Hypothalamic Integration of Circadian Rhythms

PROGRESS IN BRAIN RESEARCH VOLUME 139 VASOPRESSIN AND OXYTOCIN: FROM GENES TO CLINICAL APPLICATIONS

Other volumes in PROGRESS IN BRAIN RESEARCH Volume 111: Hypothalamic Integration of Circadian Rhythms, by R.M. Buijs, A. Kalsbeek, H.J. Romijn, C.M.A. Pennartz and M. Mirmiran (Eds.) – 1996, ISBN 0-444-82443-X. Volume 112: Extrageniculostriate Mechanisms Underlying Visually-Guided Orientation Behavior, by M. Norita, T. Bando and B.E. Stein (Eds.) – 1996, ISBN 0-444-82347-6. Volume 113: The Polymodal Receptor: A Gateway to Pathological Pain, by T. Kumazawa, L. Kruger and K. Mizumura (Eds.) – 1996, ISBN 0-444-82473-1. Volume 114: The Cerebellum: From Structure to Control, by C.I. de Zeeuw, P. Strata and J. Voogd (Eds.) – 1997, ISBN 0-444-82313-1. Volume 115: Brain Function in Hot Environment, by H.S. Sharma and J. Westman (Eds.) – 1998, ISBN 0-444-82377-8. Volume 116: The Glutamate Synapse as a Therapeutical Target: Molecular Organization and Pathology of the Glutamate Synapse, by O.P. Ottersen, I.A. Langmoen and L. Gjerstad (Eds.) – 1998, ISBN 0-444-82754-4. Volume 117: Neuronal Degeneration and Regeneration: From Basic Mechanisms to Prospects for Therapy, by F.W. van Leeuwen, A. Salehi, R.J. Giger, A.J.G.D. Holtmaat and J. Verhaagen (Eds.) – 1998, ISBN 0-444-82817-6. Volume 118: Nitric Oxide in Brain Development, Plasticity and Disease, by R.R. Mize, T.M. Dawson, V.L. Dawson and M.J. Friedlander (Eds.) – 1998, ISBN 0-444-82885-0. Volume 119: Advances in Brain Vasopressin, by I.J.A. Urban, J.P.H. Burbach and D. De Wied (Eds.) – 1999, ISBN 0-444-50080-4. Volume 120: Nucleotides and their Receptors in the Nervous System, by P. Illes and H. Zimmermann (Eds.) – 1999, ISBN 0-444-50082-0. Volume 121: Disorders of Brain, Behavior and Cognition: The Neurocomputational Perspective, by J.A. Reggia, E. Ruppin and D. Glanzman (Eds.) – 1999, ISBN 0-444-50175-4. Volume 122: The Biological Basis for Mind Body Interactions, by E.A. Mayer and C.B. Saper (Eds.) – 1999, ISBN 0-444-50049-9. Volume 123: Peripheral and Spinal Mechanisms in the Neural Control of Movement, by M.D. Binder (Ed.) – 1999, ISBN 0-444-50288-2. Volume 124: Cerebellar Modules: Molecules, Morphology and Function, by N.M. Gerrits, T.J.H. Ruigrok and C.E. De Zeeuw (Eds.) – 2000, ISBN 0-444-50108-8. Volume 125: Volume Transmission Revisited, by L.F. Agnati, K. Fuxe, C. Nicholson and E. Syková (Eds.) – 2000, ISBN 0-444-50314-5. Volume 126: Cognition, Emotion and Autonomic Responses: The Integrative Role of the Prefrontal Cortex and Limbic Structures, by H.B.M. Uylings, C.G. Van Eden, J.P.C. De Bruin, M.G.P. Feenstra and C.M.A. Pennartz (Eds.) – 2000, ISBN 0-444-50332-3. Volume 127: Neural Transplantation II. Novel Cell Therapies for CNS Disorders, by S.B. Dunnett and A. Björklund (Eds.) – 2000, ISBN 0-444-50109-6. Volume 128: Neural Plasticity and Regeneration, by F.J. Seil (Ed.) – 2000, ISBN 0-444-50209-2. Volume 129: Nervous System Plasticity and Chronic Pain, by J. Sandkühler, B. Bromm and G.F. Gebhart (Eds.) – 2000, ISBN 0-444-50509-1. Volume 130: Advances in Neural Population Coding, by M.A.L. Nicolelis (Ed.) – 2001, ISBN 0-444-50110-X. Volume 131: Concepts and Challenges in Retinal Biology, by H. Kolb, H. Ripps, and S. Wu (Eds.), – 2001, ISBN 0-444-50677-2. Volume 132: Glial Cell Function, by B. Castellano López and M. Nieto-Sampedro (Eds.) – 2001, ISBN 0-444-50508-3. Volume 133: The Maternal Brain. Neurobiological and neuroendocrine adaptation and disorders in pregnancy and post partum, by J.A. Russell, A.J. Douglas, R.J. Windle and C.D. Ingram (Eds.) – 2001, ISBN 0-444-50548-2. Volume 134: Vision: From Neurons to Cognition, by C. Casanova and M. Ptito (Eds.) – 2001, ISBN 0-44450586-5. Volume 135: Do Seizures Damage the Brain, by A. Pitkänen and T. Sutula (Eds.) – 2002, ISBN 0-444-50814-7. Volume 136: Changing Views of Cajal’s Neuron, by E.C. Azmitia, J. DeFelipe, E.G. Jones, P. Rakic and C.E. Ribak (Eds.) – 2002, ISBN 0-444-50815-5. Volume 137: Spinal Cord Trauma: Regeneration, Neural Repair and Functional Recovery, by L. McKerracher, G. Doucet and S. Rossignol (Eds.) – 2002, ISBN 0-444-50817-1. Volume 138: Plasticity in the Adult Brain: From Genes to Neurotherapy, by M.A. Hofman, G.J. Boer, A.J.G.D. Holtmaat, E.J.W. Van Someren, J. Verhaagen and D.F. Swaab (Eds.) – 2002, ISBN 0-444-50981-X.

PROGRESS IN BRAIN RESEARCH VOLUME 139

VASOPRESSIN AND OXYTOCIN: FROM GENES TO CLINICAL APPLICATIONS

EDITED BY

DOMINIQUE POULAIN STÉPHANE OLIET DIONYSIA THEODOSIS INSERM U.378, Neurobiologie Morphofonctionnelle, Institut François Magendie 1, Rue Camille St.-Saëns, 33077 Bordeaux Cedex, France

ELSEVIER AMSTERDAM – BOSTON – LONDON – NEW YORK – OXFORD – PARIS SAN DIEGO – SAN FRANCISCO – SINGAPORE – SYDNEY – TOKYO 2002

ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands © 2002 Elsevier Science B.V. All rights reserved. This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier Science via their homepage (http://www.elsevier.com) by selecting ‘Customer support’ and then ‘Permissions’. Alternatively you can send an e-mail to: [email protected], or fax to: (+44) 1865 853333. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) 978 7508400, fax: (+1) 978 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 207 631 5555, fax: (+44) 207 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier Science Global Rights Department, at the fax and e-mail addresses noted above. Notice 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 rapid advances in the medical sciences, in particular, independent verification of diagnoses and drugs dosages should be made.

First edition 2002 Library of Congress Cataloging in Publication Data World Congress on Neurohypophysial Hormones (2001: Bordeaux, France) Vasopressin and oxytocin / edited by Dominique Poulain, Stéphane Oliet, Dionysia Theodosis p. cm. – (Progress in brain research; v. 139) "The articles compromising this volume were first presented at the World Congress on Neurohypophysial Hormones, held in Bordeaux, France on September 8–12, 2001"--Pref. Includes bibliographical references index. ISBN 0-444-50982-8 1. Vasopressin--Physiological effect--Congresses. 2. Oxytocin--Physiological effect--Congresses. I. Poulain, Dominique. II. Oliet, Stéphane. III. Theodosis, Dionysia. IV. Series. QP376 P7 vol. 139 [QP572.V3] 612.8 2s--dc2l [612.4 92] 2002026512 British Library Cataloguing in Publication Data Vasopressin and oxytocin: from genes to clinical applications. - (Progress in brain research; v. 139) 1. Brain – Congresses 2. Vasopressin – Congresses 3. Oxytocin – Congresses I. Poulain, Dominique II. Oliet, Stéphane III. Theodosis, Dionysia 612.8’2 ISBN 0444509828

ISBN: ISBN: ISSN:

0-444-50982-8 (volume) 0-444-80104-9 (series) 0079-6123

∞ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).  Printed in The Netherlands.

v

List of Contributors

K.C. Abbott, Nephrology Service, Walter Reed Army Medical Center, Washington, DC 20307, USA M. Åkerlund, Department of Obstetrics and Gynecology, University Hospital, S-221 85, Lund, Sweden J. Antunes-Rodrigues, Department of Physiology, School of Medicine, University of Sao Paulo, 14049 Ribeirao Preto, Sao Paulo, Brazil U. Bahner, Klinik und Poliklinik für Innere Medizin II, Universitätsklinikum, Franz-JosefStrauss-Allee 11, 93053 Regensburg, Germany J.S. Bains, Department of Physiology and Biophysics, Neuroscience Research Group, University of Calgary, Calgary, AB, Canada G.L. Bakris, Rush University Hypertension Center, Department of Preventive Medicine, Rush Presbyterian – St. Luke’s Medical Center, 1700 W. Van Buren, Suite 470, Chicago, IL 60612, USA C. Barberis, INSERM U469, 141 rue de la Cardonille, 34094 Montpellier, Cedex 5, France C. Boudaba, Department of Cell and Molecular Biology, Neurobiology Division, Tulane University, 2000 Percival Stern Hall, New Orleans, LA 70118, USA C.W. Bourque, L7-216, Neurology Division, Centre for Research in Neuroscience, Montreal General Hospital and McGill University, 1650 Cedar Avenue, Montreal, QC H3G 1A4, Canada S.B. Bruno, Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri–Columbia, Research Park, Columbia, MO 65211, USA A.B. Brussaard, Department of Experimental Neurophysiology, Vrije University Amsterdam, de Boelelaan 1087, 1081 HV Amsterdam, The Netherlands R.M. Buijs, Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands K. Burger, Institute of Biochemistry, Johannes Gutenberg - University of Mainz, Becherweg 30, D-55099 Mainz, Germany Y. Chakfe, Neurobiology Unit, Montreal Neurological Institute, 3801 University Street, Montreal, QC H3G 2B4, Canada P. Coles, Departments of Medicine and Biochemistry, Division of Clinical and Molecular Endocrinology, Case Western Reserve University, School of Medicine, 10900 Euclid Avenue, Cleveland, OH 44106-4951, USA J.M. Coulson, Departments of Physiology and Human Anatomy and Cell Biology, University of Liverpool, Liverpool, Crown Street, Liverpool, L69 3BX, UK J.T. Cunningham, Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri–Columbia, Research Park, Columbia, MO 65211, USA G. Dayanithi, Department of Neurobiology, U432-INSERM, University of Montpellier II, Place Eugene Bataillon, F-34094 Montpellier, Cedex 5, France

vi

Y. De Keyzer, CNRS UPR 1524, Institut Cochin de Génétique Moléculaire, 24 rue du Faubourg Saint Jacques, 75014 Paris, France T.L. Dellovade, Neuroscience Department, Curis Inc., 61 Moulton Street, Cambridge, MA 02138, USA S. Di, Department of Cell and Molecular Biology, Neurobiology Division, Tulane University, 2000 Percival Stern Hall, New Orleans, LA 70118, USA T. Durroux, INSERM U469, 141 rue de la Cardonille, 34094 Montpellier, Cedex 5, France C.A. Ecelbarger, Department of Medicine, Building D, Room 232, Division of Endocrinology and Metabolism, Georgetown University, 4000 Reservoir Road NW, Washington, DC 20007, USA F. Fahrenholz, Institute of Biochemistry, Johannes Gutenberg - University of Mainz, Becherweg 30, D-55099 Mainz, Germany R.L. Fields, Laboratory of Neurochemistry, Section on Molecular Neuroscience, NINDS/ NIH, Building 36, Room 4D04, Bethesda, MD 20892-4130, USA G. Gaibelet, AVIDIS S.A., Biopôle Clermont-Limagne, 63360 Saint-Beauzire, France H. Gainer, Laboratory of Neurochemistry, Section on Molecular Neuroscience, NINDS/ NIH, Building 36, Room 4D04, Bethesda, MD 20892-4130, USA G. Garcia, Exploratory Research Department, Sanofi-Synthelabo Recherche, 371 Rue du Professeur J. Blayac, 34184 Montpellier, Cedex 04, France G. Gimpl, Institute of Biochemistry, Johannes Gutenberg - University of Mainz, Becherweg 30, D-55099 Mainz, Germany R.J. Grindstaff, Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri–Columbia, Research Park, Columbia, MO 65211, USA R.R. Grindstaff, Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri–Columbia, Research Park, Columbia, MO 65211, USA J. Gutkowska, Laboratory of Cardiovascular Biochemistry, Research Centre CHUM –Hôtel Dieu, 3850 St. Urbain Street, Masson Pavilion, Montreal, QC H2W 1T8, Canada K.H.R. Higgs, Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri–Columbia, Research Park, Columbia, MO 65211, USA M. Hirasawa, Neuroscience Research Group, and Department of Physiology and Biophysics, University of Calgary, Calgary, AB, Canada S.B. House, Laboratory of Neurochemistry, Section on Molecular Neuroscience, NINDS/ NIH, Building 36, Room 4D04, Bethesda, MD 20892-4130, USA N. Hussy, CNRS-UMR 5101, CCIPE, 141 rue de la Cardonille, 34094 Montpellier, Cedex 5, France M. Jankowski, Laboratory of Cardiovascular Biochemistry, Research Centre CHUM – Hôtel Dieu, 3850 St. Urbain Street, Masson Pavilion, Montreal, QC H2W 1T7, Canada I. Kächele, Institut für Zellbiochemie und Klinische Neurobiologie, Universität Hamburg, Martinistrasse 52, 20246 Hamburg, Germany A. Kalsbeek, Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands M.A. Knepper, Laboratory of Kidney and Electrolyte Metabolism, National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, MD, USA J.-J. Koksma, Department of Experimental Neurophysiology, Research Institute Neurosciences, Vrije Universiteit Amsterdam, de Boelelaan 1087, 1081 HV Amsterdam, The Netherlands S.B. Kombian, Faculty of Pharmacy, Kuwait University, Kuwait

vii

K.J. Kovács, Laboratory of Molecular Neuroendocrinology, Institute of Experimental Medicine, Szigony u. 43, H-1083 Budapest, Hungary G. Le Fur, Sanofi-Synthelabo Recherche, 174 Avenue de France, 75635 Paris, Cedex 13, France G. Leng, University of Edinburgh Medical School, Department of Biomedical Sciences, George Square, Edinburgh, EH8 9XD, UK Y. Li, Department of Pharmacology and Toxicology, Wright State University, 3640 Colonel Glenn Highway, Dayton, OH 45435, USA M. Ludwig, Department of Biomedical Sciences, University of Edinburgh Medical School, George Square, Edinburgh EH8 9XD, UK J.P. Maffrand, Exploratory Research Department, Sanofi-Synthelabo Recherche, 195 Route d’Espagne, 31036 Toulouse, Cedex, France D. Mazzella, Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri–Columbia, Research Park, Columbia, MO 65211, USA S.M. McCann, Pennington Biomedical Research Center (LSU), 6400 Perkins Road, Baton Rouge, LA 70808-4124, USA I. Merchenthaler, The Women’s Health Research Institute, Wyeth Research, Collegeville, PA 19426, USA E. Mohr, Institut für Zellbiochemie und Klinische Neurobiologie, Universität Hamburg, Martinistrasse 52, 20246 Hamburg, Germany D. Mouginot, CHUL Research Center, Neuroscience Unit, Laval University, Saint-Foy, QC, Canada B. Mouillac, INSERM U469, 141 rue de la Cardonille, 34094 Montpellier, Cedex 5, France F. Muders, Klinik und Poliklinik für Innere Medizin II, Universitätsklinikum, Franz-JosefStrauss-Allee 11, 93053 Regensburg, Germany C. Mullin, Institut für Zellbiochemie und Klinische Neurobiologie, Universität Hamburg, Martinistrasse 52, 20246 Hamburg, Germany T. Murase, The First Department of Internal Medicine, Nagoya University, Nagoya, Japan I.D. Neumann, Institute of Zoology, University of Regensburg, 93040 Regensburg, Germany S. Nielsen, Department of Cell Biology, University of Aarhus, Aarhus, Denmark M. Palkovits, Laboratory of Neuromorphology, Semmelweis University, Budapest, Hungary I.F. Palm, Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands M. Pascal, Exploratory Research Department, Sanofi-Synthelabo Recherche, 195 Route d’Espagne, 31036 Toulouse, Cedex, France M. Petersson, Department of Molecular Medicine, Endocrine and Diabetes Unit, Karolinska Institutet, S-171 77 Stockholm, Sweden Q.J. Pittman, Neuroscience Research Group, and Department of Physiology and Biophysics, University of Calgary, Calgary, AB, Canada P. René, CNRS UPR 1524, Institut Cochin de Génétique Moléculaire, 24 rue du Faubourg Saint Jacques, 75014 Paris, France D.S. Richards, Department of Pharmacology and Toxicology, Wright State University, 3640 Colonel Glenn Highway, Dayton, OH 45435, USA D. Richter, Institut für Zellbiochemie und Klinische Neurobiologie, Universität Hamburg, Martinistrasse 52, 20246 Hamburg, Germany

viii

G.A.J. Riegger, Klinik und Poliklinik für Innere Medizin II, Universitätsklinikum, FranzJosef-Strauss-Allee 11, 93053 Regensburg, Germany M. Rusnak, Laboratory of Neurochemistry, Section on Molecular Neuroscience, NINDS/ NIH, Building 36, Room 4D04, Bethesda, MD 20892-4130, USA J.A. Russell, Department of Biomedical Sciences, University of Edinburgh Medical School, George Square, Edinburgh, EH8 9XD, UK N. Sabatier, Department of Biomedical Sciences, University of Edinburgh Medical School, George Square, Edinburgh EH8 9XD, UK J.-C. Schellenberg, School of Biological Sciences, The University of Auckland, 3A Symonds Street, Auckland, Australia T. Sen, INSERM U469, 141 rue de la Cardonille, 34094 Montpellier, Cedex 5, France C. Serradeil-Le Gal, Exploratory Research Department, Sanofi-Synthelabo Recherche, 195 Route d’Espagne, 31036 Toulouse, Cedex, France M. Shoham, Departments of Medicine and Biochemistry, Division of Clinical and Molecular Endocrinology, Case Western University School of Medicine, 10900 Euclid Avenue, Cleveland, OH 44106-4951, USA P.J. Shughrue, Department of Neuroscience, Merck Research Laboratories, Sumneytown Pike and Broad Street, WP26A-3000, West Point, PA 19486, USA J.E. Stern, Department of Pharmacology and Toxicology, Wright State University, 3640 Colonel Glenn Highway, Dayton, OH 45435, USA M.J. Sullivan, Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri–Columbia, Research Park, Columbia, MO 65211, USA J.G. Tasker, Department of Cell and Molecular Biology, Neurobiology Division, Tulane University, 2000 Percival Stern Hall, New Orleans, LA 70118, USA A. Thibonnier, University of Michigan, School of Engineering, Ann Arbor, MI, USA M. Thibonnier, Departments of Medicine and Biochemistry, Room BRB 431, Division of Clinical and Molecular Endocrinology, Case Western University, School of Medicine, 10900 Euclid Avenue, Cleveland, OH 44106-4951, USA Y. Tian, Department of Medicine, Building D, Room 232, Division of Endocrinology and Metabolism, Georgetown University, 4000 Reservoir Road NW, Washington, DC 20007, USA G. Valette, Exploratory Research Department, Sanofi-Synthelabo Recherche, 195 Route d’Espagne, 31036 Toulouse, Cedex, France J.G. Verbalis, Department of Medicine, Building D, Room 232, Division of Endocrinology and Metabolism, Georgetown University, 4000 Reservoir Road NW, Washington, DC 20007, USA D.L. Voisin, Laboratoire de Physiologie Oro-Faciale, Faculté de Chirurgie Dentaire, Clermont Ferrand, 63000, France J. Wagnon, Exploratory Research Department, Sanofi-Synthelabo Recherche, 371 Rue du Professeur J. Blayac, 34184 Montpellier, Cedex 04, France V. Wiegand, Institute of Biochemsitry, Johannes Gutenberg - University of Mainz, Becherweg 30, D-55099 Mainz, Germany M. Yamashita, Laboratory of Neurochemistry, Section on Molecular Neuroscience, NINDS/ NIH, Building 36, Room 4D04, Bethesda, MD 20892-4130, USA

ix

Preface

The articles comprising this volume were first presented at the World Congress on Neurohypophysial Hormones held in Bordeaux, France on September 8–12, 2001. This conference brought together more than 170 scientists from 18 countries who belong to the different fields of interest representing research in the hypothalamo–neurohypophysial system. This meeting follows that held in Edinburgh in 1999 and a series of International Meetings on oxytocin or vasopressin or on the neurohypophysis in general, held sporadically but very successfully over the past five decades, since the first meeting in Bristol in 1956. The next World Congress on Neurohypophysial Hormones will be held in Kyoto, Japan, and its organization, which is in the competent hands of Professor M. Kawata, is well underway. The two neurohypophysial neurohormones, oxytocin and vasopressin, exert a variety of central and peripheral actions and thus involve different scientific domains, which too often, even to-day, do not always find the appropriate occasion to interact. The meeting offered such an opportunity. This volume, therefore, is composed of chapters dealing with topics varying from basic and clinical neurosciences and neuroendocrinology, to reproductive, renal, cardiovascular physiology and pathology. It thus encompasses all areas of current neurohypophysial research and should be of vital interest as an integrative reference volume to specialized investigators and as an excellent introductory text to students, scientists and clinicians not yet closely familiar with the field. To ensure novelty and to make sure that all topics of current importance were covered, plenary and symposium speakers as well as poster presentations concentrated on recent advances made in the last few years. In the course of the conference, a particularly tragic event in modern history occurred, that will leave long-lasting unhappy memories. The quality and friendliness of the scientific and intellectual exchanges that took place between participants from so many different countries and cultural oulooks should remind us that it is always possible, albeit difficult, to overcome the dire consequences of such events. Dominique Poulain Stéphane Oliet Dionysia Theodosis

This Page Intentionally Left Blank

xi

Acknowledgements

The Local Organizing Committee, composed of Dominique Poulain (Cochair), Dionysia Theodosis (Cochair), Stéphane Oliet (Secretary), Frank Couillaud and Jean Marc Israel, wishes to express its thanks to the members of the International Scientific Advisory Board, that included W. Armstrong, C. Barberis, L. Burrell, P. Burbach, B. Chini, T. Day, A. Fergusson, F. Grant, T. Higuchi, J.J. Legros, J. Lemos, F. Moos, D. Murphy, I. Neumann, S. Nielsen, Q. Pittman, G. Robertson, W. Rosenthal, J. Russell, E. Szczepanska-Sadowska and K. Yagi, who put together an excellent Scientific Program and helped to shape the Conference. We also thank the French Neuroscience Society, CNRS, INSERM, ARMA, and Sanofi-Synthélabo for their financial support. We are particularly grateful to the members of the INSERM Unit 378 in Bordeaux and to Clémence Fouquet, Isabelle Conjat and Jean-Francis Renaudon who assisted with the scientific, secretarial and social work and contributed to make the Conference an international success.

This Page Intentionally Left Blank

xiii

Contents

List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

1. The magnocellular neuronal phenotype: cell-specific gene expression in the hypothalamo-neurohypophysial system H. Gainer, M. Yamashita, R.L. Fields, S.B. House and M. Rusnak (Bethesda, MD, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

2. Estrogen modulates oxytocin gene expression in regions of the rat supraoptic and paraventricular nuclei that contain estrogen receptor-β P.J. Shughrue, T.L. Dellovade and I. Merchenthaler (Collegeville, PA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15

3. Short-term modulation of GABAA receptor function in the adult female rat A.B. Brussaard and J.-J. Koksma (Amsterdam, The Netherlands) . . . .

31

4. Cholesterol and steroid hormones: modulators of oxytocin receptor function G. Gimpl, V. Wiegand, K. Burger and F. Fahrenholz (Mainz, Germany) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43

5. Central vasopressin systems and steroid hormones A. Kalsbeek, I.F. Palm and R.M. Buijs (Amsterdam, The Netherlands) 57 6. Regulation of renal salt and water transporters during vasopressin escape C.A. Ecelbarger, T. Murase, Y. Tian, S. Nielsen, M.A. Knepper and J.G. Verbalis (Washington, DC, Bethesda, MD, USA, Nagoya, Japan and Aarhus, Denmark) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

75

7. Stretch-inactivated cation channels: cellular targets for modulation of osmosensitivity in supraoptic neurons C.W. Bourque, D.L. Voisin and Y. Chakfe (Montreal, QC, Canada and Clermont Ferrand, France) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

85

8. Glial cells in the hypothalamo-neurohypophysial system: key elements of the regulation of neuronal electrical and secretory activity N. Hussy (Montpellier, France) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95

xiv

9. Functional synaptic plasticity in hypothalamic magnocellular neurons J.G. Tasker, S. Di and C. Boudaba (New Orleans, LA, USA) . . . . . . . . 113 10. Postsynaptic GABAB receptors in supraoptic oxytocin and vasopressin neurons J.E. Stern, Y. Li and D.S. Richards (Dayton, OH, USA) . . . . . . . . . . . . . 121 11. Neurohypophyseal hormones in the integration of physiological responses to immune challenges K.J. Kovács (Budapest, Hungary) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 12. Involvement of the brain oxytocin system in stress coping: interactions with the hypothalamo–pituitary–adrenal axis I.D. Neumann (Regensburg, Germany) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 13. Expression of human vasopressin and oxytocin receptors in Escherichia coli B. Mouillac, T. Sen, T. Durroux, G. Gaibelet and C. Barberis (Montpellier and Saint-Beauzire, France) . . . . . . . . . . . . . . . . . . . . . . . . . . 163 14. Molecular pharmacology and modeling of vasopressin receptors M. Thibonnier, P. Coles, A. Thibonnier and M. Shoham (Cleveland, OH and Ann Arbor, MI, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 15. Nonpeptide vasopressin receptor antagonists: development of selective and orally active V1a , V2 and V1b receptor ligands C. Serradeil-Le Gal, J. Wagnon, G. Valette, G. Garcia, M. Pascal, J.P. Maffrand and G. Le Fur (Toulouse, Montpellier and Paris, France) 197 16. Rat vasopressin mRNA: a model system to characterize cis-acting elements and trans-acting factors involved in dendritic mRNA sorting E. Mohr, I. Kächele, C. Mullin and D. Richter (Hamburg, Germany) 211 17. Dendritic action potentials in magnocellular neurons J.S. Bains (Calgary, AB, Canada) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 18. Modulation of synaptic transmission by oxytocin and vasopressin in the supraoptic nucleus S.B. Kombian, M. Hirasawa, D. Mouginot and Q.J. Pittman (Kuwait, Calgary, AB, and Sainte-Foy, QC, Canada) . . . . . . . . . . . . . . . . . . . . . . . . . 235 19. The active role of dendrites in the regulation of magnocellular neurosecretory cell behavior M. Ludwig, N. Sabatier, G. Dayanithi, J.A. Russell and G. Leng (Edinburgh, UK and Montpellier, France) . . . . . . . . . . . . . . . . . . . . . . . . . . 247 20. Cardiovascular regulation of supraoptic vasopressin neurons J.T. Cunningham, S.B. Bruno, R.R. Grindstaff, R.J. Grindstaff, K.H.R. Higgs, D. Mazzella and M.J. Sullivan (Columbia, MO, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

xv

21. The central vasopressinergic system in experimental left ventricular hypertrophy and dysfunction F. Muders, G.A.J. Riegger, U. Bahner and M. Palkovits (Regensburg and Würzburg, Germany and Budapest, Hungary) . . . . . . . . . . . . . . . . . . 275 22. Cardiovascular effects of oxytocin M. Petersson (Stockholm, Sweden) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 23. Treatment of the diabetic patient: focus on cardiovascular and renal risk reduction K.C. Abbott and G.L. Bakris (Washington, DC and Chicago, IL, USA) 289 24. Oxytocin in parturition of guinea pigs, humans, and other species J.-C. Schellenberg (Geneva, Switzerland) . . . . . . . . . . . . . . . . . . . . . . . . . . 299 25. Oxytocin, vasopressin and atrial natriuretic peptide control body fluid homeostasis by action on their receptors in brain, cardiovascular system and kidney S.M. McCann, J. Antunes-Rodrigues, M. Jankowski and J. Gutkowska (Baton, Rouge, LA, USA, Sao Paulo, Brazil and Montreal, QC, Canada) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 26. Positive and negative regulators of the vasopressin gene promoter in small cell lung cancer J.M. Coulson (Liverpool, UK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 27. The vasopressin receptor of corticotroph pituitary cells P. René and Y. de Keyzer (Paris, France) . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 28. Involvement of oxytocin and vasopressin in the pathophysiology of preterm labor and primary dysmenorrhea M. Åkerlund (Lund, Sweden) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359 Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 Published by Elsevier Science B.V.

CHAPTER 1

The magnocellular neuronal phenotype: cell-specific gene expression in the hypothalamo-neurohypophysial system H. Gainer ∗ , M. Yamashita 1 , R.L. Fields, S.B. House and M. Rusnak Section on Molecular Neuroscience, Laboratory of Neurochemistry, NINDS, National Institutes of Health, Bethesda, MD 20892, USA

Abstract: The magnocellular oxytocin (OT) and vasopressin (VP) neurons of the hypothalamo-neurohypophysial system are exceptional cell biological models to study mechanisms of cell-specific gene expression and neurosecretion of neuropeptides in the central nervous system. Single cell differential gene expression experiments have further defined these phenotypes by identifying novel and distinct regulatory molecules in these neurons. Transgenic mouse studies have led to the intergenic region (IGR) hypothesis, which states that the DNA sequences between the OT- and VP-genes contain critical enhancer sites for their cell-specific expression. The recent cloning and sequencing of the human IGR, and its comparison with the mouse IGR sequence has identified conserved sequences as putative, cell-specific enhancer sites which are now being evaluated by biolistic transfections of organotypic hypothalamic cultures. With these data, it is possible to target the gene expression of specific molecules to magnocellular neurons both in vivo and in vitro, in order to perturb and/or visualize neurosecretory and other processes. Keywords: Oxytocin; Vasopressin; Cell-specific gene expression; Magnocellular neuron; Apoptosis; Biolistics; Neurosecretion

Introduction The magnocellular neurons (MCNs) of the mammalian hypothalamo-neurohypophysial system (HNS) synthesize and secrete the nonapeptides, oxytocin (OT) and vasopressin (VP) at exceptionally high rates. Because of this property, these neuroendocrine cells have served as important model systems for the study of mechanisms of peptide biosynthesis and secretion in the central nervous system (CNS). In addition, the relatively easy access to the MCN cell

∗ Correspondence to: H. Gainer, Laboratory of Neurochemistry, NINDS/NIH, Building 36, Room 4D04, Bethesda, MD 20892-4130, USA. Tel.: +1-301-496-6719; Fax: +1-301-496-1339; E-mail: [email protected] 1 Present address: Department of Biotechnology, Graduate School of Engineering, Osaka University, Yamadaoka 2-1, Suita, Osaka 565-0871, Japan.

bodies in the paraventricular (PVN) and supraoptic (SON) nuclei and axons in the median eminence by both stereotaxic and micropunch methods, and to their nerve terminals due to their presence in the posterior pituitary outside of the blood–brain barrier (Fig. 1), have made these neurons favorite subjects for many biochemical and physiological studies (Morris et al., 1978; Hatton, 1990, 1997a,b; Meister, 1993; Gainer and Wray, 1994; Armstrong, 1995; Burbach et al., 2001). Ever since it was first described in the fish central nervous system over 70 years ago (Scharrer, 1928, 1987), the HNS has been a veritable ‘rosetta stone’ for neuroendocrinology and neuroscience. Many seminal findings have been made using the HNS. These include the first discovery and characterization of neuropeptides (Du Vigneaud, 1954), the early development of peptide agonists and antagonists (Jard et al., 1987; Manning et al., 1987; Manning and Sawyer, 1989), and the first proposal of the prohormone con-

2

Fig. 1. Oxytocin- (OT) and vasopressin- (VP) synthesizing neuronal systems in the hypothalamus. These include the MCNs in the paraventricular nuclei (PVN) and supraoptic nuclei (SON), constituting the hypothalamo-neurohypophysial system (HNS), parvocellular (CRH) neurons in the PVN projecting to the median eminence (ME), and in the suprachiasmatic nuclei (SCN). The inset shows immunocytochemically identified OT and VP neurons in the PVN stained gray and black, respectively. Abbreviations: A, anterior; I, intermediate; and P, posterior regions of the pituitary. Adapted from Brownstein et al. (1980).

cept (Sachs and Takabatake, 1964; Sachs et al., 1969). In addition, MCN studies provided the first full characterization of bursting pacemaker activity in CNS neurons (Leng, 1988), the first demonstration of neuropeptide secretion from dendrites (Ludwig, 1998), novel views of glial–neuronal interactive plasticity (Theodosis and Poulain, 1987; Hatton, 1997a) and glial influence on synaptic transmission (Oliet et al., 2001), and the demonstration that peptides can produce complex behaviors (Reijmers et al., 1998; Wang et al., 1998; Insel and Young, 2001). However, despite intense investigations by many laboratories over the past 25 years to address the central question of how MCNs produce such high rates of OT and VP transcription, biosynthesis and secretion in a cell-specific manner (Burbach et al., 2001), the answer remains elusive.

The MCN phenotype The MCNs in the HNS have been classified into two distinct phenotypes, the OT and VP neurons. This historic classification (Vandersande and Dierickx, 1975; Dierickx et al., 1978) has been reinforced by many immunocytochemical (ICC) and in situ hybridization histochemical (ISHH) studies. As a result, it became the generally accepted view that expression of the OT and VP genes was mutually exclusive and occurred in separate cells in the HNS (Mohr et al., 1988). These two phenotypes in rats have also been discriminated by their distinct electrical activity patterns in response to sustained physiological stimuli, i.e. OT cells increase their action potential frequencies with regular, continuous firing, whereas VP cells generate phasic bursting patterns in response to steady depolarization (Leng, 1988).

3

These different properties have been attributed, in part, to higher levels of calcium-binding proteins (e.g. calbindin, calretinin, etc.) in OT neurons which prevent depolarizing after potentials considered critical for bursting, and to a novel non-activating outward potassium current found only in OT cells (Hatton, 1990, 1997b; Armstrong, 1995; Armstrong and Stern, 1998). The latter may underlie the milkejection high-frequency bursts observed only in OT cells. While the division of the MCNs into OT and VP phenotypes is generally accepted, these neurons also contain smaller amounts of other coexisting peptides (e.g. galanin, cholecystokinin, CRH, dynorphin, enkephalin, TRH, etc.) which can vary between cells depending on functional conditions (Meister, 1993), and suggesting a more subtle heterogeneity within the OT and VP neuronal phenotypes. Most interesting is that several laboratories, using ICC and ISHH, have reported coexistence between OT and VP peptides (Mezey and Kiss, 1991) and mRNAs (Kiyama and Emson, 1990; Glasgow et al., 1999; Xi et al., 1999) in about 1–3% of MCNs in the normal rat SON, and that this increases substantially to 17% after 2 days of lactation (Mezey and Kiss, 1991; Glasgow et al., 1999). Thus, while the segregation of OT and VP gene expression in separate cells in the HNS is the rule, it is clear that substantial coexistence of these two robustly expressed peptides can occur under some physiological conditions. Re-examination of the OT and VP MCN phenotypes by single-cell gene expression analysis In recent years, analyses of gene expression profiles in single neurons under various biological conditions (Eberwine et al., 1992) have provided new opportunities to examine the phenotypic diversity of cells in a complex organ such as the brain (Mackler and Eberwine, 1993). In a number of cases, cDNA libraries have been constructed from single neurons (Dulac and Axel, 1995; Merz et al., 1995; Korneev et al., 1996), and differential screening of cDNA libraries constructed from single olfactory neurons has led to the identification of pheromone receptors in mammals (Dulac and Axel, 1995). The magnocellular OT and VP neurons are excellent candidates for such a differential analysis. Because of their large size, and given their high rates of transcription, it

might be possible to identify unique mRNAs encoding specific transcriptional regulatory factors responsible for the cell-specific peptide gene expression in these cells by this approach. A strategy that we have used on the MCNs for this purpose is illustrated in Fig. 2. For the differential screening of OT- and VP-cell cDNA libraries, it is essential that the cDNAs being used be derived from an unequivocal cellular phenotype. We chose to use MCNs from lactating rats because OT and VP gene expression and the nonapeptide secretions are greatly upregulated in the HNS under this physiological condition. Since the chance that an isolated single cell taken at random from lactating rats is of the OT/VP coexistent phenotype is about 17% (Mezey and Kiss, 1991; Glasgow et al., 1999), we first evaluated whether the OT and VP expression phenotypes of single isolated MCNs from dissected rat SONs could be determined by RTPCR procedures. We found that even without PCR amplification, only 1–5% of a single MCN’s cDNA sample was necessary for an unequivocal assessment of the phenotypes, and the remaining cDNA could then be assayed for at least 25 other specific genes (Glasgow et al., 1999). For the preparation of cDNA libraries we amplified with PCR the cDNAs before the phenotype analysis (Fig. 2). Using the protocol described in Fig. 2, we successfully constructed separate cDNA libraries for the OT and VP neuronal phenotypes. Each library was derived from ten unequivocally characterized OT or VP MCN cDNAs which were pooled in order to minimize experimental variability. After completion of the differential screening of these cDNA libraries, we sequenced all the differentially expressed clones. A preliminary functional characterization of the clones that were obtained, based on southern blot analysis and DNA sequences, is shown in Table 1. OT and VP mRNAs were obtained only from the OT or VP cDNA libraries, respectively, thereby validating their specificities. Interestingly, most of the differentially expressed genes that we found were of unknown function, and preliminary in situ hybridization studies have shown that three of these are expressed exclusively in the HNS (not illustrated). In summary, the ability to generate single, identified, cell-specific (OT or VP neuron) cDNA libraries has allowed us an opportunity to differentially screen

4

Fig. 2. Overview of single cell-differential hybridization strategy. Individual magnocellular neurons are isolated from dissociated rat supraoptic nuclei (see inset), their mRNAs reverse transcribed into cDNA which is then PCR amplified. The individual OT and VP MCN amplified cDNAs are phenotypically characterized, and then utilized for producing phenotypically specific cDNA libraries and finally used for differential screening. See text.

for uniquely and/or preferentially expressed genes in each neuronal phenotype. Some of these could be transcriptional-activating factors, but obviously genes encoding proteins involved in other cell functions (e.g. signal transduction, membrane receptors, etc.) would also be of interest.

Cell-specific expression of OT and VP genes in MCNs: the background The DNA sequences and structures of the OT and VP genes in several mammalian species has been known for more than a decade. The OT and VP genes each contain three exons and two introns, are found on the same chromosome (chromosome 2 in the mouse and chromosome 20 in the human) in close apposition,

5 TABLE 1 Differential analysis of single OT and VP magnocellular neuronal gene expression Oxytocin cell-specific genes Gene Oxytocin Vasopressin Secretion associated proteins Putative transfactors Unknown function

No. of hits (total = 43) 19 (44%) 0 6 1 17 (40%)

Vasopressin cell-specific genes Gene Oxytocin Vasopressin Secretion associated proteins Putative transfactors Unknown function

No. of hits (total = 51) 0 10 (20%) 2 1 38 (75%)

and in opposite transcriptional orientations (Burbach et al., 2001). Despite the availability of this information, very little is known about the mechanisms that are responsible for the cell-specific expression of these genes. This is, in large part, due to the lack of homologous cell lines that express these genes which then could serve as relevant experimental models for conventional deletion construct analyses. Extensive studies using heterologous cell lines that are derived from cells that do not normally express these genes have been very useful for the identification of various regulatory cis-elements in the OT and VP genes (Burbach et al., 2001). However, heterologous cells are by definition, irrelevant models for cell-specific gene expression analyses. Consequently, most of the data pertinent to the question of cell-specific gene expression of OT and VP genes has been obtained using transgenic models and by assaying expression in vivo in the MCNs (reviewed in Waller et al., 1998; Burbach et al., 2001; Gainer and Young, 2001). Transgenic mouse studies Experiments in transgenic mice and rats (Waller et al., 1998; Burbach et al., 2001; Gainer and Young, 2001) have led to the proposal that the cell-specific enhancers for OT and VP gene expression are not located on the 5 -upstream regions of these genes, but instead are present in the IGR downstream of the VP gene (Gainer, 1998; Gainer and Young, 2001).

