Astrocytes in Brain Aging and Neurodegeneration

NEUROSCIENCE I N T E L L I G E N C E U N I T

3

Hyman M. Schipper

Astrocytes in Brain Aging and Neurodegeneration

R.G. LANDES C OM PA N Y

NEUROSCIENCE INTELLIGENCE UNIT

Astrocytes in Brain Aging and Neurodegeneration Hyman M. Schipper Department of Neurology and Neurosurgery Department of Medicine (Geriatrics) and Centre for Studies in Aging McGill University and Bloomfield Centre for Research in Aging Lady Davis Institute for Medical Research Sir Mortimer B. Davis-Jewish General Hospital Montreal, Quebec, Canada

R.G. LANDES COMPANY AUSTIN, TEXAS U.S.A.

NEUROSCIENCE INTELLIGENCE UNIT Astrocytes in Brain Aging and Neurodegeneration R.G. LANDES COMPANY Austin, Texas, U.S.A. Copyright © 1998 R.G. Landes Company All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Printed in the U.S.A. Please address all inquiries to the Publishers: R.G. Landes Company, 810 South Church Street, Georgetown, Texas, U.S.A. 78626 Phone: 512/ 863 7762; FAX: 512/ 863 0081

ISBN: 1-57059-489-9

While the authors, editors and publisher believe that drug selection and dosage and the specifications and usage of equipment and devices, as set forth in this book, are in accord with current recommendations and practice at the time of publication, they make no warranty, expressed or implied, with respect to material described in this book. In view of the ongoing research, equipment development, changes in governmental regulations and the rapid accumulation of information relating to the biomedical sciences, the reader is urged to carefully review and evaluate the information provided herein.

Library of Congress Cataloging-in-Publication Data

Astrocytes in brain aging and neurodegeneration / [edited by] Hyman M. Schipper. p. cm. -- (Neuroscience intelligence unit) ISBN 1-57059-489-9 (alk. paper) 1. Nervous system--Degeneration. 2. Nervous system--Aging. 3. Astrocytes. I. Schipper, Hyman M., 1954- . II. Series. [DNLM: 1. Neurodegenerative Diseases--physiopathology. 2. Astrocytes-physiology. 3. Brain Diseases--physiopathology. 4. Brain--physiology. 5. Aging-physiology. WL 300A859 1998] RC365.A88 1998 616.8'047--dc21 DNLM/DLC 98-26335 for Library of Congress CIP

NEUROSCIENCE INTELLIGENCE UNIT PUBLISHER’S NOTE

Astrocytes in Brain Aging and Neurodegeneration

Landes Bioscience produces books in six Intelligence Unit series: Medical, Molecular Biology, Neuroscience, Tissue Engineering, Biotechnology and Environmental. The authors of our books are acknowledged leaders in their fields. Topics are unique; almost without exception, no similar books exist on these topics. Our goal is to publish books in important and rapidly changing areas of bioscience for sophisticated researchers and clinicians. To achieve this goal, we have accelerated our publishing program to conform to the fast pace at which information grows in bioscience. Most of Department our books are of published within 90Neurosurgery to 120 days of receipt of Neurology and the manuscript. We would like to thank our readers for their Department of Medicine (Geriatrics) continuing interest and welcome any comments or suggestions they and Centre for Studies in Aging may have for future books. McGill University

Hyman M. Schipper

and Judith Kemper Bloomfield Centre for Research in Aging Lady Davis Institute for MedicalProduction Research Manager R.G. Company Sir Mortimer B. Davis-Jewish GeneralLandes Hospital Montreal, Quebec, Canada

R.G. LANDES COMPANY AUSTIN, TEXAS U.S.A.

DEDICATION To my parents, Freda and Mendel, for their unflagging devotion.

CONTENTS Part I: Biology of Astrocytes 1. Astrocyte Ontogenesis and Classification ................................................ 3 James E. Goldman Genesis of Radial Glia and Their Transformation into Astrocytes ....... 4 Genesis of Astrocytes from SVZ Cells .................................................... 5 Control of Astrocyte Differentiation ...................................................... 6 Genesis of Astrocyte Heterogeneity ........................................................ 7 Generation of Astrocytes in the Adult CNS ........................................... 8 2. Functions of Astrocytes........................................................................... 15 Harold K. Kimelberg and Michael Aschner Introduction ........................................................................................... 15 Functions of Astrocytes ......................................................................... 16 Homeostasis of the Extracellular Space ................................................ 17 Transmitter Uptake Systems ................................................................. 21 Receptors for Transmitters ................................................................... 22 Astrocytes and the Blood-Brain Barrier (BBB) .................................... 26 Astrocytes and Immune and Inflammatory Responses in the CNS ... 28 3. Astrocyte Pathophysiology in Disorders of the Central Nervous System ............................................................... 41 Michael D. Norenberg Introduction ........................................................................................... 41 Normal Functions ................................................................................. 41 General Response to Injury ................................................................... 42 Injury to Astrocytes in CNS Disorders (Passive Role) ........................ 43 Active Role of Astrocytes in CNS Disorders ........................................ 44 Clinical Considerations ......................................................................... 47 Perspectives and Conclusions ............................................................... 53 Part II: Astrocytes in Human Brain Senescence and Neurodegenerative Disorders 4. Glial Responses to Injury, Disease, and Aging ...................................... 71 Lawrence F. Eng and Yuen Ling Lee Introduction ........................................................................................... 71 Astrocyte Intermediate Filament, Glial Fibrillary Acidic Protein ....... 71 Astrocytes in Experimental Gliosis ....................................................... 73 Astrocytes in Disease ............................................................................. 73 Astrocyte Activation of GFAP in Astrogliosis ...................................... 74 Microglial Activation ............................................................................. 74 Monocyte/Macrophage Activation ....................................................... 75 Endothelial Cell Activation ................................................................... 75 Astrocytes in Normal Aging .................................................................. 75 Astrocyte Inclusions in Normal Aging ................................................. 77 Astrocyte Inclusions in Disease ............................................................. 78

5. Astrocyte Pathology in Alzheimer Disease ............................................ 91 Jerzy Wegiel and Henryk M. Wisniewski Neuropathological Changes in Alzheimer Disease .............................. 91 Relationships Between Amyloid-β, Neurons, and Glial Cells in AD ......................................................................... 91 Astrogliosis in Aging and AD ................................................................ 93 Astrocyte Degeneration in AD .............................................................. 99 6. Parkinson’s Disease ............................................................................... 111 Donato A. Di Monte Introduction ......................................................................................... 111 Idiopathic Parkinson’s Disease ........................................................... 111 MPTP-Induced Parkinsonism ............................................................ 113 Neuronal-Astrocyte Interactions in Nigrostriatal Degeneration ...... 115 Conclusion ........................................................................................... 121 7. Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases) ..................................................................................... 127 Pawel P. Liberski, Radzislaw Kordek, Paul Brown and D. Carleton Gajdusek Introduction ......................................................................................... 127 KURU ................................................................................................... 130 Creutzfeldt-Jakob Disease (CJD) and Gerstmann-Straussler-Scheinker Disease (GSS) .................... 130 GSS ....................................................................................................... 135 The Involvement of Astrocytes in Formation of Amyloid Plaques ......................................................................... 137 Scrapie, Bovine Spongiform Encephalopathy (BSE), and Chronic Wasting Disease (CWD) ........................................... 137 BSE and CWD ...................................................................................... 143 Interaction Between Astrocytes and Oligodendrocytes ..................... 143 A Particular Form of Astrocytic Reaction in TSES ............................ 145 Expression of Glial Fibrillary Acidic Protein (GFAP) and Its mRNA .................................................................................. 145 Astrocytes and the Expression of Cytokines ...................................... 149 Conclusions ......................................................................................... 153 8. Astrocytes in Other Neurodegenerative Diseases ............................... 165 Dennis W. Dickson Introduction ......................................................................................... 165 Neurofibrillary Tangles as an Archetype of Cytoskeletal Inclusions ............................................................... 167 Neurodegenerative Disorders with Filamentous Glial Inclusion Bodies .............................................................................. 169 Progressive Supranuclear Palsy (PSP) ................................................ 171 Pick’s Disease ....................................................................................... 175

Corticobasal Degeneration (CBD) ..................................................... Argyrophilic Grain Dementia (AGD) ................................................ Familial Frontotemporal Dementia and Parkinsonism Linked to Chromosome 17 (FTDP-17) ......................................... Multiple System Atrophy (MSA) ........................................................ Familial Amyotrophic Lateral Sclerosis (FALS) .................................

176 179 180 180 181

Part III: Experimental Models of Astrocyte Senescence: Implications for Neurodegenerative Disease 9. The Peroxidase-Positive Subcortical Glial System .............................. 191 Marc B. Mydlarski, James R. Brawer and Hyman M. Schipper Introduction ......................................................................................... 191 Tinctorial and Histochemical Features .............................................. 191 Topography of the Peroxidase-Positive Astroglia ............................. 192 Modulation of the Peroxidase-Positive Glial System ........................ 193 Peroxidase-Positive Astrocytes in Primary Culture ........................... 196 Subcellular Precursors of Peroxidase-Positive Astroglial Inclusions ........................................................................ 197 Summary and Conclusions ................................................................. 202 10. Astrocyte Granulogenesis and the Cellular Stress Response .............. 207 Marc B. Mydlarski and Hyman M. Schipper HSP Expression in Acutely-stressed Neural Tissues: Effects of Aging ................................................................................ 208 Stress Protein Expression in the Aging and Degenerating Human Brain .................................................... 209 A Cellular Stress Model for the Biogenesis of Astroglial Inclusions ................................................................... 210 Astrocyte Senescence and the Origin of Corpora Amylacea ............. 221 11. Glial Iron Sequestration and Neurodegeneration ............................... 235 Hyman M. Schipper The Free Radical Hypothesis of Parkinson’s Disease ........................ 235 The Redox Neurobiology of Alzheimer’s Disease .............................. 235 Iron Deposition and Neurodegenerative Disease .............................. 236 Iron Sequestration in Aging Astroglia ................................................ 237 The Role of HO-1 in Brain Iron Deposition ...................................... 239 Pro-toxin Bioactivation by Astrocytes in Primary Culture ............... 242 Pathological Glial-Neuronal Interaction in Parkinson’s Disease ..... 243 Conclusion ........................................................................................... 246 Index ................................................................................................................ 253

EDITOR Hyman M. Schipper Department of Neurology and Neurosurgery Department of Medicine (Geriatrics) and Centre for Studies in Aging McGill University and Bloomfield Centre for Research in Aging Lady Davis Institute for Medical Research Sir Mortimer B. Davis-Jewish General Hospital, Montreal, Quebec, Canada Chapters 9, 10, 11

CONTRIBUTORS Michael Aschner Department of Physiology and Pharmacology Bowman Gray School of Medicine Winston-Salem, North Carolina, U.S.A. Chapter 2 James R. Brawer Department of Anatomy and Cell Biology McGill University Montreal, Quebec, Canada Chapter 9 Paul Brown Laboratory of Central Nervous System Studies, National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, Maryland, U.S.A. Chapter 7 Donato A. Di Monte The Parkinson’s Institute Sunnyvale, California, U.S.A. Chapter 6 Dennis W. Dickson Research Department Mayo Clinic Jacksonville Jacksonville, Florida, U.S.A. Chapter 8

Lawrence F. Eng Pathology Research VAPA Health Care System Palo Alto, California and Stanford University School of Medicine Stanford, California, U.S.A. Chapter 4 James E. Goldman Department of Pathology and The Center for Neurobiology and Behavior Columbia University College of P&S New York, New York, U.S.A. Chapter 1 D. Carleton Gajdusek Laboratory of Central Nervous System Studies, National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, Maryland, U.S.A. Chapter 7 Harold K. Kimelberg Department of Pharmacology and Neuroscience Division of Neurosurgery Albany Medical College Albany, New York, U.S.A. Chapter 2

Radzislaw Kordek Laboratory of Central Nervous System Studies National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, Maryland, U.S.A. and Laboratories of Tumor Biology Laboratory of Electron Microscopy and Neuropathology Medical Academy Lodz Lodz, Poland Chapter 7 Yuen Ling Lee Pathology Research VAPA Health Care System Palo Alto, California and Stanford University School of Medicine Stanford, California, U.S.A. Chapter 4 Pawel P. Liberski Laboratory of Central Nervous System Studies National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, Maryland, U.S.A. and Laboratories of Tumor Biology Laboratory of Electron Microscopy and Neuropathology Medical Academy Lodz and Laboratory of Electron Microscopy Department of Pathology Polish Mother Memorial Hospital Lodz, Poland Chapter 7

Marc B. Mydlarski Department of Neurology and Neurosurgery McGill University and Bloomfield Centre for Research in Aging Lady Davis Institute for Medical Research Sir Mortimer B. Davis-Jewish General Hospital Montreal, Quebec, Canada Chapters 9, 10 Michael D. Norenberg Laboratory of Neuropathology Veterans Administration Medical Center and Departments of Pathology, and Biochemistry and Molecular Biology University of Miami School of Medicine Miami, Florida, U.S.A. Chapter 3 Jerzy Wegiel Department of Pathological Neurobiology New York State Institute for Basic Research in Developmental Disabilities Staten Island, New York, U.S.A. Chapter 5 Henryk M. Wisniewski Department of Pathological Neurobiology New York State Institute for Basic Research in Developmental Disabilities Staten Island, New York, U.S.A. Chapter 5

PREFACE

T

he last decade or so has witnessed a remarkable proliferation of original scientific papers, review articles and books devoted to the neuroglia and their involvement in health and disease. In the prefaces to the many excellent compendia currently available on this topic, the editors almost invariably take pains to point out that for almost 150 years the study of neuroglia in general, and astrocytes in particular, has been largely eclipsed by the effort to decipher the properties of what has traditionally been regarded as the “business” end of the nervous system, the neurons and their connections. To be sure, no one would deny the paramount importance of neurons to the workings of the brain and its ailments. Yet, there is a rapidly-growing awareness, fueled by a biotechnological prowess permitting exquisitely refined analyses of cellular behavior, that the astroglia engage in intimate, mutually-dependent interactions with virtually all neural cell types, including neurons, and subserve a multitude of adaptive functions vital to the maintenance of normal brain structure and activity. To cite but a few examples, astrocytes are known to assume pivotal roles in the establishment of the blood-brain barrier and the regulation of ion homeostasis, the elaboration of a scaffolding for neuronal migration during embryogenesis, the sequestration and metabolism of various neurotransmitters and other neuroactive substances, and the production of immunomodulatory and proinflammatory cytokines and neuropeptides. In this regard, it should come as no surprise that astrocyte dysfunction resulting from injury or disease may mediate a host of dystrophic effects within the CNS and thereby contribute to a decline in neurological status. The formation of epileptogenic scar tissue in response to CNS trauma, the release of excitotoxic amino acids following tissue hypoxia, metal exposure or oxidative stress, neoplastic transformation and malignant behavior, and the bioactivation of pro-toxins (such as MPTP) to potent neurotoxins (MPP+) are illustrative of some clinically-relevant pathophysiologic processes which directly implicate the astroglial compartment. Astrocyte hypertrophy and hyperplasia, the biosynthesis of GFAP-associated intermediate filaments (reactive gliosis) and the accumulation of discrete cytoplasmic inclusions are characteristic pathological features of the major aging-related neurodegenerative disorders, including Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis. Gliosis and inclusion body formation also figure prominently in the relatively uncommon human neurodegenerative conditions, such as Pick’s disease and corticobasal ganglionic degeneration, and occur to a lesser extent in the course of normal brain aging. The raison d’être of this monograph was to consolidate information concerning the established and putative roles of astroglia in brain aging and neurodegeneration gleaned from vast and often disparate literatures on the biology and pathology of these cells. To achieve this objective, I invited the participation of respected investigators from a mix of basic and clinical departments whose interests in the neuroglia are diverse and long-standing. In addition to providing thorough reviews of their respective fields, each team of

contributors was requested to speculate freely on the question “In the condition under consideration, do the astrocytic changes actively contribute to the degenerative process or do they merely represent passive responses to primary neuronal injury?” Given the divergence of opinion on this question, a certain degree of overlap of material covered by the authors (e.g., the role of astroglia in Alzheimer’s disease) was not only tolerated but encouraged. The chapters in this monograph are grouped in three sections: I. Biology of Astrocytes. Collectively, the chapters in this section constitute a comprehensive discussion of the origin and known functions of astroglia in the mammalian CNS and the roles these cells may play in the pathophysiology of neurological disorders. II. Astrocytes in Human Brain Senescence and Neurodegenerative Disorders. In this section, detailed accounts of the pathology of astrocytes and their involvement in human brain aging and various neurodegenerative conditions are presented. III. Experimental Models of Astrocyte Senescence: Implications for Neurodegenerative Disease. In this final part, experimental approaches to the delineation of the role of astroglia in brain aging and degeneration are described. We hope that this compendium will appeal to basic neuroscientists interested in various aspects of neuroglial biology, as well as to clinically-oriented investigators concerned with the pathogenesis of the major human neurodegenerative disorders. I am deeply grateful to the many mentors, colleagues and students at home and abroad who have helped shape my interest and refine my knowledge of the neuroglia and their place in clinical medicine. Hyman M. Schipper

Part I Biology of Astrocytes

CHAPTER 1

Astrocyte Ontogenesis and Classification James E. Goldman

A

strocytes, first named for their star-shaped appearance as visualized with heavy metal impregnations,1 in fact display a extensive variety of morphologies. All are united in their astrocyte nature, however, by common features, including multiple, thin processes, close interactions with both the neuronal and mesenchymal elements of the CNS, the presence of intermediate filaments of several types (vimentin, GFAP, nestin), and the expression of a variety of other molecules, such as S-100β and glutamine synthetase. Besides their complex, multiprocess shapes the other salient histological characteristic of astrocytes is their interactions with specific sets of other cells. First, the basal lamina that surrounds blood vessels in the brain and that lines the pial surface of the brain is covered with astrocyte end feet (the ends of astrocyte processes).2 This is indeed a very large surface area, and thus requires an exceedingly large number of astrocyte processes. Second, astrocytes intimately associate with neurons, wrapping neuronal perikarya and dendrites, contacting neurons in zones between synaptic contacts.2-4 Thus, astrocytes serve to isolate individual synapses or groups of synapses, perhaps those that share functional connections or characteristics. Such isolation of synapses makes sense in view of the astrocytes’ abilities to take up neurotransmitters with high affinities and to buffer potassium (see chapter 2). These interactions may well serve to condition and maintain astrocyte shape (see below). Astrocytes are not distributed randomly in the brain, but rather lie in separate domains with some peripheral overlap. For example, the “domain” of a neocortical astrocyte is roughly spherical with a diameter of about 100 microns.5 Similarly, “domains” of retinal astrocytes are spatially separate at 100-150 microns, with a modest degree of overlap in the peripheral processes.6 Subpial astrocytes are not spherical, but look like truncated spheres or columns.7 Thus, astrocyte development must somehow produce a matrix of astrocyte spheres which intersect only at their peripheries. It is at the periphery, incidentally, that astrocytes are connected by gap junctions, allowing movement of ions and small molecules through an astrocyte “syncytium” over many hundreds of cubic microns.8 How this regular spacing is accomplished is not known. Since glia continue to divide as they migrate through the brain (see below), sibling astrocytes begin life next to each other after a mitotic division. Do they migrate away from each other, or does the growth of the brain continue to separate related glial cells? Astrocytes display a wonderful variety of sizes and shapes. In most gray matter regions, where astrocytes have been traditionally termed “protoplasmic,” the cell body and domains of all processes roughly describe a sphere or ellipsoid. Processes branch into ever-finer twigs, more like the boughs of a tree than the rays of a star, eventually reaching tremendous numbers Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

4

Astrocytes in Brain Aging and Neurodegeneration

and microscopic size.3 In the cerebellar granule cell layer, “velate” astrocytes wrap thin sheetlike extensions about the mossy fiber glomeruli.9,10 Astrocytes in white matter (classically termed “fibrous”) display fewer processes and a less complex branching pattern than their gray matter relatives. Processes separate fascicles of axons, a characteristic particularly easily observed in spinal cord tracts and optic nerve.11,12 Astrocyte processes also contact nodes of Ranvier,13 where they may play a role in spatial ion buffering. Some astrocytes display processes oriented “radially,” perpendicular to the pial surface. These include the Bergmann glia of the cerebellar molecular layer, which retain their radially-oriented nature first established for granule cell migration.9,14 Muller glia of the retina are also first established as radially oriented cells, coursing through all retinal layers, and remain so for life. Radial type glia in periventricular regions, particularly around the third and fourth ventricles and aqueduct of Sylvius, display long processes beginning at the ventricular surface and extending for hundreds of microns into the parenchyma of the hypothalamus and brain stem.15-17

Genesis of Radial Glia and their Transformation into Astrocytes The term “radial glia” is used to describe elongated, bipolar glial cells that arise during early histogenesis of the CNS. Heavy metal impregnations and more recently, immunocytochemistry, have produced a detailed view of the radial glial scaffolding in the developing brain.18-20 Oriented “radially” (perpendicular to the pial surface), these cells extend from the ventricular zone to the pia and develop concurrently with the first sets of neurons.21-23 Not only are radial glia generated contemporaneously with some neuronal populations, but also they share a lineage with neurons, since early progenitors give rise to both radial glia and to the neurons that migrate along them.22-24 Thus, there is a glial-neuronal fate decision for a subpopulation of cells in the early ventricular zone, although how this decision is accomplished is not known. Radial glia have long been considered a form of astrocyte, based upon the expression of the intermediate filaments vimentin and nestin, and in primates, GFAP, as well as the storage of glycogen and the interactions with the pial surface, all characteristics of astrocytes. Furthermore, radial glia have been considered the source of many of the astrocytes in the mature CNS. This transformation of radial glia into astrocytes has been inferred from several observations. Radial glia disappear in cortex concurrently with the emergence of the multiprocess forms of mature astrocytes. During this time several studies have noted forms “transitional” between radial glia and astrocytes: cells with both long, radially-oriented processes and smaller branches emerging from the cell body.18,19,25,26 While undoubtedly some of these “transitional” forms reflect passing stages from radial glia to astrocytes, similar forms are produced by subventricular zone (SVZ) cells that migrate into the cortex after neurogenesis.27 Cells cultured from embryonic rodent CNS and expressing antigenic markers for radial glia begin to express GFAP in culture and assume the morphologies of cultured astrocytes.28 One dynamic study provides direct evidence for such a transformation, however.29 In work with postnatal ferret brain, the application of the lipophilic fluorescent tracer dye, diI, to the cortex initially labeled radial glia. After maturation of the brain and the disappearance of radial glia, the dye was found in astrocytes. What controls this transformation of radial glia to astrocytes and why does such transformation apparently take place in some regions (cortex, for example), but not, or to a lesser extent, in others (periventricular zones in diencephalon and brain stem)? Studies in cell culture suggest a role of extrinsic factors in promoting the change in shape from elongated to branched with many processes. Such a transformation takes place in primary cultures from embryonic forebrain,28 and can be reversibly promoted by soluble signals from the

Astrocyte Ontogenesis and Classification

5

embryonic CNS.30 Cerebellar astrocytes cocultured with granule neurons assume elongated shapes, suggesting that interactions with immature neurons helps determine astrocyte shape.31 A critical, and necessary, change in the transformation of radial glia is the loss of subpial connections. This process, which has not been examined, requires a loss of adhesion between the end of the glial process and the pial surface. Breaking such adhesion in turn may require local extracellular protease activity or redistribution of surface adhesion molecules such as integrins that may interact with mesenchymal tissue matrix, or contraction of the microfilament network in the process. Loss of adhesion to the pia does not represent a lack of adhesive properties of the cell in general, since radial glia that transform into astrocytes presumably contact blood vessels as they are detaching from the pia, or shortly thereafter.

Genesis of Astrocytes from SVZ Cells In addition to the generation of astrocytes from radial glia, astrocytes are also derived from immature cells of the subventricular zone, without apparently going through a radial intermediate stage. The genesis of astrocytes from immature cells in the forebrain SVZ was originally suggested from thymidine labeling in the postnatal rodent brain.32-34 These classic studies showed that the SVZ population is a highly proliferative one and that the thymidine label could be “chased” into mature glial cells in white matter and gray matter. More recent antigen expression studies35 and Golgi impregnations of the developing CNS7 have also supported a nonradial glial derivation of some astrocytes. Through the use of recombinant retroviruses, a direct demonstration of SVZ cell migration and differentiation into mature glia has illuminated many of the details of this developmental process.5,36-39 In these experiments, immature, cycling cells of the postnatal SVZ were labeled in vivo by stereotactic injection of retroviruses directly into the SVZ. The fates of labeled cells and their routes of migration into the striatum, overlying white matter, and neocortex could then be traced. How do astrocytes derived from the SVZ colonize the CNS? Glial colonization, to produce the distributions described above, is not a random process, but takes place in definable spatial and temporal patterns. Migration of progenitors from SVZ into white matter and cortex occurs in a coronal plane.36 Perhaps the migratory pathways are defined in part by the radial glial scaffolding. The idea that SVZ cells migrate along radial glia is supported by several observations. First, radial glia persist in the rodent neocortex through the first 1-2 postnatal weeks.20,25,26 During this period, SVZ cells distribute into white matter and cortex.5,37 By postnatal day 14 (P14), however, progenitors that migrate out of the SVZ remain in white matter and do not enter cortex.36 Thus, a restriction in migration coincides exactly with the loss of the cortical radial glial tracks. Second, we have observed progenitors from the SVZ aligning along radial glia in the cortex during early postnatal development.38 Third, progenitors from the SVZ migrate along “radial glial”-like cables in culture (Newman et al, in preparation). In contrast to the laminar colonization of neurons of the neocortex, however, astrocytes do not differentiate in a layered pattern. In fact, astrocytes derived from SVZ cells appear to differentiate at all depths of the cortex, from the pial surface to deep layers, at the same time. It is common to see radially oriented clusters of young astrocytes derived from SVZ cells, clusters we believe are clonal. At present we favor a model in which progenitors migrate into cortex, and continue to divide therein. Some of the progeny cease migration, while others continue toward the pial surface, thus leaving progeny behind at a number of cortical levels. What induces a particular progenitor to stop migrating and begin to differentiate into an astrocyte will be considered below.

6

Astrocytes in Brain Aging and Neurodegeneration

In other regions of the CNS that do not have an SVZ, the genesis of astrocytes may be different. For example, in the optic nerve, astrocytes arise prenatally, while those progenitors that migrate into and along the nerve in postnatal life do not differentiate into astrocytes, but only into oligodendrocytes.40,41 Thus, there appear to be separate lineages for astrocytes and oligodendrodcytes in this tract. Astrocytes in optic nerve likely arise from radial glial cells that formed during earlier telencephalic development and were carried into the nerve when the optic outpouching occurred. And the oligodendrocytic fate of progenitors that migrate into the nerve postnatally may be analogous to the oligodendrocytic fate of forebrain SVZ cells that settle in subcortical white matter. In culture, the postnatal progenitors (O-2A cells) can differentiate into oligodendrocytes or into astrocytes,42 showing their bipotential nature, but there is apparently an oligodendrocyte fate restriction in vivo. Astrocyte development in the spinal cord may be similar, with many of the astrocytes developing from radial glia, as suggested from antigen and morphology studies.43,44 In the cord, oligodendrocytes arise from proliferating, immature cells in the centro-ventral region,45,46 but whether astrocytes also arise from this proliferative population is not known.

Control of Astrocyte Differentiation Much recent work has utilized cell culture systems to examine the control of astrocyte differentiation, and has led to the general conclusion that cell-extrinsic factors contribute substantially to the determination of astrocyte cell fate. Oligodendrocyte progenitors isolated from the optic nerve are induced to express astrocyte genes and cease oligodendrocyte development by exposure to serum,42 serum fractions,47 and ciliary neurotrophic factor (CNTF).48 Although the nature of the serum stimulus(i) is not known, extracellular molecules isolated from endothelial and meningeal cells will also induce astrocyte genes,48,49 in combination with CNTF. This induction by matrix may well be an in vitro counterpart to astrocyte induction by cues from blood vessels and pia in vivo (see below). More recent studies identify CNTF as an attractive candidate for an important inducer of astrocytic differentiation in immature CNS cells.50-52 CNTF induces GFAP expression and a flat, astrocytic morphology in immature cortical cells via a JAK-STAT signaling pathway52 and also upregulates GFAP transcription in the CG-4 glial cell line.53 The 5' upstream region of the GFAP gene contains a consensus STAT binding site,52,54 which in transfection assays appears to be essential for the CNTF regulation of GFAP expression.52 The GFAP gene also contains consensus sequences for CREB, AP-2, and AP-1 binding,54,55 the former two possibly used for cyclic-AMP increases in GFAP transcription,54,56 the latter possibly utilized in stress-regulated increases in GFAP. Another candidate class of signaling molecules that can induce astrocyte differentiation are members of the transforming growth factor-β (TGF-β) family, in particular, the bone morphogenic proteins (BMP) 2 and 7, which cause astrocytic development and suppress oligodendrocytic development in bipotential progenitors from neonatal rat forebrain and immature cells expanded from embryonic CNS by epidermal growth factor (EGF).57,58 Notably, serum and BMPs can induce GFAP expression in progenitors that have already begun to express the early oligodendrocyte marker, O4, giving rise to a hybrid glial cell type. It is not known whether glial progenitors begin to express O4 and then become astrocytes in vivo, during normal glial development, but the possibility seems unlikely. However, under pathological conditions (such as the development of brain tumors composed of progenitor-like cells) the acquisition of astrocyte gene expression in oligodendrocyte lineage cells might occur. In contrast to BMPs, such growth factors as EGF, platelet-derived growth factor (PDGF), basic fibroblast growth factor (bFGF), thyroid hormone and insulin-like growth factor 1 (IGF1) do not promote astrocyte differentiation of O-2A progenitors. Rather, they either

Astrocyte Ontogenesis and Classification

7

promote division of progenitors without differentiation, as in the case of PDGF and bFGF,59 or are permissive for oligodendrocyte differentiation and/or survival.60,61 In most of the experiments cited above, astrocytic “differentiation” was measured by the induction of GFAP expression. While this intermediate filament protein is characteristic of astrocytes and therefore denotes the acquisition of at least one astrocyte feature, it is not clear whether there is a group of genes expressed coordinately during astrocyte development and whether in vitro systems fully capture that differentiated state. That other genes are both necessary and sufficient for astrocytic differentiation is clear from the several GFAP knockout transgenic mice, in which astrocytes do develop.62-64 In the future, control of specific receptors, transporters, and astrocyte enzymes will be required to characterize the developmental pathway. As discussed below, perhaps a progenitor makes several decisions during astrocyte development—the first to differentiate into an astrocyte and the second to acquire specific characteristics required for specific functions in the local CNS environment. Determinants of astrocyte fate in vivo has not been examined in as much detail as determinants in culture. Many of the factors suggested from the culture studies to play a role in fate determination exist in the developing brain. However, when a given progenitor becomes responsive to those signals and even whether such signals play a role in vivo is not yet known. Clues as to the nature of developmental signals may come from considering the anatomic changes that take place during astrocyte development. The peak period of astrocyte genesis coincides with the rapid growth of blood vessels65-67 and pial surface, the elaboration of dendritic arbors, and the establishment of synapses (both from cortical afferents and from intracortical circuits). For example, in the rat forebrain, thalamocortical afferents enter the cortex around P2-4 and cortico-cortical fibers around P6-8.68,69 Thus, the differentiation of astrocytes takes place during the establishment of synaptic connections and of the vascular supply. How is the development of glia coordinated with vascular and synaptic growth to assure the appropriate glial-vascular and glial-neuronal interactions? Furthermore, does the development of astrocytes and/or oligodendrocytes play a role in vascular growth or synapse formation? There is evidence for mutual interactions between astrocytes and endothelial cells. Astrocytes may participate in the formation of endothelial tight junctions, the anatomic substrate of the blood-brain barrier, and in inducing specific endothelial cell properties, such as polarization of transporters, increases in γ-glutamyl transaminase.70-72 Furthermore, the presence of astrocytes in the mammalian retina correlates with the presence of blood vessels.73 In examining the fates of progenitors from the SVZ after migration into the cortex, we have noted a close concordance between the early stages of astrocyte differentiation, as judged by an increase in intermediate filament expression and the beginnings of a complex, multiprocess cell shape and contact with blood vessels or the pial surface.39 These observations do not prove a causal relationship between astrocyte differentiation and vessel contact, but the model suggests a way in which astrocyte development can be coordinated with the tremendous growth of blood vessels and the pial surface in late gestational and postnatal CNS development.

Genesis of Astrocyte Heterogeneity Astrocytes vary both in morphology and in the expression of certain antigens from region to region. One example is the well known morphological distinctions between the “fibrous” astrocytes of white matter and the “protoplasmic” astrocytes of gray matter, the former expressing a much higher level of GFAP than the latter.74 A number of studies have clearly shown functional heterogeneity among astrocytes, although most of these experiments have been performed in vitro. Thus, astrocytes cultured from different regions of the

8

Astrocytes in Brain Aging and Neurodegeneration

CNS differ in their abilities to support process growth of neurons, in their responses to neurotransmitters, and in their expressions of proteoglycans.75-77 Astrocytes from one region appear to be matched functionally to support neurons of the same region; mesencephalic neurons grow better on mesencephalic astrocytes than on astrocytes from other regions, for example.77 In cultures from neonatal forebrain, which includes all cortical areas and white matter and some subcortical gray matter nuclei, there is a heterogeneity in the uptake of and responses to neurotransmitters within the astrocyte population.78,79 Whether this heterogeneity was determined in vivo before the cultures were established or in vitro is not clear, but the observations dramatically illustrate that astrocytes are able to acquire important functional differences. In another study, the clonal progeny of single spinal cord astrocytes in culture were examined, and both homogeneous and heterogeneous clones were observed,80 showing clearly that an individual proliferating astrocyte, or an individual progenitor, is able to generate a mixture of astrocytic forms. Less is known about heterogeneity in vivo, however, but techniques exist to study astrocyte physiology in slices, where responses to transmitters or uptake mechanisms could be studied in real time. In vivo retroviral labeling studies suggest (although do not yet prove) that different astrocyte forms can arise from a single progenitor. For example, the proximity of retrovirally labeled Bergmann glia and velate astrocytes in the cerebellar cortex suggests a clonal heterogeneity (Fig. 1.1).81,82 And, as noted above, the astrocytic progeny of a single progenitor in the neocortex probably span the entire cortical depth, and would therefore be exposed to different microenvironments. How is the heterogeneity of astrocytes determined? One model would suggest that progenitors first are induced to differentiate into astrocytes and then signals peculiar to the local environment dictate specific morphological and functional patterns. This model makes sense if an astrocyte’s functional properties must match those of the neurons in the immediate proximity. Thus, the heterogeneity of astrocytes may not be lineage related, in the sense that such heterogeneity has little to do with the astrocyte fate decision. Astrocytes can change morphology and expression of many molecules, including surface gangliosides, intermediate filaments, enzymes, and stress proteins, in response to pathological conditions (see for examples refs. 83, 84). So, even in the mature CNS, astrocytes maintain a remarkable malleability.

Generation of Astrocytes in the Adult CNS Thymidine labeling studies in the adult mammalian CNS show a low level of cell division in the mature CNS85-87 and several investigators have inferred a slow turnover of astrocytes. Genesis must be balanced by cell death, since numbers of astrocytes in the cortex do not appear to increase during adult life.88 The nature of the dividing cells is not clear; that is, astrocytes might be generated from dividing astrocytes or from dividing, immature cells that then differentiate into astrocytes. Under pathological conditions, such as trauma, astrocytes in the region of the lesion divide, although the capacity for proliferation appears limited.84 Whether new astrocytes are generated from immature cells in pathological circumstances is not known. Cycling cells in adult rat white matter, labeled with recombinant retroviruses, do not differentiate into astrocytes, either under normal conditions, demyelination, or trauma (refs. 89, 90 and our unpublished observations). This finding contrasts with studies that find a population of immature cells isolated from adult optic nerve, cord, or forebrain that can differentiate into either oligodendrocytes or astrocytes in culture (“adult O-2A progenitors”91,92). Again, there may be fate restrictions in vivo, or perhaps appropriate pathological conditions have yet to be found in vivo to induce astrocyte differentiation in cycling immature cells.

Astrocyte Ontogenesis and Classification

9

Fig. 1.1. Morphological transformations in the development of astrocytes as revealed by a Lac-Z encoding retrovirus. Newborn rat pups were injected into the forebrain SVZ or cerebellar white matter as described.5,82 Labeled cells were visualized by X-gal staining. (a) two unipolar cells in the SVZ, 1 day after injection. (b) a bipolar cell oriented radially in the cortex, 3 days after injection; such cells do not express astrocyte markers and presumably represent progenitors. (c) an early astrocyte in the cortex, 3 days after injection; one process has wrapped around a blood vessel (arrowhead); cells at this stage are expressing intermediate filament proteins. (d) two velate astrocytes in the cerebellar granule cell layer, 2 weeks after injection, displaying mature forms. (e) a Bergmann glial cell (top), with a cell body in the Purkinje cell layer and processes extending into the molecular layer, adjacent to a velate astrocyte (bottom) in the granule cell layer, 2 weeks after injection.

Acknowledgments The work from the author’s lab has been supported by NIH grant NS-17125. Many thanks to Bernetta Abramson, Cathy Chuang, JoAnn Gensert, Steven Levison, Sharon Newman, Marielba Zerlin, and Lei Zhang for all of their many major contributions to our studies.

10

Astrocytes in Brain Aging and Neurodegeneration

References 1. Andriezen WL. The neuroglia elements in the human brain. Br J Med 2:227-230. 2. Peters A, Palay SL, Webster H deF. The Fine Structure of the Nervous System, 3rd Ed. Oxford: Oxford University Press, 1991. 3. Hama K, Arii T, Kosaka T. Three-dimensional organization of neuronal and glial processes: high voltage electron microscopy. Microsc Res Tech 1993; 29:357-367. 4. Kosaka T, Hama K. Three-dimensional structure of astrocytes in the rat dentate gyrus. J Comp Neurol 1986; 249:242-260. 5. Levison SW, Goldman JE. Both oligodendrocytes and astrocytes develop from progenitors in the subventricular zone of postnatal rat forebrain. Neuron 1993; 10:201-212. 6. Chan-Ling T, Stone J. Factors determining the morphology and distribution of astrocytes in the cat retina: a ‘contact-spacing’ model of astrocyte interaction. J Comp Neurol 1991; 303:387-399. 7. Marin-Padilla M. Prenatal development of fibrous (white matter), protoplasmic (gray matter), and layer I astrocytes in the human cerebral cortex: a Golgi study. J Comp Neurol 1995; 357:554-572. 8. Dani JW, Chernjavsky A, Smith SJ. Neuronal activity triggers calcium waves in hippocampal astrocyte networks. Neuron 1992; 8:429-440. 9. Palay SL, Chan-Palay V. Cerebellar Cortex, Cytology, and Organization. New York: SpringerVerlag, 1974. 10. Chan-Palay V, Palay SL. The form of velate astrocytes in the cerebellar cortex of monkey and rat: high voltage electron microscopy of rapid Golgi preparations. Z Anat Entwicklungsgesch 1972; 138:1-19. 11. Bovolenta P, Liem RHK, Mason CA. Glial filament protein expression in astroglia in the mouse visual pathway. Brain Res 1987; 430:113-126. 12. Butt AM, Ransom BR. Visualization of oligodendrocytes and astrocytes in the intact rat optic nerve by intracellular injection of Lucifer yellow and horseradish peroxidase. Glia 1989; 2:470-475. 13. Sims TJ, Gilmore SA, Waxman SG. Radial glia give rise to perinodal processes. Brain Res 1991; 549:25-35. 14. Bovolenta P, Liem RKH, Mason CA. Development of cerebellar astroglia: transitions in form and cytoskeletal content. Dev Biol 1984; 102:248-259. 15. Seress L. Development and structure of the radial glia in the postnatal rat brain. Anat Embryol (Berl) 1980; 160:213-226. 16. Mori K, Ikeda J, Hayaishi O. Monoclonal antibody R2D5 reveals midsagittal radial glial system in postnatally developing and adult brainstem. Proc Natl Acad Sci USA 1990; 87:5489-5493. 17. Edwards MA, Yamamoto M, Caviness VS Jr. Organization of radial glia and related cells in the developing murine CNS. An analysis based upon a new monoclonal antibody marker. Neuroscience 1990; 36:121-144. 18. Choi BH, Lapham LW. Radial glia in the human fetal cerebrum: A combined Golgi, immunofluorescent, and electron microscopic study. Brain Res 1978; 148:295-311. 19. Schmechel DE, Rakic P. A Golgi study of radial glial cells in developing monkey telencephalon: morphogenesis and transformation into astrocytes. Anat Embryol (Berl.) 1979; 156:115-152. 20. Misson J-P, Edwards ME, Yamamoto M et al. Identification of radial glial cells within the developing murine central nervous system: studies based upon a new immunohistochemical marker. Dev Brain Res 1988; 44:95-108. 21. Levitt P, Cooper ML, Rakic P. Coexistence of neuronal and glial precursor cells in the cerebral ventricular zone of the fetal monkey: An ultrastructural immunoperoxidase analysis. J Neurosci 1981; 1:27-39. 22. Halliday AL, Cepko CL. Generation and migration of cells in the developing striatum. 1992; Neuron 9:384-396. 23. Gray G, Sanes J. Lineage of radial glia in the chicken optic tectum. Development 1992; 114:271-283.

Astrocyte Ontogenesis and Classification

11

24. Galileo DS, Gray GC, Owens G et al. Neurons and glia arise from a common progenitor in chicken optic tectum: demonstration with two retroviruses and cell-type-specific antibodies. Proc Natl Acad Sci USA 1990; 87:458-462. 25. Misson J-P, Takahashi T, Caviness VS Jr. Ontogeny of radial and other astroglial cells in murine cerebral cortex. Glia 1991; 4:138-148. 26. LeVine SM, Goldman JE. Embryonic divergence of oligodendrocyte and astrocyte lineages in developing rat cerebrum. J Neurosci 1988; 8:3992-4006. 27. Zerlin M, Levison SW, Goldman JE. Early patterns of migration, morphogenesis, and intermediate filament expression of subventricular zone cells in the postnatal rat forebrain. J Neuroscience 1995; 15:7238-7249. 28. Culican SM, Baumrind NL, Yamamoto M et al. Cortical radial glia: Identification in tissue culture and evidence for their transformation to astrocytes. J Neurosci 1990; 10:684-692. 29. Voigt T. Development of glial cells in the cerebral wall of ferrets: Direct tracing of their transformation from radial glia into astrocytes. J Comp Neurol 1989; 289:74-88. 30. Hunter KE, Hatten ME. Radial glial cell transformation to astrocytes is bidirectional: regulation by a diffusible factor in embryonic forebrain. Proc Natl Acad Sci USA 1995; 92:2061-2065. 31. Hatten ME, Liem RKH, Mason CA. Two forms of cerebellar glial cells interact differently with neurons in vitro. J Cell Biol 1984; 98:193-204. 32. Altman J. Proliferation and migration of undifferentiated precursor cells in the rat during postnatal gliogenesis. Exp Neurol 1966; 16:263-278. 33. Imamoto K, Paterson JA, Leblond CP. Radioautographic investigation of gliogenesis in the corpus callosum of young rats. I. Sequential changes in oligodendrocytes. J Comp Neurol 1978; 180:115-138. 34. Paterson JA, Privat A, Ling EA et al. Investigation of glial cells in semithin sections III. Transformation of subependymal cells into glial cells as shown by radioautography after 3H-thymidine injection into the lateral ventricle of the brain of young rats. J Comp Neurol 1973; 149:83-102. 35. Gressens P, Richelme C, Kadhim HJ et al. The germinative zone produces the most cortical astrocytes after neuronal migration in the developing mammalian brain. Biol Neonate 1992; 61:4-24. 36. Levison, SW, Chuang C, Abramson B et al. The migrational patterns and developmental fates of glial precursors in the rat subventricular zone are temporally regulated. Development 1993; 119:611-622. 37. Luskin MB, McDermott K. Divergent lineages for oliogodendrocytes and astrocytes originating in the neonatal forebrain subventricular zone. Glia 1994; 11:211-226. 38. Zerlin M, Levison SW, Goldman JE. Early patterns of migration, morphogenesis, and intermediate filament expression of subventricular zone cells in the postnatal rat forebrain. J Neurosci 1995; 15:7238-7249. 39. Zerlin M, Goldman JE. Interactions between glial progenitors and blood vessels during early postnatal corticogenesis: blood vessel contact represents an early stage of astrocyte differentiation. J Comp Neurol 1997; 387:537-546. 40. Skoff RP. Gliogenesis in rat optic nerve: Astrocytes are generated in a single wave before oligodendrocytes. Dev Biol 1990; 139:149-168. 41. Fulton BP, Burne JF, Raff MC. Visualization of O-2A progenitor cells in developing and adult rat optic nerve by quisqualate-stimulated cobalt uptake. J Neurosci 1995; 12:4816-4833. 42. Raff MC, Miller RH, Noble M. A glial progenitor cell that develops in vitro into an astrocyte or an oligodendrocyte depending on culture medium. Nature 1983; 303:390-396. 43. Hirano M, Goldman JE. Gliogenesis in rat spinal cord: Evidence for origin of astrocytes and oligodendrocytes from radial precursors. J Neurosci Res 1988; 21:155-167. 44. Maier CE, Miller RH Development of glial architecture in the frog spinal cord. Dev Neurosci 1995; 178:149-159. 45. Warf BC, Fok-Seang J, Miller RH. Evidence for the ventral origin of oligodendrocyte precursors in the rat spinal cord. J Neurosci 1991; 11:2477-2488.

12

Astrocytes in Brain Aging and Neurodegeneration

46. Richardson WD, Pringle NP, Yu W-P et al. Origins and early development of oligodendrocytes. In: Jessen KR, Richardson WD, eds. Glial Cell Development, Basic Principles and Clinical Relevance. Oxford, UK: Bios Scientific Publishers, 1996:53-70. 47. Levison SW, McCarthy KD. Characterization and partial purification of AIM: a plasma protein that induces rat cerebral type 2 astroglia from bipotential glial progenitors. J Neurochem 1991; 57:782-794. 48. Hughes SM, Lillien LE, Raff MC et al. Ciliary neurotrophic factor induces type-2 astrocyte differentiation in culture. Nature 1988; 335:70-72. 49. Lillien LE, Sendtner M, Raff MC. Extracellular matrix-associated molecules collaborate with ciliary neurotrophic factor to induce type-2 astrocyte development. J Cell Biol 1990; 111:635-644. 50. Gard AL, Williams WC, Burrell MR. Oligodendroblasts distinguished from O-2A glial progenitors by surface phenotype (O4+/GalC-) and response to cytokines using signal transducer LIFRβ. Dev Biol 1995; 167:596-608. 51. Johe KK, Hazel TG, Muller T et al. Single factors direct the differentiation of stem cells from fetal and adult central nervous system Genes Dev 1996; 10:3129-3140. 52. Bonni A, Sun Y, Nadal-Vicens M et al. Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway. Science 1997; 278:477-483. 53. Kahn MA, Huang CJ, Caruso A et al. Ciliary neurotrophic factor activates JAK/Stat signal transduction cascade and induces transcriptional expression of glial fibrillary acidic protein in glial cells. J Neurochem 1997; 68:1413-1423. 54. Besnard F, Brenner M, Nakatani Y et al. Multiple interacting sites regulate astrocyte-specific transcription of the human gene for tglial fibrillary acidic protein. J Biol Chem 1991; 266:18877-18883. 55. Masood K, Besnard F, Su Y et al. Analysis of a segment of the human glial fibrillary acidic protein gene that directs astrocyte-specific transcription. J Neurochem 1993; 61:160-166. 56. Shafit-Zagardo B, Iwaki AK, Goldman, JE. Astrocytes regulate GFAP mRNA levels by cAMP and protein kinase C dependent mechanisms. Glia 1988; 1:346-354 57. Gross RE, Mehler MF, Mabie PC et al. Bone morphogenetic proteins promote astroglial lineage commitment by mammalian subventricular zone progenitor cells. Neuron 1996; 17:595-606. 58. Mabie P, Mehler MF, Marmur R et al. Bone morphogenetic proteins induce astroglial differentiation of oligodendroglial-astroglial progenitor cells. J Neurosci 1997; 117:4112-4120. 59. Noble M, Murray K, Stroobant P et al. Platelet-derived growth factor promotes division and inhibits premature differentiation of the oligodendrocyte/type 2 astrocyte progenitor cell. Nature 1988; 333:560-562. 60. Behar T, McMorris FA, Novotny EA, Barker JL, Dubois-Dalcq M. Growth and differentiation properties of O-2A progenitors purified from rat cerebral hemispheres. J Neurosci Res 1988; 21:168-180. 61. Barres BA, Raff M. Axonal control of oligodendrocyte development. In: Jessen KR, Richardson WD, eds. Glial Cell Development, Basic Principles and Clinical Relevance. Oxford, UK: Bios Scientific Publishers, 1996:71-83. 62. Pekny M, Leveen P, Pekna M et al. Mice lacking glial fibrillary acidic protein display astrocytes devoid of intermediate filaments but develop and reproduce normally. EMBO J 1995; 14:1590-1598. 63. Shibuki K, Gomi H, Chen L et al. Deficient cerebellar long term depression, impaired eyeblink conditioning, and normal motor coordination in glial fibrillary acidic protein mutant mice. Neuron 1996; 16:587-599. 64. Liedtke W, Edelmann W, Bieri PL et al. GFAP is necessary for the integrity of CNS white matter architecture and long-term maintenance of myelination. Neuron 1996; 17:607-615. 65. Caley DW, Maxwell DS. Development of the blood vessels and extracellular spaces during postnatal maturation of rat cerebral cortex. J Comp Neurol 1970; 138:31-48. 66. Phelps CH. The development of glio-vascular relationships in the rat spinal cord. Z Zellforsch 1972; 128:555-563.

Astrocyte Ontogenesis and Classification

13

67. Robertson PL, Du Bois M, Bowman PD et al. Angiogenesis in developing rat brain: an in vivo and in vitro study. Dev Brain Res 1985; 23:219-223. 68. Wise SP, Jones ED. Organization and postnatal development of the commissural projection of the rat somatic sensory cortex. J Comp Neurol 1976; 168:313-343. 69. Wise SP, Jones ED. Developmental studies of thalamocortical and commissural connections in the rat somatic sensory cortex. J Comp Neurol 1978; 178:187-208. 70. DeBault LE, Cancilla PA. g-Glutamyl transpeptidase in isolated brain endothelial cells: induction by glial cells in vitro. Science 1980; 207:653-655. 71. Janzer, RC, Raff MC. Astrocytes induce blood-brain barrier properties in endothelial cells. Nature 1987; 325:253-257. 72. Beck DW, Roberts RL, Olson JJ. Glial cells influence membrane-associated enzyme activity at the blood-brain barrier. Brain Res 1986; 381:131-137. 73. Schnitzer J. Astrocytes in the guinea pig, horse, and monkey retina: their occurrence coincides with the presence of blood vessels. Glia 1988; 1:74-89. 74. Kitamura T, Nakanishi K, Watanabe S et al. GFA-protein gene expression on the astroglia in cow and rat brains. Brain Res 1987; 423:189-195. 75. Wilkin GP, Marriott DR, Cholewinski AJ. Astrocyte heterogeneity. TINS 1990; 13:43-46. 76. Garcia-Abreu J, Neto VM, Carvelho SL et al. Regionally specific properties of midbrain glia: I. Interactions with midbrain neurons. J Neurosci Res 1995; 40:471-477. 77. Denis-Donini S, Glowinski J, Prochaintz A. Glial heterogeneity may define the three-dimensional shape of mouse mesencephalic dopaminergic neurons. Nature 1984; 307:641-643. 78. Amundson RH, Goderie SK, Kimelberg HK. Uptake of [3H] serotonin and [3H] glutamate by primary astrocyte cultures. II. Differences in cultures prepared from different brain regions. Glia 1992; 6:9-18. 79. McCarthy KD, Salm AK. Pharmacologically distinct subsets of astroglia can be identified by their calcium response to neuroligands. Neuroscience 1991; 41:325-333. 80. Miller RH, Szigeti V. Clonal analysis of astrocyte diversity in neonatal rat spinal cord cultures. Development 1991; 113:353-362. 81. Miyake T, Fujiwara T, Fukunaga T et al. Glial cell lineage in vivo in the mouse cerebellum. Develop Growth Differ 1995; 37:273-285. 82. Zhang L, Goldman JE. Developmental fates and migratory pathways of dividing progenitors in the postnatal rat cerebellum. J Comp Neurol 1996; 370:536-550. 83. Eddleston M, Mucke L. Molecular profile of reactive astrocytes—implications for their role in neurologic disease. Neuroscience 1993; 54:15-36. 84. Norton WT, Aquino DA, Hozumi I et al. Quantitative aspects of reactive gliosis: a review. Neurochem Res 1992; 17:877-885. 85. Altman J Autoradiographic investigation of cell proliferation in the brains of rats and cats. Anat Rec 1963; 145:573-591. 86. Kaplan MS, Hinds JW. Gliogenesis of astrocytes and oligodendrocytes in the neocortical gray and white matter of the adult rat: electron microscopic analysis of light radioautographs. J Comp Neurol 1980; 193:711-727. 87. McCarthy GF, Leblond CP. Radoautographic evidence for slow astrocyte turnover and modest oligodendrocyte production in the corpus callosum of adult mice perfused with 3H-thymidine. J Comp Neurol 1988; 2761:589-603. 88. Ling EA, Leblond CP. 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 1973; 149:73-82. 89. Gensert JM, Goldman JE. In vivo characterization of proliferating cells in adult rat subcortical white matter. Glia 1996; 17:39-51. 90. Gensert JM, Goldman JE. Remyelination by endogenous progenitors in the adult rat CNS. Neuron 1997; 19:197-203. 91. ffrench-Constant C, Raff MC. Proliferating bipotential glial progenitor cells in adult optic nerve. Nature 1986; 319:499-502. 92. Wren D, Wolswijk G, Noble M. In vitro analysis of origin and maintenance of O-2A adult progenitor cells. J Cell Biol 1992; 116:167-176.

CHAPTER 2

Functions of Astrocytes Harold K. Kimelberg and Michael Aschner

Introduction

I

n the previous chapter, Goldman covered the structure and development of astrocytes, and that chapter should be read before reading this chapter to better understand the functional properties we will discuss. Thus, an appreciation of the complex morphologies of all types of astrocytes, their interrelationships with other cells and brain structures such as blood vessels, and the complexity of astrocyte development must surely reasonably lead, based on the principle that form reflects function, to the conclusion that astrocytes are likely to have many complex properties closely associated with many aspects of brain function. Indeed, there has been no dearth of hypotheses regarding astrocyte function emerging simply from contemplation of the complexities of astrocyte morphology and interrelationships dating from the work of Golgi, Cajal and others,1 which first showed their structures in precise detail in the late nineteenth century. For example, per Lugaro2 in 1907; “the neuronal articulation* would be the center of the chemical exchange, and this would comprise therefore in all the most proximal, vacant interstitial spaces, a region for infiltration of the protoplasmic prolongations or feathery extensions of the neuroglia, perhaps with the purpose of collecting and instantly processing the smallest amount of waste product.” Golgi and Cajal among others speculated that the roles of glia included neuronal nutrition, structural and metabolic support and involvement involved in nervous system development.1 However, these and other hypotheses could not then be tested. Experimental studies on glial function began with the work of Kuffler and his colleagues in the mid 1960s. They focused on the electrical and ion transport properties of glia in simple invertebrate nervous systems and the relatively simple preparation of the amphibian optic nerve.3 Beginning in the 1970s primary astrocyte cultures from neonatal rodent CNS began to be used extensively to study the properties of astroglia.4,5 The primary cultures prepared from neonatal rodents consist predominately of GFAPpositive astrocytes and provide preparations of cells in sufficient numbers to allow for a variety of biochemical, electrophysiological, molecular and general cell biological studies. It is still unclear as to why all the cells in these astrocyte cultures, which consist primarily of flat cells which have been proposed to be analogous to protoplasmic astrocytes, stain for GFAP whereas protoplasmic astrocytes in situ in many regions, such as the cerebral cortex, stain variably for GFAP.6 This has led to the view that the cultures may consist predominantly of reactive astrocytes, which in situ are characterized by prominent GFAP staining.7 * synapse Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

16

Astrocytes in Brain Aging and Neurodegeneration

It needs to be emphasized that the bulk of the current information on the properties of mammalian astrocytes has come from these preparations, as they are relatively easily studied. However, their properties are often imprecisely referred to as astrocytic properties, without any qualifications. Studies on primary astrocyte cultures have always had the implicit caveat that one is uncertain as to how their properties are altered by growth in vitro.8 In our view, such cultures have had two major and critical advantages. One, they led to a major expansion of studies on astrocytes, albeit mainly in these culture systems, which otherwise would probably not have been done in any other system. Second, the results of such studies suggested to neuroscientists that astrocytes and other glial cells could have a number of properties such as receptors and uptake systems for transmitters and gated and rectifying channels which, based on a few negative studies on glial cells in situ had, rather prematurely, been considered specific to neurons in the CNS. The primary cultures have now provided a long list of putative functions which need to be tested in systems more representative of the in vivo state, for the numerous differences that have now been reported between the properties of the primary astrocyte cultures and the properties of astrocytes expressed in preparations closer to the in situ state, such as brain slices, makes it impossible to use primary cultures by themselves to define astrocyte function. Thus there is currently less emphasis on such cultures and a reemphasis on in situ preparations that appear to more closely correspond to in vivo situations. The relative paucity of reliable information is presumably why astrocytes so rarely figure in discussions of brain function. In contrast, the characteristic properties of neurons have been studied in great detail and were found to lend themselves relatively easily to hypotheses of information processing by the postulation of electrically active loops and networks. Also, experimental interference with neuronal function led to clear effects on neural function, so that discussions of how the nervous system functions at the most complex levels are currently almost exclusively based on the properties of neurons.9 Thus a certain circularity of reasoning is apparent that can only be broken by sufficient rigorous study of the properties that astrocytes and other glia have in the CNS. In many respects we are still searching for experimental systems in which hypotheses advanced at the turn of this century can be rigorously tested.

Functions of Astrocytes We will arrange this section based first on the properties of astrocytes established both in cultured and acutely isolated cells or in slices of the intact brain, taking care to note the experimental systems in which they were obtained. We will also discuss potential functions that these properties indicate, and then mention the very limited number of cases in which functions have actually been demonstrated. It must be born in mind, however, that the properties studied are not only limited by the experimental systems, but have been conceptually restricted, notably by concepts based on what has been developed for neurons which have led to success in understanding nervous function. Thus, it is not surprising that many studies of astrocytes have involved investigations of their membrane electrical properties and ion channels, even though this approach has led a well-respected worker in the field of glial electrophysiology to conclude that “generation of glial electric signals is not among their (i.e., astrocytes) functions.”10 This then begs the question: What is (are) the role(s) of the predominant K+ channels seen in astrocytes and what is (are) the role(s) of the -70 to -80 mV membrane potentials that are a consequence of them?

Functions of Astrocytes

17

Homeostasis of the Extracellular Space Regulation of Extracellular K+ One answer to the question just posed was proposed by Kuffler et al.3 Namely, that the selective K+ permeability implied a role in control of extracellular potassium levels ([K+]o). The original reason for Kuffler and his colleagues embarking on their pioneering studies was that electron microscopic studies of mammalian CNS had shown, at this time, that astrocytes generally formed enlarged watery compartments seemingly obliterating the extracellular space (ECS). This led to the proposal that astrocytes actually formed the extracellular space of the brain. This would require that they would uniquely be high Na+ cells, and it was to examine this question that Kuffler and his associates studied the easily accessible glial cells in the leech nervous system and the amphibian optic nerve. In both cases, glial cells were found to have a membrane potential of -80 to -90 mV, and showed a close to Nernstian response to varying [K+]o. Thus these glial cells had to have high intracellular K+ rather than Na+ concentrations, and thus could not form the ECS. The finding of an essentially selective K+-dependent membrane potential implied that the cell membranes were operationally impermeable to sodium, and possibly chloride, and led to a mechanism for uptake of K+ released by active neurons.3 The mechanism would be that a localized release of K+ from neurons during excitation would depolarize the astrocyte at this point with a 60 mV depolarization for a 10-fold increase in [K+]o. This would set up a current loop with other nondepolarized parts of the cell and, since the membrane was permeable only to K+, there would be an inward current at the depolarized point carried by extracellular K+ crossing the membrane. Since K+ is the major electrolyte inside the cell, it would also be the major current carrier inside the cell, and the current loop would be connected by efflux of K+ at some distant point. The return part of the loop would be carried by major extracellular ions such as Na+, or Cl– in the opposite direction. This led to the concept of “K+ spatial buffering” in which K+ is transferred from a region of localized release to some distant point, traveling through the astrocyte or the astrocytic syncytium. Work since the studies of Kuffler and his colleagues has concentrated on identifying the types and location of K+ channels in different glial preparations using modern patch clamp methods. K+ channels are the most diverse ionic channel type11 and, as reviewed over the past several years, a wide variety of K+ channels have been found, predominantly using cultured astrocytes.12-15 These K+ channels include an inward rectifying K+ channel (K+in), Ca2+-dependent K+ channels (K+Ca), delayed rectifying channels (K+D) and an inactivating potassium channel (K+a). K+ channels sensitive to ATP have also been found in astrocytes, such as an ATP-regulated, strongly inward rectifying K+ channel that has been observed on Bergmann glia in situ.16 Some of these channels may be related to the K+ spatial buffering phenomenon just discussed. When there is also a significant chloride permeability (see later), net KCl uptake leading to swelling will occur when [K+]o rises. Some of these K+ channels should also be responsible for the large negative K+ diffusion potentials characteristic of astrocytes. The work of Newman17 using acutely isolated astrocytes has indicated inward K+ rectifying channels at very high densities in areas of the astrocyte where it seems to be adapted to K+ spatial buffering, namely at the capillary-facing astrocytic end-feet. If the membrane potential is very close to the K+ equilibrium potential, then the net outward leak of K+ will always be very low, but this may be increased when there is depolarization of the astrocyte caused by other than an increased [K+]o, such as by receptor activation. In this case, there will be an outward flux of potassium which would be later replenished by reuptake on the Na+/K+ pump or repolarization and reestablishment of Em ≅ Ek (see later). If K+ channels are important in astrocyte function, then it is likely that alterations in their functioning would affect astrocyte properties and they are likely to be targets of the

18

Astrocytes in Brain Aging and Neurodegeneration

activation of astrocyte receptors. This is currently an active and fruitful area of investigation. Thus, it has been shown that β-receptor activation modulates K+IR currents in cultured rat spinal cord astrocytes,18 as well as altering astrocyte proliferation in vitro.15 AMPA/kainate receptor activation blocks outward K+ currents in cultured stellate mouse cortical astrocytes.19 This was suggested as a mechanism whereby astrocytes do not lose too much K+ when they become depolarized in pathological states. The K+ currents of glial cells in situ in hippocampal brain slices have been studied from different aged animals in both nonexcitable, GFAP-negative “complex cells” from younger animals and in GFAP-positive cells from older animals (>P20). The complex cells exhibited more types of ion currents. They showed a delayed outward K+ rectifier (K+D) and a transient outward A-type K+ current. They also showed a TTX-sensitive Na+ current. In the older cells, the voltage-gated Na+ and K+ outward currents downregulated and were replaced by passive and inward rectifier K+ conductances.20 These changes are consistent with a precursor glial cell with a more complex array of ion channels changing into a mature astrocyte which exhibits K+ channels that have predominantly [K+]o regulating properties. It has also been shown that a variety of K+ channel blockers inhibit cell proliferation in cultured astrocytes.15 Recent work has also shown that application of cesium for >2 min to hippocampal slices blocks long term depression (LTD) and synchronous, interictal-like bursting in the CA1 region.21 Studies using patch-clamp electrophysiology showed this to be due to a direct blockade of the K+IR currents of astrocytes. The increase in [K+]o was considered to block the pyramidal cell activity since there was no change in the pyramidal cell conductance. This experiment is reminiscent of the 30 year old study of Krnjevic and Schwartz22 wherein they attempted to detect transmitter-induced conductance changes in glial cells from the cerebral cortex using sharp electrodes. They found no such changes, possibly due to the insensitivity of their techniques, and concluded that depolarization of the glial cells was due to a rise in [K+]o rather than a transmitter-mediated conductance change in the astrocyte

Na+ Channels This is a more controversial area because if astrocytes are nonexcitable there would appear to be no need for Na+ channels, or at least voltage-sensitive ones. Na+ currents in glia were first described in astrocytes in primary cultures.23,24 Like neuronal channels, these were sensitive to tetrodotoxin (TTX), but it was found that there were both TTX-sensitive and relatively insensitive Na+ channels which had different characteristics in terms of the depolarization required to activate them.25 The depolarizations needed to open these channels were always thought to be greater than would ever be seen in astrocytes “clamped” at a highly negative membrane potential by their large K+ conductances.5 However, recent work by Sontheimer et al26 has found that astrocytes cultured from certain regions of the brain, such as the spinal cord, have a very high density of Na+ channels which would have some open probability at the resting membrane potential of these cells. It was hypothesized that the Na+ channels may function in regulating entry of Na+ in order to activate the Na+/K+ pump when active uptake K+ is required, such as when [K+]o rises from its normal level of 3 mM to 5-10 mM during periods of sustained neuronal activity. This thus represents a self-regulating mechanism for active K+ clearance by astrocytes that does not require any special properties of the Na+/K+ pump, and such special properties have not been clearly shown (see below). The major question in regard to the Na+ channels, as with other astrocytic properties mainly described in astrocytes in culture, is whether, when and in what cells Na+ channels are expressed in situ. The type II sodium channel has been seen in astrocytes in situ in the dorsal and ventral columns of the spinal cord of the adult rat using immunocytochemistry,

Functions of Astrocytes

19

but can only assumed to be functional.27 The function(s) of these channels must at present be purely speculative. Do they provide a voltage-dependent path for Na+ entry? There are, however, several other routes for Na+ entry into astrocytes. There are cotransporters for glutamate and aspartate and other substances which utilize the energy of the Na+ gradient to actively accumulate these substances. Are the channels a source of Na+ channels for the axolemmal membrane of the node, as suggested by Ritchie?28 Do they in some way add to the ability of astrocytes to control the ionic composition of the ECS? However, these channels are only activated when the cell membrane is depolarized to at least -40 mV and the question still remains as to what extent such depolarizations do occur in astrocytes in situ.

Ca2+ Channels Voltage-gated L-type Ca2+ channels were also first identified in primary astrocyte cultures.29 This was again a surprising finding because it was thought that such channels in the CNS were specific to neurons and were responsible for such properties as Ca2+ action potentials, depolarization-induced Ca2+ influx required for exocytosis at nerve terminals and modulation of neuronal firing rates by hyperpolarization of the membrane potential via Ca2+-activated K+ channels. The question raised in regard to voltage-sensitive Na+ channels can also be raised in regard to voltage-sensitive Ca2+ channels; namely, are the astrocytes ever depolarized enough to activate these channels? Thus the function of the Ca2+ channels in astrocytes has not yet been satisfactorily explained, but the occurrence of large changes in [Ca2+]i levels in astrocytes when stimulated by receptors such as glutamate, mechanical stimulation or swelling raise the possibility that part of the intracellular Ca2+ may be entering via such channels.30 Ca2+ channels will also be necessary for the entry of Ca2+ to replenish intracellular Ca2+ stores after their depletion.30 Intracellular Ca2+ is a pleiotropic intracellular second messenger affecting processes from gene regulation to ion channel activities. The reader is referred to several reviews on this topic in astrocytes.30-35 In this context, the changes in [Ca2+]i levels can be viewed as a ubiquitous intracellular signaling mechanism present in astrocytes, as in other cells, rather than subserving a specific CNS function such as regulated release of neurotransmitters at CNS synapses.

Anion Channels As in other cells, anion channels have been less studied in astrocytes than cation channels, but it is now becoming clear that anion channels do have fundamental functions in many, if not all, cells. A number of anion channels have now been identified in cultured astrocytes, including small conductance chloride channels (5-25 pS) and a high conductance chloride channel (250-400 pS).14,15 These channels also transport Cl– and HCO3–, but the high conductance channel in astrocytes may also transport organic anions such as amino acids,35 as may some of the other channels.36 The roles of anion channels may also include the uptake of HCO3– or chloride to accompany uptake of K+; a mechanism additional to K+ spatial buffering (see above) to control [K+]o. However, this will lead to cell swelling, and many pathological states involve exaggerated astrocyte swelling.37 Cell swelling causes activation of a number of ion channels and Strange et al38 and Okada39 have recently reviewed the different anion channels that may be involved in anion or amino acid efflux in swollen cells, including astrocytes. It is important to define the types of different anion channels normally present on astrocytes, and which ones are activated during swelling or are responsible for K+-dependent Donnan swelling, because, if release of excitatory amino acids in ischemia and other pathological states occurs through a particular type of anion channel, then the identification of such channels would be of considerable practical benefit to either inhibit the swelling-induced release or prevent the swelling in the first place. The inhibitory neurotransmitters GABA

20

Astrocytes in Brain Aging and Neurodegeneration

and glycine also activate anion channels associated with some of their receptors, and such receptors appear to be present on astrocytes40-43 (see later).

Ion Carriers Carriers are distinct from channels in that a concerted movement of several ions usually occurs, rather than the independent diffusional movement of ions down their electrochemical gradients as in channels.11 A number of carriers have been identified in astrocytes. Na++K++2Cl– cotransporter One important ion carrier is the Na++K++2Cl– uptake system utilized by cells for active uptake of Cl–. This carrier has been found in astrocytes in primary culture, and intracellular Cl– in astrocyte cultures has been found to be several-fold greater than expected from electrochemical equilibrium.14 This carrier has also been localized by immunocytochemistry to Bergmann glia in situ.44 High [Cl–] in astrocytes may serve as a source to maintain extracellular Cl–, based on the finding of GABAA receptors on astrocytes both in vitro and in situ whose activation leads to efflux of Cl–.45 This was proposed as a mechanism to maintain extracellular concentrations at the same time as GABA causes influx of Cl– into neurons, which contain low Cl–. In addition, the high Cl– in astrocytes may also be required for the efflux of KCl in the process of volume regulation, as proposed by Kimelberg and Frangakis.46 Measurements with ion-specific microelectrodes in guinea-pig brain slices, however, have shown intracellular glial Cl– levels that were in equilibrium with the membrane potential and therefore <10 mM.47 However, a recent microprobe analysis has shown that the intracellular Cl– levels in CNS glia are around 25 mM.48 (Na+/K+) pump As do all animal cells, astrocytes possess an active Na+/K+ pump responsible for accumulating K+ and pumping out Na+. Na+/K+ pumps consist of isoforms of α and β subunits. The work of Sweadner and colleagues49,50 has shown that the α1, α2, and α3 forms are distributed in a complex manner among different cells of the CNS. It appears that neurons can exhibit all three isoforms, either individually or in various combinations, and astrocytes and other glia cells express α1, α2 or both, but not α3. The α1 mRNA has a broad distribution in brain, whereas the α2 mRNA is much more localized. While the β subunit is the same as an adhesion molecule on glia (AMOG), the β isoform is not specific to glia.51 These findings raise interesting questions regarding the relationship between ion transport and cell adhesion. In terms of the kinetics of the different isoforms, there is evidence both for and against a specialized role of glial Na+/K+ ATPase in uptake of K+ by astrocytes.50 As with other Na+ pumps this system seems to be driven mainly by intracellular [Na+]i. It has a high affinity for K+ on the outside (Km ~ 1 mM), and a mid-activation level for Na+ of about 10 mM on the inside. Thus, with a [K+]o of around 3 mM there will only be a small amount of activation between 3 and 10 mM K+, the saturation level for K+ activation of the Na+/K+ pump, while with a Km for Na+i of ~ 10 mM and [Na+]i ≅ 10 mM14 the pump is poised to be maximally activated by increases or decreases in [Na+]i from its normal levels. Sontheimer et al52,25 found that some astrocytes have a high density of TTX-sensitive Na+ channels, and suggested that these channels are responsible for maintaining the intracellular Na+ levels required for the functioning of the Na+/K+ pump. Since these Na+ channels are voltageactivated, increased [K+]o would depolarize the astrocyte membrane potential and regulate the influx of Na+ and thereby the Na+/K+ pump via increases in [Na+]i. This is an interesting suggestion, since it would control active K+ uptake by astrocytes without requiring any specialization of the astrocytic Na+/K+ pump.

Functions of Astrocytes

21

pH carriers Other carrier systems for Cl– or Na+ involve co- or exchange transport with the pH equivalents H+, HCO3– or OH–. These systems are the Na+/H+ and Cl–/HCO3– or OH– exchangers, and a variety of electrogenic or nonelectrogenic cotransport systems for Na+ plus nHCO3–, where n can be two to three.53-55 It has been suggested, based on the possession of these transport systems and the fact that astrocytes in situ can undergo very large pH changes in ischemia (often in the opposite direction to the extracellular pH), that astrocytes are critically important in maintaining pH homeostasis in the brain.56-58 In ischemia, the astrocytes become very acidic, whereas in spreading depression the astrocytes undergo large intracellular alkaline shifts.56 The operation of such pH transporting systems could also lead to volume changes. For example, the simultaneous operation of the Na+/H+ and Cl–/HCO3– exchangers driven by intracellular hydration of CO2 to H+ and HCO3– could lead to a net uptake of Na+ and Cl– with astrocyte swelling. Such swelling is seen, for example, in trauma or ischemia, when it is likely that the Na+/K+ pump will be compromised due to falling energy levels and therefore pump out the Na+ more slowly.59,38 Such swelling has been reproduced in vitro under such conditions using primary astrocyte cultures or C6 glioma cell lines.60

Transmitter Uptake Systems Amino Acid Carriers and Cellular Metabolic Compartmentation It has long been known that there are very powerful uptake systems on astrocytes for a number of amino acid transmitters, particularly the excitatory amino acids (EAA) glutamate and aspartate. It was proposed 25 years ago that glutamate released from terminals was taken up into astrocytes where it is converted to glutamine by the astrocyte-specific enzyme glutamine synthetase.61 This was perhaps the first example of compartmentation of metabolic functions between neurons and glia. The general concept of metabolic compartmentation between astrocytes and neurons has gained increasing experimental support with the finding that key metabolic enzymes such as pyruvate carboxylase, specific isoforms of lactic dehydrogenase and malic enzyme are specific to astrocytes.62,63 Such compartmentation has profound implications for brain chemistry. It is not clear whether the uptake of glutamate, seen in both cultured and acutely isolated astrocytes and by autoradiography and immunocytochemistry on astrocytes in situ,64,65 influences synaptic transmission. Uptake of the EAAs is likely to be slower than needed to alter synaptic transmission and therefore is more likely to serve for long term control. However, it is now thought that binding to the high density of both neuronal and astrocytic transporters contributes to the decay of the EPSP, as diffusion of glutamate is considered to be too slow.66 Since glutamine synthetase is astrocyte-specific, it is also intriguing for the functional implications of the transporters that administration of a specific inhibitor of this enzyme just prior to a learning task in day-old chicks caused significant retention loss.67 Recent work has shown that the EAA transporters also transport K+ and OH–,68 and reversal of this transporter by ischemia-induced changes in cellular ion gradients has been suggested to be responsible for some of the increased EAAs seen in ischemia.69,70 Increased glutamate content in astrocytes and decreased neuronal content during ischemia have been shown by quantitative immunocytochemistry at the electron microscope level.71 This increase in steady state glutamate levels could be due to a decreased conversion to glutamine within astrocytes, leading to a large increase in glutamate which more than offsets the increased release. Recent molecular biology studies have identified at least three members of a family of glutamate transporters of which two (GLT-1 and GLAST in the rat) are found exclusively or

22

Astrocytes in Brain Aging and Neurodegeneration

predominantly on astrocytes.72-75 The GLT-1 is the major form in rat brain. GLT-1 and GLAST coexist on the same astrocyte membranes, but do not form complexes with each other. Rather they form homo-oligomers.76 In an interesting study, Rothstein et al77 have shown that separate intraventricular administration of antisense oligonucleotides to the three different transporters led after 7-10 days to a marked reduction in all the transporter proteins. However, only antisense to the astrocyte-specific transporters GLT-1 and GLAST led to an increase in [Glu]o, as measured by microdialysis. The same group has also implicated a failure of the astrocyte transporters in the etiology of amyotrophic lateral sclerosis.78 There is also good evidence for GABA transporters on astrocytes. As with the EAA transporters, different isoforms have been found, termed GAT-1, GAT-2 and GAT-3. Using cDNA probes for in situ hybridization, it was found that GAT-1 was predominantly neuronal, while GAT-3 was also seen over glial profiles. GAT-2 was mainly seen in the meninges, and is also found outside the CNS.79 However, in the cerebellum GAT-1 is found as much in the Bergmann glia surrounding the Purkinje cells as in the Purkinje cell bodies themselves.80 It was suggested that the GAT-1 localized to the basal pole and axon hillock of the Purkinje cells are primarily involved in terminating the action of GABA at basket synapses on the Purkinje axon hillocks, a role also presumably subserved by the ensheathing GAT-1-IR positive glial processes as GABAergic cerebellar neurons did not show GAT-1-IR. An amino acid transporter very active on astrocytes is that for the amino sulphonic acid taurine. Taurine occurs at high levels in many brain regions, and is considered to be an inhibitory transmitter and to also play a role in limiting cell swelling, especially in astrocytes, by its release via swelling-activated organic anion channels.37,81

Monoamine Transporters Uptake of a number of monoamine transmitters has been reported in primary astrocyte cultures.14,82 These systems resemble the high-affinity systems found in nerve terminals, being both Na+ -dependent and inhibitable by a variety of clinically relevant antidepressants, such as fluoxetine (Prozac) for serotonin. Uptake systems for adenosine83 and histamine84 have also been described in cultured astrocytes. Unlike the amino acid transport systems, the relevance of the monoamine uptake systems has not yet been established and may play a minor role, if it plays any role at all or exists on astrocytes in vivo, compared to the very active monoamine transporters in nerve terminals. Recently, a direct comparison of pure monolayer astrocyte cultures and explant cultures from different brain regions has shown that while the high affinity norepinephrine and serotonin transporters were indeed present in the pure astrocyte cultures, such uptake into astrocytes could not be detected in the explant cultures, which also contain neurons, under identical conditions.85 As an appropriate control, the uptake of radiolabeled glutamate or GABA into astrocytes was found to be comparable in the pure astrocyte and the mixed explant cultures. This is another example of how neurons can influence astrocyte properties and how homogeneous astrocyte cultures can give idiosyncratic results. In this case it appears that neurons suppress an important function that astrocytes might otherwise express as a default property. It would be of considerable interest to see whether this effect has any correlate in vivo, such as in reactive astrogliosis which occurs as a response to neuronal damage and death.

Receptors for Transmitters Receptors in Primary Astrocyte Cultures A large number of neurotransmitter receptors have been found in monolayer primary astrocyte cultures prepared from different brain regions of neonatal rats. These include adrenergic (α1, α2, β1, β2), aminergic (5HT1, 5HT2, M1, M2, H1, H2), amino acid (mGluR,

Functions of Astrocytes

23

KA/AMPA, GABAA, GABAB), peptide (AT II, somatostatin, endothelins, bradykinins, substance P, ANP, neuropeptide Y, VIP, opioid), and purinergic (P2x, P2y, P2u, A1, A2) receptors.86 They have been identified by a variety of techniques, including stimulation of second messengers (cAMP, IP3 or [Ca2+]i), electrophysiological studies, antibody staining, radio-ligand binding studies, and most recently in situ hybridization and RT-PCR for receptor mRNA. In terms of regional specialization the receptors found so far in primary astrocyte cultures from the hippocampus which could be relevant to function in this region include glutamate (mGluR, KA/AMPA),34,87,88 5HT2A89 and adenosine90 receptors. Adrenergic (α1, β) receptors have also been found in mixed neuron-astroglial hippocampal cultures.91 Most of the work on receptor expression in astrocytes so far has been done in vitro. However, it is now clear from numerous studies that many of the properties of these cultures, including receptor expression, change with different culture conditions.55,92-99 For example, Shao and McCarthy99 demonstrated that cortical astroglia tended to lose their responsiveness to carbachol and histamine and to develop responsiveness to NE with growth in serum (FBS)-containing medium, even though they were cultured from single cells. Miller et al97 showed an upregulation of a phosphoinositide-coupled mGluR in cortical astrocytes in chemically-defined medium as compared to serum (FCS)-containing medium. This effect was found to be due to the growth factors bFGF and EGF in chemically-defined medium. They also found that exposure to thrombin reduced mGluR5 level in astrocyte cultures,100 which might be the reason for the decrease in mGluR-mediated Ca2+ responses seen in astrocytes cultured in horse serum-containing medium.101 One of our laboratories also recently showed that the proportion of cells responding to 5-HT increased when acutely isolated astrocytes were cultured in horse serum-containing medium,101,102 but did not increase in serum-free, chemically defined medium.95 These findings continue to raise questions regarding which receptors seen in primary culture exist in astrocytes in vivo, or are upregulated or selected for in culture.

Receptor Studies in Astrocytes in Brain Slices Within the past few years, attempts have been made to study receptors on astrocytes in situ. Studies in brain slices (mainly from the hippocampus) have shown that astrocytes in situ do respond to applied neurotransmitters. Techniques used to measure these responses have been mainly electrophysiological, or calcium imaging with confocal microscopy. The receptors found so far in astrocytes have been GABAA receptors in hippocampal slices from P21-42 rats,103,104 glutamate (mGluR, KA/AMPA, NMDA) receptors in hippocampal slices from P9-13 rats105 and P9-12 mice,43 and P1 and less frequently P2 purinergic receptors in astrocytes in hippocampal slices from P9-13 rats.106 P2 receptors have also been found in Bergmann glia in cerebellar slices from P6-45 day old mice107 and α1-adrenoreceptors and H1 histamine receptors in Bergmann glial cells in cerebellar slices from P20-25 mice.108 However, slice studies do present several difficulties. An important one is secondary effects on astrocytes due to the release of neurotransmitters or K+ from neurons stimulated by the applied transmitters. TTX used in slice studies can only block action potential-induced terminal release of neurotransmitters and not transmitter release induced by TTXinsensitive action potentials or extrasynaptic release from dendrites or axons. The calcium responses to iGluR agonists (KA, AMPA, NMDA) found in astrocytes in hippocampal slices, but not in our acutely isolated hippocampal astrocytes, could be due to the depolarization of astrocytes by K+ released from excited neurons, which then activate voltage-activated Ca2+ channels leading to [Ca2+]i increases in astrocytes. For activation of the AMPA/KA receptors to directly lead to an increase in [Ca2+]i they need to lack the GluR2 subunit that results in Ca2+ permeability of the AMPA receptors. This has been reported in Bergmann glia in mouse cerebellar slices.109

24

Astrocytes in Brain Aging and Neurodegeneration

Porter and McCarthy105 showed that GFAP-positive astrocytes in rat hippocampal slices responded to application of NMDA-induced with an increase in [Ca2+]i. However, this might be due to a direct activation of NMDA receptors on neurons, leading to release of glutamate, K+ or some other substance which then activates the astrocytes via some different mechanism, and not indicative that the astrocytes themselves have NMDA receptors. Porter and McCarthy106 also showed that most astrocytes in hippocampal slices responded to ATP, but this was mediated by adenosine receptors (P1) and not ATP receptors. The P1 receptors were presumably activated by adenosine produced by hydrolysis of the added ATP to adenosine by extracellular ectonucleotidases within the slice. This is another kind of indirect effect which could have led to the erroneous conclusion that most hippocampal astrocytes express ATP receptors. However, a P2-induced Ca2+ transient was found in Bergmann glia in cerebellar slices from P6-45 mice,107 and this may imply a regional or species variation of P2 receptor expression in astrocytes. Another problem with slices is slowed access to applied transmitters due to long diffusion pathways and/or uptake, which can result in an insufficiently high concentration of perfused transmitters at receptor sites within the slices. Slowed access is a particular problem for rapidly desensitizing receptors such as the GluR receptors.110,111 Thus, cells close to the center of slices are less likely to respond to glutamate than those close to the edge of the slice, and the glutamate uptake inhibitor THBA did increase the percentage of cells responding to glutamate.105

Immunohistochemical Studies on Astrocytes In Vivo Immunohistochemistry in brain sections shows the localization of receptors to astrocytes with perhaps the least involvement of confounding variables. However, such studies do not indicate whether the receptors are functional and what functions they subserve. Aoki and colleagues112,113 showed astrocytic localization of α2A and β-adrenergic and NMDA1 receptors to astrocytic profiles in visual cortex using electron microscopy with immunocytochemistry. 5HT1A,114 mGluR5,115 mGluR3,116 AMPA receptor subunits GluR1 and GluR4117 and muscarinic receptor118 immunoreactivities have also been localized to astrocytes in rat hippocampus. Paspalas and Papadopoulos119 have recently reported that fine norepinephrine-containing nerve terminals ended on astrocytes around capillaries in rat visual cortex, as well as directly on the basal lamina. However, no plasma membrane differentiation at these sites on the astrocyte membranes was detected. In situ hybridization is now beginning to be used to localize receptor mRNA in astrocytes in intact brain. mRNAs for kainate receptors have been detected in astrocytes in various brain regions120 and mRNA for the NMDA2B subunit has been localized over Bergmann glia cells.121 Recent studies have also shown mGluR subtype mRNA expression in neurons in the hippocampus, but strong labeling of astrocytes was only shown for mGluR3, predominantly in the CA1 region.122,123 Light antibody staining for mGluR5 has been observed in astrocytes in hippocampal sections.115

Receptor Studies in Acutely Isolated Astrocytes Alternative preparations that also seem to more faithfully reflect in situ properties are astrocytes studied as soon as practical after isolation from the CNS, as has been done for neurons.124 This approach, although subjecting cells to some degree of rough handling, has clear experimental advantages, principally in avoiding the problem of indirect effects and slow access, as well as generally easier experimental techniques. More controlled experiments are possible but there is, of course, no possibility of studying interactions with neurons and other cells in the brain. Also, in cells from older animals, cell processes seem to be

Functions of Astrocytes

25

lost (unpublished observations); this is a serious problem if there are differences in receptor distribution within the cell. Acutely isolated astrocytes were first used to study voltage-gated ion channels in astrocytes,12,13,125 and there has been less work on receptors. Fraser et al103 identified GABAA/benzodiazepine receptors in acutely isolated hippocampal astrocytes from 2-6 week old rats by a combination of whole-cell patch clamp, calcium imaging, immunocytochemistry and fluorescent ligand binding techniques. Duffy and MacVicar126 found that <5% of hippocampal astrocytes acutely isolated from P21-42 rats showed α1 adrenergic-mediated calcium responses, although almost all the astrocytes in slices responded to norepinephrine. Glutamate did not increase [Ca2+]i in both acutely isolated astrocytes and astrocytes in slices. Seifert and Steinhauser111 applied patch-clamp technique and single-cell RT-PCR to glial precursor cells acutely isolated from the juvenile mouse hippocampal CA1 stratum radiatum subregion and found responses due to activation of GluR2 and GluR4 AMPA receptor subunits. Kimelberg et al101 studied receptor mediated Ca2+ responses of astrocytes acutely isolated from cerebral cortex to glutamate, 5-HT and ATP, and compared these with primary astrocyte cultures from cortex. It was found that GFAP-positive astrocytes acutely isolated from the cerebral cortices of postnatal 3-10 day old rats frequently showed increased intracellular [Ca2+] responses to L-glutamate. In contrast, responses to 10 µM ATP or 10 µM 5-HT were much less frequent or absent, respectively. The same cells that failed to respond to ATP or 5-HT often responded to glutamate. Culturing acutely isolated cells in media containing 10% horse serum decreased the percentage of GFAP-positive cells responsive to glutamate, but greatly increased the percentage that were responsive to ATP and 5-HT. In primary cultures prepared from the cerebral cortices of 1 day old rats and cultured in serum-containing medium for 2-4 weeks, fewer cells responded to glutamate than in acutely isolated cells, but almost all cells responded to ATP and 5-HT. The lack of response to ATP and 5-HT in the acutely isolated cells seemed unlikely to be due to selective damage to the respective receptors during enzymatic dissociation because acutely isolated GFAP-negative cells in the same preparations showed responses to ATP, several different proteases and mechanical dissociation yielded cells which also responded to Glu but not ATP, and exposure of primary cultures to papain, the enzyme used during isolation, did not abolish Ca2+ responses to several transmitters. We have now observed a very similar profile of responses for GFAP-positive astrocytes acutely isolated from the hippocampus.102 Thus, some of the receptor responses seen in primary astrocyte cultures may not reflect receptors present in astrocytes in vivo, but are rather upregulated or selected for in response to culture conditions. Of course, acutely isolated cells as models to study astrocytes have problems. First is the low yield of cells compared to primary cultures, so that the responses of individual or small groups of cells only can be measured. Thus, some aspects of astroglial functions cannot be easily studied (e.g., transmitter uptake and release). Second, and more seriously, there is the potential possibility of selective proteolytic damage to receptors, or the shearing off of processes. The latter appears to be more of a problem in older animals of >15 days. Thus, damage during isolation always needs to be ruled out when we find that these cells respond to one transmitter but not another.101

Functional Implications It is of course of great interest to ask what effects receptor activation might have on astrocyte properties. These will, of course, initially be the activation of second messengers which can then lead to a variety of functional effects, which in the case of astrocytes are still largely unknown. The activation of the KA/AMPA glutamate receptor in cultured astrocytes elicits membrane potential depolarization and Na+ and K+ inward currents.127-129 In

26

Astrocytes in Brain Aging and Neurodegeneration

cerebellar slices, KA has been shown to activate an AMPA receptor containing the GluR2 subunit on Bergmann glia which allows Ca2+ entry.109 AMPA receptor activation has been reported to reduce cell-cell junctional conductance between Bergmann glia in mouse cerebellar slices.130 Addition of glutamate to astrocytes has also been shown to engender selfpropagating Ca2+ waves through the syncytium in primary astrocyte cultures,31-33,131,132 and in hippocampal slices from 7-12 day old rats.133 This will at a minimum allow astrocytes to signal changes over a wide region of the brain. Because this superficially resembles how a neuronal network might function, there have been suggestions that these astrocytic Ca2+ waves are involved in some way in information processing, but it could simply be a mechanism whereby the astrocytes coordinate a homeostatic function, such as K+ transport within the entire astrocytic syncytium. A recent study in hippocampal slices134 showed that, in the negative feedback by glutamate of its own release at Schaffer collateral-CA1 pyramidal cell synapses mediated by presynaptic adenosine receptors, the adenosine seems to derive from perisynaptic astrocytes and inhibited synaptic transmission on the msec time scale. The effect requires simultaneous activation on the astrocytes of β adrenergic receptors by separate norepinephrine terminals and activation of an mGluR3 receptor on astrocytes by the synaptically released Glu, to raise astrocytic cAMP, which is the source of the adenosine. Thus, this system functions as a coincidence detector. The effects of activation of β-adrenergic receptors on the shape of astrocytes in vitro135,136 implied that astrocytes may have the property of changing shape during neuronal activities, including learning. There have also been a number of studies that have shown an increase in astrocyte number in the brains of rats taught tasks as compared to nonlearning rats,137 and GFAP positive processes have been shown to increase in animals which have gone through odor preference training.138 These changes, including the recent finding that inhibition of the astrocyte-specific enzyme glutamine synthetase by administration of methionine sulfoximine prior to a learning task inhibits retention of the task in day old chicks,67 imply roles for astrocytes in behavior.139

Astrocytes and the Blood-Brain Barrier (BBB) The BBB is a specialized structure responsible for the maintenance of the neuronal microenvironment. It plays a pivotal role in tissue homeostasis, fibrinolysis and coagulation, vasotonus regulation, the vascularization of normal and neoplastic tissues, and blood cell activation and migration during physiological and pathological processes, among other functions.140-143 Such regulation of blood-tissue exchange is first accomplished by individual endothelial cells being continuously linked by occluding tight junctions (zonulae occludentes). This isolates the brain from the blood and also negates the oncotic and osmotic forces that govern blood-tissue exchange elsewhere in peripheral tissues. A number of factors determine transport across the BBB.140 In the absence of specific carriers a substance’s permeability is largely dependent upon its lipophilicity. Certain molecules needed for brain metabolism, however, penetrate the BBB more readily than one would predict based on their lipid solubility alone, and such substances cross the barrier on specific carriers. Some of these carriers are symmetrically distributed both on the luminal and abluminal membranes of the endothelial cells, while others have an asymmetric distribution.143 For example, the carriers for the essential neutral amino acids, which are required in the brain for neurotransmitter synthesis, are localized on both luminal and abluminal membranes. In contrast, the carrier for the amino acid glycine appears to be located only on the abluminal membrane. This asymmetric distribution functions to remove glycine from the CNS and to keep its concentration in the brain low. Similarly, the abluminal membrane contains more of the (Na++K+) ATPase than does the luminal membrane. This enzyme forms the basis of

Functions of Astrocytes

27

a pump that simultaneously transports Na+ out of the endothelium into the brain, and K+ out of the brain into the endothelial cell. Like glycine, K+ has a potent effect on the transmission of nerve impulses and neuron firing, and this asymmetric distribution functions to maintain low K+ concentrations in the extracellular fluid.

Astrocytes and BBB-Induction In the mammalian CNS, brain capillaries develop from solid cords of endothelial cells. These cords develop a slit-like lumen, which progressively increases its caliber.144 The newly formed cords are separated from juxtaposed neurons by a basement membrane, and they progressively become ensheathed by resident astrocytes. The most rapid capillary sprouting corresponds to the period of glial cell proliferation and neuronal dendritic development and arborization.142 It was thought at one time that the astrocytic foot-processes actually formed the restrictive barrier, since this was the most obvious distinguishing feature between brain capillaries and all other capillaries in the periphery. However, electron microscope studies in the 1950s using electron-dense markers showed that the barrier to the diffusion of these markers was due to zonulae occludentes between the endothelial cells, and that there was free passage of such markers between the astrocytic end-feet.140,143 Recent work has indicated that the astrocytic end-feet processes may, however, play an important role in the induction of the BBB. Transplantation experiments have shown that the formation of the BBB depends largely on the CNS environment, since it did not form in capillaries growing into systemic tissue transplanted into the CNS, whereas the converse was true.145 Janzer and Raff146 demonstrated that astrocytes might be responsible for this phenomenon. They showed that injection of primary astrocyte cultures into the anterior eye chamber or chorioallantoic membrane of the chick induced a permeability barrier in the endothelial cells of the capillaries of these tissues that would otherwise lack such a barrier. Another line of evidence in support of the role of astrocytes in BBB induction derives from studies by Tao-Cheng et al.147 When endothelial cells were cultured alone, their tight junctions appeared fragmentary. When cocultured with astrocytes, the length, breadth and complexity of the tight junctions between the endothelial cells was increased, more closely resembling the structures seen in vivo. Interestingly, when other cell types such as fibroblasts were substituted for astrocytes the tight junctions remained fragmentary. There are also several lines of evidence for astrocytic induction of functional properties of CNS capillaries. These include gamma glutamyl transpeptidase (γ-GT) activity, a specific marker of endothelial cells of the CNS endothelium, which was abolished by the absence of astrocytes in a coculture system.148 Addition of astrocytes to endothelial cell cultures also increased the incorporation of neutral amino acids by the endothelial cells.149 The expression of the barrier-specific GLUT-1 isoform of the glucose transporter was markedly downregulated in cultured bovine brain capillary endothelial cells in the absence of brainderived or astrocyte trophic factors in the tissue culture medium.150 Astrocyte involvement in the differentiation and angiogenesis of the endothelial cells of the BBB is indirectly supported by the observation that vascular endothelial cell growth factor (VEGF) expression is induced and strongly upregulated in human malignant glioblastoma cells.141 VEGF is an angiogenic growth factor whose expression appears to parallel embryonic brain angiogenesis. Also, morphological differentiation and induction of specific BBB proteins can be induced by primary astrocyte cultures in endothelial cells in vitro.151-153 Despite the above evidence, it still remains unsettled whether astrocytes have a general role in the induction and maintenance of the BBB in vivo. Brightman154 concluded that “the precise role of perivascular astrocytes in the induction and maintenance of brain endothelium as a structural and functional barrier has yet to be fully elaborated.” Reasons for doubting a general inductive effect of astrocytes on the BBB are that the cerebral capillaries

28

Astrocytes in Brain Aging and Neurodegeneration

of a number of elasmobranchs are ensheathed by astrocytes, but their endothelial cells do not express tight junctions. Rather, they exhibit open pores which are permeable to large molecules, including horseradish peroxidase (HRP).155,156 As further suggested by Brightman,154 “either these particular astrocytes ensure that the endothelial junctions remain open or the junctional configuration is an inherent one that is not determined by the astrocytic investment.” Astrocytes are also found in close association with pituicytes, yet the endothelium in the neural lobe is of the fenestrated phenotype and it is largely permeable to dyes such as HRP. Cloned endothelial cells in vitro can establish a barrier with a relatively high resistance of 700-800 ohm⋅cm2.157 In further studies of injection of astrocytes into the anterior chamber and chorioallantoic membrane of the eye, they failed to form richly vascularized grafts.158 Rather, grafting of the astrocytes to the chorioallantoic membrane led to an extensive inflammatory response which, in turn, led to poor delivery of tracers to the graft vasculature. The iridial vessels associated with astrocyte grafts did not change their ultrastructure to resemble brain capillaries, and the astrocyte graft vasculature also failed to express high levels of the GLUT-1 isoform of the glucose transporter, even after treatment with anti-inflammatory agents.158 Hence, the authors questioned the general utility of the anterior chamber and chorioallantoic membrane for studying BBB induction, as used by Janzer and Raff.146 As pointed out by Abbott,159 in evolutionarily lower animals such as the cephalopod mollusks, the blood-brain barrier is formed between the glial cells and not between the endothelial cells. Abbott159 suggested that during evolution the barrier in vertebrates has likely shifted from glial cells to endothelial cells, “perhaps to allow greater complexity and control of the CNS interstitial environment by the glial cells, superimposed upon a barrier which prevented interference by large and rapid changes in the blood.”

