PROGRESS IN BRAIN RESEARCH VOLUME 88 NEUROBIOLOGY OF THE LOCUS COERULEUS
Recent volumes in PROGRESS IN BRAIN RESEARCH...
9 downloads
503 Views
47MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
PROGRESS IN BRAIN RESEARCH VOLUME 88 NEUROBIOLOGY OF THE LOCUS COERULEUS
Recent volumes in PROGRESS IN BRAIN RESEARCH Volume 71 : Neural Regeneration, by F.J. Seil, E. Herbert and B.M. Carlson (Eds.)-1987 Volume 72: Neuropeptides and Brain Function, by E.R. de Kloet, V.M. Wiegant and D. de Wied (Eds.)-1987. Volume 73: Biochemical Basis of Functional Neuroteratology, by G.J. Boer, M.G.P. Feenstra, M. Mirmiran, D.F. Swaab and F. van Haaren (Eds.)-1988. Volume 74: Transduction and Cellular Mechanisms in Sensory Receptors, by W. Hamann and A. Iggo (Eds.)-1988. Volume 15: Vision within Extrageniculo-striate Systems, by T.B. Hicks and G. Benedek (Eds.)-1988. Volume 76: Vestibulospinal Control of Posture and Locomotion, by 0. Pompeiano and J.H.J. Allum (Eds.)-1988. Volume 77: Pain Modulation, by H.L. Fields and J.-M. Besson (Eds.)-1988. Volume 78: Transplantation into the Mammalian CNS, by D.M. Gash and J.R. Sladek, Jr. (Eds.)-1988. Volume 79: Nicotinic Receptors in the CNS, by A. Nordberg, K. Fuxe, B. Holmstedt and A. Sundwall (Eds.)-1989. Volume 80: Afferent Control of Posture and Locomotion, by J.H.J. Alum and M. Hulliger (Eds.)-1989. Volume 81: The Central Neural Organization of Cardiovascular Control, by J.Ciriello, M.M. Caverson and C. Polosa (Eds.)-1989. Volume 82: Neural Transplantation: From Molecular Basis to Clinical Applications, by S. Dunnett and S.-J. Richards (Eds.)-1990. Volume 83: Understanding the Brain through the Hippocampus, by J. Storm-Mathison, J. Zimmer and O.P. Ottersen (Eds.)-1990. Volume 84: Cholinergic Neurotransmission: Functional and Clinical Aspects, by S.-M. Aquilonius and P.-G. Gillberg (Eds.)-1990. Volume 85: The Prefrontal Cortex; Its Structure, Function and Pathology, by H.B.M. Uylings, C.G. van Eden, J.P.C. de Bruin, M.A. Comer and M.G.P. Feenstra (Eds.)-1991. Volume 86: Molecular and Cellular Mechanisms of Neuronal Plasticity in Normal Aging and Alzheimer’s Disease, by P.D. Coleman, G.A. Higgins and C.H. Phelps (Eds.)-1990. Volume 87: Role of the Forebrain in Sensation and Behavior, by G. Holstege (Ed.)-1991.
PROGRESS IN BRAIN RESEARCH VOLUME 88
NEUROBIOLOGY OF THE LOCUS COERULEUS EDITED BY
C.D. BARNES Department of Veterinaty and ComparativeAnatomy, Pharmacology and Physiology, Washington State Universiy, Pullman, Washington, U.S.A.
and
0. POMPEIANO Dipartimento di Fisiologia e Biochimica, Universitci di Pisa, Pisa, Italy
ELSEVIER AMSTERDAM - LONDON - OXFORD - TOKYO 1991
0
1991 Elsevier Science Publishers BV. All rights reserved.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the Publisher, Elsevier Science Publishers BV, Permissions Department, P.O. Box 521, lo00 AN Amsterdam, The Netherlands. No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of the rapid advances in the medical sciences, the Publisher recommends independent verification of diagnoses and drug dosages. Special regulations for readers in the USA. This publication has been registered with the Copyright Clearance Center, Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside the USA, should be referred to the Publisher.
ISBN: 0-444-81394-2 (volume) ISBN: 0-444-80104-9 (series)
Published by Elsevier Science Publishers BV P.O. Box 211 1000 AE Amsterdam The Netherlands
Sole distributors for the USA & Canada Elsevier Science Publishing Co. Inc. 655 Avenue of the Americas New York, NY 10010 USA
Library of Congress Cataloging in Publication Data Neurobiology of the locus coeruleus / edited by C.D. Barnes and 0. Pompeiano. cm.-- (Progress in brain research ; v. 88) p. Based on a symposium held on May 16-19, 1990 at Post Falls, Idaho; organized between the Dept. of Physiology and Biochemistry of the University of Pisa, Italy and the Dept. of Veterinary and Comparative Anatomy, Pharmacology, and Physiology at Washington State University, Pullman, Wash., USA. Includes bibliographical references and index. ISBN 0-444-81394-2 (alk. paper) 1. Locus coeruleus-Physiology-Congresses. 2. Noradrenergic neurons--Congresses. I. Barnes, Charles D., 193511. Pompeiano, 0. 111. Universith di Pisa. Dipartimento di Fisiologia e Biochimica. IV. Washington State University Dept. of Veterinary and Comparative Anatomy, Pharmacology, and Physiology. v. series. [DNLM: 1. Locus Coeruleus--physiology--congresses.2. Motor Neurons--physiology--congresses. 3. Neurons, Afferent--physiology-congresses. W1 PR667J v. 88 / WL 310 N4935 19901 QP376 N48.W vol. 88 [QP377.5] 612.8'2 s--dc20 [612.8'262] DNLM/DLC for Library of Congress
Printed in The Netherlands on acid-free paper
91-28208 CIP
List of Contributors L.M. Adams, Department of Psychiatry, University of California at San Diego Medical School, La Jolla, CA 92093, U.S.A. E.D. Abercrombie, Program in Neuroscience, Department of Psychology, Princeton University, Princeton, NJ 08544, U.S.A. H. Akaoka, Division of Behavioral Neurobiology, Department of Mental Health Sciences, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. T. Alexinsky, UniversitC RenC Descartes and LPN2 C N R S , 91198, Gif-sur-Yvette, France. P. Andre, Dipartimento di Fisologia e Biochimica, Universitl di Pisa, Via S. Zeno 31, 56127 Pisa, Italy. B. Astier, Laboratoire Neuropharmacologie, Facult6 de Pharmacie, UniversitC Claude Bernard, 8 Avenue Rockefeller, 69008 Lyon, France. G. Aston-Jones, Division of Behavioral Neurobiology,Department of Mental Health Sciences, Hahnemann University, MS 403, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. C.D. Barnes, Department of Veterinary and Comparative Anatomy, Pharmacolom and Physiology, Washington State University, Pullman, WA 99164-6520, U.S.A. C.W. Berridge, Department of Psychiatry, University of California at San Diego Medical School, La Jolla, CA 92093, U.S.A. D.H. Bobker, Vollum Institute, Oregon Health Sciences University, 3181 S.W. Sam Jackson Park Road, Portland, OR 97201, U.S.A. C.M.H. Bongers, Department of Anatomy, Erasmus University Medical School, P.O. Box 1738, 3000 DR, Rotterdam, The Netherlands. A.R. CaffC, Department of Anatomy, Erasmus University Medical School, P.O. Box 1738, 3000 DR, Rotterdam, The Netherlands. B. Cardo, Laboratoire de Psychophysiologie, URA CNRS 339, Avenue des Facultts, 33405 Talence, Cedex, France. S.M. Carlton, Department of Anatomy and Neurosciences, Marine Biomedical Institute, University of Texas Medical Branch, Galveston, TX 77660, U.S.A. J.Y.H. Chan, Department of Medical Research, Veterans General Hospital, Taipei 11217, Taiwan, China. V.L. Chan-Palay, Neurology Clinic, University Hospital, (33-8091 Zurich, Switzerland. P. CharlCty, INSERM U 171, Centre Hospitalier Lyon-Sud, Pav. 4H, Chemin de Grand Revoyet, 69310 Pierre-Benite, France. J.E.Cheun, Department of Neurobiology and Anatomy, University of Rochester Medical Center, Box 603, 601 Elmwood Avenue, Rochester, NY 14642, U.S.A. C. Chiang, Division of Behavioral Neurobiology, Department of Mental Health Sciences, Hahnemann University, MS 403, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. G. Chouvet, INSERM U 171, Centre Hospitalier Lyon-Sud, Pav. 4H, Chemin de Grand Revoyet, 69310 Pierre-Benite, France. M.J. Christie, Department of Pharmacology, University of Sydney, Sydney, N.S.W., 2006, Australia. C. Cirelli, Departimento di Fisiologia e Biochimica, Universitl di Pisa, Via S. Zeno 31, 56127 Pisa, Italy. F.M. Clark, Department of Pharmacology, University of Illinois at Chicago, P.O. Box 6998, Chicago, IL 60680, U.S.A. H. Collewijn, Department of Physiology, Erasmus University Medical School, P.O. Box 1738, DR 3000 Rotterdam, The Netherlands. A.L. Curtis, Department of Mental Health Science, Division of Behavioral Neurobiology, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. P. d'Ascanio, Dipartimento di Fisiologia e Biochimica, Universitl di Pisa, Via S. Zen0 31,56127 Pisa, Italy. G. Drolet, Division of Behavioral Neurobiology, Department of Mental Health Sciences, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. M. Ennis, Department of Physiology and Biophysics, University of Cincinnati College of Medicine, Cincinnati, OH 45267-0576, U.S.A. P.G. Finlayson, Department of Ophthalmology, University of British Columbia, Vancouver, Canada. S.L. Foote, Department of Psychiatly, University of California at San Diego Medical School, La Jolla, CA 92093, U.S.A. C.A. Fornal, Program in Neuroscience, Department of Psychology, Princeton University, Princeton, NJ 08544, U.S.A.J.M. Fritschy, Department of Neuroscience, The Johns Hopkins University School of Medicine, 72j N. Wolfe Street, Baltimore, MD 21205, U.S.A.
vi S.J. Fung, Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA 99164-6520, U.S.A. K.L. Grove, Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Wash.ington State University, Pullman, WA 99164-6520, U.S.A. R. Grzanna, Department of Neuroscience, The Johns Hopkins University School of Medicine, 725 N. Wolfe Street, Baltimore, MD 21205, U.S.A. P.G. Guyenet, Department of Pharmacology,University of Virginia School of Medicine, 1300 Jefferson-Park Avenue, Charlottesville, VA 22908, U.S.A. C.W. Harley, Department of Psychology, Memorial University of Newfoundland, St. Johns, Newfoundland, AIB 3x9, Canada. G.C. Harris, Vollum Institute, Oregon Health Sciences University, 3181 S.W. Sam Jackson Park Road, Portland, OR 97201, U.S.A. J.C. Holstege, Department of Anatomy, Erasmus University Medical School, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands. E. Horn, Department of Neurology, Institute of Neurophysiology,University of Ulm, 7900 Ulm, F.R.G. D.M. Jacobowitz, Laboratory of Clinical Science National Institutes of Mental Health, National Institutes of Health, Building 10, ROOM3D48, Bethesda, MD 20892, U.S.A. B.L. Jacobs, Program in Neuroscience, Department of Psychology, Princeton University, Princeton, NJ 08544, U.S.A. B.E. Jones, Department of Neurology and Neurosurgery, Montreal Neurological Institute, McGill University, 3801 University Street, Montreal, Quebec, H3A 2B4, Canada. C.R. Jones, Merrell Dow Research Institute, 16 Rue DAnkara, 67084 Strasbourg Cedex, France. S.L. Jones, Department of Pharmacology, 764 BMSB, College of Medicine, University of Oklahoma, 940 Stanton L. Young Blvd., Oklahoma City, OK 73190, U.S.A. T. Kasamatsu, Smith-Kettlewell Eye Research Institute, 2232 Webster Street, San Francisco, CA 94115, U.S.A. E. Kempf, Centre de Neurochimie CNRS, 5 Rue B. Pascal, 67084 Strasbourg Cedex, France. E.S. Levine, Program in Neuroscience, Department of Psychology, Princeton University, Princeton, NJ 08544, U.S.A. C.4. Lin, Department of Physiology and Biophysics, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. W. Liu, Department of Physiology and Biophysics, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. D. Manzoni, Dipartimento di Fisologia e Biochimica, Universitl di Pisa, Via S. Zeno 31, 56127 Pisa, Italy. H. Markram, Center for Neurosciences, Weizmann Institute, Rehovot 76100, Israel. K.C. Marshall, Department of Physiology, University of Ottawa, 451 Smyth Road, Ottawa, Ontario K1H 8M5, Canada. D.A. McCormick, Section of Neuroanatomy, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, U.S.A. H.C. Moises, Department of Physiology, University of Michigan Medical School, Ann Arbor, MI 48109, 1JSA. - ._._ -. D.A. Morilak, Program in Neuroscience, Department of Psychology, Princeton University, Princeton, NJ 08544, U.S.A. P. Mormbde, INRA-INSERM U 259 Rue C. St. Saens, 33077 Bordeaux, Cedex, France. S. Nakamura, Department of Physiology, Faculty of Medicine, Kanazawa University, Kanazawa 920, Japan. S. Nassif-Caudarella, Laboratoire de Psychophysiologie, URA CNRS 339, Avenue des Facultis, 33405 Talence, Cedex, France. H.-R. Olpe, Research and Development Department, Pharmaceuticals Division, CIBA-GEIGY Ltd., CH-4002 Basel, Switzerland. S.S OsmanoviC, Department of Patho-Physiology, Medical Faculty, University of Belgrade, 1000 Belgrade, Yugoslavia. J.M. Palacios, Sandoz Ltd., ch.-4002 Basle, Switzerland 61 24 88 39. H.-C. Pape, Lehrstul Neurophysiologie,Medizinische Fakultat, Ruhr-Universitat Bochum, D-4630 Bochum, F.R.G. V. Pieribone, Department of Histology and Neurobiology, Karolinksa Institutet, S-104 01 Stockholm, Sweden. J.A. Pineda, Department of Psychiatry, University of California at San Diego Medical School, La Jolla, CA 92093, U.S.A. M. Pompeiano, Departimento di Fisiologia e Biochimica, Universitl di Pisa, Via S. Zen0 31, 56127 Pisa, Italy. 0. Pompeiano, Departimento di Fisiologia e Biochimica, Universitl di Pisa, Via S. Zen0 31, 56127 Pisa, Italy. H.K. Proudfit, Department of Pharmacology, University of Illinois at Chicago, P.O. Box 6998, Chicago, IL 60680, U.S.A.
vii K Rasmussen, Lilly Research Laboratories, Eli Lilly and Co.,Lilly Corporate Center, Indianapolis, IN 46285, U.S.A. V.K. Reddy, Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA 99164-6520, U.S.A. G. Richter-Levin, Center for Neurosciences, Weizmann Institute, Rehovot 76100, Israel. B.P. Rowe, Department of Physiology, Quillen College of Medicine, East Tennessee State University, Johnson City, TN 37614-0002, U.S.A. K. Sakai, Laboratoire de Pathologie Expirimentale, Faculti de Medecine, UniversitC Claude Bernard, 8 Avenue Rockefeller, 69373 Lyon, Cedex 2, France. S.J. Sara, Departement de Psychophysiologie, Laboratoire de Physiologie Nerveuse, C.N.R.S., 91198 Gif-sur-Yvette, Cedex, France. M. Segal, Center for Neurosciences, Weizmann Institute, Rehovot 76100, Israel. F.M. Sessler, Department of Physiology and Biophysics, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. S.A. Shefner, Department of Physiology and Biophysics, University of Illinois College of Medicine, P.O. Box 6998, Chicago, IL 60680, U.S.A. R. Shiekhattar, Division of Behavioral Neurobiology, Department of Mental Health Sciences, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. M.T. Shipley, Department of Anatomy and Cell Biology, University of Cincinnati College of Medicine, Cincinnati, OH 45267-0576, U.S.A. L. Sklair, Center for Neuroscience, Weizmann Institute, Rehovot, 76100, Israel. L.S. Sorkin, Department of Anatomy and Neurosciences, Marine Biomedical Institute, University of Texas Medical Branch, Galveston, TX 77660, U.S.A. R.C. Speth, Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA 99164-6520, U.S.A. I.L. Stafford, Program in Neuroscience, Department of Psychology, Princeton University, Princeton, NJ 08544, U.S.A. M. Steinmann, Research and Development Department, Pharmaceuticals Division, CIBA-GEIGY Ltd., CH-4002 Basel, Switzerland. E.L. Sutin, Department of Psychiatry, Sleep Research Center, TD-114, Stanford University School of Medicine, Palo Alto, CA 94305, U.S.A. G. Tononi, Departimento di Fisiologia e Biochimica, Universid di Pisa, Via S. Zen0 31, 56127 Pisa, Italy. R.J. Valentino, Division of Behavioral Neurobiology, Department of Mental Health Science, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. E. van Bockstaele, Division of Behavioral Neurobiology, Department of Mental Health Sciences, Hahnemann University, Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. F.W. van Leeuwen, Department of Anatomy, Erasmus University Medical School, P.O. Box 1738,3000 DR Rotterdam, The Netherlands. J. van Neerven, Department of Physiology, Erasmus University Medical School, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands. L.A. Velley, Laboratoire de Psychophysiologie, URA CNRS 339, Avenue des FacultBs, 33405 Talence, Cedex, France. J. Velly, Institut de Pharmacologie, 11 Rue Human, 67000 Strasbourg, France. B.D. Waterhouse, Department of Physiology and Biophysics, Hahnemann University, MS 409,Broad and Vine, Philadelphia, PA 19102-1192, U.S.A. KN. Westlund, Department of Anatomy and Neurosciences, Marine Biomedical Institute, University of Texas, 200 University Boulevard, Galveston, TX 77550, U.S.A. W.D. Willis, Department of Anatomy and Neurosciences, Marine Biomedical Institute, University of Texas Medical Branch, Galveston, TX 77660, U.S.A. S.R. White, Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA 99164-6520, U.S.A. J.T. Williams, Vollum Institute, Oregon Health Sciences University, 3181 S.W. Sam Jackson Park Road, Portland, OR 97201, U.S.A. A. Williamson, Section of Neuroanatomy, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, U S A . D.J. Woodward, Department of Cell Biology, University of Texas Health Science Center, 5323 Harry Hines Boulevard, Dallas, TX 75235, U.S.A. H.H. Yeh, Department of Neurobiology and Anatomy, University of Rochester Medical Center, Box 603, 601 Elmwood Avenue, Rochester, NY 14642, U.S.A. H. Zhuo, Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA 99164-6520, U.S.A. D. Zhang, Department of Anatomy and Neurosciences, Marine Biomedical Institute, University of Texas Medical Branch, Galveston, TX 77660, U.S.A.
This Page Intentionally Left Blank
ix
Preface Noradrenergic neurons were first demonstrated in the central nervous system in 1965. The neurons were clustered in nuclear groups throughout the brainstem and were identified as A1 through A7, with A6, which correlates to the locus coeruleus (LC), being the largest. The forward projecting noradrenergic fiber system of the LC ending at cortical and subcortical levels (thalamus, hypothalamus, hippocampus) was the focus of most early experiments. As a result of these studies, the LC was associated with higher functions such as attention, orientation, anxiety, stress, learning and memory, brain plasticity, and sleep-waking activities. Moreover, the fact that noradrenergic LC cells ceased firing selectively during desynchronized sleep (DS) led to the hypothesis that the arrest of firing of LC neurons plays a causal role in DS generation, releasing from inhibition cholinergic-cholinoceptive pontine tegmental neurons, which would be directly executive for the EEG desynchronization and the postural atonia occurring during this phase of sleep. Concurrent with these studies, experiments were performed to investigate the LC influences on the cerebellum, while focus was also directed on the effects of brainstem-spinal cord aminergic systems on both dorsal and ventral horn elements. As early as 1967, experiments describing spinal cord effects from brainstem adrenergic systems were first published by Barnes and colleagues. Working together in 1971, Barnes and Pompeiano published their work describing a brainstem adrenergic system activated by vestibular stimulation and acting on the spinal cord. Experiments of unit recording were made in 1975 by Pompeiano and colleagues demonstrating that the postural activity which occurs in the decerebrate cat, i.e., the y-rigidity described by Granit, was positively correlated to the discharge of LC neurons, and that the activity of these neurons was also affected by static changes in animal position, leading to stimulation of gravity receptors. The link between the vestibular system and the LC was also tested in successive experiments showing that presumably noradrenergic LC neurons not only responded to dynamic stimulation of labyrinth (and neck) receptors but also intervened directly through the coeruleospinal pathway, or indirectly through the cholinergic-cholinoceptive pontine tegmental neurons, to modify the gain of the vestibulospinal reflexes. Finally, evidence was given that the central noradrenergic system acting through appropriate areas of the cerebellar cortex intervened in the gain regulation of the vestibulospinal reflexes as well as in the adaptive process which affects the gain of the vestibulo-ocular reflex evoked in darkness and in light. These findings raise questions as to the possible role(s) that the noradrenergic LC complex and related structures exert in the plastic changes which occur in the central nervous system during and after a long-term exposure to microgravity. It was not until the late 1970s, however, that these effects were attributed to the coeruleospinal system. Since that time a concentrated effort has been made by Barnes and colleagues to determine the specific effects of coeruleospinal activity on spinal cord elements.
X
At the same time that we were studying these effects, other laboratories were pursuing valuable research work on the neurochemical anatomy of the LC system, on the membrane properties of their neurons, on the related synaptic mechanisms involving different neurotransmitters and neuromodulators, on the effects of noradrenaline on different target systems as well as on the distribution and modality of action of different types of adrenoceptors in various areas of the CNS, on the LC control of low and high brain functions, on the role of this structure during cortical development, regeneration, plasticity and, finally, on the degenerative processes which affect the LC in several neurological and psychiatric disorders. Information was accumulating so rapidly, with some major discrepancies surfacing, that in 1989 we felt that a world conference with leading researchers would be appropriate to help develop a focus. As a result, a three-day symposium, “The Neurobiology of the Locus Coeruleus,” was held at Post Falls, Idaho, U.S.A., on May 16-19, 1990. Participants presented the results of their own research and shared ideas and goals. These presentations are included in this volume; it is hoped that they will offer ideas and challenges to those conducting research on the LC. This Symposium was organized in scientific cooperation between the Department of Physiology and Biochemistry of the University of Pisa, Italy and the Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology at Washington State University, Pullman, Washington, U.S.A. It was endorsed by the International Brain Research Organization (IBRO) and was financially supported by the National Institute of Neurological Disorders and Stroke, National Institute of Mental Health, the National Science Foundation, U.S.A. and by the Minister0 dell’universiti e della Ricerca Scientifica e Tecnologica, Roma, Italy. These contributions are gratefully acknowledged. 15 November 1990
C.D. Barnes, Pullman, Washington,’U S A .
0. Pompeiano, Pisa, Italy
xi
Contents .............................................
V
.......................................................
ix
List of Contributors Preface
Section I. Anatomy of the Locus Coedeus: Its Afferents and Efferents 1. Neurochemicals in the dorsal pontine tegmentum E.L. Sutin and D.M. Jacobowitz (Bethesda, MD, USA)
.................
3
2. Noradrenergic locus coeruleus neurons:: $heir distant connections and their relationship to neighboring (including cholinergic and GABAergic) neurons of the central gray and reticular formation B.E. Jones (Montreal, Quebec, Canada) ............................
15
3. Physiological properties and afferent connections of the locus coeruleus and adjacent tegmental neurons involved in the generation of paradoxical sleep in the cat K. Sakai(Lyon, France) .......................................
31
4. Afferent regulation of locus coeruleus neurons: anatomy, physiology and
pharmacology G. Aston-Jones, M.T. Shipley, G. Chouvet, M. Ennis, E. van Bockstaele, V. Pieribone, R. Shiekhattar, H. Akaoka, G. Drolet, B. Astier, P. Charldty, R.J. Valentino and J.T. Williams (Philadelphia, PA, Cincinnati, OH and Portland, OR, USA, Pierre-Benite and Lyon, France and Stockholm, Sweden) .
47
5. Noradrenergic innervation of somatosensory thalamus and spinal cord
K.N. Westlund, D. Zhang, S.M. Carlton, L.S. Sorkin and W.D. Willis (Galveston, TX,USA). ........................................
77
6. Efferent projections of different subpopulations of central noradrenaline
neurons R. Grzanna and J.-M. Fritschy (Baltimore, MD, USA)
..................
89
......
103
7. Pontospinal transmitters and their distribution
V.K. Reddy, S.J. Fung, H. Zhuo and C.D. Barnes (Pullman, WA, USA)
8. The projections of locus coeruleus neurons to the spinal cord H.K. Proudfit and F.M. Clark (Chicago, IL, USA) .....................
123
9. Ultrastructural aspects of the coeruleo-spinal projection J.C. Holstege and C.M.H. Bongers (Rotterdam, The Netherlands)
143
..........
xii
Section 11. Properties of Locus Coeruleus Neurons 10. Single-unit and physiological analyses of brain norepinephrine function in behaving animals B.L. Jacobs, E.D. Abercrombie, C.A. Fornal, E.S. Levine, D.A. Morilak and 159 I.L. Stafford (Princeton, NJ, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Synaptic potentials in locus coeruleus neurons in brain slices J.T. Williams, D.H. Bobker and G.C. Harris (Portland, OR, USA) . . . . . . . .
167
12. Developmental aspects of the locus coeruleus-noradrenalinesystem K.C. Marshall, M.J. Christie, P.G. Finlayson and J.T. Williams (Ottawa, Ontario, Canada, Vancouver, BC, Canada, Sydney, NSW, Australia and Portland, OR, 173 USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13. GABA, and GABA, receptors and the ionic mechanisms mediating their effects on locus coeruleus neurons S.A. Shefner and S.S. OsmanoviC (Chicago, IL, USA and Belgrade, 187 Yugoslavia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14. Mechanisms of opioid actions on neurons of the locus coeruleus M.J. Christie (Sydney, NSW, Australia) ............................
197
15. Afferent effects on locus coeruleus in opiate withdrawal K. Rasmussen (Indianapolis, IN, USA) .............................
207
16. Angiotensin I1 and the locus coeruleus R.C. Speth, K.L. Grove and B.P. Rowe (Pullman, WA and Johnson City, TN, 217 USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17. Vasopressin immunoreactive fibers and neurons in the dorsal pontine tegmentum of the rat, monkey and human A.R. Caffk, J.C. Holstege and F.W. van Leeuwen (Rotterdam and Amsterdam, 227 The Netherlands) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18. Responses of locus coeruleus neurons to neuropeptides H.-R. Olpe and M. Steinmann (Basel, Switzerland) ....................
241
19. Pharmacology of locus coeruleus spontaneous and sensory-evoked activity R.J. Valentino and A.L. Curtis (Philadelphia, PA, USA) . . . . . . . . . . . . . . . . .
249
20. Selective effects of DSP-4 on locus coeruleus axons: are there pharmacologically different types of noradrenergic axons in the central nervous system? J.-M. Fritschy and R. Grzanna (Baltimore, MD, USA) . . . . . . . . . . . . . . . . . . 257
Section 111. Noradrenergic Influences on Target Neurons 21. Autoradiography of adrenaceptors in rat and human brain: a-adrenoceptor and idazoxan binding sites C.R. Jones and J.M. Palacios (Strasbourg, France and Basel, 271 Switzerland) ................................................
...
XI11
22. Actions of norepinephrine in the cerebral cortex and thalamus: implications for function of the central noradrenergic system D.A. McCormick, H.-C. Pape and A. Williamson (New Haven, CT, USA) . . . . 293 23. Noradrenergic and locus coeruleus modulation of the perforant path-evoked potential in rat dentate gyrus supports a role for the locus coeruleus in attentional and memorial processes 307 C. Harley (St. John's, Newfoundland, Canada) ....................... 24. Actions of norepinephrine in the rat hippocampus M. Segal, H. Markram and G. Richter-Levin (Rehovot, Israel)
............
323
25. The cerebellar norepinephrine system: inhibition, modulation, and gating D.J. Woodward, H.C. Moises, B.D. Waterhouse, H.H. Yeh and J.E. Cheun (Dallas, TX, Ann Arbor, MI, Philadelpha, PA and Rochester, NY, USA) . . . . . 331 26. Norepinephrine effects on spinal motoneurons S.R. White, S.J. Fung and C.D. Barnes (Pullman, WA, USA)
. . . . . . . . . . . . . 343
27. Second messenger-mediated actions of norepinephrine on target neurons in central circuits: a new perspective on intracellular mechanisms and functional consequences B.D. Waterhouse, F.M. Sessler, W. Liu and C.-S. Lin (Philadelphia, PA, USA) . 351
Section IV. Control of Motor and Sensory Systems 28. Central noradrenergic neurons: the autonomic connection P.G. Guyenet (Charlottesville, VA, USA) ...........................
365
29. Descending noradrenergic influences on pain S.L. Jones (Oklahoma City, OK, USA) .............................
381
30. Locus coeruleus control of spinal motor output S.J. Fung, D. Manzoni, J.Y.H. Chan, 0.Pompeiano and C.D. Barnes (Pullman, 395 WA, USA, Pisa, Italy and Taipei, Taiwan, China) ..................... 31. Responses of locus coeruleus neurons to labyrinth and neck stimulation 0. Pompeiano, D. Manzoni and C.D. Barnes (Pisa, Italy and Pullman, WA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 32. Locus coeruleus and dorsal pontine reticular influences on the gain of vestibulospinal reflexes 0. Pompeiano, E. Horn and P. d'Ascanio (Pisa, Italy) . . . . . . . . . . . . . . . . . . . 435 33. Noradrenergic agents into the cerebellar anterior vermis modify the gain of vestibulospinal reflexes in the cat P. Andre, P. d'Ascanio and 0. Pompeiano (Pisa, Italy) . . . . . . . . . . . . . . . . . . 463
XiV
34. Effects of GABAergic and noradrenergic injections into the cerebellar flocculus on vestibulo-ocular reflexes in the rabbit J. van Neerven, 0.Pompeiano and H.Collewijn (Rotterdam, The Netherlands 485 andPisa,Italy) ..............................................
Section V. Locus Coeruleus Influences on Higher Functions and Plasticity
35. Discharge of noradrenergic locus coeruleus neurons in behaving rats and monkeys suggests a role in vigilance G. Aston-Jones, C. Chiang and T. Alexinsky (Philadelphia, PA, USA and 501 Gif-sur-Yvette, France) ........................................
36. Electrophysiological evidence for the involvement of the locus coeruleus in alerting, orienting, and attending S.L. Foote, C.W. Berridge, L.M. Adams and J.A. Pineda (La Jolla, CA, USA) . . 521
37. The role of noradrenergic locus coeruleus neurons and neighboring cholinergic neurons of the pontomesencephalic tegmentum in sleep-wake states B.E. Jones (Montreal, Quebec, Canada) ............................
533
38. Effects of local pontine injection of noradrenergic agents on desynchronized sleep of the cat G. Tononi, M. Pompeiano and C. Cirelli (Pisa, Italy) . . . . . . . . . . . . . . . . . . . 545
39. Facilitation of learning consecutive to electrical stimulation of the locus coeruleus: cognitive alteration or stress-reduction? L. Velley, B. Cardo, E. Kempf, P. Mormede, S. Nassif-Caudarella and J. Velly (Talence, Strasbourg and Bordeaux, France) ......................... 555
40. Plasticity of sensory responses of locus coeruleus neurons in the behaving rat: implications for cognition S.J. Sara and M. Segal (Gif sur Yvette, France and Rehovot, Israel) . . . . . . . . . 571
41. Axonal sprouting of noradrenergic locus coeruleus neurons following repeated stress and antidepressant treatment S. Nakamura (Kanazawa, Japan) .................................
587
42. Adrenergic regulation of visuocortical plasticity: a role of the locus coeruleus system T. Kasamatsu (San Francisco, CA, USA)
...........................
599
43. Regulation of the development of locus coeruleus neurons in uitro L. Sklair and M. Segal (Rehovot, Israel) ............................
617
44. Alterations in .the locus coeruleus in dementias of Alzheimer’s and Parkinson’s disease V. Chan-Palay (Zurich, Switzerland) ...............................
625
SubjectIndex ..................................................
631
SECTION I
Anatomy of the Locus Coeruleus: Its Afferents and Efferents
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Bruin Research, Vol. 88 8 1991 Hsevier Science Publishers B.V.
3
CHAPTER 1
Neurochemicals in the dorsal pontine tegmentum E.L. Sutin * and D.M. Jacobowitz Laboratory of Clinical Science, Natwnul Institute of Mental Health, Bethesda, MD, U.S.A.
Detailed maps of neurochemicals in the 'locus coeruleus and adjacent dorsal tegmental areas are discussed in this chapter. The locus coeruleus appears to be one of the most complex brain regions with six neurochemicals (acetylcholinesterase,
tyrosine hydroxylase, galanin, neuropeptide Y, neurotensin, and vasoactive intestinal protein) contained within the cell bodies.
Key words: locus coeruleus, dorsal pontine tegmentum, immunofluorescence microscopy, lateral dorsal tegmental nucleus, tyrosine hydroxylase
Introduction The rat locus coeruleus (LC) is a distinct cluster of neurons located in the dorsal pontine brain stem at the ventrolateral edge of the fourth ventricle. The dorsal tegmental area of the pons contains several neurotransmitter-specific nuclei located adjacent to each other, including the noradrenergic LC, the cholinergic laterodorsal tegmental nucleus (NTDL), and the serotonergic dorsal raphe nuclei. Therefore, these various components of the tegmentum, in addition to the dorsal tegmental nucleus of Gudden (NTD) and the parabrachial nuclei, are of interest in terms of significant circuitry, neurochemical localization, and potential physiological and clinical significance. * Present address: Stanford University School of Medicine, Department of PsychiatIy, Sleep Research Center, TD-114, Palo Alto, CA 94305, U.S.A.
Several peptides and neurotransmitter-related enzymes have been localized to the rat LC. Previous studies have revealed dopamine-p-hydroxylase (Hartman, 1973; Grzanna and Molliver, 1980a,b); tyrosine hydroxylase 0 W )(Palkovits and Jacobowitz, 1974; Pickel et al., 1975a,b); acetylcholinesterase (AChE) (Knight, 1970; Lewis and Schon, 1975; Albanese and Butcher, 1980); galanin (GAL) (Skofitsch and Jacobowitz, 1985c; Melander et al., 1986), neuropeptide Y (NPY) (Olschowka et aZ., 19811, neurotensin (NT) (Uhl et al., 1979); and vasopressin (Sofroniew, 1979)containing cell bodies within this nucleus. In addition to these neurochemicals, the LC is richly innervated with fibers containing phenylethanolamine N-methyltransferase (Milner et al., 1989) and GABA (BCrod et al., 1984). Furthermore, tracer studies have demonstrated that LC cells containing NPY, TH, and GAL project to the cerebral cortex, hypothalamus and spinal cord (Holets et al., 1988). In this chapter we present
4
detailed maps of neurochemicals in the LC and adjacent dorsal tegmental areas. Methods Male rats were processed for indirect immunofluorescence microscopy with or without colchicine (100 pg/20 p1) and were perfused with 10% formalin, postfixed 30 min and rinsed in PBS containing 20% (w/v) sucrose at 4°C for 2 days. The tissue was cut into coronal 20 pm sections in a cryostat, and consecutive sections were taken through the dorsal tegmental area at the level of the LC. Every fourth section was used for each staining method, resulting in a series of sections 80 pm apart. See Sutin and Jacobowitz (1988) for a detailed description of the immunocytochemical
4~
and histochemical procedures in addition to the antisera used. Projection drawings made from a series of consecutive 240 pm AChE sections through the dorsal tegmental region were used as maps to plot the distribution of reactive perikarya and varicose fibers for each antiserum in all rats. The results for each antiserum were collated and drawn on four representative levels through the dorsal tegmental area at the level of the LC, NTDL, "ID, parabrachial nucleus, and dorsal raphe nucleus. Results and Discussion A mapping of each neurochemical is presented in Figures 1-16. Cell bodies are shown as large
'i
B tz140
c+ 480
D+ 720
Fig. 1. Mapping of AChE-containing structures in four representational levels (A-D) through the dorsal tegmental region. Cell bodies (large filled circles) and varicosities (fine dots) are shown on the left side in schematic form and a corresponding photograph is on the right side of the figure. Abbreviations: dr, dorsal raphe nucleus; inc, nucleus incertus; kf, Kolliker-Fuse nucleus; LC,locus coeruleus; NPD, dorsal parabrachial nucleus; npv, ventral parabrachial nucleus; NTD, dorsal tegmental nucleus (Gudden); NTDL, laterodorsal tegmental nucleus; NTM, nucleus of the mesencephalic tract of the trigeminal nerve; PCS, superior cerebellar peduncle; SC, subcoeruleus; TM,mesencephalic tract of the trigeminal nerve; IV, 4th cranial nerve; nV, motor nucleus of the trigeminal nerve.
5
filled circles on the right side, and fibers and terminals are shown as fine dots on the left side of the figures. The relative densities of the fibers and cell bodies are indicated by the separation of the symbols, a convention for which allowance must be made for the illustration of small anatomical structures.
AChE Figure 1 shows the mapping of AChE-containing structures as a diagram on the left side and a corresponding photograph on the right. Beginning at the most caudal level (Fig. lA), darkly staining cell bodies were evenly distributed within the LC and a few cells were ventrally dispersed, which comprised the subcoeruleus (SC) (Fig. 1AC). AChE-containing cell bodies also existed in the motor nucleus of the fifth nerve (nV) (Fig. 1A-D) and smaller, paler staining cells were seen in the dorsal raphe nucleus (Fig. 1A-D). AChE fibers visualized as brown precipitate, surrounded the superior cerebellar peduncle (PCS), the NTD, and in the nV. Medial to the LC was a zone that appeared very light in AChE-stained sections (Fig. 1A-D). According to Wilson (19851, this distinct, oval, unstained nucleus, which is located just rostromedial to the LC and enveloped rostrally by the NTDL, is Barrington’s nucleus, the micturition reflex center, because it corresponds with a subnucleus of the NTDL labeled by Satoh et al. (1978) and Loewy et al. (1979). This light-staining region is referred to throughout this work as Barrington’s nucleus, and the area medial to this zone, but excluding the nucleus incertus and NTD, as central gray. Beginning in level C, a few large, darkly stained cell bodies were seen in the central gray just medial to Barrington’s nucleus. Because we have chosen to define the NTDL in accordance with its cholinergic cell groups (AChE and ChAT), this level constitutes the caudal aspect of the NTDL. Ventral to the LC and directly dorsal to the SC was a triangular area that contained terminals and a few darkly stained cell bodies (Fig. lC,D). This small wedge extended only through
levels C and D and then seemed to be encompassed by the NTDL. The localization of AChE in the NTDL is essentially identical to that of the ChAT-containing cells. This is consistent with the current understanding that, with a few exceptions, AChE is contained in all peripheral and CNS cholinergic neurons (Koelle, 1955). An outstanding exception to this generalization is the LC and SC, which have previously been reported to contain AChE (Palkovits and Jacobowitz, 1974). The presence of AChE in these cells may be a reflection of cholinoceptive activity. This is supported by the presence of acetylcholine and ChAT activity (Helke et al., 1980) in the LC. Whether the cholinergic innervation to the LC emanates from the cells of the NTDL is currently a matter of speculation.
ChAT Most of the ChAT-containing cell bodies correlated with the AChE-stained cells discussed above. Unfortunately, fibers and terminal fields did not fluoresce and were not mapped. Therefore only cell bodies were mapped and represented in Figure 2. Neither the LC nor SC was labeled with the ChAT antibody, although a few cell bodies were seen in proximity to the LC medially and laterally, and small cells were additionally seen dispersed through the PCS (Fig. 2A,B). The giant motor cells of the nucleus of the nV, which stain darkly for AChE, were also highly immunoreactive with the ChAT antisera (Fig. 2A-C). A few ChAT-immunoreactive cells were seen in the NTDL beginning at level C. In the triangular-shaped AChE-stained zone lateral to the NTDL, a few more ChAT-containing cells were found (Fig. 2C,D). These cholinergic cells confirm the above observation that this area is part of the NTDL and that at this level Barrington’s nucleus separates them. Beginning in level D and in more rostra1 levels, the NTDL extended both dorsomedially and ventrolaterally and projected into the central tegmental tract toward the PCS.
6
A
D
C
D
Fig. 2. The density and distribution of ChAT-immunoreactive cell bodies are mapped on one side through levels A-D
TH Staining for TH provided an impressive display of intensely fluorescent cell bodies and varicose fibers at all levels through the dorsal tegmental region (Fig. 3). Especially prominent were the LC perikarya, all of which were immunoreactive. The
Figs. 3-16. .The density and distribution of neurochemical-immunoreactive cell bodies (filled circles on the right half of the panels) and fibers (fine dots of the left side of the panels).
SC cells lying directly ventral to the LC were less densely stained and fewer in number. These SC cells continued ventrally in a narrow region just medial to the motor nucleus of the nV (Fig. 3A-C). A few scattered neurons could be found in the dorsal parabrachial nucleus (Fig. 3D). Beginning at level D a more compact group of neurons were observed ventral to the base of the PCS (Fig. 3D). TH-immunoreactive fibers could be seen radiating from the LC in all directions. The greatest amount of fluorescence was situated around the periphery of the LC. An abundant number of fibers was observed in the SC region (Fig. 3B,C) and remained within the central tegmental tract at more rostra1 levels (Fig. 3D). This density of fibers constitutes the origin of the dorsal noradrenergic bundle (Jacobowitz, 1978).
GAL At the most caudal level, intensely fluorescent cell bodies were densely packed in the LC (Fig. 4A). A few equally bright cells resided just outside the boundaries of the LC ventrally and medially into the central gray (Fig. 4A,B). The dorsal raphe nucleus contained fewer GAL-immunoreactive neurons that appeared more dispersed (Fig. 4A-H). Beginning at level c, GALlike cell bodies were seen in the vicinity of the NTDL. The predominance of the cells did not fluoresce as intensely as those of the LC. Another
Fig. 4. See text for details.
group of GALcontaining cells was observed at level D just lateral to the dorsorostral part of the PCS in the npd (Fig. 4D). Brightly reactive GAL fibers were observed in the four levels. Not only has GAL been isolated and sequenced (Rokaeus and Brownstein, 1986; Vrontakis et al., 1987; Kaplan et al., 1988), but it has been found to be biologically active (Tatemoto et al., 1983) and widely distributed in the rat central nervous system. GAL immunoreactivity has been mapped immunocytochernically in the rat central nervous system by Skofitsch and Jacobowitz (1985~)and Melander et al. (1986). One of the most remarkable features about GAL is its coexistence in all LC cell bodies (Skofitsch and Jacobowitz, 198%; Melander et al., 1986). This suggests a possible modulatory role of GAL on noradrenergic and cholinergic neuronal projections to forebrain telencephalic areas. GAL immunoreactivity was also observed in B6 raphe cell bodies. A few such cells were previously mapped (Skofitsch and Jacobowitz, 1985c), whereas Melander et al. (1986) observed large numbers of cells in the raphe dorsalis (B7). Whether GAL also coexists in the cholinergic and serotonergic cells of the NTDL needs to be studied. Neurotensin A moderate number of NT-containing cell bodies were found in the LC (Fig. 5A,B) and the NPD (Fig. 5C-D). The highest concentration of
A
Fig. 5. See text for details.
A
B
C
D
Fig. 6. See text for details.
cell bodies was seen in the ventrolateral division of the NPD (Fig. 5C,D). NT-containing fibers were found in low density in the NTDL and LC and moderate densities in the central gray and ventral parabrachial nucleus (Fig. 5A-D). A high density of fibers was observed medial to the TM and NTM, surrounding the PCS, dorsal parabrachial, and Kolliker-Fuse nuclei (Fig. 5BD). No varicosoties were seen in the nucleus incertus or NTD. The presence of NT cells in the LC is interesting in light of the observatibn that this peptide inhibits firing of half of the cells studied in this nucleus (Young et al., 1978). NPY NPY-immunoreactive cell bodies were localized only to a few areas in low density, which included the LC (Fig. 6A-D) and ventral central gray (Fig. 6A,B). The LC was filled with varicosities (Fig. 6A-D), in addition to the NPD, NTDL and Barrington’s nucleus. NPY displays amino acid sequence homologies with the members of the pancreatic polypeptide family (Lin, 1980; Tatemoto et al., 1983; Kimmel et al., 1984). It is evident from immunocytochemical maps of pancreatic polypeptide (Olschowka et al., 1981) and NPY localization (Chronwall et al., 1985) that the antisera of both peptides are localized in the same neurons. Pancreatic polypeptide or NPY
8
coexists with norepinephrine in some of the LC perikarya (Everitt et al., 1984).
VlP VIP-immunoreactivecell bodies were observed scattered along the floor of the central gray and LC (Fig. 7A-C). The cells in the LC were observed mainly in the dorsomedial quadrant (Fig. 7A,B). In level C, cell bodies were also seen in the dorsal raphe, and a few very small cells were stained along the ventral border of the NTD. Larger neurons were located ventral to the LC in the triangular AChE-stained region (Fig. 7C,D). VIP-containing cells were seen slightly medially and laterally in rostral sections and within the ventrolateral contiguous NTDL. A large, well-defined group of VIP-fluorescent cells was found in the ventrolateral division of the NPD (Fig. 7A). VIP-immunoreactive fibers were seen in a crescent of varicosities beginning in level D, which stretches from the dorsal tip of the PCS in a ventral direction to the dorsolateral edge of the ventral spinal cerebellar tract. When compared with adjacent AChE-stained sections (see Fig. lD), it is evident that this arc of varicosities does not enter the AChE-stained area in the dorsolatera1 corner of the NPD. Instead the VIP-containing fibers enveloped the AChE-stained area surrounding the ventral aspect. The band of fluores-
A
Fig. 7. See text for details.
Fig. 8. See text for details.
cence was most dense at level D, then dispersed gradually in more rostral sections. VIP, unlike most peptides, was found in the NTD slightly more dense relative to surrounding areas (Fig. 7D). The nucleus incertus was also stained. Localization of VIP within LC cells has not been definitively proven, although high levels of VIP have been demonstrated in the LC by radioimmunoassay (Eiden et al., 1982). In addition, the LC has a high density of VIP binding sites (Martin et al., 1987).
CCK No cell bodies were found in the LC, NTDL, or NTD. The maps of CCK-containing neurons showed a dense accumulation in the caudal dorsal raphe nucleus (Fig. 8A-C). A few cell bodies were also scattered throughout the central gray (Fig. SA). A discrete group of intensely fluorescent cell bodies was observed in the NPD. CCK fibers, in general, followed the standard peptide pattern described above. Like some other peptides described in this report, the varicose fibers enveloped the central gray, leaving a blank area where the NTD and nucleus incertus lie (Fig. 8A-D). The close anatomical proximity of CCK cells with raphe cells suggests a potential modulatory interaction.
9
A
B
C
D
.
..
Fig. 9. See text for details.
CRF Most caudally, a few scattered CRF-immunoreactive cell bodies were observed in the central gray (Fig. 9A). A more discrete and compact group of cells was localized just ventromedial to the LC in Barrington’s nucleus (Fig. 9B,C), which extended medially into the NTDL and ventrally through the central tegmental tract (Fig. 9D). Other CRF-positive cell bodies were found in the .NPD in groups that extended in a ventrolateral and rostral direction (Fig. 9C-D). Throughout all levels a few fine, individual CRF varicose fibers were observed in the NPD. It is interesting that CRF has been found to coexist with both SP and AChE in neurons of the NTDL (Crawley et aZ., 1985).
CGRP A few small, closely grouped cell bodies were seen in the medial aspect of the caudal NTD, and scattered somata were observed around the ventral border of the central gray and npv (Fig. 10A,B). Several CGRP-like neurons were seen within and just dorsal to the motor nucleus of the Vth nerve (nV) (Fig. 1OA-C). The greatest density of cell bodies was found in the ventrolateral portion of the NPD, just surrounding the base of the PCS (Fig. 10B-D). Just dorsal to this group, at mid levels of the NPD (Fig. 10C-D), cell
bodies were observed along the dorsal border of the PCS, and then a discrete group formed, which extended laterally at more rostral levels (Fig. 10D). Beginning at level D, large cells were seen in the NTDL. CGRP-containing varicose fibers followed the typical peptide fiber pattern. The NTDL and the central gray area contain a dense localization of CGRP receptors as shown by autoradiography (Skofitsch and Jacobowitz, 1985a). The major CGRP cluster of somata are located in the NPD and Kolliker-Fuse region, which is also known to be the “pontine taste area” (Norgren and Leonard, 1970). Fibers from these areas project to the insular cortex (gustatory cortex). This cortical region is innervated by CGRP fibers and contains abundant receptors (Skofitsch and Jacobowitz, 1985a,b; Jacobowitz and Skofitsch, 1986).
SP SP-containing cell bodies were found in a narrow strip along the dorsomedial aspect of the central gray (Fig. 11A) and within the dorsal raphe nucleus (Fig. 11A-D). In the midsection of the central gray, a large cluster of SP-immunoreactive cells were found unevenly dispersed throughout Barrington’s nucleus and toward the NTD (Fig. 11A,B). A smaller cluster was observed near the dorsal tip of the PCS (Fig.
A
B
Fig. 10. See text for details.
10
A
Fig. 11. See text for details.
11A). Cell bodies were also found in the nucleus incertus (Fig. 11B,C). A moderate number of cells was seen in the most caudal levels of the NTDL (Fig. 11C,D), which began to form a welldefined group with identical shape and location to the ChAT-containing cells (Fig. 2C,D). SP processes followed the typical fiber pattern with the notable exception of dense varicosities in the caudal nucleus incertus (Fig. 110. SP has been shown to colocalize with ChAT in the NTDL (Vincent et al., 1983; Crawley et al., 1985). Although several of the NTDL cells contained ChAT and not SP, all cells that contained SP also contained ChAT. This finding was also documented by Crawley et al. (1985) using SP and AChE. These authors also found CRF to be colocalized with SP and AChE in some NTDL neurons. The presence of SP in raphe cells is significant in that it has been found to coexist with serotonin (Chan-Palay et al., 1978; Hokfelt et al., 1978) and as a triple coexistence with serotonin and thyrotropin-releasing hormone, which was demonstrated in certain neurons to project to the spinal cord (Johansson et al., 1981).
SST Several SSTimmunoreactive cell bodies were observed in the central gray region just medial and ventral to the LC (Fig. 12A,B). A few cells were also seen along the border of the lVth
ventricle and in the nucleus incertus (Fig. 12A). Cell bodies were found clustered around the NTM and more rostrally around the nucleus of the trochlear nerve (Fig. 12D). SST-immunoreactive varicose fibers were most dense within and medial to the NTM (Fig. 12B-D) and also ventral to the most caudal extent of the PCS (Fig. 12A). A moderate density of varicosities was also noted medial to the LC and in the parabrachial nucleus (Fig. 12C-D). Sparse numbers of varicosities were seen in the dorsal raphe nucleus, nucleus incertus, and NTDL (Fig. 12A-D).
GAD GAD proved to be a virtually ubiquitous enzyme that was present in high density through the extent of every section (Fig. 13). Because of the high intensity of fluorescence, discerning cell bodies was difficult. Yet, fairly consistent with all the maps were cell bodies in the caudal dorsal raphe nucleus and ventrally along the central gray (Fig. 13A). More GAD immunoreactivity was observed in the NTD than in surrounding areas. Caudally, the most concentrated portion was observed on the dorsolateral aspect of the nucleus (Fig. 13C). More rostrally the dense concentration of varicosities was seen on the ventral part of the nucleus. (Fig. 1 3 0 . The NTD remained the most intensely staining nucleus relative to the surrounding tegmental area. Other areas of high SST A
B
Fig. 12. See text for details.
11
A
. .. ...’ Fig. 13. See text for details.
Fig. 15. See text for details.
density were along the floor of the IVth ventricle (Fig. 13B-D), in Barrington’s nucleus (Fig. 13CD), and around the PCS (Fig. 13C-D). These results provide an anatomical basis for the finding that iontophoretically applied GABA decreases the firing rate of cells in the LC (Cedarbaum and Aghajanian, 1979; Guyenet and Aghajanian, 1979).
certus, NTM, and PCS. A compact cluster of cells was observed to descend rostroventrally in the NPD (Fig. 14B-D). Dyn B-containing varicose processes followed precisely the typical fiber pattern with the exception that the dorsolateral portion of the nucleus incertus possessed a dense accumulation of varicosities (Fig. 14B).
DYn B Very few Dyn B-labeled neurons were observed in the vicinity of the LC (Fig. 14). In caudal levels (Fig. 14A,B) scattered cell bodies were found in Barrington’s nucleus, nucleus in-
A
B
C
D
Fig. 14. See text for details.
5HT Besides a few scattered cell bodies in the ventrolateral portion of the central gray (Fig. 15A,B), 5HT-positive neurons were mainly confined to the midline dorsal raphe nucleus (Fig. 15). Individual large varicose fibers were seen projecting through the NPD and throughout the central gray at all levels. The greatest density of varicosities was located in the LC, motor nucleus of the Vth nerve, and the NTM (Fig. 15A-D). Atrial natriuretic factor (ANF) In caudal levels, small-sized ANF-immunoreactive cell bodies were found densely accumulated in the dorsal raphe nucleus and also scattered throughout the central gray (Fig. 16A,B). A few larger cell bodies were localized to the LC in the first level (Fig. 16A). Beginning in level B, cell bodies were found just.media1to the LC, in Barrington’s nucleus (Fig. 16B). In levels C and D, this group spread laterally and dorsally into the NTDL in the same fashion as ChAT,
12
et al. (1986a) also found that SP coexists within a subpopulation of these ANF- and ChAT-containing neurons in the NTDL and PPT.
Conclusions
Fig. 16. See text for details.
VIP, and CRF cells at these levels. In distribution and density, ANF-immunoreactivity resembled the localization of ChAT cell bodies (Fig. 3). ANF-positive fibers did not follow the typical peptide pattern. There was not the usual density of fibers surrounding the ventral PCS and NPD (Fig. 16A-E). Instead the ANF-immunoreactive fibers seemed more evenly distributed throughout the sections, with occasional densities seen medial to the NTM (Fig. 16C,D), and within the dorsal raphe nucleus (Fig. 16A-D). ANF, also referred to as atrial natriuretic peptide, is a circulating peptide hormone involved with cardiovascular and fluid balance regulation. This peptide has been mapped in the rat brain immunocytochemically (Jacobowitz er al., 1985; Skofitsch et al., 1985; Kawata et al., 1985; Standaert et al., 1986a,b; Zamir et al., 1986; Skofitsch and Jacobowitz, 1988). In accordance with all of these reports, ANF-immunoreactive cells were found mainly in the NTDL, parabrachial nucleus, and caudal dorsal raphe nucleus. According to Standaert et al. (1986a), ANF cells corresponded exactly with cells that stained for ChAT in the NTDL and peduncular pontine nucleus (PPT), but not the group in Barrington's nucleus. The present study also observed ANF cell bodies localized to the NTDL and PPT with a distribution similar to ChAT-immunoreactive cells. Standaert
Several areas within the dorsal tegmental region contain a variety of neurochemicals. The LC appears to be one of the most complex regions, with six neurochemicals (AChE, TH,GAL,NPY, NT, and VIP) contained within cell bodies. Table 1 summarizes which of the 16 neurochemicals were found using immunocytochemical techniques in cell bodies within the LC, NTDL, Barrington's nucleus, dorsal parabrachial nucleus, and dorsal raphe nucleus. The VIP cells, however, are few in number and seemed to remain confined to the dorsomedial quadrant of the LC, whereas NT and NPY cells were randomly scattered. Due to the wide array of chemicals parceled within the LC, the suggestion can be made that this nucleus is a highly active region, playing a part in a variety of functions and behaviors. The LC is known to have widespread projections in the central nervous system and is innervated by a number of brain regions that are presumably TABLE 1 Cell bodies (+ ) in the Dorsal Tegmental Region
ChAT GAL NT CCK SP NPY DynB GAD VIP CGRP
CRF
NTDL
Barrington's nucleus
+ +
+
+
+ + + + +
SST 5HT
TH ANF
+
+
LC
NPD
dr
+ +
+ + +
+
+ + + + + + + + +
+ +
+ + +
+ + + + + +
+
Sparse numbers of cell bodies not included.
+ +
13
functionally diverse. Thus the neurochemicals within its perikarya could aid in modulating the complex circuitry that would be required. References Albanese, A. and Butcher, L.L. (1980) Acetylcholinesterase and catecholamine distribution in the locus ceruleus of the rat. Brain Res. Bull., 5: 127-134. %rod, A., Chat, M., Paut, L. and Tappaz, M. (1984) Catecholaminergic and GABAergic anatomical relationship in the rat substantia nigra, locus coeruleus, and hypothalamus median eminence: Immunocytochemicalvisualization of biosynthetic enzymes on serial semithin plastic-embedded sections. J. Histochem. Cytochem., 3 2 1331-1338. Cedarbaum, J.M. and Aghajanian, G.K. (1979) Catecholamine receptors on locus coeruleus neurons pharmacological characterization. Eur. J. Phurmacol., 44:375. Chan-Palay, V., Jonsson, G. and Palay, S.L. (1978) Serotonin and substance P coexist in neurons of the rat's central nervous system. Proc. Natl. Acad Sci. USA,75: 1582-1586. Chronwall, B.M., DiMaggio, D.A., Massari, V.J., Pickel, V.M., Ruggiero, D.A. and ODonohue, T.L. (1985) The anatomy of neuropeptide-Y-containingneurons in rat brain. Neuroscience, 15: 1159-1181. Crawley, J.N., Olschowka, J.A., Diz, D.I. and Jacobowitz, D.M. (1985) Behavioral significance of the coexistence of substance P, corticotropin releasing factor, and acetylcholinesterase in lateral dorsal tegmental neurons projecting to the medial frontal cortex of the rat. Peptides, 6 891-901. Eiden, L.E., Nilaver, G. and Palkovits, M. (1982) Distribution of vasoactive intestinal polypeptide (VIP) in the rat brain stem nuclei. Brain Res., 231: 472-477. Everitt, B.J., Hokfelt, T., Terenius, T., Tatemoto, K., Mutt, V. and Goldstein, M. (1984) Differential co-existence of neuropeptide Y (NPY)-like immunoreactivity with catecholamines in the central nervous system of the rat. Neuroscience, 11: 443-462. Grzanna, R. and Molher, M.E. (1980a) Cytoarchitecture and dendritic morphology of central noradrenergic neurons. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Revisited, Raven Press, New York, pp. 83-97. Grzanna, R. and Molliver, M.E. (1980b) The locus coeruleus in the rat: An immunohistochemical delineation. Neuroscience, 5: 21-40. Guyenet, P.G. and Aghajanian, G.K. (1979) Ach, substance P and met-enkephalin in the locus coeruleus: Pharmacological evidence for independent sites of action. Eur. J. Phurmacol., 53: 319. Hartman, B.K. (1973) Immunofluorescence of dopamine-/3hydroxylase: Application of improved methodology to the localization of the peripheral and central noradrenergic nervous system. J. Histochem. Cytochem., 21: 312-332. Helke, C.J., Sohl, B.D. and Jacobowitz, D.M. (1980) Choline acetyltransferase activity in discrete brain nuclei of DOCA-salt hypertensive rats. Bruin Res., 193 292-298.
Hokfelt, T., Ljungdahl, A., Steinbusch, H., Verhofstad, A, Nilsson, G., Brodin, E., Pernow, B. and Goldstein, M. (1978) Immunohistochemical evidence of substance P-like immunoreactivity in some 5-hydroxytryptamine-containing neurons in the rat central nervous system. Neuroscience,3: 517-538. Holets, V.R., Hokfelt, T., Rokaeus, A., Terenius, L. and Goldstein, M. (1988) Locus coeruleus neurons in the rat containing neuropeptide Y, tyrosine hydroxylase or galanin and their efferent projections to the spinal cord, cerebral cortex and hypothalamus. Neuroscience, 2 4 893-906. Jacobowitz, D.M. (1978) Monoaminergic pathways in the central nervous system. In M.A. Lipton, A. Dimacio and Killiam (Eds.), Psychopharmacology A Generation of Progress, Raven Press, New York, pp. 119-130. Jacobowitz, D.M. and Skofitsch, G. (1986) Calcitonin gene-related peptide in the central nervous system: Neuronal and receptor localization, biochemical characterization and functional studies. In T.W. Moody (Ed.), Neural and Endocrine Peptides and Receptors, Plenum Press, New York, pp. 247-288. Jacobowitz, D.M., Skofitsch, G., Keiser, H.R., Eskay, R.L. and Zamir, N. (1985) Evidence for the existence of atrial natriuretic factor-containingneurons in the rat brain. Neuroendocrinology , 4 0 92-94. Johansson, O., Hokfelt, T., Pernow, B., Jeffcoate, S.L., White, N., Steinbusch, H.W.M., Verhofstad, A.A.J., Emson, P.C. and Spindel, E. (1981) Immunohistochemical support for three putative transmitters in one neuron: Coexistence of 5-hydroxytryptamine,substance P- and thyrotropin releasing hormone-like immunoreactivity in medullary neurons projecting to the spinal cord. Neuroscience, 6 1857. Kaplan, L.M., Spindel, E.R., Isselbaches, K.J. and Chin, W.W. (1988) Tissue-specific expression of the rat galanin gene. Proc. NatL Acad. Sci. USA, 8 5 1065-1069. Kawata, M., Nakao, K., Morii, N., Kim, Y., Yamashita, H., Imura, H. and Sano, Y. (1985) Atrial natriuretic polypeptide: Topographical distribution in the rat brain by radioimmunoassay and immunohistochemistry. Neuroscience, 16: 521-546. Kimmel, J.R., Pollock, H.G., Chance, R.E., Johnson, M.G., Reeve, Jr., J.R., Taylor, I.L., Miller, C. and Shively, J.E. (1984) Pancreatic polypeptide from rat pancreas. Endocrinology, 114 1725-1731. Knight, D.P. (1970) Histochemical demonstration of catecholamines and acetylcholine esterase in the same cell bodies in the locus coeruleus (rat hind brain). hot. R. Microsc. Soc.,6 26. Koelle, G.B. (1955) The histochemical identification of acetylcholinesterase in cholinergic, adrenergic and sensory neurons. J. Pharmacol. fip. Ther., 114 167-184. Lewis, P.R and Schon, F. (1975) The localization of acetylcholinesterase in the locus coeruleus of the normal rat and after 6-hydroxydopamine treatment. J. Anat., (London), 120: 373-385. Lin, T. (1980) Pancreatic polypeptide: Isolation, chemistry and biological function. In G.B. Jerzy Glass (Ed.), Gastrointestinal Hormones, Raven Press, New York, pp. 275306.
14 Loewy, A.D., Saper, C.B. and Baker, R.P. (1979) Descending projections from the pontine micturition center. Brain Res., 172 533-538. Martin, J.-L., Dietl, M.M., Hof, P.R., Palacios, J.M. and Magistretti, P.J. (1987) Autoradiographic mapping of [mono-125I]iodo-Tyrl0,Met0171 vasoactive intestinal peptide binding sites in the rat brain. Neuroscience, 23: 539565. Melander, T., Hokfelt, T. and Rokaeus. A. (1986) Distribution of galanin like immunoreactivity in the rat central nervous system. J. Comp. Neurol., 248: 475-517. Milner, T.A., Abate, C., Reis, D.J. and Pickel, V.M. (1989) Ultrastructural localization of phenylethanolamine Nmethyltransferase-like immunoreactivity in the rat locus coeruleus. Brain Res., 478: 1-15. Norgren, R. and Leonard, C.M. (1970) Ascending central gustatory pathways. J. Comp. Neurol., 150 217-238. Olschowka, J.A., ODonohue, T.L. and Jacobowitz, D.M. (1981) The distribution of bovine pancreatis polypeptidelike immunoreactive neurons in rat brain. Peptides, 2 309-331. Palkovits, M. and Jacobowitz, D.M. (1974) Topographical atlas of catecholamines and acetylcholinesterase-containing neurons in the rat brain 11. Hindbrain (Mesencephalon, Rombencephalon). J. Comp. Neurol., 157: 29-42. Pickel, V.M., Joh, T.H. and Reis, D.J. (1975a) Immunohistochemical localization of tyrosine hydroxylase in brain by light and electron microscopy. Brain Res., 85: 295-300. Pickel, V.M., Joh, T.H. and Reis, D.J. (1975b) Ultrastructural localization of tyrosine hydroxylase in noradrenergic neurons of>rain. Proc. Natl. Acad. Sci. USA, 7 2 659-663. Rokaeus, A. and Brownstein, M.J. (1986) Construction of a porcine adrenal medullary cDNA library and nucleotide sequence analysis of two clones encoding a galaninprecursor. Proc. Natl. Acad. Sci. USA, 83: 6287-6291. Satoh, K., Tohyama, M., Sakumoto, T., Yamamoto, K. and Shimizu, N. (1978) Descending projection of the nucleus tegmentalis laterodorsalis to the spinal cord; Studied by the horseradish peroxidase method following 6-hydroxyDOPA administration. Neurmci. Lett., 8: 9-15. Skofitsch, G. and Jacobowitz, D.M. (1985a) Autoradiographic distribution of 1251 calcitonin gene-related peptide binding sites in the rat central nervous system. Peptides, 4 975-986. Skofitsch, G. and Jacobowitz, D.M. (1985b) Calcitonin generelated peptide: Detailed immunohistochemical distribution in the rat central nervous system. Peptides, 6 721-745. Skofitsch, G. and Jacobowitz, D.M. (1985~)Immunohistochemical mapping of galanin-like neurons in the rat central nervous system. Peptides, 6 509-546.
Skofitsch, G. and Jacobowitz, D.M. (1988) Atrial natriuretic peptide in the central nervous system of the rat. Cell. Mol. Neurobwl., 8: 339-391. Skofitsch, G., Jacobowitz, D.M., Eskay, R.L. and Zamir, N. (1985) Distribution of atrial natriuretic factor-like immunoreactive neurons in the rat brain. Neuroscience, 16: 917-948. Sofroniew, M.V. (1979) Vasopressin- and neurophysin-immunoreactive neurons in the septa1 region, medial amygdala and locus coeruleus in colchicine-treated rats. Neuroscience, 15: 347-358. Standaert, D.G., Needleman, P. and Saper, C.B. (1986a) Organization of atriopeptin-like immunoreactive neurons in the central nervous system of the rat. J. Comp. Neurol., 253: 315-341. Standaert, D.G., Saper, C.B., Rye, D.R. and Wainer, B.H. (1986b) Colocalization of atriopeptin-like immunoreactivity with choline acetyltransferase and substance P-like immunoreactivity in the pedunculopontine and laterodorsal tegmental nuclei in the rat. Brain Rex, 382: 163-169. Sutin, E.L. and Jacobowitz, D.M. (1988) Immunocytochemical localization of peptides and other neurochemicals in the rat laterodorsal tegmental nucleus and adjacent area. J. Comp. Neurol., 270 2f3-270. Tatemoto, K, Rokaeus, A., Jornvall, H., McDonald, T.J. and Mutt, V. (1983) Galanin - a novel biologically active peptide from porcine intestine. FEBS Lett., 164: 124-128. Uhl, G.R., Goodman, R.R. and Snyder, S.H. (1979) Neurotensin-containing cell bodies, fibers and nerve terminals in the brain stem of the rat: Immunohistochemical mapping. Brain Res., 167 77-91. Vincent, S.R., Satoh, K., Armstrong, D.M. and Fibiger, H.C. (1983) Substance P in the ascending cholinergic reticular system. Nature (London), 306: 691-699. Vrontakis, M.E., Peden, L.M., Duckworth, M.L. and Frisen, H.G. (1987) Isolation and characterization of a complementary DNA (galanin) clone from estrogen-induced pituitary tumor messenger RNA. J. Biol. Chem., 262 1675516758. Wilson, P.M. (1985) A photographic perspective on the origins, form, course and relations of the acetylcholinesterase-containing fibers of the dorsal tegmental pathway in the rat brain. Brain Res. Rev., 1 0 85-118. Young, W.S., 111, Uhl, G.R. and Kuhar, M.J. (1978) Iontophoresis of neurotensin in the area of the locus coeruleus. Brain Res., 150 431-435. Zamir, N., Skofitsch, G., Eskay, R.L. and Jacobowitz, D.M. (1986) Distribution of immunoreactive atrial natriuretic peptides in the central nervous system of the rat. Brain Res., 365: 105-111.
C.D.Barnes and 0. Pompeiano (Eds.) Profiress iir Brain Researcl,, Vol. 88 0 1WI Elsevier Science Puhlishers B.V.
15 CHAPTER 2
Noradrenergic locus coeruleus neurons: their distant connections and their relationship to neighboring ( including cholinergic and GABAergic) neurons of the central gray and reticular formation B.E. Jones Department of Neurology and NeurosuGery, McCill Unirersiiy, Montreal Neurological Institute, Unicersity Street, Montreal, Quebec, Canada
Noradrenergic LC neurons appear to be relatively unique in the brain, being unsurpassed in the divergence and ubiquity of their projections through the central nervous system. In this regard, they share certain characteristics with peripheral noradrenaline neurons of the sympathetic nervous system. As such they would be assumed to play a very general role in modulating the activity of large populations of neurons in multiple, functionally diverse systems. Like other periventricular and reticular neurons, they have the potential to receive afferent information from multiple sources via long dendrites, upon which the majority of their inputs from brainstem and forebrain may arrive. They appear closely related to the cholinergic neurons of the laterodorsal tegmental nucleus,
their neighbors that are located medial and rostra1 t o them within the periventricular gray and that have similarly oriented and positioned long dendrites that would allow reception of similar afferent input as the LC neurons and also possibly interaction with the LC neurons. As evidenced by input to the noradrenergic cell bodies in the compact portion of the nucleus, a moderate GABAergic innervation, that may derive in part from local neurons, could have a potent influence on the activity of the cells. Periventricular GABAergic cells could also serve as intermediaries to other afferent input, from a distance, terminating in the periventricular region or from local neurons such as the cholinergic cells of the laterodorsal tegmental nucleus.
Key w o r k noradrenaline, acetylcholine, GABA, laterodorsal tegmental nucleus
Introduction
Originally named in the last century for the bluish color that its cells cast from their melanin content onto the floor of the 4th ventricle, the locus coeruleus (LC) nucleus was defined by the neurons that contain pigment in the dorsolateral pons of the human brain (Olszewski and Baxter, 1954; for historical review see Russell, 1955).
Most appropriately, yet ironically, in this century, the same cells were found, when suitably treated with the Falck-Hillarp technique, to fluoresce a bluish green color, due to their catecholamine content (Dahlstrom and Fwe, 1964). As evident in Nissl-stained sections in the rat, the LC neurons lie in a cluster within the periventricular gray in the dorsolateral corner of the 4th ventricle of the pons (Fig. 1). In this species, the cell
16
Fig. 1. Photomicrograph of Nissl-stained section through the caudal compact portion of the locus coeruleus nucleus (LC) in the rat. The cell bodies lie in a tight cluster within the periventricular gray in the corner of the 4th ventricle of the pons. Located caudal and at this level, lateral to the LC is the oral pole of the medial vestibular nucleus (MVe). The large cells and rootlets of the mesencephalic trigeminal nucleus (5Me) occupy a traingular region lateral to the LC and vestibular nucleus and dorsal to the motor trigeminal nucleus (5Mo) (and supratrigeminal nucleus). Dorsal and lateral to the LC lies the superior cerebellar peduncle (scp) and narrow ventral parabrachial nucleus that lies beneath it at this level.
bodies form a compact nucleus most caudally within the periventricular gray; they disperse to a certain degree more rostrally to occupy a broader area of the gray and to extend beneath the gray into the dorsolateral pontine tegmentum, where they are commonly referred to as the LC pars a or subcoeruleus (pars a)(Meesen and Olszewski, 1949; Paxinos and Watson, 1986). As noted by some of the early anatomists, the LC neurons lie within the visceral column of the brainstem. Given the tight aggregation of the cells within the periventricular gray and their catecholamine content, the LC neurons resemble an autonomic ganglion. Their proximity to and close association with the mesencephalic trigeminal sensory ganglion cells (Fig. l), as noted by the early anatomists, suggested the possibility that like these neurons, they developed from neural crest cells and migrated into the central nervous system, there perhaps representing a central sympathetic ganglion. Indeed the very diffuse projection
system, revealed by histofluorescence to emanate from this small group of adrenergic neurons and to encompass via varicose axons the entire central nervous system (Ungerstedt, 19711, resembles a sympathetic innervation. It was even suggested (Swanson et al., 1977) that these neurons innervated blood vessels in the brain, an association, however, which proved to be very limited in extent (Swanson et al., 1978; Jones, 1982). The primary association of central noradrenergic varicosities has been found to be with other neuronal elements, even though the majority of such contiguities, as with the association of peripheral adrenergic elements with their targets (Burnstock and Costa, 1975), may not be characterized by synaptic junctions that are more typical of most central bouton connections (Descarries et al., 1977; Seguela et al., 1990). Thus multiple features of the LC cells are distinct from other neurons within the brain and suggest a similarity with peripheral sympathetic neurons. From an
17
anatomical perspective, the uniqueness of the noradrenergic LC neurons can be considered by comparing their connections to those of other central neurons within the vicinity, particularly neurons of the periventricular gray and reticular formation of the pons.
gitudinal catecholamine bundle (Jones and Friedman, 1983) intermixed with lateral fascicles of the parvicellular reticular formation (and Probst’s
Efferent projections
As studied first by histofluorescence and subsequently by immunohistochemistry, the projections of .the LC neurons have long been known to be ubiquitous through the central nervous system (Ungerstedt, 1971; Maeda and Shimuzu, 1972; Lindvall and Bjorklund, 1974; Swanson and Hartman, 1975). Autoradiographic analysis of the anterograde transport of radiolabelled protein (Jones and Moore, 1977; Jones and Yang, 1985), makes it apparent that these projections encompass regions not only divergent in space but also divergent in function. Thus within the brainstem, the LC projects to the reticular formation and also to cranial nerve nuclei, including both motor (12th, loth, 7th and 5th) and sensory (cochlear and solitary tract nuclei) and both somatic and visceral systems (Fig. 2). Other neuronal fields of the medial reticular formation that, like the LC, give rise to long descending and ascending projections also innervate cranial nerve nuclei, however only motor nuclei and somatic (and branchiomeric) motor nuclei that are predominantly involved in orientation of the eyes and ears (3rd, 4th, 6th and 7th). The major descending pathway of the noradrenergic fibers travels within the lon-
Fig. 2. Schematic diagrams of sections through the brainstem and spinal cord of the rat depicting anterograde transport of radiolabelled protein following an injection of [3H]leucine into the LC nucleus. Descending fibers pass through lateral reticular fascicles (forming the longitudinal catecholamine bundle) to reach the lower brainstem and spinal cord, and ascending fibers collect within the central reticular fasciculus to reach the upper brainstem and the forebrain. The levels correspond approximately to those of the rat atlas (Paxinos and Watson, 1986). (This figure together with Figure 3 was copied with permission from Jones and Yang, 1985.)
c5
18
+5
19
bundle) that carry locally distributing fibers from lateral reticular neurons and sensory relay systems. The locus axons thus utilize a propriobulbar system to distribute collaterals to multiple nuclei through the brainstem and then continue into the spinal cord through lateral fascicles. In the spinal cord, the major projection from the LC travels within the lateral funiculus, from which it extends fibers primarily into the intermediate zone and ventral horn, here similar to the innervation by other long projecting medial reticulospinal neurons (Fig. 2). However, efferent projections from the LC and subcoeruleus (SC) can also be shown to encompass the dorsal horn and the preganglionic parasympathetic neurons of the sacral cord, depending upon the group of-LC and SC neurons labelled by the anterograde tracer (Martin et al., 1979; Westlund et al., 1982; Jones and Yang, 1985; see also Grzanna and Fritschy and Proudfit and Clark, respectively, this volume). The brainstem and spinal projections of the LC neurons would thus appear to be much more ubiquitous (including both sensory and motor in addition to other reticular zones) than those of the long projecting reticular neurons of the medial reticular formation. The quantitatively most important projections from the LC course rostrally within the central reticular fasciculus together with fibers from other reticular neurons through the midbrain into the forebrain (Jones and Yang, 1985) (Fig. 2). A major contingent of this ascending fiber system innervates the intralaminar and midline thalamic nuclei, thus providing, like other neurons of the reticular formation, an important influence to the nonspecific thalamo-cortical projection system (Fig. 3). In addition, and in contradistinction to most other reticular neurons of the medial core, however, the locus neurons densely innervate the anterior (anteroventral) nuclei and the dorsal lateral geniculate. They also project densely into
and through the subthalamus and hypothalamus where, like other reticular neurons, they innervate the zona incerta and lateral hypothalamus (and posterior and dorsal hypothalamus), but unlike other reticular neurons, they also innervate the periventricular and arcuate nuclei, among others. In the posterior and lateral hypothalamus, the noradrenergic fibers may influence cortically projecting neurons located in that region (Saper, 1985). As from other reticular neuronal fields, fibers continue forward from the LC through the medial forebrain bundle into the basal forebrain, where they collateralize in the region of the cortically projecting neurons of the substantia innominata or nucleus basalis of Meynert (Saper, 1984; Jones and Cuello, 1989). They also continue forward into the horizontal and vertical limbs of the diagonal band and the septum where they may impinge upon neurons that project to the hippocampus. They would, thus have the capacity to influence, via this ventral extrathalamic cortical relay system, the neo-, paleo- and archi-cortex. In this regard, they do not differ from other long, projecting reticular neurons (or central gray neurons, Eberhart et al., 1985) that similarly project into and through the reticular-like, isodendritic core of the forebrain (Leontovich and Zhukova, 1963; Ramon-Moliner and Nauta, 1966; Jones and Yang, 1985; Jones and Cuello, 1989). Where the LC neurons are distinct from the latter is that they also project beyond these cortically projecting neurons to directly innervate the cerebral cortex, including the neo-, paleo-and archi-cortex (Jones and Yang, 1985). Retrograde transport analysis shows that few brainstem neurons project directly to the cerebral cortex, and these are primarily comprised of other monoamine neurons, notably the serotonin and dopamine neurons of the midbrain and some cholinergic cells (see below). Even among these cell groups, however, no one nucleus projects to as many or as
Fig. 3. Schematic diagrams of sections through the forebrain depicting the continuation of rostra1 projections from the LC (Fig. 2). Continuing from the central reticular fasciculus, fibers enter the thalamus through a dorsal leaf and ascend into the subthalamus, hypothalamus and basal forebrain through a ventral leaf predominantly in the medial forebrain bundle.
20
hypothalamus (Jones and Yang, 19851, as has been found electrophysiologicallyas well (Guyenet, 1980). Such diffuse projections would be expected to exert a very general influence upon the central nervous system and perhaps to respond to multiple sources and types of afferent input. Fig. 4. Schematic diagram of section through the pons at the level of the LC depicting retrograde transport of true blue (triangles) from spinal cord and of nuclear yellow (circles) from the subthalamus-hypothalamus.One cell within the locus coeruleus was retrogradely labeled by both tracers (represented by an asterisk). (This figure was copied with permission from Jones and Yang, 1985.)
widely divergent and distant regions as the LC nucleus (Azmitia and Segal, 1978; Moore, 1980). Even though in their efferent projections the LC neurons share a certain number of characteristics with other neuronal groups of the brainstem and reticular formation, they also differ distinctly from the latter by the ubiquity and long distance of their projections. Although the efferent projections of the LC nucleus can thus be portrayed as diffuse, when compared to the widespread but more limited projections of central gray and reticular fields, they are nonetheless somewhat segregated according to the target region of neuronal groups within the nucleus (Loughlin et aZ., 1986a,b). By retrograde transport studies, it is apparent that, for the most part, different LC neurons give rise to long, descending projections than those that give rise to long ascending projections (Fig. 41, the latter being concentrated within the ventral part of the principal nucleus and the subcoeruleus nucleus (Guyenet, 1980; Jones and Yang, 1985). In this regard, almost all reticular neurons similarly give rise to long projections either to the spinal cord or to the forebrain, the two types being intermingled within reticular fields. As in certain other reticular regions, a small number of LC neurons can be doublelabeled from, and thus have bifurcating axons with long projections to, both the spinal cord and
Afferent connections
In early studies of the afferent input to the LC using retrograde transport of horseradish peroxidase (HRP); Sakai et d.,1977; Cedarbaum and Aghajanian, 19781, it was found that many of the regions that received dense inputs from the LC neurons projected, in turn, back upon these neurons. Thus the LC nucleus was found to have wide and divergent sources of afferent input from the brainstem, spinal c$rd and forebrain. Most surprisingly, on the other hand, it was more recently reported that following injections of another tracer (wheat germ agglutinin, WGA, conjugated to HRP) into the compact caudal portion of the LC nucleus, only two cell groups were found to be consistently and strongly retrogradely labelled, the paragigantocellularis lateralis and the prepositus hypoglossi nuclei within the medulla (Aston-Jones et aZ., 1986). These anatomical results have been confirmed by parallel electrophysiological studies (Ennis and Aston-Jones, 1987). Although negative results in anatomy and physiology are never conclusive (and could perhaps be explained by failure of uptake and retrograde labelling of tracer on the one hand and synaptic or antidromic activation on the other due to anesthesia, for example), they definitely suggest significant nuances in the scheme of connections. These results appear to indicate that a limited number of afferents to the LC actually innervate the cell bodies of these neurons or at least those which are tightly packed into the caudal part of the nucleus, where the WGA-HRP injections were placed. Such a possibility was also revealed earlier by electron microscopic analysis of the LC in both the cat (Ramon-Moliner, 1974)
21
Fig. 5. Photomicrographthrough the LC and SC of a section processed for dopamine-p-hydroxylae(D/3H) immunohistochemistry using the peroxidase-antiperoxidase(PAP) technique with silver intensification of the diamino benzidine (DAB) reaction product. In this section (taken from a level slightly rostral to that shown in Figure 11, the major dendritic processes of the LC neurons extend medially within the periventricular gray. The SC neurons lying within the region of the central reticular fasciculus appear to orient their dendrites in a radial manner.
and rat (Groves and Wilson, 1980a,b); it was found that the soma of these neurons was surrounded by a glial sheath as if to insulate them from the surrounding neuropil. Input to the soma and proximal dendrites occurs onto a limited number of somatic and dendritic appendages that pierce the glial sheath. But the major input was found to occur on distal dendrites of the LC neurons (Groves and Wilson, 1980a,b). The LC neurons, like other central gray and reticular neurons, have long dendrites that extend outside the limits of the nucleus (Fig. 51, as has been noted previously (Swanson, 1976a; Grzanna and Molliver, 19801. Typical of isodendritic neurons, or more particularly leptodendritic neurons of the central gray (Ramon-Moliner, 19681, their dendrites course for long distances within the periventricular gray. Lying in the corner of the fourth ventricle, the cells extend their dendrites most prominently in a rostral and medial direction, as can be appreciated in horizontal as well as coronal sections (Fig. 5). Many of the dendritic
processes, particularly from the most dorsal cells, run very close to the ventricle within the ependyma1 zone; others running parallel to these fan out across the gray to fill the entire zone medial and rostral to the nucleus. The cells located beneath the gray in the subcoeruleus, on the other hand, appear more like other isodendritic reticular neurons with more radiating dendrites extending through the fascicles of the tegmentum (the central reticular fasciculus) as can be appreciated in sagittal and, to some extent, in coronal sections (Fig. 5). Through contact with distal dendrites, the af.ferent input to the LC neurons is thus potentially, like that to other periventricular and reticular neurons, very extensive. Indeed most of the regions originally defined as being sources of afferent input to the LC (Cedarbaum and Aghajanian, 19781, have been known by anterograde degeneration and, subsequently, anterograde transport to project into and through the periventricular gray and dorsolateral tegmentum of the pons in the
22
vicinity of the locus and subcoeruleus cells and/or their dendrites. From the spinal cord, a contingent of fibers, originally referred to as the medial spinothalamic and spinoreticular systems, ascends through the reticular formation to pass in loose tegmental fascicles through the dorsolateral pontine tegmentum (DLPT) in the region of the SC nucleus and into the central gray (Nauta and Kuypers, 1958;Anderson and Berry, 1959; Mehler et al., 1960). The most dense termination of spinoreticular fibers occurs however within the medullary reticular formation, and from this level neurons project rostrally into and through the dorsolateral pontine tegmentum in the region of the SC and also the LC (Sakai et al., 1979; Jones and Yang, 1985; Vertes et al., 1986). More recently Aston-Jones and his colleagues (1986) have shown that perhaps the most important input to the LC perikarya (at least within the caudal compact portion of the nucleus labelled in their studies) may originate more specifically from the nucleus paragigantocellularis lateralis that has long been known to receive dense spinoreticular input (Mehler, et al., 1960). It is possible that the input from the medullary reticular formation has preferential access to the soma and proximal dendrites of the LC neurons, although this possibility would need to be investigated at the electron microscopic level. From the forebrain, fibers travel down within periventricular fascicles and the medial forebrain bundle to extend into the periaqueductal and periventricular gray and dorsolateral pontine tegmentum (Nauta and Haymaker, 19691, where the LC and SC perikarya and dendrites lie. Thus from regions of the septum, preoptic region, substantia innominata and hypothalamus that receive descending input from cortical regions (including limbic, sensory, motor or association areas of archi-, paleo- or neocortex) and ascending input from the LC, fibers descend through the medial forebrain bundle to converge in the dorsolateral pontomesencephalic tegmentum, central gray and region of the LC (Swanson, 1976b; Saper et al., 1979; Swanson and Cowan, 1979; Holstege et al., 1985; Holstege,
1987; Grove, 1988; Simerly and Swanson, 1988; Jones and Cuello, 1989). Many of these fibers appear to converge in the peri-ependymal zone where the LC dendrites extend outside the nucleus, although most have also been traced into the nucleus itself. In either case, the descending fibers are usually represented in association with the more rostral or central portion of the LC nucleus. Obviously, a careful study of afferents to different portions of the LC nucleus (in addition to the compact caudal portion shown in Fig. 1) and SC nucleus is necessary, first, and careful consideration of the dendritic extensions of these cells in relation to the afferent input is important, second, before the limitations of that input are decided. A tentative summary of the current knowledge would be that LC neurons, like other leptodendritic neurons of the periventricular gray (Ramon-Moliner, 19681, receive very limited input onto their soma and proximal dendrites but extensive input onto more distal dendrites which may come directly (originating in the periventricular and reticular-like isodendritic core) or indirectly (originating in sensory-motor bulbar and spinal or cortical regions) from the multiple regions to which the LC projects. Relationship to neighboring cholinergic and GAFiAergic neurons
Concentrated more rostrally and medially than the LC and SC neurons in the rat, cholinergic neurons also lie within the periventricular gray forming the laterodorsal tegmental (LDT) nucleus and beneath the gray in what (in parallel to the LC neurons) has been called the LDT pars (Y or sub-LDT (Jones, 1990a). These cells thus also form within the visceral column of the brainstem periventricular gray and lie in the same relative position to the noradrenergic LC neurons in the pons as the autonomic cholinergic cells do to the noradrenergic and adrenergic (A2421 neurons within the periventricular gray in the medulla. The LDT and sub-LDT cholinergic cells give rise to descending projections into the pontomedul-
23
A
Fig. 6. Photomicrographs of sections taken through the LC adjacent to the section shown in Figure 1 and processed for choline acetyltransferase (ChAT) immunohistochemistry (A) and glutamic acid decarboxylase (GAD) immunohistochemistry (B). In both cases the DAB reaction product of the PAP process was intensified with silver. In A, the LC nucleus stands out as having few ChAT-immunoreactivefibers within its boundaries. A few ChAT-immunoreactive cell bodies are located on the edges (medial and ventral, and dorsal and lateral) of the nucleus. ChAT-immunoreactive processes are moderately dense within the periventricular gray ventral and medial to the LC nucleus. A dense plexus of ChAT-immunoreactive processes is evident within the ventral parabrachial nucleus and a light, but significant, number of fibers within the medial vestibular nucleus. Thick black processes emanate from the motor trigeminal neurons and extend dorsal to the nucleus into the region of the supratrigeminal and mesencephalic trigeminal nuclei. In B, the LC nucleus is characterized by a moderate density of GAD-immunoreactive terminals, that is less than that to the periventricular gray medial and ventral to the nucleus. Small GAD-immunoreactive cells are present within the medial vestibular nucleus and periventricular gray and also in very small numbers within the LC nucleus as evident at higher power (Fig. 8). (Refer to the adjacent section shown in Figure 1 for identification of structures.)
24
P0.5
L
..
Fig. 7. Schematic representation of noradrenergic (green), cholinergic (black) and GAI3Aergic (red) cells and processes within the periventricular gray and tegmentum beneath the gray in the region of the rostra1 LC and subcoeruleus (as shown in Fig. 5 ) and caudal, lateral edge of the laterodorsal tegmental nucleus. The figure was composed by superimposing camera lucida drawings of DPH-, ChAT- and GAD-immunohistochemically stained (and silver intensified), adjacent sections. The full density of each component however was not represented due to overlapping profiles. Apparent dendritic processes of the noradrenergic LC neurons extend medially through the region of cholinergic neurons whose dendrites also extend in the same horizontal plane near the noradrenergic ones. In addition, varicose, apparent axonal processes of the noradrenergic neurons are also of moderate density in the region of the cholinergic neurons and their dendrites. Reciprocally ChAT-immunoreactivevaricose fibers and varicosities are moderately dense through the region of the gray where the D/3H-immunoreactiveprocesses extend. GAD-immunoreactive cells are scattered through the periventricular gray and tegmentum beneath the gray in the region of the subcoeruleus. A high density of GAD + varicosities is present through the entire region where noradrenergic cell bodies and processes as well as cholinergic cell bodies and processes are located.
lary reticular formation but not to the spinal cord (Rye ef aL, 1988; Jones, 1990a1, and they also give rise to long, ascending projections that, like other reticular projections, innervate the nonspecific thalamocortical system and extend to a certain degree into the ventral extra-thalamic cortical relay system (Satoh and Fibiger, 1986; Hallanger ef aL, 1987; Jones and Beaudet, 1987; Hallanger and Wainer, 1988; Jones and Cuello,
1989; Jones, 199Ob). Although they do provide a certain innervation to the cerebral cortex, this projection is restricted and includes a limited portion of the medial frontal cortex and insular cortex (Jones, unpublished observations; Satoh and Fibiger, 1986). These cells thus do not provide a diffuse innervation to the entire central nervous system like the LC neurons, but they provide a widespread innervation mainly to other
25
Fig. 8. High power photomicrograph taken through the ventral caudal portion of the LC (as shown in Figures 1 and 6, but from another animal) showing GAD-immunoreactive cells and varicosities. The section was processed for PAP immunohistochemistry with silver intensification. The ependyma of the periventricular zone is in the upper left corner as evident by a high density of GAD + varicosities. The two small GAD-immunoreactive cells in the middle lie within the confines of the LC nucleus, whereas the two larger GAD-immunoreactive cells to the right lie on the lateral edge of the nucleus within the oral pole of the medial vestibular nucleus. Varicosities can be seen to surround some of the clear, spherical soma of the LC neurons in this thick (25 pm) section. (Magnification bar = 10 pm.)
reticular and reticular-like neurons of the isodendritic core of the brainstem and forebrain (Leontovich and Zhukova, 1963; Ramon-Moliner and Nauta, 19661, that in turn project to spinal or cortical regions, respectively. Cholinergic neurons do not appear to provide a significant innervation to the LC cell bodies in the rat as can be appreciated by a low magnification photomicrograph through the region of the caudal compact portion of the nucleus (Fig. 6). This observation may again corroborate the principle that few afferents influence the LC neurons via their soma. On the other hand, the cholinergic cells and processes lie in close proximity to the distal dendrites of the LC neurons (Fig. 7) (Jones, 1989). The dendrites of the cholinergic cells are, like those of the LC neurons, oriented in a prominent horizontal manner through the periventricular gray and in the rostral LC and caudal LDT, the processes from the two cell types run in parallel and close proximity to one
another, including through the ependymal and subependymal zone as well as more ventrally through the gray. The sub-LDT and SC cells have similarly radiating dendrites within the tegmentum beneath the gray, one set more rostral than the other. According to their overlapping positions, the cholinergic and noradrenergic neurons may thus receive similar afferent input. In fact, retrograde transport studies (utilizing WGAHRP) have indicated widespread sources of input to the laterodorsal tegmental nucleus (Satoh and Fibiger, 1986) that overlap considerablywith those previously revealed for the LC (Cedarbaum and Aghajanian, 1978). In addition, studies using anterograde transport, as cited above, have also shown input into the region of the latero-dorsal tegmental cell bodies as well as their dendrites from septal, preoptic, basal forebrain and hypothalamic regions (Satoh and Fibiger, 1986; Grove, 1988; Simerly and Swanson, 1988; Jones and Cuello, 1989). The cholinergic and noradren-
26
ergic neurons may also possibly influence each other by dendro-dendritic contacts or axo-dendritic contacts through intermingled dendritic processes and varicose fibers of the two cell types within the central gray. Gamma amino butyric acid (GABA)-containing or -synthesizing neurons, as evidenced by glutamic acid decarboqdase (GAD) immunostaining, are located through the periventricular zone and dorsal tegmentum of the brainstem (Mugnaini and Oertel, 1985; Jones, in preparation). A large number of GAD-immunoreactive cells are found within the medial (and superior) vestibular nucleus that, at its oral pole, abuts the LC caudally and, to a certain extent, laterally (Fig. 6). A small number of GAD +cells that lie within the vestibular nucleus are thus found on the lateral edge of the LC, and an even smaller number of (smaller) GAD + cells lie within the LC nucleus itself (Fig. 8) (as previously reported by Iijima and Ohtomo, 1987). A moderately dense innervation by GAD-immunoreactive processes is obvious within the LC nucleus and as typical of such innervations in the brainstem is, in part, characterized by a perisomatic distribution of varicosities of light-to-moderate density (as compared to other reticular and vestibular perikarya that are characterized by perisomatic varicosities of high density). As studied by GABA immunocytochemistry, the GABAergic terminals represent a proportion of axo-somatic terminals of the noradrenergic cell bodies (Iijima and Ohtomo, 1987). The GABAergic innervation has been shown to be derived in part from GABA neurons located caudal to the LC in the prepositus hypoglossi nucleus (Shipley et al., 1988; see Aston-Jones et al., this volume) but would also probably derive from more proximal GABA neurons, such as those located within the nucleus that resemble interneurons or those located lateral and caudal to the nucleus. There is also a moderate-to-high density of GAD + terminals and varicosities and GAD + cells, within the periventricular gray, medial and rostral to the LC where the LC dendrites extend. GAD + cells are present in higher num-
bers in amongst cholinergic neurons of the LDT nucleus (Fig. 71, particularly at more rostral levels. Such local GABAergic neurons may provide an intermediary between cholinergic and noradrenergic neurons or could also allow a differential effect of afferent input onto the two cell types. The other major inputs to the LC that are evident immunohistochemically (excepting peptidergic; see Sultin and Jacobowitz, this volume) are catecholaminergic and serotonergic. These would comprise the major proportion of terminals apparent on LC cell bodies and dendrites within the nucleus (Pickel et al., 1977; Uger and Descarries, 1978; Groves and Wilson, 1980a,b; Iijima and Ohtomo, 1987; Milner et al., 1989). The catecholamine input would derive from adrenergic neurons located within the prepositus hypoglossi (C3) and paragigantocellularislateralis (C1) regions (Pieribone et al., 1988) and also from the noradrenergic LC neurons themselves, although dendro-dendritic interactions between these cells may be more important (Groves and Wilson, 1980a,b). The serotoninergic input was previously documented to derive from dorsal raphe neurons in part (Sakai et al., 1977; Cedarbaum and Aghajanian, 19781, but may also come from other serotonin neurons scattered through the periventricular gray and in the region of the LC as well, identified in both the cat and the rat brain (Uger et al., 1979; Iijima, 1989). There would, thus, appear to be a considerable potential for interactions among neighboring periventricular neurons, including noradrenergic, serotoninergic, cholinergic and GABAergic possibly via dendro-dendritic and axo-dendritic contacts. Acknowledgements
I would like to express my special appreciation to Lynda Mainville who has assisted with all of the neuroanatomical and histochemical studies described in this review. The research was supported by a grant from the Medical Research Council of Canada.
27
Abbreviations used in schematic figures Spinal Cord (Fig. 2) cu: cuneate fasciculus dcs: dorsal corticospinal tract gr: gracile fasciculus lfu: lateral funiculus Vfu: ventral funiculus Brain Stem (Figs. 2 and 4) oculomotor nuc 3: 4v fourth ventricle 7: facial nuc vestibulocochlear nerve 8: bic: brachium inf colliculus central gray CG: CLi: caudal linear raphe nuc cp: cerebral peduncle D: nuc Darkschewitsch DCo: dorsal cochlear nuc DPB: dorsal parabrachial nuc DR: dorsal raphe DTg: dorsal tegmental nuc dtgx: dorsal tegmental decussation icp: inferior cerebellar peduncle IF interfascicular nuc InC interstitial nuc Cajal IP interpeduncular nuc Kolliker-Fuse nuc KF: Lc: locus coeruleus LSO: lateral superior olive lateral nuc trapezoid body LTz: lateral vestibular nuc LVe: middle cerebellar peduncle mcp: Me5: mesencephalic trigeminal nuc MGD: medial geniculate nuc, dorsal ml: medial lemniscus mlf medial longitudinal fasciculus motor trigeminal nuc Mo5: MTz: medial nuc trapezoid body MVe: medical vestibular nuc Pr5: principal sensory trigeminal PrH prepositus hypoglossal nuc pyramidal tract PY: Rgc: reticularis gigantocellularis
Rgca: Rmes: RMg: RN: Rob: Rp: RPa: Rpc: RPn: Rpo: rs:
SCD: SCI: scp:
scs: SNC SNL: SNR sp50: tz: vco: VPB: vsc: VTA vtgx:
reticularis gigantocellularis pars (Y reticularis mesencephali raphe magnus nuc red nuc raphe obscurus nuc reticularis parvicellularis raphe pallidus nuc reticularis pontis caudalis raphe pontis nuc reticularis pontis oralis rubrospinal tract superior colliculus, deep layer superior colliculus, intermediate layer superior cerebellar peduncle superior colliculus, superficial layer substantia nigra, compact substantia nigra, lateral substantia nigra, reticular spinal tract trigeminal nerve nuc, oral trapezoid body ventral cochlear nuc ventral parabrachial nuc ventral spinocerebellar tract ventral tegmental area ventral tegmental decussation
Diencephalon and Telencephalon (Fig. 3) 3v third ventricle AA: anterior amygdaloid area anterior commissure, anterior aca: accumbens nuc Acb: anterior cingulate cortex ACg: AD: anterodorsal thalamic nuc AHy: anterior hypothalamic area anteromedial thalamic nuc AM: Arc: arcuate nuc AV anteroventral thalamic nuc BL basolateral amygdaloid nuc BM: basomedial amygdaloid nuc corpus callosum cc: ce: central amygdaloid nuc cg: cingulum c1: claustrum CL centrolateral thalamic nuc CM: central medial thalamic nuc
28
m DA DG DLG: DM ec: En: EP: E fi: fmi FrPaM FrPaSS: FStr: G GP: Hi: ic: ICj: IMD: La: LH: LHb: lo: LOT LP
Ls: LV: MD: MDL Me: mfb:
MHb: ml:
mt: opt: ox: PC PCg: PLCO: PMCo: Po: PO:
PT:
caudate putamen dorsal hypothalamic area dentate gyrus dorsal lateral geniculate dorsomedial hypothalamic nuc externalcapsule , endopiriform nuc entopeduncular nuc fornix fimbria forceps minor corpus callosum frontoparietal cortex, motor area frontoparietal cortex, somatosensory area fundus striati gelatinosus nuc thalamus globus pallidus hippocampus internal capsule islands Calleja intermediodorsal thalamic nuc lateral amygdaloid nuc lateral hypothalamic area lateral habenular nuc lateral olfactory tract lateral olfactory tract nuc lateral posterior thalamic nuc lateral septal nuc lateral ventricle mediodorsal thalamic nuc mediodorsal thalamic nuc, lateral medial amygdaloid nuc medial forebrain bundle medial habenular nuc medial lemniscus mammillothalamic tract optic tract optic chiasm paracentral thalamic nuc posterior cingulate cortex posterolateral cortical amygdaloid nuc posteromedial cortical amygdaloid nuc posterior thalamic nuclear group primary olfactory cortex paratenial thalamic nuc
PVA PVP: Re: RF Rh: Rt: SCh: SI: sm:
so: st:
m Tu: vhc: VM: VMH VP: VPL VPM ZI
paraventricular thalamic nuc, anterior paraventricular thalamic nuc, posterior reunions thalamic nuc rhinal fissure rhomboid thalamic nuc reticular thalamic nuc suprachiasmatic nuc substantia innominata stria medullaris thalamus supraoptic hypothalamic nuc stria terminalis taenia tecta olfactory tubercle ventral hippocampal commissure ventromedial thalamic nuc ventromedial hypothalamic nuc ventral pallidum ventroposterior thalamic nuc, lateral ventroposterior thalamic nuc, medial zona incerta
References Anderson, F.D. and Berry, C.M. (1959) Degeneration studies of long ascending fiber systems in the cat brain stem. J. Comp. Narol., 111: 195-229. Aston-Jones, G., Ennis, M., Pieribone, B.A., Thompson Nickell, W. and Shipley, M.T. (1986) The brain nucleus locus coeruleus restricted afferent control of a broad efferent network. Science, 234-737. Azmitia, E.C. and Segal, M. (1978) An autoradiographic analysis of the differential .ascending projections of the dorsal and median raphe nuclei in the rat. J. Comp. Narol., 1979 641-668. Burnstock, G. and Costa, M. (1975) Adrenergk Neurons, John Wiley & Sons, New York, 225 pp. Cedarbaum, J.M. and Aghajanian, G.K. (1978) Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique. J. Comp. Neurol., 178 1-16. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta PhysioL Scand., 62, Suppl., 232 1-55. Descarries, L,Watkins, KC. and Lapierre, Y. (1977) Noradrenergic axon terminals in the cerebral cortex of rat. 111. Topometric ultrastructural analysis. Brain Res., 133 197222. Eberhart, J.A., Morrell, J.I., Krieger, M.S. and Pfaff, D.W. (1985) An autoradiographic study of projections ascending
29 from the midbrain central gray, and from the region lateral to it, in the rat. J. Cow. Neurd, 241: 285-310. Ennis, M. and Aston-Jones, G. (1987) Two physiologically distinct populations of neurons in the ventrolateral medulla innervate the locus coeruleus. Brain Res., 425: 275-282. Grove, E.A. (1988) Efferent connections of the substantia innominata in the rat. J. Cow. NewoL, 2773 347-364. Groves, P.M. and Wilson, C.J. (1980a) Monoaminergic presynaptic axons and dendrites in rat locus coeruleus seen in reconstructions of serial sections. J. Cow. Neurd, 193 853-862. Groves, P.M. and Wilson, C.J. (1980b) Fine structure of rat locus coeruleus. J. Cow. NeuroL, 193: 841-852. Grzanna, R. and Molliver, M.E. (1980) The locus coeruleus in the rat: An immunohistochemical delineation. Neuroscience, 5: 21-40. Guyunet, P.G. (1980) The coeruleospinal noradrenergic neurons: Anatomical and electrophysiological studies in the rat. Brain Res., 189 121-133. Hallanger, AE., Levey, A.I., Lee,HJ., Tye, D.B. and Wainer, B.H. (1987) The origins of cholinergic and other subcortical afferents to the thalamus in the rat. J. Comp. NeuroL, 262 105-124. Hallanger, A.E. and Wainer, B.H. (1988) Ascending projections from the pedunculopontine tegmental nucleus and the adjacent mesopontine tegmentum in the rat. J. Cow. NeuroL, 274 483-515. Holstege, G. (1987) Some anatomical observations on the projections from the hypothalamus to brainstem and spinal cord: An HRP and autoradiographic tracing study in the cat. J. Cow. NeuroL, 260: 98-126. Holstege G., Meiners, L. and Tan, K. (1985) Projections of the bed nucleus of the stria terminals to the mesencephalon, pons, and medulla oblongata in the cat. Ekp. Brain Res., 5 8 379-391. Iijima, K. (1989) An immunocytochemical study on the GABA-ergic and serotonin-ergic neurons in rat locus ceruleus with special reference to possible existence of the 8 7 43-57. masked indoleamine cells. Acta Hist&., Iijima, K. and Ohtomo, K. (1987) Immunocytochemicalstudy on the distribution of GABA-like immunoreactive neurons in rat locus ceruleus with karyometxy and synaptology. Acta H k t m h cyiochem,u):495-509. Jones, B.E. (1982) Relationship between catecholamine neurons and cerebral blood vessels studied by their simultaneous fluorescent revelation in the rat brainstem. Brain Res. Bull., 9 33-44. Jones, B.E. (1989) The relationship among acetylcholine, noradrenaline and GABA neurons within the pons of the rat. Anat. Rec., 223 57A. Jones, B.E. (199Oa) Immunohistochemical study of choline acetyl transferase-immunoreactive processes and cells innervating the pontomedullaryreticular formation. J. Cow. Neurol., 295: 485-514. Jones, B.E. (199Ob) Influence of the brainstem reticular formation, including intrinsic monoaminergic and cholinergic neurons, upon forebrain mechanisms of sleep and waking. In M. Mancia and G. Marini (Eds.), m e Diencephalon and Sleep, Raven Press, New York, pp. 31-48.
Jones, B.E. and Beaudet, A. (1987) Retrograde labeling of neurons in the brain stem following injections of [3Hl choline into the forebrain of the rat. @. Brain Res., 65: 437-448. Jones, B.E. and Cuello, A.C. (1989) Afferents to the basal forebrain cholinergic cell area from pontomesencephaliccatecholamiine, serotonin, and acetylcholine-neurons. Neuroscience, 31: 37-61. Jones, B.E. and Friedman, L. (1983) Atlas of catecholamine perikarya, varicosities and pathways in the brainstem of the cat. J. Cow. Newol., 215: 382-396. Jones, B.E. and Moore, R.Y. (1977) Ascending projections of the locus coeruleus in the rat. 11. Autoradiographic study. Brain Res., 127 23-53. Jones, B.E. and Yang, T.-Z. (1985) The efferent projections from the reticular formation and the locus coeruleus studied by anterograde and retrograde axonal transport in the rat. J. COW. Ne~roL,242 56-92. Uger, L. and Descarries, L. (1978) Serotonin nerve terminals in the locus coeruleus of adult rat: A radioautographic study. Brain Res., 145 1-13. Lkger, L., Wiklund, L., Descarries, L. and Persson, M. (1979) Description of an indolaminergiccell component in the cat locus coeruleus: A fluorescence histochemical and radiographic study. Brain Res., 168 43-56. Leontovich, T.A. and Zhukova, G.P. (1963) The specificity of the neuronal structure and topography of the reticular formation in the brain and spinal cord of carnivora. J. Cow.Nwo~.., 121: 347-379. Lindvall, 0. and Bjorklund, A. (1974) The organization of the ascending catecholamine neuron systems in the rat brain. Acta. PhySid. S c a d , Suppl. 412 1-48. Loughlin, S.E., Foote, S.L. and Bloom, F.E. (1986a) Efferent projections of nucleus locus coeruleus: Topographic organization of cells of origin demonstrated by three-dimensional reconstruction. Neuroscience, 1 8 291-306. Loughlin, S.E., Foote, S.L.and Grzanna, R. (1986b) Efferent projections of nucleus locus coeruleus: Morphologic subpopulations have different efferent targets. Neuroscience, 1 8 307-319. Maeda, T.and Shimizu, N. (1972) Projections ascendantes du locus coeruleus et d’autres neurones aminergiques pontiques au niveau du prosencephale du rat. Brain Res., 3 6 19-35. Martin, G.F., Humbertson, A.O.J., Laxson, L.C. and Panneton W.M. (1979) Dormlateral pontospinal systems. Possible routes for catecholamine modulation of nociception. Brain Res., 163: 333-338. Meesen, H. and Olszewski, J. (1949) A cytoarchitectonic Atlas of the Rhombencephalon of the Rabbit, S . Karger, Basel, p ~ 1-52. . Mehler, W.R., Feferman, M.E. and Nauta, W.J.H. (1960) Ascending axon degeneration following anterolateral cordotomy. An experimental study in the monkey. Brain, 83: 718-750. Milner, R.A., Abate, C., Reis, D.J; and Pickel, V.M. (1989) Ultrastructural localization of phenylethanolamine Nmethyltransferase-like immunoreactivity in the rat locus coeruleus. Brain Res., 478 1-15.
30 Moore, R.Y. (1980) The reticular formation: Monoamine neuron systems. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Revisited, Raven Press, New York, pp. 67-81. Mugnaini, I. and Oertel, W.H. (1985) An atlas of the distribution of GABAergic neurons and terminals in the rat C N S as revealed by GAD immunohistochemistry. In A Bjorklund and T. Hokfelt (Eds.), Handbook of Chemical Neuroanatomy. Vol. 4 GABA and Neuropeprides in the CNS, Part l., Elsevier Science Publishers, Amsterdam, pp. 436608. Nauta, W.J. and Haymaker, W. (1969) Hypothalamic nuclei and fiber connections. In W. Haymaker, E. Anderson and W.J.H. Nauta (Eds.), The Hypothalamus, C.C. Thomas, Springfield, Illinois, pp. 136-209. Nauta, W.J.H. and Kuypers, H.G.J.M. (1958) Some ascending pathways in the brain stem reticular formation. In H.H. Jasper, L.D. Proctor, R.S. Knighton, W.C. Noshay and R.T. Costello (Eds.), Reticular Formation of the Brain, Little, Brown & Co.,Boston, pp. 3-30. Olszewski, J. and Baxter, D. (1954) Cytoarchitecture of the Human Brain Stem, Karger, Basel, pp. 1-199. Paxinos, G. and Watson, C. (1986) The Rat Brain in Stereotaxic Coordinates,Academic Press, New York. Pickel, V.M., Joh, T.H. and Reis, D.J. (1977) A serotonergic innervation of noradrenergic neurons in nucleus locus coeruleus: Demonstration by immunocytochemical localization of the transmitter specific enzymes tyrosine and tyrptophan hydroxylase. Brain Res., 131: 197-214. Pieribone, V.A., Aston-Jones, G. and Bohn, M.C. (1988) Adrenergic and non-adrenergic neurons in the C1 and C3 areas project to locus coeruleus: A fluorescent .double labeling study. Neurosci. Lett., 85: 297-303. Ramon-Moliner, E. (1968) The Morphology of Dendrites. In G.H. Bourne (Eds.), The Structure and Function of the Nervous System, Vol. 1, Academic Press, New York, pp. 205-267. Ramon-Moliner, E. (1974) The locus coeruleus of cat. 111. Light and electron microscopic studies. Cell T h . Res., 149 205-221. Ramon-Moliner, E. and Nauta, W.J.H. (1966) The isodendritic core of the brain stem. J. Comp. Neurol., 126 311-336. Russell, G.V. (1955) The nucleus locus coeruleus (dorsolateralis tegmenti). Texas Rep. BWL Med, 13 939-988. Rye, D.B., Lee, H.J., Saper, C.B. and Wainer, B.H. (1988) Medullary and spinal efferents of the pedunculopontine tegmental nucleus and adjacent mesopontine tegmentum in the rat. J. Camp. NavoL,269 315-451. Sakai, K, Sastre, J.-P., Salvert, D., Touret, M., Tohyama, M. and Jouvet, M. (1979) Tegmentoreticular projections with special reference to the muscular atonia during paradoxical sleep in the cat: An HRP study. Brain Res., 176 233-254. Sakai, K., Touret, M., Salvert, D., Gger, L. and Jouvet, M. (1977) Afferent projections to the cat locus coeruleus as visualized by the horseradish peroxidase technique. Brain Res., 119 21-41.
Saper, C.B. (1984) Organization of cerebral afferent systems in the rat. I. Magnocellular basal nucleus. J. Comp. NeuroL, 222 313-342. Saper, C.B. (1985) Organization of cerebral cortical afferent systems in the rat 11. Hypothalamocortical projections. J. Comp.NeuroL, 237 21-46. Saper, C.B., Swanson, LS. and Cowan, W.M. (1979) An autoradiographic study of the efferent connections of the lateral hypothalamic area in the rat. J. Comp. NeuroL, 183: 689-706. Satoh, K and Fibiger, H.C. (1986) Cholinergic neurons of the laterodorsal tegmental nucleus: Efferent and afferent connections. J. Comp. Neurol..,253: 277-302. Seguela, P., Watkins, KC., Geffard, M. and Descarries, L. (1990) Noradrenaline axon terminals in adult rat neocortex: An immunocytochemical analysis in serial thin sections. Neuroscience, 35: 249-264. Shipley, M.T., Pieribone V.A., Aston-Jones, G. and Ennis, M. (1988) GABA-ergic innrevation of the rat locus coeruleus. SOC.Neurosci. Abst., 1 4 406. Simerly, R.B. and Swanson, L.W. (1988) Projections of the medial preoptic nucleus: A Phaseolus vulgaris leucoagglutinin anterograde tract-tracing study in the rat. J. Comp. NeuroL, 270 209-242. Swanson, L.W. (1976a) The locus coeruleus: A cytoarchitectonic, golgi and immunohistochemical study in the albino rat. Brain Res., 110: 39-56. Swanson, L.W. (1976b) An autoradiographic study of the efferent connections of the preoptic region in the rat. J. Comp. NeuroL, 167 227-256. Swanson, L.W., Connelly, M.A. and Hartman, B.K. (1977) Ultrastructural evidence for central monoaminergic innervation of blood vessels in the paraventricular nucleus of the hypothalamus. Brain Res., 136 166-173. Swanson, L.W., Connelly, MA. and Hartman, B.K. (1978) Further studies on the fine structure of the adrenergic innervation of the hypothalamus. Brain Res., 151: 165-174. Swanson, L.W. and Cowan, W.M. (1979) The connections of the septal region in the rat. J. Comp. NeuroL, 186 621-655. Swanson, L.W. and Hartman, B.K. (1975) The central adrenergic system. An immunofluorescence study of the location of cell bodies and their efferent connections in the rat utilizing dopamine-B-hydroxylase as a marker. J. Comp. Neurol., 163 467-506. Ungerstedt, U.(1971) Stereotaxic mapping of the monoamine pathways in the rat brain. Actu. Physiol. Scand., Suppl. 367 1-48. Vertes, R.P., Martin, G.F. and Waltzer, R. (1986) An autoradiographic analysis of ascending projections from the medullary reticular formation in the rat. Neuroscience., 19: 873-898. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1982) Descending noradrenergic projections and their spinal terminations. In H.G.J.M. Kuypers and G.F. Martin (Eds.), Descending Pathways to the Spinal Cord Progress in Brain Research, Vol. 57, Elsevier, Amsterdam, pp. 219-238.
C.D. Barnes and 0.Pompeiano (Eds.) R0gre.s~in Brain Research, Vol. 88 0 1991 Elsevier Science PublishersB.V.
31
CHAPTER 3
Physiological properties and afferent connections of the locus coeruleus and adjacent tegmental neurons involved in the generation of paradoxical sleep in the cat K. Sakai D&artement de Midecine fip‘rimentale, INSERM U 52, CNRS UA 1195, Universitte‘Claude Bernard Avenue R o c ~ e l l e r , Lyon,France
Results reported here confirm and extend those of early retrograde transport studies of the brainstem in the rat and cat. This study demonstrates substantial and multiple afferent projections to the cat locus coeruleus arising from neurons containing acetylcholine, serotonin, norepinephrine, epi-
nephrine, dopamine, histamine, and neuropeptides such as methionine, enkephaline and substance P. Further, our studies reveal notable differences in afferent projection to the noradrenergic and cholinergic regions of the locus coeruleus.
x;?y words: locus coeruleus, reticular formation, paradoxical sleep, transmitters, tegmentum
Introduction
More than two decades ago, Jouvet (1967, 1969) introduced a comprehensive “monoamine theory” of the regulation of sleep and waking in which he proposed that serotonin (5-HT)-containing neurons in the raphe nuclei play a determinant role in slow-wave sleep induction (SWS), whereas norepinephrine (NE)-containing neurons of the locus coeruleus (LC)located in the dorsolateral pons, play an essential role in the generation of paradoxical sleep (PS). This theory gave great impetus to basic and clinical sleep research, led to an extensive series of experiments and thereby contributed to a great extension of our knowledge concerning basic mechanisms of sleep (cf. Jouvet, 1972). The experimental results obtained during
the last two decades, however, do not always support the monoamine theory of sleep and this theory has been largely superseded: it is now widely accepted that NE-containing neurons of the LC do not play such a critical role in the generation of PS as was hitherto believed (for review see Steriade and Hobson, 1976; Ramm, 1979; Vertes, 1984; Sakai, 1985a). It has become also evident that the dorsal pontine tegmentum is not the only structure of the brain responsible for the induction and maintenance of this sleep state (Sakai, 1985a; Jones, 1985; Seigel, 1985; Hobson et aL, 1986). This does not imply, however, that the LC and adjacent tegmental neurons do not play an important part in the regulation of PS. On the contrary, evidence points to a large involvement of these neurons in the generation of
33
PS itself as well as in the generation of major tonic and phasic events occurring during this state of sleep (Sakai, 1985a,b; Webster and Jones, 1988). This review summarizes recent experimental evidence on the functional significance of these tegmental neurons in the regulation of PS and then presents our recent anatomical data concerning the afferent connections of the cat LC. First, however, it seems necessary to clarify the denomination of the dorsal pontine tegmental structures of the cat brain. Denomination of the LC and adjacent tegmental structures The cat LC is very heterogeneous and considerably less circumscribed than in the rat. The cat LC as determined cytoarchitectonically is composed of dorsal and ventral divisions (Fig. 1). The dorsal division is located, like the entire rat LC, in the periaqueductal grey matter (PAG) just medial to the mesencephalic root of the trigeminal nerve (mV) and lateral to the nucleus laterodorsalis tegmenti (Ldt) of Castaldi. This division of the LC is called “locus coeruleus proper” or briefly “LC” (Maeda et aZ., 1973; Sakai et aZ., 1977; Uger and Hernandez-Nicaise, 1980). The LC contains small to medium-sized green fluorescent cells (Maeda et aZ., 1973; Chu and Bloom, 1974; Jones and Moore, 1974). Like the rat LC neurons (Swanson and Hartman, 1975; Grzanna and Molliver, 19801, the cat LC neurons present both tyrosine hydroxylase (TH)and dopamine-Bhydroxylase (DpH) immunoreactivity [group A6 according to the nomenclature of Dahlstrom and Fuxe (196411. On the other hand, the adjacent Ldt contains many medium-sized to large choline acetyltransferase (ChAT)-immunoreactive or
cholinergic neurons and is thus clearly distinguishable from the LC proper (Sakai et at!, 1986; Jones and Beaudet, 1987; Vincent and Reiner, 1987). It should be mentioned, however, that several authors have included in the LC both the aminergic LC and cholinergic M t (e.g., Berman, 1968). Unlike the rat LC, the cat LC extends ventrally to the reticular formation and forms its ventral division. This division is located in the dorsolateral margin of the pontine reticular formation just ventral to the LC proper and medial to the dorsal half of the brachiwn conjunctivum (BC). This division is called “locus coeruleus pars alpha” or “locus coeruleus alpha (LCal” (Maeda et aZ., 1973; Sakai et aZ., 1977). The LCa cells are arranged either ventrolaterally (lateral part of the LCa) or ventromedially (medial part of the LCa). Although the LCa as a whole is characterized by the presence of medium-sized to large noradrenergic neurons (A6 cell group), the rostral half of the nucleus, in particular its medial part, also contains many medium-sized cholinergic and non-cholinergic neurons (Fig. 1). Ventrolateral to the LCa and medial to the nucleus parabrachialis medialis (PbM) is the nucleus locus subcoeruleus (LSC) composed of more loosely arranged small and medium-sized cells. The LSC also contains a large number of noradrenergic neurons. These neurons, together with neighboring noradrenergic neurons located in the nuclei PbM and parabrachialis lateralis and Kolliker-Fuse nucleus, constitute the A7 catecholamine cell group (Maeda et al., 1973; Chu and Bloom, 1974; Jones and Moore, 1974). The LC proper, LCa and LSC together are called “the LC complex” Just medial or medioventral to the LCa, a group of small to medium-sized cells can be seen. This corresponds to the nucleus peri-locus meruleus alpha
Fig. 1. Camera lucida drawings of coronal sections to illustrate the cytoarchitecture and histochemical nature of the rostral (A) and caudal (B)part of the cat locus coeruleus and adjacent tegmental structures. The sections are immunostained for both tyrosine hydroxylase (TH:red) and choline acetyltransferase (ChAT blue) using the two-colour double-immunostaining method. The sections are also counterstained with neutral red to demonstrate nontatecholaminergicand non-cholinergiccell bodies (green). Medial is to the left and dorsal is at the top. Abbreviations: a, locus coeruleus alpha; BC, brachium conjunctivum; LC, locus coeruleus proper; Ldt, nucleus laterodorsalis tegmenti; LSC, locus subcoeruleus; mV, mesencephalicroot of the trigeminal nerve; peri-a, peri-locus coeruleus alpha; TV,ventral nucleus of Gudden.
34
(peri-LCa; Sakai et al., 1979; Sakai, 1980). Noradrenergic neurons are very scarce in the periLCa. Instead the rostral two-thirds of the nucleus contains many medium-sized to large cholinergic neurons, while its caudal part contains a number of noncholinergic small to medium-sized neurons (Fig. 11, many of which show either corticotropin releasing factor (CRF)like or methionin-enkephalin (Met-Enkhlike immunoreactivity (Luppi et al., 1988). The most rostral and lateral part of the peri-LCa and LCa merges with the X area (Sakai, 1980) or nucleus tegmenti pedunculopontinus pars compacta (Nauta, 1979) that contains a large number of medium-sized to large cholinergic cell bodies. Many noradrenergic, cholinergic and other types of medium-sized neurons are interpenetrated therefore in the rostral region of the LCa. Functional role of dorsal pontine tegmental neurons in the generation of PS Recent lesion, stimulation and single-unit recording studies revealed that neurons in the peri-LCa and LCa, together with adjacent tegmental neurons, play a critical role in the induction and maintenance of PS as well as in the generation of major tonic and phasic phenomena occurring during this state of sleep, such as postural atonia, neocortical EEG desynchronization and ponto-geniculo-occipital(PGO) waves (Sakai, 1980, 1985a,b, 1988; Sakai et al., 1981, 199Ob; Vertes, 1984). Noradrenergic neurons have been shown to play a permissive role, whereas cholinergic and cholinoceptive neurons in the LCa and peri-LCa have been shown to play an executive role in these mechanisms. Indeed, noradrenergic neurons of the LC complex cease firing just before and throughout the period of PS (PS-off neurons) (McGinty and Sakai, 1973; Hobson et aZ., 1975; Sakaf, 1980, 1985a) and almost complete destruction of the LC complex and neighboring noradrenergic neurons does not result in the suppression of PS but facilitates the occurrence of PGO waves (Jones et aL, 1977; Sastre et al., 1979). In contrast, the peri-LCa and LCa
contain neurons showing tonic activation specific to PS (“PS-on” neurons; see below) and lesions of these nuclei lead to a permanent suppression of atonia (Sakai et aL, 1981) and to an almost complete suppression (Sastre et aL, 1979) or a marked reduction of PS (Sakai, 1985a; Webster and Jones, 1988). In addition, microinjections of a potent cholinergic agonist, carbachol, in the periLCa or adjacent LCa induce PS with short latencies and the effect is reversed completely by a local or systemic administration of atropine sulfate, a well-known muscarinic antagonist (Sakai, 1988; Vanni-Mercier et aL, 1989). Single-unit recording studies in freely moving cats revealed, in the LC complex and adjacent tegmental structures, the existence of several different populations of neurons that are involved critically in the mechanisms underlying the generation of PS, Le., “PS-off,” “PS-on,” “PGO-on” and “ascending tonic” neurons. Firing patterns of these types of neurons during the sleep-waking cycle are illustrated schematically in Figure 2. The localization of these neurons is shown in Figure 3. The noradrenergic nature of PS-off neurons recorded in the cat LC complex has been well documented (see Jacobs, 1986 for review). As shown in Figure 3, these neurons are found throughout the rostrocaudal extent of the LC complex and in the neighboring parabrachial nuclei containing noradrenergic cell bodies. The majority of PS-off neurons located in the LC
Tonic
W
sws
PS
type I
m
type 11
-il-H+t
PS-ON PS-OFF PGO-ON-
-i
w
t
Fig.2. Schematic illustration of discharge patterns during the sleep-waking cycle of neurons commonly found in the dorsal pontine tegmentum. Abbreviations: W, waking, SWS, slowwave sleep; PS, paradoxical sleep.
35
proper is invaded antidromically by the stimulation of the dorsal ascending noradrenergic bundle (Sakai, 1980, 1985a). Recent anatomical studies showed that noradrenergic descending projections to the spinal cord originate primarily in the caudal part of the LCa and in the A7 cell group (Stevens et al., 1982; Sakai, 1985b; Reddy et al., 1989).
In contrast to the noradrenergic PS-off neurons, PGO-on neurons being regarded as PGO executive neurons are located exclusively in the rostra1 part of the LCa and adjacent tegmental structures, both containing cholinergic cell bodies. The cholinergic nature of PGO-on neurons has also been well documented (Sakai, 1985b, Sakai et al., 1990b).
Fig. 3. Localizations of different populations of neurons (PS-OFF, PS-ON, Tonic Type I, Tonic type I1 and PGO-ON) found in the cat locus coeruleus and adjacent tegmental structures. Large dots in “PS-OFF” indicate PS-off neurons identified antidromically by the stimulation of the dorsal ascending norepinephrine bundle. Triangles pointing upward and downward in “PS-ON” indicate ascending and descending PS-on neurons, respectively.Tonically discharging ascending neurons (Tonic Type I) are subdivided into Type I-S (closed circles) and Type I-R (open circles). Note a marked functional heterogeneity existing within the cat locus meruleus. See text for details. Abbreviations: CNF, nucleus cuneiformis; PbM and PbL, nuclei parabrachialis medialis and lateralis, respectively; X, area X or nucleus tegmenti pedunculopontinus pars mmpacta. For other abbreviations, see Figure 1.
36
Like PGO-on neurons, tonically discharging neurons that send axons directly to the medial and intralaminar thalamic nuclei and/or posterior hypothalamus are found in the rostral part of the LCa and adjacent tewental structures containing cholinergic neurons. The tonic ascending neurons are subdivided into three groups: tonic type I slow (Type 1-9, type I rapid (Type I-R) and type 11. Both type I-S and type I-R neurons exhibit a tonic pattern and a high rate of discharge during both waking (W) and PS as compared with SWS. They are distinguishable because of their low and high rates of spontaneous discharge. Type I1 neurons are characterized by a tonic pattern of discharge highly specific to the periods of PS. These ascending tonic neurons are thought to play an important role in the induction of neocortical EEG desynchronization occurring during W and PS or only during PS (El Mansari et al., 1989; Sakai et al., 1990a; Steriade et al., 1990). The possible cholinergic nature of these ascending tonic neurons, in particular type I-S, has been described (El Mansari et al., 1990). Neurons showing no spontaneous discharge during W, but exhibiting a tonic discharge just prior to and throughout the period of PS (PS-on neurons), are also localized histologically in the rostral part of the LCa and adjacent peri-LCa. Many of them send descending axons to the nucleus reticularis magnocellularis (Mc) of the ventromedial medulla corresponding to the medullary inhibitory center first described by Magoun and Rhines (1946). These descending pontine PS-on neurons are thought to be an essential part of the brainstem circuitry necessary for the generation of muscular atonia during PS (Sakai, 1980, 1981). PS-on neurons as a whole are cholinoceptive and involved critically in the executive mechanisms of PS (Sakai, 1988). These findings, when taken together, indicate that the cat LC is functionally heterogeneous and can be divided into at least three regions: (1) the LC proper; (2) the caudal part of the LCa; and (3) the mediorostral part of the LCa. In the following part of this chapter, I will summarize
the afferent projections to these monoaminergic or cholinergic regions of the LC.
Afferent projections to the cat locus coeruleus Early anatomical studies using retrograde tracer techniques in the cat and rat reported that the afferent projections to the LC are quite diverse (Sakai et al., 1977; Cedarbaum and Aghajanian, 1978; Clavier, 1979). Early anterograde tracing studies confirmed many of these afferent projections (see, for the cat brain, Bobillier et al., 1976; Hopkins and Holstege, 1978; Loewy and Saper, 1978; King, 1980; Holstege et al., 1985). Also in line with these observations, early electrophysiological studies demonstrated that the LC receives a wide variety of excitatory and inhibitory inputs (Nakamura, 1977; Takigawa and Mogenson, 1977; Aston-Jones and Bloom, 1981). Furthermore, immunohistochemical studies revealed the existence of many neurotransmitters or neuromodulators in terminals within the LC of rats and cats. The terminals contain many peptides such as substance P (SP), Met-Enk and leu-enkephalin, dynorphin, adrenal corticotrophic hormone, CRF, and a-melanocyte stimulating hormone (Hokfelt et al., 1978; Pickel et al., 1979; Watson et al., 1982; Uger et al., 1983; Rao et al., 1987). The LC contains a high concentration of acetylcholinesterase, degradative enzyme of acetylcholine (ACh) (Albanese and Butcher, 1979; Sakai, 1980) and also contains ChAT-immunoreactive terminals (Kimura et al., 1981). The dopamine synthetic enzyme TH, as well as DPH, a synthetic enzyme unique to NE and epinephrine cells, are also present in terminals (Pickel et al., 1977; ,Grzanna and Molliver, 1980). In addition, phenylethanolamine N-methyltransferase (PNMT), the final synthetic enzyme for NE, is found in terminals within the LC (Hokfelt et al., 1974). Serotonin nerve terminals have, likewise been demonstrated in the LC (Pickel et al., 1977; Uger and Descarries, 1978; Groves and Wilson, 1980). Histamine (HA)-immunoreactive fibers have recently been revealed in the LC of the rat
31
(Panula et al., 19891, guinea pig (Airaksinen and Panula, 1988) and cat (our unpublished data). On the other hand, it has been shown that iontophoretic applications of ACh and muscarinic agonists, glutamate, SP and related peptides, all excite neurons of the rat LC. In contrast, noradrenaline, a,-adrenergic agonist clonidine, GABA, morphine, enkephalin, neurotensin and 5-HT have been reported to depress the firing of LC neurons (Korf et al., 1974; Cedarbaum and Aghajanian, 1976; Bird and Kuhar, 1977; Guyenet and Aghajanian, 1977, 1979; Young et al., 1978; Adams and Foote, 1988; Chouvet et al., 1988). These observations are confirmed in part in the cat using the microinjection method in the vicinity of the LC neurons (Abercrombie and Jacobs, 1987; Abercrombie et al., 1988). In the cat we have learned recently that carbachol microinjections into the peri-LCa or adjacent LCa result in the excitation of pontine “cholinoceptive” PS-on neurons and in the suppression of firing of noradrenergic PS-off neurons: the direct excitatory action of carbachol on noradrenergic PS-off neurons located in the L C a or adjacent peri-LCa is totally overcome when the carbachol is microinjected into the same structures thereby inducing PS with short latencies (Sakai, 1988). These physiological findings also point to the existence of multiple afferent inputs to the LC and indicate that, in order to elucidate the mechanisms underlying the generation of PS, a prerequisite is to determine afferent projections to the noradrenergic or cholinergic regions of the LC and to identify their neurotransmitter characteristics. In the present study, I will describe mainly afferent projections to the cat LC arising from cholinergic, monoaminergic and some peptidergic cell bodies that appear to play an important role in the mechanisms underlying the induction and maintenance of PS. Particular attention will be also paid to afferent projections from the medulla, since Aston-Jones et al. (1986) have reported recently that major inputs to the rat LC emanate from only two nuclei located in the rostra1
medulla: the nuclei praepositus hypoglossi (PHI and paragigantocellularis lateralis (PGCL). Methods
In eighteen adult cats, unconjugated cholera toxin (CT) B subunit (0.1 p1 of a 1% solution) was injected hydraulically into the LC and adjacent dorsal pontine tegmental structures. For detection of neurotransmitter contents of retrogradely labeled neurons, a two-colour double immunostaining method was used in colchicine-treated animals: retrograde transport of CT was revealed by a biotin/streptavidin procedure with DABnickel as chromogen and the neurotransmitter contents of CT-labeled neurons were revealed by Sternberger’s PAP procedure using DAB as chromogen. Advantages and detailed procedures of the method have been described elsewhere (Luppi et al., 1987; Yoshimoto et al., 1989). Primary antibodies used are as follows: (1) TH (Institute Jacques Boy); (2) D P H (Eugene Tech International); (3) PNMT (Eugene Tech International); (4) 5-HT (a gift from Dr. H. Kimura, Department of Anatomy, University of Shiga, Japan); (5) HA (a gift from Dr. P. Panula, Department of Anatomy, University of Helsinki, Finland); (6) ChAT (Boehringer Mannheim); (7) SP (Biogenex); and (8) Met-Enk (UCB Bioproducts). Results
In this report three representative cases will be described: in one case (Y109), the CT injection site was centered on the peri-LCa and the rostromedial part of the L C a corresponding to the most cholinergic portion of the LC. In cases K 113 and Y 114, the injection loci were centered, respectively, on the caudolateral part of the L C a and on the LC proper; both contain many noradrenergic cell bodies but are virtually devoid of cholinergic cell bodies (Fig. 4). Using a very sensitive retrograde tracer technique with unconjugated CT, we confirmed and
38
extended our early anatomical study in t h e cat (Sakai et al., 1977). Briefly, major afferent projections to the cat LC in general originated in: (1) preoptic region of the anterior hypothalamus; (2) nucleus of the stria terminalis; ( 3 ) central nucleus of the amygdala; (4) posterior hypothalamus; ( 5 ) PAG; (6) nuclei raphe dorsalis (RD) and raphe
Y 109
K 113
magnus (RM); (7) dorsolateral pontine tegmentum; and (8) nucleus reticularis parvocellularis (Pc) of the rostrolateral medulla. A substantial number of retrogradely labeled neurons was also present in the mesencephalic and pontine reticular formation when the injection site was centered on the rostromedial or caudal part of the
Y 114
Fig. 4. Localizations of cholera toxin (CT)-positive and choline acetyltransferase (ChAT; open circles)-, tyrosine hydroxylase (TH; closed circles)- or serotonin (5-HT; asterisks)-immunoreactive double-labeled neurons observed in three representative cases (Y109, K113, and Y114). Blackened areas show the central zone of the injection loci. Each symbol indicates one double-labeled neuron observed on one representative section. Abbreviations: CS, nucleus raphe centralis superior; DBC, decussation of the brachium conjunctivum; K-F, Kolliker-Fuse nucleus; MV, motor trigeminal nucleus; Poo, nucleus reticularis pontis oralis; TD, dorsal nucleus of Gudden. For other abbreviations, see Figures 1 and 2.
39
LCa. The nucleus of the solitary tract contained numerous CT-labeled neurons exclusively when the tracer was injected into the noradrenergic regions of the LC (Fig. 5). The nucleus reticularis magnocellularis (Mc; or nucleus reticularis gigantocellularis pars a in the rat brain) of the medulla also contained many CT-positive neurons after
injections of the tracer in the medial part of the L C a (Fig. 5). Although we observed consistently retrogradely labeled neurons in the nucleus reticularis PGCL, this nucleus did not seem to constitute a major afferent source in the cat brain (Fig. 5). Similarly, the nucleus PH did not appear to be a major afferent to the cat LC, since we observed
Fig. 5. A series of camera lucida drawings of 25 p m sections from rostra1 to caudal order to illustrate the distribution of retrogradely labeled neurons in the medulla after injections of the tracer in three different regions of the cat locus coeruleus. Each dot represents one retrogradely labeled neurons. Abbreviations: 7, facial nucleus; 12, hypoglossal nucleus; 12N, hypoglossal nerve; A, nucleus ambiguus; AP, area postrema, 5SP, alaminar spinal trigeminal nucleus, parvocellular division; Gc, nucleus reticularis gigantocellularis; IC, inferior olivary complex; LR, lateral reticular nucleus; Mc, nucleus reticularis magnocellularis; NTS, nucleus of the solitary tract; MLB, medial longitudinal bundle; Ivs, lateral vestibulospinal tract; P, pyramidal tract; Pc, nucleus reticularis parvocellularis; PGCL, nucleus paragigantocellularis lateralis; PH, nucleus praepositus hypoglossi; PR, paramedian reticular nucleus; RM, nucleus raphe magnus; RPa, nucleus raphe pallidus; rs, rubrospinal tract; VM, medial vestibular nucleus.
40
a significant number of CT-labeled neurons in and just around the nucleus only after CT injections in the rostromedial LCa and adjacent periLCa (Fig. 5). The following is a summary of results obtained from our double-immunostaining study.
Serotoninergic afferents Serotoninergic afferents to the cat LC have been found to originate almost exclusively in the rostra1 part of the nucleus R D [B7 cell group according to Dahlstrom and Fuxe (1964)l: minor afferents originated in the nuclei raphe intermedius and centralis superior (CS; B8 cell group; Fig. 4). No double-labeled cells were noted in other raphe nuclei such as raphe pontis (B5),
magnus (B3) and pallidus (Bl) nor in the ventrolateral bulbar reticular formation or PGCL.
Histaminergic afferents Both the LC proper and the rostromedial part of the L C a were found to receive numerous histaminergic afferent projections deriving exclusively from the tuberomamillary nucleus and neighboring lateral hypothalamic area (Fig. 6). Dopaminergic afferents TH-positive but DPH-negative, presumably dopaminergic inputs to the LC have been found to arise mainly from the ipsilateral A l l dopamine cell group located in and around the posterior
*TH oHA Fig. 6 . Localizations of cholera toxin (CT)-positive and tyrosine hydroxylase (TH)- or histamine (HA)-immunoreactive double-labeled neurons in the posterior hypothalamus after injections of the tracer in three different regions of the cat locus coeruleus. Each symbol indicates one double-labeled neuron. Abbreviations: DA, dorsal hypothalamic nucleus; f, fornix; FF, nucleus of the fields of Forel; HLA, lateral hypothalamic area; HPA, posterior hypothalamic area; In, infundibular nucleus; MM, medial mamillary nucleus; MS, supramamillary nucleus; MT, mamillothalamic tract; PP, pes pedunculi; TM, tuberomamillary nucleus.
41
hypothalamic area. Minor dopaminergic afferents originated in the A13 cell group. The rostromedial region of the LC appeared to receive the densest dopaminergic projections (Fig. 6).
jections were mostly ipsilateral and found mainly when the tracer was injected into the noradrenergic regions of the LC (Figs. 4 and 7).
Noradrenergic afferents TH- and DpH-positive, but PNMT-negative, noradrenergic cell bodies sending axons to the cat LC were found both in the pontine (A4, A5, A6 and A7) and bulbar (A1 and A2) catecholaminergic ceIl groups. These noradrenergic afferent pro-
Adrenergic afferents As shown in Figure 7, CT-positive and PNMT-immunoreactive double-labeled neurons were observed in the lateral medulla (C1 cell group) particularly after C T injections in the noradrenergic regions of the LC.
OChAT * T H
PNMT
Fig. 7. Localizations of cholera toxin (CT)-positive and choline acetyltransferase (ChAT; open circles)-, tyrosine hydroxylase (TH; closed circles)- or phenylethanolamine N-methyltransferase (PNMT; asterisks)-immunoreactive double-labeled neurons noted after injections of the tracer in three different regions of the locus coeruleus. The injection sites are indicated by blackened areas. Each symbol indicates one double-labeled neuron observed on one representative section. For abbreviations, see Figure 5.
42
Cholinergic afferents In sharp contrast to the catecholaminergic inputs, the cholinergic region of the LC has been found to receive more numerous cholinergic inputs than the noradrenergic regions of the nucleus. Indeed, after CT injections in the rostromedial part of the L C a and adjacent peri-LCa, CT- and ChAT-immunoreactive double-labeled neurons were observed bilaterally, but with an ipsilateral predominance, in peribrachial regions of the dorsal pontine tegmentum. The marginal nucleus of the BC at P4 plane especially contained many small to medium-sized doublelabeled neurons (Fig. 4). In the medulla, doublelabeled neurons were scattered in the nuclei reticularis Pc, Mc and neighboring gigantocellulark (Gc), as well as in the PGCL. A few doublelabeled neurons were also found in and just around the PH (Fig. 7). These double-labeled neurons in the medulla were medium-sized to large and their distribution was very similar to that of medullary PS-on neurons (Sakai, 1988). After CT injections in the LC proper or in the caudolateral part of the LCa, a few doublelabeled “small” neurons were observed only in the Pc (Fig. 7). Peptidergic afferents On the whole, double-labeled neurons with SP or with Met-Enk were similarly localized in three representative cases. Most of the CT- and SP-immunoreactive double-labeled neurons were observed in: (1) the perifornical region of the posterior hypothalamus; (2) the Edinger-Westphal nucleus; and (3) the mesencephalic and pontine PAG. A few double-labeled neurons were also noted in a lemniscal region near the LSC and in the dorsal raphe nucleus. On the other hand, CTand Met-Enk-immunoreactive double-labeled neurons were found mainly in: (1) the perifornical region of the posterior hypothalamus; (2) the PAG; ( 3 ) the RD; and (4) the dorsolateral pontine tegmentum. There are very few ascending peptidergic projections from neurons in the PGCL.
Conclusions
The results we report here confirmed and extended those of early retrograde transport studies in the rat and cat. The present study demonstrates substantial and multiple afferent projections to the cat LC arising from neurons containing ACh, S-HT, NE, epinephrine, dopamine, HA and neuropeptides such as Met-Enk and SP. The present results also reveal notable differences in afferent projections to the noradrenergic and cholinergic regions of the LC: The noradrenergic regions of the LC playing a permissive role in the generation of PS receive dense serotonergic, noradrenergic and adrenergic inputs, whereas they receive only a few cholinergic afferents from the pons and medulla. The cholinergic region of the LC being involved in the executive mechanisms of PS receives dense serotonergic, dopaminergic and cholinergic inputs, but received relatively few noradrenergic and adrenergic afferents. The findings in regard to the cholinergic afferents may provide the anatomical basis for the mechanisms underlying cholinergic induction of PS. It seems likely that the cells of origin of cholinergic afferents to the cholinoceptive pontine PS-on cells originate primarily in the bulbar reticular formation, since the location of the ascending bulbar cholinergic neurons is very similar to that of medullary PS-on cells and, further, some of these medullary PS-on cells are invaded antidromically by the stimulation of the rostromedial part of the L C a and adjacent peri-LCa (Sakai, 1988). On the other hand, the present observations do not seem to support directly the reciprocal-interaction hypothesis according to which PS is generated as a result of direct reciprocal interactions between noradrenergic PS-off and cholinergic PS-on cells (Hobson et nl., 1975, 1986; see also Sakai, 1988). It may be that an exactly inverse relationship in unitary activity observed between pontine PS-on and PS-off neurons throughout the sleep-waking cycle (Sakai, 1985a, 1988) is mediated by some inhibitory interneurons such as those containing GABA: this amino acid has been re-
43
ported to exert a powerful inhibitory influence on the rat LC neurons (Guyenet and Aghajanian, 1979). A powerful excitatory influence of glutamate on the LC neurons has also been demonstrated (Guyenet and Aghajanian, 1979; Ennis and Aston-Jones, 1988). The functional roles of these amino acids, as well as HA, dopamine and several neuropeptides, in the permissive and executive mechanisms of PS should be determined in future studies. Furthermore, in future anatomical studies, we must define the exact cells of origin of the excitatory and inhibitory amino acid afferents to the cat LC and determine the precise input-output organization of PS-on and PS-off neurons located in the LC and adjacent pontine tegmentum. Acknowledgements
This work was supported by INSERM U 52, CNRS UA 1195, and DRET (Grant 87/215). References Abercrombie, E.D. and Jacobs, B.L. (1987) Microinjected clonidine inhibits noradrenergic neurons of the locus coeruleus in freely moving cats. Neurosci. Lett., 76: 203208. Abercrombie, E.D., Levine, E.S. and Jacobs, B.L. (1988) Microinjected morphine suppresses the activity of locus coeruleus noradrenergic neurons in freely moving cats. Neurosci. Lett., 86: 334-339. Adams, L.M. and Foote, S.L. (1988) Effects of locally infused pharmacological agents on spontaneous and sensoryevoked activity of locus coeruleus neurons. Brain Res. Bull., 21: 395-400. Airaksinen, M.S. and Panula, P. (1988). The histaminergic system in the guinea pig central nervous system: A n immunocytochemical mapping study using an antiserum against histamine. J. Comp. Neurol., 273: 163-186. Albanese, A. and Butcher, L.L. (1979) Locus coeruleus somata contain both acetylcholinesterase and norepinephrine: Direct histochemical demonstration of the same tissue section. Neurosci. Lett., 14: 101-104. Aston-Jones, G. and Bloom, F.E. (1981) Norepinephrine-containing locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Aston-Jones, G., Ennis, M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986) The brain nucleus locus coeruleus: Restricted afferent control of a broad efferent network. Science, 234: 734-737.
Berman, A.L. (1968) The Brain Stem of the Cat. A Cytoarchitectonic Atlas with Stereotaxic Coordinates. University of Wisconsin Press, Madison, Wisconsin. Bird, S.J. and Kuhar, M.J. (1977) Iontophoretic application of opiates to the locus coeruleus. Brain Res., 122: 523-533. Bobillier, P., Seguin, S., Petitjean, F., Salvert, D., Touret, M. and Jouvet, M. (1976) The raphe nuclei of the cat brain stem: A topographical atlas of their efferent projections as revealed by autoradiography. Brain Res., 113: 449-486. Cedarbaum, J.M. and Aghajanian, G.K. (1976) Noradrenergic neurons of the locus coeruleus: Inhibition by epinephrine and activation by the a-antagonist piperoxane. Brain Res., 112: 413-419. Cedarbaum, J.M. and Aghajanian, G.K. (1978) Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique. J. Comp. Neurol., 178: 1-16. Chouvet, G., Akaoka, H. and Aston-Jones, G. (1988) Serotonin selectively decreases glutamate-induced excitation of locus coeruleus neurons. C.R. Acad. Sci. Paris, 306: 339344. Chu, N.S. and Bloom F.E. (1974) The catecholamine-containing neurons in the cat dorsolateral pontine tegmentum: Distribution of the cell bodies and some axonal projections. Brain Res., 66: 1-21. Clavier, R. (1979) Afferent projections to the self-stimulation regions of the dorsal pons, including the LC in the rat as demonstrated by HRP technique. Brain Res. Bull., 4: 497504. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Scand., 62, Suppl. 232: 1-55. El Mansari, M., Sakai, K. and Jouvet, M. (1989) Unitary characteristics of presumptive cholinergic tegmental neurons during sleep-waking cycle in freely moving cats. Exp. Brain Res., 76: 519-529. El Mansari, M., Sakai, K. and Jouvet, M. (1990) Responses of presumed cholinergic mesopontine tegmental neurons to carbachol microinjections in freely moving cats. Exp. Brain Res., 83: 115-123. Ennis, M. and Aston-Jones, G. (1988) Activation of locus coeruleus from nucleus paragigantocellularis: A new excitatory amino acid pathway in brain. J. Neurosci., 8: 36443657. Groves, P.M. and Wilson, C.J. (1980) Monoaminergic presynaptic axons and dendrites in rat locus coeruleus seen in reconstructions of serial section. J. Comp. Neurol., 193: 853-862. Grzanna, R. and Molliver, M.E. (1980) Cytoarchitecture and dendritic morphology of central noradrenergic neurons. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Recisited, Raven Press, New York, pp. 83-97. Guyenet, P.G. and Aghajanian, G.K. (1977) Excitation of neurons in the nucleus locus coeruleus by substance P and related peptides. Brain Res., 136: 178-184. Guyenet, P.G. and Aghajanian, G.K. (1979) ACh, substance P and met-enkephalin in the locus coeruleus: Pharmacologi-
44 cal evidence for independent sites of action. Eur. J. Phurmucol., 53: 319-328. Hobson, J.A., McCarley, R.M. and Wyzinski, P.W. (1975) Sleep cycle oscillation: Reciprocal discharge by two brainstem groups. Science, 189: 55-58. Hobson, J.A., Lydic, R. and Baghdoyan, H.A. (1986) Evolving concepts of sleep cycle generation: From brain centers to neuronal populations. Behuc. Bruin Sci., 9: 371-448. Hokfelt, T., Fuxe, K., Goldstein, M. and Johansson, 0. (1974) Immunohistochemical evidence for the existence of adrenaline neurons in the rat brain. Bruin Res., 66: 235251. Hokfelt, T., Elde, R., Fuxe, K., Johansson, O., Ljundahl, A,, Goldstein, M., Luft, R., Efendic, E., Nilsson, G., Terenius, L., Gauten, D., Jeffcoate, S.L., Rehfeld, J., Said, S., Perez de la Mora, M., Possani, L., Tapia, R., Teran, L. and Palacios, R. (1978) Aminergic and peptidergic pathways in the nervous system with special reference to the hypothalamus. In S. Reichlin, R. Baldessarini and J. Martin (Eds.), The Hypothalamus, Raven Press, New York, pp. 69-135. Holstege, G., Meiners, L. and Tan, L. (1985) Projections of the bed nucleus of the stria terminalis to the mesencephalon, pons, and medulla oblongata in the cat. Exp. Bruin Res., 58: 397-391. Hopkins, D.A. and Holstege, G. (1978) Amygdaloid projections to the mesencephalon, pons and medulla oblongata in the cat. Exp. Bruin Res., 32: 529-547. Jacobs, B.L. (1986) Single unit activity of locus coeruleus neurons in behaving animals. Prog. Neurobiol., 27: 183-194. Jones, B.E. (1985) Neuroanatomical and neurochemical substrates of mechanisms underlying paradoxical sleep. In D.J., Mcginty, R. Drucker-Colin, A. Morrison and L. Parrneggiani (Eds.), Bruin Mechanisms of Sleep, Raven Press, New York, pp. 139-156. Jones, B.E. and Beaudet, A. (1987) Distribution of acetylcholine and catecholamine neurons in the cat brainstem: A choline acetyltransferase and tyrosine hydroxylase immunohistochemical study. J. Comp. Neurol., 261: 15-32. Jones, B.E. and Moore, R.Y. (1974) Catecholamine-containing neurons of the nucleus locus coeruleus in the cat. J. Comp. Neurol., 157: 43-52. Jones, B.E., Harper, S.T. and Haralis, A.E. (1977) Effects of locus coeruleus lesions upon cerebral monoamine content, sleep-wakefulness states and the response to amphetamine in the cat. Bruin Res., 124: 473-496. Jouvet, M. (1967) Mechanisms of states of sleep: A neuropharmacological approach. Res. Publ. Assoc. Nerc. Ment. Dis.,45: 86-126. Jouvet, M. (1969) Biogenic amines and the states of sleep. Science, 163: 32-41. Jouvet, M. (1972) The role of monoamines and acetylcholinecontaining neurons in the regulation of the sleep-waking cycle. Ergeb. Physiol., 64: 166-307. Kimura, H., McGeer, P.L., Peng, J.H. and McGeer, E.G. (1981) The central cholinergic system: Studies by choline acetyltransferase immunohistochemistry in the cat. J. Comp. Neurol., 200: 151-201.
King, G.W. (1980) Topology of ascending brain stem projections to nucleus parabrachialis in the cats. J. Comp. Neurol,, 191: 615-638. Korf, J., Bunny, B.S. and Aghajanian, G.K. (1974) Noradrenergic neurons: Morphine inhibition of spontaneous activity. Eur. J. Phurmucol., 25: 165-169. LCger, L. and Descarries, L., (1 978) Serotonin nerve terminals in the locus coeruleus of adult rat: Radioautographic study. Brain Res., 145: 1-13. Ltger, L. and Hernandez-Nicaise, M.L. (1980) The cat locus coeruleus. Light and electron microscopic study of the neuronal somata. Anut. Embryol., 159: 181-198. &er, L., Charnay, Y., Chayvialle, J.A., Btrod, A., Dray, F., Pujol, J.F., Jouvet, M. and Dubois, P.M. (1983) Localization of substance P- and enkephaline-like immunoreactivity in relation to catecholamine-containing cell bodies in the cat dorsolateral pontine tegmentum: An immunofluorescence study. Neuroscience, 3: 525-546. ,oewy, A.D. and Saper, C.B. (1978) Edinger-Westphal nucleus: Projections to the brain stem and spinal cord in the cat. Bruin Res., 150: 1-27. Luppi, P.H., Sakai, K., Salvert, D., Fort, P. and Jouvet, M. (1987) Peptidergic hypothalamic afferents to the cat nucleus raphe pallidus as revealed by a double immunostaining technique. Bruin Res., 402: 339-345. Luppi, P.H., Sakai, K., Fort, p., Salvert, D. and Jouvet, M. (1988) The nuclei of origin of monoaminergic, peptidergic, and cholinergic afferents to the cat nucleus reticularis magnocellularis: A double-labeling study with cholera toxin as a retrograde tracer. J. Comp. Neurol., 277: 1-20. Maeda, T., Pin, C., Salvert, D., Ligier, M. et Jouvet, M. (1973) Les neurones contenant des catecholamines du tegmenturn pontjque et leurs voies de projections chez le chat. Bruin Res., 57: 119-152. Magoun, H.W. and Rhines, R. (1946) An inhibitory mechanisms in the bulbar reticular formation. J. Neurophysiol., 9: 165- 171. McGinty, D.J. and Sakai, K. (1973) Unit activity of the dorsal pontine reticular formation in the cat. Sleep R e x , 2: 33. Nakamura, S. (1977) Some electrophysiological properties of neurons of rat locus coeruleus. J. Physiol. (London), 267: 641-658. Nauta, H.J.W. (1979) Projections of the pallidal complex, An autoradiographic study in the cat. Neuroscience, 4: 18531873. Panula, P., Pirvola, U., Auvinen, S. and Airaksinen, M.S. (1989) Histamine-immunoreactive nerve fibers in the rat brain. Neuroscience, 28: 585-610. Pickel, V.M., Joh, T.H. and Reis, D.J. (1977) A serotonergic innervation of noradrenergic neurons in nucleus locus coeruleus: Demonstration by immunocytochemical localization of the transmitter specific enzymes tyrosine and tryptophan hydroxylase. Brain Res., 131: 197-214. Pickel, V.M., Joh, T.H., Reis, D.J., Leeman, S.E. and Miller, R.J. (1979) Electron microscopic localization of substance P and enkephalin in axon terminals related to dendrites of catecholaminergic neurons. Bruin Res., 160: 387-400.
45 Ramm. P. (1979) The locus coeruleus, catecholamines, and REM sleep: A critical review. Behat$. Neural. Biol., 25: 4 15-448. Rao, J.K., Hu, H., Prasa, C. and Jayaraman, A. (1987) The distribution pattern of alpha-MSH-like immunoreactivity in the cat central nervous system. Peptides, 8: 327-334. Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1989) Spinally projecting noradrenergic neurons of the dorsolatera1 pontine tegmentum: A combined immunocytochemical and retrograde labeling study. Brain Rex, 491: 144-149. Sakai, K. (1980) Some anatomical and physiological properties of ponto-mesencaphalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep in the cat. In J.A., Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Rersisited, Raven Press, New York, pp. 427-447. Sakai, K. (1985a) Anatomical and physiological basis of paradoxical sleep. In D.J. McGinty, R. Drucker-Colin, A. Morrison and L. Parmeggiani (Eds.), Brain Mechanisms of Sleep, Raven Press, New York, pp. 111-137. Sakai, K. (198Sb) Neurons responsible for paradoxical sleep. In A. Wauquier, J.M. Monti, J.M. Gaillard and M. Radulovacki (Eds.), Sleep: Neurotransmitters and Neuromodulators, Raven Press, New York, pp. 29-42. Sakai, K. (1988) Executive mechanisms of paradoxical sleep. Arch. Ital. Biol., 479: 225-240. Sakai, K., Touret, M., Salvert, M., Ltger, L. and Jouvet, M. (1977) Afferent projections to t h e cat locus coeruleus as visualized by the horseradish peroxidase technique. Brain Rex, i19: 21-41. Sakai, K., Sastre, J.P., Salvert, D., Touret, M., Tohyama, M. and Jouvet, M. (1979) Tegmentoreticular projections with special reference to the muscular atonia during paradoxical sleep in the cat: An HRP study. Brain Res., 176: 233-254. Sakai, K., Sastre, J.P., Kanamori, N. and Jouvet, M. (1981) State-specific neurons in the ponto-medullary reticular formation with special reference to the postural atonia during paradoxical sleep in the cat. In C. Ajmone-Marsan and 0. Pompeiano (Eds.), Brain Mechanisms of Perceptual Awareness and Purposeful Behacior, pp. 405-429. Sakai, K., Luppi, P.H., Salvert, D., Kimura, H., Maeda, T. and Jouvet, M. (1986) Localization of cholinergic neurons in the cat lower brain stem. C.R. Acad. Sci. Paris, 303: 317324. Sakai, K., El Mansari, M., Lin, J.S., Zhang, J.G. and VanniMercier, G . (1990a) The posterior hypothalamus in the regulation of wakefulness and paradoxical sleep. In M. Mancia and G. Marini, (Eds.), The Diencephalon and Sleep, Raven Press, New York, pp. 171-198. Sakai, K., El Mansari, M. and Jouvet, M. (1990b) Inhibition by carbachol microinjections of presumptive cholinergic
PGO-on neurons in freely moving cats. Brain Res., 527: 213-223. Sastre, J.P., Sakai, K. and Jouvet, M. (1979) Persistence du sommeil paradoxal chez le chat aprts destruction de I’aire gigantocellulaire du tegmentum pontique par I’acide ka.inique. C.R. Acad. Sci. Paris, 289: 959-964. Siegel, J.M. (1985) Pontomedullary interactions in the generation of REM sleep. In D.J. McGinty, R. Drucker-Colin, A. Morrison and L. Parmeggiani (Eds.), Brain Mechanisms of Sleep, Raven Press, New York, pp. 157-174. Steriade, M. and Hobson, J.A. (1976) Neuronal activity during the sleep-waking cycle. Prog. Neurobiol., 6: 155-376. Steriade, M., Datta, S., Part, D., Oakson, G. and Dossi, R.C. (1990) Neuronal activities in brain-stem cholinergic nuclei related to tonic activation processes in thalamocortical systems. J. Neurosci. (In press). Stevens, R.T., Hodge, C.J. and Apkarian, A.V. (1982) Kdliker-Fuse nucleus: The principal source of pontine catecholaminergic cells projecting to the lumbar spinal cord of cat. Brain Res., 239: 589-594. Swanson, L.W. and Hartman, B.K. (1975) The central adrenergic system: An immunofluorescence study of the location of cell bodies and their efferent connections in the rat utilizing dopamine-P-hydroxylase as a marker. J. Comp. Neurol., 163: 467-505. Takigawa, M. and Mogenson, G.J., (1977) A study of inputs to antidromically identified neurons of the locus coeruleus. Brain Res., 135: 217-230. Vanni-Mercier, G., Sakai, K., Lin, J.S. and Jouvet, M. (1989) Mapping of cholinoceptive brainstem structures responsible for the generation of paradoxical sleep in the cat. Arch. Ital. Biol., 127: 133-164. Vertes, R.P. (1984) Brainstem control of the events of REM sleep. Prog. Neurobiol., 22: 241-288. Vincent, S.R. and Rhiner, P.B. (1987) The immunohistochemical localization of choline acetyltransferase in the cat brain. Brain Res. Bull., 18: 371-415. Watson, S.J., Khachaturian, H., Akil, H., Coy, D.H. and Goldstein, A. (1982) Comparison of the distribution of dynorphin systems and enkephalin systems in brain. Science, 218: 1134-1136. Webster, H.H. and Jones, B.E. (1988) Neurotoxic lesions of the dorsolateral pontomesencephalic tegmentum-cholinergic cell area in the cat. 11. Effects upon sleep-waking states. Brain Rex, 458: 285-302. Yoshimoto, Y., Sakai, K., Luppi, P.H., Fort, P., Salvert, D. and Jouvet, M. (1989) Forebrain afferents to the cat posterior hypothalamus: A double labeling study. Brain Res. Bull., 23: 83-104. Young, W.S., Uhl, G.R. and Kuhar, M.J. (1978) Iontophoresis of neurotensin in the area of the locus coeruleus. Brain Res., 150: 431-435.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pornpeiaiio (Eds.) P r u u r e ~in~ Bruin Research. Vol. XN 0 1YY1 Elseviei Science Publishers B.V
47 CHAPTER 4
Afferent regulation of locus coeruleus neurons: anatomy, physiology and pharmacology
',
G. Aston-Jones M.T. Shipley 2 , G. Chouvet 3 , M. Ennis *, E. van Bockstaele V. Pieribone 4, R. Shiekhattar ',H. Akaoka G. Drolet B. Astier 5 , P. Charlity R.J. Valentino and J.T. Williams
',
'
',
3,
DiLision of Behacioral Neurobiology, Department of Mental Health Sciences, Hahnemann Uniaer.sity, Broad and Vine, Philadelphia, PA and Department of Anatomy and Cell Biology, Unicersity of Cincinnati, College of Medicine, Cincinnati, OH, U.S.A.; INSERM U 171, Centre Hospitalier Lyon-Sud, Pais. 4H, Chemin de Grand Reuoyet, 69310 Pierre-Benite, France; Department of Histology and Neurobiology, Karolinska Institutet, S-104 01 Stockholm, Sweden; -5 Laboratoire Neuropharmacologie, University Claude Bernard, Faculte de Pharmacie, 8 Are. Rockefeller, 69008 Lyon, France; Oregon Health Sciences Unicersity, Vollum Institute for Biomedical Research, 3181 SW Sam Jackson Park Road, Portland, OR, U.S.A.
'
Tract-tracing and electrophysiology studies have revealed that major inputs to the nucleus locus coeruleus (LC) are found in two structuies, the nucleus paragigantocellularis (PGi) and the perifascicular area of the nucleus prepositus hypoglossi (PrH), both located in the rostra1 medulla. Minor afferents to LC were found in the dorsal cap of the paraventricular hypothalamus and spinal lamina X. Recent studies have also revealed limited inputs from two areas nearby the LC, the caudal midbrain periaqueductal gray (PAG) and the ventromedial pericoerulear region. The pericoeruleus may provide a local circuit interface to LC neurons. Recent electron microscopic analyses have revealed that LC dendrites extend preferentially into the rostroniedial and caudal juxtaependymal pericoerulear regions. These extracoerulear LC dendrites may receive afferents in addition to those projecting to LC proper. However, single-pulse stimulation of inputs to such dendritic regions reveals little or no effect on LC neurons. Double-labeling studies have revealed that a variety of neurotransmitters impinging on LC neurons originate in its two major afferents, PGi and PrH. The LC is innervated by PGi neurons that stain for markers of adrenalin, enkephalin or corticotropin-releasing factor. Within PrH, large proportions of LC-projecting neurons stained for GABA o r met-enkephalin. Finally, in contrast to previous conclusions, the
-'
dorsal raphe does not provide the robust 5-HT innervation found in the LC. W e conclude that 5-HT inputs may derive from local 5-HT neurons in the pericoerulear area. Neuropharmacology experiments revealed that the PGi provides a potent excitatory amino acid (EAA) input to the LC, acting primarily at non-NMDA receptors in the LC. Other studies indicated that this pathway mediates certain sensory responses of LC neurons. NMDA-mediated sensory responses were also revealed during local infusion of magnesium-free solutions. Finally, adrenergic inhibition of LC from PGi could also be detected in nearly every LC neuron tested when the EAA-mediated excitation is first eliminated. In contrast to PGi, the PrH potently and consistently inhibited LC neurons via a GABAergic projection acting at GABA, receptors within LC. Such PrH stimulation also potently attenuated LC sensory responses. Finally, afferents to PGi areas that also contain LC-projecting neurons were identified. Major inputs were primarily autonomic in nature, and included the caudal medullary reticular formation, the parabrachial and Kolliker-Fuse nuclei, the PAG, NTS and certain hypothalamic areas. These results are interpreted to indicate that the LC may function in parallel to peripheral autonomic systems, providing a cognitive complement to sympathetic function.
Key words: excitatory amino acici, enkephalin, CRF, epinephrine, locus coeruleus, nucleus paragigantocellularis, nucleus prepositus hypoglossi
48
Introduction The noradrenergic nucleus locus coeruleus (LC) has been the subject of intense scrutiny and diverse functional hypotheses. Interest in this nucleus can be traced to the finding by Dahlstrijm and Fuxe (1964) that the LC provides the major source of norepinephrine (NE) to the telencephalon. Subsequent studies have revealed that the LC is the most ubiquitous of all neural projection systems in the CNS, providing innervation of all major levels of the neuraxis (reviewed in Foote et al., 1983). Many investigations have also examined the effects of NE on activity of neurons in LC target areas (described in more detail by Waterhouse et al., Woodward et al., McCormick et al., and Segal et al., in this volume), and still others have delineated the physiological and pharmacological characteristics of LC neurons from in uitro preparations (see Williams et al., and Christie, this volume) as well as discharge properties of LC neurons in behaving animals (see Aston-Jones et al., Jacobs et al., Sakai, and Sarah and Segal, this volume). Despite this intense research effort, our understanding of the NE-LC system has been hampered by a serious void; until recently, there has been no systematic input-output analysis of LC neurons. Whereas functional analysis of other brain areas is predicated upon a thorough understanding of afferent connections and the effects of such inputs on cellular discharge, most LC functional analysis has been predicated on the results of lesion and pharmacological experiments (see Aston-Jones et aL, 1984). However, a new approach to understanding the noradrenergic LC system is emerging, one based upon analysis of afferent circuits regulating the activity of LC neurons. While much remains obscure about the precise functions of the LC, recent findings concerning the connections, neurotransmitters, and functions of the brain areas that innervate the LC are yielding new insights into the stimuli and events that control the LC, and thus the role of the LC in brain function.
Our finding that the nucleus LC is strongly innervated by two medullary nuclei, the PGi and prepositus hypoglossi (PrH) (reviewed below) (Aston-Jones et al., 1986a), has been supported by several recent findings (McMahon and Wall, 1985; Deutch et al., 1986; Ennis and Aston-Jones, 1987,1989b; Guyenet and Young, 1987; Haselton and Guyenet, 1987; Saper, 1987; Grenhoff et al., 1988; Hajos and Engberg, 1988; Pieribone et al., 1988; Chen and Engberg, 1989; Engberg, 1989; Rasmussen and Aghajanian, 1989; Sesack et al., 1989; Svensson et al., 1989; Tung et al., 1989; Van Bockstaele et al., 1989b; Wallace et al., 1989; Pieribone and Aston-Jones, 1991). It is being recognized that a restricted set of afferents to the LC can make sense of many previous observations. As Saper states in a recent review (Saper, 1987): “These observations explain certain findings in earlier anterograde neuroanatomical tracing studies, in which most fibers from cell groups that were thought to project to the LC.. .were found to stop at the edge of the nucleus, curiously avoiding the core of the LC . . . the outstanding finding of a single main source for LC excitatory inputs.. .may make it much easier to assign a function to LC.” Merents to LC: tract-tracing studies Our recent review of LC afferent control details previous studies in this area (Aston-Jones et al., 1990b). Receptor binding, immunocytochemical, and pharmacological studies indicated that a multitude of transmitter systems impinge on this nucleus, implying that the LC received inputs from a wide variety of central sites (Foote et al., 1983). This belief was strongly reinforced by the fact that the LC projects throughout the neuraxis, and therefore it was felt that this nucleus was likely to be under similarly widespread afferent regulation. This viewpoint was further supported by tract-tracing studies published over a decade ago (Cedarbaum and Aghajanian, 1978; Clavier, 1979; Morgane and Jacobs, 1979) which indicated that the LC receives inputs from a wide array of brain
49
areas, including several telencephalic and other forebrain regions. However, our recent re-examination of LC afferents found that this view of multiple-source projections to LC was unwarranted (Aston-Jones et al., 1986a, 1990b; Pieribone et al., 1988, 1989; Pieribone and Aston-Jones, 1991). Using discrete iontophoretic injections of the sensitive tracttracer, WGA-HRP, into the LC proper, we found that only two areas contained numerous heavily labeled neurons: the nucleus PGi in the ventrolateral rostral medulla, and the area of medial PrH in the dorsomedial rostral medulla. Anterograde tracing from these areas confirmed that they provide major inputs to the LC. Two minor inputs (areas containing few cells weakly labeled) were also identified, in the dorsal cap of the paraventricular nucleus of the hypothalamus and in the intermediate zone of the cervical spinal cord. Similar results were also obtained using the retrograde tracer Fluoro-Gold (FG) (Pieribone and Aston-Jones, 1991). We used anterograde labeling to investigate discrepancies with earlier studies that had identified many more inputs to the LC. Injections into the central nucleus of the amygdala (CNA) (Aston-Jones et al., 1986a), prefrontal cortex (Chiang et al., 19871, dorsal raphe (Pieribone et al., 19891, dorsal spinal horn (Aston-Jones et al., 1986b), ventral tegmental area (VTA) (unpublished observations), or nucleus tractus solitarius (NTS) (Ennis and Aston-Jones, 1989b) yielded no fiber or terminal labeling in the LC but instead produced heavy labeling in adjacent, pericoerulear structures including parabrachial and pontine central grey areas. Furthermore, in the same studies single-pulse stimulation of these areas yielded no response in LC neurons, or responses that were weak and long in latency, consistent with the possibility that these areas had no direct input to the LC. In contrast, identical stimulation potently activated cells in pericoerulear areas that contained corresponding anterograde label (Aston-Jones et al., 1986a; Chiang et al., 1987; Ennis and Aston-Jones, 1989b). Thus, we surmised that
the previous reports of widespread afferents to the LC probably reflected retrograde labeling from injections that spread into pericoerulear areas that are specifically innervated by a host of structures that do not project to the LC. Indeed, when our injections encroached heavily on neighboring pericoerulear areas, retrogradely labeled neurons were found in many additional areas, similar to the previous reports. The conclusion from our anatomical and physiological investigations was that the major afferents to LC derive from the PGi and PrH, with possible minor inputs from the paraventricular hypothalamus and intermediate zone of the cervical spinal cord (Aston-Jones et al., 1986a). It should be noted that both of the retrograde tracers used, WGA-HRP and FG, yielded a substantial “halo” surrounding the injection site, so that analysis of possible input from pericoerulear areas was not feasible in our experiments, and we suggested that further study would be necessary to determine if local neurons in pericoerulear areas innervate LC (described below). It should also be noted that our more recent studies employing other tracers (WGA-apoHRP-Gold or choleratoxin) have revealed additional groups of cells retrogradely labeled from LC than were consistently seen with WGA-HRP (e.g., hypothalamus/preoptic area, Kolliker-Fuse nucleus, A5 area, B9 area). Additional studies are underway to determine the degree to which these additional areas innervate the LC nucleus vs. peri-LC regions, and their possible functional effect on LC neurons. Further analysis has revealed that the LC-projecting neurons in PrH and in PGi are not distributed equally throughout these two medullary nuclei, but rather that they reside in topographically restricted subdivisions (Aston-Jones et al., 1990b). As shown in Figures 1 and 2, LC afferent neurons are located in the very medial aspect of PrH; they extend throughout the length of this nucleus from the hypoglossal nucleus to the genu of the 7th nerve. Many cells also reside more ventrally, scattered along the lateral borders of the medial longitudinal fasciculus. As this area of
50
Fig. 1. Retrogradely labeled neurons in paragigantocellularis (PGi) (A) and prepositus hypoglossi (PrH) (B) following a WGAapoHRP-Gold injection into ipsilateral locus coeruleus (LC). Darkfield photomicrographs. A. Coronal section through PGi. Arrow marks ventral brain surface. B. Coronal section through PrH. Arrow marks floor of IVth ventricle at midline.
51
Fig. 2. LC afferent neurons in the ventrolateral rostral medulla (PGi area; panels A-D) and dorsomedial rostral medulla (PrH; panels E-H). Computer-aided plots showing retrogradely labeled neurons (filled squares) in the rostral medulla for an animal with an injection of Fluoro-Gold restricted to the nucleus LC proper. Sections (40 p m thickness) are ordered from caudal to rostral for PGi (A-D) and for PrH (E-H). Inserts in A and D. Low-power hemisections showing levels of plots in panels A and D. This injection did not incorporate peri-LC regions containing extranuclear LC dendrites.
PrH is immediately adjacent to the dorsal or lateral aspects of the medial longitudinal fasciculus, we have denoted it as the perifascicular PrH. LC-projecting cells in PGi are broadly distributed through the rostral ventrolateral medulla, but also exhibit significant topography (Pieribone and Aston-Jones, 1991). LC afferent neurons are preferentially located medially in rostral juxtafacial PGi, and are more laterally located in caudal aspects of PGi; very few neurons project to the LC from the niedial caudal PGi area (Fig. 2). This has significant implications for which afferents to PGi may be expected to contact LC-projecting neurons. For example, part of the projection from the ventrolateral PAG innervates the medial caudal aspect of PGi; this projection may not contact LC afferent cells. In contrast, fibers from the NTS or the ventromedial periaqueductal grey (supraoculomotor nucleus) terminate in regions that contain an abundance of LC-projecting neurons (Van Bockstaele et al., 1989a). Our recent investigations into the pathways taken by fibers projecting to the LC from PGi have yielded the unexpected result that at least three distinct pathways carry projections from the rostral ventrolateral medulla to the LC. Injections of PHA-L into the PGi labeled two distinct ascending fiber pathways (Van Bockstaele et al., 1989b). One was located laterally, proceeding rostrally just medial to the superior olive to pass through the Kolliker-Fuse and lateral parabrachialis areas before entering the LC from the lateral and rostral aspect. The second pathway was located medially in the medulla in the same vicinity as, but distinctly ventral to, the medullary adrenergic bundle (MB). The third pathway was revealed in our analysis of the ascending adrenergic innervation of the LC. Our immunohistochemistry (Pieribone et al., 1988; Astier et al., 1990; Pieribone and Aston-Jones, 1991), as well as that of others (e.g., Hokfelt et al., 1974, 1985; Astier et al., 1987), demonstrated that the prominent adrenergic input to LC uses the MB. We have recently confirmed this by showing that lesions of the MB decrease the number of adrenergic neu-
52
rons retrogradely labeled from the LC by 90% while decreasing the number of non-adrenergic neurons retrogradely labeled in the same area by only 48% (Astier et al., 1990). This further reveals that nearly all adrenergic projections utilize the MB, and that this pathway carries other inputs to the LC from PGi as well. Physiological characteristics of neurons antidromically identified from LC
Physiological properties of LC-projecting neurons in PGi. To confirm our anatomic results for major inputs to LC from the PGi, and to examine physiological properties of LC-projecting PGi neurons, we focally stimulated the LC and searched for antidromically activated neurons in the PGi of chloral hydrate-anesthetized rats. Twenty of 79 PGi neurons (25%) were antidromically activated from the LC (Ennis and AstonJones, 1987). The antidromic latencies for these neurons averaged 10 msec, but were bimodally distributed, reflecting slowly conducting and more rapidly conducting LC afferents. Other physiological attributes distinguished these two populations of LC afferents: the slowly conducting cells typically exhibited slow spontaneous activity and large positive spikes, while the more rapidly conducting cells were not spontaneously active and yielded small negative action potentials. In addition, the slowly conducting LC afferents were preferentially located in the ventromedial PGi. Finally, many LC-projecting PGi cells were potently activated by footpad stimulation at latencies (5-10 msec) consistent with our hypothesis that these cells mediate LC responses (typically 20 msec onset) to the same stimuli (see below). Physiological properties of LC-projecting neurons in PrH. Focal stimulation of the LC also produced antidromic activation in a high percentage of PrH neurons (17 of 61 cells, or 28%) (Ennis and Aston-Jones, 1989b). Typically these cells exhibited small, negative spikes and were spontaneously active (1-30 spikes/sec). In contrast, most
neurons detected in PrH recordings that were not antidromically activated from LC were not spontaneously active. Unlike PGi neurons, PrH cells that project to the LC do not respond to footpad stimulation. Thus, a high proportion of neurons in both the PGi and PrH were antidromically activated by focal LC stimulation, confirming the retrograde and anterograde tract-tracing results for major inputs to LC from these rostral medullary nuclei.
Attempts to antidromically activate neurons in other nuclei. Of 9 neurons in 11 penetrations through the NTS, none were antidromically activated by LC stimulation (Ennis and Aston-Jones, 1989b). This result is consistent with the findings in the above anatomic experiments that neurons in the NTS are not retrogradely labeled from the LC and is also consistent with our finding that lesions of the NTS do not alter the discharge of LC neurons. Seventeen neurons recorded in the ventral PAG, in the area of the dorsal raphe, exhibited large entirely positive spikes with impulse waveform durations of 2 to 3 msec, and a slow and regular spontaneous discharge rate ranging from 0.5 to 6 Hz. These physiological properties are similar to those described for dorsal and median raphe neurons (Aghajanian, 1978; Blier and de Montigny, 1987). None of these cells was antidromically activated from the LC, even with stimulation intensities of up to 1400 PA. In addition, none of 20 contralateral LC neurons were antidromically activated by LC stimulation (Ennis and Aston-Jones, 1989b), also consistent with our finding of no retrograde labeling between the loci coerulei. Finally, only two of 40 cells tested in the lateral reticular nucleus were found to be antidromically driven by focal LC stimulation (Ennis and Aston-Jones, 1989b). One of these was located in the rostral pole of this nucleus, at the caudal border of PGi and may be a member of this latter set of cells.
53
Inputs to the pericoerulear area: innervation of LC dendrites, local LC afferents or separate circuits?
It has been commonly thought that the regions surrounding the LC in the rat are largely composed of fiber tracts with few if any neurons. However, our recent work (Aston-Jones et af., 1986a, 1990b; Fu et al., 1988) along with that of several other groups, (e.g., Saper, 1982, 1987; Cechetto et al., 1985; Deutch et af., 1986; Sesack et af., 1989; Wallace et af., 1989), has demonstrated dense terminal fields from several brain areas to the pericoerulear region. Thus, there are dense inputs to the parabrachial area lateral to the LC from the NTS (Mantyh and Hunt, 19841, dorsal spinal horn (Cechetto et af., 1985; Standaert et al., 19861, and CNA (Aston-Jones et af., 1986a; Wallace et al., 1989) among others. The central grey medial or rostral to LC, on the other hand, receives inputs from the frontal cortex (Armten and Goldman, 1984; Chiang et af., 1987; Sesack et al., 1989), dorsal raphe (Conrad et al., 1974; Bobillier et al., 1979; Pieribone et af.,19891, CNA (Aston-Jones et af., 1986a; Wallace et al., 1989) and VTA (Deutch et af., 1986) (unpublished observations). These findings cast doubt on the view that the pericoerulear area is relatively cell-free, and prompted us to begin a systematic analysis of the neuropil in the region surrounding the LC nucleus. While these studies are fairly new, several interesting results have been obtained. In addition to Barrington’s nucleus and the laterodorsal tegmental nucleus, already well documented by others, we have found that the pericoerulear area is composed of large numbers of neurons which vary in size, shape and dendritic orientation as a function of location with respect to the LC (Fu et al., 1988). Such neurons are numerous on all borders of the LC nucleus. Of particular interest is the fact that some of these pericoerulear areas contain dense accumulations of dendrites from LC neurons (Fu et af., 1989) (described below) and are heavily targeted by fibers containing sev-
eral peptides. These pericoerulear areas have not been previously characterized, and we are presently using cytoarchitectonics and immunohistochemistry to parcellate this cell- and neurite-rich region. Although there are numerous neurons in the pericoerulear area, very little is known about whether pericoerulear neurons innervate the LC. Our previous studies using injections of retrograde tracers into the LC (Aston-Jones et af., 1986a) could not adequately address this issue because the injected tracer always produced an injection “halo” in the pericoerulear area, particularly in the rostromedial pericoerulear zone. We were able to tentatively rule out inputs from the parabrachial area (Aston-Jones et af., 1986a), but inputs from rostromedial pericoerulear areas are possible. Indeed, two recently developed anatomic techniques, PHA-L and WGA-apoHRP-Gold, have indicated that such pericoerulear neurons may innervate the LC. One region in which retrogradely labeled cells were often found, within the injection halo of WGA-HRP cases, was the caudal ventrolateral PAG, immediately rostral to the LC. Anterograde transport of PHA-L from the caudal ventrolateral PAG revealed prominent innervation of the central grey ventromedial to the LC, and scattered light innervation of the LC as well (Ennis et af., 1989, 1991). Additional electrophysiological experiments confirmed a minor input to the LC from this nearby PAG region (described below). Injections of the tracer WGA-apoHRP-Gold form discrete deposits which can easily be limited to only a subregion of the LC. In addition, this tracer produces no halo, making analysis of possible inputs from immediately adjacent pericoerulear areas feasible. In our initial use of this tracer with subtotal deposits into LC, we consistently found numerous retrogradely labeled neurons surrounding the LC, particularly in the ventromedial and rostral pericoerulear regions; some of these cells also stain immunohistochemically for 5-HT (see Fig. 8, described below). While it is
54
tempting to speculate that the pericoerulear area may provide numerous afferents to the LC area, it remains to be determined whether such local labeling is due to synaptic innervation of LC neurons, or dendrites that extend into the LC but do not innervate LC neurons.
Extranuclear dendrites of LC neurons Several reports have documented that processes of LC neurons extend outside the nucleus proper for a few hundred p m (Shimizu and Imamoto, 1970; Swanson, 1976; Shimizu et al., 1978; Grzanna and Molliver, 1980; Grzanna et al., 1980; Cintra et al., 1982). However, it has never been clear whether the extranuclear processes were dendrites, axons, or both. We recently adapted a gold-silver tone method (Gallyas et al., 1982) to intensify immunohistochemically stained processes. Using this method, our light microscopic examination of dopamine-P-hydroxylase (DPH)-positive LC neurons and dendrites revealed that extranuclear LC processes have a remarkable degree of spatial organization; the vast majority of processes ramified in two distinct, focal pericoerulear zones-rostromedial and caudal juxtaependymal peri-LC. As shown in Figure 3, light microscopic observations revealed that many processes in the medial pericoerulear area were clearly dendrites, but many others were very thin and beaded, and appeared to be axons (Fu et al., 1988). However, it was impossible on the basis of these observations to determine what proportion of these processes were axons and which were dendrites. To unambiguously identify the nature of such processes, we analyzed DPHand TH-stained material at the electron microscope level. Electron microscopic analysis indicated that out of more than 500 labelled processes in the rostromedial and caudal juxtaependymal peri-LC zones, all but three labelled processes were dendrites (the three exceptions were in the caudal zone) (Fu et al., 1989). In additional studies, we found that these extranuclear LC dendrites are heavily targeted by noncatecholaminergic afferent synapses. Thus, LC
Fig. 3. High magnification brightfield photo montage of a 25-gm-thick horizontal section through the LC and pericoerulear region stained with an antibody directed against dopamine-p-hydroxylase and Nissl counter-stained. Note profuse dendrites exiting the rostra1 edge of LC and forming a rostrally directed dendritic fascicle. Note also the fine, beaded nature of extranuclear dendrites in the rostromedial pericoerulear region. Many of these fibers resemble axons, but all were shown to be dendrites upon electron microscopic examination (Fu et al., submitted; Shipley et al., submitted). Orientation: Rostra1 is at the top, medial is to the right.
neurons have an appreciable postsynaptic surface that lies a considerable distance outside the confines of the nucleus proper. Moreover, this “re-
55
ceptive surface” is preferentially distributed in two discrete peri-LC zones. The fact that several brain areas project to these pericoerulear regions raises the possibility that some of these inputs synapse upon extranuclear dendrites of LC neurons. Both PGi and PrH project to these two pericoerulear zones as well as to the LC proper, and may terminate on LC neurons in both locations. However, at this time there are no data concerning the sources or nature of synaptic contacts onto the extranuclear dendrites of LC neurons. The rostromedial zone containing extranuclear dendrites, in particular, receives input from several brain areas that do not innervate the LC nucleus (frontal, insular and perirhinal cortices, amygdala, dorsal raphe, VTA). Although inputs to the soma-rich LC nucleus would presumably exert a greater influence on LC discharge than those impinging on distal dendrites, pericoerulear contacts onto LC processes could provide functionally important modulation of activity driven by PGi and PrH which do target the LC proper. While the foregoing histochemical analysis revealed properties of LC neurons viewed as a group, they do not distinguish possible differences among individual LC neurons. To address this issue, we have studied the morphologies and dendritic domains of individual, intracellularly filled LC neurons. These cells were filled with biocytin during the course of physiologic/ pharmacological studies in in vitro tissue slices taken in the horizontal plane (see Williams et al., this volume). Filled neurons were remarkably uniform in their morphology, with 4 or 5 major dendrites leaving the soma from all directions, and a single thin process (presumably the axon) often emanating from a proximal dendrite. As shown in Figure 4, dendrites of individual LC neurons could be seen to exit the LC nucleus for considerable distances, extending preferentially in the rostral, medial and caudal directions. This property also appears to be uniform for LC neurons; of the 20 filled cells examined to date, all have at least 1 dendrite that extends into the rostromedial peri-
lVth
Ventricle
ratrat\
\
Fig. 4. Camera-lucida drawing of an LC neuron that was intracellularly filled in a horizontal tissue slice (350 Fm-thick). Boundaries of the LC nucleus proper are indicated by the dashed lines. Note extranuclear dendrites extending rostrally and medially; a similar profile was obtained for all individual cells filled to date, with many cells exhibiting even longer extranuclear dendrites. Orientation as indicated.
LC region and many have a dendrite that also extends distally into the caudal juxtacoerulear zone as well. These properties were similar for LC neurons whose somata were located throughout the nucleus (though soma size and dendritic extent may be smaller for cells ’located caudally vs. rostrally in the nucleus). Thus, individual LC neurons often possess dendrites that extend into both of the preferred extranuclear dendritic zones; there do not appear to be separate subpopulations of LC neurons giving rise to extranuclear dendritic extensions. Neurochemical identity of afferents to LC
Several major neurotransmitter inputs to the LC have been systematically characterized by ourselves and others. Many of the neurotransmitters found in fibers in the LC or peri-LC are also found in cells in PGi, PrH or peri-LC, as indicated in Table 1. Using anatomic double-labeling techniques, we have identified sources of adren-
TABLE 1 Neurochemically characterized cells and fibers in the LC, peri-LC, PGi and PrH Transmitter
Fibers in LC
Fibers in peri-LC
Cells in PGi
ACh
-
+
+
* * Altschuler et al., 1984; Butcher and Woolf, 1984; Kimura et al., 1984; Sutin and Jacobowitz, 1988; Ruggerio et al., 1990
epinephrine
+
+
+
Hokfelt et al., 1974; Ross et al., 1981; Berod et al., 1984; Granata et al., 1985; Hokfelt et al., 1985; Kalia et al., 1985; Ruggiero et al., 1985; Astier et al., 1986; Astier et al., 1987; Haselton and Guyenet, 1987; Tucker et al., 1987; Pieribone et al., 1988; Thor and Helke, 1988; Pieribone and Aston-Jones, in press
serotonin
+
+
+
+
+
* * Figs. 7 & 8; Bowker et al., 1981; Beitz, 1982; Hunt and Lovick, 1982; Steinbusch, 1984; Thor and Helke, 1988; Pieribone et al., 1989
excitatory amino acids
+
+
+
?
+
Ottersen and Storm-Mathisen, 1984a; Ottersen and Storm-Mathisen, 1984b; Ennis and Aston-Jones, 1986; Ennis and Aston-Jones, 1988; Aston-Jones and Ennis, 1988
GABA
+
+
+
+
* * Figs. 5 & 6; Ottersen and Storm-Mathisen, 1984a; Ottersen and Storm-Mathisen, 1984b; Mugnaini and Oertel, 1985; Shipley et a/., 1988; Ennis and Aston-Jones, 1989a; Ennis and Aston-Jones, 1989b
enkephalin
+
+
+
+
Hokfelt et al., 1977; Sar et al., 1978; Hokfelt et al., 1979; Uhl e t a / . , 1979; Miller and Pickel, 1980a; Miller and Pickel, 1980b; Watson et a/., 1980; Finley et al., 1981; Hunt and Lovick, 1982; Conrath-Verrier etal., 1983; Khachaturian et al., 1983; Guthrie and Basbaum, 1984; Lynch et al., 1984; Charnay et al., 1985; Fallon and Leslie, 1986; * * Cassini et al., 1989; Drolet et al., in press
substance P
+
+
+
+
Hokfelt et al., 1978; Ljungdahl et al., 1978; Nomura et al., 1982; Triepel etal., 1985; * * Cassini et al., 1989; Sutin and Jacobowitz, 1988
neurotensin
+
+
+
+
Uhl et al., 1979; * * Beitz, 1982; Jennes et al., 1982; Minagawa et al., 1983; Triepel et al., 1984; Papadopoulos et al., 1986
VIP
+
+
+
+
* * Eiden et al., 1982;Martin et al., 1987;Sutin and Jacobowitz, 1988; Wang and Aghajanian, 1989
somatostatin
+ +
+ +
+
+ +
Johansson et al., 1984;Vincent et al., 1985
CRF
* * Unpublished observations by us.
+
Cells in PrH
+ +
Cells in peri-LC
References
* * Fig. 9; Bloom et al., 1982; Merchenthaler e t a / . , 1982; Olshcowka et al., 1982; Cummings et al., 1983; Swanson et al., 1983; Merchenthaler, 1984; Sakanaka et al., 1987
Fig. 5. Low-power darkfield photomicrographs of coronal sections taken from rostra1 (left) and caudal (right) LC, stained with an antibody against glutamate decarboxylase (GAD), a marker of GABAergic neurons. Note the dense GAD-positive terminal plexus within LC (white arrow), and surrounding the LC nucleus proper both rostrally and medially, in the region of the extranuclear dendrites (black arrow; see Figs. 3 and 4). Orientation: dorsal is at the top and medial (IVth ventricle) is at the left.
58
ergic and the GABAergic inputs to the LC, and preliminary studies also indicate sources of 5-HT, corticotropin-releasing factor (CRF) and enkephalin inputs. As summarized below, these neurochemically characterized inputs originate from PGi, PrH, or the pericoerulear area. Adrenergic afferents to the LC Using retrograde transport of FG combined with PNMT immunofluorescence, we (Pieribone et al., 1988; Pieribone and Aston-Jones, 1991) determined that most adrenergic afferents to the LC derive from the PGi. We calculated that PNMT-immunoreactive (PNMT-ir) neurons constitute 21% of LC-projecting cells from that area,
while a small number of adrenergic afferents from PrH also exist (4% of LC-projecting PrH neurons). Within the PGi, the proportion of LC afferent neurons that stained for PNMT varied topographically. In particular, although the number of LC-projecting neurons was smaller in rostral, juxtafacial PGi, approximately 80% of LC afferents from this subregion of PGi were PNMT-ir. In contrast, about 18% of LC afferents in more central PGi (e.g. , the subambigual, retrofacial aspect of this nucleus) also stain for PNMT. GABAergic afferents to L C Our physiological and pharmacological studies of the PrH-LC pathway have demonstrated a
Fig. 6 . Brightfield photomicrographs showing double-labeled GABAergic afferents to LC from PrH (examples indicated by arrows). Neurons in PrH exhibiting brown, diffuse, DAB reaction product for GAD (a marker of GABAergic neurons) immunohistochemistry and punctate black granules of WGA-apoHRP-Gold transported retrogradely from the LC. Note also numerous GAD-positive, non-LC afferent neurons. IVth ventricle is at the top, midline is to the right.
59
$
LC
Fig. 7. Computer-aided reconstruction of PHA-L-labeled fibers in the LC area after injections into the dorsal raphe. Note nearly total lack of labeling in LC proper, but robust innervation of the ventromedial pericoerulear region. Fiber labeling was also extensive in the rostra1 pericoerulear region.
strong GABAergic innervation of LC from this major afferent (Ennis and Aston-Jones, 1989a, 1989b) (described below). Immunohistochemistry for GABA or GAD through the LC area revealed a dense innervation of the LC nucleus by
GABAergic fibers and terminals (Shipley et at., 1988) (Fig. 5). Staining for GABA/GAD following colchicine treatments revealed numerous, labeled neuronal somata in the same area of PrH (medial, perifascicular) that also contains neurons that project to LC (Pieribone et al., 1990). Recent experiments using WGA-apoHRP-Gold as the retrograde tracer have demonstrated that a large number (more than 40%) of LC-projecting neurons from PrH are GAD- or GABA-positive (Pieribone et al., 1990) (Fig. 6). These findings provide anatomic confirmation of our physiological and pharmacological results (described below).
Serotonergic afferents to LC We have confirmed results of others that the LC is densely innervated by 5-HT fibers. How-
Fig. 8. High-power view of a 5-HT-positive neuron (as indicated by diffuse, opaque DAB reaction product) in the ventromedial pericoerulear area that also contains WGA-apoHRP-Gold transported from LC (black particles in somata). Most of the other WGA-apoHRP-Gold particles in this field are also contained in cells (apparent with DIC illumination) that are not 5-HT-positive.
60
61
ever, retrograde transport of WGA-HRP or FG from LC did not consistently label neurons in the dorsal raphe (Aston-Jones et al., 1986a), the previously presumed source of 5-HT to the LC. Anterograde transport of WGA-HRP or PHA-L from the dorsal raphe nucleus labeled fibers in the central grey medially adjacent to the LC, but not in the LC proper (Pieribone et af., 1989) (Fig. 7). In addition, extensive lesions of the dorsal raphe nucleus (including the lateral “wings”) failed to diminish the dense 5-HT fiber staining found in LC (Pieribone et af., 1989). Furthermore, brain hemisections in the frontal plane rostral or caudal to the LC, or LC hemi-island preparations (severing tissue around the LC on rostral, medial and caudal sides) did not decrease 5-HT staining in LC (Pieribone e f al., 1989). Most recently we have observed that 5-HT cells in the rostroventral pericoerulear region send processes into the LC (Pieribone et al., 1989). Such 5-HT neurons are also retrogradely labeled following discrete injections of WGA-apoHRP-Gold into the LC (Fig. 8). Therefore, 5-HT innervation of the LC may derive from this group of pericoerulear 5-HT neurons. Lesions of this area combined with 5-HT immunohistochemistry through the LC are needed to further substantiate this hypothesis.
CRF afferents to L C Using an antibody directed against rat/human CRF, we (Valentino et al., 1990) found numerous CRF-positive fibers in the LC and CRF-positive cells in areas that contain LC afferents (PGi, PrH, and the dorsal cap of the paraventricular hypothalamic nucleus; Fig. 9). Preliminary results using WGA-apoHRP-Gold retrograde tracing and
immunohistochemistry revealed doubly labeled neurons in the dorsal cap of the paraventricular hypothalamic nucleus and in the PGi. In addition, there are many CRF cells in the pericoerulear area (Fig. 91, raising the possibility that local, pericoerulear neurons also contribute to the CRF innervation of LC. These studies are in progress. Effects of CRF, and putative activation of CRF afferents, on LC activity, are described by Valentino and Curtis (this volume).
E n k e ~ ~ ~ a afferents Lin to LC Numerous fibers within the LC nucleus stain with an antibody raised against enkephalin. Enkephalin-positive fibers are even denser in the rostral and ventromedial pericoerulear regions containing extranuclear LC dendrites, raising the possibility that distal dendrites of LC neurons also receive enkephalin inputs. We have also observed numerous enkephalin-positive neurons in regions containing LC afferents in both PGi and PrH, as well as in the pericoerulear area. Finally, our retrograde tracing with WGA-apoHRP-Gold combined with enkephalin immunohistochemistry revealed a surprising number of doubly labeled neurons in both PGi and PrH, indicating that a substantial percentage of LC-projecting neurons in each of these rostral medullary regions contain enkephalin (Fig. 10; Drolet et al., 1990). The significance of these opioid peptide projections to the LC is the subject of ongoing research. Other neurotransmitters in PGi, PrH and L C We have examined neurons in PGi, PrH, and peri-LC to define the sources of other transmitter inputs to the LC. In addition to the neurons that stain for PNMT, 5-HT, GABA/GAD, CRF and
Fig. 9. Corticotropin-releasing factor-immunoreactive (CRF-IR) fibers and cells in the LC/peri-LC region, and in PGi. a. Photomicrograph of a coronal section showing fluorescent (rhodamine tagged) CRF-IR neurons immediately lateral to the LC nucleus proper. This animal was pretreated with colchicine to optimize cell-body staining of CRF. Arrow is in the IVth ventricle medially, and indicates the LC nucleus proper. b. CRF-IR neurons in the PGi of a colchicine-pretreated rat. Similar CRF labeling in animals with WGA-apoHRP-Gold injections in LC reveal that retrogradely labeled and CRF-IR neurons are interdigitated in PGi, and that some LC-projecting neurons in PGi are CRF-IR. Ventral brain surface visible at lower left. Dorsal is at the top and medial is to the right for both panels.
62
enkephalin reviewed above, we have identified neurons in the PGi and PrH that stain for ChAT, neurotensin, VIP, angler fish pancreatic polypeptide, dynorphin and substance P. In addition, fibers that stain for all of these putative neurotransmitter markers are densely located in periLC regions (Table l). Studies using retrograde transport combined with immunohistochemistry will determine which of these areas provides these neurochemically defined inputs to the LC. Cellular pharmacology of afferents to LC
Excitatory amino acid ( E M )pathway from PGi to LC Single pulse stimulation of PGi activated 78% of LC neurons (Ennis and Aston-Jones, 1986,
1988). Cholinergic antagonists had no effect on LC activation by PGi or sciatic nerve stimulation. This is consistent with recent observations that the LC appears to be nearly devoid of cholinergic fibers (Ruggiero et al., 1990). However, the EAA antagonists, kynurenic acid or D-glutamyl glycine, consistently blocked both PGi- and sciatic-induced activation of the LC (AstomJones and Ennis, 1988). Similar results have recently been obtained with local application of kynurenic acid or the non-NMDA antagonist CNQX (Fig. 11) (Ennis and Aston-Jones, submitted; Shiekhattar and Aston-Jones, 1991). The NMDA antagonists AP5 or AP7 were not effective on either response, indicating that PGi-induced EAA activation of LC neurons probably takes place at a non-NMDA receptor in the LC. These results
Fig. 10. Enkephalin-immunoreactive (Enk-IR) LC afferent neurons in the PGi. Brightfield photomicrograph of a coronal section through the PGi of a rat injected with WGA-apoHRP-Gold into the LC. Retrogradely transported WGA-apoHRP-Gold appears as black puncta inside neurons, whereas Enk-IR neurons contain the brown, diffuse diaminobenzidine reaction product. Neurons containing both labels are indicated by the solid arrows, while several sample Enk-IR cells that are not retrogradely labeled are indicated by the open arrows. Note the large percentage of LC-projecting neurons that are also Enk-IR. Ventral brain surface is at lower left; medial is at the right and dorsal is at the top.
63
30r
lull
I I I
680
, I
1360 Time (msec)
2040
'Or
f kn U
During-CNOX
lot 0
t
680
1360 Time (msec)
2040
30r Recovery
0
t
1360 Time ( m s e c )
680
20LO
Fig. 11. Locally applied CNQX blocks sensory responses of LC neurons. Cumulative peri-stimulus time histograms (PSTHs) revealing blockade of footpad nerve-evoked response in an LC neuron by local application of the specific nonNMDA antagonist CNQX. CNQX was applied from a composite dual-barrel micropipette as an 8 p M solution in artificial cerebrospinal fluid (ACSF). Similar blockade was observed for each of 12 cells tested. Footpad stimulation (to activate sciatic nerve) at arrows; 50 stimuli accumulated in each PSTH.
also suggested that sciatic-evoked excitation of LC neurons might be mediated through the PGiLC pathway. This was confirmed with experiments demonstrating that infusions of lidocaine, GABA or synaptic decouplers into PGi blocked or attenuated LC response to sciatic activation (described in Aston-Jones et al., this volume).
NMDA-receptor-mediated sensory response of L C neurons revealed by low-Mg++ infusion in vivo We (Shiekhattar et al., 1991) have developed a method to locally perfuse a small region of brain in vivo using slow microinfusion (30-60 nl/min) from a multibarrel micropipette. With this method, microinfusion of artificial cerebrospinal fluid (ACSF) lacking Mg++ ions onto LC neurons caused a significant increase (by about 36%) in footpad stimulation-evoked sensory responses of these cells (Fig. 12) (Shiekhattar and Aston-Jones, 1991). This increased responsiveness was prolonged, recovering about 3-5 min after termination of the Mg++-free ACSF infusion. This increase in sensory response magnitude was completely blocked by co-infusion with the specific NMDA antagonists AP5 (50 p M ; Fig. 12) or CGS 19755 (1 kM). There was no consistent effect of Mg++-free solutions on spontaneous LC discharge. These results demonstrate that NMDA receptor mechanisms exist for this sensory response of LC cells, but that they are normally occluded, presumably by Mg++ blockade of the NMDA channel. They also indicate a possible mechanism whereby EAA-mediated sensory responses may be augmented in the LC. Although little is known concerning variation of Mg++ concentrations in ciuo, these results indicate that decreased Mg could have profound effects on LC sensory responsiveness and, consequently, on noradrenergic neurotransmission throughout the brain. Finally, the finding that sensory input to LC accesses NMDA receptor mechanisms indicates that modulation of NMDA receptor activity may modulate LC function. + +
Adrenergic inhibition of L C from PGi Clear inhibitory effects of PGi stimulation on LC neurons appeared when the EAA-mediated activation of LC from the same stimulation sites in the PGi was eliminated. Under kynurenate blockade, an underlying inhibition from PGi was observed in 88% of LC cells (Ennis and AstonJones, 1988). We (Astier and Aston-Jones, 1989)
64
35
f
PRE
-
5
08. J .32'1
0
1360 Time ( r n s e c )
0
g'
2040
DUR
23 Q
0
I
680 1360 Time (rnsec)
2040
have also shown that systemic administration of the a 2 antagonist idazoxan (1 mg/kg) attenuates this inhibition in 80% of neurons tested (Fig. 13). A similar effect was seen for locally infused idazoxan (0.5-2.5 ng in 50 nl), whereas vehicle infusion was without effect. Several LC cells (15% of those tested) exhibited pure inhibition after PGi stimulation; systemic or local idazoxan was typically effective in antagonizing these responses as well. Furthermore, stimulation of the MB, which carries adrenergic fibers from the PGi to the LC, produced pure inhibition of LC activity. This response was also blocked by systemic or local idazoxan (Astier and Aston-Jones, 1989). These results provide physiological and pharmacological confirmation of our anatomic results for a prominent adrenergic projection to LC from C1 cells in PGi (described above), and indicate that this input uniformly inhibits activity of LC neurons.
** T
0
100 -. ~
C
0
+ 0
cU X
Id
50 ACSF
Mg*+-f ree ACSF
Mg++-free ACSFt AP5
Fig. 12. In ciuo perfusion of LC with MgCC-free ACSF reveals NMDA-mediated sensory response. A. PSTH of the response of a typical LC neuron to stimulation of the rear footpad (for sciatic nerve activation; at arrow) before infusion of Mg++-free ACSF.B. Additional late response to footpad stimulation is apparent during slow infusion (50 nl/min) of Mg++-free ACSF into the LC from a micropipette barrel adjacent to the recording barrel. 50 sweeps in each PSTH.C. Bar graph illustrating increase of footpad response magnitude for LC neurons during infusion of Mg++-free ACSF (ACSFM g f f ; **P < 0.01, paired t-test on absolute response magnitudes), and reversal by co-infused AP5 (ACSF-Mg + AP5), a specific NMDA-receptor antagonist. n = 12 cells for ACSF alone, 18 cells for Mg++-freeACSF, and 8 cells for Mg++-free ACSF plus AP5.
GABAergic inhibition of LC from PrH In contrast to the potent activation described above from PGi, low-frequency stimulation of the other major afferent to the LC, the PrH, potently inhibited 85% of LC neurons tested (Ennis and Aston-Jones, 1989a, 1989b). Idazoxan was without effect on this response, but picrotoxin (6 mg/kg, iv) consistently attenuated this inhibition. Bicuculline microinfused or microiontophoresed onto LC neurons during PrH stimulation also consistently antagonized this response, whereas locally applied strychnine was ineffective. Taken together, these results indicate a strong GABAergic projection from the PrH to the LC, consistent with the anatomic findings described above (Figs. 5 and 6). In addition to attenuating spontaneous LC discharge, stimulation of this GABAergic input from PrH also potently attenuated excitation of LC neurons evoked by sciatic nerve stimulation (Ennis and Aston-Jones, 1989b) (Fig. 14). Therefore, the PrH is able to significantly attenuate activation of LC neurons by their major excitatory input, the PGi.
65
Effect of activation of preuiously reported afferents on LC discharge The central nucleus of the amygdala (CNA) was previously reported to be the major afferent to the LC (Cedarbaum and Aghajanian, 19781, whereas our anatomical studies demonstrate a dense projection from the CNA to the lateral and rostromedial pericoerulear areas but no input to P G I Stim 15 I
A
I
Q
5 i
the LC proper (Aston-Jones et al., 1986a), consistent with other recent results (Wallace et al., 1989). As LC neurons extend dendrites into the extracoerulear neuropil, especially into the rostromedial pericoerulear zone, we considered the possibility that the CNA could influence LC neurons through contacts onto LC dendrites even though it does not innervate the LC proper. We used electrophysiological methods as an initial means of testing this possibility and discerning the functional impact of CNA on LC discharge (Aston-Jones et al., 1990b). High-intensity (2 mA)
PRE - KYN
1
5 t
0 V
=
P r H Stim
I
Time ( m s e c )
m
10
0
$
5
c
POST-KY N Tlrne ( rnsfc )
15
0
B
V
FS I
f
Time (msec)
m
10
i
oi
15r_
_ . 5 C
P O S T - K Y N 8. I D A
2 U 0
ldh I
n
680
1360
2060
Time ( m s e c )
0 V
l5 T C
Time (msec )
Fig. 13. Local infusion of KYN into LC reveals underlying adrenergic inhibition from PGi. A. PSTH generated during single pulse, electrical stimulation of PGi (at arrow). Activation of PGi yields short latency, potent excitation of this typical LC neuron. B. PGi-evoked excitation of the cell shown in (A) is completely attenuated 1 min after infusion of 0.01 pmol of KYN into LC, as shown in (B) post-drug. Note that blockade excitation by KYN reveals a purely inhibitory response of this neuron to PGi stimulation. Stimulation intensity in (A and B) = 300 pA. C. The inhibition shown in (B) is completely attenuated 5 min after administration of the cr2 receptor antagonist idazoxan (0.2 mg/kg, iv). Similar results are obtained with local microinfusion of idazoxan into the LC. (From Astier and Aston-Jones, 1989.)
E
PrH Stim
& FS
1°1
1111 11111
11111
II
I
Uii I
Time ( r n s e c )
Fig. 14. A. PSTH showing inhibition of discharge of an LC neuron during single pulse electrical stimulation of PrH (300 FA, at arrow). B. PSTH, for the same LC neuron in (A), generated during footpad stimulation (FS). FS (20V, at arrow) yields robust excitation of this LC neuron. C. FS-evoked excitation of the same LC neuron is potently attenuated when FS (20 V) and PrH (300 p A ) stimuli are simultaneously delivered (at arrow). (From Ennis and Aston-Jones, 1989b.l
66
stimulation of CNA elicited only weak and inconsistent synaptic activation in 5/50 LC neurons (6 rats). Five other LC neurons were antidromically activated, as expected, because the LC densely innervates the CNA (unpublished results). In cbntrast, 19/30 neurons in the adjacent parabrachial area exhibited strong, short-latency synaptic activation from CNA stimulation (Aston-Jones et al., 1990b). The weak action of high intensity CNA stimulation on a few LC neurons may reflect a polysynaptic pathway (e.g., the CNA projection to the rostromedial pericoerulear area could innervate local neurons that may contact LC cells). It is also possible that CNA inputs to this pericoerulear area synapse upon distal dendrites of LC neurons; of course, both types of connections also may exist. Electron microscopic analysis is needed to decide conclusively among these possibilities. Similar results were obtained for the NTS (Ennis and Aston-Jones, 1989b) and the medial prefrontal cortex (Chiang et al., 1987). Our anterograde tract-tracing studies confirmed our retrograde transport studies, finding that the NTS strongly innervates the medial parabrachial area just lateral to the LC. Results for NTS stimulation resemble those for CNA stimulation, in that only weak, long-latency synaptic responses were observed in the LC while short-latency, potent excitation was observed in the adjacent parabrachial zone where we find fiber innervation from the NTS (Ennis and Aston-Jones, 1989b). The findings for the prefrontal cortex are similar; as reported by Aston-Jones et al. (this volume), train stimulation of cortex may reveal indirect but functionally important influences on LC activity. Finally, recent studies have found similar results for PAG projections to the LC area. Retrograde and anterograde tracing experiments revealed that the midbrain PAG projects heavily and focally to the rostromedial pericoerulear region. In contrast, the LC proper receives only sparse fibers from PAG (Aston-Jones et al., 1989b; Ennis et al., 1989, 1991; Van Bockstaele et al., 1991). Electrophysiological experiments were
PAG Stim
"1
A
Time (msec)
"1 5
m
--
B
:
Per1- LC
10-
0
1 .-
5
m
50 ~ / U I I U U O l ~ ~ ~ ~ ~ J . I ~ ~ _ 111.11110 ) 1 IL~lUllllllilllllllllJlll~ l~Unll 0 100 300 COO 500 600 Time ( m s e c
1
Fig. 15. A. PSTH showing weak synaptic activation of an LC neuron from PAG stimulation (1000 PA, at arrow). B. The same stimulation site yielded robust activation of neurons in the adjacent rostromedial pericoerulear region, as illustrated for a typical cell (stimulation at arrow, 1000 PA). PSTHs accumulated for 100 sweeps.
consistent with these anatomic results, finding that only 6/100 PAG neurons were antidromically activated from the LC and that most LC neurons were only weakly influenced by PAG stimulation (Fig. 15; Ennis et al., 1991). Overall, the mean activation of LC neurons from the lateral PAG was 32 spikes per 100 stimuli. In contrast, the majority of rostromedial pericoerulear neurons were robustly activated from the same PAG sites, yielding an average of 75 spikes per 100 stimuli (Fig. 15; Ennis et al., 1989, 1991). Latencies for responses in LC or peri-LC were similar, about 5-7 msec. These results indicate that while the PAG influences LC activity, the major impact of PAG on the dorsolateral pontine tegmentum is upon neurons in the rostromedial pericoerulear region rather than on LC neurons. Whether PAG-evoked responses of LC neurons are due to direct synaptic contacts onto LC neurons/dendrites, or are mediated via the
67
more responsive pericoerulear cells; requires further study. The results of the above studies underline the need for combined electrophysiological and electron microscopic studies of inputs to extranuclear dendrites and of possible pericoerulear connections to LC.
Effect3 of 5-HT and NE on LC acticity and responskity Although it has long been appreciated that the LC is densely innervated by 5-HT fibers, and interactions between brain NE and 5-HT systems have been implicated in a variety of clinically relevant phenomena (e.g., sleep-waking cycle control, depression), the effects of 5-HT on discharge of NE-LC neurons has been little investigated. Using microiontophoretic and micropressure techniques, we investigated the effects of directly applied 5-HT and other agents on LC discharge (Chouvet et al., 1988; Aston-Jones et al., 1991). lontophoretic 5-HT had no consistent effect on spontaneous LC activity, decreasing activity of many cells while the discharge of other neurons was unchanged or increased in the presence of this agent. However, similar iontophoretic application of 5-HT consistently and potently attenuated responses of LC neurons to iontophoretic glutamate (GIu). In contrast, 5-HT did not attenuate (and in some cases even potentiated) responses to ACh (Fig. 16). These results were not due to an artifact of iontophoresis as (i) there was no effect of iontophoresis at similar currents through an adjacent barrel containing saline, and (ii) similar effects were found for 5-HT directly applied to LC neurons by micropressure. Additional studies indicated that the 5-HT effects observed in the LC are primarily mediated by 5-HTl;, receptors (Aston-Jones et al., 1990a; Charlety et al., 1990). The 5-HTl, agonists 83 H D P A T and buspirone mimicked the effect of 5-HT by selectively attenuating responses of LC neurons to Glu. Iontophoretic application of the 5-HT,, agonist TFMPP weakly mimicked 5-HT,
NE 1 5 n A
Glu20nAs
s s s
0
s s s s s
s s s
s
- -
5-HT 1 O O n A
AChlOOnA
5-HT 1 0 0 n A
Esxsxs3
I III
NE 1 5 n A
m m
I I m IIII
Fig. 16. Comparison of norepinephrine (NE) and 5-HT effects on responses to glutamate (Glu) and ACh for the same LC neuron. Computer integrated activity-time histograms (5 sec bin width) for activity of a typical LC neuron during microiontophoresis, as indicated. NE applied with a long pulse of low current (at striped bars) inhibits basal activity but leaves responses evoked by Glu (applied at filled circles, upper trace) or ACh (applied at solid bars, lower trace) intact. In contrast, 5-HT applied in long pulses (at open bars) does not affect basal discharge rate, but markedly attenuates responses to Glu (upper trace). Note that although responses to Glu are attenuated by 5-HT, responses to ACh of the same neuron remain intact during application of 5-HT (lower trace). Both traces recorded from the same LC neuron. Calibration bar = 2 min. (From Aston-Jones et al., 1990).
indicating a possible minor 5-HT,, component to this response. In contrast, the 5-HT, agonist DO1 did not mimic 5-HT, and the 5-HT, antagonist ketanserin did not block 5-HT’s effects, leading us to conclude that a 5-HT, receptor is not significantly involved in this response. As the response of LC neurons to sciatic nerve activation appears to be mediated by an EAA pathway from the PGi, it seemed possible that this sensory response of LC neurons would be modulated by 5-HT inputs. Indeed, we have found that pretreatment of animals with parachlorophenylalanine (PCPA) to deplete 5-HT in nerve terminals significantly increased footpad-evoked responses of LC neurons, and that injection of the 5-HT precursor 5-hydroxytryptophan (5-HTP) to increase 5-HT in the brain reduces footpad
68
stimulation-evoked response magnitudes (Shiekhattar and Aston-Jones, unpublished observations). Along with the findings for PCPA administration in naive animals, these results indicate that 5-HT tonically modulates LC sensory responses. However, these systemic injections of PCPA and 5-HTP would be expected to affect 5-HT neurotransmission throughout the nervous system, and so their site of action for affecting LC responsiveness is unclear. To investigate whether direct 5-HT inputs to LC may modulate its sensory responses, we have initiated studies using direct application of 5-HT and 5-HT agonists in combination with footpad stimulation. Local application of 5-HT onto LC somata did not selectively attenuate footpad stimulation responses (Akaoka and Aston-Jones, unpublished results), in contrast to results anticipated by the effects of 5-HT on responses to directly applied Glu, described above. This may reflect inaccessibility of 5-HT applied at the soma to reach distant dendritic synapses mediating responses from PGi, or multiple EAA receptors on LC neurons, some of which (those responsive to footpad stimulation activation) are not sensitive to 5-HT. In contrast to these effects of 5-HT, iontophoresis of NE, or of the a 2 agonist clonidine, with modest currents markedly decreased spontaneous LC discharge but did not attenuate LC responses to either Glu or ACh (Fig. 16) (Chouvet et al., 1988; Aston-Jones et al., 1991); application of these agents with high currents attenuated evoked activity as well. Nonetheless, there was a selective sensitivity of basal discharge to inhibition by a 2 receptor activation. Also, in preliminary studies systemic administration of clonidine or of morphine produced a similar selective inhibition of spontaneous activity while response to footpad (sciatic) stimulation remained intact. When considered in terms of the ratio of phasic, evoked activity to tonic, basal activity of LC neurons (termed response contrast), markedly distinct effects were seen for 5-HT vs. for NE. By selectively attenuating Glu-evoked responses,
5-HT markedly decreased the response contrast of LC neurons for Glu-evoked activity, but did not change the ACh response contrast for these cells. NE, on the other hand, by selectively attenuating basal discharge without affecting responses to either Glu or ACh, markedly augmented response contrasts of LC neurons for both Glu and ACh.
merents to afferents: inputs to the nucleus paragigantocellularis Given that major afferents to LC arise in PGi and PrH, it is clear that knowledge of afferents to these medullary structures is critical to understanding the afferent control of LC. We have begun, therefore, to study the organization of afferents to these two medullary nuclei. Retrograde transport of WGA-HRP or FG, as well as anterograde transport of PHA-L, were used to identify afferents to the PGi (Van Bockstaele et aL, 1989,). As illustrated in Figure 17, major inputs found to the PGi were the caudal medullary reticular formation, raphe magnus area, KollikerFuse/ lateral parabrachialis, NTS, PAG, paraventricular hypothalamic nucleus, and a newly
Fig. 17. Diagram illustrating major afferents to retrofacial PGi in the rat. Abbreviations: amb, nucleus arnbiguus: CC, corpus callosum; CG, central gray; IC, inferior colliculus; KF, Kolliker, Fuse nucleus; LH, lateral hypothalamus: LPb, lateral parabrachial nucleus; LV, lateral ventricle; MdD, caudal medullary reticular formation dorsal to the lateral reticular nucleus; NTS, nucleus tractus solitarius; PVN, paraventricular nucleus of the hypothalamus; PGi, nucleus paragigantocellularis; RPa, raphe pallidus; V11, VIIth nerve nucleus; 4V, fourth ventricle.
69
described but very prominent input from the supraoculomotor nucleus of the central grey. Several of these afferents exhibited clear topography within the PGi. As seen with anterograde transport of PHA-L, for example, PAG projects primarily to the medial PGi while the NTS and the supraoculomotor nucleus project to specific, but distinctly different, areas of the PGi (Van Bockstaele et al., 1991). These results indicate that afferents to the PGi are diverse and located throughout the brainstem and spinal cord, but are generally associated with autonomic and integrative functions. Neurons that project to the LC are distributed diffusely, though with some topography, in PGi (Fig. 2). These results, together with the topography of afferent projections to PGi, indicate that all, or only a select few, of the afferents to the PGi may synapse upon LC-projecting neurons. At present we do not know which extrinsic or intrinsic afferents to PGi or PrH synapse upon neurons that project to LC. This is an important issue because knowing which inputs to the PGi or PrH contact LC-projecting cells will reveal functional circuits that most directly influence the LC system. Experiments to answer these questions are planned for both the PGi and PrH. Conclusions
Studies over the last 5 years have yielded a new perspective on the afferent regulation of the noradrenergic brain nucleus LC. Major afferents to the LC are found in two rostral medullary regions, the medial, perifascicular PrH in the dorsomedial medulla, and the PGi in the rostral ventral medulla. These major afferents utilize an EAA(s) (PGi) or GABA (PrH) as the most potent neurotransmitter agents in their projections to the LC. Thus, PGi activation predominantly excites LC neurons through a non-NMDA receptor-sensitive mechanism, while PrH potently inhibits LC neurons by a GABA, receptor in the LC.
However, the LC receives inputs from a variety of neurotransmitter systems, indicating that its afferent organization is more complex than such a synopsis might suggest. Indeed, we find that multiple neurotransmitters impinge on the LC from these two major afferent nuclei. Within the PGi, for example, there are not only EAA projections to the LC, but also prominent adrenergic inputs. When activation by EAAs is blocked pharmacologically, inhibition from these adrenergic projections becomes apparent. We also have recently obtained preliminary evidence for GABA and CRF inputs to the LC from the rostral ventrolateral medulla. These findings, taken together with other results reviewed here, indicate that functionally distinct subpopulations of PGi neurons may innervate the LC: (i> physiologically distinguishable subpopulations of PGi neurons are antidromically activated from the LC, and one of these subpopulations is predominantly located in the ventromedial aspect of PGi, (ii) there exist at least three distinct pathways for projections from the PGi to the LC, and (iii) projections from the PGi to the LC, and afferents to the PGi, are topographically distributed within the PGi. In addition, sciaticevoked LC activation appears to be preferentially mediated by neurons in the ventromedial PGi (Chiang and Aston-Jones, 1989; reviewed in AstonJones et al., this volume). The possibility of functionally distinct subpopulations of afferents to the LC from the PrH is presently less clear. As the PGi and the PrH are prominent in regulating LC activity, functions previously ascribed to these two medullary areas provide important insights into the functions of the NE-LC system. The PrH has been extensively studied in cats and primates as a preoculomotor nucleus, and many of its constituent neurons are important in the control of eye movements (Baker, 1977; McCrea et al., 1979). The close correspondence of the LC with increased attention to environmental stimuli and orienting behaviors (AstonJones and Bloom, 1981a,b) may derive in part from such oculomotor circuitry, as eye move-
70
ments are an important part of behavioral orienting responses that accompany increased attentiveness to external stimuli. The PGi area has been linked to cardiovascular, nociceptive and respiratory functions (Guyenet and Les Brown, 1986; Morrison et al., 1988; Sun et al., 1988). In particular, activation of this area broadly increases activity in the peripheral sympathetic nervous system (Ross et al., 1984). The strong excitatory input from this same area to the LC provides a neurobiological substrate for the fact that sympathoexcitatory stimuli are also very effective in activation of LC neurons (Elam et al., 1984, 1986), and is strongly supportive of the proposal that the NE-LC system serves as the cognitive limb of a globally conceived sympathetic nervous system (Aston-Jones, 1985; Aston-Jones et al., 1990b). The functional implications of the PGi and PrH as afferents to the LC, particularly as related to activity of LC neurons in behaving animals and the role of the noradrenergic LC system in vigilance, are discussed in more detail in Aston-Jones et al. (this volume). Important studies for the future include further examination of the role that extranuclear dendrites of LC neurons play in differential regulation of LC activity, and the extent and functional importance of LC innervation by ‘‘local’’ pericoerulear neurons. Sources of, and functional interactions between, the multitude of transrnitters that innervate the LC are also important aspects of our current and future research. Together, recent results concerning afferent regulation of the LC are revealing the neurobiological substrates by which activity in the globally projecting NE-LC system is regulated. These results lay the groundwork for further experimentation to provide a complete cellular anatomic, physiological and pharmacological understanding of the control of activity in these neurons. Such a cellular understanding is necessary for a comprehensive input-output analysis of the NE-LC system, which is essential to understanding the system’s function at a cellular and systems level.
Acknowledgements
This work was supported by PHS grants NS24698, NS20463, HL08097, DA06214, MH40008, MH42796, O N R contract N00014-86-0493, AFOSR grant 90-0147, the Phillippe Foundation, the Simone et Cino del Duca Foundation, INSERM, the Fonds de la Recherche en Sant6 du QuCbec. References Aghajanian, G.K. (1978) Feedback regulation of central monoaminergic neurons: Evidence from single-cell recording studies. In M.B.H. Youdim, W. Lnvenberg, D.F. Sharman and J.R. Lagnado (Eds.), Neurochemisrry und Neuropharmacology, Wiley, New York, pp. 1-32. Altschuler, R., Fex, J., Parakkal, M. and Eckenstein, F. (1984) Colocalization of enkephalin-like and ChAT-like immunoreactivities in olivocochlear neurons of the guinea pig. J. Histochem. Cytochem., 32: 839-843. Arnsten, A.F. and Goldman, R.P. (1984) Selective prefrontal cortical projections to the region of the locus coeruleus and raphe nuclei in the rhesus monkey. Brain Res., 306: 9-18. Astier, B. and Aston-Jones, G. (1989) Electrophysiological evidence for medullary adrenergic inhibition of rat locus coeruleus. Soc. Neurosci. Abstr., 15: 1012. Astier, B., Kitahama, K., Denoroy, L., BLrod, A., Jouvet, M. and Renaud, B. (1986) Biochemical evidence for an interaction between adrenaline and noradrenaline neurons in the rat brainstem. Brain Res., 397: 333-340. Astier, B., Kitahama, K., Denoroy, L., Jouvet, M. and Renaud, B. (1987) Immunohistochemical evidence for the adrenergic medullary longitudinal bundle as a major ascending pathway to the locus coeruleus. Neirrosci. Lett., 74: 132-138. Astier, B., Van Bockstaele, E.J., Aston-Jones, G . and Pierihone, V.A. (1990) Anatomic evidence for multiple pathways leading from the nucleus paragigantocellularis to the locus coeruleus in the rat. Neurosci. Lett. (In press). Aston-Jones, G. (1 985) Behavioral functions of locus coeruleus derived from cellular attributes. Physiol. Psych., 13: 118126. Aston-Jones, G. and Bloom, F.E. (1981a) Activity of norepinephrine-containing locus coeruleus neurons in hehaving rats anticipates fluctuations in the sleep-waking cycle. .I. Neurosci., 1: 876-86. Aston-Jones, G. and Bloom, F.E. (1981b) Norepinephrinecontaining locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900.
71 Aston-Jones, G. and Ennis, M. (1988) Sensory-evoked activation of locus coeruleus may be mediated by a glutamate pathway from the rostra1 ventrolateral medulla. In A. Cavalheiro, J. Lehmann and L. Turski (Eds.), Frontiers in Excitatory Amino Acid Research, A.R. Liss, Inc., New York. pp. 471-478. Aston-Jones, G., Foote, S.L. and Bloom, F.E. (1984) Anatomy and physiology of locus coeruleus neurons: Functional implications. In M. Ziegler and C.R. Lake (Eds.), Norepinephrine: Frontiers of Clinical Neuroscience, Vol. 2, Williams and Wilkins, Baltimore, pp. 92-1 16. Aston-Jones, G., Ennis, M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986a) The brain nucleus locus coeruleus: Restricted afferent control of a broad efferent network. Science, 234: 734-737. Aston-Jones, G., Shipley, M.T., Nickell, W.T., Ennis, M. and Pieribone, V. (l986b) Afferents to locus coeruleus are largely restricted to two medullary nuclei: Anatomic and physiologic studies. Soc. Neurosci. Abstr., 12: 138. Aston-Jones, G., Van Bockstaele, E., Pieribone, V. and Shipley, V. (1989) Topographic projections from the periaqueductal gray (PAG) to the ventrolateral medulla (VLM) in the rat. Soc. Neurosci. Ahsrr., 15: 593. Aston-Jones, G., CharlCty, P., Akaoka, H., Shiekhattar, R. and Chouvet, G. (1990a) Serotonin acts at 5-HTla receptors to selectively attenuate glutamate-evoked responses of locus coeruleus neurons. Soc. Neurosci. Ahstr., 16: 799. Aston-Jones, G., Shipley, M.T., Ennis, M., Williams, J.T. and Pieribone, V.A. (1990h) Restricted afferent control of locus coeruleus neurons revealed by anatomic, physiologic and pharmacologic studies. In C.A. Marsden and D.J. Heal (Eds.), The Pharmacology of Noradrenaline in the Central Nervous System, Oxford University Press, G.B., Oxford, pp. 187-247. Aston-Jones, G., Akaoka, H., CharlCty, P. and Chouvet, G. (1991) Serotonin selectively attenuates glutamate-evoked activation of locus coeruleus neurons in ciuo. J. Neurosci., 11: 760-769. Baker, R. (1977) The nucleus prepositus. In B. Brooks and F. Bajandas (Eds.), Eye Movements, Plenum Press, New York, pp. 145-178. Beitz, A.J. (1982) The sites of origin of brainstem neurotensin and serotonin projections to the rodent nucleus raphe magnus. J. Neurosci., 2 (7): 829-842. BCrod, A,, Chat, M., Paut, L. and Tappaz, M. (1984) Catecholaminergic and GABAergic anatomical relationship in the rat substantia nigra, locus coeruleus, and hypothalamic median eminence: Immunocytochemical visualization of biosynthetic enzymes on serial semithin plastic-embedded sections. J. Histochem. Cytochem., 32: 1331-1338. Blier, P. and de Montigny, C. (1987) Modification of 5-HT neuron properties by su ned administration of the 5H T l A agonist gepirone: Electrophysiological studies in the rat brain. Synapse, 1: 470-480. Bloom, R.E., Battenberg, E.L.F., Rivier, J. and Vale, W. (1982) Corticotropin releasing factor (CRF): Immunoreactive neurones and fibers in rat hypothalamus. Reg. Peptides, 4: 43-48.
Bobillier, P., Seguin, S., Degueurce, A., Lewis, B.D. and Pujol, J.F. (1979) The efferent connections of the nucleus raphe centralis superior in the rat as revealed by autoradiography. Bruin Res., 166: 1-8. Bowker, R.M., Steinbusch, H.W.M. and Coulter, J.D. (1981) Serotonergic and peptidergic projections to the spinal cord demonstrated by a combined retrograde H R P histochemical and immunocytochemical staining method. Bruin Res., 211: 412-417. Butcher, L. and Woolf, N.J. (1984) Histochemical distribution of acetylcholinesterase in the central nervous system: Clues to the localization of cholinergic neurons. In A. Bjorklund, T. Hokfelt and M.J. Kuhar (Eds.), Classical Transmitters and Transmitter Receptors in the CNS. Purt I I , Elsevier Science Publ. B.V., Amsterdam, pp. 1-45. Cassini, P., Ho, R.H. and Martin, G.F. (1989) The brainstem origin of enkephalin- and substance-P-like immunoreactive axons in the spinal cord of the North American opossum. Brain Behau. Evol., 34: 212-22. Cechetto, D.F., Standaert, D.G. and Saper, C.B. (1985) Spinal and trigeminal dorsal horn projections to the parabrachial nucleus in the rat. J. Comp. Neurol., 240: 153-160. Cedarbaum, J.M. and Aghajanian, G.K. (1978) Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique. J. Comp. Neurol., 178: 1-16. CharlCty, P.J., Akaoka, H., Aston-Jones, G. and Chouvet, G. (1990) In viuo pharmacological characterization of the serotonin receptors involved in the interaction with excitatory amino acids in the nucleus locus coeruleus of the rat. Eur. J. Pharm. Suppl., 3: 30. Charnay, Y., LCger, L., R o s i e r , J., Jouvet, M. and Dubois, P.M. (1985) Evidence for synenkephalin-like immunoreactivity in pontobulbar monoaminergic neurons of the cat. Brain Res, 335: 160-4. Chen, Z. and Engberg, G. (1989) The rat nucleus paragigantocellularis as a relay station to mediate peripherally induced central effects of nicotine. Neurosci. Lett., 101: 67-71. Chiang, C. and Aston-Jones, G. (1989) Microinjection of lidocaine, GABA or synaptic decouplers into the ventrolateral medulla blocks sciatic-evoked activation of locus coeruleus. Soc. Neurosci. Abstr., 15: 1012. Chiang, C., Ennis, M., Pieribone, V. and Aston-Jones, G. (1987) Effects of prefrontal cortex stimulation on locus coeruleus discharge. Soc. Neurosci. Abstr., 13: 912. Chouvet, G., Akaoka, H. and Aston-Jones, G. (1988) Serotonin selectively decreases responses of locus coeruleus neurons to glutamate. C.R. Acud. Sci. Paris, 290: D807D810. Cintra, L., Diaz-Cintra, S., Kemper, T. and Morgane, P.J. (1982) Nucleus locus coeruleus: A morphometric golgi study in rats of three age groups. Brain Res., 247: 17-28. Clavier, R.M. (1979) Afferent projections to the self-stimulation regions of the dorsal pons, including the locus coeruleus, in the rat as demonstrated by the horseradish peroxidase technique. Brain Res. Bull., 4: 497-504. Conrad, L.C.A., Leonard, C.M. and Pfaff, D.W. (1974) Connections of the median and dorsal raphe nuclei in the rat:
72 An autoradiographic and degeneration study. J. Comp. Neurol., 156: 179-206. Conrath-Varier, M., Dietl, M., Arluison, M., Cesselin, F., Bourgoin, S. and Hamon, M. (1983) Localization of Metenkephalin-like immunoreactivity within pain-related nuclei of cervical spinal cord, brainstem and midbrain in the cat. Bruin Res. Bull., 11: 587-604. Cummings, E., Elde, R., Ells, J. and Lindall, A. (1983) Corticotropin-releasing factor immunoreactivity is widely distributed within the central nervous system of the rat: An immunohistochemical study. J. Neurosci., 3: 1355-1368. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Scand., 62: 5-55. Deutch, A.Y., Goldstein, M. and Roth, R.H. (1986) Activation of the locus coeruleus induced by selective stimulation of the ventral tegmental area. Bruin Res., 363: 307-314. Drolet, G., Akaoka, H., Van Bockstaele, E.J., Aston-Jones, G. and Shipley, M.T. (1990) Opioid afferents to the locus coeruleus from the rostral medulla as detected by retrograde transport combined with immunohistochemistry. Soc. Neurosci. Abstr., 16: 1027. Eiden, L.E., Nilaver, G. and Palkovits, M. (1982) Distribution of vasoactive intestinal polypeptide (VIP) in the rat brain stem nuclei. Brain Res., 231: 472-7. Elam, M., Yao, T., Svensson, T.H. and ThorCn, P. (1984) Regulation of locus coeruleus neurons and splanchnic, sympathetic nerves by cardiovascular afferents. Bruin Res., 290: 281-7. Elam, M., Svensson, T.H. and ThorCn, P. (1986) Locus coeruleus neurons and sympathetic nerves: Activation by cutaneous sensory afferents. Bruin Res., 366: 254-261. Engberg, G. (1989) Nicotine induced excitation of locus coeruleus neurons is mediated via release of excitatory amino acids. Life Sci., 44: 1535-40. Ennis, M. and Aston-Jones, G. (1986) A potent excitatory input to the nucleus locus coeruleus from the ventrolateral medulla. Neurosci. Lett., 71: 299-305. Ennis, M. and Aston-Jones, G. (1987) Two physiologically distinct populations of neurons in the ventrolateral medulla innervate the locus coeruleus. Brain Res., 425: 275-282. Ennis, M. and Aston-Jones, G. (1988) Activation of locus coeruleus from nucleus paragigantocellularis: A new excitatory amino acid pathway in brain. J. Neurosci., 8: 36443657. Ennis, M. and Aston-Jones, G. (1989a) GABA-mediated inhibition of locus coeruleus from the dorsomedial rostral medulla. J. Neurosci., 9: 2973-2981. Ennis, M. and Aston-Jones, G. (1989b) Potent inhibitory input to locus coeruleus from the nucleus prepositus hypoglossi. Brain Res. Bull., 22: 793-803. Ennis, M., Behbehani, M., Van Bockstaele,.E.J., Aston-Jones, G. and Shipley, M.T. (1989) Influence of periaqueductal grey on rat locus coeruleus neurons. Soc. Neurosci. Abstr., 15: 1013. Ennis, M., Behbehani, M.M., Van Bockstaele, E.J., M.T., S. and Aston-Jones, G. (1991) Projections from the periaque-
ductal gray to nucleus locus coeruleus and the pericoerulear region: Anatomic and physiologic studies. J. Comp. Neurol., 306: 480-494. Fallon, J.H. and Leslie, F.M. (1986) Distribution of dynorphin and enkephalin peptides in the rat brain. J. Comp. Neurol., 249: 293-336. Finley, J.C.W., Maderdrut, J.L. and Petrusz, P. (1981) The immunocytochemical localization of enkephalin in the central nervous system of the rat. J. Comp. Neurol., 198: 541-565. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Reu., 63: 844-914. Fu, L., Shipley, M., Aston-Jones, G., Chiang, C. and Pieribone, V. (1988) Architecture of the pericoeruliar region in the rat. Soc. Neurosci. Abstr., 14: 404. Fu, L., Shipley, M.T. and Aston-Jones, G. (1989) Dendrites of rat locus coeruleus are asymmetrically distributed: Immunocytochemical LM and EM studies. SOC. Neurosci. Abstr., 15: 1013. Gallyas, F., Gorcs, T. and Merchenthaler, I. (1982) High-grade intensification of the end-product of the diaminobenzidine reaction for peroxidase histochemistry. J. Histochem. Cytochem., 30: 183-184. Granata, A.R., Ruggiero, D.A., Park, D.H., Joh, T.H. and Reis, D.J. (1985) Brain stem area with C1 epinephrine neurons mediates baroreflex vasodepressor responses. Am. J. Physiol., 248: H247-H567. Grenhoff, J., Tung, C.S. and Svensson, T.H. (1988) The excitatory amino acid antagonist kynurenate induces pacemaker-like firing of dopamine neurons in rat ventral tegmental area in Lii'o. Acta Physiol. Scand., 134: 567-8. Grzanna, R. and Molliver, M.E. (1980) The locus coeruleus in the rat: An immunohistochemical delineation. Neuroscience, 5: 21-40. Grzanna, R., Molliver, M.E. and Coyle, J.T. (1980) Visualization of central noradrenergic neurons in thick sections by the unlabeled antibody method: A transmitter-specific Golgi image. Proc. Nutl. Acud. Sci. USA, 75: 2502-2506. Guthrie, J. and Basbaum, A.I. (1984) Colocalization of immunoreactive proenkephalin and prodynorphin products in medullary neurons of the rat. Neuropeptzdes, 4: 437-445. Guyenet, P.G. and Les Brown, D. (1986) Nucleus paragigantocellularis lateralis and lumbar sympathetic discharge in the rat. Am. J. Physiol., 250: R1981-R1094. Guyenet, P.G. and Young, B.S. (1987) Projections of nucleus paragigantocellularis lateralis to locus coeruleus and other structures in rat. Bruin Res., 406: 171-84. Hajos, M. and Engberg, G. (1988) Role of primary sensory neurons in the central effects of nicotine. Psychophurmucology (Berlin), 94: 468-70. Haselton, J.R. and Guyenet, P.G. (1987) PNMT neurons of the C1 cell group with projections to both locus coeruleus and spinal cord. Soc. Neurosci. Abstr., 13: 809. Hokfelt, T., Fuxe, K., Goldstein, M. and Johansson, 0. (1974) Immunohistochemical evidence for the existence of adrenaline neurons in the rat brain. Bruin Res., 66: 235251.
73 Hokfelt, T., Ljungdahl, A,, Terenius, L., Elde, R. and Nisson, G. (1977) Immunohistochemical analysis of peptide pathways possibly related to pain and analgesia: Enkephalin and substance P. Proc. Natl. Acad. Sci. USA, 74: 30813085. Hokfelt, T., Ljungdahl, A., Steinbusch, H., Verhofstad, A., Nilsson, G., Brodin, E., Pernow, B. and Goldstein, M. (1978) lmmunohistochemical evidence of substance P-like immunoreactivity in some 5-hydroxytryptamine-containing neurons in the rat central nervous system. Neuroscience, 3: 517-538. Hokfelt, T., Terenius, L., Kuypers, H.G.J.M. and Dann, 0. (1979) Evidence for enkephalin immunoreactive neurons in the medulla oblongata projecting to the spinal cord. Neurosci. Lett., 14: 55-60. Hokfelt, T., Johansson, 0. and Goldstein, M. (1985) Central catecholamine neurons as revealed by immunohistochemistry with special reference to adrenaline neurons. In A. Bjoklund and T. Hokfelt (Eds.), Handbook of Chemical Neuroanatomy. Volume 2: Classical Transmitters in the CNS, Elsevier Science, New York, pp. 157-276. Hunt, S.P. and Lovick, T.A. (1982) The distribution of serotonin, met-enkephalin and beta-lipotropin-like immunoreactivity in neuronal perikarya of the cat brainstem. Neurosci. Lett., 30(2): 139-45. Jennes, L., Stumpf, W.E. and Kalivas, P.W. (1982) Neurotensin: Topographical distribution in rat brain by immunohistochemistry. J. Comp. Neurol., 210: 21 1-24. Johansson, O., Hokfelt, T. and Elde, R.P. (1984) Immunohistochemical distribution of somatostatin-like immunoreactivity in the central nervous system of the adult rat. Neuroscience, 13: 265-339. Kalia, M., Fuxe, K. and Goldstein, M. (1985) Rat medulla oblongata. 111. adrenergic (C1 and C2) neurons, nerve fibers and presumptive terminal processes. J. Comp. Neurol.. 233: 333-349. Khachaturian, H., Lewis, M. and Watson, S.J. (1983) Enkephalin systems in the diencephalon and brainstem of the rat. J. Comp. Neurol., 220: 310-320. Kimura, H., McMeer, P.L. and Peng, J.-H. (1984) Choline acetyltransferase-containing neurons in the rat brain. In A. Bjorklund, T. Hokfelt and M.J. Kuhar (Ed.), Classical Transmitters and Transmitter Receptors in the CNS. Part II, Elsevier Science Publishers B.V., Amsterdam, pp. 51-65. Ljungdahl, A,, Hokfelt, T. and Nilsson, G. (1978) Distribution of substance P-like immunoreactivity in the central nervous system of the rat I. Cell bodies and nerve terminals. Neuroscience, 3: 63-80. Lynch, D.R., Strittmatter, S.M. and Snyder, S.H. (1984) Enkephalin convertase localization by [3H]guanidinoethylmercaptosuccinic acid autoradiography: Selective association with enkephalin-containing neurons. Proc. Natl. Acad. Sci. USA, 81: 6543-7. Mantyh, P.W. and Hunt, S.P. (1984) Neuropeptides are present in projection neurones at all levels in visceral and taste pathways: From periphery to sensory cortex. Brain Res., 299: 297-311. Martin, J.L., Dietl, M.M., Hof, P.R., Palacios, J.M. and Magistretti, P.J. (1987) Autoradiographic mapping of
[mono[l25I]iodo-TyrlO, MetOl7~asoactiveintestinal peptide binding sites in the rat brain. Neuroscience, 23: 539-65. McCrea, R.A., Baker, R. and Delgado-Garcia, J. (1979) Afferent and efferent organization of the prepositus hypoglossi nucleus. Prog. Brain Res., 50: 653-665. McMahon, S.B. and Wall, P.D. (1985) Electrophysiological mapping of the brainstem projections of spinal cord lamina I cells in the rat. Brain Res., 333: 19-26. Merchenthaler, I. (1984) Corticotropin releasing factor (CRF)-like immunoreactivity in the rat central nervous system. extrahypothalamic distribution. Peptides, 5: 53-69. Merchenthaler, I., Vigh, S., Petrusz, P. and Schally, A.V. (1982) Immunocytochemical localization of corticotropinreleasing factor (CRF) in the rat brain. Am. J. Anat., 165: 386-396. Miller, R.J. and Pickel, V.M. (1980a) The distribution and functions of the enkephalins. J. Histochem. Cytochem., 28: 903-17. Miller, R.J. and Pickel, V.M. (1980b) Immunohistochemical distribution of enkephalins: Interactions with catecholamine-containing systems. Adu. Biochem. Psychopharrnacol., 25: 349-59. Minagawa, H., Shiosaka, S., Inagaki, S., Sakanaka, M., Takatsuki, K., Ishimoto, I., Senba, E., Kawai, Y., Hard, Y., Matsuzaki, T. and Tohyama, M. (1983) Ontogeny of neurotensin-containing neuron system of the rat: Immunohistochemical analysis-11. Lower brainstem. Neuroscience, 8 (3): 467-486. Morgane, P.J. and Jacobs, M.S. (1979) Raphe projections to the locus coeruleus in the rat. Brain Res. Bull., 4: 519-34. Morrison, S.F., Milner, T.A. and Reis, D.J. (1988) Reticulospinal vasomotor neurons of the rat rostra1 ventrolateral medulla: Relationship to sympathetic nerve activity and the C1 adrenergic cell group. J. Neurosci., 8: 1286-1301. Mugnaini, E. and Oertel, W.H. (1985) An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry. In A. Bjorklund and T. Hokfelt (Eds.), Handbook of Chemical Neuroanatomy, Elsevier Science Publishers B.V., Amsterdam, pp. 436-608. Nomura, H., Shiosaka, S., Inagaki, S., Ishimoto, I., Senba, E., Sakanaka, M., Tekatsuki, K., Matsuzaki, T., Kubota, Y., Saito, H., Takase, S., Kogure, K. and Tohyama, M. (1982) Distribution of substance P-like immunoreactivity in the lower brainstem of the human fetus: An immunohistochemical study. Brain Res., 252 (2): 315-25. Olshcowka, J.A., O’Donohue, T.L., Mueller, G.P. and Jacobowitz, D.M. (1982) The distribution of corticotropin releasing factor-like immunoreactive neurons in rat brain. Peptides, 3: 995-1015. Ottersen, O.P. and Storm-Mathisen, J. (1984a) Glutamateand GABA-containing neurons in the mouse and rat brain, as demonstrated with a new immunocytochemical technique. J. Comp. Neurol., 229: 374-392. Ottersen, O.P. and Storm-Mathisen, J. (1984b) Neurons containing or accumulating transmitter amino acids. In A. Bjorklund, T. Hokfelt and M.J. Kuhar (Ed.), Classical Transmitters and Transmitter Receptors in the CNS. Part II, Elsevier Science Publishers B.V., Amsterdam, pp. 141-224.
74 Papadopoulos, G.C., Karamanlidis, A.N., Antonopoulos. J. and Dinopoulos, A. (1986) Neurotensin-like immunoreactive neurons in the hedgehog (Erinaceus europaeus) and the sheep (Ovis aries) central nervous system. J. Comp. Neurol., 244: 193-203. Pieribone, V.A. and Aston-Jones, G. (1991) Adrenergic innervation of the rat nucleus locus coeruleus arises from the C1 and C3 cell groups in the rostral medulla: An anatomic study combining retrograde transport and immunofluorescence . Neuroscience, 4 1 5 25 -542. Pieribone, V.A., Aston-Jones, G. and Bohn, M.C. (1988) Adrenergic and non-adrenergic neurons of the C1 and C3 areas project to locus coeruleus: A fluorescent double labeling study. Neurosci. Lett., 85: 297-303. Pieribone, V., Van Bockstaele, E., Shipley, M. and AstonJones, G. (1989) Serotonergic innervation of rat locus coeruleus derives from non-raphe brain areas. Soc. Neurosci. Abstr., 15: 420. Pieribone, V.A., Shipley, M.T., Ennis, M. and Aston-Jones, G. (1990) Anatomic evidence for GABA-ergic afferents to the rat locus coeruleus in the dorsal medial medulla: A immunocytochemical and retrograde transport study. Soc. Neurosci. Abstr., 16: 300. Rasmussen, K. and Aghajanian, G.K. (1989) Withdrawal-induced activation of locus coeruleus neurons in opiate-dependent rats: Attenuation by lesions of the nucleus paragigantocellularis. Brain Res, 505: 346-50. Ross, C., Armstrong, D., Ruggiero, D., Pickel, V., Joh, T. and Reis, D. (1981) Adrenaline neurons in the rostral ventrolateral medulla innervate thoracic spinal cord: A combined immunocytochemical and retrograde transport demonstration. Neurosci. Lett., 25: 257-262. Ross, C.A., Ruggiero, D.A., Park, D.H., Joh, T.H., Sved, A.F., Fernandez-Pardal, J., Saavedra, J.M. and Reis, D.J. (1984) Tonic vasomotor control by the rostral ventrolateral medulla: Effect of electrical or chemical stimulation of the area containing C1 adrenaline neurons on arterial pressure, heart rate, and plasma catecholamines and vasopressin. J. Neurosci., 4: 474-494. Ruggiero, D.A., Ross, C.A., Anwar. M., Park, D.H., Joh, T.H. and Reis, D.J. (1985) Distribution of neurons containing phenylethanolamine N-methyltransferase in medulla and hypothalamus of rat. J. Comp. Neurol., 239: 127-154. Ruggiero, D.A., Giuliano, R., Anwar, M., Stornetta, R. and Reis, D.J. (1990) Anatomical substrates of cholinergic-autonomic regulation in the rat. J. Comp. Neurol., 292: 1-53. Sakanaka, M., Shibasaki, T. and Lederis, K. (1987) Corticotropin releasing factor-like immunoreactivity in the rat brain as revealed by a modified cobalt-glucose oxidase-diaminobenzidine method. J. Comp. Neurol., 260: 256-298. Saper, C.B. (1982) Reciprocal parabrachial-cortical connections in the rat. Brain Res., 242: 33-40. Saper, C.B. (1987) Function of the locus coeruleus. TINS, 10: 343-344. Sar, M., Stumpf, W.E., Miller, R.J., Chang, K.J. and Cuatrecasas, P. (1978) Immunohistochemical localization of enkephalin in rat brain and spinal cord. J. Comp. Neurol., 182: 17-37.
Sesack, S.R., Deutch, A.Y., Roth, R.H. and Bunney. B.S. (1989) Topographical organization of the efferent projections of the medial prefrontal cortex in the rat: An anterograde tract-tracing study with phaseolus vulgaris leucoagglutinin. J. Comp. Neurol., 290: 213-242. Shiekhattar, R. and Aston-Jones, G. (1991) NMDA-receptormediated responses of brain noradrenergic neurones are suppressed by extracellular magnesium. Synapse. (in press). Shiekhattar, R., Chiang, C. and Aston-Jones, G. (1991) Regulation of locus coeruleus discharge by extracellular calcium ion concentration. Synapse, (in press). Shimizu, N. and Imamoto, K. (1970) Fine structure of the locus coeruleus in the rat. Arch. Histol. Jap., 31: 229-246. Shimizu, N., Ohnishi, S., Satoh, K. and Tohyama, M. (1978) Cellular organization of locus coeruleus in the rat as studied by Golgi method. Arch. Histol. Jap., 41: 103-1 12. Shipley, M., Pieribone, V., Aston-Jones, G. and Ennis, M. (1988) GABA-ergic innervation of the rat locus coeruleus. Soc. Neurosci. Ahstr., 14: 406. Standaert, D.G., Watson, S.J., Houghten, R.A. and Saper, C.B. (1986) Opioid peptide immunoreactivity in spinal and trigeminal dorsal horn neurons projecting to the parabrachial nucleus in the rat. J. Neurosci., 6: 1220-6. Steinbusch, H.W.M. (1984) Serotonin-immunoreactive neurons and their targets in the CNS. In A. Bjorklund, T. Hokfelt and M.J. Kuhar (Eds.), Classical Transmitters and Transmitter Receptors in the CNS. Part 11, Elsevier Science Publishers B.V., Amsterdam, pp. 68-1 18. Sun, M., Hackett, J.T. and Guyenet, P.G. (1988) Sympathoexcitatory neurons of rostral ventrolateral medulla exhibit pacemaker properties in the presence of a glutamate-receptor antagonist. Bruin Res., 438: 23-40. Sutin, E.L. and Jacobowitz, D.M. (1988) Immunocytochemical localization of peptides and other neurochemicals in the rat laterodorsal tegmental nucleus and adjacent area. J. Comp. Neurol., 270: 243-70. Svensson, T.H., Engberg, G., Tung, C.S. and Grenhoff, J. (1989) Pacemaker-like firing of noradrenergic locus coeruleus neurons in r i m induced by the excitatory amino acid antagonist kynurenate in the rat. Acta fhysiol. Scand., 135: 421-2. Swanson, L.W. (1976) The locus coeruleus: A cytoarchitectonic, golgi and immunohistochemical study in the albino rat. Brain Res., 110: 39-56. Swanson, L.W., Sawchenko, P.E., Rivier, J. and Vale, W.W. (1983) Organization of ovine corticotropin-releasing factor immunoreactive cells and fibers in the rat brain: An immunohistochemical study. Neuroendocrinology, 36: 165186. Thor, K.B. and Helke, C.J. (1988) Catecholamine-synthesizing neuronal projections to the nucleus tractus solitarii of the rat. J. Comp. Neurol., 268: 264-280. Triepel, J., Mader, J., Weindl, A,, Heinrich, D., Forssmann, W.G. and Metz, J. (1984) Distribution of NT-IR perikarya in the brain of the guinea pig with special reference to cardiovascular centers in the medulla oblongata. Histochemistry, 81: 509-516. Triepel, J., Weindl, A,, Kiemle, I., Mader, J., Volz, H.P.,
75 Reinecke, M. and Forssmann, W.G. (1985) Substance Pimmunoreactive neurons in the brainstem of the cat related to cardiovascular centers. Cell Tissue Res., 241: 31 -41. Tucker, D.C., Saper, C.B., Ruggiero, D.A. and Reis, D.J. (1987) Organization of central adrenergic pathways: I. Relationships of ventrolateral medullary projections to the hypothalamus and spinal cord. J . Comp. Neurol., 259: 59 1-603. Tung, C.S., Ugedo, L., Grenhoff, J., Engberg, G. and Svensson, T.H. (1989) Peripheral induction of burst firing in locus coeruleus neurons by nicotine mediated via excitatory amino acids. Synapse, 4: 313-8. Uhl, G.R., Goodman, R.R., Kuhar, M.J., Childers, S.R. and Snyder, S.H. (1979) Immunohistochemical mapping of enkephalin containing cell bodies. fibers and nerve terminals in the brainstem of the rat. Bruin Res., 166: 75-94. Valentino, R.J., Van Bockstaele, E.J. and Aston-Jones, G. (1990) Corticotropin-releasing factor-immunoreactive (CRF-IR) neurons are localized in nuclei which project to the locus coeruleus (LC). Soc. Neurosci. Abstr., 16: 519. Van Bockstaele, E.J., Pieribone, V.A. and Aston-Jones, G. (1989a) Diverse afferents converge on the nucleus paragigantocellularis in the rat ventrolateral medulla: Retrograde and anterograde tracing studies. J. Cornp. Neurol., 290: 561-584.