This view is based on two general observations. The first is that constructs containing only 5 -flanking upstream regions of these genes (from 0.5–9 kb) connected either to intact, wild-type (usually heterologous) genes or to various reporter genes, do not produce cell-specific expression of these genes in MCNs of transgenic animals. Second, the inclusion of 3 -downstream flanking regions of the VP genes to these very same constructs can produce robust cell-specific gene expression in the magnocellular neurons. Subsequent experiments using deletion constructs of the VP component of the transgene suggested that the 3 -downstream region of the VP gene (about 0.6 kb of IGR) was the likely site of an enhancer for cell-specific expression of the OT gene. Further studies on the VP gene (Grant et al., 1993; Zeng et al., 1994; Waller et al., 1996, 1998; Jeong et al., 2001) showed that by extending the 3 downstream region of the VP gene to 2–3 kb, it was possible to get cell-specific expression of these transgenes in VP neurons only. These observations have led to a hypothesis, termed the ‘intergenic region’ (IGR) hypothesis (Gainer, 1998), which states that the cis-elements in the genomic DNA responsible for the cell-specific expression of OT- and VP-genes are located downstream of exon III of the VP gene. Insights from comparative genomics A useful theoretical approach is to compare the IGR sequences in different species, in order to identify conserved sequences that could be candidate regulatory elements. The recent demonstration that isotocin genes from fish can be expressed in a cell-specific manner in OT neurons in transgenic rats (Venkatesh et al., 1997) suggested a significant evolutionary conservation of the cell-specific determinants. In an earlier attempt to identify these enhancer elements, a comparison was made between the mouse and rat IGR sequences, but these two sequences, with the exception of a 6.4-kb LINE insert present only in the rat, were found to be too similar to allow for discrimination of conserved sequences that might be relevant for cell-specific gene expression (Ratty et al., 1996). The mouse and rat are separated by only one million years of evolution, and hence there appeared to be insufficient time for the divergence of non-essential sequences. In contrast, the human and

6 TABLE 2 Conserved sequences in mouse and human IGRs Motif

Position a in mouse IGR

Position a in human IGR

Putative OT-enhancer cluster AGGGGAG GTCGTG AGGAAGCGATC AGAGAGG GGACAA AACTGCTA CTTGACC

43 107 173 321 340 781 807

24 93 130 275 294 918 685

Putative VP-enhancer cluster TTCTCTCT GCCACA GGGATGCTCTGCA CTAAGGTCAGGGG CAGAGTC CTGGGT ACAGGGACCA CTCAGG GGCAGGT TTCAATCC GTGTATGGA TCCTCT AATCAGGCT GAGTCCC CAGCTACAGAGCCAT

1293 1358 1379 1396 1479 1552 1571 1618 1693 1717 1730 1777 1832 1921 1933

1097 1167 1444 1461 1316 1428 1447 1508 1597 1621 1894 1681 1739 1844 1857

Other conserved sequences AGGCCAG GCCCCCTCAACCTCT TCCCGTTTC CAATTAGACACCAGC

3211 3222 3541 3553

9592 9603 9994 10006

Analogous transfactor b GKLF, Ik-2, MZF1, GC BOX, SP1

V-Myb ER

Ik-2 RORα1, V-ErbA

Ik-2 TCF11, AP-1 δEF1 Gfi-1

RORα1 Ik-2 Ap-4, CCCAAT box

a

Position numbers start from the first base following exon III of the VP gene in the mouse and human IGRs (i.e position 43 for the mouse would be base 43 in the mouse IGR sequence accession no. U38901 and position 24 in human would be base 24 in the human IGR sequence accession no. AF254641). b The MatInspector V2.2 program was used to search for transcription factor consensus binding sequence matches. Abbreviation: GKLF, gut enriched Kruppel-like factor (Shields and Yang, 1998); Ik-2, Ikaros genezinc finger protein (Molnar and Georgopoulos, 1994); MZF1, myeloid zinc finger protein (Morris et al., 1994); GC BOX, GC box elements (Buche, 1990); SP1, stimulating protein; V-Myb, v-Myb protein (Grotewold et al., 1994); ER, estrogen receptor; V-ErbA, viral homolog of thyroid hormone receptor α1 (Subauste and Koenig, 1995) TCF11, CNC-bZIP factor (Johnsen et al., 1998); RORα1, novel orphan hormone nuclear receptor (Giguere et al., 1994); δEF1, δ-crystallin enhancer binding protein (Sekido et al., 1994); Gfi-1, zinc finger protein (Zweidler-McKay et al., 1996); Ap-4, activator protein 4; CCCAAT box, cellular and viral CCAAT box (Buche, 1990).

mouse are separated by about 100 million years, and we reasoned that this greater evolutionary distance might allow the conserved sequences in the IGR to be identified. We subsequently cloned and sequenced the human IGR (accession no. AF254641) and compared it to the mouse IGR sequence (Gainer et al., 2001). Twenty-six sequence motifs ranging from 6 to 16 bp in length were found to be conserved between the mouse and are located in three distinct clusters

in the mouse IGR. Twelve of the 26 sequences are similar to consensus motifs for known transcription factors and five of these are zinc finger proteins (Table 2). The conserved motifs found in the first two clusters in Table 2 are locations that are consistent with the available transgenic data which suggest that the elements required for cell specific expression of OT and VP reside <1 kb and <2 kb, respectively, downstream of exon III of the VP gene (Gainer and

7

Young, 2001; Jeong et al., 2001). These conserved sequences should provide a theoretical basis for setting the boundaries of deletions in the mouse IGR for future studies of the cell specific expression of VP and OT genes in transgenic mice, or preferably, in a ‘high throughput’ system of analysis. A high throughput strategy for the study of cell-specific gene expression in the hypothalamus An in vitro model A ‘high throughput’ method to evaluate deletion constructs for the elucidation of the cis-elements involved in cell-specific gene expression in the CNS first requires in vitro models in which the distinct neuronal phenotypes found in vivo can be easily identified in the cultured system. In the case of the hypothalamus, this means that the distinct neuronal phenotypes expressing the OT and VP genes should be found in identifiable nuclei (i.e. PVN, SCN, and SON) in the in vitro model. Organotypic slice-explant cultures are known to maintain the cytoarchitectonic organization and topographic relationships of the original postnatal brain tissues (Gähwiler et al., 1997) and this is also true for organotypic cultures of hypothalamus derived from neonatal mice and rats (House et al., 1998). Applications of this technique to the rat hypothalamus have been very successful for the maintenance of OT MCNs in PVN and SON (House et al., 1998), parvocellular VP neurons in the PVN (Bertiny et al., 1993) and in the SCN (Wray et al., 1993; Belenky et al., 1996), but much less so for the VP MCNs in the rat hypothalamus (Wray et al., 1991; Vutskits et al., 1998; Rusnak et al., 2001). The finding that rat VP-MCNs do not thrive in organotypic cultures, in comparison to OT-MCNs, may be related to their relative vulnerabilities to axotomy-induced programmed cell death in vivo (Raisman, 1973; Herman et al., 1987; Dohanics et al., 1996). Dohanics et al. (1996) found that pituitary stalk compression caused 65% of the VP-MCNs as compared to 31% of the OT-MCNs in the SON to degenerate. It is likely that the axotomy of long axons which occurs in the preparation of organotypic cultures also leads to a preferential degeneration of the VP MCNs over the OT MCNs. In an insightful

paper, Vutskits et al. (1998) reported that ciliary neurotrophic factor (CNTF) was a selective survival factor for rat VP-MCNs in organotypic cultures of the rat hypothalamic paraventricular nucleus (PVN). Because of this, we examined the effects of CNTF on the survival of the MCNs in rat and mouse SONs. Our studies on the MCNs in the rat SON confirmed the survival effect of CNTF. We found that 10 ng/ml CNTF in the culture media increased the survival of the VP-MCNs by 6-fold and the OT-MCNs by 3-fold (Rusnak et al., 2002). Corresponding changes were found in both VP- and OT-mRNA levels in the SON as a result of the CNTF treatment (Fig. 3, see Rusnak et al., 2002, for quantitation). Other experiments indicated that the CNTF treatment does not increase OTand VP-gene expression, and that the increase of OT and VP mRNA observed in the SON was due to the increase in VP MCN number (Rusnak et al., 2002). The above findings all relate to studies conducted on MCNs in the rat hypothalamus. In contrast to the rat, the MCNs in the mouse hypothalamus do not appear to undergo apoptosis when placed into organotypic cultures (House et al., 1998), and longterm survival of the VP-MCNs is exceptionally good (over a 1000 cells in the SON) under normal culture conditions and in the absence of any added factors (Rusnak et al., 2002). Addition of CNTF had no effect on the survival or the mRNA content of mouse MCNs in the SON (Fig. 4). These differences between rat and mouse MCNs are surprising, and it is unclear why the mouse MCNs are so resistant to the apoptotic consequences of axotomy, which are so prominent in the rat MCNs. In any case, since the mouse MCNs survive so well in long-term culture without the addition of any extra factors, this may represent the superior model system for longterm molecular and physiological studies of MCNs in vitro. Transfection of hypothalamic neurons in organotypic culture Given the above in vitro models, the next critical requirement for a high throughput approach is to be able to transfect the cultured neurons with exogenous DNA constructs. Conventional methods such as calcium-phosphate and lipid-mediated cell transfection are not very successful when used with

8

Fig. 3. Film images of radioactive in situ hybridization for VP mRNA (left panels) and OT mRNA (right panels) in rat coronal hypothalamic slices cultured in the absence of (upper panels, Controls) and in the presence of CNTF for 15 days (lower panels). Abbreviations: PVN, paraventricular nucleus; SCN, suprachiasmatic nucleus; SON, supraoptic nucleus; ACC, accessory hypothalamic nucleus. Scale bars: 500 μm. Adapted from Rusnak et al. (2002).

organotypic cultures. This may, in part, be due to the presence of a reactive astrocyte layer which covers the cultured slice, thereby preventing access of the DNA containing vehicles to the neurons in the slice. We have found two approaches to the transfection of slice-explant cultures that were promising, the use of viral vectors and particle-mediated gene transfer. Several types of modified viral vectors have been used as efficient vehicles to transfer genes into differentiated neurons in the central nervous system in vitro and in vivo (Karpati et al., 1996; Slack and Miller, 1996). These include herpes-related viruses (Fink et al., 1996; Federoff, 1999), attenuated adenoviruses (Geddes et al., 1996, 1999), adeno-associated viruses (Kaplitt et al., 1994; Du et al., 1996), and lentiviruses (Federico, 1999). To date, only adenovirus has been used in the HNS in vivo (Geddes et al., 1996, 1997, 1999; Vascquez et al., 1998) and the adeno-associated virus has been used to transfect oxytocin cells in vitro (Keir et al., 1999). Viral vectors are typically very effective but the production of multiple viral vectors of sufficient titer accommodating the large numbers of deletion constructs necessary for promoter/enhancer analysis, would al-

most be as labor-intensive and time-consuming as the use of transgenic mice for this purpose. Hence, we turned to the second technique, particle-mediated gene transfer (or biolistics), in which a 1-μm gold particle coated with plasmid DNA is accelerated into the nuclei of cells using a burst of helium gas pressure. This method can be used to transfect any DNA construct of any size and in any configuration (Gainer et al., 2002). In principle, biolistic transfection provides a very rapid assay, comparable to the use of conventional transfection methods and cell lines for deletion construct analysis, but with the advantage of being able to be used in the more biologically relevant primary neurons. We first evaluated the particle-mediated gene transfer technique by transfecting HNS neurons in stationary organotypic cultures using OT and VP constructs that had already been shown to produce successful cell-specific expression in transgenic mouse experiments (Jeong et al., 2001). The first construct that we used contained the entire VP gene (including the three exons and two introns) plus the 3.9 kb of upstream and 2.1 kb of downstream flanking (IGR) regions, and with a CAT reporter in-

9

Fig. 4. Film images of radioactive in situ hybridization for VP mRNA (left panels) and OT mRNA (right panels) in mouse coronal hypothalamic slices cultured in the absence of (upper panels, Controls) and in the presence of CNTF for 15 days (lower panels). Abbreviations: PVN, paraventricular nucleus; SCN, suprachiasmatic nucleus; SON, supraoptic nucleus; ACC, accessory hypothalamic nucleus. Scale bar: 500 μm. Adapted from Rusnak et al. (2002).

serted in exon III. This construct, termed VP III– CAT-2, was the positive control in our biolistics experiments. Since there are so few magnocellular OT and especially VP neurons present per rat slice, and the biolistic procedure is random and with a low ‘hit’ rate, only a small number of transfected cells might be expected to be seen expressing the CAT reporter. In fact, we found a modest number of cells that had been transfected and these had the expected morphologies of OT- and VP-synthesizing neurons, and were largely located in regions of the slice (e.g. PVN, SON, and SCN) where such neurons would be expected to be found (Fig. 5A–C). The results of these experiments showed that, after biolistic transfection, the CAT was selectively expressed only in cell-types which closely resem-

bled the parvocellular and magnocellular cells that endogenously expressed VP in these organotypic hypothalamic cultures, and not the predominately multipolar neurons which were visualized when the Nerve Specific Enolase promoter was used to drive reporter expression (Gainer et al., 2002). The specificity of the expression is also indicated by the fact that biolistic transfection of the VP-III–CAT2.1 construct in hippocampal slices which have no VP-expressing cells, yielded no cells that expressed the CAT reporter after transfection. Furthermore, even though glia are the predominant cells in the slices being penetrated by the gold particles (deduced from CMV–GFP experiments; Gainer et al., 2002), no glial expression of CAT was found using the VP III–CAT-2.1 construct. Finally, double la-

10

Fig. 5. Gene expression of VP-gene constructs in organotypic cultures after biolistic transfection. VP III–CAT-2.1 (A–C) and VP-I– EGFP-2.1 (D–F) constructs were used to transfect organotypic hypothalamic cultures. Analysis of CAT or GFP expression was made by immunohistochemistry in PVN (A,D,E) or SON (B) and SCN (C,F) nuclear regions of the cultures. See text for description of constructs.

11

bel immunocytochemistry using CAT antibodies and antibodies against VP-neurophysin showed colocalization between endogenous VP and CAT in many but not all CAT-expressing neurons. Similar results were obtained using an OT-III–CAT-3.6 transgene construct (see Jeong et al., 2001 for this construct’s structure). Given this success using the positive control construct, VP III–CAT-2.1, we then tested a novel deletion construct. This VP gene deletion construct contained the same upstream and downstream sequences, and exon I as in the VP III–CAT-2.1 gene, but had exons II and III and introns I and II removed, and the CAT reporter was replaced by EGFP which was fused to exon I. This gene construct was called VP I–GFP-2.1, and as was observed with the VP III–CAT-2.1 construct, the EGFP expression was found exclusively in cells in the PVN, SON, and SCN regions (Fig. 5D,F), many of which also expressed the endogenous gene in the cultured slice. Given these data, it appears that the combined biolistic–organotypic culture approach can be a practical and effective high throughput strategy to study cell-specific gene expression mechanisms in any cell type in the central nervous system. Experiments are currently underway to construct and evaluate OT and VP deletion constructs based on the clues derived from the comparative genomics analysis (Table 2). The HNS in the new millennium The MCNs continue to serve as ‘rosetta stones’ for elucidating important issues in neuroscience, and applications of new methodologies will continue to provide new opportunities to understand this system. The differential screening method and data described here has only begun to explore the molecular differences between the OT and VP MCNs. Use of cDNA microarray technologies to analyze the amplified cDNAs from single MCNs should greatly expand our insight into these phenotypes, and analysis of the functions of these newly uncovered molecules in the MCNs should be feasible using the ‘high throughput’ strategy described earlier. In addition, the rat MCNs provide a fascinating and new in vitro model for the study of apoptosis in the central nervous system. Finally, given that the mouse HNS in organotypic culture is an exceptionally stable in vitro system,

increased use of this species for long-term molecular and physiological studies of MCNs in vitro should be seriously considered. An additional advantage of the mouse hypothalamic organotypic culture is that it permits the culturing of MCNs from transgenic or ‘knock-out’ mouse models, thereby extending the opportunities for mechanistic analyses of the MCNs in the SON. With further development of these approaches, it should soon be feasible to selectively target the gene expression of specific molecules to MCNs in order to profoundly perturb and study their biosynthetic, neurosecretory and other processes, in vivo and in vitro. Abbreviations ICC ISHH IGR LINE MCNs OT PVN SON VP

immunocytochemistry in situ hybridization histochemistry intergenic region long interspersed nuclear element magnocellular neurons oxytocin paraventricular nucleus supraoptic nucleus vasopressin

Acknowledgements We thank our colleagues in the LNC, NINDS, Drs. S.-W. Jeong, E. Glasgow, B.-J. Zhang, A. Thomas, and K. Kusano, and collaborators, Drs. W. Scott Young III and Hiroshi Arima, for their contributions to various aspects of the work discussed here. We also thank L.L. Hampton and Dr. J.F. Battey in NINDS, for their help in isolating the human IGR, J. Nagel in the NINDS DNA Sequencing Facility for help with the automated sequencing, and Dr. Alexey S. Kondrashov, National Center for Biotechnology Information and Dr. Joseph G. Hacia, National Human Genome Research Institute for their advice about genomic sequence alignment, analysis, and molecular evolution. References Armstrong, W.E. (1995) Morphological and electrophysiological classification of hypothalamic supraoptic nuclei. Prog. Neurobiol., 47: 291–339.

12 Armstrong, W.E. and Stern, J.E. (1998) Electrophysiological distinctions between oxytocin and vasopressin neurons in the supraoptic nucleus. Adv. Exp. Med. Biol., 449: 67–77. Belenky, M., Wagner, S., Yarom, Y., Matzner, H., Cohen, S. and Castel, M. (1996) The suprachiasmatic nucleus in stationary organotypic culture. Neuroscience, 70: 127–143. Bertiny, L.T., Kursner, C., Gaillard, R.C., Corder, R. and Kiss, J.Z. (1993) A tissue culture model of the hypophysiotrophic CRF producing neuronal system. Neuroendocrinology, 57: 716–728. Brownstein, M.J., Russell, J.T. and Gainer, H. (1980) Synthesis, transport, and release of posterior pituitary hormones. Science, 207: 373–378. Buche, P. (1990) Weight matrix descriptions of four eukaryotic RNA polymerase II promoter elements derived from 502 unrelated promoter sequences. J. Mol. Biol., 212: 563–578. Burbach, J.P.H., Luckman, S.M., Murphy, D. and Gainer, H. (2001) Gene regulation in the magnocellular hypothalamoneurohypophysial system. Physiol. Rev., 81: 1197–1267. Dierickx, K., Vandersande, F. and Goosens, N. (1978) The one neuron–one hormone hypothesis and the hypothalamic neurosecretory system. In: J.-D. Vincent and C. Kordon (Eds.), Cell Biology of Hypothalamic Neurosecretion. Vol. 280, Colloques CNRS, Paris, pp. 391–398. Dohanics, J., Hoffman, G.E. and Verbalis, J.G. (1996) Chronic hyponatremia reduces survival of magnocellular vasopressin and oxytocin neurons after axonal injury. J. Neurosci., 16: 2373–2380. Du, B.P., Wu, D., Boldt-Houle, M. and Terwilligier, E.F. (1996) Efficient transduction of human neurons with an adenoassociated virus vector. Gene Ther., 3: 254–261. Dulac, C. and Axel, R. (1995) A novel family of genes encoding putative pheromone receptors in mammals. Cell, 83: 195–206. Du Vigneaud, V. (1954) Hormones of the posterior pituitary gland: oxytocin and vasopressin. Harvey Lectures Ser., 50: 1–26. Eberwine, J., Yeh, H., Miyashiro, K., Cao, Y., Nair, S., Finnell, R., Zettel, M. and Coleman, P. (1992) Analysis of gene expression in single live neurons. Proc. Natl. Acad. Sci. USA, 89: 3010–3014. Federico, M. (1999) Lentiviruses as gene delivery vectors. Curr. Opin. Biotechnol., 10: 448–453. Federoff, H.J. (1999) Novel targets for gene therapy. Gene Ther., 6: 1907–1908. Fink, D.J., DeLuca, N.A., Goins, W.F. and Glorioso, J.C. (1996) Gene transfer to neurons using herpes simplex virus-based vectors. Annu. Rev. Neurosci., 19: 265–287. Gähwiler, B.H., Capogna, M., Debanne, D., McKinney, R.A. and Thompson, S.M. (1997) Organotypic slice cultures: a technique has come of age. Trends Neurosci., 20: 471–477. Gainer, H. (1998) Cell-specific gene expression in magnocellular oxytocin and vasopressin neurons. Adv. Exp. Med. Biol., 449: 15–27. Gainer, H. and Wray, S. (1994) Cellular and molecular biology of oxytocin and vasopressin. In: E. Knobil and J.D. Neill (Eds.), The Physiology of Reproduction. Raven Press, New York, pp. 1099–1129.

Gainer, H. and Young, W.S., III (2001) Transgenic models for studies of oxytocin and vasopressin. In: M.G. Castro (Ed.), Transgenic Models in Endocrinology. Kluwer-Academic Press, Boston, pp. 25–46. Gainer, H., Fields, R.L. and House, S.B. (2001) Vasopressin gene expression: experimental models and strategies. Exp. Neurol., 171: 190–199. Gainer, H., Fields, R.L. and House, S.B. (2002) Neuronal transfection using particle-mediated gene transfer. In: A. Merighi and G. Carmignoto (Eds.), Cellular and Molecular Methods in Neurobiological Research. Springer-Verlag, Berlin, in press. Geddes, B.J., Harding, T.C., Hughes, D.S., Byrnes, A.P., Lightman, S.L., Conde, G. and Uney, J.B. (1996) Persistent transgene expression in the hypothalamus following stereotaxic delivery of a recombinant adenovirus: suppression of the immune response with cyclosporin. Endocrinology, 137: 5166– 5169. Geddes, B.J., Harding, T.C., Lightman, S.L. and Uney, J.B. (1997) Long-term gene therapy in the CNS: reversal of hypothalamic diabetes insipidus in the Brattleboro rat by using an adenovirus expressing arginine vasopressin. Nat. Med., 3: 1402–1404. Geddes, B.J., Harding, T.C., Lightman, S. and Uney, J.B. (1999) Assessing viral gene therapy in neuroendocrine models. Fr. Neuroendocrinol., 20: 296–316. Giguere, V., Tini, M., Flock, G., Ong, E., Evans, R.M. and Otulakowski, G. (1994) Isoform-specific amino-terminal domains dictate DNA-binding properties of RORalpha, a novel family of orphan nuclear hormone receptors. Genes Dev., 8: 538–553. Glasgow, E., Kusano, K., Chin, H., Mezey, E., Young III, W.S. and Gainer, H. (1999) Single cell RT-PCR analysis of rat supraoptic magnocellular neurons: neuropeptide phenotypes and high voltage-gated calcium channel subtypes. Endocrinology, 140: 5391–5401. Grant, F.D., Reventos, J.J., Gordon, W., Kawabata, S., Miller, M. and Majzoub, J.A. (1993) Expression of rat arginine vasopressin gene in transgenic mice. Mol. Endocrinol., 7: 659–667. Grotewold, E., Drummond, B.J., Bowen, B. and Peterson, T. (1994) The myb-homologous P gene controls phlobaphene pigmentation in maize floral organs by directly activating a flavonoid biosynthetic gene subset. Cell, 76: 543–553. Hara, Y., Battey, J. and Gainer, H. (1990) Structure of mouse vasopressin and oxytocin genes. Mol. Brain Res., 8: 319–324. Hatton, G.I. (1990) Emerging concepts of structure–function dynamics in adult brain: the hypothalamo-neurohypophysial system. Prog. Neurobiol., 34: 437–504. Hatton, G.I. (1997a) Function-related plasticity in hypothalamus. Annu. Rev. Neurosci., 20: 375–397. Hatton, G.I. (1997b) Oxytocin and vasopressin neurones: vive la difference!. J. Physiol., 500: 284. Herman, J.P., Marciano, F.F., Wiegand, S.J. and Gash, D.M. (1987) Selective cell death of magnocellular vasopressin neurons in neurophysectomized rats following chronic administration of vasopressin. J. Neurosci., 7: 2564–2575. House, S.B., Thomas, A., Kusano, K. and Gainer, H. (1998) Stationary organotypic cultures of oxytocin and vasopressin

13 magnocellular neurons from rat and mouse hypothalamus. J. Neuroendocrinol., 10: 849–861. Insel, T.R. and Young, L.J. (2001) The neurobiology of attachment. Nat. Rev. Neurosci., 2: 129–136. Jard, S., Barbaris, C., Audigier, S. and Tribollet, E. (1987) Neurohypophysial hormone receptor systems in brain and periphery. Prog. Brain Res., 72: 173–187. Jeong, S.-W., Castel, M., Zhang, B.-J., Fields, R.L., Paras, P., Arnheiter, H., Chin, H. and Gainer, H. (2001) Cell-specific expression and subcellular localization of CAT-reporter protein expressed from oxytocin and vasopressin gene promoterdriven constructs in transgenic mice. Exp. Neurol., 171: 255– 271. Johnsen, O., Murphy, P., Prydz, H. and Kolsto, A.-B. (1998) Interaction of the CNC-bZip factor TCF11/LCR-F1/Nrf1 with MafG: binding-site selection and regulation of transcription. Nucleic Acids Res., 26: 512–520. Kaplitt, M.G., Leone, P., Samulski, R.J., Xiao, X., Pfaff, D.W., O’Malley, K.L. and During, M.J. (1994) Long term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Gen., 8: 148–154. Karpati, G., Lochmuller, H., Nalbantoglu, J. and Durham, H. (1996) The principles of gene therapy for the nervous system. Trends Neurosci., 19: 49–54. Keir, S.D., House, S.B., Li, J., Xiao, X. and Gainer, H. (1999) Gene transfer into hypothalamic organotypic cultures using an adeno-associated virus vector. Exp. Neurol., 160: 313–316. Kiyama, H. and Emson, P.D. (1990) Evidence for the coexpression of oxytocin and vasopressin messenger ribonucleic acids in magnocellular neurosecretory cells: simultaneous demonstration of two neurophysin messenger ribonucleic acids by hybridization histochemistry. J. Neuroendocrinol., 2: 257–259. Korneev, S., Blackshaw, S.E., Kaiser, K. and Davies, J.A. (1996) cDNA libraries from identified neurons. Proc. R. Soc. Lond. B, 263: 57–62. Leng, G. (1988) Pulsatility in Neuroendocrine Systems. CRC Press, Boca Raton, FL, pp. 1–261. Ludwig, M. (1998) Dendritic release of vasopressin and oxytocin. J. Neuroendocrinol., 10: 881–895. Mackler, S.A. and Eberwine, J. (1993) Diversity of glutamate receptor subunit mRNA expression within live hippocampal CA1 neurons. Mol. Pharmacol., 44: 308–315. Manning, M. and Sawyer, W.H. (1989) Discovery, development, and some uses of vasopressin and oxytocin antagonists. J. Lab. Clin. Med., 114: 617–632. Manning, M., Bankowski, K. and Sawyer, W.H. (1987) Selective agonists and antagonists of vasopressin. In: S.M. Gash and G.J. Boer (Eds.), Vasopressin. Principles and Properties. Plenum Press, New York, pp. 335–368. Meister, B. (1993) Gene expression and chemical diversity in hypothalamic neurosecretory neurons. Mol. Neurobiol., 7: 87– 110. Mezey, E. and Kiss, J.Z. (1991) Coexpression of vasopressin and oxytocin in hypothalamic supraoptic neurons of lactating rats. Endocrinology, 129: 1814–1820. Merz, D.C., Dunn, R.J. and Drapeau, P. (1995) Generating a

phage display antibody library against an identified neuron. J. Neurosci. Methods, 62: 213–219. Mohr, E., Bahnsen, U., Kiessling, C. and Richter, D. (1988) Expression of the vasopressin and oxytocin genes in rats occurs in mutually exclusive sets of hypothalamic neurons. FEBS Lett., 242: 144–148. Molnar, A. and Georgopoulos, K. (1994) The ikaros gene encodes a family of functionally diverse zinc finger DNAbinding proteins. Mol. Cell. Biol., 14: 8292–8303. Morris, J.F., Nordmann, J.J. and Dyball, R.E.J. (1978) Structure– function correlation in mammalian neurosecretion. Int. Rev. Exp. Pathol., 18: 1–95. Morris, J.F., Hromas, R. and Rauscher III, F.J. (1994) Characterization of the DNA-binding properties of the myeloid zinc finger protein MZF1: two independent DNA-binding domains recognized by two DNA consensus sequences with a common G-rich core. Mol. Cell. Biol., 14: 1786–1795. Oliet, S.H., Piet, R. and Poulain, D.A. (2001) Control of glutamate clearance and synaptic efficacy by glial coverage of neurons. Science, 292: 923–926. Raisman, G. (1973) An ultrastructural study of the effects of hypophysectomy on the supraoptic nucleus of the rat. J. Comp. Neurol., 147: 181–207. Ratty, A.K., Jeong, S.-W., Nagle, J.W., Chin, H., Gainer, H., Murphy, D. and Venkatesh, B. (1996) A systematic survey of the intergenic region between the murine oxytocin and vasopressin encoding genes. Gene, 174: 71–78. Reijmers, L.G., van Ree, J.M., Spruijt, B.M., Burbach, J.P. and De Wied, D. (1998) Vasopressin metabolites: a link between vasopressin and memory?. Prog. Brain Res., 119: 523–535. Rusnak, R., House, S.B., Arima, H. and Gainer, H. (2002) Ciliary neurotrophic factor increases the survival of magnocellular oxytocin and vasopressin neurons in rat supraoptic nucleus in organotypic culture. Microsc. Res. Tech., 56: 101–112. Sachs, H. and Takabatake, Y. (1964) Evidence for a precursor in vasopressin biosynthesis. Endocrinology, 75: 943–948. Sachs, H., Fawcett, P., Takabatake, Y. and Portanova, R. (1969) Biosynthesis and release of vasopressin and neurophysin. Recent Prog. Horm. Res., 25: 447–491. Scharrer, B. (1987) Neurosecretion: beginnings and new directions in neuropeptide research. Annu. Rev. Neurosci., 10: 1– 17. Scharrer, E. (1928) Die lichtempfindlichkeit blinder elritzen (Untersuchungen uber das zwischenhirn der fische). Z. Vergl. Physiol., 7: 1–11. Sekido, R., Murai, K., Funahashi, J., Kamachi, Y., FujisawaSehara, A., Nabeshima, Y. and Kondoh, H. (1994) The deltacrystallin enhancer-binding protein delta EF1 is a repressor of E2-box-mediated gene activation. Mol. Cell. Biol., 14: 5692– 5700. Shields, J.M. and Yang, V.W. (1998) Identification of the DNA sequence that interacts with the gut-enriched kruppel-like factor. Nucleic Acids Res., 26: 796–802. Slack, R.S. and Miller, F.D. (1996) Viral vectors for modulating gene expression in neurons. Curr. Opin. Neurobiol., 6: 576– 583. Subauste, J.S. and Koenig, R.J. (1995) Comparison of the DNA

14 binding specificity and function of v-ErbA and thyroid hormone receptor alpha 1. J. Biol. Chem., 270: 7957–7962. Theodosis, D.T. and Poulain, D.A. (1987) Oxytocin-secreting neurones: a physiological model for structural plasticity in the adult mammalian brain. Trends Neurosci., 10: 426–430. Vandersande, F. and Dierickx, K. (1975) Identification of the vasopressin producing and of the oxytocin producing neurons in the hypothalamic magnocellular neurosecretory system of the rat. Cell. Tiss. Res., 164: 153–162. Vascquez, E.C., Johnson, R.F., Beltz, T.G., Haskell, R.E., Davidson, B.L. and Johnson, A.K. (1998) Replication-deficient adenovirus vector transfer of gfp reporter gene into supraoptic nucleus and subfornical organ neurons. Exp. Neurol., 154: 353–365. Venkatesh, B., Si-Ho, S.-L., Murphy, D. and Brenner, S. (1997) Transgenic rats reveal remarkable functional conservation of regulatory controls between the fish isotocin and rat oxytocin genes. Proc. Natl. Acad. Sci. USA, 94: 12462–12466. Vutskits, L., Bartanusz, V., Schulz, M.F. and Kiss, J.Z. (1998) Magnocellular vasopressinergic neurons in explant cultures are rescued from cell death by ciliary neurotrophic factor and leukemia inhibiting factor. Neuroscience, 87: 571–582. Waller, S., Fairhall, K.M., Xu, J., Robinson, I.C.A.F. and Murphy, D. (1996) Neurohypophyseal and fluid homeostasis in transgenic rats expressing a tagged rat vasopressin prepropeptide in hypothalamic neurons. Endocrinology, 137: 5068–5077.

Waller, S.J., Ratty, A., Burbach, J.P.H. and Murphy, D. (1998) Transgenic and transcriptional studies on neurosecretory cell gene expression. Cell. Mol. Neurobiol., 18: 149–171. Wang, Z., Young, L.J., De Vries, G.J. and Insel, T.R. (1998) Voles and vasopressin: a review of molecular, cellular, and behavioral studies of pair bonding and paternal behaviors. Prog. Brain Res., 119: 483–499. Wray, S., Kusano, K. and Gainer, H. (1991) Maintenance of LHRH and oxytocin neurons in slice explants cultured in serum-free media: effects of tetrodotoxin on gene expression. Neuroendocrinology, 54: 327–339. Wray, S., Castel, M. and Gainer, H. (1993) Characterization of the suprachiasmatic nucleus in organotypic slice explant cultures. Microsc. Res. Tech., 25: 46–80. Xi, D., Kusano, K. and Gainer, H. (1999) Quantitative analysis of oxytocin and vasopressin mRNAs in single magnocellular neurons isolated from supraoptic nucleus of rat hypothalamus. Endocrinology, 140: 4677–4682. Zeng, Q., Carter, D.A. and Murphy, D. (1994) Cell specific expression of a vasopressin transgene in rats. J. Neuroendocrinol., 6: 469–477. Zweidler-McKay, P.A., Grimes, H.L., Flubacher, M.M. and Tsichlis, P.N. (1996) Gfi-1 encodes a nuclear zinc finger protein that binds DNA and functions as a transcriptional repressor. Mol. Cell. Biol., 16: 4024–4034.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 2

Estrogen modulates oxytocin gene expression in regions of the rat supraoptic and paraventricular nuclei that contain estrogen receptor-β Paul J. Shughrue ∗ , Tammy L. Dellovade and István Merchenthaler The Women’s Health Research Institute, Wyeth Research, Collegeville, PA 19426, USA

Abstract: Oxytocin is an important modulator of female reproductive functions including parturition, lactation and maternal behavior, while vasopressin regulates water balance and acts as a neurotransmitter. For decades, it has been suggested that estrogen regulates the production and/or release of oxytocin and vasopressin in the rodent brain. Although several studies demonstrated that estrogen can modulate vasopressin mRNA levels in regions known to contain estrogen receptor (ER), such as the bed nucleus of the stria terminalis and medial amygdala, data from the paraventricular and supraoptic nuclei were inconclusive. Since early immunohistochemical and in situ hybridization studies revealed few, if any, ER containing cells in these hypothalamic nuclei, it was thought that oxytocin and vasopressin were not directly regulated by estrogen. The discovery of a second ER (ER-β) in the late 1990s suggested that estrogen could act in many brain regions heretofore not considered targets for estrogen action. Initial in situ hybridization studies revealed a wide distribution of ER-β mRNA in the rat brain including neurons of the supraoptic nucleus and the parvocellular and magnocellular divisions of the paraventricular nucleus. Subsequent double-label in situ hybridization/immunocytochemistry studies showed that ER-β mRNA was present in oxytocin and vasopressin neurons, with the degree of colocalization being both neuropeptide and region specific. In an attempt to demonstrate that ER-β mRNA was translated into a biologically active protein, a series of in vivo binding studies were conducted in rats with 125 I-estrogen. These data revealed the presence of nuclear estrogen binding sites in neurons of the magnocellular system indicating that ER-β mRNA was translated into protein. Concurrent studies in mice found that the distribution of ER-β mRNA and 125 I-estrogen binding was similar to rats, although there were some notable differences. For example, ER-β mRNA and binding were not detected in the mouse supraoptic nucleus and although ER-β was the principle ER in the paraventricular nucleus, ER-α was also present. The prevalence of ERs in the mouse paraventricular nucleus was further investigated using ER-α and ER-β knockout mice for in vivo binding studies with 125 I-estrogen. The results of these studies showed that ER-β was the predominant ER in the paraventricular nucleus and confirmed the presence of ER-β in other brain regions. Moreover, our group recently generated and characterized several polyclonal antisera raised against the C-terminus of ER-β. Through the use of these antisera, we have confirmed the presence of ER-β in the rat paraventricular and supraoptic nuclei and shown that ER-β is colocalized, in part, with oxytocin and vasopressin. To assess the ability of estrogen to modulate the expression of oxytocin mRNA, ovariectomized rats were treated with vehicle or estradiol and the brains processed for in situ hybridization. The results of these studies revealed that estradiol downregulated oxytocin mRNA in the rat paraventricular nucleus within 6 h of treatment. Together these data and the observation that some of the oxytocin and vasopressin neurons contain ER-β suggest that estrogen, acting through ER-β, may directly regulate oxytocin gene expression. However, since the paraventricular nucleus has many subdivisions with different projections and the degree of colocalization of ER-β with oxytocin/vasopressin varies among subdivisions, the effects of estrogen treatment on gene expression requires further study to ascertain the role of estrogen action in this neuronal systems. Keywords: ER-α; ER-β; ERαKO; ERβKO; Estradiol; Vasopressin; Paraventricular nucleus; Supraoptic nucleus ∗ Correspondence

to: P.J. Shughrue, Department of Neuroscience, Merck Research Laboratories, Sumneytown Pike and Broad Street, WP26A-3000, West Point, PA 19486, USA. Tel.: +1-215-652-4816; Fax: +1-215-652-2075; E-mail: [email protected]

16

Oxytocin and vasopressin Oxytocin and vasopressin are neuropeptides that are primarily synthesized by the magnocellular neurons of the paraventricular, supraoptic and accessory nuclei. These peptides are then transported to the posterior lobe of the pituitary and released into the systemic circulation. In addition, both peptides are synthesized by parvocellular neurons in the paraventricular nucleus and vasopressin is also expressed in the bed nucleus of the stria terminalis, medial amygdala and other brain regions. The parvocellular neurons and accessory regions release oxytocin and vasopressin at nerve terminals throughout the brain including the hypothalamus, brainstem and spinal cord where they are believed to act as neurotransmitters (Sawchenko and Swanson, 1982; Cechetto and Saper, 1988; Wagner and Clemens, 1991). Although oxytocin and vasopressin differ from each other by only two amino acids, their physiological actions are quite different. Oxytocin released into the general circulation plays a critical role in regulating milk-ejection from the female nipple as well as the induction of uterine contractility during childbirth. Centrally, oxytocin is thought to regulate maternal and reproductive behaviors, although other physiological roles may still be elucidated. Vasopressin, or anti-diuretic hormone, plays a critical role in maintaining water balance by stimulating the re-uptake of water in the collecting tubules and ascending limb of Henle’s loop. The vasopressin that is synthesized and released within the brain is thought to be important for many functions including learning and memory. Estrogen regulates oxytocin and vasopressin Interestingly, the first studies to describe the distribution of estrogen receptors in the rat brain, using in vivo autoradiography with 3 H-estradiol (Stumpf, 1970; Pfaff and Keiner, 1973; Stumpf et al., 1975), detected estrogen binding sites in the supraoptic and paraventricular nuclei. Rhodes et al. (1981) later used a double-label in vivo autoradiography/ immunocytochemistry method to show that the oxytocin/vasopressin neurons in the paraventricular nucleus concentrated radiolabeled estrogen, the first evidence that estrogen may directly regulated these peptides. However, with the development of specific

antisera raised against the estrogen receptor, investigators noted that only a few immunoreactive cells were present in the paraventricular nucleus and no staining was seen in the supraoptic nucleus (Cintra et al., 1986; Sar and Parikh, 1986; Axelson and van Leeuwen, 1990). These data casted a shadow of doubt over the binding studies and questioned whether estrogen directly modulates oxytocin gene expression. With the cloning of the rat estrogen receptor in 1987 (Koike et al., 1987), it became possible to look at the cellular localization for estrogen receptor in the rat central nervous system with in situ hybridization. Simerly et al. (1990) provided the first comprehensive map of estrogen receptor mRNA in the rat brain, a study that clearly demonstrated that the paraventricular and supraoptic nuclei lacked estrogen receptor gene expression. Despite the lack of ER in the paraventricular and supraoptic nuclei, an accumulating body of data suggested that estrogen regulates the production and/or release of oxytocin in the rodent brain. In the 1970s, physiological studies found that the electrical activity of oxytocin neurons in the paraventricular nucleus changed over the estrus cycle and was increased following estrogen treatment of ovariectomized animals (Negoro et al., 1973; Akaishi and Sakuma, 1985). Subsequent immunocytochemical and in situ hybridization studies found that estrogen increases the amount of oxytocin immunoreactivity and mRNA in a number of brain regions including the paraventricular nucleus (Rhodes et al., 1981; Jirikowski et al., 1988; Van Tol et al., 1988; Caldwell et al., 1989; Miller et al., 1989a; Amico et al., 1995). These observations, and the discovery of a functional estrogen response element in the promoter region of the rat oxytocin gene (Ivell and Richter, 1984; Mohr et al., 1988; Burbach et al., 1990; Adan et al., 1993) were at odds with the lack of ER in the oxytocin neurons and suggested that estrogen may directly stimulate oxytocin gene expression in the rat brain. However, additional studies also began to appear in the literature that questioned whether estrogen directly modulated oxytocin expression (Jirikowski et al., 1988; Caldwell et al., 1989; Burbach et al., 1990; Kawata et al., 1991). Based on these conflicting reports, it was difficult to understand the mechanism by which estrogen modulates oxytocin levels in paraventricular and supraoptic nuclei.

17

In contrast to the oxytocin system, there has been little compelling evidence to suggest that estrogen modulates vasopressin levels in the paraventricular or supraoptic nuclei. Nevertheless, a number of studies have shown that the vasopressin neurons localized in the bed nucleus of the stria terminalis and medial amygdala are sensitive to changes in estrogen levels. For example, using immunocytochemistry and in situ hybridization, gonadectomy was shown to dramatically attenuate the level of vasopressin in the bed nucleus of the stria terminalis and medial amygdala, while estrogen treatment increases levels in both of these regions (DeVries et al., 1984, 1985 and DeVries et al., 1986; Miller et al., 1989b, 1992; Brot et al., 1993). Since estrogen receptors were shown to be concentrated in the bed nucleus of the stria terminalis and medial amygdala using in vivo autoradiography (Stumpf, 1970; Pfaff and Keiner, 1973; Stumpf et al., 1975), immunocytochemistry (Cintra et al., 1986; Sar and Parikh, 1986) and in situ hybridization (Simerly et al., 1990), a direct effect of estrogen on the vasopressin neurons seemed plausible. In 1990, Axelson and van Leeuwen (1990) showed, using double-label immunocytochemistry, that the majority of the vasopressin neurons detected in the bed nucleus and medial amygdala also co-express estrogen receptors. In contrast, this study was unable to detect any cells in the paraventricular and supraoptic nuclei that contained both vasopressin and estrogen receptors (Axelson and van Leeuwen, 1990). In agreement with these finding, gonadectomy and estrogen treatment do not alter vasopressin levels in either the paraventricular or supraoptic nuclei (DeVries et al., 1984, 1985 and DeVries et al., 1986). Taken together these observations provided strong support for the hypothesis that estrogen directly modulates vasopressin expression in the amygdala and BNST, but not in the magnocellular and parvocellular neurons of the paraventricular and supraoptic nuclei. Thus, despite more than two decades of research and the use of a myriad of methodological approaches, the role of estrogen in the paraventricular and supraoptic nuclei remained unresolved. Localization of ER-β mRNA in the brain In 1996, a second nuclear estrogen receptor, named ER-β, was cloned from the rat prostate (Kuiper et

al., 1996). This newly discovered receptor has a high degree of sequence homology with the classical ER (now called ER-α) as well as a specific binding affinity for estradiol (Kuiper et al., 1996). Cotransfection of cells with ER-β and an estrogen response element reporter construct demonstrated that estrogens, acting through ER-β, could activate gene transcription in a dose-dependent manner (see Kuiper et al., 1998). The original in situ hybridization studies in the rat central nervous system (Shughrue et al., 1996, 1997a) revealed the presence of ER-β mRNA throughout the brain including regions where ER-α was expressed (e.g. preoptic area, bed nucleus of the stria terminalis, medial amygdala), as well as regions where ER-α mRNA was sparse (e.g. cortex and hippocampus) or absent (e.g. paraventricular and supraoptic nuclei, cerebellum). In the rat diencephalon, intense hybridization signal was seen in almost every neuron of the supraoptic nucleus (Fig. 1B), extending throughout the length of the hypothalamus and in the retrochiasmatic supraoptic nucleus. In addition, numerous ER-β-containing neurons were concentrated in the paraventricular nucleus (Fig. 1A). The majority of the intensely labeled cells were located in the magnocellular subdivisions, with the posterior subnucleus having the highest number of labeled cells (Fig. 1A). Additional ER-β mRNA-containing neurons were also present in the parvocellular subdivisions of the paraventricular nucleus (Fig. 1A), although the level of hybridization signal was attenuated when compared with the magnocellular neurons. While the majority of the labeled cells in the parvocellular neurons were located in the medial division, the dorsal cap, and to a lesser extent, the lateral and periventricular parvocellular subnuclei also contained scattered labeled cells. Subsequent in situ histochemical studies by other laboratories (Laflamme et al., 1998; Osterlund et al., 1998) also detected high levels of ER-β mRNA in the rat supraoptic and paraventricular nuclei, confirming the differential localization of hybridization signal among the magnocellular and parvocellular nuclei. It is worth noting that ER-β mRNA was also seen in the mouse paraventricular nucleus, but not the supraoptic nucleus (Shughrue et al., 1997b) and that the number of neurons and regional distribution appeared to be more restricted than with the rat (Shughrue et al., 1996, 1997a). Based on the

18

Fig. 1. Representative autoradiograms of ER-β mRNA in the paraventricular (A) and supraoptic (B) nuclei of the rat hypothalamus by in situ hybridization. ER-β mRNA is concentrated in both the magnocellular neurons (PLM) of the paraventricular (A) and supraoptic (B) nuclei, although additional labeled neurons were also seen in the parvocellular subdivision (PDC, PMP and PV) of the paraventricular nucleus. Brain region comparable to bregma −1.80 mm. Interrupted line delineates the paraventricular nucleus (A). Asterisks indicate the third ventricle (A). Triangles indicate the ventral surface of the brain (B). Scale bar: 400 μm. Republished with permission from Shughrue et al. (1997a).

abundance of ER-β mRNA in the rat supraoptic and paraventricular nuclei, it seemed plausible that estrogen, acting via ER-β, could directly modulate the expression of oxytocin and/or vasopressin. Colocalization of ER-β mRNA with oxytocin and vasopressin With the discovery of ER-β and revelation that it is highly expressed in the supraoptic and paraventricular nuclei of the rat hypothalamus, a number of laboratories used a double-label in situ hybridization/ immunocytochemistry to investigate whether oxytocin and vasopressin neurons contain ER-β mRNA (Hrabovszky et al., 1998; Laflamme et al., 1998; Shughrue and Merchentaler, unpublished). In the supraoptic area, ER-β mRNA was seen in many vasopressin and oxytocin immunoreactive neurons (Hrabovszky et al., 1998; Shughrue and Merchentaler, unpublished). The percentage of vasopressin cells expressing ER-β mRNA appeared to be higher than double-labeled oxytocin neurons, but both were abundant and distributed throughout the rostrocaudal extent of the supraoptic nucleus. In the paraventricular nucleus, double-labeled cells were also detected, although only a few neurons co-expressed vaso-

pressin and ER-β mRNA (Hrabovszky et al., 1998; Laflamme et al., 1998; Shughrue and Merchentaler, unpublished). In contrast, neurons that co-expressed oxytocin and ER-β mRNA were found in both the parvocellular and magnocellular divisions (Fig. 2). The highest degree of colocalization was seen in the caudal portion of the paraventricular nucleus (medial and lateral parvocellular divisions) where as many as 93% of the oxytocin neurons have been reported to contain ER-β mRNA (Hrabovszky et al., 1998). In the medial and rostral subdivisions of the paraventricular nucleus, the majority of the doublelabeled oxytocin neurons were seen in the posterior magnocellular division, although the degree of coexpression was attenuated when compared with the caudal divisions. These findings highlight the importance of examining estrogen regulation of oxytocin expression within particular subregions of the paraventricular nucleus. It is possible that the conflicting results in previous studies are partly due to the fact that effects of estrogen were generally measured within the entire nucleus rather those regions where OT neurons co-express ER-β.

19

Fig. 2. Colocalization of oxytocin immunoreactivity (brown staining) and ER-β mRNA (silver grains) in the paraventricular nucleus of the female rat hypothalamus using a double label immunocytochemistry/in situ hybridization method. Note that the majority of the oxytocin neurons in the medial (MP) and lateral (LP) parvocellular divisions also express ER-β mRNA (black arrows in C and D), although neurons that solely express ER-β mRNA were also seen (red arrows in C and D). Brain regions in A and B are comparable to bregma levels −2.12 and −2.2 mm. Asterisks indicate the third ventricle (A). Abbreviations: LP, paraventricular nucleus, lateral parvocellular; MP, paraventricular nucleus, medial parvocellular.

Binding in the rat and mouse brain In an attempt to demonstrate that ER-β mRNA was translated into a functional protein, a series of binding studies were conducted in rats using a ligand (125 I-estrogen) that has a similar affinity for both ERs (Shughrue et al., 1999). When ovariectomized animals were injected with 125 I-estrogen, neurons with a nuclear uptake and retention of the radiolabeled estrogen were seen throughout the brain including regions that contained only ER-α, only ER-β or both ERs (Shughrue et al., 2000; Shughrue and Merchen-

thaler, unpublished observations). In particular, labeling was seen in the neurons of the supraoptic and paraventricular nuclei (Fig. 3), with a distribution that matched the topography of ER-β mRNA (Fig. 1A vs. Fig. 3A). Competition studies with 17βestradiol and DES, but not progesterone, eliminated the binding sites throughout the brain, demonstrating that 125 I-estrogen was selectively binding to the ERs (Shughrue et al., 2000). These observations and the fact that ER-α is not expressed in the rat supraoptic and paraventricular nuclei (Simerly et al., 1990; Shughrue et al., 1997a; Laflamme et al., 1998; Os-

20

21

terlund et al., 1998) provided compelling evidence that ER-β mRNA was translated into a biologically active protein. To definitively show that ER-β was translated into a protein, additional in vivo binding studies were conducted using estrogen receptor knockout (ERαKO and ERβKO) mice (Shughrue et al., 2002). When 125 I-estrogen was injected into wild-type mice (Fig. 4, left column), labeled cells were seen in brain regions that contain predominantly ER-α (ventromedial and arcuate nuclei), ER-β (paraventricular nucleus, medial tuberal nucleus, entorhinal cortex and dorsal raphe) or similar levels of both ERs (preoptic area, bed nucleus of the stria terminalis and amygdala). While binding was seen in the ERαKO brain (Fig. 4, right column), the degree of labeling was attenuated in many regions compared with wild-type animals. For example, the intensity of signal was diminished in the ERαKO preoptic area, bed nucleus of the stria terminalis, medial amygdala, arcuate nucleus and ventromedial nucleus. However, little or no change was detected in the paraventricular nucleus. A comparison of the distribution of ER-β mRNA in the ERαKO brain (Shughrue et al., 1997b) with the distribution of 125 I-estrogen binding sites (Fig. 4, right column), further revealed that binding was seen in brain regions where ER-β was expressed. Moreover, in most brain regions, there was a good correlation between the degree of 125 I-estrogen binding and the intensity of ER-β hybridization signal, with areas such as the bed nucleus of the stria terminalis, paraventricular nucleus, medial amygdala and dorsal raphe containing the highest levels of ER-β mRNA and 125 I-estrogen binding. A subsequent study utilized ERβKO mice to further investigate the presence and distribution of ER-containing neurons in the brain. As seen before, binding was detected throughout the wild-type brain including the paraventricular nucleus. Similarly, 125 I-estrogen binding was widely distributed in the ERβKO mouse forebrain (Fig. 4, center column), although a reduction in labeled cell number

was seen in regions where ER-β is highly expressed (Shughrue et al., 1997b). The most notable decrease was seen in the ERβKO paraventricular nucleus, an area in the mouse brain where ER-β is the predominant ER. In contrast, no attenuation in labeled cell number was detected in regions of the ERβKO brain where ER-β is sparse or absent, such as the ventromedial and arcuate nuclei of the hypothalamus. Together, the results of these studies demonstrated that ER-β was translated into a biologically active protein in the supraoptic nucleus and paraventricular nucleus as well as in most of the other brain regions investigated. During the course of the rat binding studies, a series of double-label experiments were also conducted to investigate the colocalization of ERβ with oxytocin and vasopressin immunoreactivity. The results of these studies clearly demonstrated that radiolabeled estrogen was concentrated in the nucleus of vasopressin and oxytocin neurons in the supraoptic and paraventricular nuclei (data not shown). Moreover, the pattern of colocalization was in agreement with previous double-label in situ hybridization/immunocytochemistry studies described above. ER-β immunoreactivity Li et al. (1997) provided the first data on the distribution of ER-β immunoreactivity in the rat forebrain using the polyclonal antiserum PA1-310. The results of this study (Li et al., 1997) and others (Alves et al., 1998) showed that ER-β immunoreactivity was present throughout the length of the supraoptic nucleus and retrochiasmatic supraoptic nucleus. In the paraventricular nucleus, ER-β was seen in the magnocellular and parvocellular divisions (Li et al., 1997; Simonian and Herbison, 1997; Alves et al., 1998), although the number of immunoreactive cells was markedly attenuated when compared with ER-β mRNA (Shughrue et al., 1997a). ER-β immunoreac-

Fig. 3. Autoradiographic images of 125 I-estrogen binding in the female rat paraventricular nucleus by in vivo autoradiography. Cells with a nuclear uptake and concentration of radiolabeled ligand were seen throughout the paraventricular nucleus, with a concentration in the rostral magnocellular divisions and in the medial and lateral parvocellular regions. Brain regions comparable to bregma −1.80 (A), bregma −1.88 (B) and bregma −2.12 (C). Abbreviations: LP, paraventricular nucleus, lateral parvocellular; MP, paraventricular nucleus, medial parvocellular; PM, paraventricular nucleus, posterior magnocellular; 3V, third ventricle.