Astrocytes and Immune and Inflammatory Responses in the CNS The CNS was regarded for decades as an “immunologically privileged” organ. This long-standing view that the brain is isolated from the effects of the immune system has been challenged with convincing experimental evidence that in response to invasion by microorganisms the CNS can mount its own defense by resident cells, such as the microglia and astrocytes.160-162 As summarized by Benveniste,163 “cells of the CNS constitutively express low levels of antigens encoded for by major histocompatibility complex (MHC) genes whose products play a fundamental role in the induction and regulation of immune responses.” However, both activated microglia and astrocytes can secrete a number of cytokines which can modulate the function of lymphoid-mononuclear cells, thus establishing an integrative communication pathway between resident cells of the CNS and those of the immune system. For a detailed discussion on the function of microglia in CNS immune mediation, the reader is referred to reviews in Graeber et al164 (also see chapter 4 of this volume). For a comprehensive review on CNS cytokines and their respective origins (i.e., astrocytes, microglia, macrophages) the reader is directed to a review chapter by Benveniste.163 As an example of how astrocytes are potentially involved in immune responses in the CNS, we will focus on astrocyte-specific cytokine elaboration and their potential role in initiation and suppression of immune responses, as well as the role of astrocyte-derived cytokines in sustaining and propagating CNS-induced damage

Do Astrocytes Modulate CNS Immune Responses? Evidence has accumulated for a role for cytokines in the CNS. For example, when directly injected into the brain, IL-1 promotes glial scarring or astrogliosis, suggesting that IL-1 may be important in mediating astrocytic hypertrophy upon neuronal injury.165 When primary astrocyte cultures derived from newborn mice are treated with lipopolysaccharide

Functions of Astrocytes

29

(LPS, E. coli) the astrocytes secrete interleukin-1 (IL-1).166 Intracerebral synthesis of IL-1 has been implicated as a prerequisite to intracerebral T cell activation, primarily because IL-1 enhances the production of IL-2 and expression of IL-2 receptors on T cells.167 It would appear, therefore, that astrocytes may be both responsive to IL-1 and capable of synthesizing it, providing for autocrine regulation of IL-1 levels within the CNS. The signals leading to the recruitment of circulating blood monocytes, and possibly resident CNS macrophages, are poorly understood. Astrocytes have been implicated as active participants in this process in view of their ability in primary culture to secrete an IL-3like factor which induces growth of cultured mouse peritoneal exudate cells (PEC) and brain tissue macrophages.168 Astrocytes also secrete granulocyte-macrophage colony-stimulating factor (GM-CSF), as evidenced by induction of colony formation in bone marrow cells and growth of FDC-P1 cells.169 GM-CSF is a cytokine necessary for growth and differentiation of macrophages and has been found to lead to an accumulation of macrophages at the site of inflammatory lesions. GM-CSF enhances a number of functional activities of mature macrophages such as their phagocytic, cytotoxic, and microbicidal activities. Migration of activated T cells across the compromised BBB in the course of CNS disease is associated with parenchymal production of interferon-γ (IFN-γ). IFN-γ interacts with astrocytes, as well as microglia, in the CNS, where it has been shown to modulate MHC gene expression and increase class I antigen expression.170 Expression of class I antigens on the astrocytic membranes increases their susceptibility to lysis by class 1-restricted cytotoxic T cells. Like IL-1, IFN-γ can lead to increased expression of adhesion molecules on astrocytes.171 Although IFN-γ does not appear to directly stimulate astrocytic cytokine production, it appears to “prime” the astrocytes to respond to other cytokines, such as TNF-α.172 The latter induces cytokine production by astrocytes, and leads to secretion of IL-6.173 The response of cultured astrocytes to IFN-γ results in increased expression of MHC antigens and costimulatory molecules (intercellular adhesion molecule-1, LFA-1 alpha) which mediate astrocyte-T cell interactions.174 IFN-γ can induce the production of the cytotoxic amino acid quinolinic acid, an NMDA agonist, and in conjunction with IL-1β it can increase NO synthetase (NOS) expression in astrocytes.175 Inducible NOS (iNOS) induction in astrocytes, as well as macrophages, has recently been postulated to contribute to HIV-related neurotoxicity.176 Recent studies also suggest that astrocyte-derived cytokines may be detrimental. As noted in the preceding paragraph, after an initial penetration of T cells into the CNS, astrocytes can further support the intracerebral T cell activation process. GM-CSF would therefore be expected to increase granulocyte and macrophage survival within the CNS and augment their activity against invading microbes.177 However, because viral replication in cultured HIV-infected monocytes is increased by GM-CSF, the potential for these cytokines to augment viral production in monocytes and microglia in HIV encephalitis exists, potentially worsening the spread of the infection within the CNS.178,179 Another cytokine, transforming growth factor (TGF) β1, was also recently implicated in facilitating the recruitment of peripheral infected monocytic cells, and thus it may contribute to HIV-1-related inflammation and the spread of the virus into the CNS.180 However, in view of the fact that many of the in vitro studies await in vivo confirmation, it is essential that future studies be directed toward determining the role of cytokines in inflammatory invasion of the CNS by blood-borne factors in vivo. There is circumstantial evidence that implicates astrocytes in mediating the neurotoxic effects of the HIV-1 soluble protein gp120. The latter has been reported to modulate several astrocyte functions, inducing intracellular signaling, ion transport, release of endogenous amino acids, and protein phosphorylation.181-185 A prominent effect of gp120 on astrocyte function includes increased efflux of K+. Since the glutamate carrier is both voltage- and ion

30

Astrocytes in Brain Aging and Neurodegeneration

gradient-dependent,68 increased [K+]o should increase [glutamate]o both because of reduced effectiveness of astrocytic glutamate uptake due to depolarization as well as swelling-induced glutamate release. A well-reproduced laboratory model for the CNS autoimmune disease, multiple sclerosis, is experimental allergic encephalomyelitis (EAE). As early as 1933, Rivers et al186 noted that when monkeys were injected with a rabbit brain extract they developed encephalitis that was characterized by destruction of the white matter. During the course of EAE, mononuclear leukocytes preferentially accumulate in the CNS. Ransohoff et al187 have recently monitored the factors that govern leukocyte trafficking in EAE. Using in situ hybridization, Ransohoff et al187 noted that astrocytes were the major source of mRNAs encoding for IP10 and JE/MCP-1, members of a family of chemoattractant cytokines. This suggested that astrocyte-derived cytokines may function as chemoattractants for inflammatory cells during EAE. For additional information on the role of cytokines in multiple sclerosis/autoimmune encephalitis the reader should consult the review by Benveniste.188 Infection of mice with the neurotropic JHM strain of mouse hepatitis virus (MHVJHM) leads within several weeks of infection to a demyelinating encephalomyelitis disease associated with prominent astrogliosis and infiltration of inflammatory cells. Analysis of infected spinal cords derived from these mice have recently revealed that three pleiotropic cytokines, TNF-α, IL-1β, and IL-6, as well as type 2 nitric oxide synthase (iNOS) are expressed by activated astrocytes localized to areas of virus infection and demyelination.189 These data also show that, by analogy to the human demyelinating disease multiple sclerosis, astrocytes are a major cellular source for these cytokines in mice with chronic MHVJHM infection and the findings are consistent with a role of astrocyte-derived cytokines and nitric oxide in the demyelinating process. It is now appreciated that, “different cytokines activate distinct functional programs in astrocytes, which may play a specific role in different brain diseases or at different stages of the same disease.”174 In addition, it appears that astrocytic responses to various cytokines largely depends on the presence or absence of neurons in the culture. Accordingly, neuronal-glial interactions may be of utmost importance in determining the activation threshold of astrocytes to inflammatory cytokines.174

Astrocytes as CNS Antigen Presenting Cells Astrocytes have been proposed to function as antigen-presenting cells (APCs), i.e., those cells with the ability to present antigens to lymphocytes.167,190 When astrocytes from Lewis rats were cocultured with a syngenic, MBP-specific, Ia-restricted T cell line of Lewis rat origin, they stimulated T cell proliferation. This process is antigen-specific and restricted to the MHC.191 During such cocultivation of T cells and astrocytes, the latter are induced by the preactivated T cells to express MHC type II molecules.160,191 Furthermore, supernatants of lectin-stimulated spleen cells containing IFN can induce murine astrocytes in culture to express Ia antigens,192 underscoring the point that astrocytes depend on the presence of Iainducing signals, such as IFN-γ, to function as APCs. However, the validity of these studies depends on the complete absence of microglia in the cultures,193 because, in both rat194 and human,195 microglia potently express MHC type II antigens in situ. Microglia constitute 5-10% of the total glial cell population and are considered to be the macrophages of the brain.164 Their major function is as a scavenger cell, ingesting cellular debris, a process which may be important for tissue modeling in the developing CNS. Microglia may also be involved in inflammation and repair in the adult CNS due to their phagocytic ability, release of neutral proteinases, and production of oxidative radicals. Microglia have been shown to express MHC antigens upon activation, and they are known to secrete a number of immunoregulatory cytokines and to respond to cytokine stimulation. At this stage, the evi-

Functions of Astrocytes

31

dence favors the microglia, and not the astrocytes, functioning as the brain’s APCs. Microglia are the more likely source of IL-1 during acute-phase brain injury because microglia are the first brain cells to appear in increased numbers at sites of trauma or infection. In addition, IL-1 appears to be produced more efficiently in microglia than in astrocytes.196 Other evidence favoring microglia as the principle APCs of the CNS includes observations on mixed glial cultures from adult human brain where only a limited number of astrocytes express MHC class II molecules, whereas the majority of the microglial cells were MHC IIpositive.197 In addition, microglia were readily induced by IFN-γ to express MHC II, whereas astrocytes were nonresponsive to the same treatment.198,197 Although earlier studies suggested that astrocytes can be induced by the preactivated T cells to express MHC type II molecules160,191 more recent studies challenge this view. Microglia treated with IFN-γ were capable of presenting MBP to MBP-specific T cells, whereas astrocytes could not fulfill such a role even in the presence of high concentrations of IFN-γ.197

Acknowledgments We would like to acknowledge support from the NIH to HKK (NS 19492 and NS 35205) and to MA (NIEHS 7331) for the period during which this chapter was written and for the experimental work cited from our laboratories. We thank Erin Grasek and Tina Giannakopoulos for help in preparing the manuscript.

References 1. Somjen GG. Nervenkitt: Notes on the history of the concept of neuroglia. Glia 1988;1:2-9. 2. Lugaro E. Sulle funzioni della nevroglia. Riv D Pat Nerv Ment 1907; 12:225-233. 3. Kuffler SW, Nicholls JG, Orkand RK. Physiological properties of glial cells in the central nervous system of amphibia. J Neurophysiol 1966; 29:768-787. 4. Fedoroff S, Vernadakis A. Astrocytes—Development, Morphology, and Regional Specialization of Astrocytes, Volumes 1-3. Orlando: Academic Press, 1986. 5. Kettenmann H, Ransom BR, eds. Neuroglia. New York: Oxford University Press, 1995. 6. Dahl D, Bjorklund H, Bignami A. Immunological markers in astrocytes. In: Fedoroff S, Vernadakis A, eds. Astrocytes: Cell Biology and Pathology of Astrocytes. Orlando: Academic Press, 1986:1-25. 7. Kimelberg HK, Norenberg MD. Astrocytic responses to central nervous system trauma. In: Salzman SK, Faden AI, eds. The Neurobiology of Central Nervous System Trauma. New York, Oxford: Oxford University Press, 1994:193-208. 8. Kimelberg HK. Primary astrocyte cultures—A key to astrocyte function. Cell and Molec Neurobiol 1983; 3:1-16. 9. Crick F. The Astonishing Hypothesis: The Scientific Search for the Soul. New York: Charles Scribner’s Sons, 1994. 10. Somjen, G. Electrophysiology of mammalian glial cells in situ. In: Kettenmann H, Ransom BR, eds. Neuroglia. New York, Oxford: Oxford University Press, 1995:319-334. 11. Hille B. Ionic Channels of Excitable Membranes. Sunderland, MA: Sinauer Assoc, Inc., 1992:1-20,115-139. 12. Barres BA, Chun LLY, Corey DP. Ion channels in vertebrate glia. Annual Rev Neurosci 1990; 13:441-474. 13. Duffy S, MacVicar BA. Voltage-dependent ionic channels in astrocytes. In: Murphy S, ed. Astrocytes, Pharmacology and Function. San Diego: Academic Press, 1993:137. 14. Kimelberg HK, Jalonen T, Walz W. Regulation of the brain microenvironment: Transmitters and ions. In: Murphy S, ed. Astrocytes, Pharmacology and Function. San Diego: Academic Press, Inc., 1993:193-228. 15. Sontheimer H. Voltage-dependent ion channels in glial cells. Glia 1994; 11:156-172. 16. Takumi T, Ishii T, Horio Y et al. A novel ATP-dependent inward rectifier potassium channel expressed predominantly in glial cells. J Biol Chem 1995; 270:16339-16346.

32

Astrocytes in Brain Aging and Neurodegeneration

17. Newman EA. High potassium conductance in astrocyte endfeet. Science 1986; 233:453-454. 18. Roy ML, Sontheimer H. β-adrenergic modulation of glial inwardly rectifying potassium channels. J Neurochem 1995; 64:1576-1584. 19. Robert A, Magistretti PJ. AMPA/kainate receptor activation blocks K+ currents via internal Na+ increase in mouse cultured stellate astrocytes. Glia 1997; 20:38-50. 20. Kressin K, Kuprijanova E, Jabs R et al. Developmental regulation of Na+ and K+ conductances in glial cells of mouse hippocampal brain slices. Glia 1995; 15:173-187. 21. Janigro D, Gasparini S, D’Ambrosio R et al. Reduction of K+ uptake in glia prevents longterm depression maintenance and causes epileptiform activity. J Neurosci 1997; 17:2813-2824. 22. Krnjevic K, Schwartz S. Some properties of unresponsive cells in the cerebral cortex. Exp Brain Res 1967; 3:306-319. 23. Bevan S, Chiu SY, Gray PTA et al. Presence of voltage-gated sodium, potassium and chloride channels in rat cultured astrocytes. Phil Trans Roy Soc of London: Biol 1985; 225:299-313. 24. Bowman CL, Kimelberg HK, Frangakis MV et al. Astrocytes in primary culture have chemically activated sodium channels. J Neurosci 1984; 4:1527-1534. 25. Sontheimer H, Ransom BR, Waxman SG. Different Na+ currents in P0- and P7-derived hippocampal astrocytes in vitro: Evidence for a switch in Na+ channel expression in vivo. Brain Res 1992; 597:24-29. 26. Sontheimer H, Waxman SG. Ion channels in spinal cord astrocytes in vitro. II. Biophysical and pharmacological analysis of two Na+ current types. J Neurophysiol 1992; 68:1001-1011. 27. Black JA, Westenbroek R, Ransom BR et al. Type II sodium channels in spinal cord astrocytes in situ: Immunocytochemical observations. Glia 1994; 12:219-227. 28. Ritchie JM. Current perspectives in glial electrophysiology. Ann N Y Acad Sci 1991; 633:331-342. 29. MacVicar BA. Voltage-dependent calcium channels in glial cells. Science 1984; 2265: 1345-1347. 30. Verkhratsky A, Kettenmann H. Calcium signaling in glial cells. Trends in Neurosciences Special Issue: glial signaling 1996; 19:346-352. 31. Cornell-Bell AH, Finkbeiner S. Ca2+ waves in astrocytes. Cell Calcium 1991; 12:185-204. 32. Finkbeiner S. Calcium waves in astrocytes—Filling in the gaps. Neuron 1992; 8:1101-1108. 33. Finkbeiner SM. Glia calcium. Glia 1993; 9:83-104. 34. Kim WT, Rioult MG, Cornell-Bell AH. Glutamate-induced calcium signaling in astrocytes. Glia 1994; 11:173-184. 35. Jalonen T. Single-channel characteristics of the large-conductance anion channel in rat cortical astrocytes in primary culture. Glia 1993; 9:227-237. 36. Roy G. Amino acid current through anion channels in cultured human glial cells. J Membr Biol 1995; 147:35-44. 37. Kimelberg HK. Current concepts of brain edema. J Neurosurg 1995; 83:1051-1059. 38. Strange K, Emma F, Jackson PS. Cellular and molecular physiology of volume-sensitive anion channels. Am J Physiol Cell Physiol 1996; 270:C711-C730. 39. Okada, Y. Volume expansion-sensing outward-rectifier Cl- channel: Fresh start to the molecular identity and volume sensor. Am J Physiol 1997; 273:C755-C789. 40. Berger T, Walz W, Schnitzer J et al. GABA- and glutamate-activated currents in glial cells of the mouse corpus callosum slice. J Neurosci Res 1992; 31:21-27. 41. Kirchhoff F, Mülhardt C, Pastor A et al. Expression of glycine receptor subunits in glial cells of the rat spinal cord. J Neurochem 1996; 66:1383-1390. 42. Pastor A, Chvátal A, Syková E et al. Glycine- and GABA-activated currents in identified glial cells of the developing rat spinal cord slice. Eur J Neurosci 1995; 7:1188-1198. 43. Steinhäuser C, Jabs R, Kettenmann H. Properties of GABA and glutamate responses in identified glial cells of the mouse hippocampal slice. Hippocampus 1994; 4:19-36. 44. Zalc B, Collet A, Monge M et al. Tamm-Horsfall protein, a kidney marker is expressed on brain sulfogalactosylceramide-positive astroglial structures. Brain Res 1984; 291:182-187.

Functions of Astrocytes

33

45. Kettenmann H, Bormann J. Patch-clamp study of gamma-aminobutyric acid receptor Cl– channels in cultured astrocytes. Proc Natl Acad Sci U S A 1988; 85:9336-9340. 46. Kimelberg HK, Frangakis MV. Furosemide- and bumetanide-sensitive ion transport and volume control in primary astrocyte cultures from rat brain. Brain Res 1985; 361:125-134. 47. Ballanyi K., Grafe P, Bruggencate GT. Ion activities and potassium uptake mechanisms of glial cells in guinea-pig olfactory cortex slices. J Physiol 1987; 382:159-174. 48. LoPachin RM Jr, Stys PK. Elemental composition and water content of rat optic nerve myelinated axons and glial cells: Effects of in vitro anoxia and reoxygenation. J Neurosci 1995; 15:6735-6746. 49. Sweadner KJ. Overlapping and diverse distribution of Na-K ATPase isozymes in neurons and glia. Can J Physiol Pharmacol 1992; 70:S255-S259. 50. Sweadner KJ. Na,-K-ATPase and its isoforms. In: Kettenmann H, Ransom BR, eds. Neuroglia. New York, Oxford: Oxford University Press, 1995:259-272. 51. Gloor S, Antonicek H, Sweadner KJ et al. The adhesion molecule on glia (AMOG) is a homologue of the B subunit of the Na, K-ATPase. J Cell Biology 1990; 110:165-174. 52. Sontheimer H, Black JA, Ransom BR et al. Ion channels in spinal cord astrocytes in vitro. I. Transient expression of high levels of Na+ and K+ channels. J Neurophysiol 1992; 68:985-1000. 53. Boyarsky G, Ganz MB, Sterzel RB et al. pH regulation in single glomerular mesangial cells I. Acid extrusion in absence and presence of HCO3. Am J Physiol: Cell Physiol 1993; 255:C844-C856. 54. Newman EA, Astion ML. Localization and stoichiometry of electrogenic sodium bicarbonate co-transport in retinal glial cells. Glia 1991; 4:424-428. 55. Pappas CA, Ransom BR. Depolarization-induced alkalinization (DIA) in rat hippocampal astrocytes. J Neurophysiol 1994; 72:2816-2826. 56. Chesler M. Regulation and modulation of pH in the nervous system. Prog Neurobiol 1990; 34:401-427. 57. Chesler M, Kraig RP. Intracellular pH transients of mammalian astrocytes. J Neurosci 1989; 9:2011-2019. 58. Grichtchenko II, Chesler M. Depolarization-induced alkalinization of astrocytes in gliotic hippocampal slices. Neurosci 1994; 62:1071-1078. 59. Kimelberg HK. Astrocytic edema in CNS trauma. J Neurotrauma 1992; 9:Suppl.1:S71-S81. 60. Kempski O, Zimmer M, Neu A et al. Control of glial cell volume in anoxia in vitro studies on ischemic cell swelling. Stroke 1987; 18:623-628. 61. Balazs R, Patel AJ, Richter D. Metabolic Compartmentation in the Brain. London: MacMillan Press, 1973:167-184. 62. Farinelli SE, Nicklas WJ. Glutamate metabolism in rat cortical astrocyte cultures. J Neurochem 1992; 58:1905-1915. 63. Shank RP, Leo GC, Zielke HR. Cerebral metabolic compartmentation as revealed by nuclear magnetic resonance analysis of D-[1- 13 C]glucose metabolism. J Neurochem 1993; 61:315-323. 64. Aas J, Berg-Johnson J, Hegstead E et al. Redistribution of glutamate and glutamine in slices of human neocortex exposed to combined hypoxia and glucose deprivation in vitro. J Cereb Blood Flow Metab 1993; 13:503-515. 65. Torp R, Andine P, Hagberg H et al. Cellular and subcellular redistribution of glutamate-, glutamine- and taurine-like immunoreactivities during forebrain ischemia: A semiquantitative electron microscopic study in rat hippocampus. Neurosci 1991; 41(2/3):433-447. 66. Diamond JS, Jahr CE. Transporters buffer synaptically released glutamate on a submillisecond time scale. J Neurosci 1997; 17:4672-4687. 67. Gibbs ME, O’Dowd BS, Hertz L et al. Inhibition of glutamine synthetase activity prevents memory consolidation. Cognit Brain Res 1996; 4:57-64. 68. Bouvier M, Szatkowski M, Amato A et al. Glial cell glutamate uptake carrier countertransports pH-changing anions. Nature 1992; 360:471-474. 69. Attwell D, Barbour B, Szatkowski M. Nonvesicular release of neurotransmitter. Neuron 1993; 11:403-407.

34

Astrocytes in Brain Aging and Neurodegeneration

70. Rutledge EM, Kimelberg HK. Release of [3H]-D-aspartate from primary astrocyte cultures in response to raised potassium. J Neurosci 1996; 16:7803-7811. 71. Ottersen OP, Laake JH, Reichelt W et al. Ischemic disruption of glutamate homeostasis in brain: Quantitative immunocytochemical analyses. J Chem Neuroanat 1996; 12:1-14. 72. Lehre KP, Levy LM, Ottersen OP et al. Differential expression of two glial glutamate transporters in the rat brain: Quantitative and immunocytochemical observations. J Neurosci 1995; 15:1835-1853. 73. Kanner BI. Glutamate transporters from brain: A novel neurotransmitter transporter family. FEBS Letters 1993; 325:95-99. 74. furuta A, Rothstein JD, Martin LJ. Glutamate transporter protein subtypes are expressed differentially during rat CNS development. J Neurosci 1997; 17:8363-8375. 75. Rothstein JD, Martin L, Levey AI et al. Localization of neuronal and glial glutamate transporters. Neuron 1994; 13:713-725. 76. Haugeto O, Ullensvang K, Levy LM et al. Brain glutamate transporter proteins form homomultimers. J Biol Chem 1996; 271:27715-27722. 77. Rothstein JD, Dykes-Hoberg M, Pardo CA et al. Knockout of glutamate transporters reveals a major role for astroglial transport in excitotoxicity and clearance of glutamate. Neuron 1996; 16:675-686. 78. Bristol LA, Rothstein JD. Glutamate transporter gene expression in amyotrophic lateral sclerosis motor cortex. Ann Neurol 1996; 39:676-679. 79. Durkin MM, Smith KE, Borden LA et al. Localization of messenger RNAs encoding three GABA transporters in rat brain: An in situ hybridization study. Mol Brain Res 1995; 33:7-21. 80. Morara S, Brecha NC, Marcotti W et al. Neuronal and glial localization of the GABA transporter GAT-1 in the cerebellar cortex. NeuroReport 1996; 7:2993-2996. 81. Sanchez-Olea R, Morales M, Garcia O et al. C1 channel blockers inhibit the volume-activated efflux of C1 and taurine in cultured neurons. J Amer Physiol Soc 1996; 270:C1703C1708. 82. Kimelberg HK, Katz DM. High-affinity uptake of serotonin into immunocytochemically identified astrocytes. Science 1985; 228:889-891. 83. Matz H, Hertz L. Effects of adenosine deaminase inhibition on active uptake and metabolism of adenosine in astrocytes in primary cultures. Brain Res 1990; 515:168-172. 84. Huszti Z, Imrik P, Madarász E. [3H]histamine uptake and release by astrocytes from rat brain: Effects of sodium deprivation, high potassium, and potassium channel blockers. Neurochem Res 1994; 19:1249-1256. 85. Hosli L, Hosli E. Autoradiographic studies on the uptake of 3H-noradrenaline and 3Hserotonin by neurones and astrocytes in explant and primary cultures of rat CNS: Effects of antidepressants. Int J Devl Neuroscience 1995; 13:897-908. 86. Kimelberg HK. Receptors on astrocytes—What possible functions. Neurochem Int 1995; 26:27-40. 87. Cornell-Bell AH, Finkbeiner SM, Cooper MS et al. Glutamate induces calcium in cultured astrocytes: Long-range glial signaling. Science 1990; 247:470-473. 88. Glaum SR, Miller RJ. Acute regulation of synaptic transmission by metabotropic glutamate receptors. In: Conn PJ, Patel J eds. The Metabotropic Glutamate Receptors. Totowa, NJ: Humana Press, 1994:147-172. 89. Deecher DC, Wilcox BD, Dave V et al. Detection of 5-hydroxytryptamine2 receptors by radioligand binding, Northern blot analysis, and Ca2+ responses in rat primary astrocyte cultures. J Neurosci Res 1993; 35:246-256. 90. Ogata T, Nakamura Y, Tsuji K et al. Adenosine enhances intracellular Ca2+ mobilization in conjunction with metabotropic glutamate receptor activation by t-ACPD in cultured hippocampal astrocytes. Neurosci Letts 1994; 170:5-8. 91. Lerea LS, McCarthy KD. Neuron-associated astroglial cells express α1 and β adrenergic receptors in vitro. Brain Res 1990; 521:7-14. 92. Barres BA, Chun LLY, Corey DP. Calcium current in cortical astrocytes: Induction by cAMP and neurotransmitters and permissive effect of serum factors. J Neurosci 1989; 9:3169-3175.

Functions of Astrocytes

35

93. Kimelberg HK, Goderie SK, Conley PA et al. Uptake of [3H]serotonin and [3H]glutamate by primary astrocyte cultures I. Effects of different sera and time in culture. Glia 1992; 6:1-8. 94. Landis DMD, Weinstein LA, Skordeles CJ. Serum influences the differentiation of membrane structure in cultured astrocytes. Glia 1990; 3:212-221. 95. Jalonen TO, Margraf RR, Wielt DB. et al. Serotonin induces inward potassium and calcium currents in rat cortical astrocytes. Brain Res 1997; 758:69-82. 96. Michler-Stuke A, Wolff JR, Bottenstein JE. Factors influencing astrocyte growth and development in defined media. Int J Devl Neuroscience 1984; 2:575-584. 97. Miller S, Romano C, Cotman CW. Growth factor upregulation of a phosphoinositidecoupled metabotropic glutamate receptor in cortical astrocytes. J Neurosci 1995; 15:6103-6109. 98. Raff MC, Miller RH, Noble M. A glial progenitor cell that develops in vitro into an astrocyte or an oligodendrocyte depending on culture medium. Nature 1983; 303:390-396. 99. Shao Y, McCarthy KD. Regulation of astroglial responsiveness to neuroligands in primary culture. Neurosci 1993; 55:991-1001. 100. Miller S, Sehati N, Romano C et al. Exposure of astrocytes to thrombin reduces levels of the metabotropic glutamate receptor mGluR5. J Neurochem 1996; 67:1435-1447. 101. Kimelberg HK, Cai Z, Rastogi P et al. Transmitter-induced calcium responses differ in astrocytes acutely isolated from rat brain and in culture. J Neurochem 1997; 68:1088-1098. 102. Cai Z, Kimelberg HK. Glutamate receptor-mediated calcium responses in acutely isolated hippocampal astrocytes. Glia 1997; 21:380-389. 103. Fraser DD, Duffy S, Angelides KJ et al. GABAa/Benzodiazepine receptors in acutely isolated hippocampal astrocytes. J Neurosci 1995; 15:2720-2732. 104. MacVicar BA, Tse FWY, Crichton SA et al. GABA-activated Cl– channels in astrocytes in hippocampal slices. J Neurosci 1989; 9(10):3577-3583. 105. Porter JT, McCarthy KD. GFAP-positive hippocampal astrocytes in situ respond to glutamatergic neuroligands with increases in [Ca2+]i.. Glia 1995; 13:101-112. 106. Porter JT, McCarthy KD. Adenosine receptors modulate [Ca2+]i in hippocampal astrocytes in situ. J Neurochem 1995; 65:1515-1523. 107. Kirischuk S, Möller T, Voitenko N et al. ATP-induced cytoplasmic calcium mobilization in Bergmann glial cells. J Neurosci 1995; 15:7861-7871. 108. Kirischuk S, Tuschick S, Verkhratsky A et al. Calcium signaling in mouse Bergmann glial cells mediated by a1-adrenoreceptors and H1 histamine receptors. Eur J Neurosci 1996; 8:1198-1208. 109. Muller T, Moller T, Berger T et al. Calcium entry through kainate receptors and resulting potassium-channel blockade in Bergmann glial cells. Science 1992; 256:1563-1566. 110. Gallo V, Patneau DK, Mayer ML et al. Excitatory amino acid receptors in glial progenitor cells: Molecular and functional properties. Glia 1994; 11:94-101. 111. Seifert G, Steinhäuser C. Glial cells in the mouse hippocampus express AMPA receptors with an intermediate Ca2+ permeability. Eur J Neurosci 1995; 7:1872-1881. 112. Aoki C, Go C-G, Venkatesan C et al. Perikaryal and synaptic localization of α2A-adrenergic receptor immunoreactivity in brain as revealed by light and electron microscopic immunocytochemistry. Brain Res 1994; 650:181-204. 113. Aoki C, Venkatesan C, Go C-G et al. Cellular and subcellular localization of NMDA-R1 subunit immunoreactivity in the visual cortex of adult and neonatal rats. J Neurosci 1994; 14:5202-5222. 114. Whitaker-Azmitia PM, Clarke C, Azmitia EC. Localization of 5-HT1A receptors to astroglial cells in adult rats: Implications for neuronal-glial interactions and psychoactive drug mechanism of action. Synapse 1993; 14:201-205. 115. Van den Pol AN, Romano C, Ghosh P. Metabotropic glutamate receptor mGluR5 subcellular distribution and developmental expression in hypothalamus. J Comp Neurol 1995; 362:134-150.

36

Astrocytes in Brain Aging and Neurodegeneration

116. Jeffery G, Sharp C, Malitschek B et al. Cellular localization of metabotropic glutamate receptors in the mammalian optic nerve: A mechanism for axon-glia communication. Brain Res 1996; 741:75-81. 117. Petralia RS, Wang YX, Zhao HM et al. Ionotropic and metabotropic glutamate receptors show unique postsynaptic, presynaptic, and glial localizations in the dorsal cochlear nucleus. J Comp Neurol 1996; 372:356-383. 118. Van der Zee EA, De Jong GI, Strosberg AD et al. Muscarinic acetylcholine receptor-expression in astrocytes in the cortex of young and aged rats. Glia 1993; 8:42-50. 119. Paspalas CD, Papadopoulos GC. Ultrastructural relationships between noradrenergic nerve fibers and non-neuronal elements in the rat cerebral cortex. Glia 1996; 17:133-146. 120. Petralia RS, Wang YX, Niedzielski AS et al. The metabotropic glutamate receptors, mGluR2 and mGluR3, show unique postsynaptic, presynaptic and glial localizations. Neurosci 1996; 71:949-976. 121. Luque JM, Richards JG. Expression of NMDA 2B receptor subunit mRNA in Bergmann glia. Glia 1995; 13:228-232. 122. Shigemoto R, Nakanishi S, Mizuno N. Distribution of the mRNA for a metabotropic glutamate receptor (mGluR1) in the central nervous system: An in situ hybridization study in adult and developing rat. J Comp Neurol 1992; 322:121-135. 123. Yokoi M, Kobayashi K, Manabe T et al. Impairment of hippocampal mossy fiber LTD in mice lacking mGluR2. Science 1996; 273:645-647. 124. Kay AR, Wong RKS. Calcium current activation kinetics in isolated pyramidal neurones of the CA1 region of the mature guinea-pig hippocampus. J Physiol (Lond) 1987; 392:603-616. 125. Tse FW, Fraser DD, Duffy S et al. Voltage-activated K+ currents in acutely isolated hippocampal astrocytes. J Neurosci 1992; 12(5):1781-1788. 126. Duffy S, MacVicar BA. Adrenergic calcium signaling in astrocyte networks within the hippocampal slice. J Neurosci 1995; 15:5535-5550. 127. Bowman CL, Kimelberg HK. Excitatory amino acids directly depolarize rat brain astrocytes in primary culture. Nature 1984; 311:656-659. 128. Kettenmann H, Schachner M. Pharmacological properties of gamma-aminobutyric acid-, glutamate- and aspartate-induced depolarizations in cultured astrocytes. J Neurosci 1985; 5:3295-3301. 129. Sontheimer H, Kettenmann H, Backus KH et al. Glutamate opens Na+/K+ channels in cultured astrocytes. Glia 1988; 1:328-336. 130. Muller T, Moller T, Neuhaus J et al. Electrical coupling among Bergmann glial cells and its modulation by glutamate receptor activation. Glia 1996; 17:274-284. 131. Van den Pol AN, Finkbeiner S, Cornell-Bell AH. Calcium excitability and oscillations in suprachiasmatic nucleus neurons and glia in vitro. J Neurosci 1992; 12:2648-2664. 132. Cornell-Bell AH, Finkbeiner SM., Cooper MS. Glutamate induces calcium in cultured astrocytes: Long-range glial signaling. Science 1990; 247:470-473. 133. Pasti L, Volterra A, Pozzan T et al. Intracellular calcium oscillations in astrocytes: A highly plastic, bidirectional form of communication between astrocytes and neurons in situ. J.Neurosci 1997; 17:7817-7830. 134. Winder DG, Ritch PS, Gereau RW et al. Novel glial-neuronal signaling by coactivation of metabotropic glutamate and β-adrenergic receptors in rat hippocampus. J Physiol (Lond) 1996; 494:743-755. 135. Narumi S, Kimelberg HK, Bourke RS. Effects of norepinephrine on the morphology and some enzyme activities of primary monolayer cultures from rat brain. J Neurochem 1978; 31:1479-1490. 136. Shain W, Forman DS, Madelian V et al. Morphology of astroglial cells is controlled by beta-adrenergic receptors. J Cell Biol 1987; 105:2307-2314. 137. Diamond MC, Law F, Rhodes H et al. Increases in cortical depth and glia numbers in rats subjected to enriched environment. J Comp Neurol 1966; 128:117-126. 138. Matsutani S, Leon M. Elaboration of glial cell processes in the rat olfactory bulb associated with early learning. Brain Res 1993; 613:317-320.

Functions of Astrocytes

37

139. Kimelberg HK, Jalonen TO, Aoki C, McCarthy K. Transmitter receptor and uptake systems in astrocytes and their relation to behaviour. In: Laming PR, Sykova E, Reichenbach A, Hatton G, Bauer H. Glial Cells: Their Role in Behaviour. Cambridge: Cambridge University Press 1998; 6:107-129. 140. Rapoport SI. Blood-Brain Barrier in Physiology and Medicine. New York: Raven Press, 1976. 141. Risau W. Molecular biology of blood-brain barrier ontogenesis and function. Acta Neurochir 1994; 60:109-112. 142. Risau W. Differentiation of endothelium. FASEB J 1995; 9:926-933. 143. Goldstein GW, Betz AL. Blood-brain barrier. Scientific Amer 1986; 255:74-83. 144. Jacobson M. Histogenesis and morphogenesis of the central nervous system. In: Jacobson M, ed. Developmental Neurobiology. New York: Plenum Press, 1978:57-114. 145. Stewart PA, Wiley MJ. Developing nervous tissue induces formation of blood-brain barrier characteristics in invading endothelial cells: A study using quail-chick transplantation chimeras. Develop Biology 1981; 84:183-192. 146. Janzer RC, Raff MC. Astrocytes induce blood-brain barrier properties in endothelial cells. Nature 1987; 325:253-257. 147. Tao-Cheng J, Nagy Z, Brightman MW. Tight junction of brain endothelium in vitro are enhanced by astroglia. J Neurosci 1987; 7:3293-3299. 148. DeBault LE, Cancilla PA. γ-Glutamyl transpepitidase in isolated brain endothelial cells: Induction by glial cells in vitro. Science 1980; 207:653-655. 149. Cancilla PA, DeBault LE. Neutral amino acid transport properties of cerebral endothelial cells in vitro. J Neuropathol Exp Neurol 1997; 191-199. 150. Boado RJ, Wang L, Pardridge WM. Enhanced expression of the blood-brain barrier GLUT1 glucose transporter gene by brain-derived factors. Mol Brain Res 1994; 22:259-267. 151. Dehouck M-P, Meresse S, Delorme P et al. An easier, reproducible, and mass-production method to study the blood-brain barrier in vitro. J Neurochem 1990; 54:1798-1801. 152. Lobrinus JA, Juillerat-Jeanneret L, Darekar P et al. Induction of the blood-brain barrier specific HT7 and neurothelin epitopes in endothelial cells of the chick chorioallantoic vessels by a soluble factor derived from astrocytes. Dev Brain Res 1992; 70:207-211. 153. Tagami M, Yamagata K, Fujino H et al. Morphological differentiation of endothelial cells co-cultured with astrocytes on type-I or type-IV collagen. Cell and Tissue Res 1992; 268:225-232. 154. Brightman MW. Implications of astroglia in the blood-brain barrier. Ann NY Acad Sci 1991; 633:343-347. 155. Brightman MW, Reese TS, Olson Y et al. Morphological aspects of the blood-brain barrier to peroxidase in elasmobranchs. Progr Neuropathol 1971; 1:146-161. 156. Bundgaard M, Cserr HF. A glial blood-brain barrier in elasmobranchs. Brain Res 1981; 226:61-73. 157. Rutten MJ, Hoover RL, Karnovsky MJ. Electrical resistance and macromolecular permeability of brain endothelial monolayer cultures. Brain Res 1987; 425:301-310. 158. Holash JA, Noden DM, Stewart PA. Re-evaluating the role of astrocytes in blood-brain barrier induction. Dev Dynamics 1993; 197:14-25. 159. Abbott NJ, Raff MC. Glial-neuronal interaction. Preface. Ann N Y Acad Sci 1991; 633:xiii-xv. 160. Fierz W, Endler B, Reske K et al. Astrocytes as antigen presenting cells: I. Induction of Ia antigen expression on astrocytes by T cells via immune interferon and its effect on antigen presentation. J Immunol 1990; 134:3785-3793. 161. Fontana A, Erb P, Pircher H et al. Astrocytes as antigen-presenting cells. Part II: Unlike H-2K-dependent cytotoxic T cells H-2Ia-restricted T cells are only stimulated in the presence of interferon-gamma. J Neuroimmunol 1986; 12:15-28. 162. Schnyder B, Weber E, Fierz W et al. On the role of astrocytes in polyclonal T cell activation. J Neuroimmunol 1986; 10:209-218. 163. Benveniste EN. Astrocyte-microglia interactions. In: Murphy S, ed. Astrocytes: Pharmacology and Function. San Diego: Academic Press, 1993:355-382. 164. Graeber MB, Kreutzberg GW, Streit WJ. Microglia. Glia 1993; 7:2-118.

38

Astrocytes in Brain Aging and Neurodegeneration

165. Giulian D. Interleukin-1 injected into mammalian brain stimulates astrogliosis and neovascularization. J Neurochem 1988; 8:2485-2490. 166. Fontana A, Kristensen F, Dubs R et al. Production of prostaglandin E and interleukin 1-like factors by cultured astrocytes and C-6 glioma cells. J Immunol 1982; 129:2413-2419. 167. Fontana A, Frei K, Bodmer S et al. Immune-mediated encephalitis: On the role of antigen presenting cells in brain tissue. Immunol Rev 1987; 100:185-201. 168. Frei K, Bodmer S, Schwerdel C et al. Astrocytes of the brain synthesize interleukin 3-like factors. J Immunol 1985; 135:4044-4047. 169. Malpiero UV, Frei K, Fontana A. Production of hemopoietic colony-stimulating factors by astrocytes. J Immunol 1990; 144:3816-3821. 170. Wong GHW, Barlett PF, Clark-Lewis I et al. Inducible expression of H-2 and Ia antigens on brain cells. Nature 1984; 310:688-691. 171. Frohman EM, Frohman TC, Dustin ML et al. The induction of intracellular adhesion molecule 1 (ICAM-1) expression on human fetal astrocytes by interferon-γ tumor necrosis factor-α, lymphotoxin, and interleukin-1: Relevance to intracerebral antigen presentation. Neuroimmunol 1989; 23:117-124. 172. Chung IY, Benveniste EN. Tumor necrosis factor-α production by astrocytes: Induction by lipopolysaccharide, IFN-gamma, and IL-1β. J Immunol 1990; 144:2999-3007. 173. Benveniste EN, Sparacio SM, Morris JG et al. Induction and regulation of interleukin-6 gene expression in rat astrocytes. J Neuroimmunol 1990; 30:201-212. 174. Aloisi F, Borsellino G, Caré A et al. Cytokine regulation of astrocyte function: In-vitro studies using cells from the human brain. Int J Dev Neurosci 1995; 13:265-274. 175. Nathan C. Nitric oxide as a secretory product of mammalian cells. FASEB 1992; 6:3051-3064. 176. Nottet HSLM, Gendelman HE. Unraveling the neuroimmune mechanisms for the HIV-1 associated cognitive/motor complex. Immunol Today 1995; 16:441-448. 177. Garland JM. Colony stimulating factors. In: Thomson A, ed. The Cytokine Handbook. San Diego: Academic Press, 1991:269-300. 178. Mucke L, Eddleston M. Astrocytes in infectious and immune-mediated diseases of the central nervous system. FASEB 1993; 7:1226-1232. 179. Tweardy DJ, Mott PL, Glazer EW. Monokine modulation of human astroglial cell production of granulocyte colony-stimulation factor and granulocyte-macrophage colony-stimulating factor. 1. Effects of IL-1 alpha and IL-1 beta. J Immunol 1990; 144:2233-2241. 180. Wahl SM, Allen JB, McCartney-Francis N et al. Macrophage- and astrocyte-derived transforming growth factor beta as a mediator of central nervous system dysfunction in acquired immune deficiency syndrome. J Exp Med 1991; 173:981-991. 181. Benos DJ, Hahn BH, Bubien JK et al. Envelope glycoprotein gp120 of human immunodeficiency virus type 1 alters ion transport in astrocytes: Implications for AIDS dementia complex. Proc Natl Acad Sci USA 1994; 91:494-498. 182. Ciardo A, Meldolesi J. Effects of the HIV-1 envelope glycoprotein gp 120 in cerebellar cultures. [Ca2+] increases in glial cell subpopulation. Euro J Neurosci 1993; 5:1711-1718. 183. Levi G, Patrizio M, Bernardo A et al. Human immunodeficiency virus coat protein gp120 inhibits the β-adrenergic regulation of astroglial and microglial functions. Proc Natl Acad Sci U S A 1993; 90:1541-1545. 184. Pulliam L, West D, Haigwood N et al. HIV-1 envelope gp 120 alters astrocytes in human brain cultures. AIDS Res Hum Retrovirus 1993; 9:439-444. 185. Schneider-Schaulies J, Schneider-Schaulies S, Brinkman R et al. HIV-1 and gp 120 receptor on CD-4-negative brain cells activates a tyrosine kinase. Virology 1992; 191:765-772. 186. Rivers TM, Sprunt DH, Berry GP. Observations on attempts to produce acute disseminated encephalomyelitis in monkeys. J Exp Med 1933; 58:39-53. 187. Ransohoff RM, Hamilton TA, Tani M et al. Astrocyte expression of mRNA encoding cytokines IP-10 and JE/MCP-1 in experimental autoimmune encephalomyelitis. FASEB J 1993; 7:592-600. 188. Benveniste EN. The role of cytokines in multiple sclerosis/autoimmune encephalitis and other neurological disorders. In: Agrawal B, Puri R, eds. Human Cytokines, Their Role in Research and Therapy. Boston: Blackwell Science Publications, 1995:195-216.

Functions of Astrocytes

39

189. Sun N, Grzybicki D, Castro RF et al. Activation of astrocytes in the spinal cord of mice chronically infected with a neurotropic coronavirus. Virology 1995; 213:482-493. 190. Erb P, Kennedy M, Hagmann I et al. Accessory cells and the activation and expression of different T cell functions. In: Feldmann M, McMichael A, eds. Regulation of Immune Gene Expression. Clifton: The Humana Press, 1986:187 191. Fontana A, Fierz W, Wekerle H. Astrocytes present myelin basic protein to encephalitogenic T-cell lines. Nature 1984; 307:273-276. 192. Hirsch M-R, Wietzerbin J, Pierres M et al. Expression of Ia antigens by cultured astrocytes treated with gamma-interferon. Neurosci Lett 1988; 41:199-204. 193. Giulian D, Baker TJ. Characterization of ameboid microglia isolated from developing mammalian brain. Neurosci 1986; 6(8):2163-2178. 194. Vass K, Lassmann H. Intrathecal application of interferon gamma. Progressive appearance of MHC antigens within the rat nervous system. Am J Pathol 1990; 137:789-800. 195. Lampson LA, Hickey WF. Monoclonal antibody analysis of MHC expression human brain biopsies: Tissue ranging from “histologically normal” to that showing different levels of glial tumor involvement. J Immunol 1986; 136:4054-4062. 196. Giulian DJ, Baker TJ, Shih L-CN et al. Interleukin 1 of the central nervous system is produced by ameboid microglia. J Exp Med 1986; 164:594-604. 197. Williams K, Bar-Or A, Ulvestad E et al. Biology of adult human microglia in culture: Comparisons with peripheral blood monocytes and astrocytes. J Neuropath Exp Neurol 1992; 51:538-549. 198. Matsumoto Y, Ohmori K, Fujiwara M. Immune regulation by brain cells in the central nervous system: Microglia but not astrocytes present myelin basic protein to encephalitogenic T cells under in vivo-mimicking conditions. Immunol 1992; 76:209-216.

CHAPTER 3

Astrocyte Pathophysiology in Disorders of the Central Nervous System Michael D. Norenberg

Introduction

A

strocytes play a prominent role in the central nervous system (CNS) response to injury. These responses may be useful in restoring the integrity of the CNS microenvironment as well as in promoting reparative and regenerative events. On the other hand, some astrocytic actions may harm the CNS and possibly impede regeneration. This article will consider the contributions of astrocytes to neurologic disease in the context of passive and active glial responses: Astrocytes may be injured and become incapable of carrying out their normal function, resulting in a gliopathy (i.e., passive response), or ostensibly “healthy” astrocytes may secrete potentially harmful molecules and thus play an active role in CNS disorders.

Normal Functions A growing body of evidence indicates that astrocytes are involved in many functions that are critical to the CNS. Their activities involve important interactions with neurons,1 oligodendrocytes,2 microglia,3 and endothelial cells.4 A particularly important function is the maintenance and regulation of the extracellular environment. Such actions include buffering of K+, H+ and Ca2+ and osmoregulation.5 While most studies have employed cell culture methods to examine glial function, recent in vivo investigations utilizing the selective gliotoxin fluoroacetate have given added support for the critical role of astrocytes in the maintenance of the extracellular environment.6,7 Other putative astroglial functions include neurotransmitter and neuromodulator uptake and release;8,9 regulation of synaptic transmission and neuronal excitability;10 provision of nutrients, energy substrates and neurotransmitter precursors;11,12 neurotrophism;13 metabolism and detoxification of ammonia, drugs and hormones;14,15 free radical scavenging;16 metal sequestration;17 development and maintenance of the blood-brain barrier;18 guidance of neuronal migration during development;19 and immune/inflammatory functions.20 The astrocyte uptake of the excitatory neurotransmitter glutamate is a particularly critical astrocytic function.21 It not only serves to recycle glutamate, regulate glutamatergic neurotransmission and detoxify ammonia, but it is also necessary to avoid excitotoxic injury.22

Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

42

Astrocytes in Brain Aging and Neurodegeneration

Glial uptake of glutamate is achieved by powerful transporters. Three glutamate transporters have been cloned: GLT-1,23 GLAST,24 EAAC1.25 In situ hybridization26 and immunohistochemical studies27-29 indicate that GLT-1 is strictly astrocytic, GLAST is mostly glial but also found in neurons, while EAAC1 is chiefly neuronal. Astrocyte cultures express chiefly the GLAST transporter.30 A defect in astrocyte glutamate transport contributes to the pathogenesis of several neurological disorders (see below).

General Response to Injury Reactive Astrocytosis Glial transformation to reactive astrocytes (gemistocytes) represents one of the earliest and dramatic responses of the CNS to tissue injury. Reactive astrocytosis (gliosis, astrogliosis) occurs following physical, chemical, ischemic, infectious, immunologic and neuro-degenerative disorders.This response is characterized by cellular hypertrophy and a profusion of new, thicker and longer cytoplasmic processes. The end product of gliosis is often referred to as a “glial scar”. Electron microscopic findings show changes consistent with enhanced metabolic activity, i.e., increased numbers of mitochondria and ribosomes, enlarged Golgi complexes and increased amounts of glycogen.31 A most striking change is the accumulation of bundles of 10 nm intermediate glial filaments, composed chiefly of glial fibrillary acidic protein (GFAP) and vimentin.32 Overexpression of GFAP is currently the most commonly used indicator of reactive astrocytosis. The reactive astrocyte produces a wide array of molecules including growth factors, extracellular matrix molecules (glycoproteins and proteoglycans), adhesion molecules, β-amyloid precursor protein (APP), proteases, protease inhibitors, and immune/inflammatory molecules (MHC molecules, cytokines, chemokines). Additionally, many enzymes are upregulated; those particularly related to CNS disorders include the inducible form of nitric oxide synthase, monoamine oxidase-B, superoxide dismutase, catalase, glutathione-S-transferase (an enzyme involved in the detoxification of various xenobiotics), kynurenine aminotransferase, 3-hydroxyanthranilic acid oxygenase, and quinolinic acid phosphoribosyltransferase. For reviews on factors produced by reactive astrocytes see refs. 33, 34. The morphologic changes of reactive astrocytes are those of metabolically activated cells. The precise significance of this activation is uncertain and there is considerable controversy regarding its beneficial or deleterious effect on the CNS. One might predict that early stages of reactive astrocytosis may be involved in the restoration of the ionic milieu, provision of nutrients, removal of toxins (including excitotoxins such as glutamate), free radical scavenging, and metal sequestration. Later, astroglial responses are perhaps geared towards promoting repair and regeneration. Nevertheless, the dominant view until recently was that gliosis created an inhospitable, non-permissive environment for neurite outgrowth, interfered with remyelination, and possibly disturbed neuronal circuitry leading to seizures. These issues will be covered below.

Astrocyte Swelling Astrocyte swelling represents one of the earliest pathological features of most CNS injuries, and at times may be the only abnormality found. It occurs following ischemia, trauma, hypoglycemia, status epilepticus, hypotonicity, and fulminant hepatic failure. It is usually best seen in the perivascular astrocytic endings, possibly because of the greater number of transport systems at that site. Various factors have been implicated in the mechanism of swelling, including osmotic changes, abnormalities in ion transport, and excessive concen-

Astrocyte Pathophysiology in Disorders of the Central Nervous System

43

trations of glutamate, lactic acid, arachidonic acid, potassium, free radicals, and ammonia. For reviews on glial swelling see ref. 35. Astrocyte swelling may lead to increased intracranial pressure and its associated complications. Swelling may also impair astrocyte integrity and function. Swollen glia depolarize and thus lose their ability to maintain the necessary ionic gradients for the uptake of glutamate and other amino acids. Moreover, swollen astrocytes release K+ and glutamate, which may result in changes in the level of excitability and contribute to excitotoxicity.36,37 The reduction in the size of the extracellular space following astrocyte swelling may also elevate extracellular ionic concentrations, which could affect neuronal excitability.38 Astrocyte swelling may also compress capillaries, contributing to a reduction in cerebral blood flow.39 Ultimately, when swelling is severe, the cell membrane may rupture, resulting in cell death. Astrocyte swelling also causes the release of large amounts of taurine.36 While such release is likely to help restore cell volume (due to loss of an osmolyte), taurine has neuroinhibitory effects which can affect the state of neuronal excitability.40 Whether this reduction in excitability is useful or not is difficult to predict. Taurine may also exert a neuroprotective effect,41 possibly through its antioxidant properties.42

Alzheimer Type II Response Alzheimer type II astrocytes are seen in a variety of metabolic encephalopathies including hepatic encephalopathy (HE), uremia, hypercapnia and the early stages of anoxia and hypoglycemia, especially in infants (for review see ref. 43). This change often occurs in the setting of elevated brain or blood ammonia. The process is occasionally referred to as protoplasmic astrocytosis or metabolic gliosis. Alzheimer type II astrocytes have enlarged, pale nuclei with peripheral margination of chromatin and often prominent nucleoli. In experimental models of HE, Alzheimer type II astrocytes contain increases in mitochondria, smooth and rough endoplasmic reticulum, and cytoplasmic glycogen. Eventually, degenerative changes characterized by hydropic changes, cytoplasmic vacuoles and degenerated mitochondria are observed.44 The initial change in the Alzheimer type II astrocyte suggests that it is a metabolically active cell responding to a perturbation in the extracellular milieu (presumably to excessive levels of ammonia). The later changes are indicative of cell injury. Culture studies, and more recently in vivo studies, have identified various functional deficits that may contribute to the encephalopathy of HE and related conditions (see below).

Injury to Astrocytes in CNS Disorders (Passive Role) Although astrocytes are more resistant to various CNS insults than neurons and oligodendrocytes, they are nevertheless vulnerable to many injurious processes and may even undergo necrosis. Injury or death of astrocytes can cause severe impairment in the regulation of extracellular potassium concentrations, amino acid levels, and extracellular pH.7 Astrocytes may occasionally be the primary target of injury (primary gliopathy). Hepatic encephalopathy, associated with the Alzheimer type II response, is probably the best example of such a process. Many factors are released in injured brain, including lactic acid, potassium, arachidonic acid, ammonia, free radicals, glutamate, nitric oxide and cytokines.45 All of these factors negatively impact on glial function. Elevation of extracellular K+ occurring in tissue injury leads to astrocyte depolarization,46 intracellular alkalinization,47 and extracellular acidification.48 Astrocytic depolarization diminishes the ability of astrocytes to take up glutamate.49 Additional effects of potassium on astrocytes include increased glycogenolysis,50,51 which may contribute to the generation of lactic acid. Elevated K+ may also be a factor in glial swelling.52

44

Astrocytes in Brain Aging and Neurodegeneration

The release of glutamate and associated excitotoxicity contributes to the pathogenesis of ischemia, trauma, hypoglycemia and various neurodegenerative conditions.53,54 Glutamate may also have profound effects on astrocytes, including depolarization, swelling, activation of phosphoinositide hydrolysis, increase in intracellular Ca2+, generation of calcium waves, morphologic changes, stimulation of glucose utilization and lactate release, enhanced glycogen synthesis, decreased cell proliferation, production of growth factors, protooncogene and transcription factor expression, inhibition of MHC class II expression, and the release of GABA, glutamate, aspartate, glycine, taurine, alanine, and serine.55,56 Lactic acid causes injury to astrocytes,57,58 contributes to glial swelling,59,60 and impairs the astrocyte’s capacity to take up glutamate.61,62 Arachidonic acid is a potent inhibitor of glial glutamate uptake and a cause of glial swelling,63,64 and free radicals are have been shown to inhibit glutamate uptake by astrocytes65,66 and to cause glial swelling.63,64

Active Role of Astrocytes in CNS Disorders This section reviews the possibility that in some circumstances astrocytes actively produce compounds that are potentially harmful to the CNS.

Excitotoxins Astrocytes have been shown to release glutamate and/or aspartate in the following conditions: treatment with kainic acid and depolarization with high concentrations of K+;67 in a culture model of hypoxia/ischemia;68 following inhibition of glycolysis;69,70 and after treatment with the HIV-1 coat protein gp120,71 mercuric chloride,72 trimethyltin,73 and aluminum.74 Astrocyte swelling can also cause glutamate to be released into the extracellular space.36 Such increases in extracellular glutamate may contribute to excitotoxic injury. Quinolinic acid, a tryptophan metabolite with excitotoxic properties, is synthesized in the CNS by glial cells. While the majority of studies indicate that astrocytes are the cells involved in this process,75 evidence for a microglial source is also available.76 The synthetic enzyme 3-hydroxyanthranilic acid oxygenase is upregulated following tissue injury.77 Inappropriate release of quinolinic acid may contribute to excitotoxic injury. However, the same metabolic pathway that generates quinolinic acid also generates kynurenic acid, which has inhibitory effects on glutamate receptors.78 The net effect of this dual release is unknown. Quinolinic acid may play a role in AIDS encephalopathy (see below), seizures and various neurodegenerative diseases.54,75 Glycine, a known activator of the NMDA receptor, can be synthesized and released by astrocytes after treatment with kainic acid, quisqualate and high concentrations of K+.67,79,80

Glutamine Astrocytes are well known to synthesize and release glutamine.81,82 Simantov showed that in mixed CNS cultures, high concentrations of glutamine are toxic.83 The potential toxicity of glutamine is likely due to its being the principal precursor of glutamate.84 Glutamine may additionally contribute to excitotoxicity through its inhibition of kynurenine aminotransferase,85 thereby preventing the synthesis of kynuretic acid, a glutamate receptor antagonist. The neuroprotective effect of methionine sulfoximine, an inhibitor of glutamine synthetase which is mainly found in astrocytes,86 could be due to its ability to abrogate the potential excitotoxic properties of glutamine.87

Lactic Acid Lactic acid can be released from astrocytes as a result of stimulation of anaerobic glycolysis. The almost exclusive localization of glycogen to astrocytes may contribute to this release.88 Lactic acid may result in extracellular acidosis and contribute to tissue injury.89

Astrocyte Pathophysiology in Disorders of the Central Nervous System

45

Recent studies have additionally shown that it reduces astrocytic glutamate uptake,61,62 possibly contributing to excitotoxicity. It also causes glial swelling.90 The deleterious effects of lactic acid, however, may be partially offset by the fact that it is a source of fuel for neighboring neurons,11,91 and has additionally been found to reduce NMDA receptor activation.92

Nitric Oxide Nitric oxide (NO) contributes to normal physiological processes,93 including those in brain.94 While NO plays a role in host defense,95,96 it may also contribute to tissue injury and may contribute to the pathogenesis of various neurological diseases.93,97 NO is synthesized from L-arginine by nitric oxide synthase (NOS). Astrocytes possess both the constitutive and inducible forms of NOS.98 The inducible form of NOS has been shown to be elevated in reactive astrocytes following CNS injury.99 The significance of this increase is uncertain. NO possesses cytotoxic properties that could contribute to neuronal death.93,97 However, because of its vasodilating effect, NO may also improve blood flow.100 While microglia are the major producers of NO following injury, astrocytes are the principal source of NO in humans.99

Free Radicals In general, it appears that astrocytes serve a protective role in mitigating the actions of free radicals.16 In certain sites, such as the hypothalamus, there is evidence that astrocytes may be a source of free radicals.101

Extracellular Matrix Molecules (ECM) Proteoglycans are complex molecules consisting of a protein core to which chains of glycosaminoglycans are covalently bound. Sulfate groups confer a high anionic charge to these molecules. Astrocytes have been shown to synthesize chondroitin sulfate proteoglycans, dermatan sulfate proteoglycan, and heparan sulfate proteoglycans.102,103 Proteoglycans appear to stimulate neurite outgrowth, guidance and remodeling during development.104,105 Despite these beneficial actions of ECMs, astrocyte-derived proteoglycans have been shown to inhibit neurite outgrowth, to contribute to the development of neuritic plaques in Alzheimer’s disease, and to play a role in establishing the epileptic focus (see below). Various astrocyte-derived glycoproteins such as tenascin,102,106,107 hyaluronate-binding protein,108 Thy-1 glycoprotein,109 and other as yet unidentified proteins110 have been shown to exert a repulsive action on neurite outgrowth and may contribute to aberrant synaptogenesis associated with the epileptic lesion (see below).

Inflammatory/Immune Molecules The inflammatory response contributes to the destruction of microorganisms and removal of necrotic debris (phagocytosis), which sets the scene for appropriate reparative responses. CNS inflammation is associated with the activation of microglia, recruitment of blood monocytes and neutrophils and increased vascular permeability. However, an overreactive inflammatory response may also be deleterious. Indeed, a fine balance is at work and the outcome of the inflammatory response may be difficult to predict.111 It has become clear that astrocytes contribute to immune/inflammatory phenomena.20 Various cytokines have been identified as products of astrocytes including IL-1β, IL-3, IL-6 and TNF-α (for reviews see refs. 112, 113). Some cytokines may contribute to tissue injury in AIDS, multiple sclerosis and experimental allergic encephalomyelitis (see below). Chemokines are recently described molecules that potently recruit inflammatory cells following injury and play a key role in wound healing.114 In the CNS they are made mostly by astrocytes and are upregulated in astrocytes following treatment with various

46

Astrocytes in Brain Aging and Neurodegeneration

cytokines.115-117 Among these chemokines include monocyte chemoattractant protein-1α (MCP-1), a chemoattractant and stimulator of monocytes and macrophage inflammatory protein 1 (MIP-1), which is strongly chemotactic for neutrophils and other leukocytes. Class II major histocompatability (MHC) antigens are critical molecules involved in antigen presentation and are vital in the initiation of immune responses. These molecules have been identified on astrocytes,118-121 although this matter is controversial.122,123 The protease cathepsin G is found in reactive astrocytes after trauma.124 Proteases may be beneficial in the destruction of toxic cytokines, neurite extension, remodeling of extracellular matrix, chemotaxis and hemostasis.125-127 They may also contribute to tissue damage and to the accumulation of β-amyloid.128 Matrix metalloproteinases can also be produced by astrocytes;129,130 these proteases break down connective tissue and have been implicated in ischemia, multiple sclerosis, and amyotrophic lateral sclerosis.131 α2-Macroglobulin, a broad spectrum protease inhibitor that can be synthesized by astrocytes,132,133 is an acute phase protein that acts to eliminate proteases and cytokines from inflammatory processes. It appears to play a role in development through an interaction with plasminogen activator and other proteases. Another protease inhibitor, α 1antichymotrypsin, appears to be involved in the genesis of neuritic plaques in Alzheimer’s disease (see below). Adhesion molecules are involved in cell-cell interaction, cell migration, neurite outgrowth and guidance, synaptogenesis, synaptic reorganization, and myelination.134 Adhesion molecules also possess potent chemotactic properties for microglia and/or leukocytes. Excessive recruitment of these inflammatory cells may produce further tissue injury. The intracellular adhesion molecule, ICAM-1, has been found on astrocytes and can be upregulated in these cells by IFN-γ and TNF-α.121,135,136 A preliminary study has also identified vascular adhesion molecule-1 in astrocytes.137 Adhesion molecules may play a role in regeneration, inflammatory disorders, Alzheimer’s disease and in the production of seizure foci (see below).

Astrocyte-Microglial Interactions Microglia are the resident histiocytes of the CNS and are the principal cells involved in inflammatory and immunological responses as well as in phagocytosis. Microglia release certain factors that are potentially harmful to the CNS including nitric oxide, superoxide anions, reactive oxygen radicals, and excitotoxins. Microglia also synthesize and release cytokines as well as various proteases. Interestingly, astrocytes appear capable of counteracting the harmful effects of microglia138,139 (see refs. 140-142 for review). Important interactions occur between astrocytes and microglia/macrophages. Microglia are a prime source of gliotic mediators (IL-1β, TNF-α, IL-6). Inhibiting microglia with colchicine decreases the extent of astrogliosis.143 Microglial-derived IL-1 stimulates nerve growth factor production by astrocytes,144,145 possibly contributing to regeneration. In turn, astrocytes are the principal producers of granulocyte-macrophage colony stimulating factor in brain (GM-CSF),146,147 which serves as a growth factor for microglia, and induces granulocytes and macrophages to migrate into inflammatory foci, thereby increasing their retention and survival.148 Since GM-CSF increases viral replication in cultured HIV infected monocytes, it has been postulated that this cytokine increases the viral load in HIV encephalitis.149 Astrocytes can also stimulate microglial proliferation through their production of IL-3150 and laminin.151 While abundant amounts of extracellular matrix are deposited in gliotic tissue, not all of it necessarily comes from astrocytes. Microglia/monocytes are capable of making chondroitin sulfate proteoglycan.152 Furthermore, macrophage/microglial release of IL-1β153 may increase the production of chondroitin sulfate proteoglycan by astrocytes in certain brain

Astrocyte Pathophysiology in Disorders of the Central Nervous System

47

lesions. Ness and David154 have additionally shown that astrocytes cocultured with meningeal cells or in the presence of conditioned media from meningeal cells inhibited neurite growth. These cocultured astrocytes had more tenascin-C and chondroitin sulfate proteoglycan and had less laminin, a factor that supports neurite outgrowth.

Clinical Considerations Regeneration Most axons in the vertebrate CNS fail to regenerate following injury. This failure has been largely attributed to a “hostile terrain” due to the presence of inhibitory molecules,155 or inadequate supply of growth factors or growth promoting molecules.156 One of the impediments to repair and regeneration has been the deposition of a glial scar, a notion originally expressed by Ramon y Cajal.157 By acting as a mechanical barrier, the glial scar could interfere with the process of axonal regeneration. Reactive astrocytes grown on nitrocellulose filters,158 as well as membranes isolated from gliotic lesions,159 have been shown to be poor substrates for neurite outgrowth. Interestingly, this inhibition was only observed in membranes derived from isomorphic gliosis, whereas no such inhibitory effect was observed in membranes derived from anisomorphic gliosis.159 As reviewed above, some of the inhibitory factors are proteoglycans and glycoproteins. Gliosis at the root entry zone blocks regeneration of dorsal roots,160,161 possibly due to the development of synaptic-like contacts with reactive astrocytes.162,163 Astrocytes may thus activate a physiologic stop pathway as opposed to inhibiting neurite growth. Before assigning too sinister a role to the astrocyte, it may be useful to remember that astrocytes can exert a beneficial role, at least in the young or immature brain.164,165 The generation of a glial scar forms a glia limitans which may isolate the lesion from the remaining viable tissue. Astrocytes have been shown to promote neuronal survival in culture and in vivo, protect neurons against excitotoxic and anoxic injury, and aid in neuronal repair and functional recovery. Astrocytes produce various neurotrophic factors, and provide an excellent substrate for neurite outgrowth, mediated by a number of adhesion molecules and extracellular matrix molecules. They are also involved in the removal of degenerating synapses and the provision of axonal guidance (see refs. 13, 166 for review). While it is clear that astrocytes, and in particular young or immature astrocytes, are supportive of neurite outgrowth, older astrocytes have a diminished capacity to do so.109,167,168 Perhaps mature cells do not provide sufficient surface molecules, extracellular matrix molecules or other substances for regeneration to occur.169 It should be recalled that meninges, fibroblasts, microglia/macrophages, and oligodendrocytes also contribute to the formation of the glial scar. Consequently, any or all of these cells may adversely affect the outcome of regeneration. Certainly, the fibroblastic/gliotic scar is impermeable to axonal growth.32,170,171 Furthermore, as noted above, microglia/macrophages play a major role in the deposition of proteoglycans and glycoproteins following CNS lesions. There is also growing evidence that oligodendrocytes/myelin contribute to the inhospitable environment following CNS injury.172,173 In summary, the issue remains unresolved whether astrocytes, or particularly reactive astrocytes, play a beneficial or detrimental role in CNS regeneration. Overall, it appears that while immature astrocytes have a positive influence on neurite outgrowth, reactive astrocytes in adults may be injurious. Although reactive astrocytes may serve to wall off an area of injury, generate various growth promoting substances and mediate a number of homeostatic functions, excessive production of certain proteoglycans and glycoproteins appear to impede the regeneration process.

48

Astrocytes in Brain Aging and Neurodegeneration

Acquired Immunodeficiency Syndrome (AIDS) The majority of people with AIDS eventually develop neurological symptoms (cognitive, motor, behavioral). Pathological changes in the CNS include inflammation, microgliosis, multinucleated giant cells, pallor and vacuolization of the white matter, cerebral atrophy, and a vacuolar myelopathy. Astrocytic gliosis is a common and early finding in HIV-1 infection.174,175 The mechanism for neuronal injury is not known, as neurons appear not to be directly involved by HIV, suggesting that the viral effects on neurons are indirect.176 A commonly held view focuses on the deleterious effects of HIV-1-infected microglia/macrophages. These cells are capable of releasing cytokines, nitric oxide, arachidonic acid and its metabolites, unknown neurotoxins, free radicals, and platelet-activating factor (PAF). Important interactions between astrocytes and microglia occur in AIDS. It appears that astroglial cells are required to activate HIV-1-infected monocytes to produce the various neurotoxic factors.177,178 Conditioned media of lipopolysaccharide-treated astrocytes increase HIV-1 gene expression in monocytes,179 and astrocyte-derived IL-6 promotes HIV-1 replication.180 Conversely, supernatants from HIV-treated macrophages induce cultured astrocytes to release TGF-β.181 There is evidence that astrocytes may be also be infected with HIV-1, particularly in the pediatric age group.182-184 Astrocyte dysfunction induced by direct HIV-1 infection could thus possibly affect CNS development. An indirect mechanism for neuronal injury has been the involvement of the HIV-1 envelope glycoprotein gp120. Pulliam and colleagues185 were the first to suggest that AIDS dementia may partially involve a perturbation of astrocyte function by gp120. These workers showed that treatment of human brain tissue with gp120 caused astrocyte alterations and death. Studies on astrocyte cultures showed decreased expression of glial fibrillary acidic protein (GFAP), as well as the diminution of a major 66 kDa phosphoprotein. Levi et al subsequently reported that gp120 inhibits β-adrenergic regulation of astroglial and microglial functions.186 Other studies have also shown that gp120 interacts with microglia and/or astrocytes to release neurotoxic compounds, some of which act synergistically with glutamate to activate NMDA receptors.187 Several studies have demonstrated that gp120 also produces abnormalities in astrocytic glutamate transport. Benos and coworkers71,188 showed that gp120 stimulated a Na+/H+ antiporter, resulting in loss of the Na+ gradient, intracellular alkalinization, activation of outward K+ conductance, membrane depolarization and increased glutamate efflux. Dreyer and Lipton189 showed that gp120 impairs astrocyte uptake of excitatory amino acids, and the resulting excess glutamate may lead to neuronal damage. They suggested that the effect of gp120 on astroglial glutamate uptake may be indirect, as a consequence of a direct effect of gp120 on macrophages, which in turn releases arachidonic acid, a known inhibitor of glutamate uptake.190-192 Quinolinic acid is a tryptophan derivative that has excitotoxic properties. A potential role for quinolinic acid in AIDS has been described.193 Since enzymes involved in the synthesis of quinolinic acid have been identified in astrocytes,75 excessive quinolinate production by astrocytes could contribute to excitotoxic damage. It should be noted, however, that microglia may also be a source of quinolinic acid76 (for general reviews on astrocytes and AIDS, see refs. 194-196).

Multiple Sclerosis Multiple sclerosis (MS) is an inflammatory demyelinating disease of unknown etiology that is widely considered to have an immune pathogenesis.197 MS is associated with prominent astrocytosis and indeed the term sclerosis refers to this astrocytic response. Re-

Astrocyte Pathophysiology in Disorders of the Central Nervous System

49

active astrocytes in the MS lesion (plaque) can be extremely large, are often binucleated and may even show atypical nuclei.197 The precise role of astrocytes in MS is not clear. Some investigators support the view that astrocytes are capable of actively promoting demyelination and suppressing remyelination,198,199 while others believe that astrocytes support remyelination.200,201 Astrocytes may also play a role in phagocytosis of myelin debris.120,202 As discussed above, reactive astrocytes contribute to the process of inflammation by virtue of their production of cytokines, proteases,203 MHC class II antigens, adhesion molecules, and nitric oxide. Reactive astrocytes associated with MS plaques also contain substance P,204 a potent mediator of vasodilatation and local immune responses.205 Experimental allergic encephalomyelitis (EAE) is an autoimmune disorder that is widely used as an experimental model of MS.197,206 Reactive astrocytosis is a prominent histopathological feature in EAE.207,208 As in MS, astrocytes have been shown to produce cytokines, chemokines and adhesion molecules,121,209-212 all of which could contribute to demyelination. There is controversy regarding the presence or absence of MHC class II antigens in astrocytes in EAE. As reviewed by Benveniste,113 such failure may be due to the time during the disease process in which they were looked for, and the fact that class II antigens are more readily downregulated in astrocytes as compared to microglia.213,214 Furthermore, astrocytes may be destroyed via an MHC class II restricted cytotoxicity.119 On the other hand, Massa and colleagues215 have shown that astrocytes from a susceptible rat strain (Lewis) express higher amounts of class II antigens in vitro after IFN or virus treatment compared to a less susceptible strain (Brown-Norway). In keeping with this observation are studies showing that EAE resistant and susceptible strains of rats also differ in their ability to express TNF.209 Despite these detrimental factors potentially contributing to demyelination, it should be noted that astrocytes can promote process outgrowth by adult human oligodendrocytes in vitro through the interaction between astrocyte derived bFGF and extracellular matrix molecules (vitronectin, fibronectin, laminin, heparan sulfate proteoglycans) (see ref. 216 and references cited therein). Additionally, insulin-like growth factor is present in astrocytes,217 is upregulated following injury218,219 and has trophic actions on oligodendrocytes.220

Alzheimer’s Disease Reactive gliosis is a prominent finding in Alzheimer’s disease (AD),221,222 which may actively contribute to its pathogenesis as opposed to merely representing a nonspecific reaction to tissue injury. Such gliosis is associated with an increase in GFAP223,224 and GFAP mRNA.225 Reactive astrogliosis is also a prominent feature of the neuritic plaque.226 Neuritic plaques represent a major component of the pathology of AD and consist of extracellular masses of amyloid intimately associated with dystrophic neurites, activated microglia and reactive astrocytes. Important advances have been made in recent years so that a clearer picture of the evolution of plaques is beginning to emerge. Fundamental to the initiation of the plaque is the formation of β-amyloid1-42 derived from the abnormal proteolytic processing of β-amyloid precursor protein (APP).227 The deposition of β-amyloid (diffuse plaque) initiates a cascade of events that culminate in the formation of the neuritic plaque. An excellent review on the genesis of neuritic plaques has recently been published228 and a detailed account of the role of astroglia in plaque formation is presented in chapter 5 of this volume. The neurotoxic β-amyloid1-42 is a 4 kDa, 42 amino acid peptide that contains a complement activation domain (initiating inflammation), glycation binding sites (which contribute to the recruitment of microglia) as well as binding sites for apolipoprotein E (Apo E), α1-antichymotrypsin and proteoglycans. Through the interaction with microglia and

50

Astrocytes in Brain Aging and Neurodegeneration

astrocytes (see below), the soluble β-amyloid is converted to an insoluble, fibrillary protein with a folded, β-pleated sheet configuration which is resistant to protease digestion and stains positive with thioflavin S and Congo red (congophilia) and is neurotoxic.229,230 An important effect of β-amyloid is the activation of microglia. This activation is associated with the release of proinflammatory cytokines, reactive oxygen species, proteases, acute phase proteins and other poorly defined toxic molecules,231-233 and represents a key pathogenetic step in the progression of AD. Activated microglia and β-amyloid are also responsible for the reactive gliosis associated with the neuritic plaque.234,235 While reactive astrocytes in the plaque may act to wall off the amyloid from the surrounding neuropil,226 it may have a more sinister role. Reactive astrocytes in this setting appear to secrete a number of molecules that are critical to the formation of neuritic plaques (see below) as well as various toxic compounds (cytokines, free radicals, nitric oxide, proteases). Reactive astrocytes in AD express elevated levels of α1-antichymotrypsin (ACT),236,237 probably secondary to microglial-derived IL-1β and TNF-α. ACT is an acute phase protein, serine protease inhibitor, capable of inhibiting cathepsin G and chymotrypsin that normally prevent excessive proteolysis during inflammation. ACT is intimately associated with β-amyloid;238,239 this association appears to promote the assembly of β-amyloid into filaments and contributes to the resistance of β-amyloid to inflammatory protease damage.240-242 Similarly, astrocytes are a source of proteoglycans (see above) which are found in neuritic plaques.243,244 Proteoglycans induce β-amyloid to form insoluble β-pleated structures (fibrils), and protect β-amyloid from proteolytic degradation.245-247 Apolipoprotein E (Apo E) is normally involved in the transport (recycling) of triglycerides and cholesterol.248 The identification of the Apo E type 4 allele is of importance, as it influences the risk of acquiring AD.249 Apo E in brain is primarily a product of astrocytes250,251 and its increased deposition has been described in AD.250,252 Apo E (presumably a mutant form) binds to β-amyloid, thereby leading to the transformation of β-amyloid to fibrillary amyloid.241,253,254 Abundant neurite sprouting and formation of dystrophic neurites is characteristic of neuritic plaques.255 Pike et al256 showed that appearance of reactive astrocytes occur earlier than dystrophic neurites in mild AD suggesting that reactive astrocytes are the cause of the dystrophic neurites. It has been suggested that ACT,238 bFGF,257 adhesion molecules258 and S-100β, all derived from astrocytes, may contribute to the formation of abnormal sprouts and dystrophic neurites found in plaques. While astrocytes play a key role in the formation of fibrillary amyloid from β-amyloid, the latter also has potent affects on astrocytes. β-amyloid may be responsible for astrocyte activation in AD.234,235,259 It also stimulates the production of matrix metalloproteinases by astrocytes,130 which may break down myelin260 and impair the BBB.131 It also stimulates the production of bFGF by astrocytes,259 possibly contributing to increased production of APP by reactive astrocytes.261,262 It additionally causes a loss of astroglial calcium homeostasis, impairment of ion transport, generation of free radicals, and a decrease in glutamate uptake (for review, see ref. 263).

Epilepsy The potential involvement of astrocytes in epilepsy dates back to the work of Penfield264 who emphasized the role of the glio-mesenchymal scar as a factor in seizure production. Gliosis is arguably the most common and consistent neuropathologic feature in the seizure focus,265-267 and the extent of gliosis often parallels the severity of seizure activity.268,269 As expected, astrocytes show an increase in GFAP, GFAP mRNA and vimentin.270-273 Other astrocytic changes have included an elevated activity of various oxidative enzymes,274 in-

Astrocyte Pathophysiology in Disorders of the Central Nervous System

51

creased number of gap junctions,275 and alterations in Ca2+ homeostasis.276 Interestingly, astrocytic alterations may precede the onset of seizures.277 The mechanism by which astrocytes contribute to the epileptic state remains speculative. An older view suggested that gliosis resulted in mechanical deformation and heightened excitability.278 It is likely, however, that perturbations of glial functions are the factors that lead to altered states of excitability. Any imbalance between excitatory and inhibitory processes is likely to result in seizure activity.279,280 In view of the key role of astrocytes in glutamate, GABA and taurine uptake and release, abnormalities in these functions may contribute to the epileptic state.281 Consistent with this possibility is the finding that a glialselective inhibitor of GABA uptake, THPO,282 can protect against experimental seizures.283 Other mechanisms by which glial dysfunction may contribute to the epileptic state include abnormalities in the metabolism of K+,284,285 H+,47,286 ammonia,287,288 and quinolinic acid.289,290 Experimental models of epilepsy are associated with aberrant synaptic sprouting that may lead to abnormal recurrent excitatory circuits that result in seizures.291-294 Many of these models are associated with the production of various astrocyte-derived molecules including growth factors (bFGF),295 fibronectin,296 vitronectin,297 tenascin,107,294 neural cell adhesion molecules,298,299 and S-100.300 It has been suggested that such molecules may be responsible for the production of the aberrant axonal sprouting.301,302

Parkinson’s Disease and MPTP Neurotoxicity The involvement of astrocytes in Parkinson’s disease is presented in detail in chapter 6 of this volume. Here we describe a recent observation on the potential involvement of astroglial glutamate uptake in the pathogenesis of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) neurotoxicity. There is evidence to suggest that excessive glutamatergic activity may play a role in Parkinson’s disease.303-305 We have recently shown that MPTP significantly diminished astroglial glutamate uptake.306 Thus, in addition to the oxidation of MPTP to the toxin MPP+ by astroglial MAO B, interference in glutamate uptake may be another means by which astrocytes contribute to MPTP neurotoxicity and parkinsonism.

Amyotrophic Lateral Sclerosis (ALS) ALS is a progressive disorder of motor neurons of the cortex, brainstem and spinal cord leading to muscle atrophy and weakness. As in other degenerative diseases, it is associated with a prominent gliosis.307 Abnormalities in glutamate metabolism and excitotoxicity have been proposed as pathogenetic mechanisms for ALS.308 Patients show a marked decrease in glutamate uptake in synaptosomes from spinal cord, motor cortex, and somatosensory cortex.309 Subsequent studies by the same research group have shown that the astrocyte-specific transporter GLT-1 was markedly decreased in ALS, in motor cortex and spinal cord.310 Additionally, loss of either GLAST or GLT-1 transporter by the use of antisense oligonucleotides to the glial glutamate transporters resulted in a progressive motor deficit in rats.311 All of these findings support the concept that defective clearing of glutamate from the extracellular space may lead to neurotoxic levels of extracellular glutamate and contribute to the neuronal damage in ALS. The mechanism for defects in glutamate uptake in ALS is not known. A mutation in Cu/Zn superoxide dismutase (SOD-1) has been identified in about 20% of patients with familial ALS.312 Eosinophilic inclusions containing SOD-1 and ubiquitin have been identified in astrocytes of patients with familial ALS.313 Several SOD-1 mutant mice have been described that show clinical and pathological features of ALS.314,315 As in cases of familial ALS, astrocytic SOD-1 and ubiquitin-containing cytoplasmic inclusions are also found.

52

Astrocytes in Brain Aging and Neurodegeneration

Importantly, these inclusions appear before the clinical onset of disease.315 This model also shows reduced expression of the GLT-1 glutamate transporter. Recent studies have also described a reduction or absence of SOD-1 in reactive astrocytes in cases of sporadic ALS.316 In view of the important role of SOD in protection against oxidative stress,317,318 a defect in scavenging free radicals may partially explain the abnormality in glial glutamate uptake.

Stroke A detailed account of the role of astrocytes in stroke is given in a recent review.319 Astrocyte swelling and cellular hypertrophy is seen as early as 1-3 hours following reversible ischemia.320 These astrocytes show increased numbers of mitochondria and rough endoplasmic reticulum in keeping with evidence of increased protein synthesis.321 The nuclei are enlarged and pale and these cells resemble Alzheimer type II astrocytes that have been described in hyperammonemia and hepatic encephalopathy.43 Whether these astroglial changes are secondary to elevated levels of ammonia that have been documented in ischemia,322,323 or are mediated through some other mechanism, is not known. In the early transitional phase of the Alzheimer type II change, astrocytes appear metabolically activated, consistent with data suggesting an increased glucose utilization by glial cells.324 Cell culture studies have shown that anoxic-ischemic injury diminishes the capacity of astrocytes for glutamate uptake,321,325,326 which can exacerbate excitotoxicity.327,328 A transient decrease in the glutamate transporter GLT-1 has been identified in the CA1 region of the hippocampus following global ischemia.329 Glutamate uptake may also be compromised because of energy failure, leading to depolarization and the inability of astrocytes to maintain ionic gradients that are necessary for glutamate transport.21,49 Alternatively, various factors are generated in the ischemic process that are known to interfere with glutamate uptake including arachidonic acid, free radicals, lactic acid, and possibly nitric oxide.330

Hepatic Encephalopathy (HE) HE is a common neurological complication of severe liver disease which occurs in acute and chronic forms. Acute HE presents with the abrupt onset of delirium, seizures, and coma. The principal cause of death in acute HE is brain edema associated with increased intracranial pressure. Chronic HE, sometimes referred to as portal-systemic encephalopathy, is characterized by altered mental state, change in personality, diminished intellectual capacity, abnormal muscle tone and tremor. The pathogenesis of HE remains poorly understood. The dominant view over many decades has been the generation of gut-derived neurotoxins, with the greatest emphasis on the role of ammonia.331 In more recent years, the involvement of heightened GABAergic neurotransmission, possibly through the action of endogenous benzodiazepines, has been stressed.332 The pathology of HE suggests that astrocytes play a crucial role in this condition. Astrocyte swelling represents the principal component of acute HE and likely contributes to the brain edema found in this condition, while the Alzheimer type II astrocytic change is the histological hallmark of chronic HE. No significant or consistent neuronal changes have been identified (for reviews of astrocyte changes in HE, see refs. 43, 333). Astrocytes are the cells in brain where ammonia is metabolized, as glutamine synthetase (GS) is predominantly located in astrocytes.86 Astrocytes are also involved in glutamate uptake, and failure of astrocytes to do so may contribute to abnormal glutamatergic neurotransmission.334 Additionally, astrocytes may be sites of action of benzodiazepines, putative factors in the pathogenesis of HE. The view we have espoused is that astrocytes are primarily injured in HE14,333 and that astrocytic dysfunction contributes to neuronal derangements, leading to encephalopathy.14

Astrocyte Pathophysiology in Disorders of the Central Nervous System

53

Administration of ammonium chloride to cultured astrocytes reproduces the pathological changes observed in HE.335,336 Such treatment also decreases glial fibrillary acidic protein (GFAP) and GFAP mRNA, consistent with the loss of GFAP observed in humans with HE.337 Additional effects of ammonia on astrocytes have included a decreased cAMP response to β-adrenergic agonists, decreased Ca2+ influx, altered protein phosphorylation, diminished glycogen levels, and reduced glutamate and GABA uptake. Ammonia also causes astrocyte swelling, alterations in energy and amino acid metabolism, and upregulation of the peripheral-type benzodiazepine receptor (for reviews, see refs. 333, 338). There is growing evidence that HE is associated with major derangements of glutamate neurotransmission resulting from an ammonia-induced alteration in glutamate metabolism.334,339 We have recently carried out studies on the effect of ammonia on glutamate uptake in cultured astrocytes.340 Chronic treatment (days) resulted in inhibition of glutamate uptake that was associated with a fall in the Vmax, suggesting that the number of glutamate transporters was decreased. Consistent with this finding is a fall in mRNA for the GLAST glutamate transporter in ammonia-treated astrocyte cultures.341 Treatment of mice with thioacetamide (a hepatotoxin) or with ammonia for 3 days caused a decrease in GLT-1 mRNA steady state levels in cerebral cortex and striatum.342 The mechanism for enhancement of GABAergic tone in HE is unclear. Our laboratory has been investigating the potential involvement of the “peripheral-type” benzodiazepine receptor (PBR). In contrast to the “central” benzodiazepine receptor that is present on the plasma membrane as part of the neuronal GABAA receptor complex, the PBR in the CNS is confined to glial cells343-345 where it is primarily located on the outer mitochondrial membrane.346 The best studied function of the PBR is the regulation of steroid biosynthesis.347 Neurosteroids, particularly tetrahydroprogesterone (THP, allopregnanolone), and tetrahydrodeoxycorticosterone (THDOC) have potent CNS depressant effects that are mediated through their actions on the GABAA receptor.348-350 Recent findings from our laboratory have shown that: the number of PBR binding sites is increased in ammonia-treated cultured astrocytes using PK 11195 as the PBR ligand; the PBR is upregulated in mice with acute HE caused by thioacetamide (TAA) as well as in hyperammonemic mice; treatment with PK 11195, a putative antagonist of the PBR, significantly attenuates ammonia toxicity in mice; pregnenolone levels are increased in TAA- and ammonia-treated animals; and brain levels of THP and THDOC are elevated in hyperammonemic mice and mice with acute liver failure produced by TAA.338,351 Upregulation of the astrocytic PBR by ammonia can potentially result in increased production of neurosteroids which have positive modulatory effects on the GABAA receptor, that in turn may lead to neuroinhibition and neurologic dysfunction.

Perspectives and Conclusions Astrocytes are active and dynamic cells involved in many aspects of CNS function and are early responders to CNS injury. They also contribute to the pathogenesis of many neurological conditions, although it is not always clear whether their effects are beneficial or detrimental. Such issues as the concentrations of various factors and the clinical setting may materially determine whether the astrocyte response is appropriate or not. Additionally, astrocytes have important interactions with microglia and other mesenchyme-derived cells that strongly influence the outcome of disease. It should also be stressed that astrocytes may be injured in disease states and thus astroglial functional failure may contribute to the pathogenesis of some CNS disorders. While astrocytes may occasionally exert deleterious actions, on balance their activity appears more geared towards generating homeostatic responses and promoting repair and

54

Astrocytes in Brain Aging and Neurodegeneration

regeneration. Indeed, the regulation of the astrocyte response may in future years provide a key strategy in influencing the outcome of CNS injury.

Acknowledgments This work was supported by research grants from the Department of Veterans Affairs and the National Institutes of Health (NS30291 and NS34951).

References 1. Abbott NJ, ed. Glial-Neuronal Interactions. New York: New York Acad Sci, 1991. 2. Gard AL. Astrocyte-oligodendrocyte interactions. In: Murphy S, ed. Astrocytes: Pharmacology and Function. San Diego: Academic Press, 1993:331-354. 3. Benveniste EN. Astrocyte-microglia interactions. In: Murphy S, ed. Astrocytes, Pharmacology and Function. New York: Academic Press, 1993:355-382. 4. Cancilla PA, Bready J, Berliner J. Astrocyte-endothelial interactions. In: Murphy S, ed. Astrocytes: Pharmacology and Function. San Diego: Academic Press, 1993:383-397. 5. Walz W. Role of glial cells in the regulation of the brain ion microenvironment. Prog Neurobiol 1989; 33:309-333. 6. Largo C, Cuevas P, Herreras O. Is glia dysfunction the initial cause of neuronal death in ischemic penumbra. Neurol.Res. 1996; 18:445-448. 7. Largo C, Cuevas P, Somjen GG et al. The effect of depressing glial function in rat brain in situ on ion homeostasis, synaptic transmission, and neuron survival. J Neurosci 1996; 16:1219-1229. 8. Schousboe A. Transport and metabolism of glutamate and GABA in neurons and glial cells. Int Rev Neurobiol 1981; 22:1-45. 9. Martin DL. Synthesis and release of neuroactive substances by glial cells. Glia 1992; 5:81-94. 10. Keyser DO, Pellmar TC. Synaptic transmission in the hippocampus: Critical role for glial cells. Glia 1994; 10:237-243. 11. Dringen R, Gebhardt R, Hamprecht B. Glycogen in astrocytes: Possible function as lactate supply for neighboring cells. Brain Res 1993; 623:208-214. 12. Westergaard N, Sonnewald U, Schousboe A. Metabolic trafficking between neurons and astrocytes: The glutamate glutamine cycle revisited. Dev Neurosci 1995; 17:203-211. 13. Muller HW, Junghans U, Kappler J. Astroglial neurotrophic and neurite promoting factors. Pharmacol Ther 1995; 65:1-18. 14. Norenberg MD. The role of astrocytes in hepatic encephalopathy. Neurochem Pathol 1987; 6:13-33. 15. Abramovitz M, Homma H, Ishigaki S et al. Characterization and localization of glutathione-S-transferases in rat brain and binding of hormones, neurotransmitters, and drugs. J Neurochem 1988; 50:50-57. 16. Makar TK, Nedergaard M, Preuss A et al. Vitamin E, ascorbate, glutathione, glutathione disulfide, and enzymes of glutathione metabolism in cultures of chick astrocytes and neurons: Evidence that astrocytes play an important role in antioxidative processes in the brain. J Neurochem 1994; 62:45-53. 17. Sawada J, Kikuchi Y, Shibutani M et al. Induction of metallothionein in astrocytes by cytokines and heavy metals. Biol Signals 1994; 3:157-168. 18. Janzer RC, Raff MC. Astrocytes induce blood-brain barrier properties in endothelial cells. Nature 1987; 325:253-257. 19. Stitt TN, Gasser UE, Hatten ME. Molecular mechanisms of glial-guided neuronal migration. Ann NY Acad Sci 1991; 633:113-121. 20. Frei K, Fontana A. Immune regulatory functions of astrocytes and microglial cells within the central nervous system. In: Goetzl EJ, Spector NH eds. Neuroimmune Networks: Physiology and Diseases. New York: Alan R. Liss, 1989:127-136. 21. Nicholls D, Attwell D. The release and uptake of excitatory amino acids. Trends Pharmacol Sci 1990; 11:462-468.