Van Bockstaele, E., Pieribone, V., Aston-Jones, G. and Shipley, M. (1989b) Multiple projection pathways from the ventrolateral medulla to locus coeruleus in rat. Soc. Neurosci. Abstr., 15: 1013. Van Bockstaele, E.J., Aston-Jones. G., Ennis, M. Shipley, M.T., and Pieribone, V.A. (1991) Subregions of the periaqueductal gray topographically innervate the rostra1 ventral medulla in the rat. J. Cornp. Neurol., 309: 1-23. Vincent, S.R., McIntosh, C.H., Buchan, A.M. and Brown, J.C. (1985) Central somatostatin systems revealed with monoclonal antibodies. J. Cornp. Neurol., 238(2): 169-186. Wallace, D.M., Magnuson, D.J. and Gray, T.S. (1989) The amygdala-brainstem pathway: Selective innervation of dopaminergic, noradrenergic and adrenergic cells in the rat. Neurosci. Lett., 97: 252-8. Wang, Y. and Aghajanian, G.K. (1989) Excitation of locus coeruleus neurons by vasoactive intestinal peptide: Evidence for a G-protein-mediated inward current. Bruin Res., 500: 107-118. Watson, S.J., Richard, C., Ciaranello, R.D. and Barchas, J.D. (1980) Interaction of opiate peptide and noradrenaline systems: Light microscopic studies. Peptides, 1: 23-30.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompriano (Eds.) Progress m Brain Research, Vol. 88 0 1Y91 Elsevier Science Publishers B.V.
CHAPTER 5
Noradrenergic innervation of somatosensory thalamus and spinal cord K.N. Westlund, D. Zhang, S.M. Carlton, L.S. Sorkin and W.D. Willis Department of Anatomy and Neurosciences, Marine Biomedical Institute, Unir ersity of Texas Medical Branch, Galceston, TX, U.S.A
Monoamine systems have been shown to be an important part
of an endogenous analgesic system of the central nervous system. Some aspects of the anatomical basis of monoamine modulation of nociceptive input were investigated in these studies. Two sites examined where monoamine systems are known to impinge on the pain transmission system included the grey matter of the somatosensory thalamus and the spinal cord. In particular, the connections of noradrenergic systems with these regions were emphasized. In the ventral posterolatera1 nucleus of the thalamus the presence of a sparse innervation by both noradrenergic and serotonergic fibers was confirmed by electron microscopy. Boutons containing markers for either serotonin or norepinephrine were observed contacting dendrites and somata in this region. The origins of
these projections were determined, by retrograde transport studies, to he primarily in the locus coeruleus and the dorsal raphe. Also examined was noradrenergic innervation of the spinothalamic tract neurons which relay information related to pain from the spinal cord. Some catecholamine boutons were observed to contact spinothalamic neurons directly. These included spinothalamic tract neurons of the wide dynamic range and the high threshold category. The presence of noradrenergic elements in the somatosensory thalamus and, in particular, the direct connection with spinothalamic tract neurons at the level of the spinal cord clearly provides an anatomical substrate for influencing sensory mechanisms related to pain.
Key words: norepinephrine, somatosensory, thalamus, spinal cord, locus coeruleus, dorsal raphe, primate
Introduction
The monoamine systems are believed to play a significant role in the neural mechanisms underlying central autonomic control, including regulation of cardiovascular dynamics, endocrine control and affective behavior, cortical awareness and sleep, somatomotor activity including locomotion, and somatosensory perception including pain and analgesia. An overwhelming number of studies have appeared in the literature describing monoamine involvement in somatosensory pro-
cessing. Many of these studies have dealt specifically with the interactions of serotonin and the spinothalamic system at the level of spinal cord integration (for review see Willis, 1982). To date, only a few studies have addressed the role of catecholamines in somatosensory processing in rat (Reddy and Yaksh, 19801, in cat (Engberg and Ryall, 1966; Engberg and Marshall, 1971; Belcher et al., 1978; Headley et al., 19781, in rabbit (Satoh et al., 1979) and in monkey (Willcockson et af., 1984). Norepinephrine has been shown to be at least
as potent as serotonin in selectively reducing the responses of multireceptive neurons in the spinal cord to noxious but not innocuous stimuli. In an earlier study, it was noted that an increased concentration of the noradrenergic breakdown product, normethanephrine, accumulated in the cord following treatments with analgesics (Shiomi and Takagi, 1974). Electrical stimulation of the nucleus locus coeruleus (LC), which provides a major portion of the noradrenergic projections to the spinal cord (Nygren and Olson, 1976; Westlund et al., 19531, produces analgesic effects similar to those produced by electrical stimulation of the nucleus raphe magnus (Mokha et al., 1983; Jones and Gebhart, 1986). Iontophoretically applied norepinephrine has recently been shown to reduce selectively the responses of multireceptive neurons in spinal cord laminae IV and V to noxious but not innocuous stimuli by an a,-mediated mechanism (Fleetwood-Walker et al., 1985). Norepinephrine had no effect on neurons responding to noxious stimuli alone. Yaksh (1979) observed that intrathecal administration of either a serotonin or a norepinephrine antagonist (methysergide or phentolamine) partially blocks periaqueductal grey-induced analgesia. Complete reversal can be achieved, however, if both antagonists are applied concurrently. Thus, it is tempting to speculate that the descending noradrenergic neurons are an integral part of a complex feedback circuit, along with neurons containing serotonin and opioid peptides, involved in monitoring and altering spinal somatosensory input, including noxious stimuli. At first glance, the anatomical organization of the descending serotonergic and noradrenergic components of the monoamine system display many parallels. Both are part of the “reticular” neuronal system described by Scheibel and Scheibel (1958) (i.e., neurons in the brainstem with long axons and highly collateralized terminal fields). Serotonin- and norepinephrine-containing neurons were among the first neurons to be chemically identified and mapped using the Falck-Hillarp histofluorescence technique (Dahl-
strom and Fuxe, 1964). Serotonergic and noradrenergic innervation of the brain and spinal cord is derived from several distinct collections of cells distributed through the brainstem (Dahlstrom and Fuxe, 1965) which were alphanumerically designated by Dahlstrom and Fuxe since their distribution pattern does not conform to classical cytoarchitectural borders. The recent availability of antibodies specific to the monoamine systems has allowed increased sensitivity and precision to the study of monoamine systems. Considerable progress has been made in defining the descending projections of monoamine systems since the advent of the Falck-Hillarp histofluorescence technique, immunocytochemistry and the development of methods combining axoplasmic transport methods with histochemical methods (for reviews see Bowker et al., 1982; Westlund et al., 1982, 1984a,b). The upshot of all of these studies is that upon closer analysis, the “widespread” projections of these reticular monoamine systems do in fact have quite specific destinations. The anatomical basis of catecholamine modulation of nociceptive input can be investigated by discerning the relationship between catecholamine and spinothalamic elements. Two sites where catecholamine pathways are known to impinge on the pain transmission system are the thalamus and the gray matter of the spinal cord. To date it is known that specific catecholaminergic cell groups with projections to the spinal cord include dopaminergic neurons in the hypothalamic cell group A11 of Dahlstrom and Fuxe (1964) (Blessing and Chalmers, 1979; Bjorklund and Skagerberg, 1979, 1982; Hokfelt et al., 1979; Swanson er al., 1981) and noradrenergic projections arising primarily from the A5-A7 cell groups in the pons (Hancock and Fougerousse, 1976; Westlund and CouIter, 1980; Blessing er a/., 1981; Westlund et al., 1983, 1984a,b). One descending pathway from the C1 group in the rostra1 medulla to the thoracic cord has been described as adrenergic (Ross er al., 1981). Although studies defining the specific origins of catecholamine projec-
79
tions to the spinal cord are well underway, studies of other sites of interaction between monoamine and somatosensory systems are needed. Another site where the monoamine systems impinge on the somesthetic system is the main somatosensory relay nucleus, the ventrobasal thalamus. Pharmacological studies indicate that systemic doses of morphine much lower than those required to depress spinal cord neurons can cause a reduction in nociceptive responses of ventrobasal thalamic neurons (Benoist et al., 1983). This suggests involvement of higher integrative systems in analgesic mechanisms. Electrophysiological studies have demonstrated inhibition of thalamic neurons following electrical stimulation in both the periaqueductal gray (Schieppati and Gritti, 1983) and t h e dorsal raphe nucleus (Anderson and Dafny, 1983). A recent study of the diencephalic projections of the raphe nuclei revealed that only the raphe medianus located in the midbrain nucleus centralis superior projects to the regions of the thalamus participating in pain perception in the rat (Peschanski and Besson, 1984). Two reports in the cat have previously included projections from the raphe magnus in the list of brainstem raphe nuclei with thalamic projections (Bobillier et al., 1976; Fields and Basbaum, 1978, 1979). A minor ascending projection from the nucleus LC to the ventrobasal complex has been reported in the rat by Jones and Moore (19771, using standard autoradiographic anterograde labeling. Swanson and Hartman (1975) report a similar distribution of noradrenergic terminals identified by the presence of the synthetic enzyme, dopamine-P-hydroxylase (DPH), suggesting noradrenergic involvement at the thalamic level as well. As mentioned previously, the innervation of brain regions by monoamine systems is typified by extensive networks of terminal varicosities. The dorsal horn of the spinal cord, the site of somatosensory integration of incoming primary afferent information, is heavily innervated by a latticework of catecholamine-containingterminal varicosities. Although it was shown that all cate-
cholaminergic fibers originate in the brainstem (Carlsson et al., 1964), the precise relationship of catecholamine-containing terminal varicosities to STT cells is unknown. It has been predicted by several groups (Engberg and Ryall, 1966; Engberg and Marshall, 1971; Belcher et al., 1978; Satoh et al., 19791, based on physiological and pharmacological studies, that norepinephrine causes inhibition of nociceptive neurons by a postsynaptic action. The iontophoretic application of either norepinephrine or dopamine has been shown to reduce the responses of primate STT cells to pulsed application of glutamate, consistent with a postsynaptic inhibitory action of catecholamines on STT cells (Willcockson et al., 1984). To summarize portions of our previous work as they pertain to the present studies, it was observed that the noradrenergic projections to the spinal cord arise almost exclusively from cell groups in the pons. Of the noradrenergic cells with projections to the spinal cord, 80% were localized in the nucleus LC and nucleus subcoeruleus (SC) (Westlund et al., 1981, 1983, 1984b). Autoradiographic analysis of the descending projections of these two regions indicate that both nuclei project extensively to the spinal gray at all levels (Westlund and Coulter, 1980). A latticework of noradrenergic terminal fibers and varicosities was visualized immunocytochemically in the spinal cord.
Monoamine terminals in the somatosensory thalamus More recent studies have examined specific anatomical connectivities and have required the development of techniques with better resolution. Light and electron microscopic observations have confirmed the presence of a sparse innervation by serotonergic and noradrenergic terminals of the ventral posterolateral nucleus of the macaque monkey (Fig. 1). The ventral posterolateral nucleus of the thalamus receives sensory information from the spinal cord. A sparse distribution was noted for both kinds of aminergic terminals,
80
with noradrenergic terminals present in slightly higher levels. En passant endings and intervarimse segments were evident at the ultrastructural level. Profiles immunocytochemically labeled for serotonin (Fig. lA,C) or for the noradrenergic marker, DPH (Fig. lB,D), were either axo-axonic or axo-dendritic. Some of the noradrenergic terminals appeared to contact dendritic spines.
-
81
Origins of brainstem projections to the thalamus Monoaminergic neurons projecting their axons to the ventral posterolateral nucleus of the thalamus were identified using a combined horseradish peroxidase and immunocytochemical double-label method for identifying serotonin and DPH (Westlund, et al., 1990a). To date, three monkeys have received injections of HRP into the thalamus, including one monkey with bilateral injections in the VPL nucleus. In the animal with the bilateral VPL injections, no spread of the HRP was evident in the adjacent internal capsule. The origins of brainstem projections to the VPL nucleus of the thalamus in this animal were localized and mapped as illustrated in Figure 2. In unilaterally injected animals, most of the backfilled neurons were confined to the side of the injection, but some cells were evident on the side opposite the injection site. A significant proportion of the neurons retrogradely labeled with HRP were localized in known monoamine cell groups. Many HRP-labeled neurons were observed in noradrenergic cell groups in the pons, including the nucleus LC (Fig. 31, the nucleus SC and the A5 cell group. The HRP-labeled cells were local-
MIDBRAIN
CON8
Fig. 2. Schematized map detailing the placement of bilateral HRP injections into the VPL nucleus of the thalamus. At the level of the midbrain, cells with axons projecting to the level of the VPL nucleus are identified by the dots. The asterisks at this level represent neurons which also contained serotonin. In the pons, the unsymmetrical section illustrates the neurons with projections to the VPL from the A5, subcoeruleus (SC) and the locus coeruleus (LC). The larger asterisks at this level represent neurons which were also stained for a noradrenergic marker. Notice the prominence of the projections from the dorsal raphe and the LC.
Fig. 1. A. Example of a thin serotonin-positive fiber with occasional small varicosities (arrowheads) associated with the cell clusters in the VPL nucleus of the thalamus. B. Example of a noradrenergic terminal interposed between two dendrites, one large and one small. The terminal synapses (S)with a small dendrite which might be a dendritic spine. A puncta adherens (open arrows) provides attachment to a larger dendrite C. At the EM level, an axon (A) as well as an en passant ending are labeled for serotonin. The axonal ending has a synapse (S)and round, clear vesicles. D. Noradrenergic fibers identified using localization of dopamine-p-hydroxylase (DpH) course among the cells clustered in the VPL. The presence of varicosities suggests that synaptic specializations might be present. Bar: 15 pm in A,465 nm in B 625 nm in C; and 30 p m in D.
82
ized primarily in outer portions of the nucleus LC. The lateral and medial parabrachial nuclei also contained HRP-labeled neurons following retrograde transport from the VPL nucleus of the thalamus. Double-label studies combining imniunocytochemical and HRP histochemistry confirmed that the HRP-labeled cells in the nucleus LC are noradrenergic. No HRP-labeled cells were observed in the vicinity of noradrenergic cell groups in the medulla. The primary serotonergic region containing HRP backfilled neurons was the nucleus raphe dorsalip. Most of the HRP-labeled cells were localized on the side of the injection in the nucleus raphe dorsalis and ventrolaterally in the midbrain reticular formation. Double-label experiments have revealed that many of the midbrain
cells backfilled with HRP from the VPL of the thalamus also contained immunoreactivity for serotonin. A few double-labeled serotonin neurons were also observed in the nucleus centralis superior and the raphe pontis. In the caudal brainstem, large numbers of HRP-labeled neurons were observed in regions known to project to the VPL of the thalamus, such as the dorsal column nuclei and the spinal trigeminal nucleus (Lund and Webster, 1967; Boivie, 1971, 1978; Berkley, 1974, 1975; Albe-Fessard et al., 1975; Berkley and Hand, 1978). No serotonergic regions in the medulla contained HRP-labeled cells. These preliminary data suggest that the source of noradrenergic projections to VPL nucleus in the monkey may be more extensive than one would predict based on previous studies in the
Fig. 3. A low-power photomicrograph of the LC is shown in A. The LC neurons have been localized immunocytochemically with the synthetic enzyme D P H . Some of the neurons at the periphery of the nucleus (arrows) are also double-stained for H R P retrogradely transported from the thalamus. At higher power in B, one of the LC neurons is shown. The open arrows indicate the dark crystalline H R P retrogradely transported from the ventral posterolateral nucleus of the thalamus. Bar: 0.2 mm in A,20 p m in B.
83
rat. Only LC (Lindvall et al., 1974; Swanson and Hartman, 1975; Jones and Moore, 1977) and medial and lateral parabrachial (Saper and Loewy, 1980, 1982) nuclei have been shown to project to the thalamus in previous studies in the rat. It is also interesting to note that only pontine noradrenergic cell groups are retrogradely labeled. The same regions are the sole source of noradrenergic projections to the spinal cord (Westlund et al., 1981, 1983, 1984a,b). Although other studies have reported projections from serotonergic regions to the thalamus in rat and cat (for review see Azmitia, 1978; Consolazione and Cuello, 1982; Steinbusch and Nieuwenhuys, 1983), only one other double-label study is available (Consolazione et al., 1984). This study, done in the rat, has reported serotonergic projections from the nucleus raphe dorsalis and pontis to the dorsomedial portions of the ventral thalamus. In regard to projections to the VPL nucleus of the thalamus, the data presented here for the monkey differ significantly from reports of the origins of brainstem projections to the VPL nucleus in other species. In the rat, Peschanski and Besson (1984) report projections to the VPL nucleus following anterograde transport of wheat-germ agglutinin-conjugated HRP from the raphe medianus and not from other raphe nuclei. In studies in the cat, no mention is made of projections from the nucleus raphe medianus or dorsalis to the VPL nucleus (Bobillier et al., 1976, 1979). Rather, a projection from the nucleus raphe magnus to the VPL nucleus was traced following injection and anterograde transport of ''C-leucine. Thus, a significant species difference in the origins of pathways to the VPL nucleus of the thalamus appears to be emerging. Direct innervation of spinothalamic tract neurons
Catecholaminergic axonal varicosities identified by immunocytochemical staining for DPH were observed at the light microscopic level apposing the somata of retrogradely labeled spinothalamic
tract neurons in the monkey spinal cord (Westlund et al., 1990b). At the electron microscopic level, the terminals were seen to contain clear, round vesicles and small, dense core vesicles (69.8 nm average diameter). Three HRP retrogradely labeled and two intracellularly labeled spinothalamic neurons were serially sectioned and examined at selected intervals at the electron microscopic level. Electron microscopic study revealed DPH-containing axonal terminals that directly contacted the somata and/or dendrites of lamina I, IV and V spinothalamic tract neurons. All of the profiles contacting one of the retrogradely labeled lamina I spinothalamic tract neurons were categorized from eight planes of section spaced at 1 p m intervals. Of the 305 terminal profiles counted which were adjacent to this soma, 17 (5.6%) stained positively for DPH. Of these 17 axonal varicosities, three were followed in serial sections to confirm that they had synaptic thickenings and alignment of vesicles along the membrane contacting the spinothalamic tract soma. This represented 5% of the synapses observed contacting this soma. A computer-reconstruction of this neuron is illustrated in Figure 4. Catecholaminergic terminals were observed contacting the somata and dendrites of two intracellularly filled STT cells characterized as high threshold and wide dynamic range neurons (Fig. 5). These observations clearly indicate a direct innervation of spinothalamic tract neurons by catecholaminergic neurons, providing anatomical data to support previous physiological findings demonstrating that catecholamines modulate pain transmission. In our studies of neurochemically defined terminal types contacting STT cells, profiles are classified according to a scheme devised from serial section analysis published by Carlton et al. (1989). Profiles are categorized as either terminals, non-terminals, or unidentifiable profile types. Terminal profiles are characterized as containing round, clear vesicles (R), round, clear vesicles with 25% flat or oval vesicles (F), round, clear vesicles with 2-4 dense core vesicles
84
(Dl), round, clear vesicles with > 5 dense core vesicles (D2), round, clear vesicles with 2-4 large (> 70 nm) dense core vesicles (Ll), or round, clear vesicles with 2 5 dense core vesicles (L2). For example, the total population of somatic contacts surrounding one STT neuron (870 profiles) in eight planes of section was classified. Terminal profiles represented 44% of the population of profiles apposing the soma. Types of non-terminal elements (NT) ( n = 488) seen contacting the STT soma included glia (32%), dendrites and non-terminal axons. Myelinated axonal profiles apposing STT neurons were rare. Of the terminal types ( n = 3 0 9 , by far the most prevalent termi-
nal type observed (63.7% of the terminal population) contacting the STT soma was the R type, representing terminal profiles with round, clear vesicles and no more than one dense core vesicle. The least prevalent terminal type observed contacting the soma of the STT neuron was the F type containing numerous flat vesicles. In a second study, eleven STT neurons were sampled from 3-8 planes of section at 1 p m intervals. The entire soma was montaged and the somatic contacts classified by the scheme described above (Fig. 6). In a comparison of all terminal types contacting STT neurons in Iamina I ( n = 5 ) versus those in lamina V ( n = 6), termi-
Fig. 4. Computer-assisted reconstruction of a retrogradely labeled spinothalamic tract neuron in lamina I or the monkey spinal cord. The representation was created from electron microscopic montages taken from eight planes of section. The blunted extensions represent proximal dendrites of this large marginal cell. The lighter colored speckles represent noradrenergic terminal profiles which were apposed to the soma. Several of the terminals were followed in serial section to confirm the presence of synapses. Bar: 13 p m .
85
Fig. 5. A. An example of a spinothalamic tract neuron situated at the lateral edge of lamina V of the lumbar spinal cord of monkey. This neuron was physiologically characterized as a wide dynamic range neuron and had an extensive dendritic arbor characteristic of spinothalamic tract neurons. One of the noradrenergic contacts observed on the dendrites of this cell is illustrated in the electron micrograph in B. Notice the synaptic specialization and the presence of secretory vesicles. Bar: 175 p m in A; 100 nm in B.
nals containing round, clear vesicles were observed in 16% versus 42%, respectively, of the total of profiles contacting STT cells in each lamina. Non-terminal profiles including dendrites, glia, and axons were a more prominent
PROFILES CONTACTING STT SOMATA
60,
I
0
i I LAMINA I
LAMINA V
20
10
0
R
F
D1
D2
L1
L2
X
4T
PROFILE CLASSIFICATION
Fig. 6 . Bar graph representing the population of profiles apposing the somata of identified spinothalamic tract neurons in lamina I and V of the monkey spinal cord. This included terminal profiles (typed as R, F, D1, D2, L1, L2), (see text), profiles which could not be characterized (X) and non-terminal profiles (NT).
(56%) population adjacent to lamina I STT neurons as compared to those contacting lamina V STT neurons (21%). Catecholaminergic profiles contacting ail of the lamina I neurons were distributed in both the R and D1 categories. In contrast, catecholaminergic terminals contacting the somata of intracellularly filled lamina IV and V STT neurons were found to be D2 type terminal profiles. The catecholaminergic population contacting the dendrites of these neurons were distributed in both the D1 and D2 categories. Efforts will continue to collect more dorsal horn neurons to study differences in innervation by catecholamine terminals and to elucidate the significance of the differences in innervation already observed. This will include the differences observed in the type of terminal innervating lamina I versus deeper layer STT neurons and differences seen in somatic versus dendritic innervation. It is clear at this time, however, that in the spinal cord, somatic and dendritic noradrenergic contacts are present on sensory elements and, in this case, directly on spinothalamic neurons. The
86
direct connectivity clearly provides an anatomical substrate for influencing spinal sensory mechanisms. Conclusions
These studies have sought to reveal a more definitive anatomical basis for the mechanisms proposed for the modulation of nociception by norepinephrine by carefully mapping the locations of catecholamine terminals in relation to STT cells in the spinal cord and their targets in the thalamus. Thus, although descending pathways which could form the anatomical basis for monoaminergic interactions with the spinothalamic tract at the spinal cord are well known, continued analysis of chemically defined anatomical connectivities must be done before we can begin to understand the role of monoamines in complex functions such as somatosensory processing. These studies have been extended into the ventral posterolateral nucleus of the thalamus to determine the sources of noradrenergic innervation to this sensory relay nucleus. It is likely that these connectivities contribute to what is known as “stimulation-produced analgesia” and for the clinical effectiveness of some pharmaceuticals used in pain control. Extension of these studies to other regions is necessary to understand the role of norepinephrine in motor and autonomic control. Acknowledgements
The authors wish to thank Scott Richardson, Judy Pettett, Elizabeth Hayes and Helen Willcockson for expert technical assistance. We thank Deatra Clay for preparation of the manuscript. This work was supported by grants NS11255, NS01445 and NS09743 from the NIH, and an Unrestricted Pain Center Grant from Bristol-Myers Squibb Co. References Albe-Fessard, D., Boivie, J., Grant, G. and Levante, A. (1975) Labelling of cells in the medulla oblongata and the spinal
cord of the monkey after injections of horseradish peroxidase in the thalamus. Neurosci. Lett., 1: 75-80. Anderson, E. and Dafny, N. (1983) An ascending serotonergic pain modulation pathway from the dorsal raphe nucleus to the parafascicularis nucleus of the thalamus. Bruin Res., 269: 57-67. Azmitia, E.C. (1978) The serotonin-producing neurones in the midbrain median and dorsal raphe nuclei. In L.L. Iversen, S.D. Iversen and S.H. Snyder (Eds.), Handbook of Psychophurrnacology, Vol. 9, Plenum Press, New York, pp. 233-314. Belcher, G., Ryall, R.W. and Schaffner, B. (1978) The differential effects of 5-hydroxytryptamine, noradrenaline and raphe stimulation on nociceptive and non-nociceptive dorsal horn interneurons in the cat. Bruin Res., 151: 307-321. Benoist, J.M., Kayers, V., Gantron, M. and Guilbaud, G. (1983) Low dose of morphine strongly depresses responses of specific nociceptive neurons in the ventrobasal complex of the rat. Pain, 15: 333-344. Berkley, K.J. (1974) Differential labeling of neural pathways converging on the ventrobasal complex of cat thalamus. Bruin Res., 66: 342-348. Berkley, K.J. (1 975) Different targets of different neurons in nucleus gracilis of the cat. J. Comp. Neurol., 163: 285-304. Berkley, K.J. and Hand, P.J. (1978) Efferent projections of the gracile nucleus in the cat. Bruin Rex, 153: 263-283. Bjorklund, A. and Skagerberg, G. (1979) Evidence for a major spinal cord prqjection from the diencephalon. A1 1 dopamine cell group in the rat using transmitter-specific fluorescent retrograde tracing. Bruin Res., 177: 170-175. Bjorklund, A. and Skagerberg, G. (1982) Descending monoaminergic projections to the spinal cord. In B. Sjiilund and A. Bjorklund (Eds.), Bruin Stem Control of Spinal Mechanisms, Elsevier Biomedical Press, Amsterdam, pp. 55-88. Blessing, W.D. and Chalmer, J.P. (1979) The projections of catecholamine (presumably dopamine)-containing neurons from hypothalamus to spinal cord. Neurosci. Lett., 11: 35-40. Blessing, W.W., Goodchild, A.K., Dampney. R.A.L. and Chalmers, J.P. (1981) Cell groups in the lower brain stem of the rabbit projecting to the spinal cord, with special reference to catecholamine-containing neurons. Brain Res., 221: 35-55. Bobillier, P., Seguin, S., Petitjean, F., Salvert, D., Touret, M. and Jouvet, M. (1976) The raphe nuclei of the cat brainstem: A topographical atlas of their efferent projections as revealed by autoradiography. Bruin Res., 113: 449-486. Bobillier, P., Seguin, S., Degueurce, A., Lewis, B.D. and Pujol, J.F. (1979) The efferent connections of the nucleus raphe centralis superior in the rat as revealed by autoradiography. Bruin Rex, 166 Boivie, J. (1971) The terminat the spinothalamic tract in the cat. An experimental study with silver impregnation methods. Exp. Bruin Res., 12: 331-353. Boivie, J. (1978) Anatomical observations on the dorsal column nuclei, their thalamic projection and the cytoarchitec-
87 ture of some somatosensory thalamic nuclei in the monkey. J. Comp. Neurol., 178: 17-48. Bowker, R.M., Westlund, K.N., Sullivan, M.C. and Coulter, J.D. (1982) The organization of descending serotonergic pathways. In G.F. Martin and H.G.J.M. Kuypers (Eds.), Descending Parhways to the Spinal Cord, Elsevier/North Holland, Amsterdam, pp. 239-265. Carlsson, A., Dahlstrom, A., Fuxe, K. and Hillarp, N.-A. (1964) Cellular localization of monoamines in the spinal cord. Acta Physiol. Scand., 60: 112-119. Carlton, S.M., LaMotte, C.C., Honda, C.N., Surmeier, D.J., Delanerolle, N. and Willis, W.D. (1989) Ultrastructural analysis of axosomatic contacts on functionally identified primate spinothalamic tract neurons. J. Comp. Neurol., 281: 555-566. Consolazione, A. and Cuello, A.C. (1982) CNS serotonin pathways. In N.N. Osborne (Ed.), Biology of Serotonergic Transmission, John Wiley and Sons, London, pp. 29-61. Consolazione, A,, Priestley, J.V. and Cuello, A.C. (1984) Serotonin-containing projections to the thalamus in the rat revealed by a horseradish peroxidase and peroxidase antiperoxidase double-staining technique. Brain Res., 322: 233-243. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Scand., 6 2 1-55. Dahlstrom, A. and Fuxe, K. (1965) Evidence for the existence of monoamine neurons in the central nervous system. 11. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Scand., 62(Suppl. 247): 1-36. Engberg, I. and Marshall, K.C. (1971) Mechanisms of noradrenaline hyperpolarization in spinal cord motoneurons of the cat. Acta Physiol. Scand., 83: 142-144. Engberg, I. and Ryall, R.W. (1966) The inhibitory action of noradrenaline and other monoamines on spinal neurones. J. Physiol. (London), 185: 298-322. Fields, H.R. and Basbaum, A.I. (1978) Brainstem control of spinal pain transmission neurons. Ann. Rev. Physiol., 40: 193-221. Fields, H.L. and Basbaum, A.I. (1979) Anatomy and physiology of a descending pain control system. In J.J. Bonica (Ed.), Advances in Pain Research and Therapy, Raven Press, New York, pp. 427-440. Fleetwood-Walker, S.M., Mitchell, R., Hope, P.J., Malony, V. and Iggo, A. (1985) An a > receptor mediates the selective inhibition by noradrenaline of nociceptive responses of identified dorsal horn neurones. Brain Res., 334: 243-254. Hancock, M.B. and Fougerousse, C.L. (1976) Spinal projections from the nucleus coeruleus and nucleus subcoeruleus in the cat and monkey as demonstrated by retrograde transport of horseradish peroxidase. Brain Res. Bull., 1: 229-234. Headley, P.M., Duggan, A.W. and Griersmith, B.T. (1978) Selective reduction by noradrenaline and S-hydroxytryptamine of nociceptive responses of cat dorsal horn neurones. Brain Res., 145: 185-189.
Hokfelt, T., Phillipson, 0. and Goldstein, M. (1979) Evidence for a dopaminergic pathway in the rat descending from the A l l cell group to the spinal cord. Acta Physiol. Scand., 107: 393-395. Jones, B.E. and Moore, R.Y. (1977) Ascending projections of the locus coeruleus in the rat. 11. Autoradiographic study. Brain Res., 127: 23-53. Jones, S.L. and Gebhart, G.F. (1986) Quantitative characterization of coeruleospinal inhibition of nociceptive transmission in the rat. L Neurophysiol., 56: 1397-1410. Lindvall, O., Bjorklund, A,, Nobin, A. and Stenevi, U. (1974) The adrenergic innervation of the rat thalamus as revealed by the glyoxylic acid fluorescence method. J. Comp. Neurol., 154: 317-348. Lund, R.D. and Webster, K.E. (1967) Thalamic afferents from the spinal cord and trigeminal nuclei. An experimental anatomical study in the rat. J. Comp. Neurol., 130: 313-328. Mokha, S.S., McMillan, J.A. and Iggo, A. (1983) Descending influences on spinal nociceptive neurons from locus coeruleus: Actions, pathways, neurotransmitters and mechanisms. In J.J. Bonica (Ed.), Aduances in Pain Research and Therapy, Raven Press, New York, pp. 387-392. Nygren, L.-G. and Olson, L. (1976) On spinal noradrenaline receptor supersensitivity: Correlation between nerve terminal densities and flexor reflexes various times after intracisternal 6-hydroxy dopamine. Brain Res., 116: 455470. Peschanski, M. and Besson, J.M. (1984) Diencephalic connections of the raphe nuclei of the rat brainstem: An anatomical study with reference to the somatosensory system. J. Comp. Neurol., 224: 509-534. Reddy, S.V.R. and Yaksh, T.L. (1980) Spinal noradrenergic termination system mediated antinociception. Brain Res., 189: 391-401. Ross, C.A., Armstrong, D.M., Ruggiero, D.A., Pickel, V.M., Joh, T.H. and Reis, D.J. (1981) Adrenaline neurons in the rostra1 ventrolateral medulla innervate thoracic spinal cord: A combined immunocytochemical and retrograde transport demonstration. Neurosci. Lett., 25: 257-262. Saper, C.B. and Loewy, A.D. (1980) Efferent connections of the parabrachial nucleus in the rat. Brain Res., 197: 291317. Saper, C.B. and Loewy, A.D. (1982) Projections of the pedunculopontine tegmental nucleus in the rat: Evidence for additional extrapyramidal circuitry. Brain Res., 252: 367372. Satoh, M., Kawajiri, S.-I., Vkai, Y. and Yamamoto, M. (1979) Selective and nonselective inhibition by enkephalins and noradrenaline of nociceptive response of lamina V type neurons in the spinal dorsal horn of the rabbit. Brain Res., 177: 384-387. Scheibel, M.E. and Scheibel, A.B. (1958) Structural substrates for integrative patterns in the brainstem reticular core. In H.H. Jasper, L.D. Proctor, R.S. Knighton, W.C. Noshay and R.T. Costello (Eds.), Reticular Formation of the Brain, Little, Brown, Boston, pp. 31-55. Schieppati, M. and Gritti, I. (1983) Influences of locus
88 coeruleus, raphe dorsalis, and periaqueductal gray matter on somatosensory-recipient thalamic nuclei. Exp. Neurol., 82: 698-705. Shiomi, H . a n d Takagi, H. (1974) Morphine analgesia and the bulbospinal noradrenergic system: Increase in the concentration of normethanephrine in the spinal cord of the rat caused by analgesics. Br. J. Pharmacol., 52: 519-526. Steinbusch, H.W.M. and Nieuwenhuys, R. (1983) The raphe nuclei of the rat brain stem: A cytoarchitectonic and immunohistochemical study. In P.C. Emson (Ed.), Serotonin: Current Aspects of Neurochemistry and Function, Cjiemical Neuroanatomy, Raven Press, New York, pp. I31-207. Swanson, L.W. and Hartman, B.K. (1975) The central adrenergic system. An immunofluorescence study of the location of cell bodies and their efferent connections in the rat utilizing dopamine-P-hydroxylase as a marker. J. Comp. Neurol., 163: 467-506. Swanson, L.W., Sawchenko, P.E., Btrod, A,, Hartman, B.K., Helle, K.B. and Vanorden, D.E. (1981) An immunohistochemical study of the organization of catecholaminergic cells and terminal fields in the paraventricular and supraoptic nuclei of the hypothalamus. J. Comp. Neurol., 196: 271-285. Westlund, K.N. and Coulter, J.D. (1980) Descending projections of the locus coeruleus and subcoeruleus/medial parabrachial nuclei in monkey: Axonal transport studies and dopamine-P-hydroxylase immunocytochemistry. Brain Res., 2: 235-264. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1981) Origins of spinal noradrenergic pathways demonstrated by retrograde transport of antibody to dopamine-P-hydroxylase. Neurosci. Lett., 25: 243-249. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1982) Descending noradrenergic projections and their
spinal terminations. In G.F. Martin and H.G.J.M. Kuypers (Eds.), Descending Pathways to the Spinal Cord, Elsevier/North Holland, Amsterdam, pp. 219-238. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of rat. Brain Res., 263: 15-31. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1984a) Organization of descending noradrenergic systems. In C.R. Lake and M.G. Ziegler (Eds.), Norepinephrine: Frontiers of Clinical Neuroscience, Vol. 4, Williams and Wilkins Co., Baltimore, pp. 55-73. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1984b) Origins and terminations of descending noradrenergic projections to the spinal cord of monkey. Brain Res., 292: 1-16. Westlund, K.N., Sorkin, L.S., Carlton, S.M., Ferrington, D.G., Willcockson, H.H. and Willis, W.D. (1990a) Serotonergic and noradrenergic projections to the ventral posterolateral nucleus of monkey thalamus. J. Comp. Neurol., 295: 197207. Westlund, K.N., Carlton, S.M., Zhang, D. and Willis, W.D. (1990b) Direct catecholaminergic innervation of primate spinothalamic tract neurons. J. Comp. Neurol., 299: 178186. Willcockson, W.S., Chung, J.M., Hori, Y., Lee, K.H. and Willis, W.D. (1984) Effects of iontophoretically released amino acids and amines on primate spinothalamic tract cells. J. Neurosci., 4: 732-740. Willis, W.D. (1982) Control of nociceptive transmission in the spinal cord. In D. Ottoson (Ed.), Progress in Sensory Physiology, Vol. 3, Springer-Verlag, Berlin, 159 pp. Yaksh, T.L. (1979) Direct evidence that spinal serotonin and noradrenaline terminals mediate the spinal antinociceptive effects of morphine in the periaqueductal gray. Brain Res., 160: 180-185.
C.D. Barnes and 0. Pompeiano (Eds.) Progress m Errrrrr Rewrrrch, Vol. 88 0 1991 Elscvier Science Publishers B.V.
89
CHAPTER 6
Efferent projections of different subpopulations of central noradrenaline neurons R. Grzanna and J.-M. Fritschy Department of Neuroscience, Johns Hopkins Unii)er.sitySchool of Medicine, Baltimore, MD, U.S.A.
Early anatomical studies of the projections of central noradrenergic (NA) neurons led to the widely accepted view of NA cells as a class of diffusely projecting neurons. This view greatly influenced the formulation of numerous hypotheses about the functional role of these neurons in the central nervous system (CNS). With the introduction of transmitterspecific retrograde and anterograde transport methods, two powerful tools became available to rigorously re-examine whether the projections of NA neurons are diffuse or topographically organized. This article summarizes some of the results of these studies in which retrograde transport of fluorescent tracers and anterograde transport of the lectin Phaseolus vulgaris leucoagglutinin (PHA-L), respectively,
were combined with immunohistochemical identification of NA neurons and their projections. The results of these studies revealed a remarkable degree of specificity in the projections of different subgroups of NA neurons. In the rat CNS, the differential distribution of NA axons of the locus coeruleus (LC) and non-coerulean NA cells is particularly striking in the spinal cord and brainstem. In these regions, NA axons of the LC are primarily distributed to sensory nuclei while NA axons of non-coerulean NA neurons are distributed to motor nuclei. The results support the proposition that NA neurons can be divided into subgroups which differ in their connections and hence represent separated anatomical entities with different functional capacities.
Key words: locus coeruleus, neurotoxin, true blue, PHA-L, spinal cord, brainstem
Introduction The functional significance of the existence of several discrete subgroups of noradrenergic (NA) neurons in the central nervous system (CNS) has remained a controversial issue. By applying topographical and morphological criteria, Dahlstrom and Fuxe (1964) delineated seven NA subgroups (the A1-A7 groups) in the rat brainstem but conceded that “ . . . it is possible-perhaps even probable-that the division is partly artificial.” The demonstration of extensive and extremely
divergent NA projections to widespread and functionally diverse regions in the CNS raised serious doubts as to whether the division of NA cells into discrete subgroups is functionally meaningful. Moore and Bloom (1979) concluded their comprehensive review of the anatomy and physiology of central NA neuron systems with the statement that NA neurons appear to have diffuse and widespread projections with little topography. These authors saw no compelling reason to distinguish more than two NA neuron systems: the locus coeruleus (LC) system and the lateral
90
tegmental system. Bjorklund and Lindvall (1986) considered NA cells in the dorsomedial medulla (the A2 group) a separate entity from the lateral tegmental group of NA neurons and thus distinguished three separate NA neuron systems: the LC system, the dorsal medullary system and the lateral tegmental system. These authors cautioned, however, that the widely dispersed projections of NA neurons should not be interpreted as evidence that the system lacks structural or functional specialization. Moreover, they stressed the existence of the principal differences between pontine and medullary NA systems with respect to the type of sensory inputs activating them. The view of NA neurons as a poorly organized class of diffusely projecting neurons had a tremendous influence on the formulation of hypotheses about their functional role in the CNS (Kety, 1970; Dismukes, 1977). Neural systems with projections as divergent as those of NA neurons are likely to exert a very global influence on CNS functions rather than having specific effects on restricted targets. Not surprisingly, the view of NA neurons as a class of broadly projecting neurons which influence target cells by nonsynaptic release of neurotransmitter (Descarries et al., 1977) was rapidly and almost universally accepted. Yet there is a growing body of experimental evidence suggesting that subgroups of central NA neurons have rather different and mostly non-overlapping projections and thus should be considered separate anatomical entities. In particular, recent anatomical studies with transmitter-specific tracing methods have provided a wealth of new data about the projections of different NA subgroups. These new findings should be taken into account in the design of future experiments into the functional role of NA subgroups, in particular the role of the LC, the largest NA subgroup in the mammalian CNS.
Efferents of NA subgroups revisited with transmitter-specific tracing methods While the histofluorescence method provided nearly all of the light microscopic data about the
distribution of NA axons within the CNS, the method could not furnish reliable information concerning the extent to which (if at all) different NA subgroups project to specific targets within the CNS. A systematic search for topography within the projections of NA subgroups was not feasible until the availability of transmitterspecific tracing methods. The use of fluorescent retrograde tracers combined with histochemical staining of retrogradely labeled neurons (Skirboll et al., 1984; Kuypers and Huisman, 1984) provided powerful tools to reliably trace the projections of NA subgroups. Similarly, the combination of anterograde transport of the lectin Phaseolus vulgaris leucoagglutinin (PHA-L) (Gerfen and Sawchenko, 1984, 1985) with immunohistochemical staining of NA axons holds great promise of adding significantly to our knowledge of NA neuron projections. Sawchenko and Swanson (1981, 1982) were among the first to combine retrograde transport of True Blue with immunohistochemistry in studies of the origin of the NA innervation of the hypothalamus. They reported remarkably restricted and largely non-overlapping projections of the Al, A2, and A6 groups to this brain region. Others have used this new methodology to identify the origin of the projections of NA neurons to the spinal cord (Loewy et al., 1979, 1986; Bjorklund and Skagerberg, 1982; Westlund et at., 1983). Byrum and Guyenet (1987) employed tract tracing and immunohistochemistry to describe the unique anatomical relationships between NA cells of the A5 group and CNS regions involved in the control of cardiovascular functions. We have used fluorescent retrograde tracers in combination with dopamine-P-hydroxylase (DP H ) immunohistochemistry and identified distinct topographic relationships between NA cells in the LC and the A5 and A7 groups and cranial nerve nuclei (Grzanna et al., 1987) and in studies of divergent projections of NA cells to brainstem and spinal cord (Lyons and Grzanna, 1988). Recently, the powerful anterograde tracing method with PHA-L (Gerfen and Sawchenko,
91
1984, 1985) has been applied for tracing the projections of NA subgroups (Fritschy et al., 1987; Cunningham and Sawchenko, 1988; Fritschy and Grzanna, 1990a,b). This method is particularly promising for studies of rat LC efferents since in this species the LC is a tightly packed cell cluster that presents an easy target for tracer injections. In addition, the PHA-L method can be combined with immunohistochemistry to establish definitively the neurochemical identity of the anterogradely labeled axons. Studies with these new transmitter-specific tracing methods have provided new data about the organization of NA efferents in the rat CNS which strongly support the proposition that different NA subgroups entertain different connections. The purpose of this chapter is to review the results of studies obtained with these new transmitter-specific tracing methods. Special emphasis will be given to the description of LC efferents.
Organization of NA projections to brainstern Levitt and Moore (1979) studied the origin and organization of the catecholamine innervation in the rat brainstem. By analyzing the effects of bilateral LC lesions on the innervation pattern and the NA content in brainstem nuclei, they identified two major NA projection systems: the LC system and the lateral tegmental neuron system comprised of NA cells of the Al, A2, A5 and A7 groups. As a general organizational plan they proposed that the LC projects to sensory and association nuclei of the brainstem, including the brainstem reticular formation, while NA cells of the lateral tegmental group innervate motor nuclei. Confirmation of this organizing principle has come from retrograde as well as anterograde transport studies. Grzanna et al. (1987) showed that the NA inputs to the motor nuclei of the trigeminal and facial nerve originate almost exdusively in cells of the A5 and A7 groups with almost no contribution from the LC. In contrast, after deposits of tracer in the rostra1 part of the
spinal trigeminal nucleus, the majority of labeled NA neurons was found in the LC. Our PHA-L anterograde transport studies provided a detailed view of the distribution of LC axons in the brainstem (Fritschy and Grzanna, 1990b) which confirms in a large measure the conclusion reached by Levitt and Moore (1979) that the LC innervates only select areas of the brainstem. These areas include the sensory trigeminal complex, particularly its spinal nucleus, the principal nucleus of the inferior olive, the cochlear nuclei, the pontine nuclei, the’ interpeduncular nucleus and the tectum of the midbrain. The most unexpected observation from the anterograde transport studies was the near absence of LC axons within the brainstem reticular formation. Comparisons between the density of PHA-L labeled LC axons in different CNS regions after injections of the tracer into the LC consistently revealed far fewer LC axons in the brainstem than in cerebral cortex and spinal cord. These comparisons leave the strong impression that the projections of the LC are primarily destined for cortical regions, the tectum and cerebellum and the spinal cord dorsal horn and that the LC innervation of the brainstem is sparse and restricted to a few select areas. We have recently documented that the NA neurotoxin DSP-4 can be used to ablate NA axons of the LC with remarkable selectivity while sparing non-coerulean NA axons (Fritschy and Grzanna, 1989, and this volume). We have taken advantage of this selective vulnerability of LC axons to visualize the differential distribution of LC and of non-coerulean NA axons in the brainstem. As shown in Figure 1, DSP-4 eliminates NA axons in the spinal trigeminal nucleus pars caudalis but has little effect on NA axons in the adjacent reticular formation. It is widely accepted that preterminal NA axons traverse the hindbrain as a single, circumscript bundle of fibers in the dorsolateral tegmentum (Jones and Friedman, 1983). NA cells of the pons and medulla are thought to feed axons to this fiber system which serves as the principal conduit of descending fibers to the brainstem and
92
Fig. 1. Darkfield photomicrograph of the distribution of NA axons in medulla of a control rat (panel A) and a DSP-4 treated rat (panel B) two weeks after drug administration. Note the nearly complete loss of NA axon staining in the spinal trigeminal nucleus. Dopamine-p-hydroxylase (DPH) staining with the peroxidase method of Hsu ef al. (1981). Bar: 200 p m .
spinal cord as well as ascending fibers from medullary NA subgroups. Analysis of sagittal sections of the brainstem stained with anti-DPH reveals an extraordinary number of longitudinally oriented NA axons. As shown in Figure 2, these longitudinally oriented NA fibers are distributed throughout the dorsoventral extent of the reticular core rather than being restricted to the catecholamine bundle. At the spinal-medullary junction, descending NA axons in this loosely textured fiber system can be traced into the spinal cord. The most dorsally located NA axons turn laterally and traverse obliquely the caudal portion of the spinal trigeminal nucleus pars caudalis to enter the dorsal horn of the spinal cord. Many of the ventrally positioned NA axons, located just dorsal to the infe-
rior olivary complex, enter the lateral and ventral funiculi of the spinal cord. Anterograde labeling of LC axons with PHA-L makes it possible to trace their position in the hindbrain en route to the spinal cord. At the level of the caudal pons and the medulla, the majority of labeled LC axons was observed in a dorsolatera1 position of the brainstem between the nucleus of the solitary tract and the dorsal part of the spinal trigeminal nucleus. Our description of the location of descending LC axons within the brainstem is at variance with the generally held view that they travel in the longitudinal catecholamine bundle. We suspect that the bulk of the fibers in the catecholamine bundle are ascending NA axons of the A1 and A2 groups and that descending LC axons are located dorsally
93
Gr
Sol
10
PY Fig. 2. Distribution of NA axons in the caudal medulla in a parasagittal (A) and a transverse section (B). In the parasagittal section, note the presence of longitudinally oriented axons scattered throughout the reticular formation rather than forming a circumscript fiber bundle. The more ventrally located fibers are notably thicker than N A axons in the dorsal portion of the medulla. The level of the parasagittal section is indicated by the arrows in the transverse section shown on the right. Immunoperoxidase staining with antibodies to D P H . Gr, gracile nucleus; 10, inferior olive; py, pyramidal tract; Sol, nucleus of the solitary tract. Bar: 200 p m .
while descending NA axons of the A5 and A7 groups are located ventrally to this bundle. Thus, there appears to exist a crude topographic arrangement of NA axons traversing the brainstem with respect to their origin and sites of termination.
Organization of descending NA projections to the rat spinal cord All levels and all subdivisions of the gray matter of the spinal cord are innervated by NA axons. In the white matter NA axons are particularly numerous in the dorsolateral funiculus and in the ventral funiculus (Fig. 3). For no other
CNS region have reports about the origin and pattern of distribution of NA projections been so conflicting as for the spinal cord (for a detailed discussion see Proudfit and Clark, this volume). The supraspinal origin of NA axon terminals has been known since the early 1960s (Magnusson and Rosengren, 1963; Carlsson et al. 1964). Dahlstrom and Fuxe (1965) reached the conclusion that there are two main descending NA pathways in the spinal cord: “One large system which runs in the anterior funiculus and in the most ventral part of the lateral funiculus, terminating in the ventral horn and another somewhat smaller system which runs in the lateral funiculus,
Fig. 3. Darkfield photomicrograph of the distribution of NA axons in a transverse section of a rat lumbar spinal cord. DPH immunohistochemical staining using the immunoperoxidase method. Bar: 200 p m .
especially in its dorsal part, terminating in the sympathetic lateral column and the dorsal horn.” Their conclusion that the A1 group and possibly the A2 group contain the cells of origin of the descending NA projections could not be confirmed by later studies.
It was not until the introduction of transmitter-specific retrograde tracing methods that projections of the LC, the subcoeruleus, and the A5 and A7 groups to the spinal cord were unequivocally established (Bjorkund and Skagerberg, 1982; Westlund et al., 1983; Byrum et al., 1984; Loewy
95
et al., 1986; Byrum and Guyenet, 1987). The multiple origins of descending NA axons raised the question of whether different NA subgroups project to functionally different regions of spinal cord and thus may differ in the functional capacity.
Even the earliest studies of the projections from different cell groups to the spinal cord SUSpected a differential distribution of descending NA axons (see Dahlstrom and Fuxe, 1965). Nygren and Olson (1977) and Commissiong et al. (1978) used the histofluorescence method to as-
B
A
D
Fig. 4. Maps of the distribution of phaseolus vulgaris leucoagglutinin-labeled locus coeruleus (LC) axons in horizontal sections of rat spinal cord. Each chart depicts the distribution of labeled LC axons within a single 40 micron section. A, section through the substantia gelatinosa; B, deep part of dorsal horn: C, intermediate zone; D and E, dorsal and ventral part of ventral horn. Bar: 500 km. (From Fritschy and Grzanna, 19YOa.)
Fig. 5. Darkfield photomicrograph of a DPH-stained section of rat lumbar spinal cord two weeks after a single intraperitoneal injection of the noradrenergic neurotoxin DSP-4. Note the complete loss of NA axon staining in the dorsal horn and the intermediate zone. Immunoperoxidase staining. Bar: 100 p m .
97
sess the effects of bilateral LC lesions on the distribution of NA in the spinal cord. They found more pronounced decreases of catecholamine fibers in the ventral horn than in the dorsal horn. The results of these studies have laid the foundation for the still widely accepted view that in the rat the LC distributes NA axons via the ventrolateral funiculus to the ventral horn, the intermediate zone and the ventral part of the dorsal horn (Bjorklund and Lindvall, 1986). In contrast, cells of the A5 group, are thought to distribute axons to the intermediolateral cell column (Loewy et al., 1979, 1986; Byrum and Guyenet, 1987). Summarizing data from biochemical and anatomical studies of descending NA projections, Bjorklund and Skagerberg (1982) presented an organizational plan depicting two separate bulbospinal NA systems: the coeruleo-spinal pathway originating in the LC and subcoeruleus which distributes axons to the ventral horn, the intermediate zone and the ventral part of the dorsal horn and the tegmento-spinal pathway originating in cells of the A5 and A7 group which distributes axons to the dorsal horn and the intermediolateral cell column. This basic plan differs fundamentally from the organizational scheme proposed for the projections of the LC and the lateral tegmental cells to the brainstem. With the PHA-L anterograde transport method it has become possible to trace directly the trajectory and distribution of the coeruleo-spinal axons. Since the rat LC is a compact cell group consisting entirely of NA projection neurons, the nucleus is an ideal target for the application of this powerful anterograde tracing method. Our PHAL anterograde tracing experiments identified two anatomical features of the coeruleo-spinal projection in the rat: (1) LC axons descend the length of the spinal cord within the superficial layers of the dorsal horn, and ( 2 ) LC axons are distributed mainly within the dorsal horn and the intermediate zone of the spinal cord (Fig. 4). Thus, the longitudinally oriented NA axons in the superficial laminae of the dorsal horn originally described by Dahlstrom and Fuxe (1965)
represent LC axons. These findings differ from previous anatomical descriptions of the coeruleospinal pathway in two important respects: (1) the majority of LC axons traverse the spinal cord within the gray matter of the dorsal horn rather than in the ventrolateral funiculus, and (2) their distribution is most dense in the superficial part of the dorsal horn, less dense in the deeper laminae of the dorsal horn and in the intermediate zone, and very sparse in the ventral horn. The preferential distribution of LC axons within the dorsal horn and the intermediate zone of the spinal cord makes it likely that cells of the A5 and A7 groups, which form the tegmento-spinal pathway, distribute axons to the ventral horn and the intermediolateral cell column. Several lines of evidence support this conclusion. In the brainstem, cells of the A5 and A7 subgroups have long been considered the principal source of the NA innervation of cranial nerve motor nuclei (Levitt and Moore, 1979; Grzanna et al., 1987). Lyons and Grzanna (1988) demonstrated that most A7 cells which innervate the motor nucleus of the trigeminal nerve also distribute collaterals to the spinal cord. It seems unlikely that NA neurons, which project to motor nuclei in the brainstem, distribute collaterals to the dorsal rather than ventral horn of the spinal cord. Furthermore, we have demonstrated that following ablation of the NA innervation of the dorsal horn and intermediate zone with the noradrenaline neurotoxin DSP-4, cells of the LC, unlike cells of the A5 and A7 groups, can no longer be labeled from the spinal cord by retrograde transport (Lyons et al., 1989). Since DSP-4 treatment does not eliminate the dense NA input to the ventral horn (Fig. 51, these retrograde transport data support our proposition of a separate NA pathway from the A5 and A7 groups to the ventral horn distinct from the coeruleo-spinal pathway. It has been suggested that coeruleo-spinal axons descend in the spinal cord ipsilaterally in the ventral and ventrolateral funiculus (Commissiong et al., 1978; Basbaum and Fields, 1979) and that A5 and A7 axons travel in the dorsolateral fu-
98
A
Fig. 6 . Summary diagram of the course and distribution of the coeruleo-spinal (A) and tegmento-spinal (B) pathways in the rat. Coeruleo-spinal projection descends the length of the spinal cord within the superficial laminae of the dorsal horn and distributes fibers within the dorsal and the intermediate zone. The majority of NA axons of the tegrnento-spinal pathway descend the spinal cord in the ventral funiculus and distribute terminals to the ventral horn and in the ventral part of the intermediate zone.
niculus. Since the vast majority of LC axons descend in the dorsal horn, we suspect that NA axons in the white matter of the lateral and ventral funiculus originate in cells of the A5 and A7 groups. Thus, at least two separate descending NA systems can be distinguished which differ in their origin, their trajectories and their sites of termination (Fig. 6). Additional arguments can be made in support of the organizational plan proposed here for the descending NA projections. The very same basic plan has previously been proposed for the projec-
tions of the LC and the lateral tegmental cell groups to the brainstem. Levitt and Moore (1979) concluded that the LC innervates sensory and association nuclei of the brainstem and that NA cells of the lateral tegmental group innervated motor and autonomic nuclei (see above). Levitt and Moore (1979) noted distinct differences in the morphology of NA axon varicosities of the LC and the lateral tegmental group. Such differences are best seen when comparing NA axon terminals in motor nuclei of the brainstem with NA axons in the sensory trigeminal nucleus or the cerebellum. Similarly, structural differences are apparent when comparing NA axons in the dorsal horn with those in the ventral horn (see Fig. 3). The anatomical data presented here about the organization of descending NA projections to the spinal cord indicate a functional segregation of the LC and non-coerulean NA axons similar to that suggested for the brainstem by Levitt and Moore (1979). Whether NA axons of the A5 and A7 groups form a single system or whether they differ in their course and termination will have to await the outcome of future studies. The present data reveal a degree of specificity in the distribution of NA axons from different NA subgroups that is incompatible with the long-held belief that the LC exerts a global influence over spinal cord functions. The differential distribution of the coeruleo-spinal and tegmento-spinal NA projections suggests that the two pathways differ in their functional capacities. The preferential distribution of LC axons to the dorsal horn and their unique intragriseal trajectory strongly suggest a role of the LC in the modulation of sensory inputs, particularly nociceptive inputs, to the dorsal horn. The proposition put forward by Kuypers (1982) and Holstege and Kuypers (1987) that the coeruleo-spinal projection may be instrumental in providing the emotional drive in the execution of movements should be re-evaluated in view of the new anatomical data about the organization of this supraspinal monoaminergic pathway. Such a role appears far more likely for the tegmento-spinal NA projection. However, as discussed by
99
Proudfit (this volume), significant species differences appear to exist in the organization of the supraspinal NA systems. Are the projections of the LC topographically organized? Early studies of the projections of the LC have characterized this nucleus as one of the most diffusely projecting centers within the CNS. Yet, in recent years several studies have demonstrated a crude but nevertheless distinct topographic order in the efferent network of the LC. These studies were prompted in part by the results of retrograde transport studies which revealed that tracer injections into different CNS regions produced different labeling patterns within the LC (Mason and Fibiger, 1979). Moreover, morphological and immunohistochemical studies of the LC revealed distinct subdivisions within the LC as well as morphologically and neurochemically different types of neurons (Swanson, 1976; Grzanna and Molliver, 1980a,b; Cintra et al., 1982; Steindler and Trosko, 1989). Even though retrograde transport studies have confirmed the existence of LC neurons with divergent projections (Ader et al., 1980; Nagai et al., 1981; Room et al., 1981; Steindler, 1981; Fallon and Loughlin, 1982), they represent only a small subset of cells. Loughlin et al. (1986a,b) provided the most extensive evidence for regional topography within the LC. They used three-dimensional reconstructions of the distribution of LC neurons retrogradely labeled by injections of horseradish peroxidase into six different CNS regions. Going beyond the conventional mapping of retrograde labeling in the LC, they correlated the morphological features of retrogradely labeled neurons with their projection targets and made the intriguing observation that morphologically different types of neurons project to different regions of the CNS (topomorphological organization). In this context, it will be of considerable interest to determine whether subpopulations of LC neurons, in which NA is co-localized with a particular neuropeptide, project to different re-
gions within the CNS (Levin et al., 1987; Holets et al., 1988). A systematic analysis of the results of retrograde as well as anterograde tracing experiments with PHA-L may provide new insights into the topography of LC projections. Conclusions
The application of transmitter-specific tracing methods has revealed an unsuspected degree of specificity in the projections of subgroups of NA cells. As Cunningham and Sawchenko pointed out (19881, the ascending projections from medullary NA subgroups are positioned to play a role in the distribution of visceral afferent inputs to the forebrain. The A1 group has been suspected to play a role in fluid and cardiovascular homeostasis (Cunningham and Sawchenko, 1988). The A2 group, based upon its connection with the parvocellular division of the PVN, has been implicated in the regulation of the hypothalamopituitary-adrenal axis. A unique relationship between the A5 group and autonomic centers has long been suspected (Loewy et al., 1979, 1986; Byrum et al., 1984; Guyenet and Byrum, 1985; Byrum and Guyenet, 1987). Moreover, our recent tracing studies suggest a special relationship between A7 cells and somatic motoneurons. The cortices and brain regions directly involved in the processing of sensory data receive their NA innervation from the LC, while subcortical regions, the core of the brainstem and motor nuclei are innervated by non-coerulean NA neurons. Since recent electronmicroscopic studies have provided unequivocal evidence that NA axons engage postsynaptic cells via discrete junctional sites (Olschowka et al., 1981; Parnavelas and Papadopoulos, 19&9),it should come as no surprise that topography is as much a structural feature of the projections of NA neurons as it is of other CNS neurons. Acknowledgements
The authors would like to thahk Mrs. Hong Fritschy for drawing the summary diagrams of the
100
descending noradrenergic pathways. This work was supported by NIH grant MH41977. References Ader, J.P., Room, P., Postema, F. and Korf, J. (1980) Bilaterally diverging axon collaterals and contralateral projections from rat locus coeruleus neurons, demonstrated by fluorescent retrograde double labeling and norepinephrine metabolism. J. Neural Transm., 49: 207-218. Basbaum, A.I. and Fields, H.L. (1979) The origin of descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat: Further studies on the anatomy of pain modulation. J. Comp. Neurol., 187: 513-532. Bjorklund, A. and Lindvall, 0. (1986) Catecholaminergic brain stem regulatory system. In V.B. Mountcastle and F.E. Bloom (Eds.), Handbook of Physiofu&,y,Section 1. American Physiological Society, Bethesda, MD, pp. 155-235. Bjorklund, A. and Skagerberg, G. (1982) Descending monoaminergic projections to the spinal cord. In B. Sjolund and A. Bjorklund (Eds.), Brain Stem Control of Spinal Mechanisms, Elsevier Biomedical Press, Amsterdam, pp. 55-88. Byrum, C.E. and Guyenet, P.G. (1987) Afferent and efferent connections of the AS noradrenergic cell group in the rat. J. Comp. Neurol., 261: 529-542. Byrum, C.E., Stornetta, R. and Guyenet, P.G. (1984) Electrophysiological properties of spinally projecting AS noradrenergic neurons. Brain Res., 303: 15-29. Carlsson, A., Falck, B., Fuxe, K. and Hillarp, N.-A. (1964) Cellular localization of monoamines in the spinal cord. Acta fhysiol. Scund., 60: 112-119. Cintra, L., Diaz-Cintra, S., Kemper, T. and Morgane, J.P. (1982) Nucleus locus coeruleus: A morphometric Golgi study in rats of three age groups. Brain Rex, 247: 17-28. Commissiong, J.W., Hellstrom, S.O. and Neff, N.H. (1978) A new projection from the locus coeruleus to the spinal ventral columns: Histochemical and hiochemical evidence. Brain Res., 148: 207-213. Cunningham, E.T. and Sawchenko, P.E. (1988) Anatomical specificity of noradrenergic inputs to the paraventricular and supraoptic nuclei of the rat hypothalamus. J. Comp. Neurol., 274: 60-76. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta fhysiol. Scand., 62: Suppl. 232, 1-55. Dahlstrom, A. and Fuxe, K. (1965) Evidence for the existence of monoamine neurons in the central nervous system. 11. Experimentally induced changes in the intraneuronal amine levels of bulbospinal neuron systems. Acla fhysiul. Scand., 64: Suppl. 247, 1-36. Descarries, L., Watkins, K.C. and Lapierre, Y. (1977) Noradrenergic axon terminals in the cerebral cortex of rat. 111. Topometric ultrastructural analysis. Bruin Res., 133: 197222.
Dismukes, K. (1977) A new look at the aminergic nervous system. Nature (London), 269: 557-558. Fallon, J.H. and Loughlin, S.E. (1982) Monoamine innervation of the forebrain: Collateralization. Brain Res. Bull., 9: 295-307. Fritschy, J.-M. and Grzanna, R. (1989) Immunohistochemical analysis of the neurotoxic effects of DSP-4 identifies two populations of noradrenergic axon terminals. Neuroscience, 30: 181-197. Fritschy, J.-M. and Grzanna, R. (1990a) Demonstration of two separate descending noradrenergic pathways to the rat spinal cord: Evidence for an intragriseal trajectory of locus coeruleus axons in the superficial layers of the dorsal horn. J. Comp. Neurol., 291: 553-582. Fritschy, J.-M. and Grzanna, R. (1990b) Distribution of locus coeruleus axons within the rat brainstem demonstrated by Phaseoleus vulgaris leucoagglutinin anterograde tracing in combination with dopamine-P-hydroxylase immunofluorescence. J. Comp. Neurol., 293: 616-631. Fritschy, J.-M., Lyons, W.E., Mullen, C.A., Kosofsky, B.E., Molliver, M.E. and Grzanna, R. (1987) Distribution of locus coeruleus axons in the rat spinal cord: A combined anterograde transport and immunohistochemical study. Brain Rex, 437: 176-180. Gerfen, C.R. and Sawchenko, P.E. (1984) An anterograde neuroanatomic tracing method that shows the detailed morphology of neurons, their axons and terminals: Immunohistochemical localization of an axonally transported plant lectin, Phaseolus vulgaris-leucoagglutinin (PHA-L). Brain Rex, 290: 219-238. Gerfen, C.R. and Sawchenko, P.E. (1985) A method for anterograde axonal tracing of chemically specified circuits in the central nervous system: Combined Phaseolus vulgaris leucoagglutinin (PHA-L) tract tracing and immunohistochemistry. Brain Res., 343: 144-150. Grzanna, R. and Molliver, M.E. (1980a) Cytoarchitecture and dendritic morphology of central noradrenergic neurons. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Revisited, Raven Press, New York, pp. 83-97. Grzanna, R. and Molliver, M.E. (1980b) The locus coeruleus in the rat: An immunohistochemical delineation. Neuruscience, 5: 21-40. Grzanna, R., Chee, W.K. and Akeyson, E.W. (1987) Noradrenergic projections to brainstem nuclei: Evidence for differential projections from noradrenergic subgroups. J. Cump. Neurol., 262: 76-91. Guyenet, P.G. and Byrum, C.E. (1985) Comparative effects of sciahic nerve stimulation, blood pressure, and morphine on the activity of AS and A6 pontine noradrenergic neurons. Brain Res., 327: 191-201. Holets, V.R., Hokfelt, T., Rokaeus, A,, Terenius, L. and Goldstein, M. (1988) Locus coeruleus neurons in the rat containing neuropeptide Y, tyrosine hydroxylase or galanin and their efferent projections to the spinal cord, cerebral cortex and hypothalamus. Neuroscience, 24: 893-906. Holstege, J.C. and Kuypers, H.G.J.M. (1987) Brainstem projections to spinal motoneurons: An update. Neuroscience, 23: 809-821.
101 Hsu, S.M., Raine, L. and Fanger, H. (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem. Cytochem., 29: 577580. Jones, B.E. and Friedman, L. (1983) Atlas of catecholamine perikarya, varicosities and pathways in the brainstem of the cat. J. Cornp. Neuroi., 215: 382-396. Kety, S.S. (1970) The biogenic amines in the central nervous system: Their possible roles in arousal, emotion, and learning. In F.O. Schmitt (Ed.), The Neurosciences, Second Study Program, Rockefeller University Press, New York, pp. 324-336. Kuypers, H.G.J.M. (1982) A new look at the organization of the motor system. Prog. Brain Rex, 57: 381-403. Kuypers, H.G.J.M. and Huisman, A.M. (1984) Fluorescent neuronal tracers. In S. Fedoroff (Ed.), AdLlances in Cellular Neurobiology, Vol. 5 , Academic Press, New York, pp. 307-340. Levin, M.C., Sawchenko, P.E., Howe, P.R.C., Bloom, S.R. and Polak, J.M. (1987) Organization of galanin-immunoreactive inputs to the paraventricular nucleus with special reference to their relationship to catecholaminergic afferents. J. Cornp. Neurol., 261: 562-582. Levitt, P. and Moore, R.Y. (1979) Origin and organization of brainstem catecholamine innervation in the rat. J. Comp. Neurol., 186: 505-528. Loewy, A.D., McKellar, S. and Saper, C.B. (1979) Direct projections from the A5 catecholamine cell group to the intermediolateral cell column. Bruin Res., 174 309-314. Loewy, A.D., Marson, L., Parkinson, D., Perry, M.A. and Sawyer, W.B. (1986) Descending noradrenergic pathways involved in the A5 depressor response. Brain Res., 386: 313-324. Loughlin, S.E., Foote, S.L. and Bloom, F.E. (1986a) Efferent projections of the nucleus locus coeruleus: 1. Topographic organization of cells of origin demonstrated by three-dimensional reconstruction. Neuroscience, 18: 291 -306. Loughlin, S.E., Foote, S.L. and Grzanna, R. (1986b) Efferent projections of nucleus locus coeruleus: 11. Morphologic subpopulations have different efferent targets. Neuroscience, 18: 307-319. Lyons, W.E. and Grzanna, R. (1988) Noradrenergic neurons with divergent projections to the motor trigeminal nucleus and the spinal cord: A double retrograde neuronal labeling study. Neuroscience, 26: 681-693. Lyons, W.E., Fritschy, J.-M. and Grzanna, R. (1989) The noradrenergic neurotoxin DSP-4 eliminates the coeruleospinal projection but spares projections of the A5 and A7 groups to the ventral horn of the rat spinal cord. J. Neurosci., 9: 1481-1489. Magnusson, T. and Rosengren, E. (1963) Catecholamines of the spinal cord normally and after transection. Experientia, 19: 229. Mason, S.T. and Fibiger, H.D. (1979) Regional topography
within noradrenergic locus coeruleus as revealed by retrograde transport of horseradish peroxidase. J. Comp. Neurol., 187: 703-724. Moore, R.Y. and Bloom, F.E. (1979) Central catecholamine neuron systems: Anatomy and physiology of the norepinephrine and epinephrine systems. Ann. Re[,. Neurosci., 2: 113-168. Nagai, T., Satoh, K., Imamoto, K. and Maeda, T. (1981) Divergent projections of catecholamine neurons of the locus coeruleus as revealed by fluorescent retrograde double labeling technique. Neurosci. Lett., 23: 117-123. Nygren, L.-G. and Olson, L. (1977) A new major projection from the locus coeruleus: The main source of noradrenergic nerve terminals in the ventral and dorsal columns of the spinal cord. Brain Res., 132 85-93. Olschowka, J.A., Molliver, M.E., Grzanna, R., Rice, F.L. and Coyle, J.T. (1981) Ultrastructural demonstration of noradrenergic synapses in the rat central nervous system by dopamine-P-hydroxylase immunocytochemistry. J. Hisfochrm. Cytochem., 29: 271 -280. Parnavelas, J.G. and Papadopoulos, G.C. (1989) The monoaminergic innervation of the cerebral cortex is not diffuse and nonspecific. TINS, 12: 315-319. Room, P., Postema, F. and Korf, J. (1981) Divergent axon collaterals of the rat locus coeruleus neurons: Demonstration by a fluorescent double labeling technique. Bruin R ~ s .221: , 219-230. Sawchenko, P.E. and Swanson, L.W. (1981) A method for tracing biochemically defined pathways in the central nervous system using combined fluorescence retrograde transport and immunohistochemical techniques. Brain Rex, 210: 31-51. Sawchenko, P.E. and Swanson, L.W. (1982) The organization of noradrenergic pathways from the brainstem to the paraventricular and supraoptic nuclei in the rat. Bruin Res. Rec., 4: 275-325. Skirboll, L., Hokfelt, T., Norell, G., Phillipson, O., Kuypers, H.G.J.M., Bentivoglio, M., Catsman-Berrevoets, C.E., Visser, T.J., Steinbusch, H., Verhofstad, A., Cuello, A.C., Goldstein, M. and Brownstein, M. (1984) A method for specific transmitter identification of retrogradely labeled neurons: lmmunofluorescence combined with fluorescence tracing. Brain Res. Reu., 8: 99-127. Steindler, D.A. (1981 Locus coeruleus neurons have axons that branch to the forebrain and cerebellum. Bruin Res., 223: 367-373. Steindler, D.A. and Trosko, B.K. (1989) Two types of locus coeruleus neurons born on different embryonic days in the mouse. Anat. Embyol., 179: 423-434. Swanson, L.W. (1976) The locus coeruleus: A cytoarchitectonic, Golgi and immunohistochemical study in the albino rat. Brain Res., 110: 39-56. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of the rat. Brain Rex, 263: 15-31.
This Page Intentionally Left Blank
C.D. Barnei and 0. Pompeiano (Eds.) Progre.vs U I Hrurn lieseurch, Vol. 88 0 1Y91 Elrevier Science Publishers B.V.
103 CHAPTER 7
Pontospinal transmitters and their distribution V.K. Reddy, S.J. Fung, H. Zhuo and C.D. Barnes Department of Veterinary and Comparutioe Anatomy, Pharmacology and Physiology, College of Veterinary Medicine, Washington State Unicersity, Pullman, WA, U.S.A.
The dorsolateral pontine tegmentum of the cat is known to contain a large population of catecholaminergic neurons. Additionally, several studies have also shown the presence of other neurochemicals (acetylcholine, enkephalin, neuropeptide Y, serotonin, somatostatin and substance P). In this study, we have employed retrograde transport of horseradish peroxidase in combination with immunocytochemistry to determine the locations of pontospinal neurons which contain catecholamine, enkephalin, neuropeptide Y, and serotonin. Furthermore, we have combined the retrograde transport of Fast Blue and immunofluorescence histochemistry to determine whether enkephalin-containing neurons are catechol-
aminergic. All pontospinal neurons, irrespective of the neurochemical content, were observed in the ventral and lateral parts of the dorsolateral pontine tegmentum at coronal levels P1.8-P4.0. These neurons were located in the nuclei locus coeruleus alpha and subcoeruleus and the Kolliker-Fuse nucleus. A high concentration of these neurons was evident in the Kolliker-Fuse nucleus when compared to the nuclei locus coeruleus alpha and subcoeruleus. Quantitative data have revealed that enkephalin is contained in a large proportion of the pontospinal catecholaminergic neurons (75%). The observations suggest that catecholaminergic neurons may contain one or more putative peptide neurotransmitters.
Key wordx enkephalin, neuropeptide Y, catecholamine, dorsolateral pontine tegmentum, spinal cord
Introduction
In mammals, the nucleus locus coeruleus (LC) lies in the dorsolateral pontine tegmentum (DLPT) and extends rostrocaudally from the level of the dorsal tegmental nucleus of Gudden to rostral levels of the motor nucleus of the trigeminal nerve. Unlike the rat, this nucleus in the cat is loosely arranged and lacks definite boundaries (Russel, 1955; Chu and Bloom, 1974; RamonMoliner, 1974). Based on cytoarchitecture, this nucleus was further divided into two parts: (1) the LC dorsalis (LCd) located in the periventricular
gray medial to the root of the mesencephalic tract of the trigeminal nerve; (2) the LC alpha (LCa) which lies ventral to the LCd and medial to the dorsal half of the rostral cerebellar peduncle (Maeda et al., 1973; Sakai et al., 1979; LCger and Hernandez-Nicaise, 1980). The DLPT of cats, similar to that of rats, is characterized by the presence of a large population of catecholaminergic neurons (Dahlstrom and Fuxe, 1964; Maeda et al., 1973; Jones and Moore, 1974; Wiklund et al., 1981). In cats, this catecholaminergic neuronal population, estimated at about 9,150 cells unilaterally (Wiklund et al., 19811, is distributed
104
in various nuclear groups of the DLPT. These catecholaminergic nuclear groups, often referred to as the LC complex, include the nuclei LC (LCd; LCa), subcoeruleus (SC), parabrachialis medialis (PBM) and lateralis (PBL) and the Kolliker-Fuse nucleus (NKF). Although in the past these nuclear groups of the cat were regarded as a homogenous system because of their catecholamine content, several studies have since described the concomitant existence of other neurotransmitters and peptides; these include: acetylcholine, enkephalin, neuropeptide Y (NPY), serotonin (5-HT), somatostatin and substance P (Charnay et al., 1981, 1982; Lkger et al., 1981, 1983; Wiklund et al., 1981; Hunt and Lovick, 1982; Jones and Beaudet, 1987; Reiner and Vincent, 1987). The LC is known to project to diverse regions of the neuraxis; many of the LC and SC axons project to the forebrain (Foote et al., 1983), through the dorsal noradrenergic (NA) bundle (Ungerstedt, 1971; Nakazato, 19871, and to the spinal cord (Kuypers and Maisky, 1975, 1977; Hancock and Fougerousse, 1976; McBride and Sutin, 1976; Basbaum and Fields, 1979; Holstege et al., 1979; Tohyama et al., 1979; Takeuchi et al., 1980; Hayes and Rustioni, 1981) via the dorsolateral, ventrolateral, and ventral funiculi (Kuypers and Maisky, 1977; Stevens et al., 1985). LC neurons that project to the spinal cord of the cat were often observed in its caudoventral part (Hancock and Fougerousse, 1976; Hayes and Rustioni, 1981; Nakazato, 19871, whereas those projecting to the forebrain were found predominantly in its rostrodorsal part (Maeda et al., 1973; Pickel et al., 1974; Mason and Fibiger, 1979; Nakazato, 1987; Jones and Cuello, 1989). Furthermore, double-labeling studies in rats have demonstrated the presence of neurons, in an area located between the dorsal and ventral parts of the LC, whose axon collaterals project to both the forebrain and spinal cord (Room et al., 1981). In contrast to the adult, in-pouch young opossums (Cabana and Martin, 1984) and day-old rats (Chen and Stanfield, 19871, the coeruleospinal neurons
are found throughout the LC including dorsal as well as ventral parts. However, the coeruleospinal neurons in a neonatal rat become restricted to its ventral portion by the end of the fourth postnatal week; this change is thought to be brought about by selective collateral elimination during postnatal development (Chen and Stanfield, 1987). Catecholaminergic axons have been demonstrated in the superficial laminae of the dorsal horn, ventral horn, intermediolateral cell column and lamina X of the spinal cord of rat and opossum (Fuxe, 1965; Martin et al., 1979, 1982). Anterograde transport studies have shown that various nuclear groups in the DLPT project to overlapping targets in the spinal cord (Westlund and Coulter, 1980; Holstege and Kuypers, 1982; Westlund et al., 1982); however, some of these nuclei are known to preferentially project to specific targets. Accordingly, in the rat, the LC (Fritschy et al., 1987) and the NKF (Clark and Proudfit, 1988) have been shown to have a dense projection to the superficial laminae of the dorsal horn. In addition, the LC is also known to project differentially to the sacral parasympathetic nuclei (Westlund and Coulter, 1980; Holstege and Kuypers, 1982), whereas the SC (Westlund and Coulter, 1980) and NKF (Holstege and Kuypers, 1982) project to the thoracic sympathetic cell column. Except for one study in the cat (Stevens et al., 19821, in which a combined retrograde transport of a fluorescent dye (Evans Blue) and glyoxylic acid histofluorescence technique was used to demonstrate the distributions of pontospinal catecholaminergic neurons, all anatomical studies to date have either demonstrated the existence of specific neurochemicals in the LC-complex neurons or the distributions of non-immunostained pontospinal neurons. As part of the long-term physiological study on the stimulation-produced effects of these cells on the cat spinal cord, we sought to determine the distribution, size and make-up of pontospinal catecholaminergic, peptidergic and serotoninergic neurons. Since the results of our initial experiments (Reddy et al.,
105
1989, 1990b) documented that a major proportion of all pontospinal neurons contain catecholamine (80-87%) and enkephalin (72-80%), we also examined the distribution of pontospinal neurons that co-contain catecholamine and enkephalin. In order to accomplish these studies we used: (1) the retrograde transport of horseradish peroxidase (HRP) in combination with either methionine-enkephalin (M-Enk), neuropeptide Y (NPY), tyrosine hydroxylase (TH), or serotonin (5-HT) immunocytochemistry; 2) the immunoff uorescence
technique to simultaneously demonstrate TH- and M-Enk-like immunoreactivity in neurons that were labeled by spinal injections of Fast Blue
(FBI. Results
Origins of pontospinal axons which contain catecholamine To identify the origins of the pontospinal catecholaminergic axons, HRP was injected into lum-
C
Fig. 1. Photomicrographs of neurons in the dorsolateral pontine tegmentum (DLPT) of the cat. A, B and C, pontospinal tyrosine hydroxylase-like (TH-LI) (double-labeled; TH-HRP) neurons from the nuclei locus coeruleus (LC), Kolliker-Fuse (NKF) and subcoeruleus (SC). D, retrogradely labeled (HRP-labeled) neuron from the nucleus LC. E, TH-LI neuron from the nucleus LC. (From Reddy et al., 1989.)
106
oxidase (PAP) method of Sternberger (1979). Figure 1 shows three types of identifiable neurons subsequent to employing the aforementioned staining paradigm: (1) pontospinal TH-like immunoreactive (-LI) or TH-HRP cells (Fig. 1 A-C); (2) retrogradely labeled or HRP-labeled cells (Fig. 1D); (3) TH-LI cells (Fig. 1E). Figure 2 plots the three types of neurons in the DLPT at AP levels P1.8-P4.0 from one cat following HRP injections into lumbar intumescence.
bar or cervical intumescentia of anaesthetized cats; following an appropriate survival period the animals were perfusion fixed (Reddy et al., 1989). The brainstem was sectioned coronally (25-30 Fm) on a freezing microtome and the sections were processed for HRP histochemistry (Bowker et al., 1982). The sections were then incubated in rabbit anti-TH antiserum (dilution 1:8,000) for 48 h at room temperature. They were further processed by means of the peroxidase-antiper-
P 1.8
0
-1
-2
H
-3
4
3
2
4
1
P 2.0
3
2
1
P4.0 O
'
I -
H
.. y o T a . k . O Om
-1
\=-2
. -3
! -4
5
4
3
L
2
1
m 4
3
2
1 -4
L
Fig. 2. A plotting showing the distribution of TH-LI (open circles), TH-HRP (closed circles) and HRP-labeled neurons (open squares). Plotting is at four stereotaxic AP planes and includes depth (H) and laterality (L) coordinates. (From Reddy et al., 1989.)
107 TABLE 1 Percentage of pontospinal cells from individual nuclei of the LC complex which contain TH following injections of HRP into cervical (C) or lumbar (L) segments of the spinal cord
sc
LC TH-HRP
=%
TH-HRP TH-HRP + HRP
Cat #1 (L)
TH-HRP + HRP _- 85.3
:E
_ :!$ - 88.4
Cat #2 (C)
- = 92.5
:$
-=
::: 92.3
Cat #3 (C)
_- 80.3
:!:
Cat #6 (C)
:_ I -- 89.5
4 60.8 _ ,;: -- 81.9
Cat #7 (L)
= 80.0
_ i:l -- 74.8
85.5 5.5 2.5
79.6 12.5 5.6
MEAN SD SEM
=
NKF TH-HRP
Yo
TH-HRP + HRP
= Yo
= 92.8 222 _
u'$- 94.9
:;:
78.5
-=
=
149 _ 16, - 82.2
iz
- --
86.9 88.5 6.4 2.8
Cat 1 had all sections counted, cats 2 and 3 had each third section counted, cats 6 and 7 had each tenth section counted. (From Reddy ef al., 1989.)
The pontospinal TH-LI neurons are present in the LC, SC, and NKF. A vast majority of these cells were found in the ventrolateral part of the DLPT and extended rostrocaudally from coronal levels P1.8-P4.0. Quantitative analysis of these data show that 80% to nearly 90% of the pontospinal cells from the LC, SC and NKF contain TH-like immunoreactivity (Table 1). The normalized frequency distribution of TH-HRP cells in the LC, SC and NKF are summarized in Table 2. The majority of the TH-HRP neurons are located in the NKF. However, we observed no significant difference in the percentage of these TH-HRP neurons in the LC (LCd and L C a ) and NKF ( P > 0.05).
Origins of pontospinal axons which contain peptides In this study, like the previous study, we combined retrograde transport of HRP and the indirect antibody PAP method of Sternberger (1979) to localize the pontospinal M-Enk-LI neurons and compare these with the locations of descend-
TABLE 2 Normalized frequency (percent) distribution of pontospinal catecholaminergic (TH-HRP) neurons in the nuclei locus coeruleus, subcoeruleus and Kolliker-Fuse in five cats following injections of HRP into cervical (C) or lumbar (L) segments of the spinal cord Cat No.
TH-HRP cells counted in DLPT
1 (L)
1258
2 (C)
590
3 (C)
147
6 (C)
363
7 (L)
392
MEAN SD SEM
Percent distribution
Lc
sc
NKF
50.06 (504) 40.00 (236) 37.75 (282) 32.78 (119) 41.84 (164)
23.85 (300) 22.37 (132) 12.45 (93) 26.17 (95) 25.77 (101)
36.09 (454) 37.63 (222) (49.8) (372) 41.05 (149) 32.39 (127)
38.49 3.50 1.57
22.12 5.61 2.5
39.39 6.6 2.95
Numbers in parentheses refer to cell counts of TH-HR neurons. Cat 1 had all sections counted, cats 2 and 3 had each third section counted, cats 6 and 7 had each tenth section counted.
108
ing TH-LI neurons of the cat (Reddy et al., 1990b). To enhance the concentration of M-Enk, colchicine (300-400 p g in saline) was administered intracerebroventricularly 24 h prior to perfusion. Following processing for HRP histochemistry, every first brainstem section in 8 or 10 sections was immunostained for M-Enk and the second section for TH. Microscopic examination of such sections revealed: (1) pontospinal M-EnkLI or TH-LI (double-labeled; M-Enk-HRP or TH-HRP) cells; (2) retrogradely labeled (HRPlabeled) cells; (3) M-Enk-LI or TH-LI cells. Figure 3 depicts the three types of neurons following M-Enk immunocytochemistry. The distribution of the three types of cells in adjacent sections of the DLPT immunostained for either M-Enk or T H from one cat, subjected
to bilateral injections of HRP into the lumbar segments of the spinal cord, is shown (Fig. 4). The M-Enk-HRP neurons, similar to the THHRP neurons, were encountered in the ventrolatera1 part of the DLPT at coronal levels P2.0-P4.0. Comparison of adjacent sections revealed a close similarity in the locations of both M-Enk-HRP and TH-HRP neurons. Morphometric data (Table 3) also show no significant difference in the mean diameters of the various types of pontospinal neurons within and between each of the three major nuclei (LC, SC, NKF ( P > 0.05)), although the cells in the NKF tend to be larger than those in the SC and LC (NKF > SC > LC). Additionally, like the TH-HRP neurons (Table 21, the majority of M-Enk-HRP neurons are located in the NKF (Table 4). Furthermore, these
A
B
Fig. 3. High-power photomicrographs of neurons in the DLPT of the cat. A, B and C, M-Enk-HRP neurons from the nuclei subcoeruleus, LCa and Kolliker-Fuse nucleus, respectively. D, HRP-labeled neuron from the nucleus LCa. E, M-Enk-LI neurons from the nucleus LCd. (From Reddy et al., 1990b.)
109
data (Table 5 ) and those presented earlier (Table 1) suggest that a large proportion of all the pontospinal neurons contain M-Enk- ( 7 2 4 0 % ) and TH-like immunoreactivity (80437%). From these observations one can only speculate that a large proportion of all pontospinal catecholaminergic neurons may also contain M-Enk. Findings from one other experiment revealed the presence of NPY-HRP neurons in the ventrolateral part of the DLPT. Similar to the locations of TH-HRP neurons, NPY-HRP neurons were encountered in the LCa, SC and NKF at coronal levels P2.0-P4.0. Approximately 20% of all descending cells from the DLPT exhibited NPY-like immunoreactivity. A major proportion of these cells approximated 18 p m in diamter, resembling the TH-HRP and M-Enk-HRP neurons.
Origins of pontospinal neurons which co-contain catecholamine and methionine-enkephalin We used the immunofluorescence technique (Wessendorf and Elde, 1985; Staines et al., 1988; Reddy et al., 1990a) to simultaneously demonstrate TH- and Enk-LI in neurons that were labeled by spinal injections of FB. Approximately 8-10 pl of a 5% solution of FB was injected bilaterally into the sixth and seventh (L6, L7) segments of the spinal cord of anaesthetized cats. Two weeks later the animals were anaesthetized, and colchicine (300-400 p g of saline) was injected into one of the lateral ventricles. The animals were perfused 24 h later and the appropriate levels of the brainstem were frozen-sectioned coronally (20 pm). The free-floating brainstem sections were incubated in the following solutions and thoroughly washed in 0.1 M phosphate buffered saline (PBS) after each incubation: (1) a mixture of primary antisera containing mouse anti-TH and rabbit anti-M-Enk in 0.1 M PBS (dilution 1 : 1 : 1,000) for 48 h at 4°C; (2) a mixture of secondary antisera containing fluoresceinlabeled goat anti-mouse IgG and lissaminerhodamine-labeled goat anti-rabbit IgG in 0.1 M PBS (dilution 1: 1:50) for 1 h at room temperature.
The sections were examined with an Olympus (BH2) fluorescence photomicroscope using appropriate filter systems to visualize FB, fluorescein, and lissamine-rhodamine. Examples of pontospinal neurons that co-contain TH- and MEnk-like immunoreactivity are shown in Figure 5. Figure 6 shows plots of the results obtained from one of the cases subjected to lumbar injections of FB. Neurons exhibiting both TH- and M-Enk-like immunoreactivity have an extensive distribution throughout the DLPT, whereas the FB-labeled cells which contained evidence for TH- and MEnk-like immunoreactivity (TH-Enk-FB; triplelabeled cells) were limited to the ventrolateral part of the DLPT. The distribution of all neurons of the DLPT that project to the spinal cord are shown in Figure 7. Included are FB-labeled neurons that are immunoreactive to either TH or M-Enk, those that are immunoreactive to both TH and Enk (same as those plotted in Fig. 6) and those that are immunoreactive to neither T H nor M-Enk. All pontospinal neurons similar to those described earlier (Figs. 2, 4) extended ventrolaterally from the distal half of the mesencephalic tract of the trigeminal nerve to the ventral margin of the rostra1 cerebellar peduncle and rostrocaudally from coronal levels P1.5-P4.0. Table 6 presents the distribution of variously labeled cell types at different levels (P1.5-P4) of the LC complex (cat #170). About 90% of all catecholaminergic neurons in the LC complex were found to have M-Enk-like immunoreactivity, while close to 77% of all enkephalinergic neurons showed TH-like immunoreactivity. In addition, approximately 75% of the pontospinal neurons exhibited both TH- and M-Enk-like immunoreactivity, 8% exhibited TH-like immunoreactivity alone, and 7% contained M-Enk alone. In about 10% of the spinally projecting neurons there was no immunostaining for either TH or M-Enk.
Origins of pontospinal axons which contain serotonin As a part of a long-term study on the neurochemical composition of bulbo- and pontospinal
110
M -ENK
TH
P 1.8
-1
-2
0
-1
P 2.5
-2
-3
H
H -4
I
1
I
I
P 3.0
-1
P 4.0
-2
-2 "
5 4 3 2 1 Fig. 4. Plottings of neurons from adjacent sections of the DLPT of the cat in relation to Horsley-Clarke's stereotaxic planes, following bilateral injections of HRP into the lumbar segments of the spinal cord (cat # 7) and subsequent M-Enk (right) or TH (left) immunocytochemistry. M-Enk-LI or TH-LI neurons (open circles); M-Enk-HRP or TH-HRP neurons (closed circles); HRP-labeled neurons (open squares). (From Reddy et al., 1990b.l
111
Fig. 5. Photomicrographs of the neurons in the nucleus locus coeruleus (A-C); the nucleus subcoeruleus (D-F); and the Kolliker-Fuse nucleus (G-I). A, D, and G show neurons labeled by Fast Blue (FB); B, E, and H demonstrate M-Enk-LI neurons; C, F, and I depict TH-LI neurons. Arrow heads indicate neurons that contain FB plus M-Enk and TH. Calibration bar in I = 30 Fm; it applies to panels A-I. (From Zhuo et al., 1991.)
112 TABLE 3
TABLE 4
Mean diameters (pm) of the pontospinal neurons in the dorsolateral pontine tegmentum following HRP injection into lumbar segments (Cat No. 7) and methionine enkephalin or tyrosine hydroxylase immunocytochemistry
Normalized frequency (percent) distribution of pontospinal enkephalinergic (M-Enk-HRP) neurons in the nuclei locus cueruleus, subcoeruleus and Kolliker-Fuse in three cats fullowing injections of HRP into cervical (C) or lumbar (L) segments of the spinal cord
Nucleus
M-Enk-HRP TH-HRP neurons neurons
Locus coeruleus 18.48 (0.48) Subcoeruleus 18.56 (0.62) 20.68 (0.62) Kolliker-Fuse
HRP-labeled neurons
Cat No.
M-Enk-HRP
Percent distribution
cells counted in DLPT
Lc
sc
NKF
l(L)
297
8(L)
276
11 (C)
369
31.7 194) 22.8 (63) 25.5 (94)
21.2 (63) 19.2 (53) 20.6 (76)
47.1 (140) 58.0 (160) 53.9 (199)
26.7 4.56 2.63
20.3 1.03 0.59
53.0 5.5 3.18
17.84 (0.63) 18.48 (0.59) 18.46 (0.50) 19.22 (0.86) 19.44 (0.74) 19.48 (0.79)
Numbers in parentheses are fSEM. N = 25 for all the neuronal types in each of the three nuclei. (From Reddy ef al., 1990b.)
Mean SD SEM
pathways in the cat, we sought to determine whether any of the neurons in the cat’s DLPT that contain 5-HT also project to the spinal cord. To accomplish this, we processed the brainstem sections using the indirect antibody PAP method (Sternberger, 1979) to immunostain for 5-HT in neuroni that were retrogradely labeled by spinal injections of HRP. Results from this preliminary study revealed the presence of 5-HT-LI neurons through the rostrocaudal extent of the DLPT. Although sparse, 5-HT-HRP neurons were present in the LCa and SC. The soma size of the vast majority of the 5-HT-LI and 5-HT-HRP neurons was not different from TH-LI or TH-HRP
Numbers in parentheses refer to cell counts of M-Enk-HRP neurons. Cells from every tenth section were counted in cats No. 7 and 8 and from every eighth section in cat No. 11. (From Reddy ef al., 1990b.)
neurons and their average diameters approximated 18 pm. (Fig. 8). Discussion In the present study we used the combined retrograde transport of HRP (Bowker et al., 1982) and the indirect antibody PAP technique of Stern-
TABLE 5 Percentage of pontospinal enkephalinergic neurons in the nuclei locus coeruleus, subcoeruleus and Kolliker-Fuse following HRP injections into the cervical (C) or lumbar (L) segments of the spinal cord Nucleus locus coeruleus
Nucleus subcoeruleus
M-Enk-HRP M-Enk-HRP + HRP Cat No. 7 (L) Cat No. 8 (L) Cat No. 11 (C) Mean % SD SEM
-,;_ - 74.0
= Yo
M-Enk-HRP + HRP 63
=
M-Enk-HRP =%
M-Enk-HRP + HRP
80.8
= 78.8
= 84.1
$ = 76.4
_- 74.5
76.4, 2.4 1.39
Nucleus Kolliker-Fuse
M-Enk-HRP
,;
79.8 4.9 2.81
=
=%
64.5
160 -
- 79.6
199
278
=
71.6 71.9 7.55 4.36
Cells from every tenth section were counted in cats No. 7 and 8 and from every eighth section in Cat No. 11. (From Reddy et al., 1990b.)
113
berger (1979) to determine the distributions of either pontospinal TH-, M-Enk-, or NPY-LI neurons. However, this method cannot directly demonstrate the coexistence of two neurochemicals in the same neurons that were retrogradely
labeled by spinal injections of HRP. We therefore employed the three-color immunofluorescence technique (Wessendorf and Elde, 1985; Staines et al., 1988) to demonstrate the pontospinal neurons that co-contain TH and M-Enk.
P3.0
P1.5
-1
H
-2
-3
5
4
3
2
1
5
4
3
2
1
114
lems, the potential interactions between the various antisera were rigorously tested and no inappropriate interactions among the primary antibodies and the secondary antisera were found. This study documents the existence and distribution of pontospinal neurons that contain: (1) either TH, M-Enk, or NPY; (2) both TH and
Staines et al. (1988) pointed out two major problems that might be encountered using this technique: (1) the possibility of cross-reaction between the primary antibodies and the inappropriate secondary antibodies; (2) the potential for cross-reaction between the secondary antibodies themselves. In order to circumvent these prob-
P 3.0
P1.5
r-rT-7- 1"
0
-1
-2
H
-3
5
4
3
P4.0
P2.0
i
~o
2
1
0
-1
H
-2
-3
5
4
3
L
2
1
L
Figure 7. Plots of four kinds of pontospinal neurons in sections from the DLPT of the cat (#I701 that contain FB plus TH and M-Enk (closed asterisks),FB plus TH (open asterisks), FB plus M-Enk (x), and FB alone (dots). (From Zhuo et al., 1991.)
115 TABLE 6 Frequency distribution of different cell types along the rostrocaudal extent (P1.5-P4) of the dorsolateral pontine tegmentum of the cat (#170) Cell type5
Total
P 1.5
P2
P3
P4
A B C D E F G
143 43 16 393 10 13 114 732
26 17 1 96 0 1 14 155
31 21 12 139 4 12 44 263
39 5 1 70 3 0 34 152
47 0 2 88 3 0 22 162
M-Enk TH FB TH-M-Enk M-Enk-FB TH-FB TH-M-Enk-FB H Total (from A-G)
I Total TH-M-Enk/Total M-Enk (D + G)/(A + D + E + G) J Total TH-M-Enk/Total TH (D + G)/(B + D + F + G) K TH-M-Enk-FB/Total FB G/(C + E + F + G) L Single FB/Total FB C/(C + E + F + G) M M-Enk-FB/Total FB (E/(C + E + F + G) N TH-FB/Total FB F/(C + E + F + G)
76.8%
81%
84%
71%
69%
90%
86%
85%
95%
100%
75%
88%
61%
89%
81%
10%
6%
17%
3%
7%
7%
0%
6%
8%
11%
8%
6%
17%
0%
0%
From Zhuo et al., 1991.
M-Enk; (3) neither TH nor M-Enk. These cells were observed in the caudoventral part of the LC and in the SC and NKF and extended rostrocau-
dally from Anterior/Posterior levels P1.5-P4.0. This cell distribution was generally comparable to those earlier reports where retrograde transport
Figure 8. Photomicrographs of the neurons in the LCa of the cat. A, serotonin-like (5-HT-LI) (arrow) and dRP-labeled (arrow head) neurons. B, 5-HT-HRP neuron. Calibration bar in B = 10 km; it applies to both A and B.
was utilized to localize the pontospinal neurons (Hancock and Fougerousse, 1976; Hayes and Rustioni, 1981; Nakazato, 1987). These data also revealed that the majority of the pontospinal THLI, M-Enk-LI, NPY-LI and TH- and M-Enk-LI neurons were concentrated in the NKF and less so in the LC and SC. Since most catecholaminergic neurons in the DLPT also co-contain M-Enk (Charnay et al., 1982; LCger et al., 1983), it is not surprising that the present findings generally agree with previous observations in which the NKF was also found to be a principal source of pontospinal catecholaminergic neurons (Stevens et al., 1982). An earlier study (Ltger and HernandezNicaise, 1980) described four types of cells in the LCd, L C a and SC: medium-sized neurons (30-50 pm), two types of intermediate-sized neurons (15-40 pm) and small neurons (10-15 pm). Furthermore, histofluorescence studies have demonstrated that medium-sized neurons are catecholaminergic (Jones and Moore, 1974) and the intermediate-sized cells are both catecholaminergic and indolaminergic (LCger et al., 1979). The morphometric data in the present study extends these findings by demonstrating that all pontospinal (TH-HRP; M-Enk-HRP; NPY-HRP; HRP-labeled) neurons fall into the intermediate cell category; their mean diameters ranged from 17.84 to 20.68 pm. There was no significant difference in the mean diameters of these neurons within or between the three nuclei LC, SC and NKF ( P > 0.05). Additionally, quantitative analysis of these data suggests that a majority of all pontospinal neurons contain TH (80-87%), MEnk ( 7 2 4 0 % ) or TH and M-Enk (75%), and in a very small percentage (10%) of the pontospinal neurons there was no immunostaining for either TH or Enk. These latter neurons may well contain other neurochemicals. Serotonin and NPY (Lai and Barnes, 1985; preliminary studies, this lab) are known to be present in some of the pontospinal neurons. The conduction velocities of coeruleospinal neurons in the rat ranged from 0.5-0.9 m/sec, a
feature characteristic of unmyelinated fibers (Guyenet, 1980); however, electrophysiological experiments in the cat present data with conduction velocities greater than 5 m/sec (Fung and Barnes, 1981; Mokha et al., 1985). In a more recent study in the cat (Nakazato, 1987), conduction velocities consistent with that of the rat ( < 1 m/sec) account for fewer than 13% of the cells reported and the conduction values in others were estimated to be 2.5-33 m/sec, correlating well with earlier data in cats (Fung and Barnes, 1981). Furthermore, the slow-conducting neurons ( < 1 m/sec) were considered to be NA and the fast-conducting neurons (2.5-33 m/sec) to be non-NA (Nakazato, 1987). In contrast to this view, quantitative analysis of our data indicates that 85% of all coeruleospinal neurons are NA and, therefore, stand a greater chance of being recruited in electrical stimulation experiments. It is possible to assume that the remaining 15% of the coeruleospinal non-NA neurons are much larger and are recruited first in such experiments; however, we can find no evidence in the cat to support this view. The morphometric data based on cell size (present study) would argue against electrical stimulation biasing toward either NA (TH-HRP) or non-NA (HRP-labeled) coeruleospinal neurons. Additionally, the cell sizes for the TH-HRP cell group of 17.84k0.63 p m mean diameter are not characteristic of unmyelinated neurons; this fact suggests that myelinated fibers may be present in the coeruleospinal system of the cat. Our finding of a dual presence of TH- and Enk-like immunoreactivity within individual coeruleospinal neurons accords well with data from gther indirect studies. There are biochemical data suggesting that norepinephrine (NE) and Enk are co-stored in the dense-cored vesicles of axonal terminals in cat stellate ganglion (Bastiaensen et al., 1988). Earlier studies by Costa and co-workers have shown that 8-Br-cyclic AMP triggered an increase in cultured bovine adrenal medullary chromaffin cell contents of proEnk mRNA (Quach et al., 1984) as well as TH activity
117
(Kumakura et af., 19791, suggesting a common mode of cellular induction of Enk and NE synthesis. Other pharmacological studies have also provided evidence of co-release of these two transmitters from the chromaffin cells in vitro (Livett et af., 1981). Although these findings tend to support the dual presence of NE and Enk in the soma and axon terminals of coeruleospinal neurons, it may be possible that in triple-labeled cells, only one antigen might be contained in the spinal trajectory while the other contributed to local, recurrent axon collaterals or other efferent pathways. The functional role of NE and Enk co-transmission in spinally projecting systems is not yet known. The LC and the adjacent SC in the cat contain pontospinal TH- and/or M-Enk-LI neurons (present study) and these nuclei are known to project to the superficial laminae of the dorsal horn of the spinal cord (Holstege et af., 1979; Holstege and Kuypers, 1982; Fritschy et af.,1987). The data from these studies, as well as that cited earlier (Stevens et al., 19821, implicate NKF as the primary source of pontospinal enkephalinergic and NA neurons; however, the specific neuronal termination sites from this nucleus remain to be identified. A recent study in rats addressed this question and found a substantial projection of catecholaminergic neurons in the NKF to laminae I and I1 of the cervical and lumbar intumescentia (Clark and Proudfit, 1988; see Proudfit, this volume). The spinal dorsal horn contains an abundance of enkephalinergic (Glazer and Basbaum, 1981) and catecholaminergic (Fuxe, 1965; Nygren and Olson, 1977; Martin et af., 1982) varicosities, with the latter being reduced following bilateral lesioning of the LC in rats (Commissiong et al., 1978). Physiological evidences indicate that Enk (Randic and Miletic, 1978) and NA LC neurons (Mokha et af., 1985) exert an inhibitory action on pain-evoked activity of dorsal horn interneurons. It appears that both Enk and NE transmitters may act in concert to produce the inhibitory response of pain-activated interneuronal activity. In corollary fashion, the co-
transmitter Enk may cooperate or amplify the postsynaptic effects of the classic NE transmitter towards its target neurons. Autoradiographic studies have shown that LC, SC, NKF and PBL project to the motoneuronal cell groups of the spinal cord (Holstege et af., 1979; Holstege and Kuypers, 1982) and each of these nuclei contains pontospinal neurons that co-contain TH and M-Enk (present study). Studies have established independently that catecholamine- (Mizukawa, 1980) and Enk-containing (Bouras et af., 1984; Honda and Lee, 1985; Atsumi and Sakamoto, 1987; Tashiro et af,, 1989) axon terminals contact spinal cord cells including motoneurons. Electrophysiological studies have indicated that both LC stimulation (Barnes et al., 1989) and exogenously applied NE (Connell and Wallis, 1989; Fung et af., 1989) excite spinal motoneurons. When applied iontophoretically, Enk is reported to produce no effect on the motoneuron resting membrane potential (Zieglgansberger and Tulloch, 1979; Duggan and Zhao, 1984), although an excitatory action is observed on Renshaw cells (Davies and Dray, 1976). Through the latter action, the Enk could reduce the segmental motor outflow indirectly. While the NE coeruleospinal action is excitatory on the motoneuron, the Enk co-transmitter may act antagonistically to limit any excessive firing of motoneurons. A recent study also points to the inhibitory p-opioid receptor regulation of NE from various terminal trajectories of LC neurons (Werling et af., 1987). Taken together, these results suggest that the co-contained classic NE and peptide Enk transmitters may interact pre- and postsynaptically upon release to maintain a coordinated motoneuronal output. Existence of NPY in the catecholaminergic coeruleospinal cells in the rat has been demonstrated indirectly on adjacent sections (Holets et af., 1988) but similar studies have not been published to date in the cat. However, our quantitative data revealed that 85% of all pontospinal neurons contain TH and approximately 20% contain NPY; we therefore suspect some of the cells
118
may contain both. Although, the physiological role of the presence of NPY in catecholaminergic neurons is not yet clear, independent applications of these neurochemicals are known to excite supraoptic vasopressin neurons, as compared with an occlusive interaction when applied simultaneously (Day et al., 1985). Conclusions
The various nuclear groups in the DLPT of the cat, referred to as the LC complex include the nuclei LC (LCd; LCa), SC, PBL, and PBM and the NKF. All pontospinal neurons, irrespective of the neurochemical content were observed in the ventrolateral part of the DLPT at coronal levels P1.8-P4.0. These neurons were located within the cytoarchitectural limits of the LCa, SC and NKF and of the three a majority of these neurons were observed in the NKF. A large proportion of the pontospinal neurons contain both TH- and M-Enk-like immunoreactivity (75%). Approximately 20% of the pontospinal neurons also exhibited NPY-like immunoreactivity. Catecholaminergic pontospinal neurons may, therefore, contain one or more putative peptide neurotransmitters. These studies provide an anatomical substrate for elucidating the physiological role of both catecholamine and peptides in the pontospinal neurons of the cat. References Atsumi, S. and Sakamoto, H. (1987) Enkephalin-like immunoreactive axon terminals make synapses with amotoneurons in the chicken. Brain Rex, 409: 187-192. Barnes, C.D., Fung, S.J. and Pompeiano, 0. (1989) Descending catecholaminergic modulation of spinal cord reflexes in cat and rat. Ann. N.Y. Acad. Sci., 563: 45-58. Basbaum, A.L. and Fields, H.L. (1979) The origin of the descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat: Further studies on the anatomy of pain modulation. J. Comp. Neurol., 187: 513532. Bastiaensen, E., Miserez, B. and Potter, W.D. (1988) SubcelMar fractionation of bovine ganglion stellatum: Co-storage of noradrenaline, Met-enkephalin and neuropeptide Y in large “dense-cored” vesicles. Bruin Res., 442: 124-130.
Bouras, C., Taban, C.H. and Constantinidis, J. (1984) Mapping of enkephalins in human brain. An immunohistofluorescence study on brain from patients with senile and presenile dementia. Neuroscience, 12: 179-190. Bowker, R.M., Westlund, K.N., Sullivan, M.D. and Coulter, J.D. (1982) A combined retrograde transport and immunocytochemical staining method for demonstrating the origin of serotonergic projections. J. Histochem. Cytochern., 30: 805-8 10. Cabana, T. and Martin, G.F. (1984) Developmental sequence in the origin of descending spinal pathways. Studies using retrograde transport techniques in the North American opossum (Didelphis virginiana). Deu. Bruin Res., 15: 247263. Charnay, Y., Ltger, L. and Dubois, P.M. (1981) Distribution de la substance P et de la methionine enkephalin dans la partie dorsolaterale du tegmentum pontique chez le chat. Etude par immunofluorescence. C.R. Acad. Sci. (Paris), 292: 435-440. Charnay, Y., Lkger, L., Dray, F., BCrod, A., Jouvet, M., Pujol, J.F. and Dubois, P.M. (1982) Evidence for the presence of enkephalin in catecholaminergic neurons of the cat locus coeruleus. Neurosci. Lett., 30: 147-151. Chen, K.S. and Stanfield, B.B. (1987) Evidence that selective collateral elimination during postnatal development results in a restriction in the distribution of locus coeruleus neurons which project to the spinal cord in rats. Bruin Re$., 410: 154-158. Chu, N .3 . and Bloom, F.E. (1974) The catecholamine-containing neurons in the cat dorsolateral pontine tegmenturn: Distribution of cell bodies and some axonal projections. Brain Res., 66: 1-21. Clark, F.M. and Proudfit, H.K. (1988) The projection of noradrenergic neurons in the Kolliker-Fuse nucleus (A7) to the spinal cord dorsal horn. SOC. Neurosci, Absir., 14: 778. Commissiong, J.W., Hellstrom, S.O. and Neff, N.H. (1978) P. new projection from locus coeruleus to the spinal ventral columns: Histochemical and biochemical evidence. Brain Res., 148: 207-213. Connell, L.A. and Wallis, D.I. (1989) 5-hydroxytryptaniine depolarizes neonatal rat motoneurons through a receptor unrelated to an identified binding site. NeuropharrnacolOD, 28: 625-634. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Actn Physiol. Scand., 62, Suppl. 232: 1-55. Davies, J. and Dray, A. (1976) Effects of enkephalin and morphine on Renshaw cells in feline spinal cord. Nature (London), 262 603-604. Day, T.A., Jhamandas, J.H. and Renaud, L.P. (1985) Comparison between the action of avian pancreatic polypeptide, neuropeptide Y and norepinephrine on excitability of rat supraoptic vasopressin neurons. Neurosci. Left., 62: 181185.
119 Duggan, A.W. and Zhao, Z.Q. (1984) Microelectrophoretic administration of naloxone near motoneurons fails to reproduce the effects of systemic naloxone in anesthetized cats. Neurosci. Lett., 45: 305-310. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Reu., 63(3): 844-914. Fritschy, J.-M., Lyons, W.E., Mullen, C.A., Kosofsky, B.E., Molliver, M.E. and Grzanna, R. (1987) Distribution of locus coeruleus axons in the rat spinal cord: A combined anterograde transport and immunohistochemical study. Brain Res., 437: 176-180. Fung, S.J. and Barnes, C.D. (1981) Evidence of facilitatory coeruleospinal action in lumbar motoneurons of cats. Brain Res., 216: 299-311. Fung, S.J., Manzoni, D., Chan, J.Y.H. and Barnes, C.D. (1989) Concordant depolarizing responses produced by locus coeruleus stimulation and norepinephrine in individual spinal motoneurons. Soc. Neurosci. Abstr., 15: 920. Fuxe, K. (1965) Evidence for the existence of monoaminecontaining neurons in the central nervous system. IV. The distribution of monoamine terminals in the central nervous system. Acta Physiol. Scand. 64 (Suppl. 247): 37-85. Glazer, E.J. and Basbaum, A.I. (1981) Immunohistochemical localization of leucine-enkephalin in the spinal cord of the cat: Enkephalin containing marginal neurons and pain modulation. J. Comp. Neurol., 196: 377-389. Guyenet, P.G. (1980) The coeruleospinal noradrenergic neurons: Anatomical and electrophysiological studies i n the rat. Brain Res., 189: 121-133. Hancock, M.B. and Fougerousse, C.L. (1976) Spinal projections from nucleus locus coeruleus and nucleus subcoeruleus in the cat and monkey as demonstrated by the retrograde transport of horseradish peroxidase. Brain Res. Bull., 1: 229-232. Hayes, N.L. and Rustioni, A. (1981) Descending projections from brainstem and sensorimotor cortex to spinal enlargements in the cat. Exp. Brain Rexb41: 89-107. Holets, V.R., Hokfelt, T., Rokaeus, A,, Terenis, L. and Goldstein, M. (1988) Locus coeruleus neurons in the rat containing neuropeptide Y, tyrosine hydroxylase or galanin and their efferent projections to the spinal cord, cerebral cortex and hypothalamus. Neuroscience, 24: 893-906. Holstege, G. and Kuypers, H.G.J.M. (1982) The anatomy of brain stem pathways to the spinal cord in the cat. A labeled amino acid tracing study. In H.G.J.M. Kuypers and G.F. Martin (Eds.), Descending Pathways to the Spinal Cord, Progress in Brain Research, Vol. 57, Elsevier Biomedical Press, New York, pp. 145-173. Holstege, G., Kuypers, H.G.J.M. and Boer, R.C. (1979) Anatomical evidence for direct brainstem projections to the motor neuronal cell groups and autonomic preganglionic cell groups in cat spinal cord. Bruin Res., 171: 329-333. Honda, C.N. and Lee, C.L. (1985) lmmunohistochemistry of synaptic input and functional characterization of neuron near the central canal. Brain Rex, 343: 120-128. Hunt, S.P. and Lovick, T.A. (1982) The distribution of sero-
tonin, metenkephalin and P-lipoprotein-like immunoreactivity in neuronal perikarya of the cat brainstem. Neurosci. Lett., 30: 139-145. Jones, B.E. and Beaudet, A. (1987) Distribution of acetylcholine and catecholamine neurons in the cat brainstem: A choline acetyltransferase and tyrosine hydroxylase immunohistochemical study. J. Comp. Neurol., 261: 15-32. Jones, B.E. and Cuello, A.C. (1989) Afferents to the basal forebrain cholinergic cell area from pontomesencephaliccatecholamine, serotonin, and acetylcholine-neurons. Neuroscience, 31: 37-61. Jones, B.E. and Moore, R.Y. (1974) Catecholamine-containing neurons of the nucleus locus coeruleus in the cat. J. Comp. Neurol., 157: 43-52. Kumakura, K., Guidotti, A. and Costa, E. (1979) Primary cultures of chromaffin cells: Molecular mechanisms for the induction of tyrosine hydroxylase mediated by 8-Br-cyclic AMP. Mol. Pharmacol., 16: 865-876. Kuypers, H.G.J.M. and Maisky,, V.A. (1975) Retrograde axonal transport of horseradish peroxidase from spinal cord to brainstem cell groups in the cat. Neurosci. Lett., 1: 9-14. Kuypers, H.G.J.M. and Maisky, V.A. (1977) Funicular trajectories of descending brainstem pathways in the cat. Brain Res., 136: 159-165. Lai, Y. and Barnes, C.D. (1985) A spinal projection of serotoninergic neurons of the locus coeruleus in the cat. Neurosci. Lett., 58: 159-164. LCger, L. and Hernandez-Nicaise, M.-L. (1980) The cat locus coeruleus: Light and electron microscopic study of the neuronal somata. Anat. Embryol., 159: 181-198. LCger, L., Wicklund, L., Descarries, L. and Persson, M. (1979) Description of an indolaminergic cell component in the cat locus coeruleus: A fluorescence histochemical and radioautographic study. Brain Res., 168: 43-56. LCger, L., Charnay, Y. and Dubois, P.M. (1981) Distribution of substance P, methionine-enkephalin and somatostatin in the dorsolateral tegmentum of the cat: An immunofluorescence study. Neurosci. Lett., Suppl., 7: 5260. G g er , L., Charnay, Y., Chayvialle, J.A., Btrod, A., Dray, F., Pujol, J.F., Jouvet, M. and Dubois, P.M. (1983) Localization of substance P- and enkephalin-like immunoreactivity in relation to catecholamine-containing cell bodies in the cat dorsolateral pontine tegmentum: An immunofluorescence study. Neuroscience, 8: 825-846. Livett, B.G., Dean, D.M., Whelan, L.G., Udenfriend, S. and Rosier, J. (1981) Co-release of enkephalin and catecholamines from cultured adrenal chromaffin cells. Nature (London), 289: 317-319. Maeda, T., Pin, C., Salvert, D., Migier, M. and Jouvet, M. (1973) Les neurones contentant des catecholamines du tegmentum pontique et leurs de proections chez le chat. Brain Res., 57: 119-152. Martin, G.F., Humbertson, A.O., Laxson, L.C., Panneton, W.M. and Ischismadia, I. (1979) Spinal projections from the mesencephalic and pontine reticular formation in the North American opossum: A study using axonal transport techniques. J. Comp. Neurol., 187: 373-400.
120 Martin, G.F., Cabana, T., DiTirro, F.J., Ho, R.H. and Humbertson, A.O. (1982) Reticular and raphe projections to the spinal cord of the North American opossum: Evidence of functional heterogeneity. In H.G.J.M. Kuypers and G.F. Martin (Eds.), Descending Puthwuys to the Spinal Cord, Progress in Bruin Research, Vol. 57, Elsevier Biomedical Press, New York, pp. 109-129. Mason, S.T. and Fibiger, H.C. (1979) Regional topography within noradrenergic locus coeruleus as revealed by retrograde transport of horseradish peroxidase. J. Comp. Neurol., 187: 703-724. McBride, R.L. and Sutin, J . (1976) Projections of the locus coeruleus and adjacent pontine tegmentum of the cat. J. Comp. Neurol., 165: 265-284. Mizukawa, K. (1980) The segmental detailed topographical distribution of monoaminergic terminals and their pathways in the spinal cord of the cat. Anut. Anz., 147: 125- 144. Mokha, S.S., McMillan, J.A. and Iggo, A. (1985) Descending control of spinal nociceptive transmission. Actions produced on spinal nociceptive neurons from the nuclei locus coeruleus (LC) and raphe magnus (NRM). Exp. Bruin Res., 58: 213-223. Nakazato, T. (1987) Locus coeruleus neurons projecting to the forebrain and the spinal cord in the cat. Neuroscience, 23: 529-538. Nygren, L. and Olson, L. (1977) A new major projection from locus coeruleus: The main source of noradrenergic nerve terminals in the ventral and dorsal columns of the spinal cord. Bruin Res., 132: 85-93. Pickel, V.M., Segal, M. and Bloom, F.E. (1974) A radiographic study of the efferent pathways of the locus coeruleus. J. Comp. Neurol., 155: 15-42. Quach, T.T., Tang, F., Kageyama, H., Mocchetti, I., Guidotti, A,, Meek, J.L., Costa, E. and Schwartz, J.P. (1984) Enkephalin biosynthesis in adrenal medulla, modulation of proenkephalin mRNA content of cultured chromaffin cells by 8-bromo-adenosine 3',5'-monophosphate. Mol. Phurmacol., 26: 255-260. Ramon-Moliner, E. (1974) The locus coeruleus of cat. 111. Light and electron microscopic studies. Cell Tissue Res., 149 205-221. Randic, M. and Miletic, V. (1978) Depressant action of methionine-enkephalin and somatostatin in cat dorsal horn neurons activated by noxious stimuli. Bruin Res., 152: 196-202. Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1989) Spinally projecting noradrenergic neurons of the dorsolatera1 pontine tegmentum: A combined immunocytochemical and retrograde labeling study. Bruin Res., 491: 144-149. Reddy, V.K., Cassini, P., Ho, R.H. and Martin, G.F. (1990a) Origins and terminations of bulbospinal axons that contain serotonin and either enkephalin or substance-P in the North American opossum. J. Comp. Neurol., 294: 96-108. Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1990b). Localization of enkephalinergic neurons in the dorsolatera1 pontine tegmentum projecting to the spinal cord of the cat. J. Comp. Neurol., 291: 195-202.
Reiner, P.B. and Vincent, S.R. (1987) Topographic relations of cholinergic and noradrenergic neurons in the pontomesencephalic tegmentum: An immunohistochemical study. Bruin Res. BUN., 19: 705-714. Room, P., Postema, F. and Korf, J. (1981) Divergent axon collaterals of rat locus coeruleus neurons: Demonstration by fluorescent double labeling technique. Brain Res., 221: 219-230. Russel, G.V. (1955) The nucleus locus coeruleus (dorsolateralis tegmenti). Tex.Rep. Biol. Med., 13: 939-988. Sakai, K., Sastre, J.P., Salvert, D., Touret, M., Tohyama, M. and Jouvet, M. (1979) Tegmentoreticular projections with special reference to the muscular atonia during paradoxical sleep in the cat: An HRP study. Bruin Rex, 176: 233-254. Staines, W.A., Meister, B., Melander, T., Nagy, J.I. and Hokfelt, T. (1988) Three-color immunofluorescence histochemistry allowing triple labeling within a single section. J. Histochem. Cytochem., 36(2), 145-151. Sternberger, L.A. (1979) Imrnunohistochemistry, John Wiley & Sons, New York, 354 pp. Stevens, R.T., Hodge, C.J., Jr. and Apkarian, A.V. (1982) Kolliker-Fuse nucleus: The principal source on pontine catecholaminergic cells projecting to the lumbar spinal cord of the cat. Bruin Res., 239: 589-594. Stevens, R.T., Apkarian, A.V. and Hodge, C.J., Jr. (1985) Funicular course of catecholaminergic fibers terminating in the lumbar spinal cord of the cat. Bruin Res., 336: 243-251. Takeuchi, Y., Uemura, M., Matsuda, K., Matsushima, R. and Mizuno, N. (1980) Parabrachial nucleus neurons projecting to the lower brainstem and the spinal cord: A study in the cat by Fink-Heimer and the horseradish peroxidase methods. Exp. Neurol., 70: 403-413. Tashiro, T., Ruda, M.A., Satoda, T., Matsushima, R. and Mizuno, N. (1989) Convergence of serotonin-, enkephalinand substance P-like immunoreactive afferent fibers onto cat medullary dorsal horn projection neurons: A triple immunocytochemical staining technique combined with the retrograde HRP-tracing method. Bruin Res., 491: 360-365. Tohyama, M., Sakai, K., Touret, M., Salvert, D. and Jouvet, M. (1979) Spinal projections from the lower brainstem in the cat as demonstrated by the horseradish peroxidase technique. 11. Projections from the dorsolateral pontine tegmentum and raphe nuclei. Bruin Rex, 176: 215-231. Ungerstedt, U. (1971) Stereotaxic mapping of the monoamine pathways in the rat brain. Actu Physiol. Scund., Suppl. 367: 1-48. Werling, L.L., Brown, S.R. and Cox, B.M. (1987) Opioid receptor regulation of the release of norepinephrine in brain. Neurophurmucology, 26: 987-996. Wessendorf, M.W. and Elde, R.P. (1985) Characterization of an immunofluorescence technique for the demonstration of coexisting neurotransmitters within nerve fibers and terminals. J. Histochem. Cytochem., 33: 984-994. Westlund, K.N. and Coulter, J.D. (1980) Descending projections of the locus coeruleus and subcoeruleus/medial parabrachial nuclei in monkey: Axonal transport studies
121 and dopamine-p-hydroxylase immunocytochemistry. Bruin Rex RW., 2: 234-264. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1982) Descending noradrenergic projections and their spinal terminations. In H.G.J.M. Kuypers and G.F. Martin (Eds.), Descending Pathways to the Spinal Cord, Progress in Bruin Research, Vol. 57, Elsevier Biomedical Press, New York, pp. 219-238. Wiklund, L., LCger, L. and P e r s o n , M. (1981) Monoamine cell distribution in the cat brainstem: A fluorescencehistochemical study with quantification of indolaminergic
and locus coeruleus cell groups. J. Comp. Neurol., 203: 613-647. Zhuo, H., Fung, S.J., Reddy, V.K. and Barnes, C.D. (1991) Immunohistochemical evidence for coexistence of methionine enkephalin and tyrosine hydroxylase in neurons of the locus coeruleus complex projecting to the spinal cord of the cat. J. Chem. Neuroanut., (in press). Zieglgansberger, W. and Tulloch, I.F. (1979) The effects of methionine-and leucine-enkephalin on spinal neurons of the cat. Brain Rex, 167: 533-564.
This Page Intentionally Left Blank
C.D. Barnea and 0. Pompeiano (Edc.) Progress hi Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
123
CHAPTER 8
The projections of locus coeruleus neurons to the spinal cord H.K. Proudfit and F.M. Clark Department of Pharmacology, University of Illinois College of Medicine at Chicago, S. Wolcott, Chicago, IL, U.S.A
Spinally projecting noradrenergic neurons located in the locus coeruleus/subcoeruleus (LC/SC) are a major source of the noradrenergic innervation of the spinal cord. However, the specific terminations of these neurons have not been clearly defined. The purpose of this chapter is to describe the results of experiments that used the anterograde tracer Phaseolus vulgaris leucoagglutinin in combination with dopamine-p-hydroxylase immunocytochemistry to more precisely determine the spinal cord terminations of neurons located in the LC/SC. The results of these experiments indicate that the axons of LC neurons are located primarily in the ipsilateral ventral funiculus and terminate most heavily in the medial part of laminae VII and VIII, the motoneuron pool of lamina
IX, and lamina X. LC neurons provide a moderately dense innervation of the ventral part of the dorsal horn, but only a very sparse innervation of the superficial dorsal horn. The SC projects ipsilaterally in the ventrolateral funiculus and terminates diffusely in the intermediate and ventral laminae of the spinal cord. Finally, the results of preliminary experiments indicate that different rat substrains may have LC neurons that exhibit qualitatively different termination patterns in the spinal cord. More specifically, LC neurons in some rat substrains innervate the dorsal horn, while those in other substrains primarily innervate the ventral horn and intermediate zone.
Key words: anterograde tracer, dopamine-p-hydroxylase, dorsal horn, Fluoro-Gold, intermediate zone, norepinephrine, Phaseolus culgaris leucoagglutinin, retrograde tracer, spinal cord, ventral horn
Introduction The innervation of the spinal cord by noradrenergic neurons is extensive, but mainly concentrated in the superficial dorsal horn, the ventral horn motoneuron pool and lamina X of all spinal cord segments, as well as the intermediolateral cell column of thoracic and sacral segments (Carlsson et al., 1964; Fuxe, 1965; Nygren and Olson, 1977; Westlund et al., 1982, 1983, 1984; Schrader and Skagerberg, 1985; Fritschy and Grzanna, 1990) (Fig. 1). Because there are no known cate-
cholamine cell bodies in the spinal cord, this innervation is presumed to arise from one or more of the brainstem catecholamine cell groups (Al-A7) first described by Dahlstrom and Fuxe (1964). The most compelling evidence indicates that only the pontine catecholamine cell groups (A5, A7 and nucleus locus coeruleus (A6)) have spinal projections (Westlund et al., 1981, 1982, 1983, 1984; Fritschy and Grzanna, 1990). The extensive innervation of the spinal cord by descending noradrenergic neurons suggests that norepinephrine is involved in a wide variety of
124
neuronal processes that take place in the spinal cord. Thus, a precise description of the descending projections of noradrenergic neurons that arise from these catecholamine cell groups is an important prerequisite to understanding the physiological role of noradrenergic neurons in the control of sensory, motor and autonomic processes in the spinal cord. This chapter will focus on the projection of locus coeruleus (LC) neurons to the spinal cord with particular emphasis on their termination sites, funicular trajectory and functional significance. Evidence for projections of LC neurons to the spinal cord
It is well-documented that a large number of LC neurons innervate forebrain structures (Foote et
al., 1983; Loughlin and Fallon, 1985). It is simiIarIy well-documented that neurons in the LC also project to the spinal cord. Indeed, the innervation of the spinal cord by noradrenergic neurons located in the nucleus LC has been studied extensively using a variety of methods such as: retrograde tracers (Kuypers and Maisky, 1975; Hancock and Fougerousse, 1976; Satoh et al., 1977; Guyenet, 1980; Nagai et al., 1981; Room et al., 1981; Westlund et al., 1981, 1982, 1983, 1984; Stevens et al., 1982, 1985; Jones and Yang, 1985; Jones et al., 1986; Kausz, 1986; Loughlin et al., 1986; Nakazato, 1987; Holets et al., 1988; Lyons and Grzanna, 1988; Lyons et al., 1989; Reddy et al., 1989; Clark and Proudfit, 1989, 1991; Fritschy and Grzanna, 1990; Proudfit et al., 1990), electrolytic lesions (Ross and Reis, 1974; Nygren and Olson, 1977; Commissiong et al., 1978; AdCr et
Fig. 1. The noradrenergic innervation of the lumbar spinal cord. Darkfield photomicrographic montage showing doparnine-p-hydroxylase-immunoreactive (DPH-ir) axons stained using the peroxidase-antiperoxidase method. The calibration bar is 500 y m .
125
al., 1979; Karoum et al., 1980; Commissiong, 1981; Clark and Proudfit, 1989, 19911, anterograde transport of tritiated amino acids (Pickel et al., 1974; Westlund and Coulter, 1980; Jones and Yang, 19851, anterograde transport of Phaseolus uulgaris leucoagglutinin (PHA-L) (Fritschy et al., 1987; Clark and Proudfit, 1989, 1991; Fritschy and Grzanna, 1990) and electrophysiological recording of LC neurons that were antidromically activated from the spinal cord (Guyenet, 1980). The LC neurons that project to the spinal cord appear to constitute a distinct subpopulation of LC neurons. This conclusion is supported by the observations that only a few LC neurons are double-labeled after dual local injections of retrograde tracers into the spinal cord and one of several forebrain regions such as the frontal cortex (Nagai et al., 1981; Room et al., 19811, thalamus (Room et al., 1981; Jones and Yang, 19851, cerebellum (Nagai et al., 1981) or hippocampus (Room el al., 1981). There is some controversy concerning the projections of LC neurons to the spinal cord and their termination sites. The majority of the evidence indicates that spinally projecting noradrenergic neurons are located in the ventral and caudal aspects of the LC, have axons that are located in the ipsilateral ventral quadrant of the spinal cord and terminate primarily in the ventral horn, the intermediate zone and lamina X near the central canal. Although most of the evidence indicates that the innervation of the dorsal horn is sparse, there is conflicting evidence that the dorsal horn is heavily innervated by descending LC neurons (Fritschy et al., 1987; Fritschy and Grzanna, 1990). The purpose of the following discussion is to present the anatomical evidence for the projections of LC neurons to the spinal cord and to discuss some recent observations that may resolve some of the conflicting results. Location of coeruleospinal neurons The location of spinally projecting noradrenergic neurons in the LC has been determined using retrograde tracers injected into the spinal cord in
conjunction with a marker for noradrenergic neurons, such as dopamine-p-hydroxylase-immunoreactivity (DpH-ir) (Hartman, 1973; Swanson and Hartman, 1975). The demonstration of neurons in the LC that contain both retrograde label and DPH-ir, is considered to be evidence for spinally projecting noradrenergic neurons. Studies using these methods have demonstrated that most spinally projecting noradrenergic neurons that originate in the LC are located in the caudal and ventral parts of this nucleus. Thus, injection of retrograde tracers into the spinal cord, such as Fast Blue (Holets et al., 19881, True Blue (Lyons and Grzanna, 1988; Lyons et al., 1989; Fritschy and Grzanna, 1990), Evans Blue (Room et al., 19811, DPH antisera (Westlund et al., 1982, 1983), horseradish peroxidase (HRP; Satoh et al., 1977; Mason and Fibiger, 19791, primuline (Nagai, 1981) or [3H]norepinephrine (Jones et al., 1986) primarily labels identified noradrenergic neurons in the caudal and ventral parts of the LC in rat (Fig. 2C and D). The location of spinally projecting noradrenergic neurons in the LC is similar in cat (Stevens et al., 1985) and monkey (Westlund et al., 1984). Spinally projecting LC neurons appear to innervate all levels of the spinal cord, since the location and distribution of labeled neurons in the LC is approximately the same regardless of whether retrograde tracer injections are made into cervical, thoracic or lumbar segments (Westlund et al., 1982; Reddy et al., 1989). The relative contribution of LC neurons to the noradrenergic innervation of the spinal cord appears to be in the range of 30 to 40% according to estimates by a number of investigators using different methods. For example, bilateral lesions of the LC reduce spinal cord norepinephrine content by 27-39% (Commissiong et al., 1978; Ad& et al., 1979; Hammond and Proudfit, 1980). Similar estimates in the range of 26-41% have been obtained by determining the percentage of noradrenergic neurons in the LC, subcoeruleus (SC), A5 and A7 nuclei that are also labeled by retrograde tracers injected into the spinal cord (Westlund et al., 1983, 1984; Fritschy and
126
PHA-L
A
B
Cresyl violet
TH
C
D
Fluoro-gold
Fig. 2. The location of a representative injection of PHA-L in the LC. Panel A shows the location of three small PHA-L deposits in the LC (arrows). Panel B shows the LC in a transverse section stained with cresyl violet. Panel C shows noradrenergic neurons in the LC stained for tyrosine hydroxylase immunoreactivity (TH-ir). Panel D shows the location of coeruleospinal neurons in the ventral LC. These neurons were retrogradely labeled by Fluoro-Gold injected into the ipsilateral side of the spinal cord ventral horn. The arrows indicate two Fluoro-Gold-labeled neurons that were not labeled with TH-ir. The calibration is 200 g m . (From Clark and Proudfit, 1991.)
127
Grzanna, 1990). These results are in reasonably good agreement, despite the use of different methods, and they indicate that noradrenergic neurons located in the LC make a significant contribution to the noradrenergic innervation of the spinal cord. Funicular trajectory of LC axons in the spinal cord The funicular trajectory of LC axons in the spinal cord has been determined using a variety of experimental approaches. The results of most studies have demonstrated that the axons of spinally projecting LC neurons in rat, cat and monkey course through the ventral and ventrolatera1 funiculi. For example, in rats, bilateral LC lesions reduce the number of axons in the ventral and ventrolateral funiculi which contain catecholamines as determined by fluorescence histochemistry (Nygren and Olson, 1977). These authors found no evidence for the existence of LC axons in the dorsolateral funiculus. The use of retrograde tracers injected into various spinal cord funiculi is another approach used to determine the funicular pathways in which descending LC axons travel. For example, studies in the rat and cat demonstrated that HRP injected into the ventrolateral funiculus produces substantial labeling in the ipsilateral LC (Guyenet, 1980; Nakazato, 1987). Similar studies in cat demonstrated that lesions of the ventral and ventrolateral funiculi eliminate the labeling of LC neurons following unilateral injections of fluorescent retrograde tracers into the lumbar spinal cord, whereas dorsolateral funiculus lesions did not reduce the retrograde labeling of LC neurons (Stevens et aL, 1985). The results of these studies indicate that the majority of spinally projecting noradrenergic neurons in the LC have axons located in the ventral quadrant of the spinal cord. Anterograde tracers have also been used to determine the funicular location of descending LC axons. For example, studies in which tritiated proline and leucine were microinjected into the LC or SC complex in monkeys demonstrate label-
ing in the ventrolateral and lateral funiculi (Westlund and Coulter, 1980). Similar studies indicate that LC axons in the rat also travel in the ventral funiculus (Pickel et al., 1974). These injections did not spread beyond the LC and included only LC neurons that were identified by fluorescence histochemistry. By comparison, in rats, larger injections of tritiated leucine that included the LC, mesencephalic trigeminal nucleus, SC, medial parabrachial, caudal laterodorsal tegmental nucleus and periventricular gray produced labeling in axons located in the lateral funiculus (Jones and Yang, 1985). Although the use of tritiated amino acids for tracing the pathways and terminations of axons has been a useful technique, its use does not allow identification of the neurotransmitter content of labeled axons. However, the combined use of the anterograde tracer PHA-L with immunocytochemistry does allow the identification of neurotransmitters in anterogradely labeled axons (Gerfen and Sawchenko, 1984). This tracer has recently been used by Fritschy and coworkers (Fritschy et al., 1987; Fritschy and Grzanna, 1990) to examine the funicular trajectory of spinally projecting LC neurons in rats. The results of these studies indicated that unilateral iontophoretic deposits of PHA-L in the LC produced the heaviest labeling of DpH-ir axons in the superficial layers (laminae I and 11) of the spinal cord dorsal horn (Fritschy and Grzanna, 1990). Labeled axons were also seen in the dorsolateral funiculus and to a lesser extent in the medial part of the ventral funiculus. These authors concluded that the spinally projecting axons from LC neurons travel mainly in laminae I and I1 of the superficial dorsal horn. These results are in sharp contrast to those obtained using less direct methods which indicate that the spinal cord projections of LC neurons travel in the ventral quadrant. We have also used anterograde transport of PHA-L to examine the spinal projections of LC neurons (Clark and Proudfit, 1989, 1990, but our results are fundamentally different from those
128
A
m B
C
Fig. 3. The projections of coeruleospinal neurons demonstrated by unilateral PHA-L injections in the LC. Three different PHA-L injection sites are illustrated (A, B and C). Enlarged drawings of each injection site are shown to the right of each brain stem section, and camera lucida drawings that illustrate the distribution of PHA-L-labeled terminals in the lumbar segments of the spinal cord are shown in the far right column. The PHA-L deposits are drawn in solid black and the location of DPH-ir neurons in the locus coeruleus (LC) are represented by the stippled region. The drawing of each spinal cord section represents a composite of 12 sections. (From Clark and Proudfit, 1991.)
129
A
Fig. 4. Anterograde transport to the spinal cord of PHA-L from injection sites located either dorsal or ventral to the LC. A. PHA-L injection that encompassed DPH-ir neurons (filled circles) in the nucleus subcoeruleus (SO. B. PHA-L injection located ventral to DPH-ir neurons in the LC and SC. C. PHA-L injection located dorsal to the DPH-ir neurons in the LC. The legend for Figure 3 contains additional details. (From Clark and Proudfit, 1991.)
reported by Fritschy and coworkers (Fritschy et al., 1987; Fritschy and Grzanna, 1990). In our studies, PHA-L was iontophoretically injected
into the area of the LC or the SC region from which spinally projecting noradrenergic neurons originate (Fig. 2). LC axons in the spinal cord
130
were identified by PHA-L labeling and D P H immunoreactivity. Figure 3 illustrates the results of three such experiments in which PHA-L injections encompassed DpH-ir neurons in the LC. These injections included most of the ventral caudal aspect of the LC. A fourth PHA-L injection was centered on the SC, but did not include DPH-ir neurons in the main body of the LC (Fig. 4A). Additional injections of PHA-L in two other rats were located ventral and dorsal to the DPH-ir LC neurons (Fig. 4, B and C). Following PHA-L injections which included DPH-ir neurons in the caudal, part of the ventral LC, PHA-L-labeled axons were seen to course primarily through the medial ventral funiculus. Only occasional PHAL-labeled axons were seen in the dorsolateral or lateral funiculus. To summarize, most studies using a variety of experimental approaches, such as LC lesions, retrograde tracers and anterograde tracers, have demonstrated that axons of noradrenergic LC neurons course through the ventral quadrant of the spinal cord in rat, cat and monkey. However, there is conflicting evidence which indicates that the axons of spinally projecting LC neurons are located almost exclusively in the superficial laminae of the dorsal horn or the dorsolateral funiculus of the rat. The reason for these conflicting results may be that there are fundamental differences in the spinal cord projections of LC neurons in different substrains of rats. A more detailed discussion of this possibility is presented in a later section of this chapter. Laterality of coeruleospinal axons There is a substantial amount of evidence which supports the conclusion that the funicular trajectory of coeruleospinal neurons is predominantly ipsilateral. For example, unilateral LC lesions in the rat produced a 65% reduction in DPH activity on the ipsilateral side of the spinal cord, but no change on the contralateral side (Ross and Reis, 1974). In addition, unilateral injections of retrograde tracers into the spinal cord produced predominantly ipsilateral retro-
grade labeling of LC neurons in the rat (Guyenet, 1980; Jones and Yang, 1985; Jones et al., 1986; Holets et al., 19881, cat (Kuypers and Maisky, 1975; Hancock and Fougerousse, 1976; Stevens et al., 1985; Kausz, 1986; Nakazato, 1987; Reddy et al., 1989) and monkey (Hancock and Fougerousse, 1976). Additional evidence for an ipsilatera1 coeruleospinal projection is provided by anterograde tracer studies in which unilateral microinjections of tritiated amino acids into the LC produced ipsilateral labeling in the spinal cord in the rat (Pickel et al., 1974; Jones and Yang, 19851, cat (Holstege and Kuypers, 1982) and monkey (Westlund and Coulter, 1980). We have done similar experiments using unilateral injections of the anterograde tracer PHA-L in the ventral caudal part of the LC. These injections included DPH-ir neurons in the ventral and caudal part of the LC (Fig. 2) and produced primarily ipsilateral labeling in the spinal cord (Fig. 3). These results were corroborated by experiments in which unilateral injections of the fluorescent retrograde tracer Fluoro-Gold were made into the spinal cord. These injections were centered on the ventral horn and included part of the intermediate zone, but did not spread to the contralateral side. As illustrated in Table 1 (see results from Sasco rats), the majority (72%) of DPH-ir neurons in the LC that were retrogradely labeled with Fluoro-Gold were found on the side ipsilateral to the tracer injection in the ventral horn. These results support the conclusion that most of the descending noradrenergic axons from the LC do not cross to the contralateral side. Although most of the evidence strongly supports an ipsilateral coeruleospinal projection, there are several reports that provide evidence for bilateral LC projections. For example, unilateral LC lesions have been reported to produce bilateral decreases in spinal cord norepinephrine (Karoum et al., 1980; Commissiong, 1981). In addition, two recent reports indicate that a unilateral injection of PHA-L in the LC produces bilateral labeling in the rat spinal cord (Fritschy et al., 1987; Fritschy and Grzanna, 1990). These
131
authors also demonstrated that unilateral injections of the retrograde tracer True Blue into the spinal cord produced bilateral labeling of DPH-ir neurons in the LC (Fritschy and Grzanna, 1990). In summary, the majority of the evidence obtained using three different animal species (rat, cat and monkey) and various experimental approaches, such as anterograde tracers, retrograde tracers and lesion-induced norepinephrine depletion, has consistently demonstrated that descending noradrenergic neurons project mainly on the ipsilateral side. However, there is some contradictory evidence obtained from rats that supports a bilateral projection. Preliminary evidence discussed in a later section suggests that the conflicting results may be explained by fundamental differences in descending LC projections in different substrains of rats.
Terminal fields of coeruleospinal neurons The location of the terminal fields in the spinal cord that are innervated by descending noradrenergic LC neurons has been described in a large number of reports. These reports generally agree that LC neurons terminate primarily in the ventral horn and intermediate zone, but only sparsely innervate the dorsal horn.
For example, most of the evidence indicates that the innervation of the dorsal horn by LC neurons is only low-to-moderate. Thus, unilateral (Commissiong et af., 1978) or bilateral (Nygren and Olson, 1977) LC lesions produced small-tomoderate decreases in norepinephrine-containing axons in the dorsal horn of rats. Furthermore, injections of tritiated amino acids into the LC of rat (Jones and Yang, 1985) or monkey (Westlund and Coulter, 1980) produced only light-to-moderate labeling in the dorsal horn. Although these results indicate that the dorsal horn does not receive a major innervation from LC neurons, Fritschy and coworkers (Fritschy et af., 1987; Fritschy and Grzanna, 1990) demonstrated a heavy projection to the dorsal horn by LC neurons in rat using combined anterograde transport of PHA-L and DPH-immunocytochemistry. We have done similar experiments using injections of the anterograde tracer PHA-L combined with DPH immunocytochemistry to determine the areas in the spinal cord where spinally projecting LC axons terminate. The anterograde tracer PHA-L was iontophoretically injected into the area of the LC and SC, and the spinal cord terminations of neurons in the injected areas were determined. Figures 3 and 4 illustrate the
TABLE 1 Number of retrogradely labeled TH-IR neurons ventral horn Vendor
Injection site
N h
Ventral Dorsal
2 2
Ventral Dorsal
2 2
a
in the LC following a unilateral injection of Fluoro-Gold into the dorsal or
Ipsilateral LC
Contralateral LC
Total
5 (80) 94+ 1 6 4 )
16 k 9(20) 80 f 2 (46)
7.5 f 13 (30) 174 3 (70)
181 f 48 (72) 51 f 2(64)
70 17 (28) 29 f 8(36)
Harlan
59,
Sasco
+
251 f 65 (76) 80 k 10 (24)
Values represent the number of double-labeled neurons in 24 transverse sections. Number of animals. Numbers in parentheses indicate the percentage of retrogradely-labeled TH-ir LC neurons that were found on the ipsilateral or contralateral side. Numbers in parentheses indicate the percentage of the total number of double-labeled LC cells that were retrogradely labeled by Fluoro-Gold injections in either the dorsal or ventral horn. a
132
A
D
8
C
E
F
133
results of four experiments in which PHA-L injections encompassed DPH-ir neurons in the LC or SC. In three of these experiments, the injections included most of the ventral caudal aspect of the LC (Fig. 3A, B and C). A fourth PHA-L injection (Fig. 4A) was centered on the SC, but did not include DPH-ir neurons in the main body of the LC. Additional injections of PHA-L in two other cases were located either ventral or dorsal to the DPH-ir LC neurons (Fig. 4B and C). Following PHA-L injections in the caudal part of the ventral LC, PHA-L-labeled axons were seen to course through the medial ventral funiculus. The terminations of these axons were found mainly on the side ipsilateral to the injection site in the medial aspect of the ventral horn (laminae VIII and IX) and intermediate zone (lamina VII) and in lamina X near the central canal (Fig. 3). There was a moderate projection to the motoneuron pool of lamina IX, but there were very few PHA-L-labeled axons in the dorsal horn (laminae I-V). However, low-to-moderate labeling was present in the ventromedial part of the dorsal horn (lamina VI). Only sparse labeling was found on the side contralateral to the injection site in the three cases illustrated in Figure 3. The distribution of labeled axons and terminals in the spinal cord following an injection of PHA-L in the SC was somewhat different from that produced by similar injections in the LC (Fig. 4). The labeled axons in the lumbar spinal cord coursed through the ipsilateral ventrolateral funiculus and terminated ipsilaterally in both the medial and lateral part of the ventral horn, including the motoneuron pool of lamina IX. In addition, there was sparse labeling in the dorsal horn, but the density of these terminals appeared
to be higher than that produced by injections in the LC. There was also some sparse labeling in the contralateral intermediate zone. In contrast, injection of PHA-L into sites ventral or dorsal to the LC/SC complex resulted in very little labeling in the spinal cord (Fig. 4). These two PHA-L injections did not encompass any DPH-ir neurons in either the LC or SC. Coeruleospinal neurons appear to project to all levels of the spinal cord. This conclusion is supported by the observation that injections of PHA-L in the LC resulted in consistent labeling in cervical, thoracic and lumbar segments of the spinal cord (Fig. 5). The labeling in the cervical segments of the spinal cord was predominantly ipsilateral and was more dense than that in the lumbar segments. However, the distribution of ipsilateral labeling was similar in all three spinal cord levels. In addition, there was moderate contralateral labeling in both the lateral aspect of the dorsal horn and the medial aspect of the ventral horn in the cervical segments. Such contralateral labeling was not present in either the thoracic or lumbar segments. Finally, very few labeled axons or terminals were seen in either the intermediolateral cell column or the dorsal horn of the thoracic segments after PHA-L injections in either the LC or SC. Several cases were examined to ascertain whether the PHA-L-labeled axons and terminals in the spinal cord also contained DpH-ir. These double-labeling experiments were done on spinal cord sections taken from the same animal illustrated in Figure 3A (Fig. 6). Nearly 80% of PHAL-labeled axons and terminals also contained DpH-ir, which indicates that the majority of PHA-L-labeled axons and terminals in the spinal
Fig. 5. The projections of coeruleospinal neurons to cervical, thoracic and lumbar segments demonstrated by unilateral PHA-L injection in the LC. A. PHA-L injection site. B. Enlarged drawing of the injection site. The solid black area represents the extent of the PHA-L injection site and the stippled area represents the location of DpH-ir neurons in the LC. C . Darkfield photomicrograph of a thoracic spinal cord section showing the distribution of PHA-L-labeled axons and terminals. D. PHA-L-labeled axons in cervical spinal cord segments. E. PHA-L-labeled axons in thoracic spinal cord segments. F. PHA-L-labeled axons in lumbar spinal cord segments. Each drawing is a composite of 12 contiguous sections. This case is the same as that illustrated in Figure 3A. (From Clark and Proudfit, 1991.)
134
cord contain catecholamines. This conclusion is supported by the observation that the distribution of double-labeled terminals was similar to the distribution of axons seen in sections only labeled with PHA-L. These results indicate that some spinally projecting non-catecholamine-containing neurons in or near the LC also transport PHA-L to the spinal cord. The number and location of these neurons was estimated by making large bilateral injections of the retrograde tracer Fluoro-Gold into the lumbar spinal cord. Figure 7 shows the location of neurons Iabeled with retrogradely transported Fluoro-Gold, those labeled with tyroA
sine hydroxylase-immunoreactivity (TH-ir), a marker for noradrenergic neurons, and those double-labeled with both Fluoro-Gold and TH-ir. At the level of the caudal LC the majority of retrogradely labeled neurons in or near the LC/SC also contained TH-ir. The retrogradely labeled TH-ir neurons were located mainly in the ventral caudal aspect of the LC and in the SC (Fig. 7B and C). At these caudal locations, only a few retrogradely labeled neurons near the LC did not contain TH-ir. At the rostra1 pole of the LC (Fig. 7A) a dense cluster of retrogradely labeled neurons that did not contain TH-ir was located just ventral and C
D
Fig. 6 . Examples of DPH-ir axons in the ventral horn that were ante rogradely labeled by an injection of PHA-L in the LC. Panel A shows an axon that was anterogradely labeled by an injection of PHPL-Linto the ventral LC and in panel B the same axon is seen to also contain DPH-ir. Panel C shows examples of two other anterogr;idely labeled axons in the ventral horn (arrows) and in panel D these two axons also exhibit DPH-ir. Several axons that exhibit Dpl44r, but were not anterogradely labeled from the LC, are also present in panel D. The calibration bar is 40 p m . (From Clark and Proudfit, 1991.)
135
-0.25
++a*,.
-0.50
D
Fig. 7. The location of neurons in the pontine tegmentum that project to the spinal cord. Spinally projecting neurons were retrogradely labeled following bilateral injections of FluoroGold into the lumbar segments of the spinal cord. The location of retrogradely labeled neurons and those that exhibited TH-ir are plotted on transverse brainstem sections (A, B and C). Crosses represent neurons retrogradely labeled with Fluoro-Gold, open circles represent TH-ir somata and filled circles represent TH-ir somata that were also labeled with Fluoro-Gold. The Fluoro-Gold injection site is shown in panel D. The numbers to the right of each section are the anteriorposterior stereotaxic coordinates for that section. (From Clark and Proudfit, 1991.)
medial to the LC. This group of neurons appears to be the lateral dorsal tegmental nucleus (TLD) first described by Barrington (1925). It is unlikely that PHA-L-labeled terminals in the ventral horn originated from these neurons since PHA-L in-
jection sites were located caudal to this cell cluster. Furthermore, the neurons located in the TLD appear to project primarily to the intermediolateral cell column in sacral segments, but not to the ventral horn in lumbar or cervical segments (Loewy et al., 1979). In view of the sparse number of spinally projecting non-catecholamine-containing neurons near the LC, it is not surprising that nearly 80% of the axons in the ventral horn that were labeled with anterogradely transported PHA-L also contained DPH-ir. The projections of LC neurons to specific regions of the spinal cord were also examined using unilateral electrolytic lesions to destroy the LC/SC complex and by determining the subsequent changes in the density of DPH-ir axons and terminals in two regions of the spinal cord (Fig. 8). In each of five cases, complete unilateral destruction of the LC/SC complex produced a consistent and statistically significant reduction in the density of DPH-ir axons and terminal segments in the ipsilateral ventral horn of the lumbar spinal cord. By comparison, the density of DpH-ir axons in the ipsilateral dorsal horn of the lumbar spinal cord was slightly, but significantly reduced in only two of five animals (P
136
ina VII and X in monkeys (Westlund and Coulter, 1980). In addition, PHA-L injections in the LC produced labeling of DPH-ir axons in lamina X (Fritschy et aL, 1987; Fritschy and Grzanna, 1990). However, there is also contradictory evidence which indicates that LC neurons have only sparse projections to lamina X (Nygren and 01son, 1977). Finally, our experiments provide direct evidence for a major projection from pontospinal LC neurons to the ventral horn of the spinal cord. These results are consistent with those of other studies using more indirect anatomical methods. For example, lesions of the LC reduced the norepinephrine content in the ventral horn of rodents by 95-100% of control, which indicates
that a large percentage of the noradrenergic innervation of the ventral horn originates from pontospinal LC neurons (Nygren and Olson, 1977; Commissiong et al., 1978; Commissiong, 1981). In addition, microinjection of tritiated amino acids into the LC produced moderate to heavy labeling in the ventral horn in rat (Pickel et al., 1974; Jones and Yang, 19851, cat (Holstege and Kuypers, 1982) and monkey (Westlund and Coulter, 1980). Finally, injections of HRP into the ventrolateral funiculus in rat (Guyenet, 1980) and cat (Nakazato, 1987) produced labeling of neurons in the ventral LC. The results of these studies and the evidence cited previously which indicates that the axons of pontospinal LC neurons are located in the ventral funiculus, also
D
**
A
6050403020100
60 50LO-
30-
Ventral horn
Dorsal horn
o Contralateral lpsilateral p
*
20100 Fig. 8. Effects of unilateral lesions that included the LC and SC on the density of DPH-ir axons in the dorsal and ventral horns of lumbar spinal cord segments. Sections A-E show five cases in which unilateral electrolytic lesions included the entire LC/SC complex. The area of the lesion is indicated by the stippled area and the location of intact DPH-ir neurons is indicated by the solid black area. The cases illustrated in sections A, B and C survived 14 days following the lesion, while those illustrated in D and E survived 40 days. The bar graphs located to the right of each section show the densities (arbitrary units) of DPH-ir axons that were determined using computer-assisted image analysis. The values for the ventral horn are shown on the left and those for the dorsal horn on the right. The open bars represent values for the contralateral side and the filled bars represent those for the ipsilateral side. Single and double asterisks indicate statistically significant differences for comparisons of average density values on the contralateral side with those on the ipsilateral side ( P < 0.01 and 0.05, respectively). (From Clark and Proudfit, 1991.)
137
support the conclusion that LC neurons send major projections to the ventral horn. There is some evidence, however, that does not support this conclusion. Thus, the results of a study using a combination of PHA-L injections in the LC and D@H-immunocytochemistry indicate that the innervation of the ventral horn by LC neurons is sparse in the rat (Fritschy et ai., 1987; Fritschy and Grzanna, 1990). Because the methods used in our studies are nearly identical to those used by Fritschy and coworkers (Fritschy et ai., 1987; Fritschy and Grzanna, 19901, it is difficult to reconcile these discrepant observations in terms of methodological differences. However, the results discussed in the following section demonstrate the existence of fundamental anatomical differences between Sprague-Dawleyderived rats obtained from different vendors that may explain these conflicting results. In summary, a majority of the anatomical evidence derived from studies using rat, cat and primate indicates that LC neurons densely innervate the medial part of the laminae VII and VIII, the motoneuron pool of laminae IX and lamina X near the central canal. LC neurons also provide a moderate innervation of the ventral part of the dorsal horn (lamina VI). However, there is conflicting evidence which indicates that spinally projecting LC neurons in rat provide a dense innervation of the dorsal horn, a lesser, but significant, innervation of the intermediate zone, but only sparse innervation of the ventral horn. Evidence for genetic differences in the projections of coeruleospinal neurons Although the majority of the evidence indicates that coeruleospinal neurons provide a major innervation of the ipsilateral ventral horn, the careful studies reported by Fritschy and colleagues (Fritschy et d., 1987; Fritschy and Grzanna, 1990) suggest that LC neurons primarily innervate the dorsal horn on both sides. Since this conflicting evidence was obtained using experimental methods similar to ours, the explanation for these divergent results does not appear
to be related to methodological differences. One likely explanation for the reported differences in the projections of LC neurons may be related to a fundamental difference in the origin and terminations of these projections in different substrains of rats. This suggestion is supported by the results of preliminary experiments in our laboratory in which the spinal cord projections of LC neurons were examined in rats obtained from two different vendors (Proudfit et al., 1990). More specifically, we examined the possibility that spinal cord projections of LC neurons in Sprague-Dawley-derived rats obtained from Harlan (Harlan Sprague Dawley, Inc.), the substrain used by Fritschy and coworkers, differ from those in Sprague-Dawleyderived rats obtained from Sasco (Sasco, Inc.) that were used in our experiments. Rats from both vendors were given a unilateral injection of the retrograde tracer Fluoro-Gold into either the lumbar dorsal or ventral horn of the spinal cord. Ten days later, the animals were perfused and brainstem sections including the LC were examined for the presence of TH-ir neurons labeled with retrogradely transported Fluoro-Gold. The results of experiments using Harlan rats indicated that LC neurons retrogradely labeled by FluoroGold injected in the dorsal horn were distributed bilaterally with approximately 50% of the labeled neurons on each side (Table 1). Furthermore, Fluoro-Gold injections into the dorsal horn of Harlan rats labeled more LC neurons than did similar injections into the ventral horn. These results are in good agreement with those of Fritschy and Grzanna (1990) who reported that LC neurons have major bilateral projections to the spinal cord dorsal horn. In contrast, the results obtained using Sprague-Dawley-derived rats from Sasco were quite different. These rats exhibited primarily ipsilateral labeling of LC neurons following injections of Fluoro-Gold into the spinal cord. More specifically, 72% of the TH-ir LC cells that were retrogradely labeled by a unilateral Fluoro-Gold injection in the ventral horn were found on the ipsilateral side (Table 1). Furthermore, ventral horn injections of Fluoro-Gold
138
labeled more LC neurons than did similar injections in the dorsal horn. These results are consistent with our previous results which indicated that LC neurons in Sasco rats have a predominant ipsilateral projection to the ventral horn (Clark and Proudfit, 1989, 1991). These preliminary results suggest that the differences between the results reported by Fritschy and coworkers (Fritschy et al., 1987; Fritschy and Grzanna, 1990) and those described by Clark and Proudfit (1989, 1991) can be explained by fundamental differences in the spinal cord projections of LC neurons in different substrains of rats (Proudfit et al., 1990). Functional significance of coeruleospinal neurons The majority of the evidence presented in this review indicates that spinally projecting noradrenergic neurons in rat, cat and monkey are located in the ventral and caudal aspects of the LC and have axons that are located in the ipsilatera1 ventral quadrant of the spinal cord. In addition, the terminal fields of LC neurons are primarily in the ventral horn, the intermediate zone and lamina X near the central canal. The prominent innervation of the ventral horn suggests that the spinally projecting noradrenergic neurons in the LC modulate motor functions. This conclusion is consistent with an extensive body of information which indicates that descending LC neurons modulate the excitability of motoneurons in the ventral horn (Strahlendorf et al., 1980; Fung and Barnes, 1981, 1984, 1987; Chan et al., 1986; Fung et aL, 1987). The dense noradrenergic innervation of lamina X and the medial part of lamina VII suggests that descending LC neurons may modulate nociception by altering the excitability of spinothalamic tract neurons located in these laminae (Trevino et al., 1972, 1973, 1974; Willis et al., 1974; Carstens and Trevino, 1978). This conclusion is supported by numerous reports which demonstrate that electricaI stimulation of LC neurons produces antinociception (Segal and
Sandberg, 1977; Sandberg and Segal, 1978; Margalit and Segal, 1979; Jones and Gebhart, 1986a; Janss et al., 1987) that can be blocked by intrathecal injection of noradrenergic antagonists (Jones and Gebhart, 1986a). Furthermore, electrical stimulation of LC neurons can also inhibit the responses of nociceptive dorsal horn neurons to noxious peripheral stimuli in rat (Jones and Gebhart, 1986b; Jones and Gebhart, 19871, cat (Hodge et al., 1981, 1983a, b; Mokha et al., 1983, 1985, 1986; Zhao and Duggan, 1988) and primate (Girardot et al., 1987). These neurons were recorded in laminae I to VI in the rat (Jones and Gebhart, 1986b, 1987), laminae 111 to VI in the cat (Hodge et al., 1981, 1983a, b; Mokha et al., 1985, 1986; Zhao and Duggan, 1988) and lamina IV in the primate (Girardot et al., 1987). Since the dendrites of neurons in laminae V and VI extend into laminae VI and VII (Brown, 19811, the presence of coeruleospinal axons in laminae VI and VII is consistent with LC-induced inhibition of nociceptive neurons in these laminae. The coeruleospinal axons activated by LC stimulation appear to be located in the ipsilateral ventral quadrant of the spinal cord. Thus, the inhibition of nociceptive dorsal horn neurons by LC stimuIation in the rat can be significantly reduced by injection of lidocaine into the ipsilatera1 ventrolateral funiculus (Jones and Gebhart, 1987). Furthermore, these authors demonstrated that neither ipsilateral nor bilateral lesions of the cervical dorsolateral funiculus that included the superficial laminae of the dorsal horn altered the inhibition of lumbar nociceptive dorsal horn neurons produced by LC stimulation (Jones and Gebhart, 1987). Similar experiments in cat demonstrated that the inhibition of nociceptive dorsal horn neurons by LC stimulation could also be blocked by ipsilateral spinal cord lesions that included the ventrolateral funiculus (Mokha et al., 1983, 1986). In addition, massive lesions that destroyed the entire dorsal half of the thoracic spinal cord did not alter the LC-induced inhibition of nociceptive dorsal horn neurons recorded from lumbar segments (Mokha et al., 1986). These
139
results provide compelling evidence that the coeruleospinal axons that modulate nociception are located in the ventrolateral funiculus, but not in the dorsolateral funiculus or the superficial laminae of the spinal cord. In summary, there is convincing evidence that coeruleospinal noradrenergic neurons modulate both nociception and motor functions in the spinal cord of both cat and rat. These physiological actions are consistent with the prominent innervation of spinal cord regions where spinothalamic tract cells and motoneurons are located. Conclusions
We have used several different approaches, including anterograde transport of PHA-L in conjunction with D P H immunocytochemistry, retrograde transport of Fluoro-Gold in conjunction with TH-immunocytochemistry, and electrolytic lesions, to demonstrate that spinally projecting noradrenergic neurons in the rat are located in the ventral and caudal aspects of the LC and have axons that are located in the ipsilateral ventral quadrant of the spinal cord. In addition, the terminal fields of LC neurons are located primarily in the ventral horn, the intermediate zone and lamina X near the central canal. These results are in excellent agreement with those obtained using more indirect methods in the rat, cat and monkey. Although the vast majority of evidence indicates that the innervation of the dorsal horn by coeruleospinal neurons is rather sparse, there is conflicting evidence that the dorsal horn of the rat is heavily innervated by descending LC neurons. However, recent evidence from our laboratory indicates that these conflicting results may be due to genetic differences between substrains of rats. The existence of different substrains of rats that have fundamentally different projections to the spinal cord may provide a powerful tool for defining the physiological functions of coeruleospinal noradrenergic neurons.
Acknowledgement
This work was supported by USPHS Grant DA 03980 from the National Institute on Drug Abuse. References Ad&, J.P., Postema, F. and Korf, J. (1979) Contribution of the locus coeruleus to the adrenergic innervation of the rat spinal cord: A biochemical study. J. Neural. Trunsm., 44: 159-173. Barrington, F.J.F. (1925) The effect of lesions of the hind-and mid-brain on micturition in the cat. Quart. J. Exp. Physiol., 15: 81-102. Brown, A.G. (1981) Organization in the Spinal Cord. Springer-Verlag, Berlin, pp. 130-153. Carlsson, A,, Dahlstrom, A,, Fuxe, K. and Hillarp, H.-A. (1964) Cellular localization of monoamines in the spinal cord. Acfu Physiol. Scund., 60: 112-119. Carstens, E. and Trevino, D.L. (1978) Anatomical and physiological properties of ipsilaterally projecting spinothalamic neurons in the second cervical segment of the cat’s spinal cord. J. Comp. Neurol., 182: 167-184. Chan, J.Y., Fung, S.J., Chan, S.H. and Barnes, C.D. (1986) Facilitation of lumbar monosynaptic reflexes by locus coeruleus in the rat. Bruin Res., 369: 103-109. Clark, F.M. and Proudfit, H.K. (1989) The projection of locus coeruleus neurons to the spinal cord in the rat. Soc. Neurosci. Abstr., 15: 547. Clark, F.M. and Proudfit, H.K. (1991) The projection of locus coeruleus neurons to the spinal cord in the rat determined by anterograde tracing combined with immunocytochemistry. Bruin Res., 538: 231-245. Commissiong, J.W. (1981) Evidence that the noradrenergic coeruleospinal projection decussates at the spinal level. Brain Res., 212: 145-151. Commissiong, J.W., Hellstrom, S.O. and Neff, N.H. (1978) A new projection from locus coeruleus to the spinal ventral columns: Histochemical and biochemical evidence. Brain Res., 148: 207-213. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoarnines in the cell bodies of brain stem neurons. Acta Physiol. Scund. (Suppl. 232), 62: 1-55. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus ceruleus: New evidence of anatomical and physiological specificity. Physiol. Rec., 63: 844-914. Fritschy, J.-M. and Grzanna, R. (1990) Demonstration of two separate descending noradrenergic pathways to the rat spinal cord: Evidence for an intragriseal trajectory of locus coeruleus axons in the superficial layers of the dorsal horn. J. Comp. Neurol., 291: 553-582. Fritschy, J.-M., Lyons, W.E., Mullen, C.A., Kosofsky, B.E., Molliver, M.E. and Grzanna, R. (1987) Distribution of locus coeruleus axons in the rat spinal cord: A combined anterograde transport and immunohistochemical study. Brain Res., 437: 176-180.
140 Fung, S.J. and Barnes, C.D. (1981) Evidence of facilitatory coeruleospinal action in lumbar motoneurons of cats. Brain Rex, 216: 299-311. Fung, S.J. and Barnes, C.D. (1984) Locus coeruleus control of spinal cord activity. In C.D. Barnes (Eds.), Bruinstem Control of Spinal Cord Function, Academic Press, Orlando, pp. 216-256. Fung, S.J. and Barnes, C.D. (1987) Membrane excitability changes in hindlimb motoneurons induced by stimulation of the locus coeruleus in cats. Brain Res., 402: 230-242. Fung, S.J., Pompeiano, 0. and Barnes, C.D. (1987) Suppression of the recurrent inhibitory pathway in lumbar cord segments during locus coeruleus stimulation in cats. Bruin Res., 402: 351-354. Fuxe, K. (1965) Evidence for the existence of monoamine neurons in the central nervous system. IV. The distribution of monoamine terminals in the central nervous system. Acta Physiol. Scund., 64, Suppl. 247: 39-85. Gerfen, C.R. and Sawchenko, P.E. (1984) An anterograde neuroanatomic tracing method that shows the detailed morphology of neurons, their axons and terminals: Immunohistochemical localization of an axonally transported plant lectin, Phaseolus vulgaris-leucoagglutinin (PHA-L). Bruin Res., 290: 219-238. Girardot, M.-N., Brennan, T.J., Martindale, M.E. and Foreman, R.D. (1987) Effects of stimulating the subcoeruleusparabrachial region on the non-noxious and noxious responses of Tl-T5 spinothalamic tract neurons in the primate. Bruin Res., 409: 19-30. Guyenet, P.G. (1980) The coeruleospinal noradrenergic neurons: Anatomical and electrophysiological studies in the rat. Brain Res., 189: 121-133. Hammond, D.L. and Proudfit, H.K. (1980) Effects of locus coeruleus lesions on morphine-induced antinociception. Brain Rex, 188: 79-91. Hancock, M.B. and Fougerousse, C.L. (1976) Spinal projections from the nucleus locus coeruleus and nucleus subcoeruleus in the cat and monkey as demonstrated by the retrograde transport of horseradish peroxidase. Brain Res. Bull., 1: 229-234. Hartman, B.K. (1973) Immunofluorescence of dopamine-6hydroxylase. Application of improved methodology to the localization of the peripheral and central noradrenergic nervous system. J. Histochem. Cytochem., 21: 312-332. Hodge, C.J., Jr., Apkarian, A.V., Stevens, R., Vogelsang, G. and Wisnicki, H.J. (1981) Locus coeruleus modulation of dorsal horn unit responses to cutaneous stimulation. Brain Res., 204 415-520. Hodge, C.J., Jr., Apkarian, A.V., Owen, M.P. and Hanson, B.S. (1983a) Changes in the effects of stimulation of locus coeruleus and nucleus raphe magnus following dorsal rhizotomy. Brain Res., 288: 325-329. Hodge, C.J., Jr., Apkarian, A.V., Stevens, R.T., Vogelsang, G.D., Brown, 0. and Franck, J.I. (1983b) Dorsolateral pontine inhibition of dorsal horn cell responses to cutaneous stimulation: Lack of dependence on catecholaminergic systems in cat. J. Neurophysiol.L 50: 1220-1235. Holets, V.R., Hokfelt, T., Rokaeus, A,, Terenius, L. and
Goldstein, M. (1988) Locus coeruleus neurons in the rat containing neuropeptide Y, tyrosine hydroxylase or galanin and their efferent projections to the spinal cord, cerebral cortex and hypothalamus. Neuroscience, 24: 893-906. Holstege, G. and Kuypers, H.G.J.M. (1982) The anatomy of brain stem pathways to t h e spinal cord in cat. A labeled amino acid tracing study. Prog. Brain Res., 57: 145-175. Janss, A.J., Jones, S.L. and Gebhart, G.F. (1987) Effect of spinal norepinephrine depletion on descending inhibition of the tail flick reflex from the locus coeruleus and lateral reticular nucleus in the rat. Bruin Res., 400: 40-52. Jones, S.L. and Gebhart, G.F. (1986a) Characterization of coeruleospinal inhibition of the nociceptive tail-flick reflex in the rat: Mediation by spinal a,-adrenoceptors. Bruin Res., 364: 315-330. Jones, S.L. and Gebhart, G.F. (1986b) Quantitative characterization of ceruleospinal inhibition of nociceptive transmission in the rat. J. Neurophysiol., 56: 1397-1410. Jones, S.L. and Gebhart, G.F. (1987) Spinal pathways mediating tonic, coeruleospinal, and raphespinal descending inhibition in the rat. J. Neurophysiol., 58: 138-159. Jones, B.E. and Yang, T.-Z. (1985) The efferent projections from the reticular formation and the locus coeruleus studied by anterograde and retrograde axonal transport in the rat. J. Comp. Neurol., 242: 56-92. Jones, B.E., Park, M. and Beaudet, A. (1986) Retrograde labeling of neurons in the brain stem following injections of [3H]choline into the rat spinal cord. Neuroscience, 18: 901-916. Karoum, F., Commissiong, J.W., Neff, N.H. and Wyatt, R.J. (1980) Biochemical evidence for uncrossed and crossed locus coeruleus projections to the spinal cord. Bruin Res., 196: 237-241. Kausz, M. (1986) Distribution of neurons in the lateral pontine tegmentum projecting to thoracic, lumbar and sacral spinal segments in the cat. J. Hirnforsch., 27: 485-493. Kuypers, H.G.J.M. and Maisky, V. (1975) Retrograde axonal transport of horseradish peroxidase from spinal cord to brain stem cell groups in the cat. Neurosci. Lett., 1: 9-14. Loewy, A.D., Saper, C.B. and Baker, R.P. (1979) Descending projections from the pontine micturition center. Bruin Rex, 172: 533-538. Loughlin, S.W. and Fallon, J.H. (1985) Locus coeruleus. In G. Paxinos (Ed.), The Rat Nervous System, Hindbrain and Spinal Cord, Vol. 2, Academic Press, Boston, pp. 79-93. Loughlin, S.E., Foote, S.L. and Grzanna, R. (1986) Efferent projections of nucleus locus coeruleus: Morphologic subpopulations have different efferent targets. Neuroscience, 18: 307-319. Lyons, W.E. and Grzanna, R. (1988) Noradrenergic neurons with divergent projections to the motor trigeminal nucleus and the spinal cord: A double retrograde neuronal labeling study. Neuroscience, 26: 681-693. Lyons, W.E., Fritschy, J.-M. and Grzanna, R. (1989) The noradrenergic neurotoxin DSP-4 eliminates the coeruleospinal projection but spares projections of the A5 and A7 groups to the ventral horn of the rat spinal cord. J. Neurosci., 9: 1481-1489.
141 Margalit, D. and Segal, M. (1979) A pharmacological study of analgesia produced by stimulation of the nucleus locus coeruleus. Psychoparmacology, 62: 169-173. Mason, S.T. and Fibiger, H.C. (1979) Regional topography within noradrenergic locus coeruleus as revealed by retrograde transport of horseradish peroxidase. J. Cornp. Neurol., 187: 703-724. Mokha, S.S., McMillan, J.A. and Iggo, A. (1983) Descending influences on spinal nociceptive neurons from locus coeruleus: Actions, pathway, neurotransmitters, and mechanisms. In J.J. Bonica, U. Lindblom and A. Iggo (Eds.), Adcances in Pain Research and Therapy, Vol. 5, Raven Press, New York, pp. 386-392. Mokha, S.S., McMillan, J.A. and Iggo, A. (1985) Descending control of spinal nociceptive transmission. Actions produced on spinal multireceptive neurones from the nuclei locus coeruleus (LC) and raphe magnus (NRM). Exp. Brain Res., 58: 213-226. Mokha, S.S., McMillan, J.A. and Iggo, A. (1986) Pathways mediating descending control of spinal nociceptive transmission from the nuclei locus coeruleus (LC) and raphe magnus (NRM) in the cat. Exp. Brain Rex, 61: 597-606. Nagai, T., Satoh, K., Imamoto, K. and Maeda, T. (1981) Divergent projections of catecholamine neurons of the locus coeruleus as revealed by fluorescent retrograde double labeling technique. Neurosci. Lett., 23: 117-123. Nakazato, T. (1987) Locus coeruleus neurons projecting to the forebrain and the spinal cord in the cat. Neuroscience, 23: 529-538. Nygren, L.-G. and Olson, L. (1977) A new major projection from locus coeruleus: The main source of noradrenergic nerve terminals in the ventral and dorsal columns of the spinal cord. Brain Res., 132: 85-93. Pickel, V.M., Segal, M. and Bloom, F.E. (1974) A radioautographic study of the efferent pathways of the nucleus locus coeruleus. J. Comp. Neurol., 155: 15-42. Proudfit, H.K., Yeomans, D.C. and Clark, F.M. (1990) Differences in coeruleospinal projections in rat suhstrains. SOC. Neurosci. Abstr., 16: 96. Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1989) Spinally projecting noradrenergic neurons of the dorsolatera1 pontine tegmentum: A combined immunocytochemical and retrograde labeling study. Brain Res., 491: 144-149. Room, P., Postema, F. and Korf, J. (1981) Divergent axon collaterals of rat locus coeruleus neurons: Demonstration by a fluorescent double labeling technique. Brain Res., 221: 219-230. Ross, R.A. and Reis, D.J. (1974) Effects of lesions of locus coeruleus on regional distribution of dopamine-P-hydroxylase activity in rat brain. Brain Res., 73: 161-166. Sandberg, D.E. and Segal, M. (1978) Pharmacologic analysis of analgesia and self-stimulation elicited by electrical stimulation of catecholamine nuclei in the rat brain. Brain Res., 152: 529-542. Satoh, K., Tohyama, M., Yamamoto, K., Sakumoto, T. and Shimizu, N. (1977) Noradrenaline innervation of the spinal cord studied by the horseradish peroxidase method combined with monoamine oxidase staining. Exp. Brain Res., 30: 175-186. Schreder, H.D. and Skagerberg, G. (1985) Catecholamine
innervation of the caudal spinal cord in the rat. J. Cornp. Neurol., 242: 358-368. Segal, M. and Sandberg, D. (1977) Analgesia produced by electrical stimulation of catecholamine nuclei in the rat brain. Brain Res., 123: 369-372. Stevens, R.T., Hodge, C.J., Jr., and Apkarian, A.V. (1982) Kolliker-Fuse nucleus: The principal source of pontine catecholaminergic cells projecting to the lumbar spinal cord of cat. Brain Res., 239: 589-594. Stevens, R.T., Apkarian, A.V. and Hodge, C.J., Jr. (1985) Funicular course of catecholamine fibers innervating the lumbar spinal cord of the cat. Brain Res., 336: 243-251. Strahlendorf, J.C., Strahlendorf, H.K., Kingsley, R.E., Gintautas, J. and Barnes, C.D. (1980) Facilitation of the lumbar monosynaptic reflexes by locus coeruleus stimulation. Neuropharmacology, 19: 225-230. Swanson, L.W. and Hartman, B.K. (1975) The central adrenergic system. An immunofluorescence study of the location of cell bodies and their efferent connections in the rat utilizing dopamine-P-hydroxylase as a marker. J. Comp. Neurol., 163: 467-506. Trevino, D.L., Maunz, R.A., Bryan, R.N. and Willis, W.D. (1972) Location of cells of origin of the spinothalamic tract in the lumbar enlargement of the cat. Exp. Neurol., 34: 64-77. Trevino, D.L., Coulter, J.D. and Willis, W.D. (1973) Location of cells or origin of spinothalamic tract in lumbar enlargement of the monkey. J. Neurophysiol., 36: 75-76. Trevino, D.L., Coulter, J.D., Maunz, R.N. and Willis, W.D. (1974) Location and functional properties of spinothalamic cells in the monkey. In J.J. Bonica, (Ed.) International Symposium on Pain. Adr;. Neurol., 4: 167-170. Westlund, K.N. and Coulter, J.D. (1980) Descending projections of locus coeruleus and subcoeruleus/medial parabrachial nuclei in monkey: Axonal transport studies and dopamine-P-hydroxylase immunocytochemistry. Brain Res. Rec., 2: 235-264. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1981) Origins of spinal noradrenergic pathways demonstrated by retrograde transport of antibody to dopamine-P-hydroxylase. Neurosci. Lett., 25: 243-249. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D (1982) Descending noradrenergic projections and their spinal terminations. Prog. Brain Res., 57: 219-238. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of the rat. Brain Res., 263: 15-31. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1984) Origins and terminations of descending noradrenergic projections to the spinal cord of monkey. Brain Res., 292: 1-16. Willis, W.D., Trevino, D.L., Coulter, J.D. and Maunz, R.N. (1974) Responses of primate spinothalamic tract neurons to natural stimulation of hindlimb. J. Neurophysiol., 37: 358-372. Zhao, Z.-Q. and Duggan, A.W. (1988) Idazoxan blocks the action of noradrenaline but not spinal inhibition from electrical stimulation of the locus coeruleus and nucleus Kolliker-Fuse of the cat. Neuroscience, 25: 997-1005.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Progress i n Brain Research, Vol. 85 0 1991 Elsevier Science Publishers B.V.
143 CHAPTER 9
Ultrastructural aspects of the coeruleo-spinal projection J.C. Holstege and C.M.H. Bongers Department of Anatomy, Erasmus Unicersity Medical School, Rotterdam, The Netherlands
Few studies have focussed on the ultrastructure of the coeruleo-spinal projection. In rat the projections from the area of the locus coeruleus (LC) and subcoeruleus (SC) to lumbar motoneuronal cell groups exhibited two different types of terminals: E-type terminals, containing many very small vesicles and S-type terminals, containing many spherical vesicles and an occasional dense-cored vesicle. These findings are in agreement with data indicating the existence of a noradrenergic (NA) and a non-NA projection from the area of the LC and SC to the spinal cord. A study on dopamine-phydroxylase (DpH)-immunoreactive terminals in lumbar motoneuronal cell groups showed that they contained several granular vesicles, which were not found in the E- and S-type terminals. Only a few immunoreactive terminals exhibited a synaptic specialization in a single, thin section. A low incidence of synaptic junctions was also found for the E-type terminals, but not for the S-type. Based on this and other data, it is suggested that the E-type terminal is NA, while the S-type may contain a non-NA transmitter, possibly acetylcholine. A low incidence of synaptic junctions in single, thin sections may indicate the presence of non-synaptic NA
~
terminals, but direct evidence from serial-section analysis is not available. In the superficial dorsal horn, terminals derived from the area of the LC and SC were identified at the ultrastructural level in two studies, one using the anterograde degeneration technique in opossum, the other (presented in this chapter) using WGA-HRP anterograde tracing in rat. It was found in both studies that most of the labeled structures were small axons (mostly unmyelinated), while few terminals were labeled. They contained mostly spherical vesicles and, according to the degeneration study, a variable number of dense-cored vesicles. The labeled terminals appeared to make regular synaptic contacts mostly with small dendrites and occasionally with spines. They were not present in glomeruli o r engaged in presynaptic arrangements. A study on NA terminals showed similar results, although large granular vesicles were not observed and fewer synapses were seen. On the few data available at present it is concluded that in the spinal superficial dorsal horn, most terminals derived from the area of the LC and SC are NA and establish conventional synapses. However, a non-NA component cannot be excluded.
Key words: coeruleo-spinal projection, noradrenergic terminals, non-noradrenergic terminals, ventral and dorsal horns, E M study
Introduction
There are, at least, two ways to obtain data on the ultrastructure of the coeruleo-spinal projections. One is to use anterograde tracing from the locus coeruleus (LC) and/or subcoeruleus (SC) in order to study the morphology of the labeled
terminals in the spinal cord. This approach provides information on the origin and morphology of the identified terminals, but not on their transmitter content. The second way is to identify noradrenergic (NA) terminals in the spinal cord by using immunocytochemistry or other transmitter-specific techniques at the ultrastructural level.
144
This approach will provide information on the morphology and synaptology of the NA terminals, but not on their origin. With either approach it is to be expected that, at least in part, the same types of terminals will be labeled, since several, but not all (see later), of the NA terminals in the spinal cord are derived from the LC and SC. In this chapter the ultrastructural data obtained with anterograde tracing techniques and those obtained with the identification of NA terminals will be described separately. After a discussion of the various data we will draw some conclusions on the ultrastructure of the coeruleo-spinal projection. Anterograde tracing studies at the ultrastructural level
Retrograde tracing studies at the light microscopic (LM) level (see e.g., Kuypers and Maisky, 1975; Hancock and Fourgerousse, 1976; Basbaum and Fields, 1979; Tohyama et al., 1979) have clearly shown the existence of a coeruleo-spinal projection. In order to establish the exact termination area of the coeruleo-spinal fibers within the spinal cord, the anterograde autoradiographic tracing technique was used at the LM level (Martin et al., 1979; Holstege and Kuypers, 1982; Jones and Yang, 1985). It was found that the LC and adjacent areas in the dorsolateral pons, projected to all levels of the spinal cord and, at each level, to all the laminae (of Rexed, 1952). More recently it was shown (Fritschy and Grzanna, 1990) by using small injections of Phaseolos uulgaris leucoagglutinin (PHA-L) that there is a certain degree of topographical organization in the coeruleo-spinal projection: the ventral part of the LC proper mainly projects to the dorsal horn, while the locus SC preferentially projects to the ventral horn (for details see Grzanna and Fritschy, this volume). It is interesting to note that such a widespread projection is aiso found in the spinal projections from the caudal raphe nuclei and that both projection systems consist of monoaminergic (noradrenaline and serotonin, respectively) and
non-monoaminergic fibers (see e.g., Bjorklund and Skagerberg, 1982; Stevens et al., 1985; Bowker and Abbott, 1990; Fritschy and Grzanna, 1990). This is important for interpreting the results obtained with anterograde tracing techniques at the ultrastructural level, since terminals which contain different transmitters may display a different morphology. Coeruleus and SC projections to spinal motoneuronal cell groups In rat lumbar motoneuronal cell groups the terminals derived from the area of the LC and SC were identified at the ultrastructural level using the anterograde transport of 3H-leucine combined with the EM autoradiography technique (Holstege and Kuypers, 1987a). The EM autoradiograms were subjected to two types of analysis: the circle method and the cluster method. These methods were developed because the analysis of EM autoradiograms is not straightforward, due to a relatively low resolution inherent to the autoradiographic technique (for details and references on the analysis of EM autoradiograms see Holstege and Vrensen, 1988). It may be briefly explained as follows. A radioactive source in the tissue (e.g., transported 3H-leucine) may give rise to a silver grain after the tissue is processed for LM or EM autoradiography (i.e., applying an emulsion layer, exposure, developing and fixing). A radioactive source radiates in all directions, leading to the formation of silver grains, some of which are located directly over the radioactive source, while others are located a short distance away from it. Thus, in some cases silver grains are located over another structural profile than the one actually containing the 3H-isotope. It has been calculated that, under certain experimental conditions, about 50% of the silver grains have their centre located within a distance of 0.5 p m away from a 3H-isotope (and vice versa). Consequently there is a 50% probability that a 3H-isotope is located within a circle with a 0.5 p m radius, which is placed over a silver grain in the EM autoradiograms. By comparing the distribu-
145
tion of these circles over the tissue with the distribution of a generated group of (hypothetical) circles, randomly distributed over the same tissue, it can be determined statistically which tissue compartments (terminals, axons, dendrites, etc.) carry a larger number of silver grains than would be expected if the distribution were random. These tissue compartments are considered to have a high probability of containing radioactivity and, depending on the criteria, may then be considered as labeled. A typical result would be that terminals in general or the terminals of a specific type are radioactively labeled, whereas, for example, dendrites are not. However, with the circle method it cannot be determined whether a particular terminal actually contains radioactivity. For this purpose another, more practical, method of analysis can be used (referred to as the cluster method); it is based on the assumption that a structure is radioactively labeled when a group of silver grains (a cluster) is centered on it. In the case of the coeruleo-spinal projections, a cluster was defined as containing six or more silver grains, but a definition of four or more silver grains has also been used (De Zeeuw et al., 1990). The circle and the cluster methods were both applied to the EM autoradiograms of the coeruleo-spinal projections (Fig. 1, c.f. Holstege and Kuypers, 1987a). It was found that after 'H-leucine injections in the area of the LC and SC (Fig. 21, two types of terminals were labeled: S-type terminals (with a large number of spherical vesicles and an occasional dense-cored vesicle) (Fig. 3) and E-type terminals (with a large number of small, and sometimes elongated, vesicles) (Figs. 4 and 5). The E-type terminals were not described before in the motoneuronal cell groups. They were distinguished as a separate type since they were frequently labeled in the coeruleo-spinal experiments even though they probably constitute no more than 1% of the terminals in rat lumbar motoneuronal cell groups (personal observation). Another striking difference between the s- and E-type terminals was the finding that in single, thin sections labeled S-type
anA ._
T
T
30 v)
z
$20
0 Se
10
S
""1
F
E
G
C
"
T
Fig. 1. Histograms showing the distribution of the radioactivity over the various types of terminals in rat lumbar spinal motoneuronal cell groups after 3H-leucine injections in the area of the locus coeruleus (LC) and subcoeruleus (SC). The upper histogram shows the results obtained with the circle analysis and the lower histogram shows the results obtained with the cluster analysis. Values indicated are percentages (i95% confidence limits) and total numbers ( n ) . (From Holstege and Kuypers, 1987a).
terminal profiles displayed a synapse in about 50% of the cases, whereas only 10% of the E-type terminals did so. Terminals of both types contacted mainly proximal dendrites (containing ribosomes) and to a much lesser extent distal dendrites and cell somata. The results obtained with the circle method (Fig. 1) showed that, besides S-and E-type terminal profiles, F-type terminals should also be considered to contain radioactivity. F-type terminals were not labeled in the cluster analysis (Fig. 1) since they never carried more than 6 silver grains. This indicated that they did not contain a large amount of radioactivity, possibly because they were derived from cells which
146
147
took up only a limited amount of the injected ’H-leucine. This, in turn, may be due to the fact that these cells were located on the periphery of the injection site where a lower concentration of 3H-leucine was present, leading to a weaker labeling of the terminals in the spinal cord. In the motoneuronal cell groups, F-type terminals are associated with an inhibitory transmitter, either GABA (Holstege and Calkoen, 1990) or glycine (Ljungdahl and Hokfelt, 1973; van den Pol and Gorcs, 1988). This would imply the existence of an inhibitory projection from the dorso-lateral pons to the spinal cord. However, it should be realized that the IabeIing of F-type terminals may also be an “artifact” of the statistical approach of the circle method. In view of this and the fact that it is unclear from which area in the periphery of the injection site the presumed inhibitory Ftype terminals originate, they will not be further discussed in this chapter. Coeruleus and subcoeruleus projections to the superficial dorsal horn We present here some preliminary results on the projections from the area of the LC and SC to the dorsal horn obtained in rat with the anterograde WGA-HRP tracing technique at the ultrastructural level. Details of the technique can be found elsewhere (de Zeeuw et al., 1988; Holstege, 1991). Briefly, the following steps were taken: In three rats, injections (20 nl of 7% WGA-HRP in saline) were made in the area of the LC and SC under deep pentobarbital anaesthesia (60 mg/kg). After 4-5 days survival, the
animals were re-anesthetized and perfused with 200 ml saline in 0.1 M phosphate buffer (pH 7.3) followed by 1 1 of 5% gluteraldehyde or 3.5% gluteraldehyde and 1% paraformaldehyde in the same buffer. After perfusion, the brainstem and lumbar spinal cord were dissected. Frozen sections were cut from the injection site, incubated with di-amino-benzidine (DAB), mounted on slides, lightly counterstained, coverslipped and viewed in the LM (Fig. 6). Vibratome sections of the lumbar spinal cord were incubated with tetramethyl benzidine (Mesulam, 19781, stabilized with DAB-Cobalt (Lemann et al., 1985), chemically dehydrated (Muller and Jacks, 1975) and embedded in Araldite. Ultrathin sections were cut from the ipsilateral dorsal horn at the lower lumbar segments, counterstained with uranyl acetate and lead citrate and viewed in a Philips EM 300 electron microscope. WGA-HRP-labeled structures were identified in the superficial (laminae I and 11) area of the dorsal horn. Some of the labeled structures could be identified as terminals ( n = 49). They contained mostly spherical vesicles (Figs. 7 and 8) with an occasional dense-cored vesicle. Several of them (approximately 40%) established asymmetric synapses (Figs. 7 and 8) mostly with small or intermediate dendritic profiles and, in a few cases, with spines. The labeled terminals were never seen to be pre-synaptic to other terminals. Other labeled structures could be identified as myelinated axons. However, in the sample analyzed so far, the largest number of labeled profiles could not be clearly identified, sometimes because of poor ul-
Fig. 2. Light microscopic autoradiograph of the rat brainstem, showing the ’H-leucine injection site in the area of the locus coeruleus and subcoeruleus. Fig. 3. Electron micrograph of an autoradiographically labeled (cluster-labeled) terminal profile in the rat L5 lumbar motoneuronal cell groups, after 3H-leucine injection in the area of the locus coeruleus and subcoeruleus. The labeled terminal profile contains many spherical vesicles and was classified as an S-type (S). Arrowheads indicate an asymmetric synaptic junction. Bar = 0.2 pm. Fig. 4. Electron micrograph of the rat L5 lumbar motoneuronal cell groups, after a 3H-leucine injection in the area of the LC and SC. The autoradiographically labeled terminal profile (cluster-labeled) contains many very small and elongated vesicles and was classified as an E-type (E). Part of the intervaricose segment is also labeled. Two other terminal profiles (not cluster-labeled) are also present. One was classified as an S-type, containing many spherical vesicles (S), the other (F) containing many flattened vesicles and classified as an F-type. Bar = 0.2 p m .
148
Fig. 5. Electron micrograph of the rat L5 lumbar motoneuronal cell groups, after a 'H-leucine injection in the area of the LC and SC. The autoradiographically labeled terminal (cluster-labeled) contains many very small and sometimes elongated vesicles (E-type). Three unlabeled terminals are present. Two of these were classified as S-type (S) and one as F-type (F). Note the difference in size and shape of the vesicles contained within the different types of terminals. The arrowheads indicate synaptic junctions. The E-type terminal profile does not exhibit a synaptic specialization. Bar = 0.2 Fm.
149
Fig. 6 . Light micrograph of the rat brainstem, showing tne WGA-HRP injection site in the area of the locus coeruleus and subcoeruleus. Figs. 7 and 8. Electron micrographs of terminal profiles labeled with WGA-HRP reaction product (arrows) in the L5 superficial dorsal horn after WGA-HRP injections in the area of the LC and SC. The labeled terminal profiles contain many spherical vesicles (S) and show asymmetric synaptic junctions (arrowheads) with a small (7) or intermediate-sized (8) dendrite. Bars = 0.2 y m .
trastructure but mostly because the labeled profiles were small and obscured by the reaction products. It seems likely that many of these small structures were unmyelinated axons, but some of these small structures contained a few vesicles, indicating that they may have been terminal profiles. Whether these small terminal profiles were E-type or contained granular vesicles could not be convincingly established. The coeruleo-spinal projection to the dorsal horn has also been studied (Goode et al., 1980) at the EM level after lesioning the dorso-lateral pons, including the LC. Degenerating structures were identified in the superficial dorsal horn (laminae I and 11). Most of the degenerating structures were axons, whereas degenerating terminals were rare and, after a lesion restricted to the LC, they were even absent. The identified degenerating terminals were described as “containing clumps of clear spherical vesicles and a population of larger vesicles with a variety of dense cores.” Some of them made synaptic contacts, which were asymmetric. These results are in general agreement with our WGA-HRP findings, except for the presence of terminals with many granular vesicles, which we did not observe.
Ultrastructural identification of noradrenergic terminals in the spinal cord In the past decades NA terminals have been identified by means of several techniques including high-affinity uptake of ‘H-noradrenalin (Aghajanian and Bloom, 1967; for review see Descarries and Beaudet, 1983), applying false transmitters (5- or 6-hydroxydopamine) (Tranzer and Thoenen, 1967; 1968) and potassium permanganate fixation (Richardson, 1966; Hokfelt, 1967; Koda and Bloom, 1977). However, the specificity of these techniques (especially with respect to the distinction between noradrenaline and dopamine) is still under debate (see Buijs, 1982 for discussion). More recently it became possible to identify NA terminals by means of immunocytochemistry with antibodies (with a
known specificity) directed against noradrenalin (Geffard et al., 1986) or its synthesizing enzyme dopamine-P-hydroxylase (DPH) (Hartman, 1973; Grzanna et al., 1977), which is also present in adrenergic terminals. All techniques mentioned above can be applied at the LM and EM levels. To our knowledge only two studies have focussed on the ultrastructural identification of NA terminals in the spinal cord. In rat lumbar motoneuronal cell groups they were identified by means of pre-embedding immunocytochemistry with antibodies directed against DPH (Kojima et al., 1985). They found that the reaction product in the labeled terminals was “often confined within small granular vesicles and in several terminals large granules also had selective peroxidase labeling for DPH.” Approximately 20% of the labeled terminals displayed a synapse mainly with dendrites. In the dorsal horn, NA terminals were identified at the ultrastructural level (Satoh et al., 1982), after glyoxylic acid-potassium permanganate fixation, which led to the formation of a dense granule within the small synaptic vesicles of NA terminals. Large dense-cored vesicles were not observed. The NA terminals in the dorsal horn were mainly located opposite small dendritic profiles or spines, but typical synaptic structures were seldom seen. In other areas of the brain, e.g, thalamus (Olschowka et al., 19811, area postrema (Sotelo, 1970, hypothalamus (Olschowka et al., 1981; Silverman et al. , 1985), cerebral cortex (Descarries and Lapierre, 1973; Olschowka et al., 19811, the motor trigeminal nucleus (Senba et al., 1981) and the dorsal raphe (Baraban and Aghajanian, 1981), identified NA terminals appeared to contain small spherical vesicles often accompanied by a variable number of large, dense-cored vesicles. The presence of dense-cored (granular) vesicles is often observed in NA terminals, but they are not invariably present (Sotelo, 1971). The percentages of the terminal profiles which displayed a synaptic structure in single, thin sections varied between 5% and over 60% in the different studies. The results obtained in serial-section analysis
151
and the possibility of non-synaptic release will be discussed below. The issue of non-synaptic transmission
The possibility of non-synaptic release of transmitter from terminals was proposed by Descarries and co-workers (Descarries et al., 1977; Beaudet and Descarries, 1978). They calculated, based on findings in a small series of ultrathin sections, that the majority of the NA terminals in the cerebral cortex lacked synaptic specializations. Since other studies on NA terminals in the cerebral cortex (Olschowka et al., 1981; Parnavelas et al., 1985; Papadopoulos et al., 1987) denied the existence of many non-synaptic NA terminals, they recently reinvestigated this issue (Sequela et al., 1990) using antibodies against noradrenalin. Also in this case, it was claimed that the majority of the NA terminals did not establish conventional synaptic contacts. Therefore, the question, whether or not non-synaptic terminals exist in the cerebral cortex, remains unsettled. Nevertheless, the initial findings of Descarries and co-workers have led to the general concept that whenever a particular group of terminals shows a low incidence of synaptic specializations in single, thin sections, this may indicate the existence of nonsynaptic terminals within that group, implying that non-synaptic transmission occurs. This concept may also apply to the motoneuronal cell groups since both the E-type and the G-type terminals (the latter of which is characterized by a large number of dense-cored vesicles and derived from the caudal raphe nuclei) show a low incidence of synaptic specializations (both k lo%, Holstege and Kuypers, 1987a). In cat spinal cord G-type terminals were investigated in serial sections (Ulfake et al., 1987). It was found that some of the serotonergic G-type terminals indeed lacked a synaptic specialization. Furthermore, in the spinal trigeminal nucleus it was observed (Zhu et al., 1986) that occasionally a large granular vesicle was fused with the limiting membrane of a terminal at a site which did not show a synaptic specialization. A similar phenomenon was ob-
served in serotonergic terminals in the supraependym plexus along the rat aqueduct (Buma, 1989). These observations indicate that non-synaptic release of transmitter does occur in the brain, especially in serotonergic terminals. In the peripheral nervous system non-synaptic release occurs in NA terminals (Thureson-Klein, 1983) but the situation for NA terminals in the central nervous system is still unclear. From a functional point of view non-synaptic release of transmitter substances would lead to a more widespread effect of these substances since they may diffuse over a longer distance before being degraded or removed from the extracellular space, whereas synaptic release may lead to much faster removal or degradation of the transmitter substances from the synaptic cleft. On the other hand, with synaptic release, a much higher concentration of transmitter may be achieved locally ( i e . , in the synaptic cleft) as compared with more diffuse non-synaptic release. The concept of non-synaptic transmission may well be compatible with the hypothesis that monoaminergic transmitters in spinal motoneuronal cell groups (Holstege and Kuypers, 1987b1, and elsewhere in the brain (McCall and Aghajanian, 1979; Waterhouse et al., 19881, are primarily involved in gain-setting mechanisms (see also White et al., this volume). Such a gain-setting mechanism would involve many of the neuronal elements in a particular area and it seems likely that their receptors can be reached more effectively through non-synaptic release of transmitter substances as compared to synaptic release. Finally it should be mentioned that even if the terminals of a particular projection system establish synaptic junctions, they may also release part of their transmitter content at non-synaptic sites, since the two release mechanisms are not necessarily mutually exclusive. The origin of the noradrenergic terminals in the spinal cord
The origin of an identified NA terminal is generally not known and can only be deduced from
152
other studies. LM studies (see Westlund et al., 1983) have shown that the NA projection to the spinal cord is derived from NA cells in the LC (A6), the nucleus SC, the medial and lateral parabrachial nuclei, the Kolliker-Fuse nuclei (together forming the A7 group) and the area near the exit of the facial nerve (A5). In a recent study in rat (Fritschy and Grzanna, 19901, using detailed anterograde tracing with PHA-L and retrograde tracing combined with D P H immunocytochemistry, it was found that the ventral part of the LC proper preferentially innervated the dorsal horn and the intermediate zone, whereas the A7 and A5 NA neurons (including those of the nucleus SC) preferentially projected to the motoneuronal cell groups and the intermediate zone of the ventral horn. However, stimulation in the A5 region results in antinociception, which effect can be blocked by intrathecal application of noradrenalin antagonists, indicating the involvement of NA fibers which project directly from the A5 area to the dorsal horn (Proudfit, 1988). Furthermore, there may be species differences in the origin of the descending NA projections to the spinal cord. (For a detailed description of the descending NA pathways, see Grzanna and Fritschy, and Proudfit and Clark, this volume). Discussion In rat lumbar motoneuronal cell groups it was found (Holstege and Kuypers, 1987a) that injections of 'H-leucine, in the area of the LC and SC labeled two types of terminals: E-type and S-type terminals. The finding that two different terminal types were labeled is consistent with data showing that there is an NA and a non-NA projection from the dorso-lateral pons to the spinal cord (Stevens et al., 1985; Fritschy and Grzanna, 1990). This leads us to ask which of the two terminal types that were autoradiographically labeled contained noradrenalin and which transmitter was present in the other terminal type. An immunocytochemical study in lumbar spinal motoneuronal cell groups with antibodies against D P H (Kojima
et al., 1985) showed that the D P H reaction product was confined within small and sometimes large granules and that about 20% of the terminals showed a synapse. The presence of granular vesicles in the NA terminals was not a prominent feature of the E-type or the S-type terminals. However, the low synaptic incidence of the NA terminals in single, thin sections ( 20%) agrees with the percentage (-t 10%) found for the E-type terminals. In this context it is interesting to note the analogy with the projections in rat from the caudal raphe nuclei and the adjoining reticular formation to the lumbar motoneuronal cell groups. These projections exhibited different terminal types (Holstege, 1987; Holstege and Kuypers, 1987a) and contained a monoaminergic transmitter (serotonin) partially co-localized with various peptides (for review see Holstege and Kuypers, 1987b) and non-monoaminergic transmitters like GABA (Holstege, 19911, glycine (Holstege and Bongers, 1990) and possibly acetylcholine (Bowker et al., 1983; Jones et al., 1986). In the motoneuronal cell groups, serotonin is contained within G-type terminals (Ulfake et al., 1987) and some characteristics of this terminal type are similar to those of the E-type terminal: both terminal types constitute less than 1% of the total number of terminals which are present in the motoneuronal cell groups, but they represent over 50% of their respective projections. In addition, they characteristically show a low incidence of synaptic contacts, indicating that some of the terminals may not establish a synaptic contact as was shown for G-type terminals in a serial section analysis (Ulfake et aL, 1987). Thus, the fact that in the raphe-spinal projection the amine (serotonin) is located in the G-type terminal may be interpreted to indicate that the amine (noradrenalin) in the coeruleo-spinal projection is located in a terminal type with similar characteristics as the G-type. This would imply that the E-type terminal is the terminal containing noradrenalin. As a consequence, the S-type terminal would be the non-NA terminal originating from neurons in the area of the LC and SC. In the
153
motoneuronal cell groups, S-type terminals account for almost half of all the terminals (Conradi, 1969; Bernstein and Bernstein, 1976). They have been shown to be derived from the cerebral cortex (Ralston and Ralston, 1989, dorsal root afferents (RaIston and Ralston, 1979; Conradi et al., 1983) and recurrent motor-axon collaterals (Lagerback et al., 19811. Hence, they have always been associated with an excitatory projection. If this were also the case in the coeruleo-spinal projection, the S-type may contain acetylcholine since it has been shown that there are several cholinergic cells in the dorsolateral pons (Mesulam et af., 1984) and that a few of these cells may project to the spinal cord (Jones et af., 1986). On the other hand, it should be mentioned that terminals with spherical vesicles have also been shown to contain noradrenalin in other areas of the brain (see Descarries and Lapierre, 1973; Card et af., 1986). Therefore, anterograde tracing techniques, combined with immunocytochemistry at the ultrastructural level (see de Zeeuw et af., 1988; Holstege, 1991) are needed to establish with certainty whether the E-type in the motoneuronal cell groups actually contains noradrenalin and whether the S-type is cholinergic. The few data which are available on the ultrastructure of the coeruleo-spinal projections to the dorsal horn were obtained from a degeneration study after lesioning the dorso-lateral pons (Goode et af., 1980) and from a study (presented in this chapter) using the anterograde transport of WGA-HRP. Both studies showed that most of the labeling was located in small axons, whereas a relatively small number of the labeled structures were identified as terminals. These findings agree with an LM study (Fritschy and Grzanna, 19901, which indicated that many fibers derived from the LC area descend through the superficial dorsal horn. The few labeled terminals observed in the WGA-HRP study contained mostly spherical vesicles and an occasional dense-cored vesicle. Similar findings were obtained in the degeneration experiments, except that the degenerating
terminals contained several large dense-cored vesicles. NA terminals were identified in the superficial dorsal horn after permanganate fixation (Satoh et af., 1982). As a result of this fixation procedure, a granule is formed within the synaptic vesicles which are normally electron-lucent. Large granular vesicles were not observed. Based on these morphological data, it seems likely that the NA terminals in the dorsal horn are similar to those labeled in the WGA-HRP tracing experiment. However, the presence of another type of terminal, possibly with granular vesicles as found in the degeneration study, could not be excluded. In the tracing studies it was observed that several of the labeled terminals (in the WGA-HRP experiments +40%) exhibited a synapse in a single section analysis, mainly with small and intermediate dendrites and occasionally with spines. NA terminals showed much fewer synapses, but this may have been due to the fixation procedure. Taken together the data seem to indicate that the majority of the terminals in the superficial dorsal horn, which are derived from the area of the LC and SC, contain noradrenaline and establish conventional synaptic contacts, mostly with distal dendrites. They do not seem to be involved in presynaptic arrangements and are not present in glomeruli. However, it should be stressed that the ultrastructural data available so far are few and sometimes contradictory. Detailed EM studies are needed to obtain more conclusive data on the coeruleo-spinal projection to the dorsal horn. It may be concluded that the terminals which are derived from the area of the LC and SC display a different morphology in the ventral horn as compared to those in the dorsal horn. The E-type terminal, which is prominent in the motoneuronal cell groups was not identified in the dorsal horn and the S-type, which is presumed to be non-NA in the motoneuronal cell groups, may possibly contain noradrenaline in the dorsal horn. Although conclusive evidence for these morphological differences is still lacking, it appears that the NA projection to the dorsal horn and the ventral horn displays a different morphology and
154
therefore, may be derived from different neurons. Anterograde tracing with PHA-L (Fritschy and Grzanna, 1990) and the differential effects of the toxin DSP-4 on the NA innervation of the dorsal and ventral horn (Lyons et al., 1989) also point in this direction. Whether it is justified to conclude (Fritschy and Grzanna, 1990) that these different neuronal populations, (e.g., neurons in the ventral part of the LC which project to the dorsal horn as opposed to neurons in the nucleus SC which project to the ventral horn) are activated independently and under different circumstances and are therefore functionally unrelated, remains to be determined. Acknowledgements
The authors would like to thank: Dr. J. Voogd for reading the manuscript; Mrs. C.M. Bijker-Biemond and Mrs. H. Goedknegt for their technical assistance; Mr. E. Dalm for his help with the surgery; Mrs. P. van Alphen for her help with the photography and Mrs. E. Klink for her expert support. This investigation was supported by grant nr 900-550-072 from the Foundation for Medical Research MEDIGON, which is subsidized by the Netherlands Research Organization (NWO). References Aghajanian, G.K. and Bloom, F.E. (1967) Electron-microscopic autoradiography of rat hypothalamus after intraventricular H3-norepinephrine. Science, 153: 308-310. Baraban, J.M. and Aghajanian, G.K. (1981) Noradrenergic innervation of serotonergic neurons in the dorsal raphe: Demonstration by electron microscopic autoradiography. Brain Res., 204: 1-11. Basbaum, A.I. and Fields, H.L. (1979) The origin of descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat: Further studies on the anatomy of pain modulation. J. Comp. Neurol., 187: 513-532. Beaudet, A. and Descarries, L. (1978) The monoamine innervation of rat cerebral cortex: Synaptic and nonsynaptic axon terminals. Neuroscience, 3: 851-860. Bernstein, J.J. and Bernstein, M.E. (1976) Ventral horn synaptology in the rat. J. Neurocytol., 5: 109-123. Bjorklund, A and Skagerberg, G. (1982) Descending monoaminergic projections to the spinal cord. In B. Sjolund and A. Bjorklund (Eds.), Brain Stem Control and Spinal Mechanisms, Vol. 3, Elsevier, Amsterdam, pp. 55-88.
Bowker, R.M. and Abbott, L.C. (1990) Quantitative re-evaluation of descending serotonergic and non-serotonergic projections from the medulla of the rodent: Evidence for extensive co-existence of serotonin and peptides in the same spinally projecting neurons, but not from the nucleus raphe magnus. Brain Res., 512: 15-25. Bowker, R.M., Westlund, K.N., Sullivan, M.C., Wilber, J.F. and Coulter, J.D. (1983) Descending serotonergic, peptidergic and cholinergic pathways from the raphe nuclei. A multiple transmitter complex. Brain Res., 288: 33-48. Buijs, R.M. (1982) The ultrastructural localization of amines, amino acids and peptides in the brain. In R.M. Buijs, P. Pevet and D.F. Swaab (Eds.), Chemical Transmission in the Brain: The Role of Amines, Amino Acids and Peptides, Progress in Brain Research, Vol., 55, Elsevier, Amsterdam, pp. 167-183. Buma, P. (1989) Ultrastructural demonstration of exocytosin in the supraependymal plexus of the rat. Neurosci. Lett., 104: 263-268. Card, J.P., Riley, J.N. and Moore, R.Y. (1986) The motor trigeminal nucleus of the rat: Analysis of neuronal structure and the synaptic organization of noradrenergic afferents. J. Comp. Neurol., 250: 469-484. Conradi, S. (1969) Ultrastructure and distribution of neuronal and glial elements on the motoneuron surface in the limbosacral cord of the adult cat. Acta Physiol. Scand., SUPPL,332: 5-48. Conradi, S., Cullheim, S., Gollvik, L. and Kellerth, J.-0, (1983) Electron microscopic observations on the synaptic contacts of group Ia muscle spindle afferents in the cat lumbosacral spinal cord. Brain Res., 265: 31 -39. De Zeeuw, C.I., Holstege, J.C., Calkoen, F., Ruigrok, T.J.H. and Voogd, J. (1988) A new combination of WGA-HRP anterograde tracing and GABA-immunocytochemistry applied to afferents of the cat inferior olive at the ultrastructural level. Brain Res., 447: 369-375. De Zeeuw, C.I., Holstege, J.C., Ruigrok, T.J.H. and Voogd, J. (1990) Mesodiencephalic and cerebellar terminals terminate upon the same dendritic spines in the glomeruli of the cat and rat inferior olive: An ultrastructural study using a combination of [3H]leucine and wheat germ agglutinin coupled horseradish peroxidase anterograde tracing. Neuroscience, 34: 645-655. Descarries, L. and Beaudet, A. (1983) The use of autoradiography for investigating transmitter-specific neurons. In A. Bjorklund and T. Hokfelt (Eds.), Handbook of Chemical Neuroanatomy, Vol. 1, Methods in Chemical Neuroanatomy, Elsevier, Amsterdam, pp. 286-364. Descarries, L. and Lapierre, Y. (1973) Noradrenergic axon terminals in the cerebellar cortex of rat. 1. Radioautographic visualization after topical application of DL['H]norepinephrine. 3raiti Res., 51: 141-160. Descarries, L., Watkins, K.C. and Lapierre, Y. (1977) Noradrenergic axon terminals in the cerebral cortex of rat. 111. Topometric ultrastructural analysis. Brain Res., 133: 197222. Fritschy, J.-M. and Grzanna, R. (1990) Demonstration of two separate descending noradrenergic pathways to the rat
155 spinal cord: Evidence for an intragriseal trajectory of locus coeruleus axons in the superficial layers of the dorsal horn. J. Comp. Neurol., 291: 553-582. Geffard, M., Patel, S., Dulluc, J. and Rock, A.-M. (1986) Specific detection of noradrenaline in the rat brain by using antibodies. Brain Res., 363: 395-400. Goode, G.E., Humbertson, A.O. and Martin, G.F. (1980) Projections from the brain stem reticular formation to laminae 1 and I1 of the spinal cord. Studies using light and electron microscopic techniques in the North American opossum. Brain Res., 189: 327-342. Grzanna, R., Morrison, J.H., Coyle, J.T. and Molliver, M.E. (1977) The immunohistochemical demonstration of noradrenergic neurons in the rat brain: The use of homologous antiserum to dopamine-P-hydroxylase. Neurosci. Lett., 4: 127-132. Hancock, M.B. and Fourgerousse, C.L. (1976) Spinal projections from the nucleus locus coeruleus and nucleus subcoeruleus in the cat and monkey as demonstrated by the retrograde transport of horseradish peroxidase. Brain Res. Bull., 1: 229-234. Hartman. B.K. (1973) Immunofluorescence of dopamine-phydroxylase. Application of improved methodology to the localization of the peripheral and central noradrenergic nervous system. J. Histochem. Cyrochem., 21: 312-332. Hokfelt, T. (1967) On the ultrastructural localization of noradrenaline in the central nervous system of the rat. Z. Zellforsch., 79: 110-117. Holstege, G. and Kuypers, H.G.J.M. (1982) The anatomy of brain stem pathways to the spinal cord in cat. A labelled amino acid tracing study. In H.G.J.M. Kuypers and G.F. Martin (Eds.), Descending Pathways to the Spinal Cord, Progress in Brain Research, Vol. 57, Elsevier, Amsterdam, pp. 145-175. Holstege, J.C. (1987) Brainstem projections to lumbar motoneurons in rat. 11. An ultrastructural study by means of the anterograde transport of wheat-germ agglutinin coupled to horseradish peroxidase and using the tetramethyl benzidine reaction. Neuroscience, 21: 369-376. Holstege, J.C. (1991) Ultrastructural evidence for GABAergic brainstem projections to spinal motoneurons in rat. J. Neurosci., 11: 1.59-167. Holstege, J.C. and Bongers, C.M.H. (1990) Ultrastructural evidence that brainstem projections to spinal motoneurons contain glycine. Eur. J. Neurosci., Suppl., 3: 96. Holstege, J.C. and Calkoen, F. (1990) The distribution of GABA in lumbar motoneuronal cell groups. A quantitative ultrastructural study in rat. Brain Res., 530: 130-137. Holstege, J.C. and Kuypers, H.G.J.M. (1987a) Brainstem projections to lumbar motoneurons in rat. I. An ultrastructural study using autoradiography and the combination of autoradiography and HRP histochemistry. Neuroscience, 21: 345-367. Holstege, J.C. and Kuypers, H.G.J.M. (1987b) Brainstem projections to spinal motoneurons: An update. Neuroscience, 23: 809-821. Holstege, J.C. and Vrensen, G.F.J.M. (1988) Anterograde
tracing in the brain using autoradiography and HRPhistochemistry. A comparison at the ultrastructural level. J. Microsc., 150: 233-243. Jones, B.E. and Yang, T.-Z. (1985) The efferent projections from the reticular formation and the locus coeruleus studied by anterograde and retrograde axonal transport in rat. J. Comp. Neurol., 242: 56-92. Jones, B.E., Park, M. and Beaudet, A. (1986) Retrograde labeling of neurons in the brain stem following injections of [3H]choline into the rat spinal cord. Neuroscience, 18: 901-916. Koda, L.Y. and Bloom, F.E. (1977) A light and electron microscopic study of noradrenergic terminals in the rat dentate gyrus. Brain Res., 120: 327-335. Kojima, M., Matsuura, T., Tanaka, A., Amagai, T., Imanishi, J. and Sano, Y. (1985) Characteristic distribution of noradrenergic terminals on the anterior horn motoneurons innervating the perineal striated muscles in the rat. Anat. Embryo[., 171: 267-273. Kuypers, H.G.J.M. and Maisky, V.A. (1975) Retrograde axonal transport of horseradish peroxidase from spinal cord to brain stem cell groups in the cat. Neurosci. Lett., 1: 9-14. Lagerback, P.-A., Ronnevi, L.-O., Cullheim, S. and Kellerth, J.-0. (1981) An ultrastructural study of the synaptic contacts of a-motoneurone axon collaterals: I. Contacts in lamina IX and with identified a-motoneurone dendrites in lamina VII. Brain Res., 207: 247-266. Lemann, W., Saper, C.B., Rye, D.B. and Wainer, B.H. (1985) Stabilization of TMB reaction product for electron microscopic retrograde and anterograde fiber tracing. Brain Res. Bull., 14: 277-281. Ljungdahl, A. and Hokfelt, T. (1973) Autoradiographic uptake patterns of 'H-GABA and 'H glycine in central nervous tissues with special reference to the cat spinal cord. Brain Res., 62: 587-595. Lyons, W.E., Fritschy, J.-M. and Grzanna, R. (1989) The noradrenergic neurotoxin DSP-4 eliminates the coeruleospinal projection but spares projections of the AS and A7 groups to the ventral horn of the rat spinal cord. J. Neurosci., 9: 1481-1489. Martin, G.F., Humbertson, A.O., Laxson, L.S., Panneton, W.M. and Tschismadia, I. (1979) Spinal projections from mesencephalic and pontine reticular formation in North American opossum: A study using axonal transport techniques. J. Comp. Neurol., 187: 373-400. McCall, R.B. and Aghajanian, G.K. (1979) Serotonergic facilitation of facial motoneuron excitation. Brain Res., 169: 11-27. Mesulam, M.-M. (1978) Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: A non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents. J. Histochem. Cytochem., 26: 106-117. Mesulam, M.-M., Mufson, E.J., Levey, A.I. and Wainer, B.H. (1984) Atlas of cholinergic neurons in the forebrain and upper brainstem of the macaque based on monoclonal
156 choline acetyltransferase immunohistochemistry and acetylcholinesterase histochemistry. Neuroscience, 12: 669-686. Muller, L.L. and Jacks, T.J. (1975) Rapid chemical dehydration of samples for electron microscopic examinations. J. Histochem. Cytochem., 23: 107-110. Olschowka, J.A., Molliver, M.E., Grzanna, R., Rice, F.L. and Coyle, J.T. (1981) Ultrastructural demonstration of noradrenergic synapses in the rat central nervous system by dopamine-P-hydroxylase immunocytochemistry. J. Histochem. Cytochem., 29: 271-280. Papadopoulos, G.C., Parnavelas, J.G. and Buijs, R. (1987) Monoaminergic fibers form conventional synapses in the cerebral cortex. Neurosci. Left.,76: 275-279. Parnavelas, J.G., Moises, H.C. and Speciale, S.G. (1985) The monoaminergic innervation of the rat visual cortex. Proc. R. SOC. London (BioL), 223: 319-329. Proudfit, H.K. (1988) Pharmacologic evidence for the modulation of nociception by noradrenergic neurons. In H.L. Fields and J.-M. Besson (Eds.), Pain Modulation, Progress in Brain Research, Vol. 77: 357-370. Ralston, D.D. and Ralston, H.J. 111 (1985) The termination of corticospinal tract axons in the macaque monkey. J. Comp. Neurol., 242: 325-337. Ralston, H.J. and Ralston, D.D. (1979) Identification of dorsal root synaptic terminals on monkey ventral horn cells by electron microscopic autoradiography. J. Neurocytol., 8: 151-166. Rexed, B. (1952) The cytoarchitectonic organization of the spinal cord in the cat. J. Comp. Neurol., 96: 415-496. Richardson, K.C. (1966) Electron microscopic identification of autonomic nerve endings. Nature (London), 210: 756. Satoh, K., Kashiba, A,, Kimura, H. and Maeda, T. (1982) Noradrenergic axon terminals in the substantia gelatinosa of the rat spinal cord. Cell Tissue Res., 222: 359-378. Seguela, P., Watkins, K.C., Geffard, M. and Descarries, L. (1990) Noradrenaline axon terminals in adult rat neocortex: An immunocytochemical analysis in serial thin sections. Neuroscience, 35: 249-264. Senba, E., Toyhama, M., Shiosaka, S., Takagi, H., Sakanaka, M., Matsuzaki, T. Takahashi, Y. and Shimizu, N. (1981) Experimental and morphological studies of the noradrenaline innervations in the nucleus tractus spinalis nervi trigemini of the rat with special reference to their fine structures. Brain Res., 206: 39-50. Silverman, A.-J., Oldfield, B., Hou-Yu, A. and Zimmerman, A. (1985) The noradrenergic innervation of vasopressin
neurons in the paraventricular nucleus of the hypothalamus: An ultrastructural study using radioautography and immunocytochemistry. Brain Res., 325: 215-229. Sotelo, C. (1971) The fine structural localization of nore~ i n e p h r i n e - ~in H the substantia nigra and area postrema of the rat. An autoradiographic study. J. Ulfrastruct. Res., 36: 824-841. Stevens, R.T., Apkarian, A.V. and Hodge, C.J., Jr. (1985) Funicular course of catecholamine fibers innervating the lumbar spinal cord of the cat. Brain Res., 336: 243-251. Thureson-Klein, A. (1983) Exocytosis from large and small dense cored vesicles in noradrenergic nerve terminals. Neuroscience, 10: 245-252. Tohyama, M., Sakai, K., Touret, M., Salvert, D. and Jouvet, M. (1979) Spinal projections from the lower brain stem in the cat as demonstrated by the horseradish peroxidase technique. 11. Projections from the dorsolateral pontine tegmentum and raphe nuclei. Brain Res., 176: 215-231. Tranzer, J.P. and Thoenen, H. (1967) Electron microscopic localization of 5-hydroxydopamine (3,4,5,-trihydroxyphenyl-ethylamine), a new “false” sympathetic transmitter. Experientia, 23: 743-745. Tranzer, J.P. and Thoenen, H. (1968) An electron microscopic study of selective acute degeneration of sympathetic nerve terminals after administration of 6-hydroxydopamine. Experientia, 24: 155-156. Ulfake, B., Arvidsson, U., Cullheim, S., Hokfelt, T., Brodin, E., Verhofstad, A. and Visser, T. (1987) An ultrastructural study of 5-hydroxytryptamine-, thyrotropin-releasing hormone- and substance P-immunoreactive axonal boutons in the motor nucleus of spinal cord segments L7-S1 in the adult cat. Neuroscience, 23: 917-929. van den Pol, A.N. and Gorcs T. (1988) Glycine and glycine receptor immunoreactivity in brain and spinal cord. J. Neurosci., 8: 472-492. Waterhouse, B.D., Sessler, F.M., Cheng, J.-T., Woodward, D.J., Azizi, S.A. and Moises, H.C. (1988) New evidence for a gating action of norepinephrine in central neuronal circuits of mammalian brain. Brain Res. Bull., 21: 425-432. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of the rat. Bruin Res., 263: 15-31. Zhu, P.C., Thureson-Klein, A. and Klein, R.L. (1986) Exocytosis from large dense cored vesicles outside the active synaptic zones of terminals within the trigeminal subnucleus caudalis: A possible mechanism for neuropeptide release. Neuroscience, 19: 43-54.
SECTION I1
Properties of the Locus Coeruleus Neurons
This Page Intentionally Left Blank
C.D. Barneh and 0. Pompeiano (Eds.) Progre%s111 Bruin Rneurch, Vol. 88 8 1991 Elsevier Science Publishers B.V.
159 CHAPTER 10
Single-unit and physiological analyses of brain norepinephrine function in behaving animals B.L. Jacobs, E.D. Abercrornbie, C.A. Fornal, E.S. Levine, D.A. Morilak and I.L. Stafford Program in Neuroscience, Department of Psychology, Princeton University, Princeton, NA U.S.A
In behaving cats, the single-unit activity of locus coeruleus noradrenergic neurons is strongly activated by a variety of challenges (stressors). For example, exposing cats to a dog or to loud white noise, dramatically increases the activity of these neurons and simultaneously produces strong activation of the sympathetic nervous system. Similarly, glucoregulatory, thermoregulatory, and cardiovascular challenges also coactivate noradrenergic neurons and the sympathetic nervous system. A related research program utilized a simple hrainstem response (the monosynaptic jaw closure reflex) to
explore the physiological significance of this response of brain noradrenergic neurons. Conditions which activate these neurons were also shown to potentiate the elicited jaw closure-reflex response. Importantly, when the noradrenergic input to the motor side of this reflex pathway was destroyed with a neurotoxin, the conditions which previously potentiated the reflex were now ineffective. These data represent the first demonstration that the release of norepinephrine, at a specific site, and under physiological conditions, facilitates behavioral output in the intact organism.
Key words: cat, locus coeruleus, neuronal activity, stressors, brainstem reflex, behavior
Introduction One of the major goals OL research on brain norepinephrine (NE) and especially the locus coeruleus (LC) is understanding the functional significance of this system. Under what conditions are LC-NE neurons activated (or inhibited), and what is the effect upon relevant CNS target structures? In the present chapter we describe our research on these two issues. What behavioral/environmental or physiological manipulations alter the activity of LC-NE neurons in be-
having cats? How does a change in the activity of these neurons affect the output of a simple CNS system that receives a dense N E input? (For this latter question, we have studied the jaw closure or masseteric reflex in behaving cats.) Our research has focused on the variables that tonically activate this system. In a more general context, most, or all, of these manipulations could be subsumed under the rubric of stress. Other chapters in this volume address the equally important issue of phasic activation of this system. Such a phasic activation is of more direct rele-
160
vance to issues such as the role of the LC-NE system in attention and orientation. Locus coeruleus neuronal activity in behaving cats Bac kground To record action potentials extracellularly in freely moving cats, we employ chronically implanted flexible microwires, which can be advanced through the brain by means of an attached microdrive. This technique allows us to maintain our recordings of the same neuron in spite of gross movements on the part of the animal. It also permits us to study the same neuron for periods of up to several days. Because the LC in the cat is neurochemically heterogeneous (ie., comprised of both NE and non-NE neurons), we employ a set of criteria for identifying LC-NE neurons on-line (Rasmussen et al., 1986a): (1) long-duration action potential (> 2 msec); (2) slow ( 1-2 spikes/sec in awake animals) and often regular discharge pattern; (3) a complete suppression of neuronal activity during REM sleep; (4) a burst of spikes followed by a pause in response to the phasic application of a painful stimulus; and (5) a complete suppression of activity in response to the systemic administration of a,-agonist drugs such as clonidine (such compounds exert their inhibitory action at a somatodendritic autoreceptor). At the end of an experiment, the localization of the electrode tracks to the area of the LC is confirmed through histological analysis. Portions of our initial studies were descriptive and aimed simply at characterizing the activity of LC-NE neurons (Rasmussen et al., 1986a,b); their activity changes dramatically as a function of the behavioral state of the cat. From a peak discharge rate of approximately 2 spikes/sec during active waking or arousal, the activity gradually declines across quiet waking, drowsy, slow-wave sleep, and, finally, falls completely silent during REM Sleep. Neuronal activity is relatively unchanged during a variety of typical feline behav-
-
iors (e.g., eating and grooming), during simple or complex movements, in response to tactile stimuli (e.g., rubbing the fur), and during feline social interactions. Presentation of simple phasic sensory stimuli, such as clicks or flashes, elicits a short-latency (50 msec) neuronal response (1-4 spikes) followed by an inhibitory period. Behavioral and environmental challenges Prior to initiating this line of investigation, we adopted a set of measures that we believed would index the stress response. Most researchers agree on the activation of the sympathetic nervous system for this purpose. Therefore, we measured changes in plasma levels of catecholamines, and also recorded tonic changes in heart rate, because this latter index can be monitored on-line on a continuous basis. In some studies we also monitored arterial blood pressure. Initially, we wanted to distinguish stimuli or conditions that were simply arousing or activating from those that were stressful. We noted that LC-NE neuronal activity in cats was not increased above an active waking baseline when two inaccessible rats or inaccessible food was introduced into the recording chamber. This was in the face of obvious signs of activation or arousal: low-voltage fast EEG, almost continuous eye movements, and orientation toward the stimuli (Rasmussen et al., 1986a). Consistent with our general hypothesis, when we examined the cat’s physiological response to the inaccessible rats, we found that this produced no significant increase in plasma NE or tonic heart rate (Abercrombie and Jacobs, 1987). Contrary to this, when cats were exposed to loud white noise (100 dB) for 15 min, restrained for 15 min, or confronted with a dog for 5 min, significant increases occurred in all three variables: tonic heart rate increased by 3050%; plasma NE levels rose by 100-200%; and LC-NE unit activity increased by 100-200% (Abercrombie and Jacobs, 1987; Wilkinson and Jacobs, 1988; Levine et al., 1990). When the stressors were removed, all three measures returned to baseline levels within minutes.
161
These studies indicate that LC-NE neurons are frequently co-activated in association with the sympathetic nervous system. Indeed, there was a very high, positive correlation between the degree of activation of LC-NE neurons and tonic heart rate in these studies (Abercrombie and Jacobs, 1987). These data provide support for the concept that this neurochemical neuronal group may play an important role in challenges or emergency situations which confront the organism. Reciprocally, these data indicate that situations that are merely arousing or attention-capturing (such as the introduction of two albino rats) do not provide tonic activation of these neurons. Physiological challenges Simultaneous with our examination of the effect of environmental challenges upon LC-NE neuronal activity, we initiated a series of studies utilizing physiological challenges. In an attempt to maximize the generality of our findings we manipulated several major physiological systems: thermoregulatory, cardiovascular, and glucoregulatory. We were particularly interested in exploring the following issues regarding LC-NE neurons. (1) Can physiological variables influence their activity? (2) If so, can this occur independent of changes in behavioral state? (3) Do these neurons appear to play a specific role in any physiological regulatory systems, or is the relationship a more general one? (4) What is the relationship of LC-NE activity to activation of the sympathoadrenal system? Two systems are present in mammals for defending against thermoregulatory challenges, heat loss and heat gain. These can be evoked, respectively, by placing animals in a heated environment and by systemic administration of an exogenous pyrogen. Comparing the results of these two manipulations can be particularly revealing since they both raise body temperature, but do so by opposite means. In response to ambient heating of the recording chamber to 43"C, LC-NE activity increased only at the point when panting was elicited (Morilak et al., 1987a). Thus, neuronal
activity was not related to environmental temperature per se, nor was it related to body temperature since most animals successfully defended against this challenge and displayed little or no increase in temperature. In response to systemic administration of a pyrogen (muramyl dipeptide, 50 pg/kg, iv), LC-NE activity increased only when the peak increase in body temperature of 2°C was reached. The fact that this increase in neuronal activity was seen in animals displaying signs of sedation and lethargy, typical of a febrile response, indicates that the LC-NE system can be activated independent of a concomitant change in behavioral state. In addition, since both manipulations increased LC-NE unit activity, despite eliciting opposite thermoregulatory responses, it appears that these neurons do not play a specific role in thermoregulation, but may participate more generally in the response to physiological challenges. In our initial attempts to manipulate the cardiovascular system, we employed systemic administration of vasoactive drugs: phenylephrine for producing hypertension and nitroprusside for hypotension (both in doses of 5-15 pg/kg, iv). It proved impossible for us to obtain valid measures of LC-NE activity during the dynamic phases of these drug actions because both produced an activating effect in the cats and thus phasically increased unit activity (Morilak et al., 1987b). However, following the first 10-20 sec, these drugs produced sustained hypertension or hypotension, and during these tonic phases, LC-NE unit activity was unchanged from baseline levels. In other studies we used the drug hydralazine (1 mg/kg, iv) as a long-lasting hypotensive stimulus and as a preferential activator of the neural component of the sympathoadrenal system. We also employed hemorrhage to decrease blood volume and to activate both the neural and hormonal components of the sympathoadrenal system. Hydralazine produced a significant increase in heart rate, plasma NE, and LC-NE unit activity (Morilak et al., 1987b). However, the increase in unit activity appeared to be of lesser magni-
162
tude and of shorter duration than the concomitant sympathetic activation. As with pyrogen administration, the increase in LC-NE unit activity produced by hydralazine occurred in the absence of any behavioral or polygraphic signs of activation or arousal. Somewhat surprisingly, hemorrhage as great as 30% of total blood volume produced no significant increase in LC-NE unit activity, even though levels of both plasma epinephrine and NE were significantly increased (Morilak et al., 1987b). These latter data provide direct evidence that sympathetic activation can occur in the absence of an activation of LC-NE neuronal activity. The final experiment in this series examined the effect of insulin-induced hypoglycemia on LC-NE unit activity in behaving cats (Morilak et al., 1987~).Insulin (2 I.U./kg iv) produced a 65% decrease in plasma glucose level and a 5-fold increase in plasma epinephrine. LC-NE unit activity was significantly increased, but only at the point where plasma glucose was at its lowest level and where, concurrently, plasma epinephrine was at its highest level. All of these changes were reversed by subsequent administration of glucose (150 mg/kg iv). As with several manipulations described above, this increase in LC-NE unit activity during hypoglycemia occurred in the absence of any behavioral or polygraphic signs of behavioral activation or arousal. These results once again indicate that the LC-NE neuronal system is activated only when extreme changes occur in physiological regulatory systems.
the possibility of examining the functional effect of such activation.
Masseteric reflex The masseteric reflex is a monosynaptic brainstem neural circuit: stretch receptors in the masseter muscle spindles, with cell bodies in the mesencephalic trigeminal nucleus (Mes V), synapse directly upon neurons in the trigeminal motor nucleus (Mo V), which innervate the masseter muscle. This reflex can be elicited experimentally in behaving animals by means of pulsatile electrical stimulation of Mes V, which directly excites Mo V neurons, causing contraction of the masseter muscle, resulting in jaw closure or jaw clenching. The amplitude of this response can be recorded easily by means of indwelling leads placed in the belly of the ipsilateral masseter muscle. This is an ideal system for studying the central actions of NE because Mo V receives dense NE inputs. (This derives from both the LC and neighboring lateral tegmental NE neuronsthe activity of these latter neurons are similar to those of LC-NE neurons.) The simplicity of the reflex circuitry, the ease with which the reflex response can be elicited and recorded in unrestrained and unanesthetized animals, the wellestablished NE innervation of the reflex circuit, the ability to locally and specifically remove that innervation with neurotoxin injections, and the opportunity to administer drugs locally to influence the reflex, all make this an ideal system for investigating the functional effects of changes in NE neuronal activity.
Modulatory actions of NE on a simple reflex
Background The studies in the preceding section provide invaluable information regarding the effect of a variety of behavioral, environmental, and physiological challenges upon the single-unit activity of LC-NE neurons. Of equal importance from our perspective is the fact that these studies provide the means for activating this system under naturalistic or physiological conditions. This affords
Pharmacological studies Prior to embarking directly upon studies of the physiological significance of the NE input to the masseteric reflex circuit, we wanted to examine whether simple pharmacological manipulations would exert the expected actions (Stafford and Jacobs, 1990a). If this were supported, it would encourage taking the next step, i.e., analyses under naturalistic or physiologically relevant conditions. In order to insure that any drug effects
163
were specific to the masseteric reflex circuit, these experiments employed local injections directly into MoV, via indwelling cannulae. Microinfusions of NE (0.125-5.0 p g in 0.5 pl) produced dose-dependent changes in the amplitude of the elicited reflex responses. The peak increase ( 200-400% above baseline) occurred in response to infusions of 1.0 p g NE. Increases were evident within 1 min post-infusion, reached maximal levels at 5 min post-infusion, and returned to baseline levels by 30 min. Pharmacological specificity of this effect is indicated by the fact that the increase seen in response to 1.0 p g NE was blocked by systemic pretreatment with the a ,-adrenergic antagonist prazosin (5 mg/kg, ip), but not by pretreatment with the 5-HT antagonist methysergide (0.5 mg/kg, ip). Further support that the a,-adrenergic receptor mediates this response was derived from experiments showing that infusions of the a,-adrenergic agonist phenylephrine (0.5 or 5.0 pg) also produced dose-dependent increases in the amplitude of the masseteric reflex, while the P-adrenergic agonist isoproterenol and the q a d r e n e r g i c agonist clonidine failed to do so (local application of clonidine actually suppressed the reflex response). These pharmacological data indicate that NE, acting at an a,-adrenergic receptor, modulates the amplitude of the masseteric reflex.
-
Physiological studies We hypothesized that if NE acted in the CNS to produce excitation of motoneurons (as pharmacological studies in the preceding section indicated), then the environmental conditions that activate NE neurons should augment the masseteric reflex response (Stafford and Jacobs, 1990b). Therefore, following baseline measures of the reflex amplitude, the reflex was again elicited while cats were exposed to loud white noise (100 db), a dog, or a series of 64 clicks each of which was presented at various intervals prior to Mes V stimulation. The first two conditions dramatically facilitated the reflex response, during exposure to
the stimuli, with the response returning to baseline levels shortly after stimulus offset. The clicks produced reflex facilitation only when presented 100 or 150 msec prior to Mes V stimulation. This latter temporal relationship is consistent with what is known about the time course that brain NE neurons are activated by clicks (see p. 160). Although these results were consistent with our hypothesis, they were, nonetheless, simply correlative. We reasoned that if activation of NE neurons by environmental stimuli was responsible for the augmentation of the reflex under those same conditions, then blockade of NE neurotransmission should block the augmentation (Stafford and Jacobs, 1990b). Two approaches were used in order to study this issue. In the first, cats were given either the a,-adrenergic antagonist prazosin (5 mg/kg, ip) or the 5-HT antagonist methysergide (0.5 mg/kg, ip) and then exposed to the white noise or the series of clicks. In all cases prazosin blocked the reflex augmentation whereas methysergide was without effect. In the second study, the catecholamine neurotoxin, 6-hydroxydopamine (4 p g in 1 pl), was microinfused directly into Mo V in order to produce unilateral NE denervation of the motor component of the reflex pathway. One week later, the reflex was examined bilaterally while the cats were exposed to white noise, a dog, or a series of clicks. The normal augmentation of the reflex produced by these conditions was either completely blocked or markedly diminished on the denervated side, with the response on the intact side remaining essentially the same as that measured prior to denervation. The specificity of the neurochemical destruction produced by 6-hydroxydopamine was indicated by a greater than 70% loss of NE in Mo V, with no significant change in 5-HT levels. Additionally, the localized precision of the destruction was shown by the fact that the NE content in areas just outside of Mo V were at control levels. Finally, pharmacological evidence of the denervation was obtained by microinfusing the same doses of NE that had been examined prior to neurotoxin administration and
164
now observing a markedly enhanced response, or denervation supersensitivity. Discussion We believe that the whole of this research program is greater than the sum of its parts. Thus, not only have we discovered a variety of conditions that activate LC-NE neurons and a variety of conditions that augment a simple brain stem reflex, but we have shown that they are the same stimuli and they exert both of these effects with a similar temporal pattern. Furthermore, through local neurochemical denervation experiments we have shown that these two phenomena are not merely correlated, but that a cause and effect relationship exists between the activation of brain NE neurons and the augmentation of this simple reflex. To our knowledge, these data represent the first demonstration that the release of NE, at a specific site, and under physiological conditions, facilitates behavioral output in the intact organism. These data, in conjunction with the previous experimental literature, allow us to postulate a functional role for brain NE. The NE systems of the CNS may act to enhance an organism’s reactivity to internal or external stimuli, that is, the processing of sensory information and of the motor responses to it. This enhancement may occur during times of stress (tonic stimuli), when interaction with environmental stimuli is most likely to occur, or, in response to phasic alerting stimuli. The present data also shed light on a number of important specific issues. First, the facilitatory effect upon motor behavior seen in response to a number of activators or stressors may be attributable almost completely to NE. This is shown by our denervation experiment where the loss of the NE input specifically to Mo V almost completely eliminated the facilitation following exposure to loud white noise or a dog. This is somewhat surprising since one might have expected such effects to be mediated by the co-activation
or confluence of a number of CNS neurochemical systems. It is also possible that neurochemical systems other than NE may be importantly involved in such effects, but that NE is necessary for their expression. Second, these data give some indication of the magnitude of NE’s facilitatory effect upon motor responses. It is possible that NE exerts a stable and reliable facilitatory influence, but one that is of small magnitude and therefore insignificant. The present results do not, however, support such an assertion since the reflex amplitude is increased approximately 2 to 6-fold by stimuli such as loud white noise. Third, these data bear on the issue of the duration of action of neurotransmitter released under physiological conditions. One might expect a neurotransmitter such as NE, which is hypothesized to exert a somewhat hormone-like neuromodulatory effect, to have a prolonged action that greatly outlives the initiating stimulus. Our data do not support this. Within the first minute following the offset of white noise or the removal of the dog, the amplitude of the reflex response returns to, or near, baseline levels. This is also true of the activation of LC-NE neuronal activity by these same stimuli. Thus, it appears that NE released under physiological conditions does not exert a particularly long-lasting effect. Finally, we discuss three new directions in which this work is taking us. The first is a direct outgrowth of our reflex studies. This research addresses the question of whether NE has a global facilitatory effect. It turns out that the answer is dependent upon the precise nature of the question. For example, if NE is microinfused into Mo V and the motoneurons for jaw opening (digastric) are activated via stimulation of the inferior alveolar (dental) nerve, this reflex is augmented (unpublished observations). However, if the reflex is elicited during the aforementioned conditions known to activate brain NE neurons, e.g., loud white noise, the reflex is not facilitated, and may even be suppressed. Thus, although NE may exert an excitatory modulatory influence
I65
upon all motoneurons, when a particular response is examined in an intact organism it is embedded in a complexity of interactions, some of which may override NE’s facilitatory influence. It would seem appropriate that under the physiological conditions which augment a jaw closure response there should not also be an augmentation of an opposing, jaw opening, response. The second new direction is to examine the neurochemical identity of the afferent inputs that control LC-NE neuronal activity. These studies employ microiontophoresis of neurotransmitter antagonist drugs in conjunction with extracellular single-unit recording in awake head-restrained cats. When cats are exposed to any of the variety of conditions known to activate or inhibit LC-NE neuronal activity, can these effects be blocked by the iontophoretic administration of antagonists to GABA, excitatory amino acids, etc.? Finally, we are also exploring the actions of NE on neurons with definable and reproducible properties, those in areas of mammalian visual cortex. How are the various response properties of these neurons, such as receptive field size, direction selectivity, etc., influenced by NE or NE antagonists? We believe that such studies will provide a critical clue to how the action of CNS NE is integrated into the overall functioning of the organism. Acknowledgements The authors’ research described in this chapter was supported by the U.S. Air Force (AFOSR 87-0301-A) and the N.I.M.H. (MH-23433). The
authors also wish to acknowledge the excellent technical assistance of Ms. Christine Metzler.
References Abercrornbie, E.D. and Jacobs, B.L. (1987) Single unit response of noradrenergic neurons in the locus coeruleus of freely moving cats. I. Acutely presented stressful and nonstressful stimuli. J. Neurosci., 7: 2837-2843. Levine, E.S., Litto, W.J. and Jacobs, B.L. (1990) Activity of cat locus coeruleus noradrenergic neurons during the defense reaction. Brain Res., 531; 189-195. Morilak, D.A., Fornal, C. and Jacobs, B.L. (1987a) Effects of physiological manipulations on locus coeruleus neuronal activity in freely moving cats: I. Thermoregulatory Challenge. Brain Res., 422: 17-23. Morilak, D.A., Fornal, C. and Jacobs, B.L. (1987b) Effects of physiological manipulations on locus coeruleus neuronal activity in freely moving cats: 11. Cardiovascular Challenge. Brain Res., 422: 24-31. Morilak, D.A., Fornal, C. and Jacobs, B.L. (1987~)Effects of physiological manipulations on locus coeruleus neuronal activity in freely moving cats: 111. Glucoregulatory Challenge. Brain Rex, 422: 32-39. Rasrnussen, K., Morilak, D.A. and Jacobs, B.L. (1986a) Single unit activity of locus coeruleus neurons in the freely moving cat: I. During naturalistic behaviors and in response to simple and complex stimuli. Brain Res., 371: 324-334. Rasmussen, K., Strecker, R.E. and Jacobs, B.L. (1986b) Single unit response of noradrenergic, serotonergic and dopaminergic neurons in freely moving cats to simple sensory stimuli. Brain Rex, 369: 336-340. Stafford, I.L. and Jacobs, B.L. (1990a) Noradrenergic modulation of the masseteric reflex in behaving cats. I. Pharmacological Studies. J. Neurosci., 10: 91-98. Stafford, I.L. and Jacobs, B.L. (1990b) Noradrenergic modulation of the masseteric reflex in behaving cats. 11. Physiological Studies. J. Neurosci., 10: 99-107. Wilkinson, L.O. and Jacobs, B.L. (1988) Lack of response of serotonergic neurons in the dorsal raphe nucleus of freely moving cats to stressful stimuli. Exp. Neurol., 101: 445-457.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Proroyrc.x$ w Brum Research, Val. 88 0 1991 Elbevier Science Publishers B.V.
167 CHAPTER 11
Synaptic potentials in locus coeruleus neurons in brain slices J.T. Williams, D.H. Bobker and G.C. Harris The Vollum Institute, Oregon Health Sciences University, S.W. Sam Jackson Park Road, Portland, OR, U.S.A.
Neurons of the locus coeruleus (LC) fire action potentials spontaneously in uitro in the absence of any stimulation. This spontaneous activity is thought to arise from intrinsic membrane properties that include a balance between at least two ion conductances. One is a persistent inward sodium current that is active near the threshold for action potential generation. The second is a calcium-dependent potassium current that is activated following the entry of calcium during the action potential, is responsible for the after-hyperpolarization following the action potential, and decays over a period of 1-2 sec following the action potential. The spontaneous activ-
ity of LC neurons can be altered by both excitatory and inhibitory synaptic inputs. O n e excitatory input has been described that is mediated by glutamate receptors of both the non-NMDA and NMDA subtypes. Inhibitory synaptic potentials include those mediated by GABA (acting on GABA,-receptors), glycine (acting on a strychnine-sensitive receptor) and noradrenaline (acting on a,-adrenoceptors). The presence of synaptic potentials mediated by these transmitters, studied in uitro, correlate with studies made in uivo and with histochemical identification of synaptic inputs to the locus coeruleus.