22

Fig. 4. Autoradiographic images of 125 I-estrogen binding in the hypothalamus of female wild-type, ERβKO, and ERαKO mice by in vivo autoradiography. In the wild-type mouse brain, 125 I-estrogen binding was seen in regions that express both ERs (medial preoptic area (MPO) and bed nucleus of the stria terminalis (BST)), as well as regions where ER-α (arcuate (AN) and ventromedial (VMN) nuclei of the hypothalamus) or ER-β (paraventricular nucleus (PVN)) are the predominant ERs. In the ERβKO brain, 125 I-estrogen binding was seen in similar brain regions, although the degree of labeling was attenuated in the preoptic area and bed nucleus of the stria terminalis and absent in the paraventricular nucleus. In the ERαKO brain, 125 I-estrogen binding was markedly reduced in the preoptic area, arcuate and ventromedial nuclei, but abundant in the paraventricular nucleus. Republished with permission from Shughrue et al. (2002).

tive cells were also seen in the bed nucleus of the stria terminalis and medial amygdala, but were absent in the cortex, medial preoptic area, zona incerta and other regions where ER-β mRNA was detected. As expected, ER-β immunoreactivity was seen in the cell nucleus in most brain regions, but surprisingly, it was also seen in the cytoplasm in the hippocam-

pus and lateral septum (Li et al., 1997; Simonian and Herbison, 1997). It is worth noting that a concentration of immunoreactivity was also detected in regions where ER-β mRNA was very sparse (e.g. lateral septum and hippocampus). Based on these data, it was clear that the PA1-310 antiserum was detecting only a fraction of the ER-β containing neurons,

23

while at the same time detecting other non-specific antigens. Alternatively, one could argue that the vast majority of ER-β mRNA may not be translated into protein. In an attempt to clarify the discrepancies between the localization of ER-β mRNA, in vivo binding and the immunoreactivity seen with PA1-310, we developed and utilized a new antiserum raised against the C-terminus of the mouse ER-β called Z8P. When Z8P was used to evaluate the topography of ER-β containing neurons in the rat brain, a broad distribution of labeled cells was revealed, although the most intense immunoreactivity was seen in the nuclei of the hypothalamus (Shughrue and Merchenthaler, 2001). In the hypothalamus, immunoreactivity was concentrated in the supraoptic and paraventricular nuclei (Fig. 5A–D). The majority of intensely labeled cells were found in the magnocellular subdivisions of the paraventricular nucleus, with the posterior subnucleus having the highest number of immunoreactive cells (Fig. 5A–B). Lightly labeled, scattered ER-β immunoreactivity neurons were also present in the parvocellular subdivisions of the paraventricular nucleus (Fig. 5A–C), including the medial parvocellular subdivision (Fig. 5A–C), dorsal cap (Fig. 5A–B), and to a lesser extent, the lateral and periventricular parvocellular subnuclei (Fig. 5A– C). Immunoreactive nuclei were also seen extending from the paraventricular nucleus towards the supraoptic nucleus in the accessory magnocellular cell groups. In addition to these hypothalamic nuclei, ER-β immunoreactivity was detected in the cerebral cortex, medial septum, preoptic area, bed nucleus of the stria terminalis, zona incerta, medial and cortical amygdaloid nuclei, cerebellum, nucleus of the solitary tract, ventral tegmental area and spinal trigeminal nucleus (Shughrue and Merchenthaler, 2001). A series of double-label immunocytochemistry/in situ hybridization studies were conducted and demonstrated that ER-β mRNA and Z8P staining were co-localized in neurons throughout the brain, including the supraoptic and paraventricular nuclei, thus confirming the specificity of the Z8P antisera. These data demonstrate that ER-β mRNA is translated into an immunoreactive protein throughout most of the rat brain, including the supraoptic and paraventricular nuclei (Fig. 5E–F).

Colocalization of ER-β with oxytocin vasopressin The generation of a good antiserum for ER-β (Z8P) has enabled us to look at the colocalization of ER-β immunoreactivity with oxytocin and vasopressin in the rat brain. While similar studies were conduced previously using the PA1-310 antiserum (Simonian and Herbison, 1997; Alves et al., 1998), both the specificity and titer of PA1-310 have generated some uncertainty about the degree of colocalization in the rat paraventricular nucleus. Additionally, one study (Simonian and Herbison, 1997) failed to detect ER-β immunoreactivity in the supraoptic nucleus, a further indication that these data required confirmation with a more specific antiserum. The results of the present dual-labeling immunocytochemistry studies further confirmed data collected using oxytocin/vasopressin immunoreactivity together with ER-β mRNA. In the paraventricular nucleus, many oxytocin magnocellular neurons expressed ER-β immunoreactivity (Fig. 6). This was particularly evident in the ER-β rich posterior subnucleus, where the vast majority, 90–95%, of the oxytocin cells was double-labeled. In the lateral magnocellular subdivision, approximately 80% of oxytocin cells expressed ER-β, but in the more anterior and medial magnocellular subdivision, very few double-labeled cells were seen. In general, the degree of co-localization of ER-β immunoreactivity with vasopressin was much lower throughout the paraventricular nucleus than was seen with oxytocin. Estrogen regulates oxytocin The double-labeling experiments discussed above have demonstrated, using a variety of methods, that ER-β is co-localized with a portion of the oxytocin neurons in the paraventricular nucleus of the rat brain. This definitive anatomical evidence has provided a foundation for trying to understanding if, and how, estrogen modulates oxytocin expression. To date, few studies have used a treatment paradigm that would allow investigators to determine whether estrogen directly modulates the level of oxytocin. That is, most studies have treated ovariectomized animals with estrogen for several days or weeks prior to evaluating changes in gene expression. While these studies have provided useful information, it has been

24

Fig. 5. The distribution of ER-β-immunoreactive neurons (A–F) and ER-β mRNA (E,F) in the paraventricular (A–C, E–F) and supraoptic (D) nuclei of the female rat hypothalamus. ER-β immunoreactivity was seen throughout the paraventricular nucleus, although the majority of the labeled cells were concentrated in the rostral magnocellular divisions (compare A, bregma −1.80 with C, bregma −2.12). Labeled cells were also seen in the parvocellular (A–C), and periventricular divisions (A–C) and were abundant in the supraoptic nucleus (D). Utilizing a double-label in situ hybridization/immunocytochemistry method, ER-β immunoreactive cells (gray staining; E,F) in the posterior magnocellular (E) and lateral parvocellular (F) divisions of the paraventricular nucleus were shown to contain ER-β mRNA (silver grains; E,F). Abbreviations: AHP, anterior hypothalamic area, posterior; DC, paraventricular nucleus, dorsal cap; LP, paraventricular nucleus, lateral parvocellular; MP, paraventricular nucleus, medial parvocellular; PM, paraventricular nucleus, posterior magnocellular; PV, paraventricular nucleus, periventricular division; SON, supraoptic nucleus; ox, optic chiasm. Scale bars: 400 μm in A–C; 200 μm in D; 30 μm in E–F. Republished with permission from Shughrue and Merchenthaler (2001).

25

Fig. 6. The distribution of ER-β immunoreactive (Texas Red) and oxytocin immunoreactive (FITC) neurons in the paraventricular nucleus of the female rat brain using a double-label immunocytochemistry method. Low power confocal images (A,B) show the similar distribution of ER-β and oxytocin immunoreactivity in the central portion (∼ bregma −1.88) of the paraventricular nucleus. High power images (C–E) clearly show that most of the oxytocin neurons in the lateral parvocellular (C,D), posterior magnocellular and accessory magnocellular (E) regions also contain ER-β, whereas neurons that contain only ER-β were frequently interspersed with these double-labeled neurons. Abbreviations: LP, paraventricular nucleus, lateral parvocellular; MP, paraventricular nucleus, medial parvocellular; PM, paraventricular nuclear, posterior magnocellular.

26

Fig. 7. Evaluation of the hybridization signal for oxytocin mRNA in the paraventricular nucleus of the female rat as detected with in situ hybridization. Note the dramatic decrease (P < 0.05) in hybridization signal when ovariectomized (Ovx) rats were treated with estradiol (5 μg/kg BW) for 6 h, as compared with the vehicle (50% DMSO/saline) treated controls. Statistical significance: * P < 0.05.

difficult to ascertain if estrogen is acting directly on the oxytocin neurons or indirectly through interneurons or other sites in the brain. Furthermore, these treatment differences may account for the inconsistencies in the outcomes of these studies. Therefore, we conducted a series of in situ hybridization studies using a treatment paradigm previously used to investigate regulation of progesterone receptor mRNA by estrogen in the female rat brain (Shughrue et al., 1997c). Ovariectomized females were treated with a single subcutaneous dose of estrogen (5 μg/kg BW) or vehicle (50% DMSO/saline), euthanized 6 h later and coronal sections processed for in situ hybridization with a cRNA probe for oxytocin. Quantitative assessment of film autoradiograms (Fig. 7) revealed that the level of oxytocin mRNA was significantly attenuated in the paraventricular nucleus of estrogentreated animals (P < 0.05), while no changes ob-

served in the supraoptic nucleus. Since these observations were in striking contrast with previous reports, a second study looked at the time-course of these changes. Once again, ovariectomized females were treated with a single subcutaneous dose of estrogen or vehicle, and the estrogen-treated animals euthanized 6, 12, 24 or 48 h after injection. As seen in the first study, the level of oxytocin mRNA in the paraventricular nucleus was significantly attenuated 6 h after estradiol administration (P < 0.001), when compared with the vehicle-treated controls (Fig. 8). At 12 and 24 h after a single injection of estradiol, the levels of oxytocin mRNA were still significantly less than controls (P < 0.05), although the magnitude was less than observed at 6 h (1.6 vs. 3.1 fold). By 48 h after estradiol treatment, no difference was detected between the treated animals and controls (Fig. 8). These findings suggest that estra-

27

Fig. 8. A comparison of the level of oxytocin mRNA in the paraventricular nucleus of the female rat after a single injection of 17β-estradiol (5 μg/kg BW) and a 6–48-h survival time. Note that the level of oxytocin mRNA was significantly attenuated (P < 0.001) 6 h after the injection, was still low at 12 and 24 h (P < 0.05) and then increased to a level similar to control ovariectomized animals at 48 h. Statistical significance * P < 0.05; *** P < 0.001.

diol is capable of directly modulating oxytocin gene expression in the rat paraventricular nucleus, while no significant effect was seen in the supraoptic nucleus. It is interesting to note that a single bolus of estradiol decreased the level of oxytocin mRNA. These observations are at odds with previous studies that showed that estrogen had no significant effect on oxytocin or increased levels in the paraventricular nucleus (Van Tol et al., 1988; Crowley et al., 1993; Dellovade et al., 1999). While the reason for these differences is unknown, it is noteworthy that the treatment paradigm used in our studies (single injection, rapid uptake) is clearly different than the methods used by most investigators (single or mul-

tiple injections; slow, sustained delivery). With this in mind, it is possible that the oxytocin system is differentially responsive to both the level of estrogen and the duration of action, a response that may involve both direct effects as well as input from other estrogen-sensitive regions and/or interneurons. Additionally, since the paraventricular nucleus is composed of many subdivisions that may or may not contain double-labeled (oxytocin/ER-β) neurons, it is likely that the effects of estrogen vary among subdivisions. Clearly, additional studies are required to elucidate the role of estrogen action in the oxytocin neuronal system as well as the vasopressin system, particularly in the supraoptic nucleus. Nevertheless,

28

the results of anatomical and molecular studies, discussed herein, have provided a foundation for understanding the role of estrogen in these systems, while also highlighting the need for future studies. With these data, we can begin to understand the myriad of effects estrogen has in the brain, particularly those that have not heretofore been considered. Abbreviations 3V AHC AN BST DC LP MP MPO ox PDC PLM PM PMP PV PVN SON VMN

third ventricle anterior hypothalamic area, central arcuate nucleus bed nucleus of the stria terminalis paraventricular nuclear, dorsal cap paraventricular nucleus, lateral parvocellular paraventricular nucleus, medial parvocellular medial preoptic area optic chiasm paraventricular nucleus, dorsal cap paraventricular nucleus, lateral magnocellular paraventricular nucleus, posterior magnocellular paraventricular nucleus, medial parvocellular paraventricular nucleus, periventricular parvocellular paraventricular nucleus supraoptic nucleus ventromedial hypothalamic nucleus

References Adan, R.A.H., Cox, J.J., Beischlag, T.V. and Burbach, J.P.H. (1993) A composite hormone response element mediates the transactivation of the rat oxytocin gene by different classes of nuclear hormone receptors. Mol. Endocrinol., 7: 47–57. Akaishi, T. and Sakuma, Y. (1985) Estrogen excites oxytocinergic, but not vasopressinergic cells in the paraventricular nucleus of the female rat hypothalamus. Brain Res., 335: 302– 305. Alves, S.E., Lopez, V., McEwen, B.S. and Weiland, N.G. (1998) Differential colocalization of estrogen receptor β (ERβ) with oxytocin and vasopressin in the paraventricular and supraoptic nuclei of the female rat brain: an immunocytochemical study. Proc. Natl. Acad. Sci. USA, 95: 3281–3286. Amico, J.A., Crowley, R.S., Insel, T.R., Thomas, A. and O’Keefe,

J.A. (1995) Effect of gonadal steroids upon hypothalamic oxytocin expression. In: R. Ivell and J. Russell (Eds.), Oxytocin. Plenum Press, New York, pp. 23–35. 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. Brot, M.D., DeVries, G.J. and Dorsa, D.M. (1993) Local implants of testosterone regulate vasopressin mRNA in sexually dimorphic nuclei of the rat brain. Peptides, 14: 933–940. Burbach, J.P.H., Adan, R.A.H., Van Tol, H.H.M., Verbeeck, M.A.E., Axelson, J.F., van Leeuwen, F.W. and Beekman, J.M. (1990) Regulation of the rat oxytocin gene by estradiol. J. Neuroendocrinol., 2: 633–639. Caldwell, J.D., Brooks, P.J., Jirikowski, G.F., Barakat, A.S., Lund, P.K. and Peterson, C.A. (1989) Estrogen alters oxytocin mRNA levels in the preoptic area. J. Neuroendocrinol., 1: 273–278. Cechetto, D.F. and Saper, C.B. (1988) Neurochemical organization of the hypothalamic projection to the spinal cord in the rat. J. Comp. Neurol., 272: 579–604. Cintra, A., Fuxe, K., Härfstrand, A., Agnati, L.F., Miller, L.S., Greene, J.L. and Gustafsson, J.-Å. (1986) On the cellular localization and distribution of estrogen receptors in the rat tel- and diencephalon using monoclonal antibodies to human estrogen receptor. Neurochem. Int., 8: 587–595. Crowley, R.S., Insel, T.R., O’Keefe, J.A. and Amico, J.A. (1993) Cytoplasmic oxytocin and vasopressin gene transcripts decline postpartum in the hypothalamus of the lactating rat. Endocrinology, 133: 2704–2710. Dellovade, T.L., Zhu, Y.-S. and Pfaff, D.W. (1999) Thyroid hormones and estrogen affect oxytocin gene expression in hypothalamic neurons. J. Neuroendocrinol., 11: 1–10. DeVries, G.J., Buijs, R.M. and Sluiter, A.A. (1984) Gonadal hormone actions on the morphology of the vasopressinergic innervation of the adult rat brain. Brain Res., 298: 141–145. DeVries, G.J., Buijs, R.M., van Leeuwen, F.W., Caffé, A.R. and Swaab, D.F. (1985) The vasopressinergic innervation of the brain in normal and castrated rats. J. Comp. Neurol., 233: 1–19. DeVries, G.J., Duetz, W., Buijs, R.M., van Heerikhuize, J. and Vreeburg, J.T.M. (1986) Effects of androgens and estrogens on the vasopressin and oxytocin innervation of the adult rat brain. Brain Res., 399: 296–302. Hrabovszky, E., Kallo, I., Hajszan, T., Shughrue, P.J., Merchenthaler, I. and Liposits, Z. (1998) Expression of estrogen receptor-β messenger ribonucleic acid in oxytocin and vasopressin neurons of the rat supraoptic and paraventricular nuclei. Endocrinology, 139: 2600–2604. Ivell, R. and Richter, D. (1984) Structure and comparison of the oxytocin and vasopressin genes from rat. Proc. Natl. Acad. Sci. USA, 81: 2006–2010. Jirikowski, G.F., Caldwell, J.D., Peterson, C.A. and Stumpf, W.E. (1988) Estradiol influences oxytocin-immunoreactive brain systems. Neuroscience, 25: 237–248. Kawata, M., McCabe, J.T., Chung, S.K., Dutt, A., Yuri, K., Hirakawa, M., Kumamoto, K., Hirayama, Y. and Pfaff, D.W. (1991) The effect of progesterone on oxytocin messenger RNA in hypothalamic neurons of estrogen-treated female rats

29 studied with quantitative in situ hybridization histochemistry. Biomedical Res., 12: 405–415. Koike, S., Sakai, M. and Muramatsu, M. (1987) Molecular cloning and characterization of rat estrogen receptor cDNA. Nucleic Acids Res., 15: 2499–2513. Kuiper, G.G.J.M., Enmark, E., Pelto-Huikko, M., Nilsson, S. and Gustafsson, J.-Å. (1996) Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. USA, 93: 5925–5930. Kuiper, G.G.J.M., Shughrue, P.J., Merchenthaler, I. and Gustafsson, J.-Å. (1998) The estrogen β subtype: a novel mediator of estrogen action in neuroendocrine systems. Front. Neuroendocrinol., 19: 253–286. Laflamme, N., Nappi, R.E., Drolet, G., Labrie, C. and Rivest, S. (1998) Expression and neuropeptidergic characterization of estrogen receptors (ERα and ERβ) throughout the rat brain: anatomical evidence of distinct roles of each subtype. J. Neurobiol., 36: 357–378. Li, X., Schwartz, P.E. and Rissman, E.F. (1997) Distribution of estrogen receptor-β-like immunoreactivity in rat forebrain. Neuroendocrinology, 66: 63–67. Miller, M.A., DeVries, G.J., Al-Shamma, H.A. and Dorsa, D.M. (1992) Decline of vasopressin immunoreactivity and mRNA levels in the bed nucleus of the stria terminalis following castration. J. Neurosci., 12: 2881–2887. Miller, F.D., Ozimek, G., Milner, R.J. and Bloom, F.E. (1989a) Regulation of neuronal oxytocin mRNA by ovarian steroids in the mature and developing hypothalamus. Proc. Natl. Acad. Sci. USA, 86: 2468–2472. Miller, M.A., Urban, J.H. and Dorsa, D.M. (1989b) Steroid dependency of vasopressin neurons in the bed nucleus of the stria terminalis by in situ hybridization. Endocrinology, 125: 2335–2340. Mohr, E., Bahnsen, U., Kiessling, C.H. and Richter, D. (1988) Expression of the vasopressin and oxytocin genes in rats occurs in mutually exclusive sets of hypothalamic neurons. FEBS Lett., 242: 144–148. Negoro, H., Visessuwan, S. and Holland, R.C. (1973) Unit activity in the paraventricular nucleus of female rats at different stages of the reproductive cycle and after ovariectomy, with or without oestrogen or progesterone treatment. J. Endocrinol., 59: 545–558. Osterlund, M., Kuiper, G.G.J.M., Gustafsson, J.-Å. and Hurd, Y.L. (1998) Differential distribution and regulation of estrogen receptor-α and -β mRNA within the female rat brain. Mol. Brain Res., 54: 175–180. Pfaff, D.W. and Keiner, M. (1973) Atlas of estradiolconcentrating cells in the central nervous system of the female rat. J. Comp. Neurol., 151: 121–158. Rhodes, C.H., Morrell, J.I. and Pfaff, D.W. (1981) Distribution of estrogen-containing, neurophysin-containing magnocellular neurons in the rat hypothalamus as demonstrated by a technique combining steroid autoradiography and immunohistology in the same tissue. Neuroendocrinology, 33: 18–23. Sar, M. and Parikh, I. (1986) Immunohistochemical localization of estrogen receptor in rat brain, pituitary and uterus with monoclonal antibodies. J. Steroid Biochem., 24: 497–503.

Sawchenko, P.E. and Swanson, L.W. (1982) Immunohistochemical identification of neurons in the paraventricular nucleus of the hypothalamus that project to the medulla or to the spinal cord in the rat. J. Comp. Neurol., 205: 260–272. Shughrue, P.J. and Merchenthaler, I. (2001) The distribution of estrogen receptor β immunoreactivity in the rat central nervous system. J. Comp. Neurol., 436: 64–81. Shughrue, P.J., Komm, B.S. and Merchenthaler, I. (1996) The distribution of estrogen receptor-β mRNA in the rat hypothalamus. Steroids, 61: 678–681. Shughrue, P.J., Lane, M.V. and Merchenthaler, I. (1997a) The comparative distribution of estrogen receptor-α and β mRNA in the rat central nervous system. J. Comp. Neurol., 388: 507– 525. Shughrue, P., Scrimo, P., Lane, M., Askew, R. and Merchenthaler, I. (1997b) The distribution of estrogen receptor-β mRNA in forebrain regions of the estrogen receptor-α knockout mouse. Endocrinology, 138: 5649–5652. Shughrue, P.J., Lane, M.V. and Merchenthaler, I. (1997c) Regulation of progesterone receptor mRNA in the rat medial preoptic nucleus by estrogenic and antiestrogenic compounds: an in situ hybridization study. Endocrinology, 138: 5476–5484. Shughrue, P.J., Lane, M.V. and Merchenthaler, I. (1999) Biologically active estrogen receptor-β: evidence from in vivo autoradiographic studies with estrogen receptor α-knockout mice. Endocrinology, 140: 2613–2620. Shughrue, P.J., Lane, M.V. and Merchenthaler, I. (2000) Evidence for novel 125 I-estrogen binding sites in the rat hippocampus. Neuroscience, 99: 605–612. Shughrue, P.J., Askew, G.R., Dellovade, T.L. and Merchenthaler, I. (2002) Estrogen-binding sites and their functional capacity in estrogen receptor double knockout mouse brain. Endocrinology, 143: 1643–1650. Simerly, R.B., Chang, C., Muramatsu, M. and Swanson, L.W. (1990) Distribution of androgen and estrogen receptor mRNAcontaining cells in the rat brain: an in situ hybridization study. J. Comp. Neurol., 294: 76–95. Simonian, S.X. and Herbison, A.E. (1997) Differential expression of estrogen receptor α and β immunoreactivity by oxytocin neurons of rat paraventricular nucleus. J. Neuroendocrinol., 9: 803–806. Stumpf, W.E. (1970) Estrogen-neurons and estrogen-neuron systems in the periventricular brain. Am. J. Anat., 129: 207–218. Stumpf, W.E., Sar, M. and Keefer, D.A. (1975) Atlas of estrogen target cells in rat brain. In: W.E. Stumpf and L.D. Grant (Eds.), Anatomical Neuroendocrinology. Karger, Basel, pp. 104–119. Van Tol, H.M., Bolwerk, E.L.M., Liu, B. and Burbach, J.P.H. (1988) Oxytocin and vasopressin gene expression in the hypothalamo-neurohypophyseal system of the rat during the estrous cycle, pregnancy, and lactation. Endocrinology, 122: 945–951. Wagner, C.K. and Clemens, L.G. (1991) Projections of the paraventricular nucleus of the hypothalamus to the sexually dimorphic lumbosacral region of the spinal cord. Brain Res., 539: 254–262.

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 3

Short-term modulation of GABAA receptor function in the adult female rat Arjen B. Brussaard ∗ and Jan-Jurjen Koksma Department of Experimental Neurophysiology, Vrije Universiteit Amsterdam, Research Institute Neurosciences and Centre for Neurogenomics and Cognitive Research, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands

Abstract: Oxytocin neurons in the supraoptic nucleus (SON) exhibit marked neuronal plasticity during each reproductive cycle. We have previously shown that this neuronal plasticity includes GABAA receptor subunit switching around the time of parturition. Here we focus on additional plasticity in short-term regulatory mechanisms of postsynaptic receptor function before and after parturition, i.e. alterations in metabotropic and allosteric modulation of GABAA receptor activity. Both short- and long-term regulation of the GABAA receptor function affects the electrical behaviour of the oxytocin neurons (Brussaard and Herbison, 2000); however, their causal linkage until recently remained unclear. Non-genomic gonadal steroid feedback to oxytocin neurons is mediated via the neurosteroid allopregnanolone (3α-OH-DHP) that is an allosteric modulator of postsynaptic GABAA receptors. We recently found evidence to support the idea that (1) neurosteroids not only potentiate GABAA receptor function but also prevent its suppression by PKC (Brussaard et al., 2000), and (2) that neurosteroid sensitivity of GABAA receptor is not regulated by subunit switching, but instead, is dependent on the balance between endogenous phosphatase and PKC activity (Koksma et al., 2002). Thus, before pregnancy, the GABAA receptors are sensitive to 3α-OH-DHP, due to a constitutively high level of phosphatase activity. At parturition, endogenous release of oxytocin within the SON shifts the intracellular balance towards a higher level of phosphorylation, leading to 3α-OH-DHP insensitivity of the GABAA receptors. Here we discuss the putative mechanisms underlying these changes in receptor physiology, their causal relations and the functional significance for the hormonal output. Keywords: GABAA receptor; Lactation; Parturition; Synaptic plasticity; Oxytocin; Progesterone-metabolite; Allopregnanolone; PKC; Phosphatases

Introduction The magnocellular oxytocin neurons reside within the supraoptic nuclei (SON) and paraventricular nuclei of the hypothalamus from where they project to the posterior pituitary and secrete oxytocin directly into the circulation. The electrical and biosynthetic activity of these neurons and consequently the

∗ Correspondence

to: A.B. Brussaard; Department of Neurophysiology, Vrije Universiteit Amsterdam, de Boelelaan 1087, 1081 HV Amsterdam, The Netherlands. Tel.: +3120-444-7098; Fax: +31-20-444-7112; E-mail: brssrd@ bio.vu.nl

oxytocin levels in blood-plasma undergo substantial changes over the reproductive cycle in the female (Summerlee, 1981; Leng et al., 1999). Oxytocin neurons exhibit a low level of firing during pregnancy but then display an abrupt transition to synchronous bursting behaviour, on top of elevated tonic firing, at the time of birth when oxytocin helps contracting the myometrium of the uterus. Shortly afterwards, the oxytocin neurons again display episodic high-frequency bursting behaviour in response to suckling, with oxytocin now acting on myoepithelial cells of the mammary glands to stimulate lactation. After the cessation of lactation, the activity of oxytocin neurons returns to a low baseline level. Thus, the hypothalamic oxytocin neurons display marked

32

changes in neuronal activity with each cycle of pregnancy and lactation. As reviewed previously (Brussaard and Herbison, 2000), it appears that different forms of plasticity underlie the cyclic changes in neuronal activity of the oxytocin neurons. Morphological analyses in the rat have revealed a broad synaptic remodelling and neuronal–glial changes during pregnancy and lactation. These include alterations in somatic and dendritic appositions between the parallel-oriented oxytocin neurons, as well as changes in the formation of gap junctions and synapses (Theodosis and Poulain, 1993; Hatton, 1997; El Majdoubi et al., 1997; Stern and Armstrong, 1998). Although numerous factors, including the retention of certain embryological features by adult oxytocin neurons, might be involved in causing neuronal changes in the adult SON, the large sustained fluctuations in circulating progesterone and oestrogen concentrations during pregnancy and lactation are likely to be the principal factors that induce and coordinate these changes. GABA regulates oxytocin neuron activity in vivo Approximately half of all axosomatic and axodendritic synapses on oxytocin neurons are GABAergic in nature, and essentially all spontaneous inhibitory postsynaptic currents (sIPSCs) in these cells arise through activation of the GABAA receptor (Randle and Renaud, 1987; Wuarin and Dudek, 1993). These GABAA receptors are composed of pentameric combinations of α1 and/or α2, β2 and/or β3 and γ2 subunits of the GABAA receptor (Fenelon et al., 1995; Fenelon and Herbison, 1996). Studies undertaken in vivo have highlighted the importance of GABAA receptor activation in enabling oxytocin neurons to exhibit intermittent high-frequency bursting behaviours during lactation (Moos, 1995; Voisin et al., 1995). In particular, microinfusion of GABAA receptor antagonists into the SON prohibits the occurrence of high-frequency bursts of action potentials by oxytocin neurons in response to suckling. Accordingly, it has been proposed that GABAA receptor activation is responsible for the periods of non-firing between bursts of action potentials, and thereby maintains the membrane potential of oxytocin neurons within a range that is compatible with bursting behaviour. Other neurochemical inputs are thought to induce the

episodes of synchronized firing (Voisin et al., 1994; Jourdain et al., 1998). Experimental difficulties have prevented detailed in vivo analysis of the role of GABAA receptors in regulating the oxytocin neurons during pregnancy and parturition. However, our studies have shown that the administration of a GABAA receptor antagonist in late pregnancy results in a massive increase in oxytocin firing in vitro, and oxytocin secretion in vivo (Brussaard et al., 1997). Unlike the situation during lactation, where the GABA-mediated input promotes bursting firing patterns by means of a dynamic interaction with excitatory inputs, during late pregnancy, the GABAA receptor input, by means of tonic inhibition, is predominantly involved in restraining the oxytocin neurons from firing. Thus GABAA receptor signalling exerts powerful, reproductive state-specific effects on the functioning of oxytocin neurons. The changes in GABA-mediated input on oxytocin neurons over the approximately 40-day pregnancy–parturition–lactation cycle in the rat have formed the trigger to study long-term plasticity of GABAA receptor-mediated synaptic transmission. GABAA receptor subunit switching affects synaptic current decay Long-term plasticity of postsynaptic GABAA receptors results from alterations in the subunit composition of these receptors expressed by oxytocin neurons (Brussaard et al., 1997, 1999; Brussaard and Herbison, 2000). In particular, it was found that the gene expression of the α1 subunit was elevated during the course of pregnancy, such that peak levels were obtained on day 19 of pregnancy and that cellular mRNA levels then declined precipitously over the final two days of pregnancy, before the onset of parturition. Electrophysiological analysis over the last two days of pregnancy revealed that this change in α1 subunit expression correlated with the presence of GABAA receptors displaying distinct kinetic and pharmacological properties. At late pregnancy, the α1 subunit dominant GABAA receptor subtype is characterized by a relatively fast monoexponential synaptic current decay (time constant 18 ± 4 ms) that is prolonged by the neurosteroid allopregnanolone (3α-OH-DHP), leading to time constant values be-

33

tween 40 and 50 ms. In contrast, two days later, at the time of parturition, the GABAA receptor subtype changes to one with a significantly slower synaptic current decay (26±6 ms, Brussaard et al., 1997), and this isoform persists throughout subsequent lactation (Brussaard et al., 1999). Time-wise in line with this, but not causally related to this subunit switch, is the alteration in pharmacological profile of the GABAA receptor: i.e. upon parturition the GABAA receptor was found to be markedly less sensitive to 3α-OHDHP (Brussaard et al., 1997; and see below). In line with classical models that suggest that the decay of sIPSCs is determined by the kinetic properties of the postsynaptic receptors (Mody et al., 1994), we substantiated our positive correlation between high α1-subunit expression and fast sIPSC decay in the oxytocin neurons by means of antisense-mediated deletion experiments (Brussaard et al., 1997). Thus, the endogenous changes in the ratio of α1 : α2-subunit mRNA in oxytocin neurons over the reproductive cycle engender a functional plasticity upon postsynaptic GABAA receptors at the level of individual synapses. Unresolved issues in GABAA receptor regulation This leaves unanswered what the functional significance of subunit switching of GABAA receptors for the electrical activity of the oxytocin might be. Current research is aimed at solving this issue. However, we speculate that changes in the type of presynaptic GABAergic cells innervating the oxytocin neurons in the SON around parturition may induce an apparent subunit switch at the level of the postsynaptic receptors. Alternatively, α2 containing GABAA receptors may represent a ‘juvenile’ (see Laurie et al., 1992) or transient (i.e. metabolically unstable) type of GABAA receptor, that is only expressed transiently during the post-parturition period. Whatever the functional significance, current thinking includes the idea that both α1 and α2 containing GABAA receptors are still expressed upon parturition, but at a different ratio compared to late pregnancy (Brussaard et al., 1997, 1999) and most likely segregated to distinct type of synapses and/or subcellular compartments (Koksma, Fritschy and Brussaard, in prep.). Here we address three other research questions: (1) is there a causal relationship between GABAA

receptor subunit switching (Brussaard et al., 1997) and alterations in short-term modulation of this receptor type by allosteric or metabotropic modulators (Brussaard and Herbison, 2000)? and, concerning the nature of this short-term modulation: (2) does allosteric interaction between the GABAA receptor and the extracellular 3α-OH-DHP molecules affect the intracellular signal transduction routes of receptor regulation? and, vice versa, (3) do changes in intracellular signal transduction cascades affect neurosteroid sensitivity? Subunit composition of GABAA receptors and neurosteroid sensitivity Interestingly, neurosteroid sensitivity of GABAA receptors differs in different brain areas (Nguyen et al., 1995; Poisbeau et al., 1997; Cooper et al., 1999), suggesting a relation with receptor subunit composition. Indeed, in vitro expression studies indicated that GABAA receptor subunit composition determines neurosteroid sensitivity (Puia et al., 1990; Lambert et al., 1995; Belelli et al., 1996). However, most subunit combinations are sensitive to the allosteric effect of neurosteroids (Lambert et al., 1995) and conflicting data remain on the few subunits that have been implicated in reducing neurosteroid sensitivity of GABAA receptors (Zhu et al., 1994; Davies et al., 1997; Whiting et al., 1997; Smith et al., 1998a,b). In our lab, in a recent series of experiments, we found that GABAA receptors in the SON neurons of WT mice and α1−/− mice are equally sensitive to potentiation by 3α-OH-DHP (Koksma et al., 2002). Thus, in mouse SON neurons, the α1 subunit is not required for allosteric modulation of GABAA receptors by 3α-OH-DHP. This implies that a downregulation of this subunit observed in rats around parturition (Brussaard et al., 1997) is not likely to be the direct cause for GABAA receptor resistance to 3α-OH-DHP. Hence, the answer to research question (1) is negative, and instead, we hypothesized that an alternative mechanism may regulate neurosteroid sensitivity in SON neurons, i.e. posttranslational modification of GABAA receptor. Phosphorylation of serine/threonine residues on β or γ subunits (McDonald and Moss, 1997) may exert direct effects on GABAA receptor properties. More in particular, phosphorylation may affect neu-

34

rosteroid modulation of GABAA receptors (Leidenheimer and Chapell, 1997; Fancsik et al., 2000). In order to answer research question (2) and (3) we investigated in detail the causal relationships between neurosteroid regulation of GABAA receptors and intracellular shifts in the balance between protein kinase and phosphatase activity. Functional expression of neurosteroid-insensitive GABAA receptors upon parturition The postsynaptic GABAA receptors in the SON are susceptible to allosteric modulation by 3α-OH-DHP during some stages of the female reproductive cycle, in particularly during pregnancy (Brussaard et al., 1997, 1999). At late pregnancy in mammals the endogenous levels of progesterone, and consequently those of 3α-OH-DHP are very high. These high concentrations of progesterone are maintained until term, when progesterone levels show an abrupt fall (Concas et al., 1998; and see Fig. 1). The concentration of progesterone comes up again during lactation. The levels of 3α-OH-DHP in the brain follow the changes in progesterone. When 3α-OH-DHP is present in pregnant rats, it causes an increase of the synaptic current decay time constant, without affecting the amplitude of the GABAergic events. This effect in itself is strong enough to effectively inhibit cell firing of oxytocin neurons during pregnancy (Brussaard et al., 1997). As reported and reviewed previously, the increasing levels of circulating 3α-OH-DHP with advancing pregnancy (Concas et al., 1998) strongly potentiate the overall synaptic efficacy of the GABA input (by around 40%), in particular by directly increasing the sIPSC decay time constant (Fig. 2A). This implies that neurosteroid potentiation of postsynaptic receptor activity is a predominant factor in the alteration of the impact of the GABA input observed during pregnancy. This marked increase in synaptic inhibition underlies the potent tonic restraining influence of the GABAergic input upon oxytocin neuron firing in late pregnancy and provides oxytocin neurons with a powerful safety switch to prevent premature release of oxytocin. Around parturition GABAA receptors in oxytocin neurons become insensitive to modulation by 3α-OHDHP (Brussaard et al., 1997; Fig. 2A). The removal

Fig. 1. Levels of progesterone and its metabolite allopregnanolone in the brain abruptly drop before parturition (P10–P21 = pregnancy days 10, 15, etc.) and rise again on post-parturition days (PPD) and during lactation (PPD2, etc.). Concentrations are measured by HPLC and presented in ng/g of cortex tissue. Dotted curves after PPD2 were extrapolated using data from other species. Horizontal solid line (grey) indicates basal non-reproductive (= oestrus) levels. Adapted from Concas et al. (1998).

of 3α-OH-DHP potentiation of sIPSCs is entirely responsible for the fall in synaptic efficacy between late pregnant and parturitient rats (Brussaard and Herbison, 2000). Most likely, this comes about both through the change in functional expression of 3αOH-DHP-insensitive GABAA receptors by oxytocin neurons around that time and the fall in circulating 3αOH-DHP concentrations prior to parturition (Fig. 1).

35

Fig. 2. Dis-inhibition of oxytocin cells upon parturition. (A) The neurosteroid 3α-OH-DHP potentiates sIPSC decay time constants during late pregnancy, but not after parturition. (B) Oxytocin receptor activation after parturition decreases sIPSC amplitude via a PKC-dependent mechanism.

36

37

The aggregate of these changes will gradually disinhibit oxytocin neurons over the last day of pregnancy, which, in association with temporally distinct increases in glutamate- and noradrenaline-mediated transmission (Herbison et al., 1997; Kombian et al., 1997; Leng et al., 1999; see also Fig. 3), enables oxytocin neuron activation at the time of parturition. Neurosteroid affects PKC-mediated modulation of GABAA receptors In addition to neurosteroid regulation of the oxytocin neurons, somatodendritic release of oxytocin within the SON acting on autoreceptors (Ludwig, 1998) may regulate the fast synaptic inhibition of these magnocellular cells, via suppression of postsynaptic GABAA receptor activity in a Ca2+ -dependent manner (Brussaard et al., 1996). We recently showed that the reduction of sIPSC amplitude triggered by oxytocin, is mediated by PKC (Brussaard et al., 2000; and see Fig. 2B). Next, since 3α-OH-DHP and oxytocin both affect the activity of the postsynaptic GABAA receptor, we also investigated whether allosteric interaction of 3α-OH-DHP with this receptor may influence its modulation by metabotropic pathways. Much to our surprise, we found that pretreatment with 3α-OHDHP interferes with the ability of oxytocin to reduce the sIPSC amplitude in oxytocin neurons (Brussaard et al., 2000; Fig. 3A). The novelty of this finding is that non-genomic, allosteric interaction of 3α-OH-DHP with the neurosteroid-sensitive GABAA receptor subtype in oxytocin neurons of the SON prevents PKC from phosphorylating the GABAA receptor itself or one of its interacting proteins. The conformational change induced by neurosteroid binding to the receptor that causes prolongation of the ion channel open time (Twyman and McDonald, 1992) makes one or more PKC phosphorylation site(s) of the GABAA receptor

inaccessible. Alternatively, it alters hitherto unknown receptor–protein interactions that may indirectly depend on PKC activity. Our findings do not necessarily demonstrate that direct phosphorylation of GABAA receptors by PKC is prevented; however, several phosphorylation sites are present at the γ2 and the β2 subunit (Kellenberger et al., 1992; Moss et al., 1992), both of which are expressed in the SON (Fenelon and Herbison, 1996). We would like to outline here that activation of PKC has variable effects on GABAA receptor activity (Sigel and Baur, 1988; Leidenheimer et al., 1992; Leidenheimer and Chapell, 1997; Poisbeau et al., 1999), which are likely to be caused by cell-specific differentiation in GABAA receptor subunit composition or by expression of different PKC isomers. In addition, differences in PKC-dependent effects may be brought about by heterogeneity in the expression of other proteins that interact or associate with the GABAA receptor from the intracellular side (Brandon et al., 1999; Kneussel et al., 1999; Wang et al., 1999). This finding has major implications for our view of the induction of oxytocin release at the onset of parturition. The physiological relevance of 3α-OHDHP signaling is largest during pregnancy, when the neurosteroid potentiates inhibition of oxytocin neurons by the combination of prolongation of the sIPSC decay and a block of PKC-dependent suppression of the GABAA receptors. This combination provides an efficient mechanism to prevent premature oxytocin release (Fig. 3A). The effect of 3α-OH-DHP on the sIPSC decay leads to a >2-fold increase in the synaptic efficacy of the GABA input (Brussaard et al., 1999). Thus, during pregnancy, the overall impact of GABAergic transmission in the SON is ‘locked’ in a potentiated mode, a condition which is not under control of oxytocin autoregulation, due to the continuous presence of allopregnanolone. As mentioned above, this is sufficient to silence the firing activity of oxytocin neurons.

Fig. 3. Bidirectional interaction between metabotropic and allosteric modulation of the GABAA receptor. (A) High levels of 3α-OH-DHP in the brain during late pregnancy prevent PKC-mediated suppression of sIPSC amplitude. (B) Fall in 3α-OH-DHP levels renders GABAA receptors susceptible to PKC modulation and increasing somatodendritic release of oxytocin gives rise to high levels of activated PKC. The reduced GABA input acts in concert with glutamatergic and noradrenergic input to excite oxytocin neurons. (C) High levels of oxytocin release make GABAA receptors resistant to 3α-OH-DHP after parturition and consequently rising levels of 3α-OH-DHP do not affect GABA currents (adapted from Gimpl and Fahrenholz, 2001).

38

Then, at parturition (Fig. 3B), the levels of 3αOH-DHP fall (Fig. 1) and the neurosteroid no longer controls the firing activity, and also fails to prevent the autoregulatory action of oxytocin within the SON (Brussaard et al., 2000). As a result, at the postpartum and subsequent lactation stage (Fig. 3C), disinhibition of oxytocin neurons occurs via a reduction in tonic GABAergic synaptic input (Brussaard et al., 1996, 2000), giving way to other excitatory synaptic input (Herbison et al., 1997). What regulates neurosteroid sensitivity of GABAA receptors? Although allosteric modulation of GABA-mediated currents by neurosteroids (Majewska, 1992) is well established, the underlying molecular mechanism is poorly understood. The naturally occurring change from neurosteroid-sensitive receptor being functionally expressed in the SON during pregnancy to the neurosteroid-resistant GABAA receptors occurring after parturition provides a good model to study how neurosteroid sensitivity of GABAA receptors is regulated in general. As indicated above, subunit composition of the GABAA receptor is not likely to play any significant role in this. However, since neurosteroids through direct interaction with the GABAA receptor may prevent PKC dependent suppression of GABAA receptor function, next, we investigated whether vice versa, PKC activity also influences neurosteroid sensitivity of GABAA receptors in oxytocin neurons. We found that the endogenous GABAA receptor sensitivity to 3α-OH-DHP in SON neurons during late pregnancy is brought about by an endogenously regulated constitutively high level of phosphatase activity, and can be suppressed temporarily by inducing a shift towards a relatively higher level of activity of PKC (Koksma et al., 2002). Vice versa, after parturition, when GABAA receptors are insensitive to neurosteroids in these cells, both inhibition of PKC and stimulation of Ca2+ -dependent phosphatase 2A, restored the 3α-OH-DHP sensitivity. We therefore propose that phosphatases and PKC have a converging effect, possibly acting on the same serine/threonine residue of one of the non-α GABAA receptor subunits (McDonald and Moss, 1997). Alternatively, PKC and phosphatases could also act on different phosphorylation sites, or on other proteins

of the postsynaptic density of the GABA synapse, including receptor-interacting proteins like RACK-1 (Brandon et al., 1999), receptor-clustering proteins like gephyrin (Kneussel et al., 2001) or phosphatasetargeting proteins like spinophilin (Hsieh-Wilson et al., 1999) that might serve as a scaffold for the action of these enzymes. This phosphorylation-dependent regulation of the effect of an allosteric modulator is distinct from direct effects of phosphorylation of residues of β and γ subunits on GABAA receptor channel desensitization properties (Krishek et al., 1994). In oxytocin neurons we did not observe any direct effect on sIPSC decay time constants of alteration in protein kinase or phosphatase activity, such as reported for hippocampal neurons (Jones and Westbrook, 1997; Poisbeau et al., 1999). Physiological significance of neurosteroid resistance of GABAA receptors Obviously, the above raises the question what causes this shift in signal transduction of the postsynaptic neuron. We hypothesized that a shift towards higher levels of PKC activity may be explained by endogenous activation of oxytocin autoreceptors upon parturition. Thus, we asked ourselves whether metabotropic modulation of the GABAA receptor induced by oxytocin (Brussaard et al., 2000) is the cause of neurosteroid resistance of GABAA receptors as observed at the postpartum stage. The answer to this research question was positive, i.e. in lactating females that normally display 3α-OH-DHPresistance of their GABAA receptors, we were able to restore receptor sensitivity to this neurosteroid by pretreating the SON with an oxytocin antagonist (Koksma et al., 2002). In summary, PKC and 3α-OH-DHP have distinct effects on the GABAA receptor and these effects are mutually exclusive. During pregnancy, when 3αOH-DHP levels are high and there is little oxytocin secretion, GABAA receptors are occupied by 3αOH-DHP preventing PKC modulation (Fig. 3A). Then, upon parturition 3α-OH-DHP levels abruptly drop and the somatodendritic release of oxytocin goes up. Now many GABAA receptors have become vulnerable to modulation by PKC (Fig. 3B). During lactation the concentration of 3α-OH-DHP

39

may rise again but high intranuclear oxytocin concentrations prevent any allosteric effect (Fig. 3C). This bidirectional interaction between metabotropic and allosteric modulation of GABAA receptors not only helps to explain the onset of parturition, but might be a mechanism occurring at GABA synapses throughout the brain.