Astrocyte Pathophysiology in Disorders of the Central Nervous System

55

22. Rosenberg PA, Amin S, Leitner M. Glutamate uptake disguises neurotoxic potency of glutamate agonists in cerebral cortex in dissociated culture. J Neurosci 1992; 12:56-61. 23. Pines G, Danbolt NC, Bjoras M et al. Cloning and expression of a rat brain L-glutamate transporter. Nature 1992; 360:464-467. 24. Stork T, Schulte S, Hofmann K et al. Structure, expression and functional analysis of a Na+ dependent glutamate/aspartate transporter from rat brain. Proc Natl Acad Sci USA 1992; 89:10955-10959. 25. Kanai Y, Hediger MA. Primary structure and functional characterization of a high affinity glutamate transporter. Nature 1992; 360:467-471. 26. Torp R, Danbolt NC, Babaie E et al. Differential expression of two glial glutamate transporters in the rat brain: An in situ hybridization study. Eur J Neurosci 1994; 6:936-942. 27. Danbolt NC, Storm-Mathisen J, Kanner BI. An [Na+ + K+] coupled L-glutamate transporter purified from brain is located in glial cell processes. Neuroscience 1992; 51:295-310. 28. Levy LM, Lehre KP, Rolstad B et al. A monoclonal antibody raised against an [Na+ + K+]coupled L-glutamate transporter purified from rat brain confirms glial cell localization. FEBS Lett 1993; 317:79-84. 29. Rothstein JD, Martin L, Levey AI et al. Localization of neuronal and glial glutamate transporters. Neuron 1994; 13:713-725. 30. Kondo K, Sashimoto H, Kitanaka J et al. Expression of glutamate transporters in cultured glial cells. Neurosci Lett 1995; 188:140-142. 31. Nathaniel EJH, Nathaniel DR. The reactive astrocyte. Adv Cell Neurobiol 1981; 2:249-301. 32. Eng LF, Reier PJ, Houle JD. Astrocyte activation and fibrous gliosis: Glial fibrillary acidic protein in CNS tissue. Prog Brain Res 1987; 71:439-455. 33. Eddleston M, Mucke L. Molecular profile of reactive astrocytes - Implications for their role in neurologic disease. Neuroscience 1993; 54:15-36. 34. Norenberg MD. Reactive astrocytosis. In: Aschner M, Kimelberg HK, eds. The Role of Glia in Neurotoxicity. Boca Raton: CRC Press, 1996:93-107. 35. Kimelberg HK. Current concepts of brain edema. J Neurosurg 1995; 83:1051-1059. 36. Kimelberg HK, Goderie SK, Higman S et al. Swelling-induced release of glutamate, aspartate, and taurine from astrocyte cultures. J Neurosci 1990; 10:1583-1591. 37. Bender AS, Norenberg MD. Calcium dependence of hypoosmotically induced potassium release in cultured astrocytes. J Neurosci 1994; 14:4237-4243. 38. Smith SJ. Do astrocytes process neural information? In: Yu ACH, Hertz L, Norenberg MD, Sykova E, Waxman S, eds. Neuronal-Astrocytic Interactions: Implications for Normal and Pathological CNS Function. Amsterdam: Elsevier, 1992:119-136. 39. Garcia JH, Liu KF, Yoshida Y et al. The brain microvessels: Factors altering their patency after the occlusion of a middle cerebral artery (Wistar rat). Am J Pathol 1994; 145:1-13. 40. Kuriyama K. Taurine as an neuromodulator. Fed Proc 1980; 39:2680-2684. 41. Schurr A, Tseng MT, West CA et al. Taurine improves the recovery of neuronal function following cerebral hypoxia: An in vitro study. Life Sci 1987; 40:2059-2066. 42. Aruoma OI, Halliwell B, Hoey BY et al. The antioxidant action of taurine, hypotaurine and their metabolic precursors. Biochem J 1988; 256:251-255. 43. Norenberg MD. The astrocyte in liver disease. In: Fedoroff S, Hertz L, eds. Advances in Cellular Neurobiology, Vol. 2. New York: Academic Press, 1981:303-352. 44. Norenberg MD. A light and electron microscopic study of experimental portal-systemic (ammonia) encephalopathy. Lab Invest 1977; 36:618-627. 45. Faden AI, Salzman S. Pharmacological strategies in CNS trauma. Trends Neurosci 1992; 13:29-35. 46. Orkand RK, Nicholls JG, Kuffler SW. Effects of nerve impulses on the membrane potential of glial cells in the central nervous system of amphibia. J Neurophysiol 1966; 29:788-806. 47. Chesler M, Kraig RP. Intracellular pH transients of mammalian astrocytes. J Neurosci 1989; 9:2011-2019. 48. Kraig RP, Pulsinelli WA, Plum F. Hydrogen ion buffering during complete brain ischemia. Brain Res 1985; 342:281-290.

56

Astrocytes in Brain Aging and Neurodegeneration

49. Barbour B, Brew H, Attwell D. Electrogenic uptake of glutamate and aspartate into glial cells isolated from the salamander (Ambystoma) retina. J Physiol (Lond) 1991; 436:169-193. 50. Cambray-Deakin M, Pearce B, Morrow C et al. Effects of extracellular potassium on glycogen stores of astrocytes in vitro. J Neurochem 1988; 51:1846-1851. 51. Subbarao KV, Stolzenburg JU, Hertz L. Pharmacological characteristics of potassium-induced, glycogenolysis in astrocytes. Neurosci Lett 1995; 196:45-48. 52. Kimelberg HK, Ransom BR. Physiological aspects of astrocyte swelling. In: Fedoroff S, Vernadakis A, eds. Astrocytes, Vol. 3. Orlando: Academic Press, 1986:129-166. 53. Choi DW. Glutamate neurotoxicity and diseases of the nervous system. Neuron 1988; 1:623-634. 54. Whetsell WO, Shapira NA. Biology of disease: Neuroexcitation, excitotoxicity and human neurological disease. Lab Invest 1993; 68:372-387. 55. Teichberg VI. Glial glutamate recepors: likely actors in brain signaling. FASEB J 1991; 5:3086-3091. 56. Hansson E, Rönnbäck L. Astrocytes in glutamate neurotransmission. FASEB J 1995; 9:343-350. 57. Norenberg MD, Mozes LW, Gregorios JB et al. Effects of lactic acid on astrocytes in primary culture. J Neuropathol Exp Neurol 1987; 46:154-166. 58. Giffard RG, Monyer H, Choi DW. Selective vulnerability of cultured glia to injury by extracellular acidosis. Brain Res 1990; 530:138-141. 59. Jakubovicz D, Klip A. Lactic acid-induced swelling in C6 glial cells via Na+/H+ exchange. Brain Res 1989; 485:215-224. 60. Staub F, Peters J, Kempski O et al. Swelling of glial cells in lactacidosis and by glutamate: Significance of Cl--transport. Brain Res 1993; 610:69-74. 61. Swanson RA, Farrell K, Simon RP. Acidosis causes failure of astrocyte glutamate uptake during hypoxia. J Cereb Blood Flow Metab 1995; 15:417-424. 62. Bender AS, Young LP, Norenberg MD. Effect of lactic acid on L-glutamate uptake in cultured astrocytes: Mechanistic considerations. Brain Res 1997; 750:59-66. 63. Chan PH, Longar S, Chen S et al. The role of arachidonic acid and oxygen radical metabolites in the pathogenesis of vasogenic brain edema and astrocytic swelling. Ann NY Acad Sci 1989; 559:237-247. 64. Staub F, Winkler A, Peters J et al. Swelling, acidosis, and irreversible damage of glial cells from exposure to arachidonic acid in vitro. J Cereb Blood Flow Metab 1994; 14:1030-1039. 65. Piani D, Frei K, Pfister H-W et al. Glutamate uptake by astrocytes is inhibited by reactive oxygen intermediates but not by other macrophage-derived molecules including cytokines, leukotrienes or platelet-activating factor. J Neuroimmunol 1993; 48:99-104. 66. Volterra A, Trotti D, Tromba C et al. Glutamate uptake inhibition by oxygen free radicals in rat cortical astrocytes. J Neurosci 1994; 14:2924-2932. 67. Levi G, Patrizio M. Astrocyte heterogeneity: Endogenous amino acid levels and release evoked by non-N-methyl-D-aspartate receptor agonists and by potassium-induced swelling in type-1 and type-2 astrocytes. J Neurochem 1992; 58:1943-1952. 68. Ogata T, Nakamura Y, Shibata T et al. Release of excitatory amino acids from cultured hippocampal astrocytes induced by a hypoxic-hypoglycemic stimulation. J Neurochem 1992; 58:1957-1959. 69. Gemba T, Oshima T, Ninomiya M. Glutamate efflux via the reversal of the sodium-dependent glutamate transporter caused by glycolytic inhibition in rat cultured astrocytes. Neuroscience 1994; 63:789-795. 70. Longuemare MC, Swanson RA. Excitatory amino acid release from astrocytes during energy failure by reversal of sodium-dependent uptake. J Neurosci Res 1995; 40:379-386. 71. Benos DJ, Hahn BH, Bubien JK et al. Envelope glycoprotein gp120 of human immunodeficiency virus type 1 alters ion transport in astrocytes: Implications for AIDS dementia complex. Proc Natl Acad Sci USA 1994; 91:494-498. 72. Mullaney KJ, Vitarella D, Albrecht J et al. Stimulation of D-aspartate efflux by mercuric chloride from rat primary astrocyte cultures. Devl Brain Res 1993; 75:261-268.

Astrocyte Pathophysiology in Disorders of the Central Nervous System

57

73. Dawson R, Jr., Patterson TA, Eppler B. Endogenous excitatory amino acid release from brain slices and astrocyte cultures evoked by trimethyltin and other neurotoxic agents. Neurochem Res 1995; 20:847-858. 74. Albrecht J, Simmons M, Dutton GR et al. Aluminum chloride stimulates the release of endogenous glutamate, taurine and adenosine from cultured rat cortical astrocytes. Neurosci Lett 1991; 127:105-107. 75. Schwarcz R, Guidetti P, Roberts RC. Quinolinic acid and kynurenic acid: Glia-derived modulators of excitotoxic brain injury. In: Aschner M, Kimelberg HK, eds. The Role of Glia in Neurotoxicity. Boca Raton: CRC Press, 1996:245-262. 76. Heyes MP, Achim CL, Wiley CA et al. Human microglia convert L-tryptophan into the neurotoxin quinolinic acid. Biochem J 1996; 320:595-597. 77. Speciale C, Okuno E, Schwarcz R. Increased quinolinic acid metabolism following neuronal degeneration in the rat hippocampus. Brain Res 1987; 436:18-24. 78. Perkins MN, Stone TW. An iontophoretic investigation of the actions of convulsant kynurenines and their interaction with the endogenous excitant quinolinic acid. Brain Res 1982; 247:183-187. 79. Holopainen I, Kontro P. Uptake and release of glycine in cerebellar granule cells and astrocytes in primary culture: Potassium-stimulated release from granule cells is calciumdependent. J Neurosci Res 1989; 24:374-383. 80. Sato K, Yoshida S, Fujiwara K et al. Glycine cleavage system in astrocytes. Brain Res 1991; 567:64-70. 81. Waniewski RA, Martin DL. Exogenous glutamate is metabolized to glutamine and exported by rat primary astrocyte cultures. J Neurochem 1986; 47:304-313. 82. Farinelli SE, Nicklas WJ. Glutamate metabolism in rat cortical astrocyte cultures. J Neurochem 1992; 58:1905-1915. 83. Simantov R. Glutamate neurotoxicity in culture depends on the presence of glutamine: Implications for the role of glial cells in normal and pathological brain development. J Neurochem 1989; 52:1694-1699. 84. Bradford HF, Ward HK. Glutamine--a major substrate for nerve endings. J Neurochem 1978; 30:1453-1459. 85. Okuno E, Nakamura M, Schwarcz R. Two kynurenine aminotransferases in human brain. Brain Res 1991; 542:307-312. 86. Norenberg MD, Martinez-Hernandez A. Fine structural localization of glutamine synthetase in astrocytes of rat brain. Brain Res 1979; 161:303-310. 87. Swanson RA, Shiraishi K, Morton MT et al. Methionine sulfoximine reduces cortical infarct size in rats after middle cerebral artery occlusion. Stroke 1990; 21:322-327. 88. Cataldo AM, Broadwell RD. Cytochemical identification of cerebral glycogen and glucose-6-phosphatase activity under normal and experimental conditions. Neurons and glia. Electron Microscop Tech 1986; 3:413-437. 89. Siesjö BK. Acidosis and ischemic brain damage. Neurochem Pathol 1988; 9:31. 90. Staub F, Baethmann A, Peters J et al. Efects of lactacidosis on glial cell volume and viability. J Cereb Blood Flow Metab 1990; 10:866-876. 91. Poitry-Yamate CL, Poitry S, Tsacopoulos M. Lactate released by Müller glial cells is metabolized by photoreceptors from mammalian retina. J Neurosci 1995; 15:5179-5191. 92. Giffard RG, Monyer H, Christine CW et al. Acidosis reduces NMDA receptor activation, glutamate neurotoxicity, and oxygen-glucose deprivation neuronal injury in cortical cultures. Brain Res 1990; 506:339-342. 93. Moncada S, Palmer RM, Higgs EA. Nitric oxide: Physiology, pathophysiology, and pharmacology. Pharmacol Rev 1991; 43:109-142. 94. Dawson TM, Dawson VL, Snyder SH. A novel neuronal messenger molecule in brain: The free radical, nitric oxide. Ann Neurol 1992; 32:297-311. 95. Lee SC, Dickson DW, Brosnan CF et al. Human astrocytes inhibit Cryptococcus neoformans growth by a nitric oxide-mediated mechanism. J Exp Med 1994; 180:365-369. 96. Peterson PK, Gekker G, Hu S et al. Human astrocytes inhibit intracellular multiplication of Toxoplasma gondii by a nitric oxide-mediated mechanism. J Infect Dis 1995; 171:516-518.

58

Astrocytes in Brain Aging and Neurodegeneration

97. Beckman JS. The double-edged role of nitric oxide in brain function and superoxide-mediated injury. J Dev Physiol 1991; 15:53-59. 98. Murphy S, Simmons ML, Agullo L et al. Synthesis of nitric oxide in CNS glial cells. Trends Neurosci 1993; 16:323-328. 99. Brosnan CF, Battistini L, Raine CS et al. Reactive nitrogen intermediates in human neuropathology: An overview. Dev Neurosci 1994; 16:152-161. 100. Goadsby P, Kaube H, Hoskin KL. Nitric oxide synthesis couples cerebral blood flow and metabolism. Brain Res 1992; 595:167-170. 101. Schipper HM. Astrocytes, brain aging, and neurodegeneration. Neurobiol Aging 1996; 17:467-480. 102. Pindzola RR, Doller C, Silver J. Putative inhibitory extracellular matrix molecules at the dorsal root entry zone of the spinal cord during development and after root and sciatic nerve lesions. Dev Biol 1993; 156:34-48. 103. DeWitt DA, Richey PL, Praprotnik D et al. Chondroitin sulfate proteoglycans are a common component of neuronal inclusions and astrocytic reaction in neurodegenerative diseases. Brain Res 1994; 656:205-209. 104. Letourneau PC, Condic ML, Snow DM. Extracellular matrix and neurite outgrowth. Curr Opin Genet Devel 1992; 2:625-634. 105. Faissner A, Steindler DA. Boundries and inhibitory molecules in developing neural tissue. Glia 1995; 13:233-254. 106. Laywell ED, Dörries U, Bartsch U et al. Enhanced expression of the developmentally regulated extracellular matrix molecule tenascin following adult brain injury. Proc Natl Acad Sci USA 1992; 89:2634-2638. 107. Scheffler B, Faissner A, Beck H et al. Hippocampal loss of tenascin boundaries in Ammon’s horn sclerosis. Glia 1997; 19:35-46. 108. Mansour H, Asher R, Dahl D et al. Permissive and non-permissive reactive astrocytes: Immunofluorescence study with antibodies to the glial hyaluronate-binding protein. J Neurosci Res 1990; 25:300-311. 109. Tiveron M-C, Barboni E, Pliego Rivero FB et al. Selective inhibition of neurite outgrowth on mature astrocytes by Thy-1 glycoprotein. Nature 1992; 355:745-748. 110. Liuzzi FJ. Proteolysis is a critical step in the physiological stop pathway: Mechanisms involved in the blockade of axonal regeneration by mammalian astrocytes. Brain Res 1990; 512:277-283. 111. Perry VH, Andersson P-B, Gordon S. Macrophages and inflammation in the central nervous system. Trends Neurosci 1993; 16:268. 112. Morganti-Kossmann MC, Kossmann T, Wahl SM. Cytokines and neuropathology. Trends Pharmacol Sci 1992; 13:286-291. 113. Benveniste EN. Inflammatory cytokines within the central nervous system: Sources, function, and mechanism of action. Am J Physiol Cell Physiol 1992; 263:C1-C16. 114. Furie MB, Randolph GJ. Chemokines and tissue injury. Am J Pathol 1995; 146:1287-1301. 115. Goldman JE, Chiu FC. DbcAMP causes intermediate filament accumulation and actin re-organization in astrocytes. Brain Res 1984; 306:85-95. 116. Faden AI et al. Experimental spinal cord injury: Effects of trauma or ischemia on TRH and muscarinic receptors. Neurology 1986; 36:723-726. 117. Ghirnikar RS, Lee YL, He TR et al. Chemokine expression in rat stab wound brain injury. J Neurosci Res 1996; 46:727-733. 118. Hickey WF, Osborn JP, Kirby WM. Expression of Ia molecules by astrocytes during acute experimental allergic encephalomyelitis in the Lewis rat. Cell Immunol 1985; 91:528-535. 119. Sun D, Wekerle H. Ia-restricted encephalitogenic T lymphocytes mediating EAE lyse autoantigen-presenting astrocytes. Nature 1986; 320:70-72. 120. Lee SC, Moore GR, Golenwsky G et al. Multiple sclerosis: A role for astroglia in active demyelination suggested by class II MHC expression and ultrastructural study. J Neuropath Exp Neurol 1990; 49:122-136.

Astrocyte Pathophysiology in Disorders of the Central Nervous System

59

121. Cannella B, Cross AH, Raine CS. Adhesion-related molecules in the central nervous system: Upregulation correlates with inflammatory cell influx during relapsing experimental autoimmune encephalomyelitis. Lab Invest 1991; 65:23-31. 122. Vass K, Lassmann H, Wekerle H et al. The distribution of Ia antigen in the lesions of rat acute experimental allergic encephalomyelitis. Acta Neuropathol 1986; 70:149-160. 123. Matsumoto Y, Kawai K, Fujiwara M. In situ Ia expression on brain cells in the rat: Autoimmune encephalomyelitis-resistant strain (BN) and susceptible strain (Lewis) compared. Immunology 1989; 66:621-627. 124. Abraham CR, Kanemaru K, Mucke L. Expression of cathepsin G-like and alpha1antichymotrypsin-like proteins in reactive astrocytes. Brain Res 1993; 621:222-232. 125. Monard D. Cell-derived proteases and protease inhibitors as regulators of neurite outgrowth. Trends Neurosci 1988; 11:541-544. 126. Neveu I, Jehan F, Jandrot Perrus M et al. Enhancement of the synthesis and secretion of nerve growth factor in primary cultures of glial cells by proteases: A possible involvement of thrombin. J Neurochem 1993; 60:858-867. 127. Eddleston M, De la Torre JC, Oldstone MB et al. Astrocytes are the primary source of tissue factor in the murine central nervous system. A role for astrocytes in cerebral hemostasis. J Clin Invest 1993; 92:349-358. 128. Abraham CR, Driscoll J, Potter H et al. A calcium-activated protease from Alzheimer’s disease brain cleaves at the N-terminus of the amyloid beta-protein. Biochem Biophys Res Commun 1991; 174:790-796. 129. Wells GMA, Catlin G, Cossins JA et al. Quantitation of matrix metalloproteinases in cultured rat astrocytes using the polymerase chain reaction with a multi-competitor cDNA standard. Glia 1996; 18:332-340. 130. Deb S, Gottschall PE. Increased production of matrix metalloproteinases in enriched astrocyte and mixed hippocampal cultures treated with β-amyloid peptides. J Neurochem 1996; 66:1641-1647. 131. Rosenberg GA. Matrix metalloproteinases in brain injury. J Neurotrauma 1995; 12:833-842. 132. Gebicke-Haerter PJ, Bauer J, Brenner A et al. Alpha2-Macroglobulin synthesis in an astrocyte subpopulation. J Neurochem 1987; 49:1139-1145. 133. Higuchi M, Ito T, Imai Y et al. Expression of the α2-macroglobulin-encoding gene in rat brain and cultured astrocytes. Gene 1994; 141:155-162. 134. Reichardt LF, Tomaselli KJ. Extracellular matrix molecules and their receptors: Functions in neural development. Annu Rev Neurosci 1991; 14:531-570. 135. Frohman EM, Frohman TC, Dustin ML et al. The induction of intracellular adhesion molecule 1 (ICAM-1) expression on human fetal astrocytes by interferon-gamma, tumor necrosis factor-alpha, lymphotoxin, and interleukin-1: Relevance to intracerebral antigen presentation. J Neuroimmunol 1989; 23:117-124. 136. Sobel RA, Mitchell ME, Fondren G. Intracellular adhesion molecule-1 (ICAM-1) in cellular immune reactions in the human central nervous system. Am J Pathol 1990; 136:1309-1316. 137. Hurwitz AA, Lyman WD, Guida MP et al. Tumor necrosis factor alpha induces adhesion molecule expression on human fetal astrocytes. J Exp Med 1992; 176:1631-1636. 138. Vaca K, Wendt E. Divergent effects of astroglial and microglial secretions on neuron growth and survival. Exp Neurol 1992; 118:62-72. 139. Giulian D, Vaca K, Corpuz M. Brain glia release factors with opposing actions upon neuronal survival. J Neurosci 1993; 13:29-37. 140. Graeber MB, Streit WJ. Microglia: Immune network in the CNS. Brain Pathol 1990; 1:2-5. 141. Kreutzberg GW. Microglia, the first line of defence in brain pathologies. Arzneimittelforschung 1995; 45:357-360. 142. Thomas WE. Brain macrophages: Evaluation of microglia and their functions. Brain Res Rev 1992; 17:61-74. 143. Giulian D, Chen J, Ingeman JE et al. The role of mononuclear phagocytes in wound healing after traumatic injury to adult mammalian brain. J Neurosci 1989; 9:4416-4429.

60

Astrocytes in Brain Aging and Neurodegeneration

144. Gadient RA, Cron KVC, Otten U. Interleukin-1 and tumor necrosis factor-alpha synergistically stimulate nerve growth factor (NGF) release from cultured astrocytes. Neurosci Lett 1990; 117:335-340. 145. Gage FH, Olejniczak P, Armstrong DM. Astrocytes are important for sprouting in the septohippocampal circuit. Exp Neurol 1988; 102:2-13. 146. Hao C, Guilbert LJ, Fedoroff S. Production of colony-stimulating factor-1 (CSF-1) by mouse astroglia in vitro. J Neurosci Res 1990; 27:314. 147. Aloisi F, Carè A, Borsellino G et al. Production of hemolymphopoietic cytokines (IL-6, IL-8, colony-stimulating factors) by normal human astrocytes in response to IL-1beta and tumor necrosis factor-alpha. J Immunol 1992; 149:2358-2366. 148. Garland JM. Colony stimulating factors. In: Thomson A, ed. The Cytokine Handbook. San Diego: Academic Press, 1991:269-300. 149. Tweardy DJ, Mott PL, Glazer EW. Monokine modulation of human astroglial cell production of granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor. I. Effects of IL-1alpha and IL-beta. J Immunol 1990; 144:2233-2241. 150. Frei K, Bodmer S, Schwerdel C et al. Astrocytes of the brain synthesize interleukin 3-like factors. J Immunol 1985; 135:4044-4047. 151. Chamak B, Mallat M. Fibronectin and laminin regulate the in vitro differentiation of microglial cells. Neuroscience 1991; 45:513-527. 152. Levitt D, Ho PL. Induction of chondroitin sulfate proteoglycan synthesis and secretion in lymphocytes and monocytes. J Cell Biol 1983; 97:351-358. 153. Johnson-Green PC, Dow KE, Riopelle RJ. Neurite growth modulation associated with astrocyte proteoglycans: Influence of activators of inflammation. Glia 1992; 5:33. 154. Ness R, David S. Leptomeningeal cells modulate the neurite growth promoting properties of astrocytes in vitro. Glia 1997; 19:47-57. 155. Schwab ME, Bartholdi D. Degeneration and regeneration of axons in the lesioned spinal cord. Physiol Rev 1996; 76:319-370. 156. Silos-Santiago I, Greenlund LJ, Johnson EM Jr et al. Molecular genetics of neuronal survival. Curr Opin Neurobiol 1995; 5:42-49. 157. Ramon y Cajal S. Degeneration and Regeneration of the Nervous System. Oxford, Oxford University Press, 1928. 158. Rudge JS, Silver J. Inhibition of neurite outgrowth on astroglial scars in vitro. J Neurosci 1990; 10:3595-3603. 159. Bovolenta P, Wandosell F, Nieto-Sampedro M. CNS glial scar tissue: A source of molecules which inhibit central neurite outgrowth. In: Yu ACH, Hertz L, Norenberg MD, Sykova E, Waxman SG, eds. Neuronal-Astrocytic Interactions: Pathological Implications. Amsterdam: Elsevier, 1992:367-379. 160. Stensaas LJ, Partlow LM, Burgess PR et al. Inhibition of regeneration: the ultrastructure of reactive astrocytes and abortive axon terminals in the transition zone of the dorsal root. Prog Brain Res 1987; 71:457-468. 161. Carlstedt T, Cullheim S, Risling M et al. Nerve fiber regeneration across the PNS-CNS interface at the root-spinal cord junction. Brain Res Bull 1989; 22:93-102. 162. Carlstedt T. Regenerating axons form nerve terminals at astrocytes. Brain Res 1985; 347:188-191. 163. Liuzzi FJ, Lasek RJ. Astrocytes block axonal regeneration in mammals by activating the physiological stop pathway. Science 1987; 237:642-645. 164. Noble M, Fok-Seang J, Cohen J. Glia are a unique substrate for the in vitro growth of central nervous system neurons. J Neurosci 1984; 4:1892-1903. 165. Fallon JR. Preferential outgrowth of central nervous system neurites on astrocytes and Schwann cells as compared with nonglial cells in vitro. J Cell Biol 1985; 100:198-207. 166. Rudge JS. Astrocyte-derived neurotrophic factors. In: Murphy S, ed. Astrocytes. Pharmacology and Function. San Diego: Academic Press, 1993:267-305. 167. Fawcett JW, Housden E, Smith-Thomas L et al. The growth of axons in three-dimensional astrocyte cultures. Dev Biol 1989; 135:449-458.

Astrocyte Pathophysiology in Disorders of the Central Nervous System

61

168. Smith GM, Rutishauser U, Silver J et al. Maturation of astrocytes in vitro alters the extent and molecular basis of neurite outgrowth. Dev Biol 1990; 138:377-390. 169. Schwartz M, Cohen A, Stein-Izsak C et al. Dichotomy of the glial cell response to axonal injury and regeneration. FASEB J 1989; 3:2371-2378. 170. Kruger S, Sievers J, Hansen C et al. Three morphologically distinct types of interface develop between adult host and fetal brain transplants: Implications for scar formation in the adult central nervous system. J Comp Neurol 1986; 249:103-116. 171. Berry M, Maxwell WL, Logan A et al. Deposition of scar tissue in the central nervous system. Acta Neurochirurg Suppl 1983; 32:31-53. 172. Caroni P, Schwab ME. Antibody against myelin-associated inhibitor of neurite outgrowth neutralizes non-permissive substrate properties of CNS white matter. Neuron 1988; 1:85-96. 173. Savio T, Schwab ME. Lesioned corticospinal tract axons regenerate in myelin-free rat spinal cord. Proc Natl Acad Sci USA 1990; 87:4130-4133. 174. Petito CK, Cho E-L, Lemann W et al. Neuropathology of acquired immunodeficiency syndrome (AIDS): an autopsy review. J Neuropathol Exp Neurol 1986; 45:635-646. 175. Gray F, Gherardi R, Scaravilli F. The neuropathology of the acquired immune deficiency syndrome (AIDS). A review. Brain 1988; 111:245-266. 176. Epstein LG, Gendelman HE. Human immunodeficiency virus type-1 infection of the nervous system: Pathogenetic mechanisms. Ann Neurol 1993; 33:429-436. 177. Genis P, Jett M, Bernton EW et al. Cytokines and arachidonic metabolites produced during human immunodeficiency virus (HIV)-infected macrophage-astroglia interactions: Implications for the neuropathogenesis of HIV disease. J Exp Med 1992; 176:1703-1718. 178. Nottet HSLM, Jett M, Flanagan CR et al. A regulatory role for astrocytes in HIV-1 encephalitis: An overexpression of eicosanoids, platelet-activating factor, and tumor necrosis factor-α by activated HIV-1-infected monocytes is attenuated by primary human astrocytes. J Immunol 1995; 154:3567-3581. 179. Vitkovic L, Kalebic T, de Cunha A et al. Astrocyte-conditioned media stimulates HIV-1 expression in chronically infected promonocyte clone. J Neuroimmunol 1990; 30:153. 180. Vitkovic L, Wood GP, Major EO et al. Human astrocytes stimulate HIV-1 expression in a chronically infected promonocyte clone via interleukin-6. AIDS Res Human Retroviruses 1991; 7:723-727. 181. Wahl SM, Allen JB, McCartney-Francis N et al. Macrophage- and astrocyte-derived transforming growth factor beta as a mediator of central nervous system dysfunction in acquired immune deficiency syndrome. J Exp Med 1991; 173:981-991. 182. Saito Y, Sharer LR, Epstein LG et al. Overexpression of nef as a marker for restricted HIV-1 infection of astrocytes in postmortem pediatric central nervous tissues. Neurology 1994; 44:474-481. 183. Tornatore C, Chandra R, Berger JR et al. HIV-1 infection of subcortical astrocytes in the pediatric central nervous system. Neurology 1994; 44:481-487. 184. Nath A, Hartloper V, Furer M et al. Infection of human fetal astrocytes with HIV-1: Viral tropism and the role of cell to cell contact in viral transmission. J Neuropath Exp Neurol 1995; 54:320-329. 185. Pulliam L, West D, Haigwood N et al. HIV-1 envelope gp120 alters astrocytes in human brain cultures. AIDS Res Hum Retroviruses 1993; 9:439-444. 186. Levi G, Patrizio M, Bernardo A et al. Human immunodeficiency virus coat protein gp120 inhibits the beta-adrenergic regulation of astroglial and microglial functions. Proc Natl Acad Sci USA 1993; 90:1541-1545. 187. Lipton SA. Models of neuronal injury in AIDS: Another role for the NMDA receptor? Trends Neurosci 1992; 15:75-79. 188. Benos DJ, McPherson S, Hahn BH et al. Cytokines and HIV envelope glycoprotein gp120 stimulate Na+/H+ exchange in astrocytes. J Biol Chem 1994; 269:13811-13816. 189. Dreyer EB, Lipton SA. The coat protein gp120 of HIV-1 inhibits astrocyte uptake of excitatory amino acids via macrophage arachidonic acid. Eur J Neurosci 1995; 7:2502-2507.

62

Astrocytes in Brain Aging and Neurodegeneration

190. Yu ACH, Chan PH, Fishman RA. Arachidonic acid inhibits uptake of glutamate and glutamine but not of GABA in cultured cerebellar granule cells. J Neurosci Res 1987; 17:424-427. 191. Barbour B, Szatkowski M, Ingledew N et al. Arachidonic acid induces a prolonged inhibition of glutamate uptake into glial cells. Nature 1989; 342:918-920. 192. Volterra A, Trotti D, Cassutti P et al. High sensitivity of glutamate uptake to extracellular free arachidonic acid levels in rat cortical synaptosomes and astrocytes. J Neurochem 1992; 59:600-606. 193. Heyes MP, Brew B, Martin A et al. Cerebrospinal fluid quinolinic acid concentrations are increased in acquired immune deficiency syndrome. Adv Exp Med Biol 1991; 294:687-690. 194. Merrill JE, Chen ISY. HIV-1, macrophages, glial cells, and cytokines in AIDS nervous system disease. FASEB J 1991; 5:2391-2397. 195. Blumberg BM, Gelbard HA, Epstein LG. HIV-1 infection of the developing nervous system: Central role of astrocytes in pathogenesis. Virus Res 1994; 32:253-267. 196. Patton HK, Benveniste EN, Benos DJ. Astrocytes and the AIDS dementia complex. J Neuro-AIDS 1996; 1:111-131. 197. Raine CS. Demyelinating diseases. In: Davis RL, Robertson DM, eds. Textbook of Neuropathology. Baltimore: Williams and Wilkins, 1997:627-714. 198. Raine CS, Bornstein MB. EAE: A light and electron microscope study of remyelination and “sclerosis” in vitro. J Neuropathol Exp Neurol 1970; 29:552-574. 199. Vick RS, Neuberger TJ, DeVries GH. Role of adult oligodendrocytes in remyelination after neural injury. J Neurotrauma 1992; 9 Suppl 1:S93-103. 200. Blakemore WF. Remyelination by Schwann cells of axons demylinated by intraspinal injection of 6-aminonicotinamide in the rat. J Neurocytol 1975; 4:745-757. 201. Franklin RJM, Crang AJ, Blakemore WF. The role of astrocytes in the remyelination of glia-free areas of demyelination. Adv Neurol 1993; 59:125-133. 202. Toshniwal P, Tiku ML. Astrocytes can degrade myelin basic protein. Neurology 1986; 36:146-147. 203. Maeda A, Sobel RA. Matrix metalloproteinases in the normal human central nervous system, microglial nodules, and multiple sclerosis lesions. J Neuropathol Exp Neurol 1996; 55:300-309. 204. Kostyk SK, Kowall NW, Hause SL. Substance P immunoreactive astrocytes are present in multiple sclerosis plaques. Brain Res 1989; 504:284-288. 205. May PC, Lampert-Etchells M, Johnson SA et al. Dynamics of gene expression for a hippocampal glycoprotein elevated in Alzheimer’s disease and in response to experimental lesions in rat. Neuron 1990; 5:831-839. 206. Wekerle H, Linington C, Lassmann H. Cellular immune reactivity within the CNS. Trends Neurosci 1986; 9:271-277. 207. Goldmuntz EA, Brosnan CF, Chiu FC et al. Astrocytic reactivity and intermediate filament metabolism in experimental autoimmune encephalomyelitis: The effect of suppression with prazosin. Brain Res 1986; 397:16-26. 208. Cammer W, Tansey FA, Brosnan CF. Reactive gliosis in the brains of Lewis rats with experimental allergic encephalomyelitis. J Neuroimmunol 1990; 27:111-120. 209. Chung IY, Norris JG, Benveniste EN. Differential tumor necrosis factor alpha expression by astrocytes from experimental allergic encephalomyelitis-susceptible and -resistant rat strains. J Exp Med 1991; 173:801-811. 210. Merrill JE, Kono DH, Clayton J et al. Inflammatory leukocytes and cytokines in the peptide-induced disease of experimental allergic encephalomyelitis in SJL and B10.PL mice. Proc Natl Acad Sci USA 1992; 89:574-578. 211. Ransohoff RM, Hamilton TA, Tani M et al. Astrocyte expression of mRNA encoding cytokines IP-10 and JE/MCP-1 in experimental autoimmune encephalomyelitis. FASEB J 1993; 7:592-600. 212. Glabinski AR, Tani M, Aras S et al. Regulation and function of central nervous system chemokines. Int J Dev Neurosci 1995; 13:153-165.

Astrocyte Pathophysiology in Disorders of the Central Nervous System

63

213. Frohman EM, Vayuvegula B, Gupta S et al. Norepinephrine inhibits gamma interferoninduced major histocompatability class II (Ia) antigen expression on cultured astrocytes via beta2-adrenergic signal transduction mechanisms. Proc Natl Acad Sci USA 1988; 85:1292-1296. 214. Frohman E, Frohman T, Vayuvegula B et al. Vasoactive intestinal polypeptide inhibits the expression of the MHC class II antigens on astrocytes. J Neurol Sci 1988; 88:339-346. 215. Massa PT, ter Meulen V, Fontana A. Hyperinductibility of Ia antigen on astrocytes correlates with strain-specific susceptibility to experimental autoimmune encephalomyelitis. Proc Natl Acad Sci USA 1987; 84:4219-4223. 216. Oh LYS, Yong VW. Astrocytes promote process outgrowth by adult human oligodendrocytes in vitro through interaction between bFGF and astrocyte extracellular matrix. Glia 1996; 17:237-253. 217. García-Segura LM, Pérez J, Pons S et al. Localization of insulin-like growth factor I (IGF-I)-like immunoreactivity in the developing and adult rat brain. Brain Res 1991; 560:167-174. 218. Garcia-Estrada J, Garcia-Segura LM, Torres-Aleman I. Expression of insulin-like growth factor I by astrocytes in response to injury. Brain Res 1992; 592:343-347. 219. Gluckman P, Klempt N, Guan J et al. A role for IGF-1 in the rescue of CNS neurons following hypoxic-ischemic injury. Biochem Biophys Res Commun 1992; 182:593-599. 220. Mozell RL, McMorris FA. Insulin-like growth factor I stimulates oligodendrocyte development and myelination in rat brain aggregate cultures. J Neurosci Res 1991; 30:382-390. 221. Schecter R, Yen S-HC, Terry RD. Fibrous astrocytes in senile dementia of the Alzheimer type. J Neuropathol Exp Neurol 1981; 40:95-101. 222. Mancardi GL, Liwnicz BH, Mandybur TI. Fibrous astrocytes in Alzheimer’s disease and senile dementia of Alzheimer type. Acta Neuropathol 1983; 61:76-80. 223. Delacourte A. General and dramatic glial reaction in Alzheimer brains. Neurology 1990; 40:33-37. 224. Jorgensen OS, Brooksbank BWL, Balazs R. Neuronal plasticity and astrocytic reaction in Down syndrome and Alzheimer’s disease. J Neurol Sci 1990; 98:63-79. 225. Le Prince G, Delaere P, Fages C et al. Alterations in glial fibrillary acidic protein mRNA level in the aging brain and in senile dementia of the Alzheimer type. Neurosci Lett 1993; 151:71-73. 226. Mandybur TI, Chuirazzi CC. Astrocytes and the plaques of Alzheimer’s disease. Neurology 1990; 40:635-639. 227. LeBlanc AC. The role of β-amyloid peptide in Alzheimer’s disease. Metab Brain Dis 1994; 9:3-31. 228. Dickson DW. The pathogenesis of senile plaques. J Neuropathol Exp Neurol 1997; 56:321-339. 229. Price DL, Borchelt DR, Walker LC et al. Toxicity of synthetic A beta peptides and modeling of Alzheimer’s disease. Neurobiol Aging 1992; 13:623-625. 230. Pike CJ, Burdick D, Walencewicz AJ et al. Neurodegeneration induced by β-amyloid peptides in vitro: The role of peptide assembly state. J Neurosci 1993; 13:1676-1687. 231. Meda L, Cassatella MA, Szendrei GI et al. Activation of microglial cells by β-amyloid protein and interferon-gamma. Nature 1995; 374:647-650. 232. Lorton D, Kocsis JM, King L et al. β-Amyloid induces increased release of interleukin-1 beta from lipopolysaccharide-activated human monocytes. J Neuroimmunol 1996; 67:21-29. 233. Giulian D, Haverkamp LJ, Yu JH et al. Specific domains of β-amyloid from Alzheimer plaque elicit neuron killing in human microglia. J Neurosci 1996; 16:6021-6037. 234. Pike CJ, Cummings BJ, Monzavi R et al. β-amyloid-induced changes in cultured astrocytes parallel reactive astrocytosis associated with senile plaques in Alzheimer’s disease. Neuroscience 1994; 63:517-531. 235. Canning DR, McKeon RJ, DeWitt DA et al. β-Amyloid of Alzheimer’s disease induces reactive gliosis that inhibits axonal outgrowth. Exp Neurol 1993; 124:289-298. 236. Pasternack JM, Abraham CR, Van Dyke BJ et al. Astrocytes in Alzheimer’s disease gray matter express α1-antichymotrypsin mRNA. Am J Pathol 1989; 135:827-834.

64

Astrocytes in Brain Aging and Neurodegeneration

237. Gollin PA, Kalaria RN, Eikelenboom P et al. Alpha1-antitrypsin and alpha1-antichymotrypsin are in the lesions of Alzheimer’s disease. NeuroReport 1992; 3:201-203. 238. Abraham CR, Shirahama T, Potter H. α1-Antichymotrypsin is associated solely with amyloid deposits containing the beta-protein. Amyloid and cell localization of 1-antichymotrypsin. Neurobiol Aging 1990; 11:123-129. 239. Rozemuller JM, Abbink JJ, Kamp AM et al. Distribution pattern and functional state of α1-antichymotrypsin in plaques and vascular amyloid in Alzheimer’s disease. An immunohistochemical study with monoclonal antibodies against native and inactivated α1-antichymotrypsin. Acta Neuropathol 1991; 82:200-207. 240. Fraser PE, Nguyen JT, McLachlan DR et al. α1-Antichymotrypsin binding to Alzheimer Aβ peptides is sequence specific and induces fibril disaggregation in vitro. J Neurochem 1993; 61:298-305. 241. Ma J, Yee A, Brewer BJ et al. Amyloid-associated proteins α1-antichymotrypsin and apolipoprotein E promote assembly of Alzheimer β-protein into filaments. Nature 1994; 372:92-94. 242. Kanemaru K, Meckelein B, Marshall DCL et al. Synthesis and secretion of active α1-antichymotrypsin by murine primary astrocytes. Neurobiol Aging 1996; 17:767-771. 243. Su JH, Cummings BJ, Cotman CW. Localization of heparan sulfate glycosaminoglycan and proteoglycan core protein in aged brain and Alzheimer’s disease. Neuroscience 1992; 51:801-813. 244. Snow AD, Nochlin D, Sekiguchi R et al. Identification and immunolocalization of a new class of proteoglycan (keratan sulfate) to the neuritic plaques of Alzheimer’s disease. Exp Neurol 1996; 138:305-317. 245. Brunden KR, Richtercook NJ, Chaturveti N et al. pH-Dependent binding of synthetic β-amyloid peptides to glycosaminoglycans. J Neurochem 1993; 61:2147-2154. 246. Crutcher KA, Anderton BH, Barger SW et al. Cellular and molecular pathology in Alzheimer’s disease. Hippocampus 1993; 3 Spec No:271-287. 247. Snow AD, Malouf AT. In vitro and in vivo models to unravel the potential roles of beta/ A4 in the pathogenesis of Alzheimer’s disease. Hippocampus 1993; 3 Spec No:257-267. 248. Mahley RW. Apolipoprotein E: Cholesterol transport protein with expanding role in cell biology. Science 1988; 240:622-630. 249. Roses AD. On the metabolism of apolipoprotein E and the Alzheimer diseases. Exp Neurol 1995; 132:149-156. 250. Diedrich JF, Minnigan H, Carp RI et al. Neuropathological changes in scrapie and Alzheimer’s disease are associated with increased expression of apolipoprotein E and cathepsin D in astrocytes. J Virol 1991; 65:4759-4768. 251. Namba Y, Tomonaga M, Kawasaki H et al. Apolipoprotein E immunoreactivity in cerebral amyloid deposits and neurofibrillary tangles in Alzheimer’s disease and kuru plaque amyloid in Creutzfeldt-Jakob disease. Brain Res 1991; 541:163-166. 252. Coria F, Castano E, Prelli F et al. Isolation and characterization of amyloid P component from Alzheimer’s disease and other types of cerebral amyloidosis. Lab Invest 1988; 58:454-458. 253. Strittmatter WJ, Saunders A, Schmechel D et al. Apolipoprotein E: High avidity binding to β-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci USA 1993; 90:1977-1981. 254. Wisniewski T, Golabek A, Matsubara E et al. Apolipoprotein E: Binding to a soluble Alzheimer’s beta-amyloid. Biochem Biophys Res Commun 1993; 192:359-365. 255. Geddes JW, Monaghan DT, Cotman CW et al. Plasticity of hippocampal circuitry in Alzheimer’s disease. Science 1985; 230:1179-1181. 256. Pike CJ, Cummings BJ, Cotman CW. Early association of reactive astrocytes with senile plaques in Alzheimer’s disease. Exp Neurol 1995; 132:172-179. 257. Gray CW, Patel AJ. Characterization of a neurotrophic factor produced by cultured astrocytes involved in the regulation of subcortical cholinergic neurons. Brain Res 1992; 574:257-265.

Astrocyte Pathophysiology in Disorders of the Central Nervous System

65

258. Akiyama H, Kawamata T, Yamada T et al. Expression of intercellular adhesion molecule (ICAM)-1 by a subset of astrocytes in Alzheimer disease and some other neurodegenerative disorders. Acta Neuropathol 1993; 85:628-634. 259. Araujo DM, Cotman CW. Beta-amyloid stimulates glial cells in vitro to produce growth factors that accumulate in senile plaques in Alzheimer’s disease. Brain Res 1992; 569:141-145. 260. Gijbels K, Proost P, Masure S et al. Gelatinase B is present in the cerebrospinal fluid during experimental autoimmune encephalomyelitis and cleaves myelin basic protein. J Neurosci Res 1993; 36:432-440. 261. Siman R, Card JP, Nelson RB et al. Expression of beta-amyloid precursor protein in reactive astrocytes following neuronal damage. Neuron 1989; 3:275-285. 262. Berkenbosch F, Refolo LM, Friedrich VL Jr et al. The Alzheimer’s amyloid precursor protein is produced by type I astrocytes in primary cultures of rat neuroglia. J Neurosci Res 1990; 25:431-440. 263. Mark RJ, Blanc EM, Mattson MP. Amyloid β-peptide and oxidative cellular injury in Alzheimer’s disease. Mol Neurobiol 1996; 12:211-224. 264. Penfield W, Humphreys S. Epileptogenic lesions of the brain. A histologic study. Arch Neurol Psychiat 1940; 43:240-261. 265. Harris AB. Cortical neuroglia in experimental epilepsy. Exp Neurol 1975; 49:691-715. 266. Westrum LE, White LEJ, Ward AAJ. Morphology of the experimental epileptic focus. J Neurosurg 1964; 21:1033-1044. 267. Armstrong DD. The neuropathology of of temporal lobe epilepsy. J Neuropathol Exp Neurol 1993; 52:433-443. 268. Fischer J, Holubar J, Malik V. Neurohistochemical study of the development of the experimental epileptogenic cortical cobalt-gelatine focus in rats and its correlation with the onset of epileptic electrical activity. Acta Neuropathol 1968; 11:45-54. 269. Payan H, Toga M, Berard-Badier M. The pathology of post-traumatic epilepsies. Epilepsia 1970; 11:81-94. 270. Orzi F, Zoli M, Passarelli F et al. Repeated electroconvulsive shock increases glial fibrillary acidic protein, ornithine decarboxylase, somatostatin and cholecystokinin immunoreactivities in the hippocampal formation of the rat. Brain Res 1990; 533:223-231. 271. Steward O. Electroconvulsive seizures upregulate astroglial gene expression selectively in the dentate gyrus. Mol Brain Res 1994; 25:217-224. 272. Stringer JL. Repeated seizures increase GFAP and vimentin in the hippocampus. Brain Res 1996; 717:147-153. 273. Khurgel M, Ivy GO. Astrocytes in kindling: Relevance to epileptogenesis. Epilepsy Res 1996; 26:163-175. 274. Brotchi J. The activated astrocyte - a histochemical approach to the epileptic focus. In: Schoffeniels E, Franck G, Towers DB, Hertz L, eds. Dynamic Properties of Glial Cells. Oxford: Pergamon, 1978:429-434. 275. Lee SH, Magge S, Spencer DD et al. Human epileptic astrocytes exhibit increased gap junction coupling. Glia 1995; 15:195-202. 276. Cornell-Bell AH, Williamson A. Hyperexcitability of neurons and astrocytes in epileptic human cortex. In: Fedoroff S, Juurlink BHJ, Doucette R, eds. Biology and Pathology of Astrocyte-Neuron Interactions. New York: Plenum Press, 1993:51-65. 277. Norenberg MD, Chu N-S. Aminophylline-induced preictal alterations in cortical astrocytes. Exp Neurol 1977; 54:340-351. 278. Ward A. Glia and epilepsy. In: Schoffeniels E, Franck G, Hertz L, Tower DB, eds. Dynamic Properties of Glial Cells. Oxford: Pergamon Press, 1978:413-427. 279. Meldrum BS. Pharmacology of GABA. Clin Neuropharmacol 1982; 5:293-316. 280. Bradford HF. Glutamate, GABA and epilepsy. Prog Neurobiol 1995; 47:477-511. 281. Schousboe A, Larsson OM, Frandsen A et al. Neuromodulatory actions of glutamate, GABA and taurine: Regulatory role of astrocytes. Adv Exp Med Biol 1991; 296:165-180. 282. Schousboe A, Thorbek P, Hertz L et al. Effects of GABA analogues of restricted conformation on GABA transport in astrocytes and brain cortex slices and on GABA receptor binding. J Neurochem 1979; 33:181-189.

66

Astrocytes in Brain Aging and Neurodegeneration

283. Schousboe A, Hjeds H, Engler J et al. Tissue distribution, metabolism, anticonvulsant efficacy and effect on brain amino acid levels of the glia-selective gamma-aminobutyric acid transport inhibitor 4,5,6,7-tetrahydroisoxazolo (4,5-c) pyridin-3-ol in mice and chicks. J Neurochem 1986; 47:758-63. 284. Pollen DA, Trachtenberg MC. Neuroglia: Gliosis and focal epilepsy. Science 1970; 167:1252-1253. 285. Lewis DV, Mutsuga N, Schuette WH et al. Potassium clearance and reactive gliosis in the alumina gel lesion. Epilepsia 1977; 18:132-136. 286. Somjen GG. Acidification of interstitial fluid in hippocampal formation caused by seizures and by spreading depression. Brain Res 1984; 311:186. 287. Benjamin AM. Ammonia. In: Lajtha A, ed. Handbook of Neurochemistry, Vol 1. New York: Plenum Press, 1982:117-137. 288. Cooper AJL, Plum F. Biochemistry and physiology of brain ammonia. Physiol Rev 1987; 67:440-519. 289. Lapin IP. Convulsant action of intracerebroventricularly administered l-kynurenine sulfate, quinolinic acid and other derivatives of succinic acid, and effects of amino acids: Structure-activity relationships. Neuropharmacology 1982; 21:1227-1233. 290. Du F, Williamson J, Bertram E et al. Kynurenine pathway enzymes in a rat model of chronic epilepsy: Immunohistochemical study of activated glial cells. Neuroscience 1993; 55:975-989. 291. Sutula TP, Cascino G, Cavazos JIP et al. Mossy fiber synaptic reorganization in the epileptic human temporal lobe. Ann Neurol 1989; 26:321-330. 292. Babb TL, Kupfer WR, Pretorius JK et al. Synaptic reorganization by mossy fibers in human epileptic fascia dentata. Neuroscience 1991; 42:351-363. 293. Represa A, Jorquera I, Le Gal La Salle G et al. Epilepsy induced collateral sprouting of hippocampal mossy fibers: Does it induce the development of ectopic synapses with granule cell dendrites? Hippocampus 1993; 3:257-268. 294. Niquet J, Jorquera I, Faissner A et al. Gliosis and axonal sprouting in the hippocampus of epileptic rats are associated with an increase of tenascin-C immunoreactivity. J Neurocytol 1995; 24:611-624. 295. Gall CM, Berschauer R, Isackson PJ. Seizures increase basic fibroblast growth factor mRNA in adult rat forebrain neurons and glia. Mol Brain Res 1994; 21:190-205. 296. Niquet J, Jorquera I, Ben-Ari Y et al. Proliferative astrocytes may express fibronectin-like protein in the hippocampus of epileptic rats. Neurosci Lett 1994; 180:13-16. 297. Niquet J, Gillian A, Ben-Ari Y et al. Reactive glial cells express a vitronectin-like protein in the hippocampus of epileptic rats. Glia 1996; 16:359-367. 298. Le Gal La Salle G, Rougon G, Valin A. The embryonic form of neural cell surface molecule (E-NCAM) in the rat hippocampus and its reexpression on glial cells following kainic acid-induced status epilepticus. J Neurosci 1992; 12:872-882. 299. Niquet J, Jorquera I, Ben-Ari Y et al. NCAM immunoreactivity on mossy fibers and reactive astrocytes in the hippocampus of epileptic rats. Brain Res 1993; 626:106-116. 300. Griffin WST, Yeralan O, Sheng JG et al. Overexpression of the neurotrophic cytokine S100 in human temporal lobe epilepsy. J Neurochem 1995; 65:228-233. 301. Tauck DL, Nadler JV. Evidence for functional mossy fiber sprouting in hippocampal formation of kaininc acid treated rats. J Neurosci 1985; 5:1016-1022. 302. Ben-Ari Y, Represa A. Brief seizure episodes induce long-term potentiation and mossy fibre sprouting in the hippocampus. Trends Neurosci 1990; 13:312-318. 303. DeLong MR. Primate models of movement disorders of basal ganglia origin. Trends Neurosci 1990; 13:281-285. 304. Turski L, Bressler K, Rettig KJ et al. Protection of substantia nigra from MPP+ neurotoxicity by N-methyl-D-aspartate antagonists. Nature 1991; 349:414-418. 305. Greenamyre JT. Glutamate-dopamine interactions in the basal ganglia: Relationship to Parkinson’s disease. J Neural Trans 1993; 91:255-269. 306. Hazell AS, Itzhak Y, Liu H et al. 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) decreases glutamate uptake in cultured astrocytes. J Neurochem 1997; 68:2216-2219.

Astrocyte Pathophysiology in Disorders of the Central Nervous System

67

307. Schiffer D, Cordera S, Cavalla P et al. Reactive astrogliosis of the spinal cord in amyotrophic lateral sclerosis. J Neurol Sci 1996; 139 Suppl.27-33. 308. Rothstein JD. Excitotoxic mechanisms in the pathogenesis of amyotrophic lateral sclerosis. Adv Neurol 1995; 68:7-20. 309. Rothstein JD, Martin LJ, Kuncl RW. Decreased glutamate transport by brain and spinal cord in amyotrophic lateral sclerosis. New Engl J Med 1992; 326:1464-1468. 310. Rothstein JD, Van Kammen M, Levey AI et al. Selective loss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann Neurol 1995; 38:73-84. 311. Rothstein JD, Dykes-Hoberg M, Pardo CA et al. Knockout of glutamate transporters reveals a major role for astroglial transport in excitotoxicity and clearance of glutamate. Neuron 1996; 16:675-686. 312. Rosen DR, Siddique T, Patterson D et al. Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 1993; 362:59-62. 313. Kato S, Shimoda M, Watanabe Y et al. 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 1996; 55:1089-1101. 314. Gurney ME, Pu H, Chiu AY et al. Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science 1994; 264:1772-1775. 315. Bruijn LI, Becher MW, Lee MK et al. ALS-linked SOD1 mutant G85R mediates damage to astrocytes and promotes rapidly progressive disease with SOD1-containing inclusions. Neuron 1997; 18:327-338. 316. O’Reilly SA, Roedica J, Nagy D et al. Motor neuron-astrocyte interactions and levels of Cu,Zn superoxide dismutase in sporadic amyotrophic lateral sclerosis. Exp Neurol 1995; 131:203-210. 317. Yu BP. Cellular defenses against damage from reactive oxygen species. Physiol Rev 1994; 74:139-161. 318. Halliwell B. Free radicals, antioxidants and human disease: Curiosity, cause, or consequence? Lancet 1994; 344:721-724. 319. Norenberg MD. Astrocytes in ischemic injury. In: Ginsberg MD, Bogousslavsky J, eds. Cerebrovascular Disease: Pathophysiology, Diagnosis, and Management. Cambridge: Blackwell, 1997. 320. Petito CK, Babiak T. Early proliferative changes in astrocytes in postischemic noninfarcted rat brain. Ann Neurol 1982; 11:510-518. 321. Hori O, Matsumoto M, Maeda Y et al. Metabolic and biosynthetic alterations in cultured astrocytes exposed to hypoxia/reoxygenation. J Neurochem 1994; 62:1489-1495. 322. Richter D, Dawson RMC. The ammonia and glutamine content of the brain. J Biol Chem 1948; 176:1119-1210. 323. Goldberg RN, Cabal LA, Sinatra FR et al. Hyperammonemia associated with perinatal asphyxia. Pediatrics 1979; 64:336-341. 324. Rischke R, Krieglstein J. Postischemic neuronal damage causes astroglial activation and increase in local cerebral glucose utilization of rat hippocampus. J Cereb Blood Flow Metab 1991; 11:106-113. 325. Yu ACH, Gregory GH, Chan PH. Hypoxia-induced dysfunction and injury of astrocytes in primary cell cultures. J Cereb Blood Flow Metab 1989; 9:20-28. 326. Swanson RA. Astrocyte glutamate uptake during chemical hypoxia in vitro. Neurosci Lett 1992; 147:143-146. 327. Rothstein JD, Jin L, Dykes-Hoberg M et al. Chronic inhibition of glutamate uptake produces a model of slow neurotoxicity. Proc Natl Acad Sci USA 1993; 90:6591-6595. 328. Robinson MB, Djali S, Buchhalter JR. Inhibition of glutamate uptake with L-transpyrrolidine-2,4-dicarboxylate potentiates glutamate neurotoxicity in primary hippocampal cultures. J Neurochem 1993; 61:2099-2103. 329. Torp R, Lekieffre D, Levy LM et al. Reduced postischemic expression of a glial glutamate transporter, GLT1, in the rat hippocampus. Exp Brain Res 1995; 103:51-58. 330. Pogun S, Dawson V, Kuhar MJ. Nitric oxide inhibits 3H-glutamate transport in synaptosomes. Synapse 1994; 18:21-26.

68

Astrocytes in Brain Aging and Neurodegeneration

331. Butterworth RF. Pathophysiology of hepatic encephalopathy; the ammonia hypothesis revisited. In: Bengtsson F, Jeppsson B, Almdal T, Vilstrup H, eds. Hepatic Encephalopathy and Metabolic Nitrogen Exchange. Boca Raton: CRC Press, 1993:9-24. 332. Basile AS, Jones EA, Skolnick P. The pathogenesis and treatment of hepatic encephalopathy: Evidence for the involvement of benzodiazepine receptor ligands. Pharmacol Rev 1991; 43:27-71. 333. Norenberg MD. Hepatic encephalopathy. In: Kettenmann H, Ransom BR, eds. Neuroglia. New York: Oxford, 1995:950-963. 334. Rao VLR, Murthy CRK, Butterworth RF. Glutamatergic synaptic dysfunction in hyperammonemic syndromes. Metab Brain Dis 1992; 7:1-20. 335. Gregorios JB, Mozes LW, Norenberg LOB et al. Morphologic effects of ammonia on primary astrocyte cultures. I. Light microscopic studies. J Neuropathol Exp Neurol 1985; 44:397-403. 336. Gregorios JB, Mozes LW, Norenberg MD. Morphologic effects of ammonia on primary astrocyte cultures. II. Electron microscopic studies. J Neuropathol Exp Neurol 1985; 44:404-414. 337. Sobel RA, DeArmond SJ, Forno LS et al. Glial fibrillary acidic protein in hepatic encephalopathy. An immunohistochemical study. J Neuropathol Exp Neurol 1981; 40:625-632. 338. Norenberg MD. Astrocytic-ammonia interactions in hepatic encephalopathy. Semin Liver Dis 1996; 16:245-253. 339. Szerb JC, Butterworth RF. Effect of ammonium ions on synaptic transmission in the mammalian central nervous system. Prog Neurobiol 1992; 39:135-153. 340. Bender AS, Norenberg MD. Effects of ammonia on L-glutamate uptake in cultured astrocytes. Neurochem Res 1996; 21:567-573. 341. Zhou B, Norenberg MD. Ammonia downregulates GLAST mRNA glutamate transporter in cultured astrocytes. Soc Neurosci Abstr 1997; 23:1461. 342. Norenberg MD, Huo Z, Neary JT et al. The glial glutamate transporter in hyperammonemia and hepatic encephalopathy: Relation to energy metabolism and glutamatergic neurotransmission. Glia 1997; 21:124-133. 343. McCarthy KD, Harden TK. Identification of two benzodiazepine binding sites on cells cultured from rat cerebral cortex. J Pharmacol Exp Ther 1981; 216:183-191. 344. Bender AS, Hertz L. Binding of (3H) R05-4864 in primary cultures of astrocytes. Brain Res 1985; 341:41-9. 345. Itzhak Y, Baker L, Norenberg MD. Characterization of the peripheral-type benzodiazepine receptor in cultured astrocytes: evidence for multiplicity. Glia 1993; 9:211-218. 346. Anholt RRH, Pedersen PL, DeSouza EB et al. The peripheral-type benzodiazepine receptor: Localization to the mitochondrial outer membrane. J Biol Chem 1986; 261:576-583. 347. Papadopoulos V. Peripheral-type benzodiazepine/diazepam binding inhibitor receptor: Biological role in steroidogenic cell function. Endocr Rev 1993; 14:222-240. 348. Mok WM, Herschkowitz S, Krieger NR. In vivo studies identify 5alpha-pregnan-3alpha-ol20-one as an active anesthetic agent. J Neurochem 1991; 57:1296-1301. 349. Bitran D, Hilvers RJ, Kellog CK. Anxiolytic effects of 3alpha-hydroxy-5alpha[beta]-pregnan20-one: Endogenous metabolites of progesterone that are active at the GABA-A receptor. Brain Res 1991; 561:157-161. 350. Wieland S, Lan NC, Mirasedeghi S et al. Anxiolytic activity of the progesterone metabolite 5alpha-pregnan-3alpha-ol-20-one. Brain Res 1991; 565:263-268. 351. Norenberg MD, Itzhak Y, Bender AS. The peripheral benzodiazepine receptor and neurosteroids in hepatic encephalopathy. Adv Exp Biol Med 1997; 420:95-111.

Part II Astrocytes in Human Brain Senescence and Neurodegenerative Disorders

CHAPTER 4

Glial Responses to Injury, Disease, and Aging Lawrence F. Eng and Yuen Ling Lee

Introduction

A

strocytes comprise 25% of the cells and 35% of the mass in the central nervous system (CNS). They have intimate contact with the pia of the brain, neurons, endothelial cells, pericytes, myelin membrane internodes, synapses, and microglia. Astrocytes have specialized functions depending on their location in the CNS and participate in a variety of important physiologic and pathologic processes. Any kind of stress or injury to the CNS induces a stereotypic response in the astrocytes termed reactive gliosis and, in the extreme case, astrogliosis and plaque or scar formation. The prominence of astroglial reactions in various injuries/diseases, the rapidity of the astroglial response and the evolutionary conservation of reactive astrogliosis indicate that reactive astrocytes fulfill important functions in the CNS.1,2 Astrogliosis is best characterized by rapid synthesis of glial fibrillary acidic protein (GFAP), a cytoskeletal intermediate filament. Numerous in vitro and in vivo studies on the molecular profiles of substances which are upregulated during astrocyte activation document the complex and varied responses of astrocytes to injury. Besides morphological changes, reactive astrocytes have also been shown to upregulate a number of different molecules including other glial markers (S-100β), cytokines (IL-1, IL-6, IFN and TNF), growth factors (FGF, NGF, NT-3, CNTF) and heat shock proteins.3 Reactive astrocytes are thought to play a role in the healing phase following CNS injury by actively monitoring and controlling the molecular and ionic contents of the extracellular space of the CNS. It has been hypothesized that activated astroglia may benefit the damaged neurons by participating in several important biological processes such as regulation of neurotransmitter levels, repair of the extracellular matrix, control of the blood-CNS interface, control of transport processes, and trophic support to the damaged cells. On the other hand, gliosis which occurs during normal aging and after injury may result in detrimental pathological effects by interfering with the residual neuronal circuits through inhibiting regeneration or preventing remyelination.

Astrocyte Intermediate Filament, Glial Fibrillary Acidic Protein As a member of the cytoskeletal protein family, GFAP is thought to be important in modulating astrocyte motility and shape by providing structural stability to extensions of astrocytic processes. It was first isolated from a multiple sclerosis (MS) plaque in 1969 (Fig. 4.1). GFAP is the principal 8-9 nm intermediate filament in mature astrocytes.2 Genomic clones have been obtained from human, mouse and rat GFAP genes. Each gene is Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

72

Astrocytes in Brain Aging and Neurodegeneration

Fig. 4.1. Diagram illustrates the isolation of GFAP from a MS plaque for amino acid analysis.

composed of nine exons distributed over about 10 kb of DNA and yields a mature mRNA of about 3 kb. The coding sequences for the three genes are highly homologous. Strong homology also extends upstream of the RNA start site for about 200 bp, recurs between about -1300 and 1700 (RNA startpoint = +1) and is present in some intronic regions. The primary sites for the initiation of RNA and protein synthesis are essentially identical for the three genes, and each contains a TAT-like sequence (CATAAA or AATAA) in the expected 5'-flanking position. In addition to GFAP-α, two additional mRNAs that start at different sites have been identified, GFAP-β and GFAP-γ. The different tissue distributions of GFAP-α, -β, and -γ mRNAs suggest that the synthesis of each is subject to unique control. All transcriptional studies to date either have explicitly measured GFAP-α or have not distinguished among the possible mRNA isotypes. GFAP transgenes have been used extensively to study signaling pathways that operate during development, disease, and injury—all states that increase GFAP gene activity.4,5

Glial Responses to Injury, Disease, and Aging

73

Astrocytes in Experimental Gliosis Increased protein content or immunostaining of GFAP has been found in experimental models involving gliosis, such as cryogenic lesions,6 stab wounds,7-15 toxic lesions16-18 and experimental allergic encephalomyelitis (EAE).19-22 Increased levels of GFAP mRNA have been found in the 6-hydroxydopamine lesion-bearing substantia nigra in the rat,17,23 mechanical lesions of rat cerebral cortex,24-26 entorhinal cortex lesions,27 corticospinal axotomy,28 EAE29 and lesions of the dentate gyrus.30

Astrocytes in Disease While reactive gliosis occurs with any type of insult to the CNS, the anatomical region, severity of gliosis, and developmental time sequence vary in amyotrophic lateral sclerosis (ALS), Gerstmann-Straussler syndrome (GSS), Huntington’s, Wilson’s, Pick’s, Parkinson’s, Alzheimer’s, and Down’s diseases.31 Astrocytic gliosis is a prominent neuropathologic change in Alzheimer’s disease (AD). Numerous reports have shown reactive astrocytes in AD brains, most frequently in association with neuritic plaques.32-38 It is not known whether gliosis precedes the appearance of β-amyloid peptide (βAP) or whether it might be induced by βAP. Moreover, cases have been reported where diffuse βAP plaques in advanced AD are not associated with reactive astrocytes.39-41 Diffuse plaques in advanced AD cases may not be identical to the early lesions which may characterize preclinical AD cases. Our approach to identifying early pathological changes in AD has been to study young cases of Down’s syndrome (DS), because individuals with DS develop the neuropathological changes of AD prematurely.42 DS cases show a very mild form of AD pathology from the second and third decades of life; however, DS cases from the fifth decade and older show fully developed AD.43 Since the amygdala is a site of neuropathologic change, including extensive gliosis, in both AD and DS,44-46 we examined the amygdala for evidence of astrocytic gliosis in young and old cases of DS and AD. We also compared the distribution of astrocytes with the distribution of βAP deposits in the amygdala to determine whether the βAP deposits were spatially related to astrocytes. Our results demonstrated that astrocytic hypertrophy is not an early change in the AD-like process of DS and that astrocyte morphology did not differ in young DS cases from that of controls. Furthermore, there was no consistent spatial relationship between the numerous βAP deposits observed in the young DS cases and astrocytes.47 In agreement with our study, Michetti et al48 did not observe a difference in the morphology of S-100 labeled astrocytes in the cerebellum of DS cases (ranging from newborn to 26 months) relative to controls. Griffin et al,49 however, have reported that S-100 labeled astrocytes were increased in size in DS cases aged 2 days, 3.5 months, and 34 years. Our data demonstrate that the βAP deposits in young DS brain, which may be similar to those in preclinical AD, are not associated with reactive gliosis.47 In a recent study of DS brains, GFAP was found to be expressed at levels significantly below those of controls, suggesting that trisomy 21 exerts a suppressive effect on GFAP gene expression.50 Moossy et al51 have described two cases of “primary dementia” not distinguishable from AD but devoid of neurofibrillary tangles. Astrocytosis was observed in several subcortical nuclei and mainly in the thalamus. Astrogliosis has been demonstrated by chemical analysis and immunocytochemistry for GFAP in Huntington’s disease.52,53 Huntington disease brains were used as positive controls for gliosis in a study of schizophrenic patient brains.53 Employing quantitative image analysis of brains immunostained for GFAP, Roberts et al54 did not find significant differences in 20 different brain areas between schizophrenic and Huntington’s disease groups, in contrast to previous reports of gliosis in schizophrenic brains. Holzer’s histological stain for glial fibrils showed increased fibrillary gliosis that affected mainly the periventricular structures of the diencephalon, the periventricular structures of the periaqueductal region of the mesencephalon, and the basal forebrain of schizophrenic

74

Astrocytes in Brain Aging and Neurodegeneration

subjects. The hypothalamus, midbrain tegmentum, and innominata were also involved.55 Astrocytes in human brain immunostain less intensely when the death-autopsy interval is prolonged, because GFAP is very sensitive to proteolysis.56,57 In contrast, Holzer’s stain prominently labels mildly gliotic tissue which may not be evident by GFAP immunostaining (L. Forno, personal communication). Significant astrogliosis is also characteristic of ALS. In a study of 13 ALS brains, gliosis was present in six different control areas and was different from that seen in AD, Pick’s disease, and Parkinson’s disease. GFAP staining within the subcortical white matter of ALS was unlike that of any other disease examined with the exception of cerebral infarction.58 The prion diseases, which include CJD, GSS, and kuru in humans and scrapie in animals, have been characterized as being “hypergliotic” because the gliosis often appears to be out of proportion to the degree of nerve cell loss or injury.59,60 Astrocyte hypertrophy and gliosis were concentrated in the cerebellum of the ataxic form of CJD.61,62 The brains from hamsters infected with the human CJD agent showed a gradual increase of GFAP and GFAP mRNA during the course of the disease.63 Molecular studies suggest that a single abnormal nerve cell protein PrPsc may cause both the neuronal degeneration and reactive gliosis.64 Finally, in Wilson’s disease and hepatic encephalopathy, gliosis occurs in demyelinated foci of the brain; however, there is a population of protoplasmic astrocytes in the gray matter (Alzheimer’s type 2 cells) that show a decrease in GFAP content.65-68

Astrocyte Activation of GFAP in Astrogliosis What are the signals that induce the upregulation of GFAP in normal aging and disease? Astrocytes can be activated by molecules expressed by microglia, monocyte/macrophages, endothelial cells, lymphocytes, by various blood proteins, ions, free radicals, neurotransmitters, enzymes, and their degradation products resulting from damaged neurons and glial cells. Whatever the factors may be, they probably induce mild or transient activation which is less intense than that seen in injury or disease.

Microglial Activation The broad distribution of microglia in the CNS is similar to that of astrocytes. Microglial cells are intimately associated with astrocytes, endothelial cells, oligodendrocytes and neurons. Microglial activation appears to be independent of the form of pathological stimulus since uniform changes occur in all models, including proliferation, transformation into phagocytic cells with macrophage morphology, and upregulation of cell surface molecules such as the MHC antigens. Microglia responds quickly to a variety of signaling molecules at a very early stage of injury.69-74 Activation often precedes reactions of any other cell type in the CNS, even before the reactive astrocytic response. They respond to changes in the brain’s structural integrity, and also to very subtle alterations in their microenvironment, such as an imbalance in homeostasis that precedes histologically detectable pathological changes.75 This may be due to the unique collection of microglial membrane channels which includes an inward-rectifying K+ channel (see below).76,77 Although little is known about the in vivo regulation of microglial proliferation and activation, both in vivo and in vitro experiments suggest an involvement of cytokines in this process.78,79 T cell derived IFN-γ is the cytokine most widely implicated in the activation of microglia. The long list of events that occur after IFN-γ treatment includes the production of reactive oxygen intermediates, increase in MHC class II expression and synthesis of complement components. Colony stimulating factors (CSF) have been identified as mediating potent activation of microglial cells. The source of these molecules in the brain in vivo is still controversial. Astrocytes have been shown to express macrophage (M)-CSF, granulocyte (G)-CSF, and granulocyte-macrophage (GM)-CSF. Neurons have been shown to secrete

Glial Responses to Injury, Disease, and Aging

75

M-CSF in vitro.80 Chemotactic effects on resident microglia and blood monocytes are exerted by TGF-β and the β-chemokines, a group of 8-10 kDa proteins with highly conserved cysteines linked by disulfide bonds. Other β-chemokines like monocyte chemoattractant protein-1 (MCP-1/JE) or macrophage inflammatory protein-1α (MIP-1α) are produced by IL-1β- and TNF-α-activated astrocytes and microglia.81 Early metabolic and ultrastructural alterations in the neurons, including disintegration of cytoskeletal proteins, synaptic membrane changes, decrease in protein synthesis, altered polyamine metabolism or elevated K+ in the intracellular space, may activate the microglia. Potassium can depolarize microglia via an inward potassium channel; however, they lack a rectifying outward current so that even a small inward current leads to membrane depolarization with unknown metabolic consequences. Activated microglia have been reported to release the following compounds: IL-1, IL-6, IL-10, TGF-β and TNF-α. Activated microglia increase with normal aging and exhibit increased expression of MHC II, leukocyte common antigen CD4, and ED1, the latter a marker of the lysosomal apparatus whose upregulation indicates activation of the endosomal-lysosomal system in aging microglia.82,83

Monocyte/Macrophage Activation Bone marrow-derived monocytes and macrophages infiltrate the CNS following injury or disease. Because of the variety of effects they mediate, monocytes and macrophages may play a crucial role in neuroimmunologic disorders such as MS and EAE. Their activation and recruitment into the CNS occur in response to chemokines and cytokines secreted by endogenous cells of the CNS (astrocytes, microglia, and endothelial cells) as well as activated lymphocytes. Activated macrophages secrete inflammatory mediators (IL-1, IL-4, IL-6, IL-8, TNF-α, TGF-β, MIP-1, MIP-2, M-CSF, MCP-1, γIP-10, GRO, RANTES), nitric oxide and proteases which serve to enhance the inflammatory response, promote vascular permeability and initiate myelin destruction.84,85 In addition, their role as antigen presenting cells has been well documented.86

Endothelial Cell Activation Interaction between various subclasses of inflammatory cells and endothelial cells (ECs) comprising the mammalian blood-brain barrier (BBB) is an early, important event in the course of CNS inflammation, as exemplified by EAE. Breakdown of the BBB through the action of vasoactive amines and other soluble inflammatory factors (cytokines) and enhanced endothelial transcytotic activity facilitates the migration of inflammatory cells in the CNS.87-89 However, studies by Knobler et al90 have suggested that ECs at the intact BBB may also actively participate in the trafficking of lymphocytes into the CNS. Vascular adhesion molecules (selectins) are elaborated in acute brain inflammation and appear to foster lymphocyte entry into the CNS.91-93 In addition, activated/injured ECs have been shown to express factors such as PDGF, plasminogen activator, prostacylin, TNF-β, IL-1 and IL-8. Several of these cytokines play a role in the adhesion of leukocytes to vascular endothelium and thereby contribute to the inflammatory response.94

Astrocytes in Normal Aging Senescent astrocytes increase in size, become fibrous and exhibit a gradual increase in GFAP and GFAP mRNA. The increase of GFAP during senescence can be seen in species from several long-separated evolutionary orders of mammals.95 The augmented GFAP mRNA largely corresponds to enhanced astrocyte hypertrophy rather than to increases in total numbers of these cells.96,97 In humans, the increase in GFAP mRNA appears to be negligible before 60 years, but increases occur in the hippocampus and in frontal and temporal cortex during an average life-span in the absence of specific neurological disease.70 GFAP mRNA

76

Astrocytes in Brain Aging and Neurodegeneration

and protein increase 2-fold after midlife in mice and rats.98-102 In a study of aging female mouse brains employing combined immunocytochemical and in situ hybridization techniques, GFAP mRNA and protein exhibited significant age-related increases in the major white matter tracts, including the corpus callosum, fimbria, stria terminalis, and optic tract. Gray matter showed large increases in GFAP mRNA with age in the thalamus and hypothalamus, areas typically expressing little GFAP in the young.103 A parallel, age-related increase in GFAP intron RNA in the hippocampus, internal capsule, and corpus callosum of male rats indicates that the regulation of GFAP expression during aging occurs at the translational level.104 In mice, rats, and humans, increases in GFAP expression with aging occur gradually and there may be significant inter-individual variability in the degree of hypertrophy.70 Thus, minor degrees of reactive gliosis accompanying early degeneration or inflammatory CNS conditions may be easily overlooked. In other disorders, such as MS or adrenoleukodystrophy, reactive gliosis is generally intense and readily distinguishable from normal background levels of GFAP expression.105-107 There are two recent studies suggesting that low grade or transient stress might induce a gradual increase in GFAP expression in the course of normal aging. The first is the spreading depression (SD) experiments in which resident microglia were shown to release factors that activate the GFAP gene without altering astrocyte morphology. Cortical SD (CSD) was elicited in rat brain for one hour by topical application of 4 M potassium chloride solution. This treatment was sufficient to induce a microglial reaction throughout the cortex at 24 hours. Activated microglial cells exhibited apparent cellular hypertrophy, increased immunostaining with macrophage/microglial antibodies (MUC 100, 102, and OX-42), and de novo expression of MHC class II antigens.75 No neuronal damage or increase in GFAP immunoreactivity was detected three days after treatment. In a second study, Kraig et al108 reported that enhanced GFAP immunostaining could be demonstrated in rat brain astrocytes two days after induction of SD (21 DC shift) by application of KCl to the parietal cortex for 3 hr. The increased GFAP staining, however, returned to normal after two weeks. Both of these studies suggest that SD may be a useful technique to delineate the cellular mechanisms subserving GFAP upregulation in reactive astrocytes. For example, SD studies may help determine which factors released by transiently activated microglia are responsible for subsequent induction of GFAP in these astrocytes. Finch and co-workers have published a series of timely papers10,98,99,103 which demonstrate that GFAP and GFAP mRNA increase with normal aging in the rodent and human brain. Moreover, they showed that this increase in GFAP mRNA could be attenuated in aging rat if they were maintained on a calorie-restricted diet109,110 and hypothesized that the aging-related increase in GFAP is due to oxidative stress. One possible source of oxidative stress in the aging brain could be reactive microglia. A 2-fold increase in activated microglia was found in the aging rat hippocampus based on immunostaining with OX6, a marker for microglia/macrophages. Both the increase in reactive microgliosis and the parallel increases in GFAP expression could be delayed by dietary restriction.111 Such results are consistent with the effects of food restriction in attenuating oxidative damage to brain membranes during aging.112,113 To test this hypothesis, glial cell cultures were treated with H202, cysteamine or bacterial lipopolysaccharide to directly activate microglia. Treatment with H2O2 and cysteamine, but not bacterial lipopolysaccharide, induced GFAP mRNA in mixed cultures containing astrocytes, oligodendrocytes, and microglia as well as purified astrocytes alone. Furthermore, the GFAP response to oxidative stress was shown to be regulated at the transcriptional level.104,114 It was suggested that this control could be mediated by transcriptional response elements in the 5' upstream promoter region of the GFAP gene that respond

Glial Responses to Injury, Disease, and Aging

77

to oxidative stress, including sites for Fos, Jun and NFκB.114 However, various soluble factors in addition to reactive oxygen species are released from activated microglia which may regulate Fos, Jun or NFκB and, therefore, GFAP expression in nearby astrocytes.

Astrocyte Inclusions in Normal Aging Corpora amylacea (CA) arise in the human CNS in the course of normal aging and in several disease states. They are well-circumscribed, rounded inclusions ranging from 5 to 20 µm in diameter, and are well demonstrated by a wide variety of stains, including hematoxylin and eosin, iodine, Nile blue sulfate, methyl violet, PAS, and Best’s carmine. Most CA are comprised of a densely staining central round zone of amorphous material surrounded by a lighter peripheral rim. Chemical analysis of purified CA showed that they consist of up to 80% glycogen-like substance bound to approximately 1% sulfate and phosphate. In one study, protein content was estimated at around 5% whereas reactions for lipids, nucleic acids and sialic acid were negative.115 In another study, purified CA were found to contain about 4% total protein. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDSPAGE) analysis showed protein bands with molecular weights ranging from 24 to 133 kDa. Four bands with molecular weights of 24, 92, 94, and 133 kDa represented the major proteins present.116 Immunocytochemical analysis of human brain and spinal cord revealed small, rounded bodies expressing epitopes of the 72 kDa heat shock protein (HSP) in both normal and neurologically-abnormal individuals. These bodies were interpreted as the precorpora amylacea (pre-CA) which gradually enlarged into mature CA. Mature CA expressed HSP epitopes chiefly in their peripheral rims while the smaller, immature pre-CA stained intensely for HSP throughout the entire structure.117 In still another SDS-PAGE study, ubiquitin and HSP27 and 70 were demonstrated in CA by immunocytochemistry.118 In a combined light and electron microscopic study, CA in control normal aging brains were found to contain tau, tubulin, ubiquitin, and amyloid serum component P protein. Energy dispersive X-ray microanalysis data on isolated CA and those found in tissue sections corroborate findings obtained by wavelength dispersive analysis.119 CA characteristically contained high levels of iron, some copper and minimal aluminum. Elemental composition analysis also revealed a high content of phosphorus.120 In a second study from the same laboratory, CA were positively immunostained with antibodies directed against myelin basic protein, proteolipid protein, galactocerebroside, ferritin, and myelin-associated glycoprotein (MAG).121 A more recent X-ray microprobe analysis of CA revealed significant concentrations of sodium, phosphorus, sulfur, and chloride.122 Ultrastructurally, Ramsay123 demonstrated that CA were intracytoplasmic bodies with a smooth outline, lacking surrounding membranes. They consist of many dense linear structures of approximately 10 nm diameter; in those with more dense central cores, fibrils appeared to intermingle with granular deposits, and glycogen particles were often present at the periphery of the inclusions. Ramsay123 identified the structures as lying within the processes of fibrous astrocytes and there is general agreement that this is their usual, albeit not exclusive, location. Identical bodies have been identified in small numbers within myelinated axons,124 in the ventral and lateral horns of the human spinal cord, and within neurites of the human orbital cortex.125 The origin of corpora amylacea is not known. They occur particularly in the subpial and subependymal zones of the cerebral hemispheres, in the cerebral white matter particularly around small blood vessels, and in the hippocampus and long tracts of the spinal cord in elderly subjects, with or without neurological disorders. They are present in great numbers in degenerating white matter tracts and are reportedly increased in many sites in numerous pathological states. In accord with their largely intra-astrocytic localization, they

78

Astrocytes in Brain Aging and Neurodegeneration

tend vary greatly in number in areas of reactive gliosis and are occasionally observed in fibrillary astrocytomas. A functional role for CA has been proposed based on the presence of classical complement pathway components, the activation product C3d, the terminal complement complex (TCC), the C3 convertase regulator cofactor protein (MCP), the fluid phase regulator S-protein, and clusterin. It has been proposed that the formation of CA confers the CNS with a degree of protection from the consequence of complement activation, and that the functional role of CA formation is to provide an efficient mechanism for isolating potentially dangerous proteins following cell death in the aging and diseased CNS.126 The presence of ubiquitin and a number of HSPs strongly suggest that they are the result of some type of low-level chronic or transient stress.127 While the astrocyte is the cell that most commonly contain CA, proteins derived from other neural cells such as neurons and oligodendrocytes have occasionally been demonstrated in CA. An ultrastructural study of the vestibular nerve in patients suffering from Meniere’s disease and vascular cross-compression syndrome of the root entry zone revealed that CA produced in astrocytes can be transported to pial connective tissue across the glial-limiting lamina. The authors of that study suggested that CA are components of a glio-pial system devoted to the clearance of substances from the CNS.128 As in the case of augmented GFAP expression in normal aging, oxidative stress has also been implicated as an important mechanism mediating CA formation129-132 (see chapter 10). A review of the structural and biochemical changes which occur in subpopulation of astroglia in the normal aging brain recently has been published.133

Astrocyte Inclusions in Disease Chin and Goldman134 have recently compiled a detailed survey of the morphologic and histochemical properties of a variety of glial cytoplasmic inclusion bodies (GCIs) which arise in astrocytes and oligodendroglia under pathological conditions. The diseases in this review include multiple system atrophy, progressive supranuclear palsy, corticobasal ganglionic degeneration,. Pick disease, and Alexander disease. GCIs stain intensely with antibodies to ubiquitin and αB-crystallin and less intensely against α- and β-tubulin.135-138 Immunostaining with various antibodies to tau and paired helical filaments tended to be weak or negative.135-137 Ari et al139 and Abe et al140 have reported strong staining with monoclonal antibodies against microtubule-associated protein (MAP).5 Staining with antibodies to actin, vimentin, desmin, cytokeratin and GFAP has been consistently negative. Among the most prominent of the stress protein-rich neural inclusions are the Rosenthal fibers (RFs). RFs are eosinophilic, cytoplasmic inclusions of astrocytes present in pilocytic astrocytomas, in astrocytic scars, in MS plaques, chronic infarcts, and most prominently in Alexander’s disease.141-144 The RFs appear to be similar among the different disorders. RFs vary from round, focal deposits of a few microns in diameter to elongated, cigar-shaped fibers one hundred microns or more in length, the latter usually residing within astrocyte processes. At the ultrastructural level, RFs appear as dense, osmiophilic masses lying on a meshwork of intermediate filaments.43 The inclusion is composed of two small molecular weight heat shock proteins, αB-crystallin and HSP27.145,146 Some of the αB-crystallin is conjugated to ubiquitin.147 Levels of αB-crystallin and HSP27 mRNA are elevated in Alexander’s disease.148,149 It has been suggested that a variety of “stressors” might induce the accumulation of RFs in astrocytes (reviewed in ref. 134). To study the behavior of astrocytes overexpressing GFAP, six lines of transgenic mice were generated which carry multiple copies of the human GFAP gene.150,151 Mice that expressed high levels of the human GFAP gene (Tg73.1, Tg73.7, and Tg73.8) died by postnatal day 8-24 days while mice that expressed lower levels of the transgene (Tg73.2, Tg73.3, and

Glial Responses to Injury, Disease, and Aging

79

Tg73.5) survived and attained adulthood. At the light microscopic level, astrocytes in the high-expressing lines were distended by aggregates of globular eosinophilic material. Ultrastructural examination of optic nerve from a 13 day old Tg73.7 mouse showed astrocytes that contained abundant cytoplasmic filaments in association with irregular osmiophilic deposits resembling Rosenthal fibers (RFs) characteristic of Alexander’s disease.141,142 Astrocyte cultures were prepared from a transgenic mouse (Tg73.2) that carries multiple copies of the human GFAP gene and from its wild type littermate. Astrocytes in the Tg73.2 cultures appear irregularly shaped and enlarged, expressed increased human GFAP and its mRNA, exhibited both human and mouse GFAP, and expressed αB-crystallin protein and mRNA, HSP27 protein, and vimentin protein. At the light microscopic level, the Tg73.2 astrocytes appeared filled with eosinophilic deposits surrounded by GFAP positive immunostaining. Many, but not all, astrocytes in 20 day Tg73.2 cultures exhibited large, oddly shaped cells that immunostained with antibody specific for human GFAP (SMI-21), while most astrocytes in Tg73.2 cultures immunostained with a polyclonal anti-bovine GFAP antiserum (R-68) that reacts with human and mouse GFAP. This demonstrated that not all astrocytes in the Tg73.2 mouse contained the human GFAP gene. Tg73.2 astrocytes in culture for 14 days immunostained for αB-crystallin, but the wild type astrocyte cultures grown for equivalent lengths of time did not. Tg73.2 cultures at 20 days contain elevated αB-crystallin message when compared to time-matched wild type cells. At 14 days in culture, both the transgenic and wild type cells immunostained sparsely for HSP27. Conventional ultrastructural examination of Tg73.2 astrocytes showed numerous osmiophilic deposits in a bed of intermediate filaments (Fig. 4.2A) identical to that seen in a case of Alexander’s disease (Fig. 4.2B). Tg73.2 astrocyte cultures exhibited double labeling with antibovine GFAP antiserum (R-68) and anti-human GFAP (SMI-21) using an immunogold technique (Fig. 4.3B). Wild type astrocytes (Fig. 4.3A) exhibited staining with R-68 but not SMI-21. R-68 immunostaining of a paraffin-embedded brain section from an infant diagnosed with Alexander’s disease152 revealed numerous astrocytic processes surrounding the blood vessels which were replete with RFs. A RF exhibiting GFAP staining in its periphery is depicted in Figure 4.4A. A sample of white matter from an infant with Alexander’s disease was homogenized with 0.05 M phosphate buffer (pH 8) and centrifuged to give a clear supernatant and an insoluble pellet. This pellet was fixed in paraformaldehyde, embedded in paraffin, and sections were immunostained with R-68. The Rosenthal fibers in the pellet were surrounded by GFAP immunoreactivity (Fig. 4.4B). Evidence has been provided to show that the abnormal deposits present in Tg73.2 astrocytes are very similar to RFs found in Alexander’s disease. We conclude that Tg73.2 mouse astrocytes in culture do not require exposure to additional stressors from external sources or contact with other neural cells to produce RFs. This suggests that the human GFAP transgene is sufficient to induce the formation of RFs and that an excess of GFAP in astrocytes may be detrimental to normal neural development. In contrast, other studies have demonstrated that a lack of GFAP in astrocytes is not detrimental to normal breeding and development.153-155 We have previously shown that the metabolic turnover of GFAP is slow.2 Based on our present findings, we hypothesize that the normal mechanism for GFAP turnover may be insufficient to handle the excess GFAP produced by transgenic expression or under pathological conditions, thus resulting in the induction and accumulation of various stress proteins. Together, the aberrant intracellular deposition of GFAP and its attendant HSP chaperones constitute the astroglial RFs which accumulate in Alexander’s disease and other neuropathological conditions.

80

Astrocytes in Brain Aging and Neurodegeneration

Fig. 4.2. (A) Astrocytes in culture for 20 days from a Tg73.2 mouse were analyzed at the ultrastructural level. Note the dense Rosenthal fibers among the glial filaments. (B) Astrocytes from a 17 month old infant brain with Alexander’s disease were examined at the ultrastructural level. Note the dense deposits in the astrocyte which are identical to those seen in the Tg73.2 astrocyte cultures.

Glial Responses to Injury, Disease, and Aging

81

Fig. 4.3. Ultrastructural analysis of wild type (A) and Tg73.2 (B) astrocytes double-labeled with R-68 (12 nm gold particle) antiserum and SMI-21 (18 nm gold particle). The wild type astrocytes exhibit R-68 immunoreactivity only, whereas the Tg73.2 astrocytes exhibit immunostaining for both R-68 and SMI-21. The astrocyte cultures were fixed in 0.3 M NaCl in 70% aqueous ethanol.

Astrocytes in Brain Aging and Neurodegeneration

82

Fig. 4.4. (A) R-68 immunostain of a paraffin-embedded brain section from an infant diagnosed with Alexander’s disease. (B) R-68 immunostain of an insoluble pellet prepared from an Alexander’s disease brain.

Acknowledgments This chapter is supported by the Office of Research and Development, Medical Research Service, Department of Veterans Affairs and by NIH grant NS-11632. Dr. L.F. Eng is the Chief of the Chemistry Section, Pathology and Laboratory Service, Department of Veterans Affairs Medical Center, Palo Alto, CA 94304 and Professor of Pathology, Stanford University School of Medicine, Stanford, CA 94305.

References 1. Eng LF, Ghirnikar RS. GFAP and gliosis. Brain Pathol 1994; 4:229-237. 2. Eng LF, Lee YL, Intermediate filaments in astrocytes. In: Kettermann H, Ransom BR, eds. Neuroglia. Oxford: Oxford University Press, 1995; 650-667. 3. Eddleston M, Mucke L. Molecular profile of reactive astrocytes: Implications for their role in neurologic disease. Neuroscience 1993; 54:15-36. 4. Brenner M. Structure and transcriptional regulation of the GFAP gene. Brain Pathol 1994; 4:245-257. 5. Brenner M, Messing A. GFAP transgenic mice. Methods: A companion to methods in enzymology. 1996; 10:352-364. 6. Amaducci L, Forno KI, Eng LF. Glial fibrillary acidic protein in cryogenic lesions of the rat brain. Neurosci. Lett 1981; 21:27-32. 7. Jeneczko K. The proliferative response of astrocytes to injury in neonatal rat brain. A combined immunocytochemical and autoradiographic study. Brain Res 1988; 456:280-285. 8. Latov N, Nilaver G, Zimmerman EA et al. Fibrillary astrocytes proliferate in response to brain injury, a study combining immunoperoxidase technique for glial fibrillary acidic protein and radioautography of tritiated thymidine. Develop Biol 1979; 72:381-384.

Glial Responses to Injury, Disease, and Aging

83

9. Mathewson AJ, Berry M Observations on the astrocyte response to a cerebral stab wound in adult rats. Brain Res 1985; 327:61-69. 10. Miyake T, Hattori T, Fukuda M et al. Quantitative studies on proliferative changes of reactive astrocytes in mouse cerebral cortex. Brain Res 1988; 451:133-138. 11. Takamiya Y, Kohsaka S, Toya S et al. Possible association of platelet-derived growth factor (PDGF) with the appearance of reactive astrocytes following brain injury in situ. Brain Res 1986; 383:305-309. 12. Takamiya Y, Kohsaka S, Otani M et al. Immunohistochemical studies on the proliferation of reactive astrocytes and the expression of cytoskeletal proteins following brain injury in rats. Develop Brain Res 1988; 38:201-210. 13. Topp KS, Faddis BT, Vijayan VK. Trauma-induced proliferation of astrocytes in the brains of young and aged rats.Glia 1989; 2:201-211. 14. Hozumi I, Chiu FC, Norton WT. Biochemical and immunocytochemical changes in glial fibrillary acidic protein after stab wounds. Brain Res 1990; 524:64-71. 15. Vijayan VK, Lee Y, Eng F. Increase in glial fibrillary acidic protein following neural trauma. Mol Chem Neuropathol 1990; 13:107-118. 16. Brock TO, O’Callaghan JP. Quantitative changes in the synaptic vesicle proteins synapsin 1 and P38 and the astrocyte-specific protein glial fibrillary acidic protein are associated with chemical-induced injury to the rat central nervous system. J Neurosci 1987;7:931-942. 17. Rataboul P, Vernier P, Biguet NF et al. Modulation of GFAP mRNA levels following toxic lesions in the basal ganglia of the rat. Brain Res Bull 1989; 22:155-161. 18. Reinhard JF Jr, Miller DB, O’Callaghan JP. The neurotoxicant MPTP (1-methyl-4-phenyl1,2,3,6-tetrahydropyridine) increases glial fibrillary acidic protein and decreases dopamine levels of the mouse striatum: Evidence for glial response to injury. Neurosci Lett 1988; 95:246-251. 19. Aquino DA, Chiu FC, Brosnan CF et al. Glial fibrillary acidic protein increases in the spinal cord of Lewis rats with acute experimental autoimmune encephalitis. J Neurochem 1988; 51:1085-1096. 20. Goldmuntz EA. Brosnan CF, Chiu FC et al. Astrocytic reactivity and intermediate filament metabolism in experimental autoimmune encephalomyelitis: The effect of suppression with prazosin. Brain Res 1986; 397:16-26. 21. Smith ME, Somera FP, Eng LF. Immunocytochemical staining for glial fibrillary acidic protein and the metabolism of cytoskeletal proteins in experimental allergic encephalomyelitis. Brain Res 1983; 264:241-253. 22. Smith ME, Somera FP, Swanson, K. et al. Glial fibrillary acidic protein in acute and chronic relapsing experimental allergic encephalomyelitis (EAE). Prog Clin Biol Res 1984; 146:139-144. 23. Rataboul P, Biguet NF, Vernier P et al. Identification of a human glial fibrillary acidic protein cDNA: A tool for the molecular analysis of reactive gliosis in the mammalian central nervous system. J Neurosci Res 1988; 20:165-175. 24. Condorelli DF, Dell’Albani P, Kaczmarek L et al. Glial fibrillary acidic protein messenger RNA and glutamine synthetase activity after nervous system injury. J Neurosci Res 1990; 26:251-257. 25. Hozumi I, Aquino DA, Norton, WT. GFAP mRNA levels following stab wounds in rat brain. Brain Res 1990; 534:291-294. 26. Landry CF, Ivy GO, Brown IR. Effect of a discrete dorsal forebrain lesion in the rat on the expression of neuronal and glial-specific genes: Induction of calmodulin, NF-L, SC1, and GFAP mRNA. J Neurosci Res 1992; 32:280-289. 27. Poirier J, Hess M, May PC et al. Astrocytic apolipoprotein E mRNA and GFAP mRNA in hippocampus after entorhinal cortex lesioning. Mol Brain Res 1991; 11:97-106. 28. Kost-Mikucki SA, Oblinger MM. Changes in glial fibrillary acidic protein mRNA expression after corticospinal axotomy in the adult hamster. J Neurosci Res 1991; 28:182-191. 29. Aquino DA, Shafit-Zagardo B, Brosnan CF et al. Expression of glial fibrillary acidic protein and neurofilament mRNA in gliosis induced by experimental autoimmune encephalomyelitis. J Neurochem 1990; 54:1398-1404.

