout the central nervous system (Messing et al., 1998). In the highest-expressing lines, GFAP levels were elevated 15- to 20-fold over controls. The mice died within a few weeks after birth with no overt symptoms other than failure to gain weight despite adequate nursing. Of particular interest was the Wnding that the murine Rosenthal Wbers appeared indistinguishable morphologically and biochemically from human Rosenthal Wbers, which led to the conclusion that formation of these complex inclusions could be initiated by a primary change in the expression of GFAP. Furthermore, given the increasing recognition of gene duplications as a mechanism in human neurological disease (Lupski et al., 1991; Sistermans et al., 1998), we proposed that GFAP should be evaluated as a candidate gene for Alexander disease. Subsequent studies conWrmed that GFAP indeed was the gene responsible for most cases of Alexander disease, but in the form of heterozygous missense mutations rather than gene duplications. Brenner et al. (2001) reported the identiWcation of heterozygous missense mutations within the coding region of GFAP in 12 out of 13 cases of Alexander disease in whom the diagnosis had been conWrmed by pathology. These mutations predicted nonconservative changes in amino acids, in each case involving an arginine (R79C, R79H, R239C, R239H, R258P, and R416W). In 7 of the original 12 cases analyzed, DNA samples were available from both parents of the aVected child, and the parental samples were always normal. These Wndings are consistent with the hypothesis that Alexander disease usually results from dominant de novo mutations in GFAP. These results have now been conWrmed in a larger group of patients, and current data show that ~95% of cases of infantile Alexander disease, and at least some of the juvenile and adult cases, are associated with mutations in GFAP (Li et al., 2002). The list of known mutations has grown and now includes 20 codons, many involving amino acids other than arginine. The GFAP mutations show extensive but incomplete homologies to disease-causing mutations in other intermediate Wlaments, as discussed in more detail by Li et al. (2002). A tabulation of published GFAP mutations is being maintained at the University of Wisconsin-Madison (www.waisman.wisc.edu/alexander) and the Human Intermediate Filament Mutation Database (www.interWl.org).

DOES ALEXANDER DISEASE REPRESENT GAIN OR LOSS OF FUNCTION IN GFAP? The Wnding that most cases of Alexander disease result from heterozygous missense mutations in GFAP places it in the family of genetic disorders of intermediate Wlaments.

ASTROCYTES AND GFAP

Among these disorders is the epidermolysis bullosa complex, also initially identiWed through analysis of mouse models that closely resembled the human phenotype and that pointed the way toward genetic analysis of the keratin genes. These other intermediate Wlament disorders are typically dominant, also involving heterozygous missense mutations. Such dominant eVects are plausible when one considers that mature Wlaments are assemblies of multiple polypeptides, dysfunction of any of which might aVect the whole. However, since the mouse phenocopies of the other human intermediate Wlament disorders are either dominant negatives or knockouts of the respective genes, the dominant mutations identiWed in the human patients are generally viewed as acting to induce loss of function (Fuchs and Cleveland, 1998). In contrast, GFAP Wlaments are present in Alexander disease patients, and GFAP null mice are fully viable and their pathology does not resemble Alexander disease (Gomi et al., 1995; Liedtke et al., 1996; McCall et al., 1996; Pekny et al., 1995). Thus the GFAP mutations in Alexander disease do not appear to be acting by reducing or eliminating normal GFAP function, but by producing a new, deleterious, activity. Further discussion on this issue and comparison with other intermediate Wlament diseases can be found in the review by Li et al. (2002). The following sections discuss the consequences of deleting or duplicating the GFAP gene in the mouse, with the goal of highlighting the strengths and limitations of these current models of Alexander disease and suggesting directions for future research.

ASTROCYTES AND GFAP Mature astrocytes have a complex morphology that includes extension of processes to contact multiple other cell types and structures, including formation of the glial limitans at the pial surface, glial end-feet ensheathing blood vessel walls or synapses between neurons, and contacts with axonal nodes of Ranvier. At each of these contacts, astrocytes are thought to play key roles, such as inducing the blood-brain barrier in endothelium and modulating neuronal function through regulation of the ionic or neurotransmitter content of the extracellular space. In addition, astrocytes produce numerous growth factors that act on oligodendrocytes and their precursors. Compromise of any or all of these functions could contribute to the phenotype of Alexander disease, as nothing is known about the pathogenetic mechanisms actually at work in human patients. GFAP is the major structural protein in astrocytes, but their progenitor neural stem cells contain the nestin intermediate Wlament and the immediate precursors of astrocytes (as well as many mature astrocytes) express vimentin. The activation of GFAP synthesis as astrocytes mature is considered a key step in their diVerentiation (Dahl, 1981; Bovolenta et al., 1984). Studies in cultured astrocytes and cell lines of forced over-expression of GFAP or its antisense inhibition suggested that GFAP directly controls process outgrowth (Hatten et al., 1991; Rutka and Smith, 1993; Weinstein et al., 1991). In response to a wide variety of insults to the CNS, astrocytes undergo the process of reactive gliosis, which consists primarily of hypertrophy accompanied by dramatic changes in gene expression (Eddleston and Mucke, 1993). Increased levels of GFAP is perhaps the most characteristic change occurring in reactive astrocytes (Amaducci et al., 1981; Bignami and Dahl, 1976; Eng and Lee, 1993; Mathewson and Berry, 1985), though there is also renewed expression of vimentin (Dahl et al., 1981; Pixley and de Vellis, 1984) and nestin (Clarke et al., 1994; Frise´n et al., 1995). Nestin, vimentin, and GFAP can coassemble in the same Wlament (Eliasson et al., 1999).

GFAP-NULL MICE One way to study the role of GFAP in astrocytes is to generate mice that are deWcient in GFAP. Given the apparent requirement for GFAP for process formation indicated by the cell culture experiments, and the likely essential requirement for these processes in brain

1117

1118

46. MODELS OF ALEXANDER DISEASE

development and function, it was presumed that GFAP null mice would be embryonic lethals, or at least severely compromised. Consistent with this supposition, no spontaneous mutants of the murine Gfap gene were known. Using standard techniques of gene targeting in embryonic stem cells, four groups independently generated GFAP-null mice (Gomi et al., 1995; Liedtke et al., 1996; McCall et al., 1996; Pekny et al., 1995). The consistent result was the surprising Wnding that GFAP-null mice are viable, with seemingly normal lifespans, reproduction, and gross motor behavior. However, alterations in neuronal synaptic function were identiWed, with impaired long-term depression in the cerebellum (Shibuki et al., 1996) and enhanced long-term potentiation in the hippocampus (McCall et al., 1996). Ultrastructural studies in optic nerve and spinal cord suggested minor changes in astrocyte morphology, with shortening and thinning of astrocyte processes (Liedtke et al., 1996; McCall et al., 1996). Subsequent studies of dye-Wlled cells also supports this conclusion (Anderova´ et al., 2001). Studies in cell culture suggest that GFAP-null astrocytes may alter their expression of several components of the extracellular matrix (Menet et al., 2001), but whether this occurs in vivo is less clear (Wang et al., 1997). There is some evidence for alterations in the blood brain barrier derived from cell culture models (Pekny et al., 1998) and to a limited extent from studies in vivo (Liedtke et al., 1996). Initial injury trials with GFAP-null mice also failed to Wnd signiWcant eVects, with the mice showing apparently normal responses to scrapie infection (Gomi et al., 1995; Tatzelt et al., 1996), kainic acid (Gomi et al., 1995), or stab wounds (Pekny et al., 1995). However, more recent studies have identiWed increased susceptibility to blunt head injury (Nawashiro et al., 1998) or ischemia (Nawashiro et al., 1998, 2000; Tanaka et al., 2002), and impaired swelling following exposure to hyptonic stress in vivo (Anderova´ et al., 2001), though the mechanisms are not known. When double mutant mice are made that are deWcient in both GFAP and vimentin, there is reduced gliosis and increased hemorrhage following traumatic injury (Pekny et al., 1999). Since there is no up-regulation of vimentin or nestin in the GFAP null mice (Gomi et al., 1995; McCall et al., 1996), this indicates that in the normal astrocyte either GFAP or vimentin is suYcient to maintain a subset of functions. Two studies directly link GFAP deWciency with abnormalities of myelination. Liedtke et al. (1996) reported that approximately half of the GFAP-null mice in their colony developed an adult-onset degeneration of myelin, with resulting hydrocephalus. Grossly visible changes in cerebral white matter did not become apparent until mice were 18 months and older, and some white matter tracts such as optic nerve were only minimally aVected, if at all. The mechanism of this myelin degeneration has not been determined, and the possible contribution of background genetics (for instance, corpus callosum defects in 129 strain mice) has not been excluded (Wahlsten, 1982). A second report from this same group determined that GFAP-null mice had increased susceptibility to experimental allergic encephalitis, with increased clinical scores and pathology (Liedtke et al., 1998). However, inXammation is thought to play little role in the pathogenesis of Alexander disease (see chapter 36 by Messing and Goldman in this volume). Overall, the relevance of GFAP-null mice to Alexander disease remains uncertain. Such mice do not develop a severe leukodystrophy and have no Rosenthal Wbers, two primary characteristics of the human disease. In addition, the GFAP-null mice do not exhibit developmental delay, spasticity, seizures, ataxia, or paralysis. Nevertheless, the changes in neuronal physiology observed in GFAP-null mice (Shibuki et al., 1996; McCall et al., 1996), and potential eVects on myelination, leaves open the possibility that at least some aspects of Alexander disease may reXect loss of function of GFAP. Furthermore, it is well known that dominant negative mutations may produce more severe phenotypes than corresponding null mutations, as recently demonstrated with keratin 10 (Reichelt et al., 2001).

GFAP TRANSGENICS As a complementary approach to gene knockouts for studying the role of GFAP in astrocytes, we sought to force constitutive over-expression of GFAP in otherwise normal astro-

GFAP TRANSGENICS

cytes to test whether GFAP up-regulation was suYcient to induce other aspects of reactive gliosis. Our rationale was that adding GFAP transgenes would elevate GFAP levels as a primary change in astrocytes, without the confounding eVects of injuries or other treatments that are typically used to induce this response. For this purpose we used a genomic human GFAP (hGFAP) clone that we had previously isolated for transcriptional studies. The surprising result was that astrocytes in these transgenics not only became hypertrophic, but also formed Rosenthal Wbers (Messing et al., 1998), the Wrst time such inclusions had been produced reliably in an in vivo model (Fig. 46.1). Like the Rosenthal Wbers in Alexander disease, those in the mice contained GFAP and the small heat shock proteins aB-crystallin and HSP27 (the mouse homolog of HSP25), each of whose expression was also up-regulated. In addition, when astrocytes were cultured from the hGFAP transgenics in isolation from neurons, Rosenthal Wbers still formed, indicating that this was a cell-autonomous phenotype (Eng et al., 1998). The Rosenthal Wbers in the hGFAP transgenics are distinct from the astrocytic inclusions associated with superoxide dismutase (SOD) mutations in transgenic mice (Bruijn et al., 1997) and in humans with amyotrophic lateral sclerosis (Kato et al., 1996), as these contain ubiquitin and SOD1 but no GFAP (Kato et al., 1997). Previously there had been considerable debate over what might precipitate formation of Rosenthal Wbers, and whether the astrocyte pathology had a central role in the disease or was merely a secondary response. The hGFAP transgenic mice show that Rosenthal Wbers can form as the result of a primary change in just one component, GFAP, and suggest that astrocyte dysfunction may have an active rather than passive role in the human disease. Although the existing lines of GFAP transgenic mice are good models for studying the etiology of Rosenthal Wbers, they are imperfect models of Alexander disease as they do not display several other features of the disease. Mice of the highest expressing lines did die at an early age, but without overt symptoms such as seizures. The lines now being maintained express the hGFAP transgene at lower levels, and while they also accumulate Rosenthal Wbers, their only overt phenotype is a small body size compared to littermates (Messing et al., manuscript in preparation). Furthermore, none of the hGFAP transgenics develop hydrocephalus or detectable white matter changes, leaving unresolved the central question of why Alexander disease in humans is often a leukodystrophy. Finally, in light of the genetic studies of human Alexander disease patients noted earlier, the GFAP-over-expressing mice do not model the true genetic lesions that have been identiWed in the human patients; the transgenics were made by adding copies of a wild-type human GFAP gene, whereas heterozygous missense mutations have been found in the human disease. The hGFAP transgenic mice might prove to be a model for GFAP gene duplication. It is not yet known whether such duplication might underlie any of the several Alexander disease cases for which no GFAP coding mutation has been found, or if it might be found in association with a diVerent human disorder. As noted earlier, the hGFAP transgenics also model the up-regulation of stress proteins found in human Alexander disease and so can serve for the further characterization of this response. In particular, aB-crystallin is a major component of the Rosenthal Wbers, and one explanation for its up-regulation is that it is part of a failed protective response of the astrocyte to aberrant expression of GFAP. The aB-crystallin is thought to function as a chaperone and promotes disassembly of GFAP Wlaments in cell-free assays (Nicholl and Quinlan, 1994). Recent studies by Koyama and Goldman show that transfecting aBcrystallin into cultured cells counteracts the formation of aggregates produced by transfection of GFAP alone (Koyama and Goldman, 1999). These Wndings suggest that Alexander disease might be ameliorated were it possible to increase the level of aB-crystallin above that which normally occurs in the disease. Johnson and colleagues recently generated a novel strain of transgenic mice that express the human placental alkaline phosphatase reporter gene under the control of a minimal antioxidant response element promoter (Johnson et al., 2002). These mice thus oVer an eYcient means of monitoring oxidative stress, and crossing the ARE-hPAP mice with the hGFAP transgenics induces marked up-regulation of the hPAP reporter gene (Fig. 46.2). Further studies of such double transgenics will help deWne the relationship between oxidative stress and inclusion body formation.

1119

1120

46. MODELS OF ALEXANDER DISEASE

FIGURE 46.1 Rosenthal Wbers in the cerebellum of an hGFAP transgenic mouse. Phloxine tartrazine stain, paraYn section.

ARE-hPAP

ARE-hPAP/hGFAP double Tg

FIGURE 46.2 Extensive up-regulation of the hPAP reporter gene in ARE-hPAP/hGFAP double transgenic mice compared to single ARE-hPAP transgenic controls, tested at 3 months of age.

NEW MOUSE MODELS

NEW MOUSE MODELS A GFAP mouse model recently produced by the Itohara laboratory (Takemura et al., (2002) bears on the question of whether Rosenthal Wber formation in the hGFAP transgenics might arise from sequence incompatibility caused by the human GFAP, rather than from simple over-expression. Although there is substantial homology between human and mouse GFAP (91% identity and 95% similarity at the amino acid level) (Brenner et al., 1990), it was still possible that the sequence diVerences could have led to the production of incompatible polypeptides with abnormalities of either assembly, disassembly, or degradation. In the Itohara mice, the entire head domain of the endogenous Gfap gene, where most of the species diVerences reside, was swapped with that of the human gene. Heterozygous mice therefore contain a normal gene copy number, carrying one allele of the normal mouse gene, and a second humanized allele with the human amino terminal end. Nevertheless, these mice developed no Rosenthal Wbers or any other pathology. As noted earlier, it is possible that the lack of leukodystrophy, seizures, and hydrocephalus in the GFAP over-expressing mice reXects a basic diVerence in genetic disease mechanism from that in Alexander disease. To try to develop a more accurate model, current eVorts are aimed at generating mice engineered with targeted point mutations analogous to those seen in the human Alexander disease. In particular, mutations of just two amino acids (R79 and R239) account for nearly half of all reported cases, and these are the Wrst mutations being modeled in mice. Preliminary observations indicate that such mice do indeed form Rosenthal Wbers, oVering formal proof that these mutations are causative for this aspect of the disease (Hagemann et al., unpublished observations). It should soon be known whether these mice exhibit the full spectrum of lesions and symptoms observed in Alexander disease.

SUMMARY Previous transgenic results show that mice are capable of building the biochemically complex Rosenthal Wbers within a short period of time in vivo. These transgenic lines also showed a signiWcant phenotype, either early death or low body weight, indicating that they are sensitive to at least some forms of astrocyte dysfunction. The existing models are thus useful for examining several aspects of Alexander disease. However, since they do not display most of the common symptoms and pathology of the disorder, mice are now being produced with targeted point mutations in their endogenous Gfap gene. It is hoped that these will provide more complete models, leading to continued progress in unraveling the pathogenesis of Alexander disease and facilitating the development and testing of strategies for potential therapy.

Acknowledgments We thank our collaborators, especially James E. Goldman, who have contributed to these studies. This work was supported by grants from the National Institutes of Health (NS-22475, NS-41803).

References Alexander, W. S. (1949). Progressive Wbrinoid degeneration of Wbrillary astrocytes associated with mental retardation in a hydrocephalic infant. Brain 72, 373–381. Amaducci, L., Forno, K. I., and Eng, L. F. (1981). Glial Wbrillary acidic protein in cryogenic lesions of the rat brain. Neurosci. Lett. 21, 27–32. Anderova´, M., Kubinova´, S., Mazel, T., Chva´tal, A., Eliasson, C., Pekny, M., and Sykova´, E. (2001). EVect of elevated Kþ, hypotonic stress, and cortical spreading depression on astrocyte swelling in GFAP-deWcient mice. Glia 35, 189–203.

1121

1122

46. MODELS OF ALEXANDER DISEASE

Bignami, A., and Dahl, D. (1976). The astroglial response to stabbing. ImmunoXuorescence studies with antibodies to astrocyte-speciWc protein (GFA) in mammalian and submammalian species. Neuropathol. Appl. Neurobiol. 2, 99–110. Borrett, D., and Becker, L. E. (1985). Alexander’s disease. A disease of astrocytes. Brain 108, 367–385. Bovolenta, P., Liem, R. K. H., and Mason, C. A. (1984). Development of cerebellar astroglia: Transitions in form and cytoskeletal content. Dev. Biol. 102, 248–259. Brenner, M., Johnson, A. B., BoespXug-Tanguy, O., Rodriguez, D., Goldman, J. E., and Messing, A. (2001). Mutations in GFAP, encoding glial Wbrillary acidic protein, are associated with Alexander disease. Nature Genet. 27, 117–120. Brenner, M., Lampel, K., Nakatani, Y., Mill, J., Banner, C., Mearow, K., Dohadwala, M., Lipsky, R., and Freese, E. (1990). Characterization of human cDNA and genomic clones for glial Wbrillary acidic protein. Mol. Brain Res. 7, 277–286. Bruijn, L. I., Becher, M. W., Lee, M. K., Anderson, K. L., Jenkins, N. A., Copeland, N. G., Sisodia, S. S., Rothstein, J. D., Borchelt, D. R., Price, D. L., and Cleveland, D. W. (1997). ALS-linked SOD1 mutant G85R mediates damage to astrocytes and promotes rapidly progressive disease with SOD1-containing inclusions. Neuron 18, 327–338. Clarke, S. R., Shetty, A. K., Bradley, J. L., and Turner, D. A. (1994). Reactive astrocytes express the embryonic intermediate neuroWlament nestin. NeuroReport 5, 1885–1888. Cox, N. R., Kwapien, R. P., Sorjonen, D. C., and Braund, K. G. (1986). Myeloencephalopathy resembling Alexander’s disease in a Scottish terrier dog. Acta Neuropathol. (Berl.) 71, 163–166. Dahl, D. (1981). The vimentin-GFA protein transition in rat neuroglia cytoskeleton occurs at the time of myelination. J. Neurosci. Res. 6, 741–748. Dahl, D., Bignami, A., Weber, K., and Osborn, M. (1981). Filament proteins in rat optic nerves undergoing ˚ Wlament protein, in normal and Wallerian degeneration: Localization of vimentin, the Wbroblastic 100-A reactive astrocytes. Exp. Neurol. 73, 496–506. Eddleston, M., and Mucke, L. (1993). Molecular proWle of reactive astrocytes–Implications for their role in neurologic disease. Neuroscience 54, 15–36. Eliasson, C., Sahlgren, C., Berthold, C. H., Stakeberg, J., Celis, J. E., Betsholtz, C., Eriksson, J. E., and Pekny, M. (1999). Intermediate Wlament protein partnership in astrocytes. J. Biol. Chem. 274, 23996–24006. Eng, L. F., and Lee, Y. L. (1993). Intermediate Wlaments in astrocytes. In ‘‘Neuroglial Cells’’ (B. R. Ransom and H. Kettenmann, eds.). Oxford University Press, New York. Eng, L. F., Lee, Y. L., Kwan, H., Brenner, M., and Messing, A. (1998). Astrocytes cultured from transgenic mice carrying the added human glial Wbrillary acidic protein gene contain Rosenthal Wbers. J. Neurosci. Res. 53, 353–360. Fankhauser, R., Fatzer, R., Bestetti, G., Deruaz, J. P., and Perentes, E. (1980). Encephalopathy with Rosenthal Wbre formation in a sheep. Acta Neuropathol. (Berl.) 50, 57–60. Frise´n, J., Johansson, C. B., To¨ro¨k, C., Risling, M., and Lendahl, U. (1995). Rapid, widespread, and longlasting induction of nestin contributes to the generation of glial scar tissue after CNS injury. J. Cell Biol. 131, 453–464. Fuchs, E., and Cleveland, D. W. (1998). A structural scaVolding of intermediate Wlaments in health and disease. Science 279, 514–519. Gomi, H., Yokoyama, T., Fujimoto, K., Ideka, T., Katoh, A., Itoh, T., and Itohara, S. (1995). Mice devoid of the glial Wbrillary acidic protein develop normally and are susceptible to scrapie prions. Neuron 14, 29–41. Hatten, M. E., Liem, R. K. H., Shelanski, M. L., and Mason, C. A. (1991). Astroglia in CNS injury. Glia 4, 233–243. Head, M. W., and Goldman, J. E. (2000). Small heat shock proteins, the cytoskeleton, and inclusion body formation. Neuropathology & Applied Neurobiology 26, 304–312. Herndon, R. M. (1999). Is Alexander’s disease a nosologic entity or a common pathologic pattern of diverse etiology? Journal of Child Neurology 14, 275–276. Herndon, R. M., Rubinstein, L. J., Freeman, J. M., and Mathieson, G. (1970). Light and electron microscopic observations on Rosenthal Wbers in Alexander’s disease and in multiple sclerosis. J. Neuropathol. Exp. Neurol. 29, 524–551. Iwaki, T., Kume-Iwaki, A., Liem, R. K. H., and Goldman, J. E. (1989). aB-Crystallin is expressed in nonlenticular tissues and accumulates in Alexander disease brain. Cell 57, 71–78. Johnson, D. A., Andrews, G. K., Xu, W., and Johnson, J. A. (2002). Activation of the antioxidant responsive element in primary cortical neuronal cultures derived from transgenic reporter mice. J. Neurochem. 81, 1233–1241. Kato, S., Hayashi, H., Nakashima, K., Nanba, E., Kato, M., Hirano, A., Nakano, I., Asayama, K., and Ohama, E. (1997). Pathological characterization of astrocytic hyaline inclusions in familial amyotrophic lateral sclerosis. Am. J. Pathol 151, 611–620. Kato, S., Shimoda, M., Watanabe, Y., Nakashima, K., Takahashi, K., and Ohama, E. (1996). Familial amyotrophic lateral sclerosis with a two base pair deletion in superoxide dismutase 1 gene: Multisystem degeneration with intracytoplasmic hyaline inclusions in astrocytes. J. Neuropathol. Exp. Neurol. 55, 1089–1101. Koyama, Y., and Goldman, J. E. (1999). Formation of GFAP cytoplasmic inclusions in astrocytes and their disaggregation by aB-crystallin. Am. J. Pathol. 154, 1563–1572. Li, R., Messing, A., Goldman, J. E., and Brenner, M. (2002). GFAP mutations in Alexander disease. Int. J. Dev. Neurosci. 20, 259–268.

SUMMARY

Liedtke, W., Edelmann, W., Bieri, P. L., Chiu, F. C., Cowan, N. J., Kucherlapati, R., and Raine, C. S. (1996). GFAP is necessary for the integrity of CNS white matter architecture and long-term maintenance of myelination. Neuron 17, 607–615. Liedtke, W., Edelmann, W., Chiu, F.-C., Kucherlapati, R., and Raine, C. S. (1998). Experimental autoimmune encephalomyelitis in mice lacking glial Wbrillary acidic protein is characterized by a more severe clinical course and an inWltrative central nervous system lesion. Am. J. Pathol. 152, 251–259. Lupski, J. R., Montes de Oca-Luna, R., Slaugenhaupt, S., Pentao, L., Guzzetta, V., Trask, B. J., SaucedoCardenas, O., Barker, D. F., Killian, J. M., Garcia, C. A., Chakravarti, A., and Patel, P. I. (1991). DNA duplication associated with Charcot-Marie-Tooth disease type 1A. Cell 66, 219–232. Mathewson, A. J., and Berry, M. (1985). Observations on the astrocyte response to a cerebral stab wound in adult rats. Brain Res. 327, 61–69. McCall, M. A., Gregg, R. G., Behringer, R. R., Brenner, M., Delaney, C. L., Galbreath, E. J., Zhang, C. L., Pearce, R. A., Chiu, S. Y., and Messing, A. (1996). Targeted deletion in astrocyte intermediate Wlament (Gfap) alters neuronal physiology. Proc. Natl. Acad. Sci. USA 93, 6361–6366. McGrath, J. T. (1980). Fibrinoid leukodystrophy (Alexander’s disease). In ‘‘Spontaneous Animal Models of Human Disease’’ (E. J. Andrews, B. C. Ward, and H. N. Altman, eds.), Vol. 2., pp. 147–148. Academic Press, New York. Menet, V., Ribotta, M. G. Y., Chauvet, N., Drain, M. J., Lannoy, J., Colucci-Guyon, E., and Privat, A. (2001). Inactivation of the glial Wbrillary acidic protein gene, but not that of vimentin, improves neuronal survival and neurite growth by modifying adhesion molecule expression. J. Neurosci. 21, 6147–6158. Messing, A., Head, M. W., Galles, K., Galbreath, E. J., Goldman, J. E., and Brenner, M. (1998). Fatal encephalopathy with astrocyte inclusions in GFAP transgenic mice. Am. J. Pathol. 152, 391–398. Nawashiro, H., Brenner, M., Fukui, S., Shima, K., and Hallenbeck, J. M. (2000). High susceptibility to cerebral ischemia in GFAP-null mice. Journal of Cerebral Blood Flow & Metabolism 20, 1040–1044. Nawashiro, H., Messing, A., Azzam, N., and Brenner, M. (1998). Mice lacking glial Wbrillary acidic protein are hypersensitive to traumatic cerebrospinal injury. NeuroReport 9, 1691–1696. Ni, Q., Johns, G. S., Manepalli, A., Martin, D. S., and Geller, T. J. (2002). Infantile Alexander’s disease: Serial neuroradiologic Wndings. J. Child Neurol. 17, 463–466. Nicholl, I. D., and Quinlan, R. A. (1994). Chaperone activity of a-crystallins modulates intermediate Wlament assembly. EMBO J. 13, 945–953. Pekny, M., Johansson, C. B., Eliasson, C., Stakeberg, J., Wallen, A., Perlmann, T., Lendahl, U., Betsholtz, C., Berthold, C. H., and Frisen, J. (1999). Abnormal reaction to central nervous system injury in mice lacking glial Wbrillary acidic protein and vimentin. J. Cell Biol. 145, 503–514. Pekny, M., Leve´en, P., Pekna, M., Eliasson, C., Berthold, C.-H., Westermark, B., and Betsholtz, C. (1995). Mice lacking glial Wbrillary acidic protein display astrocytes devoid of intermediate Wlaments but develop and reproduce normally. EMBO J. 14, 1590–1598. Pekny, M., Stanness, K. A., Eliasson, C., Betsholtz, C., and Janigro, D. (1998). Impaired induction of blood-brain barrier properties in aortic endothelial cells by astrocytes from GFAP-deWcient mice. Glia 22, 390–400. Pixley, S. K. R., and de Vellis, J. (1984). Transition between immature radial glia and mature astrocytes studied with a monoclonal antibody to vimentin. Dev. Brain Res. 15, 201–209. Reichelt, J., Bussow, H., Grund, C., and Magin, T. M. (2001). Formation of a normal epidermis supported by increased stability of keratins 5 and 14 in keratin 10 null mice. Molecular Biology of the Cell 12, 1557–1568. Richardson, J. A., Tang, K., and Burns, D. K. (1991). Myeloencephalopathy with Rosenthal Wber formation in a miniature poodle. Vet. Pathol. 28, 536–538. ¨ ber eine eigenthu¨mliche, mit syringomyelie complicirte geschwulst des ru¨ckenmarks. Rosenthal, W. (1898). U Bietr. Pathol. Anat. 23, 111–143. Russo, L. S., Jr., Aron, A., and Anderson, P. J. (1976). Alexander’s disease: A report and reappraisal. Neurology 26, 607–614. Rutka, J. T., and Smith, S. L. (1993). Transfection of human astrocytoma cells with glial Wbrillary acidic protein complementary DNA: Analysis of expression, proliferation, and tumorigenicity. Cancer Res. 53, 3624–3631. Sherwin, R. M., and Berthrong, M. (1970). Alexander’s disease with sudanophilic leukodystrophy. Archives of Pathology & Laboratory Medicine 89, 321–328. Shibuki, K., Gomi, H., Chen, L., Bao, S. W., Kim, J. S. K., Wakatsuki, H., Fujisaki, T., Fujimoto, J., Katoh, A., Ikeda, T., Chen, C., Thompson, R. F., and Itohara, S. (1996). DeWcient cerebellar long-term depression, impaired eyeblink conditioning, and normal motor coordination in GFAP mutant mice. Neuron 16, 587–599. Sistermans, E. A., Decoo, R. F. M., Dewijs, I. J., and Vanoost, B. A. (1998). Duplication of the proteolipid protein gene is the major cause of Pelizaeus-Merzbacher disease. Neurology 50, 1749–1754. Takemura, M., Nishiyama, H., and Itohara, S. (2002). Distribution of phosphorylated glial Wbrillary acidic protein in the mouse central nervous system. Genes to Cells 7, 295–307. Tanaka, H., Katoh, A., Oguro, K., Shimazaki, K., Gomi, H., Itohara, S., Masuzawa, T., and Kawai, N. (2002). Disturbance of hippocampal long-term potentiation after transient ischemia in GFAP deWcient mice. J. Neurosci. Res. 67, 11–20. Tatzelt, J., Maeda, N., Pekny, M., Yang, S. L., Betsholtz, C., Eliasson, C., Cayetano, J., Camerino, A. P., DeArmond, S. J., and Prusiner, S. B. (1996). Scrapie in mice deWcient in apolipoprotein E or glial Wbrillary acidic protein. Neurology 47, 449–453. Wahlsten, D. (1982). DeWciency of corpus callosum varies with strain and supplier of the mice. Brain Res. 239, 329–347.

1123

1124

46. MODELS OF ALEXANDER DISEASE

Wang, X., Messing, A., and David, S. (1997). Axonal and non-neuronal cell responses to spinal cord injury in mice lacking glial Wbrillary acidic protein. Exp. Neurol. 148, 568–576. Weinstein, D. E., Shelanski, M. L., and Liem, R. K. H. (1991). Suppression by antisense mRNA demonstrates a requirement for the glial Wbrillary acidic protein in the formation of stable astrocytic processes in response to neurons. J. Cell Biol. 112, 1205–1213. Weissenbo¨ck, H., Obermaier, G., and Dahme, E. (1996). Alexander’s disease in a Bernese mountain dog. Acta Neuropathol. (Berl.) 91, 200–204.