Key words: y-aminobutyric acid, noradrenaline, glutamate, glycine, brain slice, synaptic potentials
Introduction
Locus coeruleus (LC) neurons fire action potentials spontaneously both in uiuo and in uitro. The spontaneous activity of neurons in the brain slice preparation are regular in nature having little variation in the interspike interval for periods of hours. This observation suggests that the intrinsic activity of LC neurons alone could maintain basal noradrenergic tone in limbic areas, cerebral and cerebellar cortices, all areas where the primary source of noradrenergic innervation stems from the LC (Moore and Bloom, 1979). Under basal
conditions LC neurons fire action potentials asynchronously but have been shown to synchronize in response to strong synaptic input (Aston-Jones and Bloom, 1981b; Ennis and Aston-Jones, 1988). LC neurons are activated by a wide variety of both noxious and non-noxious sensory stimuli (Foote et al., 1980) and actively inhibited during rapid eye movement sleep (Aston-Jones and Bloom, 1981a). In addition, changes in the firing pattern of LC neurons correlated with changes in sensory evoked responding (Valentino and Aulisi, 1987; Valentino and Foote, 1987). Synaptic inputs that alter the spontaneous activity of LC neurons
168 A
1
Fast
400 ms 8
Fig. 1. Fast and slow synaptic potentials. A. Recording of membrane potential showing the relative time courses of fast (depolarizing) and slow (hyperpolarizing) synaptic potentials. B. Expanded time scale to show the time-course of the fast synaptic potential. In these and other figures traces are the average of 4 synaptic potentials evoked with electrical stimulation (at the arrow) at 0.1 Hz. The membrane potential was held at -70 mV.
would be expected to transiently increase or decrease noradrenergic tone in the widespread projection areas. A review of the synaptic potentials found in the rat brain slice preparation that may mediate changes in activity of LC neurons in ~1iuo is the focus of the present study. In the brain slice containing the nucleus LC there are no visually identifiable afferent pathways. Stimulating electrodes are simply placed in the area or even within the body of the LC. Electrical stimulation evokes the release of a variety of transmitters which are recorded as postsynaptic potentials. Qualitatively similar synaptic potentials are found in slices cut in both the coronal and horizontal planes. It seems however, that activation of synaptic potentials is more effective in horizontal slices with the stimulating electrodes placed caudal to the body of the LC (Grudt, Harris and Williams, unpublished results). The synaptic potentials we have found are of two forms, fast and slow (Fig. 1). Fast synaptic potentials had latencies to onset of 0.5-2 msec, peaked in 20-40 msec and decayed to the control
potential in 100-200 msec. The slow inhibitory synaptic potential 6.p.s.p.) had a latency to onset of 40-50 msec, peaked in 200-400 msec and was 2-3 sec in duration. There were three “fast” synaptic potentials, one mediated by an excitatory amino acid, a second by y-aminobutyric acid (GABA) and a third by glycine. The i.p.s.p. was mediated by noradrenaline (NA) (or adrenaline). Although the fast synaptic potentials could be evoked with lower stimulus intensity than the slow synaptic potential, afferent nerve fibers could not be stimulated selectively in the slice preparation. Each of the different types of synaptic potential was studied after blocking the others by using the appropriate receptor antagonists. e.p.s.p.s: excitatory amino acid-mediated An excitatory amino acid-mediated synaptic potential (e.p.s.p.) was determined based on the sensitivity of the synaptic potential to amino acid antagonists and by comparison of the reversal potential of the synaptic potential with exogenously applied amino acid agonists (Cherubini et af., 1988). This e.p.s.p. is the only excitatory synaptic potential found to date. An excitatory amino acid afferent pathway from the nucleus paragigantocellularis has been described in a study in uiuo (Ennis and Aston-Jones, 1988). In the slice preparation, all experiments on the e.p.s.p. were carried out in the presence of bicuculline (30 pM) or in bicuculline plus picrotoxin (100 p M ) to block the GABA-mediated synaptic potential (see below). The amplitude of the e.p.s.p. was graded with the strength of stimulation. At a given stimulus intensity the amplitude remained stable with repetitive stimulation (0.1 Hz) for periods of 1-3 h. Hyperpolarization of the membrane from near rest ( - 60 mV) to about - 90 mV resulted in an increase in the amplitude of the e.p.s.p. Further hyperpolarization resulted in little or no further increase in amplitude (Egan et af., 1983; Cherubini et al., 1988). The total duration of the e.p.s.p. was 150-200 msec at - 60 mV and decreased 2 to
169
: -80mV
’ ,
f
Fig. 2. GABA and excitatory amino acid components of the fast synaptic potential. A. In this and other figures synaptic potentials are superimposed to illustrate the effect of the indicated antagonist. Bicuculline (30 pM) decreased the amplitude of the fast synaptic potential by about 50%. B. CNQX (10 pM) caused a further reduction of the fast synaptic potential. The fast synaptic potential was decreased to about 10% of its original amplitude by the combination of GABA, and excitatory amino acid antagonists.
solutions. In Mg++-free solutions the NMDA receptor-mediated component of the e.p.s.p. was enhanced in amplitude and duration. CNQX blocked only the early component of the e.p.s.p., whereas APV blocked the remaining slow component. Superfusion of glutamate, NMDA and quisqualate all induced an inward current at - 60 mV (Cherubini et al., 1988). The NMDA-induced current amplitude was smaller at -100 mV, whereas the glutamate and quisqualate currents were larger. The reversal potential for all three agonists was about 0 mV. Kynurenate (500 p M ) blocked the inward current induced by all three agonists and APV blocked only the NMDA-induced inward current. It appears from these results that the response of LC neurons to excitatory amino acid agonists and antagonists is similar to those on neurons in other brain areas where the action has been studied extensively (Mayer and Westbrook, 1987). i.p.s.p.s: GABA-mediated
5-fold with membrane hyperpolarization to - 120 mV. The decrease in duration of the e.p.s.p. was similar to the decrease in the membrane time constant resulting from an increase in potassium conductance ( I K , l R )(Williams et af., 1984, 1989). With depolarization from rest the amplitude of the e.p.s.p. declined and reversed polarity at about 0 mV. Such experiments were carried out with CsCl filled electrodes so that the membrane could be depolarized to potentials beyond + 10 mV. Superfusion of kynurenate (500 p M to 2 mM), a non-selective excitatory amino acid antagonist, decreased the e.p.s.p. to about 25% of the control amplitude (Cherubini et al., 1988). The more selective and potent kainate/quisqualate antagonist, 6-cyano-2,3-dihydroxyy-7-nitro-quinoxaline (CNQX) was more effective, often causing a total inhibition of the e.p.s.p. (but see also Fig. 2). The NMDA selective antagonist, ~,~-2-amino-5-phosphonovaleric acid (APV) had only a small effect on the e.p.s.p. (Cherubini et al., 1988) in normal
Ennis and Aston-Jones (1989) have shown that stimulation in the area of the prepositus hypoglossi caused a bicuculline-sensitive inhibition of the firing of LC neurons in uiuo. In the presence of CNQX or kynurenate a short latency synaptic potential was found in all neurons. This synaptic potential had a time-course which was similar to the excitatory amino acid component; that is, it rose to a peak in 10-25 msec and declined to the resting potential in about 200 msec (Cherubini et af., 1988). The reversal potential of this synaptic potential was dependent on the contents of the intracellular recording electrode. With a KCI-filled electrode the extrapolated reversal potential was near -40 mV, whereas with K-acetate-filled electrodes it was closer to -70 mV suggesting that this synaptic potential results from an increased conductance to chloride ions. Superfusion with bicuculline (10 p M ) completely and reversibly blocked this synaptic potential. The results suggest that the
170
inhibition of the firing induced by stimulation in the prepositus hypoglossi is a result of a GABAmediated i.p.s.p. The response of LC neurons to exogenously applied GABA results from an increased conductance to chloride ions and has been studied extensively (Osmanovic and Shefner 1990; Shefner and Osmanovic, this volume). i.p.s.p.s: glycine-mediated A strychnine-sensitive synaptic potential was observed only after both the GABA and glutamate synaptic potentials were blocked pharmacologically and the stimulus intensity was increased (Fig. 3). The time course of this synaptic potential was similar to both the e.p.s.p. and the GABAmediated i.p.s.p. Glycine (100-500 p M ) applied by superfusion caused an inward current at potentials more negative than the threshold for action potential generation ( - 5 5 mV with KClfilled recording electrodes). Both the synaptic potential and the effect of exogenously applied glycine were blocked by strychnine (10 nM-1 pM). The anatomical origin of this glycine input has not been described. i.p.s.p.s: noradrenergic/ adrenergic-mediated
Following the fast-synaptic potentials, there was a hyperpolarization which reached a peak 200-300 msec following the stimulus and had a total duration of about 2 sec (Fig. 4; Egan et al., 1983;
Surprenant and Williams, 1987). The amplitude of the hyperpolarization was graded with the intensity of stimulation, and had a maximum amplitude of 25 mV. The reversal potential for this synaptic potential was about -110 mV in the normal superfusion solution. The reversal potential was shifted to less negative potentials as the external potassium ion content was raised, indicating that an increased conductance to potassium ions mediated the synaptic potential. The i.p.s.p. was blocked by a,-adrenoceptor antagonists, phentolamine, yohimbine and idazoxan. In addition, cocaine or desmethylimipramine, used to block catecholamine reuptake, increased the amplitude and prolonged the duration of the i.p.s.p. (Egan et al., 1983; Surprenant and Williams, 1987). The i.p.s.p. was mimicked in several ways by exogenously applied a,-adrenoceptor agonists. NA, adrenaline, dopamine, clonidine and UK 14304, all hyperpolarized LC neurons by increasing the conductance to potassium ions. This hyperpolarization was antagonized by yohimbine and idazoxan (Williams et al., 1985). Finally, the concentration response curves to NA and adrenaline were shifted to the left by cocaine and desmethylimipramine (Surprenant and Williams, 1987). Taken together these results indicate that the i.p.s.p. is mediated by either NA or adrenaline acting on a,-adrenoceptors. There are two sources of adrenergic input to the LC. One coming from paragigantocellularis demonstrated in in viuo studies (Pieribone et al., 1988) and the second from dendrites of LC neu-
Control
-70mV rr* y.a
50 m s Fig. 3. A strychnine sensitive synaptic potential. The synaptic potential remaining after superfusion with picrotoxin (100 p M ) and CNQX (10 p M ) to block the GABA and excitatory amino acid receptors was decreased in amplitude by strychnine (100 nM). Strychnine (1 pM) reduced this synaptic potential to about 25% of control (not shown).
Yohimblne ( 1 p M )
7Y Control
I LmV
W n r s
Fig. 4. Noradrenergic inhibitory synaptic potential. The hyperpolarizing synaptic potential was completely blocked by yohimbine (1 pM).
171
rons themselves (Groves and Wilson, 1980). The anatomical demonstration of dendro-dendritic contacts suggest that LC neurons may release NA onto themselves (Groves and Wilson, 1980). There are several suggestions that the NA mediating the i.p.s.p. is released from LC neurons. When stimulating electrodes were placed within the body of the LC, i.p.s.p.’s were evoked with lower stimulus intensities and were larger in amplitude than with placement outside the nucleus. Pressure ejection of high potassium or glutamate in the area of the LC often caused a hyperpolarization which was sensitive to a,-adrenoceptor antagonists (Williams et d., 1987). After-blockade of a,-adrenoceptors, glutamate and high potassium evoked only depolarizations. The hyperpolarization therefore was considered to be indirectly mediated by NA release. Finally, the amplitude of the i.p.s.p. was depressed by excitatory amino acid antagonists, suggesting that at least part of the NA release arose from the generation of action potentials that were synaptically evoked. There are numerous other transmitter candidates that may modulate the activity of LC neurons. The actions of some of these candidates, applied exogenously, have been studied in great detail. Opioids Opioids hyperpolarize LC neurons by activation of an inwardly rectifying potassium conductance mediated through a pertussis toxin-sensitive GTP-binding protein (Williams et al., 1989). This action has been reviewed by Christie (this volume). There is no evidence that endogenous opioids are released in response to electrical or high-potassium stimulation in the slice preparation even in the presence of peptidase inhibitors (Williams et al., 1987).
either the spontaneous firing rate or membrane potential. Both the glutamate and GABA synaptic potentials were however depressed by 5-HT agonists that were selective for 5-HT1-subtype receptors (Bobker and Williams, 1989). The depolarization induced by glutamate or GABA applied exogenously (superfusion) was not affected by 5-HT, suggesting that the inhibition of the synaptic potentials was the result of presynaptic inhibition of transmitter release. There are some unresolved differences between these results and those found in uiuo, where the excitation induced by iontophoresis of glutamate was depressed by co-iontophoresis of 5-HT (Chouvet et al., 1988). In each case, however there is little or no effect on the spontaneous activity of neurons. We have not been able to resolve the action of endogenous 5-HT in the slice preparation. The presence of the 5-HT terminals was suggested because inhibition of 5-HT uptake by cocaine shifted the dose response curve to 5-HT about 40-fold to the left (Bobker and Williams, 1989). Acetylcholine Acetylcholine acts on both nicotinic (Egan and North, 1986) and muscarinic (Egan and North, 1985) receptors to increase the spontaneous activity of LC neurons. In each case, the increase in activity is induced by an underlying depolarization. The only evidence for an action of endogenous acetylcholine is the increase in spontaneous activity found when cholinesterase inhibitors were superfused (Egan, unpublished results). This increase in spontaneous activity was antagonized by superfusion with atropine. Electrically evoked cholinergic fast (hexamethonium-sensitive) or slow (atropine-sensitive) synaptic potentials have not been observed. Conclusions
Serotonin (5-HT) An extensive plexus of 5-HT-immunoreactive ter-
minals are found in the LC (Steinbusch, 1981). When superfused, 5-HT agonists had no effect on
The basal spontaneous activity of LC neurons can be affected by synaptically released transmitters to cause both excitation and inhibition in the slice preparation. Rapid excitation results from release
172
of an excitatory amino acid. Inhibition can be mediated by GABA, glycine and noradrenalhe/ adrenaline. These transmitters can play a role in the firing pattern of LC in the slice preparation and presumably in uiuo, but seem not to influence the rate of spontaneous activity in unstimulated preparations. There is a long list of other putative transmitters which have actions on LC neurons but that have not been shown to mediate synaptic potentials in the brain slice. Opioids, CRF, acetylcholine, somatostatin, adenosine, NPY and 5-HT are a few such compounds. Electrical stimulation may not be an appropriate method to evoke the synaptic release of these agents or they may have subtle effects that have gone undetected thus far. It is also possible that synaptic events mediated by these compounds will be dependent on having an intact animal in order to be unequivocally identified as transmitters.
References Aston-Jones, G. and Bloom, F.E. (1981a) Activity of norepinephrine-containing locus coeruleus neurons in behaving rats anticipates fluctations in the sleep-waking cycle. J. Neurosci., 1: 876-886. Aston-Jones, G. and Bloom, F.E. (1981b) Norepinephrinecontaining locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Bobker, D.H. and Williams, J.T. (1989) Serotonin agonists inhibit synaptic potentials in the rat locus ceruleus in uitro via 5-hydroxytryptamineIA and 5-hydroxytryptamineIB receptors. J. Pharrnacol. Exp. Ther., 250: 37-43. Cherubini, E.,North, R.A. and Williams, J.T. (1988) Synaptic potentials in locus coeruleus neurones. J. Physiol. (London), 406: 431-442. Chouvet, G.,Akaoka, H. and Aston-Jones, G. (1988) Serotonin selectively decreases glutamate-induced excitation of locus coeruleus neurons. C.R. Acad. Sci. Paris, 306: 339344. Egan, T.M. and North, R.A. (1985) Acetylcholine acts on M2-muscarine receptors to excite rat locus coeruleus neurones. Br. J. Pharmacol., 85: 733-735. Egan, T.M. and North, R.A. (1986) Actions of acetylcholine and nicotine on rat locus coeruleus neurons in vitro. Neuroscience, 19: 565-571. Egan, T.M., Henderson, G., North, R.A. and Williams, J.T. (1983) Noradrenaline mediated synaptic inhibition in locus coeruleus neurones. J. Physiol. (London), 345: 477-488. Ennis, M. and Aston-Jones, G. (1988) Activation of locus
coeruleus from nucleus paragigantocellularis: A new excitatory amino acid pathway in brain. J. Neurosci., 8: 36443657. Ennis, M. and Aston-Jones, G. (1989) GABA-mediated inhibition of locus coeruleus from the dorsomedial rostra1 medulla. J. Neurosci., 9: 2973-2981. Foote, S.L., Aston-Jones, G. and Bloom, F.E. (1980) Impulse activity of locus coeruleus neurons in awake rat and monkey is a function of sensory stimulation and arousal. Proc. Natl. Acad. Sci. USA, 77: 3033-3037. Groves, P.M. and Wilson, C.J. (1980) Monoaminergic presynaptic axon and dendrites in rat locus coeruleus seen in reconstructions of serial sections. J. Cornp. Neurol., 183: 853-862. Mayer, M. and Westbrook, G. (1987) The physiology of excitatory amino acids in the vertebrate central nervous system. Prog. Neurobiol., 28: 197-276. Moore, R.Y. and Bloom, F.E. (1979) Central catecholamine neuron system: Anatomy and physiology of the norepinephrine and epinephrine system. Ann. Rec. Neurosci., 2: 1113-1168. Osmanovic, S. and Shefner, S.A. (1990) y-Aminobutyric acid responses in rat locus coeruleus neurones in vitro: A current-clamp and voltage-clamp study. J. Physiol. (London), 421: 151-170. Pieribone, V.A., Aston-Jones, G. and Bohn, M.C. (1988) Adrenergic and non-adrenergic neurons of the C1 and C3 areas project to locus coeruleus: A fluorescent double labeling study. Neurosci. Lett., 85: 297-303. Steinbusch, H.W.M. (1981) Distribution of serotonin-immunoreactivity in the central nervous system of the rat-cell bodies and terminals. Neuroscience, 6: 557-618. Surprenant, A.M. and Williams, J.T. (1987) Inhibitory synaptic potentials recorded from mammalian neurones prolonged by blockade of noradrenaline uptake. J. Physiol. (Londonl, 382: 87-103. Valentino, R.J. and Aulisi, E.F. (1987) Carbachol-induced increases in locus coeruleus spontaneous activity are associated with an altered pattern of response to sensory stimuli. Neurosci. Lett., 74: 297-303. Valentino, R.J. and Foote, S.L. (1987) Corticotropin-releasing factor disrupts sensory responses of brain noradrenergic neurons. Neuroendocrinology, 45: 28-36. Williams, J.T., North, R.A., Shefner, S.A., Nishi, S. and Egan, T.M. (1984) Membrane properties of rat locus coeruleus neurones. Neuroscience, 13: 137-156. Williams, J.T., Henderson, G. and North, R.A. (1985) Characterization of alpha,-adrenoceptors which increase potassium conductance in rat locus coeruleus neurones. Neuroscience, 14: 95-101. Williams, J.T., Christie, M.J. and North, R.A. (1987) Potentiation of enkephalin action by peptidase inhibitors in rat locus coeruleus in vitro. J. Pharmacol. Exp. Ther., 243: 397-401. Williams, J.T., North, R.A. and Tokimasa, T. (1989) Inward rectification of resting and opiate-activated potassium currents in rat locus coeruleus neurons. J. Neurosci., 8: 42994306.
C.D. Barnes and 0. Pompeiano (Eds.) Progress m Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
173 CHAPTER 12
Developmental aspects of the locus coeruleus-noradrenaline system K.C. Marshall
',M.J. Christie
2,
P.G. Finlayson and J.T. Williams
I Department of Physiology, University of Ottawa, Smyth Road Ottawa, Ontario, Canada, Department of Pharmacology, University of Sydney, Sydney, N.S. W., Australia, Department of Ophthalmology, University of British Columbia, Vancouver, Canada and Vollum Institute of Biomedical Research, Oregon Health Sciences University, S. W. Sam Jackson Park Road, Portland, OR, U.S.A.
The locus coeruleus-noradrenaline (LC-NA) system exhibits an early developmental pattern, so that its nerve terminals are present in target areas before formation of most synapses. Several properties of the source neurons in the LC change substantially during early postnatal periods: spontaneous activity patterns, responsiveness to sensory stimulation, and responsiveness to NA. The effect may be to confer enhanced responsiveness of LC neurons, and an enhanced release of NA in target areas, during early postnatal development. Developmental changes in density of adrenoceptors or adrenergic responsiveness in target areas have also been documented. The usual pattern is a progressive increase in adrenergic ligand binding, with some reduction during later
phases of development. However, there are a number of examples of receptor subtypes and region-specific transient binding during the first few weeks of postnatal life, followed by reductions to very low levels. These observations may reflect developmentally transient adrenergic responsiveness in certain target areas. NA and the LC-NA system have been implicated in the control of morphological and functional properties of neurons in target areas, and in the control of developmentally important biochemical systems (ornithine decarboxylase). NA, as well as other neurotransmitters, may individually, or in cooperation, exert important trophic influences during a restricted developmental period.
Key words: noradrenaline, locus coeruleus, development, electrophysiology, adrenoceptor binding
Introduction
A feature of locus coeruleus (LC) neurons which suggests a role in trophic control of target areas is the early schedule of development. Lauder and Bloom (1974) have shown that these cells undergo differentiation at about 12 days of gestational age in the rat, i.e., considerably earlier than cells in target areas such as hippocampus
and cerebellum. Differentiation is also complete in LC at an earlier time than in raphe nuclei or substantia nigra pars compacta, which contain serotonin and dopamine-using neurons respectively. Catecholamine fluorescence can be demonstrated in LC at embryonic day 14 (Maeda and Dresse, 1969; Olson and Seiger, 1972). Innervation of target areas by LC neurons also occurs at a relatively early developmental stage. Fibres
174
staining for catecholamine fluorescence are observed as early as embryonic day 18 in neocortex (Levitt and Moore, 1979) and hippocampus (Loy and Moore, 1979), and at least by birth in the cerebellum (Loren et al., 1976; Yamamoto et al., 1977). This represents a period prior to influx of most other afferent systems (e.g., Loy and Moore, 1979) and is consistent with a possible role for LC in regulation of neuronal development. Electrical properties of LC neurons
Developmental changes in membrane electrical properties of LC neurons were found by Williams and Marshall (1987) who described rhythmic membrane oscillations during intracellular recordings from LC neurons in brain slices from young rats. The oscillations exhibited progressively increasing frequency in slices from rats of 8 days ( < 1 Hz) to 20 days (about 3 Hz) (Fig. 1). By day 27, the rhythmic oscillations were seldom observable (Christie et al., 1989), as is the case in slices from adult animals. The rhythmic activity was insensitive to tetrodotoxin, but appeared to be calcium-dependent in that it was blocked in bathing solutions with low-calcium/high-magnesium content (Williams and Marshall, 1987). Hyperpolarizing current injection blocked action potentials generated at peaks of the rhythmic waves, but not the oscillations themselves, which continued at about the same frequency (Fig. 1). Under voltage clamp, current oscillations were observed at a similar frequency, suggesting that the origin of the rhythmic potentials was at a location remote from the recording electrode. The frequency of the oscillations could, however, be modified by bath-applied drugs: phenylephrine increased the frequency (Williams and Marshall, 1987) as did a muscarinic cholinergic agonist, while noradrenaline (NA) (through a2adrenoceptors) and p-opiates reduced the frequency (M.J. Christie, J.T. Williams and K.C. Marshall, unpublished observations). The studies described above suggest that remote dendritic regions of the LC neuron con-
#
Fig. 1. Representative intracellular recordings of spontaneous activity of locus coeruleus (LC) neurons at various ages. Action potentials (threshold about - 5 5 mV) usually arose from the peak depolarizing phase of rhythmic potential variations. Full action potential amplitude is not shown (limited by frequency response of chart recorder). The membrane was hyperpolarized at arrow by passing current (50-200 PA) through the recording electrode. The amplitude of rhythmic potential variations declined and their frequency increased with age, becoming undetectable by 27 days. (From Christie et al., 1989.)
tribute much to the electrical activity occurring at the soma. Furthermore, analysis of electrotonic potentials in LC neurons indicated the presence of two electrical compartments, which could be interpreted to suggest a large contribution of dendrites to the electrical properties (Christie et a[., 1989). These findings were also consistent with the possibility that remote electrical activity arose from neighbouring LC neurons which were electrically coupled to the impaled cell. The first evidence suggesting electrical coupling among neonatal LC neurons was the synchrony of rhythmic oscillation of membrane potential throughout the nucleus (Christie et d., 1989). This was demonstrated by simultaneous intracellular recording of pairs of LC neurons
175
which were separated by about 100 to 300 pm. As shown in Figure 2, the peaks of slow rhythmic oscillations in a pair of LC cells were fully in phase, within the resolution of the recording ( < 30 msec). Full action potentials usually arose from the peaks of depolarizations but were not necessarily synchronous. Similar observations were made in 48 pairs of LC neurons from animals less than 24 days old (Christie et al., 1989). After this age synchronous activity declined. One possible mechanism of the synchronous activity is that neonatal LC neurons are directly coupled by low-resistance pathways. This was directly demonstrated by experiments like the one shown in Figure 3. During simultaneous recording from two neurons, a large current injected into one LC neuron (sending neuron) evoked a potential change in a second LC neuron about
+5
1
1
0
m
v
I 2sJ
Fig. 2. Rhythmic potentials are fully synchronous between all neurons in the LC from a 3-day-old rat. Each pair of records represents simultaneous recordings of membrane potential from two neurons. Impalement of neuron 1 was maintained for 3 h. Neurons 2-5 were sequentially impaled with a second electrode in the ipsilateral LC. Recording electrodes were separated by about 100-300 k m . Spontaneous action potentials were prevented by passing current (50-150 pA) through the recording electrode. (From Christie et al., 1989.)
+2
I
100 ms
Fig. 3. Electrotonic coupling between LC neurons. Simultaneous recordings from two cells from a 9-day-old rat (electrode separation about 80 Km). Traces show electrotonic potentials at each electrode (average of 10) evoked by current pulses (1-2 nA, 200 msec) passed through electrode 1. The amplitude of the electrotonic potential recorded in neuron 2 was about 2% of that recorded in neuron 1. Solution contained tetraethylammonium (20 mM), magnesium ( 5 mM), barium (200 pM),and TTX (300 nM). (From Christie et al., 1989.)
100 p m away (receiving neuron). This occurred during blockade of synaptic transmission and was not observed if one electrode was moved to the outside of the impaled neuron. It was possible to detect electrotonic coupling only in animals less than 10 days old. The slow time constant of the electrotonic potential in the receiving neuron and the small coupling ratio (0.005-0.02) suggested that each LC neuron was coupled over a long electrotonic distance to about two other neurons. Only slow potential changes (less than about 1 V/sec) would be expected to propagate through such a network, synchronizing slow spontaneous oscillations, but not necessarily full action potentials. It is not yet clear whether synchronous subthreshold activity occurs in the neonatal LC in ciuo. It is conceivable that electrical coupling is artifactually induced by slice preparation. Sectioning of the dendritic field has been proposed to produce artifactual coupling in slices of guinea-pig cortex (Gutnick et al., 1985). The most likely basis of low-resistance coupling, the presence of gap-junctions, has yet to be studied in neonatal LC. However, other observations are
176
consistent with slow synchronous activity in uiuo. Mean rates of spontaneous firing in young rats exhibited a similar developmental change to that seen in the slices, progressing from below 1 Hz in 1-6 day rats to about 3 Hz at 19 days and older, including adult animals (Nakamura et al., 1987). Sakaguchi and Nakamura (1987) have also shown that LC neurons in young rats may exhibit periods of spontaneous synchronous discharges, in extracellular recordings of more than one unit. (See also Nakamura and Sakaguchi, 1990.)
!O . N!
I
. .
60 METH I
h
W . . .k - h .Q 60 CLON
80 NA
2 min 1 min
b
Adrenergic responsiveness of LC neurons The responsiveness of LC neurons to adrenergic drugs has also been found to change with development. In slices from adult rat, NA causes a hyperpolarization, mediated by an a,-adrenoceptor-activated potassium conductance increase (Williams et al., 1985). However, in explant cultures containing LC neurons, Finlayson and Marshall (1984) found that in younger cultures, the response to NA was characteristically a biphasic hyperpolarization/depolarization. The depolarizing component was mediated by a,-adrenoceptors in that it could be mimicked by a,-adrenoceptor selective agonists such as methoxamine (Fig. 4), and blocked by the a,-adrenoceptor antagonist prazosin (Fig. 5; Finlayson and Marshall, 1986, 1988). The a,-adrenoceptor response to iontophoretically applied agents appeared to be direct, since it was still present in bathing solutions containing tetrodotoxin, or low-calcium/ high-magnesium solutions. In a similar manner, the hyperpolarizing component was mimicked by clonidine (Fig. 4) and blocked by yohimbine (Fig. 51, indicating mediation by a,-adrenoceptors. Williams and Marshall (1987) tested the adrenergic responsiveness of LC neurons in brain slices from young animals and found that the a,-adrenoceptor selective agonist phenylephrine gave depolarizing responses which were blocked by prazosin. In slices from adult rats, however, phenylephrine had no effect, or evoked a small hyperpolarizing response. The response to NA
10 CLON
Fig. 4. Iontophoretic application of noradrenaline (NA) and adrenergic agonists to LC neurons in explant cultures (applications indicated by bars beneath records). (a) NA evokes biphasic response (upper left) while methoxamine gives a depolarization with no evident hyperpolarizing phase (upper right); clonidine application results in slow hyperpolarization with no apparent depolarizing phase. (b) application of clonidine during continuing series of NA applications; response to clonidine is a slow hyperpolarization; NA responses appear as simple depolarizations during maximal clonidine response and gradually return to control biphasic pattern. Time calibration bar (inset) represents 2 min for (a) and 1 min for (b). (From Finlayson and Marshall, 1986.)
was also different in slices from young animals, in that the hyperpolarization was prolonged, and the initial hyperpolarization was not maintained throughout the NA application (Fig. 6). Both effects were more pronounced in the youngest animals (8 days), and both were blocked by prazosin. In slices from adult animals, the only remnant of these effects is a small increase in the amplitude of the NA-mediated hyperpolarization by prazosin (M.J. Christie and J.T. Williams, unpublished observations). Studies of adrenergic responsiveness of LC neurons during early development have also been carried out in uiuo. Iontophoretically applied NA strongly inhibited spontaneous firing of LC neurons in neonatal rats, and even seemed to have more potent inhibitory effects than in adults (Kimura and Nakamura, 1987). This effect was
177
blocked by the preferential a,-adrenoceptor antagonist piperoxan, indicating an early expression of a,-mediated inhibitory mechanisms. Conversely, autoinhibition following antidromic activation of LC neurons was not blocked by piperoxan in 1-8 day animals, but was blocked in most cells in animals older than 15 days. This suggests that although the a,-adrenoceptor response mechanisms were present in neonates, autoinhibitory synaptic mechanisms were only established at later stages of development.
YOH PRAZ A
-
C 1pM
-
20nM
-
E luM
100nM
F 20 min washout
G 1IAM
Fig. 5. Dose-dependent antagonism of depolarizing response to NA (iontophoretic applications indicated by bars beneath records). Biphasic response to NA (a) is converted to selective and enhanced depolarization in the presence of 1 p M yohimbine (YOH) (b). The depolarization is increasingly antagonized by 20, 50, and 100 nM prazosin (PRAZ) (c-el. The faster washout of yohimbine results in a simple hyperpolarizing response to NA shortly after washout of drugs began (f). Reversal of prazosin blockade is evident during subsequent perfusion with yohimbine (g) (after 34 min of washout). (NA was applied with an iontophoretic current of 40 nA). (From Finlayson and Marshall, 1988.)
Nakamura et al. (1988) have also demonstrated excitatory a,-adrenoceptor-mediated responses in LC neurons, primarily in young animals. These responses could be evoked by iontophoretic application of small amounts of NA or of phenylephrine. Excitatory components of the response to antidromic activation of LC neurons by dorsal bundle stimulation were also found in young animals, and both these and the drugevoked excitations were blocked by the a,-adrenoceptor antagonist HEAT (2-@-(4-hydroxyphenylethylaminomethy1)-tetralone). Curiously, these excitatory responses seemed activity-related, in that they were primarily observed in cells having low levels of spontaneous activity. It is also interesting that only in young animals, application of HEAT resulted in decreased rates of spontaneous activity, suggesting that there may be a tonic excitatory influence exerted through a,adrenoceptors in early stages of development. The reason for the marked diminution of the a,-adrenoceptor-mediated excitation in later stages of development is not clear, though various possibilities have been discussed (Finlayson and Marshall, 1988; Nakamura et al., 1988; Nakamura and Sakaguchi, 1990). One possible explanation is a reduction in these receptors during postnatal development, and such effects have been demonstrated by Jones et al. (1985a) for a,-adrenoceptors in certain brain areas. The LC was, however, found to exhibit binding of iodinated HEAT in adult rat brain (Jones et al., 1985b), and appeared to exhibit progressively increasing binding during early postnatal development (Jones et aZ., 1985a). Conversely, Young and Kuhar (1980) found only low levels of a,-selective binding in LC of mature rats, using a different antagonist (H3 WB-4101). It would be desirable, therefore, to test this possibility in another manner. What influence might the developmentally transient excitatory adrenergic mechanism have on the function of the LC? From the work of Nakamura et al. (1988) cited above, it can result in elevated levels of spontaneous firing, but it could also be predicted that the LC neurons in
178 N A 100 UM
NA 30 LIM
Prozasin
'
-
Fig. 6 . NA hyperpolarizations (bath-applications indicated by bars above records) were increased in amplitude and decreased in duration following superfusion with prazosin. This experiment was carried out in the presence of cocaine (10 yM). Top, NA produced inhibition of firing, an inhibition of the slow calcium spikes, and a hyperpolarization, the duration of which wa5 dependent on the concentration applied. The hyperpolarization declined during the application of NA. Following the washout of NA, the membrane potential rehyperpolarized before returning toward the resting level. Bottom, Following superfusion with prazosin (500 nM), NA caused a membrane hyperpolarization that was greater in amplitude and shorter in duration for both concentrations of NA. This cell was from a 9-day-old animal. (From Williams and Marshall, 1987.)
young animals might be more responsive to afferent synaptic activation as a result of this mechanism. In fact, there is evidence that this is the case, also from the work of Nakamura and coworkers. From morphological studies, on the rat, it has been shown that synaptogenesis commences in the LC neurophil and on cell somata at embryonic days 19 and 20, respectively, and increases during the first two months of postnatal life (Lauder and Bloom, 1975). In keeping with this, Sakaguchi and Nakamura (1987) have found responses of LC neurons to somatosensory stimuli, even at prenatal stages of development. In anesthetized neonatal animals, the LC neurons were highly sensitive to somatosensory stimulation, in contrast to anesthetized adult rats, which respond well to noxious but poorly to non-noxious stimulation (Kimura and Nakamura, 1985; Nakamura et al., 1987). Responsiveness to auditory and visual stimuli was found after about two weeks of postnatal development. While there is no direct evidence that the enhanced responsiveness is due to the a,-adrenergic responsiveness in young animals, this appears to be a reasonable possibility in view of the demonstrations of a l adrenoceptor influence on responses to dorsal bundle stimulation and on spontaneous activity cited above.
Assuming that this enhanced responsiveness of LC neurons during neonatal and early postnatal development is also present in unanesthetized animals, the activity of these cells would be more governed by sensory activation, particularly since rates of spontaneous activity are relatively low. It is possible that this developmental property of LC neurons has some significance in the role of LC innervation of target areas during this period. The possibility should also be considered that the a,-adrenoceptor-mediated actions may be expressed at the nerve terminals of LC neurons as well as at the somata. While release of NA in target areas such as the neocortex appears to be unaffected by a,-adrenergic drugs in the adult rat (Singer, 1988), we are not aware of any tests of this type during the early postnatal period. Such a mechanism might result in enhanced levels of NA release in target areas during development. Adrenergic receptors: responsiveness in LC target areas
In most developmental studies of neurotransmitter receptors, binding studies of radiolabelled ligand in homogenized tissue, or autoradiography have been used, though there are some examples of functional tests, e.g., second messenger responses to the neurotransmitter or electrophysio-
179
logical measures of responsiveness to agonist application. The general developmental pattern is a progressive increase in receptors commencing in the prenatal or early postnatal period, and achieving adult levels after 4-6 weeks (in the rat).
P -Adrenoceptors The development of P-adrenoceptor systems has been the focus of several investigations. In the cerebral cortex, Harden et al. (1977a) found a progressive increase in iodinated hydroxylbenzylpindolol binding in homogenized tissue, to a maximum at about 15 days, with a subsequent decline of about 30% in the adult. The formation of cyclic AMP from ATP in response to L-isoproterenol in cortical slices followed a similar developmental course, as far as it was studied (to about 17 days). The content of NA in cortical tissue was also measured, and was found to follow a more gradual pattern of increase, continuing into adulthood. These results suggest some independence between the noradrenergic innervation and the development of P-adrenoceptors and the coupled adenyl cyclase system. This was confirmed in a related study which included animals in which cortical levels of NA were severely depleted, subsequent to 6-hydroxydopamine (6OHDA) treatment as neonates. Although the maxima1 levels of binding density and cyclic AMP response were significantly greater in the lesioned animals the developmental time courses of these parameters were very similar to the controls (Harden et al., 1977b). Comparable data were reported for cortical tissue using a different ligand (dihydroalprenolol), (Kawai et al., 1984). Very similar results to these have been found for development of P-adrenoceptor binding, p-stimulated cyclic AMP responses and effects of 6OHDA treatment in the pineal gland, which receives adrenergic innervation from the superior cervical ganglion (Weiss et al., 1984). Other investigators have used methods to distinguish PI selective from P,-adrenoceptor binding. Pittman et al. (1980) used displacement of iodinated hydroxypindolol by selective PI and P2
drugs, and found considerably higher density P than P,-adrenoceptor binding, but patterns of development for each subtype that were similar to that described by Harden et al. (1977a) for undifferentiated P-adrenoceptor binding. In confirmation of this pattern, Duman et al. (1989) described a progressive increase in expression of P , and @,-adrenoceptor mRNA from low levels at postnatal day 1 through to adulthood. The developmental pattern of P-adrenoceptor binding in the cerebellum was different)from that in the cerebral cortex in two ways; the P2 binding was considerably higher than P I binding, and the two followed different temporal patterns (Pittman et al. 1980). Whereas the p2 binding density increased progressively from one week to adult levels at six weeks, the P I binding peaked at about three weeks, and then declined to very low adult levels. East and Dutton (1982) used the nonselective P-adrenergic ligand dihydroalprenolol, and also reported a transient peak of binding density at about 15 days in mouse cerebellum, whereas Bartolome et al. (1987) reported progressively increasing levels of total P binding in the rat cerebellum, using iodinated pindolol. Further studies by this latter group confirmed this pattern for each of the P-adrenoceptor subtypes (Lorton et al., 19881, and also confirmed the opposite ratios of PI/& binding in the cerebral cortex and cerebellum. The difference in observed patterns of PI-adrenoceptor development in the cerebellum remains unresolved at this time. It appears therefore that in most cases, Padrenoceptors exhibit a progressive development in density toward adult levels, though there is often some decrease from maximal levels during the first several weeks of postnatal age to the adult value. In one example, however, there is evidence for a developmentally transient expression of a P-adrenergic mechanism. Lanfumey and Adrien (1988a,b) have reported that neuronal activity in the dorsal raphe nucleus of rats was inhibited by iontophoretically applied 'P-adrenoceptor antagonists in animals less than two weeks
180
old, but not in older animals. Within the 3-13 day age range more cells were inhibited in the younger animals. In animals receiving neonatal 6-OHDA lesions of the LC, however, the inhibition of activity by P-antagonists was also observed in mature rats. The effects of the lesions were selective, in that responses to the a,-adrenoceptor agonist clonidine, to 5-hydroxytryptamine and to y-aminobutyric acid were not different from those observed in control animals. It is curious that application of P-agonists does not appear to affect these neurons in the normal mature, young or LC-lesioned animals, though the agonists can overcome the inhibitory effect of the antagonists. The results suggest a developmentally transient P-mediated facilitation of dorsal raphe neurons, which is maintained to adult stages in the absence of adrenergic innervation from LC. It would be interesting to know the effects of LC lesions made after two weeks of age on the response to P-antagonists. a-Adrenoceptors Development of a-adrenoceptors and response mechanisms have also been the subject of several studies. As discussed above, in the LC itself a,-adrenoceptor-mediated inhibitory responses are present at an early age, and are maintained into adult stages, while a,-mediated excitations are only transiently apparent in early postnatal periods. In LC target areas, developmentally transient a-adrenoceptor functions have also been found. Binding studies have revealed developmentally transient a-adrenoceptor-related properties and the patterns appear to be characteristic for particular receptor subtypes and anatomical regions. Deskin et al. (1981) used incorporation of labelled orthophosphate into phospholipids in whole brain homogenates as a measure of an a-adrenergic response to NA, and found a progressive increase in the response, in parallel with the development of synaptosomal uptake of tritiated NA. In this study, the developmental pattern of a-responsiveness was markedly altered in brains of animals which had received
neonatal 6-OHDA lesions, in that NA responses were markedly depressed below control values, with some return toward normal at three weeks. This study suggests that development of this adrenergic response is at least in part dependent on the catecholamine innervation. This supports the proposal by Marshall et al., (1981) that development of a-adrenergic depolarizing responses in cultured spinal neurons is dependent on adrenergic innervation. The proposal was based on observations that neurons co-cultured with brain stem explants containing LC showed a much higher frequency of depolarizing responses to NA than those grown in the absence of such explants. In binding studies of whole brain minus cerebellum, Morris et al. (1980) observed that both a and a,-adrenoceptor binding sites increased progressively from birth to about 20 days, and then decreased to adult levels (by approximately 25%). Using the same preparation (whole brain minus cerebellum) from mouse, Maurin et al. (1985) found similar results, though the developmental peaks were higher relative to adult levels, and the peak for a,-adrenoceptor binding occurred at about five weeks. Hartley and Seeman (1983) found that there were regionally distinct patterns of development. In the “mesolimbic” area, a,-adrenoceptor binding density increased to a peak at about three weeks, and then decreased to adult levels. In other areas, (hippocampus, frontal cortex and hypothalamus) there was a biphasic increase, reaching maximal levels at about 30 days, with some subsequent decrease. The pattern for a , binding was also regionspecific; in mesolimbic areas, a biphasic increase to a maintained level of about 30 days, in the other p e a s an early rapid increase followed by a slower progressive increase to 10 weeks. Regional differences in developmental patterns for a-adrenoceptor binding were also apparent in other studies. Bartolome et al. (1987) assessed cerebral cortex, cerebellum, and midbrain plus brain stem, and found characteristic patterns for each. In cortex, a l - and a,-adrenoceptor ligands showed increasing levels of binding
,-
181
to about 32 days, followed by some decrease. In the cerebellum conversely, both exhibited relatively high levels at ten days, followed by decreases to much lower values. In midbrain and brain stem, the pattern was closer to that of the cerebellum than the cortex. In studies of a l adrenoceptor binding, Johnson et al. (1984) found increases in cortical binding from 7 to 60 days, but decreases in hypothalamus and preoptic area, while binding in the amygdala remained relatively constant. Dermon and Kouvelas (1988) carried out similar studies in chick brain, and found an early peaking of a,-adrenoceptor binding in optic lobes, brain stem and cerebellum. Dramatic differences in regional patterns of a ,-adrenoceptor development were shown by Jones et al. (1985a) in an autoradiography study using iodinated HEAT. While most regions displayed a progressively increasing pattern, the bed nucleus of stria terminalis, globus pallidus and pineal gland showed early peaks (at 1-2 weeks), and then decreases to low levels. A transient maximal binding of the same ligand in the pineal was shown by Sugden and Klein (1985). In studies of the cerebral cortex Kawai et aL, (1984) demonstrated maximal binding of a l - and a,-adrenoceptor ligands at 2-4 weeks, with subsequent declines. Schoepp et al. (1985) showed progressive increases in a 1 binding to 37 days, with increases in NA-stimulated inositol phosphates occurring over the same time period. In an interesting report on a,-adrenoceptor binding, Nomura et al. (1984) found that the developmental patterns in the cerebral cortex were markedly different for low- and high-affinity binding sites. The low-affinity binding increased from 7 to 30 days and maintained that level to maturity. However, for the high-affinity site, maximal binding was observed at 15 days, with subsequent decreases to lower levels (about 60% of the maximum).
Conclusions While the above reports show varied patterns of adrenoceptor development, there is a frequent
observation of progressive increase of binding over the first 2-4 weeks of postnatal life (in rats) followed by decreases to lower levels. In some cases, these decreases are dramatic, e.g., a , binding in cerebellum (Bartolome et aL, 1987; Dermon and Kouvelas, 19881, and a 1 binding in the globus pallidus (Jones et al., 1985a). It seems important that the developmental patterns are often distinct for specific anatomical regions, and for specific receptor subtypes. There are several possible reasons for observations of developmental maxima in binding. There might be developmental changes in affinity of the receptors, though where this has been studied, it has generally not explained the differences in binding density, ( e g , Nomura et al., 1984). It is conceivable that the reductions in binding reflect a “down regulation” of receptors, subsequent to synaptic influence by afferent noradrenergic fibres. Such changes might also be explained by the tendency to report binding density as concentration per weight of protein (East and Dutton, 1982), and the possibility that non-neuronal protein is added at a greater rate than neuronal protein after a certain stage of development. Different developmental schedules for neurons and glia in various brain areas might offer some explanation for the regional differences observed (Bartolome et al., 1987). In autoradiography studies using tritiated ligands, there is also the possibility of “quenching” due to development of myelin (Kuhar and Unnerstall, 1985). In at least some of the examples described, however, such factors seem unable to explain the transient peaks of binding in view of the magnitude of the change, or the different patterns observed for receptor subtypes in the same tissue. This leaves the possibility that in some cases transient peaks of activity reflect specific developmental functions for the associated neurotransmitter. As the large number of receptor subtypes is increasingly recognized, it may not be surprising to find that certain of them are selectively expressed during development, (e.g. , Ventura et al., 1984). The growing use of molecular biological methods for identification of receptor sub-
182
types should permit testing of this concept in the near future.
Deuelopmental role of the LC-NA system The functions of this system in development remain unclear. Morphological abnormalities have been reported in cells of LC target areas, following neonatal treatment of rats with 6-OHDA. Felten et al. (1982) found diminished length and branching of pyramidal cell dendrites in the cerebral cortex, though the abnormalities were restricted to a part of the pyramidal cell population, and almost total depletion of NA in the cortex was required before the changes were observed. Blue and Parnavelas (19821, in an electron microscopic study, found that 6-OHDA lesions in neonates resulted in increased synaptic density in the visual cortex and suggested that the noradrenergic system inhibited synaptogenesis during development. A most interesting indication of a role for NA in development stems from an extensive body of work by Kasamatsu and collaborators in the visual cortex of cats (see also Kasamatsu, this volume). Because of the detailed studies available on the physiology and functional development of neuronal responses in the visual cortex, this represents an exemplary system for the investigation of factors influencing development. The perturbation of visual function, (e.g., monocular deprivation) during a “critical period” of development can result in a dramatic loss of responsiveness of cortical neurons to input from the manipulated eye. This developmental plasticity of ocular dominance has been used as a test substrate for factors which may have trophic functions. It has been found that local (cortical) or intraventricular treatment with 6-OHDA to deplete NA is associated with a reduction of the plasticity, and that it can be at least partially restored by local infusion of NA (see Kasamatsu and Shirokawa, 1988, for a review of this work). Challenges to this concept have arisen, particularly due to the failure of other groups to repeat these
results using other methods of depleting NA, (e.g., Bear and Daniels, 1983; Daw et al., 1985), and it has been proposed that both NA and acetylcholine must be involved to disrupt the plasticity (Bear and Singer, 1986). Other studies indicate the P-adrenoceptor antagonists can also block the plasticity (Shirokawa and Kasamatsu, 1986) and that NA, acting through P-adrenoceptors is considerably more potent than muscarinic cholinergic agonists in reversing the reduced plasticity from 6-OHDA treatment (Kasamatsu et al., 1989). It is possible to conclude that NA does have a role in regulation of developmental plasticity in the cortex, but that it depends in part on interaction with or contributions by other neurotransmitter systems. A review of this area has recently been published (Gordon et al., 1988). Much work has been carried out on developmental aspects of neurotransmitter systems in the visual cortex. For the NA system, Jonsson and Kasamatsu (1983) have reported a transient increase in P-adrenoceptor binding in occipital cortex of cats during the critical period. Shaw et al. (1986) have used autoradiography to show that there is a redistribution of P-adrenoceptors during development, such that at early stages there is a concentration in layer IV, with a later shift of higher binding density to the more superficial layers (a similar pattern was noted for muscarinic receptor binding). It may be, therefore, that developmentally transient adrenergic mechanisms are involved in control of cortical plasticity, though more work is required to establish the possible relationships. While little is known of the mechanisms for developmental control, one possibility is evident from the studies of Slotkin and co-workers. NA has been found to stimulate activity of ornithine decarboxylase, a synthetic enzyme for developmentally important polyamines. In the early postnatal period (to 12 days), there is a transient and marked stimulation of this enzyme in the cerebellum, and this appears to be mediated by Pzadrenoceptors and adenyl cyclase activation (Morris and Slotkin, 1985). In addition, intracis-
183
ternally administered 6-OHDA resulted in lowered NA and polyamine levels, and reduced cerebellar growth during development (Lau et al., 1990). Regulation of polyamines, therefore, represents one possible mechanism for control of neural development by NA. Other mechanisms which may contribute to trophic functions of the LC-NA system may include the enhanced responsiveness of LC neurons during early postnatal development, as discussed in the first part of this review. In addition, the transient expression of adrenoceptors and some adrenergic responses during the postnatal period may underlie developmental functions. NA appears to have neuromodulatory functions on target cells, including the potentiation of excitatory or inhibitory neurotransmitters and it has been suggested that these potentiations may be of short or long duration (Marshall and Tsai, 1988; see also Segal et al., this volume). The selective expression of some of these modulatory functions during critical periods of development could therefore also be responsible for some trophic functions of the noradrenergic system. Acknowledgements
The assistance of Donna Mulder and Denyse LongprC in preparation of the manuscript is gratefully acknowledged. The research of the authors is supported by grants from National Institutes of Health to J.T.W. (DA 04523 and MH 45003) and from the Medical Research Council of Canada to K.C.M. References Bartolome, J.V., Kavlock, R.J., Cowdery, T., Orband-Miller, L. and Slotkin, T.A. (1987) Development of adrenergic receptor binding sites in brain regions of the neonatal rat: Effects of prenatal or postnatal exposure to methylmercury. Neurvtvxicolvgy, 8: 1-14. Bear, M.F. and Daniels, J.D. (1983) The plastic response to monocular deprivation persists in kitten visual cortex after chronic depletion of norepinephrine. J. Neurvsci., 3: 407416.
Bear, M. and Singer, W. (1986) Modulation of visual cortex plasticity by acetylcholine and noradrenaline. Nature (Lvndonf,320: 172-176. Blue, M.E. and Parnavelas, J.G. (1982) The effect of neonatal 6-hydroxydopamine treatment on synaptogenesis in the visual cortex of the rat. J. Comp. Neurvl., 205: 199-205. Christie, M.J., Williams, J.T. and North, R.A. (1989) Electrical coupling synchronizes subthreshold activity in locus coeruleus neurons in vitrv from neonatal rats. J. Neurosci., 9:3584-3589. Daw, N.W., Videen, T.O., Rader, R.K., Robertson, T.W. and Coscia, C.J. (1985) Substantial reduction of noradrenaline in kitten visual cortex by intraventricular injections of 6-hydroxydopamine does not always prevent ocular dorninance shifts after monocular deprivation. Exp. Bruin Res., 59: 30-35. Dermon, C.R. and Kouvelas, E.D. (1988) Binding properties, regional ontogeny and localization of adrenergic receptors in chick brain. Int. J. Dev. Neurvsci., 6: 471-482. Deskin, R., Seidler, F.J., Whitmore, W.L. and Slotkin, T.A. (1981) Development of a-noradrenergic and dopaminergic receptor systems depends on maturation of their presynaptic nerve terminals in the rat brain. J. Neurvchem., 36: 1683-1690. Duman, R.S., Saito, N. and Tallman, J.F. (1989) Development of P-adrenergic receptor and G protein messenger RNA in rat brain. Mol. Bruin Res., 5: 289-296. East, J.M. and Dutton, G.R. (1982) Development of betaadrenergic receptors in normal and mutant mouse cerebellum. Deu. Neurvsci., 5: 117-124. Felten, D.L., Hallman, H. and Jonsson, G. (1982) Evidence for a neurotrophic role of noradrenaline neurons in the postnatal development of rat cerebral cortex. J. Neurocytvl., 11: 119-135. Finlayson, P.G. and Marshall, K.C. (1984) Hyperpolariziiig and age-dependent depolarizing responses of cultured locus coeruleus neurons to noradrenaline. Deu. Bruin Rex, 15: 167-175. Finlayson, P.G. and Marshall, K.C. (1986) Locus coeruleus neurons in culture have a developmentally transient a , adrenergic response. Dev. Bruin Res., 25: 292-295. Finlayson, P.G. and Marshall, K.C. (1988) Characterization of a developmentally transient adrenergic depolarizing response in cultured locus coeruleus neurons. Can. J. Physiol. Phurmucvl., 66: 1547-1554. Gordon, B., Allen, E.E. and Trombley, P.Q. (1988) The role of norepinephrine in plasticity of visual cortex. Prog. Neurvbiol., 30: 171-191. Gutnick, M.J., Lobel-Yaakov, R. and Rimon, G. (1985) Incidence of neuronal dye-coupling in neocortical slices depends on the plane of section. Neuroscience, 15: 659-666. Harden, T.K., Wolfe, B.B., Sporn, J.R., Perkins, J.P. and Molinoff, P.B. (1977a) Ontogeny of P-adrenergic receptors in rat cerebral cortex. Bruin Res., 125: 99-108. Harden, T.K., Wolfe, B.B., Sporn, J.R., Poulos, B.K. and Molinoff, P.B. (1977b) Effects of 6-hydroxydopamine on the development of the beta adrenergic receptor/adenylate cyclase system in rat cerebral cortex. J. Phurmucvl. Exp. Ther., 203: 132-143.
184 Hartley, E.J. and Seeman, P. (1983) Development of receptors for dopamine and noradrenaline in rat brain. Eur. J. Pharmacol., 91: 391-397. Johnson, A.E.; Nock, B., Ryer, H. and Feder, H. (1984) Hypothalamic and preoptic area a,-receptor concentrations decrease, and cerebral cortical a,-receptor concentrations increase during postnatal maturation of male guinea pigs. Neuroendocrinology, 38: 243-247. Jones, L.S., Gauger, L.L., Davis, J.N., Slotkin, T.A. and Bartolome, J.V. (1985a) Postnatal development of brain alpha,-adrenergic receptors: In uitro autoradiography with ['251]heat in normal rats and rats treated with alpha-difluoromethylornithine, a specific, irreversible inhibitor of ornithine decarboxylase. Neuroscience, 15: 1195-1202. Jones, L.S., Gauger, L.L. and Davis, J.N. (1985b) Anatomy of brain alpha,-adrenergic receptors: In vitro autoradiography with ['251]-heat. J. Comp. Neurol., 231: 190-208. Jonsson, G. and Kasamatsu, T. (1983) Maturation of monoamine neurotransmitters and receptors in cat occipital cortex during postnatal critical period. Exp. Brain Res., 50: 449-458. Kasamatsu, T. and Shirokawa, T. (1988) Norepinephrine-dependent neuronal plasticity in kitten visual cortex. In C.D. Woody, D.L. Alkon and J.L. McGaugh (Eds.), Cellular Mechanisms of Conditioning and Behavioral Plasticity, Plenum Publishing Corporation, New York, pp. 437-444. Kasamatsu, T., Ohashi, T. and Imamura, K. (1989) Integration of adrenergic and cholinergic regulation in ocular dominance plasticity. Biomed. Res. 10, Suppl. 2: 43-53. Kawai, M., Sakaue, T., Watanabe, C., Nomura, Y. and Segawa, T. (1984) Specific binding of [3H]WB 4101, [3H]clonidine and [ 3H]dihydroalprenolol in cerebral cortical membranes in developing, adult and old rats. Jpn. J. Pharmacol., 36: 265-267. Kimura, F. and Nakamura, S. (1985) Locus coeruleus neurons in the neonatal rat: Electrical activity and responses to sensory stimulation. Deu. Brain Res., 23: 301-305. Kimura, F. and Nakamura, S. (1987) Postnatal development of a-adrenoreceptor-mediated autoinhibition in the locus coeruleus. Deu. Brain Res., 35: 21-26. Kuhar, M.J. and Unnerstall, J.R. (1985) Quantitative receptor mapping by autoradiography: Some current technical problems. TINS., 8: 49-53. Lanfumey, L. and Adrien, J. (1988a) Age-dependent P-adrenergic response of neuronal discharge in the raphe dorsalis: Evidence of a transient regulation. Int. J. Dev. Neurosci., 6: 301-307. Lanfumey, L. and Adrien, J. (1988b) Adaptive changes of /3-adrenergic receptors after neonatal locus coeruleus lesion: Regulation of serotoninergic unit activity. Synapse, 2: 644-649. Lau, C., Cameron, A., Antolick, L. and Slotkin, T.A. (1990) Trophic control of the ornithine decarboxylase/polyamine system in neonatal rat brain regions: Lesions caused by 6-hydroxydopamine produce effects selective for cerebellum. Deu. Brain Res., 52: 167-173. Lauder, J.M. and Bloom, F.E. (1974) Ontogeny of monoamine neurons in the locus coeruleus, raphe nuclei and substan-
tia nigra of the rat. I. Cell differentiation. J. Comp. Neurol., 155: 469-482. Lauder, J.M. and Bloom, F.E. (1975) Ontogeny of monoamine neurons in the locus coeruleus, raphe nuclei and substantia nigra of the rat. 11. Synaptogenesis. J. Comp. Neurol., 163: 251-264. Levitt, P. and Moore, R.Y. (1979) Catecholamine innervation of the developing neocortex in the rat. In E. Usdin, 1.J. Kopin and J. Barchas (Eds.), Catecholamines: Basic and Clinical Frontiers, VoL. I , Pergamon, Elmsford, New York, pp. 809-811. Lortn, I., Bjorklund, A. and Lindvall, 0. (1976) The catecholamine systems in the developing rat brain: Improved visualization by a modified glyoxylic acid-formaldehyde method. Brain Res., 117: 313-318. Lorton, D., Bartolome, J., Slotkin, T.A. and Davis, J.N. (1988) Development of brain beta-adrenergic receptors after neonatal 6-hydroxydopamine treatment. Brain Res. Bull., 21: 591-600. Loy, R. and Moore, R.Y. (1979) Ontogeny of the noradrenergic innervation of the rat hippocampal formation. Anat. Embryol. (Berlin), 157: 243 253. Maeda, T. and Dresse, A. (1969) Recherches sur l'development du locus coeruleus. Etude des catecholamines au microscope de fluorescence. Acta Neurol. Belg., 69: 5-10. Marshall, K.C. and Tsai, W.H. (1988) Noradrenaline induces short and long duration potentiation of glutamate excitations of cultured Purkinje neurons. Can. J. Physiol. Pharmacol., 66: 848-853. Marshall, K.C., Pun, R.Y.K., Hendelman, W.J. and Nelson, P.G. (1981) A coeruleo-spinal system in culture. Science, 213: 355-357. Maurin, Y., Lesaux, F., Graillot, C. and Baumann, N. (1985) Altered postnatal ontogeny of a,-and a,-adrenoreceptor binding sites in the brain of a convulsive mutant mouse (quaking). Deu. Brain Res., 22: 229-235. Morris, G. and Slotkin, T.A. (1985) Beta-2 adrenergic control of ornithine decarboxylase activity in brain regions of the developing rat. J. Pharmacol. Exp. Ther., 233: 141-147. Morris, M.J., Dausse, J-P., Devynck, M-A. and Meyer, P. (1980). Ontogeny of a 1 and a,-adrenoreceptors in rat brain. Brain Rex, 190: 268-271. Nakamura, S., Kimura, F. and Sakaguchi, T. (1987) Postnatal development of electrical activity in the locus coeruleus. J. Neurophysiol., 58: 510-524. Nakamura, S., Sakaguchi, T., Kimura, F. and Aoki, F. (1988) The role of alpha ,-adrenoreceptor-mediated collateral excitation in the regulation of the electrical activity of locus coeruleus neurons. Neuroscience, 27: 921-929. Nakamura, A. and Sakaguchi, T. (1990) Development and plasticity of the locus coeruleus: A review of recent physiological and pharmacological experimentation. Prog. Neurobiol., 34: 505-526. Nomura, Y., Kawai, M., Mita, K. and Segawa, T. (1984) Developmental changes in cerebral cortical [2H]clonidine binding in rats: Influences of guanine nucleotide and cations. J. Neurochem., 42: 1240-1245. Olson, L. and Seiger, A. (1972) Early prenatal ontogeny of ~
185 central monoamine neurons in the rat: Fluorescence histochemical observations. Z. Anat. Entwickl-Gesch., 137: 301-316. Pittman, R.N., Minneman, K.P. and Molinoff, P.B. (1980) Ontogeny of P,and P,adrenergic receptors in rat cerebellum and cerebral cortex. Brain Res., 188: 357-368. Sakaguchi, T. and Nakamura, S. (1987) Some in uiuo electrophysiological properties of locus coeruleus neurons in fetal rats. Exp. Brain Res., 68: 122-130. Schoepp, D.D. and Rutledge, C.O. (1985) Comparison of postnatal changes in alpha,-adrenoreceptor binding and adrenergic stimulation of phosphoinositide hydrolysis in rat cerebral cortex. Biochern. Pharrnacol., 34: 2705-271 1. Shaw, C., Wilkinson, M., Cynader, M., Needler, M.C., Aoki, C. and Hall, S.E. (1986) The laminar distributions and postnatal development of neurotransmitter and neuromodulator receptors in cat visual cortex. Brain Res. Bull., 16: 661-671. Shirokawa, T. and Kasamatsu, T. (1986) Concentration-dependent suppression by P-adrenergic antagonists of the shift in ocular dominance following monocular deprivation in kitten visual cortex. Neuroscience, 18: 1035-1046. Singer, E. (1988) Transmitter release from brain slices elicited by single pulses: A powerful method to study presynaptic mechanisms. Trends Pharrnacol. Sci., 9: 274-276.
Sugden, D. and Klein D.C. (1985) Development of the rat pineal a,-adrenoreceptor. Brain Res., 325: 345-348. Ventura, A.L.M., Klein, W.L. and De Mello, F.G. (1984) Differential ontogenesis of D , and D, dopaminergic receptors in the chick embryo retina. Deu. Brain Res., 12: 217-223. Weiss, B., Clark, M.B. and Greenberg, L.H. (1984) Modulation of catecholaminergic receptors during development and aging. In L. Abel (Ed.), Handbook of Neurochernistry, Receptors in the Nervous System, 2nd Ed. Vol. 6 , Plenum Press, New York, pp. 595-627. Williams, J.T. and Marshall, K.C. (1987) Membrane properties and adrenergic responses in locus coeruleus neurons of young rats. J. Neurosci., 7: 3687-3694. Williams, J.T., Henderson, G. and North, R.A. (1985) Characterization of a,-adrenoreceptors which increase potassium conductance in rat locus coeruleus neurons. Neuroscience, 14: 95-101. Yamamoto, T., Ishikawa, M. and Tanaka, C. (1977) Catecholaminergic terminals in the developing and adult rat cerebellum. Brain Rex, 132: 355-361. Young, W.S., 111 and Kuhar, M.J. (1980) Noradrenergic cr,-and a,-receptors: Light microscopic autoradiographic localization. Proc. Natl. Acad. Sci. USA. 77: 1696-1700.
This Page Intentionally Left Blank
C.D. Rarncs and 0. Pompeiano (Ed>.) Pro,ve.i.i i f # B u m Resuurch, Vol. 88 63 1 9 9 1 Elsevier Science Puhliahers B.V.
187
CHAPTER 13
and GABA, receptors and the ionic mechanisms mediating their effects on locus coeruleus neurons
G A B A A
S.A. Shefner and S.S. OsmanoviC 1,2 I
Department of Physiology and Biophysics, University of Illinois, College of Medicine, Chicago, IL, U S A . and Department of Patho-Physiology, Medical Faculty, Belgrade, Yugoslavia
'
Anatomical, neurochemical, and electrophysiological studies have provided evidence that y-aminobutyric acid (GABA) is an important inhibitory neurotransmitter in the locus coeruleus (LC) nucleus. We have used intracellular recording to study the actions of GABA on putative noradrenergic neurons of the rat LC, in a brain slice preparation. GABA application in the bath, or more locally by micropressure ejection inhibited spontaneous firing and increased the conductance of LC neurons. In addition, GABA could hyperpolarize or depolarize LC neurons; the size of these responses depended on the C1- gradient across t h e membrane. GABA responses were antagonized by bicuculline. These data indicate that the actions of GABA on LC neurons are primarily mediated by activation of GABA, receptors which increases the CI- conductance. When GABA is applied to LC neurons after blockade of GABA, receptors with bicucullline, a resid-
ual action mediated by GABA, type receptors can be seen. Similar responses can be obtained with the GABA ,-selective agonist baclofen. GABA, activation inhibits spontaneous firing and causes membrane hyperpolarization due to an increase in K + conductance. Single-electrode voltage clamp experiments were used to study the voltage dependency of GABA responses in LC neurons. GABA-induced current showed outward rectification. The conductance increase caused by a given amount of GABA decreased with membrane hyperpolarization. The time constant of decay of the GABA current also decreased with membrane hyperpolarization. Due to the voltage dependency of GABA responses, GABA exerts a stronger inhibitory effect on LC neurons at depolarized potentials than at hyperpolarized potentials, which could serve as a negative feedback mechanism to control excitability of these neurons.
Key words: locus coeruleus, y-aminobutyric acid, GABA, receptor, GABA, receptor, baclofen, intracellular recording
Introduction
Evidence that GABA is a neurotransmitter in the locus coeruleus nucleus Several lines of evidence indicate that yaminobutyric acid (GABA) is an important inhibitory transmitter regulating the activity of locus coeruleus (LC) neurons. Autoradiographic
studies have demonstrated GABA receptors in the LC (Palacios et al., 1981) and binding studies have shown both GABA, and GABA, receptors in the terminal fields of LC neurons, specifically on noradrenergic terminals in cortex and hippocampus (Suzdak and Gianutsos, 1985). Neurochemical studies indicate that almost one-half of the terminals in the LC can take up
188
radiolabeled GABA (Iversen and Schon, 1973) and that the enzyme which synthesizes GABA, glutamic acid decarboxylase (GAD), is present in the LC (Belin et al., 1979; Perez de la Mora et al., 1981). GAD-positive terminals have been found in the LC, closely juxtaposed to norepinephrine-containing cell bodies and dendrites, which suggests functional contacts between GABAergic and noradrenergic neurons (BCrod et al., 1984).
Functional interaction of G A B A and noradrenergic neurons Immunocytochemical studies indicate that there are no GABAergic cell bodies in the rat LC (Bkrod et al., 1984). Recently, a major source of GABAergic input to the rat LC has been shown to be the nucleus prepositus hypoglossi, which is located in the dorsomedial rostra1 medulla (Ennis and Aston-Jones, 1989). Stimulation of the nucleus prepositus hypoglossi in situ in the rat evokes GABAergic inhibition of LC neurons which can be blocked by the GABA, receptor antagonists bicuculline and picrotoxin (Ennis and Aston-Jones, 1989). Furthermore, Cherubini et al. (1988) have shown that there is a GABAergic component of synaptic potentials, evoked in LC neurons by focal stimulation in brain slices, which can be blocked by GABA, antagonists. These studies indicate that release of endogenous GABA can inhibit LC neurons by action at GABA, receptors. Iontophoretic application of both GABA itself and the GABA,-selective agonist baclofen, have been shown to inhibit singleunit firing of rat LC neurons recorded in uivo (Cedarbaum and Aghajanian, 1977; Guyenet and Aghajanian, 1979). Direct postsynaptic actions studied by application of G A B A and baclofen to LC neurons recorded intracellularly in brain slices We have studied GABA responses in presumed noradrenergic neurons of the LC nucleus in totally submerged rat brain slices to determine the receptor type and ionic mechanisms mediat-
ing GABA effects on LC neurons. Brain slices (300 p m in thickness) were cut transversely through a caudal portion of the LC where there is a high density of noradrenergic neurons in the rat. The recording chamber was of small volume (0.3 ml) and the flow rate was 2.2 mi/min. These parameters, coupled with the low deadspace in the system, permitted steady-state concentrations of drug to be reached in the bath in about 2 min or less. GABA, baclofen, and other drugs were dissolved in artificial cerebrospinal fluid and applied in known concentrations in the bath or more locally by micropressure ejection. The de-
J10 5 s
mV
,500 ms
I'
Fig. 1. Pressure ejection of y-aminobutyric acid (GABA) causes inhibition of spontaneous firing, hyperpolarization and reduced input resistance in a locus coeruleus (LC) neuron. The pressure pipette was filled with 10 mM GABA and positioned close to the surface of the slice near the recording electrode. A pressure of 10 psi was used to eject GABA for different durations. Constant current hyperpolarizing pulses (300 msec duration) were used to monitor input resistance. Note that larger reductions in resistance and longer inhibition of firing resulted as the duration of the pressure application was increased from 100 msec to 1 sec. Resting membrane potential: - 58 mV.
189
tails of the preparation of such slices and the intracellular recording techniques used have been reported previously (Osmanovii and Shefner, 1990).
Results
GABA effects studied with current clamp recording LC neurons recorded in slices of the adult rat brain fire action potentials spontaneously, at a rate of about 0.2-2 Hz. Application of GABA by pressure ejection inhibits the spontaneous firing (Fig. 1). The duration of inhibition depends on the duration of the pressure pulse. As shown in Figure 1, GABA application also hyperpolarized the neuron and reduced the input resistance. These effects were concentration-dependent, as seen when the concentration of GABA applied to the cell was raised by increasing the duration of the pressure pulse. Figure 2 shows responses to bath-application of GABA at two different concentrations. Note that the extent of inhibition of firing is concentration-dependent. A small reduction in input resistance can be seen with 300 p M GABA and a larger decrease in resistance with ~~
600 pM, which also caused a small hyperpolarization. We have previously shown that the EC50 for the GABA-induced conductance increase for LC neurons determined in the slice preparation is 1.7 mM (OsmanoviC and Shefner, 1990). It should be noted that although known concentrations of GABA are applied in the bath, uptake of GABA was not inhibited in these experiments, so the actual GABA concentrations at the receptor sites is not known. GABA uptake in the slices is expected to be extensive, since the EC50 value for GABA in our experiments is about two orders of magnitude higher than published values for cultured or dissociated neurons. As we have previously reported, GABA-induced inhibition of spontaneous firing, changes in membrane potential and input resistance are direct postsynaptic actions on LC neurons, since these responses persist in low calcium (0.25 mM) high magnesium (10 mM) media which blocks calcium-dependent transmitter release (see OsmanoviC and Shefner, 1990). The effect of GABA itself is mediated predominantly by activation of GABA, type receptors. As shown in Figure 2, the effects of bath-applied GABA on LC neurons
I l l l l l l l l l l l l l I I IIIIIIII II I I IIIl111IIIIll I I I [ 1 1 1 1 II I I I I I I I I I II1111I I I I I I O I I l l IIII IIII I I l l 1 1 1 1 1 1 I IIIIIII11I1IIIII[ Fig. 2. Responses of a typical LC neuron to bath-application of GABA and antagonism by bicuculline methiodide (BMI). Voltage records on the left show the effect of bath-application of two different GABA concentrations. These GABA concentrations were retested after 12 min of superfusion with BMI. Note complete block of the effects of 300 p M GABA and antagonism of the response to 600 p M GABA. Calibration bars apply to all records. Resting membrane potential: -58 mV. Recording electrode filled with 2 M KCI.
190
are antagonized by the GABA, receptor antagonist bicuculline methiodide (BMI). The extent of antagonism depends on the respective amounts of GABA and BMI given, as expected for a competitive antagonist. Figure 2 shows that 100 p M BMI completely blocks the effects of 300 p M GABA, and while it also antagonizes the effect of 600 p M GABA, a small GABA-induced reduction in firing rate still remains. We have found that while GABA application always increases the membrane conductance, the polarity of the resultant voltage change can be depolarization, hyperpolarization or no change in membrane potential, depending on the recording conditions (OsmanoviC and Shefner, 1990). For example, when the recording electrode was filled with 2 M KCl, which causes leakage of Cl- into the cell, GABA usually depolarized LC neurons; such responses showed a mean reversal potential of about -48 mV. By contrast, when recording electrodes were filled with 2-4 M potassium acetate, GABA usually hyperpolarized LC neurons, with a mean reversal potential of about - 71 mV. The GABA equilibrium potential (E,,,,) also shifted when the external C1- concentration in the media was altered; the shift was about -52 mV per tenfold change in C1- concentration. This is close to the shift predicted by the Nernst equation and indicates that the predominant action of GABA on LC neurons is to increase C1conductance.
Baclofen (GABA,) effects studied with current clamp recording When the effects of the GABA,-specific agonist baclofen were compared to the effects of GABA, it was noted that while GABA depolarized LC neurons recorded with KC1-filled electrodes, baclofen hyperpolarized them (OsmanoviC and Shefner, 1988). Baclofen-induced hyperpolarizations of LC neurons were small, with average maximum amplitudes of 6-8 mV, but showed clear concentration dependence. The concentration response curve for baclofen hyperpolariza-
tion of LC neurons was sigmoidal, with an EC50 of about 2.0 p M . This value is similar to the EC50 of baclofen for other cell types studied in slice preparations (Lacey et al., 1988). In addition to hyperpolarization, baclofen caused a prolonged inhibition of spontaneous firing and reduced the input resistance. The maximum conductance increase induced by baclofen, however, was much smaller than GABA-induced conductance changes. Baclofen responses in LC neurons showed stereospecificity for ( - ) baclofen and were resistant to the GABA, antagonist BMI. Furthermore, baclofen responses persisted in low-calcium/high-magnesiummedia which blocks synaptic transmission. Taken together, these findings indicate that baclofen exerts a direct postsynaptic action at GABA, receptors on LC neurons. Figure 3 shows voltage-current (V/I) curves determined in the same LC neuron under several conditions. The control V/I relation (filled circles) intersects the y axis at -60 mV indicating the resting membrane potential. Bath-application of GABA (open circles) caused membrane depolarization and reduced the slope resistance. The GABA and control curves intersect at EGABA, which was about -55 mV, close to the value of E,, when KC1-filled recording electrodes are used. By contrast, when GABA was applied in the presence of the GABA, antagonist BMI, the membrane was hyperpolarized and the decrease in slope resistance was much smaller. The equilibrium potential for the GABA response in the presence of BMI (filled diamonds) was 40 mV more negative than EGABA,i.e., had shifted toward the K+ equilibrium potential (EK).Application of baclofen to the same cell (open diamcnds) caused a response similar to the GABA plus BMI condition, but in this case the hyperpolarization and reduction in slope resistance were slightly larger. The reversal potential of the baclofen response (EBJ was slightly more negative, very close to the expected value of E, in this preparation. We have previously reported that mean values for EBacand E, determined in the same LC neurons were both about -115 mV (OsmanoviC
191
I (nA) 0
/ /
Control
GABA GABA+Bic \
Bac
MP (mV)
Fig. 3. Comparison of equilibrium potentials for responses to GABA alone (E,,), GABA in the presence of the GABA, and to the GABA agoantagonist bicuculline (E,,,,, nist baclofen (EBac).Voltage-current (V/I) curves were constructed by recording the voltage responses elicited by passing current pulses (500 msec duration) of varying amplitudes in the control condition (filled circles), in the presence of GABA (200 b M , open circles), GABA (200 W M ) and bicuculline methiodide (80 K M ) given together (filled diamonds), and baclofcn (10 p M , open diamonds). Note that E,,,, was about -55 mV, close to the value of E,, when 2 M KCI recording electrodes are used. By contrast, E,ABA+B,cand EBacare over 40 mV more negative, close to E,.
and Shefner, 1988). Furthermore, when the external K + concentration was raised, EBacshifted by 61 mV per tenfold change in K+ concentration, the same value predicted by the Nernst equation for a pure increase in K + conductance. In conclusion, our intracellular recordings from somata of LC neurons indicate the presence of functional GABA, and GABA, receptors. When GABA itself is applied, GABA,-mediated effects predominate. Responses mediated by GABA, receptors can be seen, however, when GABA is applied after blockade of GABA, receptors by bicuculline, or when GABA, receptors are selectively activated by baclofen.
Voltage clump studies: voltage dependency of GABA responses In voltage clamp experiments, GABA responses in LC neurons were recorded with a single-electrode voltage clamp circuit (Axoclamp 2A, Axon Instruments). Figure 4 shows an experiment in which the membrane potential of an LC neuron was held at -60 mV and GABA (400 p M ) was applied in the bath. GABA caused an outward current and increase in conductance which reversed with washout. Such responses indicate that outward currents underlie GABA-induced hyperpolarizations seen with potassium acetate-filled recording electrodes in current clamp studies. We have previously demonstrated that such GABA-induced current responses showed similar reversal potentials to GABA-induced voltage responses, both reversing close to the C1equilibrium potential (Ec,) (OsmanoviC and Shefner, 1990). Figure 5 shows a control current-voltage (I-V) curve (open circles) determined under voltage clamp conditions. The inset in Figure 5 shows that in the presence of GABA, the current response to a hyperpolarizing voltage pulse is not square, but “sags” and there is an outward tail current at the offset of the pulse. I-V curves (in the presence of GABA) were plotted for current values at the peak of the current response (filled circles) and at steady-state after the sagging of the response (filled diamonds). Note that the I/V curves for peak and steady-state current show
I
GABA 400 IIM
m;mmmmmmmmmmrm’i-n-
t 1 o.l
rrmiiii~mnTiiii iimmmmminnmi ri 1 iiiI ii I I i I I I I rrrirmimmmmnlo,
Fig. 4. GABA-induced outward current and conductance increase measured under voltage-clamp conditions. Upper trace: current, lower trace voltage. Note that bath-application of GABA (400 p M ) caused an outward current in this neuron recorded with a potassium acetate-filled electrode. The outward current was accompanied by an increase in input conductance and these effects reversed with washout of GABA. Conductance was monitored by observing the current deflections elicited by constant amplitude hyperpolarizing voltage steps (300 msec duration). Holding potential: - 60 mV.
192
different reversal potentials, which suggests that E,, is shifting during the hyperpolarizing voltage pulse. We have proposed that these slow relaxations (the sag and outward tail current) are due to redistribution of chloride ions through GABA-activated channels (OsmanoviC and Shefner, 1990). A similar phenomenon has been reported for GABA responses in neurons in crayfish stretch receptors (Adams et al., 1981) and bull-frog dorsal root ganglia (Akaike et al., 1987). As we have previously reported, the membranes of LC neurons show anomalous or inward rectification (OsmanoviC and Shefner, 1987). This can be seen in Figure 6A, which shows current
dl ‘
-110
-7
-90
-5
/
-1
0.4
40.6
0-
5 8
n4
“ ‘
//I
Fig. 6. Resting conductance and GABA-induced conductance in LC neurons rectify in opposite directions. LC neuron impaled with a KCI electrode was clamped at different membrane potentials. Voltage commands (12 mV, 300 msec) were applied at each holding potential to measure membrane conductance in the absence (A) and presence of 500 p M GABA (B). Note that in the control condition, the input conductance is larger at hyperpolarized membrane potentials (inward rectification) while in the presence of GABA, the input conductance becomes larger with membrane depolarization (outward rectification). GABA application caused a small inward current (0.06 nA) at the resting membrane potential (-57 mV).
-:ojo 0.2
‘peak
J0.7
Irn (nA)
Fig. 5. Membrane current responses elicited by voltage steps show sagging in the presence of GABA due to redistribution of C1-. Current-voltage (I/V) curves were determined by passing different amplitude voltage steps from the holding potential and measuring the amplitude of currents elicited by these steps in the absence (open circles) and presence of 600 p M GABA applied by superfusion (filled symbols). Note that GABA caused an inward current and increased the slope conductance. Inset shows current response to a 14 mV hyperpolarizing voltage pulse applied from the holding potential of - 63 mV. GABA equilibrium potentials were estimated from the intersection of the control I/V curve (open circles) and I/V curves of the peak current response to the voltage steps (filled circles) and of the “steady-state” current response, after sagging of the current during the voltage step (filled diamonds). The equilibrium potential at the beginning of the voltage steps (peak amplitude of the current responses) was more negative than at the end of the voltage steps (after sagging of the current responses). Electrode filled with 2 M KCI.
responses to constant amplitude hyperpolarizing pulses applied at different holding potentials. Note that the same voltage pulse causes a larger current response at more negative holding potentials. This increased conductance at hyperpolarized potentials is blockable by external cesium and barium ions, internal injection of acetate, and is of the classical potassium-dependent type well characterized in skeletal muscle (OsmanoviC and Shefner, 1987). In contrast to the behavior of the membrane under control conditions, when GABA is present, current responses to constant amplitude voltage pulses are larger at depolarized holding potentials, i.e., the membrane becomes outwardly rectifying (Fig. 6B). The opposite types of rectification in the absence and presence of GABA also can be seen on the I-V curves in Figure 5. Note that the control I-V relation is not a straight line but a curve, with higher slope conductance at more hyperpolarized potentials (inward rectification), while the I-V curves in GABA show decreased slope conductance in the hyperpolarized range (outward rectification). The net membrane current becomes outwardly rectifying in the presence of GABA,
193
because the GABA-induced conductance is voltage-dependent, as can be seen in Figure 7. In this voltage clamp experiment, a constant amount of GABA was applied by micropressure ejection and the resultant GABA-induced currents measured at different holding potentials (see Fig 7A). Application of GABA (at arrows) caused outward or inward currents depending on the holding potential. Conductance was monitored with constant amplitude hyperpolarizing voltage pulses and GABA application caused an increase in membrane conductance. It is clear from Figure 7A, that the GABA-induced conductance increase was larger when the membrane potential was held at - 48 mV than at more hyperpolarized
A
I
I
I
I.,.
-60
-80
,111.1
-100
Fig. 8. Time constant of decay of GABA currents depends on membrane potential. GABA was applied by pressure ejection (10 mM, 10 psi, 500 msec) and GABA-induced current measured while the membrane potential was held at different levels. The time course of decay of the GABA current could be fit by a single exponential function and time constants of decay were calculated for responses obtained at different holding potentials. Note that time constant was an approximately linear function of membrane potential; GABA currents decayed more slowly (had larger time constants) at depolarized holding potentials, than at hyperpolarized potentials.
I
6
IGABA(~A)
Fig. 7. GABA-induced currents show outward rectification. A, GABA was applied by pressure ejection (10 mM, 10 psi, 500 msec) at arrows when the membrane potential was held at different levels. This LC neuron was recorded with a 2 M KCI-filled electrode and the polarity of the current response reversed at about -53 mV. Conductance was monitored by observing current deflections in response to constant hyperpolarizing voltage steps of 15 mV and 300 msec duration. Note that the GABA-induced conductance is greater at -48 mV than at -87 mV. This voltage dependence of the GABA-induced conductance can also be seen in more detail on the graph in part B. Peak amplitudes of GABA-induced currents as shown in A, are plotted versus holding potential. The curve is not linear but shows outward rectification, ie., slope conductance is reduced at hyperpolarizing potentials.
holding potentials (e.g., -87 mV). Figure 7B shows a graph of the peak amplitude of GABAinduced currents as a function of holding potential for the same LC neuron, as determined from records as shown in panel A. Outward rectification of the GABA current can be seen as a decrease in slope conductance at hyperpolarizing holding potentials. From experiments like those shown in Figure 7A, we noted that the time course of decay of GABA current evoked by pressure ejection also appeared to depend on membrane potential. Decay time constants were calculated for GABA responses evoked at different holding potentials, and these time constants were plotted as a function of holding potential, as shown in Figure 8. At more depolarized holding potentials, GABA currents had longer time constants, ie., decayed more slowly. The voltage dependence of decay time constants which we found in LC neurons is qualitatively similar to voltage-dependence of GAEiA-mediated synaptic currents reported for other types of neurons (Adams et al., 1981; Collingridge et al., 1984; Barker and Harrison, 1988).
194
Conclusions Our in uitro intracellular studies of GABA responses in rat LC neurons indicate that GABA inhibits their spontaneous firing and causes a large conductance increase, which is associated with membrane hyperpolarization, depolarization, or no change in membrane potential. GABA responses are antagonized by the GABA, antagonist BMI and reverse at Ec,, indicating that GABA acts predominantly on GABA, receptors to increase C1- conductance. In the presence of BMI, the reversal potential of the residual GABA response shifts toward E,. BMI-resistant effects can be mimicked by the GABA, agonist baclofen, which inhibits spontaneous firing and hyperpolarizes LC neurons by increasing the K+ conductance. The maximum conductance increase due to GABA, activation is much smaller than that activated by GABA, receptors. Both GABA, and GABA, effects on LC neurons are concentration-dependent and result from a direct action, since they persist in low-calcium/highmagnesium media which blocks synaptic transmission. The GABA-induced conductance change is voltage-dependent, decreasing with membrane hyperpolarization. GABA-induced current shows outward rectification. The time constant of decay of GABA current decreases with membrane hyperpolarization, suggesting that membrane potential also influences the kinetics of GABA responses. Due to the voltage dependence of GABA responses, GABA exerts a stronger inhibitory effect on LC neurons at depolarized than at hyperpolarized membrane potentials. This could serve as a negative feedback mechanism to control excitability of these neurons. Acknowledgements This work was supported by US PHS grant number AA05846. We wish to thank Dr. Mark S. Brodie for his helpful comments on the manuscript.
References Adams, P.R., Constanti, A. and Banks, F.W. (1981) Voltage clamp analysis of inhibitory synaptic action in crayfish stretch receptor neurones. Fed. Proc., 40: 2637-2641. Akaike, N., Inomata, N. and Tokutomi, N. (1987) Contribution of chloride shifts to the fade of y-aminobutyric acidgated currents in frog dorsal root ganglion cells. J. Physiol. (London), 391: 219-234. Barker, J.L. and Harrison, N.L. (1988) Outward rectification of inhibitory postsynaptic currents in cultured rat hippocampal neurons. J. Physiol. (London), 403: 41-55. Belin, M.F., Aguera, M., Tappaz, M., McRae-Degueurce, A,, Bobillier, P. and Pujol, J.F. (1979) GABA-accumulating neurons in the nucleus raphe dorsalis and periaqueductal grey in the rat: A biochemical and radioautographic study. Bruin Res., 170: 279-297. BCrod, A., Chat, M., Paut, L. and Tappaz, M. (1984) Catecholaminergic and GABAergic anatomical relationship in the rat substantia nigra, locus coeruleus, and hypothalamic median eminence: Immunocytochemical visualization of biosynthetic enzymes on serial semithin plastic-embedded sections. J. Histochem. Cytochem., 32: 1331-1338. Cedarbaum, J.M. and Aghajanian, G.K. (1977) Catecholamine receptors on locus coeruleus neurons: Pharmacological characterization. Eur. J. Pharmacol., 44: 375-385. Cherubini, E., North, R.A. and Williams, J.T. (1988) Synaptic potentials in rat locus coeruleus neurones. J. Physiol. (London), 406: 431-442. Collingridge, G.L., Gage, P.W. and Robertson, B. (1984) Inhibitory post-synaptic currents in rat hippocampal CAI neurones. J. Physiol. (London), 356: 551-564. Ennis, M. and Aston-Jones, G. (1989) GABA-mediated inhibition of locus coeruleus from the dorsomedial rostra1 medulla. J. Neurosci., 9: 2973-2981. Guyenet, P.G. and Aghajanian, G.K. (1979) ACh, substance P and Met-enkephalin in the locus coeruleus: Pharmacological evidence for independent sites of action. Eur. J. Phurmacol., 53: 319-328. Iversen, L.L. and Schon, F.E. (1973). The use of autoradiographic techniques for the identification and mapping of transmitter-specific neurones in CNS. In A.J. Mandell (Ed.), New Concepts in Neurotransmitter Regulation, Plenum Press, New York, pp. 153-193. Lacey, M.G., Mercuri, N.B. and North, R.A. (1988) On the potassium conductance increase activated by GABA, and dopamine D, receptors in rat substantia nigra neurones. J. Physiol. (London), 401: 437-453. Osmanovid, S.S. and Shefner, S.A. (1987) Anomalous rectification in rat locus coeruleus neurons. Brain Re . , 417: 161- 166. OsmanoviC, S.S. and Shefner, S.A. (1988) Baclofen increases the potassium conductance of rat locus coeruleus neurons recorded in brain slices. Brain Res., 438: 124-136. Osmanovid, S.S. and Shefner, S.A. (1990) y-aminobutyric acid responses in rat locus coeruleus neurons in vitro: A current-clamp and voltage-clamp study. J. Physiol. (London), 421: 151-170.
195 Palacios, J.M., Wamsley, J.K. and Kuhar, M.J. (1981) High affinity GABA receptors - autoradiographic localization. Bruin Rex, 222: 285-307. Perez de la Mora, M., Possani, L.D., Tapia, R., Teran, L., Palacios, R., Fuxe, K., Hokfelt, T. and Ljungdahl, A. (198 I) Demonstration of central y-aminobutyrate-contain-
ing nerve terminals by means of antibodies against glutamate decarboxylase. Neuroscience, 6: 875-895. Suzdak, P.D. and Gianutsos, G. (1985) GABA-noradrenergic interaction: Evidence for differential sites of action for GABA-A and GABA-B receptors. J. Neural Trunsm. 64: 163- 172.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Ed$.) Progress in Brain Research, Vol. 88
197
0 1991 Elsevier Science Publishers B.V.
CHAPTER 14
Mechanisms of opioid actions on neurons of the locus coeruleus M.J. Christie Department of Pharmacology, University of Sydney, N.S. W., Australia
The locus coeruleus (LC) has provided a useful model for pioneering studies of the mechanisms underlying the acute and chronic actions of opioid drugs. Acutely, opioids inhibit the electrical activity of single neurons in the rat and guinea pig LC. Inhibition is due to a membrane hyperpolarisation. In these cells, opioids act on y-receptors to increase the opening of inwardly rectifying potassium channels, thus leading to hyperpolarisation. The y-receptors are coupled to potassium channels via G-proteins which are sensitive to inactivation by
pertussis toxin. This coupling process is quite direct, in that it does not involve freely diffusible intracellular second messengers. Agonists specific for other receptors, such as aZ-and somatostatin-receptors, are capable of opening the same population of potassium channels on LC neurons. Following chronic treatment of animals with morphine, a specific deficit develops in the ability of p-receptors to open potassium channels, producing reduced sensitivity of LC neurons to inhibition by opioids.
Key words: locus coeruleus, receptor opioid, potassium channel, tolerance, G-protein, chronic drug
Introduction Opioids directly inhibit the electrical activity of neurons in many regions of the nervous system, including the locus coeruleus (LC) (see Duggan and North, 1983, for review). This inhibition arises from changes in ionic conductances in the somatic membrane or nerve terminal, leading ultimately to a reduced firing of action potentials and depressed release of neurotransmitter. The ionic mechanisms underlying opioid actions were first elucidated in the LC (Williams et al., 19821, and the molecular mechanisms involved have been more thoroughly studied in these neurons than in other regions of the CNS. It is now clear that
opioids inhibit LC neurons by activating p-receptors, which couple to the opening of potassium channels via G-proteins (North et aL, 1987). Within the LC neuron, the coupling process between receptor and potassium channel is localised to a membrane area of less than a few square microns, and does not involve freely diffusible second messengers (Miyake et al., 1989). No action of endogenously released opioids has yet been clearly identified in the LC, so the physiological role in the whole animal of this action on the LC soma is still unclear. Much of the impetus to understand the acute actions of opioids on CNS neurons has arisen from the problems of tolerance and physical de-
198
pendence. The LC has been a focus of much of this work, partly because of the similarity between the behavioural effects of electrical stimulation of the LC and the opiate withdrawal syndrome. While the mechanisms underlying dependence remain elusive, toIerance involves a specific deficit in the ability of opioid receptors to open potassium channels (Christie et al., 1987). Acute actions of opioids Opioids directly inhibit LC neurons Opioids inhibit the firing of action potentials when applied to neurons of many regions of the CNS (see Duggan and North, 1983, for review). In most cases where opioids excite neurons, it has been shown to occur as a result of the disinhibition of inhibitory interneurons (e.g., Madison and Nicoll, 1987). The first in vivo studies demonstrated inhibition of sensory-evoked action potential firing following intravenous administration of morphine (Korf et at., 1974). Inhibition was specific in that it was reversed by the opioid antagonist, naloxone. Similar results were reported in cat LC (Strahlendorf et al., 1980). Inhibition of spontaneous firing was also shown in rat LC in viuo following intravenous or localised (iontophoretic) application of opioids (Bird and Kuhar, 1977; Guyenet and Aghajanian, 1977). This observation has been substantiated in uiuo and in ilitro for species in which LC neurons can be identified as a fairly uniform population of noradrenaline-containing cells (see below). Opioid receptor type on LC neurons Opioids are known to affect cells via activation of at least three distinct membrane receptors; p-, 6- and K-receptors (see Leslie, 1987, for review). With the advent of brain slice techniques to study CNS neurons in vitro, the opportunity arose to classify the opioid receptor(s1 responsible for inhibition in LC neurons. Quantitative pharmacological techniques have demonstrated that inhibition is due to activation of p-receptors in the rat
(Williams and North, 1984). In addition to the actions of non-selective agonists such as methionine-enkephalin (met-enkephalin) and D-Ala-DLeu-enkephalin (DADLE), LC neurons were directly inhibited by p-agonists such as morphine and D-Ala-methionine-enkephalin-glyol(DAGO). However, the 6-agonist, D-Pen-D-Pen-enkephalin (DPDPE) (North et al., 1987) and the K-agonist, U50488H (Williams and North, 1984) were ineffective in rat LC, even at very high concentrations, suggesting that 6- and K-receptors were not responsible for inhibition. Quantification of the equilibrium dissociation constant (K,) of an antagonist is the most reliable method to define receptor type. This is because, unlike agonist actions, it does not depend on efficiency of coupling between receptor and effector. When determined by Schild analysis, the K , of naloxone to antagonize the action of met-enkephalin was 2 nM (Williams and North, 1984). This result was very similar to the value obtained for p-receptors in biochemical assays and differed by 10 to 30-fold from the value expected €or 6-or K-receptors. Moreover, the &receptor antagonist, ICI174864, blocked the effects of met-enkephalin very poorly ( K , > 5 pM).Thus, it is clear that inhibition of LC neurons by opioids is due solely to the activation of p-receptors. p-Receptors act cia G-proteins Although direct information on the primary aminoacid structure of the p-receptor is still lacking, there is little doubt that it belongs to the family of receptors which exert their cellular effects via GTP-binding proteins (G-proteins, see Birnbaumer, 1990, for review). Evidence that p receptors act through G-proteins includes inhibition of adenylate-cyclase in various preparations including LC (Duman et al., 1988), co-purification of p-binding sites with pertussis toxin-sensitive G-proteins (Wong et al., 1989) and the ability of purified p-receptors to reconstitute with purified G, and Go (Ueda et al., 1988). In LC, treatment of rats with pertussis toxin, which inactivates &-subunits of G, and Go, prevents inhibi-
199
tion of LC neurons by opioids (Aghajanian and Wang, 1986).
p-Receptor distribution on LC neurons Opioid receptors are distributed over the plasma membrane and nerve terminal regions of LC neurons. Physiological studies with patchclamp electrodes (see below) suggest a fairly uniform distribution of receptors over the perikaryal membrane of isolated LC neurons, although this result could have been influenced by enzymatic and mechanical procedures used to dissociate the LC. Some studies have also suggested that pagonists decrease the excitability of LC axon terminals. p-agonists inhibit noradrenaline release from brain slices of LC projection areas (Werling et al., 1987) and increase the threshold for antidromic activation of LC fibers when infused into cerebral cortex in civo (Nakamura et al., 1982). However, because these studies could not be conducted in the absence of synaptic transmission, they did not rule out indirect effects of p-agonists on the LC terminal membrane. p-Receptors activate potassium channels Techniques for intracellular recording from brain slices have also facilitated the elucidation of ionic mechanisms underlying opioid inhibition. The first intracellular recordings from LC neurons demonstrated inhibition to be due to a membrane hyperpolarization in guinea-pig (Pepper and Henderson, 1980). The hyperpolarization was associated with a fall in input resistance, and persisted under conditions in which synaptic transmission was blocked, indicating that opioids act directly on the impaled neuron. Williams et al. (1982) demonstrated that the hyperpolarization was due to an increased conductance to potassium ions in rat LC neurons. This was due to opening of membrane potassium channels, shifting the membrane potential from rest (about - 55 mV) towards the potassium equilibrium potential (about - 110 mV>. The macroscopic properties of the potassium
conductance were similar to those of a class of potassium channels known as inward rectifiers, first characterised in invertebrate egg cells (for review see Hille, 1984). These channels are so named because they do not open until the membrane potential approaches the potassium equilibrium potential, thus allowing potassium currents to flow more readily in the inward than outward direction. The mid-point of the conductance activated by opioids in LC neurons (about - 60 mV) is close to action potential threshold (about -55 mV), so these channels are capable of hyperpolarizing the membrane (North and Williams, 1985; Williams et al., 1988). Opioid-activated potassium channels have been demonstrated directly in acutely dissociated rat LC neurons (Miyake et at., 1989). When patchclamp pipettes were sealed onto the cell surface, single potassium channels were observed when opioids were included in the pipette solution, but not when opioids were applied to the rest of the cell. The probability of channel opening was dependent on agonist concentration and was antagonized when naloxone was included in the pipette. The channels had a unit conductance of about 45 pS in isotonic potassium and displayed bursts of activity lasting several seconds to minutes, separated by similar closed periods. A thorough analysis of channel kinetics was precluded by these long, closed periods, but within the periods of activity there may be two closed states and one open state of the channel. The open state is prolonged when higher concentrations of opioids are tested. All agonist-filled pipettes of about the same size contained similar numbers of channels, suggesting a fairly uniform distribution over the perikaryal membrane. The number of opioidactivated channels per cell was estimated to be about 1,000, given the diameter of recording pipettes and the total surface area of the LC neuron. This result suggests that p-receptors must also be fairly uniformly distributed over the soma, because they are within the few square microns of membrane contained within the pipette.
200
Mechanism of coupling between opioid receptors and potassium channels The main inference drawn from the experiments described above was that the coupling between opioid receptor activation and potassium channel opening is localized to the patch of membrane enclosed by the patch pipette, therefore ruling out the involvement of freely diffusible second messengers. Although second messenger systems such as the adenylate cyclase to protein kinase-A cascade are inhibited by opioids in LC (see below), these are not involved in the activation of potassium channels. One report, which suggested that agents which activate protein kinase-A reversed the hyperpolarization produced by morphine (Andrade and Aghajanian, 1985) was subsequently shown to result from activation of a small inward current by these agents (Wang and Aghajanian, 1987). Activators of protein kinase-A failed to prevent opioid hyperpolarizations, or potassium currents in other experiments (North and Williams, 1985). Beyond the lack of involvement of freely diffusible second messengers, and an involvement of G-proteins sensitive to pertussis toxin, the nature of the coupling between preceptor and potassium channel is unknown. It is possible that activated G-proteins bind directly to the channel, or that biochemical events confined to a region in close proximity ( < 1 pm) to the receptor and channel mediate the effects. These issues also need to be resolved for a variety of receptors which couple to potassium channels in LC and other cells. Multiple receptors open a single population of potassium channels Agonists selective for several receptors have been demonstrated to hyperpolarize, increase a potassium conductance and/or open potassium channels on LC neurons. These include agonists selective for adrenergic az- (North and Williams, 1983, somatostatin- (Inoue et al., 1988) and GABA,- (Osmanovic and Shefner, 1988) receptors. Adenosine receptor agonists also hyperpo-
larize LC neurons, possibly by activating a potassium conductance (Shefner and Chiu, 1986). This list is not exhaustive, agonists selective for some other receptors inhibit LC (see Olpe and Steinmann, this volume), and others might not have been tested. For the only combination tested, different receptors have been shown to activate the same population of potassium channels as do p-receptors. This conclusion has been inferred from simultaneous application of different agonists to LC neurons, e.g., if sufficient p-agonist is applied to an LC neuron to fully activate the potassium conductance, a,-agonists cannot activate the conductance any further, even though an a,-agonist applied alone can activate the conductance to the same extent as the p-agonist (North and Williams, 1985). This observation held for any combinations of p - and a,-agonists. The location in the sequence of biochemical events where different receptors converge to open one population of potassium channels is not known. It might be that each receptor can activate a different population of G-proteins which can all open the same population of channels, or in the extreme alternative, convergence to a single population of G-proteins may occur. In either case, it is not surprising that different receptors can open the same population of potassium channels on LC neurons. The same is true for p- (Christie and North, 1988), and other receptors in a variety of cells (see North, 1989, for review). These receptors generally belong to a family of G-protein-coupled receptors which have also been shown to inhibit adenylate cyclase (see Birnbaumer, 1990, for review). Like in the LC, where studied in detail, this coupling has been shown to involve pertussis toxin-sensitive G-proteins, but not freely diffusable second messengers such as CAMP.
Do opioid receptors influence other ion channels on LC neurons? Not only do multiple G-protein-coupled receptors converge onto single effectors, single receptors are able to influence multiple effectors in
20 1
single cells via G-proteins (see Birnbaumer, 1990, for review). In addition to opening potassium channels, opioid receptors have been shown to influence the activity of voltage-dependent calcium channels in several cell types. Both 6- and K-receptors inhibit calcium currents in cultured cells (Macdonald and Werz, 1986; Hescheler et al., 1987), and p-receptors inhibit calcium currents in a human neuroblastoma cell-line (E. Seward and G. Henderson, personal communication). Although p-agonists inhibited the calcium component of the action potential in LC neurons, this was shown to be secondary to potassium current activation (North and Williams, 1983). When studied directly, voltage-dependent calcium currents were unaffected in LC neurons by high concentrations of the non-selective opioid, met-enkephalin (Williams, Christie and North, unpublished observations). In the same experiments, muscarinic receptor agonists inhibited the current. To date, the electrophysiological actions of opioids on the soma of LC neurons seem to be confined to the opening of potassium channels.
Presynaptic actions of opioids Opioids have been shown to inhibit synaptic potentials impinging on neurons throughout the nervous system. In the LC, studies of opioid actions on synaptic potentials are complicated by the post-synaptic actions of p-opioids, i.e. , the large post-synaptic increase in potassium conductance affects both the amplitude and time course of synaptic potentials (see Williams et at., this volume). In contrast, although K-agonists have no post-synaptic effects, they have been reported to produce a 38% inhibition of the amplitude of fast synaptic potentials which were evoked by electrical stimulation in rat brain slices (McFadzean et al., 1987). The fast synaptic potential is due to the release of at least three neurotransmitters: an excitatory aminoacid, GABA and glycine (see Williams et al., this volume). It is unknown which of these are inhibited by K-opioids, and which projections to LC are responsible.
Physiological role of opioids in LC The physiological role of p-receptor-mediated inhibition of LC neurons has yet to be established. Fibers and terminals containing products of pro-opiomelanocortin, pro-dynorphin and pro-enkephalin are present in the vicinity of the LC (see Sutin and Jacobowitz, this volume) and may make synaptic contacts with LC neurons (Pickel et al., 1979). However, attempts to detect a post-synaptic response to endogenous opioids released by electrical or high potassium stimulation within slices of LC met with no success, even in the presence of peptidase inhibitors (Williams et al., 1987). This result is not surprising, an endogenous opioid-mediated synaptic response has yet to be reported on any neuron. A possible explanation is that unusual stimulus parameters are needed to evoke release of endogenous opioids and other peptides, such as somatostatin, within the LC. It is also possible that endogenous opioids act on LC neurons in a diffuse, non-synaptic, manner. One report of a naloxone-reversible inhibition of LC firing following stimulation of the pro-opiomelanocortin containing arcuate nucleus (Strahlendorf et al., 1980) has yet to be shown to be a direct effect on LC using intracellular recording. Opioid tolerance and physical dependence in LC The notion that tolerance to opioids, i.e., reduced responsiveness to an agonist following chronic exposure to it, involves cellular adaptive processes within the neuron that bears opioid receptors is generally accepted (see Koob and Bloom, 1988, for review). The mechanisms underlying physical dependence are less clearly understood than those underlying tolerance. Physical dependence is characterised in whole animals by withdrawal signs upon removal of agonist, or administration of antagonist. Withdrawal signs are often qualitatively opposite to the acute agonist actions, e.g., hyperalgesia is a withdrawal sign opposite to the acute analgesic action of opioids. At the level of single neurons, signs of dependence would be
202
considered as the occurrence of enhanced spontaneous activity or heightened excitability. Tolerance and physical dependence following chronic exposure to opioids have been extensively investigated in the LC. This interest arose partly from the simiIarity between the behavioral effects of electrical stimulation of the LC and the opiate withdrawal syndrome. Indeed, this similarity provided the rationale for the use of clonidine in the treatment of opiate withdrawal in man (Gold et al., 1978). Clonidine, and other a,-adrenoceptor agonists, by acting through the same cellular pathway as opioids inhibits LC neurons and thereby reduces withdrawal excitation. Tolerance The electrophysiological mechanisms underlying tolerance have been studied in any detail only in LC neurons (Andrade et al., 1983; Christie et al., 1987). Tolerance was observed as a reduction in the sensitivity of LC neurons to agonists such as met-enkephalin and normorphine. The maximum potassium current induced by normorphine was also reduced (Christie et al., 1987). This reduced sensitivity to opioids was observed for many hours after removal of tissue from animals and hence withdrawal of morphine. These effects were indistinguishable from inactivation of a fraction of the p-receptors on the normal cell surface with the irreversible antagonist, p-chlornaltrexamine. It was, therefore, concluded that tolerance was equivalent to inactivation of p-receptors, This interpretation is functionally equivalent to a decrease in the ability of individual p-receptors to open channels. This conclusion was similar to those drawn from isolated organ experiments (Chavkin and Goldstein, 1984). No effects were found on the K , of naloxone, the potassium conductance itself, or the ability of aZadrenoceptor agonists to open the same population of potassium channels (Christie el al., 1987). No signs indicative of physical dependence (withdrawal, or antagonist-induced excitation) were observed (Andrade et al., 1983; Christie et al., 1987).
These studies localized tolerance specifically to coupling between p-receptors and potassium channels, because the channels themselves and the ability of other receptors to couple to them were unaffected. Tolerance must therefore be due to a reduced preceptor density on the cell surface, or reduced ability of receptors to couple to G-proteins. These two possibilities could not be distinguished in physiological studies. Examples of both possibilities have been observed in biochemical studies of p- and &receptors in cultured cells (Law et al., 1983; Werling et al., 1989) and brain membrane preparations (Rogers and El-Fakahany, 1986). Beyond this level of analysis, little is known of the mechanisms underlying opioid tolerance. Changes occur with chronic opioids in the adenylate-cyclase to protein kinase-A cascade (Duman et al., 1988; Nestler and Tallman, 1988) and Gprotein ribosylation in LC (Nestler et al., 1989); however, the physiological relationship of these adaptations to tolerance is unknown. An observation which could be related to tolerance development is that p-receptors were desensitised for 15-20 min after occupation by high concentrations of agonists (G.C. Harris and J.T. Williams, personal communication). This acute desensitisation is analogous to homologous desensitisation of other G-protein coupled receptors, e.g., padrenoceptors are phosphorylated and inactivated by p-adrenergic receptor-kinase only when occupied by an agonist (Benovic et al., 1989). Whether similar mechanisms play a role in the development of the long-term tolerance described above is unknown. Physical dependence Dependence (increase in extracellular action potential frequency) was first reported in LC neurons in vivo after iontophoretic application of the opioid antagonist, naloxone (Aghajanian, 1978). These results were replicated using intracerebroventricular application of an antagonist, naltrexone (Valentino and Wehby, 1989). However, these observations were not replicated using
203
extracellular (Andrade et af., 1983) or intracellular (Christie et af., 1987) recording in vitro, upon withdrawal of opioids, or upon application of naloxone. There are several possible explanations for this discrepancy. First, increased firing frequency in ciuo might simply be due to reversal of the effects of the agonist (morphine) which was still present in the tissue. However, the naloxoneinduced frequency was much greater than that observed in naive animals (Aghajanian, 1978; Valentino and Wehby, 1989), ruling out this possibility. Second, if dependence in LC dissipates very rapidly, it may have declined during the preparation of brain slices and impalement of neurons (usually longer than 30 min in the absence of an agonist). However, maintenance of tissue in morphine (1 FM) during brain slice preparation did not lead to the observation of withdrawal excitation upon superfusion of naloxone (J.T. Williams, personal communication), ruling out this possibility. Finally, increased frequency could have been due to altered synaptic activity, which would bc disrupted in brain slices. This would suggest that dependence develops in nerve terminals and/or cells projecting to the LC. This possibility has not been explored in vitro, but lesions to nucleus paragigantocellularis, which provides the main excitatory afferents to the LC, attenuated withdrawal excitation in uiuo (Rasmussen and Aghajanian, 1989). These studies imply that tolerance and dependence can be dissociated at the neuronal level and are, therefore, consequences of different cellular mechanisms. This interpretation was consistent with similar observations reported for isolated organ preparations (Wuster et al., 1982), contrary to earlier theories of opioid tolerance and dependence, which posited a unitary mechanism for both phenomena (Collier, 1980). This interpretation might imply that tolerance occurs in many types of opioid sensitive neuron, but dependence develops only in a subset. Some studies support this possibility. Cells in mouse dorsal root ganglion explants develop hyper-excitability following chronic treatment with morphine (Crain
et a/., 1988). Another possibility is that dependence requires intact neural networks or requires some form of synaptic interactions. This might explain the discrepancy between the observation of dependence in LC neurons in uivo and its absence in uitro. Conclusions The LC has contributed greatly to our understanding of the acute and chronic actions of opioids. Within the LC opioids have direct inhibitory effects on the soma, as well as presynaptic inhibitory effects. It is also likely that opioids have direct inhibitory effects on the terminals of LC neurons. The somatic effects are due to opening of inwardly rectifying potassium channels. p-receptors, as well as other membrane receptors, couple to these channels via the activation of pertussis toxin-sensitive G-proteids). Diffusible second messengers, such as CAMP, are not involved in the coupling process between preceptor and potassium channel. The physiological role of inhibition of adenylate cyclase by preceptors in LC is still unknown, but might involve long-term processes such as gene regulation. This understanding of the mechanisms of acute actions has facilitated analysis of the chronic effects of opioids in LC. Tolerance to opioids in LC in uitro is due either to a reduced number of preceptors on the soma, or to a reduced efficiency of coupling between preceptors and potassium channels. This deficit is specific in that a,-adrenoceptor agonists can still fully activate the same population of potassium channels. Signs of opioid withdrawal have been observed in LC neurons in uiuo, but not in brain slices in citro. The reasons for this discrepancy are unclear. The observation of tolerance, but not physical dependence, in LC neurons in uitro suggests the mechanisms underlying the two phenomena differ. The presynaptic actions of opioids are less clearly understood than the post-synaptic effects. K-receptor agonists depress fast synaptic potentials evoked in the LC by electrical stimulation.
204
The neurotransmitters involved and anatomical source of these synaptic potentials have yet to be described. A clear demonstration of synaptic responses to endogenously released opioids is lacking in the LC, as it is elsewhere in the nervous system. Acknowledgements Drs. J.T. Williams and S.M. Johnson are gratefully acknowledged for their comments and suggestions. References Aghajanian, G.K. (1978) Tolerance of locus coeruleus neurons to morphine and suppression of withdrawal response by clonidine. Nature (London), 276: 86-188. Aghajanian, G.K. and Wang, Y.Y. (1986) Pertussis toxin blocks the outward currents evoked by opiate and a, agonists in locus coeruleus neurons. Brain Res., 371: 390-394. Andrade, R. and Aghajanian, G.K. (1985) Opiate- and alpha,-adrenoceptor-induced hyperpolarizations of locus coeruleus neurons in brain slices: Reversal by cyclic adenosine 3’,5’-monophosphate analogues. J. Neurosci., 5: 2359-2364. Andrade, R., VanderMaelen, C.P. and Aghajanian, G.K. (1983) Morphine tolerance and dependence in locus coeruleus: Single cell studies in brain slices. Eur J. Pharmacol., 91: 161-169. Benovic, J.L., DeBlasi, A., Stone, W.C., Caron, M.G. and Lefkowitz, R.J. (1989) P-adrenergic receptor kinase: Primary structure delineates a multigene family. Science, 246: 235-240. Bird, S.J. and Kuhar, M.J. (1977) Iontophoretic application of opiates to the locus coeruleus. Brain Res., 122: 523-533. Birnbaumer, L. (1990) G proteins in signal transduction. Ann. Rec. Pharmacol. Toxicol., 30: 675-705. Chavkin, C. and Goldstein, A. (1984) Opioid receptor reserve in normal and morphine-tolerant guinea pig ileum myenteric plexus. Proc. Natl. Acad. Sci. USA, 81: 7253-7257. Christie, M.J. and North, R.A. (1988) Agonists at p opioid, M, muscarinic and GABA, receptors increase the same potassium conductance in rat lateral parabrachial neurons. Br. J. Pharmacol., 95: 896-902. Christie, M.J., Williams, J.T. and North, R.A. (1987) Cellular mechanisms of opioid tolerance: Studies in single brain neurons. Mol. Pharmacol., 32: 633-638. Collier, H.O.J. (1980) Cellular site of opiate dependence. Nature (London), 283: 625-629. Crain, S.M., Shen, K.-F. and Chalazonitis, A. (1988) Opioids excite rather than inhibit sensory neurons after chronic opioid exposure of spinal cord-ganglion cultures. Brain Res., 455: 99-109.
Duggan, A.W. and North, R.A. (1983) Electrophysiology of opioids. Pharmacol. Reu., 35: 219-281. Duman, R.S., Tallman, J.F. and Nestler, E.J. (1988) Acute and chronic opiate-regulation of adenylate cyclase in brain: Specific effects in locus coeruleus. J. Pharrnacol. Exp. Ther., 246: 1033-1039. Gold, M.S., Redmond, D.E. and Kleber, H.D. (1978) Clonidine in opiate withdrawal. Lancet, i: 929-930. Guyenet, P.G. and Aghajanian, G.K. (1977) Excitation of neurons in the nucleus locus coeruleus by substance-P and related peptides. Brain Rex, 136: 178-184. Hescheler, J., Rosenthal, W., Trautwein, W. and Schultz, G. (1987) The GTP-binding protein, Go, regulates neuronal calcium channels. Nature (London), 325: 445-447. Hille, B. (1984) Ionic Channels in Excitable Membranes, Sinauer, Sunderland, MA, 426 pp. Inoue, M., Nakajima, S. and Nakajima Y. (1988) Somatostatin induces an inward rectification in rat locus coeruleus neurons through a pertussis toxin-sensitive mechanism. J. Physiol. (London), 407: 177-198. Koob, G.F. and Bloom, F.E. (1988) Cellular and molecular mechanisms of drug dependence. Science, 242: 715-723. Korf, J., Bunney, B.S. and Aghajanian, G.K. (1974) Noradrenergic neurons: Morphine inhibition of spontaneous activity. Eur. J. Pharmacol., 25: 165-167. Law, P.Y., Hom, D.S. and Loh, H.H. (1983) Opiate receptor down-regulation and desensitization in neuroblastoma X glioma NG108-15 hybrid cells are two separate cellular adaptation processes. Mol. Pharmacol., 24: 413-424. Leslie, F.M. (1987) Methods used for the study of opioid receptors. Pharmacol. Reu., 39: 197-249. Macdonald, R.L. and Werz, M.A. (1986) Dynorphin A decreases voltage-dependent calcium conductance of mouse dorsal root ganglion neurons. J. Physiol. (London), 377: 237-249. Madison, D.V. and Nicoll, R.A. (1987) Enkephalin hyperpolarizes interneurons in rat hippocampus. J. Physiol. (London), 398: 123-130. McFadzean, I., Lacey, M.G., Hill, R.G. and Henderson, G. (1987) Kappa opioid receptor activation depresses synaptic input to rat locus coeruleus neurons in vitro. Neuroscience, 20: 231-239. Miyake, M., Christie, M.J. and North, R.A. (1989) Single potassium channels opened by opioids in rat locus coeruleus neurons. Proc. Natl. Acad. Sci. USA, 86: 34193422. Nakamura, S., Tepper, J.M., Joung, S.J., Ling, N. and Groves, P.M. (1982) Noradrenergic terminal excitability: Effects of opioids. Neurosci. Lett., 30: 57-62. Nestler, E.J. and Tallman, J.F. (1988) Chronic morphine treatment increases cyclic AMP-dependent protein kinase activity in the rat locus coeruleus. Mol. Pharmacol., 33: 127-132. Nestler, E.J., Erdos, J.J., Terwilliger, R., Durnan, R.S. and Tallman, J.F. (1989) Regulation of G proteins by chronic morphine in rat locus coeruleus. Brain Res., 476: 230-239. North, R.A. (1989) Drug receptors and the inhibition of nerve cells. Br. J. Pharmacol., 98: 13-28.
205 North, R.A. and Williams, J.T. (1983) Opiate activation of potassium conductance inhibits calcium action potentials in rat locus coeruleus neurons. Br. J. Pharmacol., 80: 225-228. North, R.A. and Williams, J.T. (1985) On the potassium conductance increased by opioids in locus coeruleus neurons. J. Physiol. (London), 364: 265-280. North, R.A., Williams, J.T., Surprenant, A. and Christie, M.J. (1987) p and S opioid receptors both belong to a family of receptors which couple to a potassium conductance. Proc. Natl. Acad. Sci. USA, 84: 5487-5491. Osmanovic, S.S. and Shefner, S.A. (1988) Baclofen increases the potassium conductance of rat locus coeruleus neurons recorded in brain slices. Brain Res., 438: 124-136. Pepper, C.M. and Henderson, G. (1980) Opiates and opioid peptides hyperpolarize locus coeruleus neurons in vitro. Science, 209: 394-396. Pickel, V.M., Joh, T.H., Reis, D.J., Leeman, S.E. and Miller, R.J. (1979) Electron microscopic localization of substance P and enkephalin in axon terminals related to dendrites of catecholaminergic neurons. Brain Res., 160: 387-400. Rasmussen, K. and Aghajanian, G.K. (1989) Withdrawal-induced activation of locus coeruleus neurons in opiate-dependent rats: Attenuation by lesions of the nucleus paragigantocellularis. Brain Res., 505: 346-350. Rogers, N.F. and El-Fakahany, E. (1986) Morphine-induced opioid receptor down-regulation detected in intact adult rat brain cells. Eur. J. Pharmacol., 124: 221-230. Shefner, S.A. and Chiu, T.H. (1986) Adenosine inhibits locus coeruleus neurons: A n intracellular study in a rat brain slice preparation. Brain Res., 366: 364-368. Strahlendorf, H.K., Strahlendorf, J.C. and Barnes, C.D. (1980) Endorphin mediated inhibition of locus coeruleus neurons. Brain Rex, 191: 284-288. Ueda, H., Harada, H., Nozaki, M., Katada, T., Ui, M., Satoh, M. and Takagi, H. (1988) Reconstitution of rat brain 1 opioid receptors with purified guanine nucleotide binding regulatory proteins, Gi and Go. Proc. Natl. Acad. Sci. USA, 85: 7013-7017.
Valentino, R.J. and Wehby, R.G. (1989) Locus ceruleus discharge characteristics of morphine-dependent rats: Effects of naltrexone. Brain Res., 488: 126-134. Wang, Y.Y. and Aghajanian, G.K. (1987) Excitation of locus coeruleus neurons by adenosine 3’,5’-cyclic monophosphate-activated inward current: Extracellular and intracellular studies in rat brain slices. Synapse, 1: 481-487. Werling, L.L., Brown, S.L. and Cox, B.M. (1987) Opioid receptor regulation of the release of norepinephrine in brain. Neuropharmacology, 26: 987-996. Werling, L.L., Mcmahon, P.N. and Cox, B.M. (1989) Selective changes in p opioid receptor properties induced by chronic morphine exposure. Proc. Natl. Acad. Sci. USA, 86: 63936397. Williams, J.T. and North, R.A. (1984) Opiate-receptor interactions on single locus coeruleus neurons. Mol. PharmaC O ~,. 26: 67-70. Williams, J.T., Egan, T.M. and North, R.A. (1982) Enkephalin opens potassium channels in mammalian central neurons. Nature (London), 299: 74-76. Williams, J.T., Christie, M.J., North, R.A. and Roques, B.P. (1987) Potentiation of enkephalin action by peptidase inhibitors in rat locus coeruleus in vitro. J. Pharmacol. Exp. Ther., 243: 397-401. Williams, J.T., North, R.A. and Tokimasa, T. (1988) Inward rectification of resting and receptor-linked potassium currents in rat locus coeruleus neurons. J. Neurosci., 8: 42994306. Wong, Y.H., Demoliou-Mason, C.D., Barnard, E.A. (1989) Opioid receptors in magnesium-digitonin-solubilized rat brain membranes are tightly coupled to a pertussis toxinsensitive guanine nucleotide-binding protein. J. Neurochem., 52: 999-1009. Wuster, M., Schulz, R. and Herz, A. (1982) The development of opiate tolerance may dissociate from dependence. Life Sci., 31: 1695-1698.
This Page Intentionally Left Blank
C.D. Barnea and 0. Pompeiano (Ed$.) Progress in Brain Research, Vol. 8s 0 1 Y Y 1 Elscvicr Science Publishers B.V.
207 CHAPTER 15
Afferent effects on locus coeruleus in opiate withdrawal K. Rasmussen Central Nercous System Research, Lilly Research Laboratories, Eli Lilly and Co., Lilly Corporate Center, Indianapolis, IN, U.S.A.
The locus coeruleus (LC) has been hypothesized to play an important role in opiate withdrawal. This hypothesis is supported, in part, by the finding that LC neurons greatly increase their activity during antagonist-precipitated morphine withdrawal and that this increased activity correlates temporally with withdrawal behavior. However, this withdrawal-induced increase in unit activity is not seen in citro in brain slices taken from morphine-dependent animals. indicating that afferents to the LC play an important role in the withdrawal-induced activation of these neurons. This chapter reviews data indicating: (1) the morphine-withdrawal-induced activation of LC neurons is mediated predominantly by non-N-methyl-D-aspartate (NMDA) excitatory amino acid pathways in the brain; (2) the activation of the LC during morphine withdrawal may be mediated, at least in part, by an
excitatory amino acid projection from the nucleus paragigantocellularis. The role of other excitatory amino acid pathways in the withdrawal-induced activation of the LC remains to be determined; ( 3 ) intrinsic changes in the G-protein/cyclic AMP system of LC cells may play an important role in mediating the effects of afferent inputs to the LC during morphine withdrawal; (4) NMDA antagonists (unlike the a z agonist clonidine) attenuate the behavioral signs of morphine withdrawal without blocking the withdrawal-induced increase of LC unit activity. In addition, non-competitive NMDA antagonists like MK801 may not be useful to alleviate opiate-withdrawal symptoms in man because of their PCP-like side effects. However, competitive NMDA antagonists like LY274614 could be of great benefit for alleviating opiatewithdrawal symptoms in man.
Key words: locus coeruleus, opiate withdrawal, excitatory amino acids, N-methyl-D-aspartate. kynurenic acid, MK 801, LY 274614, nucleus paragigantocellularis, G-proteins, adenylate cyclase, cyclic AMP-dependent protein kinase
Introduction
The brain noradrenergic (NA) system has been hypothesized to play an important role in opiate dependence and withdrawal (see Redmond and Krystal, 1984). Studies of opiate regulation of the NA system have focused on the locus coeruleus (LC), the largest cluster of NA neurons in the mammalian brain (Dahlstrom and Fuxe, 1965; Foote et al., 1983). Autoradiographic studies have
shown that the LC possesses a high density of opiate receptors, particularly of the mu and kappa subtype (Pert et al., 1975; Tempe1 and Zukin, 1987). Single-unit recordings in anesthetized or awake rats have shown that LC neurons are inhibited by local or systemic administration of opiates (Korf et al., 1974; Bird and Kuhar, 1977; Aghajanian, 1978; Valentino and Wehby, 1988). In addition, in awake cats, local (but not systemic) administration of opiates decreases the
208
activity of LC neurons (Rasmussen and Jacobs, 1985; Abercrombie et al., 1988) The mechanisms underlying the acute inhibition of LC neurons by opiates have been examined with electrophysiological and biochemical techniques. Intracellular recordings in vitro have suggested that acute opiate inhibition of LC neurons involves regulation of two types of ion channels: activation of a potassium channel and inhibition of a slowly depolarizing sodium channel (Aghajanian and Wang, 1987; Wang and Aghajanian, 1987; North et al., 1987). Opiate regulation of both channels is mediated through pertussis toxin-sensitive G-proteins (Aghajanian and Wang, 1986; North et al., 1987), and opiate regulation of the slowly depolarizing sodium channel may be mediated through an opiate/G-protein-induced decrease in neuronal cyclic AMP levels (Aghajanian and Wang, 1987; Wang and Aghajanian, 1987). Biochemical studies have demonstrated that, acutely, opiates inhibit adenylate cyclase activity in the LC, an effect mediated through a pertussis toxin-sensitive G-protein(s) (Duman et al., 1988; Beitner et al., 1989), and decreased cyclic AMP-dependent protein phosphorylation in this brain region (Guitart and Nestler, 1989). Activity of the locus coeruleus during opiate withdrawal
Chronic administration of morphine leads to tolerance to its inhibitory effects on LC neurons (Aghajanian, 1978). LC neurons also exhibit dependence after chronic opiate exposure, as administration of opiate receptor antagonists to opiate-dependent rats results in increased LC firing rates in uiuo (Aghajanian, 1978; Valentino and Wehby, 1989; Rasmussen and Aghajanian, 1989a). It has been proposed that this withdrawal-induced activation of LC neurons plays a role in some of the behavioral manifestations of the opiate-abstinence syndrome seen in animals and humans (see Redmond and Krystal, 1984) The time course of the withdrawal-induced activation of LC neurons has recently been exam-
ined and this activity correlates with the behavioral symptoms of opiate withdrawal (Rasmussen et al., 1990). Rats were given morphine by daily subcutaneous implantation of morphine pellets for 5 days. On the sixth day, the last two morphine pellets were removed and morphine withdrawal was induced by subcutaneous administration of naltrexone, an opiate receptor antagonist, with additional doses given 6, 24, and 72 h later. Figure 1 shows that the behavioral symptoms of opiate withdrawal and the withdrawal-induced activation of LC neurons showed virtually identical time courses. Both measures rose dramatically within minutes of naltrexone administration and showed increased levels for over 24 h. Both measures peaked within 15-30 min, decreased rapidly from this peak over the first 2-4 h, remained at relatively stable, elevated levels from 4-24 h, and returned to baseline by 72 h. The recovery from behavioral abstinence and LC withdrawal activation, therefore, appeared to follow two phases: an early, rapid phase during which time > 50% of the behavioral and electrophysiological signs of withdrawal resolved, and a later, slower phase during which time the remaining behavioral and electrophysiological signs of withdrawal required 72 h of continuous withdrawal for complete resolution. The parallel between withdrawal behavior and LC neuronal activity suggests that activation of LC neurons plays a role in the opiate abstinence syndrome. This is consistent with a number of studies that correlate opiate abstinence symptoms with LC firing rates. First, clonidine, an a*adrenergic receptor agonist, suppresses both the increased LC unit activity seen during opiate withdrawal (Aghajanian, 1978) and the severity of opiate withdrawal symptoms in animals and humans (Tseng et al., 1975; Gold et al., 1978; Taylor et al., 1988). The effectiveness of clonidine in reducing LC neuronal activity and opiate withdrawal behaviors has been reported after both systemic and local LC infusion (Aghajanian, 1978; Taylor et al., 1988). Second, electrical stimulation of the LC in non-human primates produces be-
209
haviors similar to those observed during opiate withdrawal (Grant et al., 1988). Third, local administration of opiate receptor antagonists into the LC of opiate-dependent animals elicits a number of withdrawal-like behaviors (Esposito et al., 1987). Fourth, an increase in NA function (indicated by increased turnover of norepinephrine) has been found in cerebral cortex during opiate withdrawal (Zigun e f al., 1981). This increase, which can be reversed by clonidine (Zigun et al., 1981; DiStefano and Brown, 19851, presumably reflects the increase in LC neuronal activity
0
L 1
4
6
24
72
TIME (h)
t
t
t
t
-t-
---e
1
4
6
24
Morphine Cantral
72
TIME (h)
t
t
Naltrexone
t
observed under these conditions. However, since lesions of the dorsal NA bundle, the major ascending output of the LC, do not prevent many of the behavioral manifestations of opiate withdrawal in rats (Britton et al., 19841, actions of clonidine on the descending outputs of the LC or at other sites (e.g., adrenergic neurons elsewhere in the brain, amygdaloid neurons, dorsal root ganglion/dorsai horn neurons, spinal sympathetic neurons) may be involved (see Franz et al., 1982; Freedman and Aghajanian, 1985; Crain et al., 1987). In addition, since a2 agonists that do not readily penetrate the blood-brain barrier can reduce some symptoms of opiate withdrawal, a peripheral site of action may play a role in clonidine’s ability to lessen opiate withdrawal symptoms (Taylor et al., 1988, Thollander et al., 1989). Further, since clonidine is effective on only a subset of withdrawal symptoms in humans (Charney et al., 1982; Jasinski et al., 19851, other neurotransmitter systems are likely to be involved in.
Fig. 1. Time course of abstinence behaviors and locus coeruleus (LC) neuronal activity during opiate withdrawal. Rats were treated with morphine for six days and withdrawal was precipitated by naltrexone administration. Naltrexone doses were repeated 6, 24, and 72 h after the initial injection. This protocol was shown to result in sustained, maximal levels of withdrawal. Top panel. The severity of the resulting opiate abstinence syndrome was studied by use of a composite score of 14 withdrawal behaviors. The data represent means of composite scores kS.E.M. of 6 animals at the 0-6 h time points and 4 animals at the 24 and 72 h time points. Bottom panel. The firing rates of LC neurons during opiate withdrawal were studied by single-unit recording techniques. The data represent mean firing rates + S.E.M. of 10 to 40 neurons at each time point from 2-4 rats. Statistical analyses (ANOVA scores and paired t-tests) of the behavioral and electrophysiological data revealed the following significant differences ( P < 0.002). Behavioral scores and LC firing rates of withdrawing animals, at each of the time points between 15 min and 24 h of withdrawal, were significantly different from baseline, pre-naltrexone scores (zero time point), and from control animals that had received identical naltrexone injections. In addition, the levels of peak behavioral symptoms and LC firing rates (i.e., at 15-30 min) were significantly different from those observed at 4, 6, and 24 h. Finally, repeat naltrexone injections at 6, 24, and 72 h failed to elicit increases in behavioral symptoms or LC firing rates above their respective pre-injection levels. (From Rasmussen et al., 1990.)
210
the full manifestation of opiate-withdrawal symptoms. The role of afferent inputs to the locus coeruleus in opiate-withdrawal-induced activity
Effects of anatomical or functional deafferentation of the locus coeruleus While LC neurons show a dramatic withdrawal-induced activation in vivo they do not show withdrawal-induced activation in brain slices taken from morphine-dependent rats (Andrade et al., ,19831, suggesting the possibility that the activation seen in uivo is mediated by afferents to the LC which are disconnected in vitro. In an effort to evaluate the role of LC afferents in the activation of LC neurons induced by opiate withdrawal in viuo, radio-frequency lesions were placed in two major afferents to the LC, the nucleus paragigantocellularis (PGi) and nucleus prepositus hypoglossi (PrH) (Aston-Jones et al., 1986; see Fig. 2). Since a subset of the projection from the PGi to the LC has been shown to empIoy an excitatory amino acid (Ennis and Aston-Jones, 1988), some animals were pretreated with kynurenic acid (a non-selective excitatory amino acid antagonist) prior to precipitating withdrawal. Figure 3 shows the effects of lesions of the PGi and PrH, and of intracerebroventricular (icv) kynurenic acid administration on the firing rate of LC neurons recorded in anesthetized, morphine-dependent rats. Lesions of the PrH did not attenuate or enhance the withdrawal-induced activation, indicating that this afferent is not involved in the withdrawal-induced activation of the LC. Lesions of the interpeduncular nucleus and central nucleus of the amygdala (two nuclei thought to be involved in morphine withdrawal; Geary and Wooten, 1983) also did not block withdrawal activation of the LC (not shown; Rasmussen and Aghajanian, 1989a). Lesions of the PGi, however, greatly attenuated the activation of LC neurons during naltrexone-induced morphine withdrawal. As an added control, in some of the
Fig. 2. Representative coronal sections (stained with cresyl violet) of brains from animals in which lesions (*) were made in either the paragigantocellularis (PGi) (top) or prepositus hypoglossi (PrH) (bottom). Numbers to the left of each section refer to AP coordinates according to Paxinos and Watson (1982). Note that the lesions are ovoid or spherical in shape, corresponding to the natural contour of the target nuclei. Multiple lesions were made in each animal to destroy the PGi from P2.3 to P3.3 or the PrH from P1.8 to P3.3. (From Rasmussen and Aghajanian, 1989a.)
PGi-lesioned animals, recordings were made in the LC both ipsilateral and contralateral to the lesion. This data is shown in Figure 4. After naltrexone administration, the firing rates in the LC contralateral to the lesion rose to levels which were not significantly different from unlesioned chronic morphine animals, while the firing rates in the LC ipsilateral to the lesion were significantly lower than in unlesioned chronic morphine animals, indicating that any possible nonspecific or systemic effects of PGi lesions, (e.g., blood pressure changes) did not contribute to the attenuation of the LC activation. This result is consistent with anatomical studies indicating that the
21 1
-
5
E I ”
4
=
3
2
2
Y
1
m
sham chronic morphine +PrH lesion +PGi lesion iKynurenate
n Pre NTX
Post NTX (3-15 min) (15 30 min)
Fig. 3. Naltrexone-induced withdrawal activation of LC neurons in PGi lesioned, PrH lesioned and kynurenate pretreated animals. Rats were treated with morphine for four days. Bars represent mean LC firing rates ( + S.E.M.) for animals receiving sham pellets, morphine pellets, morphine pellets and PrH lesions, morphine pellets and PGi lesions, or morphine pellets and kynurenate pretreatment, both before and 3-15 and 15-30 rnin after naltrexone (NTX) administration. (From Rasmussen and Aghajanian, 1989a.)
PGi input to the LC is predominantly ipsilateral (Aston-Jones et al., 1986). Pretreatment with kynurenic acid greatly attenuated the activation of LC neurons during morphine withdrawal, supporting the idea that an excitatory amino acid projection from the PGi plays a role in the activation of the LC during morphine withdrawal. Although the icv route of administration does not permit precise localization of its site of action,
+ PGI lesion-contra m c
i
2
0 4
1
0 Pre NTX
Post NTX
kynurenic acid has recently been shown to block the effects of excitatory amino acid agonists directly in the LC (Olpe et al., 1989). The radio-frequency lesions used in this study allow for the possibility that a destruction of fibers of passage accounts for the effects of the lesions. Although cell body lesions, (e.g., as induced by an excitotoxin such as ibotenic acid) would be useful in this regard, morphine-dependent animals have greatly increased mortality after receiving ibotenic acid injections in the PGi (Rasmussen and Aghajanian, 1989a). In addition, with cell body lesions several days are needed for a full lesion to occur, potentially affecting the development of tolerance to morphine.
How excitatory amino acid afferents mediate the opiate-withdrawal-inducedactivation of LC neurons There are several ways in which the excitatory amino acid afferents to the LC from the PGi may play a role in the activation of LC neurons during morphine withdrawal. One possibility is that excitatory PGi neurons themselves greatly increase their activity. Another possibility, since morphine has been shown to inhibit the release of excitatory amino acids (Crowder et af., 19861, is that during morphine withdrawal there is an increased release of excitatory amino acids from PGi afferent nerve terminals in the LC. A third possibility is that intrinsic changes in LC neurons may enhance the sensitivity of the LC to excitatory inputs from the PGi during morphine withdrawal. This last possibility will be discussed in more detail below.
(3 30 min)
Fig. 4. Naltrexone-induced withdrawal activation of LC neurons ipsilateral and contralateral to PGi lesions. Rats were treated with morphine for four days. Bars represent mean LC firing rates (+S.E.M.) for animals receiving sham pellets, morphine pellets, or morphine pellets and PGi lesions with recordings made in the LC both ipsilateral and contralateral to the lesion in the same animal, both before and for 30 min after naltrexone (NTX) administration. For PGi-lesioned animals the contralateral LC was recorded from 3-15 min and the ipsilateral L C from 15-30 min after naltrexone. (From Rasrnussen and Aghajanian, 1989a.)
Intrinsic changes in LC neurons In the LC, chronic opiate administration has been shown to up-regulate the G-protein/cyclic AMP system at each major step between receptor and physiological response. Chronic opiates have been shown to increase levels of Gi, and Go,, adenylate cyclase activity, cyclic AMP-dependent protein kinase activity, and a number of MARPPs
212
(morphine- and cyclic AMP-regulated phosphoproteins) in the LC, effects not observed in several other brain regions studied (Nestler and Tallman, 1987; Duman et af., 1988; Nestler et af., 1989; Guitart and Nestler, 1989, 1990; Guitart et al., 1990). In the absence of alterations in opiate receptors or opiate-regulated ion channels in the LC in response to chronic opiate administration (see Redmond and Krystal, 1984; Christie et al., 1987; Loh et al., 19881, it has been proposed that such an up-regulated G-protein/cyclic AMP system contributes to opiate tolerance, dependence, and/or withdrawal in the LC (Nestler, 1990). In order to examine this hypothesis, the levels of G-proteins, adenylate cyclase, and cyclic AMP-dependent protein kinase were examined during morphine withdrawal and compared to the time course of the behavioral manifestations of opiate withdrawal and the in uiilo activity of LC neurons as discussed above (see Fig. 1). The biochemical parameters measured in the LC appeared to correlate temporally with the early, rapid phase of withdrawal. Thus, adenylate cyclase (Fig. 51, cyclic AMP-dependent protein kinase activities and ADP-ribosylation levels of G,, and Go,, all elevated in the LC of dependent animals, remained elevated at 20 min, showed partial declines toward control levels (but remained significantly elevated) at 1 h, and were back to control levels by 6 h. Taken together, the time courses of the electrophysiological (see above; Fig. 1) and biochemical data suggest that increased levels of G-proteins and an up-regulated cyclic AMP system may contribute to the early withdrawal activation of LC neurons. In contrast to the early stages of withdrawal, the up-regulated cyclic AMP system cannot account for the sustained activation of LC neurons and withdrawal symptoms seen during later stages of withdrawal. Rather, these later signs of withdrawal could be due to various MARPPs known to be increased in opiate-dependent animals (Guitart and Nestler, 1989), to other intracellular messenger systems in the LC, and/or to various extrinsic inputs to the LC.
-20
0 1
24
72
TIME (h)
t
t
t
Naltrexone Fig. 5. Time course of adenylate cyclase activity during opiate withdrawal. Rats were treated with morphine for six days and withdrawal was precipitated by naltrexone administration. The data represent forskolin-stimulated conditions; similar results were seen under basal conditions. Adenylate cyclase levels were significantly elevated above control in dependent animals (no naltrexone administration) and after 20 min and 1 h of withdrawal ( P < 0.05 by chi square test); levels were not different from control after 6, 24, and 72 h of withdrawal. (From Rasmussen et ul., 1990.)
Source of excitatory amino acid afferents to the LC The effects of PGi lesions and kynurenic acid administration on the withdrawal-induced activity of LC neurons indicates that activation of the LC during morphine withdrawal may be mediated at least in part by an excitatory amino acid projection from the PGi. However, it is interesting that the attenuation of the withdrawal-induced activation of LC neurons produced by kynurenic acid was significantly greater than that produced by PGi lesions (see Fig. 2). One possible explanation for this difference is that the lesions did not destroy all afferents emanating from the PGi. Another possibility is that there are other excitatory amino acid afferents involved in withdrawal activation of the LC. Consistent with this possibility is the fact that PGi lesions completely fail to block somatosensory activation of LC neurons (Rasmussen and Aghajanian, 1989b1, in contrast to kynurenic acid which is able to block both somatosensory and withdrawal-induced activation
213
of the LC. Therefore, it is possible that a PGi-independent, but kynurenate-sensitive, afferent pathway to the LC is also activated during morphine withdrawal. One possible source of such a PGi-independent excitatory amino acid pathway to the LC is lamina I of the spinal cord. Consistent with this view is anatomical (Cedarbaum and Aghajanian, 1978) and electrophysiolo/gical (McMahon and Wall, 1985) evidence for a direct projection to the LC from lamina I of the spinal cord. Another possible source of a PGi-independent, but kynurenate-sensitive pathway is a recently defined excitatory amino acid pathway to the LC from the medial prefrontal cortex (Highfield and Grant, 1989). Finally, kynurenic acid could also be having an indirect effect by acting on excitatory amino acid pathways that do not project directly to the LC.
Effects of systemically administered excitatory amino acid antagonists on morphine withdrawal behavior and locus coeruleus unit activity Given that icv administration of kynurenic acid blocks the withdrawal-induced activation of the LC, could systemically administered excitatory amino acid antagonists be effective in blocking the behavioral symptoms of opiate withdrawal? In preliminary studies, systemically administered kynurenic acid was found to block, dose-dependently, some, but not all, of the behavioral signs of opiate withdrawal (Rasmussen et al., in press). However, single-unit recordings in the LC showed that the same doses of kynurenic acid that effectively attenuated the behavioral symptoms of opiate withdrawal did not block the withdrawal-induced activation of the LC. Since icv administration of kynurenic acid does block the withdrawalinduced increase in LC unit activity, a peripheral site of action for systemically administered kynurenic acid is implicated. Indeed, a recently defined glutamate receptor in the myenteric plexus (Shannon and Sawyer, 1989) could play a role in kynurenic acid’s blockade of opiatewithdrawal behaviors, especially diarrhea and weight loss. However, as yet, there are no other
functionally defined glutamate receptors in the periphery. In addition, kynurenic acid has been suggested to have some central effects when administered peripherally (Gill and Woodruff, 1990). Therefore, the exact role of peripheral vs. central glutamate receptors in kynurenic acid’s blockade of morphine-withdrawal behaviors remains to be determined. Since there is some question as to kynurenic acid’s ability to penetrate into the brain, we next examined excitatory amino acid antagonists that are known to have central effects. In preliminary studies, we have examined antagonists selective for a subtype of excitatory amino acid receptor, the N-methyl-D-aspartate (NMDA) receptor, on morphine-withdrawal behavior and the withdrawal-induced activation of LC cells (Rasmussen et al., 1991). Systemic administration of the noncompetitive NMDA antagonist MK801 dose dependently blocked the behavioral signs of withdrawal in morphine-dependent rats. However, the same doses of MK801 that blocked morphine withdrawal also simultaneously (and dose dependently) produced PCP-like behavioral effects. The competitive NMDA antagonist LY274614 also dose dependently blocked the behavioral signs of withdrawal in morphine-dependent rats but did not produce any PCP-like behavioral effects. Single-unit recordings of LC neurons from morphine-dependent animals showed that neither MK801 nor LY274614 blocked the withdrawal-induced activation of these neurons. Thus, NMDA antagonists (unlike the a2 agonist clonidine; see above) attenuate the behavioral signs of morphine withdrawal without blocking the withdrawal-induced increase of LC unit activity. Therefore, these studies indicate that the excitatory amino acid-induced activation of the LC during opiate withdrawal is not mediated through NMDA receptors. Consistent with this finding are results indicating that the excitatory amino acid projection to the LC from the PGi is mediated by non-NMDA receptors (Ennis and AstonJones, 1988). Studies aimed at identifying the subtype of glutamate receptor involved in the
214
withdrawal-induced activation of the LC are currently underway. These results also indicate that non-competitive NMDA antagonists like MK801 may not be useful to alleviate opiate-withdrawal symptoms in man because of their PCP-like side effects. However, competitive NMDA antagonists like LY274614 could be of great benefit for alleviating opiate-withdrawal symptoms in man. Acknowledgements
Some of this work was supported by PHS Grant MH-17871 and the State of Connecticut. The author would like to thank Marsha Stockton and Nancy Margiotta for their excellent technical assistance. References Abercrombie, E.D., Levine, E.S. and Jacobs, B.L. (1988) Microinjected morphine suppresses the activity of locus coeruleus noradrenergic neurons in freely moving cats. Neurosci. Lett., 86: 334-338. Aghajanian, G.K. (1978) Tolerance of locus coeruleus neurones to morphine and suppression of withdrawal response by clonidine. Nature (London), 276: 186-187. Aghajanian, G.K. and Wang, Y.-Y. (1986) Pertussis toxin blocks the outward currents evoked by opiate and a*agonists in locus coeruleus neurons. Brain Res., 371: 390394. Aghajanian, G.K. and Wang, Y.-Y. (1987) Common alpha, and opiate effector mechanisms in the locus coeruleus: Intracellular studies in brain slices. Neuropharmacology, 26: 789-800. Andrade, R., VanderMaelen, C.P. and Aghajanian, G.K. (1983) Morphine tolerance and dependence in the locus coeruleus: Single cell studies in brain slices. Eur. J. Pharmacol., 91: 161-169. Aston-Jones, G.,Ennis, M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986) The brain nucleus locus coeruleus: Restricted afferent control of a broad afferent network. Science, 234: 734-736. Beitner, D.B., Duman, R.S. and Nestler, E.J. (1989) A novel action of morphine in the rat locus coeruleus: Persistent decrease in adenylate cyclase. Mol. Phamacol., 35: 559564. Bird, S.J. and Kuhar, M.J. (1977) Iontophoretic application of opiates to the locus coeruleus. Brain Res., 1 2 2 523-533. Britton, K.T., Svensson, T.,Schwartz, J., Bloom, F.E. and Koob, G.F. (1984) Dorsal noradrenergic bundle lesions fail
to alter opiate withdrawal or suppression of opiate withdrawal by clonidine. Life Sci., 34: 133-139. Cedarbaum, J.M. and Aghajanian, G.K. (1978) Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique. J. Comp. Neurol., 178: 1-16. Charney, D.S., Riordan, C.E., Kleber, H.D., Murburg, M., Braverman, P., Sternberg, D.E., Heninger, G.R. and Redmond, D.E. (1982) Clonidine and Naltrexone: A safe, effective, and rapid treatment of abrupt withdrawal from methadone therapy. Arch. Gen. Psychiatry, 39: 1327-1332. Christie, M.J., Williams, J.T. and North, R.A. (1987) Cellular mechanisms of opioid tolerance: Studies in single brain neurons. J. Pharmacol. Exp. Ther., 32: 633-638. Crain, S.M., Crain, B. and Makman, M.H. (1987) Pertussis toxin blocks depressant effects of opioid, monoaminergic and muscarinic agonists on dorsal-horn network responses in spinal cord-ganglion cultures. Brain Res., 400: 185-190. Crowder, J.M., Norris, D.K. and Bradford, H.F. (1986) Morphine inhibition of calcium fluxes, neurotransmitter release and protein and lipid phosphorylation in brain slices and synaptosomes. Biochem. Pharmacol., 35: 2501-2507. Dahlstrom, A. and Fuxe, K. (1965) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brainstem neurons. Acta Physiol. Scand., Suppl. 232: 1-55. DiStefano, P.S. and Brown, O.M. (1985) Biochemical correlates of morphine withdrawal. 2. Effects of clonidine. J. Pharmacol. Exp. Ther., 233: 339-344. Duman, R.S., Tallman, J.F. and Nestler, E.J. (1988) Acute and chronic opiate-regulation of adenylate cyclase in brain: Specific effects in locus coeruleus. J. Pharmacol. Exp. Ther., 246: 1033-1039. Ennis, M. and Aston-Jones, G. (1988) Activation of locus coeruleus from nucleus paragigantocellularis: A new excitatory amino acid pathway in brain. J. Neurosci., 8: 36443657. Esposito, E., Kruszewska, A,, Ossowska, G. and Samanin, R. (1987) Noradrenergic and behavioral effects of naloxone injected in the locus coeruleus of morphine-dependent rats and their control by clonidine. Psychopharmacology, 93: 393-396. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Reu , 63: 844-914. Franz, D.N., Hare, B.D. and Mccloskey, K.L. (1982) Spinal sympathetic neurons: Possible sites of opiate withdrawal suppression by clonidine. Science, 215: 1643-1645. Freedman, J.E. and Aghajanian, G.K. (1985) Opiate and a,-adrenoceptor responses of rat amygdaloid neurons: Co-localization and interactions during withdrawal. J. Neurosci., 5: 3016-3024. Geary, W.A.11 and Wooten, G.F. (1983) A behavioral and 2-deoxyglucose autoradiographic study of the effects of cumulative morphine dose on naloxone precipitated withdrawal in the rat. Brain Rex, 275: 117-126. Gill, R. and Woodruff, G.N. (1990) The neuroprotective actions of kynurenic acid and MK-801 in gerbils are synergis-
215 tic and not related to hypothermia. Eur. J. Pharmacol., 176: 143-149. Gold, M.S., Redmond, D.E., Jr. and Kleber, H.D. (1978) Clonidine blocks acute opiate-withdrawal symptoms. Lancet, ii: 599-602. Grant, S.J., Huang, Y.H. and Redmond, Jr., D.E. (1988) Behavior of monkeys during opiate withdrawal and locus coeruleus stimulation. Pharmacol. Biochem. Behau., 30: 13-19. Guitart, X. and Nestler, E.J. (1989) Identification of morphine and cyclic AMP-regulated phosphoproteins (MARPPs) in the locus coe ruleus and other regions of rat brain. Regulation by acute and chronic morphine. J. Neurosci., 9: 4371-4387. Guitart, X. and Nestler, E.J. (1990) Identification of MARPP14-20, morphine- and cyclic AMP-regulated phosphoproteins of 14-20 K,, as myelin basic proteins. Evidence for their acute and chronic regulation by morphine in rat brain. Brain Res., 516: 57-65. Guitart, X., Hayward, M.D., Nisenbaum, L.K., Beitner, D.B., Haycock, J.W. and Nestler, E.J. (1990) Identification of MARPP-58, a morphine- and cyclic AMP-regulated phosphoprotein of 58 K,, as tyrosine hydroxylase: Evidence for regulation of its expression by chronic morphine in the rat locus coeruleus. J. Neurosci., 10: 2635-2645. Highfield, D. and Grant, S.J. (1989) Electrophysiological evidence for an excitatory amino acid pathway from medial prefrontal cortex to lateral dorsal tegmental nucleus and rostra1 locus coeruleus. SOC.Neurosci. Abstr., 15: 644. Jasinski, D.R., Johnson, R.E. and Kocher, T.R. (1985) Clonidine and morphine withdrawal: Differential effects on signs and symptoms. Arch. Gen. Psychiatry, 42: 1063-1066. Korf, J., Bunney, B.S. and Aghajanian, G.K. (1974) Noradrenergic neurons: Morphine inhibition of spontaneous activity. Eur. J. Pharmacol., 25: 165-169. Loh, H.H., Tao P.-L. and Smith, A.P. (1988) Invited review: Role of receptor regulation in opioid tolerance mechanisms. Synapse, 2: 457-462. McMahon, S.B. and Wall, P.D. (1985) Electrophysiological mapping of brainstem projections of spinal cord lamina I cells in the rat brain. Brain Res., 333: 19-26. Nestler, E.J. (1990) Adaptive changes in signal transduction systems: Molecular mechanisms of opiate addiction in the rat locus coeruleus. In J.M. Ritchie and P.J. Magistretti (Eds.), Progess in Cell Research, Vol. I , Allan R. Liss, New York, pp. 73-88. Nestler, E.J. and Tallman, J.F. (1987) Chronic morphine treatment increases cyclic AMP-dependent protein kinase activity in the rat locus coeruleus. Mol. Pharmacol., 33: 127- 132. Nestler, E.J., Tenvilliger, R., Erdos, J.J., Duman, R.S. and Tallman, J.F. (1989) Regulation by chronic morphine of G-proteins in the rat locus coeruleus. Brain Res. 476: 230-239. North, R.A., Williams, J.T., Suprenant, A. and Christie, M.J. (1987) Mu and delta receptors belong to a family of receptors that are coupled to potassium channels. Proc. Natl. Acad. Sci. USA, 84: 5487-5491.
Olpe, H-R., Steinmann, M.W., Brugger, F. and Pozza, M.F. (1989) Excitatory amino acid receptors in rat locus coeruleus. An extracellular in uitro study. Naunyn Schmiedebergs Arch. Pharmacol., 339: 312-314. Paxinos, G. and Watson, C. (1982) The Rat Brain in Stereotaxic Coordinates, Academic Press, New York, 237 pp. Pert, C.B., Kuhar, M.J. and Snyder, S.H. (1975) Autoradiographic localization of the opiate receptor in rat brain. Life Sci., 1 6 1849. Rasmussen, K. and Aghajanian, G.K. (1989a) Withdrawal-induced activation of locus coeruleus neurons in opiate-dependent rats: Attenuation by lesions of the nucleus paragigantocellularis. Brain Res., 505: 346-350. Rasmussen, K. and Aghajanian, G.K. (1989b) Failure to block responses of locus coeruleus neurons to somatosensory stimuli by destruction of two major afferent nuclei. Synapse, 4: 162-164. Rasmussen, K. and Jacobs, B.L. (1985) Locus coeruleus unit activity in freely moving cats is increased following systemic morphine administration. Brain Res., 344 240-248. Rasmussen, K., Beitner-Johnson, D.B., Krystal, J.H., Aghajanian, G.K. and Nestler, E.J. (1990a) Opiate withdrawal and the rat locus coeruleus: Behavioral, electrophysiological and biochemical correlates. 1.Neurosci., 10: 2308-2317. Rasmussen, K., Fuller, R.W., Stockton, M.E., Perry, K.W., Swinford, R.M. and Ornstein, P.L. (1991) NMDA receptor antagonists suppress behaviors but not norepinephrine turnover or locus coeruleus unit activity induced by opiate withdrawal. Eur. J. Pharmacol, 117: 9-16. Rasmussen, K., Krystal, J.H. and Aghajanian, G.K. Excitatory amino acids and morphine withdrawal: Differential effects of central and peripheral kynurenic acid administration, Psychopharmacology , in press. Redmond, D.E., Jr. and Krystal, J.H. (1984) Multiple mechanisms of withdrawal from opioid drugs. Ann. Rec. Neurosci., 7: 443-478. Shannon, H.E. and Sawyer, B.D. (1989) Glutamate receptors of the N-methyl-D-aspartate subtype in the myenteric plexus of the guinea pig ileum. J. Pharmacol. Exp. Ther., 251: 518-523. Taylor, J.R., Elsworth, J.D., Garcia, E.J., Grant, S.J., Roth R.H. and Redmond, D.E., Jr. (1988) Clonidine infusion into the locus coeruleus attenuates behavioral and neurochemical changes associated with naloxone-precipitated withdrawal. Psychopharmacology, 96: 121-134. Tempel, A: and Zukin, R.S. (1987) Neuroanatomical patterns of the I, N, and k opioid receptors of rat brain as determined by quantitative in uitro autoradiography. Proc. Natl. Acad. Sci. USA, 84: 4308-4312. Thollander, M., Hellstrom, P.M. and Svensson, T.H. (1989) Suppression of small intestine motility and morphine withdrawal diarrhoea by clonidine: Peripheral site of action. Acta Physiol. Scand., 137: 385-392. Tseng, L.-F., Loh, H.H. and Wei, E.T. (1975) Effects of clonidine on morphine withdrawal signs in the rat. Eur. J. Pharmacol., 30: 93-99. Valentino, R.J. and Wehby, R.G. (1988) Morphine effects on locus coeruleus neurons are dependent on the state of
216 arousal and availability of external stimuli: Studies in anesthetized and unanesthetized rats. J. Phurmacol. Exp. Ther., 244: 1178-1186. Valentino, R.J. and Wehby, R.G. (1989) Locus coeruleus discharge characteristics of morphine-dependent rats: Effects of naltrexone. Bruin Res., 488: 126-134. Wang, Y.-Y. and Aghajanian, G.K. (1987) Excitation of locus coeruleus neurons by an adenosine 3’,5’-cyclic monophos-
phate-activated inward current: Extracellular and intracelM a r studies in rat brain slices. Synapse, 1: 481-487. Zigun, J.R., Bannon, M.J. and Roth, R.H. (1981) Comparison of two a-noradrenergic agonists (clonidine and guanfacine) on norepinephrine turnover in the cortex of rats during morphine abstinence. Eur. J. Phurmacol., 70: 565570.
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Bruin Research, Vol. 88 0 1991 Elcevier Scirnce Publishers B.V.
217 CHAPTER 16
Angiotensin I1 and the locus coeruleus R.C. Speth I
K.L. Grove and B.P. Rowe
Departments of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA, U.S.A. and Department of Physiology, Quillen College of Medicine, East Tennessee State University, Johnson City, TN, U.S.A.
’
The locus coeruleus (LC) is a putative site of action for angiotensin 11 in the brain. Immunocytochemical studies have identified angiotensin II-like immunoreactive material in nerve terminals innervating the LC, and the LC contains one of the highest densities of angiotensin I1 receptor binding sites in the rat brain. Recent studies using selective neurotoxins suggest that the binding sites for angiotensin I1 in the LC are present on noradrenergic perikarya. Angiotensin I1 receptors are now known to exist as two subtypes that are distinguishable both pharmacologically and biochemically. Radioligand binding studies using agonists and antagonists
selective for these angiotensin I1 receptor subtypes indicate that the rat LC contains a mixture of the two known angiotensin I1 receptor subtypes, but that the PD123177-sensitive AII, receptor subtype is predominant. Comparisons of spontaneously hypertensive rats with normotensive rats indicates that angiotensin 11 and its receptors in the L C are elevated in the hypertensive rat strain. Studies of the biochemical and physiological actions of angiotensin I1 in the LC have not yet established an agreed-upon function for angiotensin I1 in this nucleus.
Key words: angiotensin receptors, receptor subtypes, blood pressure, brain renin-angiotensin system
Introduction
Angiotensin I1 (AII) is best known as the vasoconstrictor agent produced in the bloodstream following the release of renin from the kidney under conditions of renovascular stress. The sites of action and sources of A11 extend far beyond the bloodstream, however. This chapter will focus on the brain renin-angiotensin system (BRAS) with special focus on the locus coeruleus (LC). The concept of a BRAS is often attributed to the pioneering studies of Ganten et al. (1971) and Lewicki et al. (1978) who carried out the initial studies on the presence of renin-like enzymes and
the ultimate precursor of AII, angiotensinogen, in the brain. It is now established that tissues other than the brain also generate A11 independent of the systemic renin-angiotensin system (RAS) (Campbell, 1987). Angiotensinergic input to the locus coeruleus
Nerve terminals containing AII-immunoreactive material (AII-ir) have been observed in the LC (Fuxe et al., 1976; Weyhenmeyer and Phillips, 1982), although the densities of the AII-ir fibers were moderate-to-low in comparison to other
218
brain nuclei. No AII-ir containing nerve cell bodies have been observed in the LC (Fuxe et al., 1980; Kilcoyne et al., 1980; Brownfield et al., 1982; Weyhenmeyer and Phillips, 1982). In addition, AII-ir has been observed in micropunches containing the LC using a radioimmunoassay (Meyer et al., 1989). In situ hybridization histochemistry studies have identified angiotensinogen mRNA in the LC (Fuxe et al., 1988) and radioligand binding assays with the angiotensin converting enzyme (ACE) inhibitors 3H-captopril and 1251-351Ahave also identified ACE in this nucleus (Strittmatter et al., 1984; Chai et al., 1987). However, renin has not yet been demonstrated to occur in the LC. The cells of origin of the AII-ircontaining nerve fibers innervating the LC have not been defined, but could arise from AII-ircontaining cell bodies that have been identified in the magnocellular neurons of the paraventricular nucleus of the hypothalamus, the bed nucleus of the stria terminalis, the lateral parabrachial nucleus (LPB), or the solitary tract nucleus (NTS) (Fuxe et al., 1980; Brownfield et al., 1982; Weyhenmeyer and Phillips, 1982; Lind et al., 1985). However, it should be noted that two of these studies (Brownfield et al., 1982; Lind et al., 1985) failed to observe AII-ir-containing nerve terminals in the LC. A n alternative source of angiotensinergic input to the LC is via the ventricular system. A11 has been identified in cerebrospinal fluid (CSF) in rats (Hermann et al., 1982) and the intracerebroventricular (icv) route of administration is an extremely effective means for demonstrating the actions of A11 in the brain. It is likely that A11 is synthesized extracellularly in the brain, since there is little correspondence in the location or cell types in which the various components of the BRAS are found, e.g., A11 has been found in neurons (Fuxe et al., 1980; IQlcoyne et al., 1980; Brownfield et al., 1982; Weyhenmeyer and Phillips, 1982) while angiotensinogen has only been observed in glia (Lynch et al., 1987). An extracellular synthesis of A11 in the brain, with transport of the components via the ventricular
system, is similar to the systemic RAS where the synthetic components arise from different tissues and are circulated via the bloodstream. Located at the floor of the fourth ventricle, the LC should be readily accessible to A11 present in the CSF. In addition, since ACE is present in the LC, angiotensin I, the immediate precursor of AII, which has also been shown to be present in the CSF (Hermann et al., 1982) could access the LC and be converted to AII. The LC is highly vascularized, suggesting that the bloodstream could also be a source of angiotensinergic input. However, A11 does not cross the blood-brain-barrier (BBB), and the BBB is effective at the LC (see below for further discussions of the ability of the BBB to exclude A11 at the LC). Therefore blood-borne A11 is not a stimulus to the LC.
Angiotensin receptor binding in the locus coeruleus
Early studies of the actions of A11 on the brain focused primarily on circumventricular organs that were outside the BBB. With the acceptance of the concept that A11 could be generated in the brain, however, came the related concept that the brain parenchyma could also be a site of action for AII. This idea was firmly entrenched with the radioligand binding studies of Bennett and Snyder (1976) and Sirett et al. (19771, which demonstrated specific 1251-AIIbinding sites in brain regions other than the circumventricular organs. Localization of putative actions of A11 in the brain was further enhanced by the application of in vitro receptor autoradiographic techniques (Gehlert et al., 1984b; Healy and Printz, 1984; Israel et al., 1984; Mendelsohn et al., 1984; Speth et al., 1985; Gehlert et al., 1986a; McKinley et al., 1987), which identified specific nuclei containing putative A11 receptor binding sites. Most recently, in the rat brain, two subtypes of A11 binding sites were identified (Rowe et al., 1990a,b,
219
1991; Speth et al., 1990, 19911, which correspond to the two subtypes of AII receptors described in peripheral tissues (Chiu et al., 1989a,b; Whitebread et al., 1989; Chang and Lotti, 1990; Speth and Kim, 1990). Using two different radioligands, '251-AII (Gehlert et al., 1984a) and '251-sarcosine' A11 (Mendelsohn et al., 1984), two of the early receptor autoradiography studies identified the LC as having putative A11 receptor binding sites. Subsequently, a third antagonist radioligand, 1251sarcosine',isoleucine* AII (1251-SIAII), was also used to identify A11 receptor binding sites in the LC (Speth et al., 1987). Densitometric measurements of 1251-AIIbinding in the rat brain indicated that the LC contained the tenth highest density of binding sites among 53 brain areas surveyed (Gehlert et al., 1986a). A separate study in a different strain of rats indicated that the LC was one of the two nuclei with the highest density of '"I-AII binding sites (Gehlert et al., 1986b). More recently, selective neurotoxins were used to investigate the cellular localization of the AII binding sites in the LC (Rowe et al., 1990~). Unilateral lesioning of the LC with either 6-hydroxydopamine (to selectively lesion catecholaminergic neuronal elements) or ibotenic acid (to selectively lesion neuronal perikarya) caused a reduction in specific '251-SI A11 binding in the lesioned LC. With both neurotoxins the extent of reduction in 1251-SI A11 binding was strongly correlated with the extent of reduction in cerebrocortical norepinephrine (NE) which was used as an indicator of the extent of the lesion of the noradrenergic neurons of the LC. These observations strongly indicate that the A11 receptor binding sites in the LC are located on noradrenergic cell bodies and are not on presynaptic terminals of afferent nerves innervating the LC. Another potential explanation for the presence of A11 receptor binding sites in the LC is that they could be located on the blood vessels that permeate this nucleus. To test this possibility and to further verify the integrity of the BBB in
the LC, we administered '251-SI A11 via cardiac puncture to rats and sacrificed them 2 min later (Rowe and Speth, 1989). No Iz5I-SI A11 binding was observed in the LC, however, abundant '251-SI A11 binding was observed in the BBB-deficient circumventricular organs, e.g., area postrema, subfornical organ, organum vasculosum of the lamina terminalis (OVLT). Thus the 1251-SIA11 binding in the LC is clearly localized to nonvascular, parenchymal elements within the BBB. Until recently, brain A11 receptor binding studies carried out in our laboratory (including the lesion studies described above) and in most other laboratories included a sulfhydryl reducing agent. This practice dates back to the initial studies of brain A11 receptors (Bennett and Snyder, 1976; Sirett et al., 1977), which indicated that the '251-AIIcould be degraded by sulfhydryl reducing agent-sensitive peptidase activity present in the brain. However, we recently discovered that '=ISI A11 receptor binding in some brain areas was diminished in the presence of sulfhydryl reducing agents (Rowe et al., 1990b; Speth et al., 1990, 1991). Subsequent to this finding, we used a nonpeptidic AII receptor subtype-selective antagonist (DuP 753) to further characterize brain A11 receptor binding sites (Rowe ef al., 1990a). Brain areas in which DuP 753 was a potent competitor for '251-SI A11 binding were identical to those in which sulfhydryl reducing agents inhibited 1251-SI A11 binding. Binding of '251-SIA11 to the LC was relatively unaffected by sulfhydryl reducing agents or DuP 753 (5 x l o p 7 M), suggesting that the LC contained the subtype which we have named the AII, subtype (Speth and Kim, 1990). This subtype appears to be the same as the AII-subtype A described by Whitebread et al, (1989) and the AII-2 subtype described by Chiu et al. (1989a), which is also not inhibited by sulfhydryl reducing agents (Chiu et al., 1989b; Whitebread et al., 1989). To further establish the nature of the A11 receptor subtypes in the brain, an additional study (Rowe et al., 1991) was undertaken using a second nonpeptidic A11 receptor antagonist
220
Fig. 1. Pseudocolor representation of the binding of 12'I-SI angiotensin I1 (MI) to a coronal section from a rat brain at approximately 9.7 mm posterior to bregma, based on the atlas of Paxinos and Watson (1986). '251-SI AII binding was carried out as described by Rowe et al. (1990~).The color bar to the right was generated from standards containing known amounts of ''1 (Microscales, Amersham, Arlington Heights, IL). Note that binding in the majority of the lateral parabrachial nucleus (LPB) is M DuP 753, while 1251-SIA11 binding can still be observed in the inferior colliculus and locus coeruleus (LC)at eliminated by 10-5 M D ~ 753. P
(PD123177, also known as EXP655) which is selective for the AII-2 (AII, by our nomenclature) receptor subtype (Chiu et aL, 1989a). In this study, PD123177 M) substantially reduced 1251-AIIbinding in the LC, again suggesting an AII, receptor subtype binding profile. Figure 1 describes the effects of varying concentrations of DuP 753 (10-s-10-4 M) on 1251-SIAII binding in the LC as well as in the inferior colliculus and LPB. "'I-SI A11 binding in the LC, inferior colliculus and a small portion of the LPB was resistant to inhibition by DuP 753 at low concentrations. By contrast, 1251-SIA11 binding in the majority of the LPB was eliminated by a low concentration of DuP 753. The competition curve for DuP 753 versus lZI-SI A11 binding in the LC was biphasic, indicating that there was a heterogeneity of A11 receptor subtypes in the nucleus. A two-site analysis of the competition for 1251-SI A11 binding indicated that 34% of the '251-SI A11
binding was to the AII, subtype (AII-subtype B) by the nomenclature of Whitebread et al. (1989) and AII-1 by the nomenclature of Chiu et al. (1989a), with an IC,, of 3.6 x lo-' M, while 66% of the lZ5I-SIA11 binding was to the AII, subtype, with an IC,, of > M. A similar evaluation of the competition curve for DuP 753 versus lZ5I-SIAII binding in the ventral hippocampus (not shown) revealed a single class of binding sites with an IC50 of 1.9 X l o p 8 M, indicating that this structure contained only the AII, receptor subtype (Rowe et aL, 1991). We have recently characterized an analog of AII, p-aminophenylqlanine6 AII (PNF) as being selective for the AII, receptor subtype (Speth and Kim, 1990). Based on binding assays carried out in the presence and absence of guanosine 5'-0-(3-thriotriphosphate), this analogue appears to be an agonist; however, until a functional response for the AII, receptor subtype can be
221
120110100-0
5 2 g
=
90-
807060-
504030-
6 20-
z
100
I
I
I
1
I
Fig. 2. Competition for lZ5I-SI A11 binding in selected rat brain nuclei by p-aminophenylalanine6 A11 (PNF). The AIIp selective putative agonist PNF (Speth and Kim, 1990) was a more potent competitor for '251-SI A11 binding in the LC and inferior olive (10) than in the solitary tract nucleus (NTS) and LPB. This is consistent with classification of the LC and I 0 as having predominately AIIp receptors and the NTS and LPB as having predominately AII, receptors. Specific binding is defined as lZ5I-SIA11 binding displaced by 1 p M AIL The data represent the average values obtained from two rat brains.
defined, this cannot be proven. Figure 2 indicates that PNF selectively inhibits 1251-SIA11 binding to the LC and inferior olive, nuclei that have been characterized as containing predominately the AII, receptor subtype (Rowe et al., 1990a, 1991), but that it is a weak competitor for 12sI-SI AII binding in the NTS and LPB, nuclei that have been characterized as having predominately the AII, receptor subtype (Rowe et al., 1990a, 1991). Thus the AII-receptor binding studies using selective neurotoxins and A11 receptor subtypeselective ligands indicate that the AII-receptor binding sites on the noradrenergic perikarya in the LC are predominately of the AII, receptor subtype. As noted above, the functional response and the transduction mechanism of this receptor subtype have yet to be established. Possible functions of angiotensin I1 in the locus coeruleus
A number of physiological functions have been ascribed to A11 in the brain; they include dipso-
genesis, activation of the sympathetic nervous system, stimulation and inhibition of hormone release from the pituitary, interference with learning, regulation of body temperature, and analgesia (see review by Speth et al., 1988). At the cellular level of organization, A11 is reported to affect release of several neurotransmitters in the brain, (see review by Speth et al., 1988) including NE the primary transmitter of the LC neurons.
Actions at the cellular level The effects of A11 on brain NE release are generally considered to be stimulatory, in accord with the well-characterized action of A11 at sympathetic nerve terminals to increase the release of NE associated with neuronal firing (Starke, 1977; Zimmerman, 1981). For example, icv A11 increases NE in the CSF (Chevillard et al., 1979), and AII stimulates NE release from brain slices (Garcia-Sevilla et al., 1979; Meldrum et al., 1984; Schacht, 1984). The dipsogenic, pressor and pituitary hormone releasing actions of icv A11 are diminished by agents that deplete or damage NE-containing neurons or block adrenergic receptors (Smookler et al., 1966; Severs et al., 1971; Camacho and Phillips, 1981; Jones, 1984; Hoffman et al., 1977a; Gordon et al., 1979; Walters and Speth, 1989; Steele and Ganong, 1986). There may be regional differences in the actions of A11 on brain NE release and utilization. When administered icv, renin increased norepinephrine turnover in the dorsomedial hypothalamus, but decreased NE turnover in the paraventricular nucleus of the hypothalamus (Fuxe et al., 1980). Brain region-specific effects of A11 on NE release or utilization have been reported by others as well (Sumners and Phillips, 1983; Huang et al., 1987). The effects of A11 on NE in the LC are controversial. Sumners and Phillips (1983) reported that icv A11 increased norepinephrine utilization in the LC of rats, while Huang et al. (1987) observed no effect of A11 on NE release in slices of rat brain LC. Huang et al. (1987) did, however, observe a stimulation of NE release from slices of
222
rat brain cerebral cortex in response to AII. Since the overwhelming majority of noradrenergic fibers in the cerebral cortex arise from the LC, it is surprising that their perikarya do not express AII receptors mediating a similar function. Based on the observations of Huang et al. (1987), it is possible that the effects of icv A11 on NE utilization in the LC seen by Sumners and Phillips (1983) are indirect, affecting brain nuclei that project to and influence the activity of the LC, or that the decreases in NE levels seen in the LC by Sumners and Phillips (1983) reflect a stimulation of transport of NE from the nerve cell body to the nerve terminals. There is little known about the transduction mechanisms of the A11 receptors in the brain, including those of the LC. While A11 receptors in peripheral tissues act via second messenger systems, stimulating phosphoinositide hydrolysis or inhibiting adenylate cyclase activity (Garcia-Sainz, 1987), such actions have not been demonstrated in the brain. Studies from our laboratory indicate that A11 decreases phosphoinositide hydrolysis in the diencephalon-midbrain (Tamura and Speth, 1990) and that the dipsogenic and pressor actions of A11 are mediated by a mechanism that is not inactivated by pertussis toxin (Speth and Grove, 1991). Huang et al. (1987) demonstrated that A11 did not affect phosphoinositide hydrolysis in either the LC or cerebral cortex of the rat. Thus, there is little evidence for a biochemical transduction for the neuronal actions of A11 that is similar to its effects in peripheral tissues. Another possible mechanism for mediating the actions of A11 is through a gated ion channel. In support of this concept, the AII, receptor subtype, which predominates in the LC, is not linked to a guanosine triphosphate binding protein in a PC12 cell line (Speth and Kim, 1990). Microiontophoretically applied A11 generally has excitatory effects on spontaneous firing rates of neurons in responsive areas of the brain, although inhibitory effects of A11 on neuronal firing rates have also been observed (see review by Felix et al., 1988). The response latencies to microion-
tophoretically applied A11 are generally long and likely indicate a metabotropic (biochemical mediation) rather than an ionotropic (ion permeability) effect. There is, however, one report of short latency ( < 1 sec) as well as longer latency responses to A11 administered into the LC (Anissimov et al., 1980). These authors distinguished six different types of neuronal responses - based on latency, whether the response was excitatory, inhibitory, or biphasically excitatory, and inhibitory - to A11 in the LC of conscious rabbits. These observations suggest that A11 may have complex actions at the LC, possibly acting as a neuromodulator of the responses to various synaptic transmitters released from afferents to the LC. Physiological effects There is such a broad range of LC functions that it is difficult to isolate a specific function that could be ascribed to A11 in this nucleus. However, of the known functions of AII in the brain the one that is most likely to involve the LC is blood pressure regulation. Several studies suggest that there is an interaction between AII and the LC in blood pressure regulation. Spontaneously hypertensive rats (SHR) show an elevated pressor responsivity to icv A11 (Hoffman et al., 197%). Consistent with this greater sensitivity to AII, several studies have demonstrated a selective enhancement of A11 receptor binding in cardiovascular regulatory centers of the SHR brain (Stamler ef al., 1980; Plunkett and Saavedra, 1985; Gehlert et al., 1986b; Saavedra et al., 1986; Gutkind et al., 19881, including the LC (Gehlert et al., 1986b). In the SHR, AII antagonists administered icv decrease blood pressure, in contrast to normotensive rats that show no change in blood pressure (Phillips et al., 1979). This suggests that increased AII receptor stimulation is a causal factor in the hypertension. Consistent with this hypothesis, Meyer et al. (1989) observed an increase in the amount of AII-ir in selected brain nuclei including the LC in the SHR. Berecek ef al. (1984) observed a small pressor response to
223
A11 microinjected into the LC of normotensive rats, although this effect was substantially less than the pressor response to an equal amount of vasopressin. Since Berecek et al. (1984) did not study the SHR, it could be hypothesized that the SHR would have shown a larger pressor response to A11 microinjections into the LC.
to restore fluid homeostasis. The additional stimulation to the LC may add to the state of arousal of the nervous system of the animal increasing the likelihood that appropriate means to correct the problem are instituted.
Conclusions
A committee of the Council for High Blood Pressure Research of the American Heart Association (Bumpus et al. (1991) Hypertension, 17: 720721) recently recommended a new nomenclature for angiotensin receptor subtypes. The AII, and AII, subtypes described herein correspond to the AT, and AT, subtypes of Bumpus et al., respectively.
The LC contains a high density of A11 binding sites, receives angiotensinergic input and shows moderate, but variable, responsivity to AII. While the effects of A11 on the LC most likely are related to blood pressure regulation this is by no means an established fact. To add to the confusion regarding the function of A11 in the LC, the predominant A11 receptor subtype in the LC, the AII, subtype has not been demonstrated to mediate a functional response, and should by convention (Green, 1990; Watson and Abbott, 1990) be referred to as an undefined receptor or as an acceptor until a functional response can be demonstrated. However, since the LC shows both excitatory and inhibitory responses to AII, it could be hypothesized that one subtype mediates the excitatory actions of AII while the other mediates the inhibitory actions of AII. Since the inhibitory actions of A11 appear to predominate (Anissimov et al., 1980), this may indicate that the predominating AII, subtype mediates the inhibitory actions of A11 in the brain, and that the less abundant AII, subtype mediates the excitatory actions of AII. In line with the overall function of A11 to maintain body fluid and electrolyte homeostasis, we offer the following hypothetical function for AII at the LC. A major function of the LC is to integrate sensory information and generate arousal of the animal. It is possible that A11 conveys to the LC sensory information regarding the blood pressure status of the animal. In hypotension there may be increased angiotensinergic input to the LC as well as to autonomic nervous system control centers as RASs attempts
Note added in proof
Acknowledgements
The authors thank Michelle Carter and David Saylor for technical assistance. The work reported was supported by USPHS (NS-24388, NS21305, and BRSG SO7 RR05465) and the American Heart Association, Tennessee Affiliate. References Anissimov, J.Z., Deleva, J.I., Nicolov, N.A., Sudakov, K.V. and Sherstnev, V.V. (1980) Neuronal sensitivity of some brain regions of angiotensin I1 in rabbits. Agressologie, 21: 171-175. Bennett, J.P., Jr. and Snyder, S.H. (1976) Angiotensin I1 binding to mammalian brain membranes. J. Biol. Chem., 251: 7423-7430. Berecek, K.H., Olpe, H.R., Jones, R.S.G. and Hofbauer, K.G. (1984) Microinjection of vasopressin into the locus coeruleus of conscious rats. Am. J. Physiol., 247: H675H681. Brownfield, M.S., Reid, LA., Ganten, D. and Ganong, W.F. (1982) Differential distribution of immunoreactive angiotensin and angiotensin converting enzyme in rat brain. Neuroscience, 7 1759-1769. Camacho, A. and Phillips, M.I. (1981) Separation of drinking and pressor responses to central angiotensin by monoamines. Am. J. Physiol., 240: R106-Rl13. Campbell, D.J. (1987) Circulating and tissue angiotensin systems. f. Clin. Invest., 79: 1-6. Chai, S.Y., Mendelsohn, F.A.O. and Paxinos, G. (1987) Angiotensin converting enzyme in rat brain visualized by quantitative in vitro autoradiography. Neuroscience, 20: 615-627.
224 Chang, R.S.L. and Lotti, V.J. (1990) Two distinct angiotensin I1 receptor binding sites in rat adrenal revealed by new selective nonpeptide ligands. Mol. Pharmacol., 29: 347351. Chevillard, C., Duchene, N., Pasquier, R. and Alexandre, J.-M. (1979) Relation of the centrally evoked pressor effect of angiotensin I1 to central noradrenaline in the rabbit. Eur. J. Pharmacol., 58: 203-206. Cbiu, A.T., Herblin, W.F., McCall, D.E., Ardecky, R.J., Carini, D.J., Duncia, J.V., Pease, L.J., Wong, P.C., Wexler, R.R., Johnson, A.L. and Timmermans, P.B.M.W.M. (1989a) Identification of angiotensin I1 receptor subtypes. Biochem. Biophys. Res. Commun., 165: 196-203. Chiu, A.T., McCall, D.E., Nguyen, T.T., Carini, D.J., Duncia, J.V., Herblin, W.F., Uyeda, R.T., Wong, P.C., Wexler, R.R., Johnson, A.L. and Timmermans, P.B.M.W.M. (1989b) Discrimination of angiotensin I1 receptor subtypes by dithiothreitol. Eur. J. Pharmacol., 170: 117-118. Felix, D., Imboden, H., Schelling, P. and Harding, J.W. (1988) Electrophysiological assessment of central angiotensin function. In J.W. Harding, J.W. Wright, R.C. Speth and C.D. Barnes (Eds.), Angiotensin and Blood Pressure Regulation, Academic Press, San Diego, pp. 165-208. Fuxe, K., Ganten, D., Hokfelt, T. and Bolme, P. (1976) Immunohistochemical evidence for the existence of angiotensin 11-containing nerve terminals in the brain and spinal cord of the rat. Neurosci. Lett., 2: 229-234. Fuxe, K., Anderson, K., Ganten, D., Hokfelt, T. and Eneroth, P. (1980) Evidence for the existence of an angiotensin 11-like immunoreactive central neuron system and its interactions with the central catecholamine pathways. In F. Gross and G. Vogel (Eds.), Enzymatic Release of Vasoactiue Peptides, Raven Press, N.Y., pp. 161-168. Fuxe, K., Bunnemann, B., Aronsson, M., Tinner, B., Cintra, A., Von Euler, G., Agnati, L.F., Nakanishi, S., Ohkubo, H. and Ganten, D. (1988) Pre- and postsynaptic features of the central angiotensin systems. Indications for a role of angiotensin peptides in volume transmission and for interactions with central monoamine neurons. Clin. Exp. Hypertens. [A], 10, Suppl. 1: 143-168. Ganten, D., Minnich, J.L., Granger, P., Hayduk, K., Brecht, H.M., Barbeau, A., Boucher, R. and Genest, J. (1971) Angiotensin-forming enzyme in brain tissue. Science, 173: 64-65. Garcia-Sainz, J.A. (1987) Angiotensin 11 receptors: One type coupled to two signals or receptor subtypes?. TIPS, 8: 48-49. Garcia-Sevilla, J.A., Dubocovich, M.L. and Langer, S.Z. (1979) Angiotensin I1 facilitates the potassium-evoked release of 3H-noradrenaline from the rabbit hypothalamus. Eur. J. Pharmacol., 56: 173-176. Gehlert, D.R., Speth, R.C., Healy, D.P. and Wamsley, J.K. (1984a) Autoradiographic localization of angiotensin I1 receptors in the rat brainstem. Life Sci., 34: 1565-1571. Gehlert, D.R., Speth, R.C. and Wamsley, J.K. (1984b) Autoradiographic localization of angiotensin I1 receptors in the rat brain and kidney. Eur. J. Pharmacol., 98: 145-146. Gehlert, D.R., Speth, R.C. and Wamsley, J.K. (1986a) Distri-
bution of ['251]angiotensin I1 binding sites in the rat brain: A quantitative autoradiographic study. Neuroscience, 18: 837-856. Gehlert, D.R., Speth, R.C. and Wamsley, J.K. (1986b) Quantitative autoradiography of angiotensin I1 receptors in the SHR brain. Peptides, 7 1021-1027. Gordon, F.J., Brody, M.J., Fink, G.D., Buggy, J. and Johnson, A.K. (1979) Role of central catecholamines in the control of blood pressure and drinking behavior. Brain Res., 178: 161-173. Green, J.P. (1990) Pharmacological receptors: The need for a compendium of classification, nomenclature and structure. TIPS, 11: 13-16. Gutkind, J.S., Kuribara, M., Castren, E. and Saavedra, J.M. (1988) Increased concentration of angiotensin I1 binding sites in selected brain areas of spontaneously hypertensive rats. J. Hyperten., 6: 79-84. Healy, D.P. and Printz, M.P. (1984) Localization of angiotensin I1 binding sites in rat septum by autoradiography. Neurosci. Lett., 44: 167-172. Hermann, K., Ganten, D., Bayer, C., Unger, T., Lang, R. and Rascher, W. (1982) Definite evidence for the presence of [Ile,]-angiotensin I and [Ile,]-angiotensin I1 in the brain of rats. In D. Ganten, M. Printz, M.I. Phillips and B.A. Scholkens (Eds.), The Renin Angiotensin System in the Brain, Springer-Verlag, Berlin, pp. 192-207. Hoffman, W.E., Phillips, M.I. and Schmid, P. (1977a) The role of catecholamines in central antidiuretic and pressor mechanisms. Neuropharmacology, 16: 563-569. Hoffman, W.E., Phillips, M.I. and Schmid, P.G. (1977b) Central angiotensin 11-induced responses in spontaneously hypertensive rats. Am. J. Physiol., 234: H426-H433. Huang, Y., Rogers, J. and Henderson, G. (1987) Effects of angiotensin I1 on [3H]noradrenaline release and phosphatidylinositol hydrolysis in the parietal cortex and locus coeruleus of the rat. J. Neurochem., 49: 1541-1549. Israel, A,, Correa, F.M., Niwa, M. and Saavedra, J.M. (1984) Quantitative determination of angiotensin I1 binding sites in rat brain and pituitary gland by autoradiography. Brain Rex, 322: 341-345. Jones, D.L. (1984) Injections of phentolamine into the anterior hypothalamus-preoptic area of rats blocks both pressor and drinking responses produced by central administration of angiotensin 11. Brain Res. Bull., 13: 127-133. Kilcoyne, M.M., Hoffman, D.L. and Zimmerman, E.A. (1980) Immunocytochemical localization of angiotensin I1 and vasopressin in rat hypothalamus: Evidence for production in the same neuron. Clin. Sci., 59: 57s-60s. Lewicki, J.A., Fallon, J.H. and Printz, M.P. (1978) Regional distribution of angiotensinogen in rat brain. Brain Res., 158: 359-371. Lind, R.W., Swanson, L.W. and Ganten, G. (1985) Organization of angiotensin I1 immunoreactive cells and fibers in the rat central newous system. Neuroendocrinology, 40: 2-24. Lynch, K.R., Hawelu-Johnson, C.L. and Guyenet, P.G. (1987) Localization of brain angiotensinogen mRNA by hybridization histochemistry. Mol. Brain Res., 2: 149-158,
225 McKinley, M.J., Allen, A.M., Clevers, J., Paxinos, G. and Mendelsohn, F.A. (1987) Angiotensin receptor binding in human hypothalamus: Autoradiographic localization. Brain. Res., 420: 375-379. Meldrum, M.J., Xue, C.-S., Badino, L. and Westfall, T.C. (1984) Angiotensin facilitation of noradrenergic neurotransmission in central tissues of the rat: Effects of sodium restriction. J. Cardiovasc. Pharmacol., 6: 989-995. Mendelsohn, F.A.O., Quirion, R., Saavedra, J.M., Aguilera, G. and Catt, K.J. (1984) Autoradiographic localization of angiotensin I1 receptors in rat brain. Proc. Natl. Acad. Sci. USA, 81: 1575-1579. Meyer, J.M., Felten, D.L. and Weyhenmeyer, J.A. (1989) Levels of immunoreactive angiotensin I1 in microdissected nuclei from adult WKY and SH rat brain. Clin. Exp. Hypertens. [A], 11: 103-117. Paxinos, G. and Watson, C. (1986). The Rat Brain in Stereotaxic Coordinates, 2nd edn., Academic Press, Orlando, 237 PP. Phillips, M.I., Weyhenmeyer, J., Felix, D., Ganten, D. and Hoffman, W.E. (1979) Evidence for an endogenous brain renin-angiotensin system. Fed. Proc., 38: 2260-2266. Plunkett, L.M. and Saavedra, J.M. (1985) Increased angiotension I1 binding affinity in the nucleus tractus solitarius of spontaneously hypertensive rats. Proc. Natl. Acad. Sci. USA, 82: 7721-7724. Rowe, B.P. and Speth, R.C. (1989) Radiolabeling of the rat CNS by blood-borne 1251-Sarl,Ile8 angiotensin I1 (Iz5ISIAII). SOC.Neurosci. Abstr., 15: 230. Rowe, B.P., Grove, K.L., Saylor, D.L. and Speth, R.C. (1990a) Angiotensin I1 receptor subtypes in the rat brain. Eur. J. Pharmacol., 186: 339-342. Rowe, B.P., Grove, K.L. and Speth, R.C. (1990b) Differential effects of mercaptoethanol on angiotensin receptor binding at multiple brainstem sites. FASEB J., 4: A600. Rowe, B.P., Kalivas, P.W. and Speth, R.C. (1990~)Autoradiographic localization of angiotensin I1 receptor binding sites on noradrenergic neurons of the locus coeruleus. J. Neurochem., 55: 533-540. Rowe, B.P., Grove, K.L., Saylor, D.L. and Speth, R.C. (1991) Discrimination of angiotensin I1 receptor subtype distribution in the rat brain using non-peptidic receptor antagonists. Regul. Pept., 33: 45-53. Saavedra, J.M., Correa, F.M.A., Plunkett, L.M., Israel, A., Kurihara, M. and Shigematsu, K. (1986) Binding of angiotensin and atrial natriuretic peptide in brain of hypertensive rats. Nature (London), 320: 758-760. Schacht, U. (1984) Effects of angiotensin I1 and ace inhibitors on electrically stimulated noradrenaline release from superfused rat brain slices. Clin. Exp. Hypertens. [A], 6: 1847-1851. Severs, W.B., Summy-Long, J., Daniels-Severs, A. and Connor, J.D. (1971) Influence of adrenergic blocking drugs on central angiotensin effects. Pharmacology, 5: 205-214. Sirett, N.E., McLean, AS., Bray, J.J. and Hubbard, J.I. (1977) Distribution of angiotensin I1 receptors in rat brain. Brain Res., 122: 299-312. Smookler, H.H., Severs, W.B., Kinnard, W.J. and Buckley,
J.P. (1966) Centrally mediated cardiovascular effects of angiotensin 11. J. Pharmacol. Exp. Ther., 153: 485-494. Speth, R.C. and Kim, K.H. (1990) Discrimination of two angiotensin I1 receptor subtypes with a selective analogue of angiotensin 11, p-aminophenylalanine6 angiotensin 11. Biochem. Biophys. Res. Comm., 169: 997-1006. Speth, R.C. and Grove, K.L. (1991) Pertusis toxin blocks the dipsogenic actions of carbachol, but does not block the dipsogenic and pressor actions of angiotensin 11. Regulat. Pept., 3 2 121-128. Speth, R.C., Wamsley, J.K., Gehlert, D.R., Chernicky, C.L., Barnes, K.L. and Ferrario, C.M. (1985) Angiotensin I1 receptor localization in the canine CNS. Brain Res., 326: 137-143. Speth, R.C., Wamsley, J.K., Wright, J.W., Abhold, R.H. and Harding, J.W. (1987) Characterization of putative sites of action of angiotensin I1 in the rat central nervous system by in uitro receptor autoradiography with 1251-sarcosine, isoleucines angiotensin 11. In J.P. Buckley and C.M. Ferrario (Eds.), Brain Peptides and Catecholamines in Cardiovascular Regulation, Raven Press, New York, pp. 257-272. Speth, R.C., Wright, J.W. and Harding, J.W. (1988) Brain angiotensin receptors: Comparison of location and function. In J.W. Harding, J.W. Wright, R.C. Speth and C.D. Barnes (Eds.), Angiotensin and Blood Pressure Reguiation, Academic Press, New York, NY,pp. 1-34. Speth, R.C., Grove, K.L., Carter, M.R. and Rowe, B.P. (1990) Differential effects of mercaptoethanol on angiotensin I1 receptor binding at multiple forebrain sites. FASEB J., 4 A600. Speth, R.C., Rowe, B.P., Grove, K.L., Carter, M.R. and Saylor, D. (1991) Sulfhydryl reducing agents distinguish two subtypes of angiotensin I1 receptors in the rat brain. Brain Res., 548: 1-8. Stamler, J.F., Raizada, M.K., Fellows, R.E. and Phillips, M.I. (1980) Increased specific binding of angiotensin I1 in the organum vasculosum of the laminae terminalis area of the spontaneously hypertensive rat brain. Neurosci. Lett., 17: 173-177. Starke, K. (1977) Regulation of noradrenaline release by presynaptic receptor systems. Rev. Physiol. Biochem. Pharmacol., 77: 1-124. Steele, M.K. and Ganong, W.F. (1986) Effects of catecholamine-depleting agents and receptor blockers on basal and angiotensin 11- or norepinephrine-stimulated luteinizing hormone release in female rats. Endocrinology, 119: 2728-2736. Strittmatter, S.M., Lo, M.M.S., Javitch, J.A. and Snyder, S.H. (1984) Autoradiographic visualization of angiotensin-converting enzyme in rat brain with [3H]captopril: Localization to a striatonigral pathway. Proc. Natl. Acad. Sci. USA, 81: 1599-1603. Sumners, C. and Phillips, M.I. (1983) Central injection of angiotensin I1 alters catecholamine activity in rat brain. Am. J. Physiol., 244: R257-R263. Tamura, C.S. and Speth, R.C. (1990) Effects of angiotensin I1 on phosphatidylinositol hydrolysis in rat brain. Neurochem. Znt., in press.
226
Walters, D.E. and Speth, R.C. (1989) Monoamine depletion does not alter angiotensin I1 binding sites in the rat brain. Brain Res. Bull.,22: 283-288. Watson, S. and Abbott, A. (1990) Receptor nomenclature supplement. TIPS, 11 (Suppl): 1-30. Weyhenmeyer, J.A. and Phillips, M.I. (1982) Angiotensin-like immunoreactivity in the brain of the spontaneously hypertensive rat. Hypertension, 4: 514-523.
Whitebread, S., Mele, M., Kamber, B. and de Gasparo, M. (1989) Preliminary biochemical characterization of two angiotensin I1 receptor subtypes. Biochem. Biophys. Res. Commun.,163: 284-291. Zimmerman, B.G. (1981) Adrenergic facilitation by angiotensin: Does it serve a physiological function? Clin.Sci., 6 0 343-348.
C.D. Barnes and 0. Pompeiano (Eda.) Prop-css in Brain Resrurch, Vol. 88 0 1991 Elsevier Science Publishers B.V.
227 CHAPTER 17
Vasopressin immunoreactive fibers and neurons in the dorsal pontine tegmentum of the rat, monkey and human A.R. Caff6, J.C. Holstege and F.W. van Leeuwen
'
Erasmus Unicersity Rotterdam, Faculty of Medicine, Department of Anatomy, Rotterdam, The Netherlands and I Netherlands Institute for Brain Research, Meibergdreef;Amsterdam, The Netherlands
It is now well established that extensive extrahypothalamic vasopressin (VP) systems exist in the rat, monkey and human brain. There are marked differences between species, but in each case VP nuclei provide dense afferents to the dorsal pontine tegmentum. Here VP may play a role in the mechanisms exerted by the locus coeruleus (LC) neurons, possibly both as a neurotransmitter and as a neuromodulator. Although we are aware of some properties of V P systems, e.g., gonadal steroid dependency in the rat, major gaps characterize our knowledge of its anatomy. With regard to the interaction of VP with the LC in the brainstem of mammals
~
~~
~______
some of the questions which stand out are: (1) Is VP really being biosynthesized and transported by LC cells and, if not, what is its function within these cells? (2) Is there a structural difference between male and female LC neurons in the rat as a consequence of the sex-dimorphic VP innervation? (3) What is the origin of VP afferents in the dorsal pontine tegmentum of the (nonlhuman primate and are these afferents also controlled by gonadal steroids? Research strategies to answer these questions will provide us with information to resolve some of the current inconsistencies about the anatomy and the function of the V P and LC systems in the brain.
______
~
~
Key wordr vasopressin fibers, vasopressin neurons, locus coeruleus, dorsal pontine tegmentum, rat, monkey, human
Introduction
Vasopressin (VP) is a nanopeptide which is synthesized in the brain as part of the precursor protein propressophysin. Propressophysin is broken up into fragments during axonal transport. Apart from VP, these fragments include neurophysin and a C-terminal glycoprotein (for review see Gainer, 1983). While the chemical nature of VP was elucidated during the fifties (Du Vignaud et al., 1954), this neuropeptide was localized in the magnocellular hypothalamo-neurohypophyseal system (HNS) only at the start of the seven-
ties. Using immunocytochemical methods, Swaab et al. (197.51, and others (for review see Livett, 19781, reported the presence of VP in the paraventricular (PVN) and supraoptic hypothalamic nuclei. VP transported by the HNS is released into the blood as a hormone which acts on distant peripheral receptors (Jard, 1983). Soon it was discovered that not only cells in the HNS synthesized VP, but that VP neurons also existed in the suprachiasmatic nucleus (SCN) (e.g., VandeSande et al., 1975). The occurrence of VP cells in the HNS and SCN is a constant feature throughout mammalian species.
228
The first papers on identified extrahypothalamic VP fibers were published in the late seventies (for review see Buijs et al., 1983). The view that extrahypothalamic fibers release VP synaptically as a neurotransmitter is now supported by much data (Buijs and Van Heerikhuijze, 1982; for review see Van Leeuwen, 1987). Up until 1982 it was assumed that, except for the connections of the PVN to the medulla oblongata and spinal cord (Sofroniew and Schrell, 19821, the SCN was the origin of the extrahypothalamic VP fibers. However, using more sensitive detection methods combined with the axonal transport inhibitor agent colchicine, additional VP cell groups were revealed in the rat brain. These comprise the bed nucleus of the stria terminalis (BST), medial amygdala, dorsomedial hypothalamic nucleus, and locus coeruleus (LC) (Caffk and Van Leeuwen, 1983; Van Leeuwen and Caffk, 1983; Sofroniew, 1985). Some of these additional VP-containing nuclei were also described in other species, e.g., the presence of VP in the BST in man (Fliers et ul., 1986) and monkey (Caffk et a]., 1989). The LC, in the rat and nonhuman primate is a compact nucleus in the floor of the fourth ventricle. It is by far the most important noradrenergic center in the brain of mammals, containing about one-half of the total number of noradrenergic neurons (for review see Foote et al., 1983). Anterograde tracing with Phaseolus vulgaris leucoagglutinin confirmed the projections of the LC to a wide range of brain sites and revealed previously unrecognized patterns and density of noradrenergic innervation, especially in the brainstem and spinal cord (Grzanna and Fritschy, this volume). It has been shown that the LC is composed of different cell groups, each with their specific efferent connections (Loughlin et al., 1986a,b). In addition, parts of LC cells exhibit distinct neurochemical properties with respect to the coexistence of neuropeptides and noradrenaline (see Sutin and Jacobowitz, this volume) and the liability of their axons to DSP-4 toxicity (Grzanna and Fritschy, 1989). This intrinsic diversity, combined with differential inputs of afferent
.
systems, might account for the contradictory roles postulated for this nucleus (see for review Aston-Jones et al., 1984). There are indications for an interaction between the vasopressinergic system and the noradrenergic LC on different levels of the central nervous system of the rat. Microinjection of VP or its analogues in some terminal regions of LC fibers modulates noradrenaline metabolism (Versteeg, 1983). VP-induced behavioral changes are modified by selective lesions of the dorsal noradrenergic bundle (Kovacs et al., 1979). Furthermore, VP microiontophoresed into the LC activates local neurons (Olpe and Baltzer, 1981). For a better understanding of these mechanisms, a detailed knowledge of the VP cells within and VP innervation of the LC at the light and electron microscopic levels is required. In this chapter the origin and distribution of VP-containing fibers and cells in the dorsal pontine tegmentum of the rat, monkey and human are described. Methods
The results presented here were obtained in adult male and female albino Wistar rats and in ten adult male and female Cynomolgus monkeys (Mucuca fascicularis). Human tissue was obtained from post-mortem material. For light microscopy immunocytochemically stained VP structures in 80 p m vibratome coronal sections were analyzed. Some of the rats were treated intraventricularly with colchicine for 48 h. Others received focal thermal or neurotoxic (6-hydroxydopamine) lesions within the LC. For the immunocytochemistry, antisera raised against each of the three parts of the propressophysin molecule were used. In general, animals were perfused with 4% paraformaldehyde in 0.1 M sodium cacodylate, pH 7.4, or with 5% glutaraldehyde in the same buffer. Details of the experimental and light microscopic protocols and data on the specificity of the antisera used have been published elsewhere
229
(see Caff6 and Van Leeuwen, 1983; Fliers et al., 1986; Caffk et at., 1989). For electron microscopy three adult male, albino Wistar rats were used. They were deeply
anesthetized with pentobarbital (60 mg/kg) and transcardially perfused with saline followed by a mixture of 3% glutaraldehyde and 4% paraformaldehyde in 0.1 M sodium cacodylate buffer
1 Fig. 1. Series of transverse drawings of the rat brainstem (according to Paxinos and Watson, 1986). A survey of vasopressin (#W1) or glycoprotein (#C3 final)-immunoreactive structures is shown on the right side of the drawings. Fibers and neurons are represented by line drawings and black dots, respectively. For nomenclature of brainstem structures refer to Paxinos and Watson.
230
(pH 7.35). After postfixation for 2 h in the same fixative, 80 p m coronal vibratome sections of the LC region were cut and stained immunocytochemically with the PAP method using an antibody directed against rat neurophysin (#R4) or rat glycoprotein (#C3 final; see Van Leeuwen et
al., 1989). The sections were then postfixed for 90 min in 1.5% osmium tetroxide, chemically dehydrated with di-methoxy-propane (Muller and Jacks, 1975) and flat embedded in Araldite. After polymerization, semi-thin sections were cut from the plastic blocks and examined in the light mi-
F
16
1 Fig. 1 (continued).
231
croscope. The blocks which showed the highest amount of immunoreactivity in the LC were trimmed to pyramids. Ultrathin sections were cut, some of which were contrasted with uranyl acetate and lead citrate. The sections were analyzed in a Philips EM 300 electron microscope. Vasopressinergic innervation of the pontine tegmentum
Vasopressinergic innervation in the rat For the rat, VP afferents will be described in the pontine tegmentum starting from level Bregma -6.72 mm according to the atlas of Paxinos and Watson (1986). Their distribution is schematically indicated in a series of transverse drawings (Fig. 1). Because of the plane of sectioning, the dorsal part of the coronal sections from which the present information is derived shows more rostral levels of the brainstem compared with the ventral part. A moderate number of thin, beaded, VP fibers is present in the dorsal central grey area in the rostral pontine tegmentum at the level of the oculomotor complex of the rat (Fig. 1A). In this region the VP fibers are more or less orientated in a medial to lateral direction coursing in a distinct fiber tract at the border of the dorsal and lateral parts of the central grey. The fibers pass ventrally and dorsally to the dorsal longitudinal fascicle. The region in between these fiber tracts is only sparsely innervated. The density of VP fibers in the medial central grey is similar to the VP fiber tracts in the dorsal central grey, but these fibers are orientated parallel to the ventrodorsal axis of the aqueduct. A number of these thin, beaded VP fibers can be seen to penetrate the ependymal layer of the aqueduct. The lateral part of the central grey area of the rostral pontine tegmentum is less densely innervated by VP fibers than the dorsal part. Horizontally oriented fibers in this region pass between the large neurons of the nucleus of the mesencephalic trigeminal tract and the medial central grey. Only a few VP fibers occur lateral to the central grey in the
reticular formation of the rostral pontine tegmentum. The Edinger-Westphal nucleus and the region dorsal to the supraoculomotor gray contain large numbers of VP fibers. They merely seem to traverse the former nucleus in the midline of the brain to aggregate in the latter region. They distinctly avoid both the large and small cells of the oculomotor nuclei. This distribution is largely consistent with data reported by others (DuboisDauphin and Zakarian, 1987). Some scattered VP fibers do exist in the supra-oculomotor grey. Dorsolateral to the oculomotor complex, a moderately dense VP fiber tract can be observed to traverse the ventral part, of the deep mesencephalic nuclei ventrolaterally in the direction of the pedunculopontine tegmental nucleus. Here these fibers merge with fibers coursing along the lateral margin of the brainstem. More caudally, at the level of the trochlear nuclear complex, the pattern of VP fibers in the dorsal central grey is somewhat different from more rostral regions (Fig. 1B). Here the fibers are no longer distributed in distinct tracts, but they are scattered over the medial part of this region. The number of VP fibers in lateral, central and medial parts of the central grey and in the lateral tegmental areas is comparable to levels rostral to the trochlear nuclei. In addition, densely stained fibers on both sides of the midline of the ventral tegmentum are dispersed in the shape of a V (see De Vries et al., 1985). No VP fibers are observed to enter the trochlear nuclei. Caudal to the trochlear nuclei (Bregma - 7.80, according to Paxinos and Watson, 1986; Fig. 1C) the density of VP fibers dorsal to the aqueduct is substantially reduced as compared with more rostral levels. In contrast, the laterodorsal and lateroventral central grey areas contain many more VP fibers than rostrally. Dense innervation occurs in the medial central grey, while at this level the entire dorsal raphe displays an abundance of VP afferents. This type of VP innervation in the dorsal pontine tegmentum is maintained approximately up to the level of the rhabdoid nucleus (Fig. lD/E). Here the number of VP fibers in
232
233
the dorsal and central parts of the central grey is reduced. However, from the more ventral parts of the central grey a densely packed VP tract courses ventrolaterally between the pericentral dorsal tegmental nucleus and the cuneiform nucleus to the lateral border of the brainstem. A similarly dense VP innervation as in the dorsal raphe can also be observed at the level of the anterior LC (bregma -9.16 to -9.30, Paxinos and Watson, 1986). In this area the VP innervation is present at the ventricular margin of the periventricular tegmentum (Fig. 1F-HI. In addition, a moderate number of fibers exists in the dendritic zone of the LC which surrounds the cytoarchitectonically defined nucleus containing its perikarya (Fig. 2A). Furthermore, some scattered VP fibers traverse the entire periventricular tegmentum, except for the dorsal tegmental nucleus, which is almost devoid of VP innervation. The most dense innervation in the tegmentum at this level, however, is an area ventral to the dorsal and laterodorsal tegmental nuclei at the border with the medial longitudinal fascicle (see De Vries and Buijs, 1983). Lateral to the periventricular tegmentum, a dense VP innervation is found in the medial and lateral parabrachial nuclei. The pattern of innervation in the posterior part of the LC is similar to the anterior part. However, the number of VP fibers markedly decreases towards the posterior pole of this nucleus.
Vasopressinergic innervation in the monkey For the monkey, VP afferents will be described in the dorsal pontine tegmentum starting from level A 5.5 according to the atlas from Shantha el al. (1968). They are schematically indicated in a series of transverse drawings (Fig. 3).
Fig. 3. Series of traverse drawings of the monkey brainstem (From Shantha et al., 1968). A survey of glycoprotein (#K.1.7)-immunoreactive fibers and neurons is shown on the left side of the drawings. For nomenclature of the brainstem structures (refer t o Shantha et al.)
At the mesencephalic-pontine junction, scattered VP fibers are present (Fig. 3A). However, a dense VP innervation is present in rostra1 parts of the dorsal central grey i.e., the region around the dorsal longitudinal fascicle. Similar to the rat, the fibers are mainly grouped in a tract coursing ventrolaterally (Fig. 3B,C). A moderate number of fibers is present in the medial part of the central grey. These fibers run parallel to the aqueduct, and some penetrate the ependymal layer. Densely stained VP fibers traverse the an-
Fig. 2. A. Immunoreactive fibers (arrows) in the rat locus coeruleus stained with an antibody directed against rat glycoprotein (#C3 final). Neurons of the rat locus coeruleus (arrowheads) show faint staining after long incubation in DAB solution. B. Immunoreactive fibers (arrows) in the monkey locus coeruleus stained with an antibody directed against guinea pig glycoprotein (#K.1.7). Small amounts of neuromelanin can be observed in some neurons (arrowheads). C. Immunoreactive fibers (arrows) in the human locus coeruleus stained with an antibody directed against vasopressin (# 125). Large amounts of neuromelanin can be observed in the neurons (arrowheads).
234
nular nucleus to the ventral region of the central grey. The oculomotor nuclear complex is not innervated by VP fibers. Lateral to the central gray a few scattered VP fibers traverse the lateral reticular tegmental area. More caudally, at the level of the trochlear nuclear complex, the agglomeration of VP fibers in the dorsolateral quadrant of the central grey continues (Fig. 3C). They leave the central grey in a ventrolateral direction. At the level of the nucleus of the lateral lemniscus, these fibers meet a smaller VP fiber tract which courses by the lateral margin of the brainstem. A second dense VP fiber tract runs through the midline of the brain from the central superior nucleus in between the medial longitudinal fascicles to reach the ventral part of the central grey. The trochlear nuclei are not innervated by VP fibers. This type of innervation is maintained at more caudal levels, i.e., with a heavy innervation of the dorsal raphe. At the level of the LC strong VP innervation is found in the ventricular margin of the periventricular tegmentum (Fig. 3D). This pattern of innervation spreads out towards more ventral levels, diffusely traversing the entire periventricular tegmentum, including the dendritic field of the LC (Fig. 2B). A moderate number of fibers also penetrates the LC proper in the anterior and middle parts of the LC. The dorsal tegmental nucleus, however, is only sparsely innervated. Lateral to the LC a large number of VP fibers occurs in the lateral and medial parabrachial nuclei.
Vasopressinergic innervation in the human Data is lacking on the precise distribution of VP afferents in the pontine tegmentum in the human. Using immunocytochemical staining of post-mortem material, the existence of VP fibers was reported in the central grey and dorsal raphe (Sofroniew et al., 1981). Furthermore, dense VP innervation can be observed in the LC of the human (Fig. 2C; Fliers et al., 1986; Rossor et al., 1982). Some of these fibers form perineuronal specializations on neurons of the LC. It is unclear
whether there are rostro-caudal differences in VP innervation of the human LC. Immunoelectron microscopical analysis of vasopressin terminal profiles in the locus coeruleus of the rat
Examination of the ultrathin sections showed several immuno-stained profiles in the LC area. Many of these profiles were small and appeared as intervaricose segments, while only a few could be identified as terminal profiles. Some of the immuno-labeled structures could not be identified with certainty either because they were obscured by reaction product or because of poor ultrastructure. The immuno-labeled terminal profiles (Figs. 7-8) contained many clear vesicles in addition to a number of relatively large densecored vesicles (80- 120 nm). Terminals observed in contrasted sections were also traced in an adjacent unstained section in order to establish with certainty that the terminal was labeled. Synaptic membrane specializations could be identified in some of the labeled terminals (Figs. 7-8). They were usually asymmetrical and contacts were established with small (distal) dendrites. Synaptic contacts with large dendrites or cell somata were rarely observed. The morphological characteristics of the VP-labeled terminal profiles in the LC of the rat were similar to those observed in the rat limbic system (Buijs and Swaab, 1979). Vasopressin neurons in the locus coeruleus of the rat and monkey
In the LC of rats pretreated with colchicine some intense immunoreactive cells are found with antibodies directed against all three parts of propressophysin (Fig. 4A). These cells are multipolar and medium-sized, and possess large nuclei relative to the cytoplasm (see Caff6 et al., 1985). In the more anterior part of the LC, VP neurons are not restricted to the nucleus proper but extend ventrally into the subcoerulear region (Caffk and
235
Fig. 4. A. Neurophysin (#R4)-immunoreactive cells (arrowheads) in the locus coeruleus of the rat. IV, fourth ventricle. Bar = 30 pm. B. A glycoprotein (#K.1.7.)-immunolabeled neuron (arrowhead) in the periventricular tegmentum of the monkey. IV, fourth ventricle. Bar = 100 prn. Fig. 5. Pairs of adjacent 5 p m cryostat sections of the posterior pole of the rat locus coeruleus stained with an antibody directed against noradrenaline (#W3/30-9) (A) and vasopressin (#W1) (B). Arrowheads denote a group of cells with both noradrenaline and vasopressin irnrnunoreactivity. IV, fourth ventricle. Bar = 30 pm.
236
Van Leeuwen, 1983). Caudally the cells are confined to the compact part of the LC. When adjacent 5 p m cryostat sections are stained with either an antibody to noradrenaline or to VP, some of the cells in the rat LC display both VP and noradrenaline immunoreactivity. Most of the cells in which VP and noradrenaline coexist are situated in the caudal part of the LC complex (Fig. 5A,B). Only a small number of these cells is located in the anterior LC (Fig. 6A,B). In the colchicine-pretreated monkey we observed a few intensely stained solitary multipolar VP neurons in the periventricular tegmentum, but not in the LC proper (Fig. 4B). Faintly stained cells are also present in the LC proper, but their number is much smaller as compared with the rat. Whether VP coexists with noradrenaline in these neurons in the monkey has not yet been examined. To date, no VP cells have been identi-
fied in the region of the human LC. The possible transport of VP by efferents of the LC was investigated in the rat after thermal or neurotoxic lesioning of the compact part of the LC. Two methods were used. Sections through the whole brain were screened to detect a possible reduction of VP-immunoreactive fibers, and the level of VP and noradrenaline was radioimmunologically determined in selected regions of the brain. After appropriate survival periods, no decrease of VP-immunoreactive fibers could be observed in the brain when compared with controls. After a unilateral thermal LC lesion a significant reduction of noradrenaline levels was measured in selected brain areas, e.g., cerebral cortex, ventral hippocampus and cerebellum, ipsilateral to the lesion. There was no effect on the amount of VP. Similar results were obtained after neurotoxic LC lesion (see CaffC et al., 1988).
Fig. 6. Pairs of adjacent 5 p n sections of the anterior pole of the rat locus coeruleus stained with an antibody directed against noradrenaline (#W3/30-9) (A) and rat neurophysin (#R4) (B). Arrowheads denote a cell with both noradrenaline and neurophysin immunoreactivity. IV, fourth ventricle. Bar = 30 Fm.
237
Figs. 7-8. Electronmicrographs showing glycoprotein (#C3 final, Fig. 7)- o r neurophysin (#R4,Fig. 8)- immunoreactive terminal profiles contain many small clear vesicles and several dense-cored vesicles (white arrowheads). Presumed synaptic specializations (black arrowheads) were established with small dendrites (d). Contrasted with uranyl acetate and lead citrate (Fig. 7) or uncontrasted (Fig. 8). Bar = 0.2 p m .
Discussion Using Phaseolus vulgaris leucoagglutinin anterograde axonal tracing, Luiten et al. (1985) de-
scribed two major fiber tracts through the brainstem which originate from the PVN in the rat. The first passes dorsally through the periaqueductal grey, periventricular grey, LC, and parabrachial nuclei to the principal sensory trigeminal nucleus and the nucleus of the tractus solitarius. A second, smaller fiber tract courses laterally through the brainstem. Holstege (1987) studied PVN efferents to the brainstem and spinal cord in the cat using autoradiography. According to this study PVN neurons in the cat send their descending fibers via the medial forebrain bundle and medial central grey. At the level of the upper pons some lateral fibers shift medially. Others course dorsally in the periventricular tegmentum. Efferents from the BST in the cat were also studied using autoradiography (Holstege et al., 1985). At mesencephalic levels, fibers from the BST are distributed to the periaqueductal grey; they pass ventrolateral to the cuneiform and pedunculopontine nuclei. Caudally in the pons fibers terminate in the parabrachial nuclei, the nucleus of Kolliker-Fuse, and the LC. The origin and course of VP fibers in the pontine tegmentum of the rat can be delineated when the abovementioned anatomical data are combined with studies affecting distinct VP nuclei in this species. Using extensive BST lesions, De Vries and Buijs (1983) reported a decrease of VP fiber density in the periventricular grey and LC, among others. Furthermore, after long-term castration, VP fibers disappear in the dorsal central grey, dorsal raphe, and LC, among others (De Vries et al., 1985). VP fibers in these regions were not affected by lesions made in the PVN, SCN, or the medial amygdala (Hoorneman and Buijs, 1982; De Vries and Buijs, 1983; Caff6 et aZ., 1987). These arguments point to the BST as the source of the major contingent of VP fibers in the rat dorsal pontine tegmentum. Currently, similar data are not available for either the cat, nonhuman primate or human. The VP neurons and fibers from the BST (and medial amygdala) of the rat appeared to be under direct gonadal steroid control (De Vries et al.,
238
1983; Axelson and Van Leeuwen, 1990). VP cell numbers in these nuclei are larger and their efferents more extensive in the male as compared with the female rat (Van Leeuwen et al., 1985). After orchidectomy the VP cells in these nuclei slowly lose their VP content. This effect is reversed by testosterone replacement therapy (see De Vries et al., 1985). Quantitatively the most important VP terminal areas in the rat (and monkey) dorsal pontine tegmentum are the dorsal raphe and the rostra1 and middle parts of the LC. We have preliminary evidence that these VP fibers give rise to terminal profiles which establish synaptic contacts mainly with distal dendrites of LC cells. Whether synaptic contacts are being formed with LC cells of specific morphological or neurochemical subtype remains to be determined. The VP terminal profiles may be responsible for mediating the activation of LC cells during local VP application as reported by Olpe and Baltzer (1981). A monosynaptic BST-LC connection would further have several interesting physiological consequences. Since the density of these VP fibers is sex-dimorphic this may imply that the number of LC cells contacted would be different in the male and the female rat. Because the amount of VP in these fibers is under gonadal steroid control, the discharge rates of contacted LC cells to VP stimulation would be subject to modification over periods of days, weeks or even years. These particular situations remain to be investigated. Available immunocytochemical data clearly show that each of the three parts of propressophysin is being expressed within LC neurons of colchicine-pretreated rats when examined with refined immunocytochemical methods. Since it is unlikely that propressophysin as such, or all three constituent parts, are separately taken up by LC cells it is assumed that these molecules are being synthesized within LC neurons. However, studies designed to detect VP mRNA by in situ hybridization histochemistry did not provide evidence for their synthesis in the LC (Urban et al.,
1990). The question of VP biosynthesis within LC cells is, therefore, still a matter of conjecture. Whatever the case, the partial coexistence of VP and NA within LC cells shows the neurochemical heterogeneity of this nucleus in addition to its morphological diversity. Attempts to identify VP efferents from LC neurons have so far proved to be unsuccessful (Caffb et al., 1988). There are several possible explanations for this, one is that LC VP cells project to other than the studied brain areas. A likely candidate for a projection of VP/noradrenaline-containing LC axons remains the spinal cord, since LC axons project abundantly to this region (Westlund et al., 1983; Fritschy and Grzanna, 1990). Alternatively, VP might be processed differently in the LC when compared to the HNS. Burbach et al. (1984) found VP fragments in the rat brain which are not present in the pituitary. Thus, the various VP systems may process this peptide in a different manner. Furthermore, VP might not be transported and may exert some local metabolic function within LC neurons. In general, proteins produced in the cell soma need not undergo axonal transport. An example of a propressophysin-derived protein is reported by Van Leeuwen et al. (1989). For the LC a similar absence of axonal transport to distant projection sites was reported for neuropeptide Y (Gustafson and Moore, 19871, although there is no consensus about this observation (Holets et al., 1988). In recent years the LC has received much attention because of its unique anatomical and physiological properties, self-evidently ascribed to its noradrenergic nature. It is now becoming increasingly clear that other neurosubstances like neuropeptides may be equally important. The demonstration of both VP neuropeptide within neurons of LC and of VP terminal profiles within this nucleus raises new questions and initiates additional investigations. Progress in this research will enable a better understanding of anatomical and physiological characteristics of the LC.
239
Acknowledgements
The authors express their gratitude to Dr. J.P.H. Burbach (Rudolf Magnus Institute for Pharmacology, Utrecht), Prof. A.G. Robinson (University of Pittsburgh, NIH grant AM 16166) for their generous gifts of the antisera to the rat glycoprotein (#C3 final) and rat neurophysin (#R4), respectively. Prof. J. Voogd is acknowledged for his helpful comments during the preparation of the manuscript. Ms. C.M.H. Bongers is thanked for performing the electron microscopic work. Furthermore we are indebted to H. Stoffels and G. van der Meulen for the drawings and photographs, respectively, and to R. Hawkins for correcting the English text. References Aston-Jones, G., Foote, S L. and Bloom, F.E. (1984) Anatomy and physiology of locus coeruleus neurons: Functional implications. In M.G. Ziegler and C.R. Lake (Eds.) Norepinephrine. Frontiers of Clinical Neuroscience, Vol 2, Williams and Wilkins, Baltimore, pp. 92-116. Axelson, J.F. and Van Leeuwen, F.W. (1990) Differential localization of estrogen receptors in various vasopressin synthesizing nuclei of the rat brain. J. Neuroendocrinol., 2: 209-216. Buijs, R.M. and Swaab, D.F. (1979) Immuno-electron microscopical demonstration of vasopressin and oxytocin synapses in the limbic system of the rat. Cell Tissue Res., 204: 355-365. Buijs, R.M. and Van Heerikhuijze, J.J. (1982) Vasopressin and oxytocin release in the brain: A synaptic event. Bruin Res., 252: 71-76. Buijs, R.M., De Vries, G.J., Van Leeuwen, F.W. and Swaab, D.F. (1983) Vasopressin and oxytocin: Distribution and putative function in the brain. In B.A. Cross and G. Leng (Eds.) The Neurohypophysis: Structure, Function and Control. Progress in Brain Research, Vol. 60, Elsevier, Amsterdam, pp. 115-122. Burbach, J.P.H., Wang, X., Ten Haaf, J.A. and De Wied, D. (1984) Substances resembling C-terminal vasopressin fragments are present in the brain but not in the pituitary. Brain Res., 306: 384-387. Caff.6, A.R. and Van Leeuwen, F.W. (1983) Vasopressin-immunoreactive cells in the dorsomedial hypothalamic region, medial amygdaloid nucleus and locus coeruleus of the rat. Cell Tissue Res., 233: 23-33. Caff.6, A.R., Van Leeuwen, F.W., Buijs, R.M., De Vries, G.J. and Geffard, M. (1985) Coexistence of vasopressin, neurophysin and noradrenaline immunoreactivity in medium-
sized cells of the locus coeruleus and subcoeruleus in the rat. Bruin Res., 338: 160-164. Caff.6, A.R., Van Leeuwen, F.W. and Luiten, P.G.M. (1987) Vasopressin cells in the medial amygdala of the rat project to the lateral septum and the ventral hippocampus. J. Cornp. Neurol., 261: 237-252. CaffC, A.R., Van Leeuwen, F.W., Buijs, R.M. and Van Der Gugten, J. (1988) Vasopressin and noradrenaline coexistence in the rat locus coeruleus: differential decreases of their levels in distant brain areas after thermal and neurotoxic lesions. Brain Res., 459: 386-390. CaffC, A.R., Van Ryen, P.C., Van Der Woude, T.P. and Van Leeuwen, F.W. (1989) Vasopressin and oxytocin systems in the brain and upper spinal cord of Macaca fascicularis. J. Cornp. Neurol., 287: 302-325. De Vries, G.J. and Buijs, R.M. (1983) The origin of the vasopressinergic and oxytocinergic innervation of the rat brain with special reference to the lateral septum. Brain Res., 273: 307-317. De Vries, G.J., Best, W. and Sluiter, A.A. (1983) The influence of gonadal steroids on a sex difference in the vasopressinergic innervation of the brain. Deuelop. Brain Res., 8: 377-380. De Vries, G.J., Buijs, R.M., Van Leeuwen, F.W., Caff.6, A.R. and Swaab, D.F. (1985) The vasopressinergic innervation of the brain in normal and castrated rats. J. Cornp. Neurol., 233: 236-254. Dubois-Dauphin, M. and Zakarian, S. (1987) Distribution of the C-terminal glycopeptide of the vasopressin prohormone in rat brain: An immunocytochemical study. Neuroscience, 21: 903-921. Du Vignaud, V., Gish, D.T. and Katsoyannis, P.G. (1954) A synthetic preparation possessing biological properties associated with arginine vasopressin. J. Am. Chem. Soc., 76: 4751-4752. Fliers, E., Guldenaar, S.E.F., Van Der Wal, N. and Swaab, D.F. (1986) Extrahypothalamic vasopressin and oxytocin in the human brain; presence of vasopressin cells in the bed nucleus of the stria terminalis. Brain Res., 375: 363-367. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Rec., 63: 844-914. Fritschy, J.M. and Grzanna, R. (1990) Distribution of locus coeruleus axons within the rat brainstem demonstrated by Phaseolus vulgaris leucoagglutinin anterograde tracing in combination with dopamine-P-hydroxylase immunofluorescence. J. Cornp. Neurol., 293: 616-631. Gainer, H. (1983) Precursors of vasopressin and oxytocin. In B.A. Cross and G. Leng (Eds.) The Neurohypophysis: Structure, Function and Control. Progress in Brain Research, Vol. 60, Elsevier, Amsterdam, pp. 205-215. Grzanna, R. and Fritschy, J.M. (1989) Immunohistochemical analysis of the regeneration of noradrenergic axons following ablation by DSP-4 treatment. Eur. J. Neurosci., Suppl. 2, 230. Gustafson, E.L. and Moore, R.Y. (1987) Locus coeruleus neurons produce, but do not transport neuropeptide tyrosine (NPY). Soc. Neurosci. Abstr., 13: 298.
240 Holets, V.R., Hokfelt, T., Rokaeus, A., Terenius, L. and Goldstein, M. (1988) Locus coeruleus neurons in the rat containing neuropeptide Y, tyrosine hydroxylase or galanin and their efferent projections to the spinal cord, cerebral cortex and hypothalamus. Neuroscience, 24: 893-906. Holstege, G. (1987) Some anatomical observations on the projections from the hypothalamus to brainstem and spinal cord: An HRP and autoradiographic study in the cat. J. Comp. Neurol., 260: 98-126. Holstege, G., Meiners, L. and Tan, K. (1985) Projections of the bed nucleus of the stria terminalis to the mesencephalon, pons and medulla oblongata in the cat. Exp. Brain Res., 58: 379-391. Hoorneman, E.M.D. and Buijs, R.M. (1982) Vasopressin fiber pathways in the rat brain following suprachiasmatic nucleus lesioning. Brain Res., 243: 235-241. Jard, S. (1983) Vasopressin: Mechanisms of receptor activation. In B.A. Cross and G. Leng, (Eds.) The Neurohypophysis: Structure, Function and Control. Progress in Brain Research, Vol. 60,Elsevier, Amsterdam, pp. 383-394. Kovacs, G.L., Bohus, B. and Versteeg, D.H.G. (1979) Facilitation of memory consolidation by vasopressin: Mediation by terminals of the dorsal noradrenergic bundle? Brain Res., 172: 73-85. Livett, B.G. (1978) Immunocytochemical localization of nervous system specific proteins and peptides. Znt. Reu. Cytol., SUPPL7: 53-231. Loughlin, S.E., Foote, S.L. and Bloom, F.E. (1986a) Efferent projections of nucleus locus coeruleus: Topographic organisation of cells of origin demonstrated by three-dimensional reconstruction. Neuroscience, 18: 291-306. Loughlin, S.E., Foote, S.L. and Grzanna, R. (1986b) Efferent projections of nucleus locus coeruleus: Morphologic subpopulations have different efferent targets. Neuroscience, 18: 307-319. Luiten, P.G.M., Ter Horst, G.J. Karst, H. and Steffens, A.B. (1985) The course of paraventricular hypothalamic efferents to autonomic structures in medulla and spinal cord. Brain Res., 329: 374-318. Muller, L.L. and Jacks, T.J. (1975) Rapid chemical dehydration of samples for electron microscopic examinations. J. Histochem. Cytochem., 23: 107-110. Olpe, H. and Baltzer, V. (1981) Vasopressin activates noradrenergic neurons in the rat locus coeruleus: A microiontophoretic investigation. Eur. J. Pharmacol., 73: 377-378. Paxinos, G. and Watson, C. (1986) The Rat Brain in Stereotaxic Coordinates, Second edition. Academic Press, Sidney. 237 pp. Rossor, M.N., Hunt, S.P., Iversen, L.L., Bannister, R., Hawthorn, J., Ang, V.T.Y. and Jenkins, J.S. (1982) Extrahypothalamic vasopressin is unchanged in Parkinson’s disease and Huntington’s disease. Brain Res., 253: 341-343.
Shantha, T.R., Manocha, S.L. and Bourne, G.H. (1968) A Stereotaric Atlas of the Java Monkey Brain (Macaca irus). S. Karger, New York, 68 pp. Sofroniew, M.V. (1985) Vasopressin- and neurophysin-immunoreactive cells in the septa1 region, medial amygdala and locus coeruleus in colchicine-treated rats. Neuroscience, IS: 347-358. Sofroniew, M.V. and Schrell, U. (1982) Evidence for a direct projection from oxytocin and vasopressin neurons in the hypothalamic paraventricular nucleus to the medulla oblongata: Immunohistochemical visualization of both the horseradish peroxidase transported and the peptide produced by the same neurons. Neuroscience Lett., 22: 211217. Sofroniew, M.V., Weindl, A., Schrell, U. and Wetzstein, R. (1981) Immunohistochemistry of vasopressin, oxytocin and neurophysin in the hypothalamus and extrahypothalamic regions of the human and primate brain. Acta Histochem. 24: 79-95. Swaab, D.F., Pool, C.W. and Nijveldt, F. (1975) Immunofluorescence of vasopressin and oxytocin in the rat hypothalamo-neurohypophyseal system. J. Neural. Transm., 36: 195-215. Urban, J.H., Miller, M.A., Drake, C.T. and Dorsa, D.M. (1990) Detection of vasopressin mRNA in cells of t h e medial amygdala but not the locus coeruleus by in situ hybridization. J. Chem. Neuroanat. 3: 277-283. VandeSande, F., Dierich, K. and De Mey, J. (1975) Identification of the vasopressin-neurophysin producing neurons of the rat suprachiasmatic nuclei. Cell Tissue Res., 156: 377-380. Van Leeuwen, F.W. (1987) Vasopressin receptors in the brain and pituitary. In: D.M. Gash and G.J. Boer (Eds.) Vasopressin: Principles and Properties. Plenum Press, New York, pp. 477-496. Van Leeuwen, F.W. and Caffi, A.R. (1983) Vasopressin-immunoreactive cell bodies in the bed nucleus of the stria terminalis of the rat. Cell Tissue Res., 228: 525-534. Van Leeuwen, F.W., Caff6, A.R. and De Vries, G.J. (1985) Vasopressin cells in the bed nucleus of the stria terminalis of the rat: Sex differences and the influence of androgens. Brain Res., 325: 391-394. Van Leeuwen, F., Van Der Beek, E., Seger, M., Burbach, P. and Ivell, R. (1989) Age-related development of a heterozygous phenotype in solitary neurons of the homozygous Brattleboro rat. Proc. Natl. Acad. Sci. USA, 86: 64 17-6420. Versteeg, D.H.G. (1983) Neurohypophyseal hormones and brain neurochemistry. Pharmacol. Ther., 19: 297-325. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of the rat. Brain Res., 263: 15-31.
C.D. Barnes and 0. Pornpeiano (Eds.) Progress in Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
24 1 CHAPTER 18
Responses of locus coeruleus neurons to neuropeptides H.-R. Olpe and M. Steinmann Research and Deaelopment Department, Pharmaceuticals Division, CIBA-GEIGY Ltd., Basel, Switzerland
The knowledge on the neuronal inputs to the locus coeruleus (LC) and their roles in regulating noradrenergic (NA) cellular activity is quite advanced. In recent years, however, about ten neuropeptides were found to be localized in the area of the rodent LC; peptides which may be considered as potential transmitters o r modulators acting in this area. Electrophysiological studies performed in viuo and in Liitro have revealed that many of these peptides are able to alter LC neuronal activity. Stimulatory effects have been described with vasopressin, substance P, adrenocorticotropin hormone and corticotropin-releasing factor. Depressant effects were seen with galanin, somatostatin, neuropeptide Y and enkephalin. Variable actions were observed in the case of neurotensin. While these findings point to a possible regulatory function of these peptides in this area, precise roles
remain unclear. Important information is lacking that would conclusively demonstrate their regulatory functions. It should be determined whether the stimulation of peptidergic cells elicits synaptic effects identical to the ones observed with local exogenous peptide applications. By studying the action of blockers of these transmitter and modulator candidates, we would probably begin to understand their importance in the regulation of tonic and phasic activity components. The LC is generally considered to consist of a homogenous group of neurons. T h e recent observation that subpopulations of these cells contain peptides as in the case of neuropeptide Y, galanin and vasopressin, points to the possible existence of subgroups of neurons having different functions.
Key words: locus coeruleus, neuropeptides, electrophysiology, modulator, rat
Introduction
The locus coeruleus (LC) of the rat brain is generally considered to be a homogenous and compact nucleus containing mainly noradrenergic (NA) neurons. Intra- and extracellular electrophysiological recordings are in keeping with this notion since the biophysical properties of all cells appear to be very similar (Williams et al., 1984). Over the past ten years, a steadily increasing number of potential physiological roles have been proposed that are claimed to be linked with this
brain area (Olpe et al., 1985; Svensson, 1987). We are, therefore, facing the problem that a small number of seemingly homogeneous neurons is thought to be involved in several distinct brain functions. One theoretical approach for solving this problem is to search for a common basic action of noradrenaline that would be coupled with a number of different, more complex functions. Alternatively, a basis for the heterogeneity of functions could be found in the fact that the LC projections have recently been shown to be selective and not as widely and diffusely spread
242
TABLE 1 Neuropeptides present in the area of the locus coeruleus Peptides present in locus coeruleus
Localized in perikarya
vasopressin substance P adrenocorticotropin hormone, ACTH corticotropin-releasing factor, CRF somatostatin. SRIF neuropeptide Y enkephalin neurotensin galanin delta sleep-inducing peptide, DSIP
as originally described (Grzanna et al., 1987). Another possible suggestion for the heterogeneity of LC neurons themselves derives from recent reports that several neuropeptides are present in this brain area which have been localized in fibers and/or cell bodies. A survey of the kind of peptides described in this area is provided in Table 1 (see also Sutin and Jacobowitz, this volume). It is remarkable that the rodent LC contains not only a considerable number of different peptides, but some of them are present at rather high concentrations, as in the case of substance P (Douglas et al., 1982) and somatostatin (Palkovits et al., 1982). The presence of these peptides raises the question of their physiological functions. Electrophysiological techniques provide a means of addressing this issue, at least partly, in that the sensitivity of LC neurons to these peptides can be investigated. While such studies give valuable information on the effects of peptides on cell activity, their membrane potentials and other biophysical membrane properties, they do not necessarily solve the problem of identifying the peptides’ physiological roles. Such information may eventually be obtained from a combination of different approaches including behavioral studies. Neuropeptides localized in fibers in the LC
Localized in fibres
Authors CaffC et al., 1985 De Vries et al., 1985 Pickel et al., 1979 Watson et al., 1978 Cummings et al., 1983 Finley et al., 1981 Johansson et al., 1984 Everitt et al., 1984 Chronwall et al., 1985 Charney et al., 1982 Pickel et al., 1979 Jennes et al., 1982 Moore and Gustafson, 1989 Feldman and Kastin, 1984
may be considered as potential transmitter candidates, but the question of whether the fibers make direct synaptic contacts with NA cell bodies or dendrites is not yet known. It is also unclear whether the peptides are colocalized with any one of the two major afferent fiber systems projecting to this brain area, as described by AstonJones et al., 1986 (see also this volume). With such information available, the electrophysiological experiments could be designed in a more appropriate manner. The knowledge of the precise termination area of these fibers would point to the sites where peptides are likely to exert their actions i.e., in the cell body area of the LC proper or on their dendrites. Neuropeptides colocalized with NA LC neurons, on the other hand, may be cotransmitter candidates, i.e., they may be transported to target areas and released as cotransmitters together with noradrenaline. Among the neuropeptides shown to be colocalized with noradrenaline are galanin (Melander et al., 1986), neuropeptide Y (Everitt et al., 19841, enkephalin (Charnay et al., 1982) vasopressin (Caff6 et al., 1985) and the corticotropin-releasing factor (Cummings et al., 1983). Finally, a release of peptides via dentrites or axon collaterals located within the LC area itself is also conceivable. The
243
TABLE 2 Electrophysiological effects of neuropeptides on locus coeruleus neurons Preparation
Authors
increase increase increase
anaesthetized rat anaesthetized rat slice, gerbil
Olpe and Baltzer, 1981 Berecek eta/., 1987 Olpe et al., 1987
substance P
increase increase increase
anaesthetized rat slice, rat slice, gerbil
Guyenet and Aghajanian, 1977 Cheeseman et al., 1983 Olpe et al., 1987
adrenocorticotropin hormone, ACTH
increase variable effects increase
anaesthetized rat anaesthetized rat slice, gerbil
Olpe and Jones, 1982 A d a m and Foot, 1988 Olpe et al., 1987
corticotropin-releasing factor, CRF
increase
anaesthetized rat
Valentino et al., 1983
somatostatin, SRlF
decrease
slice, gerbil cell culture
Olpe ef al., 1987 Masuko et al., 1986
neuropeptide Y
decrease
slice, rat
Illes and Regenold, 1990
enkephalin
decrease
anaesthetized rat slice, rat
Korf et al., 1974 Williams and North, 1984 Pepper and Henderson 1980
Peptide
Effect on dicharge rate
vasopressin
Effect on membrane potential
hyperpolarisation hyperpolarisation
hyperpolarisation neurotensin
inactive decrease increase
anaesthetized rat anaesthetized rat slice, gerbil
Guyenet and Aghajanian, 1977 Young et al., 1978 Olpe et al., 1987
galanin
decrease
slice, rat
Seutin et al., 1989
peptides could thereby function to regulate LC neuronal activity at the level of NA cell bodies or dendrites. The major methodological problems linked to electrophysiological studies on peptides concern the mode of peptide administration, the peptide catabolism within the tissue and the diffusional barriers imposed on these rather large molecules by the tissue itself. Most studies published on the action of peptides in this area are complicated by these factors, with the exception possibly of cellculture studies (Masuko et al., 1986). The route of administration may be of particular importance. The mode of application of somatostatin in the hippocampus has been shown to give rather variable results with local ionophoretic administrations resulting in excitatory (Olpe et aL, 1980), and bath-applications exerting inhibitory effects
(Pittman and Siggins, 1981). Another major problem of this kind of investigation is the fact that receptor blockers of these peptides were either not available or were not used. In view of these methodological considerations and limitations, and given the fact that the cellular origin of most peptides is unknown, our understanding of the action of neuropeptides in the LC must be considered as rather preliminary. The effects of peptides on locus coeruleus neurons
Table 2 lists some of the major effects described with neuropeptides on rodent LC neurons. With the exception of a few investigations, we lack detailed information on these agents as modulators of established afferent transmitters or of
245 E ffect on firin g rate of locus coeruleus neurons neurotensin was found to be inactive in viuo N=10 (Guyenet and Aghajanian, 1977) and subse490 Neurok,inin to - A depress NA cells in a quently was observed Neurokinin - B Substance - P similar study (Young Eledoisin et al., 1978). Bath-applied neurotensin in the slice preparation of the gerbil was weakly excitatory if tested at a concentration .-C ofolu 10 p M (Olpe et al., 1987). Taken together, a rather small number of neurons has been investineurotensin was found to be inactive in viuo (Guyenet and Aghajanian, 1977) and subsequently was observed to depress NA cells in a similar study (Young et al., 1978). Bath-applied neurotensin in the Concentratlon slice preparation of the gerbil in nM was weakly excitatory if tested at a concentration Fig. 2. Mean concentration-response curves of the excitatory ofactions 10 pofMfour (Olpe et al.,on1987). Taken together, tachykinins the spontaneous firing rate aof gerbil LC neurons are depicted. The spontaneous firing rather small number of neurons has beenmean investi-
gated so far and the cellular sensitivity may not be uniform across the LC. Pronounced excitation of neuronal activity was observed however with the vasoactive intestinal peptide in an in vitro study (Wang and Aghajanian, 1989). It was concluded that this peptide induces an Na+-dependent inward current and the involvement of a pertussis toxin-sensitive G protein was suggested (Wang and Aghajanian, 1989). Considerable immunoreactivity to angiotensin I1 has been observed recently in the LC area (see the chapters by Marshall et al. and Speth et al.). The binding site for angiotensin was described as predominantly of the A 11, type (see Speth et al., this volume). It was found that angiotensin 11, while having almost no effect on the spontaneous rate recorded 10 min prior to the peptide administration was discharge frequency or on the membrane potenFig. 3. The effect of prolonged bath-administration of taken as the control value. Each value is the mean of 10 of four different tachykinins on the spontaneous firing rate tial of LC neurons, appears to attenuate glutaneurons rt S.E.M. LC neurons of the gerbil is depicted. Each peptide mate-evoked excitatory responses (see Marshall administered at a concentration of 100 WM. A quite et al., this volume). Both agents were pressure nounced tachyphylaxis is seen with substance P. neurotensin ejected. In three experiments, we tested the acneurotensin was was found found toto bebe inactive inactive inin viuo viuo (Guyenet and Aghajanian, 1977) and subsetion of bath-applied angiotensin I1 on eight rat (Guyenet and Aghajanian, 1977) and subsequently was observed to depress NA cells in a LC neurons. The peptide was applied at concenquently was observed to depress NA cells in a similar study (Young et al., 1978). Bath-applied trations of 10 and 100 p M . It did not induce any similar study (Young et al., 1978). Bath-applied neurotensin in the slice preparation of the gerbil notable change in spontaneous cell firing (OIpe, neurotensin in the slice preparation of the gerbil was weakly excitatory if tested at a concentration unpublished observation). Taken together the rewas weakly excitatory if tested at a concentration of 10 p M (Olpe et al., 1987). Taken together, a sults suggest that angiotensin I1 may act as a Fig. of prolonged bath-administration of foura of 3.10The p Meffect (Olpe et al., 1987). Taken together, ofbeen four investidifferent tachykinins on the spontaneous firing rate rather small number of neurons has neuromodulator in the LC area. rather small number of neurons has been investiLC neurons of the gerbil is depicted. Each peptide was Fig. 3. The effect of prolonged bath-administration A modulatory role has also been proposed for administered at a concentration of 100 WM. A quite profour differentenkephalin (McFadzean et al., 1987) and recently tachykinins on found the spontaneous firing rate of neurotensin was to be inactive i n viuo P. nounced tachyphylaxis is seen with substance .34 SUBSTANCE ELEDOlSlN LC neuronsP of the gerbil. 34 is depicted. Each peptide for neuropeptide Y (Illes and Regenold, 1990). (Guyenet and Aghajanian, 1977) andWM.subseof 100 A quite administered at a concentration Immunoreactivity to dynorphin, a putative ligand quentlynounced was observed tois depress NA cellsP. in a tachyphylaxis seen with substance for kappa opioid receptors, has been localized in similar study (Young et al., 1978). Bath-applied the LC area (Zamir et al., 1983). On the basis of neurotensin in the slice preparation of the gerbil was weakly excitatory if tested at a concentration electrophysiological investigations performed on of 10 p M (Olpe et al., 1987). Taken together, a a LC preparation, it has been suggested that K NEUROKlNlN A NEUROKlNlN B rather small number of neurons has been investireceptors are located presynaptically on afferent fibers that provide an excitatory input to LC. fibers that provide an excitatory input to LC. Activation of K receptors by a selective agonist Activation of K receptors by a selective agonist attenuated locally evoked excitatory postsynaptic attenuated locally evoked excitatory postsynaptic potentials (McFadzean et al., 1987). potentials (McFadzean et al., 1987). Neuropeptide Y exerts two effects. It deNeuropeptide Y exerts two effects. It depresses LC neuronal activity (Illes and Regenold, presses LC neuronal activity (Illes and Regenold, of 1990) and potentiates the hyperpolarizing effect 1990) and potentiates the hyperpolarizing effect of a,-receptor agonists in a selective manner of a,-receptor agonists in a selective manner 7
C
245 E ffect on firin g rate of locus coeruleus neurons 490 -
N=10 7
Neurok,inin - A Neurokinin - B Substance - P Eledoisin
.-C u
ol C
Concentratlon in nM
Fig. 2. Mean concentration-response curves of the excitatory actions of four tachykinins on the spontaneous firing rate of gerbil LC neurons are depicted. The spontaneous mean firing rate recorded 10 min prior to the peptide administration was taken as the control value. Each value is the mean of 10 neurons rt S.E.M.
neurotensin was found to be inactive in viuo (Guyenet and Aghajanian, 1977) and subsequently was observed to depress NA cells in a similar study (Young et al., 1978). Bath-applied neurotensin in the slice preparation of the gerbil was weakly excitatory if tested at a concentration of 10 p M (Olpe et al., 1987). Taken together, a rather small number of neurons has been investi.34
SUBSTANCE P
NEUROKlNlN A
. 34
ELEDOlSlN
NEUROKlNlN B
Fig. 3. The effect of prolonged bath-administration of four tachykinins on the spontaneous firing rate of four different LC neurons of the gerbil is depicted. Each peptide was administered at a concentration of 100 WM. A quite pronounced tachyphylaxis is seen with substance P.
gated so far and the cellular sensitivity may not be uniform across the LC. Pronounced excitation of neuronal activity was observed however with the vasoactive intestinal peptide in an in vitro study (Wang and Aghajanian, 1989). It was concluded that this peptide induces an Na+-dependent inward current and the involvement of a pertussis toxin-sensitive G protein was suggested (Wang and Aghajanian, 1989). Considerable immunoreactivity to angiotensin I1 has been observed recently in the LC area (see the chapters by Marshall et al. and Speth et al.). The binding site for angiotensin was described as predominantly of the A 11, type (see Speth et al., this volume). It was found that angiotensin 11, while having almost no effect on the spontaneous discharge frequency or on the membrane potential of LC neurons, appears to attenuate glutamate-evoked excitatory responses (see Marshall et al., this volume). Both agents were pressure ejected. In three experiments, we tested the action of bath-applied angiotensin I1 on eight rat LC neurons. The peptide was applied at concentrations of 10 and 100 p M . It did not induce any notable change in spontaneous cell firing (OIpe, unpublished observation). Taken together the results suggest that angiotensin I1 may act as a neuromodulator in the LC area. A modulatory role has also been proposed for enkephalin (McFadzean et al., 1987) and recently for neuropeptide Y (Illes and Regenold, 1990). Immunoreactivity to dynorphin, a putative ligand for kappa opioid receptors, has been localized in the LC area (Zamir et al., 1983). On the basis of electrophysiological investigations performed on a LC preparation, it has been suggested that K receptors are located presynaptically on afferent fibers that provide an excitatory input to LC. Activation of K receptors by a selective agonist attenuated locally evoked excitatory postsynaptic potentials (McFadzean et al., 1987). Neuropeptide Y exerts two effects. It depresses LC neuronal activity (Illes and Regenold, 1990) and potentiates the hyperpolarizing effect of a,-receptor agonists in a selective manner
246
(Illes and Regenold, 1990). The site and mechanism of this interaction remain to be elucidated. Discussion
The available data on neuropeptides in the LC of the rodent brain suggests that they play a physiological role in regulating neuronal activity. This hypothesis remains to be confirmed, however. There are no studies yet showing that blockers of any one of these transmitter or cotransmitter candidates affect LC neuronal activity. More importantly, it would be crucial to study the effect of stimulating the peptidergic neurons which provide fibers to the LC to see if the synaptic effects mimic the action of local or bath-applied peptides. In view of the lack of such information, our knowledge on the action and role of peptides in the LC is fragmentary and preliminary. So far, the strongest evidence for a role of various neuropeptides in this region is the histochemical demonstration that they are present. It is clear that unless we advance in our understanding of their functions, we will lack important information on the physiology of LC neurons. It was noted above that the apparent discrepancy between the homogeneity of the LC and the rather large number of postulated functions poses one of the main problems remaining to be solved. So far, the work related to peptides does not help much in approaching this issue. However, the observation that some peptides are not homogeneously distributed could be indicative of subgroups of neurons which might have different functions. Subpopulations of LC perikarya have been shown to contain neuropeptide Y and galanin (Moore and Gustafson, 1989). Coexistence of vasopressin with noradrenaline has been demonstrated mainly in the posterior part of LC (Caffk et al., 1985). In a recent study on the distribution of substance P, neuropeptide Y, neurotensin and thyrotropin-releasing hormone in the human LC it was demonstrated that these peptides are unevenly distributed (Pammer et al., 1990). The
immunoreactive neuronal networks showed the highest density in the middle, and to a lesser extent, in the caudal part of this nucleus. It is thus conceivable that the study of neuropeptides might provide a means of dividing this small nucleus into compartments of functional neuronal subgroups. It is not unlikely that these peptides will be shown to have important functions in the control of the LC and that the knowledge of their role might change our view on the functions of the LC itself. Acknowledgement
The author thanks Mrs. M. Koller for typing the manuscript. References Adams, L.M. and Foote, S.L. (1988) Effects of locally infused pharmacological agents on spontaneous and sensoryevoked activity of locus coeruleus neurons. Bruin Res. Bull., 21: 395-400. Aston-Jones, G., Ennis M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986) The brain nucleus locus coeruleus: Restricted afferent control of a broad efferent network. Science, 234: 734-737. Berecek, K.H., Olpe, H.-R., Mah, S.C. and Hofbauer, K.G. (1987) Alterations in responsiveness of noradrenergic neurons of the locus coeruleus in deoxycorticosterone acetate (DOCAI-salt hypertensive rats. Brain Res., 401: 303-31 1. Cafft, A.R., Van Leeuwen, F.W., Buijs, R.M., De Vries, G.J. and Geffard, M. (1985) Coexistence of vasopressin, neurophysin and noradrenaline in medium-sized cells of the locus coeruleus and subcoeruleus in the rat. Brain Res., 338: 160-164. Charnay, Y., LCger, L., Dray, F., Btrod, A,, Jouvet, M., Pujol, J.E. and Dubois, P.M. (1982) Evidence for the presence of enkephalin in catecholaminergic neurons of cat locus coeruleus. Neurosci. Lett., 30: 147-151. Cheeseman, H.J., Pinnock, R.D. and Henderson, G. (1983) Substance P excitation of rat locus coeruleus neurons. Eur. J. Pharmacol., 94: 93-99. Chronwall, B.M., Di Maggio, D.A., Massari, V.J., Pickel, V.M., Ruggiero, D.A. and O’Donohue, T.L. (1985) The anatomy of neuropeptide-Y-containing neurons in rat brain. Neuroscience, 15: 1159-1181. Cummings, S., Elde, R., Ells, J. and Lindall, A. (1983) Corticotropin-releasing factor immunoreactivity is widely distributed within the central nervous system of the rat: An immunohistochemical study. J. Neurosci., 3: 135551368,
247 De Vries, G.J., Buijs, R.M., Van Leeuwen, F.W., Caff6, A.R. and Swabb, D.F. (1985) The vasopressinergic innervation of the brain in normal and castrated rats. J. Comp. Neurol., 233: 236-254. Douglas, F.L., Palkovits, M. and Brownstein, M.J. (1982) Regional distribution of substance P-like immunoreactivity in the lower brainstem of the rat. Brain Res., 245: 376-378. Everitt, B.J., Hokfelt, T., Terenius, R., Tatemoto, K., Mutt, V. aild Goldstein, M. (1984) Differential co-existence of neuropeptide Y (NPY)-like immunoreactivity with catecholamines in the central nervous system of the rat. Neuroscience, 11: 443-462. Feldman, S.C. and Kastin, A.J. (1984) Localization of neurons containing immunoreactive delta sleep-inducing peptide in the rat brain: An immunocytochemical study. Neuroscience, 11: 443-462. Finley, J.C.W., Maderdrut, J.L., Roger, L.J. and Petrusz, P. (1981) The immunocytochemical localization of somatostatin-containing neurons in the rat central nervous system. Neuroscience, 6: 2173-2192. Grzanna, R., Chee, W.K. and Akeyson, E.W. (1987) Noradrenergic projections to brainstem nuclei: Evidence for differential projections from noradrenergic subgroups. J. Comp. Neurol., 263: 76-91. Guyenet, P.G. and Aghajanian, G.K. (1977) Excitation of neurons in the nucleus locus coeruleus by substance P and related peptides. Brain Res., 136: 178-188. Illes, P. and Regenold, J.T. (1990) Interaction between neuropeptide Y and noradrenaline on central catecholamine neurons. Nature (London), 344: 62-63. Jennes, L., Stumpf, W.E. and Kalivas, P.W. (1982) Neurotensin: Topographical distribution in rat brain by immunohistochemistry. J. Comp. Neurol., 210: 21 1-224. Johansson, O., Hokfelt, T. and Elde, R.P. (1984) Immunocytochemical distribution of somatostatin-like immunoreactivity in the central nervous system of the adult rat. Neuroscience, 13: 265-339. Korf, J., Bunney, B.S. and Aghajanian, G.K. (1974) Noradrenergic neurons: Morphine inhibition of spontaneous activity. Eur. J. Pharmacol., 25: 165-175. Masuko, S., Nakajima, Y., Nakajima, S. and Yamaguchi, K. (1986) Noradrenergic neurons from the locus coeruleus in dissociated cell culture: Culture methods, morphology and electrophysiology. J. Neurosci., 6: 3229-3241. McFadzean, I., Lacey, M.G., Hill, R.G. and Henderson, G. (1987) Kappa opioid receptor activation depresses excitatory synaptic input to rat locus coeruleus neurons in vitro. Neuroscience, 20: 231-239. Melander, T., Hokfelt, T., Rokaeus, A,, Cuello, A.C., Oertel, W.H., Verhofstad, A. and Goldstein, M. (1986) Coexistence of galanin-like immunoreactivity with catecholamines, 5-hydroxytryptamine, GABA and neuropeptides in the rat CNS. J. Neurosci., 6: 3640-3654. Moore, R.Y. and Gustafson, E.L. (1989) The distribution of dopamine-p-hydroxylase, neuropeptide Y and galanin in locus coeruleus neurons. J. Chem. Neuroanat., 2: 95-106. Olpe, H.-R. and Baltzer, V. (1981) Vasopressin activates
noradrenergic neurons in the rat locus coeruleus: A microionophoretic investigation. Eur. J. Pharmacol., 73: 377378. Olpe, H.-R. and Jones, R.S.G. (1982) Excitatory effects of ACTH on noradrenergic neurons of the locus coeruleus in the rat. Brain Res., 251: 177-179. Olpe, H.-R., Balcar, V.J., Bittiger, H., Rink, H. and Sieber, P. (1980) Central actions of somatostatin. Eur. J. Pharmacol., 63: 127-133. Olpe, H.-R., Steinmann, M.W. and Jones, R.S.G. (1985) Electrophysiological perspectives on locus coeruleus: Its role in cognitive versus vegetative functions. Physiol. Psycho/., 13: 179-187. Olpe, H.-R., Steinmann, M.W., Pozza, M.F. and Haas, H.L. (1987) Comparative investigations of the actions of ACTH 1-24' somatostatin, neurotensin, substance P and vasopressin on locus coeruleus neuronal activity in vitro. Naunyn Schmiedebergs Arch. Pharmacol., 336: 434-437. Palkovits, M., Epelbaum, J., Tapia-Arancibia, L. and Kordon, C. (1982) Somatostatin in catecholamine-rich nuclei of the brainstem. Neuropeptides, 3: 139-144. Pammer, C., Gorcs, T. and Palkovits, M. (1990) Peptidergic innervation of the locus coeruleus cells in the human brain. Brain Res., 515: 247-255. Pepper, C.M. and Henderson, G. (1980) Opiates and opioid peptides hyperpolarize locus coeruleus neurons in vitro. Science, 209: 394-396. Pickel, V.M., John, T.H., Reis, D.J., Leeman, S.E. and Miller, R.J. (1979) Electron microscopic localization of substance P and enkephalin in axon terminals related to dendrites of catecholaminergic neurons, Brain Res., 160: 387-400. Pittmann, Q.R. and Siggins, G.R. (1981) Somatostatin hyperpolarizes hippocampal pyramidal cells in vitro. Brain Rex, 221: 402-408. Seutin, V., Verbanck, P., Masotte, L. and Dresse, A. (1989) Galanin decreases the activity of locus coeruleus neurons in vitro. Eur. J. Pharmacol., 164: 373-376. Svensson, T.H. (1987) Peripheral, autonomic regulation of locus coeruleus noradrenergic neurons in brain: Putative implications for psychiatry and psychopharmacology. Psychopharmacology, 92: 1-7. Valentino, R.J., Foote, S.L. and Aston-Jones G. (1983). Corticotropin-releasing factor activates noradrenergic neurons of the locus coeruleus. Brain Res., 270: 363-367. Wang, Y.-Y. and Aghajanian, G.K. (1989) Excitation of locus coeruleus neurons by vasoactive intestinal peptide: Evidence for a g-protein-mediated inward current. Brain Res., 500: 107-118. Watson, S.J., Richard, C.W. and Barchas, J.D. (1978) Adrenocorticotropin in rat brain: Immunocytochemical localization in cells and axons. Science, 20: 1180-1181. Williams, J.T. and North, R.A. (1984) Opiate-receptor interactions on single locus coeruleus neurons. Mol. PharmaC O ~ . 26: , 489-497. Williams, J.T., Egan, T.M. and North, R.A. (1982) Enkephalin opens potassium channels in mammalian central neurons. Nature (London), 299: 74-76.