Plasticity in short-term modulation of GABAA receptor function Fig. 4 presents a current working model of how this bidirectional interaction between allosteric and metabotropic receptor modulation could be brought about at the level of the transmembrane receptor protein(s). For simplicity we have excluded receptorinteraction (postsynaptic density) proteins, that may

Fig. 4. Receptor model for bidirectional interaction between PKC and 3α-OH-DHP. (A) During pregnancy the balance between phosphatase and PKC activity is in the direction of dephosphorylation, rendering GABAA receptors susceptible to 3α-OH-DHP modulation. When 3α-OH-DHP is bound to the receptor PKC cannot. (B) After parturition the balance is shifted towards higher levels of PKC activity. Phosphorylation of the receptor affects the receptor in such a way that the binding site for 3α-OH-DHP is no longer exposed, or 3α-OH-DHP is no more able to alter receptor properties.

40

also be phosphorylated and/or de-phosphorylated to mediate the alterations in receptor function. We conclude that neurosteroid sensitivity of GABAA receptors in the SON can be manipulated both by alterations in phosphatase and PKC activity (Koksma et al., 2002). This suggests that phosphatases and PKC converge onto the same site of action, possibly a phosphorylation site of the γ2 or the β2 subunit receptor subunits, that are expressed in the SON (Fenelon and Herbison, 1996). In such a scenario, neurosteroid binding to the receptor prevents the intracellular action of PKC, whereas vice versa, the neurosteroid-binding site of the GABAA receptor is only exposed if the receptor is dephosphorylated by phosphatases. At parturition there is a dual effect of PKC on GABAA receptors: i.e. a reduction of their open probability (Fig. 2B) and reduction of their neurosteroid sensitivity (see Fig. 4B). This dual effect might be caused by differences in the degree to which the receptor is phosphorylated (Fig. 4B). Concluding remarks We reviewed here the functional significance and the putative underlying mechanism of the dual effect of 3α-OH-DHP on GABAA receptor function, i.e. this neurosteroid not only prolongs the synaptic current decay of GABAA receptor-mediated events, but also prevents subsequent metabotropic modulation by oxytocin autoreceptor activation within the SON (Brussaard et al., 2000). This mechanism previously was shown to be essential for silencing the oxytocin neurons throughout the entire pregnancy period (Brussaard and Herbison, 2000). Secondly, we discourage the idea that the suppression of neurosteroid potentiation of GABAA receptors is induced by alterations in receptor subunit composition. Although the shift in α1 : α2 subunit ratio of the GABAA receptors in these cells correlates well with the change in the pharmacological behaviour of these receptors in dorsomedial SON neurons around the time of parturition (Brussaard et al., 1997), there is currently no evidence that this subunit switch is causally related to neurosteroid resistance observed after parturition. Instead, third, we conclude that neurosteroid modulation of GABAA receptors depends on the balance between Ca2+ -dependent phosphatase and protein ki-

nase C activity (Koksma et al., 2002). We hypothesize that the tonic release of oxytocin, via a shift in signal transduction of the postsynaptic cell, leads to neurosteroid insensitivity upon parturition. The resultant dis-inhibition of oxytocin neurons after parturition is believed to play a determinant role in the timing of oxytocin neuron activity. A similar mechanism of GABAA receptor regulation may contribute to ethanol sensitivity, since GABAA receptor susceptibility to ethanol has been reported to be altered in transgenic mice lacking particular PKC isomers (Harris et al., 1995; Hodge et al., 1999). Thus, bidirectional interaction between metabotropic and allosteric modulation of GABAA receptors may be a phenomenon that regulates the efficacy of GABA synapses not only in the SON, but throughout the brain. Acknowledgements The authors acknowledge support by the Dutch Organization for Scientific Research (ALW-N.W.O.). References Belelli, D., Lambert, J.J., Peters, J.A., Gee, K.W. and Lan, N.C. (1996) Modulation of human recombinant GABAA receptors by pregnanediols. Neuropharmacology, 35: 1223–1231. Brandon, N.J., Uren, J.M., Kittler, J.T., Wang, H., Olsen, R., Parker, P.J. and Moss, S.J. (1999) Subunit-specific association of protein kinase C and the receptor for activated C kinase with GABA type A receptors. J. Neurosci., 19: 9228–9234. Brandon, N.J., Delmas, P., Kittler, J.T., McDonald, B.J., Sieghart, W., Brown, D.A., Smart, T.G. and Moss, S.J. (2000) GABAA receptor phosphorylation and functional modulation in cortical neurons by a protein kinase C-dependent pathway. J. Biol. Chem., 275: 38856–38862. Brooks-Kayal, A.R., Shumate, M.D., Jin, H., Rikhter, T.Y. and Coulter, D.A. (1998) Selective changes in single cell GABA(A) receptor subunit expression and function in temporal lobe epilepsy. Nat. Med., 4: 1166–1172. Brussaard, A.B. and Herbison, A.E. (2000) Long-term plasticity of postsynaptic GABAA -receptor function in the adult brain: insights from the oxytocin neurone. Trends Neurosci., 23(5): 190–195. Brussaard, A.B., Kits, K.S. and de Vlieger, T.A. (1996) Postsynaptic mechanism of depression of GABAergic synapses by oxytocin in the supraoptic nucleus of immature rat. J. Physiol., 497: 495–507. Brussaard, A.B., Kits, K.S., Baker, R.E., Willems, W.P., LeytingVermeulen, J.W., Voorn, P., Smit, A.B., Bicknell, R.J. and Herbison, A.E. (1997) Plasticity in fast synaptic inhibition of

41 adult oxytocin neurons caused by switch in GABA(A) receptor subunit expression. Neuron, 19: 1103–1114. Brussaard, A.B., Devay, P., Leyting-Vermeulen, J.L. and Kits, K.S. (1999) Changes in properties and neurosteroid regulation of GABAergic synapses in the supraoptic nucleus during the mammalian female reproductive cycle. J. Physiol., 516: 513– 524. Brussaard, A.B., Wossink, J., Lodder, J.C. and Kits, K.S. (2000) Progesterone-metabolite prevents protein kinase C-dependent modulation of gamma-aminobutyric acid type A receptors in oxytocin neurons. Proc. Natl. Acad. Sci. USA, 97: 3625–3630. Concas, A., Mostallino, M.C., Porcu, P., Follesa, P., Barbaccia, M.L., Trabucchi, M., Purdy, R.H., Grisenti, P. and Biggio, G. (1998) Role of brain allopregnanolone in the plasticity of gamma-aminobutyric acid type A receptor in rat brain during pregnancy and after delivery. Proc. Natl. Acad. Sci. USA, 95: 13284–13289. Corpechot, C., Young, J., Calvel, M., Wehrey, C., Veltz, J.N., Touyer, G., Mouren, M., Prasad, V.V., Banner, C. and Sjovall, J. et al. (1993) Neurosteroids: 3 alpha-hydroxy-5 alphapregnan-20-one and its precursors in the brain, plasma, and steroidogenic glands of male and female rats. Endocrinology, 133(3): 1003–1009. Cooper, E.J., Johnston, G.A.R. and Edwards, F.A. (1999) Effects of a naturally occurring neurosteroid on GABAA IPSCs during development in rat hippocampal or cerebellar slices. J. Physiol., 521: 437–449. Davies, P.A., Hanna, M.C., Hales, T.G. and Kirkness, E.F. (1997) Insensitivity to anaesthetic agents conferred by a class of GABA(A) receptor subunit. Nature, 385: 820–823. El Majdoubi, M., Poulain, D.A. and Theodosis, D.T. (1997) Lactation-induced plasticity in the supraoptic nucleus augments axodendritic and axosomatic GABAergic and glutamatergic synapses: an ultrastructural analysis using the dissector method. Neuroscience, 80: 1137–1147. Fancsik, A., Linn, D.M. and Tasker, J.G. (2000) Neurosteroid modulation of GABA IPSCs is phosphorylation dependent. J. Neurosci., 20: 3067–3075. Fenelon, V.S. and Herbison, A.E. (1996) Plasticity in GABAA receptor subunit mRNA expression by hypothalamic magnocellular neurons in the adult rat. J. Neurosci., 16: 4872–4880. Fenelon, V.S., Sieghart, W. and Herbison, A.E. (1995) Cellular localization and differential distribution of GABAA receptor subunit proteins and messenger RNAs within hypothalamic magnocellular neurons. Neuroscience, 64: 1129–1143. Harris, R.A., McQuilkin, S.J., Paylor, R., Abeliovich, A., Tonegawa, S. and Wehner, J.M. (1995) Mutant mice lacking the gamma isoform of protein kinase C show decreased behavioral actions of ethanol and altered function of gammaaminobutyrate type A receptors. Proc. Natl. Acad. Sci. USA, 92: 3658–3662. Hatton, G.I. (1997) Function-related plasticity in hypothalamus. Annu. Rev. Neurosci., 20: 375–397. Herbison, A.E., Voisin, D.L., Douglas, A.J. and Chapman, C. (1997) Profile of monoamine and excitatory amino acid release in rat supraoptic nucleus over parturition. Endocrinology, 138: 33–40.

Hodge, C.W., Mehmert, K.K., Kelley, S.P., McMahon, T., Haywood, A., Olive, M.F., Wang, D., Sanchez-Perez, A.M. and Messing, R.O. (1999) Supersensitivity to allosteric GABA(A) receptor modulators and alcohol in mice lacking PKC epsilon. Nat. Neurosci., 2: 997–1002. Hsieh-Wilson, L.C., Allen, P.B., Watanabe, T., Nairn, A.C. and Greengard, P. (1999) Characterization of the neuronal targeting protein spinophilin and its interactions with protein phosphatase-1. Biochemistry, 38(14): 4365–4373. Israel, J.M. and Poulain, D.A. (2000) 17-Oestradiol modulates in vitro electrical properties and responses to kainate of oxytocin neurones in lactating rats. J. Physiol., 524: 457–470. Jones, M.V. and Westbrook, G.L. (1997) Shaping of IPSCs by endogenous calcineurin activity. J. Neurosci., 17: 7626–7633. list. Jourdain, P., Israel, J.M., Dupouy, B., Oliet, S.H., Allard, M., Vitiello, S., Theodosis, D.T. and Poulain, D.A. (1998) Evidence for a hypothalamic oxytocin-sensitive pattern-generating network governing oxytocin neurons in vitro. J. Neurosci., 18: 6641–6649. Kellenberger, S., Malherbe, P. and Sigel, E. (1992) Function of the alpha 1 beta 2 gamma 2S gamma-aminobutyric acid type A receptor is modulated by protein kinase C via multiple phosphorylation sites. J. Biol. Chem., 267: 25660–25663. Kneussel, M., Brandstatter, J.H., Laube, B., Stahl, S., Muller, U. and Betz, H. (1999) Loss of postsynaptic GABA(A) receptor clustering in gephyrin-deficient mice. J. Neurosci., 19: 9289– 9297. Kneussel, M., Brandstatter, J.H., Gasnier, B., Feng, G., Sanes, J.R. and Betz, H. (2001) Gephyrin-independent clustering of postsynaptic GABA(A) receptor subtypes. Mol. Cell. Neurosci., 17(6): 973–982. Koksma, J.J., Van Kesteren, R.E., Rosahl, T.W., Zwart, R., Smith, A.B., Lüddens, H. and Brussaard, A.B. (2002) Oxytocin regulates neurosteroid modulation of GABAA receptors in nucleus around parturition, in prep. Kokama, J.J., Fritschy, J.M. and Brussaard, A.B. (2002) in prep. Kombian, S.B., Mouginot, D. and Pittman, Q.J. (1997) Dendritically released peptides act as retrograde modulators of afferent excitation in the supraoptic nucleus in vitro. Neuron, 19(4): 903–912. Krishek, B.J., Xie, X., Blackstone, C., Huganir, R.L., Moss, S.J. and Smart, T.G. (1994) Regulation of GABAA receptor function by protein kinase C phosphorylation. Neuron, 12: 1081–1095. Lambert, J.J., Belelli, D., Hill-Venning, C. and Peters, J.A. (1995) Neurosteroids and GABAA receptor function. Trends Pharm. Sci., 16: 295–303. Laurie, D.J., Seeburg, P.H. and Wisden, W. (1992) The distribution of thirteen GABAA receptor subunit mRNAs in the rat brain, III. Embryonic and postnatal development. J. Neurosci., 12: 4151–4172. Leidenheimer, N.J. and Chapell, R. (1997) Effects of PKC activation and receptor desensitization on neurosteroid modulation of GABA(A) receptors. Mol. Brain Res., 52: 173–181. Leidenheimer, N.J., McQuilkin, S.J., Hahner, L.D., Whiting, P. and Harris, A. (1992) Activation of protein kinase C selec-

42 tively inhibits the gamma-aminobutyric acid A receptor: role of desensitization. Mol. Pharm., 41: 1116–1123. Leng, G., Brown, C.H. and Russell, J.A. (1999) Physiological pathways regulating the activity of magnocellular neurosecretory cells. Prog. Neurobiol., 57: 625–655. Ludwig, M. (1998) Dendritic release of vasopressin and oxytocin. J. Neuroendocrinol., 10: 881–895. Majewska, M.D. (1992) Neurosteroids: endogenous bimodal modulators of the GABAA receptor. Mechanism of action and physiological significance. Prog. Neurobiol., 38: 379–385. McDonald, B.J. and Moss, S.J. (1997) Conserved phosphorylation of the intracellular domains of GABA(A) receptor beta2 and beta3 subunits by cAMP-dependent protein kinase, cGMP-dependent protein kinase protein kinase C and Ca2+ /calmodulin type II-dependent protein kinase. Neuropharmacology, 36: 1377–1385. Mody, I., De Koninck, Y., Otis, T.S. and Soltesz, I. (1994) Bridging the cleft at GABA synapses in the brain. Trends Neurosci., 17: 517–524. Moos, F.C. (1995) GABA-induced facilitation of the periodic bursting activity of oxytocin neurones in suckled rats. J. Physiol., 488: 103–114. Moss, S.J., Doherty, C.A. and Huganir, R.L. (1992) Identification of the cAMP-dependent protein kinase and protein kinase C phosphorylation sites within the major intracellular domains of the beta 1, gamma 2S, and gamma 2L subunits of the gamma-aminobutyric acid type A receptor. J. Biol. Chem., 267: 14470–14476. Nguyen, Q., Sapp, D.W., van Ness, P.C. and Olsen, R.W. (1995) Modulation of GABAA receptor binding in human brain by neuroactive steroids: species and brain regional differences. Synapse, 19: 77–87. Poisbeau, P., Felz, P. and Schlichter, R. (1997) Modulation of GABAA receptor-mediated IPSCs by neuroactive steroids in a rat hypothalamo-hypophyseal coculture model. J. Physiol., 500: 475–485. Poisbeau, P., Cheney, M.C., Browning, M.D. and Mody, I. (1999) Modulation of synaptic GABAA receptor function by PKA and PKC in adult hippocampal neurons. J. Neurosci., 19: 674–683. Puia, G., Santi, M.R., Vicini, S., Pritchett, D.B., Purdy, R.H., Paul, S.M., Seeburg, P.H. and Costa, E. (1990) Neurosteroids act on recombinant human GABAA receptors. Neuron, 4(5): 759–765. Randle, J.C.R. and Renaud, L.P. (1987) Actions of gammaaminobutyric acid on rat supraoptic nucleus neurosecretory neurones in vitro. J. Physiol., 387: 629–647. Sigel, E. and Baur, R. (1988) Activation of protein kinase C differentially modulates neuronal Na+ , Ca2+ , and gammaaminobutyrate type A channels. Proc. Natl. Acad. Sci. USA, 85: 6192–6196.

Smith, S.S., Gong, Q.H., Hsu, F.C., Markowitz, R.S., FfrenchMullen, J.M. and Li, X. (1998a) GABAA receptor α4 subunit suppression prevents withdrawal properties of an endogenous steroid. Nature, 392: 926–930. Smith, S.S., Gong, Q.H., Li, X., Moran, M.H., Bitran, D., Frye, C.A. and Hsu, F.C. (1998b) Withdrawal from 3α-OH-5αpregnan-20-one using a pseudopregnancy model alters the kinetics of hippocampal GABAA -gated current and increases the GABAA receptor α4 subunit in association with increased anxiety. J. Neurosci., 18: 5275–5284. Stern, J.E. and Armstrong, W.E. (1998) Reorganization of the dendritic trees of oxytocin and vasopressin neurons of the rat supraoptic nucleus during lactation. J. Neurosci., 18: 841–853. Summerlee, A. (1981) Extracellular recordings from oxytocin neurones during the expulsive phase of birth in unanaesthesized rats. J. Physiol., 321: 1–9. Theodosis, D.T. and Poulain, D.A. (1993) Activity-dependent neuronal–glial and synaptic plasticity in the adult mammalian hypothalamus. Neuroscience, 57: 501–535. Twyman, R.E. and McDonald, R.L. (1992) Neurosteroid regulation of GABAA receptor single-channel kinetic properties of mouse spinal cord neurons in culture. J. Physiol., 456: 215– 245. Voisin, D.L., Chapman, C., Poulain, D.A. and Herbison, A.E. (1994) Extracellular GABA concentrations in rat supraoptic nucleus during lactation and following haemodynamic changes: an in vivo microdialysis study. Neuroscience, 63: 547–558. Voisin, D.L., Herbison, A.E. and Poulain, D.A. (1995) Central inhibitory effects of muscimol and bicuculline on the milk ejection reflex in the anaesthetized rat. J. Physiol., 483: 211– 224. Wang, H., Bedford, F.K., Brandon, N.J., Moss, S.J. and Olsen, R.W. (1999) GABA(A)-receptor-associated protein links GABA(A) receptors and the cytoskeleton. Nature, 397: 69–72. Whiting, P.J., McAllister, G., Vassilatis, D., Bonnert, T.P., Heavens, R.P., Smith, D.W., Hewson, L., O’Donnell, R., Rigby, M.R., Sirinathsinghji, D.J., Marshall, G., Thompson, S.A. and Wafford, K.A. (1997) Neuronally restricted RNA splicing regulates the expression of a novel GABAA receptor subunit conferring atypical functional properties. J. Neurosci., 17: 5027– 5037. Wuarin, J.-P. and Dudek, F.E. (1993) Patch-clamp analysis of spontaneous synaptic currents in supraoptic neuroendocrine cells of the rat hypothalamus. J. Neurosci., 13: 2322–2331. Zhu, W.J., Wang, J.F., Krueger, K.E. and Vicini, S. (1994) Delta subunit inhibits neurosteroid modulation of GABAA receptors. J. Neurosci., 16: 6648–6656.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 4

Cholesterol and steroid hormones: modulators of oxytocin receptor function Gerald Gimpl ∗ , Volker Wiegand, Katja Burger and Falk Fahrenholz Institute of Biochemistry, Johannes Gutenberg-University of Mainz, Becherweg 30, D-55099 Mainz, Germany

Abstract: The function and physiological regulation of the oxytocin-receptor system is strongly steroid-dependent. This is, unexpectedly, only partially reflected by the promoter sequences in the oxytocin receptor and favors the idea that posttranscriptional mechanisms may also play a significant role for the physiological regulation of the oxytocin-receptor system. Our data indicate that cholesterol acts as an allosteric modulator of the oxytocin receptor and stabilizes both membrane-associated and solubilized OT receptors in a high-affinity state for agonists and antagonists. Moreover, highaffinity OT receptors are 2-fold enriched in cholesterol-rich plasma membrane domains in HEK293 fibroblasts stably expressing the human OT receptor. Biochemical data suggest a direct and cooperative molecular interaction of cholesterol molecules with OT receptors. To localize the cholesterol interacting domain of the oxytocin receptor the C-terminal part including the last two transmembrane domains have been exchanged by the corresponding sequences of the cholecystokinin type B receptor, which is functionally not dependent on cholesterol. Concerning its ligand-binding behavior this chimeric receptor protein showed the same dependence on cholesterol and its analogues as the wild type oxytocin receptor. From mutagenesis experiments and studies with receptor chimera between the OTR and cholecystokinin type B receptor, we conclude that a major part of the cholesterol interacting domain may be localized in the first part of the oxytocin receptor, possibly in a domain nearby the agonist binding site. Progesterone is considered to be essential to maintain the uterine quiescence. High concentrations of progesterone (>10 μM) attenuate or block the signaling of several GPCRs, including the OT receptor via a fast, reversible and non-genomic pathway. Progesterone is known to inhibit both cholesterol biosynthesis and the intracellular trafficking of cholesterol. We therefore test the hypothesis that progesterone affects the signal transduction and subdomain localization of receptors via its influence on cholesterol trafficking. Since cholesterol-rich subdomains (rafts) are considered to be organization centers for cellular signal transduction, changes of the level or distribution of cholesterol may have profound effects on receptor-mediated signaling in general. Using fluorescence recovery after photobleaching (FRAP) measurements with GFP-tagged oxytocin receptors the influence of steroids on the mobility and distribution of the oxytocin receptor in the plasma membrane was analyzed. Progesterone had no effect on the lateral mobility of the oxytocin receptor, but it led to marked inhibition of cellular motility such as vesicle trafficking and movements of filopodia. Non-genomic effects of progesterone and estradiol with respect to receptor signaling as well as the influence of cholesterol on signal transduction will be discussed in more detail. Keywords: Oxytocin receptor; Cholecystokinin receptor; Cholesterol; Progesterone; Steroids; Non-genomic effects; Signal transduction; Affinity state

∗ Correspondence to: G. Gimpl, Institute of Biochemistry, Johannes Gutenberg-University of Mainz, Becherweg 30, D-55099 Mainz, Germany. Tel.: +49-6131-3923829; Fax: +49-6131-3925348; E-mail: [email protected]

44

Oxytocin receptor and cholesterol Cholesterol and its effect on the affinity state of the oxytocin receptor Oxytocin (OT) receptors require at least two essential components for high-affinity OT-binding, divalent cations such as Mg2+ and cholesterol. This is observed in different systems with both membraneembedded and solubilized receptors. Following solubilization with the detergent Chapso, OT receptors lose characteristic binding properties, the affinity for OT becomes lower and/or additional low-affinity state receptors appear in the extract. Solubilization with Chapso leads to a substantial cholesterol depletion of the soluble extract and we could demonstrate that substitution with cholesterol markedly enhanced the OT binding of soluble OT receptors (Fahrenholz et al., 1995; Klein et al., 1995). This became first evident in reconstitution of soluble OT receptors using liposomes of defined composition. A saturable high-affinity OT binding was obtained only with liposomes that contained a critical amount of cholesterol. Moreover, when OT receptors were expressed in insect cells, which naturally have plasma membranes with low cholesterol content, the receptors are mainly in a low-affinity state (K d > 100 nM). Following addition of cholesterol to the culture medium, a fraction of OT receptors is converted from a low- to a high-affinity state (K d ∼ 1 nM) (Gimpl et al., 1995). The low-affinity state was identified as a physiologically active receptor state and the conversion of the affinity states to each other is to a certain degree reversible. The interaction of cholesterol with OT receptors is of high specificity and is not due to mere changes of membrane fluidity (Gimpl et al., 1997). Furthermore, cholesterol stabilizes both membrane-associated and solubilized OT receptors against thermal denaturation (Gimpl and Fahrenholz, 2000). Taken together, the data provide evidence for a direct and cooperative molecular interaction of cholesterol with OT receptors: Cholesterol acts as an allosteric modulator and stabilizes the receptor in a high-affinity state for agonists and antagonists. In some cell systems, populations of highand low-affinity OT receptors have been observed (Pliska et al., 1986; Crankshaw et al., 1990; Di-Scala and Strosser, 1995; Gimpl and Fahrenholz, 2000).

This could reflect uneven cholesterol distributions within the plasma membrane of these cells. Highaffinity state OT receptors are expected to be preferentially localized in cholesterol-rich subdomains of the plasma membrane. In fact, in HEK293 fibroblasts stably expressing the human OT receptor the high-affinity state receptors were found to be about 2-fold enriched in cholesterol-rich plasma membrane domains (Gimpl and Fahrenholz, 2000). Receptor heterogeneity with respect to affinity states may be much more pronounced in cell systems with abundant cholesterol-rich microdomains such as lipid ‘rafts’ or caveolae structures. Interestingly, Broderick and Broderick (1990) noticed a conspicuous abundance of caveolae in myometrial cells at term. Cholesterol acts as an allosteric modulator of the oxytocin receptor Our previous studies suggest a direct molecular interaction between the oxytocin receptor and cholesterol. Since the ligand-binding activity of the OT receptor declined sharply when the cholesterol amounts were reduced below a critical level (about 50% of the cholesterol content found in untreated membranes), the molecular cholesterol–receptor interaction might be explained by a cooperative mechanism. Hill analysis of cholesterol content versus OT binding suggests that the OT receptor binds several molecules of cholesterol (n = 6) in a positive cooperative manner. Alternatively, at a critical cholesterol level, the distribution of the cholesterol molecules in the vicinity of the receptor could change significantly, e.g. by alterations in the formations of cholesterol dimers or by rearrangement of cholesterol between both leaflets of the membrane bilayer. Divalent metal ions like Mg2+ were found to increase both the OT binding capacity and the affinity state of the OT receptor (Soloff, 1990). This is surprisingly similar to what cholesterol does. Mg2+ increases the potency of OT analogs in stimulating uterine contractions and has been proposed to display its effect on the OT-receptor interaction by influencing positive cooperativity (Pliska and Kohlhauf, 1991). Conclusively, cholesterol and Mg2+ are essential allosteric modulators of the OT receptor and may be involved in the regulation of OT-mediated signaling functions.

45

Do these allosteric modulators play a role for the regulation of OT-related physiological processes? Particularly in reproductive tissues, the steroid and cholesterol concentrations are highly dynamic. Using freeze-fracture cytochemistry with the cholesterolbinding filipin, marked increases in cholesterol have been found in rat uterine epithelial cells at the time of blastocyst implantation (Murphy and Dwarte, 1987). In the human placental syncytiotrophoblast basal membrane, Sen et al. (1998) observed a steady decrease in cholesterol/phospholipid ratio in correlation with an increase in membrane fluidity during placental development. At term, however, the cholesterol to phospholipids ratio in syncytiotrophoblast membranes was found to be increased as compared with the cholesterol/phospholipid ratio in early placentas (Mazzanti et al., 1994). We previously found that cholesterol can modulate receptor function by both changes of the membrane fluidity and direct binding effects, e.g. in case of the oxytocin receptor (Gimpl et al., 1997). Plasma membranes with lowered cholesterol content showed a decreased capacity (Bmax ) of binding sites and/or a decreased affinity (K d ) of ligand-receptor binding. Interestingly, Lopez et al. (1995) reported that pregnancy in humans was associated with increases in both density and affinity of OT receptors. However, to draw further conclusions, more studies are required. The cholesterol binding site(s) of the oxytocin receptor To obtain further information about cholesterol binding site(s) of the OT receptor, a molecular modeling approach was employed. Previous studies revealed that the cholecystokinin type B (CCKB) receptor, a G protein coupled receptor with some homology to the OT receptor, does not depend on cholesterol. We therefore used the CCKB receptor as a negative control in the modeling approach. Based on these assumptions, a candidate cholesterol docking domain has been identified in the OT receptor but was clearly absent in the corresponding region of the CCKB receptor. Within this cholesterol docking site key positions of the human OT receptor such as P197, Y200, W203, M296 and W297 have been suggested as important residues which may be putatively involved in cholesterol binding. This is illustrated in Fig. 1. Ac-

cording to these suggestions the human OT receptor has been mutated at single amino acid residues (e.g. W203) as well as by exchange of C-terminal receptor fragments for corresponding fragments of the CCKB receptor, i.e. producing receptor chimeras. The results showed that even the OTR-CCKBR chimeras which possess the C-terminus of the CCKBR including about half of the third intracellular loop, the two transmembrane regions and the cytosolic part behave like the OT receptor with respect to its cholesterol dependence. This suggests that the cholesterol binding site is not localized in this C-terminal part of the OT receptor. Further mutations resulted in an OT receptor which is functionally inactive due to misfolding or incorrect transport to the plasma membrane. Misfolding or mistranslocation of receptor mutants is a very common observation, e.g. with mutants of the vasopressin V2 receptor leading to diabetes insipidus. On the other hand, it is possible that the OT receptor requires cholesterol to be correctly folded and transported to the plasma membrane. According to this interpretation, the third transmembrane domain and/or the residue W203 form a cholesterol docking site as proposed by the molecular modeling approach. But this has to be substantiated by further mutations. We assume that the major cholesterol binding site which has to be occupied to create a stable and high-affinity receptor conformation is localized at or nearby the oxytocin binding site, i.e. in the N-terminal part of the OT receptor. Taken together, it seems likely that the receptor’s agonist binding site is formed in a stable conformation only when one or more cholesterol molecules including a Mg2+ ion are present. The employment of further technical strategies is necessary to prove this hypothesis. Cholesterol acts as stabilizer of the oxytocin receptor The high stability against thermal inactivation of the OT receptor in the presence of cholesterol is another interesting receptor property. This is illustrated in Fig. 2A. Oxytocin receptors localized in cholesterolpoor membrane domains (i.e. HDM fractions) were significantly less stable than receptors residing in the cholesterol-rich low-density membrane fractions (Cav-LDM). Receptors in HDM domains could be

46 Fig. 1. Schematic model of the human oxytocin receptor indicating amino acid residues which are putatively involved in ligand-binding, cholesterol-binding and associated signal transduction events. The glutamine and lysine residues highly conserved within the vasopressin/oxytocin receptor family may partly define an agonist-binding pocket which is common to all the different subtypes of this receptor family. According to a molecular modeling approach (Fanelli et al., 1999), an oxytocin docking site has been proposed (marked by arrows). In the inactive receptor conformation, the highly conserved arginine (R137) may be constrained in a pocket which is formed by polar residues (indicated by asterisks). Following agonist binding this arginine side chain may be shifted out of the ‘polar pocket’ thereby unmasking a G protein binding site. Receptor domains putatively interacting with oxytocin, peptide oxytocin antagonists and Gαq are marked by lines and dashed boxes as indicated in the symbol legend (Postina et al., 1996; Hoare et al., 1999; Breton et al., 2001). The amino acid residues marked by a circle with asterisk edge have been predicted to form a cholesterol docking domain according to molecular modeling (Gimpl et al., 2000). However, biochemical data do not fully support this prediction (see text).

47

(a)

(b)

Fig. 2. Effect of cholesterol on the stability of the oxytocin receptor against thermal inactivation (A) and proteolytical degradation (B). (A) Caveolin-enriched low-density membrane domains (Cav-LDM) and cholesterol-poor high-density membrane domains (HDM) were incubated at 37°C for indicated periods of time. In a parallel set of assays the HDM membranes were first pretreated for 10 min at 30°C with cholesterol-MβCD (0.3 mM). After that time, the membranes were pelleted and were washed once with binding buffer. Then, the membranes were incubated for various times at 37°C. The high-affinity oxytocin binding activity was determined in a radioligand-binding assay using 5 nM of [3 H]oxytocin. The data represent the remaining binding activity in percent of control binding at time t = 0. The data are given as means ± S.D. of three experiments. (B) Immunoblot analysis of the degradation process of the oxytocin receptor in membranes of HEK293 cells stably expressing the FLAG-tagged oxytocin receptor–GFP conjugate. The membranes were incubated at 37°C for 0 min or 4 h in the absence (lanes 1, 3) or presence (lanes 2, 4) of 0.3 mM Chol-MβCD. At the indicated time, the membranes were processed for SDS–PAGE (reducing conditions) and subsequent immunoblot. For specific receptor detection the anti-FLAG antibody M2 was used. The sharp band at ∼75 kDa corresponds to the deglycosylated receptor.

stabilized against thermal inactivation when additional cholesterol was supplied to the membranes (Fig. 2A). We further explored this receptor property and asked whether cholesterol is also capable to protect the receptor from proteolytic degradation. This is in fact the case as shown in Fig. 2B. The assumption of direct cholesterol–receptor binding is the most straightforward interpretation of the high resistance to proteolytic degradation of the oxytocin receptor in the presence of additional cholesterol. Does the proteolytical degradation of the oxytocin receptor play a physiological role, e.g. in case of normal receptor turnover or signaling offset as observed for the vasopressin 2 receptor (Kojro and Fahrenholz, 1995). If so, then oxytocin receptors residing in a cholesterol-rich microdomain are expected to show prolonged receptor signaling. We are engaged to prove this hypothesis.

Cholesterol and its effect on the mobility of OT receptors It is well established that cholesterol functions as the main regulator of membrane fluidity in eukaryotic cells. Alterations in the cholesterol content should therefore induce changes of the receptor mobility if the receptor floats passively in the lipid bilayer. Fluorescence recovery after photobleaching (FRAP) experiments were performed to address this question in more detail. After a brief 1-s spot bleaching of the GFP-tagged oxytocin receptor stably expressed in untreated HEK293 cells, the recovery normally occurs with a t1/2 of about 20 s with a mobile fraction of about 80% of the receptors (Table 1). This is about 10 times faster than the recovery process of oxytocin receptors localized in vesicles inside the cell. Cholesterol depletion below a threshold concentration (<50% of control) using cyclodextrins led to an increased membrane fluidity and slightly accelerated this recovery process for a small amount of oxytocin receptors (∼27%), whereas the major-

48

Oxytocin receptor and progesterone

TABLE 1 FRAP measurements of the lateral mobility of the oxytocin receptor Treatment

Mobile fraction (%)

t1/2 (s)

None − Cholesterol (CD 40 min) + Cholesterol (Chol-CD 20 min) + Cholesterol–cholesterol (Chol-CD 20 min, CD 20 min) + Progesterone (5 min, 100 μM)

80.5 ± 5.6 26.6 ± 7.1 79.1 ± 6.2 67.0 ± 8.1

19.3 ± 4.8 14.3 ± 5.8 26.7 ± 6.3 37.3 ± 8.5

74.0 ± 5.2

20.3 ± 4.1

The values are means ± SD (n = 12 cells).

ity of receptors became immobile. Thus, following cholesterol depletion only a population of OT receptors appears to reside in a more fluid environment. The immobilization of the majority of OT receptors may be caused by phase separation within the lipid bilayer and/or unknown alterations of the cytoskeleton. In contrast, cholesterol enrichment increased the t1/2 of FRAP and had no significant effect on the fraction of mobile receptors. This behavior is not unexpected and reflects more or less the changes of fluidity which are controlled by cholesterol. However, if the cholesterol-enriched membranes are again depleted of cholesterol, the t1/2 of FRAP of oxytocin receptors unexpectedly further increases (Table 1). This treatment may mimick to a certain extent the alterations of cholesterol in a highly dynamic membrane and/or under some physiological conditions. Thus, under these experimental conditions, the mobility of the oxytocin receptor does not reflect the changes in membrane fluidity and this may have yet unpredictable consequences on receptor signaling. We found that acute administration of progesterone (up to 100 μM) had no significant effect neither on the fraction of mobile receptors nor on FRAP time. However, the cellular motility was markedly inhibited following short (1–5 min) progesterone treatment of 293 fibroblasts (unpublished observation). The molecular mechanisms which underlie these rapid progesterone-induced effects could be the same as those responsible for the inhibitory effect on receptor signaling (see below).

The steroid hormone progesterone is essential to maintain uterine quiescence. Progesterone withdrawal in combination with increasing levels of estrogen leads to increased expression of several genes involved in the onset of parturition. Moreover, posttranscriptional actions of progesterone are considered to play an important role for promoting uterine smooth muscle relaxation. Grazzini et al. (1998) previously claimed that progesterone could act as a direct negative modulator of the OT receptor at concentrations in the nanomolar range and thus offered a plausible mechanism of how progesterone could contribute to uterine quiescence. In contrast, we observed that much higher concentrations of progesterone (>10 μM) are required to attenuate or block the signaling of the OT receptor (Burger et al., 1999). Using these high concentrations, however, progesterone is capable to attenuate or block the signaling function of all G protein coupled receptors that we have tested (Burger et al., 2000). The progesterone effects occurred within minutes, were reversible and could not be blocked by a protein synthesis inhibitor. Overall, the action of progesterone was found to be more cell type-specific than receptor-specific. Obviously, progesterone displays an inhibitory effect on receptor signaling via a more general mechanism. The question arises whether under physiological conditions those high micromolar concentrations of progesterone are found in cells in vivo. The progesterone levels in plasma or in nonsteroidogenic tissues are normally much lower. In steroidogenic tissues, however, very high amounts of progesterone have been measured. Near term, the human placenta secretes upward of 300 mg of progesterone daily. The progesterone content of this organ was shown to be 7 μg/g wet tissue (Simpson and Burkhart, 1980). In human corpus luteum, progesterone concentrations reached peak levels of about 25 μg/g tissue shortly after ovulation and in the early luteal phase (Swanston et al., 1977). Thus, in steroidogenic cells as well as in their environment, progesterone might be present at high doses which are able to attenuate the signaling of many receptors including the oxytocin receptor. What are the molecular mechanisms underlying this non-genomic progesterone action which is pre-

49

sumably operative in all cell types? Progesterone is long known to profoundly affect the intracellular trafficking of cholesterol. For example, cell biologists routinely use this steroid hormone and certain hydrophobic amines to induce a Niemann–Pick-type C-like distribution of cholesterol in cells. In this respect, progesterone is the most potent of natural steroids, with maximal effects occurring at 30 μM (Butler et al., 1992). Although it generally takes several hours of incubation with progesterone to induce this pathophysiological cholesterol distribution in cells, progesterone may start to disturb the cholesterol trafficking machinery within minutes. Presently, there are no adequate methods to detect these early changes of cholesterol trafficking. One candidate progesterone-binding protein is the multidrug resistance P-glycoprotein (Qian and Beck, 1990). This protein does not only play a role in detoxification, but appears also be involved in intracellular cholesterol transport. It has been demonstrated that progesterone inhibits both the cholesterol esterification and the transport of cholesterol to and from the plasma membrane. In particular, progesterone reduces the cholesterol pool residing in caveolae (Smart et al., 1996). Paradoxically, at the same time progesterone stimulates the activity of HMG-CoA reductase, the key enzyme of de novo cholesterol biosynthesis. Hence, cholesterol precursors like lanosterol begin to enrich in the membranes of the cell. On the basis of these progesterone-induced events the following scenario (see model in Fig. 3) may be operative. As mentioned above, the OT receptor needs a cholesterol-rich microenvironment to become stabilized in its high-affinity state. Since the cholesterol precursors, particularly lanosterol, are completely inactive to support the OT receptor in its high-affinity state (Gimpl et al., 1997), the responsiveness of the OT system may not be fully operative during the continuous presence of high progesterone concentrations. According to this scenario, progesterone withdrawal would restore the cholesterol transport so that the highly enriched amounts of cholesterol precursors would now become rapidly converted to cholesterol. This would lead to a sudden rise of cholesterol and should push the responsiveness to OT since low-affinity OT receptors could now be converted into their high-affinity state. According to this postulated mechanism, progesterone could af-

fect the signaling of all those receptors which are functionally dependent on cholesterol. Nevertheless, these events could only occur in cells which ‘see’ micromolar doses of progesterone, e.g. in steroidogenic tissues and their environment. Most likely, progesterone acts in these tissues via both genomic and non-genomic pathways together with other steroids to control receptor activity. General aspects of steroids and signaling Cholesterol in signal transduction Membrane cholesterol is considered to play a crucial role for signal transduction processes in multiple ways, e.g. (1) by its control of the fluidity state and biophysical properties of the lipid bilayer, (2) by formation of lipid rafts and caveolae, (3) in the hedgehog signaling pathway, (4) by activation of proteins via sterol-sensing domains, (5) as direct modulators of membrane proteins, e.g. GPCRs and ion channels. Caveolae are small invaginations of the plasma membrane and are considered to be a specialized form of rafts that contain the cholesterol-binding structural protein caveolin. Several membrane receptors such as platelet-derived growth factor, receptors for epidermal growth factor, adrenalin, bradykinin, endothelin, cholecystokinin, and oxytocin are enriched in caveolae or caveolae-like domains. In caveolae, receptors seem to be functionally connected to molecules of the signaling cascades, e.g. G protein α subunits, protein kinase Cα, mitogen-activated protein kinase, adenylyl cyclase and phosphoinositides (reviewed in Okamoto et al., 1998). Isshiki et al. (1998) demonstrated that in endothelial cells, calcium waves upon intracellular calcium release originate from caveolae-rich cell edges. Caveolin and caveolae are involved in intracellular cholesterol transport and cholesterol efflux to extracellular carriers (Bist et al., 1997; Liu et al., 1999). Thus, accumulating evidence suggests that dynamic cholesterol-dependent regulation of specific signal transduction pathways occurs within caveolae. Unlike caveolae, which are abundantly present in many but not in all cells, sphingolipid-cholesterol subdomains, often designated as rafts, may be present in every cell although there is some controversy on

50

Fig. 3. Schematic model of non-genomic inhibitory effects of progesterone. Progesterone inhibits both the signal transduction of Gq -coupled receptors (as shown here for the oxytocin receptor) and the intracellular trafficking of cholesterol. Principally, eukaryotic cells can obtain the required cholesterol (Chol, grey ellipses) by two sources: endogenously by de novo synthesis of cholesterol and exogenously by uptake of cholesteryl-ester (CE)-rich LDL (low-density lipoprotein) particles via receptor-mediated (R) endocytosis. De novo synthesized cholesterol first arrives at cholesterol-rich domains in the plasma membrane (caveolae and/or ‘lipid-rafts’), which may function as cholesterol-‘sorting centers’ within the plasma membrane. Progesterone blocks several intracellular transport pathways of cholesterol (black bars) except for the LDL receptor-mediated uptake of cholesterol. Moreover, cholesterol esterification does not occur in the presence of progesterone, presumably due to the lack of cholesterol substrate for acyl-CoA : cholesterol acetyltransferase (ACAT). As a consequence, unesterified cholesterol accumulates in lysosomes (or late endosomes) and lysosome-like compartments (designated as ‘lamellar bodies’). The key enzyme for the cholesterol de novo synthesis, 3-hydroxy-3-methylglutaryl-CoA reductase (HMG-CoA Red), is stimulated in the presence of progesterone. But the cholesterol biosynthesis stops at the level of precursors. Overall, progesterone induces a state of cholesterol auxotrophy. However, following progesterone withdrawal, the accumulated precursors will be rapidly converted to cholesterol. Thus, cells will become overloaded with cholesterol for a certain period of time after which the cholesterol homeostasis will be reestablished. These reversible progesterone-induced changes of the cholesterol trafficking could have a strong influence on signal transduction processes, particularly in case of the OT receptor (OTRH and OTRL , high-affinity and low-affinity OT receptor). Adapted from Gimpl and Fahrenholz (2001).

their existence in vivo. They could provide platforms for efficient initiation of signaling cascades by bringing together receptors with their second messenger effectors (Incardona and Eaton, 2000; Simons and Toomre, 2000). As a component of the recently identified hedgehog pathway which is of fundamental importance, e.g. for the patterning of the nervous system, cholesterol plays a central role for ligand biogenesis and for signal transduction in receiving cells. Hedgehog proteins are morphogens posttranslationally modified with a covalently attached cholesterol molecule which anchors the protein in the outer leaflet of the

membrane bilayer and mediates its association with lipid rafts. The hedgehog receptor complex consists of Patched, a polytopic transmembrane protein with structural similarity (sterol sensing domain) to the Niemann–Pick C1 (NPC1) protein and a second protein, termed Smoothened, a heptahelical organized receptor like a GPCR. Patched acts as the ligand-binding subunit and regulates the activity of Smoothened, which functions as the signaling subunit. In the absence of ligand, Patched inhibits a tonic activity of Smoothened. The inhibition is released upon hedgehog binding thus activating a signal transduction cascade. Conclusively, in the hedgehog

51

signaling complex cholesterol participates in signal generation by controlling the expression and patterning of the hedgehog ligand, as well as in signal reception due to the relationship of Patched to NPC1 and other proteins with sterol sensing domains. Reduced hedgehog signaling dramatically manifests in holoprosencephaly, a severe brain malformation. In addition, other environmental and genetic causes of holoprosencephaly are related to sterols or sterol metabolism (Incardona and Eaton, 2000). Changes of the cholesterol level influence receptor-mediated signal transduction at several levels. Cholesterol depletion led to constitutively activation of the p42/44 MAP kinase cascade, and epidermal growth factor caused hyperactivation of ERK in cholesterol-depleted cells (Furuchi and Anderson, 1998; Galbiati et al., 1998). Similarly, Visconti et al. (1999) found that removal of cholesterol is able to initiate transmembrane signaling leading to subsequent stimulation of protein tyrosine phosphorylation and sperm capacitation. At the receptor level, there is evidence for a direct cholesterol interaction with a number of receptors such as the oxytocin receptor, galanin receptor (Pang et al., 1999), opioid receptor (Lagane et al., 2000), nicotinic acetylcholin receptor (Rankin et al., 1997), and rhodopsin (Albert et al., 1996). Dependent on the receptor type, cholesterol is able to modulate the affinity state, the binding capacity, the ligand-binding specificity, and/or the conformational stability of the receptor. Receptors which are not directly interacting with cholesterol may be affected in their binding properties by cholesterols’ effect on the membrane fluidity as shown for the cholecystokinin B receptor (Gimpl et al., 1997). Our recent analysis of ligand-induced calcium responses at different cholesterol contents of the cells revealed that many GPCRs require cholesterol for optimal signal transduction (Burger et al., 2000). Since both clathrin coat- and caveolae-mediated endocytosis require cholesterol, signal transduction is also expected to be cholesterol-dependent at the level of receptor internalization and downregulation (Subtil et al., 1999). Steroid hormones and signaling There is a growing number of examples of rapid, nongenomic signaling by steroids such as aldosteron-in-

duced increases in intracellular calcium in vascular smooth muscle cells, vitamin D-induced increases in intracellular calcium in osteosarcoma cells (Caffrey and Farach-Carson, 1989), and progesteronemediated maturation of amphibian and fish oocytes. 17-β-Estradiol has been found to induce Ca2+ dependent translocation of eNOS (Goetz et al., 1999) and might contribute to smooth muscle relaxation via direct binding to the β-subunit of the Maxi-K channel, which confers the channel with a higher Ca2+ sensitivity (Valverde et al., 1999). Valera et al. (1992) observed an inhibitory effect of progesterone on the neuronal nicotinic acetylcholine receptor. Administration of progesterone has been demonstrated to inhibit the transmembrane Ca2+ entry, to promote Ca2+ release from intracellular stores and membrane hyperpolarization subsequently activating K+ channels (Mironneau et al., 1981). As mentioned above, progesterone and closely related analogues inhibited the Ca2+ mobilization induced by stimulation of many GPCRs (Burger et al., 1999, 2000) in CHO and HEK293 cells. Oxytocin- and endothelin-1-induced increases in [Ca2+ ]i were also attenuated in progesterone-treated myometrial cells (Fomin et al., 1999). Table 2 gives a list of some selected presumably non-genomic effects induced by the steroid hormones progesterone and 17-βestradiol. A number of steroids act as important modulators in the central nervous system where several of them are de novo synthesized as neurosteroids. Modulatory effects of neurosteroids have been reported on the neuronal nicotinic acetylcholine receptor (see above), glycine receptors (Wu et al., 1990), nonNMDA glutamate receptors (Wu et al., 1991), the GABAA receptor (Majewska et al., 1986), and 5HT3 serotonin receptors (Wetzel et al., 1998). Thus, neurosteroids seem to function as general modulators of ligand-gated ion channels. On the other hand, there is still controversy about the specificity of these effects (Rupprecht and Holsboer, 1999). One crucial question is whether these various nongenomic actions of steroids are caused by interaction with the corresponding proteins or, alternatively, are due to the interference with a common and basic cellular mechanism. Administration of steroids may for example disturb the Ca2+ metabolism or directly interact with components of the cytoskele-