84

Astrocytes in Brain Aging and Neurodegeneration

30. Steward O, Torre ER, Phillips LL et al. The process of reinnervation in the dentate gyrus of adult rats: Time course of increases in mRNA for glial fibrillary acidic protein. J Neurosci 1990; 10:2373-2384. 31. Calne DB. Neurodegenerative Diseases. Philadelphia: WB Saunders, 1994. 32. Duffy PE, Rapoport M, Graf L. Glial fibrillary acidic protein and Alzheimer-type senile dementia. Neurology 1980; 30:778-782. 33. Schechter R, Yen S-HC, Terry RD. Fibrous astrocytes in senile dementia of the Alzheimer type. J Neuropathol Exp Neurol 1981; 40:95-101. 34. Mancardi GL, Liwnicz BH, Mandybur TI. Fibrous astrocytes in Alzheimer’s disease and senile dementia of Alzheimer’s type. An immunohistochemical and ultrastructural study. Acta Neuropathol (Berl) 1983; 61:76-80. 35. Beach TB, McGeer EG. Lamina-specific arrangement of astrocytic gliosis and senile plaques in Alzheimer disease visual cortex. Brain Res 1988; 463:357-361. 36. Beach TG, Walker R, McGeer EG. Patterns of gliosis in Alzheimer’s disease and aging cerebrum. Glia 1989; 2:420-436. 37. Mandybur TI. Cerebral amyloid angiopathy and astrocytic gliosis in Alzheimer’s disease. Acta Neuropathol (Berlin) 1989; 78:329-331. 38. Vijayan V, Geddes JW, Anderson KJ et al. Astrocyte hypertrophy in the Alzheimer’s disease hippocampal formation. Exp Neurol 1991; 112:72-78. 39. Joachim CL, Morris JH, Selkoe DJ. Diffuse senile plaques occur commonly in the cerebellum in Alzheimer’s disease. Am J Pathol 1989; 135:309-319. 40. Rozemuller JM, Eikelenboom P, Stam FC, Beyreuther K, Masters CL. A4 protein in Alzheimer’s disease: Primary and secondary cellular events in extracellular amyloid deposition. J Neuropathol Exp Neurol 1989; 48:674-691. 41. Suenaga T, Hirano A, Llena JF et al. Modified immunocytochemical studies in cerebellar plaques in Alzheimer’s disease. J Neuropathol Exp Neurol 1990; 49:31-40. 42. Malamud N. Neuropathology of organic brain syndromes associated with aging. In: Gaitz C, ed. Aging and the Brain. New York: Plenum, 1972:63-87. 43. Burger PC, Vogel S. The development of pathologic changes of Alzheimer’s disease and senile dementia in patients with Down’s syndrome. Am J Pathol 1973; 73: 457-476. 44. Mann DMA. The pathological association between Down’s syndrome and Alzheimer’s disease. Mech Aging Dev 1988; 43:99-136. 45. Murphy GM, Eng LF, Ellis WG et al. Antigenic profile of plaques and neurofibrillary tangles in the amygdala in Down’s syndrome: A comparison with Alzheimer’s disease. Brain Res 1990; 537:02-108. 46. Murphy GM Jr, Murphy E, Greenberg BD et al. Alzheimer’s disease: β-amyloid precursor protein expression in plaques varies among cytoarchitectonic areas of the medial temporal lobe. Neurosci Lett 1991; 131:100-104. 47. Murphy GM Jr, Ellis WG, Lee YL et al. Astrocytic gliosis in the amygdala in Down’s syndrome and Alzheimer’s disease. In: Yu ACH, Hertz L, Norenberg MD, Sykova E, Waxman SG, eds. Progress in Brain Research, Vol. 94. Amsterdam: Elsevier Science Publishers B.V., 1992:475-483. 48. Michetti F, Larocca LM, Rinelli A et al. Immunocytochemical distribution of S-100 protein in patients with Down’s syndrome. Acta Neuropathol (Berlin) 1990; 80:475-478. 49. Griffin WST, Stanley LC, Ling C et al. Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc Natl Acad Sci USA 1989; 86:7611-7615. 50. Goodison KL, Parhad IM, White CL III et al. Neuronal and glial gene expression in neocortex of Down’s syndrome and Alzheimer’s disease. J Neuropathol Exp Neurol 1993; 52:192-198. 51. Moossy J, Martinez AJ, Hanin I et al. Thalamic and subcortical gliosis with dementia. Arch Neurol 1987; 44:510-513. 52. Selkoe DJ, Salazar FJ, Abraham C et al. Huntington’s disease: Changes in striatal proteins reflect astrocytic gliosis. Brain Res 1982; 245:117-125.

Glial Responses to Injury, Disease, and Aging

85

53. Stevens CD, Altshuler LL, Bogerts B et al. Quantitative study of gliosis in schizophrenia and Huntington’s chorea. Biol Psychiat 1988; 24:697-700. 54. Roberts GW, Colter N, Lofthouse R et al. Gliosis in schizophrenia: A survey. Biol Psychiat 1986; 2:1043-1050. 55. Stevens JR. Neuropathology of schizophrenia. Arch Gen Psychiatry 1982; 39:1131-1139. 56. Bigbee JW, Bigner DD, Pegram C et al. Study of glial fibrillary acidic protein in a human glioma cell line grown in culture and as a solid tumor. J Neurochem 1983; 40:460-467. 57. DeArmond SJ, Fajardo M, Naughton SA et al. Degradation of glial fibrillary acidic protein by a calcium dependent proteinase: An electroblot study. Brain Res 1983; 262:275-282. 58. Kushner PD, Stephenson DT, Wright S. Reactive astrogliosis is widespread in the subcortical white matter of amyotrophic lateral sclerosis brain. J Neuropathol Exp Neurol 1991; 50:263-277. 59. Dormont D, Delpech A, Courcel M-N et al. Hyperproduction de proteine glio-fibrillaire acide (GFA) au cours de l’évolution de la tremblante experimentale de la souris. C R Acad Sci (Paris) 1981; 293:53-56. 60. Mackenzie A. Immunohistochemical demonstration of glial fibrillary acidic protein in scrapie. J Comp Path 1983; 3:251-259. 61. Gomori AF, Partnow MJ, Horoupian DS et al. The ataxic form of Creutzfeldt-Jakob disease. Arch Neurol 1973; 29:318-323. 62. Lafarga M, Berciano MT, Suarez I et al. Reactive astroglia-neuron relationships in the human cerebellar cortex: A quantitative, morphological and immunocytochemical study in Creutzfeldt-Jakob disease. Int J Devl Neurosci 1993; 11:199-213. 63. Manuelidis L, Tesin DM, Sklaviadis T et al. Astrocyte gene expression in Creutzfeldt-Jakob disease. Proc Natl Acad Sci USA 1987; 84:5937-5941. 64. DeArmond SJ, Kristensson K, Bowler RP. In: Yu ACH, Hertz L, Norenberg MD, Sykova E, Waxman SG, eds. Progress in Brain Research, Vol. 94. Amsterdam: Elsevier Science Publishers 1992:437-446. 65. Sobel RA, DeArmond SJ, Forno LS et al. Glial firillary acidic protein in hepatic encephalopathy. An immunohistochemical study. J Neuropathol Exp Neurol 1981;40: 625-632. 66. Kretzschmar HA, DeArmond SJ, Forno LS. Measurement of GFAP in hepatic encephalopathy by ELISA and transblots. J Neuropathol Exp Neurol 1985; 44:459-471. 67. Norenberg MD. The role of astrocytes in hepatic encephalopathy. Neurochem Pathol 1987; 6:13-33. 68. Ma KC, Ye ZR, Fang J et al. Glial fibrillary acidic protein immunohistochemical study of Alzheimer I & II astrogliosis in Wilson’s disease. Acta Neurol Scand 1988; 78:290-296. 69. Perry VH, Andersson PB, Gordon S. Macrophages and inflammation in the central nervous system. Trends Neurosci 1993; 16:268-273. 70. Gehrmann J, Matsumoto Y, Kreutzberg GW. Microglia: Intrinsic immunoeffector cells of the brain. Brain Res Rev 1995; 20:269-287. 71. Streit WJ, Graeber MB, Kreutzberg GW. Functional plasticity of microglia: A review. Glia 1988; 1:301-307. 72. Dickson DW, Mattiace LA, Kure K et al. Microglia in human disease, with an emphasis on acquired immunodeficiency syndrome. Lab Invest 1991; 64:135-156. 73. Giulian D, Chen J, Ingwman JE et al. The role of mononuclear phagocytes in wound healing after traumatic injury to adult mammalian brain. J Neurosci 1989; 9:4416-4429. 74. McGeer PL, Itagaki S, McGeer EG. Expression of the histocompatibility glycoprotein HLADR in neurological disease. Acta Neuropathol 1988; 76:550-557 75. Gehrmann J, Mies G, Bonnekoh P et al. Microglial reaction in the rat cerebral cortex induced by cortical spreading depression. Brain Pathol 1993; 3:11-17. 76. Kettenmann H, Hoppe D, Gottman K et al. Cultured microglial cells have a distinct pattern of membrane channels different from peritoneal macrophages. J Neurosci Res 1990; 26:278-287. 77. Kettenmann H, Banati R, Walz W. Electrophysiological behavior of microglia. Glia 1993; 7:93-101.

86

Astrocytes in Brain Aging and Neurodegeneration

78. Imamura K, Suzumura A, Sawada M et al. Induction of MHC class II antigen expression on murine microglia by interleukin 3. J Neuroimmunol 1994; 55:119-125. 79. Eng LF, Ghirnikar RS, Lee YL. Inflammation in EAE: Role of chemokine/cytokine expression by resident and infiltrating cells. Neurochem Res 1966; 21:511-525. 80. Nohava K, Malipiero U, Frei K et al. Neurons and neuroblastoma as a source of macrophage colony-stimulating factor. Eur J Immunol 1992; 22:2539-2545. 81. Hayashi M, Luo Y, Laning JH et al. Production and function of monocyte chemoattractant protein-1 nd other β-chemokines in murine glial cells. J Neuroimminol 1995; 60:143-150 82. Perry VH, Matyszak MK, Fearn S. Altered antigen expression in the aged rodent CNS. Glia 1993; 7:60-59. 83. Ogura K, Ogawa M, Yoshida M. Effect of aging on microglia in the normal rat brain: Immunohistochemical observations. Neuroreport 1994; 5:1224-1226. 84. Dijkstra CD, DeGroot CJ, Huitinga I. The role of macrophages in demyelination. J Neuroimmunol 1992; 40:183-188. 85. Hulkower K, Brosnan C F, Aquino DA et al. Expression of CSF-1, c-fms, and MCP-1 in the central nervous system of rats with experimental allergic encephalomyelitis. J Immunol 1993; 150:2525-2533. 86. McCombe PA, de Jersey J, Pender MP. Inflammatory cells, microglia and MHC class II antigen-positive cells in the spinal cord of Lewis rats with acute and chronic relapsing experimental autoimmune encephalomyelitis. J Neuroimmunol 1994; 51:153-167. 87. Norton WT, Brosnan CF, Cammer W et al. Mechanisms and suppression of inflammatory demyelination. Acta Neurobiol Exp 190; 50:225-235. 88. Claudio L, Kress Y, Factor J et al. Mechanisms of edema formation in experimental autoimmune encephalomyelitis. The contribution of inflammatory cells. Am J Pathol 1990; 137:1033-1045. 89. Claudio L, Martiney J A, Brosnan CF. Ultrastructural studies of the blood-retina barrier after exposure to interleukin-1 beta or tumor necrosis factor-alpha. Lab Invest 1994; 70:850-861. 90. Knobler RL, Marini JC, Goldowitz D et al. Distribution of the blood brain barrier in heterotopic brain transplants and its relationship to the lesions of EAE. J Neuropathol Exp Neurol 1992; 51:36-39. 91. Cannella B, Cross AH, Raine CS. Relapsing autoimmune demyelination: A role for vascular addressins. J Neuroimmunol 1991; 35:295-300. 92. Cannella B, Cross AH, Raine CS. Adhesion related molecules in the central nervous system. Upregulation correlates with inflammatory cell influx during relapsing experimental autoimmune encephalomyelitis. Lab Invest 1991; 65:23-31. 93. Raine CS, Cannella B, Duijvestijn A M et al. Homing to central nervous system vasculature by antigen-specific lymphocytes. II. Lymphocyte/endothelial cell adhesion during the initial stages of autoimmune demyelination. Lab Invest 1990; 63:476-489. 94. Seki T, Joh K, Oh-ishi T. Augmented production of interleukin-8 in cerebrospinal fluid in bacterial menengitis. Immunology 1993; 80:333-335. 95. Nichols NR, Day JR, Laping NJ et. al. GFAP mRNA increases with age in rat and human brain. Neurobiol Aging 1993; 14:421-429. 96. Landfield PW, Waymire JL, Lynch GS. Hippocampal aging and adrenocorticoids: Quantitative correlations. Science 1978; 202:1098-1102. 97. Gordon MN, Morgan DG. Increased GFAP expression in the aged brain does not result from increased astrocyte density. Soc Neurosci Abstr.1993;19:1042. 98. Goss JR, Finch CE, Morgan DG. GFAP RNA prevalence is increased in aging and in wasting mice. Exp Neurol 1990; 108:266-268. 99. Goss JR, Finch CE, Morgan DG. Age-related changes in glial fibrillary acidic protein mRNA in mouse brain. Neurobiol Aging 1991; 12:165-170. 100. O’Callaghan JP, Miller DB. The concentration of glial fibrillary acidic protein increases with age in the mouse and rat brain. Neurobiol Aging 1991; 12:171-174. 101. Bronson RT, Lipman RD, Harrison DE. Age-related gliosis in the white matter of mice. Brain Res 1993; 609:124-128.

Glial Responses to Injury, Disease, and Aging

87

102. Laping NJ, Morgan TE, Nichols NR et al. Transforming growth factor-1 induces neuronal and astrocyte genes; Tubulin alpha-1, glial fibrillary acidic protein, and clusterin. Neuroscience 1994; 58:563-572. 103. Kohama SG, Goss JR, Finch CE et al. Increases of glial fibrillary acidic protein in the aged female mouse brain. Neurobiol Aging 1995; 16:59-67 104. Yoshida T, Goldsmith SK, Morgan TE et al. Transcription supports age-related increases of GFAP gene expression in the male rat brain. Neurosci Ltt 1966; 215:107-110. 105. Hallpike JF, Adams CWM, Tourtellotte WW. Multiple Sclerosis. Pathology, Diagnosis and Management. Baltimore: Williams & Wilkins, 1983. 106. Schaumburg HH, Powers JM, Raine CS et al. Adrenoleukodystrophy: A clinical and pathological study of 17 cases. Arch Neurol 1975; 33: 577-591. 107. Goldman JE, Schaumburg HH, Norton WT. Isolation and characterization of glial filaments from human brain. J Cell Biol 1978; 78:426-440. 108. Kraig RP, Dong L, Thisted R. Spreading depression increases immunohistochemical staining of glial fibrillary acidic protein J Neurosci 1991; 11:2187-2198. 109. Major DE, Kessiak JP, Cotman CW, et.al. Life-long dietary restriction attenuates age-related increases in glial fibrillary acidic protein. GFAP mRNA in the rat hippocampus. Soc Neurosci abstr 1994; 20:50 110. Nichols RN, Caleb EF, Nelson JF. Food restriction delays the age-related increase in GFAP mRNA in rat hypothalamus. Neurobiol Aging 1995; 16:105-110. 111. Morgan TE, Rozovshy I, Goldsmith SK et al. Increased transcription of the astrocyte gene GFAP during middle-age is attenuated by food restriction: Implications for the role of oxidative stress. Free Rad Biol Med 1997; 23:524-528. 112. Tacconi MT, Lligona L, Salmona M et al. Aging and food restriction: Effects on lipids of the cerebral cortex. Neurobiol Aging 1991; 12:55-59. 113. Choi JH, Yu BP. Brain synaposomal aging: Free radicals and membrane fluidity. Free Rad Biol Med 1995; 18:133-139. 114. Laping NJ, Teter B, Nichols NR et al. Glial fibrillary acidic protein: Regulation by hormones, cytokines, and growth factors. Brain Pathol 1994b; 4:259-275. 115. Stam FC, Roukema PA. Histochemical and biochemical aspects of corpora amylacea. Acta Neuropathol 1973; 25:95-102. 116. Steyaert A, Cissé S, Merhi Y et al. Purification and polypeptide composition of corpora amylacea from aged human brain. J Neurosci Meth 1990; 31:59-64. 117. Martin JE, Mather K, Swash M et al. Heat-shock protein expression in corpora amylacea in the central nervous system—clues to their origin. Neuropathol Appl Neurobiol 1991; 17:113-119. 118. Cissé S, Perry G, Lacoste-Royal G et al. Immunochemical identification of ubiquitin and heat-shock proteins in corpora amylacea from normal aged and Alzheimer’s disease brains. Acta Neuropathol 1993; 85:233-240. 119. Sakai M, Austin J, Witmer F et al. Studies of corpora amylacea. Arch Neurol 1969; 21:526-544. 120. Singhrao SK, Neal JW, Newman GR. Corpora amylacea could be an indicatior of neurodegeneration. Neuropathol Appl Neurbiol 1993; 19:269-276. 121. Singhrao SK, Neal JW, Piddlesdent SJ et al. New immunocytochemical evidence for neuronal/oligodendroglial origin for corpora amylacea. Neuropathol Appl Neurol 1994; 20:66-73. 122. Tokutake H, Nagase TH, Morisaki S et al. X-ray microprobe analysis of corpora amylacea. Neuropathol Appl Neurobiol 1995; 21:269-273. 123. Ramsay HJ. Ultrastructure of corpora amylacea. J Neuropath Exp Neurol 1956; 24:25-39. 124. Takahashi K, Agari M, Nakamura H. Intra-axonal corpora amylacea in ventral and lateral horns of the human spinal cord. Acta Neuropathol (Berlin) 1975; 31:151-158. 125. Anzil AP, Herrlinger H, Blinzinger Ket al. Intraneuritic corpora amylacea. Demonstration in orbital cortex of elderly subjects by means of early post-mortem brain sampling and electron microscopy. Virchows Archives of Pathological Anatomy and Histology 1974; 364:297-302.

88

Astrocytes in Brain Aging and Neurodegeneration

126. Singhrao SK, Morgan BP, Neal JW et al. A functional role for corpora amylacea based on evidence from complement studies. Neurodegeneration 1995; 4:335-345. 127. Cissé S, Lacoste-Royal G, Laperriere G et al. Ubiquitin is a component of polypeptides purified from corpora amylacea of aged human brain. Neurochemical Res 1991; 16:429-433 128. Sbarbati A, Carner M, Colletti V et al. Extrustion of corpora amylacea from the marginal glia at the vestibular root entry zone. J Neuropath Exp Neurol 1996; 55:196-201. 129. Schipper HM, Cissé S. Mitochondrial constituents of corpora amylacea and autofluorescent astrocytic inclusions in senescent human brain. Glia 1995; 14:55-64 130. Schipper HM, Cissé S, Stopa EG. Expression of heme oxygenase-1 in the senescent and Alzheimer-diseased brain. Ann Neurol 1995; 37:758-768. 131. Chopra VS, Chalifour LE, Schipper HM. Differential effects of cysteamine on heat shock protein induction and cytoplasmic granulation in astrocytes and glioma cells. Molecular Brain Res 1995; 31:173-18. 132. Cissé S, Schipper HM. Experimental induction of corpora amylacea-like inclusions in rat astroglia. Neuropathol Appl Neurobiol 1995; 21:423-431. 133. Schipper HM. Astrocytes, brain aging, and neurodegeneration. Neurobiol Aging 1996; 17:467-480. 134. Chin SM, Goldman JE. Glia inclusions in CNS degenerative diseases. J Neuropathol Exp Neurol 1996; 55:499-508. 135. Papp MI, Kahn JE, Lantos PL. Glial cytoplasmic inclusions in the CNS of patients with multiple system atrophy (striatonigral degeneration, olivopontocerebellar atrophy and ShyDrager syndrome). J Neurol Sci 1989; 94:79-100. 136. Nakazato Y, Yamazaki H, Hirato J et al. Oligodendroglial microtubular tangles in olivopontocerebellar atrophy. J Neuropath Exp Neurol 1990; 49:521-530. 137. Kato S, Nakamura H, Hirano A et al. Argyrophilic ubiquitinated cytoplasmic inclusions of Leu-7-positive glial cells in olivopontocerebellar atrophy (multiple system atrophy). Acta Neuropath 1991; 82:488-493. 138. Murayama S, Arima K, Nakazato Y et al. Immunocytochemical and ultrastructural studies of neuronal and oligodendroglial cytoplasmic inclusions in multiple system atrophy. 2. Oligodendroglial cytoplasmic inclusions. Acta Neuropath 1992; 84:32-38. 139. Arai N, Nishimura M, Oda M et al. Immunohistochemical expression of microtubule-associated protein 5 (MAP5) in glial cells in multiple system atrophy. J Neurol 1992; 109;102-106. 140. Abe H, Yagishita S, Amano N et al. Argyrophilic glial intracytoplasmic inclusions in multiple system atrophy: Immunocytochemical and ultrastructural study. Acta Neuropathol 1992; 273-277. 141. Alexander WS. Progressive fibrinoid degeneration of fibrillary astrocytes associated with mental retardation in a hydrocephalic infant. Brain 1949; 72:373-381. 142. Herndon RM, Rubinstein LV, Freeman JM, Mathieson G. Light and electron microscopic observations on Rosenthal fibres in Alexander’s disease and in multiple sclerosis. J Neuropathol Exp Neurol 1970; 29:524-550. 143. Grcevic N, Yates PO. Rosenthal fibres in tumours of the central nervous system. J Pathol Bact 1957; 73:467-472. 144. Borrett G, Becker LE. Alexander’s disease: A disease of astrocytes. Brain 1985; 108:367-385. 145. Iwaki T, Kume-Iwaki A, Liem RKH, Goldman J. αB-crystallin is expressed in non-lenticular tissues and accumulates in Alexander’s disease. Cell 1989; 57:71-78. 146. Tomokane N, Iwaki T, Tateishi J et al. Rosenthal fibers share epitopes with αB-crystallin, glial fibrillary acidic protein and ubiquitin, but not with vimentin: Immunoelectron microscopy with colloidal gold. Am J Pathol 1991; 138:875-885. 147. Goldman JE, Corbin E. Rosenthal fibers contain ubiquitinated αB-crystallin. Am J Pathol 1991; 139:933-938. 148. Head M, Corbin E, Goldman JE. Overexpression and abnormal modification of the stress proteins αB-crystallin and Hsp27 in Alexander’s disease. Am J Pathol 1993; 143:1743-1753. 149. Head MW, Corbin E, Goldman JE. Coordinate and independent expression of αB-crystallin and hsp27. J Cell Physiol 1994; 159:41-50.

Glial Responses to Injury, Disease, and Aging

89

150. Messing A., Galbreath EJ, Sijapati KK, Brenner M. Overexpression of GFAP in transgenic mice. J Neuropath Exp Neurol 1996; 55:620 (abstract). 151. Messing A, Head MW, Galles K, Galbreath EJ, Goldman UE, Brenner M. Fatal encephalopothy with astrocyte inclusions in FGAP transgenic mice. Am J Pathol 1998; 152:391-398. 152. Ramsay P, Norman M, Eng LF. Chemical study of an Alexander brain. Trans Am Soc Neurochem 10:125, 1979, Abstract. 153. Gomi H., Yokoyama T, Fujimoto K et al. Properties of astrocytes from mice devoid of GFAP. Mice devoid of the glial fibrillary acidic protein develop normally and are susceptible to scrapie prions. Neuron 1995; 14:29-41. 154. Pekny M, Leveen P, Pekna M et al. Mice lacking glial fibrillary acidic protein display astrocytes devoid of intermediate filaments but develop and reproduce normally. EMBO J 1995; 14:1590-1598. 155. McCall. MA, Gregg. RG, Behringer RR et al. Targeted deletion in astrocyte intermediate filament (GFAP) alters neuronal physiology. Proc Natl Acad Sci USA 1996; 93:6361-6366.

CHAPTER 5

Astrocyte Pathology in Alzheimer Disease Jerzy Wegiel and Henryk M. Wisniewski

Neuropathological Changes in Alzheimer Disease

A

lzheimer disease (AD) is a degenerative cerebral disorder with progressive dementia. In the brain of subjects with Alzheimer-type dementia, several percent of the volume of the gray matter is infiltrated with 39- to 43-amino-acid amyloid-β (Aβ), which forms parenchymal diffuse nonfibrillar deposits and fibrillar neuritic plaques. In AD, microglial cells change morphology, immunophenotype, function, and distribution. The subpopulation of microglial cells produces fibrillar amyloid deposits in classical and primitive plaques.1-3 Aβ deposits in the wall of capillaries are the product of perivascular cells and perivascular microglial cells,4-5 both of which are of monocyte/microglial cell lineage.4,6-8 Neurons are the source of parenchymal nonfibrillar amyloid in diffuse plaques.9-11 Nonfibrillar and fibrillar amyloid deposits in the tunica media of leptomeningeal and parenchymal arteries and veins are the product of smooth muscle cells.12-13 Fibrillar parenchymal amyloidosis is associated with such secondary changes as neuronal degeneration and astrocytosis.14 The effect of vascular amyloidosis is local impairment in blood circulation, neuronal degeneration and loss, and astrocytosis.4 The second hallmark of AD is neurofibrillary degeneration with cytoplasmic accumulation of abnormally phosphorylated tau-protein and paired helical filaments (PHFs), which cause severe, structure-specific neuronal loss reaching up to 80-90% in the hippocampal formation15 and up to 50% in the neocortex.16 During the course of AD, the number of astrocytes increases by several times.17 In part, these changes are related to the hallmarks of AD—amyloidosis-β and neurofibrillary degeneration of neurons. Activated astrocytes are engaged in the dispersion, degradation, and removal of Aβ14 as well as in the removal of ghost tangles and dark neurons. Astrocyte pathology in AD comprises not only their proliferation and activation but also several forms of degeneration, such as cytoplasmic accumulation of PHFs, Rosenthal fibers, anchorage densities with desmosome-like structures, eosinophilic inclusions, and corpora amylacea.18 This complex Alzheimer-type pathology is the result of numerous interactions between amyloid-β, neurons affected and not affected by neurofibrillary changes, microglial cells, and astrocytes.

Relationships Between Amyloid-β, Neurons, and Glial Cells in AD Neurons, astrocytes, microglial cells, and oligodendrocytes maintain a complex parenchymal milieu in the normal brain. Amyloidosis-β and neurofibrillary pathology modify Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

92

Astrocytes in Brain Aging and Neurodegeneration

the brain environment and initiate several reinforcing feedbacks loops, which change the interneuronal and interglial balance. Astrocytes produce apolipoprotein E (Apo E),19-20 interleukin-1,21-22 endothelin-1,23 prostaglandin E, and a subset of these cells expresses intercellular adhesion molecule-1.24 In part, their function depends on microglial cells, which produce growth factors including interleukin-1,25-27 which in turn induces astrogliosis and activates astrocytes.28-29 The activation of astrocytes is associated indirectly with another form of microglial cell function, namely, with the production of several cytotoxins—glutamate, tumor necrosis factor α (TNF~α), nitric oxide, hydrogen peroxide, and oxygen-containing free radicals—which affect neurons.30-31 Neuronal degeneration and death cause both activation of astrocytes and astrocytosis.

Interleukin 1 Microglial cells of the neuritic plaques are activated, and they overexpress acute-phase cytokine interleukin-1. Interleukin-1 beta has a strong mitogenic impact on cultured astrocytes.32 In tissue, this cytokine activates astrocytes and induces expression of the astrocytederived cytokine, S-100β, which increases the intraneuronal free calcium level and may cause neuronal injury and death. Interleukin-1 upregulates expression and processing of βAPP, favoring amyloid-β deposition,33 and induces expression of protease inhibitor alpha-1antichymotrypsin, thromboplastin, complement protein C, and Apo E, all of which are present in neuritic plaques. Interleukin-1 induces increased synthesis of alpha-1antichymotrypsin, which acts as a pathological chaperone,34-35 binding to the beta protein and strongly promoting its polymerization into amyloid filaments in vitro.36

Apolipoprotein E Both astrocytes and microglia produce Apo E.19-20 The synthesis of Apo E increases after neuronal injury in the central37 and peripheral nervous system.38 Apo E interacts with both normal soluble Aβ and fibrillar amyloid in plaques.39 More than 60% of persons with one allelic form of Apo E4 suffer from AD by the time they reach 75 years of age, and more than 90% of the subjects with two copies of the Apo E4 gene have the disease by 75 years of age.40 Apo E4 is an important pathological chaperon protein in soluble amyloid-beta protein fibrillization and tau phosphorylation.39,41 In vitro studies show that Apo E4 binds Aβ faster and with a different pH dependence than Apo E3.42 Persons who inherit two Apo E4 genes bind more Aβ to form plaques.41

Amyloid-β The amyloidogenic processing of APP might be upregulated by extracellular Aβ43-44 through the cellular receptor for Aβ.45-46 Astrocytes recognize extracellular amyloid and respond to this antigen.14 In vitro studies indicate that Aβ enhances the secretion of interleukin-147-48 and basic fibroblast growth factor (bFGF) from cultured microglia and astrocytes as well as proliferation of microglial cells. That interleukin-1 and bFGF elevate the synthesis of βAPP suggests that this cascade effect contributes to plaque development.47 In cultured astrocytes, all three major transcripts of βAPP are expressed, with the ratio for APP 695, APP 751, and APP 770 isoform mRNAs being 1:4:2. Treatment with transforming growth factor beta 1 (TGF-beta 1) produces about a 6-fold increase in total APP mRNA.32

Endothelin Endothelin-1 immunoreactive astrocytes are very rare in non-AD cases but are very numerous in AD brains in the periphery of plaques, in the molecular layer of the cerebral cortex, in the subcortical white matter, and the folia of cerebellum.23 Endothelin-1 expres-

Astrocyte Pathology in Alzheimer Disease

93

sion increases not only in AD but also in infarcts and traumatic injuries.49 Astrocytes grown in vitro also release endothelin-1 and -3 into the culture medium.50-53 In vitro endothelin acts as a mitogen for astrocytes and tumor cells.51,54 Receptors to endothelin-1 were identified in the plasma membrane of cultured astrocytes of rats and mice51 and to endothelin-3 in rat astrocytes.55-56 It has been suggested that endothelins may modulate neuronal activity.57-59 Endothelin-1 is considered the most potent and long-lasting vasoconstrictor peptide known to date.60-64 Endothelins discharged from reactive astrocytes in AD brains may induce constriction of arterioles and contribute to the local reduction of blood flow.65-66 Simultaneous expression of endothelin and endothelin receptors in primary astrocytes implies the presence of an autocrine control mechanism in astrocytes.52 These cytokines and the molecular and cellular processes that they support form a complex of interactions that may be capable of self-propagation facilitated by means of several reinforcing feedback loops (Fig. 5.1).

Astrogliosis in Aging and AD Astrocytic gliosis, characterized by cellular hypertrophy and augmented GFAP expression, is a morphological marker of cerebral aging.67 These changes occur not only in the brains of aged, nondemented people and in individuals with Alzheimer disease,68 but also in the brains of aged monkeys,69 rats,70-74 and mice.67,75 Astrogliosis in brains of species free of plaques, such as mice, the great topographical variability of astrogliosis, and the presence of astrocytosis in many other pathological processes indicate that astrocyte proliferation is a regional reaction elicited by many different factors. In the brain of subjects with Alzheimer disease, focal and diffuse astrocytosis develops.17,76-84 In advanced stages of AD, the number of astrocytes increases approximately 4-fold.17 In AD brains, the astrocytes appear in plaques, around ghost tangles, dark neurons, capillaries obliterated by Aβ, and in areas of ischemic changes. Astrocytes reveal specific morphological properties in each of the above-listed pathological interactions, which indicates that their role in each of these events may be different. Astrocytes surrounding plaques are usually so rich in glial fibrillary acidic protein-positive fibers17,80 that the number of plaques demonstrated by GFAP is usually higher than the number obtained with other techniques.83 Astrocytic accumulation in plaque is considered the reaction to focal extracellular accumulation of Aβ.1,67,79,83,85 Because almost all fibrillar plaques develop in gray matter, Aβ-associated astrocytosis is restricted to the cerebral cortex and subcortical gray matter.

Neuritic Plaques Light and electron microscopic studies show that neuritic plaques consist of amyloid; dystrophic, degenerating and regenerating neuronal processes; astrocytes; and microglial cells. Plaques with a central amyloid core, called also amyloid star, surrounded by six to eight microglial cells, degenerated neurites, and a ring of astrocytic processes, are called classical plaques. Primitive plaques do not have amyloid star and are composed of wisps of amyloid associated with one to four microglial cells, and chaotically distributed bundles of degenerated neurites and astrocytic processes. Two or three astrocytes are involved directly in classical plaque formation.14,17,86 Rozemuller and coworkers87 hypothesized that the proteolytic cleavage of amyloidogenic proteins and formation of amyloid fibrils are related to astroglia. However, ultrastructural studies indicate that microglial cells are the cells engaged in fibrillar Aβ deposition in classical and primitive plaques.1-4 The spatial distribution of proliferating astrocytic processes in the plaque indicates that astrocytic reaction in the classical plaque relates mainly to the amyloid deposits. Astrocyte receptors recognize many molecules,88-89 and extracellular amyloid appears also to be detectable by astrocytes. Astrocytic processes proliferate in the plaque periphery, isolate Aβ

Fig. 5.1. Schematic diagram showing the relationships between astrocytes, microglial cells, and neurons in AD.

94 Astrocytes in Brain Aging and Neurodegeneration

Astrocyte Pathology in Alzheimer Disease

95

aggregates, and divide the latter into smaller clusters (Fig. 5.2). Transformation of fibrillar amyloid in floccular and amorphous material in areas of proliferating astrocytic processes suggests that astrocyte ectoenzymes degrade Aβ. The proportion 1:9.2 between the volume of amyloid and the volume of proliferating astrocytic processes appears to be the measure of astrocytic response to Aβ in classical plaques.

Diffuse Plaques Astrocytic reaction is undetectable in diffuse plaques, which are nonfibrillar and thioflavin S and Congo red negative. Their formation is associated with neuronal release of Aβ peptide. Nonfibrillar amyloid is present in relatively low concentrations, is probably bound to chaperon proteins thereby preventing fibrillization, and appears not to be able to activate surrounding astrocytes. As a result, in diffuse plaques, there is no sign of local increases in the numbers of astrocytes or astrocyte activation.

Astrogliosis in the Area of Amyloid Angiopathy Deposition of amyloid in the wall of the capillary vessels by perivascular cells and perivascular microglial cells is the cause of vessel obliteration. Amyloid deposits and remnants of the capillary wall are surrounded by astrocytic processes and degraded. The presence of only amyloid deposits indicates that this material is more resistant to degradation than endothelial cell residues.4

Astrocyte Response to Neurofibrillary Changes Neurofibrillary degeneration appears to be the major cause of neuronal loss in such brain structures as the hippocampal formation,15 but much less so in the isocortex.16 Light and electron microscopic studies indicate that in the end stage of neurofibrillary changes neurons are degraded by astrocytic processes that penetrate ghost tangles,84,90 separate aggregates of cellular debris and bundles of PHF. The pathological fibrillar component is the most resistant residue in neurons showing neurofibrillary changes. Scars formed by astrocytes and residues of NFTs are often encountered in cortical biopsy specimens derived from AD subjects (Fig. 5.3). In the end stage of ghost tangle resolution, the characteristic features of PHF disappear, and fibrils are not detectable with antibodies to abnormally phosphorylated tau. They are detectable with antibody (mAb 3-39) to ubiquitin bound to PHF (50-65 aa of ubiquitin). Degradation of the ghost tangles is probably a very slow process. The differences in the number of ghost tangles—very few in the isocortex16 and very numerous in the cornu Ammonis or subiculum15—suggest that the rate of neuronal death associated with neurofibrillary changes and the local reaction of astrocytes vary and are structure specific.

Astrocytes Response to Ischemia-Related Neuronal Degeneration The second factor contributing to neuronal loss in AD is amyloid angiopathy. The deposition of amyloid-β in the wall of capillary vessels causes their obliteration and regional ischemic changes, with neuronal degeneration and death.4-5 These changes also are associated with astrocyte proliferation and activation. Interaction between astrocytes and dark neurons similar to the interaction between astrocytes and ghost tangles suggests that astrocytes are involved in removal of bodies and processes of necrotic neurons with and without neurofibrillary changes (Fig. 5.4). The number and size of astrocytic processes surrounding dark neuronal perikarya and neurites increases. The volume of cytoplasm in the neuronal perikarya is reduced and the hyaloplasm is condensed, whereas astrocytic processes become distended.

96

Astrocytes in Brain Aging and Neurodegeneration

Fig. 5.2. Periphery of a classical plaque with aggregates of fibrillar amyloid-β (asterisks), which are separated from the neuronal processes by a dense network of astrocytic processes (As). Neuronal processes (np), which are in direct contact (arrows) with amyloid deposits (asterisks), reveal degenerative changes, with an accumulation of abnormal mitochondria, osmophilic bodies, and vacuoles and an increase in the diameter of the processes.

Astrocyte Pathology in Alzheimer Disease

97

Fig. 5.3. Tangential (arrows) and longitudinal (double arrows) sections of bundles of fibrils of the remnants of a ghost tangle between astrocytic processes (As). In this stage of ghost tangle degradation, the twisted structure of the PHFs is indistinguishable.

98

Astrocytes in Brain Aging and Neurodegeneration

Fig. 5.4. Neuronal pathology in cerebral cortex affected by severe amyloid angiopathy and obliteration of capillary vessels. There is proliferation of edematous astrocytic processes (As) around the body of a dark neuron (N) exhibiting condensed nucleoplasm and aggregated chromatin (arrow) in the deformed cell nucleus, vacuolation of endoplasmic reticulum (er), and degenerative changes in mitochondria (m).

Astrocyte Pathology in Alzheimer Disease

99

Astrocyte Degeneration in AD In brains of patients with severe Alzheimer-type pathology, some astrocytes degenerate with accumulation of abnormally phosphorylated tau, Rosenthal fibers, eosinophilic inclusions, anchorage densities, or corpora amylacea. Interindividual differences in the type, distribution, and severity of these changes suggest that they may represent responses to several different factors. They might be a nonspecific expression of chronic oversaturation of the cell environment with extracellular proteins. The nature of accumulated proteins varies, depending on the pathological conditions to which astrocytes are exposed, and as a result, different morphological forms of degeneration are observed. Abnormal tau protein phosphorylation with deposition of fibrillar tau-positive inclusions might be the effect of a cellular imbalance between the process of phosphorylation, dependent upon specific kinases, and dephosphorylation, dependent upon specific phosphatases. As a result, hyperphosphorylated tau accumulates in the neurons, and in some cases in astrocytes and/or oligodendrocytes of individuals with AD.

Tau-Positive Inclusions The abnormal fibrils, which resemble PHFs in neurons, were described also in astrocytes,91-93 where they appear as fibrils with periodical constrictions (twisted tubules) or straight tubules. The presence of tau-positive twisted and nontwisted tubules in astrocytes and oligodendrocytes94 in AD and other progressive disorders, including progressive supranuclear palsy95-98 and Pick’s disease,99 suggests that glial elements also are affected by insults similar to those that affect neurons. The presence of tau protein in astrocytes and oligodendrocytes100-101 indicates that tau can no longer be considered a neuron-specific protein. Tau-positive inclusions in neurons and glial cells have been recently recognized in many neurodegenerative diseases including Alzheimer disease, Pick disease, progressive supranuclear palsy, subacute sclerosing panencephalitis, and corticobasal degeneration.92,96-98,102-104 In corticobasal degeneration, they are even more numerous in astrocytes and oligodendroglia than in neurons.104 Taupositive inclusions share common phosphorylation characteristics irrespective of the underlying disease or cell type in which they occur.103 In comparison with neuronal tangles, glial inclusions show some morphological and immunocytochemical differences. In corticobasal degeneration, they are tau- and Gallyaspositive but Bielschowsky-negative.104 NFTs in neurons and tau-positive inclusions in astrocytes in progressive supranuclear palsy are composed mainly of straight tubules103 and have less phosphorylated tau than in AD.105

Rosenthal Fibers In three of the six examined cortical biopsies from patients with AD, Rosenthal fibers were found in the astrocyte body and cytoplasmic processes.18 In some cells, Rosenthal fibers are associated with glial filaments (Fig. 5.5), but in many astrocytes, there is no spatial relationship between intermediate filaments and Rosenthal fibers. The proportion of cortical astrocytes affected by Rosenthal fibers ranges from 5 to 40%. This form of astrocyte degeneration is associated with condensation of cell cytoplasm, deformation of the cell nucleus, and aggregation of the chromatin and nucleoplasm. This form of astrocyte degeneration is detectable not only in the area of AD pathology but also in the surrounding, morphologically unchanged neuropil. Astrocytes with Rosenthal fibers involved in ghost tangle removal are seen sporadically (Fig. 5.6). Rosenthal fibers develop in human astrocytes, in astrocytic tumors, in Alexander’s disease, and in astrocytes where reactive gliosis has been present for a long time.106-113 They were described also in astrocytes of sheep.114-116 Rosenthal fibers have a heterogenous chemical

100

Astrocytes in Brain Aging and Neurodegeneration

Fig. 5.5. Astrocyte degeneration with accumulation of Rosenthal fibers (RF) between cytoplasmic intermediate filaments (if), condensation of nucleoplasm, aggregation of chromatin (ch), and deformation of cell nucleus. Neuronal degeneration with PHF accumulation in cell processes is also depicted.

Astrocyte Pathology in Alzheimer Disease

101

Fig. 5.6. Remnants of ghost tangle (arrows) in deep cytoplasmic invagination of an astrocyte with numerous Rosenthal fibers (RF) in the cytoplasm and aggregated chromatin (ch) in the nucleus (Nu).

structure and share epitopes with αB-crystallin, GFAP, and ubiquitin.111,117-118 Because of spatial relationships to intermediate filaments, overproduction or incomplete degradation of glial filaments was considered the cause of Rosenthal fiber formation.119 In AD, in many astrocytes this contact is undetectable, which may indicate that these changes are not related to GFAP (but see chapter 4). Rosenthal fibers are ubiquitinated from the earliest steps of their formation111 but ubiquitination is considered a secondary reaction.111,118 Ubiquitin binds to abnormal proteins destined for ATP-dependent proteolysis.120-121 In experimental studies, Rosenthal fiber formation has been attributed to the intake of foreign proteins by astrocytes and their cytoplasmic pathology in degradation and disposal of these proteins.116,122

Eosinophilic Inclusions This form of degeneration manifests with the accumulation of eosinophilic material in the cytoplasm of astrocytes in AD. The inclusions contain two morphologically distinguishable components:

102

Astrocytes in Brain Aging and Neurodegeneration

Fig. 5.7. Numerous eosinophilic inclusions (ei) in the cytoplasm of an astrocyte. There is condensed nucleoplasm and aggregated chromatin (ch) in the deformed cell nucleus (Nu).

Astrocyte Pathology in Alzheimer Disease

103

Fig. 5.8. Anchorage densities (arrows) with hemidesmosome-like structures (double arrow) in end-feet of perivascular-reactive astrocyte. There is pathological reinforcement of the astrocytic interface between the vessel and neuropil. bm = basement membrane; E = endothelial cell.

1. electron-dense granular or floccular material, which is very condensed in some astrocytes (Fig. 5.7) and more loosely arranged in other cells, especially in those with cytoplasm that is watery and poor in organelles; and 2. the remnants of cytoplasmic organelles in different stages of disintegration. The presence of fragments of rough endoplasmic reticulum, clusters of ribosomes, and small vesicles embedded in electron-dense granular or floccular material suggests nonlysosomal degradation of cell components in this form of astrocyte degeneration. The distribution, size, and shape of the inclusions in astrocytes vary. They prevail in the cell body, but in cells with numerous inclusions, they are also present in cell processes. In severely affected astrocytes, the cytoplasm is occupied by numerous inclusions 6-12 µm in diameter. In other cells, tiny aggregates of granular or floccular osmophilic material prevail. This form of astrocyte degeneration was observed originally in the brain of a 5 year old child with Aicardi’s syndrome123 and in a 20 year old man with cerebral palsy, mental retardation and brain malformation.124 However, development of eosinophilic inclusions in the

104

Astrocytes in Brain Aging and Neurodegeneration

brain of aged persons with AD indicates that eosinophilic degeneration of astrocytes is not restricted to brain developmental malformations or congenital astrocytic dysfunction.

Anchorage Densities Anchorage densities associated with hemidesmosome-like structures develop in perivascular reactive astrocytes (Fig. 5.8). The fusion of hundreds of hemidesmosomes produces continuous hemidesmosomes, which are coated by thickened basal lamina. Anchorage densities appear about 200-300 nm away from hemidesmosomes. Hemidesmosomes are connected to anchorage densities by numerous fibrils. The anchorage densities, in turn, are often contiguous with intermediate cytoplasmic filaments. This complex structure appears to reinforce the cell membrane facing the perivascular space. Astrocytic anchorage densities associated with hemidesmosome-like structures were described in two cases of brain atrophy.92 The presence of the same pathological changes in two cases of AD18 and in one case of Gerstmann-Sträussler-Scheinker disease (unpublished observation in material obtained from Dr. Budka, Vienna), all affected by severe brain atrophy, indicates that astrocytic anchorage densities associated with hemidesmosome-like structures may develop in many pathological conditions associated with brain atrophy and astrogliosis.

Corpora Amylacea Corpora amylacea (CA) are cytoplasmic inclusions of human astrocytes and neurons and are often deposited in the extracellular space following disintegration of the host cell. They are associated with normal aging and neurodegenerative diseases such as Alzheimer’s disease, Lafora disease, and progressive supranuclear palsy. CA are discussed in considerable detail in chapters 4 and 10 of this volume.

Acknowledgment The authors wish to thank Dr. B. Lach and Dr. A.P. Anzil for biopsy material, Dr. K.C. Wang for assistance in preparing the material for electron microscopic studies, and Ms. M. Stoddard Marlow for copy-editing the manuscript. The study was supported by funds from the New York State Office of Mental Retardation and Developmental Disabilities and a grant from the National Institutes of Health, National Institute of Aging No. PO1-AG11531.

References 1. Wisniewski HM, Wegiel J, Wang KC et al. Ultrastructural studies of the cells forming amyloid fibers in classical plaques. Can J Neurol Sci 1989; 16:535-542. 2. Wisniewski HM, Vorbrodt AW, Wegiel J et al. Ultrastructure of the cells forming amyloid fibers in Alzheimer disease and scrapie. Am J Med Genet, Suppl 1990; 7:287-297. 3. Wegiel J, Wisniewski HM. The complex of microglial cells and amyloid star in three- dimensional reconstruction. Acta Neuropathol 1990; 81:116-124. 4. Wisniewski HM, Wegiel J, Wang K-C et al. Ultrastructural studies of the cells forming amyloid in the vessel wall in Alzheimer disease. Acta Neuropathol 1992; 84:117-127. 5. Wisniewski HM, Wegiel J. Migration of perivascular cells into the neuropil and their involvement in β-amyloid plaque formation. Acta Neuropathol 1993; 85:586-595. 6. Hickey WF, Kimura H. Perivascular microglial cells of the CNS are bone marrow-derived and present antigen in vivo. Science 1988; 239:290-292. 7. Perry VH, Gordon S. Modulation of CD4 antigen on macrophages and microglia in rat brain. J Exp Med 1987; 166:1138-1143. 8. Perry VH, Crocker RP, Gordon S. The blood-brain barrier regulates the expression of a macrophage sialic acid-binding receptor on microglia. J Cell Sci 1992; 101:201-207.

Astrocyte Pathology in Alzheimer Disease

105

9. Probst A, Langui D, Ipsen S et al. Deposition of beta/A4 protein along neuronal plasma membranes in diffuse senile plaques. Acta Neuropathol 1991; 83:21-29. 10. Pappolla MA, Omar RA, Vinters HV. Image analysis microspectroscopy shows that neurons participate in the genesis of early primitive (diffuse) senile plaques. Am J Pathol 1991; 139:599-607. 11. Wisniewski HM, Wegiel J, Kotula L. Some neuropathological aspects of Alzheimer’s disease and its relevance to other disciplines. Neuropath Appl Neurobiol 1996; 22:3-11. 12. Wisniewski HM, Wegiel J. β-amyloid formation by myocytes of leptomeningeal vessels. Acta Neuropathol 1994; 87:233-241. 13. Wegiel J, Wisniewski HM, Dziewiatkowski J et al. The origin of amyloid in cerebral vessels of aged dogs. Brain Res 1995; 705:225-234. 14. Wisniewski HM, Wegiel J. Spatial relationships between astrocytes and classical plaque components. Neurobiol Aging 1991; 12:593-600. 15. Bobinski M, Wegiel J, Wisniewski HM et al. Neurofibrillary pathology—correlation with hippocampal formation atrophy in Alzheimer disease. Neurobiol Aging 1996; 17:909-919. 16. Gomez-Isla T, Hollister R, West H et al. Neuronal loss correlates with but exceeds neurofibrillary tangles in Alzheimer’s disease. Ann Neurol 1997; 41:17-24. 17. Schechter R, Yen SHC, Terry RD. Fibrous astrocytes in senile dementia of Alzheimer type. J Neuropath Exp Neurol 1981; 40:95-101. 18. Wegiel J, Wisniewski HM. Rosenthal fibers, eosinophilic inclusions, and anchorage densities with desmosome-like structures in astrocytes in Alzheimer disease. Acta Neuropathol 1993; 87:355-361. 19. Pitas RE, Boyle JK, Lee et al. Astrocytes synthesize apolipoprotein E and metabolize apolipoprotein E-containing lipoproteins. Biochem Biophys Acta 1987; 917:148-161. 20. Poirier J, Hess M, May PC et al. Astrocytic apolipoprotein E mRNA and GFAP mRNA in hippocampus after entorhinal cortex lesioning. Mol Brain Res 1991; 11:97-106. 21. Fontana A, Kristensen F, Dubs R et al. Production of prostaglandin E and an interleukin1 like factor by cultured astrocytes and C6 glioma cells. J Immunol 1982; 129:2413-2419. 22. Bakhit C, Armanini M, Bennett GL et al. Increase in glia-derived nerve growth factor following destruction of hippocampal neurons. Brain Res 1991; 560:76-83. 23. Zhang WW, Badonic T, Höög A et al. Astrocytes in Alzheimer’s disease express immunoreactivity to the vaso-constrictor endothelin-1. J Neurol Sci 1994; 122:90-96. 24. Akiyama H, Kawamata T, Yamada T et al. Expression of intercellular adhesion molecule (ICAM)-1 by a subset of astrocytes in Alzheimer disease and some other degenerative neurological disorders. Acta Neuropathol 1993; 85:628-634. 25. Giulian D, Baker TJ. Characterization of ameboid microglia isolated from developing mammalian brain. J Neurosci 1986; 6:2163-2178. 26. Giulian D, Baker TJ, Shih LN et al. Interleukin-1 of the central nervous system is produced by ameboid microglia. J Exp Biol Med 1986; 164:594-604. 27. Giulian D, Chen J, Ingeman JE et al. The role of mononuclear phagocytes in wound healing after traumatic injury to adult mammalian brain. J Neurosci 1989; 9:4416-4429. 28. Giulian D, Lachman LB. Interleukin-1 stimulation of astroglial proliferation after brain injury. Science 1985; 228:497-499. 29. Giulian D, Woodward J, Young DG et al. Interleukin-1 injected into mammalian brain stimulates astrogliosis and neovascularization. J Neurosci 1988; 8:2485-2490. 30. Boje KM, Arora PK. Microglial-produced nitric oxide and reactive nitrogen oxides mediate neuronal cell death. Brain Res 1992; 587:250-256. 31. Chao CC, Hu S, Molitor TW et al. Activated microglia mediate neuronal cell injury via a nitric oxide mechanism. J Immunol 1992; 149:2736-2741. 32. Gray CW, Patel AJ. Regulation of beta-amyloid precursor protein isoform mRNAs by transforming growth factor-beta 1 and interleukin-1 beta in astrocytes. Brain Res Mol Brain Res 1993; 19:251-256. 33. Sheng JG, Ito K, Skinner RD et al. In vivo and in vitro evidence supporting a role for the inflammatory cytokine interleukin-1 as a driving force in Alzheimer pathogenesis. Neurobiol Aging 1996; 17:761-766.

106

Astrocytes in Brain Aging and Neurodegeneration

34. Picken MM, Larrondo-Lillo M, Coria F et al. Distribution of the protease inhibitor α1antichymotrypsin in cerebral and systemic amyloid. J Neuropathol Exp Neurol 1990; 49:41-48. 35. Potter H, Abraham CR, Dressler D. The Alzheimer amyloid components α1-antichymotrypsin and β-protein form a stable complex in vitro. In: Iqbal K, McLachlan DRC, Winblad B et al, eds. Alzheimer’s Disease. Chichester: John Wiley and Sons Ltd, 1991:275-279. 36. Das S, Potter H. Expression of the Alzheimer amyloid-promoting factor antichymotrypsin is induced in human astrocytes by IL-1. Neuron 1995; 14:447-456. 37. Snipes GJ, McGuire CB, Norden JJ et al. Nerve injury stimulates the secretion of apolipoprotein E by nonneuronal cells. Proc Natl Acad Sci U S A 1986; 83:1130-1134. 38. Boyles JK, Zoellner CD, Anderson LJ et al. A role for Apolipoprotein E, Apolipoprotein A1, low density lipoprotein receptors in cholesterol transport during regeneration and remyelination of the rat sciatic nerve. J Clin Invest 1989; 83:1015-1031. 39. Wisniewski T, Golabek A, Matsubara E et al. Apolipoprotein E: Binding to soluble Alzheimer’s β-amyloid. Biochem Biophys Res Commun 1993; 192:359-365. 40. Schmechel DE, Saunders AM, Strittmatter WJ et al. Increased amyloid β-deposition as a consequence of apolipoprotein E genotype in late-onset Alzheimer’s disease. Proc Natl Acad Sci U S A 1993; 90:9649-9653. 41. Strittmatter WJ, Weisgraber KH, Goedert M et al. Hypothesis: Microtubule instability and paired helical filament formation in the Alzheimer disease brain are related to apolipoprotein E genotype. Exp Neurol 1994; 125:163-171. 42. Strittmatter WJ, Weisgraber KH, Huang et al. Binding of human apolipoprotein E to βA4 peptide: Isoform-specific effects and implications for late-onset Alzheimer disease. Proc Natl Acad Sci U S A 1993; 90:8098-8102. 43. Yang AJ, Knauer M, Burdick DA et al. Intracellular Aβ1-42 aggregates stimulate the accumulation of stable, insoluble amyloidogenic fragments of the amyloid precursor protein in transfected cells. J Biol Chem 1995; 270:14786-14792. 44. Davis-Salinas J, Saporito-Irvin SM, Cotman CW et al. Amyloid β-protein induces its own production in cultured degenerating cerebrovascular smooth muscle cells. J Neurochem 1995; 65:931-934. 45. Yan SD, Chen X, Fu J et al. RAGE and amyloid-β peptide neurotoxicity in Alzheimer’s disease. Nature 1996; 382:685-691. 46. Khoury EJ, Hickman SE, Thomas CA et al. Scavenger receptor-mediated adhesion of microglia to β-amyloid fibrils. Nature 1996; 382:716-719. 47. Araujo DM, Cotman CW. Beta-amyloid stimulates glial cells in vitro to produce growth factors that accumulate in senile plaques in Alzheimer’s disease. Brain Res 1992; 569:141-145. 48. Del Bo R, Angeretti N, Lucca E. Reciprocal control of inflammatory cytokines, IL-1 and IL-6, and beta-amyloid production in cultures. Neurosci Lett 1995; 188:70-74. 49. Jiang MH, Höög A, Ma K et al. Endothelin-1-like immune-reactivity is expressed in reactive astrocytes of the human brain. Neuroreport 1993; 4:935-937. 50. Miller RC, Pelton JT, Huggins JP. Endothelins—from receptors to medicine. Trends Pharmacol Sci 1993; 14:54-60. 51. MacCumber M, Ross CA, Snyder H. Endothelin in brain: Receptors, mitogenesis, and biosynthesis in glial cells. Proc Natl Acad Sci U S A 1990; 87:2359-2363. 52. Ehrenreich H, Anderson RW, Fox CH et al. Endothelins, peptides with potent vasoactive properties are produced by human macrophages. J Exp Med 1990; 172:1741-1748. 53. Ehrenreich H, Kehrl JH, Anderson RW et al. A vasoactive peptide, endothelin-3, is produced by and specifically binds to primary astrocytes. Brain Res 1991; 538:54-58. 54. Supattapone S, Simpson AWM, Ashley CC. Free calcium rise and mitogenesis in glial cells caused by endothelin. Biochem Biophys Res Commun 1989; 165:1115-1122. 55. Hösli E, Hösli L. Autoradiographic evidence for endothelin receptors on astrocytes in cultures of rat cerebellum, brainstem and spinal cord. Neurosci Lett 1991; 129:55-58. 56. Couraud P-O, Durieu-Trautmann O, Mahe E et al. Comparison of binding characteristics of endothelin receptors on subpopulations of astrocytes. Life Sci 1991; 49:1471-1476.

Astrocyte Pathology in Alzheimer Disease

107

57. Giaid A, Gibson S, Ibrahim N et al. Endothelin 1 and endothelium-derived peptide is expressed in neurons of the human spinal cord and dorsal root ganglia. Proc Natl Acad Sci U S A 1989; 86:7634-7638. 58. Giaid A, Gibson SJ, Herrero MT et al. Topographical localisation of endothelin mRNA and peptide immunoreactivity in neurones of the human brain. Histochemistry 1991; 95:303-314. 59. Yamaji T, Johshita H, Ishibashi M et al. Endothelin family in human plasma and cerebrospinal fluid. J Clin Endocrinol Metab 1990; 71:1611-1615. 60. Shigeno T, Mima T. A new vasoconstrictor peptide, endothelin: Profiles as vasoconstrictor and neuropeptide. Cerebrovasc Brain Metab Rev 1990; 2:227-239. 61. Yanagisawa M, Kurihara H, Kimura S et al. A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature 1988; 332:411-415. 62. Ide K, Yamakawa K, Nakagomi T et al. The role of endothelin in the pathogenesis of vasospasm following subarachnoid hemorrhage. Neurol Res 1989; 11:101-104. 63. Kauser K, Rubanyi G, Harder D. Endothelium dependent modulation of endothelin-induced vasoconstriction and membrane depolarization in cat cerebral arteries. J Pharmacol Exp Ther 1989; 252:93-97. 64. Lee M-E, Monte SMDL, Ng S-C et al. Expression of the potent vasoconstrictor endothelin in the human central nervous system. J Clin Invest 1990; 86:141-147. 65. O’Brien JT, Eagger S, Syed GMS et al. A study of regional cerebral blood flow and cognitive performance in Alzheimer’s disease. J Neurol Neurosurg Psychiatry 1992; 55:1182-1187. 66. Bonte FJ, Tintner R, Weiner MF et al. Brain blood flow in the dementias: SPECT with histopathologic correlation. Radiology 1993; 186:361-365. 67. Mandybur TI, Ormsby I, Zemlan FP. Cerebral aging: A quantitative study of gliosis in old nude mice. Acta Neuropathol 1989; 77:507-513. 68. Delacourte A. General and dramatic glial reaction in Alzheimer brains. Neurology 1990; 40:33-37. 69. O’Kusky J, Colonier M. Postnatal changes in the number of astrocytes, oligodendrocytes, and microglia in the visual cortex (area 17) of the macaque monkey: A stereological analysis in normal and monocularly deprived animals. J Comp Neurol 1982; 210:307-315. 70. Adams I, Jones DG. Synaptic remodelling and astrocytic hypertrophy in rat cerebral cortex from early to late adulthood. Neurobiol Aging 1983; 3:179-186. 71. De la Roza C, Cano J, Reinoso-Suarez F. An electron microscopic study of astroglia and oligodendroglia in the lateral geniculate nucleus of aged rats. Mech Ageing Dev 1985; 29:267-281. 72. Geinisman Y, Bondareff W, Dodge JT. Hypertrophy of astroglial processes in the dentate gyrus of the senescent rat. Am J Anat 1978; 153:537-544. 73. Landfield PW, Rose G, Sandles L et al. Patterns of astroglial hypertrophy and neuronal degeneration in the hippocampus of aged, memory-deficient rats. J Gerontol 1977; 32:3-12. 74. Lindsay JD, Landfield PW, Lynch G. Early onset and topographical distribution of hypertrophied astrocytes in hippocampus of aging rats: A quantitative study. J Gerontol 1979; 34:661-671. 75. Lamar CH, Hinsman EJ, Henrickson CK. Alterations in the hippocampus of aged mice. Acta Neuropathol 1976; 36:387-391. 76. von Braunmuhl A. Alterserkrankungen des Zentralnervensystems. Senile Involution. Senile Demenz. Alzheimerische Krankheit. In: Lubarsch O, Henke F, Rossle R, eds. Handbuch der Speziellen Pathologischen Anatomie und Histologie XIII/1. Berlin: Springer Verlag, 1957; 13:337-539. 77. Senitz D, Goertchen R. Über Astrozytenveränderungen in der orbitofrontalen Hirnrinde bei seniler Demenz. Zentral Allg Pathol 1978; 122:515-521. 78. Hansen LA, Armstrong DM, Terry RD. An immunohistochemical quantification of fibrous astrocytes in the aging human cerebral cortex. Neurobiol Aging 1987; 8:1-6. 79. Dickson DW, Farlo J, Davies P. Alzheimer’s disease. A double-labeling immunohistochemical study of senile plaques. Am J Pathol 1988; 132:86-101.

108

Astrocytes in Brain Aging and Neurodegeneration

80. Duffy PE, Rapoport M, Graf L. Glial fibrillary acidic protein and Alzheimer-type senile dementia. Neurology 1980; 30:778-780. 81. Fischer O. Die presbyophrene Demenz, deren anatomische Grundlage und klinische Abgrenzung. Z Gesamte Neurol Psychiat 1910; 3:371-471. 82. Mancardi GL, Liwnicz BH, Mandybur TI. Fibrous astrocytes in Alzheimer’s disease and senile dementia of Alzheimer’s type. An immunohistochemical and ultrastructural study. Acta Neuropathol 1983; 61:76-80. 83. Mandybur TI. Cerebral amyloid angiopathy and astrocytic gliosis in Alzheimer’s disease. Acta Neuropathol 1989; 78:329-331. 84. Probst A, Ulrich J, Heitz PU. Senile dementia of Alzheimer type: Astroglial reaction to extracellular neurofibrillary tangles in the hippocampus. Acta Neuropathol 1982; 57:75-79. 85. Wisniewski HM, Sinatra RS, Iqbal K et al. Neurofibrillary and synaptic pathology in the aged brain. In: Johnson JE, ed. Aging and Cell Structure. New York: Plenum Publishing, 1981:105-142. 86. Itagaki S, McGeer PL, Akiyama H et al. Relationship of microglia and astrocytes to amyloid deposits of Alzheimer disease. J Neuroimmunol 1989; 24:173-182. 87. Rozemuller JM, Eikelenboom P, Stam FC et al. A4 protein in Alzheimer’s disease: Primary and secondary cellular events in extracellular amyloid deposition. J Neuropathol Exp Neurol 1989; 48:674-691. 88. Ebersolt C, Perez M, Bockaernt J. Neuronal, glial, and monoglial localization of neurotransmitter: Sensitive adenylate cyclases in cerebral cortex of mice. Brain Res 1981; 213: 139-150. 89. Whitaker-Azmitia PM, Azmitia EC. Autoregulation of fetal serotonergic neuronal development: role of high affinity serotonin receptors. Neurosci Lett 1986; 67:307-312. 90. Yamaguchi H, Morimatsu M, Hirai S et al. Alzheimer’s neurofibrillary tangles are penetrated by astroglial processes and appear eosinophilic in their final stages. Acta Neuropathol 1987; 72:214-217. 91. Ikeda K, Haga C, Akiyama H et al. Coexistence of paired helical filaments and glial filaments in astrocytic processes within ghost tangles. Neurosci Lett 1992; 148:126-128. 92. Nakano I, Iwatsubo T, Otsuka N et al. Paired helical filaments in astrocytes: electron microscopy and immunocytochemistry in a case of atypical Alzheimer’s disease. Acta Neuropathol 1992; 83:228-232. 93. Yamazaki M, Nakano I, Imazu O et al. Paired helical filaments and straight tubules in astrocytes: An electron microscopic study in dementia of the Alzheimer type. Acta Neuropathol 1995; 90:31-36. 94. Nishimura M, Tomimoto H, Suenaga T et al. Immunocytochemical characterization of glial fibrillary tangles in Alzheimer’s disease brain. Am J Pathol 1995; 146:1052-1058. 95. Nishimura M, Namba Y, Ikeda K et al. Glial fibrillary tangles with straight tubules in the brains of patients with progressive supranuclear palsy. Neurosci Lett 1992; 143:35-38. 96. Yamada T, McGeer PL. Oligodendroglial microtubular masses: An abnormality observed in some human neurodegenerative diseases. Neurosci Lett 1990; 120:163-166. 97. Yamada T, McGeer PL, McGeer EG. Appearance of paired nucleated, tau-positive glia in patients with progressive supranuclear palsy brain tissue. Neurosci Lett 1992; 135:99-102. 98. Yamada T, Calne DB, Akiyama H et al. Further observations on tau-positive glia in the brains with progressive supranuclear palsy. Acta Neuropathol 1993; 85:308-315. 99. Yamazaki M, Nakano I, Imazu O et al. Astrocytic straight tubules in the brain of patient with Pick’s disease. Acta Neuropathol 1994; 88:587-591. 100. Migheli A, Butler M, Brown M et al. Light and electron microscope localization of the microtubule-associated tau protein in rat brain. J Neurosci 1988; 8:1846-1851. 101. Papasozomenos SC, Binder LI. Phosphorylation determines two distinct species of tau in the central nervous system. Cell Motil Cytoskeleton 1987; 8:210-226. 102. Wakabayashi K, Oyanagi K, Makifuchi T et al. Corticobasal degeneration: Etiopathological significance of the cytoskeletal alterations. Acta Neuropathol 1994; 87:545-553. 103. Iwatsubo T, Hasegawa M, Ihara Y. Neuronal and glial tau-positive inclusions in diverse neurologic diseases share common phosphorylation characteristics. Acta Neuropathol 1994; 88:129-136.

Astrocyte Pathology in Alzheimer Disease

109

104. Horoupian DS, Chu PL. Unusual case of corticobasal degeneration with tau /Gallyas-positive neuronal and glial cells. Acta Neuropathol 1994; 88:592-598. 105. Flament S, Delacourte A, Verny M et al. Abnormal tau proteins in progressive supranuclear palsy. Similarities and differences with the neurofibrillary degeneration of the Alzheimer type. Acta Neuropathol 1991; 81:591-596. 106. Alexander WS. Progressive fibrinoid degeneration of fibrillary astrocytes associated with mental retardation in a hydrocephalic infant. Brain 1949; 71:373-381. 107. Borrett D, Becker LE. Alexander’s disease: A disease of astrocytes. Brain 1985; 108:367-385. 108. Crome L. Megaloencephaly associated with hyaline pan-neuropathy. Brain 1953; 76:215-228. 109. Grcevic N, Yates PO. Rosenthal fibers in tumors of the central nervous system. J Pathol Bacteriol 1957; 73:467-472. 110. Herndon RM, Rubenstein LJ, Freeman JM et al. Light and electron microscopic observations on Rosenthal fibers in Alexander’s disease and in multiple sclerosis. J Neuropath Exp Neurol 1970; 29:524-551. 111. Lach B, Sikorska M, Rippstein P et al. Immunoelectron microscopy of Rosenthal fibers. Acta Neuropathol 1991; 81:503-509. 112. Ogasawara N. Multiple Sclerose mit Rosenthalschen Fasern. Acta Neuropathol 1956; 5:61-68. 113. Russell DS, Rubinstein LJ. In: Pathology of Tumors of the CNS. Baltimore: Williams and Wilkins 1989; 977:155-167. 114. Frankhauser R, Fatzer R, Bestetti JP et al. Encephalopathy with Rosenthal fiber formation in a sheep. Acta Neuropathol 1980; 50:57-60. 115. McGrath JT. Spontaneous animal models of human disease. In: Andrews EJ, Ward C, Altman NH, eds. Spontaneous Animal Models of Human Disease. New York: Academic Press, 1979; 2:147-148. 116. Horoupian DS, Kress Y, Yen SH et al. Nickel induced changes and reappraisal of Rosenthal fibers in focal CNS lesions. J Neuropathol Exp Neurol 1982; 41:664-675. 117. Iwaki T, Kume-Iwaki A, Corbin E et al. Expression of the B-chain of α-crystallin in CNS glia. J Neuropathol Exp Neurol 1990; 49:344. 118. Tomokane N, Iwaki T, Tateishi J et al. Rosenthal fibers share epitopes with αB-crystallin, glial fibrillary acidic protein, and ubiquitin, but not with vimentin. Immunoelectron microscopy with colloidal gold. Am J Pathol 1991; 138:875-885. 119. Goldman JE, Corbin E. Isolation of major protein component of Rosenthal fibers. Am J Pathol 1988; 130:569-578. 120. Hershko A, Ciechanover A, Heller H et al. Proposed role of ATP in protein breakdown: Conjugation of proteins with multiple chains of the early peptide of ATP-dependent proteolysis. Proc Natl Acad Sci U S A 1980; 77:1783-1786. 121. Rechsteiner M. Ubiquitin-mediated pathways for intracellular proteolysis. Ann Rev Cell Biol 1987; 3:1-30. 122. Kress Y, Gaskin F, Horoupian DS et al. Nickel induction of Rosenthal fibers in rat brain. Brain Res 1981; 210:419-425. 123. Abe H, Yagashita S, Itoh K et al. Novel eosinophilic inclusion in astrocytes. Acta Neuropathol 1992; 83:659-663. 124. Minagawa M, Shioda K, Shimizu Y et al. Inclusion bodies in cerebral cortical astrocytes: A new change of astrocytes. Acta Neuropathol 1992; 84:113-116.

CHAPTER 6

Parkinson’s Disease Donato A. Di Monte

Introduction

A

strocytes have often been seen as mere “supportive” elements of the brain architecture and have generally received little consideration in discussions on the pathogenesis of neurodegeneration. This view is slowly but inevitably being reconsidered. The more we know about biochemical, pharmacologic and pathological interactions between neuronal cells and astrocytes, the more the concept of a “hierarchic supremacy” of neurons versus astrocytes becomes outdated. Neuron-astrocyte interaction is an essential component of brain function and, as such, would also be expected to play a role in pathological processes. Indeed, evidence pointing to astrocytes as “active” participants in the toxic events leading to neurodegeneration is quite convincing, starting with the simple observation of gliosis as a common astrocyte response to injury. In the case of Parkinson’s disease, astrocytes form glial scars in the areas of neurodegeneration.1 It would be a mistake, however, to view formation of these scars as the only neuron-astrocyte interaction, secondary to nerve cell loss. Most likely, glial scars are the ultimate manifestation of an ongoing interaction which has been active during the entire course of the neurodegenerative process in the Parkinsonian brain. Both the pathogenesis of nigrostriatal degeneration and the role of astrocytes in Parkinson’s disease are still poorly understood. Hypothetical scenarios can be visualized based on the current knowledge of mechanisms of nigrostriatal injury. For example, important clues can be derived from studies of models of toxicant-induced parkinsonism. In particular, the mechanisms of neuronal loss caused by the parkinsonism-inducing agent 1methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) have confirmed an important role of astrocytes and directed our research toward specific aspects of neuron-astrocyte interactions that are most likely involved in dopaminergic degeneration. In this paper, I will first present a general overview on idiopathic Parkinson’s disease and the MPTP model of parkinsonism. Then, in the second part of the review, specific examples of known or suggested neuron-astrocyte interactions will be discussed in relation to nigrostriatal dopaminergic degeneration.

Idiopathic Parkinson’s Disease Idiopathic Parkinson’s disease is one of the most common neurodegenerative disorders of aging. The age-specific prevalence is estimated to be 41/100,000 in the population at age 50, but it increases dramatically to 1,518/100,000 by age 80.2 From the clinical point of view, this age-related neurodegenerative disorder is characterized by the classic symptomatic triad of tremor, rigidity and bradykinesia as well as a number of other symptoms and Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

112

Astrocytes in Brain Aging and Neurodegeneration

signs such as postural instability and loss of mimicry. The primary pathological feature of Parkinson’s disease is the loss of pigmented neurons in the substantia nigra pars compacta.3 The degenerative process, however, also involves neurons in other anatomical regions and, in particular, in the locus ceruleus and nucleus basalis of Meynert. Reactive microglia and glial scars, often made up of astrocytes with delicate glial fibrils, are usually observed in the areas most severely affected by neurodegeneration.1 Since nigral neurons project their axons to the striatum and use dopamine as their neurotransmitter, depletion of dopamine in the nigrostriatal pathway represents the most relevant neurochemical alteration in the Parkinsonian brain.4 It is also the basis for the primary treatment of the disease with L-dopa which, after crossing the blood-brain barrier, is decarboxylated to and replenishes the lacking dopamine. As already mentioned, the toxic mechanisms underlying nigrostriatal degeneration remain to be identified. However, at least two hypotheses concerning dopaminergic cell injury are often debated and will be discussed here because of their implications for possible astrocyte involvement (see section below on Neuron-Astrocyte Interactions in Nigrostriatal Degeneration). It has been suggested that dopamine, in addition to being the neurotransmitter used by nigrostriatal neurons, may also act as an endogenous toxin. As a corollary of this hypothesis, oxygen radical-induced oxidative stress could be responsible for or contribute to neurodegeneration.5 The scheme in Figure 6.1 summarizes the relationship between dopamine, oxidative stress and neuronal injury. The enzymatic catabolism of dopamine is catalyzed by monoamine oxidase (MAO) and generates the 3,4-dihydroxyphenyl-acetaldehyde metabolite, as well as hydrogen peroxide (H2O2). H2O2 is an oxidizing agent which could damage cells both directly and after its further reduction to the hydroxyl radical, a reaction catalyzed by transition metals (Fenton reaction). Nonenzymatic conversion of dopamine could also lead to oxidative stress via the formation of 6-hydroxydopamine and/ or toxic quinone metabolites.6 Several lines of evidence support a role of dopamine-induced oxidative stress in nigrostriatal degeneration. This evidence includes a finding of increased oxidized glutathione (GSSG) levels as a consequence of higher dopamine turnover in animals treated with reserpine or haloperidol.7-8 GSSG production is likely to reflect oxidative stress since GSSG is formed from the reaction between reduced glutathione (GSH) and H2O2, catalyzed by glutathione peroxidase. Additional evidence derives from studies in humans indicating, for example, an increased total iron content in nigral tissue of patients dying of Parkinson’s disease.9-10 As a transition metal, iron could be a catalyst in the Fenton reaction, generating the highly toxic hydroxyl radical (OH•) from H2O2. The other hypothesis that is often put forward to explain neuronal loss in Parkinson’s disease is that abnormalities in energy metabolism and, in particular, in mitochondrial oxidative phosphorylation may make the nigrostriatal tissue vulnerable to neurodegeneration. Findings in vitro support this hypothesis, showing that dopaminergic neurons are relatively more susceptible than other neuronal cell types to the toxic effects of energy deprivation.11 Convincing evidence also derives from studies in humans revealing a decrease in mitochondrial complex I activity in the brain of patients with idiopathic Parkinsonism.12-13 This decrease appears to be selective for the substantia nigra, as it was not found in the caudate nucleus, medial and lateral globus pallidus, cerebral cortex or cerebellum.14 More recently, Swerdlow and colleagues performed an experiment in which a clonal line of human neuroblastoma cells containing no mitochondrial DNA was repopulated with mitochondria from control subjects or patients with Parkinson’s disease.15 Cell lines from patients showed decreased complex I activity, increased oxygen radical production and a greater susceptibility to MPP+-induced cell death as compared to control cell lines. These results seem to support the hypothesis that abnormalities in mitochondrial DNA underlie the complex I impairment in Parkinson’s disease.

Parkinson’s Disease

113

Fig. 6.1. Mechanisms of dopaminergic degeneration inferred from the MPTP and dopamine models of neurotoxicity.

It is important to emphasize that the oxidative stress and mitochondrial hypotheses of nigrostriatal degeneration should not be seen as antithetical. In fact, it is most likely that a link exists between impairment of mitochondrial activity and oxidative stress since mitochondrial damage may ultimately cause an increased generation of oxygen radicals, and, vice versa, mitochondria are critical targets for oxidative stress.16 Thus, both toxic mechanisms may contribute to neuronal damage, perhaps in different phases in the course of Parkinson’s disease and/or to various extents in different patient populations.

MPTP-Induced Parkinsonism The description 15 years ago of the clinical syndrome that abruptly developed in young drug addicts who injected themselves with MPTP-contaminated illicit drugs represents an historical hallmark for scientists working in the field of neurodegenerative disorders.17 MPTP poisoning mimicked as closely as anyone could have anticipated the clinical features of Parkinson’s disease18 and has since become a valuable model for in vitro and animal studies on nigrostriatal degeneration. The intense basic science work that followed the MPTP discovery has confirmed a striking number of similarities between MPTP-induced parkinsonism and the idiopathic disease. In particular, neurochemical measurements in the monkey and rodent brain have demonstrated dramatic MPTP-induced depletion of nigrostriatal dopamine.19-20 From a pathological point of view, examination of the monkey brain after MPTP exposure has revealed a rather selective action of the neurotoxicant toward the nigrostriatal system.19,21 Furthermore, similar to observations in the Parkinsonian brain, dopaminergic cell loss in MPTP-treated monkeys is accompanied by definite glial scars.1 The mechanisms of MPTP neurotoxicity have also been considerably, though not completely, clarified by studies in the last 15 years. A seemingly concerted sequence of metabolic, biochemical and toxic events appears to be triggered by MPTP exposure. A brief

114

Astrocytes in Brain Aging and Neurodegeneration

review of these events represents a necessary background for our discussion on neuronalastrocyte interactions in nigrostriatal degeneration. Shortly after its discovery, MPTP was found to be metabolized by MAO,22 an enzyme localized to the outer membrane of mitochondria. Of the two forms of this enzyme, MAO type B appears to be much more efficient than type A in catalyzing MPTP conversion.23 This conversion occurs via a two-step process: First, MPTP is oxidized to the 1-methyl-4-phenyl-2,3-dihydropyridinium (MPDP+) intermediate, and then MPDP+ is further oxidized to the final product, the 1-methyl-4phenylpyridinium (MPP+) metabolite. MAO B is responsible for the conversion of MPTP to MPDP+, while the rapid oxidation of MPDP+ to MPP+ is unlikely to be mediated by either MAO or other enzymatic activities.24 The biotransformation of MPTP to MPP+ is an essential step leading to neurotoxicity, since MPP+ is thought to be the ultimate mediator of the biologic effects of MPTP. Indeed, MPTP toxicity can be prevented by MAO B inhibitors25-26 and MPP+ itself causes toxic effects similar to those seen after MPTP administration.27-28 The next critical step following MPTP bioactivation and leading to neurotoxicity is the accumulation of MPP+ by nigrostriatal dopaminergic neurons. This is an active process occurring via the catecholamine uptake system and is thought to play an important role in the selective action of MPTP.29 Dopaminergic neurons are significantly more vulnerable to MPTP toxicity because they are exposed to higher levels of MPP+ for a prolonged period of time. The mechanism of MPP+-induced cell death has also been extensively studied. In 1985, a report by Nicklas and colleagues,30 showing that MPP+ was an inhibitor of mitochondrial complex I, provided the first experimental evidence linking MPP+ cytotoxicity to an impairment of mitochondrial energy metabolism. A number of studies have since supported the now widely accepted view that, upon reaching the intracellular space, MPP+: 1. accumulates in mitochondria;31 2. blocks mitochondrial respiratory chain activity at the level of complex I;30 and 3. causes a dramatic depletion of the cellular energy substrate, ATP.24,32-34 Oxygen radicals could also be generated as a consequence of mitochondrial electron flow inhibition.35 Although the occurrence of these sequential events remains to be directly demonstrated in nigrostriatal dopaminergic neurons, MPP+-induced ATP depletion has been linked to cytotoxicity in a variety of in vitro models24,32-33 and an ATP decrease has been observed in the striatum and ventral mesencephalon of mice injected with MPTP.34 The impairment of energy supplies caused by MPP+ has prompted scientists to suggest that excitatory amino acids (EAAs) may contribute to its neurotoxicity. ATP depletion and excitotoxicity could be linked by one or more of the following biochemical events: 1. the reduction of ATP levels could affect the normal uptake and inactivation of EAAs, resulting in their accumulation in the synaptic cleft;36 2. the membrane depolarization resulting from ATP depletion could relieve the voltage-dependent Mg2+ block of NMDA channels.37 This would facilitate the activation of EAA receptors and thus increase the sensitivity of neuronal cells to EAAmediated damage; 3. overstimulation of the EAA receptors could itself cause membrane depolarization leading to a further increase in energy consumption;38-39 4. one of the consequences of NMDA receptor activation is an influx of calcium and thus a rise in cytosolic calcium. In order to counteract this potentially cytotoxic condition, calcium is: a. actively taken up by both mitochondria and the endoplasmic reticulum,40 and b. pumped out of the cells via the plasma membrane calcium translocase.41 Mitochondrial uptake and membrane translocation of calcium are energy-dependent processes and would significantly contribute to ATP consumption.

Parkinson’s Disease

115

Thus, impaired mitochondrial activity and excitotoxicity might trigger a toxic cycle with failing energy metabolism, leading to EAA receptor activation which, in turn, causes further depletion of energy supplies. That excitotoxicity plays a role in MPTP-induced nigrostriatal injury remains to be convincingly demonstrated. While some studies have shown protection against MPTP neurotoxicity by antagonists of the N-methyl-D-aspartate (NMDA) excitotoxic receptor,42-43 other investigators have failed to achieve such protective effects.44-45 The energy impairment and ATP loss caused by MPP+ could lead to an increase in cytosolic Ca2+ either directly or indirectly via NMDA receptor activation (Fig. 6.1). The cytotoxic consequences of this Ca2+ perturbation have been related to the stimulation of enzyme activities and, in particular, attention has recently been focused on the activity of Ca2+/calmodulin-dependent nitric oxide synthase (NOS) in the CNS. Stimulation of NOS would generate excessive amounts of nitric oxide (NO), a neurotransmitter with potential cytotoxic properties.46 The possibility that NOS stimulation may play a role in MPTP-induced nigrostriatal injury is suggested by studies in which the NOS inhibitor, 7-nitroindazole (7-NI), prevented neurotoxicity.47-49 This inhibitor has been found to counteract dopamine depletion and neuronal cell loss in the striatum and substantia nigra of mice47-48 and to protect against the behavioral, neurochemical and pathological effects of MPTP in the primate model.49

Neuronal-Astrocyte Interactions in Nigrostriatal Degeneration The scheme in Figure 6.1 summarizes the pathways leading to nigrostriatal degeneration as suggested by evidence from studies on idiopathic and MPTP-induced Parkinsonism reviewed in the previous paragraphs. Next, the involvement of astrocytes in these toxic pathways will be evaluated. Experimental findings will be discussed together with more hypothetical mechanisms in order to provide a general view of current knowledge as well as possible directions for future research.

Astrocytes, MAO and Oxidative Stress Reactions catalyzed by MAO seem to be involved in different toxic pathways leading to nigrostriatal damage. As shown in Figure 6.1, MAO has been implicated in the oxidative stress hypothesis of neurodegeneration because of the formation of H2O2, a product of dopamine catabolism. This involvement of MAO in dopamine metabolism suggests that astrocytes play an important role in neurodegeneration in light of the fact that these cells may comprise a major compartment for the extraneuronal metabolism of dopamine. A recent study in monkeys has revealed that, after administration of the dopamine precursor L-DOPA, a significant proportion (approximately 50%) of dopamine deamination is mediated through MAO B.50 This reaction does not occur within dopaminergic neurons, which express only MAO A,51 but could involve astrocytes because: 1. these cells contain both MAO A and MAO B;52-53 and 2. they possess uptake sites for dopamine on their plasma membranes.54 Thus, MAO-mediated catabolism of dopamine within astrocytes could represent a significant source of H2O2. Hydrogen peroxide readily traverses cell membranes55 and may promote neuronal degeneration. Oxidative stress as a mechanism of cell injury only occurs when the production of oxidizing species (e.g., H2O2) overwhelms anti-oxidant defense mechanisms. In the case of H2O2, a primary anti-oxidant effect is achieved by the activity of glutathione peroxidase. This enzyme reduces H2O2 to H2O at the expense of glutathione, which is converted from its reduced state (GSH) to its oxidized form (GSSG). Interestingly, astrocytes are known to contain relatively higher levels of GSH than neuronal cells,56-57 raising the possibility that H2O2 generated either intra- or extraneuronally may be scavenged within astrocytes.

116

Astrocytes in Brain Aging and Neurodegeneration

Decreased levels of glutathione have been reported in the substantia nigra of patients with Parkinson’s disease as compared to control subjects.9,58 This decrease may reflect specific neuron-astrocyte interactions, with intraglial glutathione playing an important role against H2O2 accumulation and oxidative stress. An imbalance between pro-oxidant generating reactions and anti-oxidant defenses may ultimately contribute to neurodegeneration in the Parkinsonian brain.

Astrocytes, MAO and MPTP Another pathway by which MAO-catalyzed reactions could lead to nigrostriatal damage is through the bioactivation of MPTP-like endogenous and/or exogenous toxins (Fig. 6.2). The MPTP model of parkinsonism also provides a clear example of a relationship between MAO, astrocytes and dopaminergic cell injury. As already mentioned, immunocytochemical studies have shown that brain MAO B is localized to serotonergic neurons and astrocytes, but not to dopamine-containing neuronal groups.51-52 Therefore, although dopaminergic neurons of the nigrostriatal system are the main targets for MPTP neurotoxicity, they may be incapable of mediating the bioactivation of MPTP to its toxic metabolite, MPP+.59 The possibility that serotonergic neurons may play a significant role in the production of MPP+ in the brain in vivo has been ruled out by studies indicating that lesions of these neurons do not attenuate MPTP neurotoxicity.60 Thus, indirect evidence points to astrocytes as a primary locus for the MAO B-catalyzed conversion of MPTP to MPP+. The role of astrocytes in MPTP biotransformation has been extensively documented by in vitro studies showing that: 1. glial cells in culture are indeed able to oxidize MPTP, first to MPDP+ via MAO and then to MPP+;61 and 2. neuronal cells which are killed by MPP+ but not by MPTP, become sensitive to the toxic effects of MPTP when cocultured with glial cells.62 If astrocytes represent the major source of MPP+, one might expect them to be particularly vulnerable to its cytotoxic properties. However, an astrocytic reaction, rather than overt glial damage, is a predominant feature of MPTP exposure in vivo. Changes in the expression of glial fibrillary acidic protein (GFAP) have been observed in the rodent brain after systemic administration of MPTP.63 Furthermore, neuropathological examination of the brain of monkeys with relatively long survival times after MPTP exposure (1 to 4 years) has revealed proliferation of cell processes and glial filaments and focal glial scars localized to the ventral and lateral cell groups of the substantia nigra.1 Recent studies in vitro using primary cultures of astrocytes have provided clues that may explain the apparent resistance of glial cells to MPTP toxicity.64-65 In order to cause cytotoxicity, relatively high intracellular levels of MPP+ have to be generated and maintained within astrocytes. Two mechanisms may prevent this from happening. First, MPP+ may be produced extracellularly rather than within astrocytes as a consequence of the following events: 1. MPTP is oxidized to MPDP+ by intraglial MAO; 2. MPDP+ crosses the plasma membranes of astrocytes in its lipophilic form of 1,2MPDP; and 3. MPDP+ generates MPP+ in the extracellular space, possibly via autoxidation. Second, even if produced within glial cells and despite its charged chemical structure, MPP+ appears capable of crossing astrocyte membranes and gaining access to the extracellular compartment;65 thus, MPP+ may not reach concentrations great enough and/or may not persist long enough to cause irreversible injury to astrocytes. In contrast, the active accumulation of MPP+ via the catecholamine uptake system29 (see section on MPTP-in-

Parkinson’s Disease

117

Fig. 6.2. Bioactivation of MPTP by astrocytic MAO B and accumulation of MPP+ in dopaminergic neurons.

duced Parkinsonism, above) explains its selective toxic effects on dopaminergic neurons. A schematic view of the interaction between astrocytes and dopaminergic neurons leading to MPTP bioactivation and MPP+ neurotoxicity is presented in Figure 6.2.

Astrocytes, MAO and Aging Onset of Parkinson’s disease before the age of 50 is quite rare, while the incidence of the disease increases with age in the over-50 population. Therefore, aging represents the most evident risk factor for Parkinson’s disease. The age-related mechanisms underlying neurodegeneration, though still unknown, may involve astrocytes and MAO. This possibility is supported by studies with the MPTP model of parkinsonism. It was noted that, in the mouse, the sensitivity to MPTP-induced dopamine depletion and degeneration of dopaminergic neurons increased with age.66-68 However, while MPTP-induced striatal dopamine depletion was more pronounced in older mice, the effects of direct exposure to MPP+ did not seem to be age-related.69 These findings suggested that differences in MPTP neurotoxicity with age involved changes in its conversion to the toxic MPP+ metabolite, catalyzed by MAO B. Subsequent studies proved a direct correlation between susceptibility to MPTPinduced nigrostriatal damage and MAO B activity; both increase in mice between 2 and 10 and between 10 and 16 months of age.20 The involvement of MAO B in the age-related toxicity of MPTP is unlikely to be peculiar to the mouse model. In fact, an increase in MAO B activity is thought to be a common feature of the aging mammalian brain and has also been described in humans.70-72 This increase is generally attributed to a greater proportion of MAO B-containing astrocytes in the aging brain.70-71 Thus, if MAO B catalyzes the metabolic activation of MPTP-like neurotoxins, astrocytes are likely to play an important role in rendering the aging brain increasingly susceptible to neurodegeneration. The age-related increase in MAO B activity within astrocytes may also lead to enhanced generation of H2O2, a product of any reaction catalyzed by MAO B. As previously discussed, dopamine itself may be a substrate for astrocytic MAO B. The contribution of this

118

Astrocytes in Brain Aging and Neurodegeneration

extraneuronal pathway of dopamine deamination could therefore increase with age, and H2O2 generated within astrocytes may become a relatively more prominent factor for neurodegeneration in the aging striatum and substantia nigra.

Astrocytes, Iron and Aging Another possible link between astrocytes, aging and neurodegeneration arises from the following observations: 1. increased total iron content has been found in nigral tissue of patients dying with Parkinson’s disease;9-10 2. a significant proportion of the excess iron reported in the Parkinsonian brain appears to be localized within astrocytes;73 and 3. an age-dependent increase in iron-containing astrocytic inclusions has been described in different areas of the brain, including the striatum.74 The role of iron as a catalyst of cytotoxic reactions may be attributed to its ability to reduce H2O2 to OH• in the Fenton reaction (Fig. 6.1). However, it is improbable that the generation of OH• within astrocytes would damage adjacent neuronal cells; OH• is a highly reactive oxygen species which, unlike H2O2, would not be expected to cross cell membranes. Thus, the mechanisms by which astrocytic iron may contribute to neurodegeneration must involve the formation of toxic metabolites other than OH•. An important requirement for iron to be involved in toxic reactions is its presence in a redox-active form. In this form, iron can act like a nonenzymatic peroxidase, oxidizing substrates and transferring electrons to H2O2.75 It is noteworthy therefore that inclusions observed within striatal astrocytes of the aging rodent and human brain exhibit peroxidase activity, most likely mediated by ferrous iron.74 This suggests that an increase in astrocytic iron content with age and possibly in pathological conditions may lead to the pseudoperoxidase-dependent formation of reactive metabolites and thus contribute to neurodegeneration (see chapter 11). A potential substrate for iron-catalyzed oxidation is dopamine itself. Indeed, experimental evidence indicates that catecholamines like dopamine are substrates for peroxidase activity.76-78 Furthermore, addition of iron to a solution of dopamine has been shown to cause increased dopamine autoxidation as measured by aminochrome (a dopamine-derived quinone) formation.79 Products of dopamine oxidation such as quinone derivatives could cross astrocyte membranes and inflict damage to neuronal cells via oxidative stress and/or binding to cellular macromolecules.6 Iron-catalyzed reactions could also play a role in the bioactivation of MPTP-like neurotoxins. Although the conversion of MPTP to MPP+ is mostly dependent on the activity of MAO B (see section on MPTP-induced Parkinsonism, above), recent results have shown that a small proportion of MPP+ formation still occurs in astrocyte cultures in the presence of MAO inhibitors and appears to require the catalytic activity of transition metals such as iron.80 A possible mechanism for this MAO-independent MPP+ generation involves the following reactions: 1. formation of the superoxide radical from oxidation of ferrous iron in the presence of oxygen;81 2. reaction of the superoxide radical with MPTP to produce a reactive intermediate which in turn would generate MPDP+;81 and 3. rapid nonenzymatic oxidation of MPDP+ to MPP+.24 The contribution of ironmediated activation to the overall conversion of MPTP, although minor under normal conditions, may become more significant with aging and in the Parkinsonian brain due to iron accumulation within neuronal and astrocytic cells.

Parkinson’s Disease

119

Astrocytes, Energy Metabolism and Excitotoxicity Prompted at least in part by the finding that MPTP is a mitochondrial poison, research on the pathogenesis of nigrostriatal damage has focused on the role of energy metabolism. A deficiency in mitochondrial oxidative phosphorylation and, more generally, in the ability of neurons to utilize glucose and to convert it into energy-rich pyrophosphate bonds in ATP would certainly be expected to place neurons at serious risk for degenerative processes. If this is the case for Parkinson’s disease, then an important aspect of the relationship between energy deficiency and neurodegeneration would be the interaction of neuronal cells with astrocytes. There is little doubt that the metabolic requirements of neuronal cells are dependent upon reactions that occur within astrocytes.82 These glial cells have been shown to protect neurons in culture against anoxia,83 and glycogen stores within astrocytes appear to enable neighboring neurons to survive glucose deprivation.84 Therefore, any future study aimed at elucidating the role of energy deficiency in nigrostriatal degeneration should approach the issue from the point of view of neuron-astrocyte interactions. We have recently investigated a specific mechanism by which a perturbation of energy metabolism in astrocytes may ultimately cause neuronal damage. This involves the ability of astrocytes to take up glutamate through a high-affinity glutamate-aspartate carrier.85 Maintenance of low extracellular concentrations of glutamate is a critical function for protecting neurons against the cytotoxic effects that result from a sustained activation of excitatory amino acid receptors (e.g., the NMDA receptor). Indeed, as illustrated in Figure 6.1, excitotoxicity has been proposed as a mechanism of nigrostriatal degeneration in the MPTP model of neurotoxicity as well as in idiopathic Parkinsonism. A series of experiments were conducted in our laboratory in order to test the consequences of energy impairment on glutamate uptake from the extracellular space in primary cultures of astrocytes. In one of these experiments, astrocytes were preincubated for 5 hours in the absence or presence of MPP+. Then, glutamate (500 µM) was added and its extracellular levels were monitored at 45, 90 and 150 min. As reported in Table 6.1, glutamate was efficiently removed by control astrocytes (without MPP+) and only 8% of its initial concentration was present extracellularly at the 150 min time point. In contrast, the ability of astrocytes to remove glutamate was significantly impaired after preincubation with MPP+ and, with 50 µM MPP+, 65% of the initial concentration of glutamate was still measured at 150 min in the extracellular compartment (Table 6.1). Since: 1. no sign of cytotoxicity was observed in MPP+-treated cultures during the time of the experiment; and 2. results similar to those seen with MPP+ could be obtained in cultures pretreated with rotenone (an inhibitor of mitochondrial complex I activity), it can be implied that any perturbation of energy metabolism in astrocytes significantly impairs their critical function of removing extracellular glutamate. The following sequence of events may be hypothesized that would link neurodegeneration to specific neuron/astrocyte interactions. Changes in energy metabolism take place in the astrocyte population because of: 1. aging (a decline in mitochondrial oxidative phosphorylation appears to occur in the aging brain of primates86); 2. pathological processes (a decrease in mitochondrial complex I activity has been reported in the Parkinsonian substantia nigra12-14); and/or 3. toxic exposure (MPTP is metabolically activated to the mitochondrial poison MPP+ within astrocytes). Astrocytes would then become unable to take up efficiently glutamate from the extracellular space, leading to a sustained activation of NMDA receptors and ultimately to excitotoxic injury.