C H A P T E R

47 Models of Pelizeaus-Merzbacher-Disease

au1

Klaus-Armin Nave and Ian R. Griffiths

INTRODUCTION Among inherited neurological disorders, Pelizaeus-Merzbacher disease (PMD; McKusick #312080) is a leukodystrophy well suited for modelling in rodent mutants. Myelin-deWcient mice and rats with altered expression of the X chromosome-linked gene for proteolipid protein (Plp/Dm20) have been investigated over many years and reXect a wide spectrum of phenotypic disease expression, that is similar in many respects to human PMD (see Chapter 37) and its milder form, spastic paraplegia type 2 (SPG-2; McKusick #312920), deWned by mutation of the human PLP/DM20 gene. The virtual identity of the mouse and human myelin protein in amino acid sequence may explain, at least in part, the similarity of loss-of-function and gain-of-function eVects that characterize mutant oligodendrocytes in the CNS of these species. The detailed comparison of several natural and transgenic Plp mutants has helped to dissect the molecular pathology of PMD at the subcellular level. Several recent reviews have covered the genetics of PMD as a human leukodystrophy (Garbern et al., 1999; Hodes et al., 1994; Nave and BoespXug-Tanguy, 1996; Seitelberger et al., 1996; Woodward and Malcolm, 1999, 2001; Yool et al., 2000) and have drawn attention to the homologous animal models (Bradl and Linington, 1996; GriYths, 1996; GriYths et al., 1995a, 1998a; Nadon and West, 1998; Nave and BoespXug-Tanguy, 1996; Scherer, 1997; SkoV, 1995; Vela et al., 1998; Werner et al., 1998). Recent observations in these mice have revealed an unexpected role of PLP-expressing oligodendrocytes in maintaining the long-term integrity of myelinated CNS axons. It is hoped that animal models of PMD will aid in the development of future treatment strategies (GriYths, 1996; GriYths et al., 1995b, 1998a).

au2

au3

PROTEOLIPID PROTEIN IN MOUSE AND MAN Proteolipid protein (PLP) is one of the major structural proteins of CNS myelin. As reviewed in more detail in Chapter 16, PLP is an integral membrane protein of 276 amino acids and a molecular weight of 30kD (Milner et al., 1985; StoVel et al., 1984), comprising about 50% of total myelin protein (Lees and BrostoV, 1984). By alternative RNA splicing, a second protein isoform (termed ‘‘DM20’’) is generated that lacks 35 residues from an internal loop region of the polytopic membrane protein (Nave et al., 1987). Unexpectedly for a structural protein of myelin, the primary sequence of PLP and DM20 is 100% conserved between rodents (mouse, rat) and humans (Diehl et al., 1986; Hudson et al., 1987; Milner et al., 1985; Morello et al. 1986), and very few amino acid diVerences are

Myelin Biology and Disorders, Volume 2

1125

Copyright 2004, Elsevier (USA). All rights reserved.

au4

1126

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

found in other mammalian species. This suggests that PLP engages with additional proteins, and possibly with itself, in multiple space-restricted protein interactions. Although no bona Wde protein structural data are available, a widely accepted model places PLP/DM20 into the family of ‘‘4-helix bundle’’ proteins with four transmembrane domains and intracellular protrusion of the amino- and carboxyl-termini (Popot et al., 1991; Wahle and StoVel, 1998; Weimbs and StoVel, 1992). The identiWcation of acylated cysteine residues has conWrmed the orientation of the amino and carboxyl termini. Presumably, these acyl residues anchor the cytosolic part of the polypeptide chain close to the membrane. Posttranslational palmitylation also explains the unusual hydrophobicity of PLP and DM20 that has hampered its biochemical analysis (Laursen et al., 1984; StoVel et al., 1984). Recently, PLP has been identiWed as a component of CHAPS-insoluble membrane lipid rafts in oligodendrocytes, and a direct binding of cholesterol to PLP has been demonstrated by UV crosslinking (Simons et al., 2000). DisulWde bonds stabilize the extracellular loop regions and overall topology (Weimbs et al., 1992, 1994). There is indirect evidence that PLP/DM20 assembles into an oligomeric complex prior to reaching the myelin compartment, but the stoichiometry of this hypothetical PLP oligomer is not known (Jung et al., 1996; Jung and Nave, unpublished results; McLaughlin et al., 2002). A functional diVerence between PLP and DM20 is not known. However, immunostaining of oligodendrocytes with isoform-speciWc antibodies has suggested that the ‘‘PLP-speciWc’’ region harbors a signal for the eYcient sorting of PLP into the myelin compartment (Anderson et al., 1997; McLaughlin et al., 2002; Trapp et al., 1997). The rodent Plp gene is predominantly expressed by myelin-forming oligodendrocytes, but cells in this lineage transcribe the Plp gene well before myelination starts, already in a subset of bipotential neuron/glia precursor cells (Dickinson et al., 1996; Goebbels, Zalc, and Nave, unpublished data; Genoud et al., 2002; Spassky et al., 1998; Yu et al., 1994). While this has suggested a protein function in earliest oligodendrocyte diVerentiation, there has been little supportive evidence from the phenotypic analysis of the available Plp mutants. Recent evidence suggests that a subset of CNS neurons, including brain stem motor neurons, express a previously unrecognized vesicular isoform of PLP (sr-PLP, sr-DM20 for ‘‘soma-restricted’’) that diVers from myelin PLP by an N-terminal extension of 12 residues and its absence from myelin (Bongarzone et al., 1999, 2001a, 2001b). So far, the srPLP encoding exon (located within the Wrst intron) has been found only in mice. Its possible biological function remains to be determined. Rodent Schwann cells also express the Plp gene at a low level, with a predominating DM20 isoform that is not found in peripheral myelin (GriYths et al., 1995a; Pucket et al., 1987). The mechanism of the intracellular retention of PLP/DM20 in Schwann cells is unknown, but in PLP-overexpressing mice the protein becomes a major component of compact myelin (Anderson et al., 1997). Human PLP mutations have been associated with a yet unexplained peripheral neuropathy suggesting indeed a role for PLP also in the peripheral nervous system (Garbern et al., 1997; GriYths et al., 1995a; Shy et al., 2003). Mechanistically, a normal cellular function of PLP has been diYcult to deWne, and the complex phenotype of Plp mutant mice was not immediately helpful. Clearly, the incorporation of PLP into compact myelin contributes to the stable association of adjacent myelin membranes that form the ‘‘intraperiod line’’ of compacted myelin (Boison et al., 1995; Klugmann et al., 1997; Rosenbluth et al., 1996). As detailed later, it appears that PLP also contributes to the integrity of myelinated axons, a function that could be related to its association with membrane lipid rafts and possibly the transport of raft-associated signaling molecules. The unknown cellular function of PLP/DM20 is likely related to that of two PLP-related tetraspan proteins, termed M6A and M6B (Kitagawa et al., 1993; Werner et al., 2001; Yan et al., 1993). These proteolipids are expressed in both neurons and glial cells, but the mutation of either gene has no obvious developmental defect in the mouse (Werner et al., unpublished results). Structurally, both M6A and M6B correspond to DM20, and additional M6B isoforms are speciWed by extensive alternative mRNA splicing (Werner et al., 2001). PLP and M6B may have an overlapping function in developing oligodendrocytes that is only obvious in the corresponding double mutant mice (Werner et al., unpublished data)

V. ANIMAL MODELS OF HUMAN DISEASE

au5

1127

THE CLINICAL SPECTRUM OF PLP GENE MUTATIONS

The PLP gene has been linked to the X chromosome in mouse and humans (Willard and Riordan, 1985) and could later be Wne-mapped to human Xq22.3 (Mattei et al., 1986). The X-linkage of PLP sparked an immediate interest in its possible role in X-linked myelin disorders, including Pelizaeus-Merzbacher disease in humans and the jimpy mutation of mice (for an early review see Nave and Milner, 1989).

THE CLINICAL SPECTRUM OF PLP GENE MUTATIONS Mouse models are most valuable if they phenocopy the human disorder both clinically and with respect to the underlying pathology. As detailed in Chapter 37, PMD was described more than 100 years ago (Merzbacher, 1910; Pelizaeus, 1885) and is clinically characterized by impaired motor development, an early onset before 1 year, and an X-linked mode of transmission. Current diagnostic criteria were deWned by Boulloche and Aicardi (1986) and include electrophysiological and magnetic resonance imaging (MRI) techniques, to be applied at any time during the life of a patient. As Pelizaeus already noticed, impaired motor development presents very early, and motor handicaps are greater than the retardation of psycho-intellectual development. In the most severe type II (‘‘connatal’’) form of PMD, aVected males show virtually no psychomotor development and death occurs in the Wrst decade. In the more frequent type I (‘‘classical’’) form, patients reach a few motor developmental milestones up to 10 to 12 years of age, but after a plateau, have a slow deterioration (at ages 15 to 20) and death occurrs often in the fourth decade (Nave and BoespXug-Tanguy, 1996). Motor performance varies between families (i.e., PLP mutations) but also between aVected males of the same PMD family. This suggests the existence of yet unidentiWed disease modiWer genes that have also been noted for Plp mutations in diVerent lines of inbred mice. Histologically, the neuropathological Wndings in PMD are the severe hypomyelination, which is limited to the CNS, in the absence of inXammation. Islets of preserved myelin around blood vessels may result in the patchy appearence of white matter. Staining neutral lipid material in white matter areas with sudan black led to the classiWcation of PMD as a ‘‘sudanophilic leukodystrophy.’’ Using these neuropathological characteristics, Seitelberger et al. (1970) subdivided several forms of PMD, taking into account the X-linked mode of transmission, age of onset, and more recently also molecular genetics (Seitelberger et al., 1996). PMD type I is the slowly progressive ‘‘classical’’ form, described Wrst by Pelizaeus; the more severe form PMD type II is ‘‘connatal’’ and shows rapid progression and death within 10 years of life. Clinically distinct from either form of PMD is X-linked spastic paraplegia type 2 (SPG-2) (Johnston and Mackusick, 1962) which is a milder, allelic form at the PLP locus (Saugier-Veber et al., 1994). SPG-2 patients have progressive gait abnormalities after an almost normal motor development in the Wrst 2 years. This makes SPG-2 a late onset disease that, nevertheless, shares clinical features with PMD (nystagmus, cerebellar ataxia, pyramidal syndrome). However, hypotonia, choreoathetosis or the bobbing movements of head and trunk, which are frequent in PMD, are rarely observed. SPG-2 patients walk unaided for many years, and most of them are able to attend school. Myelination may not be severely aVected and progressive axonal involvement may be largely responsible for the clinical picture of SPG-2. Nearly the same clinical spectrum has been recognized in mouse mutants of the Plp gene.

NATURAL MUTATIONS OF THE RODENT PLP GENE PROVIDE MODELS OF PELIZAEUS-MERZBACHER DISEASE Spontaneous neurological mutations in the mouse that are associated with a hypomyelinated phenotype of the CNS have been known for a long time. They were studied before the molecular basis of any human leukodystrophy was known. The subgroup of X-linked

V. ANIMAL MODELS OF HUMAN DISEASE

au6

1128

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

recessive myelin mutants was identiWed by formal genetic criteria (i.e., aVected males born to healthy female carriers). These mutants were allelic to each other and a mutation of the X chromosome-linked Plp gene has been identiWed in these lines. The following is a brief overview of the clinical phenotype of various Plp mutants, followed by a discussion of the possible disease mechanism in the natural and engineered Plp mutants, both of which are models for human PMD. Not covered in this chapter are Plp mutations in nonrodent animal species, including the shaking pup, a dog model (Nadon et al., 1990), paralytic tremor, a rabbit model (Tosic et al., 1994), and a congenital tremor in pig (Baumgartner and Brenig, 1996).

The Jimpy Mouse The jimpy mutation in the mouse (genetic symbol: Plpjp) provided the Wrst animal model related to PMD. This natural mutation was discovered by Phillips (1954) as a lethal X-linked trait (Sidman et al., 1964) and has been maintained by the Jackson Laboratories on a CBA background (B6CBACa Aw-J/A-Plpjp). For a long time, the genetically linked coat color gene Tabby (15 cM genetic distance) was the only marker of hemizygous jimpy males prior to the onset of symptoms, but has now been replaced by PCR-based genetic testing (Schneider et al., 1995). The clinical signs of aVected males begin in the second postnatal week, coinciding with the onset of CNS myelination in wild-type mice. Ataxia, generalized body tremor that precedes locomotor activity, and seizures are similar in this and other myelin-deWcient mouse mutants (Fig. 47.1). Before the fourth week of age, prolonged convulsions result in premature death (Sidman et al., 1964). The Wnal cause of the animals’ death may related to the toxicity of the protein expressed in speciWc brain stem neurons (discussed later). The molecular defect of jimpy mice is an A-to-G point mutation in the 7 exon Plp gene. It destroys the splice acceptor site of intron 4, thereby causing the abnormal loss of exon 5 from mature PLP and DM20 mRNA. Consequently, both PLP isoforms have truncated carboxyl-termini and lack the fourth transmembrane domain (Dautigny et al., 1986; Hudson et al., 1987; Macklin et al, 1987; Morello et al., 1986; Nave et al., 1986, 198 ). The histological jimpy phenotype is the almost complete lack of myelin in the CNS, associated with the lack of mature oligodendrocytes (Hirano et al., 1969; Krauss-Ruppert et al., 1973; Meier and BischoV, 1974, 1975; Meier and BischoV, 1975) (Fig. 47.2). Increased rate of oligodendrocyte cell death (SkoV, 1982, 1995) resembles apoptosis in some but not all aspects (Knapp et al., 1999a). A small percentage of jimpy oligodendrocytes survive and generate small patches of myelin for 1 to 2% of axons. Ultrastructurally, the intraperiod line (IPL) of myelin, normally a double-line with irregular appearance, appears fused in jimpy to a single electron-dense structure, indistinguishable from the major dense line (MDL). This ultrastructural abnormality suggests the lack of a ‘‘strut’’-type rather than a ‘‘glue’’-type molecule in compacted myelin (Duncan et al., 1989; Schneider et al., 1995). Biochemical studies have failed to puriWy myelin fractions from jimpy mice, and the mutant PLP is nearly undetectable in total brain extracts, possibly because of instability and degradation (Benjamins et al., 1994; Fannon et al., 1994). As discussed further below, PLP is made in mutant oligodendrocytes but does not reach the cell surface or the myelin compartment. Peripheral myelin appears unaVected in jimpy, possibly reXecting the CNS-speciWc assembly, if not expression, of PLP/DM20. AVected jimpy mice are generally smaller in size than wild-type males, and have other pathological features that are not obviously related to the primary gene defect or the lack of myelin (Gotow et al., 1999; Knapp et al., 1998, 1999b; Le Goascogne et al., 2000; SkoV et al., 1976; Vela et al., 1996; Wu et al., 2000). Some of these features may also be present in other mice with Plp point mutations, in all PLP-dependent myelin disorders, or even in unrelated dysmyelinated mice.

The Jimpy-msd Mouse The mouse mutant msd (Plpjp-msd; for myelin synthesis deWcient) was Wrst described by Meier and MacPike (1970) and soon identiWed as allelic to jimpy (Eicher and Hoppe, 1973). The

V. ANIMAL MODELS OF HUMAN DISEASE

au7

au8

au9

NATURAL MUTATIONS OF THE RODENT

FIGURE 47.1

FIGURE 47.2 Light micrograph of resin section taken from the white matter of the spinal cord of a jimpy mouse showing many of the features associated with spontaneous Plp gene mutations in mice. There is an overall paucity of myelin; many axons are devoid of sheaths or associated with thin sheaths. A pyknotic nucleus is present, evidence of apoptosis, probably of an oligodendrocyte.

mouse is phenotypically, and by its reduced lifespan of 3 to 4 weeks, indistinguishable from jimpy, although slightly more spinal cord axons are ensheathed (Billings-Gagliardi et al., 1980). Gencic et al. (1990) identiWed a point mutation in exon 7 of the Plp gene with a rather conservative amino acid substitution (A242V), which alters the fourth transmembrane domain of the protein. The msd mouse oVers some experimental advantages relating to the preserved carboxylterminus of the mutant protein, to which several high-aYnity antibodies are available (Gow et al., 1998; Jung et al., 1996; Yamamura et al., 1991). Interestingly, the same amino acid change (A242V) has been identiWed in one human family with connatal PMD (Komaki et al., 1999), and a related human mutation (A242E) has been described by Seeman et al., (2002).

The Jimpy-4J Mouse The spontaneous mutation jimpy-4J (Plp jp-4J) was identiWed in the Jackson laboratory as a new jimpy allele with the most severe histological phenotype described (Billings-Gagliardi et al., 1995). Jimpy-4J mice show the complete absence of CNS myelin, a dramatic loss of oligodendrocytes, and live for about 24 days. The mutation has been found in exon 2, predicting the substitution Ala-38-Ser in Wrst hydrophilic loop of PLP (Pearsall et al., 1997).

V. ANIMAL MODELS OF HUMAN DISEASE

1129

1130

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

The Myelin-Deficient Rat The myelin-deWcient md rat has been identiWed by Csiza (Csiza and Lahunta, 1979; Dentinger et al. 1982) and also provides a model for the connatal form of PMD associated with a lack of PLP (Koeppen et al., 1987). Boison and StoVel (1989) identiWed the primary defect in the rat Plp gene, associated with the nonconservative substitution T74P and predicting a break in the second transmembrane helix of PLP. The complete lack of mutant PLP from compact myelin results in a characteristic decrease of myelin periodicity—that is, closer apposition of myelin membranes (Duncan et al., 1987)—later conWrmed in jimpy mice (Duncan et al., 1995; Schneider et al., 1995), as well as subtle axonal abnormalities (Barron et al., 1987; Dentinger et al., 1985). Clinically, md rats are similar to the Plp mutant mice, with tremors, tonic seizures, and premature death. Recently, a longer-lived mutant substrain has been isolated in which aVected males survive for about 80 days (Duncan et al., 1995), suggesting the importance of yet unknown modiWer genes. As larger animals, md rats are well suited for cell transplantation experiments (Bruestle et al., 1999), for electrophysiological studies on the consequences of myelin-deWciency (Utzschneider et al., 1992; Young et al., 1989), and for research on the underlying cause of premature death (Miller et al., 2003).

The Rumpshaker Mouse The rumpshaker (Plpjp-rsh) mutation predicts the substitution Ile-186-Thr in PLP and DM20, and is associated with a relatively mild phenotype (Schneider et al., 1992). The preservation of oligodendrocytes with considerably reduced cell death, when compared to jimpy, results in substantially improved myelination (Fig. 47.3) (GriYths et al., 1990). The optic nerve contains both amyelinated and normally myelinated axons, whereas spinal cord axons are only thinly myelinated with little growth in thickness during development (Fanarraga et al., 1992). DiVerent from jimpy and jimpy-msd mice, rumpshaker mutant myelin also incorporates signiWcant amounts of proteolipids, with the DM20 isoform predominating (Fanarraga et al., 1992; Karthigasan et al., 1996). However, the altered ratio of PLP/DM20 is not responsible by itself for a dysmyelinated phenotype (Uschkureit et al., 2001). Karthigasan et al. (1996) reported that the biochemical abnormalities in rumpshaker myelin correlate with a wider periodicity and less stable packing of the layers. Rumpshaker mice also demonstrated that apoptosis in other Plp mutants is not caused by dysmyelination per se (Schneider et al., 1992), but as discussed later, rather by the toxicity of mutant PLP and DM20 isoforms when retained in the endoplasmic reticulum of oligodendrocytes (Gow and Lazzarini, 1996; Gow et al., 1998). Rumpshaker mice are long lived and reproduce well, although showing ataxia and mild tremor, but little seizure activity. Interestingly, the severity of their phenotype is markedly inXuenced by the background strain—that is, unknown modiWer genes (Al-Saktawi et al., 2003) (Fig. 47.4). Rumpshaker mice are a good animal model for SPG-2 in humans, in fact the same point mutation (I186T) was identiWed later in the index family of this human disease (Kobayashi et al., 1994; Naidu et al., 1997).

PLP/DM20-DEPENDENT DISEASE MECHANISMS IN NATURAL AND ENGINEERED RODENT MODELS OF PMD AND SPG-2 Pelizaeus-Merzbacher disease and SPG-2 have been associated with (and are now genetically deWned by) mutations of the human PLP locus, speciWcally (1) various point mutations in the coding region of the PLP gene and consequently the expression of mutant proteolipids, (2) with the loss of PLP function in null alleles or equivalent mutations, and (3) with the overexpression of PLP in patients that carry a gene duplication. For an updated listing of speciWc mutations in the PLP gene, see www.med.wayne.edu/Neurology/plp.html. At the molecular level, it appears that PMD/SPG-2 is associated with three distinct pathomechanisms that act overlappingly. They have been revealed by dissecting the disease process at the cellular level in cultured cells and in corresponding mouse mutants.

V. ANIMAL MODELS OF HUMAN DISEASE

au10

PLP/DM20-DEPENDENT DISEASE MECHANISMS

FIGURE 47.3 Electron micrograph of white matter from the spinal cord of a rumpshaker mouse. There is moderate dysmyelination with axons either naked or surrounded by a disproportionately thin myelin sheath. The outline of the sheaths is undulating and vacuoles are present but the majority of the myelin is compacted, although when viewed at high magnification the periodicity is reduced (not shown).

FIGURE 47.4 Area of white matter from the spinal cord of a rumpshaker mouse with the mutation on the C57BL/6 genetic background. When the mutation is expressed on this background the phenotype is considerably more severe than when on the C3H background (Al-Shaktawi et al., 2003). The section is immunostained with the CC-1 antibody (green) to demonstrate oligodendrocytes and with the apoptotic marker, caspase-3 (red). One double labeled oligodendrocyte (orange) is undergoing apoptosis; another unidentiWed cell is single labeled for caspase-3 (red). Apoptotic cells are rare when the rumpshaker mutation is expressed on the C3H background but common when on the C57BL/6 background and also with most other spontaneous mutations of the Plp gene, such as jimpy.

Plp Point Mutations and Abnormal Oligodendrocyte Death The ‘‘classical’’ neurological mouse mutant jimpy lacks CNS myelin almost completely. The underlying histological feature is a paucity of mature oligodendrocytes and signs of abnormally increased oligodendrocyte death, such as pyknotic nuclei and TUNEL positive oligodendroglial nuclei (SkoV, 1982; Knapp et al., 1986). Jimpy oligodendrocytes die even

V. ANIMAL MODELS OF HUMAN DISEASE

1131

1132

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

when maintained in cell culture and it has been impossible so far to deWne survival factors that prevent cell death (Bartlett et al., 1998; Williams and Gard, 1997; Wolf and Holden, 1969). More recent studies suggest that abnormal cell death in jimpy is a combination of both apoptotic and nonapoptotic mechanisms (Cerghet et al., 2001; Southwood and Gow, 2001; Thomson et al., 1999). Transplantation experiments of oligodendrocytes from jimpy donor brains into wild-type (or shiverer mutant) host brains have conWrmed that cell death is cell autonomous at an early window of postnatal development (Lachapelle et al., 1994). It coincides with increasing expression of the mutant PLP gene, albeit the abnormal protein never accumulates to signiWcant levels (Vermeesch et al., 1990). This issue has been re-investigated at the single cell level in both jimpy and jimpy-msd mice, and also in the md-rat, with high-aYnity PLP/DM20 antibodies, TUNEL staining, confocal immunoXuorescence microscopy, in situ hybridization, and electron microscopy (Gow et al., 1998; Grinspan et al., 1998; Lipsitz et al., 1998; Thomson et al., 1999). Collectively, these studies have provided evidence that mutant oligodendrocytes mature normally up to the stage of expressing myelin proteins and other late diVerentiation markers, but undergo apoptosis later, presumably following a threshold expression of mutant PLP. In normal development, oligodendrocytes are generated in excess and their Wnal number is regulated by extrinsic signals that counteract apoptosis (Barres et al., 1992; Trapp et al., 1997). Theoretically, oligodendrocyte death in PLP mutants could result from disturbed glial diVerentiation or from the loss of a PLP/DM20 function in oligodendrocyte development. The latter seems unlikely because the complete loss of PLP in a genetic null mutant mouse does not obviously diminish oligodendrocyte survival (StoVel et al., 1984; Klugmann et al., 1997; Yool et al., 2001). Moreover, mutant oligodendrocytes express most of the late antigenic marker genes before they die (Gow et al., 1998; Grinspan et al., 1998; Pringle et al., 1997). Nevertheless, Beesley et al. (2001) found that mutant oligodendrocytes (cultured from md rats) failed to express the myelin oligodendrocyte glycoprotein (MOG). Even when oligodendrocyte apoptosis was suppressed with a caspase-3 inhibitor, the expression of MOG could not be detected, suggesting a rather direct eVect of mutant PLP on oligodendrocyte diVerentiation. At present, oligodendrocyte death in PLP mutant mice is best explained by a ‘‘toxic’’ eVect of aberrant Plp expression on oligodendrocyte survival. Using heterologous expression of Plp in Wbroblastoid cells, Gow et al. (1994) observed the abnormal retention of mutant, but not wild-type, PLP and DM20 in the endoplasmic reticulum of transfected COS-7 cells. Using this cDNA expression system, a genotype-phenotype correlation between the degree of intracellular retainment of mutant proteolipid and disease severity (in PMD and the natural mouse mutants) has been established (Gow and Lazzarini, 1996). PLP and DM20 from mutant mice is indeed physically misfolded protein, as shown by the monoclonal antibody O10 that deWnes a conformation-sensitive epitope that is lost in PLP/ DM20 isoforms derived from dysmyelinated mice (Jung et al., 1996). Expression of misfolded membrane proteins that are subsequently retained in the ER, is known to induce a cellular stress response (Ron, 2002; Sherman and Goldberg, 2001; that may lead to apoptotic cell death in vulnerable cells, such as developing oligodendrocytes. The induction of various down-stream eVector genes of the ER response, including chop1, has been recently described by Gow and coworkers for Plp-mutant rumpshaker mice (Southwood et al., 2002). Whether the unexpected eVect of Chop1-deWciency worsening the rumpshaker phenotype is directly atributable to Chop1 function in stressed oligodendrocytes (Southwood et al., 2002), or the eVect of unknown modiWer genes associated with this and other Plp mutations (Al-Saktawi et al., 2003; Billings-Gagliardi et al., 2001; Duncan et al., 1995) remains to be determined. An important physiological question is the ultimate cause of premature death of these mice after several weeks of survival. The comparison with other myelin mutants suggests that Plp mutants do not die as a result of dysmyelination per se (Billings-Gagliardi et al., 1999; Wolf et al., 1999). In the md rat it has recently been shown that the expression and toxicity of mutant PLP, detectable in a subset of glutamatergic brainstem neurons (not oligodendroglia) of the nucleus hypoglossus, perturbs the normal function of these nuclei

V. ANIMAL MODELS OF HUMAN DISEASE

1133

PLP/DM20-DEPENDENT DISEASE MECHANISMS

in breathing control. Miller et al. (2003) found that normal ventilatory functions, following prolonged epileptic seizures and hypoxia, are not properly used for compensation. Thus, it may be a speciWc neuronal dysfunction and hypoxic ventilatory depression, rather than the loss of myelin, that ultimately leads to premature death.

Transgenic Plp Overexpression as a Model for PLP Gene Duplication The second disease mechanism underlying PMD involves the duplication of the entire X-linked PLP/DM20 gene—that is, a mere twofold overexpression of wild-type PLP and DM20 protein. It has now been recognized that PLP gene duplications, rather than point mutations, underlie the majority of familial cases with PMD (Cremers, et al., 1987; Ellis and Malcolm, 1994; Mimault et al., 1999; Nave and BoespXug-Tanguy, 1996; Sistermans et al., 1998). Turning a normal cellular gene into a lethal disease gene by two-fold overexpression is a challenge to understand at the cellular level, possibly related to overexpression of a duplicated PMP22 gene in Charcot-Marie-Tooth disease type 1A (see Chapters 39 and 48). Mouse models for this type of PMD have been generated inadvertantly, originally in an attempt to complement the defect of jimpy mice with a wild type Plp transgene (Kagawa et al., 1994; Readhead et al., 1994). An unexpected observation was a severely dysmyelinated phenotype of all those transgenic mice that overexpressed wild-type Plp at least two-fold (as quantiWed at the RNA level) (Fig. 47.5). In the study of Readhead et al. (1994), such mutants were homozygous or double-heterozygous for either one of two cosmid-derived full-length Plp transgenes (line #72 or #66). The novel phenotype was related to that of jimpy mice, but less severe, with later onset, and with signiWcantly longer survival times (up to 6 months). DiVerent from the dysmyelinated jimpy mouse, the histological phenotype of Plp overexpressors revealed a complex mixture of dys- and demyelinating features, explaining the later onset of behavioural symptoms. Abnormal oligodendrocyte death (Fig. 47.6) was less obvious in the homozygous lines #72 and #66 of Readhead et al. (1994), but pronounced in the shorter-lived line #4e of Kagawa et al. (1994), which overexpressed Plp at a higher level, and independently in a line of Plp -transgenic rats (Bradl et al., 1999). The mechanisms of abnormal cell death appear diVerent from those in Plp-mutant mice (Cerghet et al., 2001). A follow-up study of the PLP-transgenic rat demonstrated an intracellular accumulation of PLP together with other membrane proteins, such as MAG and MOG (but not with MBP), suggesting a rather widespread eVect on the export of other myelin proteins, but also nonmyelin membrane proteins (Bauer et al., 2002). ER distensions, unfolded protein response (with the induction of chaperones), and occcasional cell death was most obvious in grey matter oligodendrocytes. It is unlikely, however, that this striking cellular phenoytype is the same in human PMD, because the level of PLP overexpression is only twofold in duplication patients. Why is PLP overexpression not tolerated? It is possible that PLP interacts stoichiometrically with cellular proteins that are required for proper PLP export. If such a stochiometry were disturbed, excess PLP may accumulate and thus mimick the retention of misfolded PLP in mutant mice. Another hypothesis has been put forward by Simons et al. (2002), who found that the known interaction of PLP with cholesterol in membrane lipid rafts causes PLP-overexpressing oligodendrocytes to deliver both PLP and cholesterol into the lysosomal compartment. Here, the abnormal appearance of cholesterol could conceivably alter normal lysosomal functions and thus aVect general cell metabolism. Ultrastructurally, oligodendrocytes in PLP overexpressing mice contain numerous autophagic vacuoles (Readhead et al., 1994) (Fig. 47.5), which is in agreement with this hypothesis. Again it is not proven that these Wndings can be fully extrapolated to oligodendrocytes in PMD patients with only twofold PLP gene dosage. Interestingly, cultured cells that are vulnerable to PLP overexpression appear less aVected by overexpression of the longer srPLP isoform (Bongarzone et al., 2001b). Mice with a single copy of the Plp transgene (in line #66 or #72) that overexpress PLP and DM20 less than twofold at the transcriptional level (Readhead et al., 1994), develop

V. ANIMAL MODELS OF HUMAN DISEASE

au11

1134

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

FIGURE 47.5 Electron micrograph of white matter from the spinal cord of a transgenic mouse overexpressing wild type PLP (#66, Readhead et al., 1994). Dysmyelinated axons surround a central oligodendrocyte cell body in which numerous autophagic vacuoles containing membranous structures, are evident.

FIGURE 47.6 Immunostained sagittal section of cerebrum of a Plp overexpressing transgenic mouse showing the corpus callosum (lower aspect of Wgure) and the adjacent cortex (upper). The section is stained for MAG (red) which strongly labels the oligodendrocyte cell bodies. Note the paucity of myelin staining in the corpus callosum. Costaining for caspase-3 (green) indicates several oligodendrocyes undergoing apoptosis.

normally. However, after 12 months of age, they show progressive ataxia and behavioural abnormalities. This aspect of PLP-dependent disease is caused by a (noninXammatory) demyelination with clear axonal involvement, most obvious in long tracts of the spinal cord (Anderson et al., 1998). CNS demyelination has Wrst been observed in adult mice overexpressing multiple copies of a DM20 cDNA transgene driven by a myelin-speciWc promoter (Mastronardi et al., 1993; Simons-Johnson, et al., 1995). The mechanism underlying this late-onset neurodegeneration is unclear and its relationship to the developmental

V. ANIMAL MODELS OF HUMAN DISEASE

PLP/DM20-DEPENDENT DISEASE MECHANISMS

disorder in homozygous Plp -transgenics remains to be determined. It is likely that a similar axonal involvement contributes to the progressive nature of human PMD in patients with a PLP gene duplication.

PLP/DM20-Deficient Mice: Proteolipid Function and Axonal Degeneration Several laboratories have used gene targeting technologies and homologous recombination of in mouse embryonic stem cells to ablate the expression of PLP and DM20 (Boison and StoVel, 1994; Klugmann et al., 1997; Sporkel et al., 2002; Stecca et al., 2000). These mice should provide novel models of PMD, when associated with a deletion of the entire PLP gene (Raskind et al., 1991) or an equivalent null mutation (Sistermans et al., 1996), associated with a milder clinical course of disease. Moreover, null mutants should diVerentiate between true loss-of-function and aberrant gain-of-function eVects that overlap in most mice and human patients with point mutations in this gene. Isoform-speciWc gene targeting should also help to understand functional diVerences between PLP and DM20. Surprisingly, the interpretation of all these data, when combined, is more diYcult than anticipated, but the mice have clearly shown that PLP is not required for myelin formation. A toxic gain-of-function of PLP, when misfolded or overexpressed, emerges as the major cause of dysmyelination and cell death. StoVel and coworkers reported the Wrst targeting of the Plp gene and obtained mice with a severe perturbation of Plp gene expression (close to a null mutation), resulting from the combination of a splice defect and anti-sense silencing by an inserted promoter sequence (Boison et al., 1994). This mutation did not interfere with myelination, but the ultrastructure of CNS myelin was highly disordered. Myelin lamellae were loosely wrapped and lacked the apposition of the extracytoplasmic membrane surfaces (intraperiod line). This led to a profound reduction of conductance velocities of CNS axons (Gutierrez et al., 1996), impairments in neuromotor coordination, and behavioral changes. Large diameter axons were loosely myelinated, and small axons were even unmyelinated. This led to the conclusion that adhesion properties of PLP are responsible for the tight apposition of membranes in compact myelin (Boison et al., 1995). The creation of a complete PLP/ DM20 null mutation in mice by Klugmann et al. (1997) conWrmed most of these Wndings, but showed that the defect at the intraperiod line was highly variable, even within a single sheath with absent, condensed, and normal intraperiod lines all being observed (Klugmann et al., 1997; Yool et al., 2002). A tendency toward increased Wxation artefacts (myelin delamination) in mutant brains was concluded and interpreted as a sign of reduced physical stability (see also Rosenbluth et al., 1996). There were no motor deWcits in young adult PLP-deWcient mice and also the peripheral nervous system was normal, at least within the short lifespan of mice (Klugmann et al., 1997), a remarkable diVerence from human patients with the PLP null mutation who develop a demyelinating peripheral neuropathy (Garbern et al., 1997). A second observation, important for the pathomechanism of PMD, was the Wnding that although myelin is assembled in the absence of PLP/DM20, after several months widespread axonal swellings cause neurodegeneration, predominantly of small-caliber axons (GriYths et al., 1998b) (Fig. 47.7). This Wber degeneration is probably secondary to impaired axonal transport, and is direct proof that myelinated axons require local oligodendroglial support (and that PLP/DM20 is required for this oligodendroglial function). Length-dependent axonal degenerations of the CNS have subsequently been found in both PLP-deWcient mice and human patients with PMD (Garbern et al., 2002). Such swellings and axonal degeneration proWles are not a feature of myelin mutants in general. They have been noted recently in the myelinated CNS of mice that carry a mutation of the oligodendrocyte-speciWc Cnp1 (CNPase) gene (Lappe-Siefke et al., 2003). Thus, several genes (including Plp) may be required by oligodendrocytes to provide axonal support, independent of myelin assembly itself. Loss of such a supportive glial function may explain the progressive axonal involvement that can be observed in several inherited and acquired myelin diseases, both in the CNS and PNS. The molecular nature of the underlying axonglial interaction is completely unknown.

V. ANIMAL MODELS OF HUMAN DISEASE

1135

1136

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

FIGURE 47.7 Electron micrograph of (A) longitudinal and (B) transverse sections of myelinated axons from a Plp null mouse. The axoplasm contains numerous membranous organelles, dense bodies and mitochondria. These typically accumulate initially distal to the nodal complex and are related to defects in retrograde axonal transport.

What is the molecular function and functional diVerence of PLP and DM20? Targeting the PLP/DM20 gene in combination with the Cre-Lox system, Stecca et al. (2000) have generated mice that lack alternative splicing and express, at normal levels, only the DM20 isoform. These mice showed the same phenoytype as reported for the combined PLP/ DM20 null mutants, including the late onset axonal degeneration desribed by GriYths et al. (1998b). This would suggest that PLP or PLP and DM20, but not DM20 alone, are required for maintaining axonal integrity. In contrast, Spo¨rkel et al. (2002) who have independently created a DM20-only mouse mutant, found minimal phenotypical diVerence from wild-type mice, which would suggest that DM20 is suYcient for oligodendrocytes to maintain axonal integrity, and that there is little functional diVerence between both isoforms. This discrepancy cannot be adequately explained at the moment. However, already GriYths et al. (1998b) noted that either PLP or DM20 cDNA based transgenes, when expressed in PLP/DM20 null mice, were both able to reduce the extent of axonal pathology, which agrees better with the Wndings of Spo¨rkel et al. (2002). The priciple cellular function of PLP and DM20 remains to be identiWed. It is likely that in oligodendrocytes a DM20-related proteolipid, termed M6B, serves in part a redundnant function with PLP/DM20. This is suggested because both PLP-deWcient and M6B-deWcient mutants are myelinated with few ultrastructural defects, whereas PLP/M6B double deWcient mice have major developmental defects of CNS myelin (Werner et al., in preparation).

V. ANIMAL MODELS OF HUMAN DISEASE

CONCLUSIONS

CONCLUSIONS Mouse models of Pelizaeus-Merzbacher disease and SPG-2 have helped us to better understand the genetic basis and molecular pathology of the human myelin disorder and its wide clinical spectrum. This includes numerous histological, ultrastructural, and biochemical investigation of dysmyelination and demyelination in the CNS of PLP mutant mice, tracing back to the natural mutations in the 1960s and 1970s, and followed by the analysis of experimentally derived mutant lines in recent years. In conjunction with subcellular observations on mutant protein traYcking, in vivo and in vitro, a causal relationship of myelin protein misfolding, loss-of-function, abnormal gain-of-function eVects, and lethal dysmyelination can be reconstructed. The greatest challenge remains in understanding the principle role that proteolipids in glia play in myelin assembly and long-term axonglia interactions, the latter playing a major role in the progressive nature of PMD.