52 TABLE 2 Multiple actions of progesterone and estradiol presumably involving nongenomic mechanisms Steroid

Effect

References

Progesterone

Maturation of oocytes

Progesterone Progesterone Progesterone Progesterone Progesterone Progesterone Progesterone Progesterone Estradiol

Acrosome reaction of sperm Inhibitory modulation of nAChR Inhibitory effect on GPCR signaling Ligand of sigma receptor Cholesterol trafficking defects Cholesterol synthesis inhibitor Production of meiosis-activating sterols Counteracting of pregnenolone-induced stimulation of microtubuli assembly Neuropsychopharmacological actions (anesthetic, sedative, anxiolytic) Activation/translocation of eNOS

Estradiol Estradiol Estradiol Estradiol

Activating of MAP kinase Binding to Maxi-K channel β subunit Modulation fo GPCR signaling Neuroprotection

Blondeau and Baulieu (1984); Bayaa et al. (2000); Tian et al. (2000) Blackmore (1999) a Valera et al. (1992) Burger et al. (1999) Su et al. (1988) Butler et al. (1992) Metherall et al. (1996) Lindenthal et al. (2001) Murakami et al. (2000) Rupprecht and Holsboer (1999) a Goetz et al. (1999); Hisamoto et al. (2001) Russell et al. (2000) Valverde et al. (1999) Kelly and Wagner (1999) a Moosmann and Behl (1999)

a

Review articles summarizing many original findings to that topic.

ton. In this respect, pregnenolone and progesterone have been shown to affect microtubuli assembly and both steroids were identified as high-affinity ligands for microtubuli-associated protein 2 in neuronal hippocampal cells (Murakami et al., 2000). Recently, a connection between neurosteroids and the mysterious sigma receptor has been proposed. A number of findings suggest that sigma receptors mediate or at least influence the effects of systemically administered neurosteroids. Progesterone is to date the only natural ligand for the sigma1 receptor which surprisingly, exhibits a high homology to a fungal sterol isomerase (Hanner et al., 1996). The physiological significance of this homology is yet unclear. Perhaps sigma receptors are enzymatically involved in the synthesis of certain neurosteroid modulators. While the classical steroid hormones are produced in specialized steroidogenic cells starting from cholesterol, another group of steroids with putative reproductive functions have been identified as C29 intermediates of the de novo cholesterol synthesis pathway at the postsqualene level. Due to their effects these sterols are summarized as meiosis-activating sterols. Interestingly, Lindenthal et al. (2001) recently observed that administration of progesterone to a hepatoma cell line and rat testes

led to an accumulation of these meiosis-activating sterols via inhibition of certain postsqualene steps of cholesterol biosynthesis. These findings underline the putative importance of cholesterol precursors in reproductive tissue and a new possible role for progestins in this process. These few examples demonstrate that much effort will be necessary in the future to clarify the biochemical mechanisms of the multiple functions of steroids. Abbreviations Cav-LDM

caveolin-enriched low density membranes CCKBR cholecystokinin type B receptor eNOS endothelial NO synthase FRAP fluorescence recovery after photobleaching GPCR G protein coupled receptor HDM high-density membranes MAP kinase mitogen-activated protein kinase nAChR nicotinic acetylcholin receptor OT oxytocin OTR oxytocin receptor

53

Acknowledgements This work was supported by grants from the Deutsche Forschungsgemeinschaft (Fa 122-5 and Gi 201/23,2-4). We thank Christa Wolpert for technical assistance. References Albert, A.D., Young, J.E. and Yeagle, P.L. (1996) Rhodopsin– cholesterol interactions in bovine rod outer segment disk membranes. Biochim. Biophys. Acta, 1285: 47–55. Bayaa, M., Booth, R.A., Sheng, Y. and Liu, X.J. (2000) The classical progesterone receptor mediates Xenopus oocyte maturation through a nongenomic mechanism. Proc. Natl. Acad. Sci. USA, 97: 12607–12612. Bist, A., Fielding, P.E. and Fielding, C.J. (1997) Two sterol regulatory element-like sequences mediate up-regulation of caveolin gene transcription in response to low density lipoprotein free cholesterol. Proc. Natl. Acad. Sci. USA, 94: 10693– 10698. Blackmore, P.F. (1999) Extragenomic actions of progesterone in human sperm and progesterone metabolites in human platelets. Steroids, 64: 149–156. Blondeau, J.P. and Baulieu, E.E. (1984) Progesterone receptor characterized by photoaffinity labelling in the plasma membrane of Xenopus laevis oocytes. Biochem. J., 219: 785–792. Breton, C., Chellil, H., Kabbaj-Benmansour, M., Carnazzi, E., Seyer, R., Phalipou, S., Morin, D., Durroux, T., Zingg, H., Barberis, C. and Mouillac, B. (2001) Direct identification of human oxytocin receptor-binding domains using a photoactivatable cyclic peptide antagonist. Comparison with the human V1a vasopressin receptor. J. Biol. Chem., 276: 26931–26941. Broderick, R. and Broderick, K.A. (1990) Uterine function. In: Molecular and Cellular Aspects. Plenum Press, New York, NY, pp. 1–33. Burger, K., Fahrenholz, F. and Gimpl, G. (1999) Non-genomic effects of progesterone on the signaling function of G proteincoupled receptors. FEBS Lett., 464: 25–29. Burger, K., Gimpl, G. and Fahrenholz, F. (2000) Regulation of receptor function by cholesterol. Cell. Mol. Life Sci., 57: 1577–1592. Butler, J.D., Blanchette-Mackie, J., Goldin, E., O’Neill, R.R., Carstea, G., Roff, C.F., Patterson, M.C., Patel, S., Comly, M.E., Cooney, A., Vanier, M.T., Brady, R.O. and Pentchev, P.G. (1992) Progesterone blocks cholesterol translocation from lysosomes. J. Biol. Chem., 267: 23797–23805. Caffrey, J.M. and Farach-Carson, M.C. (1989) Vitamin D3 metabolites modulate dihydropyridine-sensitive calcium currents in clonal rat osteosarcoma cells. J. Biol. Chem., 264: 20265–20274. Crankshaw, D., Gaspar, V. and Pliska, V. (1990) Multiple [3 H]oxytocin binding sites in rat myometrial plasma membranes. J. Recept. Res., 10: 269–285. Di-Scala, G.D. and Strosser, M.T. (1995) Downregulation of the

oxytocin receptor on cultured astroglial cells. Am. J. Physiol., 268: C413–C418. Fahrenholz, F., Klein, U. and Gimpl, G. (1995) Conversion of the myometrial oxytocin receptor from low to high affinity state by cholesterol. Adv. Exp. Med. Biol., 395: 311–319. Fanelli, F., Barbier, P., Zanchetta, D., De Benedetti, P.G. and Chini, B. (1999) Activation mechanism of human oxytocin receptor: a combined study of experimental and computersimulated mutagenesis. Mol. Pharmacol., 56: 214–225. Fomin, V.P., Cox, B.E. and Word, R.A. (1999) Effect of progesterone on intracellular Ca2+ homeostasis in human myometrial smooth muscle cells. Am. J. Physiol., 276: C379–C385. Furuchi, T. and Anderson, R.G. (1998) Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-related kinase (ERK). J. Biol. Chem., 273: 21099–21104. Galbiati, F., Volonte, D., Engelman, J.A., Watanabe, G., Burk, R., Pestell, R.G. and Lisanti, M.P. (1998) Targeted downregulation of caveolin-1 is sufficient to drive cell transformation and hyperactivate the p42/44 MAP kinase cascade. EMBO J., 17: 6633–6648. Gimpl, G. and Fahrenholz, F. (2000) Human oxytocin receptors in cholesterol-rich vs. cholesterol-poor microdomains of the plasma membrane. Eur. J. Biochem., 267: 2483–2497. Gimpl, G. and Fahrenholz, F. (2001) The oxytocin receptor system: structure, function, and regulation. Physiol. Rev., 81: 629–683. Gimpl, G., Klein, U., Reilaender, H. and Fahrenholz, F. (1995) Expression of the human oxytocin receptor in baculovirusinfected insect cells: high-affinity binding is induced by a cholesterol–cyclodextrin complex. Biochemistry, 34: 13794– 13801. Gimpl, G., Burger, K. and Fahrenholz, F. (1997) Cholesterol as modulator of receptor function. Biochemistry, 36: 10959– 10974. Gimpl, G., Burger, K., Politowska, E., Ciarkowski, J. and Fahrenholz, F. (2000) Oxytocin receptors and cholesterol: interaction and regulation. Exp. Physiol., 85: 41S–49S. Goetz, R.M., Thatte, H.S., Prabhakar, P., Cho, M.R., Michel, T. and Golan, D.E. (1999) Estradiol induces the calciumdependent translocation of endothelial nitric oxide synthase. Proc. Natl. Acad. Sci. USA, 96: 2788–2793. Grazzini, E., Guillon, G., Mouillac, B. and Zingg, H.H. (1998) Inhibition of oxytocin receptor function by direct binding of progesterone. Nature, 392: 509–512. Hanner, M., Moebius, F.F., Flandorfer, A., Knaus, H.G., Striessnig, J., Kempner, E. and Glossmann, H. (1996) Purification, molecular cloning, and expression of the mammalian sigma1binding site. Proc. Natl. Acad. Sci. USA, 93: 8072–8077. Hisamoto, K., Ohmichi, M., Kurachi, H., Hayakawa, J., Kanda, Y., Nishio, Y., Adachi, K., Tasaka, K., Miyoshi, E., Fujiwara, N., Taniguchi, N. and Murata, Y. (2001) Estrogen induces the Akt-dependent activation of endothelial nitric-oxide synthase in vascular endothelial cells. J. Biol. Chem., 276: 3459–3467. Hoare, S., Copland, J.A., Strakova, Z., Ives, K., Jeng, Y.J., Hellmich, M.R. and Soloff, M.S. (1999) The proximal portion of the COOH terminus of the oxytocin receptor is required for coupling to g(q), but not g(i). Independent mechanisms for el-

54 evating intracellular calcium concentrations from intracellular stores. J. Biol. Chem., 274: 28682–28689. Incardona, J.P. and Eaton, S. (2000) Cholesterol in signal transduction. Curr. Opin. Cell. Biol., 12: 193–203. Isshiki, M., Ando, J., Korenaga, R., Kogo, H., Fujimoto, T., Fujita, T. and Kamiya, A. (1998) Endothelial Ca2+ waves preferentially originate at specific loci in caveolin-rich cell edges. Proc. Natl. Acad. Sci. USA, 95: 5009–5014. Kelly, M.J. and Wagner, E.J. (1999) Estrogen modulation of G-protein-coupled receptors. Trends Endocrinol. Metab., 10: 369–374. Klein, U., Gimpl, G. and Fahrenholz, F. (1995) Alteration of the myometrial plasma membrane cholesterol content with betacyclodextrin modulates the binding affinity of the oxytocin receptor. Biochemistry, 34: 13784–13793. Kojro, E. and Fahrenholz, F. (1995) Ligand-induced cleavage of the V2 vasopressin receptor by a plasma membrane metalloproteinase. J. Biol. Chem., 270: 6476–6481. Lagane, B., Gaibelet, G., Meiloc, E., Masson, J.-M., Cezanne, L. and Lopez, A. (2000) Role of sterols in modulating the human mu-opioid receptor function in Saccharomyces cerevisiae. J. Biol. Chem., 275: 33197–33200. Lindenthal, B., Holleran, A.L., Aldaghlas, T.A., Ruan, B., Schroepfer Jr., G.J., Wilson, W.K. and Kelleher, J.K. (2001) Progestins block cholesterol synthesis to produce meiosisactivating sterols. FASEB J., 15: 775–784. Liu, P., Li, W.P., Machleidt, T. and Anderson, R.G. (1999) Identification of caveolin-1 in lipoprotein particles secreted by exocrine cells. Nat. Cell. Biol., 1: 369–375. Lopez, B.A., Rivera, J., Europe, F.G., Phaneuf, S. and Asboth, G. (1995) Parturition: activation of stimulatory pathways or loss of uterine quiescence?. Adv. Exp. Med. Biol., 395: 435–451. Majewska, M.D., Harrison, N.L., Schwartz, R.D., Barker, J.L. and Paul, S.M. (1986) Steroid hormone metabolites are barbiturate-like modulators of the GABA receptor. Science, 232: 1004–1007. Mazzanti, L., Staffolani, R., Cester, N., Romanini, C., Pugnaloni, A., Belmonte, M.M., Salvolini, E., Brunelli, M.A. and Biagini, G. (1994) A biochemical–morphological study on microvillus plasma membrane development. Biochim. Biophys. Acta, 1192: 101–106. Metherall, J.E., Waugh, K. and Li, H. (1996) Progesterone inhibits cholesterol biosynthesis in cultured cells. Accumulation of cholesterol precursors. J. Biol. Chem., 271: 2627–2633. Mironneau, J., Savineau, J.P. and Mironneau, C. (1981) Fast outward current controlling electrical activity in rat uterine smooth muscle during gestation. J. Physiol. (Paris), 77: 851– 859. Moosmann, B. and Behl, C. (1999) The antioxidant neuroprotective effects of estrogens and phenolic compounds are independent from their estrogenic properties. Proc. Natl. Acad. Sci. USA, 96: 8867–8872. Murakami, K., Fellous, A., Baulieu, E.E. and Robel, P. (2000) Pregnenolone binds to microtubule-associated protein 2 and stimulates microtubule assembly. Proc. Natl. Acad. Sci. USA, 97: 3579–3584. Murphy, C.R. and Dwarte, D.M. (1987) Increase in cholesterol in

the apical plasma membrane of uterine epithelial cells during early pregnancy in the rat. Acta Anat. Basel, 128: 76–79. Okamoto, T., Schlegel, A., Scherer, P.E. and Lisanti, M.P. (1998) Caveolins, a family of scaffolding proteins for organizing ‘preassembled signaling complexes’ at the plasma membrane. J. Biol. Chem., 273: 5419–5422. Pang, L., Graziano, M. and Wang, S. (1999) Membrane cholesterol modulates galanin–GalR2 interaction. Biochemistry, 38: 12003–12011. Pliska, V. and Kohlhauf, A.H. (1991) Effect of Mg2+ on the binding of oxytocin to sheep myometrial cells. Biochem. J., 277: 97–101. Pliska, V., Heiniger, J., Muller, L.A., Pliska, P.B. and Ekberg, G. (1986) Binding of oxytocin to uterine cells in vitro. Occurrence of several binding site populations and reidentification of oxytocin receptors. J. Biol. Chem., 261: 16984–16989. Postina, R., Kojro, E. and Fahrenholz, F. (1996) Separate agonist and peptide antagonist binding sites of the oxytocin receptor defined by their transfer into the V2 vasopressin receptor. J. Biol. Chem., 271: 31593–31601. Qian, X.D. and Beck, W.T. (1990) Progesterone photoaffinity labels P-glycoprotein in multidrug-resistant human leukemic lymphoblasts. J. Biol. Chem., 265: 18753–18756. Rankin, S.E., Addona, G.H., Kloczewiak, M.A., Bugge, B. and Miller, K.W. (1997) The cholesterol dependence of activation and fast desensitization of the nicotinic acetylcholine receptor. Biophys. J., 73: 2446–2455. Rupprecht, R. and Holsboer, F. (1999) Neuroactive steroids: mechanisms of action and neuropsychopharmacological perspectives. Trends Neurosci., 22: 410–416. Russell, K.S., Haynes, M.P., Sinha, D., Clerisme, E. and Bender, J.R. (2000) Human vascular endothelial cells contain membrane binding sites for estradiol, which mediate rapid intracellular signaling. Proc. Natl. Acad. Sci. USA, 97: 5930– 5935. Sen, A., Ghosh, P.K. and Mukherjea, M. (1998) Changes in lipid composition and fluidity of human placental basal membrane and modulation of bilayer protein functions with progress of gestation. Mol. Cell. Biochem., 187: 183–190. Simons, K. and Toomre, D. (2000) Lipid rafts and signal transduction. Nat. Rev. Mol. Cell. Biol., 1: 31–39. Simpson, E.R. and Burkhart, M.F. (1980) Acyl CoA : cholesterol acyl transferase activity in human placental microsomes: inhibition by progesterone. Arch. Biochem. Biophys., 200: 79– 85. Smart, E.J., Ying, Y., Donzell, W.C. and Anderson, R.G. (1996) A role for caveolin in transport of cholesterol from endoplasmic reticulum to plasma membrane. J. Biol. Chem., 271: 29427–29435. Soloff, M.S. (1990) Uterine function. In: Molecular and Cellular Aspects. Plenum Press, New York, NY, pp. 373–392. Su, T.P., London, E.D. and Jaffe, J.H. (1988) Steroid binding at sigma receptors suggests a link between endocrine, nervous, and immune systems. Science, 240: 219–221. Subtil, A., Gaidarov, I., Kobylarz, K., Lampson, M.A., Keen, J.H. and McGraw, T.E. (1999) Acute cholesterol depletion inhibits

55 clathrin-coated pit budding. Proc. Natl. Acad. Sci. USA, 96: 6775–6780. Swanston, I.A., McNatty, K.P. and Baird, D.T. (1977) Concentration of prostaglandin F2alpha and steroids in the human corpus luteum. J. Endocrinol., 73: 115–122. Tian, J., Kim, S., Heilig, E. and Ruderman, J.V. (2000) Identification of XPR-1, a progesterone receptor required for Xenopus oocyte activation. Proc. Natl. Acad. Sci. USA, 97: 14358– 14363. Valera, S., Ballivet, M. and Bertrand, D. (1992) Progesterone modulates a neuronal nicotinic acetylcholine receptor. Proc. Natl. Acad. Sci. USA, 89: 9949–9953. Valverde, M.A., Rojas, P., Amigo, J., Cosmelli, D., Orio, P., Bahamonde, M.I., Mann, G.E., Vergara, C. and Latorre, R. (1999) Acute activation of Maxi-K channels (hSlo) by estradiol binding to the beta subunit. Science, 285: 1929–1931.

Visconti, P.E., Galantino-Homer, H., Ning, X., Moore, G.D., Valenzuela, J.P., Jorgez, C.J., Alvarez, J.G. and Kopf, G.S. (1999) Cholesterol efflux-mediated signal transduction in mammalian sperm. J. Biol. Chem., 274: 3235–3242. Wetzel, C.H., Hermann, B., Behl, C., Pestel, E., Rammes, G., Zieglgansberger, W., Holsboer, F. and Rupprecht, R. (1998) Functional antagonism of gonadal steroids at the 5hydroxytryptamine type 3 receptor. Mol. Endocrinol., 12: 1441–1451. Wu, F.S., Gibbs, T.T. and Farb, D.H. (1990) Inverse modulation of gamma-aminobutyric acid- and glycine-induced currents by progesterone. Mol. Pharmacol., 37: 597–602. Wu, F.S., Gibbs, T.T. and Farb, D.H. (1991) Pregnenolone sulfate: a positive allosteric modulator at the N-methyl- Daspartate receptor. Mol. Pharmacol., 40: 333–336.

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 5

Central vasopressin systems and steroid hormones Andries Kalsbeek ∗ , Inge F. Palm and Ruud M. Buijs Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands

Keywords: Circadian; Corticosterone; Estrogen; Luteinizing hormone; Paraventricular nucleus; Supachiasmatic nucleus; Sexual dimorphism; Vasopressin

Vasopressin systems in the brain The physiological effects of vasopressin (VP) as a peripheral hormone were first reported as early as 1895 (Oliver and Schäfer, 1895). However, it was almost 60 years before the probable source of VP released in the general circulation was described by Bargmann and Scharrer (1951). They proposed that VP was released from neurons residing in the supraoptic (SON) and paraventricular nucleus (PVN) that projected to the posterior pituitary. Only some 15 years after the localization of VP in the brain, a possible central function for VP was described for the first time by De Wied (1965). At first the central effects of VP were attributed to a ‘feedback’ action of VP released as a hormone from the neurohypophysis. The first immunocytochemical studies using VP antibodies were mainly intended to investigate in more detail the magnocellular VP neurons in the SON and PVN (Swaab and Pool, 1975; Vandesande and Dierickx, 1975; Vandesande et al., 1975), but in these initial studies VP was also detected in the suprachiasmatic nucleus (SCN). Subsequent, more sensitive immunocytochemical methods revealed an extensive distribution of VP-containing fibers outside the hypothalamus (Buijs, 1978; Buijs et al., ∗ Correspondence to: A. Kalsbeek, Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands. Tel.: +31-20-566-5500; Fax: +31-20-696-1006; E-mail: [email protected]

1978; Buijs and Swaab, 1979) and a neurotransmitter role for VP could be hypothesized. Initially the different intra- and extrahypothalamic VP projections were all thought to originate from the PVN and SON neurons and the SCN. However, using specific brain lesions or colchicine pretreatment to block axonal transport a considerable part of the VP innervation was found to be derived from additional VP-producing neurons located in the bed nucleus of the stria terminalis (BNST) and the medial amygdala (AME) (Hoorneman and Buijs, 1982; De Vries and Buijs, 1983; Van Leeuwen and Caffé, 1983). Consequently, using a wide variety of different techniques, the largest part of the VPergic innervation in the brain could be traced back to four major VP systems: (1) the sexually dimorphic system, with VP neurons localized in the BNST and the AME; (2) the autonomic and (3) endocrine system originating from parvocellular VP neurons located in the PVN; and (4) the biological clock system derived from VP neurons located in the SCN. It is still a matter of debate whether the VP system discovered first, i.e. the magnocellular VP neurons in the PVN and SON that produce VP as a hormone, also contribute to the central projections. Some double-labelled neurons in the PVN after tracer injections in both the neural lobe and spinal cord seem to indicate the latter. Steroid hormones and central VP systems have been found to interact in a variety of ways. The clearest example is presented by the sexually dimorphic system, which shows a complete disappearance

58

in the absence of gonadal steroids. The VP neurons of the endocrine system show an opposite response in the sense that VP synthesis in these neurons is upregulated in the absence of the adrenal glucocorticoids. On the other hand, the VP neurons of the biological clock system and the autonomic system do not seem to be affected by circulating steroid hormones, but VP released from the biological clock system is involved in the control of corticosterone release and together with estrogen controls the release of luteinizing hormone. Next, the various interactions of steroid hormones and VP will be discussed in more detail for each of the central VP systems described above. Steroid hormone effects on vasopressin systems The magnocellular VP system A large number of experiments have indicated an effect of gonadal steroids on the VP system that projects to the bloodstream. For instance, under basal conditions, plasma VP and 24-h urinary excretion of VP are higher in male than in female rats (Stone et al., 1989). Immobilization stress results in a marked stimulation of plasma VP concentrations in females, whereas in males there is no significant VP response (Carter and Lightman, 1986). In vitro, the osmotically stimulated VP release from a hypothalamic slice and the VP mRNA increase are inhibited by testosterone (Swenson and Sladek, 1997). In the female rat the activity of the neurohypophysial VP system is affected by the estrous cycle (Swaab and Jongkind, 1970; Skowsky et al., 1979; Forsling et al., 1991). Also the intrahypothalamic release of VP in response to the elevation of plasma osmolality is greater in female rats than in males (Ota et al., 1994). During the 48 h preceding and following parturition dramatic changes occur in the magnocellular VP neurons of the PVN and SON of the rat as reflected in its activity and mRNA levels (Swaab and Jongkind, 1970; Thomas et al., 1996). Also in humans a sexual dimorphism in the control of plasma VP has been demonstrated (Rhodes and Rubin, 1999). The effects of glucocorticoids on the magnocellular VP system are less clear. VP release from hypothalamic slices containing both the PVN and

SON is inhibited in a dose-dependent manner by corticosterone (Liu and Chen, 1995), but in vivo, blockade of glucocorticoid receptors resulted in decreased VP mRNA levels in the SON and decreased plasma VP levels (Pesonen et al., 1992). On the other hand, it has been demonstrated several times that adrenalectomy does not alter magnocellular VP activity (Herman, 1995). The effects, described above, on the magnocellular VP system may either result from the direct influence of steroid hormones on magnocellular VP neurons or involve indirect effects of steroid hormones affecting the central regulation of VP secretion from the posterior pituitary. Indeed, initially no receptors for either gonadal steroids (Axelson and Van Leeuwen, 1990; Zhou et al., 1994) or adrenal steroids (Morimoto et al., 1996) could be detected on magnocellular VP neurons. Recently, however, the presence of estrogen receptor β was reported in magnocellular VP neurons (Alves et al., 1998). The sexually dimorphic VP system The effects of gonadal steroids on the VP systems in the BNST and the AME (Fig. 1) are among the most dramatic reported for a neurotransmitter system. Several other neuropeptide systems also show dramatic fluctuations in their expression under the influence of gonadal steroids; however, a complete elimination of the expression of a particular neuropeptide by gonadectomy has been reported only for the VP cells of the BNST and AME (De Vries et al., 1986). Those pronounced effects of sex steroids have also resulted in some very clear sexual dimorphisms, with male rats having twice as many VP-ir cells in the BNST than females. Since these VP neurons project to many forebrain areas, such as the lateral habenular nucleus and the lateral septum, consequently these target areas also contain many more VP-ir fibers in males than in females (Fig. 2; De Vries et al., 1981; De Vries and Buijs, 1983; Van Leeuwen et al., 1984; Bamshad et al., 1993; Wang et al., 1996). Following gonadectomy of adult male or female rats, VP fiber density in the lateral septum and lateral habenular nucleus decreases gradually until, after about 10 weeks, virtually no fibers can be detected anymore (De Vries et al., 1984; Dubois-Dauphin et al., 1994). Similar changes

59

Fig. 1. The vasopressin sexually dimorphic system as illustrated by a sagittal scheme of the main vasopressin pathways arising from the BNST and AME. LS, lateral septum; VDBB, diagonal band of Broca; LH, lateral habenula; HIP, hippocampus; CG, central gray; DR, dorsal raphe; LC, locus coeruleus; OT, olfactory tubercle.

are observed when there is a natural cause for the decreased gonadal activity, i.e. during aging (Fliers et al., 1985; Goudsmit et al., 1988) or during the non-breeding season in seasonal species (Buijs et al., 1986; Hermes et al., 1990; Lakhdar-Ghazal et al., 1995; Bittman et al., 1996). In addition to altering the activity of VP cells in the adult rat, gonadal steroids also have a perinatal ‘organizational’ effect, which results in a sexual dimorphism of these extrahypothalamic VP systems (De Vries et al., 1984; Wang et al., 1993). Concomitant with VP-ir also VP mRNA levels decline, indicating that gonadal steroids do not only affect VP peptide content but also VP gene expression (Miller et al., 1989; Carter and Murphy, 1993; Szot and Dorsa, 1994). Disappearance of VP-ir cells and fibers and VP mRNA is prevented in females by estrogen and in males by testosterone replacement. In contrast with the gonadal steroids, adrenal steroids do not seem to be a strong signal for the sexually dimorphic VP system. Although dexamethasone

is able to suppress VP gene expression in the BNST and AME, it does so apparently by its suppressive effect on circulating testosterone levels (Urban et al., 1991). An opposite effect is seen in Sprague–Dawley rats, i.e. removal of circulating glucocorticoids causes a decrease of circulating testosterone levels and a subsequent decrease of VP mRNA in the AME (Viau et al., 2001); apparently the sensitivity of the gonadal system to adrenal manipulation is strain-dependent. The endocrine VP system The parvocellular VP neurons in the PVN belong to either the endocrine or the autonomic VP system as described above (Fig. 3). Vasopressin from neurons of the endocrine VP system is released in the median eminence and in most of these neurons VP is co-localized with corticotropin-releasing hormone (CRH). It has previously been shown that various stimuli causing an activation of the hypothalamic– pituitary–adrenal axis, such as immobilization, nov-

60

Fig. 2. Sex difference in the vasopressin innervation of the lateral septum. Males (B) have many more vasopressin-immunoreactive fibers than females (A). Modified from De Vries et al. (1981).

elty, insulin and high doses of colchicine, go along with a depletion of the CRH and VP content in the external zone of the median eminence (Berkenbosch et al., 1989; De Goeij et al., 1992; Bartanusz et al., 1993; Romero et al., 1993). The co-release of VP strongly potentiates the ACTH-releasing effect of CRH (Rivier and Vale, 1983; Rivier et al., 1984). The emerging view is that VP is the principal regulated variable that puts a situation-specific drive on the axis, whereas CRH serves mainly to impose a stimulatory tone (Kovacs et al., 2000). Under resting conditions the parvocellular VP neurons of the PVN

do not contain detectable amounts of VP immunoreactivity or mRNA. However, in the early 80s it was shown that both CRH and VP expression are strongly upregulated after adrenalectomy (Sawchenko et al., 1984; Whitnall, 1988). The increased expression of VP and CRH in the parvocellular PVN neurons apparently is due to a disappearance of glucocorticoid feedback, since corticosterone replacement or dexamethasone administration prevented these effects of adrenalectomy (Kovacs et al., 1986; Carter et al., 1993; Herman, 1995). Although local implants of dexamethasone into the region of the PVN are suffi-

61

Fig. 3. The autonomic and endocrine vasopressin system as illustrated by a sagittal scheme of the main vasopressin pathways arising from the parvocellular PVN neurons. ME, median eminence; PBN, parabrachial nucleus; A1, A1 group; DVC, dorsal vagal complex; IML, intermediolateral column of the spinal cord.

cient to block adrenalectomy-induced CRH and AVP increases, and glucocorticoid receptors are present on CRH-containing neurons (Liposits et al., 1987; Ceccatelli et al., 1989), local effects of glucocorticoids on parvocellular neurons may not be the sole source of inhibitory feedback. Several brain regions other than the PVN, such as the lateral septum and the hippocampus, also contain high levels of glucocorticoid receptors and exert a profound inhibitory effect on the activity of the HPA axis (Moberg et al., 1971; Seggie et al., 1974; Baldino et al., 1988). The adrenalectomy-induced increase of VP (but not CRH) synthesis in the PVN is abolished by a concomitant gonadectomy (Viau et al., 1999, 2001), suggesting a permissive role for gonadal steroids on VP synthesis that is revealed in the absence of glucocorticoids. A stimulatory effect of estrogen on VP and/or CRH synthesis might be part of the explanation for the well-known sexual dimorphism in HPA activity (Nicholson et al., 1985; Handa et al., 1994; Atkinson and Waddell, 1997), with higher

plasma corticosterone levels consistently reported in females. Indeed, females show higher bioassayable levels of hypothalamic CRH (Hiroshige et al., 1973) and increased levels of CRH mRNA in the PVN (Watts and Swanson, 1989). However, since androgen receptors in the PVN are essentially located in those neurons in the PVN that give rise to the descending projections (Simerly et al., 1990; Zhou et al., 1994), the feedback actions of the gonadal steroids on the endocrine VP system are probably neurogenic in nature, mediated upstream from the PVN (Viau and Meaney, 1996). The autonomic VP system Vasopressin-containing parvocellular neurons in the PVN also project to the brainstem and spinal cord, where they terminate within the nucleus of the solitary tract, the dorsal motor nucleus of the vagus and the intermediolateral column (Swanson, 1977; Buijs, 1978; Lang et al., 1983; Sawchenko, 1987), i.e.

62

nuclei containing either the parasympathetic or sympathetic preganglionic neurons. Therefore, this part of the PVN VP system may play an important role in the regulation of autonomic mechanisms and is labeled the autonomic VP system (Fig. 3). Double retrograde tracing studies have established that the populations of cells in the PVN that project to the median eminence are almost completely separated from those that give rise to the long descending projections (Swanson and Kuypers, 1980; Swanson et al., 1980). The absence of increased immunostaining after adrenalectomy in PVN neurons with descending projections (Sawchenko, 1987) affirms this clear separation of neurons belonging to either the endocrine or the autonomic VP system in the PVN. Both glucocorticoid and gonadal steroid receptors have been found in the parvocellular PVN neurons (Sar and Stumpf, 1980; Rhodes et al., 1982; Ceccatelli et al., 1989; Jirikowski et al., 1993; Morimoto et al., 1996); however, only for the gonadal steroid receptors a co-localization within neurons with descending projections could be shown (Corodimas and Morrell, 1990; Wagner et al., 1993; Zhou et al., 1994). Most of them apparently contain oxytocin (Axelson and Van Leeuwen, 1990; Zhou et al., 1994; Simonian and Herbison, 1997; Alves et al., 1998). Clear effects of either glucocorticoids or gonadal steroids on the autonomic VP system have not been described. The biological clock VP system The main component of the mammalian biological clock system, i.e. the endogenous pacemaker, is localized in the SCN. The SCN is a bilaterally paired nucleus in the ventral part of the anterior hypothalamus, situated next to the third ventricle and on top of the optic chiasm. It consists of several classes of neurons which can be identified by their electrophysiological properties (Pennartz et al., 1998a,b) or their transmitter content (Romijn et al., 1996, 1997). Vasopressin-containing neurons constitute an important subpopulation of this heterogeneous nucleus, i.e. between 10% and 30% of the ∼10,000 SCN neurons may contain VP (Sofroniew and Weindl, 1980; Moore and Speh, 1993; Madeira et al., 1997), a feature which is maintained across many mammalian species (Sofroniew and Weindl, 1980; Cassone et al., 1988; Reuss et al., 1989; Goel et al., 1999; Smale

and Boverhof, 1999). SCN neurons contain an internal pacemaker producing an endogenous rhythm in electrical activity with peak values found during subjective daytime even when the SCN neurons are completely isolated from the surrounding brain tissue (Inouye and Kawamura, 1979; Groos and Hendriks, 1982; Bos and Mirmiran, 1990). Interestingly, VP secretion follows the general pattern of electrical and metabolic activity of the SCN neurons (Gillette and Reppert, 1987). Amongst others the medial preoptic area (MPOA), the periventricular PVN and the dorsomedial hypothalamus (DMH) have been identified as projection areas of the SCN VP neurons (Fig. 4) by using different combinations of lesion, tracer and immunocytochemical techniques (Hoorneman and Buijs, 1982; Watts and Swanson, 1987; Kalsbeek et al., 1993; Leak and Moore, 2001). In contrast with the projection areas of the VP neurons in the BNST and the AME, VP fiber densities in SCN projection areas do not respond to changes in the level of circulating gonadal steroids (Fig. 5; De Vries et al., 1984; Viau et al., 2001). This lack of effect of gonadal steroids on the biological clock VP system is in good agreement with the relative paucity of either estrogen or androgen receptors in the SCN (Simerly et al., 1990; Zhou et al., 1994; Shughrue et al., 1997; Alves et al., 1998). On the other hand, sexual dimorphisms of its projections have been reported (Crenshaw et al., 1992; LakhdarGhazal et al., 1992; Horvath et al., 1998) and several additional data suggest that gonadal steroids are able to affect the (output of the) biological clock (Daan et al., 1975; Morin et al., 1977; Albers, 1981; Kow and Pfaff, 1984; Krajnak et al., 1998). At present, however, it seems most probable that these changes are neurogenic, i.e. mediated by steroid-sensitive afferent inputs to the SCN (De La Iglesia et al., 1999; Horvath et al., 1999). Like gonadal steroid receptors, glucocorticoid receptors, too, show a relative paucity in the adult mammalian SCN (Van Eekelen et al., 1987; Yi et al., 1994) and the effects of glucocorticoids on the SCN VP system are at best ambiguous (Vandesande et al., 1974; Carter and Murphy, 1989; Gozes et al., 1994; Larsen et al., 1994; Maurel et al., 2000; Viau et al., 2001). The apparent insensitivity of the biological clock to adrenal steroids has recently been strengthened further by its lack of effect on circadian clock

63

Fig. 4. Photomicrographs of the thalamic area around the third ventricle (III) in a control (B) and castrated rat (A). Note that castration eliminates the staining of vasopressin fibers in the lateral habenula (LH) that derive from the BNST, but not the staining of fibers in the paraventricular nucleus of the thalamus (PV), which are derived from the SCN. Modified from De Vries et al. (1984).

genes in the SCN as opposed to those in peripheral tissues (Balsalobre et al., 2000). Vasopressin effects on steroid hormones Vasopressin and the corticosterone rhythm The prominent rhythm of VP release from SCN terminals and the close proximity of its PVN projections to the CRH-containing neuroendocrine motorneurons led us to investigate a possible causal relation (Kalsbeek and Buijs, 1992). Our initial experiments revealed a strong inhibitory effect of SCNderived VP on corticosterone release when delivered at the level of the PVN/DMH area in SCN-lesioned animals (Kalsbeek et al., 1992). Subsequent studies

using the application of VP and a VP antagonist in the PVN/DMH area of SCN-intact animals proved the inhibition to be caused by endogenous VP, released from SCN terminals (Fig. 6). Additional experiments using timed VP antagonist administration at different times of the L/D cycle further clarified the control mechanism of the daily corticosterone surge (Kalsbeek et al., 1996b). The increased release of VP during the first part of the light period ensures basal corticosterone levels during the initial part of the sleep period. Subsequently, the concomitant arrest of VP release and the increased release of an additional SCN transmitter stimulating corticosterone release during the second part of the light period results in the daily corticosterone surge just before awakening. After the onset of the dark period also

64

Fig. 5. The biological clock vasopressin system as illustrated by a sagittal scheme of the main projections arising from the vasopressinergic SCN neurons. MPOA, medial part of the preoptic area; Pe, periventricular part of the PVN; DMH, dorsomedial nucleus of the hypothalamus; PV, paraventricular nucleus of the thalamus.

Fig. 6. Effects of a 1-h administration of vasopressin, vasopressin antagonist or Ringer (all from t = 0 onwards) at the level of the PVN/ DMH area on plasma corticosterone values. Note the pronounced dis-inhibitory effect of the vasopressin antagonist, whereas vasopressin and Ringer do not notably affect basal corticosterone levels. The shaded area delineates corticosterone levels of control animals without hypothalamic infusions. Modified from Kalsbeek et al. (1996a).

65

the release of the stimulatory SCN transmitter will vanish and corticosterone levels will slowly decline. Although in the meantime the inhibitory effect of VP on corticosterone release as just described has been replicated by different laboratories (Wotjak et al., 1996; Gomez et al., 1997) its direction is completely opposite to the previously described and well-known stimulatory effect of VP on ACTH and corticosterone release (Gillies et al., 1982; Rivier and Vale, 1983; Rivier et al., 1984). The logical explanation of course is that the different effects are caused by VP derived from two completely different systems. The inhibitory effect concerns VPergic SCN neurons that synapse at the level of the PVN and DMH, the stimulatory effect is due to VP derived from parvocellular PVN neurons and acting at the level of the median eminence. At present it is not clear which are the exact target neurons for the inhibitory VPergic projections. Since VP generally has a stimulatory effect on its postsynaptic target neuron (Raggenbass, 2001) it is unlikely that the VP-containing fibers directly contact CRH-containing neurons. Indeed, a direct projection is also not supported by the sparse SCN projections to the hypophysiotropic neurons (Buijs et al., 1993; Vrang et al., 1995). Most probably the inhibitory effect of VP is mediated by GABAergic interneurons in the DMH and subparaventricular PVN (Hermes et al., 2000). However, when comparing the corticosterone and ACTH responses induced by the hypothalamic administrations of the VP antagonist it is clear that the SCN control of corticosterone release cannot be explained solely by its effect on the HPA axis, i.e. via the subsequent release of CRH and ACTH. Comparing the response of ACTH with that of corticosterone in different experimental conditions reveals that the circadian release of corticosterone does not depend heavily on the release of ACTH, contrary to, for instance, the stress-related release of corticosterone (Kalsbeek et al., 1996b; Buijs et al., 1997). Instead of a stimulation of ACTH release it seems that the daily corticosterone peak is mainly caused by an increased sensitivity of the adrenal cortex to ACTH (Jasper and Engeland, 1994; Buijs et al., 1999). Indeed, using the transneuronal virus tracing technique a neuronal pathway between SCN, dorsal PVN, spinal cord and adrenal cortex could be established (Buijs et al., 1999). Therefore, via the GABA-containing interneurons in the DMH

and the subparaventricular PVN, the VPergic projection from the SCN seems to contact not only the endocrine CRH-containing neurons but also the autonomic neurons in the dorsal PVN. Therefore, via its effect on the GABAergic interneurons the daytime peak of VP release from SCN terminals in the PVN/DMH area might ensure basal corticosterone levels via two separate mechanisms: (1) it inhibits the release of ACTH via its effect on the endocrine CRH neurons, and (2) it inhibits the sensitivity of the adrenal cortex for ACTH via its effect on the autonomic PVN neurons. Evidently, the purpose of this rhythm is not only to drive the daily corticosterone peak in mammals but also to maintain, for example, stress-induced corticosterone activations within the physiological range appropriate for the stress and appropriate for that moment of the L/D cycle (Buijs et al., 1997). Vasopressin and the luteinizing hormone surge Daily rhythms in plasma levels of gonadal steroids are usually not very pronounced, except during human puberty (Norjavaara et al., 1996; Mitamura et al., 1999; Gupta et al., 2000). Nevertheless there is a clear relation between the mammalian biological clock and many aspects of sexual behavior: for instance the temporal organization of pulsatile activity in the hypothalamic–pituitary–adrenal (HPG) axis is essential for successful reproduction. Lesion studies have shown that two brain structures are indispensable for generating the preovulatory surge of luteinizing hormone (LH): firstly, the MPOA containing the dense concentration of estrogen receptors necessary for the positive estrogen feedback and, secondly, the SCN providing the timing signal for the LH surge on the day of proestrus. Early anatomical studies already indicated a dense VP innervation in the MPOA, which probably derives from the SCN since it was not sensitive to gonadal hormones (Hoorneman and Buijs, 1982; De Vries et al., 1984). More recent studies showed that estrogen receptor containing neurons in the MPOA receive direct synaptic contacts from SCN fibers, probably containing VP as a neurotransmitter (De La Iglesia et al., 1995; Watson et al., 1995) and that VP receptor mRNA is expressed in MPOA neurons (Ostrowski et al., 1994; Funabashi et al., 2000b). In addition,

66

Fig. 7. Daily levels of plasma LH in intact (shaded area) and ovariectomized and estrogen-treated animals. The ovariectomized and estrogen-treated animals were administered vasopressin in the MPOA during 5 h. Vasopressin-treated animals were divided in animals with a high LH surge amplitude (•; ∼1/3 of the animals) and those with an LH surge comparable to that of control animals (◦; ∼2/3 of the animals). Modified from Palm et al. (2001).

some early works of Södersten et al. (1983, 1985, 1986) indicated an interesting relationship between female sexual behavior and SCN-derived VP, although the effect had not been localized to a specific SCN target area. We hypothesized that the MPOA functions as an intermediate brain area for the transmission of circadian information from the SCN to the HPG axis. Indeed, an increase in extracellular VP levels (by reverse microdialysis) in the MPOA of SCN-intact animals had a stimulatory effect on the LH surge (Fig. 7), whereas it did not affect plasma corticosterone levels (Palm et al., 2001). The stimulatory effect of VP was restricted to a specific time-period that coincided with the sensitive time window for a daily neuronal signal prior to the LH surge (Everett and Sawyer, 1950), and also with the peak of VP secretion by SCN neurons (Schwartz et al., 1983; Schwartz and Reppert, 1985; Gillette and Reppert, 1987; Kalsbeek et al., 1995). The important role of SCN-derived VP in the initiation of the LH surge was further emphasized by our experiments in SCN-lesioned animals. The complete absence of any circadian output from the SCN induces basal, nonfluctuating LH levels, but a 2-h administration of VP in the MPOA is sufficient to reinstate a complete LH surge that is comparable to the estrogen-induced

surges in SCN-intact animals, both in shape and amplitude (Palm et al., 1999). Therefore, in our view the high VP secretion by SCN terminals in the MPOA, occurring during the sensitive time window prior to the surge, is the circadian signal essential for the generation of an LH surge. So, by the simultaneous secretion of VP in different target areas, the SCN may reduce the stress response in one area (PVN/DMH), while it stimulates sexual receptivity in another (MPOA). In addition, it seems physiologically relevant that the stimulatory role of VP on the HPG axis coincides with an inhibitory role of SCN-derived VP on stress hormone secretion. Once every 4 or 5 days the combination of a positive estrogen feedback and circadian VP input, probably by MPOA neurons bearing both estrogen and VP receptors, enables the MPOA to trigger a concerted action of gonadotropin-releasing hormone (GnRH) containing neurons. The resulting massive release of GnRH in the median eminence will initiate the next LH surge. Using a completely different experimental setup a similar conclusion was reached by Funabashi et al. (2000a). Next to this indirect control of the SCN direct projections from the SCN to the endocrine GnRH motorneurons, although sparse, have also been reported (Van Der Beek et al., 1993,

67

1997), i.e. an anatomical situation very much comparable with the circadian control of the HPA axis. The use of intermediate brain structures such as the DMH and MPOA, of course, has the great advantage of enabling the integration of information from different sources (such as circadian, hormonal, metabolic, stress, etc.) before a definitive signal is sent to the endocrine motorneuron.