Astrocytes in Brain Aging and Neurodegeneration

120

Table 6.1. Glutamate uptake by primary cultures of astrocytes in the presence and absence of MPP+ Pretreatment

None MPP+ (5 µM) MPP+ (50 µM)

Time (min) 0

45

90

150

100 ± 2.8 100 ± 6.8 100 ± 6.3

81.3 ± 3.9 101.3 ± 4.8* 95.1 ± 9.4*

41.7 ± 0.9 64.7 ± 2.0** 81.0 ± 4.2**

7.9 ± 0.8 32.4 ± 2.1** 65.9 ± 6.0**

Astrocytes were preincubated for 5 hours in the absence or presence of MPP+. Then glutamate (500 µM) was added, and its extracellular levels were monitored at 45, 90 and 150 min. Data (± SEM) are expressed as per cent of glutamate concentration measured at the 0 time point. * Statistically different (p<0.05) from the control group preincubated in the absence of MPP+. ** Statistically different (p<0.05) from the other two experimental groups.

Astrocytes and Nitric Oxide If the neurodegenerative process underlying Parkinson’s disease involves excitotoxicity, then it is possible that neuronal injury is mediated at least in part through the generation of nitric oxide. As already mentioned, when NMDA receptors are activated, the NMDA channel allows an influx of calcium into the cells. Several lines of evidence indicate that, after sustained receptor activation, an increase in cytosolic calcium concentrations stimulates calmodulin-regulated NOS activity and may lead to neuronal damage via excessive NO production. Indeed, NMDA-induced toxicity in neuronal cultures can be prevented by inhibitors of NOS.87 NOS inhibitors are also effective in protecting against ischemic cerebral damage, which is thought to be NMDA receptor-mediated in animal models.88 The involvement of NO in the pathogenesis of nigrostriatal damage in experimental and idiopathic Parkinsonism is suggested not only by findings concerning the MPTP and methamphetamine models of dopaminergic injury47-49,89 (see section on MPTP-induced Parkinsonism, above), but also by observations in the Parkinsonian brain. Hunot and colleagues have recently reported that immunoreactivity for the inducible form of NOS (iNOS) is increased in dopaminergic brain regions (i.e., the substantia nigra pars compacta, the ventral tegmental area and the A8 catecholaminergic cell group) of Parkinsonian patients as compared to control subjects.90 Interestingly, iNOS appeared to be localized to nonneuronal elements, possibly activated macrophages and/or astrocytes. Whether astrocyte-generated NO plays a causal role in dopaminergic degeneration or whether iNOS stimulation in glial cells occurs as a consequence of neuronal damage remains to be ascertained. It is already known, however, that astrocytes express iNOS when exposed to specific inducers, such as lipopolysaccharide or tumor necrosis factor-α (TNF-α).91 The effects of TNF-α may be of particular importance, since TNF-immunoreactive glial cells have been described in the substantia nigra of patients with Parkinson’s disease but not in control subjects.92 Thus, it is possible that induction of astrocytic NOS by cytokines like TNF-α may contribute to nigrostriatal damage, providing another mechanism by which astrocyte/neuronal interactions may lead to neurodegeneration in Parkinson’s disease.

Parkinson’s Disease

121

Astrocytes and Neurotrophic Factors It is not surprising that neurotrophic factors have increasingly become the subject of intense studies concerning neurodegenerative disorders. Their important role in neuronal development and subsistence has raised great expectations for their potential therapeutic efficacy against progressive cell loss and/or in support of neuronal regeneration. Nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophin 3 and 5 (NT-3 and NT-5) and basic fibroblast growth factor (bFGF) have all been tested as potential neuroprotective agents against dopaminergic degeneration; in the MPP+ model of neurotoxicity, they have shown various degrees of effectiveness.93 However, the identification of a glial cell line-derived neurotrophic factor (GDNF) in 1993 has significantly redirected the focus of scientists in the field of Parkinson’s disease to this new member of the transforming growth factor-β superfamily.94 This is because GDNF appears to be both more specific and more potent than other factors toward dopaminergic neurons. GDNF is expressed in the developing striatum and promotes the survival of mesencephalic dopaminergic neurons in culture.94-95 When used in animal models of nigrostriatal injury, it has been reported not only to protect against dopaminergic damage but also to induce repair mechanisms. In C57Bl/6 mice exposed to MPTP, GDNF injected over the substantia nigra or in the striatum prior to MPTP protected dopaminergic cell bodies and terminals.96 Interestingly, GDNF was partially effective against dopamine depletion even when injected 1 week after MPTP. This ability of GDNF to improve the recovery of dopaminergic neurons after toxic injury has also been documented in vitro. Short-term exposure of mesencephalic cultures to MPP+ for 1 hour resulted in a continuous loss of cells over the ensuing 5 days.97 However, if GDNF was added after MPP+ removal, it significantly prevented neuronal depletion and stimulated the regrowth of dopaminergic fibers. The therapeutic potential of GDNF in the treatment of Parkinson’s disease is already being explored.98 One major challenge is the need for a delivery strategy that would allow GDNF to reach target areas of the brain, bypassing the blood-brain barrier. Possible solutions to this problem may emerge from studies on the mechanisms of induction, production and secretion of neurotrophic factors by astrocytes during neuronal development and in the course of neurotoxic injury. These studies may lead to treatment of nigrostriatal degeneration with agents that stimulate GDNF production in situ, rendering its direct administration unnecessary.

Conclusion Secretion of neurotrophic factors by astrocytes as a mechanism of neuronal pruning and differentiation during development and as a possible signal for neuronal sprouting and regeneration during pathological/toxic events provides a clear illustration of astrocyte/neuron interactions. In this review article, we have discussed a number of other mechanisms by which such interactions may contribute to the pathogenesis of idiopathic and toxicantinduced Parkinsonism. In some cases, astrocytes could play a protective role against the development and progression of nigrostriatal degeneration. For example, they may counteract oxidative stress by virtue of their relatively high content of GSH and support tissue recovery via the production of neurotrophic factors. On the other hand, astrocytes may be directly involved in the biochemical/pathological changes leading to neurodegeneration. As previously seen, astrocytes may bioactivate MPTP-like neurotoxins and promote, for example, the generation of NO. It is also possible that neuron/astrocyte interactions underlie the increased vulnerability to nigrostriatal damage with aging. Changes in astrocyte levels of MAO and/or iron may contribute to this effect. Neurodegeneration is likely to result from an imbalance between toxic events and defense capabilities within the nigrostriatal tissue. Both toxic and defense mechanisms would

122

Astrocytes in Brain Aging and Neurodegeneration

therefore be expected to involve astrocyte/neuron interactions, accounting for the role of astrocytes in both counteracting and promoting neuronal injury. It is also important to reemphasize that the precise sequence of events leading to death of nigrostriatal dopaminergic neurons remains to be unraveled. For this reason, discussion on the role of astrocytes (and neurons, for that matter) remains somewhat speculative in nature. Perhaps the single most important goal of future research on Parkinson’s disease is to evaluate all of the hypothesized mechanisms of neurodegeneration (e.g., oxidative stress, mitochondrial failure and excitotoxicity) in order to identify causative factors. Astrocyte/neuron interactions should be an important component of these future studies, because there is little doubt that astrocytes are indeed active participants in both physiologic and pathological processes of the CNS.

Acknowledgments The author wishes to thank Dr. S.A. Jewell for her comments on the manuscript. This work was supported by The Parkinson’s Institute.

References 1. Forno LS, DeLanney LE, Irwin I et al. Astrocytes and Parkinson’s disease. In: Yu ACH, Hertz L, Norenberg MD, Sykova E, Waxman SG, eds. Progress in Brain Research. Vol 94. Amsterdam: Elsevier Science Publishers, 1992:429-436. 2. Tanner CM, Thelen JA, Offord KP et al. Parkinson’s disease (PD) incidence in Olmsted County, MN: 1935-1988. Neurology 1992; 42 (suppl 3): 194. 3. Forno LS. Neuropathology of Parkinson’s disease. J Neuropathol Exp Neurol 1996; 55:259-272. 4. Bernheimer H, Birkmeyer W, Hornykiewicz O. Brain dopamine and the syndromes of Parkinson and Huntington. J Neurol Sci 1973; 20:925-964. 5. Fahn S, Cohen G. The oxidant stress hypothesis in Parkinson’s disease: Evidence supporting it. Ann Neurol 1992; 32:804-812. 6. Graham DG. Oxidative pathways for catecholamines in the genesis of neuromelanin and cytotoxic quinones. Mol Pharmacol 1978; 14:633-643. 7. Spina MB, Cohen G. Dopamine turnover and glutathione oxidation: Implications for Parkinson’s disease. Proc Natl Acad Sci USA 1989; 86:1398-1400. 8. Cohen G, Spina MB. Deprenyl suppresses the oxidant stress associated with increased dopamine turnover. Ann Neurol 1989; 26:689-690. 9. Riederer P, Sofic E, Rausch WD et al. Transition metals, ferritin, glutathione, and ascorbic acid in Parkinsonian brains. J Neurochem 1989; 52:515-520. 10. Dexter DT, Weils FR, Lees AJ et al. Increased nigral iron content and alterations in other metal ions occurring in brain in Parkinson’s disease. J Neurochem 1989; 52:1830-1836. 11. Marey-Semper I, Gelman M, Levi-Strauss M. A selective toxicity toward cultured mesencephalic dopaminergic neurons is induced by the synergistic effects of energetic metabolism impairment and NMDA receptor activation. J Neurosci 1995; 15:5912-5918. 12. Schapira AHV, Cooper JM, Dexter DT et al. Mitochondrial complex I deficiency in Parkinson’s disease. J Neurochem 1990; 54:823-827. 13. Mizuno Y, Otha S, Tanaka M et al. Deficiencies in complex I subunits of the respiratory chain in Parkinson’s disease. Biochem Biophys Res Commun 1989; 163:1450-1455. 14. Schapira AHV, Mann VM, Cooper JM et al. Anatomic and disease specificity of NADH CoQ1 reductase (complex I) deficiency in Parkinson’s disease. J Neurochem 1990; 55:2142-2145. 15. Swerllow RH, Parks JK, Miller SW et al. Origin and functional consequences of the complex I defect in Parkinson’s disease. Ann Neurol 1996; 40:663-671. 16. Di Monte DA, Chan P, Sandy MS. Glutathione in Parkinson’s disease: A link between oxidative stress and mitochondrial damage? Ann Neurol 1992; 32: S111-S115.

Parkinson’s Disease

123

17. Langston JW, Ballard PA, Tetrud JW et al. Chronic Parkinsonism in humans due to a product of meperidine analog synthesis. Science 1983; 219:979-980. 18. Langston JW. MPTP and Parkinson’s disease. TINS 1985; 8:79-83. 19. Burns RS, Chiueh CC, Markey SP et al. A primate model of Parkinsonism: Selective destruction of dopaminergic neurons in the pars compacta of the substantia nigra by N-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. Proc Natl Acad Sci USA 1983; 80:4546-4550. 20. Irwin I, Finnegan KT, DeLanney LE et al. The relationships between aging, monoamine oxidase, striatal dopamine and the effects of MPTP in C57BL/6 mice: A critical reassessment. Brain Res 1992; 572:224-231. 21. Langston JW, Forno LS, Rebert CS et al. Selective nigral toxicity after systemic administration of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) in the squirrel monkey. Brain Res 1984; 292:390-394. 22. Chiba K, Trevor AJ, Castagnoli N Jr. Metabolism of the neurotoxic tertiary amine, MPTP, by brain monoamine oxidase. Biochem Biophys Res Commun 1984; 120:574-578. 23. Salach JI, Singer TP, Castagnoli N Jr et al. Oxidation of the neurotoxic amine 1-methyl-4phenyl-1,2,3,6-tetrahydropyridine (MPTP) by monoamine oxidases A and B and suicide inactivation of the enzymes by MPTP. Biochem Biophys Res Commun 1984; 125:831-835. 24. Wu EY, Langston JW, Di Monte DA. Toxicity of the 1-methyl-4-phenyl-2,3-dihydropyridinium and 1-methyl-4-phenylpyridinium species in primary cultures of mouse astrocytes. J Pharmacol Exp Ther 1992; 262:225-230. 25. Langston JW, Irwin I, Langston EB et al. Pargyline prevents MPTP-induced Parkinsonism in primates. Science 1984; 225:1480-1482. 26. Heikkila RE, Manzino L, Cabbat FS et al. Protection against the dopaminergic neurotoxicity of 1-methyl-4-phenyl-1,2,5,6-tetrahydropyridine by monoamine oxidase inhibitors. Nature 1984; 311:467-469. 27. Markey SP, Johannessen JN, Chiueh CC et al. Intraneuronal generation of a pyridinium metabolite may cause drug-induced Parkinsonism. Nature 1984; 311:464-467. 28. Bradbury AJ, Costall B, Domeney AM et al. 1-Methyl-4-phenylpyridine is neurotoxic to the nigrostriatal dopamine pathway. Nature 1986; 319:56-57. 29. Javitch JA, D’Amato RJ, Strittmatter SM et al. Parkinsonism-inducing neurotoxin, N-methyl-4-phenyl-1,2,3,6-tetrahydropyridine: Uptake of the metabolite N-methyl-4-phenylpyridine by dopamine neurons explains selective toxicity. Proc Natl Acad Sci USA 1985; 82:2173-2177. 30. Nicklas WJ, Vyas I, Heikkila RE. Inhibition of NADH-linked oxidation in brain mitochondria by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. Life Sci 1985; 36:2503-2508. 31. Ramsay RR., Singer TP. Energy-dependent uptake of N-methyl-4-phenylpyridinium, the neurotoxic metabolite of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine, by mitochondria. J Biol Chem 1986; 261:7585-7587. 32. Di Monte DA, Jewell SA, Ekstrom G et al. 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and 1-methyl-4-phenylpyridine (MPP+) cause rapid ATP depletion in isolated hepatocytes. Biochem Biophys Res Commun 1986; 137:310-315. 33. Denton T, Howard BD. A dopaminergic cell line variant resistant to the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. J Neurochem 1987; 49:622-629. 34. Chan P, Langston JW, Irwin I et al. 2-Deoxyglucose enhances 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine (MPTP)-induced ATP loss in the mouse brain. J Neurochem 1993; 61:610-616. 35. Cleeter MWJ, Cooper JM, Schapira AHV. Irreversible inhibition of complex I by 1-methyl-4-phenylpyridinium: Evidence for free radical involvement. J Neurochem 1992; 58:786-789. 36. Nicholls D, Attwell D. The release and uptake of excitatory amino acids. TIPS 1990; 11:462-468. 37. Henneberry RC, Novelli A, Vigano MA et al. Energy-related neurotoxicity at the NMDA receptor: A possible role in Alzheimer’s disease and related disorders. Prog Clin Biol Res 1989; 317:143-156.

124

Astrocytes in Brain Aging and Neurodegeneration

38. Tanaka S, Sako K, Tanaka T et al. Uncoupling of local blood flow and metabolism in the hippocampal CA3 in kainic acid-induced limbic seizure status. Neuroscience 1990; 36:339-48. 39. Tuong MDT, Brion F, Schwartz JC. Stimulation of deoxy[3H]glucose uptake into slice from cerebral cortex elicited by excitatory amino acids. Neuroscience 1984; 12:385-393. 40. Carafoli E, Crompton M. The regulation of intracellular calcium by mitochondria. Ann NY Acad Sci 1978; 307:269-284. 41. Kraus-Friedmann N, Biber J, Murer H et al. Calcium uptake in isolated hepatic plasmamembrane vesicles. Eur J Biochem 1982; 129:7-12. 42. Turski L, Bressler K, Rettig K et al. Protection of substantia nigra from MPP+ neurotoxicity by N-methyl-D-aspartate antagonists. Nature 1991; 349:414-418. 43. Zuddas A, Oberto G, Vaglini F et al. MK-801 prevents 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine-induced Parkinsonism in primates. J Neurochem 1992; 59:733-739. 44. Sonsalla PK, Zeevalk GD, Manzino L et al. MK-801 fails to protect against the dopaminergic neuropathology produced by systemic 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine in mice or intranigral 1-methyl-4-phenylpyridinium in rats. J Neurochem 1992; 58:1979-1982. 45. Chan P, Langston JW, Di Monte DA. MK-801 temporarily prevents MPTP-induced acute dopamine depletion and MPP+ elimination in the mouse striatum. J Pharmacol Exp Ther 1993; 267:1515-1520. 46. Varner PD, Beckman JS. Nitric oxide toxicity in neuronal injury and degeneration. In: Vincent S, ed. Nitric Oxide in the Nervous System. San Diego: Academic Press, 1995:191-206. 47. Schultz JB, Matthews RT, Muqit MMK et al. Inhibition of neuronal nitric oxide synthase by 7-nitroindazole protects against MPTP-induced neurotoxicity in mice. J Neurochem 1995; 64:936-939. 48. Przedborski S, Jackson-Lewis V, Yokoyama R et al. Role of neuronal nitric oxide in MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine)-induced dopaminergic neurotoxicity. Proc Natl Acad Sci USA 1996; 93:4565-4571. 49. Hantraye P, Brouillet E, Ferrante R et al. Inhibition of neuronal nitric oxide synthase prevents MPTP-induced Parkinsonism in baboons. Nature Med 1996; 2:1017-1022. 50. Di Monte DA, DeLanney LE, Irwin I et al. Monoamine oxidase-dependent metabolism of dopamine in the striatum and substantia nigra of L-dopa-treated monkeys. Brain Res 1996; 738:53-59. 51. Westlund KN, Denney RM, Kochersperger LM et al. Distinct monoamine oxidase A and B populations in primate brain. Science 1985; 230:181-183. 52. Levitt P, Pintar JE, Breakefield XO. Immunocytochemical demonstration of monoamine oxidase B in brain astrocytes and serotonergic neurons. Proc Natl Acad Sci USA 1982; 79:6385-6389. 53. Yu PH, Hertz L. Differential expression of type A and type B monoamine oxidase of mouse astrocytes in primary cultures. J Neurochem 1982; 39:1492-1495. 54. Semenoff D, Kimelberg HK. Autoradiography of high affinity uptake of catecholamine by primary astrocyte cultures. Brain Res 1985; 348:125-136. 55. Halliwell B, Gutteridge JMC. Oxygen toxicity, oxygen radicals, transition metals and disease. Biochem J 1984; 219:1-14. 56. Slivka A, Mytilineou C, Cohen G. Histochemical evaluation of glutathione in brain. Brain Res 1987; 409:275-284. 57. Raps SP, Lai JCK, Hertz L et al. Glutathione is present in high concentrations in cultured astrocytes but not in cultured neurons. Brain Res 1989; 493:398-401. 58. Perry TL, Godin DV, Hansen S. Parkinson’s disease: A disorder due to nigral glutathione deficiency? Neurosci Lett 1982; 33:305-310. 59. Vincent SR. Histochemical localization of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine oxidation in the mouse brain. Neuroscience 1989; 28:189-199. 60. Melamed E, Pikarski E, Goldberg A et al. Effect of serotonergic, corticostriatal and kainic acid lesions on the dopaminergic neurotoxicity of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) in mice. Brain Res 1986; 399:178-180.

Parkinson’s Disease

125

61. Di Monte DA, Wu EY, Irwin I et al. Biotransformation of 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine in primary cultures of mouse astrocytes. J Pharmacol Exp Ther 1991; 258:594-600. 62. Notter MSD, Irwin I, Langston JW et al. Neurotoxicity of MPTP and MPP+ in vitro: Characterization using specific cell lines. Brain Res 1988; 456:254-262. 63. Reinhard JF Jr, Miller D, O’Callaghan JP. The neurotoxicant MPTP (1-methyl-4-phenyl1,2,3,6-tetrahydropyridine) increases glial fibrillary acidic protein and decreases dopamine levels of the mouse striatum: Evidence for glial response to injury. Neurosci Lett 1988; 95:246-251. 64. Di Monte DA, Wu EY, DeLanney LE et al. Toxicity of 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine in primary cultures of mouse astrocytes. J Pharmacol Exp Ther 1992; 261:44-49. 65. Di Monte DA, Wu EY, Irwin I et al. Production and disposition of 1-methyl-4phenylpyridinium in primary cultures of mouse astrocytes. Glia 1992; 5:48-55. 66. Jarvis MF, Wagner GC. Age-dependent effects of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). Neuropharmacology 1985; 24:581-583. 67. Gupta M, Gupta BK, Thomas R et al. Aged mice are more sensitive to 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine than young adults. Neurosci Lett 1986; 70:326-331. 68. Ricaurte GA, Irwin I, Forno LS et al. Aging and MPTP-induced degeneration of dopaminergic neurons in the substantia nigra. Brain Res 1987; 403:43-51. 69. Irwin I, Ricaurte GA, DeLanney LE et al. The sensitivity of nigrostriatal dopamine neurons to MPP+ does not increase with age. Neurosci Lett 1988; 87:51-56. 70. Benedetti MS, Keane PE. Differential changes in monoamine oxidase A and B activity in the aging rat brain. J Neurochem 1980; 35:1026-1032. 71. Fowler J, Wilberg A, Oreland L et al. The effect of age on the activity and molecular properties of human brain monoamine oxidase. J Neural Transm 1980; 49:1-20. 72. Robinson DS, Davis JM, Nies A et al. Relation of sex and aging to monoamine oxidase activity of human brain, plasma, and platelets. Arch Gen Psychiatry 1971; 24:536-539. 73. Jellinger K, Paulus W, Grundke-Iqbal I et al. Brain iron and ferritin in Parkinson’s and Alzheimer diseases. J Neural Transm 1990; 2:327-340. 74. Schipper HM. Gomori-positive astrocytes: Biological properties and implications for neurologic and neuroendocrine disorders. Glia 1991; 4:365-377. 75. Miller DM, Buettner GR, Aust SD. Transition metals as catalysts of “autoxidation” reactions. Free Radical Biol Med 1990; 8:95-108. 76. Kalyanaraman B, Felix CC, Sealy RC. Peroxidatic oxidation of catecholamines: A kinetic electron spin resonance investigation using the spin stabilization approach. J Biol Chem 1984; 259:7584-7589. 77. Metodiewa D., Dunford HB. The role of myeloperoxidase in the oxidation of biologically active polyhydroxyphenols (substitutes catechols). Eur J Biochem 1990; 193:445-448. 78. Schipper HM, Kotake Y, Janzen EG. Catechol oxidation by peroxidase-positive astrocytes in primary culture: An electron spin resonance study. J Neurosci 1991; 11:2170-2176. 79. Poirier J, Donaldson J, Barbeau A. The specific vulnerability of the substantia nigra to MPTP is related to the presence of transition metals. Biochem Biophys Res Commun 1985; 128:25-33. 80. Di Monte DA, Schipper HM, Hetts S et al. Iron-mediated bioactivation of 1-methyl-4phenyl-1,2,3,6-tetrahydropyridine (MPTP) in glial cultures. Glia 1995; 15:203-206. 81. Poirier J, Barbeau A. A catalyst function for MPTP in superoxide formation. Biochem Biophys Res Commun 1985; 131:1284-1289. 82. Hamprecht B, Dringen R. Energy metabolism. In: Kettenman H, Ransom BR, eds. Neuroglia. New York: Oxford University Press, 1995:473-487. 83. Vibulsreth S, Hefti F, Ginsberg MD et al. Astrocytes protect cultured neurons from degeneration induced by anoxia. Brain Res 1987; 422:303-311. 84. Swanson RA, Choi DW. Glial glycogen stores affect neuronal survival during glucose deprivation in vitro. J Cereb Blood Flow Metab 1993; 13:162-169.

126

Astrocytes in Brain Aging and Neurodegeneration

85. Schousboe A, Westergaard N. Transport of neuroactive amino acids in astrocytes. In: Kettenman H, Ransom BR, eds. Neuroglia. New York: Oxford University Press, 1995:246-258. 86. Di Monte DA, Sandy MS, DeLanney LE et al. Age-dependent changes in mitochondrial energy production in striatum and cerebellum of the monkey brain. Neurodegeneration 1993; 2:93-99. 87. Dawson VL, Dawson TM, Bartley DA et al. Mechanisms of nitric oxide-mediated neurotoxicity in primary brain cultures. J Neurosci 1993; 13:2651-2661. 88. Nowicki JP, Duval D, Poignet H et al. Nitric oxide mediates neuronal death after focal cerebral ischemia in the mouse. Eur J Pharmacol 1991; 204:339-340. 89. Di Monte DA, Royland J, Jakowec MW et al. Role of nitric oxide in methamphetamine neurotoxicity: Protection by 7-nitroindazole, an inhibitor of neuronal nitric oxide synthase. J Neurochem 1996; 67:2443-2450. 90. Hunot S, Boissiere F, Faucheux B et al. Nitric oxide synthase and neuronal vulnerability in Parkinson’s disease. Neuroscience 1996; 72:355-363. 91. Vigne P, Damais C, Frelin C. IL1 and TNFα induce cGMP formation in C6 astrocytoma cells via the nitridergic pathway. Brain Res 1993; 606:332-336. 92. Boka G, Anglade P, Wallach D et al. Immunocytochemical analysis of tumor necrosis factor and its receptors in Parkinson’s disease. Neurosci Lett 1994; 172:151-154. 93. Kirschner PB, Jenkins BG, Schulz JB et al. NGF, BDNF and NT-5, but not NT-3 protect against MPP+ toxicity and oxidative stress in neonatal animals. Brain Res 1996; 713:178-185. 94. Lin L-FH, Doherty DH, Lile JD et al. GDNF: A glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science 1993; 260:1130-1132. 95. Schaar DG, Sieber B-A, Dreyfus CF et al. Regional and cell-specific expression of GDNF in rat brain. Exp Neurol 1993; 124:368-371. 96. Tomac A, Lindqvist E, Lin L-FH et al. Protection and repair of the nigrostriatal dopaminergic system by GDNF in vivo. Nature 1995; 373:335-339. 97. Hou J-GG, Lin L-FH, Mytilineou C. Glial cell line-derived neurotrophic factor exerts neurotrophic effects on dopaminergic neurons in vitro and promotes their survival and regrowth after damage by 1-methyl-4-phenylpyridinium. J Neurochem 1996; 66:74-82. 98. Gash DM, Zhang Z, Ovadla A et al. Functional recovery in Parkinsonian monkeys treated with GDNF. Nature 1996; 380:252-255.

CHAPTER 7

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases) Pawel P. Liberski, Radzislaw Kordek, Paul Brown and D. Carleton Gajdusek

Introduction

T

he transmissible spongiform encephalopathies (TSE) or prion diseases are a group of neurodegenerative disorders which include kuru (Fig. 7.1), Creutzfeldt-Jakob disease (CJD), Gerstmann-Straussler-Scheinker (GSS) disease, and fatal familial insomnia (FFI) in man; natural scrapie in sheep, goats and mufflons; transmissible mink encephalopathy (TME) in ranch-reared mink; chronic wasting disease (CWD) of captive and free living mule deer and elk in the USA; bovine spongiform encephalopathy (BSE) or “mad cow disease” and its analogues in several exotic species of ungulates, a puma and several cheetahs from British zoological gardens; and feline spongiform encephalopathy in domestic cats.1-5 The status of spongiform encephalopathy in ostrich is still unclear but it is probably unrelated to TSEs.6 Recently, a new variant of CJD (vCJD) was described among young people in the UK.7 This vCJD is characterized by florid PrP plaques surrounded by a corona of vacuoles (Fig. 7.2). As virtually identical plaques were reproduced in macaques inoculated with BSE material8 and the same glycosylation9 pattern is observed in BSE and vCJD, the transmission from BSE to humans seems increasingly likely. TSEs are caused by a still incompletely understood pathogen variously referred to as a virus (usually with the adjectives slow, unconventional or atypical), agent, “prion” or “virino”. These names reflect, in part, different views on the molecular structure of the pathogen and, by the same token, our ignorance of its nature.10 Those who prefer to view this pathogen as composed “predominantly or entirely” of one protein, PrP, use the term “prion” hence the term “prion disorders”.11-12 The last term, however, implies more than a semantic preference: It suggests that neuropathologically confirmed cases of TSE are only the “tip of the iceberg” of poorly delineated conditions (“dementias without a characteristic pathology”) which share abnormalities in the PRNP gene (the gene which encodes for PrP).13 To evaluate the validity of this claim, one of us searched for PrP in 46 cases of “nonspongiform” dementias; none was positive.14 It was concluded that “for all intents and purposes ‘prion dementia’ and ‘spongiform encephalopathy’ are one and the same”. Irrespective of the nature of the agent, it is widely accepted that the abnormal isoform (probably also an abnormal conformer) PrPsc plays a crucial role in the pathogenesis of the whole group of TSEs.15-16 This protein accumulates in all TSE-affected brains, either as Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

128

Astrocytes in Brain Aging and Neurodegeneration

Fig. 7.1. A preadolescent child, totally incapacitated by kuru in 1957. The child had such severe dysarthria that he could no longer communicate by word, but was still intelligent and alert. He had spastic strabismus. He could not stand, sit without support, or even roll over; he had been ill for less than six months and died within a few months of the time of photography.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

129

Fig. 7.2. Astrocytes within the perimeter of florid plaques in a case of the new variant of CJD. Inset, a typical florid plaque surrounded by a corona of spongiform change. GFAP immunohistochemistry, original magnification x 1000; inset, H & E. By courtesy of Dr. James Ironside, Edinburgh, Scotland.

amyloid plaques or plaque-like deposits or in the so-called synaptic forms.17 Furthermore, alterations (point mutations, insertions and deletions) in the gene encoding PrP (PRNP in humans) cosegregate with various phenotypic presentations of CJD, GSS or FFI.18 The “virino” hypothesis suggests that the pathogen is an unprecedented molecular chimera composed of a still to be discovered nucleic acid and a shell protein which is hostencoded (possibly even PrP).19 The virus hypothesis simply suggests that the pathogen is yet to be purified, as there is no conclusive evidence to prove that it is outside the spectrum of conventional viruses.20 The “unified theory” of Weissmann suggests that, not unlike the virino theory,19 the agent is a molecular chimera,21-22 in which PrPsc confers infectivity, while a still undetected oligonucleotide specifies strain characteristics; in other words, the agent has a genome that is unnecessary for infectivity. Astrocytosis or reactive gliosis is a prominent feature of naturally occurring and experimentally produced TSEs.23-25 It is also a feature of a new entity “familial progressive subcortical gliosis” in which PrP accumulates in the brain.26 In contrast to familial CJD and GSS cases, which are linked to chromosome 20, familial progressive gliosis is linked to chromosome 17. In this chapter we shall review diverse aspects of astrocytic gliosis in naturally occurring and experimentally induced TSE.

130

Astrocytes in Brain Aging and Neurodegeneration

KURU Natural Disease Historically, astrocytic hypertrophy and proliferation have been stressed as a hallmark of kuru27-28 and supported by more recent systematic immunohistochemical studies.29 Astrocytic proliferation was widespread and more abundant in gray than in white matter. Usually, astrocytosis paralleled neuronal destruction, but it has also been observed in regions with only minimal brain pathology. In the pons, severe gliosis was observed in the tegmental and basal portions, with a conspicuous sparing of the pyramidal tracts and medial lemnisci. Gliosis was severe in the midbrain, basal ganglia, thalamus, subcortical white matter and in the cerebellum where the vermis was mostly affected. Conspicuously, Fañanás cell proliferation has been noticed. In the cerebral cortex, proliferation of astrocytes was in excess of other pathological changes. Furthermore, astrocytosis was diffusely present in the anterior horns of the spinal cord. Some astrocytes showed clasmatodendrosis. In contrast, Neuman et al30 did not observe severe astrocytosis in three kuru patients, while Scrimgeour at al31 found only mild astrocytic changes with the presence of rare binucleated forms in the cerebral cortex.

Experimental Studies Experimental kuru in chimpanzees is characterized by widespread astrocytosis, and both hypertrophy and proliferation of astrocytes have been observed.32-34 Gliosis seems to parallel the severity of spongiform change and neuronal loss, being most abundant in markedly vacuolated sensory cortex and less so in better preserved motor areas. Striatum, diencephalon, cerebral white matter and cerebellum show severe gliosis. Ultrastructurally, astrocytes show focal and, in our opinion, artifactual, attenuation of the cytoplasm and accumulations of glycogen granules.34 In a separate unpublished study of early changes in New World monkeys infected with kuru, Liberski, Brown and Gajdusek found, using GFAPimmunohistochemistry and electron microscopy, only moderate astrocytosis. Interestingly, astrocytes were observed adjacent to cerebellar granule cells undergoing faulty myelination. The biological significance of this phenomenon is unknown.

Creutzfeldt-Jakob Disease (CJD) and Gerstmann-Straussler-Scheinker Disease (GSS) CJD Natural disease Variably severe astrocytosis is observed among almost all neurodegenerative conditions, and CJD is no exception. Hypertrophic astrocytes, detected by means of metal techniques (Holzer, Kanzler or Cajal’s stains) or more recently by immunostaining against glial fibrillary acid protein (GFAP), are seen in all vacuolated areas (Figs. 7.3, 7.4). In cerebral cortex they are particularly prominent in deeper cortical layers. Gemistocytic forms are frequently observed (Fig. 7.5). When destruction is so severe as to lead to the collapse of vacuolated neuropil, proliferating astrocytes may virtually replace all other cellular elements. In such a situation the spongiform changes may no longer be recognizable. In the cerebellum, the proliferation of Bergman glia is frequently observed. Marin and Vial35 were the first to report on the ultrastructure of human CJD. They described hypertrophic astrocytes, a proportion of which suffered from suboptimal fixation and produced areas of “watery” cytoplasm. Astrocytic cytoplasm contained numerous lipofuscin granules. A similar distinction between hypertrophic astrocytes containing glial

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

131

Fig. 7.3. Whole mount section of kuru brain stained with Kanzler stain to detect astrocytosis. The distribution of fibrillary gliosis is similar in the middle and deep layers of parasagittal and interhemispheric neocortex and cingulate cortex but more diffuse in the thalamus and underneath the insular and temporal cortices. Reprinted with permission from Hainfellner JA et al, Brain Pathol 1997; 7:547-553.

fibrils and those of “watery” cytoplasm was made by Gonatas et al36 Torrack,37-38 Brion et al,39 Bubis et al40 and Ribadeau-Dumas and Escourolle.41-42 Astrocytic nuclei frequently contain inclusion bodies.43 The first type consists of granular and filamentous profiles, frequently forming paracrystalline arrays, and most probably represents deformed chromatin. The other type corresponds to the IVth type of “nuclear bodies” (Fig. 7.6) according to the classification of Bouteille et al.44 The first type of inclusion results most probably from suboptimal fixation, but the type IV nuclear body has been frequently found in infectious and neoplastic disorders and is regarded as a nonspecific reaction of cells to noxious stimuli.45 Still another type of intranuclear inclusion reported by Jellinger46 corresponds to type A nuclear inclusions. As these inclusions have been reported in numerous viral and degenerative conditions, as well as a result of abnormal mitoses, they also most probably represent nonspecific changes. There are only two overlapping morphometric studies of astrogliosis in the cerebella (both Bergmann and velate astrocytes) in two cases of the ataxic form of CJD.47-48 Astrocytes increased from 192.76 + 117.98 cells per mm2 in controls to 278.08 + 137.73 per mm2 in CJD. An increase in the cross-sectioned nuclear area of Bergmann glia (32.72 + 6.8 mm2 vs. 42.75 + 9.61 mm2) and of velate astrocytes (34.86 + 7.29 mm2 vs. 39.37 + 7.10 mm2) was seen when control values were compared with those of CJD values. Of note, the basic threedimensional geometry of the astrocytic scaffold of the cerebellum was maintained despite severe loss of granule cells. Electron microscopy revealed several subcellular organelles, rare but otherwise typical for reactive astrocytes, single cilia consisting of ciliary shafts (Fig. 7.7),

132

Astrocytes in Brain Aging and Neurodegeneration

a b

Fig. 7.4. Natural CJD. Severe astrocytosis detected by Cajal gold sublimate impregnation method (a) and GFAP immunohistochemistry (b). Note binucleated astrocytes (arrows) in (b). Original magnification, x 1000.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

133

Fig. 7.5. Numerous GFAP-immunopositive gemistocytic astrocytes in the cerebral cortex of CJDaffected brain. x 400.

clusters of interchromatin and perichromatin granules, various adhesive plaque junctions (Fig. 7.8) and simple and granular nuclear bodies. Of particular interest is the presence of infoldings of plasma membranes in the perivascular regions of astrocytic end-feet. These infoldings were covered by an interrupted or continuous electron-dense undercoat of 30-60 nm in diameter. The latter observation is in agreement with the earlier freeze-etching study of Dubois-Dalcq et al49 who showed an increased number of astrocyte-specific particles, as opposed to their depletion on membranes, forming vacuoles. Experimental studies Astrocytosis represents a substantial component of the neuropathological profile of experimental CJD. In the first reported transmission experiment, Beck et al50 found moderate to severe astrocytosis in both biopsy and necropsy specimens of CJD virus-infected chimpanzees. In the cerebral cortex, the hypertrophic astrocytes completely distorted the neuronal architecture. Many astrocytes were of the gemistocytic type, similar to those in human CJD. Severe glial reaction was also seen in the striatum, diencephalon and cerebellar cortex. Beck et al50 raised the problem of astrocytes as a primary target for CJD agent, in other words, the location of CJD within a vague spectrum of so-called “glial dystrophies”. This notion was based primarily on observed discrepancies between the severity of astrocytosis and neuronal damage. While such differences have been unequivocally noted, it must be stressed that in most situations the most severely vacuolated brain regions also presented the highest level of astrocytosis. Manuelidis and colleagues51-54 found particularly severe astrocytosis in experimental CJD in guinea pigs, hamsters and mice. In CJD affected-hamsters “clusters of these cells appear almost as pure astrocytic cultures; this collection is far in excess of what classically in human and experimental neuropathology is known as reactive astrocytosis”. To further substantiate the notion of a primary involvement of astrocytes in CJD, Manuelidis and Manuelidis reported that astrocytes from CJD-affected brains could be maintained in vivo (immortalized) for a long time.54 In contrast, those established from

Astrocytes in Brain Aging and Neurodegeneration

134

a b

Fig. 7.6. (a, b). Two examples of nuclear bodies. Lead citrate and uranyl acetate, original magnification, x 30,000.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

135

Fig. 7.7. An astrocyte containing a cilium (arrow). Lead citrate and uranyl acetate, original magnification, x 12,000.

uninfected brains all died after a short period of time. This problem was further addressed in a serial killing experiment, in which we found by means of electron microscopy that astrocytosis paralleled spongiform change in parietal cortex and adjacent corpus callosum of mice infected with the Fujisaki strain of CJD agent.55 As dilated and swollen astrocytic processes were found occasionally in both CJD-infected and sham-inoculated animals, these were regarded as a result of local suboptimal fixation and not as part of CJD neuropathology as reported by others.3,4,51-52 Recently, the neuropathology of TSE has been partially reproduced in transgenic mice created by microinjection of the chimeric murine cosmid containing a codon 101 Pro to Leu substitution in the ORF of mouse PrP gene.56 The 101 codon substitution is regarded as an equivalent to that found in GSS (a deletion of codon 55 in the mouse PrP gene).57-58 Transgenic mice presented severe spongiform change but rather mild or moderate astrocytosis, except in the cerebellum where severe Bergmann radial gliosis was observed. Again, spongiform change rather than astrocytosis seems to be the primary neuropathological phenomenon in TSEs.

GSS While astrocytic gliosis is a prominent and ubiquitous finding in CJD, kuru, scrapie and bovine spongiform encephalopathy, it still remains a controversial issue in GSS. Hudson et al59 found mild astrocytosis associated mostly with amyloid plaques in 3 cases of GSS. In a case reported by Kuzuhara et al,60 moderate astrocytosis of the cerebellar white matter was found, while a severe astrocytic reaction was seen in the inferior colliculus. Vinters et al61 reported astrocytic gliosis throughout the neocortex while, in contrast, Tateishi et al62 found astrocytosis only in an area of concomitant infarct. Similarly, Ghetti et al,63 Nochlin et al64 and Pearlman et al65 reported severe gliosis only in areas where numerous plaques and neuronal loss were also seen. In 3 cases of GSS studied by us at the Laboratory of Central Nervous System Studies (LCNSS), National Institutes of Health in Bethesda, and in the Neurological Institute of the University of Vienna, astrocytosis was found throughout the cerebral and

Astrocytes in Brain Aging and Neurodegeneration

136

a b

Fig. 7.8. (a) Symmetric desmosome-like adhesive plaque junction (arrows); (b) long tortuous adhesive plaque junctions (arrows) in an astrocyte in the subependymal space. Note that one of these junctions opens to the extracellular space. Lead citrate and uranyl acetate, original magnification, x 50,000.

cerebellar cortex, but the severity of this change never approached that found in CJD cases. In particular, gemistocytic astrocytes were never seen in these cases; rather, the astrocytes were characteristically elongated and slender, reminiscent of pilocytic astrocytes. However, in a recent case from the original Austrian GSS family, astrocytosis in the cerebral cortex approached that of CJD brains and innumerable gemistocytic astrocytes were seen.66-68 Thus, the diversity of neuropathology of GSS is perhaps of the same magnitude as that of CJD.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

137

The Involvement of Astrocytes in Formation of Amyloid Plaques Amyloid plaques are a neuropathological feature of TSEs.69 There is an “unnecessarily complex”69 classification of amyloid deposits in humans into several partially overlapping categories.70 Kuru plaques and multicentric plaques are characteristic features of kuru (or CJD) and GSS, respectively (Fig. 7.9). Cortical kuru (unicentric) plaques of GSS consist of amyloid fibrils within a narrow extracellular space between distended astrocytic processes (Fig. 7.10).71-73 Amyloid fibrils invaginated “deeply the surrounding profiles of astrocytes so that the filaments sometimes seemed to be intracellular”. Such peripheral accumulations of astrocytic processes in close connection with the amyloid fibrils was noted even in the earliest amyloid plaques.71-73 This intimate association of amyloid and astrocytes in GSS led Boellaard et al72 to coin the term glial plaques. Glial plaques are plaques of TSEs and contrast with neuritic plaques of Alzheimer’s disease,74 but, like the latter, are invaded by microglial cells.75-76 A systematic immunohistochemical approach disclosed that 30% to 50% of unicentric plaques contain microglia, while astrocytes are located around these plaques with long processes penetrating them.75 In contrast, only some of the multicentric plaques contain microglial cells, but the pattern of astrocytic involvement is practically the same. These authors also discriminate “cores with satellite deposits” (a variant of multicentric plaques). Seventy to 80% of the latter contain microglia cells; the pattern of astrocyte involvement remains unchanged. In contrast, diffuse (primitive) plaques in mice exhibit neither amyloid cores nor amyloid filaments.77-80 They are infiltrated by neither astrocytes nor microglial cells. However, as plaques mature and the PrP within them fibrilizes, the number of both microglial cells and astrocytes tends to increase. Thus, it seems that both categories of glial cells may merely be reactive cells, and elegant immunogold studies have shown that PrP is indeed localized to lysosomal compartments of these cells.

Scrapie, Bovine Spongiform Encephalopathy (BSE), and Chronic Wasting Disease (CWD) Scrapie Natural disease Generally, the degenerative brain pathology of natural sheep and goat scrapie consists of spongiform change and astrocytosis. The latter change is highly variable; many cases of natural scrapie in sheep show inconspicuous or undetectable astrocytosis. In the majority of textbook descriptions “neuronal loss” is also mentioned. However, the only reference supporting this statement is that of Beck et al81 who studied brain areas which are characterized by highly variable numbers of neurons. In contrast, neuronal loss clearly does occur in BSE.82 The true nature of the astrocytic changes in scrapie remains poorly understood. It still remains to be resolved, for example, whether astrocytic proliferation (hyperplasia), astrocytic hypertrophy or both are responsible for the apparent increases in numbers of astrocytes observed in scrapie brain under light microscopy.83 Hadlow,84 studying scrapie-infected dairy goats, reported that both hypertrophy and, to a lesser extent, hyperplasia brought about an overall increase in the apparent number of astrocytes seen in sections. The estimation of the number of astrocytes was difficult, however, as the Cajal method used also stains different proportions of astrocytes in normal goat brains. The “scrapie” astrocytes were not always easily discriminated from pleomorphic “normal” astrocytes, but typically they measured up to 14 mm in diameter and contained a few chromatin granules. In particular, the

138

Astrocytes in Brain Aging and Neurodegeneration

Fig. 7.9. General view of GSS pathology. Note a multicentric plaque (circle) and three kuru (unicentric) plaques (squares). A microglial cell (arrow) is visible at the periphery. Lead citrate and uranyl acetate, original magnification, x 4400.

presence of kidney-shaped, elongated and irregularly lobulated nuclei, frequently noted in clusters of 3 to 4 cells reminiscent of those in Alzheimer II cells or the “naked nuclei” of Alzheimer, have been reported. Hypertrophy and proliferation of astrocytes were confined to the affected (vacuolated) gray matter. The adjacent white matter was involved only occasionally. However, in the midbrain and several thalamic nuclei, astrocytosis was more severe than vacuolation and cerebral cortex characterized by minimal spongiform change occasionally presented disproportionately spectacular astrocytosis. Topographically, various brain regions were involved to different degrees. The lesions tended to be bilaterally symmetrical and the boundaries between affected and unaffected

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

139

Fig. 7.10. Light microscopy of kuru (unicentric) plaques of GSS. GFAPimmunoreactive astrocytes are seen around the unstained (star) plaque;, GFAP-immunohistochemistry; x 400. By courtesy of Dr. Maria Barcikowska, Medical Research Centre, the Polish Academy of Sciences, Warsaw.

regions were remarkably sharp. Dense astrocytosis was observed in the pallidum, septal nuclei, and diencephalon. Moderate astrocytosis was seen in the striatum and the brain stem, where hypertrophy prevailed above hyperplasia. Minimal astrocytic hypertrophy with slight proliferation was seen in deeper layers of cerebral and cerebellar cortex which otherwise remained largely unaffected by the disease. In the cerebellar cortex, radially arranged glial processes and clusters of Bergmann glial cells constituted distinctive features. The hippocampus formation was unaffected, but diffuse astrocytosis was evident, mostly between pyramidal cells and the alveus. Hypertrophy and “undoubted” proliferation has been also detected in natural scrapie in goats.85 Topographically, astrocytosis of both experimental and natural caprine scrapie were alike, except that the striatum, pallidum and septal nuclei were only slightly affected in the latter. A similar increase in the number of hypertrophic astrocytes in sheep with natural scrapie was reported recently.86 Of note, the number of astrocytes diminished with age both in controls and in scrapie-affected sheep. In the latter group, however, this decrease was not as pronounced. No association was found between the degree of astrocytosis and duration of clinical disease or severity of spongiform change. Experimental studies Pattison and Jones stated that astrocyte hypertrophy, but not proliferation, was a feature of rats infected with the Chandler strain of scrapie agent.87 Astrocytosis mostly paralleled spongiform change and was greater after intracerebral inoculation than after intraperitoneal inoculation. Astrocytosis preceded the vacuolation by 14 days. Astrocytic end-plates were hypertrophic, and in the later stages of disease “capillaries appeared to be embedded in swollen, darkly staining astrocytic cytoplasm.” The problem of hypertrophy versus hyperplasia has been studied by Fraser and colleagues88-89 in several models of murine scrapie and by Liberski and colleagues90-93 in hamsters infected with the 263K strain of scrapie agent. Fraser88 coined the term gliocytosis to denote proliferation of astrocytes accompanied by some changes in their morphology and substantial proliferation of rod-like microglial cells. In murine scrapie, gliocytosis, encountered in the hippocampus and the thalamus, is an extremely rare phenomenon found in approximately 3% of ten thousand murine scrapie-affected brains. Gliocytosis occurs in a wide range of scrapie isolates passaged in different strains of mice, but almost exclusively after intracerebral inoculation (256 examples of 260 studied brains with gliocytosis88). In more detailed studies of gliocytosis (sclerosis) of the hippocampus formation, Scott and Fraser89 found that its presence paralleled that of severe vacuolation. Liberski and colleagues found “gliocytosis” in hamsters infected with the 263K strain of scrapie in much higher proportion than that of murine scrapie models.90-93 Both astro-

140

Astrocytes in Brain Aging and Neurodegeneration

Fig. 7.11. Experimental scrapie in hamsters. A dividing astrocyte. Note chromatin (arrows) and a part of apparatus (arrowheads) including a centriole (open arrow). Lead citrate and uranyl acetate, original magnification, x 7000.

cytic hypertrophy and proliferation were observed. Astrocytosis apparently correlated with spongiform change but not with neuronal loss. In the hippocampus, astrocytic changes were seen in both the pyramidal cell layer and the granular cell layer of fascia dentata. Astrocytic hyperplasia was evident and different stages of mitosis were recognized (Fig. 7.11). Many astrocytes were similar to “naked nuclei” of Alzheimer II cells (Fig. 7.12). Others contained lobulated and bizarre nuclei more reminiscent of those which characterize hypertrophic reactive astrocytes or astrocytes encountered in multifocal leukoencephalopathy. The presence of glial fibers and Rosenthal fibers, regarded as products of gliofilament condensation and degeneration, were frequently noted. Proliferation of astrocytes was accompanied by the presence of rod-like microglial cells. To clarify the proliferative potential of astrocytes in TSEs, we have studied immunohistochemically the immunoreactivity of proliferating cell nuclear antigen (PCNA) δ which is active in DNA leading-strand (Fig. 7.13),94 an auxillary protein of polymerase-δ synthesis, (an established marker for cell proliferation95-99 in experimental scrapie and CJD and in human cases of kuru, CJD and GSS). In CJD-infected mouse and scrapie-infected hamster brains, astrocytes expressing PCNA exhibited homogeneously stained, intensely black nuclei. Astrocyte PCNA-specific immunostaining was confined entirely to the cell nuclei. Faint cytoplasmic staining detected in many hypertrophic astrocytes was regarded as nonspecific in the absence of nuclear staining but proved useful for identification of cell morphology. During the early stages of experimental CJD with minimal spongiform change, PCNA-immunopositive nuclei were occasionally observed in the subependymal zone (PCNA labeling index, PCNA LI, 0% to 1.0%). From 18 weeks postinoculation, PCNAimmunopositive astrocytes were most frequent in the corpus callosum (PCNA LI, 0% to 3.6%) and cerebellar white matter (0% to 3.7%), regions which characteristically exhibit robust vacuolation. The gray matter lesions were practically devoid of PCNA-immunopositive

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

141

Fig. 7.12. “Naked nuclei” in scrapie-affected hamster brain. Original magnification, x 1000.

Fig. 7.13. PCNA-immunoreactive astrocytic nuclei (arrows) in CJD-affected mouse brains. PCNA immunohistochemistry. Original magnification, x 1000.

142

Astrocytes in Brain Aging and Neurodegeneration

astrocytes except in the deep cortical layers of the parietal cortex adjacent to the corpus callosum (PCNA LI, 0% to 4.5%). In the latter region, PCNA LIs even exceeded those of the corpus callosum. No other cells, particularly no ferritin-immunopositive cells with the morphology of ramified microglia,76,100 expressed PCNA, and no PCNA expression was observed in the brains of control animals. In CJD-affected mouse brains, PCNA LIs correlate significantly with the grade of astrocytosis in both deep layers of the cerebral cortex and the corpus callosum (r = 0.78 and 0.5; p<0.01 and p<0.05, respectively) but not in the subependymal zone or in the cerebellar white matter. The correlation of PCNA LI and incubation time (measured in weeks) was statistically significant only in the subependymal zone (r = 0.41; p<0.05) while the grade of astrocytosis correlates significantly with incubation period only in the deep layers of cerebral cortex and in the subependymal zone (r = 0.47 and 0.51; p<0.05 and p<0.01, respectively). In CJD-affected mice of all stages, the number of PCNA-immunopositive astrocytes was low, less than 5% of the visible population of astrocytes (the highest PCNA LI, 4.5%). By contrast, in brain tissues from human patients with kuru, CJD and GSS, in which abundant PrP-immunopostive plaques were seen,76,100 no PCNA-immunopositive cells were detected despite the presence of numerous microglial cells and reactive astrocytes, which were clearly identified on adjacent sections following immunostaining with antibodies against ferritin and glial fibrillary acidic protein (GFAP), respectively. Ultrastructural studies are in general agreement concerning the glial changes.101-103 Astrocytes did not show any features which discriminated them from reactive astrocytes found in a plethora of neurodegenerative disorders. A few “serial killing” experiments performed so far provide conflicting data on whether astrocytosis appears before or after vacuolation. Marsh and Kimberlin104 found hypertrophic astrocytes in scrapie-infected hamsters 9 weeks after intracerebral inoculation and preceding vacuolation by 2 weeks. This initial astrocytic hypertrophy was first observed at the piaarachnoid surfaces and adjacent to the ventricles. In contrast, Liberski and Alwasiak demonstrated that astrocytosis actually followed vacuolation in hamsters infected with the 263K strain of scrapie agent.105 Scrapie-specific vacuoles appeared 8 weeks postinoculation, while at that time astrocytosis unequivocally surpassed vacuolation. Masters et al106 found astrocytosis detectable at week 7 or 5 by means of routine neuropathological staining or indirect immunofluorescence. Unequivocal spongiform change appeared in this model 7-8 weeks after inoculation. While the spongiform change stabilized in intensity at week 9-10 postinoculation, the number of astrocytes increased steadily until the clinical phase of disease and thus paralleled the progressive increase in the infectivity titers. This correlation may implicate astrocytes as a target for the replicating agent rather than merely constituting passively reactive cells. Unfortunately, both experimental studies suffered from the obvious weakness of the use of “poorly vacuolated” models. Thus, the problem of whether astrocytosis is merely a reaction toward the destruction of neuronal elements, or whether astrocytes undergo primary proliferation and hypertrophy, cannot be settled. Scrapie passaged to cattle produce prominent but moderate astrocytosis and only minimal or no spongiform change in numerous brain structures.107-108 The latter discriminate scrapie in cattle from BSE in this species. Astrocytes frequently appeared in clusters and the topographic distribution of astrocytosis was reminiscent of sheep scrapies, i.e., septal nuclei were prominently affected while the other forebrain structures were not. The other affected structures included the thalamus, midbrain tegmentum (especially the periaqueductal gray matter), pontine nuclei, and the nucleus of the solitary tract. In the cerebellum, Bergmann glia increased in number. The moderate degree of astrocytosis was confirmed by GFAPimmunohistochemistry. Analogous findings were reported for cattle infected with TME.109

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

143

BSE and CWD The data on astrocytic reaction in bovine spongiform encephalopathy (BSE) and chronic wasting disease (CWD) are very limited. In the first report of BSE by Wells et al,110 only mild gliosis was noted in BSE-affected cattle brains and this finding has been subsequently confirmed.111-112 Liberski et al113-114 found that numerous hypertrophic astrocytes, not infrequently binucleated ones containing abundant glial filaments, accompanied the neuronal degeneration. Jeffrey et al82 reported on astrocytosis in BSE-infected mice, but as the astrocytic response is highly variable among different experimental models, this study has little relevance to the problem of astrocytosis in natural BSE. In a captive puma (Felis concolor) infected with BSE, both astro- and microgliosis was readily apparent; the latter formed typical microglial nodules.115 Analogously, in chronic wasting disease (CWD) in mule deer, hybrids of mule deer and white tailed deer and Rocky Mountain elk, numerous hypertrophic astrocytes were noted.116-120

Interaction Between Astrocytes and Oligodendrocytes Interactions between astrocytes and oligodendrocytes121 have been previously reported only in early lesions of multiple sclerosis (MS) and a few other conditions.122-123 Its presence in both naturally occurring and experimentally induced TSE suggests an early cellular event that may trigger further tissue destruction (see section: Astrocytes and expression of cytokines, this chapter). In the brain biopsy of a patient with CJD,124 low-power electron microscopy revealed numerous examples of astrocytes and oligodendroglial cells in close apposition to one another; cellular membranes of one cell type were often molded on those of another (Fig. 7.14). Occasionally two oligodendroglial cells were seen in close contact with the same astrocyte. At higher magnification, both types of cells were connected by rare adhesive plaque junctions (Fig. 7.15). These subcellular organelles were composed of two symmetric or asymmetric subplasmalemmal densities (attachment plaques) collectively forming “attenuated desmosomes” or “desmosome-like” structures. In both the 263K and 22C-H hamster models, similar phenomena were observed. A narrow intercellular space between these attachment plaques was visible, containing one or two intermediate lines. More complex structures were also seen in both hamster models. Astrocytic cytoplasm was penetrated by a few oligodendroglial processes, or oligodendroglial cells were completely surrounded by astrocytic processes which formed multilayered onion-like “collars” around the former. Such interactions were previously reported in early lesions of MS122 and at that time they were regarded as unusual and possibly specific for this demyelinating process. In a subsequent detailed study, however, Wu and Raine123 showed that such interactions, while frequently encountered in MS lesions, are nonspecific, being observed in other neurological disorders including Krabbe’s disease, Toxoplasma encephalitis and brain infarcts. The common denominator in all these processes is the presence of inflammatory lymphocytic infiltrates, which tend to be minimal or totally absent in TSEs.125,126 It is of further interest that astrocytes and oligodendrocytes show weak electric coupling in vitro, which has been interpreted as evidence that these cells are physically connected.127,128 The significance of the interactions between astrocytes and oligodendrocytes is unclear at the present time. As in MS and other brain lesions in which it has been studied, this interaction is not associated with a response to any infectious pathogen. Rather, it may be an event that triggers brain tissue destruction mediated by pro-inflammatory cytokines secreted from astrocytes, lymphocytes and macrophages.

144

Astrocytes in Brain Aging and Neurodegeneration

Fig. 7.14. A general view of complex structure formed by astrocytes (circles) and oligodendrocytes (squares). Lead citrate and uranyl acetate, original magnification, x 12,000.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

145

Fig. 7.15. An astrocyte (star) and oligodendrocyte (square) connected by well-formed adhesive plaque junction (arrows). Lead citrate and uranyl acetate, original magnification, x 50,000.

A Particular Form of Astrocytic Reaction in TSES The majority of TSEs are polioencephalopathies (diseases of the gray matter) and corresponding fine structural changes are relatively well described. However, the panencephalopathic type of CJD characterized by the predominant involvement of the white matter has also been reported129 and axonal and myelin pathology at the ultrastructural level has been described.55,130-132 Myelinated axons presented various pathological lesions. While these changes were observed simultaneously in different areas of the same sample, the following description is organized as if it followed a sequence of events. Initially, the myelin sheath was separated by cytoplasmic tongues into several concentric bands. Cellular processes penetrated between layers of myelin and lifted away the outermost lamellae. Then a complicated labyrinth of concentric cellular processes, clearly belonging to either astrocytes or macrophages, invested myelinated axons (Fig. 7.16). In terminal stages, axons completely denuded of myelin were seen in the center of a concentric network of cellular processes, or spirals of myelin were seen surrounded by such processes. The myelin fragments penetrated into astrocytes or macrophages, where they underwent final digestion (Fig. 7.17). Macrophages containing fragments of axons, paracrystalline lamellar bodies and myelin debris were abundant in this model.

Expression of Glial Fibrillary Acidic Protein (GFAP) and Its MRNA First isolated from mature multiple sclerosis plaques,133-134 glial fibrillary acidic protein (GFAP) is a major protein component of glial filaments, a class of intermediate filaments specific for astrocytes. On polyacrylamide gel electrophoresis (PAGE), GFAP generally appears as a 49 kDa protein accompanied by proteolytic cleavage products down to a

146

Astrocytes in Brain Aging and Neurodegeneration

Fig. 7.16. The panencephalopathic type of CJD. A myelinated fiber (star) invested by astrocytic process (arrowheads). Lead citrate and uranyl acetate, original magnification, x 12,000.

molecular weight of 40 kDa. It is regarded as a useful marker for normal, hypertrophic and neoplastic astrocytes. Mackenzie studied Compton mice infected with the “Chandler” (139A) strain of scrapie by means of GFAP-immunohistochemistry.135 The use of GFAP as an astrocytic marker proved to be extremely useful, particularly for quantitative estimation of astrocytosis. The latter had previously been complicated by the insensitivity of routine H & E staining90,92 and the capriciousness of the Cajal metal impregnation technique.88 Abundant GFAP-immunopositive astrocytes were seen in corpus callosum, hippocampus, cerebellum and spinal cord. This location was disease-specific, as a different pattern of GFAP immunoreactivity was observed in mice infected with Semliki forest virus or mice intoxicated with cuprizone. Furthermore, GFAP immunoreactive astrocytes were readily detected in scrapie-affected sheep. It is noteworthy that there was no correlation between clinical signs of scrapie and GFAP immunoreactivity in the brain stem, nor between the distribution of spongiform change and GFAP immunoreactivity. The overproduction of GFAP was recently confirmed in mice infected with the C506 strain of scrapie virus and in scrapie-affected sheep.136-139 Furthermore, GFAP mRNA paralleled the GFAP increase in natural scrapie of sheep.138 Astrocytic reactions characterized by robust GFAP immunostaining were termed “hypergliotic”139 when they were regarded as being out of proportion to the degree of neuronal damage. Furthermore, the regional distribution of GFAP-immunoreactive astrocytes paralleled that of PrP.140-142 GFAP concentrations, measured in homogenates of whole scrapieaffected hamster brain, were increased 20 to 30 days following intracerebral inoculation.140 The initial rise was slow and accelerated some 60 days postinoculation when the first signs of clinical scrapie were also observed. PrP 27-30 was first detectable approximately 45 days postinoculation; thus, the accelerated increase of GFAP concentration clearly followed that of PrP. A similar rise of PrP followed by that of GFAP was observed in selected brain re-

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

147

gions. For instance, in the thalamus, GFAP concentrations were increased at 50 to 55 days postinoculation and were preceded by increased concentrations of PrP 33-35sc (at 5-10 days postincubation). It thus seemed that PrP induces reactive gliosis. To test this hypothesis directly, the influence of PrP on astrocytic growth in vitro has been studied.140 Primary astrocytic cultures (more than 90% astrocytes) exhibited a significant increase in cell numbers after 3 days of PrP exposure. Furthermore, a dramatic increase in GFAP-immunopositive glial filaments was observed following PrP supplementation of the culture medium. Such an increase in GFAP concentration was paralleled by an increase in GFAP mRNA. In conclusion, PrP was found to be a potent stimulant for astrocytes. It has been suggested that PrP released into extracellular spaces induces reactive astrocytic gliosis. PrPsc has also been localized to astrocytes.143 In serial experiments, PrPsc was detected in astrocytes 8 weeks following inoculation, then increased to the stage where it was detectable diffusely throughout the neuropil. The accumulation of PrPsc within astrocytes preceded both astrocytosis, which was first observed 12 weeks following inoculation and the appearance of scrapie amyloid 16 weeks postinoculation. From the above discussion it would appear that PrP, GFAP, and astrocytosis are functionally related and that PrP is a growth factor for astrocytes. Molecular studies of GFAP have had an interesting twist. Weitgrefe et al144 constructed a cDNA from purified poly(A+)RNA from scrapie-infected mouse brain. For differential hybridization, this cDNA library was screened by [32P]-labeled cDNA reverse-transcribed from poly(A+)RNA of scrapie-infected and control brains. One clone (Scr-1) hybridized preferentially to scrapie-infected brains. However, in dot-blot experiments, Scr-1 was shown to also hybridize to control material, although the extent was 20-fold less. On Northern blots, Scr-1 hybridized to the 3.3 kb RNA species. In in situ hybridization experiments, Scr-1 was located to neurons, mostly in scrapie-affected brains, and to dystrophic neurites within neuritic plaques in human brains with Alzheimer’s disease and rare senile plaques of multiinfarct dementia brains.145 While the significance of the Scr-1 gene was unknown at the time of its discovery, it has subsequently been established that the Scr-1 clone merely represented the 3' noncoding region of Gfap.145,146 The Scr-1 cDNA sequence is 98% homologous to the 3' untranslated region of the mouse Gfap cDNA. Indeed, Scr-1 was further used as a probe to examine the expression of GFAP mRNA in CJD-infected hamsters.146 The mRNA for GFAP was studied in regions that show no spongiform change and compared with those exhibiting severe vacuolation. An increased amount of GFAP mRNA was found in the cerebral cortex toward the end stage of disease, and its increase preceded the appearance of spongiform change. However, in some cerebral areas which had prominent vacuolation, its increase was not readily apparent. Conversely, a large increase of GFAP mRNA was noticed in the cerebellum, in which spongiform changes were absent. Analogous data were provided for scrapie-infected newborn mice.147 Andreas-Barquin et al138 found upregulation at the protein and mRNA level of both GFAP and glutamine synthetase (GS). The latter finding may suggest that the traffic of glutamate and glutamine is distorted in scrapie, but it was not confirmed in a subsequent study. GFAP is not, however, necessary for scrapie infection.148-149 Mice in which the first exon of the Gfap gene was disrupted by replacing it with lacZ gene (Gfap–/–) are susceptible to scrapie infection, develop typical pathology (including astrocytosis), and exhibit the same level and distribution (by histoblots) of PrPsc. Mutated astrocytes showed subtle differences in immunostaining with antibodies against vimentin and S-100 protein. Both vimentin and S-100 protein signals tend to be granular and limited in the perinuclear space, as opposed to wild type astrocytes where both signals are rather filamentous and fill the whole cytoplasm. Indeed, in a recently described Brazilian family with spongiform encephalopathy and a novel mutation at codon 183 of the PRNP gene, gliosis was minimal or even absent throughout the brain.149a

Astrocytes in Brain Aging and Neurodegeneration

148

a Fig. 7.17a.

Moreover, it was demonstrated that not only GFAP but several biologically active substances localized to astrocytes (methallothionein, crystallins, apolipoprotein E, cathepsin D and various lymphokines; see below) are upregulated in scrapie150-156 and some of these compounds are also upregulated in another neurodegenerative disorder, Alzheimer’s disease, suggesting a convergent pathological mechanism.145,147,148 Finally, tissue factor, the tissue activator of the coagulation protease cascade, was also upregulated in astrocytes of scrapieinfected brains.157 In experimental scrapie, the upregulation of astrocytic enzymes precedes the development of neuropathological lesions but follows the rise of PrP.148 Apolipoprotein E4, which is a risk factor for Alzheimer’s disease,158 may increase when astrocytes assume

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

149

b Fig. 7.17. (a, opposite) General view of brain lesions of the panencephalopathic type of CJD. Note several macrophages (stars) and an astrocyte (circle) digesting a myelinated fiber (arrow); (b, above) remnants of myelinated fiber (arrow) within the cytoplasm of an astrocyte. Lead citrate and uranyl acetate, original magnification: (a) x 7000; (b) x 12,000.

the role of macrophages, as demonstrated in experimental CJD,55,130-132 and crystallin, being a heat shock protein, may participate in the early response to CNS damage.151-152 Collectively, these data suggest that astrocytes induced by PrP may play a substantially more important role in the pathogenesis of TSE than merely reacting passively to other brain tissue lesions.

Astrocytes and the Expression of Cytokines In the brain, a potential role of cytokines in the pathogenesis of neurological disorders was first considered in multiple sclerosis (MS) and related demyelinating disorders.159 In contrast to TSEs, the neuropathology of MS is characterized by focal perivascular lymphocytic infiltrates and macrophages associated with areas of demyelination, disappearance of oligodendroglial cells and proliferation of astrocytes. Similar features are observed in experimental allergic encephalomyelitis (EAE)—a disorder induced by active immunization with myelin basic protein (MBP), which may be passively transferred by T cells.160 Numerous data indicate a pivotal role for tumor necrosis factor-α (TNF-α) in the pathogenesis of EAE,161-166 and several studies suggest a similar pathomechanism for MS.167-174 Myelin ballooning accompanied by oligodendrocyte degeneration and astrocytic hypertrophy has been produced in mouse spinal cord cultures treated with recombinant human TNF-α,175 and cytotoxic activity of TNF-α and lymphotoxin on oligodendroglial cells was reported. Similar findings in EAE have prompted the hypothesis that TNF-α, a cytokine released from activated microglia/macrophages176 and astrocytes,177,178 is directly involved in myelin breakdown in several demyelinating disorders, presumably by interacting with sodium channels on the axolemma.179 This hypothesis has been further substantiated by immunohistochemical detection of TNF-α expression in astrocytes in brain tissues of patients with MS168,174 as well as by blocking of passive transfer of EAE by anti-TNF-α neutralizing antibodies164 or by the potent TNF-α inhibitor pentoxyfilline.180 Thus, at least in demyelinating disorders, TNF-α

150

Astrocytes in Brain Aging and Neurodegeneration

may play a crucial role in the activation of immune cells. Although immunocyte activation has not been directly implicated in the pathogenesis of the TSEs,181 the data accruing from the aforementioned MS and EAE studies raise the possibility that TNF-α and other proinflammatory cytokines may play important roles in the development of the former conditions, in particular those panencephalopathic types of TSE with severe white matter involvement.

Tumor Necrosis Factor-α (TNF-α) TNF-α was first described in 1975 when Carswell and colleagues182 found, that in serum of mice infected with bacillus Calmette-Guerin (BCG) and treated with endotoxin (lipopolysaccharide W Escherichia coli), there was an additional toxin capable of inducing necrosis of certain neoplasms. This toxin is also cytostatic or cytotoxic to a variety of human cell lines in vitro.183 Discovered earlier, cachectin (a toxin causing cachexia by blocking lipoprotein lipase) was shown to be identical to TNF-α.184,185 In contrast, lymphotoxin—a biologically active compound released by T cells in immunized rats and causing cytolysis of mice L929 fibroblasts—is a different molecule designated TNF-β. In the central nervous system (CNS), TNF-α is produced by both microglial cells and astrocytes.178,186 Reactive astrocytes also produce prostaglandins and IL-1.187 These cells may also express class II major histocompatibility complex (MHC) antigens following exposure to interferon-γ (IFN-γ), probably by increasing expression of a receptor for TNF-α.188,189 Astrocytes are postulated to act as antigen presenting cells in the CNS, presenting antigens to T cell active clones, in an MHC-restricted fashion, upon expression of class II molecules.190 TNF-α damages oligodendroglial cells in vitro and is a mitogen for astrocytes but not for oligodendrocytes.191,192 TNF-α is merely the violin section of a large “orchestra” of cytokine molecules which may act synergically or in opposition to one another. TNF-α induces production of IL-1, IL-6, TGF-α and other cytokines, and may thus activate multiple cell types in the course of an inflammatory reaction.193 In addition, IL-1 may induce production of TNF-α194 and further amplify the reaction. These cytokines are produced within the CNS by both astrocytes and microglial cells150 Transgenic mice with IL-6 overexpression show hippocampal lesions similar to those observed in scrapie,195 although IL-6 was not shown to be overexpressed in this disease.150 Intracerebral inoculation with IL-1α or IF-γ causes extensive astrocytosis.196-198 Expression of these cytokines may increase after activation of IF-γ released from lymphocytes172 but other cytokines, such as IL-10 or TGF-β, may act antagonistically and block secretion of TNF-α and IL-1α.199 An example of synergistically acting cytokines are TNF-α and IL-1α, which induce the reciprocal release of cytokines.150,194,200,201 IL-1α, one of the first cytokines described (lymphocyte activating factor, endogenous pyrogen, mitogenic protein), exhibits multiple biological activities, with the activation of T cells being the most important.202 Other activities of IL-1α include activation of natural killer (NK) cells, induction of IL-2 receptor expression, induction of B cell proliferation and induction of colony stimulating factor (G-CSF, M-CSF, GM-CSF) secretion.203-205 Moreover, IL-1 may induce TNF-α production.194 There are two forms of IL-1: IL-1α and β which, although products of two distinct genes, are characterized by a similar molecular weight of 17 kDa. Both gene products independently bind to two distinct classes of IL-1 receptor (I and II). A number of cell types produce IL-1: monocytes and macrophages, Langerhan’s and dendritic cells, NK cells, endothelial cells and, within the CNS, astrocytes and microglial cells.203 IL-1α is mitogenic for astrocytes, but is similar to TNF-α in causing a decrease in the number of oligodendrocytes.170 TNF-α may be responsible for some of the neuronal cell loss which occurs in CJD, scrapie and BSE. TNF-α may play a role in the induction of apoptosis.206 In experimental

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

151

Fig. 7.18. Large intramyelin vacuole (star) and numerous astrocytic processes (squares) within an optic nerve following intraocular injection of recombinant TNF-α. Lead citrate and uranyl acetate, original magnification, x 7000

CJD, in scrapie207 and in experimental BSE,82 neuronal autophagic vacuoles, similar to those observed in cells undergoing apoptosis, were observed. Although Selmaj et al208 failed to demonstrate DNA fragmentation typical of apoptosis in cultured oligodendroglial cells treated with TNF-α, DNA fragmentation was found after application of lymphotoxin.

TNF-α in TSEs The potential role of TNF-α in the development of neuropathological changes has been studied in the TSEs.209 Recombinant murine TNF-α injected into the vitreous of the mouse eye produced myelin ballooning in the optic nerve (Fig. 7.18).210 This myelin dilatation was ultrastructurally indistinguishable from that observed in the panencephalopathic type of CJD. Furthermore, TNF-α immunoreactivity in astrocytes of scrapie- and CJD-infected mouse brain has been shown.209,211 Campbell et al150 in their sequential study of experimental scrapie demonstrated that the mRNA of TNF-α, IL-1α and IL-1β are overexpressed in scrapie-affected brains. On the other hand the expression of brain TNF-α mRNA and IL-4, IL-5, IFN-γ, IL-2, IL-6 and IL-3 mRNAs were not altered during scrapie infection. The expression of TNF-α, IL-1α, IL-1β, and other cytokine mRNAs in the kidneys, spleen, and liver were not altered by scrapie infection, indicating that overexpression of these cytokines is brain-specific. Detailed time course experiments showed that significant increase of TNF-α, IL-1α and IL-1β mRNAs occurred by week 15 postinoculation and increased progressively until end-stage disease at week 25. This study showed not only that there is pronounced activation of cerebral TNF-α, IL-1α and IL-1β gene product expression in scrapie but also that the increased expression of

152

Astrocytes in Brain Aging and Neurodegeneration

Fig. 7.19. RT-PCR products for TNF-α, IL-1α, GFAP, and β-actin. Lanes 1-11, controls, 2 weeks apart from 2nd to 22nd week; lanes 12-33, CJD-infected animals, one week apart from 1st to 22nd week after inoculation; Lane 23, 100 bp ladder.

these cytokines’ mRNA correlates well with the progression of the clinical disease and molecular neuropathological changes. This correlation is suggestive of a causal relationship, but this remains unproven. Similar results were obtained in another experimental TSE—the panencephalopathic type of CJD.211 Here, the brain tissues from CJD virus-infected mice were examined at 1 week intervals postinoculation for TNF-α and IL-1α transcript expression using reverse transcriptase-directed polymerase chain reaction (RT-PCR). TNF-α expression was also examined by Western and Northern blot analyses and by immunocytochemistry. Total RNA samples from control brains were diluted to minimize banding intensity and this dilution

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

153

Fig. 7.20. Agarose gel with RT-PCR products for TNF-α, IL-1α, GFAP, PrP, and β-actin. For controls (lanes 1, 3, 5, 7, 9) and CJD-infected animals, week 22 postinoculation (lanes 2, 4, 6, 8, 10). Lane 11, 100 bp ladder.

(0.4 mg/ml) was used throughout the study (comparative PCR). Similar to the results obtained by Campbell and colleagues,150 low intensity bands for TNF-α increased significantly after 15 week postinoculation, being unchanged in the brains of control animals (Figs. 7.19 and 7.20). Analogous results were obtained for IL-1α. In contrast, bands for PrP and β-actin derived from brains of infected animals were similar to those of controls and did not show any correlation with disease progression. These results were subsequently confirmed by Western and Northern blot analysis and by immunohistochemistry, implicating hypertrophic astrocytes as TNF-α-secreting cells. Collectively, the aforementioned studies demonstrate that TNF-α and IL-1α are upregulated in CJD-affected brains and suggest that these cytokines may serve as molecular mediators of white matter degeneration in experimental CJD. On the other hand, the overexpression of TNF-α in diseases as diverse as CJD, AIDS vacuolar myelopathy and MS may suggest that pro-inflammatory lymphokines may merely act as final end-stage mediators of axon and myelin damage, irrespective of its cause. Accumulation of PrP may initiate a cascade of events inducing the release of TNF-α, IL-1α and β which, in turn, amplify other cytokine responses, produce secondary astrocytosis and microglial infiltration, and culminate in oligodendroglial injury and white matter damage. As mentioned above, overproduction of these cytokines correlates with the progression of the clinical features and molecular neuropathological changes, but although this correlation is suggestive of a causal relationship, this still remains to be proven.

Conclusions At present, it is very difficult, or even impossible, to answer the question, “What is the role of astrocytes in TSEs?” In general, astrocytes undergo proliferative and hypertrophic

154

Astrocytes in Brain Aging and Neurodegeneration

changes in the course of TSEs. Also, and this is particularly true for the natural diseases, the presence and extent of astrocyte hypertrophy are highly variable. In some cases there is a focal astrocytosis, in others the astrocytic reaction is diffuse, and occasionally it is nonexistent. The same appears to be true for numerous models of scrapie in mice and animal models of CJD, GSS and BSE. Astrocytes participate in PrP plaque formation, but probably only its later stages. In an exceptional model of the panencephalopathic type of CJD, astrocytes take part in myelin stripping and, finally, in myelin digestion. Astrocytes secrete cytokines, in particular TNF-α, which may amplify further tissue damage. In all these activities, they do not seem to differ from astrocytes in other brain lesions; the reader may find further examples in any textbook of brain pathology, including this volume. In the TSEs, astrocytes do not appear to be the primary targets of the infectious agent. Their responses are predominantly reactive in nature, although they may participate in the perpetuation of tissue injury.

Acknowledgments The research performed by PPL and RD is supported by the Maria Sklodowska-Curie Foundation in Poland and by Fogarty International Center and the Kosciuszko Foundation in the USA. It is a part of European Community Concerted Action “Biomed 1 and 2”— “Prion diseases: From neuropathology to pathobiology and molecular genetics” awarded to Professor Herbert Budka, Vienna, Austria.

References 1. Gajdusek DC. Unconventional viruses and the origin and disappearance of kuru. Science 1977; 197:943-960. 2. Gajdusek DC. Transmissible amyloidoses of the brain. In: Liberski PP, ed.Light and Electron Microscopic Neuropathology of Slow Virus Disorders. Boca Raton: CRC Press, 1993:1-31. 3. Liberski PP. The enigma of slow viruses: Facts and artifacts. Arch Virol 1993 (suppl 6):1-267. 4. Liberski PP. Prions, β-sheets and transmissible dementias: Is there still something missing? Acta Neuropathol (Berl) 1995; 90:113-125. 5. Kretzschmar HA. Human prion diseases (spongiform encephalopathies). Arch Virol [suppl] 1993; 7:261-293. 6. Schoon HA, Brunckhorst D, Pohlenz J. Spongiform encephalopathie beim Rothalsstrauss (Struthio camelus). Ein kasuistischer Beitrag. Therarztl Prax 1991; 19:263-265. 7. Will RG, Ironside JW, Zeidler M et al. A new variant of Creutzfeldt-Jakob disease in UK. Lancet 1996; 347:921-925. 8. Lasmezas CI, Deslys JP, Demaimay R et al. BSE transmission to macaques. Nature 1996; 381:743-744. 9. Collinge J, Sidle KCL, Meads J et al. Molecular analysis of prion strain variation and the aetiology of “new variant” CJD. Nature 1996; 383:685-688. 10. Gajdusek DC. Infectious amyloids: Subacute spongiform encephalopathies as transmissible cerebral amyloidoses. In: Fields BN, Knipe DM, Howley PM et al, eds, Fields Virology, 3rd ed. New York: Lippincott-Raven Publishers, 1997:2851-2900. 11. Prusiner SB. Prions. In: Fields BN, Knipe DM, Howley PM et al, eds. Fields Virology, 3rd ed. New York: Lippincott-Raven Publishers, 1997:2901-2950. 12. Prusiner SB. Novel proteinaceous infection particles cause scrapie. Science 1982; 216:136-144. 13. Collinge J, Owen F, Poulter M et al. Prion dementia without characteristic pathology. Lancet 1990; 336:7-9. 14. Brown P, Kaur P, Sulima M et al. Real and imagined clinicopathological limits of “prion dementia”. Lancet 193; 341:127-129.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

155

15. Kocisko DA, Come JH, Priola SA et al. Cell-free formation of protease-resistant prion protein. Nature 1994; 370:471-4. 16. Lansbury PT, Caughey B. The chemistry of scrapie infection: Implications of the “ice 9” metaphor. Chem Biol 1995; 2:1-5. 17. Budka H, Aguzzi A, Brown P et al. Neuropathological diagnostic criteria for CreutzfeldtJakob disease (CJD) and other human spongiform encephalopathies. Brain Pathol 1995; 4:459-466. 18. Brown P, Goldfarb LG, Gajdusek DC. The new biology of spongiform encephalopathy: Infectious amyloidoses with a genetic twist. Lancet 1991; 337:1019-1022. 19. Kimberlin RH. Scrapie and possible relationships with viroids. Sem Virol 1990; 1:153-162. 20. Diringer H. Hidden amyloidoses. Exp Clin Immunogenet 1993; 9:212-229. 21. Weissmann C. Sheep disease in human clothing. Nature 1989; 338:298-299. 22. Weissmann C. A “unified theory” of prion propagation. Nature 1991; 352:679-683. 23. Liberski PP, Budka H. An overview of neuropathology of the slow unconventional virus infections. In: Liberski PP, ed. Light and Electron Microscopic Neuropathology of Slow Virus Disorders. Boca Raton: CRC Press, 1992:111-151. 24. Liberski PP, Guiroy DC, Yanagihara R et al. Astrocytic changes. In: Liberski PP, ed. Light and Electron Microscopic Neuropathology of Slow Virus Disorders. Boca Raton: CRC Press, 1992:187-232. 25. Liberski PP, Nerurkar V, Yanagihara R et al. Immunohistochemistry of astrocytic reaction. In: Liberski PP, ed. Light and Electron Microscopic Neuropathology of Slow Virus Disorders. Boca Raton: CRC Press, 1992:233-250. 26. Petersen RB, Tabaton M, Chen SG et al. Familial progressive subcortical gliosis: Presence of prions and linkage to chromosome 17. Neurology 1995; 45:1062-1067. 27. Klatzo I, Gajusek DC. Pathology of kuru. Lab Invest 1959; 8:799-847. 28. Klatzo I, Gajusek DC, Zigas V. Evaluation of pathological findings in twelve cases of kuru. In: Van Boagert L, Radermecker J, Hozay J, Lowenthal A, eds. Encephalitides. Amsterdam: Elsevier Publ Comp, 1959:172-190. 29. Hainfellner JA, Liberski PP, Guiroy DC et al. Pathology and immunocytochemistry of a kuru brain. Brain Pathol 1997; 7:547-553. 30. Neuman MA, Gajdusek DC, Zigas V. Neuropathologic findings in exotic neurologic disorder among natives of the highlands of New Guinea. J Neuropathol Exp Neurol 1964; 23:486-507. 31. Scrimgeour EM, Masters CL, Alpers MP et al. A clinico-pathological study of a case of kuru. J Neurol Sci 1983; 59:265-275. 32. Beck E, Daniel PM, Asher DM et al. Experimental kuru in chimpanzee: A neuropathological study. Brain 1973; 96:31-38. 33. Beck E, Daniel PM, Alpers M et al. Experimental “kuru” in chimpanzees. A pathological report. Lancet 1966; 2:1056-1059. 34. Lampert PW, Earle KM, Gibbs CJ Jr et al. Experimental kuru encephalopathy in chimpanzees and spider monkeys. Electron microscopic studies. J Neuropathol Exp Neurol 1969; 28:353-370. 35. Marin O, Vial JD. Neuropathological and ultrastructural findings in two cases of subacute spongiform encephalopathy. Acta Neuropathol (Berl) 1964; 4:218-229. 36. Gonatas NK, Terry RD, Weiss M. Electron microscopic study in two cases of JakobCreutzfeldt disease. J Neuropathol Exp Neurol 1965; 24:575-598. 37. Torack RM. Ultrastructure and histochemical studies in a cse of progressive dementia and its relationship to protein metabolism. Am J Pathol 1966; 49:77-97. 38. Torack RM. Ultrastructural and histochemical studies of cortical biopsies in subacute dementia. Acta Neuropathol (Berl) 1969; 13:43-55. 39. Brion S. La efermedad de Creutzfeldt-Jakob. Neurol Neurocir Psiq (Mexico) 1968; 9:103-115. 40. Bubis JJ, Goldhammer Y, Braham J. Subacute spongiform encephalopathy. Electron microscopic studies. J Neurol Neurosurg Psychiatr 1965; 35:881-887.

156

Astrocytes in Brain Aging and Neurodegeneration

41. Ribadeau-Dumas JL, Escourolle R. The Creutzfeldt-Jakob syndrome. A neuropathological and electron-microscopic study. In: Subirana A, Espadaler JM, Burrows EH, eds. Neurology. Proc of the Xth Int Cong Neurol. Amsterdam: Excerpta Medica, 1974:315-329. 42. Ribadeau-Dumas JL, Escourolle R, Castaigne P. Syndrome de Creutzfeldt-Jakob: etude ultrastructurale de trois observations. Rev Neurolog (Paris) 1969; 121:405-422. 43. Payne CM, Sibley WA. Intranuclear inclusions in a case of Creutzfeldt-Jakob disease. Acta Neuropathol (Berl) 1975; 31:353-361. 44. Bouteille M, Kalifat S, Delarue J. Ultrastructural variations of nuclear bodies in human diseases J Ultrastruct Res 1967; 19:474-486. 45. Liberski PP, Papierz W, Alwasiak J et al. Diagnostic difficulties in a case of subacute sclerosing encephalitis. Light and electron microscopic studies. Neuropatol Polska 1988; 26:98-110. 46. Jellinger K. Nuclear inclusions in subacute spongiform encephalopathy. Acta Neuropathol (Berl) 1971; 17:283-286. 47. Lafarga M, Berciano MT, Suarez I et al. Cytology and organization of reactive astroglia in human cerebellar cortex with severe loss of granule cells: A study on the ataxic form of Creutzfeldt-Jakob disease. Neuroscience 1991; 40:337-352. 48. Lafarga M, Berciano MT, Suarez I et al. Reactive astroglia-neuron relationships in the human cerebellar cortex: A quantitative, morphological and immunohistochemical study in Creutzfeldt-Jakob disease. Int J Devl Neuroscience 1993; 11:199-213. 49. Dubois-Dalcq M, Rodriguez M, Reese TS et al. Search for a specific marker in the neural membranes of scrapie mice. Lab Invest 1977; 36:547-553. 50. Beck E, Daniel PM, Matthews WB et al. Creutzfeldt-Jakob disease. The neuropathology of a transmission experiment. Brain 1969; 92:699-716. 51. Kim JH, Manuelidis EE. Serial ultrastructural study of experimental Creutzfeldt-Jakob disease in guinea pigs. Acta Neuropathol (Berl) 1986; 69:81-90. 52. Kim JH, Manuelidis EE. Ultrastructural findings in experimental Creutzfeldt-Jakob disease in guinea pigs. J Neuropathol Exp Neurol 1983; 42:29-43. 53. Manuelidis EE, Manuelidis L. Clinical and morphological aspects of of transmissible Creutzfeldt-Jakob disease. In: Zimmerman HM, ed.Progress in Neuropathology. Vol 4. New York: Raven Press, 1979:1-26. 54. Manuelidis EE, Manuelidis L. Observations on Creutzfeldt-Jakob disease propagated in small rodents. In: Prusiner SB, Hadlow WJ, eds.Slow Transmissible Diseases of the Nervous System. Vol 2. New York: Academic Press, 1979:147-172. 55. Liberski PP, Yanagihara R, Asher DM et al. Reevaluation of the ultrastructural pathology of experimental Creutzfeldt-Jakob disease. Brain 1990; 113:121-137. 56. Hsiao K, Scott M, Foster D et al. Spontaneous neurodegeneration in transgenic mice with mutant prion protein. Science 1990; 250:1587-1590. 57. Hsiao K, Baker HF, Crow TJ et al. Linkage of a prion protein missense variant to Gerstmann-Straussler syndrome. Nature 1989; 338:342-345. 58. Hsiao K, Doh-Ura K, Kitamoto T et al. A prion protein amino acid substitution in ataxic Gerstmann-Straussler syndrome. Ann Neurol 1989; 26:106. 59. Hudson AJ, Farrell MA, Kalnins R et al. Gerstmann-Straussler-Scheinker disease with coincidental familial onset. Ann Neurol 1983; 14:670-678. 60. Kuzuhara S, Kanazawa I, Sasaki H et al.Gerstmann-Straussler-Scheinker’s disease. Ann Neurol 1983; 14:216-225. 61. Vinters HV, Hudson AJ, Kaufmann JCE. Gerstmann-Straussler-Scheinker disease: Autopsy study of a familial case. Ann Neurol 1986; 20:540-543. 62. Tateishi J, Kitamoto T, Tashima T et al. Gerstmann-Straussler-Scheinker’s disease: Dissimilarities in pathology and transmission. In: Court LA, Dormont D, Brown P, Kingsbury DT, eds.Unconventional Virus Diseases of the Central Nervous System. Paris: Commisariat a’ l’Energie Atomique, 1989:151-157. 63. Ghetti B, Tagliavini F, Masters CL et al. Gerstmann-Straussler-Scheinker disease. II. Neurofibrillary tangles and plaques coexist in an affected family. Neurology 1989; 39:1453-1461.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

157

64. Nochlin D, Sumi SM, Bird TD et al. Familial dementia with PrP positive amyloid plaques: A variant of Gerstmann-Straussler syndrome. Neurology 1989; 39:910-918. 65. Pearlman RL, Towfighi J, Pezeshkpour GH et al. Clinical significance of types of cerebellar amyloid plaques in human spongiform encephalopathies. Neurology 1988; 38:1249-1254. 66. Liberski PP, Budka H. Ultrastructural pathology of Gerstmann-Straussler-Scheinker disease. Ultrastr Pathol 1995; 19:23-36. 67. Hainfellner J, Brantner-Inhaler S, Cervenakova L et al. The original Gerstmann-StrausslerScheinker family of Austria: Divergent clinicopathological phenotypes but constant PrP genotype. Brain Pathol 1995; 5:201-213. 68. Liberski PP. Reappraisal of ultrastructural diversity of amyloid plaques in GerstmannStraussler-Scheinker syndrome. Pat Polska 1995; 46:33-41. 69. Jeffrey M, Goodbrand IA, Goodsir A. Pathology of the transmissible spongiform encephalopathies with special emphasis on ultrastructure. Micron 1995; 26:277-298. 70. Watanabe R, Duchen L. Cerebellar amyloid in human prion diseases. Neuropathol Appl neurobiol 1993; 19:253-260. 71. Boellaard JW, Schlote W. Subakute spongiforme Encephalopathie mit multiformer Plaquebildung. Eigenartige familiar-hereditare Kranknheit des Zentralnervensystems [spinocerebellare Atrophie mit Demenz, Plaques and plaqueahnlichen im Klein- and Grosshirn (Gerstmann, Straussler, Scheinker)]. Acta Neuropathol (Berl) 1980; 49:205-212. 72. Boellaard JW, Schlote W. Glial plaques: Amyloid deposits characteristic of slow transmissible encephalopathies. Virchows Arch [Cell pathol] 1981; 37:337-341. 73. Boellaard JW, Schlote W, Heldt N. The development of amyloid plaques in human transmissible encephalopathy and the role of astrocytes in their formation. In: Court LA, Dormont D, Brown P, Kingsbury DT, eds. Unconventional Virus Diseases of the Central Nervous System. Paris: Commissariat a’ l’Energie Atomique, 1989:162-171. 74. Wisniewski HM, Terry RD. Re-examination of the pathogenesis of the senile plaques. In: Zimmerman HM, ed.Progress in Neuropathology, Vol. 2. New York: Grune & Stratton, 1973:1-26. 75. Miyazono M, Iwaki T, Kitamoto T et al. A comparative immunohistochemical study of kuru and senile plaques with a special reference to glial reactions at various stages of amyloid plaque formation. Am J Pathol 1991; 589-598. 76. Barcikowska M, Liberski PP, Boellaard Jet al. Microglia is a component of the prion protein amyloid plaque in the Gerstmann-Straussler-Scheinker syndrome. Acta Neuropathol (Berl.) 1993; 85:623-627. 77. Jeffrey M, Goodsir CM, Bruce ME et al. Infection specific prion protein (PrP) accumulates on neuronal plasmalemma in scrapie infected mice. Neurosci Lett 1992; 147:106-109. 78. Jeffrey M, Goodsir C, Bruce ME et al. Morphogenesis of amyloid plaque in 87V murine scrapie. Neuropathol Appl Neurobiol 1994; 20:535-542. 79. Jeffrey M, Goodsir CM, Bruce ME et al. Correlative light and electron microscopic studies of PrP localization in 87V mice. Brain Res 1994; 656:329-343. 80. Jeffrey M, Goodsir CM, Fowler N et al. Ultrastructural immuno-localization of synthetic prion protein peptide antibodies in 87V murine scrapie. Neurodegeneration 1996; 5:101-109. 81. Beck E, Daniel PM, Parry HB. Degeneration of the cerebellar and hypothalamo-neurophysial systems in sheep with scrapie; and its relationship to human system degenerations. Brain 1964; 87:153-176. 82. Jeffrey M, Scott JR, Williams A et al. Ultrastructural features of spongiform encephalopathy transmitted to mice from three species of bovidae. Acta Neuropathol (Berl) 1992; 84:559-569. 83. Field EJ. The significance of astroglial hypertrophy in scrapie, kuru, multiple sclerosis and old age, together with a note on the possible nature of the scrapie agent. Deutsch Z Nervenheil 1967; 192:265-274. 84. Hadlow WJ. The pathology of experimental scrapie in dairy goats. Res Vet Sci 1961; 2:289-315. 85. Hadlow WJ, Kennedy RC, Race RE et al. Virologic and histologic findings in dairy goats affected with natural scrapie. Vet Pathol 1980; 17:187-199.