248 Williams, J.T., North, R.A., Shefner, S.A., Nishi, S. and Egan, T.M. (1984) Membrane properties of rat locus coeruleus neurons. Neuroscience, 13: 137-156. Young, W.S., Uhl, G.R. and Kuhar, M.J. (1978) Ionophoresis of neurotensin in the area of the locus coeruleus. Brain Res.. 150: 431-435.
Zamir, N., Palkovits, M. and Brownstein, M.J. (1983) Distribution of immunoreactive dynorphin in the central nervous system of the rat. Bruin Rex, 280: 81-93.
C.D. Barnes and 0. Pompeiano (Eds.) Projiress in Bruin Research, Vol. 88
249
0 l Y Y l Elsevier Science Publishers B.V.
CHAPTER 19
Pharmacology of locus coeruleus spontaneous and sensory-evoked activity R.J. Valentino and A.L. Curtis Department of Mental Health Science, Diuision of Behaciorul Neuvobiology, Hahnemann Uniuersity, Broad and Vine, Philadelphia, PA, U.S.A.
Neuroenabcrine and catecholamine dysfunctions in depression may be linked by corticotropin-releasing factor (CRF) effects on locus coeruleus (LC) neurons. One consequence of C R F hypersecretion in depression would be persistent elevated levels of LC discharge and diminished responses to phasic sensory stimuli. The hypothesis that antidepressants could reverse these changes was tested by characterizing effects of pharmacologically distinct antidepressants o n L C sensory-evoked discharge, LC activation by stress, and LC activation by CRF. The most consistent effect of all of the antidepressants tested was a decrease in LC sensory-evoked discharge after acute administration. However, tolerance occurs to these effects after chronic administration. With chronic administration each of the antidepressants produced
effects which could potentially interfere with C R F function in the LC. Desmethylimipramine and mianserin attenuated LC activation by a stressor which requires endogenous CRF, suggesting that these antidepressants attenuate stress-elicited release of C R F and perhaps the hypersecretion that occurs in depression. The serotonin reuptake inhibitor, sertraline (SER), enhanced the signal-to-noise ratio of the LC sensory response, an effect opposite to that of CRF. Thus, SER could serve as a functional antagonist of C R F that is hypersecreted in depression. The finding that three pharmacologically distinct antidepressants share the potential to interfere with C R F function in the LC implies that this may be an important common mechanism for antidepressant activity.
Key words: locus coeruleus, antidepressant, desmethylaminpramine, sertraline, mianserine, stress, corticotropin releasing factor
Introduction Depression is associated with dysfunctions in monoamine systems (Bunney and Davis, 1965; Schildkraut, 1965) and in the hypothalamic pituitary axis (HPA) (Stokes and Sikes, 1987, for review). Although all of the monoamine systems have been implicated in depression, most theories on the biological etiology of depression have focused on alterations in noradrenergic and/or
serotonergic function primarily because most clinically effective antidepressants affect these two neurotransmitter systems. Inconsistencies in the monoamine theories of depression have been pointed out. Thus, the finding that not all drugs which increase noradrenergic or serotonergic function are effective antidepressants (Post et al., 1974; Richelson and Pfenning, 1984), and the lack of temporal correlation between the pharmacological and clinical effects of antidepressants
250
(Overall et al., 1962; Oswald et al., 1972) has led to the formulation of new theories to explain the efficacy of these agents. Thus far, the effect which is most consistently produced by different antidepressants and which is most temporally correlated to their clinical efficacy is alteration in P-receptor function (see Heninger and Charney, 1987 for review). However, the behavioral correlate of this change has yet to be established. HPA dysfunction in depression is characterized by increased plasma levels of cortisol (Sachar et al., 1970; Lesch et al., 19881, ACTH (Lesch et al., 1988), and abnormal regulation by exogenously administered glucocorticoids (Carrol, 1982). The isolation and characterization of corticotropin-releasing factor (CRF) as the primary neurohormone responsible for ACTH release during stress (Vale et al., 1981) has provided a new tool for the study of HPA dysfunction in depression, and many of the endocrine abnormalities associated with depression are now thought to result from CRF hypersecretion (Gold et al., 1986; Holsboer et al., 1986; Lesch et al., 1988). The finding that CRF levels in the cerebrospinal fluid (CSF) of depressed patients are increased compared to nondepressed subjects supports the concept of CRF hypersecretion in depression (Nemeroff et al., 1984). Monoamine and HPA dysfunctions in depression may be linked by CRF actions on monoaminergic neurons. CRF-immunoreactive fibers and cells are localized in several extrahypophyseal regions including the serotonergic raphe nuclei and the noradrenergic nucleus locus coeruleus (LC) (Swanson et al., 1983; Cummings et al., 1983; Sakanaka et al., 1987; Foote and Cha, 1988). Hypersecretion of CRF in these regions could potentially affect monoamine function. This is particularly likely in the LC where substantial anatomic and physiological evidence suggests that CRF may serve as a neurotransmitter (Valentino, 1988). For example, CRF administered icv or directly into the LC by micropressure techniques alters LC discharge in anesthetized and unanesthetized rats such that spontaneous discharge is
increased and responses to phasic sensory stimuli are decreased (Valentino et al., 1983; Valentino and Foote, 1987, Valentino and Foote, 1988). This effect is dose-dependent and occurs with doses (1.0 -3.0 pg) that mimic behavioral (Sutton et al., 1982; Britton et al., 1985) and autonomic (Brown et al., 1982; Fisher et al., 1982) components of stress responses. Analogues of CRF that do not elicit ACTH release do not increase LC spontaneous discharge, suggesting that receptors for these effects have similar structural requirements (Valentino et al., 1983; Valentino and Foote, 1987, 1988). The most compelling evidence for a neurotransmitter role of CRF in the LC comes from studies demonstrating that LC activation elicited by hemodynamic stress (produced by infusion of nitroprusside) may be prevented by prior icv administration of the CRF antagonist, alpha helical CRF,-,,, but not by dexamethasone which blocks hypophyseal CRF release (Plotsky and Vale, 1984) indicating that the source of CRF mediating this effect is extrahypophyseal (Valentino and Wehby, 1988). Taken together with the localization of CRF-immunoreactive fibers in LC, these results have provided strong evidence for a neurotransmitter role in the LC during hemodynamic stress. The ability of various stressors to activate the LC noradrenergic system has been well documented. For example, shock, conditioned stimuli associated with shock, and hemorrhage increase norepinephrine turnover in brain regions whose sole source of norepinephrine is from the LC (Thierry et al., 1968; Korf et al. , 1973; Cassens et al., 1980; Solomon et al., 1986). Moreover, several physiological stressors including hypovolemia, hypercapnea, and bladder distention increase LC spontaneous discharge rate (Svenson, 1987). The neural pathways and neurotransmitters involved in stress-induced LC activation have yet to be established. However, based on the studies mentioned above, CRF is a strong candidate for a mediator of stress-induced LC activation. Considering the effects of CRF on LC discharge, CRF hypersecretion in depression would
25 1
be predicted to result in altered LC discharge characterized by increased tonic or basal activity and diminished responses to phasic sensory stimuli. Consistent with this, responses to discrete stimuli have been reported to be decreased in depressed patients (Maas et al., 1971; Buchsbaum et af., 1981). One mode of action of antidepressants may be ‘tb reverse the neuronal consequences of CRF hypersecretion by either: (1) Pharmacologically antagonizing CRF in the LC; (2) Producing effects on LC cells which oppose those of CRF (functional antagonism); or (3) Blocking CRF release or hypersecretion. This hypothesis was recently tested by a characterization of the acute and chronic effects of three pharmacologically distinct antidepressants on LC discharge (Valentino et af., 1990; Curtis and Valentino, 1990). Previous studies of the effects of antidepressants on LC discharge focused on spontaneous discharge rate only, and most investigations involved only acute effects (Nyback et al., 1975; Svenson and Usdin, 1978; Olpe et aZ., 1983). In the present study, the acute and chronic effects of three antidepressants on (1) LC spontaneous discharge, (2) LC sensory-evoked discharge, (3) LC activation by a stressor, and (4)LC activation by CRF were quantified. The three antidepressants included desmethylimipramine (DMI), a norepinephrine reuptake inhibitor, sertraline (SER), a serotonin reuptake inhibitor, and mianserin (MIA), an atypical antidepressant that does not block reuptake of either norepinephrine or serotonin. Effects of antidepressants on LC spontaneous discharge
The effects of antidepressants on LC spontaneous discharge rate are variable and probably unrelated to their clinical efficacy. For example, DMI (0.1 and 0.3 mg/kg, iv) decreases LC spontaneous discharge rates in anesthetized rats (Nyback et al., 1975; Svensson and Usdin, 1978; Olpe et al., 1983; Valentino et al., 1990) and chronic administration (10 mg/kg/day, ip for 21
days) results in tolerance to this effect (ScuveeMoree and Dresse, 1979; Valentino et al., 1990). In contrast, MIA (0.01-1.0 mg/kg, iv) increases (Curtis and Valentino, 19901, and SER (1.0 and 3.0 mg/kg, iv) has no effect on LC spontaneous discharge rate (Valentino et aZ., 1990). Chronic administration of MIA (10 mg/kg/day, ip for 21 days) also results in tolerance to its effects on LC spontaneous discharge (Curtis and Valentino, 1990). The finding that these three antidepressants do not consistently alter LC spontaneous discharge rate, and that tolerance usually occurs with chronic administration, suggests that effects of antidepressants on LC spontaneous discharge are unrelated to their clinical efficacy. Effects of antidepressants on LC sensory-evoked discharge
The response of LC neurons to phasic sensory stimuli is a characteristic that has been a basis for hypothesizing a role for the LC in arousal (Foote et aZ., 1980; Aston-Jones and Bloom, 1981; Foote et al., 1983). Generally, in anesthetized animals a potentially noxious stimulus such as footshock is required to evoke LC discharge, while many modes of non-noxious stimuli evoke LC discharge in unanesthetized rats (Aston-Jones and Bloom, 1981). Anatomic and physiological studies indicate that LC activation by footshock is mediated by an excitatory amino acid pathway from the nucleus paragigantocellularis (Ennis and AstonJones, 1988). The typical pattern of LC discharge in response to phasic sensory stimuli is characterized by a brief (80-100 msec) period of increased discharge occurring 15-20 msec after the stimulus, followed by a longer period of relatively decreased discharge (the postactivation pause). All of the antidepressants tested decrease LC discharge evoked by footshock when administered acutely (Fig. 1). For example, DMI decreases LC discharge evoked by footshock, but only at doses that also decrease LC spontaneous discharge rate (Fig. 1). Tonic LC discharge is usually more sensitive to inhibition by DMI than
252
is evoked discharge (Fig. 1). This results in an enhanced signal-to-noise ratio of the LC sensory response. Interestingly, this effect is opposite to that produced by CRF (Valentino and Foote, 1987). SER selectively decreases LC discharge evoked by repeated footshock (Fig. 1). MIA has effects which are similar to those produced by icv CRF, i.e., MIA increases tonic LC discharge and decreases evoked discharge (Fig. 1). The effects of MIA are probabIy due to antagonism of 5HT2, because 5HT2 agonists have the opposite effect (Rasmussen and Aghajanian, 1986). However, the site of action of this effect, i.e., whether it is directly within the LC or on neurons that project to the LC, is not known. While the acute effects of antidepressants are of interest, the chronic effects of antidepressants are more likely to be related to the mechanism of their clinical effect. After chronic administration, 300 1
"
T
D S M
TONIC
O
S
M
EVOKED
D S M
S:N
Fig. 1. Acute effects of antidepressants on locus coeruleus (LC) sensory-evoked discharge. Bars indicate LC spontaneous discharge rate (tonic), discharge rate evoked by repeated footshock (evoked), and the ratio of evoked-to-tonic discharge rate (S:N) expressed as a percentage of the pre-drug rates during trials of repeated footshock (1.3 mA, 0.1 Hz, 0.5 msec duration). The drugs used were desmethylimipramine (DMI) (D, 0.3 mg/kg, n = 171, sertraline (SER) (S, 3.0 rng/kg, n = 7), and mianserin (MIA) (M, 0.1 mg/kg, n = 7). Vertical lines indicate ? 1 S.E.M. **P < 0.005, *P < 0.05. (Modified from Valentino et al., 1990.)
m D
S
M
TONIC
EVOKED
S:N
Fig. 2. Chronic effects of antidepressants on LC sensoryevoked discharge. Open bars indicate the mean spontaneous discharge rate (tonic), mean evoked discharge rate (evoked), and mean ratio of evoked-to-tonic discharge rate (S:N) during trials of repeated footshock of groups of control rats. Solid bars represent the corresponding values for matched groups of rats chronically administered DMI (D, 10 mg/kg, ip for 21 days; n = 28 controls vs 23 chronic rats), SER (S, 10 mg/kg, ip for 21 days; n = 16 controls vs 8 chronic rats) or MIA (M, 10 mg/kg, ip for 21 days; n = 4 1 controls vs 16 chronic rats). Recordings were generated 12-24 h after the last dose. Vertical lines indicate f lS.E.M. *P < 0.05. (Modified from Valentino et al., 1990.)
tolerance develops to the effects of DMI and MIA on LC tonic and sensory-evoked discharge (Fig. 2). Thus, the pattern of LC discharge elicited by repeated footshock is similar in untreated and chronically treated rats. Moreover, administration of the antidepressants to chronically treated rats does not result in significant changes in spontaneous or sensory-evoked discharge (Valentino et al., 1990; Curtis and Valentino, 1990). Similar to rats chronically treated with DMI and MIA, tolerance develops with chronic SER administration to effects on LC sensory-evoked discharge (Fig. 2). Interestingly, in rats chronically treated with SER, LC evoked-discharge is somewhat greater than matched controls and the signal-to-noise
253
ratio of the sensory response is significantly greater (Fig. 2). The enhanced signal-to-noise, observed only after chronic treatment with SER, is opposite to that produced by CRF (Valentino and Foote, 1987, 19881, suggesting that the effects of CRF on LC sensory-evoked discharge might be opposed in rats chronically treated with SER. Thus, SER has the potential to act as a functional antagonist of CRF.
s !-
B
0
w
8
2 8 2 W W
c w I-
Effects of antidepressants on LC discharge elicited by stress
Nitroprusside infusion (10 ~ g / 3 0pI/min) increases LC spontaneous discharge and disrupts LC responses to phasic sensory stimuli in a manner similar to CRF (Valentino and Wehby, 1988). These effects are temporally correlated to hypotension and are probably mediated by endogenous CRF because they are prevented by prior administration of a CRF antagonist (Valentino and Wehby, 1988; Valentino, 1989). Although acute administration of DMI (0.1 mg/kg, iv) decreases LC spontaneous discharge rate, it does not inhibit LC activation by nitroprusside (Fig. 3). In contrast, the neuronal effects of nitroprusside are greatly attenuated in rats chronically administered DMI (10 mg/kg/day for 21 days) (Fig. 3). This is not due to changes in the magnitude of the stressor because nitroprusside produces a similar magnitude of hypotension in chronic DMI rats as in untreated rats (Valentino et al., 1990). Nor is the attenuation of LC activation due to repeated injections and handling of rats, because nitroprusside increases LC spontaneous discharge of rats chronically administered saline (Fig. 3). Finally, the attenuation of nitroprusside-induced LC activation is not due to antagonism of CRF at binding sites because the efficacy of CRF in increasing LC spontaneous discharge is comparable in untreated and chronic DMI rats (Fig. 4). These results suggest that chronic administration of DMI may diminish stress-elicited CRF release. If this effect generalizes to other stressors it could have important clinical implications, partic-
d
w
8
a
I0
2 Y .
--
CON
SAL
aDMl
CDMl
cSER
aMA
cMA
Fig. 3. Effect of nitroprusside infusion on LC spontaneous discharge rate in rats treated with different antidepressants. Bars indicate the effect of nitroprusside on LC spontaneous discharge rate expressed as a percentage of the preinfusion rate, 6 min after the initiation of infusion in 15 untreated rats (CON), 12 rats chronically administered saline (SAL), 8 rats acutely administered DMI (aDMI), 12 rats chronically administered DMI (cDMI), 9 rats chronically administered SER (cSER), 11 rats acutely administered MIA (aMIA), and 8 rats chronically administered MIA (cMIA). Chronic administration of DMI and acute and chronic administration of MIA attenuated nitroprusside-elicited LC activation. Vertical lines indicate k 1S.E.M. *P < 0.05. (Modified from Valentino et al., 1990.)
ularly if abnormal CRF secretion is a primary defect in depression. Like DMI, MIA also attenuates stress-elicited LC activation (Fig. 3). However, MIA is unique in that it attenuates both nitroprusside- and CRFinduced LC activation when administered acutely (Curtis and Valentino, 1990). Preliminary studies indicate that MIA is not an antagonist at CRF binding sites (DeSouza, personal communication). Thus, acute administration of MIA decreases LC activation elicited by a variety of stimuli which include footshock, hemodynamic stress, and CRF. Although inhibition of nitroprusside’s effect is probably related to inhibition of CRF, it is not clear whether similar mechanisms are involved in attenuation of footshock. It is possible that MIA
80
References
IT
"
CON
III SAL
DMI
SER
MA
Fig. 4. Effect of icv corticotropin-releasing factor (CRF) (3.0 F g ) on LC spontaneous discharge rate of rats chronically administered different antidepressants. Bars indicate the increase in LC spontaneous discharge rate produced by icv CRF expressed as a percentage of the pre-CRF discharge rate, in 12 untreated rats (CON), 6 rats chronically administered saline (SAL), 24 rats chronically administered DMI, 5 rats chronically administered SER, and 8 rats chronically administered MIA. Chronic administration of the antidepressants does not alter the efficacy of CRF. Vertical lines indicate f lS.E.M.
decreases transmission in the final common pathway for all of these stimuli. Alternatively, MIA may alter LC neurons in such a way that they are less responsive to a variety of stimuli. Chronic administration of MIA results in tolerance to all of its effects except attenuation of nitroprussideinduced LC activation (Fig. 3). Therefore, chronic administration of MIA is similar to DMI in that both attenuate stress-elicited LC activation which requires endogenous CRF. It is of interest that chronic administration of SER does not affect LC activation by nitroprusside. Acknowledgements
This work was supported by PHS Grants MH40008 and MH42796 and a NARSAD fellowship to Dr. Curtis.
Aston-Jones, G. and Bloom, F.E. (1981) Norepinephrine-containing neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Britton, K., Morgan, J., Rivier, J., Vale, W. and Koob, G. (1985) Chlordiazepoxide attenuates CRF-induced response suppression in the conflict test. Psychopharmacolo m , 86: 170-174. Brown, M.R., Fisher, L.A., Spiess, J., Rivier, C. and Vale, W. (1982) Corticotropin-releasing factor: Actions on the sympathetic nervous system and metabolism. Endocrinology, 111: 928-931. Buchsbaum, M.S., Muscettola, G. and Goodwin, F.K. (1981) Urinary MHPG, stress response, personality factors and somatosensory evoked potentials in normal subjects and patients with major affective disorders. Neuropsychobiology, 7: 212-224. Bunney, W.E., Jr. and Davis, J.M. (1965) Norepinephrine in depressive reactions. Arch. Gen. Psychiatry, 13: 483-494. Carroll, B. (1982) The dexamethasone suppression test for melancholia. Br. J. Psychiatry, 140: 292-304. Cassens, G., Roffman, M., Kuruc, A., Orsulak, P.J. and Schildkraut, J.J. (1980) Alterations in brain norepinephrine metabolism induced by environmental stimuli previously paired with inescapable shock. Science, 209: 11381139. Cummings, S., Elde, R., Ells, J. and Lindall, A. (1983) Corticotropin-releasing factor immunoreactivity is widely distributed within the central nervous system of the rat: An immunohistochemical study. J. Neurosci., 3: 1355-1368. Curtis, A.L. and Valentino, R.J. (1990) Mianserin attenuates locus coeruleus (LC) activation elicited by phasic sensory stimuli, hemodynamic stress, and i.c.v. corticotropin-releasing factor (CRF): Possible mode of antidepressant action. SOC. Neurosci. Abstr., 16: 913. Ennis, M. and Aston-Jones, G. (1988) Activation of locus coeruleus from nucleus paragigantocellularis: A new excitatory amino acid pathway in brain. J. Neurosci., 8: 36443657. Fisher, L.A., Rivier, J., Rivier, C., Spiess, J., Vale, W. and Brown, M.R. (1982) Corticotropin-releasing factor (CRF): Central effects on mean arterial pressure and heart rate in rats. Endocrinology, 110: 2222-2224. Foote, S.L. and Cha, C.1. (1988) Distribution of corticotropin-releasing factor-like immunoreactivity in brainstem of two monkey species (saimiri sciureus and macaca fasiculark): An immunohistochemical study. J. Comp. Neurol., 216: 239-264. Foote, S.L., Aston-Jones, G. and Bloom, F.E. (1980) Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proc. Natl. Acad. Sci. USA, 17: 3033-3037. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Reo., 63: 844-914.
255 Gold, P.W., Loriaux, D.L., Roy, A., Kling, M.A., Calabrese, J.R.. Kellner, C.H., Nieman, L.K., Post, R.M., Pickar, D., Gallucci, W., Averginos, P., Paul, S . , Oldfield, E.H., Cutler, G.B., Jr. and Chrousos, P.G. (1986) Responses to cortico-releasing hormone in the hypercortisolism of depression and Cushings disease. N. Engl. J. Med., 314 1329-1342. Heninger, G.R. and Charney, D.S. (1987) Mechanism of action of antidepressant treatments: Implications for the etiology and treatment of depressive disorders. In H.Y. Meltzer (Ed.), Psychopharmacology: The Third Generation of Progress, Raven Press, New York, pp. 535-544. Holsboer, F., Gerken, A,, von Bardeleben, U., Grimm, W., Beyer, H., Muller, O.A. and Stalla, G.K. (1986) Human corticotropin-releasing hormone in depression: Correlation with thyrotropin secretion following thyrotropin-releasing hormone. Bid. Psychiatry, 21: 601-611. Korf, J., Aghajanian, G.K. and Roth, R.H. (1973) Increased turnover of norepinephrine in the rat cerebral cortex during stress: Role of the locus coeruleus. Neuropharmacology, 1 2 933-938. Lesch, K.P., Laux, G., Schulte, H.M., Pfuller, H. and Beckrnann, H. (1988) Corticotropin and cortisol response to human CRH as a probe for HPA system integrity in major depressive disorder. Psychiatry Res., 24: 25-34. Maas, J.W., Dekirmenjian, H. and Fawcett, J. (1971) Catecholamine metabolism, depression and stress. Nature (London), 230: 330-331. Nemeroff, C., Widerlov, E., Bissette, G.T., Walleus, H., Karlson, I., Eklund, K. Kilts, C., Loosen, P. and Vale, W. (1984) Elevated concentrations of CSF corticotropin-releasing factor-like immunoreactivity in depressed patients. Science, 226: 1342-1344. Nyback, H.V., Walters, J.R. and Aghajanian, G.K. (1975) Tricyclic antidepressants: Effects on the firing rate of brain noradrenergic neurons. Eur. J. Pharmacol., 3 2 302312. Olpe, H.R., Jones, R.S.G. and Steinmann, M.W. (1983) The locus ceruleus: Actions of psychoactive drugs. Experentia, 39: 242-249. Oswald, I., Brezinova, V. and Dunleavy, D.L.F. (1972) On the slowness of action of tricyclic antidepressant drugs. Br. J. Psychiatry, 120: 673-677. Overall, J.E., Hollister, L.E., Pokorny, A.D., Casey, J.F. and Katz, G. (1962) Drug therapy in depressions. Clin. Pharmacol. Ther., 3: 16-22. Plotsky, P.M. and Vale, W. (1984) Hemorrhage-induced secretion of corticotropin-releasing factor-like immunoreactivity into the rat hypophyseal portal circulation and its inhibition by glucocorticoids. Endocrinology, 114: 164-169. Post, R.M., Kotin, J. and Goodwin, F.K. (1974) The effects of cocaine on depressed patients. Am. J. Psychiatry, 131: 511-517. Rasmussen. K. and Aghajanian, G.K. (1986) Effects of hallucinogens on spontaneous and sensory-evoked locus coeruleus unit activity in the rat: Reversal by selective 5-HT2 antagonists. Brain Rex, 385: 395-400.
Richelson, E. and Pfenning, M. (1984) Blockade by antidepressants and related compounds of biogenic amine uptake into rat brain synaptosomes: Most antidepressants selectively block norepinephrine uptake. Eur. J. Pharmacol., 104: 277-286. Sakanaka, M., Shibasaki, T. and Lederes, K. (1987) Corticotropin-releasing factor-like immunoreactivity in the rat brain as revealed by a modified cobalt-glucose oxide-diaminobenzidene method. J. Comp. Neurol., 260: 256-298. Sachar, E.J., Hellman, L., Fukushima, D.K. and Gallagher, T.F. (1970) Cortisol production in depressive illness. Arch. Gen. Psychiatry, 23: 289-298. Schildkraut, J.J. (1965) The catecholamine hypothesis of affective disorders: A review of supporting evidence. Am. J. Psychiatry, 122: 509-522. Scuvee-Moreau, J.J. and Dresse, A.E. (1979) Effect of various antidepressant drugs on the spontaneous firing rate of locus coeruleus and raphe dorsalis neurons of the rat. Eur. . I . Pharmacol., 57: 219-225. Solomon, R.A., McCormack, B.M., Lovitz, R.N., Swift, D.M. and Hegemann, M.T. (1986) Elevation of brain norepinephrine concentration after experimental subarachnoid hemorrhage. Neurosurgery, 19: 363-366. Stokes, P.E. and Sikes, C.R. (1987) Hypothalamic-pituitaryadrenal axis in affective disorders. In H.Y. Meltzer, (Ed.), Psychopharmacology: The Third Generation of Progress, Raven Press, New York, pp. 589-607. Sutton, R.E., Koob, G.F., LeMoal, M., Rivier, J. and Vale, W. (1982) Corticotropin-releasing factor produces behavioral activation in rats. Nature (London), 297: 331-333. Svensson, T.H. (1987) Peripheral, autonomic regulation of locus coeruleus noradrenergic neurons in brain: Putative implications for psychiatry and psychopharmacology. Psychopharmacology, 9 2 1-7. Svensson, T.H. and Usdin, T. (1978) Feedback inhibition of brain noradrenaline neurons by tricyclic antidepressants: Receptor mediation. Science, 202: 1089-1091. Swanson, L.W., Sawchenko, P.E., Rivier, J. and Vale, W.W. (1983) Organization of ovine corticotropin-releasing factor immunoreactive cells and fibers in rat brain: An immunohistochemical study. Neuroendocrinology, 36: 165-186. Thierry, A.M., Javoy, F., Glowinski, J. and Kety, S.S. (1968) Effects of stress on the metabolism of norepinephrine, dopamine and serotonin in the central nervous system of the rat: Modification of norepinephrine turnover. J. Pharmacol. Exp. Ther., 163: 163-171. Vale, W., Spiess, J. and Rivier, C. (1981) Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin. Science, 213: 1394-1397. Valentino, R.J. (1988) CRF effects on central noradrenergic neurons: Relationship to stress. In G.P. Chrousos, D.L. Loriaux and P.W. Gold (Eds.), Mechanisms of Physical and Emotional Stress, Plenum Press, New York, pp. 47-64. Valentino, R.J. (1989) Corticotropin-releasing factor: Putative neurotransmitter in the noradrenergic nucleus locus ceruleus. Psychopharmacol. Bull., 25: 306-3 11.
256 Valentino, R.J. and Foote, S.L. (1987) Corticotropin-releasing factor disrupts sensory responses of brain noradrenergic neurons. Neuroendocrinology, 45: 28-36. Valentino, R.J. and Foote, S.L. (1988) Corticotropin-releasing hormone increases tonic hut not sensory-evoked activity of noradrenergic locus ceruleus neurons in unanesthetized rats. J. Neurosci., 8: 1016-1025. Valentino, R.J. and Wehby, R.G. (1988) Corticotropin-releasing factor: Evidence for a neurotransmitter role in the
locus ceruleus during hemodynamic stress. Neuroendocrinology , 48: 614-671. Valentino, R.J., Foote, S.L. and Aston-Jones, G. (1983) Corticotropin-releasing factor activates noradrenergic neurons of the locus ceruleus. Brain Rex, 270: 363-367. Valentino, R.J., Curtis, A.L., Parris, D.G. and Wehby, R.G. (1990) Antidepressant actions of brain noradrenergic neurons. J. Pharmacol. Exp. Ther., 253: 833-840.
C.D. Barnes and 0. Pompeiano (Eds.) Prugress in Bruin Research, Vol. 88 0 1991 tlsevier Science Publishers B.V.
257 CHAPTER 20
Selective effects of DSP-4 on locus coeruleus axons: are there pharmacologically different types of noradrenergic axons in the central nervous system? J.-M. Fritschy and R.Grzanna Department of Neuroscience, Johns Hopkins Unirxrsity School of Medicine, N. Wove St., Baltimore, MD, U.S.A.
There is considerable evidence from biochemical studies that the transmitter-depleting action of drugs and neurotoxins which act upon central noradrenergic (NA) axon terminals is not uniform in different brain regions. Among NA axons, those originating in the locus coeruleus (LC) have been proposed to be most susceptible to the action of NA neurotoxins such as N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine (DSP-4). The studies described here were conducted to determine whether this differential susceptibility to DSP-4 reflects a pharmacological heterogeneity between different populations of NA axons. To determine whether DSP-4 acts selectively upon LC axons, we have characterized the effects of this drug on NA axons in different brain regions, by using noradrenaline and dopamine-P-hydroxylase ( D P H ) immunohistochemistry. Following systemic administration of DSP-4, there was an almost complete loss of noradrenaline and D P H
staining in brain regions innervated by LC axons. No effects of the drug treatment were detected in brain regions innervated primarily by non-coerulean N A axons. These results demonstrate that both the transmitter-depleting and the neurodegenerative action of DSP-4 are restricted to N A axons originating in the LC. To explore the basis for this selectivity, noradrenaline uptake studies were conducted using synaptosomes from brain regions in which NA axons differ in their response to DSP-4. The results reveal a significant difference in the affinity of DSP-4 for the noradrenaline uptake carrier in cortical and hypothalamic synaptosomes. This finding is compatible with the hypothesis that the noradrenaline uptake carrier is pharmacologically distinct in LC and non-coerulean NA axons. This heterogeneity in noradrenaline uptake raises the question whether other drugs may also have differential actions on LC and non-coerulean NA neurons.
Key words: neurotoxicity, degeneration, immunohistochemistry, noradrenaline uptake, locus coeruleus
Introduction
Many widely prescribed drugs exert their therapeutic effects by interaction with noradrenergic (NA) neurons. Observations that central NA neurons may differ in their response to pharmacological manipulation date back to the pioneering
studies by Dahlstrom and Fuxe and their colleagues (1964; Carlsson et al., 1966; Anden, 1967). Using drugs which influence noradrenaline synthesis, storage and release, they observed nonuniform effects on NA axons in different brain regions. For example, Carlsson et al. (1966) documented that the noradrenaline-releasing effects
258
of amphetamine are more pronounced in cerebral cortex than in hypothalamus and brainstem. Despite the considerable implications of these observations for understanding the functional organization of central NA neurons, these studies did not stimulate further efforts to characterize the differential responses of NA neurons to therapeutic drugs. Pronounced regional differences in the response of NA axons were also observed with catecholamine neurotoxins (Iversen and Uretsky, 1970; Sachs and Jonsson, 1972;Jonsson and Sachs, 1976; Pappas et al., 1976). After systemic injections of 6-hydroxydopa,Jacobowitz and Kostrzewa (1971) observed long-lasting reductions in noradrenaline levels in cerebral cortex, hippocampus and cerebellum, but found few effects in the pons, medulla, hypothalamus, preoptic area and septum. In their comprehensive review on the pharmacology of 6-hydroxydopamine, Kostrzewa and Jacobowitz (1974) emphasized, among the NA axons, that LC axons are most susceptible to the effects of this drug. A striking example for distinct responses of LC and non-coerulean NA axons to 6-hydroxydopamine was reported by McBride et al. (1985). These authors demonstrated that systemic administration of this neurotoxin to neonatal rats results in selective elimination of the coeruleo-spinal projection, whereas descending axons of A5 and A7 NA neurons were spared. Further evidence that the action of neurotoxins may be more pronounced on LC axons has been reported for 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine(MPTP; Hallman et al., 1985; Forno et al., 1986) and N-(2-chloroethy1)N-ethyl-2-bromobenzylamine (DSP-4; Jonsson et al., 1981; Jonsson, 1983; Logue et al., 1985). For instance, these studies provided evidence that the noradrenaline-depleting effects of DSP-4 are more pronounced in cerebral cortex, cerebellum and spinal cord than in hypothalamus and brainstem. Commenting on the correlation between the regional effects of neurotoxins on noradrenaline levels and the distribution of LC axons, several
authors stressed that these drugs do not have an absolute specificity for LC axons, and that noncoerulean NA fibers are also affected, but to a lesser degree (Jonsson and Sachs, 1973; Jonsson et al., 1981; Kostrzewa, 1988). These investigators argued that the pharmacological properties of NA axons are identical in all brain regions, and that structural features account for the differential response of NA axons to neurotoxins. Most investigators have employed biochemical assays to evaluate the effects of NA neurotoxins in various brain areas. However, in regions where a drug causes only partial depletion of noradrenaline, it is not possible to determine with biochemical assays whether a partial decrease in transmitter levels reflects an incomplete loss of noradrenaline from all NA axons or a complete loss from a subpopulation of NA axons. Thus a potential selectivity of neurotoxins for different populations of NA axons may have escaped detection. Unlike biochemical assays, immunohistochemistry permits direct assessment of the effects of drugs on individual NA axons. The recent introduction of specific antibodies to noradrenaline (Geffard et al., 1986) makes it possible to visualize NA axons directly based upon their neurotransmitter content. The studies described below were conducted to evaluate, with an anatomical method, the responses of NA axons to DSP-4. This drug was chosen because of its unique ability to cross the blood-brain barrier and to induce selective loss of NA neuron markers without affecting dopaminergic neurons (Ross, 1976; Ross and Renyi, 1976). To determine whether DSP-4 acts selectively upon NA axons of the LC, we have characterized the transmitter-depleting and the neurodegenerative actions of this neurotoxin on NA axons in different brain regions. Both noradrenaline and dopamine-P-hydroxylase (DPH) immunohistochemistry were used at the light microscopy level to visualize the effects of DSP-4 on these two markers of NA axons over a two-week period. The rapid loss of noradrenaline induced by DSP-4 is likely to reflect an interaction between the drug
259
and NA axon terminals (Landa et al., 1984). In contrast, DPH is a membrane-bound protein and may serve as an index of the structural integrity of NA axons. By characterizing the sequence of events that follows exposure to DSP-4, we sought to determine whether the selectivity of this drug for LC axons can be observed in relation to its transmitter-releasing effects as well as its neurodegenerative action. Furthermore, by using an in vitro uptake assay, we explored whether the preferential action of DSP-4 on LC axons reflects
a pharmacological difference between LC and non-coerulean NA axons. Rats were treated with a single systemic injection of DSP-4 (50 mg/kg). Noradrenaline and DPH were visualized in brain sections using specific antibodies in combination with the avidinbiotin peroxidase method of Hsu et al. (1981) (see Fritschy and Grzanna, 1989, and Fritschy et al., 1990 for details). Antibodies to noradrenaline were kindly provided by Dr. M. Geffard, Bordeaux, France (Geffard et al., 1986); antibodies
Fig. 1. A pair of dark-field photomicrographs illustrating the acute effects of DSP-4 on NA axons in cingulate cortex. A, control; B, 24 h after drug administration (50 mg/kg). Sections were stained with antibodies to noradrenaline in combination with the imrnunoperoxidase method. Although NA axons are barely detectable in drug-treated animals, their morphology is similar to that seen in control. Bar = 100 pm.
260
to rat D P H were prepared and characterized in our laboratory (Grzanna and Coyle, 1976). Acute effects of DSP-4 evidence for a selective loss of noradrenaline from locus coeruleus axon terminals
Changes in noradrenaline levels were measured by high performance liquid chromatography in four brain regions at 6 h, 24 h and 2 weeks after DSP-4 treatment (Grzanna et al., 1989). This biochemical analysis confirmed that DSP-4 induces rapid and profound loss of noradrenaline in brain regions innervated by the LC (e.g., cerebral cortex and cerebellum), but has only moderate effects in brain regions innervated by noncoerulean NA axons (ventral forebrain and hypothalamus). Processing of brain sections with noradrenaline immunohistochemistry revealed that DSP-4 induces profound loss of NA axon staining within hours after treatment (Grzanna et al., 1989; Fritschy et al., 1990). In sections from untreated rats, NA axons form extensive plexuses of fine but intensely stained varicose fibers. Six hours after DSP-4 treatment, NA axons were only faintly stained; after 24 h, they were barely detectable with noradrenaline immunohistochemistry (Fig. 1). The most remarkable characteristic of this effect of DSP-4 is its striking regional specificity. The loss of NA axon staining was restricted to brain regions innervated by the LC. These regions include the entire neocortex and hippocampus (Figs. 1-2), olfactory bulb, thalamus (except its paraventricular nucleus), tectum (Fig. 3A,B), cerebellum, brainstem sensory nuclei and spinal cord dorsal horn. In contrast, NA axon staining seemed unaffected by the drug treatment in regions innervated by non-coerulean NA axons. The staining remained unchanged in the basal forebrain, hypothalamus, brainstem motor and autonomic nuclei (Fig. 3C,D), raphe nuclei, spinal cord intermedio-lateral cell column and ventral horn. Few brain regions were only partially depleted. They include the septum, amygdala, pon-
tine nuclei and the principal trigeminal sensory nucleus. The loss of NA axon staining induced within hours by DSP-4 in regions innervated by the LC persisted for the 2-week time period studied here. After 4 days, however, a few degenerating NA axons were seen in these areas. Unlike control, these fibers were thick and swollen. At 7 days, many of these abnormal axons were studded with fine axonal sprouts (Fig. 4). After 10-14 days, most of these axons had disappeared and the loss of NA axon staining was complete. The demonstration that non-coerulean NA axons are insensitive to the action of DSP-4 offers a plausible explanation for the partial depletion of noradrenaline induced by DSP-4 in the ventral forebrain, hypothalamus and brainstem, as revealed by biochemical studies (Jaim-Etcheverry and Zieher, 1980; Jonsson et al., 1981; Hallman et al., 1984; Logue et al., 1985; Grzanna et aL, 1989). Long-term effects of DSP-4 degeneration of LC axon terminals
In contrast to the profound loss of noradrenaline immunoreactivity induced within hours by DSP-4, the structural integrity of NA axons, as visualized by D P H immunohistochemistry, was unchanged until day 4 after drug treatment (Fritschy et al., 1990). During this time span, no effect of the drug treatment could be observed on NA axons by D P H staining. Thereafter, loss of D P H immunoreactivity was abrupt and exhibited the same regional pattern as the loss of noradrenaline described above (see Fritschy and Grzanna, 1989, for dqtails). After this rapid disappearance of DPH staining, swollen and distorted fibers appeared transiently. The distribution and morphology of these degenerating axons was similar to those seen with antibodies to noradrenaline (Fig. 4). These fibers were mainly found along the distal end of ascending LC axon pathways, notably in the cingulum, around the genu of the corpus callosum, the fornix and the stria
261
Fig. 2. A pair of dark-field photomicrographs depicting the DSP-4-induced loss of NA axon staining in the hippocampus. A, control; B, 7 days after drug treatment. Note that the loss of NA axons is uniform and virtually complete, even in the densely innervated hilus of the dentate gyrus. Immunoperoxidase staining with anti-noradrenaline antibodies. Bar = 100 Km.
262
Fig. 3. Dark-field photographs illustrating the differential effects of DSP-4 on LC and non-coerulean NA axons. A and C, control; B and D, 7 days after drug treatment. The top row depicts the loss of NA axons in the inferior colliculus. These axons originate in the LC. The bottom row depicts the sparing of the NA innervation in the motor nucleus of the trigeminal nerve. These axons originate in the A7 group. Sections shown in B and D are from the same animal. Immunoperoxidase staining with anti-noradrena-
263
swellings and intense staining of these axonal segments presumably indicate accumulation of anterogradely transported material proximal to the lesion induced by DSP-4. Two weeks after DSP-4 treatment, most of these abnormal fibers were no longer seen, leaving CNS regions innervated by the LC virtually devoid of NA axons. The loss of NA axon staining by noradrenaline and DPH immunohistochemistry suggests that the effects of DSP-4 are initially restricted to the terminal portion of LC axons. In our studies, staining of LC neurons and their preterminal axons appeared unaffected during the first two weeks after drug treatment. In Nissl-stained sections, the morphology of LC perikarya was similar to control. However, a quantitative analysis of the number of LC neurons two weeks after drug treatment revealed a 10-20% loss of cell bodies (Fritschy and Grzanna, in preparation). Moreover, in a preliminary study, we observed more pronounced cell loss in the LC (40% on average six months after DSP-4 treatment; Fritschy and Grzanna, 1991). What is the basis for the selectivity of DSP-4 for locus coeruleus axons?
Fig. 4. Bright-field photomicrograph illustrating a degenerating NA axon in cerebral cortex visualized by noradrenaline immunohistochemistry 7 days after DSP-4 treatment. Note the numerous fine axonal sprouts. Immunoperoxidase staining. Bar = 20 Fm.
medullaris. The distribution of swollen axons suggested that they represent the distal portion of preterminal LC axons (Fritschy et aL, 1990). The
A striking property of DSP-4, revealed by noradrenaline as well as DPH immunohistochemistry, is the regional specificity of its effects on NA axons. At all levels of the neuraxis, there is a strong correlation between the pattern of DSP-4induced NA axon degeneration and the distribution of LC fibers (Fritschy and Grzanna, 1990a,b; see Foote et al., 1983; Bjorklund and Lindvall, 1986, for reviews). This correlation strongly suggests that the effects of this drug are restricted to LC axons. As noted in the Introduction, several authors suggested that the differential effects of neurotoxins on NA axons do not reflect differences in the pharmacological or biochemical properties of these fibers (Jonsson and Sachs, 1973; Jonsson et al., 1981; Kostrzewa, 1988). Based on the observation that LC and non-coerulean NA axons are
264
265
morphologically distinct, Sachs and Jonsson (1975) suggested that different surface/volume ratios of varicosities may account for their differential susceptibility. It has also been proposed that the density of NA axons within a given terminal field may influence their vulnerability (Kostrzewa, 1988). This suggestion stems from the observation that non-coerulean NA axons tend to form dense fiber plexuses, notably in the hypothalamus, basal forebrain and raphe nuclei, while terminal LC axons are more sparsely distributed. Our findings indicate, however, that the density of NA axons is not necessarily correlated with their susceptibility to DSP-4. For instance, LC axons form dense plexuses in the hilus of the dentate gyrus of the hippocampus (Fig. 2) and in the anteroventral nucleus of the thalamus. These axons are, nevertheless, almost completely eliminated by the drug. Conversely, non-coerulean NA axons form relatively scattered plexuses of terminal branches in the septum and in the brainstem reticular formation, for example, regions in which they are spared by the drug treatment. That factors such as axonal morphology and distribution density do not play a critical role in the differential toxicity of DSP-4 is illustrated in Figure 5 , which depicts different responses of NA axons to DSP-4 in two adjacent nuclei of the amygdaloid complex. DSP-4 treatment produces a profound loss of NA axons in the basolateral nucleus but spares the NA innervation of the adjacent central nucleus (see Fig. 6 for orientation). Differences in endogenous transmitter concentration have also been suggested as a factor contributing to the differential vulnerability of NA axons to neurotoxins. Sachs and Jonsson (1975) suggested that the toxicity of 6-hydroxydopamine is influenced by the amount of noradrenaline
Fig. 6 . Schematic drawing of a hemisection through the rat brain illustrating the location of the photographic fields depicted in Figure 5. (From Paxinos and Watson, 1986). Abbreviations: BL, basolateral amygdaloid nucleus; Ce, central amygdaloid nucleus; CPu, caudate putamen; ic, internal capsule; mfb, medial forebrain bundle; ot, optic tract; Pir, piriform cortex; VMH, ventromedial hypothalamic nucleus; VPL, ventral posterolateral thalamic nucleus.
present in the cytoplasmic pool. Drugs, such as tyrosine hydroxylase inhibitors, which reduce the concentration of transmitter in this pool potentiate 6-hydroxydopamine toxicity in sympathetic nerve terminals (Sachs et al., 1975). According to Sachs and Jonsson (1975), the greater susceptibility of LC axons to 6-hydroxydopamine may be due to the lower levels of endogenous noradrenaline present in these fibers compared to non-coerulean NA axons. In the case of DSP-4, neither the vesicular nor the cytoplasmic pool of noradrenaline appear to play a significant role in the toxicity of this drug, since its effects are not influenced by pretreatment with amphetamine (Hallman and Jonsson, 1984) or with reserpine (5
Fig. 5. A pair of dark-field photomicrographs of the amygdala depicting the regional selectivity of the effects of DSP-4 on NA axons. Orientation landmarks are provided in Figure 6. A, control. Note that the distribution density and the morphology of NA axons are similar in both the central nucleus (left) and basolateral nucleus (right). B, treated rat 14 days after drug administration. DSP-4 induces selective loss of NA axons in the basolateral nucleus, leaving those innervating the central nucleus largely unaffected. Immunoperoxidase staining with antibodies to DPH. Bar = 200 p m .
266
mg/kg, 24 h, personal observations). DSP-4 treatment increases calcium-dependent release of transmitter evoked in vitro by electrical stimulation, but blocking this effect does not influence the neurotoxic action of the drug (Landa et al., 1988). These observations suggest that the level of endogenous noradrenaline within NA axon terminals is unlikely to be a critical factor contributing to the differential effects of DSP-4 on NA axons. By comparing the effects of DSP-4 on noradrenaline and DPH staining at different time points, we were able to distinguish two phases in the action of DSP-4. The acute depletion of transmitter from LC axons, which occurs within hours after drug treatment, is likely to be related to the interaction of DSP-4 with the noradrenaline uptake carrier. The subsequent loss of DPH staining, which occurs several days later, represents the neurodegenerative phase of DSP-4. The structural disintegration of LC axons is evidenced by the rapid disappearance of DPH staining and by the transient appearance of degenerating fibers (Fig. 4). The massive loss of noradrenaline induced by DSP-4 represents a first, and perhaps crucial step in the neurotoxic action of the drug. Since DSP-4 does not induce loss of transmitter from non-coerulean NA axons, our findings suggest that they may be pharmacologically distinct from LC axons. Evidence that the differential effects of DSP-4 may be due to differences in the pharmacological properties of locus coeruleus and non-coerulean noradrenergic axons
An interaction of DSP-4 with the noradrenaline uptake carrier is required for the drug to exert its neurotoxic effects (Hallman and Jonsson, 1984; Ross, 1987). Several studies have related the toxicity of DSP-4 to its alkylating properties and suggested that the key target of this drug is the noradrenaline uptake carrier itself (Ross, 1976; Lee et al., 1982). It is unknown whether the destruction of the noradrenaline transporter by
DSP-4 is sufficient to induce NA axon degeneration. Such a mechanism has recently been proposed by Roffler-Tarlov et al. (1990) for the loss of dopaminergic axons in the striatum of weaver mice. Since the noradrenaline uptake carrier is a principal target of DSP-4 acting on NA axons, one might argue that the selectivity of this drug is best explained by a pharmacological difference in noradrenaline uptake in LC and non-coerulean NA axons. The profound difference in the susceptibility to DSP-4 of LC axons in cerebral cortex and non-coerulean NA axons in the hypothalamus prompted us to test the hypothesis that the noradrenaline uptake carriers in these two regions may be pharmacologically distinct (Zaczek et al., 1990). The characteristics of noradrenaline uptake were compared in synaptosomal preparations from cerebral cortex and hypothalamus. The inhibitory effects of two concentrations of DSP-4 (100 and 300 nM) on the uptake of [ 3H]noradrenaline were measured following a 5 min incubation period at 37°C. Two major findings emerged from this study: (i) the inhibitory action of DSP-4 on noradrenaline transport is considerably more potent in cortical synaptosomes than in hypothalamic synaptosomes ( K , = 179 A 39 and 460 k 35 nM, respectively); (ii) during this 5 min period, DSP-4 acts as a competitive inhibitor of noradrenaline uptake. Altogether, these findings are compatible with the hypothesis that DSP-4 is transported into synaptosomes (Jonsson et al., 1983, and that more DSP-4 enters into cortical NA axon terminals (which originate in the LC) than into hypothalamic NA axon terminals (which originate in non-coerulean NA neurons). The results of these uptake studies demonstrate a significant difference in the affinity of DSP-4 for the noradrenaline uptake carrier in cortical and hypothalamic synaptosomes. Whether this pharmacological distinction is indeed the basis for the selective action of DSP-4 upon LC axons remains to be determined. It would be of considerable interest to study the pharmacology of noradrenaline transport into synaptosomes iso-
267
lated from other brain regions and to analyze whether other NA neurotoxins exhibit a similar differential affinity for the NA uptake carrier. Conclusions
By demonstrating that DSP-4 acts specifically upon LC axons, the results suggest the existence of subsets of central NA neurons with different pharmacological properties. The evidence summarized here, that NA axons are pharmacologically heterogeneous, implies that other drugs may also have differential actions on LC and noncoerulean NA neurons. A reinvestigation of the effects of neurotoxins such as 6-hydroxydopa or MPTP on NA axons using immunohistochemical methods may confirm the present conclusions. DSP-4 has gained little popularity presumably because it does not produce uniform depletions of noradrenaline throughout the brain. Our results suggest that the differential effects of this drug, rather than being an inconvenience, represent a valuable tool to uncover the functional organization of central NA systems. Moreover, the confirmation of differential responses of NA neurons to other drugs may lead to more specific treatment protocols of neuropsychiatric disorders. Acknowledgements
This research was supported by USPHS Grant NIMH MH-41977 and by a Biomedical Research Support Grant #RR5378. We are grateful to Dr. S.B. Ross, Astra Lakemedel AB, Sweden for a generous gift of DSP-4 and to Dr. M. Geffard, Bordeaux for providing antibodies to noradrenaline. J.M.F. is a recipient of a fellowship from the Swiss National Fund for Scientific Research. References Anden, N.E. (1967) Effects of reserpine and a tyrosine hydroxylase inhibitor on the monoamine levels in different
regions of the rat central nervous system. Eur. J. Pharmacol., 1: 1-5. Bjorklund, A. and Lindvall, 0. (1986) Catecholaminergic brain stem regulatory systems. In V.B. Mountcastle, F.E. Bloom and S.R. Geiger (Eds), Handbook of Physiology. Section I , Vol. IV: The Nervous System, American Physiological Society, Bethesda, pp. 155-235. Carlsson, A,, Lindqvist, M., Fuxe, K. and Hamberger, B. (1966) The effect of (+)-amphetamine on various central and peripheral catecholamine-containing neurons. J. Pharm. Pharmacol., 18: 128-132. Dahlstrom, A,, and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central newous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Scand., 62, suppl., 232: 1-55. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus ceruleus: New evidence of anatomical and physiological specificity. Phyiol. Reu., 63: 844-915. Forno, L.S., Langston, J.W., DeLanney, L.E., Irwin, I. and Ricaurte, G.A. (1986) Locus coeruleus lesions and eosinophilic inclusions in MPTP-treated monkeys. Ann. Neurol., 20: 449-455. Fritschy, J.-M. and Grzanna, R. (1989) Immunohistochemical analysis of the neurotoxic effects of DSP-4 identifies two populations or noradrenergic axon terminals. Neuroscience, 30: 181-197. Fritschy J.-M. and Grzanna, R. (1990a) Demonstration of two separate descending noradrenergic pathways to the rat spinal cord: Evidence for an intragriseal trajectory of locus coeruleus axons in the superficial layers of the dorsal horn. J. Comp. Neurol., 291: 553-582. Fritschy J.-M. and Grzanna, R. (1990b) Distribution of locus coeruleus axons within the rat brainstem demonstrated by Phaseolus uulgaris leucoagglutinin anterograde tracing in combination with dopamine-P-hydroxylase. J. Comp. Neurol., 293: 616-631. Fritschy J.-M. and Grzanna, R. (1991) Experimentally induced neuron loss in the locus coeruleus of adult rats. Exp. Neurol., 111: 123-127. Fritschy, J.-M., Geffard, M. and Grzanna, R. (1990) The response of noradrenergic axons to systemically administered DSP-4 in the rat: An immunohistochemical study using antibodies to noradrenaline and dopamine-P-hydroxylase. J . Chem. Neuroanat., 3: 309-321. Geffard, M., Patel, S., Dulluc, J. and Rock, A.M. (1986) Specific detection of noradrenaline in the rat brain by using antibodies. Brain Rex, 363: 395-400. Grzanna, R. and Coyle, J.T. (1976) Rat adrenal dopamine-phydroxylase: Purification and immunologic characteristics. J. Neurochem., 27: 1091-1096. Grzanna, R., Berger U., Fritschy, J.-M. and Geffard, M. (1989) The acute action of DSP-4 on central norepinephrine axons: Biochemical and immunohistochemical evidence for differential effects. J. Histochem. Cytochem., 37: 1435-1442. Hallman, H. and Jonsson, G. (1984) Pharmacological modifications of the neurotoxic action of the noradrenaline neu-
268 rotoxin DSP4 on central noradrenaline neurons. Eur. J. Pharmacol., 103: 269-278. Hallman, H., Sundstrom, E. and Jonsson, G. (1984) Effects of the noradrenaline neurotoxin DSP4 on monoamine neurons and their transmitter turnover in rat CNS. J. Neural Transm., 60: 89-102. Hallman, H., Lange, J., Olson, L., Stromberg, I. and Jonsson, G. (1985) Neurochemical and histochemical characterization of neurotoxic effects of l-methyl-4-phenyl-l,2,3,6tetrahydropyridine on brain catecholamine neurones in the mouse. J. Neurochem., 44: 117-127. Hsu, S.M., Raine, L. and Fanger, H. (1981) Use of avidinbiotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem. Cytochem., 29: 577-580. Iverson, L.L. and Uretsky, N.J. (1970) Regional effects of 6-hydroxydopamine on catecholamine containing neurons in rat brain and spinal cord. Brain Rex, 24: 364-367. Jacobowitz, D. and Kostrzewa R. (1971) Selective action of 6-hydroxydopa on noradrenergic terminals: Mapping of preterminal axons of the brain. Life Sci., 10: 1329-1342. Jaim-Etcheverry, G. and Zieher, L.M. (1980) DSP-4: A novel compound with neurotoxic effects on noradrenergic neurons of adult and developing rats. Brain Res., 1988: 513523. Jonsson, G. (1983) Chemical lesioning techniques: monoamine neurotoxins. In A. Bjorklund and T. Kokfelt (Eds), Handbook of Chemical Neuroanatomy, Vol. 1. Methods in Chemical Neuroanatomy. Elsevier, Amsterdam, pp 463-507. Jonsson, G. and Sachs, C. (1973) Pharmacological modifications of the 6-hydroxy-dopa induced degeneration of central noradrenaline neurons. Biochem. Pharmacol., 22: 1709-1716. Jonsson, G. and Sachs, C. (1976) Regional changes in [3H]noradrenaline uptake, catecholamines and catecholamine synthetic and catabolic enzymes in rat brain following neonatal 6-hydroxydapamine treatment. Med. Biol., 54: 286-297. Jonsson, G., Hallman, H., Ponzio, F. and Ross, S. (1981) DSP4 (N-(chloroethyl)-N-ethyl-2-bromobenzylamine)-a useful denervation tool for central and peripheral noradrenaline neurons. Eur. J. Pharmacol.,72: 173-188. Jonsson, G., Ross, S.B. and Sundstrom, E. (1985) Uptake and accumulation of 3H-DSP4, a noradrenaline neurotoxin, in central and peripheral noradrenaline neurons. J. Neurochem., 44: S184. Kostrzewa, R.M. (1988) Reorganization of noradrenergic neuronal systems following neonatal chemical and surgical injury. Prog. Brain Res., 73: 405-423. Kostrzewa, R.M. and Jacobowitz, D.M. (1974) Pharmacological actions of 6-hydroxydopamine. Pharmacol. Rec., 26: 199-288. Landa, M.E., Rubio, M.C. and Jaim-Etcheverry, G. (1984) The neurotoxic compound N-(2-chloroethyl)-N-ethyl-Z-
bromobenzylamine hydrochloride (DSP4) depletes endogenous norepinephrine and enhances release of ['HI norepinephrine from rat cortical slices. J. Pharmacol. Exp Ther., 231: 131-136. Landa, M.E., Rubio, M.C. and Jaim-Etcheverry, G. (1988) N-2-chloroethyl-N-ethyl-2-brornobenzylamine (DSP4) increases the 3H-noradrenaline release evoked by nerve depolarization. Acta Physiol. Pharmacol. Latinoam., 38: 167-180. Lee, C.M., Javitch, J.A. and Snyder, S.H. (1982) Characterization of ['HI desipramine binding associated with neuronal noradrenaline uptake sites in rat brain membranes. J. Neurosci., 2: 1515-1525. Logue, M.P., Growdon, J.H., Coviella, I.L.G. and Wurtman, R.J. (1985) Differential effects of DSP-4 administration on regional brain norepinephrine turnover in rats. Life Sci., 37: 403-409. McBride, R.L., Ozment R.V. and Sutin J. (1985) Neonatal 6-hydroxydopamine destroys spinal cord noradrenergic axons from the locus coeruleus, but not those from lateral tegmental cell groups. J. Comp. Neurol., 235: 375-383. Pappas, B.A., Saari, M. and Peters, D.A.V. (1976) Regional brain catecholamine levels after intraventricular 6-hydroxydopamine in the neonatal rat. Res. Commun. Chem. Pathol. Pharmacol., 14: 751-754. Paxinos, G. and Watson, C. (1986) The Rat Brain in Stereotaxic Coordinates, 2nd edn., Academic Press, Orlando. Roffler-Tarlov, S., Pugatch, D. and Graybiel, A.M. (1990) Patterns of cell and fiber vulnerability in the mesostriatal system of the mutant mouse weaver. 11. High affinity uptake sites for dopamine. J. Neurosci., 10: 734-740. Ross, S.B. (1976) Long-term effects of N-2-chloroethyl-Nethyl-2-bromobenzylamine hydrochloride on noradrenergic neurones in the rat brain and heart. Br. J. Pharmacol,, 58: 521-527. Ross, S.B. (1987) Pharmacological and toxicological exploitation of amine transporters. TIPS, 8: 227-231. Ross, S.B. and Renyi, A.L. (1976) On the long-lasting inhibitory effect of N - ( 2 - c h l o r o e t h y l ) - N - e t h y l - 2 bromobenzylamine (DSP 4) on the active uptake of noradrenaline. J. Pharm. Pharmacol., 28: 458-459. Sachs, C. and Jonsson, G. (1972) Degeneration of central and peripheral noradrenaline neurons produced by 6-hydroxydopa. J. Neurochem., 19: 1561-1575. Sachs, C. and Jonsson, G. (1975) Mechanisms of action of 6-hydroxydopamine. Biochem. Pharmacol., 24: 1-8. Sachs, C., Jonsson, G., Heikkila, R. and Cohen, G. (1975) Control of the neurotoxicity of 6-hydroxydopamine by intraneuronal noradrenaline in rat iris. Acta Physiol. Scand., 93: 345-351. Zaczek, R., Fritschy, J.-M., Culp, S., De Souza, E.B. and Grzanna, R. (1990) Differential effects of DSP-4 on noradrenaline axons in cerebral cortex and hypothalamus may reflect heterogeneity of noradrenaline uptake sites. Brain Res., 522: 308-314.
SECTION I11
Noradrenergic Influences on Target Neurons
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Brain Research, Vol. 88
271
0 1991 Elsevier Science Publishers B.V.
CHAPTER 21
Autoradiography of adrenoceptors in rat and human brain: a-adrenoceptor and idazoxan binding sites C.R. Jones and J.M. Palacios
' Merrell Dow Research Institute, Rue d'Ankara, Strasbourg, France and
This chapter reviews the current classification of adrenoceptors, and notes the difficulties of combining the molecular biological and pharmacological classifications of adrenoceptors. Possibilities for mapping the distribution of the proposed subtypes of adrenoceptors using currently available ligands are discussed, and the autoradiographic visualisation of the az-, PI-. and P,-adrenoceptors in the broad subtypes of a]-, rat, monkey and human brain described and illustrated. The
Pre-Clinical Research, Sandoz Ltd., Basel, Switzerland
non-selectivity of ligands currently being used to label aadrenoceptors is shown; we compare the distribution of [ 3H]idazoxan binding sites with the distribution of a,-adrenoceptors visualised using other ligands. Resolution limitations of current autoradiographic approaches are considered and we shown how in siiu hybridisation can complement data from receptor labelling studies used to localise receptors to pre- or postsynaptic sites.
Key words: adrenoceptors, autoradiographic localisation, rat brain, human brain, idazoxan
Introduction Historically, adrenoceptors have been characterised and classified using pharmacological techniques supporting the existence of a fourway classification. Each division in Ahlquist's original 1948 classification ( a and p ) (Ahlquist, 1948) has subsequently been divided into two subtypes, a1 and a2 (Langer, 19741, PI and p2 (Lands et al., 1967) (Table 1). Recent advances in molecular biology have led to the cloning of novel adrenoceptors which form part of a large family of G proteincoupled receptors (Dixon et al., 1986; Frielle, 1987; KobiIka et al., 1987; Cotecchia et al., 1988; Regan et al., 1988; Guyer et al., 1990; Lomasney
et al., 1990; Schwinn et al., 1990) and have resulted in additional subdivisions in the classification of adrenoceptors (Table 1). Some difficulties remain in matching the proposed molecular biological classification with the pharmacological subtypes (Table 2); still, Ki values are now available for many drugs at both cloned and pharmacologically characterised receptors, and it is theoretically possible to perform mapping studies to investigate the localisation of most of these novel subtypes in the brain. Molecular biology has also produced novel adrenoceptor subtypes, some of which have yet to be characterised pharmacologically; others cannot be differentiated with currently available pharmacological tools. When
272 TABLE 1 Classification of adrenoceptors. The apparent correspondence between the molecular biological and pharmacological classifications is given underneath the pharmacological classification in parentheses, note the existence of more than one clone for the aZb receptor and the absence of receptors for some clones.
alpha
r-lbq'a "2
"1
1
-
PI
Pharmacological
P2
1
alpha
"L
"L
(C4 species)
(?)
"la
(?Rat)
(DDT1) (C10) (C4/?)
PI
(pi)
P2
P3
(p2)
(p3)
(CloneS)
I Molecular Biology
I
I
"1
- - m
DDT-1 Bovine Rat
1
"2
C2 C4 clone 5 C10
compared with traditional homogenate binding methods (Kuhar, 1986; Palacios, 19841, the autoradiographic study of adrenoceptors has the advantage of increased anatomical resolution and sensitivity. It has also provided the possibility of constructing atlases of receptors in given neuronal regions and circuits which should have value in determining the mechanisms of actions of drugs (Kuhar, 1978). Since it is possible to use human post mortem material, in vitro studies have been performed on pathological tissue to define the role of receptors in disease (Mackay et al., 1982; Palacios et al., 1990a,b; Waeber et al., 1991) and to compare the distribution of receptors with that of neurotransmitters. The general methodological aspects have been extensively reviewed (Kuhar et al., 1986) and topics specifically relevant to the localisation of adrenoceptors in the CNS recently reviewed (Jones et al., 1990).
beta
p 2 P3
Current classification of adrenoceptors a ,-Adrenoceptors Two a,-adrenoceptors were separated on the basis of their pharmacological characteristics (Morrow and Creese, 1986; Han et al., 1987). One receptor is resistant to inactivation by chlorethylclonidine (a,,) and has higher affinity for the antagonist WB4101 (Table 2). This pharmacological classification agrees well with recent data from molecular biology since more than one a,-adrenoceptor has been cloned and expressed. However, there are some important differences between the pharmacology of naturally expressed and cloned a,-adrenoceptors. These differences may relate to tissue-specific posttranslational modifications or to specific tissue expression of G-proteins that could alter agonist binding interactions. The first cloned a,-adrenoceptor used
273 TABLE 2 Comparison of Ki (nM) values for cloned and pharmacologically defined adrenoceptors a 2 receptors
Compound Corynanthine Phentolamine Prazosin Rauwolscine SKF 104,078
u2/c10 1188 6.2 2237 7.1 97
a2a
144 4.4 338 0.56 28.1
a2/C2 1002 9.2 293 11 105
a 2b
91.2 4.3 3.7 0.22 12.8
a2/C4
182 14.4 67.7 2.1 41
a2= 28.1 9.7 15.1 0.045 42.6
"2d
1548 3.3 457 18.6 309
a , receptors #lal a
WB4101 Prazosin Phentolamine Norepinephrine Chlorethyl clonidine
0.1 0.09 2.06 -
0
p receptors "'p2 human heart I-Propranolol d-Propranolol Practolol Procaterol IC1118551 CGP20712a CYP
1.o 131 -
0.67 27.54 0.04
bovine a, 0.55 0.27 4.8 16500 68% #4p2c~0 1.3 23 4.5 46 2.3 # 5 0.035
#*%
13.4 1.2 107 6 300 70%
Hamster a I 8.5 0.26 1.55 9 600 95%
# ~ P ~ C H O p3 adipocyte -
257 2300 490
values for rat brain (Graziadei et al., 1989); #'alb K , for rat liver (Han et al., 1987); # 3 K i (Buxton et al., 1987); (Ermorine et al., 1987); #' (Ermorine et al., 1989), # 5 K i(Bouvier et al., 1987). No satisfactory binding data are available on Po binding to adipocytes since at the concentrations of cyanopindolol necessary non-specific binding is high and the cells also express both P I and p2 receptors which obscure the P3 receptor, a labelled p3 antagonist is eagerly awaited. #lala K ,
#4
sequence data from the a,-adrenoceptor purified from hamster vas deferens to design oligonucleotide probes (Cotecchia et al., 1988). Interestingly the hamster vas deferens is pharmacologically a mixed population of alaand (Ylb subtypes. When expressed, the pharmacology of this receptor resembles that of an a l b subtype with a low affinity for WB4101, inactivation by chlorethyl clonidine and coupling to phosphatidyl inositol metabolism (Table 2). A cloned bovine a,-adrenoceptor, when expressed, has high affinity for WB4101 (Schwinn et al. 1990), suggesting an a l a subtype, but in contrast to the pharmacologically defined receptor, it is partially inactivated (68%) by chlorethlyclonidine. Furthermore, the tissue distribution of this receptor has been shown to be limited to the rabbit liver and human dentate gyrus. Since this distribution does not match that of the pharmacological receptor, it suggests a
further novel subtype or species variant which has not yet been identified pharmacologically. A new a,-adrenoceptor gene has recently been cloned from the rat, and although the pharmacology is at present unknown, Northern analysis revealed that the tissue distribution for this receptor at least matches the distribution of the ala subtype (vas deferens > hippocampus > cortex > heart > kidney). a ,-Adrenoceptors Four subtypes of a,-adrenoceptors have been characterised pharmacologically using tissues that have been proposed to express only one receptor subtype (Table 1). The ratio of selectivities for the two a,-adrenoceptor ligands yohimbine and prazosin distinguishes between the aZaand aZb subtypes (Table 2). a,,-Adrenoceptors are found in human platelets, the HT29 cell line, cerebral
274
cortex, and rat brain stem as an apparently homogeneous population of receptors. These receptors have low (338 nM) affinity for prazosin and high affinity for yohimbine (ratio 1:600) (Bylund et al., 1988). The a,,-adrenoceptor subtype is found in neonatal rat lung, and NG108-15 cells, and has high affinity for prazosin (4-5 nM) and yohimbine: prazosin ratio of 1:lO (Latipifour et al., 1982; Bylund et al., 1988). The a,,-adrenoceptor subtype was defined from the characteristics of ligands binding to opossum kidney cells (Murphy and Bylund 1988). Pharmacologically, this receptor is between the and a,,-adrenoceptor subtypes, with an intermediate affinity for prazosin (15 nM). The examination of the affinities of a range of compounds supports the existence of this subtype as a distinct category (Bylund et al., 1990). Finally, an additional subtype aZdhas been characterised in bovine pineal glands using the same methodology (Bylund et al., 1990). Although the pharmacology resembles that of an u,,-adrenoceptor (Table 2), correlation analysis supported a separate category for this receptor (Bylund et al., 1990). The a,-adrenoceptor occurs at both pre- and post-synaptic sites. Based on the differential interactions of SK & F 104078 (Hiebele et al., 1988) it has been proposed that the pre- and postsynaptic receptors are distinct. However in binding studies with SK&F 104078, the maximum difference in relative affinities between the respective subtypes is only 25-fold (Table 21, whereas in uitro the compound displays 100-fold selectivity for the post- versus the presynaptic receptors; thus it is unclear which category the presynaptic receptor belongs to in the (Y2a,b,c,d ciassification.
Molecular biology of a,-adrenoceptors The pharmacological subdivision of a,-adrenoceptors is supported by molecular biology. The probing of human genomic DNA with a fragment from the cloned platelet adrenoceptor revealed three bands on Southern blots (Kobilka et al., 1987). As predicted, three distinct qadrenoceptors have now been cloned and located in differ-
ent human chromosomes (Kobilka et al., 1987; Regan et al., 1988; Lomasney et al., 1990); recently an additional clone was added to the series (Weinshank et al., 1990). The pharmacology of the cloned receptors expressed in COS-7 cells and of the naturally occurring adrenoceptors is summarised in Table 2. The aZapharmacology corresponds with that of the a,C10 receptor and RNA from HT29 cells and rat cortex (sources of a,,-adrenoceptors) hybridize with probes from the a,C10 receptor to give signals of appropriate size (3.8 kb) (Lorenz et al., 1990). However, there are some inconsistencies between the two classifications. For example, the affinity of prazosin is 4-fold lower in the Cos-7 expression system than that reported from pharmacological studies (Table 2). It is also difficult to find the pharmacological counterpart of the a,C2 receptor, for although it has the pharmacology of an a,,-adrenoceptor with a high affinity for prazosin (Table 1) and lacks glycosylation sites like the a 2 p 2 receptor (Lanier et al., 1988). Northern analysis with a,C2 probes detected no signal in neonatal rat lung (a,,-adrenoceptor prototype), HT29 cells (aZaprototype) or opossum kidney cells (a,, prototype) (Lorenz et al., 1990). Thus, at the present time, the a,C2 clone probably represents a receptor not encompassed by the current pharmacological classification (Table 1). Additionally, RNA from the neonatal rat lung (a,, prototype) does not hybridise with the a,C4 probe although Lorenz et al. (1990) propose that this represents an a,,, receptor. However, other sources of a,,adrenoceptors (rat brain, and neuroblastoma NG 108-15) cells do hybridize with the a,C4 probe, and probes of the rat homologue of the a,C4 receptor hybridise with RNA from neonatal rat lung (Zeng et al., 1990). The pharmacologically defined a,,-adrenoceptor seems to be heterogenous since the recently isolated human clone 5 also has pharmacology (Weinshank et al., 1990). The a,C4 probes also hybridize with RNA from opossum cells ( a , c ) (Lorenz et al., 1990). Despite the apparent pharmacological separation of a!,,' and a,,-adrenoceptors, Lorenz et al.
215
(1990) propose, from the results of hybridisation experiments, that the opossum a,,-adrenoceptor and the CNS a,,-adrenoceptors are equivalent and species variants of the a,C4 receptor (Table 1). The expression of the a,C2 receptor in rat kidney and liver raises an additional controversy concerning the classification of adrenoceptors. Recently, Zonnenschein et al. (1990) identified in the rat liver idazoxan binding sites that arc not displaced by endogenous agonists (noradrenaline or adrenaline) although the site binds guanidino structures (guanabenz and guanoxan) and imidazoline structures (naphazoline and UK14304). Since there are no rauwolscine binding sites in rat liver (Zonnenschein et al., 19901, it is possible that the a,C2 probe is cross reacting with the imidazoline receptor mRNA, although the high affinity of the cloned receptors for rauwolscine (Table 2), and the lack of atypical idazoxan binding sites in COS-7 cells transfected with a , receptors is evidence against one of the a,-adrenoceptor clones being the idazoxan receptor itself. Furthermore, solubilisation studies support the existence of separate receptor proteins binding idazoxan that are distinct from a , receptors (Wikberg and Uhlen, 1990; Parini et al., 1989). As a ligand [3H]idazoxan frequently labels more sites than do other a,-adrenoceptor ligands, in both peripheral and central tissues of several species (Boyajian et al., 1987, Yakubu et al., 1988, Michel ef al., 1988, Brown et al., 1990). Several explanations have been proposed to account for this: First, that there are two adrenoceptors which have equal affinity for idazoxan but one of which, (labelled Ri) by Boyajian et al. (1987) is insensitive to rauwolscine. However, the naturally occurring a,-adrenoceptors have similar affinity for both ligands (Table 2); Second, the greater number of idazoxan binding sites may be a consequence of buffer composition (Weinshank et al., 19901, perhaps relating to an allosteric modulation of idazoxan but not rauwolscine binding (Guyer et al., 1990); Third, idazoxan may be
labelling a non-adrenergic site in addition to the a2 receptor. P-Adrenoceptors The pharmacology of P-adrenoceptors is less controversial than that of a-adrenoceptors. In most tissues the existence of two populations of receptors (PI and P,) explains the pharmacology. However, in certain peripheral tissues (ileum, colon, smooth muscle and notably the adipocyte) the pharmacology has been atypical with low pA2 values for both P , and P2 selective antagonists (Emorine et al., 1989). The recent cloning of a novel P-adrenoceptor ( P 3 )explains somewhat this atypical pharmacology (Zaagsma and Nahorski, 1990), although the expressed P3 receptor has low affinity for both ICI 118,551 and CGP 20712A suggesting that it is of neither the P,- nor P2adrenoceptor subtypes. There are some inconsistencies; for example the ligand [ 125Ilcyanopindolo1 does not label the rat adipocyte receptor, yet it has high affinity for the expressed receptor in CHO cells ( K , 500 pM). To date, the expression of this receptor, as determined by Northern analysis, is limited to peripheral tissues. It remains to be seen whether the greater sensitivity of the in situ hybridisation technique will allow detection of this receptor in the central nervous system. Autoradiographic strategy
Based on the pharmacology of natural and cloned receptors, it is possible to devise strategies to label adrenoceptor subpopulations using autoradiographic techniques. This approach is feasible using drugs to mask receptor subtypes when using non-selective ligands. Thus it is possible to visualise subtypes even if no selective radiolabelled compounds are available for a given receptor. This technique has been well-established for mapping the distributions of PI- and p,-adrenoceptors since there are highly selective antagonists for the two subtypes (Buxton et al., 1987; Waeber et al., 1991). More recently, we applied the mask-
276 TABLE 3 Conditions for labelling adrenoceptor subtypes in autoradiographic studies Receptor a1.d
alh
a2ahcd
*
a2ad aZd
*
a 2a "2ahd
PI P2
P' *
*I
Ligand
Additions
Non-specific
[ 12511BE2254 [I2'1]HEAT [ 'Hlyohimbine ['Hjyohimbine ['Hlyohimbine [ 'H15methylurapidil [ 'Hlyohimbine [ '2sl]cyanopindolol [ I]cyanopindolol [ 12sI]cyanopindolol
chlorethyl clonidine (100 pM) WB 4101 (30 nM)
prazosin 0.1 p M prazosin 0.1 y M phentolamine 10 p M phentolamine 10 p M phentolamine 10 p M phentolamine 10 p M phentolamine 10 p M propranolol(1 y M ) propranolol(1 y M ) pindolol/noradrenaline (1 p M )
*
2
~
~
~
*25HT,prazosin *25HT,corynanthine *25HT *25HT,rauwolscine **5HT, CGP 20712 *3a/or ICI 89,406 * 4 *25HT;ICI 118,551 * 4 *'5HT; ICI 118,551 *4, CGP20712A *3
* I The lower affinity of ['2'I]cyanopindolol for the P3 receptor requires higher concentrations of the ligand similar to those used to label 5HTlb receptors (Pazos et ul. 1985a) the reported higher affinity of pindolol may make [1251]pindolola more suitable ligand; ** lo-' M; * 3 100 nM; * 4 70 nM (Buxton et ul. 1987). * * This table refers to possible protocols based on the known selectivity of the ligands (Table 2).
ing technique to a,-adrenoceptors. We used [12sl]HEATas a non-selective ligand for the a,adrenoceptor subtypes and either appropriate concentrations of the irreversible antagonist chlorethyl clonidine to remove a ,,-adrenoceptor binding or WB4101 to remove a ,,-adrenoceptor binding. Evidence was provided for a,-adrenoceptor heterogeneity in the human hippocampus with the different a , subtypes localised to different hippocampal regions (Hoyer et al., 1990; Gross et a/., 1989), even though the selectivity of these agents is only 50-fold for the subtypes (affinities of WB4101 and chlorethyl clonidine are given in Table 2). This approach could be applied to the other adrenoceptors. For a,-adrenoceptors the total population of aZa-,a2,,-, aZc-,and aZdadrenoceptor subtypes would be labelled by [ 'H]yohimbine, and various combinations of drugs (Table 2) could be used, together with image subtraction techniques, to visualise the distribution of the respective subtypes (Table 3). Similarly P,-adrenoceptors could be visualised using [12511pindolol( K , p3 11 pM) together with ICI
118, 551 and CGP 20712a to block p,- and P2adrenoceptors, although a selective Ps ligand is eagerly awaited. Distribution of adrenoceptor populations a ,-Adrenoceptors The distribution of a ,-adrenoceptors has been studied autoradiographically using the non-selective ligands ['251]HEAT and ['Hlprazosin. In rat brain (Fig. l), the decreasing rank order of densities (in fmol/mg protein) was the external plexiform layer of the olfactory bulb (2301, thalamus (1401, cortex (lamina V showing a double band), dentate gyrus, and septohippocampal area (50). The basal ganglia and cerebellum had low densities of specific a,-adrenoceptor binding. In the brainstem high densities were found in the facial nerve nucleus, solitary tract and periaqueductal grey matter (120-140 fmol/mg protein) (Jones et al., 1985a,b; Minneman et al., 1981). However, there are marked species differences in the distribution of a,-adrenoceptors between primates and
Fig. 1. Distribution of a-adrenoceptors ( a , panels A - F a , G-H) in sagittal sections of rat brain with different radioligands. Left hand panels (A,C,E,G) represent total binding and right hand panels represeilt nonspecific binding (10 y M phentolamine); [3H]rauwolscine (A,B), [ 'HJidazoxan (C,D), [ 1251]iodoclonidine(E,F), [ 'Z51]HEAT (G-H); Cx, cortex; Hip, hippocampus; Cpu, caudate putamen; Th, thalamus; Cb, cerebellum.
211
218
B
Fig. 2. Sections of human spinal cord labelled with ['251]HEAT. showing a,-adrenoceptor distribution. Panels A,C, total binding; panels B,D, nonspecific binding in the presence of 10 pM phentolamine.
humans compared with other species (Palacios et al., 1987). In man the highest densities were found in the hippocampus, particularly the dentate gyrus and CA3 region. In the cortex the laminar distribution is aIso altered with the highest levels localised to the most external laminae
layers 1-11 and layer VI. The a,-adrenoceptor density in the human geniculate body was low compared to other species. In human brains the supraoptic and paraventricular hypothalamic nuclei had high levels of receptors, whereas intermediate densities were found in the amygdala
Fig. 3. Sections of the human spinal cord, showing a,-adrenoceptor distribution (panels A-D), and monkey spinal cord panels (E-H), labelled with [3H]idazoxan (A,B,E,F), and [ 'H]rauwolscine (C,D,G,H). Total binding panels, A,C,E,G; nonspecific binding panels, B,D,F,H in the presence of 10 p M phentolamine.
279
280
and stria terminalis. In contrast to the distribution of a,-adrenoceptors in the brainstem, low levels of a,-adrenoceptors were found in the locus coeruleus. In the spinal cord the highest densities were in the central gray matter with no laminar pattern in the dorsal horn. a,-adrenoceptors were seen overlying the motor neurons, consistent with the influence of catecholamines on the noradrenergic coerulospinal system (Pompeiano, 1988) (Fig. 2), but higher resolution studies are needed to document the precise cellular localisation of these receptors. a,-Adrenoceptors a,-Adrenoceptors are also widely distributed, but little information on the distribution of the subtypes is available since autoradiographic studies to date have used approaches which label all the subtypes. Although radioligand binding studies in the rat support the existence of equal proportions of aZaand a,,-adrenoceptors in the cerebral cortex, hippocampus and corpus striatum (Bylund, 1985), autoradiographic studies with [ 3H]clonidine and para-[ 3H]aminoclonidine (Unnerstall et aL, 1984; Young and Kuhar, 1980) visualised high densities of specific binding in thalamic nuclei, the hypothalamus, hippocampus, amygdala and brain stem nuclei, such as the locus coeruleus and nucleus tractus solitarius (Palacios et al., 1981). In the locus coeruleus, functional studies suggest that the a,-adrenoceptors are located presynaptically (Cederbaum and Aghajanian, 1977; Svensson and Usdin, 1978). In man, high densities of a,-adrenoceptors were found in the visual cortex (270 fmol/mg protein), claustrum (95 fmol/mg protein) and cerebellar cortex.
The levels and distribution in human and monkey brain stem nuclei were similar to those found in the rat: nucleus tractus solitarius (133 fmol/mg protein) and locus coeruleus (166 fmol/mg protein) (Probst et al., 1985) (Fig. 3). In the spinal cord, there is a marked laminar pattern of distribution with high concentrations overlying the substantia gelatinosa (Fig. 3). In several brain regions a similar distribution of a,-adrenoceptors and opiate receptors (reviewed by Pazos, 1988) supports data showing functional interactions between the two systems (van Giersbergen et al., 1989). The overlapping distributions of a,-adrenoceptors and opiate receptors is prominent in both the spinal cord and the neuronal circuit involved in the self-administration of opiates (the hippocam pal-accumbens-amygdala-entorhinal
cortex circuit). In the rat, the distribution of [ 'Hlidazoxan binding sites (Boyajian et al., 1987) is similar to that reported with [ 3H]clonidine (Young and Kuhar, 1981) and p~ra-[~H]aminoclonidine (Unnerstall et al., 1984); however, the ligand [ 'H]rauwolscine labelled fewer sites in most brain regions (Boyajian et al., 1987) and the distribution of the sites resembled that seen with the dopamine ligand [ 3H]spiroperidol (the ligand was not labelling dopamine sites from analysis of competition curves with dopamine and other catecholamines). To examine whether the discrepancy between the distribution of sites labelled with a,-adrenoceptor ligands existed in primates and man, we investigated the labelling patterns of the three ligands, [ 3H]idazoxan, [ 3H]rauwolscine and [ 3H]yohimbine using serial sections of rat, human, and monkey brain (Table 4). To exclude the
TABLE 4 Radioligand binding conditions for tissue sections comparing the distributions of three a 2 ligands (incubation times were taken from Stephenson and Summers (1985)) Ligand
Conc.
Preincubation
Incubation
Washout
[3H]Rauwolscine [3H]Idazoxan [3H]Yohimbine
0.6 nM 0.8 nM 1.0 nM
30 min GppNHPlOO p M 30 min GppNHPlOO p M 30 min GppNHPlOO p M
Krebs phosphate pH7.3 at 25°C 2 h Krebs phosphate pH7.3 at 25°C 2 h Krebs phosphate pH7.3 at 25°C 2 h
2 X 5 min 50 mM Tris pH7.4 at 0°C 2 X 5 min 50 mM Tris pH7.4 at 0°C 2 X 5 min 50 mM Tris pH7.4 at 0°C
28 1
contribution from SHTla receptors (Convents et al., 1989) all incubations were carried out in the presence of 1 y M 5HT. In the sagittal rat brain
sections a similar distribution was seen to that found by Boyajian and Leslie (1987), with a greater number of idazoxan compared to rau-
Fig. 4. Sections at the level of the basal ganglia from the monkey showing a,-adrenoceptors distribution. The coronal sections were from the same animal and labelled with ['H]idazoxan (A$), ['Hlyohimbine (C,D), and ['H]rauwolscine (E,F). The pattern of labelling differed between the ligands; panels A,C,E represent total binding and panels B,D,F nonspecific binding in the presence of 10 p M phentolamine. Cd, caudate; C, claustrum; Cx, cortex; Put, putamen; Th, thalamus; Gpi, Gpm, globus pallidus; Hip, hippocampus.
282
283
.
..
,.
Fig. 6. Sections of human hippocampus showing az-adrenoceptor distribution, serial sections were labelled with [3H]idazoxan (A,B), [ 3H]rauwolscine (C,D) and [ 3Hlyohimbine(E,F). Panels A,C,E, total binding; panels B,D,F represent nonspecific binding in the presence of 10 pM phentolamine. The pattern of labelling is similar with all three ligands, DG dentate gyrus; Hip hippocampus; CA1 and CA3 refer to hippocampal regions; the arrow overlays the stratum molecular in panel C.
Fig. 5. Section of the human caudate at two levels with serial sections labelled with [ 3H]idazoxan(A,B,C) [3H]rauwolscine(D,E,F) and [3Hlyohimbine(G,H,I). Panels A,C,D,F,G,I, total binding; and panels B,E,H, nonspecific binding in the presence of 10 F M phentolamine. ACC,accumbens; C, claustrum; Cd, caudate; Cx,cortex; Put, putamen.
284
Fig. 7. Sections at the level of the nucleus tractus solitarius (NTS) in the monkey brainstem, showing a,-adrenoceptors distribution. Note the high level of specific binding in the granular layer of the cerebellar cortex. The coronal sections were from the same animal and labelled with [3H]idazoxan (A,B), and [3H]yohirnbine (C,D). The patterns of labelling were similar; panels A and C represent total binding, panels B and D represent nonspecific binding in the presence of 10 p M phentolamine.
wolscine binding sites (especially in the olfactory areas; Fig. 1; panels A and C). In the monkey brain this difference was seen in some areas such as the cortex, but in the basal ganglia of monkey and man, human brain stem and hippocampus there were more rauwolscine than idazoxan sites (Figs. 4, 5 and 6). In the monkey cortex the density of idazoxan sites was 70 vs 50 fmol/mg protein for rauwolscine. In the monkey basal ganglia the higher densities of rauwolscine than idazoxan binding sites were seen in the putamen: density of [ 3H]rauwolscine sites 127 fmol/mg
protein compared to 47.8 fmol/mg protein for [3H]idazoxan (Fig. 4; panel A). In the human caudate (Fig. 5; panel A) the level of nonspecific binding defined by phentolamine was high, but there were more sites labelled with [ 3Hlrauwolscine than with [3H]idazoxan (47 vs 18 fmol/mg protein). In other brain areas the density of sites for the two ligands was similar. In the monkey brain stem, at the level of the inferior olive, both ligands had high concentrations of binding sites overlying the nucleus tractus solitarius (Fig. 7). These findings parallel the distribution of specific
Fig. 8. Sections at the level of the facial nerve nucleus (NVII) labelled with [3H]idazoxan (A,B), [3H]rauwolscine (C,D) and [3H]yohimbine (E,F). The labelling pattern is similar but lower levels of binding are seen in the section labelled with [3H]idazoxan. Total binding panels A,C,E; nonspecific binding in the presence of 10 p M phentolamine panels B,D,F.
286
para-aminoclonidine binding described previously (Probst et al., 1985). In contrast to the monkey (Fig. 4), the density of [3H]rauwolscine sites in
the stratum molecular of the human hippocampus was much higher than the density of [3H]idazoxan sites (Fig. 6). In the human brain
Fig. 9. Distribution of P-adrenoceptors in sagittal sections of rat brain treated with saline (A,B), imipramine (C,D) or 80H-DPAT and imipramine together (E,F). Sections were labelled with [ '251]cyanopindolol.The left hand panels show total binding and the right hand panels nonspecific binding in the presence of 1 p M propranolol. There is an additive down regulation induced by 80H-DPAT and imipramine of cortical P , receptors, (Cx,cortex; Cp, caudate putamen; Th, thalamus).
281
stem at the level of the facial nucleus there were again more sites labelled with [ 3H]rauwolscine (60 vs < 10 fmol/mg protein; Fig. 81, yet in the cervical spinal cord the localisation was similar to that seen with para-aminoclonidine (Unnerstall et al., 1984, Probst et al., 1983, and similar densities of sites were seen with both ligands (Fig. 3). The binding sites were localised over the substantia gelatinosa and central grey matter (Fig. 3). The distribution of a,-adrenoceptors in the spinal cord contrasts with the more diffuse distribution of a ,-adrenoceptors which are localised over the motor neurons and intermediolateral columns (Fig. 2) and is similar to the localisation of opiate receptors in laminae 1/11 (Dashwood et al., 1985). P-Adrenoceptors The distribution of P-adrenoceptors in the rat brain is illustrated in Figures 9 and 10. The comparative distribution between species has recently been reviewed (Jones et aZ., 1990). High concentrations of receptors are found in the neocortex (Layers 1-110, nucleus accumbens, olfactory tubercle, globus pallidus, dorsal subiculum, substantia nigra (pars reticularis), molecular layer of the cerebellum and cingulate cortex (Fig. 9). In view of the reduced exposure time of iodinated radioligands many publications have used ['251]cyanopindolol (CUP) as a ligand for studies on P-adrenoceptors. However, this ligand will also label 5HTlb receptors (Pazos et al., 1985a). Figure 9 shows the distribution of 5HTlb receptors labelled with [1251]CYPin the presence of propranolol 1 p M (panels B, D, F), and localises high concentrations of 5HTlb binding sites in the subiculum and substantia nigra. Because of the availability of compounds with high selectivity, early studies on the distribution of P-adrenoceptors were able to map the distribution of the subtypes (Palacios and Kuhar, 1982; Rainbow et al., 1984; Pazos et al., 1985b). In the rat 80-90% of the binding in the cerebral cortex, medial nuclei of the thalamus, and all regions of the hippocampus is to PI-adrenoceptors. In the cere-
Fig. 10. Coronal sections of rat cerebellum labelled by coincubating ['251]cyanopindolol (50 pM) in the presence of either 100 nM CGP20712a (A) to visualise p2 receptor distribution or 1 pM propranolol (nonspecific binding) (B) and p2 receptor mRNA visualised by in situ hybridisation using cDNA labelled oligonucleotide probe for p2 receptor mRNA (NEN Du Pont) (Ventimiglia et al., 1987). (D) There is a mismatch between the localisation of the receptors to the molecular (mol) layer and the RNA to the granular (gr) layer of the cerebellum.
bellum (Fig. lo), 80-90% of the binding is to P,-adrenoceptors. In layer VI of the cerebral cortex, the olfactory tubercle and superficial layer
288
of the superior colliculus, 60-70% of the receptors are of the p,-adrenoceptor subtype. Why the P,-adrenoceptor is the main receptor expressed in some regions is unclear since there is generally a poor correlation between the levels of p-adrenoceptors and the extent of noradrenergic innervation in different brain regions (Pahcios and Kuhar, 1982). The distribution of these receptors, which have higher affinity for adrenaline as a transmitter, far exceeds the possible diffusion area of adrenaline released from adrenergic neurons (Pieribone et al., 1988). Recent studies have localised p,-adrenoceptors to barrel field arrays in the somatosensory cortex. These neurons, in single-unit recordings, respond to stimulation of the rats vibrissae and provide some insight into the role of p-adrenoceptors in normal neuronal processing. The removal of whiskers disturbed the development of the pattern of padrenoceptors in the somatosensory cortex, yet administration of propranolol did not block the stimulation of [deoxy-1 4 C l g l u ~ utilisation ~~e by whisker stimulation. Additional functions for these p,-adrenoceptors may be related to roles in neuronal development, since P-adrenoceptors increase the activity of ornithine decarboxylase (Morris et al., 1983) the rate limiting step in spermidine synthesis, and their glial localisation may indicate other functions (Trimmer and McCarthy, 1986). Further higher resolution studies are necessary to resolve the precise cellular location of these receptors to help identify their functional roles. Future studies using in situ hybridisation techniques may also help to define the precise cellular location of these receptors. For example, we used the in situ hybridisation method to study the localisation of p,-adrenoceptor mRNA in the rat cerebellum using a cDNA probe for the p,-adrenoceptor mRNA (Ventimiglia et al., 1987). The probe localised mRNA production to the granule cell layer of the cerebellum (Fig. 10; panel C). Therefore in the cerebellum there is a mismatch between the localisation of p,-adrenoceptors to the molecular layer (Fig. 10; panel A)
and the localisation of the mRNA for the receptors to the granule cell layer. This mismatch indicates that the receptors are synthesised in the granular cell layer and, therefore, are presynaptically located on the dendritic fields of the Purkinje cells. Conclusions The future autoradiographic study of adrenoceptors together with in situ hybridisation offers the possibility of distinguishing between a pre- and postsynaptic location of adrenoceptors, and allows us to study the regulation of these receptors at the gene level. Future studies will be facilitated by the development of ligands selective for each of the subtypes and the cellular resolution increased through the use of specific antibody staining techniques. Molecular biology has identified receptor heterogeneity where there are insufficient tools to identify this heterogeneity pharmacologically. At the present time, the mapping of the central nervous system distribution of adrenoceptor subtypes, and their localisation to specific cell types, awaits detailed analysis. The extra “alpha adrenergic binding sites” in the rat seen with the ligand [ 3H]idazoxan (compared to [3H]rauwolscine) are not found in most areas of the human brain. This finding exemplifies the marked species differences apparent in the distribution of adrenoceptors, which both makes it difficult to extrapolate findings from animals to man and complicates the interpretation of behavioural studies in rodents. Species differences in receptor distribution are further complicated by species variations in the pharmacology of adrenoceptors. With the eventual cloning of the full family of adrenoceptor genes it will be possible to construct an accurate map of the distribution of the adrenoceptor subtypes and to fully characterise their pharmacology in animal and human brains. These future studies should provide new insights into the mechanisms of drug action and help to identify novel drug targets.
289
References Ahlquist, R.P. (1948) A study of the adrenotropic receptors. Am. J. Physiol., 153: 586-600. Bouvier, M., Hnaatowich, M., Collins, S., Kobilka, B.K., De Blasi, A., Lefkowitz, R.J. and Caron, M.G. (1987) Expression of a human cDNA encoding the P,-adrenergic receptor in Chinese hamster fibroblasts (CHW): Functionality and regulation of the expressed receptors. Mol. PharmaCOL, 33: 133-139. Boyajian, C.L. and Leslie, F.M. (1987) Pharmacological evidence for a2-adrenoceptor heterogeneity: Differential binding properties of [3H]rauwolscine and ['Hlidazoxan in rat brain. J. Pharmacol. Exp. Ther., 241: 1092-1098. Boyajian, C.L., Loughlin, S.E. and Leslie, F.M. (1987) Anatomical evidence for 01-2 adrenoceptor heterogeneity: Differential autoradiographic distributions of [3H]rauwolscine and ['Hlidazoxan in rat brain. J. Pharmacol. Exp. Ther., 241: 1079-1091. Brown, C.M., MacKinnon, A.C., McGrath, J.C. and Kilpatrick, A.T. (1990) a,-adrenoceptor subtypes and imidazoline like binding sites in rat brain. BY. J. Pharmacol., 99: 803-809. Buxton, B.F., Jones, C.R., Molenaar, P. and Summers, R.J. (1987) Characterisation and autoradiographic localisation of P-adrenoceptor subtypes in human cardiac tissues. Br. J. Pharmacol., 92: 299-310. Bylund, D.B. (1985) Heterogeneity of a - 2 adrenergic receptors. Pharmacol. Biochem. Behau., 22: 835-843. Bylund, D.B., Prenger, C.R and Murphy, T.J. (1988) aZaand adrenergic receptor subtypes: Antagonist binding in tissues in cell lines containing only one subtype. J. Pharmacol. Exp. Ther., 245: 600-607. Bylund, D.B, Blaxall, H.S., Murphy, T.J. and Simmoneaux, V. (1990) Pharmacological evidence for and a Z dadrenergic receptor subtypes. In Szabadi (Ed), The Pharmacology of Adrenoceptors, Birkhausen Verlag, London, in press. Cedarbaum, J.M. and Aghajanian, G.J. (1977) Catecholamine receptors on locus coeruleus: Pharmacological characterisation. Eur. J. Pharmacol., 44: 375-385. Convents, A,, Convents, D., De Backer, J.P., De Keyser, J. and Vauquelin, G. (1989) High affinity of ['H]rauwolscine and ['H]RX781094 to a p adrenergic receptors and non stereoselective sites in human and rabbit brain cortex membranes. Biochem. Pharmacol., 38: 455-463. Cotecchia, S., Schwinn, D.A., Randall, R.R., Lefkowitz, R.J., Caron, M.G. and Kobilka, B.K. (1988) Molecular cloning and expression of the cDNA for hamster a,-adrenergic receptor. Proc. Natl. Acad. Sci., USA, 85: 7159-7163. Dashwood, M.R., Fleetwood-Walker, S.M., Gilbey, M.P. Mitchell, R. and Spyer, K.M. (19x5) Distribution of a adrenergic, muscarinic and opiate receptors within the lumbar spinal cord of the cat. J. Physiol., 362: 42P. Dixon, R.A.F., Kobilka, B.K., Strader, D.J., Benovic, J.L., Dohlman, H.G., Frielle, T., Bolanowski, M.A., Bennet, C.D., Rands, E., Diehl, R.E., Mumford, R.A., Slater, E.E.,
Sigal, I.S., Caron, M.G., Lefkowitz, R.J. and Strader, C.D. (1986) Cloning of the gene and cDNA for mammalian P-adrenergic receptor and homology with rhodopsin. Nature, 321: 75-79. Emorine, L.J., Marullo, S., Delavier-Klutchko, C., Kaveri, S.V., Durieu-Trautman, 0. and Strosberg, A.D. (1987) Structure of the gene for human beta 2-adrenergic receptor: Expression and promoter characterization. Proc. Natl. Acad. Sci., USA, 84: 6995-6999. Emorine, L.J., Marullo, S.G, Briend-Sutren, M.M., Patey, G., Tate, K., Klutchko, D.C. and Strosberg, A.D. (1989) Molecular characterisation of the human P,-adrenergic receptor. Science, 245: 1118-1121. Frielle, T., Collins, S., Kiefer, W.D., Caron, M.G., Lefkowitz, R.J. and Kobilka, B.K. (1987) Cloning of the cDNA for the human P1 adrenergic receptor. Proc. Natl. Acad. Sci., USA, 84: 7920-7924. Graziadei, I., Zernig, G., Boer, R. and Glossman, H. (1989) Stereoselective binding of niguldipine enantiomers to a l a adrenoceptors labelled with [ 3H]5-methyl urapidil. Eur. J. Pharmacol., 172: 329-337. Gross, G., Hanft, G. and Mehdorn, H.M. (1989) Demonstration of a , a- and a I P-adrenoceptor binding sites in human brain tissue. Eur. J. Pharmacol., 169: 325-328. Guyer, C.A., Horstman, D.A., Wilson, A.L., Clark, J.D., Cragoe, E.J. and Limbird, L.E. (1990) Cloning sequencing and expression of the gene encoding the porcine a 2 adrenergic, receptor: Allosteric modulation by Na + , H + , and amiloride analogs. J. Biol. Chem., 265: 17307-17317. Han, C., Abel, P.W. and Minneman, K.P. (1987) Heterogeneity of a,-adrenergic receptors revealed by chlorethylclonidine. J. Pharmacol. Exp. Ther., 32: 505-510. Hiebele, J.P., Sulpizo, A.C., Nichols, A.J., Willette, R.N. and Ruffolo, R.R., Jr. (1988) Pharmacologic characterisation of SK&F 104078, a novel IY-2adrenoceptor antagonist which discriminates between pre- and postjunctional a-2 adrenoceptors. J. Pharmacol. Exp. Ther., 247: 645-652. Hoyer, D., Jones, C.R., Ford, W. and Palacios, J.M.P. (1990) Subtypes of a,-adrenoceptors in hippocampus of pigs, guinea-pigs, calves and humans: Regional differences. Eur. J. Pharmacol., 188: 9-16. Jones, C.R., Hoyer, D. and Palacios, J.M.P. (1990) Adrenoceptor autoradiography. In Heal D.J. and Marsden C.A. (Eds.), The Pharmacology of Noradrenaline in the Neruous Sysrem, Oxford Medical, Oxford, pp. 41-75. Jones, S.L., Miller, G., Gauger, L.L. and Davis, J.N. (1985a) Regional distribution of rat brain a,-adrenergic receptors: Correlation between [ '251]HEAT membrane binding and in citro autoradiography. Life Sci., 36: 45-51. Jones, S.L., Gauger, L.L. and Davis, J.N. (1985b) Anatomy of brain a,-adrenergic receptors: In oitro autoradiography with [1251]HEAT.J. Comp. Neurol., 231: 190-208. Kobilka, B.K., Matsui, H., Kobilka, T.S., Yang-Feng, T.L., Francke, U., Caron, M.G., Lefkowitz, R.J. and Regan, J.W. (1987) Cloning, sequencing and expression of the gene coding for the human platelet a2-adrenergic receptor. Science, 238: 660-656.
290 Kuhar, M.J. (1978) Histochemical localization of opiate receptors and opioid peptides. Fed. Proc., 37: 153-157. Kuhar, M.J. (1986) Quantitative receptor autoradiography: An overview. In C.A. Boast, E.W. Snowhill and C.A. Altar (Eds.), Quantitative Receptor Autoradwgraphy , Alan R. Liss, New York, pp. 1-12. Kuhar, M.J., De Souza, E.B. and Unnerstall, J.R. (1986) Neurotransmitter receptor mapping by autoradiography and other methods. Ann. Reu. Neurosci., 9: 27-60. Lands, A.M., Arnold, A,, McAuliff, J.P., Luduena, F.P. and Brown, T.G., (1967) Differentiation of receptor systems activated by sympathomimetic amines. Nature, 215: 597598. Langer, S.Z. (1974) Presynaptic regulation of catecholamine release. Biochem. Pharmacol., 23: 1793-1800. Lanier, S.M., Homcy, C.J., Patenaude, C. and Graham, R.M. (1988) Identification of structurally distinct qadrenergic receptors. J. Biol. Chem., 264: 14491-14496. Latipifour, J., Jones, S.B. and Bylund, D.B. (1982) Characterisation of [3H]Yohimbine binding to putative a-2 adrenergic receptors in neonatal rat lung. J. Pharmacol. Exp. Ther., 223: 606-611. mnasney, J.W., Lorenz, W., Allen, L.F., King, K., Regan, J.W., Yang-Feng, T.L., Caron, M.G. and Lefkowitz, R.J. (1990) Expansion of the a 2 adrenergic receptor family: Cloning and characterisation of a human a2 adrenergic receptor subtype the gene for which is located on chromosome 2. Proc. Natl. Acad. Sci., USA, 87: 5094-5098. m e w , W., Lomasney, J.W., Collins, S., Regan, J.W., Caron, M.G. and Lefkowitz, R.J. (1990) Expression of three alpha 2-adrenergic receptor subtypes in rat tissues: implications for alpha 2 receptor classification. Mol. Pharmacol., 38: 599-603. Mackay, A.V., Iversen, L.L., Rossor, M., Spokes, E., Bird, E., Arregui, A., Creese, I. and Snyder, S.H. (1982) Increased brain dopamine and dopamine receptors in schizophrenia. Arch. Gen. Psychiatry, 39: 991-997. Michel, C., Brodde, O.E., Schepnel, B., Behrendt, J., Tschada, R., Motulsky, H.J. and Insel, P.A. (1988) [3H]idazoxan and some other a,-adrenergic drugs also bind with high affinity to a non adrenergic site. Mol. Pharmacol., 35: 324-330. Minnernan, K.P., Piffman, R.N. and Molinoff, P.B. (1981) P-adrenergic receptor subtypes: Properties distribution and regulation. Ann. Rev. Neurosci., 4: 419-462. Morris, G., Seidler, F. and Slotkin, T. (1983) Stimulation of ornithine decarboxylase by histamine or norepinephrine in brain regions of the developing rat: Evidence for biogenic amines as trophic agents in neonatal brain development. Life Sci., 32: 1565-1571. Morrow, A. and Creese, I. (1986) Characterisation of aladrenergic receptor subtypes in rat brain: A reevaluation of [3H]WB4104 and [3H]Prazosin binding. J. Pharmacol. Exp. Ther., 29: 321-330. Murphy, T.J. and Bylund, D.B. (1988) Characterisation of a,-adrenergic receptors in the OK cell line. J. Pharmacol. Exp. Ther., 244: 571-578. Palacios, J.M.P. (1984) Receptor autoradiography, 10 years later. J. Recept. Res., 4: 633-644. Palacios, J.M.P. and Kuhar, M.J. (1982) Beta adrenergic re-
ceptor localisation in rat brain by light microscopic autoradiography. Neurochem. Int., 4, 473-490. Palacios, J.M., Niehoff, D.L. and Kuhar, M.J. (1981) Receptor autoradiography with tritium sensitive film: Potential for computerised densitometry. Neurosci. Lett., 24: 111-116. Palacios, J.M.P., Hoyer, D. and Cortes, R. (1987) a,-adrenoceptors in the mammalian brain: similar pharmacology but different distribution in rodents and primates. Brain Res., 419: 65-75. Palacios, J.M., Rigo, M., Chinaglia, G. and Probst, A. (1990a) A reduced density of striatal somatostatin receptors in Huntington’s chorea. Brain Res., 522: 342-346. Palacios, J.M., Mengod, G., Vilaro, M.T., Wiederhold, K.H, Bodeke, H., Alvarez, F.J., Chinaglia, G. and Probst, A. (1990b) Cholinergic receptors in the rat and human brain microscopic visualisation. In Aquilonius S.M and Gillberg P.G. (Eds.) Progress in Brain Research, Vol. 84, pp. 243253. Parini, A,, Coupry, I., Graham, R.M., Uzielli, I., Atlas, D. and Lanier, S.M., (1989) Characterisation of an imidazoline/guanidinium receptive site distinct from the a,-adrenergic receptor. J. Biol. Chem., 264: 11874-11878. Pazos, A. (1988) Anatomical distribution of opioid receptors and adrenoceptors in the brain of mammals. Reu. Pharmacol. Clin. Exp. Ther., 5: 11-17. Pazos, A,, Engel, G. and Palacios, J.M.P. (1985a) Beta adrenoceptor blocking agents recognise a subpopulation of serotonin receptors in the rat brain. Brain Res., 343: 403-408. Pazos, A,, Probst, A. and Palacios, J.M.P. (1985b) P-adrenoceptor subtypes in the human brain: Autoradiographic localisation. Brain Res., 358: 324-328. Pieribone, V.A., Aston-Jones, G. and Bohn, M.C. (1988) Adrenergic and non-adrenergic neurons in the C1 and C3 areas project to locus coeruleus: A fluorescent double labelling study. Neurosci. Lett., 85: 297-303. Pompeiano, 0. (1988) Influence of the noradrenergic coerulospinal system on recurrent inhibition in the spinal cord and its role during postural reflexes. In H. Flohr (Ed.) Post-Lesion Neural Plasticity, Springer Verlag, Berlin, pp. 259-278. Probst, A,, Cortes, R. and Palacios, J.M.P. (1985) Distribution of a,-adrenergic receptors in the human brainstem: An autoradiographic study using [ 3H]p-aminoclonidine. Eur. J. Pharmacol., 106: 471-488. Rainbow, T.C., Parsons, B. and Wolfe, B.B. (1984) Quantitative autoradiography of betal- and beta2-adrenergic receptors in rat brain. Proc. Natl. Acad. Sci., USA, 81: 1585-9. Regan, J.W., Kobilka, T.S., Yang-Feng, T.L., Caron, M.G., Lefkowitz, R.J. and Kobilka, B.K. (1988) Cloning and expression of a human kidney cDNA for an a,-adrenergic subtype. Proc. Natl. Acad. Sci., USA, 85: 6301-6305. Schwinn, D.A., Lornasney, J.W., Lorenz, W., Szklut, P.J., Fremeau, R.T., Yang-Feng, T.L., Caron, M.G., Lefkowitz, R.J. and Cotecchia, S. (1990) Molecular cloning and expression of the cDNA for a novel a,-adrenergic receptor subtype. J. Biol. Chem., 265: 8183-8189. Stephenson, J.A. and Summers, R.J. (1985) Light microscopic
29 1 autoradiography of the distribution of [‘H]rauwolscine binding to a,-adrenoceptors in rat kidney. Eur. J. Pharrnacoi., 116: 271-278. Svensson, T.H. and Usdin, T. (1978) Feedback inhibition of brain noradrenaline neurons by tricyclic antidepressants: a-receptor inhibition. Science, 202: 1089-92. Trimmer, P.A. and McCarthy, K.D. (1986) Immunocytochemically defined astroglia from foetal newborn and young adult rats express b adrenergic receptors in vitro. Brain Res.. 392: 151-165. Unnerstall, J.R., Kopajtic, T.A. and Kuhar, M.J. (1984) Distribution of a 2 agonist binding sites in the rat and human central nervous system: Analysis of some functional, anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agents. Brain Res. Reu., 7: 69-101. van Giersbergen, P.L., Tierney, A.V.S., Wiegant, V.M. and de Jong, W. (1989) Possible involvement of brain opioid peptides in clonidine-induced hypotension in spontaneously hypertensive rats. Hypertension, 13: 83-90. Ventimiglia, R., Greene, M.I. and Geller, H.M. (1987) Localisation of P-adrenergic receptors on differentiated cells of the central nervous system in culture. Proc. Natl. Acad. Sci., USA, 84 5073-5077. Waeber, C., Rigo, M., Chinaglia, G., Probst, A. and Palacios, J.M.P (1991) Beta adrenergic receptor subtypes in the basal ganglia of patients with Huntingtons chorea and Parkinson’s disease. Brain Rex, in press. Weinshank, R.L., Zgombick, J.M., Macchi, M., Adham, N.,
Lichtblau, H., Branchek, T.A. and Hartig, P.R. (1990) Cloning, expression and pharmacological characterisation of a human a,,-adrenergic receptor. Mol. Pharmacol., 38: 681-688. Wikberg, J.E.S. and Uhlen, S. (1990) Further characterisation of the guinea pig cerebral cortex idazoxan receptor: Solubilisation, distinction from the imidazole site, and demonstration of cirazoline as an idazoxan receptor-selective drug. J. Nrurochem , 55: 192-203. Yakubu, M.A., Hamilton, C.A., Howie, C.A. and Reid, J.L. (1988) Idazoxan and brain a,-adrenoceptors in the rabbit. Brain Res., 463: 289-295. Young, W.S. and Kuhar, M.J. (1980) Noradrenergic a , - and a,-receptors: Light microscopic autoradiographic localisation. Proc. Natl. Acad. Sci., USA, 77: 1696-1700. Young, W.S. and Kuhar, M.J. (1981) Anatomical mapping of clonidine (alpha-2 noradrenergic) receptors in rat brain: relationship to function. Prog. Clin. Biol. Rex, 71: 41-52. Zaagsma, J. and Nahorski, S.R. (1990) Is the adipocyte betaadrenoceptor a prototype for the recently cloned atypical beta 3-adrenoceptor? Trends Pharmacol. Sci., 11: 3-7. Zeng, D.J.K., Harrison, D.D., D’Angelo, C.M., Barber, A.L., Tucker, Z.L. and Lynch, K.R. (1990) Molecular characterisation of a rat]a,b-adrenergic receptor. Proc. Natl. Acad. Sci., USA, 87: 3102-3106. Zonnenschein, R., Diamant, S. and Atlas, D. (1990) Imidazoline receptors in rat liver cells: A novel receptor or a subtype of a,-adrenoceptors? EY J. Pharmacol., 190: 203-21 5.
This Page Intentionally Left Blank
C D. Barnes and 0. Pompeiano (Eds.) Pr0gre.m ,n Bruin Research, Vol. XX 0 1991 Elsevier Science Publishers B.V.
293 CHAPTER 22
Actions of norepinephrine in the cerebral cortex and thalamus: implications for function of the central noradrenergic system D.A. McCormick, H.-C. Pape and A. Williamson Section of Neurobiology, Yale University School of Medicine, Cedar Street, New Hai,en, C r U.S.A.
Norepinephrine (NE) has potent and long-lasting ionic effects on cortical and thalamic neurons. In cortical pyramidal cells, activation of P-adrenergic receptors results in an enhanced excitability and responsiveness to depolarizing inputs. This enhanced excitability is expressed as a reduction in spike frequency adaptation and is mediated by a marked suppression of a slow Ca++-activated potassium current known as I,,,. In the thalamus, application of NE results in the suppression of ongoing rhythmic burst activity and a switch to the single spike firing mode of action potential generation.
This effect is mediated through an a,-adrenergic suppression of a resting leak potassium current, I,,, and through a P-adrenoceptor-mediated enhancement of the hyperpolarization activated cation current I,,. Together with the actions of other neuromodulatory neurotransmitters (i.e., acetylcholine, histamine, serotonin) these effects facilitate the switch of these neurons from a state of rhythmic oscillation and low excitability during drowsiness and slow-wave sleep to a state of increased excitability and responsiveness during periods of waking, attentiveness and cognition.
Key words: noradrenaline, neocortex, arousal, sleep, neuromodulation
Introduction
The mammalian cerebral cortex and thalamus are densely innervated by noradrenergic fibers arising almost exclusively from the brainstem nucleus, the locus coeruleus (LC) (Lindvall et al., 1974; Foote et al., 1983; Hughes and Mullikin, 1984; Morrison and Foote, 1986). In addition, p, a l , and a,-adrenoceptors have been localized within both the cerebral cortex and thalamus (Rainbow et al., 1984; Jones et al., 1985; Palacios et al., 1987). Stimulation of the LC modulates, through
the activation of adrenoceptors, the electrical activity generated by neurons in the cortex and thalamus (Rogawski and Aghajanian, 1980; Kayama et al., 1982; Sat0 et al., 1989). These, and other findings indicate that the central noradrenergic system is intimately involved in determining both the level of neuronal excitability and the pattern of activity generated by neurons in thalamocortical systems. Postsynaptic actions of norepinephrine (NE) have been determined in a number of CNS regions. In the brainstem, hypothalamus, and spinal cord, activation of a1
294
adrenoceptors is associated with a slow depolarization mediated by a decrease in a Kf conductance (Aghajanian, 1984,1985; Ma and Dun, 1985; Randle et at., 1986; reviewed by Supranant, 1990). In contrast, activation of p-adrenergic receptors in the hippocampus is associated with an increase in excitability through a reduction in spike frequency adaptation (Haas and Konnerth, 1983; Madison and Nicoll, 1986; Nicoll et al., 1990). The response to activation of a,-adrenoceptors has been extensively studied in the LC and is associated with inhibition mediated by an increase in membrane K + conductance (Williams et al., 1985; North, 1989). The precise ionic mechanisms by which NE modulates the pattern of activity generated by thalamic and cortical neurons is the subject of this chapter. Actions of NE in the cerebral cortex
liamson, 1989; Foehring et al., 1989); the specific ionic responses mediated by a 1 and a 2 adrenoceptors remain to be investigated. Activation of P-adrenergic receptors on cortical pyramidal cells results in a marked reduction in spike frequency adaptation through a block of the Ca+’-activated potassium current known as I,, (McCormick and Williamson, 1989; Foehring et al., 1989). In the absence of agonists, intracellular injection of a constant depolarizing current pulse into a cortical pyramidal cell results in the generation of a train of action potentials which slow down in frequency over time, a process known as spike frequency adaptation (Fig. 1, control). Spike frequency adaptation appears to result from the activation of a number of different potassium currents, of which two, I,, and I,, play a major role (Madison and Nicoll, 1984). IAHP is a potassium current which is activated by increases in the intracellular concentration of Ca into the micromolar range (Pennefather et al., 1985). It was named I,, because its activation by Ca++ entry, occurring during a train of action potentials, results in an afterhyperpolarization of the membrane potential following the train. I,, in contrast, is not activated by Ca++, but rather is sensitive to the membrane potential of the cell (Brown and Adams, 1980). Depolarization of the membrane above approximately - 75 mV results in increasing activation of I , which subsequently reduces the firing rate of the neuron. Together, these two potassium currents slow down the rate of action potential generation during a constant depolarizing current pulse (Fig. 1, control). Examination of the effects of NE, or other neurotransmitters, on I,, can be performed with a technique known as hybrid voltage clamp. In this technique, the cell is first induced to generate a train of action potentials with intracellular injection of a depolarizing current pulse. At the end of the pulse, the microelectrode amplifier is switched to voltage clamp mode and the voltage is “clamped” at the desired holding potential (typically -60 mV). The resulting current which the amplifier must supply in order to keep the +
Iontophoretic applications of NE to cerebral cortical neurons recorded extracellularly in vivo have revealed a wide variety of postsynaptic responses including weak inhibition, slow excitation and “modulation” (Bevan et al., 1977; Waterhouse et at., 1981; Waterhouse et al., 1988; Sat0 et al., 1989). In general, the weak inhibitory responses are localized throughout all layers and are mediated by p-adrenergic receptors while the slow excitatory responses are more prominent in the middle or deeper layers and are mediated by a-adrenoceptors, probably of the a I subtype (see below; Bevan et al., 1977; Waterhouse et al., 1981; Armstrong-James and Fox, 1983). Modulatory responses to NE have been reported to take the form of increases in the so-called “signal-tonoise” ratio; in other words, an increase in the ratio between the stimulus-specific response to baseline firing rate. These modulatory responses appear to be mediated by both a- and P-adrenergic receptors. Intracellular investigations of the actions of NE on cortical pyramidal cells have so far only revealed an ionic action in response to stimulation of p-adrenoceptors (McCormick and Wil-
+
A
Histamine
Control I
/
/
I
I
Recovery I
Methacholine
C
Norepinephrine
D
,,
E
I
n-Current Pulse
F
G,
Serotonin
J
L
Fig. 1. Histamine (HA), methacholine (MCh), norepinephrine (NE), and serotonin 6 H T ) all reduce spike frequency adaptation and IAHP in human neocortical neurons. A-D. Intracellular injection of a depolarizing current pulse results in the generation of a train of action potentials which shows spike frequency adaptation (Control). Application of HA (A; 500 p M in pipette), MCh (B; 1 mM), NE (C; SO0 pM) or 5-HT (D; 300-500 pM) results in a reversible reduction of adaptation. E-H. Intracellular injection of a current pulse was used to generate from 6 to 10 action potentials (current trace only shown). After the cessation of the pulse, the cell was switched to voltage clamp mode (held at -60 mV) and the after current examined. Application of all four agents reduced a slow outward component (IAHp) of this after current. HA and MCh both caused, in addition, an apparent inward current and became progressively therefore the traces in E and F were offset to match the pre-drug baseline for illustrative purposes. I,,,, smaller throughout the course of the experiment due in part to repeated applications of the four agents. All data, except C, were obtained from the same layer 111 human cortical neuron from the anterior temporal lobe. Although this cell displayed spike frequency adaptation to NE, this data was not suitable for illustration due to filtering (cutoff at 100 Hz) for examination of I,,,. Data in C was obtained from another anterior temporal human cortical neuron. (From McCormick and Williamson, 1989.)
membrane potential at this level is equal to that which is flowing through the ionic channels activated by the train of action potentials (Fig. 1E-H). A major component of this current is I,, (Fig. 1E). Application of NE, or the P-adrenoceptor agonist isoprenaline, results in the complete sup-
pression of I,, (Fig. 1G) and a marked reduction in spike frequency adaptation (Fig. 1C). This response appears to be a very general mechanism underlying noradrenergic influence on cortical pyramidal cells for it has been found in nearly every cortical pyramidal cell exhibiting spike fre-
296
quency adaptation in widely diverse regions of the cortex, including the cingulate, sensorimotor, and visual areas, and within CA, pyramidal cells and dentate granule cells of the hippocampus (Madison and Nicoll, 1986; Haas and Rose 1987; McCormick and Williamson, 1989; Foehring et al., 1989). Similarly, this response has been demonstrated in cortical pyramidal cells from a number of different species including rat, guinea pig, cat and human (Madison and Nicoll, 1986; McCormick and Williamson, 1989; Foehring et al., 1989). The exact cellular mechanisms by which NE reduces I, are not yet completely known, although the work by Madison and Nicoll in the hippocampus reveal that it probably involves the P-adrenergic stimulation of adenylyl cyclase (Madison and Nicoll, 1986). Presumably, activation of P-adrenergic receptors increases the production of cAMP through stimulation of adenylyl cyclase. This increase in intracellular concentrations of cAMP may then result in a decrease in the amplitude of I,,,, through a CAMP-dependent protein kinase. One possibility is that this A-kinase phosphorylates some critical portion of the I,, channel and results in a decrease in their sensitivity to the intracellular levels of Ca++. Interestingly, NE is not the only neurotransmitter which can reduce I, and therefore reduce spike frequency adaptation. For example, we have shown in the human neocortical pyramidal cells that histamine, acetylcholine, and serotonin are also capable of reducing I, even when examined in the same neuron (Fig. 1). This result indicates a remarkable convergence of neurotransmitter action between the various neuromodulatory transmitter agents, a property which is now known to be widespread in the central and peripheral nervous system (North, 1989; McCormick and Williamson, 1989; NicolI et al., 1990). The ability of NE to block spike frequency adaptation in cortical pyramidal cells would be expected to have very specific consequences for action potential generation in v i m If the neuron
were silent, blocking of I,, by NE may have no effect at all, since this current will most likely not be active. However, if the neuron were to be suddenly depolarized while P-adrenoceptors were activated, then the resulting train of action potentials will occur at a rate higher than that which would be generated in the absence of NE, particularly after the first 50 msec or so. Indeed, the longer the cell is depolarized, the greater the enhancement of spike firing resulting from suppression of I,,. If the neuron were tonically active, reduction of ,I will result in an increase in firing rate of the neuron, due to the reduction of tonic activation of this potassium current. In this manner, NE, acting through @-receptors,will selectively and potently enhance the response of cortical pyramidal cells to prolonged depolarizations, such as those associated with stimulation of a sensory receptive field. Although this ability to enhance the response of cortical pyramidal cells to trains of excitatory inputs may underlie the apparent ability of NE to enhance the response to stimulation of peripheral receptive fields in some neurons (e.g., Waterhouse et al., 1988), it does not explain two other findings: (1) the apparent weak inhibitory effects of NE in the cerebral cortex; and (2) the a-adrenoceptor-mediated slow, excitatory responses (Bevan et al., 1977; Waterhouse et d., 1981; Armstrong-James and Fox, 1983). Unfortunately, the cellular mechanisms underlying either of these effects are not yet known. However, the responses known to occur to NE in other regions of the central and peripheral nervous system suggest some possibilities. Madison and Nicoll found in hippocampal pyramidal cells that application of NE occasionally resulted in a hyperpolarization of the membrane potential in addition to the reduction in spike frequency adaptation. This small hyperpolarization was suggested to underlie the weak inhibitory influence of NE previously demonstrated in vivo (Madison and Nicoll, 1986). The hyperpolarization was associated with an increase in membrane conductance suggesting that it was mediated by opening of either K+ or C1-
297
synaptically or presynaptically, in the cerebral cortex. The ability of NE to excite cortical pyramidal cells through a-adrenoceptors probably results from a reduction of a resting potassium current, since this response has distinct from I, or I,,,, been demonstrated in a variety of other central neurons, including thalamic relay cells (see below; reviewed by Nicoll et d., 1990). This hypothesis remains to be tested thoroughly.
channels, although the specific ionic mechanisms have not yet been determined. A second possibility is suggested by the actions of N E in smooth muscle cells. In these cells, activation of P-adrenoceptors results in a selective enhancement of the M-current (Sims et al., 1988). If a similar response occurred in the cerebral cortex, it would be expected to reduce the ability of the neuron to generate ongoing activity (i.e., inhibition of spike activity). However, the voltage-dependent nature of I, would predict that the resulting inhibitory effects should be larger for large depolarizations than for spontaneous activity, a prediction which is opposite to that reported in viuo. Finally, a third possibility is that NE reduces the effectiveness of excitatory neurotransmission, either post-
Actions of NE in the thalamus Extracellular iontophoretic applications of N E onto thalamocortical relay neurons in the rodent LGNd or nucleus reticularis result in a potent
Cerebral Cortex
A
Normal
NE
Pyramidal Cell
Y
B
c
I I I I I I I I I I I I I I I I I I I l L L l
+9K
?
A Retina
Fig. 2. Postsynaptic actions of NE in the thalamus and cerebral cortex. A. Application of NE to a human cortical pyramidal cell in uitro greatly reduces spike frequency adaptation and results in an enhanced discharge. This effect is associated with a block of I A H P (see Fig. I). B. Application of NE to a thalamic neuron during the periodic injection of depolarizing current pulses (top trace). NE application results in a slow depolarization due to a reduction in membrane potassium conductance. The slow depolarization shifts the neuron from the burst firing mode (pre; expanded for detail) to the single spike mode (post). The cell in (€3) is a relay cell from the parataenial thalamic nucleus. Similar results were obtained in LGNd relay neurons. C . NE causes the same response in thalamic reticular (nRt) cells. The ionic actions of N E on cortical and thalamic interneurons are not known. (From McCormick, 1989.)
298
and prolonged excitation or increase in neuronal excitability (Rogawski and Aghajanian, 1980; Kayama et al., 1982). This response appears to be mediated though a,-adrenoceptors and can be activated by electrical stimulation of the LC (Rogawski and Aghajanian, 1980; Kayama et al., 1982). Stimulation of the LC in the cat also results in prolonged excitation of LGNd relay cells (Nakai and Takaori, 19741, although some authors report that iontophoretic application of NE to these neurons results in weak inhibition (Phillis et al., 1967). The reasons for these species differences are not known. The lack of significant species differences in cat and rodent LGNd in vitro (McCormick and Prince, 19881, indicates that the observed differences in viuo may result from differences in methodology. Application of NE to thalamocortical relay neurons in vitro results in two main effects, depending upon the pharmacological subtype of receptor stimulated. Activation of a,-adrenoceptors results in a marked slow depolarization resulting from the reduction of a resting "leak" potassium current which we have termed I,, (Fig. 2B). This slow depolarization can be 10-20 mV in amplitude and up to 2 min in duration, even after a single pulse application of NE (Fig. 2B). The widespread presence of a,-adrenoceptors in the thalamus (Jones et al., 1985) parallels the presence of this slow depolarizing response throughout the thalamus including the lateral and medial geniculate nuclei, the nucleus reticularis, the anteroventral nucleus, and the parataenial nucleus (McCormick and Prince, 1988). The slow depolarizing response to NE appears to be mediated by a G-protein, since the intracellular injection of a non-hydrolyzable GTP analogue, GTPy - S , results in responses which do not return to baseline, (e.g., they do not recover once activated). However, this slow depolarizing response is not blocked by pertussis toxin, indicating that this G-protein is not pertussis toxin sensitive and therefore not G, or Go (McCormick, unpublished observations). (The effectiveness of pertussis toxin treatment was confirmed in these cells by the
-
4
40 S
NE
4
80 s
WT
I InAI
v
IrnVI
"-....".
C0"tlOl
--05
(,
NE
- -1
5
Fig. 3. Responses of thalamic neurons to NE and 5-HT after pharmacological block of a, and a 2 adrenoceptors. A. Application of NE to a dorsal lateral geniculate neuron during the injection of hyperpolarizing constant current pulses (top trace) results in a small depolarizing and a large increase in apparent input conductance (bottom trace). Upward going lines spikes (indicated by asterisk). B. represent rebound Ca Application of 5-HT to a medial geniculate neuron has a similar effect. C,D. The I-V relationships obtained in voltage clamp before and after application of NE and 5-HT show a progressive increase in inward current at membrane potentials negative to approximately -60 mV. (From Pape and McCormick, 1989.) +
+
finding that it had blocked responses to the GABA, agonist baclofen.) The second messenger system (if any) through which NE results in a reduction in I,, is not known, although it does not appear to involve CAMP (see below). Interestingly, we have found that stimulation of muscarinic or histaminergic H receptors on thalamocortical neurons can also result in reduction of I (McCormick and Prince, 1987; McCormick and Feeser, unpublished results). Thus, as in cortical pyramidal cells, there appears to be a remarkable convergence of transmitter action in thalamic neurons. Activation of 6-adrenoceptors on thalamocortical relay cells results in a novel and interesting response: the membrane potential appears to change only slightly while the responsiveness of the cell to a hyperpolarizing current pulse is markedly reduced (Fig. 3A). Examination of this effect in voltage clamp revealed that stimulation
299
of P-adrenoceptors results in an enhancement of inward rectification at membrane potentials negative to approximately -60 mV (Fig. 3C). Further examination of this @-adrenoceptor effect revealed that it was mediated by an enhancement of the hyperpolarization-activated cation current 1, (Fig. 4). Hyperpolarization of thalamic relay neurons (and practically all principal neurons in the CNS and PNS) results in the activation of a current, I,, which is carried by both Na+ and K+ ions (Pape and McCormick, 1989). The activation of this current depolarizes the membrane potential back towards the reversal potential of I, ( - 43 mV). Subsequently, the activation of this current results in a depolarizing “sag” of hyperpolarizing responses, a feature which has previously been referred to as “anomalous rectification.” The P-adrenergic enhancement of I ,, subsequently, partially offsets the ability of the neuron to hyperpolarize in response to other inputs. Activation of P-adrenergic receptors on thalamocortical relay neurons results in a shift of the activation curve for 1, such that at any given voltage, a larger percentage of I, is active (Fig. 4). The resulting increase in the amount of I,
A
+0.5 V (mV)
- 90
-80
i
active at resting membrane potential results in an overall increase in the “leak” conductance of the membrane, thereby resulting in a decrease in the response of the cell to hyperpolarizing current pulses (Fig. 3A). In contrast, the response of the neuron to depolarizing current pulses appears to be largely unaltered, or in some cases, even slightly potentiated (unpublished observations). Interestingly, we have also found that stimulation of serotonergic receptors of an, as yet, unknown subtype can also result in an enhancement of I, (Fig. 3B, D), as can activation of H, histaminergic receptors (McCormick and Feeser, unpublished observations). These results indicate that I, is under the control of at least three different neurotransmitter systems in thalamocortical relay neurons. Enhancement of I, by NE, 5-HT and HA may occur through the activation of adenylyl cyclase since this response is mimicked by applicatim of the membrane permeable CAMP analogue 8bromo-CAMP, by activation of adenylyl cyclase directly with applications of forskolin (application of the inactive forskolin analogue 1,9-dideoxy-forskolin is without effect), or by inhibition of phos-
I
- 70
1s
Fig. 4. NE enhances a current activated by hyperpolarization. A. Graph of the instantaneous (circles) and steady-state (triangles) currents in response to a hyperpolarizing voltage step before (open symbols) and after (filled symbols) application of NE. Original traces are shown in B. Application of N E results in a marked enhancement of the inward current activated by hyperpolarization with only a small increase in instantaneous conductance. (From Pape and McCormick, 1989.)
300
phodiesterase activity with IBMX (Pape and McCormick, 1989). The possibility that these receptors may couple to adenylyl cyclase through G, remains to be tested. These results suggest that the sensitivity of I, to the voltage across the membrane is constantly under the control of the intracellular concentration of CAMP, which itself is under the control of stimulatory, and presumably inhibitory, G-proteins which are coupled to neurotransmitter receptors. Exactly how increases in intracellular concentrations of cAMP lead to an increase in the voltage sensitivity of I, is not known. However, one can easily imagine a process by which cAMP activates a protein kinase which phosphorylates some critical portion of I, channels, resulting in a shift in the electrical dipole of the voltage-sensing portion of the channel, and therefore an increase in sensitivity to the charge across the membrane. This ongoing equilibrium between phosphorylation and de-phosphorylation (controlled by the state of stimulation of @-adrenergic and other receptors and intracellular biochemical processes) would then result in the constant adjustment of the amplitude of 1,. Functional consequences of noradrenergic actions in the thalamus
Thalamic neurons display two very distinct states of action potential generation both in uivo and in uitro. During periods of slow-wave sleep and EEG synchronization, thalamocortical relay neurons display high-frequency (200-500 Hz) burst discharges, while during periods of waking and attentiveness, thalamic relay neurons display activity characterized by the occurrence of single spikes (see Fig. 5 ) (McCarley et al., 1983; Steriade and Deschcnes, 1984; Steriade and Llinis, 1988). Burst firing in thalamic neurons is due to the presence of a large, low-threshold Ca current (Jahnsen and LlinAs, 1984a,b; Coulter et al., 1989; Hernandez-Cruz and Pape, 1989). The voltagedependent properties of this current are such that when the membrane potential is negative to +
+
6 S sleep-
P sleep
S sleep
Z P sleep
U
k I
i
I
\
’
\
1
” 1
i15
100 msec
Fig. 5. Firing properties of thalamic neurons and their alteration with shifts in sleep-wake cycle. A. At -75 mV, thalamic cells respond to a depolarizing current pulse with a slow Ca++ spike (arrow) that triggers a burst of three Na+-dependent action potentials. This type of cellular activity is known as burst firing. Depolarizing the cell to -63 mV inhibits burst firing by inactivating the low threshold C a f + current. The response of the cell is now entirely passive,. Upon depolarization to -53 mV, the same amplitude current pulse generated a train of four action potentials. This latter pattern of spike activity is known as the single spike or transfer mode of action potential generation. Decreasing resting membrane potassium conductance (gK), as with NE stimulation of a1 adrenoceptors, is very effective in shifting the neuron from the burst to single spike firing mode. B. lntracellular recording from an LGNd neuron during transition from slow-wave sleep (S-sleep) to paradoxical sleep @sleep) and vice versa. SPOL (sommeil phasique a ondes lentes) is an intermediary stage between S-sleep and P-sleep and is characterized by ponto-geniculateoccipital (PGO) waves (Hirsch et al., 1983). Slow-wave sleep is characterized by the presence of rhythmic burst discharges (expanded in C1) and a relatively hyperpolarized membrane potential. Paradoxical sleep is characterized by single spike activity (C2) and a relatively depolarized membrane potential. Similar shifts in membrane potential are found in the transition from slow-wave sleep to waking. Depolarization of the membrane potential in paradoxical sleep may represent the influence of increased activity of brainstem cholinergic neurons while the depolarizing influence upon waking may be due to a mixture of cholinergic, noradrenergic, serotonergic and histaminergic influences. (B and C from Hirsch et al., 1983, with permission.)
301
-65 mV, thalamic neurons can exhibit a lowthreshold Ca++ spike that generates a highfrequency burst of fast Na+/K+-mediated action potentials (Fig. 5A, -75 mV). If the neuron is tonically depolarized positive to -65 mV, then the more traditional single spike, or transfer, mode of neuronal activity is generated (Fig. 5A, -53 mV). These two firing modes have very different functional consequences for the transfer of synaptic inputs to the cerebral cortex. The burst firing mode appears to be ideally suited for the generation of intrathalamic and thalamocortical rhythms, while the single spike firing mode is ideally suited for the faithful transmittal of fast synaptic inputs arriving as a consequence of the activation of peripheral receptive fields. For example, in the burst firing mode, the output of the relay neuron is highly non-linear in relation to the input that triggers it, while in the single spike firing mode, input and output are much more linearly related (McCormick and Feeser, 1990). The transfer of information through the thalamus to the cerebral cortex is accurate only when the thalamic neuron is in the transfer mode and not in the burst firing mode (Livingstone and Hubel, 1981; Steriade and LlinBs, 1988). The high correlation between the mode of action potential generation in single thalamic neurons and the state of the EEG in the animal gives rise to the exciting possibility that observations on thalamic neuronal membrane properties in vitro can be directly related to behavioral observations. The ability of neurotransmitters to change the firing mode of thalamic neurons from one state to the other is of particular interest because it may underlie, in part, the animal’s shift from a state of EEG synchronization (e.g., drowsiness, inattentiveness or slow-wave sleep) to one of EEG-desynchronization (e.g., arousal, alertness, and paradoxical sleep). Indeed, application of NE to thalamocortical relay neurons results in a potent and prolonged shift of the firing mode of these cells from burst firing to the generation of tonic activity (Fig. 2B). Intracellular recordings indicate that this shift in
firing mode is generated in large part by a large depolarization of the membrane potential resulting from reduction of I KL (McCormick, unpublished observations). In addition, we have also found that enhancement of I,, alone, either through the activation of P-adrenoceptors or serotonergic receptors, also results in a decrease in the ability of the cell to oscillate. However, although this P-receptor-mediated effect may suppress ongoing rhythmic activity, it only mildly promotes the generation of single-spike activity (Pape and McCormick, 1989; unpublished observations). Functional implications of noradrenergic responses
Based upon this and other information we would like to propose the following scenario for the contribution of the ascending noradrenergic system in activation of the thalamus and cortex. The transition from synchronized, rhythmic thalamocortical activity during slow-wave sleep to desynchronized, high-frequency activity during arousal and attentiveness is associated with an average increase in firing rate of noradrenergic, as well as cholinergic, serotonergic, and histaminergic, neurons (Trulson and Jacobs, 1979; Aston-Jones and Bloom, 1981; Vanni-Mercier et al., 1984; Lamour et al., 1986). The increased release of NE (as well as by ACh, 5-HT, and HA) will have numerous and complex actions on forebrain neurons (e.g., see McCormick, 1989; Nicoll et al., 1990). In the thalamus, the increased release of NE will cause a slow depolarization of thalamocortical relay cells as well as neurons in the nucleus reticularis (Fig. 2). In addition, increases in the release of ACh and HA onto thalamocortical relay neurons will also add to this slow depolarizing influence (McCormick and Prince, 1987; McCormick and Feeser, unpublished observations). The resulting slow depolarization will inactivate the low-threshold C a f + spike and move the membrane potential out of the voltage range in which I, is active. Consequently, the depolariza-
302
tion towards single spike firing threshold, the increase in specific membrane resistance, and the reduced responsiveness to hyperpolarizations (through enhancement of Ih), will all increase the likelihood that phasic EPSPs will trigger action potentials in a one-to-one manner (McCormick and Feeser, 1990) and thus result in an increase in the faithful transfer of incoming spike trains to the cerebral cortex. The actions of NE, and other modulatory neu-’ rotransmitters, in the cerebral cortex will contribute to the determination of the fate of these synaptic potentials arriving from the thalamus. In cortical pyramidal cells, the reduction of the Ca++-activated K + current, I,,,, by NE (as well as by ACh, 5-HT, and HA) and the voltageactivated potassium current I by ACh and 5-HT (McCormick and Williamson, 1989) will enable these cells to respond faithfully to the incoming trains of EPSPs. Selective activation of individual NE fibers may allow for activation of localized regions of the cerebral cortex, or functionally related groups of cells, which may correspond to those which are activated in the thalamus. Together, these modulatory actions in both thalamic and cortical neurons may facilitate the pattern of neuronal processing which is associated with cognition. Evidence for and against a role of the central noradrenergic system in arousal
Anatomical, electrophysiological, and pharmacological data indicate that the central noradrenergic system arising from the LC operates by modulating the excitability and “state” of neurons in divergent regions of the CNS, probably in a coordinated and unified fashion, in order to achieve some goal. The question then becomes, what goal? The most accurate answer to this question would entail a complete detailing of all features of this neurotransmitter system including the precise anatomical, pharmacological and physiological properties of every synapse which it makes and receives, as well as the interactions of these
synapses with the physiological properties of each of the pre- and post-synaptic elements involved. Barring such a comprehensive knowledge of this system, we are forced to resort to forming generalizations, sometimes using vague and potentially inaccurate terms. One such term is “arousal.” The normal nervous system appears to exist within a continuum of three broad states: aroused, synchronized sleep, and desynchronized (also known as REM or paradoxical) sleep. Arousal and rapid eye movement (REM) sleep are associated with an EEG which exhibits a preponderance of higher frequencies, while synchronized sleep is characterized by the preponderance of slower (1-12 Hz) frequencies in the EEG. These two states of the EEG, in turn, correspond to a preponderance of neuronal activity associated with the ongoing interactions of large number of synapses, neurons, and neuronal circuits during arousal and REM sleep, or to the relatively slow, synchronous firing of groups of neurons in thalamocortical circuits during slow-wave sleep. Here we use the term arousal to indicate a state in which the nervous system is activated and the animal is awake and is either directly interacting with the world around it, or has the potential to quickly do so in response to some stimulus. Thus, we do not refer to the desynchronization of the EEG which occurs during REM sleep as arousal, per se, but rather as EEG activation. Although specific exceptions to this generalization can be raised, these terms are useful linguistic tools for discussion. A role of the central noradrenergic system in “arousal” is suggested largely upon the findings of physiological and pharmacological studies. The activity of central noradrenergic neurons are known to increase in anticipation of increases in ‘‘arousal’’ (Aston-Jones, 1985; Jacobs, 1986). In addition, presentation of a novel stimulus which attracts the animal’s attention is associated with a strong burst of action potentials in LC neurons (Aston-Jones, 1985). Intraventricular injections of NE have a potent “activating” influence upon the EEG and “arousing” influence upon the animal’s behavior (Matsuda, 1968, 1969; Cordeau er aZ.,
303
1971), while central administration of antagonists facilitate the appearance of EEG-synchronization and behavioral sleep (Matsuda, 1968, 1969). The specific ionic actions of NE on thalamic and cortical neurons also suggest a role in “arousal” since these actions result in a shift in the responsiveness and firing properties of these neurons similar to those which occur during the shift from synchronized sleep to waking (see above). The arguments against a role for the central NE pathways in arousal typically center around two findings: (1) noradrenergic neurons in the LC are specifically inactive during REM sleep even though the EEG is desynchronized (Aston-Jones, 1985) and (2) extensive lesions of the LC do not impair the animal’s ability to exhibit relatively normal sleep-wake cycles (Jones et al., 1977). Although a lack of a role for the central NE-system in “arousal” is one possible interpretation of these results, a more parsimonious explanation is suggested by the finding that there are at least five other systems which also contribute to activation of the nervous system: (1) brainstem cholinergic projections to thalamus; (2) basal forebrain cholinergic projections to cortex and thalamus; (3) brainstem 5-HT system; (4) hypothalamic histaminergic projections to thalamus and cortex; ( 5 ) hypothalamic projections to cortex which contain an a-MSH-like peptide. During REM sleep, extracellular recordings indicate that brainstem cholinergic neurons become markedly active, while NE, 5-HT and perhaps HA neurons fall silent (Trulson and Jacobs, 1979; Vanni-Mercier et al., 1984; Aston-Jones, 1985; Jacobs, 1986; Lamour et al., 1986). Thus, the activation of the EEG in this state may result largely from the central actions of ACh, although it is unlikely that this transmitter is solely responsible for this state. Similarly, the lack of pronounced alterations in the sleep-wake cycle after large lesions of the LC is probably due to the presence of the five other systems which are capable of generating activation of the nervous system and behavioral arousal in the absence of the noradrenergic system. In this general sense,
the lesion results do not reveal any evidence as to whether or not the LC is involved in inducing the cellular phenomena which underlie arousal, but rather merely state that it is probably not essential for it to occur when all of the other systems are left intact. Conclusions
The anatomical and physiological features of the central noradrenergic system suggest that it contributes to the control of the “processing state” of the central nervous system. The postsynaptic actions of N E in the thalamus and cerebral cortex are consistent with a role in the control of neuronal excitability and responsiveness to other, more phasic, postsynaptic potentials. These noradrenergic actions suggest that activation of the LC will lead to an abolition of slow rhythmic oscillations in thalamocortical systems and an increase in responsiveness of these neurons to inputs from peripheral sensory receptors. In this manner, the central noradrenergic system, with the other modulatory transmitter systems, may control the level of “arousal” and readiness of the CNS. Acknowledgements
Supported by NIH, the Klingenstein Fund, Pattison Fund, the Sloan Foundation, and a fellowship from the Deutsche Forschungemeinschaft. References Aghajanian, G.K. (1984) The physiology of central alpha- and beta-adrenoceptors. In Catecholamines: Neuropharmacology and Central Nervous System - Theoretical Aspects, Alan R. Liss, New York, pp. 85-92. Aghanjanian, G.K. (1985) Modulation of a transient outward current in serotonergic neurones by 0,-adrenoceptors. Nalure (London), 315: 501-503. Armstrong-James, M. and Fox, K. (1983) Effects of ionophoresed noradrenaline on the spontaneous activity of neurones in rat primary somatosensory cortex. J. Physiol. (London), 335: 427-447. Aston-Jones, G. (1985) Behavioral function of locus coeruleus derived from cellular attributes. Physiol. Psychol., 13: 118126.
304 Aston-Jones, G. and Bloom, F.E. (1981) Norepinephrine containing locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 876-886. Bevan, P., Bradshaw, C.M. and Szabadi, E. (1977) The pharmacology of adrenergic neuronal responses in cerebral cortex: Evidence for excitatory a- and inhibitory 0-receptors. Br. J. Pharmacol., 59: 635-641. Brown, D.A. and Adams, P. (1980) Muscarinic suppression of a novel voltage-sensitive K + current in a vertebrate neurone. Nature (London), 283: 673-676. Coulter, D., Huguenard, J and Prince, D.A. (1989) Calcium currents in rat thalamocortical relay neurones: Kinetic properties of the transient, low-threshold current. J. Physiol. (London), 414: 587-604. Cordeau, J.P., Champlain, J. De and Jacks, B. (1971) Excitation and prolonged waking produced by catecholamines injected into the ventricular system of cats. Can. J. Physiol. Pharmacol., 49: 627-631. Foehring, R.C., Schwindt, P.C. and Crill, W.E. (1989) Norepinephrine selectively reduces slow Ca++- and Na+-mediated K+ currents in cat neocortical neurons. J. Neurophysiol., 61: 245-256. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence for anatomical and physiological specificity. Physiol. Reu, 63: 844-914. Haas, H:L. and Konnerth, A. (1983) Histamine and noradrenaline decrease calcium-activated potassium conductance in hippocampal pyramidal cells. Nature (London), 302: 432-434. Haas, H.L. and Rose, G.M. (1987) Noradrenaline blocks potassium conductance in rat dentate granule cells in citro. Neurosci. Lett., 78: 171-174. Hernandez-Cruz, A. and Pape, H.-C. (1989) Identification of two calcium currents in acutely dissociated neurons from the rat lateral geniculate nucleus. J. Neurophysiol., 61: 1270-1283. Hirsch, J.C., Fourment, A. and Marc, M.E. (1983) Sleep-related variations of membrane potential in the lateral geniculate body neurons of the cat. Brain Res., 259: 308312. Hughes, H.C. and Mullikin, W.H. (1984) Brainstem afferents to the lateral geniculate nucleus of the cat. Exp. Brain Res., 54: 253-258. Jacobs, B.L. (1986) Single unit activity of locus coeruleus neurons in behaving animals. Prog. Neurohiol., 27: 183-194. Jahnsen, H. and LlinCs, R. (1984a) Electrophysiological properties of guinea-pig thalamic neurones: An in uitro study. J. Physiol. (London), 349: 205-226. Jahnsen, H. and Llinis, R. (1984b) Ionic basis for the electroresponsiveness and oscillatory properties of guinea-pig thalamic neurones in uitro. J. Physiol. (London), 349: 227-247. Jones, B.E., Harper, S.T. and Halaris, A.E. (1977) Effects of locus coeruleus lesions upon cerebral monoamine content, sleep-wakefulness states and the response to amphetamine in the cat. Brain Res., 124: 473-496.
Jones, L.S., Gauger, L.L. and Davis, J.N. (1985) Anatomy of brain alpha,-adrenergic receptors: In citro autoradiography with [125I]-HEAT. J. Comp. Neurol., 231: 190-208. Kayama, Y. Negi, T., Sugitani, N. and Iwama, K. (1982) Effects of locus coeruleus stimulation on neuronal activities of dorsal lateral geniculate nucleus and perigeniculate nucleus of the rat. Neuroscience, 7 655-666. Lamour, Y., Dutar, P., Rascol, 0. and Jobert, M. (1986) Basal forebrain neurons projecting to the rat frontoparietal cortex: Electrophysiological and pharmacological properties. Brain Res., 362: 122-131. Lindvall, O., Bjorklund, A,, Nobin, A. and Stenevi, U. (1974) The adrenergic innervation of the rat thalamus as revealed by the glyoxylic acid fluorescence method. J. Comp. NeuroL, 154: 317-348. Livingstone, M.S. and Hubel, D.H. (1981) Effects of sleep and arousal on the processing of visual information in the cat. Nature (London), 291: 554-561. Ma, R.C. and Dun, N.J. (1985) Norepinephrine depolarizes lateral horn cells of neonatal rat spinal cord in citro. Neurosci. Lett., 60: 163-168. Madison, D.V. and Nicoll, R.A. (1984) Control of repetitive discharge of rat CA, neurones, in citro. J. Physiol. (London)., 354: 319-331. Madison, D.V. and Nicoll, R.A. (1986) Actions of noradrenaline recorded intracellularly in rat hippocampal CA pyramidal cells, in uitro. J. Physiol. (London),372: 221-244. Matsuda, Y. (1968) Effects of intraventricularly administered adrenaline on rabbit’s EEG and their modification by adrenergic blocking agents. Jpn. J. Pharmacol., 18: 139152. Matsuda, Y. (1969) Effects of some sympathomimetic amines administered intraventricularly on rabbit’s EEG. Jpn. J. Pharmacol., 19: 102-109. McCarley, R.W., Benoit, 0. and Barrionuevo, Ci. (1983) Lateral geniculate nucleus unitary discharge in sleep and waking: State- and rate-specific aspects. J. Neurophysiol., 50: 798-818. McCormick, D.A. (1989) Cholinergic and noradrenergic modulation of thalamocortical processing. TINS, 12: 215-221. McCormick, D.A. and Feeser, H.R. (1990) Functional consequences of burst firing in lateral geniculate relay neurons. Neuroscience, 39: 103-113. McCormick, D.A. and Prince, D.A. (1987) Actions of acetylcholine in the guinea-pig and cat medial and lateral geniculate nuclei, in uitro. J. Physiol. (London), 392: 147-165. McCormick, D.A. and Prince, D.A. (1988) Noradrenergic modulation of firing pattern in guinea pig and cat thalamic neurons, in citro. J. Neurophysiol., 59: 978-996. McCormick, D.A. and Williamson, A. (1989) Convergence and divergence of neurotransmitter action in the human cerebral cortex. Proc. Natl. Acad. Sci. USA, 86: 8098-8102. Morrison, J.H. and Foote, S.L. (1986) Noradrenergic and serotonergic innervations of cortical, thalamic, and tectal visual structures in old and new world monkeys. J. Comp. Neurol., 43: 117-138. Nakai, Y. and Takaori, S. (1974) Influence of norepinephrine-
,
305 containing neurons derived from the locus coeruleus on lateral geniculate neuronal activity of cats. Brain Rex, 1: 47-60. Nicoll, R.A., Malenka, R.C. and Kauer, J.A. (1990) Functional comparison of neurotransmitter receptor subtypes in mammalian central nervous system. Physiol. Rec., 70: 5 13-565. North, R.A. (1989) Drug receptors and inhibition of nerve cells. Br. J. Pharmacol., 98: 13-28. Palacios, J.M., Hoyer, D. and Cortes, R. (1987) a,-adrenoceptors in the mammalian brain: Similar pharmacology but different distribution in rodents and primates. Brain Rex, 419: 65-75. Pape, H.-C. and McCormick, D.A. (1989) Noradrenaline and serotonin selectively modulate thalamic burst firing by enhancing a hyperpolarization activated cation conductance. Nature (London), 340: 715-718. Pennefather, P., Lancaster, B., Adarns, P.R. and Nicoll, R.A. (1985) Two distinct CA-dependent K currents in bullfrog sympathetic ganglion cells. Proc. Natl. Acnd. Sci USA, 8 2 3040-3044. Phillis, J.W., Tebecis, A.K. and York, D.H. (1967) The inhibitory action of monoamines on lateral geniculate neurones. J. Physiol. (London), 190: 563-581. Rainbow, T.C., Parsons, B. and Wolfe, B.B. (1984) Quantitative autoradiography of beta,- and beta,-adrenergic receptors in rat brain. Proc. Natl. Acad. Sci. USA, 81: 158551589, Randle, J.C.R., Bourque, C.W. and Renaud, L.P. (1986) a,Adrenergic receptor activation depolarizes supraoptic neurosecretory neurones in ritro. Am. J. Physiol., 251: R569-R574. Rogawski, M.A. and Aghajanian, G.K. (1980) Modulation of lateral geniculate neurone excitability by noradrenaline rnicroiontophoresis or locus coeruleus stimulation. Nature (London), 287: 731-734.
Sato, H., Fox, K. and Daw, N.W. (1989) Effect of electrical stimulation of locus coeruleus on the activity of neurons in the cat visual cortex. J. Neurophysiol., 62: 946-958. Sims, S.M., Singer, J.J. and Walsh, J.V. (1988) Antagonistic adrenergic-muscarinic regulation of M current in smooth muscle cells. Science, 239: 190-193. Steriade, M. and Deschhes, M. (1984) The thalamus as a neuronal oscillator. Brain Rex, 320: 1-63. Steriade, M. and LlinPs, R. (1988) The functional states of the thalamus and the associated neuronal interplay. Physiol. Rec., 68: 649-742. Suprenant, A. (1990) The neurotransmitter noradrenaline and its receptors. Semin. Neurosci., 1: 125-136. Trulson, M.E. and Jacobs, B.L. (1979) Raphe unit activity in freely moving cats: Correlation with level of behavioral arousal. Brain Rex, 163: 135-150. Vanni-Mercier, G., Sakai, K. and Jouvet, A. (1984) Specific neurons for wakefulness in the posterior hypothalamus in the cat. C.R. Acad. Sci. Ser. (Paris) 111, 298: 195-200. Waterhouse, B.D., Moises, H.C. and Woodward, D.J. (1981) Alpha-receptor-mediated facilitation of somatosensory cortical neuronal responses to excitatory synaptic inputs and iontophoretically applied acetylcholine. NeuropharmaC O ~ O ~ 20: Y , 907-920. Waterhouse, B.D., Sessler, F.M., Cheng, J.-T., Woodward, D.J., Azizi, S.A. and Moises, H.C. (1988) New evidence of a gating action of norepinephrine in central neuronal circuits of mammalian brain. Brain Res. Bull., 21: 425-432. Williams, J.T., Henderson, G. and North, R.A. (1985) Characterization of alpha,-adrenoceptors which increase potassium conductance in rat locus coeruleus neurones. Neuroscience, 14: 95-101.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
307 CHAPTER 23
Noradrenergic and locus coeruleus modulation of the perforant path-evoked potential in rat dentate gyrus supports a role for the locus coeruleus in attentional and memorial processes C. Harley Department of Psychology, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada
The perforant path-dentate gyrus synapse has provided a model system for functional neural plasticity in adult mammalian brain. NMDA-dependent long-term changes in neural connectivity occur at this synapse in response to highfrequency input. Norepinephrine (NE) applied exogenously or released endogenously can initiate both a short- and a long-term potentiation (LTP) of the dentate gyrus response to perforant path input. Triggering of the potentiated response depends on preceptor activation and does not require a high-frequency stimulus. An increase in locus coeruleus (LC) activity can initiate both short and LTP of the perforant path
response, although a reduction in LC activity does not alter baseline perforant path responses. This chapter considers differences between NE modulation in uitro and in vivo, differences and similarities between NE-LTP and frequencyinduced LTP, and the surprising specificity of NE effects at the perforant path synapse. Studies of NE in the dentate gyrus support a role for the LC in promoting both short- and long-term enhancement of responses to complex sensory inputs and are consistent with a role for the LC in memorial as well as attentional processes.
Key words: dentate gyrus, hippocampus, locus coeruleus, P-receptors, norepinephrine, long-term potentiation
Introduction During the last two decades the model system characterized by perforant path activation of the dentate gyrus-evoked potential has proved seminal for a new neurobiology of learning and memory. The demonstration that the perforant path dentate gyms synapse is capable of long-term potentiation (LTP) by repetitive high-frequency stimulation of the perforant path (Bliss and Mmo,
1973), and the subsequent insight that activation of a new class of voltage and chemically dependent synaptic channels, the NMDA receptor complex, was the trigger event for LTP has provided physiological support for the Hebbian model of associative learning (Wigstrom and Gustafsson, 1985). The present review discusses evidence that another synapse, the noradrenergic synapse originating from the locus coeruleus (LO, participates
308
in initiating long-term changes in the response of the dentate gyrus to perforant path input. After reviewing the effects of norepinephrine (NE) application and LC activation on the dentate gyrusevoked potential, we will consider briefly how these effects might be understood at the cellular level and, finalIy, speculate as to the role of these effects in the functioning of the dentate gyrus. Dentate gyms-evoked potential modulation by NE B.
Long-lasting potentiation In 1983 we reported that NE, iontophoresed for 2-5 min at the dentate gyrus cell body layer in the anesthetized rat, produced a significant potentiation of the perforant path evoked population spike which lasted for many minutes (Neuman and Harley, 1983). A single experiment documenting the evoked potential change induced by NE is shown in Figure 1. The profile of NE-induced long-term change averaged over 16 experiments is shown in Figure 2. Although the experiment in Figure 1 shows an example of an increase in the field EPSP, there were no consistent changes in EPSP amplitude over the 16 experiments of Figure 2. Population spike increases were reliable however, and whether or not the induced changes in population spike were long-lasting, an initial short-lasting potentiation was typical. ,%-Receptormediation Using the brain slice preparation we were able to confirm that NE applied at 10 p M for 10 min could consistently produce potentiation of the perforant path-evoked population spike (Lacaille and Harley, 1985). In 25% of these experiments the changes were long-lasting. Increases in the EPSP slope were seen in 50% of the experiments, but they did not account for the entirety of the population spike increases even in those experiments. In the brain slice, it was possible to identify the &receptor as the adrenergic receptor critical for NE's potentiation effect. Isoproterenol, a P-agonist, produced the population
w
NE, 100 nA
: 50 rn Sec.
C.
D.
25 min after B
E
'
I
\ Chart time 1 min
50 rnin after B
Fig. 1. Top panel (left). Schematic of the recording and iontophoretic site in the dentate gyrus and of the data display system. Top panel (right). Tracing of perforant path-evoked potential indicating measurement of the population spike amplitude. Panel A is a polygraph writeout of the control evoked potentials taken prior to the application of norepinephrine (NE) in panel B. Three to four min following initiation of N E iontophoresis there is a clear increase in population spike amplitude as seen in panel C. Panel E demonstrates the long-lasting nature of the potentiation. An increase in EPSP height is also observable. The 50 msec calibration is the "real-time" calibration for the field potential. (From Neuman and Harley, 1983.)
spike potentiation, and timolol, a &antagonist, blocked the occurrence of NE-induced potentiation. a,-Receptor activation had no effect or a depressant effect on the population spike.
Similarities of NE modulation and high-frequencyinduced L TP Sarvey's group replicated our observation of long-lasting facilitation by NE in the brain slice
309
14~r
-3
NE
T
I
9
I
15
I
24
T -
r
I
30
Time (min)
Fig. 2. Average potentiation of population spike amplitude in 16 observations of NE-induced long-lasting potentiation. The perforant path was stimulated once every 10 sec. Each point represents an average of the 10 preceding population spikes. Bars represent standard errors of the mean. (From Neuman and Harley, 1983.)
preparation and made several important additional observations. In their experiments 5-50 p M NE applied for 30 min invariably produced a long-lasting potentiation of the population spike. They also found, measuring the EPSP in the dendritic layer, that long-lasting EPSP slope increases consistently accompanied long-lasting potentiation of the population spike (Stanton and Sarvey, 1987). P-Receptor activation was again found to be the critical event for initiating both short- and long-lasting potentiation (Stanton and Sarvey, 1985a). Of particular interest is their series of slice experiments demonstrating critical links between high frequency-induced LTP and the long-lasting potentiation induced by NE. First, they provided evidence that high frequency-induced LTP is “virtually eliminated” in the dentate gyrus if NE is depleted by 6-OHDA (Stanton and Sarvey, 1985a). Similarly the P,-blocker, metoprolol, also eliminated high frequency-induced LTP as it had NE-induced LTP (Stanton and Sarvey, 1985a). Second, they showed that NE-induced long-lasting effects and high frequency-induced LTP can both be blocked by applications of protein synthesis inhibitors (Stanton and Sarvey, 1984,1985~) and that both initiate increases in CAMP concentrations (Stanton and Sarvey, 1985b). Third, and most importantly, they showed that application of
an NMDA receptor blocker prevents short and long-lasting NE-induced potentiation just as it prevents high frequency-induced LTP (Burgard et al., 1989; see Fig. 3). Bliss’s group has also provided evidence of a parallel between high frequency-induced and NE-induced potentiation. Both manipulations produce a significant increase in potassium-evoked glutamate release in the dentate gyrus (Lynch et al., 1985; Lynch and Bliss, 1986). The ability to block NE-induced potentiation and high frequency-induced LTP by similar manipulations, and in particular by the use of an NMDA blocker, suggests that these two forms of potentiation may converge on the same plasticity mechanism. Stanton and Heinemann (1986) demonstrated that one effect of NE in the slice is to increase the entry of calcium into cells of the dentate gyrus during a high-frequency stimulus. This effect was blocked by propranolol. Curiously 240
0 LTP or
200
--
1 /IM ontogonist
0
y
a E
NELLP
180
a 10#M ontogonist 160
0
-E
140
0
120
a 0
m
.-“ 100 80 0-APV HFT
0-APV
50gM
CPP
NE
Fig. 3. The effects on NMDA antagonists on long-term potentiation (LTP) and NELLP (norepinephrine-induced long-lasting potentiation). Open bars show potentiation of population spike amplitude produced by a high frequency stimulus train (HFT) or 50 p M NE. Black bars show the effects of a 1 Km concentration of either D( -)APV or CPP on the potentiation produced by HFT or NE. Hatched bars show t h e effect of 10 p M D(-)APV or CPP on potentiation. Each bar is the mean% baseline amplitude fS.E.M. of the population spike taken at the end of the final wash period. The number of experiments in each group appears within the bar. Asterisk denotes a significant difference ( P < 0.05, ANOVA plus Duncan’s test) in the mean compared to control potentiation. (From Burgard et al., 1989.)
310
the increase in calcium influx was not observed in the molecular layer, only in the cell layer. The entry of calcium is known to be pivotal in the cascade of events initiated by NMDA receptor activation (Sarvey et al., 1989). Gray and Johnston (1987) have shown that NE and p-agonists increase voltage-dependent calcium currents in whole clamped granule cells, while Radhakrishnan and Albuquerque (1989) have recently reported that propranolol reduces NMDA-activated currents. The model of NE action which emerges from the work of Sarvey and Stanton is one in which NE makes more probable the activation of NMDA receptors by single perforant path pulses. Lacaille and Schwartzkroin (1988) reported that one action of NE via @receptor activation is a relatively prolonged, but mild, depolarization of granule cell membranes accompanied by an increase in membrane resistance - probably mediated by a decreased potassium conductance. Stanton, Mody and Heinemann confirmed the observation of a p-mediated depolarization with increased membrane resistance for dentate granule cells (Stanton et al., 1989). In their report such depolarizations were long-lasting and could continue for more than an hour after NE washout. The long-lasting depolarizations were blocked by an NMDA receptor blocker, but a short-lasting depolarization was still seen. Since neither Sarvey nor Stanton see a short-lasting increase in the population spike induced by NE when an NMDA receptor blocker is present, it does not seem that the depolarization effect contributes to the enhanced population spike. It would appear that the actions of NE which are important for the enhanced population spike are those which enhance or activate NMDA currents. The depolarization effect would provide a condition favorable for NMDA receptor activation by glutamate pulses, in addition, as previously mentioned, NMDA currents would themselves be enhanced. One puzzling and possibly contradictory observation is the report that iontophoresed NE increases theta cell activity in the dentate gyrus
(Rose and Pang, 1989). Given the presumed inhibitory role of theta cells, one might expect a reduced probability of LTP-like effects with NE. Since this does not seem to be the case, it may be that cellular, not dendritic, inhibition is enhanced by NE’s effect on theta cells. It has been observed that cellular inhibition has little effect on the probability of LTP, although it would be expected to enhance the specificity of circuit output, while dendritic inhibition can effectively prevent LTP-like changes (Douglas, personal communication). Finally, in the Pang and Rose study, the effect of iontophoresing the P-agonist, isoproterenol, was an increase in granule cell firing, as predicted from the depolarizing effects of preceptor activation, as well as an increase in theta cell firing. If NE in the dentate gyrus acts as a facilitator of NMDA-induced plasticity then it is possible to argue that NE potentiation should be specific to synapses activated during the interval when NE is exerting its effects and that NE potentiation of the evoked potential participates in the Hebbian properties of LTP, ie., the requirement for temporal contiguity of synaptic inputs. Experiments have not yet demonstrated the synaptic specificity of NE potentiation. In an early attempt to look at specificity, using the slice preparation, we assessed the requirement for temporal contiguity by comparing potentiation induced by pairing perforant path activation and NE application with potentiation induced by applying NE during a period of no perforant path stimulation. Similar levels of potentiation were seen in the washout period in both experiments (Lacaille and Harley, 1985). However given the likelihood that NE-induced depolarization was still present at the onset of the washout, the results seemed inconclusive as regards the pairing specificity of NE potentiation. More recently, however, Dahl and Sarvey applied the P-agonist isoproteronol for 30 min and then waited for a 30 min washout period without stimulation; adrenergic potentiation of EPSP slope (population spikes were not evoked) was as great
31 1
as that seen when stimulation occurred at regular intervals throughout the period of NE agonist application (Dahl and Sarvey, 1990). This seems to contradict the hypothesis that NE long-lasting effects are dependent on NMDA receptor activation. Can NE enhance response to synaptic inputs for an extended period, whether or not they are active at the time of NE release? If so, how? The slice work provides support for the hypothesis that LTP and NE-induced potentiation are interdependent and converge on common mechanisms in the dentate gyrus, however a parallel story has not yet been developed in uiuo. Evidence for a role of NE in dentate gyrus LTP processes in uiuo is, at best, equivocal. Depletion of NE in viuo has been reported in one study to attenuate the EPSP slope increase produced by high-frequency stimulation in the dentate gyrus but not to prevent LTP of the population spike (Bliss et al., 1983). In a second study NE depletion affected baseline responding and appeared to increase LTP of the EPSP slope while reducing LTP of the population spike (Robinson and Racine, 1985). When the effects of depletion on baseline responses were taken into account, no overall effect of NE depletion on dentate gyrus LTP was observed. We have also failed to observe attenuation of LTP when NE release is blocked by clonidine (unpublished observations). This may not be a serious discrepancy since it appears more difficult to elicit dentate LTP in the slice than in uiuo. Facilitating factors, e.g., blockade of GABA inhibition (Wigstrom and Gustafsson, 1983) and, possibly, depolarization by NE may be required for reliable induction of dentate LTP in the slice. If N E operates to promote long-lasting synaptic change in the dentate gyrus by facilitating LTP, then it operates in conjunction with a number of potential facilitating events. NE’s role may be more or less important depending on those other events. NE has also been shown to play a facilitatory role in the LTP of EPSP slope in area CA3 of the hippocampus (Hopkins and Johnston, 1984), although NE alone is not sufficient in CA3 to
produce LTP-like changes if only single pulses are used as stimuli. This is a P-receptor-mediated effect and p-blockers prevent CA3 LTP from occurring in the slice. Interestingly, blockade of GABA inhibition overcomes this P-blocker effect and LTP can again be induced. Here LTP is not produced via an NMDA mechanism so that it is clear NE’s role in facilitating long-term change need not be restricted to NMDA synapses. Johnston et al. (1988) hypothesized that the increase in voltage-dependent calcium conductance which they have observed in granule cells and the decrease in calcium-activated potassium conductances which has also been reported to be produced by NE in granule and pyramidal cells (Haas and Rose, 1987; Madison and Nicoll, 1982) may act together in CA3 cells to promote a calcium induction of LTP. Winson and Dahl (1985) replicated our report of NE-induced long-lasting facilitation in uivo using 5 min of iontophoresed NE. The development of the effect in their figures parallels the development seen here in Figures 1 and 2. Winson and Dahl’s pharmacological effects in the dentate gyrus, however, indicate a much more complex pattern of receptor effects than that reported with bath-application of adrenergic blockers. In their study iontophoresis of the Pblocker, sotalol, at the cell body layer produced significant potentiation of the population spike. We had also observed this effect during our own iontophoretic studies (unpublished observations). In general a agonists, in their study, produced population spike increases and P-agonists produced decreases. This is the opposite of what had been reported with NE agonists and antagonists in the slice, and makes it clear that it will be necessary to evaluate the effects of synaptic NE release in order to understand the working system. NE, itself, only produced an effect at the cell body layer if iontophoresed for 5 min. The change seen, as already mentioned, was a long-lasting potentiation of the population spike. In all cases any effects on the population spike of NE ago-
312
nists and antagonists were only seen in a minority of experiments. This is reminiscent of Lacaille and Schwartzkroin’s (1988) experiments with microdrop NE application in the slice, where less than half of the NE applications produced changes. Such variability is generally ascribed to difficulties in appropriate localization of the applied substance. EPSP slope changes were common in the Winson and Dahl(1985) experiments, and these varied as a function of iontophoretic location. EPSP slope and amplitude were decreased if NE or its agonists and antagonists were iontophoresed in the mid-third of the dendritic region. The population spike was usually decreased by applications at that level. EPSP slope was increased if NE was iontophoresed directly above the cell body layer, but the population spike was still decreased. No consistent EPSP changes were seen with NE iontophoresed at the cell body layer (Winson and Dahl, 1985). Again, the question of interest is: which of these “localized” effects embodies the actual effect of NE release?
Selectivity of NE effects in dentate gyms A recent report of NE effects in the slice by Dahl and Sarvey (1989) is of particular interest when considering functional effects of NE release. They provide evidence that the potentiating action of NE is only seen for the evoked potential produced by stimulation of the medial perforant path, the lateral perforant path-evoked potential is significantly depressed by NE application. Both effects are produced by the P-agonist, isoproterenol and both can be blocked by propranolol or by an NMDA blocker. Given the diffuse nature of NE innervation in the hippocampus it is of some interest that NE modulation is so selective as to differentially effect glutamate synapses on the same dendritic tree. There is also other evidence that the lateral and medial pathways can be independently modulated. Naloxone has been reported to block LTP only on the lateral and not on the medial perforant path (Bramham et al., 1988). Most of the work on NE-induced
potentiation of the perforant path has been done on the medial perforant path potential. The observation of selective potentiating and depressant effects of NE on the two perforant path inputs should be investigated in viuo. Evoked potential potentiation by LC activation
We have been particularly concerned with whether or not release of NE by LC terminals in the dentate gyrus would produce the same short and LTP effects as had been seen with direct application of NE in this area.
Electrical stimulation of LC Early short reports of the effects of LC electrical stimulation suggested that, while single pulse LC stimulation had little discernable effect on the perforant path-evoked potential, trains of pulses were effective in producing potentiation of the dentate gyrus potential. In one study both the pre- and postsynaptic components of the evoked potential were enhanced (Bliss and Wendlandt, 1977); in the other study, in which recordings were made at the dendritic and the cell body layer, only the population spike, not the EPSP slope, was increased by LC stimulation (Assaf et al., 1979). Dahl and Winson provided the first full-length report of the effects of electrical stimulation in the LC area on the dentate gyrus-evoked potential (Dahl and Winson, 1985). In an earlier exploration of the brainstem with 1000 Hz triple pulses Winson had not observed any effects on the dentate gyrus potential with 200 p A 100 pS pulses in the LC region (Winson, 1980). In the 1985 study using a train of 6-12 50 Hz pulses which ended 50 msec before the perforant path pulse, Dahl and Winson found a potentiation of the population spike recorded at the cell body layer and a decrease of the EPSP recorded at the dendritic layer. The field EPSP measured at the cell layer was unaffected, but the field EPSP measured in the dendritic layer was consistently decreased. Dahl and Winson suggest that this indicates NE
313
.-. --7
165
2
a
spike amplitude spike onset nepsp slope
N.10 1L5
135-
E 125115-
5
0
2 165
{ 1351 E 125,
% 105:
lOmin
20min
30mln
LOmin
Time from LC stirnulotion onset
,r 7 1
: g 5 1 -6 20 .-"
75.1 Lton
p--...-~.e-
t
f -?--k - - y--j ~
1
LO 60 80 LC-PP IS1 i n milliseconds
Fig. 4. The lower graph illustrates potentiation of the population spike amplitude produced by a 15 msec burst of LC 333 H z stimulation at varying LC-PP interstimulus intervals (ISIS) from 20-80 msec ( n = 7). Note the absence of a potentiating effect on EPSP slope and population spike latency. The upper graph shows the pattern of LTP of the population spike amplitude in ten experiments in which 50 pairs of LC-PP stimulation were given using a 40 msec ISI. The baseline potential was based on the preceding 10 min of 0.1 Hz perforant path-evoked responses. Note the apparent decrease in EPSP slope during LC-PP pairing. Mean and standard errors for both sets of data are shown. (From Harley et al., 1989.)
may radically alter source and sink distribution in the granule cells. More than six pairings of LC and PP stimulation resulted in an attenuation of LC potentiation effects. Identification of LC stimulation effects was related to histological placement within 300 p M of the LC. One thousand (1000) Hz trains were again ineffective in producing these changes. We found that either the 15 msec 333 Hz LC train which had produced population spike potentiation in the study of Assaf et al. (1979) or the 10 Hz train employed by Bliss and Wendlandt (19771, was effective in producing a significant potentiation of the perforant path-evoked potential (Harley et al., 1989). As reported by Assaf et al. (1979) the optimal interval for maximum potentiation with the 5 pulse train required initiating stimulation of the LC 40 msec prior to activation of the perforant path. See the lower panel in Figure 4. This interval was consistent with a reported antidromic conduction time from the hip-
pocampus to the LC of 20-70 msec (Nakamura and Iwama, 1975). It should be mentioned as a caveat that a currently unidentified system running near the median raphe also produces dentate gyrus population spike potentiation with similar raphe-perforant path stimulation intervals (Assaf and Miller, 1978). This system has been shown to suppress granule cell firing during the period of enhancement. Not only did the short-term potentiation of the population spike match what we had seen with direct applications of N E but, of equal interest, it was also possible to produce LTP by repeatedly pairing electrical stimulation of the LC and activation of the perforant path. See the upper panel of Figure 4. In 60% of the experiments there was a significant decrease in EPSP slope, while in the remaining 40% there was either no change or an increase. Surprisingly, contralateral "LC" stimulation also appeared to produce population spike potentiation at a similar interval. Difficulties arose when we attempted to block the effects of electrical stimulation of the LC with the /?-blocker, propranolol. N o attenuation of the potentiation occurred (Fig. 5). Our difficulty was reminiscent of the earlier story of the rewarding effects of electrical stimulation in the vicinity of the LC. While rewarding effects could be shown to be rather well localized to the vicinity of the nucleus, and while electrodes supporting such
6or al
U 3
LC electrical
130
'r: -
a
6
120
2-
110
ln
100
a
90
Before
20 min
40 min
6 0 m i n after
I
propmnolol
Fig. 5 . The lack of effect of propranolol (20-30 mg/kg ip), a 0-receptor blocker, on the immediate potentiation effect produced by LC electrical stimulation 40 msec prior to perforant path activation. (From Harley et a l , 1989 )
314
I4O
tI
selective enough to allow us to study LC activation per se. We decided to change tactics and use micropipette ejections of glutamate to stimulate the nucleus more selectively.
N.20
Glutamate activation of locus coeruleus Glutamate activation of LC produces a clear potentiation of the perforant path-evoked population spike (Fig. 6; Harley and Milway, 1986). In some experiments the potentiation is long-lasting (Fig. 7). The effects of glutamate can be blocked by systemic propranolol (20-30 mg/kg) or by timolol or propranolol delivered by cannulae in the dentate gyrus (Harley and Evans, 1988). A direct comparison of glutamate ejection in the LC alternating with electrical stimulation from an electrode placed within 100 p m of the pipette tip can be seen in Figure 8. Both manipulations produce potentiation of the perforant path population spike, but only the glutamate activation effects are blocked by subsequent propranolol administration. It is, no doubt, possible to induce NE release by electrical stimulation in the vicinity of the LC nucleus. NE release in the forebrain has been demonstrated following periods of electrical stimulation in the vicinity of the LC (Tanaka et al.,
I II
GLU
90 -5
I
I
0
5
I
10
Time (rnin)
Fig. 6. The effects of 100-150 nl 0.5 M glutamate ejected in the LC on the perforant path-evoked population spike in the dentate gyrus averaged over 20 animals. (From Harley and Milway, 1986.)
effects had been shown, in pioneering studies by Segal (Segal and Bloom, 1976a,b) to affect cell activity in the hippocampus, it has not been possible in subsequent work to demonstrate their dependence on the functioning of LC (Clavier et al., 1976; Corbett et al., 1977; Corbett and Wise, 1979). We were concerned that electrical stimulation in the fiber-rich LC area was simply not Mean Control Values
e
Population spike EPSP slope
amplitude 6 . 7 m v 3.2 mv/ms
Pop spike onset
latency
T M 18
3.4ms
.
.. . . . . -.. ... . .. 9
u o"\
120
80-
GLU
t
.
.
-*
.*
*
.
.. ..
a
.
5 mins
Fig. 7. An experiment in which glutamate ejection in the LC initiated a long-term increase in both population spike amplitude and EPSP slope. Note the apparent similarity to the time course of potentiation development seen in Figure 2. (From Harley and Evans, 1988.)
315 I LC Glutamate
eZa LC Electrical
1601
3
(4)-
140
0 c -
130
Q
E, 0
Y
'20 110
Q
Ln
100
90
Before 20min propranolol
40rnin
60 rnin a f t e r
Fig. 8. Effective propranolol (20-30 mg/kg iv) attenuation of the potentiation of population spike amplitude elicited by glutamate activation of the LC. At the same time potentiation produced by LC electrical stimulation was not attentuated by propranolol. Each animal was tested using a coupled electrode and micropipette assembly in the LC. Numbers in parentheses indicate animals used in each average. (From Harley et al., 1989.)
1976). Furthermore Washburn and Moises (1989) have recently reported, using the stimulation parameters of Assaf et al., (1979) that potentiation of the dentate gyrus population spike was blockable by systemic propranolol and could also be blocked by the a,-agonist, clonidine. These are encouraging results. We would prefer to use electrical stimulation to produce NE release in the dentate gyrus as it is a much easier methodology for the chronic studies which would enable us to investigate the time course of NE-induced long-term changes in the behaving animal. We have had some success in obtaining potentiation in awake animals with glutamate ejection near the LC via cannulae, but the technique is awkward and unsatisfactory for long-term study. In the Washburn and Moises (1989) study, they were careful to use LC currents which only produced half-maximal potentiation of the population spike. It is possible that our stimulation was too effective and that the NE release evoked competed successfully for binding sites with propranolol. Alternatively, our stimulation may have released peptides as well as NE, and the peptide release itself was adequate to produce potentiation. It has been shown that low-frequency stimu-
lation of monoamine pathways preferentially promotes monoamine release while high-frequency stimulation induces the additional release of peptides which often have a similar effect on the target cell (Lundberg et al., 1986). Only the monoamine effects are reduced by postsynaptic monoamine blockers. An important aspect of the Washburn and Moises study was their care in evaluating EPSP slope changes in the dendrites. They found no significant change in EPSP slope either in the dendrites or at the cell body layer. The issue of EPSP increases may be an important one for the hypothesis that NE promotes LTP effects, since increases in EPSP slope are generally considered a hallmark of LTP. While EPSP slope increases may be demonstrated reliably in the slice it is not clear that this is so with LC activation. In the two in vivo studies which have taken care to monitor dendritic EPSP slope changes, the site at which consistent EPSP increases have been reported in the slice, neither has found reliable increases in EPSP slope. It would appear that LC activation most frequently potentiates the coupling of the EPSP and population spike rather than potentiating the EPSP slope itself. Such enhanced population spike effects are also seen with frequency-induced LTP (Bliss and Gardner-Medwin, 1973; Taube and Schwartzkroin, 1988). They do not rule out the hypothesis that NE promotes LTP, but suggest that for some reason LC release of NE in the dentate favors increases in E-S coupling rather than in EPSP slope. This difference between LC effects in v i m and NE effects in the slice makes it particularly important to evaluate NMDA blockade effects on LC-induced potentiation. As another approach to a more selective LC stimulation paradigm we have taken advantage of the work (Ennis and Aston-Jones, 1986) showing excitatory inputs to the LC from the paragigantocellularis (PGi) region of the medulla. We have found that 2-4 0.5 msec pulses of 200 to 800 FA at 333 Hz in the PGi produce a clear potentiation of the dentate gyrus population spike at 30-40
316
PGI-PP spikes
PP spikes
Fig. 9.. The potentiating effect of two 0.5 msec pulses of paragigantocellularis stimulation at 800 F A and 333 Hz 30 msec prior to a perforant path (PP)pulse. This potentiating effect was completely blocked by 4 pg/kg clonidine, an a*blocker. Recovery of the stimulation effect occurred about 30 min after the injection of clonidine.
msec latencies (Fig. 91. Longer intervals are not effective. In the experiment shown, 4 pg/kg iv of clonidine blocked the PGi-induced potentiation (Babstock and Harley, unpublished observations). The effective latency range and the clonidine results are consistent with the hypothesis that the potentiating effects we observed are being mediated via LC activation and NE release in the dentate gyrus. Long-lasting effects of repeated PGI stimulation have also been observed. Locus coeruleus unit activity and dentate gyrus-evoked potential modulation
In other recent experiments on the relationship between LC activity and dentate gyrus-evoked potential modulation, Susan Sara and I have recorded from LC cells, using a variety of methods to increase or decrease LC firing, while concomitantly monitoring the perforant path-evoked potential (Harley and Sara, 1990). We found that silencing the LC with iv injections of clonidine had no observable effect on the dentate gyrusevoked potential. This was also reported by Washburn and Moises (1989) although they did not directly evaluate LC silence. Tail pinch, while briefly increasing LC activity, typically did not
modulate the evoked potential, however intravenous catheterization of the penile vein was associated at times with increases in population spike amplitude (see Fig. 11). This manipulation might be expected to produce more potent changes in LC activity. Injections of 50 nM glutamate in the LC characteristically produced a rush of firing followed by a loss of cell recording due, we have assumed, to depolarization block as has been described previously (Adams and Foote, 1988). In the experiment illustrated in Figure 11, cells appeared to continue to fire at elevated rates after the initial rush of activity; however, the height of the recorded waveforms diminished below the spike window of the ratemeter and the activity was not counted. The waveform shape appeared unchanged. Typically the rate of single unit firing after glutamate went from 1.2 Hz to 6-10 Hz for 10-20 sec and then we were unable to count the cell or cells accurately. Within 5-10 min after each glutamate ejection in Figure 10, we found nearby LC cells that had normal rates of activity. The oscillograph photographs (Fig. 10) show that the cells recorded prior to the time of glutamate ejection all met the criterion of a stereotyped excitation and inhibition pattern following tail pinch. It can also be seen that this particular LC recording displayed a rhythmic character prior to the second glutamate ejection which was also seen in several other experiments. The second glutamate ejection disrupted the rhythmicity but the rhythmicity returned within 5 min of the ejection. The pattern of acute excitement followed in a matter of minutes by a return to basal activity patterns was observed in all the effective glutamate experiments. These data suggest that the initial phasic increases in the dentate gyrusevoked potential depend on an abrupt and substantial release of NE in the hippocampus. The subsequent long-lasting increases in the population spike, once triggered, continue in the presence of basal LC activity, just as such increases can be observed following the washout of NE from the slice preparation. In general, in the
t
A
Perforant path potential each 10 SecS
~-
___
Fig. 10. Oscilloscope tracings of multiunit LC baseline activity recorded during the experiment of Figure 10. The lower trace illustrates the characteristic response to tail pinch seen with LC units. The upward line between the excitatory and inhibitory phases of the pinch response is an artifact. The upper trace is a recording of rhythmic activity observed at the same site. This activity was disrupted by glutamate ejections but returned within 5 min of the ejection.
glutamate experiments we were able to trigger both a short-term potentiation by increasing LC cell firing and a characteristic LTP which did not appear to depend on continued LC activity. The EPSP slope was increased in parallel in some, but not all. of these cases. Implications for hippocampal functioning: NE facilitation of synaptic plasticity
What is the relation of LC modulation of the dentate gyrus-evoked potential to the real world functioning of the hippocampus? Cellular recording evidence strongly supports the hypothesis that
7
miorant pain poieni~aieach 1osecs
IJpop"la"o~" splke=
Lc-cirachvlt;
. .-
Fig. 11. The effects of 50 flM ejections of glutamate in the LC o n the perforant path-evoked population spike and on multiunit cell firing in the LC recorded simultaneously. During the period shown in the top panel two glutamate ejections were made. Each evoked an increase in L C cell firing, as seen in the lower panels, which immediately preceded a significant increase in population spike amplitude. A third phasic increase in population spike amplitude was noted during an iv catherization procedure.
the hippocampus generates a cellular spatial map (O'Keefe and Nadel, 1978). Hippocampal cells have been shown to respond selectively to specific places in the environment. They apparently become coupled to a set of distal environmental cues such that any subset of those cues is effective in activating the map. The environmental cue information reaches the hippocampus by way of
318
the perforant path connections from the entorhinal cortex. McNaughton and Morris (1987) have proposed a simplified model of the operation of the spatial hippocampal network based on the assumption that the synaptic inputs carrying environmental information become coupled to a given set of hippocampal cells through a process of Hebbian association. Hippocampal neurons encode environmental information from the entorhinal cortex in a matrix of weak and strong synaptic connections. Given partial or complete environmental input into such a matrix, the appropriate cell firing pattern is generated. The synaptic weights are determined by the coupling of inputs with a set of cell activations from detonator synapses which provide the strong depolarization condition necessary for NMDAinduced long-term plasticity. The process of Hebbian association then generates increases in the synaptic weights of the concurrently active environmental inputs. The detonator concept seems unlikely since it presumes hardwired templates of cells which fire in a given environment. I would suggest that, in lieu of a set of prewired detonators, the hippocampus receives sets of facilitatory inputs which when active, singly or in concert, can increase the likelihood of strengthening a set of synaptic inputs activated by a particular environmental configuration. The representation of that input set then becomes, as postulated, fixed in the hippocampal matrix. The LC would function as one of those facilitatory systems, but there appear to be multiple facilitating inputs. Any diffuse input which, for example, reduced the probability of inhibition would increase the likelihood of NMDA-mediated changes in synaptic weights. It has been shown that other diffuse systems such as the massive supramammillary input to the dentate gyrus are also capable of potentiating perforant path input (Mizumori et al., 1989). It is not yet certain, at the cellular level, by what single or multiple mechanisms the LC promotes long-lasting modification of the hippocampal network.
Reduction of inhibition appears unlikely given the reports of increased activation of theta units by NE; on the other hand, as discussed earlier, NE does produce an increase in glutamate release, an increase in Ca’+currents, a reduction in Ca++-mediated K+currents and depolarization, all of which are cellular changes consistent with a greater probability of dentate response to input. While I have focused on the evidence for a role of the LC in promoting long-lasting potentiation of the glutamatergic perforant path input to the dentate gyrus, it appears from our work that a short-lasting potentiation, as in other areas of the central nervous system, is more likely with brief LC activation. It seems fitting that a mechanism which can produce transient potentiation of cellular responses to a putative sensory input should also, on an apparent continuum of degree or duration of activation, be linked to an increased future response to that input. The short term effect would presumably increase the intensity of experience of or promote attention to the input while the long-term effect could be viewed as the promotion of memory. The two processes of attention and memory have long been functionally linked in the psychological literature. It now appears that they are linked at the mechanistic level as well. These ideas were discussed two decades ago by Kety (Harley, 1987). The picture of LC function which emerges from the study of the dentate gyrusevoked potential supports the hypothesis that the LC is part of a system which increases the likelihood of the physiological changes which underly the functions “noticing” and “remembering.” It does not, however, argue that these events will not happen in the absence of the LC since even in current models there are diverse ways to enhance the functioning of glutamatergic synapses. Increases in LC activity appear to occur upon introduction into novel environments in particular (see Aston-Jones et al., Foote et al., and Sara and Segal, this volume) and might underly the enduringly enhanced population spike reported in novel, enriched environments by Sharp et al.
319
(1985) or the enduringly enhanced population spike observed during discriminative operant conditioning (Skelton el aL, 1987). Given only a facilitator role, it is probable that relatively subtle and specific behavioral assessment will be necessary to dissect the contribution of LC modulation to normal hippocampal functioning. Nonetheless this postulated role is consistent with the observation that loss of multiple facilitator systems, including the LC, as appears to occur in Alzheimer's disease, is devastating for memory function (McGeer, 1984). Finally, while I have suggested that diffuse systems may act, in general, as facilitators of long-term changes in the hippocampal network, it is increasingly clear that even in that structure the targets of NE action may be quite specific. Thus, for example, in the slice it appears that N E s effects on dentate gyrus circuitry may be restricted to potentiation of only the medial perforant path. Anatomical data suggest that the medial perforant path carries visual and auditory information from the environment while the lateral perforant path carries primarily olfactory input (Swanson et aZ., 1987). Interestingly, after ablation of the fornix which significantly disrupts subcortical inputs to the dorsal hippocampus, including those of the LC, place cell responding is dominated by local olfactory cues rather than distal environmental cues (Shapiro et al., 1989). Spatial behavior appears much less flexible in such animals. It is possible that NE acts to enhance distal environmental cue coupling to the hippocampal network while reduced NE activity, for example, may favor olfactory cue coupling. Interestingly, in that regard, blockers that interfere with noradrenergic and cholinergic inputs in the dentate, and which might now be expected to attenuate LTP processes in the medial perforant pathway, interfere with spatial learning (Gill et al., 1989); while blockers which interfere with the opiate system that promotes LTP in the lateral perforant pathway promote spatial learning (Decker et al., 1989). Do granule neurons function as members of two different and, possibly,
conflicting networks with participation in each being, in part, gated by LC input? The LC system, which has generally been regarded as diffuse and nonspecific, may have a selective and distinct role in promoting certain forms of coding in the hippocampal network while attenuating others. It is unlikely that multiple, diffuse systems will play synonymous roles in all forms of attentional and memorial processes, Acknowledgements
This research was supported by the Natural Sciences and Engineering Research Council of Canada and by the French Ministry of Research and Technology. References Adams, L.M. and Foote, S.L. (1988) Effects of locally infused pharmacological agents on spontaneous and sensoryevoked activity of locus coeruleus neurons. Bruin Res. Bull., 21: 395-400. Assaf, S.Y. and Miller, J.J. (1978) Neuronal transmission in the dentate gyrus: Role of inhibitory mechanisms. Brain Rex, 151: 587-592. Assaf, S.Y., Mason, S.T. and Miller, J.J. (1979) Noradrenergic modulation of neuronal transmission between the entorhinal cortex and the dentate gyrus of the rat. J. Physiol. (London), 292: 52P. Bliss, T.V.P. and Gardner-Medwin, A.R. (1973) Long-lasting potentiation of synaptic transmission in the dentate area of the unanaesthetized rabbit following stimulation of the perforant path. J. Physiol. (London), 232: 357-374. Bliss, T.V.P and L m o , T. (1973) Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J. Physiol. (London),232: 331-356. Bliss, T.V.P. and Wendlandt, S. (1977) Effects of stimulation of locus coeruleus on synaptic transmission in the hippocampus. Proc. Znt. Union Physiol. Sci., 13: 81. Bliss, T.V.P., Goddard, G.V. and Riives, M. (1983) Reduction of long-term potentiation in the dentate gyrus of the rat following selective depletion of monoamines. J. Physiol. (London), 334: 475-491. Bramham, C.R., Errington, M.L. and Bliss, T.V. (1988) Naloxone blocks the induction of long-term potentiation in the lateral but not in the medial perforant pathway in the anesthetized rat. Brain Res., 449: 352-356. Burgard, E.C., Decker, G. and Sarvey, J.M. (1989) NMDA receptor antagonists block norepinephrine-induced longlasting potentiation and long-term potentiation in rat dentate gyrus. Bruin Res., 482: 351-355.
320 Clavier, R.M., Fibiger, H.C. and Phillips, A.G. (1976) Evidence that self-stimulation of the region of the locus coeruleus does not depend upon noradrenergic projections to telencephalon. Brain Res., 113: 71-82. Corbett, D. and Wise, R.A. (1979) Intracranial self-stimulation in relation to the ascending noradrenergic fiber system of the pontine tegmentum and caudal midbrain: A moveable electrode mapping study. Brain Res., 177: 423426. Corbett, D., Skelton, R.W. and Wise, R.A. (1977) Dorsal noradrenergic bundle lesions fail to disrupt self-stimulation from the region of the locus coeruleus. Brain Res., 137: 37-44. Dahl, D. and Sarvey, J.M. (1989) Norepinephrine induces pathway-specific long-lasting potentiation and depression in the hippocampal dentate gyrus. Proc. Natl. Acad. Sci. USA, 86: 4776-4780. Dahl, D. and Sarvey, J.M. (1990) P-Adrenergic agonist-induced long-lasting synaptic modifications in hippocampal dentate gyrus require activation of NMDA receptors, but not electrical activation of afferents. Brain Res. (In press). Dahl, D. and Winson, J. (1985) Action of norepinephrine in the dentate gyrus. I. Stimulation of locus coeruleus. Exp. Brain Res., 59: 491-496. Decker, M.W., Introini-Collison, I.B. and McGaugh, J.L. (1989) Effects of naloxone on Morris water maze learning in th? rat: Enhanced acquisition with pretraining but not post-training administration. Psychobiology, 17: 270-275. Ennis, M. and Aston-Jones, G. (1986) A potent excitatory input to the nucleus locus coeruleus from the ventrolateral medulla. Neurosci. Lett., 71: 299-305. Gill, T.M., Decker, M.W. and McGaugh, J.L. (1989) Concurrent muscarinic and beta-adrenergic blockade impairs inhibitory (passive) avoidance and Morris water maze performance. SOC.Neurosci. Abstr., 15: 733. Gray, R. and Johnston, D. (1987) Noradrenaline and betaadrenoceptor agonists increase activity of voltage-dependent calcium channels in hippocampal neurons. Nature (London), 327: 620-622. Haas, H.L. and Rose, G.M. (1987) Noradrenaline blocks potassium conductance in rat dentate granule cells in uitro. Neurosci. Lett., 78: 171-174. Harley, C.W. (1987) A role for norepinephrine in arousal, emotion and learning? Limbic modulation by norepinephrine and the Kety hypothesis. Prog. Neuro-Psychopharmacol. Biol. Psychiatty., 11: 419-458. Harley, C.W. and Evans, S. (1988) Locus-coeruleus-induced enhancement of the perforant-path evoked potential: Effects of intradentate beta blockers. In C.D. Woody, D.L. Alkon and J.L. McGaugh (Eds.), Cellular Mechanisms of Conditioning and Behavioral Plasticity, Plenum, New York, pp. 415-423. Harley, C.W. and Milway, J.S. (1986) Glutamate ejection in the locus coeruleus enhances the perforant path-evoked population spike in the dentate gyrus. Exp. Brain Res., 63: 143-150. Harley, C.W. and Sara, S.J. (1990) Locus coeruleus cell activ-
ity and potentiation of the perforant path evoked dentate gyrus population spike. Sac. Neurosci. Abstr., 16: 264 Harley, C., Milway, J.S. and Lacaille, J.C. (1989) Locus coeruleus potentiation of dentate gyrus responses: Evidence for two systems. Brain Res. Bull., 22: 643-650. Hopkins, W.F. and Johnston, D. (1984) Frequency-dependent noradrenergic modulation of long-term potentiation in the hippocampus. Science, 226: 350-352. Johnston, D., Hopkins, W.F. and Gray, R. (1988) Noradrenergic enhancement of long-term synaptic potentiation. In P.W. Landfield and S.A. Deadwyler (Eds.) Long-Term Potentiation: From Biophysics to Behauior. Alan R. Liss, New York, pp. 355-376. Lacaille, J-C. and Harley, C.W. (1985) The action of norepinephrine in the dentate gyrus: Beta-mediated facilitation of evoked potentials in uitro. Brain Res., 358: 210-220. Lacaille, J-C. and Schwartzkroin, P.A. (1988) Intracellular responses of rat hippocampal granule cells in iitro to discrete application of norepinephrine. Neurosci. Lett., 89: 176-181. Lundberg, J.M., Rudehill, A,, Sollevi, A,, TheodorssonNorheim, E. and Hamberger, B. (1986) Frequency-dependent and reserpine-dependent chemical coding of sympathetic neurotransmission-Differential release of noradrenaline and neuropeptide Y from pig spleen. Neurosci. Lett., 109: 341-348. Lynch, M.A. and Bliss, T.V.P. (1986) Noradrenaline modulates the release of [14C] glutamate from dentate but not from CAl/CA3 slices of rat hippocampus. Neuropharmacology, 25: 171-176. Lynch, M.A., Errington, M.L. and Bliss, T.V.P. (1985) Longterm potentiation of synaptic transmission in t h e dentate gryus: Increased release of [14C]glutamate without increase in receptor binding. Neurosci. Lett., 62: 123-129. Madison, D.V. and Nicoll, R.A. (1982) Noradrenaline blocks accomodation of pyramidal cell discharges in the hippocampus. Nature (London), 299: 636-638. McGeer, P.L. (1984) The 12th J.A.F. Stevenson Memorial Lecture: Aging, Alzheimer’s disease, and the cholinergic system. Can. J. Physiol. Pharmacol., 62: 741-754. McNaughton, B.L. and Morris, R.G.M. (1987) Hippocampal synaptic enhancement and information storage within a distributed memory system. TINS, 10: 408-415. Mizumori, S.J.Y., McNaughton, B.L. and Barnes, C.A. (1989) A comparison of supramammillary and medial septa1 influences on hippocampal field potentials and single unit activity. J. Neurophysiol., 61: 1-17. Nakamura, S. and Iwama, K. (1975) Antidromic activation of the rat locus coeruleus neurons from hippocampus, cerebral and cerebellar cortices. Brain Rex, 99: 372-376. Neuman, R.S. and Harley, C.W. (1983) Long-lasting potentiation of the dentate gyrus population spike by norepinephrine. Brain Res., 273: 162-165. O’Keefe, J. and Nadel, L. (1978) The Hippocampus as a Cognitive Map, Clarendon Press, Oxford, 570 pp. Radhakrishnan, V. and Albuquerque, E.X. (1989) Effect of beta adrenergic blockers on N-methyl-D-aspartate
321 (NMDAhctivated channels of rat hippocampal neurons. Soc. Neurosci. Abstr., 15: 828. Robinson, G.B. and Racine, R.J. (1985) Long-term potentiation in the dentate-gyrus: Effects of noradrenaline depletion in the awake rat. Brain Res., 325: 71-78. Rose, G.M. and Pang, K.C.H. (1989) Differential effect of norepinephrine upon granule cells and interneurons in the dentate gyrus. Brain Res., 488: 353-356. Sarvey, J.M., Burgard, E.C. and Decker, G. (1989) Long-term potentiation: Studies in the hippocampal slice. J. Neurosci. Methods, 28: 109-124. Segal, M. and Bloom, F.E. (1976a) The action of norepinephrine in the rat hippocampus. 111. Hippocampal cellular response to locus coeruleus stimulation in the awake rat. Brain Rex, 107: 499-511. Segal, M. and Bloom, F.E. (1976b) The action of norepinephrine in the rat hippocampus. IV. The effects of locus coeruleus stimulation on evoked hippocampal unit activity. Brain Res., 107: 513-525. Shapiro, M.L., Simon, D.K., Olton, D.S., Gage, F.H., 111, Nilsson, 0. and Bjorklund, A. (1989) Intrahippocampal grafts of fetal basal forebrain tissue alter place fields in the hippocampus of rats with fimbria-fornix lesions. Neuroscience, 32: 1- 18. Sharp, P.E., McNaughton, B.L. and Barnes, C.A. (1985) Enhancement of hippocampal field potentials in rats exposed to a novel, complex environment. Brain Rex, 339: 361-365. Skelton, R.W., Scarth, A.S., Wilkie, D.M., Miller, J.J. and Philips, A.G. (1987) Long-term increases in dentate granule cell responsivity accompany operant conditioning. J. Neurosci., 7: 3081-3087. Stanton, P.K. and Heinemann, U. (1986) Norepinephrine enhances stimulus-evoked calcium and potassium concentration changes in dentate granule cell layer. Neurosci. Lett., 67: 233-238. Stanton. P.K. and Sarvey, J.M. (1984) Blockade of long-term potentiation in rat hippocampal CA1 region by inhibitors of protein synthesis. J. Neurosci., 4: 3080-3088. Stanton, P.K. and Sarvey, J.M. (1985a) Depletion of norepinephrine, but not serotonin, reduces long-term potentiation in the dentate gyrus of rat hippocampal slices. J. Neurosci., 8: 2169-2176. Stanton, P.K. and Sarvey, J.M. (1985b) The effect of highfrequency electrical stimulation and norepinephrine on
cyclic AMP levels in normal versus norepinephrine-depleted rat hippocampal slices. Brain Res., 358: 343-348. Stanton, P.K. and Sarvey, J.M. (1985~)Blockade of norepinephrine-induced long-lasting potentiation in the hippocampal dentate gyrus by an inhibitor of protein synthesis. Brain Res., 361: 276-283. Stanton, P.K. and Sarvey, J.M. (1987) Norepinephrine regulates long-term potentiation of both the population spike and dendritic EPSP in hippocampal dentate gyrus. Brain Res. Bull., 18: 115-119. Stanton, P.K., Mody, I. and Heinemann, U. (1989) A role for N-methyl-D-aspartate receptors in norepinephrine-induced long-lasting potentiation in the dentate gyrus. Exp. Brain Res., 77: 517-530. Swanson, L.W., Kohler, C. and Bjorklund, A. (1987) The limbic region. I. The septohippocampal system. In A. Bjorklund, T. Hokfelt and L.W. Swanson (Eds.), Handbook of Chemical Neuroanatomy, Vol. 5, Integrated systems of the CNS, Part I , Elsevier, Amsterdam, pp. 125-227. Tanaka, C., Inagaki, C. and Fujiwara, H. (1976) Labeled noradrenaline release from rat cortex following electrical stimulation of locus coeruleus. Brain Res., 106: 384-389. Taube, J.S. and Schwartzkroin, P.A. (1988) Mechanisms of long-term potentiation: EPSP/spike dissociation, intradendritic recordings, and glutamate sensitivity. J. Neurosci., 8: 1632-1644. Washburn, M. and Moises, H.C. (1989) Electrophysiological correlates of presynaptic alpha 2-receptor-mediated inhibition of norepinephrine release at locus coeruleus synapses in dentate gyrus. J. Neurosci., 9: 2131-2140. Wigstrom, H. and Gustafsson, B. (1983) Large long-lasting potentiation in the dentate gyrus in uitro during blockade of inhibition. Brain Res., 275: 153-158. Wigstrom, H. and Gustafsson, B. (1985) On a long-lasting potentiation in the hippocampus: A proposed mechanism for its dependence on coincident pre- and postsynaptic activity. Acta Physiol. Scand., 123: 519-522. Winson, J. (1980) Influence of raphe nuclei on neuronal transmission from perforant path through dentate gyrus. J. Neurophysiol., 44: 937-950. Winson, J. and Dahl, D. (1985) Action of norepinephrine in the dentate gyrus. 11. Iontophoretic studies. Exp. Brain R ~ s .59: , 497-506.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Pronrcss w Rrmn Research, Vol. 88 0 1991 tlscvier Science Publishers B.V.
323 CHAPTER 24
Actions of norepinephrine in the rat hippocampus M. Segal, H. Markram and G. Richter-Levin Center for Neurosciences, Weizrnann Institute, Rehoilot, Israel
Acting at postsynaptic a l - and /?-receptors, norepinephrine (NE) exerts a complex action in rat hippocampus. It is currently believed that p,-receptor activation enhances excitability of recorded neurons, whereas a , activation suppresses reactivity to afferent stimulation. These reported effects of n-agonists are not consistent with a I effects found elsewhere in the brain. We have conducted experiments in the anesthetized rat and found that a n amphetamine-induced
increase in the dentate gyrus population spike can be blocked by a p-antagonist but also by an a,-antagonist. We have conducted experiments in the brain slide preparation and found that an a-agonist, phenylephrine (PHE), selectively enhances responses to N-methyl-D-aspartate (NMDA) but not to quisqualate. We propose that the product of activation of both a- and p-receptor types will enhance reactivity of hippocampal cells to afferent stimulation.
Key words: hippocampus, a-receptor, norepinephrine, dentate gyrus
Introduction
The hippocampus is one of the major targets for the efferents of the nucleus locus coeruleus (LC) (for review see Foote et al., 1983). The association of the hippocampus with cognitive functions of the brain and the assumed involvement of the LC in these higher brain functions (Sara et al., 1988) made the hippocampus a preferred target for analysis of norepinephrine's (NE) action in higher brain structures (Segal and Bloom, 1974, 1976). Indeed, numerous studies in the past decade focussed on the physiological effects of NE in the hippocampus (Table 1); many utilized the in i'itro brain slice preparation and described an array of actions of a number of recently developed noradrenergic ligands in the hippocampus.
A parallel advance in the understanding of second messenger involvement in the action of catecholamines (Exton, 1988) also contributed to understanding the roles of NE in hippocampal functioning. Norepinephrine in the hippocampus: P-adrenergic actions
While the P-receptors constitute a minority of the total NE receptors in the hippocampus (Rainbow et al., 1984) they were studied extensively in various preparations. In fact, most of the reported effects of NE in the hippocampus are mediated by a P-receptor; among these are facilitation of population spike response of the dentate gyrus to perforant path stimulation in the
324
slice as well as the intact brain preparation (Table l), an increase in the spontaneous activity of theta cells (putative interneurons? (Pang and Rose, 1987)), a blockade of the slow, Ca-dependent K current found in pyramidal and granule neurons of the hippocampus and elsewhere (Madison and Nicoll, 1982; Haas and Konnerth, 1983) a facilitation of reactivity of CA3 cells to mossy fiber stimulation (Hopkins and Johnson, 1988), a phenomenon resembling long-term potentiation studied in these and other hippocampal neurons, and an increase in efficacy of perforant path stimulation in producing a rise in [K], and a reduction in [Ca], (Stanton and Heinemann, 1986). Thus, there is general agreement that a P-adrenergic effect is facilitatory to reactivity of the hippocampus to afferent stimula-
tion. This action of enhancing reactivity to afferent stimulation is likely to involve a change in postsynaptic sensitivity to the neurotransmitter substance, as Segal (1982) has shown a P-mediated enhanced response to topical application of glutamate. Whenever tested in the hippocampus, the p-adrenergic effect is of the @,-type (Madison and Nicoll, 1986; Fowler and O'Donnel, 1988). Only few results are not consistent with this consensus: in one, done in anesthetized rats, Winson and Dahl (1985) found a P-adrenergic reduction in the response of dentate gyrus to medial perforant path stimulation, whereas in another study Dahl and Sarvey (1989) found a P-adrenergic increase in reactivity of the dentate gyrus, recorded in a slice, to stimulation of the medial perforant path, but a decrease in the
TABLE 1 Physiology of norepinephrine in the hippocampus Preparation
Region CA1 DG DG DG CA3/1 CA1 CA1,3 CA 1 CA1 CA3 CA1-3 CA1 DG DG DG CA1 DG CA1 DG CAI CA1 CAI DG
Results
Reference Fowler & O'Donnel '88 Stratton ef al. '88 Dahl & Sarvey '89 Washburn & Moises '89 Curet & deMontigny '89 Bijak '89 Curet & deMontigny '88 Munlieff & Dunwiddie '88 Madison & Nicoll '88 Hopkins & Johnston '88 Pang & Rose '87 Dunwiddie & Alford '87 Haas & Rose '87 Stanton & Heinemann '86 Harley & Milway '86 Madison & Nicoll '86 Lacaille & Harley '85 Sah et al. '85 Winson & Dahl '85 Haas & Konnerth '83 Madison & Nicoll '82 Segal '82 Harley ef al. '89
Code: A, intact rat; B, slice; C, iontophoresis/pressure; D, superfusion; E, LC stimulation; F, extracellular (single unit); G, extracellular (population); H, extracellular (ion selective electrode); I, intracellular recording; T , facilitation; I, suppression. Abbreviations: spont. act., spontaneous activity; mpp, medial perforant path; Lpp, lateral perforant path; POP, population spike.
325
response to lateral perforant path stimulation. Also, if indeed there is a @-adrenergic stimulation of theta interneurons, one can expect an increase in the local inhibitory tone in the hippocampus, which has not as yet been reported. Finally, while the @ excitation is widespread and pronounced, it is probably not the primary effect of LC activation; when the noradrenergic afferents are removed (by the injection of NE neurotoxins), the main reported result is an increase in the seizure susceptibility of the affected hippocampus (Kokaia et al., 1989). The @,-adrenergic excitatory action, whether being rather specific (e.g., an effect on LTP in CA3 neurons) or general (e.g., an effect on dentate population response) is likely to be mediated by a second messenger system which causes an increase in CAMPproduction (Segal et al., 1981). Indeed, activation of adenylate cyclase mimics some actions of NE in the hippocampus (Madison and Nicoll, 1986; Statton et al., 1988). It is not clear as yet how adenylate cyclase is mediating the various P-adrenergic effects in the hippocampus and elsewhere. The ionic mechanisms underlying the P-adrenergic excitation have also been studied extensively. The original reports on the blockade of afterhyperpolarization by @-ligands(Madison and Nicoll, 1982; Haas and Konnerth, 1983) considered it a unique action on a @-receptor,yet more recent studies suggest that NE reduces other potassium conductances as well (Foehring et al., 1989). These effects are clearly smaller in magnitude than the effect of acetylcholine which has been shown to affect similar potassium conductances (Segal, 19881, leaving the question of the specific ionic basis of NE action still open. One compounding factor is the consistent difference in the magnitude of action of NE seen in the intact brain and the slide preparation, suggesting that perhaps different mechanisms dominate the cellular responses under these two different conditions.
Norepinephrine in the hippocampus: a-adrenergic actions
By comparison, the a-adrenergic action in the hippocampus is more confusing. Most studies tend to agree that the main effect of a-agonists is to depress spontaneous or evoked activity of hippocampal neurons recorded extracellularly or intracellularly in the intact and the sliced hippocampus (Table 1). For example, Mynlieff and Dunwiddie (1988) found that a-adrenergic action suppresses population spike in region CAI of the hippocampal slice, and Curet and deMontigny (1988) found an a-adrenergic suppression of spontaneous hippocampal activity in the intact brain, in line with a previous observation by Pang and Rose (1987) and a later observation by Bijak (1989) in the hippocampal slice. While this is a general observation seen in a number of laboratories, there is no consensus as to which a-receptor species is involved in this action; Curet and deMontigny (1988) suggest that the action that they observed in the intact brain is mediated by an a,-receptor, yet most other studies suggest it is an a,-receptor (see Table 1). There are, however, exceptions to this rule; Madison and Nicoll (1988) found that a-adrenergic agents suppress IPSP’s and suggested that this action underlies the facilitatory action of NE on reactivity of CA1 neurons to afferent stimulation. Comparing results from extracellular population and single-cell recording in the intact brain and intracellular recording in the slice, it is quite striking to see how little direct effects on membrane properties a-agonists have; neither clonidine nor phenylephrine (PHE) exert any marked hyperpolarizing action in hippocampal cells in a slice (Segal, unpublished observations). While the primary action of N E at the a,-receptor in the hippocampus is reported to be inhibitory, its effects in most other brain regions studied are reported to be excitatory; a slow depolarization associated with a decrease in input
326
resistance, implying a reduction in resting K conductance, is reported in dorsal raphe cells, in spinal cord cells and in hypothalamic cells (Ma and Dun, 1985; Ogata and Matsuo, 1986; Freedman and Aghajanian, 1987; Fukuda et al., 1987; Yamashita et al., 1987). The reason for this marked difference between hippocampal and other neurons in the response to a,-adrenergic stimulation remains elusive. The biochemical action of the activated al-receptor has been studied extensively; a-activation causes an increase in inositol phosphate (PI) turnover (Exton, 1988; Wilson and Minneman, 1989). An activation of the PI system leads to an increase in intracellular calcium concentration which is associated with a host of cellular functions. The physiological correlates of the activation of PI turnover in the hippocampus are not entirely clear. Another adrenergic receptor residing in the hippocampus is presynaptic and likely modulates release of neurotransmitters from their terminals. A prominent receptor, likely to regulate release of NE, is an a,-receptor. Several studies have demonstrated that a,-ligands modulate the reactivity of the hippocampus to stimulation of the LC (Curet and deMontigny, 1989). Yet it is not clear if these receptors are solely associated with the noradrenergic terminals in the hippocampus. Effects of enhanced NE release on hippocampal reactivity to afferent stimulation
We have conducted experiments with the intact hippocampus to examine some possible a-adrenergic effects on reactivity of the hippocampus to afferent stimulation. Intraperitoneal injection of the a,-antagonist idazoxan caused an increase in population spike response of the dentate gyrus to stimulation of the perforant path. This action is presumed to be caused by the blockade of inhibitory input on the LC cells, which will, in turn, cause a rise in LC cellular firing rate and consequently an increase in NE release in target tissue. This effect of idazoxan was attenuated by prior
depletion of NE in the brain (Richter-Levin et al., 1991). Interestingly, associated with the increase in population spike in the dentate gyrus, there was a decrease in the population EPSP, representing the synaptic currents of the perforant path synapse. This indicates that idazoxan may have additional effects in the hippocampus other than causing the enhanced release of NE. At any rate, NE appears to enhance the EPSP/spike coupling in dentate granular cells. The a,-antagonist prazosin did not share this effect of idazoxan in the hippocampus and was ineffective when applied alone. Release of NE from LC terminals could also be enhanced by amphetamine, a drug know to cause release of catecholamines (Sanger, 1988). Following intraperitoneal injection of amphetamine, we found an increase in population spike response to perforant path stimulation without a consistent change in the population EPSP. This enhanced reactivity was blocked by sotalol, a P-antagonist but also by prazosin, indicating that the enhanced reactivity is mediated by both a,- and p noradrenergic receptors (Fig. 1). a-Adrenergic modulation of hippocampal reactivity to N-methyl-D-aspartate (NMDA)
Experiments were conducted with the hippocampal slide preparation to examine the possible roles of the a-adrenergic receptor in reactivity of the hippocampus to afferent stimulation. We recorded intracellular activity of CA1 pyramidal neurons and examined the effects of the P-receptor agonist isoproterenol, and the a-agonists PHE and clonidine on these cells. These ligands were applied by iontophoresis near the apical dendritic region of the cell, in stratum radiatum. In addition, we examined the effects of these ligands on reactivity of the recorded neurons to iontophoretic application of excitatory amino acids, NMDA and quisqualate. While having little effect on resting properties of the recorded neurons, PHE produced a marked potentiation of their reactivity to NMDA (Fig. 2). This effect was
327 0
b
Amphetom i n e
Baseline Prozosin
1000
5
200
milliseconds
C
Baseline 2 0 0 Amphetarnine ( 3 mg/kg Prozosin ( 2 m g /kg 1 3t p c 0 . 0 3
[
*
.-2
10
.-.. )
d
U
oseline
%
0 Amphetamine ( 3 m g / k g )
E
\
I5F
a P r o z o s ~ n( 2 m g / k g 1
al a
-n0
n m
a
5
Q v)
a
w
0
0
P
0
Baseline
-
f
Fig. 1. The amphetamine-induced increase in hippocampal reactivity to afferent stimulation is mediated by both a- and P-receptors. Rats were anesthetized and prepared for experiments as detailed elsewhere (Richter-Levin and Segal, 1990). A. Averaged responses of dentate gyrus to stimulation of perforant path before and 20 min after (asterisk) intraperitoneal injection of 3 mg/kg o-amphetamine. B. Lack of effect of prazosin (2 mg/kg) on population spike size measured in five different rats. C,D. Selective effects of amphetamine on population spike and lack of effect on EPSP slope (D) and its blockade by prazosine. Asterisk indicates a significant effect of amphetamine. E,F. Sotalol (2 mg/kg) blocks the effects of amphetamine in four rats. As seen in C and D, amphetamine increases the population spike size without affecting the EPSP. Sotalol blocks the effects of amphetamine on population spike.
shared to a small extent by clonidine and isoproterenol. The response to quisqualate was not affected by PHE. The potentiation of the response to NMDA was attenuated by prazosin but not by idazoxan. Since the response to NMDA is known to be
voltage dependent (Collingridge and Lester, 19891, it is possible that the observed effects of PHE are actually mediated by a shift in the voltage during the response to NMDA. To examine this, we voltage clamped the recorded neurons and examined the current responses to
328
r
L
P -
.
, ," ,, , 10.5nA 20sec
Fig. 2. Effects of phenylephrine (PHE) on reactivity of hippocampal neurons to iontophoretic application of N-methylD-aspartate (NMDA). Hippocampal neurons were recorded intracellularly with K-acetate-containing micropipettes in a slice preparation. A. Voltage (top record) responses to 20 nA current pulses of NMDA (bottom record) are markedly potentiated during concurrent application of PHE. The responses to NMDA are associated with an increase in conductance, estimated by the response to intercellular application of 0.5 Hz, 100 msec current pulses (middle trace) across the recorded cell membrane. B. In another cell, using the same paradigm, isoproterenol (ISO) causes a smaller and longer lasting increase in the number of action potential discharges in response to pulse application of NMDA.C. In another CA1 neuron, voltage-damped at resting potential ( = -65 mV, top record), the current responses to NMDA (middle record) are markedly enhanced by PHE. This effect is short-lasting as seen above. I, current; V, potential.
NMDA. PHE caused a marked increase in the reactivity to NMDA in a cell clamped at resting potential (Fig. 2C). These studies indicate that an a,-adrenergic action in the hippocampus selectively enhances the responses to an acidic amino acid acting at the NMDA receptor. Since an a,-adrenergic receptor activates PI turnover, we examined the possibility that myo-inositol-1,4,5-trisphosphate
(IP3) shares this action of PHE. We recorded from neurons with microelectrodes containing IP3, and found that here was a gradual increase in the reactivity of these neurons to NMDA, as the drug was leaking out of the pipette into the recorded cell. Activation of another second messenger mechanism associated with the noradrenergic receptor, the CAMP system, did not share this effect of IP3 and PHE. Instead, CAMP did cause a long-lasting increase in the depolarizing responses to quisqualic acid. IP3 releases calcium from intracellular stores, and indeed PHE was found to cause a rise in [Ca], following its application to cultured hippocampal -eurons (Murphy and Miller, 1989). In recent experiments, using a Ca-sensitive dye Fluo-3 in a confocal microscope, we found that PHE potentiates an NMDA-induced rise of [Ca], in cultured hippocampal neurons (Segal, unpublished observations). Our results illustrate a novel a-adrenergic effect in the hippocampus, which is related to the second messenger system associated with this receptor in the hippocampus and elsewhere. While region CA1 of the hippocampus is relatively poor in a,-receptor density, compared to some regions of the neocortex or the thalamus (Rainbow and Biegon, 19831, it is one of the highest in its association with the PI second messenger system (Johnson and Minneman, 1985). The ability of NE acting at an a,-receptor to enhance NMDA responses in these other structures needs to be analyzed. In the cerebellum, NE, acting at an a,-receptor, has a short-term enhancing action on the reactivity to glutamate (Marshall and Tsai, 1988). Within the hippocampus, there is only one other report of a direct facilitating action for PHE on evoked responses to afferent stimulation (Winson and Dahl, 198.5). While most studies associating long-term modulation of reactivity to afferent stimulation focussed on the P-adrenergic action which does cause a long-lasting effect, the possible contribution of the short acting a,ligands should be considered especially in view of the importance of NMDA in the initiation of LTP in region CA1 and in the dentate gyrus of
329
the hippocampus (Collingridge and Lester, 1989). The two receptor types, activated by N E and recruiting the two major known second rnessenger systems should thus have a major role in regulating neuronal plasticity.
References Bijak, M. (1989) Antidepressant drugs potentiate the a , adenoreceptor effect in hippocampal slices. Eur. J. PharmaC O ~ 166: . , 183-191. Collingridge, G.L. and Lester, R.A. (1989) Excitatory amino acid receptors in the vertebrate central nervous system. Pharinucol. Rw., 40: 143-210. Curet, 0. and deMontigny, C. (1988) Electrophysiological characterization of adrenoceptors in the rat dorsal hippocampus. 1. Receptors mediating the effect of microiontophoretically applied norepinephrine. Bruin Rrs., 475: 35-46. Curet, 0. and deMontigny, C. (1989) Electrophysiological characterization of adrenoreceptors in the rat dorsal hippocampus. 111. Evidence for the physiological role of terminal 0,-adrenergic autoreceptors. Bruin Res., 499: 18-26. Dahl, D. and Sarvey, J.M. (1989) Norepinephrine induces pathway-specific long lasting potentiation and depression in thc hippocampal dentate gyrus. Proc. Natl. Acad. Sci. USA, 86: 4776-4780. Dunwiddie, T.V. and Alford, C. (1987) Electrophysiological actions of phenycyclidine in hippocampal slices from the rat. Neuropharmacology, 26: 1267-1273. Exton, J.H. (1988) The roles of calcium and phosphoionositides in the mechanisms of a,-adrenergic and other agonists. Re[,. Physiol. Biochem. Pharmacol., 111: 117-224. Foehring. R.C., Schwindt, P.D. and Crill, W.E. (1989) Norepinephrine selectively reduces slow CAZf and Na-mediated K currents in cat neocortical neurons. J. Neurophysiol., 61: 245-256. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Ret,., 63: 844-914. Fowler, J.C. and O’Donnell, J.M. (1988) Antagonism of the responses to isoproterenol in the rat hippocampal slide with subtype-selective antagonists. Eur. J. Pharmucol., 153: 105-110. Freedman, J.E. and Aghajanian, G.K. (1987) Role of phosphoinositide metabolites in the prolongation of afterhyperpolarizations by a ,-adrenoreceptors in rat dorsal raphe neurons. J. Neurosci., 7: 3897-3906. Fukuda, A., Minami, T., Nabekura, J. and Oomura, Y . (1987) The effects of noradrenaline on neurons in the rat dorsal motor nucleus of the vagus, in titro. J. Physiol. (London), 393: 213-231. Haas, H.L. and Konnerth, A. (1983) Histamine and noradrenaline decrease calcium-activated potassium conductance in hippocampal pyramidal cells. Nature, 302: 432434.
Haas, H.L. and Rose, G.M. (1987) Noradrenaline blocks potassium conductance in rat dentate granule cells in vitro. Neurosci. Lett., 78: 171-174. Harley, C., Milway, J.S. and Lacaille, J.C. (1989) Locus coeruleus potentiation of dentate gyrus responses: Evidence for two systems. Bruin Res. Bull., 22: 643-650. Harley, C.W. and Milway, J.S. (1986) Glutamate ejection in the locus coeruleus enhances the perforant path-evoked population spike in the dentate gyrus. Exp. Bruin Res., 63: 143-150. Hopkins, W.F. and Johnston, D. (1988) Noradrenergic enhancement of long-term potentiation at mossy fiber synapses in the hippocampus. J. Neurophysiol., 59: 667-687. Johnson, R.D. and Minneman, K.P. (1985) a,-adrenergic receptors and stimulation of [3H]inositol metabolism in rat brain: Regional distribution and parallel inactivation. Brain Res., 341: 7-15. Kokaia, M., Bengzon, J., Kalen, P. and Lindvall, 0. (1989) Noradrenergic mechanisms in hippocampal kindling with rapidly recurring seizures. Bruin Rex, 491: 398-402. Lacaille, J.C. and Harley, C.W. (1985) The action of norepinephrine in the dentate gyrus: @-mediated facilitation of evoked potentials in vitro. Bruin Rex, 358: 210-220. Ma, R.C. and Dun, N.J. (1985) Norepinephrine depolarizes lateral horn cells of neonatal rat spinal cord in uitro. Neumsci. Lett., 60: 163-168. Madison, D.V. and Nicoll, R.A. (1982) Noradrenaline blocks accommodation of pyramidal cell discharge in the hippocampus. Nature, 299: 636-638. Madison, D.V. and Nicoll, R.A. (1986) Actions of noradrenaline recorded intracellularly in rat hippocampal CAI pyramidal neurones, in ritro. J. Physiol. (London), 373: 221-244. Madison, D.V. and Nicoll, R.A. (1986) Cyclic adenosine 3’5’ monophosphate mediates @-receptor actions of noradrenaline in rat hippocampal pyramidal cells. J. Physiol. (London), 372: 245-259. Madison, D.V. and Nicoll, R.A. (1988) Norepinephrine decreases synaptic inhibition in the rat hippocampus. Bruin R ~ s .442: , 131-138. Marshall, K.C. and Tsai, W.H. (1988) Noradrenaline induces short and long duration potentiation of glutamate excitations of cultured Purkinje neurons. Can. J. Physiol. Pharrnacol., 66: 848-853. Murphy, S.N. and Miller, R.J. (1989) Two distinct quisqualate receptors regulate Ca 2 + homeostasis in hippocampal neurons in riitro. Mol. Pharmacol., 35: 671-680. Mynlieff, M. and Dunwiddie, T.V. (1988) Noradrenergic depression of synaptic responses in hippocampus of rat: Evidence for mediation by a,-receptors. Neuropharmacolom, 27: 391-398. Ogata, N. and Matsuo, T. (1986) The effects of catecholamines on electrical activity of neurons in the guinea pig supraoptic nucleus in uitro. Brain Res., 385: 122-135. Pang, K. and Rose, G.M. (1987) Differential effects of norepinephrine on hippocampal complex-spike and theta-neurons. Brain Res., 425: 146-158.
330 Rainbow, T. and Biegon, A. (1983) Quantitative autoradiography of 3-H-prazosin binding sites in rat forebrain. Neurosci. Lett., 40: 221 -226. Rainbow, T.C., Parsons, B. and Wolfe, B.B. (1984) Quantitative autoradiography of p , and p2 adrenergic receptors in rat brain. Proc. Natl. Acad. Sci. USA, 81: 1585-1589. Richter-Levin, G . and Segal, M. (1990) Effects of serotonin releasers on dentate granule cell excitability in the rat. Exp. Brain Res., 82: 199-207 Richter-Levin, G., Segal, M. and Sara, S. (1991) An a 2 antagonist, idazoxan, enhances EPSP-spike coupling in the rat dentate gyrus. Brain Res., 540: 291-294.. Sah, P., French, C.R. and Gage, P.W. (1985) Effects of noradrenaline on some potassium currents in CAI neurones in rat hippocampal slices. Neurosci. Lett., 60: 295300. Sanger, D.J. (1988) Behavioural effects of the a2-adrenoceptor antagonists idazoxan and yohimbine in rats: Comparisons with amphetamine. Psychopharmacology, 96: 243249. Sara, S.J., Devauges, V. and Segal, M. (1988) Locus coeruleus engagement in memory retrieval and attention. In Progress in Catecholamine Research, Part B: Central Aspects. Alan R. Liss, New York, pp. 155-161. Segal, M. (1982) Norepinephrine modulates reactivity of hippocampal cells to chemical stimulation in uitro. Exp. Neurol., 17: 86-93. Segal, M. (1988) Synaptic activation of a cholinergic receptor in rat hippocampus. Brain Rex, 42: 79-86.
Segal, M. and Bloom, F.E. (1974) Norepinephrine in the rat hippocampus: Iontophoresis studies. Brain Res., 82: 79-97. Segal, M. and Bloom, F.E. (1976) The action of norepinephrine in rat hippocampus: IV. The effects of locus coeruleus stimulation on evoked hippocampal unit activity. Brain Rex, 107: 513-525. Segal, M., Greenberger, V. and Hofstein, R. (1981) Cyclic AMP generating systems in rat hippocampal slices. Brain Rex, 213: 351-364. Stanton, P.K. and Heinemann, U. (1986) Norepinephrine enhances stimulus evoked calcium and potassium concentration changes in dentante granular layer. Neurosci. Lett., 67: 233-238. Statton, K.R., Baraban, J.M. and Worley, P.F. (1988) Norepinephrine stimulation of adenylate cyclase potentiates protein kinase C action: Electrophysiological studies in the dentate gyrus. Synapse, 2: 614-618. Washburn, M. and Moises, H.C. (1989) Electrophysiological correlates of presynaptic a,-receptor-mediated inhibition of norepinephrine release at locus coeruleus synapses in dentate gyrus. J. Neurosci., 9: 131-140. Wilson, K.M. and Minneman, K.P. (1989) Regional variations in a,-adrenergic receptor subtypes in rat brain. J. Neurochem., 53: 1782-1786. Winson, J. and Dahl, D. (1985) Action of norepinephrine in the dentate gyrus. 11. Iontophoretic studies. Exp. Brain Res., 59: 497-506. Yamashita, H., Inenaga, K. and Kanna, H. (1987) Depolarizing effect of noradrenaline on neurons of the rat supraoptic nucleus in c'itro. Brain Res., 405: 348-352.
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
331 CHAPTER 25
The cerebellar norepinephrine system: inhibition, modulation, and gating D.J. Woodward
’,H.C. Moises
2,
B.D. Waterhouse
3,
H.H. Yeh and J.E. Cheun
Department of Cell Biology, University of Texas Southwestern Medical Center at Dallas, Harry Hines Bouleuard, Dallas, TX, Department of Physiology, University of Michigan Medical School, Ann Arbor, MI, Department of Physiologv and Biophysics, Hahnemann University, Broad and Vine, Philadelphia, PA and Department of Neurobiology and Anatomy, University of Rochester Medical Center, Elmwood Avenue, Rochester, U.S.A.
A series of studies has been conducted to determine the mode of action on the cerebellar cortical circuitry of the norepinephrine (NE)-containing afferents from the locus coeruleus. NE has been known to exert an “inhibitory” action on the background firing observed in Purkinje cells, due presumably to a shift in conductances favoring hyperpolarization. An additional independent action at low threshold appears to be an enhancement of GABA, the inhibitoj transmitter of cerebellar interneurons. Recent whole-cell
patch-clamp studies on isolated Purkinje cells indicate that exposure to NE increases the chloride current caused by transient pulses of GABA applied iontophoretically. NE applied to Purkinje cells in the parafloccular lobule during stimulation by moving visual patterns revealed the capacity either to “gate” signals initially not expressed, or to amplify the gain of phasic excitations. The control of emergent circuit functions may be the functional consequence of the multiple modulatory functions of NE.
Key words: norepinephrine, GABA, modulation, gating
Introduction
The cerebellum has long been regarded as among the most promising model systems available for clarifying of actions of the norepinephrine (NE)containing system of afferents from the locus coeruleus (LC). The presumption has been that the basic cellular mechanisms of action of NE, as they are revealed, will reflect to a great extent the processes that occur elsewhere in the central nervous system (CNS). In addition, the functional consequences of the synaptic action of NE on the cerebellar circuits should be expected to bear
some relation to functions of NE in other brain regions. The cerebellum, as a major component of the motor systems of the CNS, is of course, itself intrinsically important to understand. Using the cerebellum as a model system, we have high expectations that insights at all levels of function will generalize to other structures in the CNS. NE as an inhibitory transmitter Experiments providing initial insights into the actions of NE in the cerebellum employed the technique of multibarrel drug iontophoresis
332
through micropipettes. This technique became practical in the late 1960s when circuits were developed to prevent electrical transients from directly stimulating neurons during drug application. As with most neurons in the CNS, the cerebellar Purkinje cells of anesthetized preparations uniformly exhibited an inhibition of their characteristic 10-40/sec discharge of spontaneous firing following brief iontophoretic applications of NE. The appellation “Inhibitory Transmitter” was applied to NE largely on the basis of this single effect. Additional weight was given to this idea when it was found (Hoffer et aZ., 1971b) that mean firing rates of Purkinje cells were elevated from 35/sec, seen under normal conditions, to about 50/sec after destruction of NE fibers by administration of the neurotoxin 6-hydroxydopamine (6-OHDA). A major advance in understanding NE’s mechanism of action occurred in the early 1970s when arguments were made that NE did not act as a conventional transmitter by directly gating the activity of ion channels but elevated the levels of a second messenger, cyclic AMP (CAMP),through activation of the enzyme adenyl cyclase (Siggins et al., 1969; Hoffer et aZ., 1971a). The phosphorylation of A kinase, dependent on CAMP,presumably led to further phosphorylations which more directly produced the physiological effects. Drugs that prolonged the actions of CAMP, i.e., the methyl xanthines, enhanced the actions of NE as would be expected if the actions of the neurotransmitter were mediated via the adenylyl cyclase-CAMP cascade. The important notion that transmitters such as NE can work without direct increases in conductance came about when initial results were reported in rat cerebellar Purkinje cells (Siggins et al., 1971). Those studies employed specially constructed pipettes in which a single, fine-tipped microelectrode used for intracellular recording was glued to a multibarrel pipette so that drugs could be applied in v i m to regions of single neurons. In this way, simultaneous measurements were made of membrane voltage and resistance
changes during NE application. The data showed that a suppression of firing by NE was accompanied by hyperpolarization. However, membrane conductance clearly did not decrease, as after administration of GABA, but either remained unchanged or was increased in some instances up to 100%. The conclusion advanced by Siggins and coworkers (1971) was that an unidentified mechanism was at work which could modify or eliminate a large fraction of the resting conductance in order to yield the inhibitory effect. Release of endogenous NE by stimulation of the LC region yielded similar results, as did direct NE administration. That is, there occurred a hyperpolarization and resistance increase correlated with a suppression of firing. Effects evoked by a 5-10 sec train of stimuli lasted up to a minute, consistent with a long-lasting action of a second messenger system (Hoffer et aZ., 1973). Still, in these studies the effects were quite variable; it was not possible to identify what fractions of which conductances were modified to yield these effects, and the possibility of interactions with synaptic mechanisms was still untested. NE as a modulatory neurotransmitter
The initial laboratory observation (D.J.W.) that NE could have a modulatory function was found in 1970 when initial exploratory studies of 6OHDA-treated animals revealed weak inhibitory responses of Purkinje cells to surface stimulation of the parallel fiber system. It was clear, but puzzling, that the GABA-mediated inhibition normally caused by activation of basket and stellate cells was attenuated in the absence of the NE system. It became evident at that time that a thorough physiological study would be required of the range of interactions of NE with all of the synaptic systems in the cerebellar circuitry. A pioneering study by Foote et aZ. (1975) suggested that the inhibitory action of NE in the cerebral cortex would have the effect of suppressing spontaneous discharge yet leaving responses to strong excitatory input intact. The net effect of
334
1983). Interactions with glutamate were somewhat less clear, but excitations were clearly preserved relative to the suppression of background activity by NE. Stimulation of the LC with trains of pulses to cause synaptic release of NE caused a similar increase in GABA action (Fig. 11, lasting for tens of seconds, and also enhanced the
tic conductances. More generally, the aim was to fully characterize the interaction of NE with the physiological properties of Purkinje cell function. NE applied iontophoretically was found to enhance the inhibitory action of GABA even when levels of NE were below that which caused direct inhibition of background activity (Moises et al., '
1Contrd
2NE
'
- '
3Recoveiy
" h Rprnvprv
5 sec Fig. 2. Antagonistic effect of iontophoretically applied norepinephrine (NE) on Purkinje cell inhibition induced by taurine and p-alanine. Responses of two different neurons are illustrated in (a) and (b). Ratemeter record in (a) shows inhibition of Purkinje cell discharge produced by iontophoretic pulses of taurine of 60 nA (solid bar) before, during and after administration of NE at 10 nA (dotted line). Corresponding drug response histograms show Purkinje cell inhibitions produced by taurine reduced from 53 to 48% during NE (10 nA) iontophoresis, despite the concomitant decrease in background discharge. Similar records from another cell (b) show that inhibitions produced by p-alanine at 68 nA were antagonized (54-43%) by NE at 50 nA. All histograms contain 6 sweeps. Vertical calibrations for ratemeter records are in spikes/sec and those for histogram records are in counts per address. 5 sec (From Yeh et al., 1981.) Fig. 2. Antagonistic effect of iontophoretically applied norepinephrine (NE) on Purkinje cell inhibition induced by taurine and p-alanine. Responses of two different neurons are illustrated in (a) and (b). Ratemeter record in (a) shows inhibition of Purkinje tic generally,pulses the of aim was ofto60 nA (solid 1983). withafterglutamate were somebar)Interactions before, during and administration of NE at 10 cell conductances. discharge producedMore by iontophoretic taurine nA (dotted line). Corresponding drug response histograms show Purkinje inhibitions produced by taurine reduced from pre53 to fully characterize the interaction of NE with the whatcell less clear, but excitations were clearly 48% during NE (10 nA) iontophoresis, despite the concomitant decrease in background discharge. Similar records from another physiological properties of Purkinje cell function. served relative to the suppression of background 68 histograms nAofwere antagonized cell (b) show that inhibitions produced by p-alanine atAll contain 6 (54-43%) by NE at 50 nA. All histograms contain 6 NE applied iontophoretically was found to enactivity by for NE. Stimulation LC with trains sweeps. Vertical calibrations for ratemeter records arethe spikes/sec and those histogram records of are the in counts per address. LC with trains inincounts per address. al., 1981.) action of GABA even (From hanceYeh theetinhibitory pulses to cause synaptic release of NE caused whenof NE of release caused
a similar levels of NE were below that which causedaction direct(Fig. 11, last- increase in GABA action (Fig. 11, lastinhibition of background activity (Moises also et al.,enhanced ing for the tens of seconds, and also enhanced the
334
'
1Contrd
2NE
'
- '
3Recoveiy
" h Rprnvprv
5 sec Fig. 2. Antagonistic effect of iontophoretically applied norepinephrine (NE) on Purkinje cell inhibition induced by taurine and p-alanine. Responses of two different neurons are illustrated in (a) and (b). Ratemeter record in (a) shows inhibition of Purkinje cell discharge produced by iontophoretic pulses of taurine of 60 nA (solid bar) before, during and after administration of NE at 10 nA (dotted line). Corresponding drug response histograms show Purkinje cell inhibitions produced by taurine reduced from 53 to 48% during NE (10 nA) iontophoresis, despite the concomitant decrease in background discharge. Similar records from another cell (b) show that inhibitions produced by p-alanine at 68 nA were antagonized (54-43%) by NE at 50 nA. All histograms contain 6 sweeps. Vertical calibrations for ratemeter records are in spikes/sec and those for histogram records are in counts per address. (From Yeh et al., 1981.)
tic conductances. More generally, the aim was to fully characterize the interaction of NE with the physiological properties of Purkinje cell function. NE applied iontophoretically was found to enhance the inhibitory action of GABA even when levels of NE were below that which caused direct inhibition of background activity (Moises et al.,
1983). Interactions with glutamate were somewhat less clear, but excitations were clearly preserved relative to the suppression of background activity by NE. Stimulation of the LC with trains of pulses to cause synaptic release of NE caused a similar increase in GABA action (Fig. 11, lasting for tens of seconds, and also enhanced the
335
excitation caused by climbing (Fig. 1) and mossy fiber input (Woodward et al., 1979; Moises and Woodward, 1980; Moises et al., 1981, 1983). NE also enhanced the inhibitory action of muscimol, but not taurine or p-alanine (Fig. 2) or glycine (Yeh et at., 1981). Other effects have been characterized. For example, these modulatory-type actions of NE have been found early in development before dendritic maturation of Purkinje cells (Yeh and Woodward, 1983b,c). The use of selective agonists and antagonists for &- and /?,-subreceptors argue for involvement of p ,-receptor subtypes in inducing the enhancement of GABA (Yeh and Woodward, 1983a). Destruction of NE fibers with 6-OHDA results in a supersensitivity of the modulation of GABA by NE (Moises et al., 1980). In addition, chronic drug treatment such as daily injections of flurazepam (Waterhouse et al., 1983) or desmethylimipramine (Yeh and Woodward, 1982) results in a selective elimination of the NE-GABA enhancement after eight days of adaptation to the drug. These types of results suggested that the lowest threshold, @,-receptormediated actions of NE on GABA efficacy appeared to occur independent of the direct inhibitory actions of NE on Purkinje cells which were observed at slightly higher levels of administration. It is perhaps not surprising that multiple actions might be found, since the noradrenergic actions mediated via a rise in intracellular CAMP alone might be expected to involve complex phosphorylation processes. Overall, the action of NE was found to modulate conventional synaptic actions of amino acid transmitters. It does not appear that the resistance increase or shift in conductances responsible for the suppression of background firing could done account in any simple way for the combinations of interactions found between iontophoretically or synaptically released NE and putative excitatory and inhibitory amino acid transmitters. It must be emphasized that the primary value of such work on the in v i m cerebellum is in validating the idea of modulation in a preparation with
a relatively undisturbed state of cellular physiology. Indeed, it has been shown by West and Woodward (1984) that iontophoretically applied NE causes a marked potentiation of GABA inhibition in Purkinje cells in awake locomoting rats. This observation is critically important for demonstrating that modulations, presumably based on the postulated cascade of events in protein phosphorylation, employ normal physiological mechanisms. The iontophoretic data clearly are not optimal for revealing the channel alterations responsible for these effects. The development of techniques for voltage clamp of the whole-cell and patch-clamp techniques now offer more direct methods for testing such issues. Recent preliminary findings by Cheun and Yeh at the University of Rochester have shown the feasibility of isolating single Purkinje cells from postnatal rats after the modulatory actions of NE are fully established. Whole-cell patch-clamp studies have progressed considerably toward a verification of concepts derived from extracellular studies. Application of NE and GABA to an isolated single Purkinje cell revealed an amplification of GABA-induced chloride currents, at levels of NE having no other membrane effects (Fig. 3). Administration of forskolin, which acts to directly activate adenyl cyclase, also amplified the currents induced by GABA pulses in these kinds of experiments (Fig. 4). Such studies suggest that conditions can be established, suitable for study of such phenomena, in isolated cells free from the complexity of intact circuitry. Similar results from studies conducted in cerebellar tissue slices by Sessler et al. (1989) also validate the findings of earlier iontophoretic studies on intact preparations. Recent advances in studies on the GABA, receptor have demonstrated a , p, 6 and y subunits with multiple subspecies for many of the units. In situ hybridization and binding studies indicate as many as ten distinct genetic variations of the GABA receptor may exist in different neurons in the CNS (R. Olson, personal communication). Our simplest concept of NE modula-
336
tion, formulated a decade ago (Woodward et aZ., 19791, held that NE might directly alter the GABA receptor system, presumably by the known CAMPmessenger system. In view of the diversity of the GABA receptor structure, it would seem possible that enhanced channel opening could occur by means of an A-kinase, CAMP-dependent, mechanism for phosphorylating a site on the GABA receptor or related structure near the inner surface of the membrane (Schofield et al., 1987; Sweetnam et al., 1987). Considerable variation could be expected in the degree of physiological regulation of GABA receptor efficacy from one region to another, and this might account in part for regional differences in NE actions.
NE as gating regulator
A major task has been to define the physiological parameter that is regulated by the N E systems. The cerebellar cortex includes a feed-forward excitation via the mossy fiber excitation of granule cells and then of Purkinje cell dendrites. A feed-forward inhibition is mediated if parallel fiber axons excite basket and stellate cells which then inhibit Purkinje cells via release of GABA. Temporally dispersed signals allow interactions to be observed between these effects in a more natural mode than when synaptic responses are elicited by pulsed electrical stimulation of cerebellar afferents. The cerebellum is known to re-
Fig. 3. NE modulates Purkinje cell responses to GABA by augmenting GABA-activated current amplitude and prolonging relaxation time. Left. A solitary Purkinje cell obtained following gentle enzymatic dissociation of rat cerebellum. The truncated axon and dendritic arbor and immunoreactivity to PEP-19, a marker for Purkinje cells, aided in its identification. Right (Top). A family of inward and outward currents typically seen in freshly dissociated Purkinje cells. The cells are viable, and whole-cell recordings showed adequate space clamp characteristics. In this example, currents were elicited by 10 mV depolarizing command steps from a holding potential of - 65 mV which was near resting potential based on zero current measurements. Bar = 5 msec/0.5 nA. (Bottom). GABA (10 FM) were applied to the same cell and the resultant inward currents were recorded before (A), during (B and C) and after (D and A) application of 250 pM NE. In (A), current traces during the control and full recovery periods overlap. Full recovery from the NE-induced effect on GABA-activated currents took approximately 3-5 min. Calibration: 50 msec/O.l nA.
337
I
n
A
i
30 sec GABA
Isec
p-
. ..
.:
For s kolin
. ..
Fig. 4. Forskolin potentiates I,,,. A. Continuous penwriter in a Purkinje cell held at -65 mV record illustrating I, before, during, and after forskolin (200 p M ) application. Pressure pulses (200 msec) of GABA (10 pM) was applied every 30 sec (arrowheads above penwriter record). The solid bar above the penwriter record indicates the period of recovered to control levels 3 min forskolin application. I,, after cessation of forskolin application. B. The current responses following forskolin (as shown in A) were digitized and superimposed on those records during the control (pre-forskolin) and recovery (post-forskolin) periods. The solid traces from control and recovery correspond to averaged I, immediately periods, while the dotted trace represents I after exposure to forskolin.
ceive widespread inputs, but few of these, particularly from the cerebral cortex, brain stem and spinal cord are conveniently and easily controlled by experimentally defined conditions. An advance was made in this regard by the discovery that the secondary visual cortex in rat projects onto the parafloccular lobule (Burne et al., 1978; Burne and Woodward, 1983). It was possible therefore to activate the cerebellum by moving, visual patterns controlled by a digitally generated ramp waveform. Neurons in the visual pathway from retina to lateral geniculate, primary and secondary visual cortices become activated and project signals down to the dorsolateral region of the pons and then to the paraflocculus. Studies by Moises et al. (1990) revealed that a
variety of responses could be found in the cerebellum when moving, visual stimuli were employed. A remarkable finding was that a class of Purkinje cells which exhibited little response to moving visual patterns, when presented under control conditions (i.e., prior to exogenous NE application), became responsive after administration of NE via microiontophoresis (Fig. 5). After NE administration, the visual signals could be characterized as having a “strong effect,” although before NE nothing at all was observed. Such results can hardly be characterized as increased “signal to noise” since there was no NE action on the background firing rate and no “signal” prior to beginning iontophoresis of the catecholamine. Other Purkinje neurons (Fig. 6) that were responsive to visual stimuli under initial control conditions revealed an absolute increase in excitation at low levels of NE administration. Our view is that the effect of causing a signal to arise from a stable background is better characterized as an action of NE to alter the “gating properties of the circuit.” One might imagine a situation where feed-forward excitation and inhibition are equally balanced between the direct parallel fiber and basket stellate cell actions on Purkinje cells. The net action of an increase in GABA receptor function caused by NE would be to unbalance the input and “gate in” an effective response. Other similar gating control functions might be found depending on the normal setpoint or operating level of a particular neuronal circuit.
Conclusions
An understanding of how NE might regulate cerebellar functions is begining to emerge. NE actions on transmitter functions appear to lead to an increase in synaptic efficacy. An enhancement of GABA receptor functions is a likely explanation for the enhanced synaptic inhibition found after NE. A consequence of these elevated actions may be to regulate circuit gating as an emergent function.
338
operation has yet to emerge to define the context for understanding NE action further. Little progress has yet been made to correlate physiological release of NE during normal behavior with either its identified cellular actions or with meaningful changes in the operating properties of the cerebellar circuits. The relations between these short-term effects and the possibility of
Many questions remain unanswered regarding both the underlying basic cellular mechanisms and global action of NE. Actions of synaptically released NE have yet to be characterized as well as the interactions with excitatory mechanisms, calcium currents and subsequent interactions with receptor function as well as spike initiation zones. An integrated view of the Purkinje neuron in
OTempora'
-0.5 0
Seconds
5.5 Vert. bar
NoreDineghrine 5 0 n A
Recoverv 9 - 1 2 m i n
Recovery 0 - L rnin
Spikes/s
236-
236-
-0.5
0
Seconds
5.5
-0.5
0
Seconds
5.5
Fig. 5. Expression of an inhibitory response of a neuron to visual stimulation by NE iontophoresis. A. Prior to administration of NE, this Purkinje cell showed no appreciable response to presentations of a vertically oriented light bar stimulus (9" X 903 moving from left to right (upslope of trapezoidal waveform) and back through its receptive field. B. During iontophoresis of NE 50 Na, the cell responded in an inhibitory manner to presentations of the visual stimulus with a preference for movement in the forward (left-to-right, indicated by small arrow) direction. Note that the expression of the inhibitory response by NE was accompanied by an elevation in background discharge of the neuron. C,E. Note also that the cell continued to show inhibitory responses to visual stimulation for upwards of 8 min following the termination of NE administration. D. The movement trajectory of the light bar stimulus (long arrow) through the inhibitory zone of the cell's receptive field (shading). (From Moises et aL, 1990.)
339
D GABAlOnA
A Control 11"x176" Vert bar
Vert. b a r
... ...-..-..-_--.....-.-."r" ..-.. ..... ............................ .......................... .:..:. ......................................... ..... .-. ........... -..-.... ..... .......... --. ....... ........... ...................... -.-. ............ ...................................... ................... .-... ........ .................. -................... . - ............... ......... . ..- -" -..-.-. -.. .. _....-...................................... ........... -. .................................. ...-..-......-..................... .......................................... ................................ ................................. - . . I .
-0.5 0
6.5
Seconds
Norepinephrine lOnA
.................. --..-........... -. ............. . . ". . .... . . ........-. .-..... ..-..-. .. -...... ............... ............ ..... . .-....... ................. ---. _.-_. . . .. ... .. . ...-..... ... . . . . . . . . . . ............ ........... ..---.. -..--.... .. . ..-.-.. ... . ..... .... . ......... .. .... ._......... -. .. .._ ......... -.. .-.. .......... . ......... -. ... . . ......... .--..... ._. . . . . ..... .... ..... ... .... .-........... ........... . .. . .--. . -. -.... ......-.-. .....- ..............-.
-0.5
0
6.5
Seconds
Recovery
U
Spikes/s
. . C
6.5
Norepinephrine lOnA + Sotalol 25nA
s
Temporal
B I N : 20 m s
-05 0
t
90
Spikes/s ia.4 -
:
45
.