References Al-Saktawi, K., McLaughlin, M., Klugmann, M., Schneider, A., Barrie, J. A., McCulloch, M. C., Montague, P., Kirkham, D., Nave, K.-A., GriYths, I. R., 2003, Genetic background determines phenotypic severity of the Plp rumpshaker mutation. J. Neurosci. Res. 72:12–24. Anderson, T. J., Montague, P., Nadon, N., Nave, K-A., GriYths, I. R.,1997, ModiWcation of Schwann cell phenotype with PLP transgenes: Evidence that the PLP and DM20. J. Neurosci. Res. 50:13–22. Anderson, T. J., Schneider, A., Barrie, J. A., Klugmann, M., McCulloch, M. C., Kirkham, D., Kyriakides, E., Nave, K.-A., GriYths, I. R.,1998, Late-onset neurodegeneration in mice with increased dosage of the proteolipid protein gene. J. Comp. Neurol. 394:506–519. Barres, B. A., Hart, I. K., Coles, H. S., Burne, J. F, Voyvodic, J. T., Richardson, W. D., RaV, M. C., 1992, Cell death and control of cell survival in the oligodendrocyte lineage. Cell 70:31–46. Barron, K. D., Dentinger, M. P., Csiza, C. K., Keegan, S. M., Mankes, R., 1987, Abnormalities of central axons in a dysmyelinative rat mutant. Exp. Mol. Pathol. 47:125–142. Bartlett, W. P., Knapp, P. E., SkoV, R. P 1998, Glial conditioned medium enables jimpy oligodendrocytes to express properties of normal oligodendrocytes: Production of myelin antigens and membranes. Glia 1:253–259. Bauer, J., Bradl, M., Klein, M., Leisser, M., Deckwerth, T. L., Wekerle, H., Lassmann, H., 2002, Endoplasmic reticulum stress in PLP-overexpressing transgenic rats: Gray matter oligodendrocytes are more vulnerable than white matter oligodendrocytes. J Neuropathol Exp Neurol. 61:12–22. Baumgartner, B. G., Brenig, B., 1996, [The role of proteolipid proteins in the development of congenital tremors type AIII: A review], Dtsch. Tierarztl. Wochenschr. 103:404–407. Beesley, J. S., Lavy, L., Eraydin, N. B., Siman, R., Grinspan, J. B., 2001, Caspase-3 activation in oligodendrocytes from the myelin-deWcient rat. J Neurosci Res. 64:371–379. Benjamins, J. A., Studzinski, D. M., SkoV, R. P., 1994, Analysis of myelin proteolipid protein and F0 ATPase subunit 9 in normal and jimpy CNS. Neurochem Res. 19:1013–1022. Billings-Gagliardi, S., Adcock, L. H., Wolf, M. K., 1980, Hypomyelinated mutant mice: Description of jpmsd and comparison with jp and qk on their present genetic backgrounds. Brain Res. 194:325–338. Billings-Gagliardi, S., Nunnari, J. N., Nadon, N. L., Wolf, M. K., 1999, Evidence that CNS hypomyelination does not cause death of jimpy-msd mutant mice. Dev Neurosci. 21:473–82. Billings-Gagliardi, S., Nunnari, J. J., Wolf, M. K., 2001, Rumpshaker behaves like juvenile-lethal Plp mutations when combined with shiverer in double mutant mice. Dev Neurosci. 23:7–16. Billings-Gagliardi, S., Kirschner, D. A., Nadon, N. L., DiBenedetto, L. M., Karthigasan, J., Lane, P., Pearsall, G. B., Wolf, M. K., 1995, Jimpy 4J: A new X-linked mouse mutation producing severe CNS hypomyelination. Dev Neurosci. 17:300–310. Boison, D., StoVel, W., 1989, Myelin-deWcient rat: A point mutation in exon III (A!C, Thr75!Pro) of the myelin proteolipid protein causes dysmyelination and oligodendrocyte death. EMBO J. 8:3295–3302. Boison, D., StoVel,W., 1994, Disruption of the compacted myelin sheath of axons of the central nervous system in proteolipid protein-deWcient mice. Proc. Natl. Acad. Sci. USA 91:11709–11713. Boison, D., Bu¨ssow, H., D’Urso, D., Mu¨ller, H. W., and StoVel, W.,1995, Adhesive properties of proteolipid protein are responsible for the compaction of CNS myelin sheaths. J. Neurosci. 15:5502–5513. Bongarzone, E. R., Campagnoni, C. W., Kampf, K., Jacobs, E. C., Handley, V. W., Schonmann, V., Campagnoni, A. T., 1999, IdentiWcation of a new exon in the myelin proteolipid protein gene encoding novel protein isoforms that are restricted to the somata of oligodendrocytes and neurons. J. Neurosci. 19:8349–8357. Bongarzone, E. R., Jacobs, E., Schonmann, V., Campagnoni, A. T., 2001a, Classic and soma-restricted proteolipids are targeted to diVerent subcellular compartments in oligodendrocytes. J Neurosci. Res. 65:477–484. Bongarzone, E. R., Jacobs, E., Schonmann, V., Kampf, K., Campagnoni, C. W., Campagnoni, A. T., 2001b, DiVerential sensitivity in the survival of oligodendrocyte cell lines to overexpression of myelin proteolipid protein gene products. J Neurosci Res. 65:485–492.

V. ANIMAL MODELS OF HUMAN DISEASE

1137

1138

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

Boulloche, J., Aicardi, J., 1986, Pelizaeus-Merzbacher disease: Clinical and nosological study. J. Child Neurol. 1:233–239. Bradl, M., Bauer, J., Inomata,T., Zielasek, J., Nave, K.-A., Toyka, K., Lassmann, H., Wekerle, H., 1999, Transgenic Lewis rats overexpressing the proteolipid protein gene: Myelin degeneration and its eVect on T cell-mediated experimental autoimmune encephalomyelitis. Acta Neuropathol. 97:595–606. Bradl, M., Linington, C., 1996, Animal models of demyelination. Brain Pathol. 6:303–311. Brustle, O., Jones, K. N., Learish, R. D., Karram, K., Choudhary, K., Wiestler, O. D., Duncan, I. D., McKay, R. D., 1999, Embryonic stem cell-derived glial precursors: A source of myelinating transplants. Science 285, 754–756. Cerghet, M., Bessert, D. A., Nave, K.-A., and SkoV, R. P.,2001, DiVerential expression of apoptotic markers in jimpy and in PLP overexpressors: Evidence for diVerent apoptotic pathways. J. Neurocytol. 30:841–855. Cremers, F. P. M., PfeiVer, R. A., van de Pol, T. J. R., Hofker, M. H., Kruse, T. A., Wieringa, B., and Ropers, H. H.,1987, An interstitial duplication of the X chromosome in a male allows physical Wne mapping of probes from the Xq13-q22 region. Hum. Genet. 77:23–27. Csiza, C. K., de Lahunta,A., 1979, Myelin deWciency (md): A neurologic mutant in the Wistar rat. Am J Pathol. 95:215–223. Dautigny, A., Mattei, M. G., Morello, D., Alliel, P. M., Pham-Dinh, D., Amar, L., Arnaud, D., Simon, D., Mattei, J. F, Guenet, J. L., et al., 1986, The structural gene coding for myelin-associated proteolipid protein is mutated in jimpy mice. Nature 321:867–329. Dentinger, M. P., Barron, K. D., Csiza, C. K., 1982, Ultrastructure of the central nervous system in a myelin deWcient rat. J. Neurocytol. 11:671–691. Dentinger, M. P., Barron, K. D., Csiza, C. K., 1985, Glial and axonal development in optic nerve of myelin deWcient rat mutant. Brain Res. 344:255–266. Dickinson, P. J., Fanarraga, M. L., GriYths, I. R., Barrie, J. A., Kyriakides, E., Montague, P.,1996, Oligodendrocyte progenitors in the embryonic spinal cord express DM-20. Neuropath. Appl. Neurobiol. 22:188–198. Diehl, H. J., Schaich, M., Budzinski, R. M., StoVel, W., 1986, Individual exons encode the integral membrane domains of human myelin proteolipid protein. Proc. Natl. Acad. Sci U S A. 83:9807–9811. Duncan, I. D., Hammang, J. P., Goda, S., Quarles, R. H., 1989, Myelination in the jimpy mouse in the absence of proteolipid protein. Glia 2:148–154. Duncan, I. D., Hammang, J. P.,and Trapp, B. D., 1987, Abnormal compact myelin in the myelin-deWcient rat: Absence of proteolipid protein correlates with a defect in the intraperiod line. Proc. Natl. Acad. Sci. USA 84:6287–6291. Duncan, I. D., Nadon, N. L., HoVman, R. L., Lunn, K. F., Csiza, C., Wells, M. R., 1995, Oligodendrocyte survival and function in the long-lived strain of the myelin deWcient rat. J Neurocytol. 24:745–762. Eicher, E. M., Hoppe, P. C., 1973, Use of chimeras to transmit lethal genes in the mouse and to demonstrate allelism of the two X-linked male lethal genes jp and msd. J Exp. Zool. 183:181–184. Ellis, D., Malcom S., 1994, Proteolipid protein gene dosage eVect in Pelizaeus-Merzbacher disease. Nat. Genet. 6: 333–334. Fanarraga, M. L., GriYths, I. R., McCulloch, M. C., Barrie, J. A., Kennedy, P. G. E., and Brophy, P. J., 1992, Rumpshaker: An X-linked mutation causing hypomyelination: Developmental diVerences in myelination and glial cells between the optic nerve and spinal cord. Glia 5:161–170. Fannon, A. M., Mastronardi, F. G., Moscarello, M. A., 1994, Isolation and identiWcation of proteolipid proteins in jimpy mouse brain. Neurochem Res. 19:1005–1012. Filbin, M. T., Walsh, F. S., Trapp, B. D., Pizzey, J. A., and Tennekoon, G. I.,1990, Role of myelin P0 protein as a homophilic adhesion molecule. Nature 344:871–872. Garbern, J. Y., Cambi, F., Tang, X. M., Sima, A. A., Vallat, J. M., Bosch, E. P., Lewis, R., Shy, M., Sohi, J., Kraft, G., Chen, K. L., Joshi, I., Leonard, D. G., Johnson, W., Raskind, W., Dlouhy, S. R., Pratt, V., Hodes, M. E., Bird, T., and Kamholz, J., 1997, Proteolipid protein is necessary in peripheral as well as central myelin. Neuron 19:205–218. Garbern, J., Cambi, F., Shy, M., Kamholz, J., 1999, The molecular pathogenesis of Pelizaeus-Merzbacher disease. Arch Neurol. 56:1210–1214. Garbern, J. Y., Yool, D. A., Moore, G. J., Wilds, I. B., Faulk, M. W., Klugmann, M., Nave, K.-A., Sistermans, E. A., van der Knaap, M. S., Bird, T. D., Shy, M. E., Kamholz, J. A., GriYths, I. R., 2002, Patients lacking the major CNS myelin protein, proteolipid protein 1, develop length-dependent axonal degeneration in the absence of demyelination and inXammation. Brain 125:551–561. Gencic,S., Hudson, L., 1990, Conservative amino acid substitution in the myelin proteolipid protein of jimpymsd mice. J. Neurosci. 10:117–124. Genoud, S., Lappe-Siefke, C., Goebbels, S., Radtke, F., Aguet, M., Scherer, S., Suter, U., Nave, K.-A., Mantei, N., 2002, Notch-1 control of oligodendrocyte diVerentiation. J. Cell Biol. 158, 709–718. Gotow, T., Leterrier, J. F., Ohsawa, Y., Watanabe, T., Isahara, K., Shibata, R., Ikenaka, K., Uchiyama, Y., 1999, Abnormal expression of neuroWlament proteins in dysmyelinating axons located in the central nervous system of jimpy mutant mice. Eur J Neurosci. 11:3893–3903. Gow, A., Friedrich, V. L., Lazzarini, R. A., 1994, Many naturally occuring mutations of myelin proteolipid protein impair its intracellular transport. J. Neurosci. Res. 37:574–583. Gow, A., and Lazzarini, R. A., 1996, A cellular mechanism governing the severity of Pelizaeus-Merzbacher disease. Nature Genetics 13, 422–428. Gow, A., Southwood, C. M., Lazzarini, R. A., 1998, Disrupted proteolipid protein traYcking results in oligodendrocyte apoptosis in an animal model of Pelizaeus-Merzbacher disease. J Cell Biol. 140:925–934.

V. ANIMAL MODELS OF HUMAN DISEASE

CONCLUSIONS

GriYths, I. R., 1996, Myelin mutants: Model systems to study normal and abnormal myelination. BioEssays 18:789–797. GriYths, I. R., Dickinson, P., Montague, P., 1995a, Expression of the proteolipid protein gene in glial cells of the post-natal peripheral nervous system of rodents. Neuropath. Appl. Neurobiol. 21:97–110. GriYths, I. R., Klugmann, M., Anderson, T. J., Thomson, C. E., Vouyiouklis, D. A., Nave, K.-A., 1998a, Current concepts of PLP and its role in the nervous system. Microsc. Res. Tech. 41:344–358. GriYths, I. R., Klugmann, M., Anderson, T., Yool, D., Thomson, C.,Schwab, M. H., Schneider, A., Zimmermann, F., McCulloch, M., Nadon, N., and Nave, K.-A.,1998b, Axonal swellings and degeneration in mice lacking myelin proteolipid protein. Science 280:1610–1613. GriYths, I. R., Schneider, A., Anderson, J., Nave, K.-A., 1995b, Transgenic and natural mouse models of proteolipid protein (PLP) related dysmyelination and demyelination. Brain Pathol. 5:275–281. GriYths, I. R., Scott, I., McCulloch, M. C., Barrie, J. A., McPhilemy, K., Cattanach, B. M., 1990, Rumpshaker mouse: A new X-linked mutation aVecting myelination: evidence for a defect in PLP expression. J Neurocytol. 19:273–283. Grinspan, J. B., Coulalaglou, M., Beesley, J. S., Carpio, D. F, Scherer, S. S., 1998, Maturation-dependent apoptotic cell death of oligodendrocytes in myelin-deWcient rats. J Neurosci Res. 54:623–634. Gutierrez, R., Boison, D., Heinemann, U., StoVel, W., 1996, Decompaction of CNS myelin leads to a reduction of the conduction velocity of action potentials in optic nerve. Neurosci Lett. 195:93–96. Hirano, A., Sax, D. S., Zimmerman, H. M., 1969, The Wne structure of the cerebella of jimpy mice and their ‘‘normal’’ litter mates. J. Neuropathol. Exp. Neurol. 28:388–400. Hodes, M. E., Pratt, V. M., Dlouhy, S. R., 1994, Genetics of Pelizaeus-Merzbacher disease. Dev. Neurosci. 15:383–394. Hudson, L. D., Berndt, J. A., Puckett, C., Kozak, C. A., Lazzarini, R. A., 1987, Aberrant splicing of proteolipid protein mRNA in the dysmyelinating jimpy mutant mouse. Proc. Natl. Acad. Sci. USA 84:1454–1458. Hudson, L. D., Puckett, C., Berndt, J., Chan, J., Gencic, S., 1989, Mutation of the proteolipid protein gene (PLP) in a human X chromosome-linked disorder. Proc. Natl. Acad. Sci. USA 86:8128–8131. Johnston, A. W., MacKusick, V. A., 1962, A sex-linked recessive form of spastic paraplegia. Am J. Hum. Genet. 14:83–94. Jung, M., Sommer, I., Schachner, M., and Nave K-A, 1996, Monoclonal antibody O10 deWnes a conformationally sensitive cell-surface epitope of proteolipid protein (PLP): Evidence that PLP misfolding underlies dysmyelination in mutant mice. J. Neurosci. 24:7920–7929. Kagawa, T., Ikenaka, K., Inoue, Y., Kuriyama, S., Tsujii, T., Nakajima, J., Nakao, J., Aruga, J., Okano, H., and Mikoshiba, K., 1994, Glial cell degeneration and hypomyelination caused by overexpression of the myelin proteolipid protein gene. Neuron 13:427–442. Karthigasan, J., Evans, E. L., Vouyiouklis,D. A., Inouye, H., Borenshteyn, N., Ramamurthy, G. V., Kirschner, D. A., 1996, EVects of rumpshaker mutation on CNS myelin composition and structure. J Neurochem. 66:338–345. Kitagawa, K., Sinoway, M. P., Yang, C., Gould, R. M., Colman, D. R., 1993, A proteolipid protein gene family: Expression in sharks and rays and possible evolution from an ancestral gene encoding a pore-forming polypetide. Neuron 11:433–448. Klugmann, M., Schwab, M., Pu¨hlhofer, A., Schneider, A., Zimmermann, F., GriYths, I. R., Nave, K.-A., 1997, Assembly of CNS myelin in the absence of proteolipid protein. Neuron 18:59–70. Knapp, P. E., Bartlett, W. P., Williams, L. A., Yamada, M., Ikenaka, K., SkoV, R. P., 1999a, Programmed cell death without DNA fragmentation in the jimpy mouse: Secreted factors can enhance survival. Cell Death DiVer. 6:136–145. Knapp, P. E., Ismaili, S., Hauser, K. F., Ghandour, M. S., 1999b, Abnormal Ca(2þ) regulation in oligodendrocytes from the dysmyelinating jimpy mouse. Brain Res 847:332–337. Knapp, P. E., Maderspach, K., Hauser, K. F., 1998, Endogenous opioid system in developing normal and jimpy oligodendrocytes: Mu and kappa opioid receptors mediate diVerential mitogenic and growth responses. Glia 22:189–201. Knapp, P. E., SkoV, R. P., 1987, A defect in the cell cycle of neuroglia in the myelin deWcient jimpy mouse. Dev. Brain Res. 35:301–306. Knapp, P. E., SkoV, R. P., Redstone, D. W., 1986, Oligodendroglial cell death in jimpy mice: An explanation for the myelin deWcit. J. Neurosci. 6:2813–2822. Kobayashi, H., HoVman, E., Marks, H., 1994, The rumpshaker mutation in spastic paraplegia. Nat. Genet. 7:351–352. Koeppen, A. H., Ronca, N. A., GreenWeld, E. A., Hans, M. B., 1987, Defective biosynthesis of proteolipid protein in pelizaeus-Merzbacher disease. Ann. Neurol. 21:159–170. Komaki, H., Sasaki, M., Yamamoto, T., Iai, M., Takashima, S., 1999, Connatal Pelizaeus-Merzbacher disease associated with the jimpy(msd) mice mutation. Pediatr Neurol. 20:309–311. Kraus-Ruppert, R., Herschkowitz, N., Fu¨rst, S., 1973, Morphological studies on neuroglial cells in the corpus callosum of the Jimpy mutant mouse. J Neuropathol Exp Neurol. 32:197–202. Lachapelle, F., Gumpel, M., Baumann, N., 1994, Contribution of transplantations to the understanding of the role of the PLP gene. Neurochem. Res. 19:1083–1090. Lappe-Siefke, C., Goebbels, S., Gravel, M., Nicksch, E., Lee, J., Braun, P. E., GriYths, I. R., Nave, K.-A., 2003, Disruption of the Cnp1 gene uncouples oligodendroglial functions in axonal support and myelination. Nat. Genet. 33:366– 374.

V. ANIMAL MODELS OF HUMAN DISEASE

1139

1140

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

Laursen, R. A., Samiullah, M., Lees, M. B., 1984, The structure of bovine brain myelin proteolipid and its organization in myelin. Proc Natl Acad Sci USA 8:2912–2916. Lees, M,, BrostoV, S. L., 1984, Proteins of myelin. In ‘‘Myelin’’ (P. Morell, ed.), 2nd. ed., pp. 197–224. Plenum Press, New York. Le Goascogne, C., Eychenne, B., Tonon, M. C., Lachapelle, F., Baumann, N., Robel, P., 2000, Neurosteroid progesterone is up-regulated in the brain of jimpy and shiverer mice. Glia 29:14–24. Lipsitz, D., Goetz, B. D., Duncan, I. D., 1998, Apoptotic glial cell death and kinetics in the spinal cord of the myelin-deWcient rat. J. Neurosci. Res. 51:497–507. Macklin, W. B., Gardinier, M. V., King, K. D., Kampf, K., 1987, An AG-GG transition at a splice site in the myelin proteolipid protein gene in jimpy mice results in the removal of an exon. FEBS Lett. 223:417–421. Mastronardi, F. G., Ackerley, C. A., Arsenault, L., Roots, B. I., Moscarello, M. A., 1993, Demyelination in a transgenic mouse: A model for multiple sclerosis. J Neurosci Res. 36:315–324. Mattei, M. G., Alliel, P. M., Dautigny, A., Passage, E., Pham-Dinh, D., Mattei, J. F., Jolles, P., 1986, The gene encoding for the major brain proteolipid (PLP) maps on the q-22 band of the human X chromosome. Hum. Genet. 72:352–353. McLaughlin, M., Hunter, D. J. B., Thomson, C. E., Yool, D., Kirkham, D., Freer, A. A., GriYths, I. R., 2002, Evidence for possible interactions between PLP and DM20 within the myelin sheath. Glia 39:31–36. Meier, C., BischoV, A., 1974, Dysmyelination in ‘‘jimpy’’ mouse. Electron microscopic study. J. Neuropathol. Exp. Neurol. 33:343–353. Meier C, BischoV A., 1975, Oligodendroglial cell development in jimpy mice and controls. An electron-microscopic study in the optic nerve. J Neurol. Sci. 26:517–528. Meier, C., Herschkowitz, N., BischoV, A., 1974, Morphological and biochemical observations in the Jimpy spinal cord. Acta Neuropathol (Berl) 27:349–362. Meier, H., MacPike, A. D., 1970, A neurological mutation (msd) of the mouse causing a deWciency of myelin synthesis. Exp Brain Res. 10:512–525. Merzbacher, L., 1910, Eine eigenartige familia¨re Erkrankungsform (Aplasia axialis extracorticolis congenita). Z. Ges. Neurol. Psychiatr. 3:1–138. Miller, M. J., Haxhiu, M. A., Georgiadis, P., Gudz, T. I., Kangas, C. D., Macklin, W. B., 2003, Proteolipid protein gene mutation induces altered ventilatory response to hypoxia in the myelin-deWcient rat. J Neurosci. 23:2265–2273. Milner, R. J., Lai, C., Nave, K.-A., Lenoir, D., Ogata, J., SutcliVe, J. G., 1985, Nucleotide sequences of two mRNAs for rat brain myelin proteolipid protein. Cell 42:931–939. Mimault, C., Giraud, G., Courtois, V., Cailloux, F., Boire, J. Y., Dastugue, B., BoespXug-Tanguy, O., 1999, Proteolipoprotein gene analysis in 82 patients with sporadic Pelizaeus-Merzbacher Disease: Duplications, the major cause of the disease, originate more frequently in male germ cells, but point mutations do not. The Clinical European Network on Brain Dysmyelinating Disease. Am. J. Hum Genet. 65:360–369. Morello, D., Dautigny, A., Pham-Dinh, D., Jolles, P., 1986, Myelin proteolipid protein (PLP and DM-20) transcripts are deleted in jimpy mutant mice. EMBO J. 5:3489–3493. Nadon, N. L., Arnheiter, H., Hudson, L. D., 1994, A combination of PLP and DM20 transgenes promotes partial myelination in the jimpy mouse. J. Neurochem. 63:822–833. Nadon, N. L., Duncan, I. D., Hudson, L. D., 1990, A point mutation in the proteolipid protein gene of the ‘shaking pup’ interrupts oligodendrocyte development. Development 110:529–537. Nadon, N. L., West, M., 1998, Myelin proteolipid protein: Function in myelin structure is distinct from its role in oligodendrocyte development. Dev Neurosci. 20:533–539. Naidu, S., Dlouhy, S. R., Geraghty, M. T., Hodes, M. E., 1997, A male child with the rumpshaker mutation, X-linked spastic paraplegia/Pelizaeus-Merzbacher disease and lysinuria. J Inherit Metab Dis. 20:811–816. Nave, K.-A., 1994, Neurological mouse mutants and genes of myelin. J. Neurosci. Res. 38:607–612. Nave, K.-A., Lai, C., Bloom, F. E., Milner, R. J., 1986, Jimpy mutant mouse: A 74-base deletion in the mRNA for myelin proteolipid protein and evidence for a primary defect in RNA splicing. Proc. Natl. Acad. Sci. USA 83:9264–9268. Nave, K.-A., Lai, C., Bloom, F. E., and Milner, R. J., 1987, Splice site selection in the proteolipid protein (PLP) gene transcript and primary structure of the DM-20 protein of central nervous system myelin. Proc. Natl. Acad. Sci. USA 84:5665–5669. Nave, K.-A., and BoespXug-Tanguy, O., 1996, Developmental defects of myelin formation: From X-linked mutations to human dysmyelinating diseases. The Neuroscientist 2:33–43. Nave, K.-A., Lai, C., Bloom, F. E., Milner, R. J., 1987, Splice site selection in the proteolipid protein (PLP) gene transcript and primary structure of the DM-20 protein of central nervous system myelin. Proc. Natl. Acad. Sci. USA 84:5665–5669. Nave, K.-A., Milner, R. J., 1989, Proteolipid proteins: Structure and genetic expression in normal and myelindeWcient mutant mice. Crit. Rev. Neurobiol. 5:65–91. Pearsall, G. B., Nadon, N. L., Wolf, M. K., Billings-Gagliardi, S., 1997, Jimpy-4J mouse has a missense mutation in exon 2 of the Plp gene. Dev Neurosci. 19:337–341. ¨ ber eine eigenthu¨mliche Form spastischer La¨hmung mit Cerebral erscheinungen auf Pelizaeus, F., 1885, U heredita¨rer Grundlage (Multiple Sklerose). Arch. Psychiatr. 16:698–710. Phillips, R. J., 1954, Jimpy, a new totally sex-linked gene in the house mouse. Z. Induk. Abstamm. Vererbunsl. 86:322 Popot, J,L., Pham-Dinh, D., Dautigny, A., 1991, Major myelin proteolipid: The 4-a-helix topology. J. Membrane Biol. 120:233–246.

V. ANIMAL MODELS OF HUMAN DISEASE

CONCLUSIONS

Pringle, N..P, Nadon, N. L., Rhode, D. M., Richardson, W. D., Duncan, I. D., 1997, Normal temporal and spatial distribution of oligodendrocyte progenitors in the myelin-deWcient (md) rat. J Neurosci Res. 47:264–270. Pucket, C., Hudson, L., Ono, K., Friedrich, V., Benecke, J., Dubois-Dalcq, M., Lazzarini, R. A., 1987, Myelin speciWc proteolipid protein is expressed in myelinating Schwann cells but not incorporated into myelin sheaths. J. Neurosci. Res. 18:511–518. Raskind, W. H., Williams, C. A., Hudson, L. D., Bird, T. D., 1991, Complete deletion of the proteolipid protein gene (PLP) in a family with X-linked Pelizaeus-Merzbacher disease. Am. J. Med. Genet. 49:1355–1360. Readhead, C., Schneider, A., GriYths, I., Nave, K.-A., 1994, Premature arrest of myelin formation in transgenic mice with increased proteolipoprotein gene dosage. Neuron 12:583–595. Ron, D., 2002, Translational control in the endoplasmic reticulum stress response. J Clin. Invest. 110:1383–1388. Rosenbluth, J., StoVel, W., and SchiV, R., 1996, Myelin structure in proteolipid protein (PLP)-null mouse spinal cord. J. Comp. Neurol. 371:336–344. Saugier-Veber, P., Munnich, A., Bonneau, D., Rozet, J. M., Le Merrer, M., Gil, R., BoespXug-Tanguy, O., 1994, X-linked spastic paraplegia and Pelizaeus-Merzbacher disease are allelic disorders at the proteolipid protein locus. Nat. Genet. 6:257–261. Scherer, S. S., 1997, Molecular genetics of demyelination: New wrinkles on an old membrane. Neuron 18:13–16. Schneider, A., GriYths, I. R., Readhead, C., Nave, K.-A., 1995, Dominant-negative action of the jimpy mutation in mice complemented with an autosomal transgene for myelin proteolipid protein. Proc. Natl. Acad. Sci. USA 92:4447–4451. Schneider, A., Montague, P., GriYths, I., Fanarraga, M., Kennedy, P., Brophy, P., and Nave, K.-A., 1992, Uncoupling of hypomyelination and glial cell death by a mutation in the proteolipid protein gene. Nature 358:758–761. Seeman, P., Paderova, K., Benes, V. Jr, Sistermans, E. A., 2002, A severe connatal form of Pelizaeus Merzbacher disease in a Czech boy caused by a novel mutation (725C>A, Ala242Glu) at the ‘jimpy(msd) codon’ in the PLP gene. Int. J. Mol. Med. 9:125–129. Seitelberger, F. 1970, Pelizaeus-Merzbacher disease. In ‘‘Handbook of Clinical Neurology’’ (P. J. Vinken and G. W. Bruyn, eds.), Vol 10, pp. 150–202. North Holland, Amsterdam. Seitelberger, F., Urbanitz, I., and Nave, K.-A. 1996, Pelizaeus-Merzbacher Disease. In ‘‘Handbook of Clinical Neurology’’ (P. J. Vinken and G. W. Bruyn, eds.), Vol. 22 (Ch. 26), pp. 559–579. Elsevier Science B. V. Amsterdam. Sherman, M. Y., Goldberg, A. L., 2001, Cellular defenses against unfolded proteins: A cell biologist thinks about neurodegenerative diseases. Neuron 29:15–32. Shy, M. E., Hobson, G., Jain, M., BoespXug-Tanguy, O., Garbern, J., Sperle, K., Li, W., Gow, A., Rodriguez, D., Bertini, E., Mancias, P., Krajewski, K., Lewis, R., Kamholz, J., 2003, Schwann cell expression of PLP1 but not DM20 is necessary to prevent neuropathy. Ann Neurol. 53:354–365. Sidman, R. L., Dickie, M. M., Appel, S. H., 1964, Mutant mice (quaking and jimpy) with deWcient myelination in the central nervous system. Science 144:309–311. Simons, M., Kra¨mer, E. M., Macchi, P., Rathke-Hartlieb, S., Trotter, J., Nave, K.-A., Schulz, J. B., 2002, Overexpression of the myelin proteolipid protein leads to accumulation of cholesterol and proteolipid protein in endosomes/lysosomes: Implications for Pelizaeus-Merzbacher disease. J Cell Biol. 157:327–336. Simons, M., Kra¨mer, E. M., Thiele, C., StoVel, W., Trotter, J., 2000, Assembly of myelin by association of proteolipid protein with cholesterol- and galactosylceramide-rich membrane domains. J Cell Biol. 151:143–154. Simons-Johnson, R., Roder, J. C., Riordan, J. R., 1995, Over-expression of the DM-20 myelin proteolipid causes central nervous system demyelination in transgenic mice. J. Neurochem. 64:967–976. Sistermans, E. A., de Coo, R. F., De Wijs, I. J., Van Oost, B. A., 1998, Duplication of the proteolipid protein gene is the major cause of Pelizaeus-Merzbacher disease. Neurology 50:1749–1754. Sistermans, E. A., de Wijs, I. J., de Coo, R..F, Smit, L. M., Menko, F. H., van Oost, B. A., 1996, A (G-to-A) mutation in the initiation codon of the proteolipid protein gene causing a relatively mild form of PelizaeusMerzbacher disease in a Dutch family. Hum Genet. 97:337–339. SkoV, R. P., 1976, Myelin deWcit in the Jimpy mouse may be due to cellular abnormalities in astroglia. Nature 264:560–562. SkoV, R. P., 1982, Increased proliferation of oligodendrocytes in the hypomyelinated mouse mutant-jimpy. Brain Res. 248:19–31. SkoV, R. P., 1995, Programmed cell death in the dysmyelinating mutants. Brain Pathol. 5:283–288. Snipes, G. J., Suter, U., Shooter, E., 1993, The genetics of myelin. Curr. Op. Neurobiol. 3:694–702. Southwood, C., Gow, A., 2001, Molecular pathways of oligodendrocyte apoptosis revealed by mutations in the proteolipid protein gene. Microsc Res Tech. 52:700–708. Southwood, C. M., Garbern, J., Jiang, W., Gow, A., 2002, The unfolded protein response modulates disease severity in Pelizaeus-Merzbacher disease. Neuron 36:585–596. Spassky, N., Goujet-Zalc, C., Parmantier, E., Olivier, C., Martinez, S., Ivanova, A., Ikenaka, K., Macklin, W., Cerruti, I., Zalc, B., Thomas, J. L., 1998, Multiple restricted origin of oligodendrocytes. J. Neurosci. 18:8331–8343. Sporkel, O., Uschkureit, T., Bussow, H., StoVel, W., 2002, Oligodendrocytes expressing exclusively the DM20 isoform of the proteolipid protein gene: Myelination and development. Glia 37:19–30.