158

Astrocytes in Brain Aging and Neurodegeneration

86. Georgsson G, Gisladottir E, Arnadottir S. Quantitative assessment of the astrocytic response in natural scrapie of sheep. J Comp Pathol 1994; 108:229-240. 87. Pattison IH, Jones KM. The astrocytic reaction in experimental scrapie in the rat. Res Vet Sci 1967; 8:160-165. 88. Fraser H. Neuropathology of scrapie: The precision of the lesions and their significance. In: Prusiner SB, Hadlow WJ, eds. Slow Transmissible Diseases of the Nervous System. Vol. 1. New York: Academic Press, 1979:387-406. 89. Scott JR, Fraser H. Degenerative hippocampal pathology in mice infected with scrapie. Acta Neuropathol (Berl) 1984; 65:62-68. 90. Liberski PP. Astrocytic reaction in experimental scrapie. J Comp Pathol 1987; 97:73-78. 91. Liberski PP. The brain fine structure in experimental scrapie. The 263K strain in golden syrian hamsters. Neuropatol Polska 1987; 25:35-51. 92. Liberski PP. Gliocytosis in experimental scrapie (263K strain of scrapie) in golden syrian hamsters. Neuropatol. Polska 1986; 24:221-230. 93. Liberski PP, Asher DM, Yanagihara R et al. Serial ultrastructural studies of scrapie in hamsters. J Comp Pathol 1989; 101:429-442. 94. Biernat W, Liberski PP, Guiroy DC et al. Proliferating cell nuclear antigen immunohistochemistry in astrocytes in experimental Creutzfeldt-Jakob disease and in human kuru, Creutzfeldt-Jakob disease and Gerstmann-Straussler-Scheinker syndrome. Neurodegeneration 1995; 4:195-201. 95. Bravo R, Frank R, Blundell PA et al. Cyclin/PCNA is the auxillary protein of DNA polymerase-δ. Nature 1987; 326:515-517. 96. Bravo R, Macdonald-Bravo H. Changes in the nuclear distribution of cyclin (PCNA) but not its synthesis depend on DNA replication. EMBO J 1985; 4:655-661. 97. Bravo R, Macdonald-Bravo H. Existence of two populations of cyclin/proliferating cell nuclear antigen during the cell cycle: association with DNA replication sites. J Cell Biol 1987; 105:1549-1554. 98. Matthews MB, Benbstein RM, Franza BR et al. Identity of the proliferating cell nuclear antigen. Nature 1984; 309:374-376. 99. Prelich G, Cheng-Keat T, Kostura M et al. Functional identity of proliferating cell nuclear antigen and a DNA polymerase-δ auxillary protein. Nature 1987; 326:517-520. 100. Guiroy DC, Wakayama I, Liberski PP et al. Relationship of microglia and scrapie amyloidimmunoreactive plaques in kuru, Creutzfeldt-Jakob disease and Gerstmann-Straussler syndrome. Acta Neuropathol. (Berl) 1994; 87:526-530. 101. Chandler RL Electron microscopic and other observations on the ependyma and small cerebral blood vessels in mice and rats infected with scrapie. Res Vet Sci 1967; 8:166-169. 102. Chandler RL Ultrastructural pathology of scrapie in the mouse: An electron microscopic study of spinal cord and cerebellar areas. Br J Exp Pathol 1968; 49:52-59. 103. Chandler RL Ultrastructural observations on scrapie in the gerbil. Res Vet Sci 1968; 15:322-328. 104. Marsh RF, Kimberlin RH. Comparison of scrapie and transmissible mink encephalopathy in hamsters. II. Clinical signs, pathology and pathogenesis. J Infect Dis 1975; 131:104-110. 105. Liberski PP, Alwasiak J. Neuropathology of experimental transmissible spongiform encephalopathy. The 263K strain of scrapie in golden syrian hamsters. Neuropatol Polska 1983; 21:378-392. 106. Masters CL, Rohwer RG, Franko MC et al. The sequential development of spongiform change and gliosis of scrapie in the golden syrian hamsters. J Neuropathol Exp Neurol 1984; 43:242-252. 107. Clark WW, Hourigan JL, Hadlow WJ. Encephalopathy in cattle experimentally infected with the scrapie agent. Am J vet Res 1995; 56:606-612. 108. Cutlip RC, Miller JM, Race RE et al. Intracerebral transmission of scrapie to cattle. J Inf Dis 1994; 169:814-820. 109. Robinson MM, Hadlow WJ, Knowles DP et al. Experimental infection of cattle with the agents of transmissible mink encephalopathy and scrapie. J Comp Pathol 1995; 113:241-251.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

159

110. Wells GAH, Scott AC, Johnson CT et al. A novel progressive spongiform encephalopathy in cattle. Vet Rec 1987; 121:419-420. 111. Wells GAH, Wilesmith JW, McGill IS. Bovine spongiform encephalopathy. Brain Pathol 1991; 1:69-78. 112. Wells GAH, Wilesmith JW. The neuropathology and epidemiology of bovine spongiform encephalopathy. Brain Pathol 1995; 5:91-103. 113. Liberski PP. Ultrastructural neuropathologic features of bovine spongiform encephalopathy. In: Bolis L, Gibbs CJ Jr. Proceedings of an International Roundtable on Bovine Spongiform Encephalopathy. J AmVet Med Assoc 1990; 196:1682. 114. Liberski PP, Yanagihara R, Wells GAH et al.. Comparative ultrastructural studies of bovine spongiform encephalopathy, scrapie and Creutzfeldt-Jakob disease. J Comp Pathol 1992; 106:361-38. 115. Willoughby K, Kelly DF, Lyon DG et al. Spongiform encephalopathy in a captive puma (Felis concolor). Vet Rec 1992; 131:431-434. 116. Williams ES, Young S. Chronic wasting disease of captive mule deer: A spongiform encephalopathy. J Wildlife Dis 1980; 16:89-98. 117. Williams ES, Young S. Spongiform encephalopathy of Rocky Mountain elk. J Wildlife Dis 1982; 18:465-471. 118. Williams ES, Young S. Spongiform encepalopathies in Cervidae. Review of Scientific Technology Office and International Epizootiology 1992; 11:551-567. 119. Guiroy DC, Williams ES, Liberski PP et al. Ultrastructural pathology of chronic wasting disease in captive mule deer. Acta Neuropathol. (Berl) 1993; 85:437-444. 120. Guiroy DC, Liberski PP, Williams ES et al. Electron microscopic findings in brain of Rocky Mountain elk with chronic wasting disease. Folia Neuropathol. (Warsaw) 1994; 32:171-173. 121. Liberski PP, Brown P, Cervenakova L, Gajdusek DC. Interactions between astrocytes and oligodendroglia in human and experimental Creutzfeldt-Jakob disease and scrapie. Exp Neurol 1997; 144:227-234. 122. Prineas JW, Kwon EE, Goldenberg PZ et al. Interaction of astrocytes and newly formed oligodendrocytes in resolving multiple sclerosis lesions. Lab Invest 1990; 63:624-636. 123. Wu E, Raine CS. Interactions between oligodendrocytes and hypertrophic astrocytes and their occurrence in other, nondemyelinating conditions. Lab Invest 1992; 67:88-99. 124. Liberski PP, Sklodowski P, Klimek A. Creutzfeldt-Jakob disease: Ultrastructural study of brain biopsy: Unusual interaction between astrocytes and oligo- and microglia. Folia Neuropathol (Warsaw) 1995; 33:85-92. 125. Museteanu C, Diringer H. Perivascular infiltrates of leukocytes in brains of scrapie-infected mice. Nature 1991; 294:360-361. 126. Williams AE, Ryder S, Blakemore WF. Monocyte recruitment into the scrapie-affected brain. Acta Neuropathol (Berl) 1995; 90:164-169. 127. Ransom BR, Kettemann H Electrical coupling, without dye coupling, between mammalian astrocytes and oligodendrocytes in cell culture. Glia 1990; 3:258-266. 128. Couraud P-O. Interactions between lymphocytes, macrophages, and central nervous system. J Leukocyte Biol 1994; 56:407-415. 129. Tateishi J, Ohta M, Koga M, Sato Y, Kuroiwa Y. Transmission of chronic spongiform encephalopathy with kuru plaques from humans to rodents. Ann Neurol 1978; 24:35-40. 130. Liberski PP, Yanagihara R, Gibbs CJ Jr, Gajdusek DC. White matter ultrastructural pathology of experimental Creutzfeldt-Jakob disease in mice. Acta Neuropathol (Berl.) 1989b; 79:1-9. 131. Liberski PP, Yanagihara R, Gibbs CJ Jr, Gajdusek DC. Mechanism of the damage to myelinated axons in experimental Creutzfeldt-Jakob disease in mice: An ultrastructural study. Eur J Epidemiol 1991; 7:545-550. 132. Liberski PP, Gajdusek DC. Myelinated axon undergoes complete demyelination in the panencephalopathic but it is merely subjected to the Wallerian degeneration in the polioencephalopathic-type of transmissible spongiform encephalopathies. Polish J Pathol 1997; 48:163-171.

160

Astrocytes in Brain Aging and Neurodegeneration

133. Eng LF, Vanderhaegen JJ, Bignami A et al. An acidic protein isolated from fibrous astrocytes. Brain Res 1971; 28:351-354. 134. Eng LE, DeArmond SJ. Immunochemistry of the glial fibrillary acidic protein. In: Zimmerman HM, ed. Progress in Neuropathology. Vol 5. New York: Raven Press, 1983:19-39. 135. Mackenzie A. Immunohistochemical demonstration of glial fibrillary acidic protein in scrapie. J Comp Pathol 1983; 93:251-259. 136. Lefrancois T, Tardy M. L’astrocyte, une cellule cle’ dans la pathogenie des encephalopathies spongiformes. Rec Med Vet 1995; 171:29-32. 137. Lefrancois T, Fages C, Brugere-Picoux et al. Astroglial reactivity in natural scrapie of sheep. Microbial Pathogenesis 1994; 17:283-289. 138. Andreas-Barquin PJ, le Prince G, Fages C et al. Expression of glial fibrillary acidic protein and glutamine synthetase genes in the natural scrapie of sheep. Mole Chem Neuropathol 1994; 22:57-65. 139. Dormont D, Delpech B, Delpech A et al. Hyperproduction de proteine gliofibrillaire acide (GFA) au cors de l’evolution de la tremblante experimentale de la Souris. C R Acad Sci Paris 1981; 293:53-56. 140. DeArmond SJ, Gonzales M, Mobley WC et al. PrPsc in scrapie-infected hamster brain is spatially and temporally related to histopathology and infectivity titer. In: Iqbal K, Wisniewski HM, Winblad B, eds. Alzheimer’s Disease and Related Disorders. New York: Alan R. Liss, Inc., 1989:601-618. 141. DeArmond SJ, Kretzschmar HA, McKInley MP et al. Molecular pathology of prion disease. In: Prusiner SB, McKinley MP, eds. Prions. Novel Infectious Pathogens causing Scrapie and Creutzfeldt-Jakob Disease. New York: Academic Press, 1987:387-414. 142. DeArmond SJ, Mobley WC, DeMott DL et al. Changes in the localization of brain prion proteins during scrapie infection. Neurology 1987; 37:1721-1280. 143. Brown HR, Goller NL, Rudelli RD et al. The mRNA encoding the scrapie agent protein is present in a variety of non-neuronal cells. Acta Neuropathol (Berl) 1987; 80:1-6. 144. Weitgrefe S, Zupanic M, Haase A et al. Cloning of a gene whose expression is increased in scrapie and in senile plaques in human brain. Science 1985; 230:1177-1179. 145. Diedrich J, Wietgrafe S, Zupancic M et al. The molecular pathogenesis in scrapie and Alzheimer’s disease. Microb Pathogen 1987; 2:435-442. 146. Manuelidis L, Tesin DM, Sklaviadis T et al. Astrocyte gene expression in Creutzfeldt-Jakob disease. Proc Natl Acad Sci USA 1987; 84:5937-5941. 147. Lazarini F, Deslys JP, Dormont D. Variations in prion protein and glial fibrillary acidic protein mRNAs in the brain of scrapie-infected newborn mouse. J Gen Virol 1992; 73:1645-1648. 148. Gomi H, Yokoyama T, Fujimoto K et al. Mice devoid of the glial fibrillary acidic protein develop normally and are susceptible to scrapie prions. Cell 1995; 14:29-41. 149. Tatzelt J, Maeda N, Pekny M et al. Scrapie in mice deficient in apolipoprotein E or glial fibrillary acidic protein. Neurology 1996; 47:449-453. 149a.Nitrini R, Rosemberg S, Passos-Bueo MR et al. Familial spongiform encephalopathy with distinct clinico-pathological features associated with a novel prion protein gene mutation at codon 183. Ann Neurol 1997; 42:138-146. 150. Campbell IL, Eddleston M, Kemper P. Activation of cerebral cytokine gene expression and its correlation with onset of reactive astrocyte and acute-phase response gene expression in scrapie. J Virol 1994; 68:2383-2387. 151. Diedrich J, Duguid JR, Haase AT. The role of astrocytes in the neuropathology of scrapie and Alzheimer’s disease. Sem Virol 1991; 2:233-238. 152. Diedrich J, Minnigan H, Carp RI. Neuropathological changes in scrapie and Alzheimer’s disease are associated with increased expression of apolipoprotein E and cathepsin D in astrocytes. J Virol 1991; 65:4759-4768. 153. Duguid JR, Bohmont CW, Liu N et al. Changes in gene expression shared by scrapie and Alzheimer disease. Proc Natl Acad Sci USA 1987; 86:7260-7264.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

161

154. Williams AE, van Dam AM, Man-A-Hing WK et al. Cytokines, prostaglandins and lipocortin-1 are present in the brains of scrapie-infected mice. Brain Res 1994; 654:200-206. 155. Renkawek K, de Jong WW, Merck KB et al. A,b-crystallin is present in reactive glia in Creutzfeldt-Jakob disease. Acta Neuropathol (berl) 1992; 83:324-327. 156. Kenward N, Hope J, Landon M et al. Expression of polyubiquitin and heat-shock protein 70 genes increases in the latter stages of disease progression in scrapie-infected mouse brain. J Neurochem 1994; 62:1870-1877. 157. Eddleston M, de la Torre JC, Oldstone MBA et al. Astrocytes are the primary source of tissue factor in the murine central nervous system. A role of astrocytes in cerebral hemostasis. J Clin Invest 1993; 92:349-358. 158. Roses AD, Einstein G, Gilbert J et al. Morphological, biochemical and genetic support for an apolipoprotein E effect on microtubular metabolism. Ann New York Acad Sci 1996; 777:146-57. 159. Brosnan CF, Selmaj K, Raine CS. Hypothesis: A role of tumor necrosis factor in immunemediated demyelination and its relevance to multiple sclerosis. J Neuroimmunol 1988; 18: 87-94. 160. Ben-nun A, Wekerle H, Cohen I. The rapid isolation of clonable antigen specific lines capable of mediating autoimmune encephalomyelitis. Eur J Immunol 1981; 11:195-199. 161. Chung IY, Norris J G, Benveniste EN. Differential TNF expression by astrocytes from EAEsusceptible and -resistant rat strains. J Exp Med 1991; 173:801-811. 162. Martin D, Near S L, Bendele A et al. Inhibition of tumor necrosis factor is protective against neurologic dysfunction after active immunization of Lewis rats with myelin basic protein. Exp Neurol 1995; 131:221-228. 163. Powell M B, Mitchell D, Lederman J et al. Lymphotoxin and TNF production by myelin basic protein-specific T cell clones correlates with encephalitogenicity. Int Immunol 1990; 2:539-544. 164. Selmaj KW, Raine CS, Cross A H. Anti-tumor necrosis factor therapy abrogates autoimmune demyelination. Ann Neurol 1991; 30:694-700. 165. Selmaj K, Papierz W, Glabinski A et al. Prevention of chronic relapsing experimental autoimmune encephalomyelitis by soluble tumor necrosis factor receptor I. J Neuroimmunol 1995; 56:135-141. 166. Selmaj KW, Raine CS. Experimental autoimmune encephalomyelitis: immunotherapy with anti-tumor necrosis factor antibodies and soluble tumor necrosis factor receptors. Neurology 1995; 45, Suppl. 6: S 44-49. 167. Benvenuto R, Paroli M, Buttinelli C et al. TNF synthesis by CSF-derived T cell clones from patients with multiple sclerosis. Clin Exp Immunol 1991; 84:97-102. 168. Hoffman FM, Hinton DR, Johnson K et al. TNF-α identified in multiple sclerosis brain. J Exp Med 1989; 170, 607-612. 169. Glabiñski A, Mirecka M, Pokoca L. Tumor necrosis factor α but not lymphotoxin is overproduced by blood mononuclear cells in multiple sclerosis. Acta Neurol Scand 1995; 91:276-279. 170. Merrill JE. Proinflamatory and antiinflamatory cytokines in multiple sclerosis and central nervous system acquired immunodeficiency syndrome. J Immunother 1992; 12:167-170. 171. Merrill JE, Strom SR, Ellison GW et al. In vitro study of mediators of inflammation in multiple sclerosis. J Clin Immunol 1989; 9:84-96. 172. Rieckmann P, Albrecht M, Ehrenreich H, Weber T et al. Semi-quantitative analysis of cytokine gene expression in blood and cerebrospinal fluid cells by reverse transcriptase polymerase chain reaction. Res Exp MedBerl 1995; 195:17-29. 173. Rieckmann P, Albrecht M, Kitze B et al. Tumor necrosis factor α messenger RNA expression in patients with relapsing-remitting multiple sclerosis is associated with disease activity. Ann Neurol 1995; 37:82-88. 174. Selmaj KW, Raine CS, Cannella B et al. Identification of lymphotoxin and tumor necrosis factor in multiple sclerosis lesions. J Clin Invest 1991; 87:949-954. 175. Selmaj K W, Raine CS. Tumor necrosis factor mediates myelin and oligodendrocyte damage in vitro. Ann Neurol 1988; 23:339-346.

162

Astrocytes in Brain Aging and Neurodegeneration

176. Hetier E, AyalaJ, BousseauA et al. Ameboid microglial cells and not astrocytes synthesize TNF-α in Swiss mouse brain cell cultures. Eur J Neurosci 1990; 2:762-768. 177. Chung IY, Benveniste EN. TNF-α production by astrocytes. Induction by polysaccharide, interferon-α and interleukin-1α. J Immunol 1990; 144:2999-3007. 178. Robbins D.S, Shirazi Y, Drysdale B-E et al. Production of cytotoxic factor for oligodendrocytes by stimulated astrocytes. J Immunol 1987; 139:2593-2597. 179. Barna BP, Estes ML, Jacobs BS et al. Human astrocytes proliferate in response to tumor necrosis factor-α. J Neuroimmunol 1990; 30:239-243. 180. Nataf S, Louboutin JP, Feve JR et al. Pentoxifylline inhibits experimental allergic encephalomyelitis. Acta Neurol Scandinav 1993; 88:97-99. 181. Brown P. The phantasmagoric immunology of transmissible spongiform encephalopathy. In: Waksman BH, ed.Immunologic Mechanisms in Neurologic and Psychiatric Disease. New York: Raven Press, Ltd., 1990:305-313. 182. Carswell EA, Old LJ, Kassel RL et al. An endotoxin-induced serum factor that causes necrosis of tumors. Proc Natl Acad Sci USA 1975; 72:3666-3669. 183. Sugarman BJ, Aggarwal BB, Hass PE et al. Recombinant human tumor necrosis factoralpha: Effects on proliferation of normal and transformed cells in vitro. Science 1985; 230:943-945. 184. Kawakami M, Pekala PH, Lane MD et al. Lipoprotein lipase suppression in 3T3-L1 cells by an endotoxin-induced mediator from exudate cells. Proc Natl Acad Sci 1982; 79:912-915. 185. Beutler B, Cerami A. Cachectin and tumor necrosis factor as two sides of the same biological coin. Nature 1986; 320:584-588. 186. Lieberman AP, Pitha PM, Shin HS et al. Production of TNF-α and other cytokines by astrocytes stimulated with lipopolysaccharide or a neurotropic virus. Proc Natl Acad Sci USA 1989; 86:6348-6352. 187. Fontana A, Kristensen F, Dubs R et al. Production of prostaglandin E and an interleukin1 like factor by cultured astrocytes and C6 glioma cells. J Immunol 1982; 129:2413-2419. 188. Fierz W, Endler B, Reske K. Astrocytes as antigen presenting cells. I. Induction of Ia antigen expression on astrocytes by T cells via immune interferon and its effect on antigen presentation. J Immunol 1985; 134:3785-3793. 189. Benveniste EN, Sparacio SM, Bethea JR. Tumor necrosis factor-α enhances interferon-γmediated class II antigen expression on astrocytes. J Neuroimmunol 1989; 25:209-219. 190. Takiguchi M, Frelinger JA. Induction of antigen presentation ability in purified cultures of astroglia by interferon-γ. J Mol Cell Immunol 1986; 2:269-280. 191. Lachman LB, Brown DC, Dinarello CA. Growth promoting effects of recombinant IL-1 and TNF for a human astrocytoma cell line. J Immunol 1987; 138:2913-2916. 192. Selmaj K, Shafit-Zagardo B, Aquino DA et al. Tumor necrosis factor-induced proliferation of astrocytes from mature brain is associated with downregulation of glial fibrillary acidic protein mRNA. J Neurochem 1991; 57:823-830. 193. Philip R, Epstein LB. TNF-α as immunomodulator and mediator of monocyte cytotoxicity induced by itself, gamma interferon and interleukin 1. Nature 1986; 323:86-89. 194. Campbell IL, Abraham CR, Mazliah E et al. Neurologic disease induced in transgenic mice by cerebral overexpression of IL-6. Proc Natl Acad Sci USA 1993; 90:10061-10065. 195. Giulian D, Woodward J, Young DG et al. Interleukin-1 injected into mammalian brain stimulates astrogliosis and neovascularisation. J Neurosci 1988, 8:2485-2490. 196. Giulian D, Lachman LB. Interleukin-1 stimulation of astroglial proliferation after brain injury. Science 1985, 228:497-499. 197. Yong VW, Moumdjian R, Yong FP et al. Gamma-interferon promotes proliferation of adult human astrocytes in vitro and reactive gliosis in the adult mouse brain in vivo. Proc Natl Acad Sci USA 1991; 88:7016-7020. 198. Olsson T. Cytokine-producing cells in experimental autoimmune encephalomyelitis and multiple sclerosis. Neurology 1995; 45:11-15. 199. Le J, Vilcek J. Tumor necrosis factor and interleukin 1: Cytokines with multiple overlapping biological activities. Lab Invest 1987; 56:234-248.

Astrocytes in Transmissible Spongiform Encephalopathies (Prion Diseases)

163

200. Dinarello CA, Cannon JG, Wolff SM et al. Tumor necrosis factor (cachectin) is an endogenous pyrogen and induces production of interleukin 1. J Exp Med 1986; 163:1433-1450. 201. Dinarello CA. Biology of interleukin-1. FASEB J 1988; 2:108-115. 202. Dinarello CA. Biologic basis for interleukin in disease. Blood 1996; 87:2095-147. 203. Lai Ch-F, Baumann H. Interleukin-1β induces production of granulocyte colony-stimulating factor in human hepatoma cells. Blood 1996; 87:4143-4148. 204. Lee M, Segal GM, Bagby GC. Interleukin-1 induces human bone marrow-derived fibroblasts to produce multilineage hematopoietic growth factors. Exp Hematol 1987; 15:983-988. 205. Jeffrey M, Fraser JR, Halliday WG et al. Early unsuspected neuron and axon terminal loss in scrapie-infected mice revealed by morphometry and immunocytochemistry. Neuropathol Appl Neurobiol 1995; 21:41-49. 206. Robaye B, Mosselmans R, Fiers W et al. Tumor necrosis factor induces apoptosis (programmed cell death) in normal endothelial cells in vivo. Am J Pathol 1991; 138:447-453. 207. Liberski PP, Yanagihara R, Gibbs CJ Jr, Gajdusek DC. Neuronal autophagic vacuoles in experimental scrapie and Creutzfeldt-Jakob disease. Acta Neuropathol (Berl) 1992; 83:134-139. 208. Selmaj K, Raine CS, Faroq M et al. Cytokine cytotoxicity against oligodendrocytes. Apoptosis induced by lymphotoxin. J Immunol 1991; 147:1522-1529. 209. Liberski PP, Nerurkar VR, Yanagihara R et al. Tumor necrosis factor: Cytokine-mediated myelin vacuolation in experimental Creutzfeldt-Jakob disease. Abstract no P 68-15 in Abstracts of the VIIIth International Congress of Virology, Berlin, West Germany, August 26-31, 1990, 421. 210. Liberski PP, Yanagihara R, Nerurkar VR et al. Tumor necrosis factor produces CJD-like lesions in vivo. Neurodegeneration 1933; 2:215-225. 211. Kordek R, Nerurkar VR, Liberski PP et al. Heightened expression of tumor necrosis-α, interleukin 1α, and glial fibrillary acidic protein in experimental Creutzfeldt-Jakob disease in mice. Proc Natl Acad Sci USA 1996; 93:9754-9758.

CHAPTER 8

Astrocytes in Other Neurodegenerative Diseases Dennis W. Dickson

Introduction

G

lial pathology is increasingly recognized in several neurodegenerative diseases. The relationship of the glial changes to neurodegeneration is uncertain, but the discovery of glial inclusion bodies in select neurodegenerative diseases suggests that glial dysfunction may contribute to disease pathogenesis. While it has not been specifically studied in the disorders under consideration, basic studies have provided evidence for cross talk between glia and neurons with production of mutual trophic factors and their receptors. It is thus possible that glial pathology may contribute to, or be a direct consequence of, neurodegeneration rather than a curious epiphenomenon. It would indeed be groundbreaking if some of the disorders that are currently considered to be neurodegenerative diseases were in fact due to primary abnormalities in glia—i.e., gliodegenerative diseases. Only further research into fundamental biology of glia and their interactions with neurons will produce answers to these questions. All neurodegenerative diseases are associated with reactive gliosis that usually is topographically coincident with neuronal degeneration and loss. Gliosis in this setting is undoubtedly an important pathologic finding, and in the case of spongiform encephalopathies, such as Creutzfeldt-Jakob disease, a cardinal histopathologic feature of the disease. On the other hand, gliosis in the setting of neurodegeneration offers few clues to disease pathogenesis since there is no way to know if it is anything other than a reactive or secondary change due to the processes that lead to neuronal degeneration and loss. Increasingly, the genetic bases for many of the disorders are being discovered, but the pathogenesis of even the most common of the disorders, namely Alzheimer’s disease, is unresolved and the focus of many current investigations. Particular among neurodegenerative disorders are those in which glial cytoplasmic inclusions are composed of filamentous aggregates; these disorders will be the focus of the present discussion. Other types of glial inclusions, such as Rosenthal fibers and autofluorescent inclusions that are seen in tumors, storage diseases and aging, are discussed in detail in other chapters of this book. A recent review covers many of these same topics.1 By convention, when we speak of neurodegenerative disorders we mean those disorders associated with progressive and selective loss of neurons, whose etiology in most cases remains uncertain. The particular subset of neurons that is vulnerable to cell loss defines the condition. For example, in idiopathic Parkinson’s disease, selective loss of neurons in the pars compacta of the substantia nigra in the midbrain leads to depletion of dopaminergic nerve Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

166

Astrocytes in Brain Aging and Neurodegeneration

terminals in the basal ganglia, which manifests as a characteristic movement disorder. Inclusion bodies composed of neurofilament in the cytoplasm of vulnerable neurons, which are referred to as Lewy bodies,2 define the pathology of idiopathic Parkinson’s disease, but glial inclusions are not widely recognized in Lewy body disease. Only recently has a report appeared that Gallyas-positive glial inclusions can be detected in some cases of idiopathic Parkinson’s disease.3 The inclusions were within astrocytes, since they were colabeled with antibodies to glial fibrillary acidic protein, but they were negative for tau and tubulin. No such inclusions were observed in Alzheimer’s disease or aged controls. Further studies are needed to confirm these observations and to determine if these argyrophilic astrocytic inclusions are specific to Lewy body disease. Evidence suggests that most of the filamentous aggregates within glia in many neurodegenerative diseases are derived from cytoskeletal elements, in particular, the microtubule-associated protein tau. Tau protein is also the major structural protein of neurofibrillary tangles (NFT), one of the major histopathological hallmarks of Alzheimer’s disease.4 NFT are numerous and widespread in the brain in Alzheimer’s disease; they are restricted in distribution in normal aging.5,6 NFT or NFT-like neuronal inclusions can also be found in certain other neurodegenerative diseases. In the great majority of the latter conditions, inclusions are found within astrocytes or oligodendrocytes, or both, as well as in neurons. In Alzheimer’s disease, glial inclusions, though occasionally detected,7 are far less common. NFT and glial tau inclusions are argyrophilic, which means that they are intensely stained with silver impregnation methods, such as the Bielschowsky, Bodian and Gallyas stains. This property has been traced to the presence of highly charged molecules, in particular, highly phosphorylated proteins.8 The inclusion bodies are also variably stained with histochemical methods for amyloid, such as thioflavin-S. This presumably reflects a highly ordered secondary structure that permits intercalation of the chromogens into the filaments. In contrast to NFT, which are intensely positive with amyloid stains, glial tau inclusions are far more difficult to detect with amyloid stains. These and other results discussed subsequently suggest that tau proteins in glial inclusions are not identical to tau proteins in NFT. Currently, the only means of directly studying the composition of glial inclusions in brain tissue is with descriptive methods, such as histochemical and immunohistochemical stains and electron microscopy. Methods to separately analyze tau protein in neurons and glia with direct brain tissue biochemical methods are currently not possible. Unfortunately, all published biochemical studies of these disorders do not distinguish the cell of origin for the protein in question. Studies of cytoskeletal proteins in cultured cells may provide a clue to the nature of proteinaceous inclusions in neurodegenerative diseases, but such results must be interpreted conservatively. The simple and defined environment in which cultured glial cells grow is far different from the complex and highly ordered environment of brain tissue. There is no guarantee that biochemical features observed for glia in vitro are comparable to those in vivo, especially in the diseased brain. Nevertheless, tau protein in neurodegenerative diseases and tau protein in cultured glial cells, especially oligodendroglia, have some intriguing similarities (see below). A great deal of effort has been devoted to studying the biochemical composition of NFT in Alzheimer’s disease, but much less is known about the biochemical basis of inclusions in other neurodegenerative disorders. One may infer by analogy that what is known about NFT may have some relevance to glial inclusions. It must, however, be acknowledged that even with all the advances in understanding the molecular biology of NFT, we do not know what induces NFT formation. Depletion of trophic factors, aberrant expression of developmental antigens and excitotoxicity are a few of the hypotheses that are under current investigation. There being no animal model for NFT, research is guided largely by de-

Astrocytes in Other Neurodegenerative Diseases

167

scriptive analyses. A brief discussion of NFT seems warranted as a background for understanding the nature of glial inclusions. Subsequently, glial inclusions will be discussed with reference to the different diagnostic entities in which they have been described. Common themes will be emphasized where possible.

Neurofibrillary Tangles as an Archetype of Cytoskeletal Inclusions NFT are composed of aggregates of filaments that appear to be composed of pairs of 10 nm-diameter filaments with a helical arrangement.9 Although recent fine structural studies with atomic force microscopy suggest that a more accurate model may be that of twisted ribbon,10 the filamentous inclusions that make up NFT are commonly referred to as paired helical filaments (PHF).9 PHF have a diameter of about 22 nm with cross-over points in the PHF at about every 80 nm. In disorders with tau-positive glial inclusions the filaments are usually straight, rather than twisted, ranging in diameter from 13 nm to 18 nm. In some disorders with neuronal and glial tau inclusions, twisted wider filaments (about 24 nm in diameter) are present with a distance between crossover points that is almost twice (160 nm) that of Alzheimer-type PHF.11,12 While precise cell type identification at the fine structural level is often difficult in autopsy tissue in the absence of double-immunolabeling with cell type-specific markers, most results suggest that the wider filaments with longer periodicity are neuronal rather than glial.12 Alzheimer-type PHF have not been described in glia. In addition to filaments, NFT invariably contain poorly characterized granular material. Comparable granular material is also present in glial inclusions, where it has been described as an integral component of the filaments (“granule-coated filaments”13). There are currently no clues as to the composition of the granular material. Given the great abundance of NFT in AD and the unusual solubility properties of PHF, PHF can be purified to near homogeneity and biochemically characterized.14,15 Biochemical and immunochemical studies have demonstrated that PHF are primarily composed of microtubule-associated protein tau.14-19 NFT are immunoreactive with antibodies to epitopes spanning the full length of tau,20 suggesting that full length tau is present in NFT. Similar studies of tau-positive glial inclusions also suggest that full length tau protein is present;21 however, there is suggestive evidence in some of the disorders with abundant glial inclusions that certain splice forms of tau may preferentially accumulate.12,22 Tau protein is a phosphoprotein with multiple isoforms derived from alternative splicing of a single gene on chromosome 1723 and also from posttranslational modification (reviewed in refs. 24 and 25). The best studied of the posttranslational modifications is phosphorylation. The phosphorylation state of tau determines its ability to promote polymerization of tubulin and to stabilize microtubules. More highly phosphorylated forms show decreased ability to promote polymerization of tubulin and to stabilize microtubules. Tau protein in PHF has increased and abnormal phosphorylation based on indirect immunochemical and immunocytochemical methods16 and direct chemical analysis of phosphate content.26 Hyperphosphorylated tau protein is incompetent with respect to tubulin assembly, and it is this form that accumulates in brains of neurodegenerative diseases.27 The current theory is that tau protein which has dissociated from tubulin due to increased phosphorylation undergoes self-assembly to form filaments.28 The tau protein in glial inclusions is also hyperphosphorylated based on immunostaining with antibodies to multiple phosphorylation sites in the tau molecule.21 Direct phosphate analysis of tau from glia in these disorders has not been reported. Whether phosphorylation of tau has any impact on microtubules in glia has not been investigated, but in neurons it is speculated that microtubule instability leads to impaired axoplasmic and dendritic transport and may thereby contribute directly to neurodegeneration.28 In contrast to neurons where microtubules are abundant in all cell domains

168

Astrocytes in Brain Aging and Neurodegeneration

(especially axons) and where they are crucial to essential cellular functions such as axoplasmic transport, the presence and function of microtubules in glia is far less clear. Microtubules are sparse in normal astrocytes,29 but more abundant in normal oligodendroglia.30 Furthermore, the major microtubule associated protein (MAP) in astrocytes is not tau, but rather a specific splice variant of MAP2.31-34 MAP2 expression is increased in astrocytes subjected to cell injury or stress.32,33 While a number of antibody based methods have indicated that inclusion bodies in astrocytes contain tau protein, direct proof for tau protein in astrocytes is lacking. All available evidence is based upon indirect immunocytochemical analyses. The fact that antibodies to multiple epitopes in tau stain the inclusions is fairly definitive evidence that tau is present in astrocytes, but not all studies have used more than one antibody to tau. It must be realized that microtubule-associated proteins share certain structural motifs and even conserved domains. For example, the microtubule-binding domain of MAP2 and tau share significant sequence homology.24 Additional studies are warranted to determine if any of the immunoreactivity observed in astrocytes in these disorders is related to crossreactivity with MAP2. Although tau was once felt to be restricted to neurons,35 more recent studies suggest that tau can be detected in normal glial cells; however, only trace amounts of tau can be detected at basal levels in normal human (unpublished data), bovine and rodent astrocytes.31,36,37 Even in cultured human fetal astrocytes exposed to activating conditions (e.g., interleukin-1β),38 or phosphatase inhibitors (e.g., okadaic acid),36 tau is difficult to detect (unpublished data). These results suggest that additional factors lead to production and aggregation of tau protein within astrocytes in neurodegenerative diseases or that adult astrocytes have properties that are distinct from fetal astrocytes. Tau is far easier to detect in oligodendrocytes, especially in brains that have been subjected to injury or stress, such as ischemia.40,41 It is also readily detected in cultured oligodendrocytes,36,37 where, interestingly, it appears to be composed of restricted tau isoforms, similar to isoform restriction that occurs in neurodegenerative diseases with glial tau inclusions.36 Specifically, exon 3 of tau is an alternatively spliced exon that contributes to the heterogeneity in tau isoforms.24 The function of this domain is not known, but tau in neurodegenerative diseases with abundant glial tau-positive inclusions have accumulation of tau splice forms that preferentially lack exon 3 based on lack of immunostaining with exon 3 specific antibodies.42, 43 The same is true for cultured oligodendrocytes.36 Native tau protein is a soluble protein that does not readily form filaments. Structural analysis indicates that it is an elongated molecule, but one that does not have a regular shape, which is an unusual feature for a soluble protein.25 There are multiple discrete domains in tau, with 3 or 4 conserved repeats in a domain involved in binding of tau protein to microtubules.25 Alternative RNA splicing generates tau proteins with either 3 or 4 repeats in the microtubule-binding domain.24 The microtubule-binding domain of tau appears to be essential for assembly into pathological filaments, since recombinant tau protein composed of little more than the microtubule-binding domain spontaneously assembles into PHFlike structures.44 Full length recombinant tau molecules do not spontaneously form filaments.45 This may indicate that proteolysis may play a role in producing tau fragments that spontaneously form filaments. On the other hand, recent studies have shown that full length recombinant tau protein forms filaments resembling the pathological filaments in brain diseases when it is mixed with acidic polymers such as heparan sulfate proteoglycan45 or lipids such as arachidonic acid. 46 Additional factors, such as protein crosslinking, ubiquitination,47 glycation,48-50 or association with polymers or other proteins, appear to be essential in the formation and aggregation of PHF (reviewed in ref. 51). Whether or not these processes are relevant to glial tau inclusions awaits further study.

Astrocytes in Other Neurodegenerative Diseases

169

Clearly, the phosphorylation state is one way in which pathological tau differs from normal tau. Given this observation, there has been much interest in identifying the kinases that may be responsible for catalyzing tau phosphorylation. Alternatively, increased phosphate content in tau may also be due to decreased activity or inhibition of specific phosphatases.52 It is of interest that freshly isolated tau proteins undergo rapid dephosphorylation,53 which suggests that phosphorylation of tau protein is under tight control. Both kinases and phosphatases are likely to be important in maintaining this tight control. On the other hand, experimental paradigms suggest that tau aggregates occur in the absence of abnormal phosphorylation and that phosphate-dependent epitopes may appear subsequently, contributing to stabilization of the aggregates.54 Regardless, phosphorylation remains an important phenotypic difference between normal tau and abnormal tau in most cytoskeletal inclusions. Among the several kinases that have been implicated in phosphorylation of tau protein (reviewed in ref. 55), proline-directed kinases have drawn a great deal of attention. This is because there are multiple such phosphorylation sites in tau and because phosphorylation at these sites affects the function of tau. Cyclin-dependent kinases are proline-directed kinases that can phosphorylate tau protein in vitro.56,57 While many different kinases can be shown to phosphorylate tau in a test tube, only the cyclin-dependent kinases have been shown to consistently colocalize with NFT and even to copurify with PHF.56-58 Cyclin-dependent kinases are members of the family of kinases that are involved in cell cycle regulation, and their expression in differentiated cells has been considered to be aberrant.57,59 Aberrant expression of such kinases has been implicated in programmed cell death or apoptosis,59 which is the mode of neuronal (and glial?) loss in most neurodegenerative diseases in which it has been studied. Of known cyclin-dependent kinases, one particular species that may be relevant to tau phosphorylation in glia is similar to cdc2 and has been referred to as KKIALRE.60 The latter terminology refers to an amino acid sequence in the carboxyl terminus of the kinase that is unique and distinguishes KKIALRE from authentic cdc2 kinase. Antibodies to the KKIALRE domain of this cdc2-like kinase label astrocytes in gray and white matter of human brain (Fig. 8.1). The expression of KKIARLE is increased in diseased brains in regions with gliosis, as seen in Alzheimer’s disease and other neurodegenerative diseases. Further studies are needed on the distribution of this and other kinases in neurodegenerative disorders associated with glial tau pathology.

Neurodegenerative Disorders with Filamentous Glial Inclusion Bodies The most common disorders with glial tau pathology (Table 8.1) include progressive supranuclear palsy (PSP), corticobasal degeneration (CBD) and Pick’s disease. Less frequently, tau-positive astrocytic inclusions have been detected in Guam Parkinson-dementia complex (GPDC), postencephalitic Parkinsonism and other neurodegenerative disorders with NFT. Rarely, tau-positive glial inclusions are found in AD. Tau-immunoreactive astrocytes are also detected in aged human brains, but in this context the immunoreactivity is not associated with inclusion bodies. Mixed and variable numbers of glial and neuronal tau-positive inclusions are seen in familial frontotemporal dementia with Parkinsonism linked to chromosome 17 (FTDP-17). Predominantly oligodendroglial inclusions are seen in dementia with argyrophilic grains (AGD). Inclusion bodies reminiscent of Lewy bodies have recently been described in astrocytes in some families with amyotrophic lateral sclerosis (FALS) and, interestingly, also in transgenic animals bearing the same mutation as FALS. Finally, glial inclusions have come to be the defining histopathologic hallmark of multiple system atrophy (MSA). These inclusions are found in oligodendroglia rather than astrocytes, and differ

Astrocytes in Brain Aging and Neurodegeneration

170

Fig. 8.1. Astrocytes in cortical gray matter of Alzheimer’s disease are immunoreactive with a cdc2-related (KKIARLE) antibody.

Table 8.1. Classification of glial pathology Glial Cell Type

Name

Composition

Associated Disorders

Astrocytes

Tufted astrocyte and thorn shaped astrocytes

Hyperphosphorylated tau

PSP, PD, PDC, PEP, SSPE

Astrocytes

Astrocytic plaque

Hyperphosphorylated tau

CBD

Oligodendrocytes Coiled bodies; oligodendrocyte Hyperphosphorylated tau microtubule bodies; glial fibrillary tangles

CBD, PSP, AGD, AD

Oligodendrocytes Glial cytoplasmic inclusion

Normal tau; Ubiquitin

MSA

Astrocytes

Superoxide dismutase

FALS

Lewy body-like inclusion

Abbreviations: PSP = progressive supranuclear palsy; PD = Pick’s disease; PDC = Parkinson dementia complex of Guam; PEP = postencephalitic Parkinsonism; SSPE = subacute sclerosing panencephalitis; CBD = corticobasal degeneration; AGD; argyrophilic grain dementia; AD = Alzheimer’s disease; MSA = multiple system atrophy (Shy-Drager syndrome); FALS = familial amyotrophic lateral sclerosis.

Astrocytes in Other Neurodegenerative Diseases

171

from the other inclusions noted above in their inconsistent immunoreactivity with antibodies to phosphate-dependent tau epitopes.61 A scheme for classifying neurodegenerative disorders based on presence and type of inclusion body has recently been proposed62 and modified here to fit with new information (Fig. 8.2). The major inclusion bodies in astrocytes in the aged brain are corpora amylacea, which are composed of glycosidic polymers rather than cytoskeletal proteins.29 (Corpora amylacea are discussed in chapters 4 and 10 of this volume.) A much less recognized change in glia in aging is tau immunoreactivity. This can be demonstrated with immunocytochemical methods and specific tau antibodies.63,64 Initial studies of tau protein distribution and immunoreactivity failed to recognize tau in glial cells. The fact that tau was considered to be a neuron-specific molecule restricted to the axoplasmic domain65 no doubt influenced interpretation of early studies. Redistribution of tau epitopes, especially phosphorylated tau epitopes, to the somatodendritic domain of neurons was considered a pathological trait.64 Presence of phospho-tau within the soma of small cells was more apt to be considered a pathological process in neurons than in glia. Rigorous double-labeling immunocytochemical studies with cell type-specific markers, such as glial fibrillary acidic protein, were needed before pathological tau was widely recognized in glia. While antibodies to tau protein do not stain astrocytes in brains of young adults and of other species, they do stain astrocytes in elderly individuals and especially in Alzheimer’s disease (Fig. 8.3). Certain astrocytes consistently show tau immunoreactivity in aged and Alzheimer’s disease brains. Tau-positive astrocytes are located in subpial regions at the base of the brain (e.g., basal forebrain), in the medial temporal lobe (e.g., amygdala) and subependymal regions (e.g., temporal horn of lateral ventricle). That the immunoreactivity is within astrocytes is unambiguous since there is staining of astrocytic end-feet at the glia limitans and around blood vessels (Fig. 8.4). The astrocytic tau is cytoplasmic, without formation of discrete inclusion bodies. Tau-positive astrocytes in aging and Alzheimer’s disease have not been well studied and no ultrastructural studies have been reported. More recently, glial inclusion bodies have been described in Alzheimer’s disease, but these are not a common finding compared to the other conditions under consideration.7 The tau-positive glia in Alzheimer’s disease are labeled by antibodies against transferrin and 2'3'-cyclic nucleotide 3'-phosphohydrolase, which are markers for oligodendrocytes. Ultrastructurally, they were composed of bundles of straight filaments about 16 nm in diameter.

Progressive Supranuclear Palsy (PSP) PSP is a sporadic degenerative disease associated with axial rigidity, vertical eye movement abnormalities and subcortical dementia, first described by Steele, Richardson and Olszewski.66,67 The pathology of PSP is characterized by neuronal loss and gliosis in a number of interrelated subcortical nuclei of the extrapyramidal system, including basal ganglia, motor nuclei of the thalamus, dopaminergic and other nuclei in the midbrain, noradrenergic neurons and neurons in the pontine base, inferior olivary nucleus and cerebellar dentate nucleus. In most of these locations neurons have NFT, and tau aggregates are detected within cell processes.68 The NFT in PSP differ from those in Alzheimer’s disease by the presence of 15 nm to 18 nm diameter straight, rather than twisted, filaments.69-71 Nevertheless, the neuronal inclusions in PSP contain tau protein that is very similar to PHF in Alzheimer’s disease with immunocytochemical methods.72 On the other hand, biochemical studies of tau protein in PSP reveal differences from those in Alzheimer’s disease73, 74 that may to some extent reflect the fact that pathological tau is derived from both neurons and glia in PSP. Specifically, the abnormal tau in PSP is composed predominantly of two isoforms while in Alzheimer’s disease PHF-tau is composed of three major isoforms. In the scheme proposed for classifying neurodegenerative disorders in Figure 8.2 these two forms are referred to as

172

Astrocytes in Brain Aging and Neurodegeneration

Fig. 8.2. Diagram illustrating a scheme for classifying neurodegenerative diseases depending on the presence and type of inclusion body. Note that glial inclusions are found in several branches of the tree, but that tau-positive glial inclusions are most frequent in those disorders with tau isoform restriction. Glial tau pathology is particularly prominent in those disorders with a pathological tau “doublet” isoform. Tau-positive glial inclusions are also seen in “triplet” tau disorders, but they are a minor or inconspicuous feature in most cases.

Astrocytes in Other Neurodegenerative Diseases

173

Fig. 8.3. Astrocytes in cortical gray matter are immunoreactive with an antibody to tau protein that recognizes an epitope near the carboxyl terminus of tau. Tissue was fixed in periodate-lysine paraformaldehyde and sectioned with a Vibratome.

“doublet” and “triplet” tau. In PSP pathological tau, some evidence suggests that this isoform pattern may reflect preferential accumulation of specific splice forms of tau. This observation, along with information about tau in glia, might suggest that observed biochemical differences in tau in PSP compared to Alzheimer’s disease may be related in part to the greater contribution of glial tau in PSP. Since tau was considered to be a neuron-specific protein, tau-positive inclusions in PSP were initially considered to be NFT in small neurons. Double labeling methods have now demonstrated unequivocally that many of the tau-positive small cells in PSP, especially in the basal ganglia, are astrocytes.1,21,75-81 Immunocytochemical studies with antibodies to tau protein also revealed abnormal filamentous profiles in cell processes in affected regions of gray matter and also white matter (e.g., pencil fibers in the caudate and putamen) in PSP (Fig. 8.5). These so-called “neuropil threads”81,82 were initially interpreted to be within neuronal processes, but ultrastructural immunolabeling studies have now shown that many of these profiles are within the cytoplasm of oligodendrocytes and also within loops of myelin sheaths, which are extensions of the cytoplasm of oligodendroglia.83 Oligodendroglial tau-positive inclusions have become a recognized pathological feature of PSP (Fig. 8.5). Similar oligodendroglial inclusions can be found in other neurodegenerative disorders, where they are often more abundant than PSP. They have been referred to as “coiled bodies.”84 Coiled bodies were first described in argyrophilic grain dementia (AGD), an uncommon, or at least under-recognized, neurodegenerative disorder described by the Braaks.84-86 AGD is named for grain-like lesions within the neuropil that can be detected with silver stains or tau immunostains. The grains correspond to filamentous aggregates within segmental domains of cell processes. The latter are mostly neuronal

174

Astrocytes in Brain Aging and Neurodegeneration

Fig. 8.4. Subpial astrocytes at the base of the brain near the basal forebrain show tau immunoreactivity. Note absence of immunostaining of corpora amylacea in some of the processes.

processes (i.e., dendrites) based upon immunoelectron microscopic studies, but some are also clearly within glial cell processes.86 Like oligodendroglial coiled bodies, grains are not disease-specific and can be found in several neurodegenerative disorders, most notably corticobasal degeneration (see below). Oligodendroglial coiled bodies should be distinguished from the glial cytoplasmic inclusions (GCI) within oligodendroglia that are the hallmark of multiple system atrophy (MSA).87-90 Oligodendroglial coiled bodies contain phosphorylated tau epitopes and are actually best recognized with immunocytochemical methods. In contrast, GCI are essentially limited to MSA and have a different morphology and antigenic makeup. Phospho-tau antibodies either fail to stain GCI or they do so inconsistently. Recent evidence suggests that tau in GCI has properties closer to normal tau than to the abnormal tau that aggregates in neuronal and glial inclusions in other disorders.61 GCI are also intensely immunoreactive with antibodies to ubiquitin, which is a small heat shock molecule involved in ATP-dependent proteolysis of abnormal or denatured proteins.87-90 In contrast, oligodendroglial coiled bodies and tau-positive argyrophilic inclusions in astrocytes in other disorders are very weakly immunoreactive for ubiquitin. As mentioned previously, astrocytes are invariably affected in PSP. The tau-positive inclusions in neurons and glia in PSP are composed of straight filaments at the ultrastructural level. A variety of names have been attached to the abnormal astrocytes in PSP, such as “tufted astrocytes” or “thorn-shaped astocytes”76-79,91 (Fig. 8.5). This nomenclature is purely descriptive and has not proven to be useful in discriminating astrocytic lesions in one disorder from another. Within a given disorder there is morphologic diversity of abnormal astro-

Astrocytes in Other Neurodegenerative Diseases

175

Fig. 8.5. Astrocytes and oligodendroglia have tau-positive inclusions in PSP. (A), (B) gray matter astrocytes [(A) cortex, (B) basal ganglia] and (C), (D) white matter glial inclusions [(C) cortex, (D) basal ganglia]. Note variability in tufted astrocyte morphology [(A) and (B)] depending upon anatomical region. Coiled bodies are evident in the cerebral white matter and the pencil fibers of the basal ganglia.

cytes. There is also variability in the appearance of the astrocytes depending upon the anatomical region in which they reside (Fig. 8.5). Tufted astrocytes similar to those in PSP can be seen in Pick’s disease,78,79,92 Guam Parkinson dementia complex,93 postencephalitic Parkinsonism94 and subacute sclerosing panencephalitis.95 All of these disorders have both neuronal (i.e., NFT) and glial tau inclusions. The major means of differentiating the disorders is not by the appearance of the astrocyte inclusions, but rather by the clinical presentation and pathological findings, as well as the distribution and nature of the glial and neuronal pathology. Biochemical analysis of the major tau isoforms that accumulate in brain tissue may offer another means of differentiating the disorders (see Fig. 8.2).

Pick’s Disease Pick’s disease is a rare late-life degenerative disorder presenting as dementia and personality deterioration due to circumscribed (“lobar”) atrophy with marked neuronal loss and gliosis in the frontal and anterior temporal lobes.96 While there is no universal agreement as to the defining feature of Pick’s disease, the presence of argyrophilic, round inclusion bodies within neurons is an increasingly accepted pathological hallmark. These

176

Astrocytes in Brain Aging and Neurodegeneration

inclusion bodies were originally described by Alzheimer97 and have come to be known as Pick bodies. Pick bodies are filamentous inclusions composed of altered tau protein that is highly phosphorylated, similar to tau proteins in Alzheimer’s disease.98 Few biochemical studies have been performed on cytoskeletal proteins in Pick’s disease, but those that have been reported suggest that the tau protein abnormalities in Pick’s disease are distinct from Alzheimer’s disease and closer to those in PSP, with expression of two major pathological tau isoforms.99 Pick bodies differ most from NFT in their distinct anatomical distribution and characteristic microscopic appearance. At the ultrastructural level, Pick bodies contain filaments of variable morphology, including wide and long-period, twisted filaments or straight filaments.100-102 Astrocytic inclusions in Pick’s disease have a distinctive appearance, but show morphologic overlap with lesions in the various disorders under consideration. In Pick’s disease and PSP the inclusions are most often of the tufted or thorn-shaped appearance (Fig. 8.6). In Pick’s disease, abnormal astrocytes are found in affected cortical regions, while in PSP they are most abundant in the basal ganglia. In the cortex in PSP they are confined to motor and premotor cortex. Thus, glial pathology in these disorders parallels the distribution of the other more widely recognized neuronal pathology. The tufted astrocytes in PSP often have filamentous aggregates that are often displaced into the cell processes with less staining in the perinuclear region, which accounts for the tufted appearance. In contrast, in Pick’s disease the filamentous aggregates are often in a more proximal perinuclear cellular domain (Fig. 8.6). Despite this general observation, the overlap in appearance precludes a morphological basis for neuropathological diagnosis. Far more important is the distribution of the pathology.105 White matter pathology is a well-known feature of Pick’s disease, with loss of myelinated fibers in a distribution that parallels the areas with most marked cortical atrophy. The degree of myelin loss correlates with the severity of the cortical atrophy and is reflected in loss of myelin-related lipids.103 These observations strongly suggest that white matter pathology is a type of Wallerian degeneration secondary to cortical neuronal loss; however, the recent discovery of cytoskeletal inclusion bodies in oligodendroglia and astrocytes in white matter in Pick’s disease raises the possibility that the white matter pathology may be an integral part of the disease rather than a secondary change. An interesting phenotype of oligodendroglial inclusions in Pick’s disease includes round inclusion bodies that are highly reminiscent of neuronal Pick bodies.78 More often oligodendroglial lesions in Pick’s disease have the appearance of coiled bodies or oligodendroglial microtubular inclusions similar to those of PSP and other disorders21,104 (Fig. 8.6).

Corticobasal Degeneration (CBD) CBD is a rare, sporadic disorder whose classical clinical picture is one of asymmetrical rigidity, dystonia and apraxia, with mild or inapparent cognitive deterioration.106,107 It is becoming clear that the clinical phenotype is broader than originally described. Many subjects have progressive dysphasia, reflecting the fact that the brunt of the pathology is often in the dominant cerebral hemisphere.100,108,109 The hallmark lesion of CBG is the achromatic neuron.106 Achromatic neurons are swollen and weakly stained with routine histochemical methods, hence their name. They are also referred to as ballooned or swollen chromatolytic neurons.110 Ballooned neurons lack diagnostic specificity when found in limbic areas, but are highly characteristic of CBD when found in the convexity cerebral cortex, especially in the superior frontal and parietal lobes. They are intensely immunoreactive with phosphorylated neurofilament antibodies and variably stained with tau antibodies.111 Tau antibodies also reveal a host of other pathologies in CBD, foremost among them being inclusions in both oligodendroglia and astrocytes.12,21,42,43,78,79,112-117

Astrocytes in Other Neurodegenerative Diseases

177

Fig. 8.6. Astrocytic and neuronal inclusions in Pick’s disease are immunoreactive with a tau antibody. (A) Note neuronal Pick body (arrow head) and two taupositive astrocytes (arrows). (B) Astrocytic nature of the glial inclusions is obvious in that some of the cells form endfeet to blood vessels (arrow). (C) In white matter some of the glial inclusions in apparent oligodendrocytes appear similar to small Pick bodies. Reprinted with permission from Vincent I et al, J Cell Biol 1996; 132:413-425.

Glial inclusions are more numerous in CBD than in any of the other disorders discussed, which may account for the many reports in recent years on this disorder.12,21,42,43,78,79,112-117 While morphological distinctions may not differentiate the disorders, the relative abundance of neuronal versus glial inclusions, and the distribution in gray versus white matter and forebrain vs. hindbrain, clearly differs in PSP, Pick’s disease and CBD.105 In particular, Pick’s disease is predominantly a neuronal disorder, CBD predominantly glia and PSP both. PSP affects mostly deep gray matter, Picks’ disease mostly cortical and CBD both. In CBD the predominant distribution of abnormal tau protein is within cell processes of glia and neurons with abundant white matter disease. Neither Picks’ disease nor PSP have white matter pathology as marked as that of CBD. If any of the astrocytic lesions is diagnostically useful, perhaps the so-called “astrocytic plaque” of CBD comes closest to meeting this criterion (Fig. 8.7). The astrocytic plaque is most apparent in the affected cortical regions, with fewer lesions in deep gray matter. In the astrocytic plaque abnormal tau accumulates in distal cellular processes of reactive astrocytes, forming an annular arrangement of miliary structures in the neuropil.105,115 The appearance is suggestive of a neuritic plaque in Alzheimer’s disease, but the central core does not contain amyloid. Instead, specific cellular markers demonstrate a central astrocyte with dilated distal processes containing tau immunoreactivity. Ultrastructural studies demonstrate that filaments within the cell body of astrocytes resemble glial filaments in fibrous astrocytes, while those in distal segmental cell processes are thicker. The filaments in the cell bodies are not immunostained with tau and PHF antibodies, while the ones in distal processes are positive (Fig. 8.8). In considering the predominant cellular domain of the astrocyte affected by tau aggregates, Pick’s disease, PSP and CBD appear to differ. In Pick’s disease aggregates are in or close to the cell body, in PSP proximal processes tend to be affected and in CBD distal segmental domains are preferentially affected. Cerebral white matter shows marked pathology in CBD characterized by numerous tau-positive thread-like processes as well as many tau-positive glial cells (i.e., coiled bodies) (Fig. 8.7). The abnormal white matter tau aggregates have been shown with double labeling methods to reside in both neuronal axons and glial processes.115 The white matter tau

178

Astrocytes in Brain Aging and Neurodegeneration

Fig. 8.7. Tau-positive glia are numerous in CBD. (A) Many tau-positive astrocytic plaques are visible in this low power image of superior frontal cortex. (B) Double stained sections (monoclonal antibodies to vimentin and to tau) show a central astrocyte in the middle of the clusters of stubby tau-positive processes. (C) The cerebral white matter has many glial inclusions in oligodendrocytes (coiled bodies). (D) White matter processes are numerous in certain cortical and subcortical fiber tracts. Other studies have demonstrated that these processes are within glial and neuronal processes. Ultrastructural studies suggest that predominantly oligodendrocytes produce this lesion.

pathology in CBD is greater than in any of the other conditions. It is most marked in areas that also show the greatest cortical pathology, but is also very severe in certain diencephalic fiber tracts where neuronal pathology is not marked, most notably the thalamic fasciculus. In some cases the pattern of white matter tau-positive pathology follows defined anatomic pathways, such as the corticospinal tract in the cerebral peduncle and medullary pyramid. Given the abundance of oligodendroglial and astrocytic tau pathology in CBD, if biochemistry of brain samples reflects glial changes in any disorder it is most likely to be representative of glial changes in CBD. In this disorder biochemical studies show an abnormal tau pattern with two major tau isoforms, probably due to preferential alternative splicing of tau DNA.12,118 Exon 3 is also underrepresented in pathological tau in CBD.42,43 There is evidence to suggest that exon 10 is abnormally expressed in the abnormal tau protein of CBD.12,22 Exon 10 is another alternatively spliced exon in tau,24 which determines whether tau will have three or four repeats in the microtubule-binding domain. It is suggested that preferential tau splicing contributes to the restricted isoform pattern in CBD. What remains to be

Astrocytes in Other Neurodegenerative Diseases

179

Fig. 8.8. An astrocyte from the cortex in CBD shows bundles of glial intermediate filaments in the cytoplasm. In distal processes the filaments are thicker and more loosely spaced. While it is impossible to state with certainty the cell of origin of the small distal processes, they contain bundles of straight filaments. (This section was immunostained with an antibody to PHF before embedding in plastic, with detection using a peroxidase method). The glial filaments are about 5 nm to 8 nm in diameter and the immunolabeled filaments are more than twice as thick (about 25 nm). (Note that immunoperoxidase decoration increases the outer diameter. Filaments that are not immuno-decorated have a thickness of about 18 nm to 20 nm.) (Scale bar is 833 nm, and 540 nm for inset)

determined is if this represents tau derived from glia (e.g., oligodendroglia36) where this is the normal expression pattern of tau or if it reflects an abnormal phenotype of tau specific to the disease.

Argyrophilic Grain Dementia (AGD) The clinical phenotype of AGD, also referred to as Braak’s disease, is not clear.84 While many of the individuals with this pathology have dementia, this does not seem to be invariable. In some cases memory disorders seem to predominate, while in others grain-type pathology is detected in clinically asymptomatic individuals. The changes in the brain are anatomically restricted and characterized by tau aggregates within short segments of distal cell processes of neurons and glia.85,86 The major pathology is found in the limbic gray matter, with very few pathological changes outside of this region.84 In the original description, the Braaks emphasized white matter pathology with coiled inclusion bodies in oligodendroglia that were positive with silver stains and antibody methods for tau protein. The inclusions were composed of fibrillar material that appeared to encircle the nucleus and to extend for variable distances into the cell processes. Affected cells were shown to be oligodendrocytes with cell type-specific markers. Astrocytic tau inclusions are far less common

180

Astrocytes in Brain Aging and Neurodegeneration

and have not been emphasized in the descriptions of AGD. Disorders that have been described as being an overlap between PSP and AGD,119 having many glial tau-positive inclusions, likely represent CBD on second consideration, given the fact that ballooned neurons are consistently present in these cases (E. Masliah, personal communication).

Familial Frontotemporal Dementia and Parkinsonism Linked to Chromosome 17 (FTDP-17) Frontotemporal dementia describes a clinical phenotype with variable pathology. The clinical syndrome includes dementia with variable personality changes, frontal lobe signs, motor involvement and language disturbance. There is often asymmetry, with the left hemisphere more often affected than the right. Some cases of FTD have Pick’s disease, while most have dementia lacking distinctive histology.120,121 A variable degree of Parkinsonism is not uncommon in FTD and about 10% of cases are hereditary.121 A recent consensus conference proposed the term FTDP as an umbrella term to describe this group of disorders that had previously gone under a wide diversity of terms.122 All familial cases included in the report were variably linked to a gene of chromosome 17.122 Interestingly, the tau gene is located within the consensus region. Closer inspection of many of the cases of FTDP-17 has shown that tau-positive inclusions in neurons and glia are common.22,123 In some cases tau pathology is marked, with numerous inclusions in neurons, oligodendroglia and astrocytes. This has prompted the designation for the disease in one of the families as a “tau-opathy”.122 Such cases have a great deal of overlap with PSP and CBD.22 Furthermore, recent studies suggest that PSP is also linked to a polymorphism in the tau gene on chromosome 17.124,125 The polymorphism is within an intron in tau (between exons 9 and 10) that is subject to alternative splicing.124 Glial pathology in FTDP-17 includes coiled bodies, white matter threads, tufted astrocytes and in some cases astrocytic plaques similar to those in CBD. While CBD is usually a nonfamilial disorder, the great similarity in pathology between some cases of FTDP-17 and CBD suggests that some cases may be a familial form of CBD.

Multiple System Atrophy (MSA) Multiple system atrophy refers to a symptom complex that includes cerebellar ataxia, Parkinsonism, orthostatic hypotension and autonomic dysfunction.126 The pathology is variable and reflects the predominant clinical phenotype. The pathological diagnoses that are subsumed under the rubric of MSA include sporadic olivopontocerebellar degeneration, striatonigral degeneration and Shy-Drager syndrome.87,126 The hallmark of MSA is the glial cytoplasmic inclusion (GCI) which is a round or crescent-shaped inclusion that is intensely argyrophilic and positive with ubiquitin antibodies, but variably stained or negative with tau antibodies61,87-90 (Fig. 8.9). Recent studies suggest that GCI contain tau protein that is not as highly phosphorylated as tau in PHF and more analogous to normal tau.61 This alone indicates that aggregation and polymerization of cytoskeletal elements, and tau in particular, are not dependent on high phosphate content of the cytoskeletal components. The discovery of glial inclusions in MSA was a breakthrough in the nosology and classification of the various spinocerebellar degenerations. The presence of glial cytoplasmic inclusions has become a means of differentiating sporadic from hereditary forms of spinocerebellar degeneration.127 The latter, which are often trinucleotide repeat disorders,128 are only rarely associated with glial cytoplasmic inclusions that are positive with tau immunocytochemistry.129 On the other hand, a different type of inclusion body has recently been identified in many (if not all) of the trinucleotide repeat disorders, including Huntington’s disease, and that is the intranuclear inclusion.130 The latter are most often within nuclei of neurons, but some of these disorders, as well as sporadic disorders, have similar hyaline inclusions with glial nuclei.131 In these disorders, evidence suggests that the protein that

Astrocytes in Other Neurodegenerative Diseases

181

Fig. 8.9. Glial cytoplasmic inclusions are the hallmark lesions of multiple system atrophy. In this section the white matter glia from cerebellum (A) and pontine base (B) have crescent-shaped inclusion bodies [arrows in (A)]. These inclusions were completely negative with phospho-tau antibodies. Occasional thread-like profiles are also visible with ubiquitin staining [arrowhead in (B)], which also probably represents inclusions in oligodendroglial cytoplasm.

aggregates is also the mutated protein, but further investigation is clearly needed in this emerging area. It is unknown how, or if, nuclear inclusions in neurons and glia affect cellular function.

Familial Amyotrophic Lateral Sclerosis (FALS) ALS is usually a sporadic condition associated with progressive degeneration of motor neurons in brainstem and spinal cord. The clinical syndrome is one of progressive weakness and muscle atrophy with preservation of higher cognitive funcitons.132 The pathology is that of selective neuronal loss and gliosis relatively confined to the upper and lower motor neurons and denervation atrophy of skeletal muscle.133 It is variably associated with degeneration in other systems, such as the spinocerebellar and somatic sensory pathways, especially in familial forms.133 There are few histological hallmarks that are specific to the disorder. The exception is the Bunina body, an inclusion in affected neurons in ALS that appears to be derived from membranous organelles, possibly related to the endoplasmic reticulum.108 In familial cases a number of investigators have reported hyaline cytoplasmic inclusions in affected neuronal populations in the motor cortex, brainstem and anterior horn of the spinal cord.134-136 The neuronal inclusions in familial ALS resemble Lewy bodies of idiopathic Parkinson’s disease (for review see ref. 2). These spherical hyaline cytoplasmic inclusions contain neurofilament protein and are highly ubiquitinated.134-136 More recently, Lewy body-like inclusions have also been described in astrocytes13,137 in certain familial forms of amyotrophic lateral sclerosis associated with mutations in Cu/Zn superoxide dismutase.138 Comparable inclusion bodies are also detected in transgenic mice carrying the mutation in Cu/Zn superoxide dismutase.139 Developmental studies in this animal model suggest that glial pathology may actually precede neuronal changes. This may be a valuable model for exploring the role of glial pathology in neurodegeneration. At the ultrastructural level the inclusions in both the transgenic animals and in astrocytes in humans with FALS are granule-coated filaments.13,138 Since they are immunoreactive

182

Astrocytes in Brain Aging and Neurodegeneration

with antibodies to SOD, they may be composed of polymers of this molecule, but this remains to be demonstrated by direct biochemical analysis. This would be unprecedented. All other common filamentous inclusions in neurons and glia are derived from cytoskeletal proteins. On the other hand, the lesson from trinucleotide repeat disorders and from various amyloidoses is that pathological fibrils may be formed from proteins of diverse molecular nature, particularly if they have an altered conformation that favors the low energy state of stable filaments.

Acknowledgments Immunocytochemical studies of Drs. Mel Feany and Linda Mattiace were important in defining glial pathology in non-AD disorders. Biochemical studies were performed by Dr. Hanna Ksiezak-Reding. Tissue culture studies of human fetal astrocytes were performed by Dr. Deke He. Additional tissue culture studies were performed by Drs. Sunhee Lee and Meng-Liang Zhao. Discussion of various aspects of this work with Drs. Peter Davies, Bridget Shafit-Zagardo and Shu-Hui Yen are acknowledged. Yvonne Kress assisted with ultrastructural studies. The efforts of these individuals at Albert Einstein College of Medicine is gratefully acknowledged. Support for this research was provided by NIA AG06803.

References 1. Chin SS, Goldman JE. Glial inclusions in CNS degenerative diseases. J Neuropathol Exp Neurol 1996; 55:499-508. 2. Pollanen MS, Dickson DW, Bergeron C. Pathology and biology of the Lewy body. J Neuropathol Exp Neurol 1993; 52:183-191. 3. Wakabayashi K, Takahashi H. Gallyas-positive, tau-negative glial inclusions in Parkinson’s disease midbrain. Neurosci Lett 1996; 217:133-136. 4. Goedert M, Spillantini MG, Jakes R et al. Multiple isoforms of human microtubule-associated protein tau: Sequences and localization in neurofibrillary tangles of Alzheimer’s disease. Neuron 1989; 3:519-526. 5. Morrison JH, Hof PR. Life and death of neurons in the aging brain. Science 1997; 278:413-419. 6. Dickson DW, Crystal HA, Mattiace LA et al. Identification of normal and pathological aging in prospectively studied nondemented elderly humans. Neurobiol Aging 1992; 13:1-11. 7. Nishimura M, Tomimoto H, Suenaga T et al. Immunocytochemical characterization of glial fibrillary tangles in Alzheimer’s disease brain. Am J Pathol 1995; 146:1052-1058. 8. Phillips LL, Autilio-Gambetti L, Lasek RJ. Bodian’s silver method reveals molecular variation in the evolution of neurofilament proteins. Brain Res 1983; 278:219-223. 9. Wisniewski HM, Narang HK, Terry RD. Neurofibrillary tangles of paired helical filaments. J Neurol Sci 1976; 27:173-181. 10. Pollanen MS, Markiewicz P, Goh MC. Paired helical filaments are twisted ribbons composed of two parallel and aligned components: Image reconstruction and modeling of filament structure using atomic force microscopy. J Neuropathol Exp Neurol 1997; 56:79-85. 11. Yen S-H, Dickson DW, Peterson C et al. Cytoskeletal abnormalities in neuropathology. In: Zimmerman HM, ed. Progress in Neuropathology, Vol 6. New York: Raven Press, 1986:63-90. 12. Ksiezak-Reding H, Morgan K, Mattiace LA et al. Ultrastructure and biochemical composition of paired helical filaments in corticobasal degeneration. Am J Pathol 1994; 145:1496-1508. 13. Kato S, Shimoda M, Watanabe Y et al. 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 1996; 55:1089-1101. 14. Greenberg SG, Davies P. A preparation of Alzheimer paired helical filaments that displays distinct tau proteins by polyacrylamide gel electrophoresis. Proc Natl Acad Sci USA 1990; 87:5827-5831.

Astrocytes in Other Neurodegenerative Diseases

183

15. Lee VM, Balin BJ, Otvos L Jr et al. A68: A major subunit of paired helical filaments and derivatized forms of normal tau. Science 1991; 251:675-678. 16. Grundke-Iqbal I, Iqbal K, Tung YC et al. Abnormal phosphorylation of the microtubuleassociated protein tau (τ) in Alzheimer cytoskeletal pathology. Proc Natl Acad Sci USA 1986; 83:4913-1917. 17. Yen S-H, Dickson DW, Crowe A et al. Alzheimer neurofibrillary tangles contain unique epitopes and epitopes in common with heat-stable microtubule-associated proteins, MAP2 and tau. Am J Pathol 1987; 126:63-73. 18. Goedert M, Wischik CM, Crowther RA et al. Cloning and sequencing of the cDNA encoding a core protein of the paired helical filament of Alzheimer disease: Identification as the microtubule-associated protein tau. Proc Natl Acad Sci USA 1988; 85:4051-4055. 19. Goedert M, Spillantini MG, Cairns NJ et al. Tau proteins of Alzheimer paired helical filaments: Abnormal phosphorylation of all six brain isoforms. Neuron 1992; 8:159-168. 20. Kosik KS, Orecchio LD, Binder L et al. Epitopes that span the tau molecule are shared with paired helical filaments. Neuron 1988; 1:817-825. 21. Iwatsubo T, Hasegawa M, Ihara Y. Neuronal and glial tau-positive inclusions in diverse neurologic diseases share common phosphorylation characteristics. Acta Neuropathol 1994; 88:129-136. 22. Spillantini MG, Goedert M, Crowther RA et al. Familial multiple system tauopathy with presenile dementia: A disease with abundant neuronal and glial tau filaments. Proc Natl Acad Sci USA 1997; 94:4113-4118. 23. Neve RL, Harris P, Kosik KS et al. Identification of cDNA clones for the human microtubule-associated protein tau and chromosomal localization of the genes for tau and microtubule-associated protein 2. Brain Res 1986; 387:271-280. 24. Goedert M, Crowther RA, Garner CC. Molecular characterization of microtubule-associated proteins tau and MAP2. TINS 1991; 14:193-199. 25. Mandelkow E-W, Schweers O, Drewes G et al. Structure, microtubule interactions and phosphorylation of tau protein. In: Wurtman RJ, Corkin S, Growdon JH, Nitsch RM, eds. The Neurobiology of Alzheimer’s Disease. New York: New York Academy of Science, 1996:96-106. 26. Ksiezak-Reding H, Liu WK, Yen SH. Phosphate analysis and dephosphorylation of modified tau associated with paired helical filaments. Brain Res 1992; 597:209-219. 27. Bramblett GT, Trojanowski JQ, Lee VM. Regions with abundant neurofibrillary pathology in human brain exhibit a selective reduction in levels of binding-competent tau and accumulation of abnormal tau-isoforms (A68 proteins). Lab Invest 19992; 66:212-222. 28. Trojanowski JQ, Schmidt ML, Shin R-W et al. PHFτ (A68): From pathological marker to potential mediator of neuronal dysfunction and degeneration in Alzheimer’s disease. Clin Neuroscience 1993; 1:184-191. 29. Hirano A. Neurons and astrocytes. In: Davis RL, Robertson DM, eds. Textbook of Neuropathology. Baltimore: Williams & Wilkins, 1997:1-110. 30. Raine CS. Oligodendrocytes and central nervous system myelin. In: Davis RL, Robertson DM, eds. Textbook of Neuropathology. Baltimore: Williams & Wilkins, 1997:111-164. 31. Zientek GM, Herman MM, Katsetos CD et al. Absence of neuron-associated microtubule proteins in the rat C-6 glioma cell line. A comparative immunoblot and immunohistochemical study. Neuropathol Appl Neurobiol 1993; 19:346-349. 32. Geisert EE Jr, Johnson HG, Binder LI. Expression of microtubule-associated protein 2 by reactive astrocytes. Proc Natl Acad Sci USA 1990; 87:3967-3971. 33. Lin RC, Matesic DF. Immunohistochemical demonstration of neuron-specific enolase and microtubule-associated protein 2 in reactive astrocytes after injury in the adult forebrain. Neuroscience 1994; 60:11-16. 34. Doll T, Meichsner M, Riederer BM et al. An isoform of microtubule-associated protein 2 (MAP2) containing four repeats of the tubulin-binding motif. J Cell Sci 1993; 106:633-639. 35. Binder LI, Frankfurter A, Rebhun LI. The distribution of tau in the mammalian central nervous system. J Cell Biol 1985; 101:1371-1378.

184

Astrocytes in Brain Aging and Neurodegeneration

36. Ksiezak-Reding H, Farooq M, Norton W et al. A distinct subset of tau isoforms is expressed in bovine brain oligodendrocytes. Mol Biol Cell 1997; 8:263a. 37. Muller R, Heinrich M, Heck S et al. Expression of microtubule-associated proteins MAP2 and tau in cultured rat brain oligodendrocytes. Cell Tissue Res 1997; 288:239-249. 38. Liu W, Shafit-Zagardo B, Aquino D et al. Cytoskeletal alterations in human fetal astrocytes induced by IL-1β. J Neurochem 1994; 63:1625-1634. 39. Arias C, Sharma N, Davies P et al. Okadaic acid induces early changes in microtubuleassociated protein 2 and tau phosphorylation prior to neurodegeneration in cultured cortical neurons. J Neurochem 1993; 61:673-682. 40. Dewar D, Dawson D. Tau protein is altered by focal cerebral ischemia in the rat: An immunohistochemical and immunoblotting study. Brain Research 1995; 684:70-78. 41. Irving EA, Yatsushiro K, McCulloch J et al. Rapid alteration of tau in oligodendrocytes after focal ischemic injury in the rat: involvement of free radicals. J Cerebral Blood Flow Met 1997;17:612-622. 42. Feany MB, Ksiezak-Reding H, Liu WK et al. Epitope expression and hyperphosphorylation of tau protein in corticobasal degeneration: Differentiation from progressive supranuclear palsy. Acta Neuropathol 1995; 90:37-43. 43. Nishimura T, Ikeda K, Akiyama H et al. Glial tau-positive structures lack the sequence encoded by exon 3 of the tau protein gene. Neurosci Lett 1997; 224:169-172. 44. Wille H, Drewes G, Biernat J et al. Alzheimer-like paired helical filaments and antiparallel dimers formed from microtubule-associated protein tau in vitro. J Cell Biol 1992; 118:573-584. 45. Goedert M, Jakes R, Spillantini MG et al. Assembly of microtubule-associated protein tau into Alzheimer-like filaments induced by sulfated glycosaminoglycans. Nature 1996; 383:550-553. 46. Wilson DM, Binder LI. Free fatty acids stimulate the polymerization of tau and amyloid beta peptides. In vitro evidence for a common effector of pathogenesis in Alzheimer’s disease. Am J Pathol 1997; 150:2181-2195. 47. Iqbal K, Grundke-Iqbal I. Ubiquitination and abnormal phosphorylation of paired helical filaments in Alzheimer’s disease. Molec Neurobiol 1991;5:399-410. 48. Smith MA, Taneda S, Richey PL et al. Advanced Maillard reaction end products are associated with Alzheimer disease pathology. Proc Natl Acad Sci USA 1994; 91:5710-5714. 49. Yan S-D, Chen X, Schmid A-M et al. Glycated tau protein in Alzheimer disease: A mechanism for induction of oxidant stress. Proc Natl Acad Sci USA 1994; 91:7787-7791. 50. Ledesma MD, Bonay P, Colaco C et al. Analysis of microtubule-associated protein tau glycation in paired helical filaments. J Biol Chem 1994; 269:21614-21619. 51. Yen S-H, Liu W-K, Hall FL et al. Alzheimer neurofibrillary lesions: Molecular nature and potential roles of different components. Neurobiol Aging 1995; 16:381-387. 52. Goedert M, Jakes R, Qi Z et al. Protein phosphatase 2A is the major enzyme in brain that dephosphorylates tau protein phosphorylated by proline-directed protein kinases or cyclic AMP-dependent protein kinase. J Neurochem 1995; 65:2804-2807. 53. Garver TD, Harris KA, Lehman RA et al. Tau phosphorylation in human, primate, and rat brain: Evidence that a pool of tau is highly phosphorylated in vivo and is rapidly dephosphorylated in vitro. J Neurochem 19994; 63:2279-2287. 54. Savory J, Huang Y, Herman MM et al. Quantitative image analysis of temporal changes in tau and neurofilament proteins during the course of acute experimental neurofibrillary degeneration; Non-phosphorylated epitopes precede phosphorylation. Brain Res 1996; 707:272-281. 55. Goedert M. Tau protein and the neurofibrillary pathology of Alzheimer’s disease. In: Wurtman RJ, Corkin S, Growdon JH, Nitsch RM, eds. The Neurobiology of Alzheimer’s Disease. New York: Ann NY Acad Sci 1996; 777:121-131. 56. Liu W-K, Williams RT, Hall FL et al. Detection of a cdc2-related kinase associated with Alzheimer paired helical filaments. Am J Pathol 1995; 146:228-238. 57. Vincent I, Jicha G, Rosado M et al. Aberrant expression of mitotic cdc2/cyclin B1 kinase in degenerating neurons of Alzheimer’s disease brain. J Neuroscience 1997; 17:3588-3598.

Astrocytes in Other Neurodegenerative Diseases

185

58. Vincent I, Rosado M, Davies P. Mitotic mechanisms in Alzheimer’s disease? J Cell Biol 1996; 132:413-425. 59. Ross ME. Cell division and the nervous system: Regulating the cycle from neural differentiation to death. TINS 1996; 19:62-68. 60. Yen SH, Kenessey A, Lee SC et al. The distribution and biochemical properties of a Cdc2related kinase, KKIALRE, in normal and Alzheimer brains. J Neurochem 1995; 65:2577-2584. 61. Cairns NJ, Atkinson PF, Hanger DP et al. Tau protein in the glial cytoplasmic inclusions of multiple system atrophy can be distinguished from abnormal tau in Alzheimer’s disease. Neurosci Lett 1997; 230:49-52. 62. Dickson DW. Neurodegenerative diseases with cytoskeletal pathology: A biochemical classification. Ann Neurol 1997; 42:541-544. 63. Dickson DW, Ksiezak-Reding H, Liu W-K et al. Immunocytochemistry of neurofibrillary tangles (NFT) with antibodies to subregions of tau protein: Identification of hidden and cleaved epitopes and a new phosphorylation site. Acta Neuropathol 1992; 84:596-605. 64. Papasozomenos SC, Binder LI. Phosphorylation determines two distinct species of Tau in the central nervous system. Cell Motil Cytoskeleton 1987; 8:210-226. 65. Dotti CG, Banker GA, Binder LI. The expression and distribution of the microtubule-associated proteins tau and microtubule-associated protein 2 in hippocampal neurons in the rat in situ and in cell culture. Neuroscience 1987; 23:121-130. 66. Steele JC, Richardson JC, Olszewski J. Progressive supranuclear palsy: A heterogeneous degeneration involving the brain stem, basal ganglia and cerebellum, with vertical gaze and pseudobulbar palsy, nuchal dystonia and dementia. Arch Neurol 1964; 10:333-339. 67. Litvan I, Agid Y. Progressive Supranuclear Palsy: Clinical and Research Approaches. New York: Oxford University Press, 1992. 68. Jellinger K, Riederer P, Tomonaga M. Progressive supranuclear palsy: Clinico-pathological and biochemical studies. J Neural Transm 1980; 16(Suppl):111-128. 69. Roy S, Datta CK, Hirano A et al. Electron microscopic study of neurofibrillary tangles in Steele-Richardson-Olszewski syndrome. Acta Neuropathol 1974; 29:175-179. 70. Tellez-Nagel I, Wisniewski HM. Ultrastructure of neurofibrillary tangles in SteeleRichardson-Olszewski syndrome. Arch Neurol 1973; 29:324-327. 71. Yagishita S, Itoh Y, Amano N et al. Ultrastructure of neurofibrillary tangles in progressive supranuclear palsy. Acta Neuropathol 1979; 48:27-30. 72. Schmidt ML, Lee VM-Y, Hurtig H et al. Properties of antigenic determinants that distinguish neurofibrillary tangles in progressive supranuclear palsy and Alzheimer’s disease. Lab Invest 1988; 59:460-466. 73. Flament S, Delacourte A, Verny M et al. Abnormal tau proteins in progressive supranuclear palsy. Similarities and differences with the neurofibrillary degeneration of the Alzheimer type. Acta Neuropathol 1991; 81:591-596. 74. Vermersch P, Robitaille Y, Berneir L et al. Biochemical mapping of neurofibrillary degeneration in a case of progressive supranuclear palsy: Evidence for general cortical involvement. Acta Neuropathol 1994; 87:572-577. 75. Nishimura M, Namba Y, Ideda K et al. Glial fibrillary tangles with straight tubules in the brains of patients with progressive supranuclear palsy. Neurosci Lett 1992; 143:35-38. 76. Yamada T, Calne DB, Akiyama H et al. Further observations on tau-positive glia in the brains with progressive supranuclear palsy. Acta Neuropathol 1993; 85:308-315. 77. Nishimura T, Ikeda K, Akiyama H et al. Immunohistochemical investigation of tau-positive structures in the cerebral cortex of patients with progressive supranuclear palsy. Neurosci Lett 1995; 201:123-126. 78. Feany MB, Mattiace LA, Dickson DW. Neuropathologic overlap of progressive supranuclear palsy, Pick’s disease and corticobasal degeneration. J Neuropathol Exp Neurol 1996; 55:53-67. 79. Dickson DW, Feany MB, Yen SH et al. Cytoskeletal pathology in non-Alzheimer degenerative dementia: New lesions in diffuse Lewy body disease, Pick’s disease, and corticobasal degeneration. J Neural Transm. Suppl 1996;47:31-46.

186

Astrocytes in Brain Aging and Neurodegeneration

80. Li F, Iseki E, Kosaka K et al. Progressive supranuclear palsy with fronto-temporal atrophy and various tau-positive abnormal structures. Clin Neuropathol 1996; 15:319-323. 81. Probst A, Langui D, Lautenschlager C et al. Progressive supranuclear palsy: Extensive neuropil threads in addition to neurofibrillary tangles. Acta Neuropathol 1988; 77:61-68. 82. Nelson SJ, Yen S-H, Davies P et al. Basal ganglia neuropil threads in progressive supranuclear palsy. J Neuropathol Exp Neurol 1989; 48:324. 83. Arima K, Nakamura M, Sunohara N et al. Ultrastructural characterization of the tau-immunoreactive tubules in the oligodendroglial perikarya and their inner loop processes in progressive supranuclear palsy. Acta Neuropathol 1997; 93:558-566. 84. Braak H, Braak E. Argyrophilic grains: Characteristic pathology of cerebral cortex in cases of adult onset dementia without Alzheimer changes. Neurosci Lett 1987; 76:124-127. 85. Tolnay M, Spillantini MG, Goedert M et al. Argyrophilic grain disease: Widespread hyperphosphorylation of tau protein in limbic neurons. Acta Neuropathol 1997; 93:477-484. 86. Ikeda K, Akiyama H, Kondo H et al. A study of dementia with argyrophilic grains. Possible cytoskeletal abnormality in dendrospinal portion of neurons and oligodendroglia. Acta Neuropathol 1995; 89:409-414. 87. Papp MI, Kahn JE, Lantos PL. Glial cytoplasmic inclusions in the CNS of patients with multiple system atrophy (striatonigral degeneration, olivopontocerebellar atrophy and ShyDrager syndrome). J Neurol Sci 1989; 94:79-100. 88. Horoupian DS, Dickson DW. Striatonigral degeneration, olivoponto-cerebellar atrophy and “atypical” Pick disease. Acta Neuropathol 1991; 81:287-295. 89. Kato S, Nakamura H. Cytoplasmic argyrophilic inclusions in neurons of pontine nuclei in patients with olivopontocerebellar atrophy: Immunohistochemical and ultrastructural studies. Acta Neuropathol 1990; 79:584-594. 90. Tamaoka A, Mizusawa H, Mori H et al. Ubiquitinated alpha B-crystallin in glial cytoplasmic inclusions from the brain of a patient with multiple system atrophy. J Neurol Sci 1995; 129:192-198. 91. Ikeda K, Akiyama H, Kondo H et al. Thorn-shaped astrocytes: Possibly secondarily induced tau-positive glial fibrillary tangles. Acta Neuropathol 1995; 90:620-625. 92. Yamazaki M, Nakano I, Imazu O et al. Astrocytic straight tubules in the brain of a patient with Pick’s disease. Acta Neuropathol 1994; 88:587-591. 93. Oyanagi K, Makifuchi T, Ohtoh T et al. Distinct pathological features of the Gallyas- and tau-positive glia in the Parkinsonism-dementia complex and amyotrophic lateral sclerosis of Guam. J Neuropathol Exp Neurol 1997; 56:308-316. 94. Ikeda K, Akiyama H, Kondo H et al. Anti-tau-positive glial fibrillary tangles in the brain of postencephalitic Parkinsonism of Economo type. Neurosci Lett 1993; 162:176-178. 95. Ikeda K, Akiyama H, Kondo H et al. Numerous glial fibrillary tangles in oligodendroglia in cases of subacute sclerosing panencephalitis with neurofibrillary tangles. Neurosci Lett 1995; 194:133-135. 96. Constantinidis J. Pick dementia: Anatomoclinical correlations and pathophysiological considerations. In: Rose FC, ed. Modern Approaches to the Dementias, Part I: Etiology and Pathophysiology. Basel: Karger, 1985:72-97. 97. Alzheimer A. Über eigenartige Krankheitsfälle des späteren Alters. Z ges Neurol Psychiat 1911; 4:356-385. 98. Love S, Saitoh T, Quijada S et al. Alz-50, ubiquitin and tau immunoreactivity of neurofibrillary tangles, Pick bodies and Lewy bodies. J Neuropathol Exp Neurol 1988; 47:393-405. 99. Delacourte A, Robitaille Y, Sergeant N et al. Specific pathological tau protein variants characterize Pick’s disease. J Neuropathol Exp Neurol 1996; 55:159-168. 100. Murayama S, Mori H, Ihara Y et al. Immunocytochemical and ultrastructural studies of Pick’s disease. Ann Neurol 1990; 27:394-405. 101. Wisniewski HM, Coblentz JM, Terry RD. Pick’s disease: A clinical and ultrastructural study. Arch Neurol 1972; 26:97-108. 102. Takauchi S, Hosomi M, Marasigan S et al. An ultrastructural study of Pick bodies. Acta Neuropathol 1984; 64:344-348.

Astrocytes in Other Neurodegenerative Diseases

187

103. Scicutella A, Davies P. Marked loss of cerebral galactolipids in Pick’s disease. Ann Neurol 1987; 22:606-609. 104. Yamada T, McGeer PL. Oligodendroglial microtubular masses: An abnormality observed in some human neurodegenerative diseases. Neurosci Lett 1990; 120:163-166. 105. Feany MB, Dickson DW. Neurodegenerative disorders with extensive tau pathology: A comparative study and review. Ann Neurol 1996; 40:139-148. 106. Rebeiz JJ, Kolodny EH, Richardson EP. Corticodentatonigral degeneration with neuronal achromasia. Arch Neurol 1968; 18:20-33. 107. Gibb WRG, Luthert PJ, Marsden CD. Corticobasal degeneration. Brain 1989; 112:1171-1192. 108. Ikeda K, Akiyama H, Iritani S, Corticobasal degeneration with primary progressive aphasia and accentuated cortical lesion in superior temporal gyrus: Case report and review. Acta Neuropathol 1996; 92:534-539. 109. Kertesz A, Hudson L, Mackenzie IR et al. The pathology and nosology of primary progressive aphasia. Neurology 1994; 44:2065-2072. 110. Clark AW, Manz HJ, White CL III et al. Cortical degeneration with swollen chromatolytic neurons: Its relationship to Pick’s disease. J Neuropathol Exp Neurol 1986; 45:268-284. 111. Dickson DW, Yen S-H, Suzuki KI et al. Ballooned neurons in select neurodegenerative diseases contain phosphorylated neurofilament epitopes. Acta Neuropathol 1986; 71:216-223. 112. Horoupian DS, Chu PL. Unusual case of corticobasal degeneration with tau/Gallyas-positive neuronal and glial tangles. Acta Neuropathol 1994; 88:592-598. 113. Mori H, Nishimura M, Namba Y et al. Corticobasal degeneration: A disease with widespread appearance of abnormal tau and neurofibrillary tangles, and its relation to progressive supranuclear palsy. Acta Neuropathol 1994; 88:113-121. 114. Wakabayashi K, Oyanagi K, Makifuchi T et al. Corticobasal degeneration: Etiopathological significance of the cytoskeletal alterations. Acta Neuropathol 1994; 87:545-553. 115. Feany MB, Dickson DW. Widespread cytoskeletal pathology characterizes corticobasal degeneration. Am J Pathol 1995; 146:1388-1396. 116. Takahashi T, Amano N, Hanihara T et al. Corticobasal degeneration: Widespread argentophilic threads and glia in addition to neurofibrillary tangles. Similarities of cytoskeletal abnormalities in corticobasal degeneration and progressive supranuclear palsy. J Neurol Sci 1996; 138:66-77. 117. Bergeron C, Pollanen MS, Weyer L et al. Cortical degeneration in progressive supranuclear palsy. A comparison with cortical-basal ganglionic degeneration. J Neuropathol Exp Neurol 1997; 56:726-734. 118. Buee Scherrer V, Hof PR, Buee L et al. Hyperphosphorylated tau proteins differentiate corticobasal degeneration and Pick’s disease. Acta Neuropathol 1996; 91:351-359. 119. Masliah E, Hansen LA, Quijada S et al. Late onset dementia with argyrophilic grains and subcortical tangles or atypical progressive supranuclear palsy? Ann Neurol 1992; 29:389-396. 120. Mann DMA, South PW, Snowden JS et al. Dementia of frontal lobe type: Neuropathology and immunohistochemistry. J Neurol Neurosurg Psychiatry 1993; 56:605-614. 121. Knopman DS, Mastri AR, Frey F et al. Dementia lacking distinctive histologic features: A common non-Alzheimer degenerative dementia. Neurology 1990; 40:251-256. 122. Foster NL, Wilhelmsen K, Sima AAF et al. Frontotemporal dementia and Parkinsonism linked to chromosome 17: A consensus conference. Ann Neurology 1997; 41:706-715. 123. Sima AA, Defendini R, Keohane C et al. The neuropathology of chromosome 17-linked dementia. Ann Neurol 1996; 39:734-743. 124. Conrad C, Andreadis A, Trojanowski JQ et al. Genetic evidence for the involvement of τ in progressive supranuclear palsy. Ann Neurol 1997; 41:277-281. 125. Lazzarini AM, Golbe LI, Dickson DW et al. Tau intronic polymorphism in Parkinson’s disease and progressive supranuclear palsy. Neurology 1997; 48:A427. 126. Wenning GK, Tison F, Ben Shlomo Y et al. Multiple system atrophy: A review of 203 pathologically proven cases. Movement Dis 1997; 12:133-147. 127. Harding AE. Inherited ataxias. Curr Opinion Neurol 1995; 8:306-309. 128. Paulson HL, Fischbeck KH. Trinucleotide repeats in neurogenetic disorders. Ann Rev Neurosci 1996; 19:79-107.

188

Astrocytes in Brain Aging and Neurodegeneration

129. Gilman S, Sima AA, Junck L et al. Spinocerebellar ataxia type 1 with multiple system degeneration and glial cytoplasmic inclusions. Ann Neurol 1996; 39:241-255. 130. DiFiglia M, Sapp E, Chase KO et al. Aggregation of huntingtin in neuronal intranuclear inclusions and dystrophic neurons in brain. Science 1997; 277:1990-1993. 131. Weidenheim KM, Dickson DW. Intranuclear inclusion bodies in an elderly demented woman: A form of intranuclear inclusion body disease. Clin Neuropathol 1995; 14:93-99. 132. Bradley WG Overview of motor neuron disease: Classification and nomenclature. Clin Neurosci 1995; 3:323-326. 133. Hirano A. Neuropathology of ALS: An overview. Neurology 1996; 47(Suppl 2):S63-66. 134. Mizusawa H, Matsumoto S, Yen SH et al. Focal accumulation of phosphorylated neurofilaments within anterior horn cell in familial amyotrophic lateral sclerosis. Acta Neuropathol 1989; 79:37-43. 135. Murayama S, Ookawa Y, Mori H et al. Immunocytochemical and ultrastructural study of Lewy body-like hyaline inclusions in familial amyotrophic lateral sclerosis. Acta Neuropathol 1989; 78:143-152. 136. Lowe J, Aldridge F, Lennox G et al. Inclusion bodies in motor cortex and brainstem of patients with motor neurone disease are detected by immunocytochemical localization of ubiquitin. Neurosci Lett 1989; 105:7-13. 137. Kato S, Hayashi H, Nakashima K et al. Pathological characterization of astrocytic hyaline inclusions in familial amyotrophic lateral sclerosis. Am J Pathol 1997; 151:611-620. 138. Rosen DR, Siddique T, Patterson D et al. Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 1993; 362:59–62. 139. Bruijn LI, Becher MW, Lee MK et al. ALS-linked SOD1 mutant G85R mediates damage to astrocytes and promotes rapidly progressive disease with SOD1-containing inclusions. Neuron 1997; 18:327-338.

Part III Experimental Models of Astrocyte Senescence: Implications for Neurodegenerative Disease

CHAPTER 9

The Peroxidase-Positive Subcortical Glial System Mark B. Mydlarski, James R. Brawer and Hyman M. Schipper

Introduction

A

subpopulation of astrocytes bearing unique cytoplasmic inclusions which progressively accumulate with advancing age has been described in limbic and periventricular brain regions of all vertebrates thus far examined, including frogs,1 rats,2 mice,3,4 dogs,5 cats,5 monkeys6 and humans.7-10 These cells were initially identified as astrocytes by electron microscopy on the basis of their attenuated cytoplasm, ellipsoidal, euchromatic nuclei and bundles of intermediate filaments.3,11-13 The cytoplasmic granules that distinguish these cells are round to angular in shape, of varying dimensions and intensely osmiophilic (Fig. 9.1). The inclusions are often invested with limiting membranes and occasionally appear contiguous with short, tubular elements filled with material of similar electron density.3,12,14 At the light microscopic level, this glial subpopulation is shown to advantage by dual labeling for endogenous peroxidase activity (see below) and the astrocyte marker, GFAP.