Pe PV PVN

Conclusion

References

The present overview shows that there are a number of interactions between central VP systems and circulating steroid hormones. For understanding the type of interaction, however, it is essential not to consider the central VP system as a uniform entity, but to appreciate its variation. In addition to the fact that the distribution of steroid receptors may be very site-specific, also the neurotransmitter functions of VP are closely linked to its site of origin and production. Finally, of course, the function of VP in the target area will depend heavily on the function of that region. Abbreviations ACTH AME BNST CG CRH DMH DR DVC GABA GnRH HIP HPA HPG IML ir LC L/D LH ME MPOA OT PBN

adrenocorticotropic hormone medial amygdala bed nucleus of the stria terminalis central gray corticotropin-releasing hormone dorsomedial nucleus of the hypothalamus dorsal raphe dorsal vagal complex gamma aminobutyric acid gonadotropin-releasing hormone hippocampus hypothalamic–pituitary–adrenal hypothalamic–pituitary–gonadal intermediolateral column of the spinal cord immunoreactivity locus coeruleus light/dark luteinizing hormone median eminence medial preoptic area olfactory tubercle parabrachial nucleus

SCN SON VDBB VP

periventricular area paraventricular nucleus of the thalamus paraventricular nucleus of the hypothalamus suprachiasmatic nucleus supraoptic nucleus diagonal band of Broca (ventral part) vasopressin

Albers, H.E. (1981) Gonadal hormones organize and modulate the circadian system of the rat. Am. J. Physiol., 241: R62–R66. Alves, S.E., Lopez, V., Mcewen, B.S. and Weiland, N.G. (1998) Differential colocalization of estrogen receptor β (ERβ) with oxytocin and vasopressin in the paraventricular and supraoptic nuclei of the female rat brain: An immunocytochemical study. Proc. Natl. Acad. Sci. USA, 95: 3281–3286. Atkinson, H.C. and Waddell, B.J. (1997) Circadian variation in basal plasma corticosterone and adrenocorticotropin in the rat: Sexual dimorphism and changes across the estrous cycle. Endocrinology, 138: 3842–3848. 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. Baldino, F., O’Kane, T.M., Fitzpatrick-McElligott, S. and Wolfson, B. (1988) Coordinate hormonal and synaptic regulation of vasopressin messenger RNA. Science, 241: 978–981. Balsalobre, A., Brown, S.A., Marcacci, L., Tronche, F., Kellendonk, C., Reichardt, H.M., Schütz, G. and Schibler, U. (2000) Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science, 289: 2344–2347. Bamshad, M., Novak, M.A. and De Vries, G.J. (1993) Sex and species differences in the vasopressin innervation of sexually naive and parental Prairie voles, Microtus ochrogaster and Meadow voles, Microtus pennsylvanicus. J. Neuroendocrinol., 5: 247–255. Bargmann, W. and Scharrer, E. (1951) The site of origin of the hormones of the posterior pituitary. Am. Sci., 39: 255–259. Bartanusz, V., Jezova, D., Bertini, L.T., Tilders, F.J.H., Aubry, J.M. and Kiss, J.Z. (1993) Stress-induced increase in vasopressin and corticotropin-releasing factor expression in hypophysiotropic paraventricular neurons. Endocrinology, 132: 895–902. Berkenbosch, F., De Goeij, D.C.E. and Tilders, F.J.H. (1989) Hypoglycemia enhances turnover of corticotropin-releasing factor and of vasopressin in the zona externa of the rat median eminence. Endocrinology, 125: 28–34. Bittman, E.L., Jetton, A.E., Villalba, C. and De Vries, G.J. (1996) Effects of photoperiod and androgen on pituitary function and neuropeptide staining in Siberian hamsters. Am. J. Physiol., 40: R64–R72. Bos, N.P.A. and Mirmiran, M. (1990) Circadian rhythms in spon-

68 taneous neuronal discharges of the cultured suprachiasmatic nucleus. Brain Res., 511: 158–162. Buijs, R.M. (1978) Intra- and extrahypothalamic vasopressin and oxytocin pathways in the rat. Pathways to the limbic system, medulla oblongata and spinal cord. Cell Tissue Res., 192: 423–435. 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., Swaab, D.F., Dogterom, J. and Van Leeuwen, F.W. (1978) Intra- and Extrahypothalamic vasopressin and oxytocin pathways in the rat. Cell Tissue Res., 186: 423–433. Buijs, R.M., Pévet, P., Masson-Pévet, M., Pool, C.W., De Vries, G., Canguilhem, B. and Vivien-Roels, B. (1986) Seasonal variation in vasopressin innervation in the brain of the European hamster (Cricetus cricetus). Brain Res., 371: 193–196. Buijs, R.M., Markman, M., Nunes-Cardoso, B., Hou, Y.-X. and Shinn, S. (1993) Projections of the suprachiasmatic nucleus to stress-related areas in the rat hypothalamus: A light and electron microscopic study. J. Comp. Neurol., 335: 42–54. Buijs, R.M., Wortel, J., Van Heerikhuize, J.J. and Kalsbeek, A. (1997) Novel environment induced inhibition of corticosterone secretion: Physiological evidence for a suprachiasmatic nucleus mediated neuronal hypothalamo–adrenal cortex pathway. Brain Res., 758: 229–236. Buijs, R.M., Wortel, J., Van Heerikhuize, J.J., Feenstra, M.G.P., Ter Horst, G.J., Romijn, H.J. and Kalsbeek, A. (1999) Anatomical and functional demonstration of a multisynaptic suprachiasmatic nucleus adrenal (cortex) pathway. Eur. J. Neurosci., 11: 1535–1544. Carter, D.A. and Lightman, S.L. (1986) Diurnal pattern of stressevoked neurohypophyseal hormone secretion: sexual dimorphism in rats. Neurosci. Lett., 71: 252–255. Carter, D.A. and Murphy, D. (1989) Diurnal rhythm of vasopressin mRNA species in the rat suprachiasmatic nucleus: independence of neuroendocrine modulation and maintenance in explant culture. Mol. Brain Res., 6: 233–239. Carter, D.A. and Murphy, D. (1993) Regulation of vasopressin (VP) gene expression in the bed nucleus of the stria terminalis — Gonadal steroid-dependent changes in VP mRNA accumulation are associated with alterations in mRNA poly (A) tail length but are independent of the rate of VP gene transcription. J. Neuroendocrinol., 5: 509–515. Carter, D.A., Pardy, K. and Murphy, D. (1993) Regulation of vasopressin gene expression — Changes in the level, but not the size, of vasopressin messenger RNA following endocrine manipulations. Cell. Mol. Neurobiol., 13: 87–95. Cassone, V.M., Speh, J.C., Card, J.P. and Moore, R.Y. (1988) Comparative anatomy of the mammalian hypothalamic suprachiasmatic nucleus. J. Biol. Rhythms, 3: 71–91. Ceccatelli, S., Cintra, A., Hökfelt, T., Fuxe, K., Wikstrom, A.C. and Gustafsson, J.A. (1989) Coexistence of glucocorticoid receptor-like immunoreactivity with neuropeptides in the hypothalamic paraventricular nucleus. Exp. Brain Res., 78: 33– 42. Corodimas, K.P. and Morrell, J.I. (1990) Estradiol-concentrating

forebrain and midbrain neurons project directly to the medulla. J. Comp. Neurol., 291: 609–620. Crenshaw, B.J., De Vries, G.J. and Yahr, P. (1992) Vasopressin innervation of sexually dimorphic structures of the Gerbil forebrain under various hormonal conditions. J. Comp. Neurol., 322: 589–598. Daan, S., Damassa, D., Pittendrigh, C.S. and Smith, E.R. (1975) An effect of castration and testosterone replacement on a circadian pacemaker in mice (Mus musculus). Proc. Natl. Acad. Sci. USA, 72: 3744–3747. De Goeij, D.C.E., Jezova, D. and Tilders, F.J.H. (1992) Repeated stress enhances vasopressin synthesis in corticotropin releasing factor neurons in the paraventricular nucleus. Brain Res., 577: 165–168. De La Iglesia, H.O., Blaustein, J.D. and Bittman, E.L. (1995) The suprachiasmatic area in the female hamster projects to neurons containing estrogen receptors and GnRH. Neuroreport, 6: 1715–1722. De La Iglesia, H.O., Blaustein, J.D. and Bittman, E.L. (1999) Oestrogen receptor-α-immunoreactive neurones project to the suprachiasmatic nucleus of the female Syrian hamster. J. Neuroendocrinol., 11: 481–490. 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: 303–317. De Vries, G.J., Buijs, R.M. and Swaab, D.F. (1981) Ontogeny of the vasopressinergic neurons of the suprachiasmatic nucleus and their extrahypothalamic projections in the rat brain — presence of a sex difference in the lateral septum. Brain Res., 218: 67–78. De Vries, G.J., Buijs, R.M. and Sluiter, A.A. (1984) Gonadal hormone actions on the morphology of the vasopressinergic innervation of the adult rat brain. Brain Res., 298: 141–145. De Vries, G.J., Duetz, W., Buijs, R.M., Van Heerikhuize, J.J. and Vreeburg, J.T.M. (1986) Effects of androgens and estrogens on the vasopressin and oxytocin innervation of the adult rat brain. Brain Res., 399: 296–302. De Wied, D. (1965) The influence of the posterior and intermediate lobe of the pituitary and pituitary peptides on the maintenance of a conditioned avoidance response in rats. Int. J. Neuropharmacol., 4: 157–167. Dubois-Dauphin, M., Theler, J.M., Ouarour, A., Pévet, P., Barberis, C. and Dreifuss, J.J. (1994) Regional differences in testosterone effects on vasopressin receptors and on vasopressin immunoreactivity in intact and castrated Siberian hamsters. Brain Res., 638: 267–276. Everett, J.W. and Sawyer, C.H. (1950) A 24-hour periodicity in the ‘LH release apparatus’ of female rats, disclosed by barbiturate sedation. Endocrinology, 47: 198–218. Fliers, E., De Vries, G.J. and Swaab, D.F. (1985) Changes with aging in the vasopressin and oxytocin innervation of the rat brain. Brain Res., 348: 1–8. Forsling, M.L., Kelestimur, H. and Windle, R. (1991) The influence of reproductive status on vasopressin release in the rat. J. Endocrinol., 130: 387–393. Funabashi, T., Shinohara, K., Mitsushima, D. and Kimura, F.

69 (2000a) Gonadotropin-releasing hormone exhibits circadian rhythm in phase with arginine–vasopressin in co-cultures of the female rat preoptic area and suprachiasmatic nucleus. J. Neuroendocrinol., 12: 521–528. Funabashi, T., Shinohara, K., Mitsushima, D. and Kimura, F. (2000b) Estrogen increases arginine–vasopressin V1a receptor mRNA in the preoptic area of young but not of middle-aged female rats. Neurosci. Lett., 285: 205–208. Gillette, M.U. and Reppert, S.M. (1987) The hypothalamic suprachiasmatic nuclei: circadian patterns of vasopressin secretion and neuronal activity in vitro. Brain Res. Bull., 19: 135–139. Gillies, G.E., Linton, E.A. and Lowry, P.J. (1982) Corticotropin releasing activity of the new CRF is potentiated several times by vasopressin. Nature, 299: 355–357. Goel, N., Lee, T.M. and Smale, L. (1999) Suprachiasmatic nucleus and intergeniculate leaflet in the diurnal rodent Octodon degus: Retinal projections and immunocytochemical characterization. Neuroscience, 92: 1491–1509. Gomez, F., Chapleur, M., Fernette, B., Burlet, C., Nicolas, J.P. and Burlet, A. (1997) Arginine vasopressin (AVP) depletion in neurons of the suprachiasmatic nuclei affects the AVP content of the paraventricular neurons and stimulates adrenocorticotrophic hormone release. J. Neurosci. Res., 50: 565– 574. Goudsmit, E., Fliers, E. and Swaab, D.F. (1988) Testosterone supplementation restores vasopressin innervation in the senescent rat brain. Brain Res., 473: 306–313. Gozes, I., Avidor, R., Giladi, E., Shani, Y., Mcewen, B.S., Dussaillant, M. and Rostene, W. (1994) Adrenalectomy decreases vasoactive intestinal peptide mRNA levels in the rat suprachiasmatic nucleus. Neurosci. Lett., 167: 24–28. Groos, G.A. and Hendriks, J. (1982) Circadian rhythms in electrical discharge of rat suprachiasmatic neurones recorded in vitro. Neurosci. Lett., 34: 283–288. Gupta, S.K., Lindemulder, E.A. and Sathyan, G. (2000) Modeling of circadian testosterone in healthy men and hypogonadal men. J. Clin. Pharmacol., 40: 731–738. Handa, R.J., Burgess, L.H., Kerr, J.E. and O’Keefe, J.A. (1994) Gonadal steroid hormone receptors and sex differences in the hypothalamo–pituitary–adrenal axis. Horm. Behav., 28: 464– 476. Herman, J.P. (1995) In situ hybridization analysis of vasopressin gene transcription in the paraventricular and supraoptic nuclei of the rat: Regulation by stress and glucocorticoids. J. Comp. Neurol., 363: 15–27. Hermes, M.L.H.J., Buijs, R.M., Masson-Pévet, M. and Pévet, P. (1990) Seasonal changes in vasopressin in the brain of the garden dormouse (Eliomys quercinus L.). J. Comp. Neurol., 293: 340–346. Hermes, M.L.H.J., Ruijter, J.M., Klop, A., Buijs, R.M. and Renaud, L.P. (2000) Vasopressin increases GABAergic inhibition of rat hypothalamic paraventricular nucleus neurons in vitro. J. Neurophysiol., 83: 705–711. Hiroshige, T., Abe, K., Wada, S. and Kaneko, M. (1973) Sex differences in circadian periodicity of CRF activity in the rat hypothalamus. Neuroendocrinology, 11: 306–320.

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. Horvath, T.L., Cela, V. and Van Der Beek, E.M. (1998) Genderspecific apposition between vasoactive intestinal peptidecontaining axons and gonadotrophin-releasing hormoneproducing neurons in the rat. Brain Res., 795: 277–281. Horvath, T.L., Diano, S., Sakamoto, H., Shughrue, P.J. and Merchenthaler, I. (1999) Estrogen receptor β and progesterone receptor mRNA in the intergeniculate leaflet of the female rat. Brain Res., 844: 196–200. Inouye, S.I.T. and Kawamura, H. (1979) Persistence of circadian rhythmicity in a mammalian hypothalamic ‘island’ containing the suprachiasmatic nucleus. Proc. Natl. Acad. Sci. USA, 76: 5962–5966. Jasper, M.S. and Engeland, W.C. (1994) Splanchnic neural activity modulates ultradian and circadian rhythms in adrenocortical secretion in awake rats. Neuroendocrinology, 59: 97– 109. Jirikowski, G.F., Mcgimsey, W.C., Caldwell, J.D. and Sar, M. (1993) Distribution of oxytocinergic glucocorticoid target neurons in the rat hypothalamus. Horm. Metab. Res., 25: 543– 544. Kalsbeek, A. and Buijs, R.M. (1992) Peptidergic transmitters of the suprachiasmatic nuclei and the control of circadian rhythmicity. In: J. Joosse, R.M. Buijs and F.J.H. Tilders (Eds.), The Peptidergic Neuron. Progress in Brain Research, Vol. 92, Elsevier, Amsterdam, pp. 321–333. Kalsbeek, A., Buijs, R.M., Van Heerikhuize, J.J., Arts, M. and Van Der Woude, T.P. (1992) Vasopressin-containing neurons of the suprachiasmatic nuclei inhibit corticosterone release. Brain Res., 580: 62–67. Kalsbeek, A., Teclemariam-Mesbah, R. and Pévet, P. (1993) Efferent projections of the suprachiasmatic nucleus in the golden hamster (Mesocricetus auratus). J. Comp. Neurol., 332: 293–314. Kalsbeek, A., Buijs, R.M., Engelmann, M., Wotjak, C.T. and Landgraf, R. (1995) In vivo measurement of a diurnal variation in vasopressin release in the rat suprachiasmatic nucleus. Brain Res., 682: 75–82. Kalsbeek, A., Van Der Vliet, J. and Buijs, R.M. (1996a) Decrease of endogenous vasopressin release necessary for expression of the circadian rise in plasma corticosterone: a reverse microdialysis study. J. Neuroendocrinol., 8: 299–307. Kalsbeek, A., Van Heerikhuize, J.J., Wortel, J. and Buijs, R.M. (1996b) A diurnal rhythm of stimulatory input to the hypothalamo–pituitary–adrenal system as revealed by timed intrahypothalamic administration of the vasopressin V1 antagonist. J. Neurosci., 16: 5555–5565. Kovacs, K., Kiss, J.Z. and Makara, G.B. (1986) Glucocorticoid implants around the hypothalamic paraventricular nucleus prevent the increase of corticotropin-releasing factor and arginine vasopressin immunostaining induced by adrenalectomy. Neuroendocrinology, 44: 229–234. Kovacs, K.J., Foldes, A. and Sawchenko, P.E. (2000) Glucocorticoid negative feedback selectively targets vasopressin tran-

70 scription in parvocellular neurosecretory neurons. J. Neurosci., 20: 3843–3852. Kow, L.-M. and Pfaff, D.W. (1984) Suprachiasmatic neurons in tissue slices from ovariectomized rats: Electrophysiological and neuropharmacological characterization and the effects of estrogen treatment. Brain Res., 297: 275–286. Krajnak, K., Kashon, M.L., Rosewell, K.L. and Wise, P.M. (1998) Sex differences in the daily rhythm of vasoactive intestinal polypeptide but not arginine vasopressin messenger ribonucleic acid in the suprachiasmatic nuclei. Endocrinology, 139: 4189–4196. Lakhdar-Ghazal, N., Kalsbeek, A. and Pévet, P. (1992) Sexual differences and seasonal variations in vasoactive intestinal peptide immunoreactivity in the suprachiasmatic nucleus of Jerboa (Jaculus orientalis). Neurosci. Lett., 144: 29–33. Lakhdar-Ghazal, N., Dubois-Dauphin, M., Hermes, M.L.H.J., Buijs, R.M., Bengelloun, W.A. and Pévet, P. (1995) Vasopressin in the brain of a desert hibernator, the jerboa (Jaculus orientalis): Presence of sexual dimorphism and seasonal variation. J. Comp. Neurol., 358: 499–517. Lang, R.E., Heil, J., Ganten, D., Hermann, K., Rascher, W. and Unger, T. (1983) Effects of lesions in the paraventricular nucleus of the hypothalamus on vasopressin and oxytocin contents in brainstem and spinal cord of rat. Brain Res., 260: 326–329. Larsen, P.J., Vrang, N., Møller, M., Jessop, D.S., Lightman, S.L., Chowdrey, H.S. and Mikkelsen, J.D. (1994) The diurnal expression of genes encoding vasopressin and vasoactive peptide within the rat suprachiasmatic nucleus is influenced by circulating glucocorticoids. Mol. Brain Res., 27: 342–346. Leak, R.K. and Moore, R.Y. (2001) Topographic organization of suprachiasmatic nucleus projection neurons. J. Comp. Neurol., 433: 312–334. Liposits, Zs., Uht, R.M., Harrison, R.W., Gibbs, F.P., Paull, W.K. and Bohn, M.C. (1987) Ultrastructural localization of glucocorticoid receptor (GR) in hypothalamic paraventricular neurons synthesizing corticotropin releasing factor (CRF). Histochemistry, 87: 407–412. Liu, X. and Chen, Y.Z. (1995) Membrane-mediated inhibition of corticosterone on the release of arginine vasopressin from rat hypothalamic slices. Brain Res., 704: 19–22. Madeira, M.D., Andrade, J.P., Lieberman, A.R., Sousa, N., Almeida, O.F.X. and Paula-Barbosa, M.M. (1997) Chronic alcohol consumption and withdrawal do not induce cell death in the suprachiasmatic nucleus, but lead to irreversible depression of peptide immunoreactivity and mRNA levels. J. Neurosci., 17: 1302–1319. Maurel, D., Sage, D., Mekaouche, M. and Bosler, O. (2000) Glucocorticoids up-regulate the expression of glial fibrillary acidic protein in the rat suprachiasmatic nucleus. Glia, 29: 212–221. Miller, M.A., Urban, J.H. and Dorsa, D.M. (1989) Steroid dependency of vasopressin neurons in the bed nucleus of the stria terminalis by in situ hybridization. Endocrinology, 125: 2335–2340. Mitamura, R., Yano, K., Suzuki, N., Ito, Y., Makita, Y. and Okuno, A. (1999) Diurnal rhythms of luteinizing hormone,

follicle-stimulating hormone, and testosterone secretion before the onset of male puberty. J. Clin. Endocr. Metab., 84: 29–37. Moberg, G.P., Scapagnini, U., De Groot, J. and Ganong, W.F. (1971) Effect of sectioning the fornix on diurnal fluctuation in plasma corticosterone levels in the rat. Neuroendocrinology, 7: 11–15. Moore, R.Y. and Speh, J.C. (1993) GABA Is the principal neurotransmitter of the circadian system. Neurosci. Lett., 150: 112–116. Morimoto, M., Morita, N., Ozawa, H., Yokoyama, K. and Kawata, M. (1996) Distribution of glucocorticoid receptor immunoreactivity and mRNA in the rat brain: An immunohistochemical and in situ hybridization study. Neurosci. Res., 26: 235–269. Morin, L.P., Fitzgerald, K.M. and Zucker, I. (1977) Estradiol shortens the period of hamster circadian rhythms. Science, 196: 305–307. Nicholson, S., Lin, J.-H., Mahmoud, S., Campbell, E., Gillham, B. and Jones, M. (1985) Diurnal variations in responsiveness of the hypothalamo–pituitary–adrenocortical axis of the rat. Neuroendocrinology, 40: 217–224. Norjavaara, E., Ankarberg, C. and Albertsson-Wikland, K. (1996) Diurnal rhythm of 17 beta-estradiol secretion throughout pubertal development in healthy girls: Evaluation by a sensitive radioimmunoassay. J. Clin. Endocr. Metab., 81: 4095– 4102. Oliver, G. and Schäfer, E.A. (1895) On the physiological effects of extracts of the supra-renal capsules. J. Physiol. (Lond.), 8: 277–279. Ostrowski, N.L., Lolait, S.J. and Young, W.S. (1994) Cellular localization of vasopressin V1a receptor messenger ribonucleic acid in adult male rat brain, pineal, and brain vasculature. Endocrinology, 135: 1511–1528. Ota, M., Crofton, J.T., Liu, H.W., Festavan, G. and Share, L. (1994) Increased plasma osmolality stimulates peripheral and central vasopressin release in male and female rats. Am. J. Physiol., 36: R923–R928. Palm, I.F., Van Der Beek, E.M., Wiegant, V.M., Buijs, R.M. and Kalsbeek, A. (1999) Vasopressin induces an LH surge in ovariectomized, estradiol-treated rats with lesion of the suprachiasmatic nucleus. Neuroscience, 93: 659–666. Palm, I.F., Van Der Beek, E.M., Wiegant, V.M., Buijs, R.M. and Kalsbeek, A. (2001) The stimulatory effect of vasopressin on the luteinizing hormone surge in ovariectomized, estradioltreated rats is time-dependent. Brain Res., 901: 109–116. Pennartz, C.M.A., Bos, N.P.A., de Jeu, M.T.G., Geurtsen, A.M.S., Mirmiran, M., Sluiter, A.A. and Buijs, R.M. (1998a) Membrane properties and morphology of vasopressin neurons in slices of rat suprachiasmatic nucleus. J. Neurophysiol., 80: 2710–2717. Pennartz, C.M.A., de Jeu, M.T.G., Geurtsen, A.M.S., Sluiter, A.A. and Hermes, M.L.H.J. (1998b) Electrophysiological and morphological heterogeneity of neurons in slices of rat suprachiasmatic nucleus. J. Physiol. (Lond.), 506: 775–793. Pesonen, U., Jolkkonen, J., Huupponen, R. and Koulu, M. (1992) Differential hypothalamic arginine vasopressin response to

71 glucocorticoid receptor antagonism in lean and obese Zucker rats. Neuroendocrinology, 56: 271–277. Raggenbass, M. (2001) Vasopressin- and oxytocin-induced activity in the central nervous system: electrophysiological studies using in-vitro systems. Prog. Neurobiol., 64: 307–326. Reuss, S., Hurlbut, E.C., Speh, J.C. and Moore, R.Y. (1989) Immunohistochemical evidence for the presence of neuropeptides in the hypothalamic suprachiasmatic nucleus of ground squirrels. Anat. Rec., 225: 341–346. Rhodes, C.H., Morrell, J.I. and Pfaff, D.W. (1982) Estrogenconcentrating neurophysin-containing hypothalamic magnocellular neurons in the vasopressin-deficient (Brattleboro) rat: a study combining steroid autoradiography and immunocytochemistry. J. Neurosci., 12: 1718–1724. Rhodes, M.E. and Rubin, R.T. (1999) Functional sex differences (‘sexual diergism’) of central nervous system cholinergic systems, vasopressin, and hypothalamic–pituitary–adrenal axis activity in mammals: a selective review. Brain Res. Rev., 30: 135–152. Rivier, C. and Vale, W. (1983) Modulation of stressinduced ACTH release by corticotropin-releasing factor, catecholamines and vasopressin. Nature, 305: 325–327. Rivier, C., Riveir, J., Mormede, P. and Vale, W. (1984) Studies of the nature of the interaction between vasopressin and corticotropin-releasing factor on adrenocorticotropin release in the rat. Endocrinology, 115: 882–886. Romero, L.M., Plotsky, P.M. and Sapolsky, R.M. (1993) Patterns of adrenocorticotropin secretagogue release with hypoglycemia, novelty, and restraint after colchicine blockade of axonal transport. Endocrinology, 132: 199–204. Romijn, H.J., Sluiter, A.A., Pool, C.W., Wortel, J. and Buijs, R.M. (1996) Differences in colocalization between Fos and PHI, GRP, VIP and VP in neurons of the rat suprachiasmatic nucleus after a light stimulus during the phase delay versus the phase advance period of the night. J. Comp. Neurol., 372: 1–8. Romijn, H.J., Sluiter, A.A., Pool, C.W., Wortel, J. and Buijs, R.M. (1997) Evidence from confocal fluorescence microscopy for a dense, reciprocal innervation between AVP-, somatostatin-, VIP/PHI-, GRP- and VIP/PHI/GRPimmunoreactive neurons in the rat suprachiasmatic nucleus. Eur. J. Neurosci., 9: 2613–2623. Sar, M. and Stumpf, W.E. (1980) Simultaneous localization of [3 H]estradiol and neurophysin I or arginine vasopressin in hypothalamic neurons demonstrated by a combined technique of dry-mount autoradiography and immunohistochemistry. Neurosci. Lett., 17: 179–184. Sawchenko, P.E. (1987) Evidence for differential regulation of corticotropin-releasing factor and vasopressin immunoreactivities in parvocellular neurosecretory and autonomic-related projections of the paraventricular nucleus. Brain Res., 437: 253–263. Sawchenko, P.E., Swanson, L.W. and Vale, W.W. (1984) Coexpression of corticotropin-releasing factor and vasopressin immunoreactivity in parvocellular neurosecretory neurons of the adrenalectomized rat. Proc. Natl. Acad. Sci. USA, 81: 1883–1887.

Schwartz, W.J. and Reppert, S.M. (1985) Neural regulation of the circadian vasopressin rhythm in cerebrospinal fluid: A preeminent role for the suprachiasmatic nuclei. J. Neurosci., 5: 2771–2778. Schwartz, W.J., Coleman, R.J. and Reppert, S.M. (1983) A daily vasopressin rhythm in rat cerebrospinal fluid. Brain Res., 263: 105–112. Seggie, J., Shaw, B., Uhlir, I. and Brown, G.M. (1974) Baseline 24-hour plasma corticosterone rhythm in normal, shamoperated and septally-lesioned rats. Neuroendocrinology, 15: 51–61. Shughrue, P.J., Lane, M.V. and Merchenthaler, I. (1997) Comparative distribution of estrogen receptor-α and -β mRNA in the rat central nervous system. J. Comp. Neurol., 388: 507–525. Simerly, R.B., Chang, C., Muramatsu, M. and Swanson, L.W. (1990) Distribution of androgen and estrogen receptor mRNAcontaining cells in the rat brain: an in situ hybridization study. J. Comp. Neurol., 294: 76–95. Simonian, S.X. and Herbison, A.E. (1997) Differential expression of estrogen receptor α and β immunoreactivity by oxytocin neurons of rat paraventricular nucleus. J. Neuroendocrinol., 9: 803–806. Skowsky, W.R., Swan, L. and Smith, P. (1979) Effects of sex steroid hormones on arginine vasopressin in intact and castrated male and female rats. Endocrinology, 104: 105–108. Smale, L. and Boverhof, J. (1999) The suprachiasmatic nucleus and intergeniculate leaflet of Arvicanthis niloticus, a diurnal murid rodent from east Africa. J. Comp. Neurol., 403: 190– 208. Södersten, P., Henning, M., Melin, P. and Ludin, S. (1983) Vasopressin alters female sexual behaviour by acting on the brain independently of alterations in blood pressure. Nature, 301: 608–610. Södersten, P., De Vries, G.J., Buijs, R.M. and Melin, P. (1985) A daily rhythm in behavioral vasopressin sensitivity and brain vasopressin concentrations. Neurosci. Lett., 58: 37–41. Södersten, P., Boer, G.J., De Vries, G.J., Buijs, R.M. and Melin, P. (1986) Effects of vasopressin on female sexual behavior in male rat. Neurosci. Lett., 69: 188–191. Sofroniew, M.V. and Weindl, A. (1980) Identification of parvocellular vasopressin and neurophysin neurons in the suprachiasmatic nucleus of a variety of mammals including primates. J. Comp. Neurol., 193: 659–675. Stone, J.D., Crofton, J.T. and Share, L. (1989) Sex differences in central adrenergic control of vasopressin release. Am. J. Physiol., 257: R1040–R1045. Swaab, D.F. and Jongkind, J.F. (1970) The hypothalamic neurosecretory activity during the oestrous cycle, pregnancy, parturition, lactation, and persistent oestrus, and after gonadectomy, in the rat. Neuroendocrinology, 6: 133–145. Swaab, D.F. and Pool, C.W. (1975) Specificity of oxytocin and vasopressin immunofluorescence. J. Endocrinol., 66: 263–272. Swanson, L.W. (1977) Immunohistochemical evidence for a neurophysin-containing autonomic pathway arising in the paraventricular nucleus of the hypothalamus. Brain Res., 128: 346–353. Swanson, L.W. and Kuypers, H.G.J.M. (1980) The paraventric-

72 ular nucleus of the hypothalamus: Cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J. Comp. Neurol., 194: 555–570. Swanson, L.W., Sawchenko, P.E., Wiegand, S.J. and Price, J.L. (1980) Separate neurons in the paraventricular nucleus project to the median eminence and to the medulla or spinal cord. Brain Res., 198: 190–195. Swenson, K.L. and Sladek, C.D. (1997) Gonadal steroid modulation of vasopressin secretion in response to osmotic stimulation. Endocrinology, 138: 2089–2097. Szot, P. and Dorsa, D.M. (1994) Expression of cytoplasmic and nuclear vasopressin RNA Following castration and testosterone replacement — Evidence for transcriptional regulation. Mol. Cell. Neurosci., 5: 1–10. Thomas, A., Kim, N.B. and Amico, J.A. (1996) Differential regulation of oxytocin and vasopressin messenger ribonucleic acid levels by gonadal steroids in postpartum rats. Brain Res., 738: 48–52. Urban, J.H., Miller, M.A. and Dorsa, D.M. (1991) Dexamethasone-induced suppression of vasopressin gene expression in the bed nucleus of the stria terminalis and medial amygdala is mediated by changes in testosterone. Endocrinology, 129: 109–116. Van Der Beek, E.M., Wiegant, V.M., Van Der Donk, H.A., Van Den Hurk, R. and Buijs, R.M. (1993) Lesions of the suprachiasmatic nucleus indicate the presence of a direct vasoactive intestinal polypeptide-containing projection to gonadotrophinreleasing hormone neurons in the female rat. J. Neuroendocrinol., 5: 137–144. Van Der Beek, E.M., Horvath, T.L., Wiegant, V.M., Van Den Hurk, R. and Buijs, R.M. (1997) Evidence for a direct neuronal pathway from the suprachiasmatic nucleus to the gonadotropin-releasing hormone system: Combined tracing and light and electron microscopic immunocytochemical studies. J. Comp. Neurol., 384: 569–579. Van Eekelen, J.A.M., Rosenfeld, P., Levine, S., Westphal, H.M. and De Kloet, E.R. (1987) Postnatal disappearance of glucocorticoid receptor immunoreactivity in the suprachiasmatic nucleus of the rat. Neurosci. Res. Commun., 1: 129–133. Van Leeuwen, F.W. and Caffé, A.R. (1983) Vasopressinimmunoreactive cell bodies in the bed nucleus of the stria terminalis of the rat. Cell Tissue Res., 228: 525–534. Van Leeuwen, F.W., Caffé, A.R. and De Vries, G.J. (1984) 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. Vandesande, F. and Dierickx, K. (1975) Identification of the vasopressin producing and of the oxytocin producing neurons in the hypothalamic magnocellular neurosecretory system of the rat. Cell Tissue Res., 164: 153–162. Vandesande, F., De Mey, J. and Dierickx, K. (1974) Identification of neurophysin producing cells, I. The origin of the neurophysin-like substance-containing nerve fibres of the external region of the median eminence of the rat. Cell Tissue Res., 151: 187–200.

Vandesande, F., Dierickx, K. and De Mey, J. (1975) Identification of the vasopressin-neurophysin producing neurons of the rat suprachiasmatic nuclei. Cell Tissue Res., 156: 377–380. Viau, V. and Meaney, M.J. (1996) The inhibitory effect of testosterone on hypothalamic–pituitary–adrenal responses to stress is mediated by the medial preoptic area. J. Neurosci., 16: 1866–1876. Viau, V., Chu, A., Soriano, L. and Dallman, M.F. (1999) Independent and overlapping effects of corticosterone and testosterone on corticotropin-releasing hormone and arginine vasopressin mRNA expression in the paraventricular nucleus of the hypothalamus and stress-induced adrenocorticotropic hormone release. J. Neurosci., 19: 6684–6693. Viau, V., Soriano, L. and Dallman, M.F. (2001) Androgens alter corticotropin releasing hormone and arginine vasopressin mRNA within forebrain sites known to regulate activity in the hypothalamic–pituitary–adrenal axis. J. Neuroendocrinol., 13: 442–452. Vrang, N., Larsen, P.J. and Mikkelsen, J.D. (1995) Direct projection from the suprachiasmatic nucleus to hypophysiotrophic corticotropin-releasing factor immunoreactive cells in the paraventricular nucleus of the hypothalamus demonstrated by means of Phaseolus vulgaris-leucoagglutinin tract tracing. Brain Res., 684: 61–69. Wagner, C.K., Sisk, C.L. and Clemens, L.G. (1993) Neurons in the paraventricular nucleus of the hypothalamus that project to the sexually dimorphic lower lumbar spinal cord concentrate 3 H-estradiol in the male rat. J. Neuroendocrinol., 5: 545–551. Wang, Z.X., Bullock, N.A. and De Vries, G.J. (1993) Sexual differentiation of vasopressin projections of the bed nucleus of the stria terminals and medial amygdaloid nucleus in rats. Endocrinology, 132: 2299–2306. Wang, Z.X., Zhou, L., Hulihan, T.J. and Insel, T.R. (1996) Immunoreactivity of central vasopressin and oxytocin pathways in microtine rodents: A quantitative comparative study. J. Comp. Neurol., 366: 726–737. Watson, R.E., Langub, M.C., Engle, M.G. and Maley, B.E. (1995) Estrogen-receptive neurons in the anteroventral periventricular nucleus are synaptic targets of the suprachiasmatic nucleus and peri-suprachiasmatic region. Brain Res., 689: 254–264. Watts, A.G. and Swanson, L.W. (1987) Efferent projections of the suprachiasmatic nucleus, II. Studies using retrograde transport of fluorescent dyes and simultaneous peptide immunohistochemistry in the rat. J. Comp. Neurol., 258: 230–252. Watts, A.G. and Swanson, L.W. (1989) Diurnal variations in the content of preprocorticotropin-releasing hormone messenger ribonucleic acids in the hypothalamus paraventricular nucleus of rats of both sexes as measured by in situ hybridization. Endocrinology, 125: 1734–1738. Whitnall, M.H. (1988) Distributions of pro-vasopressin deficient CRH neurons in the paraventricular hypothalamic nucleus of colchicine-treated normal and adrenalectomized rats. J. Comp. Neurol., 275: 13–28. Wotjak, C.T., Kubota, M., Liebsch, G., Montkowski, A., Holsboer, F., Neumann, I. and Landgraf, R. (1996) Release of vasopressin within the rat paraventricular nucleus in response

73 to emotional stress: A novel mechanism of regulating adrenocorticotropic hormone secretion?. J. Neurosci., 16: 7725–7732. Yi, S.J., Masters, J.N. and Baram, T.Z. (1994) Glucocorticoid receptor mRNA ontogeny in the fetal and postnatal rat forebrain. Mol. Cell. Neurosci., 5: 385–393.

Zhou, L., Blaustein, J.D. and De Vries, G.J. (1994) Distribution of androgen receptor immunoreactivity in vasopressinand oxytocin-immunoreactive neurons in the male rat brain. Endocrinology, 134: 2622–2627.

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 6

Regulation of renal salt and water transporters during vasopressin escape Carolyn A. Ecelbarger 1,∗ , Takashi Murase 2, Ying Tian 1 , Soren Nielsen 3 , Mark A. Knepper 4 and Joseph G. Verbalis 1 1

Department of Medicine, Division of Endocrinology and Metabolism, Georgetown University, Washington, DC 20007, USA 2 The First Department of Internal Medicine, Nagoya University, Nagoya, Japan 3 Department of Cell Biology, University of Aarhus, Aarhus, Denmark 4 Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD, USA

Abstract: Hyponatremia, defined as a serum sodium < 135 mmol/l, is one of the most commonly encountered and serious electrolyte disorders of clinical medicine. The predominant cause of hyponatremia is an inappropriate elevation of circulating vasopressin levels relative to serum osmolality or the ‘syndrome of inappropriate antidiuretic hormone secretion’ (SIADH). Fortunately, the degree of the hyponatremia is limited by a process that counters the water-retaining action of vasopressin, namely ‘vasopressin escape’. Vasopressin escape is characterized by a sudden increase in urine volume with a decrease in urine osmolality independent of circulating vasopressin levels. Until recently, little was known about the molecular mechanisms underlying escape. In the 1980s, we developed an animal model for vasopressin escape in which male Sprague–Dawley rats were infused with dDAVP, a V2-receptor-selective agonist of vasopressin, while being fed a liquid diet. Rats drank a lot of water in order to get the calories they desired. Using this model, we demonstrated that the onset of vasopressin escape (increased urine volume coupled to decreased urine osmolality) coincided temporally with a marked decrease in renal aquaporin-2 (water channel) protein and mRNA expression in renal collecting ducts. This protein reduction was reversible and correlated to decreased water permeability of the collecting duct. Studies examining the mechanisms underlying AQP2 decrease revealed a decrease in V2-receptor mRNA expression and binding, as well as a decrease in cyclic AMP production in response to acute-dDAVP challenge in collecting duct suspensions from these escape animals. Additional studies showed an increase in sodium transporters of the distal tubule. These changes, hypothetically, might help to attenuate the hyponatremia. Future studies are needed to fully elucidate systemic, intra-organ, and cellular signaling responsible for the physiological phenomenon of vasopressin escape. Keywords: Aquaporins; SIADH; Thiazide-sensitive Na-Cl cotransporter; Epithelial sodium channel; Hyponatremia; Bumetanide-sensitive Na-K-2Cl cotransporter

∗ Correspondence to: C.A. Ecelbarger, Building D, Room 232, 4000 Reservoir Road NW, Georgetown University, Washington, DC 20007, USA. Tel.: +1-202-687-0453; Fax: +1-202-687-2040; E-mail: [email protected]

76

Background Hyponatremia and vasopressin escape Hyponatremia, defined as a serum sodium <135 mmol/l, is one of the most commonly encountered, at 15–30% of hospitalized patients (Flear et al., 1981; DeVita et al., 1990), and serious electrolyte disorders of clinical medicine. The predominant cause of hypoosmotic serum or hyponatremia is an inappropriate elevation of circulating vasopressin levels relative to serum osmolality or the ‘syndrome of inappropriate antidiuretic hormone secretion’ (SIADH). This condition has been associated with several disease states such as cirrhosis and congestive heart failure (Bichet et al., 1992). In addition, ectopic production of vasopressin-like substances can occur from tumors, e.g., olfactory neuroblastoma (Osterman et al., 1986). Drug-induced hyponatremia is also very common and can result from treatment with chlorpropamide, carbamazepine, diuretics, and some antineoplastic agents (Kinzie, 1987). Fortunately, the degree of the hyponatremia is limited by a process that counters the water-retaining action of vasopressin, namely ‘vasopressin escape’. Vasopressin escape is characterized by a sudden increase in urine volume with a decrease in urine osmolality independent of circulating vasopressin levels. Preceding the diuresis of vasopressin escape, a rapid natriuresis occurs. In the late 1950s, Levinsky et al. (1959) administered intravenous AVP and water to dogs and then measured GFR and urinary sodium and potassium excretion. They described a natriuresis which peaked almost immediately upon administration of the vasopressin and water and which was clearly independent of the diuresis. The shape of the natriuresis peak would suggest that this event could be considered biphasic, i.e., in the early phase sodium reabsorption is rapidly and dramatically decreased resulting in a nearly immediate natriuresis, in the late phase, sodium excretion returns to baseline levels as sodium reabsorption is normalized. Development of an animal model A good animal model of vasopressin escape has proved difficult to achieve. Animals that are directly

infused with AVP will simply decrease water intake (Verbalis and Drutarosky, 1988). Therefore, several complicated models were designed to overcome this problem (Chan, 1973; Gross and Anderson, 1982; Ayus et al., 1989). Generally they involved administration of vasopressin concurrent with either intermittent or continuous infusion of hypotonic fluids. These models had high mortality rates associated with them and generally required infusion of nonphysiological volumes of infusates. In the late 1980s, Verbalis and Drutarosky (1988) developed an animal model that more closely resembled the physiological characteristics of clinical SIADH. In this model, rats are subcutaneously implanted with osmotic minipumps to administer dDAVP (1-deamino-[8D -arginine]-vasopressin), a V2-selective vasopressin agonist, while being offered a diet high in water content (liquid-formula diet). The rats consume a high volume of water in order to get the calories they need. This model has virtually 0% mortality and low morbidity associated with it. The rats initially retain water and become rapidly hyponatremic. However, after about 2 days, they begin to ‘escape’ by excreting significantly higher volumes of increasingly more dilute urine. This model has been successfully used to study the adaptation of the brain to hyponatremia (Verbalis and Gullans, 1991) and rapid correction of hyponatremic states by elimination or accumulation of electrolytes and osmolytes, respectively (Verbalis and Gullans, 1993). Additional studies (Verbalis, 1994) examined contribution of water accumulation versus sodium loss to the hyponatremia, as well as differences between male and female rats in their responses to hyponatremia (Verbalis, 1993). Molecular mechanisms underlying vasopressin escape Until recently, little was known about the molecular mechanisms underlying escape. Several potential explanations have been offered over many years of study, including alterations in renal tubular epithelial permeability (Levinsky et al., 1959), dissipation of the corticomedullary osmotic gradient (Chan, 1973), renal hemodynamic changes as a result of increased renal artery pressure (Hall et al., 1986), and enhanced synthesis of renal prostaglandin E2 (Gross et al., 1983). However, many of these studies have

77

reported conflicting results and to date none of these proposed explanations has provided a comprehensive understanding of the cellular mechanisms underlying this process at the level of the renal tubule. Elegant studies, performed by Cowley and associates (Cowley et al., 1984) in which body weight and total body water were maintained constant by infusion using a servo-controlled system, demonstrated that volume expansion was necessary for vasopressin escape. In addition, Hall et al. (1986) were able to show that increased renal perfusion pressure, in particular, was a prerequisite for the escape. The signals arising from the increase in renal perfusion pressure that ultimately lead to renal escape are not yet clear. However, several of the key signaling pathways and down-stream proteins in the kidney that are important in mediating the escape are beginning to be elucidated. In particular, we (Ecelbarger et al., 1997, 1998; Murase et al., 1999) and others (Jonassen et al., 2000; Ishikawa et al., 2001) have demonstrated the critical role that down-regulation of the water channel protein aquaporin-2 plays in mediating the escape. Our studies, which further examine the molecular mechanisms underlying escape, are described in the subsequent sections. Methods For studies described here (Ecelbarger et al., 1997, 1998, 2001; Murase et al., 1998, 1999; Tian et al., 2000), male Sprague–Dawley rats were infused subcutaneously, by minipump, with dDAVP (5–20 ng/h) for 5–11 total days. After a 4-day baseline period, rats were divided into two groups; for 1–7 additional days, ‘control’ rats received dDAVP, dry diet and ad libitum water while ‘water-loaded’ rats received dDAVP and a high-water diet, i.e., either a liquid-formula diet (Verbalis, 1994; Ecelbarger et al., 1997; Murase et al., 1998, 1999; Tian et al., 2000) or a diet made with the same dry ingredients, but with a high volume of water added, and then gelled by mixing with heated agar (Ecelbarger et al., 1998, 2001). Water-loaded rats were offered 50–60 ml of water per day incorporated into the diet. The rats consumed most of the water on the first day and slightly lesser amounts on subsequent days.

Decreased expression of the renal water channel, aquaporin-2 A significant increase in urine volume and decrease in urine osmolality in the water-loaded rats was observed by the second day of water loading, indicating onset of vasopressin escape (Ecelbarger et al., 1997, 1998, 2001; Murase et al., 1998; Tian et al., 2000). Furthermore, water-loaded rats developed a marked hyponatremia (plasma sodium range approximately 98–122 mmol/l) (Ecelbarger et al., 1997, 2001). The onset of escape coincided temporally with a marked decrease in renal aquaporin-2 protein as measured by semiquantitative immunoblotting (Fig. 1A) (Ecelbarger et al., 1997), as well as decreased mRNA expression, as assessed by Northern blotting (Fig. 1B) (Ecelbarger et al., 1997). Furthermore, immunohistocytochemical analysis (Fig. 2) and differential centrifugation studies (Ecelbarger et al., 1997) demonstrated that trafficking of aquaporin-2 to the plasma membrane remained intact during vasopressin escape, therefore, was unlikely to contribute to the escape. In contrast to aquaporin2, there were no decreases in the renal expression of aquaporins 1, 3, or 4 (Ecelbarger et al., 1997). In fact, aquaporin-3 abundance was significantly increased in the water-loaded rats (Ecelbarger et al., 1997, 2001). These results suggest that escape from vasopressin-induced antidiuresis is attributable, at least in part, to a selective vasopressin-independent decrease in aquaporin-2 water channel expression in the renal collecting duct. In contrast, the escape phenomenon does not appear to be associated with an impairment of the short-term process by which vasopressin regulates collecting duct water permeability, namely vasopressin-induced trafficking of aquaporin-2 to the apical plasma membrane. Aquaporin-2 is the only water channel known to be present in the apical plasma membrane of the collecting duct principal cells. Since the apical plasma membrane is the rate-limiting barrier for transepithelial water transport (Flamion and Spring, 1990), the fall in aquaporin-2 expression seen with vasopressin escape would be predicted to correspond to a marked decrease in collecting duct water permeability. In fact, this was found to be the case. In subsequent studies (Ecelbarger et al., 1998) we measured the osmotic water permeability of inner

78

Fig. 1. Aquaporin-2 expression in whole kidney homogenates after 3 days of water-loading during vasopressin escape (Ecelbarger et al., 1997). Each lane is loaded with a sample from a different animal (n = 6 rats/treatment). (A) Aquaporin-2 protein blot is probed with anti-aquaporin-2 antibody L127 (Nielsen et al., 1993). Bands at 29 and 35–40 kDa are specific bands for aquaporin-2. (B) Aquaporin-2 mRNA. A significant decrease in AQP2 protein abundance and expression of aquaporin-2 mRNA was observed (unpaired t-test, P < 0.05 considered significant).

medullary collecting ducts (IMCDs) from six waterloaded rats and six controls using the isolated perfused tubule technique (Burg et al., 1966). We found that, with 100 nM vasopressin in the peritubular bath, the osmotic water permeabilities of the IMCDs from the water-loaded rats were reduced on average to 46% of their corresponding controls (waterloaded, 240 ± 25 μm/s; control, 527 ± 63 μm/s; P < 0.002). In our initial studies, the escaped animals were also markedly hypoosmolar compared to controls as a result of water loading during sustained antidiuresis. This raised the possibility that one factor responsible for the down-regulation of aquaporin-2 during escape might be a decrease in osmolality, either systemically or interstitially, in the kidney. Subsequent studies (Murase et al., 1999), therefore, evaluated whether changes in systemic or local osmolality contributed to the down-regulation of aquaporin-2 expression in this model. In the first set of studies, two groups of dDAVP-infused rats were water-loaded, but after establishment of escape, one group was then water-restricted for 1–4 days to reverse the es-

cape, whereas the other groups continued the water loading. Fig. 3 shows a summary of immunoblot density data acquired from escape rats (either subsequently water-restricted or maintained on the high water diet) sacrificed after 1, 2, 3, or 4 days as compared to dDAVP-treated controls. Aquaporin-2 abundance in the whole kidney returned to control levels after 4 days of water restriction. However, rats remained markedly hyponatremic: 115 ± 4 mmol/l on day 4 versus 107 ± 1 mmol/l before the start of water restriction. Thus, serum hypoosmolality did not appear to be an important mediator of the downregulation of aquaporin-2 abundance in the kidney. In another group of rats, we (Murase et al., 1999) were also able to deduce that renal interstitial tissue osmolality, likewise, did not appear to be important in the regulation of aquaporin-2 abundance. We examined the regulation of aquaporin-2 during vasopressin escape in the three major areas of the kidney. Despite very different interstitial osmolalities, significant down-regulation of aquaporin-2 abundance in the water-loaded rats compared to dDAVP-infused controls was seen in all regions of the kidney: cor-

Fig. 2. Distribution of aquaporin-2 in collecting duct principal cells during vasopressin escape (Ecelbarger et al., 1997). (A) Immunoperoxidase labeling of thin cryosections from rat inner medullas of dDAVP-treated control rats. (B,C) Labeling in dDAVP-treated rats after water loading for 3 days. Note the reduced aquaporin-2 labeling in collecting ducts of water-loaded rats. In collecting duct cells from water-loaded rats (B,C), aquaporin-2 labeling was chiefly associated with the apical plasma membrane domain and vesicles in the immediate subapical region, consistent with the expected action of dDAVP to shuttle aquaporin-2 to the apical plasma membrane domain. Thus, vasopressin escape was apparently not due to a failure of dDAVP to induce trafficking of aquaporin-2 vesicles to the apical region. (D) Immunolabeling control. Use of non-immune IgG as primary antibody reveals a complete absence of labeling.