L
4
45
TempoNral
90
-45
Seconds
6.5
Fig. 6. Comparison of the effects of iontophoresis of N E and GABA on visually evoked and spontaneous simple spike discharge of a Purkinje cell. A. This neuron was stimulated by movement of a vertical bar of light (11" X 176") from right to left (upslope of trapezoidal waveform) and in the reverse (preferred) direction through its receptive field. B. During iontophoresis of N E 10 nA, spontaneous discharge was markedly reduced, whereas activity evoked by stimulus movement in the preferred direction (left-to-right) was only slightly decreased, resulting in the enhancement of the ratio of signal to noise. Note that N E had the additional effect of enhancing the cell's excitatory response to movement of the stimulus in the non-preferred direction (from right to left). C . This enhancement of the visual responses by NE was blocked when the specific P-adrenergic antagonist sotalol 25 n A was applied concurrently to the cell by iontophoresis. Note that sotalol antagonized both the depression in spontaneous activity by N E and the enhancement in the response to right-to-left movement of the visual stimulus. Drug iontophoresis was suspended for 20 min following the generation of this histogram record to allow for complete recovery to the control levels of visually evoked and spontaneous activity before further testing with GABA was initiated. D. In contrast to the effects observed during NE, iontophoresis of GABA 10 nA produced a smaller suppression in background firing and yet virtually eliminated the responses to visual stimulation. Recovery of the visually evoked responses to near control levels was observed following the cessation of N E ( C ) and GABA iontophoresis (E). F. The receptive field location of the excitatory zone (shading) and forward movement trajectory (arrow) of the light bar stimulus. (From Moises et al., 1990.)
340
long-term plastic changes also remain to be clarified. Many encouraging results relative to these issues are forthcoming from the new generation of research discussed by other authors in this volume. Acknowledgements
Supported by DA-02338, AA-03901, and Biological Humanics Foundation (DJW); NS-24830 and NS-01340 (HHY). References Burne, R.A. and Woodward, D.J. (1983) Visual cortical projections to the paraflocculus in the rat: An electrophysiologic study. Exp. Brain Res., 49: 55-67. Burne, R.A., Mihailoff, G.A. and Woodward, D.J. (1978) Visual corticopontine input to the paraflocculus: A combined autoradiographic and horseradish peroxidase study. Brain Res., 143: 139-146. Foote, S.L., Freedman, R. and Oliver, A.P. (1975) Effects of putative transmitters on neuronal activity in monkey cortex. Brain Res., 86: 229-242, Freedman, R., Hoffer, B.J., Woodward, D.J. and Puro, D. (1977) Interaction of norepinephrine with cerebellar activity evoked by mossy and climbing fibers. Exp. Neurol., 55: 269-288. Hoffer, B.J., Siggins, G.R. and Bloom, F.E. (1971a) Studies on norepinephrine containing afferents to Purkinje cells of rat cerebellum. 11. Sensitivity of Purkinje cells to norepinephrine and related substances administered by microiontophoresis. Brain Res., 25: 523-534. Hoffer, B.J., Siggins, G.R., Woodward, D.J. and Bloom, F.E. (1971b) Spontaneous discharge of Purkinje neurons after destruction of catecholamine-containing afferents by 6-hydroxydopamine. Brain Res., 30: 425-430. Hoffer, B.J., Siggins, G.R., Oliver, A.P. and Bloom, F.E. (1973) Activation of the pathway from locus coeruleus to rat cerebellar Purkinje neurons: Pharmacological evidence of noradrenergic central inhibition. J. Pharmacol. Exp. Ther., 184: 553-569. Moises, H.C. and Woodward, D.J. (1980) Potentiation of GABA inhibitory action in cerebellum by locus coeruleus stimulation. Brain Rex, 182: 327-344. Moises, H.C., Woodward, D.J., Hoffer, B.J. and Freedman, R. (1979) Interactions of norepinephrine with Purkinje cell responses to putative amino acid neurotransmitters applied by microiontophoresis. Exp. Neurol., 64: 489-515. Moises, H.C., Hoffer, B.J. and Woodward, D.J. (1980) GABA facilitation by noradrenaline shows supersensitivity in cerebellum after 6-hydroxy-dopamine. Naunyn Schmiedehergs Arch. Pharmacol., 315: 37-46. Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1981)
Locus coeruleus stimulations potentiate Purkinje cell responses to afferent input: The climbing fiber system. Brain Res., 222: 43-64. Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1983) Locus coeruleus stimulations potentiate local inhibitory processes in rat cerebellum. Brain Res. Bull., 10: 795-804. Moises, H.C., Burne, R.A. and Woodward, D.J. (1990) Modification of the visual response properties of cerebellar neurons by norepinephrine. Brain Res., 514: 259-275. Schofield, P.R., Darlison, M.G., Fujita, N., Burt, D.R., Stephenson, F.A., Rodriguez, H., Rhee, L.M., Ramachandran, J., Reale, V., Glencorse, T.A., Seeburg, R.H. and Barnard, E.A. (1987) Sequence and functional expression of the GABA A receptor shows a ligand-gated receptor superfamily. Nature (London), 328: 221-227. Sessler, F.M., Mouradian, R.D., Cheng, J.-T., Yeh, H.H., Liu, W., and Waterhouse, B.D. (1989) Noradrenergic potentiation of cerebellar Purkinje cell responses to GABA: Evidence for mediation through the p-adrenoceptor-coupled cyclic AMP system. Brain Res., 499: 27-38. Siggins, G.R., Hoffer, B.J. and Bloom, F.E. (1969) Cyclic adenosine monophosphate possible mediator for norepinephrine effects on cerebellar Purkinje cells. Science, 165: 1018-1020. Siggins, G.R., Oliver, A.P., Hoffer, B.J. and Bloom, F.E. (1971) Cyclic adenosine monophosphate and norepinephrine: Effects on transmembrane properties of cerebellar Purkinje cells. Science, 171: 192-194. Sweetnam, P.M., Tallman, J.F., Gallager, D.W. and Nestler, E.J. (1987) Phosphorylation of a subunit of the GABA/ benzodiazepine receptor by a receptor-associated protein kinase. Soc. Neurosci. Abstr., 13: 964. Waterhouse, B.D., Moises, H.M., Yeh, H.H., Geller, H.M. and Woodward, D.J. (1983) Evidence that benzodiazepines and norepinephrine act through a common mechanism to facilitate Purkinje cell responses to GABA. J. Pharmacol. Exp. Ther., 228: 257-267. Waterhouse, B.D., Sessler, F.M., Cheng, J.-T., Woodward, D.J., Azizi, S.A. and Moises, H.C. (1988) New evidence for a gating action of norepinephrine in central neuronal circuits of mammalian brain. BrainRes. Bull., 21: 425-432. West, M.O. and Woodward, D.J. (1984) Iontophoresis in the freely moving rat: Norepinephrine in the cerebellum. Soc. Neurosci. Abstr., 10: 71. Woodward, D.J., Waterhouse, B.D., Hoffer, B.J. and Freedman, R. (1979) Modulatory actions of norepinephrine in the central nervous system. Fed. Proc., 38: 2109-2116. Yeh, Y.H. and Woodward, D.J. (1982) Alterations in padrenergic response characteristics following long-term treatment with desmethyl-imipramine: Interaction between norepinephrine and y-aminobutyric acid in rat cerebellum. J. Pharmacol. Exp. Ther., 226: 126-134. Yeh, H.H. and Woodward, D.J. (1983a) p-1adrenergic receptors mediate noradrenergic facilitation of Purkinje cell responses to y-aminobutyric acid in cerebellum of rat. Neuropharmacology , 22: 629-639. Yeh, H.H. and Woodward, D.J. (1983b) Noradrenergic action in the developing rat cerebellum: Interaction between
341 norepinephrine and y-aminobutyric acid applied microiontophoretically to immature Purkinje cells. Dec. Brain Rex, 10: 49-62. Yeh, H.H. and Woodward, D.J. ( 1 9 8 3 ~ )Alterations in p adrenergic physiological response characteristics after long-term treatment with desmethyl imipramine: Interaction between norepinephrine and y-aminobutyric acid in rat cerebellum. J. Pharmacol. Exp. Ther., 226: 126-134.
Yeh, H.H., Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1981) Modulatory interactions between norepinephrine and taurine, p-alanine, y-aminobutyric acid and muscimol, applied iontophoretically to cerebellar Purkinje cells. Neuropharmacology, 20: 549-560.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Prr0gre.s in Bruin Research, Val. ti8 0 1991 Elsevier Science Publishers B.V
343 CHAPTER 26
Norepinephrine effects on spinal motoneurons S.R. White, S.J. Fung and C.D. Barnes Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, WA, U.S.A.
Intracellular recordings from cat spinal motoneurons in situ demonstrated that microiontophoretic application of N E with tow-intensity ejection currents produces a slowly developing, small-amplitude depolarization of the cells, in contrast to early reports of NE-induced hyperpolarization. This depolarization was associated with an increase in excitability of the cells and a decrease in membrane conductance. These observations are consistent with the hypothesis that N E reduces potassium conductance in spinal motoneurons as has been
proposed for facial motoneurons (VanderMaelen and Aghajanian, 1980) and thalamic neurons (McCormick and Prince, 1988). The time course of the facilitatory effects of N E on cat motoneuron excitability recorded intracellularly agreed very closely with the time course of NE-induced facilitation of glutamate-evoked excitability in rat spinal motoneurons recorded extracellularly. The similarity of the observations in rats and cats suggests that N E functions generally to enhance mammalian motoneuron responsiveness to excitatory input.
Key words: norepinephrine, spinal motoneuron, microiontophoresis, spinal cord
Introduction
The ventral horn of the mammalian spinal cord contains norepinephrine (NE) fiber plexuses at cervical through sacral levels, and some of these NE fibers appear to surround the large somata of the spinal motoneurons (CarIsson et al., 1964; Anden, 1965; Dahlstrom and Fuxe, 1965; Konishi, 1968; Commissiong et al., 1978; Mizukawa, 1980; Westlund and Coulter, 1980, Westlund et al., 1983). Although there is disagreement about the exact origin of ventral horn NE terminals, reports agree that most spinal cord NE terminals derive from cell bodies in the pons: the locus coeruleus (LC) and adjacent subcoeruleus, the
medial and lateral parabrachial nuclei, the Kolliker-Fuse nucleus and a region dorsolateral to the superior olivary nucleus (Westlund and Coulter, 1980; Westlund et al., 1983; Fritschy et al., 1987; Lyons et al., 1989; Reddy et al., 1989). Electrical stimulation in the region of the LC enhances spinal somatomotor outflow in both cats and rats, an effect that is attenuated by a-adrenergic blocking drugs (Strahlendorf et al., 1980; Fung and Barnes, 1981, 1987; Chan et al., 1986; Lai et af., 1989). This suggests that NE that is released in the spinal cord may enhance spinal motoneuron excitability. However, local application of NE directly into the vicinity of motoneurons has been reported to have opposite effects
344
in cats and rats. Microiontophoretic application of NE into the vicinity of cat lumbar spinal motoneurons consistently hyperpolarized these cells or decreased the excitability of the cells (Engberg and Ryall, 1966; Weight and Salmoiraghi, 1967; Phillis et al., 1968; Engberg and Marshall, 1971; Engberg et al., 1976; Marshall and Engberg, 1979). The hyperpolarization that accompanied NE application was rapid in onset and recovered at the offset of current application (Engberg and Marshall, 1971; Marshall and Engberg, 1979). Attempts to pharmacologically antagonize the hyperpolarization were, however, unsuccessful. Indeed, both a- and P-adrenergic antagonists mimicked the hyperpolarization produced by NE, suggesting that the hyperpolarization may be a nonspecific effect (Engberg et al., 1976). In apparently direct contradiction to the inhibitory effects of NE on cat spinal motoneurons, microiontophoretic application of NE enhanced extracellularly recorded firing of rat spinal and facial motoneurons and guinea pig trigeminal motoneurons (Barasi and Roberts, 1977; McCall and Aghajanian, 1979; White and Neuman, 1980, 1983; Katakura and Chandler, 1990). Furthermore, NE depolarized rat facial motoneurons recorded intracellularly and depolarized ventral roots of neonatal rat hemisected spinal cord preparations (Evans and Watkins, 1978; VanderMaelen and Aghajanian, 1980; Kitazawa et al., 1985; Neuman, 1985; Connell et al., 1989). The enhanced motoneuron excitability developed slowly during NE application, outlasted the offset of NE application and was attenuated or blocked by a-adrenergic antagonists. These contradictory findings in rats and cats may reflect a true species difference in motoneuron responses to NE. However, it is more likely that the contradiction represents differences in methodology and/or a difference in the portion of the total time course of the NE response that was chasen for analysis. The purpose of the studies presented here was to reinvestigate the effects of NE on cat spinal motoneurons with particular emphasis on determining whether NE may induce slowly developing
increases in excitability that resemble the slowly developing effects of NE on rat spinal motoneurons. Motoneuron recording Lumbar spinal cord motoneuron responses to microiontophoretically applied NE were examined in urethane-anesthetized rats and in decerebrate or pentobarbital-anesthetized cats. Motoneurons were identified by antidromic stimulation of the ventral root in rats (White and Neuman, 1980) and by antidromic stimulation of the ventral root or peripheral muscle nerve in cats (Fung and Barnes, 1987). Glass micropipettes (7-barrel) were used for extracellular recording from single motoneurons in combination with extracellular microiontophoresis. Intracellular recording combined with extracellular microiontophoresis was accomplished in cats using a "piggyback" electrode configuration (White and Fung, 1989). The single-barrel intracellular recording pipette protruded about 70 p m beyond the tip of the 5-barre1 extracellular microiontophoresis electrode. Micropipette barrels contained 3.0 M KCl for intracellular recording, 4.0 M NaCl for extracellular recording and automatic current neutralization, 0.1 M NE bitartrate in distilled water with 0.01 M ascorbic acid to retard oxidation (pH 4 - 4 3 , 0.1 M phenylephrine hydrochloride (pH 4-4.51, 0.2 M monosodium glutamate (pH 6.8) to drive rat motoneurons, and 0.01 M acetic acid in saline (pH 4) for acid control. The time course of NE effects on glutamateevoked motoneuron firing in the rat was compared to the time course of NE effects on motoneuron membrane potential and motoneuron excitability in the cat. Rat motoneurons were driven at low firing levels by cycled glutamate applications (5 sec on, 15 sec off). Following establishment of stable glutamate-evoked firing levels, effects of NE, phenylephrine (PHI and control solutions on glutamate-evoked firing were measured. In cat motoneurons, effects of NE and the acid control solution on resting membrane
345 NE 15 -n
Pn
25
n
PH
20 -
-
300T
PH 25
NE 20
0 G 80
G150
40
80
120
160
200
i 240
2 MIN
Time ( s e c )
Fig. 1. Effects of norepinephrine (NE) and phenylephrine (PHI on glutamate-evoked firing in three rat lumbar motoneurons. Glutamate (G) was cycled 5 sec on, 15 sec off throughout the records at ejection currents indicated at the arrowheads below the records. Ejection times and currents for NE and PH are indicated by the lines and numbers above the records. Records are oscillograph pen traces of cumulative action potentials during each G application cycle. Onset of each G ejection period reset the spike counter and the pen. (Note that the G ejection became subthreshold for eliciting spikes prior to N E application in the bottom left record.)
potentials, responses to intracellular depolarizing current pulses, and membrane input resistances were tested. All motoneurons included in the analysis had spike heights greater than 60 mV and resting membrane potentials greater than - 5 5 mV. Effects of NE on spinal motoneuron excitability
In the anesthetized rat, NE (10-20 nA, 60 sec) produced a slowly developing enhancement of
Fig. 2. Time course of changes in glutamate-evoked firing in 15 rat spinal motoneurons during and after N E application (10 nA, 60 sec). Means and standard errors are plotted. N E was applied at the line below the record. The number of action potentials elicited during each glutamate cycle during and after N E application was compared to the average number of spikes occurring during the three glutamate cycles immediately preceding NE onset for each cell. Glutamate was cycled 5 sec on, 15 sec off at a current sufficient for eliciting a stable low level of baseline firing in each cell.
glutamate-evoked motoneuron firing that outlasted the ejection period (Fig. l), confirming previous findings for rat spinal motoneurons (White and Neuman, 1980, 1983) and rat facial motoneurons (McCall and Aghajanian, 1979). The a-adrenergic agonist, PH mimicked the effect of NE (Fig. 1). Occasionally, motoneuron excitability was reduced following a period of PH-induced increased firing (Fig. 1, bottom), but N E still retained the ability to enhance cell firing at this time. Although the magnitude of the NE-induced
NE L 0 n A
-62 mV
*..**.................................. -c
10 s
Fig. 3. N E depolarization of spinal motoneurons in a pentobarbital-anesthetized cat. The records are from two different anterior biceps-semimembranosus (ABSM) motoneurons. Ejection periods for N E (40 nA, 60 sec) are indicated by the lines above the records. The membrane potential records were drawn by a Grass oscillograph pen. The small pen deflections that are superimposed on the membrane potential records resulted from short intracellular current pulses that were delivered at one Hz to test input resistance (top) and rheobase (bottom). The dotted lines indicate the pre-NE membrane potential level.
346
Fig. 4. Depolarizations induced by three successive NE applications in a decerebrate cat ABSM motoneuron. The record is an oscillograph pen tracing of the membrane potential. Periods of NE application are indicated by the bars below the record. The antidromic spike size for this ABSM cell was 75 mV.
excitability increase varied among motoneurons, NE produced a slowly developing increase in excitability that outlasted the period of current application in all cells that were tested (Fig. 2). Equivalent current applied to the acid control solution had no effect on motoneuron firing whereas high acid ejection currents (40 nA or greater) often decreased glutamate-evoked firing. Intracellular recordings from decerebrate or pentobarbital-anesthetized cats revealed that microiontophoretically applied NE (30-80 nA, 3060 sec) produced a slowly developing, small-amplitude depolarization of the motoneurons (Fig. 3, 4). This depolarization was replicated 37 times in 20 different motoneurons, and was observed in both decerebrate and pentobarbital-anesthetized cats. Most of the cells were anterior biceps-semimembranosus (ABSM, 8) motoneurons, but NE depolarized common peroneal (CP, 5 ) and posterior biceps-semitendinosus (PBST, 4) motoneurons, as well as motoneurons that were identified only by antidromic stimulation of the ventral root (3). Equivalent currents (30-80 nA, 30-60 sec) applied to the acid control solution had no effect on membrane potential or produced a slight hyperpolarization. During a few NE applications (about 15%), brief hyperpolarization appeared to precede the depolarization, but hyperpolarization alone was not observed. When initial short-lasting hyperpolarization did occur at the onset of NE ejection for a cell, other NE applications to the same cell produced only depolarization. Figure 4 shows responses to three successive NE applications in a
motoneuron with a less stable resting membrane potential than the cells in Figure 3. Hyperpolarization accompanied the onset of the third NE ejection period, but depolarization rapidly supervened. Depolarization was the consistently reproducible effect of NE on all of the motoneurons, whether or not initial hyperpolarization may have occurred. The slow onset and long duration of the NE-induced depolarizations in the cat spinal motoneuron closely resembled the slowly developing, long-lasting time course of NE effects on glutamate-evoked excitability of rat spinal motoneurons (Fig. 2). The NE-induced depolarization of rat spinal motoneurons appeared to be mediated by aadrenergic receptors because it was mimicked by PH and reduced by phentolamine (White and Neuman, 1983). Phentolamine and prazosin also antagonized NE-induced ventral root depolarizations in neonatal rat hemisected spinal cord preparations (Connell et al., 1989). Slow excitatory effects of NE and LC stimulation on lateral geniculate neurons were also attenuated by aadrenergic antagonists (Rogawski and Aghajanian, 1980). However, NE-induced depolarization of cat neocortical neurons appears to be mediated by P-adrenergic receptors (Foehring et al., 1989). The small-amplitude depolarizations that were produced by NE rarely brought the motoneurons to firing level, but were associated with increased responsiveness to other excitatory input. The amplitude of the intracellular depolarizing pulse that was required to evoke an action potential was reduced by NE in 11/11 cat spinal motoneurons. Increased excitability to intracellular depolarizing pulses developed slowly during NE application, became maximal at ejection offset, and persisted for several minutes following ejection offset (Fig. 5). Application of NE also mimicked the ability of small intracellular depolarizing pulses to convert partial or initial segment spikes evoked by antidromic stimulation to full soma-dendritic spikes (Fig. 6, top). Similarly, fractionation of antidromi-
347 ABSM Pre-NE
23s post-onset
25s post-off set
AMM
A
JL
Control
t2nA
F
75spost-onset 170s E NE(80nA.BOs) p o s t - o n s c l g A-D. 1 rns E-H.5rns
-l20rnV 5ms
Fig. 5. ABSM motoneuron responses to intracellular depolarizing pulses are enhanced by NE (40 nA, 35 sec). Each record is a photograph of 5 superimposed oscilloscope traces. Depolarizing pulses were delivered intracellularly (1 Hz) at current intensities indicated below each record. Prior to NE application, 7.5 nA pulses failed to fire the cell. Within 23 sec of NE ejection onset, each depolarizing pulse fired the cell. By 25 sec post-ejection offset, the cell fired immediately at the onset of each 7.5 nA depolarizing pulse, and 5.4 nA depolarizing pulses could now fire the cell. Excitability began to decrease by 7 min post-NE offset and had recovered to pre-NE levels by 10 min post-offset. (The bridge was partially unbalanced for the top three records, and the upward deflection at the end of each record is artifact at stimulus offset).
cally evoked full spikes to partial spikes by intracellular hyperpolarizing pulses could be overcome by NE application (Fig. 6, bottom). Conversion of partial spikes to full spikes by NE application was seen in 7/7 cells tested. Again, this effect of NE was slow to develop and was long-lasting. The small depolarization of the motoneurons that accompanied NE application could account for the ability of NE both to convert partial spikes to full spikes and to overcome the inhibitory effect of small intracellular hyperpolarizing pulses. As a further index of NE-induced increases in excitability, long intracellular depolarizing pulses that evoked one or two spikes prior to NE application were found to evoke repetitive firing during and after NE application in 4/4 cells tested. For the cell illustrated in Figure 7, multiple spikes first occurred 22 sec after the onset of NE ejection. Although spiking peaked just at ejection offset, excitability did not return to pre-NE level for several minutes. NE-induced repetitive firing has also been reported for cat sympathetic preganglionic neurons (Yoshimura et al., 19861, hip-
Control
-6.7nA
19s post-onset NE(40nA,60s )
220s post-onset
Fig. 6. NE converted partial or initial segment (IS) spikes to full soma-dendritic (SD) spikes. Each record is a photograph of 10 superimposed oscilloscope traces. A. Record from an ABSM motoneuron that responded to antidromic stimulation (1 Hz) with IS spikes. B. intracellular depolarizing pulses converted the IS spikes to full spikes. C. NE application also converted the IS spikes to full spikes. Full spikes began to appear by 30 sec post-NE onset and occurred to every antidromic stimulus by 15 sec post-NE offset. D. antidromic stimulation evoked a full spike in only 1/10 trials by 110 sec post-NE offset. E. Record from a different ABSM motoneuron that responded to antidromic stimulation with full spikes. F. The full spikes were fractionated to IS spikes by intracellular hyperpolarizing pulses. G. NE application converted the fractionated spikes to full spikes. H. By 160 sec post-NE offset, all the spikes were again fractionated by the hyperpolarizing pulses.
ABSM
Pre-NE (80nA. L 5 s )
5s post-offset
2 2 s post-onset
L 5 rnin post -offset
-//20rn" 200 ms
Fig. 7. NE induced repetitive firing to an intracellular depolarizing pulse (15 nA, 1 sec) that elicited only two spikes prior to NE application. Multiple spiking began 22 sec following the onset of NE ejection and became maximal at ejection offset. Motoneuron excitability remained elevated for more than 4 min post-NE ejection offset.
348 A
pocampal pyramidal cells (Madison and Nicoll, 19861, thalamic parataenial nucleus cells (McCormick and Prince, 1988) and neocortical pyramidal cells (Foehring el aL, 1989). All of these latter studies utilized slice preparations and the repetitive firing was shown to be associated with decreases in the slow component of the spike after-hyperpolarizing potential. Membrane input resistance was measured before and during NE-induced depolarization in two motoneurons by measuring voltage drops across the membrane to intracellular depolarizing pulses. The NE-induced depolarization was associated with an increase in input resistance in both cells, one of which is shown in Figure 8. Decreased conductance associated with a slow depolarization following NE application has previously been reported for facial motoneurons and thalamic neurons (VanderMaelen and Aghajanian, 1980; McCormick and Prince, 1988). 0 -
b -
‘L 10 m s
Fig. 8. NE-induced depolarization was accompanied by an increase in input resistance in an ABSM motoneuron. Top. Oscilloscope record of membrane potential. N E (40 nA) application is indicated by the line below the membrane potential record. Bottom. Intracellular depolarizing pulses (2.8 nA, 30 ms, 1 Hz) were applied and voltage drops across the membrane were measured. Sixteen consecutive traces of the voltage drop were averaged before (a) and after (b) NE application. The superimposed averaged traces are shown. The average input resistance following NE (b) was 30.8% greater than the input resistance during the control period (a). The bridge was underbalanced by the amount indicated by the arrow.
n5-
A
__
- 73
lmin
6
E 0
5HT 5 5 n A
30s
c
Fig. 9. Serotonin (SHT) effects on decerebrate cat spinal motoneurons. A. High-gain oscillograph pen tracing of membrane potential change produced by 5HT application indicated by the line above the record. B. Low-gain oscilloscope record of the same SHT application to illustrate the spikes that occurred in this cell near the end and after the 5HT ejection period. C. Input resistance increase produced in a different motoneuron by 5HT application (SO nA, 60 sec). Voltage drops to intracellular hyperpolarizing pulses (2 nA, 20 ms, 1 Hz) collected 20 sec before (a, control), 45 sec after (b, peak effect), and 100 sec after (c, recovery) 5HT application are shown. Eight consecutive traces were averaged for each of the three superimposed records. (Modified from White and Fung, 1989.)
NE effects on cat spinal motoneurons, e.g., slowly developing depolarization, increased cellular excitability, and increased input resistance, were similar to the effects of serotonin (5HT) on these cells (Fig. 9) as reported recently by White and Fung (1989). Similar observations have been made for both NE and 5HT in rat facial motoneurons in situ by VanderMaelen and Aghajanian (19801, in brainstem slices by Rasmussen and Aghajanian (19901, and in cat and guinea pig thalamus slices by McCormick and Prince (1988). Although the mechanisms by which NE and 5HT produce these effects on spinal motoneurons have not yet been identified, the observations in both cats and rats are consistent with the hypothesis that these monoamines decrease motoneuron potassium conductance (VanderMaelen and Aghajanian, 1982). Both NE and 5HT appear to
349
act within the vicinity of mammalian motoneurons to enhance motoneuron responsiveness to excitatory input.
Acknowledgements The authors wish to thank Jocelyn Penner and Jackie Raymond for careful technical assistance. This research was supported by NIH grant NS24388.
References Anden, N.E. (1965) Distribution of monoamines and dihydroxyphenylalanine decarboxylase activity in the spinal cord. Acta Physiol. Scand., 64: 197-203. Barasi, B. and Roberts, M.H.T. (1977) Responses of motoneurones to electrophoretically applied dopamine. Br. J. Pharm a d , 60: 29-34. Carlsson, A,, Falck, B., Fuxe, K. and Hillarp, N.-A. (1964) Cellular localization of monoamines in the spinal cord. Acta Physiol. Scand., 60: 112-119. Chan, J.Y.H., Fung, S.J., Chan, S.H.H. and Barnes, C.D. (1986) Facilitation of lumbar monosynaptic reflexes by locus coeruleus in the rat. Brain Res., 369: 103-109. Commissiong, J.W., Hellstrom, S.O. and Neff, N.H. (1978) A new projection from locus coeruleus to the spinal ventral columns: Histochemical and biochemical evidence. Brain Res., 148: 207-213. Connell, L.A., Majid, A. and Wallis, D.I. (1989) Involvement of al-adrenoreceptors in the depolarizing but not the hyperpolarizing responses of motorneurones in the neonate rat to noradrenaline. Neuropharmacology, 28: 1399-1404. Dahlstrom, A. and Fuxe, K. (1965) Evidence for the existence of monoamine neurons in the central nervous system. 11: Experimentally induced changes in the intraneuronal amine levels of the bulbospinal neuron systems. Acta Physiol. Scand., 64 (Suppl. 247): 1-36. Engberg. I. and Marshall, K.C. (1971) Mechanism of noradrenaline hyperpolarization in spinal cord motoneurons of the cat. Acta Physiol. Scand., 83: 142-144. Engberg, I. and Ryall, R.W. (1966) The inhibitory action of noradrenaline and other monoamines on spinal neurones. J. Physiol. (London), 185: 298-322. Engberg, I., Flatman, J.A. and Kadzielawa, K. (1976) Lack of specificity of motoneurone responses to microiontophoretically applied phenolic amines. Acfa Physiol. Scand., 96: 137- 139. Evans, R.H. and Watkins, J.C. (1978) Specific antagonism of excitant amino acids in the isolated spinal cord of the neonatal rat. Eur. .I Pharmacol., . 50: 123-129. Foehring. R.C., Schwindt, P.C. and Crill, W.E. (1989) Norepinephrine selectively reduces slow calcium- and sodiummediated potassium currents in cat neocortical neurons. J. Neurophysiol., 61: 245-256.
Fritschy, J.-M., Lyons, W.E., Mullen, C.A., Kosofsky, B.E., Molliver, M.E. and Grzanna, R. (1987) Distribution of locus coeruleus axons in the rat spinal cord: A combined anterograde transport and immunohistochemical study. Brain Res., 437: 176-180. Fung, S.J. and Barnes, C.D. (1981) Evidence of facilitatory coerulospinal action in lumbar motoneurons of cats. Brain Rex, 216: 299-311. Fung, S.J. and Barnes, C.D. (1987) Membrane excitability changes in hindlimb motoneurons induced by stimulation of the locus coeruleus in cats. Brain Res., 402: 230-242. Katakura, N. and Chandler, S.H. (1990) An iontophoretic analysis of the pharmacologic mechanisms responsible for trigeminal motoneuronal discharge during masticatory-like activity in the guinea pig. J. Neurophysiol., 63: 356-369. Kitazawa, T., Saito, K. and Ohga, A. (1985) Effects of catecholamines on spinal motoneurones and spinal reflex discharges in the isolated spinal cord of the newborn rat. Dec Brain Rex, 19: 31-36. Konishi, M. (1968) Fluorescence microscopy of the spinal cord of the dog, with special reference to the autonomic lateral horn cells. Arch. Histol. Jpn., 30: 33-44. Lai, Y-Y., Strahlendorf, H.K., Fung, S.J. and Barnes, C.D. (1989) The actions of two monoamines on spinal motoneurons from stimulation of the locus coeruleus in the cat. Brain Res., 484: 268-272. Lyons, W.E., Fritschy, .I.-M. and Grzanna, R. (1989) The noradrenergic neurotoxin DSP-4 eliminates the coerulospinal projection but spares projections of the A5 and A7 groups to the ventral horn of the rat spinal cord. J. Neurosci., 9: 1481-1489. Madison, D.V. and Nicoll, R.A. (1986) Actions of noradrenaline recorded intracellularly in rat hippocampal CA1 pyramidal neurones, in citro. J. Physiol. (London), 372: 221-244. Marshall, K.C. and Engberg, I. (1979) Reversal potential for noradrenaline-induced hyperpolarization of spinal motoneurons. Science, 205: 422-424. McCall, R.B. and Aghajanian, G.K. (1979) Serotonergic facilitation of facial motoneuron excitation. Brain Res., 169: 11-27. McCormick, D.A. and Prince, D.A. (1988) Noradrenergic modulation of firing pattern in guinea pig and cat thalamic neurons, in citro. J. Neurophysiol., 59: 978-996. Mizukawa, K. (1980) The segmental detailed topographical distribution of monoaminergic terminals and their pathways in the spinal cord of the cat. Anaf. Anz., 147: 125- 144. Neuman, R.S. (1985) Action of serotonin and norepinephrine on spinal motoneurones following blockade of synaptic transmission. Can. J. Physiol. Pharmacol., 63: 735-738. Phillis, J.W., Tebecis, A.K. and York, D.H. (1968) Depression of spinal rnotoneurones by noradrenaline, 5-hydroxytryptamine and histamine. Eur. J. Pharmacol., 4: 471-475. Rasmussen, K. and Aghajanian, G.K. (1990) Serotonin excitation of facial motoneurons: Receptor subtype characterization. Synapse, 5: 324-332. Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1989) Spinally projecting noradrenergic neurons of the dorsolat-
350 era1 pontine tegmentum: A combined immunocytochemical and retrograde labeling study. Brain Res., 491: 144-149. Rogawski, M.A. and Aghajanian, G.K. (1980) Modulation of lateral geniculate neurone excitability by noradrenaline microiontophoresis or locus coeruleus stimulation. Nature (London), 87: 731-734. Strahlendorf, J.C., Strahlendorf, H.D., Kingsley, R.E., Gintautas, J. and Barnes, C.D. (1980) Facilitation of the lumbar monosynaptic reflexes by locus coeruleus stirnulation. Neuropharmucology, 19: 225-230. VanderMaelen, C.P. and Aghajanian, G.K. (1980) Intracellular studies showing modulation of facial motoneurone excitability by serotonin. Nature, (London), 287: 346-347. VanderMaelen, C.P. and Aghajanian, G.K. (1982) Serotonininduced depolarization of rat facial motoneurons in cico: Comparison with amino acid transmitters. Brain Res., 239: 139-152. Weight, F.F. and Salmoiraghi, G.C. (1967) Motoneurone depression by norepinephrine. Nature (London), 213: 12291230. Westlund, K.N. and Coulter, J.D. (1980) Descending projec-
tions of the locus coeruleus and subcoeruleus/medial parabrachial nuclei in monkey: Axonal transport studies and dopamine-b-hydroxylase immunocytochemistry. Bruin Res. Rec;., 2: 235-264. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of the rat. Brain Res., 263: 15-31. White, S.R. and Fung, S.J. (1989) Serotonin depolarizes cat spinal rnotoneurons in situ and decreases motoneuron afterhyperpolarizing potentials. Brain Res., 502: 205-213. White, S.R. and Neuman, R.S. (1980) Facilitation of spinal motoneuron excitability by 5-hydroxytryptamine and noradrenaline. Brain Rex, 188: 119-127. White, S.R. and Neuman, R.S. (1983) Pharmacological antagonism of facilitatory but not inhibitory effects of serotonin and norepinephrine on excitability of spinal motoneurona Neuropharmacology , 22: 489-494. Yoshimura, M., Polosa, C. and Nishi, S. (1986) Noradrenaline modifies sympathetic preganglionic neuron spike and afterpotential. Bruin Res., 362: 370-374.
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
35 1 CHAPTER 27
Second messenger-mediated actions of norepinephrine on target neurons in central circuits: a new perspective on intracellular mechanisms and functional consequences B.D. Waterhouse, F.M. Sessler, W. Liu and C.-S. Lin Department of Physiology and Biophysics, Hahnemann Uniaersiry, Broad and Vine, Philadelphia, PA, U.S.A.
Ever since the initial demonstration of a widespread distribution of noradrenergic fibers to functionally diverse regions of the mammalian forebrain, there has been considerable interest in determining the electrophysiological effects of norepinephrine (NE) on individual neurons within these target areas. While early studies showed that N E could directly inhibit cell firing via increased intracellular levels of cyclic AMP, more recent work has revealed a spectrum of noradrenergic actions, which are more accurately characterized as neuromodulatory. More specifically, numerous experimental conditions have been described where NE at levels subthreshold for producing direct depressant effects on spontaneous firing can facilitate neuronal responses to both excitatory and inhibitory synaptic stimuli. The goal of this report is to review
recent evidence which suggests that the various modulatory actions of N E on central neurons result from the activation of different adrenoceptor-linked second messenger systems. In particular, we have focused on the candidate signal transduction mechanisms that may underlie NE’s ability to augment cerebellar and cortical neuronal responsiveness to GABAergic synaptic inputs. The consequences of such NE-induced changes in synaptic efficacy are considered not only with respect to their influences on feature extraction properties of individual sensory cortical neurons but also with regard to the potential impact such actions would have on the signal processing capabilities of a network of noradrenergically innervated cortical cells.
Key words: norepinephrine, target neurons, second messengers, GABA, phosphorylation, neuromodulation
Introduction Despite initial proposals that it acted as an inhibitory neurotransmitter at central synapses, the effects of norepinephrine (NE) on single cell activity and the means through which such effects can occur have continued as subjects of considerable investigation by numerous laboratories. Indeed, the mechanisms and consequences of NE’s
action on central neurons are two of several key issues essential to understanding the role of the central noradrenergic system in brain function. While many early studies combined extracellular single-unit recording with direct microiontophoretic application of NE and NE-like compounds in anesthetized whole animal preparations to characterize the potential synaptic actions of NE, more recent studies have employed
352
intracellular and patch-clamp techniques under a variety of in vitro conditions to address questions concerning the fundamental mechanisms underlying NE’s effects. During the course of this work NE has been found to alter several different parameters of neuronal membrane function including resting membrane potential, calcium-dependent potassium conductances and GABA-induced chloride flux. Many of these actions are believed to be mediated by specific adrenoceptors which are linked to various intracellular second messenger systems. While an in-depth review of this work is beyond the scope of this chapter, we will attempt to establish a link between the results of such mechanistic studies and data from other experiments designed to assess the potential influence of synaptically released NE on neuronal circuit function. In this regard it is important to remember that while elucidation of the mechanism is an important concern, it is equally significant to show that such processes can account for the spectrum of actions which have been observed under physiological conditions. Effects of NE on spontaneous and stimulus-evoked spike train activity Early investigations viewed NE as a putative neurotransmitter and were designed to determine its ability to increase or decrease the firing rate of individual neurons in noradrenergic terminal fields. For the most part, local application of NE by microiontophoresis or stimulation of the input pathway from the LC suppressed the spontaneous discharge of neurons in the cerebellum (Hoffer et al., 1971; 1973), cerebral cortex (Krynjevic and Phillis, 1963; Stone, 1973; Olpe et al., 1980), hippocampus (Segal and Bloom, 1974a,b), thalamus (Phillis et aZ., 1967; Nakai and Takaori, 19741, and hypothalamus (Miyahara and Oomura, 1982). These depressant responses were mimicked and blocked by a variety of adrenoceptor agonists and antagonists, respectively. Additional studies indicated that noradrenergic depressant
responses were mediated by p-receptor-induced increases in intracellular cyclic AMP (Siggins et aZ., 1971a,b). Taken together these findings reinforced the notion that NE might function primarily as an inhibitory transmitter at central synapses via an intracellular second messenger system. Despite these early findings, it is important to point out that the demonstration of an agent’s ability to suppress spontaneous firing rate in individual cells does not necessarily explain its role in the operation of a neuronal circuit nor does it mean that the lowest threshold physiological action of the substance has been determined. For this reason a strategy emerged whereby iontophoretically applied NE was interacted with neuronal responses to stimulation of non-noradrenergic afferent pathways. The goal of these studies was to approximate conditions where cells would be simultaneously receiving noradrenergic and non-noradrenergic inputs and determine the potential for NE to influence the synaptic transfer of information between neurons. In the first of these studies (Foote et al., 19751, responses of individual auditory cortical neurons to species-specific vocalizations were examined in awake monkeys before, during, and after NE microiontophoresis. Examination of post-stimulus time histograms revealed that during NE application, the background firing of these cells was suppressed more than auditory-evoked excitatory discharges. Similar results were obtained by Freedman et al. (1977) when iontophoretically applied NE was interacted with cerebellar Purkinje cell responses to mossy and climbing fiber pathway stimulation in anesthetized rats. Specifically, stimulus-induced excitations were preserved relative to a depression of spontaneous discharge such that there was a net enhancement of evoked responses. Likewise, stimulus-bound inhibition was augmented relative to changes in background firing. Thus, in both auditory cortex and cerebellum, NE exerted a differential depressant effect on evoked versus spontaneous discharge such that the “signal-to-noise” ratios of synaptically mediated responses were increased.
353
An in-depth analysis of this net facilitating effect of NE on synaptic transmission was subsequently carried out in cerebellum and cerebral cortex by Woodward and colleagues (1979) with the intent of characterizing the nature of this noradrenergic action at levels of synaptically or iontophoretically released NE which had minimal or no effect on spontaneous discharge. Under these conditions numerous instances were reported where synaptically mediated and transmitter-induced patterns of cell discharge were increased well above control levels during iontophoretic application of NE or noradrenergic pathway stimulation (Moises et al., 1979; 1983a; 1983b; Moises and Woodward, 1980; Waterhouse and Woodward 1980; Waterhouse et al., 1980a,b). More recently, a similar series of studies was conducted in the rat lateral hypothalamus (LH) by Waterhouse and colleagues (Cheng et al., 1988; Sessler et al., 1988) and revealed an identical array of NE-mediated facilitating actions on LH cell responsiveness to synaptic inputs and putative amino acid transmitters. These findings, in conjunction with those of others (Rogawski and Aghajanian, 1980a,b; Kasamatsu and Heggelund, 1982; Videen et al., 1984; Kossl and Vater, 19891, suggest that a prominent physiological function of central noradrenergic pathways might be to enhance the efficacy of non-monoamine excitatory and inhibitory synaptic transmission within target neuronal circuits rather than directly suppress cell firing. In primary sensory areas of the brain, such actions might lead to improved signal processing capabilities of the local circuitry and, as such, be beneficial to an organism during periods requiring increased vigilance or orientation to novel environmental stimuli (Aston-Jones, 1985). In addition to continuing efforts to determine the precise parameters of sensory neuronal function which are regulated by NE, other studies are beginning to focus on the specific receptor-linked signal transduction mechanisms which are responsible for mediating noradrenergic modulatory effects.
Noradrenergic potentiation of GABA responses
The demonstration of NE's ability to enhance neuronal responsiveness to GABA application (Moises et al., 1979; Waterhouse et al., 1980b; Yeh et al., 1981) and GABAergic pathway stimulation (Moises et al. 1983b, Sessler et al., 1988), via P-adrenergic receptor activation (Waterhouse et al., 1982; Yeh and Woodward, 1983; Cheng et al., 1988), prompted a series of experiments designed to evaluate the potential role of cyclic AMP as an intracellular mediator of GABA potentiation by extracellularly applied NE. In intact, anesthetized rats, iontophoretically applied 8-bromo-cyclic AMP (a membrane permeant analog of 3 ' 3 ' cyclic AMP) was found to potentiate GABA-induced inhibition of Purkinje cell discharge in a manner identical to that observed with NE (Sessler et al., 1989). Specificity of this effect for GABA was demonstrated by showing that a similar augmentation of transmitter-induced inhibition of Purkinje cell firing did not routinely occur when 8-bromo-cyclic AMP was interacted with p-alanine, an amino acid which, like GABA, is capable of hyperpolarizing neuronal membranes via increases in chloride conductance (Sessler et al., 1989). Other experiments (Sessler et al., 1989) conducted in cerebellar tissue slices have confirmed this result and also shown that agents which increase intracellular levels of cyclic AMP, i.e., forskolin (direct activator of adenyl cyclase) and IBMX (phosphodiesterase inhibitor) in combination with NE, can also augment GABA responses of presumed Purkinje neurons (see Fig. 1). Neither adenosine which could be a by-product of extracellular metabolism of 8-bromo-cyclic AMP nor 1,9-dideoxy-forskolin, which lacks the ability to activate adenyl cyclase, were capable of potentiating GABA responses. Collectively, these results provide evidence that NE augmentation of GABA efficacy in central neurons may be mediated by P-receptor-coupled increases in intracelM a r cyclic AMP.
354 TABLE 1 Evidence for noradrenergic potentiation of GABAergic responses Preparation
Brain region
Mode of NE application
Investigators
Acutely dissociated neurons (whole cell patch clamp)
Cerebellar Purkinje cells
Bath application
Yeh and Cheun, unpublished
Tissue slice (intracellular)
Layer IV/V pyramidal cells (somatosensory cortex)
Bath application, Microdrop
Sessler et al., 1990
Tissue slice (extracellular)
Cerebellar cortex Lateral hypothalamus
Iontophoretic
Cheng e f al., 1988 Sessler et al., 1988
Anesthetized, whole animal (extracellular)
Cerebellar Purkinje cells Lateral hypothalamus Cerebral cortex (somatosensory)
Iontophoretic, LC stimulation
Moises et al., 1979; 1983; Sessler et al., 1988; Waterhouse et al., 1980
Awake, freely moving animal (extracellular)
Cerebellar Purkinje cells
Iontophoretic
West and Woodward, unpublished
With regard to the physiological significance of these findings, it is important to point out that, to date, the phenomenon of potentiation of GABA responses by NE and related adrenergic agents has not only been observed in cerebellar Purkinje cells recorded from in uitro tissue slice preparations and intact, anesthetized animals (see Table 1) but also from awake, freely moving rats (West and Woodward, unpublished).
Noradrenergic influences on GABA-induced membrane conductance changes A question which emerges from the observation that iontophoretically or synaptically released NE can augment neuronal responses to extracellularly applied GABA is whether there are any well-established membrane actions of NE which can account for such effects. Experiments in hippocampus (Madison and Nicoll, 1986a,b), thalamus (McCormick and Prince, 19881, cerebellum (Siggins et a[., 1971b) and cerebral cortex (Foehring et al., 1989) have shown both depolarizing and hyperpolarizing effects of NE on neuronal membranes. Studies in hippocampus (Madison and Nicoll, 1986a,b), thalamus (McCormick and Prince, 1988) and cerebral cortex (Foehring et a[., 1989) have also demonstrated NE-induced clo-
sure of a calcium-activated potassium conductance which leads to blockade of accommodation and prolongation of depolarizing current-evoked excitatory discharges. This latter action could account for noradrenergic potentiation of threshold level excitatory synaptic responses, but is not adequate to explain NE-induced augmentation of GABAergic inhibitory responses. Likewise, the membrane hyperpolarization produced by NE is also an unlikely candidate mechanism since other agents which can hyperpolarize cell membranes like GABA and serotonin exert qualitatively different effects on central neuron responses to GABA (Yeh, 1982; Waterhouse et al., 1986). In our own laboratory we have been examining the transmembrane electrophysiological events which are associated with interactions between NE and GABA in layer IV/V pyramidal neurons of rat somatosensory cortex. Intracellular recordings were made from cortical cells in a submerged tissue slice preparation before, during, and after bath-application of GABA, NE and isoproterenol (ISO), alone or in combination. Superfusion of GABA (0.5-3 mM) produced small decreases in resting membrane potential associated with a reduction (22%) in membrane input resistance. NE and I S 0 also produced small membrane depolarizations (1-4 mV) but no concomitant changes in
355 NE 35nA
GABA 3 0 n A
IS0
Control
GABA 2 5 n A
GABA 12 nA
Control
Recovery
8 - B r o m o CAMP
..............................................
Forskolin 2 5 p M
............................. ..-..
NE
NE + IBMX
Recovery
Fig. I . Potentiation of GABA-induced inhibitory actions via activation of a P-receptor-linked second messenger system. Illustrated here are ratemeter (A,C,D) and perievent histogram (B,E) records obtained from individual rat cerebellar Purkinje cells under a variety of experimental conditions. In each case inhibitory responses to iontophoretic pulses (10 sec duration, 40 sec intervals) of GABA (solid bars) were potentiated well above control levels during concomitant administration of agents which are known to elevate or, in the case of IBMX, prevent the enzymatic breakdown of intracellular cyclic AMP. These changes in GABA responsiveness were observed in the absence of any direct effects on spontaneous firing rate. A. In oitro tissue slice; ionotophoretic norepinephrine (NE) 35 nA vs GABA 30 nA. B. Halothane-anesthetized intact animal; iontophoretic isoproterenol (ISO) 21 nA vs GABA 23 nA. C. I n ritro tissue slice; iontophoretic 8-bromo-cyclic AMP 2 nA vs GABA 25 nA. D. In citro tissue slice; bath application of forskolin 25 FM vs GABA 12 nA. E. I n uitro tissue slice; iontophoretic NE 10 n A + IBMX 20 nA (phosphodiesterase inhibitor) vs GABA 50 nA. In B and E each histogram sums unit activity during an equivalent number of GABA applications.
membrane conductance. Simultaneous application of NE and GABA potentiated amino acid changes in input resistance in 36% of the cases tested. However, when the a blocker, phentol-
amine, was added to the medium (10-20 pM), NE-induced enhancement of the GABA response was observed in 83% of the cells examined; suggesting a potential masking of this effect via a-receptor activation. Consistent with this interpretation was the finding that the P-agonist, ISO, produced net increases in GABA-induced conductance changes in the majority (92%) of cases tested (see example, Fig. 2). The potentiating effect of N E and I S 0 on GABA-induced membrane responses was mimicked by the adenyl cyclase activator, forskolin, and the membrane permeant analog of cyclic AMP, 8-bromo-cyclic AMP (see Fig. 2). Furthermore, identical facilitating interactions were observed when the GABA, specific agonist, muscimol, was substituted for GABA. This latter result suggests two conclusions concerning the noradrencrgic effect on GABA responsiveness: (1) it does not involve GABA reuptake mechanisms, since muscimol is not a substrate for this process and (2) it is specific for the GABA, receptor since muscimol is not an agonist at the CABA, receptor site. Taken together these findings suggest that noradrenergic potentiation of GABA inhibition may be mediated by preceptor and cyclic AMP-linked effects on mechanisms which regulate GABA, receptor-induced changes in membrane conductance. A second important consideration in these experiments has been the identity of the neurons studied. Accordingly, intracellularly recorded cells whose membrane responses to GABA can be enhanced by adrenergic agonists have been classified as “regular spiking” and “bursting” according to electrophysiological criteria established previously by McCormick et al. (1985). In addition, a number of these neurons have also been morphologically identified as layer V pyramidal cells by intracellular injection of Lucifer yellow or biocytin (see Fig. 3). The goal of this aspect of the work is to correlate the observed membrane actions of N E with specific cellular components of an identified neocortical circuit.
355 NE 35nA
GABA 3 0 n A
IS0
Control
GABA 2 5 n A
GABA 12 nA
Control
Recovery
8 - B r o m o CAMP
..............................................
Forskolin 2 5 p M
............................. ..-..
NE
NE + IBMX
Recovery
Fig. I . Potentiation of GABA-induced inhibitory actions via activation of a P-receptor-linked second messenger system. Illustrated here are ratemeter (A,C,D) and perievent histogram (B,E) records obtained from individual rat cerebellar Purkinje cells under a variety of experimental conditions. In each case inhibitory responses to iontophoretic pulses (10 sec duration, 40 sec intervals) of GABA (solid bars) were potentiated well above control levels during concomitant administration of agents which are known to elevate or, in the case of IBMX, prevent the enzymatic breakdown of intracellular cyclic AMP. These changes in GABA responsiveness were observed in the absence of any direct effects on spontaneous firing rate. A. In oitro tissue slice; ionotophoretic norepinephrine (NE) 35 nA vs GABA 30 nA. B. Halothane-anesthetized intact animal; iontophoretic isoproterenol (ISO) 21 nA vs GABA 23 nA. C. I n ritro tissue slice; iontophoretic 8-bromo-cyclic AMP 2 nA vs GABA 25 nA. D. In citro tissue slice; bath application of forskolin 25 FM vs GABA 12 nA. E. I n uitro tissue slice; iontophoretic NE 10 n A + IBMX 20 nA (phosphodiesterase inhibitor) vs GABA 50 nA. In B and E each histogram sums unit activity during an equivalent number of GABA applications. membrane conductance. Simultaneous applica-
tion of NE and GABA potentiated amino acid changes in input resistance in 36% of the cases tested. However, when the a blocker, phentolmembrane conductance. Simultaneous application of NE and GABA potentiated amino acid changes in input resistance in 36% of the cases tested. However, when the a blocker, phentol-
amine, was added to the medium (10-20 pM), NE-induced enhancement of the GABA response was observed in 83% of the cells examined; suggesting a potential masking of this effect via a-receptor activation. Consistent with this interpretation was the finding that the P-agonist, ISO, produced net increases in GABA-induced conductance changes in the majority (92%) of cases tested (see example, Fig. 2). The potentiating effect of N E and I S 0 on GABA-induced membrane responses was mimicked by the adenyl cyclase activator, forskolin, and the membrane permeant analog of cyclic AMP, 8-bromo-cyclic AMP (see Fig. 2). Furthermore, identical facilitating interactions were observed when the GABA, specific agonist, muscimol, was substituted for GABA. This latter result suggests two conclusions concerning the noradrencrgic effect on GABA responsiveness: (1) it does not involve GABA reuptake mechanisms, since muscimol is not a substrate for this process and (2) it is specific for the GABA, receptor since muscimol is not an agonist at the CABA, receptor site. Taken together these findings suggest that noradrenergic potentiation of GABA inhibition may be mediated by preceptor and cyclic AMP-linked effects on mechanisms which regulate GABA, receptor-induced changes in membrane conductance. A second important consideration in these experiments has been the identity of the neurons studied. Accordingly, intracellularly recorded cells whose membrane responses to GABA can be enhanced by adrenergic agonists have been classified as “regular spiking” and “bursting” according to electrophysiological criteria established previously by McCormick et al. (1985). In addition, a number of these neurons have also been morphologically identified as layer V pyramidal cells by intracellular injection of Lucifer yellow or biocytin (see Fig. 3). The goal of this aspect of the work is to correlate the observed membrane actions of N E with specific cellular components of an identified neocortical circuit.
357
rect actions via an intracellular second messenger system. Examination of the amino acid sequences for the various GABA, receptor subunits has revealed several potential sites for regulation of GABA, receptor function. One particularly intriguing idea is that the @-subunitof the receptor complex includes an intracellularly directed loop which contains a consensus site for protein phosphorylation (Schofield et al., 1987). Several laboratories (Kirkness et al., 1989; Browning et al., 1990) have, in fact, demonstrated under in vitro conditions that the GABA, receptor can be phosphorylated by a cyclic AMP-dependent kinase. Since noradrenergic activation of the @-receptor can result in intracellular increases in cyclic AMP, it appears that a means for phosphorylating the GABA, receptor and, thus, effecting a change in its function via an extracellular chemical signal is possible. A similar mechanism for regulating GABA receptor function has been proposed by Sweetnam er al. (1988) who have described the potential for phosphorylation of the a-subunit of the GABA, receptor complex by a receptor associated protein kinase. In this case, however, the kinase does not appear to be cyclic AMP-dependent. Overall, these findings suggest that the GABA, receptor may have multiple phosphorylation sites, each of which could serve a regulatory function subject to second messenger control. Although these results provide an intriguing set of possibilities, the physiological significance of such second messenger-initiated actions on the GABA, receptor remains to be established. In one recent experiment (Heuschneider and Schwartz, 1989) 8-bromo-cyclic AMP and forskolin were found to block GABA-induced efflux of chloride ions from synaptoneurosomes (intact pre- and postsynaptic membranes prepared from rat cerebral cortex). The implication of this data is that a cyclic AMP-dependent action, possibly phosphorylation, would lead to a reduced capacity of GABA to open chloride channels. Another
study in cultured neurons has shown that diffusion of the catalytic subunit of the A kinase into cells through a patch-clamp electrode also decreases GABA-induced chloride conductance (Porter et a/., 1989). Here again the implication is that activation of the A kinase initiates an action which reduces GABA efficacy. Other in vitro electrophysiological studies have demonstrated that phosphorylating factors are necessary for preventing “run down” of whole cell current responses to GABA over the course of repeated amino acid applications (Gyenes et al., 1988; Stelzer et al., 1988). While evidence of “run down” of GABA, receptor efficacy in the absence of phosphorylating factors implies a maintenance function for protein phosphorylation of this receptor, it also raises the possibility that under physiological conditions the GABA, receptor exists in a dynamic functional state which can be up- or down-regulated through rapidly responding biochemical mechanisms. Our contention is that N E may be one of perhaps several endogenous substances capable of enhancing the efficacy of GABAergic synaptic transmission via a cyclic AMP-dependent phosphorylation reaction. Functional consequences of NE enhancement of GABAergic synaptic efficacy
Aside from establishing the mechanism(s) responsible for NE enhancement of GABA efficacy is the issue of the impact of such an action on neurons and circuits receiving noradrenergic projections. Recently published reports by Moises et al. (1990) and Waterhouse et al. (1990) describe the effects of iontophoretically applied NE on the complex responses of parafloccular Purkinje cells and visual cortical area 17 neurons to stimulation of the visual field with moving bars of light. One frequently observed effect of NE in both studies was the enhancement of inhibitory troughs (“inhibitory side bands”) flanking periods of stimulus-evoked excitation (see example, Fig. 4). The net effect of this action was a more sharply defined transition between stimulus-induced inhibi-
358
Control
I
M
SHUT
. . . . . . .. .. . .. . .
.,:
_ ' '
. .. .. . . . .. ... . .... .. . .. . .. . .
.
..
.
I
SPIKES/S
183
. .
, ;..... ':,... .. .<j::...... ... ...j... :..;.,.,. : . . .......... .. ..... .. .
753deg/s
-
............ 0.0
12 01,
Seconds
60
IS0 9 n A
107-
00
Seconds
45%
-
60
Recovery
. . ..... . . . . .... . .. . ... .... . - ... .. .... .... ........ .. . .. .... ..... ... .... ... .... . . .... . . . . . .. .. .. . . . ... . .... .. .. .. . .,. . ... . . . .. .. . .. . . . . . . .. . ... . .. I .
....
13.3
00
I'
'
Seconds
60
Fig. 4. ISO-induced augmentation o f an inhibitory side hand. Cumulative raster and perievent histograni records obtained from an area 17 visual cortical neuron in a halothaneanesthetized rat illustrate the responses of the cell to a moving bar of light before, during, and after microiontophoretic application of I S 0 9 nA. The uppermost line depicts the opening (step-up) and closing (step-down) of the shutter controlling the stimulus display. The trapezoid waveform indicates the position of the stimulus in visual space as it moves in the temporal (T) to nasal (N) direction and back along the same path in the reverse direction. Time is indicated along the horizontal axis of the histograms and discharge frequency of the cell indicated along the vertical axis. Each histogram sunis unit activity during 15 trials; bin width = IS msec. The control response (upper panel) of the cell to movement (75.3 deg/sec) of a horizontally oriented bar of light (shown at left) was an increase in firing rate preceded by inhibition as the stimulus traversed (75.3) the visual field. The inhibitory component of the control response (broken bar beneath histogram) was quantitatively expressed as a 12% suppression of spontaneous firing rate (dotted line beneath histogram). During I S 0 iontophoresis (middle panel) the stimulus-bound inhibition was augmented from the control level (12-45%) with little concomitant change in background firing rate. Recovery (lower panel) to the control pattern of response was seen after termination of the I S 0 ejection current.
tion and excitation which could be interpreted as an enhancement of the cell's ability to detect stimulus movement across its receptive field borders. An additional observation in these studies was that NE was capable of revealing inhibitory responses to otherwise subthreshold synaptic inputs. in these circumstances visual stimulusevoked inhibition of cell firing was only evident during concomitant iontophoretic administration of NE (cf. Fig. 6, Waterhouse et al., 1990). Such results suggest that while inhibitory afferent signals may normally arrive at central synapses, they may go undetected by postsynaptic neurons unless facilitated by NE. The presence of GABA-containing neurons in the cortical circuitry and the influence of iontophoretic GABA on the receptive field properties of sensory cortical neurons is well documented (Sillito, 1975, 1977; Lin et al., 1986). Thus fluctuations in the output of noradrenergic neurons as might occur during changing behavioral states (Aston-Jones and Bloom, 1981a,b) could regulate receptive field properties of sensory cortical neurons through an influence on GABA efficacy. While it is at present difficult to predict the outcome of such effects on an ensemble of cortical neurons processing the same sensory afferent signal, the possibility must now be considered that synaptically released NE could transiently improve the feature extraction properties of primary sensory circuits within the brain to the extent that the perception of sensory stimuli could be heightened. Mechanisms underlying NE interactions with excitatory synaptic transmission A recent report of the effects of N E on transmembrane electrophysiological properties of cerebrocortical neurons (Foehring et al., 1989) describes not only noradrenergically induced decreases in afterhyperpolarization leading to blockade of accommodation but also a consistent NE-mediated slow depolarizing (3-5 mV) shift in
359
resting membrane potential associated with no change in membrane conductance. Similar effects of NE on hippocampal pyramidal (Madison and Nicoll, 1986a) and thalamic (McCormick and Prince, 1988) neurons have been described previously. For hippocampal pyramidal neurons, the observed decrease in afterhyperpolarization was found to be due to a reduction in calcium-dependent potassium conductance. As was described earlier, such actions could reasonably account for the NE-mediated potentiations of threshold level excitatory synaptic responses which have been observed under in uizw conditions. In each of these studies the membrane actions elicited by NE were blocked by @-antagonistsand mimicked by application of I S 0 and 8-bromocyclic AMP. These latter findings suggest that cyclic AMP is the intracellular second messenger responsible for mediating actions leading to increased neuronal responsiveness to excitatory synaptic inputs and putative transmitter substances. At present the means through which cyclic AMP would bring about such a change in neuronal excitability have not been elucidated. In ciz9o studies of rat somatosensory cortical neurons have shown that noradrenergic facilitation of excitatory responses to afferent synaptic inputs and putative transmitter substances can be mimicked and blocked by a-receptor agonists and antagonists, respectively (Waterhouse et al., 1981). Similar results have been reported for cells recorded from the dorsal lateral geniculate of intact, anesthetized rat (Rogawski and Aghajanian, 1980a). Recent in vitro studies of somatosensory cortical neurons in our own laboratory (Mouradian et al., 1989) have confirmed this pharmacology and, furthermore, have demonstrated that 8-bromo-cyclic AMP is incapable of mimicking such noradrenergic facilitating actions on neuronal excitability. Thus, in contrast to other results obtained in vitro, our most recent findings are consistent with the pharmacology observed under in viuo conditions and suggest that the facilitating effects of NE on somatosensory cortical neuron responses to excitatory synaptic events
do not involve cyclic AMP but rather rely upon a different cascade of intracellular events initiated by activation of a-type adrenoceptors. Impact of synaptically released NE on functional neuronal arrays
The demonstration that NE can regulate central neuronal responsiveness to afferent synaptic inputs through separate receptors and the likelihood that multiple intracellular biochemical processes are involved, provides several insights into the biophysical mechanisms through which such regulation occurs, but also raises many more questions concerning the potential sites for physiological control of this modulatory system. One important consideration is that receptor-linked second messenger systems are capable of producing numerous changes in cell physiology via the phosphorylation of multiple, specific intracellular substrate proteins (Nestler and Greengard, 1984). Increased intracellular cyclic AMP, for example, can also lead to phosphorylation and subsequent desensitization of the @-receptorcomplex (Sibley et al., 1986). Thus, the net physiological response of a neuron to an activated second messenger system is ultimately determined by the cellular expression and function of its target proteins. Differential expression of ion channel proteins, receptor subunits and intracellular enzymes could, therefore, explain the variety of noradrenergic actions which have been observed in different brain regions (Rogawski and Aghajanian, 1980a; Waterhouse et al., 1981; 1982; Mueller et al., 1982; B a d e and Dunwiddie, 1984; Parfitt et at., 1988). Such concerns re-emphasize the need to identify the neurons capable of specific responses to NE in complex circuits such as those within the neocortex. As reviewed elsewhere (Siggins and Gruol, 19861, the cell types which are targets of noradrenergic innervation, the cellular location of NE synapses, the receptor complement of target cells and the nature of their receptor-linked signal transduction mechanisms are all critical for
360 Coerulo- cortical innervation
Y
F-l
Target neurons 7 Location of NE synapses? Receptor complement Signal t ransduction I t n kages 7
NE Fig. 5. Hypothetical scheme indicating the potential cellular and subcellular distribution of noradrenergic synapses within the cerebrocortical circuitry. Abbreviations: P, layer V pyramidal cell; p, layer II/III pyramidal cell; s, spiny stellate cell; AS, aspiny stellate cell.
determining the net response of a circuit to synaptically released NE. As illustrated in Figure 5, the contribution of NE to the afferent signal processing capabilities of cerebrocortical circuits depends upon which circuit elements are targets of noradrenergic innervation and what type of noradrenergic response they can elicit. While such information is currently unavailable for cerebral cortex, it is clear that differential influences of NE on sub-classes of cortical neurons, e.g., stellate versus pyramidal, would lead to qualitatively different predictions about the role of the noradrenergic system in cerebrocortical function. Moreover, it is now obvious that there are multiple sites where aberrations in the synthesis of specific proteins involved in a signal transduction pathway could lead to physiological deficits. Thus, identification of the intracellular cascade of events associated with specific NE actions in target neuronal circuits could facilitate the search for the molecular determinants of those genetically based neurological (Waterhouse, 1986) and behavioral (Hornykiewicz, 1986) disorders which may involve the noradrenergic system. Acknowledgements
The work described in this report was supported by AFOSR-87-0138. BDW is also the recipient of
an NIH Research Career Development Award, NS 01233. The authors would like to thank Ms. Joy Hudson for typing the manuscript. References Aston-Jones, G. (1985) Behavioral functions of locus coeruleus derived from cellular attributes. Physiol. Psychol., 13: 118126. Aston-Jones, G. and Bloom, F.E. (1981a) Activity of norepinephrine containing locus coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle. J. Neurosci., 1: 876-886. Aston-Jones, G. and Bloom, F.E. (1981b) Norepinephrinecontaining locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Basile, A S . and Dunwiddie, T.V. (1984) Norepinephrine elicits both excitatory and inhibitory responses from Purkinje cells in the in vitro rat cerebellar slice. Brain Rex, 296: 15-25. Browning, M.D., Bureau, M., Barnes, E. and Olsen, R.W. (1990) Protein kinase C and CAMP-dependent protein kinase phosphorylate the P-subunit of the purified GABA, receptor. Proc. Natl. Acad. Sci. USA, 87: 13151318. Cheng, J-T., Sessler, F.M., Azizi, S.A., Chapin, J.K. and Waterhouse, B.D. (1988) Electrophysiological actions of norepinephrine in rat lateral hypothalamus: 11. An in Litro study of the effects of iontophoretically applied norepinephrine on LH neuronal responses to y-aminobutyric acid (GABA). Brain Rex, 446: 90-105. Foehring, R.C., Schwindt, P.C. and Crill, W.E. (1989) Norepinephrine selectively reduces slow Ca2+- and Na +-mediated K, currents in cat neocortical neurons. X Neurophysiol., 61: 245-256 Foote, S.L., Freedman, R. and Oliver, A.P. (1975) Effects of putative neurotransmitters on neuronal activity in monkey auditory cortex. Brain Res., 86: 229-42. Freedman, R., Hoffer, B.J., Woodward, D.J. and Puro, D. (1977) Interaction of norepinephrine with cerebellar activity evoked by mossy and climbing fibers. Exp. Neurol., 55: 269-288. Gyenes, M., Farrant, M. and Farb, D.H. (1988) “Run-Down’’ of y-Aminobutyric acid, receptor function during wholecell recording: A possible role for phosphorylation. Am. SOC.Pharmacol. Exp. Ther., 34: 719-723. Heuschneider, G. and Schwartz, R. (1989) CAMPand forskolin decrease y-aminobutyric acid-gated chloride flux in rat brain synaptoneurosomes. Proc. Natl. Acad. Sci. USA, 86: 2938-2942. Hoffer, B.J., Siggins, G.R. and Bloom, F.E. (1971) Studies on norepinephrine-containing afferents to Purkinje cells of rat cerebellum. 11. Sensitivity of Purkinje cells to norepinephrine and related substances administered by microiontophoresis. Brain Res., 25: 522-34. Hoffer, B.J., Siggins, G.R., Oliver, A.P. and Bloom, F.E. (1973) Activation of the pathway from locus coeruleus to
361 rat cerebellar Purkinje neurons: Pharmacological evidence of noradrenergic central inhibition. J. Phurmucol. Exp. Ther., 184: 553-569. Hornykiewicz, 0. (1986) Brain noradrenaline and schizophrenia. In J.M. van Ree and S. Matthysse (Eds.), Psychiatric Disorders: Neurotransmitters and Neuropeptides, Progress in Bruin Research, Vol. 65, Elsevier, Amsterdam, pp. 29-39. Kasamatsu, T. and Heggelund, P. (1982) Single cell responses in cat visual cortex to visual stimulation during iontophoresis of noradrenaline. Exp. Bruin Rex, 45: 317-327. Kirkness, E.F., Bovenkerk, C.F., Ueda, T. and Turner, A.J. (1989) Phosphorylation of y-aminobutyrate (GABA)/benzodiazepine receptors by cyclic AMP-dependent protein kinase. Biochem. J., 259: 613-616. Kossl, M. and Vater, M. (1989) Noradrenaline enhances temporal auditory contrast and neuronal timing precision in the cochlear nucleus of the mustached bat. J. Neurosci., 9: 4169-4178. Krynjevic, K. and Phillis, J.W. (1963) Actions of certain amines on cerebral cortical neurones. Br. J. Phurmucol., 2 0 471490. Lin, C.-S., Lu, S.M. and Schmechel, D.E. (1986) Glutamic acid decarboxylase and somatostatin immunoreactivities in rat visual cortex. J. Cornp. Neurol., 244: 369-383. Madison, D.V. and Nicoll, R.A. (1986a) Actions of noradrenaline recorded intracellularly in rat hippocampal CAI pyramidal neurones, in uitro. J. Physiol. (London), 372: 221-244. Madison, D.V. and Nicoll, R.A. (1986b) Cyclic adenosine 3',5'-monophosphate mediates preceptor actions of noradrenaline in rat hippocampal pyramidal cells. J. Physiol. (London), 373: 245-259. McCormick, D.A. and Prince, D.A. (1988) Noradrenergic modulation of firing pattern in guinea pig and cat thalamic neurons, in citro. J. Neurophysiol., 59: 978-996. McCormick, D.A., Connors, B.W., Lighthall, J.W. and Prince, D.A. (1985) Comparative electrophysiology of pyramidal and sparsely spiny stellate neurons of the neocortex. J. Neurophysiol., 54: 782-806. Moises, H.C. and Woodward, D.J. (1980) Potentiation of GABA inhibitory action in cerebellum by locus coeruleus stimulation. Bruin Rex, 182: 327-344. Moises, H.C., Woodward, D.J., Hoffer, B.J. and Freedman, R. (1979) Interactions of norepinephrine with Purkinje cell responses to putative amino acid neurotransmitters applied by microiontophoresis. Exp. Neurol., 64: 489-515. Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1983a) Locus coeruleus stimulation potentiates Purkinje cell responses to afferent input: The climbing fiber system. Bruin Res., 222: 43-64. Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1983b) Locus coeruleus stimulation potentiates local inhibitory processes in rat cerebellum. Bruin Res. Bull., 10: 795-804. Moises, H.C., Burne, R.A. and Woodward, D.J. (1990) Modification of the visual response properties of cerebellar neurons by norepinephrine. Bruin Res., 514: 259-275. Mouradian, R.D., Sessler, F.M. and Waterhouse, B.D. (1989) Noradrenergic modulation of cortical neuronal excitability:
Pharmacologic characterization in terms of adrenergic and amino acid receptor subtypes. Soc. Neurosci. Abstr., 15: 1012 Mueller, A.L., Palmer, M.R., Hoffer, B.J. and Dunwiddie, T.V. (1982) Hippocampal noradrenergic responses in oiuo and in 6:itro characterization of alpha and beta components. Nuunyn Schmiedebergs Arch. Phurrnucol., 318: 259266. Miyahara, S. and Oomura, Y. (1982) Inhibitory action of the ventral noradrenergic bundle on the lateral hypothalamic neurons through alpha noradrenergic mechanisms in the rat. Bruin Res., 234: 459-463. Nakai, Y. and Takaori, S. (1974) Influence of norepinephrine-containing neurons derived from the locus coeruleus on lateral geniculate neuronal activities of cats. Bruin Res., 71: 47-60. Nestler, E.J., Greengard, P. (1984) Protein Phosphorylation in the Nervous System. John Wiley & Sons, New York: 398 PP. Olpe, H., Glatt, A., Lazlo, J. and Schellengerg, A. (1980) Some electrophysiological and pharmacological properties of the cortical, noradrenergic projection of the locus coeruleus in the rat. Bruin Res., 186: 9-19. Olsen, R.W. and Tobin, A.J. (1990) Molecular biology of GABA, receptors. FASEB J., 4: 1469-80. Parfitt, K.D., Freedman, R. and Bickford-Wimer, P.C. (1988) Electrophysiological effects of locally applied noradrenergic agents at cerebellar Purkinje neurons: Receptor specificity. Bruin Res., 462: 242-51. Phillis, H.W., Tebecis, A.K. and York, D.H. (1967) The inhibitory action of monoamines on lateral geniculate neurones. J. Physiol. (London), 190: 563-581. Porter, N.M., Twyman, R.E., Uhler, M.D. and Macdonald, R.L. (1989) Cyclic AMP dependent protein kinase decreases GABA, receptor current in mouse spinal neurons. Sac. Neurosci. Abstr., 15: 1150. Rogawski, M.A. and Aghajanian, G.K. (1980a) Activation of lateral geniculate neurons by norepinephrine: Mediation by an alpha-adrenergic receptor. Bruin Res., 182: 345-359. Rogawski, M.A. and Aghajanian, G.K. (1980b) Norepinephrine and serotonin: Opposite effects on the activity of lateral geniculate neurons evoked by optic pathway stimulation. Exp. Neurol., 69: 678-684. Schofield, P.R., Darlison, M.G., Fujita, N., Burt, D.R., Stephenson, F.A., Rodriguez, H., Rhee, L.M., Rarnachandran, J., Reale, V., Glencorse, T.A., Seeburg, P.H. and Barnard, E.A. (1987) Sequence and functional expression of the GABAA receptor shows a ligand-gated receptor super-family. Nature (London), 328: 221 -227. Segal, M. and Bloom, F.E. (1974a) The action of norepinephrine in the rat hippocampus. I. Iontophoretic studies. Bruin Res., 107: 513-525. Segal, M. and Bloom, F.E. (1974b) The action of norepinephrine in the rat hippocampus. 11. Activation of the input pathway. Bruin Res., 107: 99-114. Sessler, F.M., Cheng, J-T. and Waterhouse, B.D. (1988) Electrophysiological actions of norepinephrine in rat lateral hypothalamus. I. Norepinephrine induced modulation of
362 LH neuronal responsiveness to afferent synaptic inputs and putative neurotransmitters. Brain Res., 446: 77-89. Sessler, F.M., Mouradian, R.D., Cheng, J.-T., Yeh, H.H., Liu, Weimin and Waterhouse, B.D. (1989) Noradrenergic potentiation of cerebellar Purkinje cell responses to GABA: Evidence for mediation through the P-adrenoceptor-coupled cyclic AMP system. Brain Res., 499 27-38. Sibley, D.R., Strasser, R.H., Benovic, J.L., Daniel, K. and Lefkowitz, R.L. (1986) Phosphorylation/dephosphorylation of the beta-adrenergic receptor regulates it functional coupling to adenylate cyclase and subcellular distribution. Proc. Natl. Acad. Sci. USA, 83: 9408-9412. Siggins, G.R. and Gruol, D.L. (1986) Mechanisms of transmitter action in the vertebrate central nervous system. In F.E. Bloom (Ed.), Handbook of Physiology. Vol. 4, American Physiological Society, Bethesda, MD., pp. 1-1 14. Siggins, G.R., Hoffer, B.J. and Bloom, F.E. (1971a) Studies on norepinephrine-containing afferents to Purkinje cells of rat cerebellum 111. Evidence for mediation of effects by cyclic 3',5'-adenosine monophosphate. Brain Rex, 25: 535-553. Siggins, G.R., Oliver, A.P., Hoffer, B.J. and Bloom, F.E. (1971b) Cyclic adenosine monophosphate and norepinephrine: Effects on transmembrane properties of cerebellar Purkinje c~lls.Science, 171: 192-194. Sillito, A.M. (1975) The contribution of inhibitory mechanisms to the receptive field properties of neurons in the striate cortex of the cat. J. Physiol. (London), 250: 305-329. Sillito, A.M. (1977) Inhibitoiy processes underlying the directional specificity of simple, complex and hypercomplex cells in the cat's visual cortex. J. Physiol. (London), 271: 305-329. Stelzer, A,, Kay, A.R. and Wong, R.K.S. (1988) GABAA-receptor function in hippocampal cells is maintained by phosphorylation factors. Science, 241: 339-341. Stone, T.W. (1973) Pharmacology of pyramidal tract cells in the cerebral cortex. Naunyn Schmiedebergs Arch. PharmaC O ~ . ,278: 333-346. Sweetnam, P.M., Lloyd, J., Gallombardo, P., Malison, R.T., Gallager, D.W., Tallman, J.F. and Nestler, E.J. (1988) Phosphorylation of the GABA, /benzodiazepine receptor (Y subunit by a receptor-associated protein kinase. J. Neurochem., 51: 1274-1288. Videen, T.O., Daw, N.W. and Rader, R.K. (1984) The effect of norepinephrine on visual cortical neurons in kittens and adult cats. J. Neurosci., 4: 1607-1617. Waterhouse, B.D. (1986) Electrophysiological assessment of
monoamine synaptic function in neuronal circuits of seizure susceptible brains. Life Sci., 39: 807-818. Waterhouse, B.D. and Woodward, D.J. (1980) Interaction of norepinephrine with cerebrocortical activity evoked by stimulation of somatosensory afferent pathways in the rat. Exp. Neurol. 67: 11-34. Waterhouse, B.D., Moises, H.C. and Woodward, D.J. (1980a) Locus coeruleus stimulation potentiates somatosensory cortical neuronal responses to afferent synaptic inputs. Soc. Neurosci. Abstr., 6: 448. Waterhouse, B.D., Moises, H.C. and Woodward, D.J. (1980b) Noradrenergic modulation of somatosensory cortical neuronal responses to iontophoretically applied putative neurotransmitters. Exp. Neurol., 69: 30-49. Waterhouse, B.D., Moises, H.C. and Woodward, D.J. (1981) Alpha receptor mediated facilitation of somatosensory cortical neuronal responses to excitatory synaptic inputs and iontophoretically applied acetylcholine. Neurvpharmacology, 20: 907-920. Waterhouse, B.D., Moises. H.C., Yeh, H.H. and Woodward, D.J. (1982) Norepinephrine enhancement of inhibitory synaptic mechanisms in cerebellum and cerebral cortex: Mediation by beta adrenergic receptors. J. Phurmacol. Exp. Ther., 221: 495-506. Waterhouse, B.D., Moises, H.C. and Woodward, D.J. (1986) Interaction of serotonin with somatosensory cortical neuronal responses to afferent synaptic inputs and putative neurotransmitters. Brain Res. Bull., 17: 507-518. Waterhouse, B.D., Azizi, S.A., Burne, R.A. and Woodward, D.J. (1990) Modulation of rat cortical area 17 neuronal responses to moving visual stimuli during norepinephrine and serotonin microiontophoresis. Bruin Res., 5 14: 276292. Woodward, D.J., Waterhouse, B.D., Hoffer, B.J. and Freedman, R. (1979) Modulatory actions of norepinephrine in the central nervous system. Fed. Proc., 38: 2109-2216. Yeh, H.H. (1982) Doctoral dissertation, University of Texas Health Science Center. Yeh, H.H. and Woodward, D.J. (1983) Beta-1 adrenergic receptors mediate noradrenergic facilitation of Purkinje cell responses to gamma-aminobutyric acid in cerebellum of rat. Neuropharmacology, 22: 629-639. Yeh, H.H., Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1981) Comparison of noradrenergic modulatory actions on Purkinje cell responses to iontophoresis of GABA, taurine, beta-alanine and muscimol. Neuropharmacology, 20: 549-560. ~
SECTION IV
Control of Motor and Sensory Systems
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Brain Research, Vol. 88 0 1YY1 Elsevier Science Publishers B.V.
365 CHAPTER 28
Central noradrenergic neurons: the autonomic connection P.G. Guyenet UniGersity of Virginia School of Medicine, Department of Pharmacology, Jefferson-Park Aoenue, Charlottesuille, VA, U.S.A.
Most CNS noradrenergic (NE) cell groups reside in portions of the medulla oblongata primarily involved in autonomic control (Al, A2,A5) and even the pontine locus coeruleus (A61 receives a major innervation from these medullary areas. This review examines the neuroanatomical and neurophysiological literature relevant to the issue of the role of CNS NE neurons in central autonomic control (with emphasis on cardiovascular control). It is concluded that NE cells, with the
possible exception of certain A5 and A1 neurons, have relatively weak or no inputs from visceral cardiovascular afferents but provide a complex “open loop” control over non-aminergic circuits which are more specialized in the processing of cardiovascular and other autonomic reflexes. The question of whether the C1 “adrenergic” cells of the rostra1 medulla oblongata actually use noradrenaline as a neurotransmitter is also briefly addressed.
Key words: CNS NE neurons, medulla oblongata, pons, ventrolateral medulla, nucleus tractus solitarius, central autonomic control
Introduction The Al, A2 and A5 groups of noradrenergic (NE) neurons (which together with the A7 pontine cluster, constitute the lateral tegmental group of NE cells, Levitt and Moore, 1979) are located in portions of the medulla oblongata which are primarily involved in autonomic and endocrine regulations, i.e., nucleus of the solitary tract (NTS), periambigual area and rostroventrolateral medullary reticular formation (RVLM, see Fig. 1). These three cell groups also innervate spinal, brainstem, diencephalic and basal forebrain regions primarily involved in autonomic and endocrine functions, NTS, ventrolateral medulla, dorsolateral pons, central grey matter, thoracolumbar intermediolateral cell column, paraven-
tricular, supraoptic and other hypothalamic nuclei, some midline thalamic nuclei, septum, bed nucleus stria terminalis, central nucleus of amygdala etc., (Levitt and Moore, 1979; Grzanna et al., 1987; Lyons et al., 1989). Thus the input-output characteristics of the lateral tegmental group (with the possible exception of the pontine A7 group about which very little is known) appear to set these cells apart from the locus coeruleus (LC) which alone innervates the cerebral cortex and whose subcortical projections are preferentially targeted to primary sensory and association nuclei (dorsal horn of spinal cord, spinal trigeminal nucleus, colliculi etc., Levitt and Moore, 1979; Grzanna et al., 1987; Lyons et al., 1989). Yet there is mounting evidence that the LC also receives major inputs from the ventrolateral and
366
AGNE
Fig. 1. Location of the A1-A6 groups of noradrenergic cells in the rat. Noradrenergic cells (circles) and adrenergic cells (triangles) are represented on a parasagittal section of the lower brainstem to emphasize their close proximity to structures involved in autonomic regulation. The various subdivisions of nucleus ambiguus are according to Bieger and Hopkins (1987). Abbreviations: Ac, compact subdivision of nucleus ambiguus; Ae, external formation of nucleus ambiguus; Al, loose formation of nucleus amb.; Asc, subcompact division; CVLM, caudal ventrolateral medulla (represents external formation of nucleus ambiguus overlying lateral reticular nucleus, Iat. RN); RVLM, rostralventrolateral medulla, a major source of input to both the thoracolumbar sympathetic neurons and the locus coeruleus (A6NE1; NTS, nucleus of the solitary tract.
dorsomedial medulla, two areas also primarily involved in cardiorespiratory and other autonomic regulations (Aston-Jones et al., 1986; Guyenet et al., 1989; Ruggiero et al., 1989). Thus the anatomical characteristics of central NE neurons convey a dual message. On one hand it is clear that each NE cluster exhibits territorial specialization in terms of its projection areas. On the other hand, these different NE cell groups all have in common the characteristic of either residing in an area which is specialized in endocrine and autonomic regulations (Al, A2, A5) or receiving major inputs from such areas (LC; Aston-Jones et al., 1986). These common characteristics raise the question of whether the agenda of the lateral tegmental NE cells is really fundamentally different from that of the LC (A6). One of the goals of this review will be to analyze the possible significance of these shared autonomic connections.
Because of the presence of NE in many brain areas involved in cardiovascular control, the role of N E in the maintenance of arterial pressure and in the operation of cardiovascular reflexes has been extensively investigated, and a significant portion of this review will be devoted to this topic. The problem has been generally approached in one of four main ways. One approach has been to measure how NE release (in various brain nuclei) or the unit activity of NE neurons is affected by changes in systemic pressure or other hemodynamic or respiratory variables. Second, intraparenchymal injections of NE in various brain areas or stimulation of NE cell bodies has been performed to determine the effects of these perturbations on the hemodynamic and respiratory status of experimental animals. Third, more or less restricted lesions of NE terminal fields or cell bodies have been performed in an attempt to measure the role of central NE neurons in the operation of selected cardiovascular reflexes. Finally, the electrophysiological effects of NE on specific CNS neurons involved in cardiovascular regulations have been investigated in several instances (e.g., preganglionic cells). The AS noradrenergic cell group This bilateral cluster of NE neurons is located dorsolateral and rostra1 to the superior olivary complex at the pontomedullary transition zone (Fig. 1). Of the various groups of NE cells, the A5 cluster appears most likely to have a role in controlling the sympathetic outflow although the modalities of its involvement are still far from clear. Input-output relationships of the A5 cell group The most persuasive argument for a role of A5 cells in central autohomic control is still neuroanatomic. Virtually all A5 cells project to the spinal cord, and the projection is generally described as specifically targeted to the thoracolumbar intermediolateral cell column and associated autonomic areas (lamina X and “intercalate nu-
367
clew,” Loewy et al., 1979b; Strack et al., 1989). One dissenting view is that of Grzanna and colleagues who have argued that A5 cells may also innervate ventral horn motoneurons, undefined lower lumbar structures (perhaps autonomic areas) and certain cranial motor nuclei (trigeminal and facial, Grzanna et al., 1987). The discrepancy could be explained by the existence of distinct cell clusters within the A5 cell group, each with partially separate projection areas, or possibly by a technical problem with the retrograde transport technique used by Grzanna et al. (ie., uptake by fibres of passage) or by strain differences (see Proudfit and Clark, this volume). Other known projections of A5 cells read like a classic list of brainstem areas involved in autonomic and neuroendocrine control, e.g., NTS, ventrolateral medulla, parabrachial nuclei and midbrain central grey (Byrum and Guyenet, 1987). At least 40% of A5 cells also project to still undefined hypothalamic nuclei (Byrum and Guyenet, 1987). Thus the main targets of A5 innervation appear globally similar to those of the more caudally located adrenergic cells (C1 group, Fig. 1) whose involvement in cardiovascular control is considered fairly well established (for a review of brain adrenergic systems see Hokfelt et al., 1984 and Ruggiero et al., 1989). The putative inputs to the A5 cells have been identified by retrograde tracing techniques and also include areas of the brain predominantly involved in autonomic control (NTS, parabrachial nuclei, perifornical area of the hypothalamus, ventrolateral medullary reticular formation, paraventricular nucleus of hypothalamus, Woodruff et al., 1986b; Byrum and Guyenet, 1987). These results are based on tract-tracing technology at the light microscopic level; definitive identification of these inputs must await electron microscopic plus electrophysiological evidence. A5 NE neurons, contrary to A1 cells but like A2 neurons, do not contain neuropeptide Y (NPY)-immunoreactive material (Everitt et al., 1984). A
small amount of this peptide may also be present in LC cells as well. Electrophysiological properties of A5 neurons A5 N E cells are the only cells in this portion of the ventrolateral medulla/pons which project to the thoracic spinal cord. Consequently, they can be convincingly identified on the basis of antidromic activation from the spinal cord (Byrum, et al., 1984). Under anesthesia (chloral hydrate or halothane) these neurons display a slow, regular rate of discharge, large triphasic extracellular spikes similar in shape to that of LC cells and, like the latter, they are silenced by minute doses of the a,-adrenergic agonist clonidine. This effect of clonidine is, as in the case of the LC, probably due in large part to the direct activation of somatodendritic receptors as suggested by iontophoretic experiments (Andrade and Aghajanian, 1982). A5 and A6 cells are also inhibited by systemic administration of the N E uptake blocker desmethylimipramine (Byrum et al., 1984). The presence of a,-adrenergic receptors on LC cells was originally viewed as mediating collateral inhibition between these cells (Aghajanian et at., 1977J but these receptors could also be involved in mediating an inhibitory input from the C1 adrenergic cells (PNMT-immunoreactive) of the rostra1 ventrolateral medulla (Astier and AstonJones 1989). The A5 area also contains PNMTimmunoreactive terminals (Hokfelt et al., 1984) suggesting that both A5 and A4 cells may be regulated by C1 adrenergic cells. The firing rate of many A5 cells is reduced when arterial pressure is elevated by infusion of a pressor agent (Andrade and Aghajanian 1982) and this effect is severely attenuated by combined section of the vagus, superior laryngeal, carotid sinus, aortic depressor and cervical sympathetic nerves (Guyenet, 1984). While suggestive, this evidence is still insufficient to definitively establish that the effect of pressor agents on A5 cells is due to the activation of baroreceptors. Also, the
368
precise type of baroreceptor involved (arterial, cardiopulmonary) needs to be established.
Autonomic effects of lesions or stimulations of the A5 cell group Electrical stimulation of the “A5 area” produces pressor effects and sympathoexcitation in rat and rabbit. In both species these pressor effects reportedly disappear after CNS treatment with the catecholamine-selective cytotoxic agent 6-hydroxydopamine (6-OHDA; icv in rats, Loewy et al., 1979a, local in A5 area in rabbit, Woodruff et al., 1986a). In contrast, the microinjection of excitatory amino acid (EAA) receptor agonists into the “A5 area” produces a depressor response (also in anesthetized rats, Loewy et al., 1986). This response is associated with a redistribution of peripheral blood flow in favor of hindlimb skeletal muscles and an ipsilateral reduction. in cerebral blood flow (Stanek et al., 1984). The depressor response is not observed after icv pretreatment of the rats with 6-OHDA (Neil and Loewy, 1982). The hypotensive effect produced by A5 stimulation with E M S is only modestly altered by the nearly complete destruction of the heavy projection of A5 cells to the cord and yields only to extensive combined 6OHDA lesions of the nucleus tractus solitarius and the spinal cord (Loewy et al., 1986). This result is in keeping with the extensive collateralization of A5 cells to the NTS, parabrachial area, medullary raphe and other areas involved in elaborating the vasomotor outflow. The discrepancy between the results of electrical and chemical stimulation of the A5 area has not been entirely resolved. The most plausible interpretation is that the pressor effects produced by electrical stimulation of the A5 region are due to the activation of fibers of passage which connect pressor areas of the medulla and pons (e.g., rostral medulla and dorsal pontine area, see Guyenet and Young, 1987). Yet it remains unclear why the A5 pressor effects were attenuated by 6-OHDA in two separate studies.
In freely behaving rats previously subjected to bilateral 6-OHDA lesions of the A5 area ( > 85% cell loss), cardioacceleration in response to hypotension was significantly attenuated while resting arterial pressure and heart rate were normal (controls received sham injections in the A5 area). A pharmacological dissection of the phenomenon suggested that the parasympathetic component of the reflex was massively depressed while the sympathetic component was, in fact, slightly exaggerated (Stornetta et al., 1986). This indicates that the activation of A5 cells could contribute to lowering vascular resistance under certain conditions as suggested by Stanek et af., (1984). Effects of noradrenaline on spinal preganglionic neurons and rostral medullary vasomotor cells One way to investigate the role of A5 cells in central autonomic regulation is to examine the effect of NE on CNS neurons which have a well-documented role in vasomotor control and are presumed targets of A5 neurons. Spinal preganglionic neurons are most likely in this category (Strack et al., 1989) although electronmicroscopic proof of monosynaptic A5 input to preganglionic neurons is still lacking (previous studies have repeatedly demonstrated monosynaptic contact between catecholaminergic boutons and preganglionic neurons, e.g., Chiba and Matsuko, 1986, but so far the only definitively characterized monosynaptic input is that from C1 adrenergic cells, Milner et al., 1988). In vitro, the most commonly observed effect of catecholamines on preganglionic cells is the activation of a -receptors which produces a variety of effects including depolarization, reduction in after-hyperpolarization and production of an afterdepolarizing event (species: cat, Yoshimura et al., 1987; for a review see Yoshimura et al., 1989). Long-term application of catecholamines even induced repetitive bursting behavior. These changes in the cells’ intrinsic properties amount to a gain-setting mechanism whereby the cells produce a burst of action potentials in response to a brief depolarizing pulse which would only have triggered single
369
spikes in the absence of NE. This mechanism may multiply the effect produced by other transmitters (such as glutamate) which work via ligand-gated channels. These excitatory effects of catecholamines may explain several previous observations (reviewed by McCall, 1988). In cats, the sympathetic vasomotor outflow (spontaneous in intact cats, or evoked by stimulation of the dorsolateral funiculus in spinal preparations) is reduced by systemic administration of a,-adrenergic antagonists or increased by NE-releasing agents. Also sympathoexcitation is produced by intrathecal administration of a,-adrenergic agonists in rats (Shi et al., 1988). It is worth stressing that no amount of a,-catecholamine receptor antagonist will produce more than about 30% reduction in the sympathetic vasomotor outflow in the cat (the effect is even less in the rat) indicating that the catecholaminergic innervation of the preganglionic neurons cannot provide the major part of the excitatory drive to these cells. a,-Receptor-mediated excitation is not the only effect of catecholamines on preganglionic neurons however. The prevalent effect of iontophoretic application of catecholamines in uivo is a,-receptor-mediated inhibition of unit discharge (e.g., Guyenet and Cabot, 1981). Accordingly, a substantial fraction of these cells in uitro is also hyperpolarized by catecholamines, often the same cells which also display an excitatory (depolarizing) response. The hyperpolarization has been attributed to an increase in potassium conductance similar to that found previously at other a,-receptors such as in the LC and the myenteric plexus (Yoshimura et al., 1989). The presence of both excitatory and inhibitory effects of catecholamines on target cells is not uncommon (a similar situation has been described for magnocellular cells of the paraventricular nucleus, purkinje cells, and spinal a-motoneurons, Yoshimura et af., 1989). In the case of preganglionic neurons, it remains to be established if both a,- and a,-receptors are accessible to NE released from the same terminals. As indicated before, preganglionic cells also receive a monosy-
naptic input from C1 adrenergic cells (Milner et al., 1988), which raises yet another possibility, namely that NE released from AS cells activates one type of a receptor while the catecholamine released by C1 cells (adrenaline or noradrenaline or a mixture of both, Sved, 1989) activates another. If this is so, the correct match is still unknown. The rostra1 ventrolateral medulla contains a collection of cells with projections to the intermediolateral cell column whose ongoing activity is essential to maintain the sympathetic tone and for the integration of a variety of vasomotor reflexes (for review, see Guyenet et al., 1989). Some of these neurons are tonically active in tissue slices as a result of intrinsic pacemaker properties. This population is most likely not adrenergic (for review see Guyenet et af., 1989) and possibly releases glutamate (for review see Morrison et al., 1989). Their pacemaker discharge is increased 50-70% by the activation of P-adrenergic receptors which are located on their cell bodies (Sun and Guyenet, 1990). The source of the catecholaminergic innervation of the rostralventrolateral medulla is not well established. Despite prior suggestion that it might originate from the A1 cells of the caudal ventrolateral medulla, this hypothesis has not been supported by positive neuroanatomical evidence (for review see Blessing and Li, 1989). A more likely source of catecholaminergic innervation is the AS cell group, although this point has also been disputed (Blessing and Li, 1989). Other possibilities include NE or epinephrine released in the rostralventrolateral medulla by C1 adrenergic cells or perhaps by inputs from the LC or A7 cell groups (Sun and Guyenet, 1986).
Conclusions The collection of sometimes contradictory data presented above illustrates the difficulty in obtaining a clear picture of the agenda of AS cells in the control of the vasomotor outflow. It appears probable that these cells are not part of the core circuitry responsible for the operation of the
370
%$GABA
/
EAA Fig. 2. Pathway of the baroreflex. Recent evidence strongly suggests that this reflex does not involve noradrenergic neurons but rather cells which utilize excitatory amino acids or GABA as transmitters (reviewed in Guyenet et al., 1989; Ruggiero et al., 1989). Whether the RVLM sympathoexcitatory cells are phenotypically adrenergic or not is still debated. There is evidence that this core circuitry can be modulated by NE both in the NTS and the RVLM.
baroreflex and other autonomic vasomotor reflexes under anesthesia since relatively minor effects on the sympathetic outflow result from blocking central catecholaminergic receptors or depleting central NE stores. In our view, another -very persuasive reason for this statement is that far more likely candidates for that role have been identified in the medulla: these are neurons which presumably release primarily GABA or EAA (depicted in Fig. 2; for review see Guyenet et al., 1989). Thus A5 neurons most likely play a “modulatory” role on networks involved in autonomic control, perhaps resulting in blood flow redistribution in favor of the skeletal muscles and an upregulation of the cardiovagal parasympathetic outflow. Such an agenda would require the facilitation of some preganglionic pathways and the relative inhibition of others, in theoretical agreement with the multiple and apparently contradictory effects of catecholamines on these cells. The fact that NE can produce both inhibition of preganglionic cells (a,-receptors) and facilitation of excitatory transmission (via a,-receptors) could perhaps also explain the opposite effects produced by chemical and electrical stimulation of A5 cells, since, depending on the pattern of neu-
ronal stimulation, one or the other effect might take precedence in terms of the overall discharge rate of the target cells. Whether the role of A5 cells is limited to the vasomotor portion of the autonomic system or whether these cells also influence other aspects of the autonomic outflow (e.g., sudomotor and piloerector outflows, gut motricity, adipose tissue, pupil, etc.) is entirely unexplored although these hypotheses remain, so far, equally plausible from a neuroanatomical standpoint.
The “adrenergic” cell group of the rostralventrolateral medulla The role of this cell group in central autonomic regulation would appear to be outside the topic of this review since the presence of immunoreactive phenylethanolamine-N-methyltransferase in these cells has long justified the belief that their primary transmitter might be adrenaline and not noradrenaline (for a review see Milner et al., 1989; Ruggiero et al., 1989). However recent evidence that there simply is no detectable adrenaline in the spinal cord suggests that the main catecholamine released by these cells at their spinal synaptic endings may in fact be N E (Sved, 1989). These cells are insensitive to 6-OHDA (e.g., Sun and Guyenet, 1990) and therefore may not have any amine uptake capability. C1 adrenergic cells express a very large variety of peptides (NPY, enkephalins, substance P, galanin; Levin et al., 1987; Milner et al., 1989). These cells may belong to the population of tonically active reticulospinal cells of the rostroventrolateral medulla which are barosensitive and provide an excitatory drive to the vasomotor sympathetic outflow (reviewed by Guyenet et al., 19891, but the evidence is far from definitive. The predominant effect of catecholamines microinjected into the ventrolateral medulla in civo is sympathoinhibition. The receptors involved in this effect appear to be of the a , subtype, but their precise cellular location is unknown (Punnen et al., 1987). This and autoradiographic evidence of
371
a moderate level of a,-receptor binding in the rostral medulla (Unnerstall et al., 1984) have led to the commonly held view that the rostral medulla is a major target for the central sympatholytic effect of systemically administered clonidine. A still undemonstrated possibility is that a,-receptors might be specifically associated with the adrenergic component of the reticulospinal projection of the ventrolateral medulla (Guyenet et al., 1989).
The A1 noradrenergic cell group and autonomic regulations It is quite clear that the “A1 area” is prominently involved in cardiorespiratory integration, regulation of hypothalamic hormones and the control of the upper GI tract and airways (Ross et al., 1985; Bieger and Hopkins, 1987; Ellenberger and Feldman, 1990). Yet, in our opinion, the evidence that A1 cells mediate specific cardiorespiratory or endocrine reflexes is still equivocal. By “mediating” a response, it is implied here that the cell has a rather restricted input-output function such as conveying a fairly limited type of information (e.g., baroreceptor- or chemoreceptor-related) from a restricted number of cells to a limited set of others. The term is contrasted with the concept of a modulatory effect (gating being an extreme form) exerted on the activity of other pathways through which modality-specific information is relayed.
Anatomical data The A1 N E cell group is located at and caudal to the obex in what has been defined as the “external formation” of the nucleus ambiguus (Bieger and Hopkins, 1987). A large fraction of A1 N E cells exhibit immunoreactivity for NPY (Everitt et al., 1984) and a smaller proportion stain for galanin (Levin et al., 1987). Whether these peptides are ever in the same cells is unresolved. They form a slender column within an area which contains an abundance of vagal motoneurons, the ventral respiratory group of bul-
bospinal respiratory premotoneurons and other respiratory-related neurons, and propriomedullary neurons involved in cardiorespiratory control (Ross et al., 1985). The specific inputs to A1 cells could possibly be identified only by combined use of anterograde tracers and electron microscopy since these cells and their dendrites are intermingled with numerous other neurons. No study of that type is in existence and thus currently available data represent a list of inputs to the “A1 area” at large. As in the case of the A5 cells, these represent a collection of brain structures predominantly involved in autonomic regulations (e.g., Ross et al., 1985; Ellenberger and Feldman, 1990). A comprehensive list of all possible targets of A1 cells is known from anterograde tracing studies in which tracer deposits have been placed in the general area where these cells are located (Loewy, et al., 1981; Sawchenko and Swanson, 1982). Again, only a limited number of these possible targets have been confirmed by retrograde transport experiments although that to the paraventricular nucleus of the hypothalamus has been examined in great detail (e.g., Sawchenko and Swanson, 1982; Levin et at., 1987). A1 cells innervate predominantly the vasopressin-rich component of magnocellular neurosecretary cells. The other projections of A1 cells include various other hypothalamic nuclei (probably including the dorsomedial nucleus, median eminence, lateral hypothalamic area), the median preoptic area, the median preoptic nucleus, a few basal forebrain structures such as the medial septum and bed nucleus of stria terminalis, the paraventricular nucleus of the hypothalamus, and caudally, the NTS, (e.g., Tucker et al., 1987). While the literature contains a few dissenting results, the general consensus of most recent investigators is that A1 cells innervate neither the rostralventrolateral medulla nor the spinal cord (Blessing and Li, 1989).
Electrophysiological and electrochemical data In rats, “A1 area” cells antidromically activated from the lateral hypothalamus (or median
372
preoptic area) fall into at least two main types (slow- and fast-conducting, for review see Yamashita et al., 1989) and probably three (Kaba et al., 1986). Fast-conducting units (ca. 2.2 m/sec) are normally silent but are activated by nociceptive stimuli. Slow-conducting cells (ca. 0.3 m/sec) display large triphasic action potentials and a slow, ongoing rate of discharge. Around 50% of these cells are most likely A1 NE cells since their antidromic activation is selectively blocked by application of 6-OHDA (but not 5,7-dihydroxytryptamine, a 5-HT selective neurotoxin) to their axons (Kaba et al., 1983). These 6-OHDA-sensitive cells are not excited by strong somatic noxious stimuli (Kaba et al,, 1986) and appear to be estrogen-sensitive (Kaba et aL, 1983). It seems probable that the slow-conducting group of A1 area cells with hypothalamic projection also contains 6-OHDA insensitive neurons which, according to (Kaba et al., 19861, tend to be located slightly rostral to the 6-OHDA-sensitive (i.e., most likely NA) cells. A good possibility is that the overlapping group of 6-OHDA-insensitive cells might belong to the C1 adrenergic cluster. Other investigators, who have focused on visceral inputs to A1 area slow-conducting cells have not distinguished between 6-OHDA sensitive and insensitive cells (McAllen and Blessing, 1987; Yoshimura et al., 1989) and thus the NE nature of the units which were recorded remains in doubt. In these studies, slow-conducting units were consistently excited by vagal stimulation. Most were inhibited by low doses of clonidine but this property may not discriminate between NE and adrenergic cells (Guyenet et al., 1989). About one-half of the tonically active A1 area slow-conducting cells which project to the hypothalamus are inhibited when systemic pressure is elevated by iv injection of a vasoconstrictor (rat, Yamashita et al., 1989). In the rabbit the majority of tonically active A1 area cells, backfired from the region of the supraoptic nucleus, responded to increases in afterload (aortic coarctation, which probably stimulates primarily arterial but also cardiac and pulmonary baroreceptors) and slightly more than
half were inhibited by aortic depressor nerve stimulation. Also about one-half were excited by decreasing cardiac return (McAllen and Blessing, 1987). Thus the constant finding in rat and rabbit is that around one-half of the tonically active, slow-conducting A1 area cells which project to the hypothalamus are inhibited when arterial baroreceptors are activated. Retrograde tracing experiments also indicate that only 50% of the A1 area cells which project to the hypothalamus are NE. Part of the remaining 50% may be C1 adrenergic cells which are present as far back as the level of the obex and, at least in the rostralventrolateral medulla, appear to display prominent barosensitivity (Guyenet et al., 1989). C1 adrenergic cells are insensitive to 6-OHDA (Granata et al., 1986; Sun and Guyenet, 1990) in keeping with the finding of (Kaba et aL, 1986) that only about half of the slow-conducting cells are affected by this toxin. In summary, the available electrophysiological data suggest that the A1 area contains tonically active cells with direct hypothalamic projections. Many such cells receive excitatory vagal and inhibitory baroreceptor inputs, but it is not clear which fraction, if any, of those barosensitive units is NE. A strong catechol electrochemical oxidation current is recorded with fine carbon electrodes in the A1 area (Quintin et af., 1987). This signal has been attributed to the extracellular presence of DOPAC released from the somatodendritic portion of the cells and its magnitude appears related to the level of NE synthesis. This indicator is dramatically reduced by clonidine and by hypocapnia due to hyperventilation. It is modestly reduced (ca. 25%) by raising systemic pressure with phenylephrine to around 135 mmHg or by volume load in previously hypovolemic rats. On the other hand the signal is considerably increased (up to 100%) by hypotensive hemorrhage (45 mmHg) or by long-duration infusions of a vasodilator (sodium nitroprusside). In these experiments, the effect of hemorrhage apparently could not be attributed to alterations in blood gases (and thus to peripheral chemoreceptor acti-
373
vation) since these remained unchanged. The effect of hemorrhage was greatly attenuated by sinoaortic deafferentation (ninth and tenth nerves cut at the foramen laceratum). Using a similar approach, Thrivikraman et al. (1989) also found that NE release was increased in the paraventricular nucleus (terminal field for A1 cells) during a hypotensive hemorrhage in anesthetized cats. Thus, both the electrochemical and unit recording evidence suggest that hypotension, and especially hypotensive hemorrhage, increase catecholamine release at the level of both terminals and cell bodies of A1 cells. It should be noted that the effect of hemorrhage on NE neurons is not at all specific to the A1 cell group since the turnover of NE everywhere in the CNS is also considerably enhanced by hypotensive hemorrhage. One representative study among many is that of Conlay et al. (1986) in the rat spinal cord (whose NE innervation originates in part from the LC and the rest from A5 but none of which originates from Al). Accordingly, the unit activity of the LC is also clearly enhanced by hemorrhage or long-term hypotension with sodium nitroprusside (Elam, 1985). In this case the afferent limb of the reflex appears to be vagal.
Functional data Electrolytic lesions of the A1 area have variously been described as effective or ineffective in attenuating the sympathetic baroreflex (Blessing and Li, 1989). The discrepancies are probably related to differences in the rostrocaudal level of the ventrolateral reticular formation which had been impaired by the lesions. The critical area responsible for mediating the sympathetic baroreflex appears to be located in a restricted portion of the external formation of the nucleus ambiguus at or slightly rostra1 to the obex (e.g., Willette et al., 1984; Guyenet et aL, 1987). This area is clearly more limited rostrocaudally than the territory emcompassed by the entire A1 cell group which extends to the pyramidal decussation. Other lines of evidence suggest that this restricted area may contain propriomedullary
GABAergic interneurons which convey baroreceptor-related information to the reticulospinal premotorneurons of the rostralventrolateral medulla (Fig. 2; for review see Guyenet et al., 1989). In addition, the A1 cells do not have the appropriate projections to mediate the sympathetic baroreflex since they do not project to the vicinity of the sympathetic premotoneurons of the rostralventrolateral medulla (Blessing and Li, 1989). Finally, while pharmacological blockade of catecholaminergic receptors in the rostralventrolateral medulla effectively antagonizes the effects of exogenously applied or endogenously released catecholamines (via tyramine), there is no evidence that these treatments significantly alter the sympathetic baroreflex (Granata et al., 1986). Thus, as in the case of A5 cells, there is no evidence that A1 cells are part of the core medullary circuit responsible for the operation of the sympathetic baroreflex and other reflexes. A role of A1 cells in “modulating” the activity of this core circuitry (“open-loop control”) is not to be dismissed, however, since a number of studies have demonstrated that the intraparenchymal injection of catecholaminergic agonists or antagonists into the NTS affects arterial pressure or modulates baroreflexes (e.g., Snyder et al., 1978; Smith et al., 1982; Granata et al., 1986). Of course this evidence does not specifically implicate A1 cells rather than A2 or A5 or even C1 cells, since all of these catecholaminergic cell groups also project to the NTS. The paraventricular (PVH) and supraoptic (SON) nuclei receive direct projections from the A1 cell group and magnocellular cells are excited by both NE (a,-receptors) and NPY (Levin et al., 1987; Day, 1989). An inhibitory effect of catecholamines has also been observed and has been attributed to P-receptors or to a,-receptors. This dual effect of catecholamines is reminiscent of the sympathetic preganglionic cells described above which also exhibit a,-adrenergic receptormediated excitatory and q m e d i a t e d inhibitory responses. There is excellent evidence that extensive le-
314
sions of the NE innervation of the brain rostra1 to the mesencephalon impair hemorrhage-induced release of vasopressin (Lightman et a),, 1984) and the inhibition of SON vasopressin cells by baroreceptor activation (Banks and Harris, 1984). Yet, two main pieces of data strongly suggest that the direct projection of A1 cells to the magnocellular neurons is not the major conduit through which baroreceptor- or chemoreceptor-afferent activity reaches these cells. First, the selective but complete destruction of the N E innervation of the SON does not prevent baroreceptor inhibition of the vasopressin cells (Day and Renaud, 1984) and, second, this inhibition yields entirely to iontophoretic application of a GABA antagonist (Jhamandas and Renaud, 1987). Electrical stimulation of the A1 area produces activation of vasopressin-releasing neurons of the supraoptic nucleus with a mean onset latency of 38 msec (Day, 1989). This latency is very close to the average antidromic latency of the tonically active slow-conducting cells recorded in the A1 area (Yamashita et at., 1989). In addition, the excitatory effect of A1 area stimulation on SON cells is greatly attenuated in rats in which the SON has been depleted of NE by microinjection of 6-OHDA (Day and Renaud, 1984). At first sight these three observations would appear to constitute strong evidence in support of a role of A1 NE neurons in mediating the excitatory effect of A1 area stimulation on SON vasopressin cells. Unfortunately the criteria of pharmacological blockade, indispensable for transmitter identification, has not been met, since the excitatory effect of A1 area stimulation cannot be blocked by even the highest doses of a ,-adrenergic receptor antagonist and the effect of NPY appears inappropriately small to account for the response (Day, 1989). Also the brief duration of the excitation produced by A1 area stimulation may be irreconcilable with an effect mediated by a,-receptor activation which requires the production of inositolpolyphosphate second messenger and, therefore, might be expected to last substantially longer. Thus, while it has been possible to
demonstrate satisfactorily that the short-lasting excitation of SON cells triggered by electrical stimulation of the ventrolateral medulla requires the integrity of the NE input to the SON it has not been possible to demonstrate that the effect is mediated by the release of N E in the SON. A possible interpretation is that the presence of NE is permissive for effective synaptic communication between medulla and SON to occur via nonNE inputs. Another possibility is that A1 neurons release an unknown fast-acting neurotransmitter or that NE activates a novel receptor which is neither a nor /3 (for discussion see Day, 1989).
Conclusions While the input-output relationships of the A1 area seem to implicate these N E cells in the control of neuroendocrine and autonomic functions, the agenda of these neurons is still unclear. Unit recording of A1 area cells somewhat supports the notion that at least some of them may receive inhibitory inputs from baroreceptors, but it does not seem that these cells are the major conduit through which baroreceptor-related information influences the discharge of vasopressin releasing cells. The hypothesis that they mediate baroreceptor inhibition of the vasomotor sympathetic outflow is even less clearly substantiated. Noradrenergic cells of the nucleus tractus solitarius and area postrema We will very briefly describe some of the available information relative to two groups of catecholaminergic cells of the dorsal medulla, the A2 NE cell group and the area postrema.
The A2 cells A2 neurons are located at and caudal to the obex in the median and commissuralis subnuclei of the NTS (Kalia et al., 1985; Riche et al., 1990). Their projections are extensive and may include most of the NTS, various hypothalamic nuclei including the oxytocin and vasopressin cells of the SON and PVH nuclei, and various basal fore-
315
brain areas (Riche et aL, 1990). In the NTS, caudal to the area postrema, a large majority of NTS neurons which projects to or through the lateral hypothalamus belong to the A2 group (rat, Moore and Guyenet 1983a; Riche et al., 1990). Based on this information, neurons with projections to the level of the median forebrain bundle have been recorded in the commissuralis subnucleus of the NTS and these units were considered likely to be the A2 cells. The vast majority have a slow, regular rate of discharge, large triphasic extracellular spikes, are inhibited by minute doses of clonidine iv (2-10 pg/kg), and by a,-adrenergic agonists applied via microiontophoresis (Moore and Guyenet, 1983b). Their activity is reduced by brief elevations of arterial pressure caused by increases in afterload, and this inhibition is attenuated in animals subjected to sectioning of the vagal and glossopharyngeal nerves. Finally, these cells are uniformly excited by vagal stimulation (Moore and Guyenet, 1985). These few characteristics of putative A2 cells (especially the action potential shape, slow regular firing and exquisite sensitivity to clonidine) are clearly reminiscent of the properties of LC, A5 and putative A1 neurons. Stimulation of the NTS in the region of the A2 cells produces excitation of SON vasopressin and oxytocin cells. The effect on vasopressin cells appears to be relayed in the ventrolateral medulla and, therefore, cannot be due to the direct projection of the A2 N E cells to the SON (Day, 1989). On the other hand, the excitation of oxytocin cells produced by NTS stimulation appears not to involve a relay in the same area of the ventrolateral medulla (Raby and Renaud 19891, but the theory that this excitation is attributable to a monosynaptic NE input from A2 cells needs to be substantiated. The area postrema The area postrema is a circumventricular organ which contains a large population of tyrosine-hydroxylase-immunoreactive neurons These appear to be largely NE (Kalia et al.,
1985). It is possible that some of these neurons project to the rostralventrolateral medulla (Blessing et al., 1987) although this conclusion requires confirmation with anterograde tract-tracing techniques. The other projections of the area postrema are also to brainstem structures involved in autonomic regulations (primarily the NTS and parabrachial nuclei; Shapiro and Miselis, 1985). While area postrema stimulation or ablation produces hemodynamic changes ( e g , Ferguson and Marcus, 1988; Mangiapane et al., 1989) the specific contribution of the NE cells in regulating the cardiovascular system has not yet been explored. Locus coeruleus and autonomic control
Because of the extensive projections of the A6 cell group and the nearly ubiquitous effect of catecholamines in all areas of the brain, it is clear that these NE cells must have a role in some aspect of autonomic control. While electrical stimulation of the LC elevates arterial pressure, this appears predominantly due to the activation of axons of passage (for references see Sved and Felsten, 1987). In contrast, chemical stimulation of the LC with either glutamate or carbachol produces atropine-resistant (i.e., presumably sympathetically mediated) hypotension in anesthetized animals and atropine-sensitive bradycardia. These effects have been attributed to the stimulation of NE cells since they completely disappear after destruction of the NE cells with 6-OHDA (Sved and Felsten, 1987). While these experiments demonstrate that the LC can affect the vasomotor sympathetic and parasympathetic outflow, the precise mechanism is unknown and may well remain unexplainable due to the enormous range of LC projections. In addition, the LC also appears to receive inputs of visceral origin. It receives a monosynaptic input from C1 adrenergic and other cells of the rostralventrolateral medulla (Aston-Jones et al., 1986; Ennis and Aston-Jones 1986, 1987; Milner et al., 1989) and additional inputs from the
316
dorsomedial medulla (Ennis and Aston-Jones, 1989). Stimulation of the ventrolateral medulla produces a glutamate receptor-mediated excitation of the A6 cells and an q a d r e n e r g i c receptor-mediated inhibition (perhaps from C1 cells, Astier and Aston-Jones, 1989). C1 cells in the same area also project to the intermediolateral cell column and a range of other brain structures involved in autonomic control which gives credence to the autonomic connection of the LC. On the other hand the input-output characteristics of the rostralventrolateral medullary cells, which are thought to provide an excitatory input to the LC, have not been examined and thus it is possible that despite their location, the primary role of these cells might not be to regulate autonomic functions or, at least not specifically, the vasomotor outflow. In anesthetized animals, LC cells respond to arterial pressure elevation only with transient inhibition (Elam, 1985). This effect is eliminated by vagotomy but it is not affected by sections of the arterial baroreceptor afferents. This is in clear contrast with the vasomotor sympathetic output and the sympathetic premotorneurons of the rostralventrolateral medulla which are tonically inhibited for as long as pressure remains elevated (Guyenet et af., 1989). A6 cells are consistently inhibited by volume loading and excited by hypotensive hemorrhage or sodium nitroprusside (a vasodilator agent) in anesthetized rats. This is a sustained effect which is also eliminated by bilateral vagotomy (Elam, 1985). While these experiments support the view that activation of lowpressure baroreceptors (atrial/cardiopulmonary) produces inhibition of the LC, this conclusion is not yet backed by direct evidence and relies on the effects of vagotomy. It is, therefore, possible that other types of receptors might be involved specially since the most effective stimuli to raise LC activity is severe hypotension (produced by hemorrhage or vasodilators) which, when combined with anesthesia probably results in severe peripheral tissue hypoxia. LC cell activity is also activated by hypoxia and hypercapnia (Elam,
1985). In unanesthetized cats, LC cells exhibit a modest degree of EKG-related rhythmicity, a criteria often considered as necessary to establish the presence of baroreceptor-related inputs (Morilak et al., 19871, but also a very artefact prone index in freely behaving animals. Indeed arterial pressure fluctuations seem to have very limited effects on the activity of these cells in the unanesthetized state (for review see Jacobs, 1986). Thus, at the present stage of investigation (largely based on experimentation in anesthetized animals, in the case of the lateral tegmental NE cells), it appears that the responsiveness of the LC to changes in peripheral hemodynamics may not differ greatly from that of the NE cell groups of the medulla (A5, A2, A l ) with the possible exception of A5 cells which appear to respond in more sustained and time-locked fashion to elevations of resting pressure. In the case of the LC, an input from arterial baroreceptors appears unlikely and for the other cell groups definitive experiments have either not been done (A2 and A5) or uncertainties concerning the identification of the cells do not allow us to draw conclusions with certainty (Al). LC, A2 and A1 cells respond to vagal stimulation by an excitation (effect on A5 unknown). Somatic nerve stimulation is very effective in activating the locus, modestly effective in activating A5 cells, and reportedly ineffective on presumed A2 cells. General conclusions The notion that medullary NE cells, just like the LC, provide a facilitatory (gain setting) or even permissive (gating) influence on certain target cells via the activation of a1 neuronal receptors (vasopressin cells, sympathetic preganglionic neurons), q r e c e p t o r s or P-receptors (RVLM neurons) is compatible with currently available data. The range of mechanisms which mediate the effect of NE released by the IateraI tegmental cell group is, so far as we know, similar to those mediating the effects of the LC on its targets. It is also well established that norepinephrine can
377
profoundly affect numerous types of endocrine and autonomic regulation (including cardiovascular function and respiration), when it is applied to brain sites which are the target of these NE cells. Moreover the selective removal of NE terminals affects, at times dramatically, the physiological functions performed by these brain nuclei (e.g., Snyder et al., 1978; Day and Renaud, 1984; McRae-Degueurce et al., 1986). Yet, evidence that medullary NE cells mediate specific autonomic reflexes is simply not convincing. This is especially obvious in the case of the respiratory system because, while numerous effects of catecholamines have been described, interaction between brainstem neurons involved in respiratory rhythmogenesis occur on short time scales totally incompatible with the slow conduction velocity of catecholaminergic neurons. As discussed above, the situation is somewhat less clear in the case of the brainstem circuits involved in controlling the sympathetic outflow and the release of vasopressin but, in our estimation, the bulk of the available evidence also tends to favor the view that N E neurons are not mediating specific autonomic reflexes, but rather that these cells modulate the activity of other, more specialized neurons. A classic view of the role of central NE neurons (largely based on studies performed on the LC system, for review see Jacobs, 1986) is that these cells respond to or anticipate localized or global changes in brain activity by carrying out a global or regional “CNS-autonomic” type of agenda consisting of (i) adjusting neuronal excitability (via a l - and az- and /?-receptors, resulting in a range of effects from simple facilitation to gating), (ii) adjusting the metabolic performance of glial cells and neurons (via /?,-receptors; Stone and Ariano, 19891, (iii) participating in local blood flow regulation (via p2- or a2-receptors in microvessels; Kalaria et al., 1989). This interpretation is consistent with the fact that these few hundred NE cells have been found to play a role in virtually every aspect of brain function and it also accounts for the nearly ubiquitous presence of NE innervation in the CNS and the
extreme degree of collateralization of these cells. Viewed in this light, it is conceivable that the agenda of all CNS NE cell groups might be rather similar, the major difference between the LC and the lateral tegmental group being the relative strength of visceral vs. somatic inputs to individual cell groups and a territorial division of the brain in terms of projection areas (somatic sensorimotor for the LC, visceroautonomic and endocrine for the others). In other words it is legitimate to speculate that the A l , A2 and A5 N E cells perform the same range of “housekeeping” duties as the LC but are primarily dedicated to the autonomic and neuroendocrine parts of the brain. It would be crucial to establish whether, like that of the LC (Jacobs, 19861, the activity of the lateral tegmental field NE neurons is statedependent. References Aghajanian, G.K., Cedarbaum, J.M. and Wang, R.Y. (1977) Evidence for norepinephrine-mediated collateral inhibition of locus coeruleus neurons. Brain Res., 136: 570-577. Andrade, R. and Aghajanian, G.K. (1982) Single cell activity in the noradrenergic AS region: Response to drugs and peripheral manipulations of blood pressure. Brain Res., 242: 125-135. Astier, B. and Aston-Jones, G. (1989) Electrophysiological evidence for medullary adrenergic inhibition of rat locus coeruleus neurons. SOC. Neurosci. Abstr., 15, Part 2: p. 1012. Aston-Jones, G., Ennis, M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986) The brain nucleus locus coeruleus: Restricted afferent control of a broad efferent network. Science, 234: 734-737. Banks, D. and Harris, M.C. (1984) Lesions of the locus coeruleus abolish baroreceptor-induced depression of supraoptic neurones in the rat. J. Physiol. (London), 355: 383-398. Bieger, D. and Hopkins, D.A. (1987) Viscerotopic representation of the upper alimentary tract in the medulla oblongata of the rat: The nucleus ambiguus. J. Comp. Neurol. 162 546-562. Blessing, W.W. and Li, Y.-W. (1989) Inhibitory vasomotor neurons in the caudal ventrolateral region of the medulla oblongata. Prog. Bruin Res., 81: 83-97. Blessing, W.W., Hedger, S.C., Joh, T.H. and Willoughby, J.O. (1987) Neurons in the area postrema are the only catecholamine-synthesizing cells in the medulla or pons with projections to the rostra1 ventrolateral medulla (C1 area) in the rabbit. Bruin Res., 419: 336-340.
378 Byrum, C.E. and Guyenet, P.G (1987) Afferent and efferent connections of the A5 noradrenergic cell group in the rat. J. Comp. Neurol. 261: 529-542. Byrum, C.E., Stornetta, R.L. and Guyenet, P.G. (1984) Electrophysiological properties of spinally projecting A5 noradrenergic neurons. Brain Rex, 303: 15-29. Chiba, T. and Masuko, S. (1986) Direct synaptic contact of catecholamine axons on the preganglionic sympathetic neurons in the thoracic spinal cord. Bruin Res., 380: 405408. Conlay, L.A., Maher, T.J., Godley, B.F. and Wurtman, R.J. (1986) Spinal cord noradrenergic neurons are activated in hypotension. Bruin Res., 375: 210-213. Day, T.A. (1989) Control of neurosecretory vasopressin cells by noradrenergic projections of the caudal ventrolateral medulla. In J. Ciriello, M.M. Caverson and C. Polosa (Eds.), The Central Neural Organization of Cardiocasculur Control, Progress in Bruin Res., Vol. 81, Elsevier, Amsterdam, pp. 303-317. Day, T.A. and Renaud, L.P. (1984) Electrophysiological evidence that noradrenergic afferents selectively facilitate the activity of supraoptic vasopressin neurons. Bruin Res., 303: 233-240. Elam, M. (1985) On the Physiological Regulation of Brain Norepinephrine Neurons in Rat Locus Coeruleus. Kompendietryckeriet, Kallered, Goteborg, Sweden. Ellenberger, H.H. and Feldman, J.L. (1990) Brainstem connections of the rostral ventral respiratory group of the rat. Bruin Res., 513: 35-42. Ennis, M. and Aston-Jones, G. (1986) A potent excitatory input to the nucleus locus coeruleus from the ventrolateral medulla. Neurosci. Lett. 71: 299-305. Ennis, M. and Aston-Jones, G. (1987) Two physiologically distinct populations of neurons in the ventrolateral medulla innervate the locus coeruleus. Bruin Res., 425: 275-282. Ennis, M. and Aston-Jones, G. (1989) GABA-mediated inhibition of locus coeruleus from the dorsomedial rostral medulla. J. Neurosci., 9: 1973-2981. Everitt, B.J., Hokfelt, T., Terenius, L., Tatemoto, K., Mutt, V. and Goldstein, M. (1984) Differential coexistence of neuropeptide Y (NPY)-like immunoreactivity with catecholamines in the central nervous system of the rat. Neuroscience, 11: 443-462. Ferguson, A.V. and Marcus, P. (1988) Area postrema stimulation induced cardiovascular changes in the rat. Am. 1. Physiol., 255: R855-R860. Granata, A.R., Numao, Y., Kumada, M. and Reis, D.J. (1986) A1 noradrenergic neurons tonically inhibit sympathoexcitatory neurons of C1 area in rat brainstem. Bruin Res., 377: 127-146. Grzanna, R., Chee, W.K. and Akeyson, E.W. (1987) Noradrenergic projections to brainstem nuclei: Evidence for differential projections from noradrenergic subgroups. J. Comp. Neurol., 263: 76-91. Guyenet, P.G. (1984) Baroreceptor-mediated inhibition of AS noradrenergic neurons. Bruin Rex, 303: 31-40. Guyenet, P.G. and Cabot, J.B. (1981) Inhibition of sympathetic preganglionic neurones by catecholamines and
clonidine: Mediation by an alpha-adrenergic receptor. J. Neurosci., 1: 908-917. Guyenet, P.G. and Young, B.S. (1987) Projections of nucleus paragigantocellularis lateralis to locus coeruleus and other structures in the rat. Brain Res., 406: 171-184. Guyenet, P.G., Filtz, T.M. and Donaldson, S.R. (1987) Role of excitatory amino acids in rat vagal and sympathetic baroreflexes. Brain Res., 407: 272-284. Guyenet, P.G., Haselton, J.R. and Sun, M.-K. (1989) Sympathoexcitatory neurons of the rostroventrolateral medulla and the origin of the sympathetic vasomotor tone. In J. Ciriello, M.M. Caverson and C. Polosa (Eds.), The Central Neural Organization of Central Cardiovascular Control, Progress in Brain Research, Vol. 81, Elsevier, Amsterdam, pp. 105-116. Hokfelt, T., Johansson, 0. and Goldstein, M. (1984) Central catecholamine neurons as revealed by immunohistochemistry with special reference to adrenaline neurons. In A. Bjorklund and T. Hokfelt (Eds.), Handbook of Chemical Neurounutomy, VoI.2: Classical Transmitters in the CNS. Part I , Elsevier, Amsterdam, pp. 157-276. Jacobs, B.L. (1986) Single-unit activity of locus coeruleus neurons in behaving animals. Prog. Neurobiol., 27: 183-194. Jhamandas, J.H. and Renaud, L.P. (1987) Bicuculline blocks an inhibitory baroreflex input to supraoptic vasopressin neurons. Am. J. Physiol., 252: R947-R952. Kaba, H., Saito, H., Otsuka, K., Seto, K. and Kawakami. M. (1983) Effects of estrogen on the excitability of neurons projecting from the noradrenergic A1 region to the preoptic and anterior hypothalamic area. Brain Res., 274: 156159. Kaba, H., Saito, H., Otsuka, K. and Seto, K. (1986) Ventrolatera1 medullary neurons projecting to the medial preopticanterior hypothalamic area through the medial forebrain bundle: An electrophysiological study in the rat. Exp. Brain Res., 63: 369-374. Kalaria, R.N., Stockmeier, C. and Harik, S. (1989) Brain microvessels are innervated by locus coeruleus noradrenergic neurons. Neurosci. Lett., 97: 203-208. Kalia, M., Fuxe, K. and Goldstein, M. (1985) Rat medulla oblongata. 11. Dopaminergic, noradrenergic (A1 and A2) and adrenergic neurons, nerve fibers, and presumptive terminal processes. J. Cornp. Neurol., 233: 308-332. Levin, M.C., Sawchenko, P.E., Howe, P.R.C., Bloom, S.R. and Polak, J.M. (1987) Organization of galanin-immunoreactive inputs to the paraventricular nucleus with special reference to their relationship to catecholaminergic afferents. J. Comp. Neurol., 261: 562-582. Levitt, P. and Moore, R.Y. (1979) Origin and organization of brainstem catecholamine innervation in the rat. J. Comp. Neurol., 186: 505-528. Lightman, S.L, Todd, K. and Everitt, B.J. (1984) Ascending noradrenergic projections from the brainstem: evidence for a major role in the regulation of blood pressure and vasopressin secretion. Exp. Bruin Res., 5 5 : 145-157. Loewy, A.D., et al. (1979a) Electrophysiological evidence that the A5 catecholaminergic cell group is a vasomotor center. Brain Res., 178: 196-200.
379 Loewy. A.D., et al. (1979b) Direct projections from the A5 catecholamine cell group to the intermediolateral cell column. Brain Res., 174: 309-314. Loewy. A.D., Wallach, J.H. and McKellar, S. (1981) Efferent connections of the ventral medulla oblongata in the rat. Brain Res. Re[*.,3: 63-80. Loewy. A.D., Marson, L., Parkinson, D., Perry, M.A. and Sawyer, W.B. (1986) Descending noradrenergic pathways involved in the A5 depressor response. Brain Res., 386: 313-324. Lyons, W.E.. Fritschy, J.-M. and Grzanna, R. (1989) The noradrenergic neurotoxin DSP-4 eliminates the coeruleospinal projection but spares projections of the A5 and A7 groups to the ventral horn of the rat spinal cord. J. Neurosci., 9: 1481-1489. Mangiapane, M.L., Skoog, K.M., Rittenhouse, P., Blair, M.L., and Sladek, C.D. (1989) Lesion of the area postrema region attenuates hypertension in spontaneously hypertensive rats. Circ. Res. 64: 129-135. McAllen, R.M. and Blessing, W.W. (1987) Neurons (presumably A1 cells) projecting from the caudal ventrolateral medulla to the region of the supraoptic nucleus respond to baroreceptor inputs in the rabbit. Neurosci. Lett., 73: 247252. McCall, R.B. (1988) Effects of putative neurotransmitters on sympathetic preganglionic neurons. Annu. Rec. Physiol., 50: 553-564. McRae-Degueurce, A,, Bellin, S.I., Landas, S.K. and Johnson, A.K. (1986) Fetal noradrenergic transplants into amine depleted basal forebrain nuclei restore drinking to angiotensin. Brain Res., 374: 162-166. Milner, T.A., Morrison, S.F., Abate, C. and Reis, D.J. (1988) Phenylethanolamine N-methyltransferase-containing terminals synapse directly on sympathetic preganglionic neurons in the rat. Brain Res., 448: 205-222. Milner, T.A., Pickel, V.M., Morrison, S.F. and Reis, D.J. (1989) Adrenergic neurons of the rostralventrolateral medulla: Ultrastructure and synaptic relations with other transmitter identified neurons. In J . Ciriello, M.M. Caverson and C. Polosa (Eds.), The Central Neural Organization of Cardiocascular Control, Prog. Brain Res., Vol. 81, Elsevier, Amsterdam, pp. 29-48. Moore. S.D. and Guyenet, P.G. (1983a) An electrophysiological study of the forebrain projection of nucleus commissura h : Preliminary identification of presumed A2 catecholaminergic neurons. Brain Res., 263: 211-222. Moore, S.D. and Guyenet, P.G. (1983b) Alpha-receptor mediated inhibition of A2 noradrenergic neurons. Brain Res., 276: 188-191. Moore, S.D. and Guyenet, P.G. (1985) Effect of blood pressure on A2 noradrenergic meurons. Brain Res., 338: 169172. Morilak, D.A., Fornal, C.A. and Jacobs, B.L. (1987) Effects of physiological manipulations on locus coeruleus neuronal activity in freely moving cats. 11. Cardiovascular challenge. Brain Res., 422: 24-31. Morrison, S.F., Ernsberger, P., Milner, T.A., Callaway, J., Gong, A. and Reis, D.J. (1989) A glutamate mechanism in
the intermediolateral nucleus mediates sympathoexcitatory responses to stimulation of the rostralventrolateral medulla. In J. Ciriello, M.M. Caverson and C. Polosa (Eds.), The Central Neural Organization of Cardiovascular Control, Prog. Brain Res., Vol. 81, Elsevier, Amsterdam, pp. 159170. Neil, J.J. and Loewy, A.D. (1982) Decreases in blood pressure in response to L-glutamate microinjections into the A5 catecholaminergic cell group. Brain Res., 241: 271-278. Punnen, S., Urbanski, R., Krieger, A.J., Sapru, H.N. (1987) Ventrolateral medullary pressor area: site of hypotensive action of clonidine. Brain Res. 422: 336-346. Quintin, L.. Gillon, J.-Y., Ghigone, M., Renaud, B. and Pujol, J.-F. (1987) Baroreflex-linked variations of catecholamine metabolism in the caudal ventrolateral medulla: An "in vivo" electrochemical study. Brain Res., 425: 319-336. Raby, W.N. and Renaud, L.P (1989) Nucleus tractus solitarius innervation of supraoptic nucleus: Anatomical and electrophysiological studies in the rat suggest differential innervation of oxytocin and vasopressin neurons. Prog. Brain Res., 81: 319-327. Riche, D., De Pommery, J. and Menetrey, D. (1990) Neuropeptides and catecholamines in efferent projections of the nuclei of the solitary tract in the rat. J. Comp. Neurol., 293: 399-424. Ross, C.A., Ruggiero, D.A. and Reis, D.J. (1985) Projections from the nucleus tractus solitarii to the rostral ventrolatera1 medulla. J. Comp. Neurol., 242: 511-534. Ruggiero, D.A., Cravo, S.L., Arango, V. and Reis, D.J. (1989) Central control of the circulation by the rostral ventrolatera1 reticular nucleus: Anatomical substrates. In J. Ciriello, M.Caverson and C. Polosa (Eds.), The Central Neural Organization of Cardiovascular Control, Progress in Brain Research, Vol. 81, Elsevier, Amsterdam, pp. 49-79. Sawchenko, P.E. and Swanson, L.W. (1982) The organization of noradrenergic pathways from the brainstem to the apraventricular and supraoptic nucleus in the rat. Brain Res. Rec., 4: 275-325. Shapiro, R.E. and Miselis, R.R. (1985) The central connections of the area postrema in the rat. J. Comp. Neurol., 234: 344-364. Shi, H., Lewis, D.I. and Coote, J.H. (1988) Effects of activating spinal-adrenoreceptors on sympathetic nerve activity in the rat. J. Auton. Nerc. Syst., 23: 69-78. Smith, W.L., Egle, J.L. and Adams, M.D. (1982) Adrenergic receptors in the nucleus tractus solitarii of the rat. Eur. J. Pharmacol., 81: 11-19. Snyder, D.W., Nathan, M.A. and Reis, D.J. (1978). Chronic lability of arterial pressure produced by selective destruction of the catecholamine innervation of the nucleus tractus solitarii in the rat. Circ. Res., 43: 662-671. Stanek, K.A., Neil, J.J., Sawyer, W.B. and Loewy, A.D. (1984) Changes in regional blood flow and cardiac output after I-glutamate stimulation of A5 cell group. Am. J. Physiol., 246: H44-H51. Stone, E.A. and Ariano, M.A. (1989) Are glial cells targets of the central noradrenergic system? A review of the evidence. Brain Res. Reu., 14: 297-309.
380 Stornetta, R.L., Guyenet, P.G. and McCarthy, R.M. (1986) Modulation of autonomic outflow by pontine AS noradrenergic neurons. In K. Nakamura (Ed.), Brain and Blood Pressure Control, Elsevier, Amsterdam, pp. 23-28. Strack, A.M., Sawyer, W.B., Hughes, J.H., Platt, K.B. and Loewy, A.D. (1989) A general pattern of CNS innervation of the sympathetic outflow demonstrated by transneuronal pseudorabies viral infections. Brain Res., 491: 156- 162. Sun, M.-K. and Guyenet, P.G. (1986) Effect of clonidine and gamma-amino butyric acid on the discharges of medullospinal symapathoexcitatory neurons in the rat. Brain Res., 368: 1-17. Sun, M.-K. and Guyenet, P.G. (1990) Excitation of rostral medullary pacemaker neurons with putative sympathoexcitatory function by cyclic AMP and beta-adrenoceptor agonists in citro. Brain Res., 511: 30-40. Sved, A.F. (1989) PNMT-containing catecholaminergic neurons are not necessarily adrenergic. Brain Res., 481: 113118. Sved, A.F. and Felsten, G . (1987) Stimulation of the locus coeruleus decreases arterial pressure. Brain Res., 414: 119-132. Thrivikraman, D.A., Bereiter, D.A. and Gann, D.S. (1989) Catecholamine activity in paraventricular hypothalamus after hemorrhage in cats. Am. J. Physiof., 257: H370H376. Tucker, D.C., Saper, C.B., Ruggiero, D.A. and Reis, D.J. (1987) Organization of central adrenergic pathways: I. Relationships of ventrolateral medullary projections to the hypothalamus and spinal cord. J. Comp. Neurol., 259: 591-603. Unnerstall, J.R., Kopaitic, T.A. and Kuhar, M.J. (1984) Distribution of alpha-2 agonist binding sites in the art and
human central nervous system: Analysis of some functional, anatomic correlates of the pharmacological effects of clonidine and related adrenergic agents. Brain Res. Reu., 7: 69-101. Willette, R.N., Punnen, S., Krieger, A.J. and Sapru, H.N. (1984) Interdependence of rostral and caudal ventrolateral medullary areas in the control of blood pressure. Bruin Res., 321: 169-174. Woodruff, M.L., Baisden, R.H. and Whittington, D.L. (1986a) Effects of electrical stimulation of the pontine AS cell group on blood pressure and heart rate in the rabbit. Brain Res., 379: 10-23. Woodruff, M.L., Baisden, R.H., Whittington, D.L. and Kelly, J.E. (1986b) Inputs to the pontine AS noradrenergic cell group: A horseradish peroxidase study. Exp. Neurol., 94: 782-787. Yamashita, H., Kannan, H. and Ueta, Y. (1989) Involvement of caudal ventrolateral medulla in mediating visceroreceptive information to the hypothalamic paraventricular nucleus. In J. Ciriello, M.M. Caverson and C. Polosa (Eds.), The Central Neural Organization of Cardiovascular Control, Prog. Brain Rex, Vol. 81, Elsevier, Amsterdam, pp. 293302. Yoshimura, M., Polosa, C. and Nishi, S. (1987) Noradrenaline-induced afterdepolarization in cat sympathetic preganglionic neurons in citro. J. Neurophysiol., 57: 13141324. Yoshimura, M., Polosa, C. and Nishi, S. (1989) Multiple actions of noradrenaline on sympathetic preganglionic neurons of the cat studied in the spinal cord slice. In J. Ciriello, M.M. Caverson and C.Polosa (Eds.), The Central Organization of Cardiouascular Control, Progress in Brain Research, Vol. 81, Amsterdam, pp. 181-190.
C.D. Barnes and 0. Pompeiano (Eds.1 Progress in Brain Research, Val. 88 0 1991 Elsevier Science Publishers B.V.
381 CHAPTER 29
Descending noradrenergic influences on pain S.L. Jones Department of Pharmacology, College of Medicine, University of Oklahoma, Oklahoma City, OK, U.S.A.
Multiple separate and distinct supraspinally organized descending inhibitory systems have been identified which are capable of powerfully modulating spinal nociceptive transmission. Until recently, brainstem sites known to be involved in the centrifugal modulation of spinal nociceptive transmission were few in number, being limited to midline structures in the midbrain and medulla (e.g., periaqueductal gray and nucleus raphe magnus). However, with continued investigation, that number has increased and brainstem sites previously thought to be primarily involved in cardiovascular function and autonomic regulation (e.g., nucleus tractus solitarius; locus coeruleus/subcoeruleus (LC/SC); A5 cell group; lateral reticular nucleus) also have been demonstrated to play a role in the modulation of spinal nociceptive transmission. Spinal monoamines (norepinephrine (NE) and serotonin) have been shown to mediate stimulation-produced descending inhibition of nociceptive transmission from these brainstem sites. The majority of NE-containing fibers and terminations in the spinal cord arise from supraspinal sources; thus, the
LC/SC, the parabrachial nuclei, the Kolliker-Fuse nucleus and the A5 cell group have all been suggested as possible sources of the spinal noradrenergic (NA) innervation involved in the centrifugal modulation of spinal nociceptive transmission. Several lines of evidence suggest that the LC/SC plays a significant role in a functionally important descending inhibitory NA system. Focal electrical stimulation in the LC produces an antinociception and increases significantly the spinal content of NA metabolites. The inhibition of the nociceptive tail-flick withdrawal reflex produced by electrical stimulation in the LC/SC has been demonstrated to be mediated by postsynaptic a,-adrenoceptors in the lumbar spinal cord. Similarly, electrical or chemical stimulation given in the LC/SC inhibits noxious-evoked dorsal horn neuronal activity. Thus, results reported in electrophysiological experiments confirm those reported in functional studies and the NA coeruleospinal system appears to play a significant role in spinal nociceptive processing.
Key words: antinociception, descending inhibition, analgesia, electrical stimulation, dorsal horn, spinal cord
Introduction Multiple separate and distinct supraspinally organized descending inhibitory systems have been identified which, when activated either by electrical stimulation or by the microinjection of drugs (e.g., morphine or glutamate) into selected regions of the brain (see Yaksh and Rudy, 1978; Gebhart, 1982 and Hammond, 1986 for reviews),
powerfully modulate spinal nociceptive transmission (Besson et al., 1975; Handwerker et al., 1975; Duggan et aZ., 1977 and Soja and Sinclair, 1983). The relevance of these descending inhibitory systems to antinociception and analgesia became clear with the demonstration that both spinal nociceptive reflexes and complex behaviors elicited by nociceptive stimuli are inhibited in rats, cats and monkeys (e.g., see Mayer, 1979 for
382
review) and analgesia is produced in man (e.g., Hosobuchi ef al., 1977) by stimulation or opioids given into these same brain regions. Until recently, brainstem sites known to be involved in the centrifugal modulation of spinal nociceptive transmission were few in number, being limited to midline structures in the midbrain and medulla (e.g., the periaqueductal gray and nucleus raphe magnus (NRM)). However, with continued investigation, that number has increased and brainstem sites previously thought to be primarily involved in cardiovascular function and autonomic regulation, (e.g., nucleus tractus solitarius; locus coeruleus (LC); A5 cell group; lateral reticular nucleus) also have been shown to play a role in the modulation of spinal nociceptive transmission. Spinal monoamines (norepinephrine (NE) and serotonin) have been demonstrated to mediate stimulation-produced descending inhibition of nociceptive transmission from these brainstem sites (see Proudfit, 1988 for review). This chapter will review the role that monoamines play in spinal nociceptive processing and will focus on the descending coeruleospinal system. Adrenoceptor agonists and antinociception It has been known since the 1940s that monoamines are involved in the modulation of pain and analgesia when it was demonstrated that systemically administered sympathomimetic agents (e.g., amphetamine) produced an analgesia in man (Burill et al., 1944). It became clear, however, that not all sympathomimetic agents produce an antinociception (e.g., oxymetazoline) when administered systemically. It was soon recognized that the failure of such agents to produce antinociceptive effects was due to their inability to cross the blood-brain barrier and enter the central nervous system; when administered intracerebrally, such agents also produce antinociceptive effects (see Yaksh, 1985 for review). That monoamines produce their antinociceptive effects via spinal sites of action has been
established. Monoaminergic agonists, including serotonin (Yaksh and Wilson, 1979) and NE (Kuraishi et al., 1979; Reddy et al., 19801, when administered intrathecally directly into the spinal subarachnoid space produce powerful antinociceptive effects. Systematic studies have revealed that the structure-activity series for intrathecally administered adrenoceptor agonists for the hot plate and tail-flick (TF) analgesiometric tests is: ST-91 (2-[2,6-diethyl-phenylamine]-2-imidazoline; (a,)= NE > methoxamine (a,) > > isoproterenol ( p ) = 0 (Yaksh, 1985). In the primate, using a shock titration task in which the animal defines a nociceptive threshold, similar results have been reported (Yaksh, 1985). Thus, the ability of intrathecally administered adrenoceptor agonists, including NE, to elevate nociceptive thresholds appears to be mediated by a adrenoceptors. To assess the relative role of spinal a,-versus spinal a,-adrenoceptors in antinociception, dose-response relationships have been generated examining the ability of selective a,-and qadrenoceptor antagonists to alter the antinociceptive effects produced by the intrathecal administration of a adrenoceptor agonists. The rank order of potency for antagonizing the antinociceptive effects of the a,-adrenoceptor agonist ST-91 is: yohimbine (a,),rauwolscine (a,),prazosin (a,), phentolamine (a, and a2),corynanthine (a,), propranolol ( p ) = 0 (Yaksh, 1985). Thus, a population of spinal a,-adrenoceptors specifically appears to mediate a-adrenoceptor agonist-produced antinociception. Modulation of spinal nociceptive transmission by a-adrenoceptor agonists Behavioral studies suggest that intrathecally administered NE is selective with regard to its antinociceptive effects; doses of NE required for maximal antinociceptive effects produce no significant signs of motor dysfunction (e.g., Reddy et al., 1980; Howe et al., 1983; Yaksh, 1985). NE also has been demonstrated to modulate sensory transmission in the dorsal horn; however, the
383
N E applied via microiontophoresis has been demonstrated to inhibit selectively unit activity of nociceptive-specific neurons in deep laminae of the cat dorsal horn (Belcher et al., 1978; Headley
effects of NE on evoked dorsal horn neuronal activity, and the selectivity of those effects for spinal nociceptive transmission, is less clear (see Table 1). TABLE 1
Effects of iontophoretically administered norepinephrine o n spinal cord dorsal horn neurons Lami- Neuron nae' Type
n
Activity/stimulus
Inhibition
Engberg & Ryall, 1966 Cat/Decerebrate
-
204
Elec. Stim.
96
Weight & Salmoiraghi. Cat/Decerebrate Cat/Ether 1966
-
92 36
Belcher er al., 1978
-
Reference
Species/Anesthetic
Cat /Fluothane/ a-Chl
a
HT
9 6
25 Spontaneous 10 Bradykinin ',heat 14 D,L-homocysteic acid 45 Spontaneous 12 Brush, movement 18 u,L-homocysteic acid
LT
Headley er ul., 1978
Cat/a-Chl
IV-v
MULTI
23
Satoh et al., 1979
Rabbit/Ether
V
MULTI
20 Bradykinin 11 Bradykinin, brush
Todd & Millar, 1983
Cat/Pentobarbitone
1-111
47
Willcockson et ul.,
Primate/Hal/N20/ a-Chl
I-v
19
1984
Flee twood-Walker
Cat/Hal/a-Chl
-
Rat/Hal/Urethane
1-11
et al. 19x5 ~
Howe & Zieglgansberger, 1987
MULTI LT
46 10
deep ~
~
~
~
~
~
Pinch Brush Glutamate
~
No Effect 108
30 6
53 24 5 1 2
2 1 2
35 10 14
4 19 12
8 4
7 5
17
3 2
25 2
12
43 Heat Spontaneous, brush, DL-H 2
25 Pinch Pinch, Brush Brush 39 Pinch Pinch, Brush Brush 30 Proprioceptive Deep Pinch
111
~
Heat Heat, brush
20 9 12 8 1 2
Excitation
3 8
5
1 1
3
4 1
2 '
5 10 6' 14 4 1 9
1
9k 5'
4
~
Abbrer,iutions: a-Chl, a-chloralose; Hal, halothane; N,O, nitrous oxide. Laminae in which neurons were recorded and norepinephrine iontophoresed; - indicates that laminae were not described. Neuron Type: HT, high threshold; LT, low threshold; MULTI, multireceptive. n , number of spinal units studied and number in which responses to stimuli were inhibited, excited or unaffected by iontophoretically applied norepinephrine. Electrical stimulation of dorsal roots and cutaneous, muscle and joint afferent nerves. Bradykinin (5-15 f i g ) was injected into the blood supply of the peripheral receptive fields. Noxious heat, usually radiant (45-50°C). The inhibition of heat-evoked unit activity produced by iontophoretically applied norepinephrine was selective for nociceptiveevoked activity; spontaneous and non-noxious-evoked unit activity was unaffected for these same 43 units. ' One unit exhibited biphasic inhibition/excitation. I Six units exhibited biphasic inhibition/excitation. One unit exhibited biphasic inhibition/excitation. I One unit exhibited biphasic inhibition/excitation.
384
et al., 1978). Fleetwood-Walker and colleagues (1985) have established further that the selective inhibition of nociceptive-evoked unit activity produced by the iontophoresis of NE onto spinocervical tract or dorsal column neurons in the cat is mediated by a,-adrenoceptors. The q a d r e n o c eptor agonists, clonidine and metaraminol, mimicked the selective inhibition of nociceptiveevoked unit activity produced by NE; the a l adrenoceptor agonist, phenylephrine, and the p adrenoceptor agonist, isoprenaline, did not. Additionally, the a,-adrenoceptor antagonists, yohimbine and idazoxan, either reversed or reduced the potency of the inhibition of nociceptiveevoked unit activity produced by NE. In the primate, iontophoretically applied NE has been reported to inhibit glutamate-evoked unit activity of dorsal horn neurons specifically characterized as having ascending projections to the thalamus. The inhibition, however, was not selective for nociceptive-specific neurons; both nociceptive- and nonnociceptive-evoked unit activity was inhibited by NE (Willcockson et al., 1984). In contrast, Todd and Millar (1983) reported that iontophoretically applied NE excited approximately 50% of the neurons examined in laminae I and 11, but had no effect on units in laminae 111; no correlation was found between effects on unit activity and neuron modality. Howe and Zieglgansberger (1987) similarly reported that iontophoresed NE inhibited or excited dorsal horn neuronal activity; NE had exclusively inhibitory effects on low threshold neurons in laminae I and 11, and multireceptive neurons in lamina 111. In an in vitro intracellular study in the rat spinal cord slice preparation, North and Yoshimura (1984) reported that NE, applied either by superfusion or pressure ejection, hyperpolarized 80% of the neurons examined in lamina 11. The hyperpolarization was associated with an increase in potassium conductance, and was blocked by phentolamine or yohimbine but not by propranolol or prazosin. However, due to limitations of the in uitro preparation, it was not possible to determine whether the neurons influenced by NE received nociceptive afferent input.
Thus, electrophysiological evidence supports behavioral data which suggests that NE modulates sensory transmission in the spinal cord dorsal horn. The bulk of evidence suggests that iontophoretically applied NE has predominantly inhibitory effects on dorsal horn neuronal activity; however, excitatory effects also have been reported. From the data currently available, it is difficult to conclude that NE has selective inhibitory effects on spinal nociceptive transmission; the physiological characterization of the neurons influenced by NE (i.e., low threshold, multireceptive or high threshold) often is unclear, as is the dorsaI horn laminae in which they were located (see Table 1). The coeruleospinal projection
Transection studies have revealed that spinal cord NE and serotonin content are depleted significantly caudal, but not rostral, to the level of a complete transection of the spinal cord, indicating that the source of spinal cord monoaminergic innervation is organized supraspinally (e.g., Carlsson et al., 1963; Magnusson and Rosengren, 1963). In the rat, pontine NA cell groups have been demonstrated to be the primary source of NA nerve terminals in the spinal cord (Nygren and Olson, 1977; Moore and Bloom, 1979; Bjorklund and Skagerberg, 1982; Westlund et al., 1983). Neurons immunohistochemically labeled with retrogradely transported dopamine-P-hydroxylase (DPH) antiserum from the spinal cord have been localized in the nucleus LC, the nucleus subcoeruleus (SC), the parabrachial nuclei, the Kolliker-Fuse nucleus and the region of the superior olivary nucleus (the A5 cell group) (Westlund et al., 1983). Numerous studies, utilizing a variety of techniques, have demonstrated that a direct coeruleospinal projection exists in the rat, cat and primate (Olson and Fuxe, 1972; Amaral and Sinnamon, 1977; Nygren and Olson, 1977; Commissiong et al., 1978; Ader et al., 1979; Guyenet, 1980; Westlund and Coulter, 1980; Foote et al., 1983; Westlund et al., 1983; Carlton et al., 1985;
385
Loughlin et al., 1986a,b; Fritschy et al., 1987). Westlund et al. (1983) reported that 86% of neurons retrogradely labeled with D P H antibody from the spinal cord in the rat are located in the LC/SC. Bilateral lesions of the LC/SC produce a 30-40% reduction in spinal cord NE content in the rat, and an 80% reduction in spinal cord NE content in the ventral horn of the spinal cord in the cat (see Bjorklund and Skagerberg, 1982 for review). NA coeruleospinal efferents originate primarily from ventral portions of the LC and the SC (e.g., Westlund et aL, 1983 Loughlin et al., 1986a,b); they can be traced the full length of the spinal cord and decussate at all levels of the neuraxis (Nygren and Olson, 1977; Commissiong et al., 1978; Westlund et al., 1983; Carlton et al., 1985; Jones and Yang, 1985). Coeruleospinal fibers descend the spinal cord predominantly in the ipsilateral, ventrolateral quadrant; however, axons and terminals also have been identified which extend into the dorsolateral quadrant of the spinal cord (Nygren and Olson, 1977; Commissiong et al., 1978; Bjorklund and Skagerberg, 1982; Westlund et al., 1983; Jones and Yang, 1985). It should be noted that significant anatomical differences exist between the rat and cat with regard to the organization and spinopetal efferents of the NE-containing dorsolateral pontine nuclei. The rat and cat are the two most commonly used species in pharmacological and electrophysiological studies of descending inhibition. In contrast to the rat, it has been reported that in the cat, although catecholamine-containing neurons in the LC do project to the spinal cord (Jones and Moore, 1974; Kuypers and Maisky, 19751, the primary source of spinal NE-containing nerve terminals is not the LC, but rather the Kolliker-Fuse nucleus. Many cells that do project to the spinal cord from the LC do not contain NE (Stevens et al., 1982, 1985); serotonin-containing cells in the LC have been shown to contribute to the monoaminergic content of the ventral horn of the spinal cord (Lai and Barnes, 1985). These anatomical differences between species have
added confusion to the literature and, as a result, the extrapolation of data obtained in pharmacological, neurochemical and behavioral studies in the rat, to data obtained in electrophysiological studies in the cat, may not necessarily be valid. Termination patterns in the spinal cord High densities of DpH-labeled fibers have been identified in the superficial laminae of the dorsal horn, in the ventral horn around large motoneurons, around the central gray and in the intermediolateral spinal gray matter in the thoracic and sacral spinal cord (e.g., Westlund et aL, 1983). Both a l - and a,-adrenoceptors are present in the spinal gray matter of the rat spinal cord (Seybold and Elde, 1984; Unnerstall et al., 1984). Autoradiographic studies have revealed dense distributions of p-[3H]aminoclonidine binding sites (i.e. , a,) in lamina I1 of the rat spinal cord, around the central canal and in the intermediolateral cell column (see Seybold, 1986 for review). Nociceptive-specific primary afferents also have been demonstrated to terminate primarily in the superficial laminae of the spinal cord dorsal horn (Light and Perl, 1979a,b); thus, NE-containing nerve terminals and NA binding sites are ideally situated to contribute significantly to the modulation of spinal nociceptive processing. A recent study examined the termination patterns of NA coeruleospinal fibers within the rat dorsal horn by using the anterograde tracer Phaseolus vulgaris leucoagglutinin (PHA-L) in combination with DPH immunohistochemistry (Fritschy et al., 1987). Contrary to what previous studies have reported utilizing retrograde-tracing techniques (e.g., Jones and Yang, 1983, dense projections of coeruleospinal axons were not observed in the ventral horn and deep laminae of the spinal cord dorsal horn. Rather, dense coeruleospinal projections were observed in the superficial laminae of the dorsal horn, particularly in the substantia gelatinosa ( i e . , lamina 11; however, see Proudfit and Clark, this volume). More than 80% of the PHA-L labeled fibers in
386
J )loo/*
P 3.0
I
... &&:::m.
\ - 25-50
I
MOO/* P 3.5
Fig. 1. Diagrammatic summaries of stimulation thresholds for inhibition of the tail-flick (TF) reflex drawn on representative coronal brain sections. The thresholds (PA) for inhibition of the TF reflex are indicated: > loo/*, indicates no inhibition of the TF reflex at stimulation intensities of > 100 p A and up to 200 p A or non-antinociceptive effects of stimulation at stimulation intensities ranging between 6.25 and 200 p A . These summary diagrams were constructed from 14 electrode tracks through the pons (v). (From Jones and Gebhart, 1986a, with permission.)
the spinal cord also exhibited positive immunostaining for DPH, indicating that the majority of fibers projecting to the spinal cord from the LC in the rat utilize NE as their neurotransmitter (Fritschy et al., 1987). Modulation of spinal nociceptive transmission by coeruleospinal efferents
As discussed above, NE has been shown to be involved in both antinociception and inhibition of spinal nociceptive transmission. Since the majority of NE-containing fibers and terminations arise from supraspinal sources, the LC/SC, the parabrachial nuclei, the Kolliker-Fuse nucleus and the A5 cell group have all been suggested as
possible sources of the spinal NA nerve terminals involved in the centrifugal modulation of spinal nociceptive transmission. Several lines of evidence suggest that the LC/SC plays a significant role in a functionally important descending inhibitory NA system. In the rat, focal electrical stimulation in the LC produces an antinociception (Segal and Sandberg, 1977; Margalit and Segal, 1979) and increases significantly the spinal content of NE metabolites (Crawley et af.,1979). Recent studies have examined systematically the role of the LC/SC in the modulation of spinal nociceptive reflexes and spinal nociceptive transmission in the rat and cat. Systematic mapping studies have revealed that, in the lightly pentobarbital-anesthetized rat, inhi-
387 Phentolornine
Yohirnbine
Prazosin
Naloxone
Methysergide
100
40
20
0
Fig. 2. Summary of the effects of intrathecally administered phentolamine ( n = 6), yohirnbine ( n = 101, prazosin ( n = 61, naloxone ( n = 4) and methysergide ( n = 6 ) on the stimulation thresholds in the locus coeruleus/subcoeruleus (LC/SC) required to inhibit the TF reflex. Mean % increases (kS.E.M.) in the stimulation threshold following cumulative intrathecal doses of antagonists. *, P 2 0.01 VS. pre-treatment LC/SC stimulation threshold (paired t-test).
bition of the nociceptive T F withdrawal reflex is produced by electrical stimulation throughout a wide region of the dorsolateral pons; however, stimulation sites requiring the lowest intensities of stimulation to inhibit the T F reflex (12.5-25 pA) are in the LC/SC (Fig. 1). S-glutamate microinjections and stimulation strength-duration determinations indicated that inhibition of the T F reflex produced by stimulation in the LC/SC results from the activation of presumably NA cell bodies (Jones and Gebhart, 1986a). The intrathecal administration of pharmacological antagonists (phentolamine, yohimbine, prazosin, naloxone, methysergide, atropine or bicuculline) revealed that only the non-selective a-adrenoceptor antagonist, phentolamine, and the selective a,-adrenoceptor antagonist, yohimbine, significantly increased stimulation thresholds in the LC/SC to inhibit the T F reflex (Fig. 2). A cumulative intrathecal dose of 30 p g of phentolamine produced a mean 83.1 16.3% increase in the inhibitory threshold; a cumulative 30 p g dose of yohimbine produced a mean 93.9 t- 13.2% increase in the stimulation threshold to inhibit the TF reflex (Jones and Gebhart, 1986a). Thus, inhibition of the spinal nociceptive TF reflex produced by electrical stimulation in the LC/SC appears to be mediated by postsynaptic a,-adrenoceptors in the lumbar spinal cord.
*
Similarly, electrical stimulation in the dorsolatera1 pons has been demonstrated to inhibit, significantly, spinal nociceptive transmission. Systematic tracking studies in pentobarbital-anesthetized rats revealed that the site of maximal inhibition of noxious heat-evoked (50°C) dorsal horn neuronal activity produced by 100 p A of electrical stimulation was in the ventral LC and the SC (Fig. 3). Electrical stimulation in either the ipsilateral or contralateral LC/SC was equally effective in inhibiting heat-evoked dorsal horn unit activity (Fig. 3). The inhibition was intensity-, pulse duration-, and frequency-dependent. Inhibition of heat-evoked unit activity to 50% of control was produced at a significantly lower intensity of stimulation using a pulse duration of 400 ps (54.0 k 8.8 FA) compared with stimulation using a pulse duration of 100 ps (82.1 11.2 PA); electrical stimulation at a frequency of 100 Hz resulted in maximal inhibition of heat-evoked dorsal horn neuronal activity (Jones and Gebhart, 1986b). Microinjections of S-glutamate or kainic acid into the LC/SC suggest that the inhibition of dorsal horn neuronal responses to noxious heating of the skin produced by electrical stimulation in the LC/SC is the result of the excitation of cell bodies in the LC/SC, which are presumably NA (Jones and Gebhart, 1986b). Antibodies to DPH have been demonstrated
+
388
A
B
CONTRALATERAL rsrpanrs to skin healing
depth. mm
105
P3 0
20
,so
m
0
d
.
-
b
C
Llyu
o m % control respanse lo skin heating
Fig. 3. Histological reconstructions of electrode tracks (vertical lines) through the ipsilateral and contralateral LC/SC. A. The units’ responses to skin heating are shown as a percentage of control to the right of each track. At each stimulation site ( b ) the effects of 100 p A stimulation on heat-evoked activity was determined. Two example peristimulus time histograms are shown for each track at sites indicated by the matching letters; onset and termination of brain stimulation are indicated by upward arrows and downward arrows, respectively. Beneath the histograms the black bar illustrates the duration of skin heating. In both the ipsilateral and contralateral tracks the site of maximal inhibition by 100 F A stimulation was in the SC where heat-evoked activity was inhibited to 19.7 and 22.8% of control, respectively. B. Three additional tracks lateral to, through, and medial to the LC/SC; at each stimulation site ( b ) the effects of 100 p A stimulation on heat-evoked spinal unit activity was determined. The unit’s responses to skin heating are shown as a percentage of control below each track corresponding to the matching letters (a, b, and c). (From Jones and Gebhart, 1986b, with permission.)
to be selectively incorporated and retrogradely transported by NA nerve terminals (Fillenz et al., 1976; Ziegler et aL, 1976; Westlund et al., 1983). Microinjections of DOH into the lumbar spinal cord into the same region in which dorsal horn neurons were recorded in the above studies resulted in DpH-labeled cells in the LC/SC (Fig. 4). Thus, NA, coeruleospinal terminals are located in the same region of the spinal cord in which dorsal horn unit recordings were made, supporting the supposition that LC/SC stimula-
tion-produced inhibition of noxious-evoked dorsal horn unit activity is an NA-mediated effect. Mic,roinjections of the local anesthetic, lidocaine, and transection techniques to interrupt neuronal transmission have been used to examine the funicular trajectories of the coeruleospinal fibers functionally important in the modulation of spinal nociceptive transmission (Jones and Gebhart, 1987). Microinjections (0.5 pl) of lidocaine to reversibly interrupt neuronal transmission were made into the ipsilateral and/or contralateral
389
ventrolateral quadrants (VLFs) of the cervical spinal cord and the effects on LC/SC stimulation-produced inhibition of heat-evoked dorsal horn neuronal activity were determined; the results are summarized in Figure 5A. Prior to the microinjection of lidocaine, focal electrical stimulation in the LC/SC at a mean intensity of 101.4 & 5.1 pA ( n = 18) inhibited heat-evoked dorsal horn unit activity to 31.5 f 3.0% of control. Following the microinjection of lidocaine into the ipsilateral VLF, the efficacy of the same intensity of LC/SC stimulation to inhibit heat-evoked unit activity was decreased significantly to 66.5 & 4.7% of control. Subsequently blocking the contralat-
Fig. 4. LC/SC neurons containing retrogradely transported dopamine-O-hydroxylase (DPH) antibody from the lumbar spinal cord. A. Low power (4x1 photomicrograph of retrogradely filled DOH-containing neurons in the LC/SC. B. Higher magnification (16 X ) of DOH-labeled neurons in the LC/SC from another animal. Abbreviations: LC/SC, locus coeruleus/subcoeruleus; IV, fourth cerebral ventricle; PB, parabrachial nuclei.
era1 VLF, thereby producing a bilateral blockade of neuronal transmission in the VLFs of the cervical spinal cord, further decreased significantly the efficacy of LC/SC stimulation-produced inhibition to 71.1 f 6.8% of control. The mean duration of effect of lidocaine was 54.9 f 5.5 min, after which time LC/SC stimulation at the same mean intensity again inhibited heat-evoked unit activity to a mean 37.8 k 4.0% of control. Contrariwise, irreversible transections of the dorsolateral quadrants of the cervical spinal cord failed to significantly affect LC/SC stimulationproduced inhibition of heat-evoked dorsal horn neuronal activity (pre-, 40.0 f 2.8% of control; ipsi-, 51.5 f 5.6% of control; bi-, 53.7 f 8.4% of control; n = 9; Fig. 5B). Thus, these results suggest that a significant portion of coeruleospinal fibers mediating LC/SC stimulation-produced inhibition descend to the lumbar spinal cord in the ipsilateral ventrolateral quadrant; at the cervical level, coeruleospinal fibers in the dorsolateral quadrants of the spinal cord are not functionally involved in LC/SC stimulation-produced inhibition of nociceptive transmission. Thus, results reported in electrophysiological experiments confirm those reported in functional studies using the nociceptive TF withdrawal analgesiometric model. The NA coeruleospinal system appears to play a significant role in spinal nociceptive processing. In analogous studies in the cat, dorsal horn neuronal activity evoked by non-noxious and noxious cutaneous stimuli (Hodge et al., 1981; 1983) and by electrical nerve stimulation at intensities sufficient to activate A-6- and C-fibers (Mokha et al., 1985; 1986) similarly has been shown to be inhibited by electrical stimulation in the LC. A recent intracellular analysis suggests that inhibition of spinal nociceptive transmission from the LC involves both pre- and postsynaptic mechanisms (Mokha and Iggo, 1987). Transections of the spinal cord at the lower thoracic level, involving a part or whole of the ipsilateral ventral quadrant, reduced the inhibition produced by stimulation in the LC, suggesting that in the cat,
300
coeruleospinal fibers also descend the spinal cord predominantly in the ipsilateral ventrolateral funiculus (Mokha et al., 1986). In contrast to the rat, Hodge et al. (1983) have reported that in the cat LC stimulation-produced inhibition of dorsal horn neuronal activity is not mediated by spinal NE. The destruction of spinal NA terminals by the selective NA neurotoxin 6-hydroxydopamine (6-OHDA) failed to alter the efficacy of LC stimulation to inhibit dorsal horn neuronal activity evoked by noxious stimuli. How-
ever, unaccounted for in this study (and many studies involving neurotransmitter depletion) is the possible development of receptor supersensitivity. In the report by Hodge et al. (1983), cats were studied 8-10 days following the first intrathecal dose of 6-OHDA, a time likely well past the development of receptor supersensitivity. Although depIeted significantly, N E was not totally absent; thus, it is possible that released N E acting at supersensitive receptors could produce inhibitory effects on spinal neurons qualitatively
LC-VLF LIDO % control
p
re*
80
*1
pre-
bi-
ipsi
recov
LC-DLF-X
% control
0
f
60
pre-
ipsi-
bi -
Fig. 5. Summary data of the effects of ipsilateral and bilateral' ventrolateral funiculus (VLF) lidocaine microinjections and dorsolateral funiculus transections (DLF-X) on locus coeruleus stimulation (LCS)-produced inhibition of heat-evoked dorsal horn unit activity. In A and B are shown mean dorsal horn unit responses to skin heating (50°C) as a percentage of the control response during LCS before lidocaine or DLF-X (pre-) and after ipsilateral (ipsi-) and bilateral (bi-) lidocaine microinjections into the VLF (A) or DLF-X (B); sites of stimulation in the LC/SC are shown to the right. In A, recov. represents the mean inhibition by LCS following dissipation of the effects of lidocaine. * represents a significant difference compared with pre-VLF lidocaine values. * * represents a significant difference compared with post-ipsilateral VLF lidocaine values. (From Jones and Gebhart, 1987, with permission.)
39 1
and quantitatively indistinguishable from that seen in intact, vehicle-treated animals. Support for this hypothesis has been provided in a study by Janss et al. (1987) in which the intrathecal administration of 6-OHDA in the rat, which depleted spinal NE content to approximately 12% of control, failed to affect the thresholds of stimulation in the LC/SC and medullary lateral reticular nucleus for inhibition of the nociceptive T F withdrawal reflex. Significant pharmacological a,-adrenoceptor supersensitivity developed within 3 days following the intrathecal administration of 6-OHDA, an observation supported by a concomitant significant increase in the number of a,-adrenoceptor binding sites in the lumbar spinal cord. Thus, the development of receptor supersensitivity is an additional factor which must be considered when estimating the contribution of spinal NE to LC/SC stimulation-produced inhibition of nociceptive transmission in both the rat and cat using neurotransmitter depletion techniques.
Physiological conclusions of LC / SC Clearly, an abundance of evidence indicates that endogenous descending inhibitory systems exist which, when activated by electrical stimulation or drugs, can modulate spinal nociceptive transmission. What peripheral stimuli activate these descending systems “naturally,” and whether they play an important physiological role in the modulation of nociceptive transmission, however, is not known. It is hypothesized that ascending nociceptive projection neurons (i.e., spinothalamic, spinoreticular and spinocervical neurons) send collaterals to brainstem nuclei involved in descending inhibition/antinociception; stimulation of peripheral nociceptors generates a positive feedback loop and neuronal activity is increased and maintained in supraspinal brainstem regions, which in turn leads to the enhancement of descending inhibition onto second order dorsal horn interneurons. The existence of a positive feedback loop between the spinal cord and the
medullary NRM has been demonstrated by Cervero and Wolstencroft (1 984). In decerebrate or decerebellate cats, extracellular recordings were made from spinal cord neurons located in or close to lamina VII of the dorsal horn; all neurons examined were excited by electrical stimulation in the NRM and/or adjacent reticular formation, had axonal projections to the NRM and/or reticular formation via spinal cord pathways in the ventrolateral quadrant, and were affected by intense (i.e., noxious) pressure applied to deep tissues of the limbs. Reciprocal excitatory connections were demonstrated between the NRM and neurons in lamina VII, establishing a positive feedback loop. That noxious peripheral stimuli can activate descending inhibitory systems also has been demonstrated. Electrical stimulation of the sciatic nerve in the cat, at intensities sufficient to activate A-6- and C-fibers, markedly increased the release of serotonin and NE in spinal cord superfusates. The increase in monoamine levels in the superfusate was attenuated significantly by a cold-block of the cervical spinal cord, suggesting activation of descending serotonergic and NA systems. The spinal release of serotonin and NE also could be evoked by stimulation of the infraorbital branch of the trigeminal nerve, indicating that small fiber afferent input from the entire body can likely drive activity in spinopetal monoaminergic systems (Tyce and Yaksh, 1981). Although the above-mentioned studies have not elucidated the supraspinal source of NE mediating the release of spinal NE and its antinociceptive effects, the LC/SC likely is involved since anatomical studies have demonstrated it to be the primary source of NE-containing nerve terminals in the lumbar spinal cord.
Acknowledgements Special thanks is extended to G.F. Gebhart for his comments regarding the manuscript. The author’s data reported here were supported by USPHS awards DA02879 and NS19912.
392
References Ader, J-P., Postema, F. and Korf, J. (1979) Contribution of the locus coeruleus to the adrenergic innervation of the rat spinal cord: A biochemical study. J. Neural Trunsm., 44: 159-173. Amaral, D.G. and Sinnamon, H.M. (1977) The locus coeruleus: Neurobiology of a central noradrenergic nucleus. Prog. Neurobiol. 9: 147-196. Belcher, G., Ryall, R.W. and Schaffner, R. (1978) The differential effects of 5-hydroxytryptamine, noradrenaline, and raphe stimulation on nociceptive and non-nociceptive dorsal horn interneurons in the cat. Bruin Res., 151: 307-321. Besson, J.M., Guilbaud, G.and LeBars, D. (1975) Descending inhibitory influences exerted by the brainstem upon the activities of dorsal horn lamina V cells induced by intraarterial bradykinin into the limbs. J. Physiol. (London), 248: 725-739. Bjorklund, A. and Skagerberg, G. (1982) Descending monoaminergic projections to the spinal cord. In B. Sjolund and A. Bjorklund (Eds.), Brain Stem Control of Spinal Mechanisms, Elsevier Biomedical Press, Amsterdam, pp. 55-88. Burill, P.Y., Goetzl, F.Y. and Ivy, A.C. (1944) The pain threshold raising effect of amphetamine. J. Dent. Res., 23: 337-344. Carlsson, A,, Magnusson, T. and Rosengren, E. (1963) 5-Hydroxytryptamine of the spinal cord normally and after transection. Experientiu, 19: 359. Carlton, S.M., Chung, J.M., Leonard, R.B. and Willis, W.D. (1985) Funicular trajectories of brainstem neurons projecting to the lumbar spinal cord in the monkey (Macaca fasicularis): A retrograde labeling study. J. Comp. Neurol., 241: 382-404. Cervero, F. and Wolstencroft, J.H. (1984) A positive feedback loop between spinal cord nociceptive pathways and antinociceptive areas of the cat’s brain stem. Pain, 20: 125-138. Commissiong, J.W., Hellstrom, S.O. and Neff, N.H. (1978) A new projection from the locus coeruleus to the spinal ventral columns: Histochemical and biochemical evidence. Bruin Res., 148: 207-213. Crawley, J.N., Roth, R.H. and Mass, J.W. (1979) Locus coeruleus stimulation increases noradrenergic metabolite levels in rat spinal cord. Brain Res., 166: 180-184. Duggan, A.W., Hall, J.G., Headley, P.M. and Griersmith, B.T. (1977) The effect of naloxone on the excitation of dorsal horn neurones of the cat by noxious and non-noxious cutaneous stimuli. Bruin Res., 138: 185-189. Engberg, I. and Ryall, R.W. (1966) The inhibitory action of noradrenaline and other monoamines on spinal neurones. J. Physiol, (London), 185: 298-322. Fillenz, M., Gagnon, C., Stoeckel, K. and Thoenen, H. (1976) Selective uptake and retrograde axonal transport of dopamine-P-hydroxylase antibodies in peripheral adrenergic neurons. Brain Rex, 114: 293-303. Fleetwood-Walker, S.M., Mitchell, R., Hope, P.J., Molony, V. and Iggo, A. (1985) An a,-receptor mediates the selective
inhibition by noradrenaline of nociceptive responses of identified dorsal horn neurons. Bruin Rex, 334: 243-254. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus ceruleus: New evidence of anatomical and physiological specificity. Physiol. Reu., 63: 844-914. Fritschy, J-M., Lyons, W.E., Mullen, C.A., Kosofsky, B.E., Molliver, M.E. and Grzanna, R. (1987) Distribution of locus coeruleus axons in the rat spinal cord: Combined anterograde transport and immunohistochemical study. Brain Res., 437: 176-180. Gebhart, G.F. (1982) Opiate and opioid peptide effects on brain stem neurons: Relevance to nociception and antinociceptive mechanisms. Pain, 12: 93-140. Guyenet, P.G. (1980) The coeruleospinal noradrenergic neurons: An anatomical and electrophysiological study in the rat. Bruin Res., 189: 121-133. Hammond, D.L. (1986) Control systems for nociceptive afferent processing: The descending inhibitory pathways. In T.L. Yaksh (Ed.), Spinul Afferent Processing, Plenum Press, New York, pp. 363-390. Handwerker, H.O., Iggo, A. and Zimmermann, M. (1975) Segmental and supraspinal actions on dorsal horn neurons responding to noxious and non-noxious skin stimuli. Pain, 1: 147-165. Headley, P.M., Duggan, A.W. and Griersmith, B.T. (1978) Selective reduction by noradrenaline and 5-hydroxytryptamine of nociceptive responses of cat dorsal horn neurons. Bruin Res., 145: 185-189. Hodge, C.J., Apkarian, A.V., Stevens, R.T., Vogelsang, G.D. and Wisnicki, H.J. (1981) Locus coeruleus modulation of dorsal horn unit responses to cutaneous stimulation. Bruin Res., 240: 415-420. Hodge, C.J., Apkarian, A.V., Stevens, R.T., Vogelsang, G.D., Brown, 0. and Franck, J.I. (1983) Dorsolateral pontine inhibition of dorsal horn cell responses to cutaneous stimulation: Lack of dependence on catecholaminergic systems in cat. J. Neurophysiol., 50: 1220-1235. Hosobuchi, Y., Adams, J.E. and Linchitz, R. (1977) Pain relief by electrical stimulation of the central gray matter in humans and its reversal by naloxone. Science, 197: 183-186. Howe, J.R. and Zieglgansberger, W. (1987) Responses of rat dorsal horn neurons to natural stimulation and to iontophoretically applied norepinephrine. J. Comp. Neurol., 255: 1-17. Howe, J.R., Wang, J-Y. and Yaksh, T.L. (1983) Selective antagonism of the antinociceptive effect of intrathecally applied alpha adrenergic agonists by intrathecal prazosin and intrathecal yohimbine. J. Phurmucol. Exp. Ther., 224: 552-558. Janss, A.J., Jones, S.L. and Gebhart, G.F. (1987) Effect of spinal norepinephrine depletion on descending inhibition of the tail-flick reflex from the locus coeruleus and lateral reticular nucleus in the rat. Brain Res., 400: 40-52. Jones, B.E. and Moore, R.Y. (1974) Catecholamine-containing neurons of the nucleus locus coeruleus in the cat. J. Comp. Neurol., 157: 43-52. Jones, B.E. and Yang, T-Z. (1985) The efferent projections from the reticular formation and the locus coeruleus stud-
393 ies by anterograde and retrograde axonal transport in the rat. J. Comp. Neurol. 242: 56-92. Jones, S.L. and Gebhart, G.F. (1986a) Characterization of coeruleospinal inhibition of the nociceptive tail-flick reflex in the rat: Mediation by spinal a,-adrenoceptors. Bruin Res., 364: 315-330. Jones, S.L. and Gebhart, G.F. (1986b) Quantitative characterization of ceruleospinal inhibition of nociceptive transmission in the rat. J. Neurophysiol., 56: 1397-1410. Jones, S.L. and Gebhart, G.F. (1987) Spinal pathways mediating tonic, coeruleospinal, and raphe-spinal descending inhibition in the rat. J. Neurophysiol., 58: 138-159. Kuraishi, Y., Harada, Y. and Takagi, H. (1979) Noradrenaline regulation of pain-transmission in the spinal cord is mediated by a-adrenoceptors. Brain Res., 174: 333-336. Kuypers, H.G.J.M. and Maisky, V.A. (1975) Retrograde axonal transport of horseradish peroxidase from spinal cord to brainstem cell groups in the cat. Neurosci. Lett., 1: 9-14. Lai, Y-Y. and Barnes, C.D. (1985) A spinal projection of serotonergic neurons of the locus coeruleus in the cat. Neurosci. Lett., 58: 159-164. Light, A.R. and Perl, E.R. (1979a) Reexamination of the dorsal root projection to the spinal dorsal horn including observations on the differential termination of coarse and fine fibers. J. Comp. Neurol., 186: 117-132. Light, A.R. and Perl, E.R. (1979b) Spinal terminations of functionally identified primary afferent neurones within slowly conducting myelinated fibers. J. Comp. Neurol., 186: 133- 1SO. Loughlin, S.E., Foote, S.L. and Bloom, F.E. (1986a) Efferent projections of nucleus locus coeruleus: Topographic organization of cells of origin demonstrated by three-dimensional reconstruction. Neuroscience, 18: 291-306. Loughlin, S.E., Foote, S.L. and Grzanna, R. (1986b) Efferent projections of nucleus locus coeruleus: Morphologic subpopulations have different efferent targets. Neuroscience, 18: 307-319. Magnusson, T. and Rosengren, E. (1963) Catecholamines of the spinal cord normally and after transection. Experientiu, 19: 229. Margalit, D. and Segal, M. (1979) A pharmacologic study of analgesia produced by stimulation of the nucleus locus coeruleus. Psychophurmucology, 62: 169-173. Mayer, D.J. (1979) Endogenous analgesia systems: Neural and behavioral mechanisms. In J.C. Liebeskind and D.G. Albe-Fessard (Eds.), Advances in Pain Research and Therapy, Voi. 3, Raven Press, New York, pp.385-410. Mokha, S.S. and Iggo, A. (1987) Mechanisms mediating the brain stem control of somatosensory transmission in the dorsal horn of the cat’s spinal cord: An intracellular analysis. Exp. Bruin Res., 69: 93-106. Mokha, S.S., McMillan, J.A. and Iggo, A. (1985) Descending control of spinal nociceptive transmission. Actions produced on spinal multireceptive neurones from the nuclei locus coeruleus (LC) and raphe magnus (NRM). Exp. Bruin Res., 58: 213-225. Mokha, S.S., McMillan, J.A. and Iggo, A. (1986) Pathways
mediating descending control of spinal nociceptive transmission from the nuclei locus coeruleus (LC) and raphe magnus (NRM) in the cat. Exp. Bruin Res., 61: 597-606. Moore, R.Y. and Bloom, F.E. (1979) Central catecholamine neuron systems: Anatomy and physiology of the norepinephrine and epinephrine systems. Ann. Reu. Neurosci., 2: 113-168. North, R.A. and Yoshimura, M. (1984) The actions of noradrenaline on neurones of the rat substantia gelatinosa in uitro. J. Physiol. (London), 349: 43-55. Nygren, L.G. and Olson, L.A. (1977) A new major projection from locus coeruleus: The main source of noradrenergic nerve terminals in the ventral and dorsal columns of the spinal cord. Bruin Res., 132: 85-93. Olson, L. and Fuxe, K. (1972) Further mapping out of central noradrenaline neuron systems: Projections of the “subcoeruleus” area. Bruin Res., 43: 289-295. Proudfit, H.K. (1988) Pharmacologic evidence for the modulation of nociception by noradrenergic neurons. In H.L. Fields and J.M. Besson (Eds.), Progress in Bruin Research, Vol. 77, Elsevier, Amsterdam, pp. 357-370. Reddy, S.V.R., Maderdrut, J.L. and Yaksh, T.L. (1980) Spinal cord pharmacology of adrenergic agonist-mediated antinociception. J. Phurmucol. Exp. Ther., 213: 525-533. Segal, M. and Sandberg, D. (1977) Analgesia produced by electrical stimulation of catecholamine nuclei in the rat brain. Brain Res., 123: 369-372. Seybold, V.S. (1986) Neurotransmitter receptor sites in the spinal cord. In T.L. Yaksh (Ed.), Spinal Afferent Processing, Plenum Press, New York, pp. 117-139. Seybold, V.S. and Elde, R.P. (1984) Receptor autoradiography in thoracic spinal cord: Correlation of neurotransmitter binding sites with sympathoadrenal neurons. J. Neurosci., 4: 2533-2542. Soja, P.J. and Sinclair, J.G. (1983) Tonic descending influences on cat spinal dorsal horn neurons. Somufosens. Res., 1: 83-93. Stevens, R.T., Hodge, C.J. and Apkarian, A.V. (1982) Kolliker-Fuse nucleus: The principal source of pontine catecholaminergic cells projecting to the lumbar spinal cord of the cat. Bruin Res., 39: 589-594. Stevens, R.T., Apkarian, A.V. and Hodge, C.J. (1985) Funicular course of catecholamine fibers innervating the lumbar spinal cord of the cat. Bruin Res., 336: 243-251. Todd, A.J. and Millar, J. (1983) Receptive fields and responses to iontophoretically applied noradrenaline and 5-hydroxytryptamine of units recorded in laminae 1-111 of cat dorsal horn. Bruin Res., 288: 159-167. Tyce, G.M. and Yaksh, T.L. (1981) Monoamine release from cat spinal cord by somatic stimuli: An intrinsic modulatory system. J. Physiol. (London), 314: 513-529. Unnerstall, J.R., Kopajtic, T.R. and Kuhar, M.J. (1984) Distribution of a 2 agonist binding sites in the rat and human central nervous system: Analysis of some functional anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agonists. Bruin Res. Reu., 7: 69- 101. Westlund, K.N. and Coulter, J.D. (1980) Descending projec-
394 tions of the locus coeruleus and subcoeruleus/medial parabrachial nuclei in monkey: Axonal transport studies and dopamine-P-hydroxylase immunocytochemistry. Bruin Res. Reu., 2: 235-264. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of the rat. Bruin Rex, 263: 15-31. Willcockson, W.S., Chung, J.M., Hori, Y., Lee, K.H. and Willis, W.D. (1984) Effects of iontophoretically released amino acids and amines on primate spinothalamic cells. J. Neurosci., 4: 732-740. Yaksh, T.L. (1985) Pharmacology of spinal adrenergic systems
which modulate spinal nociceptive processing. Pharmncol. Biochem. Behau. 22: 845-858. Yaksh, T.L. and Rudy, T.A. (1978) Narcotic analgetics: CNS sites and mechanisms of action as revealed by intracerebra1 injection techniques. Pain, 4: 299-359. Yaksh, T.L. and Wilson, P.R. (1979) Spinal serotonin terminal system mediates antinociception. J. Pharrnacol. Exp. Ther., 208: 446-453. Ziegler, M.G., Thomas, J.A. and Jacobowitz, D.M. (1976) Retrograde axonal transport of antibody to dopamine-Phydroxylase. Brain Res., 104: 390-395.
C.D. Barnes and 0. Pompeiano (Eds.) Progres, in Bruin Research, Vol. 88 0 1YY1 hlsevier Science Publishera B.V.
395 CHAPTER 30
Locus coeruleus control of spinal motor output S.J. Fung
D. Manzoni
2,
J.Y.H. Chan 3 , 0. Pompeiano
and C.D. Barnes
’
Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, College of Veterinary Medicine, Washington State Uniuersity, Pullinan, WA, U.S.A., Department of Physiology and Biochemistry, University of Pisa, Via S. Zeno, Pisa, Italy and Department of Medical Research, Veterans General Hospital, Taipei, Taiwan, China
Using electrophysiological techniques, we investigated the functional properties of the coeruleospinal system for regulating the somatomotor outflow at lumbar cord levels. Many of the fast-conducting, antidromically activated coeruleospinal units were shown to exhibit the a,-receptor response common to noradrenergic locus coeruleus (LC) neurons. Electrically activating the coeruleospinal system potentiated the lumbar monosynaptic reflex and depolarized hindlimb flexor and extensor motoneurons via an a,-receptor mechanism. The latter synaptically induced m e m b r a n e depolarization was
mimicked by norepinephrine applied iontophoretically to motoneurons. That LC inhibited Renshaw cell activity and induced a positive dorsal root potential at the lumbar cord also reinforced LC’s action on motor excitation. We conclude that LC augments the somatomotor output, at least in part, via an a,-adrenoceptor-mediated excitation of ventral horn motoneurons. Such process is being strengthened by LC’s suppression of the recurrent inhibition pathway as well as by its presynaptic facilitation of afferent impulse transmission at the spinal cord level.
Key words: locus coeruleus, norepinephrine, spinal cord, motoneuron, Renshaw cell, presynaptic facilitation
Introduction
Early stimulation studies indicated that the brainstem, although hierarchically lower than the cortical centers, plays a role in influencing the spinal motor pathways. In 1932, Ingram et al. reported a wide region within the pontomesencephalic tegmentum which, when stimulated, produced muscle movements in the anesthetized cat. Because the current spread in this study was broad, as judged by various stimulation-produced autonomic and motor responses, it was difficult to localize the supraspinal source(s) of movement
control. Using a stimulus strength which was subthreshold for inducing somatic motor movement, Rhines and Magoun (1946) found the corresponding brainstem reticular core to facilitate both cortically and reflexly induced fore- and hindlimb movements in anesthetized cats and monkeys. The notion that spinal regulatory systems exist in the brainstem, as well as in higher neural centers, was clearly an important advance. Over the past four decades, it has become clear that there is no single, central (reticular) core of brainstem regulatory systems. Rather, multiple, although ultimately interacting, systems
3 96
are involved in the descending control of spinal motor mechanisms. The dorsolateral pontine coeruleospinal system, in accord with recent anatomical (Kuypers, 1982; Holstege and Kuypers, 1987) and physiological (Marshall, 1983; Fung and Barnes, 1984; Barnes et al., 1989) evidence, appears to play a pivotal role in controlling the spinal motor pathway of mammals. In our laboratories the search for a functional link between the locus coeruleus (LC) and spinal cord motoneurons has yielded some answers but has also raised new questions. This chapter sum-
marizes established facts, and introduces new findings that point to the complexity of the descending LC control mechanism. Electrophysiological verification of coeruleospinal projections
Early in our investigations we sought to verify, physiologically, the existence of coeruleospinal projections in cats (for anatomical reviews, see Kuypers, 1982; Holstege and Kuypers, 1987) by utilizing antidromic activation techniques. This 4 If 2
A
F
C
H
I
I
., . ..
,I
* . *
.' . ... . .. .. .... ... .. : . ... a .
E
+
1
. . ..
a
.
Fig. 1. Firing properties of coeruleospinal neurons. A-C are for one neuron and D-J for another neuron. A. Spontaneous activity. B. Pinching applied to ipsilateral hind paw throughout the 20-sec sweep. C. Immediately following B. D-E. Raster-dot displays of the spontaneous activity and response to sural shock (at arrow), respectively. Note excitation-inhibition pattern in E. Each picture shows 32 sweeps running successively from the top downwards. F. Four consecutive responses at 100 Hz cord stimulation as numbered. Fractionation occurs in all but the first trace. G. Antidromic response to cord stimulation (at dot). H. A spontaneous spike occurring outside the critical period. Collision begins when a spontaneous spike falls within the critical period either before (I) or after (J) the cord stimulation, blocking the occurrence of the antidromic response (at arrows). Upward deflections are positive. Calibrations: 2 sec, 200 pV for A-C; 100 msec for D-E; 2 msec, 200 pV for F; 10 msec, 500 pV for G-J. (From Fung and Barnes, 1981.)
397
physiological approach is not only best suited for such verification purposes, but it also provides information about the axonal conduction velocity of the descending tracts, as well as the pharmacological properties peculiar to the coeruleospinal neurons. In our previous studies on decerebrate cats (Barnes et al., 1980; Fung and Barnes, 19811, single shocks (0.1 msec pulses, 1/sec, 15-1000 p A intensities) were delivered to the ventral horn of the 7th lumbar segment via a concentric bipolar stimulating electrode (0.5 mm interpolar distance, core pole being negative) while unitary activity was sought from the ipsilateral LC region. LC units were identified by: (1) their slow and sporadic spontaneous firing rates (e.g., Fig. 1A, 3 Hz; Fig. 1D, 3.6 Hz) similar to those described by others (cat: Chu and Bloom, 1974; Hobson et al., 1975; Sakai, 1980; Watabe et al., 1982; rat: Guyenet, 1980; Foote et at., 1983; monkey: Grant and Redmond, 1984; Aston-Jones et al., 1985; Grant et al., 1988); (2) the postexcitation-inhibition response upon leg pinching (Fig. 1B-C) or high-intensity sural nerve shock (Fig. 1E). The postactivation pause in firing is characteristic of a,-receptor-mediated hyperpolarization of noradrenergic LC neurons in viuo (Aghajanian and VanderMaelen, 1982; Aghajanian et al., 1983). Non-coerulear units are either nonresponsive (e.g., mesencephalic trigeminal neurons) or without the autoinhibition pattern (e.g., pontine reticular neurons; Wolstencroft, 1964) to the pinch stimulus. The identified LC units were subsequently shown to project to the lumbar ventral horn, according to tests for collision of the antidromic response by a conditioning spontaneous spike within a critical period (Fig. 1H-J). According to Fuller and Schlag (19761, such a collision period is attributable to the sum of the antidromic latency and the time of recovery ( i e . , refractory period) at the point of stimulation in the cord. In corollary, both spontaneous and antidromic responses were recordable from the LC soma at an interval which slightly exceeded (i.e., 1 msec) twice the antidromic latency (Fung and
Barnes, 1984). Other features such as constant antidromic latency (Fig. 1G,H) and highfrequency following (e.g., 100 Hz; Fig. 1F) with frequent fractionation of the antidromic spike waveform (Fig. 1F) were secondary as far as antidromicity is concerned (Lipski, 1981). Using similar methods, other authors have identified the coeruleospinal units in cats (Nakazato, 1987; Pompeiano et al., 1990) and rats (Guyenet, 1980). Finding the antidromically activated LC units in cats complements the anatomical anterograde (Holstege et al., 1979) and retrograde (Kuypers and Maisky, 1975; Hancock and Fougerousse, 1976; Basbaum and Fields, 1979; Hayes and Rustioni, 1981; Stevens et al., 1982; Kausz, 1986; Reddy et al., 1989,1990) tract-tracing studies, in that coeruleospinal axons course ipsilaterally to innervate at least the lumbar cord of cats. The spread of retrograde tracer, in most studies, has included various portions of the spinal gray and the contiguous white matter, making it difficult to determine the extent of the coeruleospinal terminal fields. Anterograde labeling techniques, however, have shown that coeruleospinal axons terminate in the dorsal, intermediate, and ventral gray columns in cats (Holstege et al., 1979). In our studies (Barnes et al., 1980; Fung and Barnes, 1981) there were instances where LC units could be activated with very low current pulses (ie., 15 p A) delivered to the ventrolateral gray column of the lumbar cord; this suggested that those coeruleospinal axons probably traverse close to the motoneuron region of the cord. The feline coeruleospinal system is known to comprise both crossed and uncrossed descending fibers (Hancock and Fougerousse, 1976; Kuypers and Maisky, 1977; Hayes and Rustioni, 1981; Stevens et al., 1982; Kausz, 19861, with ipsilateral predominance. Such a bilateral innervation pattern has also been tested electrophysiologically; tests revealed that stimulation of either side of the LC is effective in potentiating the lumbar monosynaptic reflex (MSR) in decerebrate cats (Strahlendorf et al., 1980; Lai et al., 1989). To examine the pharmacological responsive-
-
398
ness of the coeruleospinal neurons, two a,adrenergic agonists (cyclobenzaprine and clonidine) and one antagonist (piperoxane) were used to test the antidromic response in decerebrate cats (Barnes et al., 1980; Fung and Barnes, 1984). The choices of clonidine and piperoxane were based upon their known a,-adrenergic agonistic and antagonistic properties, respectively, in rat LC neurons (Aghajanian and VanderMaelen, 1982). Cyclobenzaprine was also of interest for the following reasons: Based on catecholamine depletion (by reserpine and tetrabenazine) and adrenergic blockade (by phenoxybenzamine) studies, it appears that an intermediary adrenergic descending system mediates the vestibular (8th nerve) volley-induced facilitation of the extensor MSR in decerebrate cats (Barnes and Pompeiano, 1971). This is consistent with the recent finding of such an anatomical link (i.e., the vestibulo-coerulear projection) is likely to occur in cats and rats (Fung et at., 1987b). Cyclobenzaprine’s lack of effect on spinal motoneuron activity has also led to the notion that the drug probably acts at the supraspinal level to produce its antispastic effects in cats (Esplin and Capek, 1979). Studies by Barnes et al. (1980) showed that the noradrenergic coeruleospinal system may well be affected by this drug. As depicted in Figure 2, cyclobenzaprine administration (0.5 mg/kg, iv) clearly suppressed both spontaneous (top panels) and pinch-evoked responses (middle panels). Our current results are at variance with another study (Commissiong et al., 1981) which showed an excitatory effect of cyclobenzaprine on spontaneous LC unit activity in anesthetized rats. If the drug is an effective muscle relaxant, this latter finding is difficult to reconcile with the known LC-induced facilitation of lumbar MSR in both the rat (Chan et al., 1986) and cat (Strahlendorf et al., 1980; Lai et al., 1989). Further, without the standard a*adrenergic agonist and antagonist tests (see below), it is uncertain if Commissiong et al.’s reported cells are of non-coerulear origin. Despite potent inhibition by the drug, the recorded coeruleospinal unit was found to respond with
C
CBZ
A
0531796
I
I
Fig. 2. Effect of cyclobenzaprine (CBZ) on locus coeruleus (LC) cell with long descending axon. Column C: control recordings before any drug. Column CBZ: 15-18 min after 0.5 mg/kg of CBZ.A. Spontaneous activity. B. Cell response to pinching between the toes of the right hind foot during the period of the underlying bar. C. Five superimposed traces of the cell being activated antidromically by a 0.1 msec, 100 p A stimulus applied to the left ventral horn of L7. Calibration: A and B, 2 sec, 200 pV; C, 2 msec, 100 pV. (From Barnes et al., 1980.)
diminished amplitude to the antidromic (lumbar cord) volley (see Fig. 2, bottom panels). As with cyclobenzaprine, clonidine administration (5 kg/kg, iv) reduced the antidromic spike height. Subsequent piperoxane injection (0.5 mg/kg, iv) reversed clonidine’s effect and increased the amplitude and upstroke of the antidromic LC response to higher levels than those present before clonidine treatment (Fung and Barnes, 1984). Since no change in antidromic response latencies occurred following any of these drug treatments, one can assume that the effective interacting loci are not located on the axon but reside on the soma membrane. This is important because binding studies have revealed that a,-receptors are present in the LC (Young and Kuhar, 1980) as well as in the spinal cord, including the dorsal and the ventral horns (Simmons and Jones, 1988). Intracellular studies have presented evidence of an a,-receptor-mediated hyperpolarization of rat noradrenergic LC neurons in civo after clonidine administration (Aghajanian and VanderMaelen, 1982). These findings support the notion that coeruleospinal neurons are probably similar to
39Y
the noradrenergic LC neurons in terms of their intrinsic regulation of firing via the a,-receptor mechanism. This is consistent with the rat data in which the responses of the antidromically activated coeruleospinal units to a,-agonists were identical to those of physiologically identified LC neurons (Guyenet, 1980). The typical a,-agonistantagonist responses elicited by our antidromic unitary responses further ascertained that they were indeed noradrenergic coeruleospinal units rather than other pontospinal (e.g., reticulospinal) cells. Since a,-adrenergic agonists have been found to bind also with noncatecholaminergic receptor sites in the medulla (Bricca et al., 1989), our data must be interpreted with caution unless the nonaminergic LC and pontine reticulospinal neurons are proven to be devoid of any significant a,-mediated reaction. Our antidromic activation study also provides information about the conduction speed of the coeruleospinal axons (Fung and Barnes, 1981). From a sample of 25 coeruleospinal units, mean velocities were calculated to be 20 5 8 m/sec (kS.D.), with a range of 7-32 m/sec. A n even wider spectrum of conduction velocities corresponding to both myelinated and nonmyelinated coeruleospinal axons is also evident from another study in cats (Nakazato, 1987). Using the same (antidromic activation) technique, a third group of workers also reported a high (48.7 m/sec; Pompeiano et al., 1990) average conduction speed for the feline coeruleospinal axons. While the actual size of the coeruleospinal axons has not been investigated, morphometric analyses in the cat revealed the mean diameter of the tyrosine hydroxylase-immunoreactive coeruleospinal neurons to be 18 p m size, compared to 20 p m size for non - tyrosine hyd roxylase-immunore active coeruleospinal counterparts (Reddy et al., 1989). Our results strongly suggest that the feline coeruleospinal fibers are probably larger in diameter (and hence conduct faster) than earlier reported for the rat. The recent ultrastructural finding that many (88%) of the rat serotonergic spinal fibers are myelinated (Araneda et al., 1989)
also does not conform to the classical concept that most, if not all, aminergic fibers are nonmyelinated (Dahlstrom and Fuxe, 1965) and consequently conduct at speeds of less than 1 m/sec. In support of our interpretation, a myelinated coeruleospinal tract in humans has briefly been mentioned (Papez, 1925). By contrast, the coeruleospinal system in the rat has been shown to have much slower conduction rates (less than 1 m/sec; Guyenet, 19801, similar to those characteristic of the ascending LC projections in rats (Foote et al., 1983), cats (McBride and Sutin, 1976; Watabe et al., 1982) and monkeys (German and Fetz, 1976). Potentiation of lumbar MSRs Several lines of evidence have suggested a descending noradrenergic coeruleospinal facilitation of motor activity. As discussed earlier, Barnes and Pompeiano (1971) presented evidence of a descending adrenergic system which acts in parallel to the vestibulospinal activation of hindlimb extensor tone in decerebrate cats. Our findings of the presumptive excitatory coeruleospinal neurons being suppressed by cyclobenzaprine (Barnes et al., 1980) is compatible with the drug’s known muscle relaxant effect. The drastic diminution in postdecapitation reflex following prior LC lesioning in rats (Suenaga et al., 1979; Pappas et al., 1980) also points to the involvement of the excitatory, rather than an inhibitory, coeruleospinal system in motor control. Though it has previously been shown that the catecholamine precursor Ldopa potentiates cat spinal MSR (Baker and Anderson, 19701, later studies have suggested possible transmitter release other than norepinephrine (NE) (for details, see Marshall, 1983). In this context, it has been shown that the L-dopa action on lumbosacral MSR is sensitive to dopamine blockade (by pimozide) in spinal cats (Geber and Dupelj, 1977; Dupelj and Geber, 1981). As far as dopamine is concerned, a similar study has reported instead a dopamine receptor-mediated de-
400
pression of MSR in spinal cats (Carp and Anderson, 1982). In addition to dopamine’s involvement, serotonin has been reported to be released endogenously following systemic L-dopa (Butcher et al., 1972). In view of these findings, Strahlendorf et al. (1980) first described a common facilitation of flexor and extensor MSRs upon stimulating the LC in decerebrate cats. That this facilitation was sensitive to the mixed aminergic blockers phenoxybenzamine and chlorpromazine, but not the dopamine blocker haloperidol, led to the suggestion that it was mediated by noradrenergic rather than dopaminergic coeruleospinal synapses. Using a more selective a,-adrenergic antagonist (prazosin), studies have reaffirmed a noradrenergic role in the LC facilitation of lumbar MSRs (Chan et al., 1986; Lai et al., 1989). The notion of an excitatory a,-noradrenergic coeruleospinal action agrees with the evidence in other central neurons discussed by Szabadi (1979). Anatomically, i; is known that the feline coeruleospinal system is comprised of a major noradrenergic (Reddy et al., 1989, 1990), as well as a minor serotonergic, component (Lai and Barnes, 1985; see Reddy et al., this volume). The latter finding is supported by the study, in decerebrate cats, which revealed a decremental change in LC-induced potentiation in lumbar MSRs following sequential blockade by prazosin and methysergide (a serotonin antagonist) (Lai et al., 1989). That both NE and serotonin have been shown to excite, intracellularly, cat spinal motoneurons (White and Fung, 1989; White et al., this volume) accords with these findings. In addition to the LC source, the possibility also exists that such serotonergic facilitatory action could be mediated partially through the intermediary raphe-spinal system. In this regard, a serotonergic raphe-spinal motor system (Anderson, 1972; Kuypers, 1982) as well as the raphe-spinal-induced motoneuron depolarization have already been established based on studies in cats (Holtman et al., 1986; Lalley, 1986; Fung and Barnes, 1989; Edamura and Aoki, 1989).
Motoneuron responses to synaptically and iontophoretically released NE
Findings from reflex studies intensified our desire to elucidate the cellular mechanisms attending the coeruleospinal enhancement of motor outflow. The intracellular recording combined with extracellular iontophoresis techniques have made it possible to examine the effects of synaptically (via LC stimulation) and locally released NE on motoneuron membrane potential and excitability changes in decerebrate cats (for methodological details and histological controls of stimulation sites, two papers should be consulted: Fung and Barnes, 1987a; White and Fung, 1989). In all stimulation studies, our stringent verification procedures regarding LC electrode placements (Fung and Barnes, 1984, 1987a) reduce, but do not eliminate, the errors introduced by non-coerulear source(s) of spinal action caused by inadvertent co-activation of fibers en passage (for problems of current spread, see Fung and Barnes, 1987a). It has been demonstrated that chemically inactivating the LC perikarya with the a,-agonist clonidine results in hypotonus of the forelimb extensors in decerebrate cats (Pompeiano et al., 1987). This strengthens the notion that the excitatory spinal motor action derives supraspinally, at least in part, from coeruleospinal neurons. In our studies, although stimulation sites were variously localized in the LC dorsalis and L C a regions, there was an unequivocal depolarizing membrane response of motoneurons, both flexors and extensors alike, to LC stimulation (Fung and Barnes, 1987a). The fact that it continues to be depolarizing when recorded with microelectrodes filled with chloride-free electrolytes (Fung and Barnes, 1987a), and that it functionally reduces the motoneuron firing threshold from other inputs on stimulation (Fung and Barnes, 1986, 1987b), shows it to be an excitatory postsynaptic potential (EPSP) simiIar to that produced by muscle afferents on cat spinal motoneurons (Eccles, 1964). In contrast to tetanic LC stimuli, a
40 1
I
1
Fig. 3. Excitatory motoneuronal responses evoked by single (A,C) and repetitive (B,D) LC stimuli (dots). A,B. Eight traces of LC-induced synaptic responses. C,D. Averaged waveforms of A and B, respectively. Note the simple, low-amplitude excitatory postsynaptic potential (EPSP) (C) following single shock at 200 p A to the LC and the larger-sized step-like EPSP (arrows, D) resulting from 6 stimuli of 1.3 msec interpulse interval at 300 p A to the same site. Gastrocnemius medialis a-motoneuron: antidromic spike = 84 mV. Calibrations: 2 msec, 1 mV for A and C; 2 msec, 5 mV for B and D.
single shock was much less consistent in producing an EPSP of significant amplitude from the motoneuron (Fig. 3A,C). There was an all-or-none EPSP (amplitude = 0.6 & 0.2 mV, range = 0.2-0.9 mV; latency = 4.6 + 0.4 msec, range = 4.1-5.2 msec; n = 10 cells) of a simple time course similar to the one depicted in Figure 3C. Such inconsistency in motoneuron response has been documented for other descending systems, including the monosynaptic vestibulospinal (Grillner et al., 1970) and medial longitudinal fasciculus in cats (Grillner et al., 1971) and rubrospinal paths in monkeys (Shapovalov et al., 1971). Tetanic stimuli, in this and a few other cells, evoked a step-like EPSP as shown in Figure 3D. Others have reported that there is a marked increase in amplitudes of successive monosynaptic EPSPs in response to tetanic volleys to the baboon pyramidal tract (Langren et al., 19621, the monkey precen-
tral gyrus (Porter and Hore, 1969; Muir and Porter, 19731, and the cat medial longitudinal fasciculus (Wilson and Yoshida, 1969). In addition, temporal summation, with a typical increase in EPSP amplitudes in response to successive shocks, has been demonstrated in the polysynaptic pathways linking the red nucleus to the hindlimb motoneurons (Hongo et al., 1969) and the labyrinthine receptors to the forelimb motoneurons in cat (Maeda et d., 1975). Thus, the question remains whether the single shock-induced EPSP and the tetanic stimuli-induced temporal summation features shown in Figure 3 could be entirely attributable to either the mono- or oligosynaptic coeruleospinal pathways. Despite these uncertainties, other evidence supports the existence of a monosynaptic link for this system. Based on the identical latencies of the EPSP responses to single and tetanic LC stimuli (5.4 msec in Fig. 3C,D) and the high-frequency following of successive EPSPs to tetanic LC volleys (1.3 msec interpulse interval or 770 Hz approximately) (Fig. 3D), it is possible to infer a monosynaptic connection from the coeruleospinal preterminals to the motoneuron (Berry and Pentreath, 1976). The ability of coeruleospinal neurons to be activated antidromically by ventral horn stimulation (Barnes et al., 1980; Fung and Barnes, 1981) is in accord with this suggestion. Other supportive evidence derives from fluorescence, immunocytochemical and ultrastructural findings which demonstrate that some aminergic axonal terminals, presumably of supraspinal origin, synapse with cell bodies, inter a h , of a-motoneurons in the ventral horn of rat (Dahlstrom and Fuxe, 1965; Commissiong et al., 1978; Zhan et al., 1989) and cat (Jordan et al., 1977; Mizukawa, 1980; Mizukawa and Takeuchi, 1982). This is compatible with anatomical findings of aminergic coeruleospinal neurons (Lai and Barnes, 1985; Reddy et al., 1989, 1990, this volume) innervating spinal neurons, including motoneuron groups in cats (Holstege et al., 1979). Previously we have described four major patterns of motoneuronal responses upon LC stimu-
402
lation: a simple EPSP succeeded by weak, prolonged depolarization, a double-peak EPSP, an EPSP succeeded by a weak hyperpolarization, or a slow-rising EPSP (Fung and Barnes, 1987a). Among all motoneurons tested, the commonly induced, initial LC EPSP has been determined to be accompanied by an enhanced membrane excitability, including a reduced rheobase (Fung and Barnes, 1987a) and an enhanced initial segment-somatodendritic coupling for spike generation (Fung and Barnes, 1986). A similar excitatory LC drive, upon motoneuron repetitive firing, has also been shown to involve an inhibition of LC on post-spike after-hyperpolarizing potentials (Fung and Barnes, 1987a,b). In view of the immunostaining and physiological (single unit and MSR) findings discussed, the possible mediation of the stimulation-produced excitation of motoneurons via aminergic coeruleospinal synapses was tested intracellularly in one motoneuron by systemically injecting the serotonin antagonist methysergide (2 mg/kg) followed by the mixed adrenoceptor antagonist phenoxybenzamine (2 mg/kg). Methysergide lowered slightly the LC-EPSP amplitude and half-width as in Figure 4 (middle panel); phenoxybenzamine eliminated the EPSP completely (Fig. 4, bottom panel). While the aminergic coeruleospinal synapses were blocked functionally by these agents, the motoneuron remained excitable to antidromic activation from its axon (spike height = 69 mV after the second drug injections). These data suggest that the descending LC effect is dominated by the noradrenergic synapses but mediated synergistically by serotonin transmission. This is in line with data showing that serotonincontaining LC cells are scarce in cat (Wiklund et al., 1981) and constitute only a minor component within the feline coeruleospinal system (Lai and Barnes, 1985; Reddy et al., this volume). Phenoxybenzamine was tested in another motoneuron and observed to reduce the LC EPSP with partial recovery. In two other motoneurons, systemic administration of the selective a ,-antagonist prazosin (20 pg/kg) similarly reduced the LC
Fig. 4. Antagonistic effects of the serotonergic (methysergide) and a-adrenergic (phenoxybenzamine) blockers on LC-induced EPSP in a gastrocnemius medialis a-motoneuron (same cell as in Fig. 3). All traces are averaged high-gain D C recordings ( n = 16) following the LC stimulation (200 PA, 6 pulses; arrows). Top panel: pre-drug control of LC-EPSP. Middle panel: slight diminution in LC-EPSP 6 min after an accumulated dose of 2 mg/kg iv of methysergide. Time elapsed in this record was 35 min from the initial methysergide injection. Bottom panel: abolition of LC-EPSP due to an accumulated dose of 3 mg/kg iv of phenoxybenzamine. This record was taken 30 min after the middle panel record. Antidromic spike height was 69 mV after the bottom panel record was taken. Calibrations: 50 msec. 5 mV.
EPSP. These intracellular results are in accord with previous stimulation studies, in cat (Lai et al., 1989) and rat (Chan et al., 19861, which demonstrated that the LC-induced potentiation of lumbar MSR is markedly sensitive to prazosin blockade, suggesting their mediation via, at least in part, the a,-receptor mechanism. Although the LC's excitation of motoneurons partially involves the noradrenergic synapses, the underlying ionic mechanism for such an effect is as yet unknown. Membrane resistance changes underlying the four categories of LC EPSPs have been examined in cat hindlimb motoneurons of flexor and extensor origins (for methodological
403
LC 6 0 p A
30s
I dl
-NE45
WJ
Fig. 5. Parallel depolarizing changes, within the same motoneuron, produced by tetanic LC stimuli (84 pulses at 500 Hz) and iontophoretically applied norepinephrine (NE) (0.1 M, pFI 4). The initial hyperpolarization in the right top panel was inconsistent and non-reproducible throughout successive NE trials. Note differential changes in motoneuron input resistance (bottom panels) accompanying the LC- (left top panel) and NE-induced (right top panel) depolarizing responses. All traces (except the right top) are averages of 20 traces. Input resistance was determined by passing hyperpolarizing constant current (-3.1 nA) pulses (at 1 Hz) of 140 msec duration through the intracellular electrode. In the left bottom panel, the LC trace was obtained by graphically subtracting the LC-EPSP from the apparent voltage-drop during the LC-EPSP. Antidromic motoneuron spike height = 90 mV.
details, see Fung and Barnes, 1987a). A general decrease of 21.2 i 8.1% (range = 11-37%, n = 11 cells) in motoneuron input resistance was observed during the LC EPSP. In one recent study, however, iontophoretically released NE was shown to cause a marked increase ( i e . , up to 30%) in motoneuron input resistance (White et al., this volume). When these two manipulations were tested simultaneously in three individual motoneurons, results similar to the above were observed. Figure 5 illustrates the differential changes in motoneuron input resistance (bottom panels) accompanying the LC-(left top panel) and NE-induced (right top panel) depolarizing responses. This cell exhibited a 17% decrease and 27% increase in input resistance due to LC stimulation and iontophoretically released NE, respectively. The NE-induced depolarization, accompa-
nied by an increase in membrane resistance, is evident in many neurons including the dorsal horn substantia gelatinosa neurons (North and Yoshimura, 19841, cultured spinal cord neurons (Pun et al., 1985; Legendre et al., 1988), vesicle parasympathetic neurons (Akasu et al., 1989, dorsal motor nucleus of vagus (Fukuda et al., 19871, hippocampal granule cells (Lacaille and Schwartzkroin, 19881, thalamocortical neurons (McCormick and Prince, 1988), and dorsal root ganglion cells (Todorovic and Anderson, 1990). A decrease in potassium conductance is implicated in the NE-induced membrane depolarization. In contrast, N E has also been reported to depolarize the neocortical (Foehring et al., 1989) and hypothalamic supraoptic (Yamashita et al., 1987) neurons with no change or a decrease in input resistance, respectively. In our study, the apparent differences between LC- and NE-induced input resistance changes are easily reconciled by taking into account the mixed activation (via LC stimulation) of noradrenergic plus other putative LC transmitters which caused a net (dominant) decrease rather than the previously observed NE-induced increase in motoneuron membrane resistance. In this context, it is known that the chemical organization of the feline coeruleospinal system is very heterogeneous, comprising at least NE, enkephalin, neuropeptide Y, and serotonin (Reddy et al., this volume). In 8 of 37 cultured mouse spinal neurons tested, enkephalin caused abrupt depolarizations of membrane potential accompanied by an increase in inward current (Barker et al., 1978). In our study, if the cotransmitter enkephalin increased the conductance to sodium or calcium ions an N E EPSP would still be evident, but the net membrane resistance could be reversed. Also, by LC stimulation, it is possible to include non-coerulear (via co-activation) or intermediary descending systems which then caused a net decrease in motoneuron resistance. Additionally, the latter result might be due to the hyperpolarizing potentials frequently admixed with the EPSP following LC stimulation (Fung and Barnes, 1987a).
404
Antagonism of recurrent inhibitory pathway
Presynaptic facilitation
The Renshaw cell-spinal motoneuron system is known to be affected by both segmental and supraspinal inputs (for review, see Pompeiano, 1984). Early pharmacological studies reported that the aminergic system(s) tonically suppresses the recurrent inhibition of the extensor quadriceps MSR in decerebrate cats (Sinclair and Sastry, 1974; Sastry and Sinclair, 1976). In the same species, we have shown that the coeruleospinal system antagonizes the recurrently inhibited MSRs of extensor and flexor origins (Fung et al., 1987a, 1988). In these studies, the disinhibition of MSRs was found to exceed the concurrent facilitation in magnitude and was significantly correlated with that of recurrent inhibition. Direct recordings from Renshaw cells further revealed LC's disinhibitory process to involve the suppression of Renshaw cell discharges (Fig. 6). Our studies have not revealed the nature of transmitter species involved. One likely candidate is NE, due to its known inhibitory effect on Renshaw cell activity when released iontophoretically (Biscoe and Curtis, 1966; Engberg and Ryall, 1966; Weight and Salmoiraghi, 1966).
A presynaptic facilitatory role of LC on segmental transmission of low-threshold afferents has been demonstrated in decerebrate cats (Fung and Barnes, 1987~).This role is predicated upon findings of (a) LC-induced positive dorsal root potentials (DRPs; Fig. 7); (b) diminution of peripherally evoked negative DRPs after interactions with LC-induced positive DRPs; and (c) reduced excitability in large muscle and cutaneous afferent terminals following LC-conditioning volleys. Other authors have described, instead, a negative DRP upon stimulating the LC in anesthetized cats (Apkarian et al., 1984; Mokha et al., 1986; Mokha and Iggo, 1987). These latter findings are due, presumably, to the effect of anesthesia which has been documented to readily abolish the positive DRP produced by small diameter afferent fibers (Mendell and Wall, 1964; Mendell, 1970). It is possible that, in our study (Fig. 7), the antecedent LC-induced positive DRPs may partially obscure the ensuing negative D R P that resembles those reported by others. More important, the presence of a tonically active presynaptic inhibition among the flexion reflex afferents in the decerebrate but not the anesthetized cat may account for our observation of the LC-induced early positive DRPs. Since a negative DRP corre-
a
. . . . . . . . . . ... ... ..... ... ... ..... ... ... .. ... ... . . . .. ................ ........... ... .....
. .. .. ... .. .. ..... .. .. .... . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ....... .. .. .... ... ... ... ... ... ... ... ... ... ... . ._
#
+
. .... .. .. ....... . :,..;.
~.. .. . . . . . ......... ..,.....,.,..... ...... . .......... . ., . . .. ..... $
4-
Fig. 6. Inhibitory LC control on Renshaw cell firing. a,c. Raster-dot displays of 16 consecutive sweeps (from top downwards) of control and conditioned Renshaw cell discharges evoked by single ventral root volleys, respectively. Corresponding intracellular records were shown in single sweeps in b and d. Single arrows: ventral root shocks. Double arrows: LC stimuli (100 FA, 4 pulses). Calibrations: 10 msec, 5 mV. (From Fung et al., 1987a.l
5
C a t 60
Fig. 7. Averaged record ( n = 16 consecutive traces) of LCevoked dorsal root potentials (DRPs), consisting of a fast-rising early and late positive D R P succeeded by a smaller-sized, prolonged negative DRP. LC stimuli (vertical lines at beginning of trace): 250 FA, 4 pulses. (From Fung and Barnes, 1987c.)
405
lates with the primary afferent depolarization, our results may be attributable to the LC’s removal of the tonic depolarization of primary afferents in decerebrate cats. Conclusions The extracellular, antidromic activation studies have confirmed anatomically reported ipsilateral LC projections to lumbar ventral gray columns. In particular, the electrophysiological (postactivation inhibition response to noxious inputs) and a,-receptor properties are characteristic of noradrenergic coeruleospinal units across species. In addition to slow-conducting fiber components (Nakazato, 1987), studies by ourselves and others (Fung and Barnes, 1981; Pompeiano et aL, 1990) have provided evidence for a fast-conducting feline coeruleospinal system which correlates well with cell size measurements typical of tyrosine hydroxylase-immunoreactive, spinally projecting LC neurons (Reddy et al., 1989). In conjunction with the establishment of a direct coeruleospinal pathway in cats, our stimulation studies have further proven the functional connectivity between the LC and lumbar motoneuron sites. The general facilitation of flexor and extensor MSRs following LC stimuli is well substantiated by the LC’s activation of motoneuron discharges. That both changes are sensitive to a,-receptor blockade suggests that noradrenergic synapses are at least partially involved in coeruleospinal control of segmental somatomotor outflow. Additional synaptic controls are known to involve LC’s suppression (through undetermined transmitter mechanisms) of both ventral (e.g., Renshaw cell) and dorsal horn interneurons (presumably those tonically active having axoaxonic contacts on large muscle and cutaneous afferents); this produces motor excitation through functionally disinhibiting and presynaptically facilitating the motoneurons, respectively. In accord with the contention of being the principal aminergic motor system (Kuypers, 1982; Holstege and Kuypers, 19871, the noradrenergic coeruleospinal pathway has also
been postulated to influence posture as well as the response gain (sensitivity) of ipsilateral limb extensors during postural changes induced experimentally by tilting the decerebrate cat (d’Ascanio et al., 1985; Pompeiano et aZ., 1987). With its multifaceted synaptic controls over spinal reflex activity, the coeruleospinal system is strategically suited for coordinating the motor performance of animals. Acknowledgements We acknowledge the editorial assistance of M.C. Smith and the word processing assistance of R.S. Thompson and P.M. Perron. This work was supported by NIH grants NS24388 and NS07685-22 as well as by grants of the Minister0 dell’Universita, and the Agenzia Spaziale Italiana, Roma, Italy. References Aghajanian, G.K. and VanderMaelen, C.P. (1982) a,-Adrenoceptor-mediated hyperpolarization of locus coeruleus neurons: Intracellular studies in uiuo. Science, 215: 13941396. Aghajanian, G.K., VanderMaelen, C.P. and Andrade, R. (1983) Intracellular studies on the role of calcium in regulating the activity and reactivity of locus coeruleus in uiuo. Brain Rex, 273: 237-243. Akasu, T., Gallagher, J.P., Nakamura, T., Shinnick-Gallagher, P. and Yoshimura, M. (1985) Noradrenaline hyperpolarization and depolarization in cat vesical parasympathetic neurones. J. Physiol. (London), 361: 165-184. Anderson, E.G. (1972) Bulbospinal serotonin-containing neurons and motor control. Fed. Proc., 31: 107-112. Apkarian, A.V., Hodge, C.J., Jr., Stevens, R.T. and Franck, J.I. (1984) Lumbar dorsal root potentials elicited by stimulation of nucleus locus coeruleus. Exp. Neurol., 85: 202208. Araneda, S., Magoul, R. and Calas, A. (1989) [3H]-serotonin retrograde labelling in serotonergic fibers. Bruin Res. Bull., 22: 951-958. Aston-Jones, G., Foote, S.L. and Segal, M. (1985) Impulse conduction properties of noradrenergic locus coeruleus axons projecting to monkey cerebrocortex. Neuroscience, 15: 765-777. Baker, R.G. and Anderson, E.G. (1970) The antagonism of the effects of L-3, 4-dihydroxyphenylalanineon spinal reflexes by adrenergic blocking agents. J. Phurmacol. Exp. Ther., 173: 224-231.
406 Barker, J.L., Smith, T.G., Jr. and Neale, J.H. (1978) Multiple membrane action of enkephalin revealed using cultured spinal neurons. Brain Rex, 154: 153-158. Barnes, C.D. and Pompeiano, 0. (1971) The interaction of brain stem adrenergic systems and Vlllth nerve stimulation on the spinal cord. NeurophurmucoloAy, 10: 437-446. Barnes, C.D., Fung, S.J. and Gintautas, J. (1980) Brainstem noradrenergic system depression by cyclobenzaprine. Neuropharmacoloky, 19: 221-224. Barnes, C.D., Fung, S.J. and Pompeiano, 0. (1989) Descending catecholaminergic modulation of spinal cord reflexes in cat and rat. Ann. N.Y. Acud. Sci., 563: 45-58. Basbaum, A.1. and Fields, H.L. (1979) The origin of descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat: Further studies on the anatomy of pain modulation. J. Cornp. Neurol., 187: 513-523. Berry, M.S. and Pentreath, V.W. (1976) Criteria for distinguishing between monosynaptic and polysynaptic transmission. Bruin Rex, 105: 1-20. Biscoe, T.J. and Curtis, D.R. (1966) Noradrenaline and inhibition of Renshaw cells. Science, 12: 1230-1231. Bricca, G., Dontenwill, M., Molines, A,, Feldman, J., Belcourt, A. and Bousquet, P. (1989) The imidazoline preferring receptor: Binding studies in bovine, rat and human brainstem. Eur. J. Phurrnacol., 162: 1-9. Butcher, L.L., Engel, J. and Fuxe, K. (1972) Behavioral, biochemical, and histochemical analyses of the central effects of monoamine precursors after peripheral decarboxylase inhibition. Bruin Res., 41: 387-411. Carp, J.S. and Anderson, R.J. (1982) Dopamine receptormediated depression of spinal monosynaptic transmission. Bruin Res., 242: 247-254. Chan, J.Y.H., Fung, S.J., Chan, S.H.H. and Barnes, C.D. (1986) Facilitation of lumbar monosynaptic reflexes by locus coeruleus i n the rat. Bruin Res., 369: 103-109. Chu, N. and Bloom, F.E. (1974) Activity patterns of catecholamine-containing pontine neurons in the dorso-lateral tegmentum of unrestrained cats. J. Neurohiol., 5: 527-544. Commissiong, J.W., Hellstrom, S.O. and Neff, N.H. (1978) A new projection from locus coeruleus to the spinal ventral columns: Histochemical and biochemical evidence. Bruin Rex, 148: 207-213. Commissiong, J.W., Karoum, F., Reiffenstein, R.J. and Neff, N.H. (1981) Cyclobenzaprinr: A possible mechanism of action for its muscle relaxant effect. Cun. J. Physiol. Phurrnucol., 59: 37-44. d’Ascanio, P., Bettini, E. and Pompeiano, 0. (1985) Tonic inhibitory influences of locus coeruleus on the response gain of limb extensors to sinusoidal labyrinth and neck stimulations. Arch. Itul. B i d , 123: 69-100. Dahlstrom, A. and Fuxe, K. (1965) Experimentally induced changes in the intraneuronal amine levels of bulbospinal neuron systems. Actu Physrol. Scand., 64: 1-36. Dupelj, M. and Geber, J. (1981) Dopamine as a possible neurotransmitter in the spinal cord. Neuropharmucology, 20: 145-148. Eccles, J.C. (1964) The Physiology of Synapses, Academic Press, New York. 316 pp.
Edamura, M. and Aoki, M. (1989) A biphasic excitability changes in hindlimh motoneurons evoked by stimulation of the nucleus raphe magnus in the cat. Cornp. Biochem. Physiol. [A], 93: 71 1-716. Engberg, I . and Ryall, R.W. (1966) The inhibitory action of noradrenaline and other monoamines on spinal neurones. J. Physiol. (London), 185: 298-322. Esplin, B. and Capek, R. (1979) Effects of cyclobenzaprine a n spinal synaptic transmission. Neuropharrnucology, 18: 559564. Foehring, R.C., Schwindt, P.C. and Crill, W.E. (1989) Norepinephrine selectively reduces slow Cazc- and Na+-mediated K + currents in cat neocortical neurons. J. Neurophysiol., 61: 245-256. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus ceruleus: New evidence of anatomical and physiological specificity. Physiol. Re(,., 63: 844-914. Fukuda, A., Minami, T., Nabekura, J. and Oomura, Y. (1987) The effects of noradrenaline on neurones in the rat dorsal motor nucleus of the vagus, in t i t r o . J. Physiol. (Londonl. 393: 213-231. Fuller, J.H. and Schlag, J.D. (1976) Determination of antidromic excitation by the collision test: Problems of interpretation. Bruin Res., 112: 283-298. Fung, S.J. and Barnes, C.D. (1981) Evidence of facilitatory coerulospinal action in lumbar motoneurons of cats. Bruin R ~ s .216: , 299-31 1. Fung, S.J. and Barnes, C.D. (1984) Locus coeruleus control of spinal cord activity. Brainstem Control of Spinal Cord in PhpsiFunction. In C.D. Barnes (Ed.), Research TOP~CS olo~y,Academic Press, New York, pp, 215-255. Fung, S.J. and Barnes, C.D. (1986) Increased efficacy of antidromic and orthodromic activation of cat LY motoneurons upon arrival of coerulospinal volleys. Arch. Ital. B i d , 124: 229-243. Fung, S.J. and Barnes, C.D. (1987a) Membrane excitability changes in hindlimb motoneurons induced by stimulation of the locus coeruleus in cats. Bruin Res., 402: 230-242. Fung, S.J. and Barnes, C.D. (1987h) Coerulospinal enhancement of repetitive firing with correlative changes in postspike afterhyperpolarization of cat spinal motoneurons. Arch. Ital. Biol., 125: 171-186. Fung, S.J. and Barnes, C.D. (1987~)Presynaptic facilitatory action of locus coeruleus stimulation upon hindlimb sensory impulse transmission in decerebrate cats. Arch. Ital. Biol., 125: 187-200. Fung, S.J. and Barnes, C.D. (1989) RaphC-produced excitation of spinal cord motoneurons in the cat. Newosci. Left., 103: 185-190. Fung, S.J., Pompeiano, 0. and Barnes, C.D. (1987a) Suppression of the recurrent inhibitory pathway in lumbar cord segments during locus coeruleus stimulation in cats. Bruin R ~ s . 402: , 351-354. Fung, S.J., Reddy, V.K., Bowker, R.M. and Barnes, C.D. (1987b) Differential labeling of the vestibular complex following unilateral injections of horseradish peroxidase into the cat and rat locus coeruleus. Bruin RKS., 401: 347-352.
407
Fung, S.J., Pompeiano, 0. and Barnes, C.D. (1988) Coerulospinal influence on recurrent inhibition of spinal motonuclei innervating antagonistic hindleg muscles of the cat. Pfliigers. Arch., 412 346-353. Geber, J. and Dupelj, M. (1977) The effect of [.-Dopa on the spinal monosynaptic mass reflex. Experientia, 33: 10741075. German, D.C. and Fetz, E.E. (1976) Responses of primate locus coeruleus and subcoeruleus neurons to stimulation at reinforcing brain sites and to natural reinforcers. Brain Res., 109: 497-514. Grant, S.J. and Redmond, D.E., Jr. (1984) Neuronal activity of the locus ceruleus in awake Macaca arctoides. Exp. Neurol.. 84: 701-708. Grant, S.J., Aston-Jones, G . and Redmond, D.E., Jr. (1988) Responses of primate locus coeruleus neurons to simple and complex sensory stimuli. Brain Rex Bull., 21: 401-410. Grillner, S., Hongo, T. and Lund, S. (1970) The vestibulospinal tract. Effects on a-motoneurones in the lumbosacral spinal cord in the cat. Exp. Brain Res., 10: 94-120. Grillner, S., Hongo, T. and Lund, S. (1971) Convergent effects on a-motoneurones from the vestibulospinal tract and a pathway descending in the medial longitudinal fasciculus. Exp. Brain Res., 12: 457-479. Guyenet, P.G. (1980) The coeruleospinal noradrenergic neurons: Anatomical and electrophysiological studies in the rat. Bruin Res., 189: 121-133. Hancock, M.B. and Fougerousse, C.L. (1976) Spinal projections from the nucleus locus coeruleus and nucleus subcoeruleus in the cat and monkey as demonstrated by the retrograde transport of horseradish peroxidase. Brain Res. Bull., 1; 229-234. Hayes, N.L. and Rustioni, A. (1981) Descending projections from brainstem and sensorimotor cortex to spinal enlargements in the cat. Exp. Brain Res., 41: 89-107. Hobson, J.A., McCarley, R.W. and Wyzinski, P.W. (1975) Sleep cycle oscillation: Reciprocal discharge by two brainstem neuronal groups. Science, 189: 55-58. Holstege, J.C. and Kuypers, H.G.J.M. (1987) Brainstem projections to spinal motoneurons: An update. Neuroscience, 23: 809-821. Holstege, G., Kuypers, H.G.J.M. and Boer, R.C. (1979) Anatomical evidence for direct brain stem projection to the somatic motoneuronal cell groups and autonomic preganglionic cell groups in cat spinal cord. Brain Res., 171: 329-333. Holtman, J.R., Dick, T.E. and Berger, A.J. (1986) Involvement of serotonin in the excitation of phrenic motoneurons evoked by stimulation of the raphe obscurus. J. Neurosci., 6: 1185-1193. Hongo, T., Jankowska, E. and Lundberg, A. (lY6Y) The rubrospinal tract. I. Effects on a-motoneurones innervating hindlimb muscles in cats. Exp. Brain Res., 7: 344-364. Ingram, W.R., Ranson, S.W., Hannett, F.I., Zeiss, F.R. and Terwilliger, E.H. (1932) Results of stimulation of the tegmentum with the Horsley-Clarke stereotaxic apparatus. Arch. Neurol. Psychiat., 28: 513-541. Jordan, L.M., McCrea, D.A., Steeves, J.D. and Menzies, J.E.
(1977) Noradrenergic synapses and effects of noradrenaline on interneurons in the ventral horn of the cat spinal cord. Can. J. Physiol., 55: 399-412. Kausz, M. (1986) Distribution of neurons in the lateral pontine tegmentum projecting to thoracic, lumbar and sacral spinal segments in the cat. J. Hirnfor~sch.,27: 485-493. Kuypers, H.G.J.M. (1982) A new look at the organization of the motor system. In Kuypers, H.G.J.M. and Martin, G.F. (Eds.), Progress in Brain Research, Vol. 57, Anatomy of Descending Pathways to the Spinal Cord, Elsevier Biomedical Press, Amsterdam/New York/Oxford, pp. 381-403. Kuypers, H.G.J.M. and Maisky, V.A. (1975) Retrograde axonal transport of horseradish peroxidase from spinal cord to brain stem cell groups in the cat. Neurosci. Lett., 1: 9-14, Kuypers, H.G.J.M. and Maisky, V.A. (1977) Funicular trajectories of descending brain stem pathways in cat. Brain Res., 136: 159-165. Lacaille, J.-C. and Schwartzkroin, P.A. (1988) Intracellular responses of rat hippocampal granule cells in d r o to discrete applications of norepinephrine. Neurosci. Lett., 89: 176-181. Lai, Y.-Y. and Barnes, C.D. (1985) A spinal projection of serotonergic neurons of the locus coeruleus in the cat. Neurosci. Lett., 58: 159-164. Lai, Y.-Y., Strahlendorf, H.K., Fung, S.J. and Barnes, C.D. (1989) The actions of two monoamines on spinal motoneurons from stimulation of the locus coeruleus in the cat. Brain Res., 484: 268-272. Lalley, P.M. (1986) Responses of phrenic motoneurones of the cat to stimulation of medullary raphe nuclei. J. Physiol. (London), 380: 349-371. Landgren, S., Phillips, C.G. and Porter, R. (1962) Minimal synaptic actions of pyramidal impulses on some a motoneurones of the baboon's hand and forearm. J. Physiol. (London), 161: 91-111. Legendre, P., Dupouy, B. and Vincent, J.D. (1988) Excitatory effect of noradrenaline on pacemaker cells in spinal cord primary cultures. Neuroscience, 24: 647-658. Lipski, J. (1981) Antidromic activation of neurones as an analytic tool in the study of the central nervous system. J. Neurosci. Meth., 4: 1-32. Maeda, M., Maunz, R.A. and Wilson, V.J. (1975) Labyrinthine influence on cat forelimb motoneurons. Exp. Bruin Res., 22: 69-86. Marshall, K.C. (1983) Catecholamines and their actions in the spinal cord. In Davidoff, R.A. (Ed.), Handbook of the Spinal Cord, Vol. 1, Pharmacology, Marcel Dekker, Inc., New York, pp. 275-328. McBride, R.L. and Sutin, J. (1976) Projections of the locus coeruleus and adjacent pontine tegmentum in the cat. J. Comp. Neurol., 165: 265-284. McCormick, D.A. and Prince, D.A. (1988) Noradrenergic modulation of firing pattern in guinea pig and cat thalamic neurons, in uitro. J. Neurophysiol., 59: Y78-996. Mendell, L.M. (1970) Positive dorsal root potentials produced by stimulation of small diameter muscle afferents. Brain Rex, 18: 375-379.
408 Mendell, L.M. and Wall, P.D. (1964) Presynaptic hyperpolarization: A role for fine afferent fibres. J. Physiol. (London), 1 7 2 274-294. Mizukawa, K. .(1980) The segmental detailed topographical distribution of monoaminergic terminals and their pathways in the spinal cord of the cat. Anar. Anz., 147: 125- 144. Mizukawa, K. and Takeuchi, Y. (1982) Electron microscopic investigation of monoaminergic terminals to a-motoneurones in the anterior horn of the cat spinal cord. Acfa Med. Okayama, 36: 85-93. Mokha, S.S. and Iggo, A. (1987) Mechanisms mediating the brain stem control of somatosensory transmission in the dorsal horn of the cat's spinal cord: An intracellular analysis. Exp. Brain Res., 69: 93-106. Mokha, S.S., McMillan, J.A. and Iggo, A. (1986) Pathways mediating descending control of spinal nociceptive transmission from the nuclei locus coeruleus (LC) and raphe magnus (NRM) in the cat. Exp. Bruin Res., 61: 597-606. Muir, R.B. and Porter, R. (1973) The effect of a preceding stimulus on temporal facilitation at corticornotoneuronal synapses. J. Physiol. (London), 228: 749-763. Nakazato, T. (1987) Locus coeruleus neurons projecting to the forebrain and the spinal cord in the cat. Neurorcience, 23: 529-538. North, R.A. and Yoshimura, M. (1984) The actions of noradrenaline on neurones of the rat substance gelatinosa in i)itro. J. Physiol., 349: 43-55. Papez, J.W. (1925) A locospinal tract. Anat. Rec., 29: 392. Pappas, B.A., Breese, G.R., Mailman, R.B. and Mueller, R.A. (1980) Importance of the locus coeruleus and involvement of a-adrenergic receptors in the post-decapitation reflex in the rat. Psychopharmacology, 69: 163-171. Pompeiano, 0. (1984) Recurrent inhibition. In Davidoff, R.A. (Ed.), Handbook of the Spinal Cord, Vols. 2 and 3, Marcel Dekker, New York, pp. 461-557. Pompeiano, O., D'Ascanio, P., Horn, E. and Stampacchia, G. (1987) Effects of local injection of the a,-adrenergic agonist clonidine into the locus coeruleus complex on the gain of vestibulospinal and cervicospinal reflexes in decerebrate cats. Arch. Ifal. Biol., 125: 225-269. Pompeiano, O., Manzoni, D., Barnes, C.D., Stampacchia, G. and D'Ascanio, P. (1990) Responses of locus coeruleus and subcoeruleus neurons to sinusoidal stimulation of labyrinth receptors. Neuroscience, 35: 227-248. Porter, R. and Hore, J. (1969) Time course of minimal corticomotoneuronal excitatory postsynaptic potentials in lumbar motoneurons of the monkey. J. Neurophysiol., 32: 443-45 1. Pun, R.Y.K., Marshall, K.C., Hendelman, W.J., Guthrie, P.B. and Nelson, P.G. (1985) Noradrenergic responses of spinal neurons in locus coeruleus-spinal cord co-cultures. J. Neurosci., 5: 181-191. Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1989) Spinally projecting noradrenergic neurons of the dorsolatera1 pontine tegmentum: A combined immunocytochemical and retrograde labeling study. Brain Res., 491: 144-149. Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1990)
Localization of enkephalinergic neurons in the dorsolatera1 pontine tegmentum projecting to the spinal cord of the cat. J. Comp. Neurol., 291: 195-202. Rhines, R. and Magoun, H.W. (1946) Brain stem facilitation of cortical motor response. J. Neurophysiol., 9: 219-229. Sakai, K. (1980) Some anatomical and physiological properties of ponto-mesencephalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep in the cat. In Hobson, J.A. and Brazier, M.A.B. (Eds.), The Reticular Formation ReL'isited. Raven Press, New York, pp. 427-447. Sastry, B.S.R. and Sinclair, J.G. (1976) Tonic inhibitoiy influence of a supraspinal monoaminergic system on recurrent inhibition of an extensor monosynaptic reflex. Brain Rex, 117: 69-76. Shapovalov, A.I., Kardmjan, O.A., Kurchavyi, G.G. and Repina, Z.A. (1971) Synaptic actions evoked from the red nucleus on the spinal a-motoneurons in the rhesus monkey. Brain Res., 32: 325-348. Simmons, R.M.A. and Jones, D.J. (1988) Binding of ['Hlprazosin and [3H]p-aminoclonidine to a-adrenoceptors in rat spinal cord. Brain Res., 445: 338-349. Sinclair, J.G. and Sastry, B.S.R. (1974) Supraspinal monoaminergic involvement in the blockade of recurrent inhibition of the monosynaptic reflex. Neuropharmacology, 13: 741-747. Stevens, R.T., Hodge, C.J., Jr. and Apkarian, A.V. (1982) Kolliker-Fuse nucleus: The principal source of pontine catecholaminergic cells projecting to the lumbar spinal cord of cat. Brain Res., 2 3 9 589-594. Strahlendorf, J.C., Strahlendorf, H.K., Kingsley, R.E. and Barnes, C.D. (1980) Facilitation of the lumbar monosynaptic reflexes by locus coeruleus stimulation. NeuropharmaC O / O R ~ ,19: 225-230. Suenaga, N., Oishi, R. and Fukuda, T. (1979) The role of locus coeruleus in decapitation convulsions of rats. Brain Res., 177: 83-93. Szabadi, E. (1979) Adrenoceptors on central neurones: Microelectrophoretic studies. Neuropharmacology, 18: 831-843. Todorovic, S. and Anderson, E.G. (1990) 5-HT2 and 5-HT3 receptors mediate two distinct depolarizing responses in rat dorsal root ganglion neurons. Brain Rex, 511: 71-79. Watabe, K., Nakai, K. and Kasamatsu, T. (1982) Visual afferents to norepinephrine-containing neurons in cat locus coeruleus. Exp. Bruin Res., 48: 66-80. Weight, F.F. and Salmoiraghi, G.C. (1966) Adrenergic responses of Renshaw cells. J, Pharmacol. Exp. Ther., 154: 391-397. White, S.R. and Fung, S.J. (1989) Serotonin depolarizes cat spinal cord motoneurons in situ and decreases motoneuron afterhyperpolarizing potentials. Brain Res., 502: 205213. Wiklund, L., LCger, L. and Persson, M. (1981) Monoamine cell distribution in the cat brain stem. A fluorescence histochemical study with quantification of indolaminergic and locus coeruleus cell groups. J. Comp. Neurol., 203: 613-647. Wilson, V.J. and Yoshida, M. (1969) Comparison of effects of
40Y
stimulation of Deiters’ nucleus and medial longitudinal fasciculus on neck, forelimb, and hindlimb motoneurons. J. Neurophysiol., 32: 743-758. Wolstencroft, J.H. (1964) Reticulospinal neurones. J. Physiol. (London), 174: 91-108. Yamashita, H., Inenaga, K. and Kannan, H. (1987) Depolarizing effect of noradrenaline on neurons of the rat supraoptic nucleus in citro. Brain Res.. 405: 348-353.
Young, W.S.111 and Kuhar, M.J. (1980) Noradrenergic a , and a,-receptors: Light microscopic autoradiographic localization. Proc. Natl. Acad. Sci. USA, 77: 1696-1700. Zhan, W.-Z., Ellenberger, H.H. and Feldman, J.L. (1989) Monoaminergic and gabaergic terminations in phrenic nucleus of rat identified by immunohistochemical labeling. Neuroscience, 31: 105-1 13.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pornpeiano (Ed,.) f'rogreci in Brain Research, Vol. 88 b 1991 Elscvier Science Publishers B.V.
411 CHAPTER 31
Responses of locus coeruleus neurons to labyrinth and neck stimulation 0. Pompeiano
', D. Manzoni
'
and C.D. Barnes
'
Department of Physiology and Biochemistry, Unicersity of Pisa, I%:aS. Zeno, Pisa, Italy and Department of Veterinary and Comparatiiv Anatomy, Pharmacolom and Physiology, College of Veterinary Medicine, Washington State Unicersity, Pullman, WA, U.S.A.
The electrical activity of a large population of locus coeruleus (LC)-complex neurons, some of which were antidromically activated by stimulation of the spinal cord at T12-L1, was recorded in precollicular decerebrate cats during labyrinth and neck stimulation. Some of these neurons showed physiological characteristics attributed to norepinephrine (NEI-containing LC neurons, i.e., (i) a slow and regular resting discharge; (ii) a typical biphasic response to compression of the paws consisting of short impulse bursts followed by a silent period, which was attributed to recurrent and/or lateral inhibition of the corresponding neurons; and (iii) a suppression of the resting discharge during episodes of postural atonia, associated with rapid eye movements (REM), induced by systemic injection of an anticholinesterase, a finding which closely resembled that occurring in intact animals during desynchronized sleep. Among the neurons tested, 80 of 141 (i.e., 56.7%) responded to the labyrinth input elicited by sinusoidal tilt about the longitudinal axis of the whole animal at the standard parameters of 0.15 Hz, 5 lo", and 73 of 99 (i.e., 73.7%) responded to the neck input elicited by rotation of the body about the longitudinal axis at the same parameters, while maintaining the head stationary. A periodic modulation of firing rate of the units was observed during the sinusoidal stimuli. In particular, most of the LC-complex units were maximally excited during side-up tilt of the animal and sidedown neck rotation, the response peak occurring with an average phase lead of about + 17.9" and +34.2" with respect to the extreme animal and neck displacements, respectively. Similar results were also obtained from the antidromically identified coeruleospinal (CS) neurons. The degree of conver-
gence and the modalities of interaction of vestibular and neck inputs on LC-complex neurons were also investigated. In addition to the results described above, the LC-complex neurons were also tested to changing parameters of stimulation. In particular, both static and dynamic components of single unit responses were elicited by increasing frequencies of animal tilt and neck rotation. Moreover, the relative stability of the phase angle of the responses evaluated with respect to the animal position in most of the units tested at increasing frequencies of tilt allowed the conclusion to attribute these responses to the properties of macular ultricular receptors. This conclusion is supported by the results of experiments showing that LC-complex neurons displayed steady changes in their discharge rate during static tilt of the animal. These findings are discussed in relation to the facts that: (i) the predominant patterns of response of the LC-complex neurons as well as the CS neurons to animal tilt and neck rotation are opposite to those of the vestibulospinal (VS) neurons projecting to the same segments of the spinal cord, and (ii) the former neurons exert an inhibitory influence on Renshaw cells, while the latter exert an excitatory influence on them, due to activation of ipsilateral limb extensor motoneurons and their recurrent collaterals. We postulated that periodic changes in firing rate of the antidromically identified CS neurons during animal tilt and neck rotation might change the functional c o u p h g of the ipsilateral limb extensor motoneurons with their own Renshaw cells, thus intervening in the gain regulation of the corresponding VS and cervicospinal reflexes. As to the unidentified LC-complex neurons responsive to the same stimuli, they might be involved in the labyrinthine and cervical control of functions other than posture.