Tinctorial and Histochemical Features These cells are commonly referred to as Gomori-positive or peroxidase-positive astrocytes on the basis of their tinctorial and histochemical characteristics. The cytoplasmic inclusions exhibit metachromasia in toluidine blue-stained sections6 and have an affinity for the Gomori stains, aldehyde fuchsin and chrome alum hematoxylin.15 Gomori stains were originally used to identify pancreatic β cells, and a high sulfur content of β cells (proinsulin disulfide bonds) was thought to account for their Gomoriphilia.16 Histochemical and microprobe analyses have confirmed that the Gomori-positive astrocyte granules are indeed rich in sulfhydryl groups.17,18 However, the Gomori stains are relatively nonspecific and, under certain circumstances, will complex with sulfuric acid esters and with sulfonic, aldehyde, carboxyl and phosphate groups in neuronal and other nonastrocytic substrates.2,6,19-21 Thus, in addition to the glial granules, aldehyde fuchsin stains oxidized neuronal lipofuscin, neuromelanin,19 neuronal dense bodies, neurosecretory material of the hypothalamo-hypophyseal system6 and corpora amylacea.19 On the basis of their tinctorial properties and propensity to accumulate with aging (see below), the Gomoriphilic astrocyte granules were initially regarded as a form of the senescent pigment, lipofuscin,22 or as phagocytosed neurosecretion.21,23 However, both views were challenged in the face of studies demonstrating that: 1. few Gomori-positive astrocyte granules are present in the supraoptic nucleus, where neurons replete with sulfur-rich neurosecretory material (neurophysins) abound;24 Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

192

Astrocytes in Brain Aging and Neurodegeneration

2. in unstained sections viewed under light microscopy, the astrocytic inclusions reveal no visible pigment characteristic of lipofuscin;25 3. under transmission EM, lipofuscin exhibits heterogeneous electron-lucent and dense regions in osmicated preparations,26 whereas osmicated Gomori-positive astrocyte granules tend to be uniformly electron-dense;3,12 4. unlike lipofuscin, the glial granules are not labeled with the conventional lipid markers Sudan Black, Sudan III25 or oil red O;27 and 5. the glial inclusions emit an orange-red autofluorescence (610-640 nm) consistent with the presence of porphyrins24,28 or oxidized flavoproteins29,30 and distinct from the green or yellow-orange autofluorescence (400-545 nm) typically emitted by lipofuscins in situ.28,31 Perhaps most importantly, the Gomori-positive gliosomes are rich in iron,32-34 may contain other transition metals such as copper14 and chromium14 and express the metalbinding protein metallothionein.35 The gliosomes stain intensely with diaminobenzidine (DAB), a marker of endogenous peroxidase activity.36-38 In these cells, DAB staining persists after tissue preheating, at extremes of pH and in the presence of the catalase inhibitor aminotriazole. 9,39 The peroxidase activity is therefore nonenzymatic in nature (pseudoperoxidase) and is most likely mediated by ferrous iron or other redox-active transition metals.40

Topography of the Peroxidase-Positive Astroglia Peroxidase-positive astrocytes are relatively abundant in the subependymal zone throughout the neuraxis and in blood-brain barrier-deficient regions including all of the circumventricular organs.6,17,25,36 In the rat telencephalon, relatively high concentrations of these cells are found in the olfactory bulb, the caudate nucleus adjacent to the lateral ventricle, the putamen-globus pallidus, and the hippocampus.36,41 A highly stratified distribution of these cells was delineated within the dorsal hippocampus of adult rats in a study employing dual histochemical/immunohistochemical labeling to identify peroxidase (DAB)positive cytoplasmic granules within GFAP-positive astrocytes. In this region, numerous DAB-positive astrocytes are confined to the hilus of the dentate gyrus and the lacunosum molecular layer and stratum oriens of subsectors CA1-3. Other hippocampal layers exhibit GFAP-positive astrocytes with little or no detectable DAB reaction product, such as the granule cell and inner molecular layers of the dentate gyrus, and the pyramidal cell layer and stratum radiatum of CA1-3.41 In the diencephalon, peroxidase-positive astroglia are prominent in the arcuate nucleus and ventral premammillary area of the basal hypothalamus, in the third ventricular subependymal zone and in the organum vasculosum of the lamina terminalis. In the mesencephalon, these cells are frequently observed in the periaqueductal gray, dorsal to the raphe nuclear complex and in the superficial aspect of the superior colliculi. Although few DAB-positive astrocytes are seen in the substantia nigra of 3 month old rats, this area becomes heavily populated with these cells by 15 months of age.41a In the rhombencephalon, peroxidase-positive astrocytes are most numerous in the area postrema and occur, to a lesser extent, in the nucleus gracilis, dorsal motor nucleus of the vagus, locus coeruleus, olivary nuclear complex and lateral cerebellum. In the spinal cord, small numbers of these cells have been identified in Rexed’s laminae 1 and 2 of the dorsal horn.36 Further details concerning the topography of these cells within the rat neuraxis are presented in Keefer and Christ,36 Schipper8 and Schipper and Mateescu-Cantuniari.41 In serial sections derived from adult human autopsy material, topographically superimposable chrome alum hematoxylin-positive and DAB-positive astrocytes were found throughout the periventricular forebrain, in the optic tract and globus pallidus and within the diencephalon.7,8 In the latter, these cells appear to be concentrated in the organum

The Peroxidase-Positive Subcortical Glial System

193

Fig. 9.1. (A) Ultrastructure of Gomori-positive astrocyte in the hypothalamic arcuate nucleus of a normal adult female rat. Osmiophilic, Gomori-positive gliosomes (g) and a bundle of intermediate filaments (arrows) are depicted. x17,220 magnification. (B) Gomori-positive astrocyte in the hypothalamic arcuate nucleus of an adult female rat rendered anovulatory with estradiol valerate. This treatment induces a marked accumulation of Gomori-positive cytoplasmic inclusions (g) in close proximity to degenerating dendritic profiles (arrow near bottom). N, astrocyte nucleus; F, intermediate filaments: x17,220. Reprinted with permission from Brawer JR et al, Endocrinol 1978; 103:501-512.

vasculosum of the lamina terminalis, infundibular region and capsule of the mammillary body. As in other vertebrates, DAB staining in the human astrocytes occurred over a wide range of pH and was resistant to tissue preheating and aminotriazole, indicating that staining was due to a nonenzymatic (likely iron-mediated) pseudoperoxidase reaction.

Modulation of the Peroxidase-Positive Glial System Aging Gomori-positive astroglia first appear at the end of the first week in the postnatal rabbit hypothalamus42 and in the fourth week in rats and mice.15,43 In the hypothalamus of rats and mice, numbers of peroxidase-positive astrocytes and their granule content progressively increase between 6 and 14 months of age.44 In female rodents, early ovariectomy significantly attenuates the senescence-dependent proliferation of these gliosomes in the hypothalamic arcuate nucleus (Fig 9.2), indicating that, in this neuroendocrine locus, the accumulation of peroxidase-positive astrocytic inclusions is influenced by exposure to circulating sex hormones (see below).44 Relative to other subcortical brain regions, few peroxidase-positive glial granules are present in the rat substantia nigra at 3 months of age. By 15 months, however, many nigral astroglia are replete with large, DAB-positive inclusions which increase in abundance by a factor of four in comparison with 3 month old animals.41a In humans, numbers of peroxidase-positive astrocyte granules increase in the basal ganglia and throughout the periventricular forebrain between the ages of 3 and 69 years.7 Despite their consistent increase with aging, the results of various histochemical and morphological studies (described above) indicate that the peroxidase-positive granules are a unique form of glial inclusion constitutively different from the aging pigment lipofuscin.

194

Astrocytes in Brain Aging and Neurodegeneration

Fig. 9.2. Effects of aging and long-term gonadectomy on numbers of Gomori-positive astrocytic granules in the arcuate nucleus of male and female rats. (A) Females: Numbers of astrocytic granules increase significantly with advancing age (fine crosshatching). Early ovariectomy markedly attenuates this aging effect (heavy crosshatching). (B) Male rats: The age-related increase in numbers of astrocytic granules in male rats is less robust than in females and early castration in the former does not significantly suppress this aging phenomenon. Reprinted with permission from Schipper HM et al, Biol Reprod 1981; 25:413-419.

Sex Hormones and Reproductive Senescence In adult female rodents, the persistent estrus (PE) state, an anovulatory syndrome characterized by the development of polycystic ovaries and persistent vaginal cornification, spontaneously develops as a function of advancing age.44,45 In young postpubertal female rats, ablation of the preoptic-suprachiasmatic region of the diencephalon or transection of its projections(s) into the medial basal hypothalamus abolishes sexual cyclicity and elicits the PE state.46 The PE state also arises in adult female rats: 1. exposed to continuous illumination;47 2. treated neonatally with systemic or intrahypothalamic testosterone;48,49 and 3. following the intramuscular injection of estradiol valerate.11,50 Concomitant with the induction of premature reproductive failure, constant light treatment50 and multiple12 or single11,51 intramuscular injections of 2 mg EV greatly accelerate the aging-related accumulation of peroxidase-positive astrocyte granules in the hypothalamic arcuate nucleus, a neuroendocrine locus rich in estrogen receptors.12,13,50,51 Furthermore, EV treatment promotes the collapse of dendritic profiles, degeneration of axon terminals, synaptic loss and remodeling, depletion of neuronal b-endorphin, formation of myelin figures and the accumulation of phagocytic microglial cells within the arcuate nucleus.11,12,50,52-55 The degenerating neuronal processes often occur in close proximity to hypertrophic astrocytes exhibiting a massive proliferation of electron-dense (metal-rich) cytoplasmic inclusions (Fig. 9.1). Using numbers of reactive microglia and astrocytic granules as quantitative indices of steroid-related neural damage, we demonstrated that the EV and light-induced pathological changes in the arcuate nucleus are completely abrogated by prior ovariectomy.50 Thus, an ovarian product, and not the EV or constant light treatment per se, is responsible for the progressive development of degenerative changes in the arcuate nuclei of PE rats.50 Repeated monthly injections of EV in male rats50 and tonic, high-physiologic levels of unconjugated estradiol maintained by Silastic implants in gonadectomized

The Peroxidase-Positive Subcortical Glial System

195

Fig. 9.3. Effects of sex hormones on numbers of Gomori-positive astrocytic granules in the arcuate nucleus of castrated female rats. All animals received Silastic implants designed to release high-physiologic levels of steroid hormones or vehicle only (control) for 3 months. (a) control, (b) estradiol-17β, (c) testosterone, (d) dihydrotestosterone, (e) estradiol plus testosterone, (f) estradiol plus dihydrotestosterone. Estradiol induces a massive accumulation of Gomori-positive astrocytic granules. Dihydrotestosterone, and to a lesser extent testosterone, suppress this estradiol effect. Reprinted with permission from Brawer JR et al, Endocrinol 1983; 112:194-199.

females56 produce identical histopathological changes in the arcuate nucleus. In contradistinction to estrogens, the administration of androgens56 or progestins57 fails to elicit similar glial reactions in this brain region (Fig. 9.3). In normal aging female rats and mice, the progressive accumulation of peroxidase-positive astrocyte granules within the arcuate nucleus can be blocked by early gonadectomy.44 Taken together, these observations indicate that aging of the neuroendocrine hypothalamus may be hastened by abnormal patterns of circulating ovarian estradiol resulting from EV or constant light exposure.44,58 Conceivably, estrogen-induced neurodegeneration within the arcuate nucleus compromises the integrity of gonadotropin-regulating neural circuitry in this brain region with ensuing anovulatory sterility.59 Our histomorphological observations are consistent with earlier physiological studies demonstrating that E2 withdrawal by early ovariectomy enables female rats to cycle (young or old) ovarian grafts at very advanced ages relative to sham-operated littermates.45 Along similar lines, corticosterone administration enhances, whereas adrenalectomy attenuates, senescence-dependent gliosis in glucocorticoid receptor-rich regions of the rodent hippocampus.60,61 Thus, during aging, several classes of steroid hormones may render dysfunctional the neural circuitry subserving their own regulation. Metal-mediated peroxidase reactions within arcuate astroglia may play a pivotal role in the development of estrogen-related hypothalamic injury during aging. In various estrogen target tissues, the oxidative metabolism of estrogen incurs the formation of cytotoxic semiquinones and other free radical intermediates.62-65 The mammalian hypothalamus contains estrogen 2/4-hydroxylase and peroxidases which promote the conversion of estradiol to 2- or 4-hydroxyestradiol (catecholestrogen).66-68 Highly reactive semiquinone radicals are formed when catecholestrogens are oxidized further in peroxidase/H2O2-catalyzed reactions. Catecholestrogens may also undergo spontaneous autoxidation, resulting in the generation of semiquinone radicals and reactive oxygen species including H2O2 and superoxide anion.63,64 In peripheral sex steroid target tissues, estrogen-derived semiquinones and reactive oxygen species have been shown to facilitate membrane lipid peroxidation and DNA damage, which may in part account for the teratogenic and carcinogenic effects of estrogen in these tissues.62,65,69,70 In an analogous fashion, glial peroxidase activity in the hypothalamic arcuate nucleus may promote the bioactivation of estrogens and catecholestrogens to

196

Astrocytes in Brain Aging and Neurodegeneration

cytotoxic semiquinone radicals and reactive oxygen species. The latter, in turn, may be directly responsible for the axodendritic pathology and loss of β-endorphin observed in the arcuate nucleus of PE rats. In support of this hypothesis, we demonstrated that dietary supplementation with potent antioxidants such as α-tocopherol55 or the 21-aminosteroid U7438971 blocks the depletion of hypothalamic β-endorphin in EV-treated rats. Further evidence implicating peroxidase-positive astroglia in oxidative neural injury is derived from studies on the metabolism of catecholestrogens and catecholamines by these cells in primary brain cell cultures (see chapter 11).

X-Irradiation and Trauma Increased numbers of Gomori-positive glia have been documented in the rat arcuate nucleus and third ventricular subependymal zone following exposure to cranial X-irradiation.38 The author of that study suggested that the accumulation of peroxidase-positive granules in glia inhabiting periventricular brain regions may contribute to the protection of blood-brain barrier-deficient regions by degrading cytotoxic, blood-borne substances.38 Recent studies have indicated, however, that the metal-rich glial inclusions, rather than conferring protection to the surrounding neuropil, may promote the generation of neurotoxic free radical intermediates (see chapter 11). In 1990, Noble and coworkers72 reported prominent augmentation of GFAP staining and endogenous glial peroxidase activity in rat spinal cord at two weeks following contusion injury. They conjectured that heme-derived compounds ingested from the extracellular space may be the source of endogenous peroxidase activity in the cells. As described below and in chapter 11, nonheme iron sequestered within “stressed” astroglial mitochondria is responsible for nonenzymatic DAB oxidation in these cells.

Peroxidase-Positive Astrocytes in Primary Culture In Vitro “Aging” In the early 1970s, Srebro and Macinska reported that Gomori-positive cells with tinctorial and fluorescent attributes akin to those of periventricular astrocytes in situ are present in cultures of rodent embryonic and human fetal brain tissue.10,34 Gomori-positive glia in hamster and mouse diencephalic explants were first observed on day 14 in vitro and progressively accumulated over the ensuing 2-3 weeks.34 In periventricular brain explants derived from a 6 week old human embryo, Gomori-positive glia appeared after 5 weeks in culture and increased in number thereafter.10 Over the last ten years we have been investigating the structural, histochemical and functional properties of peroxidase-positive astrocytes in dissociated embryonic and neonatal rat brain primary cell cultures.27,41 As in the rat hypothalamus,51 peroxidase-positive inclusions were localized to GFAP-positive astrocytes in culture by combining DAB histochemistry with anti-GFAP immunohistochemistry. The numbers of peroxidase-positive astrocytes and their granule content progressively increase between days 10 and 46 in vitro, consistent with earlier observations in diencephalic explants and in the intact rodent hypothalamus. In contrast with 10 day old cultures where DAB-positive astrocytes represented fewer than 1% of all cells, older cultures exhibited numerous peroxidase-positive granules deposited within the cytoplasm of isolated flat and stellate astroglia and in astrocytes forming confluent monolayers. In unstained preparations, the astrocytic inclusions are invisible and phase-dark under light and phase-contrast microscopy, respectively. As in situ, the astrocytic inclusions are Gomoriphilic, emit an orange-red autofluorescence, and exhibit nonenzymatic peroxidase activity resistant to aminotriazole, tissue preheating or broad pH modification.8,27,73

The Peroxidase-Positive Subcortical Glial System

197

The Cysteamine Model We demonstrated that exposure to the sulfhydryl agent, 2-mercaptoethylamine or cysteamine (CSH; 880 µM in culture medium administered twice weekly from in vitro days 6-18) induces a marked accumulation of peroxidase-positive astrocytes in primary culture. Cystamine, the oxidized disulfide of CSH, generated a similar glial response at relatively low doses (8.8-88 µM) but was pancytotoxic at 880 µM concentrations. Equimolar concentrations of L-cysteine or ethanolamine, which differ from CSH by single functional group modifications, did not stimulate the accumulation of Gomori-positive cytoplasmic inclusions in cultured astroglia.27 The CSH-treated astrocytes exhibit orange-red autofluorescent granules and nonenzymatic peroxidase activity indistinguishable from that of Gomori-positive astrocytes in unstimulated, older cultures and in senescent subcortical brain regions in situ (Fig. 9.4).25,27,28 At the ultrastructural level, 18 day old CSH-treated astrocytes contain numerous membrane-bound cytoplasmic inclusions which are variable in size and round or ovoid in shape. The inclusions consistently exhibit an intensely osmiophilic matrix identical to that observed in senescent subcortical astroglia in situ.11-33 Occasionally, concentric stacks of membrane reminiscent of myelin-like figures or fingerprint bodies are seen at the periphery of larger inclusions (Fig. 9.5). As noted in situ, clusters of the dense inclusions infrequently appear contiguous with cisternal elements filled with a similar electron-opaque substance.3,11,12,33 In nonosmicated material, the DAB reaction product is visualized as a moderately dense, granular precipitate deposited within many, but not all, of the inclusions. Within strongly-labeled inclusions, the DAB reaction product is either dispersed homogeneously throughout the granule matrix or is restricted to discrete intraorganellar compartments. Elemental iron is detected in the inclusions by energy dispersive X-ray microanalysis, and the presence and concentration of the metal correlates closely with the presence and intensity of DAB staining (Fig. 9.6).33 These astrocyte granules exhibit little or no affinity for Prussian blue, a marker of ferric and hemosiderin iron, arguing that ferrous iron is responsible for the nonenzymatic peroxidase activity in these cells.27

Subcellular Precursors of Peroxidase-Positive Astroglial Inclusions Fine structural, cytochemical and X-ray microprobe analyses of cultured neonatal rat astroglia were performed at various time points following CSH exposure in an effort to delineate the subcellular precursors of the peroxidase-positive, cytoplasmic inclusions.74 In CSH-treated astroglia, the earliest morphological changes appear restricted to the mitochondrial compartment. Within six hours of treatment, many (but not all) of the astrocytes contained mitochondria with irregular swollen cristae which often assumed tubular or saccular morphologies. By 24-72 hours, numerous mitochondrial profiles were characterized by double membranes completely devoid of organized cristae encompassing homogeneous, dense matrices (Fig. 9.7). In some cases, multiple concentric stacks of membrane surrounded the mitochondrial matrix for variable distances along its perimeter. Occasional profiles exhibited narrow, tail-like extensions of the double membrane which tended to terminate as small bulbs. Typically, the acristic mitochondrial profiles formed large clusters intermixed with normal-appearing mitochondria. The alterations in mitochondrial morphology at the different CSH exposure intervals were paralleled by changes in elemental composition. Within 6 hours of CSH exposure, many mitochondria with distorted or absent cristae displayed X-ray emission peaks for chromium. These mitochondrial forms contained no detectable iron and were consistently DAB-negative. Within 24-72 hours many acristic mitochondria probed positively for both chromium and iron and exhibited variable DAB reaction product (visualized in nonosmicated preparations). The intensity of DAB staining correlated with the size of the iron peaks further indicating that (ferrous) iron is the likely source of nonenzymatic per-

198

Astrocytes in Brain Aging and Neurodegeneration

Fig. 9.4. Embryonic day 17 rat brain cell cultures (18 days in vitro). (A) Untreated control. DAB stain for endogenous peroxidase activity. Astrocytes devoid of DAB-positive granules are observed. Methyl green counterstain. x292. (B) Effects of cysteamine (880 µM twice weekly in medium from day 6). DAB stain. Astrocytes exhibit a massive accumulation of cytoplasmic peroxidase-positive inclusions. Methyl green counterstain. x292. (C) Untreated embryonic day 17 rat brain cell culture photographed for autofluorescence. Astrocytes exhibit faint or no orange-red autofluorescence. x292. (D) Cysteamine-treated brain cell culture. Cysteamine-induced astrocyte granules and cytoplasm emit intense orange-red autofluorescence. x292. Reprinted with permission from Schipper HM et al, Dev Brain Res 1990; 54:71-79.

Fig. 9.5.(A) Transmission electron microscopy of astrocytes from control culture (unexposed to cysteamine). Segments of four cells are visible. Polysomes, short cisternae of rough endoplasmic reticulum and bundles of intermediate filaments are scattered throughout the cytoplasm. A small Golgi apparatus (arrowheads) and a dense inclusion body (arrow) are depicted. Bar = 0.83 µm. (B) Astrocyte from cysteamine-treated culture. The cytoplasm is replete with large osmiophilic inclusions. Many of the gliosomes exhibit concentric stacks of membrane along segments of their periphery (arrows).Bar = 0.45 µm. Reprinted with permission from McLaren J et al, J Histochem Cytochem 1992; 40:1887-1897.

The Peroxidase-Positive Subcortical Glial System

199

Fig. 9.6.(a) Peroxidase activity in astrocyte from cysteamine-treated culture. Inclusions within this nonosmicated cell exhibit varying degrees of peroxidase activity indicated by the granular DAB reaction product. Inclusions D and E show little DAB reaction product, whereas inclusions C and F are strongly positive. The DAB precipitate in the latter appears localized to specific intraorganellar compartments. Bar = 0.4 µm. (b) X-ray emission spectra derived from cell depicted in (a). Emission peaks indicating the various elemental constituents are labeled. The concentration of a given element is proportional to the area under the peak(s) for that element. The large peak at the right of each histogram indicates copper resulting from the use of copper grids. (A) This emission spectrum was generated by a region of clear cytoplasm. Note the absence of a peak for iron. (B) This spectrum was generated by a euchromatic region of nucleus. Note the absence of a peak for iron. (C) This spectrum was generated by inclusion C in (a). The two iron peaks (arrows) are indicative of a high iron concentration within this inclusion. (D) Spectrum for inclusion D in (a). There is only a single small iron peak (arrow), indicative of a relatively low concentration of iron. (E) Spectrum for inclusion E in (a) The single small iron peak (arrow) indicates a low iron concentration. (F) Spectrum for inclusion F in (a). The twin peaks (arrows) indicate a high concentration of iron. Reprinted with permission from McLaren J et al, J Histochem Cytochem 1992; 40:1887-1897.

200

Astrocytes in Brain Aging and Neurodegeneration

Fig. 9.7. Cysteamine-induced mitochondrial pathology in cultured astroglia. (A) Mitochondrial swelling and dissolution of cristae after 24 h of CSH exposure. Bar = 300 nm. (B) Mitochondrial macroautophagy after 12 days of CSH exposure. Bar = 500 nm. Reprinted with permission from Brawer JR et al, Brain Res 1994; 633:9-20.

oxidase activity in these cells. X-ray microanalysis of peroxidase-positive astroglial granules in several subcortical brain regions of adult rats have revealed the presence of elemental copper rather than iron and chromium.14,14a Copper, a transition metal also capable of promoting pseudoperoxidase reactions,74 was not detected in control and CSH-treated astroglial cultures.75 These observations suggest that various redox-active metals may be sequestered in senescent astroglial mitochondria, and that the relative abundance of a particular metal in these organelles reflects, at least in part, its bioavailability within specific brain regions and neural cell cultures. In support of this notion, rat astroglial mitochondria have been shown to accumulate exogenous iron,76 chromium,77 lead78 and manganese79 following administration of these metals to the culture media. In young adult rats, subcutaneous CSH injections (150-300 mg/kg twice weekly for 3 weeks) elicit striking astrocyte hypertrophy (gliosis) and 2- to 3-fold increases in numbers of peroxidase-positive astrocyte granules in hippocampus, striatum, and other subcortical brain regions related to vehicle-injected controls (Fig. 9.8).80 As in the case of CSH-treated cultures, peroxidase-positive glial granules in the intact rat and human brain invariably exhibit mitochondrial epitopes in immunohistochemical preparations.9,14 In rat brain, Young and co-workers recently demonstrated that immunoreactive acyl-CoA binding protein81 and brain fatty acid binding protein (FABP)82 are most prominent in brain regions replete with peroxidase-positive astrocytes, and that these proteins colocalize significantly to this glial population. In rat hepatocytes, chemically-mediated inhibition of mitochondrial β-fatty acid oxidation results in enhanced expression of these lipid-binding proteins and increased cytoplasmic lipid levels.83 Thus changes in mitochondrial β-oxidation may represent a biochemical lesion corresponding to the pathological alterations in mitochondrial structure observed in senescent subcortical astroglia. Whether or not CSH exposure incurs similar changes in β-oxidation and levels of fatty acid binding proteins in cultures of immature astroglia remains to be determined. Although mitochondria appear to be the fundamental subcellular precursors of the CSH-induced glial granules, other organelles participate in the biogenesis of these inclusions to varying degrees. Progressively over a period of 3-12 days of CSH exposure, many aberrant astroglial mitochondria (incipient peroxidase-positive inclusions) become incor-

The Peroxidase-Positive Subcortical Glial System

201

Fig. 9.8. Effects of subcutaneous CSH administration (see text for treatment regimen) on numbers of DAB-positive astrocyte and ependymal granules in various brain regions. Columns and vertical bars represent means and standard errors of the means, respectively. Asterisks denote statistically significant increases in granule numbers relative to untreated controls (p<0.05). Reprinted with permission from Schipper HM et al, J Neuropathol Exp Neurol 1993; 52:399-410.

porated along with other cytoplasmic elements in apparent autophagosomes (Fig. 9.7). Many of the latter exhibit acid phosphatase activity indicative of the participation of lysosomes.75 Using a panel of FITC-labelled antibodies directed against organelle-specific proteins and laser scanning confocal microscopy, we confirmed partial colocalization of lysosomes, and to a lesser extent early endosomes and rough endoplasmic reticulum, to the red autofluorescent (peroxidase-positive) granules induced in cultured astroglia by CSH exposure.84 We determined that CSH stimulates the differential expression of specific lysosomal hydrolases in cultured astroglia. CSH suppresses cathepsin B mRNA levels and immunoreactive protein, whereas cathepsin D mRNA and protein levels are significantly augmented in these cells. Moreover, cathepsin D but not cathepsin B exhibits robust colocalization to the autofluorescent glial inclusions.85 Furthermore, concordant with our in vitro observations, cathepsin B immunoreactivity is prominent in the hypothalamic ventromedial nucleus, which accumulates few peroxidase-positive glial inclusions during aging, and is relatively inapparent in the heavily-granulated hypothalamic arcuate nucleus. Cathepsin D, on the other hand, is heavily expressed in the aging arcuate nucleus, where it colocalizes to the autofluorescent glial inclusions and exhibits scant immunoreactivity in the adjacent ventromedial nuclear complex.85 These findings further support our contention that the astroglial

202

Astrocytes in Brain Aging and Neurodegeneration

inclusions induced by CSH in vitro are identical to those which naturally accumulate in the aging subcortical brain.

Summary and Conclusions A subpopulation of astrocytes in hippocampus, striatum and other subcortical brain regions accumulates cytoplasmic inclusions with advancing age that are histochemically and morphologically distinct from lipofuscin. The gliosomes exhibit an affinity for Gomori stains, orange-red autofluorescence, and intense nonenzymatic peroxidase activity mediated by iron and other redox-active transition metals. In the hypothalamic arcuate nucleus of adult rats and mice, chronic estrogen exposure induces the proliferation of peroxidase-positive astrocytic inclusions in close proximity to degenerating neuritic processes. The glial peroxidase activity may promote oxidative damage of adjacent neuropil constituents in this brain region by metabolizing catecholestrogens and catecholamines to potentially neurotoxic free radical derivatives. In support of this hypothesis, dietary supplementation with potent antioxidants (α-tocopherol, 21-aminosteroids) prevents estradiol-related depletion of hypothalamic β-endorphin, a marker of estrogen toxicity. The sulfhydryl agent, cysteamine (CSH), induces the accumulation of peroxidase-positive astrocytic inclusions in situ and in primary brain cell cultures. In the latter, electron microprobe analysis in conjunction with diaminobenzidine cytochemistry confirmed that redox-active (likely ferrous) iron is largely responsible for the manifest nonenzymatic peroxidase activity in these cells. In CSH-treated glial cultures and in the aging subcortical brain, the peroxidase-positive glial inclusions are derived from effete, metal-laden mitochondria which become incorporated with cathepsin D-positive lysosomes in a complex autophagic process. Taken together, the morphological and histochemical data indicate that CSH accelerates the appearance of a senescent phenotype in subpopulations of astroglia both in situ and in primary culture. As such, the aminothiol compound may serve as a useful tool to delineate further the structural and biochemical correlates of astroglial aging and facilitate our understanding of inclusion biogenesis in the senescent and degenerating CNS. In the following chapter, cellular mechanisms implicated in the formation of peroxidase-positive astroglial inclusions and the relationship of the latter to human corpora amylacea are reviewed.

Acknowledgments The authors thank Mrs. Kay Berckmans and Mrs. Adrienne Liberman for assistance with the preparation of this manuscript. This work is supported by grants from the Medical Research Council of Canada (HMS.JRB) and the Fonds de la Recherche en Santé du Québec (HMS).

References 1. Srebro Z, Lach H, Krawczyk S et al. Observations on the presence of Gomori-positive cells in the telencephalon of various forms of green frogs. Acta Biologica Cracoviensia 1975; 18:203-212. 2. Wislocki GB, Leduc EH. The cytology of the subcommissural organ, Reissner’s fibre, periventricular glial cells and posterior collicular recess of the rat’s brain. J Comp Neurol 1954; 101:283-309. 3. Srebro Z, Lach H. The ultrastructure of the periventricular glia in the brains of rats and mice. Folia Biologica (Krakow) 1987; 35:131-136. 4. Srebro Z, Cichocki T, Godula J. Peroxidase activity in glia and ependymocytes of the mouse brain. Acta Morphologica Academiae Scientiarum Hungaricae 1971; 19:389-395.

The Peroxidase-Positive Subcortical Glial System

203

5. Diepen R, Engelhardt F, Smith-Agreda V. Uber Ort und Art der Entstehung des Neurosekrets im supraoptico-hypophysaren system bei Hund und Katze. Verh Anat Ges Anat Anz (suppl) 1954; 101:276-286. 6. Creswell GF, Reis DJ, MacLean PD. Aldehyde-fuchsin positive material in brain of squirrel monkey (Saimiri sciureus). Am J Anat 1964; 115:543-558. 7. Schipper HM, Stopa EG, Wheelock T. Gomori-positive periventricular glia in the human brain. Society for Neuroscience, Phoenix, AZ (Abstract 232.7), 1988. 8. Schipper HM. Gomori-positive astrocytes: Biological properties and implications for neurologic and neuroendocrine disorders. Glia 1991; 4:365-377. 9. Schipper HM, Cissé S. Mitochondrial constituents of corpora amylacea and autofluorescent astrocytic inclusions in senescent human brain. Glia 1995; 14:55-64. 10. Srebro Z, Macinska A. Gomori-positive glial cells in in vitro cultures of human fetal brain tissue. Acta Biologica Cracoviensia 1973; 16:147-152. 11. Brawer JR, Naftolin F, Martin J et al. Effects of a single injection of estradiol valerate on the hypothalamic arcuate nucleus and on reproductive function in the female rat. Endocrinol 1978; 103:501-512. 12. Brawer JR, Sonnenschein C. Cytopathological effects of estradiol on the arcuate nucleus of the female rat. A possible mechanism for pituitary tumorigenesis. Am J Anat 1975; 144:57-88. 13. Casanueva F, Cocci D, Locatelli V et al. Defective central nervous system dopaminergic function in rats with estrogen-induced pituitary tumors, as assessed by plasma prolactin concentrations. Endocrinol 1982; 110:590-599. 14. Brawer JR, Stein R, Small L et al. Composition of Gomori-positive inclusions in astrocytes of the hypothalamic arcuate nucleus. Anat Rec 1994; 240:407-415. 15. Koritsanszky S, Vigh B, Aros B. Studies on the Gomori-positive glial cells. I. Changes in the Gomori-positive glial cells in rats of various ages. Acta Biologica Hungaricae 1967; 18:9-19. 16. Gomori F. A differential stain for cell types in the pancreatic islets. Am J Pathol 1939; 15:497-499. 17. Srebro Z, Cichocki T. A system of periventricular glia in brain characterized by large peroxisome-like organelles. Acta Histochem Bd 1971; 41:108-114. 18. Wagner HJ, Pilgrim C. A microprobe analysis of Gomori-positive glial cells in the rat arcuate nucleus. Histochem 1978; 55:147-157. 19. Barden H. Further histochemical studies characterizing the lipofuscin component of human neuromelanin. J Neuropath Exp Neurol 1978; 37:437-451. 20. Barden H. The oxidative generation of sulfonic acid groups in neuro-melanin and lipofuscin in the human brain. J Histochem Cytochem 1984; 32:329-336. 21. Noda H. On the Gomorophil findings other than the neurosecretory system in the observation of the hypothalamo-hypophysial system. Gunma Journal of Medical Sciences 1959; 8:223-232. 22. Röhlich P, Vigh B, Teichmann I et al. Electron microscopic examination of the median eminence of the rat. Acta Biologica Hungaricae 1965; 15:431-457. 23. Löfgreen F. The glial-vascular apparatus in the floor of the infundibular cavity. Acta Universitatis Lundensis 1961; 57:1-18. 24. Cottle MKW, Silver A. Fluorescent granules in the Guinea-pig hypothalamus and their possible relation to neurosecretory substance. Zeitschrist fur Zellforschung, Bd 1970; 103:559-569. 25. Goldgefter L. Studies on the structure and function of Gomori-positive glial cells in the rat hypothalamus. Acta Anat 1976; 95:545-557. 26. Sohal RS, Wolfe LS. Lipofuscin: Characteristics and significance. Progress in Brain Res 1986; 70:171-183. 27. Schipper HM, Scarborough DE, Lechan RM et al. Gomori-positive astrocytes in primary culture: Effects of in vitro age and cysteamine exposure. Dev Brain Res 1990; 54:71-79. 28. Goldgefter L, Schejter AS, Gill D. Structure and microspectrofluorometric studies on glial cells from the periventricular and arcuate nuclei of the rat hypothalamus. Cell Tissue Res 1980; 211:503-510.

204

Astrocytes in Brain Aging and Neurodegeneration

29. Duchen MR, Biscoe TJ. Mitochondrial function in type I cells isolated from arterial chemoreceptors. J Physiol 1992; 450:13-31. 30. Kohler M, Fromter E. Identification of mitochondrial-rich cells in unstained vital preparation of epithelia by autofluorescence. Eur J Physiol 1985; 403:47-49. 31. Udenfriend S. Fluorescence assay in biology and medicine. New York and London: Academic Press, 1962; 293-294. 32. Hill J, Switzer R. The regional distribution and cellular localization of iron in the rat brain. Neurosci 1984; 3:595-603. 33. McLaren J, Brawer JR, Schipper HM. Iron content correlates with peroxidase actvity in cysteamine-induced astroglial organelles. J Histochem Cytochem 1992; 40:1887-1897. 34. Srebro Z, Macinska A. Cytochemical demonstration of ferric iron and fluorescence microscopy observations on Gomori-positive glia grown in vitro. Brain Res 1972; 42:53-58. 35. Young JK, Garvey JS, Huang PC. Glial immunoreactivity for metallothionein in the rat brain. Glia 1991; 4:602-610. 36. Keefer DA, Christ JF. Distribution of endogenous diaminobenzidine-staining cells in the normal rat brain. Brain Res 1976; 116:312-316. 37. Sherlock DA, Field PM, Raisman G. Retrograde transport of horseradish peroxidase in the magnocellular neurosecretory system of the rat. Brain Res 1975; 88:403-414. 38. Srebro Z. Periventricular Gomori-positive glia in brains of X-irradiated rats. Brain Res 1971; 35:463-468. 39. Kumamoto T. Histochemical study on endogenous diaminobenzidine-positive granules in the glia cell of rat brain. Acta Histochem Cytochem 1981; 14:173-185. 40. Goldfischer S, Villaverde H, Forschirm R. The demonstration of acid hydrolase, thermostable-reduced diphosphopyridine nucleotide tetrazolium reductase and peroxidase activities in human lipofuscin pigment granules. J Histochem Cytochem 1966; 14:641-652. 41. Schipper HM, Mateescu-Cantuniari A. Identification of peroxidase-positive astrocytes by combined histochemical and immunolabeling techniques in situ and in cell culture. J Histochem Cytochem 1991; 39:1009-1016. 41a. Schipper HM, Vininsky R, Brull R, Small L, Brawer JR. Astrocyte mitochondria: A substrate for iron deposition in the aging rat substantia nigra. Exp Neurol 1998; in press. 42. Srebro Z, Slebodzinski A. Periventricular Gomori-positive glial cells in the hypothalamus of the rabbit. Folia Biologica (Krakow) 1966; 14:391-395. 43. Maksymowicz K, Srebro Z. Gormori-positive glia in the mouse: Post-natal development and topographical distribution. Folia Biologica (Krakow) 1972; 20:135-146. 44. Schipper HM, Brawer JR, Nelson JF et al. Role of the gonads in the histologic aging of the hypothalamic arcuate nucleus. Biol Reprod 1981; 25:413-419. 45. Aschheim P. Aging in the hypothalamic-hypophyseal ovarian axis in the rat. In: Everitt AV, Burgess JA, eds. Hypothalamus, Pituitary and Aging. Springfield: Charles C. Thomas 1976:376-418. 46. Blake C, Weiner R, Gorski R et al. Secretion of pituitary luteinizing hormone and follicle stimulating hormone in female rats made persistently estrous or diestrous by hypothalamic deafferentation. Endocrinol 1972:90:855-861. 47. Singh, KB, Greenwald GS. Effects of continuous light on the reproductive cycle of the female rat: Induction of ovulation and pituitary gonadotrophins during persistent oestrus. J. Endocrinol 1967; 38:389-394. 48. Hahn DH, McGuire JL. The androgen-sterilized rat: Induction of ovulation and implantation by luteinizing hormone-releasing hormone. Endocrinol 1978; 102:1741-1747. 49. Nadler RD. Intrahypothalamic locus for induction of androgen sterilization in neonatal female rats. Neuroendocrinol 1972; 8:349-357. 50. Brawer JR, Schipper HM, Naftolin F. Ovary-dependent degeneration in the hypothalamic arcuate nucleus. Endocrinol 1980; 107:274-279. 51. Schipper HM, Lechan RM, Reichlin S. Glial peroxidase activity in the hypothalamic arcuate nucleus: Effects of estradiol valerate-induced persistent estrus. Brain Res 1990; 507:200-207.

The Peroxidase-Positive Subcortical Glial System

205

52. Olmos G, Naftolin F, Perez J et al. Synaptic remodeling in the rat arcuate nucleus during the estrous cycle. Neurosci 1989; 32:663-667. 53. Perez J, Naftolin F, Garcia S et al. Sexual differentiation of synaptic connectivity and neuronal plasma membrane in the arcuate nucleus of the rat hypothalamus. Brain Res 1990; 527:116-122. 54. Naftolin F, Garcia-Segura LM, Keefe D et al. Estrogen effects on the synaptology and neural membranes of the rat hypothalamic arcuate nucleus. Biol Reprod 1990; 42:21-28. 55. Desjardins GC, Beaudet A, Schipper HM et al. Vitamin E protects hypothalamic β-endorphin neurons from estradiol neurotoxicity. Endocrinol 1992; 131:2482-2484. 56. Brawer JR, Schipper HM, Robaire B. Effects of long-term androgen and estradiol exposure on the hypothalamus. Endocrinol 1983; 112:194-199. 57. Schipper HM, Piotte M, Brawer JR. Effects of progestins on the estradiol-related accumulation of astrocytic granules in the hypothalamic arcuate nucleus. Brain Res 1990; 527:176-179. 58. Schipper HM. Role of peroxidase-positive astrocytes in estradiol-related hypothalamic damage. In: Fedoroff S, Juurlink B, Doucette, R, eds. Biology and Pathology of Astrocyte-neuron Interactions. New York: Plenum Publishing Corp., 1993:125-139. 59. Brawer JR, Beaudet A, Desjardins GC et al. Pathologic effect of estradiol on the hypothalamus. Biol Reprod 1993; 49:647-652. 60. Landfield PW. An endocrine hypothesis of brain aging and studies on brain-endocrine correlations and monosynaptic physiology during aging. In: Finch CE, Potter DE, Kenny AD, eds. Parkinson’s Disease II. New York: Plenum Publishing Corp., 1978:179-199. 61. Landfield PW, Waymire JC, Lynch G. Hippocampal aging and adrenocorticoids: Quantitative correlations. Sci 1978; 202:1098-1102. 62. Horning EC, Thenot J-P, Helton E. Toxic agents resulting from the oxidative metabolism of steroid hormones and drugs. J Toxicol Environ Health 1978; 4:341-361. 63. Kalyanaraman B, Sealy R, Sivarajah K. An electron spin resonance study of o-semiquinones formed during the enzymatic and autoxidation of catechol estrogens. J Biol Chem 1984; 259:14018-14022. 64. Kalyanaraman B, Felix CC, Sealy RC. Semiquinone anion radicals of catechol(amine)s, catechol estrogens, and their metal ion complexes. Environ Health Perspect 1985; 64:185-198. 65. Liehr JG, Roy D. Free radical generation by redox cycling of estrogens. Free Radic Biol 1990; 8:415-423. 66. Ball P, Knuppen R. Formation of 2- and 4-hydroxy-estrogens by brain, pituitary, and liver of the human fetus. J Clin Endocrinol Metab 1978; 47:732-737. 67. MacLusky NJ, Barnea ER, Clark CR et al. Catechol estrogens and estrogen receptors. In: Merriam GR, Lipsett MB, eds. Catechol Estrogens New York: Raven Press 1983:151-165. 68. Mondschein J, Hersey R, Weisz J. Purification and characterization of estrogen-2/4-hydroxylase activity from rabbit hypothalamus: Peroxidase-mediated catechol estrogen formation. Endocrinol 1986; 119:1105-1112. 69. Jellinck PH, Fletcher R. Peroxidase-catalyzed conjugation of [4-14C]-estradiol with albumin and thiols. Can J Biochem 1970; 48:1192-1198. 70. Metzler M, McLachlan JA. Peroxidase-mediated oxidation, a possible pathway for the metabolic activation of diethylstilbestrol. Biochem Biophys Res Commun 1978; 85:874-884. 71. Schipper HM, Desjardins GC, Beaudet A et al. The 21-aminosteroid antioxidant, U74389F, prevents estradiol-induced depletion of hypothalamic β-endorphin in adult female rats. Brain Res 1994; 652:161-163. 72. Noble LJ, Cortez SC, Ellison JA. Endogenous peroxidatic activity in astrocytes after spinal cord injury. J Comp Neurol 1990; 296:674-685. 73. Cissé S, Schipper HM. Isolation of pseudoperoxidase-positive astrocyte granules from intact rat brain and cysteamine-treated neuroglial cultures. Brain Res 1993; 615:141-146. 74. Stohs SJ, Bagchi D. Oxidative mechanisms in the toxicity of metal ions. Free Radical Biol Med 1995; 18:321-336. 75. Brawer JR, Reichard G, Small L et al. The origin and composition of peroxidase-positive granules in cysteamine-treated astrocytes in culture. Brain Res 1994; 633:9-20.

206

Astrocytes in Brain Aging and Neurodegeneration

76. Wang X, Manganaro F, Schipper HM. A cellular stress model for the sequestration of redox-active glial iron in the aging and degenerating nervous system. J Neurochem 1995; 64:1868-1877. 77. Brawer JR, Small L, Wang X et al. Uptake and subcellular distribution of 15Cr in Gomoripositive astrocytes in primary culture. Neurotoxicol 1995; 16:327-336. 78. Holtzman D, Olson J, de Vries C et al. Lead toxicity in primary cultured cerebral astrocytes and cerebellar granular neurons. Toxicol Appl Pharmacol 1987; 89:211-235. 79. Wedler FC, Vichnin MC, Ley BW et al. Effects of Ca(II) ions on Mn(II) dynamics in chick glia and rat astrocytes—Potential regulation of glutamine synthetase. Neurochem Res 1994; 19:145-151. 80. Schipper HM, Mydlarski MB, Wang X. Cysteamine gliopathy in situ: A cellular stress model for the biogenesis of astrocytic inclusions. J Neuropathol Exp Neurol 1993; 52:399-410. 81. Young JK. Immunoreactivity for diazepam binding inhibitor in Gomori-positive astrocytes. Regulatory Peptides 1994; 50:159-165. 82. Young JK, Baker JH, Muller T. Immunoreactivity for brain-fatty acid binding protein in Gomori-positive astrocytes. Glia 1996; 16:218-226. 83. Vanden Heuvel JP, Sterchele PF, Nesbit, DJ et al. Coordinate induction of acyl-CoA binding protein, fatty acid binding protein and peroxisomal β -oxidation by peroxisome proliferators. Biochem Biophys Acta 1993; 1177:183-190. 84. Schipper HM, Cissé S, Walton PA. Colocalization of organelle-specific proteins to autofluorescent astrocyte granules by laser scanning confocal microscopy. Exp Cell Res 1993; 207:62-67. 85. Chopra VS, Moozar KL, Mehindate K, Schipper HM. A cellular stress model for the differential expression of glial lysosomal cathepsins in the aging nervous system. Exp Neurol 147:221-228.

CHAPTER 10

Astrocyte Granulogenesis and the Cellular Stress Response Mark B. Mydlarski and Hyman M. Schipper

A

host of cellular insults, including sublethal exposure to heat, reactive oxygen species, metal ions, amino acid analogues, denatured proteins and sulfhydryl agents, stimulate both prokaryotic and eukaryotic cells to elaborate a number of highly-conserved stress proteins. The superfamily of stress proteins includes high molecular weight heat shock proteins (HSPs) such as HSP90 and 72, certain low molecular weight peptides (e.g., HSP27) and a group of glucose-regulated proteins (e.g., GRP94). The latter appear to respond to a more restricted range of stimuli such as glucose deprivation and calcium ionophores but not to generalized intracellular oxidative stress. The transcription of heat shock genes is regulated by cis-acting heat shock elements in the promoter regions and trans-acting heat shock factors. In mammalian cells, heat shock factors undergo posttranslational modification after heat shock or exposure to other stressors and thereby acquire DNA-binding capability. In addition to the classic HSPs, genes coding for heme oxygenase-1 (HO-1), ubiquitin and some α-crystallins contain heat shock element consensus sequences and may be upregulated with the former in a concerted cellular stress response. HSPs are thought to protect cells undergoing stress by preventing damage to the translational apparatus, by maintenance of lipid membrane integrity, by accelerating degradation of misfolded or denatured proteins and by obviating deleterious protein aggregation by binding to exposed hydrophobic surfaces. Ubiquitin binds to normal and abnormal short-lived proteins and targets them for ATP-dependent proteolysis. In addition, ubiquitin may exhibit complex interactions with heat shock factors and thereby coordinate transcription of HSP genes. In response to acute cellular stress, induction of HO-1 may also protect cells by catabolizing pro-oxidant metalloporphyrins, such as heme, to bile pigments with free radical-scavenging capabilities. In this chapter, we shall consider the role of the cellular stress response in the biogenesis of astrocytic inclusions and establishment of gliosis in aging and degenerating neural tissues. The chapter is subdivided into four major sections: First, the behavior of neural HSPs under conditions of acute stress is discussed, with emphasis on the participation of astrocytes and the impact of aging on these processes. Next, we review literature implicating various stress proteins in the formation of neural inclusions in the aging and degenerating human CNS. In a penultimate section, the role of the cellular heat shock response in the formation of peroxidase-positive astroglial inclusions is explored in considerable detail. We conclude by providing a cellular stress model for the formation of corpora amylacea (CA) in the aging human brain.

Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

208

Astrocytes in Brain Aging and Neurodegeneration

HSP Expression in Acutely-stressed Neural Tissues: Effects of Aging An understanding of the behavior of neural HSPs under conditions of acute stress, such as that incurred by ischemia or hyperthermia, may shed light on the role(s) of stress proteins in brain aging and neurodegeneration which, on account of their indolent nature, often defy direct, methodical analysis. In brain, elevated demands for oxygen and energy accompany an enhanced sensitivity to metabolic stress compared with other tissues. Heat shock protein (HSP)70c is constitutively expressed to a high degree in the CNS and represents approximately 1% of total axonal protein.1 Furthermore, numerous studies have demonstrated that the CNS responds to a diverse array of insults, including hyperthermia, ischemia, and physical and chemical stressors, by transcriptionally upregulating various HSPs, especially members of the HSP70 family.2-4 In a rat model of global ischemia, although strong immunoreactivity for HSP72 was evidenced in surviving hippocampal CA3 and CA4 pyramidal neurons, the stress protein was also markedly expressed in CA1 neurons which exhibit pronounced cytotoxicity under hypoxic conditions.5,6 Similarly, enhanced accumulation of Ub-protein conjugates in the ischemia-sensitive CA1 region was shown to occur with increasing severity of ischemic insult.7 The authors conjectured that accumulation of ubiquitin (Ub)-protein conjugates may precede, and play a role in, the development of ischemia-related neurotoxicity. Others have cited the role of Ub in developmental apoptosis and suggested that during pathogenic processes this function of Ub is reactivated.8 Hayashi et al7 nonetheless conceded that it was not clear whether the observed ubiquitination of proteins promoted or resulted from neuronal cell death. The enhanced resistance to forebrain ischemic cytotoxicity observed in rats preconditioned with mild, transient heat shock has been attributed to the heat-induced upregulation of stress protein expression.9 Additionally, widespread induction of neural HSP72 following mild ischemia in gerbils correlated with tolerance to subsequent severe ischemic challenge.10 Following hyperthermia, HSP72 immunoreactivity is most prominent in astrocytes and vascular endothelia in situ, and is observed to a lesser extent in neurons.11 Similarly, relative to cultured astrocytes, neurons exhibit an attenuated heat shock response in vitro.11,12 Large numbers of astroglia and ependymal cells upregulate HO-1 expression in situ following thermal stress, whereas the neuronal HO-1 response is relatively muted.13 Furthermore, treatment of rats with chemical depletors of the antioxidant glutathione results in robust induction of HO-1 in astrocytes, ependymal cells, Bergmann glia and leptomeninges, but not in neurons.14 Cultured astrocytes, but not neurons, strongly overexpress HO-1 following exposure to oxidative stress.15 Induction of HO-1 may promote the accumulation of free radical-scavenging bile pigments,16 and thereby fortify the brain’s antioxidant potential. The more pronounced heat shock response exhibited by astrocytes in comparison with neurons may underlie the former’s relative resistance to a host of noxious stimuli including hyperthemia,12 oxidative stress15 and ischemia.17 The aforementioned studies have given rise to the concept of a “hierarchy of vulnerability” of cells in an ischemic territory.17 Brief focal ischemia results in enhanced transcription and elaboration of the HSP70 gene product restricted to hippocampal pyramidal neurons residing in the ischemic region. In contrast, ischemia of longer duration causes upregulation of HSP70 within a rim of astrocytes surrounding the damaged neurons, and apparently results in translational inhibition of the latter. Finally, severe ischemia sufficient to cause hippocampal infarction results in HSP70 overexpression in capillary endothelial cells, but not in neurons or astroglia, within the ischemic core. Based on these observations, it has been suggested that translational blockade resulting from ischemic injury occurs first in neurons, and then in astroglia. Capillary endothelial cells, on the other hand, appear relatively resistant to similar levels of hypoxia, and may be the last cell type in an infarcted area to cease expression of stress proteins. The induction of HSP70 in astrocytes and neu-

Astrocyte Granulogenesis and the Cellular Stress Response

209

rons within the CNS has also been documented in various studies following stereotaxic injection of kainic acid,18 flurothyl-induced status epilepticus,19 cortical stab wounds,20 and spinal cord trauma.21 Several studies have examined the relationship between aging and the ability to mount a cellular stress response in the brain. For example, relative to younger rats, older animals exhibit an age-related deficiency in HSP70 mRNA induction in brain (and other tissues) following heat stress.22 However, the heat-induced increase in colonic temperature of the aged rats in this study was lower than that observed in younger controls. Based on this observation, the authors concluded that impaired upregulation of HSP70 in heat-stressed, older animals resulted from an age-associated decline in heat-generating capacity rather than an age-related failure to mount a heat shock response. Subsequently, Pardue et al23 performed a comparable study with young and old rats which controlled for potential differences in body temperature between the groups. Relative to younger thermal-stressed animals, induction of HSP70 mRNA was again blunted in dentate gyrus granule cells and pyramidal cells of the hippocampus in heat-stressed, older rats. Few studies have examined the normal distribution of HSP expression in the CNS of young versus old, unstressed animals. In one such study, Tytell et al24 assayed retinal HSP70 expression in young and old rats and could not detect any significant age-related differences. However, following hyperthermia, levels of inducible HSP70 in older animals were significantly attenuated relative to younger controls.24,25 There is a well documented reduction in the ability of older organisms to cope with stress and maintain vital homeostatic mechanisms under adverse conditions.26 Age-related impairments in the heat shock response system may contribute to the decreased ability of senescent organisms to mount adaptive responses under stressful conditions.27 It has been suggested that inability of older cells to adequately promote the posttranslational conversion of inactive heat shock factor (HSF) to its oligomeric DNA-binding form underlies this functional deficit.27 A reduced capacity of senescent neural tissues to mount a cytoprotective heat shock response may be a factor predisposing the aging CNS to neurodegeneration.

Stress Protein Expression in the Aging and Degenerating Human Brain A number of studies have examined the distribution of various stress proteins in normal and diseased human neural tissues. Pappolla et al28 demonstrated punctate deposits of Ub-immunoreactivity distributed throughout the white matter of normal aged (≥70 years), but not normal young (≤33 years), human brain. Although ultrastructural evaluation of the Ub deposits was not performed, Ub immunoreactivity appeared to localize to axonal elements by light microscopy. In another study, immunoelectron microscopy detected granular Ub-immunoreactivity within glia and myelin lamellae of white matter.29 Ub-positive dystrophic neurites in cortical areas and axonal spheroids in the substantia nigra and striatum were also found to accumulate with advancing age in normal subjects.29 More recently, an age-dependent accumulation of vacuole-laden, Ub-immunopositive astrocytes within the globus pallidus of normal human brain was demonstrated.30 The authors suggested that this was a normal, age-related effect because the numbers of these astrocytes were not further increased in basal ganglia derived from subjects with Alzheimer’s disease (AD), Parkinson’s disease (PD), multiple system atrophy, or multiple sclerosis. The expression of Ub and various HSPs has been associated with a large number of pathologic structures found in neurodegenerative states. Of these, the neurofibrillary pathology of AD has been most extensively studied. Ub levels in the AD brain are reportedly higher than in nondemented, age-matched controls.31 Several studies revealed that Ub is a component of the neurofibrillary tangles (NFT) and senile plaques characteristic of AD.32,33

210

Astrocytes in Brain Aging and Neurodegeneration

Relative to age-matched controls, HSP72 is dramatically overexpressed in AD brain, and, like Ub, exhibits colocalization to neuritic plaques and NFT.34 HSP27 is similarly overexpressed in AD relative to nondemented controls.35 In AD, HSP27 was immunolocalized to degenerating reactive astrocytes, particularly in areas rich in senile plaques, and exhibited occasional colocalization to NFT.35 In another study, HSP28 (HSP27) was increased relative to controls in the temporal, frontal and parietal lobes of AD, and staining appeared localized to damaged neuronal elements and senile plaques.36 αB-crystallin-immunopositive hypertrophic astrocytes and microglia appeared more numerous in AD in comparison with controls and were concentrated in areas replete with plaques and NFT.37 In the AD brain, HO-1 immunostaining is dramatically increased in cortical and hippocampal neurons and astrocytes relative to age-matched, nondemented controls, and is found in association with neuritic plaques, NFT and corpora amylacea.38,39 Several other neuropathological states are characterized by the accumulation of aberrant, cytoplasmic inclusions in neural cells (see chapter 8). Many of these inclusions, including Lewy bodies in PD, Lewy body-like inclusions in ALS, as well as Pick bodies and ballooned neurons in Pick’s disease, are ubiquitinated and may be associated with other stress proteins. In brains of patients with multiple system atrophy, glial cytoplasmic inclusions accumulate within oligodendroglia and appear to contain ubiquitinated forms of αBcrystallin.40 This finding is reminiscent of an earlier report in which ubiquitinated αB-crystallin was localized to Rosenthal fibers.41 The latter represent eosinophilic, GFAP-containing, HSP27-positive inclusions which accumulate within astrocytic processes in long-standing gliosis, certain cerebellar astrocytomas, and in Alexander’s disease (see chapter 4). Hypoxic encephalopathy during infancy is associated with the accretion of gliofibrillary inclusions, predominantly in white matter. This astrocytic inclusion body appears related to Rosenthal fibers and similarly contains GFAP, Ub and αB-crystallin, but diverges from the latter in the ultrastructural pattern of intermediate filament deposition.42 In certain neurodegenerative states, astrocytes may occasionally exhibit cytopathological changes reminiscent of those typically encountered in affected neuronal populations. For example, gliofibrillary tangles associated with PHF-tau have been reported in white matter astrocytes in subjects with corticobasal ganglionic degeneration and progressive supranuclear palsy (see chapter 8).42 A summary of stress protein-containing neural inclusion bodies is presented in Table 10.1. It should be borne in mind that many of the neurodegenerative changes associated with HSP expression represent end-stage or “graveyard” histopathology. In such cases, there is considerable difficulty resolving cause and effect relationships between HSP induction and the development of specific neuropathologic features. Regarding the ubiquitination of abnormal structures in neurodegenerative states, Wilkinson43 stated that “It is not known whether ubiquitin is present at the time of deposition or only detected late in the process as a reflection of the cells’ attempt to deal with this pathological situation.” Due to the inherent difficulties in studying inclusion biogenesis using human tissues, there is considerable value in the generation of tissue culture and animal models in which the development of stressrelated inclusions can be thoroughly investigated. The work described in the following section was performed with the view of elucidating, in prospective fashion, the role of the cellular stress (heat shock) response in the biogenesis of astroglial inclusions.

A Cellular Stress Model for the Biogenesis of Astroglial Inclusions Cell Stress and Astrocyte Granulation in Primary Culture Using the cysteamine (CSH) model for accelerated astrocyte granulation (chapter 9), we determined that this aminothiol compound induces stress protein expression in astroglia akin to its effects on rat liver.44 Astrocyte cultures exposed to CSH for 6 h exhibit a robust

Astrocyte Granulogenesis and the Cellular Stress Response

211

cellular stress response characterized by the enhanced expression of HO-1 and HSPs 27, 72 and 90, long before increases in numbers of DAB (peroxidase)-positive cytoplasmic granules become apparent at the LM level.45 Furthermore, in comparison to control cultures, CSH-pretreated cells exhibited enhanced resistance to both trypsinization-related cell death and killing by H2O2 exposure, providing physiologic evidence that CSH induces a cellular heat shock response in cultured astroglia. Thus, activation of a cellular stress response precedes, and may be a prerequisite for, the formation of peroxidase-positive astrocyte granules in CSH-treated glial cultures. In this respect, the astrocytic inclusions may represent a type of “stress granule” reminiscent of heat shock granules which have been shown to arise in other cell types following sustained stress.46,47

Cell Stress and Astrocyte Granulation In Situ Further studies were undertaken to determine whether CSH administration to young adult rats accelerates the aging-related accumulation of peroxidase-positive astrocyte granules in situ, and whether this also occurs in the context of a glial cellular stress response.48 We noted that subcutaneous CSH treatment (150-300 mg/kg body weight twice weekly x 3 weeks) produced 2- to 3-fold increases in numbers of DAB-positive astrocyte granules within the dorsal hippocampus, corpus callosum, striatum and the third ventricular subependymal zone. Brain regions normally containing few or no peroxidase-positive glia, such as the cerebellum and cerebral cortex, exhibited little or no response to CSH treatment (see chapter 9, Fig. 9.8). Thus, systemic CSH administration appears to accelerate the appearance of a normal “aging” phenotype in subpopulations of rat astroglia. Furthermore, using dual label immunohistochemistry, we observed that both acute (24 h) and prolonged (3 weeks) CSH exposure induced the expression of HSPs 27, 72 and 90 and GRP94 by GFAP-positive astrocytes residing precisely within those brain regions shown to be susceptible to CSHinduced glial granulation. The upregulation of GRP94 in situ probably represents an indirect effect of CSH (perhaps mediated by perturbations in glucose or calcium homeostasis) because, in contradistinction to its direct stimulatory action on other HSPs, this aminothiol does not appreciably alter patterns of GRP94 expression in primary astrocyte cultures.45 Taken together, our findings indicate that, as in primary astroglial cultures, induction of the cellular heat shock response is a proximal event in the biogenesis of peroxidase-positive astrocyte granules in the aging subcortical brain. In addition to promoting glial stress protein biosynthesis and cytoplasmic granulation, prolonged exposure to CSH elicited robust astrocytic hypertrophy and GFAP expression (gliosis) in the corpus callosum, ventral hippocampal commissure and striatum of adult rats. Akin to the effects of CSH, multiple systemic injections of estradiol valerate (EV) induce overexpression of glial HSPs and the accumulation of peroxidase-positive astroglial granules in estrogen receptor-rich brain regions (see below).49 However, unlike chronic CSH treatment, long term EV exposure did not promote astrogliosis and enhanced GFAP expression in situ. Taken together, these results indicate that the accumulation of peroxidasepositive astrocyte inclusions may occur within the context of, or entirely independently of, classical astrocyte hypertrophy (gliosis). In contrast to the striking effects of chronic CSH exposure on glial morphology and histochemistry in situ, this treatment regimen engendered no overt neuronal pathology at the LM level. Specifically, there was no evidence of neuronal damage, demyelination or lipofuscin accumulation in preparations stained with cresyl violet, hematoxylin-eosin, modified Bielschowsky’s silver method, Luxol fast blue, or periodic acid-Schiff.48 Prolonged CSH exposure did reduce somatostatin immunoreactivity in fiber tracts of the ventral striatum48 as had been reported in earlier short term CSH studies.50 However, these findings may not represent true depletion of somatostatin levels or loss of somatostatin-containing neurons.

DLBD

DLBD

Dystrophic neurites

Cortical Lewy Bodies

limbic & para-limbic cortices; amygdala

CA2/3; basal forebrain; brainstem nuclei

Associated with senile plaques

Aging AD & prion diseases

Dystrophic neurites

NF

NF

Unknown

Unknown

Unknown

Substantia nigra Dorsal column nuclei; substantia nigra; globus pallidus

Aging

Tau, NF

PHF-tau

GFAP

GFAP

Unknown

Structural Protein

Hippocampus

Neuroaxonal dystrophy Aging

Marinesco bodies

Granulovacuolar bodies Aging; AD

NEURONS

White matter

CBG; PSP

Gliofibrillarytangles

Subpial and perivascular regions; white matter White matter

Alexander’s disease

Rosenthal fibers

Subpial and periventricular regions

Location

Gliofibrillary inclusion Infantile hypoxic encephalopathies

Aging; AD

Characteristic Clinical Condition

Corpora amylacea

ASTROCYTES

Inclusion Body

Cell Type

Table 10.1. Stress protein-containing neural inclusion bodies

Ubiquitin; αB-crystallin;

Ubiquitin

Ubiquitin; HO-1b

Ubiquitin

Ubiquitin

Ubiquitin;HO-1c

Ubiquitin

Ubiquitin; αB-crystallin

Ubiquitin; αB-crystallin; HSP27

Ubiquitina; HSP27a72a; HO-1b

Stress Protein

212 Astrocytes in Brain Aging and Neurodegeneration

Cortex; hippocampus

ALS-sporadic and familial

AD;PSP

Pick’s disease

CBG; Dementia with grains

Dementia lacking distinctive histopathology

Lewy body-like inclusions in ALS

Neurofibrillary tangles & neuropil threads

Pick bodies

Neuropil grains

Ubiquitin reactive inclusions

White matter

Glial cytoplasmic inclusions

PHF-tau

Unknown

Unknown

PHF-tau

PHF-tau

PHF-tau

NF

NF

occasionally NF

NF

Ubiquitin; αB-crystallinf

Ubiquitin

Ubiquitin

Ubiquitin

Ubiquitin

Ubiquitin; HO-1b

Ubiquitin; αB-crystallin

Ubiquitin; αB-crystallin

Ubiquitine

Ubiquitin; αB-crystallin; HSP27b

Modified after Dickson and Yen, 1994. Additional data are referenced. Abbreviations: AD, Alzheimer’s disease; ALS, amyotrophic lateral sclerosis; CBG, corticobasal ganglionic degeneration; DLBD, diffuse Lewy body disease; GFAP, glial fibrillary acidic protein; NF, neurofilament; PD, Parkinson’s disease; PHF, paired helical filaments; PSP, progressive supranuclear palsy. References: a, Cissé et al, 1993; b, Schipper et al, 1995; c, Smith et al, 1994; d, Kato et al, 1992; e, Migheli et al, 1994; f, Tamaoka et al, 1995

Multisystem atrophy

Brain and spinal cord myelin

Cortex (small neurons in layer2); dentate fascia

Cortex; hippocampus; basal forebrain; brainstem nuclei

Anterior horn cells

Substantia nigra; brainstem monoaminergic; basal forebrain

Granular Aging degeneration of myelin

OLIGODENDROGLIAL

Hippocampal dentate fascia; cerebral cortex

PD:DLBD

Lewy Bodies

Anterior horn cells

ALS

Bunina bodies

limbic & para-limbic cortices; amygdala

CBG; Pick’s disease

Ballooned Neurons

Astrocyte Granulogenesis and the Cellular Stress Response 213

214

Astrocytes in Brain Aging and Neurodegeneration

Rather, by virtue of disulfide bond interactions, CSH may chemically alter somatostatin molecules, rendering them undetectable by conventional immunohistochemistry.51 The absence of overt neuronal pathology, in conjunction with observations that CSH-induced HSP expression and cytoplasmic granulation are restricted to astrocytes (and certain third ventricular ependymal cells) in situ and in pure neuroglial cultures,45 strongly suggest that this aminothiol compound elicits a primary “gliopathy” independent of any antecedent neuronal injury. The development of neuropathological features, including basal ganglia necrosis and patchy demyelination, has been documented in young patients chronically receiving CSH for the treatment of nephropathic cystinosis.52 However, these degenerative changes appear to derive from the disease process itself, rather than from iatrogenic CSH toxicity, because similar neuropathologic profiles have been observed in rare individuals with nephropathic cystinosis who have survived into the second or third decades without CSH treatment. Whether or not sustained CSH treatment induces astrocyte hypertrophy, HSP elaboration and cytoplasmic granulation in the brains of these patients (as it does in rats) remains to be determined.

Stress Protein Content of Peroxidase-Positive Astrocyte Granules Experiments were performed to ascertain whether HSPs are actual constituents of the peroxidase-positive astrocyte granules in adult rat brain sections and in CSH-treated astrocyte cultures.53 Using fluorescein-conjugated antibodies directed against a host of stress proteins in conjunction with laser scanning confocal microscopy, we determined that HSP27 exhibits intense colocalization to the red autofluorescent (peroxidase-positive) astrocyte inclusions which derive from degenerative mitochondria both in situ and in vitro (Figs. 10.1-10.3). GRP94 exhibited partial colocalization to peripheral regions of the astrocytic inclusions in both preparations. Colocalization of GRP94, an endoplasmic reticulum (ER)-derived stress protein, suggests that elements of the ER may participate in the biogenesis of peroxidase-positive astrocyte granules. Using a panel of antibodies directed against organelle-specific proteins, it was shown that the ER exhibits occasional colocalization to the astrocytic inclusions in CSH-treated cultures.48 Furthermore, ultrastructural studies of these gliosomes in periventricular brain regions of aging and estrogen-treated rodents54,55 and in CSH-treated glial cultures56 revealed infrequent contiguity of the inclusions with electron-dense cisternal elements reminiscent of ER. HO-1, another ER-derived stress protein, also exhibits partial immunolocalization to the autofluorescent astrocyte granules in CSH-treated cultures and in the intact rat brain (Mydlarski and Schipper, unpublished results). In contrast to the aforementioned stress proteins, HSP72 was a minor constituent of the astrocytic inclusions in situ and in culture, and the glial granules appeared consistently devoid of HSP90 and α B-crystallin 53 (Fig. 10.4). Thus, patterns of stress protein immunolocalization to astroglial inclusions in subcortical regions of the adult rat brain and in CSH-treated glial cultures are virtually indistinguishable from each other (Table 10.2). These findings greatly extend previous histochemical and morphological data underscoring the identical origin of the CSH-induced astroglial inclusions and those which spontaneously accumulate in the aging subcortical brain.

Role of Ubiquitin in the Biogenesis of Astroglial Inclusions After 6 days of primary culture, control (untreated) neonatal rat astroglia exhibited weak, diffuse Ub-immunostaining. In contrast, short term (6 h) CSH treatment induced a shift to intense, granular deposition of Ub53 suggestive of the formation of Ub-protein conjugates.57 The induction of Ub and the formation of high molecular weight conjugates is known to occur in a variety of tissues following heat shock or oxidative stress.58-60 Activa-

Astrocyte Granulogenesis and the Cellular Stress Response

215

Table 10.2. Gomori-positive astrocyte granules: Stress protein expression patterns Stress Proteins

HSP27 HSP72 HSP90 GRP94 Ubiquitin αB-crystallin

Astrocyte Granules 12-Week Rat Brain

12-Day CSH-Treated Cultures

intense, larger granules occasional no intense, granule periphery intense, larger granules no

intense, larger granules infrequent no; nuclear translocation intense, granule periphery intense, larger granules no

GRP = glucose-regulated protein; HSP = heat shock protein

tion of the Ub system in our model represents an early event in the biogenesis of the glial inclusions. These results contest the “common view of ubiquitin being involved in [neuropathological] reactions that are both secondary and late”.61 Furthermore, our results demonstrate that the Ub system is activated shortly after CSH exposure (6 h), concomitant with the upregulation of HSPs 27, 72, 90 and HO-1.45 This finding argues against the view that activation of the Ub system is contingent on prior failure of the HSPs in their attempt to renature damaged polypeptides and reconfer normal protein homeostasis.43 As in the case of HSP27, ubiquitin exhibits intense colocalization to the autofluorescent astrocytic inclusions both in CSH-treated glial cultures and in situ (Fig. 10.5).53 Peroxidasepositive astrocyte granules often appear to be delimited by membranes under transmission EM,56 and the larger inclusions are heavily labeled with lysosome-specific markers.48,62 Ubiquitination of mature, autofluorescent inclusions53 is therefore consistent with earlier studies demonstrating the accumulation of free Ub and Ub-protein conjugates within lysosomes,63-65 and contradicts the view that Ub-immunoreactivity in neural and other tissues is restricted to nonmembrane-bound, nonlysosomal inclusions.66 Additional studies will be required to determine whether Ub or other specific HSPs colocalize to aberrant mitochondria prior to autophagy, or if incorporation of the stress proteins into astroglial inclusions occurs during or after lysosomal fusion (see chapter 9). Ub has been associated with a number of senescence and disease-related neural inclusions, including granulovacuolar bodies, Marinesco bodies, granular degeneration of myelin, dystrophic axons and neurites, and corpora amylacea.42,67 Of these inclusions, only corpora amylacea have previously been shown to predominate in astrocytes of the normal aging human brain.68-70 The aforementioned results, along with those of a study71 depicting the presence of Gomori-positive astroglial granules in aging human neural tissues, establish the ubiquitinated, peroxidase-positive glial inclusions as a second, highly consistent biomarker of astrocyte senescence in the mammalian brain. The presence of Ub in these inclusions may provide important clues concerning their mode of formation and chemical constituents. The accumulation of aberrant proteins and protein aggregates within a variety of cytoplasmic inclusions is characteristic of many senescence and disease-related neurodegenerative changes.42 Shang and Taylor72 demonstrated that H2O2-related oxidative stress compromises the Ub conjugation activity of cultured mammalian cells with a resultant reduction in proteolytic activity. Moreover, their results indicated that activation of the Ub system only occurs following removal of the stress and cell recovery. The activities of the Ub-activating and conjugating enzymes, E1 and E2,

216

Astrocytes in Brain Aging and Neurodegeneration

Fig. 10.1. (See color plate 1 for color representation of these figures.) Identification of Gomoripositive astrocytes by DAB-GFAP double label immunohistochemistry. Hypothalamic arcuate nucleus. Long arrow: brown reaction product (endogenous peroxidase activity) within pink (GFAP-positive) astrocyte. Short arrow: astrocytic process replete with endogenous peroxidase activity. Arrowhead: astrocyte largely devoid of DAB-positive inclusions. 40 micron section; x630. Reprinted with permission from Schipper HM et al, Brain Res 1990; 507:200-207. Fig. 10.2. Laser scanning confocal micrograph of adult arcuate nucleus stained with the mitochondrial marker CLSO. Consistent colocalization of the mitochondrial marker (green) to the autofluorescent glial granules (red) produces yellow fluorescence. Bar = 10 µM. Reprinted with permission from Brawer JR et al, Anat Rec 1994; 240:407-415.

Astrocyte Granulogenesis and the Cellular Stress Response

217

Figs. 10.3-10.5 (opposite). Immunolocalization of stress proteins to autofluorescent astrocyte granules in rat brain sections and CSH-treated glial cultures. Fig. 10.3: HSP27 shows intense colocalization (yellow fluorescence) to astrocyte granules in situ (10.3A; empty arrows) and in culture (10.3B; empty arrows). Solid arrows indicate smaller granules devoid of HSP27-immunoreactivity [bars = 100 µm for (A); 10 µm for (B)]. Fig. 10.4. In brain sections (A), strong αB-crystallin staining can be seen in cells along the third ventricular wall (V). Both in situ (A) and in vivo (B), α B-crystallin manifests strong immunolabeling of astroglia but no colocalization to the autofluorescent inclusions (arrows) [bars = 10 µm for (A) and (B)]. Fig. 10.5. Within the subependymal zone of the third ventricle (V), ubiquitin exhibits strong colocalization (yellow fluorescence) to the autofluorescent granules (A), arrows. [Bar = 25 µm]. In vivo, the larger autofluorescent granules are ubiquitinated (B), empty arrows, whereas many smaller granules are not ubiquitin-immunoreactive (solid arrow). Occasional ubiquitin staining of granule-free cytoplasm is shown in (B) (arrowhead) [bar = 10 µm]. Reprinted with permission from Mydlarski MB et al, Brain Res 1993; 627:113-121.

respectively, are dependent on their constituent free thiol groups.73,74 The authors suggested that oxidative modification of these thiol groups may inactivate Ub-dependent proteolysis and contribute to the intracellular accrual of damaged proteins. In a similar fashion, unremitting oxidative stress in the aging and diseased nervous system, and in primary astrocyte cultures repeatedly exposed to CSH (see below), may prevent Ub-dependent degradation of aberrant proteins and promote their accumulation within astrocytic inclusions. Furthermore, Ub-dependent proteolysis and the release of renatured target proteins from most HSPs require ATP hydrolysis.75,76 Although we have not measured ATP concentrations in CSH-treated cells, others have shown that ATP levels in mammalian cells decline rapidly following heat shock.77-79 Thus, ATP depletion, as a direct consequence of cellular stress or resulting from a progressive, aging-related decline in mitochondrial function, could contribute to the accrual of HSP-bound proteins and compromise the availability of free HSPs necessary for the chaperoning of newly-damaged polypeptides. The role of oxidative stress and the Ub system in the accumulation of aberrant proteins and protein aggregates in the aging and CNS is currently an active area of research with important ramifications for the elucidation (and possible treatment) of various human neurodegenerative disorders. The use of CSH-exposed astroglia and other well-characterized models of neural cell senescence should facilitate this line of inquiry by providing opportunities to test salient hypotheses in simplified, but biologically-relevant, contexts.

Estrogen-Related Astrocyte Granulation As described in chapter 9, administration of estrogen to adult female rats accelerates the accretion of peroxidase-positive astrocyte granules in periventricular brain regions expressing sex steroid receptors.54,80-83 We subsequently determined that three monthly intramuscular injections of estradiol valerate (EV; 0.2 or 2.0 mg) elicit the overexpression of HSPs 27, 72, and 90 and augment cytoplasmic granulation in GFAP-positive astrocytes of the arcuate nucleus and third ventricular subependymal zone. In contrast, long term EV treatment induced little or no HSP upregulation or astrocyte granulation in estrogen receptor-deficient brain regions such as the caudate-putamen and corpus callosum (Table 10.3).49 Olazábal and coworkers have previously reported the induction of HSPs 70 and 90 in rodent hypothalamic neurons following estrogen treatment.84-88 Our study, on the other hand, demonstrated estrogen-related upregulation of HSPs in astrocytes. Of significance, short term administration of EV (48 h) induced similar HSP expression, but no concomitant cytoplasmic granulation, in arcuate astrocytes.49 These findings, in conjunction with results derived from CSH- and H2O2-treated astrocyte cultures45,53 and CSH-exposed rats,48 further

6.9 7.8 1.9 14.4

26.3 ± 13.2 ± 19.5 ± 16.1 ±

26.0 16.6 15.4 22.4

Control

16.7 ± 7.6 ± 2.7 ± 16.0 ±

38.2 ± 24.9 9.3 ± 9.4 12.2 ± 8.1 13.1 ± 11.0

EV 0.2 mg

CC

68.2 ± 7.1* 17.2 ± 18.1 37.8 ± 15.5* 23.5 ± 24.4

Ev 0.2 mg

46.8 ± 22.3 15.2 ± 4.8 14.9 ± 9.1 5.8 ± 2.9

EV 2.0 mg

56.2 ± 27.2 23.8 ± 11.7 37.1 ± 6.4* 15.6 ± 18.6

EV 2.0 mg 4.4 3.4 7.9 13.9

33.0 ± 7.4 ± 3.7 ± 8.3 ±

4.5 10.2 3.1 4.1

Control

28.5 ± 4.4 ± 6.1 ± 12.0 ±

Control 14.2* 10.9 9.0* 17.0

36.6 ± 4.6 ± 4.8 ± 9.8 ±

12.7 4.8 6.0 5.6

EV 0.2 mg

CP

58.6 ± 16.8 ± 35.5 ± 24.5 ±

EV 0.2 mg

Peri-III

Abbreviations: ARC, arcuate nucleus; EV, estradiol valerate; CC, corpus callosum; CP, caudate-putamen; HSP, heat shock protein; Peri-III, third periventricular region. * Significantly increased from control values (p<0.05).

HSP27 HSP72 HSP90 GRP94

HSP27 HSP72 HSP90 GRP94

Control

ARC

29.3 5.9* 6.0* 25.2

43.3 ± 2.7 ± 12.9 ± 7.6 ±

8.4 4.0 5.3* 4.0

EV 2.0 mg

55.9 ± 30.6 ± 32.8 ± 25.7 ±

EV 2.0 mg

Table 10.3.Percent of GFAP-positive astrocytes (mean ±SD) per region expressing HSP27, 72, or 90 or GRP94 after long term EV treatment

218 Astrocytes in Brain Aging and Neurodegeneration

Astrocyte Granulogenesis and the Cellular Stress Response

219

indicate that induction of a cellular stress (heat shock) response precedes, and is not a consequence of, the development of iron-laden astroglial inclusions.

Intracellular Oxidative Stress: A “Final Common Pathway” for the Biogenesis of Astrocytic Inclusions Oxidation of CSH in the presence of transition metals generates several pro-oxidant species including H2O2, and the superoxide, hydroxyl and thiyl radicals.87 HO-1, Ub, and the various HSPs induced by CSH in cultured astroglia45,53 are commonly upregulated in response to hyperthermic challenge as well as oxidative stress. In contrast, addition of CSH to the glial monolayers did not enhance the expression of GRP94, a stress protein known to respond to glucose deprivation and calcium ionophores, but not to heat shock or oxidative stress.88,89 Systemic administration of CSH also resulted in overexpression of these redoxsensitive HSPs in GFAP-positive astroglia in situ48 providing further, albeit indirect, evidence of a free radical mechanism of CSH action. The gene coding for manganese superoxide dismutase (MnSOD) is modulated in bacteria and in mammalian cells by oxidative stress.90,91 This mitochondrial enzyme, which catalyzes the dismutation of superoxide anion to H2O2, protects mitochondria from inordinate or inadvertent superoxide radical generation during the normal electron transport process and following exposure to mitochondrial toxins. MnSOD gene expression and activity are significantly enhanced in CSH-treated glial cultures and in the intact diencephalon of rats given subcutaneous injections of CSH relative to vehicle-injected controls.92 Increased MnSOD activity in liver mitochondria derived from aged humans has been proposed as a mechanism whereby senescent tissues cope with an increased oxidative burden.93 Elevated MnSOD levels reported in the substantia nigra of subjects with Parkinson’s disease94 have similarly been interpreted as a response to excessive oxidative challenge in this condition.95 Oxidative stress is thus a likely mediator of increased MnSOD gene transcription and enzymatic activity observed in CSH-exposed astroglia.92 As described above, EV-related astrocyte granulation in the hypothalamic arcuate nucleus occurs in the context of an antecedent cellular stress response.52 We conjectured that, analogously to the action of CSH, estradiol may promote a concerted HSP response in ARC astroglia via the generation of pro-oxidant intermediates. The mammalian hypothalamus contains estrogen 2/4-hydroxylases which catalyze the conversion of estradiol to 2- and 4-hydroxyestradiol (catecholestrogens).96,97 Subsequent peroxidase-catalyzed reactions transform catecholestrogens to highly reactive semiquinone radicals.98 Spontaneous autoxidation of catechol groups may, additionally, generate pro-oxidant species including H2O2 and superoxide anion.98 Thus, although native estradiol-17β has been shown under certain circumstances to possess antioxidant properties,99 catecholestrogen-derived free radical species may mediate a cellular stress response and the accumulation of iron-rich cytoplasmic granules in subpopulations of hypothalamic astroglia. A vicious cycle may then ensue whereby iron-mediated oxidation of catechol moieties within astroglia produces oxyradicals which further stimulate HSP overexpression and the formation of redox-active cytoplasmic inclusions. In summary, prolonged or repeated exposure to oxidative stress may be the “final common pathway” responsible for activation of the cellular stress (heat shock) response and subsequent biogenesis of peroxidase-positive astroglial inclusions in vitro and in the intact aging nervous system. This hypothesis is supported by the fact that X-irradiation, a known generator of intracellular free radical intermediates, increases numbers of peroxidase-positive glial granules in the rat hypothalamus in a dose-dependent manner.100 Direct evidence implicating oxidative stress in the generation of peroxidase-positive astroglial inclusions is presented in the following section.

220

Astrocytes in Brain Aging and Neurodegeneration

Pro-oxidant Effects of CSH on Astroglial Mitochondria Effete, iron-laden mitochondria are the primary subcellular precursors of peroxidasepositive cytoplasmic inclusions in CSH-exposed astroglial cultures, and these glial inclusions invariably exhibit mitochondrial epitopes in the intact aging rat and human brain (chapter 9). We determined that mitochondrial distension and disorganization of cristae are the earliest morphological changes visible in cultured rat astroglia by transmission electron microscopy following CSH exposure (chapter 9). Mitochondrial swelling is acknowledged to be an important bio-marker of intracellular oxidative stress101 and it occurs with aging,102 following heat shock,103 and under conditions of increased osmotic pressure due to compromised membrane integrity.104 Direct evidence of oxidative damage in biological systems can be obtained by measuring the accumulation of oxidized proteins, lipids, and nucleic acids in whole cells and in various subcellular compartments.105 Significant increases in lipid peroxide levels and oxidative DNA lesions have been amply documented in mammalian mitochondria as a function of advancing age.93,106,107 Of all subcellular compartments, mitochondria normally represent the most abundant source of endogenous prooxidants.108 In young tissues, free radicals generated in the inner mitochondrial membrane by step-wise reduction of molecular oxygen during oxidative phosphorylation are normally tightly bound to the cytochromes of the electron transport chain and produce relatively little oxidative damage. In aging cells, on the other hand, fidelity of electron transport is progressively compromised, resulting in increased oxidative damage to the mitochondrial membranes, to mitochondrial DNA and other cellular constituents. The mutated mitochondrial genome, in turn, codes for aberrant electron transport chain proteins, resulting in a vicious spiral of further free radical “leakage” and oxidative injury.108 In addition, various hemoproteins constituting the mitochondrial electron transport system sustain thiol oxidation reactions with concomitant generation of cytotoxic pro-oxidant species.86 In light of the above, we determined whether oxidative stress is an important mechanism mediating CSH-related injury to isolated astroglial mitochondria (granule precursors).92 Administration of CSH (600-1000 µM) to purified mitochondrial suspensions derived from cultured rat astroglia resulted in significant mitochondrial lipid peroxidation relative to non-CSH-treated (control) preparations. Conceivably, abundant mitochondrial heme ferrous iron sustains the autoxidation of CSH (to cystamine) with concurrent generation of reactive oxygen species.87 The latter, in turn, may initiate or exacerbate oxidative damage to the mitochondrial compartment. These observations are consistent with earlier data indicating that inhibition of CSH autoxidation prevents CSH-induced lipid peroxidation of rat liver mitochondria.109 Addition of catalase significantly attenuates CSH-related mitochondrial lipid peroxidation, suggesting that H2O2 plays an important role in the transformation of normal astroglial mitochondria to metal-laden cytoplasmic inclusions in vitro and in astrocytes of the aging subcortical brain. Indeed, prolonged H2O2 exposure promotes the accumulation of (mitochondria-derived) peroxidase-positive inclusions in cultured rat astroglia akin to the effects of CSH.45

CSH (Paradoxically) Confers Cytoprotection to Astroglia Concomitant with Mitochondrial Injury Although purified astroglial mitochondria exhibited enhanced sensitivity to lipid peroxidation in the presence of CSH,92 whole cell lysates derived from CSH-treated astroglial cultures consistently manifested lower levels of lipid peroxidation relative to untreated controls.83 This latter observation is consistent with the observation that CSH reduces lipid peroxidation in rat liver microsomes in vitro.110 The apparent discrepancy in the behavior of CSH is reconciled by the fact that aminothiols may act either as oxidizing or reducing agents depending on the redox status of their microenvironments.87,111 Redox potentials of

Astrocyte Granulogenesis and the Cellular Stress Response

221

different subcellular compartments have been shown to vary within a given cell.112 In cellular compartments containing relatively low amounts of redox-active transition metals, the antioxidant properties of CSH may prevail over its oxyradical-generating capacity. Indeed, CSH has been used for many years in experimental oncology to protect against excessive radiation-induced tissue damage.113 CSH may confer cytoprotection under these circumstances by direct scavenging of reactive oxygen species111,114 or by chelation of catalytic transition metals.109,115 We observed that short term (6 hr) exposure of cultured astroglia to CSH confers enhanced resistance to mechanoenzymatic stress (trypsinization) in comparison with untreated controls. Prolonged (12 day) CSH treatment additionally protects cultured astrocytes from subsequent H2O2 toxicity relative to control preparations (Fig. 10.6).45 Since glial CSH is no longer detectable by HPLC at the end of a 24 h washout period preceding the H2O2 challenge,45 it would seem that the aminothiol does not serve as a direct protectant in the cytotoxicity assays. It is far more likely that the upregulation of MnSOD gene expression and the elaboration of various stress proteins in cultured astroglia following CSH exposure (vide supra) are important mediators of this glioprotective effect. The augmented MnSOD activity may curtail the accumulation of cytotoxic superoxide anion, while the HSPs may function to prevent deleterious aggregation of unfolded or aberrant proteins and protect lipid membranes and the translational apparatus from stress-induced damage.116-119 Furthermore, rapid activation of Ub in CSH-exposed astroglia may facilitate glial recovery and survival by assisting in the degradation of denatured protein complexes and the re-establishment of normal protein homeostasis. Finally, the enhanced HO-1 activity may confer some degree of cytoprotection by augmenting the degradation of pro-oxidant metalloporphyrins (heme) to antioxidant bile pigments,16 thereby promoting the restoration of a more favorable redox microenvironment. That similar glioprotective responses may be manifest in situ is supported by the observations that: 1. glial cultures are easier to establish when the cells are harvested from postmortem Alzheimer brain than from age-matched, nondemented controls (A. LeBlanc, personal communication); and 2. astrocytes derived from rodent hippocampi previously lesioned (“stressed”) with ibotenic acid exhibit enhanced survival and proliferation in vitro relative to those procured from nonlesioned controls.120 As alluded to throughout this volume, astrocyte hypertrophy, GFAP biosynthesis, and possibly astroglial hyperplasia (reactive gliosis) are fundamental pathological features of virtually all major human neurodegenerative disorders and occur, albeit to a lesser extent, in the course of normal brain senescence. By promoting glial survival under these conditions, stress-induced upregulation of cytoprotective mechanisms in astrocytes (simulated by CSH exposure) may facilitate the establishment of gliosis and the accumulation of intracellular inclusions in the face of concomitant neuronal depletion.

Astrocyte Senescence and the Origin of Corpora Amylacea Corpora amylacea (CA) are glycoproteinacious, cytoplasmic inclusions that accumulate in subpial and periventricular regions of human brain in the course of normal aging. Numbers of CA are reportedly increased in AD121,122 and other neurodegenerative conditions123-126 relative to age-matched, normal controls. CA are most frequently encountered within astroglia or as extracellular deposits.68,70,127 CA may occasionally arise within neuritic processes68,128-131 and they have also been reported in a variety of nonneural tissues.132-134 Many of the tinctorial and histochemical properties of CA have been delineated (see chapter 4). Yet, their subcellular origin and the mechanism(s) responsible for their biogenesis remain enigmatic. Human CA share many topographical, histochemical and antigenic features with the peroxidase-positive astrocytic granules considered in this and the preceding

222

Astrocytes in Brain Aging and Neurodegeneration

Fig. 10.6. Automated MTT cell viability assay depicts cytotoxic effects of trypsinization and H2O2 exposure on fixed numbers (40,000 cells per well) of control and CSHpretreated astrocytes. Optical density correlates directly with cell viability. (A) Long term CSH exposure (880 µM DIV 6-18). CSH-pretreated cells (♦) exhibit increased resistance to mechanical trauma and H2O2 exposure relative to controls ( ). (B) Short term CSH exposure (880 µM x 6 h). As in (A), CSH-pretreated cells (♦) exhibit robust resistance to mechanical trauma relative to controls ( ). However, normalizing for the effects of mechanical stress, both groups show similar declines in cell viability with increasing H2O2 concentrations. An asterisk denotes a significant difference from control values (p<0.050 by Student-Newman-Keuls post hoc test). An open star denotes first significant decline in cell viability relative to respective conditions at 0 µM H2O2 (p<0.050 by Student-Newman-Keuls). Reprinted with permission from Mydlarski MB et al, J Neurochem 1993; 61:1755-1765.

chapter. Both types of inclusion predominate in periventricular brain regions,101,135,136 progressively accumulate with aging,81,120,137 and exhibit affinities for periodic acid-Schiff (PAS) and Gomori’s chrome alum hematoxylin stains,138 metachromasia with toluidine blue,138,139 and nonenzymatic peroxidase activity.82,135,138 In addition, both CA and the peroxidasepositive glial granules are ubiquitinated and contain several heat shock proteins.53,71,67,140 We recently demonstrated consistent immunolocalization of two mitochondrial proteins, sulfite oxidase141 and HSP60,142,143 to the red autofluorescent (peroxidase-positive) astrocyte granules and CA in subependymal regions of senescent and Alzheimer-diseased human brain, as well as in smears of purified human CA.71 In addition, CA in situ and in the purified fractions were noted to contain nucleic acids on the basis of anti-DNA staining.

Astrocyte Granulogenesis and the Cellular Stress Response

223

That these nucleic acids are of mitochondrial origin is strongly suggested by the robust colocalization of sulfite oxidase and DNA within CA in dual-labeled preparations. These data are consistent with earlier electron micrographs depicting damaged mitochondrial components within CA of the optic nerve144 and within “granular glycogen bodies” in senescent human astrocytes.145 Thus far, the only major divergent histochemical feature between peroxidase-positive astrocyte granules and CA is the presence of orange-red autofluorescence in the former and the absence of endogenous fluorescence in the latter. On the basis of the evidence available to date, we submit that: 1. the peroxidase-positive astrocyte granules may be structural precursors of CA in senescent human brain; and 2. degenerate mitochondria within periventricular astrocytes are a major source of autofluorescent cytoplasmic inclusions and CA in the aging human brain. Conceivably, during the putative maturation of peroxidase-positive granules to CA, progressive glycation of autofluorescent mitochondrial substrates may be responsible for the quenching of endogenous fluorescence in the larger inclusions.71 The histochemistry and morphology of CA and Gomori (peroxidase)-positive astrocyte granules are summarized in Table 10.4. As described above, CSH-derived free radical intermediates stimulate the transformation of normal mitochondria to peroxidase-positive, autofluorescent inclusions in primary astrocyte cultures62,146 and in the intact rat brain.48,147 More recently, we observed that long term (90 day) exposure of neonatal rat glial cultures to CSH148 and subcutaneous administration of CSH to adult albino rats148a results in the formation of large spherical, PAS-positive astrocytic inclusions which are highly reminiscent of, if not identical to, human CA. As in the case of human CA, the CSH-induced CA-like inclusions lack endogenous fluorescence and exhibit nonenzymatic peroxidase activity and consistent immunostaining for the mitochondrial protein sulfite oxidase (Fig. 10.7).148,148a These findings further support our contention that mitochondrial damage and autophagy play an important role in the biogenesis of CA (and peroxidase-positive granules) in astrocytes of the aging periventricular brain. Taken together with observations reported in chapter 9, the data reviewed herein suggest a mechanism for the biogenesis of CA in aging astrocytes whereby sustained or repeated intracellular oxidative stress serves as a “final common pathway” mediating the following sequence of events (Fig. 10.8).148 Stage 1: Mitochondria swell, become autofluorescent, and sequester redox-active iron (nonenzymatic peroxidase activity).56,62 The astrocytes undergo a cellular stress reaction as evidenced by upregulation of HSP27, 72 and 90, ubiquitin, and HO-1.45,53,135 Stage 2: The abnormal mitochondria fuse with lysosomes, undergo macroautophagy and incorporate HSP27 and ubiquitin (formation of Gomori-positive “stress granules”). Other stress proteins exhibit partial or no colocalization to the autofluorescent inclusions and remain largely confined to granule-free cytoplasm (HSP72, HO-1) or undergo translocation to the nucleus (HSP90).48,53 Stage 3: Granule constituents (proteins) become glycosylated,149,150 with quenching of autofluorescence and displacement of mitochondrial components to the inclusion periphery (nascent CA). It is conceivable that during the formation of CA, progressive glycosylation of damaged astrocyte mitochondria may abrogate free radical generation by iron-containing mitochondrial proteins. The glycosylation of redox-active mitochondria may thus represent a protective mechanism which serves to limit oxidative injury within the aging nervous system. Stage 4: Many stressed astroglia eventually degenerate and mature, residual CA are deposited in the extracellular space. In accord with this model, excessive oxidative stress reported in the brains of Alzheimer subjects39,151-154 may exacerbate senescence-related injury

Yes

0.5-10 µm

Yes

Accumulation with age

Yes ? Yes

Yes (red)

No

Yes Yes Yes

Yes (red)

Yes 3-10.5 No No

Visible pigment

Inductions Cysteamine Estrogen X-irradiation

Autofluorescence

Peroxidase activity pH range AT inhibition Heat inactivation

Yes 3-11 No No

No

Pleom, round

0.5-10 µm

Pleom, round

Size

Shape

—

PV, limbic

Distribution

Astrocytes (ependyma)

Astrocytes (ependyma)

Cultures GAI

Cell origin

Sections GAI

Rat Brain

Yes 3-11 No No

Yes (red)

? ? ?

No

Pleom, round

0.5-10 µm

Yes

PV, limbic

Astrocytes (ependyma)

Sections GAI

Yes 3-10.5 No No

No

? ? ?

No

Round

10-50 µm

Yes

PV, limbic, subpial

Astrocytes and neurons(?ependyma)

Sections CA

Human Brain

Yes 3-10.5 No No

No

? ? ?

No

Round

10-50 µm

—

—

Astrocytes and neurons (?ependyma)

Cell Fractions CA

Table 10.4. Histochemistry and morphology of CA and Gomori-positive astrocyte inclusions in human and rat brain tissuesa

224 Astrocytes in Brain Aging and Neurodegeneration

Yes

Mitochondrial Antigens

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

?