V. ANIMAL MODELS OF HUMAN DISEASE

1141

1142

47. MODELS OF PELIZEAUS-MERZBACHER-DISEASE

Stecca, B., Southwood, C. M., Gragerov, A., Kelley, K. A., Friedrich, V. L., Gow, A., 2000, The evolution of lipophilin genes from invertebrates to tetrapods: DM-20 cannot replace proteolipid protein in CNS myelin. J Neurosci. 20:4002–4010. StoVel, W., Hillen, H., Giersiefen, H., 1984, Structure and molecular arrangement of proteolipid protein of central nervous system myelin. Proc. Natl. Acad. Sci. U S A. 81:5012–5016. Thomson, C. E., Anderson, T. J., McCulloch, M. C., Dickinson, P. J., Vouyiouklis, D. A., GriYths, I. R., 1999, The early phenotype associated with the jimpy mutation of the proteolipid protein gene. J. Neurocytol. 28:207–221. Tosic, M., Dolivo, M., Domanska-Janik, K., Matthieu, J. M., 1994, Paralytic tremor (pt): A new allele of the proteolipid protein gene in rabbits. J Neurochem. 63:2210–2216. Trapp, B. D., Nishiyama, A., Cheng, D., Macklin, W., 1997, DiVerentiation and death of premyelinating oligodendrocytes in developing rodent brain. J Cell Biol. 137:459–468. Uschkureit, T., Sporkel, O., Bussow, H., StoVel, W., 2001, Rumpshaker-like proteolipid protein (PLP) ratio in a mouse model with unperturbed structural and functional integrity of the myelin sheath and axons in the central nervous system. Glia 35:63–71. Utzschneider, D., Black, J. A., Kocsis, J. D., 1992, Conduction properties of spinal cord axons in the myelindeWcient rat mutant. Neuroscience 49:221–228. Vela, J. M., Dalmau, I., Gonzalez, B., Castellano, B., 1996, The microglial reaction in spinal cords of jimpy mice is related to apoptotic oligodendrocytes. Brain Res. 712:134–142. Vela, J. M., Gonzalez, B., Castellano, B., 1998, Understanding glial abnormalities associated with myelin deWciency in the jimpy mutant mouse. Brain Res. Rev. 26:29–42. Vermeesch, M. K., Knapp, P. E., SkoV, R. P., Studzinski, D. M., Benjamins, J. A., 1990, Death of individual oligodendrocytes in jimpy brain precedes expression of proteolipid protein. Dev. Neurosci. 12:303–315. Wahle, S., StoVel, W., 1998, Cotranslational integration of myelin proteolipid protein (PLP) into the membrane of endoplasmic reticulum: Analysis of topology by glycosylation scanning and protease domain protection assay. Glia 24:226–235. Weimbs, T., StoVel, W., 1992, Proteolipid protein (PLP) of CNS myelin: Positions of free, disulWde-bonded, and fatty acid thioester-linked cysteine residues and implications for the membrane topology of PLP. Biochemistry 31:12289–12296. Weimbs,T., StoVel, W., 1994, Topology of CNS myelin proteolipid protein: Evidence for the nonenzymatic glycosylation of extracytoplasmic domains in normal and diabetic animals. Biochemistry 33:10408–10415. Werner, H., Dimou, L., Klugmann, M., PfeiVer, S., Nave, K.-A., 2001, Multiple splice isoforms of proteolipid M6B in neurons and oligodendrocytes. Mol. Cell. Neurosci. 18:593–605. Werner, H., Jung, M., Klugmann, M., Sereda, M., GriYths, I., Nave, K.-A., 1998, Mouse models of myelin diseases. Brain Pathol. 8:771–793. Willard, H. F., Riordan, J. R., 1985, Assignment of the gene for myelin proteolipid protein to the X chromosome: Implications for X-linked myelin disorders. Science 230:940–942. Williams, W. C., Gard, A. L., 1997, In vitro death of jimpy oligodendrocytes: Correlation with onset of DM-20/ PLP expression and resistance to oligodendrogliotrophic factors. J Neurosci Res. 50:177–189. Wolf, M. K., Holden, A. B., 1969, Tissue culture analysis of the inherited defect of central nervous system myelination in jimpy mice. J Neuropathol Exp Neurol. 28:195–213. Wolf, M. K., Nunnari, J. N., Billings-Gagliardi, S., 1999, Quaking*shiverer double-mutant mice survive for at least 100 days with no CNS myelin. Dev Neurosci. 21:483–490. Woodward, K., Malcolm, S., 1999, Proteolipid protein gene: Pelizaeus-Merzbacher disease in humans and neurodegeneration in mice. Trends Genet. 15:125–128. Woodward, K., Malcolm, S., 2001, CNS myelination and PLP gene dosage. Pharmacogenomics 2:263–272. Wu, Q., Miller, R. H., RansohoV, R. M., Robinson, S., Bu, J., Nishiyama, A., 2000, Elevated levels of the chemokine GRO-1 correlate with elevated oligodendrocyte progenitor proliferation in the jimpy mutant. J Neurosci. 20:2609–2617. Yamamura, T., Konola, J. T., Wekerle, H., Lees, M. B., 1991, Monoclonal antibodies against myelin proteolipid protein: IdentiWcation and characterization of two major determinants. J Neurochem. 57:1671–1680. Yan, Y., Lagenaur, C., Narayanan, V., 1993, Molecular cloning of M6: IdentiWcation of a PLP/DM-20 gene family. Neuron 11:423–431. Yool, D. A., Edgar, J. M., Montague, P., Malcolm, S., 2000, The proteolipid protein gene and myelin disorders in man and animal models. Hum. Mol. Genet. 9:987–992 Yool, D. A., Klugmann, M., McLaughlin, M., Vouyiouklis, D. A., Dimou, L., Barrie, J. A., McCulloch, M. C., Nave, K.-A., GriYths, I. R., 2001, Myelin proteolipid proteins promote the interaction of oligodendrocytes and axons. J. Neurosci. Res. 63:151–164. Yool, D., Klugmann, M., Barrie, J. A., McCulloch, M. C., Nave, K.-A., GriYths, I. R., 2002, Observations on the structure of myelin lacking the major proteolipid protein. Neuropath. Appl. Neurobiol. 28:75–78. Young, W., Rosenbluth, J., Wojak, J. C., Sakatani, K., 1989, Extracellular potassium activity and axonal conduction in spinal cord of the myelin-deWcient mutant rat. Exp. Neurol. 106:41–51. Yu, W. P., Collarini, E. J., Pringle, N. P., Richardson, W. D., 1994, Embryonic expression of myelin genes: Evidence for a focal source of oligodendrocyte precursors in the ventricular zone of the neural tube. Neuron 12:1353–1362.

V. ANIMAL MODELS OF HUMAN DISEASE

C H A P T E R

48 Models of Charcot-Marie-Tooth Disease Lawrence Wrabetz, Maria Laura Feltri, and Ueli Suter

INTRODUCTION Charcot Marie Tooth (CMT) hereditary neuropathies (including De´je´rine-Sottas syndrome [DSS], congenital hypomyelinating neuropathy [CH], and hereditary neuropathy with liability to pressure palsies [HNPP]) are collectively the most common inherited neuromuscular disorder. They are inherited as dominant or recessive traits and consist of demyelinating (CMT1) or axonal (CMT2) types. Roughly one-half of CMT neuropathies are caused by a 1.4 megabase internal duplication in chromosome 17, containing PMP22 (peripheral myelin protein 22kD gene). Another 20% are caused by speciWc mutations of PMP22, MPZ (myelin protein zero gene), or CX32/GJb1 (Connexin 32 (Cx32) or gap junction beta 1 gene), and rarer missense, nonsense, frameshift, splice site, or other mutations in 11 additional genes. About one-third of all CMT neuropathies have yet to be associated with disease genes (see Chapter 39, ‘‘Inherited Neuropathies,’’ and Berger et al., 2002; Kleopa and Scherer, 2002; Lupski and Garcia, 2001; Wrabetz et al., 2001; for recent reviews and http:// molgen-www.uia.ac.be/CMTMutations for a catalog of mutations). Even if genetically heterogeneous, the CMT phenotype is similar among the majority of patients, including progressive distal weakness and wasting in the limbs with less evident sensory loss, deformities in the feet, and reduced or absent tendon reXexes. Despite the identiWcation of 14 disease genes, our understanding of the diverse pathogeneses of various CMT neuropathies, and how they lead to this common phenotype, is limited and has yet to reveal treatment strategies. Disability correlates best with axonal damage in CMT neuropathies (Berciano et al., 2000; Dyck et al., 1989; Gabreels-Festen et al., 1992; Krajewski et al., 2000), even if the majority begin as demyelinating disorders. Studies of developing nerve suggest that Schwann cells and axons are reciprocally interdependent for trophic support and the determination of their respective phenotypes (see Chapters 1, 2, and 13 for details). Essentially, most if not all CMTs evolve to a disturbance of this Schwann cell/axon unit, rather than isolated damage to myelin-forming Schwann cells or axons. Therefore, to decipher the pathomechanisms of CMT neuropathy, animal models have taken center stage, as it is there that the normal three-dimensional and reciprocal relationships between Schwann cells and axons are obtained, and that chronic secondary changes may be followed. There are four types of animal models of CMT: those resulting from naturally occurring mutations, those resulting from alterations produced through random chemical mutagenesis, those resulting from targeted mutations introduced by homologous recombination in mouse embryonic stem cells (all of these alterations reside at the endogenous gene locus), and random insertion transgenes (here extra copies of the altered gene are inserted away

Myelin Biology and Disorders, Volume 2

1143

Copyright 2004, Elsevier (USA). All rights reserved.

1144

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

from the endogenous locus). Each type of model has distinct advantages, useful in order to examine the two fundamental questions raised by hereditary neuropathies: (1) How are such diverse phenotypes like CMT1, DSS, CH, and CMT2 produced from the same gene? (2) Conversely, how do mutations in these diVerent genes, some of which are expressed speciWcally in Schwann cells, lead to similar axonopathy—the development of which correlates best with disability? Here we review the animal models of CMT neuropathies, the information they have provided about natural history and pathogenesis, and treatment perspectives. See Table 48.1 for a summary list of models.

RODENT MODELS OF CMT PMP22 Animal Models Introduction Mutations aVecting the PMP22 gene (see Chapter 21, ‘‘PMP22 Gene’’) have been associated with various disorders of the peripheral nervous system (PNS) including CMT1A (the most common form of hereditary neuropathies), DSS, CH, and HNPP (see Chapter 39). An intrachromosomal duplication of 1.4 megabases on chromosome 17p11.2-p12, encompassing the PMP22 gene, is the most frequently encountered genetic alteration (Inoue et al., 2001; Nelis et al., 1996). Deletions of the same region and various frame-shift, splice site, and nonsense mutations in the PMP22 gene result in the HNPP phenotype. Some PMP22 missense mutations are associated with CMT1A, or rarely HNPP, but the majority has been linked to the more severe neuropathies of the DSS or CH type. Various animal models for PMP22 point mutations are currently available (Tab. 48.1). Some of them have been found in normal mouse colonies and they provided the initial key to the discovery that PMP22 is a disease gene for motor and sensory neuropathies in human (Suter et al., 1993). Others were isolated as part of a screen for mutagen-induced mutations (Isaacs et al., 2000). Furthermore, a particular mouse line with a large in-frame deletion of PMP22 has been described (Suh et al., 1997). ArtiWcially generated transgenic lines include mice and rats overexpressing PMP22 (Huxley et al., 1996; Huxley et al., 1998; Magyar et al., 1996; Niemann et al., 2000; Norreel et al., 2001; Robertson et al., 2002b; Sancho et al., 1999; Sancho et al., 2001; Sereda et al., 1996) and, although no exact human correlate has been described so far, homozygous Pmp22-null mice which display also a severe demyelinating neuropathy (Adlkofer et al., 1995). In addition, heterozygous Pmp22null (Adlkofer et al., 1997a) and PMP22-antisense mRNA (Maycox et al., 1997) mice have been established as models of HNPP. The generation and analysis of these animals provided the formal proof that PMP22 is the disease-causing and dosage-sensitive gene in CMT1A and HNPP (Suter and Nave, 1999). Moreover, these animals serve as models for the diVerent forms of the disease and are valuable tools to examine disease mechanisms and potential treatment strategies. PMP22 Mutations in Natural Mouse Lines Trembler mouse The spontaneous mouse mutant Trembler (Tr) was described more than 50 years ago (Falconer, 1951); soon thereafter, a second mouse line with a comparable phenotype was isolated at the Jackson Laboratory, Trembler-J (Tr-J) (Henry et al., 1983; Henry and Sidman, 1983). Both mutants show autosomal-dominant inheritance, unsteady gait, and weakness of the hindlimbs. Furthermore, they are aVected by an axial tremor and show stress-induced convulsions—hence the name ‘‘Trembler.’’ However, no abnormalities could be detected in the central nervous system (CNS) of Tr animals (Braverman, 1953). Recent evidence suggests that the ‘‘trembling’’ behavior is caused by neuromyotonia, an increased muscle stiVness due to hyperactivity of motor units, that can also be found in other animal models of demyelinating CMT (Toyka et al., 1997; Zielasek et al., 2000). Interest-

1145

RODENT MODELS OF CMT

TABLE 48.1 Mouse Models of CMT Neuropathies* CMT gene

Mouse (rat)

Mutation

Model

Reference

PMP22

Null(þ/
). or (
/
) Pmp22Antisense PMP22OE rat PMP22OE mouse C61, C22 My41 Conditional Pmp22 OE Trembler TremblerJ TremblerNcnp Trembler m1H Trembler m2H

neoR-ins P0-rPmp22AS transgene Pmp22wt Pmp22wt PMP22wt Pmp22wt Pmp22 cDNAwt G150D L16P delExonIV H12R Y153X

HNPP HNPP CMT1A or CH CH CMT1A or CH CH CMT1A DSS CMT1A CMT1A/DSS? DSS CMT1A/DSS?

(Adlkofer et al., 1997a, 1995) (Maycox et al., 1997) (Sereda et al., 1996) (Magyar et al., 1996) (Huxley et al., 1996, 1998) (Robertson et al., 2002b) (Perea et al., 2001) (Falconer, 1951; Suter et al., 1992b) (Henry et al., 1983; Suter et al., 1992a) (Suh et al., 1997) (Isaacs et al., 2000) (Isaacs et al., 2000)

MPZ

P0 Null (þ/
) P0 Null (
/
) P0OE P0myc S63del-random-ins

Tandem MpzneoR-ins Tandem MpzneoR-ins Mpzwt transgene Mpz5’myc transgene MpzS63del transgene

CMT1B (mild) DSS CH CMT1B with tomacula CMT1B

(Martini et al., 1995) (Giese et al., 1992) (Wrabetz et al., 2000; Yin et al., 2000) (Previtali et al., 2000) (Wrabetz et al., 2002)

GJb1/CX32

Cx32 Null (y/
) R142W-random-ins 175fs

Cx32neoR-ins P0-Cx32R142W transgene P0-Cx32-175fs transgene

CMT1X CMT1X normal

(Anzini et al., 1997; Scherer et al., 1998) (Scherer et al., 1999) (Abel et al., 1999)

EGR2/KROX20

Krox20 Null (
/
)

LacZ/neoR-ins or Cre-ins

CH

(Topilko et al., 1994; Voiculescu et al., 2000)

Krox20lo/lo Krox20(XoxNeo)

loxP(neoR)loxP-ins intron 1 loxP(exon2)loxP(NeoR)loxP

CH ?(hypomorph)

(Nagarajan et al., 2002) (Taillebourg et al., 2002)

PRX

Prx Null (
/
)

PrxneoR-ins

CMT4F

(Gillespie et al., 2000)

KIF1B

KIF1B Null (þ/
)

Kif1BneoR-ins

CMT2A

(Zhao et al., 2001)

NEFL

L394P-random ins

MSV-NeXL394P transgene

CMT2E (severe)

(Lee et al., 1994)

LMNA

Lmna null (
/
)

LmnaneoR-ins

AR-CMT2

(De Sandre-Giovannoli et al., 2002; Sullivan et al., 1999)

???

Clawpaw

unknown

CH with arthrogryposis

(Henry et al., 1991)

*

See text or references for details of transgene structures or targeting alleles. Also see text for follow-up studies on these mice.

ingly, tremor is not uncommon in CMT patients (Cardoso and Jankovic, 1993), and severe muscle cramps have been reported as an associated feature (Hahn et al., 1991). On the molecular level, the Tr mutation is due to the replacement of the small, neutral amino acid glycine by the charged and bulky amino acid aspartic acid at position 150 in the last hydrophobic domain of the PMP22 protein (G150D, Suter et al., 1992b). The same mutation has been found in a family diagnosed with a severe DSS neuropathy (Ionasescu et al., 1997). Intracellularly, the mutation leads to a PMP22 protein traYcking defect (D’Urso et al., 1998; Naef et al., 1997). Furthermore, a particular propensity for aggregation of the Tr protein has been reported, which may contribute to the cellular basis of the phenotype (Tobler et al., 2002). Adult Tr mutants show severe hypomyelination of peripheral nerves (Ayers and Anderson, 1973, 1975, 1976; Henry et al., 1983; Low, 1976a, 1976b, 1977), increased Schwann cell number and Schwann cell proliferation (Perkins et al., 1981; Sancho et al., 2001), and reduced NCV (Low and McLeod, 1975). During development (Fig. 48.1), the onset of myelination is delayed (Henry et al., 1983). Furthermore, structural abnormalities in myelin of heterozygous Tr mice have been described (Kirschner and Sidman, 1976). Homozygous Tr mice have a normal life span, although they lack myelinated Wbers almost completely (Henry and Sidman, 1988). Hypomyelination in Tr is associated with a general downregulation of myelin protein components, including PMP22 (Adlkofer et al., 1997b; Bascles et al., 1992; Costaglioli et al., 2001; Garbay et al., 1995; Garbay and Bonnet, 1992; Maier et al., 2002a; Vallat et al.,

1146

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

FIGURE 48.1 SpeciWc pathological alterations of CMT1A due to PMP22G150D, CMT1A with duplication, and HNPP are reproduced in Trembler mouse (A), the Pmp22 overexpressing rat (B), and heterozygous Pmp22-null mouse (C and D), respectively. (A) Severe hypomyelination is revealed in a semithin section of P18 Trembler peripheral nerve. (B). Typical onion bulbs are detected in peripheral nerve of a 2-month-old rat. (C). An electron micrograph and (D) teased Wber preparation of the quadriceps nerve of adult heterozygous Pmp22-null mice show tomacula. Bar (in D) ¼ 12 mm (A), 4 mm (B), 5 mm (C), and 75 mm (D).

1999) and mislocalization of MAG (Vallat et al., 1999). In addition, the lipid metabolism of myelinated peripheral nerves is strongly altered (Boiron-Sargueil et al., 1995; Garbay et al., 1998; Garbay and Cassagne, 1994; Heape et al., 1995; Heape et al., 1996; Salles et al., 2002; Sargueil et al., 1999). Grafting experiments have demonstrated that the myelination defect in Tr is largely Schwann cell-autonomous (Aguayo et al., 1977), consistent with the high expression of PMP22 in normal myelinating Schwann cells. Nevertheless, there is also a strong eVect of the mutant Tr Schwann cells on the axon (Maier et al., 2002b). This includes decreased axonal caliber, increased neuroWlament density, and decreased slow axonal transport (de Waegh and Brady, 1990; de Waegh and Brady, 1991; de Waegh et al., 1992). In addition, the microtubule cytoskeleton and its associated proteins are altered in composition and phosphorylation (Kirkpatrick and Brady, 1994). Interestingly, xenograft experiments of nerve biopsies from a patient carrying a PMP22 point mutation into nude mice have yielded similar results (Chance and Dyck, 1998; Sahenk et al., 1998), suggesting only a minor role (if any) of neuronally expressed PMP22 in the disease process (De Leon et al., 1994; Parmantier et al., 1997). The Tr mouse has also been used to show that compact myelin is not required for the initial clustering of voltage-gated sodium channels, ankyrinG 480/270 kDa and cell adhesion molecules leading to the formation of the node of Ranvier in peripheral nerves (Lambert et al., 1997). However, the juxtaparanodal potassium channels Kv1.1 and Kv1.2 are redistributed in Tr nerves (Wang et al., 1995). The inXuence of demyelination and dysmyelination on axonal proteins in CMT1A models has not been investigated in detail, although it certainly contributes to the phenotype (Neuberg et al., 1999).

RODENT MODELS OF CMT

Trembler-J mouse A missense mutation exchanging a proline residue for a leucine at position 16 in the Wrst hydrophobic domain of PMP22 has been found in Tr-J mice (L16P, Suter et al., 1992a). The same mutation has also been described in a family with severe Charcot-Marie-Tooth disease type 1 (Valentijn et al., 1992). The Tr-J protein appears to reach the intermediate compartment between the endoplasmic reticulum and the Golgi apparatus but cannot be transported further (D’Urso et al., 1998; Tobler et al., 1999). The mutant protein forms multimers with wild-type PMP22 protein, accumulates partially in aggresomes, and is degraded via the lysosomal and proteasomal pathways (Notterpek et al., 1997; Ryan et al., 2002; Tobler et al., 2002). Furthermore, the Tr-J protein shows a prolonged interaction with the chaperone calnexin and sequesters calnexin into myelin-like intracellular Wgures (Dickson et al., 2002). Thus, the cellular basis for the resulting disease may involve the loss of this folding partner for other proteins. The pathology in adult heterozygous Tr-J mice is qualitatively similar but less pronounced than in Tr (Henry et al., 1983). Thinly myelinated axons, Schwann cell onion bulb formations, abnormalities in myelin compaction, and signs of altered axon-glia interactions are the main features (Heath et al., 1991; Notterpek et al., 1997; Robertson et al., 1997). Whereas the levels of protein components of compact myelin, including PMP22, P0 and MBP, are decreased, MAG protein, thought to mediate Schwann cell/axon interactions, shows only altered glycosylation (Bartoszewics et al., 1996; Inuzuka et al., 1985; Notterpek et al., 1997). Trembler-Ncnp mouse A third spontaneous PMP22 mouse mutant named Tr-Ncnp (Sakai et al., 1999; Suh et al., 1997) is associated with an in-frame deletion of exon IV (Suter et al., 1994). Thus, the resulting PMP22 protein lacks the second and part of the third hydrophobic domain (Taylor et al., 2000). Tr-Ncnp diVers from Tr and Tr-J in that giant vacuoles are present in the sciatic nerve of homozygous animals resembling swollen endoplasmic reticulum of Schwann cells. Furthermore, signiWcant cell death was found in the nerves of these animals, a feature that is also present in other PMP22 mutants (Sancho et al., 2001). ArtiWcial PMP22 Mutants Mutagen-induced PMP22 mutants A large scale N-ethyl-N-nitrosurea (ENU) mutagenesis project in the mouse has yielded two new mouse lines with PMP22 point mutations (Isaacs et al., 2000). The Wrst mutation replaces a histidine at position 12 by an arginine residue (H12R). A mutation at the same position (H12Q) is associated with DSS in humans (Valentijn et al., 1995). As expected, the phenotype of this mouse is pronounced with severe tremor and paucity of myelination—only occasional thin myelin sheaths are found. The second mutant line carries a nonsense mutation at position 153 (Y153X) with no human genetic correlate. These mice are less aVected clinically but have also strongly hypomyelinated peripheral nerves. Models with Increased PMP22 Gene Dosage PMP22 trangenic mouse DiVerent strategies have been employed to generate animal models for CMT1A associated with increased PMP22 gene dosage. Transgenic mice and rats have been generated and analyzed that carry varying numbers of extra copies of the mouse PMP22 gene. Mice with 16 or 30 additional copies display a severe congenital hypomyelinating neuropathy with an almost complete lack of myelinated large caliber axons, associated with pronounced slowing of nerve conduction velocity (NCV) (Magyar et al., 1996). AVected nerves contain an increased number of Schwann cells that are aligned along axons. Cellular onion bulbs are rare but empty basal laminae, probably remnants of degenerated supernumerary Schwann cells and their processes, are common. The mutant Schwann cells are in a promyelination-like state since they express the early Schwann cell markers p75NTR, N-CAM and L1. Thus, mutant Schwann cells are unable to diVerentiate toward the myelinating phenotype. Increased Schwann cell proliferation and Schwann cell death were also found in postnatal development of PMP22 mutant nerves (Sancho et al., 2001). The degree of cell death correlated with increased proliferation (as in homozygous Pmp22-null and Tr mice),

1147

1148

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

implicating general mechanisms that normally regulate Schwann cell number in developing nerve. Thus, a primary inXuence of PMP22 on proliferation and cell death, as suggested by cell culture experiments, seems unlikely in postnatal development. Interestingly, if postnatal rat Schwann cells are tranduced with a viral construct to overexpress PMP22, and allowed to myelinate dorsal root ganglia (DRG) axons in culture, initial myelin spiraling and myelin compaction are normal (D’Urso et al., 1997), whereas animal models suggest that PMP22 overexpression can alter or even prevent normal myelination. Thus, expression of PMP22 in embryonic Schwann cells may be critical (Hagedorn et al., 1999; Paratore et al., 2002), although diVerences in the levels of overexpression in vitro and in vivo can not be excluded as an explanation for these observations. Further analysis revealed that the mutant mice are also aVected by a distally accentuated axonopathy (Sancho et al., 1999). Such axonal loss, most probably due to altered Schwann cell-axon interactions, has also been observed in homozygous Pmp22-null and heterozygous Mpz-null mice (Frei et al., 1999). Degenerating axons seem to be preferentially associated with demyelinating or dysmyelinating Schwann cells. Thus, these animals provide a valuable model in which to identify the responsible signals. A pronounced muscular phenotype is also observed, most likely as a secondary eVect (Maier et al., 2002b). Single atrophied, triangle-shaped muscle Wbers and massive loss of muscle Wbers leading to fascicular muscle atrophy are seen. In atrophied muscles, ultraterminal sprouting and elongated endplates with abnormal morphology as well as polyinnervation of the neuromuscular junction are frequent. It will be an important issue to characterize the nature of this muscular atrophy and the precise process of denervation and reinvervation at the cellular and molecular level. PMP22 transgenic rat The same PMP22 transgene described earlier was used to generate PMP22 overexpressor rats (Niemann et al., 2000; Sereda et al., 1996). A single line with three transgenic Pmp22 copies was derived. The animals develop gait abnormalities caused by a peripheral hypomyelination, Schwann cell hypertrophy, and muscle weakness. In particular, these mutant rats are the only CMT1A model with abundant onion bulb formation (Fig. 48.1), suggesting active demyelination in peripheral nerves, similar to nerves of patients. Intracellular myelin-like Wgures that might be related to the structures described by Dickson et al. (2002) in PMP22-transfected cells (discussed earlier), and in Tr-J, and Tr nerves (Adlkofer et al., 1997b), are also present (Niemann et al., 2000). Myelin abnormalities are more pronounced in ventral than in dorsal roots, as has been noted in other animal models for CMT and CH (Martini, 1997; Wrabetz et al., 2000). If the PMP22 gene dosage is doubled in homozygous transgenic rats, no myelin is formed. PMP22 appears to be blocked in a late Golgi compartment, while the program of myelin gene expression is not aVected (Niemann et al., 2000). Similar to CMT1A patients the visible, electrophysiological, and pathological phenotypes in heterozygous and homozygous PMP22-transgenic rats are highly variable from animal to animal and correlate with variability in PMP22 mRNA overexpression. The molecular basis for this observation remains unclear, but the data are consistent with the results obtained from PMP22 transgenic mice that carry diVerent copy numbers of a yeast artiWcial chromosome (YAC) containing the human PMP22 gene (Huxley et al., 1996). In these mice, varying levels of PMP22 expression correlate with the degree of demyelination and reduced NCV (Huxley et al., 1998; Norreel et al., 2001). As measured by its nuclear localization, cyclin D1 is active as Schwann cells proliferate after nerve injury, but not during development (Atanasoski et al., 2001). What about in CMT1A? Analysis in PMP22-transgenic rats revealed that proliferating Schwann cells associated with demyelinated axons show nuclear cyclin D1 expression (Atanasoski et al., 2002). Thus, these demyelinated axons produce signals in Schwann cells more like those of nerve injury than development. Conditional PMP22 overexpressing mouse Perea and colleagues have recently generated a mouse model with conditional PMP22 overexpression (Perea et al., 2001; Robertson et al., 2002a). They used a tetracyline-regulatable system to demonstrate that PMP22

RODENT MODELS OF CMT

overexpression rapidly induces demyelination even in adult mice that have developed with normal PMP22 expression. This experiment shows conclusively that adult, fully diVerentiated Schwann cells are vulnerable to PMP22 overexpression. Conversely, adult animals that developed with PMP22 overexpression showed rapid, partial remyelination when PMP22 expression was normalized. These results are encouraging from a disease point of view, since they suggest that mutant Schwann cells may not be irreversibly damaged in adult CMT1A patient and might be able to remyelinate if PMP22 overexpression can be corrected. Due to technical limitations of the experimental system, however, it was not possible to examine whether remyelination was speciWc to Schwann cells with normalized PMP22 expression and whether this eVect would also lead to an amelioration of the axonopathy, muscle atrophy, and clinical behavior of the mice. Mice with decreased PMP22 gene dosage Mice with genetic disruption of the PMP22 gene are viable but develop walking diYculties due to progressive weakness of the hind limbs (Adlkofer et al., 1995). Onset of myelination is slightly delayed indicating a crucial role of PMP22 in the initial steps of myelination, a conclusion that is supported by the morphological analysis of PMP22 transgenic and Tr-J mice (Robertson et al., 1999). Myelination is altered by the characteristic formation of mainly paranodal, but also internodal tomacula. Furthermore, the myelin structure appears to be more sensitive to myelin splitting artefacts. Hypermyelination structures (tomacula) are unstable and degenerate with age. Demyelination and remyelination, indicated by thinly myelinated axons and Schwann cell onion bulbs, and associated with very slow NCV, are the predominant feature in older animals although some tomacula can still be found (Adlkofer et al., 1995; Sancho et al., 1999). Homozygous Pmp22 null mice also develop a distal axonopathy with deWnitive signs of active axonal degeneration seen in older animals (Sancho et al., 1999). As a consequence of the neural phenotype, a muscular atrophy arises as indicated by extensive type grouping of muscle Wbers. No obvious pathological alterations have been detected in other tissues that normally express PMP22 although major functions of PMP22 in regulating cell death (Wilson et al., 2002) and as a component of tight junctions have been described (Notterpek et al., 2001). Whether this is due to more subtle eVects yet to be discovered or possible compensatory eVects by other members of the PMP22/EMP/MP20 family remains to be determined. Heterozygous ‘‘knock-out’’ PMP22 (þ/0) mice are an animal model for HNPP and show the characteristic tomacula as observed in human nerves (Adlkofer et al., 1997a, and Fig. 48.1). These structures develop already in young PMP22 (þ/0) mice. With age, some electrophysiological abnormalities can be detected accompanied by a signiWcant number of abnormally swollen and degenerating tomacula. Thinly myelinated axons and Schwann cell onion bulbs appear, indicating ongoing demyelination and remyelination. Such Wndings have also been reproduced in another transgenic HNPP model using the P0 promoter to drive rat antisense PMP22 mRNA expression (Maycox et al., 1997). Similar to the observed progressive degenerative events in the mouse model, some aging HNPP patients develop a chronic form of neuropathy with similarities to CMT1 (Windebank, 1993). Whether the intrinsic instability of tomaculous myelin is also involved in the transient symptoms experienced by HNPP patients after nerve trauma remains to be determined. Finally, available PMP22 alleles were exploited in cross-breeding experiments to generate more information about the genetic mechanism of missense mutants of PMP22 (Adlkofer et al., 1997b). These studies revealed that the Tr allele is able to act as a true gain-of-function mutation on a null background (Tr/0) as well as in homozygous Tr animals (Tr/Tr). That is, the PMP22Tr is able to produce dose-dependent, toxic gain of function in Schwann cells, independent of any eVect on wild-type PMP22 (see Chapter 39 for pathogenetic details).

P0 Animal Models Introduction P0 glycoprotein is the most abundant protein in peripheral nerve myelin (GreenWeld et al., 1973, see Chapter 20, ‘‘The P0 Protein Gene,’’ for details). It is a single pass

1149

1150

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

transmembrane protein with an immunoglobulin(Ig)-like fold, predicting a role in adhesion (Lemke and Axel, 1985). Aggregation assays of transfected cells showed that the extracellular domain of P0 can mediate homophilic intermolecular interactions (D’Urso et al., 1990; Filbin et al., 1990; Schneider-Schaulies et al., 1990). Crystallographic analysis of the extracellular domain suggests that four P0 molecules form a tetramer in cis that interacts homophilically in trans with tetramers in apposing myelin lamellae (Shapiro et al., 1996) to form the intraperiod line. In fact, complete disruption of P0 expression in homozygous Mpz-null mice leads to uncompaction of the myelin sheath, including disruption of the intraperiod line (Giese et al., 1992). MPZ in human maps to chromosome 1 and is mutated in CMT1B. CMT1B cases tend to manifest a more severe phenotype than the median CMT1; more rapidly progressive, with more profoundly slowed NCV (Bird, 1999). Pathological features include hypomyelination, onion bulbs, rarely tomacula, and loss of large Wbers. Thus far, more than 80 mutations have been identiWed in MPZ. They account for 4 to 14% of CMT not due to chromosome 17 duplication in various series (see Table 39.2 in Chapter 39). Missense, nonsense, and frameshift mutations have been reported (reviewed in Nelis et al., 1999), but no deletion or duplication of the gene. The mutations are distributed primarily in the extracellular domain, but also within the transmembrane and intracellular domains. Mutations in MPZ generate perhaps the widest diversity of CMT phenotypes—ranging from mild CMT1B, typical CMT1, or HNPP-like, to the more severe DSS and CH or even CMT2—as well as pathologies—including demyelination, dysmyelination, Xorid tomacula, myelin outfolding, or axonal loss with clusters of regenerating axons. Commonly, speciWc mutations reproducibly generate certain phenotypes. For example, mutations S44F, T124M, D61G, and Y119C usually cause CMT2-like neuropathy (Marrosu et al., 1998; Senderek et al., 2000). How these mutations in a myelin protein inXuence primarly axons is an interesting question. Such diverse phenotypes suggest mutation-speciWc cellular pathomechanisms that can be readily investigated in animal models. Loss-of-Function Models of MPZ-Related Neuropathies Most MPZ mutations produce neuropathy in heterozygotes. Thus, one possible explanation for diverse phenotypes is variable loss of function. At least one MPZ mutation, V102FS, probably represents a complete null allele (Pareyson et al., 1999; Warner et al., 1996), as virtual translation predicts a protein of only 78 aminoacids that contains no transmembrane domain. However, in an extended family, a phenotype in heterozygous parents and grandparents was recognized only after homozygous children presented with DSS. The children showed delayed motor milestones, severe weakness, and very slow NCV of less than 4m/s. Nerve biopsies showed markedly reduced numbers of myelinated Wbers, thin myelin sheaths, and numerous basal lamina onion bulbs. Heterozygous relatives, instead, were asymptomatic with only mildly slowed NCV (median motor 40 to 48m/s) (Sghirlanzoni et al., 1992). These two situations are well modelled by heterozygous Mpz-null and homozygous Mpz-null mice, respectively. Heterozygous-null mice develop a late-onset neuropathy, with thin myelin sheaths and progressive demyelination and subtle onion bulb formation (Fig. 48.2) starting at 4 months of age (Martini et al., 1995). In contrast, homozygousnull mice show severe hypomyelination, with poorly compacted myelin sheaths, and some axons surrounded by reduced or absent wraps of Schwann cell membrane, redundant basal lamina, and with age the formation of rudimentary onion bulbs (Giese et al., 1992; Martini et al., 1995). Consistent with this, NCV are only mildly decreased in heterozygous-null mice, but are severely reduced in homozygous-null mice, with low amplitude compound motor action potentials (CMAP) (Zielasek et al., 1996). Therefore, the majority of MPZrelated neuropathies probably fall in an interval between loss of function of one (mild CMT1B) and two (DSS-like) alleles. This is consistent with dominant-negative gain of abnormal function (Martini, 1997); that is, the mutant allele loses adhesive function and provokes varying loss of function in the partner wild-type allele. A series of experiments in vitro and in vivo (discussed later) provide further support for this hypothesis.

RODENT MODELS OF CMT

FIGURE 48.2 DiVerent pathogenetic mechanisms in transgenic mice reproduce diverse pathological alterations as described in MPZ-related neuropathies. (A) A quadriceps nerve from a 13-month-old heterozygous Mpz-null mouse contains a typical onion bulb with supernumerary Schwann cell processes encircling a thin myelin sheath. (B) A low-power electron micrograph of P0 overexpressor sciatic nerve at P28 shows paucity of myelin, with many Schwann cells arrested at the promyelin stage, occasional bundles of unsorted naked axons and redundant basal lamina (arrow). P0myc mice manifest both tomacula, seen in a teased Wber (C) and uncompaction, as represented in a micrograph (D). Finally, semithin section analysis of P0S63del sciatic nerves at 18 months old reveals Xorid onion bulbs (E). Bar (in E) ¼ 4mm (A), 1.5mm (B), 20mm (C), 1.5mm (D), and 25mm (E).