Key words: neck-vestibular inputs, locus coeruleus, subcoeruleus, unit responses, postural reflexes
412
Introduction
The locus coeruleus (LC) and the locus subcoeruleus (SC) represent a rather small nuclear complex composed mainly of norepinephrine (NE)-containing neurons (Dahlstrom and Fuxe, 1964), located in the dorsolateral pontine tegmentum. These neurons, identified in different animal species including the cat (cf. Jones and Friedman, 19831, fire regularly at a low rate in the animal at rest and show a pronounced inhibition following antidromic or synaptic activation. These findings are attributed, in part at least, to self-inhibitory synapses which act on a,-adrenoceptors by utilizing mechanisms of recurrent and/or lateral inhibition (cf. Foote et al., 1983; Ennis and Aston-Jones, 1986a). An additional possibility, however, is that this post-stimulation inhibition arises directly from depolarization, via a calcium-activated increase in potassium conductance (cf. Aghajanian et al., 1983; Williams et al., 1984). Within the LC complex there are subpopulations of neurons, which give rise to different ascending and descending projections (cf. Foote et al., 1983; Loughlin et at., 1986a,b), thus implying a diversity of the postulated role for this structure. Several lines of evidence indicate that the LC complex exerts a facilitatory role on posture: (1) Anatomical or functional inactivation of the LC complex elicited either by electrolytic lesion of this structure (d’Ascanio et al., 1985) or by local injection into the LC complex of the a,-adrenergic agonist clonidine (Pompeiano et al., 19871, which suppresses the discharge of the NE-containing neurons (Svensson et al., 1975; Guyenet, 1980; cf. Foote et al., 19831, greatly decreased the postural activity particularly in the ipsilateral limbs of decerebrate cats. On the other hand, just the opposite result was obtained after injection into the LC complex of the cholinergic agonist carbachol (Stampacchia et al., 1987a), which acts on excitatory cholinergic synapses located on the LC neurons (cf. Foote et al., 19831, where muscarinic receptors are present (Egan and North,
1985). (2) Stimulation experiments have shown that the coeruleo- and the subcoeruleospinal (CS) pathway, which sends axons to the ventral horn region of the spinal cord (cf. Westlund and Coulter, 1980; Kuypers and Huisman, 1982; Westlund et al., 1982, 1983, 1984; Holstege and Kuypers, 1989), exerts an excitatory influence on ipsilateral limb extensor (and flexor) motoneurons (Fung and Barnes, 1981, 1987; Chan et al., 19861, an effect which was due, at least in part, to suppression of the tonic discharge of inhibitory Renshaw cells acting on spinal motoneurons (Fung et al., 1987, 1988). (3) Finally, experiments of unit recording have shown that in intact (cf. Sakai, 1980; Foote et al., 1983; Hosbon and Steriade, 19861, as well as in decerebrate cats (Pompeiano and Hoshino, l975,1976a,b; cf. Pompeiano, 19801, the postural activity is present as long as the LC-complex neurons fire regularly; however, as soon as their discharge decreased either during desynchronized or rapid eye movement (REM) sleep in intact animals, or during the REM episodes induced by systemic injection of an anticholinesterase in decerebrate cats, the postural activity also decreased and disappeared. In addition to static changes in posture, the LC complex may also contribute to the dynamic control of posture during the vestibulospinal (VS) and cervicospinal reflexes. It is known that in decerebrate cats, displacement of the head after neck deafferentation (Lindsay et at., 1976) or rotation about the longitudinal axis of the whole animal (Schor and Miller, 1981; Manzoni et al., 1983a; Ezure and Wilson, 19841, leading to stimulation of vestibular receptors, as well as displacement of the neck with a stationary head, leading to stimulation of neck receptors (Magnus, 1924; McCouch et al., 1951; Lindsay et al., 1976; Manzoni et al., 1983a; Ezure and Wilson, 1983, 1984) produced asymmetric responses in limb extensors. In particular, side-down rotation of the head or of the animal produced contraction, whereas side-up rotation resulted in relaxation of ipsilatera1 forelimb extensors; on the other hand, patterns of response of opposite sign occurred for
413
the same directions of neck rotation. These findings explained why directional changes in head position in the intact preparation, i.e., when both tonic labyrinth and neck reflexes act in opposition, leave the position of the limbs unmodified (cf. von Holst and Mittelstaedt, 1950; Lindsay et al., 1976; Manzoni et al., 1983a; Ezure and Wilson, 1984). The reciprocal changes in postural activity of the forelimbs, produced by animal tilt and neck rotation have been attributed to an integration of opposite influences arising from vestibular and neck receptors, converging through independent channels on the lateral vestibular nucleus (LVN) of Deiters. There is in fact evidence that the LVN neurons, which exert a monosynaptic and/or a polysynaptic excitatory influence on ipsilateral limb extensor motoneurons (Lund and Pompeiano, 1968; cf. Pompeiano, 19751, receive convergent inputs from both types of receptors (Boyle and Pompeiano, 1981; Stampacchia et al., 1987b). In particular, most of the VS neurons, including those which were antidromically identified as projecting to the lumbosacral segments of the spinal cord, were excited during side-down tilt of the animal and side-up neck rotation (cf. also Boyle and Pompeiano, 1980a,b; Schor and Miller, 1982; Marchand et al., 1987). Moreover, when both types of receptors were costimulated during head rotation, the response characteristics of the VS neurons closely corresponded to those predicted by a vectorial summation of the individual responses (Boyle and Pompeiano, 1981). In .order to understand the influences that the vestibular and the neck inputs exert on the LC complex, a first group of experiments were performed in precollicular decerebrate cats, to find out: (1) whether the LC-complex neurons, including the CS neurons, having the characteristics attributed to noradrenergic neurons, responded to sinusoidal stimulation of vestibular and neck receptors; (2) whether their predominant response patterns were in-phase or out-of-phase with respect to those of the VS neurons; (3) whether the degree of convergence and the
modality of interaction of both labyrinth and neck inputs impinging on the LC-complex neurons were comparable to those reported for the VS neurons and, finally, (4) whether the responses of the LC-complex neurons to labyrinth and neck stimulation contributed to the postural adjustments of limb extensors during the VS and the cervicospinal reflexes (Barnes et al., 1989; Manzoni et al., 1989; Pompeiano et al., 1990). These observations will be integrated by the results of a second group of experiments, performed in decerebrate cats (Pompeiano and Hoshino, 1975, 1976a,b), aimed to find out whether the LC-complex neurons, which were originally tested to labyrinth stimulation, showed a suppression of their firing rate during the episodes of postural atonia induced by an anticholinesterase (cf. Pompeiano, 19801, similar to that occurring in intact animals during desynchronized sleep (cf. Sakai, 1980; Foote et al., 1983; Hobson and Steriade, 1986). It would then be possible to prove that the LC-complex units, responsive to the labyrinth input, display one of the main properties usually attributed to the NE-containing LC neurons. Methods The first group of experiments was performed in precollicular decerebrate cats operated under ether anesthesia (Barnes et al., 1989; Manzoni et al., 1989; Pompeiano et al., 1990). The head of the animal was fixed through the stereotaxic frame in a horizontal position and pitched lo" nosedown, while the spinous process of the second cervical vertebra was held by a clamp secured on a tilting table; the lower part of the trunk was rigidly fixed to a spinal cord frame and both foreand hindlimbs were extended and clamped. The system was driven sinusoidally; in particular, rotation about the longitudinal axis of the whole animal produced selective stimulation of vestibular receptors, while rotation of the neck clamp and table simultaneously in both directions of the coronal plane beneath a stationary head pro-
414
duced selective stimulation of neck receptors located within the atlanto-occipital and atlanto-axial joints and/or the perivertebral dorsal muscles. Moreover, rotation of the head while the vertebral clamp remained fixed on the table in a horizontal position elicited costimulation of both vestibular and neck receptors. Sinusoidal stimuli at the standard parameters of 0.15 Hz, +lo" peak amplitude of displacement were used. In some experiments, however, the peak amplitude varied from 2.5" to 20" at 0.15 Hz and the frequency from 0.008 to 0.32 Hz, at 5 10". Single-unit activity was recorded extracellularly from the dorsolateral pontine tegmentum at the coordinates of P 1.5 to 4.0, L or R 2.5 to 3.0, H - 1.5 to 2.5, by using glass micropipettes (5-10 MR impedance) filled with 0.5 M sodium chloride saturated with pontamine sky blue. Two stimulating electrodes, made of thin tungsten wires completely insulated except at the tip were implanted into the ventral quadrant of the spinal cord of each side, and used for the identification of CS neurons, following the collision test. Each unit was also tested during pinching of the paws with toothed forceps. After application of the antidromic and orthodromic tests, the unit activity was converted to standard pulses and analyzed by a digital signal averager (Correlatron 1024, Laben). Sequential pulse density histograms were obtained by averaging 18 sweeps, each containing the responses to two successive cycles of rotation (128 bins, 0.1 sec bin width, for the standard test of 0.15 Hz, 100). The analog output of the signal averager was plotted on an X-Y plotter and the digital data were processed on-line with a computer system which performed a Fourier analysis (cf. Boyle and Pompeiano, 1981). A spectral analysis of the angular input (table or head rotation) and of the output (unit activity) was performed; it was then possible to evaluate the gain (absolute change of the mean discharge rate per degree of displacement, in imp./sec/deg), the sensitivity (percentage change of the mean discharge rate per degree, in %/deg) and the phase angle of the
+
first harmonic of the response, expressed in degrees with respect of the peak of stimulation of vestibular receptors (i.e., downward displacement of the whole animal toward the recording side) or neck receptors (i.e., upward displacement of the table side ipsilateral to the recorded unit). The mean discharge rate or base frequency (imp./sec), evaluated during each test, usually corresponded to the mean firing rate of the same unit recorded at rest. All the experiments were performed in animals immobilized with pancuronium bromide (Pavulon, Organon, The Netherlands, 0.6 mg/ kg/ h, iv) and artificially ventilated. At the end of each penetration, the location of the recorded units was marked by passing cathodal current through the microelectrode tip (Hellon, 1971). The position of the dye mark was then identified on histological serial sections of the brainstem counterstained with neutral red. The dorsolateral pontine structures were outlined on a series of drawings, following the anatomical criteria described by Sakai (1980). The second group of experiments was also performed in precollicular decerebrate cats (Pompeiano and Hoshino, 1975,1976a,b). In these instances the single-unit activity of LC-complex neurons was recorded with tungsten microelectrodes (3-6 M R impedance) and the root mean square activity of the corresponding units was displayed on a Beckman polygraph together with the electromyographic (EMG) activity of the triceps brachii and the electro-oculogram (EOG). The unit activity as well as the EMG and the EOG were also stored on different channels of a tape recorder (Philips), to be analyzed for statistical evaluation of the results. Episodes of postural atonia were elicited by intravenous injection of 0.1 mg/kg of eserine sulfate, an anticholinesterase which passes through the blood-brain barrier. The same units were also submitted to 15" displacement about the longitudinal axis of the animal and their responses to static tilt were tested during postural rigidity as well as during the episodes of postural atonia induced by the
415
anticholinesterase. Small electrolytic lesions were made at the end point of the most representative tracks to identify the position of specific sites of recording. The lesions were produced by passing anodal current through the tip of the recording electrode (0.5 mA for 10-20 sec). Histological controls were then made at the end of the experiments to identify the position of the recorded
units following the criteria described by Pompeiano and Hoshino (1976a,b). Results
General properties of the recorded LC-complex neurons In the main group of experiments we recorded the activity of 141 LC-complex neurons: in partic-
P 2.0
Fig. 1. Anatomical localization of locus coeruleus (LC)-complex neurons tested to sinusoidal tilt of the whole animal. Among the 141 units recorded and histologically identified, 80 units (25/41 LCd, 31/67 L C a and 24/33 SC neurons) responded to sinusoidal tilt about the longitudinal axis of the animal at 0.15 Hz, ilo"; 20 units increased their firing rate during side-down tilt with a phase angle of the first harmonic of the response ranging from + 75 to -45" (a), 45 units increased their firing rate during side-up tilt with a phase angle ranging from - 105 to 135" ( A ) and 15 units showed a phase angle that varied from + 75 to 135" and from -45 to - 105" (e), The remaining 61 units were unaffected by the stimulus (X). These units have been plotted on 6 representative transverse sections of the pons taken at the stereotaxic planes of P1.5 to P4.0 and labeled progressively from rostra1 to caudal levels. (From Pompeiano et al., 1990.)
+
+
416
ular, 41 units were histologically located in the dorsal part of the LC (LCd), 67 in the ventral part (LCa) and 33 in the SC (Fig. 1). This explored area actually corresponded to the region in which spinally projecting neurons are located, as shown by the retrograde cell marker horseradish peroxidase (cf. Reddy et al., 1989). Demonstrating that all the recorded units are confined within the LC and SC area does not necessarily prove that they belong to NE-containing neurons. In fact, while these noradrenergic neurons form in the rat a compact cluster of cells, they appear in the cat to be sparse and intermingled with non-noradrenergic ones (cf. Foote et al., 1983). For this reason particular care was made in our experiments to find out whether the recorded neurons had the characteristics usually attributed to NE-containing neurons previously recorded in the rat (cf. Foote et al., 1983; Ennis and Aston-Jones, 1986a). In fact: (1) most of the recorded neurons had a typical positive-negative extracellular spike of long duration ( 2 1.5 msec); (2) these units had a regular resting discharge. The mean firing rate of the LC-complex units (9.6 & 13.3, S.D., imp./sec; n = 141) was higher than that reported in the same animal species by previous authors (cf. Sakai, 1980; Foote et al., 1983; Hobson and Steriade, 1986), a finding which may depend upon different preparations used (decerebrate vs. intact animals). However, 68 out of 141 recorded neurons (i.e., 48.2%) fired at a frequency lower than 5 imp./sec (mean firing rate 2.5 f 1.4, S.D., imp./sec); (3) among the 134 LC-complex units tested, 113 (84.3%) were influenced by pinch stimuli applied to the paws; moreover, the majority of these units (89 units, i.e., 66.4%) showed a burst of excitation followed by a period of quiescence, which occurred in spite of the persistence of the noxious stimulus, as reported in rats (Fig. 2A); on the other hand, 24 units (17.9%) showed an inhibition followed by a rebound effect characterized by bursts of excitation followed by delayed inhibitions (Fig. 2B). In addition to these characteristics, 16 out of the recorded LC-complex units were antidromically
I
B
5 sec
lo,
5 sec
Fig. 2. Characteristic patterns of response of LC-complex units to a nociceptive input evoked by pinch stimuli applied to the foot of the ipsilateral forelimb. Individual units located in the SC which were either responsive (A) or unresponsive (B) to animal tilt. PSTHs averaged over three sweeps (124 bins with 0.4s bin width) showing in (A) a burst of excitation followed by inhibition during pinch stimuli applied to the ipsilateral forepaw, while in B there was only a pause of inhibition during the pinch stimulus followed by rebound effects; these were characterized by bursts of excitation (asterisks) separated by a delayed depression. The underlines represent duration of the applied stimulus. (From Pompeiano ct al., 1990.)
activated from the ipsilateral spinal cord at T12L1. Although the cell bodies of these neurons were located in the ventral part of both the LCd ( n = 2) and the LC ( n = 4) as well as in the SC ( n = lo), the conduction velocity of their axons (48.7 k 22.4, S.D., m/sec) was on the average higher than that reported for noradrenergic CS neurons in the rat (Guyenet, 1980).
Response characteristics of LC-complex neurons to sinusoidal tilt of the animal Pompeiano et al. (1990) tested the whole population of 141 LC-complex neurons to roll tilt of the animal at the parameters of 0.15 Hz, *lo". Among these recorded units, 80 (56.7%) showed a periodic modulation of their firing rate in response to the sinusoidal input; in these instances, the average gain and sensitivity of the first harmonic of responses corresponded to 0.19 k 0.26,
41 7
S.D., imp./sec/deg and 2.7 f 2.7, %/deg, respectively. However, the proportion of responsive units as well as the gain (and sensitivity) values of their responses were on the average higher in the SC (24/33, i.e., 72.7% and 0.32 i 0.36, S.D., imp./sec/deg) and the LCd (25/41, i.e., 61.0% and 0.19 k 0.23, S.D., imp./sec/deg) than in the LCa (31/67, i.e., 46.3% and 0.09 i 0.13, S.D., imp./sec/deg). These findings indicate that the labyrinth input was not homogeneously distributed within the LC complex. If we consider the phase angle of the responses, two main groups of units were observed (Fig. 3A). The first group of units (45/80, i.e., 56.25%) were excited during side-up tilt of the animal and showed an average phase lag of - 162.3 k 27.4, S.D., deg, corresponding to an average lead of +17.7" with respect to the extreme side-up position (@-responses). An example of these unit responses is shown in Figure 4A (upper trace). On the other hand, the second group of units (20/80, it., 25.0%) were excited during side-down tilt of the animal and showed an average phase lead of + 18.5 k 32.7, S.D., deg (a-responses). In addition to these two populations of units, there were 15 units (i.e., 18.75%) whose phase angle of responses was intermediate between the values reported above. It is of interest that among the positional-sensitive units, which were located in the SC and the LCd, those excited by side-up tilt of the animal showed more than a twofold larger gain and sensitivity ( n = 45:0.22 i 0.24, S.D., imp./ sec/ deg and 3.6 i 3.2, S.D., %/deg) with respect to the units excited by side-down tilt ( n = 20:0.09 f 0.11, S.D., imp./sec/deg and 1.5 kO.9, S.D., %/deg) (t-test between the means, P < 0.05 and P < 0.01 for differences in gain and sensitivity values of the responses, respectively). Similar response properties were also found in 11 out of the 16 antidromically identified CS neurons which responded to animal tilt. In fact, the majority of these responsive neurons were located in the SC and the LCd ( n = 10 units) and showed large amplitude @-responses ( n = 8 units).
A L C complex
NUMBER OF UNITS 15
1
J
LABYRINTH INPUT N=80
10
U
NECK INPUT N-73
LAG
PHASE (DEGI
LtAD
Fig. 3. Distribution of the phase angle of the first harmonic of responses of LC-complex neurons tested during roll tilt of the animal (A) and neck rotation (B) at 0.15 Hz, +lo". Positive numbers in the abscissas indicate, in degrees, the phase lead, whereas negative numbers indicate the phase lag of responses with respect to the extreme side-down position either of the animal or of the neck, as indicated by 0". Responses of LC-complex neurons to tilt o r neck rotation, underlined by horizontal bars, have been used to evaluate the average phase angle of units excited during or near the side-down (0') or side-up (180") animal or neck displacement. Most of the units were excited during side-up tilt of the animal (45/80 neurons) and side-down neck rotation (40/73 neurons). (From Pompeiano et al., 1990 in A and Barnes et al., 1989 in B.)
Since the low resting discharge represents one of the main characteristics of the NE-containing LC neurons, we selected among the whole population of recorded neurons 68 units (21 located in the LCd, 37 in L C a and 10 in SC) which showed a base frequency lower than 5 imp./sec; 6 out of these units were antidromically identified as CS neurons (mean conduction velocity, 35.3 k 19.5, S.D., m/sec). Consonant with the results obtained from the whole population of tested units, a large proportion of these slow discharging LC-complex units
418
(38/68, ie., 56.0%) responded to animal tilt at 0.15 Hz, k 10" and showed an average gain and sensitivity of 0.11 f 0.12, S.D., imp./sec/deg and 3.6 k 2.6, S.D., %/deg, respectively. The majority of these responsive units (23/38, i.e., 60.5%) were excited during side-up tilt of the animal, while a smaller group of units (10/38, i.e., 26.3%) were excited during side-down tilt; five units (i.e., 13.2%) showed intermediate responses. Moreover, among the units which were excited by the extreme animal displacement, those activated by side-up tilt of the animal showed more than a twofold gain and sensitivity with respect to the units excited by side-down tilt. It appears there-
fore that the response characteristics to tilt similar to those described for the whole population of responsive units also affected the subpopulation of slow discharging LC-complex units. The amplitude and frequency characteristics of responses of LC-complex neurons to sinusoidal labyrinth stimulation were also investigated in several units, some of which were antidromically identified as CS neurons. The gain and phase angle of the responses did not vary appreciably as the amplitude ranged from 5 to 20" at 0.15 Hz, suggesting that the responses were linear ( n = 4 units). By varying the frequency of stimulation from 0.008 to 0.32 Hz at the fixed amplitude of
B
\
/M
0.1.
90'
2 sec
Fig. 4. Example of neck-vestibular interaction in a SC neuron. A. Sequential pulse density histogram averaged over 20 sweeps (128 bins, 0.1 sec bin width), showing the responses of the unit to sinusoidal stimulation of vestibular receptors, neck receptors and combined vestibular and neck receptors at the standard parameters of 0.15 Hz, & lo". Phase relation of responses to each input is indicated. The lower trace indicates the position of the head o r the neck. The response gain of this unit corresponded to 0.41 imp./sec/deg (labyrinth input), 0.31 imp./sec/deg (neck input) and 0.05 imp./sec/deg (neck labyrinth inputs); mean firing rate 3.3 imp./sec. B. The parameters of these responses are plotted vectorially; the scale along the 90" meridian refers to the gain. The computed theoretical vector (V) of response to both neck (N) and labyrinth (L) inputs showed a phase lag of - 131.4" and a gain of 0.11 imp./sec/deg. (From Manzoni et al., 1989.)
+
+
419
lo", the LC-complex neurons ( n = 8 units) showed either no change (static responses) or only a slight increase in the response gain (dynamic responses). The phase angle of the responses remained stable over the frequency range in four units, thus being attributed to stimulation of macular utricular receptors. In these instances, static and dynamic responses were observed in separate units, due to static and dynamic properties of otolith receptors. There were also units which showed an increase in phase lead of the responses at increasing frequency of tilt, probably due to some contribution of the canal input ( n = 2 units); moreover, two additional units showed a decrease in phase lead as a function of frequency. The mean gain and phase angle of the responses of all the individual units to various frequencies of tilt are shown diagrammatically in Figure 5 (upper and lower diagrams, respectively: filled circles).
Response characteristics of LC-complex neurons to sinusoidal neck rotation Barnes et al. (1989) have recorded the frequency response of 99 LC-complex neurons elicited during neck rotation at the parameters of 0.15 Hz, t10". From the total population of recorded neurons, 73 units (73.7%) displayed a response characterized by a periodic modulation of the discharge frequency in relation to the sinusoidal input. Among the units tested, 17/26 LCd neurons (65.4%), 33/46 L C a neurons (71.7%) and 23/27 SC neurons (85.2%) were responsive to neck rotation. The average gain and sensitivity of the first harmonic of responses corresponded to 0.26 k 0.40, S.D., imp./sec/deg and 2.9 -t 2.6, S.D., %/deg, respectively. However, the corresponding gain (and sensitivity) values were on the average higher for the SC (0.46 k 0.48, S.D., imp./sec/deg) and the LCd neurons (0.21 f 0.31, S.D., imp./sec/deg) than for the L C a neurons (0.16 -t 0.33, S.D., imp./sec/deg). The phase angle of the first harmonic of responses was also evaluated (Fig. 3B). Most of the responding units (40/73, i.e., 54.8%) were maxi-
j/
.LABYRINTH INPUT N=8 oNECK INPUT N=7
L C complex
T
1
0008
0015
0026
0051
015
032
O.dO8
0015
0,026
0051
0.15
0.32
$1 -45
FREQUENCY (Hzl
Fig. 5. Changes in gain (A) and phase angle (€3) of responses of LC-complex neurons to increasing frequency of sinusoidal tilt of the animal or neck rotation at the peak amplitude of lo". The filled circles show the average gain and phase angle of responses of 8 LC-complex neurons to increasing frequencies of tilt. All these units were excited by side-up and depressed during side-down tilt. On the whole, the average gain of the responses, normalized with respect to the value obtained at 0.026 Hz, increased from 0.99+0.38, S.D., at 0.015 Hz to 1.47k0.33, S.D., at 0.32 Hz (paired t-test between the means, P < 0.05). Moreover, the average phase angle of the responses evaluated with respect to the extreme side-up animal position varied from 28.7k 27.0, S.D., deg at 0.015 Hz to +31.5+30.0, S.D., at 0.32 Hz ( P > 0.05, not significant). The open circles show the average gain and phase angle of responses of 7 LC-complex neurons to increasing frequencies of neck rotation. Among these units, 5 were excited by sidedown and depressed during side-up neck rotation, while 2 units showed the opposite response pattern. The average gain of the responses, normalized with respect to the value obtained at 0.026 Hz, increased from 0.69 0.24, S.D., imp./sec/deg at 0.015 Hz to 1.46k0.63, S.D., imp./sec/deg at 0.32 Hz ( P< 0.05). Moreover, if we exclude one unit, the average phase angle of the responses evaluated with respect to the extreme side-down or side-up neck displacement decreased from +58.3 14.9, S.D., deg at 0.015 Hz to 12.7k 23.9, S.D., deg at 0.32 Hz ( P < 0.05). (From Pompeiano et aL, 1989 and Barnes et al., 1989.)
+
+
mally excited during side-down neck rotation and showed an average phase lead of +32.3 k 24.3
420
deg with respect to the extreme side-down neck displacement. A n example of these unit responses is shown in Figure 4A (middle trace). On the other hand, a much smaller group of units (18/73 units, 24.7%) were excited during side-up neck rotation and showed an average phase lag of - 141.5 f 26.0 deg, corresponding to a lead of +38.5" with respect to the extreme side-up neck displacement. In addition to these main populations of units, there were 15 units (i.e., 20.5%) which showed intermediate phase angle of the responses. Among the positional-sensitive units, which were located in the SC and the LCd, those excited by side-down neck rotation showed higher gain and sensitivity of the responses ( n = 40:0.38 f 0.50, S.D., imp./sec/deg and 3.8 k 3.7, S.D., %/deg) than those excited by side-up neck rotation ( n = 18:0.15 f 0.18, S.D., imp./sec/deg and 1.4 f 0.9, S.D., %/deg) (t test between the means P > 0.05 and P < 0.01 for differences in gain and sensitivity values of the responses, respectively). Similar response properties were found in 11 out of 14 antidromically identified CS neurons which responded to neck rotation. In fact, the majority of these responsive neurons were located in the SC ( n = 10 units) and showed large amplitude responses characterized by excitation during side-down neck rotation ( n = 9 units). Among the whole population of recorded neurons, 49 units (13 located in the LCd, 27 in LCa and 9 in SC) showed a base frequency lower than 5 imp./sec; 7 out of these units were antidromically identified as CS neurons (mean conduction velocity, 34.1 f 18.1, S.D., m/sec). Most of these slow discharging units (33/49, i.e., 67.3%) responded to neck rotation at 0.15 Hz, +lo" and showed an average gain and sensitivity of 0.12 f 0.14, S.D., imp./sec/deg and 3.2 k 2.8, S.D., %/deg, respectively. Moreover, the majority of these responsive units (18/33, i.e., 54.6%) were excited during side-down neck rotation, while a smaller group of units (7/33, i.e., 21.2%) were excited during side-up neck rotation; eight units (i.e., 24.2%) showed intermediate responses. Moreover, among the units which were excited by
the extreme neck displacement, those excited by side-down neck rotation showed more than a twofold larger gain and sensitivity with respect to the units excited during side-up neck rotation. These findings were similar to those described for the whole population of responsive units. The effects of changing amplitude and frequency of neck rotation were also investigated on several LC-complex units, some of which antidromically identified as CS neurons. The amplitude of modulation typically increased by increasing the amplitude of neck displacement from 2.5 to lo"; however, the average gain slightly decreased ( n = 4 units). These changes in amplitude, however, did not modify the phase angle of the response. When the frequency of stimulation was varied from 0.008 to 0.32 Hz at the fixed amplitude of lo", the LC-complex neurons ( n = 7 units) either did not show changes in the response gain (static responses) or showed a slight increase in the response gain (dynamic responses). The phase angle of the responses, evaluated with respect to the extreme neck displacement, usually showed a decrease in phase lead of the responses. The mean gain and phase angle of the responses to various frequencies of neck rotation are shown diagrammatically in Figure 5 (upper and lower diagrams, respectively; empty circles).
Convergence and interaction of labyrinth and neck inputs on LC-complex neurons The degree of convergence and the modality of interaction of both labyrinth and neck inputs on LC-complex neurons were also investigated (Manzoni et al., 1989). Among a population of 90 LC-complex neurons tested during sinusoidal stimulation of vestibular and neck receptors at standard parameter, 52 units (57.8%) responded to independent stimulation of both types of receptors and 10 out of these units were antidromically activated by stimulation of the spinal cord at T12-Ll. Histological controls indicated that the proportion of convergent units was higher in the SC (20/26, i.e., 76.9%) than in the LC (32/64, i.e., 50.0%), the value found in the LCd (12/24,
421
i.e., 50.0%) being equal to that in the L C a (20/40, i.e., 50.0%). The base frequency of all the convergent units corresponded on the average to 9.6 -t 11.8, S.D., imp./sec. The average gain and sensitivity of all the convergent units were slightly higher for the neck (0.34 k 0.45, S.D., imp./sec/deg and 3.55 -t 2.82, S.D., %/deg) than for the labyrinth responses (0.23 f 0.29, S.D., imp./sec/deg and 3.13 -t 3.04, S.D., %/deg; paired t-test between the means, P < 0.001 for differences in both gain and sensitivity values). Moreover, the average gain (and sensitivity) values of all the convergent SC neurons to neck (0.51 k 0.50, S.D., imp./sec/deg, n = 20) or labyrinth stimulation (0.37 k 0.37, S.D., imp./sec/deg, n = 20) were about twice as high as the corresponding values obtained from all the convergent LCd and LCa neurons in response to the neck (0.23 k 0.39, S.D., imp./sec/deg, n = 32) or the labyrinth input (0.14 k 0.18, S.D., imp./sec/deg, n = 321, the difference being statistically significant (t test between the means, P < 0.05 and P < 0.01 for gain values of neck and labyrinth responses, respectively; similar differences were also obtained for the sensitivity values). With respect to the phase angle of the responses, the majority of the convergent units were excited by side-down neck rotation (29/52 units, i e . , 55.8%) and by side-up tilt of the whole animal (29/52 units, i.e., 55.8%). The corresponding phase angle values were on the average similar to those obtained from the whole population of responsive units. The convergent units were subdivided according to the difference in phase angle of responses to the individual neck and labyrinth inputs elicited by standard parameters of stimulation ( A 4 ) . Most of these units, 10 of which were antidromically activated by stimulation of the spinal cord at T12-L1, were characterized by reciprocal responses, i.e., A 4 > 90" (43/52, i.e., 82.7%), while the other units showed parallel responses to both inputs, i.e., A 4 < 90" (9/52, i.e., 17.3%). In particular, the average difference in the phase angle of the responses to both inputs corresponded to
158.9 -t 19.8, S.D., deg for the former population of units and to 45.5 -t 21.5, S.D., deg for the latter population. Moreover, among the units characterized by reciprocal responses, the majority were excited by side-down neck rotation and side-up tilt of the whole animal; in contrast, the units characterized by parallel responses did not show any preferential distribution of the phase angle of the responses. Among the 52 LC-complex neurons receiving convergent neck and vestibular inputs, 29 were submitted not only to roll tilt of the animal and neck rotation, but also to head rotation, a procedure which led to costimulation of neck and labyrinth receptors. The main result of these experiments was that the gain and the phase angle of the responses of individual units to combined stimulation of both types of receptors were similar to the component values of the vector predicted by a linear summation of the separate labyrinth and neck responses. Some nonlinearities of the responses to head rotation, however, were recorded from the majority of LC neurons and concerned both the gain as well as the phase angle values of these responses. In fact, a large number of unit responses to head rotation as evaluated experimentally showed a smaller gain, but a more prominent phase lead with respect to the values obtained on the basis of the vectorial analysis of the responses. Figure 4 illustrates an example of these units displaying the most common pattern of reciprocal convergence. In addition to the mean gain and phase angle of the individual neck and vestibular responses, the polar diagram (B) shows the corresponding values of the responses to combined stimulation of both types of receptors obtained experimentally (see also A, lower trace), or predicted according to the vectorial analysis.
Additional properties of the LC-complex neurons Although the observations reported in the first section of the Results suggest that at least a proportion of the recorded units responsive to sinusoidal stimulation of labyrinth and neck re-
422
ceptors belong to NE-containing neurons, further experiments are required to support this conclusion. One of the main characteristics of the noradrenergic LC neurons is that in intact, unanesthetized animals, these units undergo spontaneous fluctuations of their discharge rate in relation to the levels of vigilance. In particular, the LC neurons exhibit a slow regular discharge during waking, which declines during synchronized sleep and virtually disappears during the episodes of desynchronized or REM sleep (cf. Sakai, 1980; Foote et al., 1983; Hobson and Steriade, 1986). These episodes are characterized not only by EEG desynchronization and bursts of REM, but also by a complete suppression of the postural activity (Jouvet, 1962) associated with a tonic
depression of the spinal reflexes (Pompeiano, 1966, 1967a). Since REM episodes associated with postural atonia (Matsuzaki, 1969; Magherini et al., 1972; Seguin et al., 1973; Mergner et al., 1976; cf. Pompeiano, 1980) and with a tonic depression of the spinal reflexes similar to those occurring spontaneously during desynchronized sleep could also be elicited in precollicular decerebrate cats after systemic injection of an anticholinesterase (cf. Pompeiano, 19801, experiments were performed by Pompeiano and Hoshino (1975, 1976a,b) to find out: (1) whether even after decerebration the LC-complex neurons showed a state-dependent behavior in relation to different levels of postural activity, and (2) whether the same units also responded to labyrinth stimulation.
A 1
A
2
Yiii
i
i i u
U\
ii
i
20 sec
Fig. 6 . Polygraph records of a LC neuron during the state of postural activity as well as during the episode of postural atonia induced by an anticholinesterase. Precollicular decerebrate cat 1, electromyogram (EMG) of the ipsilateral triceps brachii; 2, electro-oculogram (EOG); 3, root mean square activity of the LC neuron. A. Regular discharge of the unit in the presence of a good extensor rigidity. B,C. Inhibition of the unit activity during the episode of postural atonia produced by iv injection of 0.1 mg/kg of eserine sulfate, and complete recovery of the unit discharge following the reappearance of the postural rigidity induced by pressure applied to the skin of the ipsilateral forelimb (arrow). (From Pompeiano and Hoshino, 1976a.)
423
Among 251 units recorded from different pontine regions and tested before, during, and after episodes of postural atonia produced by iv injection of 0.1 mg/kg of the anticholinesterase eserine sulfate, 31 showed a regular activity in the
decerebrate preparation. Although these units discharged at an average firing rate higher than that reported in intact animals (see also above), their activity was suppressed during the episodes of postural atonia induced by the anticholin-
A I
2
3 4
8
.
C
D
\
-
5 sec Fig. 7. Effects of static tilt of the whole animal at the peak amplitude of 15" on a dorsal pontine reticular neuron during the control period as well as during the episode of postural atonia produced by an anticholinesterase. Precollicular decerebrate cat. 1, EMG of the ipsilateral triceps brachii; 2, EOG; 3, root mean square activity of the pontine neuron recorded from the right central tegmental field; 4, position of the tilt table plotted from the output of a pontentiometric recorder; upward deflection corresponds to tilt to the left side, downward deflection tilt to the right side. A,B. Records taken during the control period, showing an increased discharge during contralateral tilt (A) and a reduced discharge during ipsilateral tilt (B). C,D. Suppression of the static responses of the unit to both contralateral (C) and ipsilateral tilt (D) during the episode of postural atonia produced by iv injection of 0.1 mg/kg of eserine sulfate. Note the short-lasting discharge of the unit at the beginning of the contralateral tilt. (From Pompeiano and Hoshino, 1976a.)
424
esterase. This finding preceded by 5-60 sec the abolition of the postural tonus and lasted throughout the cataleptic episode, although recovering as soon as the decerebrate rigidity reappeared either spontaneously or following somatosensory stimulation ( i e . , compression of the skin). Figure 6 illustrates one of these units. Histological controls demonstrated that most of these neurons were located in the LC and SC, as well as in the neighbouring pontine reticular formation. It is of interest that all these units, which showed a suppression of their discharge during the episodes of postural atonia induced by the anticholinesterase, displayed steady changes in their discharge rate when tested during static tilt of the animal ( f lSo), a finding which persisted throughout the tilted position, thus being attributed to stimulation of macular, utricular receptors. In particular, most of the units showed a steady increase in their discharge rate during side-up tilt and a decrease during side-down tilt (p-responses). An example of these units is illustrated in Figure 7, A,B. In this case the average discharge frequency, which corresponded to 9.0 imp./sec when the animal was kept in a horizontal position, increased to 23.9 imp./sec during side-up tilt (A) and decreased to 6.7 imp./sec during side-down tilt (B). The results described above agree with the observations reported in earlier sections, insofar as: (1) the mean discharge rate of the LC neurons recorded in decerebrate cats was apparently higher than that reported in intact animals, and (2) the predominant response pattern of the LCcomplex neurons and the neighbouring pontine reticular neurons to static tilt of the animal closely corresponded to that obtained during dynamic (sinusoidal) tilt (see Pompeiano et al., 1990). Moreover, the observation that in decerebrate cats the discharge of the LC-complex units and the neighbouring pontine reticular neurons showed a state-dependent behavior during the episodes of postural atonia induced by the anticholinesterase, similar to that occurring in intact animals during desynchronized sleep, supports
the conclusion that these neurons were probably noradrenergic in nature. A final observation by Pompeiano and Hoshino (1976a,b) was that the same labyrinth input which produced reciprocal changes in firing rate of the units if applied on a background of postural rigidity, was unable to elicit any responses during the cataplectic episodes induced by the anticholinesterase. Figure 7C,D illustrates the great depression of the spontaneous discharge of a dorsal pontine neuron during the cataplectic episode induced by iv injection of 0.1 mg/kg of eserine sulfate (compare with Fig. 7A,B), as well as the suppression of the responses of the same unit to animal tilt. Since the LVN neurons, which transmit the labyrinth input to the brainstem structures, did not show any changes in their discharge rate during the cataplectic episodes induced by the anticholinesterase (Thoden et al., 19721, it is reasonable to conclude that the inability of the LC-complex neurons and the neighbouring neurons to respond to the labyrinth input during the induced cataplectic episode did not depend upon a reduced efficacy of the orthodromic labyrinthine volleys in their course through the LVN, but rather upon a suppression of the responses of the noradrenergic and NE-sensitive LC-complex neurons to the labyrinth input, due to postsynaptic inhibitory mechanisms. Discussion Population of recorded LC-complex neurons In the experiments reported in earlier sections of the Results, we selected units which were histologically found in the LC complex, i.e., in the LCd, the L C a and the SC, where NE-containing neurons are located (cf. Jones and Friedman, 1983). Moreover, a proportion of these neurons were physiologically identified as CS neurons projecting to the lumbosacral segments of the ipsilatera1 spinal cord. There is in fact evidence that the LC complex projects axons to the spinal cord, as shown by employing both anatomical (cf. West-
425
lund and Coulter, 1980; Holstege and Kuypers, 1982; Kuypers and Huisman, 1982; Westlund et af., 1982, 1983, 19841, as well as physiological methods (Guyenet, 1980; Fung and Barnes, 1981; Nakazato, 1987). Demonstrating that all the recorded units were confined within the LC and SC area does not prove that they belonged to NE-containing neurons. In fact, the physiological characteristics of noradrenergic neurons can be determined mainly in the rat, where the LC consists of a compact cluster of these neurons (cf. Foote et af., 1983). In the cat, however, in which the noradrenergic LC-complex neurons are sparse and intermingled with non-noradrenergic ones, the neurochemical identity of the recorded neurons can be assessed only by combining intracellular recording experiments with histoimmunochemistry. Unfortunately this approach could hardly be applied to our experiments, because of the difficulty in obtaining stable intracellular recording from LC-complex neurons during our dynamic test stimulations. As reported above, some of the LC-complex neurons recorded in our experiments were found to be antidromically activated from the ipsilateral cord at T12-L1. In these instances the conduction velocity of the corresponding CS axons was, on the average, higher than that reported in the rat (Guyenet, 1980). This finding, which fits with the results of previous experiments (Fung and Barnes, 19811, should be compared with that obtained in the same animal species by Nakazato (1987), who reported conduction velocities from less than 1 to 33 m/sec for CS neurons antidromically activated by stimulating the high cervical cord at Cl-C2. The assumption put forward by Nakazato (19871, ix.,that only unmyelinated fibers with conduction velocities of less than 1 m/sec, which corresponded to less than 10% of the recorded units, were noradrenergic would exclude the noradrenergic nature of our recorded units. This conclusion, however, is not supported by the fact that about 85% of all CS neurons in the cat contain NE (Reddy et af., 1989). The demonstration that most of the CS neurons originated from the ven-
tral part of the LC and the SC (Westlund et al., 1983, 1984), where medium-sized and large, multipolar cells are located (Loughlin et al., 1986a,b), explains why in our experiments the conduction velocity of the spinally projecting axons was higher than originally anticipated. Direct experiments are in progress to find out whether and to what extent the NE-containing CS neurons are provided with myelinated axons in the cat. Obviously the units, which were not antidromically activated by spinal cord stimulation at T12-L1, could either reach the spinal cord at a higher level or project to supraspinal structures. In addition to a typical positive-negative extracellular spike of long duration (1.5-2 msec), the NE-containing LC-complex units, recorded in rats, showed the following physiological characteristics, namely: (1) a slow and regular resting discharge, and (2) a response to noxious stimuli characterized by a transient excitatory response followed by a reduced discharge, a finding which has been attributed, in part at least, to recurrent and/or lateral inhibition of the noradrenergic neurons (cf. Foote et al., 1983; Ennis and AstonJones, 1986a; see Introduction). In our decerebrate cats (see above) the mean firing rate of the LC-complex units (9.6 imp./sec) was higher than that reported in intact cats (cf. Sakai, 1980; Foote et al., 1983; Hobson and Steriade, 19861, but comparable to that obtained in previous experiments, which were also performed in decerebrate preparations (Pompeiano and Hishino, 1975, 1976a,b). This finding can be explained by assuming that decerebration interrupts a descending pathway exerting a tonic inhibitory influence on the LC neurons. That the resulting increase in resting discharge of LC-complex neurons actually contributes to the y-rigidity which occurs after decerebration (cf. Granit, 1970), is supported by the fact that the LC neurons exert not only a facilitatory influence on extensor amotoneurons, but also an inhibitory influence on Renshaw cells (Fung et af., 1987, 1988>,a finding which would lead to disinhibition and thus to an increased discharge of both the y-motoneurons
426
and the small-size a-motoneurons (cf. Pompeiano, 1984). An additional observation by Pompeiano and Hoshino (1976a,b), reported in the last section of the Results, was that a decrease or suppression of the resting discharge of LC-complex neurons occurred in decerebrate cats during the episodes of postural atonia induced by systemic injection of an anticholinesterase (cf. Pompeiano, 1980). These episodes closely resembled those occurring in intact animals during desynchronized sleep (Jouvet, 1962). In both instances, in fact, the postural activity was associated with bursts of REM. Moreover, a suppression of the resting discharge of the LC-complex neurons, similar to that induced by the anticholinesterase in decerebrate cats, was also observed in intact animals during desynchronized sleep (cf. Sakai, 1980; Foote et al., 1983; Hobson and Steriade, 1986). This finding actually represents one of the main characteristics generally attributed to the NEcontaining neurons. The demonstration that the LC-complex units, which showed a suppression of their firing rate during the episode of postural atonia induced by the anticholinesterase, also responded to static tilt of the animal leading to stimulation of macular receptors (see Fig. 7), suggests that these neurons are of noradrenergic origin. As to the typical pattern of response of the LC-complex units to compression of the paws, it was found that even after decerebration a large proportion of units responded with a burst of excitation followed by a period of quiescence, as reported previously in rats (cf. Foote et al., 1983; Ennis and Aston-Jones, 1986a). The units which were simply inhibited by the stimulus probably escaped the direct excitatory input from the periphery, being submitted only to recurrent and/or lateral inhibition of neighbouring LC-complex neurons driven by the pinch stimulus. In conclusion, it appears that at least a proportion of the LC-complex neurons recorded in decerebrate cat showed the physiological characteristics which are typical of the NE-containing neurons.
Response characteristics of LC-complex neurons, including CS neurons, to labyrinth and neck stimulation The main result of our experiments was that a large proportion of LC-complex neurons (80/141 units, i.e., 56.7%) particularly located in the caudal part of the LCd and rostra1 part of the SC, some of which were antidromically identified as CS neurons, exhibited a periodic modulation of their firing rate during sinusoidal tilt of the animal at the standard parameters of 0.15 Hz, k lo", being either excited during side-up (56.2%) or during side-down (25.0%) animal tilt. Moreover, these two populations of neurons showed an average phase lead of +17.9" with respect to the extreme animal displacements, thus being attributed to stimulation of macular, utricular receptors. By increasing frequency of stimulation from 0.008 to 0.32 Hz at the peak amplitude of lo", which raised the maximum angular acceleration from 0.025 to 41.7"/s2, four out of the eight tested units showed either a stability or a decrease in phase lead of the responses. These findings, which occurred in spite of the increase in peak angular acceleration above threshold for canal-induced responses of vestibular nuclei neurons, can be attributed to stimulation of macular receptors, since a relative stability in the phase angle of the responses was also obtained in cats (Anderson et al., 1978) and monkey (Fernandez and Goldberg, 1976) by recording the responses of individual utricular afferents to increasing frequency of sinusoidal tilt. Moreover, the response gain of the four LC neurons either did not change or showed only a slight increase by increasing frequency of stimulation, suggesting that both static and dynamic components of the response were due to the static and dynamic sensitivity of otolith receptors. In addition to these units, two units exhibited a decrease in phase lead of the responses, as a function of frequency, a finding which could still be attributed to stimulation of otolith receptors if we assume that some interaction had occurred between the afferent volleys
427
originating from both labyrinths. Finally the last two units exhibited an increase in phase lead of the responses that tended to be related to the velocity signal during increase in head angular acceleration; this increase in phase lead was also associated with an increase in response gain, as expected if both macular and canal inputs converged on the same unit. Similar to the findings reported above, a large proportion of LC-complex neurons (73/99 units, ie., 73.7%), some of which were antidromically identified as CS neurons, responded to sinusoidal rotation of the neck about the longitudinal axis at 0.15 Hz, k lo". Most of these neurons responded to the direction of neck orientation, being either excited during side-down (54.8%) or during sideup (24.7%) neck rotation. Moreover, these populations of neurons showed an average phase lead of +34.2" with respect to neck position suggesting than neck angular displacement, rather than neck angular velocity, contributed to the neuronal responses. A n increase in frequency of neck rotation from 0.008 to 0.32 Hz at the peak amplitude of lo", while producing a slight increase in response gain, greatly decreased the average phase lead of the responses relative to the neck displacement. This finding could be attributed in part at least to vectorial interaction of the neck inputs originating from both the ipsilateral and the contralateral side. Since in the present experiments the skin of the neck was denervated and the dorsal neck muscles were disconnected bilaterally, the responses of LC-complex neurons to neck rotation depended on stimulation of deep receptors innervated by the upper cervical afferents, which could originate either from the neck joint and tendon organ receptors (McCouch et al., 1951; Cohen, 1961) or else from primary endings of neck muscle spindles located in the small perivertebral muscles, closely related to the atlanto-occipital and atlanto-axial membranes, as shown in both anatomical (Backer and Richmond, 1982; Richmond and Backer, 1982) and physiological studies
(Richmond and Abrahams, 1979; Chan et al., 1987).
Comparison between responses of CS and lateral VS neurons to animal tilt and neck rotation The responses of LC-complex neurons, including the CS neurons, to animal tilt and to neck rotation should be compared with those recorded in decerebrate cats from LVN neurons (Boyle and Pompeiano, 1980a,b; Schor and Miller, 1982). If we consider the activity of VS neurons antidromically activated by stimulation of the spinal cord at T12-L1, thus projecting to the lumbosacral segments (Marchand et al., 1987), it appears that the proportions of VS neurons responsive to roll tilt of the animal and neck rotation (58.9% and 68.8%) were almost comparable to those found in the present experiments for the LC-complex neurons (56.7% and 73.7%, respectively). An additional finding concerns the predominant pattern of response of the VS neurons to animal tilt and to neck rotation. If we consider the responses of the VS neurons to animal tilt, it appears that most of the VS neurons activated by the extreme animal displacements (51/66, i.e., 77.3%) were excited by side-down tilt (a-responses, Fig. 8A), a finding consistent with the pattern of utricular input (cf. Wilson and Melvill Jones, 1979). Moreover, most of these units showed a larger gain than that of the units displaying the opposite response pattern (Marchand et al., 1987). In contrast to these findings, the majority of the LC-complex neurons activated by the extreme animal displacements (45/65, i.e., 69.2%) were excited during side-up tilt of the animal (@-responses, Fig. 3A). Moreover, these units showed more than a twofold larger gain with respect to the units excited by side-down tilt. As to the responses of the VS neurons to the neck input (Marchand et al., 19871, it appeared that the majority of these neurons activated by the extreme neck displacements were excited during side-up neck rotation (see Fig. 8B) 69/67, i.e., 88.1%); moreover, most of these units had a
428
A
NUMBER OF UNITS
LVN
+135_-:180
-135
-9b
-45
-1nSf339.S D
0
'45
+%
113-0
-250f238SD
R U
I
NECK INPUT N=75
'135
-180
-135
-128 5tq8 1 S D
LAG
-90
-45
0
+45
+k
+135
& l
'559!210SD
PHASE (DEGI
LEAD
Fig. 8. Distribution of the phase angle of the first harmonic of responses of'LVN neurons tested during roll tilt of the animal (A) and neck rotation (B) at 0.026 Hz, k 10". All the neurons were antidromically activated by stimulation of the spinal cord at T12-L1, thus projecting to the lumbosacral segments of the spinal cord. Scale of abscissas as in Figure 3. Most of the units were excited during side-down tilt of the animal 6 1 / 7 6 neurons) and side-up neck rotation 69/75 neurons). (From Marchand et al., 1987.)
larger gain than that of the units displaying the opposite response pattern. In contrast to these findings, the majority of the LC-complex neurons affected by the extreme animal displacements (40/58, i.e., 69.0%) showed the opposite response pattern, being excited during side-down neck rotation (Fig. 3B). Moreover, these units showed more than a twofold larger gain with respect to the units excited by side-up neck rotation. We postulate that the vestibular input of one side is transmitted not only to the ipsilateral LVN, but also via a crossed pathway, to the
ventral aspect of the contralateral medullary reticular formation (Kubin et al., 1980; Manzoni et al., 1983b), whose neurons project at least in part, to the LC complex (Aston-Jones et al., 1986) to exert a prominent excitatory influence on the NE-containing neurons (Ennis and Aston-Jones, 1986b, 1987). One of the main channels transmitting the vestibular input of one side to the contralateral medullary reticular structures is represented by the lateral VS tract acting on neurons of the crossed spinoreticular pathway (Corvaja et al., 1977; cf. Pompeiano, 1979). In this way the labyrinthine volleys originating from macuIar receptors of one side during side-down tilt would be responsible not only for the a-responses of VS neurons originating from the ipsilateral LVN, but also for the P-responses of CS neurons originating from the contralateral LC complex. Even the responses of the LC-complex neurons to the neck input, in particular those characterized by excitation during side-down neck rotation, could at least in part be elicited by volleys passing through the medullary RF. There is, in fact, evidence that the proprioceptive neck input of one side impinges on the ventral aspect of the medullary R F (Kubin et al., 1981a; Srivastava et al., 19841, whose neurons may, in part at least, project to the ipsilateral LC-complex where they exert an excitatory influence as reported above. The main channel transmitting the neck input of one side to the ipsilateral medullary reticular structure is represented by the uncrossed spinoreticular pathway (Corvaja et al., 1977; cf. Pompeiano, 1979). The reciprocal behavior of the VS neurons with respect to the LC-complex neurons during neck rotation suggests that the neck input of one side, while exciting the ipsilateral medullary reticular structure and the related LC-complex neurons, inhibits the VS neurons of the same side. This effect has been attributed to the fact that the neck input of one side, driven during side-down neck rotation, excites not only reticulocerebellar neurons located in the ipsilatera1 lateral reticular nucleus (Kubin et al., 1981a)
429
but also the related Purkinje cells of the cerebellar hemivermis (De Noth et al., 1980), which in turn project to the dorsal aspect of LVN (Corvaja and Pompeiano, 1979), thus inhibiting their neurons (cf. Pompeiano, 1967b; Ito, 1984). In this way the afferent volleys originating from neck receptors during side-down neck displacement could be responsible not only for the increased discharge of ipsilateral LC-complex neurons, but also for the reduced discharge of ipsilateral VS neurons which occurs for this direction of neck orientation. However, the possible contribution of crossed projections transmitting the neck input of one side to the contralateral pontine tegmental and vestibular structures cannot be excluded. It is of interest that a large proportion of LC-complex neurons, including the CS neurons projecting to the lumbosacral segments of the spinal cord, received a convergent input from both labyrinth and neck receptors, as shown for the LVN neurons, including the VS neurons projecting to the lower segments of the spinal cord (Boyle and Pompeiano, 1981; Stampacchia et at., 1987). Moreover, the majority of these convergent responses were organized reciprocally, i.e., outof-phase, and were characterized by excitation during side-down neck rotation and side-up tilt of the animal for the LC-complex neurons, an effect which was just opposite to that found for the VS units. The observation that the labyrinth and neck inputs act in opposition on the majority of the CS and VS neurons, and that the responses of CS are about 180" out-of-phase with respect to those of the excitatory VS neurons, may explain why changes in position of the head leave the position of the limbs unchanged (cf. von Holst and Mittelstaedt, 1950; Lindsay et al., 1976; Manzoni et al., 1983a; Ezure and Wilson, 1984). It is worth mentioning that a cancellation of the labyrinth and neck responses of LC neurons during head rotation, leading to combined stimulation of both types of receptors, occurred very rarely, as in most instances the labyrinth and neck responses had different gains and were not
precisely in opposition of phase. In these instances the responses of LC-complex neurons to head rotation were similar to those obtained by vectorial summation of the responses to the individual inputs. Similar results were also observed at the level of the VS neurons (Boyle and Pompeiano, 1981) as well as of medullary reticular neurons (Kubin et al., 1981b; Pompeiano et al., 1984). The persistence of some residual response of CS neurons during combined neck and vestibular stimulations would counteract the persistence of some residual response of VS neurons, thus leading to a resolution of the conflicting influences at the motoneuronal level.
Functional considerations The responses of LC and SC neurons to animal tilt and neck rotation could be better understood if we knew the neuronal systems receiving their efferent projections. However, the demonstration that a proportion of responsive units presumed to be, at least in part, noradrenergic were antidromically activated by stimulating the spinal cord at T12-L1, thus projecting to the lumbosacral segments, and that part of the unidentified units could still be CS neurons projecting to the upper segments of the spinal cord, led us to suggest that the LC complex might be involved in the postural adjustments during the tonic labyrinth and neck reflexes. It is known that the lateral VS neurons, which send axons to the ventral horn region of the spinal cord, where they terminate particularly in laminae VII-IX (Shinoda et al., 19861, exert a direct excitatory influence on ipsilateral limb extensor motoneurons (Lund and Pompeiano, 1968; cf. Pompeiano, 1975) and through their recurrent collaterals on the related Renshaw cells. Even the CS neurons, whose aminergic terminals are in close contact with ventral horn neurons, as shown either by histofluorescent techniques (Fuxe, 1965; Lackner, 1980; Mizukawa, 19801, by immunocytochemical staining for dopamine-P-hydroxylase (Westlund and Coulter, 1980; Westlund et al., 1983, 1984) and tyrosine-hydroxylase (Reddy et
430
al., 1989) or by anterograde autoradiographic tracing following injections of labeled amino acid into the LC and SC (McBride and Sutin, 1976; Holstege et al., 1979; Westlund and Coulter, 1980; Jones and Yang, 1985; cf. Holstege and Kuypers, 1982; Westlund et al., 19821, exert an excitatory influence on ipsilateral limb extensor (and flexor) motoneurons (Fung and Barnes, 1981,1987; Chan et al., 1986). This effect, though probably directly mediated (see White et al., this volume) can be attributed, at least in part, to disinhibition of spinal cord motoneurons (Fung et al., 1987,1988). Stimulation of the LC, performed in precollicular decerebrate cats, may in fact decrease the recurrent inhibition acting on both extensor and flexor motoneurons (Fung et al., 19881, an effect which can be attributed to CS inhibition of Renshaw cell activity, as shown by experiments of unit recording (Fung et al., 1987). The descending influences on a-motoneurons and Renshaw cells described above may originate from noradrenergic LC and SC neurons, since iontophoretic application of NE depolarizes the a-motoneurons (White et al., this volume), but inhibits the Renshaw cell activity (Biscoe and Curtis, 1966; Engberg and Ryall, 1966; Weight and Salmoiraghi, 1966). It would be of interest to find out whether these effects are mediated through different types of adrenoceptors. As a result of these findings, it appears that while both VS as well as CS neurons exert a synergistic influence on limb extensor motoneurons, they exert an antagonistic influence on the corresponding Renshaw cells (R-cells). The demonstration that the LC inhibits R-cells linked with both extensor and flexor motoneurons suggested that this structure could, by increasing or decreasing their neuronal discharge, modify the functional coupling between R-cells and the corresponding motoneurons, thus altering the input-output relation of a-motoneurons to a given excitatory input (Hultborn et al., 1979; cf. Pompeiano, 1984). If we consider the changes in firing rate of the VS and the CS neurons projecting to the lumbosacral segments of the spinal cord which occur
during the VS and the cervicospinal reflexes, it appears that the contraction of ipsilateral limb extensors during side-down animal tilt and side-up neck rotation (see Introduction), depends on the increased discharge of excitatory VS neurons (Boyle and Pompeiano, 1980a,b; Schor and Miller, 1982; Marchand et al., 1987). This would increase the activity of limb extensor motoneurons and the related Renshaw cells. However, the reduced discharge of the LC-complex neurons, particularly of the CS neurons projecting to the same segments of the spinal cord, to the same direction of animal and neck orientation, would lead not only to disfacilitation of extensor motoneurons but also to disinhibition of the Renshaw cells anatomically coupled with these extensor motoneurons, a finding which would increase the functional linkage of these inhibitory interneurons with their own motoneurons. The increased recruitment of the Renshaw cells just at the time in which the motoneuronal discharge is elicited by the VS volleys might explain why in precollicular decerebrate cats the amplitude of the EMG modulation and thus the response gain of limb extensors to labyrinth and neck stimulation was quite small in the forelimbs and almost absent in the hindlimbs (Manzoni et al., 1983a, 1984). This interpretation of the experimental findings is supported by the results of experiments performed in decerebrate cats, showing that anatomical (d’Ascanio et al., 1985) or functional inactivation of some of the noradrenergic LCcomplex neurons, produced by local injection of minute doses of the a,-adrenergic agonist clonidine (Pompeiano et al., 1987), which acts on the somato-dendritic a,-adrenoceptors by enhaixing recurrent and/or lateral inhibition of the corresponding NE-containing neurons (cf. Foote et al., * 1983), decreased the postural activity in the animal at rest. However, the reduced coupling of R-cells with the extensor motoneurons during side-down animal tilt and side-up neck rotation would lead to derecruitment of the Renshaw cells for these directions of animal tilt and neck rotation, thus enhancing the response gain of the
431
corresponding muscles to labyrinth and neck stimulations. Experiments of units recording from Renshaw cells anatomically linked with limb extensor motoneurons in normal decerebrate cats, as well as under conditions in which the activity pf LC neurons was decreased following injection of an anticholinesterase, support this conclusion (Pompeiano et al., 1985a,b). It is worth noticing that when the anatomical or functional inactivation of the noradrenergic LC-compiex neurons was prominent enough to produce episodes of postural atonia in the animal at rest (d’Ascanio et al., 1989; Pompeiano, 0. et al., unpublished observation), the response gain of limb extensors to roll tilt of the animal (at 0.15 Hz, klO0) was so depressed that no modulation occurred in the EMG. An almost complete inactivation of the noradrenergic LC-complex neurons would, in fact, lead not only to a complete suppression of the responses of these neurons to labyrinth stimulation, as shown in Figure 7C and D, but also to a steady disinhibition of the Renshaw cells, leading to a prominent postsynaptic inhibition of the related extensor motoneurons. This would suppress not only the postural activity in the animal at rest, but also the responses of the corresponding limb extensor motoneurons to activation of the VS reflex arc. We postulate, therefore, that the LC-complex neurons play a prominent role in the control of posture as well as in the gain regulation of the VS reflexes. Since the LC-complex neurons undergo spontaneous fluctuations in their firing rate leading to changes in postural activity during the sleep-waking cycle (cf. Sakai, 1980; Foote et al., 1983; Hobson and Steriade, 19861, they may intervene in order to adapt to the animal state the response gain of limb extensors to labyrinth and neck stimulations. Acknowledgements
This work was supported by National Institute of Neurological and Communicative Disorders and Stroke Research Grant NS 07685-22 and by
grants of the Minister0 dell’Universit6 e della Ricerca Scientifica e Tecnologica, and the Agenzia Spaziale Italiana, Roma, Italy. References Aghajanian, G.K., Van der Maelen, C.P. and Andrade, R. (1983) Intracellular studies on the role of calcium in regulating activity and reactivity of locus coeruleus neurons tn uiL)o.Brain Rex, 273: 237-243. Anderson, J.H., Blanks, R.H.I. and Precht, W. (1978) Response characteristics of semicircular canal and otolith systems in cat. 1. Dynamic response of primary vestibular fibers. Exp. Brain Res., 32: 491-507. Aston-Jones, G., Ennis, M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986) The brain nucleus locus coeruleus: Restricted afferent control of a broad efferent network. Science, 234: 734-737. Bakker, D.A. and Richmond, F.J.R. (1982) Muscle spindle complexes in muscles around upper cervical vertebrae in the cat. J. Neurophysiol., 49: 62-74. Barnes, C.D., Manzoni, D., Pornpeiano, O., Stampacchia, G. and d’Ascdnio, P. (1989) Responses of locus coeruleus and subcoeruleus neurons to sinusoidal neck rotation in decerebrate cat. Neuroscience, 31: 371-392. Biscoe, T.J. and Curtis, D.R. (1966) Noradrenaline and inhibition of Renshaw cells. Science, 151: 1230-1231. Boyle, R. and Pornpeiano, 0. (1980a) Reciprocal responses to sinusoidal tilt of neurons in Deiters’ nucleus and their dynamic characteristics. Arch. Ztal. Bid., 118: 1-32. Boyle, R. and Pompeiano, 0. (1980b) Responses of vestibulospinal neurons to sinusoidal rotation of the neck. J. Neurophysiol., 44: 633-649. Boyle, R. and Pompeiano, 0. (1981) Convergence and interaction of neck and macular vestibular inputs on vestibulospinal neurons. J. Neurophysiol., 45: 852-868. Chan, J.Y.H., Fung, S.J., Chan, S.H.H. and Barnes, C.D. (1986) Facilitation of lumbar monosynaptic reflexes by locus coeruleus in the rat. Brain Res., 369: 103-109. Chan, Y.S., Kasper, J. and Willis, V.J. (1987) Dynamics and directional sensitivity of neck muscle spindle receptors to head rotation. J. Neurophysiol., 57: 1716-1729. Cohen, L.A. (1961) Role of eye and neck proprioceptive mechanisms in body orientation and motor coordination. J. Neurophpiol., 2 4 1-1 1. Corvaja, N. and Pompeiano, 0. (1979) Identification of cerebellar corticovestibular neurons retrogradely labeled with horseradish peroxidase. Neuroscience, 4: 507-515. Corvaja, N., Grofovi, I., Pompeiano, 0. and Walberg, F (1977) The lateral reticular nucleus in the cat. I. An experimental anatomical study of its spinal and supraspinal afferent connections. Neuroscience, 52: 537-553. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Scand., 62, Suppl., 232: 1-55.
432 d'Ascanio, P. Bettini, E. and Pompeiano, 0. (1985) Tonic inhibitory influences of locus coeruleus on the response gain of limb extensors to sinusoidal labyrinth and neck stimulations. Arch. Ital. Biol., 123: 69-100. d'Ascanio, P., Pompeiano, M. and Tononi, G. (1989) Inhibition of vestibulospinal reflexes during the episodes of postural atonia induced by unilateral lesion of the locus coeruleus in the decerebrate cat. Arch. Ital. Biol., 127: 63-79. Denoth, F., Magherini, P.C., Pompeiano, 0. and StanojeviC, M. (1980) Responses of Purkinje cells of cerebellar vermis to sinusoidal rotation of neck. J. Neurophysiol., 43: 46-59. Egan, T.M. and North, R.A. (1985) Acetylcholine acts on m2-muscarinic receptors to excite rat locus coeruleus neurons. Br. J. Pharmacol., 85: 733-735. Engberg, I. and Ryall, R.W. (1966) The inhibitory action of noradrenaline and other monoamines on spinal neurones. J. Physiol. (London), 185: 298-322. Ennis, M. and Aston-Jones, G. (1986a) Evidence for self- and neighbor-mediated postactivation inhibition of locus coeruleus neurons. Brain Res., 374: 299-305. Ennis, M. and Aston-Jones, G. (1986b) A potent excitatory input to nucleus locus coeruleus from the ventrolateral medulla. Neurosci. Lett., 71: 299-305. Ennis, M. and Aston-Jones, G. (1987) Two physiologically distinct populations of neurons in the ventrolateral medulla innervate the locus coeruleus. Brain Res., 425: 275-282. Ezure, K. and Wilson, V.J. (1983) Dynamics of neck-to-forelimb reflexes in the decerebrate cat. J. Neurophysiol., 50: 688-695. Ezure, K. and Wilson, V.J. (1984) Interaction of tonic neck and vestibular reflexes in the forelimb of the decerebrate cat. Exp. Brain Res., 54: 289-292. Fernlndez, C. and Goldberg, J.M. (1976) Physiology of peripheral neurons innervating otolith organs of the squirrel monkey. 111. Response dynamics. J. Neurophysiol., 39: 996-1008. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Rec., 93: 844-914. Fung, S.J. and Barnes, C.D. (1981) Evidence of facilitatory coeruleospinal action in lumbar motoneurons of cats. Brain Res., 216: 299-311. Fung, S.J. and Barnes, C.D. (1987) Membrane excitability changes in hindlimb motoneurons induced by stimulation of the locus coeruleus in cats. Brain Res., 402: 230-242. Fung, S.J., Pompeiano, 0. and Barnes, C.D. (1987) Suppression of the recurrent inhibitory pathway in lumbar cord segments during locus coeruleus stimulation in cats. Brain Res., 402 351-354. Fung, S.J., Pompeiano, 0. and Barnes, C.D. (1988) Coeruleospinal influence on recurrent inhibition of spinal motonuclei innervating antagonistic hindleg muscles of the cat. l'flugers Arch., 412: 346-353. Fuxe, K. (1965) Evidence for the existence of monoamine neurons in the central nervous system. IV. Distribution of monoamine terminals in the central nervous system. Acta Physivl. Scand., 64, Suppl., 247: 37-85.
Granit, R. (1970) The Basis of Motor Control. Academic Press, New York, pp. VII-346. Guyenet, P.G. (1980) The coeruleospinal noradrenergic neurons: Anatomical and electrophysiological studies in the rat. Bruin Rex, 189: 121-133. Hellon, R.F. (1971) The marking of electrode tip positions in nervous tissue. J. Physiol. (London), 214: 12P. Hobson, J.A. and Steriade, M. (1986) Neuronal basis of behavioral state control. In F.E. Bloom (Ed.), Intrinsic Regulatory System of the Brain, Section I. The Nervous System, Vol. IV, American Physiological Society, Bethesda, Maryland, pp. 701-823. Holstege, G. and Kuypers, H.G.J.M. (1982) The anatomy of brain stem pathways to the spinal cord in the cat. A labeled amino acid tracing study. In K.G.J.M. Kuypers and G.F. Martin (Eds.), Anatomy of Descending Pathways to the Spinal Cord, Progress in Brain Research, Vol. 57, Elsevier, Amsterdam, New York, pp. 145-175. Holstege, G., Kuypers, K.G.J.M. and Boer, R.C. (1979) Anatomical evidence for direct brainstem projections to the somatic motoneuronal cell groups and anatomical preganglionic cell groups in cat spinal cord. Brain Res., 171: 329-333. Hultborn, H., Lindstrom, S. and Wigstrom, H. (1979) On the function of recurrent inhibition in the spinal cord. Exp. Brain Res., 37: 399-403. Ito, M. (1984) The Cerebellum and Neural Control, Raven Press, New York, pp. XVII-580. Jones, B.E. and Friedman, L. (1983) Atlas of catecholamine perikarya varicosities and pathways in the brainstem of the cat. J. Comp. Neurvl., 215: 382-396. Jones, B.E. and Yang, T.-2. (1985) The efferent projections from the reticular formation and the locus coeruleus studied by anterograde and retrograde axonal transport in the rat. J. Comp. Neurvl., 242: 56-92. Jouvet, M. (1962) Recherches sur les structures nerveuses et les mecanismes responsables des differentes phases du sommeil physiologique. Arch. Ital. B i d , 100: 125-206. Kubin, L., Magherini, P.C., Manzoni, D. and Pompeiano, 0 . (1980) Responses of lateral reticular neurons to sinusoidal stimulation of labyrinth receptors in decerebrate cat. J. Neurophysiol., 44: 922-936. Kubin, L., Magherini, P.C., Manzoni, D. and Pompeiano, 0. (1981a) Responses of lateral reticular neurons to sinusoidal rotation of neck in decerebrate cat. Neuroscience, 6: 1277- 1290. Kubin, L., Manzoni, D. and Pompeiano, 0. (1981b) Responses of lateral reticular neurons to convergent neck and macular vestibular inputs. J. Neurophysiol., 46: 48-64. Kuypers, H.G.J.M. and Huisman, A.M. (1982) The new anatomy of the descending brain pathways. In B. Sjolund and A. Bjorklund (Eds.), Brain Stem Control of Spinal Mechanisms, Elsevier/North Holland, Amsterdam, pp. 29-54. Lackner, K.J. (1980) Mapping of monoamine neurones and fibers in the cat lower brainstem and spinal cord. Anat. Embryvl., 161: 169-195. Lindsay, K.W., Roberts, T.D.M. and Rosenberg, J.R. (1976)
433 Asymmetric tonic labyrinth reflexes and their interaction with neck reflexes in the decerebrate cat. J. Physiol. (London), 261: 583-601. Loughlin, S.E., Foote, S.L. and Bloom, F.E. (1986a) Efferent projections of nucleus locus coeruleus: Topographic organization of cells of origin demonstrated by three-dimensional reconstruction. Neuroscience, 18: 291-306. Loughlin, S.E., Foote, S.E. and Grzanna, R. (1986b) Efferent projections of nucleus locus coeruleus: Morphologic subpopulations have different efferent targets. Neuroscience, 18: 307-319. Lund, S. and Pompeiano, 0. (1968) Monosynaptic excitation of alpha-motoneurons from supraspinal structures in the cat. Acfa Physiol. Scand., 73: 1-21. Magherini, P.C., Pompeiano, 0. and Thoden, U. (1972) Cholinergic mechanisms related to REM sleep. I. Rhythmic activity of the vestibulo-oculomotor system induced by an anticholinesterase in the decerebrate cat. Arch. Ital. Biol., 110: 234-259. Magnus, R. (1924) Korperstellung, Springer, Berlin, Heidelberg, New York, pp. XIII-740. Manzoni, D., Pompeiano, O., Srivastava, U.C. and Stampacchia, G. (1983a) Responses of forelimb extensors to sinusoidal stimulation of macular labyrinth and neck receptors. Arch. Ital. B i d , 121: 205-214. Manzoni, D., Pompeiano, O., Stampacchia, G. and Srivastava, U.C. (1983b) Responses of medullary reticulospinal neurons to sinusoidal stimulation of labyrinth receptors in decerebrate cat. J. Neurophysiol., 50: 1059-1079. Manzoni, D., Pompeiano, O., Srivastava, U.C. and Stampacchia, G. (1984) Gain regulation of vestibular reflexes in fore-and hindlimb muscles evoked by roll tilt. Boll. Soc. Ital. Biol. Sper., 60, Suppl., 3: 9-10. Manzoni, D., Pompeiano, O., Barnes, C.D., Stampacchia, G. and d’Ascanio, P. (1989) Convergence and interaction of neck and macular vestibular inputs on locus coeruleus and subcoeruleus neurons. Pfliigers Arch., 413: 580-598. Marchand, A.R., Manzoni, D., Pompeiano, 0. and Stampacchia, G. (1987) Effects of stimulation of vestibular and neck receptors on Deiters neurons projecting in the lumbosacral cord. Pfugers Arch., 409: 13-23. Matsuzaki, M. (1969) Differential effects of sodium butyrate and physostigmine upon the activities of para-sleep in acute brain stem preparations. Brain Rex, 13: 247-265. McBride, R.L. and Sutin, J. (1976) Projections of the locus coeruleus and adjacent pontine tegmentum in the cat. J. Comp. Neurol., 165: 265-284. McCouch, G.P., Deering, I.D. and Ling, T.H. (1951) Location of receptors for tonic neck reflexes. J. Neurophysiol., 14: 191- 195. Mergner, T., Magherini, P.C. and Pompeiano, 0. (1976) Temporal distribution of rapid eye movements and related monophasic potentials in the brain stem following injection of an anticholinesterase. Arch. Ital. Biol.,114: 75-99. Mizukawa, K. (1980) The segmental detailed topographical distribution of monoaminergic terminals and their pathways in the spinal cord in the cat. Anat. Anz., 147: 125-144. Nakazato, T. (1987) Locus coeruleus neurons projecting to the
forebrain and the spinal cord in the cat. Neuroscience, 23: 529-538. Pompeiano, 0. (1967a) The neurophysiological mechanisms of the postural and motor events during desynchronized sleep. Res. Publ. Assoc. Res. Neru. Ment. Dis.,45: 351-423. Pompeiano, 0. (1967b) Functional organization of the cerebellar projections to the spinal cord. In C.A. Fox and R.S. Snyder (Eds.), The Cerebellum, Progress in Brain Research, Vol. 25, Elsevier, Amsterdam, pp. 282-321. Pompeiano, 0. (1975) Vestibulospinal relationships. In R.F. Naunton (Ed.), The Vestibular System, Academic Press, New York, London, pp. 147-180. Pompeiano, 0. (1979) Neck and macular labyrinth influences on the cervical spinoreticulocerebellar pathway. I n R. Granit and 0. Pompeiano (Eds.), Reflex Control of Posture and Movement, Progress in Brain Research, Vol. 50, Elsevier/North Holland, Amsterdam, New York, pp. 501-514. Pompeiano, 0. (1980) Cholinergic activation of reticular and vestibular mechanisms controlling posture and eye rnovements. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Revisited, IBRO Monograph Series, Vol. 6 , Raven Press, New York, pp. 473-512. Pompeiano, 0. (1984) Recurrent inhibition. In R.A. Davidoff (Ed.), Handbook of the Spinal Cord, Vols. 2 and 3, Marcell Dekker, New York, pp. 461-557. Pompeiano, 0. and Hoshino, K. (1975) Tonic depression of pontine neurons during the episodes of postural atonia induced by an anticholinesterase in the decerebrate cat. Sleep Res., 4: 42. Pompeiano, 0. and Hoshino, K. (1976a) Tonic inhibition of dorsal pontine neurons during the postural atonia produced by an anticholinesterase in the decerebrate cat. Arch. Ital. B i d , 114: 310-340. Pompeiano, 0. and Hoshino, K. (1976b) Central control of posture: Reciprocal discharge by two pontine neuronal groups leading to suppression of decerebrate rigidity. Brain Rex, 116: 131-138. Pompeiano, O., Manzoni, D., Srivastava, U.C. and Stampacchia, G. (1984) Convergence and interaction of neck and macular vestibular inputs on reticulospinal neurons. Neuroscience, 12: 111-128. Pompeiano, O., Wand, P. and Srivastava, U.C. (1985a) Responses of Renshaw cells coupled with hindlimb extensor motoneurons to sinusoidal stimulation of labyrinth receptors in the decerebrate cat. Pflugers Arch., 403: 245-257. Pompeiano, O., Wand, P. and Srivastava, U.C. (1985b) Influence of Renshaw cells on the gain of hindlimb extensor muscles to sinusoidal labyrinth stimulation. Pfliigers Arch., 404: 107-118. Pompeiano, O., d’Ascanio, P., Horn, E. and Stampacchia, G. (1987) Effects of local injection of the u2-adrenergic agonist clonidine in the locus coeruleus complex on the gain of vestibulospinal and cervicospinal reflexes in decerebrate cats. Arch. Ital. Biol., 125: 225-269. Pompeiano, O., Manzoni, D., Barnes, C.D., Stampacchia, G. and d’Ascanio, P. (1990) Responses of locus coeruleus and subcoeruleus neurons to sinusoidal stimulation of labyrinth receptors. Neuroscience, 35: 227-248.
434 Reddy, V.K., Fung, S.J., Zhuo, H. and Barnes, C.D. (1989) Spinally projecting noradrenergic neurons of the dorsolatera1 pontine tegmentum: A combined immunocytochemical and retrograde labeling study. Brain Res., 491: 144-149. Richmond, F.J.R. and Abraham, V.C. (1979) Physiological properties of muscle spindles in dorsal neck muscles of the cat. J. Neurophysiol., 42: 604-617. Richmond, F.J.R. and Bakker, D.A. (1982) Anatomical organization and sensory receptor content of soft tissues surrounding upper cervical vertebrae in cat. J. Neurophysiol., 48: 49-61. Sakai, K. (1980) Some anatomical and physiological properties of pontomesencephalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Reuisited, IBRO Monograph Series, Vol. 6, Raven Press, New York, pp. 427-447. Schor, R.H. and Miller, A.D. (1981) Vestibular reflexes in neck and forelimb muscles evoked by roll tilt. J. Neurophysiol., 46: 167-178. Schor, R.H. and Miller, A.D. (1982) Relationship of cat vestibular neurons to otilith-spinal reflexes. Enp. Brain Res., 47: 137-144. Seguin, J.J., Magherini, P.C. and Pompeiano, 0. (1973) Cholinergic mechanisms related to REM sleep. 111. Tonic and phasic inhibition of monosynaptic reflexes induced by an anticholinesterase in the decerebrate cat. Arch. Ital. Biol., 111: 1-23. Shinoda, Y.,Ohgaki, T. and Futami, T. (1986) The morphology of single lateral vestibulospinal tract axons in the lower cervical cord of the cat. J. Comp. NeuroI., 249: 226-24 1. Srivastava, U.C., Manzoni, D., Pompeiano, 0. and Stampacchia, G. (1984) Responses of medullary reticulospinal neurons to sinusoidal rotation of neck in the decerebrate cat. Neuroscience, 11: 473-486. Stampacchia, G., Barnes, C.D., d’Ascanio, P. and Pompeiano, 0. (1987a) Effects of microinjection of a cholinergic agonist into the locus coeruleus on the gain of vestibulospinal reflexes in decerebrate cats. Arch. Ital. BioL, 125: 107-138. Stampacchia, G., Manzoni, D., Marchand, A.R. and Pompeiano, 0. (1987b) Convergence of neck and macular
vestibular inputs on vestibulospinal neurons projecting to the lumbosacral segments of the spinal cord. Arch. Ital. Biol., 125: 201-224. Svensson, T.H., Bunney, B.S. and Aghanjanian, G.K. (1975) Inhibition of both noradrenergic and serotonergic neurons in brain by the a-adrenergic agonist clonidine. Brain Res., 92: 291-306. Thoden, U., Magherini, P.C. and Pompeiano, 0. (1972) Cholinergic mechanisms related to REM sleep. 11. Effects of an anticholinesterase on the discharge of central vestibular neurons in the decerebrate cat. Arch. Ital. Biol., 110: 260-283. von Holst, E. and Mittelstaedt, H. (1950) Das Reafferenzprinzip. Wechselwirkungen zwischen Zentralnervensystem und Peripherie. Natunvissenschaften, 37: 464-476. Weight F.F. and Salmoiraghi, G.C. (1966) Adrenergic responses of Renshaw cells. J. Pharmacol. Exp. Ther., 154: 391-397. Westlund, K.N. and Coulter, J.C. (1980) Descending projections of the locus coeruleus and subcoeruleus/medial parabrachial nuclei in monkey: Axonal transport studies and dopamine-beta-hydroxylase immunocytochemistry. Brain Res., 2: 235-264. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1982) Descending noradrenergic projections and their spinal terminations. In H.G.J.M. Kuypers and G.F. Martin (Eds.), Anatomy of Descending Pathways to the Spinal Cord, Progress in Brain Research, Vol. 57, Elsevier, Amsterdam, New York, pp. 219-238. Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1983) Noradrenergic projections to the spinal cord of the rat. Brain Res., 263: 15-31, Westlund, K.N., Bowker, R.M., Ziegler, M.G. and Coulter, J.D. (1984) Origins and terminations of descending noradrenergic projections to the spinal cord of monkey. Brain Res., 292: 1-16, Williams, J.T., North, R.A., Shefner, S.A., Nishi, S . and Egan, T.M. (1984) Membrane properties of rat locus coeruleus neurons. Neuroscience, 13: 137-156. Wilson, V.J. and Melvill Jones, G. (1979) Mammalian Vestibular Physiology. Plenum Press, New York, pp. XI-365.
C.D. Barnes and 0. Pompeiano (Eds.) P r o ~ r e111. ~Rruin Research, Vol. 88 0 1991 Elsrvier Science Publishcrs B.V.
435 CHAPTER 32
Locus coeruleus and dorsal pontine reticular influences on the gain of vestibulospinal reflexes 0. Pompeiano, E. Horn and P. d’Ascanio Department of Physiology and Biochemistry, Unluersit),of Pisa, Via S. Zeno, Pisa, Italy
Experimental anatomical and physiological studies have shown that noradrenergic locus coeruleus (LC) neurons, which are NE-sensitive due to inhibitory adrenoceptors, send inhibitory afferents to neurons of the peri-LCa and the adjacent dorsal pontine reticular formation (pRF); on the other hand these tegmental neurons, which are, in part at least, cholinergic as well as cholinoceptive, send excitatory afferents to the medullary inhibitory reticulospinal (RS) system. Experiments performed in precollicular decerebrate cats indicate that these pontine structures exert a regulatory influence on posture as well as on the gain of vestibulospinal (VS) reflexes. In particular, the increased discharge of dorsal pontine reticular neurons, and the related inhibitory RS neurons induced by microinjection of cholinergic agonists into the peri-LCa and the adjacent p R F of one side, decreased the postural activity, but greatly increased the response gain of the ipsilateral triceps brachii in response to stimulation of labyrinth receptors resulting from roll tilt of the animal (at 0.15 Hz, +1W). Similar results were also obtainea when the discharge of these pontine and medullary reticular neurons was raised, either by local injection into the peri-LCa and the dorsal pRF of the P-adrenergic antagonist propranolol, which blocked the inhibitory influence of the noradrenergic LC neurons on these structures, or by local injection into the LC complex of an a2- or P-adrenergic agonist (clonidine or isoproterenol) which led to functional inactivation of the noradrenergic neurons; in the latter case the effects were
bilateral. Just the opposite results were obtained after microinjection into the LC of a cholinergic agonist, leading to activation of the corresponding neurons. Evidence was also presented indicating that the cholinergic excitatory afferents to the LC originated from the ipsilateral dorsal pRF. The effects described above were dose-dependent and site-specific, as shown by histological controls. Under given conditions, the decrease in postural activity induced either by direct activation of presumptive cholinergic and cholinoceptive p R F neurons or by inactivation of noradrenergic and NE-sensitive LC neurons was followed by transient episodes of postural atonia which lasted several minutes and affected the ipsilateral and sometimes also the contralatera1 limbs. In these instances, the EMG modulation of the corresponding triceps brachii to animal tilt was suppressed. These findings suggest two different ranges of operation for the noradrenergic and cholinergic structures located in the dorsolateral pontine tegmentum, leading either to a decrease or to an increase in gain of the VS reflexes. The cellular basis of these gain changes is discussed. In conclusion, the pontine structures described above operate as a oariable gain regulator acting at the motoneuronal level during the VS reflexes. Since the same structures are also responsible for the spontaneous fluctuations in posture related to the sleep-waking cycle, they may well intervene as a control system in order to adapt to the animal state the response gain of limb extensors to labyrinth stimulation.
Key words: locus coeruleus, dorsal pontine tegmentum, noradrenergic and cholinergic systems, regulation, vestibulospinal reflexes
Introduction
Experiments performed in cats have shown that the norepinephrine (NE)-containing neurons lo-
cated in the locus coeruleus (LC) complex, which includes both the dorsal (LCd) and ventral (LCa) portions of this structure as well as the locus subcoeruleus (cf. Sakai, this volume), exert a fa-
436
cilitatory influence on posture. In fact, stimulation of this nuclear region activates ipsilateral limb extensor (and flexor) motoneurons, an effect which is due, in part to direct coerulospinal (CS) excitatory influences on spinal motoneurons (cf. Fung et al., this volume), in part to inhibition of inhibitory interneurons like the Renshaw cells acting on these motoneurons (Fung et al., 1988). Moreover, iontophoretic application of N E not only depolarizes the a-motoneurons (cf. White et al., this volume), but also inhibits the Renshaw cell activity (Biscoe and Curtis, 1966; Engberg and Ryall, 1966; Weight and Salmoiraghi, 19661, these opposite effects being probably mediated by different types of adrenoceptors. In addition to the direct CS projection, the LC complex may influence posture by utilizing indirect projections. In particular, their neurons, which are not only noradrenergic but also NEsensitive (cf. Foote et al., 1983; Ennis and AstonJones, 19861, send efferent projections to dorsal pontine structures, such as the peri-LCa and the adjacent dorsal pontine reticular formation (pRF) (Sakai et al., 1977; Jones and Yang, 1985; cf. also Jones and Friedman, 19831, on which they exert a tonic inhibitory influence (see for ref. Pompeiano, 1980, Sakai, 1980, 1988; Foote et al., 1983; Hobson and Steriade, 1986). On the other hand, these tegmental structures which contain not only cholinergic, ie., choline acetyltransferase-immunoreactive neurons (Mesulam et al., 1984; Sakai et al., 1986; Jones and Beaudet, 1987; Vincent and Reiner, 1987; Jones, 1990; Shiromani et al., 1990; cf. Jones and Sakai, this volume), but also cholinoceptive neurons (see below), exert an excitatory influence on the inhibitory areas of the medullary reticular formation (mRF) (Magoun and Rhines, 19471, thus leading to suppression of posture (see for ref. Pompeiano, 1980; Sakai, 1980, 1988; Foote et al., 1983; Hobson and Steriade, 1986). This effect is apparently mediated by a descending tegmento-reticular projection ending on two distinct regions of the medial medulla, the nucleus magnocellularis of the rostral medulla (Sakai et al., 1979; Jones and Yang,
1985; Luppi et al., 1988; cf. Jones and Sakai, this volume) and the nucleus paramedianus of the caudal medulla (Lai and Siegel, 1988; Rye et al., 1988; Shiromani et al., 1990; cf. Jones and Sakai, this volume). These are actually the areas which have massive reticulospinal (RS) projections (cf. Ohta et al., 1988) and produce IPSPs in cervical (Peterson et al., 1979) and lumbar (Lliniis and Terzuolo, 1964; Jankowska et al., 1968) motoneurons. Moreover, there is evidence that the nucleus paramedianus is made up of cholino-sensitive neurons (Lai and Siegel, 1988) thus receiving an excitatory input from cholinergic pontine neurons (Rye et al., 1988; Shiromani et al., 1990), while the nucleus magnocellularis contains glutamate-sensitive neurons, receiving an excitatory input from glutamatergic neurons (Lai and Siegel, 1988). Although the pontine atonia-generating area is sensitive to both acetylcholine and glutamate (Lai and Siegel, 1988), it is still unknown whether the corresponding neurons are responsive to both substances or consist of two distinct, but anatomically intermingled, groups of neurons, one group being responsive to acetylcholine and the other to glutamate. The presumptive cholinergic and cholinoceptive neurons located in the peri-LCa and the adjacent dorsal pRF play an executive role in the induction and maintenance of desynchronized sleep (DS), which is known to be characterized by a complete suppression of posture (cf. Sakai, 1980, 1985a,b, 1988, this volume; Vertes, 1984). The same structures also contribute to the maintenance of postural activity in decerebrate animals (cf. Pompeiano, 1980, 1985; Lai and Siegel, 1988). In fact, lesion of the pontine structures led to a markep reduction, or a suppression, not only of the DS episodes in intact animals (Jouvet and Delorme, 1965; Henley and Morrison, 1974; Sakai, 1985a; Webster and Jones, 1988), but also of the cataplectic episodes induced by systemic injection of an anticholinesterase in decerebrate cats (d’Ascanio et al., 1985b; cf. Pompeiano, 1985). On the other hand, microinjection of the cholinergic agonist carbachol in that area induced
438
asymmetric responses in limb extensors (Lindsay et al., 1976; Schor and Miller, 1981; Manzoni et al., 1983a). In particular, slow, side-down rotation of the head, or of the animal, produced contraction, whereas side-up rotation resulted in relaxation of ipsilateral forelimb extensors, as measured in the medial head of the triceps brachii (a-responses). The peak of the responses was closely related to the extreme head or animal displacement attributable to stimulation of position-sensitive macular (gravity) receptors. Responses of the same pattern but smaller in amplitude could also affect ipsilateral hindlimb extensors as measured in the triceps surae (Boyle and Pompeiano, 1984; Manzoni et al., 1984; d'Ascanio et al., 1985a; Pompeiano et al., 1985a). In order to understand the role that both the excitatory VS neurons, as well as the medullary inhibitory RS neurons, exert during VS reflexes, the discharge of neurons originating from the lateral vestibular nucleus (LVN) (Boyle and Pompeiano, 1980; Schor and Miller, 1982; Marchand et af., 1987) as well as the inhibitory area of the mRF (Manzoni et al., 1983~)was recorded in precollicular decerebrate cats during roll tilt of the animal at the frequency of 0.026 Hz, f lo". We will summarize here only the results of experiments in which the recorded units were found to project to the lumbosacral segments of the spinal cord. Marchand et al. (1987) found that among the 129 VS neurons located in the LVN and antidromically activated by stimulation of the spinal cord at T12-L1 (mean conduction velocity, 90.0 f 21.5, S.D. m/sec), 76 units (58.9%) showed a periodic modulation of their firing rate during animal tilt. In these instances the average gain of the first harmonic of the unit responses, expressed in absolute change of the firing rate per degree of displacement, corresponded to 0.47 ? 0.44, S.D. imp./sec/deg, while the average sensitivity expressed in percentage change of the firing rate per degree of displacement corresponded to 3.24 f 3.15, S.D.%/deg. Most of the units re-
sponded to the extreme animal displacement (Fig. 2A). In particular, 51 units (67.1%) were excited during side-down tilt of the animal (a-responses) while 15 units (19.7%) by side-up tilt (@-responses). Both populations of neurons fired with an average phase lead of the first harmonic of responses of +21.0 f 27.2, S.D. deg with respect to the extreme animal displacements. The remaining 10 units (13.2%) showed an intermediate phase angle of the responses. It should be noted that the gain of the VS units that showed a-responses (0.57 -t 0.49, S.D. imp./sec/deg) was on an average higher than that of the units displaying @-responses(0.26 f 0.19, S.D. imp./sec/deg). The phase angle of responses of the VS neurons to animal tilt either remained unmodified or showed a decrease in phase lead by increasing frequency of tilt from 0.008 to 0.32 Hz (at f lo"), thus indicating that the responses were due to stimulation of macular receptors; in some instances, however, the phase lead of the responses increased as if the units received a convergent input from the canal receptors (cf. Manzoni et af., 1987; Kasper et al., 1988). In addition to these findings, Manzoni et af. (1983~)observed that a large proportion of neurons (64/93, i.e., 68.8%) located in the inhibitory area of the mRF and antidromically activated by stimulation of the spinal cord at T12-L1 (mean conduction velocity, 68.5 & 27.6, S.D. m/sec) responded to animal tilt. The average gain and sensitivity of the unit responses corresponded to 0.30 f 0.32, S.D. imp./sec/deg and 3.66 f 3.64, S.D.%/deg, respectively. Most of the units responded to the extreme animal displacements (Fig. 2B). However, 37 units (57.8%) were excited during side-up tilt of the animal (@-responses) while 21 units (32.8%) were excited by side-down tilt (a-responses). Both populations of neurons fired with an average phase lead of the first harmonic of responses of 25.7 f 24.4, S.D. deg with respect to the extreme animal displacements. The remaining 6 units (9.4%) showed an intermediate phase angle of the responses. Finally, the gain of the RS units that showed @-re-
+
438
asymmetric responses in limb extensors (Lindsay et al., 1976; Schor and Miller, 1981; Manzoni et al., 1983a). In particular, slow, side-down rotation of the head, or of the animal, produced contraction, whereas side-up rotation resulted in relaxation of ipsilateral forelimb extensors, as measured in the medial head of the triceps brachii (a-responses). The peak of the responses was closely related to the extreme head or animal displacement attributable to stimulation of position-sensitive macular (gravity) receptors. Responses of the same pattern but smaller in amplitude could also affect ipsilateral hindlimb extensors as measured in the triceps surae (Boyle and Pompeiano, 1984; Manzoni et al., 1984; d'Ascanio et al., 1985a; Pompeiano et al., 1985a). In order to understand the role that both the excitatory VS neurons, as well as the medullary inhibitory RS neurons, exert during VS reflexes, the discharge of neurons originating from the lateral vestibular nucleus (LVN) (Boyle and Pompeiano, 1980; Schor and Miller, 1982; Marchand et af., 1987) as well as the inhibitory area of the mRF (Manzoni et al., 1983~)was recorded in precollicular decerebrate cats during roll tilt of the animal at the frequency of 0.026 Hz, f lo". We will summarize here only the results of experiments in which the recorded units were found to project to the lumbosacral segments of the spinal cord. Marchand et al. (1987) found that among the 129 VS neurons located in the LVN and antidromically activated by stimulation of the spinal cord at T12-L1 (mean conduction velocity, 90.0 f 21.5, S.D. m/sec), 76 units (58.9%) showed a periodic modulation of their firing rate during animal tilt. In these instances the average gain of the first harmonic of the unit responses, expressed in absolute change of the firing rate per degree of displacement, corresponded to 0.47 ? 0.44, S.D. imp./sec/deg, while the average sensitivity expressed in percentage change of the firing rate per degree of displacement corresponded to 3.24 f 3.15, S.D.%/deg. Most of the units re-
sponded to the extreme animal displacement (Fig. 2A). In particular, 51 units (67.1%) were excited during side-down tilt of the animal (a-responses) while 15 units (19.7%) by side-up tilt (@-responses). Both populations of neurons fired with an average phase lead of the first harmonic of responses of +21.0 f 27.2, S.D. deg with respect to the extreme animal displacements. The remaining 10 units (13.2%) showed an intermediate phase angle of the responses. It should be noted that the gain of the VS units that showed a-responses (0.57 -t 0.49, S.D. imp./sec/deg) was on an average higher than that of the units displaying @-responses(0.26 f 0.19, S.D. imp./sec/deg). The phase angle of responses of the VS neurons to animal tilt either remained unmodified or showed a decrease in phase lead by increasing frequency of tilt from 0.008 to 0.32 Hz (at f lo"), thus indicating that the responses were due to stimulation of macular receptors; in some instances, however, the phase lead of the responses increased as if the units received a convergent input from the canal receptors (cf. Manzoni et af., 1987; Kasper et al., 1988). In addition to these findings, Manzoni et af. (1983~)observed that a large proportion of neurons (64/93, i.e., 68.8%) located in the inhibitory area of the mRF and antidromically activated by stimulation of the spinal cord at T12-L1 (mean conduction velocity, 68.5 & 27.6, S.D. m/sec) responded to animal tilt. The average gain and sensitivity of the unit responses corresponded to 0.30 f 0.32, S.D. imp./sec/deg and 3.66 f 3.64, S.D.%/deg, respectively. Most of the units responded to the extreme animal displacements (Fig. 2B). However, 37 units (57.8%) were excited during side-up tilt of the animal (@-responses) while 21 units (32.8%) were excited by side-down tilt (a-responses). Both populations of neurons fired with an average phase lead of the first harmonic of responses of 25.7 f 24.4, S.D. deg with respect to the extreme animal displacements. The remaining 6 units (9.4%) showed an intermediate phase angle of the responses. Finally, the gain of the RS units that showed @-re-
+
439
sponses (0.37 f 0.39, S.D. imp./sec/deg) was on an average higher than that of the units displaying a-responses (0.19 0.17, S.D. imp./sec/deg). The phase angle of response of the RS neurons to animal tilt, which was closely related to the extreme animal position at low frequency of tilt, due to stimulation of macular receptors, showed an increase in phase lead by increasing
*
A
M A C U L A R INPUT .VESTIBULOSPINAL NEURONS
N 76
frequency of tilt from 0.008 to 0.32 Hz (at f lW), thus indicating that the responses received a contribution from the canal receptors. The observation that most of the RS neurons projecting to the lumbosacral segments of the spinal cord were excited during side-up tilt @responses) differs from that obtained from the VS neurons projecting to the same segments of the spinal cord, the majority of which were excited during side-down tilt (a-responses) (Marchand et al., 1987). This predominant pattern of response corresponds to that obtained at the level of the utricular afferents, thus in apparent agreement with the morphological polarization .of the majority of the corresponding receptors. It appears, therefore, that the macular input of one side, responsible for the a-responses of lateral VS neurons to tilt, is transmitted via a crossed pathway to the ventral aspect of the contralateral mRF, thus being responsible for the /3-responses of the corresponding RS neurons. One of the main channels transmitting the macular input of one side to the contralateral mRF is the lateral VS tract, acting on neurons of the crossed spinoreticular pathway (cf. Pompeiano, 1979).
Fig. 2. Polar diagrams showing the gain and the phase angle of the first harmonic of responses of VS (A) and medullary RS (B) to sinusoidal roll tilt of the animal at 0.026 Hz, T 10". All the neurons were activated antidromically by stimulation of the spinal cord at T12 and LI, thus projecting to the lumbosacral segments of the spinal cord. The response gain of each unit is indicated by the distance of the corresponding dot from the center of the diagram (see the scale along the vertical meridian); 6 units in A and 2 units in B had a gain higher than 1.0 imp./sec/deg. The relative position of the dot with respect to 0" meridian indicates in degrees the phase lead (positive values) or the phase lag (negative values) of responses with respect to the extreme side-down position of the animal. T h e dashed lines outline the standard deviation of the phase angle of response of the two main populations of VS and RS units that showed a positional sensitivity. In particular, the mean phase angle of the VS units excited during side-down animal tilt (A) corresponded to + 25.0+ 23.8, S.D., deg ( n = 51) while that of the RS units excited during side-up animal tilt (B) corresponded to - 156.6k24.8, S.D., deg ( n = 37). (From Marchand et al. (1987) and Manzoni et al. (1983~1.)
440
In order to understand the role that the VS and the RS neurons exert during VS reflexes, we should consider that the VS neurons exert a direct excitatory influence on ipsilateral limb extensor motoneurons (Lund and Pompeiano, 1968; cf. Pompeiano, 1975) and then, through their recurrent collaterals, on the related Renshaw cells. On the other hand, the medullary reticular area, from which the activity of the RS neurons was recorded, exerts an inhibitory influence on ipsilateral limb extensor (and flexor) motoneurons (Llinhs and Terzuolo, 1964; Jankowska et al., 1968; Peterson et al., 19791, an effect which has been attributed, in part at least, to direct excitation of the corresponding Renshaw cells (cf. Pompeiano, 1984). Since the postural adjustments induced by labyrinth stimulation are characterized by an increased contraction of ipsilateral limb extensors during side-down tilt of the animal, we hypothesized that for this direction of animal orientation the motoneurons innervating the ipsilateral limb extensors are not only excited by an increased discharge of VS neurons, but also disinhibited by a reduced discharge of RS neurons, leading to disfacilitation of the corresponding Renshaw cells (cf. Pompeiano et al., 1985b). Just the opposite result would occur during side-up tilt. In these instances we should expect that for a given amount of labyrinth input, the amplitude of the EMG modulation of limb extensors depends on the background discharge of the medullary inhibitory RS neurons. In precollicular decerebrate cats, in which the discharge of the LC neurons (Pompeiano and Hoshino, 1976a,b) tonically inhibits the activity of peri-LCa and the adjacent dorsal pRF neurons, as well as of the related medullary inhibitory RS neurons (Hoshino and Pompeiano, 1976; Srivastava et al., 1982), thus giving rise to postural rigidity, a small amount of disinhibition would affect the limb extensor motoneurons during side-down animal tilt. These motoneurons would then respond weakly to the excitatory VS volleys elicited by given parameters of animal displacement, giving rise to a small-am-
plitude modulation of the EMG activity and, thus, to a small gain of responses of limb extensors to labyrinth stimulation. Indeed, in spite of the good physiological conditions of the animals and the detectable modulation of a large proportion of VS and RS neurons to tilt, the gain of the VS reflexes was quite small in forelimb extensors (Manzoni et al., 1983a) and almost negligible or absent in hindlimb extensors (Boyle and Pompeiano, 1984; Manzoni et al., 1984; d’Ascanio et al., 1985a; Pompeiano et al., 1985a). However, if the background discharge of the medullary inhibitory RS neurons increases, due either to direct activation of the peri-LCa and dorsal pRF neurons or to inactivation of the LC leading to disinhibition of the underlying tegmental structures, their activity would lead either to reduction or suppression of the ipsilateral postural activity in the animal at rest. In the first instance the increased discharge of the VS neurons during side-down tilt would be associated with a great reduction in firing rate of the medullary inhibitory RS neurons for this direction of animal orientation. This would lead to a prominent disinhibition of the ipsilateral limb extensor motoneurons, thus increasing the response gain of the corresponding limb extensors to labyrinth stimulation. In the second instance, however, the increased discharge of the peri-LCa and the dorsal pRF neurons, as well as of the related medullary inhibitory RS neurons, would be so high as to suppress not only the postural activity of the ipsilateral limbs in the animal at rest, but also the responses of the limb extensors to animal tilt, due to prominent postsynaptic inhibition of the corresponding motoneurons. In the present report we will summarize the results of experiments using a “chemical microstimulation” technique to analyze the role played by the pontine neuronal circuits in the static, as well as in the dynamic, control of posture induced by labyrinth stimulation. In particular, both the postural activity as well as the response gain of a forelimb extensor to animal tilt were evaluated before and after local injection into the LC com-
441
plex and the adjacent tegmental structures of appropriate neurotransmitter agonists or antagonists, which led either to direct activation of presumptive cholinergic and cholinoceptive pontine reticular neurons or to inactivation of noradrenergic and NE-sensitive LC neurons, thus releasing from inhibition the activity of the dorsal pontine reticular neurons and the related medullary RS neurons. The experiments were performed in precollicular decerebrate cats. The multiunit EMG activity of the medial head of the triceps brachii was sampled by a digital signal averager, during roll tilt of the animal at 0.15 Hz, flo", leading to sinusoidal stimulation of labyrinth receptors. A computer analysis of the responses allowed evaluation of the base frequency (mean firing rate of the multiunit EMG activity, recorded during each test, in imp./sec), the gain (imp./ sec/ deg) and the phase angle (deg) of the first harmonic of the output with respect to the peak of the side-down stimulus displacement ipsilateral to the recording side (Manzoni et al., 1983a). Groups of six averaged EMG responses were recorded at regular intervals of a few minutes for several hours before and after microinjection of various transmitter agonists or antagonists either into the periLCa and the dorsal pRF (at the stereotaxic coordinates of P2.5 or 3, L2.5 or 2.8, H-2.5 or -3.5) or into the LC complex (at P2 or 3, L2.5 or 2.8, H-1.5 or -2.5). Usually, 0.10-0.25 pl of solutions of active substances at different concentrations in sterile saline (pH 7.4 f 0.21, marked with 5% pontamine sky blue, were injected into dorsal pontine structures of one side by using a thin stainless steel cannula (0.5 mm diameter). At the end of the experiments, the tip of the cannula used for each penetration, as well as the extent of the nerve tissue stained with the blue dye, were identified on serial sections of the brainstem, counterstained with neutral red (for the identification and abbreviations of different brain structures illustrated in the schematic drawings of Figures 3-7, see Berman, 1968).
Noradrenergic and cholinergic pontine mechanisms leading to an increase in gain of uestibulospinal reflexes Effects of local injection of cholinergic agonists into the peri-LCa and the adjacent dorsal pontine reticular structures ( p W ) Preliminary experiments performed in decerebrate cats have shown that systemic injection of the anticholinesterase eserine sulfate, at a dose of 0.05-0.075 mg/kg, iv, which reduced postural activity, increased the gain of the multiunit EMG responses of the triceps brachii (Pompeiano et al., 1983) and the triceps surae (Pompeiano et al., 1985b) to animal tilt. However, no significant changes in pattern and phase angle of the responses were observed. The increase in response gain induced by the anticholinesterase started 5-10 min after injection, reached the highest value in about one hour and then slowly declined. The effects described above were attributed to a central action of the anticholinesterase on the dorsal pRF, since electrolytic lesion limited to this region not only decreased the response gain of the triceps brachii to labyrinth stimulation but also prevented this gain from increasing after systemic injection of small doses of eserine sulfate (d'Ascanio et al., 1985b; cf. Pompeiano, 1985). The results of these experiments were extended by Barnes et al. (1987), who showed that in decerebrate cats microinjection into the dorsal part of the pRF of one side, of 0.1-0.2 pl of the cholinergic agonist carbachol (at 0.01-0.2 p g / p l of saline), produced a postural asymmetry characterized by a decrease of the extensor tonus of the ipsilateral limbs, while the decerebrate rigidity either remained unmodified or slightly decreased in the contralateral limbs. In the same experiments the amplitude of modulation, and thus the gain of the multiunit EMG responses of the ipsilateral triceps brachii to animal tilt at 0.15 Hz, f lo", increased. This finding was associated with a slight decrease in phase lead of the responses.
442
Imp./rec
Imp./*ec
'1 50
roo, 25
SIDE DOWN
SIDE UP
SIDE UP I
2 sec
. 2 sec
Fig. 3. Increase in the response gain of the triceps brachii of one side to animal tilt after local injection of the muscarinic agonist bethanechol into the ipsilateral dorsal pRF. Precollicular decerebrate cat (Exp. 12). Schematic representation of a transverse section of the pons, in which the dot indicates the position of the tip of the cannula used for the injection (at P3, L2.8, H-3) of 0.25 p1 of bethanechol solution at 0.1 pg/pl of saline (1). The site was located within the dorsal pRF of the left side (right side of the figure). Below the scheme there are sequential pulse density histograms (SPDHs) showing the multiunit EMG responses of the left triceps brachii to animal tilt at 0.15 Hz, f 10" (average of 6 sweeps); the lower records indicate the animal displacement. Upper traces. Control records showing an average base frequency (BF) of 59.3 imp./sec, a response gain of the first harmonic of 0.42 imp./sec/deg and a phase lag of -3.5" with respect to the peak of the ipsilateral side-down tilt. Lower traces. Responses recorded 1 and 3 min after local injection of bethanechol into the ipsilateral dorsal pRF showing an average BF of 57.0 imp./sec, a response gain of 1.93 imp./sec/deg, and a phase lag of -32.7". (From Barnes et al. (1987).)
However, no changes in the dynamic characteristics of the responses were found in the contralatera1 triceps brachii. The changes in posture as well as in response gain produced by carbachol injection appeared suddenly and persisted for about 3 h after the injection before disappearing. The effects of carbachol, which acts on both muscarinic and nicotinic receptors, were also reproduced in other experiments by local injection into the pRF of an
equal dose of bethanechol, a pure muscarinic agonist (Fig. 3). Moreover, injection into the same structure of 0.25 ~1 of a solution of the muscarinic blocker atropine sulfate (at 6 p g /p l of saline) not only produced a recovery of the postural activity in the ipsilateral limbs, but also suppressed the increased gain as well as the slight changes in phase angle of the responses produced by the cholinergic substances, returning them to the control values.
443
The effects described above were not due to mechanical stimulation of the pontine neurons following injection of the fluid, since neither changes in posture nor in gain of the VS reflexes were observed following administration of an equal volume of saline prior to the injection of the active drugs. Moreover, the effects were dose-dependent and site-specific; in fact, the cholinoceptive neurons were located in the periLCa and the adjacent dorsal pontine reticular structures (Fig. 3, upper scheme). This area corresponds to the region where presumptive cholinergic and cholinoceptive neurons are located and give rise to a tegmento-reticular projection acting on the medullary inhibitory RS neurons (see Introduction). It appears also that the direct excitatory action of the cholinergic agonists on noradrenergic neurons located in the peri-LCa and adjacent tegmental structures is totally overcome when cholinergic agonists are microinjected into these structures. In order to understand the influence that the cholinergic system exerts on the static and dynamic control of posture, we should consider that cholinergic agents increase the background discharge of the dorsal pRF neurons and the medullary inhibitory RS neurons, as shown in experiments of unit recording performed after systemic injection of the anticholinesterase (Hoshino and Pompeiano, 1976; Srivastava et al., 1982). This finding would first lead to a tonic inhibitory influence on ipsilateral limb extensor motoneurons (cf. Morales et al., 1987), thus decreasing the extensor rigidity (Matsuzaki et al., 1968; cf. Jouvet, 1972; Pompeiano, 1976, 1980). Moreover, for a given labyrinth signal, the higher the firing rate of these RS neurons in the animal at rest, the greater the disinhibition affecting the extensor motoneurons during side-down animal tilt. These motoneurons would then respond more efficiently to the same excitatory VS volleys elicited by given parameters of stimulation, resulting in an increased gain of the EMG responses of limb extensors to labyrinth stimulation. The opposite result would occur following
inactivation of the dorsal pontine tegmental region by the muscarinic blocker atropine sulfate, decreasing the resting discharge of the inhibitory RS neurons (Barnes et al., 1987). In conclusion, it appears that local activation of neurons in the peri-LCa and the dorsal pontine reticular formation, which are presumably cholinergic but also cholinoceptive due to the existence of muscarinic receptors, decreased the postural activity in the ipsilateral limbs but greatly enhanced the amplitude of the EMG modulation of ipsilateral limb extensors to labyrinth stimulation. These pontine reticular neurons thus exert a critical role in the gain regulation of VS reflexes. Effects of local injection of the P-adrenergic antagonist propranolol into the peri-LCa and the adjacent dorsal pontine reticular structures (pRF) The demonstration that unilateral injection of cholinergic agonists into the pRF, leading to activation of presumptive cholinergic and cholinoceptive neurons, not only reduced the postural activity in the ipsilateral limb extensors but also increased the gain of the EMG responses of the corresponding triceps brachii to labyrinth stimulation, led us to investigate whether similar results could also be elicited by inactivation of the noradrenergic afferents, which exert an inhibitory influence on the dorsal pRF neurons. Experiments reported by using binding methods (Pompeiano, M. et al., 1989) as well as histofluorescent (Hess, 1979) or autoradiographic methods (Rainbow et al., 1984), have demonstrated the existence of P-adrenoceptors in the dorsal pontine tegmentum, where reticular structures are located. The possibility that p-adrenoceptors located on dorsal p R F neurons are involved in the noradrenergic-mediated control of posture and VS reflexes was tested by d’Ascanio et al. (1989b), who studied in decerebrate cats the effects of local injection, into the dorsal pRF, of the P-adrenergic antagonist propranolol on posture as well as on the response gain of the triceps brachii to labyrinth stimulation. Microinjection into the p R F of one side of
444
4.0-
3.0-
3.0.
8 -
\
f a
.
.
. -
2.0.
. ..
1.0.
I;,
0
-100
I I 0 -50
0,
,
+50
+lo0 ,
+150min ,
I:., -100
-50
.*
t
..
I 0,
+50 ;
+I00 ~,
+%Omin
Fig. 4. Effects of local injection of the non-selective P-adrenergic antagonist propranolol into the dorsal pRF of one side o n the response gain of the ipsilateral and the contralateral triceps brachii to animal tilt. Precollicular decerebrate cat (Exp. 1). The gain values of the averaged multiunit responses of the triceps brachii of both sides to animal tilt at 0.15 Hz, 10" were evaluated at different time intervals before and after local injection (at P2.5, R2.8, H-3.5) of 0.25 p1 of propranolol solution a t 4.5 p g / p l of saline stained with 5% pontamine (I). The site was located within the pRF of the right side (left side of the inset, hatched area). Each dot represents the mean value of 6 averaged responses. The arrow indicates the time of injection (0 m i d . The response gain of the right (IPSI) triceps brachii increased after injection of propranolol into the ipsilateral pRF, while the response gain of the left (CONTRA) triceps brachii slightly decreased. (From d'Ascanio et al. (1989bI.I
+
0.25 pl of the non-selective P-antagonist propranolol (at 4.5-9 pg/pl of saline), which acts on both PI- and P,-adrenoceptors, produced a postural asymmetry, characterized by a prominent decrease of the extensor tonus in the ipsilateral limbs, while the decerebrate rigidity either increased or slightly decreased in the contralateral limbs. Most interestingly, the gain of the multiunit EMG responses of the ipsilateral triceps brachii to animal tilt (at 0.15 Hz, &lo") increased, while that of the contralateral muscle slightly decreased (Fig. 4). There was only a slight decrease in the phase lead of the responses on the ipsilateral side, which were always characterized by an increased EMG activity during sidedown tilt (a-responses), while on the contralatera1 side the response pattern reversed in some instances, as shown by an increased EMG activity during side-up tilt (p-responses).
The changes in posture as well as in the VS reflexes were first observed 10-20 min after the propranolol injection and reached the maximum values in about 1-2 h; they were then followed for more than 3 h after the injection. Histological controls showed that the effective region was located in the dorsal pontine tegmentum, immediately ventral to the LC (Fig. 4, upper scheme). This area actually corresponds to the peri-LCa and the adjacent dorsal pontine reticular structures, from which a tegmento-reticular projection ending on the medullary inhibitory area originates (see Introduction). Moreover, control experiments indicated that the gain of the VS reflexes was not influenced by injecting an equal volume of saline within the same structures. We postulated that the P-antagonist propranolol blocks the inhibitory influence exerted by the noradrenergic LC neurons on the ipsilateral
445
dorsal pontine structures, where presumptive cholinergic and self-excitatory cholinoceptive neurons are located. The increased discharge of these pontine neurons and the related medullary inhibitory RS neurons would lead to a decreased postural activity in the ipsilateral limbs. Moreover, for a given labyrinth signal, the higher the firing rate of these inhibitory RS neurons in the animal at rest, the greater the disinhibition which affects the ipsilateral limb extensor motoneurons during side-down tilt. These motoneurons would then respond more efficiently to the same excitatory volleys elicited by given parameters of stimulation, thus leading to an increased gain of the EMG responses of forelimb extensor muscles to labyrinth stimulation. In contrast to these ipsilateral effects, injection of the P-antagonist into the dorsal pRF of one side decreased the gain of the EMG responses and, in some instances, reversed the response pattern of the contralateral extensor muscle to labyrinth stimulation. These crossed effects, which were not observed after injection of cholinergic agonists (see above), can be attributed to the existence of reciprocal mechanisms by which the NE-sensitive pRF neurons of one side act on pontine structures of the contralateral side (cf. d'Ascanio et al., 1989b, for details). In conclusion, it appears that the prominent changes in posture, as well as in gain of the VS reflexes which occurred after local injection of the non-selective P-adrenergic antagonist propranolol into the peri-LC a and the adjacent dorsal pRF were due to suppression of the inhibitory influences that the noradrenergic LC neurons exert on dorsal pontine tegmental neurons and the related medullary inhibitory RS system of the injected side, by acting through P-adrenoceptors. Effects of local injection of the a,-adrenergic agonist clonidine into the locus coeruleus complex Since the pontine reticular neurons and the related medullary inhibitory RS system are under the tonic inhibitory control of noradrenergic fibers, experiments were designed to find out
whether inactivation of the NE-containing neurons located in the LC complex, which apparently contribute to these afferents, could modify the postural activity as well as the response gain of limb extensor muscles to labyrinth stimulation, as shown by direct activation of the cholinergic pontine system. It is known that iontophoretic application or local microinjection into the LC of NE or epinephrine (Cedarbaum and Aghajanian, 1976, 19771, as well as of the a,-adrenergic agonist clonidine (Svensson et al., 1975; Guyenet, 1980; Abercrombie and Jacobs, 1987b) inhibits the activity of the corresponding NE-containing neurons. Moreover, immunohistochemical observations, as well as in vitro autoradiography have shown the existence of catecholaminergic terminals (Hokfelt et al., 1974; Pickel et aL, 1977; Groves and Wilson, 1980; Grzanna and Molliver, 1980; Jones and Friedman, 1983) as well as of a,-adrenergic receptors within the LC (Young and Kuhar, 1979, 1980; Unnerstall et al., 1984; cf. also Palacios and Wamsley, 1984). These and other data have led to the suggestion that the activity of LC cells is dampened by a self-inhibitory circuit, represented by recurrent collaterals from the LC axons themselves (Cedarbaum and Aghajanian, 1976, 1977; Aghajanian et al., 1977). In addition, NE contained in proximal dendrites and somata of the LC-complex neurons can be released as a consequence of impulse activity (Koda et al., 1980) and, due to dendro-dendritic and dendro-somatic contacts between neighbouring neurons (Shimizu et al., 1979; Groves and Wilson, 19801, lead to lateral suppression of LC cells activity (cf. Foote et al., 1983). Experiments performed by Pompeiano et al. (1987) have shown that microinjection of 0.25 pl of clonidine (at 0.012-0.15 p g / p l of saline) into the LC complex of one side not only decreased decerebrate rigidity in the ipsilateral limbs, but also increased the amplitude of modulation and thus the response gain of the multiunit EMG responses of the ipsilateral triceps brachii to animal tilt (at 0.15 Hz, f 10'). Moreover, a slight
446
Exp. 25
' 10
SIDE UP
-
SIDE UP
2 sec
2 sec
Fig. 5. Effects of local injection of the a*-adrenergic agonist clonidine into the LC of one side on the response gain of the ipsilateral and the contralateral triceps brachii to animal tilt. Precollicular decerebrate cat (Exp. 25). Schematic representation of a transverse section of the pons, in which the shaded area indicates the site of injection (at P3, L2.8, H-2) of 0.25 pl of clonidine solution at 0.012 p g / p l of saline stained with 5% pontamine. This site (I) was located within the LC of the left side (right side of the figure). Below this scheme there are SPDHs showing the multiunit EMG responses of the left (IPSI) and the right (CONTRA) triceps brachii to animal tilt at 0.15 Hz, 10" (average of 6 sweeps); the lower records indicate the animal displacement. Upper traces: control records from the left muscle showing an average BF of 85.8 imp./sec, a response gain of 0.41 imp./sec/deg and a phase lead of +45.2", while that from the right muscle showed a BF of 122.7 imp./sec, a gain of 0.87 imp./sec/deg and a phase lead of 14.9". Lower traces: 6 and 10 min after injection of clonidine into the left LC, the left muscle showed an average BF of 136.1 imp./sec, a response gain of 1.71 imp./sec/deg and a phase lead of +22.7", while the right muscle showed a BF of 127.6 imp./sec, a gain of 2.80 imp./sec/deg and a phase lead of +8.7". (From Pompeiano et a/. (19871.1
+
+
decrease in phase lead of the responses was obtained (Fig. 5, left traces). The same injection also produced either no change or only a slight decrease in postural activity in the contralateral limbs, which was also associated with an increase in the response gain of the triceps brachii of that side to the same parameters of labyrinth stimulation (Fig. 5, right traces). This effect, however, was usually associated with a slight increase in phase lead of the responses. These findings were
first observed 10-15 min after injection of clonidine, reached the highest value in about 30-60 min, and persisted for more than 2 h after the injection. The effects of clonidine were not due to irritative phenomena, since no postural or reflex changes were observed following local injection of saline into the LC complex. On the other hand, the amount of these changes depended on the dose of clonidine injected. Only slight transi-
447
tory changes in systemic blood pressure and heart rate were observed after injection of clonidine at the concentrations reported above. Histological controls demonstrated that in some experiments the injection site was located within the dorsal and ventral part of the LC (Fig. 5, upper scheme). In these instances the results could be attributed to inactivation of LC complex neurons, since: (1) the postural and reflex changes following microinjection of the a,-adrenergic agonist clonidine into the LC complex were similar to those obtained in decerebrate cats after selective electrolytic lesion of the LC of one side (d’Ascanio et al., 1985a); moreover, (2) microinjection of clonidine produced a complete suppression of the spontaneous activity of noradrenergic neurons of the LC in freely moving cats (Abercrombie and Jacobs, 1987b). This effect actually lasted for about 90 min and was followed by a recovery of the spontaneous unit activity which returned to the control levels by 120 min, a period of time comparable to the duration of the postural and reflex changes observed in our decerebrate cats. In other experiments, however, the injection site was located within the LC and the locus subcoeruleus, as well as in the peri-LCa and the adjacent pRF. In these instances, clonidine might have acted on a-adrenoceptors located not only on the cell bodies, but also on the terminals of noradrenergic LC neurons projecting to the periLCa and the surrounding pRF, thus releasing from inhibition the corresponding target neurons. The possibility that clonidine acted on a*adrenoceptors located either presynaptically on cholinergic terminals by inhibiting the release of acetylcholine (Beani et al., 1978; Bianchi et al., 1979; Vizi, 1980; Moroni et aZ., 19831, or postsynaptically on presumptive cholinergic and cholinoceptive pRF neurons by suppressing their activity (Greene et al., 1989; Bier et al., 19901, is made unlikely by the fact that: (1) microinjection of clonidine into the LC region, which inhibited the noradrenergic neurons, had no consistent effect on the activity of neighbouring non-noradrenergic
neurons (Abercrombie and Jacobs, 1987b); (2) the postural and reflex changes induced by local injection of clonidine into the LC complex and the adjacent dorsal p R F closely resembled those elicited in other experiments by local administration of a cholinergic agonist in the pRF (see above); and (3) the resulting changes were greatly reduced or suppressed by local injection into the same pontine region, of 0.25 pl of atropine sulfate (at 6 pg/pl of saline), an effect which was strictly ipsilateral. It should also be mentioned that in decerebrate cats the firing rate of the presumptive cholinergic and cholinoceptive pRF neurons is apparently depressed by a relatively high resting discharge of the LC neurons (Pompeiano and Hoshino, 1976a,b; Hoshino and Pompeiano, 1976; cf. Pompeiano, 19801, so that a local depressive influence, possibly exerted by clonidine on the cholinergic system, could easily be overcome by the prominent disinhibition of a larger population of dorsal pontine neurons, resulting from a divergent projection of the LC neurons to the pontine reticular area. In conclusion, local injection of the a,-adrenergic agonist clonidine into the LC complex of one side as well as in the adjacent pontine tegmental region decreased the postural activity, but greatly increased the gain of the VS reflexes acting on the forelimb extensors of both sides. The decreased postural activity in the ipsilateral, and to a lesser extent also in the contralateral, limbs following clonidine injection can be attributed to inactivation not only of the noradrenergic CS neurons, which exert an excitatory influence on ipsilateral limb extensor motoneurons (cf. Fung et al., this volume), but also of the NE-containing LC neurons, which inhibit the activity of the dorsal pRF neurons and the related medullary inhibitory RS neurons of both sides. The resulting increase in discharge rate of these pontine and medullary reticular neurons led also to an increased gain of the VS reflexes which utilized cholinergic mechanisms. In fact, when the effects were elicited by local spread of the a,-adrenergic agonist from the LC to the dorsal
448
part of the pRF, they were suppressed by administration into the same structures of the muscarinic blocker atropine. It appears, therefore, that the NE-containing LC-complex neurons play a permissive role in the control of posture as well as in the gain regulation of the VS reflexes.
Effects of local injection of the P-adrenergic agonist isoproterenol into the locus coeruleus complex The existence of P-adrenoceptors in the LC complex has been hypothesized by the results of (1) microiontophoretic studies, in which the spontaneous discharge of LC neurons could be inhibited not only by the a,-adrenergic agonist clonidine, but also by the P-adrenergic agonist isoproterenol, this effect being only partially antagonized by local administration of the adrenergic blocker sotalol (Cedarbaum and Aghajanian, 1976, 19771, and (2) light microscopic studies, in which several markers were used to visualize and localize in the LC these adrenoceptors (Atlas and Segal, 1977; Atlas and Melamed, 1978; Hess, 1979; Palacios and Kuhar, 1980, 19821, including P,- and &-subtypes (Rainbow et al., 1984). We decided, therefore, to study whether functional inactivation of the NE-containing LC and subcoerulear neurons, elicited by local injection into the LC complex of minute doses of a p-adrenergic agonist, could modify posture as well as the gain of the VS reflexes elicited by sinusoidal stimulation of labyrinth receptors, just as shown in clonidine experiments (see above). Indeed, experiments performed by d'Ascanio et al. (1989a) have proved that microinjection of 0.25 p l of isoproterenol (at 4.5-9.0 pg/pl of saline) into the LC complex of one side decreased the extensor rigidity in the ipsilateral limbs and to a lesser extent also in the contralateral limbs. The same injection also increased the amplitude of modulation and thus the response gain of the ipsilateral and to a lesser extent also of the contralateral triceps brachii to animal tilt at 0.15 Hz, *lo" (Fig. 6, upper and middle traces). Moreover, slight changes in phase lead of the responses were observed on both sides. These find-
ings appeared within 5-10 min after the injection of isoproterenol, fully developed within 20-30 min, and persisted for about 2-3 h after the injection. The changes in posture, as well as in gain of VS reflexes induced by the P-adrenergic agonist isoproterenol could be suppressed by local injection into the LC complex of 0.25 p l of the Padrenergic antagonist propranolol (at 4.5-9.0 pg/pl of saline) (Fig. 6, lower traces). No significant changes in systemic blood pressure and heart rate were observed following microinjection of isoproterenol at the concentrations reported above. The effects described above were site-specific. Histological controls indicated, in fact, that the effective injections of isoproterenol were located in the region of the LC complex, particularly at the caudal level (Fig. 6, upper scheme). Both the dorsal and ventral part of this structure were involved in these experiments. In conclusion, the postural and reflex changes following local injection of isoproterenol into the LC complex of one side were similar to those induced by clonidine injection. These changes were attributed to partial inactivation not only of the CS neurons, which exert a facilitatory influence on ipsilateral limb extensor motoneurons, but also of the LC neurons which inhibit the dorsal pontine tegmental neurons and the related medullary inhibitory RS neurons of both sides. The resulting increase in discharge rate of both these pontine and medullary reticular neurons would then be responsible not only for the decrease in postural activity but also for the increased gain of the VS reflexes acting on the forelimb extensors of both sides.
Noradrenergic and cholinergic pontine mechanisms leading to a decrease in gain of uestibulospinal reflexes The results described in the previous section were obtained in decerebrate cats after injection of small doses: (1) of a cholinergic agonist or a
449
Exp 4
Imp./sec
501
IPS1
CONTRA
hp./sec
sol
100
1
50
5
Fig. 6. Effects of local injections of the non-selective p-adrenergic agonist isoproterenol and antagonist propranolol into the LC complex of one side on the response gain of the ipsilateral and the contralateral triceps brachii to animal tilt. Precollicular decerebrate cat (Exp. 4). Schematic representation of a transverse section of the pons, in which the shaded area indicates the site of injection (at P2.5, R2.5, H-3.5) of 0.25 pl of isoproterenol solution at 4.5 p g / p l of saline stained with 5% pontamine (II), followed after some time by injection into the same area of 0.25 pI of propranolol at 9 p g / p l of saline (111). The site was located within the LC of the right side (left side of the figure). Below this scheme there are SPDHs showing the multiunit E M G responses of the right (IPSI) and the left (CONTRA) triceps brachii to animal tilt at 0.15 Hz, f lo" (average of 6 sweeps); the lower records indicate the animal displacement. Upper traces. Control record from the right muscles showing an average B F of 90.8 imp./sec, a response gain of 0.64 imp./sec/deg and a phase lead of 13.3", while that from the left muscle showed a BF of 80.2 imp./sec, a gain of 0.79 imp./sec/deg, and a phase lag of -2.8". Middle traces. 28 and 50 min after injection of isoproterenol into the right LC, the right muscle showed an average BF of 89.4 imp./sec, a response gain of 10.11 imp./sec/deg and a phase lead of +27.2", while the left muscle showed a BF of 93.4 imp./sec, a gain of 7.75 imp./sec/deg and a phase lead of +38.6". Lower traces. Complete recovery of the small-amplitude responses recorded 40-45 min after injection of propranolol into the same LC region. (From d'Ascanio et al. (1989a).)
+
p-adrenergic antagonist into the peri-LCa and the adjacent dorsal pRF, and (2) of an a*-or p-adrenergic agonist into the LC complex. At the
doses used in these experiments only a decreased postural activity in the ipsilateral limbs, sometimes associated with slight changes in postural
450
activity in the contralateral limbs, were observed. Moreover, in all these experiments, the increased gain of VS reflexes acting on the corresponding triceps brachii did not depend on the decreased postural activity, since it was still observed if the reduced activity of the extensor muscle was compensated by an increased static stretch of the muscle. Interestingly, when high doses of these agents were administered, the postural changes described above were interrupted from time to time by transient episodes of postural atonia, which lasted several minutes and affected the ipsilateral and occasionally also the contralateral limbs. In these instances, the EMG modulation of the corresponding triceps brachii to animal tilt at 0.15 Hz, &lo" was suppressed. These episodes were originally observed after electrolytic lesion of the LC (d'Ascanio et al., 1989c), but they also appeared after local injection into this structure either of the a,-adrenergic agonist clonidine (unpublished observations) or of the P-adrenergic agonist isoproterenol (d'Ascanio et al., 1989a). Similar results were also obtained after injection into the peri-LCa and adjacent dorsal pRF of the p-adrenergic antagonist propranolol (d'Ascanio et al., 1989b) as well as the cholinergic agonist carbachol (Barnes et aL, 1987; d'Ascanio et al., 1988). The effects described above were attributed to prominent disinhibition or direct excitation of the dorsal pontine reticular neurons and the related medullary RS neurons, leading to intense postsynaptic inhibition of the extensor a-motoneurons (cf. Morales et al., 1987). These episodes closely resembled those obtained either in intact animals during DS (Pompeiano, 1967; Jouvet, 1972) or in decerebrate cats after systemic injection of 0.10-0.15 mg/kg, iv of eserine sulfate (Pompeiano, 1980). In this instance, the postural atonia induced by the anticholinesterase was also associated with a suppression of the VS reflexes recorded from the triceps brachii during tilt (Manzoni et al., 1983b).
Interaction between cholinergic and noradrenergic pontine mechanisms Effects of focal injection of a chofinergic agonist into the locus coeruleus complex In addition to inhibitory transmitters like NE, the LC neurons are also subjected to the direct influence of excitatory transmitters like acetylcholine, as shown by iontophoretic experiments (cf. Egan and North, 1985). There is also evidence that cholinergic transmission in the LC is muscarinic in nature (cf. Wamsley et al., 1981; Egan and North, 1985). These findings are consistent with the presence of cholinergic fibers (Raichle et al., 1975) and terminals in the LC (Cheney et al., 1975; Kimura et al., 1981). Moreover, both somata and proximal dendrites of LC neurons (cf. Albanese and Butcher, 1980; Kimura et al., 1981; Kimura and Maeda, 19821, as well as presynaptic terminals ending on LC cells (Ishii, 1981) stain positively for acetylcholinesterase, the degradative enzyme of acetylcholine. On the basis of these findings, experiments were performed by Stampacchia et al. (1987) to determine whether in decerebrate cats, cholinergic activation of the NE-containing LC neurons could modify the postural activity as well as the gain of VS reflexes in the way opposite to that elicited by activation of the inhibitory noradrenergic synapses. In particular, microinjection of 0.1-0.4 pl (usually 0.25 p l ) of carbachol solution (at 0.02-0.2 pg/pl of saline) into the LC complex of one side, which slightly increased the postural activity of the ipsilateral limbs, decreased the amplitude of modulation and thus the response gain of the corresponding extensor triceps brachii to animal tilt at 0.15 Hz, &lo" (Fig. 7, upper and middle traces). This effect occurred even when the Iimb position was adjusted to produce the same background discharge of the muscle as in the control situation. In addition to the changes in gain described above, there was also an increase in phase lag of the
45 1
150
125
SIDE DOWN
SIDE DOWN
Fig. 7. Effects of local injections of the muscarinic-nicotinic agonist carbachol into different dorsal pontine tegmental structures of one side on the response gain of the ipsilateral triceps brachii to animal tilt. Precollicular decerebrate cat (Exp. 10). Schematic representation of transverse sections of the brainstem showing the position of the tip of the cannula used for the injection (at P3, L2.8, H-2.5) of 0.10 pl of carbachol solution at 0.02 p g / p l of saline (I, left scheme), followed 17,100 and 113 min afterwards by three injections (at P3, L2.8, H-3.5) of about 0.15 p1 of carbachol at 0.02, 0.2 and 0.2 pg/pl of saline, respectively (11-IV, right scheme). The corresponding sites were located within the LC of the left side (left scheme) and the dorsal tegmental reticular area immediately ventral to it (right scheme). Below these schemes there are SPDHs showing the multiunit EMG responses of the left triceps brachii to animal tilt at 0.15 Hz, f lo" (average of 6 sweeps); the lower records indicate the animal displacement. Upper traces. Control records showing an average BF of 95.2 imp./sec, a response gain of 1.09 imp./sec/deg and a phase lead of +4.3". Middle traces. Five and 7 min after injection of carbachol into the left LC, the average BF corresponded to 99.9 imp./sec, the gain decreased to 0.44 imp./sec/deg, while the phase angle corresponded to a lead of 16.6". Lower traces. Responses recorded 4 and 7 min after the last injection of carbachol into the ipsilateral tegmental reticular area. In this case the average BF corresponded to 97.0 imp./sec, the gain raised to 1.96 imp./sec/deg, while the phase angle corresponded to a lead of 11.5". (From Stampacchia et al. (1987).)
+
+
452
responses, which appeared at a threshold dose lower than that required to decrease the response gain of this extensor muscle. The effects described above, which did not affect the contralateral side, began a few min after the injection, became most prominent within the first 20-30 min, and persisted for a few hours following the injection. Moreover, the observed changes increased in magnitude by increasing the dose of carbachol. Histological controls indicated that the structure responsible for the postural and reflex changes described above corresponded to the LC (Fig. 7, left scheme). The site-specificity of the responses was shown by the fact that just the opposite effects, namely a recovery in amplitude of the EMG modulation of the forelimb extensor during tilt (Fig. 7, lower traces), was obtained in the same experiments when an equal amount of carbachol was injected into the dorsal pRF immediately ventral to the LC (Fig. 7, right scheme). The increased postural activity in the ipsilateral limbs following unilateral injection of carbachol into the LC can be attributed to cholinergic activation not only of the direct CS neurons, which exert an excitatory influence on extensor motoneurons (Fung et al., this volume), but also of the LC neurons which exert an inhibitory influence on the peri-LCa and the adjacent dorsal pRF as well as on the related inhibitory RS system. The reduced discharge of these pontine and medullary reticular neurons would also decrease the response gain of the ipsilateral triceps brachii to animal tilt. In fact, the lower the firing rate of the RS neurons is in the animal at rest, the smaller the disinhibition which affects the limb extensor motoneurons during side-down animal tilt. These motoneurons would then respond less efficiently to the same excitatory volleys elicited by given parameters of labyrinth stimulation, thus leading to a reduced gain of the EMG response of limb extensors to animal tilt. The conclusion of these experiments, i.e., that the effects of local injection of carbachol into the LC depend upon cholinergic activation of norad-
renergic LC neurons, is supported by the results of experiments showing that local infusion of the cholinergic agonist bethanechol, obtained by introducing a cannula lateral to the LC in halothane-anesthetized rats, activated LC neurons for several minutes (Adams and Foote, 1988; Foote et al., this volume). It should be mentioned, however, that in the cat the noradrenergic and the cholinergic neurons located in the dorsal pontine tegmentum are not segregated in separate groups, as in the rat, but are to some extent intermingled with a predominance of noradrenergic neurons in the LC complex and cholinergic neurons in the peri-LCa and the adjacent dorsal pRF. Therefore, when in our experiments the tip of the cannula was located in the LC, it appeared as if the direct excitatory action of carbachol on the cholinergic neurons was largely overcome by the excitatory influence on the noradrenergic neurons, which thus contributed to suppress the discharge of the pRF neurons particularly located outside the injected area. It is of interest that the effects of local injection of carbachol into the LC were mainly ipsilateral, while those of clonidine were bilateral (see above). This probably means that the LC neurons provided with cholinoreceptors project mainly ipsilaterally, while those provided with a,-adrenoceptors project bilaterally, thus affecting the periLCa and the adjacent pRF neurons of both sides.
Pontine reticular origin of cholinergic excitatory afferents to the locus coeruleus complex controlling the gain of uestibulospinal reflexes In order to identify the origin of the cholinergic excitatory afferents to the LC complex acting on V$ reflexes, we postulated that within the cholinergic pontine tegmental region, which drives the medullary inhibitory RS system, there are cells which send afferents to the LC neurons, thus providing the pRF with a negative feedback system via the LC. This possibility is supported by the fact that the LC complex receives afferents from the dorsal pontine tegmental region (Cedarbaum and Aghajanian, 1978; Clavier, 1979; Jones
453
and Yang, 1985; Sakai, this volume) where presumptive cholinergic and cholinoceptive neurons are located (cf. Jones and Sakai, this volume). If this hypothesis were correct, functional inactivation of this cholinergic afferent system to the LC following injection of the muscarinic blocker atropine sulfate, in appropriate pontine reticular structures, might reduce the discharge of the LC neurons, a finding which on the basis of clonidine experiments would lead to a decreased postural activity in these ipsilateral limbs as well as to an increased gain of response of the forelimb extensors to labyrinth stimulation. Experiments performed in decerebrate cats by Horn et al. (1987) have shown that microinjection of 0.25 pl of atropine sulfate (at 6 p g / p l of saline) into the dorsal part of the pRF did not greatly modify the postural activity in the ipsilatera1 limbs nor the gain and the phase angle of the multiunit EMG response of the corresponding triceps brachii to roll tilt of the animal at 0.15 Hz, *lo". However, if the same dose of atropine sulfate were injected into the dorsal pRF after administration of the a,-adrenergic agonist clonidine into the ipsilateral LC, the decrease in postural activity as well as the increase in amplitude of modulation and, thus, in response gain of the ipsilateral triceps brachii to labyrinth stimulation following clonidine injection became more prominent. Moreover, a further decrease in phase lead and, actually, a phase lag of the responses with respect to those obtained after clonidine administration were observed after atropine injection. It is of interest that while injection of clonidine into the LC of one side increased the gain of VS reflexes of both sides, the following injection of atropine into the ipsilateral dorsal pRF further increased the gain of the VS reflexes of the corresponding side only, while that of the contralateral side was either unmodified or slightly decreased. This finding suggests that the dorsal pRF region involved by the atropine injection affected mainly, if not exclusively, the ipsilateral LC. The reticular structures whose selective inac-
tivation by the muscarinic blocker increased the response gain of the triceps brachii to labyrinth stimulation corresponded to the pontine tegmental field located ventromedially to the LC. The results of these experiments should be compared with those reported in a previous section, showing that injection of atropine sulfate into the dorsal pRF reduced the gain of the VS reflexes only when the firing rate of the corresponding reticular neurons was locally raised either by injection of cholinergic agonists, which activated the cholinoceptive pontine neurons (Barnes et al., 1987) or by local administration of the a,-adrenergic agonist clonidine, which probably blocked the inhibitory influence exerted by the noradrenergic terminals on the dorsal pRF neurons (Pompeiano et al., 1987). In the experiments reported in this section, in which clonidine was injected directly into the LC, thus leading to a disinhibition of the dorsal p R F neurons, a successive injection of atropine sulfate into the medial aspect of the pontine reticular area greatly increased the gain of the VS reflexes. This finding can be attributed to the fact that inactivation by atropine sulfate of the presumptive cholinergic and cholinoceptive p R F neurons acting on the LC further reduced the discharge rate of the LC neurons already attenuated by the clonidine injection, as well as their inhibitory influence on the dorsal pRF and the related medullary RS neurons, thus enhancing the gain of the VS reflexes. The fact that in decerebrate cats injection of atropine sulfate into the same pontine reticular area did not modify the gain of the VS reflexes can easily be understood, since in this preparation the activity of the cholinoceptive pontine reticular neurons (Hoshino and Pompeiano, 1976) and the related medullary RS neurons (Manzoni et al., 1 9 8 3 ~is ) very low, probably due to the prominent inhibitory influence exerted on them by the LC neurons, which fire at a relatively high rate after decerebration (Pompeiano and Hoshino, 1976a,b; Pompeiano et al., 1990, 1991).
454
Discussion The experiments reported above have shown that the VS reflexes elicited by slow roll tilt of the animal, leading to stimulation of macular (gravity) receptors, are characterized by contraction of ipsilateral limb extensors during side-down tilt of the whole animal and relaxation during side-up tilt (Lindsay et ab, 1976; Schor and Miller, 1981; Manzoni et al., 1983a). These postural adjustments were originally attributed to the activity of the lateral VS neurons, which exert a direct excitatory influence on ipsilateral limb extensor motoneurons (Lund and Pompeiano, 1968; cf. Pompeiano, 1979). In fact, most of the VS neurons, including those projecting to the lumbosacral segments of the spinal cord, showed an increased discharge during side-down tilt and a decreased discharge during side-up tilt (a-responses) (Boyle and Pompeiano, 1980; Schor and Miller, 1983; Marchand et al., 1987). Additional experiments have shown that a large proportion of RS neurons, including those projecting to the lumbrosacral segments of the spinal cord, which originate from the medullary inhibitory area, responded to animal tilt (Manzoni et al., 1983~). However, most of these units showed a response pattern opposite to that displayed by the VS neurons (p-responses). These findings suggested that the contraction of limb extensors during side-down tilt, and thus the amplitude of this response, depended not only upon an increased discharge of excitatory VS neurons, but also upon a reduced discharge of presumptive inhibitory RS neurons. In order to test this hypothesis, experiments were performed to investigate whether changes in the discharge rate of these medullary inhibitory RS neurons could modify not only the postural activity but also the gain of the VS reflexes. It is known that the descending inhibitory RS neurons are under the tonic excitatory control of presumptive cholinergic and cholinoceptive neurons located in the peri-LCa and the adjacent dorsal pRF. On the other hand, these pontine regions
are under the inhibitory control of noradrenergic and NE-sensitive neurons located in the LC complex (see Introduction and Fig. 1). Using a technique by which small amounts of drugs could be applied in small volumes to produce neuropharmacologically specific effects upon different neuronal groups, attempts were made to modify the tonic discharge of the medullary inhibitory RS neurons and thus the gain of the VS reflexes by injecting appropriate cholinergic and noradrenergic agents either in the peri-LCa and the neighbouring dorsal pRF or in the LC area. In particular, in a first group of experiments (Barnes et aZ., 19871, unilateral injection into the peri-LCa and the adjacent dorsal p R F of small doses of cholinergic agonists, like carbachol which acts on both muscarinic and nicotinic receptors, or bethanechol which acts on pure muscarinic receptors, while exciting the corresponding cholinoceptive neurons, decreased the postural activity in the ipsilateral limbs but greatly increased the response gain of the triceps brachii to labyrinth stimulation. These changes were suppressed after local injection into the same region of small doses of the muscarinic blocker atropine sulfate. The effects of carbachol and bethanechol did not depend on the reduced postural activity, since they were still observed when the EMG activity was kept constant by appropriate adjustments of the static stretch of the muscle. Moreover, the results obtained were dose-dependent and site-specific, as shown by histological controls. The conclusion of these experiments, i.e., that the mechanisms responsible for the postural and reflex changes described above were cholinergically mediated, is in line with the results of previous experiments showing that in decerebrate cats: (1) systemic injection of small doses of an anticholinesterase such as eserine sulfate, which decreased the postural activity, greatly increased the gain of the VS reflexes, these findings being suppressed by iv administration of atropine sulfate (Pompeiano et aZ., 1983); (2) electrolytic lesion of dorsal pontine reticular structures of one
455
side increased the postural activity in the ipsilatera1 limbs, due to a reduced discharge of the inhibitory mRS system, but greatly decreased the amplitude of the EMG responses of the triceps brachii to labyrinth stimulation (d’Ascanio et af., 1985b; Pompeiano, 1985); 3) the same lesion also suppressed the decreased postural activity as well as the increased gain of these EMG responses after systemic injection of small doses of the anticholinesterase eserine sulfate (d’Ascanio et af., 198Sb; Pompeiano, 1985). The influence that the cholinoceptive pontine tegmental neurons exert on the static and dynamic control of posture can be understood if we consider that in decerebrate cats cholinergic activation of the peri-LCa and the adjacent pRF neurons tonically excites the medullary inhibitory RS neurons. This was experimentally shown by recording the changes in firing rate of these dorsal pontine reticular neurons (Hoshino and Pompeiano, 19761, as well as of medullary RS neurons (Srivastava et af., 1982) which occur in decerebrate cats after systemic injection of an anticholinesterase. In the experiments by Barnes et al. (19871, the increased discharge of these inhibitory RS neurons induced by local injection into the pRF of cholinergic agonists would first exert a tonic inhibitory influence on extensor motoneurons (Morales et al., 19871, thus decreasing the extensor rigidity (cf. also Mori, 1989). Moreover, since the inhibitory RS neurons fire out of phase with respect to the excitatory VS neurons acting on ipsilateral limb extensor motoneurons, we can assume that for a given labyrinth signal, the higher the firing rate of these RS neurons in the animal at rest, the greater the disinhibition affecting the extensor motoneurons during side-down animal tilt. These motoneurons would then respond more efficiently to the same excitatory VS volleys elicited by given parameters of stimulation, thus giving rise to an increased gain of the EMG responses of forelimb extensor muscles to labyrinth stimulation. On the other hand, injection of the muscarinic antagonist atropine sulfate into the dorsal p R F suppressed the
postural hypotonia in the ipsilateral limbs induced by local injection of a cholinergic agonist as well as the increase in the response gain of the ipsilateral triceps brachii to animal tilt, which reached values lower than those obtained in the control records (Barnes et af., 1987). The finding that presumptive cholinergic and cholinoceptive pontine reticular neurons as well as the related medullary RS neurons are under the tonic inhibitory control of the LC, led us to expect that inactivation of the NE-containing LC system produced changes in posture as well as in gain of the VS reflexes similar to those obtained by direct activation of the dorsal pontine reticular system. In fact, unilateral injection into the periL C a and the dorsal p R F of the P-adrenergic antagonist propranolol (d’Ascanio et al., 1989b), which blocked the inhibitory influence of noradrenergic LC neurons on the pontine reticular system, or else injection into the LC complex of one side of the a,-adrenergic agonist clonidine (Pompeiano et af., 1987) or the P-adrenergic agonist isoproterenol (d’Ascanio et af., 1989a), which led to functional inactivation of noradrenergic and NE-sensitive LC neurons, not only reduced the postural activity in the limb extensors, but also increased the amplitude of modulation and thus the gain of the multiunit EMG responses of the triceps brachii to labyrinth stimulation. These effects, which were mainly ipsilateral after injection of propranolol into the pRF, or bilateral after injection of clonidine and isoproterenol into the LC complex, were attributed to disinhibition of presumptive cholinergic and cholinosensitive pontine reticular neurons, since they were, under given conditions, suppressed by local administration into the peri-LCa and the adjacent dorsal pRF structures of the muscarinic blocker atropine sulfate. Therefore, the same explanation of the postural and reflex changes induced by local injection of cholinergic agonists into the peri-LCa and the neighbouring pRF could also apply to the results of experiments leading to inactivation of the noradrenergic LC system. It is worth mentioning that the decerebrate
456
A
C
8 RESTING
D
ACTIVITY
LC-cs NEURONS
PRF NEURONS
LIMB E X 1
J- -
RESPONSES T O ANIMAL TILT
LIMB E X 1 ANIMAL PoSIT’oN
side down
Fig. 8. Scheme illustrating the “primary” and “secondary” range of operation of the LC-complex neurons and the adjacent dorsal pontine tegmental neurons, leading to “reciprocal” or “parallel” changes in posture and gain of the VS reflexes, respectively. LC-CS neurons, resting discharge of neurons projecting to the spinal cord either through the p R F (LC) or directly, through the coeruleospinal projection (CS); p R F neurons, discharge of neurons located in the peri-LCa and the adjacent dorsal pontine reticular formation; mRS neurons, discharge of medullary inhibitory reticulospinal neurons; limb ext., E M G activity of the triceps brachii in the animal at rest (upper traces) or during animal tilt at 0.15 Hz, f lo” (lower traces). A. In precollicular decerebrate cats, the maintained discharged of the LC-CS neurons is associated with a reduced activity of the p R F and the related mRS neurons; in this instance the prominent extensor rigidity is associated with a small amplitude EMG modulation of the limb extensor during tilt. B,C. A reduced discharge of the LC-CS neurons associated with an increased activity of the pRF-mRS neurons reduces the postural activity, while the amplitude of the EMG modulation of the limb extensor during tilt increases. D. A further decrease o r suppression of the discharge of the LC-CS neurons associated with a prominent activity of the pRF-mRS neurons leads to a postural atonia, associated with suppression of the VS reflexes. (Adapted from Pompeiano et al. (1983) and Manzoni et al. (1983b1.1
cats, in contrast to intact animals, show on an average a higher resting discharge of the LC neurons, probably due to interruption of a supramesencephalic descending pathway exerting a tonic inhibitory influence on the LC neurons (Pompeiano and Hoshino, 1976a,b; Pompeiano et aL, 1990, this volume). This finding would lead to
a prominent inhibition of the dorsal pRF neurons and the related medullary inhibitory RS neurons, which might account not only for the increased postural activity in the decerebrate animals, but also for the low gain of response of the ipsilateral limb extensors to labyrinth stimulation (see Fig. 8A). In our experiments, an increase in gain of
457
VS reflexes occurred only after local injection of small doses of noradrenergic or cholinergic agents, which presumably led either to a small decrease in firing rate of the LC-complex neurons or to a moderate increase in discharge of the dorsal pRF neurons, as indicated by the slight hypotonia which occurred after the injections (see Fig. 8 from A to C): this situation can be referred to a “primary” range of operation of the system. However, when higher doses of these agents were injected, thus leading either to suppression of the discharge of the LC-complex neurons or to a great increase in discharge of the pontine and related medullary reticular neurons, the postsynaptic inhibition of the extensor motoneurons became so prominent (Morales et aL, 1987) to suppress posture, as well as the EMG responses of the limb extensors to labyrinth stimulation (Barnes et al., 1987; d’Ascanio et al., 1988, 1989a,b; cf. also d’Ascanio et al., 1 9 8 9 ~(see ) Fig. 8D); this situation can be attributed to a “secondary” range of operation of the system. Results similar to those described above might occur in intact animals under strict physiological conditions. Experiments of unit recording have shown that in these preparations the resting discharge of the LC neurons which is quite low during quiet waking, thus being comparable to the situation illustrated in Figure 8B and/or C, increases during alertness (cf. Foote et al., 1983) or stress (Abercrombie and Jacobs, 1987a), but disappears during desynchronized sleep (cf. Sakai, l980,1985a, 1988; Foote et al., 1983; Aston-Jones, 1985; Hobson and Steriade, 1986; Jacobs, 1986). While in the first instance we should expect an increase in postural activity associated with a decreased gain of VS reflexes, in the second instance the arrest of the tonic discharge of LC neurons would lead to such a prominent increase in activity of dorsal pontine reticular neurons (see for ref. Sakai, 1980, 1988; Foote et al., 1983; Hobson and Steriade, 1986) and the related medullary inhibitory RS neurons (Siege1 et al., 1979; Kanamori et al., 1980; Chase el al., 1981; cf. Sakai, this volume) to suppress posture (cf.
Pompeiano, 1967; Jouvet, 19721, as well as the responses of limb extensors to labyrinth stimulation. This hypothesis is supported by the fact that during the cataplectic episodes induced by systemic injection of an anticholinesterase in decerebrate cats, in which the discharge rate of the LC neurons disappeared as during the episodes of desynchronized sleep (Pompeiano and Hoshino, 1976a,b; cf. Pompeiano, 19801, while that of the dorsal pontine reticular neurons and the related medullary inhibitory RS neurons greatly increased (Hoshino and Pompeiano, 1976; Srivastava et al., 1982; cf. Pompeiano, 19801, there was a complete suppression not only of posture (Matsuzaki et al., 1968; cf. Pompeiano, 1976, 1980), but also of VS reflexes (Manzoni et al., 1983b). In conclusion, while an increased discharge of the LC neurons during the “primary” range of operation of the system would reduce the gain of the VS reflexes (see Fig. 8, from C to A), an increased discharge of the same neurons during the “secondary” range would produce the opposite effect (see Fig. 8, from D to C). This operational model suggests that the neuronal circuit made by the inhibitory LC neurons, as well as the excitatory pRF neurons, acts as a variable gain regulator, which may adapt to the animal state the response gain of limb extensors to labyrinth stimulation. The same system could also play a prominent role in the compensation of the postural and reflex deficits induced by labyrinthectomy, as well as in the adaptive processes which occur during exposure to micro-gravity and subsequent re-adaptation to a normal environment. It is of interest that in addition to inhibitory transmitters, like NE, the LC neurons are also influenced by excitatory transmitters, like acetylcholine. These influences, which can be transmitted through cholinergic fibers (Raichle et al., 1975) and terminals (Cheney et al., 1975; Kimura et al., 1981) originating in part at least from the dorsal pRF, would counteract the inhibitory influences exerted on the same LC neurons by the noradrenergic synapses. In fact, in contrast to the
458
results obtained after local inactivation of the LC neurons by a2- and P-adrenergic agonists, activation of these neurons by a cholinergic agonist not only increased the postural activity in the ipsilatera1 limbs, but also reduced the response gain of the corresponding triceps brachii to labyrinth stimulation (Stampacchia et al., 1987). As a result of these findings it appears that the presumptive cholinergic pRF neurons, which are not only self-excitatory (in fact they are cholinoceptive) but also excitatory on the medullary inhibitory RS neurons, also send excitatory afferents to the LC. On the other hand, the NE-containing LC neurons are not only self-inhibitory, due to mechanisms of recurrent or lateral inhibition, but also inhibitory on the adjacent dorsal pRF neurons. These interconnections between putative cholinergic and noradrenergic neurons closely correspond to those postulated by the reciprocal interaction model to be responsible for the occurrence of desynchronized sleep (cf. Hobson and Steriade, 1986; Hobson et al., 1986). However, according to these authors the presumptive cholinergic reticular system was considered to be located in the gigantocellular tegmental field, a region different from the peri-LCcu and the adjacent dorsal pRF (cf. Jones and Sakai, this volume). Experiments are in progress to investigate whether other excitatory (i.e., glutamatergic) or inhibitory (i.e., GABAergic) systems act on the neuronal circuit described above, to control posture as well as the gain of the VS reflexes.
Acknowledgements This work was supported by National Institute of Neurological and Communicative Disorders and Stroke Research Grant NS 07685-22 and by grants of the Minister0 dell’Universit5 e della Ricerca Scientifica e Tecnologica, and the Agenzia Spaziale Italiana, Roma, Italy.
References Abercrombie, E.D. and Jacobs, B.L. (1987a) Single-unit response of noradrenergic neurons in the locus coeruleus of
freely moving cats. I. Acutely presented stressful and nonstressful stimuli. J. Neurosci., 7: 2837-2843. Abercrombie, E.D. and Jacobs, B.L. (1987b) Microinjected clonidine inhibits noradrenergic neurons of the locus coeruleus in freely moving cats. Neurosci. Lett., 76: 203208. Adams, L.M. and Foote, S.L. (1988). Effects of locally infused pharmacological agents on spontaneous and sensoryevoked activity of locus coeruleus neurons. Brain Rex Bull., 21: 395-400. Aghajanian, G.K., Cedarbaum, J.M. and Wang, R.Y. (1977) Evidence for norepinephrine-mediated collateral inhibition of locus coeruleus neurons. Bruin Res., 136: 570-577. Albanese, A. and Butcher, L.L. (1980) Acetylcholinesterase and catecholamine distribution in the locus coeruleus of the rat. Brain Res. Bull., 5: 127-134. Aston-Jones, G. (1985) Behavioral function of locus coeruleus derived from cellular attributes. Physiol. Psychol., 13: 118126. Atlas, D. and Melamed, E. (1978) Direct mapping of @-adrenergic receptors in the rat central nervous system by a novel fluorescent @-blocker. Brain Res., 150: 377-385. Atlas, D. and Segal, M. (1977) Simultaneous visualization of noradrenergic fibers and P-adrenoreceptors in pre- and postsynaptic regions in rat brain. Bruin Res., 135: 347-350. Barnes, C.D., d’Ascanio, P., Pompeiano. 0. and Stampacchia, G. (1987) Effects of microinjections of cholinergic agonists into the pontine reticular formation on the gain of the vestibulospinal reflexes in decerebrate cats. Arch. Ital. BioL, 125: 71-105. Beani, L., Bianchi, C., Giacomelli, A. and Tamberi, F. (1978) Noradrenaline inhibition of acetylcholine release from guinea-pig brain. Eur. J. Pharrnacol., 48: 179-193. Berman, A.L. (1968) The Brain Stern of the Cat. A Cytoarchitectonic Atlas with Stereotuxic Coordinates, University of Wisconsin Press, Madison, pp. XVI-175. Bianchi, C., Spidalieri, G., Guandalini, P., Tanganelli, S. and Beani, L. (1979) Inhibition of aceytlcholine outflow from guinea-pig cerebral cortex following locus coeruleus stimulation. Neurosci. Lett., 14: 97-100. Bier, M.J., Greene, R.W. and McCarley, R.W. (1990) Norepinephrine mediated responses in the pontine reticular formation: In uitro and in oiuo studies. Proc. 10th Eur. Congr. Sleep Res., Strasbourg, May 20-25, 333 pp. Biscoe, T.J. and Curtis, D.R. (1966) Noradrenaline and inhibition of Renshaw cells. Science, 151: 1230-1231. Boyle, R. and Pompeiano, 0. (1980) Reciprocal responses to sinusoidal tilt of neurons in Deiters’ nucleus and their dynamic characteristics. Arch. Ital. Biol., 118: 1-32. Boyle, R. and Pompeiano, 0. (1984) Discharge activity of spindle afferents from the gastroenemius-soleus muscle during head rotation in the decerebrate cat. pfligers Arch., 400: 140-150. Cedarbaum, J.M. and Aghajanian, G.K. (1976) Noradrenergic neurons of the locus coeruleus: Inhibition by epinephrine and activation by the a-antagonist piperoxane. Bruin Res., 112: 413-419. Cedarbaum, J.M. and Aghajanian, G.K. (1977) Catecholamine
4.59 receptors on locus coeruleus neurons: Pharmacological characterization. Eur. J. Pharmacpl., 44: 375-385. Cedarbaum, J.M. and Aghajanian, G.K. (1978) Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique. L Comp. Neurol., 178: 1-16. Chase, M.N., Enomoto, S., Murakami, T., Nakamura, Y. and Taira, M. (1981) Intracellular potential of medullary reticular neurons during sleep and wakefulness. Exp. Neurol., 71: 226-233. Cheney. D.L., LeFevre, H.F. and Racagni, G. (1975) Choline acetyltransferase activity and mass fragmentographic measurement of acetylcholine in specific nuclei and tracts of rat brain. Neuropharmacology, 14: 801-809. Clavier, R.M. (1979) Afferent projections to the self-stimulation regions of the dorsal pons, including the locus coeruleus, in the rat as demonstrated by the horseradish peroxidase technique. Brain Res. Bull., 4: 497-504. d’Ascanio, P., Bettini, E. and Pompeiano, 0. (1985a) Tonic inhibitory influences of locus coeruleus on the response gain of limb extensors to sinusoidal labyrinth and neck stimulations. Arch. Ital. Biol., 123: 69-100. d’Ascanio, P., Bettini, E. and Pompeiano, 0. (198%) Tonic facilitatory influences of dorsal pontine reticular structures on the response gain of limb extensors to sinusoidal labyrinth and neck stimulations. Arch. Ira/. Biol., 123: 101-132. d’Ascanio, P., Pompeiano, O., Stampacchia, G. and Tononi, G. (1988) Inhibition of vestibulospinal reflexes following cholinergic activation of the dorsal pontine reticular formation. Arch. Ital. BioL, 126: 291-316. d’Ascanio, P., Horn, E., Pompeiano, 0. and Stampacchia, G. (1989a) Injections of P-adrenergic substances in the locus coeruleus affect the gain of vestibulospinal reflexes in decerebrate cats. Arch. Ital. Biol., 127: 187-218. d’Ascanio, P., Horn, E., Pompeiano, 0. and Stampacchia, G. (1989b) Injections of a P-adrenergic antagonist in pontine reticular structures modify the gain of vestibulospinal reflexes in decerebrate cats. Arch. Ital. Biol., 127: 275-303. d’Ascanio, P., Pompeiano, M. and Tononi, G. (1989~)Inhibition of vestibulospinal reflexes during the episodes of postural atonia induced by unilateral lesion of the locus coeruleus in the decerebrate cat. Arch. Ital. Biol., 127: 81-97. Egan, T.M. and North, R.A. (1985) Acetylcholine acts on m z muscarinic receptors to excite rat locus coeruleus neurons. Br. J. Pharmacol., 85: 733-735. El Mansari, M., Sakai, K. and Jouvet, M. (1989) Unitary characterisitics of presumptive cholinergic tegmental neurons during sleep-waking cycle in freely moving cats. Exp. Brain Res., 76: 519-529. Engberg, I. and Ryall, R.W. (1966) The inhibitow action of noradrenaline and other monoamines on spinal neurons. . I Physiol. . (London), 185: 298-322. Ennis, M. and Aston-Jones, G. (1986) Evidence for self- and neighbor-mediated post-activation inhibition of locus coeruleus neurons. Brain Res., 374: 299-305. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Rec., 63: 844-914.
Fung, S.J., Pompeiano, 0. and Barnes, C.D. (1988) Coeruleospinal influence on recurrent inhibition of spinal motonuclei innervating antagonistic hindleg muscles of the cat. PfZugers Arch., 412 346-353. Greene, R.W., Gerber, U., Haas, H.L. and McCarley, R.W. (1989) Noradrenergic actions on neurons of the medial pontine reticular formation in citro. Sleep Res., 18: 11. Groves, P.M. and Wilson, C.J. (1980) Monoaminergic presynaptic axons and dendrites in rat locus coeruleus seen in reconstructions of serial sections. J. Comp. Neurol., 193: 853-862. Grzanna, R. and Molliver, M.E. (1980) Cytoarchitecture and dendritic morphology of central noradrenergic neurons. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Recisited, IBRO Monograph Series, Vol. 6 , Raven Press, New York, pp. 83-97. Guyenet, P.G. (1980) The coeruleospinal noradrenergic neurons: Anatomical and electrophysiological studies in the rat. Brain Res., 189: 121-133. Henley, K. and Morrison, A.R. (1974) A re-evaluation of the effects of lesions of the pontine tegmentum and locus coeruleus on phenomena of paradoxical sleep in the cat. Acta Neurobiol. Exp., 34: 215-232. Hess, A. (1979) Visualization of P-adrenergic receptor sites with fluorescent P-adrenergic blocker probes- or autofluorescent granules? Brain Rex, 160: 533-538. Hobson, J.A. and Steriade, M. (1986) Neuronal basis of behavioral state control. In F.E. Bloom (Ed.), Intrinsic Regulatory System of the Brain, Handbook of Physiology, Section I , The Nervous System, Vol. W , American Physiological Society, Bethesda, Maryland, pp. 701-823. Hobson, J.A., Lydic, R. and Baghdoyan, H.A. (1986) Evolving concepts of sleep cycle generation: From brain centers to neuronal populations. Behac. Brain Res., 9: 371-448. Hokfelt, T., Fuxe, K., Goldstein, M. and Johansson, 0. (1974) Immunohistochemical evidence for the existence of adrenaline neurons in the rat brain. Brain Res., 66: 235251. Horn, E., d’Ascanio, P., Pompeiano, 0. and Stampacchia, G. (1987) Pontine reticular origin of cholinergic excitatory afferents to the locus coeruleus controlling the gain of vestibulospinal and cervicospinal reflexes in decerebrate cats. Arch. Ital. Biol., 125: 273-304. Hoshino, K. and Pompeiano, 0. (1976) Selective discharge of pontine neurons during the postural atonia produced by an anticholinesterase in the decerebrate cat. Arch. Ital. Biol., 114: 244-277. Ishii, S. (1971) Morphological studies on the mechanism of adrenergic transmission in the central nervous system. I. Granulated vesicles. In 0. Eranko (Ed.), Histochemistry of Nervous Transmission, Progress in Brain Research, Vol. 34, Elsevier, Amsterdam, pp. 187-206. Jacobs, B.L. (1986) Single unit neurons in behaving animals. Prog. Neurobiol., 27: 183-194. Jankowska, E., Lund, S., Lundberg, A. and Pompeiano, 0. (1968) Inhibitory effects evoked through ventral reticulospinal pathways. Arch. Ital. Btol., 106: 124-140. Jones, B.E. (1990) Immunohistochemical study of choline acetyltransferase immunoreactive processes and cells in-
460 nervating the pontomedullary reticular formation in the rat. J. Comp. Neurol., 295: 485-514. Jones, B.E. and Beaudet, A. (1987) Distribution of acetylcholine and catecholamine neurons in the cat brainstem: A choline acetyltransferase and tyrosine hydroxylase immunohistochemical study. J. Comp. Neurol., 261: 15-32. Jones, B.E. and Friedmart, L. (1983) Atlas of catecholamine pericarya varicosities and pathways in the brainstem of the cat. J. Comp. Neurol., 215: 382-396. Jones, B.E. and Yang, T.-Z. (1985) The efferent projections from the reticular formation and the locus coeruleus studied by anterograde and retrograde axonal transport in the rat. J. Comp. Neurol., 242: 56-92. Jouvet, M. (1972) The role of monoamines and acetylcholinecontaining neurons in the regulation of the sleep-waking cycle. Ergebn. Physiol., 64: 166-307. Jouvet, M. and Delorme, F. (1965) Locus coeruleus et sommeil paradoxal. C.R. Soc. B i d , 159: 895-899. Kanamori, N., Sakai, K. and Jouvet, M. (1980) Neuronal activity specific to paradoxical sleep in the ventromedial medullary reticular formation of unrestrained cats. Brain Res., 189: 251-255. Kasper, J., Schor, R.H. and Wilson, V.J. (1988) Response of vestibular neurons to head rotations in vertical planes. I. Responses to vestibular stimulation. J. Neurophysiol., 60: 1753- 1764. Kimura, H. and Maeda, T. (1982) Aminergic and cholinergic systems in the dorsolateral pontine tegmentum. Brain Res. Bull., 9: 493-499. Kimura, H., McGeer, P.L., Peng, J.H. and McGeer, E.G. (1981) The central cholinergic system studied by choline acetyltransferase immunohistochemistry in the cat. J. Comp. Neurol., 200: 151-201. Koda, L.Y., Aston-Jones, G. and Bloom, F.E. (1980) Small granular vesicles in the locus coeruleus may indicate dendritic release of norepinephrine. Soc. Neurosci. Abstr., 6: 446. Lai, Y.Y. and Siegel, J. (1988) Medullary regions mediating atonia. J. Neurosci., 8 4790-4796. Lindsay, K.W., Roberts, T.D.M. and Rosenberg, J.R. (1976) Asymmetric tonic labyrinth reflexes and their interaction with neck reflexes in the decerebrate cat. J. Physiol. (London), 261: 583-601. Llinls, R. and Terzuolo, C.A. (1964) Mechanisms of supraspinal actions upon spinal cord activities. Reticular inhibitory mechanisms on alpha extensor motoneurons. J . Neurophysiol., 27: 5 79-5 91. Lund, S. and Pompeiano, 0. (1968) Monosynaptic excitation of a-motoneurons from supraspinal structures in the cat. Act0 Physiol. Scand., 73: 1-21. Luppi, P.-H., Sakai, K., Fort, P., Salvert D. and Jouvet, M. (1988) The nuclei of origin of monoaminergic, peptidergic and cholinergic afferents to the cat nucleus reticularis magnocellularis: A double-labeling study with cholera toxin as a retrograde tracer. J. Comp. Neurol., 277: 1-20. Magoun, H.W. and Rhines, R. (1947) Spasticity. The StretchReflex and Extrapyramidal Systems, C.C. Thomas, Springfield, 59 pp.
Manzoni, D., Pompeiano, O., Srivastava, U.C. and Stampacchia, G. (1983a) Responses of forelimb extensors to sinusoidal stimulation of macular labyrinth and neck receptors. Arch. Ital. Biol., 121: 205-214. Manzoni, D., Pompeiano, O., Srivastava, U.C. and Stampacchia, G. (1983b) Inhibition of vestibular and neck reflexes in forelimb extensor muscles during the episodes of postural atonia induced by an anticholinesterase in decerebrate cat. Arch. Ital. Biol., 121: 267-283. Manzoni, D., Pompeiano, O., Stampacchia, G. and Srivastava, U.C. (1983~)Responses of medullary reticulospinal neurons to sinusoidal stimulation of labyrinth receptors in decerebrate cat. J. Neurophysiol., 50: 1059-1079. Manzoni, D., Pompeiano, O., Srivastava, U.C. and Stampacchia, G. (1984) Gain regulation of vestibular reflexes in fore- and hindlimb muscles evoked by roll tilt. Boll. Soc. Ital. Biol. Sper., 60, Suppl.: 9-10. Manzoni, D., Marchand, A.R., Pompeiano, 0. and Stampacchia, G. (1987) Coding of the velocity signal by lumbar projecting vestibulospinal neurons in decerebrate cats. Neuroscience, 22, Suppl.: S739. Marchand, A.R., Manzoni, D., Pompeiano, 0. and Stampacchia, G. (1987) Effects of stimulation of vestibular and neck receptors on Deiters’ neurons projecting to the lumbosacral cord. pfliigers Arch., 409: 13-23. Matsuzaki, M., Okada, Y. and Shuto, S. (1968) Cholinergic agents related to para-sleep state in acute brain stem preparation. Brain Res., 9: 253-267. Mesulam, M.-M., Mufson E.J., Levey, A.I. and Wainer, B.H. (1984) Atlas of cholinergic neurons in the forebrain and upper brainstem of the macaque based on monoclonal choline acetyltransferase immunohistochemistry and acetylcholinesterase histochemistry. Neuroscience, 12: 669686. Morales, F.R., Engelhardt, J.K., Soja, P.J., Pereda, A.E. and Chase, M.H. (1987) Motoneurons properties during motor inhibition produced by microinjection of carbachol into the pontine reticular formation of the decerebrate cat. J. Neurophysiol., 57: 1118-1129. Mori, S. (1989) Contribution of postural muscle tone to full expression of posture and locomotor movements: Multifaceted analyses of its setting brainstem-spinal cord mechanisms in the cat. Jpn. J. Physiol, 39: 785-809. Moroni, F., Tanganelli, S., Antonelli, T., Carla, V., Bianchi, C. and Beani, L. (1983) Modulation of cortical acetylcholine and y-amino-butyric acid release in freely moving guinea pigs: Effects of clonidine and other adrenergic drugs. J. Pharmacol. Exp. Ther., 227: 435-440. Ohta, Y . ,Mori, S. and Kimura, H. (1988) Neuronal structures of the brainstem participating in postural suppression in cats. Neurosci. Res., 5: 181-202. Palacios, J.M. and Kuhar, M.J. (1980) Beta-adrenergic receptor localization by light microscopic autoradiography. Science, 208: 1378-1380 Palacios, J.M. and Kuhar, M.J. (1982) Beta-adrenergic receptor localization in the rat brain by light microscopic autoradiography. Neurochem. Int., 4: 473-490. Palacios, J.M. and Wamsley, J.K. (1984) Catecholamine re-
461 ceptors. In A. Bjorklund, T. Hokfelt and M.J. Kuhar (Eds.), Classical Transmitters and Transmitter Receptors in the CNS. Part IL Handbook of Chemical Anatomy, Vol. 3, Elsevier, Amsterdam, pp. 325-351. Peterson, B.W., Pitts, N.G. and Fukushima, K. (1979) Reticulospinal connections with limb and axial motoneurons. Exp. Brain Rex, 36: 1-20. Pickel, V.M., Joh, T.H. and Reis, D.J. (1977) A serotonergic innervation of noradrenergic neurons in nucleus locus coeruleus: Demonstration by immunocytochemical localization of the transmitter specific enzymes tyrosine or tryptophan hydroxylase. Brain Res., 131: 197-214. Pompeiano, M., Ronca-Testoni, S. and Pompeiano, 0. (1989) Distribution of P-adrenergic receptors in different brainstem regions controlling posture in the cat. Arch. Ital. Biol., 127: 13-23. Pompeiano, 0. (1967) The neurophysiological mechanisms of the postural and motor events during desynchronized sleep. Res. Publ. Assoc. Nerc. Ment. Dis., 45: 351-423. Pompeiano, 0. (1975) Vestibulo-spinal relationships. In R.F. Naunton (Ed.), The Vestibular System, Academic Press, New York, San Francisco, London, pp. 147-180. Pompeiano, 0. (1976) Mechanisms responsible for spinal inhibition during desynchronized sleep: Experimental study. In C. Guilleminault, W.C. Dement and P. Passouant (Eds.), Narcolepsy, Adcances in Sleep Research, Vol. 3, Spectrum, New York, pp. 41 1-449. Pompeiano, 0. (1979) Neck and macular labyrinthine influences on the cervical spino-reticulocerebellar pathway. In R. Granit and 0. Pompeiano (Eds.), Reflex Control of Posture and Mouement, Progress in Brain Research, Vol. 50, Elsevier, Amsterdam, pp. 501-514. Pompeiano, 0. (1980) Cholinergic activation of reticular and vestibular mechanisms controlling posture and eye movements. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Reuisited, IBRO monograph Series, Vol. 6, Raven Press, New York, pp. 473-512. Pompeiano, 0. (1984) Recurrent inhibition. In R.A. Davidoff (Ed.), Handbook of the Spinal Cord, Vol. 2-3, Marcel Dekker Inc., New York, pp. 461-557. Pompeiano, 0. (1985) Cholinergic mechanisms involved in the gain regulation of postural reflexes. In A. Waquier, J.M. Gaillard, J.M. Monti and M. Radulovacki (Eds.), Sleep: Neurotransmitters and Neuromodulators, Raven Press, New York, pp. 165-184. Pompeiano, 0. and Hoshino, K. (1976a) Central control of posture: Reciprocal discharge by two pontine neuronal groups leading to suppression of decerebrate rigidity. Brain Res., 116: 131-138. Pompeiano, 0. and Hoshino, K. (1976b) Tonic inhibition of dorsal pontine neurons during the postural atonia produced by an anticholinesterase in the decerebrate cat. Arch. Ital. Biol., 114: 310-340. Pompeiano, O., Manzoni, D., Srivastava, U.C. and Stampacchia, G. (1983) Cholinergic mechanisms controlling the response gain of forelimb extensor muscles to sinusoidal stimulation of macular labyrinth and neck receptors. Arch. Ital. Biol., 121: 285-303.
Pompeiano, O., Wand, P. and Srivastava, U.C. (1985a) Responses of Renshaw cells coupled with hindlimb extensor motoneurons to sinusoidal stimulation of labyrinth receptors in the decerebrate cat. Pfligers Arch., 403: 245-257. Pompeiano, O., Wand, P. and Srivastava, U.C. (1985b) Influence of Renshaw cells on the gain of hindlimb extensor muscles to sinusoidal labyrinth stimulation. PJliigers Arch., 404: 107-118. Pompeiano, O., d’Ascanio, P., Horn, E. and Stampacchia, G. (1987) Effects of local injection of the a,-adrenergic agonist clonidine into the locus coeruleus complex on the gain of vestibulospinal and cervicospinal reflexes in decerebrate cats. Arch. Ital. Biol., 125: 225-269. Pompeiano, O., Manzoni, D., Barnes, C.D., Stampacchia, G. and d’Ascanio, P. (1990) Responses of locus coeruleus and subcoeruleus neurons to sinusoidal stimulation of labyrinth receptors. Neuroscience, 35: 227-248. Raichle, M.E., Hartman, B.K., Eichling, J.O. and Sharpe, L.G. (1975) Central noradrenergic regulation of cerebral blood flow and vascular permeability. Proc. Natl. Acad. Sci., USA, 72: 3726-3730. Rainbow, T.C., Parsons, B. and Wolfe, B.B. (1984) Quantitative autoradiography of p,- and P,-adrenergic receptors in rat brain. Proc. Natl. Acad. Sci., USA, 81: 1585-1589. Rye, D.B., Lee, H.J., Saper, C.B. and Wainer, B.H. (1988) Medullary and spinal efferents of the pedunculopontine tegmental nucleus and adjacent mesopontine tegmentum in the rat. J. Comp. Neurol., 269: 315-341. Sakai, K. (1980) Some anatomical and physiological properties of ponto-mesencephalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep. In J.A. Hobson and M.A.B. Brazier (Eds.), The Reticular Formation Recisited, IBRO Monograph Series, Vol. 6, Raven Press, New York, pp. 427-447. Sakai, K. (1985a) Anatomical and physiological basis of paradoxical sleep. In D.J. McGinty, R. Drucker-Colin, A. Morrison and L. Parmeggiani (Eds.), Brain Mechanisms of Sleep, Raven Press, New York, pp. 111-137. Sakai, K. (1985b) Neurons responsible for paradoxical sleep. In A. Waquier, J.M. Gaillard, J.M. Monti and M. Radulovacki (Eds.), Sleep: Neurotransmitters and Neuromodulators, Raven Press, New York, pp. 29-42. Sakai, K. (1988) Executive mechanisms of paradoxical sleep. Arch. Ital. Biol., 126: 239-257. Sakai, K., Touret, M., Salvert, D., Leger, L. and Jouvet, M. (1977) Afferent projections to the cat locus coeruleus as visualized by horseradish peroxidase technique. Brain Res., 119: 21-41. Sakai, K., Sastre, J.-P., Salvert, D., Touret, M., Tohyama, M. and Jouvet, M. (1979) Tegmentoreticular projections with special reference to the muscular atonia during paradoxical sleep in the cat: An HRP study. Brain Res., 176: 233-254. Sakai, K., Luppi, P.-H., Salvert, D., Kimura, H., Maeda, T. and Jouvet, M. (1986) Localisation des neurones cholinergiques dans le tronc cerebral inferieur chez le chat. C.R. Acad. Sci. Paris, Series III, 303: 317-324. Schor, R.H. and Miller, A.D. (1981) Vestibular reflexes in
462 neck and forelimb muscles evoked by roll tilt. J. Neurophysiol., 46: 167-178. Schor, R.H. and Miller, A.D. (1982) Relationship of cat vestibular neurons to otolith-spinal reflexes. Exp. Bruin Res., 47: 137-144. Shirornani, P.J. and McGinty, D.J. (1986) Pontine neuronal response to local cholinergic microinfusion: Relation to REM sleep. Bruin Res., 386: 20-31. Shirornani, P.J., Lai, Y.Y. and Siegel, J.M. (1990) Descending projections from the dorsolateral pontine tegrnentum to the paramedian reticular nucleus of the caudal medulla in the cat. Bruin Res., 517: 224-228. Shimizu, N., Katoh, Y., Hida, T. and Satoh, K. (1979) The fine structural organization of the locus coeruleus in the rat with reference to noradrenaline contents. Exp. Bruin Res., 37: 139-148. Siegel, J.M., Wheeler, R.L. and McGinty, D.J. (1979) Activity of medullary reticular formation neurons in the unrestrained cat during waking and sleep. Bruin Res., 179: 49-60. Srivastava, U.C., Manzoni, D., Pompeiano, 0. and Starnpacchia, G. (1982) State-dependent properties of medullary reticular neurons involved during the labyrinth and neck reflexes. Neurosci. Lett., 10: S461. Stampacchia, G., Barnes, C.D., d’Ascanio, P. and Pompeiano, 0. (1987) Effects of microinjection of a cholinergic agonist into the locus coeruleus on the gain of vestibulospinal reflexes in decerebrate cats. Arch. Itul. Biol., 125: 107-138. Svensson, T.H., Bunney, B.S. and Aghajanian, G.K. (1975) Inhibition of both noradrenergic and serotonergic neurons in brain by the a-adrenergic agonist clonidine. Bruin Res., 92: 291-306. Unnerstall, J.R., Kopajtic, T.A. and Kuhar, M.J. (1984) Distri-
bution of a,-agonist binding sites in the rat and human central nervous system: Analysis of some functional, anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agents. Bruin Res. Rec., 7: 69-101. Vanni-Mercier, G., Sakai, K., Lin, J.S. and Jouvet, M. (1989) Mapping of cholinoceptive brainstem structures responsible for the generation of paradoxical sleep in the cat. Arch. Itul. Biol., 127: 133-164. Vertes, R.P. (1984) Brainstem control of the events of REM sleep. Prog. Neurobiol., 22: 241-288. Vincent, S.R. and Reiner, P.B. (1987) The immunohistochemical localization of choline acetyltransferase in the cat brain. Bruin Res. Bull., 18: 371-415. Vizi, E.S. (1980) Modulation of cortical release of acetylcholine by noradrenaline released from nerves arising from the rat locus coeruleus. Neuroscience, 5: 2139-2144. Wamsley, J.K., Lewis, M.S., Young, W.S. and Kuhar, M.J. (1981) Autoradiographic localization of rnuscarinic cholinergic receptors in rat brainstern. J. Neurosci., 1: 176-191. Webster, H.H. and Jones, B.E. (1988) Neurotoxic lesions of the dorsolateral pontomesencephalic tegmentumcholinergic cell area in the cat. I1 Effects upon sleep-waking states. Bruin Res., 458: 285-302. Weight, F.F. and Salmoiraghi, G.C. (1966) Adrenergic responses of Renshaw cells. J. Phurmucol. Exp. Ther., 154: 391-397. Young, W.S. and Kuhar, M.J. (1979) Noradrenergic a 1 and a 2 receptors: Autoradiographic visualization. Eur. J. Phurmucol., 59: 3 17-3 19. Young, W.S. and Kuhar, M.J. (1980) Noradrenergic a 1 and a 2 receptors: Light microscopic autoradiographic localization. Proc. Nutl. Acud. Sci., USA, 77: 1696-1700.
C.D. Barnes and 0. Pompeiano (Eds.) Progre.ss in Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
463
CHAPTER 33
Noradrenergic agents into the cerebellar anterior vermis modify the gain of vestibulospinal reflexes in the cat P. Andre, P. d'Ascanio and 0. Pompeiano Department of Physiology and Biochernistiy, Unii5ersityof Pisa, Via S. Zeno, Pisa, Italy
The noradrenergic (NA) afferent projection to the cerebellar cortex, which originates mainly from the locus coeruleus (LO, may act on the target neurons by utilizing both a- and P-adrenoceptors. Experiments performed in decerebrate cats have shown that unilateral injection into the vermal cortex of the cerebellar anterior lobe of 0.25 p1 of the a,-adrenergic agonist metoxamine or the a,-agonist clonidine (at 2-8 p g / p l of saline) as well as of the non-selective P-agonist isoproterenol (at 8-16 p g / p I ) decreased the postural activity in the ipsilateral forelimb, while the extensor tonus either remained unmodified or slightly increased on the contralateral side. The same agents also increased the gain of the vestibulospinal (VS) reflexes elicited by recording the multiunit EMG responses of the ipsilateral and the contralateral triceps brachii to roll tilt of the animal (at 0.15 Hz, _+lo"), leading to sinusoidal stimulation of labyrinth receptors. The crossed effects were more prominent for the a*-than for the a,- and P-agonists. Only slight changes in the phase angle of the responses were observed. The effects described above appeared 5-10 min after the injection, reached the peak values after 15-30 min and disappeared within 2 h. The effective area was located within the third and/or the fourth folium of the culmen rostra1 to the fissura prima, 1.4-1.8 mm lateral to the midline. This area corresponded to zone B of the cerebellar cortex, which projects to the ipsilatera1 lateral vestibular nucleus (LVN), on which it exerts a prominent inhibitory influence. In fact, monopolar stimulation of this area with three negative pulses (at 300/sec) performed prior to the local injection inhibited the spontaneous EMG activity of the ipsilateral triceps brachii. The effects described above were dose-dependent; injection of an equal volume of saline was ineffective. All changes in posture and reflexes elicited by metoxamine or clonidine were impaired by previous injection into the same corticocerebellar area of the corresponding a,-o r a2adrenergic antagonist prazosin or yohimbine, respectively (0.25 pI at 8-16 p g / p I ) . However, cross-interactions between a,and a,-adrenergic agonists and antagonists were also ob-
served. In fact, injection of the a,-adrenergic antagonist yohimbine prevented the occurrence of all the metoxamine effects, while administration of the a,-adrenergic antagonist prazosin prevented the occurrence of the ipsilateral, but not of the contralateral effects induced by clonidine injection. In conclusion, while the ipsilateral effects induced by metoxamine and clonidine were almost equally depressed by the aland the a,-adrenergic antagonists, the contralateral effects were more prominently affected by the a z - than by the a,-antagonist. Injection in other experiments of 0.25-0.50 pl of the non-selective P-adrenergic antagonist propranolol not only increased the extensor tonus in the Ipsilateral limbs, while slightly reducing the extensor tonus in the contralateral limbs, but also decreased the response gain of the ipsilateral, and to a smaller extent of the contralateral, triceps brachii to animal tilt. The same injection also reduced or suppressed the postural and reflex changes induced by previous administration into the same corticocerebellar area of the non-selective P-adrenergic agonist isoproterenol. An additional finding was that the propranolol injection slightly modified, but did not prevent, the increase in the response gain of the ipsilateral triceps brachii to labyrinth stimulation following administration into the same corticocerebellar area of the a I - or the a,-adrenergic agonist metoxamine or clonidine, while the contralateral effects (particularly induced by clonidine) were suppressed. These findings suggest differences in the degree of colocalization of a- and P-adrenoceptors on the corticocerebellar neurons involved in the gain regulation of the ipsilateral and the contralateral triceps brachii to labyrinth stimulation. Since during animal tilt most of the Purkinje (PI-cells of the cerebellar vermis fire out-of-phase with respect to the related VS neurons, we postulated that both the a,- and a,-adrenergic agonists, as well as the P-adrenergic agonist, act on these P-cells by enhancing the amplitude of their modulation to labyrinth stimulation, thus exerting a positive influence on the gain of the VS reflexes.
Key words: vestibulospinal reflex, gain regulation, cerebellar vermis, noradrenergic agents, a,-, a2- and P-adrenoceptors
464
Introduction The cerebellar cortex receives not only the classical afferent systems terminating as mossy fibers (MF) and climbing fibers (CF), but also a third afferent system originating from norepinephrine (NE)-containing neurons located particularly in the locus coeruleus (LC; Bloom et al., 1971; 01son and Fuxe, 1971; Chu and Bloom, 1974; Pickel et al,, 1974; Pasquier et al., 1980; cf. Foote et al., 1983). This coeruleo-cerebellar projection, which is mainly ipsilateral, reaches the whole area of the cerebellar cortex including the cerebellar vermis (cf. Dietrichs, 1988). The early observation that the noradrenergic (NA) afferents to the cerebellar cortex terminate on Purkinje (PI-neurons (Fuxe, 1965; Hokfelt and Fuxe, 1969) has been confirmed by later studies showing that these fibers make synaptic contacts primarily on P-cell dendrites in the molecular layer and, to a lesser extent, on the P-cell body and superficial granule cell layers (Bloom et al., 1971; Olson and Fuxe, 1971; Chu and Bloom, 1974; Pickel et al., 1974; Landis and Bloom, 1975; Yamamoto et al., 1977; Kimoto et al., 1981; cf. Powers et al., 1989, in humans). Experiments of electrolytic lesion of the LC (d’Ascanio et al., 1985, 1989c) or functional inactivation of the corresponding neurons, induced by local microinjection either of the a,-adrenergic agonist clonidine (Pompeiano et al., 1987) or of the P-adrenergic agonist isoproterenol (d’Ascanio et al., 1989a1, have shown that this structure controls the dynamics of the vestibulospinal (VS) reflexes elicited by sinusoidal roll tilt of the animal; in particular, under given conditions, an increase or a decrease in the response gain of forelimb extensors to animal tilt was observed. These findings were attributed to involvement of direct (cf. Pompeiano, 1989) or indirect projections of the LC to the spinal cord (cf. d’Ascanio et al., 1989b). The demonstration, however, that the LC also sends NA afferents to the cerebellar vermis, whose P-cells projecting to the lateral vestibular nucleus (LVN) (Corvaja and Pom-
peiano, 1979; Trott and Armstrong, 1987; Voogd, 1989) respond to animal tilt (Denoth et al., 19791, raises the question whether this projection contributes to the gain regulation of the VS reflexes. The NA system may act on the cerebellar cortex through different types of adrenoceptors. In particular, radioligand binding studies have revealed the presence of both a,- and a2-(Bylund and U’Prichard, 1983) as well as P I - and P2adrenoceptors (Minneman et al., 1979, 1981; Lefkowitz et al., 1983; Pompeiano, M. et al., 1989) in the cerebellum of mammals. Moreover, in vitro autoradiography has shown a fairly high level of specific binding for a,-receptors in the molecular layer (Jones et al., 1985; Palacios et al., 1987) and a low level of specific binding for a,-receptors in the granule cell layer of the cerebellar cortex (Unnerstall et al., 1984; Bruning et al., 1987). P-adrenoceptors also appeared to be located in the molecular layer (Palacios and Kuhar, 1980, 1982; Rainbow et al., 1984; Lorton and Davis, 19871, as well as in “patches” over small groups of P-cells somata (Sutin and Minneman, 1985). Moreover, quantitative autoradiography revealed that the molecular layer had a higher concentration of P2- than of PI-receptors (Nahorski, 1978; Minneman et al., 1979, 1981; Rainbow et al., 1984). A limitation in these studies on the localization of receptors is whether the binding sites relate to cell bodies or fiber terminals of target neurons. In fact, by using in situ hybridization histochemistry, it was shown that in the cerebellar cortex of rat neurons expressing P-adrenoceptors (namely @,-receptors) corresponded to the granule cells (Palacios, 1988). The aim of our experiments was to investigate whether in decerebrate cats the dynamic characteristics of responses of forelimb extensors to sinusoidal stimulation of labyrinth receptors could be modified by local injection of specific a- and P-adrenergic agonists and antagonists into the vermal cortex of the cerebellar anterior lobe. The results of these experiments have been reported in detail in recent studies (Andre et al., 1991a,b).
465
Methods The experiments were performed in 25 cats, which were decerebrated at the precollicular level under ether anesthesia. The multiunit EMG activity of forelimb extensors, namely the medial head of the triceps brachii of both sides, was recorded during roll tilt of the animals at 0.15 Hz, +lo" (cf. Manzoni et al., 1983). Sequential pulse density histograms (SPDHs) were obtained by averaging data of 6 sweeps, each containing the responses to two successive cycles (128 bins, 0.1 sec bin width). These stimulation sequences were repeated at intervals of 4-10 min for several hours, before and after pressure microinjection of NA agonists or antagonists into the hemivermal cortex of the cerebellar anterior lobe of one side. The digital data of these averaged responses were processed on-line with a computer system (PET, 2001-8C), which performed a fast Fourier transform. In particular, the gain (expressed in imp./sec/deg) and the phase angle of the first harmonic component of the responses (expressed in degrees with respect to the peak of the sidedown displacement of the animal) were evaluated. The base frequency (BF, in imp./sec), which corresponded to the DC value obtained from the harmonic analysis of the responses, was also evaluated; this value was comparable to the mean frequency of the multiunit discharge recorded at rest. For the injection of drugs, a vertically oriented stainless steel cannula with the outer diameter of 200-300 pm, connected to a Hamilton 1 pl syringe, was lowered into the vermal cortex of the cerebellar anterior lobe passing through the third or the fourth folium rostra1 to the fissura prima (culmen) at 1.4-1.8 mm lateral to the midline and at the depth of about 4 mm below the surface. The a ,-adrenergic agonist (metoxamine, Sigma) or antagonist (prazosin, Pfeizer), the a,-adrenergic agonist (clonidine, Sigma) or antagonist (yohimbine, Sigma), and the non-selective padrenergic agonist ( f-isoproterenol hydrochloride, Sigma) or antagonist (DL-propranolol hydro-
chloride, Sigma), which act on both PI- and &receptors, were used during the experiments. We injected usually 0.25 pI solutions of a-adrenergic agonists (at 2-8 pg/pl) or antagonists (at 8-16 pg/pl) and a p-adrenergic agonist or antagonist (at 8-16 pg/pl), dissolved in sterile saline stained with pontamine sky blue (5%) as a marker. The pH of the solution was appropriately adjusted to reach the value of pH 7.4 -+ 0.2. In control experiments, a local injection of 0.25 p1 of saline was performed. The corticocerebellar area chosen for injection was first submitted to electrical stimulation with three cathodal pulses (at 300/sec, 0.2 msec in duration, 0.1-10 volts). These short trains were applied monopolarly at the repetitive rate of 0.52/sec by using stainless steel wire of 300 p m in size, electrolytically sharpened and completely insulated except at the tip. It was then possible to document whether the selected region was able to inhibit the spontaneous EMG activity of the ipsilateral triceps brachii. At the end of each experiment the localization of the tip of the cannula for each penetration, as well as the extent of the blue-stained tissue were identified on cerebellar frontal sections counterstained with neutral red.
Results Unilateral injection of 0.25 pl of the al- and a,-adrenergic agonists metoxamine and clonidine (at 4-16 pg/pl) as well as of the P-adrenergic agonist isoproterenol (at 8-16 pg/pl) into the vermal cortex of the cerebellar anterior lobe produced a postural asymmetry characterized by a reduced extensor tonus in the ipsilateral limbs, while the decerebrate rigidity either remained unmodified or slightly increased in the contralatera1 limbs. Just the opposite changes in posture were obtained in other experiments after unilateral injection into the same corticocerebellar area of 0.25-0.50 pl of the a,- and a,-adrenergic antagonists prazosin and yohimbine, and the padrenergic antagonist propranolol (at 8-16
466
pg/pl). These effects persisted for about 2 h after the injection, before disappearing.
Increase in the response gain of a limb extensor to animal tilt after unilateral injection of adrenergic agonists into the vermal cortex of the cerebellar anterior lobe Rotation about the longitudinal axis of the animal at 0.15 Hz, f lo" produced a sinusoidal modulation of the multiunit EMG activity of the triceps brachii, characterized by an increased activity during side-down tilt and a decreased activity during side-up tilt (cf. Schor and Miller, 1981; Manzoni et al., 1983). The peak of these responses was related to the extreme animal displacement, thus being attributed to stimulation of macular, utricular receptors; moreover, the response gain was rather small in the decerebrate preparations, as shown in previous experiments (Manzoni et al., 1983). The EMG responses of the triceps brachii of both sides to animal tilt were averaged and recorded at regular intervals before and after unilateral injection of different adrenergic agonists into the vermal cortex of the cerebellar anterior lobe. In order to compensate for the decrease in spontaneous EMG activity following these injections, the ipsilateral forelimb was passively flexed to increase the static stretch of the muscle; the opposite was performed on the contralateral side if the spontaneous EMG activity slightly increased.
Injection of the a,-adrenergic agonist, metoxamine Unilateral injection of metoxamine into the vermal cortex of the cerebellar anterior lobe increased the amplitude of the EMG modulation and thus the response gain of the ipsilateral forelimb extensor to labyrinth stimulation (Fig. 1A). Similar results were also obtained in some instances on the contralateral side. However, only slight changes in the phase angle of the responses were observed.
Table 1A documents the average BF, gain and phase angle values of the first harmonic of responses of the ipsilateral and the contralateral triceps brachii to animal tilt in two representative experiments (Exps. 7 and 81, before and after individual injections of 0.25 pl of metoxamine (at 4-8 pg/pl) into the anterior hemivermis of one side. In these experiments, in which the BF was only slightly modified with respect to the control values, the mean gain of the averaged responses elicited after metoxamine injection increased significantly to 180.3% of the control value on the ipsilateral side (t-test, P < O.OOl), while a smaller increase to 137.4% of the control value occurred on the contralateral side (t-test, N.S.). Although the increase in the response gain was statistically significant only in Exp 8 (t-test, P < 0.001, for gain values of both the ipsilateral and the contralateral responses), an important finding was obtained in Exp. 7, i.e., that the proportion of averaged records which showed responses to animal tilt increased after injection of metoxamine from 4/11 (i.e., 36.4%) in the control situation to 14/17 (i.e., 82.4%) in the ipsilateral limb and from 6/12 (i.e., 50.0%) in the control situation to 15/16 (i.e., 93.8%) in the contralateral limb. In this experiment, therefore, the a,-adrenergic agonist increased the probability of eliciting EMG responses to labyrinth stimulation. As to the phase angle of the averaged responses recorded both ipsilaterally and contralaterally to the side of the metoxamine injections, it appears that the corresponding values did not significantly change with respect to the controls (Table 1A). The increased gain of the EMG responses of the ipsilateral triceps brachii to animal tilt described above was first observed 5-10 min after injection of the a,-adrenergic agonist and progressively increased to reach the highest values after 20 min; it was then followed for 1.5-2 h after the injection before disappearing. The gain changes of the EMG responses of the contralatera1 triceps brachii to animal tilt, if present (Exp. 8), showed a similar time course.
467
Injection of the a,-adrenergic agonist, clonidine Unilateral administration of clonidine into the anterior vermis greatly increased the amplitude of modulation and thus the response gain of the ipsilateral (Fig. 1B) as well as the contralateral triceps brachii to labyrinth stimulation. Moreover, only slight changes in phase angle of the responses were observed.
SIDE
DOWN
SIDE DOWN
10
SIDE UP
Table 1B illustrates the results of two experiments (Exps. 1 and 2) showing that individual injections of 0.25 pl of clonidine (at 2-4 p g / p l ) into the vermal cortex of the cerebellar anterior lobe on one side increased significantly the mean gain of the responses of the ipsilateral as well as of the contralateral triceps brachii to labyrinth stimulation to 171.2% and 190.4% of the control
1
2 sec
SIDE UP
2 sec
Fig. 1. Increase of the response gain of the triceps brachii to animal tilt after injection of an cu,-(A), cu,-(B) and p-(C) adrenergic agonist into the ipsilateral vermal cortex of the cerebellar anterior lobe. Precollicular, decerebrate'cats (Exps. 2, 8 and 16). SPDHs showing the averaged multiunit responses of the triceps brachii of one side to animal tilt at 0.15 Hz, If- lo". Each record is the average of 6 sweeps, using 128 bins with 0.1 sec bin width. The lower trace in each column indicates the animal displacement; "side-down" indicates the displacement of the animal towards side of recording. The traces on the left side were taken before, while those on the right side after individual (A,B) or multiple (three in C) injections of 0.25 pI of the following solutions: methoxamine (4 p g / p I of saline) in A (Exp. 8), clonidine (4 p g / p l of saline) in B (Exp. 21, and isoproterenol (16 p g / p l of saline) in C (Exp. 16). The cannula passed through the second or third folium rostra1 to the fissura prima, 1.4-1.6 mm lateral to the midline (L 1.4-1.6) and reached the depth of 2-4 mm below the surface (H - 2 to -4). In these experiments the response gain increased, on the average, from 0.61 to 2.08 imp./sec/deg in A, from 0.60 to 1.94 imp./sec/deg in B and from 1.24 to 2.72 imp./sec/deg. (From Andre et al., 1991a,b.)
TABLE 1 Dynamic characteristics of the averaged multiunit responses of the ipsilateral and the contralateral triceps brachii to animal tilt (at 0.15 Hz, k 103 elicited in different experiments before and after unilateral ifijections into the vermal cortex of the cerebellar anterior lobe of an a,-(A), a2-(B), P-adrenergic agonist (C) or a P-adrenergic antagonist (D) Groups of averaged records
Base frequency
responsive
(imp./sec)
Side of recording
Time of recording relative to injection
Number of experiments
A. Ipsi.
Before After Before After
Exps. 7, 8 Exps. 7, 8 Exps. 7, 8 Exps. 7, 8
15 22 18 23
(65.2%) (88.0%) (75.0%) (95.8%)
8(34.8%) 3 (12.0%) 6 (25.0%) 1(4.2%)
Before After Before After
Exps. Exps. Exps. Exps.
1, 2 1, 2 1, 2 1, 2
11 20 16 21
(45.8%) (66.7%) (61.5%) (95.5%)
13 (54.2%) 10 (33.3%) 10 (38.5%) 1 (4.5%)
Before After Before After
Exps. 8, 9, Exps. 8, 9, Exps. 8, 9, Exps. 8, 9,
93 120 60 67
(89.4%) (92.3%) (93.8%) (87.0%)
11 (10.6%) 10 (7.7%) 4 (6.2%) 10 (13.0%)
Before After Before
Exps. 1, 6, 12, 13 Exps. 1, 6, 12, 13 Exps. 1, 6, 12, 13
47 (100%) 65 (91.5%) 41 (83.7%)
-
After
Exps. 1, 6, 12, 13
46 (86.8%)
7 (13.2%)
Contra. B. Ipsi. Contra.
C. Ipsi. Contra. D. Ipsi.
Contra.
14, 15, 16, 20 14, 15, 16, 20 14, 20 14, 20
unresponsive
6 (8.5%) 8 (16.3%)
< ''05 N'S' N'S'
NS' N'S' N'S' NS
132.3f 26.1 154.9f 32.0 144.7f 17.3 150.4 f 22.7
Gain (imp./sec/deg)
< '.Ool N,S,
0.71 k0.18 1.28f0.56 1.15f0.44 1.58f0.82
0.59 f0.23 1.01k 0.38 0.83 f0.31 < 0'001 1.58f0.56
142.3f 49.5 150.9f53.2 160.8f42.2 171.5f44.9
< '.Oo1 < 0.02
P < 0.01
1.25 f 0.64 1.68 f0.85 1.53k 0.63 1.85 f0.99
N,S,
+ 32.3 f22.1 +17.5+11.7 + 16.3k 14.4 p
N,S,
< O.OO1 < 0'02 N.S.
0.82 f0.48
- 3.6 f 19.5
+ 13.8f 19.5 + 7.8 f 15.3 + 6.3 f 14.5
p<0'05
1.80 f0.63
< 0'001 1.31f0.68 0.97 k 0.54 N.S.
125.2f 23.7
N.S.
< 0,01
137.1k 34.0 153.7f20.1 150.7f 38.1 134.1 f 35.4
160.5f 33.7 149.6f31.9 141.3f24.5
Phage angle relative to animal position (deg)
+ 13.6f 15.1 + 14.7f 16.5 + 13.4+ 10.7 + 19.6f + 16.1f + 10.4f
13.8
9.0 14.1 + 10.0+ 14.7 ( n = 38) + 15.3&16.2
Individual or multiple doses of 0.25-0.50 p1 of the following solutions of (A) metoxamine at 4-8 p g / p I , (B) clonidine at 2-4 p g / p I , (C) isoproterenol at 8-16 p g / p I and (D) propranolol at 8-16 p g / p I of saline, were injected into the hemivermal cortex of the cerebellar anterior lobe of one side, namely within the second or third folium rostra1 to the fissura prima. The base frequency, gain and phase angle values (mean fS.D.) refer to the groups of averaged records showing coherent responses to tilt. The phase angle of the responses is expressed in deg with respect to the extreme side-down animal displacement (0"); plus indicates lead, minus indicates lag of the responses. Student's t-test was used for statistical calculations. Significant differences between averaged values are indicated on the table; N.S., difference between mean values is not significant ( P > 0.05). Experiments numbered in A and B are different from those in C and D.
469
values, respectively (t-test, P < 0.01 and P < 0.001, in the two instances). In these experiments the BF remained comparable to the control Values. Moreover, the proportion of averaged records which showed responses to animal tilt increased after injection of clonidine, both ipsilaterally and contralaterally to the side of the injection. In conclusion, injection into the cerebellar vermis of the a,-adrenergic agonist clonidine increased the response gain of the ipsilateral as well as of the contralateral triceps brachii to Cloni d ine I
I
animal tilt, the crossed effects being more prominent than those reported in the previous section after injection of the a,-adrenergic agonist metoxamine. As to the phase angle of the averaged responses to animal tilt, it appeared that the mean value of the responses was only slightly modified by the clonidine solution, both ipsilaterally and contralaterally to the side of the injection (Table 1B). The time course of the gain changes of the EMG responses of the triceps brachii of both
Yohimbine Clonidine I t
nm
:
1.5-
h
(3, (1,
'D
. .
1
1.0-
u
aJ UI
\
.. .
a
-E
0.5-
.. .. 60
0
60
.... . .- ...
120
180
240 min
180
240 min
B
h
I
t
-
t
+goo
(3,
a, -0
v
a,
0-
(3,
C
0
$ -90"0 I:
60
0
60
120
Fig. 2. Time course of gain changes of the responses of the ipsilateral triceps brachii to animal tilt induced by local injection into the vermal cortex of the cerebellar anterior lobe of the a,-adrenergic agonist clonidine, before and after local administration of the corresponding antagonist yohimbine. Precollicular, decerebrate cat (Exp. 2). The changes in gain (A) and phase angle (B) of the averaged responses of the right triceps brachii to animal tilt at 0.15 Hz, f lo" induced by local injection into the right hemivermis of 0.25 pl of a clonidine solution (4 pg/pI of saline) were evaluated before and after administration into the same corticocerebellar area of 0.25 pI of a yohimbine solution (8 pg/pI of saline). For histology see Figure 5. A. The first injection of clonidine (I) enhanced for about one h the response gain of the ipsilateral right triceps brachii to animal tilt. However, after injection of yohimbine (II), the increase in the response gain induced by a successive administration in the same cortical area of clonidine (111) was negligible; moreover, many traces showed lack of responses. B. Only slight changes in the phase angle of the responses were observed. (From Andre et al., 1991a.)
470
mained almost unmodified in these experiments. Table 1C indicates the results of six experiments (Exps. 8, 9, 14, 15, 16 and 20) showing that individual or multiple administration of isoproterenol(O.25 pl at 8-16 pg/pl) into the hemiverma1 cortex of the cerebellar anterior lobe slightly but significantly increased the mean gain of the EMG responses of the ipsilateral and, to a lesser extent also, of the contralateral triceps brachii to 134.4% and to 120.9% of the control values, respectively (t-test, P < 0.001 and P < 0.02 in the two instances). On the other hand, the BF remained almost unmodified on both sides. It is worth noticing that among the experi-
sides to animal tilt following unilateral injection of clonidine into the anterior vermis was similar to that described for the metoxamine injections (see Fig. 2A,I, for the ipsilateral responses recorded in Exp. 2). Injection of the P-adrenergic agonist, isoproterenol Unilateral injection of isoproterenol into the vermal cortex of the cerebellar anterior lobe increased the amplitude of the EMG modulation and thus the response gain of the ipsilateral (Fig. 1 0 , and in some instances also of the contralateral, triceps brachii to labyrinth stimulation. Moreover, the phase angle of the responses re-
-? 20
1sop.oterenoi
1
1soprotereno1
I
I
I
I1
1
1
I
I1
3 0.
. . ... . .
2 0.
.. ..
~. . .. .
. . . .. .. .' . ..
-
- . .
0.
0
1
60
*
10-
I
60
-
120
"71"
1
I
I
+ 9c
....
0
- ..... ....
.... ;*,..- -........
:
*
- 90
60
0
60
170
min
120
60
0
60
120 min
Fig. 3. Increase of the response gain of the triceps brachii to animal tilt after injection of isoproterenol into the ipsilateral vermal cortex of the cerebellar anterior lobe. Precollicular decerebrate cats (A, Exp. 14 and B, Exp. 16). The gain (upper diagrams) and the phase angle (lower diagrams) of the averaged responses of the triceps brachii of one side to animal tilt at 0.15 Hz, f lo" were evaluated before and after individual (A) or multiple injections (B) of 0.25 pl of an isoproterenol solution (16 p g / p l of saline) into the ipsilateral hemivermis of the anterior lobe. In A the first injection of isoproterenol (1) into the left hemivermis at the level of the third folium rostral to the fissura prima, L 1.2, H -4 was ineffective. However, the second injection (111, within the same folium and at the same depth but at L 1.5, enhanced the response gain of the ipsilateral triceps brachii to animal tilt. For histology see Figure 5. In B, a slight increase of the response gain of the ipsilateral triceps brachii to animal tilt occurred after 3 injections of isoproterenol into the right hemivermis between the third and the fourth folium rostral to the fissura prima, L 1.4, H - 3 to - 5 (I). A more prominent increase of the effects occurred after three successive injections of isoproterenol within the same folium and at the same depths, but at L 1.8 (11). In both the experiments the isoproterenol injections did not modify the phase angle of the responses. (From Andre er al., 1991b.I
47 1
ments referred to in Table 1C, the average gain of the responses recorded from the ipsilateral triceps brachii increased significantly in three experiments (Exps. 8, 14 and 16) but not in others (Exps. 9, 15 and 20). The reason for these differences will become clear in the next section, in which the localization of the injection sites will be discussed in detail. The changes in gain observed in the first group of experiments were also associated with minor changes in the phase angle of the averaged responses which decreased on the ipsilateral side after isoproterenol injection (from an average lead of + 11.7 k 11.2, S.D. deg to +6.5 6.7, S.D., deg; t-test, P < O.OOl), but increased on the contralateral side (from an average lead of + 15.5 _+ 10.2, S.D. deg to +21.4 _+ 7.4, S.D. deg; t-test, P < 0.001). The increased gain of the EMG responses of the ipsilateral triceps brachii to animal tilt described in the first group of experiments appeared 5-10 min after injection of the P-adrenergic agonist and reached the highest value after 20-30 min; the increased gain remained for about 2 h after the injection before returning to normal. A similar time course was also observed in those instances in which the response gain of the contralateral triceps brachii to animal tilt increased. Figure 3 illustrates in two experiments the time course of the changes in gain (upper diagrams) and phase angle values of the responses (lower diagrams) recorded ipsilaterally to the side of isoproterenol injections. In A (Exp. 14) the first injection performed into the hemivermal cortex of the culmen 1.2 mm lateral to the midline was ineffective (I), while the second injection made 0.3 mm lateral with respect to the first one proved effective. On the other hand, in B (Exp. 16) the response gain showed a slight increase soon after the first injection performed in the same corticocerebeliar area, but 1.4 mm lateral to the midline (I), and further increased after the successive injection made 0.4 mm more laterally (11). In both instances, the phase angle of the responses remained stable throughout the experiments.
Site-specificity and dose-dependence of the responses The sites, which after injection of the adrenergic agonists not only decreased the postural activity in the ipsilateral limbs but also increased the gain of the multiunit EMG responses of the ipsilateral and to some extent also of the contralatera1 triceps brachii to labyrinth stimulation, were located within the lateral part of the vermal cortex of the culmen, particularly at the level of the second and third folium (or the third and fourth folium) rostra1 to the fissura prima, 1.4-1.8 mm lateral to the midline and at the depth of 3-4 mm below the surface. Histological controls made at the end of the experiments showed that injection of 0.25 ,ul of a solution of adrenergic agonists, stained with pontamine 5%, extended for 1-2 mm along the longitudinal axis of the cannula and for about 200-300 p m medially and laterally to the axis of penetration. This area corresponded to the longitudinal zone B of the cerebellar cortex which projects to the ipsilateral LVN (Corvaja and Pompeiano, 1979; Trott and Armstrong, 1987; Voogd, 1989) and shows antidromic field potentials following single shock stimulation of the underlying LVN (Denoth et al., 1979). Moreover, repetitive electrical stimulation of the same corticocerebellar area with cathodal pulses, performed prior to the injection, inhibited the spontaneous EMG activity of the ipsilateral triceps brachii (see Fig. 4A), a finding which can be attributed to suppression of the tonic facilitatory influence that the LVN exerts on posture (cf. Pompeiano, 1967; Ito, 1984). The site-specificity of the effects was documented by the fact that injection of an equal dose of adrenergic agonist through a cannula which penetrated within the second folium located caudally to the fissura prima, at the same IateraIity and depth indicated above, was ineffective. In this case repetitive stimulation of this corticocerebellar area with the same parameters which applied in the effective areas inhibited the postural activity in the ipsilateral limbs did not produce any effect, as illustrated in Figure 4B.
472
The critical parameters which determine the effectiveness of the injection of the adrenergic agonists into the cerebellar vermis are represented by the localization of the tip of the cannula not only in a rostro-caudal but also in a medio-lateral direction. In several experiments, in fact, injections of adrenergic agonists into the second or third folium rostral to the fissura prima, but 0.2-0.5 mm more laterally or medially with respect to the effective parameters (as for instance in Exp. 9, 15 and 20) did not produce the prominent postural and reflex changes described
above. Figure 5 illustrates the localization of the injection sites in two representative experiments. In the first case (Exp. 21, the effective injection of clonidine was located unilaterally into the vermal cortex of the cerebellar anterior lobe at the effective laterality. In the second case (Exp. 141, the first dose of isoproterenol injected in the left hemivermis at the level of the third folium rostral to the fissura prima, 1.2 mm lateral to the midline and at the depth of 4 mm below the surface (I), was ineffective. However, injection of the same dose of isoproterenol within the same folium and
B
A
Imp./sec
Imp./sec
2.OT
400,
2.0 T
400,
3.5 T
-
-
7
20 msec Fig. 4. Effects of electrical stimulation of the cerebellar vermal cortex on postural activity recorded from the ipsilateral triceps brachii. Precollicular, decerebrate cat (Exp. 5). Monopolar stimulation of the left hemivermis (culmen) with a stainless steel wire of 300 pm, electrolytically sharpened and completely insulated except at the tip (25 p m in size). A. SPDHs showing the averaged responses of the left triceps brachii to stimulation of the ipsilateral effective corticocerebellar area. The tip of the electrode was located within the second folium rostral to the fissura prima, L 1.4, H -3. B. SPDHs showing the averaged activity of the same triceps brachii following stimulation of an ipsilateral corticocerebellar area corresponding to the second folium caudal to the fissura prima, L 1.4, H -3. In both instances the stimulation consisted of 3 negative pulses at 300/sec, 0.2 msec in duration, 2.0 and 3.5 times the threshold (TI for inhibition of the spontaneous EMG activity, as evaluated in the effective area (A); the duration of the stimulus is indicated by horizontal bars. Each record is the average of 30 sweeps, obtained at the repetition rate of 2/sec. For each sweep 128 bins with 2 msec bin width were used; the terminal part of the averaged records was deleted. (From Andre et al., 1991a.I
473 Clonidine
Exp. 2
lsoproterenol
Exp.14
Fig. 5 . Schematic representation of frontal sections of the cerebellar cortex of the anterior lobe showing the localization of the sites of injection of noradrenergic agents within the third folium rostra1 to the fissura prima, 3-4 mm below the surface and at the laterality indicated below. In both experiments the shaded areas indicate the localization of the injected solution stained with 5% pontamine. In Exp. 2 a first dose (I) of 0.25 pl of clonidine solution (4 p g / p l of saline) was injected 1.4 mm lateral to the midline; a second dose (11) of 0.25 pl of yohimbine (8 p g / p l ) followed by a third dose (111) of clonidine similar to the first were also injected into the same spot. In Exp. 14 a first dose of 0.25 p1 of isoproterenol solution (16 p g / p l of saline), injected 1.2 m m lateral to the midline, was virtually ineffective, while a second dose (11) similar to the first one, injected at the laterality of 1.5 mm, was quite effective. L, left side; R, right side. (From Andre et al., 1991a,b.)
at the same depth but at 1.5 mm lateral to the midline (11) increased the response gain of the ipsilateral triceps brachii to animal tilt. Details of this experiment were illustrated in the previous section (see Fig. 3A, I and 11). The effects induced by local injection of adrenergic agonists into the vermal cortex of the cerebellar anterior lobe were not due to irritative phenomena following injection of the fluid, since neither changes in posture nor in the response gain of the triceps brachii to labyrinth stimulation were observed following injection of an equal volume of saline into the same corticocerebellar area, prior to the effective injections. Therefore, the effects described above depended on the adrenergic agent. Evidence was also presented indicating that the increase of the response gain induced by local injection of metoxamine or clonidine was dose-dependent (Exps. 1 and 10).
Decrease in the response gain of a limb extensor to animal tilt after unilateral injection of adrenergic antagonists into the vermal cortex of the cerebellar anterior lobe Since injection of adrenergic agonists into the vermal cortex of the cerebellar anterior lobe increased the gain of the ipsilateral and to some extent also of the contralateral triceps brachii to labyrinth stimulation, we expected that just the opposite effects should have been elicited in other experiments after injection of adrenergic antagonists. This approach, however, was limited by the fact that in decerebrate cats the gain of the VS reflexes acting on forelimb extensors was quite small in amplitude (cf. Manzoni et al., 1983), SO that injection of adrenergic antagonists into the cerebellar vermis of the anterior lobe should have produced a barely detectable depression of the
474
response gain in the triceps brachii. We decided, therefore: (i) to select experiments in which the response gain of the triceps brachii to animal tilt (at 0.15 Hz, +loo) was more prominent than usual in the control situation, and (ii) to investigate systematically only the effects of a non-selective P-adrenergic antagonist. As reported at the beginning of the Results, injection of 0.25-0.50 p l of a propranolol solution (at 8-16 pg/pl) produced a postural asymmetry characterized by an increased extensor tonus in the ipsilateral limbs, while the decerebrate rigidity either remained unmodified or slightly decreased in the contralateral limbs. Individual or multiple injections of the padrenergic antagonist propranolol into the hemivermal cortex of the cerebellar anterior lobe of one side decreased the amplitude of modulation and thus the response gain of the ipsilateral forelimb extensor to labyrinth stimulation. There was also an increase in number of the unresponsive averaged traces with respect to the control. Only in one out of four experiments were similar results obtained on the contralateral side. The reduced gain of the ipsilateral responses was observed even in the absence of any significant change in the BF which was maintained constant due to appropriate changes in limb position.
- iOr
Propronolol t I
A
P
. Li
20-
D
. a
-E
10-
0
I
II
-
.. ..
0-
. .. .
- .
. . . . .. ..
.. .. . ...
1 .
t
Moreover, only slight changes in the phase angle of the responses were observed. Figure 6 shows that in one experiment (Exp. 12) two injections of 0.50 pl of a propranolol solution (at 16 pg/pl) decreased the response gain of the ipsilateral triceps brachii to animal tilt from the average value of 1.93 imp./sec/deg in the control situation to 0.67 imp./sec/deg, while the phase angle of the responses remained almost unmodified. Table 1D illustrates the results of four experiments (Exps. 1, 6, 12 and 13) showing that individual or multiple injections of propranolol (0.25-0.50 pl at 8-16 pg/p1) in the hemivermal cortex of the cerebellar anterior lobe slightly but significantly decreased the mean gain of the responses of the ipsilateral triceps brachii to animal tilt to 72.8% of the control value (t-test, P < 0.001), while the BF remained almost unmodified. It is of interest that the higher the response gain obtained in the control situation, the more prominent was the decrease of the corresponding value following propranolol injections. Moreover, the phase lead of the responses slightly decreased ipsilaterally to the side of the injection. No significant change in the mean gain and phase angle of the responses was obtained contralaterally to the side of the injection. The reduced gain of the EMG responses of
*
.. ..
B
--
I
m
D
0
*
I
t90"-
0-
m
. .. .. -*'
*
. *.-.. &.. .. *. . .:. ... :.. . .
'..a''
0
-90'1 0
a
Fig. 6 . Decrease of the response gain of the triceps brachii to animal tilt after injection of propranolol into the ipsilateral vermal cortex of the cerebellar anterior lobe. Precollicular decerebrate cat (Exp. 12). The gain (A) and the phase angle (B) of the averaged responses of the right triceps brachii to animal tilt at 0.15 Hz, k lo" were evaluated before and after injections of propranolol into the ipsilateral hemivermis of the anterior lobe. In particular, injections of 0.50 pl of a propranolol solution (16 y g / p l of saline) into the right hemivermis at the level of the third folium rostra1 to the fissura prima, L 1.4 mm (I) and 1.8 mm (111, H - 4 decreased the response gain of the ipsilateral triceps brachii to animal tilt, but did not modify the phase angle of the responses. (From Andre et af., 1991b.)
475
the triceps brachii of one side to animal tilt after propranolol injection (Fig. 6) followed the same time course of the increased gain induced in other experiments by local injection of isoproterenol into the same corticocerebellar area.
Effects of unilateral injection of adrenergic antagonists on the postural and reflex changes induced by administration into the cermal cortex of the cerebellar anterior lobe of a giuen adrenergic agonist Unilateral injection into the vermal cortex of the cerebellar anterior lobe of 0.25 pl of the a l - or a,-adrenergic antagonist prazosin or yohimbine (at 8-16 pg/pI) did not greatly modify the increase in the response gain of the ipsilateral and the contralateral triceps brachii to labyrinth stimulation induced by previous injection into the same corticocerebellar area of 0.25 pl of the corresponding a or a2-adrenergic agonist metoxamine or clonidine (at 4-8 pg/pl), respectively. These negative findings were obtained when the antagonist was injected shortly after the corresponding agonist had produced the maximum increase in gain. However, if the a I - or the a,-adrenergic antagonist was injected 2-3 h after
,-
I
-
D
administration of the corresponding agonist, it prevented the occurrence of the gain changes induced by a successive injection of the selective agonist. Thus, the effects of metoxamine were suppressed by a previous injection of prazosin (Exp. 7), while the effects of clonidine were prevented by a previous administration of yohimbine (Exp. 2). Figure 2A, I1 and 111 illustrates in detail the results obtained in the latter experiment. Cross interactions between a l - and a,-adrenergic agonists and antagonists were also investigated. It was found, in particular, that the azadrenergic antagonist yohimbine suppressed both the ipsilateral and, if present, also the contralateral effects induced by metoxamine (Exp. 101, while the a,-adrenergic antagonist prazosin suppressed only the ipsilateral but not the contralateral effects elicited by clonidine (Exp. 5). These findings indicate that the a,-adrenergic antagonist prazosin was more effective on the ipsilateral than on the contralateral responses, just as shown for the effects induced by the corresponding a l agonist (see Table 1A), while the q a d r e n e r g i c antagonist yohimbine was almost equally effective on the limb extensors of both sides, in analogy with the results elicited by the a,-adrenergic agonist (see Table 1B).
B
40 -
8 . " . m
.
30-
L%
20-
z 0
OL
I
I
. -
,
.
. ..
* :*
* t 2
.. . .-.. 60
0
I
I
-90"
8 a
60
120
180
240rnin
60
0
60
120
180
24Ornin
Fig. 7. Effects of local injection of the P-adrenergic antagonist propranolol into the vermal cortex of the cerebellar anterior lobe of one side on the metoxamine-induced increase of the response gain of the ipsilateral triceps brachii to animal tilt. Precollicular decerebrate cat (Exp. 10). The changes in gain (A) and phase angle (B) of the averaged responses of the left triceps brachii to animal tilt at 0.15 Hz, f lo" induced by two local injections (11 and 111) into the left hemivermis of 0.25 pl of a metoxamine solution (8 pg/pl of saline) were evaluated after injection into the same corticocerebellar area of 0.25 p1 of a propranolol solution (8 pg/pl of saline). The tip of the cannula was located within the third foliurn rostra1 to the fissura prima, L 1.4, H - 4. A. injection of propranolol (I), which slightly increased the number of unresponsive traces of the left triceps brachii to animal tilt with respect to the control, did not prevent the occurrence of an increased gain of responses of the same muscle to tilt following administrations of metoxamine. B. The phase angle of the responses remained stable throughout the experiment. (From Andre et al., 1991a.)
476
viously enhanced by the isoproterenol injection (Exps. 8 and 9). On the other hand, the gain of the responses recorded contralaterally to the side of the injection continued to increase after the propranolol injection, probably as a result of the previous administration of isoproterenol, and started to decrease later. The effects of propranolol injection described above occurred only when the P-antagonist was injected at least 1.5 h after the isoproterenol injection. Individual or multiple (three) injections
Following the experimental procedures reported above, attempts were also made to discover whether the changes in gain of the VS reflexes induced by unilateral injection of the P-adrenergic agonist isoproterenol into the cerebellar anterior vermis were modified by successive administration into the same corticocerebellar area of the P-antagonist propranolol. Injections of 0.25-0.50 p1 of propranolol (at 8-16 pg/pI) reduced the response gain of the ipsilatera1 triceps brachii to animal tilt, which was pre-
PropranololC'0"ip ine c , o ~ i ne I
-
I
40
II
m
..
U
aJ
. ..
. . .. . .. .. .
v)
.
10
.
..
.
.-C
0 0
0
*
*
*
*
*
*
:
* 1
60
0
60
I
I
120
180
240 min
6
I
I
. . aJ
:: -90" c
Q
t
..
1
60
0
60
120
180
240 min
Fig. 8. Effects of local injection of the P-adrenergic antagonist propranolol into the vermal cortex of the cerebellar anterior lobe of one side on the clonidine-induced increase of the response gain of the ipsilateral triceps brachii to animal tilt. Precollicular decerebrate cat (Exp. 6). The changes in gain (A) and phase angle (B) of the averaged responses of the right triceps brachii to animal tilt at 0.15 Hz, rfr lo", induced by two local injections into the right hemivermis of 0.25 yl of a clonidine solution at 4 y g / y l (11) and 16 y g / p l (111) of saline were evaluated after injection into the same corticocerebellar area of 0.25 yl of a propranolol solution (8 y g / y l of saline). The tip of the cannula was located within the third folium rostra1 to the fissura prima, L 1.4 mm, H -4 mm. A. Injection of propranolol (I) slightly decreased the response gain of the right triceps brachii to animal tilt, but did not prevent the occurrence of some increase in the response gain of the same muscle to tilt following administrations of clonidine. B. Notice the relative stability of the phase angle of the responses. (From Andre et a/., 1991a.)
of 0.25 pl of propranolol performed only 30-60 min after administration in the effective area of 0.25 pl of isoproterenol at the same concentration (16 pg/p1) did not prevent the increase in gain of both the ipsilateral and the contralateral responses induced by the previous administration of this P-adrenergic agonist (Exps. 14 and 20). The selectivity of the results described above is documented by the fact that unilateral injection into the cerebellar vermis of 0.25 pl of the padrenergic antagonist propranolol (at 8 pg/pl) did not prevent the increase in the response gain of the ipsilateral triceps brachii following successive administrations into the same corticocerebellar area either of metoxamine (two injections of 0.25 pl of a solution at 8 pg/pl in Exp. 10; Fig. 7) or of clonidine (two injections of 0.25 pI of solution at 4 and 16 p g / p l in Exp. 6; Fig. 8). However, the typical time course of the gain changes which affect the ipsilateral triceps brachii to labyrinth stimulation after clonidine injection, characterized by a sudden increase of the response gain followed by a slow decay (as shown in Fig. 2A, I), was now substituted by a slow and progressive increase in the amplitude of modulation (Fig. 8A, I1 and 111). Moreover, the increase in the response gain of the contralateral triceps brachii to labyrinth stimulation, which usually occurred after clonidine injection, was not observed. This finding can be attributed to crossed blocking influences, exerted by the propranolol injection, which were not clearly seen in normal decerebrate cats (see Table lD), but appeared only after tonic activation of the q a d r e n e r g i c system by clonidine, which leads to a prominent increase in the response gain not only of the ipsilateral but also of the contralateral triceps brachii to labyrinth stimulation. It is of interest that in the two experiments using propranolol pretreatment (Exps. 6 and 10) the phase lead of the responses recorded before and after injection of the al- and the a,-adrenergic agonists remained, on the average, unmodified with respect to the control values (Exps. 1, 2 and 8) in the ipsilateral limbs but significantly increased in the
contralateral limbs (t-test, P < 0.001 for both groups of responses recorded before and after administration of the a-agonists). Discussion
Effects of local injection of adrenergic agonists in the cerebellar vermis The present experiments have shown that in decerebrate cats microinjection into the vermal cortex of the cerebellar anterior lobe of one side of either a,- and a,-adrenergic agonists or a non-selective P-adrenergic agonist decreased the postural activity in the ipsilateral limbs while that of the contralateral limbs either remained unmodified or slightly increased. The same agents also enhanced the gain of the multiunit EMG responses to the ipsilateral and in some instances also of the contralateral triceps brachii to roll tilt of the animal, leading to stimulation of labyrinth receptors. There were only slight changes in the phase angle of the responses, which remained always related to the extreme animal position, as expected if the responses were mainly due to stimulation of macular receptors. A comparison of the responses obtained in different groups of experiments indicates that both the a l - (metoxamine) and the a,-adrenergic agonist (clonidine) produced ipsilateral effects. As to the contralateral effects, they were more prominent after injection of clonidine than of metoxamine. Similar to the effects induced by clonidine, the p-adrenergic agonist isoproterenol produced ipsilateral as well as contralateral effects; however, the increase in the response gain was smaller than that induced by clonidine. The increase in the response gain of the forelimb extensors to animal tilt started 5-10 min after the injection, reached the peak within 15-30 min and slowly decreased to disappear within 2 h after the injections. The corticocerebellar area, which after injection of the adrenergic agonists elicited the effects described above, corresponded to a parasagittal zone of the cerebellar anterior vermis located
478
1.4-1.8 mm lateral to the midline, within the second or third folium of the culmen rostra1 to the fissura prima. This area, which belongs to the forelimb region of the anterior vermis (cf. Pompeiano, 1967; Ito, 19841, corresponds to zone B of the cerebellar cortex which projects to the ipsilateral LVN, as shown in both experimental anatomical (Corvaja and Pompeiano, 1979; Trott and Armstrong, 1987; Voogd, 1989) and physiological observations (Denoth et al., 1979), and exerts a direct inhibitory influence on the corresponding VS neurons (cf. Pompeiano, 1967; Ito, 1984). This conclusion is supported by the fact that repetitive electrical stimulation of this corticocerebellar area, performed prior to the injection, suppressed the postural activity in the ipsilateral triceps brachii. Neither changes in posture nor in gain of the VS reflexes were obtained after injection of an equal dose of adrenergic agonists within the second folium located caudally to the fissura prima; moreover, only slight or negligible effects were obtained when the injection site was located either medially or laterally to the effective area. The site-specificity of the responses was also documented by the fact that after injection of a 0.25 pl solution of adrenergic agents, the pontamine sky blue dye, which was used to stain the injected solution, spread in the transverse direction for about 400-600 pm, as shown in histological frontal sections of the cerebellum counterstained with neutral red. Specificity of the effects As reported in the Introduction, autoradiographic methods have shown the existence of both a,- and a,-receptors in the cerebellar cortex, with a fairly high level of specific binding for the a,-receptors in the molecular layer (Jones et al., 1985; Palacios et al., 1987) and a low level of specific binding for a,-receptors in the granule cell layer of the cerebellar cortex (Unnerstall et al., 1984; Bruning et al., 1987). These findings may explain why injection of metoxamine in the cerebellar hemivermis usually increased the gain of the ipsilateral triceps brachii to labyrinth stim-
ulation with weak, if any, effect on the contralateral side, while injection into the same corticocerebellar area of clonidine produced bilateral effects. It is likely, in fact, that metoxamine acts on a,-adrenoceptors located mainly on the dendrites of P-cells, thus involving the ipsilateral limbs, while clonidine may act on a,-adrenoceptors located not only on P-cells but also on granule cells. Since their parallel fibers are consistently longer than the width of single efferent zones (Mugnaini, 1983), they can modulate the activity of P-cells of the contralateral hemivermis, either directly or through local interneurons, thus accounting for the appearance not only of ipsilateral but also of contralateral effects. The specificity of the results is supported by the fact that the effects of metoxamine were greatly reduced or suppressed by previous injection into the cerebellar vermis of the a,-adrenergic antagonist prazosin, while those induced by clonidine were prevented by local injection of the a,-adrenergic antagonist yohimbine. An additional finding, however, was that the effects of metoxamine-which were mainly ipsilateral-could also be suppressed by previous injection into the cerebellar vermis of the a,-adrenergic antagonist yohimbine, while injection of prazosin prevented the occurrence of the ipsilateral (but not of the contralateral) effects following injection of the a,-adrenergic agonist clonidine. These findings do not necessarily mean that at the concentration used in the present experiments the putative a l or a,-adrenergic antagonists exerted some blocking properties on both types of a-receptors (cf. U'Prichard et al., 1977; Basile and Dunwiddie, 1984). An alternative possibility, in fact, is that both a l - and a,-adrenoceptors are located on the same corticocerebellar neurons as the P-cells, so that administration of either one of the two aadrenergic blockers may modify the responsiveness of the same neurons to the adrenergic agonist acting on the other subtype of a-receptors. In addition to a-adrenoceptors, in Llitro ' autoradiography studies have shown the existence of 0-adrenoceptors in the cerebellar cortex (see
479
Introduction). In particular, these receptors were located in the molecular layer (Palacios and Kuhar, 1980, 1982; Rainbow et al., 1984; Lorton and Davis, 1987) as well as in “patches” over small groups of P-cells somata (Sutin and Minneman, 1985); it appeared also that in the molecular layer the &-receptors predominated over the &-receptors (Nahorski, 1978; Minneman et al. , 1979, 1981; Rainbow et al., 1984). More recently, however, by using in situ hybridization histochemistry it was reported that in the cerebellar cortex of rat neurons expressing P,-adrenoceptors corresponded to the granule cells (Palacios, 1988). This finding suggests that at least some of the P-adrenoceptors identified in autoradiographic studies in the molecular layer may actually be located presynaptically on the terminals of parallel fibers, axons of granule cells. The increase in the response gain of the triceps brachii of both sides to labyrinth stimulation after unilateral injection of the non-selective Padrenergic agonist isoproterenol into the cerebellar vermis can thus be due to activation of padrenoceptors located in the molecular layer and acting either directly or through terminals of parallel fibers on the dendrites of P-cells projecting to the ipsilateral LVN. The same agent may also act on neurons of the granule layer interconnecting the paramedial zones of the cerebellar vermis of both sides. It is of interest that injection of the non-selective P-adrenergic antagonist propranolol into the vermal cortex of the cerebellar anterior lobe decreased the response gain of the ipsilateral and, to a lesser extent, of the contralatera1 triceps brachii to labyrinth stimulation, and also depressed the effects of a previous injection of isoproterenol into the same corticocerebellar area. Further experiments are required to determine the relative contribution of the &- and the PI-adrenoceptors in modifying the gain of the VS reflexes. The postural and reflex changes elicited by local administration of a P-adrenergic agonist or antagonist into the hemivermal cortex of the cerebellar anterior lobe resemble, to some extent,
those elicited by unilateral injection into the same corticocerebellar area of a-adrenergic agonists or antagonists. Since a cross-reactivity of /3-adrenergic agents with a-mechanisms has been reported in the literature (U’Prichard et aZ., 1977; B a d e and Dunwiddie, 1984; Granholm and Palmer, 19881, it can be argued that at the concentration used in the present experiments the P-adrenergic agents might have acted also through a-adrenergic mechanisms. This hypothesis is made unlikely by the fact that in our experiments a preinjection of the P-adrenergic antagonist propranolol into the vermal cortex of the cerebellar anterior lobe of one side did not prevent the increase in the response gain of the ipsilateral triceps brachii to labyrinth stimulation elicited by local injection of a selective a , - or a,-adrenergic agonist; only the smaller increase in the response gain of the contralateral muscle to labyrinth stimulation induced by a-adrenergic agonists was impaired by the propranolol injection. These findings exclude the possibility that in these experiments the P-adrenergic blocker propranolol displayed a local anesthetic action on the P-cells, as postulated in previous studies (cf. Parfitt et al., 1988). It is worth mentioning that p-adrenoceptors are located not only on neurons but also on glial cells (cf. Stone and Ariano, 1989). In this respect, the effects of P-adrenergic agents described in the present study could be attributed not only to a direct action on neuronal P-receptors, but also to some indirect influence on @receptors located on adjacent glial cells. Cerebellar mechanisms In order to understand the postural changes induced by unilateral injection of adrenergic agents into the vermal cortex of the cerebellar anterior lobe we should mention that the postural activity depends, in part at least, on the discharge of VS neurons originating from the LVN (cf. Pompeiano, 1975). These neurons, which exert a direct excitatory influence on ipsilateral limb extensor motoneurons (Lund and Pompeiano, 19681, are activated by the ipsilateral labyrinth input,
480
but are inhibited by P-cells of the cerebellar vermis, which contribute to the direct corticocerebellar vestibular projection to the LVN (cf. Pompeiano, 1967; Ito, 1984). It is well known that NA agents may modify the discharge of P-cells. Experiments performed in situ have shown that microiontophoretic application of norepinephrine (NE) (Hoffer et aL, 1971, 1974; Freedman et al., 1975, 1977; Moises et al., 1979b), as well as LC stimulation (Hoffer et al., 1973a; Moises and Woodward, 19801, decreased the spontaneous firing rate of P-cells. This effect was associated with hyperpolarization of the P-cells membrane (cf. Waterhouse et al., 19821, which was coupled with an increase in their input resistance, suggesting that these effects were probably mediated through CAMP (Siggins et al., 1969, 1971a,b, 1973; Hoffer et al., 1973a). On the other hand, depletion of NE in the cerebellum was associated with slight increases in P-cell background activity (Hoffer et al., 1973b; McElligott et al., 1986). The effects described above were originally attributed to P-receptors, since they could be mimicked by local injection of the 0-adrenergic agonist isoproterenol (Hoffer et al., 1971; Waterhouse et al., 1982) and blocked by the putative P-antagonist sotalol (Hoffer et al., 1971; Moises et al., 1979b). However, investigations performed in uitro (cerebellar slices, B a d e and Dunwiddie, 1984) as well as in oculo (intraocular cerebellar grafts, Granholm and Palmer, 1988) have revealed that NE could induce not only inhibition but also excitation of P-cell spontaneous activity, the former effects being mediated by a-receptors, while the latter were mediated by P-receptors (cf. also Crepe1 et al., 1987) *. Because of these dis-
* It is of interest that both cr- and P-receptors may play not only a postsynaptic role on the cerebellar P-cells, but also a presynaptic role by increasing (through &-receptors) or decreasing (through the @-receptors)the release of glutamate at the level of parallel fibers and possibly also of CF terminals, as shown in cerebellar slices (Dolphin, 1982).
crepancies, the effects of local application of NE in anesthetized rats have been reinvestigated by Parfitt et al. (19881, who found that pressure microejection of NE elicited a dose-dependent inhibition of P-cell firing rate that was apparently blocked by phentolamine, an a-adrenergic antagonist, but was rarely affected by the @-antagonist timolol. Although the effects of NE appeared to be mediated primarily by a-receptors, both aand P-agonists applied by pressure microejection were able to inhibit P-cell activity. The fact that adrenergic agonists inhibit the P-cell spontaneous activity led us to expect that unilateral injection of these agents would increase the contraction of the ipsilateral limb extensors, due to disinhibition of the corresponding LVN; on the other hand, a reduced contraction of the contralateral limbs might occur due to a reciprocal mechanism utilizing a crossed pathway. In our experiments, however, unilateral injection of a- or P-adrenergic agonists reduced the postural activity in the ipsilateral limbs, while that of the contralateral limbs increased. Since the NA system may act not only on P-cells, but also on GABAergic interneurons (cf. Landis and Bloom, 1979, we postulated that local administration of a- and/or P-adrenergic agonists would lead to a reduced discharge of these interneurons and thus to disinhibition of the related P-cells, a finding which might overcome the slight depression of the firing rate induced by the direct action of the a- or P-adrenergic agonist on the P-cells. The increased discharge of the P-cells could then reduce the contraction of the ipsilateral limb extensors, thus accounting for the results obtained in the present experiments. In addition to changes in posture, unilateral injection of a- or P-adrenergic agents into the vermal cortex of the cerebellar anterior lobe also produced changes in gain of the VS reflexes. It is of interest that the contraction of limb extensors which occurs during side-down tilt of the animal (cf. Schor and Miller, 1981; Manzoni et al., 1983) depends upon an increase discharge of LVN neurons, while just the opposite result occurs during
48 1
side-up tilt (cf. Boyle and Pompeiano, 1980). There is evidence that the discharge of VS neurons during tilt is submitted to a cerebellar control. Experiments of unit recording have, in fact, shown that P-cells located in the longitudinal paramedial zone of the anterior vermis, which projects to the ipsilateral LVN, undergo a sinusoidal modulation of their MF and CF discharge during roll tilt of the animal (Denoth et al., 1979). In particular, most of the P-cells showed a modulation of their simple spike activity, characterized by a decreased discharge during side-down tilt and an increased discharge during side-up tilt. This response pattern, which was opposite in phase with respect to that of the LVN neurons, indicates that the increased activity of the VS neurons, and thus the contraction of the corresponding limb extensors during side-down tilt, depends not only upon an increased excitatory input originating from ipsilateral labyrinth receptors, but also on disinhibition of the same neurons resulting from a reduced discharge of the overlying P-cells; this interaction will then lead to a positive influence on the VS reflex gain. Observations made in the in uiuo preparations have shown that local application of NE (Siggins et aZ., 1971b; Freedman et al., 1976, 1977; Woodward et al., 1979; Moises et al., 1979a, 1990) as well as LC stimulation (Hoffer et al., 1973a; Moises et al., 1979a, 1981, 1983; Moises and Woodward, 1980), while depressing the spontaneous activity of the P-cells, enhanced the responses of these cells to both excitatory (MF and CF) as well as inhibitory (basket and stellate cells) inputs. Similarly, iontophoretically applied NE (Freedman et al., 1975; Moises et al., 1979b; Woodward, et al., 1979; Waterhouse et al., 1982; Yeh et al., 1981; Marshall and Tsai, 1988) or LC stimulation (Moises and Woodward, 1980; Moises et al., 1983) enhanced the responsiveness of P-cells to excitatory (glutamate and aspartate) and inhibitory (GABA) neurotransmitters of the cerebellar cortex. These effects have been reported to be mediated by P-adrenergic receptors (Moises et al., 1981, 1983, 1990; Waterhouse et al., 1982; Yeh
and Woodward, 1983); however, it cannot be ruled out that a-adrenoceptors also contributed to these effects. As a result of these findings it has been postulated that one of the main functions of the NEcontaining input in cerebellar operation is to augment target neuron responsiveness to conventional afferent systems which are directly concerned with detailed information transfer, thus increasing the signal-to-noise ratio of the evoked versus spontaneous activity (cf. Woodward et al., 1979; Waterhouse et al., 1988). The same input could also act to gate the efficacy of subliminal synaptic inputs conveyed by classical afferent systems (Moises et al., 1990). In our experiments the increase in gain of the VS reflex, elicited by local injection of a- or p-agonists, into the cerebellar vermis, can be explained by assuming that NE increased the amplitude of modulation of the MF as well as of the CF discharge of the P-cells of the cerebellar vermis to given parameters of labyrinth stimulation. Since, as reported above, the simple spike discharge of the P-cells is outof-phase with respect to the activity of the VS neurons, an increased responsiveness of these P-cells to the same labyrinth input, as a result of a - or p-adrenergic actions, might contribute to an increased gain of the VS reflex as described above. The possibility that adrenergic receptors are also located on granule cells, which may transmit the afferent signal from one side of the cerebellar vermis to the P-cells of the opposite side, explains why local administration of adrenergic agents may affect the response gain not only of the ipsilateral but also of the contralateral limb extensors to labyrinth stimulation. Interestingly the background discharge of the LC neurons increases in various animals according to the different levels of alertness (cf. Foote et al., 1983) or stress (Abercrombie and Jacobs, 1987). We postulate, therefore, that the NA afferents to the cerebellum may act on the neuronal network, driven by the labyrinthine input, to function appropriately during the different animal states.
482
Acknowledgements
This work was supported by National Institute of Neurological and Communicative Disorders and Stroke Research Grant NS 07685-22 and by grants of the Minister0 dell’universith e della Ricerca Scientifica e Tecnologica, and the Agenzia Spaziale Italiana, Roma, Italy. References Abercrombie, E.D. and Jacobs, B.L. (1987) Single-unit response of noradrenergic neurons in the locus coeruleus of freely moving cats. I. Acutely presented stressful and nonstressful stimuli. J. Neurosci., 7: 2837-2843. Andre, P., d’Ascanio, P., Gennari, A,, Pirodda, A. and Pompeiano, 0. (1991a) Microinjections of 0 1 , - and a z noradrenergic substances in the cerebellar vermis of decerebrate cats affect the gain of the vestibulospinal reflexes. Arch. Ital. Biol., 129: 113-160. Andre, P., d’Ascanio, P., Manzoni, D. and Pompeiano, 0. (1991b) Microinjections of P-noradrenergic substances in the cerebellar vermis of decerebrate cats modify the gain of the vestibulospinal reflexes. Arch. Ital. Biol., 129: 161197. Basile, A.S. and Dunwiddie, T.V. (1984) Norepinephrine elicits both excitatory and inhibitory responses from Purkinje cells in the in uitro rat cerebellar slice. Brain Res., 296: 15-25. Bloom, F.E., Hoffer, B.J. and Siggins, G.R. (1971) Studies on norepinephrine-containing afferents to Purkinje cells of rat cerebellum. I. Localization of the fibers and their synapses. Brain Res., 25: 501-521. Boyle, R. and Pompeiano, 0. (1980) Reciprocal responses to sinusoidal tilt of neurons in Deiters’ nucleus and their dynamic characteristics. Arch. Ital. B i d , 118: 1-32. Briining, G., Kaulen, P. and Baumgarten, H.G. (1987) Quantitative autoradiographic localization of a*-antagonist binding sites in rat brain using L3H] idazoxan. Neurosci. Lett., 83: 333-337. Bylund, D.B. and U’Prichard, D.C. (1983) Characterization of a,- and a,-adrenergic receptors. Int. Reu. Neurobiol., 24: 343-431. Chu, N.-S. and Bloom, F.E. (1974) The catecholamine-containing neurons in the cat dorsolateral pontine tegmentum: Distribution of the cell bodies and some axonal projections. Brain Res., 66: 1-21. Corvaja, N. and Pompeiano, 0. (1979) Identification of cerebellar corticovestibular neurons retrogradely labeled with horseradish peroxidase. Neuroscience, 4: 507-5 15. Crepel, F., Debono, M. and Flores, R. (1987) a-Adrenergic inhibition of rat cerebellar Purkinje cells in vitro: A voltage-clamp study. J. Physiol. (London), 383: 487-498. d’Ascanio, P., Bettini, E. and Pompeiano, 0. (1985) Tonic inhibitory influences of locus coeruleus on the response
gain of limb extensors to sinusoidal labyrinth and neck stimulations. Arch. ftul. B i d , 123: 69-100. d’Ascanio, P., Horn, E., Pompeiano, 0. and Stampacchia, G. (1989a) Injections of P-adrenergic substances in the locus coeruleus affect the gain of vestibulospinal reflexes in decerebrate cats. Arch. Ifal. Bid., 127: 187-218. d’Ascanio, P., Horn, E., Pompeiano, 0. and Stampacchia, G. (1989b) Injections of a P-adrenergic antagonist in pontine reticular structures modify the gain of vestibulospinal reflexes in decerebrate cats. Arch. Ifal. B i d , 127: 275-303. d’Ascanio, P., Pompeiano, M. and Tononi, G. (1989~)Inhibition of vestibulospinal reflexes during the episodes of postural atonia induced by unilateral lesion of the locus coeruleus in the decerebrate cat. Arch. Ifal. Biol., 127: 81-97. Denoth, F., Magherini, P.C., Pompeiano, 0. and StanojeviC, M. (1979) Responses of Purkinje cells of the cerebellar vermis to neck and macular vestibular inputs. Pflugers Arch., 381: 87-98. Dietrichs, E. (1988) Cerebellar cortical and nuclear afferents from the feline locus coeruleus complex. Neuroscience, 27: 77-91. Dolphin, A.C. (1982) Noradrenergic modulation of glutamate release in the cerebellum. Brain Rex, 252: 111-116. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Rei,., 63: 844-914. Freedman, R., Hoffer, B.J. and Woodward, D.J. (1975) A quantitative microiontophoretic analysis of the responses of cerebral neurons to norepinephrine: Interactions with cobalt, manganese, verapamil and dichloroisoproterenol. Br. J. Pharrnacol., 54: 529-539. Freedman, R., Hoffer, B.J., Puro, D. and Woodward, D.J. (1976) Noradrenaline modulation of the responses of the cerebellar Purkinje cell to afferent synaptic activity. Br. J. Pharmacol., 57: 603-605. Freedman, R., Hoffer, B.J., Woodward, D.J. and Puro, D. (1977) Interaction of norepinephrine with cerebellar activity evoked by mossy and climbing fibers. Exp. Neurol., 55: 269-288. Fuxe, K. (1965) Evidence for the existence of monoamine neurons in the central nervous system. IV. The distribution of monoamine terminals in the central nervous system. Acta Physiol. Scand., Suppl., 247: 37-85. Granholm, A.-C.E. and Palmer, M.R. (1988) Electrophysiological effects of norepinephrine on Purkinje neurons in intraocular grafts: a- versus P-specificity. Brain Res., 459: 256-264. Hoffer, B.J., Siggins, G.R. and Bloom, F.E. (1971) Studies on norepinephrine-containing afferents to Purkinje cells of rat cerebellum. 11. Sensitivity of Purkinje cells to norepinephrine and related substances administered by microiontophoresis. Brain Res., 25: 523-534. Hoffer, B.J., Siggins, G.R., Oliver, A.P. and Bloom, F.E. (1973a) Activation of the pathway from locus coeruleus to rat cerebellar Purkinje neurons: Pharmacological evidence of noradrenergic central inhibition. J. Pharmacol. Exp. Ther., 184: 553-569.
483 Hoffer, B.J., Siggins, G.R., Woodward, D.J. and Bloom, F.E. (197313 Spontaneous discharge of Purkinje neurons after destruction of catecholamine-containing afferents by 6-hydroxydopamine. Brain Rex, 30: 425-430. Hoffer, B.J., Seiger, A., Ljungberg, T. and Olson, L. (1974) Electrophysiological and cytological studies of brain homogenate in the anterior chamber of the eye: Maturation of cerebellar cortex in oculo. Brain Rex, 79: 165-184. Hokfelt, T. and Fuxe, K. (1969) Cerebellar monoamine nerve terminals, a new type of afferent fibers to the cortex cerebelli. Exp. Brain Res., 9: 63-72. Ito, M. (1984) The Cerebellum and Neural Control. New York, Raven Press, XVII-580 pp. Jones, L.S., Gauger, L.L. and Davis, J.N. (1985) Anatomy of brain a,-adrenergic receptors: In litro autoradiography with ['251]-HEAT. J. Comp. Neurol., 231: 190-208. Kimoto, Y., Tohyama, M., Satch, K., Sakumoto, T., Takahashi, Y. and Shimizu, N. (1981) Fine structure of rat cerebellar noradrenaline terminals as visualized by potassium permenganate in situ perfusion fixation method. Neuroscience, 6: 47-58. Landis, S.C. and Bloom, F.E. (1975) Ultrastructure identification of noradrenergic boutons in mutant and normal mouse cerebellar cortex. Bruin Rex, 96: 299-305. Lefkowitz, R.J., Stadel, J.M. and Caron, M.G. (1983) Adenylate cyclase-coupled beta-adrenergic receptors: Structure and mechanisms of activation and desensitization. Ann. Re[.. Biochem., 52: 159-186. Lorton, D. and Davis, J.N. (1987) The distribution of beta-I and beta-2-adrenergic receptors of normal and reeler mouse brain: An in iitro autoradiographic study. Neuroscience, 23: 199-210. Lund, S. and Pompeiano, 0. (1968) Monosynaptic excitation of a-motoneurons from supraspinal structures in the cat. Actu Physiol. Scund., 73: 1-21. Manzoni, D., Pompeiano, O., Srivastava, U.C. and Stampacchia, G. (1983) Responses of forelimb extensors to sinusoidal stimulation of macular labyrinth and neck receptors. Arch. ltul. Biol., 121: 205-214. Marshall, K.C. and Tsai, W.H. (1988) Noradrenaline induces short and long duration potentiation of glutamate excitations of cultured Purkinje neurons. Can. J. Physiol. Pharmacol., 66: 848-853. McElligott, J.G., Ebner, T.J. and Bloedel, J.R. (1986) Reduction of cerebellar norepinephrine alters climbing fiber enhancement of mossy fiber input to the Purkinje cell. Bruin Res., 397: 245-252. Minneman, K.P., Hegstrand, L.R. and Molinoff, P.B. (1979) Simultaneous determination of beta-1 and beta-2-adrenergic receptors in tissues containing both receptor subtypes. Mol. Pharmacol., 16: 34-46. Minneman, K.P., Pittman, R.N. and Molinoff, P.B. (1981) P-adrenergic receptor subtypes: Properties, distribution and regulation. Ann. Rec. Neurosci., 4: 419-461. Moises, H.C. and Woodward, D.J. (1980) Potentiation of GABA inhibitory action in cerebellum by locus coeruleus stimulation. Bruin Res., 182: 327-344. Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1979a)
Potentiation of monosynaptic Purkinje cell excitation and inhibition following locus coeruleus activation. Soc. Neurusci. Abstr., 5: 345. Moises, H.C., Woodward, D.J., Hoffer, B.J. and Freedman, R. (1979b) Interactions of norepinephrine with Purkinje cell responses to putative amino acid neurotransmitters applied by microiontophoresis. Exp. Neurol., 64: 493-515. Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1981) Locus coeruleus stimulation potentiates Purkinje cell responses to afferent input: The climbing fiber system. Bruin Rus., 222: 43-64. Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1983) Locus coeruleus stimulation potentiates local inhibitory processes in rat cerebellum. Brain Res. Bull., 10 795-804. Moises, H.C., Burne, R.A. and Woodward, D.J. (1990) Modification of t h e visual response properties of cerebellar neurons by norepinephrine. Brain Res., 514: 259-275. Mugnaini, E. (1983) The length of cerebellar parallel fibers in chicken and Rhesus monkey. J. Comp. Neurol., 220: 7-15. Nahorski, S.R. (1978) Heterogeneity of cerebral beta-adrenergic binding sites in various vertebrate species. Eur. J. Pharmacol., 51: 199-209. Olson, L. and Fuxe, K. (1971) On the projections from the locus coeruleus noradrenaline neurons: The cerebellar innervation. Brain Rex, 28: 165-171. Palacios, J.M. (1988) Mapping brain receptors by autoradiography. ISI Atlas of Science. Pharmacology, 2: 71-77. Palacios, J.M. and Kuhar, M.J. (1980) Beta adrenergic receptor localization by light microscopic autoradiography. Science, 208: 1378-1380. Palacios, J.M. and Kuhar, M.J. (1982) Beta adrenergic receptor localization in rat brain by light microscopic autoradiography. Neurochem. Int., 4: 473-490. Palacios, J.M., Hoyer, D. and CortCs, R. (1987) a,-Adrenoceptors in the mammalian brain: Similar pharmacology but different distribution in rodents and primates. Brain Res., 419: 65-75. Parfitt, K.D., Freedman, R. and Bickford-Wimer, P.C. (1988) Electrophysiological effects of locally applied noradrenergic agents at cerebellar Purkinje neurons: Receptor specificity. Brain Res., 462: 242-251. Pasquier, D.A., Gold, M.A. and Jacobowitz, D.M. (1980) Noradrenergic perikarya (A5-A7, subcoeruleus) projections to the rat cerebellum. Brain Res., 196: 270-275. Pickel, V.M., Segal, M. and Bloom, F.E. (1974) A radioautographic study of the efferent pathways of the nucleus locus coeruleus. J. Comp. Neurol., 155: 15-42. Pompeiano, M., Galbani, P. and Ronca-Testoni, S. (1989) Distribution of P-adrenergic receptors in different cortical and nuclear regions of cat cerebellum, as revealed by binding studies. Arch. Ital. B i d , 127: 115-132. Pompeiano, 0. (1967) Functional organization of the cerebellar projections to the spinal cord. In C.A. Fox and R.S. Snider (Eds.), The Cerebellum. Progress in Brain Research, Vol. 25, Elsevier, Amsterdam, pp. 282-321. Pompeiano, 0. (1975) Vestibulospinal relationships. In R.F. Naunton (Ed.), The Vestibular System, Academic Press, New York, pp. 147-180.