?

Yes

?

Yes

Yes

Yes (larger)

Yes

Yes

Yes

Yes

?

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

?

?

?

Yes

Yes

Yes

aAT, aminotriazole; CAH, chrome alum hematoxylin; GAI, Gomori-poitive astrocyte inclusions; PAS, periodic acid Schiff; pleom, pleomorphic; PV, periventricular.

Yes

Ubiquitin

Yes

Yes

Yes

Yes

Iron-rich

Sulfur

Yes

Toluidine blue metachromasia

Heat shock proteins

Yes

Yes

Argyrophilia

Yes

Yes

Yes

Yes

CAH

Yes (larger)

Yes (larger)

DNA

PAS

Astrocyte Granulogenesis and the Cellular Stress Response 225

226

Astrocytes in Brain Aging and Neurodegeneration

Astrocyte Granulogenesis and the Cellular Stress Response

227

Fig. 10.7 (opposite-See color plate 2 for color representation of these figures.). Experimental induction of astrocyte corpora amylacea (CA). (A) PAS staining of control astroglial monolayer. No PAS-positive cytoplasmic inclusions are visible. (B) CSH-treated astrocyte culture. Long term CSH exposure induces the accumulation of large, spherical PAS-positive cytoplasmic inclusions. Some of the inclusions are intensely and homogeneously PAS-positive, whereas others exhibit faintly stained peripheral rims. In addition, CSH-treated astrocytes often contain much smaller, PAS-positive cytoplasmic granules (arrowhead) which are rarely encountered in control preparations. The large, PAS-positive inclusions observed in CSH-treated astrocyte cultures are morphologically similar, if not identical, to PAS-positive CA isolated from senescent human brain (C). Bar = 25 µm. (D) DAB staining of CSH-treated astrocyte culture. The CSH-induced inclusions are DAB-positive, indicative of endogenous peroxidase activity (arrows). In the preparations doubly stained for PAS and DAB (insert), PAS-positive (arrowheads) and DAB-positive (arrow) inclusions are occasionally encountered within the cytoplasm of individual cells. Bar = 25 µm. (E)-(G) Confocal microscopic images of CSH-treated astrocyte culture double labeled with PAS (emits red fluorescence) and FITC-tagged anti-CLSO antibody (green fluorescence). Many CA-like inclusions emit homogenous yellow fluorescence (E), indicating robust colocalization of PAS and the mitochondrial marker. Some inclusions exhibit a finely stippled pattern of red and yellow fluorescence in the inclusion periphery (F) or throughout the entire structure (G), indicating partial colocalizaton of the two markers. Bars = 25 µm. Reprinted with permission from Cissé S et al, Neuropathol Appl Neurobiol 1995; 21:423-431.

Fig. 10.8. A model for the biogenesis of corpora amylacea in senescent astroglia. Fe, Iron. G, Glycosylation. HO-1, Heme oxygenase-1. HSP, Heat shock protein. L, Lysosome. M, Mitochondria. Ub, Ubiquitin. Sun symbol, Autofluorescence. Reprinted with permission from Cissé S et al, Neuropathol Appl Neurobiol 1995; 21:423-431.

228

Astrocytes in Brain Aging and Neurodegeneration

to astrocyte mitochondria which, in turn, gives rise to the plethora of CA reported in this condition. Our mitochondrial hypothesis of CA biogenesis does not preclude the formation of these inclusions within nonastrocytic substrates. The relative preponderance of CA in senescent astroglia may be due to the sheer abundance of these cells as well as to their unique metabolic properties. In contradistinction to neurons, astroglia exposed to mitochondrial toxins exhibit long term survival by converting to robust anaerobic metabolism.155 This could allow sufficient time for the gradual transformation of damaged mitochondria to CA in these cells. Conversely, CA may be less often encountered in neuronal processes and other nonastrocytic substrates, because in these tissues the toxicity “window” permitting both sufficient mitochondrial injury and sustained cell viability may be relatively narrow.148 The data reviewed in chapters 9 and 10 of this volume provide, in our estimation, compelling evidence that exposure to the simple aminothiol compound CSH fully recapitulates many of the morphological and biochemical changes incurred by populations of subcortical astroglia as these cells naturally age. As such, CSH-treated astroglia should continue to serve as a useful model to delineate further the role of the cellular stress (heat shock) response in the biogenesis of glial inclusions and the establishment of reactive gliosis in the aging and degenerating CNS.

Acknowledgments The authors thank Mrs. Kay Berckmans and Mrs. Adrienne Liberman for assistance with the preparation of this manuscript. This work is supported by grants from the Medical Research Council of Canada (HMS.JRB) and the Fonds de la Recherche en Santé du Québec (HMS).

References 1. De Waegh S, Brady ST. Axonal transport of a clathrin uncoating ATPase (HSC70): A role for HSC70 in the modulation of coated vesicle assembly in vivo. J Neurosci Res 1989; 23:433-440. 2. Brown IR. Induction of heat shock (stress) genes in the mammalian brain by hyperthermia and other traumatic events: A current perspective. J Neurosci Res 1990; 27:247-255. 3. Dienel GA, Kiessling M, Jacewicz M, Pulsinelli W. Synthesis of heat shock proteins in rat brain cortex after transient ischemia. J Cereb Blood Flow Metab 1986; 6:505-510. 4. Tytell M, Barbe MF, Brown IR. Stress (heat shock) protein accumulation in the central nervous system. Adv Neurol 1993; 59:293-303. 5. Chopp M, Li Y, Dereski MO et al. Neuronal injury and expression of 72-kDa heat-shock protein after forebrain ischemia in the rat. Acta Neuropathol 1991; 83:66-71. 6. Simon RP, Cho H, Gwinn R, Lowenstein DH. The temporal profile of 72-kDa heat-shock protein expression following global ischemia. J Neurosci 1991; 11:881-889. 7. Hayashi T, Takada K, Matsuda M. Subcellular distribution of ubiquitin-protein conjugates in the hippocampus following transient ischemia. J Neurosci Res 1992; 31:561-564. 8. Mayer RJ, Arnold J, Laszlo L et al. Ubiquitin in health and disease. Biochim Biophys Acta 1991; 1089:141-157. 9. Chopp M, Chen H, Ho KL et al. Transient hyperthermia protects against subsequent forebrain ischemic cell damage in the rat. Neurol 1989; 39:1396-1398. 10. Kitagawa K, Matsumoto M, Kuwabara K et al. Ischemia tolerance phenomenon detected in various brain regions. Brain Res 1991; 561:203-211. 11. Marini AM, Kozuka M, Lipsky RL, Nowak TS Jr. 70-Kilodalton heat shock protein induction in cerebellar astrocytes and cerebellar granule cells in vitro: Comparison with immunocytochemical localization after hyperthermia in vivo. J Neurochem 1990; 54:1509-1516. 12. Nishimura RN, Dwyer BE, Clegg K et al. Comparison of the heat shock response in cultured cortical neurons and astrocytes. Mol Brain Res 1991; 9:39-45.

Astrocyte Granulogenesis and the Cellular Stress Response

229

13. Ewing JF, Maines MD. Rapid induction of heme oxygenase 1 mRNA and protein by hyperthermia in rat brain: Heme oxygenase 2 is not a heat shock protein. Proc Natl Acad Sci USA 1991; 88:5364-5368. 14. Ewing JF, Maines MD. Glutathione depletion induces heme oxygenase-1 (HSP32) mRNA and protein in rat brain. J. Neurochem 1993; 60:1512-1519. 15. Dwyer BE, Nishimura RN, Lu SY. Differential expression of heme oxygenase-1 in cultured cortical neurons and astrocytes determined by the aid of a new heme oxygenase antibody. Response to oxidative stress. Mol Brain Res 1995; 30:37-47. 16. Stocker R. Induction of haem oxygenase as a defense against oxidative stress. Free Radic Res Comm 1990; 9:101-112. 17. Sharp FR, Sagar SM. Alterations in gene expression as an index of neuronal injury: Heat shock and the immediate early gene response. Neurotoxicol 1994; 15:51-60. 18. Uney JB, Leigh PN, Marsden CD et al. Stereotaxic injection of kainic acid into the striatum of rats induces synthesis of mRNA for heat shock protein 70. FEBS Lett 1988; 235:215-218. 19. Lowenstein DH, Simon RP, Sharp FR. The pattern of 72-kDa heat shock protein-like immunoreactivity in the rat brain following flurothyl-induced status epilepticus. Brain Res 1990; 531:173-182. 20. Brown IR, Rush S, Ivy GO. Induction of a heat shock gene at the site of tissue injury in the rat brain. Neuron 1989; 2:1559-1564. 21. Gower DJ, Hollman C, Lee S, Tytell M. Spinal cord injury and the stress protein response. J Neurosurg 1989; 70:605-611. 22. Blake MJ, Fargnoli J, Gershon D, Holbrook NJ. Concomitant decline in heat-induced hyperthermia and hsp70 mRNA expression in aged rats. Am J Physiol 1991; 260:R663-667. 23. Pardue S, Groshan K, Raese JD, Morrison-Bogorad M. Hsp70 mRNA induction is reduced in neurons of aged rat hippocampus after thermal stress. Neurobiol Aging 1992; 13:661-672. 24. Tytell M, Yamaguchi K, Yamaguchi K. Role of heat shock protein 70 (hsp70) in photoreceptor cell survival in the aged rat. In: Hollyfield JG, LaVail MM, Anderson RE, eds. Retinal Degeneration: Clinical and Laboratory Applications. New York: Plenum Publishing Corp., 1993. 25. Tytell M. Heat shock proteins in the retina and optic nerve. In: Mayor RJ, Brown IR, eds. Heat Shock Proteins in the Nervous System. San Diego: Academic Press, 1994:83-100. 26. Shock NW. Systems integration. In: Finch CE, Hayflick L, eds. Handbook of the Biology of Aging. New York: Van Nostrand Reinhold Co., 1977:639-665. 27. Heydari AR, Takahashi R, Gutsmann A et al. Hsp70 and aging. Experientia 1994; 50:1092-1098. 28. Pappolla MA, Omar R, Saran B. Abnormal ubiquitinilated deposits highlight an age-related protein change. Am J Pathol 1989; 135:585-591. 29. Dickson DW, Wertkin A, Kress Y et al. Ubiquitin immunoreactive structures in normal human brains. Lab Invest 1990; 63:87-99. 30. Abe H, Mehraein P, Weis S. Unusual ubiquitin-positive glial cells in the globus pallidus of normal elderly human brains. Neuropathol Appl Neurobiol 1994;20:609-613. 31. Wang GP, Khatoon S, Iqbal K, Grundke-Iqbal I. Brain ubiquitin is markedly elevated in Alzheimer disease. Brain Res 1991; 566:146-151. 32. Mori H, Kondo J, Ihara Y. Ibiquitin is a component of paired helical filaments in Alzheimer’s disease. Science 1987; 235:1641-1644. 33. Perry G, Friedman R, Shaw G, Chau V. Ubiquitin is detected in neurofibrillary tangles and senile plaque neurites of Alzheimer disease brains. Proc Natl Acad Sci USA 1987; 84:3033-3036. 34. Hamos JE, Oblas B, Pulaski-Salo D et al. Expression of heat shock proteins in Alzheimer’s disease. Neurol 1991; 41:345-350. 35. Renkawek K, Bosman GJCGM, de Jong WW. Expression of small heat-shock protein hsp27 in reactive gliosis in Alzheimer disease and other types of dementia. Acta Neuropathol 1994; 87:511-519.

230

Astrocytes in Brain Aging and Neurodegeneration

36. Shinohara H, Inaguma Y, Goto S et al. αB-crystallin and HSP28 are enhanced in the cerebral cortex of patients with Alzheimer’s disease. J Neurol Sci 1993; 119:203-208. 37. Renkawek K, Voorter CEM, Bosman GJCGM et al. Expression of α B-crystallin in Alzheimer’s disease. Acta Neuropathol 1994; 87:155-160. 38. Smith MA, Kutty RK, Richey PL et al. Heme oxygenase-1 is associated with the neurofibrillary pathology of Alzheimer’s disease. Am J Pathol 1994; 145:42-47. 39. Schipper HM, Cissé S, Stopa EG. Expression of heme oxygenase-1 in the senescent and Alzheimer-diseased brain. Ann Neurol 1995; 37:758-768. 40. Tamaoka A, Mizusawa H, Mori H, Shoji S. Ubiquinated αB-crystallin in glial cytoplasmic inclusions from the brain of a patient with multiple system atrophy. J Neurol Sci 1995; 129:192-198. 41. Goldman JE, Corbin E. Rosenthal fibers contain ubiquitinated αB-crystallin. Am J Pathol 1991; 139:933-938. 42. Dickson DW, Yen SHC. Ubiquitin, the cytoskeleton and neurodegenerative diseases. In: Mayer J, Brown I, eds. Heat Shock Proteins in the Nervous System. San Diego: Academic Press, 1994:235-262. 43. Wilkinson KD. Cellular roles of ubiquitin. In: Mayor RJ, Brown IR, eds. Heat Shock Proteins in the Nervous System. San Diego: Academic Press, 1994:191-234. 44. Peterson TC, Peterson MR, Williams CN. The role of heme-oxygenase and aryl hydrocarbon hydroxylase in the protection by cysteamine from acetaminophen hepatotoxicity. Toxicol Appl Pharmacol 1989; 97:430-439. 45. Mydlarski MB, Liang J-J, Schipper HM. Role of the cellular stress response in the biogenesis of cysteamine-induced astrocytic inclusions in primary culture. J Neurochem 1993; 61:1755-1765. 46. Collier NC, Heuser J, Levy MA, Schlesinger MJ. Ultrastructural and biochemical analysis of the stress granula in chicken embryo fibroblasts. J Cell Biol 1988; 106:1131-1139. 47. Nover L, Scharf KD, Neumann D. Cytoplasmic heat shock granules are formed from precursor particles and are associated with a specific set of mRNAs. Mol Cell Biol 1989; 9:1298-1308. 48. Schipper HM, Mydlarski MB, Wang X. Cysteamine gliopathy in situ: A celular stress model for the biogenesis of astrocytic inclusions. J Neuropathol Exp Neurol 1993; 52:399-410. 49. Mydlarski MB, Liberman A, Schipper HM. Estrogen induction of glial heat shock proteins: Implications for hypothalamic aging. Neurobiol Aging 1995; 16:977-981. 50. Ceccatelli S, Hokfelt T, Hallman H et al. Immunohistochemical analysis of the effects of cysteamine on somatostatin-like immunoreactivity in the rat central nervous system. Peptides 1987; 8:371-384. 51. Kwok RPS, Cameron JL, Faller DV, Fernstrom JD. Effects of cysteamine administration on somatostatin biosynthesis and levels in rat hypothalamus. Endocrinol 1992; 131:2999-3009. 52. Vogel DG, Malekzadeh MH, Cornford ME et al. Central nervous system involvement in nephropathic cystinosis. J Neuropathol Exp Neurol 1990; 49:591-599. 53. Mydlarski MB, Schipper HM. Stress protein co-localization to autofluorescent astrocytic inclusion in situ and in cysteamine-treated glial cultures. Brain Res 1993; 627:113-121. 54. Brawer JR, Sonnenschein C. Cytopathological effects of estradiol on the arcuate nucleus of the female rat. A possible mechanism for pituitary tumorigenesis. Am J Anatomy 1975; 144:57-88. 55. Srebro Z, Lach H. The ultrastructure of the periventricular glia in the brains of rats and mice. Folia Biol (Krakow) 1987; 35:131-136. 56. McLaren J, Brawer JR, Schipper HM. Iron content correlates with peroxidase activity in cysteamine-induced astroglial organelles. J Histochem Cytochem 1992; 40:1887-1897. 57. Chronopoulos S, Lembo P, Alizadeh-Khiavi K, Ali-Khan Z. Ubiquitin: Its potential significance in murine AA amyloidogenesis. J Pathol 1992; 167:249-259. 58. Carlson N, Rogers S, Rechsteiner M. Microinjection of ubiquitin: Changes in protein degradation in HeLa cells subjected to heat-shock. J Cell Biol 1987; 104:547-555. 59. Finley D, Ozkaynak E, Varshavsky A. The yeast polyubiquitin gene is essential for resistance to high temperature, starvation and other stresses. Cell 1987; 49:1035-1046.

Astrocyte Granulogenesis and the Cellular Stress Response

231

60. Haas AL, Bright PM. The immunochemical detection and quantitation of intracellular ubiquitin-protein conjugates. J Biol Chem 1985; 260:12464-12473. 61. Landon M, Lowe J, Mayer RJ. Ubiquitin, endosomes-lysosomes and neurodegenerative diseases. In: Mayer J, Brown I, eds. Heat Shock Proteins in the Nervous System. San Diego: Academic Press, 1994:263-287. 62. Brawer JR, Reichard G, Small L, Schipper HM. The origin and composition of peroxidasepositive granules in cysteamine-treated astrocytes in culture. Brain Res 1994; 633:9-20. 63. Doherty FJ, Osborn NU, Wassell JA et al. Ubiquitin-protein conjugates accumulate in the lysosomal system of fibroblasts treated with cysteine proteinase inhibitors. Biochem J 1989; 263:47-55. 64. Laszlo L, Doherty FJ, Osborn NU, Mayer RJ. Ubiquitinated-protein conjugates are specifically enriched in the lysosomal system of fibroblasts. FEBS Lett 261:365-368. 65. Schwartz AL, Ciechanover A, Brandt RA, Geuze HJ. Immunoelectron microscopic localization of ubiquitin in hepatoma cells. EMBO J 1988; 7:2961-2966. 66. Manetto V, Abdul-Karim FW, Perry G et al. Selective presence of ubiquitin in intracellular inclusions. Am J Pathol 1989; 134:505-513. 67. Cissé S, Perry G, Lacoste-Royal G et al. Immunochemical identification of ubiquitin and heat-shock proteins in corpora amylacea from normal aged and Alzheimer’s disease brains. Acta Neuropathol 1993; 85:233-240. 68. Anzil AP, Herrlinger H, Blinzinger K, Kronski D. Intraneuritic corpora amylacea. Virchows Arch [A] 1974; 364:297-304. 69. Palmucci L, Anzil AP, Luh S. Intra-astrocytic glycogen granules and corpora amylacea stain positively for polyglucosan: A cytochemical contribution on the fine structural polymorphism of particulate polysaccharides. Acta Neuropathol (Berl) 1982; 57:99-102. 70. Ramsay HJ. Ultrastructure of corpora amylacea. J Neuropathol Exp Neurol 1965; 24:25-39. 71. Schipper HM, Cissé S. Mitochondrial constituents of corpora amylacea and autofluorescent astrocytic inclusions in senescent human brain. Glia 1995; 14:55-64. 72. Shang F, Taylor A. Oxidative stress and recovery from oxidative stress are associated with altered ubiquitin conjugating and proteolytic activities in bovine lens epithelial cells. Biochem J 1995; 307:297-303. 73. Haas AL, Warms JV, Hershko A, Rose IA. Ubiquitin-activating enzyme. Mechanism and role in protein-ubiquitin conjugation. J Biol Chem 1982; 257:2543-2548. 74. Herskho A, Heller H, Elias S, Ciechanover A. Components of ubiquitin-protein ligase system. Resolution, affinity purification, and role in protein breakdown. J Biol Chem 1983; 258:8206-8214. 75. Becker J, Craig EA. Heat-shock proteins as molecular chaperones. Eur J Biochem 1994; 219:11-23. 76. Welch WJ. Mammalian stress response: cell physiology, structure/function of stress proteins, and implications for medicine and disease. Physiol Rev 1992; 72:1063-1081. 77. Findly RC, Gillies RJ, Shulman RG. In vivo phosphorous-31 nuclear magnetic resonance reveals lowered ATP during heat shock of Tetrahymena. Science 1983; 219:1223-1225. 78. Stevenson MA, Calderwoud SK, Hahn GM. Rapid increases in inositol trisphosphate and intracellular Ca++ after heat shock. Biochem Biophys Res Commun 1981; 137:826-833. 79. Weitzel G, Pilatus U, Rensing L. Similar dose response of heat shock protein synthesis and intracellular pH change in yeast. Exp Cell Res 1985; 159:252-256. 80. Brawer J, Schipper HM, Robaire B. Effects of long-term androgen and estradiol exposure on the hypothalamus. Endocrinology 1983; 112:194-199. 81. Schipper HM, Brawer JR, Nelson JF, Felicio LS, Finch CE. Role of the gonads in the histologic aging of the hypothalamic arcuate nucleus. Biol Reprod 1981; 25:413-419. 82. Schipper HM, Lechan RM, Reichlin S. Glial peroxidase activity in the hypothalamic arcuate nucleus: Effects of estradiol valerate-induced persistent estrus. Brain Res 1990; 507:200-207. 83. Schipper HM. Role of peroxidase-positive astrocytes in estradiol-related hypothalmic damage. In: Fedoroff S, Juurlink B, Doucette R. eds. Biology and Pathology of Astrocyte-neuron Interactions. New York: Plenum Publishing Corporation, 1993:125-139.

232

Astrocytes in Brain Aging and Neurodegeneration

84. Kleopoulos SP, Olazabal UE, Lauber AH et al. Heat-shock proteins 90 Kd and 70Kd in rat brain and uterus: Cellular localization by immunocytochemistry and in situ hybridization. Society for Neuroscience 1991; 17:432 (abstr). 85. Olazabal UE, Pfaff DW, Mobbs CV. Estrogenic regulation of heat shock protein 90 kDa in the rat ventromedial hypothalamus and uterus. Mol Cell Endocrinol 1991; 84:174-183. 86. Olazabal UE, Pfaff DW, Mobbs CV. Sex differences in the regulation of heat shock protein 70 kDa and 90 kDa in the rat ventromedial hypothalamus by estrogen. Brain Res 1992; 596; 311-314. 87. Munday R. Toxicity of thiols and disulphides: Involvement of free-radical species. Free Rad Biol Med 1989; 7:659-673. 88. Lee AS. Coordinated regulation of a set of genes by glucose and calcium ionophores in mammalian cells. TIBS 1987; 12:20-23. 89. Lee AS. Mammalian stress response: Induction of the glucose-regulated protein family. Curr Op Cell Biol 1992; 4:267-273. 90. Liochev SI, Fridovich I. Superoxide radical in Escherichia coli. In: Scandalios JG, ed. Current Communications in Cell and Molecular Biology. Molecular Biology of Free Radical Scavenging Systems. New York: Cold Spring Harbor Laboratory Press, 1992:213-229. 91. Wong GHW, Kamb A, Tartaglia LA, Goeddel DV. Possible protective mechanisms of tumour necrosis factors against oxidative stress. In: Scandalios JG, ed. Current Communications in Cell and Molecular Biology. Molecular Biology of Free Radical Scavenging Systems. New York: Cold Spring Harbor Laboratory Press, 1992:69-96. 92. Manganaro F, Chopra VS, Mydlarski MB et al. Redox perturbations in cysteamine-stressed astroglia: Implications for inclusion formation and gliosis in the aging brain. Free Rad Biol Med 1995; 19:823-835. 93. Yen T-C, King K-L, Lee H-C et al. Age-dependent increase of mitochondrial DNA deletions together with lipid peroxides and superoxide dismutase in human liver mitochondria. Free Rad Biol Med 1994; 16:207-214. 94. Saggu H, Cooksey J, Dexter D et al. A selective increase in particulate superoxide dismutase activity in parkinsonian substantia nigra. J Neurochem 1989; 53:692-697. 95. Jenner P. What process causes nigral cell death in Parkinson’s disease? In: Cedarbaum JM, Gancher ST, eds. Neurological Clinics; Parkinson’s Disease. Montreal: WB Saunders Co., 1992:387-404. 96. Ball P, Knuppen R. Formation of 2- and 4-hydroxyestrogens by brain, pituitary, and liver of the human fetus. J Clin Endocrinol Metab 1978; 47:732-737. 97. Inoue K, Yoshizawa I. Immunocytochemical localization of catecholestrogens in the rat pituitary gland. II. Median eminence. Acta Histochem Cytochem 1989; 22:65-76. 98. Kalyanaraman B, Felix CC, Sealy RC. Semiquinone anion radicals of catechol(amine)s, catechol estrogens, and their metal ion complexes. Environ Health Perspect 1985; 64:185-198. 99. Goodman Y, Bruce AJ, Cheng B, Mattson MP. Estrogens attenuate and corticosterone exacerbates excitotoxicity, oxidative injury and amyloid β-peptide toxicity in hippocampal neurons. J Neurochem 1996; 66:1836-1844. 100. Srebro Z. Periventricular Gomori-positive glia in brains of X-irradiated rats. Brain Res 1971; 35:463. 101. Castilho RF, Kowaltowski AJ, Meinicke AR et al. Permeabilization of the inner mitochondrial membrane by Ca2+ ions is stimulated by t-butylhydroperoxide and mediated by reactive oxygen species generated by mitochondria. Free Rad Biol Med 1995; 18:479-486. 102. Wilson PD, Franks LM. The effect of age on mitochondrial ultrastructure and enzymes. Adv Exp Med Biol 1975; 53:171-183. 103. Welch WJ, Suhan JP. Morphological study of the mammalian stress response: Characterization of changes in cytoplasmic organelles, cytoskeleton, and nucleoli, and appearance of intranuclear actin filaments in rat fibroblasts after heat-shock treatment. J Cell Biol 1985; 101:1198-1211. 104. Skrede S. Effects of cystamine and cysteamine on the adenosine-triphosphatase activity and oxidative phosphorylation of rat-liver mitochondria. Biochem J 1966; 98:702-710. 105. Gutteridge JMC, Halliwell B. The measurement and mechanism of lipid peroxidation in biological systems. TIBS 1990; 15:129-135.

Astrocyte Granulogenesis and the Cellular Stress Response

233

106. Laganiere S, Yu VP. Modulation of membrane phospholipid fatty acid composition by age and food restriction. Gerontology 1993; 39:7-18. 107. Yu BP, Suescum EA, Yang SY. Effect of age-related lipid peroxidation on membrane fluidity and phospholipase A2: Modulation by dietary restriction. Mech Aging Dev 1992; 65:17-33. 108. Shigenaga MK, Hagen TM, Ames BN. Oxidative damage and mitochondrial decay in aging. Proc Natl Acad Sci USA 1994; 91:10771-10778. 109. Skrede S, Christophersen BO. Effects of cystamine and cysteamine on the peroxidation of lipids and the release of proteins from mitochondria. Biochem J 1966; 101:37-41. 110. Haenen GRMM, Vermeulen NPE, Timmerman H, Bast A. Effect of thiols on lipid peroxidation in rat liver microsomes. Chem Biol Interaction 1989; 71:201-212. 111. Weiss KF, Kumar KS. Antioxidant mechanisms in radiation injury and radioprotection. In: Chow CK, ed. Cellular Antioxidant Defense Mechanisms. Vol. II. Boca Raton: CRC Press Inc., 1994:163-189. 112. Hwang C, Sinskey AJ, Lodish HF. Oxidized redox state of glutathione in the endoplasmic reticulum. Science 1992; 257:1496-1502. 113. Bacq ZM. Chemical protection against X- and γ-ray. In: Bacq ZM, Alexander P, eds. Fundamentals of Radiobiology. London: Butterworths Scientific Publications, 1955:290-327. 114. Halliwell B, Gutteridge JMC, Cross CE. Review article: Free radicals, antioxidants, and human disease: Where are we now? J Lab Clin Med 1992; 119:598-620. 115. Knoblock EC, Purdy EC. Apparent instability of 2-mercaptoethylamine complexes. Radiat Res 1961; 15:94-96. 116. Ananthan J, Boldberg AL, Voellmy R. Abnormal proteins serve as eukaryotic stress signals and trigger the activation of heat shock genes. Science 1986; 232:522-524. 117. Burdon RH, Gill VM, Rice-Evans C. Oxidative stress and heatshock protein induction in human cells. Free Rad Res Commun 1987; 3:129-139. 118. Finley D, Ciechanover A, Varshavsky A. Thermolability of ubiquitin-activating enzyme from the mammalian cell cycle mutant ts85. Cell 1984; 37:43-55. 119. Liu R, Li X, Li GC. Expression of human hsp70 in rat fibroblasts enhances cell survival and facilitates recovery from translational and transcriptional inhibition following heat shock. Cancer Rec 1992; 52:3667-3673. 120. Condorelli DF, Bellurado N, Avola R et al. Effect of trophic factors, released after hippocampal injury, on astroglial cell proliferation. Metab Brain Dis 1989; 4:41-46. 121. Fleming PD, Rogers J. Neuropathology of olfactory system in Alzheimer’s disease and normal aging. Soc Neurosci Abstract 1986; 12:1314. 122. Fleming PD, Cordoza ME, Woods SG et al. Corpora amylacea increased in Alzheimer’s disease. Neurology (Suppl 1) 1987; 37:157. 123. Behrman S, Carrol JD, Janota I, Matthews WB. Progressive supranuclear palsy. Clinicopathological study of four cases. Brain 1969; 92:663-678. 124. Gambetti P, DiMauro S, Hirt L, Blume RP. Myoclonic epilepsy with Lafora bodies: Some ultrastructural, histochemical and biochemical aspects. Arch Neurol 1971; 25:483-493. 125. Janeway R, Ravens JR, Pearce LA et al. Progressive myoclonic epilepsy with Lafora inclusion bodies. I. Clinical, genetic, histopathologic and biochemical aspects. Arch Neurol 1967; 16:565-582. 126. Robitaille Y, Carpenter S, Karpati G, DiMauro S. A distinct form of adult polyglucosan body disease with massive involvement of central and peripheral neuronal processes and astrocytes. Brain 1980; 103:315-336. 127. Palmucci L, Anzil AP, Luh S. Intra-astrocytic glycogen granules and corpora amylacea stain positively for polyglucosan: A cytochemical contribution on the fine structural polymorphism of particulate polysaccharides. Acta Neuropathol (Berl) 1982; 57:99-102. 128. Takahashi K, Agari M, Nakamura H. Intra-axonal corpora amylacea in ventral and lateral horns of the spinal cord. Acta Neuropathol (Berl) 1975; 31:151-158. 129. Wolozin BL, Prucnki A, Dickson DW, Davies P. Neuronal antigen in the brains of Alzheimer patients. Science 1986; 232:648-650. 130. Yagashita S. Ultrastructural observations on axonal swelling in the human gracile nucleus. Virch Arch Path and Hist 1979; 382:217-226.

234

Astrocytes in Brain Aging and Neurodegeneration

131. Yagashita S, Itoh Y. Corpora amylacea in the peripheral nerve axon. Acta Neuropathol (Berl) 1977; 37:73-76. 132. Hollander DM, Hutchins GM. Central spherules in pulmonary corpora amylacea. Arch Pathol Lab Med 1978; 102:629-630. 133. Seman G. Fibrillar bodies in human prostate. The Prostate 1983; 54:179. 134. Sun CN. Ultrastructural study of corpora amylacea in human thyroid gland. Exp Path 1983; 23:219-225. 135. Schipper HM, Mateescu-Cantuniari A. Identification of peroxidase-positive astrocytes by combined histochemical and immunolabeling techniques in situ and in cell culture. J Histochem Cytochem 1991; 39:1009-1016. 136. Wislocki GB, Leduc EH. The cytology of the subcommissural organ, Reissner’s fiber, periventricular glial cells and posterior collicular recess of the rat’s brain. J Comp Neurol 1954; 101:283-309. 137. Koritsanszky S, Vigh B, Aros B. Studies on the Gomori-positive glial cells. I. Changes in the Gomori-positive glial cells in rats of various ages. Acta Biol Hung 1967; 18:9-19. 138. Schipper HM. Gomori-positive astrocytes: Biological properties and implications for neurologic and neuroendocrine disorders. Glia 1991; 4:365-377. 139. Stam FC, Roukema PA. Histochemical and biochemical aspects of corpora amylacea. Acta Neuropathol (Berlin) 1973; 25:95-102. 140. Cissé S, Lacoste-Royal G, Laperriere J et al. Ubiquitin is a component of polypeptides purified from corpora amylacea of aged human brain. Neurochem Res 1991; 16:429-433. 141. Rajagopalan KV. Sulfite oxidase. In: Coughlan MP, ed. Molybdenum and Molybdenumcontaining Enzymes. New York: Pergamon Press, 1980:241-272. 142. Ellis RJ, van der Vies SM. Molecular chaperones. Ann Rev Biochem 1991; 60:321-347. 143. Lawrence EH. Heat shock, stress proteins chaperones, and proteotoxicity. Cell 1991; 66:191-197. 144. Woodford B, Tso MOM. An ultrastructural study of the corpora amylacea of the optic nerve head and retina. Am J Ophthalmol 1980; 90:492-502. 145. Gertz HJ, Cervos-Navarros J, Frydl V, Shultz F. Glycogen accumulation of the aging human brain. Mech Ageing Dev 1985; 35:25-35. 146. Schipper HM, Scarborough DE, Lechan RM, Reichlin S. Gomori-positive astrocytes in primary culture: Effects of in vitro age and cysteamine exposure. Dev Brain Res 1990; 54:71-79. 147. Brawer JR, Stein R, Small L et al. Composition of Gomori-positive inclusions in astrocytes of the hypothalamic arcuate nucleus. Anat Rec 1994; 240:407-415. 148. Cissé S, Schipper HM. Experimental induction of corpora amylacea-like inclusions in rat astroglia. Neuropathol Appl Neurobiol 1995; 21:423-431. 148a.Schipper HM. Experimental indcution of corpora amylacea in adult rat brain. Microsc Res Techniq, 1998; (in press). 149. Cerami A. Aging of proteins and nucleic acids: What is the role of glucose? Trends Biochem Sci 1986; 11:311-314. 150. Yan S-D, Chen X, Schmidt A-M, Brett J. Non-enzymatic glycation of Tau in neurofibrillary tangles of Alzheimer’s disease: a mechanism for aggregation and neurotoxicity. Neurology 1994; 44 (Supple 2):A371, Abstract 960S. 151. Balazs L, Leon M. Evidence of an oxidative challenge in the Alzheimer’s brain. Neurochem Res 1994; 19:1131-1137. 152. Butterfield AD, Hensley K, Harris M et al. β-amyloid peptide free radical fragments initiate synaptosomal lipoperoxidation in a sequence-specific fashion: Implications to Alzheimer’s disease. Biochem Biophys Res Comm 1994; 200:710-715. 153. Palmer AM, Burns MA. Selective increase in lipid peroxidation in the inferior temporal cortex in Alzheimer’s disease. Brain Res 1994; 645:338-342. 154. Subbarao KV, Richardson JS, Ang LC. Autopsy samples of Alzheimer’s cortex show increased peroxidation in vitro. J Neurochem 1990; 55:342-345. 155. Bolanos JP, Peuchen S, Heales SJR et al. Nitric oxide-mediated inhibition of the mitochondrial respiratory chain in cultured astrocytes. J Neurochem 1994; 63:910-916.

CHAPTER 11

Glial Iron Sequestration and Neurodegeneration Hyman M. Schipper

The Free Radical Hypothesis of Parkinson’s Disease

I

diopathic Parkinson’s disease (PD) is a movement disorder of uncertain etiology characterized by the accelerated loss of dopaminergic (DA) neurons in the pars compacta of the substantia nigra.1 Although dissenting opinions exist,2 there is currently a broad consensus implicating oxidative stress as a major factor in the pathogenesis of PD.3-6 The free radical hypothesis of PD draws support from the following observations: 1. The accelerated oxidative deamination of DA by monoamine oxidase B (MAO B) in idiopathic and experimental Parkinsonism subjects the nigrostriatal projections to excessive concentrations of hydrogen peroxide (H2O2).3,7,8 2. The neurotoxins, 6-hydroxydopamine, manganese and, to some extent, MPTP, induce parkinsonism in animals via the generation of free radicals.8-12 3. Basal lipid peroxidation in the substantia nigra of postmortem human PD brain was found to be significantly elevated relative to non-Parkinsonian controls matched for age and postmortem interval.13 4. Free radical scavenger enzymes (such as catalase) and intracellular reducing substances (such as reduced glutathione) are reportedly deficient in the basal ganglia of patients with Parkinson’s disease.14,15 In contrast, the mitochondrial antioxidant manganese superoxide dismutase (MnSOD) appears to be augmented in the basal ganglia of PD subjects and may represent an adaptive response to oxidative stress and mitochondrial injury in this condition.16,17 5. Although still controversial, results of a large multicenter clinical trial (DATATOP) and a more recent study suggest that treatment of early PD with the MAO B inhibitor 1-deprenyl may slow the progression of this neurodegenerative disorder by curtailing the production of dopamine-derived H2O2.18-20

The Redox Neurobiology of Alzheimer’s Disease Alzheimer’s disease (AD) is characterized by progressive neuronal degeneration, gliosis, and the accumulation of intracellular inclusions (neurofibrillary tangles [NFTs]) and extracellular deposits of amyloid (senile plaques [SPs]) in discrete regions of the basal forebrain, hippocampus, and association cortices.21 Although the etiology of sporadic AD remains unknown, evidence amassed over the last 5 years has implicated free radicals and oxidative stress in the pathogenesis of this dementing illness. For example, end products of lipid peroxidation are elevated in the brains of AD subjects22,23 and various antioxidant Astrocytes in Brain Aging and Neurodegeneration, edited by Hyman M. Schipper. ©1998 R.G. Landes Company.

236

Astrocytes in Brain Aging and Neurodegeneration

defenses are reportedly deranged in AD brain and peripheral tissues.24-26 The excessive generation of free radicals may promote both paired helical filament formation and amyloid deposition in the AD brain.26,27 Moreover, the results of recent biochemical studies suggest that the neurotoxic effects of certain amyloid fragments may, in part, be mediated by free radical intermediates.28,29 A growing body of evidence suggests that brain cell mitochondria may be prime targets of chronic oxidative injury in Alzheimer-affected tissues and that bioenergetic failure (mitochondrial insufficiency) may play an important role in the pathogenesis of this disease.30,31 In support of the latter, cytochrome C oxidase and complex V activities,32 the pyruvate dehydrogenase complex, and various Krebs cycle intermediates are purportedly deficient in AD brain,30,32,33 and excessive mutations in mitochondrial genes encoding subunits of complex I and IV have been reported in the CNS and blood of AD subjects.34,35 It has been suggested that further increases in free radical generation resulting from infidelity of electron transport within the inner membranes of damaged mitochondria may perpetuate oxidative neuropil injury in the AD brain long after initiating neurotoxic insults have dissipated.31,36 As in the case of idiopathic PD, epigenetic factors contributing to the excessive oxidative stress and mitochondrial electron transport chain deficits in the brains of AD subjects remain poorly understood.

Iron Deposition and Neurodegenerative Disease The pathological sequestration of redox-active brain iron has been implicated as a major generator of reactive oxygen species in PD, AD, and other aging-related neurodegenerative disorders.

Parkinson’s Disease Abnormally high levels of tissue iron have been consistently reported in the substantia nigra and basal ganglia of PD subjects.5,6,17,37 In PD, the excessive iron deposition primarily affects the zona compacta of the substantia nigra and correlates with loss of dopaminergic neurons in this brain region.6,38,39 Using conventional histochemical stains, the excessive nigral iron appears to be predominantly deposited within astrocytes, microglia, macrophages and microvessels within areas depleted of neuromelanin-containing (dopaminergic) neurons. Although minor concentrations of iron have been detected in neuronal neuromelanin using micro-analytical techniques,6,40 histochemical evidence for substantial iron deposition in PD-affected nigral or striatal neurons is scant or nonexistent.6,41,42 Thus, glia and other nonneuronal cells may represent the chief substrates of excessive iron sequestration in the basal ganglia of PD subjects. In the latter, the augmented tissue iron levels are accompanied by alterations in the expression of several important iron-binding proteins and their receptors. In general, increased expression of tissue ferritin, the major intracellular sequestrant of ferric iron, parallels the distribution of the excess iron and largely implicates nonneuronal (glial) cellular compartments.6 The iron-binding protein transferrin is responsible for the extracellular transport of ferric iron and its delivery to virtually all mammalian tissues. After binding to transferrin receptors within the plasma membrane, the transferrin-transferrin receptor complex is internalized via endocytosis, free iron is liberated from the complex by a temperature and energy-dependent process involving endosomal acidification, the iron translocates to the cytosol and is sequestered in ferritin, and the apotransferrin-transferrin receptor complex is recycled to the cell surface, where it dissociates.43-45 To maintain tissue iron homeostasis, plasma membrane transferrin receptor densities and intracellular ferritin concentrations are tightly regulated (at transcriptional and posttranscriptional levels) by iron bioavailability and intracellular iron stores.43-45 In normal rat and human brain tissues, there appears to be an overt mismatch between local brain iron concentrations and the densities in cell surface transferrin binding sites.46-48 Moreover,

Glial Iron Sequestration and Neurodegeneration

237

in glaring contrast to the ferritin data, the density of transferrin binding sites remains unchanged or varies inversely with augmented iron stores in the substantia nigra and striatum of PD subjects.42,46,49,50 An important interpretation of these findings is that, in contradistinction to most peripheral tissues, transferrin and its receptor play a limited role, if any, in the sequestration of iron by aging and degenerating CNS tissues.42,46,49,50 Indeed, attention is shifting to alternative iron transport mechanisms such as that mediated by lactoferrin and the lactoferrin receptor, which are reportedly augmented in neurons, astrocytes, and blood vessels of PD-affected, iron-laden neural tissues.51,52

Alzheimer’s Disease As in the case of PD, abnormalities of iron homeostasis and excessive deposition of this transition metal are characteristic of Alzheimer-affected brain tissues. In AD, increases in bulk brain iron have been reported in both gray and white matter regions.53 Interestingly, although a significant proportion of ferritin iron in normal human brain is stored within oligodendroglia, in AD white matter there appears to be a shift towards pathological iron trapping within the astrocytic compartment.53 In the AD hippocampus, augmented deposition of nonheme iron has been shown to occur in NFT-bearing neurons, astrocytes, microglia, and in the vicinity of neuritic plaques.37,42,53,54 Although transferrin immunoreactivity has been noted in AD astrocytes and senile plaques, there appears to be an overall decrease in levels of immunoreactive transferrin in AD-affected cortical and subcortical brain tissue relative to that in age-matched, nondemented controls.53 Moreover, transferrin receptor densities (determined by [125I]-transferrin binding) are significantly reduced in postmortem hippocampus and temporal cortex derived from AD subjects relative to controls.50 This apparent mismatch of brain iron and transferrin/transferrin receptor is similar to that observed in the substantia nigra of PD subjects (vide supra) and further suggests that the transferrin pathway of iron mobilization may contribute little to the pathological sequestration of brain iron observed in the major aging-related neurodegenerative disorders. As in the case of the PD nigra, increased lactoferrin and/or lactoferrin receptor immunoreactivity has been reported in neurons, glia and extracellular amyloid plaques within brain regions undergoing degeneration in AD, Down’s syndrome, Pick’s disease, and ALSParkinsonism/dementia complex of Guam.52,55-58 Unlike transferrin, lactoferrin binding to its receptor is not affected by degrees of tissue iron saturation and could theoretically permit toxic levels of this metal to accumulate in these degenerating neural tissues.52 By participating in Fenton reactions, the aberrantly-sequestered brain iron could promote oxidative stress and lipid peroxidation and thereby directly contribute to the neurodegenerative process. Furthermore, the amyloid precursor protein gene contains iron response elementlike consensus sequences, raising the possibility that brain amyloid deposition in AD and other human neurodegenerative disorders may be iron-sensitive.42

Iron Sequestration in Aging Astroglia Efforts to ameliorate iron-mediated neuronal injury in AD and PD presupposes some understanding of the regulatory mechanisms subserving iron metabolism and sequestration in the aging and degenerating CNS. The following important, but as yet unanswered, questions in this regard arise from the pathological studies considered in the previous section: 1. What is the role of heme vs. nonheme iron in aging-related neurodegenerative conditions? 2. Which cell type(s) and subcellular compartments are responsible for the abnormal sequestration of brain iron in these degenerative disorders? 3. Does induction of a cellular stress (heat shock) response facilitate trapping of redox-active iron in neural tissues? and

238

Astrocytes in Brain Aging and Neurodegeneration

4. What is the role of the iron-transport protein, transferrin, in this process? We have begun to explore these and related issues by focusing on the mechanisms responsible for the accumulation of iron-rich cytoplasmic inclusions in aging subcortical astrocytes and in astroglial cultures subjected to oxidative stress. As described in chapter 9, the sulfhydryl agent cysteamine (CSH) accelerates the aging-related accumulation of ironrich cytoplasmic inclusions in hippocampal, striatal and other subcortical astroglia in situ and in primary neuroglial cultures. Evidence was also provided that these iron-laden glial inclusions and related corpora amylacea are derived from oxidatively-damaged mitochondria in the context of a cellular stress (heat shock) response (chapters 9 and 10). Several laboratories including our own59,60 have previously concluded on the basis of histochemical and spectrofluorometric data that porphyrins and heme ferrous iron are responsible, respectively, for the orange-red autofluorescence and nonenzymatic peroxidase activity in these glial inclusions. However, we subsequently determined that CSH suppresses the incorporation of the heme precursors ∆-amino[14C]-levulinic acid (ALA) and [14C]-glycine into astroglial porphyrin and heme in primary culture, prior to and during the time when increased iron content is detectable in swollen astrocyte mitochondria by microprobe analysis (Fig.11.1).61,62 Thus, contrary to hypothesis, de novo biosynthesis of porphyrins and heme is not responsible for the increased mitochondrial iron content, autofluorescence, and peroxidase activity observed in cultured astroglia following CSH exposure. Because the CSHinduced astroglial inclusions are morphologically and histochemically identical to the ironladen astrocyte granules that normally accumulate in the aging periventricular brain, it would seem highly unlikely that augmentation of porphyrin-heme biosynthesis plays a role in the biogenesis of the latter as well. Oxidized mitochondrial flavoproteins exhibit fluorescence emission spectra that may be difficult to distinguish from porphyrins63,64 and are likely mediators of orange-red autofluorescence in these astrocytic inclusions. Following inhibition of porphyrin-heme biosynthesis, CSH augments the incorporation of 59Fe (or 55Fe) into astroglial mitochondria without significantly affecting transfer of the metal into whole-cell and lysosomal compartments (Fig. 11.2).62 This CSH effect was clearly demonstrable when inorganic [59Fe]Cl3, but not [59Fe]-diferric transferrin (Fig. 11.3), served as the metal donor. These findings are consistent with previous reports that intracellular transport of low molecular weight, inorganic iron may be 5- to 10-fold more efficient than that of transferrin-bound iron in various tissues, including melanoma cells,65-67 Chinese hamster ovary cells,68 and K562 cells.69 Our observations support the conclusion of Adams and coworkers70 that inhibition of heme biosynthesis stimulates the selective transport of low molecular weight iron from the cytoplasm to the mitochondrial compartment. Recent work from our laboratory suggests that dopamine may be an important endogenous stressor mediating nigrostriatal glial iron trapping in PD and, to a lesser extent, in the course of normal aging. Akin to the effects of CSH, physiologically-relevant concentrations of dopamine (1 µM) stimulate the sequestration of nontransferrin-bound 55Fe in the mitochondrial compartment of cultured astroglia without affecting the disposition of transferrin-derived 55Fe. L-DOPA (25 µM) weakly recapitulated the effects of dopamine on glial iron sequestration, whereas equimolar concentrations of norepinephrine were entirely inert in this regard.71 The effects of dopamine on glial iron trapping were abrogated by coadministration of ascorbate (200 µM) but not by the D1 and D2 antagonists, SCH 23390 and sulpiride, respectively, suggesting that, analogous to the CSH mechanism of action, dopamine-derived free radicals promote the sequestration of nontransferrin-derived iron within astroglial mitochondria. That such dopamine-astrocyte interactions may be operational in vivo is supported by: 1. a recent nuclear microscopy study demonstrating increased elemental iron in the substantia nigra of 6-hydroxydopamine-lesioned rats;72 in conjunction with

Glial Iron Sequestration and Neurodegeneration

239

Fig. 11.1. Incorporation of [14C]-ALA into (A) uroporphyrin, (B) coproporphyrin, (C) protoporphyrin, and (D) hemin in control untreated (O) and CSH-treated (●) astrocytes. Data are presented as mean ± SD (bars) of three to six observations. *p<0.05, **p<0.01 for significance of difference relative to untreated controls. CSH suppresses porphyrin-heme biosynthesis in cultured astroglia. Reprinted with permission from Wang X et al, J Neurochem 1995; 64:1868-1877.

2. immunocytochemical evidence of direct contact between tyrosine hydroxylase-positive (dopaminergic) processes and Gomori-positive (metal-laden) astrocytes in the rat arcuate nucleus73 and basal ganglia (Schipper, unpublished results).

The Role of HO-1 in Brain Iron Deposition As described in chapter 10, HO-1 is a 32 kDa member of the stress protein superfamily that catalyses the rapid conversion of heme to biliverdin in brain and other tissues. In response to oxidative stress, induction of HO-1 may protect cells by catabolizing prooxidant metalloporphyrins such as heme to bile pigments (biliverdin, bilirubin) with free radicalscavenging capabilities.74 On the other hand, HO-1-catalyzed heme degradation liberates free iron and carbon monoxide (CO) which exacerbate intracellular oxidative stress by stimulating oxyradical generation within the mitochondrial compartment.75 We74 and others76 have recently shown that HO-1 is massively upregulated in neurons and astrocytes of Alzheimer-diseased human temporal cortex and hippocampus (but not in unaffected substantia nigra) relative to age-matched, nondemented controls. Conversely, the percentage of GFAP-positive astrocytes expressing HO-1 in substantia nigra (but not in other brain regions) of PD subjects is significantly increased in comparison with age-matched controls76a (Fig. 11.4). Although HO-1 upregulation in these conditions may confer some degree of

240

Fig. 11.2. Iron-59 uptake in control (O) and CSH-treated (●) astrocytes exposed to [59Fe]Cl3: (A) total cell, (B) lysosomal, and (C) mitochondrial fractions. Data are presented as mean ± SD (bars) of three to six observations. *p<0.05 for significance of difference relative to untreated controls. CSH promotes the sequestration of nontransferrin derived iron within the mitochondrial compartment. Reprinted with permission from Wang X et al, J Neurochem 1995; 64:1868-1877.

Astrocytes in Brain Aging and Neurodegeneration

Fig. 11.3. Iron-59 uptake in cultured control untreated (O) and CSH-treated (●) astrocytes exposed to [59Fe]-transferrin: (A) total cell, (B) lysosomal, and (C) mitochondrial fractions. Data are mean ± SD (bars) values of three to six observations. CSH has no significant effect on the disposition of transferrin-derived iron in these cells. Reprinted with permission from Wang X et al, J Neurochem 1995; 64:1868-1877.

Glial Iron Sequestration and Neurodegeneration

241

Fig. 11.4. Percentage of glial fibrillary acid protein (GFAP)-positive astrocytes concomitantly expressing HO-1 in various brain regions of control and PD subjects. Vertical lines denote standard errors of the mean and asterisks denote statistical significance (p<0.05). ( ) = number of specimens per group. Reprinted with permission from Schipper HM et al. Exp Neurol 1998; 150:60-68.

cytoprotection by degrading pro-oxidant heme to anti-oxidant bile pigments, heme-derived free iron and CO may contribute, at least in part, to the development of mitochondrial electron transport chain deficiencies and excess mitochondrial DNA mutations reported in the brains of AD and PD subjects.77,78 The upregulation of HO-1 may have important implications for the biogenesis of mitochondria-derived astrocytic inclusions in senescent and oxidatively challenged astroglia. Within 6 h of CSH exposure, cultured astroglia exhibit 4- to 10-fold increases in HO-1 mRNA and protein levels, robust HO-1 immunofluorescent staining, and a 3-fold increase in HO enzymatic activity.79-81 As in the case of CSH, H2O2, menadione and dopamine (but not norepinephrine) consistently upregulate HO-1 in cultured astroglia prior to promoting the sequestration of nontransferrin-bound 55Fe by the mitochondrial compartment.71,82 In dopamine-exposed glial cultures and in senescent subcortical astroglia in situ, the liberation of free iron and CO resulting from HO-1-catalyzed heme degradation may promote early oxidative injury to mitochondrial membranes and thereby facilitate the transformation of normal astrocyte mitochondria to iron-rich cytoplasmic inclusions. In support of this contention, we observed that dopamine-induced sequestration of mitochondrial iron

242

Astrocytes in Brain Aging and Neurodegeneration

in cultured rat astroglia is prevented by coadministration of the competitive heme oxygenase inhibitor tin-mesoporphyrin, or the HO-1 transcriptional suppressor dexamethasone (Schipper and Bernier, manuscript in preparation).

Pro-toxin Bioactivation by Astrocytes in Primary Culture Electron spin resonance spectroscopy (ESR) with magnesium spin stabilization was used to determine whether CSH-induced peroxidase activity (mitochondrial iron deposition) in cultured astroglia is capable of oxidizing catecholestrogens and catecholamines to their respective orthosemiquinone radicals.83 Incubation of 2-hydroxyestradiol with homogenates derived from untreated (control) astroglial monolayers in the presence of H2O2 and NADPH (pH 7.0) yielded no or barely detectable o-semiquinone spectra. In contrast, intense o-semiquinone spectra indicative of robust catechol oxidation were consistently observed following incubation of equimolar concentrations of 2-hydroxyestradiol with homogenates obtained from CSH-pretreated (iron-enriched) astrocyte monolayers in the presence of appropriate cofactors (Fig. 11.5). In the absence of H2O2 substrate, there was a marked reduction in signal amplitude, attesting to the important role of glial peroxidase activity in the augmentation of catecholestrogen metabolism in our system.83 The results of the ESR experiments, in conjunction with the protective effects of α-tocopherol and 21-aminosteroids on estradiol-induced depletion of hypothalamic β-endorphin (chapter 9), support the notion that free radical generation by iron-laden hypothalamic astrocytes may mediate, at least in part, the dystrophic effects of estradiol in this brain region. As in the case of 2-hydroxyestradiol, we demonstrated that the iron-dependent peroxidase activity induced in cultured astroglia by CSH exposure significantly enhances the oxidation of the catecholamine, dopamine, to its dopamine-o-semiquinone derivative in the presence of H2O2.83 This observation is consistent with previous reports that dopamine and norepinephrine are readily oxidized to semiquinones with proven neurotoxic activity in vitro via peroxidase-mediated reactions.84 Because aging subcortical astrocytes may exhibit both enhanced MAO B activity (see chapter 6) and abundant mitochondrial iron, it is conceivable that H2O2 produced by MAO B oxidation of dopamine serves as a cofactor for further dopamine oxidation (to potentially neurotoxic ortho-semiquinones) by peroxidasemediated reactions. In addition to dopamine, redox-active glial iron may also facilitate the nonenzymatic oxidation of: 1. the pro-toxin MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) to the dopaminergic toxin MPP+ in the presence of MAO inhibitors;85 and 2. the dopamine precursor DOPA to 2,4,5-trihydroxyphenylalanine (TOPA) and the non-NMDA excitotoxin TOPA-quinone.86 The high stress protein content of peroxidase-positive astrocytes and a compensatory upregulation of manganese superoxide dismutase (MnSOD) observed in these cells could serve to limit the extent of oxidative injury within the glia themselves (see chapter 10). However, the dystrophic effects of reactive oxygen species need not be confined to the cellular compartment in which they are generated. For example, H2O2 is lipid soluble and can easily traverse plasma membranes to reach the intercellular space, while superoxide (potentially generated in our model by semiquinone-quinone redox cycling or by infidelity of electron transport in damaged inner mitochondrial membranes) can be extruded from cells via anion channels.87 In support of this formulation, we recently observed that catecholamine-secreting PC12 cells grown atop monolayers of CSH-pretreated (iron-enriched) astrocytes are far more susceptible to dopamine/H2O2-related killing than PC12 cells cocultured with nonpretreated (DAB-negative) astroglia. In both coculture paradigms, astroglial death, determined by ethidium monoazide bromide nuclear staining and GFAP immunofluorescence, was not significantly augmented by dopamine-H2O2 exposure.88 On the basis of these

Glial Iron Sequestration and Neurodegeneration

243

Fig. 11.5. ESR spectra of magnesium-complexed semiquinones from the peroxidase-H2O2 oxidation of 2-hyd rox y ( c a te ch o l ) e s t r a d i o l . (A) Top: autoxidation of 2hydroxyestradiol in serum-free medium in the absence of cells following alkalinization to pH 10.0 with NaOH. Medium contained 2-hydroxyestradiol (10–2 M), MgCl2 (0.5 M) and NaOH in DMEM. The characteristic osemiquinone spectrum of oxidized 2-hydroxyestradiol is shown. Bottom: computersimulated spectrum of the 2hydroxyestradiol o-semiquinone derived from measured hyperfine coupling constants. (B) Incubation of 2-hydroxyestradiol (10 –2 M), MgCl 2 (0.5 M), NADPH (0.3 M), and H2O2 (0.1 mM) with tissue homogenate derived from untreated (control) brain cell culture (pH 7.0). The gain settings in (B) and (C) are identical, permitting direct amplitude comparisons. (C) Incubation as in (B) with tissue homogenate derived from cysteamine pretreated (peroxidase-enriched) brain cell culture. An intense osemiquinone signal is observed with hyperfine structure identical to the pattern obtained in the cell-free 2-hydroxyestradiol autoxidation experiment (A). The peroxidase activity induced in astrocytes by cysteamine catalyses catechol oxidation to o-semiquinone radicals. Reprinted with permission from Schipper HM et al, J Neurosci 1991; 11:2170.

in vitro findings, we hypothesize that, in the intact basal ganglia, leakage of free radicals from peroxidase-positive astrocytes into the surrounding neuropil may promote lipid peroxidation and degeneration of nearby dopaminergic terminals and other vulnerable neuronal constituents. In this regard, the progressive increase in numbers of peroxidasepositive astrocytes which have been documented in the basal ganglia and other subcortical regions of the aging rodent and human brain may render the latter particularly prone to Parkinsonism and other free radical-related neurodegenerations.

Pathological Glial-Neuronal Interaction in Parkinson’s Disease Our observations on CSH-stressed astroglia suggest a model for inclusion formation, iron sequestration, and the perpetuation of oxidative injury in the aging and degenerating nervous system (Fig. 11.7):

Astrocytes in Brain Aging and Neurodegeneration

244 Fig. 11.6. ESR spectra of magnesiumcomplexed semiquinones derived from the peroxidase-H 2 O 2 oxidation of dopamine. (A) Top: Autoxidation of dopamine in serum-free medium in the absence of cells at pH 10.0. Medium contained dopamine (1 mM), MgCl2 (0.2 M), and NaOH in DMEM. The characteristic dopamine-o-semiquinone spectrum is observed. Bottom: Computer-simulated spectrum of the dopamine-o-semiquinone derived from the measured hyperfine coupling constants. (B) Incubation of dopamine (1 mM), MgCl2 (0.2 M), NADPH (0.3 M), and H2O2 (0.1 mM) with tissue homogenate derived from an untreated (control) astrocyte culture (pH 7.0). The gain settings in (B) and (C) are identical. (C) Incubation as in (B) with tissue derived from cysteaminepretreated (peroxidase-enriched) astrocyte culture. ESR spectra amplitudes are approximately 2.5-fold greater than those observed in (B). The cysteamine-induced peroxidase activity catalyzes catechol oxidation to osemiquinone radicals. Reprinted with permission from Schipper HM et al, J Neurosci 1991; 11:2170.

1. In the senescent basal ganglia and other subcortical brain regions, dopamine and/ or other unidentified oxidative stressors (simulated by CSH exposure) induce a cellular stress response in subpopulations of astroglia, characterized by upregulation of various HSP and HO-1. Free iron and CO derived from HO-1-mediated heme degradation may initiate or potentiate injury to the mitochondrial compartment. 2. Stress-related inhibition of porphyrin-heme biosynthesis and/or direct oxidative damage to mitochondrial membranes promotes the selective transport of nontransferrin-derived, nonheme iron into the mitochondrial compartment. Com-

Glial Iron Sequestration and Neurodegeneration

245

Fig. 11.7. A model for glial inclusion formation, iron sequestration, and oxidative injury in the aging and degenerating nervous system.

pensatory upregulation of MnSOD may provide some degree of protection to the mitochondrial compartment by limiting the accumulation of superoxide. 3. By promoting further oxidative stress, the redox-active mitochondrial iron participates in a vicious cycle of pathologic events whereby damage to glial mitochondria as well as to the surrounding neuropil is perpetuated. This model of astrocyte senescence is consistent with the Mitochondrial Hypothesis of Aging, which states that oxidative damage to mitochondria results in bioenergetic failure, a vicious spiral of augmented mitochondrial free radical generation and injury and progressive

Astrocytes in Brain Aging and Neurodegeneration

246

tissue aging.19,89-91 Our model also accounts for the observation that mosaicism for specific mitochondrial DNA mutations in the normal aging human brain is most striking in regions particularly rich in intracellular iron such as the caudate, putamen and substantia nigra.92 Our findings recapitulate the discordant pattern of iron/ transferrin receptor localization observed in the PD nigra (see above) and raise the possibility that exacerbation of stress-related trapping of nontransferrin-derived iron by astroglial mitochondria may be an important mechanism underlying the pathological accumulation of this redox-active metal in the basal ganglia of PD subjects. Such iron could conceivably originate from degenerating neurons, glia, or myelin or from the cerebrospinal fluid (CSF). Micromolar quantities of chelatable, low-molecular-weight iron are present in normal CSF, and the concentration of this metal in CSF has been shown to increase under neuropathological conditions.93 As described above, a portion of this chelatable iron may be derived from HO-1mediated degradation of cellular heme within oxidatively-challenged neural tissues. Consistent with our model are reports that a significant proportion of the excess iron in PD brain may indeed be localized to astroglial mitochondria,37,94,95 and that deficiencies of mitochondrial electron transport are prevalent in the brains of PD subjects.77,78 By oxidizing dopamine and environmentally-derived xenobiotics to neurotoxic intermediates, the redox-active glial iron could serve as a “final common pathway” perpetuating nigrostriatal degeneration initiated by as yet undetermined genetic and epigenetic factors in patients with PD.

Conclusion There is considerable evidence implicating excessive basal ganglia iron and catecholamine-derived free radicals in the pathogenesis of idiopathic PD. Yet, the mechanisms responsible for the pathological sequestration of brain iron in this and other debilitating neurodegenerative conditions remain enigmatic. The progressive accumulation of iron-rich (peroxidase-positive) astrocytic granules represents a fundamental and highly consistent biomarker of aging in the vertebrate CNS. Although these glial inclusions were first identified almost half a century ago on the basis of their affinity for Gomori stains, it is only in recent years, and largely through the advent of in vitro toxicologic modeling of inclusion biogenesis, that we have begun to elucidate the subcellular origin of these inclusions, the mechanism(s) governing their formation, and their potential role in brain aging and neurodegeneration. The current state of our knowledge indicates that these gliosomes are “stress granules” which ultimately derive from effete, metal-laden mitochondria engaged in a complex autophagic process. Determination of the topography of these glial inclusions may permit “mapping” of CNS regions at increased risk for chronic oxidative injury during normal aging and under pathological conditions. More importantly, the ability to experimentally recapitulate the development of this senescent glial phenotype in primary culture provides a powerful model to investigate: 1. the role of HO-1 and other heat shock proteins in the biogenesis of potentially deleterious neural inclusions; 2. stress-related (dys)regulation of MnSOD and other antioxidant enzymes in the aging and degenerating nervous system; and 3. mechanisms of pathological brain iron sequestration and mitochondrial insufficiency characteristic of aging and degenerating neural tissues. Our findings support the notion that stress-related trapping of nonheme, nontransferrin-bound iron by astroglial mitochondria is a primary mechanism underlying the pathological accumulation of redox-active iron in the basal ganglia of PD subjects. We have provided evidence that the nonenzymatic peroxidase activity manifest in senescent,

Glial Iron Sequestration and Neurodegeneration

247

iron-laden astroglia promotes the bioactivation of endogenous catechols and environmentally-derived xenobiotics to potential neurotoxins which, in turn, may perpetuate “secondary” neural damage long after initiating neurotoxic insults have dissipated. If the latter is true, attempts to pharmacologically inhibit metal sequestration by “stressed” astroglial mitochondria (e.g., using HO-1 inhibitors and centrally-active iron chelators) may constitute a rational and effective strategy in the management of Parkinson’s disease and other agingrelated neurodegenerative afflictions.

Acknowledgments The authors thank Mrs. Kay Berckmans and Mrs. Adrienne Liberman for assistance with the preparation of this manuscript. This work is supported by grants from the Medical Research Council of Canada and the Fonds de la Recherche en Santé du Québec.

References 1. Hornykiewicz O. Neurochemical pathology and etiology of Parkinson’s disease: Basic facts and hypothetical possibilities. Mt Sinai J Med 1988; 45:35-43. 2. Calne DB. Initiating treatment for idiopathic Parkinsonism (Review). Neurology 1994; 44(7 Suppl 6):S19-22. 3. Cohen G, Werner P. Free radicals, oxidative stress, and neurodegeneration. In: Calne DB, ed. Neurodegenerative Diseases. Philadelphia: WB Saunders Co., 1994:139-161. 4. Fahn S, Cohen G. The oxidant stress hypothesis in Parkinson’s disease; evidence supporting it (Review). Ann Neurol 1992; 32:804-812. 5. Jenner P. What process causes nigral cell death in Parkinson’s disease. In: Cederbaum JM, Gancher ST, eds. Neurological Clinics, Part 2: Parkinson’s Disease. Philadelphia: Saunders Co., 1992:387-403. 6. Youdim MBH. Inorganic neurotoxins in neurodegenerative disorders without primary dementia. In: Calne DB, ed. Neurodegenerative Diseases. Philadelphia: Saunders WB Co., 1994:251-276. 7. Hefti F, Melamed E, Wurtman R. Partial lesions of the dopanergic nigrostriatal system in the rat brain: Biochemical characterization. Brain Res 1980; 195:123-137. 8. Zigmond M, Berger TW, Grace AA, Stricker BM. Compensatory responses to nigrostriatal bundle injury. Studies with 6-hydroxydopamine in an animal model of Parkinsonism (Review). Mol Chem Neuropath 1989; 10:185-200. 9. Barbeau A. Etiology of Parkinson’s disease: A research strategy. Can J Neurol Sci 1984; 11:24-28. 10. Cadet JL, Katz M, Jackson-Lewis V, Fahn S. Vitamin E attenuates the toxic effects of intrastriatal injection of 6-hydroxydopamine (6-OHDA) in rats: Behavioral and biochemical evidence. Brain Res 1989; 476:10-15. 11. Cohen G. Oxygen radicals and Parkinson’s disease. Upjohn symposium on Oxygen Radicals, Augusta, MI 1987; 130-135. 12. Parenti M, Rusconi L, Cappabianca V et al. Role of dopamine in manganese neurotoxicity. Brain Res 1988; 473:236-240. 13. Dexter DT, Carter CJ, Wells FR et al. Basal lipid peroxidation in substantia nigra is increased in Parkinson’s disease. J Neurochem 1989; 52:381-389. 14. Ambani LM, Van Woert MH, Murphy S. Brain peroxidase and catalase in Parkinson disease. Arch Neurol 1975; 32:114-118. 15. Perry TL, Godin DV, Hansen S. Parkinson’s disease: A disorder due to nigral glutathione deficiency? Neurosci Lett 1982; 33:305-310. 16. Kalra J, Rajput AH, Mantha SV et al. Oxygen free radical producing activity of polymorphonuclear leukocytes in patients with Parkinson’s disease. Mol Cell Biochem 1992; 112:181-186. 17. Riederer P, Sofic E, Rausch WD et al. Transition metals, ferritin, glutathione, and ascorbic acid in Parkinsonian brains. J Neurochem 1989; 52:515-520.

248

Astrocytes in Brain Aging and Neurodegeneration

18. Olanow CW, Hauser RA, Gauger L et al. The effects of deprenyl and levodopa on the progression of Parkinson’s disease. Ann Neurol 1995; 38:771-777. 19. Parkinson Study Group. Effects of deprenyl on the progression of disability in early Parkinson’s disease. N Engl J Med 1989; 321:1364-1371. 20. Parkinson Study Group. Effects of tocopherol and deprenyl on the progression of disability in early Parkinson’s disease. N Engl J Med 1993; 328:176-183. 21. Selkoe DJ. The molecular pathology of Alzheimer’s disease. Neuron 1991; 6:487-498. 22. Subbarao KV, Richardson JS, Ang LC. Autopsy samples of Alzheimer’s cortex show increased peroxidation in vitro. J Neurochem 1990; 55:342-345. 23. Palmer AM, Burns MA. Selective increase in lipid peroxidation in the inferior temporal cortex in Alzheimer’s disease. Brain Res 1994; 645:338-342. 24. Perrin R, Briancon S, Jeandel C et al. Blood activity of Cu/Zn superoxide dismutase, glutathione peroxidase and catalase in Alzheimer’s disease. A case-control study. Gerontology 1990; 36:306-313. 25. Jeandel C, Nicolas MB, Dubois F et al. Lipid peroxidation and free radical scavengers in Alzheimer’s disease. Gerontology 1989; 35:275-282. 26. Zemlan FP, Thienhaus OJ, Bosmann HB. Superoxide dismutase activity in Alzheimer’s disease: Possible mechanism for paired helical filament formation. Brain Res 1989; 476:160-162. 27. Dyrks T, Weidemann A, Multhaup G et al. Identification, transmembrane orientation and biogenesis of the amyloid A4 precursor of Alzheimer’s disease. EMBO J 1988; 7:949-957. 28. Butterfield DA, Hensley K, Harris M et al. β-Amyloid peptide free radical fragments initiate synaptosomal lipoperoxidation in a sequence-specific fashion: Implications to Alzheimer’s disease. Biochem Biophys Res Commun 1994; 200:710-715. 29. Hensley K, Carney JM, Mattson MP et al. A model for β-amyloid aggregation and neurotoxicity based on free radical generation by the peptide: Relevance to Alzheimer disease. Proc Natl Acad Sci USA 1994; 91:3270-3274. 30. Leblanc A. Increased production of 4 kDa amyloid beta peptide in serum deprived human primary neuron cultures: possible involvement of apoptosis. J Neurosci 1995; 15:7837-7846. 31. Beal MF. Does impairment of energy metabolism result in excitotoxic neuronal death in neurodegenerative illnesses? (Review) Ann Neurol 1992; 31:119-130. 32. Schagger H, Diekmann S, Ohm TG. Oxphos protein defects in Alzheimer’s disease, in Parkinson’s disease and in mitochondrial encephalomyopathies revealed by blue native electrophoresis. Soc Neurosci Abstract 1995; 21:1492. 33. Gonzalez-Lma F, Matos-Collazo S, Garrosa M. Cytochrome oxidase activity in the human inferior calliculus: Quantification at cellular and tissue levels and comparison to Alzheimer’s brains. Soc Neurosci. Abstract 1995; 21:739. 34. Fukuyama R, Hatanpaa K, Rapoport SI, Chandrasekaran K. Gene expression of ND4, a subunit of complex I of oxidative phosphorylation in mitochondria, is decreased in temporal cortex of brain of Alzheimer’s disease patients. Brain Res 1996; 713:290-293. 35. Davis RE, Miller S, Herrnstadt C et al. Mutations in mitochondrial cytochrome-c oxidase genes segregate with late-onset Alzheimer disease. Proc Natl Acad Sci USA 1997; 94:4526-4531. 36. Benzi G, Moretti A. Are reactive oxygen species involved in Alzheimer’s disease? Neurobiol Aging 1995; 16:661-674. 37. Jellinger P, Paulus W, Grundke-Iqbal I et al. Brain iron and ferritin in Parkinson’s and Alzheimer’s diseases. J Neural Transm Park Dis Dement Sect 1990; 2:327-340. 38. Youdim MBH, Ben-Shachar D, Riederer P. The role of iron in etiopathology of Parkinson’s disease. Mov Disord 1993; 8:1-12. 39. Janetzky B, Reichman H, Youdim MBN et al. Iron and oxydative damage in neurodegenerative diseases. In: Beal MF, Howell N, Bodis-Walker I, eds. Mitochondria and Free Radicals in Neurodegenerative Diseases. New York: Wiley-Liss 1997; 20:407-421. 40. Good PF, Olanow CW, Perl DP. Neuromelanin-containing neurons of the substantia nigra accumulate iron and aluminum in Parkinson’s disease: A LAMMA study. Brain Res 1992; 593:343-346.

Glial Iron Sequestration and Neurodegeneration

249

41. Morris CM, Edwardson JA. Iron histochemistry of the substantia nigra in Parkinson’s disease. Neurodegen 1994; 3:277-282. 42. Gelman BB. Iron in CNS disease. J Neuropath Exp Neurol 1995; 54:477-486. 43. Thiel EC. Regulation of ferritin and transferrin receptor mRNAs. J Biol Chem 1990; 265:4771-4774. 44. Klausner RD, Rouault TA, Harford JB. Regulating the fate of mRNA: The control of cellular iron metabolism. Cell 1993; 72:19-28. 45. Ponka, P. Physiology and pathophysiology of iron metabolism: Implications for iron chelation therapy in iron overload. In: Bergeron RJ, Brittehham GM, eds. The Development of Iron Chelators for Clinical Use. Boca Raton: CRC Press, 1994. 46. Faucheux BA, Mirch EC, Villares T et al. Distribution of 125I-ferrotransferrin binding sites in the mesencephalon of control subjects and patients with Parkinson’s disease. J Neurochem 1993; 60:2338-2341. 47. Dwork AJ, Schon EA, Herbert J. Nonidentical distribution of transferrin and ferric iron in human brain. Neurosci 1988; 27:333-345. 48. Mash DC, Sanchez-Ramos J, Weiner WJ. Transferrin receptor regulation in Parkinson’s disease and MPTP-treated mice. In: Narabayashi H, Nagatsu T, Xanagisawa N, Maurino Y, eds. Advances in Neurology. New York: Raven Press Ltd., 1993; 60:133-139. 49. Mash DC, Pablo J, Buck BE et al. Distribution and number of transferrin receptors in Parkinson’s disease and in MPTP-treated mice. Exp Neurol 1991; 114:73-81. 50. Kalaria RN, Sromek SM, Grahovac I, Harik SI. Transferrin receptors of rat and human brain and cerebral microvessels and their status in Alzheimer’s disease. Brain Res 1992; 87-93. 51. Faucheux BA, Nillesse N, Damier P et al. Expression of lactoferrin receptors is increased in the mesencephalon of patients with Parkinson disease. Proc Natl Acad Sci USA 1995; 92:9603-9607. 52. Leveugle B, Fauxheux BA, Bouras C et al. Cellular distribution of the iron-binding protein lactotransferrin in the mesencephalon of Parkinson’s disease cases. Acta Neuropath 1996; 91:566-572. 53. Connor JR, Menzies SL, St. Martin SM, Mufson EJ. A histochemical study of iron, transferrin, and ferritin in Alzheimer’s diseased brains. J Neurosci Res 1992; 31:75-83. 54. Morris CM, Kerwin JM, Edwarsan JA. Non-haem iron histochemistry of the normal and Alzheimer’s disease hippocampus. Neurodegen 1994; 3:267-275. 55. Kawamata T, Tooyama I, Yamada T et al. Lactotransferrin immunocytochemistry in Alzheimer and normal human brains. Am J Pathol 1993; 142:1574-1585. 56. Leveugle B, Spik G, Perl DB et al. The iron binding protein lactotransferrin is present in pathologic lesions in a variety of neurodegenerative disorders: A comparative immunohistochemical analysis. Brain Res 1994; 650:20-31. 57. Osmand AP, Switzer RC III. Differential distribution of lactoferrin and Alz-50 immunoreactivities in neuritic plaques and neurofibrillary tangles in Alzheimer’s disease. In: Iqbal K, McLachlan RC, Winblad B, Wisniewski HM, eds. Alzheimer’s Disease: Basic Mechanisms, Diagnosis and Therapeutic Strategies. New York: Wiley 1991:219-227. 58. Rebeck GW, Harr SD, Strickland KD, Hyman BT. Multiple, diverse senile plaque-associated proteins are ligands of an apolipoprotein E receptor, the alpha-macroglobulin receptor/low density lipoprotein receptor-related protein. Ann Neurol 1995; 37:211-217. 59. Goldgefter L, Schejter AS, Gill D. Structural and microspectrofluorometric studies on glial cells from the periventricular and arcuate nuclei of the rat hypothalamus. Cell Tissue Res 1980; 211:503. 60. Schipper HM. Gomori-positive astrocytes: Biological properties and implications for neurologic and neuroendocrine disorders. Glia 1991; 4:365. 61. Brawer JR, Reichard G, Small L, Schipper HM. The origin and composition of peroxidasepositive granules in cysteamine-treated astrocytes in culture. Brain Res 1994; 633:9-20. 62. Wang X, Manganaro F, Schipper HM. A cellular stress model for the sequestration of redox-active glial iron in the aging and degenerating nervous system. J Neurochem 1995; 64:1868-1877.

250

Astrocytes in Brain Aging and Neurodegeneration

63. Duchen MR, Biscoe TJ. Mitochondrial function in type I cells isolated from arterial chemoreceptors. J Physiol 1992:450:13-31. 64. Kohler M, Fromter E. Identification of mitochondrial-rich cells in unstained preparation of epithelia by autofluorescence. Eur J Physiol 1985; 403:47-49. 65. Richardson DR, Baker E. The uptake of inorganic iron complexes by human melanoma cells. Biochim Biophys Acta 1991; 1093:20-28. 66. Richardson DR, Baker E. The uptake of iron and transferrin by human malignant malanoma cells. Biochim Biophys Acta 1990; 1053:1-12. 67. Richardson DR, Baker E. Two mechanisms of iron uptake from transferrin by melanoma cells. The effects of desferrioxamine and ferric ammonium citrate. J Biol Chem 1992; 267:13972-13979. 68. Chan RYY, Ponka P, Schulman HM. Transferrin-receptor-independent but iron-dependent proliferation of variant Chinese hamster ovary cells. Exp Cell Res 1992; 202:326-336. 69. Inman RS, Wessling-Resnick M. Characterization of transferrin-independent iron transport in K562 cells. Unique properties provide evidence of multiple pathways of iron uptake. J Biol Chem 1993; 268:8521-8528. 70. Adams ML, Ostapiuk I, Grasso JA. The effects of inhibition of heme synthesis on the intracellular localization of iron in rat reticulocytes. Biochim Biophys Acta 1989; 1012:243-253. 71. Schipper HM, Bernier L, Bernatchez G. Pathological glial-neuronal interaction in Parkinson’s disease. Soc Neuro Abst 1996; 22:219. 72. He Y, Thong PSP, Lee T et al. Increased iron in the substantia nigra of 6-OHDA induced Parkinsonian rats: A nuclear microscopy study. Brain Res 1996; 735:149-153. 73. Young JK, McKenzie JC, Baker JH. Association of iron-containing astrocytes with dopaminergic neurons of the arcuate nucleus. J Neurosci Res 1990; 25:204-213. 74. Schipper HM, Cissé S, Stopa EG. Expression of heme oxygenase-1 in senescent and Alzheimer-diseased brain. Ann Neurol 1995; 37:758-768. 75. Zhang J, Piantadosi CA. Mitochondrial oxydative stress after carbon monoxide hypoxia in the rat brain. J Clin Invest 1992; 90:1193-1199. 76. Premkumar DRD, Smith MA, Pichey PL et al. Induction of heme oxygenase-1 mRNA and protein in neocortex and cerebral vessels in Alzheimer’s disease. J Neurochem 1995; 65:1399-1402. 76a. Schipper HM, Liberman A, Stopa EG. Neural heme oxygenase-1 expression in idiopathic Parkinson's disease. Exp Neurol 1998; 150:60-68. 77. Beal MF. Mitochondrial Dysfunction and Oxidative Damage in Neurodegenerative Diseases. Austin: RG Landes, 1995:1-128. 78. Reichmann H, Riederer P. Mitochondrial disturbances in neurodegeneration. In: Calne DB, ed. Neurodegenerative Diseases. Philadelphia: Saunders 1994:195-204. 79. Chopra VS, Chalifour LE, Schipper HM. Differential effects of cysteamine on heat shock protein induction and cytoplasmic granulation in astrocytes and glioma cells. Mol Brain Res 1995; 31:173-184. 80. Manganaro F, Chopra VS, Mydlarski MB et al. Redox perturbations in cysteamine-stressed astroglia: Implications for inclusion formation and gliosis in the aging brain. Free Rad Biol Med 1995; 19:823-835. 81. Mydlarski MB, Liberman A, Schipper HM. Estrogen induction of glial heat shock proteins: Implications for hypothalamic aging. Neurobiol Aging 1995:16:977-981. 82. Bernier L, Bernachez G, Schipper HM. Oxidative stress promotes mitochondrial iron sequestration in cultured astroglia. Soc Neuroscience Abst 1996; 22:1495. 83. Schipper HM, Kotake Y, Janzen EG. Catechol oxidation by peroxidase-positive astrocytes in primary culture: An electron spin resonance study. J Neurosci 1991; 11:2170. 84. Metodiewa D, Reszka K, Dunford H. Evidence for a peroxidatic oxidation of norepinephrine, a catecholamine, by lactoperoxidase. Biochem Biophys Res Commun 1989; 160:1183. 85. DiMonte DA, Schipper HM, Hetts S, Langston JW. Iron-mediated bioactivation of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) in glial cultures. Glia 1995; 15:203-206. 86. Newcomer TA, Rosenberg PA, Aizenman E. Iron-mediated oxidation of 3,4-dihydrophenylalanine to an excitotoxin. J Neurochem 1995; 64:1742-1748.

Glial Iron Sequestration and Neurodegeneration

251

87. Kontos H, Wei W, Ellis E et al. Appearance of superoxide anion radical in cerebral extracellular space during increased prostaglandin synthesis in cats. Cir Res 1985; 57:142. 88. Frankel D, Schipper HM. Does astroglial senescence facilitate oxidative neuronal injury? Soc Neurosci Abstr 1997; 23:1371. 89. Bowling AC, Mutisya EM, Walker LC et al. Age-dependent impairment of mitochondrial function in primate brain. J Neurochem 1993; 60:1964-1967. 90. Linnane AW, Zhang C, Baumer A, Nagley P. Mitochondrial DNA mutation and the aging process: Bioenergy and pharmacological intervention. Mutat Res 1992; 275:195-208. 91. Miquel J, Fleming J. Theoretical and experimental support for an “oxygen radical-mitochondrial injury” hypothesis of cell aging. In: Johnson JE Jr, ed. Free Radicals, Aging, and Degenerative Diseases. New York: Alan R. Liss, 1986:51-74. 92. Shigenaga MK, Hagen TM, Ames BN. Oxidative damage and mitochondrial decay in aging. Proc Natl Acad Sci USA 1994; 91:1077-10778. 93. Soong NW, Hinton DR, Cortopassi G, Arnheim M. Mosaicism for specific somatic mitochondrial DNA mutation in adult human brain. Nat Genet 1992; 2:318-323. 94. Gutteridge JM. Iron and oxygen radicals in brain. Ann Neurol (Suppl.) 1992; 32:S16-S21. 95. Connor JR, Menzles S, St. Martin SM, Mufson EJ. Cellular distribution of transferrin, ferritin and iron in normal and aged human brains. J Neurosci 1990; 27:595-611. 96. Olanow CW. Magnetic resonance imaging in Parkinsonism. In: Cederbaum JM, Gancher ST, eds. Neurological Clinics Par 2, Vol. 10. Philadelphia: Saunders, 1992; 405-420.

Index A αB-crystallin 78, 79, 210, 212-215, 217

Achromatic (ballooned) neuron 176, 180, 210, 213 Acyl-CoA binding protein 200 Adenosine triphosphate (ATP) 17, 24, 25, 101, 114, 115, 119, 174, 207, 217 Adhesion molecule 5, 20, 29, 42, 46, 47, 49, 50, 51, 75, 92 Aging 71, 74-78, 93, 104, 111, 117-119, 121, 165, 166, 171, 191, 193-196, 201, 202, 207-209, 211-215, 217, 219-223, 228, 236-238, 242, 243, 245-247 Aicardi’s syndrome 103 Alexander’s disease 78-80, 82, 99, 210 Alpha-1-antichymotrypsin 46, 49, 50, 92 Alzheimer type II astrocytes/cells 43, 52 Alzheimer’s disease (AD) 45, 46, 49, 50, 73, 74, 91-95, 99, 101, 104, 137, 147, 148, 165-167, 169-171, 173, 176, 177, 182, 209, 210, 212, 213, 235-237, 241 Ammonia 41, 43, 51-53 AMPA/kainate receptor 18 Amyloid (Aβ) 42, 46, 49, 50, 73, 77, 91-93, 95, 96, 98, 129, 135, 137, 147, 166, 177, 182, 235-237 Amyotrophic lateral sclerosis (ALS) 22, 46, 51, 52, 73, 74, 169, 170, 181, 210, 213, 237 Anchorage densities 91, 99, 103, 104 Antioxidant 43, 196, 202, 208, 219, 221, 235, 246 Apolipoprotein E (Apo E) 49, 50, 92, 148 Arcuate nucleus 192-196, 201, 202, 216, 217, 219, 239 Argyrophilic inclusions 174 Ascorbate 238 Aspartate 19, 21, 44, 115, 119 Astrocyte 3-9, 15-31, 41, 43-56, 71, 73-79, 81, 82, 91-93, 95, 97, 99, 101-104, 111, 112, 114-122, 137, 139, 140, 142, 143, 145-158, 168, 169, 174, 177, 180-187, 197-209, 212, 213, 215, 217, 219-223, 228-231, 233, 234, 242, 245, 249-251 Astrocytoma 78, 210 Astrogliosis (gliosis) 22, 28, 30, 42, 43, 46-51, 71, 73, 74, 76, 78, 92, 93, 95, 99, 104, 111, 129-131, 135, 143, 147, 165, 169, 171, 175, 181, 195, 200, 207, 210, 211, 221, 228, 235

Autofluorescence 192, 196, 198, 202, 223, 224, 227, 238 Autoxidation 116, 118, 195, 219, 220, 243, 244

B Bergmann glia 4, 8, 9, 17, 20, 22-24, 26, 131, 139, 142, 208 Blood-brain barrier (BBB) 7, 28, 29, 41, 50, 75, 112, 121, 192, 196 Bone morphogenic proteins (BMP) 6 Bovine spongiform encephalopathy (BSE) 127, 135, 137, 142, 143, 150, 151, 154

C Carbon monoxide (CO) 239, 241, 244 Catalase 42, 192, 220, 235 Catecholamine 114, 116, 118, 120, 196, 202, 242, 246 Catecholestrogen 195, 196, 202, 219, 242 Cathepsin 46, 50, 148, 201, 202 Chemokines 42, 45, 46, 49, 75 Chronic wasting disease (CWD) 127, 137, 143 Ciliary neurotrophic factor (CNTF) 6, 71 Coiled bodies 170, 173-178, 180 Copper 77, 199, 200 Corpora amylacea (CA) 77, 78, 91, 99, 104, 171, 174, 191, 202, 207, 210, 215, 221-224, 227, 228, 238 Corticobasal degeneration (CBD) 99, 169, 170, 174, 176-180 Creutzfeldt-Jakob disease (CJD) 74, 127, 129-133, 135-137, 140-143, 145-147, 149-154, 165 Cysteamine 76, 197-200, 202, 210, 224, 238, 243, 244 Cytochrome C oxidase 236 Cytokine 28-30, 42, 43, 45, 46, 48-50, 71, 74, 75, 92, 93, 120, 143, 149-154 Cytoprotection 220, 221, 241

D Dopamine 112, 113, 115-118, 121, 235, 238, 241, 242, 244, 246 Down’s syndrome (DS) 73, 237

254

Astrocytes in Brain Aging and Neurodegeneration

E

H

Electron spin resonance (ESR) spectroscopy 242-244 Endothelin 23, 92, 93 Eosinophilic inclusions 51, 91, 99, 101-103 Epidermal growth factor (EGF) 6, 23 Epilepsy 50, 51 Estrogen 194, 195, 202, 211, 214, 217, 219, 224 Experimental allergic encephalomyelitis (EAE) 30, 45, 49, 73, 75, 149, 150 Extracellular matrix molecules (ECM) 45, 47, 49

Heat shock factor (HSF) 207, 209 Heat shock protein (HSP) 71, 77-79, 149, 207-215, 217, 219, 221-223, 225, 227, 244, 246 Heme 196, 207, 220, 221, 227, 237-239, 241, 242, 244, 246 Heme oxygenase-1 (HO-1) 207, 208, 210-212, 214, 215, 219, 221, 223, 227, 239, 241, 242, 244, 246, 247 Hepatic encephalopathy (HE) 43, 52, 53, 74 Hippocampus 23-25, 52, 75-77, 139, 140, 146, 192, 195, 200, 202, 209, 211-213, 235, 237, 239 HIV-1 29, 44, 48 Huntington’s disease 73, 180 Hydrogen peroxide (H2O2) 76, 92, 112, 115-118, 195, 211, 215, 217, 219-222, 235, 241-244 Hypothalamus 4, 45, 74, 76, 192-196, 219

F Fatal familial insomnia (FFI) 127, 129 Fatty acid binding protein (FABP) 200 Ferritin 77, 142, 236, 237 Fibrils 50, 73, 77, 93, 95, 97, 99, 104, 112, 131, 137, 182 Fibrous astrocyte 77, 177 Frontotemporal dementia (FTDP-17) 169, 180

G Gemistocytic astrocytes 133, 136 Gerstmann-Straussler-Scheinker disease (GSS) 73, 74, 127, 129, 130, 135-140, 142, 154 Glial cytoplasmic inclusion (GCI) 78, 165, 170, 174, 180, 181, 210 Glial derived neurotrophic factor (GDNF) 121 Glial fibrillary acidic protein (GFAP) 3, 4, 6, 7, 15, 18, 24-26, 42, 48-50, 53, 71-79, 93, 101, 116, 129, 130, 132, 133, 139, 142, 145-148, 152, 153, 166, 171, 191, 192, 196, 210-213, 216, 217, 219, 221, 239, 241, 242 Glucose-regulated protein (GRP) 207, 211, 214, 215, 219 Glutamate 19, 21-26, 29, 30, 41-45, 48, 50-53, 92, 119, 120, 147 Glutamine 3, 21, 26, 44, 52, 147 Glutathione (GSH) 42, 112, 115, 116, 121, 208, 235 Glycogen 4, 42, 43, 44, 53, 77, 119, 130, 223 Glycosylation 127, 223, 227 Gray matter 3, 4, 5, 7, 8, 74, 76, 91, 93, 138, 140, 142, 145, 170, 173, 175, 177, 179 Guam Parkinson-dementia complex (GPDC) 169, 170, 175, 237

I Interferon-γ (IFN-γ) 29-31, 46, 74, 150, 151 Interleukin-1 (IL-1) 28-31, 45, 46, 50, 71, 75, 92, 150-153, 168 Ion channel 16, 18, 19, 25 Iron (Fe) 77, 112, 118, 121, 192, 193, 196, 197, 199, 200, 202, 219, 220, 223, 225, 227, 236-247 Ischemia 19, 21, 42, 44, 46, 52, 95, 168, 208

K Kuru 74, 127, 128, 130, 131, 135, 137-140, 142

L L-Dopa 112, 115, 238 Lactic acid 43, 44, 45, 52 Lactoferrin 237 Lafora disease 104 Lipid peroxidation 195, 220, 235, 237, 243 Lipofuscin 130, 191, 192, 202, 211 Lymphocyte 30, 74, 75, 143, 150 Lysosome 201, 202, 215, 223, 227

Index

255

M

P

Manganese superoxide dismutase (MnSOD) 219, 221, 235, 242, 245, 246 Metalloproteinase 46, 50 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) 51, 111, 113-121, 235, 242 Microglia 28-31, 41, 44-50, 53, 71, 74-77, 91-95, 112, 137-140, 142, 143, 149, 150, 153, 194, 210, 236, 237 Microtubule associated protein (MAP) 78, 168 Mitochondria 42, 43, 52, 53, 96, 98, 112-115, 119, 122, 196, 197, 200, 202, 214-217, 219, 220, 222, 223, 225, 227, 228, 235, 236, 238-242, 244-247 Monoamine oxidase (MAO) 42, 51, 112, 114118, 121, 235, 242 Monocyte/macrophage 74, 75 Muller glia 4 Multiple sclerosis (MS) 30, 45, 46, 48, 49, 71, 72, 75, 76, 78, 143, 145, 149, 150, 153, 209 Multiple system atrophy (MSA) 78, 169, 170, 174, 180, 181, 209, 210

Paired helical filaments (PHF) 78, 91, 95, 97, 99, 100, 167-169, 171, 177, 179, 180, 210, 213, 236 Parkinson’s disease (PD) 51, 73, 74, 111-113, 116-122, 165, 166, 181, 209, 210, 213, 219, 235-239, 241, 243, 246, 247 PC12 cells 242 Peroxidase-positive astrocyte 191-197, 200, 211, 214, 215, 217, 223, 242, 243 Phosphorylation 29, 53, 92, 99, 112, 119, 167, 169, 220 Pick bodies 176, 177, 210, 213 Pick’s disease 74, 99, 169, 170, 175-177, 180, 210, 213, 237 Pilocytic astrocytes 136 Porphyrin 238, 239, 244 Potassium (K+) 3, 16-21, 23-27, 29, 30, 41, 43, 44, 48, 51, 74-76 Prion protein (PrP) 74, 127, 129, 135, 137, 142, 146-149, 153, 154 Progressive supranuclear palsy (PSP) 78, 99, 104, 169-171, 173-177, 180, 210, 212, 213 Protoplasmic astrocyte 15, 74 Pyruvate dehydrogenase 236

N Nerve growth factor (NGF) 46, 71, 121 Neuritic plaque 45, 46, 49, 50, 73, 91-93, 137, 147, 177, 210, 237 Neurofibrillary tangles (NFT) 73, 95, 99, 166, 167, 169, 171, 173, 175, 176, 209, 210, 235, 237 Neuropil threads 173 Neurotoxin 48, 52, 117, 118, 121, 235, 247 Nitric oxide (NO) 29, 30, 42, 43, 45, 46, 48-50, 52, 75, 92, 115, 120, 121 NMDA receptor 24, 44, 45, 48, 114, 115, 119, 120 Nuclear inclusions 131, 181

Q

O

S-100b 3, 50, 71, 92 Scrapie 74, 127, 135, 137, 139-142, 146-148, 150, 151, 154 Semiquinone 195, 196, 219, 242-244 Stress granule 211, 223, 246 Substantia nigra 73, 112, 115, 116, 118-121, 165, 192, 193, 209, 212, 213, 219, 235-239, 246 Subventricular zone (SVZ) 4-7, 9

Oligodendrocyte 6-8, 41, 43, 47, 49, 74, 76, 78, 91, 99, 143-145, 149, 150, 166, 168, 170, 171, 173, 177-179 Oxidative stress 52, 76-78, 112, 113, 115, 116, 118, 121, 122, 207, 208, 214, 215, 217, 219, 220, 223, 235-239, 244, 245

Quinolinic acid 29, 42, 44, 48, 51

R Radial glia 4-6 Reactive oxygen species 50, 77, 118, 195, 196, 207, 220, 221, 236, 242 Rosenthal fibers (RFs) 78-80, 91, 99, 100, 101, 140, 165, 210

S

256

Astrocytes in Brain Aging and Neurodegeneration

T

U

Tau 77, 78, 91, 92, 95, 99, 166-181, 210, 212, 213 Tau-positive inclusions 99, 168, 169, 173-175, 180 Taurine 22, 43, 44, 51 Thorn shaped astrocytes 170 Transferrin 171, 236-238, 240, 246 Transforming growth factor-β (TGF-β) 6, 48, 75, 92, 121, 150 Transmissible mink encephalopathy (TME) 127, 142 Transmissible spongiform encephalopathies (TSE) 127, 129, 135, 137, 140, 143, 145, 149-154 Trauma 8, 21, 31, 42, 44, 46, 196, 209, 222 Tufted astrocyte 170, 174-176, 180 Tumor necrosis factor α (TNF α) 29, 30, 45, 46, 92, 120, 149-154

Ubiquitin (Ub) 51, 77, 78, 95, 101, 170, 174, 180, 181, 207-210, 212, 214, 215, 217, 219, 221-223, 225, 227

V Velate astrocyte 8, 9, 131 Vimentin 3, 4, 42, 50, 78, 79, 147, 178

W Wilson’s disease 73, 74

X X-irradiation 196, 219, 224