Since progression of disability in CMT neuropathy correlates with axonal damage, a model of the more severe DSS neuropathy might manifest axonal changes. That CMAP amplitudes were reduced to less than 25% of normal in homozygous-null mice was also suggestive (Martini et al., 1995; Zielasek et al., 1996). Martini and colleagues conWrmed this anatomically, as homozygous-null mice develop a distal axonopathy with loss of distal axons and sensory Merkel cells, and angulated Wbers in muscle suggesting denervation (Frei et al., 1999). Interestingly, an inXammatory inWltration consisting of T lymphocytes and macrophages was also detected in nerves of heterozygous Mpz-null mice (Martini et al., 1995; Shy et al., 1997). Increasing inWltration paralleled progressive demyelination. Indeed, by crossing these mice with mice deWcient in T lymphocytes (Rag1-null or T cell receptor a-null) or in macrophage activation (Osteopetrosis/macrophage colony stimulating factordeWcient), demyelination was ameliorated (Carenini et al., 2001; Schmid et al., 2000). Moreover, reconstitution of the immune system by transplant of wild-type bone marrow into Mpz(þ/
)/Rag1(
/
) mice reproduced the worse demyelinating phenotype of native Mpz(þ/
) mice (Maurer et al., 2001). Taken together with the T lymphocyte and macrophage inWltration also described in Cx32 deWcient mice (Kobsar et al., 2002), these data suggest that immune response may generally modulate the phenotype of hereditary neuropathies. Therefore, immunomodulation therapy in CMT patients may be worth exploring. In addition, by reconstituting speciWc elements (e.g., subsets of antigen-speciWc T cells) of the immune system in Mpz(þ/
)/Rag1(
/
) mice, it may be possible to more speciWcally target immunotherapy (Maurer et al., 2001, 2002). Gain-of-Function Models of MPZ-Related Neuropathies Given that complete haploinsuYcience of P0 in human and mouse produces only mild neuropathy, many MPZ-related neuropathies probably include an additional gain of abnormal function mechanism—either a dominant-negative eVect, most likely originating in the myelin sheath, or a ‘‘toxic’’ eVect of the mutant protein initiating from any intracellular location during the synthesis and traYcking of P0. But how can such gain of function mechanisms be correlated with the diverse MPZ-related phenotypes?

1151

1152

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

Evidence from human nerves, in vitro experiments, and transgenic mice support the notion of gain of function. SpeciWc myelin packing defects aVecting the intraperiod line have been detected in CMT1B nerves associated with several extracellular domain mutations (S63del, R98H, R98C, Kirschner et al., 1996), suggesting that the mutant proteins are expressed and inserted in the myelin sheath. When coexpressed with wild-type P0, either P0 truncated in the cytoplasmic domain or P0 unable to form a disulWde bond in the extracellular domain impaired the adhesion normally mediated by wild-type P0 between transfected CHO cells, demonstrating a dominant-negative gain of function (Wong and Filbin, 1996; Zhang and Filbin, 1998). In contrast to PMP22, there are no naturally occuring mouse mutants with Mpz point mutations. To test for diverse gain of function mechanisms, various MPZ mutations were engineered in the Mpz gene, which was then inserted as a random transgene, in addition to the two endogenous Mpz alleles, in mice. In this way loss of mutant P0 function in these mice should be invisible. As a control, it has been shown that additional copies of wild-type Mpz as a randomly inserted transgene produce P0 overexpression (P0OE) and dose-dependent congenital hypomyelinating neuropathy (CHN) in mice (Wrabetz et al., 2000). In P0OE nerves, many myelin-forming Schwann cells had appropriately ensheathed single axons, but the membrane spiraling around the axon (inner mesaxon) was arrested after 1 to 2 turns (Fig. 48.2) by adhesion and reduced space between the Wrst two wraps, associated with premature arrival of P0 (Yin et al., 2000). This represents a gain of normal function, in that it is homophilic adhesion located in the forming myelin sheath, but at the wrong time. This study also determined that less than 50% overexpression of P0 permitted normal myelination. Also, random insertion of a transgene encoding P0 with a myc epitope tag at the mature amino terminus (P0myc) produced a severe demyelinating neuropathy, even when overexpressed at only 50% relative to endogenous P0 (Previtali et al., 2000). P0myc mice had nerve morphology reminiscent of two subgroups of CMT1B neuropathies, both including amino acid substitutions predicted by the crystal structure of the P0 extracellular domain to be near the amino terminus. Pathological features included tomacula and redundant myelin loops, and uncompaction of the myelin sheath (Fig. 48.2)—neither were seen with P0 wild-type overexpression. P0myc was located by immunoelectronmicroscopy in the myelin sheath, in areas of uncompacted myelin. Moreover, the intraperiod line (where opposing P0ECD interact in trans to compact myelin) was widened, suggesting a steric, dominant-negative, gain of function mechanism. Finally, mice that contain the CMT1B allele MpzS63del (deletion of serine 63) as a random insertion transgene develop adult-onset, hypertrophic, demyelinating neuropathy, with motor NCV reduced to 24 m/sec (Wrabetz et al., 2002, and manuscript in preparation). Numerous, large onion bulbs were observed, perhaps the most Xorid described in an animal model of CMT neuropathy (Fig. 48.2). Of note, onion bulbs were never observed in P0 overexpressor or P0myc mice. Even when the total Mpz dosage was reduced to normal by crossing the MpzS63del transgene into the Mpz(þ/
) background, neuropathy remained, conWrming that it was speciWed by the S63del mutation. This shows that a gain of abnormal function not related to overexpression is suYcient to produce a CMT1B neuropathy in mice. Surprisingly, further analysis showed that the P0S63del mutant protein is retained in the endoplasmic reticulum, is subsequently partially degraded, and does not arrive to the myelin sheath (Wrabetz et al., 2002, and manuscript in preparation). Thus, the P0S63del mouse provides another good model (in addition to Trembler and TremblerJ) in which to ask how mutant myelin proteins can provoke demyelination from an intracellular location away from myelin. Taken together, these gain of function Mpz models strongly suggest that MPZ mutations act through various pathogenetic mechanisms, some mutation speciWc, that originate from various intracellular locations in Schwann cells, in order to produce the remarkably diverse MPZ-related phenotypes.

RODENT MODELS OF CMT

Cx32 Animal Models CX32 mutations were Wrst described in X-linked CMT disease (CMT1X) by BergoVen and colleagues (1993) based on genetic mapping experiments and the discovery of high-level Cx32 expression by myelinating Schwann cells. CMT1X accounts for 7 to 16% of all forms of CMT (Table 39.2, Chapter 39) and is particularly characterized by prominent axonal loss and regenerative clusters, in addition to the classically observed variable demyelination and remyelination (Hahn et al., 1990). This speciWc feature has led to the diagnosis of axonal CMT (CMT2) in some patients (Scherer and Fischbeck, 1999). In peripheral nerves, Cx32 is found in uncompacted domains along the myelin internode—the paranodes and Schmidt-Lanterman incisures (Scherer et al., 1995a). The incisures traverse the compact myelin sheath as a conical spiral and generate a cytoplasmic connection between the outer, perinuclear and the inner, periaxonal Schwann cell cytoplasm. It is thought that supply and retrieval of biosynthetic components and ions take place at the myelin-cytoplasmic channel interface as well as between outer, abaxonal and inner, adaxonal Schwann cell cytoplasm. Functional, reXexive gap junction channels have been described in incisures suggesting the formation of a more direct radial pathway across the myelin sheath (Balice-Gordon et al., 1998). However, the rate of (injected) dye diVusion across the sheath was not disturbed in Cx32-null mice suggesting that other gap junction proteins might be expressed and functional in the myelin sheath of peripheral nerves. Indeed, mouse Cx29 (putative human orthologue: Cx31.3) has been detected in the innermost aspects of the myelin sheath, the paranode, the juxtaparanode, the inner mesaxon and the Schmidt-Lanterman incisures in adult sciatic nerves (Altevogt et al., 2002). Whether this protein is required for eYcient transport through incisures or myelin integrity awaits the corresponding mouse mutant or the discovery of Cx31.3 mutations in human pedigrees. Some CX32 mutations seem to result in null alleles (Fischbeck et al., 1999; Scherer et al., 1999, see Chapter 39); and thus, Cx32-null mice have been analyzed as potential models of CMT1X (Anzini et al., 1997; Nelles et al., 1996; Scherer et al., 1998). This analysis revealed near normal NCV, but with aging, Cx32-null mice developed a late-onset, progressive peripheral neuropathy with characteristics similar to CMT1X. Cellular onion bulbs as signs of demyelination and remyelination were present and teased nerve Wber analysis revealed segmental demyelination. As observed in other CMT1 models, the motor nerves were more aVected than sensory branches. Conspicuously abnormal, enlarged periaxonal collars (Figure 48.3) and cytoplasmic thickenings of other noncompacted compartments, such as Schmidt-Lanterman incisures and paranodal loops, were frequently seen. Disorganization of the inner Schwann cell compartment in Cx32-deWcient mice may reXect the consequence of disturbed communication between the outer and inner cytoplasmic aspects of myelinating Schwann cells. Furthermore, regenerative clusters indicative of axonal degeneration are present. Heterozygous female animals show segmental loss of Cx32 expression and less signs of demyelination than age-matched homozygous mice, consistent with the Wnding that the X-linked Cx32 gene is randomly inactivated. As Wrst observed in Mpz-deWcient mice, macrophages modulate the demyelinating phenotype of Cx32-null mice (Kobsar et al., 2002; Maurer et al., 2002). Gene-expression proWling of Cx32-null mice revealed marked upregulation of the cytoskeletal protein glial Wbrillary acidic protein (GFAP) of yet unknown signiWcance (Nicholson et al., 2001). Altered molecular architecture of peripheral nerves in Cx32-null mice has been reported, but the localization of the axonal proteins Caspr and Kv1.1 and 1.2 appears not to be disrupted (Arroyo et al., 1999; Neuberg et al., 1999). Although Cx32-deWcient mice reXect the CMT1X phenotype quite accurately, there seem to be considerable diVerences among species in the role of Cx32 outside of the PNS. Some CX32 mutations also produce CNS symptoms (Kleopa et al., 2002; Paulson et al., 2002), but no other consistent phenotype has been described. In the Cx32-null mouse, however, disrupted metabolic cooperativity among hepatocytes results in reduced glucose release upon stimulation of the sympathetic nerves in the liver (Nelles et al., 1996).

1153

1154

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

FIGURE 48.3 Alterations at the axoglial junction of CMT1X are reproduced in Cx32-null mice. Note the enlarged periaxonal collar typical of CMT1X Wbers (arrows). Bar ¼ 2 mm.

Furthermore, adult Cx32-deWcient mice are more susceptible to chemically induced tumors and display more spontaneous liver tumors than age-matched wild-type mice (Temme et al., 1997; and for discussion see Willecke et al., 1999). Potential gain-of-function disease mechanisms of several CX32 mutations have been tested by expressing (under the control of the P0 promoter) the mutant allele in myelinating Schwann cells of transgenic mice. This analysis revealed that a frame-shifting mutation at codon 175 (175fs) was only expressed at the mRNA level, apparently without other eVects, consistent with a loss-of-function mechanism (Abel et al., 1999). In contrast, in R142W transgenic mice, the mutant Cx32 accumulated in the perinuclear cytoplasm without reaching the Schmidt-Lanterman incisures or the paranodes (Scherer et al., 1999). In addition, endogenous Cx32 expression was reduced and wild-type Cx32 was retained intracellularly, suggesting a ‘‘dominant-negative’’ eVect of the mutated protein (Fischbeck et al., 1999). Similar to mice completely deWcient in Cx32, R142W-mutant mice developed a late-onset demyelinating neuropathy. It should be noted, however, that a direct ‘‘dominant-negative’’ eVect between mutant and wild-type Cx32 is unlikely to occur in CMT1X patients since there is only one allele expressed per Schwann cell. Nevertheless, the mutated protein may interact with other connexins and generate an adverse eVect. In other myelinopathies, aberrant traYcking of other myelin proteins like proteolipid protein (PLP) (Gow et al., 1998), PMP22 (discussed earlier) and P0 (discussed earlier) has been described and may contribute to the pathogenesis (Southwood et al., 2002). Thus, intracellular accumulation and sequestration of chaperones by mutated proteins (Dickson et al., 2002) might be a common denominator in the pathophysiology of diseases aVecting myelinating cells.

EGR2/Krox20 Animal Models Altered dosage of various myelin genes (including PMP22, MPZ in the PNS, and PLP in the CNS) perturbs normal myelination. So it is not surprising that mutations in a transcription factor, encoded by EGR2, are associated with CMT. EGR2 is a member of the early growth response gene family and encodes the zinc-Wnger transcription factor Krox20. As predicted by segment-speciWc expression of Krox20 in developing hindbrain, Krox20null mice had defective formation of rhombomeres 3 and 5 (Schneider-Maunoury et al., 1993; Swiatek and Gridley, 1993). Unexpectedly, Krox20 null mice also manifested dysmyelination in peripheral nerve (Topilko et al., 1994). Most Krox20-null mice die shortly after birth, before signiWcant peripheral myelination has begun. However, a few homozygous-null mice can survive for as much as 2 weeks after birth (Schneider-Maunoury et al., 1993; Topilko et al., 1994), in a strain-dependent fashion (L. Wrabetz, personal observation). In outlier Krox20-null mice, myelination is blocked at an early stage, in which Schwann cells begin to wrap axons, express early markers such as myelin associated glycoprotein (MAG), but do not induce the expression of other myelin

RODENT MODELS OF CMT

genes, including Mpz and Mbp (myelin basic protein gene) (Topilko et al., 1994). This observation has been further conWrmed in mice homozygous for the hypomorphic Krox20lo gene, a ‘‘knock-down’’ allele created by inserting a neoR gene into the Egr2 intron. Homozygotes routinely survive for 3 to 4 weeks after birth, but still manage essentially no peripheral nerve myelination (Nagarajan et al., 2002, and J. Svaren personal communication). Krox20 is thought to link axonal signals to induction of myelin gene expression and myelination by Schwann cells (Murphy et al., 1996). Indeed, adenoviral overexpression of Krox20 in cultured Schwann cells markedly activates expression of the program of myelinrelated genes, including Pmp22 and Mpz within 24 hours (Nagarajan et al., 2001). Like other EGR family members, Krox20 also contains a conserved R1 repressor domain, bound by the NAB corepressor proteins, that serves to modulate transcriptional activation of Krox20 target genes (Svaren et al., 1998). Six of seven EGR2 mutations are dominant, associated with CMT1, DSS and CH, and fall in the highly conserved zinc Wnger DNA binding domains, supporting that Krox20 transcriptional function is important for myelination, whereas one recessive mutation (I268N), causing CH, falls in the R1 domain preventing interactions with NAB corepressors and superactivating a synthetic Krox20 target promoter by 15-fold in vitro (Warner et al., 1998, 1999, and see Chapter 39). Heterozygous Krox20-null mice show normal myelination (Topilko et al., 1994), whereas dominant EGR2 mutations manifest in heterozygotes, suggesting that these mutations act at least in part though gain of function. In keeping with this idea, the S382R/D383Y mutant inhibits the activation of endogenous myelin genes by wild-type Krox20 in cotransduced Schwann cells (Nagarajan et al., 2001). Therefore, to study the pathomechanisms of these mutations in vivo, it will be necessary to engineer mice with authentic mutations introduced by homologous recombination in ES cells. The recessive I268N mutation is associated with CH in the homozygous state and therefore could be modelled by homozygous Krox20 null mouse. However, unlike the mouse, I268N produces no evidence of brain stem or bone abnormalities in patients, thus it is unlikely to represent a null allele. Instead, I268N may be released from NAB regulation, and derepress a Krox20 target gene (15-fold), such that Warner et al. (1998) proposed a gene dosage neuropathy. PMP22 and MPZ would be good candidate targets, as signiWcant overexpression of either produces a CH-like neuropathy in mice (discussed earlier).

Periaxin-Null Mice Model CMT4F The autosomal recessive demyelinating neuropathies (CMT4) are less common than CMT1 and tend to be more severe (Chapter 39). Out of eight CMT4 loci, Wve disease genes have been identiWed (CMT4A:GDAP1, CMT4B1:MTMR2, CMT4B2:MTMR13/ SBF2, CMT4D:NDRG1, and CMT4F:PRX), and most mutations are thought to produce loss of function. Periaxin is a PDZ domain protein that was identiWed as a Schwann cell-speciWc cytoskeletal protein (Gillespie et al., 1994). Periaxin expression is polarized during development Wrst to the inner cytoplasmic domain of the Schwann cell, facing the axon, and then in mature nerves to the outer cytoplasmic domain, away from the axon (Scherer et al., 1995b). Here it interacts via dystrophin related protein -2 (DRP-2) with the dystroglycan complex, which in turn links laminin-2 in the basal lamina to the actin cytoskeleton in the cytoplasm (Sherman et al., 2001). Furthermore, this complex may also activate signal transduction from laminin to the Schwann cell (Wrabetz and Feltri, 2001). Homozygous or compound heterozygous nonsense and frameshift mutations in PRX have been associated with CMT4F or DSS, as well as milder CMT1-like phenotypes (Boerkoel et al., 2001; Guilbot et al., 2001; Takashima et al., 2002); all are predicted to produce loss of function, conWrmed thus far in one patient whose sural nerve biopsy contained no periaxin (Guilbot et al., 2001; Takashima et al., 2002). Therefore, the Prxnull mouse is a potential animal model of CMT4F due to PRX mutations. Prx-null mice develop a phenotype with peculiar characteristics remarkably similar to those of CMT4F patients (Gillespie et al., 2000). Prx-null mice form normal myelin, but

1155

1156

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

the myelin sheath is unstable and after 6 weeks of age mice develop myelin infolding/ thickening, followed by hypomyelination, segmental demyelination with onion bulbs, and tomacula around ‘‘compressed’’ axons by 6 to 8 months of age. Interestingly, these mice have a more prominent sensory neuropathy relative to motor involvement, in contrast to other myelin mutant mice. Allodynia, hyperalgesia, and neuropathic pain (Gillespie et al., 2000) were documented by reduced conduction velocity of aVerent myelinated Wbers in the saphenous nerve after von Frey Wlament stimulation and increased withdrawal response to mechanical and thermal stimulation. Similarly, CMT4F patients develop a severe neuropathy, with both motor involvement as well as pronounced sensory abnormalities and neuropathic pain, and morphologically, loss of myelinated Wbers, onion bulbs, myelin out- and infolding and axonal compression (see Figure 6 in Guilbot et al., 2001; Takashima et al., 2002).

CMT2 Animal Models Introduction In CMT2 hereditary neuropathies, axons are primarily and more severely aVected. CMT1 and 2 are distinguished by clinical, electrophysiological, and histological characteristics; in particular, by the NCV in the forearm (below 38m/sec in CMT1 and above 38m/sec in CMT2). At least 13 genetic loci have been described, and disease genes have been identiWed in three autosomal dominant forms, KIF1B for CMT2A, RAB7 for CMT2B, and NEFL for CMT2E, and in one autosomal recessive form, LMNA for autosomal recessive (AR)CMT2A. In addition, certain MPZ mutations are associated with a CMT2-like phenotype, and GJb1 mutations produce a CMT1X phenotype with axonal characteristics that in part overlap CMT2 (see Chapter 39 for further discussion). Animal models have been described for KIF1B, NEFL, and LMNA. Kif1B-Null Mice Model CMT2A Kif1B is a member of the kinesin gene superfamily, and produces two alternatively spliced variants, termed a and b, that encode a common N-terminal ATPase motor domain fused to one of two divergent C-terminal cargo domains. A loss of function Q98L mutation in the ATP binding site in the motor domain causes CMT2A (Zhao et al., 2001). Homozygous Kif1b-null mice die from apnea due to loss of CNS neurons. Heterozygous-null mice develop a late-onset, axonal neuropathy conWrming that haploinsuYciency can explain neuropathy in CMT2A. Neuropathy was characterized by impaired motor (rotarod performance) but not sensory (hotplate test) capacity, with reduced CMAP and preserved NCV; no nerve histology was described (Zhao et al., 2001). Kinesin 1Ba carries mitochondria, and via PDZ-containing linker proteins such as PSD-95 and synaptic scaVolding molecule-90, interacts with potassium channels, NMDA receptor subunits, and neuronal nitric oxide synathase; whereas Kinesin 1Bß carries synaptic vesicle precursor proteins—heterozygous-null mice showed decreased axonal transport of the latter (Zhao et al., 2001). It remains to be determined which cargo is important to the pathogenesis of CMT2A (Mok et al., 2002), but in any case this provides a potentially satisfying explanation for length-dependent axonal damage typical of CMT2; the distal part of longer axons should require more eYcient axonal transport. Do NeX-Null Mice Model CMT2E? NeuroWlaments are the major intermediate Wlaments of neurons and axons and are assembled by copolymerization of heavy, medium, and low molecular weight monomers, NF-H, M, and L. NeuroWlaments are an important component of axonal cytoskeleton and determinant of axonal caliber. Decreased neuroWlament phosphorylation and spacing, axonal caliber, and slow axonal transport were detected in dysmyelinated axons of Trembler or MAG-null mice (de Waegh et al., 1992; Yin et al., 1998). In demyelinating neuropathies, these eVects are non-cell autonomous, as they are detected in wild-type mouse axons regenerated through sural nerve xenografts from patients with CMT (Sahenk, 1999; Sahenk et al., 1999). It is perhaps not surprising, therefore, that NEFL, encoding the NF-L subunit, could be a CMT2 disease gene. Multiple NEFL mutations have been identiWed in CMT2E, most

MODELS OF SYNDROMIC NEUROPATHIES WITH CMT-LIKE FEATURES

missense, and several segregating faithfully with axonal neuropathy in multigeneration families (De Jonghe et al., 2001; Jordanova et al., 2003; Mersiyanova et al., 2000; Yoshihara et al., 2002). The genetic mechanism is likely to be gain of abnormal function, as CMT2E segregates as a dominant trait, most mutations are missense, and two CMT2E mutant NF-L subunits disrupt assembly and axonal transport of neuroWlaments in transfected DRG neurons (Brownlees et al., 2002; Perez-Olle et al., 2002). Mouse models relevant to NeX support this idea, as heterozygous NeX-null mice manifest reduced axonal caliber, but do not develop a CMT2-like phenotype (Zhu et al., 1997). In contrast, the expression of an NF-L missense mutant (L394P) in transgenic mice, predicted to alter the coil 2B domain and disrupt neuroWlament assembly, causes a very severe neuronopathy/neuropathy with abnormal gait and weakness (Lee et al., 1994). Finally, a series of transgenic studies of neurodegenerative diseases such as Amyotrophic Lateral Sclerosis have revealed that gain of function mechanisms are far more important than loss of function mechanisms in neuroWlament-related pathological processes (reviewed in Julien et al., 1998). Thus, valid animal models of CMT2E will probably be produced only via introduction of missense NEFL mutations by homologous recombination in ES cells. It is worth emphasizing that disability in CMT neuropathies correlates with axonal damage (see the introduction). Both primary demyelinating and axonal neuropathies produce similar neuroWlament alterations, with consequential eVects on axonal caliber, nerve conduction velocity, and axonal transport that could account for length-dependent axonal damage. Thus, this point of convergence in the pathogenesis of some axonal and demyelinating neuropathies is a valid therapeutic target. Lmna–Null Mice Model AR-CMT2A Neuropathy Lamin A/C isoforms, encoded by LMNA, are widely expressed nuclear envelope proteins. Mutations in LMNA have been associated with various diseases, including limb-girdle and Emery-Dreifuss muscular dystrophies, dilated cardiomyopathy 1A, partial lipodystrophy, and mandibuloacral dysplasia (reviewed in Moir and Spann, 2001). Recently, a homozygous R298C mutation was reported in three families with AR-CMT2A (Chaouch et al., 2003; De Sandre-Giovannoli et al., 2002). R298C probably produces lamin A/C loss of function in nerve, as reevaluation of the homozygous Lmna-null mouse (Sullivan et al., 1999) revealed a peripheral neuropathy, with abnormally stiV gait, weak grip, and loss of myelinated Wbers and more amyelinated axons of increased caliber (De Sandre-Giovannoli et al., 2002). Further characterization will determine whether this fully models the human neuropathy characterized by weakness and atrophy of both distal and proximal muscles, distal loss of sensation, normal motor NCV and reduced or absent sensory nerve action potentials, and histology showing profound loss of large myelinated Wbers, regenerating clusters of axons, and the absence of onion bulbs (Chaouch et al., 2003; De SandreGiovannoli et al., 2002). Most of the other LMNA mutations produce diverse diseases in heterozygosity, suggesting varying gain of function eVects.

MODELS OF SYNDROMIC NEUROPATHIES WITH CMT-LIKE FEATURES Several syndromes include neuropathies with CMT-like features (see Chapter 39 by Wrabetz, et al., for discussion of syndromic neuropathies). The presence or absence of neuropathy in mouse models with diverse mutations in the corresponding genes may provide clues as to how the original mutations in human cause neuropathy.

Desert Hedgehog-Null Mice Desert Hedgehog, together with Sonic Hedgehog and Indian Hedgehog are homologous to the Drosophila Hedgehog segment polarity gene and belong to a family of signaling molecules crucially important for early development (reviewed in Ingham and McMahon,

1157

1158

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

2001). Desert Hedgehog is secreted by Schwann cells in the endoneurium, from where it acts on perineurial cells through the Patched receptor, to regulate perineurial formation (Parmantier et al., 1999). Umehara et al. (2000) described a family with members homozygous for DHH-null mutations, who develop gonadal dysgenesis and a minifascicular neuropathy remarkably similar to that of Dhh-null mice. Numerous minifascicles subdivide the endoneurium, each containing several Schwann cell-axon units surrounded by perineurial-like cells. Further analysis in Dhh-null mice revealed that the formation of epineurium and perineurium nerve sheaths is defective. Perineurial tight junctions are abnormal, resulting in increased permeability of the perineurial blood/nerve barrier. Abnormalities in the endoneurial environment may cause secondary degeneration of nerve Wbers, as patients show a mild reduction in the density of myelinated Wbers (Umehara et al., 2000).

PLP-Null Mice In contrast, the peripheral neuropathy associated with some other human syndromic myelinopathies is not reproduced in animal models. For example, PLP produces two alternatively spliced transcripts that encode PLP and DM20, together the major component of CNS myelin. PLP contains a speciWc 35 amino acid domain not found in DM20. PLP is also expressed by myelin-forming Schwann cells in peripheral nerve, albeit at low levels. Mutations in PLP/DM-20 cause Pelizeaus-Merzbacher disease, an X-linked leukodystrophy (see Chapter 37 for details). Patients with mutations that eliminate the expression of both PLP and DM20 also develop a peripheral neuropathy (Garbern et al., 1997; Shy et al., 2003). Of note, mutations predicted to eliminate PLP, but maintain expression of DM20, cause neuropathy, whereas mutations that preserve the PLP-speciWc domain do not (Shy et al., 2003). This suggests that the PLP-speciWc domain is required for peripheral nerve function. However, even though Plp produces similar isoforms in mouse, Plp/DM20null mice do not develop a peripheral neuropathy (GriYths et al., 1998; Klugmann et al., 1997). Perhaps other lipophilin family members, such as M6B (Werner et al., 2001), may compensate for the absence of PLP in mouse, but not human, nerve.

Heterozygous Sox10-Null Mice and Myelinopathy Sox10 is a transcription factor required for normal development of several neural crest derivatives. Sox10 is expressed at all stages of Schwann cell lineage (Kuhlbrodt et al., 1998), during which it cooperates with other transcription factors, such as Oct6, Pax3, and Egr2, and ultimately activates myelin-speciWc genes as Schwann cells terminally diVerentiate (Kuhlbrodt et al., 1998; Peirano et al., 2000). Homozygous Sox10-null mice do not develop peripheral glia, including Schwann cells (Britsch et al., 2001). In addition, Sox10 is required for normal diVerentiation of oligodendrocytes (Stolt et al., 2002). Heterozygous mutations in SOX10 in humans give rise to Waardenburg-Shah (WS4) syndrome, which presents with intestinal aganglionosis (Hirschprung disease), depigmentation, and deafness (Waardenburg syndrome) (Pingault et al., 1998). Absence of several neural crest derivatives, such as enteric ganglion cells and melanocytes, account for the phenotype, but these patients do not develop myelin-related diseases. Instead, several WS patients with putative dominant-negative mutations (DNA binding domain retained; activation domain altered) in SOX10 had peripheral and sometimes central dysmyelination similar to Pelizeus-Merzbacher disease and CMT1 or CH (Inoue et al., 2002; 1999; Pingault et al., 2001). One hypothesis is that Waardenburg/Hirschprung symptoms derive from haploinsuYciency, whereas the myelinopathies derive from a dominant-negative gain of function for mutated SOX10. In mouse, a spontaneous mutation in Sox10 causes a similar syndrome called Dominant Megacolon (dom). Heterozygous Sox10-null mice manifest pigmentation and megacolon defects, conWrming that these symptoms in dom mice and Waardenburg/ Hirschprung patients represent haploinsuYciency; no peripheral nerve defects have been reported in heterozygous Sox10-null mice (Britsch et al., 2001). As for many CMT

ANIMAL MODELS WITH PATHOLOGICAL SIMILARITY TO HEREDITARY NEUROPATHIES

neuropathies, transgenic mice with myelinopathy-speciWc Sox10 mutations will probably be required to test the dominant-negative hypothesis.

ANIMAL MODELS WITH PATHOLOGICAL SIMILARITY TO HEREDITARY NEUROPATHIES Several animal models phenocopy human neuropathies, but result from genetic alterations not yet described in humans, or are not even associated with a disease gene yet. An example of the former, Mpz overexpression in mice phenocopies CH, with many Schwann cells unable to sort axons, arrested in the promyelin stage, or associated with a thin myelin sheath. These nerves also contain increased endoneurial collagen, and redundant Schwann cell basal lamina, without evidence of myelin destruction or Schwann cell onion bulbs (Fig. 48.2). Finally, the myelination defect improves somewhat with time. All of these features compare well with CH in human (Phillips et al., 1999). However, no corresponding human case of MPZ duplication has been documented, although the pathogenesis of EGR2 mutant I268N may include myelin gene overexpression (see the earlier discussion and Wrabetz et al., 2000). A fascinating example of a genetically unidentiWed model is Claw Paw (clp), a spontaneous autosomal recessive mutant that arose in the C57BL/6J-obese strain and has CH with abnormal posture of the forelimbs (sometimes called arthrogryposis, but in clp/clp mice it does not initially represent Wxed joint contractures, as they retain the full range of passive motion). Nerve histology showed diVuse hypomyelination, with Schwann cells temporarily blocked at the promyelinating stage, followed by improvement and progressive myelination of a subset of axons after the Wrst 2 to 4 weeks of life (Henry et al., 1991). In the adult, clp/clp is one of the few myelin mutants that shows equal or worse dysmyelination in the dorsal as compared to ventral roots. The pathological eVect of clp is at least in part Schwann cell autonomous, where it functions downstream of or in parallel with the Pou3f1/Oct6/SCIP transcription factor (Darbas et al., 2002). Also, clp maps to mouse chromosome 7 (Henry et al., 1991) and has been localized to an interval of less than 1 megabase (J. Bermingham and D. Meijer, personal communication). Screening of candidate genes is underway, and may produce a disease gene for CH in human.

GENERALIZATIONS FROM ANIMAL MODELS What generalizations emerge from the animal models of CMT? First, dys- or demyelinating animals phenocopy corresponding human CMT neuropathies reasonably well (models of axonal neuropathies are too few and too sparsely characterized to form generalizations). SpeciWc morphological alterations are produced in myelin (e.g., tomacula in heterozygous PMP22 null or myelin infolding in homozygous periaxin-null mice) that produce electrophysiological and behavioural alterations consistent with clinical characteristics of associated CMT neuropathies. Moreover, the onset and progression are consistent with, but also Xesh out, the partial data set deduced from clinical studies. So for example, we have learned that tomacula form and then degenerate to produce a more general demyelinating neuropathy in HNPP model mice. Second, CMT animal models have proven genetic cause and revealed cell biological pathogenesis. For example, it is clear that PMP22 overexpression or haploinsuYcience are the relevant causes of CMT1A and HNPP related to chromosome 17 duplication or deletion, respectively. Also some missense mutants of PMP22 or P0 produce a toxic gain of abnormal function related to retention at various points in the biosynthetic pathway, suggesting an important role for altered protein quality control in CMT neuropathies. Third, the analysis of CMT models is beginning to uncover common consequences that may serve as general therapeutic targets. So the immune response to myelin damage is present in CMT1X and CMT1B models and modulates the phenotype. Also, as in human,

1159

1160

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

probably all demyelinating mouse models develop secondary axonal damage. However, it must be noted that axonal damage is underwhelming in these mice as compared to humans, perhaps because mice have shorter nerves and life spans (see Martini, 1997, 2000, and 2001 for further discussion). As mouse models of axonal CMT2 neuropathies are analyzed further, it will be important to compare axonal damage in them to the corresponding human CMT2. Finally, studies of animal models have also indicated potential diYculties for the treatment of CMT neuropathies. Thus, where pathogenesis includes loss of function, replacement therapy will require precise control. Dosage of most myelin related genes is tightly regulated. For example, over- or underexpression of PMP22 or MPZ by only about 50% itself produces neuropathy—a real challenge for the gene therapy vectors currently in use. Second, the cell-autonomous pathogenesis, even for alterations of the same gene, is many times mutation speciWc. Thus, an important challenge is to identify further common consequences that could represent therapeutic targets for all CMT neuropathies.

CONCLUSIONS Animals models relevant to 8 of 14 CMT genes have been characterized. The majority are behaviorally, electrophysiologically, and pathologically similar to the corresponding human diseases. These models have produced new information about the normal function of CMT genes, as well as the natural history of CMT neuropathies, and emphasize that the pathogenesis usually begins cell-autonomously and is often mutation speciWc, even for diverse alterations of the same gene. However, corollary immune responses and initial axonal damage may be common features of CMT neuropathies that are reproduced in animal models of CMT and represent promising targets for the development of new therapies.

Acknowledgments L. W. and M. L. F. have been supported by the National Institutes of Health (NS41319 and NS45630); Telethon, Italy; Great Britain Multiple Sclerosis Society; and the European Community. U. S. has been supported by the Swiss National Science Foundation, the Swiss Muscle Disease Foundation, the National Center of Competence in Research ‘‘Neural Plasticity and Repair,’’ and the Swiss Bundesamt for Science related to the Commission of the European Communities, speciWc Research, Technological Development and Demonstration program ‘‘Quality of Life and Management of Living Resources,’’ QLK6-CT-2000-00179. We thank our many colleagues whose work has contributed to this area. We also thank Rudolf Martini, Igor Kobsar, and Michael Sereda for the gift of micrographs.

References Abel, A., Bone, L. J., Messing, A., Scherer, S. S., and Fischbeck, K. H. (1999). Studies in transgenic mice indicate a loss of connexin32 function in X-linked Charcot-Marie-Tooth disease. J Neuropathol Exp Neurol, 58, 702–710. Adlkofer, K., Frei, R., Neuberg, D. H., Zielasek, J., Toyka, K. V., and Suter, U. (1997a). Heterozygous peripheral myelin protein 22-deWcient mice are aVected by a progressive demyelinating tomaculous neuropathy. J Neurosci, 17, 4662–4671. Adlkofer, K., Martini, R., Aguzzi, A., Zielasek, J., Toyka, K. V., and Suter, U. (1995). Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deWcient mice. Nat Genet, 11, 274–280. Adlkofer, K., Naef, R., and Suter, U. (1997b). Analysis of compound heterozygous mice reveals that the Trembler mutation can behave as a gain-of-function allele. J Neurosci Res, 49, 671–680. Aguayo, A. J., Attiwell, M., Trecarten, J., Perkins, S., and Bray, G. M. (1977). Abnormal myelination in transplanted Trembler mouse Schwann cells. Nature, 265, 73–75. Altevogt, B. M., Kleopa, K. A., Postma, F. R., Scherer, S. S., and Paul, D. L. (2002). Connexin29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J Neurosci, 22, 6458–6470.