484 Pompeiano, 0. (1989) Relationship of noradrenergic locus coeruleus neurones to vestibulospinal reflexes. In J.H.J. Allum and M. Hulliger (Eds.), Afferent Control of Posture and Locomotion. Progress in Brain Research, Vol. 80, Elsevier, Amsterdam, pp. 329-343. Pompeiano, O., d’Ascanio, P., Horn, E. and Stampacchia, G. (1987) Effects of local injection of the a,-adrenergic agonist clonidine in the locus coeruleus complex on the gain of vestibulospinal and cervicospinal reflexes in decerebrate cats. Arch. Ital. Biol., 125: 225-269. Powers, R.E., O’Connors, D.T. and Price, D.L. (1989) Noradrenergic systems in human cerebellum. Brain Res., 481: 194- 199. Rainbow, T.C., Parsons, B. and Wolfe, B.B. (1984) Quantitative autoradiography of P I - and P2-adrenergic receptors in rat brain. Proc. Natl. Acad. Sci. USA, 81: 1585-1589. Schor, R.H. and Miller, A.D. (1981) Vestibular reflexes in neck and forelimb muscles evoked by roll tilt. J. Neurophysiol., 46: 167-178. Siggins, G.R., Hoffer, B.J. and Bloom, F.E. (1969) Cyclic adenosine monophosphate: Possible mediator for norepinephrine effects of cerebellar Purkinje cells. Science, 165: 1018- 1020. Siggins, G.R., Hoffer, B.J. and Bloom, F.E. (1971a) Studies on norepinephrine-containing afferents to Purkinje cells of rat cerebellum. 111. Evidence for mediation of norepinephrine effects by cyclic 3’-5‘-adenosine monophosphate. Bruin Res., 25: 535-553. Siggins, G.R., Oliver, A.P., Hoffer, B.J. and Bloom, F.E. (1971b) Cyclic adenosine monophosphate and norepinephrine: Effects on transmembrane properties of cerebellar Purkinje cells. Science, 171: 192-194. Siggins, G.R., Battenberg, E.F., Hoffer, B.J., Bloom, F.E. and Steiner, A.L. (1973) Noradrenergic stimulation of cyclic adenosine monophosphate in rat Purkinje neurons: An immunocytochemical study. Science, 179: 585-588. Stone, E.A. and Ariano, M.A. (1989) Are glial cells targets of the central noradrenergic system? A review of the evidence. Brain Res. Rer.., 14: 297-309. Sutin, J. and Minneman, K.P. (1985) Adrenergic beta receptors are not uniformly distributed in the cerebellar cortex. J. Comp. Neurol., 236: 547-554. Trott, J.R. and Armstrong, D.M. (1987) The cerebellar corti-
conuclear projection from lobule Vb/c of the cat anterior lobe: A combined electrophysiological and autoradiographic study. 11. Projections from the vermis. Exp. Bruin Rex, 68: 339-354. Unnerstall, J.R., Kopajtic, T.A. and Kuhar, M.J. (1984) Distribution of a 2 agonist binding sites in the rat and human central nervous system: Analysis of some functional, anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agents. Bruin Res. Rea., 7: 69-101. U’Prichard, D.C., Greenberg, D.A. and Snyder, S.H. (1977) Binding characteristics of a radiolabelled agonist and antagonist at central nervous system (Y noradrenergic receptors. ,4401. Pharmacol., 13: 454-473. Voogd, J. (1989) Parasagittal zones and compartments of the anterior vermis of the cat cerebellum. Exp. Bruin Res. Ser., 17: 3-19. Waterhouse, B.D., Moises, H.C., Yeh, H.H. and Woodward, D.J. (1982) Norepinephrine enhancement of inhibitory synaptic mechanisms in cerebellum and cerebral cortex: Mediation by beta adrenergic receptors. J. Pharmucol. Exp. Ther., 221: 495-506. Waterhouse, D.B., Sessler, F.M., Cheng, J.T., Woodward, J.D., Azizi, S.A. and Moises, H.D. (1988) New evidence for a gating action of norepinephrine in central neuronal circuits of mammalian brain. Brain Res. Bull., 21: 425-432. Woodward, D.J., Moises, H.C., Waterhouse, B.D., Hoffer, B.J. and Freedman, R. (1979) Modulatory actions of norepinephrine in the central nervous system. Fed. Proc., 38: 2109-2116. Yamamoto, T., Ishikawa, M. and Tanaka, C. (1977) Catecholaminergic terminals in the developing and adult rat cerebellum. Brain Rex, 132: 355-361. Yeh, H.H. and Woodward, D.J. (1983) Beta-1 adrenergic receptors mediate noradrenergic facilitation of Purkinje cell responses to gamma-amino-butyric acid in cerebellum of rat. Neuropharmacology, 22: 629-639. Yeh, H.H., Moises, H.C., Waterhouse, B.D. and Woodward, D.J. (1981) Modulatory interactions between norepinephrine and taurine, beta-alanine, gamma-aminobutyric and muscimol, applied iontophoretically to cerebellar Purkinje cells. Neuropharmacology, 20: 549-560.
C.D. Barnes and 0. Pompeiano (Eds.) Progress i n Brain Research. Vol. 8X 0 1991 Ekevicr Science Publishers B.V.
485 CHAPTER 34
Effects of GABAergic and noradrenergic injections into the cerebellar flocculus on vestibulo-ocular reflexes in the rabbit J. van Neerven
0. Pompeiano
and H. Collewijn
Department of Physiology I, Erasmus Uniuersity Rotterdam, Rotterdam, The Netherlands and Department of Physiology and Biochemistry, University of Pisa, via S. Zeno, Pisa, Italy
The role of the vestibule-cerebellum of the rabbit in the control of the vestibulo-ocular response (VOR) and optokinetic response (OKR) reflexes was investigated by bilateral microinjections, into the flocculus, of substances affecting GABAergic or noradrenergic neurotransmission. GABA, the main transmitter through which cerebellar interneurons inhibit Purkinje cells directly or indirectly, acts normally through GABA, receptors (mainly located in the granular layer) and GABA receptors (predominantly located in the molecular layer). Despite this different distribution, floccular injections of the GABA, agonist muscimol and of the GABA, agonist baclofen had a similar effect, presumably by profound inhibition of Purkinje cells. This effect consisted of a reduction in the gain of the VOR (in darkness and in light) as well as of the OKR by at least 50%. This provides firm evidence that the net effect of normal Purkinje-cell activity in the flocculus is to enhance the VOR and OKR, rather than to inhibit these responses, as is sometimes supposed. Intrafloccular injections of the P-noradrenergic agonist
,
isoproterenol or the P-noradrenergic antagonist sotalol did not affect the basic magnitude of the VOR and OKR. However, these substances markedly affected the adaptive processes, which cause the VOR and OKR to change its magnitude when this is no longer adequate in stabilizing the retinal image. By a suitable combination of vestibular and optokinetic stimuli, consistent upward changes in the gain of these reflexes could be reliably and reproducibly induced in uninjected animals. Floccular injections of sotalol impaired these adaptive changes markedly, whereas injections of isoproterenol enhanced the adaptation, particularly of the VOR measured in darkness. These findings strongly suggest that the effectuation of adaptive changes of vestibular, and possibly other, motor control systems is strongly facilitated by the noradrenergic innervation of the flocculus, which is normally provided by the locus coeruleus (LC), by way of the P-receptor system, although the activity of this system does not directly affect the signal transmission supporting the basic reflexes as such.
Key words: vestibule-ocular reflex, adaptation, flocculus, noradrenaline, GABA, receptors, GABA
Introduction The nucleus locus coeruleus (LC) has been shown to contain cells which send their efferents all over the central nervous system (Olson and Fuxe, 1971). These widespread projections suggest an enormous variability in actions of the noradrener-
,receptors, P-adrenoceptors
gic system, and indeed the noradrenergic system has been shown to affect a number of very different functions in the central nervous system. Experimental studies have indicated the involvement of this system in sleep, memory and plasticity. In the last decade it has become clear that
486
noradrenaline (NA) does not exert a clearcut excitatory or inhibitory influence on target neurons, but that rather it enhances the effects of the other inputs to the neuron by increasing the signal-to-noise ratio of evoked versus spontaneous activity. These findings have led to the hypothesis that the noradrenergic system exerts a modulatory action on its target neurons throughout the central nervous system (for reviews see Woodward et al., 1979 and Waterhouse et al., 1988). One important finding was the demonstration of a modulatory effect of the noradrenergic system on the plasticity of motor systems. For example, the acquisition of a locomotor task was under noradrenergic control (Watson and McElligot, 1983, 1984). NA also affected other motor functions, such as the vestibulo-spinal reflex (see Pompeiano et al., 1987, 1990; Pompeiano, 1989). This finding led us to question whether NA would also influence another vestibular reflex system, ie., the vestibulo-ocular reflex (VOR). The VOR serves to stabilize retinal images of the surroundings during movements of the head or the whole body. The input to the reflex loop, a vestibular signal, results in a compensatory eye movement which drives both eyes in the opposite direction to the head movement. The most elementary VOR can therefore only be demonstrated in total darkness, presenting only vestibular stimuli. The eye and head movement should ideally be of equal magnitude in order to assure optimal stability of the retinal image. In darkness, this is generally not the case, but in normal conditions the VOR is usually active in lighted surroundings. In such situations, both vestibular and visual input signals are used to generate compensatory eye movements. This is called visuo-vestibular interaction. The VOR in darkness acts as an open loop system, which means that the output of the reflex loop cannot be checked at the input level. However, in the light, the effectiveness of the compensatory eye movements can be measured and they can be recalibrated in order to keep them of adequate magnitude. Such
an adjustment of the intrinsic input-output relations is called adaptation. In everyday life, adaptation of the VOR takes place whenever the vestibular system is damaged by trauma or by aging. In laboratory conditions adaptation of the VOR can be induced by subjecting an animal to long-lasting, unusual combinations of visual and vestibular inputs. Vestibular stimuli, visual stimuli and compensatory eye movements can be quantified fairly easily and, therefore, the VOR is very suitable as an experimental paradigm to test both steadystate input-output relations as well as adaptive changes in VOR output. In this way studying VOR adaptation is one means to investigate plasticity in motor systems. In the present study, we chose to test the effects of the central noradrenergic system on either of these situations. The possibility of a noradrenergic influence on adaptation of the VOR has been studied in the past by the use of techniques which depleted the central stores of NA. This was done by intracisternal injections of 6-hydroxydopamine (6OHDA). Cats pre-treated with this substance had a reduced ability to generate plastic changes in VOR output (Keller and Smith, 1983; McElligott and Freedman, 1988a,b). The possible effects of noradrenergic depletion on the steady-state input-output relations of the VOR were not extensively studied by these authors. Moreover, these findings contrast with a report by Miyashita and Watanabe (1984) who found that no loss in VOR adaptation occurred after a similar 6-OHDA injection in pigmented rabbits. Such a loss was, however, obtained in rabbits after depletion of both NA and serotonin (5-HT) by intraventricular injection of 5,7-dihydroxytryptamine (5,7-DHT). Miyashita and Watanabe concluded from these experiments that 5-HT and not NA, is crucial in maintaining adaptive modifiability of the VOR, at least in the rabbit. Apart from this contradiction these general depletion techniques have an important drawback: they do not identify a specific site where NA may affect adaptability of the VOR. More
487
seriously, it is not certain whether the effects described are due to a specific impairment of long-term visuo-vestibular interactions or to more general impairments related to the overall depression of the noradrenergic system. The present study intended to demonstrate a local noradrenergic effect on the VOR. One of the sites where the noradrenergic system might specifically affect the VOR would be in the cerebellar flocculus. Although experimental evidence has been presented which either supports (It0 et al., 1974; 1982) or conflicts with (see Miles and Lisberger, 1981) the hypothesis that the floccular cerebellar cortex is the actual site of the neuronal changes instrumental in adaptation, it is generally agreed that damage or removal of the flocculus prevents adaptation of the VOR gain (It0 et al., 1982; Nagao, 1983). Moreover, the projection of the LC to the flocculus (Kimoto et al., 1978; Somana and Walberg, 1978; Langer et al., 1985) offers one obvious pathway for such an influence on VOR adaptation. The present study was designed to determine whether the noradrenergic system acts through the cerebellar flocculus to modify either the steady-state input-output relations of the VOR, adaptation of the VOR, or both. The presumed actions of the noradrenergic system were investigated by floccular injection of several agonists and antagonists, thus imitating or blocking natural noradrenergic action. Although both P,- and pz- (Minneman et al., 1979, 1981; M. Pompeiano et al., 1989) as well as a*- and a,-noradrenergic receptors (Young and Kuhar, 1980; Jones et at., 1987) have been demonstrated in the cerebellum, we will only discuss in this paper the results obtained after injection of P-agonists and blockers (see also van Neerven et al., 1990). As a brief introduction to these results, we will present data obtained after GABAergic injections into the flocculus (van Neerven et al., 19891, which demonstrate the involvement of the floccular Purkinje cells in the generation of compensatory eye movements.
Methods
Recording of eye movements Young adult (6-12 months) male and female Dutch belted rabbits with a weight of about 2 kg were used in all experiments. All rabbits had scleral search coils permanently implanted on both eyes for the measurement of their eye movements in an a.c. magnetic field. For a more elaborate description of this electromagnetic method for measuring eye movements, see Collewijn (1977). The eye coils consisted of 5 loops of a multistranded, teflon-coated, stainless steel wire (type AS 632, Cooner, Chatsworth, CA) which was woven underneath the inferior and superior rectus muscles. The implantation took place under general anaesthesia by a mixture of ketamine, acepromazide and xylazine (see van Neerven et al., 1989 for specific dosages). The lead wires of the coils were inserted beneath the skin and soldered to a small connector which was anchored, together with three head bolts, to the skull with small screws and dental acrylic. This construction was used to fixate the rabbit’s head in a headholder in order to prevent spontaneous movements during the experiments and to connect the eye coils to the detection apparatus (see Fig. 1). Implantation of guide cannulas All rabbits were also provided with bilaterally implanted guide cannulas, through which injection cannulas could be inserted into the cerebellar flocculus. To enable the implantation of these cannulas, small openings were made in the skull overlying the paramedian cerebellar lobe under general anaesthesia. A wall of dental acrylic was built to enable fixation of the cannulas to the skull. The outer cannulas were implanted under the guidance of electrophysiological recordings with glass micropipettes (Fig. 11, which identified floccular Purkinje cells (see Simpson et al., 1981). A site where the Purkinje cells responded to horizontal movement of a visual stimulus pattern
488
microelectrode outer cannula
eye c - left
flocculus
Fig. 1. Schematic drawing of the method used to localize the flocculus. (From van Neerven et al., 1990.)
was chosen as the preferred target for the injection of the various pharmacological substances. Once such a point had been localized, the outer guide cannula, which surrounded the recording electrode, was lowered until it almost touched the surface of the cortex, and then permanently fixed to the skull with dental acrylic. As the acrylic hardened, the glass electrode was marked just above the upper border of the outer cannula and then retracted. The length of the recording electrode below this mark was taken as the appropriate length for the injection cannula, which fitted exactly inside the outer cannula.
Procedures for non-adaptation experiments All experiments in which steady-state VOR measurements were taken are referred to as non-adaptation experiments. An experimental apparatus was designed to generate vestibular and optokinetic stimuli; the same apparatus could also be used to evoke adaptive changes in the VOR gain (the ratio of the eye movement amplitude and the stimulus amplitude). It consisted of a circular platform (diameter 0.75 m) surrounded by an optokinetic drum of the same diameter. The drum and the platform could be driven separately or simultaneously by velocity-controlled motors (Fig. 2).
The VOR could be evoked by sinusoidal oscillation of the platform in darkness. Oscillation of the optokinetic drum alone resulted in an optokinetic response (OKR), which was studied only in the non-adaptation experiments. Lastly, by oscillating the platform in the light within the stationary drum, the synergistic interaction between the VOR and OKR was investigated. During the experiments the rabbits were secured in a hammock and mounted on a small bench, which was placed on the platform, with their eyes centered at the axis of rotation. None of the supporting fixtures intruded into the visual field. In order to investigate the effects of the injected substances on the steady-state input-output relations of the VOR, the rabbits were oscillated on the platform at a frequency of 0.15 Hz and an amplitude of 5 degrees. The drum was kept stationary and the platform moved only during the control measurements. In each control experiment, the VOR in the light and in darkness as well as the OKR were tested. Three experi-
\ Fig. 2. Arrangement of the experimental devices: an optokinetic drum (cut out in the diagram to show the interior of the drum), decorated with random dots, which completely surrounded the platform. The drum was illuminated from the inside. The devices were separately driven by servo-controlled motors (M). The rabbit was fixated on a bench, the head secured to a head holder. The eye coils, which were soldered to a connector on the head were connected to the detection apparatus by a flexible cable.
489
ments were performed in each rabbit, in random order, to determine the time course of (a) changes in the VOR and OKR gain without injection; (b) changes in the VOR and OKR gain after injection of the appropriate agonist; and (c) changes in the VOR and OKR gain after injection of the appropriate antagonist. In the experiments where we injected GABAergic agonists into the flocculus, we used two frequency-amplitude combinations: 0.25 Hz with an amplitude of 10 degrees and 0.10 Hz with an amplitude of 2.5 degrees.
Procedure for adaptation experiments Adaptive enhancement of the VOR gain was elicited by oscillating the platform sinusoidally, at a frequency of 0.15 Hz and an amplitude of 5 degrees, while the optokinetic drum moved in counter-phase at the same frequency but with an amplitude of 2.5 degrees (see Ito et al., 1974, 1979). When this stimulus was continuously presented for 3 h this increase in retinal slip eventually resulted in the induction of an adaptive increase in the VOR gain in light as well as in darkness. Each experiment was started with four baseline measurements of the VOR during passive oscillation in light as well as in darkness, before any substance was injected into the flocculus. The platform was stationary in between these baseline measurements. As soon as the injection was completed, motion of the platform and the drum in counter-phase was started and the first control measurement was taken. From that moment on, recordings of the VOR in the light as well as in darkness were taken every 15 min. In each rabbit, the experiment described above was performed in three conditions: (a) control adaptation without any injection, (b) adaptation after injection of the appropriate agonist and (c) adaptation after injection of the appropriate antagonist. These three experiments were performed in a random order. To allow the rabbit to recover fully from the previous experiment, there was always at least one day of rest between two sessions.
Injected solutions and histology In the first series of experiments we injected solutions of baclofen, (Ciba Geigy) 5 p g / p l as a GABA, agonist and muscimol, (Sigma, USA) 16 pg/pI as the GABA, agonist. Furthermore, we used isoproterenol HCI (Sigma, USA) 16 pg/pl, an aselective P-agonist and sotalol (Bristol Myers, Holland) 4 pg/pl, an aselective P-antagonist. All substances were dissolved in saline, after which the pH was adjusted to 7.2-7.4. Injections were always made bilaterally, using a stainless steel injection cannula which was connected to a Hamilton 1.0 p l syringe. The solvent (saline) was injected alone in a few control experiments. At the end of the last experiment of a series, each rabbit was injected with a dye (Pontamine sky blue or ink), through the same injection cannula as used in the former injections, to verify the injection site by histological examination. Results
Effects of GABAergic injections on steady-state VOR characteristics To demonstrate the functional role of the cerebellar flocculus in the generation of steady-state VOR compensatory eye movements, the effects of floccular injections of GABAergic agonists were tested. Since GABA is an inhibitory neurotransmitter, application of a GABAergic agonist was expected to generate inhibition of floccular Purkinje cells. Both the GABA, agonist muscimol and the GABA, agonist baclofen were used in these experiments.
Baclofen injections The effects of baclofen on steady-state inputoutput relations of the VOR were tested in seven rabbits and in two frequency/amplitude combinations. The absolute baseline gains of the VOR and OKR had rather consistent values, which did not differ markedly in the seven rabbits at either test frequency. The average baseline gain of the VOR in light was about 0.70 for both frequencies.
-
490
c 0 a,
70-
-*-
- + -VORD
VORL
\
b\
a c
50
--. ..
b’
30 20
The effects of baclofen were fairly similar for the two frequencies tested. After 3 h, the gain of the VOR in light and in darkness was reduced by about 50%, and the gain of the OKR was reduced by about 60-70% at both frequencies (Fig. 3). Thus, the decline of the OKR was, in general, more pronounced than the decline of the VOR. The time course of the VOR gain was identical in the light and in the dark, and for the two frequencies. Only the time course of the decrease in gain of the OKR showed an effect of stimulus frequency: the decline was steeper and stronger for the stimulus with a frequency of 0.25 Hz and an amplitude of 5 degrees (peak velocity 7.9 deg/sec), than for the stimulus with a frequency of 0.10 Hz and an amplitude of 2.5 degrees (peak velocity 1.6 deg/sec). At both frequencies, the effects reached nearly asymptotic values in the first 2 h. Recovery from the depression was slow; the original baseline values were not reobtained until the next day.
;
0
30
60
90
120
150
180
30
60
90
120
150
180
I
20 0
Time (minutes)
Fig. 3. Time course of the effects of bilateral floccular injection of 1.0 pI baclofen (5.0 fig/fil) on the VOR in light, the VOR in darkness and the OKR. Upper graph shows the effects at the 0.25 Hz frequency, lower graph at 0.1 Hz. The average value of 4 baseline measurements was normalized to 100%. Mean values of 7 rabbits. (From van Neerven et al., 1989.)
The average gains of the VOR in darkness and of the OKR were both about 0.45 at 0.10 Hz. At 0.25 Hz, the average baseline gain of the VOR in darkness was 0.60, while the OKR gain was only 0.09. The overall result of bilateral baclofen injections into the flocculus was a strong decrease in gain of the VOR (both in light and darkness) as well as of the OKR, at either test frequency. Figure 3 shows the time course of these decreases in gain at the two test frequencies. This figure shows the gain as a percentage of the baseline values (normalized at 100%).
Muscimol injections Floccular injections of the GABA, agonist muscimol were made in six rabbits. The absolute baseline values of the rabbits used in these experiments were comparable to those obtained in the baclofen experiments. As in the baclofen experiments, the effect of the muscimol injection was a large decrease in gain of the VOR in light and in darkness, and an even stronger decrease in gain of the OKR. Figure 4 shows the time course of these effects in a way similar to that shown in Figure 3. As in the baclofen experiments, we found little difference between the effects of the injection on gain when comparing the two test frequencies. After 3 h, the reduction of the gain was about 50% for the VOR in light and in darkness, and about 65% for the OKR. These values, and the time course at which they were reached, were similar to those after the baclofen injections. Once more, the decline of OKR was steeper at 0.25 Hz than at 0.10 Hz (Fig. 4). With both substances, the largest part of the decrease occurred during
-
20' 0
'
49 1
'
'
--*-CXR
-+-VORD
VORL
'
'
'
'
'
'
'
'
'
30
60
90
120
150
180
30
60
90
120
150
180
100
90
ar
0
80
4-
70
$
a c
$
60 50 40
30
In neither the baclofen nor the muscimol experiments did we find any significant change in the phase of the eye movement responses with respect to the stimulus, although in all cases the gain decreased dramatically. In summary, the effects of baclofen injection, a GABA, agonist, and muscimol, a GABA, agonist into the cerebellar flocculus were highly similar. Either substance caused an impressive decrement in the gain of both the VOR in light and dark, and the OKR, at both of the two frequencies. If the assumption that GABA agonists inhibit floccular Purkinje cells is correct, we can thus conclude that the floccular Purkinje cells contribute positively to the steady-state inputoutput relations of the VOR; this is in line with earlier evidence (It0 et al., 1974). Moreover, these results assure that this method for floccular injections of pharmacological substances works satisfactorily.
20 0
Time (minutes)
Fig. 4. Time course of the effects of bilateral floccular injection of 1.0 pl rnuscimol(16.0 p g / p l ) o n the VOR in light, the VOR in darkness and the OKR. Upper graph shows the effects at the 0.25 Hz frequency, lower graph at 0.1 Hz. The average value of 4 baseline measurements was normalized to 100%. Mean values of 6 rabbits. (From van Neerven et al., 1989.)
the first 2 h of the testing period, and the OKR was affected more severely than the VOR. In fact, the only difference between the substances was that the rabbits injected with baclofen showed less variability in the amount of gain reduction than the muscimol-injected rabbits, as demonstrated by smaller standard deviations. Muscimol, like baclofen, did have a long-lasting effect. The rabbits did not recover their original gain until the next day. Some rabbits showed minor symptoms of a cerebellar syndrome such as ataxia and a slight tremor, 3 h after the muscimol injection, but recovered spontaneously within the next 12 h. These behavioral effects were never seen after a baclofen injection.
Effects of injection of a P-adrenoceptor agonist and antagonist on steady-state VOR and on VOR adaptation
Sixteen rabbits were prepared with eye coils and permanently implanted cannulas. Six were tested in non-adaptation experiments in which the effects of isoproterenol and sotalol on steady-state VOR input-output relations were investigated. Ten rabbits were used in experiments where adaptive changes of the VOR gain were tested (adaptation experiments). Non-adaptation experiments In six rabbits, the OKR as well as the VOR in the light and in darkness were recorded every 15 min for 2.5 h. For these experiments, only one stimulus with a frequency of 0.15 Hz and a 5 degree amplitude was employed. In this group of rabbits, the absolute baseline gain of the VOR was 0.92 5 0.08 (S.D.) in the light, and 0.63 5 0.14 (S.D.) in darkness. The baseline of the OKR was 0.30 0.08 (S.D.). The data gathered in these
+
492
-0.10
'
0
'
'
'
'
'
'
'
50
'
'
'
'
'
1
150
1m
VCX IN DARKNESS
T
age change in gain, AG, of the VOR was 0.03 & 0.05 (S.D.) in the light and 0.04 k 0.15 (S.D.) in darkness, while the average AG of the OKR was 0.04 k 0.04 (S.D.). None of the AG values measured over 2.5 h were significantly different from zero, according to a multiple analysis of variance (MANOVA). No differences in average change in gain were found after injection of the P-agonist isoproterenol or after injection of the pantagonist sotalol. Overall, the gain changes demonstrated in the injected rabbits were very similar to those found in the non-injected animals; thus it can be concluded that the pnoradrenergic injections did not affect the basic VOR or the OKR gain characteristics. Adaptation experiments These experiments were done in ten rabbits. The average baseline values of the VOR gain were similar to those obtained in all other experiments. The average gain of the VOR in the light was 0.86 k 0.16 (S.D.), whereas in the dark the average value was 0.68 f 0.21 (S.D.). In the normal adaptation experiments, the average AG after 1.5 h was 0.26 *0.09 (S.D.) in the light and 0.10 0.04 (S.D.) in darkness. The final gain increase after 3 h of adaptation was 0.26 f O . l l (S.D.) in the light and 0.14 0.07 (S.D.) in darkness. These data are presented in Figure 6. In most rabbits, administration of the P-agonist isoproterenol did not result in a clear effect on the VOR measured in the light. Only two rabbits showed a distinct enhancement of their VOR adaptation (+O.lO), The other rabbits showed no or only marginal gain changes compared to the normal adaptation experiment. However, the values of the VOR gain measured in darkness were systematically higher after floccular injection of isoproterenol than during normal adaptation. This enhancement was present in 9 of 10 rabbits. As in the normal adaptation, the enhancement took place mostly during the first 1.5 h, in light as well as in darkness. Injection of the P-blocker affected adaptation negatively, in the light and in darkness. In both conditions, 7 of the 10 rabbits
+
.
-o.loL--' 0
50
'
.
.
' 100
.
'
'
'
1
150
Fig. 5. Graphs of the mean time course of the changes in gain in the non-adaptation experiments comparing injections of isoproterenol and sotalol to the control non-adaptation experiments. The upper graph represents the effects on the V O R in light, the middle graph shows the VOR in darkness, and the lower graph represents the OKR. Mean values of 6 rabbits; bars represent 1 S.D. (From van Neerven et a!., 1990.)
non-adaptation experiments are plotted in Figure 5. In the experiments without injection, there was virtually no change in the gain of either the VOR or the OKR during the testing period. At the end of the experimental period (after 2.5 h), the aver-
493 ---
t normal
iswrot
T
VOR IN LIGHT
sotaiol
A
T
0.30 W
(I)
: Q
0.20
z Q
0.10
(7
-I!
0.ood
"I i
"
0
"
"
'
I
"
'
1
180
120
60
VOR IN DARKNESS 030
F
O . O O Y 4 ' 0
T
"
60
T
'
'
T
.
.
120
.
, 180
TIME (minutes)
Fig. 6 . Graphs of the mean time course of the changes in gain in the adaptation experiments comparing injections of isoproterenol and sotalol to the control adaptation experiments. The upper graph represents the effects in the light, the lower graph in darkness. Mean values of 10 rabbits; bars represent 1 S.D. (From van Neerven ef al., 1990.)
showed a clear decrease in adaptive gain change after sotalol injection. The remaining animals showed only a slight or no reduction of adaptation. In no case was adaptation increased by sotalol. The overall effects of sotalol were statistically significant. The mean AG of the compensatory eye movements was lower than that obtained in the control adaptation experiments, in the light (P<0.009, MANOVA) as well as in darkness (P< 0.003, MANOVA). Contrary to normal adaptation the rabbits showed little increase in gain during the first hour after the injection of sotalol, and a considerable part of the gain change was acquired in the last hour, when the effect of the @blocker may have been suffi-
ciently reduced to permit the resumption of the adaptive process. In conclusion, after the injection of sotalol, the adaptive change in gain was mostly smaller than, and sometimes equal to, the baseline values in the light as well as in darkness; a relatively large part of any gain increase was delayed until the third hour of adaptation. Isoproterenol injections had little effect in the light, whereas in darkness a consistent enhancement of the gain change was found, particularly in the first 1.5 h. Since we did not find a noradrenergic influence on the steady-state VOR input-output relations, while there was a clear effect on adaptive changes in the VOR output, we may conclude that these results demonstrate a noradrenergic influence on the plasticity of this oculomotor system.
Injections of the solvent In two rabbits, a bilateral injection of the solvent (saline) was made using a volume equal to that administered during the adaptation experiments (1.0 ~1 bilaterally). These experiments were carried out in order to exclude a possible mechanical effect of the injected volume on the VOR gain. One rabbit was tested in an adaptation experiment, the other in a control experiment without adaptation. These injections caused neither a significant modification of the AG compared to the normal adaptation experiment, nor changes in the gain of the VOR and OKR as compared to the baseline values. It can thus be concluded that the effects observed during the adaptation experiments were due to the pharmacological properties of the injected substances. Discussion The first part of this study showed that floccular injection of both GABA, and GABA, agonists resulted in a strong depression of the gain of the VOR and the OKR. Although the two receptor types act in very different ways and are located in different cerebellar layers, the overall effects
494
seemed very similar. GABA transmissions are known to be inhibitory in nature, and thus injection of GABAergic agonists may lead - either directly, or indirectly by way of the cerebellar interneurons - to reversible inactivation of the floccular Purkinje cells. The outcome of the experiments showed that GABAergic floccular injection depressed the VOR gain. This can be understood if we consider the function of the flocculus as part of the parallel side-path to the VOR loop, receiving both vestibular and visual input. The output of the flocculus, consisting of Purkinje cell axons, which are well known to be inhibitory, contributes to the output of the VOR. Because vestibular neurons and floccular Purkinje cells fire out-of-phase, the activation of vestibular neurons is accompanied by an inhibition of floccular Purkinje cells. This cooperation results in disinhibition of the floccular output, which thus explains how the flocculus contributes positively to the VOR and OKR output. Our observations agree well with those by Keller and Precht (1978) as well as Ito et al. (1974, 1982) who found that cerebellectomy or flocculectomy depressed the output of the VOR and OKR. The role of the flocculus in adaptive changes in the VOR gain has been demonstrated many times by flocculectomy experiments (It0 et al., 1974; Ito 1982). The present study was designed to test the hypothesis that the central noradrenergic system acts on the steady state VOR gain or on adaptive changes in the VOR gain through the cerebellar flocculus. The results have demonstrated that floccular injections of a p-adrenoceptor agonist or antagonist affect only the adaptability of the VOR gain, while the steady state gain values remain unaffected. Adaptation was enhanced by injection of isoproterenol, a P-agonist, while it was depressed by sotalol, a P-antagonist. In the control, non-adaptation experiments, injection of the P-agonist isoproterenol or the P-antagonist sotalol caused no significant changes in the basic gain of the VOR or the OKR. Moreover, injections of the solvent alone had no effect.
The increases in gain without injection after 3 h of adaptation are in agreement with those found in other studies with regard to the magnitude as well as to the time course (It0 et al., 1982; Nagao 1983). As for the effects of P-agonists and antagonists on adaptation of the VOR gain, our study agrees well with the earlier work done in this field by Keller and Smith (1983) and McEIligot and Freedman (1988a). These authors showed that the adaptive change in the VOR gain could be reduced by generalized depletion of NA in the cat. In a recent review, McElligot and Freedman (1988b) have reported that the depletion of cerebellar NA by injection of 6-OHDA in the coeruleo-cerebellar pathways had the same effect as generalized depletion. Hence, they concluded that the noradrenergic influence on VOR adaptation takes place in the cerebellum. However, they could not determine the exact localization of the effects because the coeruleo-cerebelIar afferents terminate in the whole cerebellar cortex and in all of the nuclei, as shown in anatomical (Dietrichs, 1988) and binding studies (M. Pompeiano et al., 1989). Furthermore, they did not exclude the possibility that the effects were due to changes in the basic VOR characteristics rather than to changes in adaptive capacity. The present results deal with both of these problems and also provide evidence on the responsible receptor type. On the other hand, the present results contradict those described by Miyashita and Watanabe (1984). These authors could not demonstrate a decreased adaptability after depletion of NA, but they did demonstrate a serotonergic effect. The conclusion of their work was that the serotoncrgic and not the noradrenergic system is involved in VOR adaptation. In separate experiments, which are not described in this paper, we have found, however, that floccular injection of 5-HT does not affect basic characteristics or adaptive changes of the VOR gain. The present evidence was achieved by injection of aselective P-adrenoceptor substances. The existence of P-adrenoceptors in the cerebellum is
495
supported by several studies, which involved binding methods (Minnemann et al., 1981; M. Pompeiano et al., 1989) as well as histofluorescence and autoradiographic methods (Palacios and Kuhar, 1980; 1982; Rainbow et al., 1984; Sutin and Minneman, 1985; Lorton and Davis, 1987). In the cerebellar cortex of rat, the majority of P-adrenoceptors is of the &-subtype (Minneman et al., 1979, 1981; M. Pompeiano et al., 1989). This subtype is prominently present in the molecular layer where, conversely, the density of the P,-subtype is low (Rainbow et al., 1984). Moreover, both PI- and &-receptors seem to surround the Purkinje cell somata, with a threefold relative density of pz with respect to @,-receptors (Sutin and Minneman, 1985). More recently, by using in situ hybridization histochemistry, it was shown that in the cerebellar cortex neurons expressing P,-adrenoceptors correspond to the granule cells (see Palacios, 1988). These findings, taken together, suggest that p-adrenoceptors identified in the molecular layer by autoradiography, may be located not only postsynaptically on Purkinje cells, but also presynaptically on the terminals of parallel fibers, axons of granule cells. As isoproterenol and sotalol bind aselectively to both PI- and @,-receptors, more experiments are needed to find out which P-receptor subtype is responsible for the changes in the adaptation of the VOR gain. Because the floccular injection sites were electrophysiologically identified and histologically verified after the last experiment, we can conclude that the effects found after injection of the P-adrenergic substances were elicited in the cerebellar flocculus. Spread of the injected substances to nearby areas was minimized by injection of 1 ~1 in 1 min, because low infusion volumes (Myers, 1966; Silberman et at., 1980) and infusion rates (Edwards and Hendrickson, 1981) are known to be essential to ensure anatomical specificity. According to the histological material, the dye solutions diffused to a maximum diameter of 3 mm in about 30 min, as judged from the frozen sections. Yet, it is impossible to say whether the injected
drugs might have affected other VOR-related areas such as the vestibular nuclei, because it is unknown over what distance and time the drug remain active. Nevertheless, we presume to be on the safe side when taking into consideration only those effects which started within 15 min after inject ion. We found quite some variability in adaptation of the VOR gain among rabbits as well as in the increase or decrease of the adaptation after injection of isoproterenol or sotalol. For example, some rabbits showed a prominent increase in gain in the experiments without injections, in the light as well as in darkness, whereas others displayed smaller increases. It may well be that the VOR gain change during adaptation depends partly on the resting discharge of the central noradrenergic neurons projecting to the cerebellum, and thus on the basic level of NA in the animals. However, we did not observe clear differences in the effects of injection of the P-agonist, depending on the division of the animals into high- and low-gain modifier groups, as described by McElligot and Freedman (1988a,b). The fact that the P-noradrenergic injections affected only the plasticity of the VOR, while the steady-state levels remained unaffected, is in agreement with several other investigations. It has been shown that depletion of the noradrenergic system abolishes the learning of a new task while earlier-learned tasks remained unaffected (Watson and McElligot, 1984). The present evidence is in line with the hypothesis on the modulatory role of the noradrenergic system in plasticity. Moreover, these findings fit nicely into the major theories of “cerebellar learning” which have been proposed to explain the occurrence of adaptive changes in the cerebellum. These theories suggest that adaptive changes are established by interaction between the climbing and the mossy fiber systems at the Purkinje cell level, presumably at the synaptic junctions between the parallel fibers and the Purkinje cell dendrites (Marr, 1969; Albus, 1971). According to Gilbert (1979, the noradrenergic
496
afferents to the cerebellum would be activated in order to consolidate the newly acquired information. Although no definitive proof for these hypotheses is available at the moment, they are certainly attractive and worth studying at the level of individual Purkinje cells which undergo plastic changes. Acknowledgements
These investigations were supported by the Foundation for Medical Research MEDIGON, grant no. 900-550-092, by NIH grant NS 07685-22 and by grants of the Minister0 dell’Universiti, and the Agenzia Spaziale Italiana, Roma, Italy. References Albus, J.S. (1971) A theory of cerebellar function. Math. Biosci., 10: 25-61. Collewijn, H. (1977) Eye-and head movements in freely moving rabbits. J. Physiol. (London), 266: 471-498. Dietrichs, E. (1988) Cerebellar cortical and nuclear afferents from the feline locus coeruleus complex. Neuroscience, 27: 77-92. Edwards, S.B. and Hendrickson, A. (1981) Neuroanatornical Tract-tracing Methods, Plenum Press, New York, 567 pp. Gilbert, P. (1975) How the cerebellum could memorize movements. Nature, 254: 688-689. Ito, M., Shiida, T., Yagi, N. and Yamamoto, M. (1974) Visual influence on rabbit horizontal vestibulo-ocular reflex presumably effected via the cerebellar flocculus. Brain Res., 65: 170-174. Ito, M., Jastreboff, P.J. and Miyashita, Y. (1979) Adaptive modification of the rabbit’s horizontal vestibulo-ocular reflex during sustained vestibular and optokinetic simulation. Exp. Brain Res., 37: 17-30. Ito, M., Jastreboff, P.J. and Miyashita, Y. (1982) Specific effects of unilateral lesions in the flocculus upon eye movements in albino rabbits. Exp. Brain Res., 45: 233-242. Jones, L.S., Gauger, L.L. and Davis, J.N. (1987) Anatomy of brain al-adrenergic receptors: In uitro autoradiography with [’2sI]-HEAT. J. Comp. Neurol., 231: 190-208. Keller, E.L. and Precht, W. (1978) Persistence of visual response in vestibular nucleus neurons in cerecellectomized cats. Exp. Brain Res., 3 2 591-594. Keller, E.L. and Smith, M.J. (1983) Suppressed visual adaptation of the vestibulo-ocular reflex in catecholamine depleted cats. Brain Res., 258: 323-327. Kmoto, Y., Satoh, K., Sakumoto, T., Tohyama, M. and Shimizu, N. (1978) Afferent fiber connections from the lower brain stem to the rat cerebellum by the horseradish
peroxidase method combined with M A 0 staining, with special reference to noradrenergic neurons. J. Hirnforsch., 19: 85-100. Kimoto, Y., Tohyama, M., Satoh, K., Sakumoto, T., Takahashi, Y. and Shimizu, N. (1981) Fine structure of rat cerebellar noradrenaline terminals as visualized by potassium permanganate “in situ” perfusion, fiwation and method. Neuroscience, 6: 47-58. Langer, T., Fuchs, A.F., Scudder, C.A. and Chubb, M.C. (1985) Afferents to the flocculus of the cerebellum in the rhesus macaque as revealed by retrograde transport of horseradish peroxidase. J. Comp. Neurol., 235: 1-25. Lorton, D. and Davis, J.N. (1987) The distribution of p l and P2-adrenergic receptors of normal and reeler mouse brain: An in L’itro autoradiographic study. Neuroscience, 23: 199210. Marr, D. (1969) A theory of cerebellar cortex. J. Physiol. (London), 202: 437-470. McElligot, J.G. and Freedman, W. (1988a) Vestibulo-ocular reflex adaptation in cats before and after depletion of norepinephrine. Exp. Brain Res., 69: 509-521. McElligot, J.G. and Freedman, W. (1988b) Central and cerebellar norepinephrine depletion and vestibulo-ocular reflex (VOR) adaptation. In: H. Flohr (Ed.), Post-lesion Neuronal Plasticity, Springer Verlag, Berlin, pp. 661-674. Miles, F.A. and Lisberger, S.G. (1981) Plasticity in the vestibulo-ocular reflex: A new hypothesis. Ann. Rev. Neurosci., 4: 273-299. Minneman, K.P., Hegstrand, L.R. and Molinoff, P.B. (1979) Simultaneous determination of P , and P,-adrenergic receptors in tissues containing both receptor types. Mol. Pharmacol., 16: 34-46. Minneman, K.P., Pittman, R.N. and Molinoff, P.B. (1981) p-adrenergic receptor subtypes: Properties, distribution and regulation. Ann. Rec. Neurosci., 4: 419-461. Miyashita, Y. and Watanabe, E. (1984) Loss of visual-guided adaptation of the vestibulo-ocular reflex after depletion of brain serotonin in the rabbit. Neurosci. Lett., 51: 177-182. Myers, R.D. (1966) Injection of solutions into cerebral tissue: Relation between volume and diffusion. Physiol. Behac., 1: 171-174. Nagao, S. (1983) Effects of vestibulo-cerebellar lesions upon dynamic characteristics and adaptation of vestibulo-ocular and optokinetic responses in pigmented rabbits. Exp. Brain Rex, 53: 36-46. Olson, L. and Fuxe, K. (1971) On the projections from the locus coeruleus noradrenaline neurons: The cerebellar innervation. Brain Rex, 28: 165-171. Palacios, J.M. (1988) Mapping brain receptors by autoradiography. ISI Atlas of Science: Pharmacology, 2: 71-77. Palacios, J.M. and Kuhar, M.J. (1980) p-adrenergic receptor localization by light microscopic autoradiography. Science, 208: 1378-1380. Palacios, J.M. and Kuhar, M.J. (1982) p-adrenergic receptor localization in rat brain by light microscopic autoradiography. Neurochern. Int., 4: 473-490. Pompeiano, M., Galbani, P. and Ronca-Testoni, S. (1989) Distribution of p-adrenergic receptors in different cortical
497 and nuclear regions of cat cerebellum, as revealed by binding studies. Arch. Ital. Biol., 127: 115-132. Pompeiano, 0. (1989) Relationship of noradrenergic locus coeruleus neurones to vestibulospinal reflexes. In J.H.J. Allum and M. Hulliger (Eds.), Afferent Control of Posture and Locomotion, Progress in Brain Research, Vol. 80, Elsevier, Amsterdam, pp. 329-343. Pompeiano, O., d'Ascanio, P., Horn, E. and Stampacchia, G. (1987) Effects of local injection of the a,-adrenergic agonist clonidine in the locus coeruleus complex on the gain of vestibulospinal and cervicospinal reflexes in decerebrate cats. Arch. Ital. Biol.,125: 225-269. Pompeiano, O., Manzoni, D., Barnes, C.D., Stampacchia, G. and d'Ascanio, P. (1990) Responses of locus coeruleus and subcoeruleus neurons to sinusoidal stimulation of labyrinth receptors. Neuroscience, 35: 227-248. Rainbow, T.C., Parsons, B. and Wolfe, B.B. (1984) Quantitative autoradiography of pi- and P,-adrenergic receptors in rat brain. Proc. Natl. Acad. Sci. USA, 81: 1585-1589. Silberman, E.K., Vivaldi, E., Garfield, J., McCarley, R.W. and Hobson, J.A. (1980) Carbachol triggering of desynchronized sleep phenomena. Enhancement via small volume infusions. Brain Res., 191: 215-224. Simpson, J.I., Graf, W. and Leonard, C. (1981) The coordinate system of visual climbing fibers to the flocculus. In A. Fuchs and W. Becker (Eds.) Progress in Oculomotor Research, Dei)elopments in Neuroscience, Vol. 12, Elsevier, Amsterdam, pp. 475-484. Somana, R. and Walherg, F. (1978) The cerebellar projection from locus coeruleus as studied with retrograde transport of horseradish peroxidase in the cat. Anat. Embryo!., 155: 87-94.
Sutin, J. and Minneman, K.P. (1985) Adrenergic @-receptors are not uniformly distributed in the cerebellar cortex. J. Comp. Neurol., 236: 547-554. van Neerven, J., Pompeiano, 0. and Collewijn, H. (1989) Depression of the vestibulo-ocular and optokinetic responses by intrafloccular microinjection of GABA, and GABA, agonists in the rabbit. Arch. Ital. Biol., 127: 243-263. van Neerven, J., Pompeiano, O., Collewijn, H. and van der Steen, J. (1990) Injections of @-noradrenergic substances in the flocculus of rabbits affect adaptation of the VOR gain. Exp. Brain Res., 79: 249-260. Waterhouse, B.D., Sessler, F.M., Cheng, J.T., Woodward, J.D., Azizi, S.A. and Moises, H.C. (1988) New evidence for a gating action of norepinephrine in central neuronal circuits of mammalian brain. Bruin Res. Bull., 21: 425-432. Watson, M. and McElligott, J.G. (1983) 6-OHDA induced effects upon the acquisition and performance of specific locomotor tasks in rats. Pharmacol. Behac., 18: 927-934. Watson, M. and McElligott, J.G. (1984) Cerebellar norepinephrine depletion and impaired acquisition of specific locomotor tasks in rats. Brain Rex, 296: 129-138. Woodward, D.J., Moises, H.C., Waterhouse, B.D., Hoffer, B.J. and Freedman, R. (1979) Modulatory actions of norepinephrine in the central nervous system. Fed. Proc., 38: 2109-2116. Young, W.S. and Kuhar, M.J. (1980) Noradrenergic a 1 and a 2 receptors: Light microscopic autoradiographic localization. Proc. Natl. Acad. Sci. USA, 77: 1696-1700.
This Page Intentionally Left Blank
SECTION V
Locus Coeruleus Influences on Higher Functions and Plasticity
This Page Intentionally Left Blank
C.D. Barnes and 0.Pompeiano (Eds.) Progress 117 Bruin Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
501
CHAPTER 35
Discharge of noradrenergic locus coeruleus neurons in behaving rats and monkeys suggests a role in vigilance G. Aston-Jones
’, C. Chiang
and T. Alexinsky
Dillision of Behacioral Neurobiology, Department of Mental Health Sciences, Huhnemann Unir‘ersity, Broad and Vine, Philadelphia, PA, U.S.A. and Uniuersitr‘Renk Descartes, Department de Psychophysiologie, Luhoratoire de Physiologie Nerueuse 2 CNRS, Gif-sur-Yvette, Cedex, France
’
Recordings from noradrenergic locus coeruleus (LC) neurons in behaving rats and monkeys revealed that these cells decrease tonic discharge during sleep and also during certain high arousal behaviors (grooming and consumption) when attention (vigilance) was low. Sensory stimuli of many modalities phasically activated LC neurons. Response magnitudes varied with vigilance, similar to results for tonic activity. The most effective and reliable stimuli for eliciting LC responses were those that disrupted behavior and evoked orienting responses. Similar results were observed in behaving monkeys except that more intense stimuli were required for LC responses. Our more recent studies have examined LC activity in monkeys performing an “oddball” visual discrimination task. Monkeys were trained to release a lever after a target cue light that occurred randomly on 10% of trials; animals had to withhold responding during non-target cues. LC neurons selectively responded to the target caes during this task. During reversal training, LC neurons lost their response to the previous target cue and began responding to the new target light in parallel with behavioral reversal. Cortical event-related potentials were elicited in this task selectively by the same stimuli that evoked LC responses. Injections of lidocaine, GABA, or a synaptic decoupling
solution into the nucleus paragigantocellularis in the rostra1 ventrolateral medulla, the major afferent to LC, eliminated responses of LC neurons to sciatic nerve stimulation or footor tail-pinch. This indicates that certain sensory information is relayed to LC through the excitatory amino acid (EAA) input from the ventrolateral medulla. The effect of prefrontal cortex (PFC) activation on LC neurons was examined in anesthetized rats. Single pulse PFC stimulation had no pronounced effect on LC neurons, consistent with our findings that this area does not innervate the LC nucleus. However, trains of PFC stimulation substantially activated most LC neurons. Thus, projections from the PFC may activate LC indirectly o r through distal dendrites, suggesting a circuit whereby complex stimuli may influence LC neurons. The above results, in view of previous findings for postsynaptic effects of norepinephrine, are interpreted to reveal a role for the LC system in regulating attentional state or vigilance. The roles of major inputs to L C from the ventrolateral and dorsomedial medulla in sympathetic control and behavioral orienting responses, respectively, are integrated into this view of the L C system. It is proposed that the LC provides the cognitive complement to sympathetic function.
Key words: locus coeruleus, vigilance, monkey, rat, behaving, cortex
Introduction
The noradrenergic locus coeruleus (LC) system has been proposed to be involved in almost as
many brain and behavioral phenomena as there are investigators who study this structure. These neurons have been implicated in fundamental brain processes ranging from vegetative activities
502
such as sleep and cerebral blood flow, to more complex, cognitive phenomena such as selective attention and memory (reviewed in Aston-Jones et al., 1984). Correspondingly, this system has been a favorite suspect in the “who done it” of clinical etiology, being implicated in disorders ranging from anxiety and panic to dementia and schizophrenia. How can so few cells (estimated to be 15,000 per hemisphere in humans) (Foote et al., 1983) do so many things? While some may think that investigators have waxed overly enthusiastic in some claims about this enigmatic little nucleus buried in the pontine brainstem, it is possible that the LC system serves a fundamental, general function in brain activity and thereby is involved in a number of brain and behavioral processes. The thesis of this paper is that, indeed, the LC provides a very general function, which is to regulate attention to the broad range of environmental stimuli and the degree to which behavior is engaged by the ever-changing sensory surround. The hypothesis to be developed in this report is that the LC system functions to control vigilance, defined as the surveillance of the environment, or readiness to respond to unexpected environmental events. We will develop this model using a host of cellular attributes of the LC system, particularly anatomic, physiological and pharmacological properties of these neurons. We will extend this hypothesis by incorporating recent findings pertaining to the neural systems that are afferent to LC neurons, and by using known functions of these major inputs to expand our thinking about the LC’s role in brain and behavior. We will also describe recent results of LC impulse activity in waking monkeys performing an “oddball” visual discrimination task designed to manipulate and measure vigilance. Finally, we will define a very general perspective from which to examine LC function, in which this system is viewed as a “random search generator” whose activation serves to disrupt stable activity in neural loops and associated behaviors and generate a search for a new activity that is more consistent with the most recently sampled sensory
information. This view may be useful to general nervous system models and neural network analyses. The ultimate goal of this functional analysis is to derive an algorithm for LC function, so that given a certain input, one could predict a functional outcome of LC system activation: We believe that the cellular physiological and anatomic properties of the LC system are key elements in such an analysis.
Background
Efferent organization The LC system has been the subject of intense study for more than two decades. The interest in this nucleus began in earnest when Swedish researchers (Dahlstrom and Fuxe, 1964; Ungerstedt, 1971) discovered that these cells give rise to an enormously divergent set of efferent projections; it is noteworthy that this small nucleus innervates more different brain areas than any other single nucleus yet described. This, and the fact that these neurons appeared to use the then recently discovered neurotransmitter norepinephrine (NE) to communicate with their target cells, generated great enthusiasm for understanding their possible function(s). Although some investigators have argued that NE may be released from LC fibers in a non-synaptic manner, providing a hormone-like, paracrine influence on many neurons within a diffusionlimited area (Beaudet and Descarries, 1978), more recent studies have shown that LC terminals in several brain structures make conventional synapse-like appositions with postsynaptic specializations on target neurons (Koda et al., 1978; Olschowka et al., 1981; Papadopoulos et al., 1989; Papadopoulos and Parnavelas, 1990). There is, in fact, a great deal of both regional and laminar specificity in the innervation of target structures by LC axons (e.g., Morrison et a/., 1982). Finally, although there have been reports of N E fiber apposition to blood vessels (Edvinsson et al., 1973; Hartman, 1973; Swanson et al., 1977), more re-
503
cent studies in LC terminal areas do not find a preference for apposition of dopamine-p-hydroxylase fibers on capillaries (Olschowka et al., 1981; Papadopoulos et al., 1989; Papadopoulos and Parnavelas, 1990). While such findings cannot rule out a possible involvement of LC projections in blood flow and metabolism in target areas, they indicate that this system is most prominently structured to provide conventional synaptic input to brain neurons.
Postsynaptic effects Using microiontophoresis, early studies (Hoffer et al., 1973; Segal and Bloom, 1974) found that N E inhibited basal discharge of cerebellar or hippocampal neurons in anesthetized rats, effects which appeared to be mediated by P-adrenergic receptor activation of a cyclic AMP second messenger system (Siggins et al., 1971; Foote er al., 1983). However, subsequent experiments by Foote, Segal and colleagues (Foote et al., 1975; Segal and Bloom, 1976) found that, in addition to decreasing basal discharge, N E may also enhance the selectivity of target cell discharge, so that in the presence of this neurotransmitter neurons respond with increased preference to their most strongly determined inputs. In these and studies by others that extended these findings (see contributions by Waterhouse et al., and by Woodward et al., this volume), N E acting at presumed p receptors decreased spontaneous impulse activity to a greater extent than activity evoked by afferent or sensory stimulation. This effect has recently been described for motor-related activity in primate cortex as well (Sawaguchi et al., 19901, indicating that it is not limited to sensory areas. It is noteworthy that in many cases N E has been found to augment evoked activity (either excitatory or inhibitory) while decreasing spontaneous discharge of the same neuron (Waterhouse and Woodward, 1980; Waterhouse et al., 1980, 1984). Such selective enhancement of responses to strong inputs relative to low-level or basal activity has been likened to an increase in the “signal-tonoise” ratio of target neurons by NE. Although
other effects of NE have .been described for various target areas, such biasing of target cells to respond preferentially to their strongest inputs is most significant for the present analysis.
LC discharge in unanesthetized rats and monkeys Before considering data concerning when LC neurons are active in behaving animals, it is pertinent to point out some important technical issues. Our recordings of LC discharge have utilized species (rat and monkey) whose LC is composed entirely of noradrenergic neurons. Thus, in these species one can record from known NEcontaining LC neurons by using simple histological verification of recording sites. This is an important consideration, since the wide interest in LC stems from its noradrenergic cell population. Similar experiments in other species (e.g., cat) in which the nucleus LC is composed of interdigitated N E and non-NE neurons could only positively ascribe discharge to NE-containing neurons if intracellular staining and double-labeling is carried out, a procedure not reported for any such study to date.
Spontaneous LC discharge and the sleep-waking cycle One predominant hypothesis of LC function is that these neurons control various stages of the sleep-waking cycle (Jouvet, 1969; Hobson et al., 1975; McCarley and Hobson, 1975). We found that spontaneous LC discharge covaries consistently with stages of the sleep-waking cycle, firing fastest during waking, more slowly during slowwave sleep, and becoming virtually silent during paradoxical sleep (PS) (Aston-Jones and Bloom, 1981a). In rat, the nearly total lack of activity in this nucleus during PS is evident not only from the consistent quiescence of single neurons, but especially when several neurons in the densely packed noradrenergic cell group are recorded simultaneously. In such cases, the entire population typically becomes silent, with a prominent
504
decrease in “background noise” as well. These observations, the first of their kind for known NE-containing neurons, support previous proposals that a similar subpopulation of unidentified cat LC neurons may be noradrenergic (Hobson et af., 1975; Rasmussen et al., 1986). However, other activity profiles of purported noradrenergic neurons have been reported in cat LC (Chu and Bloom, 1973, 1974). We also recorded spontaneously occurring field potentials from rat LC during sleep and waking. These slow potentials are synchronous with bursts of unit activity during waking and slow-wave sleep, but occur at their highest rates during PS in the absence of unit activity (Aston-Jones and Bloom, 1981a). These observations indicate that the absence of LC discharge during PS is due to active inhibition of these neurons, not simply disfacilitation. This is consistent with results demonstrating that LC cells in brain slices are auto-active, in the absence of synaptic inputs (Aghajanian et al., 1983; Williams et al., 1984). Further analysis revealed that LC impulse activity also changes within stages of the sleep-waking cycle, in anticipation of the subsequent stage (Fig. 1). Thus, during waking, LC neurons progressively decrease in activity as slow-wave sleep approaches, and likewise during slow-wave sleep before the onset of PS (Hobson et al., 1975; Aston-Jones and Bloom, 1981a). If waking rather than PS follows slow-wave sleep, LC neurons abruptly emit phasically robust activity 100-500 msec prior to waking. As indicated in Figure 1, the one exception to such stage-anticipation in LC discharge occurs for the PS-to-waking transition. Rat LC neurons return to waking activity either coincident with or slightly after the cessation of PS as measured by the EEG (theta activity). Thus, although anticipatory LC activity during most stage transitions is consistent with a role in generating the subsequent stage, this nucleus cannot be responsible for the termination of PS (Aston-Jones and Bloom, 1981a; Aston-Jones et af., 1984; however, see also Hobson et af., 1975). We have also monitored discharge of LC neu-
--
I IX -
(I
z =
10
-4-2 0 2 4 6
I
r
:
2 0 t-
-
S E C f r o m E M G transition PS
101-
.
w
-
l.40.J -. w L & I 80 80 9 40 80 0 4 8
I I
0
40
I
‘1. stage --+
w
I
sws
’
PS
I J
ISec---c W
Fig. 1. Locus coeruleus (LC) discharge rate during sleep-waking cycle (SWC) progression. Mean discharge rates for LC neurons in behaving rats during epochs normalized for the percentage of SWC stage completion are plotted consecutively for complete SWCs. Note that when paradoxical sleep (PSI-to-waking transitions are judged by the EEG (main plot), cellular activity does not anticipate the transition, and that LC cannot be the primary agent terminating PS. However, discharge is enhanced in anticipation of these same transitions scored by EMG criteria (inset). (From Aston-Jones and Bloom, 1981a.)
rons in unanesthetized, chair-restrained primates (Foote et al., 1980; Aston-Jones et al., 1988; Grant et al., 1988). Although these animals do not exhibit normal sleep and waking under our experimental conditions, we have observed LC activity during alertness and drowsiness as measured by EEG. As described above for rat LC, monkey LC neurons vary their activity closely with the state of arousal, even during unambiguous waking. Thus, periods of drowsiness are accompanied by decreased LC discharge, while alertness is consistently associated with elevated LC activity. Also as in rat, such changes in LC activity preceded the corresponding changes in EEG state by a few hundred msec. PS has not been observed in our chair-restrained monkeys. Spontaneous LC discharge and waking behacior We further observed that LC discharge is altered during certain spontaneous waking behaviors. During both grooming and consumption of a glucose solution, rat LC discharge decreases compared to that in other epochs of similar EEG arousal (Aston-Jones and Bloom, 1981a). Similar results were obtained for LC activity in behaving
505
primates (Grant et al., 1988). These results indicate that LC discharge is reduced not only for periods of low arousal (drowsiness or sleep), but also during certain behaviors (grooming and consumption) when animals are in an active waking, but inattentive (nonvigilant) state (see below). LC discharge also varies strongly with orienting behavior. In both rat (Aston-Jones and Bloom, 1981a,b) and monkey (Foote et al., 1980; AstonJones et al., 1988; Grant et al., 19881, the highest discharge rates we observed for LC neurons were consistently associated with spontaneous or evoked behavioral orienting responses. LC discharge associated with orienting behavior is phasically most intense when automatic, tonic behaviors (sleep, grooming, or consumption) are suddenly disrupted and the animal orients toward the external environment (Aston-Jones and Bloom, 1981a,b). Thus, as found following sleep, grooming, or consumption, there is close correspondence between spontaneous bursts of discharge and interruption of automatic, preprogrammed behaviors with an increase in attentiveness and vigilance (see below).
LC sensory responsiveness In addition to the above fluctuations in LC spontaneous discharge, we found that these neurons in unanesthetized rats and monkeys were responsive to non-noxious environmental stimuli (Fig. 2a,b; Foote et al., 1980; Aston-Jones and Bloom, 1981b). In waking rats, LC activity is markedly phasic, yielding short-latency (15-50 msec) responses to simple stimuli in every modality tested (auditory, visual, somatosensory, and olfactory). Responses were most consistently evoked by intense, conspicuous stimuli, though sporadic responses were also observed for nonconspicuous stimuli as well. These responses were similar for the different sensory modalities, and consisted of a brief excitation followed by diminished activity lasting a few hundred msec (Fig. 2a,b). While sensory responsiveness was qualitatively similar for LC neurons in rat and monkey, there
were important differences as well. In rat, any of a variety of intense stimuli evoked LC responses in a majority of sensory trials. In contrast, monkey LC was less strongly influenced by such stimuli, with responses fading after the first few trials. However, more complex stimuli, such as a new face or a meaningful but unexpected stimulus (see below), was consistently capable of eliciting LC responses in monkey (Aston-Jones et al., 1988; Grant et al., 1988). More generally, we found that stimuli effective in eliciting LC responses were also those that disrupted ongoing behavior and elicited a behavioral orienting response by the monkey. It appears, then, that the difference in stimulus-responsiveness in rat vs. monkey LC is closely related to the difference in behavioral responses evoked by stimuli in the two species. Compared to rats, monkeys require much more complex stimuli to interrupt behavior and evoke an orienting response; their LC responsiveness follows the same pattern. We quantified this linkage between behavioral disruption/orientation and LC sensory responses for rat LC neurons (Aston-Jones and Bloom, 1981b). As illustrated in Figure 2c, the largest responses were elicited by stimuli that caused an abrupt transition from sleep to waking, with associated behavioral orientation. Responses evoked during uninterrupted slow-wave sleep were much smaller in magnitude, whereas no response occurred during uninterrupted PS. In addition to these results for sleep, we found that response magnitudes during uninterrupted grooming or consumption of sweet water were reduced, whereas stimuli that disrupted such activity and generated orienting behavior elicited strong responses. Thus, there was a strong correspondence in rat, as in monkey, between sensory-evoked responsiveness in behavior and LC discharge, and a common factor for stimulus-responsivity in the two species is behavioral disruption and re-orientation. In sum, in both rat and monkey, stimuli that disrupt behavior and evoke an orienting response evoke LC response.
506
Habituation of LC responsiveness independent of habituation of behavioral state did not occur; response magnitudes for stimuli after 100 or more presentations were similar to those for initial stimuli when analyzed for similar behavioral states and orienting responses. . . . .. . . .. .. .. .. .. ., .. . . . ... . . , . . .. .. ... ... . ... .
.
LC neurons in monkeys respond to meaningful stimuli during an “oddball” discrimination / vigilance task The above results suggested that the essential property of stimuli to elicit LC responses was meaningfulness, so that intense stimuli elicited responses because their intensity made them meaningful but that non-intense, meaningful stimuli may also reliably elicit responses of LC cells. To explicitly test this possibility, we have recently begun recording LC activity in unanesthetized primates trained in an “oddball” visual
,
60.00
5
m
LO.00
a
-5 m
20.00
V 0
0.00 Time ( s e c ) b
16.001
z ;ii 12.00u a
II 1111 000
068
U
205
137
273
3L1
Tirne(sec) C
-
15.0
-
N
I
*
C
0 ._ +
trials
combined (NzL5)
(N-16)
(N:16)
(Nz16)
LfO
Fig. 2. Sensory responses of LC neurons are multimodal and state-determined. Panel a. Tone pip-evoked response of LC impulse activity in a behaving rat. Upper part is the oscilloscope recording of impulse activity, middle part is a raster display of activity for tone pip trials ordered consecutively from top to bottom, and lower part is a cumulative poststimulus time histogram (PSTH) accumulated for 50 tone pip trials (tone pips, 20 msec duration, presented at arrow for all parts of panel). Note phasic activation followed by postactivation inhibition (the latter is due to feedback mechanisms within LC) (Aghajanian, 1978; Aghajanian et al., 1977; Ennis and Aston-Jones, 1986). Panel b. PSTH of LC activity evoked by touching the tail of a behaving rat (touch to rostra1 tail, approximately as indicated at arrow). Note similarity of this response to that evoked by tone pips in panel a. Twenty-five trials accumulated in the PSTH. Panel c. Bar graph illustrating mean tone pip-evoked response magnitudes for rat LC neurons as a function of behavioral state. Note that responses are greatest for tones that awaken the animal (SWS/W trials), while tones presented during uninterrupted sleep (SWS/SWS trials) elicit substantially reduced responses; tones given during waking (W/W trials) elicit an intermediate response magnitude. Similarly, stimuli that interrupted grooming or consumption behaviors elicited large response in LC activity, and responses were reduced for similar stimuli that did not interrupt these behaviors. Response magnitude for SWS/W trials > W/W trials, **P> 0.0005; W/W magnitude > SWS/SWS magnitude, * * *P < 0.0005; SWS/W magnitude > magnitude for all trials combined > SWS/SWS magnitude, P < 0.0005 for each; paired t tests were used. (From Aston-Jones and Bloom, 1981b.)
507 ! a
II
d
i
I
I I I I I
I
I C
Fig. 3. PSTH displays for a cell in the LC area of a behaving monkey, illustrating selective response to (red) target stimulus. Panels a-c. A response is evoked by the red target stimuli (panel a) but not by correct lever release (panel b) or juice presentation (panel c). Increased activity preceding lever and juice in panels b and c correspond to stimulus presentation. Panels d and e. There is no response to the green non-target light (panel d) or to incorrect lever release to the green stimulus (panel e). Panel f. There is no response to low-intensity tone-pip presentation (see text). All stimuli and lever releases (b) occur at the arrows. Time calibration = 1 sec.
Recordings during reversal training revealed that these responses were specifically related to the meaningfulness of the stimuli, not to their physical attributes. As illustrated in Figure 4, after reversal training neurons in the LC region reversed their stimulus preference, so that responses were selectively elicited for the new S + (previous S - ) while responses for the old S + (new S - ) faded. A second period of reversal training rapidly re-established the original stimulus selectivity of primate LC neurons. Interestingly, these changes varied closely with behavioral performance, so that responses to the new S + increased (and responses to the new S decreased) as the percentage of correct behavioral responses to the new S + increased (and behavioral responses to the new S - decreased). In addition, cortical activity exhibited a similar set of properties. As shown in Figure 5 , AERPs recorded from the frontal and parietal cortices at latencies of 200-300 msec post-stimulation were selectively augmented by S cues, as reported by
+
Green
a
discrimination task (Aston-Jones et al, 1988; Alexinsky and Aston-Jones, 1990; Alexinsky et al., 1990). The task involves discriminating differently colored light cues for juice reward. A target stimulus (S + ) is presented on 10% of trials, intermixed in a semi-random fashion with nontarget lights of a different color (S - ). Neurons in the LC area were recorded along with cortical surface slow waves (averaged event-related potentials; AERPs) and behavioral responses (hits, misses, false alarms, and correct omissions). While some cells in the LC area showed responses that were purely sensory or motor in nature, most neurons exhibited activity specifically linked to the target stimulus. That is, responses for most cells were evoked selectively by S + stimuli but not S - stimuli, bar release or reward (Fig. 3).
:
Yellow
b
t
REVERSAL
C
Fig. 4. A reversal procedure in the “oddball” discrimination task reveals responses of a putatively noradrenergic neuron in the LC of a behaving monkey specific to meaningful stimuli. (a,b) PSTHs for response of a neuron to green (target), but not to yellow(non-target), stimuli. (c,d) Similar PSTHs for the same LC neuron but after reversal training, so that target stimuli are now yellow, and non-target stimuli are green. Note that green stimuli (c) no longer elicit responses, while yellow stimuli (d) now elicit a small response. Thus, the response is selectively elicited by meaningful stimuli. Stimuli at dotted lines in all panels; bar = 1 sec.
508
,
,
I
,
,
Y-
contribution by Foote et al. (this volume). Further studies are underway to better define the role of the LC system in such adaptive ,behavioral capacity.
, Percent #
1
Afferent circuits responsible for discharge properties of LC neurons
GC
Y-
Y’ G-
Fig. 5. Averaged event-related potentials (AERPs; 50 trials) for a monkey performing the “oddball” discrimination task, recorded from frontoparietal electrodes. AERPs at left were taken when the target was green (G +) and the non-target was yellow (Y - ); those on the right were taken for the opposite stimulus meanings, as indicated. Note that for both yellow and green stimuli, AERPs were larger when the stimulus was a target compared to its use as non-target.
others in both human (Hansen and Hillyard, 1984; Hillyard, 1985; Hillyard and Picton, 1979; Ruchkin et al., 1980) and non-human primates (see Foote et al., this volume); there is evidence that these potentials are similar to “P300” potentials in man (Pineda et al., 1987, 1988). During reversal training, the AERPs altered their selectivity in a manner similar to neurons in the LC area, to become selectively responsive to the new S + , and no longer respond to the previous S +(new S - ; Fig. 5). As with the neurons, these changes in cortical-evoked activity followed a time course that closely paralleled behavioral discrimination performance during reversal. Therefore, there is a close relationship among neurons in the LC area, cortical activity, and behavioral discrimination during a task requiring sustained attention to sensory cues (Alexinsky and Aston-Jones, 1990; Alexinsky et al., 1990). These results are consistent with the hypothesis that LC neurons function to promote attentiveness and adaptive behavioral responding to changing stimuli (Aston-Jones, 1985). These results also support the proposal that LC neurons may be responsible in part for the attention-related AERPs recorded in this paradigm. Additional evidence for this possibility is found in the
We have recently described the major afferents to the LC in rodent (Aston-Jones et a/., 1986, 1990); we detail these results elsewhere in this volume (Aston-Jones et al.). In brief, major afferents are found in two rostral medullary nuclei, the paragigantocellularis (PGi) in the ventrolatera1 rostral medulla, and the area of the prepositus hypoglossi (PrH) in the dorsomedial rostral medulla. Our stimulation and pharmacological analyses have revealed that the PGi predominantly excites LC cells via an excitatory amino acid projection, though inhibitory adrenergic projections exist as well. Conversely, the PrH potently and purely inhibits the LC by way of GABA projections. As a major excitatory input to LC, the PGi is a natural candidate for relaying the multimodal sensory-evoked activation of LC neurons described above. This possibility is supported by the parallel pharmacological sensitivity of LC responses evoked by PGi or by footpad (or sciatic nerve) stimulation. Broad spectrum EAA antagonists simultaneously attenuate both PGi- and sciatic-evoked responses, while antagonists of NMDA or cholinergic receptors have no effect on either response (Ennis and Aston-Jones, 1988). These results have now been replicated by several groups (Chen and Engberg, 1989; Rasmussen and Aghajanian, 1989a; Svensson et al., 1989; Tung et al., 1989). To test the hypothesis that sciaticevoked activation of LC is mediated through PGi, we (Chiang and Aston-Jones, 1989) recorded LC neurons while stimulating the contralateral footpad subcutaneously to activate the sciatic nerve, and slowly infused lidocaine (100-400 nl) into the PGi region. Such lidocaine infusions consistently blocked responses of LC neurons to sciatic nerve
509
r
Pre- inject ion
0
4
25
u
I
680 1360 Time ( m s e c )
30 sec
2040
posi-inject
a
u
8 1
0
25
m
1360
2040
1360
2040
r ,
0
Fig. 6 . Local microinfusion of lidocaine into paragigantocellulark (PGi) attenuates sensory response of an LC neuron. Upper panel. PSTH taken 2 min before microinjection. Note typical excitation elicited by stimulation of the rear footpad (for sciatic nerve activation, at arrow) followed by postactivation inhibition, which typically follows all LC neuronal excitation. Middle panel. PSTH of this neuron 30 sec after lidocaine (125 nl, 2% solution) was microinfused into PGi. Note abolition of response to footpad stimulation. Lower panel. PSTH of this LC neuron 10 min after microinfusion. Note partial recovery of response to footpad stimulation. Stimulation for all PSTHs = 2 mA, 0.5 ms pulses, presented at 0.5 Hz. Each PSTH is a collection of 50 stimuli. Similar attenuation of sciatic-evoked LC activity was obtained with microinjections of GABA or a Mn/Cd “synaptic decoupling” solution into PGi. (From Chiang and Aston-Jones, 1989.)
activation (Fig. 6). Similar infusions of GABA, or of a synaptic decoupling solution, 10 mM Cd++ plus 20 mM Mg++ (to antagonize Ca++ effects and prevent transmitter release), produced similar attenuation of footpad responses in LC neurons. These results indicate that PGi forms a critical synaptic link in this sensory response. It is
noteworthy that electrolytic lesions of the PGi area by others have failed to block sciatic-evoked activation of the LC (Rasmussen and Aghajanian, 1989a). This result may reflect topographic specificity within the PGi for LC-projecting neurons that mediate responses to sciatic stimulation. Indeed, infusions of lidocaine, GABA or the Cd++/Mn++ solution were all most effective when injected into the ventromedial retrofacial PGi area (Chiang and Aston-Jones, 1989). Therefore, electrolytic lesions of the PGi that spared this ventral, juxtaolivary region may fail to attenuate this sensory response of LC neurons. Further experiments are underway to test the hypothesis that other modalities of sensory responses in LC are also mediated by EAA inputs from the PGi. Indeed, experiments by others have found that the EAA pathway from PGi to LC mediates the LC response to systemic nicotine administration (Chen and Engberg, 1989; Engberg, 1989), while we (Akaoka et al., 1990, 1991) and others (Rasmussen and Aghajanian, 1989b) have found that hyperactivity of LC cells during opiate withdrawal is also mediated by this medullary afferent. Effects of excitatory amino acid antagonists on morphine withdrawal behaviors
It has long been known that LC neurons are hyperactive during morphine withdrawal (Korf et al., 1974; Aghajanian and Wang, 1987). In 1983 it was found that this hyperactivity does not occur in a slice preparation of LC neurons (Andrade et al., 1983). This indicated that withdrawal hyperactivity of LC neurons was not a consequence of altered opiate mechanisms within the LC, but instead may reflect a change in afferent control of the LC. Our recent studies revealing major afferents to the LC from the PGi and PrH suggested that one of these inputs may generate opiate withdrawal hyperactivity in the LC (Aston-Jones et al., 1990). Indeed, Rasmussen and Aghajanian (1989b) have found that lesions of the PGi, or antagonism of the amino acid
510
pathway from the PGi to the LC, blocked the morphine withdrawal-induced activation of the LC. We (Akaoka et al., 1990, 1991) and others (Tung et al., 1990) have confirmed their results As LC may play a role in certain opiate withdrawal symptoms, these results indicated that EAA antagonists may be effective in attenuating the opiate withdrawal syndrome. With this possibility in mind, we studied the effects of several EAA antagonists during withdrawal from mor-
phine. Rats were pretreated continuously for 6 days with morphine delivered from chronically implanted osmotic minipumps (Alza Corp.; 34 mg/kg/day). Animals were then given an EAA antagonist or vehicle and subsequently administered naltrexone (1 mg/kg, ip) to precipitate withdrawal. Several behavioral indices of opiate withdrawal were scored, including jumping, wet dog shakes, head shakes, teeth chattering, chewing, diarrhea, rhinorrhea, lacrimation, ptosis, and CNQX
Kynurenate
0
Vehicle KYN
E
T
I-
In
v
\ ul
* C
0 Vehicle 0 CNQX
10
U 0
C
0
t
0 Jumps
Jumps
Wet shakes Head shakes
CGS19755
AP5
40L
0 Vehicle
40L
Wet shakes Head shakes
0AP5
0 Vehicle
T
0 CGS19755 T
T
In v \
Y
20
3
s 5
f
10
0 Jumps
Wet shakes Head shakes
Jumps
Wet shakes Head shakes
Fig. 7. Effects of EAA antagonists on naltrexone-precipitated morphine withdrawal behaviors. Rats ( n = 2 for each drug tested) were administered morphine continuously over 6 days via osmotic minipump implanted subcutaneously. All drugs were administered icv 5 min before naltrexone injection (1.0 mg/kg, ip). The broad spectrum EAA antagonist kynurenate (11 mmol to produce about 32 mM in CSF), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) (3.5 nmol to produce about 10 mM in CSF), the specific NMDA antagonist 2-amino-5-phosophonopentanoate (AP5) (20 nmol to produce about 57 mM in CSF), and the more potent NMDA antagonist CGS19755 (1.2 nmol to produce about 3.5 mM in CSF),as indicated, all showed the ability to increase escape behavior and wet shakes above control levels, but did not affect other withdrawal signs such as head shakes, oral behavior, diarrhea, rhinorrhea, lacrimation, ptosis, and piloerection. Higher doses of each drug usually produced ataxia. Overall, none of these antagonists were effective in reducing signs of behavioral withdrawal. Approximate concentrations in CSF were calculated using a value of 350 ml for CSF volume. (From Ennis and Aston-Jones, 1988.)
511
piloerection. Results for some of these measures are summarized in Figure 7. Naltrexone alone, or with the vehicle control instead of an EAA antagonist, consistently elicited robust withdrawal in all of the behavioral measures. However, none of the EAA antagonists tested by intracerebroventricular (icv) (kynurenate, 2-amino-5-phosophonopentanoate (AP5), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), or CGS19755 (Lehmann et al., 1988)) or intraperitoneal administration (CGS19755) had consistent effects on most of the withdrawal signs. The only signs that were affected in these studies were jumping and wet dog shakes, both of which appeared to be increased by each of the EAA antagonists given icv (Fig. 7); further studies are needed to confirm these early results. Overall, these results clearly indicate that EAA antagonists of either the NMDA or non-NMDA receptor do not prevent naltrexone-precipitated withdrawal from morphine. Thus, central EAA neurotransmitters may not be importantly involved in these withdrawal signs. Also, LC hyperactivity during this state is attenuated by kynurenate (Rasmussen and Aghajanian, 1989b; Akaoka et al., 1990; Tung et al., 1990) or CNQX (Akaoka and Aston-Jones, 1991); this indicates that at least these components of the morphine withdrawal syndrome are not dependent on LC hyperactivity. This is consistent with observations (Britton et al., 1984) that lesions of the ascending NE projections from LC do not attenuate such behavioral signs of opiate withdrawal. Although LC may not mediate these behavioral manifestations of opiate withdrawal, its hyperactivity may be important for other withdrawal phenomena. Additional studies are needed to determine what components of opiate withdrawal are induced by, or are dependent upon, the hyperactivity seen in LC neurons. Effects of prefrontal cortex stimulation on LC activity
While these results seem adequate to explain certain properties of LC neurons, other charac-
teristics do not fit easily into this framework. In particular, it is difficult to understand how inputs from only two medullary structures could be responsible for the selective responsiveness of primate LC neurons to meaningful stimuli during discrimination behavior (described above). Such complex behavior is generally associated with cortical structures, yet there were no cortical inputs to LC in our tract-tracing analysis. The prefrontal cortex (PFC) has been linked previously with high level cognitive and attentional processes. Others (Arnsten and Goldman, 1984) have reported projections to the LC area of primates from the PFC. Although there were no retrogradely labeled cells in cortex following injections of tracers into LC, we examined descending projections of the PFC in rat using anterograde transport of WGA-HRP from the medial PFC (Chiang et al., 1987). Such injections yielded remarkably specific innervation patterns in the dorsal pons, with dense innervation of the central grey rostra1 and medial to LC but no fibers or terminals within central LC proper. As described in Aston-Jones et al. (this volume), this region is densely innervated by extranuclear dendrites of LC neurons. These results confirmed those reported for monkey (Arnsten and Goldman, 19841, and indicated that PFC could influence LC neurons via innervation of distal extranuclear dendrites, or less directly by innervation of neurons in the pericoerulear region that, in turn, may innervate LC or its extranuclear dendrites (see Aston-Jones et al., this volume). To examine the effect of PFC activity on LC neurons, we (Chiang et at., 1987) stimulated the medial PFC while recording single LC neurons in anesthetized rats. The most consistent response of LC neurons following cortical stimulation was antidromic activation (9 of 27 cells), as expected. In subsequent subjects, 6-hydroxydopamine was infused into the midbrain dorsal noradrenergic bundle 1 week prior to experiments to lesion ascending LC projections and eliminate this confounding effect. As illustrated in Figure 8a, lowfrequency stimulation of PFC in such animals
512
f
m
L
c a
ui
c
3 U 0
0
4 i
z I
m
L
(1,
a
40p 26r
680 1360 Time ( m s e c )
2040
1360
2040
2720 Time (msec)
4080
I
UI * C
3
131
u
1 0
t
“OTirne
(rnsec)
t L
c a
-
16
I
(0
3 C
u
c I
1360
Fig. 8. Train, but not low-frequency stimulation, of prefrontal cortex potently activates LC neurons. PSTHs illustrating responses of LC (a and c) and lateral dorsal tegmental neurons (b) to stimulation of prefrontal cortex (PFCx) in rats. Panel a. PSTH showing response of an LC neuron to PFCx stimulation at 0.5 stimuli/sec (1.0 mA). Twenty-three of 58 cells exhibited weak excitation at relatively long latencies (mean onset = 41 msec). In addition, 9/58 cells were weakly inhibited at long latencies (mean onset = 63 msec; not shown). Forty-four percent of LC neurons exhibited no response to PFCx stimulation. Panel b. PSTH using similar stimulation but recording from a cell in the lateral dorsal tegmental nucleus (LDT) where anterograde label is seen from PFCx injections. Note the more robust, shorter latency excitatory response. Panel c. Train stimulation of PFCx (10 pulses at 20 Hz) consistently produced significant activation of LC neurons (55/64 cells; mean onset = 120 msec, mean duration = 227 msec). Taken together, these data are consistent with an indirect or distal dendritic influence of PFCx on LC neurons. Stimulation at arrows in all PSTHs; 50 sweeps in panels a and b, and 25 sweeps in panel c.
yielded only weak effects on 23/58 LC neurons, and no significant effect on the remaining cells. In contrast to this weak influence on LC neurons,
Figure 8b shows that similar stimulation of PFC potently activated cells in the laterodorsal tegmental nucleus, in the same area that was heavily labeled by anterograde transport of WGA-HRP (described above). This indicated that the PFC was preferentially influencing neurons in pericoerulear areas that are densely innervated with PFC fibers and terminals, and only weakly (perhaps indirectly) influencing LC neurons. Consistent with this possibility, in additional experiments we found that train stimulation of the PFC activated LC cells much more potently than single-pulse stimulation (Fig. 8c). In addition, we found that infusions of the local anesthetic, lidocaine, into the PGi partially attenuated the influence of PFC stimulation on LC (Chiang et al., 1987). These results suggest that cognitive activity at least partially accesses the LC indirectly through the PGi; additional influence may also arise through possible connections onto distal dendrites or local “interneurons” in the pericoerulear region. These results, together with those indicating that PGi mediates at least some sensory responses of LC cells, indicate that brainstem connections play an extremely important and broad integrative role in regulating the outflow of the LC broadcast system. In addition, the cellular characteristics of these major afferents (summarized below) shed additional light on LC function, and how adaptive behavioral responses to a changing sensory environment are generated. A view of LC function based on cellular attributes: the vigilance / response initiation hypothesis
As outlined at the beginning of this article, a cellular anatomic and physiological understanding of the LC system requires knowledge of (1) the efferent projections of LC neurons, (2) the effects of NE released from LC terminals on target neuron activity, (3) the conditions under which LC neurons are active and releasing their transmitter, and (4) the factors controlling LC
513
discharge. When viewed together, these cellular properties have broad functional implications. First, the widespread efferent trajectory of the LC system implies that its function is a very global one, with physically distant and functionally disparate brain areas receiving innervation from (often individual) LC neurons. This notion is underscored by our physiological studies, revealing that LC neurons are markedly homogeneous in their discharge characteristics; for example, LC neurons throughout the nucleus exhibit very similar rates and patterns of spontaneous or sensory-evoked impulse activity (Aston-Jones and Bloom, 1981a,b). Thus, our data, in combination with the efferent anatomic results reviewed above, indicate that robust LC discharge results in globally synchronized release of N E onto target neurons located throughout the neuraxis. Postsynaptically, NE influences target cells so as to relatively promote responses to other, strong afferent input while reducing spontaneous or low-level activity. Such an enhancement of postsynaptic “signal-to-noise” ratios can lead to increased selectivity of target cell discharge to favor specific aspects of their response profiles, as discussed in this volume by Waterhouse, Woodward and others. In the context of these previous findings, the specific conditions of LC activation in unanesthetized behaving animals lead us to an hypothesis for LC function, suggesting a role of this system in the control of vigilance and initiation of adaptive behavioral responses (Aston-Jones and Bloom, 1981a,b; Aston-Jones et al., 1984). We have proposed that the LC is strongly influenced by two general classes of extrinsic afferents (L,.& possibly derived from two or more separate groups of neurons): excitatory inputs mediating sensoryevoked (or state transition-related) activity in LC neurons, and a more tonically active set of inhibitory afferents serving to modulate overall LC excitability in accordance with the vigilance state associated with the concurrent behavior. The more recent findings that PGi and PrH are major excitatory and inhibitory afferents to LC suggest
that they may provide these two classes of inputs; however, further work is needed to test this possibility. The level of LC activity at any time may be a consequence of the relative influence of each of these two classes of inputs. Strong tonic inhibition (such as found during PS) could serve to prevent LC neurons from responding to environmental stimuli during this state, so that LC inactivity permits PS to occur. Conversely, we propose that intense LC activity interrupts automatic, internally driven or vegetative behaviors (such as sleep, grooming, or consumption) that are incompatible with phasic, adaptive behavioral responding and instead engages a mode of activity characterized by a high degree of vigilance and interaction with diverse environmental stimuli. This theoretic framework is consistent with our observation that LC activity is most intense just before interruption of low-vigilance behaviors such as sleep, grooming or consumption, giving rise to alert orienting behaviors. Intense LC activation may occur when either tonic inhibition of LC neurons (engaged for automatic or vegetative behaviors) has suddenly decreased, or when excitation impinges on these cells (in response to a strong, unexpected sensory event) that is sufficiently intense to overcome concurrent tonic inhibitory inputs. Conversely, low vigilance programs may predominate in behavior when either LC discharge is effectively inhibited from responding to unexpected external stimuli, or when strong unexpected stimuli are not present in the environment. In this way, the LC may serve as a gate to determine the relative influences of two mutually exclusive sets of behavioral programs. In general terms, the LC may function to influence the overall orientation of behavior or mode of sensorimotor activities, to favor either automatic, vegetative behavioral programs, or phasic adaptive responding to salient environmental stimuli. An alternative, but equivalent, expression of this proposed role for the LC in the regulation of vigilance is a role in the initiation of adaptive
514
behavioral responses. Pronounced LC activity is associated with abrupt attention to external stimuli, which itself immediately precedes, and is a necessary component of, initiation of adaptive motoric response to salient external stimuli. Thus, in our analysis, the LC can logically fit into either sensory or motor functions, as it is not solely or directly related to either. Analysis of LC function in terms of state regulation is an alternative, and perhaps more appropriate, framework. This overall hypothesis of LC function can be stated in more abstract terms of nervous system operation. One view of heightened vigilance (e.g., startle, awakening, or stimulus-evoked disruption of ongoing behavior) is that this state is associated with conflicting patterns of neural activity, brought about, for example, by a disrupting stimulus that is inconsistent with (conflicts with) the set of stimuli that are predicted or expected to accompany the ongoing behavioral paradigm. The ensuing state of heightened vigilance consists, in this view (Fuller and Putnam, 19661, of a set of behaviors aimed at reducing or resolving this conflict, so that impinging stimuli are once again predicted by behavior. The mode of achieving this resolution involves investigating or exploring different behaviors in the animal’s repertoire (that may have had a weak relationship to a similar stimulus in the past). This “internal exploration” activity can be likened to a random search process, exploring the field of possible behavioral responses to the unexpected stimulus event. In our hypothesis, illustrated in Figure 9, robust LC discharge accompanying such a stimulus would engage a random search process by terminating ongoing low-vigilance activity and rearranging neural priorities via enhancement of signal-tonoise ratios of target neurons. This latter effect would result in preferential transmission of nervous information concerning salient stimuli and events, thereby favoring responses to the most urgent current stimuli. This proposed role of the LC as a random search generator is consistent with (and is simply a restatement of) the pro-
Behavior 1 -Random
search -Behavior2
n
I\
r i
Sensory
Integrative
Motor
LC
+
LC
I t
-
LC
f t I
f
Unpredicted stimulus
Fig. 9. Schematic illustrating proposed role of LC system as a random search generator. Left panel. Stable behavior represented by self-reinforcing (self-consistent) neural loops representing different aspects of a sensory-motor ensemble for a particular set of stimuli. Middle panel. An unexpected, meaningful stimulus occurs that activates LC (via PGi or prepositus hypoglossi (PrH)). The global release of norepinephrine from LC terminals suppresses low-level non-driven activity in neural loops or relatively enhances responses to phasically strong inputs, destabilizing and disrupting current ongoing behavior. This destabilized state by default “searches” for circuits that are sufficiently self-reinforcing and self-consistent to establish a new set of stable neural loops. Right panel. Activity driven by the currently strongest sensory events (and associated motor acts) establishes a new steady state condition of stable neural loop activity representing a new behavior o r state.
posed roles in vigilance regulation and adaptive response initiation.
Functional implications of major innervation of LC by PGi and PrH The new findings concerning afferent control of LC described here and in Aston-Jones et al. (this volume) have led to advances in understanding other properties of this system. One example is from the work of Engberg and colleagues who
515
found several years ago that systemic nicotine potently activates LC neurons via an unknown, indirect influence (Engberg and Svensson, 1980; Svensson and Engberg, 1980). Recent work has found that the PGi is the critical link in this response. Their evidence indicates that nicotine stimulates peripheral sensory (presumably visceral) afferents which in turn activate the excitatory amino acid pathway from PGi to LC (Hajos and Engberg, 1988; Chen and Engberg, 1989; Engberg, 1989). These results, and the close connection of PGi to visceral stimuli (via, e.g., vagal inputs to NTS to PGi) indicates that other drugs may affect the LC in a similar way, and suggests a new pathway for some psychopharmacological effects. One other example of potent drug effects on LC neurons that has been found to be mediated through the PGi is hyperactivity of LC neurons during withdrawal from morphine, described above. Such knowledge of the mechanisms of drug influences on LC neurons is a significant advance as it opens the way for modulation of these effects, which are thought to be important for the psychological and behavioral impact of many systemically administered drugs. The recent findings for major afferents to LC from the PrH and PGi have also prompted us to extend our theoretic framework to include functional attributes of these rostra1 medullary regions, as described below (AstomJones et al., 1990). The PrH is classically known to be a preoculomotor area. It has strong projections to oculomotor nuclei of the brainstem and many of its cells discharge closely in relation to eye movements (Baker, 1977; McCrea et al., 1979; Brodal, 1983; McCrea and Baker, 1985). However, as illustrated in Figure 10, this nucleus also has connections to pinnae motor areas (Henkel, 1981) and to vestibular nuclei that influence head movement (Cazin et al., 1982, 1984). Also, many PrH neurons exhibit activity that does not readily fit into a strict oculomotor framework (Lopez-Barneo et al., 1982; Lannou et al., 1984). It is important to note that the LC-projecting neurons in PrH are
Fig. 10. Functional attributes of the PrH. Many PrH neurons project to brain areas associated with ocular, pinnae and head movements, all components of orienting behavior. Thus, stimuli that are sufficiently intense to elicit an orienting response may do so in part through PrH circuitry, which at the same time may participate in the activation of LC at such times. This model predicts that LC-projecting neurons in PrH would decrease activity during orientation and release LC from tonic inhibition. Thus, LC would be disinhibited and prepotent for responding to stimuli at the same time that sensoria are oriented towards the most salient stimuli, helping to increase attentiveness to such stimuli. It is important to note that many aspects of this model remain to be tested.
restricted to the medial and perifasicular aspect of this nucleus, and that the PrH has received little attention in the rat. These properties, and the fact that the medial PrH is a major input to the LC, may indicate a somewhat broader function for the PrH (or, in particular, those PrH neurons that innervate the LC) than only oculomotor control. In an admittedly conjectural view, the PrH may be concerned with the initiation and
516
coordination of holistic orientation responses, rather than just the ocular components. In this framework (illustrated in Fig. lo), unexpected, urgent stimuli may influence the PrH to (i) orient the sensoria towards salient stimuli, and (ii) coordinate other central processes important in generating adaptive responses to imperative stimuli (e.g., increase LC excitability). As the PrH potently inhibits the LC via a GABA pathway (Ennis and Aston-Jones, 19891, this model predicts that the robust LC activity accompanying orienting behaviors results, at least in part, from disinhibition of LC from PrH (Aston-Jones et al., 1990). The PGi is a key sympathoexcitatory brain region, sending strong projections to directly innervate preganglionic sympathetic neurons in the spinal cord (Milner et al., 1988; Ross et al., 1981; Ruggiero et al., 1985). Thus, it is an important brain region for preparing the body to respond to urgent stimuli in the environment (defense response, “fight or flight” response) as sympathetic responses to such stimuli may be mediated, at least in part, through this area. As such unexpected or urgent stimuli are also the most reliable stimuli for activating LC neurons in rats or monkeys (described above), this function for the PGi suggests that it may be involved in activating LC neurons as well as peripheral sympathetic neurons in response to such stimuli. In fact, there is a remarkable temporal correlation between evoked LC discharge and sympathetic nerve activity (Elam et al., 1981, 1984, 1985, 1986; Reiner, 1986). This led us to test whether sensory responses of LC neurons are mediated through the PGi. Indeed, as described above, we found that blockade of the EAA pathway from the PGi eliminated responses to sciatic nerve activation (AstomJones and Ennis, 1988; Ennis and AstonJones, 19881, as did disruption of synaptic transmission within the PGi (Aston-Jones and Ennis, 1988; Chiang and Aston-Jones, 1989; Aston-Jones et al., 1990). Thus, as illustrated in Figure 11, it appears that the peripheral sympathetic system is activated in parallel with the central LC system
Fig. 11. Functional attributes of the PGi. The PGi is a key sympathoexcitatory region, reflecting its strong connections to sympathetic preganglionic neurons of the intermediolateral cell column (IML) of the spinal cord. It is known that stimuli that activate the peripheral sympathetic system also activate the LC (see text); we propose that this co-activation reflects parallel projections from PGi to IML and LC. Thus, activation of the peripheral sympathetic system prepares the animal physically for adaptive phasic responses to urgent stimuli, while parallel activation of LC increases vigilance and attentiveness, preparing the animal cognitively for adaptive responsiveness to such stimuli. It is proposed that LC serves as the cognitive limb of the global sympathetic nervous system, and that cognitive and peripheral sympathetic responses are integrated and coordinated through PGi.
by projections to both from the PGi area. Preliminary studies indicate that projections to sympathetic neurons and LC arise from separate but intermingled cells in PGi. Nonetheless, the PGi appears to be a key area for integration and coordination of activities in the LC and the sympathetic systems.
517
This analysis of the PGi has led us to extend our hypothesis of LC function, from serving to control vigilance to acting as the cognitive limb of a global sympathetic system, serving to optimize the animal’s behavioral state (via heightened attention to environmental stimuli) for making adaptive decisions concerning phasic behavioral responses at the same time that the peripheral sympathetic system prepares the animal physically to execute phasic responses to urgent stimuli (Fig. 11). One possible, though admittedly speculative, extension of our recent research concerns the nature of neural processing necessary to specify unexpectedness or novelty. The stimuli that best activate LC neurons possess the attributes of unexpectedness or novelty, and typically cause both a sympathetic and behavioral orienting response. By investigating the circuits whereby sensory stimuli are processed and transferred to PGi for activation of LC, we will begin to elucidate the neural mechanisms that are used to compute the stimulus qualities of expectedness, novelty and urgency. This has broad implications for neurobiological studies involving attention, stimulusgating and preparatory behavioral set. Acknowledgements
This work was supported by PHS grants NS24698, DA06214, ONR contract N00014-86-K-0493,and A F OSR grant 90-0147. References Aghajanian, G.K. (1978) Feedback regulation of central monoaminergic neurons: Evidence from single-cell recording studies. In M.B.H. Youdim, W. Lovenberg, D.F. Sharman and J.R. Lagnado (Eds.), Essays in Neurochemistry and Neuropharmacology, Wiley, New York. pp. 1-32. Aghajanian, G.K. and Wang, Y.Y. (1987) Common alpha 2and opiate effector mechanisms in the locus coeruleus: Intracellular studies in brain slices. Neuropharmacology , 26: 793-799 Aghajanian, G.K., VanderMaelen, C.P. and Andrade, R. (1983) Intracellular studies on the role of calcium in regulating the activity and reactivity of locus coeruleus neurons in uii.0. Brain Res., 273: 237-243.
Akaoka, H. and Aston-Jones, G. (1991) Opiate withdrawal-induced hyperactivity of locus coeruleus neurons is substantially mediated by augmented excitatory amino acid input. J. Neurosci. (in press). Akaoka, H., Drolet, G., Chiang, C. and Aston-Jones, G. (1990) Local, naloxone-precipitated withdrawal in the ventrolateral medulla activates locus coeruleus neurons via an excitatory amino acid pathway. Soc. Neurosci. Abstr., 16: 1027. Alexinsky, T. and Aston-Jones, G. (1990) Physiological correlates of adaptive behavior in the reversal of a light discrimination task in monkeys. Eur. J. Pharm. Suppl., 3: 149. Alexinsky, T., Aston-Jones, G., Rajkowski, J. and Revay, R.S. (1990) Physiological correlates of adaptive behavior in a visual discrimination task in monkeys. Soc. Neurosci. Abstr., 16: 164. Andrade, R., VanderMaelen, C.P. and Aghajanian, G.K. (1983) Morphine tolerance and dependence in the locus coeruleus: Single cell studies in brain slices. Eur. J. Pharmacol., 91: 161-169. Arnsten, A.F. and Goldman, R.P. (1984) Selective prefrontal cortical projections to the region of the locus coeruleus and raphe nuclei in the rhesus monkey. Brain Res., 306: 9-18. Aston-Jones, G. (1985) Behavioral functions of locus coeruleus derived from cellular attributes. Physiol. Psychol., 13: 118126. Aston-Jones, G. and Bloom, F.E. (1981a) Activity of norepinephrine-containing locus coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle. J. Neurosci., 1: 876-886. Aston-Jones, G. and Bloom, F.E. (1981b) Norepinephrinecontaining locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Aston-Jones, G. and Ennis, M. (1988) Sensory-evoked activation of locus coeruleus may be mediated by a glutamate pathway from the rostra1 ventrolateral medulla. In A. Cavalheiro, J. Lehmann and L. Turski (Eds.), Frontiers in Excitatory Amino Acid Research, A.R. Liss, New York, pp. 471-478. Aston-Jones, G., Foote, S.L. and Bloom, F.E. (1984) Anatomy and physiology of locus coeruleus neurons: Functional implications. In M. Ziegler and C.R. Lake (Eds.), Norepinephrine, Frontiers of Clinical Neuroscience, Vol. 2, Williams and Wilkins, Baltimore, pp. 92-116. Aston-Jones, G., Ennis, M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986) The brain nucleus locus coeruleus: Restricted afferent control of a broad efferent network. Science, 234: 734-737. Aston-Jones, G., Alexinsky, T. and Grant, S. (1988) Activity of locus coeruleus neurons in behaving primates: Relationship with vigilance. Soc. Neurosci. Abstr. 14: 407. Aston-Jones, G., Shipley, M.T., Ennis, M., Williams, J.T. and Pieribone, V.A. (1990) Restricted afferent control of locus coeruleus neurons revealed by anatomic, physiologic and pharmacologic studies. In C.A. Marsden and D.J. Heal (Eds.), The Pharmacology of Noradrenaline in the Central
518
Nerrvus System, Oxford University Press, Oxford, pp. 187247. Baker, R. (1977) The nucleus prepositus. In B. Brooks and F. Bajandas (Eds.), Eye Mocwnents, Plenum Press, New York, pp. 145-178. Beaudet, A. and Descarries, L. (1978) The monoamine innervation of rat cerebral cortex: Synaptic and nonsynaptic axons terminals. Neuroscience, 3: 85 1-860. Britton, K.T., Svensson, T, Schwartz, J., Bloom, F.E. and Koob, G.F. (1984) Dorsal noradrenergic bundle lesions fail to alter opiate withdrawal or suppression of opiate withdrawal by clonidine. Life Sci., 34: 133-139. Brodal, A. (1983) The perihypoglossal nuclei in the macaque monkey and the chimpanzee. 1. Comp. Neurol., 218: 257269. Cazin, L., Magnin, M. and Lannou, J. (1982) Non-cerebellar visual afferents to the vestibular nuclei involving the prepositus gypoglossal complex: An autoradiographic study in the rat. Exp. Bruin Res., 48: 309-313. Cazin, L., Lannou, J. and Precht, W. (1984) An electrophysiological study of pathways mediating optokinetic responses to the vestibular nucleus in the rat. Exp. Bruin Res., 54: 337-348. Chen, Z. and Engberg, G. (1989) The rat nucleus paragigantocellularis as a relay station to mediate peripherally induced central effects of nicotine. Neurosci. Lett., 101: 67-71. Chiang, C. and Aston-Jones, G. (1989) Microinjection of lidocaine, GABA or synaptic sidecouplers into the ventrolateral medulla blocks sciatic-evoked activation of locus coeruleus. Soc. Neurosci. Abstr., 15: 1012. Chiang, C., Ennis, M., Pieribone, V.A. and Aston-Jones, G . (1987) Effects of prefrontal cortex stimulation on locus coeruleus discharge. Soc. Neurosci. Ahstr., 13: 912. Chu, N.S. and Bloom, F.E. (1973) Norepinephrine-containing neurons: Changes in spontaneous discharge patterns during sleeping and waking. Science, 179: 908-910. Chu, N.S. and Bloom, F.E. (1974) Activity patterns of catecholamine-containing pontine neurons in the dorsolateral tegmentum of unrestrained cats. J. Neurohiol., 5: 527-544. Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. 1. Demonstration of monoamines in the cell bodies of brain stem neurons. Actu Physiol. Scand., 62: 5-55. Edvinsson, L., Lindvall, M., Nielsen, K.C. and Owman, C.H. (1973) Are brain vessels innervated also by central (nonsympathetic) adrenergic neurones? Bruin Res., 63: 496-499. Elam, M., Yao, T., ThorCn, P. and Svensson, T.H. (1981) Hypercapnia and hypoxia: Chemoreceptor-mediated control of locus coeruleus neurons and splanchnic, sympathetic nerves. Bruin Res., 222: 373-381. Elam, M., Yao, T., Svensson, T.H. and ThorCn, P. (1984) Regulation of locus coeruleus neurons and splanchnic, sympathetic nerves by cardiovascular afferents. Bruin Res., 290: 281-287. Elam, M., Svensson, T.H. and ThorCn, P. (1985) Differentiated cardiovascular afferent regulation of locus coeruleus neurons and sympathetic nerves. Bruin Rex, 358: 77-84.
Elam, M., Svensson, T.H. and ThorCn, P. (1986) Locus coeruleus neurons and sympathetic nerves: Activation by cutaneous sensory afferents. Bruin Res., 366: 254-261. Engberg, G. (1989) Nicotine induced excitation of locus coeruleus neurons is mediated via release of excitatoiy amino acids. Life Sci., 44: 1535-1540. Engberg, G. and Svensson, T.H. (1980) Pharmacological analysis of a cholinergic receptor mediated regulation of brain norepinephrine neurons. J. Neural. Trunsm., 49: 137-1 50. Ennis, M. and Aston-Jones, G. (1986) Evidence for self- and neighbor-mediated postactivation inhibition of locus coeruleus neurons. Bruin Res., 374: 299-305. Ennis, M. and Aston-Jones, G. (1988) Activation of locus coeruleus from nucleus paragigantocellularis: A new excitatory amino acid pathway in brain. J. Neurosci., 8: 36443657. Ennis, M. and Aston-Jones, G. (1989) GABA-mediated inhibition of locus coeruleus from the dorsomedial rostra1 medulla. J. Nrurosci., 9: 2973-2981. Foote, S.L., Freedman, R. and Oliver, A.P. (1975) Effects of putative neurotransmitters on neuronal activity in monkey auditory cortex. Bruin R e x , 86: 229-242. Foote, S.L., Aston, J.G. and Bloom, F.E. (1980) Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proc. Nutl. Acad. Sci. USA, 77: 3033-3037. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruelus: New evidence of anatomical and physiological specificity. Physiol. Rev., 63: 844-914. Fuller, R.W. and Putnam, P. (1966) On the origin of order in behavior. Gen. Syst., 11: 99-112. Grant, S.J., Aston-Jones, G. and Redmond, D.J. (1988) Responses of primate locus coeruleus neurons to simple and complex sensory stimuli. Bruin Res. Bull., 21: 401-410. Hajos, M. and Engberg, G . (1988) Role of primary sensory neurons in the central effects of nicotine. Psychophurmacology (Berlin), 94(4): 468-470. Hansen, J.C. and Hillyard, S.A. (1984) Endogenous brain potentials associated with selective auditory attention. Electroencephulogr. Clin. Neurophysiol., 49: 277-290. Hartman, B.K. (1973) The innervation of cerebral blood vessels by central noradrenergic neurons. In E. Usdin and S.H. Snyder (Eds.), Frontiers in Cutecholamine Research, Pergamon Press, New York, pp. 91-96. Henkel, C.K. (1981) Afferent sources of a lateral midbrain tegmental zone associated with the pinnae in the cat as mapped by retrograde transport of horseradish peroxidase. J. Comp. Neurol., 203: 213-226. Hillyard, S.A. (1985) Electrophysiology of human selective at ten tion. Trends Neurosci., 400-405. Hillyard, S.A. and Picton, T.W. (1979) Event-related brain potentials and selective information processing in man. In J. Desmedt (Ed.), Progress in Clinical Neurophysiology: Cognitice Components in Cerebral Euent-Refuted Potentials and Selectiise Attention, Karger, Basel, 319 pp. Hobson, J., McCarley, R. and Wyzinski, P. (1975) Sleep cycle oscillation: Reciprocal discharge by two brainstem groups. Science, 189: 55-58.
519 Hoffer, B.J., Siggins, G.R., Oliver, A.P. and Bloom, F.E. (1973) Activation of the pathway from locus coeruleus to rat cerebellar Purkinje neurons: Pharmacological evidence of noradrenergic central inhibition. J. Pharmacol. Exp. Ther., 184: 553-569. Jouvet, M. (1969) Biogenic amines and the states of sleep. Science, 163: 32-41. Koda, L.Y., Schulman, J.A. and Bloom, F.E. (1978) Ultrastructural identification of noradrenergic terminals in the rat hippocampus: Unilateral destruction of the locus coeruleus with 6-hydroxydopamine. Brain Res., 145: 190195. Korf, J., Bunney, B.S. and Aghajanian, G.K. (1974) Noradrenergic neurons: Morphine inhibition of spontaneous activity. Eur. J. Pharmacol., 25: 165-169. Lannou. J., Cazin, L., Precht, W. and Le Taillanter, M. (1984) Responses of prepositus hypoglossi neurons to optokinetic and vestibular stimulation in the rat. Bruin Rex, 301: 39-45. Lehmann, J., Hutchison, A.J., Mcpherson, S.E., Mondasori, C., Schmutz, M., Sinton, C.M., Tsai, C., Murphy, D.E., Steel, D.J., Williams, M., Cheney, D.L. and Wood, P.L. (1988) CGS 19755, a selective and competitive N-methylD-aspartate-type excitatory amino acid receptor antagonist. J. Pharmacol. Exp. Ther., 264: 65-75. Lopez-Barneo, J., Darlot, C., Berthoz, A. and Baker, R. (1982) Neuronal activity in prepositus nucleus correlated with eye movement in the alert cat. J . Neurophysiol., 47: 329-352. McCarley, R.W. and Hobson, J.A. (1975) Neuronal excitability modulation over the sleep cycle: A structural and mathematical model. Science, 189: 58-60, McCrea, R.A. and Baker, R. (1985) Anatomical connections of the nucleus prepositus of the cat. J. Comp. Neurol., 237: 377-407. McCrea, R.A., Baker, R. and Delgado-Garcia, J. (1979) Afferent and efferent organization of the prepositus hypoglossi nucleus. Prog. Brain Res., 50: 653-665. Milner, T.A., Morrison, S.F., Abate, C. and Reis, D.J. (1988) Phenylethanolamine N-methyltransferase-containing terminals synapse directly on sympathetic preganglionic neurons in the rat. Bruin Res., 448: 205-222. Morrison, J.H., Foote, S.L., O’Connor, D. and Bloom, F.E. (1982) Laminar, tangential and regional organization of the noradrenergic innervation of monkey cortex: Dopamine-P-hydroxylase immunohistochemistry. Brain Res. Bull., 9: 309-319. Olschowka, J.A., Molliver, M.E., Grzanna, R., Rice, F.L. and Coyle, J.T. (1981) Ultrastructural demonstration of noradrenergic synapses in the rat central nervous system by dopamine-beta-hydroxylase immunocytochemistry. J. Histochern. Cytochern., 29: 271-280. Papadopoulos, G.C. and Parnavelas, J.G. (1990) Distribution and synaptic organization of serotoninergic and noradrenergic axons in the lateral geniculate nucleus of the rat. J. Comp. Neurol., 294: 345-355. Papadopoulos, G.C., Parnavelas, J.G. and Buijs, R.M. (1989)
Light and electron microscopic immunocytochemical analysis of the noradrenaline innervation of the rat visual cortex [published erratum appears]. J. Neurocytol., 18: 1-10. Pineda, J.A., Foote, S.L. and Neville, H.J. (1987) Long-latency event-related potentials in squirrel monkeys: Further characterization of waveform morphology, topography, and functional properties. Electroencephalogr. Clin. Neurophysiol., 67: 77-90. Pineda, J.A., Foote, S.L., Neville, H.J. and Holmes, T. (1988) Endogenous event-related potential in squirrel monkeys: The role of task relevance, stimulus probability, and behavioral response. Electroencephalogr. Clin. Neurophysiol., 70: 155-171. Rasmussen, K. and Aghajanian, G.K. (1989a) Failure to block responses of locus coeruleus neurons to somatosensory stimuli by destruction of two major afferent nuclei. Synapse, 4: 162-164. Rasmussen, K. and Aghajanian, G.K. (1989b) Withdrawal-induced activation of locus coeruleus neurons in opiate-dependent rats: Attenuation by lesions of the nucleus paragigantocellularis. Brain Rex, 505: 346-350. Rasmussen, K., Morilak, D.A. and Jacobs, B.L. (1986) Single unit activity of locus coeruleus neurons in the freely moving cat. I. During naturalistic behaviors and in response to simple and complex stimuli. Brain Res., 371: 324-334. Reiner, P.B. (1986) Correlational analysis of central noradrenergic neuronal activity and sympathetic tone in behaving cats. Brain Rex, 378: 86-96. Ross, C., Armstrong, D., Ruggiero, D., Pickel, V., Joh, T. and Reis, D. (1981) Adrenaline neurons in the rostra1 ventrolateral medulla innervate thoracic spinal cord: A combined immunocytochemical and retrograde transport demonstration. Neurosci. Lett., 25: 257-262. Ruchkin, D.S., Sutton, S., Kietzman, M.L. and Silver, K. (1980) Slow wave and P300 on signal detection. Electroencephulogr. Clin. Neurophysiol., 50: 35-47. Ruggiero, D.A., Ross, C.A., Anwar, M., Park, D.H., Joh, T.H. and Reis, D.J. (1985) Distribution of neurons containing phenylethanolamine N-methyltransferase in medulla and hypothalamus of rat. J. Comp. Neurol., 239: 127-154. Sawaguchi, T., Matsumura, M. and Kubota, K. (1990) Catecholaminergic effects on neuronal activity related to a delayed response task in monkey prefrontal cortex. J. Neurophysiol., 63: 1385-1400. Segal, M. and Bloom, F.E. (1974) The action of norepinephrine in the rat hippocampus. I. Iontophoretic studies. Brain Res., 72: 79-97. Segal, M. and Bloom, F.E. (1976) The action of norepinephrine in the rat hippocampus. IV. The effects of locus coeruleus stimulation on evoked hippocampal unit activity. Brain Res., 107: 513-525. Siggins, G.R., Hoffer, B.J. and Bloom, F.E. (1971) Studies on norepinephrine-containing afferents to Purkinje cells of the cerebellum. 111. Evidence for mediation of norepinephrine effects by cyclic 3’3‘-adenosine monophosphate. Brain Rex, 25: 535-553.
520 Svensson, T.H. and Engberg, G. (1980) Effect of nicotine on single cell activity in the noradrenergic nucleus locus coeruleus. Acta Physiol. Scand., Suppl., 479: 31-34. Svensson, T.H., Engberg, G., Tung, C.S. and Grenhoff, J. (1989) Pacemaker-like firing of noradrenergic locus coeruleus neurons in uioo induced by the excitatory amino acid antagonist kynurenate in the rat. Acta Physiol. Scand., 135: 421-422. Swanson, L.W., Connelly, M.A. and Hartman, B.K. (1977) Ultrastructural evidence for central monoaminergic innervation of blood vessels in the paraventricular nucleus of the hypothalamus. Brain Res., 136: 166-173. Tung, C.S., Ugedo, L., Grenhoff, J., Engberg, G. and Svensson, T.H. (1989) Peripheral induction of burst firing in locus coeruleus neurons by nicotine mediated via excitatory amino acids. Synapse, 4: 313-318. Tung, C.S., Grenhoff, J. and Svensson, T.H. (1990) Morphine withdrawal responses of rat locus coeruleus neurons are blocked by an excitatory amino-acid antagonist. Acta Physiol. Scand.. 138: 581-582.
Ungestedt, U. (1971) Sterotaxic mapping of the monoamine pathways in the rat brain. Acta Physiol. Scand. Szcpp., 367: 1-48. Waterhouse, B.D. and Woodward, D.J. (1980) Interaction of norepinephrine with cerebrocortical activity evoked by stimulation of somatosensory afferent pathways in the rat. Exp. Neurol., 67: 11-34. Waterhouse, B.D., Moises, H.C. and Woodward, D.J. (1980) Noradrenergic modulation of somatosensory cortical neuronal responses to iontophoretically applied putative neurotransmitters. Exp. Neurol., 69: 30-49. Waterhouse, B.D., Moises, H.C., Yeh, H.H., Geller, H.M. and Woodward, D.J. (1984) Comparison of norepinephrineand benzodiazepine-induced augmentation of Purkinje cell responses to c-aminobutyric acid (GABA). J. Pharmacol. Exp. Ther., 228: 257-267. Williams, J.T., North, R.A., Shefner, S.A., Nishi, S. and Egan, T.M. (1984) Membrane properties of rat locus coeruleus neurones. Neuroscience. 13: 137-156.
C.D. Barnes and 0. Pompeiano (Eds.) Prugres.5 in Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
521
CHAPTER 36
Electrophysiological evidence for the involvement of the locus coeruleus in alerting, orienting, and attending S.L. Foote, C.W. Berridge, L.M. Adams and J.A. Pineda Department of Psychiatry (0603), School of Medicine, University of California, Sun Diego, La Jolla, CA, U.S.A.
In this chapter, we describe recent observations from our laboratory which support the thesis that the locus coeruleus ( L O , via its massively divergent efferent projections, participates in generating a generalized brain state that can be characterized as “alertness.” The first of these observations suggests that LC activation can convert the electroencephalographic (EEG) activity of the forebrain from patterns characteristic of a non-alert state to those characteristic of an alert state. The second observation indicates that LC activation alters sensory responses of individual neocortical neurons in a way that is compatible with the general thesis presented here, suggesting that LC-induced alterations in cortical neuronal activity may be an integral component of a hypothesized
participation of the LC in cortically mediated attentional processes. The third observation indicates that LC may modulate forebrain components of orienting responses that are indexed by event-related potentials (ERPs). Thus, the experiments described below involve electrophysiological assessment of forebrain information processing at three different levels of organization: activity of individual neurons in the millisecond range, neuronal ensemble activity persisting for 10-200 msec as indexed by ERPs, and ensemble/regional activity sustained for seconds to minutes as indicated by E E G measures. These observations suggest that alterations induced in forebrain function by manipulations of LC activity are evident at all three of these levels.
Key words: attention, cortex, hippocampus, EEG, arousal
Introduction
A number of observations suggest that the noradrenergic nucleus locus coeruleus (LC) participates in processes such as arousal and alerting. Many anatomical and physiological characteristics of the LC-noradrenergic system, including its massively divergent efferent projections and the ability of norepinephrine (NE) to enhance stimulus-elicited responses in target neurons, are compatible with this hypothesis (see Foote and Morrison, 1987). In addition, LC neurons exhibit en-
hanced discharge rates just before and during periods of arousal as indicated by behavioral and forebrain electroencephalographic (EEG) measures (Hobson et al., 1975; Foote et al., 1980; Aston-Jones and Bloom, 1981). However, these physiological and anatomical observations are correlational and do not provide strong demonstrations of a causal relationship between LC and forebrain activity. This chapter describes the effects of experimental manipulations of LC activity on various types of cortical electrophysiological measures.
522
These experiments begin to address the issue of whether changes in LC activity are necessary and/or sufficient to induce changes in these forebrain measures.
LC activation alters forebrain EEG characteristics In halothane-anesthetized rats, cortical EEG (ECoG) and hippocampal EEG (HEEG) typically exhibit activity similar to that of slow-wave sleep. However, periods of EEG activity closely resembling that seen during waking are sometimes observed spontaneously and are always observed following a stimulus such as a tail-pinch, even though the animal is still at a surgical level of anesthesia and does not overtly respond to any such stimulation (all of these phenomena are also observed in humans). We have utilized this preparation to ask whether activation of the LC would reliably alter forebrain E E G status (Berridge and Foote, in press). One advantage of the halothane-anesthetized preparation for such a study is that the “baseline” EEG status of the
LC TRIGGER
-
LC 2OGCOUNTS lOSEC 0-
,,L
animal is stable for prolonged periods, so that the effect of a particular manipulation can be reliablq assessed. Also, after a period of “arousal,” EEG signs usually return to their baseline levels so thal repeated tests can be performed. The LC is difficult to activate selectively bj means of electrical stimulation because it is small and surrounded by major fiber tracts originating from other nuclei. A method we had previouslq developed (Adams and Foote, 1988) was utilized in this study. A recording/infusion probe consisting of a microelectrode and infusion cannula wa$ lowered, using physiological landmarks, so thal the microelectrode was in the LC and the infusion cannula was 200-400 microns lateral to its lateral edge with the beveled tip oriented toward the LC. As in our previously published study, infusion of the cholinergic agonist bethanechol was used to repeatedly activate the LC for a few minutes at a time, as verified by recordings of the electrophysiological activity of individual or multiple LC neurons through the microelectrode. LC activation typically consisted of a peak increase in discharge rates of 3-5 times resting discharge
“,, /
*
Fig. 1. Relationship of LC activity to ECoG (top panel) and, in a separate experiment, HEEG (bottom panel) before, during, and after peri-LC bethanechol infusions. Bethanechol was infused at a constant rate throughout the interval indicated. EEG activity is shown in the top trace of each panel, the raw trigger output from LC activity in the middle trace, and the integrated trigger output (counts per 10-sec interval) in the bottom trace. In the top panel, LC activity is seen to increase during the latter part of the infusion, and several seconds later reduced amplitude and increased frequency become evident in the ECoG trace. As LC activity begins to decrease following the infusion, ECoG amplitude begins to increase and its frequency decreases. In the bottom panel, enhanced LC activity becomes evident in the latter part of the infusion period, and several seconds later theta rhythm begins to dominate the HEEG trace. For the remainder of the trace, LC activity remains elevated and theta rhythm predominates (From Berridge and Foote, in press.)
523
LC activation as a crucial mediating event in producing the EEG effects that follow the bethanechol infusions. These EEG changes have been quantified and statistically verified using power-spectrum analyses (see Fig. 2). It has long been known that LC electrophysiological activity is correlated with measures of forebrain EEG activation (see Foote et al., 1983 and Aston-Jones et al., this volume, for reviews). The present observations are especially interesting because they provide evidence that LC activation may initiate changes in cortical and hippocampal EEG patterns. Specifically, LC activation may be sufficient, although possibly not necessary, to induce ECoG desynchronization and a predominance of theta rhythm in HEEG activity. Obviously, however, certain caveats must apply to the interpretation of these results since they have been obtained from anesthetized preparations.
rates for a 3-5 min period. Our experience indicates that the infusion volumes, drug concentrations, and the infusion rates used activate a substantial majority of the LC neurons ipsilateral to the infusion. ECoG activity was recorded from sites in frontal neocortex (HEEG) and dorsal hippocampus. This experiment has now been performed in 28 animals with the following findings: (1) LC activation is consistently followed, within 5-20 sec, by ECoG desynchronization ( i e . , decreased amplitude and increased frequency) and a preponderance of theta activity in the HEEG (see Figs. 1 and 2); (2) if the recording/infusion probe is located so that the infusion is not effective in activating LC neurons (e.g., is placed more than 500-600 p M outside the LC), no such forebrain EEG effects are produced by the infusion; (3) following infusion-induced activation, forebrain EEG returns to pre-infusion patterns with about the same time course as the recovery of LC activity (10-20 min for complete recovery); (4) whether infusions are made from sites medial or lateral to LC, forebrain EEG changes invariably follow LC activation. These observations point to
t-
LC activation alters sensory responses of neocortical neurons This recent experiment (Adams and Foote, in preparation) was motivated by numerous studies,
PSA+
t---PSA 0 20
08
9.8
18.8
27.9
HZ
369
459
+
0 20
08
98
188
279 HZ
36.9
459
08
9.8
18.8
27.9
36.9
459
HZ
Fig. 2. Raw EEG data and power-spectrum analyses from the pre-infusion, post-infusion, and recovery periods from carbachol of a typical experiment. The exact ll-sec interval utilized for each power-spectrum analysis is indicated by the bar marked “PSA’ below each trace. There is an 80-second gap between the pre- and post-infusion traces, and a 5-min 20-sec gap between the post-infusion and recovery traces. The most striking post-infusion changes are the loss of activity at approximately 1 Hz in both traces and the emergence of dominant theta activity in the HEEG trace and power spectrum. (From Berry and Foote, in press.)
524
by ourselves and others, of the effects of norepinephrine (NE), the putative neurotransmitter released by LC axons, on presumed LC target neurons in neocortex, hippocampus, cerebellum, and numerous other brain sites (reviewed in Foote et af., 1983; Foote and Morrison, 1987). For example, in our own previous work, the effects of NE were tested on auditory cortex neurons which were activated acoustically by species-specific vocalizations in awake squirrel monkeys (Saimiri sciureus) (Foote et al., 1975). Five-barrel glass electrodes were used to record the activity of individual neurons in the superior temporal gyrus and to microiontophoretically apply NE or other putative neurotransmitters. Peristimulus-time histograms and raster displays of spontaneous activity and responses to the vocalizations were computed before, during, and after iontophoresis. With NE application, dose-dependent reduction of spontaneous and stimulus-evoked discharge was observed. For a given dose, the percent reduction of spontaneous activity was greater than the percent reduction of stimulus-elicited activity. Quantitative analysis also revealed that within periods of evoked activity, the percent reduction was inversely related to discharge rate, i.e., the percent reduction was least for episodes of most intense activity. We proposed that the presumed tonic inhibition resulting from synaptically released NE “would enhance the difference between background and stimulus-bound activity,” i.e., NE would act to enhance the “signal-tonoise” ratio of the cells. Extensive studies of the effects of NE on evoked and spontaneous activity have now been made in numerous other brain sites, both in uivo and in vitro, and other chapters in this volume detail much of this work. We wished to extend our previous observations by evaluating the effects of LC activation, rather than NE application, on sensory responses of neocortical neurons. These studies, conducted in halothane-anesthetized rats, involved the pharmacological activation of the LC, using the recording/infusion probe described above, during the simultaneous
recording of spontaneous and stimulus-elicited activity from neurons in primary somatosensory cortex. LC activation effects might be expected to differ from those of iontophoretically applied NE because the amount and spatial distribution of synaptically released NE would differ from that resulting from iontophoretic application. Also, LC activation would release NE in all the brain areas to which the LC projects, for example, somatosensory thalamus, and the effects measured at cortex would be the net result of NE release at all these sites. Presumably, LC activation more realistically mimics the effects of physiological activation of LC neurons than does NE iontophoresis. Our neocortical recordings were obtained from the hindlimb region of primary somatosensory cortex. Air-puff or electrical stimulation of the appropriate receptive field was used to activate somatosensory neurons. Peristimulus-time histograms were computed before, during, and after the repeated LC activations. Under baseline conditions, somatosensory neurons exhibited the type of stimulus-elicited activity previously reported by others. A shortlatency (15-22 msec), consistent, brief activation was followed by a longer duration (50-70 msec) suppression of firing (to below spontaneous levels) and a gradual return to baseline discharge rates (see Fig. 3). The effect of LC activation on each response component was consistent over infusions for a given animal and was also very similar in different animals. The initial, brief activation was reduced, the subsequent, longer-duration pause was converted into an activation, and spontaneous discharge rates were reduced. All of these cortical effects exhibited recovery when LC activity returned to baseline levels. In experiments designed to control for nonspecific effects of the LC infusion, it was found that lowering the infusion probe to a site 400 p m below the LC and doing the same infusions did not result in any effect on somatosensory responses. Also, systemic infusion of the a-adrenergic agonist clonidine, which inhibits LC activity,
525 RAT 60
RAT 90
RAT 00
RAT 230
I
BASELINE
LLL
INFUSION
RECOVERY
1
I 1000
I
2000
1
~ ~ ~ ~ ( m s e c )
I 1000
I
2000
1
TIME (msec)
I 1000
I
2000
~ ~ ~ ~ ( m s e c )
TIME (msec)
Fig. 3. Peristimulus-time histograms of cortical activity during baseline, infusion, and recovery conditions from NE. Note that although the overall levels of activity differ between animals, the nature of the baseline response is similar in each case, and LC activation affects each response similarly: the short-latency activation is reduced, the post-activation pause is changed to a sustained activation, and spontaneous activity is reduced. The S below the trace indicates the time of occurrence of the somatosensory stimulus. (From Adams and Foote, in press.)
in doses sufficient to neutralize the activation of LC also negated the infusion effects on cortical activity. Quantitative analyses revealed that the suppression of the initial peak, the reversal of the normal reduction in activity into a sustained enhancement of activity (above spontaneous levels), and the reduction in spontaneous activity were statistically significant (see Fig. 4). The absolute magnitude of the total response was increased, and the ratio of evoked activity to spontaneous activity was enhanced to an even greater extent. These observations indicate that LC activation and NE iontophoresis effects in cortex exhibit certain similarities and certain differences. The overall “signal-to-noise” effect is present with LC activation, but there is a loss of the temporal specificity of the phasic sensory response. The sharp onset and post-activation inhibition which clearly circumscribe the response under baseline conditions are reduced and reversed, respectively. This results in a response which is of greater
magnitude in terms of the absolute number of spikes elicited, and in an even greater enhancement of the ratio of evoked to spontaneous activity. These electrophysiological effects may have a behavioral counterpart in that there is sometimes a “trade-off” in arousal states whereby stimulus
-100‘
Short-Latency Excitation
N
Long-Latency
I
Total Evoked Total Evoked Response Spontaneous
PRE-DRUG 0 1-MIN AFTER 0 6-MIN AFTER N 12-MIN AFTER
Fig. 4. Changes induced by NE in somatosensory responses and in spontaneous cortical activity by LC activation. All rates are expressed as percentages of pre-infusion, spontaneous discharge rates except the far-right bar which is expressed as a percentage of post-infusion spontaneous rate. (From Adams and Foote, in preparation.)
526
detectability is enhanced, but aspects of discriminability are simultaneously reduced (e.g., see Posner, 1978). LC lesions alter monkey P3 event-related potentials
Event-related potentials (ERPs) are another electrophysiological index of forebrain information processing. They have been extensively studied in humans because they can be recorded non-invasively and offer the hope of sampling brain activity with temporal resolution in the millisecond range. In order to study the neural substrates of these potentials, we and others have developed animal models of particular ERP components. For example, we have previously demonstrated that monkeys exhibit certain ERP components with latency, polarity, and contingency similarities to those observed in humans (Neville and Foote, 1984; Pineda et al., 1987; 1988). These compounds will be indicated as positive (P) or negative (N) values following numbers corresponding to their latency in msec. One of the most intensively studied components of the human ERP is P300 or P3 which is elicited in response to novel and/or task-related stimuli. Two types of hypothesis regarding the neural origins of P3-like potentials, particularly P3b, are evident in the literature: those implicating a single neural source whose field potentials are volume conducted throughout large regions of the brain and those suggesting the existence of multiple sources. Single-source hypotheses have been challenged by recent evidence that steep potential gradients and polarity inversions in human P3-like activity can be recorded intracranially in
several sites, such as the frontal cortex (Wood and McCarthy, 19861, hippocampus and amygdala (Halgren et al., 1980; Wood et al,, 1980; McCarthy et al., 19821, and thalamus (Yingling and Hosobuchi, 1984). The alternative view of multiple active sites is also supported by the wide distribution of P3-like potentials recorded intracranially in animals (cingulate cortex: Gabriel et al., 1983; suprasylvian and marginal gyri: O’Connor and Starr, 1985). While the multiplesource hypothesis could account for a widespread distribution of P3 activity, it does not account for the uniformity of latencies at such spatially distributed sites. Therefore, some modification of the multiple-source proposition would appear to be necessary to formulate a more adequate hypothesis of P3 neurogenesis. One possibility is that a widely distributed neural system synchronously impacts on a number of forebrain areas. Several such candidate systems have been characterized, among them is the LC which exhibits the anatomic, physiologic, and functional properties (reviewed in Foote et al., 1983; Foote and Morrison, 1987) necessary to subserve such a role. The evidence concerning widely divergent LC projections, homogeneous activity of source neurons, transmitter effects on target neurons, and LC effects on EEG activity suggests that LC is activated during alerting or arousal, which leads to NE release onto target neurons in many brain regions. Thus, our hypothesis suggests that novel, surprising, and attentioneliciting events increase LC discharge activity and enhance “attentiveness.” In the context of the present experiment, we further hypothesize that such LC activation may be a necessary preccndition for the elicitation or enhancement of slow
Fig. 5. Sagittal sections through the brain stem of a control and a lesioned monkey (SM24). A. Section reacted with anti-dopaminep-hydroxylase (anti-DpH) from a normal monkey shows the darkly labeled noradrenergic cells of the LC nucleus and its anterior pole (AP). B. Section from lesioned monkey SM24, reacted with anti-DpH, is in approximately the same plane as the control section (A). The dashed lines demarcate the boundaries of the electrolytic lesion. C. Adjacent Nissl-stained section also showing the extent of the electrolytic lesion. The dorsal and rostra1 portions of the sections are to the top and left, respectively. BC, brachium conjunctivum; IV, trochlear nerve; dIV, decussation of trochlear nerve; mV, tract of mesencephalic nucleus of trigeminus. (From Lewis et al., 1987.)
521
528
interruption of dorsal bundle (DB) fibers (Pineda et al., 1989). Stimuli consisted of 2 and 6 kHz tone pips (40 msec duration, 60 dB above nHL) presented at 1 Hz in random order. In most sessions, one tone constituted 90% of the stimuli and the other tone lo%, while in some sessions
endogenous potentials or may directly produce synchronized slow electrical activity that may be evident as surface-recorded P3-like potentials. ERPs were recorded from untrained squirrel monkeys (Saimiri sciureus) twice a week for 4 weeks before and after bilateral LC lesions and
SM12
SM16
SM24
N106 N250-900
P372
-Pre LC Lesion 0
200
400
600
800
............ Post LC Lesion
msec Fig. 6. Averaged ERPs from individual subjects recorded at FZ, P3, and P4 in response to infrequent (10% probability) tones before (solid lines) and after (dotted lines) lesions. Subjects (SM12, SM16, SM24) with extensive cell-body and fiber lesions are shown. (Modified from Pineda et a/., 1989.)
529
tones were made equiprobable to test the effects of manipulating stimulus probability. LC and DB lesions were made under general anesthesia by first localizing the nucleus with a microelectrode, using electrophysiological landmarks, and then creating an electrolytic lesion. The electrode was then placed at the anterior pole of the nucleus and a knife cut effected. The extent of damage to LC perikarya and ascending axons was assessed by reconstructing lesions from Nissl-stained sagittal sections through the brain stems (see Fig. 5). The efficacy and selectivity of lesions in eliminating cortical noradrenergic axons were immunohistochemically verified utilizing antisera directed against dopamine-P-hydroxylase and tyrosine hydroxylase to label noradrenergic and dopaminergic axons, respectively. Prelesion ERPs resembled those previously reported for squirrel monkey in this paradigm (Neville and Foote, 1984; Pineda et al., 1987). As illustrated in the grand average ERPs for individual subjects shown in Figure 6, a large triphasic response centered over midfrontal areas was evident in the first 200 msec following either frequent or infrequent tones. This response consisted of a large positivity (mean latency, 52 msec) followed by a large negativity (mean latency, 106 msec). At longer latencies, and most prominently at lateral parietal electrodes, a broad positivity with a bipeaked morphology (mean latencies of 239 and 372 msec, respectively) was elicited in response to infrequent tones. A long-duration negativity was also recorded over frontal cortex (N250-900), temporally overlapping with the late positive components and peaking in the 500-600 msec range. Analyses of P52, P176, and N250-900 areas did not reveal any effects of, or interactions with, lesion type. In contrast, monkeys with damage to LC cell bodies and dorsal bundle fibers showed decreased P239 and P372 areas in the 90-10 blocks relative to their prelesion measures, while monkeys in which damage was restricted to DB fibers did not show such a marked decrease, (P239; lesion type X lesion F(1,3) = 10.44, P < 0.05; P372; lesion type X lesion, F(1,3) = 49.57,
P < 0.01). Monkeys with 40-70% of the LC damaged exhibited P239 and P372 area decreases greater than 60%. In contrast, little (14%) or no damage to the LC resulted in small decreases ( < 42%) or even increases in area. A Spearman rank-order correlation coefficient based on all subjects’ data indicated a statistically significant relationship between the extent of LC damage and the percentage decreases in P239 and P372 area ( r S= 0.90, P < 0.05). Several factors suggest caution in interpreting these data. First, despite the homogeneity of LC neurons and the small size of the electrolytic lesions, it is clear that areas in the trajectory of the microelectrode (e.g., cortex, colliculi), areas adjacent to the LC nucleus, and fibers of passage were also affected. Thus, systems other than the one targeted have been damaged, including systems possibly involved in the perception of sound. Second, the inherent variability in the extent of lesions plus the small number of subjects used reduces the power of any statistical evaluation. Several means of controlling for these factors are available and await future investigations. Investigations of these ERP observations suggest that the LC plays a modulatory role in the regulation of cortical responsiveness to sensory stimulation. LC activity may contribute to the production of a behavioral state in which novel sensory stimuli are more effectively processed, and this behavioral state may be a necessary precondition for the elicitation of long-latency, P3-like potentials. Alternatively, volleys of LC activity may directly produce slow waves that are recorded as P3-like activity. The decreased magnitude of monkey P3-like responses following lesions of the LC and ascending NE fibers is consistent with either of these hypotheses. Also, these results do not indicate whether NE-LC is both necessary and sufficient for the generation of such activity. Results from the present investigation are also consistent with behavioral studies, involving lesions of the LC or of DB fibers in rodents, which have reported decreases in startle responses
530
(Adams and Geyer, 1981), enhanced behavioral responsiveness to novelty (Britton et al., 19841, and increased distractibility and exploratory behavior toward novel objects (Roberts er d., 1976; Oke and Adams, 1978; Koob et al., 1984). The demonstration that P3-like generation or modulation may be dependent upon the integrity of the NE-LC system could also link the observations that patients with clinical disorders such as dementia, alcoholism, schizophrenia, and affective disorders (Schildkraut and Kety, 1967; Dongen, 1981; Mair and McEntee, 1983) exhibit pathology of the NE-LC, with demonstrations that these conditions are also associated with altered P3 components (Diner et al., 1985; Mirsky and Duncan, 1986; Polich et al., 1986). Interpretations and hypotheses
The possibility that LC activation can induce ECoG desynchronization and HEEG activity dominated by theta rhythm is compatible with a proposed involvement of the LC in “alerting.” Certainly these experiments must be extended to unanesthetized preparations in order for this proposal to receive more solid support. However, even were similar results obtained in unanesthetized preparations, two immediate complications would present themselves. First, dominant forebrain EEG “arousal” is evident during rapideye-movement (REM) sleep, a state during which LC neurons are silent (see Foote et al., 1983, for review). Second, there may be certain behavioral states characterized by the ECoG and/or HEEG, signs that appear to be induced by LC activation even though LC neurons are silent during these states (see Aston-Jones and Bloom, 1981). These observations possibly indicate that LC activation is sufficient to initiate these EEG signs, at least under some circumstances, but LC activation is not necessary for these signs to occur, possibly because other systems are also sufficient, but not necessary, for these processes. The effects of LC lesions on monkey P3 activity can be interpreted in light of our previous
neuroanatomical studies (Morrison and Foote, 1986) demonstrating a very dense NE innervation of monkey parietal cortex. Electrodes over this cortical region exhibited the largest P3 activity prior to LC lesions and showed the largest decrease following LC lesions. It is of interest to note the congruence between these monkey observations and the localization of an important alerting component of a “posterior attention system” in humans to a presumably homologous neocortical area using PET scanning methods (Posner and Petersen, 1990). Our studies of the effects of LC activation on neocortical sensory responsiveness also offer a possible mechanism for certain observations in studies of human attention. In studies of the effects of alertness on responding to discriminated stimuli, Posner (1978) has observed that enhanced alertness leads to more rapid responding, but that incorrect responses are also facilitated. This may be a correlate of our observation that neuronal responses are enhanced in magnitude but at the cost of some of the temporal and perhaps spatial resolution of these responses. The most obvious hypothetical scheme that would incorporate all three of the observations described above is that the LC participates in inducing alertness throughout many brain regions, as manifested by its ability to initiate ECoG desynchronization and intense theta rhythm in HEEG activity. This alert state is composed of changes in responsiveness of individual neurons throughout the brain, including sensory cortices, an example being the changes observed in the spontaneous discharge activity and sensory responses of somatosensory cortical neurons. The alert state is also a precondition, perhaps in some complex way, for subsequent appropriate orienting, attending or other information processing tasks. Acknowledgements
This research was supported by PHS Grant MH40008, AFOSR Grant 90-0325 and by a grant
531
from the John D. and Catherine T. MacArthur Foundation. References Adams, L.M. and Foote, S.L. (1988) Effects of locally infused pharmacological agents on spontaneous and sensoryevoked activity of locus coeruleus neurons. Bruin Res. Bull.. 21: 395-400. Adams. L.M. and Foote, S.L. Effects of locus coeruleus activation on spontaneous and sensory-evoked activity of somatosensory neocortical neurons. In preparation. Adams. L.M. and Geyer, M.A. (1981) Effects of 6-hydroxydopamine lesions of locus coeruleus on startle in rats. Psychopharmacvlv~,73: 394-398. Aston-Jones, G. and Bloom, F.E. (1981) Activity of norepinephrine-containing locus coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle. J. Neurosci., 1: 876-886. Berridge, C.W. and Foote, S.L. (1991) Effects of locus coeruleus activation on electroencephalographic activity in neocortex and hippocampus. Neuroscience, in press. Britton, D.R., Ksir, C., Britton, K.T., Young, D. and Koob, G.F. (1984) Brain, norepinephrine depleting lesions selectively enhance behavioral responsiveness to novelty. Physiol. Behui>.,33: 473-478. Diner, B.C., Holcomb, P.J. and Dykman, R.A. (1985) P300 in major depressive disorder. Psychiat. Res., 15: 175-184. Dongen, P.A.M. (1981) The human locus coeruleus in neurology and psychiatry. Progr. Neurvbiol. 17: 97-139. Foote, S.L. and Morrison, J.H. (1987) Extrathalamic modulation of neocortical function. Ann. Rec. Neurosci. 10: 67-95. Foote, S.L., Freedman, R. and Oliver, A.P. (1975) Effects of putative neurotransmitters on neuronal activity in monkey auditory cortex. Brain Res., 86: 229-242. Foote, S.L., Aston-Jones, G. and Bloom, F.E. (1980) Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proc. Natl. Acad. Sci. USA, 77: 3033-3037. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) The nucleus locus coerleus: new evidence of anatomical and physiological specificity. Physiol. Rec., 63: 844-914. Gabriel. M., Sparenborg, S.P. and Donchin, E. (1983) Macropotentials recorded from the cingulate cortex and anterior thalamus in rabbits during the “oddball” paradigm used to elicit P300 in normal human subjects. Svc. Neurvsci. Ahstr., 9: 1200. Halgren, E., Squires, N.K., Wilson, C.L., Rohrbaugh, J.W., Babh, T.L. and Crandall, P.H. (1980) Endogenous potentials generated in human hippocampal formation and amygdala by infrequent events. Science, 210: 803-805. Hobson. J.A., McCarley, R.W. and Wyzinski, P.W. (1975) Sleep cycle oscillation: Reciprocal discharge by two brainstem neuronal groups. Science 189: 55-58. Koob, G.F., Thatcher-Britton, K., Britton, D.R., Roberts, D.C.S. and Bloom, F.E. (1984) Destruction of the locus coeruleus or the dorsal N E bundle does not alter the
release of punished responding by ethanol and chlordiazepoxide. Physiol. Behau., 33: 479-485. Lewis, D.A., Campbell, M.J., Foote, S.L., Goldstein, M. and Morrison, J.H. (1Y87) The distribution of tyrosine hydroxylase-immunoreactive fibers in primate cortex is widespread but regionally specific. J. Neurosci., 7: 279-290. Mair, R.G. and McEntee, W.J. (1983) Korsakoffs psychosis: Noradrenergic systems and cognitive impairment. Behau. Brain Res., 9: 1-32. McCarthy, G., Wood, C.C., Allison, T., Goff, W.R., Williamson, P.D. and Spencer, D.D. (1982) Intracranial recordings of event-related potentials in humans engaged in cognitive tasks. Svc. Neurosci. Abstr., 8: 976. Mirsky, A.F. and Duncan, C.C. (1986) Etiology and expression of schizophrenia: Neurobiological and psychosocial factors. Ann. Rec Psychol., 37: 291-319. Morrison, J.H. and Foote, S.L. (1986) Noradrenergic and serotonergic innervation of cortical, thalamic, and tectal visual structures in old and new world monkeys. J. Cvmp. Neurol. 243: 117- 138. Neville, H.J. and Foote, S.L. (1984) Auditory event-related potentials in the squirrel monkey: Parallels to human late wave responses. Bruin Res. 298: 107-1 16. O’Connor, T.A. and Starr, A. (1985) Intracranial potentials correlated with an event-related potential, P300, in the cat. Bruin Res., 339: 27-38. Oke, A.F. and Adams, R.N. (1Y78) Selective attention dysfunctions in adult rats neonatally treated with 6-hydroxydopamine. Pharmacol. Bivchem. Brhai;., 9: 429-432. Pineda, J.A., Foote, S.L. and Neville, H.J. (1987) Long-latency event-related potentials in squirrel monkeys: Further characterization of waveform morphology, topography, and functional properties. Electroencephalogr. Clin. Neurophysid., 67: 77-90. Pineda, J.A., Foote, S.L., Neville, H.J. and Holmes, T. (1988) Endogenous event-related potentials in monkeys: The role of task relevance, stimulus probability, and behavioral response. Electroencephalogr. Clin. Neurophysiol., 70: 155171. Pineda, J.A., Foote, S.L. and Neville, H.J. (1989) Effects of locus coeruleus lesions on auditory, long-latency, event-related potentials in monkey. J. Neurosci., 9: 81-93. Polich, J., Ehlers, C.L., Otis, S., Mandell, A.J. and Bloom, F.E. (1986) P300 latency reflects the degree of cognitive decline in dementing illness. Electroencephalogr. Clin. Neurvphysiol., 63: 138-144. Posner, M.I. (1978) Chronometric Explorations of Mind, Erlbaum, Englewood Heights, New Jersey, 271 pp. Posner, M.I. and Petersen, S.E. (1990) The attention system of the human brain. Ann. Rec. Neurosci., 13: 25-42. Roberts, D.C.S., Price, M.T.C. and Fibiger, H.C. (1976) The dorsal tegmental noradrenergic projection: Analysis of its role in maze learning. J. Cvmp. Physivl. Psycho/., 90: 363-372. Schildkraut, J.J. and Kety, S.S. (1967) Biogenic amines and emotion. Science, 156: 21-30. Wood, C.C. and McCarthy, G. (1986) A possible frontal lobe contribution to scalp P300. In J.W. Rohrbaugh, R. John~
532 son, Jr. and R. Parasuraman (Eds.), Eighth International Conference on Event-Related Potentials of the Brain (EPIC CIII): Research Reports, Stanford, CA. Wood, C.C., Allison, T., Goff, W.R., Williamson, P.D. and Spencer, D.D. (1980) On the neural origin of P300 in man. Prog. Brain Res., 54: 51-56.
Yingling, C.D. and Hosobuchi, Y.A. (1984) Subcortical correlate of P300 in man. Electroencephalogr. Clin. Neurophysiol., 59: 72-76.
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Bratn Research, Vol. 88 0 1991 Ebevier Science Publishers B.V.
533
CHAPTER 37
The role of noradrenergic locus coeruleus neurons and neighboring cholinergic neurons of the pontomesencephalic tegmenturn in sleep-wake states B.E. Jones Department of Neurology and Neurosurgery, McCill University, Montreal Neurological Institute, University Street, Montreal, Quebec, Canada
Despite early suppositions that the noradrenergic (NA) locus coeruleus (LC) neurons play a critical role in the generation and tonic maintenance of wakefulness and paradoxical sleep, further studies indicated that these cells play a nonessential modulatory role in the regulation of these states. Thus, based upon evidence from pharmacological, lesion and single-unit recording studies, it now appears that NA neurons may be important for enhanced periods of attention or stress during wakefulness, though they are not necessary for the tonic maintenance of cortical activation or behavioral arousal during that state. From similar examinations, it has been found that the cessation of activity of NA LC neurons may
~
~~
normally be important in permitting the occurrence of the state of paradoxical sleep. Neighboring cholinergic neurons of the pontomesencephalic tegmentum may also be active during waking and play a role in facilitating thalamocortical activity and transmission, like NA neurons during that state. However, unlike the NA neurons, the cholinergic neurons play an active and essential role in the generation of the state of paradoxical sleep. Generation of the state of paradoxical sleep may depend upon the simultaneous activation of cholinergic neurons and cessation of NA LC neurons, that could be brought about by the intermediary action of local GABA neurons.
~~
Key words. waking, paradoxical sleep, noradrenaline, acetylcholine, GABA
Introduction
With the prior knowledge that neurons within the pontomesencephalic tegmentum were integrally involved in the initiation and maintenance of the states of waking and paradoxical sleep (Moruzzi
and Magoun, 1949; Jouvet, 19621, Jouvet was struck in the early 1960’s by the identification of noradrenergic neurons within the locus coeruleus (LC) nucleus (Dahlstrom and Fuxe, 1964) and their distribution through the dorsolateral pontine tegmentum in the cat (Pin et aL, 1968; Jones,
534
Fig. 1. Schematic representation of the distribution of noradrenergic neurons (open circles) and cholinergic neurons (solid circles) in the pontomesencephalic tegmentum of the cat. The cells were plotted from adjacent sections processed by peroxidase-antiperoxidase (PAP) immunohistochemistry for tyrosine hydroxylase (TH) and choline acetyl transferase (ChAT), respectively. (From Jones and Beaudet, 1987.)
53s
1969). In a pioneering multidisciplinary approach involving neuropharmacological, neurochemical, and neurophysiological techniques, he and his colleagues were to bring evidence that the noradrenergic LC neurons could be critically involved in mechanisms of wakefulness and paradoxical
sleep (see for review Jouvet, 1972). These original studies revealed the potential importance of these neurons, and their neighbors, in the mediation of sleep-wake states, even though subsequent research altered the conclusions of these early studies in a progressive unveiling of the precise role
Abbreviations (Fig. 1) AQ bc bic bP CB cll CLR CNF CP crf CS CV dbc dic DR dsc dts FTC FTG FTL FTP IC IP
KF LC LDT 11 LLD LLV Sm 5M SME MG
ml mlf mp 3 3n 4n
aqueduct of Sylvius brachium conjunctivum brachium of the inferior colliculus brachium pontis cerebellar cortex commissure of the lemniscus lateralis of Probst central linear raphe cuneiform nucleus cerebral peduncle central reticular fasciculus central superior nucleus ventral cochlear nucleus decussation of the brachium conjunctivum decussation of the inferior colliculus dorsal raphe nucleus decussation of the superior colliculus dorsal tegmental decussation central tegmental field gigantocellular tegmental field lateral tegmental field paralemniscal tegmental field inferior colliculus interpeduncular nucleus Kolliker-Fuse nucleus locus coeruleus nucleus laterodorsal tegmental nucleus lateral lemniscus dorsal nucleus of the lateral lemniscus ventral nucleus of the lateral lemniscus tract of the mesencephalic trigeminal nucleus motor trigeminal nucleus mesencephalic trigeminal nucleus medial geniculate medial lemniscus medial longitudinal fasciculus mammillary peduncle oculomotor nucleus 3rd nerve 4th nerve
4 5n 6n 7n 8n P SP PAC PB PBG pf PG pp PPT R rb rc RM RPo RR rs
sc SNC SNR SOL SOM sst T tb TD TR TV uf V4 vm
VMN VS vsc VTA
trochlear nucleus 5th nerve 6th nerve 7th nerve 8th nerve pyramidal tract principal sensory trigeminal nucleus periaqueductal gray parabrachial nuclei parabigeminal nucleus predorsal fasciculus pontine gray pes pedunculi pedunculopontine tegmental nucleus red nucleus restiform body reticulocerebellar fibers raphe magnus nucleus raphe pontis nucleus retrorubral nucleus rubrospinal tract superior colliculus substantia nigra pars compacta substantia nigra pars reticulata lateral nucleus of the superior olive medial nucleus of the superior olive spinal trigeminal tract nucleus of the trapezoid body trapezoid body dorsal tegmental nucleus tegmental reticular nucleus ventral tegmental nucleus uncinate fasciculus 4th ventricle vestibulomesencephalic fibers medial vestibular nucleus superior vestibular nucleus ventral spinocerebellar tract ventral tegmental area
536
that the noradrenergic neurons may play in the modulation of states (see for review Jacobs and Jones, 1978). Distribution of noradrenergic neurons in the pontomesencephalic tegmentum of the cat
Differing from that in the rat brain, the distribution of noradrenergic neurons in the cat brain was found to encompass a broad area of the dorsolateral pontine tegmentum (Fig. 1; Pin et al., 1968; Jones, 1969; Maeda et al., 1973; Jones and Moore, 1974). The LC complex, which would correspond to both the principal LC and the subcoeruleus nuclei (including pars alpha, dorsal and ventral) in the rat, includes noradrenergic neurons dispersed within the periventricular gray and through the parabrachial nuclei (Jones and Moore, 1974). These cells are most numerous and concentrated in the caudal pons, but also extend rostrally into the oral pons to the level of the isthmus. Here they extend into the region of the laterodorsal tegmental nucleus and pedunculopontine tegmental nucleus (Fig. 1). More recently with the development of immunohistochemical staining for choline acetyl transferase (ChAT), cholinergic neurons have also been identified in the dorsolateral pontomesencephalic tegmentum of the cat (Fig. 1; Jones and Beaudet, 1987). Differing again from that in the rat brain, the distribution of cholinergic cells in the cat brain encompasses a broad area that overlaps with that of the noradrenergic neurons in the pons. The ChAT-immunoreactive cells are concentrated within the laterodorsal and pedunculopontine tegmental nuclei in the caudal mesencephalon and oral pons but extend caudally as well into the region of the LC and parabrachial nuclei. In the oral pons, where the overlap of these two populations is the greatest, the cholinergic neurons are intermingled with the noradrenergic neurons within the periventricular gray and tegmentum beneath the gray (Fig. 1; P2). Thus, whereas in the rat the two cell groups are segregated into discrete nuclei, in the cat they are
intermixed in overlapping regions. On the other hand, the dendritic processes of the two cell types appear in close proximity within the periventricular gray (in the region of the LC and LDT) and tegmentum beneath the gray (in the region of the LC, and LDT,) in both species (see B.E. Jones this volume). The role of noradrenergic and cholinergic pontomesencephalic neurons in waking
From early pharmacological studies, it was known that central catecholamines could enhance and prolong arousal and wakefulness (Carlsson et al., 1957). In the first studies in which electrolytic lesions were placed in the region of the ascending noradrenergic pathway through the midbrain from the LC, it was found that the amount of waking (as measured polygraphically over 24 h recording periods) was significantly reduced, and it was correlated with the amount of noradrenaline remaining in the forebrain (Jones el al,, 1973). Less marked but significant reductions in waking were also found in the rat after 6-hydroxydopamine-induced, and therefore more selective, destruction of the ascending noradrenergic bundle (Lidbrink, 1974). Pharmacological depletion with a-methylpara-tyrosine (AMPT) of the catecholamines also led to decreases in the amount of waking that were less severe than those produced by lesions, but nonetheless significant (King and Jewett, 1971). These results converged to suggest that noradrenergic neurons of the LC by virtue of their diffuse innervation of the central nervous system and cerebral cortex, represented an integral component of an ascending activating system (Jones et al., 1973) that is critical for the tonic maintenance of cortical activation of waking. Unfortunately these original conclusions were not substantiated by further studies aimed at more direct proof of the importance of LC cells for the state of waking. Whereas, lesions placed in the most rostra1 tip of the noradrenergic neurons in the cat, the region from which the ascending fibers also collected, led to a decrease in
537
waking and an increase in sleep (Jones et al., 19731, lesions focused in the more caudal region, where the noradrenergic cells are concentrated within the LC and parabrachial nuclei, did not reduce the amount of waking in the cat (Jones et al., 1977). The latter lesions were associated with a mean depletion of 85%, and greater than 95% in two cats, of noradrenaline in the cerebral cortex. It was thus necessary to conclude that although noradrenergic LC neurons may play a role in enhancing and prolonging cortical activation of waking, as pharmacological evidence has suggested, they are not necessary for the tonic maintenance of wakefulness. The disproof of an essential role for noradrenergic neurons in waking necessitated a reinterpretation of the effects of previous lesions in the oral pons and midbrain that had resulted in a reduction in waking. Obviously, this reduction must have been due to the destruction of neurons other than, or in addition to, the noradrenergic neurons. In fact, with the discovery of cholinergic neurons in a more rostra1 yet overlapping position to the noradrenergic neurons in the dorsolatera1 pontomesencephalic tegmentum, it appeared likely that these cholinergic cells could be involved in activating processes. From the classical studies of Shute and Lewis (1967) utilizing acetylcholinesterase staining, ascending pathways of putative cholinergic neurons had been traced into the forebrain from the reticular formation, such as to suggest the existence of an “ascending cholinergic reticular system.” Indeed, pharmacological studies had also revealed that cortical activation was dependent upon cholinergic transmission (Longo, 1966). In an attempt to destroy cholinergic neurons in a selective manner relative to noradrenergic neurons of the dorsolateral pontomesencephalic tegmentum, a more recent study was undertaken employing kainic acid as a neurotoxin in this region (Jones and Webster, 1988; Webster and Jones, 1988). This toxin had previously been shown to destroy effectively cholinergic neurons in the forebrain, yet to have little effect upon
noradrenergic LC neurons in the pons (Kohler and Schwarcz, 1983). Multiple microinjections of kainic acid into the dorsolateral pontomesencephalic tegmentum of cats, which were previously operated for implantation of chronic indwelling electrodes for polygraphic recording, produced a broad area of cell loss and gliosis through the ChAT-TH cell area (Fig. 1). Across 11 cats, a mean destruction of the majority (60%) of the ChAT-immunoreactive neurons was effected with a range of 25-85%. The destruction of the cholinergic perikarya was accompanied by a decrease or loss of the cholinergic innervation of the forebrain and brainstem. In the same animals, a mean loss of the minority (35%) of the noradrenergic neurons occurred with a range of 0-50%. Destruction of the majority of pontomesencephalic cholinergic neurons did not reduce the amount of waking as measured polygraphically in any of the cats, including those with the greatest cell destruction (Fig. 2). These results thus also necessitated the conclusion that cholinergic pontomesencephalic neurons, like noradrenergic neurons, may play a role in wakefulness and arousal, as pharmacological results would suggest, but that they are not necessary for the tonic maintenance of cortical activation and wakefulness. In fact, both of these systems would represent small components of a very massive system comprised of hundreds of thousands of neurons within the pontomesencephalic reticular formation that project forward into the forebrain (Jones and Yang, 1985) and represent the ascending reticular activating system, as it was originally conceived by Moruzzi and Magoun (1949) and their colleagues (Lindsley et al., 1950). Based upon single-unit recording studies through the dorsolateral pontomesencephalic tegmentum of the freely behaving cat, the noradrenergic LC neurons are known to be active during waking and to fire at highest rates in demanding situations (Rasmussen et al., 1986; see also Jacobs et al., this volume). As had also been found to be the case in the rat (Aston-Jones
538 EEG
EEG
EEG
Fig. 2. Trivariate computer-graphic representation of one day’s recorded minute epochs projected and rotated in a simulated three-dimensional space from a normal cat (top row) and the same cat two weeks following a lesion of the dorsolateral pontomesencephalic tegmentum (bottom row). Transformed and scaled values of average E E G amplitude, average.EMG amplitude and PGO spike rate per minute were projected by computer graphics onto a two-dimensional plane and rotated 42” to the left and right to reveal the position of the points in three-dimensions. Top: In the normal cat, the data points are distributed into three major groups and partitioned by cluster analysis into three clusters that have been shown to correspond to traditionally classified wakefulness (W, small dots), slow-wave sleep (SWS, open circles) and paradoxical sleep (PS, closed circles) (Friedman and Jones, 1984). Bottom: Two weeks following neurotoxic lesions of the dorsolateral pontomesencephalic tegmentum that destroyed the majority of the cholinergic neurons and the minority of the noradrenergic neurons in the region (Fig. 1). Whereas the clusters of points indicative of waking and slow-wave sleep are still present in similar number and position (on the EEG amplitude axis), the cluster corresponding to paradoxical sleep is absent. Similarly, epochs with P G O spiking in association with low-voltage EEG and low-voltage E M G (neck muscle atonia) were not present in the polygraphic record, indicating a total loss of the state of paradoxical sleep during two weeks post-lesion. (From Jones, 1990c; including data from Webster and Jones, 1988.)
and Bloom, 1981a and b; see also Aston-Jones et al., this volume) and monkey (Foote et al., 19801, it would appear that the activity of these cells may be important under particular conditions demanding a high level of attention and/or eliciting a certain degree of stress. In this regard, the central noradrenergic neurons may, like peripheral sympathetic noradrenergic neurons which
they resemble from an anatomical point of view (see B.E. Jones, this volume), be important in the response of the organism to challenging situations. Years ago, Cannon and his colleagues (1929) determined that the peripheral sympathetic nervous system was not necessary to the life of the animal in the laboratory milieu, yet that it would be important for the adaptation and
539
survival of the animal within its natural, more challenging environment. Similarly, the noradrenergic LC neurons may be neither necessary nor sufficient for cortical activation or wakefulness, yet very important for the adaptability and survival of the animal in its normally challenging environment. Although the cholinergic neurons are close neighbors of the noradrenergic neurons in the dorsolateral pontomesencephalic tegmentum and, accordingly, may receive similar afferent input in both the cat and rat (see B.E. Jones, this volume), their efferent projections are very different, those of the former being more similar to the parasympathetic nervous system, in their widespread but nonetheless limited distribution in the rat (Hallanger et al., 1987; Hallanger and Wainer, 1988; Jones and Cuello, 1989; Jones, 1990a) and the cat (Jones and Webster, 1988; Park et al., 1988; Smith et al., 1988). The major forebrain target of the pontomesencephalic cholinergic neurons is the thalamus where, not unlike the noradrenergic neurons, they innervate the midline and intralaminar thalamic nuclei and, also, the lateral geniculate, anterior nuclei and the nucleus reticularis. Acetylcholine also has a similar action to noradrenaline in the thalamus, enhancing excitability and transmission within thalamocortical projection systems (McCormick, 1989; see McCormick et al., this volume). Few recording studies have as yet been performed in order to describe the behavioral context of the activity of these cells, yet putative cholinergic neurons have been found to be active during waking and most active during attentive or active periods of waking (El Mansari et al., 1989; Steriade et al., 1990). Like the noradrenergic neurons, these cells fire at a low, tonic rate during quiet waking, and thus may participate (albeit in a nonessential manner) in the maintenance of activity as well as in the enhancement of activity in thalamocortical systems. Whereas the behavior and role of the two may be very similar during waking, the profile of activity of the cholinergic neurons appears to be very different from that of noradrenergic neurons
through the remainder of the sleep-wake cycle and most notably during paradoxical sleep, when their roles may in fact be in opposition rather than in synergism. The role of noradrenergic and cholinergic pontomesencephalic neurons in paradoxical sleep
Early transection and lesion studies had identified the pontine tegmentum as the important region of the brainstem for the generation of paradoxical sleep (Jouvet, 1962). It was subsequently found by Jouvet and Delorme (1965) that electrolytic lesions of the region of the noradrenergic LC neurons within the dorsolateral pontine tegmentum resulted in a loss of paradoxical sleep and/or the muscle atonia that accompanies this state (Jouvet, 1972). It was thus concluded that the LC neurons could represent the executive neurons for the generation of paradoxical sleep and the motor inhibition that occurs during that state. Unfortunately these original results and conclusions did not bear up to subsequent experimental tests involving more selective lesions, pharmacological manipulations and single-unit recording. First, lesions more limited to the principal LC (Henley and Morrison, 1974) or to the noradrenergic neurons concentrated caudally within the LC and parabrachial nuclei (Jones et al., 1977) did not have the same long-lasting effect of larger lesions that included neighboring structures in the pons. Second, pharmacological depletion of the catecholamines with AMPT did not reduce paradoxical sleep but instead increased it, in association with the hypersomnia produced by this drug (King and Jewett, 1971). Third, single-unit recording actually revealed that the putative noradrenergic LC neurons ceased firing during the state of paradoxical sleep in the cat (Hobson et al., 1975). The latter result was confirmed in the rat in which the identification of noradrenergic LC neurons is almost certain, given the homogeneity of the noradrenergic cells within the compact portion of the nucleus in this species
540
(Aston-Jones and Bloom, 1981b). Thus, it appeared that instead of generating paradoxical sleep, the activity of the noradrenergic LC neurons could actually prevent or would in any event not be consonant with the appearance of paradoxical sleep. To explain the effects of lesions within the area of the noradrenergic LC neurons, it became evident that other non-noradrenergic perikarya or fibers located within or passing through the dorsolateral pontomesencephalic tegmentum could be responsible for the generation of this state and its loss following large electrolytic or thermolytic lesions of the area. With the discovery of cholinergic neurons in an overlapping position to the noradrenergic neurons, the possibility arose that these cells and/or their processes were important in the state of paradoxical sleep. Such a possibility was substantiated by early pharmacological evidence that cholinergic transmission was important for the appearance of paradoxical sleep (Jouvet, 1962; Domino et al., 1968; Hazra, 1970). Moreover, local injection of the combined muscarinic/nicotinic agonist, carbachol, into the pontine tegmentum has been shown repeatedly to elicit a state of paradoxical sleep, including neck muscle atonia in association with rapid eye movements and EEG desynchrony (George et d.,1964; Baghdoyan et al., 1984). In studies employing kainic acid to destroy the cholingeric neurons of the dorsolateral pontomesencephalic tegmentum (above), it was found that there was a marked effect upon the state of paradoxical sleep (Fig. 2). Immediately following the lesions, there was a loss of the state of paradoxical sleep in those animals with extensive destruction of the cholinergic neurons. In all animals recorded over a four-week period after the lesion, there were certain paradoxical sleep-like components which reappeared in the polygraphic record during the third to fourth weeks. Thus, during a limited number of 1 min epochs, small, isolated PGO spikes and eye movements appeared in association with a low-voltage fast-EEG activity indicative of incipient paradoxical sleep.
TABLE 1 Correlation between the extent of pontomesencephalic lesions and paradoxical sleep a Lesion Extent
%PSb
Total Volume Ch Cell Area Volume ChAt + Cells TH + Cells
-0.53 - 0.72 * 0.69 * -0.18
PGOSpike Rate -0.32 - 0.59 0.66 * -0.27
EMG AmP 0.64 * 0.73 * - 0.41 0.15
The correlation of histological variables with paradoxical sleep variables was examined by Pearson’s Product Moment Correlation for 11 cats (*, P < 0.05, r > 0.60). Taken from Webster and Jones (1988). PS as percent of total recording time on day 28 post-lesion. Average PGO spike rate during PS. Average EMG amplitude (uv) during PS. Total Volume (mm3) of tissue affected in pons and mesencephalon. Volume (mm3) of tissue affected within the cholinergic cell area. Number of ChAT-positive neurons. Number of TH-positive neurons.
a
Neck muscle atonia was not present during these epochs or during any period of the post-lesion recording, although a reduction of EMG amplitude was present in some cats during paradoxical sleep-like epochs. In order to assess the importance of the cholinergic as compared to the noradrenergic and other neurons in the region of the neurotoxic lesion in the degree of recovery of the state of paradoxical sleep and its component variables, the degree of correlation between area or cell destruction and PS or its components was assessed (Table 1). First, the amount of paradoxical sleep and both the PGO spike rate and EMG amplitude were all correlated significantly with the volume of the lesion within the ChAT-TH cell area of the pontomesencephalic tegmentum. For the first time, these results showed that the cells (and not simply the fibers of passage) in this region are critically involved in the generation of the state of paradoxical sleep and its component variables. Second, within this area, the number of cholinergic neurons, and not that of the noradrenergic neurons, were significantly correlated with the state of paradoxical sleep and PGO spike rate. Thus, these results clearly indicated
541
the importance of cholinergic neurons in the dorsolateral pontomesencephalic tegmentum and in the generation of paradoxical sleep and its phasic phenomena. It may be by their extensive innervation of other neurons within the pontomedullary reticular formation, as well as the forebrain, that the cholinergic neurons could generate the state of paradoxical sleep (Jones, 1990b,c). Conclusions Despite the evident importance of cholinergic neurons in the generation of paradoxical deep, as brought out by pharmacological and now lesion results, they represent one component of a necessarily complex system that underlies the generation of a state. As originally revealed by pharmacological results (Karczmar ef al., 1970) and subsequently by single-unit recording studies (McGinty et al., 1974; Hobson et aL, 1979, the pharmacological depletion of monoamines (by reserpine) in experimental manipulation and the cessation of activity of monoamine neurons in the natural cycle are necessary for the elicitation of paradoxical sleep by cholinergic agonists (physostigmine) or activity of cholinergic neurons. In single-unit recording studies conducted by Sakai (1985) and his colleagues; (El Mansari ef al., 1989; see Sakai, this volume), it appears that putative cholinergic neurons become active during paradoxical sleep (PS-on) immediately upon the cessation of activity of the putative noradrenergic neurons, that (like serotonergic raphe neurons) are inactive during paradoxical sleep (PSoff). These profiles suggest instead of a reciprocal interaction, as formally proposed by McCarley and Hobson (1975) for reticular (gigantocellular tegmental field) and LC neurons, a mutually inhibitory interaction between the cholinergic and noradrenergic neurons, as proposed more recently by Sakai (1985). Such an interaction is unlikely to be direct, particularly in view of the paucity of cholinergic fibers in the region of the LC cell bodies (see B.E. Jones, this volume) and the excitation that acetylcholine has been found
to produce on LC neurons (Egan and North, 1986; Guyenet and Aghajanian, 1979). But such an interaction could be possible via an intermediary of inhibitory interneurons, such as the GABA cells that have been identified within the periventricular gray in the region of the cholinergic and noradrenergic neurons and their processes (Jones, 1990c; see B.E. Jones, this volume). Evidence for an active inhibition of LC neurons during paradoxical sleep has indeed been brought forward (Aston-Jones and Bloom, 1981b) and would likely involve GABAergic transmission in view of the moderately dense GABAergic innervation of the LC cell bodies (see B.E. Jones, this volume) and the potent inhibitory effect of GABA and its agonists on these cells (Guyunet and Aghajanian, 1979; Osmanovic and Shefner, 1988). Acknowledgements I would like to express my gratitude to the individuals who contributed to the research reviewed in this article, namely Lynda Mainville, Lee Friedman, Harry Webster and Colin Holmes. The research was supported by a grant from the Medical Research Council of Canada. References Aston-Jones, G. and Bloom, F.E. (1981a) Activity of norepinephrine-containing locus coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle. J. Neurosci., 1: 876-886. Aston-Jones, G. and Bloom, F.E. (1981b) Norepinephrinecontaining locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Baghdoyan, H.A., Rodrigo-Angulo, M.L., McCarley, R.W. and Hobson, J.A. (1984) Site-specific enhancement and suppression of desynchronized sleep signs following cholinergic stimulation of three brainstem regions. Brain Res., 306: 39-52. Cannon, W.B., Newton, H.F., Bright, E.M., Menkin, V and Moore, R.M. (1929) Some aspects of the physiology of animals surviving complete exclusion of sympathetic nerve impulses. Am. 1.Physiol., 89: 84-107. Carlsson, A,, Lindqvist, M. and Magnusson, T. (1957) 3,4-dihydroxyphenylalanine and 5-hydroxytryptophan as reserpine antagonists. Nature (London), 180: 1200.
542 Dahlstrom, A. and Fuxe, K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Scand. (Suppl.), 232: 1-55. Domino, E.F., Yamamoto, K. and Dren, A.T. (1968) Role of cholinergic mechanisms in states of wakefulness and sleep. Prog. Brain Res., 28: 113-133. Egan, T.M. and North, R.A. (1986) Actions of acetylcholine and nicotine on rat locus coeruleus neurons in uitro. Neuroscience, 19: 565-571. El Mansari, M., Sakai, M. and Jouvet, M. (1989) Unitary characteristics of presumptive cholinergic tegmental neurons during the sleep-waking cycle in freely moving cats. Exp. Brain. Res., 76: 519-529. Foote, S.L., Aston-Jones, G. and Bloom, F.E. (1980) Impulse activity of locus coeruleus neurons in awake rat and monkey is a function of sensory stimulation and arousal. Proc. Nail. Acad. Sci. USA, 77: 3033-3037. George, R., Haslett, W. and Jenden, D. (1964) A cholinergic mechanism in the brainstem reticular formation: Induction of paradoxical sleep. Int. J. Neuropharmacol., 3: 541-552. Guyenet, P.G. and Aghajanian, G.K. (1979) ACh, substance P and met-enkephalin in the locus coeruleus: Pharmacological evidence for independent sites of action. Eur. J. Pharmacol., 53: 319-328. Hallanger, A.E. and Wainer, B.H. (1988) Ascending projections from the pedunculopontine tegmental nucleus and the adjacent mesopontine tegmentum in the rat. J. Comp. Neurol., 274: 483-515. Hallanger, A.E., Levey, A.I., Lee, H.J., Rye, D.B. and Wainer, B.H. (1987) The origins of cholinergic and other subcortical afferents to the thalamus in the rats. J. Comp. Neurol., 262: 105-124. Hazra, J. (1970) Effect of hemocholinium-3 on slow wave and paradoxical sleep of cat. Eur. J. Pharmacol., 11: 395-397. Henley, K. and Morrison, A.R. (1974) A re-evaluation of the effects of lesions of the pontine tegmentum and locus coeruleus on phenomena of paradoxical sleep in the cat. Acta Neurobiol. Exp. (Warszawa)., 34: 215-232. Hobson, J.A., McCarley, R.W. and Wyzinski, P.W. (1975) Sleep cycle oscillation: Reciprocal discharge by two brainstem neuronal groups. Science, 189: 55-58. Jacobs, B.L. and Jones, B.E. (1978) The role of central monoamine ,and acetylcholine systems in sleep-wakefulness &;es: Ihediation 05 ulation. In L.L. Butcher (Ed.), Cholinergic-Monami,nergicInteractions in the Brain, Academic Press, New York, pp. 271-290. Jones, B.E. (1969) Catecholamine-Containing Neurons in the Brainstem of the Cat and Their Role in Waking, Imprimerie des Beaux-Arts, Lyon, pp. 1-87. Jones, B.E. (1990a) Immunohistochemical study of choline acetyl transferase-immunoreactive processes and cells innervating the pontomedullary reticular formation. J. Comp. Neurol., 295: 485-514. Jones, B.E. (1990b) Influence of the brainstem reticular formation, including intrinsic monoaminergic and cholinergic neurons, upon forebrain mechanisms of sleep and waking.
In M. Mancia and G. Marini (Eds.), The Diencephalon and Sleep, Raven Press, New York, pp. 31-48. Jones, B.E. (1990) Commentary: Paradoxical sleep and its chemical/structural substrates in the brain. Neuroscience, (In press). Jones, B.E. and Beaudet, A. (1987) Distribution of acetylcholine and catecholamine neurons in the cat brain stem studied by choline acetyltransferase and tyrosine hydroxylase immunohistochemistry. J. Camp. Neurol., 261: 15-32, Jones, B.E. and Cuello, A.C. (1989) Afferents to the basal forebrain cholinergic cell area from ‘pontomesencephalic -catecholamine, serotonin, and acetylcholine-neurons. Neuroscience, 31: 37-61. Jones, B.E. and Moore, R.Y. (1974) Catecholamine-containing neurons of the nucleus locus coeruleus in the cat. J. Comp. Neurol., 157: 43-52. Jones, B.E. and Webster, H.H. (1988) Neurotoxic lesion’s of the dorsolateral pontomesencephalic tegmentum-cholinergic cell area in the cat. I. Effects upon the cholinergic ’ innervation of the brain. Brain Res., 451: 13-32. Jones, B.E. and Yang, T.-Z. (1985) The efferent projections from the reticular formation and the locus coeruleus studied by anterograde and retrograde axonal transport in the rat. J. Comp. Neurol., 242: 56-92. Jones, B.E., Bobillier, P., Pin, C. and Jouvet, M. (1973) The effect of lesions of catecholamine-containing neurons upon monoamine content of the brain and E E G and behavioral waking in the cat. Brain Res., 58: 157-177. Jones, B.E., Harper, S.T., and Halaris, A.E. (1977) Effects of locus coeruleus lesions upon cerebral monoamine content, sleep-wakefulness states and the response to amphetamine in the cat. Brain Res., 124: 473-496. Jouvet, M. (1962) Recherches sur les structures nerveuses et les mecanismes responsables des differentes phases du sommeil physiologique. Arch. Ital. Biol., 100: 125-206. Jouvet, M. (1972) The role of monoamines and acetylcholinecontaining neurons in the regulation of the sleep-waking cycle. Ergeb. Physiol., 64: 166-307. Jouvet, M. and Delorme, F. (1965) Locus coeruleus et sommeil paradoxal. C.R. Soc. Biol. (Paris), 159: 895-899. Karczmar, A.G., Longo, V.G. and Scotti d e Carolis, A. (1970) A pharmacological model of paradoxical sleep: The role of cholinergic and monoamine systems. Physiol. Behac., 5: 175-182. King, C.D. and Jewett, R.E. (1971) The effects of a-methyltyrosine on sleep and brain norepinephrine in cats. J. Pharamcol. Exp. Ther., 177: 188-195. Kohler, C. and Schwarcz, R. (1983) Comparison of ibotenate and kainate neurotoxicity in rat brain: A histological study. Neuroscience, 8: 819-835. Lidbrink, P. (1974) The effect of lesions of ascending noradrenaline dathways on sleep and waking in the rat. Brain Res., 74: 19-40. Lindsley, D.B., Schreiner, L.H., Knowles, W.B. and Magoun, H.W. (1950) Behavioral and E E G changes following chronic brain stem lesions in the cat. Electroencephalogr. Clin. Neurophysiol., 2: 483-498. Longo, V.G. (1966) Behavioral and electroencephalographic
543 effects of atropine and related compounds. Phurmucol. Re(,., 18: 965-996. Maeda, T., Pin, C., Salvert, D., Ligier, M. and Jouvet, M. (1973) Les neurones contenant des catecholamines du tegmentum pontique et leurs voies de projection chez le chat. Bruin Res., 57: 119-152. McCarley, R.W. and Hobson, J.A. (1975) Neuronal excitability modulation over the sleep cycle: A structural and mathematical model. Science, 189: 58-60. McCormick, D.A. (1989) Cholinergic and noradrenergic modulation of thalamocortical processing. TINS, 12: 215-221. McGinty. D.J., Harper, R.M. and Fairbanks, M.K. (1974) Neuronal unit activity and the control of sleep states. In E.D. Weitzman (Ed.), Adrwices in Sleep Research, Spectrum Publications, New York, pp. 173-216. Moruzzi, G. and Magoun, H.W. (1949) Brain stem reticular formation and activation of the EEG. Electrvencephulogr. Clin. Neurophysiol., 1: 455-473. Osmanovic, S.S. and Shefner, S.A. (1988) Baclofen increases the potassium conductance of rat locus coeruleus neurons recorded in brain slices. Bruin Res., 438: 124-136. ParC. E., Smith, Y., Parent, A. and Steriade, M. (1988) Projections of brainstem core cholinergic and non-cholinergic neurons of cat to intralaminar and reticular thalamic nuclei. Neuroscience, 25: 69-86. Pin, C.. Jones, B. and Jouvet, M. (1968) Topographie des
neurones monoaminergiques du tronc cerebrale du chat: Etude par histofluouescence. C.R. Soc. Biol. (Puris), 162: 2136-2141. Rasmussen, K., Morilak, D.A. and Jacobs, B.L. (1986) Single unit activity of locus coeruleus neurons in the freely moving cat: I. During naturalistic behaviors and in response to simple and complex stimuli. Bruin Res., 371: 324-334. Sakai, K. (1985) Neurons responsible for paradoxical sleep. In A. Wauquier, J.M. Gaillar, J.M. Monti and M. Radulovacki (Eds.), Sleep: Neurotrunsrnitters and Neurvmodulutors, Raven Press, New York, pp. 29-42. Shute, C.C.D. and Lewis, P.R. (1967) The ascending cholinergic reticular system: Neocortical, olfactory and subcortical projection. Bruin Res., 90: 497-520. Smith, Y., ParC, D., Deschenes, M., Parent, A. and Steriade, M. (1988) Cholinergic and non-cholinergic projections from the upper brainstem core to the visual thalamus in the cat. Exp. Bruin Res., 70: 166-180. Steriade, M., Datta, S., ParC, D., Oakson, G. and Curro Dossi, R. (1990) Neuronal activities in brain-stem cholinergic nuclei related to tonic activation processes in thalamocortical systems. J. Neurosci. (In press). Webster, H.H. and Jones, B.E. (1988) Neurotoxic lesions of the dorsolateral pontomesencephalic tegmentum-cholinergic cell area in the cat. 11. Effects upon sleep-waking states. Bruin Res., 458: 285-302.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Prosress m Brain Research, Vol. 88 0 1991 Elsevier Science Publishers B.V.
545
CHAPTER 38
Effects of local pontine injection of noradrenergic agents on desynchronized sleep of the cat G. Tononi, M. Pompeiano and C. Cirelli Department of Physiology and Biochemistry, Uniuersity of Pisa, Via S. Zeno, Pisa, Italy
Brain noradrenergic (NA) systems have often been implicated in the regulation of desynchronized sleep (DS). The present experiments investigate the effects on DS of the microinjection, into the cat dorsal pontine tegmentum (DPT), of the a,-agonist clonidine (CLON), the P-agonist isoproterenol and the P-antagonist propranolol. The DPT comprises most NA neurons belonging to the locus coeruleus (LC) complex, as well as other cell groups thought to be crucially involved in DS generation. Cats were implanted with standard electrodes (electroencephalogram, electrooculogram and electromyogram, PGO waves, hippocampal activity) and with guide tubes aimed at the DPT. Unilateral or bilateral injections (0.25 pl) were performed by way of thin cannulae inserted through the guide tubes. Polygraphic activity was then recorded in daily sessions lasting 4 h and scored according to standard criteria. Bilateral injections of CLON into the DPT greatly reduced
DS, while unilateral injections were much less effective. Since CLON is known to powerfully inhibit NA LC neurons, its effect was thus opposite to that expected on the basis of the reciprocal interaction model of DS generation, which postulates that NA neurons in the LC inhibit DS-executive cells located in the pontine reticular formation. Bilateral injections of the P-agonist isoproterenol also reduced DS, while the P-antagonist propranolol consistently enhanced it, the latter largely due to an increased number of DS episodes. These effects were dose-dependent and strictly site-specific, since injections in immediately neighboring structures were ineffective. These results: (i) confirm that cell groups located in the DPT play a key role in the generation of DS, and (ii) indicate that they undergo a strong NA modulation, being inhibited by a*-and P-receptor stimulation and disinhibited by @-receptor blockade.
Key words: desynchronized sleep, locus coeruleus, pontine reticular formation, adrenoceptors, clonidine, isoproterenol, propranolol
Introduction The hypothesis that the noradrenergic (NA) system is involved in the regulation of desynchronized sleep (DS) was proposed several years ago, but the relationship between this system and DS is still unclear. The fact that NA cells cease firing selectively during DS (Foote et al., 1983) has prompted many investigators to claim that the
correlation between NA activity and DS is negative. In particular, the reciprocal interaction model of DS generation (Hobson et al., 1986) postulates that the arrest of firing of NA locus coeruleus (LC) neurons (DS-off cells) plays a causal role in DS generation, releasing from inhibition cholinergic-cholinoceptive neurons which would be directly executive for DS signs (DS-on cells). The cholinergic side of the model is broadly
546
supported. Systemic administration of cholinergic agents increases DS, and the direct injection of cholinergic agonists into the pontine tegmentum can produce sustained episodes of DS and related phenomena (Hobson et al., 1986), presumably through muscarinic M,-receptors (Gerber et al., 1989). On the other hand, the postulated NA-mediated inhibition of DS-executive cells is still controversial (Gaillard, 1985; Stenberg and Hilakivi, 1985). For instance, NA neurons in the LC are strongly inhibited by a,-adrenergic agonists like clonidine (CLON) (Foote et al., 1983). Thus, one should expect CLON to increase DS. Yet, CLON powerfully decreases DS when injected systemically in several species (Gaillard, 1985). Nevertheless, the results of systemic injections are extremely difficult to interpret, due to several peripheral and/or secondary drug ac-
tions. The experiments summarized in the present report (Tononi and Pompeiano, 1988; Tononi et al., 1988a,b, 1989, 1991) were thus aimed at clarifying the role of a,- and P-adrenergic receptors in the regulation of DS by using the direct microinjection technique (the effects of cy agonists and antagonists are currently being investigated). Alpha,- and P-agonists and antagonists were injected into the dorsal pontine tegmentum (DPT) of freely moving, unanesthetized cats (see Fig. 1). The DPT, which anatomically contains the whole LC complex (LCd, LCa, peri-LCa and locus subcoeruleus), seems to be especially important for DS generation, with respect to both supposedly permissive and executive components. In fact: (i) it contains most NA DS-off cells, closely intermingled with cholinergic neurons (Jones et al., 1987). Further-
,-
Fig. 1. Anatomical localization of the injection sites in the dorsal pontine tegmentum. Sagittal section of the brain stem at the stereotaxic laterality of 2.8 mm, illustrating the location of the cannula in one representative experiment. The tip of the cannula is surrounded by a shaded area which corresponds to the halo of pontamine diffusion (5% in 0.25 pl sterile saline). Abbreviations: BC, brachium conjunctivum; BCM, marginal nucleus of the brachium conjunctivum; CBM, medial cerebellar nucleus; CUR, cuneate nucleus, rostra1 division; FTG, gigantocellular tegmental field; FTL, lateral tegmental field; F'TP, paralemniscal tegmental field; IVN, inferior vestibular nucleus; LCd locus coeruleus, pars dorsalis; L C a , locus coeruleus a ; LVN, lateral vestibular nucleus; MET, mesencephalic trigeminal tract; MVN, medial vestibular nucleus; PGL, pontine grey, lateral division; PGR, pontine grey, rostroventral division; peri-LCa, peri-locus coeruleus a ; RN, red nucleus; S, solitary tract; SL, lateral nucleus of the solitary tract; SM, medial nucleus of the solitary tract; SN, substantia nigra; SOM, medial nucleus of the superior olive; SVN, superior vestibular nucleus; T, nucleus of the trapezoid body; TB, trapezoid body; TRP, tegmental reticular nucleus, pericentral division; 3N, oculomotor nerve; 5ME, mesencephalic trigeminal nucleus; 7N, facial nerve. (From Tononi el af., 1989.)
547
more, some highly specific DS-on cells have been recorded in this area (Sakai, 1985a,b, 19881, (ii) electrolytic, thermolytic or chemical destruction of this region is particularly effective in disrupting DS (Jones, 1985; Sakai, 1988; Webster and Jones, 19881, and (iii) the cholinergic agonist carbachol appears to induce DS most reliably when injected into the DPT (Quattrochi et al., 1989; VanniMercier et al., 1989). Methods
Chronic electrodes for recording the electroencephalogram, electrooculogram and electromyogram from dorsal neck muscles were implanted in 15 adult cats under pentobarbital anesthesia. Bipolar electrodes were placed in the lateral geniculate nucleus and in the dorsal hippocampus. Two, parallel, 24-gauge stainless steel guide tubes were implanted aiming at the DPT. After recovery from surgery, the cats were accustomed to an acoustically insulated and dimly illuminated recording box until sleep-wakefulness variables became stable and within published norms. All recording sessions lasted for 4 h and took place between 2 and 9 p.m. After two or more baseline control recording sessions, injections were performed every 4 days by way of a 31-gauge stainless steel cannula inserted through the guide tubes and connected to a manually driven 1 p1 syringe. All injections were 0.25 p l in volume delivered over 60 sec. The cannula was left in place for an additional 60 sec to avoid or limit back-diffusion. In case of bilateral administration of drugs, the interval between right and left injections was less than 3 min. On control recording sessions, saline injections were performed or simply simulated to avoid damage to the target region. After the injections, following a rapid assessment of the ongoing behavior, posture and reflexes, the cats were immediately put in the recording box and observed by way of a TV camera. The polygraphic records were scored in 30 sec epochs for waking (W), synchronized sleep (SS) and DS according to standard polygraphic criteria (Ursin
and Sterman, 1981). The following variables were calculated from the polygraphic recordings: percent of the recording time spent in DS, SS and W during the four recording hours, both cumulatively and separately for each hour, latency to DS onset, mean number and duration of DS episodes, duration of the longest DS episode, and ratio between DS and SS. Statistical tests of differences between the results of specific injections and the controls were accomplished only when the histological examination established that the bilateral injections were strictly symmetrical. A two-tailed Student’s t-test was used to compare the differences between group means in the case of sleep stage percentages. The non-parametric Mann-Whitney U-test was used for the rest of the data. After a series of several injections, a volume of 0.25 pl of saline stained with 5% pontamine was injected bilaterally at the stereotaxic coordinates where the injected substances had elicited the strongest effects on sleep. The animals were then sacrificed and the brain fixed in 10% formaline. The exact localization of each penetration, as well as the extent of pontamine diffusion, were then identified on serial frozen sections of the brainstem mounted on glass slides and stained with neutral red as shown in Figure 1. The following drugs were used: CLON-HC1 (Sigma), (+)isoproterenol-HC1 (Sigma), phentolamine-HC1 (Ciba-Pharm), ( f)-propranolol-HC1 (ICI), salbutamol-sulphate (Glaxo) and yohimbine-HCI (Sigma). They were dissolved in 0.25 p1 of buffered sterile saline (pH 7.4). Results
Controls There were no differences between baseline recordings and recordings which followed the actual or simulated injection of saline, so they were collectively treated as controls.
Clonidine As shown in Table 1A (Fig. 21, bilateral injections of the a,-agonist CLON into the DPT (4 p g
54H TABLE IA Effects on sleep-waking parameters of clonidine infused into the dorsnl pontine tegmentum (4 pg in 0.25 pl of saline bilaterally and symmetrically)
a-Agonist
Controls
W
ss DS
( n = 10)
Clonidine ( n = 10)
45.5 13.2 36.2 i2.9 18.21 1.1
64.8 k 0.6 * 33.1 +S.1 2.0k0.8 * *
* P < 0.01. * * P < 0.001.
in 0.25 pl of saline, stereotaxic coordinates of P 1.0-3.0, LR 2.5-2.8, H -2.0 to -3.0, n = 10 experiments in 7 animals) dramatically decreased DS, which dropped from 18.2-t 1.1% (mean k S.E.) of the total recording time in the controls to 2.0+ 0.8%. On the other hand, SS was almost unchanged, passing from 36.2 -t 2.9% to 33.1 S.1%, so that the ratio between DS and SS dropped from 0.53 t 0.05 in the controls to 0.05 iO.02 after CLON. DS latency was greatly increased, while the number and mean duration of DS episodes decreased. In three experiments DS was completely suppressed for more than 4 h. The effects of CLON were dose-dependent (data not shown). Bilateral injections of CLON 1-2 mm out of the critical spot in a rostro-dorsal position ( n = 3 injections) resulted in normal amounts of DS. Bilateral injections into the gi-
+
%
gantocellular tegmental field (FTG) (P 3.0, LR 2.5, H -6.0, n = 2 injections) were equally ineffective. In two cats, where histology showed that injection sites were not symmetrical, intermediate effects were observed ( n = 5 injections). A pretreatment (15 min before) with the non selective a-adrenergic antagonist phentolamine (8 p g in 0.25 pl of saline bilaterally) prevented the effects of a subsequent injection of CLON (4 p g in 0.25 pl). However, in the last case the distribution of DS was not homogeneous over the four recording hours, due to a complete suppression during the first 2 h and a rebound during the last two. In all the experiments reported here no significant behavioral modifications were observed, either immediately after the injections or during the recording sessions.
Isoproterenol Following bilateral injections into the DPT of the /3-adrenergic agonist isoproterenol (IS01 (4 p g in 0.25 p1 of saline, n = 8 experiments) DS decreased on the average from 21.2 f 0.7% (mean -t S.E.) of the total recording time in the controls to 3.0 & 0.9% (Table 1B; Fig. 3). DS latency increased while the mean number and duration of episodes decreased; in three experiments DS was suppressed throughout the 4 h of the recording session; SS was reduced much less TABLE 1B
30 2 47 7
22 1
w
619
S
37 7
Effects on sleep-waking parameters of isoproterenol and propranol infused into the dorsal pontine tegmentum (4 pg in 0.25 p1 of saline bilaterally and symmetrically) &Agents
W
ss DS 0
1
2 HOURS
3
'
Controls (n=16)
Isoproterenol (N=8)
Propranolol ( n = 6)
37.2k2.0 4 1 . 6 i 1.8 21.2k0.7
71.5i9.4 24.3i8.4 3.010.9 * *
24.6k3.7 * 40.9 i2.7 34.5k2.8 * *
4
Fig. 2. Hypnograms of two individual recording sessions from one cat. A. After the injection of saline (0.25 pl). B. After the injection of clonidine (4 pg in 0.25 pI saline). The injections were located bilaterally and symmetrically in the same spot of the dorsal pontine tegmentum (P 2.0, I,R 2.5, H - 2.0). (From Tononi et al., 1991.)
* P < 0.01. * * P < 0.001. Controls include baseline recordings, saline infusions and simulated saline infusions. The values are means+ S.E. and refer to recording sessions of 4 h. Student's t-test was used for the percentages of sleep stages. In parentheses ( 1 2 ) is the number of experiments.
549
W
JY 4
Sb
(DS = 2.5% of the total recording time, mean of two trials).
37 7
DS
Propranobl After the bilateral injection into the DPT of the P-adrenergic antagonist PRP (4 p g in 0.25 pl W 79 4 ss 16 5 of saline, n = 6 experiments), DS was significantly enhanced (Table 1B), on the average from 21.2 f 41 0.7% (mean S.E.) of the total recording time in W the controls to 34.5 Ifr 2.8% after PRP. This incre31 5 SS ment was largely due to a higher number of DS LIR 7 DS 39 8 episodes (15.3 2.5 against 9.7 0.7 in the controls) and to an increase of maximum duration ' ~ 0 1 2 3 4 (25 min after PRP against 13.5 in the controls), HOURS while the mean duration of DS episodes was Fig. 3. Hypnograms of three individual recording sessions from one cat. A. After the injection of saline (0.25 PI). B. comparable to the control value. There was a After the injection of isoproterenol (4 pg in 0.25 pl saline). C. slight reduction of the latency to the first DS After the injection of propranolol (4 pg in 0.25 p1 saline). The episode (from 37.5 & 3.7 min to 25.2 k 3.2 min, injections were located bilaterally and symmetrically in the P = 0.06, U-test), which approached, but did not same spot of the dorsal pontine tegrnentum ( P 2.0, LR 2.5, H -2.0). (From Tononi ef al., l Y 8 Y . ) reach, statistical significance. Furthermore, DS enhancement was equally distributed during a11 than DS, from 41.6 1.8 to 24.3 i 8.4, so that the four recording hours. The amount of SS reratio between DS and SS decreased on the avermained unchanged, so that the ratio between DS age from 0.53 0.03 in the controls to 0.18 0.09 and SS increased on the average from 0.53 0.03 after ISO. In five out of eight experiments, howto 0.87 -t 0.11, while W decreased slightly to acever, SS did not show any significant change, count for the increased sleep time. A normal while DS was sharply reduced. This fact seems to indicate that the decrease in DS was not simply due to an aspecific increase in W. The effects 25 were dose-dependent (Fig. 4). When I S 0 was injected bilaterally 1-2 mm away from the critical 0 a, region in a rostro-dorsal or caudo-ventral posia, C tion, a normal percentage of DS was observed u, ( n = 9 injections). In addition, no effect was dea m tected after bilateral injections into the FTG ( n n = 4 injections). Following a pre-treatment (15 min before) with the a-adrenergic antagonist propranolol (PRP) (8 p g in 0.25 pl of saline bilaterally) the effect of a subsequent injection of I S 0 (4 Dose of IS0 lpg/025pl bilaterally) p g in 0.25 pl of saline bilaterally) was abolished, Fig. 4. Relationship between DS percentage and dose of for DS reached 18.1% of the recording time isoproterenol. Three doses of isoproterenol (1, 2 and 4 pg in (mean of two trials). On the other hand, after a 0.25 pI saline bilaterally), injected in the dorsal pontine tegmentum of one cat always in the same spot, produced a pretreatment with the a-adrenergic blocker graded reduction of DS (14.6, 10.0, 1.7% of the total recordphentolamine (4 p g in 0.25 pl of saline bilaterally ing time, respectively; control = 19.4%). Correlation coeffiinto the same spot), I S 0 strongly reduced DS cient = -0.99. (From Tononi et al., 1989.) 22 9
1
I
*
*
-Lpppl
L~--
*
*
*
7
4-
550
percentage of DS was observed when PRP was injected bilaterally 1-2 mm away from the critical region ( n = 7 injections), or when one of the two injection sites was not symmetrically placed and missed the critical region ( n = 7 injections). In addition, no effect was detected after bilateral PRP injection into the FTG ( n = 3 injections). Histology
Histological examination of the sections revealed that the sites where bilateral injections of the q a g o n i s t CLON, the P-agonist I S 0 and the P-antagonist PRP had strongly affected the generation of DS were all located in a limited region within the DPT, including the peri-LCa region as described by Sakai (1980) as well as the pontine reticular formation immediately ventral to it and medial to the motor nucleus of the fifth nerve (field tegmentalis lateralis, FTL), corresponding to the stereotaxic coordinates P 1.0-3.0, LR 2.52.8, H -2.0 to -3.0 (Fig. 1). Injection sites displaced by only 1-2 rnm either more rostrally and dorsally or more caudally and ventrally from the critical region were ineffective; similarly, no effect was observed after injections into the FTG. Discussion a,-Receptors In the cat, the region of the DPT extending from P 1.0 to P 3.0 contains a high proportion of the tyrosine-hydroxylase positive NA neurons belonging to the LC complex (Jones and Bebudet, 1987). These neurons cease discharging during DS and, according to the reciprocal interaction model, this fact plays a permissive role in the execution of DS (Hobson et al., 1986). The implication is that the inactivation of these neurons should enhance DS. There have been previous efforts to study the effect on DS of the inactivation of LC NA cells by electrolytic lesions and cooling, but the results are controversial. Bilateral electrolytic lesion of the LC (Jones et al., 19771, which caused an 85% depletion of cerebral
NA, scarcely affected the sleep-waking pattern of the cat, thus challenging the involvement of NA in sleep-wakefulness mechanisms. However, unilateral electrolytic lesions of the LC complex in four cats were reported to slightly increase DS from 9 to 12%, starting from the second week after operation (Caballero and De Andres, 1986). Furthermore, cooling of the dorsal LC was rapidly followed, though only in 40-50% of the cases, by SS and DS (Cespuglio et al., 1982). In this study, we tried to inactivate LC NA cells through local injections of the specific q a g o n i s t CLON. In fact, there is compelling evidence that LC NA cells are powerfully inhibited by CLON due to potentiation of NA mechanisms involving recurrent and/or lateral inhibition (Foote et al., 1983). This has also been shown in the behaving cat, where the infusion of CLON (1 pg/pl) completely and persistently suppressed their discharge, while the effects on other neighboring neurons were weak and unpredictable (Abercrombie and Jacobs, 1987). The results reported here indicate that DS, rather than being enhanced as predicted by the reciprocal interaction model, was instead strongly reduced. This effect seems to be relatively state-specific, since slow-wave sleep was virtually unaffected, and dose-dependent. These findings are consistent with the results of systemic injection experiments, which have shown that CLON reliably decreases DS in several animal species, including cats (Putkonen et al., 1977; Putkonen, 1978; Leppavuori and Putkonen, 1980). Though microinjection techniques are surely preferable to electrolytic lesions, and perhaps also to cooling, thanks to their neurochemical specificity, the present results should still be interpreted with great care. For instance, due to the small injection volumes, it is very likely that not all NA cells were silenced, particularly some located more dorsally in the LC complex as well as those located in its most rostra1 and caudal parts. Nevertheless, it is intriguing that over several injection trials, in several animals, and with several different CLON concentrations, we have never ob-
55 1
served even a modest increase of DS. Thus, as proposed by Sakai (1988), the inhibition of DS-off cells might be less critical to DS generation than the excitation of DS-on cells. That is to say, DS-on cells would be modulated in their excitability in relative independence of DS-off cells, while the converse would probably not be true. However, the unexpected result was that DS not only failed to increase, but was instead strongly decreased. A possible explanation is that CLON exerted a DS suppressive effect by acting additionally on cell groups positively implicated in the generation of DS, and located within the DPT intermingled with NA cells. Though the best known and most powerful effect of CLON is to inhibit LC NA cells, activation of a,-receptors may also affect non-NA neurons. For instance, in the slice preparation some neurons in the medial pontine reticular formation appear to be inhibited by CLON (Greene et al., 1989). Thus, in defense of the reciprocal interaction model, it can still be argued that the effect of the inhibition of NA cells might have been masked by additional effects on DS-executive cells. P-Receptors The precise effects of P-receptor agonists and antagonists on the various cell types involved in DS generation are still largely unknown. Yet, the presence of P-adrenoceptors in the DPT has been clearly demonstrated by autoradiographic studies in the rat (Palacios and Kuhar, 1980; Rainbow et al., 1984) and binding studies in the cat (Pompeiano et al., 1989). The results obtained with the microinjection of I S 0 and PRP indicate that, on the whole, cell groups located in the DPT and implicated in the generation of DS are strongly affected by P-adrenoceptor stimulation or blockade, which are followed by the suppression or the enhancement of DS, respectively. These effects are dose-dependent and site-specific. In this case, there is no agreement with systemic injection experiments, according to which P-adrenergic antagonists tend to decrease DS (Hilakivi, 1983; Lanfumey et al., 1985). Microin-
jection results also confirm that the DPT is more critical than the FTG in the generation of DS (Sastre et al., 1981; Drucker-Colin and BernalPedraza, 1983). It should be mentioned, however, that a DS-enhancing effect of unilateral PRP injections in the FTG area has been reported (Vivaldi et al., 1980; Tononi et al., 1989). Finally, these findings extend the observations of Masserano and King (1982), who injected epinephrine and phentolamine in the LC region and obtained a reduction and an enhancement of DS, respectively. In view of the present results, it is likely that the first effect was due to a stimulation of both a,- and P-receptors. Whether. a,-receptors also play a role, and whether the second effect was due to blockade of a,-, a2- or both kind of receptors will be established by microinjections experiments with a,-agonists and antagonists which are now being pursued. Acknowledgements
This work was supported by National Institute of Neurological and Communicative Disorders and Stroke Research Grant NS 07685-22 and by grants of the Minister0 dell’Universitii, and the Agenzia Spaziale Italiana, Roma, Italy. References Abercrombie, E.D. and Jacobs, B.L. (1987) Microinjected clonidine inhibits noradrenergic neurons of the locus coeruleus in freely moving cats. Neurosci. Lett., 7: 203-208. Caballero, A. and De Andres, I. (1986) Unilateral lesions in locus coeruleus area enhance paradoxical sleep. Electroencephalogr. Clin.Neurophysiol., 64: 339-346. Cespuglio, R., Gomez, M.E., Faradji, H. and Jouvet, M. (1982) Alterations in the sleep-waking cycle induced by cooling of the locus coeruleus area. Electroencephalogr. Clin. Neurophysiol., 54: 570-578. Drucker-Colin, R. and Bernal-Pedraza, J.G. (1983) Kainic acid lesion of gigantocellular tegmental field (FTG) neurons does not abolish REM sleep. Brain Res., 272: 387-391. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Reu., 63: 844-914. Gaillard, J.-M. (1985) Involvement of noradrenaline in wakefulness and paradoxical sleep. In: A. Wauquier, J.-M. Gaillard, J.M. Monti and M. Radulovacki (Eds.), Sleep,
552
Neurotransmitters and Neuromodulators, Raven Press, New York, pp. 57-67. Gerber, U., Greene, R.W. and McCarley, R.W. (1989) Muscarinic receptors modulate neuronal membrane properties in medial pontine reticular formation. Sleep Res., 18: 10. Greene, R.W., Gerber, U., Haas, H.L. and McCarley, R.W. (1989) Noradrenergic actions on neurons of the medial pontine reticular formation in ritro. Sleep Res., 18: I I . Hilakivi, I. (1983) The role of a- and P-adrenoceptors in the regulation of the stages of the sleep-waking cycle in the cat. Brain Res., 277: 109-118. Hobson, J.A., Lydic, R. and Baghdoyan, H.A. (1986) Evolving concepts of sleep cycle generation: From brain centers to neuronal populations. Behar.. Bruin Sci., 9: 371-488. Jones, B.E. (1985) Neuroanatomical and neurochemical substrates of mechanisms underlying paradoxical sleep. In: D.J. McGinty, R.R. Drucker-Colin, A. Morrison and P.L. Parmeggiani (Eds.), Brain Mechanisms of Sleep, Raven Press, New York, pp. 139-156. Jones, B.E., Beaudet, A. (1987) Distribution of acetylcholine and catecholamine neurons in the cat brainstem: A choline-acetyltransferase and tyrosine-hydroxylase immunohistochemical study. J. Comp. Neurol., 261: 15-32. Jones, B.E., Harper, S.T. and Halaris, A.E. (1977) Effects of locus coeruleus lesions upon cerebral monoamine content, sleep-wakefulness states and the response to amphetamine in the cat. Brain Res., 124: 473-496. Lanfumey, L., Dugovic, C. and Adrien, J. (1985) p , and p2 adrenergic receptors: Their role in the regulation of paradoxical sleep in the rat. Electroencephalogr. Clin. Neurophysiol., 60: 558-567. Leppavuori, A. and Putkonen, P.T.S. (1980) a-adrenoceptive influences on the control of the sleep-waking cycle in the cat. Brain Res., 193: 95-115. Masserano, J.M. and King, C. (1982) Effects on sleep of phentolarnine and epinephrine infused into the locus coeruleus of cat. Eur. J. Pharmacol., 84: 199-244. Palacios, J.M. and Kuhar, M.J. (1980) P-adrenergic receptor localization by light microscopic autoradiography. Science, ,208: 1378-1380. Pompeiano, M., Ronca-Testoni, S. and Pompeiano, 0. (1989) Distribution of P-adrenergic receptors in different brainstem regions controlling posture in the cat. Arch. Iral. Biol., 127: 13-23. Putkonen, P.T.S. (1978) a and P adrenergic mechanisms in the control of sleep stages. In: Proc. Northern Eur. Symp. Sleep Res., Basle, September 1978. In: R.G. Priest, A. Pletscher and J. Ward (Eds.), Sleep Research, MTP Press, Lancaster, pp. 19-34. Putkonen, P.T.S., Leppavuori, A. and Stenberg, D. (1977) Paradoxical sleep inhibition by central a-adrenoceptor stimulant clonidine antagonized by a-receptor blocker yohirnbine. Life Sci., 21: 1059-1066. Quattrochi. J.J., Mamelak, A.N., Madison, R.D., Macklis, J.D. and Hobson, J.A. (1989) Mapping neuronal inputs to R E M sleep induction sites with carbachol-fluorescent microspheres. Science, 245: 984-986.
Rainbow, T.C., Parsons, B. and Wolfe, B.B. (1984) Quantitative autoradiography of a I -and a,-adrenergic receptors in rat brain. Proc. Natl. Acad. Sci. USA, 81: 1585-1589. Sakai, K. (1980) Some anatomical and physiological properties of ponto-mesencephalic tegmental neurons with special reference to the P G O waves and postural atonia during paradoxical sleep. In: J.A. Hobson, M.A.B. Brazier (Eds.), The Reticular Formation Rec,isited, IBRO Monograph Series, Vol. 6 , Raven Press, New York, pp. 427-447. Sakai, K. (1985a) Neurons responsible for paradoxical sleep. In: A. Wauquier, J.M. Gaillard, J.M. Monti and M. Radulovacki (Eds.), Sleep, Neurotransmitters and Neuromodulators, Raven Press, New York, pp. 29-42. Sakai, K. (1985b) Anatomical and physiological basis of paradoxical sleep. In: D.J. McGinty, R. Drucker-Colin, A. Morrison and P.L. Parmeggiani (Eds.), Brain Mechanisms of Sleep, Raven Press, New York, pp. 111-137. Sakai, K. (1988) Executive mechanisms of paradoxical sleep. Arch. Ital. B i d , 126: 239-257. Sastre, J.-P., Sakai, K. and Jouvet, M. (1981) Are the gigantocellular tegmental field neurons responsible for paradoxical sleep? Brain Res., 229: 147-161. Stenberg, D. and Hilakivi, I. (1985) a 1 and a 2 adrenergic modulation of vigilance and sleep. In: A. Wauquier, J.-M. Gaillard, J.M. Monti and M. Radulovacki (Eds.), Sleep, Neurotransmitters and Neuromodulators, Raven Press, New York, pp. 69-77. Tononi, G. and Pompeiano, M. (1988) Local injection of the a,-adrenergic agonist clonidine into the locus coeruleus complex: Effects on desynchronized sleep. In: Proc. 9th Eur. Congr. Sleep Res., Jerusalem, September 4-9. Tononi, G., Pompeiano, M., Gianni, S. and Pompeiano, 0. (1988a) Enhancement of desynchronized sleep signs after microinjection of the P-adrenergic antagonist propranolol in the dorsal pontine tegmentum. Arch. Ital. Biol., 126: 119-123. Tononi, G., Pompeiano, M. and Pompeiano, 0. (1988b) Desynchronized sleep suppression after microinjection of the P-adrenergic agonist isoproterenol in the dorsal pontine tegmentum. Arch. Ital. B i d , 126: 125-128. Tononi, G., Pompeiano, M. and Pompeiano, 0. (1989) Modulation of desynchronized sleep through microinjection of P-adrenergic agonists and antagonists in the dorsal pontine tegmentum of the cat. Pfliigers Arch., 415: 142-149. Tononi,, G., Pompeiano, M., Cirelli, C. (1991) Suppression of desynchronized sleep through microinjection of the a,adrenergic agonist clonidine in the dorsal pontine tegmenturn of the cat. Pf7iiger.Y Arch., 418: 512-518. Ursin, R. and Sterman, M.B. (1981) A Manual for Standardized Scoring of Sleep-Waking States in the Adult Cat, Brain Information Service, Brain Research Institute, University of California, Los Angeles, 103 pp. Vanni-Mercier, G., Sakai, K., Lin, J.S. and Jouvet, M. (1989) Mapping of cholinoceptive brainstem structures responsible for the generation of paradoxical sleep in the cat. Arch. Ital. B i d , 129: 133-164. Vivaldi, E., McCarley, R.W. and Hobson, J.A. (1980) Evoca-
553 tion of desynchronized sleep signs by chemical microstimulation of the pontine brainstem. In: J.A. Hohson, M.A.B. Brazier (Eds.), The Reticular Formation Rei)isited, IBRO Monogruph Series, Yo/. 6 , Raven Press, New York, pp. 513-529.
Webster, H.H. and Jones, B.E. (1988) Neurotoxic lesions of the dorsolaterdi pontomesencephalic tegmentum-cholinergic cell area in the cat. 11. Effects upon sleep-waking states. Bruin Rrs., 458: 285-302.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Pro~re.ss1 , ~Brain Researrh, Vol. 88 0 1991 Elscvier Science Publishers B.V.
555
CHAPTER 39
Facilitation of learning consecutive to electrical stimulation of the locus coeruleus: cognitive alteration or stress-reduction? L. Velley
',B. Cardo ', E. Kempf
2,
P. Mormede ', S. Nassif-Caudarella and J. Velly
'
Laboratoire de Psychophysiologie, U R A CNRS 339, Auenue des Facultes, Talence, Centre de Neurochimie CNRS, Rue 8. Pascal, Strasbourg, INRA-INSERM U 259, Rue C. St Saens, Bordeaux and Institut de Pharrnacologie, Rue Human, Strasbourg, France
This chapter summarizes behavioral and neurochemical data on the delayed effect of locus coeruleus stimulation on learning capabilities in the rat. The initial observation showed that electrical stimulation of the locus coeruleus of a 15-day-old-rat improved the early stages of acquisition and extinction of a food-reinforced task performed 4 weeks later. Neurochemical lesion of the dorsal noradrenergic bundle performed 10 days before the stimulation did not attenuate the behavioral effect, whereas the lesion of the locus coeruleus proper suppressed the subsequent behavioral improvement. More recently we showed that the increase of adrenocorticotrophin release consecutive to a moderate stressful situation was significantly lower in previously stimulated rats than in implanted nonstimulated animals. Furthermore, we demonstrated that the neurochemical lesion of the locus coeruleus increased neophobia in the open-field as well as in a specific exploration task. Taken together these data strongly suggest that the long-term improvement in acquisition and extinction of locus coeruleus-stimulated rats results mainly from an attenuated stress reaction when these animals are confronted with a new environment (beginning of acquisition) or a new situation (beginning of extinction). Finally, we were interested in investigating the possibility of some long-term neurochemical modifications that could be related to the observed behavioral effects. The most significant modification observed concerned certain subpopulations of adrenoceptors in specific brain re-
gions. By using specific ligands of the p-, al- and a,-adrenoceptors, we studied the long-term effect (4 weeks) of the locus coeruleus stimulation on the kinetic characteristics of these three sub-types of receptors in four brain areas (the cortex, hippocampus, hypothalamus and brainstem). No significant alteration in the density of p binding sites was observed in any of the four structures analyzed; likewise locus coeruleus stimulation did not modify the density or affinity of the p-, a l - and a,-receptors in the brainstem. The density of a l - and a,-receptors was significantly increased in the cortex whereas in the hippocampus only the density of the a,-receptors was increased. Finally, a very large increase of the density of a,-adrenoceptors was observed in the hypothalamus (113%). In each case the increase in receptor density was also associated with a decreased affinity. A behavioral counterpart of these changes in the kinetic properties of the a,-receptors has been observed by using a pharmacological approach. The injection of small doses of clonidine induced, in stimulated rats (4 weeks after the stimulation), a delayed hyperactivity in the open-field 24 h after the injection whereas such hyperactivity was never observed in implanted non-stimulated animals. These neurochemical data are discussed. The stress-reduction hypothesis formulated above is tentatively related to the binding data which showed that the most significant increase of the a,-receptors was observed in the hypothalamus.
Key words: electrical stimulation, locus coeruleus, acquisition, extinction, food-reinforced operant task, neurochemical lesions, neophobia, adrenoceptors, clonidine
556
Introduction
on the acquisition and reversal of a light-dark
The studies summarized here originated from an initial observation showing that electrical selfstimulation elicited from the lateral hypothalamus of the young rat developed progressively between the ages of 15 and 45 days. Whereas in 15-day-old rats self-stimulation was observed to be at a very low level, a significant increase in self-stimulation performance was observed several days later, suggesting some maturation of the reinforcement processes in the lateral hypothalamus (LH). Thus, in order to investigate the possibility of modifying these reinforcement processes during this period of maturation, we decided to stimulate bilaterally (in an imposed fashion), the LH of 15-day-old rats and to analyze, 4 weeks later, the influence of this stimulation on the reinforcement processes tested using various learning procedures. We first demonstrated that LH electrical stimulation improved both the acquisition and extinction rates of a food-reinforced task (Velley and Cardo, 1977). In subsequent experiments, we observed the same general effect
discrimination, using either appetitive or aversive reinforcement, as well as on the acquisition of a one-way avoidance task (Velley et al., 1981). Two different questions were raised by these findings. First, is the learning improvement dependent on the age of the animal at the time of stimulation? This enhancement does not appear to be age-dependent since it was also observed following the stimulation of 90-day-old rats testing in an operant task 4 weeks later. Second, we wanted to know whether the same learning improvement could be obtained following stimulation of other brain regions particularly the extrahypothalamic areas known to support selfstimulation or whether the phenomenon was limited to the LH. In order to answer this question we stimulated, in different groups of rats, various extrahypothalamic brain regions, namely the locus coeruleus (LC), the dorsal and median raphe nuclei, the substantia nigra, the nucleus accumbens and the parietal cortex. We found that the post-stimulation increase in learning performance was not limited to the LH, but was dependent on
Fig. 1. Schematic representation of the parasagittal section (lat. 1.4 mm) of the rat brain (after Paxinos and Watson, 1986), showing the brain areas from which the electrical stimulation produced ( + I 4 weeks after the treatment an improvement of performance in the acquisition of an operant task. The sign - indicates brain areas from which the stimulation had no effect. Acb, n. accumbens; DR, dorsal raphe; LC, locus coeruleus: LH, lateral hypothalamus; mfb,medial forebrain bundle: MR, median raphe; SNC and SNR, substantia nigra. The two raphe nuclei are not in this sagittal plane.
the site of stimulation. As shown in Figure 1, three brain areas located in the brainstem induced a learning improvement 4 weeks after the stimulation, the dorsal and median raphe nuclei and the LC. In contrast, the stimulation of the substantia nigra (mainly the pars compacta), as well as the stimulation of the nucleus accumbens or of the parietal cortex (not shown) did not produce any modification of learning ability (Velley and Cardo, 1979, 1982). It is worth noting that although the substantia nigra and the nucleus accumbens are known to be sites for self-stimulation their imposed stimulation did not produce any modification of learning capacity. In contrast, the stimulation of the LC produced a significant improvement of learning whereas this nucleus does not support self-stimulation. Thus, these findings show that the increased learning capacity of stimulated rats is not due to an alteration of the reinforcement processes and suggest that the learning enhancement obtained following the stimulation of the LH could, in fact, result from the stimulation of noradrenergic and serotoninergic fibers. Finally, a global comparison of the data indicated that the rank order of efficiency of each brain structure, in terms of induced learning enhancement was as follows: LC, LH, dorsal raphe and median raphe. Therefore, all our more recent work has been centered on the analysis of the role of the coerulean system in this particular behavioral modulation. Stimulation studies
Detailed descriptions of the methods used have been published elsewhere (Velley and Cardo, 1977, 1979; Velley et nl., 1981, 1983). In all experiments, male Sprague-Dawley rats were used. Stimulation electrodes were implanted bilaterally in the LC under stereotaxic control when the rats were 13 days of age. Each electrode consisted of two twisted platinum-iridium wires, 0.09 mm in diameter. Stimulation was applied on days 15 and 16, over 15 min sessions separated by intervals of
.C
I
15
21
2s
li
h'
TIME (min)
A
B
Fig. 2. Effects of LC stimulation at 15 and 16 days of age on the acquisition and extinction of a continuous reinforced task performed 4 weeks later. A. Acquisition curves of S ( n = 201, I ( n = 18) and C ( n = 20) rats during the 1 h session (I vs C, n.s.; S vs I, P < 0.0001). B. Mean latencies of the 10th response in acquisition (I vs C, n.s.; S vs I, P < 0.001); C. Extinction curves of S ( n = 81, 1 ( n = lo), C (C = 10) rats. ( 5 ) indicates a 5-min session with reinforcement before the extinction session (S vs I vs C, n.s.). For the extinction curves, I vs C, ns.; S vs I, P < 0.0001).
30 min. The stimulation parameters were as follows: stimulations per hour, 2,580; duration, 200 msec; sinusoidal current, 100 Hz, 60 p A (peak to peak). Each rat was stimulated for a total of 4 h, 2 h on day 15 and 2 h on day 16. The electrodes were removed 1 day later under light anesthesia. Three groups of rats were used: stimulated rats (S), implanted but not stimulated rats (I), and non-implanted rats (0. Figure 2 summarizes the results of LC stimulation on the acquisition and extinction of a foodreinforced bar press conditioning (Skinner box) task performed 4 weeks later. For the acquisition session (A), each rat was placed in the test chamber for 1 h on a continuous reinforcement schedule (CRF) without previous training or shaping. Two measures of the animal's performance were taken: the number of lever presses per 5 min interval (A) and the latency (in sec) to obtain the
558
10th pellet (B). Following this acquisition session, the rats underwent one 15 min session daily until their performance reached a criterion of at least 25 responses/5 min. The last session was followed 24 h later by a 30 rnin extinction session immediately after an initial 5 min period with reinforcement (C). All behavioral tests were conducted blind. In all experiments, the t-test and analysis of variance with repeated measures were used to analyze the data. These results showed that the stimulation of the LC 4 weeks prior to the task produces a significant improvement of performance in the acquisition of an appetitively reinforced CRF task. This enhancement appeared from the outset of acquisition, as shown by the difference in the number of responses during the first 5 min of the test (Fig. 2A) and by the difference in the mean latencies to receive the 10th pellet (Fig. 2B). Extinction in S rats was also more rapid than that of the two other groups, in spite of the fact that during the first 5 min with continuous reinforcement the performance of the three groups was not significantly different. Locus coeruleus projections mediating the behavioral effects of stimulation
The findings of our stimulation studies strongly suggested that the LC system plays an important role in the observed behavioral effects. However, given the widespread projections of the LC (Lindvall and Bjorklund, 1974; Moore and Bloom, 19791, the effects could be due either to a longterm, pre- or post-synaptic modification in all the regions innervated by the nucleus, or to a modification located in a particular region of the brain. The largest ascending projection of the LC is the dorsal noradrenergic bundle (DNB). Consequently, in a first experiment, the DNB of 25-dayold rats was destroyed bilaterally at the level of mesencephalon by local injections of 6-hydroxydopamine (6-OHDA, 4 p g in 2 ~1 on each side). During the same operation, two electrodes were bilaterally implanted in the LC. The stimulation
was applied 8 days later. As in previous experiments, the acquisition of the operant task took place 4 weeks later. Four groups of rats were included: unlesioned and stimulated (NL-S), lesioned and stimulated (L-S), unlesioned and non-stimulated (NL-NS) and lesioned and nonstimulated (L-NS). The results of this experiment showed that, despite a near total loss of noradrenaline (NE) in the hippocampus and cortex (residual content: 3.1 and 1.1%,respectively) the lesion of the DNB did not have any clear and consistent behavioral effects. The overall performance of the L-S rats was lower than that of the NL-S rats, but better than that of the L-NS and NL-NS groups. Moreover, during the first 20 min of acquisition, the response rates of the NL-S and L-S groups were the same and the mean latency of the 10th response was not different in the two groups (data not shown). Consequently, in another experiment, we lesioned the LC proper by local injection of 6OHDA (4 pug in 1 p1 on each side). Two weeks Iater, the stimulation electrodes were implanted in the region of the LC. All the other aspects of the experiment (stimulation and behavioral testing) were the same as in the first experiment. Likewise, the same four groups of rats were used: L-S, NL-S, L-NS and NL-NS. The results of the acquisition session are summarized in Figure 3. LC stimulation alone again produced a significant improvement in performance. However, in contrast with the small effect of the DNB lesion, the destruction of the LC proper without stimulation (L-NS group) resulted in very poor performance during the acquisi:ion session. The overall performance of the L-S group was significantly better than the performance of the L-NS group. The residual content of NE after neurochemical lesion of the LC was measured in the hippocampus, the cortex, and the brainstem. Although the loss of NE in the three structures was relatively small compared with that found after DNB lesions, the differences between the NE content of the control group (NL-NS) and of
559 LO
.
NL-NS .NL-S
.
L-”S
.L-5
35
30
.E
E .
J
25
v1
; v)
20
Z
g 3 R
15
I
10
5
0
5
15
25
35
45
55
N5
L 5
TIME (min)
Fig. 3. Effects of the LC lesion on the learning improvement produced by the stimulation of the LC region. Number of rats: NL-S, 7; L-S, 9; NL-NS, 8; L-NS, 10. A. Acquisition curves of the 4 groups of rats. Statistical significance: NL-S vs NL-NS. P < 0.001; L-NS vs NL-NS, n.s.; L-S vs NL-S, P < 0.001; L-S vs L-NS, P < 0.02. B. Mean latencies of the 10th responses (NL-S vs NL-NS, P < 0.01; L-NS vs NL-NS, P = 0.05; L-S vs NL-S, P < 0.001; L-S vs L-NS, P < 0.01).
the two lesioned groups (L-NS and L-S) were significant (percentage of control: cortex, 37.2%; hippocampus, 60%; brainstem, 72.3%). In conclusion, these results confirm the involvement of the LC system in the long-term effect of the stimulation, but show that the rostra1 projections passing through the dorsal bundle, in front of the lesion, do not appear to be critically implicated in the observed post-stimulation effects (Velley et al., 1983). Given the behavioral data summarized above, we wanted to answer the following main questions: (1) What kind of behavioral function(s) was (were) improved by the LC stimulation? and (2) Is it possible to find some long-term neurochemical modifications that could be related to the observed behavioral effects? Cognitive enhancement or decreased neophobia? Some years ago, several theories postulated that the coerulean system was necessary, or was even
the sole brain system responsible for various behavioral functions such as ingestive behavior, arousal, sleep, reinforcement, learning and attention (see reviews in Amaral and Sinnamon, 1977; Clark, 1979; McNaughton and Mason, 1980). Two of these general hypotheses, the learning and attentional hypotheses, respectively, could possibly account for our data. However, despite early optimistic reports (Anlezark et al., 1973), more recent work has failed to confirm a noradrenergic component in the acquisition of many behavioral tasks (Amaral and Foss, 1975; Roberts et al., 1976; Sessions et al., 1976; Heybach et al., 1978; Koob et al., 1978). Numerous electrophysiological data suggest a role for the LC nucleus in the control of sensory inputs (Segal and Bloom, 1976; Cedarbaum and Aghajanian, 1978; Woodward et al., 1979; Foote et al., 1980; Waterhouse et al., 1988). For example, Aston-Jones and Bloom (1981) found that LC neurons in awake rats exhibit strong responses to certain environmental stimuli and concluded that the LC system “may function to facilitate transition between behavioral states.” This general hypothesis could also explain our observations, since the enhanced performance of the stimulated rats was observed particularly during behavioral transitions: the enhancement was first seen at the beginning of the first learning session (Fig. 2A). Likewise, after removal of the reinforcement during the extinction session the most significant decrease of response rate of stimulated rats took place during the first 10 min of the session (Fig. 2 0 . However, when the task was well learned no difference between stimulated and control rats was observed (Fig. 2C, first point of the abscissa). Thus, it seems that the behavioral enhancement appeared only when a parameter of the experimental situation was abruptly modified. If we suppose that the functioning of the coerulean system is improved in stimulated rats we might suggest that at each transitional point in a learning situation the processing of novel stimuli will be more efficient in these rats than it will be in non-stimulated animals.
5 60
However, given the widespread projections of the LC the role of this system in the transitional situations can be viewed from many different perspectives depending on the function of the particular brain area to which the noradrenergic fibers project. Most electrophysiological studies suggesting a role for the LC nucleus in the control of sensory inputs have been performed in the cortex or hippocampus. Likewise, recent data demonstrated that the electrical stimulation of the LC in adult cats restores plasticity in the ocular dominance of visual cortical cells (Kasamatsu et al., 1985). All these modulations, detected electrophysiologically either in the cortex or hippocampus, are the result of influences of the DNB fibers since the entire noradrenergic innervation of these forebrain areas transits via the fibers of this bundle. It is possible that such noradrenergic modulation of the target cells in the cortex and hippocampus is involved in some cognitive function as, for example, attentional processes. This attentional hypothesis was proposed by Mason (see review in Mason and Iversen, 1979) but was not later supported by findings of the effects of DNB lesions (Pisa and Fibiger, 1980; Owen et al., 1982; Carli ef al., 1983; Cole and Robbins, 1987; Pisa et al., 1988). Our data showing that the DNB lesion, in contrast to the LC nucleus lesion, did not seriously impair the acquisition of an appetitive operant task and did not suppress the positive influence of the LC stimulation agree with these negative results. Although the acquisition of this task is not a specific test of attention, it is difficult to imagine that acquisition of the task by a totally naive rat does not require some degree of attentional processes. Taken together, all the previous studies indicate that certain discrepancies appear, for the moment, between the electrophysiological and the behavioral observations, the former showing a clear modulating influence of N E on the reactivity of target cells in forebrain areas, the latter demonstrating a nearly total lack of influence of the DNB lesion on various behaviors, although
this same lesion produces an almost total noradrenergic denervation of these forebrain areas. A second possibility which could account for our behavioral results is to consider that each transitional point in a learning situation constitutes a stressful event and that the role of the coerulean system is to reduce this stress. The stress-reduction hypothesis, clearly formulated by Amaral and Sinnamon (1977) synthesizes two different kinds of data. First, the particular reactivity of the LC system to stressful situations is well documented (Stone, 1975; Anisman, 1978; Weiss et al., 1981; see review in Glavin, 1985). This observation is confirmed by studies of single-unit response of the LC neurons in awake animals showing that the most robust response from these neurons was to stressful stimuli (Abercrombie and Jacobs, 1987; Grant et al., 1988). Second, it is also well known that all transitional states are stressful and, in particular, produce an activation of the pituitary-adrenal axis. The introduction of an animal into a new environment or the omission of the reinforcement during extinction both produce an increase of corticosterone and adrenocorticotrophin (ACTH) levels (Coover, et al., 1971; Levine et al., 1972; Bassett et al., 1973; Davis et al., 1976; Hennessy et al., 1979; Pfister, 1979; File and Peet, 1980). Given these data, we suspected that the behavioral superiority of the LC-stimulated rats during each transitional state (beginning of acquisition or extinction) might result from such an attenuated stress reaction and not primarily from some cognitive modification. We have, in fact, obtained some indirect confirmation of this hypothesis. First, we compared the reactivity of the pituitary-adrenal axis of stimulated and implanted rats immediately after each rat was introduced for the first time for 10 min in an open-field (Velley and Mormede, unpublished data). The results are reported in the Table 1. The main observation is that the increase of ACTH levels in the stimulated rats to this moderately stressful situation was significantly lower than the increase found in implanted non-stimu-
56 I TABLE 1 Basal levels of corticosteroids and ACTH and modifications to these levels produced 4 weeks after electrical stimulation of the LC
I(n
Corticosteroids (ng/ml)
(Pg/ m 1)
Basal
Basal
12) 8 . 0 f 4 S ( n = 12) 3.2+0.8 =
Post-Exp.
ACTH Post-Exp.
174.6+32 16.7+2.3 414.7+66.6 155.8k 19.5 13.222.6 133 13 * *
Half the rats (6 S and 6 I) not tested in the open-field were used to estimate the basal levels of the two hormones. The other rats ( 6 S and 6 I) were introduced for 10 min in the open-field and sacrificed immediately after to measure the levels of the hormones (post exp.); * * P < 0.01.
lated rats. However, the basal levels of corticosteroids and ACTH measured in 6 S and 6 I rats not tested in the open-field were not significantly different. Although these data require confirmation they indicate that the LC stimulation induces a long-term hyporeactivity of the pituitary-adrenal axis when the animal is placed in a moderately stressful situation. In order to verify the implication of the LC proper in the reactivity to stressful situations, we analyzed the effects of neurochemical lesions of the nucleus on neophobia. We observed first, in agreement with other data (Heybac et al., 1978; Wendlandt and File, 1979; Britton et al., 19841, that this lesion decreased locomotor activity in the open-field test in a constant fashion. However, the activity of the lesioned rats increased progressively from days 7 to 10 and from the 12th day onwards, the locomotor activity of lesioned rats reached the level of control animals (NassifCaudarella et al., 1986). One of the main criticisms against the openfield however is the fact that this test is a forced exposure task, in which initial activity can be motivated either by escape or approach tendencies (discussed in Robbins, 1977; Corey, 1978). Consequently, the locomotor activity decrease after LC lesions could be due either to a decrease of exploration, i.e., increased neophobia, or to a decrease of escape tendencies, namely decreased
reactivity to a stressful event. In order to overcome this interpretative difficulty we recently compared the effect of a neurochemical LC lesion on the locomotor activity in the open-field and on exploration in a choice-task, the test of Hughes (1965). Briefly, the Hughes apparatus consisted of a plastic box that could be divided in half by means of temporary partitions. During the experiment which lasted 5 days, each rat lived permanently in one of the two boxes (the familiar box). Each day the partition was removed and the rat was allowed to explore the novel half of t h e box during 15 min. Thus, since in this task the rat could either enter the novel box or remain in the familiar box the level of exploration of the novel box could be more confidently attributed to approach tendencies. Figure 4 shows the main result of this experiment (Velley et al., 1988). Three groups of rats were used: lesioned rats (L) for which each LC was destroyed by local injection of 6-OHDA (4 p g in 1 PI), vehicle-injected rats (V> which received in each LC 1 pl of the vehicle, and control non-operated rats (C). The lesion and the vehicle injection were performed 4 weeks before the beginning of the behavioral tests and the rats were left undisturbed during this period. The top part of Figure 4 shows the mean time, out of 900 s, spent in the novel half of the Hughes apparatus during the five consecutive daily trials. The exploration time of the vehicle and control rats was about the same and did not vary significantly across the five trials. The exploration of the novel box by the lesioned rats was very low but slowly and progressively increased from the first to the fifth trial. The bottom part of Figure 4 shows the locomotor activity in the open-field of the same three groups of rats recorded over five consecutive daily trials. Despite the fact that the open-field test was performed after the test of Hughes, the locomotor activity of the lesioned rats was clearly inferior to the activity of the two other groups. The LC lesion induced a 53% loss of NE in the hypothalamus and a 78% loss in the hippocampus.
800
...
600-
0
-
-
u)
y
400-
2
200
-
02
1
2
3
4
5
3 4 TRIALS
5
Fig. 4. Effect of the bilateral neurochemical lesion of the LC on the exploration in the Hughes test and on locomotor activity in the open-field. Top: Mean time (+SEMI spent in the novel box of the Hughes apparatus by the rats lesioned 4 weeks before the first trial (L), by vehicle-injected (V) and by non-operated rats (C). Abscissa. The five trials. Ordinate. Time in sec out of 900 sec. Bottom. Mean number ( + SEMI of squares crossed during 5 min (ordinate) in the open-field. Abscissa. The 5 trials. L, n = 12; I , n = 6; C, n = 6. In the 2 tests the differences between the L rats and the two other groups were significant ( P < 0.001).
Taken together these data demonstrate convincingly that the neurochemical lesion of the LC increases neophobia and it may be suspected that the poor performance shown by the LC-lesioned rats in the acquisition of the operant task (Fig. 3) was due mainly to a decrease of exploration when these animals were introduced for the first time in the Skinner box. However, in the same experiment we studied the effect of the LC lesion on the reactivity of the pituitary-adrenal axis to a moderately stressful situation (introduction to a new environment). We did not observe significant modifications of corticosterone, ACTH or prolactin levels whether
the rats were previously tested in the Hughes apparatus and the open-field or were left undisturbed (Velley et af., 1988). This observation is in some disagreement with our stimulation findings, showing a clear decrease of ACTH levels in stimulated rats, consecutive to a moderate stress (Table 1). This discrepancy is presently difficult to explain. However, it is known that the hypothalamic noradrenergic innervation from the LC is limited and that the major source of this innervation is from the lower brainstem (Moore and Bloom, 1979; Moore and Card, 1984). About 90% of the N E supply of the paraventricular nucleus is provided by the A1 and A2 neuronal complexes, whereas LC participation is estimated to be about 8% (Sawchenko and Swanson, 1981). In agreement with these data it was demonstrated that the neurochemical lesion of the ventral noradrenergic bundle inhibits the ACTH stress-response (Szarfarczyk et al., 1985). However, this result was not confirmed by others showing that the same lesion or the direct infusion of 6-OHDA into the paraventricular nucleus did not modify the corticosterone response to stress (Feldman et al., 1986, 1988; Castagnt et al., 1990). Despite these conflicting data, we suggest that the absence of effect of the LC lesion on the hormone levels consecutive to a stress situation would be due to the fact that most fibers of the A1 and A2 neuronal complexes supplying the paraventricular nucleus were spared by the lesion. In contrast, the long-term effect of the LC stimulation on the ACTH release could result from the stimulation of these fibers, with some consecutive modulation of the noradrenergic input to the paraventricular nucleus. Does the LC stimulation induce long term modqications of biochemical parameters? Given the neurochemical homogeneity of the coerulean system, we looked for any long-term noradrenergic modifications that might be produced and which might be related to the behavioral changes. Two different experiments were performed.
563
TABLE 2 Brain noradrenaline levels and turnover in the LC of stimulated and non-stimulated rats 4 days and 4 weeks after the stimulation
4 days 4 weeks
N E Content WLC)
N E Decline after a-MpT
Turnover rat (ng/LC/h)
NS S
4.51 f 1.06 (8) 5.09 f 1.43 (8)
0.76 i 0 . 0 3 (16) 0.142 f 0.036 *** (16)
0.78 f 0.35 1.61 f 0.61
NS
9.83 f 1.45 (7) 8.83 f 1.76 (8)
0.109 i 0.03 (15) 0.039 f 0.01 **+ (16)
2.46 f 0.70 0.79 f 0.17
S
Turnover parameters were determined by blocking the synthesis of N E by intraperitoneal injection of a-methyl-purcr-tyrosine (a-MpT) 300 mg/kg. Regression lines were calculated from the least squares fit. * * * P < 0.001.
Alteration of the noradrenaline turnover following stimulation of the LC In this experiment, we stimulated the LC in exactly the same conditions as for the behavioral experiment and examined the N E levels and turnover in the cortex, hippocampus and LC 4 days and 4 weeks after the stimulation (Table 2). Levels of NE in the three structures studied were not significantly modified either 4 days or 4 weeks after the stimulation. However, 4 days after the treatment the turnover of NE was signifi-
cantly increased in the LC only. In contrast, 4 weeks after the stimulation a significant decrease of the NE turnover was observed in the LC.
Modulation of rat brain a-adrenoceptor populations after stimulation of the LC The aim of these experiments was to determine whether adrenergic receptors were involved in some way in the long-lasting behavioral modifications. We studied the long-term effect of LC stimulation on the characteristics of the a,-, aZ-
TABLE 3 Kinetic constants of a,-, a z - and p-adrenoceptors in different regions of the brain 4 weeks after LC stimulation. n, number of separate assays. * P < 0.05; * * P < 0.01; * * * P < 0.001
p (ICYP) Kd
C+I S
Hippocampus
C +I S
Hypothalamus
C +I S
Brainstem
C S
+I
Bnlm (fmol/mg prot.)
c~,[~H]Yohimbine
Kd
B ma x
Kd
Bmax
(pM)
(fmol/mg prot.)
(nM)
(fmol/mg prot.)
67.3 f 3.2 n=8 93.3 i 2.7 * * * n=4
231.0 f 13.5 n=8 374.3 f 57.4 * * n=4
3.5 k 0.2 n=6 6.2 f 0.4 * * * n=4
73.3 f 3.1 n=6 136.8 f 6.7 * * * n=4
97.7 f 2.7 n=3
94.4 i 3.2 n=5 104.0 i 9.2 n=3
156.6 i 4.0 n=5 148.0 f 11.5 n=3
2.3 f 0.2 n=5 6.0 f 0.1 * * n=3
111.4f 5.3 n=5 148.7 f 7.9 *** n=3
76.3 i 2.6 n=3 69.5 n=2
130.7 f 5.4 n=3 124.0 n=2
69.0 i 3.4 n=7 97.0 f 3.0 * * * n=4
201.0 i 21.0 n=7 254.0 f 34.0 n=4
6.2 f 0.6 n=7 10.5 f 0.9 * * n=4
172.0 i 14 n=7 366.0 f 25.5 * * * n=4
56.9 f 5.6 n=7 55.8f3.1 n=4
46.4 f 2.7 n=7 43.3 f 5 . 1
79.3 f 3.5 n=7 90.5 f 2.6 n=4
77.0 i 3.1 n=7 89.5 f 5.7 n=4
4.0 0.3 n=7 5.9 f 0.6 n=4
48.4 k 4.9 n=7 55.0 i 5.3 n=4
(PM) Cortex
a,[’HIPrazosin
76.6 f 4.4 n=10 82.6 f 6.8 n=5 69.0 f 6.3 n=5 77.3 i 6.2 n=3
78.2 f 2.0 n=lO
80.8 f 4.6 n=5 103.0 f 8.9 n=5
n=4
564 CLOl
and p-adrenoceptors, in 4 brain areas, the cortex, hippocampus, hypothalamus and brainstem (Velley et al., 1986). The ligands used were [12511iodocyanopindolol (ICYP) for /3-adrenoceptors, [ 3H]prazosin for a,-adrenoceptors and [3H]yohimbine for the a,-adrenoceptors. The results observed 4 weeks after the LC stimulation are shown in Table 3 and enable us to draw the following observations. No significant alteration in the number of pbinding sites was observed in any of the four structures analyzed; likewise, the LC stimulation did not modify the number or affinity of the p-, a I - and a,-receptors in the brainstem. In the cortex, the number of a,- and-a,-receptors was significantly increased (68 and 86%, respectively), but their affinity was simultaneously decreased. In the hippocampus, only the a,-adrenoceptor population was modified (increase of the B,,, and of the Kd). Finally, a very large increase of the density of a,-adrenoceptors was observed in the hypothalamus (113%) and associated as above with decrease in their affinity. In order to get a better idea of the time-course of the a-adrenoceptor increase, we performed another binding experiment in which the rats were sacrificed 2 weeks after the LC stimulation instead of after 4 weeks. Using this interval, the only significant modifications observed were an increase of a,-adrenoceptors in the cortex (19.4%) and in the hypothalamus (54%).
Confirmatory pharmacological investigations The binding data summarized above indicated that certain adrenoceptors located in certain brain areas were modified by the LC stimulation performed 4 weeks previously. The p-adrenoceptor populations were not modified in any of the regions analyzed and the three subtypes of the adrenoceptors in the brainstem were also not modified. In contrast, the density of a-adrenoceptors, mainly the a,-subtype, was apparently increased in the cortex, the hippocampus and the hypothalamus, whereas their affinity was decreased. Given this main effect of the stimulation
I
6
7
CL02.5
CLO5
1
t
8
3
10
CLO 10
t
11
12
18
is
m
21
22
DAYS
Fig. 5. Effects of clonidine (yg/kg) on open-field locomotor activity. Hatched bars: stimulated rats; open bars: implanted rats. Abscissa. Days of testing. Ordinate. Activity changes calculated as percentages of the mean activity levels of each group of rats during the first 5 days. * P < 0.05; * * P < 0.01; * * * P < 0.001.
on the a,-adrenoceptors, we decided to verify if a behavioral counterpart of the changes in the kinetic properties of these receptors could be observed. For this purpose we used a pharmacological challenge approach by comparing the effects of clonidine injections on locomotor activity in S and I rats in the open-field. To preferentially stimulate the a,-adrenoceptors, relatively low doses of the drug were used, namely 1, 2.5, 5 and 10 pg/kg. Four weeks after the LC stimulation the locomotor activity of the two groups of rats (S: n = 18; I: n = 10) was measured over a period of 5 min per day during 5 consecutive days. No drug was injected before these 5 trials. It was observed that the locomotor activity of the two groups was not different (number of squares crossed in 5 min: S, 73 f 5; I, 71 3). This observation indicates that the LC stimulTtion produces neither hyper- nor hypoactivity in the open-field. Figure 5 shows the effect of increasing doses of clonidine injected as indicated (arrows). Two main effects can be observed. The well-known sedative effect of clonidine was the same in both groups of rats when locomotor activity was measured 30 min after the drug injection. However, only stimulated rats exhibited a delayed rebound
565
hyperactivity recorded 24 h after the injection. For the 5 pg/kg dose, this rebound of activity was still detectable 8 days after the injection. During this period of rebound of activity the locomotor movements of the stimulated rats remained co-ordinated. The increase was mainly due to a higher rate (speed) of locomotion. Forward movements were often interrupted by rearing or exploration of the open-field. No stereotypy or signs of enhanced sympathetic activity were observed. These data suggest that the rebound activity observed in stimulated rats following clonidine is in some way dependent on the action of clonidine on particular a-adrenoceptor populations which are modified by the stimulation of the LC. In order to define the type of receptor implicated in this delayed effect of clonidine, we tested the influence of yohimbine (an antagonist of the a,-adrenoceptors) and the effects of prazosin (an antagonist of the a,-adrenoceptors) on the rebound of activity induced by clonidine in stimulated rats (Velley et al., 1982). At the doses used, the two antagonists had no effect on the locomotor activity of non-stimulated rats. However, the rebound of activity observed in stimulated rats was suppressed by prazosin during 24 h, whereas yohimbine did not significantly modify this rebound. Location of the a,-adrenoceptors involved The interpretation of the binding and pharmacological data summarized above is difficult given that the a,-adrenoceptors can be located both presynaptically on noradrenergic terminals and cell bodies and postsynaptically. Binding studies have shown that after central noradrenergic denervation the number of a,-binding sites is either increased or not modified in most of the blain regions analyzed (Pilc and Vetulani, 1982; U’Prichard, 1984; Gross, et al., 1985). Moreover, a,-adrenoceptors have been reported to be located on serotoninergic neurons in the rat cerebral cortex and hippocampus (Maura et al., 1982; Reinhard and Roth, 1982; Benkirane et al., 1985). In order to better define the location of the
1 m 120
1
t
\
I
8060.
40
,
‘O-
1
.
CIV LC
AYI
, OOR oJ 120
,
I
5
10
so 1m
5
10
50
-
100-
8060
-
40
.
20-
A LC-YR 0 m.Dn
100
msEs( pg / b )
Fig. 6 . Effect of increasing doses of clonidine on the open-field activity of different groups of rats. Abscissa. Doses of clonidine (log. scale). Ordinate. Activity changes calculated as percentages of the mean activity levels of each group of rats during the last two days before the first injection of clonidine. C-V. Unoperated and vehicle-injected rats ( n = 12). LC. Locus coeruleus-lesioned rats (6-OHDA, 4 pg/l PI in each LC) ( n = 6). DR, dorsal raphe-lesioned rats (5.7 DHT, 10 g g / 2 11.1) ( n = 6); MR, median raphe-lesioned rats (5.7 DHT: 10 p g / 2 4)( n = 8). LC-MR ( n = 8) and MR-DR ( n = 8), double-lesioned rats. LC-MR-DR, triple-lesioned rats ( n = 8).
a,-adrenoceptors which produce sedation after clonidine injection, we performed single or combined neurochemical lesions of the LC, the dorsal and median raphe nuclei and we then tested the sedative effects of increasing doses of clonidine in the open-field (Nassif-Caudarella et al., 1986). As shown in Figure 6 neither single nor combined lesions modified the response to clonidine and the slope of the decrease of activity produced by increasing doses of clonidine was the same in all groups whether lesioned or not, despite highly significant losses of NE and/or 5HT in the cortex, hippocampus and brainstem. In conclusion, the major result observed 4 weeks after the LC stimulation is a retarded
566
modulation of certain adrenoceptor subpopulations in particular brain regions. This modulation presents some regional and subtype specificity since neither the P-adrenoceptors nor the different adrenoceptors in the brainstem were modified. Moreover, a comparison of the binding data obtained both 2 and 4 weeks after the LC stimulation indicates that the maximum effect is not reached at the shorter time interval, suggesting that in our experimental conditions the a-adrenoceptor response to the stimulation is a rather slow and time-dependent process. The molecular mechanism underlying the increase of the a-receptors observed after LC stimulation remains to be elucidated. As already indicated, the particular localization of the a,-receptors is not yet known. Our behavioral data showing that the sedative action of clonidine is not modified by either single or combined lesions of the LC and the mesencephalic raphe nuclei (Fig. 6) confirm that the a-adrenoceptors involved in the sedative effect of this drug are localized neither on the noradrenergic nor on the serotoninergic fibers originating in these nuclei. Moreover, recent pharmacological as well as molecular cloning techniques clearly indicate the presence of at least 2 subtypes of the a , - and 3 subtypes of the a,-adrenoceptors (Bylund, 1988). Our analysis of the changes in locomotion of LC-stimulated rats produced by clonidine indicates that the rebound of activity is mainly due to activation of a,-adrenoceptors, but this effect may be dependent on the previous activation of the a,-adrenoceptors by clonidine. These data suggest that the LC stimulation induces, or enhances, a possible interaction between the two subtypes of a-adrenoceptors. Interactions between different adrenoceptors have been observed in the brain in several binding studies (Maggi et al., 1980; Swann et al., 1981), but it is presently unclear how these can account for our results. As shown in Table 3, the delayed a-adrenoceptor modulation was observed in three brain areas: a , and- a,-receptors were increased in the cortex and hypothalamus and a,-receptors in the
hippocampus. However, as already indicated, the learning improvement produced by the stimulation was not attenuated by the DNB lesion. Thus it is unlikely that the increase of the a-receptors in these DNB projection areas is involved in the observed learning enhancement. The second main result concerns the hypothalamus. In agreement with classical binding results (Leibowitz et al.,, 1982) as well as with quantitative autoradiographic measures (Palacios and Wamsley, 1984; pnnerstall et al., 1984), our results show that the hypothalamus has a high density of a,-reqeptor sites. Four weeks after the LC stimulation, dhe density of these sites is greatly increased. Can these modulations of hypothalamic adrenoceptors explain our behavioral data showing an improvement of learning capabilities of treated rats? The stress reduction hypothesis proposed above implies a major role for the hypothalamus and it is consequently possible that the reduction of the stress reaction in stimulated rats results from modifications of the noradrenergic influences on hypothalamic areas, consecutive to the increase of the a,-adrenoceptor populations. Although the effects of stress on these receptors are not well documented, a limited number of observations implicate a,-adrenoceptors in response to stress (Simson et al., 1986; Weiss et al., 1986; Nukina et al., 1987a,b; Stanford and Salmon, 1989). However there is little information on the role of these receptors in adaptation to stress. Some data suggest that N E may inhibit ACTH secretion by an action on the hypothalamus (Ganong, 1980; Suda et al., 1987) and it seems that the inhibition of the corticotrophin releasing factor by NE is mediated by an a-adrenoceptor mechanism (Buckingham and Hodges, 1979; Shimizu, 1984). Our data showing that the ACTH increase in response to a stress was significantly lower in LC-stimulated rats than in the control animals seems to be in agreement with this hypothesis. It may be supposed, for example, that the increased density of a,-adrenoceptors located on the corticotrophin releasing factor cells amplifies a noradrenergically medi~
567
ated inhibition. However, although promising, the present findings remain somewhat speculative, and require careful confirmation before the stress reduction hypothesis can be accepted. Acknowledgements These investigations were supported by grants from INSERM 78.1.011.6 and 82.6.017 and by the CNRS (URA 339). References Abercrombie, E.D. and Jacobs, B.L. (1987) Single-unit response of noradrenergic neurons in the locus coeruleus of freely moving cats. I. Acutely presented stressful and nonstressful stimuli. J. Neurosci., 7: 2837-2843. Amaral, D.G. and Foss, J.A. (1975) Locus coeruleus lesions and learning. Science, 188: 377-378. Amaral, D.G. and Sinnamon, H.M. (1977) The locus coeruleus: Neurobiology of a central noradrenergic nucleus. Prog. Neurobiol., 9: 147-196. Anisman, H. (1978) Neurochemical changes elicited by stress. Behavioral correlates. In A. Anisman and C. Bignami (Eds.), Psychopharmacology of Auersiidy Motivated Behauior, Plenum Press, New York, pp. 119-172. Anlezark, G.M., Crow, T.J. and Greenway, A.P. (1973) Impaired learning and decreased norepinephrine after bilateral locus coeruleus lesions. Science, 181: 682-684. Aston-Jones, G. and Bloom, F.E. (1981) Norepinephrine-containing locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Bassett, J.R., Cairncross, K.D. and King, M.G. (1973) Parameters of novelty, shock predictability and response contingency in corticosterone release in the rat. Physiol. Behau., 10: 901-907. Benkirane, S., Arbilla, S. and Langer, S.Z. (1985) Supersensitivity of a2-adrenoceptors modulating ( 3 H K H T release after noradrenergic denervation with DSP4. Eur. J. Pharmccol., 119: 131-133. Britton, D.C., Ksir, C., Thatcher-Britton, K., Young, D. and Koob, G.F. (1984) Brain norepinephrine depleting lesions selectively enhance behavioral responsiveness to novelty. Physiol. Behac., 33: 473-478. Buckingham, J.C. and Hodges, J.R. (1979) Hypothalamic receptors influencing the secretion of corticotrophin releasing hormone in the rat. J. Physiol. (London), 290: 421-431. Bylund, D.B. (1988) Subtypes of a,-adrenoceptors: Pharmacological and molecular biological evidence converge. TIPS, 9: 356-361. Carli, M., Robbins, T.W., Evenden, J.L. and Everitt, B.J. (1983) Effects of lesions to ascending noradrenergic neurones on performance of a 5-choice serial reaction task in
rats; implications for theories of dorsal noradrenergic bundle function based on selective attention and arousal. Behac. Brain Res., 9: 361-380. Castagnk, V., Rivet, J.M. and Mormtde, P. (1990) The integrity of the ventral noradrenergic bundle (VNAB) is not necessary for a normal neuroendocrine stress response. Brain Res., 511: 349-352. Cedarbaum, J.M. and Aghajanian, C.K. (1978) Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique. J. Comp. Neurol., 178: 1-16. Clark, T.K. (1979) The locus coeruleus in behavioral regulation: Evidence for behavior-specific versus general involvement. Behau. Neural Biol., 25: 271-300. Cole, B.J. and Robbins, T.W. (1987) Dissociable effects of lesions to the dorsal or ventral noradrenergic bundle on the acquisition, performance, and extinction of aversive conditioning. Behau. Neurosci., 101: 476-488. Coover, G.D., Goldman, L. and Levine, S: (1971) Plasma corticosterone increases produced by extinction of operant behavior in rats. Physiol. Behac., 6: 261-263. Corey, D.T. (1978) The determinants of exploration and neophobia. Neurosci. Biobehau. Rec., 2: 235-253. Davis, H., Memmott, J., McFadden, L. and Levine, S. (1976) Pituitary-adrenal activity under different appetitive extinction procedures. Physiol. Behau., 17: 687-690. Feldman, S. Conforti, N. and Melamed, E. (1986) Norepinephrine depletion in the paraventricular nucleus inhibits the adrenocortical responses to neural stimuli. Neurosci. Lett., 64: 191-195. Feldman, S. Conforti, N. and Melamed, E. (1988) Involvement of ventral noradrenergic bundle in corticosterone secretion following neural stimuli. Neuropharmacology, 27: 129- 133. File, S.A. and Peet, L.A. (1980) The sensitivity of the rat corticosterone response to environmental manipulations and to chronic chlordiazepoxyde treatment. Physiol. Behau., 25: 753-758. Foote, S.L., Aston-Jones, G. and Bloom, F.E. (1980) Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proc. Natl. Acad. Sci., USA, 77: 3033-3037. Ganong, W.F. (1980) Neurotransmitters and pituitary function: Regulation of ACTH secretion. Fed. Proc., 39: 29232930. Glavin, G.B. (1985) Stress and brain noradrenaline: A review. Neurosci. Biobehac. Rec., 9: 233-243. Grant, S.J., Aston-Jones, G. and Redmond, D.E., Jr. (1988) Responses of primate locus coeruleus neurons to simple and complex sensory stimuli. Brain Res. Bull., 21: 401-410. Gross, G., Gothert, M., Glapa, U., Engel, G. and Schiimann, H.J. (1985) Lesioning of serotoninergic and noradrenergic nerve fibres of the rat brain does not decrease binding of 3H-clonidine and 3H-rauwolscine to cortical membranes. Naunyn Schmiedebergs Arch. Pharmacol., 328: 229-235. Hennessy, M.B., Heybach, J.P., Vernicos, J. and Levine, S. (1979) Plasma corticosterone concentrations sensitively reflect levels of stimulus intensity in the rat. Physiol. Behau., 22: 821-825.