79

80

Fig. 3. Summary of immunoblotting data for aquaporin-2 protein in rats after the initiation of vasopressin escape followed by 1 to 4 days of water-restriction (open circles) or continual high water diet (water-loaded) (solid circles). Data are plotted as a percent of control rats (dDAVP-treated, dry diet) mean values (Murase et al., 1999).

tex (32 ± 2% of control), outer medulla (70 ± 9% of control), inner medulla (67 ± 6% of control). Decreased expression and binding of vasopressin V2 receptors during escape The decrease in aquaporin-2 expression in the renal collecting ducts from the water-loaded, i.e. ‘vasopressin-escape’, rats is likely to be due in part to decreased signaling through the vasopressin V2 receptor. Tian et al. (2000) found a marked reduction in vasopressin V2-receptor binding in the inner medulla of waterloaded (vasopressin-escape) rats (Fig. 4). In these studies inner medullary homogenates were prepared from untreated rats, control and waterloaded rats. V2 binding capacity was assessed by incubating the samples with a radioiodinated V2 antagonist, d(CH2 )5 [D-Ile2 ,Ile4 ,Tyr-NH92 ]AVP. dDAVP infusion alone decreased V2-receptor binding capacity (Bmax ) to 81% of untreated levels after 7 days of dDAVP infusion (n = 8, P < 0.05). Furthermore, by the third day of water-loading, vasopressin V2receptor binding decreased to 43% of the dDAVPcontrol rats (n = 10, P < 0.01). No differences were observed between the three treatment groups for binding affinity (K d). Complementary Northern-blotting studies by Murase et al. (1998) also demonstrated

decreased expression of V2-receptor mRNA in the water-loaded rats. In support of these findings, we demonstrated, in parallel studies (Ecelbarger et al., 1998) a decrease in cyclic AMP generating capacity in inner medullary collecting duct (IMCD) suspensions prepared from water-loaded rats (Fig. 5). IMCD suspensions were prepared from six water-loaded (3-day) rats and six corresponding controls. An individual suspension was prepared for each animal, and 50-μl aliquots of each rat’s suspension were incubated under eight conditions: 0, 0.1, 1.0, and 10 nM dDAVP in the presence or absence of 0.25 mM IBMX (isobutylmethylxanthine), a phosphodiesterase inhibitor. After 5 min at 37°C, the incubation was terminated by adding 200 μl of 10% trichloroacetic acid and placing the sample tube on ice. Cyclic AMP levels in the supernatant were determined by a radioimmunoassay. Protein content of each suspension aliquot was measured for normalization. Whether or not IBMX was included in the incubation solution, there was a significant reduction in the cyclic AMP accumulation in tubules from waterloaded (vasopressin-escape) rats. Because the difference between the treatments was no less apparent in the presence of the phosphodiesterase inhibitor, IBMX, we concluded that the collecting ducts from

81

Fig. 4. Effect of vasopressin escape on V2 vasopressin receptor binding in the inner medulla of rats after 3 days of water-loading (7 days of dDAVP infusion). Circles = untreated rats; triangles = control/dDAVP-treated; squares = water-loaded/dDAVP-treated (Tian et al., 2000). Water loading led to a decrease in V2 vasopressin binding capacity.

Fig. 5. Cyclic AMP production during vasopressin escape in inner medullary collecting duct suspensions treated with IBMX after 3 days of water loading (Ecelbarger et al., 1998). Circles = controls; dots = water-loaded. Water loading resulted in significantly decreased cyclic AMP accumulation in the suspension fluid after the 5-min incubation period.

the vasopressin-escape rats have a decreased ability to produce cyclic AMP, i.e., a relative vasopressin resistance. Our observation of decreased cyclic AMP generating capacity of the inner medullary collecting duct from rats undergoing vasopressin escape, as well

as the reduced expression of aquaporin-2, might be explained by the measured reduction in V2-receptor expression and binding capacity. However, additional unknown mechanisms, both V2-receptor-dependent or V2-receptor-independent, may play a role.

82

Fig. 6. Abundance of renal distal sodium transporters and channels as assessed by semiquantitative immunoblotting after 2 days of water-loading or vehicle. (A) Immunoblot loaded with whole kidney homogenate samples probed with antithiazide-sensitive Na-Cl cotransporter (NCC) antibody L573 (Kim et al., 1998). Each lane is loaded with a sample from a different animal (n = 6 rats/treatment). (B) Summary of densitometries obtained from similar blots over the time course of escape. (C) Immunoblot of whole kidney homogenates from the same rats as in (A) probed with anti-epithelial sodium channel (ENaC) antibody (α-subunit) L766 (Masilamani et al., 1999). (D) Summary of densitometry over time course. Water loading increased the abundance of both NCC and α-ENaC (Ecelbarger et al., 2001).

Increased abundance of renal distal sodium transporters More recent studies (Ecelbarger et al., 2001) have shown that several distal sodium reabsorptive mechanisms are upregulated during vasopressin escape. The abundances of the thiazide-sensitive NaCl cotransporter of the distal convoluted tubule (Fig. 6A,B), the α-subunit of the epithelial sodium channel (ENaC) of the collecting duct (Fig. 6C,D) and the 70-kDa band of the γ-subunit of ENaC (Fig. 7) were all significantly increased in whole kidney homogenates from the water-loaded rats. No changes in abundance were observed for the βsubunit of the epithelial sodium channel. Similar protein changes have recently been associated with

elevated aldosterone levels in rats (Kim et al., 1998; Masilamani et al., 1999). However, plasma aldosterone levels were significantly suppressed in this model (Ecelbarger et al., 2001). These results are not surprising and are, in fact, in agreement with what many others have observed (Fichman et al., 1974; Craven et al., 1986; Ishikawa et al., 1996). On the other hand, vasopressin escape may lead to an aldosterone-like pattern of sodium transporter and channel protein abundances by somehow affecting 11-β-dehydroxysteroid dehydrogenase-2 (11-βHSD-2) activity, the enzyme responsible for cortisol degradation in these cells. Increased sodium load to the distal nephron as a result of the early natriuresis might also play a role in the upregulation of distal sodium transporters, since increased sodium load

83

Fig. 7. Immunoblot of whole kidney homogenates from 6 control and 6 water-loaded rats probed with anti-epithelial sodium channel antibody (γ-subunit) L550 (Masilamani et al., 1999) after 2 days of treatment. Water loading resulted in the appearance of a second band on immunoblots for γ-ENaC at around 70 kDa (Ecelbarger et al., 2001).

has been shown to increase the sodium transport capacity of the distal tubule (Stanton and Kaissling, 1988). Regardless of the mechanism, these protein changes would be predicted to result in increased sodium reabsorptive capacity of the distal convoluted tubule and collecting duct during vasopressin escape, which could potentially ameliorate the resulting hyponatremia. Thus, the changes in distal sodium transporters and channels that we observed are more likely to represent adaptive responses that allow for the conservation of sodium. These responses most likely occur as a result of either increased sodium load delivered to the distal tubule and/or other factors that arise from perhaps hyponatremia that activate mineralocorticoid-like mechanisms.

Abbreviations SIADH

syndrome of inappropriate antidiuretic hormone secretion AVP arginine vasopressin dDAVP 1-deamino-[8-D-arginine]vasopressin V2 receptor vasopressin receptor subtype 2 AQP aquaporin IMCD inner medullary collecting duct IBMX isobutylmethylxanthine cAMP cyclic adenosine monophosphate 11-β-HSD-2 11-β-dehydroxysteroid dehydrogenase ENaC epithelial sodium channel NCC thiazide-sensitive Na-Cl cotransporter V2RA vasopressin receptor antagonist

Summary and conclusions References Overall, the abundances of several critical renal salt and water transporters and channels are regulated during the physiological phenomenon of vasopressin escape. The down-regulation of aquaporin-2 protein is likely to be a primary means for the onset and maintenance of physiological vasopressin escape. A reduction in V2-receptor binding capacity may facilitate decreased cyclic AMP generation and decreased aquaporin-2 expression. Regulation of aquaporin-2 protein abundance as well as changes in sodium transporter expression would be predicted to modify the transport capacity of various affected renal tubules, such as the collecting duct, distal convoluted tubule and thick ascending limb. These adaptations likely are critical in allowing the animal to maintain salt and water homeostasis when vasopressin levels are inappropriately elevated.

Ayus, J.C., Krothapalli, R.K., Armstrong, D.L. and Norton, H.J. (1989) Sympathetic hyponatremia in rats: effect of treatment on mortality and brain lesions. Am. J. Physiol., 257: F18–F22. Bichet, D.G., Kluge, R., Howard, R.L. and Schrier, R.W. (1992) Hyponatremic states. In: D.W. Seldin and G. Giebisch (Eds.), The Kidney: Physiology and Pathophysiology. Raven, New York, NY, pp. 1727–1751. Burg, M.B., Grantham, J.J., Abramow, M. and Orloff, J. (1966) Preparation and study of fragments of single rabbit nephrons. Am. J. Physiol., 210: 1293–1298. Chan, W.Y. (1973) A study of the mechanism of vasopressin escape: effects of chronic vasopressin and overhydration on renal tissue osmolality and electrolytes in dogs. J. Pharmacol. Exp. Ther., 184: 244–252. Cowley Jr., A.W., Merrill, D.C., Quillen Jr., E.W. and Skelton, M.M. (1984) Long-term blood pressure and metabolic effects of vasopressin with servo-controlled fluid volume. Am. J. Physiol., 247(3 Pt 2): R537–R545. Craven, P.A., Verbalis, J.G. and DeRubertis, F.R. (1986) In-

84 creased urinary excretion of PGE2 during inappropriate antidiuresis induced by DDAVP. Kidney Int., 29: 1110–1115. DeVita, M.V., Gardenswartz, M.H., Konecky, A. and Zabetakis, P.M. (1990) Incidence and etiology of hyponatremia in an intensive care unit. Clin. Nephrol., 34: 163–166. Ecelbarger, C.A., Nielsen, S., Olson, B., Murase, T., Baker, E.A., Knepper, M.A. and Verbalis, J.G. (1997) Role of renal aquaporins in escape from vasopressin-induced antidiuresis in rat. J. Clin. Invest., 99(8): 1852–1863. Ecelbarger, C.A., Chou, C.-L., Lee, A.J., DiGiovanni, S.R., Verbalis, J.G. and Knepper, M.A. (1998) Escape from vasopressin-induced antidiuresis: role of vasopressin resistance of the collecting duct. Am. J. Physiol., 274: F1161–F1166. Ecelbarger, C.A., Verbalis, J.G. and Knepper, M.A. (2001) Increased abundance of distal sodium transporters in rat kidney during vasopressin escape. J. Am. Soc. Nephrol., 12: 207–217. Fichman, M.P., Michaelakis, A.M. and Horton, R. (1974) Regulation of aldosterone in the syndrome of inappropriate antidiuretic hormone secretion (SIADH). J. Clin. Endocrinol. Metab., 39: 136–144. Flamion, B. and Spring, K.R. (1990) Water permeability of apical and basolateral cell membranes of rat inner medullary collecting duct. Am. J. Physiol., 259: F986–F999. Flear, C.T., Gill, G.V. and Burn, J. (1981) Hyponatremia: mechanisms and management. Lancet, 2: 26–31. Gross, P.A. and Anderson, R.J. (1982) Effects of DDAVP and AVP on sodium and water balance in conscious rats. Am. J. Physiol., 243: R512–R591. Gross, P.A., Kim, J.K. and Anderson, R.J. (1983) Mechanisms of escape from desmopressin in the rat. Circ. Res., 53: 794–804. Hall, J.E., Montani, J.P., Woods, L.L. and Mizelle, H.L. (1986) Renal escape from vasopressin: role of pressure diuresis. Am. J. Physiol., 250(5 Pt 2): F907–F916. Ishikawa, S., Fujita, N., Fujisawa, G., Tsuboi, Y., Sakuma, N., Okada, K. and Saito, T. (1996) Involvement of arginine vasopressin and renal sodium handling in pathogenesis of hyponatremia in elderly patients. Endocrinol. J., 43: 101–108. Ishikawa, S., Saito, T., Fukagawa, A., Higashiyama, M., Nakamura, T., Kusaka, I., Nagasaka, S., Honda, K. and Saito, T. (2001) Close association of urinary excretion of aquaporin2 with appropriate and inappropriate arginine vasopressindependent antidiuresis in hyponatremia in elderly subjects. J. Clin. Endocrinol. Metab., 86(4): 1665–1671. Jonassen, T.E., Christensen, S., Kwon, T.H., Langhoff, S., Salling, N. and Nielsen, S. (2000) Renal water handling in rats with decompensated liver cirrhosis. Am. J. Physiol., 279(6): F1101–F1109. Kim, G.-H., Masilamani, S., Turner, R., Mitchell, C., Wade, J.B. and Knepper, M.A. (1998) The thiazide-sensitive Na-Cl cotransporter is an aldosterone-induced protein. Proc. Natl. Acad. Sci. USA, 95: 14552–14557.

Kinzie, B.J. (1987) Management of the syndrome of inappropriate secretion of antidiuretic hormone. Clin. Pharmacol., 6(8): 625–633. Levinsky, N.O., Davidson, D.G. and Berliner, R.W. (1959) Changes in urine concentration during prolonged administration of vasopressin and water. Am. J. Physiol., 196(2): 451– 456. Masilamani, S., Kim, G.-H., Mitchell, C., Wade, J.B. and Knepper, M.A. (1999) Aldosterone-mediated regulation of ENaC α, β and γ subunit proteins in rat kidney. J. Clin. Invest., 104: R19–R23. Murase, T., Baker, E.A., Tian, Y., Knepper, M.A. and Verbalis, J.G. (1998) Down-regulation of kidney AVP V2 receptor expression during renal escape from vasopressin-induced antidiuresis [Abstr.]. J. Am. Soc. Nephrol., 9: 23A. Murase, T., Ecelbarger, C.A., Baker, E.A., Tian, Y., Knepper, M.A. and Verbalis, J.G. (1999) Kidney aquaporin-2 expression during escape from antidiuresis is not related to plasma or tissue osmolality. J. Am. Soc. Nephrol., 10(10): 2067–2075. Nielsen, S., DiGiovanni, S.R., Christensen, E.I., Knepper, M.A. and Harris, H.W. (1993) Cellular and subcellular immunolocalization of vasopressin-regulated water channel in rat kidney. Proc. Natl. Acad. Sci. USA, 90: 1663–11667. Osterman, J., Calhoun, A., Dunham, M., Cullum Jr., U.X., Clark, R.M., Stewart, D.D., Scheithauer, B.W., Zimmerman, E.A., Defendini, R., Zang, X. and Robinson, A.G. (1986) Chronic syndrome of inappropriate antidiuretic hormone secretion and hypertension in a patient with olfactory neuroblastoma. Arch. Intern. Med., 146: 1731–1735. Stanton, B.A. and Kaissling, B. (1988) Adaptation of distal tubule and collecting duct to increased Na delivery, II. Na+ and K+ transport. Am. J. Physiol., 255: F1269–F1275. Tian, Y., Sandberg, K., Murase, T., Baker, E.A., Speth, R.C. and Verbalis, J.G. (2000) Vasopressin V2 receptor binding is down-regulated during renal escape from vasopressin-induced antidiuresis. Endocrinology, 141: 307–314. Verbalis, J.G. (1993) Hyponatremia induced by vasopressin or desmopressin in female and male rats. J. Am. Soc. Nephrol., 3: 1600–1606. Verbalis, J.G. (1994) Pathogenesis of hyponatremia in an experimental model of the syndrome of inappropriate antidiuresis. Am. J. Physiol., 267: R1617–R1625. Verbalis, J.G. and Drutarosky, M.D. (1988) Adaptation to chronic hypoosmolality in rats. Kidney Int., 34: 351–360. Verbalis, J.G. and Gullans, S.R. (1991) Hyponatremia causes large sustained reductions in brain content of multiple organic osmolytes in rats. Brain Res., 567(2): 274–282. Verbalis, J.G. and Gullans, S.R. (1993) Rapid correction of hyponatremia produces differential effects on brain osmolyte and electrolyte reaccumulation in rats. Brain Res., 606(1): 19– 27.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 7

Stretch-inactivated cation channels: cellular targets for modulation of osmosensitivity in supraoptic neurons Charles W. Bourque 1,∗ , Daniel L. Voisin 2 and Yassar Chakfe 3 1

Centre for Research in Neuroscience, Montreal General Hospital and McGill University, 1650 Cedar Avenue, Montreal, QC H3G 1A4, Canada 2 Laboratoire de Physiologie Oro-Faciale, Faculté de Chirurgie Dentaire, Clermont Ferrand, 63000, France 3 Neurobiology Unit, Montreal Neurological Institute, 3801 University Street, Montreal, QC H3A 2B4 Canada

Abstract: Rat magnocellular neurosecretory cells (MNCs) show an intrinsic sensitivity to acute changes in fluid osmolality. Experiments in acutely isolated supraoptic MNCs have shown that these responses are due to in part to the cell volumedependent modulation of gadolinium-sensitive 33 pS stretch-inactivated cation (SIC) channels. Previous studies in vivo have shown that the slope (i.e. gain) of the ‘osmosensory’ relation between VP release and plasma osmolality can be increased or decreased under various physiological and pathological conditions. Here, we review recent work that shows how changes in external [Na] and excitatory neuropeptides such as angiotensin II (Ang II), cholecystokinin (CCK) and neurotensin (NT), may influence osmosensory gain in acutely isolated MNCs. Whole-cell and single-channel recording experiments have revealed that changes in external Na cause proportional changes in osmosensory gain as a result of modified SIC channel permeability and not by affecting mechanotransduction. In contrast, Ang II, CCK, or NT appear to convergently, and directly, stimulate the osmosensory cation conductance in MNCs. Preliminary analysis in current clamp further suggests that osmosensory gain may be increased upon exposure to these excitatory peptides. Whether such mechanisms contribute to the modulation of osmosensory gain in vivo remains to be established. Keywords: Supraoptic nucleus; Peptide; Angiotensin II; Modulation; Sodium; Osmosensitivity; Electrophysiology; Vasopressin

Introduction Magnocellular neurosecretory cells (MNCs) in the supraoptic and paraventricular nuclei of the hypothalamus synthesize either oxytocin (OT) or vasopressin (VP). These hormones are involved in the control of a variety of physiological functions including, but not limited to: lactation (OT), parturition (OT), natriuresis (OT) and diuresis (VP) (Poulain and Wakerley,

∗ Correspondence to: C. Bourque, L7-216, Neurology Division, Montreal General Hospital, 1650 Cedar Avenue, Montreal QC H3G 1A4, Canada. Tel.: +1-514- 934-8094; Fax: +1-514-934-8265; E-mail: [email protected]

1982; Bourque and Renaud, 1990). Following synthesis, the peptide product of each particular cell is packaged into secretory vesicles that are then transported through the axon to numerous nerve endings in the neurohypophysis. Here, peptide secretion is mediated by Ca2+ -dependent exocytosis subsequent to voltage-gated Ca2+ influx triggered by the discharge of action potentials (Bicknell, 1988; Bourque, 1991; Fisher and Bourque, 1996). Since the neurohypophysial nerve endings cannot fire repetitively during sustained depolarization of the local membrane (Bourque, 1990), secretion normally requires the arrival into the terminals of action potentials individually initiated in the soma. Thus, as for other types of neuron, changes in secretion at the nerve

86

endings of MNCs are mediated through changes in somatically generated electrical activity (Renaud and Bourque, 1991). Classic electrophysiological studies in vivo have revealed essential information concerning the physiological and pathological conditions under which changes in the electrical and secretory activity of MNCs come about (for review see Poulain and Wakerley, 1982; Bourque and Renaud, 1990). A challenge that remains, however, is to define the anatomical pathways, as well as molecular and cellular mechanisms, that underlie the regulation of MNCs during physiological perturbation. Moreover, it is recognized that certain physiological stimuli can modulate (i.e. increase or decrease) the responsiveness of MNCs to other challenges in ways that enhance homeostasis (Chakfe and Bourque, 2001). A description of the central mechanisms regulating the activity of MNCs, therefore, must also account for the integration and interaction of signals derived from multiple sensory systems. In this review we provide an overview of the mechanisms though which changes in CSF [Na2+ ]o and synaptically released neuropeptides may contribute to the modulation of osmotically regulated VP secretion. We begin by reviewing some of the mechanisms known to underlie the osmotic regulation of MNCs. The electrical activity of VP MNCs is proportional to plasma osmolality Under isotonic conditions, VP-releasing MNCs in the hypothalamus typically fire action potentials at a basal rate of 1–3 Hz (Poulain and Wakerley, 1982). In agreement with this observation, VP levels of 1–5 pg/ml can be detected in the blood of normal rats under isotonic conditions (Dunn et al., 1973). These concentrations lie near the middle zone of the dose–response sensitivity of the renal concentrating mechanism and are therefore sufficient to maintain functionally relevant levels of antidiuresis under basal conditions (Robertson et al., 1976). When plasma osmolality is reduced to values below the set-point (∼295 mosmol/kg in the rat), the firing rate of MNCs diminishes and plasma VP concentration falls below the limit of detection by radioimmunoassay (Bourque, 1998). The resulting diuresis immediately produces a homeostatic enhancement of

plasma osmolality. Conversely, if plasma osmolality rises above the set-point, the firing rate of MNCs increases proportionally, as does the concentration of VP in blood (Bourque, 1998). The resultant enhancement of water retention helps restore the osmotic pressure of the plasma toward set-point as additional water is ingested. The bi-directional osmotic control of VP-MNCs, therefore, represents a key mechanism for the maintenance of extracellular fluid osmolality (Bourque et al., 1994). The osmosensitivity of MNCs involves extrinsic and intrinsic factors Work performed previously has shown that the osmotic regulation of MNCs is dependent on intrinsic and extrinsic factors (Bourque, 1998). The extrinsic influences comprise at least two parts. First, monoand polysynaptic inputs originating from peripheral and central osmoreceptors are known to impinge upon MNCs to bias the rate of electrical activity through the release of neurotransmitters (Leng et al., 1989; Bourque et al., 1994). Axonal projections from osmosensitive neurons in the organum vasculosum lamina terminalis (OVLT), for example, have been shown to regulate MNCs through changes in the frequency of glutamatergic excitatory postsynaptic potentials (EPSPs) generated in MNCs of the supraoptic nucleus (Richard and Bourque, 1995, 2001). This control is exerted over the entire dynamic range of osmotic pressures as hypotonic stimuli applied to the OVLT reduce, whereas hypertonic stimuli enhance, the basal frequency of spontaneous EPSPs. A tonic excitatory synaptic drive originating from osmosensitive OVLT neurons therefore contributes both to the basal activity of MNCs and to osmotically evoked changes in firing rate. It is quite likely that excitatory and inhibitory inputs originating from osmosensitive neurons in other regions also contribute to the osmotic control of MNCs. A second extrinsic mechanism involves the glial cells that are intimately associated with MNCs in the hypothalamus. Work by Hussy et al. (2000) has revealed that hypoosmotic challenges provoke a robust and sustained release of taurine from the astrocytes surrounding MNCs (Deleuze et al., 1998). This response is sensitive to blockers of anion channels and may therefore be mediated through volume-regulated

87

anion channels (Bres et al., 2000). Interestingly, hyperosmotic stimuli were found to suppress basal release, suggesting that the release of taurine is regulated bi-directionally around the osmotic set point. The involvement of taurine in the control of electrical activity derives from the fact that at the concentrations involved, it behaves as a potent agonist of the strychnine-sensitive glycine receptors expressed in MNCs (Hussy et al., 1997). Since activated glycine receptors normally inhibit MNCs through the elevation of Cl− conductance, hyperosmotic stimuli depolarize MNCs via a deactivation of tonically engaged glycine receptors, whereas hypotonic stimuli hyperpolarize MNCs through the increased activation of glycine receptors. The combined actions of extrinsic factors is believed to play an important role in mediating the osmotic control of MNCs in situ (Hussy et al., 2000). Stretch-inactivated cation (SIC) channels mediate the intrinsic osmosensitivity of MNCs Experiments in hypothalamic slices provided the first evidence for the existence of an intrinsic osmosensory mechanism in MNCs. Indeed, Mason (1980) originally reported that MNCs were depolarized in response to hyperosmotic stimuli and that this response was retained in low Ca2+ , high Mg2+ solutions. Since such solutions block synaptic transmission, it was presumed that the response was generated through an endogenous mechanism. However, the recently described taurine-dependent mechanism mentioned above might well be retained in hypothalamic slices and, therefore, might have been involved in the generation of depolarizing responses in low Ca2+ high Mg2+ containing solutions. Although this issue remains to be investigated, further evidence for the existence of an intrinsic mechanism was provided by recordings from MNCs in hypothalamic explants which revealed that the hyperosmotic depolarization was associated with an increase of membrane conductance and the generation of a non-selective cation current (Bourque, 1989). Even more direct evidence for the existence of an intrinsic mechanism was ultimately provided when Oliet and Bourque (1992) revealed that MNCs acutely isolated from any synaptic or glial contacts can display depolarizing and hyperpolarizing responses to hypertonic and

hypotonic stimuli. Subsequent experiments involving imaging of cell volume and whole-cell voltageclamp measurements in this preparation established that these responses were due to the bi-directional, cell-volume dependent modulation of a non-selective cation conductance (Oliet and Bourque, 1993a,b, 1996). Single-channel analysis revealed that these responses were due to the presence, in MNCs, of 33 pS Gd3+ -sensitive SIC channels. Because these channels display a basal activity at rest, cell shrinking in the presence of hypertonic solutions disinhibits channel activity, which leads to depolarization. Conversely, cell swelling induced by hypoosmotic solutions inhibits basal channel activity and leads to hyperpolarization. It is estimated from the average maximal conductance decrease provoked by strongly hypotonic solutions, by direct cell inflation, or by the application of Gd3+ , that basally active SIC channels contribute as much as 50% of the resting membrane input conductance of acutely isolated MNCs under isotonic conditions. Not surprisingly, therefore, changes in the probability of opening (Po ) of these channels have been found to have an important impact on membrane potential during acute osmotic stimulation. Thus, in concert with the effects of the extrinsically generated responses described above, cell-volume mediated changes in SIC channel activity may play an important role in the osmotic regulation of MNCs in situ (Bourque and Oliet, 1997; Bourque, 1998). The relation between VP secretion and osmolality is subject to modulation While osmotic perturbations play a dominant role in the regulation of VP-releasing MNCs (and OT MNCs in the rat), a number of studies have shown that osmotically evoked VP secretion can be modulated by a variety of other stimuli (Chakfe and Bourque, 2001). Well documented examples of this sort of interaction include the effects of CSF sodium concentration ([Na+ ]o ), or changes in systemic blood pressure or volume, on the osmotic regulation of VP secretion. Indeed, concerted increases or decreases in CSF [Na+ ]o respectively enhance or blunt the secretory VP response to a constant hypertonic challenge (McKinley et al., 1978). Analogously, decreases or increases in total blood volume respec-

88

tively enhance and attenuate the slope of the relation between plasma VP concentration and plasma osmolality (Dunn et al., 1973). Although these responses are physiologically and pathologically important, little is known of the neural basis by which they are achieved. Since SIC channels endogenous to MNCs appear to play an important role in the control of the membrane potential and in the osmosensitivity of MNCs, we recently examined whether these particular channels might serve as targets for the modulation of osmotically regulated VP release. SIC channels operate as intrinsic Na+ sensors Since SIC channels are permeant to both Na+ and K+ ions (Oliet and Bourque, 1993a,b), we reasoned that increases in [Na+ ]o might significantly enhance the amplitude of the inward current evoked by hyperosmotic stimulation at a fixed voltage through an increase of the driving force (Voisin et al., 1999). Conversely, a reduction in [Na+ ]o might attenuate current responses through a decrease of the driving force across the channels. We thus tested this hypothesis by measuring the reversal potential (E rev ) of the membrane current evoked in acutely isolated MNCs by a standard stimulus of +30 mosmol/kg (delivered as excess mannitol) in the presence of solutions comprising different [Na+ ]o . Although values of E rev indeed became more positive as [Na+ ]o was increased, the recordings indicated that changes in E rev were more important than those predicted by the Goldman equation for a conductance featuring a constant ratio of sodium to potassium permeability (PNa /PK ). In fact, the results indicated that under the conditions of these experiments (i.e. [Na+ ]i , [K+ ]i and [K]o = 3, 150 and 3 mM, respectively), PNa /PK increased from a value near 0.18 at 105 mM [Na+ ]o to a value of 0.32 at 165 mM [Na+ ]o . Similar findings were obtained using cell-attached patch-clamp recordings from single SIC channels. The results therefore indicate that when exposed to changes in [Na+ ]o , SIC channels undergo a proportional change in PNa /PK . Thus as [Na+ ]o rises from a low to a high value, the driving force across SIC channels not only increases because of the increase in sodium concentration outside the cell, but also because the channels increase their preference for Na+ over K+ as the permeant ion.

The functional significance of this effect could be demonstrated in a variety of ways. First, although the average increase in membrane conductance caused by a +30 mosmol/kg stimulus was not different in MNCs bathed in different values of [Na+ ]o , the average amplitude of the osmoreceptor current recorded at −60 mV varied in proportion with the prevailing [Na+ ]o (Fig. 1). Second, electrical (depolarizing and action potential) responses evoked by a +30 mosmol/kg stimulus were greater when evoked as excess NaCl than when presented as excess mannitol (a membrane impermeant disaccharide). Third, because SIC channels have a non-zero Po at rest, changes in [Na+ ]o not accompanied by changes in osmolality (e.g. by substituting NaCl for mannitol) produced an ionic current resulting exclusively from changes in driving force even though SIC channel Po was unchanged. Indeed, MNCs held near resting potential displayed fast sustained inward currents under voltage clamp (or depolarization under current clamp) upon exposure to a +20 mM increase in [Na+ ]o . These responses were enhanced by prior cell shrinking evoked by applying suction to the patch pipette, and were inhibited when tested in the presence of hypotonic solutions or Gd3+ , indicating that they were effectively mediated through SIC channels. Thus the biophysical properties of SIC channels enable MNCs to operate as sensitive sensors of changes in external [Na+ ] or osmolality. More importantly, the proportional amplification of osmotic responses that occurs upon coincident increases in [Na+ ]o mirrors the enhancement or blunting of hypertonicity-evoked VP responses during concurrent increases and decreases in CSF [Na+ ] in vivo (e.g. McKinley et al., 1978). This effective modulation of osmosensory gain may therefore be of importance for the concerted regulation of MNCs during the concerted changes in CSF [Na+ ] and osmotic pressure that take place under physiological conditions (Fig. 1). Neuropeptides as modulators of MNC activity Fast synaptic neurotransmission plays a key role in the regulation of MNCs (Bourque and Renaud, 1990) and appears to rely mostly on synapses featuring glutamate or GABA as chemical transmitters (Renaud and Bourque, 1991; Renaud et al., 1992).

89

Fig. 1. Modulation of intrinsic osmosensitivity in MNCs. Changes in osmolality cause changes in cell volume that alter the probability of opening of SIC channels, possibly through mechanical effects mediated via the cytoskeleton. Changes in SIC channel activity alter the membrane potential and firing rate of MNCs. As a result, exocytotic peptide secretion is altered. Changes in [Na+ ]o modulate osmoreceptor currents by affecting the driving force through the channel and by altering the relative permeability to Na+ ions. Osmotic stimuli are normally associated with proportional changes in CSF [Na+ ] (dashed line). Numerous excitatory peptides, particularly those mediating their actions through Gq/11 , appear to enhance osmosensory gain. This effect might be mediated by peptide-evoked changes in cell volume, cytoskeleton properties or SIC channel gating.

Some of the neural pathways conveying osmosensitive and non-osmosensitive information to hypothalamic MNCs, however, are also known to contain neuropeptides as putative neurotransmitters (Renaud and Bourque, 1991; Renaud et al., 1992). Subfornical organ (SFO) neurons for example, were recently shown to be intrinsically osmosensitive (Anderson et al., 2000) and to increase their firing rate in response to decreases in blood pressure subsequent to increases in circulating levels of angiotensin II (Gutman et al., 1988; Ciriello, 1997). Since SFO neurons send direct projections to MNCs (Jhamandas et al., 1989), this circuit may be well poised to contribute to the osmotic and cardiovascular regulation of VP release through an effect on MNCs. While SFO neurons appear to relay fast excitatory signals to MNCs through the release of glutamate, activation of this pathway also leads to a slower

excitatory effect mediated via the release of angiotensin II (Ang) onto MNCs (Jhamandas et al., 1989; Ferguson and Washburn, 1998). The finding that neuropeptides are frequently and prominently featured in afferents targeting hypothalamic MNCs prompted several studies aimed at identifying the cellular effects of these agents on MNCs in vitro, where high resolution electrophysiological analysis can be performed. Early studies using sharp electrode intracellular recordings revealed that the excitatory effects of several peptides, including Ang (Yang et al., 1992), cholecystokinin (CCK; Jarvis et al., 1992), neurotensin (NT; Kirkpatrick and Bourque, 1995) and polypeptide 7B2 (Senatorov et al., 1993) were associated with increases in membrane conductance. Moreover, these experiments indicated that the ionic current stimulated by these excitatory peptides re-

90

versed near −30 mV, a value which did not correspond to the equilibrium potential for Na+ (E Na ), K+ (E K ) or Cl− (E Cl ) in the conditions of these experiments. Moreover, changing E Cl through the use of KCl-filled micropipettes failed to affect E rev , suggesting that the current might reflect the activation of non-selective cation permeable channels. Given the apparent ubiquity of this mechanism, we hypothesized that different excitatory neuropeptides might be targeting a common population of ionic channels. Moreover, since the value of E rev of the peptide-stimulated response was similar to that of the intrinsic osmosensory response of these cells, the possibility that these peptides actually modulated SIC channels became worthy of consideration. Excitatory peptides convergently regulate SIC channels in MNCs In order to characterize and compare the basis for the effects of several excitatory peptides, we obtained whole-cell patch-clamp recordings from MNCs acutely isolated from the supraoptic nucleus of adult rats (Chakfe and Bourque, 2000, 2001). Saturating concentrations (1–5 μM) of Ang, CCK or NT were bath-applied for 0.5–5 min and the effects were observed under current and voltage clamp. Under current clamp all three peptides elicited a marked and reversible membrane depolarization from an initial baseline voltage near −65 mV in about 80% of the cells tested, consistent with previous observations using sharp electrode recordings in hypothalamic explants. When spike threshold was exceeded, this response was associated with the discharge of action potentials. In most cases the excitatory effects of the peptides reached a peak about 40 s after the addition of the drug and the cells recovered within a few minutes after washout. Under voltage clamp (VH = −65 mV), the peptides evoked a reversible inward current that followed a time course similar to that of the voltage changes observed under current clamp. Steady-state current–voltage (I –V ) analysis of the peptide-evoked currents was performed by subtracting current responses to slow (16 mV/s) voltage ramps applied in the absence and presence of peptide. This protocol revealed that Ang, CCK and NT all cause increases in membrane conductance and that for each peptide the evoked current varies as

a linear function of voltage between −100 and −20 mV with an E rev near −40 mV. Values of E rev measured in control conditions were not significantly different for any of the three peptides tested, nor were they significantly changed by reducing [Cl− ]o from 147 to 7 mM. Values of E rev , however, were collectively depolarized by increasing external [K+ ] from 3 to 6 mM and hyperpolarized by lowering external [Na+ ] from 144 to 74 mM. At the macroscopic level, therefore, Ang, CCK and NT elicit membrane depolarization by activating a similar type of nonselective cation conductance. Interestingly, the average increase in membrane conductance evoked by simultaneous application of Ang, CCK and NT was not significantly greater when evoked by peptides applied individually (Chakfe and Bourque, 2000). Since saturating concentrations of the peptides were used, these findings suggested that maximal activation of one receptor could occlude the effects of other receptors. Single-channel analysis was performed using the cell-attached configuration of the patch-clamp technique. Blockers of common voltage gated Na+ and K+ channels were included in the pipette to reduce the activity of other types of channels. Following the establishment of a gigaseal, single-channel activity was recorded for 1–5 min at a holding potential below −60 mV and saturating concentrations of Ang, CCK or NT were applied to the bath. When the observed single-channel activity was analyzed in consecutive 10-s bins, values of Po were found to increase gradually, reaching a peak between 30 and 40 s. For all three peptides, the increases in Po were specifically associated with proportional decreases in mean closed time (average interval between openings) rather than with changes in mean open time. In addition to similar gating properties, single channels stimulated by different neuropeptides displayed identical conductances (∼35 pS) and ionic permeation properties. Moreover, although the values of E rev of the channels stimulated by Ang, CCK and NT did not differ from each other when recorded in control solution, the values were equally depolarized by increasing external [K+ ] from 3 to 6 mM, and hyperpolarized by decreasing external [Na+ ] from 140 to 70 mM. Thus the channels stimulated by Ang, CCK and NT have gating, conductance and permeation properties that are indistinguishable from

91

each other, and that are consistent with the macroscopic currents evoked during whole-cell recording. In seven patches in which double-channel openings were never observed, and in which the duration of the recording allowed us to establish that the chances that two channels were nonetheless present in the patch was less than 5% (i.e. patches presumed to contain only one channel; Chakfe and Bourque, 2000), reversible increases in channel activity could be evoked by consecutive applications of different peptides, indicating that the receptors for Ang, CCK and NT not only target a single population of ion channels, but that individual channels can be modulated by multiple receptors (Fig. 1). The identity of the peptide-stimulated channels was further probed by examining their sensitivity to changes in pipette pressure. Values of Po were found to vary as a bell-shaped function of absolute pipette pressure with maximum activity near 0 mm Hg. This activity profile identified the peptide-stimulated channels as being stretch-inactivated (Oliet and Bourque, 1993a, 1996). Further evidence supporting this finding was provided by the observation that peptide-stimulated channels could be blocked by Gd3+ (Chakfe and Bourque, 2000) with a sensitivity that was indistinguishable from that of the mechanically stimulated SIC channels mediating intrinsic osmosensory responses (IC50 ∼40 μM; Oliet and Bourque, 1996). Finally, since macroscopic current and voltage responses recorded under whole-cell conditions could also be blocked by Gd3+ , it was concluded that the excitation of MNCs mediated by Ang, CCK and NT results specifically from an increase in the activity of SIC channels. Interactions between excitatory peptides and osmotic responses The involvement of SIC channels in mediating the intrinsic osmosensory responses of MNCs, as well as their excitatory responses to Ang, CCK and NT, suggests that simultaneous activation of receptorand osmotically mediated mechanisms might result in functionally relevant interactions. In agreement with this hypothesis, bath application of Ang was found to be ineffective when MNCs were incubated in a hypoosmotic medium. This observation suggests that the stretch-induced inhibition of SIC channels

mediated by cell swelling can overcome the excitatory effect of peptidergic receptor activation. The effectiveness of some peptidergic inputs onto MNCs might therefore be gated by systemic osmolality in situ. Conversely, responses evoked by a constant hypertonic stimulus (+30 mosmol/kg; as excess mannitol) were consistently greater when evoked in the presence of a neuropeptide, suggesting that the activation of Ang, CCK or NT receptors may be able to enhance the mechanosensitivity of the SIC channels and, in so-doing, increase the gain of the intrinsic osmosensory response. Discussion The release of VP by hypothalamic MNCs normally decreases when the osmolality of the extracellular fluid (ECF) drops to values below the physiological set-point (∼295 mosmol/kg in the rat). This response is important because by down-regulating the kidney’s ability to retain free water through the production of a dilute urine, reduced VP levels in the plasma immediately promote a compensatory increase in ECF osmolality. Conversely, increases in VP release during hyperosmotic conditions promote water retention by the kidney and thus contribute to a restoration of normal ECF osmolality as additional water is ingested. The inhibitory and excitatory electrical responses which lead to osmotically evoked decreases and increases in VP secretion result from a complex interaction of intrinsic and extrinsic features (Bourque, 1998; Hussy et al., 2000). Currently available electrophysiological evidence suggests that the hypotonic inhibition of MNCs results from: (1) a decrease in synaptic excitation mediated by glutamatergic inputs from OVLT neurons; (2) the hyperpolarizing effect of glycine receptor activation in response to taurine release from astrocytes surrounding MNCs; and (3) from a hyperpolarizing effect due to the suppression of a basal inward current flowing through SIC channels which undergo a decrease in Po as a result of osmotic swelling. Conversely, hypertonic excitation of MNCs results from: (1) an increase in glutamatergic synaptic excitation from OVLT neurons; (2) a reduction of basal glycine receptor activation subsequent to the suppression of basal taurine release from neighboring astrocytes; and (3) an increase in the total inward current medi-

92

ated through SIC channels upon cell shrinking. The involvement of synaptic inhibitory mechanisms, if any, remains to be established. Mechanisms underlying the effects of external Na+ on MNCs The results described above indicated that MNCs are intrinsically capable of detecting subtle changes in external [Na+ ]. When CSF [Na+ ] increases, the enhanced driving force through SIC channels augments the amplitude of the cationic current flowing through the SIC channels that are active in the steady-state. As a result, MNCs exposed to an isoosmotic CSF hypernatremia will be depolarized and their excitability will be increased. Conversely, CSF hyponatremia will reduce the driving force across SIC channels and attenuate the amplitude of the basal inward current flowing through active channels. Sodium detection through this mechanism, therefore, is conditional on the existence of a basal level of SIC channel activity. Under hypotonic conditions, when SIC channels are inactive, MNCs should not be responsive to changes in [Na+ ]o . The sensitivity of MNCs to changes in [Na+ ]o , however, would be increased under hypertonic conditions. Although it is not clear if MNCs would ever become exposed to isotonic changes in [Na+ ]o under physiological conditions, changes in CSF [Na+ ]o are known to accompany fluctuations in plasma osmolality, regardless of the nature of the impermeant solute added experimentally. It is safe to assume, therefore, that parallel changes in CSF [Na+ ]o and osmolality accompany systemic osmotic perturbations under physiological conditions. The coincidence of these stimuli should serve to reinforce the responsiveness of MNCs to osmotic stimuli; i.e. increase osmosensory gain. Effects of neuropeptides on osmosensory gain Hypothalamic MNCs exposed to excitatory neuropeptides generally display a slow membrane depolarization accompanied by the induction of action potential firing or by an increase in firing rate. The experiments reviewed above have shown that at least three of these peptides (Ang, CCK and NT) produce their depolarizing effect through the stimulation of SIC channels. The biochemical

mechanisms by which peptidergic receptor activation leads to the stimulation of the channels remains to be determined. Molecular studies, however, have already shown that AT-1 receptors (Murphy et al., 1991), CCK-B receptors (Wank et al., 1992; Lee et al., 1993) and NT S1/S2 receptors (Vincent et al., 1999) all comprise the seven transmembrane domains that are characteristic of G-protein coupled effectors. Moreover, AT-1 receptors (Hunyady et al., 1996), CCK-B (Kopin et al., 1992; Lee et al., 1993) receptors and NT S1/S2 receptors (Vincent et al., 1999) have been shown to stimulate the activation of phospholipase C via binding to the G-protein Gq/11 , to provoke the production of inositol-1,4,5triphosphate, and to mobilize Ca2+ from intracellular stores. It appears likely, therefore, that the convergent activation of the stretch-inactivated cation channels in MNCs may result from a commonality of the second messenger systems responsible for the effects of Ang II, CCK and NT (Fig. 1). If this is the case, it is likely that increases in the Po of stretch-inactivated cation channels of MNCs will eventually be found to mediate the cationic conductances known to be stimulated by other excitatory peptides, such as activin-A (Oliet et al., 1995), PACAP (Shibuya et al., 1998) and peptide 7B2 (Senatorov et al., 1993). Indeed, the response of MNCs to PACAP has already been shown to be sensitive to Gd3+ (Shibuya et al., 1998). The activation of enzymatic pathways leading to channel phosphorylation or dephosphorylation and altered gating properties may well be involved in the modulatory effects of Ang II, CCK and NT (Fig. 1). However, since mechanical gating of stretch-sensitive ion channels likely involves the cytoskeleton, the modulatory effects of excitatory neuropeptides could also be exerted through actions on one or more components of the cortical cytoskeleton (Fig. 1). In this context, it is interesting to note that the kinetic basis for the peptidergic modulation of channel Po (i.e. a selective modulation of mean closed time with no effect on mean open time; Chakfe and Bourque, 2000) is the same as that which has been reported to underlie changes in channel activity evoked by osmotic or mechanical stimuli (Oliet and Bourque, 1996). Additional studies will be required to define the molecular basis for the modulatory effects of excitatory neuropeptides.