CONCLUSIONS

Anzini, P., Neuberg, D. H., Schachner, M., Nelles, E., Willecke, K., Zielasek, J., Toyka, K. V., Suter, U., and Martini, R. (1997). Structural abnormalities and deWcient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin 32. J Neurosci, 17, 4545–4551. Arroyo, E. J., Xu, Y. T., Zhou, L., Messing, A., Peles, E., Chiu, S. Y., and Scherer, S. S. (1999). Myelinating Schwann cells determine the internodal localization of Kv1.1, Kv1.2, Kvbeta2, and Caspr. J Neurocytol, 28, 333–347. Atanasoski, S., Scherer, S. S., Nave, K. A., and Suter, U. (2002). Proliferation of Schwann cells and regulation of cyclin D1 expression in an animal model of Charcot-Marie-Tooth disease type 1A. J Neurosci Res, 67, 443–449. Atanasoski, S., Shumas, S., Dickson, C., Scherer, S. S., and Suter, U. (2001). DiVerential cyclin D1 requirements of proliferating Schwann cells during development and after injury. Mol Cell Neurosci, 18, 581–592. Ayers, M. M., and Anderson, R. M. (1973). Onion bulb neuropathy in the Trembler mouse: A model of hypertrophic interstitial neuropathy (Dejerine-Sottas). in man. Acta Neuropathol, 25, 54–70. Ayers, M. M., and Anderson, R. M. (1975). Development of onion bulb neuropathy in the Trembler mouse. Comparison with normal nerve maturation. Acta Neuropathol, 32, 43–59. Ayers, M. M., and Anderson, R. M. (1976). Development of onion bulb neuropathy in the Trembler mouse. Morphometric study. Acta Neuropathol, 36, 137–152. Balice-Gordon, R. J., Bone, L. J., and Scherer, S. S. (1998). Functional gap junctions in the schwann cell myelin sheath. J Cell Biol, 142, 1095–1104. Bartoszewics, Z. P., Lauter, C. J., and Quarles, R. H. (1996). The myelin-associated glycoprotein of the peripheral nervous system in trembler mutants contains increased a2–3 sialic acid and galactose. J Neurosci Res, 43, 587–593. Bascles, L., Bonnet, J., and Garbay, B. (1992). Expression of the PMP-22 gene in Trembler mutant mice: Comparison with the other myelin protein genes. Dev Neurosci, 14, 336–341. Berciano, J., Garcia, A., Calleja, J., and Combarros, O. (2000). Clinico-electrophysiological correlation of extensor digitorum brevis muscle atrophy in children with charcot-marie-tooth disease 1A duplication. Neuromuscul Disord, 10, 419–424. Berger, P., Young, P., and Suter, U. (2002). Molecular cell biology of Charcot-Marie-Tooth disease. Neurogenetics, 4, 1–15. BergoVen, J., Scherer, S. S., Wang, S., Scott, M. O., Bone, L. J., Paul, D. L., Chen, K., Lensch, M. W., Chance, P. F., and Fischbeck, K. H. (1993). Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science, 262, 2039–2042. Bird, T. D. (1999). Historical perspective of deWning Charcot-Marie-Tooth type 1B. Ann N Y Acad Sci, 883, 6–13. Boerkoel, C. F., Takashima, H., Stankiewicz, P., Garcia, C. A., Leber, S. M., Rhee-Morris, L., and Lupski, J. R. (2001). Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am J Hum Genet, 68, 325–333. Boiron-Sargueil, F., Heape, A., Garbay, B., and Cassagne, C. (1995). Cerebroside formation in the peripheral nervous system of normal and Trembler mice. Neurosci Lett, 186, 21–24. Braverman, I. M. (1953). Neurological actions caused by the mutant gene ‘‘Trembler’’ in the house mouse (Mus musculus L.): An investigation. J Neuropathol Exp Neurol, 12, 64–72. Britsch, S., Goerich, D. E., Riethmacher, D., Peirano, R. I., Rossner, M., Nave, K. A., Birchmeier, C., and Wegner, M. (2001). The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev, 15, 66–78. Brownlees, J., Ackerley, S., Grierson, A. J., Jacobsen, N. J., Shea, K., Anderton, B. H., Leigh, P. N., Shaw, C. E., and Miller, C. C. (2002). Charcot-Marie-Tooth disease neuroWlament mutations disrupt neuroWlament assembly and axonal transport. Hum Mol Genet, 11, 2837–2844. Cardoso, F. E., and Jankovic, J. (1993). Hereditary motor-sensory neuropathy and movement disorders. Muscle Nerve, 16, 904–910. Carenini, S., Maurer, M., Werner, A., Blazyca, H., Toyka, K. V., Schmid, C. D., Raivich, G., and Martini, R. (2001). The role of macrophages in demyelinating peripheral nervous system of mice heterozygously deWcient in P0. J Cell Biol, 152, 301–308. Chance, P. F., and Dyck, P. J. (1998). Hereditary neuropathy with liability to pressure palsies: A patient’s point mutation in a mouse model. Neurology, 51, 664–665. Chaouch, M., Allal, Y., De Sandre-Giovannoli, A., Vallat, J. M., Amer-el-Khedoud, A., Kassouri, N., Chaouch, A., Sindou, P., Hammadouche, T., Tazir, M., Levy, N., and Grid, D. (2003). The phenotypic manifestations of autosomal recessive axonalCharcot-Marie-Tooth due to a mutation in Lamin A/C gene. Neuromuscul Disord, 13, 60–67. Costaglioli, P., Come, C., Knoll-Gellida, A., Salles, J., Cassagne, C., and Garbay, B. (2001). The homeotic protein dlk is expressed during peripheral nerve development. FEBS Lett, 509, 413–416. D’Urso, D., Brophy, P. J., Staugaitis, S. M., Gillespie, C. S., Frey, A. B., Stempak, J. G., and Colman, D. R. (1990). Protein zero of peripheral nerve myelin: Biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron, 4, 449–460. D’Urso, D., Prior, R., Greiner-Petter, R., Gabreels-Festen, A. A., and Muller, H. W. (1998). Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutations of the peripheral myelin protein PMP22. J Neurosci, 18, 731–740. D’Urso, D., Schmalenbach, C., Zoidl, G., Prior, R., and Muller, H. W. (1997). Studies on the eVects of altered PMP22 expression during myelination in vitro. J Neurosci Res, 48, 31–42.

1161

1162

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

Darbas, A., Jaegle, M., Walbeehm, E., Grosveld F, E., and Meijer, D. (2002). Schwann cell autonomy of the claw paw gene. Glia, 38, S68. De Jonghe, P., Mersivanova, I., Nelis, E., Del Favero, J., Martin, J. J., Van Broeckhoven, C., Evgrafov, O., and Timmerman, V. (2001). Further evidence that neuroWlament light chain gene mutations can cause CharcotMarie-Tooth disease type 2E. Ann Neurol, 49, 245–249. De Leon, M., Nahin, R. L., Mendoza, M. E., and Ruda, M. A. (1994). SR13/PMP-22 expression in rat nervous system, in PC12 cells, and C6 glial cell lines. J Neurosci Res, 38, 167–181. De Sandre-Giovannoli, A., Chaouch, M., Kozlov, S., Vallat, J. M., Tazir, M., Kassouri, N., Szepetowski, P., Hammadouche, T., Vandenberghe, A., Stewart, C. L., Grid, D., and Levy, N. (2002). Homozygous defects in LMNA, encoding lamin A/C nuclear-envelope proteins, cause autosomal recessive axonal neuropathy in human (Charcot-Marie-Tooth disorder type 2)., and mouse. Am J Hum Genet, 70, 726–736. de Waegh, S., and Brady, S. T. (1990). Altered slow axonal transport and regeneration in a myelin-deWcient mutant mouse: The trembler as an in vivo model for Schwann cell-axon interactions. J Neurosci, 10, 1855–1865. de Waegh, S. M., and Brady, S. T. (1991). Local control of axonal properties by Schwann cells: NeuroWlaments and axonal transport in homologous and heterologous nerve grafts. J Neurosci Res, 30, 201–212. de Waegh, S. M., Lee, V. M., and Brady, S. T. (1992). Local modulation of neuroWlament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell, 68, 451–463. Dickson, K. M., Bergeron, J. J., Shames, I., Colby, J., Nguyen, D. T., Chevet, E., Thomas, D. Y., and Snipes, G. J. (2002). Association of calnexin with mutant peripheral myelin protein-22 ex vivo: A basis for ‘‘gain-offunction’’ ER diseases. Proc Natl Acad Sci U S A, 99, 9852–9857. Dyck, P. J., Karnes, J. L., and Lambert, E. H. (1989). Longitudinal study of neuropathic deWcits and nerve conduction abnormalities in hereditary motor and sensory neuropathy type 1. Neurology, 39, 1302–1308. Falconer, D. S. (1951). Two new mutants, ‘‘Trembler’’ and ‘‘Reeler’’ with neurological actions in the house mouse. Genetics, 50, 192–201. Filbin, M., Walsh, F., Trapp, B., Pizzey, J., and Tennekoon, G. (1990). Role of myelin Po protein as a homophilic adhesion molecule. Nature, 344, 871–872. Fischbeck, K. H., Abel, A., Lin, G. S., and Scherer, S. S. (1999). X-linked Charcot-Marie-Tooth disease and connexin32. Ann N Y Acad Sci, 883, 36–41. Frei, R., Motzing, S., Kinkelin, I., Schachner, M., Koltzenburg, M., and Martini, R. (1999). Loss of distal axons and sensory Merkel cells and features indicative of muscle denervation in hindlimbs of P0-deWcient mice. J Neurosci, 19, 6058–6067. Gabreels-Festen, A. A., Joosten, E. M., Gabreels, F. J., Jennekens, F. G., and Janssen-van Kempen, T. W. (1992). Early morphological features in dominantly inherited demyelinating motor and sensory neuropathy (HMSN type I). J Neurol Sci, 107, 145–154. Garbay, B., Boiron-Sargueil, F., and Cassagne, C. (1995). Expression of the exon 1A-containing PMP22 transcript is altered in the trembler mouse. Neurosci Lett, 198, 157–160. Garbay, B., Boiron-Sargueil, F., Shy, M., Chbihi, T., Jiang, H., Kamholz, J., and Cassagne, C. (1998). Regulation of oleoyl-CoA synthesis in the peripheral nervous system: Demonstration of a link with myelin synthesis. J Neurochem, 71, 1719–1726. Garbay, B., and Bonnet, J. (1992). P0 protein in normal, trembler heterozygous/homozygous mice during active PNS myelination. Neuroreport, 3, 594–596. Garbay, B., and Cassagne, C. (1994). Expression of the ceramide galactosyltransferase gene during myelination of the mouse nervous system. Comparison with the genes encoding myelin basic proteins, choline kinase and CTP:phosphocholine cytidyltransferase. Dev Brain Res, 83, 119–124. Garbern, J. Y., Cambi, F., Tang, X. M., Sima, A. A., Vallat, J. M., Bosch, E. P., Lewis, R., Shy, M., Sohi, J., Kraft, G., Chen, K. L., Joshi, I., Leonard, D. G., Johnson, W., Raskind, W., Dlouhy, S. R., Pratt, V., Hodes, M. E., Bird, T., and Kamholz, J. (1997). Proteolipid protein is necessary in peripheral as well as central myelin. Neuron, 19, 205–218. Giese, K. P., Martini, R., Lemke, G., Soriano, P., and Schachner, M. (1992). Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell, 71, 565–576. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J. (1994). Periaxin, a novel protein of myelinating Schwann cells with a possible role in axonal ensheathment. Neuron, 12, 497–508. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., Cottrell, D. F., Tait, S., Garry, E. M., Wallace, V. C., Ure, J., GriYths, I. R., Smith, A., and Brophy, P. J. (2000). Peripheral demyelination and neuropathic pain behavior in periaxin-deWcient mice. Neuron, 26, 523–531. Gow, A., Southwood, C. M., and Lazzarini, R. A. (1998). Disrupted proteolipid protein traYcking results in oligodendrocyte apoptosis in an animal model of Pelizaeus-Merzbacher disease. J Cell Biol, 140, 925–934. GreenWeld, S., BrostoV, S., Eylar, E. H., and Morell, P. (1973). Protein composition of myelin of the peripheral nervous system. J. Neurochem, 20, 1207–1216. GriYths, I., Klugmann, M., Anderson, T., Yool, D., Thomson, C., Schwab, M. H., Schneider, A., Zimmermann, F., McCulloch, M., Nadon, N., and Nave, K. A. (1998). Axonal swellings and degeneration in mice lacking the major proteolipid of myelin. Science, 280, 1610–1613. Guilbot, A., Williams, A., Ravise, N., Verny, C., Brice, A., Sherman, D. L., Brophy, P. J., LeGuern, E., Delague, V., Bareil, C., Megarbane, A., and Claustres, M. (2001). A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum Mol Genet, 10, 415–421.

CONCLUSIONS

Hagedorn, L., Suter, U., and Sommer, L. (1999). P0 and PMP22 mark a multipotent neural crest-derived cell type that displays community eVects in response to TGF-beta family factors. Development, 126, 3781–3794. Hahn, A. F., Parkes, A. W., Bolton, C. F., and Stewart, S. A. (1991). Neuromyotonia in hereditary motor neuropathy. J Neurol Neurosurg Psychiatry, 54, 230–235. Heape, A. M., Bessoule, J. J., Boiron-Sargueil, F., and Cassagne, C. (1996). Pathways of incorporation of fatty acid into glycerolipids of the murine peripheral nervous system in vivo: Alterations in the dysmyelinating mutant trembler mouse. Neurochem Int, 29, 607–622. Heape, A. M., Bessoule, J.-J., Boiron-Sargueil, F., Garbay, B., and Cassagne, C. (1995). Sphingolipid metabolic disorders in Trembler mouse peripheral nerves in vivo result from an abnormal substrate supply. J Neurochem, 65, 1665–1673. Heath, J. W., Inuzuka, T., Quarles, R. H., and Trapp, B. D. (1991). Distribution of P0 protein and the myelinassociated glycoprotein in peripheral nerves from Trembler mice. J Neurocytol, 20, 439–449. Henry, E. W., Cowen, J. S., and Sidman, R. L. (1983). Comparison of trembler and trembler-J phenotypes: Varying severity of peripheral hypomyelination. J Neuropathol Exp Neurol, 42, 688–706. Henry, E. W., Eicher, E. M., and Sidman, R. L. (1991). The mouse mutation claw paw: Forelimb deformity and delayed myelination throughout the peripheral nervous system. J Hered, 82, 287–294. Henry, E. W., and Sidman, R. L. (1983). The murine mutation trembler-J: Proof of semi-dominant expression by the use of the linked vestigial tail marker. J Neurogen, 1, 39–52. Henry, E. W., and Sidman, R. L. (1988). Long lives for homozygous trembler mutant mice despite virtual absence of peripheral myelin. Science, 241, 344–346. Huxley, C., Passage, E., Manson, A., Putzu, G., Figarella-Branger, D., Pellissier, J. F., and Fontes, M. (1996). Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear injection of human YAC DNA. Hum Mol Genet, 5, 563–569. Huxley, C., Passage, E., Robertson, A. M., Youl, B., Huston, S., Manson, A., Saberan-Djoniedi, D., FigarellaBranger, D., Pellissier, J. F., Thomas, P. K., and Fontes, M. (1998). Correlation between varying levels of PMP22 expression and the degree of demyelination and reduction in nerve conduction velocity in transgenic mice. Hum Mol Genet, 7, 449–458. Ingham, P. W., and McMahon, A. P. (2001). Hedgehog signaling in animal development: Paradigms and principles. Genes Dev, 15, 3059–3087. Inoue, K., Dewar, K., Katsanis, N., Reiter, L. T., Lander, E. S., Devon, K. L., Wyman, D. W., Lupski, J. R., and Birren, B. (2001). The 1.4-Mb CMT1A duplication/HNPP deletion genomic region reveals unique genome architectural features and provides insights into the recent evolution of new genes. Genome Res, 11, 1018–1033. Inoue, K., Shilo, K., Boerkoel, C. F., Crowe, C., Sawady, J., Lupski, J. R., and Agamanolis, D. P. (2002). Congenital hypomyelinating neuropathy, central dysmyelination, and Waardenburg-Hirschsprung disease: Phenotypes linked by SOX10 mutation. Ann Neurol, 52, 836–842. Inoue, K., Tanabe, Y., and Lupski, J. R. (1999). Myelin deWciencies in both the central and the peripheral nervous systems associated with a SOX10 mutation. Ann Neurol, 46, 313–318. Inuzuka, T., Quarles, R. H., Heath, J., and Trapp, B. D. (1985). Myelin-associated glycoprotein and other proteins in Trembler mice. J Neurochem, 44, 793–797. Ionasescu, V. V., Searby, C. C., Ionasescu, R., Chatkupt, S., Patel, N., and Koenigsberger, R. (1997). DejerineSottas neuropathy in mother and son with same point mutation of PMP22 gene. Muscle Nerve, 20, 97–99. Isaacs, A. M., Davies, K. E., Hunter, A. J., Nolan, P. M., Vizor, L., Peters, J., Gale, D. G., Kelsell, D. P., Latham, I. D., Chase, J. M., Fisher, E. M., Bouzyk, M. M., Potter, A., Masih, M., Walsh, F. S., Sims, M. A., Doncaster, K. E., Parsons, C. A., Martin, J., Brown, S. D., Rastan, S., Spurr, N. K., and Gray, I. C. (2000). IdentiWcation of two new Pmp22 mouse mutants using large-scale mutagenesis and a novel rapid mapping strategy. Hum Mol Genet, 9, 1865–1871. Jordanova, A., De Jonghe, P., Boerkoel, C. F., Takashima, H., De Vriendt, E., Ceuterick, C., Martin, J. J., Butler, I. J., Mancias, P., Papasozomenos, S. C., Terespolsky, D., Potocki, L., Brown, C. W., Shy, M., Rita, D. A., Tournev, I., Kremensky, I., Lupski, J. R., and Timmerman, V. (2003). Mutations in the neuroWlament light chain gene (NEFL). cause early onset severe Charcot-Marie-Tooth disease. Brain, 126, 590–597. Julien, J. P., Couillard-Despres, S., and Meier, J. (1998). Transgenic mice in the study of ALS: The role of neuroWlaments. Brain Pathol, 8, 759–769. Kirkpatrick, L. L., and Brady, S. T. (1994). Modulation of the axonal microtubule cytoskeleton by myelinating Schwann cells. J Neurosci, 14, 7440–7450. Kirschner, D. A., and Sidman, R. L. (1976). X-ray diVraction study of myelin structure in immature and mutant mice. Biochim Biophys Acta, 448, 73–87. Kirschner, D. A., Szumowski, K., Gabreel-Festen, A. A. W. M., J. E., H., and P. A., B. (1996). Inherited demyelinating peripheral neuropathies: Relating myelin packing abnormalities to Po molecular defects. J. Neurosci. Res., 46, 502–508. Kleopa, K. A., and Scherer, S. S. (2002). Inherited neuropathies. Neurol Clin, 20, 679–709. Kleopa, K. A., Yum, S. W., and Scherer, S. S. (2002). Cellular mechanisms of connexin32 mutations associated with CNS manifestations. J Neurosci Res, 68, 522–534. Klugmann, M., Schwab, M. H., Puhlhofer, A., Schneider, A., Zimmermann, F., GriYths, I. R., and Nave, K. A. (1997). Assembly of CNS myelin in the absence of proteolipid protein. Neuron, 18, 59–70. Kobsar, I., Maurer, M., Ott, T., and Martini, R. (2002). Macrophage-related demyelination in peripheral nerves of mice deWcient in the gap junction protein connexin 32. Neurosci Lett, 320, 17–20.

1163

1164

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

Krajewski, K. M., Lewis, R. A., Fuerst, D. R., Turansky, C., Hinderer, S. R., Garbern, J., Kamholz, J., and Shy, M. E. (2000). Neurological dysfunction and axonal degeneration in Charcot-Marie-Tooth disease type 1A. Brain, 123, 1516–1527. Kuhlbrodt, K., Herbarth, B., Sock, E., Hermans-Borgmeyer, I., and Wegner, M. (1998). Sox10, a novel transcriptional modulator in glial cells. J Neurosci, 18, 237–250. Lambert, S., Davis, J. Q., and Bennett, V. (1997). Morphogenesis of the node of Ranvier: Co-clusters of ankyrin and ankyrin-binding integral proteins deWne early developmental intermediates. J Neurosci, 17, 7025–7036. Lee, M. K., Marszalek, J. R., and Cleveland, D. W. (1994). A mutant neuroWlament subunit causes massive, selective motor neuron death: Implications for the pathogenesis of human motor neuron disease. Neuron, 13, 975–988. Lemke, G., and Axel, R. (1985). Isolation and sequence of a cDNA encoding the major structural protein of peripheral nerve myelin. Cell, 40, 501–508. Low, P. A. (1976a). Hereditary hypertrophic neuropathy in the trembler mouse: Part 1: Histopathological studies: Light microscopy. J Neurol Sci, 30, 327–341. Low, P. A. (1976b). Hereditary hypertrophic neuropathy in the trembler mouse: Part 2: Histopathological studies: Electron microscopy. J Neurol Sci, 30, 343–368. Low, P. A. (1977). The evolution of ‘onion bulbs’ in the hereditary hypertrophic neuropathy of the Trembler mouse. Neuropath Appl Neurobiol, 3, 81–92. Low, P. A., and McLeod, J. G. (1975). Hereditary demyelinating neuropathy in the trembler mouse. J Neurol Sci, 26, 565–575. Lupski, J., and Garcia, C. (2001). Charcot-Marie-Tooth peripheral neuropathies and related disorders. In ‘‘The Metabolic & Molecular Bases of Inherited Disease’’ (C. R. Scriver, A. L. Beaudet, W. S. Sly, D. Valle, B. Childs, K. W. Kinzler, and B. Vogelstein, eds.), Vol. 3, pp. 5759–5788. McGraw-Hill, New York. Magyar, J. P., Martini, R., Ruelicke, T., Aguzzi, A., Adlkofer, K., Dembic, Z., Zielasek, J., Toyka, K. V., and Suter, U. (1996). Impaired diVerentiation of Schwann cells in transgenic mice with increased PMP22 gene dosage. J Neurosci, 16, 5351–5360. Maier, M., Berger, P., Nave, K. A., and Suter, U. (2002a). IdentiWcation of the regulatory region of the peripheral myelin protein 22 (PMP22). gene that directs temporal and spatial expression in development and regeneration of peripheral nerves. Mol Cell Neurosci, 20, 93–109. Maier, M., Berger, P., and Suter, U. (2002b). Understanding Schwann cell-neurone interactions: The key to Charcot-Marie-Tooth disease? J Anat, 200, 357–366. Marrosu, M. G., Vaccargiu, S., Marrosu, G., Vannelli, A., Cianchetti, C., and Muntoni, F. (1998). CharcotMarie-Tooth disease type 2 associated with mutation of the myelin protein zero gene. Neurology, 50, 1397–1401. Martini, R. (1997). Animal models for inherited peripheral neuropathies. J Anat, 191, 321–336. Martini, R. (2000). Animal models for inherited peripheral neuropathies: Chances to Wnd treatment strategies? J Neurosci Res, 61, 244–250. Martini, R. (2001). The eVect of myelinating Schwann cells on axons. Muscle Nerve, 24, 456–466. Martini, R., Zielasek, J., Toyka, K. V., Giese, K. P., and Schachner, M. (1995). Protein zero (P0)-deWcient mice show myelin degeneration in peripheral nerves characteristic of inherited human neuropathies. Nat. Genet., 11, 281–286. Maurer, M., Kobsar, I., BerghoV, M., Schmid, C. D., Carenini, S., and Martini, R. (2002). Role of immune cells in animal models for inherited neuropathies: Facts and visions. J Anat, 200, 405–414. Maurer, M., Schmid, C. D., Bootz, F., Zielasek, J., Toyka, K. V., Oehen, S., and Martini, R. (2001). Bone marrow transfer from wild-type mice reverts the beneWcial eVect of genetically mediated immune deWciency in myelin mutants. Mol Cell Neurosci, 17, 1094–1101. Maycox, P. R., Ortuno, D., Burrola, P., Kuhn, R., Bieri, P. L., Arrezo, J. C., and Lemke, G. (1997). A transgenic mouse model for human hereditary neuropathy with liability to pressure palsies. Mol Cell Neurosci, 8, 405–416. Mersiyanova, I. V., Perepelov, A. V., Polyakov, A. V., Sitnikov, V. F., Dadali, E. L., Oparin, R. B., Petrin, A. N., and Evgrafov, O. V. (2000). A new variant of Charcot-Marie-Tooth disease type 2 is probably the result of a mutation in the neuroWlament-light gene. Am J Hum Genet, 67, 37–46. Moir, R. D., and Spann, T. P. (2001). The structure and function of nuclear lamins: Implications for disease. Cell Mol Life Sci, 58, 1748–1757. Mok, H., Shin, H., Kim, S., Lee, J. R., Yoon, J., and Kim, E. (2002). Association of the kinesin superfamily motor protein KIF1Balpha with postsynaptic density-95 (PSD-95), synapse-associated protein-97, and synaptic scaVolding molecule PSD-95/discs large/zona occludens-1 proteins. J Neurosci, 22, 5253–5258. Murphy, P., Topilko, P., Schneider-Maunoury, S., Seitanidou, T., Baron-Van Evercooren, A., and Charnay, P. (1996). The regulation of Krox-20 expression reveals important steps in the control of peripheral glial cell development. Development, 122, 2847–2857. Naef, R., Adlkofer, K., Lescher, B., and Suter, U. (1997). Aberrant protein traYcking in Trembler suggests a disease mechanism for hereditary human peripheral neuropathies. Mol Cell Neurosci, 9, 13–25. Nagarajan, R., Le, N., Mahoney, H., Araki, T., and Milbrandt, J. (2002). Deciphering peripheral nerve myelination by using Schwann cell expression proWling. Proc Natl Acad Sci U S A, 99, 8998–9003. Nagarajan, R., Svaren, J., Le, N., Araki, T., Watson, M., and Milbrandt, J. (2001). EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron, 30, 355–368.

CONCLUSIONS

Nelis, E., Haites, N., and Van Broeckhoven, C. (1999). Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum Mutat, 13, 11–28. Nelis, E., Van Broeckhoven, C., De Jonghe, P., Lofgren, A., Vandenberghe, A., Latour, P., Le Guern, E., Brice, A., Mostacciuolo, M. L., Schiavon, F., Palau, F., Bort, S., Upadhyaya, M., Rocchi, M., Archidiacono, N., Mandich, P., Bellone, E., Silander, K., Savontaus, M. L., Navon, R., Goldberg-Stern, H., Estivill, X., Volpini, V., Friedl, W., Gal, A., et al. (1996). Estimation of the mutation frequencies in Charcot-Marie-Tooth disease type 1 and hereditary neuropathy with liability to pressure palsies: A European collaborative study. Eur J Hum Genet, 4, 25–33. Nelles, E., Butzler, C., Jung, D., Temme, A., Gabriel, H. D., Dahl, U., Traub, O., Stumpel, F., Jungermann, K., Zielasek, J., Toyka, K. V., Dermietzel, R., and Willecke, K. (1996). Defective propagation of signals generated by sympathetic nerve stimulation in the liver of connexin32-deWcient mice. Proc Natl Acad Sci U S A, 93, 9565–9570. Neuberg, D. H., Sancho, S., and Suter, U. (1999). Altered molecular architecture of peripheral nerves in mice lacking the peripheral myelin protein 22 or connexin32. J Neurosci Res, 58, 612–623. Nicholson, S. M., Gomes, D., de Nechaud, B., and Bruzzone, R. (2001). Altered gene expression in Schwann cells of connexin32 knockout animals. J Neurosci Res, 66, 23–36. Niemann, S., Sereda, M. W., Suter, U., GriYths, I. R., and Nave, K. A. (2000). Uncoupling of myelin assembly and schwann cell diVerentiation by transgenic overexpression of peripheral myelin protein 22. J Neurosci, 20, 4120–4128. Norreel, J. C., Jamon, M., Riviere, G., Passage, E., Fontes, M., and Clarac, F. (2001). Behavioural proWling of a murine Charcot-Marie-Tooth disease type 1A model. Eur J Neurosci, 13, 1625–1634. Notterpek, L., Roux, K. J., Amici, S. A., Yazdanpour, A., Rahner, C., and Fletcher, B. S. (2001). Peripheral myelin protein 22 is a constituent of intercellular junctions in epithelia. Proc Natl Acad Sci U S A, 98, 14404–14409. Notterpek, L., Shooter, E. M., and Snipes, G. J. (1997). Upregulation of the endosomal-lysosomal pathway in the trembler-J neuropathy. J Neurosci, 17, 4190–4200. Paratore, P., Hagedorn, L., Floris, J., Hari, L., Kleber, M., Suter, U., and Sommer, L. (2002). Cell-intrinsic and cell-extrinsic cues regulating lineage decisions in multipotent neural crest-derived progenitor cells. Int J Dev Biol 46: 193–200 (2002), 46, 193–200. Pareyson, D., Menichella, D., Botti, S., Sghirlanzoni, A., Fallica, E., Mora, M., Ciano, C., Shy, M. E., and Taroni, F. (1999). Heterozygous null mutation in the P0 gene associated with mild Charcot-Marie-Tooth disease. Ann N Y Acad Sci, 883, 477–480. Parmantier, E., Braun, C., Thomas, J. L., Peyron, F., Martinez, S., and Zalc, B. (1997). PMP-22 expression in the central nervous system of the embryonic mouse deWnes potential transverse segments and longitudinal columns. J Comp Neurol, 378, 159–172. Parmantier, E., Lynn, B., Lawson, D., Turmaine, M., Namini, S. S., Chakrabarti, L., McMahon, A. P., Jessen, K. R., and Mirsky, R. (1999). Schwann cell-derived Desert hedgehog controls the development of peripheral nerve sheaths. Neuron, 23, 713–724. Paulson, H. L., Garbern, J. Y., Hoban, T. F., Krajewski, K. M., Lewis, R. A., Fischbeck, K. H., Grossman, R. I., Lenkinski, R., Kamholz, J. A., and Shy, M. E. (2002). Transient central nervous system white matter abnormality in X-linked Charcot-Marie-Tooth disease. Ann Neurol, 52, 429–434. Peirano, R. I., Goerich, D. E., Riethmacher, D., and Wegner, M. (2000). Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol Cell Biol, 20, 3198–3209. Perea, J., Robertson, A., Tolmachova, T., Muddle, J., King, R. H., Ponsford, S., Thomas, P. K., and Huxley, C. (2001). Induced myelination and demyelination in a conditional mouse model of Charcot-Marie-Tooth disease type 1A. Hum Mol Genet, 10, 1007–1018. Perez-Olle, R., Leung, C. L., and Liem, R. K. (2002). EVects of Charcot-Marie-Tooth-linked mutations of the neuroWlament light subunit on intermediate Wlament formation. J Cell Sci, 115, 4937–4946. Perkins, C. S., Aguayo, A. J., and Bray, G. M. (1981). Schwann cell multiplication in Trembler mice. Neuropathol Appl Neurobiol, 7, 115–126. Phillips, J. P., Warner, L. E., Lupski, J. R., and Garg, B. P. (1999). Congenital hypomyelinating neuropathy: Two patients with long-term follow-up. Pediatr. Neurol., 20, 226–232. Pingault, V., Bondurand, N., Kuhlbrodt, K., Goerich, D. E., Prehu, M. O., Puliti, A., Herbarth, B., HermansBorgmeyer, I., Legius, E., Matthijs, G., Amiel, J., Lyonnet, S., Ceccherini, I., Romeo, G., Smith, J. C., Read, A. P., Wegner, M., and Goossens, M. (1998). SOX10 mutations in patients with Waardenburg-Hirschsprung disease. Nat Genet, 18, 171–173. Pingault, V., Bondurand, N., Le Caignec, C., Tardieu, S., Lemort, N., Dubourg, O., Le Guern, E., Goossens, M., and BoespXug-Tanguy, O. (2001). The SOX10 transcription factor: Evaluation as a candidate gene for central and peripheral hereditary myelin disorders. J Neurol, 248, 496–499. Previtali, S. C., Quattrini, A., Fasolini, M., Panzeri, M. C., Villa, A., Filbin, M. T., Li, W., Chiu, S. Y., Messing, A., Wrabetz, L., and Feltri, M. L. (2000). Epitope-tagged P(0). glycoprotein causes charcot-marie-tooth-like neuropathy in transgenic mice. J. Cell Biol., 151, 1035–1046. Robertson, A., Perea, J., Tolmachova, T., Thomas, P. K., and Huxley, C. (2002a). EVects of mouse strain, position of integration and tetracycline analogue on the tetracycline conditional system in transgenic mice. Gene, 282, 65–74. Robertson, A. M., Huxley, C., King, R. H., and Thomas, P. K. (1999). Development of early postnatal peripheral nerve abnormalities in Trembler-J and PMP22 transgenic mice. J Anat, 195, 331–339.