568 Heybach, J.P., Coover, G.D. and Lints, C.E. (1978) Behavioral effects of neurotoxic lesions of the ascending monoamine pathways in the rat brain. J. Comp. Physiol. Psychol., 92: 58-70. Hughes, R.N. (1965) Food deprivation and locomotor exploration in the white rat. h i m . Behau., 13: 30-32. Kasamatsu, T., Watabe, K., Heggelund, P. and Scholler, E. (1985) Plasticity in cat visual cortex restored by electrical stimulation of the locus coeruleus. Neurosci. Res., 2 365386. Koob, G., Kelley, A.E. and Mason, S.T. (1978) Locus coeruleus lesions: Learning and extinction. Physiol. Behau., 20: 709716. Leibowitz, S.F., Jhanwar-Uniyal, M., Dvorkin, B. and Makman, M.H. (1982) Distribution of a-adrenergic, p-adrenergic and dopaminergic receptors in discrete hypothalamic areas of rat. Brain Res., 233: 97-114. Levine, S., Goldman, L. and Cooper, G.D. (1972) Expectancy and the pituitary-adrenal system. In Ciba Foundation Symposium, Vol. 8, Elsevier, Amsterdam, pp. 281-291. Lindvall, 0. and Bjorklund, A. (1974) The organization of the ascending catecholamine neuron systems in the rat brain as revealed by the glyoxylic acid fluorescence method. Acta Physiol. Scand., Suppl., 412: 1-48. Maggi, A, U'Prichard, D.C. and Enna, S.J. (1980) p-adrenergic regulation of a2-adrenergic receptors in the central nervous system. Science, 207: 645-647. Mason, S.T. and Iversen, S.D. (1979) Theories of the dorsal bundle extinction effect. Brain Res. Rev. 1: 107-137. Maura, G., Gemignani, A. and Raiteri, M. (1982) Noradrenaline inhibits central serotonin release through a2adrenoceptors located on serotonergic nerve terminals. Naunyn Schmiedebergs Arch. Pharmacol., 320: 272-274. McNaughton, N. and Mason, S.T. (1980) The neurophysiology and neuropharmacology of the dorsal ascending noradrenergic bundle. Prog. Neurobiol., 14: 157-219. Moore, R.Y. and Bloom, F.E. (1979) Central catecholamine neuron systems: Anatomy and physiology of the norepinephrine and epinephrine systems. Ann. Reu. Neurosci., 2: 113-168. Moore, R.Y. and Card, J.P. (1984) Noradrenaline-containing neuron systems. In A. Bjorklund and T. Hokfelt (Eds.), Handbook of Chemical Neuroanatomy. Classical Transmitters in the CNS, Vol. 2, Part 1, Elsevier, Amsterdam, pp. 123-156. Nassif-Caudarella, S., Kempf, E. and Velley, L. (1986) Clonidine-induced sedation is not modified by single or combined neurochemical lesions of the locus coeruleus, the median and dorsal raphe nuclei. Pharmacol. Biochem. Behau., 25: 1211-1216. Nukina, I., Glavin, G.B. and La Bella, F.S. (1987a) Chronic stress affects a,-adrenoceptors in brain regions of the rat. Res. Comm. Psychol. Psychiatry Behau., 12: 53-60. Nukina, I., Glavin, G.B. and La Bella, F.S. (1987b) Acute cold-restraint stress affects a,-adrenoceptors in specific brain regions of the rat. Brain Res., 401: 30-33. Owen, S.R., Boarder, M., Gray, J.A. and Fillenz, M. (1982)
Acquisition and extinction of continuous and partially reinforced running in rats with lesions of the dorsal noradrenergic bundle. be ha^^ Brain Res., 5: 11-41. Palacios, J.M. and Wamsley, J.K. (1984) Catecholamine receptors. In A. Bjorklund, T. Hokfelt and M.I. Kuhar (Eds.), Handbook of Chemical Neuroanatomy, Vol. 3, Elsevier, Amsterdam, pp. 325-351. Paxinos, G. and Watson, C. (1986) The Rat Brain in Stereotaxic Coordinates. Academic Press, New York. 237 pp. Pfister, H.P. (1979) The glucocorticosterone response to novelty as a psychological stressor. Physiol. Behau., 23: 649652. Pilc, A. and Vetulani, J. (1982) Attenuation by chronic imipramine treatment of ('H)-clonidine binding to cortical membranes and of clonidine-induced hypothermia: The influence of central chemosympathectomy. Brain Res., 238: 499-504. Pisa, M. and Fibiger, H.C. (1980) Noradrenaline and discrimination learning: Failure to support the attentional hypothesis. Soc. Neurosci. Abstr., 6: 724. Pisa, M., Martin-Iverson, M.T. and Fibiger, H.C. (1988) On the role of the dorsal noradrenergic bundle in learning and habituation to novelty. Pharmacol. Biochem. Behau., 30: 835-845. Reinhard, J.F., Jr. and Roth, R.H. (1982) Noradrenergic modulation of serotonin synthesis and metabolism. I. Inhibition by clonidine in uiuo. J. Pharmacol. Exp. Ther., 221: 541-546. Robbins, T.W. (1977) A critique of the methods available for the measurement of spontaneous motor activity. In L.L. Iversen, S.D. Iversen and S.J.H. Snyder (Eds.), Handbook of Psychopharmacology, Principles of Behavioral Pharmacology, Vol. 7, Plenum Press, New York, pp. 37-82. Roberts, D.C.S., Price, M.T.C. and Fibiger, H.C. (1976) The dorsal tegmental noradrenergic projection: An analysis of its role in maze learning. J. Comp. Physiol. Psychol., 90: 363-372. Sawchenko, P.E. and Swanson, L.W. (1981) Central noradrenergic pathways for the integration of hypothalamic neuroendocrine and autonomic responses. Science, 214: 685687. Segal, M. and Bloom, F.E. (1976) The action of norepinephrine in the rat hippocampus. IV. The effect of locus coeruleus stimulation on evoked hippocampal unit activity. Brain Res., 107: 513-525. Sessions, C.R., Kant, G.J. and Koob, G.F. (1976) Locus coeruleus lesions and learning in the rat. Physiol. Behac., 17: 853-859. Shimizu, K. (1984) Effects of a , and a,-adrenoceptor agonists and antagonists on ACTH secretion in intact and hypothalamic deafferented rats. Jpn. J. Pharmacol., 36: 23-33. Simson, P.G., Weiss, J.M., Hoffman, L.J. and Ambrose, M.J. (1986) Reversal of behavioral depression by infusion of an a-2 adrenergic agonist into the locus coeruleus. Neuropharmacology, 25: 385-389. Stanford, S.C. and Salmon, P. (1989) Neurochemical correlates of behavioural responses to frustrative nonreward in
569 the rat: Implications for the role of central noradrenergic neurones in behavioural adaptation to stress. Exp. Brain Res.. 75: 133-138. Stone, E.A. (1975) Stress and catecholamines. In A.J. Friedhoff (Ed.), Catecholamines and Behavior, Vol. 2, Plenum Press, New York, pp. 31-72. Suda, T., Yajima, F., Tomori, N., Sumitomo, T., Nakagami, Y., Ushiyama, T., Demura, H. and Shizume, K. (1987) Inhibitory effect of norepinephrine on immunoreactive corticotropin-release factor release from the rat hypothalamus in citro. Life Sci., 40: 1645-1649. Swann, A.C., Grant, S.T., Hattox, S.E. and Maas, J.W. (1981) Adrenoceptor regulation in rat brain: Chronic effects of a , and a,-receptor blockers. Eur. J. Pharmacol., 73: 301305. Szafarczyk, A., Alonso, G., Ixart, G., MalaVal, F. and Assenmacher, I. (1985) Diurnal-stimulated and stress-induced ACTH release in rats is mediated by ventral noradrenergic bundle. Am. J. Physiol., 249: 219-226. U’Prichard, D.C. (1984) Biochemical characteristics and regulation of brain a,-adrenoceptors. Ann. N.Y. Acad. Sci., 340: 55-75. Unnerstall, J.R., Kopajtic, T.A. and Kuhar, M.J. (1984) Distribution of a,-agonist binding sites in the rat and human central nervous system; analysis of some functional, anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agents. Brain Res. Reu., 7: 69- 101. Velley, L. and Cardo, B. (1977) A long-term effect of an early stimulation of the lateral hypothalamus on the acquisition and extinction of a food reinforced operant conditioning in the rat. Neurosci. Lett., 5: 221-226. Velley, L. and Cardo, B. (1979) Long-term improvement of learning after early electrical stimulation of some central nervous structures: Is the effect structure and age-dependent? Brain Res. Bull., 4: 459-466. Velley, L. and Cardo, B. (1982) Facilitation of acquisition and extinction of an operant task four weeks after stimulation of brainstem aminergic nuclei of the rat. Behau. Neural Biol., 35: 395-407. Velley, L., Chassaing, J.M. and Cardo, B. (1981) Learning improvement of appetitively or aversively reinforced lightdark discrimination and reversal four weeks after electrical stimulation of the lateral hypothalamus of the rat. Brain Res. Bull., 6: 377-383.
Velley, L., Kempf, E. and Cardo, B. (1982) Locomotor activity of rats after stimulation of the nucleus locus coeruleus region or after lesion of the dorsal noradrenergic bundle: Effects of clonidine, prazosin and yohimbine. Psychopharmacology, 78: 239-244. Velley, L., Nassif, S., Kempf, E. and Cardo, B. (1983) Enhancement of learning four weeks after stimulation of the nucleus locus coeruleus in the rat: Differential effects of dorsal noradrenergic bundle lesion and lesion of the locus coeruleus proper. Brain Res., 265: 273-282. Velley, L., MormZde, P. and Kempf, E. (1988) Neurochemical lesion of the nucleus locus coeruleus increases neophobia in a specific exploration task but does not modify endocrine response to moderate stress. Pharmacol. Biochem. Behau., 29: 1-7. Velly, J., Cardo, B. and Velley, L. (1986) Delayed up-regulation of a-adrenoceptor populations in particular regions of the rat brain after stimulation of the nucleus locus coeruleus. Neuroscience, 18: 321-328. Waterhouse, B.D., Sessler, F.M., Cheng, J.T., Woodward, D., Azizi, S.A. and Moises, H.C. (1988) New evidence for a gating action of norepinephrine in central neuronal circuits of mammalian brain. Brain Res. Bull., 21: 425-432. Weiss, J.M., Goodman, P.A., Losito, B.G., Corrigan, S., Charry, J.M. and Bailey, W.H. (1981) Behavioral depression produced by an uncontrollable stressor: Relationship to norepinephrine, dopamine and serotonin levels in various regions of the rat brain. Brain Res. Reu., 3: 167-205. Weiss, J.M., Simson, P.G., Hoffman, L.J., Ambrose, M.J., Cooper, S. and Webster, A. (1986) Infusion of adrenergic receptor agonists and antagonists into the locus coeruleus and ventricular system of the brain. Effects on swimmotivated and spontaneous motor activity. Neuropharmacology, 25: 367-384. Wendlandt, S. and File, S.E. (1979) Behavioral effects of lesions of the locus coeruleus noradrenergic system combined with adrenalectomy. Behau. Neural Biol., 26: 189201. Woodward, D.J., Moises, H.C., Waterhouse, B.D., Hoffer, B.J. and Freedman, R. (1979) Modulatory actions of norepinephrine in the central nervous system. Fed. Proc., 38: 2109-2116.
This Page Intentionally Left Blank
C.D. Barnes and 0. Pompeiano (Eds.) Progress in Brain Research, Vol. 88 <3 19Y1 Elsevier Science Publishers B.V.
571
CHAPTER 40
Plasticity of sensory responses of locus coeruleus neurons in the behaving rat: implications for cognition
'
S.J. Sara and M. Segal
' Departement de Psychophysiologie, Laboratoire de Physiologie Nerveuse C.N.R.S. Gif sur Yvette, France and
Center for
Neuroscience, Weizmann Institute, Rehouot, Israel
The gating and tuning actions of noradrenaline (NA) at post-synaptic sites have been highly suggestive of an important role for the locus coeruleus (LC) in attention, learning and memory. By recording the activity of single units in the LC in behaving rats in a strictly controlled conditioning paradigm, direct evidence was provided that this nucleus is engaged during specific aspects of learning. The neuronal response to a discrete sensory stimulus was monitored as a function of the changing significance of the stimulus i.e., when it was novel, during habituation, associative learning, reversal and extinction. Both appetitive and aversive paradigms were used. We consistently observed differential conditioned re-
sponding with food reinforcement, while when footshock reinforcement was used, there was an increase in response to both C S + and CS-. In both paradigms, the LC response disappeared when the conditioning was expressed at a behavioral level, to reappear vigorously as soon as the stimulus reinforcement contingencies were changed, i.e., during reversal or extinction. These results suggest that the LC does not mediate specific sensory or associative information necessary for ongoing performance but shows remarkable plasticity of sensory responding as a function of changing cognitive significance of the stimulus.
Key words: locus coeruleus, single unit activity, plasticity, learning, habituation, extinction
Introduction In spite of the abundance of electrophysiological evidence that noradrenaline (NA) is a potent neuromodulator, involved in enhancing and gating signals (Madar and Segal, 1980; Rogawski and Agahjanian, 1980; Waterhouse and Woodward, 1980; Waterhouse et al., 1988; Kayama et al., 1982) and promoting synaptic plasticity throughout the nervous system (Neuman and Harley, 1983; Stanton and Sarvey, 1986; Hopkins and
Johnston, 1988; Dahl and Sarvey, 1989; see Harley, 1987 for review), the functional significance of noradrenergic action must ultimately be studied at a behavioral level. There the evidence for an essential role for NA in learning and memory is scarce and far from compelling. In awake, behaving animals the first demonstration of the neuromodulatory properties of NA was in a study showing that iontophoretic application of NA to the auditory cortex of a monkey enhanced signal/noise in response to the vocalization of a
512
conspecific (Foote et al., 1975). In the rat, electrical stimulation of the locus coeruleus (LC) selectively enhanced hippocampal neuronal responses to tones associated with food, while increasing inhibition to the same tones before they had acquired a significance (Segal and BIoom, 1976). While these two studies demonstrated in awake animals that NA could modify neuronal responses to discrete stimuli, especially if they had a biological significance for the organism, they do not provide evidence that this modulation actually improves cognitive performance. The obvious strategy in addressing this question lies in enhancing or inhibiting the release of NA and evaluating its effect on evoked potentials at postsynaptic sites and at the same time on learning or memory. The localization of the entire population of cortically projecting NA neurons in one small pontine nucleus LC simplifies the first part of the problem to some extent. Neurotoxic or electrolytic lesion to the LC or its forebrain projection fiber system, the dorsal bundle, depletes the forebrain of NA. Most behavioral investigators have addressed the question of the role of NA in learning and memory by using specific neurotoxic lesions to the coeruleocortical system. This has generated substantial literature which has been thoughtfully reviewed on several occasions (Robbins, 1984; Robbins and Everitt, 1987). For the most part these studies have failed to provide unequivocal support for the role of this system in attention, learning and memory (Pisa and Fibiger, 1983; Robbins, 1984; but see also Pineda et al., 1989; Selden et al., 1990a,b). Although deficits do exist, animals with very little NA in the forebrain do well on many challenging cognitive tasks (Sara et al., 1984; Sara, 1985a,b; Robbins and Everitt, 1987). Electrical or pharmacological stimulation of neurons within the LC increases release of NA in the forebrain (Tanaka, et al., 1976). Since the activity of the cells of the LC is regulated by inhibitory autoreceptors, agonists and antagonists of these receptors are useful pharmacological tools to excite or inhibit the system and can be
used during behavioral tasks. Systemic injections of idazoxan, the a,-adrenoceptor antagonist, increases firing of LC cells by about 80% at 2 mg/kg (Sara and Devauges, 1989). We used idazoxan to assess the effects of stimulation of the NA system in complex behavioral situations. Idazoxan is particularly effective in facilitating performance in tasks which require shifts in attention between sensory modalities, accompanied by changes in behavioral strategies. The same stimulation enhances behavioral response to novelty and change in response-reinforcement contingencies (Sara et al., 1988; Devauges and Sara, 1990). The relatively few behavioral studies using electrical stimulation of the LC have shown facilitation of successive discrimination reversals (Segal and Edelson, 1978) and of memory retrieval after a long retention interval (Sara and Devauges, 1988). Stimulation in the region of the dorsal bundle produced similar memory facilitation (Sara et al., 1980; Dekeyne et al., 1987). Recent evidence suggests that the effect is mediated through a @-receptor, since the animals pretreated with the @-receptor antagonist propranolol are not facilitated by the LC stimulation (Devauges and Sara, 1991). While these studies show that LC stimulation is most effective in facilitating cognition, if it is applied when the animal has to adapt to novelty or change in significance of a stimulus, to further determine the role of LC in learning it is important to know when in the learning process the LC is active and engaged. We suggest that the stimulation is effective because it enhances a physiological event, i.e., activation of LC cells in a particular cognitive context. Therefore the LC should be preferentially engaged during learning when stimuli are novel or when there is a change in the significance of the stimulus. We recorded activity of LC cells in response to novel to-beconditioned stimuli, habituation, and associative conditioning, reversal and extinction, to ascertain when in the learning process the LC is actually active and engaged. A Pavlovian differential conditioning paradigm was used in order to precisely
573
control the conditioned stimuli (CSs) and their relation to the reinforcement and to establish a strict temporal relation between the neuronal and the behavioral response. Both appetitive and aversive versions of the paradigm were used, because Rasmussen and Jacobs (1986) reported that, in cats, there was no conditioning to a stimulus associated with food reinforcement, but good differential responding in a conditioned emotional response (CER) protocol, concluding that LC is involved in learning exclusively when the level of stress and anxiety is high. In the aversive conditioning studies, the CSs were preceded by, and contained within, a flashing light which served as an experimentally controlled context. Interest in the LC response to context lies in the fact that previous experiments have shown that contextual cues can serve as very effective memory facilitators (Deweer et al., 1980). We have suggested that the mechanism of action might lie in a conditioned response of LC neurons to the context associated with the prior learning (Sara, 1985b). Lesion studies from two different laboratories have led to diametrically opposite conclusions concerning the role of the coeruleocortical system in mediating context effects (Archer et al., 1982; Mohammed et al., 1986; Selden et al., 1990).
Methods Particular care is taken in these experiments to assure that we are recording from a noradrenergic cell of the LC. For this reason we have taken the costly option of recording only one cell from each rat, in order to have clear histological verification of the placement of the electrode. As a pharmacological control, we injected either clonidine or idazoxan after the behavioral analysis. Male Sprague-Dawley rats weighing between 280-400 g were used in these experiments. They were food deprived for 24 h before the appetitive experiments and handled for two or three days before surgery. They were exposed to the experi-
mental cage for 15 min on the day before surgery, and in the appetitive conditioning experiments they were shaped to drink sweet water from a dipper during this session. Single-unit recordings were made through etched tungsten microelectrodes (2-5 M a impedance) glued to a movable microdrive. A 60 p m wire, insulated except for the tip, implanted in the midbrain, served as a reference, with an uninsulated wire wrapped around stainless steel skull screws as ground. In later experiments two microelectrodes of the same impedance were glued together with a distance of less than 200 p m between the tips. Differential recording between the two provided much more stable recording with practically no muscle or movement artefacts. In the aversive conditioning experiments, animals were prepared with electrodes attached to the caudal neck muscles to measure myogram (EMG) responses to the CSs and UCSs. Rats were anesthetized with pentobarbital (50 mg/kg, supplemented as necessary). They were fixed in a stereotaxic apparatus with the nose angled down ( - 10 mm from the interaural 0). The skull was exposed and holes were drilled for skull screws and for the reference electrode. The movable microdrive assembly was advanced into a hole drilled above the LC and the electrical activity was monitored. The signal was fed through a probe to a differential amplifier (Grass P5 series), filtered (400-10 kHz band pass), displayed on an oscilloscope and through an audio monitor. LC cells were identified as discharging broad spikes (0.8-1 msec) at about 1-2 Hz, encountered just below the fourth ventricle, about 5.4-6 mm from the surface of the brain. The cells characteristically responded to a paw pinch with an excitation followed by a longer (500 msec) inhibition, as described by Cedarbaum and Aghajanian (1978). The entire microdrive assembly was fixed to the skull with dental cement. The head set containing the ground wires, reference electrodes, leads for micro electrode attachment, and EMG electrodes was also cemented to the skull.
574
Behauioral apparatus and procedures Experiments were carried out in a 26 X 26 X 60 cm Plexiglas chamber with a grid floor through which electric shock could be delivered. A loudspeaker was located on the wall of the chamber and a light bulb was fixed to one side. The cage had a small window opening to which was fixed an automated dipper apparatus which delivered sweet water. The entire chamber was enclosed in a sound attenuated, electrically shielded box fitted with a transparent window through which the rat could be observed. The behavioral response was the EMG, quantified through a trigger device, which proved to be a reliable measure of conditioned response. In the appetitive conditioning experiments, the behavioral response was measured as a head movement by means of a wire attached to the recording cable. The signal was amplified and quantified as in the aversive experiments. Chronic recording procedure The animal was connected to a cable to which was fixed a field effect transistor, leading to a Grass differential amplifier. The signals were filtered, amplified and monitored as during implantation of the electrode which was advanced in increments of 20 pm. The rat was held gently in the hand of the experimenter while this manipulation was taking place and the procedure did not cause any visible stress to the rat. When a stable well-differentiated unit was encountered, its spontaneous firing pattern was observed as a function of behavior. LC cells are inhibited during sleep, increase spontaneous firing rate during exploratory behavior and are excited by tactile and auditory stimulation. If a cell had these characteristics, conditioning proceeded. Conditioning protocol The experiment began with habituation to the CSs. Habituation trials were limited to 20, because we found that more pre-exposure to the to-be-conditioned stimulus impaired subsequent conditioning (cf. Lubow, 1973). It was imperative
to carry out the whole experiment as rapidly as possible and in a single session. The differential conditioning procedure, using high- (HF) and low-frequency (LF) tones (8000 and 3000 Hz, *SO db on a 40 db background), provided a control for increases which might be due to sensitization rather than to the association of the CS with the reinforcement. After conditioning was well-established at a behavioral level, the significance of the CSs was reversed, i.e., the former CS - was now followed by the dipper and the reinforcement, or by the footshock, in the case of the aversive experiments. After 40 reversal trials, reinforcement was withheld until there was full extinction of the unit response. Then pharmacological experiments were performed to verify the noradrenergic nature of the cell. Drugs used were the a,-antagonists which increase firing rate of LC cells, yohimbine (1 mg/kg) or idazoxan (2 mg/kg), or the a,-agonist, clonidine (10 pg/kg), which inhibits LC firing. During the appetitive conditioning session one of the tones, 400 msec duration, was followed 200 msec after offset by the appearance of the dipper (CS 1, which itself made a slight noise. The other tone was not reinforced (CS - 1. (In some experiments the LF tone was the CS + 1. In the aversive conditioning the experiments, the H F and LF tones were preceded by, and contained within, a light stimulus. The offset of the light and CS tone was followed immediately by a foot shock (FS) of 0.5 mA intensity.
+
+
Data collection and analysis The output of the spike amplitude discriminator was fed to the computer. One channel counted cell firing, the other digitalized movement or EMG. Peristimulus time histograms in 100 or 50 msec bins were generated from these data. For quantitative analysis, each behavioral or cellular response to the light or the CS tones was transformed into a score reflecting its difference from mean baseline counts of the 1 sec period immediately preceding the stimulus on that trial. Statistical analyses were performed between the mean
575
counts during the stimulus and the baseline, in the case of the response to the light (paired t-test). To evaluate the conditioning to CS + and CS - , these data were submitted to analysis of variance for repeated measures. A main effect of trials in the absence of interaction indicates conditioning to both CS + and CS - ; a significant trials x CS interaction indicates differential conditioning. Paired t-tests were applied where appropriate. Histology The rats were perfused with 10% formalin with the recording electrode in place. The brain was sectioned and stained with cresyl violet and electrode tips were identified under a compound microscope.
Results The data reported here are based on eleven single units recorded from eleven rats having histologically confirmed electrode placement in the LC and, when possible (n = 61, pharmacological verification by injection of yohimbine, idazoxan or clonidine immediately after the conditioning experiments. The electrodes were located in the dorso-anterior portion of the LC; an example of the electrode placement is illustrated in Figure 1. Unit activity as a function of behauior In general, we observed relationships between LC spontaneous and evoked activity and behavior similar to those reported previously (Aston-Jones and Bloom, 1981). The cells responded to tones and tactile stimulation, but not always to flashing light. The amplitude and persistence of the response to sensory stimulation varied greatly from one cell to another. The LC cells were always inhibited during sleep. In most animals the cells were inhibited during grooming, eating and drinking, as reported previously (Aston-Jones and Bloom, 19811, but in four animals, there was no
Fig. 1. Typical electrode placement in the dorsal part of the locus coeruleus.
relationship between cell activity and consumatory or grooming behavior. Appetitiue conditioning Of the twelve rats recorded in the appetitive experiments, six had electrodes in the LC. Five of the six showed significant differential behavioral conditioning, usually beginning at the third block of 20 trials. There were differences in the magnitude and persistence of the initial unit response to the tones, with some cells rapidly habituating to one or the other, or both of the CSs, and others showing little habituation over the 20 trials. As subsequent conditioning proceeded, the response to the CS - generally decreased and response to
576
RAT 21
A REVERSAL
ACQUISITION
CS
trials
1-10
+ (LF
CS
TONE)
- (HF TONE)
CS + (HF TONE)
CS - (LF TONE)
1
1
11-20
51 1
1.51
21-30
1
,L
)& ..
. .. ..
-.-
. . .. .. . . ..
. . .
.. . ..
.. .. ..
. .. . ...
tone 'dipper
.5
.I
15
3'-40
I
51-60
. .. ...
B
5 l1
h
... ... .. .. .... ...
... .... ... ... ...
... .... ..... ... ...
.4 spikes/ 1OOrns
:u
15
41-50
.
... .. .
1
.9
i
5
I 51 l
:L tone
dipper
tone
dipper
ACQUISITION
REVERSAL
577 R 1 7 CSS
interval
90-
x
I
.5.A
.
o . " l
A
-
€
C
0
0
.A
a
-
L 0,
-
.
-
0 .t o -
o
1 a "
03
-
a
--'
n
-.5
~~~
1-20
21-40
41-60
61-80
C1
C2
C3
C4
reversal R1
C1
C2
C3
C4
R1
R2
reversal
R2
Fig. 3. Histograms, mean behavioral responses to CS + and CS - . Curves, mean unit responses to CS + and CS - (triangles, 3A) and to the 200 msec delay following the CSs (circles 3B), during conditioning and reversal. There is a significant difference between CS + and CS - at C3 ( P = 0.05); the difference between responses to CS + and CS - offset was significant for the last three conditioning blocks ( P = 0.002) and the increase in response during the first block of reversals ( P < 0.05). Values were calculated as for Figure 2. Note that the unit activity is entirely independent of the movement.
the CS + increased, independently of the magnitude of the initial response and its resistance to habituation. An example is shown in Figure 2 (rat 21). The initial unit response to the CS - is greater than that to the CS + . As the CS is associated with reinforcement, there is an increase in response to the CS + and a decrease to the CS - . The interaction over the first four blocks (80 trials) is significant (F = 3.8; P < 0.01). By the time the conditioned response is well established at a behavioral level, the LC response to both CSs is greatly diminished, as can be seen in the figure. This marked decrease in responding to both CSs during overtraining was observed in
+
all rats receiving overtraining trials ( n = 3; Figs. 2,3,4). On the other hand, when the significance of the tones is reversed, the cell immediately begins to fire to both of them on the very first reversal trials. This was observed for the two rats for whom reversal trials were run (Figs. 2,3). This high level of responding occurs at the beginning of the reversal; the increase in unit response from the last block of conditioning trials to the first block of reversals is significant (F(1,39) = 13.01; P < 0.0021, Fig. 2). There was a significant decrease in response rate from the first to the second block of reversal trials, for both tones (F(1,39) = 10.1; P < 0.005).
Fig. 2. A. Peristimulus time histograms of a single unit response in the LC during appetitive conditioning and reversal. Data are presented in 100 msec bins in blocks of 20 trials; at the left are means of 10 CS + trials (LF tone) and at the right, means of 10 CS trials (HF tone). First eight bins, baseline; next four, 400 msec tone, followed by 200 msec delay; last five, water dipper reinforcement. Note that during the initial trials, the cell responds more to the H F tone than to the LF tone. As conditioning proceeds there is decrease in response to the nonrewarded tone and an increase in response to the CS + . During overtraining, after eighty trials (last two blocks) there is very little cell response to either tone. When the significance of the tones is changed during reversal, the unit immediately responds to both tones. B. Quantification of the data presented in A. The mean number of counts per bin and per period were calculated and the response to the tone was subtracted from the mean baseline value for each trial. The graph represents the mean (kSEM) of this difference for each block of 10 CS + and CS - trials, during the 120 conditioning trials and the 40 reversal trials. * P = 0.01; * * * difference between C6 and R1 ( P = 0.002).
578
RAT 13 CELL
:.iL cs +
trials
1-20
1
BEHAVIOR
cs -
cs -
cs + 39 30
1 L
LL
391
21-40
41-60
:'I L
6 1-80
I
~
L,
L
LrL
L
one dipper
one
lipper
tone dipper
tone
dipper
Fig. 4. Peristimulus time histograms of LC unit response (left) and head movement response (right) during conditioning. Note the absence of response at the beginning of conditioning, the rapid acquisition of discrimination shown at a cellular level in the second block of 20 trials, which decreases as the differential behavioral response is established (third block). During overtraining, the behavioral response is to the dipper itself.
The discrimination was also expressed as an increase in responding during the 200 msec interval between the CS offset and the delivery of reinforcement. This is illustrated by rat 17 in Figure 3. There were no initial differences in unit responses to the two tones in this rat, and there was significantly greater response to CS + than to CS - on the third block of trials. But the differential response was even more robust during the 200 msec delay in this animal. Here the discrimination appears at a cellular level as early as the second block of trials and is maintained for the entire conditioning period (F = 12.81; P < 0.002). Moreover, the interaction CSs X trials is significant for the last training trial and the first reversal trial (F = 4.2, P = 0.051, indicating that the change in significance of the tones is recognized very rapidly at a cellular level and is expressed as a change in response to tone offset. A similar pattern of results was observed in rat 11
+
in which the unit was held for 40 conditioning trials, which was not sufficient for significant differential conditioning at a behavioral level. There was, however, a differential cellular response during the CS - US interval (F(1,79) = 4.05; P = 0.0591, which was significant on the fourth block of trials ( t = 2.3; P = 0.04, data not shown). As in rat 21, there was the same tendency for the differential unit response to the CS itself, but less robust than the response to CS offset. The unit conditioned response to the CS + was expressed earlier in the conditioning trials than the behavioral response. This is illustrated in Rat 13, Figure 4, where there was good habituation to the two tones before conditioning, both at a cellular and a behavioral level, as can be seen from the absence of responding in the early conditioning trials. The unit, depicted on the left of the figure, shows clear differential responding to the CS during the second block of trials (t(9) =
+ +
+
HA B I T VAT ION
A
7 1-10
,
5 3
.
.
..
1 gJ l,I
11-20
. .
.
.
. :. .
,
.
.
.
1
L--l---
.
.
.
.
u; .
.
.
.
.
1
.
.I
-
I
,
,
.
CONDITIONING
11 - 20
trlals
L F tone
7
31-40
5
.
1 I
l\,
7 5
.
..
. . : .
. . . .
.. ..
.
.
3
1
i ".tone light
.
,
.
11-20
...
I
. .
. .
. .
. ..
..
.
.
,$ shbck
.
EXTINCTION
:LF:
,.
.
1
.
. .'
:
iHFl
.. . ..
..
. .
. ...
1
I t . ,
,
shock
:
'
t x e
light
31-40
9 7 3 1
,
.. .
. .
!dl .
.
... .
.
... .
l l ;. J .; I , .. .., . . . .
.
..
..
.
. ..
. . . ,ll;
. . ..
.
,,
.
.L1-'
'tone
v light
... .
.. .
.
:
I
.
.
;
.. ,
.
. . . .
..
,,,
,
M-lL21- 3 0
-
/
. .
...
. . . .. .. .. .. ... ... .,'nYlk, . .
..
..
.
: : t
.
... .. .. ..
..
,,;*;,
..
JI,,
. ll
..
. ...
7
:
.
.
..
9
71-80
.
.
,,:
4
..
.
.
3 1
I_Lu
.
..
.
..
1-10
..
.
Ill, .
..
7
.. . .
.
. :
.
.
.
'
31
1
. .. .. ..
GI[,,
,
LF
7 5 3
. .. .. ..
1
.
HF tone
.. .. ..
_
.
HF
.
REVERSAL
7-10
1 '; kL j 5
3
1-10
,
RAT 108
L F tone
HF tone
trials
579
,
,
; . , '
.
.
I,
:cT.
"t o n e
light
Fig. 5. A. Peristimulus time histograms during habituation, conditioning, reversal, and extinction of LC unit responses to CS + and CS - . During conditioning, the LF tone is followed immediately by footshock (note that the response during footshock is not shown, all bins are empty). During reversal the HF tone is followed by footshock; at extinction, the tones are presented in the same sequence, but there is no footshock. Both tones are preceded by one sec by a light which persists until the tone offset. Note the robust responses to the novel tones, their rapid hibituation, and the renewal of the response to both tones as soon as the stimulus-shock contingencies are introduced. The response is to both CS and CS - and does not persist during overtraining. There is little change in response to tones during reversal, but the unit immediately begins firing to both CSs during extinction. Unit response to the light is very small, and extinguishes rapidly, but shows the same pattern of response to change in stimulus-shock contingencies as the response to tones. During reversal, where there was no change in response to tones, there is a renewed response to the light.
+
580
B
.8
-
.6
-
.2
-
u)
E
. s 0
v)
+
n
T
X
a
Q
0-
--
10
10
20
HABITUATION
.25
-
20
30
40
50
--
10
CONDITIONING
20
30
40
10
REVERSAL
20
30
40
trlrlr
EXTINCTION
C RESPONSE TO CONTEXT
E
0
ln
>
.20-
0
Y
Q
U
1
.lo-
.05]
.02
4----r+ 10
20
1-1
HABITUATION
10
20
30
40
CONDITIONING
-
10
20
30
REVERSAL
40
-
10
20
30
40 trials
EXTINCTION
Fig. 5. B and C . Quantification of the data for tones and light, respectively. Note the differences in the scales by the two graphs, reflecting the differences in magnitude of the responses. The calculations were as for Figure 2. * Significantly different from the previous block of trials.
2.7; P = 0.024) which disappears during overtraining. At a behavioral level, there is a general increase in activity during the second block of
trials, but the differential responding only appears during the third block (t(9) = 2.3; P = 0.045). Note that during overtraining, the appear-
581
HABITUATION
ance of the dipper, itself, controls the response. Another example of this phenomenon appears in the reversal data of Rat 17. As shown in Figure 3, during reversal there is no behavioral change in relative responding to CS + and CS - until the second block of trials, while at a cellular level the change in significance of the CSs is expressed at the very first reversal trials. Thus the change in significance of the stimuli is recognized at a cellular level before the information is used by the rat to adapt the behavioral response.
Auersive conditioning (CER) Of the eight rats recorded in the CER experiments, five had electrodes confirmed histologically to be in the LC. All rats showed significant differential behavioral conditioning to CS + and CS - . None of the LC units displayed differential conditioning to CS + and CS - in the CER situation. There was a great deal of difference in the individual unit responses to the two tones and the light and in the habituation of those responses; subsequent conditioning was a function of this initial response. An example of a rapidly habituating LC cell is shown in Figure 5. The cell showed habituation of responses to both CSs, an immediate increase in responding to both tones at the onset of conditioning, which did not persist over conditioning trials. There was no increase in responding when the significance of the stimuli was reversed (which is not surprising in the absence of differential responding at a cellular level), but as soon as the reinforcement was withheld, the cell resumed responding vigorously to both CSs and was remarkably resistant to extinction. Figure 6 illustrates an example of a cell which showed sensitization to both the HF and LF tones, as shown in the increase in response between the first two blocks of habituation trials. There was no habituation to either tone, even after 60 trials. During conditioning, the cell continued to respond to both CSs throughout the 80 trials. On the other hand, the response to the light rapidly habituated, with an immediate reap-
L
41-60
LJ4-L llght tone
CONDITIONING llght tone+ SHOCK
61-80
Fig. 6. Peristimulus time histograms during habituation and conditioning of the LC cell which showed no habituation to CSs. Note the sensitization to the CSs during the second block of habituation trials. The response to both CS+ and CSsustains throughout the 8 conditioning trials. The response to the light, on the other hand, rapidly habituates to resume again as soon as conditioning trials are introduced. The response drops out during overtraining. The light and tone were each of 2 sec duration and the shock was 1 sec. The response to shock is not shown and the bins are empty.
pearance of the previously habituated response on the earliest conditioning trials, and disappearance during overtraining trials. The CER to light was studied in Rat 10; the
582
results are shown in Figure 5C. There is complete and rapid habituation to the light in twenty trials, an increase in response as soon as conditioning begins ( t = 2.1; P = 0.09, still another increase which is sustained over thirty reversal trials with a significant decrease on the last block of reversals ( t = 3.6; P = 0.002). When reinforcement is withheld (first extinction block) there is again an increase is responding to the light ( t = 2.3; P < 0.05), followed by a relatively rapid extinction. In this rat, as in the previous one, there was a response to both CS and CS - tones throughout the experiment, a response which did not habituate and did not change according to the stimulus-reinforcement contingencies. One recording was made from a cell in the pontine reticular cell group, rostra1 and ventral to the LC. This cell responded very little to either tone during habituation trials, but showed significant differential responding by the second block of conditioning trials, which sustained for the remaining thirty trials (t(19) = 2.1, P < 0.05). The differential CER disappeared after just ten reversal trials. Extinction to both CSs was likewise very rapid. The animal expressed differential CER at a behavioral level by total inhibition in the EMG, 200 msec before the tone which was followed by the shock. A similar pattern of results was observed for a cell recorded from the lateral vestibular nucleus, just ventral and posterior to the LC. There was no cellular response to idazoxan in either of these animals.
+
Discussion The most striking and consistent observation in these studies was the immediate response of LC cells to any change in stimulus-reinforcement contingencies in both appetitive and CER conditioning. This response to change was even more reliable than response to a novel stimulus and is likely to be of important functional significance in understanding the role of the LC in cognitive processes. The LC response occurred when the stimulus was first associated with reinforcement,
and when the reinforcement was withheld (extinction). The response to reversal in significance of the CS + and CS - only occurred if the cell had initially responded differentially during conditioning. While this was observed only during appetitive conditioning, it does not seem to be a question of the nature of the reinforcement or the amount of stress in the situation, because there was a renewed response to the contextual stimulus (the light) when the significance of the tones was reversed in two CER experiments (Figs. 5B, 6). Here the LC units responded to the reversal of CS + and CS - , but it was expressed as a renewed response to the light, which preceded both the CS + and the CS - . Given the homogeneity of the cells of the LC in the rat, and reports of previous investigators (Aston-Jones and Bloom, 19811, the variability in the magnitude and persistence of responses to auditory and visual stimuli was unexpected. Some cells showed robust responses to both tones and did not habituate to either one (e.g., Fig. 6). Others responded vigorously before conditioning to the HF tone and little to the LF tone (Fig. 2), while other cells preferred the LF tone. The response to the light was generally smaller and more readily habituated than the response to the tones, in some instances there was no response to the light as a novel stimulus. Fluctuations in vigilance state cannot account for the variability of the responses and their habituation rates (Aston-Jones and Bloom, 1981). If the vigilance level is reflected by the baseline rate of firing in the LC (Foote et al., 1980) it is clear from the data in Figure 5 that the unit responds vigorously to both tones after 60 habituation trials, even though there is a decrease in vigilance. In fact, the marked sensitization seen in the second block of habituation trials, occurs while the spontaneous firing rate of the cell is decreasing. At the same time, on the same tonic background, there is clear habituation to the light. It is also clear in the record of rat 10, Figure 5A, that the changes in phasic response to the tones proceed independently of the tonic background activity. These
583
differences might be attributed to relative salience of the stimuli for the individual rat. Species differences in salience of stimuli have been suggested to account for differences in LC responses to various environmental stimuli in rat, cat and monkey (Jacobs, 1986; Grant et al., 1988). Within species, differences might be accounted for by differential pre-experimental experience with the stimuli in question. The important point is that the subsequent conditioned responding of the cell is partly a function of the amplitude and persistence of the initial response to the to-be-conditioned stimulus, especially in the CER paradigm. LC units showed differential responding in an appetitive situation early in the conditioning trials, sometimes before the behavioral expression of the discrimination (Figs. 3,4). The LC response disappears when the behavioral response is well established. The response to change in stimulusreinforcement contingencies was immediate and robust. In the aversive situation, there was increased general responding to both CS + and CS - during conditioning, which likewise did not persist during overtraining. The difference between appetitive and aversive conditioning concerning differential responding to CS and CS appears to be a function of the amount of stress in the situation, as there was differential responding of the LC unit when the FS level was very low. In addition to the nature of the reinforcement and the resultant amount of stress in the situation, another difference between the two protocols was the presence in the CER experiments of the contextual light stimulus preceding and containing the tone CSs. This could conceivably account for the lack of differential responding at a cellular level, although it did not prevent expression of the discrimination at a behavioral level. The observation that the LC neurons respond to the contextual stimulus, especially during change in the cognitive environment, does not resolve the problem of the discrepancies concerning the role of this system in the treatment of contextual information, found in the lesion stud-
+
ies (Mohammed et al., 1986; Selden et al., 1990a,b). One group found that lesioned rats have a broadened scope of attention and condition more to contextual cues and less to the CS, suggesting that the presence of NA promotes a focussing of attention on relevant stimuli (Selden et al., 1990). The other group observed in a series of experiments, that lesioned animals fail to notice contextual information and thus concluded NA depletion induces a narrowed attention field (Archer et al., 1982; Mohammed et al., 1986). Such issues will have to be resolved by studying the effects of NA release at postsynaptic sites, on the local electrophysiological response and, simultaneously, on behavioral adaptation to the change in stimulus-reinforcement contingencies. Nevertheless, the conditioned LC response to the light contextual stimulus in a controlled conditioning situation here, reinforces our hypothesis concerning the mediating mechanism of contextual cue reminder effects seen in several previous experiments (Deweer et al., 1980; Sara et al., 1980; Gisquet and Alexinsky, 1986; Dekeyne et al., 1987). We suggested that the context presented before a retention test can act as a CS to elicit a neuronal response in the LC. In this way the LC participates in memory retrieval processes by providing the necessary modulatory influences in brain regions more explicitly involved in stimulus selection and response output. Rasmussen and Jacobs (1986) found, in the cat, that LC neurons do not condition to stimuli associated with food reward, but show good differential CERs, and from this they concluded that LC neurons play a role in anxiety and adaptation to environmental challenges involving stress. Redmond and Huang (1979) drew a similar conclusion from observations in monkeys. Our LC stimulation and recording experiments suggest that the LC is involved in learning and cognitive adaptation on a much broader scale. The responses of LC neurons to changes in significance of stimuli are consistent with previous results indicating that stimulation of the system is particularly effective in behavioral situations re-
584
quiring shifts in attention and/or behavioral strategies (Segal and Edelson, 1978; Velly, et al., 1985; Sara and Devauges, 1988, 1989; Devauges and Sara, 1990). The vigorous and prompt response of LC neurons to a change in stimulus-reinforcement contingencies, suggests that the stimulation provided an extra “boost” to a physiological response (LC cell firing) already triggered by the cognitive demands of the situation. The results of the LC stimulation studies and the analysis of the postsynaptic effects of NA have been highly suggestive of a role for the coeruleo-noradrenergic system in attention and cognitive processes dependent upon attention. Recording the activity of these cells in the behaving animal in a strictly controlled conditioning paradigm has provided direct evidence for this role. The present results show the LC does not mediate specific sensory or associative information necessary for ongoing performance, but shows remarkable plasticity of sensory responding as a function of the changing cognitive significance of the stimulus. Acknowledgements This research was supported by a twinning grant from the European Science Foundation and by a grant from the CNRS “Action Initiative Europe”. The authors thank G. Lefloch and G. Dutrieux for technical assistance and L. Collet for drawing the figures. References Archer, T., Cotic, T. and Jarbe, T. (1982) Attenuation of the context effect and lack of unconditioned stimulus pre-exposure effect in taste-aversion learning following treatment with DSP4, the selective noradrenaline neurotoxin. Behau. Neural Biol., 35: 159-173. Aston-Jones G. and Bloom F.E. (1981) Norepinephrine-containing locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environment stimuli. J. Neurosci., 1: 172-176. Cedarbaum, J. and Aghajanian, G. (1978) Activation of locus coeruleus neurons by peripheral stimuli: Modulation by a collateral inhibitory mechanism. Life Sci., 23: 1383-1392.
Dahl, D. and Sarvey, J.M. (1989) Norepinephrine induces pathway specific long-lasting potentiation and depression in the hippocampal dentate gyrus. Proc. Natl. Acad. Sci. USA, 86: 4776-4780. Dekeyne, A,, Deweer, B. and Sara, S.J. (1987) Background stimuli as a reminder after spontaneous forgetting: Potentiation by stimulation of the mesencephalic reticular formation. Psychobiology, 15: 161-166. Devauges, V. and Sara, S.J. (1990) Activation of the noradrenergic system facilitates an attentional shift in the rat. Behav. Brain Res., 39: 19-28. Devauges, V. and Sara, S.J. (1991) Memory retrieval enhancement by locus coeruleus stimulation: Evidence for mediation by /3-receptors. Behau. Brain Res., 43: 93-97. Deweer, B., Sara, S.J. and Hars, B. (1980) Contextual cues and memory retrieval in rats: Alleviation of forgetting by a pretest exposure to background stimuli. Anim. Learn. Behau., 8: 265-272. Foote, S., Freedman, R. and Oliver, A. (1975) Effects of putative neurotransmitters on neuronal activity in monkey auditory cortex. Brain Res., 86: 229-242. Foote, S., Aston-Jones, G. and Bloom, F. (1980) Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proc. Natl. Adpd. Sci USA, 77: 3033-3037. Gisquet, P. and Alexinsky, T. (1986) Does contextual change determine long-term forgetting? Anim. Learn. Behau., 14: 349-358. Grant, S., Aston-Jones, G. and Redmond, D. (1988) Response of primate locus coeruleus neurons to simple and complex stimuli. Bruin Res. Bull., 21: 401-410. Harley, C. (1987) A role for norepinephrine in arousal, emotion and learning?: Limbic modulation by norepinephrine and the Kety hypothesis. Prog. Neuropsychopharmacol. Biol. Psychiatry, 11: 419-458. Hopkins, W. and Johnston, D. (1988) Noradrenergic enhancement of long term potentiation at mossy fiber synapses in the hippocampus. J. Neurophysiol., 59: 667-687. Jacobs, B. (1986) Single unit activity of locus coeruleus neurons in behaving animals. Prog. Neurobiol., 27: 183-194. Kayama, Y., Negi, T., Sugitani, M. and Iwama, K. (1982) Effects of locus coeruleus stimulation on neuronal activities of dorsal lateral geniculate nucleus and perigeniculate reticular nucleus of the rat. Neuroscience, 7: 655-666. Lubow, R. (1973) Latent Inhibition. Psychol. Bull., 79: 398407. Madar, Y.and Segal, M. (1980) The functional role of the noradrenergic system in the visual cortex. Activation of the noradrenergic pathway. Exp. Brain Res., 41: 814. Mohammed, A,, Callenholm, N., Jarbe, T., Swedberg, M., Danysz, W., Robbins, T. and Archer, T. (1986) Role of central noradrenaline neurons in the contextual control of latent inhibition of taste aversion learning. Behac. Brain Res., 21: 109-118. Neumann, R.S. and Harley, C . (1983) Long-lasting potentiation of the dentate gyrus population spike by norepinephrine. Brain Res., 273: 162-165. Pineda, J., Foote, S. and Neville, H. (1989) Effects of locus
585 coeruleus lesions on auditory, long-latency, event-related potentials in monkey. J. Neurosci., 9: 81-93. Pisa, M. and Fibiger, H. (1983) Intact selective attention in rats with lesions of the dorsal noradrenergic bundle. Behac. Neurosci., 4: 519-529. Rogawski, M. and Aghajanian, G. (1980) Modulation of lateral geniculate neurone excitability by NE microiontophoresis or locus coeruleus stimulation. Nature (London), 287: 731-734. Rasmussen, K. and Jacobs, B. (1986) Single unit activity in the freely moving cat. 11. Conditioning and pharmacological studies. Brain Rex, 371: 335-344. Redmond, D. and Huang, Y. (1979) New evidence for locus coeruleus-norepinephrine connection with anxiety. Life Sci., 25: 2149-2162. Robbins, T. (1984) Cortical noradrenaline, attention and arousal. Psychol. Med., 14: 13-21. Robbins, T. and Evebitt, B. (1987) Comparative functions of the central noradrenergic, dopaminergic and cholinergic systems. Neuropharmacology, 26: 893-901. Sara, S.J. (1985a) The locus coeruleus and cognitive function: Attempts to relate noradrenergic enhancement of signal/noise in the brain. Physiol. Psychol., 13(3): 151-162. Sara, S.J. (198%) Noradrenergic modulation of selective attention: Its role in memory retrieval. In D. Olton, E. Gamzu and S. Corkin (Eds.), Memory dysfunctions: A n integration of animal and human research from clinical and preclinical perspectives. Ann. N.Y. Acad. Sci., 444: 178-193. Sara, S.J. and Devauges, V. (1988) Priming stimulation of locus coeruleus facilitates memory retrieval in the rat. Brain Res., 438: 299-303. Sara, S.J. and Devauges, S. (1989) Idazoxan, an a,-antagonist, facilitates memory retrieval in the rat. Behau. Neural Biol., 51: 401-411. Sara, S.J., Deweer, B. and Hars, B. (1980) Reticular stimulation facilitates retrieval of a “forgotten” maze habit. Neurosci. Lett., 18: 211-217. Sara, S.J., Dekeyne, A,, Raquin, L. and Brouillet, E. (1984) Approach and avoidance discrimination learning and retention after treatment with the noradrenergic neurotoxin DSP4. Neurosci. Lett., Suppl., 18: 119.
Sara, S.J., Devauges, V. and Segal, M. (1988) Locus coeruleus engagement in memory retrieval and attention. In M. Sandler, A. Dahlstrom and R. Belmaker (Eds.), Progress in Catecholamine Research. Alan R. Liss, New York, pp. 155-161. Segal, M. and Bloom, F. (1976) The action of norepinephrine in the rat hippocampus.111 Hippocampal cellular responses to locus coeruleus stimulation in the awake rat. Brain Rex, 107: 499-512. Segal, M. and Edelson, A. (1978) Effects of priming stimulation of catecholamine containing nuclei in rat brain on runway performance. Brain Res. Bull., 3: 203-206. Selden, N., Robbins, T. and Everitt, B. (1990a) Enhanced behavioral conditioning to context and impaired behavioral and neuroendocrine responses to conditioned stimuli following ceruleocortical noradrenergic lesions: Support for an attentional hypothesis of central noradrenergic function. J. Neurosci., 10: 531-539. Selden, N., Cole, B., Everitt, B. and Robbins, T. (1990b) Damage to ceruleo-cortical noradrenergic projections impairs locally cued but enhances spatially cues water maze acquisition. Behau. Brain Rex, 39: 29-51. Stanton P.K. and Sarvey J.M. (1986) Norepinephrine regulates long-term potentiation of both the population spike and dendritic EPSP in hippocampal dentate gyrus. Brain Res. Bull., 18: 115-119. Tanaka, C., Inagaki, C. and Fujiwara, H. (1976) Labeled noradrenaline release from rat cortex following electrical stimulation of locus coeruleus. Brain Res., 106: 384-389. Velly, J., Kempf, E., Cardo, B. and Velley, L. (1985) Long-term modulation of learning following locus coeruleus stimulation: Behavioral and neurochemical data. Physiol. Psychol., 13(3): 163-172. Waterhouse, B.D. and Woodward, D. (1980) Interaction of norepinephrine with cerebrocortical activity evoked by stimulation of somatosensory afferent pathways in the rat. Exp. Neurol., 67: 11-34. Waterhouse, B.D., Sessler, F.M., Cheng, J.-T., Woodward, D.J., Azizi, S.A. and Moises, H.C. (1988) New evidence for a gating action of norepinephrine in central neuronal circuits of mammalian brain. Brain Res., 21: 425-432.
This Page Intentionally Left Blank
C.D. B a r n o and 0. Pompeiano (Eds.) Progress in Bruin Research, Vol. 88 0 19Y1 Elsevier Science Publishers B.V.
587 CHAPTER 41
Axonal sprouting of noradrenergic locus coeruleus neurons following repeated stress and antidepressant treatment S. Nakamura Department of Physiology, Faculty of Medicine, Kanazawa Uniuersiw, Kanazawa, Japan
Plastic changes in axon terminals of NA LC neurons following repeated stress and antidepressant treatments were examined using electrophysiological or morphological methods. For stress treatment, rats restrained in a small cage were immersed up to the neck in warm water for 10 min daily. Electrophysiological experiments were performed under urethane anesthesia on the day following the termination of stress treatment. To quantify the density of cortical axon terminals arising in the LC. the percentage of LC neurons activated antidromically from the cerebral cortex was assessed. The percentage of LC neurons showing antidromic response to cortical stimulation was increased in the animals stressed for two weeks but not for one week. Since threshold currents for antidromic activation were not changed by the stress treatment, the observed changes were interpreted as morphological (axonal sprouting) rather than physiological consequences in NA axon terminals of LC neurons. To test the ability of antidepressants to induce the regeneration of central NA axons, local injections of 6-OHDA were made
bilaterally into the symmetrical sites of the FC. Two weeks after the 6-OHDA injections, the same cortical site of one hemisphere was infused with the antidepressant MPL, DMI, or MIA, and the corresponding site of the other hemisphere with SAL. The density of glyoxylic acid-induced catecholamine fibers was greater in the cortical hemisphere infused with the antidepressants than that infused with SAL. These findings indicate that repeated mild stress and antidepressant treatments induce sprouting of NA LC axons in the cerebral cortex. Axonal sprouting of LC neurons can explain both the delayed onset of the clinical response to antidepressants and subsensitivity of P-adrenoceptors following repeated stress and antidepressant treatments, and may be a common mechanism for the clinical efficacy of antidepressant drugs and electroconvulsive shock. Furthermore, the findings suggest the possibility that axonal retraction or degeneration of central NA neurons may be involved, at least in part, in the pathology of clinical depression.
Key words: axonal sprouting, stress, antidepressant, cortex, locus coeruleus
Introduction The axons of noradrenergic (NA) locus coeruleus (LC) neurons are known to have a great capacity for axonal regeneration in response to brain dam-
age (Moore et al., 1974; Bjorklund and Stenevi, 1979). Although the plastic changes of NA LC axons occur preferentially in the developing brain, NA axonal regeneration has also been demonstrated in the mature brain (Moore et al., 1971;
588
Stenevi et al., 1972; Pickel et al., 1974; Bjorklund and Stenevi, 1979; Gage et al., 1983a,b; Sakaguchi et al., 1984; Haring et al., 1986). Most previous studies have been concerned with axonal regeneration following brain lesions. However, recent electrophysiological studies have suggested that the morphology of axon terminals arising from LC neurons dynamically changes in response to various environmental stimuli (Nakamura et aI., 1989; Sakaguchi and Nakamura, 1989, 1990; Nakamura and Sakaguchi, 1990). In this report, a plastic change of LC axon terminals following stress and antidepressant treatments will be demonstrated based on electrophysiological and morphological techniques. Both stress and antidepressants are known to cause a variety of neurochemical changes in brain NA system (Glowinski and Axelrod, 1964; Schildkraut, 1965; Schildkraut et al., 1970; Vetulani and Sulser, 1975; Banerjee et al., 1977; Tang et al., 1978; Kvetansky, 1980; Stone and McCarty, 1983; Glavin, 1985; Anisman and Zacharko, 1986). Of particular interest is the fact that some neurochemical changes, such as subsensitivity of padrenoceptors, become manifest only after two or three weeks from the start of these treatments (Vetulani and Sulser, 1975; Banerjee et al., 1977; Stone, 1983a; Stanford et al., 1984). Furthermore, the clinical response to antidepressants is well known to take two or three weeks. These findings indicate that repeated stress and antidepressant treatments cause some gradual alteration in central NA neurons. Taking account of a great capacity for axonal sprouting of LC neurons, it is conceivable that the gradual alteration caused by repeated stress and antidepressant treatments involves morphological changes of NA axon terminals in the brain. The first section of this report discusses electrophysiological evidence for axonal sprouting of LC neurons following repeated mild stress, and in the remaining sections, morphological evidence for the ability of antidepressants to induce axonal sprouting of central NA neurons in the mature brain is discussed.
Electrophysiological evidence for axonal sprouting of LC neurons following repeated mild stress
The most difficult problem in trying to confirm a plastic change in LC axon terminals is to quantify the density of NA LC axons in a given brain site. Although NA axon terminals in the brain can be visualized by appropriate histofluorescence or immunohistochemical methods, a quantitative comparison of the density of LC axon terminals in the brain, particularly among different animals, is difficult. To overcome this problem, an electrophysiological method has been used for quantifying the density of terminal axons of LC neurons (Nakamura et al., 1984a,b; Sakaguchi et al., 1984; Sakaguchi and Nakamura, 1987, 1989, 1990; Nakamura et al., 1989). The percentage of LC neurons activated antidromically from electrical stimulation of a given brain site, taking the total number of LC neurons recorded as 100, was considered as a physiological index for the density of LC axons innervating that brain site. Hereafter, this physiological index will be called “projection (P)-index.” In a previous experiment, the reliability of the “P-index” to assess the density of LC axons in the brain was tested by confirming a morphological study showing evidence for axonal sprouting of LC neurons following partial brain damage (Sakaguchi et al., 1984). Using the histofluorescence method, Stenevi et al. (1972) have demonstrated that NA axon terminals in the dorsal lateral geniculate nucleus (LGN) are increased after ablation of the visual cortex (VC). Since the NA axon terminals in the LGN are known to originate in the LC, the P-index of LC neurons for the LGN was expected to be higher in animals with VC ablation than in intact animals. This has been confirmed by electrophysiological experiments (Sakaguchi et al., 1984). To ascertain the occurrence of plastic changes of LC axons (axonal sprouting or retraction) following repeated stress, the P-index for the cerebral cortex was assessed in control and stressed rats. Male Sprague-Dawley rats (8-14 weeks of
589
age) were used. The animals were divided into two groups, control and stressed. For stress treatment, rats restrained in a small cage were immersed in warm water (36-37°C) up to the neck for 10 min daily (Nakamura et al., 1989). The animals received the stress treatment for either 1 or 2 weeks. Electrophysiological experiments were made under urethane anesthesia on the day following the termination of stress treatments. For electrical stimulation experiments, stimulating electrodes, consisting of two insulated stainless-steel wires (approximately diameter, 200 pm) with an exposed tip of approximately 0.5 mm, were implanted in the frontal cortex (FC) and occipital cortex (OC). Electrical stimulation consisted of pulses (1 ms duration) with currents ranging from 0.1 to 5 mA. The rate of stimulation was 1 Hz in all experiments. Single-unit activity of LC neurons was recorded extracellularly by a glass micropipette filled with 2 M NaCl. The location of the LC was determined by appearance of field responses evoked by stimulation of the dorsal NA bundle arising from the LC (Nakamura and Iwama, 1975; Nakamura, 1977). The singleunit activity of LC neurons was recorded superimposed upon the field response. When a single LC neuron was encountered, each stimulating site was examined for antidromic activation. Responses of LC neurons to cortical stimulation were considered to be antidromic if the responses revealed fixed latency, the ability to follow highfrequency stimulation ( > 200 Hz) and collision with spontaneous spikes. In each animal, 40 to 70 LC neurons were recorded to assess the P-index. The P-indices for the cortex in the control rats varied from animal to animal (Fig. 1). The mean P-indices (+S.E.) for the FC and OC were not different between the control (FC, 38.8 f 3.3; OC, 23.5 k2.9) and one week stress group (FC, 42.0 k 3.8; OC, 30.5 & 1.9), whereas those for the FC and OC were significantly higher in the two week stress group (FC, 50.8 2.9; OC, 33.5 k 3.0) than the control group (Newman Keuls test, P < 0.05). The P-indices for the cortex in the two week stress group did not differ from those in the one
+
70
frontal cortex
60
50
occipital cortex
8
8 4 -
00
0
00
0
~
0 .-C
40 +
O
0
0 0
c:B
0
20
0
I0 0
control I-week 2-week
stress
stress
&o[
1-week 2-week stress stress
Fig. 1. Electrophysiological evidence for terminal sprouting of locus coeruleus (LC) axons following repeated mild stress. The percentage of LC neurons activated antidromically (projection-index, P-index) from the frontal cortex (FC) and occipital cortex (OC) was used as a physiological index for the density of LC axons projecting to the FC and OC. In stress groups, rats restrained in a small cage were immersed in warm water for 10 min daily. The mean P-indices for the FC and OC in the two week stress groups were significantly higher than those in the control group (Newman Keuls test, P < 0.05). The P-indices for the FC and OC did not differ between the control and one week stress groups. Bar indicates the mean value of P-index. (From Nakamura et al., 1989.)
week stress group. These results indicated that the density of the projection from the LC to the cerebral cortex increased gradually, only reaching statistical significance after two weeks of stress. To see if the observed change in the P-index was due to a change in the excitability of terminal axons of LC neurons, threshold currents for antidromic activation, which were defined as currents sufficient to elicit antidromic responses on 100% of the non-collision trials, were measured. The threshold currents for antidromic activation from the FC were not different between the control (mean k S.E., 1.7 k 0.1 mA) and stress groups (1 week stress, 1.9 i-0.1 mA; 2 week stress, 1.5 0.1 mA) (Newman-Keuls test, P > 0.05). In addition, in several hundred cases of antidromically activated LC neurons reported previously, sufficient currents have always produced 100% antidromic response without impulse conduction failure (Nakamura et al., 1981, 1982a,b; Ryan et al., 1985). Thus, the changes in the P-index for
590
the cortex following repeated mild stress treatment are not attributable to a change in the terminal excitability nor differences in the safety factor for impulse conduction along the axon. Therefore, the increase in the P-index for the cortex is interpreted as morphological (axonal sprouting) rather than physiological. Functional significance of stress-induced sprouting of noradrenergic LC axons
The changes in the P-index following repeated mild stress, strongly suggest that LC axons in the cerebral cortex cause sprouting in response to stressful stimuli. The question then arises as to the functional significance of axonal sprouting of LC neurons caused by repeated stress treatment. Most animals that repeatedly confront stressful stimuli can adapt themselves to stress that is not extremely severe. Although the mechanism underlying behavioral adaptation to stress is not clear at present, it is likely that axonal sprouting of NA LC neurons is involved in the stress-adaptation mechanism. Stone (1983a) has proposed the hypothesis that subsensitivity of P-adrenoceptors underlies the mechanism of adaptation to stress. This view is based mainly on the similarity of the time course between the occurrence of P-adrenoceptor subsensitivity and the adaptation of animals to stress. It is well known that when the number of brain NA axons is reduced using the specific catecholamine neurotoxin 6-hydroxydopamine (6-OHDA), the number of brain Padrenoceptors, which are located on postsynaptic membranes, is increased due to reduced NA input to postsynaptic cells (denervation supersensitivity) (Sporn et al., 1976). Therefore, in repeatedly stressed animals, in which the number of brain P-adrenoceptors is reduced (Bergstrom and Keller, 1979; Pandy et al., 1979; Stanford et al., 19841, a situation opposite to that observed in 6-OHDA-treated animals may occur. In chronically stressed rats, axonal sprouting of NA LC neurons may cause enhanced NA input to target cells, thereby resulting in the decreased density of
postsynaptic P-adrenoceptors. Therefore, it is very likely that P-adrenoceptor subsensitivity and axonal sprouting of NA LC neurons occur simultaneously while animals become adapted to stressful stimuli. The effects of antidepressants upon axonal regeneration of noradrenergic LC neurons
Although repeated administration of antidepressants exerts a therapeutic effect on clinical depression, the mechanism of the clinical efficacy of antidepressants remains unknown. Another effective treatment for depression is repeated administration of electroconvulsive shock stress. This treatment is also known to cause a reduction in the sensitivity of brain P-adrenoceptors (Bergstrom and Keller, 1979; Pandy et al., 1979). Therefore, assuming that the clinical response to antidepressants involves a common mechanism with electroconvulsive shocks, it was likely that repeated administration of antidepressants, like repeated stress treatment, could cause axonal sprouting of NA LC neurons. This possibility has been examined directly using the histofluorescence method (Nakamura, 1990). First, the antidepressant maprotiline (MPL) was continuously infused directly into the FC of one side by means of an osmotic minipump (Alzet, model ZOOZ), while the corresponding cortical site of the other side was infused with saline (SAL). Female Sprague-Dawley rats, 8-16 weeks of age, were used. Under chloral hydrate anesthesia (350 mg/kg, ip), the infusion cannula (30 gauge), connected to osmotic minipumps with polyethylene tubing, was implanted into the cartical site and fixed to the skull with dental cement and cyanoacrylate glue. The osmotic minipumps were placed subcutaneously in the neck of the animals. Infusions of the antidepressant and SAL were made at a rate of 0.5 p l / h for more than two weeks (14-18 days). After more than two weeks from the start of drug infusion, the density of glyoxylic acid-induced fluorescent catecholamine fibers in the
591
S A L - injection slde
(2W)
-
MPL
- injection side (IrnM, 2W)
0.1rnrn
0.1rnrn
Fig. 2. The effects of direct infusions of antidepressants upon sprouting of catecholaminergic axons in the cerebral cortex. The cortical site of one hemisphere was infused with 1 mM maprotiline (MPL) and the corresponding site of the other hemisphere with saline (SAL). The drugs were continuously delivered from the tip of infusion cannulas (30G) connected to osmotic minipumps (Alzet, model 2002) with polyethylene tubing. Two weeks after the infusion, the density of glyoxylic acid-induced fluorescent fibers was compared between the two hemispheres. The number of fluorescent fibers in the cortex surrounding the infusion sites of both hemispheres was reduced as compared with the cortex of intact animals, due to chronic implantation of the infusion cannulas. However, the density of cortical fluorescent fibers was apparently higher in the MPL- than SAL-infused cortex. Left, medial cortex. Right, lateral cortex.
cortex was compared between the MPL- and SAL-infused cortex. The density of catecholamine fibers appeared higher in the MPLthan SAL-infused cortex (Fig. 2). However, the number of catecholamine fibers in the hemispheres infused with the antidepressant and SAL was decreased as compared with the cortex of intact animals free from infusions. The reduced density of catecholamine fibers in the infused cortex could be attributed to the degeneration of catecholamine fibers caused by nonspecific lesions due to chronic implantation of infusion cannulas. In spite of the observed effect of the antidepressant on the catecholamine fiber density it is not known whether the antidepressant in-
duced sprouting of catecholamine fibers or prevented catecholamine fibers from degenerating. To test the ability of antidepressants to induce sprouting of catecholamine fibers in the cortex, symmetrical cortical sites of two hemispheres were pretreated with 6-OHDA (2 pg/O.5 p l ) (Nakamura, 1990). One or two weeks after the 6-OHDA pretreatment, antidepressants (1 mM) and SAL were continuously infused in the cortex in the same way as described above. The same cortical site of one hemisphere as that pretreated with 6-OHDA was infused with antidepressants and the corresponding site of the other side with SAL. Infusions of the antidepressants and SAL were made at a rate of 0.5 p l / h either for less
592
than one week (5-7 days) or for more than two weeks (14-18 days). The density and distribution of glyoxylic acid-induced fluorescent catecholamine fibers were compared between the SAL- and antidepressant-infused cortices. In addition to MPL, desipramine (DMI) and -mianserin (MIA) were tested. The distribution of fluorescent catecholamine fibers revealed a marked difference between the control hemisphere and that infused with the antidepressants for more than two weeks (Figs. 3,4, and 5). In the SAL-infused hemisphere, the cortex from the inA
fusion site to the caudal cortex remained almost completely denervated of catecholamine fibers. In contrast, in the hemisphere infused with one of the above antidepressants, the denervation of catecholamine fibers was restricted to a relatively small region, and a number of catecholamine fibers appeared particularly in the superficial layer of the dorsal cortex. The asymmetrical distribution of cortical catecholamine fibers between the two hemispheres was most apparent in the region 1-3 mm caudal to the infusion site (upper in Fig. 41, and became less evident in the cortex
S A L - Infused MPL- Infused
01
18 d a y s
0.5 mm
B1
Fig. 3. The effect of MPL upon catecholamine fiber regeneration in the cerebral cortex denervated by pretreatment with 6-hydroxydopamine (6-OHDA). A. A drawing of fluorescent catecholamine terminals in the superficial layer of the cerebral cortex. This was reproduced from the coronal section obtained from the brain region approximately 2.5 mm caudal to the infusion site. Two weeks after local infusion of 6-OHDA (2 pg/0.5 pI) into symmetrical sites in the FC in both hemispheres, 1 mM MPL and SAL were infused into the same sites as the 6-OHDA infusion with osmotic minipumps for 18 days. In the SAL-infused side, a large cortical region remained denervated. In contrast, in the corresponding cortical region of the MPL-infused side, a number of fluorescent catecholamine fibers regenerated. B. Photomicrographs of glyoxylic acid-induced fluorescent catecholamine fibers in the cortical regions indicated by the arrows in A. B1, SAL-infused side. B2, MPL-infused side. (From Nakamura, 1990.)
593 DMI - i n f u s e d S A L - infused
SAL
18 days <3mm
DMI - i n f u s e d
- Infused
Fig. 4.The effect of desipramine (DMI) (1 mM) upon cortical catecholamine fiber regeneration. The method was the same as that described in Figure 3. DMI and SAL were infused for 18 days in this experiment. Upper, a drawing of fluorescent catecholamine terminals reproduced from a coronal section obtained from the brain region less than 3 mm caudal to the infusion site. Lower, this drawing was reproduced from the brain region more than 5 mm caudal to the infusion site. Note that the asymmetrical distribution of cortical catecholamine fibers between the DMI- and SAL-infused hemisphere was apparent in the rostra1 but not in the caudal cortex.
M I A - Infused
S A L - Infused
17 days
05mm
Fig. 5. The effect of 1 mM mianserin (MIA) upon cortical catecholamine fiber regeneration. In this experiment, MIA and SAL were infused for 17 days.
6 - O H D A - Infused
LLrJ
6 - OHDA- inf used
9 weeks
0.5rnm
Fig. 6. The denervation of catecholamine fibers in the cerebral cortex infused with 6-OHDA without additional treatment. Nine weeks after 6-OHDA (2 pg/0.5 pl) was infused into symmetrical sites of the frontal cortex of the two hemispheres, the denervation of catecholamine fibers was observed to nearly the same extent in the two hemispheres.
594
more than 5 mm caudal to' the infusion site (lower in Fig. 4). In animals which received infusions of MPL for less than one week, the asymmetrical distribution of catecholamine fibers between the antidepressant- and SAL-infused hemisphere was not apparent. In addition, when the two hemispheres received 6-OHDA infusion without additional treatment, the denervation of catecholamine fibers was observed to nearly the same extent between the two hemispheres (Fig. 6). To quantitatively assess the effects of the antidepressants upon the regeneration of catecholamine fibers in the cerebral cortex pretreated
6-OHDA 3 - 9 W (1.9 f 0.3) (left)
.-
anlma'
c
1 2 3 4 5
MPL 1 W (1.9 f 0.3)
MPL 2 W (0.7 f 0.2)
6 7 8 9
1011121314
SAL 1 W (1.8 f 0.3)
SAL 2 W (1.8 f 0.3)
with 6-OHDA, the mediolateral length of the denervated cortex for each animal was measured on a coronal section from a cortical region approximately 1 mm caudal to the end of the nonspecific lesion (Fig. 7). In all animals that received infusions of the antidepressants for more than two weeks, the extent of denervation was smaller in the antidepressant- than SAL-infused cortex. The cortical denervation size following 6-OHDA infusion, without additional treatment, was not different between the two hemispheres, and was almost the same as that of the cortex infused with MPL or SAL for less than one week, or with SAL for more than two weeks.
DMI 2 W (0.5 f 0.2)
MIA 2 W (0.6 f 0.2)
15 16 1 7 1 8
19 20 21 2223 2425
SAL 2 W (2.2 f 0.3)
SAL 2 W (2.1 f 0.3)
4 '
6-OHDA 3 - 9 W ( 2 . l f 0.3)
(right)
Fig. 7. The extent of the catecholamine fiber denervation in the cerebral cortex. The extent of denervation was calculated by measuring the mediolateral length of the cortical region where no fluorescent fibers appeared, although in some cases a few fibers with short branches occurred sporadically in the denemation region. The medial and lateral edge of denetvation region were determined by the appearance of fluorescent fibers longer than 100 prn. The values in the parentheses indicate the mean mediolateral length of denervation in the hemisphere following each treatment. The mean of denemation size between the two hemispheres was significantly different in animals infused with the antidepressants for more than two weeks (r-test, P < 0.01). The denervation size of the cortex infused with MPL, DMI, or MIA for more than two weeks was significantly smaller than that of the cortex with different treatments (6-OHDA for 3-9 weeks, SAL for less than one week or more than two weeks, and MPL for less than one week) (Newman-Keuls test, P < 0.05). All animals except #13, #17, #22, #23 and #24 received infusions of the antidepressants into the left hemisphere. The denervation size of the antidepressant-infused cortex in #11, #16, #19 and #20 was measured as 0. (From Nakamura, 1990.)
595
The functional implication of antidepressant-induced axonal sprouting
Repeated administration of antidepressants is known to cause P-adrenoceptor subsensitivity (Schildkraut et al., 1970; Banerjee et al., 1977; Sulser, 1979; Maj et al., 1984). As discussed for the stress-related change of LC axons, axonal sprouting of NA LC neurons in the cerebral cortex is considered to be related to a decrease in the number of cortical P-adrenoceptors due to an increased NA input to the target cells. Thus, the antidcpressant-induced p-adrenoceptor subsensitivity and NA axon sprouting remains undetermined. From electrophysiological evidence for the stress-induced sprouting of LC axons, electroconvulsive shock stress, which also results in the subsensitivity of P-adrenoceptors (Bergstrom and Keller, 1979; Pandy et al., 1979), is considered to cause axonal sprouting of NA neurons. Thus, sprouting of NA LC axons can account for the occurrence of P-adrenoceptor subsensitivity following both repeated stress and antidepressant treatments, and may be involved in the mechanism of clinical efficacy of both electroconvulsive shock and antidepressant treatments. In animals treated chronically with stress and antidepressants, the influence of NA LC neurons is considered to become more extended spatially, as their terminal fields increase in size due to axonal sprouting. Target cells primarily innervated by NA axons may become more strongly affected by the newly formed NA terminals, while those primarily devoid of NA input before the occurrence of axonal sprouting may become innervated by new terminals of NA neurons. Therefore, even though the target cells reduce the sensitivity of P-adrenoceptors located on their postsynaptic membranes, the NA influence upon the target cells would become substantially enhanced by spatially extended NA terminals. In the present experiments, the antidepressants DMI and MPL, which are known to be potent reuptake blockers of norepinephrine (NE), were found to induce axonal sprouting of central
NA neurons in the cerebral cortex. In addition to DMI and MPL, the atypical antidepressant MIA, which is not a reuptake blocker of NE, also had t h e ability to induce sprouting of NA axons. Although further experiments are needed to see if most, or possibly all, antidepressants can induce sprouting of NA axons, the finding suggests the possibility that a morphological change (axonal retraction or degeneration) is involved in the pathology of clinical depression. Recently, immunohistochemical studies on the human brain have demonstrated that LC cell loss is greater in the brains of Alzheimer’s and Parkinson’s patients with depressive symptoms, as well as a chronically depressed patient, than in those of patients without depression (Chan-Palay and Asan, 1989a,b). This finding, though not directly implying an involvement of axonal degeneration of LC neurons in the cause of depressive symptoms, is consistent with the view that the pathology of clinical depression is related to axonal retraction or degeneration of central NA neurons. Stress is known to be involved in the etiology of depression (Stone, 1983b), whereas electroconvulsive shock stress is an effective treatment of depression. Therefore, it seems likely that stress is, on one hand, a cause of depression and, on the other hand, it has a therapeutic effect on the disease. Our electrophysiological study suggested that restraint stress induced dual effects, either axonal sprouting or retraction, depending on the duration of stress treatments (Sakaguchi and Nakamura, 1989, 1990): the P-indices for the cerebral cortex increased in animals restrained for one h daily for two weeks, whereas those for the cortex decreased in animals restrained for 6 h daily for two weeks. The duration-dependent effects of stress upon LC projections may contribute, at least in part, to stress-related behavioral changes of animals. The terminal sprouting of LC neurons following the short-duration stress may be related to adaptation to stress, while the terminal retraction following the long-duration stress may have relevance to maladaptation of
596
animals to stress (possibly depression). Further experiments are necessary to confirm this hypothesis based on some reliable techniques that can distinguish adaptation from maladaptation to stress. The mechanisms for axonal sprouting of central noradrenergic neurons
The mechanism of the stress- and antidepressant-induced sprouting of NA LC axons remains unknown. Regarding stress, it is likely that a stress-induced increase in the release of some endogenous neurochemical may be involved. NE released from the axon terminals of LC neurons in response to environmental stress (Korf et al., 1973; Abercrombie and Jacobs, 1987) is one of the candidates for the stress-induced changes. Since it is known that NE has both neurotrophic and neurotoxic action (Felten et al., 1982; Kasamatsu, 1983; Rosenberg, 19881, the dual effects of stress may be explained through mediation of NE. Corticosterone, whose blood concentration is increased in response to stress, is another potential candidate that could mediate the dual effects of repeated stress treatment. This hormone has also been reported to have neurotrophic and neurotoxic action upon brain neurons (Sapolsky et al., 1985; Sloviter et al., 1989). Although there are no available data for elucidating the mechanism of the antidepressant-induced sprouting of NA axons, it is thought to involve a molecular process in common with the mechanism of the stress-induced sprouting. Therefore, studies on the molecular processes of the neurotrophic and neurotoxic actions of N E and corticosterone may be useful in understanding the mechanism of the antidepressant-induced sprouting of NA neurons and the cause of clinical depression. Acknowledgements
I am grateful to Dr. C. Yamamoto for valuable comments on the manuscript, Mr. T. Hiramatsu
for making the figures and Ms. C. Mitani for typing the manuscript. I owe many thanks to Dr. C.D. Barnes for his editing the manuscript and improving the English. This research was partly supported by a grant provided by the Ichiro Kanehara Foundation. References Abercrombie, E.D. and Jacobs, B.L. (1987) Single-unit response of noradrenergic neurons in the locus coeruleus of freely moving cats. 1. Acutely presented stressful and nonstressful stimuli. J. Neurosci., 7: 2837-2843. Anisman, H. and Zacharko, R.M. (1986) Behavioral and neurochemical consequences associated with stressors. Ann. N.Y. Acad. Sci., 467: 205-225. Banerjee, S.P., Kung, L.S., Riggi, S.J. and Chanda, S.K. (1977) Development of p-adrenergic receptor subsensitivity by antidepressants. Nature (London), 268: 455-457. Bergstrom, D.A. and Keller, K.J. (1979) Effect of electroconvulsive shock on monoaminergic receptor binding sites in rat brain. Nuture (London), 278: 464-466. Bjorklund, A. and Stenevi, U. (1979) Regeneration of monoaminergic and cholinergic neurons in the mammalian central nervous system. Physiot. Ref,., SO: 62-100. Chan-Palay, V. and Asan, E. (198Ya) Quantitation of catecholamine neurons in the locus coeruleus in human brains of normal young and older adults and in depression. J. Cornp. NeuroL, 287: 357-372. Chan-Palay, V. and Asan, E. (1989b) Alterations in catecholamine neurons of the locus coeruleus in senile dementia of the Alzheimer type and Parkinson's disease with and without dementia and depression. J. Cornp. Neurol., 287: 373-392. Felten, D.L., Hallman, H. and Jonsson, G. (1982) Evidence for a neurotrophic role of noradrenaline neurons in the postnatal development of rat cerebral cortex. J. Neurocytol., 11: 119-135. Gage, F.H., Bjorklund, A. and Stenevi, U. (1983a) Reinnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and noradrenergic afferents. Bruin Re.$., 268: 27-37. Gage, F.H., Bjorklund, A,, Stenevi, U. and Dunnett, S.B. (1983b) Functional correlates of compensatory collateral sprouting by aminergic and cholinergic afferents in the hippocampal formation. Bruin Res., 268: 39-47. Glavin, G.B. (1985) Stress and brain noradrenaline: A review. Neurosci. Biobehur. Rei,., 9: 233-243. Glowinski, J. and Axelrod, J. (1964) Inhibition of uptake of tritiated-noradrenaline in the intact rat brain by imipramine and structurally related compounds. Nuture (London), 204: 1318- 1319. Haring, J.H., Miller, G.D. and Davis, J.N. (1986) Changes in the noradrenergic innervation of the area dentata after axotomy of coeruleohippocampal projections or unilateral lesion of the locus coeruleus. Brain Res., 368: 233-238.
597 Kasamatsu, T. (1983) Neuronal plasticity maintained by the central norepinephrine system in the cat visual cortex. In J.J.M. Sprague and A.N. Epstein (Eds.), Progress in Psychobiology und Physivlogicul Psychology, Vol. 10, Academic Press, New York, pp. 1-112. Korf, J., Aghajanian, G.K. and Roth, R.H. (1973) Increased turnover of norepinephrine in the rat cerebral cortex during stress: Role of the locus coeruleus. Neurophurmacolom. 12: 933-938. Kvetansky, R . (1980) Recent progress in catecholamines under stress. In E. Usdin, R. Kvetansky and I.J. Kopin (Eds.), Catecholamine and Stress: Recent AdL)ances, Elsevier, Amsterdam, pp. 7-18. Maj, J., Przegaluntki, E. and Mogilnicka, E. (1Y84) Hypotheses concerning the mechanism of action of antidepressant drugs. Rec. Physiol. Biochem. Phurmucol., 100: 1-74. Moore. R.Y., Bjorklund, A. and Stenevi, U. (1971) Plastic changes in the adrenergic innervation of rat septa1 area in response to denervation. Brain Rex, 33: 13-35. Moore, R.Y., Bjorklund, A. and Stenevi, U. (1974) Growth and plasticity of adrenergic neurons. In F.O. Schmitt and F.G. Worden (Eds.), The Neuroscience Third Study Program, MIT Press, Cambridge, MA, pp. 961-977. Nakamura, S. (1977) Some electrophysiological properties of neurons of rat locus coeruleus. J. Physiol. (London), 267: 641-658. Nakamura, S. (1990) Antidepressants induce regeneration of catecholaminergic axon terminals in the rat cerebral cortex. Neurosci. Lett., 111: 64-68. Nakamura, S. and Iwama, K. (1975) Antidrornic activation of the rat locus coeruleus neurons from hippocampus, cerebral and cerebellar cortices. Bruin Rex, 99: 372-376. Nakamura, S. and Sakaguchi, T. (1990) Development and plasticity of the locus coeruleus: A review of recent physiological and pharmacological experimentation. Prog. Neurohid., 34: 505-526. Nakamura, S., Tepper, J.M., Young, S.J. and Groves, P.M. (1981) Neurophysiological consequences of presynaptic receptor activation: Changes in noradrenergic terminal excitability. Brain Res., 226: 115-170. Nakamura, S., Tepper, J.M., Young, S.J. and Groves, P.M. (1982a) Noradrenergic terminal excitability: Effects of opioids. Neurosci. Lett., 30: 57-62. Nakamura, S., Tepper, J.M., Young, S.J. and Groves, P.M. (1982b) Changes in noradrenergic terminal excitability induced by amphetamine and their relation to impulse traffic. Neuroscience, 7: 2217-2224. Nakamura, S., Shirokawa, T. and Sakaguchi, T. (1984a) Increased projection from the locus coeruleus to the lateral geniculate nucleus and visual cortex in young adult rats following unilateral enucleation. Neurosci. Lett., 49: 77-80. Nakamura, S., Shirokawa, T. and Sakaguchi, T. (1984b) Increased adrenergic projection from the locus coeruleus to the lateral geniculate nucleus of rats following one-eye removal at birth. Der:. Brain Rex, 15: 283-285. Nakamura, S., Sakaguchi, T. and Aoki, F. (1989) Electrophysiological evidence for terminal sprouting of locus coeruleus
neurons following repeated mild stress. Neurosci. Lett., 100: 147-152. Pandy, G.G., Heinze, W.J., Brown, B.D. and Davis, J.H. (1979) Electroconvulsive shock treatment decreases padrenergic receptor sensitivity in rat brain. Nufure (London), 280: 234-235. Pickel, V.M., Segal, M. and Bloom, F.E. (1974) Axonal proliferation following lesions of cerebellar peduncles. A combined fluorescence microscopic and autoradiographic study. J. Comp. Neurol., 155: 43-60. Rosenberg P.A. (1988) Catecholamine toxicity in cerebral cortex in dissociated cell culture. J. Neurosci., 8: 28872894. Ryan, L.J., Tepper, J.M., Sawyer, S.F., Young,S.J. and Groves, P.M. (1 985) Autoreceptor activation in central monoamine neurons: Modulation of neurotransmitter release is not mediated by intermittent axonal conduction. Neuroscience, 15: 925-931. Sakaguchi, T. and Nakamura, S. (1987) The mode of projections of single locus coeruleus neurons to the cerebral cortex in rats. Neuroscience, 20: 221-230. Sakaguchi, T. and Nakamura, S. (1989) Remodeling of noradrenergic innervation in rat cerebral cortex induced by repeated stress. Soc. Neurosci. Ahstr., 15: 781. Sakaguchi, T. and Nakamura, S. (1990) Duration-dependent effects of repeated restraint stress on cortical projections of locus coeruleus neurons. Neurosci. Lett., 118: 193-196. Sakaguchi, T., Shirokawa, T. and Nakamura, S. (1984) Changes in projection from locus coeruleus to lateral geniculate nucleus following ablation of visual cortex in adult rats. Bruin Res., 312: 319-322. Sapolsky, R.M., Krey, L.C. and McEwen, B.S. (1985) Prolonged glucocorticoid exposure reduces hippocampal neuron number: Implication for aging. J. Neurosci. 5: 12221227. Schildkraut, J.J. (1965) The catecholamine hypothesis of affective disorders: A review of supporting evidence. Am. J. Psychiutry., 122: 509-522. Schildkraut, J.J., Winokur, A. and Applegate, C.W. (1970) Norepinephrine turnover and metabolism in rat brain after long-term administration of imipramine. Science, 168: 867-869. Sloviter, R.S., Valiquette, G., Abrams, G.M., Ronk, E.C., Solias, A.L., Paul, L.A. and Neubort, S. (1989) Selective loss of hippocampal granule cells in the mature rat brain after adrenalectomy. Science, 234: 535-538. Sporn, J.R., Harden, T.K., Wolfe, B.B. and Molinoff, P.B. (1976) P-Adrenergic receptor involvement in 6-hydroxydopamine-induced supersensitivity in rat cerebral cortex. Science, 194: 624-625. Stanford, C., Fillenx, M. and Ellen, R. (1984) The effect of repeated mild stress on cerebral cortical adrenoceptors and noradrenaline synthesis in the rat. Neurosci. Lett., 45: 163-1 67. Stenevi, U., Bjorklund, A. and Moore, R.Y. (1972) Growth of intact adrenergic axons in the denervated lateral geniculate body. Exp. Neurol., 35: 290-299.
598 Stone, E.A. (1983a) Adaptation to stress and brain noradrenergic receptors. Neurosci. Biobehuc3. Rctr., 7: 503-509. Stone. E.A. (198%) Problems with current catecholamine hypotheses of antidepressant agents. Behut,. Brain Sci., 6: 535-577. Stone, E.A. and McCarty, R. (1983) Adaptation to stress: Tyrosine hydroxylase activity'and catecholamine release. Neurosci. Biohehai'. Re[,.,7: 29-34. Sulser, F. (1979) New perspectives on the mode of action of antidepressant drugs. TIPS, 1: 92-94.
Tang, S.W., Helmeste, D.M. and Stancer, H.C. (1978) The effects of acute and chronic desipramine and amitriptyline Nuunyn o n rat brain total 3-methoxy-4-hydrophenylglycol. Schmeidebergs Arch. Phurniacol., 305: 207-21 1. Vetulani, J. and Sulser, F. (1975) Action of various antidepressant treatments reduces reactivity of noradrenergic cyclic AMP-generating system in limic forebrain. Nurure (London), 257: 495-496.
C.D. Barnes and 0. Pompeiano (Ed\.) Prove.\\ 111 Brain Research, Vol. X X I‘ l Y 9 l Elsevier Science Publishers B.V.
599
CHAPTER 42
Adrenergic regulation of visuocortical plasticity: a role of the locus coeruleus system T. Kasamatsu Smith-Kettlewell Eye Research Institute, Webster Street, San Francisco, CA, U.S.A
Noradrenaiine-P-adrenoceptor-mediated neural plasticity in cat visual cortex exemplifies clearly established roles of the locus coeruleus system in brain function. The prime role of the noradrenaline-P-adrenoceptor system in the regulation of ocular dominance plasticity is discussed in this chapter and includes a newly invented paradigm of ocular dominance changes under anesthesia and paralysis without benefit of visual attention. Based on our recent findings, we have sought to integrate positive contributions of muscarinic cholinergic
receptors to the P-adrenoceptor-mediated regulatory processes. The issue of “activity dependency” is important and we recognize the necessity of designing new studies in which relationships between activity dependency within the visual pathway and global neurochemical/cellular factors can be tested directly. Further, we critically reviewed the involvement of y-aminobutyric acid, type receptors and N-methyl-Daspartate receptors in the regulation of ocular dominance plasticity.
Key words: P-adrenoceptor, muscarinic cholinergic receptor, GABA, receptor, N-methyl-o-aspartate receptor, second messengers, activity dependency, trophic effects, non-lemniscal afferents, neurotransmitter release, ocular dominance plasticity
Introduction The noradrenaline (NA)-containing nerve terminals originate from a relatively small number of NAergic cells in the locus coeruleus (LC) in the brainstem and extend their innervation widely throughout the brain including the neocortex. This long-distance monosynaptic innervation is carried mostly by slow-conducting unmyelinated or thinly myelinated axons. Catecholamine (CAI histochemistry, introduced by the Swedish school in the early 60s, made it possible to identify each neuron based on its chemical signature or transmitter phenotype. The recent advent in specific
immunological probes has aided the current explosion of our knowledge about brain organization based on chemical identity of various cell types. Yet, the NA system seems to remain among others such as the mesocortical dopamine (DA), pontine serotonin (5HT) and basal forebrain acetylcholine (ACh) projection systems one of the best characterized non-lemniscal corticopetal fiber projections within the brain. The question arises; why do we need these long-distance ascending projection fibers to the neocortex in addition to those belonging to the specific lemniscal system? Various answers have been suggested (see other chapters). This review
600
presents “neural plasticity in the visual cortex of kittens and cats” as a model for assigning a clearly established role to the LC. Our strategy in this inquiry is as follows: we take advantage both of the accumulated knowledge about single-cell physiology and connectivity among functionally characterized cells in cat visual cortex and the equally well described properties of the NA system within the visual cortex. Next, we study, physiologically and behaviorally, the neural consequences of pharmacological manipulations of the NA system in the brain. Basically we look for “modulatory” roles of the NA system in the developing, as well as mature, visual cortex. Our study is so designed because: (1) the six cortical laminae in cat visual cortex are abundant with NA terminals (Itakura et al., 19811, (2) the corresponding laminar distribution of /3-adrenoceptors (Shaw et al., 1984; Nakai et al., 1987) and their normal ontogeny are known (Jonsson and Kasamatsu, 1983; Shaw et al., 19841, (3) the “modulatory” effects of iontophoretically applied NA on excitability of visually activated single cells have been shown (Kasamatsu and Heggelund, 1982; Videen et al., 19841, and (4) the ionic mechanisms of N A s action in the brain has been basically worked out (Madison and Nicoll, 1986; McCormick and Prince, 1988). What we need, therefore, is a hypothesis on the function of the LC to be rigorously tested.
Visual cortical plasticity: hypothesis and tests Developmentally regulated neuronal plasticity has been one of the central concerns among modern brain scientists including Ramon-y-Cajal. Using the binocular cells in immature visual cortex of kittens as a model, a paradigm of monocular deprivation was devised in the early 60s to study the “experience dependency” of developmental plasticity (Wiesel and Hubel, 1963). They found that when a kitten is subjected to monocular vision for even a few days, many binocular cells in the visual cortex lose then-functioning connections with the lid-sutured eye, the majority of
cells thus becoming excitable only by stimulation from the “experienced” eye (ocular dominance plasticity). If monocular vision lasts beyond a certain time limit during the early postnatal days (ie., postnatal susceptible period) monocularly deprived kittens become behaviorally “blind” when tested with the “inexperienced” eye. Effects of monocular deprivation are easy to reproduce, and their physiological consequences, such as a decrease in binocularity and an eventual shift in ocular dominance toward the open eye, seem to give us reliable indices to measure the extent of ocular dominance plasticity. Under certain conditions, structural changes in the organization of the geniculocortical projection have been shown; i.e., shrinkage of the width of ocular dominance columns in layer IV of the visual cortex (Hubel et al., 1977; Shatz and Stryker, 1978) and a decrease in the size of somata in geniculate laminae which correspond to the monocularly deprived eye (Guillery, 1972; 1973). The development and plasticity of the mammalian visual system have been reviewed extensively (Movshon and Van Sluyters, 1981; Sherman and Spear, 1982; Wiesel 1982; Kasamatsu 1982; 1983; 1987; FrCgnac and Imbert, 1984; Singer 1984; Kasamatsu et al., 1984; Hirsch and Tieman, 1987; Cooper 1987; Frhgnac 1987; Bear et al., 1987; Rauschecker 1987; Stryker 1989). Some years ago we proposed that the CA-containing system in the kitten brain, especially NAergic, may critically contribute to the regulation of ocular dominance plasticity (Kasamatsu and Pettigrew, 1976; 1979; Pettigrew and Kasamatsu, 1978; Kasamatsu et al., 1979, 1981b; Kasamatsu 1982, 1983; Kasamatsu et al., 1984; Kuppermann and Kasamatsu, 1984). Our early findings are summarized here: (1) the plasticity in kitten visual cortex was not expressed when the release of endogenous NA was maximally and continuously suppressed by destruction of NA terminals with a CA-related neurotoxin 6-hydroxydopamine (6-OHDA) (Kostrzewa and Jacobowitz, 1974; Jonsson, 19801, (2) a direct infusion of NA into the aplastic cortex restored the
60 I
plasticity, thus causing a shift in ocular dominance following brief monocular deprivation, and (3) cortical recovery from the effects of monocular deprivation was suppressed by a direct infusion of 6-OHDA, while enhanced by a NA infusion. The last finding strongly suggests that NA availability increases the modifiability of ocular dominance regardless of the direction of changes which are primarily dictated by the experimental design. Early investigators obtained physiological evidence of neural plasticity in kitten visual cortex in which the endogenous content of NA had been significantly reduced by various means (Bear and Daniels, 1983; Bear et al., 1983; Daw et al., 1984; 1985a,b,c; Adrien et al., 1985; Trombley et al., 1986). We have discussed elsewhere conceptual issues, as well as technical problems, which were innate to the design of these studies and which thus muddled the proper interpretation of their findings (Kasamatsu, 1982; 1983; 1987; 1989; Kasamatsu et al., 1984; Kasamatsu and Shirokawa, 1988). In fact, there is little disagreement between results obtained by others (Bear et al., 1983; Daw et al., 1983; Paradiso et al., 1983) and ourselves (Kasamatsu et al., 1979; 1981b) insofar as the plasticity-suppressing effects of cortically infused 6-OHDA is concerned. However, there were significant differences in their interpretations. One group suggested that “non-specific” effects of 6-OHDA, rather than its neurotoxic effects on CA terminals, might account for our original findings of reduced plasticity in the 6OHDA-treated cortex (Daw et al., 1984; 1985abc; Trombley et al., 1986; Gordon et al., 1988). This argument’s most serious shortcoming is failure to consider the “dynamic” nature of the NA system itself when NA was not totally depleted. They assumed significant depletion of endogenous NA, say down to 25% of control, to be an assurance of an equally significant decrease in physiological and behavioral consequences of the decrease in endogenous NA. There were also some comments on reliability of biochemical assays in their study. In subsequent studies we showed that com-
pensation for the NA depletion indeed took place within the NA system either through an increase in the density of NA-related receptor binding (up-regulation) or through reinnervation by NA terminals of the 6-OHDA-infused visual cortex (regenerative NA fibers; Nakai et al., 1987; Nakai, 1987). These results are consistent with compensatory changes found either in remaining NA neurons in neocortex of rats injected with 6OHDA neonatally (Jonsson et at., 1979) or in DA neurons after lesions with 6-OHDA in the nigrostriatal system (Stricker and Zigmond, 1986). It was also shown recently that the endogenous NA content in hippocampal dialysates did not decrease due to a lesion with 6-OHDA unless the tissue content of NA became less than 50% of control (Abercrombie and Zigmond, 1989). Another school suggested that the compensatory process for the loss of NA terminals from the 6-OHDA-infused cortex takes place outside the NA system. This matter will be dealt with in detail below, when we explore the involvement of various receptors in cortical plasticity beside the NA-activated P-adrenoceptors. P-Adrenoceptor as a key step
Our early work using physiological (Kasamatsu 1979; 1980) and biochemical methods (Jonsson and Kasamatsu, 1983) strongly suggested the involvement of P-adrenoceptors, in the NA-mediated enhancement of visuocortical plasticity in kitten visual cortex. Especially, the latter study indicated that the normal ontogeny of P-adrenoceptors, rather than that of endogenous NA, might be a better indicator of the profile of postnatal susceptibility to monocular deprivation. In a series of studies we established that a P-adrenoceptor antagonist, propranolol, infused directly into kitten visual cortex indeed suppressed the expected shift in ocular dominance in a concentration-dependent manner (Fig. 1, Kasamatsu and Shirokawa, 1985; Shirokawa and Kasamatsu, 1986). It is remarkable that standard deviations for the propranolol data shown in Figure l
602
--
O 101 ,
,
D L - Propranolol Scopolamine
.
,
io-J
,
, lo-’
comparable experimental conditions, we noted earlier that the involvement, if any, of a-adrenoceptors is minimal in suppression of cortical plasticity in kitten visual cortex and that, due to presumed low concentrations at the recording sites, the effects of propranolol could not be explained by its non-specific action equivalent to local anesthetics (Kasamatsu and Shirokawa, 1985). In short, activation and inactivation of P-adrenoceptors seem to occupy a key step in biochemical cascades underlying the regulation of visuocortical plasticity.
Involvement of muscarinic acetylcholine (ACh) receptors
Conc e nt ra t 10n of Drugs
iM
in an osmotic pump)
Fig. 1. Concentration-dependent suppression of ocular dominance shift following monocular deprivation in propranololinfused kitten visual cortex. Drugs were continuously infused (1 k l / h ) into the cortex according to the method described (Kasamatsu et al., 1981a). There was a positive relation between drug concentrations (abscissa) and binocularity (ordinate). Ocular dominance groupings were determined based on the 7-group scheme of Hubel and Wiesel (1962). Open circles with vertical bars, mean binocularity (average of 2-9 determinations per point) B and its standard deviation/ range for a propranolol infusion. Closed circles are for a scopolamine infusion (2-3 determinations per point). No vertical bar was shown at 10- M scopolamine since binocularity from two kittens happened to be the same. C and r refer to the lower end of the 95% confidence intervals (B 3 0.69) and the 5% rejection intervals (B > 0.53) of normal binocularity, respectively (Kasamatsu et al., 1985). (From Imamura and Kasamatsu, 1989.)
’
are very small as compared to results with other drugs such as lithium, for example, used in related studies. Later, we also showed that the plasticity, once significantly decreased due to the propranolol infusion, recovered when the in uiuo availability of endogenous NA increased spontaneously some time after the termination of the propranolol infusion (Shirokawa and Kasamatsu, 1987). This recovery of cortical plasticity was accelerated by a cortical infusion of NA and further delayed by an infusion of tunicamycin, an inhibitor of protein glycosylation, the process needed to induce new P-adrenoceptors. Under
In addition to P-adrenoceptors, the involvement of other receptor types within the visual cortex is plausible in the regulation of ocular dominance plasticity, since the suppression of cortical plasticity by propranolol was not complete. A possible role of ACh afferents in this matter was suggested based on a study in which a decrease in the plasticity of kitten visual cortex was shown by placing lesions concurrently at the ascending NA bundles and ACh-containing somata in the basal forebrain (Bear and Singer, 1986). We scrutinized this proposal by asking if muscarinic activation of ACh receptors restored the plasticity (as previously shown by an NA infusion to kitten visual cortex which had been rendered aplastic due to inactivation of the NA-P-adrenoceptor system). We found the following. (1) A cortical infusion of the muscarinic ACh agonist, bethanechol, restored the plasticity (i.e., susceptibility to monocular deprivation) to the 6-OHDA-infused, thus aplastic, visual cortex of kittens. The “plasticityenhancing” effect of bethanechol was confined to an area near its infusion. (2) There was a positive correlation between concentrations of bethanechol and the strength of the thus-restored plasticity. ( 3 ) In restoring the plasticity, bethanechol was at least 100-fold less effective than that of NA under comparable conditions. (4) Conversely, the plasticity-suppressing effect of a muscarinic an-
603
tagonist, scopolamine, was much weaker than that of propranolol (Fig. 1). ( 5 ) Finally, bethanechol infusion failed to restore the plasticity to kitten visual cortex rendered aplastic due to a preceding infusion of propranolol. Taken together, these results suggest that the availability of P-adrenoceptors seems to be a prerequisite for integrating the plasticity-enhancing effect of activation of the NA and ACh systems, the former playing a primary role helped by the latter under certain conditions (Imamura and Kasamatsu, 1989). A possible lithium-sensitive mechanism for synergistic integration of the two components in the regulation of ocular dominance plasticity has been discussed (Kasamatsu et al., 1989b; Ohashi and Kasamatsu, 1989; Kasamatsu and Imamura, 1990). NA activation of P-adrenoceptors in adult cortex Results summarized above clearly indicate that the activation of the NA-P-adrenoceptors satisfies, at least in part, both “necessary and sufficient” conditions to maintain ocular dominance plasticity in immature visual cortex. Will this paradigm also “fit” when we study the “aplastic” cortex of adult animals? Our preliminary studies using a continuous infusion with either dibutyryl cyclic AMP (dbc AMP) or a stimulant of adenylate cyclase suggested possible, at least partial, restoration of visuocortical plasticity to the mature cortex; in response to brief ( < 1 week) monocular deprivation, many binocular cells became undetectable, thus resulting in the ocular dominance distribution dominated by monocular cells (Kasamatsu, 1980; 1986). Previously, no obvious changes in ocular dominance were reported in adult cats after brief monocular deprivation, although a recent abstract showed that visuocortical ceIls in layer II/III of adult cats tended to remain sensitive to monocular deprivation when it lasted for three months (Fox et al., 1989b). Furthermore, in adult cats which had been dark-reared throughout the susceptible postnatal period until adulthood (thus preserving physiological plasticity in
their visual cortex; Cynader and Mitchell, 1980), a continuous infusion of NA into the visual cortex caused an obvious shift of ocular dominance in response to monocular lid suture for 4 weeks (Shirokawa et al., 1989). Conversely, a cortical infusion of metoprolol, a p ,-adrenoceptor antagonist, suppressed this change. The paradigm of the present study was as follows: Cats (17 weeks to over 3 years old) were monocularly exposed to the usual laboratory environment for 2 h daily for 6 days successively. Concurrently, the LC and its vicinity were electrically stimulated with a train of 4 square pulses (0.05 msec width, 50 Hz rate, 1.5 mA intensity) every 3.3 sec. No gross changes in behavior were noted during the LC stimulation, since the current intensity was set just below the level with which slight twitches of masseter muscles, whiskers and pinae were observed. The cats were kept in the dark for the remaining 22 h. Visual receptive fields were recorded on the seventh day. We found that: (1) the overwhelming majority of cells were visually responsive as expected, indicating no discernible effects of dark rearing for a total of 132 h, (2) the proportion of binocular cells was about half that of the control, with the ocular dominance histogram strongly dominated by monocular cells, (3) under comparable conditions, these changes were not obtained in visual cortex of LC-stimulated cats, if the cortex had been infused prior to experimentation with either 6-OHDA (Kasamatsu et al., 1985) or propranolol (unpublished observation). We credited these results to the plasticity-enhancing effects of endogenous NA released within the visual cortex in response to electric stimulation of NA-containing cells in the LC of adult cats.
-
Acute preparations: a new paradigm While we were studying the effects of iontophoretically injected NA on excitability of visuocortical cells, in response to a properly oriented, moving light slit, we observed that, throughout overnight experiments, the majority of recorded
604 a
b
Rec I
Z
1
C
R e c Ill
u e c I1
3
5
7
1
3
ocular dominance qroups Rec I
5
-7
1
R e c II
3
5
7
RPC 111
squint c o r r e ‘ f e d
0
30
( ‘Ioyrs )
Q,
Fig. 2. Changes in ocular dominance of normal adult cats depend primarily on the combination of a continuous NA infusion and convergence/divergence of the visual axes under anesthesia and paralysis. The ocular dominance distribution was dominated by binocular cells in the first recording (Rec I) under the “usual” acute recording condition (a). The incidence of binocular cells decreased in the second recording (Rec 11) which took place under the presence of spontaneous “acute paralysis squint” after the start of 0.5 mM NA infusion (b). After “optical correction” of the paralysis squint, many binocular cells were again encountered in the third recording (Rec 111) from the same hemisphere. Temporal relations among various treatments are shown in d. All three recording tracks were placed in an area 2-3 mm from the cannulation site, in each 30 visually active cells ( n = 30) being recorded at an average spacing of 100 p m . Ocular dominance groupings are based on the Hubel and Wiesel scheme (1962); group I and 7, monocular cells responding exclusively to stimulation of the contralateral and ipsilateral eye, respectively. Group 4 cells receive balanced excitatory input from the two eyes. Other groups fall in between with different degrees of influence on the two eyes. Binocularity (B) is defined as the proportion of group 2-6 cells divided by the total number of visually active cells recorded ( n = 30). (Modified from Heggelund et al., 1987.)
cells was monocular, even though the animals we used were either normal adult cats or 6-OHDAinfused kittens (Heggelund and Kasamatsu, 1981). Recently, we took a closer look at our data. We hypothesized as follows: the observed preponderance of monocular cells in “normal” visual cortex may be the result of two interacting factors; (1) the higher-than-usual level of plasticity present in a given cortex due to excess NA introduced by repeated injections of concentrated (0.5 M) NA, and (2) divergence of the two visual axes brought about by anesthesia and paralysis, the necessary conditions for receptive-field recording. In other words, effects of a small amount of disparity ( < 5 deg) among pairs of receptive fields for the two eyes, which are usually tolerated by the majority of binocular cells under acute experimental conditions, become exaggerated due to the enhanced plasticity in the NA-excess cortex and thus cause a loss of binocular cells as typically seen in chronically strabismic cats (acute paralysis squint).
We tested this hypothesis in the visual cortex of normal adult cats which had been continuously infused with 0.05-0.5 mM L-NA during receptive-field recording under general anesthesia and paralysis. Divergence and convergence of the visual axes were controlled by placing ophthalmic prisms in front of the cat’s eye. We were able to show the presence of both U-shaped and normal distribution of ocular dominance in the same hemisphere of the same cat, depending on the optical divergence and convergence of the visual axes, respectively (Heggelund et al., 1987). Such changes were not obtained, if the visual cortex was infused with saline rather than L-NA. The results are exemplified in Figure 2. Alternations between the two patterns of ocular dominance were shown repeatedly in the same cortex. These results suggested a causative relationship between the availability of NA or activation of P-adrenoceptors, and the high level of visuocortical plasticity restored to the aplastic cortex of adult cats under anesthesia and paralysis. Since
605 A 215-
5
Drug-free
B
05mM
N = 2 n.60
I-NA
N = 2 n.60
8.062
(u
m GLU
‘ o - 1 2 3 4 5 6 7
1
2
0 Monocular
3
4
5
6
7
GLU
0 Monocular
Fig. 3. Ocular dominance shift was induced by monocular exposure for 22 h in the NA-infused visual cortex of kittens ( n = 2 ) under an acute experimental condition. In control cortex infused with the vehicle solution (0.4% ascorbate saline). the majority of recorded cells ( n = 60) was binocular (A). The ocular dominance distribution became clearly dominated by group-7 monocular cells which exclusively responded to stimulation of the “exposed” eye (open circle, B). The sequence of treatments is shown at the bottom. GL, geniculate axon spikes. U, visually unresponsive cells. A vertical bar at the top of each column refers to the range of the number of units per column (unpublished data).
-
we manipulated one of the two variables, the quality of visual afferent to binocular cells and the strength of visuocortical plasticity, at a time, no explanations other than those offered here are plausible. The above conclusion prompts this question: is the plasticity restored by an NA infusion into adult cortex a continuum of the plasticity fully expressed in immature cortex? We believe that we have obtained an answer to this query. Corroborating the above results in adult cats, we recently demonstrated an obvious shift of ocular dominance in the NA-infused visual cortex of young kittens, following monocular exposure to moving visual targets for 22 h under general anesthesia and paralysis (Fig. 3; Imamura and Kasamatsu, 1988). Again the proper combination of the two factors was necessary to induce acutely a shift in ocular dominance; the high level of cortical plasticity was assured by a continuous NA infusion and the effective activation of visual cortical cells using oriented, moving targets, in this case uninterrupted commercial TV programs.
-
Other combinations, such as the TV programs and NA-infused adult cortices or NA-infused kitten cortex and stimulation with electronic noise on the same TV screen, failed to induce the ocular dominance shift. The common denominator between the above two acute paradigms, acute paralysis squint in adult cats and monocular exposure in kittens, was the activation of p-adrenoceptors within the visual cortex by exogenous NA. Biochemical cascades which follow activation of p-adrenoceptors seem to have primary responsibility for maintaining visuocortical plasticity at a high level. The acute experimental condition of “passive” eye movement of the visually exposed eye by mechanically pulling for 12 h while the other eye was covered, caused a significant loss of binocular cells in kitten visual cortex (Freeman and Bonds, 1979). Freeman and Bonds interpreted the finding as suggesting that ocular dominance plasticity is normally controlled by ocuIomotor/proprioceptive afferent activity which is suppressed under anesthesia and paralysis. In that study it was not immediately clear how nociceptive stimuli, a powerful activator of NA cells in the LC, were controlled. In another acute study, monocular exposure for 15-18 h to a rotating disk of high-contrast gratings conditioned with electrical stimulation of the non-specific thalamic nuclei or mesencephalic reticular formation (MRF) was reported as causing a change in ocular dominance toward the stimulated eye (Singer and Rauschecker, 1982). Taken together with results from an accompanying paper on effects of unilateral lesions aimed at the non-specific thalamic nuclei (Singer, 1982), it was concluded that the MRF and the non-specific thalamic nuclei are the site of “gating” function in the regulation of visuocortical plasticity (Singer and Rauschecker, 1982). There are difficulties in data interpretation, however. The majority of recorded cells in their stimulation experiment remained binocular (binocularity 67%) despite monocular exposure. Although the ocular dominance distribution was relatively in favor of the
606
exposed eye, no one calls this distribution as “shifted” toward the exposed eye. Furthermore, for unexplained reasons, an unusually high proportion (43%) of recorded cells had abnormal receptive fields. Based on their studies on strabismic cats and cats which experienced a combined manipulation of intraorbital rotation of one eye and monocular lid suture of the other eye (Singer et a f . , 1979a, b), Singer proposed that “visual attention” plays an important role in gating experience-dependent modification of cortical connectivity in behaving animals. If this scheme is plausible, neural consequences of “visual attention” should be carried through activation not of the ACh system in response to stimulation of the MRF as thought by Singer and associates, but of the NA cells in the LC (Foote et af., 1983; Aston-Jones, 1985; Jacobs, 1986). In the present study, we showed that ocular dominance shift was indeed induced in the NA-infused kitten visual cortex despite anesthesia and paralysis. Conceptually, there is no reason to believe that the localized cortical infusion of NA would have moved eyeballs or activated the non-specific thalamic nuclei under general anesthesia and paralysis. Rather, it is most likely that some changes in ocular dominance reported in the previous studies on kittens under acute conditions (Freeman and Bonds, 1979; Singer and Rauschecker, 1982) were due to the LC system activation which resulted in the partial restoration of visuocortical plasticity in the acute preparations. Thus, the present results from acutely anesthetized and paralyzed adult cats, as well as kittens, strongly indicate a unique role of the NA-P-adrenoceptor system in the regulation of visuocortical plasticity. In this regard the difference between the immature and mature cortex seems to be quantitative rather than qualitative, with the former more sensitive to altered experience than the latter. Neural mechanisms of ocular dominance changes: critical evaluation Wiesel and Hubel (1965) first proposed that in visual cortex of monocularly deprived kittens, a
set of comparable visual afferents from the two eyes compete with each other for the available postsynaptic sites or trophic factors which are probably limited in number or amount (binocular competition). This remains a leading explanation of “experience- and age-dependent” modification of ocular dominance, though its cellular bases need to be refined further. A scheme along the same line was later presented with special reference to possible sprouting and retraction of geniculate terminal arbors in layer IVc of normally developing monkeys during ontogenic segregation of ocular dominance columns (Hubel et a f . , 1977). The scheme however, falls short of explaining the “dynamic” nature of the matter, as typically seen in the U- or W-shaped distribution of ocular dominance following brief monocular lid suture (Hubel and Wiesel, 1970; Olson and Freeman, 1975; 1980) and the exaggerated expression of dominancy by the contralateral eye in NA-infused, otherwise normal visual cortex of kittens (Kuppermann and Kasamatsu, 1984). The scheme also fails to explain morphological findings such as recapture of lost territory in layer IV by the previously deprived eye under a reversesuture paradigm (Le Vay et d., 1980) and the lack of “morphological plasticity” in dark-reared cats (see below). Further discussions continue below centering on the issue of “activity dependency,” as it relates in one way or another to the possible contribution of chemically defined neurons. Acticity dependency Activity-dependent processes underlie devtlopment and synaptic modification in many neural systems such as the neuromuscular junctions of newborn mammals (see Harris, 1981; Van Essen, 1982 for reviews), and the formation of the developing/regenerative retinotectal map in goldfish (Meyer, 1983; Schmidt and Edwards, 1983) and three-eyed frogs (Cline et a f . , 1987). In cultured neurons from the snail Hefisoma, action potentials induced by direct stimulation of individual
607
somata suppressed neurite elongation and growth cone motility (Cohan and Kater, 1986). The determinant role of retinal afferent in normal segregation of ocular dominance columns was primarily studied in layer IV of kitten visual cortex by Stryker and Harris (1986). They showed in physiology and autoradiography studies that a sodium-channel blocker, tetrodotoxin (TTX), suppressed the age-dependent segregation of binocular input convergence into ocular dominance columns. They interpreted the results as suggesting that electrical activity of the retinogeniculate pathway is needed for formation or maturation of ocular dominance columns in layer IV of the visual cortex. However, the interpretation may not be totally free from some pitfalls. For example, as evidenced by the very low level (about one fifth of control) of label density in the TTX-injected animal, in addition to the lack of electrical activity, other metabolic activity such as axoplasmic transport of various macromolecules might have been impaired (Edwards and Grafstein, 1983) under their experimental conditions, thus effectively contributing to the observed lack of density fluctuation. Their physiological findings such as prevalence of binocular cells, above the level of 2 weeks of age, may suggest that intravitreal injections of TTX have, in fact, caused “programmed ingrowth” of geniculate axonal arbors to stop or even become retarded. It would be interesting if “inborn” monocular cells in layer IV, which remain in the visual cortex of TTX-treated kittens, are cluatered at regular intervals thus keeping a special relation to “imaginary” columns yet to be formed. Lack of visually evoked activity may interfere with the age-dependent formation of ocular dominance columns. In dark-reared cats, anatomical segregation of geniculate afferents seems to have taken place, as evidenced by peak-to-trough fluctuation in the density of label through layer IV (Mower et al., 1985; Stryker and Harris, 1986; Swindale, 1988). Compared to normal, however, the density fluctuation was reduced with high
irregularity in shape and width. Mower et af. (1985) concluded that the capacity of immature brain to segregate geniculate afferents into ocular dominance columns disappears with time and the anatomical susceptibility to monocular deprivation is lost during several weeks of dark rearing. Thus, physiological effects of dark-rearing, i.e., prolongation of the postnatal susceptible period into adulthood (as shown by Cynader and Mitchell, 1980; Mower et al., 198l), are not necessarily mediated through cellular and neurochemical mechanisms responsible for the anatomical segregation of ocular dominance columns in layer IV. To further test the proposed role of “electric activity” in the segregation of geniculate afferents, Stryker et al. carried out the following 3 studies. (1) Young kittens were briefly subjected to lid suture of one eye and TTX injections of the other eye. The kittens were left in total darkness, thus creating an imbalance in “electric activity” of the two retinae without involving visually evoked pattern vision (Chapman et al., 1986). A shift or bias of ocular dominance toward t h e relatively “active” eye with tonic retinal discharges was found (see also Greuel et al. (1987) for “negative” results). The results seem to reinforce the idea of a vital role of binocular competition or imbalance in determining the ocular dominance distribution in immature cortex. ( 2 ) Using kittens subjected to binocular impulse blockade with intravitreal TTX injections, effects on ocular dominance of alternative electric stimulation of the two optic nerves were compared to those of simultaneous optic tract stimulation. In the former, the proportion of monocular cells was high, the degree similar to cortical effects of strabismus. The latter produced the ocular dominance pattern indistinguishable from that of kittens treated with TTX alone and even more binocular than the age-matched control. These contrasting results were interpreted as indicating that simultaneous electrical stimulation of the two optic nerves effectively blocked the
608
usual segregation of geniculate afferents, eliminating chances of binocular competition (Stryker, 1989). Spontaneous discharges of retinal ganglion cells are, indeed, not random in nature but have a pattern (Mastronarde, 1983a,b). Stryker, therefore, concluded that the usual formation of ocular dominance is “instructively” controlled by the “patterned” electric activity in tonic retinal discharges (dark discharges). (3) More recently, Reiter and Stryker (1988) reported a unique expression of ocular dominance plasticity in kitten visual cortex which had been directly infused with 10 mM muscimol, yaminobutyric acid (GABA) agonist. Recording visual receptive fields after cortical cells recovered from the muscimol-induced blockade of cell firing, they found a consistent bias of ocular dominance toward the less-active, closed eye. The majority of recorded cells remained binocular (B = 0.61), however. The resultant distribution was distinct from the “normal” distribution, which was obtained from kitten visual cortex infused with 10 p M TTX (Reiter et al., 1986). Although binocular cells prevailed in both preparations, there was a basic difference between them: muscimol inhibits postsynaptic activity alone and TTX totally silences cortical activity including geniculate afferents. The authors thus suggested that the direction of changes in ocular dominance depends on “postsynaptic membrane conductance” but not postsynaptic spike activity. A crucial question is to elucidate neurochemical factors contributing to this hypothetical “threshold” mechanism based on fluctuation of membrane conductance. It is time t o integrate the two separately pursued avenues of research in the regulation of cortical plasticity: activity dependency in the visual pathway on the one hand and global neurochemical factors such as the NA and ACh system on the other.
GABA, receptors and (NMDA) receptors
N-methyl-u-aspartate
In the preceding section, we discussed the crucial role of P-adrenoceptors, helped by muscarinic
ACh receptors, in the regulation of ocular dominance plasticity. In addition, the involvement of two other types of receptors has been suggested.
GABA. receptor Earlier, an involvement of GABA receptormediated processes in visuocortical plasticity was suggested (Duffy et al., 1976). Duffy and colleagues reported that intravenous injections of a subconvulsive dose of bicuculline (BCC), a GABA, antagonist, recovered the lost receptive fields for the monocularly deprived eye in the visual cortex of adult cats subjected to monocular lid suture since birth. Several follow-up studies by other investigators showed that for some cells BCC indeed reversed the effects of monocular deprivation. However, none of the recovered receptive fields were dominated by input from the previously deprived eye, suggesting that the observed effect which lasted very briefly during either iontophoretic or intravenous applications of BCC, could not be interpreted as genuine restoration of ocular dominance plasticity (Sillito et al., 1981). This may be related to the fact that the GABA system is a main constituent of neural networks needed for visual processing of specific afferents rather than being involved in the regulation of the plasticity. As expected, many cells showed abnormal receptive-field properties, though binocular, insofar as BCC was effective. Ramoa et al. (1988) directly infused kitten visual cortex with 5 mM BCC methiodide concurrently with monocular deprivation for a week. The expected shift in ocular dominance was significantly reduced compared to results from the saline-infused control hemispheres. The authors’ explanations were as follows: loss of selectivity due to BCC caused in effect an increase in coupling between presynaptic and postsynaptic activity, which in turn satisfied a necessary condition for Hebbian synapses of binocular cells to be strengthened, thus effectively counteracting with the usual consequence, binocular imbalance. There is a serious drawback in the design of this study. As was the case for a continuous infusion of concentrated glutamate (0.5 M, Shaw and Cy-
609
nader, 19841, the majority of visual cortical cells either lost selectivity or became visually unresponsive in the BCC-infused cortex. Conceptually, therefore, the situation cannot be distinguished from that of de facto binocular deprivation which interferes little with ocular dominance distribution despite its deleterious effects on the response quality of visuocortical cells (Wiesel and Hubel, 1965). Autoradiographic studies on GABA-replaceable [ 'H]muscimol binding sites in the visual cortex of monocularly deprived kittens have so far produced ambiguous results. Mower et al. (1986) found no differences between normal and monocularly deprived cats in number as well as affinity of GABA, receptor binding. The binding sites are scattered more or less evenly throughout the six cortical laminae and this laminar pattern stayed the same after monocular deprivation. Using a comparable method but with 15-fold concentrated [ 3H]muscimol, however, Shaw and Cynader (1988) reached the opposite conclusion. In normal cats the laminar distribution of GABA, binding peaked in the central lamina flanked by a lower number of binding sites toward the cortical surface and the white matter. They found an over 90% increase in the number of muscimol binding sites in monocularly deprived cats, without changing binding affinity and their laminar distribution pattern. The cause of this increase remains unexplained. An immunocytochemical and biochemical study showed that monocular deprivation affected neither the glutamic acid decarboxylase (GAD, a GABA synthesizing enzyme) immunoreactivity pattern throughout the six cortical laminae in kitten visual cortex nor GAD activity in homogenates of the cortex (Bear et al., 1985). An autoradiographic study of muscimol binding showed a rather complicated time course of maturation. The binding surpassed the adult value at 4 weeks of age, became slightly lower than at 8-9 weeks, and showed a second and larger peak at 13-14 weeks postnatally (Shaw et al., 1986). Binding affinity also fluctuated widely, showing the
highest value at birth and the lowest 8-9 weeks postnatally (difference > 4-fold). The problem with this seemingly straightforward study was its sparsity of datum points covering different age groups. More recently, a biochemical study clearly showed that GAD activity in areas 17 and 18 of cats increases gradually during the first postnatal month and reaches the adult level at 5-6 weeks postnatally without making any peaks (Fosse et al., 1989). Taken together, it is evident that the GABA, receptor system is not well suited for the regulation of cortical plasticity, though it may be involved in the expression of the plasticity as an indispensable building block of functioning neural network in uiuo.
NMDA receptors A unique biophysical feature of NMDA receptors (i.e., voltage-dependent gating of C A + + influx, which is antagonistically controlled by Mg++ ion) appears to give an ideal ionic mechanism to initiate long-term potentiation (LTP) in the CA1 region of mammalian hippocampus (e.g., Harris et al., 1984; Wigstrom and Gustaffson, 1984). Involvement of this type of glutamate receptors has been implied, indirectly (Rauschecker and Hahn, 1987; Tsumoto et al., 1987) or directly (Kleinschmidt et al., 19871, in the regulation of cortical plasticity. The latter authors reported that in visual cortex infused with 50 mM D,L-2amino-5-phosphonovaleric acid (APV), a specific and competitive antagonist of NMDA receptors, many cortical cells remained binocular despite concurrent monocular deprivation. Our recent study under comparable conditions simply did not confirm their claim (Kasamatsu et al., 1989a,b). However, we obtained a distinct effect of APV on cortical cells' selectivity; a substantial proportion of visually active cells became non-selective for orientation/direction. This part of our findings is consistent with those of Kleinschmidt et al. (1987). The decrease in selectivity raises a conceptual difficulty in placing NMDA receptors in the center of regulatory mechanisms of ocular dominance plasticity. Miller et al. (1989) found that
visual cortical cells were totally silenced in the presence of APV (50 mM in minipump). Furthermore, iontophoretically injected APV reduced visually evoked responses in layer II/III of cat visual cortex throughout adulthood, although the APV’s effect disappeared by 6 weeks of age, at the height of critical period sensitivity, in layer IV-VI (Fox et al., 1989a). Taken together, available data indicate that presently no specific role may be assigned to NMDA receptors in the regulation of ocular dominance plasticity. Rather, it seems most likely that activation of NMDA receptors is an indispensable part of normal impulse transmission within the visual cortex (Miller et al., 1989). Intriguingly in the hippocampal dentate gyrus, NMDA receptors seem necessary for NA to induce, via /3,-adrenoceptors, activity-independent LTP without tetanic stimulation, with persistent depolarization and an increase in input resistance (Stanton et al., 1989). It was also noted that NA specifically enhanced NMDA-gated influx of CAtf in the granule cell body layer. Thus, LTP in the dentate gyrus seems to be synergistically controlled by P,-adrenoceptors and NMDA receptors.
Other factors Trophic effects The contribution of “spontaneous” tonic retinal discharges to the regulation of excitability of cells in the central visual pathway has been recognized for some time (Chang, 1952; Arduini and Hirao, 1960; Kasamatsu et al., 1967). A TTX-induced blockade of retinal activity in one eye caused major changes in geniculate cell size within one week of the treatment (Kuppermann and Kasamatsu, 1983). Since the change occurred throughout the geniculate nucleus, including the monocular segment, this result was interpreted as suggesting the presence of trophic factors continuously transported by retinal discharges under normal conditions. Intravitreally injected TTX was also found to
interfere substantially with the maturation of geniculate physiology in kittens (Dubin et al., 1986) and synaptogenesis of pyramidal cells in the visual cortex of developing rat pups (Riccio and Mathews, 1985a,b). Intravitreal injections of TTX reduced “transiently expressed” acetylcholinesterase activity in visual cortex of immature rats (Robertson et al., 1989). Membrane depolarization has been known to regulate the accumulation of macromolecules in excitable cells. In sympathetic neurons, for example, preganglionic electric stimulation is known to increase biosynthesis of tyrosine hydroxylase (Ip and Zigmond, 1984) and accumulate its mRNA (Black et al., 1985; Jonakait and Black, 1986). The similar increase in tyrosine hydroxylase mRNA was shown in LC neurons in vico as well as in ritro (Black et al., 1987). In short, it is a distinct possibility that visual afferents impinging on cortical cells play a crucial role in maintaining cellular metabolisms and thus releasing trophic factors.
Fiber projections to non-specific systems It is unlikely that effects of patterned impulses induced in the retinogeniculate projection, especially their tonic influence in a slow and cumulative manner, are confined within the traditional boundary of the central visual pathway. For example, NA cells in the LC receive visual afferents via polysynaptic pathways (Aston-Jones and Bloom, 1981; Watabe et al., 1982). The MRF receives mono- or oligo-synaptic afferents from various sensory systems, including the visual centers (Bell et al., 1964; Magni and Willis, 1964; Kasamatsu, 1970a,b). Therefore, it is quite conceivable that blockade or activation of retinal afferents exerts their influences on cortical plasticity not only through phasic impulse transmission within the ascending visual pathway but also indirectly and in a tonic manner through many non-lemniscal structures. Release of transmitter /modulator There is a further possibility of “cross talk” among different transmitter/modulator systems,
61 1
thus synergistically contributing to the regulation of cortical plasticity. The presence of presynaptic receptors may serve to regulate transmitter release heterosynaptically. There is ample evidence for “local control” of transmitter release (Romo et ul., 1986; Cheramy et al., 1986). Marrocco et al. (1987) reported that visual stimulation with either light flash or rotating gratings caused a transient and slow increase in oxidation currents which are related to CA release detected electrochemically in area 17 of anesthetized and paralyzed monkeys and cats. Specificity of the observed increase in oxidation currents, in terms of their chemical origin and spatial localization, is not currently satisfactorily clarified. Yet, it is plausible that a synchronous volley of ascending impulses, along the geniculocortical pathway and others, triggered local accumulation of K + in extracellular space (Singer and Lux, 1975; Skinner and Molnar, 1983) which in turn caused a slow release of NA, besides other molecules, via an increase in CA t i influx into NA-containing terminal boutons iii the neighborhood of retinotopically stimulated cortical cells. Second messenger systems such as protein kinase C (Allgaier et al., 1986; Shuntoh and Tanaka, 1986; Malenka et al., 1987; Nichols et al., 1987; a review by Kaczmarek, 1987) and C a t + calmodulin-dependent PK I1 (Nichols et al., 1990) seem to be involved in transmitter release in the mammalian brain. Another far-fetched possibility of released transmitters in cortical plasticity has been suggested: motility of growth cones of 5 HT-containing neurons of Helisoma in culture was inhibited by 5HT, thus left unable to form electrical synapses with neighboring neurons (Haydon et al., 1984). Conclusions
We critically reviewed here various neurochemical factors which have been proposed to be involved in the regulation of ocular dominance plasticity. A basic strategy commonly used in the field is described here: receptor antagonists/
agonists are continuously infused directly into the visual cortex of kittens or adult cats, respectively, concurrently with brief monocular deprivation, in hopes that such treatments would modify the expected changes (kittens) or their lack (adult cats) in the ocular dominance distribution. (1) We first reviewed evidence for the essential role of NA-P-adrenoceptors in the regulation of cortical plasticity. Included are suppression of the plasticity by 6-OHDA or /3-adrenoceptor antagonists, restoration of the plasticity by NA to the 6-OHDA-infused kitten cortex or the aplastic adult cortex, the plasticity restoration to the adult cortex by electrical stimulation of the LC, and the plasticity-enhancing effect of NA in anesthetized and paralyzed preparations. (2) A contribution of ACh was also evaluated. We presented a unified view that the NA-Padrenoceptor system is primarily responsible for the regulation of cortical plasticity, being supplemented by muscarinic ACh receptors. ( 3 ) Despite the high expectation mainly based on NMDA receptors’ involvement in hippocampal LTP, we failed to confirm a specific role for them in the regulation of cortical plasticity. (4) Involvement of GABA, receptors in this matter was also discussed. We suggest that GABA, receptors, like NMDA receptors, cannot be primary carriers of regulatory mechanisms of cortical plasticity, though they may be needed for its expression. (5) Finally, we discussed the issue of “activity dependency” in changes of ocular dominance. It is hoped that interactions among the two separately pursued avenues of research, the nature of visual afferents per se which determines activity dependency, and modulatory roles of NAactivated P-adrenoceptors should be directly studied in the same visual cortex.
Acknowledgements
Supported by USPHS grant EY06733 and Core Grant EY 06883.
612
References Abercrombie, E.D. and Zigmond, M.J. (1989) Partial injury to central noradrenergic neurons: Reduction of tissue norepinephrine content is greater than reduction of extracellular norepinephrine measured by microdialysis. J. Neurosci., 9: 4062-4067. Adrien, J., Blanc, G., Buisseret, P., FrCgnac, Y., Gary-Bobo, E., Imbert, M., Tassin, J.-P. and Trotter, Y. (1985) Noradrenaline and functional plasticity in kitten visual cortex: A reexamination. J. Physiol. (London), 367: 73-98. Allgaier, C., von Kugelgen, 0. and Hertling, G. (1986) Enhancement of noradrenaline release by 12-0-tetradecanoyl phorbol-13-acetate, an activator of protein kinase C. Eur. J. Pharmacol., 129: 389-392. Arduini, A. and Hirao, T. (1960) Enhancement of evoked responses in the visual system during reversible retinal inactivation. Arch. Ilal. Biol., 98: 182-205. Aston-Jones, G. (1985) Behavioral functions of locus coeruleus derived from cellular attributes. Physiol. Psychol., 13: 118126. Aston-Jones, G. and Bloom, F.E. (1981) Norepinephrine-containing locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci., 1: 887-900. Bear, M.F. and Daniels, J.D. (1983) The plastic response to monocular deprivation persists in kitten visual cortex after chronic depletion of norepinephrine. J. Neurosci., 3: 407416. Bear, M.F. and Singer, W. (1986) Modulation of visual cortical plasticity by acetylcholine and noradrenaline. Nature (London), 320: 172-176. Bear, M.F., Paradiso, M.A., Schwartz, M., Nelson, S.B., Carnes, K.M. and Daniels, J.D. (1983) Two methods of catecholamine depletion in kitten visual cortex yield different effects on plasticity. Nature (London), 302: 245-247. Bear, M.F., Schmechel, D.E. and Ebner, F.F. (1985) Glutamic acid decarboxylase in the striate cortex of normal and monocularly deprived kittens. J. Neurosci., 5: 1262-1275. Bear, M.F., Cooper, L.N. and Ebner, F.F. (1987) A physiological basis for a theory of synaptic modification. Science, 237: 42-48. Bell, C., Sierra, G., Buendia, N. and Segundo, J.P. (1964) Sensory properties of neurons in the mesencephalic reticular formation. J. Neurophysiol., 27: 961 -987. Black, I.B., Chikarashi, D.M. and Lewis, E.J. (1985) Transsynaptic increase in RNA coding for tyrosine hydroxylase in a rat sympathetic ganglion. Brain Res., 339: 151-153. Black, I.B., Adler, J.E., Dreyfus, C.F., Friedman, W.F., LdGamma, E.F. and Roach, A.H. (1987) Biochemistry of information storage in the nervous system. Science, 236: 1263- 1268. Chang, H.T. (1952) Cortical response to stimulation of lateral geniculate body and potentiation thereof by continuous illumination of retina. J. Neurophysiol., 15: 5-26. Chapman, B., Jacobson, M.D., Reiter, H.O. and Stryker, M.P. (1986) Ocular dominance shift in kitten visual cortex caused
by imbalance in retinal electrical activity. Nature (London), 324: 154-156. Cheramy, A,, Romo, R., Godeheu, P. and Glowinski, J. (1986) In fiuo presynaptic control of dopamine release in the cat caudate nucleus - 11. Facilitatory or inhibitoiy influence of L-glu tamate. Neuroscience, 19: 1081- 1090. Cline, H.T., Debski, E.A. and Constantine-Paton, M. (1987) N-methyl-D-aspartate receptor antagonist desegregates eye-specific stripes. Proc. Nut/. Acud. Sci. U S A , 84: 43424345. Cohan, C.S. and Kater, S.B. (1986) Suppression of neurite elongation and growth cone motility by electrical activity. Science, 232: 1638-1640. Cooper, L.N. (1987) Cortical plasticity: Theoretical analysis, experimental results. In J.P. Rauschecker and P.R. Marler (Eds.), Imprinting and Cortical Plasticity, Cornparatice Aspects of Sensitice Periods, John Wiley & Son, New York, pp. 177-192. Cynader, M. and Mitchell, D.E. (1980) Prolonged sensitivity to monocular deprivation in dark reared cats. J. Neurophysiol., 43: 1026-1040. Daw, N.W., Rader, R.K., Robertson, T.W. and Ariel, M. (1983) Effects of 6-hydroxydopamine on visual deprivation in kitten striate cortex. J. Neurosci., 3: 907-914. Daw, N.W., Robertson, T.W., Rader, R.K., Videen, T.O. and Coscia, C.J. (1984) Substantial reduction of cortical noradrenaline by lesions of adrenergic pathway does not prevent effects of monocular deprivation. J. Neurosci., 4: 1354-1360. Daw, N.W., Videen, T.O., Parkinson, D. and Rader, R.K. (1985a) DSP-4 (N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine) depletes noradrenaline in kitten visual cortex without altering the effects of monocular deprivation. J. Neurosci., 5: 1925-1933. Daw, N.W., Videen, T.O., Rader, R.K., Robertson, T.W. and Coscia, C.J. (1985b) Substantial reduction of noradrenaline in kitten visual cortex by intraventricular injections of 6-hydroxydopamine does not always prevent ocular dominance shifts after monocular deprivation. E2p. Brain Res., 59: 30-35. Daw, N.W., Videen, T.O., Robertson, T. and Rader, R.K. (198.5~)An evaluation of the hypothesis that noradrenaline affects plasticity in the developing visual cortex. In A. Fein (Ed.), The Visual System, Alan R. Liss, New York, pp. 133- 144. Dubin, M.W., Stark, L.A. and Archer, S.M. (lY86) A role for action-potential activity in the development o f neuronal connections in the kitten retinogeniculate pathway. J. Neurosci!, 6: 1021-1036. Duffy, F.H., Snodgrass, S.R., Burchfiel, J.L. and Conway, J.L. (1976) Bicuculline reversal of deprivation amblyopia in the cat. Nature (London), 260: 256-257. Edwards, D.L. and Grafstein, B. (1983) InIraocular tetrodotoxin in goldfish hinders optic nerve regeneration. Brain Res., 269: 1-14. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Rei,., 63: 844-914.
613 Fosse. V.M., Heggelund, P. and Fonnum, F. (1989) Postnatal development of glutamatergic, GABAergic, and cholinergic neurotransmitter phenotypes in the visual cortex, lateral geniculate nucleus, pulvinar, and superior colliculus in cats. J. Neurosci., 9: 426-435. Fox. K., Sato, H. and Daw, N. (1989a) The location and function of NMDA receptors in cat and kitten visual cortex. J. Neurosci., 9: 2443-2454. Fox, K., Daw, N. and Sato, H. (1989b) Plasticity in adult and adolescent cat visual cortex. Soc. Neurosci. Abstr., 15: 796. Freeman, R.D. and Bonds, A.B. (1979) Cortical plasticity in monocularly deprived immobilized kittens depends on eye movement. Science, 206: 1093-1095. Frignac, Y . (1987) Cellular mechanisms of epigenesis in cat visual cortex. In J.P. Rauschecker and P. Marler (Eds.), Imprinting and Cortical Plasticity, John Wiley & Sons, New York, pp. 221-266. Frtgnac, Y . and Imbert, M. (1984) Development of neuronal selectivity in primary visual cortex of cat. Physiol. Rert., 64: 325-434. Gordon, B., Allen, E.E. and ‘Trombley, P.Q. (1988) The role of norepinephrine in plasticity of visual cortex. Prog. Neurohiol., 30: 171-191. Greuel, J.M., Luhmann, H.J. and Singer, W. (1987) Evidence for a threshold in experience-dependent long-term changes of kitten visual cortex. Der. Brain Res., 34: 141-149. Guillery, R.W. (1972) Binocular competition in the control of geniculate growth. J. Comp. Neurol., 144: 117-130. Guillery, R.W. (1973) The effects of lid suture upon the growth of cells in the dorsal lateral geniculate nucleus of kittens. J. Comp. Nrurol.. 148: 417-422. Harris, W.A. (1981) Neural activity and development. Ann. Rei. Physiol., 43: 689-710. Harris, E.W., Ganong, A.H. and Cotman, C.W. (1984) Longterm potentiation in the hippocampus involves activation of N-methyl-D-aspartate receptors. Bruin Rex, 323: 132137. Haydon, P.G., McCobb, D.P. and Kater, S.B. (1984) Serotonin selectively inhibits growth cone motility and synaptogenesis of specific identified neurons. Science, 226: 561-564. Heggelund, P. and Kasamatsu, T. (1981) Exogenous noradrenaline increases the neuronal plasticity in cat visual cortex: Lucalized, continuous microperfusion and iontophoresis. In 0. Feher and F. Joo (Eds.), Cellular Analogues of Conditioning and Neural Plasticiw, Adu. Physiol. Sci., Vol. 36, Pergamon Press, Oxford, pp. 233-242. Heggelund, P., Imamura, K. and Kasamatsu, T. (1987) Reduced binocularity in the noradrenaline-infused striate cortex of acutely anesthetized and paralyzed, otherwise normal cats. Exp. Brain Rex, 68: 593-605. Hirsch, H.V.B. and Tieman, S.B. (1987) Perceptual development and experience-dependent changes in cat visual cortex. In M. Bornstein (Ed.), Sensitive Periods in Deuelopment: Interdisciplinary Perspecfiues, Lawrence Erlbaum Assoc., Hillsdale, NJ, pp. 39-79. Hubel, D.H. and Wiesel, T.N. (1962) Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. J. Physiol. (London), 160: 106-154.
Hubel, D.H. and Wiesel, T.N. (1970) The period of susceptibility to the physiological effects of unilateral eye closure in kittens. J. Physiol. (London), 206: 419-436. Hubel, D.H., Wiesel, T.N. and LeVay, S. (1977) Plasticity of ocular dominance columns in monkey striate cortex. Phi10s. Trans. R. SOC. Lond. (Biol.), 278: 377-409. Imamura, K. and Kasamatsu, T. (1988) Acutely induced shift in ocular dominance during brief monocular exposure: Effects of cortical noradrenaline infusion. Neurosci. Lett., 88: 57-62. Imamura, K. and Kasamatsu, T. (1989) Interaction of noradrenergic and cholinergic systems in regulation of ocular dominance plasticity. Neurosci. Res., 6: 5 19-536. Ip, N. and Zigmond, R.E. (1984) Pattern of presynaptic nerve activity can determine the type of neurotransmitter regulating a postsynaptic event. Nuture (London),311: 472-474. Itakurd, T., Kasamatsu, T. and Pettigrew, J.D. (1981) Norepinephrine-containing terminals in kitten visual cortex: Laminar distribution and ultrastructure. Neuroscience, 6: 159-175. Jacobs, B.L. (1986) Single unit activity of locus coeruleus neurons in behaving animals. Prog. Neurobiol., 27: 183-194. Jonakait, G.M. and Black, I.B. (1986) Neurotransmitter phenotypic plasticity in the mammalian embryo. Curr. Top. Dell. Biol., 20: 165-175. Jonsson, G. (1980) Chemical neurotoxins as denervation tools in neurobiology. Ann. Reo. Neurosci., 3: 169-187. Jonsson, G. and Kasamatsu, T. (1983) Maturation of monoamine neurotransmitters and receptors in cat occipital cortex during postnatal critical period. Exp. Brain Res., 5 0 449-458. Jonsson, G., Wiesel, F.-A. and Hallman, H. (1979) Developmental plasticity of central noradrenaline neurons after neonatal damage-changes in transmitter functions. J. Neurobiol., 10: 337-353. Kaczmarek, L.K. (1987) The role of protein kinase C in the regulation of ion channels and neurotransmitter release. TINS, 10: 30-34. Kasamatsu, T. (1970a) Maintained and evoked unit activity in the mesencephalic reticular formation of the freely behaving cat. Exp. Neurol., 28: 450-470. Kasamatsu, T. (1970b) Effects of visual deafferentation on mesencephalic reticular activity in freely behaving cats. Exp. Neurol., 29: 251-267. Kasamatsu, T. (1979) Involvement of the P-adrenergic receptor in cortical plasticity. ARVO Abstr. Suppl. Inuest. Ophthalmof. Vis. Sci., 18: 135. Kasamatsu, T. (1980) A possible role for cyclic nucleotides in plasticity of visual cortex. SOC.Neurosci. Abstr., 6: 494. Kasamatsu, T. (1982) A role of the central norepinephrine system in regulation of neuronal plasticity in cat visual cortex. In A. Kaneko, N. Tsukahara and R. Uchizono (Eds.), Neurotransmitters in the Retina and the Visual Center~s, Biomed. Res. Suppl. 3, Biomed Res. Foundation, Tokyo, pp. 87-93. Kasamatsu, T. (1983) Neuronal plasticity maintained by the central norepinephrine system in the cat visual cortex. In
614 Prog. Psychobiol. Physiol. Psych., Vol. 10, Academic Press, New York, pp. 1-112. Kasamatsu, T. (1986) Changes in ocular dominance of adult cats following monocular lid suture: The effect of directly infused forskolin. ARVO Ahstr. Suppl. lni’est. Ophthuinzol. [’is. Sci., 27: 153. Kasamatsu, T. (1987) Norepinephrine hypothesis for visual cortical plasticity: Thesis, antithesis, and recent development. In Current Topics in Deidopmental Biology, Vol. 21, Academic Press, New York, pp. 367-389. Kasamatsu, T. (1989) Experimental amblyopia ex anopsia in kittens: Neuropharmacological approach. In P. Kellaway and J.L. Noebels (Eds.), Problems and Concepts in Decelopmental Neurophysiology, Johns Hopkins Univ. Press, Baltimore, pp. 75-91. Kasamatsu, T. and Heggelund, P. (1982) Single cell responses in cat visual cortex to visual stimulation during iontophoresis of noradrenaline. Exp. Bruin Rex, 45: 317-327. Kasamatsu, T. and Imamura, K. (1990) Ocular dominance plasticity in kitten visual cortex: Integration of noradrenergic and cholinergic regulation. In R.T. Richardson (Ed.), Actiiwtion to Acqui.sition: Functional Aspects of the Basal Forebrain Cholinergic Sysrem, Birkhauser, Boston, pp. 289-324. Kasamatsu, T. and Pettigrew, J.D. (1976) Depletion of brain catecholamines: Failure of ocular dominance shift after monocular occlusion in kittens. Science, 194: 206-209. Kasamatsu, T. and Pettigrew, J.D. (1979) Preservation of binocularity after monocular deprivation in the striate cortex of kittens treated with 6-hydroxydopamine. J. Comp. Neurol., 185: 139-162. Kasamatsu, T. and Shirokawa, T (1985) Involvement of p adrenoreceptors in the shift of ocular dominance after monocular deprivation. Exp. Bruin Res., 59: 507-514. Kasamatsu, T. and Shirokawa, T. (1988) Norepinephrine-dependent neuronal plasticity in kitten visual cortex. In C.D. Woody, D.L. Alkon and J.L. McGaugh (Eds.), CeNular Mechunlmis of Conditioning and Behaiiorul Plasticity, Plenum, New Yurk, pp. 437-444. Kasamatsu, T., Kiyono, S. and Iwama, K. (1967) Electrical activities of the visual cortex in chronically blinded cats. Tohoku J. Exp. Med., 93: 139-152. Kasamatsu, T., Pettigrew, J.D. and Ary, M. (1979) Restoration of visual cortical plasticity by local microperfusion of norepinephrine. J . Comp. Neurol., 185: 163-182. Kasamatsu, T., Itakura, T. and Jonsson, G. (1981a) Intracortical spread of exogenous catecholamines: Effective concentration for modifying cortical plasticity. J. Phurmacol. Exp. Ther., 217: 841-850. Kasamatsu, T., Pettigrew, J.D. and Ary, M. (1981b) Cortical recovery from effects of monocular deprivation: Acceleration with norepinephrine and suppression with 6-hydroxydopamine. J. Neurophysiol., 45: 254-266. Kasamatsu, T., Itakura, T., Jonsson, G., Heggelund, P., Pettigrew, J.D., Nakai, K., Watabe, K., Kuppermann, B.D. and Ary, M. (1984) Neuronal plasticity in cat visual cortex: A proposed role for the central noradrenaline system. In L. Descarries, T. Reader and H.H. Jasper (Eds.), Monoamine
1nneri:ution of Cerehrul Cortex, Alan R. Liss, New York, pp. 301-319. Kasamatsu, T., Watabe, K., Heggelund, P. and Schiiller, E. (1985) Plasticity in cat visual cortex restored by electrical stimulation of the locus coeruleus. Neurosci. Res., 2: 365386. Kasamatsu, T., Imamura, K., Hartveit, E., Heggelund, P. and Heggelund, U. (1989a) Do NMDA receptors account for cortical plasticity in kittens? ARVO Abstr. h e s t . Ophthalmol. Ks. Sci. Suppl., 30: 377. Kasamatsu, T., Ohashi, T. and Imamura, K. (lYX9b) Integration of adrenergic and cholinergic regulation in ocular dominance plasticity. Proc. 14th Seiriken Conference on “The Neural Basis of Cortical Plasticity,” Biomed. Res. 10 Suppl., 2: 43-53. Kleinschmidt, A,, Bear, M.F. and Singer, W. (1987) Blockade of “NMDA” receptors disrupts experience-dependent plasticity of kitten visual cortex. Science, 238: 3.55-358. Kostrzewa, R.M. and Jacobowitz, D.M. (1974) Pharmacological actions of 6-hydroxydopamine. Pharmacol. Re(,., 26: 199-288. Kuppermann, B.D. and Kasamatsu, T. (1983) Changes in geniculate cell size following brief monocular blockade of retinal activity in kittens. Nature (London), 306: 465-468. Kuppermann, B.D. and Kasamatsu, T. (1984) Enhanced binocular interaction in the visual cortex of nornial kittens subjected to intracorticiil norepinephrine perfusion. Brain Res., 302: 91-99. LeVay, S., Wiesel, T.N. and Hubel, D.H. (1980) The development of ocular dominance columns in normal and visually deprived monkeys. J . Comp. Neurol., 191: 1-51. Madison, D.V. and Nicoll. R.A. (1986) Actions of noradrenaline recorded intracellularly in rat hippocampal C A I pyramidal neurones, in iitro. J. Physiol. (Londord, 372: 221-244. Magni, F. and Willis, W.D. (1964) Cortical control of brain stem reticular neurons. Arch. Ital. Biol., 102: 418-433. Malenka, R.C., Ayoub, G.S. and Nicoll, R.A. (1987) Phorbol esters enhance transmitter release in rat hippocampal slices. Bruin Res., 403: 198-203. Marrocco, R.T., Lane, R.F., McClurkin, J.W., Blaha, C.D. and Alkire, M.F. (1987) Release of cortical catecholamines by visual stimulation requires activity in thalamocortical afferents of monkey and cat. J. Neurosci.. 7: 2756-2767. Mastronarde, D.N. (1983a) Correlated firing of cat retinal ganglion cells. I. Spontaneously active inputs to X- and Y-cells. J. Neurophysiol., 49: 303-324. Mastroriarde, D.N. (1983b) Correlated firing of cat retinal ganglion cells. 11. Responses of X- and Y-cells to single quanta1 events. J. Nruroptiysiol., 49: 325-349. McCormick, D.A. and Prince, D.A. (1988) Noradrenergic modulation of firing pattern in guinea pig and cat thalamic neurons, in iiitro. J. Neurophysiol., 59: 978-996. Meyer, R.L. (1983) Tetrodotoxin inhibits the formation of refined retinotopography in goldfish. Dei,. Brain Res., 6: 293-298. Miller, K.D., Chapman, 8. and Stryker, M.P. (1989) Visual responses in adult cat visual cortex depend on N-methyl-
615
D-aspartate receptors. Proc. Natl. Acad. Sci. USA, 86:
5 1X3 -5 187. Movshon, J.A. and Van Sluyters, R.C. (1981) Visual neural development. Ann. Rec. Psychol., 32: 477-522. Mower, G.D., Berry, D., Burchfiel, J.L. and Duffy, F.H. (1981) Comparison of the effects of dark-rearing and binocular suture on development and plasticity of cat visual cortex. Bruin Ra., 220: 255-267. Mower, G.D., Caplan, C.J., Christen, W.G. and Duffy, F.H. (1985) Dark rearing prolongs physiological but not anatomical plasticity of the cat visual cortex. J. Comp. Neurol., 235: 448-466. Mower, G.D., White, W.F. and Rustad, R. (1986) ['HI Muscimol binding of GABA receptors in the visual cortex of normal and monocularly deprived cats. Brain Res., 380: 253-260. Nakai. K. (1987) Regenerative catecholamine-containing terminals in kitten visual cortex: An ultrastructural study. Neurosci. Res., 4: 475-485. Nakai, K., Jonsson, G. and Kasamatsu, T. (1987) Norepinephrinergic reinnervation of cat occipital cortex following localized lesions with 6-hydroxydopamine. Neurosci. Res., 4: 433-453. Nichols, R.A., Haycock, J.W., Wang, J.K.T. and Greengard, P. (1987) Phorbol ester enhancement of neurotransmitter release from rat brain synaptosomes. J. Neurochem., 48: 615-621. Nichols, R.A., Sihra, T.S., Czernik, A.J., Nairn, A.C. and Greengard, P. (1990) Calcium/calmodulin-dependent protein kinase I1 increases glutamate and noradrenaline release from synaptosomes. Nature (London), 343: 647-651. Ohashi, T. and Kasamatsu, T. (1989) Concentration-dependent suppression by lithium of effects on monocular deprivation in kittens. Soc. Neurosci. Abstr., 15: 795. Olson, C.R. and Freeman, R.D. (1975) Progressive changes in kitten visual cortex during monocular vision. J. Neurophysiol.? 38: 26-32. Olson, C.R. and Freeman, R.D. (1980) Profile of the sensitive period for monocular deprivation in kittens. Exp. Bruin Res., 39: 17-21. Paradiso, M.A., Bear, M.F. and Daniels, J.D. (1983) Effects of intracortical infusion of 6-hydroxydopamine on the response of kitten visual cortex to monocular deprivation. Exp. Bruin Rex, 51: 413-422. Pettigrew, J.D. and Kasamatsu, T. (1978) Local perfusion of noradrenaline maintains visual cortical plasticity. Nature (London), 271: 761-763. Ramoa, A.S., Paradiso, M.A. and Freeman, R.D. (1988) Blockade of intracortical inhibition in kitten striate cortex: Effects on receptive field properties and associated loss of ocular dominance plasticity. Exp. Bruin Rrs., 73: 285-296. Rauschecker, J.P. (1987) What signals are responsible for synaptic changes in visual cortical plasticity. In J.P. Rauschecker and P. Maler (Eds.), Imprinting and Cortical Plasticiw, C0mpuratii.e Aspects of Sensitiue Periods, John Wiley & Sons, New York, pp. 193-220. Rauschecker, J.P. and Hahn, S. (1987) Ketamine-xylazine anesthesia blocks consolidation of ocular dominance
changes in kitten visual cortex. Nuture (London), 326: 183- 185. Reiter, H.O. and Stryker, M.P. (1988) Neural plasticity without postsynaptic action potentials: Less-active inputs become dominant when kitten visual cortical cells are pharmacologically inhibited. Proc. Nutl. Acud. Sci. USA, 85: 3623-3627. Reiter, H.O., Waitzman, D.M. and Stryker, M.P. (1986) Cortical activity blockade prevents ocular dominance plasticity in the kitten visual cortex. Exp. Brain Res., 65: 182-188. Riccio, R.V. and Matthews, M.A. (1985a) Effects of intraocular tetrodotoxin on dendritic spines in the developing rat visual cortex: A Golgi analysis. Deu. Bruin Res., 19: 173182. Riccio, R.V. and Matthews, M.A. (198%) T h e postnatal development of the rat primary visual cortex during optic nerve impulse blockade by intraocular tetrodotoxin: A quantitative electron microscopic analysis. Deu. Brain Rex, 20: 55-68. Robertson, R.T., Ambe, R.K. and Yu, J. (1989) Intraocular injections of tetrodotoxin reduce transiently expressed acetylcholinesterase activity in developing rat visual cortex. Der,. Brain Res., 46: 69-84. Romo, R., Cheramy, A,, Godeheu, G. and Glowinski, J. (1986) In uiuo presynaptic control of dopamine release in cat caudate nucleus. I. Opposite changes in neuronal activity and release evoked from thalamic motor nuclei. Neuroscience, 19: 1067-1079. Schmidt, J.T. and Edwards, D.L. (1983) Activity sharpens the map during regeneration of the retinotectal projection in gold fish. Bruin Res., 269: 29-39. Shatz, C.J. and Stryker, M.P. (1978) Ocular dominance in layer 1V of the cat's visual cortex and the effects of monocular deprivation. J. Physiol. (London), 281: 267-283. Shaw, C. and Cynader, M. (1984) Disruption of cortical activity prevents ocular dominance changes in monocularly deprived kittens. Nature (London), 308: 731-734. Shaw, C. and Cynader, M. (1988) Unilateral eyelid suture increases GABA, receptors in cat visual cortex. Deu. Bruin Res., 40: 148-153. Shaw, C., Needler, M.C. and Cynader, M. (1984) Ontogenesis of muscarinic acetylcholine binding sites in cat visual cortex: Reversal of specific laminar distribution during the critical period. Deu. Brain Rex, 14: 295-299. Shaw, C., Wilkinson, W., Cynader, M., Needler, M.C., Aoki, C. and Hall, S.E. (1986) The laminar distributions and postnatal development of neurotransmitter and neuromodulator receptors in cat visual cortex. Bruin Res. Bull., 16: 661-671. Sherman, S.M. and Spear, P.D. (1982) Organization of visual pathways in normal and visually deprived cats. Psychol. Reu., 62: 738-855. Shirokawa, T. and Kasamatsu, T. (1986) Concentration-dependent suppression by P-adrenergic antagonists of the shift in ocular dominance following monocular deprivation in kitten visual cortex. Neuroscience, 18: 1035-1046. Shirokawa, T. and Kasamatsu, T. (1987) Reemergence of ocular dominance plasticity during recovery from the ef-
616 fects of propranolol infused in kitten visual cortex. Exp. Brain Res., 68: 466-476. Shirokawa, T., Kasamatsu, T., Kuppermann, B.D. and Ramachandran, V.S. (1989) Noradrenergic control of ocular dominance plasticity in the visual cortex of dark-reared cats. Deo. Brain Res., 47: 303-308. Shuntoh, H. and Tanaka, C. (1986) Activation of protein kinase C potentiates norepinephrine release from sinus node. Am. J. Physiol., 251: C833-C840. Sillito, A.M., Kemp, J.A. and Blakemore, C. (1981) The role of GABAergic inhibition in the cortical effects of monocular deprivation. Nature (London), 291: 318-320. Singer, W. (1982) Central-core control of developmental plasticity in the kitten visual cortex: I. Diencephalic lesions. Exp. Brain Res., 47: 209-222. Singer, W. (1984) Learning to see: Mechanisms in experience-dependent development. In P.R. Marler and H.S. Terrace (Eds.), The Biology of Learning, Dahlem Konferenzen, Springer-Verlag, Berlin, New York, pp. 461477 Singer, W. and Lux, H.D. (1975) Potassium gradients in receptive fields of cat visual cortex. Brain Res., 96: 378-383. Singer, W. and Rauschecker, J.P. (1982) Central core control of developmental plasticity in the kitten visual cortex: 11. Electrical activation of mesencephalic and diencephalic projections. Exp. Brain Res., 47: 223-233 Singer, W., von Gruenau, M. and Rauschecker, J. (1979a) Requirements for the disruption of binocularity in the visual cortex of strabismic kittens. Brain Res., 171: 536-540. Singer, W., Yinon, U. and Tretter, F. (1979b) Inverted monocular vision prevents ocular dominance shift in kittens and impairs the functional state of visual cortex in adult cats. Brain Res., 164: 294-299. Skinner, J.E. and Molnar, M. (1983) Event-related extracellular potassium ion activity changes in frontal cortex of the conscious cat. J. Neurophysiol., 49: 204-215. Stanton, P.K., Mody, I. and Heinemann, U. (1989) A role for N-methyl-D-aspartate receptors in norepinephrine-induced long-lasting potentiation in the dentate gyms. Exp. Brain Res., 77: 517-530. Stricker, E.M. and Zigmond, M.J. (1986) Brain monoamines, homeostasis, and adaptive behavior. In V.B. Mountcastle, F.E. Bloom and S.R. Geiger (Eds.), Handbook of Physiology: The Nervous System IV, Bethesda, Am. Physiol. SOC., pp. 677-700.
Stryker, M.P. (1989) Evidence for a possible role of spontaneous electrical activity in the development of the mammalian visual cortex. In P. Kellaway and J.L. Noebels (Eds.), Problems and Concepts in Developmental Neurophysiology, Johns Hopkins Univ. Press, Baltimore, pp. 110-130. Stryker, M.P. and Harris, W.A. (1986) Binocular impulse blockade prevents the formation of ocular dominance columns in cat visual cortex. J. Neurosci., 6: 21 17-2133. Swindale, N.V. (1988) Role of visual experience in promoting segregation of eye dominance patches in the visual cortex of the cat. J. Comp. Neurol., 267: 472-488. Trombley, P., Allen, E.E., Soyke, J., Blaha, C.D., Lane, R.F. and Gordon, B. (1986) Doses of 6-hydroxydopamine sufficient to deplete norepinephrine are not sufficient to decrease plasticity in the visual cortex. J. Neurosci., 6: 266273. Tsumoto, T., Hagihara, K., Sato, H. and Hata, Y . (1987) NMDA receptors in the visual cortex of young kittens are more effective than those of adult cats. Nature (London), 327: 513-514. Van Essen, D.C. (1982) Neuromuscular synapse elimination. Structural, functional, and mechanistic aspects. In N.C. Spitzer (Ed.), Neuronal Decelopment, Plenum, New York and London, pp. 333-376. Videen, T.O., Daw, N.W. and Rader, R.K. (1984) The effect of norepinephrine on visual cortical neurons in kitten and adult cats. J. Neurosci., 4: 1607-1617. Watabe, K., Nakai, K. and Kasamatsu, T. (1982) Visual afferents to norepinephrine-containing neurons in cat locus coeruleus. Exp. Brain Rex, 48: 66-80 Wiesel, T.N. (1982) Postnatal development of the visual cortex and the influence of environment. Nature (London), 299: 583-591. Wiesel, T.N. and Hubel, D.H. (1963) Single-cell responses in striate cortex of kittens deprived of vision in one eye. J. Neurophysiol., 26: 1003- 1017. Wiesel, T.N. and Hubel, D.H. (1965) Comparison of the effects of unilateral and bilateral eye closure on cortical unit responses in kitten. J. Neurophysiol., 28: 1029-1040. Wigstrom, H. and Gustaffson, B. (1984) A possible correlate of the postsynaptic condition for long-lasting potentiation in the guinea pig hippocampus in vitro. Neurosci. Lett., 44: 327-332.
C.D. Barnes and 0. Pompeiano (Eda.) Pro&ve.?s in Brain Research, Vol. 88 0 1991 Elsevier Science Publishrrs B.V.
617 CHAPTER 43
Regulation of the development of locus coeruleus neurons in uitro L. Sklair and M. Segal Center for Neuroscience, Weizmann Institute, Rehoi?ot, Israel
Primary dissociated cultures of embryonic locus coeruleus
(LC) neurons were established. Norepinephrine (NE) uptake was used as an index of maturation of the noradrenergic (NA) neurons from the LC. When LC cells were cocultured with a low density of hippocampal target cells, NE uptake was stimulated. However, increasing the concentration of hippocampal cells resulted in a significant decrease in NE uptake. The target stimulatory effect was mediated by both neurons and glia, whereas the inhibitory effect was mediated by direct
~
~
contact between target glia and LC neurons and detected only i n the presence of serum. In addition to the effect of target, we also tested the effect of elevated intracellular cAMP levels on NE uptake versus GABA uptake. GABA uptake was an index of development of non-NA cells. Increasing intracellular CAMP resulted in selective stimulation of N E uptake. These studies illustrate the potential of dissociated LC cultures in studying the regulation of NA axonal outgrowth.
~
Key words: locus coeruleus, culture, development, target regulation, cAMP
Introduction
The central noradrenergic (NA) neurons of the locus coeruleus (LC) are extensively collateralized, innervating wide areas of the neuroaxis, from the olfactory bulb to the spinal cord (for reviews see Amaral and Sinnamon, 1977; Foote et al., 1983). The LC perikarya appear very early during ontogeny, and the general outline of axonal systems is present at birth. The postnatal development of the LC system consists mainly of proliferation and maturation of the nerve terminal network in the innervation regions. The developing NA neurons are capable of striking plas-
ticity in response to environmental modifications or injury. NA plasticity is demonstrated by the effects of 6-hydroxydopamine (6-OHDA) administration, in neonates, on the distribution of terminal plexuses of LC neurons (Jonsson et al., 1974; Jonsson and Sachs, 1976). This treatment appears to preferentially alter the development of LC neurons and leads to an almost complete and permanent NA denervation of the cerebral cortex and spinal cord with a concomitant NA hyperinnervation of the nearby regions, e.g., of the pons-medulla and cerebellum. Since these studies indicated a close relationship between 6-OHDAinduced denervation and the hyperinnervation
618
(Sachs and Jonsson, 1975), it has been suggested that the effect of the neonatal 6-OHDA treatment is due to a “pruning response”; i e . , the permanent denervation in some areas of the brain (neocortex and spinal cord) leads to a compensatory outgrowth of terminals of intact collaterals resulting in a hyperinnervation (in the ponsmedulla and cerebellum), reflecting a strict developmental program of NA neurons with respect to expression of a certain quantity of nerve terminal arborization (Jonsson and Sachs, 1976). However, in light of the specificity with which the hyperinnervation occurs, it was suggested that the target structures may play a major role in controlling the pattern of innervation (Schmidt and Bhatnagar, 1979; Gustafson and Moore, 1987). Even in adult animals, there is considerable evidence that the LC efferents are particularly plastic and can regenerate. Regrowth of LC axons into a target tissue after denervation is dramatic. Damage to a portion of the LC projection to the cerebellum triggers an increase in NA terminals that exceeds the original innervation (Pickel et al., 1973). The same damage also causes a proliferation of LC terminals in the hippocampus (Pickel et al., 1974). A similar expansion of LC fibers in the lateral geniculate nucleus has been described after visual cortex ablation (Sakaguchi et al., 1984), and septa1 lesion induced a significant increase in LC innervation of the cerebellum (Harring et al., 1986). The sprouting of LC axons innervating targets remote to the lesioned pathway has been attributed to the collateralization of intact fibers. The induction of collateral expansion is similar to the “pruning” response, which is an inherent characteristic of the damaged neuron and is a target-independent event, whereas the final connectivity and density of reinnervation are regulated by target neurons. NA innervation of the cerebral cortex is an example of a specific, probably target-regulated, pattern of innervation. NA fibers are observed in most areas of the cerebral cortex; however, a typical heterogeneous distribution of fibers can be defined with only a few obliquely oriented
fibers in layer V, and the highest density of NA fibers oriented in a distinct band is found in layer I (Fuxe et al., 1968; Levitt and Moore, 1978). Another aspect of the plasticity of the NA system is the adaptive changes in NA receptors in response to NA denervation or hyperinnervation. Lesion studies allow us to examine the response of adrenergic binding sites to different types of deafferentation. NA denervation of the cerebral cortex by 6-OHDA injection induced an increased density of P-receptor binding sites as well as an elevated ability of isoproterenol to stimulate adenylate cyclase (Sporn et al., 1976; Skolnick et al., 1978). The supersensitivity of p receptors following norepinephrine (NE) depletion is not confined to the cerebral cortex. U’Prichard et al. (1979) found that after 6-OHDA lesions there was an increase in the density of P-receptors both in the hippocampus and the thalamus. In the cerebellum as well as in other target organs (Battisti et al., 1989) &binding sites exhibit both upregulation and downreguIation which is correlated with the density of the NA projection. The sensitivity of a-adrenergic receptors can also be affected by altering NE levels in some brain areas. Reciprocal changes in a,-receptor sensitivity are induced in a similar manner as observed for @-receptors (U’Prichard et al., 1979) Biochemical studies on developing cells from the LC explants have been carried out (Schlumpf et al., 1977) and electrophysiological techniques have allowed their differentiation to be followed in the presence of hippocampal neurons, one of their target cells (Dreyfus et al., 1979). Differentiation of NA neurons was facilitated by coculturing them with target cells taken from embryonic cerebellum (Di Porzio and Estenoz, 1984). However, the signals that regulate the expression of the unique plastic properties of LC cells are still unknown. Primary culture of LC neurons is a reliable tool for studying the cellular mechanisms underlying the enormous capacity of NA ncurite outgrowth. In the complex milieu of the CNS numerous signals can influence neuronal out-
619
growth and survival. The most intensively studied of these signals have been the growth factors (Levi-Montalcini and Angeletti, 1968; Berg, 1984) and extracellular matrix molecules (Letourneau, 1985). In addition, recent studies have demonstrated that neurotransmitters and electrical activity can regulate, in quite specific ways, neuronal outgrowth and survival during development (for review see Mattson, 1988). The cellular mechanisms by which primary signals including hormones, growth factors and neurotransmitters exert their biological activity involve second messengers (Schubert et al., 1978; Kupferman, 1980; Rasmussen, 1981). Recent studies have indicated that calcium, cyclic nucleotides, and inositol phospholipid-derived second messengers can mediate neurotransmitter effects on neurite outgrowth (Mattson, 1988). We used the in citro approach to identify some of the signals that regulate the NA neuronal arborization. In the present study, dissociated LC neurons were cultured in a micro-culture system of 96-well plates that enabled immunocytochernical staining of LC neurons and determination of 13H]NE uptake. 13H]NE uptake is an index of the density of NA nerve terminals and neuronal maturation (Jonsson and Sachs, 1972; Coyle and Henry, 1973). We report here that target-derived factors as well as the CAMP generating system can regulate the development of NA neurite arborization.
Target regulation of 13H1NE uptake Dissociated LC cells from 14-day-old rat embryos (El41 were grown alone or in the presence of E18- 19 hippocampal target cells in serum-containing medium for 7 days. The development of NA neurons was monitored by their ability to take up [ 'H]NE. Hippocampal target cells were cocultured with LC cells to determine their ability to affect the uptake of [H'INE by LC cells. Stimulation of [3H]NE uptake was observed only with a low number of hippocampal cells cocultured with LC cells. However, when LC cells were cocultured with increasing amounts of hip-
i/
2 1 2 a
500
3
-t
04
08
12
16
L c Cells x 1FVwell
Fig. 1. The effect of cocultures of hippocampal target and LC afferent cells on ['HINE uptake. Cells were cultured for 7 days in serum containing medium. El5 LC cells were cultured alone ( 0 ) at increasing plating densities, and in the presence of 0.65X 10' ( 0 ) or 1 . 3 X 1 0 5 ( A ) El8 hippocampal cells. Each point is the meankS.E.M. of at least three separate cultures. Statistical comparison with LC-alone cultures, Student's t-test: * P < 0.05. (From Sklair and Segal, 1990.)
pocampal cells, a significant decrease in 13H]NE uptake was observed (Fig. 1). T o test the possibility that a high density of target cells influences survival of NA cells, the number of NA neurons was quantified in sister LC cultures. The number of NA cells in the cultures was counted following immunocytochemical staining using tyrosine hydroxylase (TH) antibody (Fig. 2). Under all plating densities the number of TH-positive cells in cocultures did not differ from the number in LC cultures. Thus, the reduction in [ 'H]NE uptake in cocultures was due to a decrease in the uptake ability of individual NA cells. Inhibitory effects of target cells, expressed at the high plating densities of afferent LC and hippocampal neurons, suggested the existence of density-derived neurotoxic activity (Grau-Wagemans et al., 1984) unrelated to the hippocampal target tissue. However, we have shown that, by itself, high-plating density in LC cells was not as effective as the hippocampal target cells in inhibiting [3H]NE uptake. Thus the possibility of a non-specific, density-derived toxic effect was ruled out.
620
Fig. 2. Tyrosine hydroxylase (TH)-positive cell in LC cultures. Cells were cultured for 7 days in serum-containing medium and then stained with a TH antibody. This is a typical NA cell with extensive neurites spread all around the culture dish. The cell body and processes are covered with many fine, wispy extensions. Calibration bar = 20 pM.
The role of glial cells in mediating the target inhibitory effect was examined using several complementary approaches. To test the direct effect of glial cells, we maintained the cocultures in a serum-free medium which favored neuronal, but not glial, survival. Under these culture conditions the stimulatory target effect was evident, but the inhibitory effect was absent. Therefore, the target stimulatory effect was not dependent on the presence of glial cells, whereas for the expression of target inhibitory effect glial cells were required. It could well be, however, that the loss of the inhibitory effect in serum-free medium might have been due to the absence of a serum-inhibitory
factor. The mediator of the glial-inhibitory effect could be a soluble molecule or intercellular contact. To test the possibility that a soluble factor was involved, conditioned medium (CM) from hippocampal glial cultures was applied to LC cultures. Glial CM stimulated ['HINE uptake up to 2 or 3-fold. Thus, the target-inhibitory effect was not mediated by glial soluble factors. The protein nature of the active factor in the CM is suggested on the basis of heat treatment results which eliminated all CM activity. These results suggest that target glial cells played a role in stimulating a [ 'HINE uptake system of afferent cells. Since glial CM had no inhibitory effect, the remaining possibilities were that the target-inhibitory effect could be mediated either by cell contact between target glia and afferent neurons, or via a serum factor. To examine the involvement of intercellular contact and the role of serum factors in the target-inhibitory effect, LC cells were plated on a feeder-layer of hippocampal glia, rather than on poly-L-lysine, in the presence or in the absence of serum. The glial feeder-layer exerted a pronounced inhibitory effect on NE uptake in serum-containing medium. In serum-free medium there was no decrease in NE uptake of LC cells plated on glia as compared to cells plated on poly-L-lysine. Therefore, serum was required for the expression of a glial inhibitory effect. A full account of these results was recently published (Sklair and Segal, 1990). The effect of the intracellular CAMP level on 13HINE uptake versus 13H1GABA uptake
In LC cultures NA cells differ strikingly from the other neurons in their extensive neuritic arborization (Fig. 2). These morphological differences are reflected in the uptake ability of the two neuronal subpopulations in LC cultures. L3H]NE and [3H]GABA uptake were measured in sister LC cultures, and the amount of neurotransmitters accumulated by individual cells was calculated; a single NA cell accumulates 0.46 fmol NE, whereas the non-NA neuron accumulates only 0.053 fmol.
339
D GABAlOnA
A Control 11"x176" Vert bar
Vert. b a r
... ...-..-..-_--.....-.-."r" ..-.. ..... ............................ .......................... .:..:. ......................................... ..... .-. ........... .... - ..... .......... --. ....... -..-.- -.-. .......... ...................... ............ ...................................... ................... .-... ........ .................. -. . ............... ......... . ..- -" -..-.-. -.. .. _....-.................... .................................... ........... -. .................................. ...-..-......-..................... .......................................... ................................ ................................. - . . I .
-0.5 0
6.5
Seconds
Norepinephrine lOnA
.................. --..-........... -. . . ". . . .. ... ........-. .-..-..-. .. -...... ............... ............ ...... . .-....... ................. ---. . . .. ... .. . ...-..... ..-._.-_. ..-................ . ........... . . . . . .... ............. ..---.. - -.... .. . ..-.-.. ... . ..... .... . ......... .. .... ._..-. .. ._ -.. .-.. .......... --........ .......... ...........--.... . . .. ......... .... ._. .......... .... .. . . . . . . . . . .. . .--. . . -.........-. ......-.-. .-. ........... .... .....- . .. ... . .. .. . ..
-0.5
0
6.5
Seconds
Recovery
U
Spikes/s
. . C
6.5
Norepinephrine lOnA + Sotalol 25nA
s
Temporal
B I N : 20 m s
-05 0
t
90
Spikes/s ia.4 -
:
45
.
L
4
45
TempoNral
90
-45
Seconds
6.5
Fig. 6. Comparison of the effects of iontophoresis of N E and GABA on visually evoked and spontaneous simple spike discharge of a Purkinje cell. A. This neuron was stimulated by movement of a vertical bar of light (11" X 176") from right to left (upslope of trapezoidal waveform) and in the reverse (preferred) direction through its receptive field. B. During iontophoresis of N E 10 nA, spontaneous discharge was markedly reduced, whereas activity evoked by stimulus movement in the preferred direction (left-to-right) was only slightly decreased, resulting in the enhancement of the ratio of signal to noise. Note that N E had the additional effect of enhancing the cell's excitatory response to movement of the stimulus in the non-preferred direction (from right to left). C . This enhancement of the visual responses by NE was blocked when the specific P-adrenergic antagonist sotalol 25 n A was applied concurrently to the cell by iontophoresis. Note that sotalol antagonized both the depression in spontaneous activity by N E and the enhancement in the response to right-to-left movement of the visual stimulus. Drug iontophoresis was suspended for 20 min following the generation of this histogram record to allow for complete recovery to the control levels of visually evoked and spontaneous activity before further testing with GABA was initiated. D. In contrast to the effects observed during NE, iontophoresis of GABA 10 nA produced a smaller suppression in background firing and yet virtually eliminated the responses to visual stimulation. Recovery of the visually evoked responses to near control levels was observed following the cessation of N E ( C ) and GABA iontophoresis (E). F. The receptive field location of the excitatory zone (shading) and forward movement trajectory (arrnw) of the lieht bar stimiilus. (From Moises Pt nl.. 1990.)
622
Several studies have dealt with the involvement of cAMP in the regulation of neuronal outgrowth and synaptogenesis (Mattson, 19881. In different neuronal types, cAMP has been found either to enhance or to inhibit neurite outgrowth. Increasing intracellular cAMP levels enhanced neurite extension in P12 cells (Schubert et al., 19781, cortical cells (Shapiro, 1973) and in neuroblastoma cell line (Nirenberg and Wilson, 1984). However, in Helisoma elevations in cAMP inhibited neurite elongation (Mattson et al., 19881. PKC activation by phorbol esters was found to promote neurite outgrowth in sensory ganglion neurons and neuroblastoma cells (Mattson, 19881. In addition to cAMP and PKC there is now considerable evidence to suggest a central role for Ca in the regulation of neurite outgrowth and in the control of neuronal cytoarchitecture (Mattson et al., 1988). Additional findings show that the different second messengers can act synergistically or antagonistically in developmental processes. Thus, although there is considerable evidence on the role of second messengers in regulation of neuronal cytoarchitecture, the primary signals which trigger the intracellular events are yet to be explored. We are currently proceeding with our studies on the LC cells along these lines. References Amaral, D.G. and Sinnamon, H.M. (1977) The locus coeruleus: Neurobiology of a central noradrenergic nucleus. Prog. Neurohiol., 9: 147-196. Azmitia, E.C. and Whitaker-Azmitia, P.M. (1987) Target cell stimulation of dissociated serotonergic neurons in culture. Neuroscience, 20: 47-63. Battisti, W.P., Artymyshyn, R.P. and Murray, M. (1989) 01and P2-adrenergic '251-Pindolol binding sites in the interpeduncular nucleus of the rat: Normal distribution and the effects of deafferentation. J. Neurosci., 9: 2509-2518. Berg, D.K. (1984) New neuronal growth factors. Ann. Rev. Neurosci., 7: 149- 170. Coyle, J. and Henry, D. (1973) Catecholamine in fetal newborn rat brain. J. Neurosci., 21: 61-67. Denis-Donini, S., Glowinski, J. and Prochiantz, A. (1983) Specific influence of striatal target neurons on the in v i m outgrowth of mesencephalic dopaminergic neurites: A morphological quantitative study. J. Neurosci., 3: 22922299.
Di Porzio, U. and Estenoz. M. (1984) Positive control of target cerebellar cells on norepinephrine uptake in embryonic brainstem cultures in serum free medium. Da.. Brain Rex, 16: 147-157. Di Porzio, U., Daguet, M.C., Glowinski. J., and Prochiantz, A. (1980) Effect of striatal cells on in r,itro maturation of mesencephalic dopaminergic neurons grown in serum-free conditions. Nature (London), 288: 370-373. Dreyfus, C.F., Gershon, M.D. and Crain, S.M. (1979) Innervation of hippocampal explants by central catecholaminergic neurons in co-cultured fetal mouse brain stem explants. Bruin Res., 161: 431-445. Foote, S.L., Bloom, F.E. and Aston-Jones, G. (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. Physiol. Rer., 63: 844-914. Fuxe, K., Hamberger, B. and Hokfelt, T. (1968) Distribution of noradrenaline nerve terminals in cortical areas of the rat. Bruin Rex, 8: 125-131. Goldowitz, P., Seiger, A. and Olson, L. (1984) Degree of hyperinnervation of area dentata by locus coeruleus in the presence of septum and entorhinal cortex as studied by sequential intraocular triple transplantation. Exp. Brain Rex, 56: 351-360. Grau-Wagemans, M.P., Selak, I., Lefehvre, P.P. and Moonen, G. (1984) Cerebellar macroneurons in serum-free cultures: Evidence for intrinsic neuronotrophic and neuronotoxic activities. Dec. Bruin Res., 15: 11-19. Gustafson, E.L. and Moore, R.Y. (1987) Noradrenaline neuron plasticity in developing a brain: Effects of neonatal 6-hydroxydopamine demonstrated by dopamine-p-hydroxylase immunocytochemistry. DeL. Bruin Rex, 37: 143-155. Haring, J.H., Miller, G.D. and Davis, J.N. (1986) Changes in the noradrenergic innervation of the area dentata after axotomy of coeruleohippocampal projections or unilateral lesion the locus coeruleus. Bruin Res., 368: 233-238. Jonsson, G. and Sachs, C. (1972) Neurochemical properties of adrenergic nerve regeneration after 6-hydroxydopamine. J. Neurochem., 19: 2557-2585. Jonsson, G., and Sachs, C. (1976) Regional changes in [3H]noradrenaline uptake, catecholamines and catecholamine synthetic and catabolic enzymes in rat brain following neonatal 6-hydroxydopamine treatment. Med. Biol., 54: 286-297. Jonsson, G., Pycock, C. and Fuxe, K. (1974) Changes in the development of central noradrenaline neurons following neonatal administration of 6-hydroxydopamine. J. Neurochem., 22: 621-626. Kater, S.B. (1985) Dynamic regulators of neuronal form and connectivity in the adult snail Helisoma. In J.A. Selverston (Eds.), Model Neural Networks and Behar,ior, Plenum, New York, 548 pp. Kupferman, I. (1980) The role of cyclic nucleotides in excitable cells. Ann. Rec. Physiol., 4 2 629-641. Letourneau, P.C. (1985) Axonal growth and guidance. I n G.M. Edelman, W.E. Gall and M. Cowan (Eds.), Molecular Bases of Neural Development, Wiley, New-York, pp. 269-294.
623
Levi-Montalcini, R. and Angeletti, P.U. (1968) Nerve growth factor. Physiol. Re[..,48: 534-569. Levitt. P. and Moore, R.Y. (1978) Noradrenaline neuron innervation of the cerebral cortex in the rat. Brain RPS.. 139: 219-231. Mattson. M.P. (198X) Neurotransmitters in the regulation of neuronal cytoarchitecture. Brain Re.s. Re!,., 13: 179-212. Mattson. M.P., Taylor-Hunter, A. and Kater, S.B. (1988) Neurite outgrowth in individual neurons of neuronal population is differentially regulated by calcium and cyclic AMP. J. Neurosci., 8: 1704-1711. Nirenberg, M. and Wilson, S. (1984) Modulation of synapse formation by cyclic adenosine monophosphate. Science. 222: 194-790. Pickel. V.M., Krebs. H. and Bloom, E.F. (1973) Proliferation of norepinephrine-containing axoiis in rat cerebellar cortex after peduncle lesion. Bruin Res., 59: 169-179. Pickcl, V.M., Segal, M. and Bloom, F.E. (1974) Axonal proliferation following lesions of cerebellar peduncles. A comhincd fluorescence microscopic and autoradiographic study. J. Comp Ncwol., 155: 43-60. Prochiantz. A., Di-Porzio, U., Katu, A,, Bergen, B. and Glouinski, J. (1979) In r,itro maturation of mesencephalic dopaminergic neurons from mouse embryos is enhanced in the presence of their target cells. Proc. Natl. Acarf. Sci., USA, 76: 538775391, Prochimtz. A,, Daguet, M.-C., Herbert, A. and Glowinsky, J. ( l 9 f l f Specific stimulation of in r.itru maturation of mesencephalic dopaminergic neurons by striatal membranes. Nutitre (London), 293: 570-572. Rasniussen,JX. (1981) Calcrum irnd CAMP u s Synarchic Messengers, Wiley, New York, 370 pp. Sachs, C. and Jonsson G. (1975) Effects of h-hydroxydopamine on central noradrenaline neurons during ontogeny. Bruin Res., 99: 277-291. Sakaguchi, T., Shirokawa, T. and Nakamura, S. (1984) Changes i n the projection from locus coeruleus to lateral geniculate nucleus following ablation of visual cortex in adult rats.
Schlumpf, M., Shoemaker, W.J. and Bloom. F.E. (1977) Explant cultures of catecholamine-containing neurons from rat brain: Biochemical, histofluorescent and electromicroscopic studies. Proc. Null. Acud. Sci., USA, 74: 4471-4475. Schmidt, R.M. and Bhatnagar, R.K. (1979) Assessment of the effects of neonatal subcutaneous 6-hydroxydopamine on noradrenergic and dopaminergic innervation of the cerebral cortex. Bruin Rex, 166: 309-319. Schubert, D., La Corbiere, M., Whiltlock, C. and Stullcup, W. (1978) Alterations in the surface properties of cells responsive to nerve growth factor. Nature (London), 273: 718-723.
Shapiro, D.L. (1973) Morphological and biochemical alterations in foetal brain cells cultured in the presence of monobutyiyl cyclic AMP. Nature (London), 241: 203-204. Sklair, L. and Segal, M. (1990) Target cell stimulation and inhibition of norepinephrine uptake in dissociated rat locus coeruleus cultures. Dei,. Bruit2 Res., 52: 191-199. Skolnick, P., Stalvey, L.P., Daly, J.W., Hoyler, E. and Davis, J.N. (1978) Binding of cy and p adrenergic ligands to cerehral cortical membranes: Effect of 6-hydroxydopaniine treatment and relationship to the responsiveness of cyclicAMP generating systems in two rat strains. Eur. J. Phurmacol., 47: 201-210. Sporn, J.. Harden, T., Wolf, B. and Molinoff, P.B. (1976) Beta-adrenergic receptor involvement in h-hydroxydopamine induced supersensitivity in rat cerebral cortex. Scic~nce,194: 624-626. U’Prichard, P., Reisine, T., Mason, S., Fibiger, II. and Yamaniura, 1 i . I . (1979) Modulation of rat brain 01 and p adrenergic receptor populations by lesion of the dorsal noradrenergic bundle. Brain Rex, 187: 142-153. Whitaker-Azmitia, P.M. and Azmitia, E.C. (1989) Stimulation of astroglial serotonin receptors produces culture media which regulates growth of serotonergic neurons. Brum Res.. 497: 80-85.
This Page Intentionally Left Blank
C.D Barnes and 0. Pompeiano (Eds.) Propress in Bruin Research, Vol. 88 0 1991 Elsevier Science Publishers B.V
625 CHAPTER 44
Alterations in the locus coeruleus in dementias of Alzheimer’s and Parkinson’s disease V. Chan-Palay Neurology Clinic, Unicersity Hospital, Frauenkhnikstrusse, Zurich, Switzerland
For diagnostic purposes, a differentiation can be made between the locus coeruleus (LC) in normal brain and the LC, in senile dementia of the Alzheimer’s type (SDAT) and Parkinson’s disease (PD). The differentiation is based on
findings concerning the morphological alterations of the TH-immunoreactive; neurons, on the topographical distribution o f neuron loss within the length of the LC, and on the total reduction in cell number.
Key words: catecholamine, immunocytochemistIy, galanin neuropeptides, plasticity, norepinephrine
Introduction
Senile dementia of the Alzheimer’s type (SDAT) is neuropathologically characterized by severe cortical atrophy and cell loss as well as a high index of dementia as measured by numbers of neurofibrillary tangles (NFT) and neuritic plaques (NP) in neocortex and hippocampus. In addition, several subcortical afferent projection systems are disturbed in the disease, namely those based on acetylcholine, norepinephrine (NE) and serotonin. The occurrence of extrapyramidal signs in SDAT suggest involvement of dopaminergic pathways in some cases. Investigations of the functional role of the locus coeruleus ( L O N E system in SDAT have previously focused on the study of the LC cellular neuropathology and measurements of N E content in the various cortical projection areas of the LC. Quantitative investi-
gations using the neuromelanin pigment as a marker for NE neurons have demonstrated a reduction of neuron numbers in the LC in most cases of SDAT with a high incidence of neuropathological markers like NFT, NP and, occasionally, Lewy bodies in the remaining neurons. Recent studies using catecholamine biosynthetic enzymes have also demonstrated a loss of LC-NE neurons in SDAT, though with different results as to the total neuron numbers counted in control and SDAT cases. This cell loss from the LC was reported to be topographically arranged. NE level, doparnine-P-hydroxylase (DPH)-activity and the levels of several other NE markers have been shown to be decreased in LC projection areas, both in ante-mortem biopsies and in postmortem brain tissue, indicating a deficiency of the NEtransmission in SDAT. Correlation between cortical plaque formation and cortical NE levels and
626
LC neuron loss in the anterior and central regions of the LC known to project to these areas in animals have been reported. Also, the severity of cortical plaque incidence and the degree of reduction of LC neuron number has been observed to be correlated, though no direct correlation between the severity of dementia and the extent of LC damage has been demonstrated. A recent study, however, has shown a positive correlation of the occurrence of depression in SDAT and the decrease in LC neuron number. Lesions of brainstem nuclei including the LC, with neuropathological changes such as Lewy bodies and NFT in Parkinson’s disease (PD) were described many years ago; a Parkinsonian state and LC lesions similar to those found in the disease are caused by the administration of the neurotoxin l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine in the macaque monkey. Even though PD is principally a disorder of locomotion, it is now generally accepted that in a number of patients, progressive mental impairment occurs in the course of the disease. Some authors have reported dementia in more than 50% of PD cases. According to the responsiveness to L-dopa treatment it has been postulated that two separate disorders can be distinguished in PD: an exclusive motor disorder occurring in a younger population with a longer and more “benign” course and a better response to L-dopa; and another, a motor followed by a cognitive disorder occurring in an older population with a more fulminant course and a poorer response to L-dopa. There has been a controversy over the distinction of a “cortical” dementia found in SDAT and a “subcortical” dementia present in PD patients. Several authors have claimed that in neuropsychological tests the dementia of SDAT, characterized mainly by aphasia, amnesia, agnosia and apraxia, can be distinguished from that found in PD patients, where the dementia is characterized by slowness of mental processing, forgetfulness, impaired cognition, apathy and depression, while no psychopathological difference between demented P D and SDAT patients was found by
others. Some investigators have suggested that the dementia in PD displays a pattern of impairment typical for a lesion of the frontal lobe and a laterality of the disease has been suggested based on the finding that patients with greater disease involvement on the left side of the body showed greater neuropsychological impairment than those more affected on the right body side. The question of whether the incidence of cortical plaques and tangles is correlated to the severity of dementia in PD is still also somewhat controversial. Early reports have shown more frequent occurrence of NP and NFT in demented P D patients than in non-demented, suggesting coincidental SDAT in these patients. Other authors could not demonstrate a positive correlation between NFT and NP formation and dementia, but reported a severe cell loss in the LC more frequently in demented P D patients that in those without symptoms of dementia. A correlation between the coeruleocortical-NE system and dementia has also been suggested based on modifications in the number of adrenergic receptors in demented PD patients. Material and methods Normal control included the brains of 11 patients, 4 male and 7 female, ranging in age from 43 to 89 years, with no clinical history of neurological or psychiatric disease as confirmed by postmortem gross and microscopic neuropathological examination. Vital data and selection criteria for all control cases were described in detail (Chan-Palay and Asan, 1989a). Data assembled in three paradigm control cases in the age group of the patients in the Alzheimer’s and Parkinson’s groups served as control values for quantitative analyses. Appropriate levels of normal mental function in patients was shown by results of between 22 and 26 from a possible 30 points in the last available mini-mental status tests. In the group of SDAT cases, the brains of 8 patients that had been clinically diagnosed according to DSM 111-R criteria were studied, 2 male and 6
627
female cases with ages ranging from 71 to 85 years. Cases of dementia due to other neurological disorders, such as ischemia, multiple infarcts, Pick’s disease, etc., were excluded. Postmortem delays ranged from 3 to 16 h, with postmortem delays of 5 h and less in 5 cases. Counts of NP and NFT were made on Bodian-Silver-stained preparations. For all cases in this group, the counts yielded moderate to high indices of NP and NFT in the examined areas, which is indicative of SDAT. Seriously impaired mental function in these patients was indicated by a score of 0 to 5 points in the last available mini-mental status tests. In the study of cases with PD seven diagnosed patients were included, ranging in age from 76 to 90 years, three male and four female cases. Clinically, two of the patients had PD responsive to L-dopa treatment without symptoms of dementia (P-D), five patients had histories of rapidly progressive dementia, with onset 2-3 years before death. Of these five demented patients, three were responsive to treatment with L-dopa (P + D); two were atypical and their Parkinsonian symptoms did not respond to L-dopa treatment (P D/L-dopa nonresponsive). The postmortem delays ranged from 3.5 to 21 h, and were less than 7 h in four of the cases. The clinical diagnosis of PD was confirmed at autopsy by both gross and microscopic neuropathological examination. The substantia nigra showed considerable cell loss, loss of pigmentation, numerous Lewy bodies, and gliosis pathognomic of PD in every case. Counts of NFT and NP were performed as described for the SDAT cases and yielded indices slightly higher in the age-matched control and non-demented Parkinson’s cases than for the demented Parkinson’s patients. The last available mini-mental status test scores were 22 to 26 for the non-demented PD group and 11 to 16 for the demented patients.
+
Fixation and immunocytochemistry
The protocols used for fixation of the studied brainstems and immunocytochemistry were de-
scribed in detail in preceding papers (Chan-Palay and Asan, 1989 a,b; Chan-Palay et al., 1990). Computer-assisted quantitative morphological analyses
The computer system and the recording procedures used have been described in detail (ChanPalay and Asan, 1989 a,b; Chan-Palay et al., 1990). Briefly, the immunocytochemically stained serial brainstem sections were reassembled in the correct anatomical order, and the LC outline in the coronal plane, its rostra1 and caudal borders were determined, and its rostrocaudal length on both sides of the brainstem was calculated. For the computer-assisted measurements of morphological parameters of TH-immunoreactive LC neurons, and for the mapping and counting of neurons and the three-dimensional reconstruction procedure for the analysis of neuron distribution in the LC an IBM-AT mouse-based userinteractive image analysis system with the Cellmate program (Bioquant, TN) was used. Outlines of individual cell somata and dendritic arbors were recorded for calculations of soma areas and dendritic arbor length. Plots of these recordings served to illustrate alterations in individual neuron morphology. To ensure comparability, all quantitative measurements of neuronal parameters were carried out on immunoreactive neurons stained with the peroxidase-antiperoxidase (PAP)-method. Cell counts were performed on all reassembled sections of one TH-immunoreacted series of sections by cursor-marking the localization of whole cell bodies using different symbols signifying different morphological classes of cells (see below) in all focus levels throughout the entire extent of the LCs of both sides of the brainstem. TH-immunoreactive cells of the locus subcoeruleus and the pars cerebellaris loci coerulei were recorded but not counted. Neuron numbers on partially damaged sections were approximated either from the numbers counted on immediately adjacent TH-immunoreacted sections or from those counted in the contralateral
628
LC of the same section and the ipsilateral LCs of the preceding and following sections of the same series. Total neuron numbers were calculated by the computer for the entire LCs and, by their differing recording symbols, differentiated according to the four neuron classes: Large multipolar (LM), large bipolar (LB), small multipolar (SM) and small bipolar (SB) neurons. The recordings of the reassembled sections were then aligned to match as closely as possible the situation in the intact brains, and an image of the three-dimensional distribution of the neurons was created by the computer. Results Identification of NE-producing neurons is done by immunocytochemical demonstration of two NE biosynthetic enzymes, T H and DPH, and immunoreactions are visualized by the PAP and immunogold-silver staining methods. It was demonstrated that the reactions with antisera against TH and D P H yield equivalent results and that both immunocytochemical visualization methods allow detailed analyses of neuronal morphology. The neurons of the human LC fall into four distinct classes: large multipolar neurons with round or multiangular somata (LM), large elliptical “bipolar” neurons (LB), neurons with round or multiangular somata (SM) and small ovoid “bipolar” (SB) neurons. Though most of the neurons contain neuromelanin pigment, some of the neurons of the larger type lack pigmentation. Dendritic arborization in all neuron types is extensive, and computer-assisted quantitative measurements of the neuronal structure parameters, soma size, dendritic arbor length, surface area and volume are given. Comparison of neuronal morphology in different age groups shows that even though the soma areas of LC neurons of all four classes are decreased in older normal adult brain, the dendritic arborization is equally extensive. Detailed mapping of the immunoreactive neurons and computer-assisted three-dimensional reconstruction of the LCs are used to analyze the
morphology of the nucleus as a whole. According to cellular distribution patterns, the LC is divided into rostral, middle and caudal parts with neurons scattered over a large area rostrally, tightly clustered in the middle and very densely packed in the caudal part. Small neurons predominate in all parts, but the relative contribution of larger cells decreases in a rostro-caudal direction. SB neurons are the most frequent cells in the caudal part and display distinct dorsomedial-ventrolateral orientation. These general morphological characteristics are the same in all age groups, but cell density in rostral and middle parts is decreased in old age, while the relative frequency of large cells is increased especially in the rostral LC. No age-dependent decrease in nuclear length is observed. Assessment of neuron numbers documents a cell loss of 27 to 55% in older adult brains. Cell loss is topographically arranged, being highest in the rostral part, lower in the middle and virtually absent in the caudal part. Quantitative assessment of the distribution of the different morphological neuron classes confirms the observations mentioned above, suggesting that, especially in the rostral part of the LCs of older adult brains, loss of smaller cells is comparatively higher than that of larger cells. The computer-generated three-dimensional reconstruction provides the possibility of visualizing LC shape and cell distribution closely approximating the situation in the intact brain and facilitates the detection of morphological differences of the LC in individual brains (see Fig. la-f). After the studies of the controls, a detailed qualitative and quantitative investigation of the morphology and distribution of the N E neurons in the human LC in two classes of neurodegenerative disorders involving demeitia, the SDAT and PD was undertaken. In SDAT, the four basic LC neuron classes found in the normal human brain are recognizable in the remaining cells, but the cell somata are generally larger, the cell bodies are swollen and misshapen, and the dendrites are foreshortened and thick and less branched than in neurons of control LCs. Quantitative analysis confirms the qualita-
629
tive observations. The reduction of absolute numbers of LC-NE neurons in paradigm cases of SDAT and PD as compared to controls are shown in Table 1. la
+
TABLE 1
e
f
R
R
L
Fig. 1. Three-dimensional computer reconstruction of the LC of a younger control case, 55 years (a); an older control case, 78 years (b); a case of mild SDAT with comparatively little cell loss, 78 years (c); a severe SDAT case with extreme cell loss, 77 years (d); a P D case without dementia, 76 years (e); and a P D case with dementia, L-dopa non-responsive, 82 years (f). The reconstruction is viewed from dorsal, shifted in a 25" angle from the plane of the figure. R = right, L = left LC. The outline of the fourth ventricle is drawn on every fourth section, and each TH-immunoreactive neuron on all the recorded sections is marked by a dot. Note the cell loss which occurs mainly in the rostral part of the older control case (b) as compared to the younger control case (a). Note also the high neuronal loss present predominantly in the rostral and middle parts in both SDAT cases (c) and (d). In the P D case with dementia/L-dopa non-responsive (f) cell loss is present throughout the nucleus, but predominantly in the caudal aspects, as is true also for the P D patient (e).
Neuron number x 10'
Age
Control Control
79 78
47.5 40.9
SDAT (Mild) SDAT (Severe) SDAT (Severe)
78 74 77
34.0 18.8 5.7
P P+D P D/L-dopa nonresponsive
76 83
31.1 23.3
79
2.5
+
C
Sex
Case
A reduction of total 'neuron numbers in the LC of between 3.5% and 87.5% as compared to age-matched controls is found in SDAT. This neuron loss is topographically arranged; in the rostral part of the LC, the reduction is greatest, being more than 28% in the case least affected in this part, and 97% in the case most severely affected. The middle part is less, and the caudal part least affected by cell loss in all cases. The average rostrocaudal nuclear length in SDAT cases is reduced as compared to controls (13 mm and 14.9 mm, respectively). In PD cases, the neuronal morphology is generally more severely altered than in SDAT cases. The four neuron classes are hardly distinguishable, the cell bodies are swollen, and frequently contain Lewy bodies. The dendrites are short and thin, and arborizations are reduced or virtually absent. In the neuropil surrounding the remaining neurons cell remnants and masses of extraneuronal pigment are found. The neuron loss is more severe than in SDAT (26.4%-94.4%). While cell reduction varies within each group, a difference in the topographical arrangement of cell loss can be recognized between P D and P D/L-dopa non-responsive: in P D/L-dopa non-responsive the neuron loss is equally great or greater in the middle and caudal part. The average rostrocaudal length in PD cases is less than in SDAT and the controls (12.4 mm). Further, we have done a study which demonstrates the presence of neuropeptide Y
+
+
+
630
(NPY)-, C-terminal flanking peptide of NPY (CPON) and galanin-immunoreactive (GA-i) neurons and axons in the LC in normal adult human brain and in brains of patients with SDAT and PD, as well as evidence for coexistence of peptides with catecholamines in individual neurons. In the controls, C-PON-i and NPY-i neurons are found, in numbers, highest rostrally, lower in the middle, and lowest in the caudal LC. GA-i neurons are more frequent caudally. C-PON-i small to medium-sized LC neuron numbers are decreased in the older normal adult brain compared to the younger. NPY-i and GA-i neuron numbers are very low in all age groups. The axonal networks are densest rostrally. C-PON-i and NPY-i axons possess larger varicosities and thicker intervaricose segments than GA-i axons. Both the peptidergic neuronal systems and the innervation pattern are altered detrimentally in SDAT and PD, worse in the latter than in the former cases. In SDAT, some C-PON-i neurons display morphological alterations resembling those encountered in LC TH-i neurons in the same case, with blurred somatic outlines, misshapen cell bodies and foreshortened dendrites. However, the alterations are generally less severe than those of the TH-i neurons, and many C-PON-i neurons appear to be entirely normal. Compared to the number of TH-i neurons, the numbers of peptide-i neurons are higher in SDAT than in controls, and often more neurons are located cau-
dally. Peptide-i axonal networks are less dense, and the remaining C-PON-i and NPY-i axons are tortuous and have larger varicosities than normal. In PD, the changes of C-PON-i neuronal morphology are as severe as in the TH-i neuron population, with rounded cell bodies and loss of dendritic arbors; the peptide-i systems of the LC, especially the increase in neuron numbers in SDAT, may be interpreted in terms of a plasticity of the aged brain. Acknowledgements We thank the Swiss National Foundation and the Sandoz Foundation for gerontological research, the Geigy Jubilaumsstiftung and Astra for research support. References Chan-Palay, V. and Asan, E. (1989a). Quantitation of catecholamine neurons in the locus coeruleus in human brains of young and older adults. J. Comp. Neurol., 287: 257-372. Chan-Palay, V. and Asan, E. (1989b). Alterations in catecholamine neurons of the locus coeruleus in senile dementia of the Alzheimer’s type, and in Parkinson’s disease with and without dementia. J. ?amp. Neurol., 287: 373-392. Chan-Palay, V., Jentsch, B., Lang, W., Hichli, M. and Asan, E. (1990). Distribution of neuropeptide Y, C-Terminal flanking peptide of NPY and galanin and coexistence with catecholamine in the locus coeruleus of normal human,. Alzheimer’s dementia and Parkinson’s disease brains. Demenfia, 1: 18-31.
631
Subject Index
Acetylcholine in cell bodies of peri-locus coeruleus a (peri-LCa) 34 See also, cholinergic neurons Acetylcholinesterase (AChE) in cell bodies of dorsal pontine tegrnentum (DPT) 5 Barrington’s nucleus 5 locus coeruleus (LC) complex 3, 5 raphe nuclei 5 in fibers to the LC 36 Adaptation (See Vestibule-ocular reflexes (VOR)) effects of P-adrenoceptors 492-493 Adenosine receptor agonists action on LC neurons 200 Adrenergic cell group C1 41, 167-170 colocalized peptides 370 efferents to LC complex 41 efferents to spinal cord 78 Adrenoceptors a-adrenoceptors action of preganglionic neurons 369 agonist PHE 326 antagonist phentolarnine 78, 276-279, 284, 346, 355, 382, 384, 387, 480, 548 development in bed nucleus of stria terminalis 181 cerebral cortex 180-181 cerebellum 180-181 mesolimbic area 180 a,-adrenoceptors 272 action in the raphe 326 thalamus 298 agonist corynanthine 382 methoxamine 176, 465-470 phenylephrine (PH) 176 antagonist HEAT 177, 276 prazosin 163, 176, 273, 326-327, 346, 382, 384, 387, 400, 465, 475-476, 478, 565 distribution in the amygdala 278 basal ganglia 276 brainstem 274, 276 cerebellum 276, 464
cerebral cortex 273, 276-278 hippocampus 278 hypothalamus 278 olfactory bulb 276 stria terminalis 280 a2-adrenoceptors agonist activate potassium conductance 176 clonidine 37, 68, 170, 176, 202, 208, 280, 315, 326, 367, 371-372,384,401,412,430,445-448,453,455, 465-479, 524,547-54a,s64-565 effect on early L C oscillations 174 isoprenaline 295 metaraminol 384 UK 14304 170 antagonist idazoxan 275-285, 326, 384, 572 piperoxan 177, 398 yohimbine 176, 273, 276, 382, 384, 387, 465, 469, 475, 565 distribution in amygdala 280 brainstem 280 cerebral cortex 280-284 cerebellum 280, 464 corpus striaturn 280 hippocampus 280-286 hypothalrnus 280 nucleus tractus solitarius (NTS) 280-284 thalamus 280 role in paradoxical (REM) sleep 547-548, 550-551 role in spinal antinociception 382 supersensitivity 391 P-adrenoceptors action in cerebral cortex 294-297 thalamus 298-301 agonist dihydroalprenolol 179 hydroxylbenzylpindolol 179 hydroxypindolol 179 isoproterenol 179, 308, 326, 354-355, 384, 448, 465-479, 491-496, 548-549 antagonist metoprolol ( p , ) 309 propranolol 276, 287-288, 313-314, 382, 384, 443-445, 448, 465, 468, 473-477, 549-550, 572, 601-602 sotalol 326-327, 448, 480, 489-495 timolol 308, 314, 480 distribution in
632 cerebral cortex 179, 287-288 cerebellum 179, 287-288, 464, 491-493 globus pallidus 287 hippocampus 308, 323-325 nucleus accumbens 287 olfactory tubercle 287 substantia nigra 287 thalamus 287 role in neural plasticity 599-611 role in paradoxical (REM) sleep 548, 511 Adrenocorticotropin (ACTH) in cell bodies of the LC complex 242 increases LC neuron discharge rate 243 Alertness role of LC 530 Alzheimer’s disease 625-630 y-Aminobutyric acid See GABA Amphetamine 326,327 produces analgesia 382 Amygdala a,-adrenoceptors 278 a,-adrenoceptors 280 efferent projections to dorsal pontine tegmentum (DPT) 49, 64-66 locus coeruleus (LC) 38, 49 Analgesia amphetamine produced 382 blocked by intrathecal methasergide 78 phentolamine 78 Angiotensin 11 (MI) afferents to the LC 217-218 in cells of parabrachial nucleus 218 receptors in the LC 218-228 Angler fish pancreatic polypeptide (AFPP) in nucleus paragigantocellularis (PGi) 61 in nucleus prepositus hypoglossi (PrH) 61 Anticholinesterase effect on LC neurons 222-224 Antidepressants attenuate stress-elicited LC activation 253 desmethylimipramine (DMI) 251, 591-596 effect on LC neuron discharge rate 251 induce axonal sprouting of LC neurons 590-596 maprotiline (MPL) 590-596 mianserin (MIA) 251 sertraline (SER) 251, 591-596 Antinociception role of LC 386-391 APV (NMDA receptor antagonist) 169 Area postrema 375 Arousal role of noradrenergic system 302-303 role of thalamic neurons 302-303 Atrial natriuretic factor (ANF) in cell bodies of dorsal pontine tegmentum (DPT) 11 Barrington’s nucleus 11
locus coeruleus (LC) complex 11 parabrachial nucleus 12 raphe 11 Atropine (muscarinic cholinoceptor antagonist) 387,442, 447, 452-453, 455 Autoinhibition of LC neurons blocked by piperoxan 177 Autonomic control role of central noradrenergic neurons 365-377 Axonal sprouting of LC neurons 618 following antidepressant treatment 590-596 following stress 588-590
Baclofen (GABA, receptor agonist) 188, 190-191, 194, 490-491 Barrington’s nucleus neurochemical agents in cell bodies or‘ acetylcholinesterase (AChE) 5 atrial natriuretic factor (ANF) 11 corticotropin releasing factor (CRF) 9 dynorphin p (Dynp) 11 glutamic acid decarboxylase (GAD) 10 neuropeptide Y (NPY) 7 substance P (SP) 9 Basal ganglia a,-adrenoceptors 276 Bed nucleus of stria terminalis angiotensin I1 (AII) afferents to LC 218 development of a-adrenoceptors 181 neurons containing vasopressin (VP) 228 Bethanechol (muscarinic cholinoceptor agonist) 442, 452, 602-603 Bicuculline (GABA, receptor antagonist) 168-169, 188, 387 BMI (GABA, receptor antagonist) 190, 194 Brainstem a,-adrenoceptors 274, 276 a,-adrenoceptors 280 dopamine-B-hydroxylase (DPH) in axons 92 Buspirone (5-HT1, receptor agonist) effect on LC neurons 67
Carbachol (cholinoceptor agonist) 34, 37, 375, 412, 441-443, 450-452, 522-523 Cerebellum action of NE 331-339 adrenoceptors in a,-adrenoceptors 276 a,-adrenoceptors 280 p;adrenoceptors 179, 287-288 afferent innervation from LC 464 contribution to posture control by adrenoceptors 463-481 following microinjection of a,-adrenoceptor agonist methoxamine 465-470 a,-adrenoceptor antagonist prazosin 465, 475-476, 478
633 a,-adrenoceptor agonist clonidine 465-479 a,-adrenoceptor antagonist yohimbine 465, 469, 475 P-adrenoceptor agonist isoproterenol 465-471, 476 P-adrenoceptor antagonist propranolol 465, 468, 473-477 control of vestibulo-ocular reflexes (VOR) 485-496 effect of p-adrenoceptor agonist isoproterenol 491 -493 antagonist sotalol 491-493 effect of GABAergic injections 489 GABA, agonist muscimol 490-491 GABA agonist baclofen 489-490 control of vestibulospinal reflexes 463-481 effect of adrenoceptor agonists 463-481 effect of adrenoceptor antagonists 463-481 development of a-adrenoceptors 180-181 Cerebral cortex actions of norepinephrine (NE) 294-297, 525-530 activation by LC stimulation 522-524 a,-adrenoceptors 273, 276-278 a,-adrenoceptors 280-284 P-adrenoceptors 179, 287-288 mediate neural plasticity 599-611 axonal sprouting of LC neurons following antidepressant treatment 590-596 stress 588-590 development of a-adrenoceptors 180-181 efferent projections to locus coeruleus (LC) 49, 511 event-related potentials modified by LC lesions 526-530 norepinephrine (NE) increase during opiate withdrawal 209 Cervicospinal reflex role of LC 41 1-431 /3-Chlornaltrexamine (fi-opioid receptor antagonist) 202 Chlorpromazine 400 Cholecystokinin (CCK) in cell bodies of dorsal pontine tegmentum (DPT) 8 raphe nuclei 8 Choline acetyltransferase (ChAT) in cells of dorsal pontine tegmentum (DPT) 5, 22, 42, 534, 536 locus coeruleus (LC) 3, 536 parabrachial nucleus 42, 536 nucleus paragigantocellularis (PGi) 61 nucleus prepositus hypoglossi (PrH) 61 Cholinergc neurons distribution in dorsal pontine tegmentum (DPT) 5 , 22-24, 42, 435-437, 534, 536 sleep-waking cycle role in paradoxical (REM) sleep 539-541 role in waking 536-539 Cholinoceptor agonist carbachol 34, 37, 375, 441-443, 450-452, 522-523 Clonidine (a,-adrenoceptor agonist) 37, 68, 170, 176, 202,
208, 280, 315, 326, 367, 371-372, 384, 401, 412, 430, 445-448,453, 455,465-479,524, 547-548, 564-565 CNQX (kainate/quisqualate receptor antagonist) 169 Coeruleospinal pathway 19, 91, 93, 104, 124-139, 144, 384, 387 action on spinal cord 395-405 dorsal horn neurons 138, 387 nociception transmission 386-391 motoneurons 117, 343, 400-403, 412 primary afferents 404 Renshaw cells 404, 412, 429-431 bifurcating axons ascending to hypothalamus 20 cellular properties criteria for identification 396, 415 response to neck stimulation 419-421, 426-427 response to vestibular stimulation 415-421, 426-427 course in spinal cord 97-98, 127-131, 385 conduction velocity 399, 416, 425 termination in dorsal horn 19, 98, 144-154 intermediate zone 19, 131-137, 144-154 intermediolateral column 19 ventral horn 19, 131-137, 144-154, 385, 412 Coexistence (colocalization) of neurochemical agents in cell bodies in A1 noradrenergic cell groups norepinephrine (NE) and galanin (GAL) 371 neuropeptide Y (NPY) 371 C1 adrenergic cell group epinephrine and enkephalin (Enk) 370 galanin (GAL) 371 neuropeptide Y (NPY) 370 substance P (SP) 370 dorsal pontine tegmentum (DPT) catecholamine and enkephalin (Enk) 105, 109 neuropeptide Y (NPY) 117 locus coeruleus (LC) complex norepinephrine (NE) and corticotropin releasing factor (CRF) 242 enkephalin (Enk) 242 galanin (GAL) 242 neuropeptide Y (NPY) 242 vasopressin (VP) 236, 242 raphe nuclei serotonin (5-HT) and substance P (SP) 10 thyrotropin releasing hormone (TRH) 10 Corpus striatum a,-adrenoceptors 280 Corticotropin releasing factor ( C W ) as neurotransmitter in LC 250 effect on LC neuron discharge 243, 250-251 fibers containing in the LC 36, 55-56, 61, 250
634 in the raphe 250 hypersecretion during stress and depression 250 in cells of dorsal pontine tegmentum (DPT) 9, 61 Barrington’s nucleus Y locus coeruleus (LC) 242 peri-locus coeruleus (Y (peri-LCa) 34 Corynanthine (a,-adrenoceptor agonist) 382 Cyclobenzaprine (CBZ) 398-399
DADLE (opioid receptor agonist) 198 DAGO (p-opioid receptor agonist) 198 Delta sleep-inducing peptide (DSIP) in cell bodies of the LC complex 242 Dementia of Alzheimer’s and Parkinson’s disease alterations in locus coeruleus (LC) 625-630 decrease in TH-containing neurons 625-630 Depression dysfunction of hypothalamic pituitary axis 250 hyposecretion of CRF 250 Desmethylimipramine (DMI) (norepinephrine reuptake inhibitor) 335, 367 antidepressant effect on LC neuron activity 251 induces LC axonal sprouting 591-596 Desynchronized sleep (See Paradoxical (REM) sleep) Development of brain site a-adrenoceptors 180- 181 P-adrenoceptors 179-180 LC cells adrenergic responsiveness 176-178 differentiate 173 electrical properties 174 electronic coupling 174-176 regulation in vitro 617-622 stain for catecholamine 173 Dihydroalprenolol (P-adrenoceptor agonist) 179 DO1 (S-HT, receptor agonist) effect on LC neurons 67 Dopamine effect of LC neurons 170 in afferents to LC 40 receptor antagonists haloperidol 400 pimozide 399 Dopamine-P -hydroxylase (DpH) in axons of the brainstem 92 spinal cord 94, 125, 130-137 apposing spinothalamic tract (STT) neurons 83 in cell bodies of the Kolliker-fuse nucleus 384-385 locus coeruleus (LC) 3, 33, 130-135 Dopaminergic cell groups A l l 40 A13 41
Dorsal horn neurons effect of locus coeruleus (LC) stimulation 138, 387 Dorsal pontine tegmentum (DPT) anatomical organization afferent projections from 20, 36 forebrain 22, 104 amygdala 49, 64-66 hypothalamus 22 prefrontal cortex 49, 53 preoptjc region 27 septum 22 substantia innominata 22 medullary reticular formation 22, 42 nucleus tractus solitarius (NTS) 49, 53, 66 periaqueductal gray (PAC) 66 raphe 49, 53 spinal cord 22, 49, 53 neurochemical nature of afferents catecholaminergic 26 cholinergic 25 GABAergic 3, 10, 26 serotonergic 26 vasopressinergic 231-234 cytoarchitecture 15, 33, 103 dendritic organization 21 relationship between noradrenergic and cholinergic neurons 22-24, 42, 435-437 GABAergic neurons 24-26 function role in paradoxical (REM) sleep 34, 435-437, 545 neurochemical agents in cell bodies acetylcholinesterase (AChE) 5 angiotensin II (AH) 218 atrial natriuretic factor (ANF) 11 calcitonin gene-related peptide (CGRP) 9 cholecystokinin (CCK) 8 choline acetyltransferase (ChAT) 5, 22, 42, 534, 536 corticotropin releasing factor (CRF) 9, 61 dynorphin p (Dynp) 11 enkephafin (Enk) 61, 104, 105, 108-118 galanin (GAL) 6 glutamic acid decarboxylase (GAD) 10, 24-26 neuropeptide Y (NPY) 7, 104, 109 neurotensin (NT) 7 serotonin (5-HT) 11, 59, 104, 112 somatostatin (SST) 10, 104 substance P (SP) 9, 104 tyrosine hydroxylase (TH), 6, 106-118, 534 vasoactive intestinal peptide (VIP) 8 DSP-4 (neurotoxin) effect on NE axons of the LC 91, 257-267 Dynorphin in afferents to LC 36 in cell bodies of dorsal pontine tegmentum (DPT) 11 Barrington’s nucleus 11 nucleus paragigantocellularis (PGi) 61 nucleus prepositus hypoglossi (PrH) 61
635 Enkephalin (Enk) decreases LC neuron discharge 37, 243 hyperpolarizes LC neuron 243 in afferents to LC 36, 42, 55, 61, 242 in cells of dorsal pontine tegmentum (DPT) 61, 104, 105, 108-118 Kolliker-Fuse nucleus 108-118 peri-locus coeruleus a (peri-LCa) 34 nucleus paragigangtocellularis (PGi) 56, 61 nucleus prepositus hypoglossi (PrH) 56, 61 Epinephrine colocalized with enkephalin (Enk) 370 galanin (GAL) 371 neuropeptide Y (NPY) 370 substance P (SP) 370 effect on LC neurons 170 in cells of nucleus paragigantocellularis (PGi) 56, 58 nucleus prepositus hypoglossi (PrH) 56, 58 Eserine sulphate (anticholinesterase) effect on locus coeruleus (LC) neurons 422-424 Event-related potentials modified by LC lesions 526-530 Excitatory amino acid ( E M ) in cells of nucleus paragjgantocellularis (PGi) 56, 62, 508 receptor antagonist kynurenate 169, 210 synaptic input to LC effect on neurons 168-169 Flocculus (See Cerebellum, Vestibule-ocular reflexes (VOR)) Flurazepam 335 Forskolin (direct activator of adenyl cyclase) 335, 353, 355, 357, 621 GABA ( y -aminobutyric acid) action modulated by NE 353-358 action on LC neurons 37, 64, 169-170, 189-194 as transmitter in the LC 187-192 in cell bodies of nucleus paragigantocellularis (PGi) 56 nucleus prepositus hypoglossi (PrH) 56, 58, 64 in fibers of the LC 3, 10, 26, 55, 188 GABA receptors GABA, receptor agonist muscimol 355, 489-491 antagonists bicuculline 168-169, 188, 387 BMI 190-194 picrotoxin 188 role in expression of neural plasticity 608-609 GABA receptor agonist baclofen 188, 190-191, 194, 489-490 on noradrenergic terminals 187 Galanin (GAL) effects on LC neurons 243 in afferents to LC 242 in cells of
dorsal pontine tegmentum (DPT) 6 locus coeruleus (LC) 3-6, 242, 630 raphe I Globus pallidus P-adrenoceptors 287 development of a-adrenoceptors 181 Glutamate effect on LC neurons 37, 67, 169, 316-317,375 Glutamic acid decarboxylase (GAD) in cells of dorsal pontine tegmentum (DPT) 10, 24-26 locus coeruleus (LC) 10, 24-26 raphe 10 Glycine effect on LC neurons 170 receptor antagonist strychnine 170
Habituation 579-584 Haloperidol (dopamine receptor antagonist) 400 HEAT (a,-adrenoceptor antagonist) 177, 276 Hippocampus action of NE on 307-312, 323-329, 352 a-adrenoceptors action in hippocampus 325-329 a,-adrenoceptors 273, 276-278 a,-adrenoceptor 280 p-adrenoceptors 308 actions in hippocampus 323-325 long-term potentiation (LPT) 308-312 afferent projections from LC 19 theta rhythm produced by LC stimulation 522-523 Histamine effect on neurons of cortex 295 thalamus 299 in afferents to the LC 36, 40 in efferents from lateral hypothalamic area 40 tuberomamillary nucleus 40 Hydralazine to produce hypotension 161 6-Hydroxydopamine (6-OHDA) produces NE denetvation 163, 179-180, 182 Hydroxylbenzylpindolol ( p -adrenoceptor agonist) 179 Hydroxypindolol ( p -adrenoceptor agonist) 179 Hypothalamic-pituitary axis dysfunction in depression 250 Hypothalamus a-adrenoceptor development 276 a,-adrenoceptors 278, 325-326 a,-adrenoceptors 280 afferent projections from noradrenergic cell groups 150 parabrachial nucleus 82 effects of norepinephrine (NE) 352 efferent projections to dorsal pontine tegmentum (DPT) 22 locus coeruleus (LC) 22, 38, 40, 49, 218
636 nucleus paragigantocellularis (PGi) 68 neurons containing vasopressin (VP) 227 origin of angiotensin I1 afferent to LC 218 receiving bifurcating LC axons to spinal curd 20 IBMX (phosphodiesterase inhibitor) 353 Idazoxan (a,-adrenoceptor antagonist) 275-285, 326, 384, 572 Isoprenaline (p-adrenoceptor agonist) 295 Isoproterenol (B-adrenoceptor agonist) 179, 308, 326, 354355,384,448,465-471,476,491-496, 548-549 Ketanserin (5-HT2 receptor antagonist) effect on LC neurons 67 Kolliker-Fuse nucleus neurochemical agents in cell bodies 9 dopamine-P-hydroxylase (DPH) 384-385 Met-enkephalin (M-Enk) 108-1 18 tyrosine hydroxylase (TH) 107-118 neurochemical agents in fibers vasopressin (VP) 237 Kynurenate (excitatory amino acid receptor antagonist) 169, 210 and opiate withdrawal 211-214, 510 Learning effected by LC stimulation 555-567, 571-573 Locus coeruleus complex (LC) a,-adrenoceptors 280 anatomical connections afferent projections from 20, 36, 48 dorsolateral pontine tegmentum 38, 42 forebrain 22 amygdala 38, 49 cortex 49, 51 1 hypothalamus 22, 38, 40 paraventricular nucleus 49, 218 preoptic region 22 septum 22 stria terminalis 38, 218 substantia innominata 22 medullary reticular formation 22, 38 neurochemical substance in afferents acetylcholine (ACh) 25, 42 acetylcholinesterase (AChE) 36 adrenocorticotrophic hormone (ACTH) 36, 242 y-aminobutyric acid (GABA) 3, 10, 26, 55, 188 angiotensin I1 (AH) 217 catecholamine 26, 40, 41 corticotropin releasing factor (CRF) 36, 55, 56, 61, 250 dopamine 40 dynorphin 36 enkephalin (Enk) 36, 42, 55, 61, 242 epinephrine 41, 51, 55, 58 excitatory amino acid ( E M ) 62, 212-213 galanin (GAL) 242 glutamic acid decarboxylase (GAD) 26, 188 histamine 36, 40 a-melanocyte stimulating hormone (a-MSH) 36
neuropeptide Y (NPY) 242 neurotensin (NT) 56, 242 norepinephrine (NE) 41 serotonin (5-HT) 11, 36, 40, 55, 59 somatostatin (SST) 56, 242 substance P (SP) 36, 42, 56, 242 vasoactive intestinal peptide (VIP) 56, 245 vasopressin (VP) 234, 242 nucleus paragigantocellularis (PGi) 20, 22, 26, 39, 49, 58, 508-509 and opiate withdrawal 210-214 nucleus prepositus hypoglossi (PrH) 20, 26, 39, 49, 508 nucleus tractus solitarius (NTS) 38, 49, 218 pontine reticular formation 38 spinal cord 22, 49 efferent projections to brainstem reticular formation 17, 91 cerebellum 91 cerebral cortex 19, 91 axon sprouting following stress and antidepressants 583-596 cranial nerve nuclei motor 17 trigeminal 162 sensory 17, 91 cochlear 17 nucleus tractus solitarius (NTS) 17 hippocampus 19 hypothalamus 19 arcuate 19 dorsal 19 lateral 19 p3raventricular 19 posterior 19 zona incerta 19 nucleus basalis of Meynert 19 spinal cord 19, 91, 93, 104, 124-139, 144, 384, 387 funicular trajectory 127-131, 385 conduction velocity 399, 416, 425 terminal fields 131-137, 144-154, 385, 412 thalamic nuclei intralaminar 19 lateral geniculate 19 midline 19 noradrenergjc terminals 79 ventral posterior lateral (VPL) 81 cellular properties criteria for identifying LC-NE neurons 160, 396, 415, 421 development of electrical properties 174-181 early age oscillations increased by muscarinic agonists 174 increased by phenylephrine (PHI 174 insensitive to tetrodotoxin 174 reduced by a-adrenoceptor agonist 174 reduced by p-opioid 174 electronic coupling 174-176 response to chemical agents acetylcholine (ACh) 37, 67, 171 adenosine receptor agonists 200
637 adrenocorticotropin (ACTH) 243 a,-adrenoceptor agonists 200, 202 clonidine 37, 68, 170, 176, 202 UK 14304 170 y-aminobutyric acid (GABA) 37, 64, 169-170, 189-194 angiotensin I1 (AII) 245 anticholinesterase 422-424 baclofen 188, 190-191 carbachol 375 corticotropin releasing factor (CRF) 243, 250-251 desmethylimipramine (DMI) 251 dopamine 170 DSP-4 258 enkephalin (Enk) 37, 243 epinephrine 170 galanin (GAL) 243 glutamate 37, 67, 169, 316-317, 375 glycine 170 hydralazine 161 insulin 162 N-methyl-D-aspartate (NMDA) 169 antagonist AP5 63, 169 CGS 19755 63 mianserin (MIA) 251 morphine 37 MPTP 258 muscarinic agonists 37 neuropeptide Y (NPY) 243, 245 neurotensin (NT) 37, 243 nitroprusside 253 norepinephrine (NE) 37, 67, 170, 176 opioids 197-204, 207-214 quisqualate 169 serotonin (5-HT) 37, 67, 171 depleter p-chlorophenylalinine (PCPA) 63 5-HT,, agonist 8-OHDPAT 67 buspirone 67 5-HT,, agonist TFMPP 67 5-HT, agonist DO1 67 5-HT2 antagonist ketanserin 67 sertraline (SER) 251 somatostatin (SST) 200, 243 substance P (SP) 37, 243-245 tachykinins 244 vasopressin (VP) 228, 243 response to cortex stimulation 511-512 response to neck stimulation 419-421, 426-427 response to stress 160-165, 588-590 response to synaptic input EPSP excitatory amino acid ( E M ) 168-169
IPSP y-aminobutyric acid (GABA) 169-170, 189 catecholamine 170, 171 blocked by a,-adrenoceptor antagonists 170 glycine 170 response to vestibular stimulation 416-421, 424, 426-427 cytoarchitecture 15, 33, 55, 399 altered morphology in dementia 625-630 dendritic organization 21, 53-55 development of LC cells regulation in vitro 617-622 development of target cells (differentiation) 173 functional aspects contribution to posture 411-431, 435-458 development of opiate tolerance 202 effects produced by chemical stimulation on cortex 522-524 on vestibulospinal reflex 445-453 electrical stimulation on adrenoceptor number and binding 563-564 on cerebellar Purkinje cells 332-333, 480 on dorsal horn neurons 138, 387 on learning 555-568,571-573 on nociceptive transmission 138, 386-391 on primary afferents 404 on spinal motoneurons 117, 343, 400-403, 412 on Renshaw cells 404, 412, 429-431 on spinal cord 395-405 similar to opiate withdrawal 209 response to sensory stimuli 505-508 role in alertness 530 antinocicept ion 386-39 1 autonomic control 375-377 cervicospinal reflexes 411-431 cortical activation 522-523 depression 250 evoked potential potentiation in hippocampus 312-317 neural plasticity 599-611 paradoxical (REM) sleep 34, 436, 539-541, 545-551 unit anaIysis 34, 160 sleep-waking cycle 412, 422-424, 503-505, 545-551 stress 250, 562 vestibulospinal reflexes 411-431, 435-458 vigilance 514-517 role of spinal cord projection in antinociception 386-391 in control of motot output 395-405 motoneurons 117, 343, 400-403, 412 primary afferents 404 Renshaw cells 404, 412 in control of posture 412-431, 435-458 trophic functions 182-183 neurochemical agents in cell bodies acetylcholinesterase (AChE) 3,5 adrenocorticotrophic hormone (ACTH) 242
638 atrial natriuretic factor (ANF) 11 choline acetyltransferase (ChAT) 3, 536 corticotropin releasing factor (CRF) 242 delta sleep-inducing peptide (DSIP) 242 dopamine-P-hydroxylase (DPH) 3, 33, 130-135 enkephalin (Enk) 108-118, 242, 403 galanin (GAL) 3-6, 242, 630 glutamic acid decarboxylase (GAD) 10, 24-26 neuropeptide Y (NPY) 3-7, 242, 403, 629-630 neurotensin (NT) 3-7, 242 serotonin (5-HT) 112, 403 somatostatin (SST) 242 tyrosine hydroxylase (TH) 3, 6, 33, 106-118, 134, 625-630 vasoactive intestinal peptide (VIP) 8 vasopressin (VP) 3, 228, 234, 242 Long-term potentiation (LTP) 307-319,324-325 by NE and LC in hippocampus 307-319 role of NMDA receptors 609-610 Maprotiline (MPL) antidepressant induces LC axonal sprouting 590-591 Masseteric reflex increased by phenylephrine (PH) 163 Medullary reticular inhibitory center afferent projection from peri locus coeruleus a (peri-LCa) 36, 436 Mesolimbic area development of a,-adrenoceptors 180 Metaraminol (a2-adrenoceptor agonist) 384 Methacholine effect on cortical neurons 295 Methasergide (5-HT receptor antagonist) 78, 163,387, 400 Methoxamine (a,-adrenoceptor agonist) 176, 465-470 N-Methyl-o-aspartate (NMDA) receptor 307-312, 318 role in long-term potentiation (LTP) 609-610 role in neuronal plasticity 609-610 Metoprolol (PI-adrenoceptor antagonist) 309 Mianserin (MIA) antidepressant 251 effect on LC neuron activity 251 Morphine (F-opioid receptor agonist) 198, 203, 208, 510 Motoneurons effect of LC stimulation 117, 343, 400-403 effect of norepinephrine (NE) 343-349, 403 Muscarinic cholinoceptor agonist - bethanechol 442, 452, 602-603 antagonist- atropine 387, 442, 447, 452-453, 455 increase frequency of early development LC oscillations 174 Muscimol (GABA, receptor agonist) 355, 489-491 Naloxone (opioid receptor antagonist) 198, 202, 387 Naltrexone (opioid receptor antagonist) 202, 208, 212, 511 Neck stimulation effect on locus coeruleus (LC) neurons 419-421, 426-427 vestibulospinal neurons 427-429
Neuropeptide Y (NPY) effect on LC neuron discharge rate 243, 245 in cell bodies of dorsal pontine tegmentum (DPT) 242 Barrington’s nucleus 7 locus coeruleus (LC) 3-7, 242, 403, 629-630 Neurotensin (NT) effect on LC neuron discharge rate 243 in cells of dorsal pontine tegmentum (DPT) 7 locus coeruleus (LC) 7 parabrachial nucleus 7 Neurotransmitter-related enzymes localized to axons and terminals in brainstem dopamine+ hydroxylase (DPH) 92 locus coerulus (LC) acetylcholinesterase (AChE) 36 choline acetyltransferase (ChAT) 42 glutamic acid decarboxylase (GAD) 11, 26 phenylethanolamine N-methyltransferase (PNMT) 3, 36 spinal cord dopamine-p hydroxylase (DOH) 83 localized to cell bodies in Barrington’s nucleus acetylcholinesterase (AChE) 5 glutamic acid decarboxylase (GAD) 10 dorsal pontine tegmentum (DPT) acetylcholinesterase (AChE) 5 choline acetyltransferase (ChAT) 5, 22, 42, 534, 536 glutamic acid decarboxylase (GAD) 10, 24-26 tyrosine hydroxylase (TH) 6, 106-118, 534 Kolliker-Fuse nucleus dopamine-p hydroxylase (DPH) 384-385 tyrosine hydroxylase (TH) 107-1 18 locus coeruleus (LC) acetylcholinesterase (AChE) 3 choline acetyltransferase (ChAT) 33, 42 dopamine-p hydroxylase (DOH) 3, 33 tyrosine hydroxylase (TH) 3, 33 nucleus paragigantocellularis (PGi) choline acetyltransferase (ChAT) 61 nucleus prepositus hypoglossi (PrH) choline acetyltransferase (ChAT) 61 parabrachial nucleus choline acetyltransferase (ChAT) 42, 536 tyrosine hydroxylase (TH) 6, 536 raphe nuclei glutamic acid decarboxylase (GAD) 10 Nitroprusside produces hypotension 160, 253 Nociception modulation of spinal transmission by coeruleospinal efferents 386-391 descending narodrenergic systems 381-391 spinal a,-adrenoceptors 382 stimulation of LC 138
639 Noradrenergic cell groups 41, 90 A1 41, 365 A2 41, 365 A4 41 A5 41, 365-370 autonomic effects 368 efferent projection to VPL 81 electrophysiologjcal properties 367 input-output relationships 366-367 stimulation produced antinociception 152 A6 See LC complex A7 41, 365 area postrema 375 distribution in dorsal pontine tegmentum (DPT) 534, 536 role in paradoxical (REM) sleep 34, 436, 539-541, 545-551 role in waking 536-539 efferent projection to brainstem 91 area postrema 150 dorsal raphe 150 locus coeruleus (LC) 41 motor trigeminal nucleus 150 cerebral cortex 150 hypothalamus 150 spinal cord 78, 125 funicular trajectory 93-98 terminal field 83, 94-98 o n blood vessels 502-503 on motoneurons 152 on spinothalamic tract (STT) neurons 83 involvement with autonomic function 365-377 Norepinephrine (NE) action on cerebellum 331-339, 352 cerebral cortex 294-297, 352 hippocampus 307-312, 323-329, 352 hypothalamus 352 long-term potentiation (LTP) in hippocampus 307-312 raphe nucleus 326 spinal cord dorsal horn neurons 383 motoneurons 343-349, 403 preganglionic neurons 368-369 Renshaw cells 404, 430 thalamus 297-303, 352 as a neuromodulator 351-360 in cells of the peri-LCa 34 microinfusion and microiontophoresis effect on cortex 524 hippocampus 503 masseteric reflex 163 Purkinje cells 334-336, 503 spinal cord dorsal horn neurons 383 motoneurons 344-348, 403 potentiation of GABA 332-336, 353-358 turnover in cortex during opiate withdrawal 209
Normorphine (opioid receptor agonist) 202 Nucleus accumbens P-adrenoceptors 287 Nucleus paragigantocellularis (PGi) afferent projections from fiber course 51 hypothalamus 68 Kolliker-Fuse nucleus 68 medullary reticular formation 68 nucleus tractus solitarius (NTS) 51, 68 periaqueductal gray (PAG) 51, 68 raphe magnus 68 supraoculomotor nucleus 68 efferent projections to locus coeruleus (LC) 49, 58-66, 212 neurochemical agents in cell bodies 210-214 acetylcholine (ACh) 56 y-aminobutyric acid (GABA) 56 angler fish pancreatic polypeptide (AFPP) 61 choline acetyltransferase (ChAT) 61 corticotropin releasing factor (CRF) 56, 61 dynorphin 61 enkephalin (Enk) 56, 61 epinephrine 56. 58 excitatory amino acid (EAA) 56, 62, 508 neurotensin (NT) 56, 61 serotonin (5-HT) 56 somatostatin (SST) 56 substance P (SP) 56, 61 vasoactive intestinal peptide (VIP) 56, 61 physiological properties 52, 508-509 on hippocampal evoked potentials 315-316 on opiate withdrawal 210-214, 509-511 role in vigilance 514-517 Nucleus prepositus hypoglossi (PrH) efferent projections to locus coeruleus (LC) 49, 58-66, 188 neurochemical agents in cell bodies y-aminobutyric acid (GABA) 56, 58, 508 angler fish pancreatic polypeptide (AFPP) 61 choline acetyltransferase (ChAT) 61 corticotropin releasing factor (CRF) 56, 61 dynorphin 61 enkephalin (Enk) 56, 61 epinephrine 56, 58 neurotensin (NT) 56, 61 somatostatin (SST) 56 substance P (SP) 56, 61 vasoactive intestinal peptide (VIP) 61 physiological properties 52, 508 role in vigilance 514-516 Nucleus tractus solitarius (NTS) a,-adrenoceptors 280-284 efferents to parabrachial nucleus 53, 66 input from AS cell group 368
8-OHDPAT (5-HT,, receptor agonist) effect on LC neurons 67
640 Olfactory bulb a,-adrenoceptors 276 Olfactory tubercle P-adrenoceptors 287 Opiate analgesia partially blocked by intrathecal methasergide 78 phentolamine 78 withdrawal effect of kynurenate 211-214, 510 effects reduced by clonidine 208 increases cortex NE 209 role of nucleus paragigantocellularis (PGi) 210-214 Opioids action on neurons of the LC 197-204, 207-214 receptor agonists DADLE 198 Met-enkephalin (M-Enk) 198, 202 normorphine 202 receptor antagonists naloxone 198, 202, 387 naltrexone 202, 208, 212, 511 tolerance 202 p-Opioids receptor agonists DAGO 198 morphine 198, 203, 208, 510 receptor antagonist P-chlornaltrexamine 202 receptor distribution 199, 207, 280 reduce frequency of early developed LC oscillations 174 Parabrachial nucleus afferent projections from nucleus tractus solitarius (NTS) 53, 66 spinal cord 53 efferent projections to spinal cord 384-385 ventral posterior lateral (VPL) thalamus 82 neurochemical agents in cell bodies angiotensin I1 (AII) 218 atrial natriuretic factor (ANF) 12 cboline acetyltransferase (ChAT) 42, 536 neurotensin (NT) 7 tyrosine hydroxylase (TH) 6, 536 Paradoxical (REM) sleep activity of LC-NE cells during 34, 160 role of adrenoceptors 550-551 dorsal pontine tegrnentum (DPT) 34,436,539-541, 545-551 decreased by clonidine 547-548 decreased by isoproterenol 548-549 increased by propranolol 549-550 noradrenergic and cholinergic neurons 539-541 Parkinson’s disease 625-630 Peptides (See individual listing) Periaqueductal gray (PAG) efferents to DPT 66
electrical stimulation inhibits thalamus 79 Peri-locus coeruleus a (peri-LCa )33 cell distribution 34 efferent projections to locus coeruleus (LC) 42 medullary inhibitory center 36, 436 neurochemical agents in cell bodies acetylcholine (ACh) 34 corticotropin releasing factor (CRF) 34 met-enkephalin (M-Enk) 34 norepinephrine (NE) 34 role in paradoxical (REM) sleep and postural control 34, 436, 539-541, 545-551 microinjection of a,-adrenoceptor agonist clonidine 547-548 microinjection of P-adrenoceptor agonist isoproterenol 548-549 microinjection of P-adrenoceptor antagonist propranolol 549-550 microinjection of cholinoceptor agonist bethanechol 442 carbachol 34, 37, 437, 441 PHE (a-adrenoceptor agonist) 326 Phenoxybenzamine 400 Phentolamine fa -adrenoceptor antagonist) 78, 276-279, 284, 346, 355, 382, 384, 387, 480, 548 Phenylephrine (PH) (a,-adrenoceptor agonist) 161, 163, 174, 176,344-345, 372,384 Picrotoxin (GABA, receptor antagonist) 188 Pimozide (dopamine receptor antagonist) 399 Pindolol 276 Piperoxan (a,-adrenoceptor antagonist) 177,398 Plasticity of LC efferents 618-619 visuocortex regulation by p-adrenoceptors 601-602 regulation by muscarinic cholinoceptors 602-603 role of GABA, receptors in expression 608-609 role of NMDA receptors 609-610 Posture contribution of locus coeruleus (LC) 411-431, 435-458 peri-LCa 435-458 Prazosin (a,-adrenoceptor antagonist) 163,176,273,326-327, 346,382,384,387,400, 465, 475476,478, 565 Preganglionic neurons a,-adrenoceptor action on 368-369 Primary afferents effect of LC stimulation 404 Propranolol ( p -adrenoceptor antagonist) 276, 287-288, 313314, 382, 384, 443-445, 448, 465, 468, 473-477, 549-550, 572, 601-602 Raphe nucleus action of a,-adrenoceptors 326 efferent projection to locus coeruleus (LC) complex 26, 40 spinal cord 144 neurochemical agents in cell bodies
64 1 acetylcholinesterase (AChE) 5 atrial natriuretic factor (ANF) 11 cholecystokinin (CCK) 8 galanin (GAL) 7 glutamic acid decarboxylase (GAD) 10 serotonin (5-HT) 11 substance P (SP) 9 thyrotropin releasing hormone (TRH) 10 vasoactive intestinal peptide (VIP) 8 neurochemical agents in fibers corticotropin releasing factor (CRF) 250 vasopressin (VP) 238 stimulation of inhibits thalamus 79 Rauwolscine 2 7 5 2 8 5 , 3 8 2 REM sleep (See Paradoxical (REM) sleep) Renshaw cell 429-431 effect of LC stimulation 404, 412 Reticulospinal neurons afferents from peri-LCa 36, 436 response to vestibular stimulation 437-441 Serotonin (5-HT) colocalized with substance P (SP) 10 thyrotropin releasing hormone (TRH) 10 depleter p-chlorophenylalanine (PCPA) 63 effects on neurons cortex 295 locus coeruleus (LC) 37, 67, 171 spinal motoneurons 348-349 thalamus 299 in afferent fibers to dorsal pontine tegmentum (DPT) 26 locus coeruleus (LC) 11, 36, 40, 55, 59 thalamus 79 in cells of dorsal pontine tegmentum (DPT) 11, 59, 104, 112 locus coeruleus (LC) 112, 403 nucleus paragingantocellularis (PGi) 56 raphe 11 receptor 5-HT receptor antagonist methasergide 78, 163, 387, 400 5-HT,, receptor agonists buspirone 67 8-OHDPAT 67 5-HT,, receptor agonist TFMPP 67 5-HT, receptor agonist DO1 67 5-HTz receptor antagonist ketanserin 67 Sertraline (SER) antidepressant 251 effect on LC neuron activity 251 induces LC axonal sprouting 591-596
Sleep-waking cycle paradoxical (REM) decreased by clonidine into DPT 547-548 isoproterenol into DPT 548-549 increased by propranolol into DPT 549-550 LC neuron activity during 412, 422-424, 503-505 role of NE and ACh neurons of DPT 536-539 slow-wave activity of thalamic neurons 300 Somatostatin (SST) action on LC neurons 200, 243 in cell bodies of dorsal pontine tegmentum (DPT) 10, 104 locus coeruleus (LC) 242 nucleus paragigantocellularis (PGi) 56 nucleus prepositus hypoglossi (PrH) 56 Sotalol (B-adrenoceptor antagonist) 326-327, 448, 480, 489-495 Spinal cord afferents from locus coeruleus (LC) 19, 91, 93, 104, 124-139, 144, 384, 387 funicular trajectory 97-98, 127-131, 385 conduction velocity 399, 416, 425 terminal fields 19, 98, 131-137, 144-154, 385, 412 parabrachial nucleus 384-385 cells action of a,-adrenoceptors 326 efferents to parabrachial nucleus 53 motoneurons action of locus coeruleus (LC) 117, 343, 400-403, 412 action of norepinephrine (NE) 343-349, 403 action of serotonin (5-HT) 348-349 Renshaw cells action of locus coeruleus (LC) 404, 412 Spinothalamic tract (S'IT) neurons innervation by axons containing D P H 83 innervation terminal types 83-85 Spiroperidol 280 Stress and antidepressants on LC 253 effect on LC neuron activity 160-165 elicited LC activation 253 hypersecretion of CRF during 250 induced LC axonal sprouting 588-590 Stria terminalis (See Bed nucleus of) a,-adrenoceptor 280 Strychnine (glycine receptor antagonist) 170 Substance P (SP) effects on LC neurons 37. 243-245 in afferent projections to the LC 36, 42, 56, 242 in cell bodies of the dorsal pontine tegmentum (DPT) 9, 104 raphe 9 Substantia nigra P-adrenoceptors 287
642 Supersensitivity a,-adrenoceptors develop after 6-OHDA 335, 391 Synaptogenesis in locus coeruleus (LC) 178 Tachykinins effects on LC neurons 244 Terminal types 145-154 Tetrodotoxin 176 young LC neurons insensitive to 174 TFMPP (5-HT,, receptor agonist) effect on LC neurons 67 Thalamus action of norepinephrine (NE) 297-303 activity during sleep 300-301 a,-adrenoceptors 276 a,-adrenoceptors 286 P-adrenoceptors 287 afferent projection from locus coeruleus (LC) 19 effect of histamine on 299 effect of serotonin (5-HT) on 299 inhibited by electrical stimulation of dorsal raphC 79 periaqueductal gray (PAG) 79 innervated by fibers containing serotonin (5-HT) 79 from raphC 82 norepinephrine (NE) 79 from A5 cell group 81 from LC complex 81 ventral posterior lateral (VPL) nucleus afferents from spinal cord 82 Timolol (P-adrenoceptor antagonist) 308, 314, 480 Tyrosine hydroxylase (TH) in cell bodies of the dorsal pontine tegrnenturn (DPT) 6, 106-118, 534 Kolliker-Fuse nucleus 107-118 locus coeruleus (LC) complex 3, 6, 33, 106-118, 134, 625-630 parabrachial nucleus 6, 536 Vasoactive intestinal peptide (VIP) in afferents to LC 56, 245 in cell bodies of the locus coeruleus (LC) 8 nucleus paragigantocellularis (PGi) 56, 61 nucleus prepositus hypoglossi (PrH) 61 raphe 8 Vasopressin (VP)227-239 colocalized with norepinephrine (NE) 236 in afferents to the Kolliker-Fuse nucleus 237 locus coeruleus (LC) 234 raphe 238
in cell bodies of the bed nucleus of stria terminalis 228 hypothalamus 227 Vestibular stimulation effect on locus coeruleus (LC)neurons 415-421, 424, 426-427 reticulospinal neurons 437-441 vestibulospinal neurons 427-429, 437-441 Vestibulo-ocular reflexes NOR) 485-496 control by cerebellum 485-496 injected with P-adrenoceptor agents 491-493 injected with GABAergic agents 489-491 control by cerebellum on adaptation 485-486 injected with P-adrenoceptor agents 492-493 Vestihulospinal neurons response to vestibular stimulation 427-429, 437-441 Vestihulospinal reflex mechanisms for gain control 411-431, 435-458, 463-481 role of cerebellum in controlling gain 463-481 injected with a,-adrenoceptor agonist methoxamine 466-470 a,-adrenoceptor agonist clonidine 466-479 P-adrenoceptor agonist isoproterenol 470-471, 476 adrenoceptor antagonists 468, 473-477 role of locus coeruleus (LC) in controlling gain 411-431, 435-458 injected with a,-adrenoceptor agonist clonidine 445-448 P-adrenoceptor agonist isoproterenol 448 cholinoceptor agonist carbachol 450-452 role of peri-LCa in controlling gain 441-445 injected with a,-adrenoceptor agonist clonidine 445-448 P-adrenoceptor antagonist propranolol 443-445 cholinoceptor agonist bethanechol 442 carbachol 441 cholinoceptor antagonists atropine 452-453 Vigilance role of noradrenergic locus coeruleus (NE-LC) 501-517 Waking role of noreadrenergic and cholineargic DPT neurons 536-539 Yohimbine (a,-adrenoceptor antagonist) 176, 273, 276, 382, 384,387, 465, 469, 475, 565