1165

1166

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

Robertson, A. M., King, R. H., Muddle, J. R., and Thomas, P. K. (1997). Abnormal Schwann cell/axon interactions in the Trembler-J mouse. J Anat, 190, 423–432. Robertson, A. M., Perea, J., McGuigan, A., King, R. H., Muddle, J. R., Gabreels-Festen, A. A., Thomas, P. K., and Huxley, C. (2002b). Comparison of a new pmp22 transgenic mouse line with other mouse models and human patients with CMT1A. J Anat, 200, 377–390. Ryan, M. C., Shooter, E. M., and Notterpek, L. (2002). Aggresome formation in neuropathy models based on peripheral myelin protein 22 mutations. Neurobiol Dis, 10, 109–118. Sahenk, Z. (1999). Abnormal Schwann cell-axon interactions in CMT neuropathies. The eVects of mutant Schwann cells on the axonal cytoskeleton and regeneration-associated myelination. Ann N Y Acad Sci, 883, 415–426. Sahenk, Z., Chen, L., and Freimer, M. (1998). A novel PMP22 point mutation causing HNPP phenotype: Studies on nerve xenografts. Neurology, 51, 702–707. Sahenk, Z., Chen, L., and Mendell, J. R. (1999). EVects of PMP22 duplication and deletions on the axonal cytoskeleton. Ann Neurol, 45, 16–24. Sakai, Y., Nakabayashi, O., Kikuchi, T., and Wada, K. (1999). IdentiWcation of break points in mutated PMP22 gene in a new Trembler (Tr-Ncnp). mouse. Neuroscience, 88, 989–991. Salles, J., Sargueil, F., Knoll-Gellida, A., Witters, L. A., Shy, M., Jiang, H., Cassagne, C., and Garbay, B. (2002). Fatty acid synthase expression during peripheral nervous system myelination. Brain Res Mol Brain Res, 101, 52–58. Sancho, S., Magyar, J. P., Aguzzi, A., and Suter, U. (1999). Distal axonopathy in peripheral nerves of PMP22mutant mice. Brain, 122, 1563–1577. Sancho, S., Young, P., and Suter, U. (2001). Regulation of Schwann cell proliferation and apoptosis in PMP22deWcient mice and mouse models of Charcot-Marie-Tooth disease type 1A. Brain, 124, 2177–2187. Sargueil, F., Knoll, A., Salles, J., Garbay, B., and Cassagne, C. (1999). High metabolism and subsequent elongation of 3-hydroxyeicosanoyl-CoA in very-long-chain fatty acid deWcient PNS of Trembler mice. Neurosci Lett, 273, 29–32. Scherer, S. S., Bone, L. J., Deschenes, S. M., Abel, A., Balice-Gordon, R. J., and Fischbeck, K. H. (1999). The role of the gap junction protein connexin32 in the pathogenesis of X-linked Charcot-Marie-Tooth disease. Novartis Found Symp, 219, 175–185. Scherer, S. S., Deschenes, S. M., Xu, Y. T., Grinspan, J. B., Fischbeck, K. H., and Paul, D. L. (1995a). Connexin32 is a myelin-related protein in the PNS and CNS. J Neurosci, 15, 8281–8294. Scherer, S. S., and Fischbeck, K. H. (1999). Is CMTX an axonopathy? Neurology, 52, 432–433. Scherer, S. S., Xu, Y. T., Bannerman, P. G., Sherman, D. L., and Brophy, P. J. (1995b). Periaxin expression in myelinating Schwann cells: Modulation by axon-glial interactions and polarized localization during development. Development, 121, 4265–4273. Scherer, S. S., Xu, Y. T., Nelles, E., Fischbeck, K., Willecke, K., and Bone, L. J. (1998). Connexin32-null mice develop demyelinating peripheral neuropathy. Glia, 24, 8–20. Schmid, C. D., Stienekemeier, M., Oehen, S., Bootz, F., Zielasek, J., Gold, R., Toyka, K. V., Schachner, M., and Martini, R. (2000). Immune deWciency in mouse models for inherited peripheral neuropathies leads to improved myelin maintenance. J Neurosci, 20, 729–735. Schneider-Maunoury, S., Topilko, P., Seitandou, T., Levi, G., Cohen-Tannoudji, M., Pournin, S., Babinet, C., and Charnay, P. (1993). Disruption of Krox-20 results in alteration of rhombomeres 3 and 5 in the developing hindbrain. Cell, 75, 1199–1214. Schneider-Schaulies, J., Brunn, A. V., and Schachner, M. (1990). Recombinant peripheral myelin protein Po confers both adhesion and neurite outgrowth-promoting properties. J. Neuroscience Research, 27, 286–297. Senderek, J., Hermanns, B., Lehmann, U., Bergmann, C., Marx, G., Kabus, C., Timmerman, V., StoltenburgDidinger, G., and Schroder, J. M. (2000). Charcot-Marie-Tooth neuropathy type 2 and P0 point mutations: Two novel amino acid substitutions (Asp61Gly; Tyr119Cys)., and a possible ‘‘hotspot’’ on Thr124Met. Brain Pathol, 10, 235–248. Sereda, M., GriYths, I., Puhlhofer, A., Stewart, H., Rossner, M. J., Zimmerman, F., Magyar, J. P., Schneider, A., Hund, E., Meinck, H. M., Suter, U., and Nave, K. A. (1996). A transgenic rat model of Charcot-Marie-Tooth disease. Neuron, 16, 1049–1060. Sghirlanzoni, A., Pareyson, D., Balestrini, M. R., Bellone, E., Berta, E., Ciano, C., Mandich, P., and Marazzi, R. (1992). HMSN III phenotype due to homozygous expression of a dominant HMSN II gene. Neurology, 42, 2201–2204. Shapiro, L., Doyle, J. P., Hensley, P., Colman, D. R., and Hendrickson, W. A. (1996). Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron, 17, 435–449. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J. (2001). SpeciWc disruption of a schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron, 30, 677–687. Shy, M. E., Arroyo, E., Sladky, J., Menichella, D., Jiang, H., Xu, W., Kamholz, J., and Scherer, S. S. (1997). Heterozygous P0 knockout mice develop a peripheral neuropathy that resembles chronic inXammatory demyelinating polyneuropathy (CIDP). J Neuropathol Exp Neurol, 56, 811–821. Shy, M. E., Hobson, G., Jain, M., BoespXug-Tanguy, O., Garbern, J., Sperle, K., Li, W., Gow, A., Rodriguez, D., Bertini, E., Mancias, P., Krajewski, K., Lewis, R., and Kamholz, J. (2003). Schwann cell expression of PLP1 but not DM20 is necessary to prevent neuropathy. Ann Neurol, 53, 354–365. Southwood, C. M., Garbern, J., Jiang, W., and Gow, A. (2002). The unfolded protein response modulates disease severity in Pelizaeus-Merzbacher disease. Neuron, 36, 585–596.

CONCLUSIONS

Stolt, C. C., Rehberg, S., Ader, M., Lommes, P., Riethmacher, D., Schachner, M., Bartsch, U., and Wegner, M. (2002). Terminal diVerentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev, 16, 165–170. Suh, J. G., Ichihara, N., Saigoh, K., Nakabayashi, O., Yamanishi, T., Tanaka, K., Wada, K., and Kikuchi, T. (1997). An in-frame deletion in peripheral myelin protein-22 gene causes hypomyelination and cell death of the Schwann cells in the new Trembler mutant mice. Neuroscience, 79, 735–744. Sullivan, T., Escalante-Alcalde, D., Bhatt, H., Anver, M., Bhat, N., Nagashima, K., Stewart, C. L., and Burke, B. (1999). Loss of A-type lamin expression compromises nuclear envelope integrity leading to muscular dystrophy. J Cell Biol, 147, 913–920. Suter, U., Moskow, J. J., Welcher, A. A., Snipes, G. J., Kosaras, B., Sidman, R. L., Buchberg, A. M., and Shooter, E. M. (1992a). A leucine-to-proline mutation in the putative Wrst transmembrane domain of the 22kDa peripheral myelin protein in the trembler-J mouse. Proc Natl Acad Sci U S A, 89, 4382–4386. Suter, U., and Nave, K. A. (1999). Transgenic mouse models of CMT1A and HNPP. Ann N Y Acad Sci, 883, 247–253. Suter, U., Snipes, G. J., Schoener-Scott, R., Welcher, A. A., Pareek, S., Lupski, J. R., Murphy, R. A., Shooter, E. M., and Patel, P. I. (1994). Regulation of tissue-speciWc expression of alternative peripheral myelin protein-22 (PMP22). gene transcripts by two promoters. J Biol Chem, 269, 25795–25808. Suter, U., Welcher, A. A., Ozcelik, T., Snipes, G. J., Kosaras, B., Francke, U., Billings-Gagliardi, S., Sidman, R. L., and Shooter, E. M. (1992b). Trembler mouse carries a point mutation in a myelin gene. Nature, 356, 241–244. Suter, U., Welcher, A. A., and Snipes, G. J. (1993). Progress in the molecular understanding of hereditary peripheral neuropathies reveals new insights into the biology of the peripheral nervous system. Trends Neurosci, 16, 50–56. Svaren, J., Sevetson, B. R., Golda, T., Stanton, J. J., SwirnoV, A. H., and Milbrandt, J. (1998). Novel mutants of NAB corepressors enhance activation by Egr transactivators. Embo J, 17, 6010–6019. Swiatek, P. J., and Gridley, T. (1993). Perinatal lethality and defects in hindbrain development in mice homozygous for a targeted mutation of the zinc Wnger gene Krox20. Genes Dev, 7, 2071–2084. Taillebourg, E., Buart, S., and Charnay, P. (2002). Conditional, Xoxed allele of the Krox20 gene. Genesis, 32, 112–113. Takashima, H., Boerkoel, C. F., De Jonghe, P., Ceuterick, C., Martin, J. J., Voit, T., Schroder, J. M., Williams, A., Brophy, P. J., Timmerman, V., and Lupski, J. R. (2002). Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann Neurol, 51, 709–715. Taylor, V., Zgraggen, C., Naef, R., and Suter, U. (2000). Membrane topology of peripheral myelin protein 22. J Neurosci Res, 62, 15–27. Temme, A., Buchmann, A., Gabriel, H. D., Nelles, E., Schwarz, M., and Willecke, K. (1997). High incidence of spontaneous and chemically induced liver tumors in mice deWcient for connexin32. Curr Biol, 7, 713–716. Tobler, A. R., Liu, N., Mueller, L., and Shooter, E. M. (2002). DiVerential aggregation of the Trembler and Trembler J mutants of peripheral myelin protein 22. Proc Natl Acad Sci U S A, 99, 483–488. Tobler, A. R., Notterpek, L., Naef, R., Taylor, V., Suter, U., and Shooter, E. M. (1999). Transport of Trembler-J mutant peripheral myelin protein 22 is blocked in the intermediate compartment and aVects the transport of the wild-type protein by direct interaction. J Neurosci, 19, 2027–2036. Topilko, P., Schneider-Maunoury, S., Levi, G., Baron-Van Evercooren, A., ChennouW, A. B., Seitanidou, T., Babinet, C., and Charnay, P. (1994). Krox-20 controls myelination in the peripheral nervous system. Nature, 371, 796–799. Toyka, K. V., Zielasek, J., Ricker, K., Adlkofer, K., and Suter, U. (1997). Hereditary neuromyotonia: A mouse model associated with deWciency or increased gene dosage of the PMP22 gene. J Neurol Neurosurg Psychiatry, 63, 812–813. Umehara, F., Tate, G., Itoh, K., Yamaguchi, N., Douchi, T., Mitsuya, T., and Osame, M. (2000). A novel mutation of desert hedgehog in a patient with 46,XY partial gonadal dysgenesis accompanied by minifascicular neuropathy. Am J Hum Genet, 67, 1302–1305. Valentijn, L. J., Baas, F., Wolterman, R. A., Hoogendijk, J. E., van den Bosch, N. H., Zorn, I., Gabreels-Festen, A. W., de Visser, M., and Bolhuis, P. A. (1992). Identical point mutations of PMP-22 in Trembler-J mouse and Charcot-Marie-Tooth disease type 1A. Nat Genet, 2, 288–291. Valentijn, L. J., Ouvrier, R. A., van den Bosch, N. H., Bolhuis, P. A., Baas, F., and Nicholson, G. A. (1995). Dejerine-Sottas neuropathy is associated with a de novo PMP22 mutation. Hum Mutat, 5, 76–80. Vallat, J. M., Sindou, P., Garbay, B., Preux, P. M., Anani, T., Richard, L., and Diot, M. (1999). Expression of myelin proteins in the adult heterozygous Trembler mouse. Acta Neuropathol (Berl), 98, 281–287. Voiculescu, O., Charnay, P., and Schneider-Maunoury, S. (2000). Expression pattern of a Krox-20/Cre knock-in allele in the developing hindbrain, bones, and peripheral nervous system. Genesis, 26, 123–126. Wang, H., Allen, M. L., Grigg, J. J., Noebels, J. L., and Tempel, B. L. (1995). Hypomyelination alters Kþ channel expression in mouse mutants shiverer and Trembler. Neuron, 15, 1337–1347. Warner, L. E., Hilz, M. J., Appel, S. T., Killian, J. M., Kolodny, E. H., Karpati, G., Carpenter, S., Watters, G. V., Wheeler, C., Witt, D., Bodell, A., Nelis, E., Broeckhoven, C. V., and Lupski, J. R. (1996). Clinical phenotypes of diVerent MPZ (Po) mutations may include Charcot-Marie-Tooth type 1b, Dejerine-Sottas, and Congenital Hypomyelination. Neuron, 17, 451–460.

1167

1168

48. MODELS OF CHARCOT-MARIE-TOOTH DISEASE

Warner, L. E., Mancias, P., Butler, I. J., McDonald, C. M., Keppen, L., Koob, K. G., and Lupski, J. R. (1998). Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat Genet, 18, 382–384. Warner, L. E., Svaren, J., Milbrandt, J., and Lupski, J. R. (1999). Functional consequences of mutations in the early growth response 2 gene (EGR2). correlate with severity of human myelinopathies. Hum Mol Genet, 8, 1245–1251. Werner, H., Dimou, L., Klugmann, M., PfeiVer, S., and Nave, K. A. (2001). Multiple splice isoforms of proteolipid M6B in neurons and oligodendrocytes. Mol Cell Neurosci, 18, 593–605. Willecke, K., Temme, A., Teubner, B., and Ott, T. (1999). Characterization of targeted connexin32-deWcient mice: A model for the human Charcot-Marie-Tooth (X-type). inherited disease. Ann N Y Acad Sci, 883, 302–309. Wilson, H. L., Wilson, S. A., Surprenant, A., and North, R. A. (2002). Epithelial membrane proteins induce membrane blebbing and interact with the P2X7 receptor C-terminus. J Biol Chem, in press. Windebank, A. J. (1993). Inherited recurrent neuropathy. In ‘‘Peripheral Neuropathy’’ (P. J. Dyck, P. K. Thomas, J. W. GriYn, P. A. Low, and J. F. Poduslo, eds.), pp. 1094–1136. WB Saunders, Philadelphia. Wong, M. H., and Filbin, M. T. (1996). Dominant-negative eVect on adhesion by myelin Po protein truncated in its cytoplasmic domain. J. Cell Biol., 134, 1531–1541. Wrabetz, L., D’Antonio, M., Dati, G., Fratta, P., Previtali, S., Imperiale, D., Zielasek, J., Toyka, K., Messing, A., Feltri, M.-L., and Quattrini, A. (2002). Transgenic mice expressing the CMT1B mutant Mpz(delser63). develop demyelinating neuropathy. 2002 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience, (Online), Program No. 523.527. Wrabetz, L., and Feltri, M. L. (2001). Do Schwann cells stop, DR(o)P2, and roll? Neuron, 30, 642–644. Wrabetz, L., Feltri, M. L., Hanemann, C. O., and Mu¨ller, H. W. (2001). The molecular genetics of hereditary demyelinating neuropathies. In ‘‘Glial Cell Development: Basic Principles and Clinical Relevance’’ (K. R. Jessen, and W. R. Richardson, eds.), pp. 327–351. Oxford University Press, Oxford. Wrabetz, L., Feltri, M. L., Quattrini, A., Imperiale, D., Previtali, S., D’Antonio, M., Martini, R., Yin, X., Trapp, B. D., Zhou, L., Chiu, S. Y., and Messing, A. (2000). P(0) glycoprotein overexpression causes congenital hypomyelination of peripheral nerves. J Cell Biol, 148, 1021–1034. Yin, X., Crawford, T. O., GriYn, J. W., Tu, P., Lee, V. M., Li, C., Roder, J., and Trapp, B. D. (1998). Myelinassociated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J Neurosci, 18, 1953–1962. Yin, X., Kidd, G. J., Wrabetz, L., Feltri, M. L., Messing, A., and Trapp, B. D. (2000). Schwann cell myelination requires timely and precise targeting of P(0). protein. J Cell Biol, 148, 1009–1020. Yoshihara, T., Yamamoto, M., Hattori, N., Misu, K., Mori, K., Koike, H., and Sobue, G. (2002). IdentiWcation of novel sequence variants in the neuroWlament-light gene in a Japanese population: Analysis of CharcotMarie-Tooth disease patients and normal individuals. J Peripher Nerv Syst, 7, 221–224. Zhang, K., and Filbin, M. T. (1998). Myelin Po protein mutated at Cys21 has a dominant-negative eVect on adhesion of wild type Po. J. Neurosci Res., 53, 1–6. Zhao, C., Takita, J., Tanaka, Y., Setou, M., Nakagawa, T., Takeda, S., Yang, H. W., Terada, S., Nakata, T., Takei, Y., Saito, M., Tsuji, S., Hayashi, Y., and Hirokawa, N. (2001). Charcot-Marie-Tooth disease type 2A caused by mutation in a microtubule motor KIF1Bbeta. Cell, 105, 587–597. Zhu, Q., Couillard-Despres, S., and Julien, J. P. (1997). Delayed maturation of regenerating myelinated axons in mice lacking neuroWlaments. Exp Neurol, 148, 299–316. Zielasek, J., Martini, R., Suter, U., and Toyka, K. V. (2000). Neuromyotonia in mice with hereditary myelinopathies. Muscle Nerve, 23, 696–701. Zielasek, J., Martini, R., and Toyka, K. V. (1996). Functional Abnormalities in P-0-DeWcient Mice Resemble Human Hereditary Neuropathies Linked to P-0 Gene Mutations. Muscle & Nerve, 19, 946–952.

Index

Action potential conduction clinical status relationship with abnormalities, 124–125 continuous conduction in demyelinated axons, 125–128 demyelinated axon abnormalities, 121–124 impedance mismatching in demyelinated axons, 130–131 lability in multiple sclerosis, 124 myelinated axons, 121 pharmacological restoration in demyelinated axons, 131–133 recovery in demyelinated axons, 125 remyelination and restoration, 133–134 Acute disseminated encephalomyelitis (ADEM) acute hemorrhagic necrotizing leukoencephalopathy, 963 clinical features chickenpox, 961 Epstein–Barr virus, 962 human immunodeWciency virus, 961–962 measles, 958–960 mumps, 963 rubella, 961 smallpox, 960–961 deWnition, 955 epidemiology, 955 immunobiology, 792–793 pathogenesis, 957–958 pathology, 955–957 prevention, 963 treatment, 963 virus-induced demyelination mechanisms, 953–954 Acute inXammatory demyelinating polyneuropathy, see Guillain–Barre syndrome AD, see Alzheimer’s disease ADEM, see Acute disseminated encephalomyelitis Adherens junctions, paranodal regions, 99 Adrenoleukodystrophy (ALD) animal models of X-linked disease, 825 clinical features of X-linked disease Addison disease only phenotype, 813 adolescent cerebral disease, 811 adrenomyeneuropathy, 811 adrenomyeneuropathy-cerebral phenotype, 811 adult cerebral disease, 813 asymptomatic males, 813

Myelin Biology and Disorders, Volume 1

1169

childhood cerebral disease, 809, 811 epidemiology, 808–809 manifesting heterozygote females, 814 olivocerebellar atrophy, 813 comparison with other leukodystrophies, 675, 681–683, 687 demyelination versus dysmyelination, 667–668 diagnosis of X-linked disease mass neonatal screening, 828 prenatal diagnosis, 827–828 screening of family members, 827 symptomatic patients, 826–827 forms, 807 gene mutations in X-linked disease genotype-phenotype correlations, 819–820 locus, 809, 817 mutation analysis, 817, 819 protein function, 817, 822–823 genetic counseling, 832 history of study, 807–809 pathogenesis adrenomyeneuropathy, 823–824 inXammatory response, 824–825 pathology of X-linked disease adrenal cortex, 815 adrenomyeneuropathy, 814–815 cerebral forms, 814 peripheral nerve lesions, 815 testes changes, 815 treatment adrenal steroid replacement therapy, 828 bone marrow transplantation, 828–829 dietary therapy, 829–830 gene therapy, 832 immunosuppression therapy, 830–831 lovastatin, 831–832 4-phenylbutyrate, 830 very long chain fatty acid accumulation histopathology, 815–817 impaired peroxisomal b-oxidation, 820–821 inXuence on X-linked disease, 821–822 synthesis, 822 Adynoglia, characteristics, 224 ALD, see Adrenoleukodystrophy Alexander disease animal models

Copyright 2004, Elsevier Science (USA). All rights reserved.

1170

INDEX

Alexander disease (Continued ) glial Wbrillary acidic protein knockout mouse, 1117–1118 glial Wbrillary acidic protein transgenic mice, 1118–1119 mouse model prospects, 1121 spontaneous models, 1116 clinical presentation, 851–852 comparison with other leukodystrophies, 675, 681 diagnosis, 854–855 familial cases, 858–859 glial Wbrillary acidic protein astrocyte expression, 1117 interacting protein mutations, 859 mutations in disease, 855–858, 1116 pathogenetic mechanisms, 860–862, 1116–1117 history of study, 851, 1115 pathology, 852–854, 1115 Alzheimer’s disease (AD), unfolded protein response, 1028 4-Aminopyridine, restoration of impulse conduction in demyelinated axons, 132 AnkyrinG sodium channel interactions, 91–93 spectrin interactions, 93–94 Astrocyte brain volume, 223 cell-cell interactions, 311 development adult central nervous system, 319–320 cerebellum, 318–319 lineage relationships with other central nervous system cells, 320–321 migration of progenitors, 316–317 molecular regulation, 321–323 pathways and astrocyte types, 317–318 precursors, overview, 243 prospects for study, 323–324 radial glia, 312–314 spinal cord, 319 supraventricular zone cells, 314–316 zones, 311–312 forms in central nervous system, 311 glial Wbrillary acidic protein expression, 1117 olfactory ensheathing cell interactions, 380–381 types, 224, 317–318 Axon action potential conduction, see Action potential conduction glia guidance in invertebrates, 211–213 mouse hepatitis virus damage mechanisms, 1093–1094 multiple sclerosis degeneration and neuroprotection, 134, 749–750 myelin-associated glycoprotein axonal pathology, 21, 438–439 growth and regeneration regulation inhibition in adults, 444–446 promotion in development, 445 myelination in maturation and survival, 19–22 Pelizaeus–Merzbacher disease damage, 879–880 peripheral myelin protein-22, axonal pathology in loss of function, 21 proteolipid protein , axonal pathology in loss of function, 21–22 safety factors demyelinated axons, 121, 124 myelinated axons, 121

Theiler’s murine encephalomyelitis virus damage mechanisms, 1087–1088 Azathioprine, secondary progressive multiple sclerosis trials, 800

BBB, see Blood–brain barrier B-cell multiple sclerosis inXammatory response, 738 myelin oligodendrocyte glycoprotein autoantibody response, 480–482 BDNF, see Brain-derived neurotrophic factor B-FABP, see Brain-speciWc fatty acid-binding protein Blood–brain barrier (BBB) Drosophila glia, 214–215 magnetic resonance imaging of breakdown, 771–773 BMPs, see Bone morphogetic proteins Bone morphogetic proteins (BMPs), astrocyte development role, 322 Brain-derived neurotrophic factor (BDNF) multiple sclerosis inXammatory response, 741–742 Schwann cell myelination regulation, 353 Brain-speciWc fatty acid-binding protein (B-FABP), Schwann cell diVerentiation marker, 334, 337

C1q, myelin oligodendrocyte glycoprotein binding, 490–491 CADASIL, see Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukencephalopathy Campath, relapsing-remitting multiple sclerosis trials, 799 Caspr, see Neurexin IV/Caspr/Paranodin family CD-9, Schwann cell diVerentiation marker, 338–339 Central nervous system barbotage, remyelination studies, 178 Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukencephalopathy (CADASIL), misclassiWcation as leukodystrophy, 665–666 Cerebrospinal Xuid (CSF), multiple sclerosis Wndings, 693 CF, see Cystic Wbrosis Charcot–Marie–Tooth disease (CMT) animal models generalizations from studies, 1159–1160 pathogenesis studies, 1159 autosomal recessive CMT2, 926, 931–932, 1157 candidate gene discovery, 934 classiWcation of types cellular basis of classiWcation, 909, 911 clinical classiWcation, 905–909 CMT1A features, 913 CMT1X features, 922 CMT2 features, 929–930 connexin 32 mutations in CMT1X animal models, 1153–1154 genotype–phenotype correlations, 922–923 loss-of-function mutations, 923–924 overview, 600, 602–604, 921, 1153–1154 GDAP1 mutations in CMT4A, 926 gene mutation frequency distribution, 918 genetic testing, 933–934 KIF1B mutations in CMT2A knockout mouse model, 1156 overview, 930

INDEX

Krox20 mutations animal models, 1154–1155 overview, 924–925 LITAF mutations in CMT1C, 925 LMNA mutations CMT2B1, 931–932 knockout mouse, 1157 MTMR13 mutations in CMT4B2, 927–928 MTMR2 mutations in CMT4B1, 926–927 NDRG1 mutations in CMT4D, 928 NEFL mutations in CMT2E knockout mouse model, 1156–1157 overview, 930–931 P0 mutations in CMT1B animal models gain-of-function models, 1151–1152 loss-of function models, 1150–1151 genotype–phenotype correlations, 919–920 overview, 917, 919, 1149–1150 pathogenesis of inherited axonal neuropathies, 932–933 periaxin mutations in CMT4F knockout mouse model, 1155–1156 overview, 640–641, 928 peripheral myelin protein-22 mutations in CMT1A animal models conditional overexpression in mice, 1148–1149 knockout mouse, 1149 mutagen induced mutants, 1147 overexpression in transgenic mice, 1147–1148 overexpression in transgenic rat, 1148 Trembler mouse, 1144–1147 Trembler-J mouse, 1147 Trembler-Ncnp mouse, 1147 overview, 912–916, 1144 RAB7 mutations in CMT2B, 930 syndromic neuropathy animal models Desert Hedgehog knockout mouse, 1157–1158 proteolipid protein knockout mouse, 1158 Sox10 knockout mouse, 1158–1159 unfolded protein response, 1026 Chemokines, multiple sclerosis inXammatory response, 740–741 Chickenpox, acute disseminated encephalomyelitis, 961 Chlamydia pneumoniae, multiple scleroisis studies, 977–978 Cholesterol biosynthesis and transport, 65 rafts, see Lipid rafts Cidofovir, progressive multifocal leukencephalopathy management, 969 Ciliary neurotrophic factor (CNTF), astrocyte development role, 321, 323 Claudin 11 tight junctions blood–testis barrier, 571 Claudin family overview, 566 endocochlear potential generation role, 569–571 knockout mouse phenotype, 566, 572–573 myelin tight junctions claudin 11 distribution, 571–572 components, 571 functions, 572–576 tissue distribution, 567–568 CMT, see Charcot–Marie–Tooth disease CNP, see Cyclic nucleotide 30 -phosphodiesterase CNTF, see Ciliary neurotrophic factor Compact myelin

proteins, see speciWc proteins structure and function, 6 Complicated spastic paraplegia (SPG2) clinical presentation and course, 873–876 diagnosis, 873–875 molecular pathogenesis, 877–880 proteolipid protein mutations, 869–870 Conduction block demyelinated Wbers, 122 sodium channel inhibition, 123–124 Conduction velocity, myelination eVects, 121–122 Conformational diseases, see also Pelizaeus–Merzbacher disease; Unfolded protein response amyloidosis, 1009 expanded trinucleotide repeat diseases, 1013–1014 neurodegenerative cytoplasmic inclusion diseases, 1010, 1012–1013 prion diseases, 1011 serinopathies, 1011–1012 Connexin 29 gene regulation of expression, 606 structure, 605–606 sequence homology with connexin 32, 606 tissue distribution, 606 Connexin 32 gene regulation of expression, 601–602, 604–605 structure, 600 mutations in Charcot–Marie–Tooth disease-1X animal models, 1153–1154 genotype–phenotype correlations, 922–923 loss-of-function mutations, 923–924 overview, 600, 602–604, 921, 1153–1154 sequence homology with connexin 29, 606 structure, 600–601 tissue distribution, 604 CSF, see Cerebrospinal Xuid Cyclic nucleotide 3’-phosphodiesterase (CNP) assays, 499–500 compact myelin function, 10 domains, 503–504 Wsh homolog, 500 function studies, 513–516 gene comparative structure between species, 506–507 developmental expression patterns, 509–510 mouse loci, 507 promoter, 510–513 regulation of expression, 510–513 transcription splice variants, 508–509 history of study, 499 isoforms, 502, 507–509 kinetic parameters, 504 lipid raft association, 76 membrane-binding properties, 505–506 myelin assembly role, 47 non-compact myelin composition and function, 11 pathology of demyelinating disorders, 516 post-translational modiWcation, 503 puriWcation, 500–502 reaction mechanism, 504 sequence homology between species, 501–504, 506 three-dimensional structure, 504–505 tissue distribution, 500, 508–509

1171

1172

INDEX

Cyclophosphamide, secondary progressive multiple sclerosis trials, 800–801 Cystic Wbrosis (CF), unfolded protein response, 1026–1027 Cytarabine, progressive multifocal leukencephalopathy management, 969 Cytokines, multiple sclerosis inXammatory response, 741

Dejerine–Sottas syndrome (DSS) genetic testing, 933–934 Krox20 mutations, 924–925 P0 mutations, 917 peripheral mylein protein-22 mutations, 913–914 Desert Hedgehog (DHH) knockout mouse, 1157 Schwann cell diVerentiation marker, 338 DHH, see Desert Hedgehog DM20, see Proteolipid protein Drosophila glia adult glia central nervous system, 201, 203 origins, 206–207 peripheral nervous system, 203–204 axonal conductance, 217 axonal insulation, 215 blood–brain barrier, 214–215 developmental functions axon guidance, 211–213 nervous system morphogenesis, 211 neuronal interaction and glial survival, 214 neuronal proliferation, 214 neuronal survival, 213 embryonic and Wrst instar glia central nervous system, 200 origins, 204–206 peripheral nervous system, 200–201 invertebrate glia comparisons, 199–200 molecular mechanisms in determination midline glia, 207 Notch signaling, 208–209 precursor cells, 209–211 neuron feeding, 216–217 neurotransmitter uptake, 215–216 vertebrate glia comparisons, 217–218 DSS, see Dejerine–Sottas syndrome Dystroglycan L-periaxin complex axon–glia interaction role, 637–638 composition, 636–637 laminin binding, 357 neurexin binding, 584–585

EAE, see Experimental autoimmune encephalitis EBV, see Epstein–Barr virus Epitope spreading autoimmunity propagation, 1053–1054 mouse hepatitis virus, 1094 Theiler’s murine encephalomyelitis virus, 1086–1087 Epstein–Barr virus (EBV) acute disseminated encephalomyelitis, 962 multiple scleroisis studies, 976 Ethidium bromide demyelination in animal models, 149 remyelination studies, 177–178

Excitotoxicity experimental autoimmune encephalitis, 1059 ischemic white matter damage, 995–997 T-cell mediated tissue damage, 753 Experimental autoimmune encephalitis (EAE) antigen-spreading in autoimmunity propagation, 1053–1054 clearance of brain inXammation, 1052–1053 glial cell transplantation model, 152 historical perspective, 1039–1040 induction and pathogenesis acute sensitization, 1042–1043 co-transfer experimental autoimmune encephalitis, 1041–1042 passive transfer of autoreactive T-lymphocytes, 1040–1041 leukocyte migration, 1050–1051 major histocompatibility complex molecule expression and T-cell activation, 1051–1052 myelin oligodendrocyte glycoprotein genetic susceptibility and myelin oligodendrocyte glycoprotein polymorphisms, 486–487 multiple sclerosis signiWcance, 487–488 pathophysiology, 469, 480, 1063 protective autoimmunity, 1061–1062 remyelination studies, 177, 1060–1061 target antigens antibody-mediated demyelination, 1045–1046 T-cell-mediated autoimmunity, 1043–1045 T-cell populations in inXammation class I major histocompatibility complexrestricted cytotoxic T-cells, 1047–1049 Th-1 cells, 1046 Th-2 cells, 1046–1047 tissue damage mechanisms antibody-mediated tissue injury, 1056, 1063 ischemia, 1059–1060 macrophage-mediated injury excitotoxins, 1059 free radicals, 1058–1059 proteases, 1056–1058 tumor necrosis factor-a, 1058 overview, 1054–1055 T-cell-mediated cytotoxicity, 1055–1056

FGF-2, see Fibroblast growth factor-2 Fibroblast growth factor-2 (FGF-2) integrin interactions in myelinating glia, 619 oligodendrocyte progenitor cell recruitment, 180–181, 261–262 Fly glia, see Drosophila glia fMRI, see Functional magnetic resonance imaging Functional magnetic resonance imaging (fMRI) brain adaptive reorganization in multiple sclerosis, 780–782 principles, 768

Gangliosides biosynthesis and transport, 63–64 myelin-associated glycoprotein receptors, 448–450 oligodendrocyte physiology regulation axo-glial communication and nerve regeneration, 73–74 diseases, 74

INDEX

ganglioside expression as a function of cell development, 72–73 GD3 and cell migration, 73 GM3 and cell diVerentiation, 73 overview, 71–72 survival inXuences, 73 Gap junctions, see also Connexin 29; Connexin 32 connexins, 599–600 paranodal regions, 99 GAP-43, see Neuromodulin GBS, see Guillain–Barre syndrome GDAP1, mutations in Charcot–Marie–Tooth disease-4A, 926 GFAP, see Glial Wbrillary acidic protein Glatiramer acetate, relapsing-remitting multiple sclerosis trials, 798–799 Glial cell transplantation action potential conduction restoration, 133–134 animal models jimpy mouse, 147–148 lesion models antibody-mediated demyelination, 149–150 diVuse lesions, 151–152 ethidium bromide demyelination, 149 experimental autoimmune encephalitis, 152 focal lesions, 148–151 lysolecithin demyelination, 149 optic nerve, 152 X-irradiated lesions, 150–151 Long Evans shaker rat, 147 MAG-fyn knockout mouse, 148 myelin-deWcient rat, 147 neonatal mice and rats, 146–147 proteolipid protein transgenic mouse, 148 shaking pup, 148 shiverer mouse, 147 clinical considerations acquired demyelination disorders, 280–281 cell availability, 162 colonization, 162 congenital demyelination disorders, 279–280 indications, 161–162 prospects, 163–164 rejection, 162 remyelination potential, 163 safety, 162 donor sources adult brain sources, 271 fetal brain sources, 270–271 neural precursor cells adult, 156 fetal, 153–156 olfactory glia, 156–157, 278–279, 376–381 oligodendrocytes, 156 pluripotent stem cells, 158–159, 279 propagated oligospheres, 271 Schwann cells, 157–158, 278 eYcacy assessment, 161 fate of transplanted cells, 144–146 historical perspective, 143 human studies neural and oligodendrocyte precursors, 159–160, 269–270 Schwann cells, 160 oligodendrocyte progenitor cell transplantation studies in humans adult brain sources, 271

congenitally-dysmyelinated brain studies, 274 diVerentiation potential after implantation, 277–278 fetal brain sources, 270–271 fetal versus adult cell behavior, 274–277 integration of grafts, 271–272 migratory characteristics, 272 myelin synthesis, 272, 274 overview, 159–160, 269–270 propagated oligospheres, 271 principles, 144 Glial Wbrillary acidic protein (GFAP) Alexander disease interacting protein mutations, 859 mutations in disease, 855–858, 1116 pathogenetic mechanisms, 860–862, 1116–1117 animal models knockout mouse, 1117–1118 mouse model prospects, 1121 transgenic mice, 1118–1119 astrocyte expression, 1117 Glial-restricted precursor cell (GRP) diVerentiation, 243–244 lineage relationships between glial progenitors, 244–246 Globoid cell leukodystrophy, see Krabbe’s disease Glucocorticoids, relapsing-remitting multiple sclerosis trials, 799 Golli proteins functions oligodendrocyte physiology, 395 signaling, 394 transcriptional regulation in nucleus, 394 gene, see Myelin basic protein nervous system distribution, 393 subcellular localization, 393–394 GRP, see Glial-restricted precursor cell Guillain–Barre syndrome (GBS) antecedant infections, 892–893 clinical manifestations, 889–890 history of study, 887–888 molecular mimicry and acute motor axon neuropathy syndrome, 896–897 multiple sclerosis relevance, 897–898 pathogenesis of acute inXammatory demyelinating polyneuropathy, 893–895 pathology of acute inXammatory demyelinating polyneuropathy, 890–892

Hereditary motor neuropathy (HMN), types, 908, 932 Hereditary neuropathy with liability to pressure palsies (HNPP) clinical features, 912–913 genetic testing, 933–934 peripheral myelin protein-22 mutations, 912–915 Hereditary sensory neuropathy (HSN), types and gene mutations, 908, 932 HERVs, see Human endogenous retroviruses HHV6, see Human herpesvirus 6 HIV, see Human immunodeWciency virus HMN, see Hereditary motor neuropathy HNPP, see Hereditary neuropathy with liability to pressure palsies HSN, see Hereditary sensory neuropathy Human endogenous retroviruses (HERVs), multiple scleroisis studies, 977

1173

1174

INDEX

Human herpesvirus 6 (HHV6), multiple scleroisis studies, 976–977 Human immunodeWciency virus (HIV) acute disseminated encephalomyelitis, 961–962 progressive multifocal leukencephalopathy management, 969–970

IFN-b, see Interferon-b IFN-g, see Interferon-g IGF-1, see Insulin-like growth factor-1 Impedance mismatch, demyelinated axons, 130–131 Insulin-like growth factor-1 (IGF-1) oligoneogenesis role, 238–239, 262–263, 294 Schwann cell myelination regulation, 352 Integrins growth factor interactions in myelinating glia, 618–622 nervous system development and repair role, 611–612 oligodendrocytes expression of integrins, 613–614 functions of integrins diVerentiation, 617 migration, 616 proliferation, 616 survival, 617–618 receptor structure and signaling, 610–611 Schwann cells expression of integrins, 613 functions of integrins diVerentiation, 615–616 migration, 614–615 proliferation, 615 survival, 616 laminin binding, 355–356 signaling impairment in myelin diseases, 622–624 Interferon-b (IFN-b) relapsing-remitting multiple sclerosis trials interferon-b-1a, 797–798 interferon-b-1b, 796–797 secondary progressive multiple sclerosis trials, 800 Interferon-g (IFN-g), relapsing-remitting multiple sclerosis trials, 799 Internode, myelin function, 5–6 organization, 3–5 Intravenous immunoglobulin (IVIg), relapsingremitting multiple sclerosis trials, 799 Invertebrate glia, see Drosophila glia Ischemic white matter damage energy deprivation early consequences, 987–989 injury mechanisms, 989, 991–995 energy metabolism in central nervous system, 986–987 epidemiology, 985–986 excitotoxicity, 995–997 experimental autoimmune encephalitis, 1059–1060 leukodystrophies, 663, 665 neuroprotective strategies, 998–1002 reperfusion injury, 994–995 IVIg, see Intravenous immunoglobulin

JC virus, see Progressive multifocal leukencephalopathy

Jimpy mouse glial cell transplantation model, 147–148 Pelizaeus–Merzbacher disease models jimpy mouse, 1130–1131 jimpy-4j mouse, 1131 jimpy-msd mouse, 1131 Juxtaparanodes, see Nodes of Ranvier

KIF1B, mutations in Charcot–Marie–Tooth disease2A knockout mouse model, 1156 overview, 930 Krabbe’s disease animal models, 848 clinical manifestations infantile disease, 842 late-onset disease, 842–843 comparison with other leukodystrophies, 675, 681–682, 684, 686 galactoceramidase gene mutations, 847–848 polymorphisms, 848 structure, 847 galactoceramide analysis, 843–844 history of study, 841 incidence, 841–842 mouse models applications bone marrow transplantation, 1107–1108 gene therapy, 1108–1109 glial cell transplantation, 1108 nerve grafting, 1107 pathogenesis studies, 1106–1107 pregnancy studies, 1109 substrate reduction, 1109–1110 galactoceramidase mutant transgenic mice, 1104–1105 overview, 1101 saposin A knockout mouse, 1105–1106, 1109 twitcher mouse biochemistry, 1103–1104 pathology, 1102–1103 phenotype, 1102 myelin galactoceramide metabolism, 844–845 pathogenesis globoid cell reaction, 846 psychosine hypothesis, 846 uniWed theory, 845–847 pathology, 843 treatment, 848 Krox20 Charcot–Marie–Tooth disease mutations animal models, 1154–1155 overview, 924–925 Schwann cell myelination regulation, 349–351 Krox24, Schwann cell myelination regulation, 352

Laminin forms, 354 Schwann cell receptors dystroglycan, 357 integrins, 355–356 leprosy implications, 357 Leukencephalopathy with vanishing white matter (VWM), unfolded protein response, 1027

INDEX

Leukocyte inhibitory factor (LIF), astrocyte development role, 321–323 Leukodystrophies, see also speciWc diseases clinical similarities and diVerences, 675 demyelination versus dysmyelination, 666–668, 670 ischemic diseases, 663, 665 myelin degradation, 670–671 myelin lesion types, 663–664 neuropathologic similarities and diVerences, 675–678, 680–681 pathogenetic criteria, 665–666, 668 pathogenetic similarities and diVerences, 681–684, 686–687 primary diseases of myelin, 668–675 sudanophilic leukodystrophies, 669, 687 LIF, see Leukocyte inhibitory factor Lipid rafts deWnition, 76 imaging, 77 myelin protein association, 76–77 prospects for study, 80 protein partitioning, 74–76 protein traYcking role, 77–79 signal transduction role, 79–80 Lipids, myelin biosynthesis and transport cholesterol, 65 phospholipids, 65–66 sphingolipids, 59–65 central nervous system myelin composition, 57–58 oligodendrocyte physiology regulation by gangliosides axo-glial communication and nerve regeneration, 73–74 diseases, 74 ganglioside expression as a function of cell development, 72–73 GD3 and cell migration, 73 GM3 and cell diVerentiation, 73 overview, 71–72 survival inXuences, 73 oligodendrocyte proliferation and terminal diVerentiation regulation by glycosphingolipids, 68–71 peripheral versus central nervous system myelin lipids, 58–59 rafts, see Lipid rafts turnover, 67–68 LITAF, mutations in Charcot–Marie–Tooth disease1C, 925 LMNA knockout mouse, 1157 mutations in Charcot–Marie–Tooth disease-2B1, 931–932 Long Evans shaker rat, glial cell transplantation model, 147 Lovastatin, adrenoleukodystrophy management, 831–832 Lysolecithin, demyelination in animal models, 149

Macrophage experimental autoimmune encephalitis injury excitotoxins, 1059 free radicals, 1058–1059 proteases, 1056–1058 tumor necrosis factor-a, 1058

multiple sclerosis inXammatory response, 738–739 injury mediation, 753–754 MAG, see Myelin-associated glycoprotein Magnetic resonance imaging (MRI), multiple sclerosis atrophy of brain and spinal cord, 777 diagnostics, 692–694, 763, 769–770 diVuse white matter abnormalities, 776–778 diVusion-weighted imaging, 767 disease progression imaging blood–brain barrier breakdown, 771–773 demyelination, 773–776 gliosis, 773 inXammation, 773 functional magnetic resonance imaging adaptive reorganization in brain, 780–782 principles, 768 gadolinium-enhanced imaging, 765 gray matter pathology, 778–779 magnetization transfer imaging, 766 matrix destruction in lesions, 773–774 prelesional change visualization, 779–780 proton density-weighted imaging, 765, 773 T1-weighted imaging, 767, 773–774 T2-weighted imaging, 676, 764–765, 771–773 Magnetic resonance spectroscopy (MRS) diVuse axonal injury and loss, 776 myelin lesions, 676–677 principles, 767–768 MBP, see Myelin basic protein Measles, acute disseminated encephalomyelitis, 958–960 Metachromatic leukodystrophy (MLD) comparison with other leukodystrophies, 675, 681–682, 684, 686 demyelination versus dysmyelination, 667–668 MHV, see Mouse hepatitis virus Microtubule cytoskeleton role in diVerentiation and myelin biogenesis oligodendrocytes, 41–42 Schwann cells, 37–40 structure, 37 Mitoxantrone, secondary progressive multiple sclerosis trials, 802 MLD, see Metachromatic leukodystrophy MOG, see Myelin oligodendrocyte glycoprotein Mouse hepatitis virus (MHV) axonal damage mechanisms, 1093–1094 demyelination mechanisms, 1093–1094 disease pathology, 1090–1091 genome, 1089 immune response, 1091–1093 life cycle, 1089–1090 MRI, see Magnetic resonance imaging, multiple sclerosis MRS, see Magnetic resonance spectroscopy MS, see Multiple sclerosis MTMR2, mutations in Charcot–Marie–Tooth disease-4B1, 926–927 MTMR13, mutations in Charcot–Marie–Tooth disease-4B2, 927–928 Multiple sclerosis (MS) action potential conduction clinical status relationship, 124–125 lability, 124 recovery, 125

1175

1176

INDEX

Multiple sclerosis (MS) (Continued ) axonal degeneration and neuroprotection, 134 classiWcation benign, 697 malignant, 698 overview, 792 primary progressive, 695, 697 progressive-relapsing, 697 relapse-remitting, 694 secondary progressive, 697 deWnition, 970 demyelinated plaques axonal injury, 749–750 features, 733–735 gray matter lesions and cortical plaques, 750–751 diagnosis, 692–694 economic impact, 691 epidemiology environmental triggers, 707–708 epidemics, 706–707, 710, 972 geographical distribution, 702, 704, 971 incidence, 691 infection studies, 708–709, 972–973 migrant studies, 704–706, 971–972 pattern analysis, 709–710 pregnancy studies, 709 prevalence, 691, 702, 704 sex diVerences, 707, 971 study methodology, 701–702 genetic heterogeneity, 723–726 genetic inXuences on clinical course, 721–723 genetic susceptibility candidate genes, 717–718, 725 class II major histocompatibility complex alleles, 717, 724–725 complex trait analysis, 712–717 familial disease, 710–712, 971 linkage genome screens, 718–719 prospects for study, 725–726 whole genome association screening, 719–721 Guillain–Barre syndrome relevance, 897–898 history of study, 691, 791 hypoxia-like tissue damage, 754–756 imaging, see Magnetic resonance imaging, multiple sclerosis infection role demyelination mechanisms of viruses, 953–954 epidemiology studies, 708–709, 972–973 exacerbation, 973 isolation reports, 975–976 pathogens Chlamydia pneumoniae, 977–978 Epstein–Barr virus, 976 human endogenous retroviruses, 977 human herpesvirus 6, 976–977 pathology studies, 973 serological studies, 974 inXammatory response adhesion molecules, 740 B-cells, 738 brain-derived neurotrophic factor, 741–742 brain inXammation, 742–743 cerebral vessel eVects, 742 chemokines and receptors, 740–741 costimulatory molecules, 741 cytokines, 741 diVerential susceptibility of target tissue, 756

macrophages, 738–739 major histocompatibility complex antigen expression, 739–740 overview, 735–736 plasma cells, 738 prospects for study, 756–757 T-cells, 736–737, 793 oligodendrocyte injury within demyelinating lesions, 743–745 progression, 691–692 recurrence immunobiology, 794–795 remission immunobiology, 794 remyelination, 176, 745–747 secondary progression immunobiology, 795 sodium channelopathy, 129 T-cell mediated tissue damage antibody modulation, 754 mechanisms activated macrophage mediated injury, 753–754 complement activation, 753 excitotoxins, 753 free radicals, 753 proteases, 752 tumor necrosis factor, 752 overview, 751–752, 793–794 treatment trials relapsing-remitting disease Campath, 799 glatiramer acetate, 798–799 glucocorticoids, 799 interferon-b-1a, 797–798 interferon-b-1b, 796–797 interferon-g, 799 intravenous immunoglobulin, 799 prospects, 799 secondary progressive disease azathioprine, 800 cyclophosphamide, 800–801 immunoablation followed by autologous stem cell rescue, 801 interferon-b, 800 mitoxantrone, 802 neural directed therapy, 801 symptomatic treatment ataxia, 803 fatigue, 802 focal weakness, 802 pain, 802 paroxysmal symptoms, 802 spasticity, 802 Mumps, acute disseminated encephalomyelitis, 963 Myelin assembly cytoskeleton role Schwann cells, 37–40 oligodendrocytes, 41–42 layers, 117 lipid transport and assembly, see Lipids, myelin oligodendrocyte diVerentiation and initiation of myelination, 42–45 oligodendrocyte relationship with myelin sheath, 118 overview of steps, 32–33, 48–49, 609 protein functions cyclic nucleotide 3’-phosphodiesterase, 47 myelin basic protein, 47–48 myelin-associated glycoprotein, 45–47 proteolipid protein, 45

INDEX

Schwann cell diVerentiation, 33–34 Schwann cell myelination, 34–37 vesicular transport and membrane biogenesis in polarized cells, 29–32 Myelin-associated glycoprotein (MAG) adhesive functions, 433–435 autoimmune neuropathy, 454–456 axon growth and regeneration regulation inhibition in adults, 444–446 promotion in development, 445 axonal pathology in loss of function, 21 cytoskeletal interactions, 40 discovery, 421 distribution in nervous system, 422, 424 gene exon structure, 426 mutations in humans, 454 promoter, 425–426 transcription splice variants and isoforms, 46, 424, 426–427 glial signaling, 453–454 glycosylation, 428–429 knockout mouse phenotype axon morphology, 438–439 delayed myelin formation, 436 glial cell abnormalities, 437–438 L-MAG mutants, 439, 332 multiple knockout studies MAG/fyn knockout mouse, 148, 443–444 MAG/N-CAM knockout mouse, 443 MAG/P0, 443 MAG/UDP-galactose:ceramide galactosyltransferase knockout mouse, 443 proteolipid protein and/or myelin basic protein knockouts, 442 multiple myelin wrappings, 437 neurological and behavioral defects, 439 overview, 435–436, 440–441 periaxonal cytoplasmic collar collapse, 437 signal transduction, 450–453 myelin assembly role, 45–47 nomenclature, 421–422 non-compact myelin composition and function, 11–12, 34 phosphorylation, 427–428 prospects for study, 456–457 quaking mouse, 435 receptors gangliosides, 448–449 NgR, 446–448 Schwann cell myelination events, 34–37, 422 siglec activity, 431–433 structure, 45–46, 429–433 targeting, 32, 46 Myelin basic protein (MBP) compact myelin composition and function, 6–9 functions, 391–393 gene golli protein encoding, 388 promoters, 390 structure, 388–389 transcriptional regulatory elements tss1, 391 tss2, 391 tss3, 390–391 transcriptional splice variants, 387–290

history of study, 387–388 isoforms, 391 messenger RNA translocation, 47–48 mutant animal models for glial cell transplantation Long Evans shaker rat, 147 shiverer mouse, 147 myelin assembly role, 47–48 phosphorylation, 392 prospects for study, 395–396 Myelin oligodendrocyte glycoprotein (MOG) adhesion molecule ativity, 488–489 autoantibody response, 480–482 C1q binding, 490–491 encephalitogenic T-cell responses, 482–486 experimental autoimmune encephalitis genetic susceptibility and myelin oligodendrocyte glycoprotein polymorphisms, 486–487 multiple sclerosis signiWcance, 487–488 pathophysiology, 469, 480 extracellular domain, 480 gene butyrophilin-like gene family, 477 cloning, 470 locus in major histocompatibility complex, 475, 477 promoter, 475 regulation of expression, 474–475 structure in human versus mouse, 470–473 transcription splice variants, 473–474 history of study, 469–470 lipid raft association, 76, 78, 489 receptor activity and signaling, 489 structure, 477, 479–480, 489–490 targeting, 32 Myelin-deWcient rat glial cell transplantation model, 147 Pelizaeus–Merzbacher disease model, 1132

NCP family, see Neurexin IV/Caspr/Paranodin family NDRG1, mutations in Charcot–Marie–Tooth disease4D, 928 NEFL, mutations in CMT2E knockout mouse model, 1156–1157 overview, 930–931 Nervous system development glia, see speciWc glial cells glial precursors, 242–246 integrin roles, 611–612 neural induction and central–peripheral nervous system diVerentiation, 225–227 neuroepithelial cell features and diVerentiation, 229–231, 233–235, 247–249 neurogliogenesis zones and growth factors, 235–239 QKI role, 654–655 radial glial cell diVerentiation, 233–235 regionalization of undiVerentiated cells, 227–230 stages, 223 stem cells apoptosis versus proliferation, 249–252 diVerentiation models, 239–241 glial precursor transition, 246–249 Neural stem cell apoptosis versus proliferation, 249–252 diVerentiation models, 239–241 engraftment, see Glial cell transplantation

1177

1178

INDEX

Neuregulins integrin interactions in myelinating glia, 619–622 neuregulin-1 in Schwann cells development role, 341–343 signaling, 343–344 oligoneogenesis role, 262 Neurexin IV/Caspr/Paranodin (NCP) family Caspr member functions, 588–589 functional conservation and evolutionary signiWcance, 593 overview, 585–586 septate junctions neurexin IV in Drosophila, 586 NCP1 role at paranodal junctions and domain organization at node of Ranvier, 589–592 vertebrates, 586, 588 cytoskeleton linking through Band 4.1 proteins, 592 Neurexins, see also Neurexin IV/Caspr/Paranodin family domains, 580–581 dystroglycan binding, 584–585 genes expression, 581 promoters, 581 structures of superfamily members, 579–580, 582–583 transcription splice variants, 582 invertebrate homologs, 582–584 a-latroxin binding, 580 LNS domain structure, 585 neurexuophilin binding, 584 neurolignin binding, 584 synapse formation, 585 synaptotagmin binding, 582 Neuromodulin, Schwann cell diVerentiation marker, 338 Neurotrophin-3 (NT3) oligoneogenesis role, 262, 294 Schwann cell myelination regulation, 353 Nodes of Ranvier ankyrinG interactions sodium channels, 91–93 spectrin, 93–94 central nervous system specializations, 95–96 developmental assembly, 103–104 internodal region, 102–103 juxtaparanodes functions, 105 potassium channels, 105, 120 specializations, 101 structure, 16–17 length and diameter, 14 paranodal regions functions, 104–105 specializations adherens junctions, 99 gap junctions, 99 septate-like junctions, 99–101 tight junctions, 97–99 structure, 16 pathology, 105–106 peripheral nervous system specializations, 94–95 sodium channels, 15, 89–91, 104, 118, 120 sodium/potassium ATPase, 94, 120 transmission electron microscopy, 15–16

Non-compact myelin, structure and function, 10–13 Notch astrocyte development signaling, 322–323 glial determination in Drosophila, 208–209 NT3, see Neurotrophin-3

O2A cell, see Oligodendrocyte-type 2 astrocyte cell Oct-6 Schwann cell diVerentiation marker, 338 Schwann cell myelination regulation, 349–351 OEC, see Olfactory ensheathing cell Olfactory ensheathing cell (OEC) anatomy of peripheral olfactory system, 371–372 astrocyte interactions, 380–381 central–peripheral nervous system interface, 373–374 clinical potential, 380–381 development, 374 engraftment, 156–157, 278–279, 376–381 growth factor responses, 375–376 heterogeneity, 374–375 myelination Schwann cell myelination similarity, 378–379 transplanted cells, 376–380 olfactory system neurogenesis, 372–373 Oligodendrocyte cytoskeleton role in cell diVerentiation and myelin biogenesis, 41–42 diVerentiation and initiation of myelination, 42–45 engraftment, see Glial cell transplantation ganglioside regulation axo-glial communication and nerve regeneration, 73–74 diseases, 74 ganglioside expression as a function of cell development, 72–73 GD3 and cell migration, 73 GM3 and cell diVerentiation, 73 overview, 71–72 survival inXuences, 73 injury within demyelinating lesions in multiple sclerosis, 743–745 integrins expression of integrins, 613–614 functions of integrins diVerentiation, 617 migration, 616 proliferation, 616 survival, 617–618 lineage development, see also Oligodendrocyte-type 2 astrocyte cell; Oligodendrocyte progenitor cell adult brain cells characterization and distribution, 300–301 functions, 303–304 maintenance in undiVerentiated state, 302–303 origins, 302 proliferation, 303 response to demyelination, 304 in vitro studies, 298, 300 in vivo morphological studies, 300 in vitro characterization biochemical and morphological characteristics, 290 cell number control, 293–294 diVerentiation potential, 290–292 migration, 293

INDEX

overview, 289 proliferation, 292–293 in vivo characterization of development cell number control, 297–298 molecular control of precursor speciWcation, 296–297 overview, 294 zones of development, 294–296 myelin synthesis rate, 57 proliferation and terminal diVerentiation regulation by glycosphingolipids, 68–71 remyelination, see Remyelination Oligodendrocyte progenitor cell (OPC) adult brain cells antigenic recognition, 260, 262–263, 266–267 characterization and distribution, 300–301 functions, 303–304 humoral control of oligoneogenesis, 261–264 isolation, 264–267 lineage potential, 260–261 maintenance in undiVerentiated state, 302–303 multipotential progenitors in human white matter, 267–269 origins, 302 overview, 259 proliferation, 303 response to demyelination, 304 turnover, 260 in vitro studies, 298, 300 in vivo morphological studies, 300 white matter cell distribution and heterogeneity, 269 lineage relationships between glial progenitors, 244–246 remyelination diVerentiation failure, 184 mediators, 181–182 dysregulation hypothesis in failure, 184–185 recruitment failure, 183–184 mediators, 180–181 remyelinating cell origins, 178–179 transplantation studies in humans adult brain sources, 271 congenitally-dysmyelinated brain studies, 274 diVerentiation potential after implantation, 277–278 fetal brain sources, 270–271 fetal versus adult cell behavior, 274–277 integration of grafts, 271–272 migratory characteristics, 272 myelin synthesis, 272, 274 overview, 159–160, 269–270 propagated oligospheres, 271 Oligodendrocyte speciWc protein (OSP) functions, 565–566 knockout mouse phenotype, 565 Oligodendrocyte-type 2 astrocyte (O2A) cell features, 242–243 lineage relationships between glial progenitors, 244–246 OPC, see Oligodendrocyte progenitor cell OSP, see Oligodendrocyte speciWc protein

P0 adhesion functions double intraperiod line formation, 531–532 heterophilic adhesion function, 533 intermembrane space comparisons across species, 528–530, 536–537 major dense line formation, 532–533 compact myelin composition and function, 6–8, 34 cytoskeletal interactions, 40 distribution in nervous system, 526 gene cloning, 523–524 developmental expression, 526–527 promoter, 527–528 regulation of expression, 526–528 structure, 524–525 homologs, 525–526 homophilic adhesion, 8–9 mutations Charcot–Marie–Tooth disease-1B gain-of-function animal models, 1151–1152 genotype–phenotype correlations, 919–920 loss-of function animal models, 1150–1151 overview, 917, 919, 1149–1150 genotype/phenotype comparisons and hereditary neuropathies, 536–538 loss- and gain-of-function eVects, 920–921 phenotypes, 530–531 post-translational modiWcations, 528, 534 prospects for study, 538–539 Schwann cells diVerentiation marker, 337–338 myelination events, 34–37 signaling, 533–534 structure, 8–9, 523–524, 534–536 targeting, 32, 528 traYcking, 528 P2 compact myelin composition and function, 6–8, 10 species distribution, 10 P22, compact myelin composition and function, 6–8 Paranodes, see Nodes of Ranvier Parkinson’s disease (PD), unfolded protein response, 1027–1028 PD, see Parkinson’s disease PDGF, see Platelet-derived growth factor Pelizaeus–Merzbacher disease (PMD) animal models jimpy mouse, 1130–1131 jimpy-4j mouse, 1131 jimpy-msd mouse, 1131 myelin-deWcient rat, 1132 overview, 1015–1016 proteolipid protein knockout mouse, 1135–1136 proteolipid protein overexpression in transgenic mice, 1134–1135 proteolipid protein point mutations, 1133–1134 rumpshaker mouse, 1132 axonal damage, 879–880 clinical presentation and course, 873–876 comparison with other leukodystrophies, 675, 686–687 diagnosis, 873–875 forms, 1130 genetic testing, 876 history of study, 867 pathogenesis

1179

1180

INDEX

Pelizaeus–Merzbacher disease (PMD) (Continued ) evidence for three distinct mechanisms, 1016 molecular pathogenesis, 877–880, 1016–1019 protein misfolding hypothesis, 1017–1019 unfolded protein response activation of response, 1023–1024 animal model studies, 1024–1025 proteolipid protein mutation studies, 1025 peripheral neuropathy, 876 proteolipid protein mutations duplications, 870, 1015 gain-of-function mutations, 873 genotype–phenotype correlations, 876–877, 1014–1015 loss-of-function mutations, 871–872, 1015 overview, 869–870 Periaxin gene structure, 633–634 transcription splice variants, 633 isoforms and structures, 633–635 knockout mouse phenotype, 638–640 L-periaxin dystroglycan complex axon–glia interaction role, 637–638 composition, 636–637 embryonic expression and localization, 635–636 mutations in Charcot–Marie–Tooth disease-4F knockout mouse model, 1155–115 overview, 640–641, 928 Peripheral myelin protein-22 (PMP-22) axonal pathology in loss of function, 21 compact myelin composition and function, 6–8, 10 dosage and disease, 10 functions, 556–557 gene cloning, 547 regulation of expression, 550–551 structure, 550 knockout mouse phenotype, 555 mutations Charcot–Marie–Tooth disease-1A conditional overexpression in mice, 1148–1149 knockout mouse, 1149 mutagen induced mutants, 1147 overexpression in transgenic mice, 1147–1148 overexpression in transgenic rat, 1148 overview, 912–916, 1144 Trembler mouse, 1144–1147 Trembler-J mouse, 1147 Trembler-Ncnp mouse, 1147 diseases, 551–552, 555 duplication and gain-of-function, 915–916 point mutations loss-of-function, 916–917 structure/function analysis, 552–555 nervous system distribution, 548 Schwann cell diVerentiation marker, 339 small vertebrate integral membrane glycoprotein family, 547 structure, 549–550 Periventricular leukomalacia (PVL), glial cell transplantation, 279–280 4-Phenylbutyrate, adrenoleukodystrophy management, 830 Phospholipids, biosynthesis and transport, 65–66

Plasma cell, multiple sclerosis inXammatory response, 738 Platelet-derived growth factor (PDGF) integrin interactions in myelinating glia, 619–622 oligodendrocyte progenitor cell recruitment, 180, 261, 293 PLP, see Proteolipid protein PMD, see Pelizaeus–Merzbacher disease PML, see Progressive multifocal leukencephalopathy PMP-22, see Peripheral myelin protein-22 Potassium channels inhibition and restoration of impulse conduction in demyelinated axons, 131–132 juxtaparanodes, 105, 120 Progesterone, Schwann cell myelination regulation, 353 Progressive multifocal leukencephalopathy (PML) clinical features, 968 deWnition, 963–964 epidemiology, 965–966 JC virus features, 964–965 management, 969–970 pathogenesis, 966 pathology, 966–968 radiographic Wndings, 968 virus-induced demyelination mechanisms, 953–954 Protein misfolding, see Conformational diseases; Unfolded protein response Proteolipid protein (PLP) axonal pathology in loss of function, 21–22 compact myelin composition and function, 6–10, 1127–1129 DM20 isoform, 401–404, 1125–1127 dosage and disease, 10, 45 expression patterns in nervous system, 412 gene evolutionary conservation, 404–405 loci, 1129 post-transcriptional regulation, 404 promoter, 402–403 structure, 1014 transcription regulation, 402–404 transcription splice variants, 401–402 history of study, 401 knockout mouse phenotype, 412–413, 1135–1136, 1158 lipid raft association, 76–78 lipophilin family homology, 405–409 membrane pore proteins and sensor functions, 414 mutant animal models for glial cell transplantation jimpy mouse, 147–148 myelin-deWcient rat, 147 shaking pup, 148 transgenic mouse, 148 mutation and disease clinical spectrum of mutations, 1129–1130 overview, 867–868 Pelizaeus–Merzbacher disease, see Pelizaeus–Merzbacher disease SPG2, see Complicated spastic paraplegia myelin assembly role, 45 overexpression in transgenic mice, 1134–1135 post-translational modiWcations, 411–412 processing, 45 prospects for study, 415 Schwann cell diVerentiation marker, 339 secreted forms, 413–414

INDEX

structure, 9, 45, 1125–1127 targeting, 32 traYcking, 410–411 translation and folding, 409–410 PVL, see Periventricular leukomalacia qkI alternative splicing and regulation, 645, 652–653 expression defects in quaking mouse, 644–645, 648 developmental expression in oligodendrocytes, 647–648 nervous system, 646–647 homologs, 646 mouse model, see Quaking mouse protein functions apoptosis modulation by QKI7, 650, 653–654 cardiovascular system development, 655–656 human immunodeWciency virus replication regulation, 656 nervous system development, 654–655 nuclear localization of QKI5, 649–650 RNA binding and translocation, 648–652 structure, 645

Quaking mouse myelin protein abnormalities, 435, 644 phenotype and biochemistry, 643–644 qkI gene expression defects, 644–645, 648

RAB7, mutations in Charcot–Marie–Tooth disease2B, 930 Radial component, central nervous system myelin, 19 Radial glia astrocyte development, 312–314 diVerentiation, 233–235 Rafts, see Lipid rafts Refsum disease, demyelinating neuropathy, 929 Remyelination experimental autoimmune encephalitis, 177, 1060–1061 experimental models central nervous system barbotage, 178 experimental autoimmune encephalitis, 177 mouse viral models, 177 toxin models, 177–178 inXammatory response role, 182–183 morphological features, 173, 175–176 multiple sclerosis, 176, 745–747 oligodendrocyte progenitor cells diVerentiation failure, 184 mediators, 181–182 dysregulation hypothesis in failure, 184–185 recruitment failure, 183–184 mediators, 180–181 remyelinating cell origins, 177–178 Schwann cells markers, 185–186 origins, 187–188 speciWcity of remyelination, 186–187 stages, 178 Theiler’s murine encephalomyelitis virus studies, 1088–1089 therapeutic promotion

growth factors, 188 immunoglobulins, 188–189 prospects, 189 transplantation studies, see Glial cell transplantation Rubella, acute disseminated encephalomyelitis, 961 Rumpshaker mouse, Pelizaeus–Merzbacher disease model, 1132

Safety factor demyelinated axons, 121, 124 myelinated axons, 121 Saltatory conduction, myelinated axons, 121 Saposin A, knockout mouse as Krabbe’s disease model, 1105–1106, 1109 Schmidt–Lanterman incisures functions, 18 structure, 17–18 Schwann cell cytoskeleton role in diVerentiation and myelin biogenesis, 37–40, 354–357 development diVerentiation, 33–34 early peripheral nervous system glia developmental potential, 339–341 markers of diVerentiation, 336–339 neural crest origins, 333–335 neuregulin-1 role, 341–344 overview, 329–330, 332 precursors, 335–336 regulation of precursor–Schwann cell transition, 344–345 survival signals, 345–347 engraftment, see Glial cell transplantation integrins expression of integrins, 613 functions of integrins diVerentiation, 615–616 migration, 614–615 proliferation, 615 survival, 616 laminin binding, 355–356 laminin receptors, 354–357 myelination, 34–37, 349–353 neural activity, 353–354 proliferation, 347–349 remyelination, see Remyelination Septate junctions Neurexin IV/Caspr/Paranodin family proteins cytoskeleton linking through Band 4.1 proteins, 592 NCP1 role at paranodal junctions and domain organization at node of Ranvier, 589–592 vertebrates, 586, 588 neurexin IV in Drosophila, 586 paranodal regions, 99–101 Shaking pup, glial cell transplantation model, 148 Shiverer mouse, glial cell transplantation model, 147 Siglec-4a, see Myelin-associated glycoprotein Smallpox, acute disseminated encephalomyelitis, 960–961 SNAREs, vesicular transport and membrane biogenesis in polarized cells, 30–32

1181

1182

INDEX

Sodium channels ankyrinG interactions, 91–93 continuous conduction in demyelinated axons, 126–127 inhibition conduction block, 123–124 neuroprotection, 134 membrane reorganization in demyelinated axons, 126–129 multiple sclerosis channelopathy, 129 nodes of Ranvier, 15, 89–91, 104, 118, 120 Sodium/potassium ATPase inhibition and restoration of impulse conduction in demyelinated axons, 131, 133 nodes of Ranvier, 94, 120 Sox10 knockout mouse phenotype, 1158–1159 Schwann cell myelination regulation, 351 Spastic paraplegia type 2, see Complicated spastic paraplegia SPG2, see Complicated spastic paraplegia Sphingolipids biosynthesis and transport, 59–65 oligodendrocyte proliferation and terminal diVerentiation regulation by glycosphingolipids, 68–71 rafts, see Lipid rafts

T-cell experimental autoimmune encephalitis leukocyte migration, 1050–1051 major histocompoatibility complex molecule expression and T-cell activation, 1051–1052 populations in inXammation class I major histocompatibility complexrestricted cytotoxic T-cells, 1047–1049 Th-1 cells, 1046 Th-2 cells, 1046–1047 target antigens, 1043–1045 tissue damage mechanisms, 1055–1056 multiple sclerosis tissue damage antibody modulation, 754 mechanisms activated macrophage mediated injury, 753–754 complement activation, 753 excitotoxins, 753 free radicals, 753 proteases, 752 tumor necrosis factor, 752 overview, 751–752, 793–794 myelin oligodendrocyte glycoprotein encephalitogenic T-cell responses, 482–486 Theiler’s murine encephalomyelitis virus (TMEV) axonal damage mechanisms, 1087–1088 capsid structure, 1074–1075

demyelination mechanisms bystander eVects, 1085–1086 direct virus eVects on oligodendrocytes, 1083–1085 epitope spreading, 1086–1087 molecular mimicry, 1087 disease pathology, 1077–1078, 1080 genetic resistance and susceptibility, 1082–1083 genome, 1075 life cycle, 1075–1076 persistence of infection, 1080–1081 receptor, 1076–1077 remyelination studies, 1088–1089 subgroups, 1074 Tight junctions, see also Claudin 11 tight junctions cytoplasmic scaVold proteins, 566 paranodal regions, 97–99 TMEV, see Theiler’s murine encephalomyelitis virus Trembler mouse, Charcot–Marie–Tooth disease-1A models Trembler mouse, 1144–1147 Trembler-J mouse, 1147 Trembler-Ncnp mouse, 1147 Triodothyronine, oligoneogenesis role, 262 Twitcher mouse biochemistry, 1103–1104 pathology, 1102–1103 phenotype, 1102

Unfolded protein response (UPR) Alzheimer’s disease, 1028 Charcot–Marie–Tooth disease, 1026 cystic Wbrosis, 1026–1027 disease screening criteria dominant alleles, 1029 null alleles, 1028–1029 recessive alleles, 1029 eVector protein constitutive expression, 1022–1023 leukencephalopathy with vanishing white matter, 1027 Parkinson’s disease, 1027–1028 Pelizaeus–Merzbacher disease activation of response, 1023–1024 animal model studies, 1024–1025 proteolipid protein mutation studies, 1025 signaling components, 1019–1022 UPR, see Unfolded protein response

VEP, see Visual evoked potential Very long chain fatty acids, see Adrenoleukodystrophy Visual evoked potential (VEP), multiple sclerosis diagnosis, 692–693 VWM, see Leukencephalopathy with vanishing white matter