93

Finally, since responses to hypertonic stimuli are enhanced in the presence of excitatory peptides, anatomical pathways using Ang (e.g. the SFO), CCK or NT may functionally enhance osmosensory gain by releasing such neuromodulators onto MNCs under specific conditions. The synergistic effects of Ang and hyperosmotic stimuli on MNC firing in vivo (Akaishi et al., 1980) and VP release in vitro (Sladek et al., 1982), for example, might be important physiologically when SFO neurons become activated by a hypovolemic stimulus. Under these conditions, the facilitated intrinsic osmoreception that would result from the modulatory effect of Ang on SIC channels might provide a basis for the steeper relation between plasma osmolality and VP release which is known to prevail under such conditions (Dunn et al., 1973). Abbreviations Ang II AT-1 CCK CSF ECF E Cl EK E Na EPSP E rev IC50 I –V MNCS NT OT OVLT PACAP PNa /PK Po SFO SIC VH VP

angiotensin II angiotensin II type 1 (receptor) cholecystokinin cerebrospinal fluid extracellular fluid equilibrium potential for Cl− ions equilibrium potential for K+ ions equilibrium potential for Na+ ions excitatory postsynaptic potential reversal potential half-maximal inhibitory concentration current–voltage magnocellular neurosecretory cells neurotensin oxytocin organum vasculosum lamina terminalis pancreatic adenylate cyclase activating polypeptide permeability to Na+ ions relative to K+ ions probability of opening subfornical organ stretch-inactivated cation holding voltage vasopressin

References Akaishi, T., Negoro, H. and Kobayashi, S. (1980) Responses of paraventricular and supraoptic units to angiotensin II, SAR1 ILE8 -angiotensin II, and hypertonic NaCl administered into the cerebral ventricle. Brain Res., 188: 499–511. Anderson, J.W., Washburn, D.L. and Ferguson, A.V. (2000) Intrinsic osmosensitivity of subfornical organ neurons. Neuroscience, 100: 539–547. Bicknell, R.J. (1988) Optimizing release from peptide hormone secretory nerve terminals. J. Exp. Biol., 139: 51–65. Bourque, C.W. (1989) Ionic basis for the intrinsic activation of rat supraoptic neurones by hyperosmotic stimuli. J. Physiol., 417: 265–278. Bourque, C.W. (1990) Intraterminal recordings from the rat neurohypophysis in vitro. J. Physiol., 421: 247–262. Bourque, C.W. (1991) Activity-dependent modulation of nerve terminal excitation in a mammalian peptidergic system. Trends Neurosci., 14: 28–30. Bourque, C.W. (1998) Osmoregulation of vasopressin neurons: a synergy of intrinsic and synaptic processes. Prog. Brain Res., 119: 59–76. Bourque, C.W. and Oliet, S.H.R. (1997) Osmoreceptors in the central nervous system. Annu. Rev. Physiol., 59: 601–619. Bourque, C.W. and Renaud, L.P. (1990) Electrophysiology of mammalian vasopressin and oxytocin neurosecretory neurons. Front. Neuroendocrinol., 11: 183–212. Bourque, C.W. and Richard, D. (2001) Axonal projections from the organum vasculosum lamina terminalis to the supraoptic nucleus: functional analysis and presynaptic modulation. Clin. Exp. Pharm. Physiol., 28: 570–574. Bourque, C.W., Richard, D. and Oliet, S.H.R. (1994) Osmoreceptors, osmoreception, and osmoregulation. Front. Neuroendocrinol., 15: 231–274. Bres, V., Hurbin, A., Duvoid, A., Orcel, H., Moos, F.C., Rabie, A. and Hussy, N. (2000) Pharmacological characterization of volume-sensitive, taurine permeable anion channels in rat supraoptic glial cells. Br. J. Pharmacol., 130: 1976–1982. Chakfe, Y. and Bourque, C.W. (2000) Excitatory peptides and osmotic pressure modulate mechanosensitive cation channels in concert. Nat. Neurosci., 3: 572–579. Chakfe, Y. and Bourque, C.W. (2001) Peptidergic excitation of supraoptic nucleus neurons: involvement of stretch-inactivated cation channels. Exp. Neurol., 171: 210–218. Ciriello, J. (1997) Afferent renal inputs onto subfornical organ neurons responsive to angiotensin II. Am. J. Physiol., 272: R1684–R1689. Deleuze, C., Duvoid, A. and Hussy, N. (1998) Properties and glial origin of osmotic-dependent release of taurine from the rat supraoptic nucleus. J. Physiol., 507: 463–471. Dunn, F.L., Brennan, T.J., Nelson, A.E. and Robertson, G.L. (1973) The role of blood osmolality and volume in regulating vasopressin secretion in the rat. J. Clin. Invest., 52: 3212– 3219. Ferguson, A.V. and Washburn, D.L. (1998) Angiotensin II: a peptidergic neurotransmitter in central autonomic pathways. Prog. Neurobiol., 54: 169–192.

94 Fisher, T.E. and Bourque, C.W. (1996) Calcium channel subtypes in the somata and axon terminals of the magnocellular neurosecretory cells of the rat supraoptic nucleus. Trends Neurosci., 19: 440–444. Gutman, M.B., Ciriello, J. and Mogenson, G.J. (1988) Effects of plasma angiotensin II and hypernatremia on subfornical organ neurons. Am. J. Physiol., 254: R746–R754. Hunyady, L., Balla, T. and Catt, K.J. (1996) The ligand binding site of the angiotensin AT1 receptor. Trends Pharmacol. Sci., 17: 135–140. Hussy, N., Deleuze, C., Pantaloni, A., Desarmenien, M.G. and Moos, F.C. (1997) Agonist action of taurine on glycine receptors in rat supraoptic magnocellular neurones: possible role in osmoregulation. J. Physiol., 502: 609–621. Hussy, N., Deleuze, C., Desarmenien, M.G. and Moos, F.C. (2000) Osmotic regulation of neuronal activity: a new role for taurine and glial cells in a hypothalamic neuroendocrine structure. Prog. Neurobiol., 62: 113–134. Jarvis, C.R., Bourque, C.W. and Renaud, L.P. (1992) Depolarizing action of cholecystokinin on rat supraoptic neurones in vitro. J. Physiol., 458: 621–632. Jhamandas, J.H., Lind, R.W. and Renaud, L.P. (1989) Angiotensin II may mediate excitatory neurotransmission from the subfornical organ to the hypothalamic supraoptic nucleus: an anatomical and electrophysiological study in the rat. Brain Res., 487: 52–61. Kirkpatrick, K. and Bourque, C.W. (1995) Effects of neurotensin on rat supraoptic nucleus neurones in vitro. J. Physiol., 482: 373–381. Kopin, A.S., Lee, Y.M., McBride, E.W., Miller, L.J., Lu, M., Lin, H.Y., Kolakowski Jr., L.F. and Beinborn, M. (1992) Expression cloning and characterization of the canine parietal cell gastrin receptor. Proc. Natl. Acad. Sci. USA, 89: 3605–3609. Lee, Y.M., Beinborn, M., McBride, E.W., Lu, M., Kolakowski Jr., L.F. and Kopin, A.S. (1993) The human brain cholecystokinin-B/gastrin receptor. Cloning and characterization. J. Biol. Chem., 268: 8164–8169. Leng, G., Blackburn, R.E., Dyball, R.E.J. and Russell, J.A. (1989) Role of the anterior peri-third ventricular structures in the regulation of supraoptic neuronal activity and neurohypophysial hormone secretion in the rat. J. Neuroendocrinol., 1: 35–46. Mason, W.T. (1980) Supraoptic neurones of rat hypothalamus are osmosensitive. Nature, 287: 154–157. McKinley, M.J., Denton, D.A. and Weisinger, R.S. (1978) Sensors for antidiuresis and thirst-osmoreceptors or CSF sodium detectors. Brain Res., 141: 89–103. Murphy, T.J., Alexander, R.W., Griendling, K.K., Runge, M.S. and Bernstein, K.E. (1991) Isolation of a cDNA encoding the vascular type 1 angiotensin II receptor. Nature, 351: 233–236. Oliet, S.H.R. and Bourque, C.W. (1992) Properties of supraoptic magnocellular neurones isolated from the adult rat. J. Physiol., 455: 291–306. Oliet, S.H.R. and Bourque, C.W. (1993a) Mechanosensitive chan-

nels transduce osmosensitivity in supraoptic neurons. Nature, 364: 341–343. Oliet, S.H.R. and Bourque, C.W. (1993b) Steady-state osmotic modulation of cationic conductance in neurons of the rat supraoptic nucleus. Am. J. Physiol., 265: R1475–R1479. Oliet, S.H.R. and Bourque, C.W. (1996) Gadolinium uncouples mechanical detection and osmoreceptor potential in supraoptic neurons. Neuron, 16: 175–181. Oliet, S.H.R., Plotsky, P.M. and Bourque, C.W. (1995) Effects of activin-A on neurons acutely isolated from the rat supraoptic nucleus. J. Neuroendocrinol., 7: 661–663. Poulain, D.A. and Wakerley, J.B. (1982) Electrophysiology of hypothalamic magnocellular neurones secreting oxytocin and vasopressin. Neuroscience, 7: 773–808. Renaud, L.P. and Bourque, C.W. (1991) Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog. Neurobiol., 36: 131– 169. Renaud, L.P., Allen, A.M., Cunningham, J.T., Jarvis, C.R., Johnson, S.A., Nissen, R., Sullivan, M.J., Van Vulpen, E. and Yang, C.R. (1992) Synaptic and neurotransmitter regulation of activity in mammalian hypothalamic magnocellular neurosecretory cells. Prog. Brain Res., 92: 277–288. Richard, D. and Bourque, C.W. (1995) Synaptic control of rat supraoptic neurones during osmotic stimulation of the organum vasculosum lamina terminalis in vitro. J. Physiol., 489: 567–577. Robertson, G.L., Shelton, R.L. and Athar, S. (1976) The osmoregulation of vasopressin. Kidney Int., 10: 25–37. Senatorov, V.V., Yang, C.R., Marcinkievicz, M., Chrétien, M. and Renaud, L.P. (1993) Depolarizing action of secretory granule protein 7B2 on rat supraoptic neurosecretory neurons. J. Neuroendocrinol., 5: 533–536. Shibuya, I., Kabashima, N., Tanaka, K., Setiadji, V.S., Noguchi, J., Harayama, N., Ueta, Y. and Yamashita, H. (1998) Patchclamp analysis of the mechanism of PACAP-induced excitation in rat supraoptic neurones. J. Neuroendocrinol., 10: 759– 768. Sladek, C.D., Blair, M.L. and Ramsay, D.J. (1982) Further studies on the role of angiotensin in the osmotic control of vasopressin release by the organ-cultured rat hypothalamoneurohypophysial system. Endocrinology, 111: 599–607. Vincent, J.P., Mazella, J. and Kitabgi, P. (1999) Neurotensin and neurotensin receptors. Trends Pharmacol. Sci., 20: 302–309. Voisin, D.L., Chakfe, Y. and Bourque, C.W. (1999) Coincident detection of CSF Na+ and osmotic pressure in osmoregulatory neurons of the supraoptic nucleus. Neuron, 24: 453–460. Wank, S.A., Pisegna, J.R. and De Weerth, A. (1992) Brain and gastrointestinal cholecystokinin receptor family: structure and functional expression. Proc. Natl. Acad. Sci. USA, 89: 8691– 8695. Yang, C.R., Phillips, M.I. and Renaud, L.P. (1992) Angiotensin II receptor activation depolarizes rat supraoptic neurons in vitro. Am. J. Physiol., 262: R1333–R1338.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 8

Glial cells in the hypothalamo-neurohypophysial system: key elements of the regulation of neuronal electrical and secretory activity Nicolas Hussy * CNRS-UMR 5101, CCIPE, 141 rue de la Cardonille, 34094 Montpellier Cedex 5, France

Keywords: Glial cell; Hypothalamus; Taurine; Glycine receptor; Plasticity

Introduction Glial cells have now gained increasing recognition of their active participation in the control of neuronal activity. Because glial cells are incapable of generating and propagating electrical signals, they were long considered as unexcitable. This encouraged many neuroscientists to perceive their role mainly as nutritional and trophic support of neurons, neuronal waste disposal, or regulation of neuronal ionic environment. It has been shown recently that glial cells (mainly astrocytes) can respond to synaptically released neurotransmitters by intracellular Ca2+ signals, which can propagate among the electrically silent astrocyte population, and which in turn induce the release of various neuroactive substances from glial cells to modulate synaptic transmission (Carmignoto, 2000; Araque et al., 2001; Haydon, 2001). Such bi-directional signaling was demonstrated in cortical or hippocampal astrocytes both in culture and in situ, and also in retinal glial cells and the perijunctional Schwann cells in in situ preparations (Araque et al., 2001; Haydon, 2001). This ∗ Correspondence to: N. Hussy, CNRS-UMR 5101, CCIPE, 141 rue de la Cardonille, 34094 Montpellier Cedex 5, France. Tel.: +33 (4) 6714-2966; Fax: +33 (4) 6754-2432; E-mail: [email protected]

has finally brought glial cells under the spotlight at the front scene of neurobiology, prompting the functional implication of glial cells in the integration and processing of neural information to be revisited. This new fascinating concept has come to light due to the recent development of cell imaging techniques allowing the visualization of the dynamics of intracellular Ca2+ signaling in astrocytes, and providing compelling evidence for the active participation of glial cells in neuronal processing. However, the idea of such a role for glial cells is certainly older. For instance, in the hypothalamo-neurohypophysial system (HNS), evidence for this has been accumulating since the late seventies, and the existence of bi-directional communication between glial cells and neurons in the control of electrical and secretory activities of the hypothalamic neurons was already proposed in the early eighties (van Leeuwen et al., 1983; Hatton, 1988). The highly dynamic morphological relationship between neurons and astrocytes, the numerous and various nerve terminals making synaptoid contacts onto astrocytes, as well as the expression by glial cells of several physiologically relevant neurotransmitter receptors, are as many indications of an important involvement of these cells in the regulation of neuroendocrine activity. We recently provided evidence for an acute glia-to-neuron communication in this system, where astrocytes play a sensory role in

96

the osmoregulation of neuroendocrine cells (Hussy et al., 2000a,b). It is likely that the role of glial cells in the control of HNS neuronal activity is still greatly underestimated. The HNS appears highly suited to study the functional implication of glial cells in the regulation of neuronal processing. This is notably due to the relative simplicity of the morphological organization of the system, which has been extensively characterized both from a neuronal and glial stand point. In addition, we know much about the physiological processes the HNS is involved in, and one can easily stimulate it, and record the output at many different levels, both in vivo and in vitro. All these features allow the issue of the respective contribution of the various cell populations to the physiology of the system to be addressed, and as such the HNS represents a very attractive model. In this review, I will summarize the present knowledge of the role of glial cells in the regulation of neuronal activity in the HNS. This is, however, not meant to be an exhaustive account of the voluminous amount of available data, and the reader is referred to other more complete reviews on each aspect (Hatton, 1990, 1997, 1999; Theodosis and Poulain, 1993; Theodosis et al., 1998; Hussy et al., 2000b). Rather, I would like to place these results in the context of the growing emphasis given to the recently recognized participation of glial cells in neuronal processing, as I believe that understanding the extent of this participation today represents a major issue in neurobiology. The hypothalamo-neurohypophysial system Detailed morphological description of the hypothalamo-neurohypophysial system can be found in several review articles (Hatton, 1990, 1999; Armstrong, 1995). I will only briefly summarize the general organization of the system, as the specific anatomical relationships between glia and neurons are directly correlated to the functional properties of the system. The HNS consists of two neuronal populations that synthesize the neurohormones vasopressin (VP) or oxytocin (OT). The cell bodies of these magnocellular neurons are mainly localized in the hypothalamic supraoptic (SON) and paraventricular nuclei (PVN), although many of them can also be found as accessory groups of neurons scattered between the two

nuclei. The PVN, located on each side of the third ventricle, is a complex structure that, in addition to VP and OT magnocellular neurons, also comprises parvocellular neurons mainly involved in the hypothalamo–pituitary–adrenal axis, as well as other non-endocrine neurons. The SON, which lines the optic chiasm and tracts, has a relatively simpler organization, as the only neurons present there belong to the two magnocellular neuron populations. As such, and because of its easy access at the bottom of the hypothalamus, the SON has been more extensively characterized, and will be the focus of the present paper. Roughly equal numbers of VP and OT cells soma are found to occupy the dorsal part of the SON. Dendrites mainly projecting ventrally form the dendritic zone of the SON, which is separated from the subarachnoid space by the ventral glia limitans. The latter consists of the soma and processes of astrocytes, in contact with the most ventrally projecting dendrites. Astrocytes also send many processes dorsally, through the dendritic and somatic zones, where they literally enwrap neuronal elements, physically separating many neuronal somata. They also make typical end-feet contacts with the numerous blood vessels that irrigate the SON. Axons of magnocellular neurons leave the SON dorsally, join the hypothalamoneurohypophysial tract that runs through the medial eminence to terminate in the neurohypophysis. There, the axons branch extensively to form the axonal endings, which are heavily charged with dense core vesicles containing the neuropeptides, and abut onto the abundant vasculature of the neurohypophysis forming neurohemal junctions. Many neurovascular contacts are also made ‘en passant’, so that a single branch can present multiple sites of neurohormone release. Also typical of the neurohypophysis are the large axonal swellings, or Herring bodies, that accumulate large amounts of VP or OT. These structures probably constitute important storage of neurohormones, although release from Herring bodies has also been reported. At the neurohemal junction, the neurohormones cross the basal lamina and enter the general circulation through fenestrated capillaries. Because of these fenestrated capillaries, the neurohypophysis lies outside the blood–brain barrier. Besides OT and VP terminals, the neurohypophysis also contains a large number of non-secretory axons

97

of various origin, as well as specialized astrocytes named pituicytes. These pituicytes show a very intricate relationship with axon endings, engulfing many of them including Herring bodies. Under basal conditions, pituicytes processes often physically separate the nerve terminals from the basal lamina, occupying a high proportion of the vascular surface and therefore limiting the number of axovascular contacts. Axons, both neurosecretory and non-secretory, have also been reported to make synaptoid contacts with pituicytes (see below). Such contacts differ from classical synapses only by the absence of postsynaptic densities (Wittkowski, 1998; Hatton, 1999). VP and OT are mainly involved in the regulation of body fluid balance. VP has an antidiuretic action at the level of the kidney, and contributes to vasoconstriction of arterioles. The function of OT, besides its well characterized involvement in the female during parturition (inducing uterus smooth muscle contraction) and lactation (responsible for contraction of myoepithelial cells of the mammary glands), is less well understood, but OT has been recognized to have a natriuretic role. The mode of release of both neurohormones in the blood depends on the firing pattern of hypothalamic neurons. The electrical activity of VP neurons is typically phasic, with alternate periods of activity and silence. In contrast, OT neurons fire continuously at a slow rate. Hyperosmotic or hypovolemic activation of the HNS reinforce both the phasic and tonic activity of VP and OT neurons, respectively, resulting in an increased neurohormone secretion. Hypoosmotic and hypervolemic stimuli have the opposite effect. During parturition and suckling, OT neurons show a particular mode of activity, characterized by periodic, brief high-frequency bursts of action potentials on top of the tonic firing. These bursts are synchronous in all OT neurons of SON and PVN, triggering a bolus release of OT in the blood responsible for uterus contraction or milk ejection. Glial cell implication in the regulation of neuroendocrine activity Introduction Three main pieces of evidence provide the basis of our present understanding of the physiological impli-

cation of glial cells in the control of neuroendocrine function in the HNS: (1) The HNS is well known for the activitydependent plasticity of the morphological relationships between neurons and glial cells. This mainly results in long-term changes in the functional properties of the neuronal network, by modifying the well established functions of glial cells such as regulation of ionic composition or uptake of neurotransmitters, but also by regulating other less classical neuron–glia interactions like those involving release of neuroactive substances by glial cells. This process involves neuron-to-glia communication that takes places on a relatively slow time scale (minutes to hours). (2) Neurohypophysial pituicytes receive several synaptoid contacts from neurosecretory nerve endings, as well as from many other terminals. As they also express many receptors of the neurotransmitters/peptides present in the terminals, this suggests an active fast-scale neuron-to-glia communication. This remark most probably also applies to hypothalamic astrocytes of the HNS. This points to the potential dynamic interactions between neurons and glia serving short-term regulation of neuronal activity and hormone release. (3) We have recently shown that HNS glial cells act as sensory elements, transmitting the osmotic information to neurons via the release of a ‘gliotransmitter’, taurine, which activates glycine receptors on the neurons to regulate their electrical and secretory activity. This demonstrates that rapid glia-to-neuron communication is directly involved in a physiological process. Long-term regulation of glia–neuron interactions: the morphological plasticity of the HNS It has been known for a long time that the morphological relationships between neurons and glial cells in the HNS are highly plastic and undergo reversible changes under physiological conditions that demand an increase in neurohormone secretion. This remodeling of the morphological organization of the HNS and the associated mechanisms have been the subject of numerous reviews over the past years, and will therefore be treated only briefly here (Hatton, 1988, 1990, 1997, 1999; Theodosis and Poulain,

98

1993; Theodosis and MacVicar, 1996; Theodosis et al., 1998; El Majdoubi et al., 2000; Salm, 2000). Upon stimulation by either dehydration or systemic injection of hypertonic saline, or in parturient and lactating rat, there is a retraction of glial processes in the hypothalamic magnocellular nuclei and the neurohypophysis (Theodosis and Poulain, 1993; Hatton, 1997). In the SON and PVN, this translates into a decrease in the number of glial processes interposed between neuronal elements and therefore an increase in the surface of juxtaposed neuronal membrane both in the somatic and dendritic zones of the nuclei. This is accompanied by an increase in the number of synapses, notably with a strong increase in the incidence of shared synapses (a single presynaptic element contacting two or more postsynaptic elements). The density of synaptic contacts does not change, however, since the augmentation in synaptic incidence parallels an enlargement of somatic size induced by stimulation (Modney and Hatton, 1989; Gies and Theodosis, 1994; El Majdoubi et al., 1996). The change in synapse number does not appear to specifically concern a particular synapse type, as it is observed for GABAergic, glutamatergic or noradrenergic contacts (Theodosis and Poulain, 1993; Gies and Theodosis, 1994; El Majdoubi et al., 1996; Michaloudi et al., 1997). In addition, reorganization of the dendritic tree of SON neurons takes place during lactation (Stern and Armstrong, 1998). Glial retraction in the supraoptic nucleus is also accompanied by a specific and reversible decrease in GFAP immunostaining, a decrease in the thickness of the ventral glia limitans, and a reorientation of radial, dorsally projecting astrocyte processes, that take upon dehydration a direction parallel to the pial surface (Salm, 2000). Are OT and VP neurons similarly affected by these plastic changes? While it seems generally accepted that parturition and lactation induce modifications that primarily affect the OT system (Theodosis et al., 1986a; Theodosis and Poulain, 1993; Hatton, 1997), various studies disagree as to the degree of plasticity undergone by VP neurons and associated glia upon dehydration or hypertonic stimulation. According to Theodosis’ group, the specificity towards OT neurons applies as well during dehydration (Chapman et al., 1986; Theodosis and Poulain, 1993), whereas Hatton’s group has reported that this type of stimulation results

in morphological changes that affect both neuronal populations, albeit with some differences (Marzban et al., 1992; Hatton, 1997). In the neurohypophysis, activated conditions are characterized by a partial retraction of pituicytes from the basal lamina, considerably increasing the extent of neural membrane in contact with the basal lamina, and a reduction in the number of axons engulfed by pituicytes (Hatton, 1997, 1999). Of interest is the report of an accompanying increase in the number of neuroglial synaptoid contacts (Wittkowski and Brinkmann, 1974; Wittkowski, 1998). The morphological changes take place within tens of minutes, with significant remodeling in the SON observed less than 30 min after a single injection of hypertonic saline (Tweedle et al., 1993). They are also completely reversible, with kinetics that depend on the duration of the stimulus (Theodosis and Poulain, 1993; Hatton, 1997; Salm, 2000). What factors contribute to the induction of these glial rearrangements? Activation of β-adrenergic receptors have been shown to trigger stellation of cultured pituicytes (Bicknell et al., 1989) and to induce morphological changes in isolated neurohypophysis that mimic those normally observed during dehydration and lactation (Smithson et al., 1990; Luckman and Bicknell, 1990). Similar effects can be induced by the membrane permeable analogue of cAMP or by activation of adenylate cyclase by forskolin (Ramsell and Cobbett, 1996). Interestingly, this effect involves the adrenalin-preferring β2 receptor on pituicytes (Hatton et al., 1991), and in vivo morphological changes in the neurohypophysis induced by hypertonic stimulation can be prevented by prior adrenal medullectomy (Beagley and Hatton, 1994). This suggests than the natural agonist of β2 receptors may be circulating adrenalin secreted by the adrenals that enters the neurohypophysis through the fenestrated capillaries. Alternatively, activation of these receptors could depend on noradrenergic fibers innervating the neural lobe (Garten et al., 1989). The high expression of β2-adrenergic receptors in SON astrocytes and their specific up-regulation by dehydration also suggest the involvement of these receptors in the glial cell plasticity in the hypothalamic nuclei (Lafarga et al., 1992). A role for nitric oxide (NO) has also been recently proposed, because NO can induce stellation of cultured pituicytes

99

(Ramsell and Cobbett, 1996), and a blocker of NO synthase significantly decreases the morphological changes induced by hypertonic stimulation in vivo (Beagley and Cobbett, 1997). The effects of NO can be reproduced by a membrane-permeable analogue of cGMP and blocked by inhibition of guanylate cyclase (Ramsell and Cobbett, 1996), indicating that elevation of either cAMP or cGMP can trigger the structural plasticity of pituicytes. Within this view, it is noteworthy that atrial natriuretic peptide (ANP) has been shown to induce an increase in cGMP in pituicytes in situ (Rambotti et al., 1994). Lastly, glial coverage diminution and synaptic remodeling in the SON can be induced in non-lactating female rat by intraventricular infusion with OT, suggesting that OT released within the hypothalamic nuclei may be able to trigger morphological plasticity (Theodosis et al., 1986b; Montagnese et al., 1990). This effect nevertheless requires permissive estrogen conditions (Montagnese et al., 1990). Part of the action of OT could be directly on the glial cells, which may express OT receptors as indicated by a study on cultured hypothalamic astrocytes (Di Scala-Guenot et al., 1994). It appears, therefore, that these changes may depend either or both on neuronal afferent inputs to the HNS such as the noradrenergic input, which would influence neurons as well as glial cells, or on the activity of magnocellular neurons themselves, which would release neuropeptides to trigger the reorganization of the system. The consequences of glial retraction are manifold (Fig. 1). A major function of glial cells is the regulation of ionic composition of the external medium, notably the maintenance of a low K+ concentration around neurons. Glial retraction should therefore result in a reduced capacity to take up K+ flowing out of neurons during electrical activity. A small increase in external K+ will depolarize neurons and axon terminals and increase their excitability. Because of the increased neuron–neuron contacts, the activity of one neuron may depolarize its neighboring neurons. This could facilitate a coordinated increase in activity of magnocellular neurons, and may even participate in the synchronization of OT neuron activity observed during parturition and lactation (Belin et al., 1984). Part of this may involve neuropeptides released intranuclearly. Indeed, secretion granules containing VP or OT together with

other various peptides (see above) also accumulate in the soma and dendrites of neuroendocrine cells, and can be released in an activity-dependent manner within the hypothalamic nuclei (Moos et al., 1989; Pow and Morris, 1989; Ludwig, 1998). Such a release is known to regulate the activity of the neurons as an autocontrol mechanism that fosters the expression of a pattern of activity that is optimal for neurohormone release (Moos and Richard, 1989; Gouzènes et al., 1998; Moos et al., 1998). Removal of astrocytic processes should facilitate the paracrine action of the neuropeptides released by one neuron onto its neighbors as well as onto presynaptic terminals (Pittman et al., 2000). Also, there is an increase in direct coupling between neurons permitted by the close juxtaposition of neuronal membrane, most probably via the formation of gap junction channels. This is indicated by the enhanced dye coupling among SON neurons in activated conditions (Hatton, 1997, 1999). Astrocytes also regulate the kinetics and concentration of neurotransmitter in the synaptic cleft. It has been recently demonstrated that, in lactating rats, glutamate transmission in the SON is reduced compared to virgin rats (Oliet et al., 2001). This is due to a decreased uptake of glutamate by astrocytes as a result of retraction of glial processes from around synapses. Glutamate can then spillover from the cleft and activate presynaptic metabotropic glutamate receptors to decrease glutamate release. Although apparently this would tend to decrease the excitability of SON neurons, it has been argued that such an inhibition could serve to filter out low frequency inputs while letting through high frequency bursts of synaptic potentials, possibly facilitating signal-to-noise ratio of high frequency afferent inputs linked to activation of OT neurons (Oliet et al., 2001). A similar mechanism may also be present on inhibitory synapses, since activation of presynaptic GABAB receptors has been shown to inhibit GABAergic synaptic transmission (Kabashima et al., 1997). Because glutamatergic and GABAergic presynaptic afferents also express GABAB and metabotropic glutamate receptors, respectively (Kabashima et al., 1997; Schrader and Tasker, 1997), cross regulation of synaptic activity is even plausible. In the neurohypophysis, pituicyte retraction from the basal lamina will favor axovascular contacts, therefore increasing the access of neurohor-

100

Fig. 1. Functional consequences of activity-dependent glial retraction in the SON. Schematic drawings showing some major implications of glial cells in the control of neuronal activity. Top: under non-activated conditions, astrocytes that normally surround and separate neurons are responsible for uptake of K+ that accumulate in the extracellular space during neuronal electrical activity, for release of the inhibitory agent taurine, and for reuptake of glutamate released in the synaptic cleft. Although not firmly demonstrated, astrocytes could also participate to GABA reuptake, and respond to VP or OT released by neurons (or to other neuropeptides co-released with VP or OT in secretion granules). Bottom: glial retraction upon activation of the HNS is accompanied by increased neuron–neuron juxtaposition and an increased number of both excitatory and inhibitory synapses, notably with the enhanced incidence of shared synapses. This would result in: (1) a decreased capacity of K+ uptake favoring neuron depolarization; (2) a decreased accessibility of taurine to its neuronal receptor (GlyR) relieving its inhibitory influence; (3) the removal of the physical barrier to a paracrine action of released neuropeptides; and (4) a decreased reuptake of glutamate and also possibly GABA, which favors activation of presynaptic metabotropic receptors that modulates synaptic transmission. Cross-activation of metabotropic receptors by glutamate and GABA onto GABAergic and glutamatergic terminals, respectively, is also possible if these synapses are close enough to one another. GLT-1, glial glutamate transporter; GAT, GABA transporter.

101

mone to the blood circulation. The reduced presence of glial cells should also diminish the influence of taurine, an inhibitory factor released by glial cells, by limiting its access to neuronal glycine receptors (see below). These observations about the plastic changes of neuron–glia morphological interactions highlight the functional importance of the communication from neurons to glial cells in the regulation of neuroendocrine activity. The morphological reorganization results in relatively long-term modulation of the ‘classical’ regulatory roles of glial cells, such as regulation of extracellular ionic composition, uptake of neurotransmitter, or physical interposition between neurons or between neurons and basal lamina. Similarly, it should affect glia-to-neuron communications, such as that carried by taurine. All these will have major consequences on the functioning of the neuronal network implicated in the corresponding physiological process. Rapid neuron–glia communication in the neurohypophysis: evidence from synaptoid contacts and receptor expression. Besides the long-term modulation of neuronal function by the plastic rearrangements described above, there are also indications of the existence of rapid communication between neurons and glial cells that does not trigger glial retraction. Apart from the βadrenergic receptors described above, glial cells of the HNS express many receptors for various neurotransmitters and peptides (see reviews by Boersma and van Leeuwen, 1994; Hatton, 1999). Most studies have, however, focused on pituicytes, either in culture or in situ. These cells express receptors for VP (Hatton et al., 1992), κ-opioid (Lightman et al., 1983; Herkenham et al., 1986; Bicknell et al., 1989), nucleotides (Nakai et al., 1999; Troadec et al., 1999), GABA (Bunn et al., 1986), PACAP (LutzBucher et al., 1996), ANP (Luckman and Bicknell, 1991) as well as serotonin, bradykinin, angiotensin II, endothelins, eicosanoids (Nakai et al., 1999) or interleukin-1β (Christensen et al., 1999). Dopamine receptors are likely to be expressed (Meister et al., 1989) as are some subtypes of glutamate receptors (Kiyama et al., 1993). The activation of many of these receptors triggers increases in intracellular

Ca2+ (Hatton et al., 1992; Boersma et al., 1993c; Nakai et al., 1999; Troadec et al., 1999). Others are coupled to cyclic nucleotides, notably triggering the stimulation-induced morphological changes (see above). Although a number of these receptors have only been identified on cultured pituicytes and their presence still has to be confirmed in situ, these data constitute evidence for important neuron-to-glia communication in the neurohypophysis. What is the evidence for a physiological role for these receptors? VP and OT are of course released in the neural lobe. Moreover, VP and OT axons in the neurohypophysis make synaptoid contacts with pituicytes (Wittkowski and Brinkmann, 1974; Boersma et al., 1993b). Synaptoid contacts are characterized by the presence of both dense core secretion granules, that contain the neurohormones, and small vesicles, the content of which is unknown. These structural features suggest that released neurohormone will have the opportunity to act on pituicytes. Furthermore, together with VP or OT, these neuroendocrine cells co-express multiple peptides and/or neurotransmitters. Evidence for the co-localization of VP or OT with dynorphin, Metenkephalin, CCK, galanin, neuropeptide FF, NPY, PACAP, glutamate, ATP, CRF, interleukin-1β have been reported, with various degrees of specificity towards VP or OT (Bondy et al., 1989; Meeker et al., 1991; Meister, 1993; Boersma et al., 1993a; Arimura and Shioda, 1995; Landry et al., 1997; Sheikh et al., 1998; Watt and Hobbs, 2000). Some have been localized in dense-core secretion granules together with VP or OT (opioids, CCK. . . ), whereas others are specifically concentrated in small secretory vesicles (glutamate, interleukin-1β. . . ). At least a proportion of these peptides and transmitters might activate receptors on pituicytes. Finally, other neuronal fibers in the neurohypophysis also form synaptoid contacts with pituicytes. This has been shown for enkephalin-containing (van Leeuwen et al., 1983) and GABAergic fibers (Fig. 2, Buijs et al., 1987), the latter systematically co-localizing dopamine (Vuillez et al., 1987). That synaptoid contacts are likely to represent true functional connections is indicated by a study of glial cells from another part of the pituitary, the pars intermedia. These glial cells are also innervated by GABA/dopamine fibers forming similar synaptoid contacts, and stimulation of

102

103

the pituitary stalk triggers GABAA and dopamine D2 receptor-mediated currents in these cells (Fig. 2, Mudrick-Donnon et al., 1993). In cultured pituicytes, activation of VP but not OT receptors results in a transient elevation of intracellular Ca2+ released from internal stores (Hatton et al., 1992). Activation of numerous other transmitter and peptide receptors induce similar Ca2+ signals (Nakai et al., 1999). These Ca2+ signals could spread to other pituicytes as Ca2+ waves, as seen with many other astrocyte populations in the brain, either through gap junctions, which are present between pituicytes (Hatton, 1997), or via the release of a factor such ATP, which mediates Ca2+ wave propagation in cultured astrocytes (Guthrie et al., 1999). The latter mechanism is feasible since pituicytes express purinergic P2Y receptors linked to IP3 production (Troadec et al., 1999). Interestingly, VP-induced Ca2+ transients in pituicytes can be reduced by prolonged exposure to dynorphin, an opioid co-released with VP, indicating dynamic modulation of glial activity (Boersma et al., 1993c). The consequences of Ca2+ elevation in pituicytes induced by VP or other transmitters are still unknown. It could activate an increase in a Ca2+ -dependent K+ conductance, as shown for the ATP-evoked Ca2+ rise in cultured pituicytes (Troadec et al., 2000), and the resulting K+ efflux may depolarize neurosecretory terminals. Alternatively or additionally, it could trigger in pituicytes the Ca2+ -dependent release of neuroactive substances, which could then act on the terminals (neuroendocrine or non-neuroendocrine) to modulate neurohormone secretion. One such neuroactive substance could be ATP, which has been shown to stimulate release of VP through activation of P2X purinoceptor on the terminals (Troadec et al., 1998).

Another candidate is glutamate: it is present in the pituicytes (Meeker et al., 1991), is known to be released in many brain astrocytes upon Ca2+ elevation (Araque et al., 2001; Haydon, 2001), and glutamate receptors have been visualized on the neurosecretory terminals (Kiyama et al., 1993). Other possible factors released by pituicytes to modulate neurohormone secretion include some neuropeptides (proenkephalin mRNA has been detected in pituicytes, Schafer et al., 1990, and astrocytes are known to synthesize many peptides, Martin, 1992), the amino acid taurine (Hussy et al., 2001, see below), or other diffusible factors such as arachidonic acid (Stuenkel et al., 1996) or NO (Lutz-Bucher and Koch, 1994), but this list is by no means exhaustive. An intriguing example of a potential acute participation of pituicytes in the regulation of hormone secretion is suggested by the combined results of recent studies of cytokine function in the neural lobe. Interleukin-1β (IL-1β) is found in both VP and OT neurosecretory axons, where it is concentrated in small vesicles, suggesting it can be released by neuronal activity (Watt and Hobbs, 2000). SON and PVN magnocellular neurons express receptors for IL-1 (Diana et al., 1999), which are likely candidates for mediating IL-1β-induced stimulation of VP and OT release (Christensen et al., 1990). In addition, pituicytes also express IL-1 receptors, which activation by IL-1β has been shown to stimulate the release of another cytokine, IL-6 (Christensen et al., 1999). IL-6 can further stimulate VP and OT release (Yasin et al., 1994), and possibly potentiate the evoked release of GABA from neurohypophysial GABAergic terminals (De Laurentiis et al., 2000). This points to a complex pattern of modulation of hormone release by cy-

Fig. 2. Functional synaptoid contacts between nerve fibers and glial cells in the hypophysis. (A) Electron microscopy photograph of GABAergic terminals in the neurohypophysis (labeled with an antibody against GABA) making synaptoid contacts (black arrows) onto pituicytes (P). These fibers also co-localize dopamine (not shown). Ax, neurosecretory axon; bl, basal lamina. (B) Similar synaptoid contacts are found in the pars intermedia of the hypophysis, where GABAergic (also co-localizing dopamine) fibers contact glial stellate cells (S). EC, endocrine cell. C, Such synaptoid contacts are functional as indicated by electrophysiological recordings of stellate cells from acutely isolated hypophysis. Responses in stellate cells were induced by electrical stimulation of the pituitary stalk (small arrows) to activate afferent fibers. These responses consisted of a rapid and short depolarization (large arrow) blocked by the GABAA receptor antagonist bicuculline, followed by a slower and longer lasting hyperpolarization blocked by the D2 receptor antagonist sulpiride. Only the GABAA component is associated with a decreased membrane resistance, as indicated by the downward deflections induced by hyperpolarizing current injections which are reduced during the transient response. Pictures in A and B were kindly provided by G. Alonso. C is adapted from Mudrick-Donnon et al. (1993), with permission.

104

tokines, which would involve activation of pituicytes and release of neuroactive factors. The picture is even further complicated by the presence of an IL-1 receptor antagonist co-expressed with IL-1 receptor (and IL-1β) in neurosecretory cells (Diana et al., 1999). Further work is required to ascertain the role of these different factors in the physiological regulation of neurohormone secretion. The data summarized above strongly suggest that neurons in the HNS communicate with glial cells using an array of transmitters and receptors, highlighting the potential for short-term bi-directional signaling between neurons and glia. Most of these studies have dealt with neurohypophysial pituicytes, but it seems reasonable to speculate that a similar or even a more complex picture will be found in the SON and PVN. The diversity of these relationships and their functional implication in the physiology of the system can only be speculated today, but future studies will likely uncover the as yet unsuspected importance of glial cells in the dynamic regulation of HNS neuroendocrine activity. Rapid glia-to-neuron communication in the HNS: implication in the osmoregulation of neurons Osmoregulation of OT and VP neurons is a complex mechanism, the detailed description of which can be found in recent review articles (Bourque et al., 1994; Oliet and Bourque, 1994; Bourque and Oliet, 1997; Bourque, 1998; Hussy et al., 2000b). It involves both the afferent sensory inputs coming from peripheral and central osmoreceptors (notably located in the hypothalamic circumventricular organs), and an intrinsic osmoregulation of magnocellular neurons within the hypothalamic nuclei. We also recently showed the osmotic modulation of neurohormone release within the neurohypophysis (Hussy et al., 2001). It is the coordinated action of all the various osmosensory elements that permits the appropriate neuroendocrine response, e.g. the modulation of the circulating concentrations of OT and VP (Fig. 3). Two different mechanisms are responsible for the intrinsic osmosensitivity of the HNS. First, magnocellular neurons express mechanoreceptors of a particular kind on their plasma membrane (Bourque et al., 1994; Oliet and Bourque, 1994; Bourque and Oliet, 1997; Bourque, 1998). These are stretch-

inactivated cationic channels permeable to both Na+ and K+ . They are active in isoosmotic conditions, albeit with a low probability of opening, and therefore contribute to the basal conductance and resting membrane potential of the neurons. Probability of opening is increased by cell shrinking that occurs during hyperosmotic stimuli, and this results in membrane depolarization and facilitation of firing. Conversely, swelling induced by hypoosmotic stimuli closes the channels, and thus has an inhibitory influence. The sensitivity of the mechanoreceptors is very high, as their activity is significantly modulated by changes in osmotic pressure as low as 2%. In addition to their sensitivity to osmotic pressure, and thus to cell volume, the mechanoreceptors also transduce information about the external Na+ concentration. Apart from the fact that the channels are permeable to Na+ and therefore are influenced by the chemical gradient for this ion, a change in Na+ concentration also alters the permeability ratio between Na+ and K+ , favoring Na+ over K+ as the concentration of Na+ goes up (Voisin et al., 1999). Hence, because a natural hypertonic stimulus such as dehydration is characterized by a rise in both osmolarity and Na+ concentration, the subsequent depolarizing cationic current through the mechanoreceptors results from both the increased probability of opening of the channels and the increased permeability to Na+ . The second mechanism taking part in the osmoregulation of magnocellular neurons involves glial cells. Astrocytes in the SON and pituicytes in the neurohypophysis have been shown to selectively accumulate the amino acid taurine (Pow, 1993; Decavel and Hatton, 1995; Miyata et al., 1997). Taurine is the degradation product of the sulfur amino acids methionine and cysteine, notably through the activity of the enzyme cysteine sulfate decarboxylase, which is selectively expressed in glial cells in the nervous system (Reymond et al., 1996). The cellular localization of taurine also depends on uptake via a taurine transporter, so that taurine is also found in neurons in various other regions of the nervous system. The role of taurine in the brain is still a matter of debate, but the osmodependent release of taurine has led to the general consensus for its participation in the regulation of cell volume (Pasantes-Morales and Schousboe, 1997). However, in the HNS, the function of taurine appears quite different, as was first

105

Fig. 3. Glial cell involvement in the osmoregulation of HNS neuroendocrine function. Astrocytes in the SON and pituicytes in the neurohypophysis release taurine in an osmodependent manner. Taurine inhibits electrical and secretory activities of SON neurons by activating glycine receptors (GlyR) located on the neuronal soma in the SON and on the nerve terminals in the neurohypophysis. Therefore, glial cells act as osmodetectors, transmitting the osmotic status of extracellular fluids to the neurons by way of the release of a neuroactive agent. This process complements the direct osmosensitivity of the neurons due to the presence of excitatory mechanoreceptors on the neuronal soma. The opposite regulation of these two systems by osmotic pressure confers them a cooperative action.

postulated by Hatton (1990). Taurine is constantly released from HNS glial cells under basal osmotic conditions, and efflux is enhanced by hypoosmotic and inhibited by hyperosmotic stimuli (Deleuze et al., 1998, 2000; Brès et al., 2000; Hussy et al., 2001). The osmosensitivity of taurine release is very

high, with significant changes observed with as low as a 3% decrease in osmolarity (Hussy et al., 2001), i.e. within the window of physiological variations in osmotic pressure. Taurine efflux in the HNS is a passive process occurring through the opening of volume-sensitive, taurine-permeable anion channels

106

(Brès et al., 2000; Hussy et al., 2001), very similar to those described in numerous other cell preparations (Nilius et al., 1997). The net level of taurine released is actually the result of the combined efflux from and re-uptake into the glial compartment. Because taurine efflux is already active in basal osmotic conditions, and because enhanced release by mild hypoosmotic stimulus is sustained over at least an hour (Deleuze et al., 1998), the release of taurine in the HNS is most probably not linked to cell volume regulation, at least in the physiological range of osmolarity of the external medium. Moreover, taurine is not an inactive compound as would be expected for a simple osmolyte. Indeed, taurine is a known agonist of inhibitory ligand-gated channels, most notably the strychnine-sensitive glycine receptors. Magnocellular neurons of the HNS express high levels of glycine receptors both on their soma and dendrites in the SON (Alonso and Hussy, unpublished observation), and on their terminals in the neurohypophysis (Hussy et al., 2001). Activation of glycine receptors by taurine opens the associated Cl− channels, with an apparent affinity of 0.4 mM (Hussy et al., 1997). Given the basal concentration of taurine in extracellular fluids (a few tens of μM) and its very high intracellular concentration (a few tens of mM, Huxtable, 1992), high enough concentration of taurine should be easily reached in the narrow extracellular space of the HNS (Hatton, 1999). Glycine receptor activation inhibits the activity of SON VP neurons in vivo (Hussy et al., 1997), and prevents the depolarization-induced entry of Ca2+ into acutely isolated neurohypophysial nerve terminals, as well as the associated release of VP (Hussy et al., 2001). We also showed that these receptors participate to the osmoregulation of neuroendocrine activity in the HNS. Indeed, a peripheral hypoosmotic stimulus is known to decrease the activity of SON neurons. Such an effect on VP neurons in vivo can be prevented by local application of strychnine in the vicinity of the neurons in the SON (Hussy et al., 1997). Interestingly, strychnine also increases the activity of VP neurons in isotonic conditions, indicative of a sustained inhibitory influence through basal activation of glycine receptors. This is in agreement with the sustained basal release of taurine in normal osmotic medium mentioned above. Local osmoregulation through activation of glycine

receptors also takes place in the neurohypophysis since the depolarization-induced release of VP in acutely isolated whole neurohypophysis is reduced by hypotonic stimulation, in a strychnine-sensitive manner (Hussy et al., 2001). What is the evidence that taurine is actually the true endogenous agonist of glycine receptors in the HNS? First, among the potential glycine receptor agonists, i.e. glycine, β-alanine and taurine, taurine is the only amino acid that shows an increased level of release following hypotonic stimulation of the SON (Hussy et al., 1997). Second, no glycine afferent fibers can be detected either in the SON or the neurohypophysis that could account for the high expression of glycine receptors in these structures (Pow, 1993; Rampon et al., 1996). Indeed, all spontaneous and evoked inhibitory synaptic potentials and currents recorded in SON neurons can be blocked by GABAA receptor antagonists (Randle et al., 1986; Wuarin and Dudek, 1993; Kabashima et al., 1997). This indicates that few if any glycinergic synapses are present, and suggests an extrasynaptic localization of glycine receptors in the SON. Lastly, and by far the most compelling argument, the strychnine-sensitive inhibition of VP release by hypoosmotic stimulus can be completely abolished by previous depletion of taurine from the neurohypophysis using a treatment with a blocker of the taurine transporter (Hussy et al., 2001). This latter observation clearly identifies glial taurine as the natural factor involved in the osmosensitive activation of neurohypophysial glycine receptors. The intrinsic osmosensitivity of magnocellular neurons in the SON can then be considered as the combined action of these two systems (Hussy et al., 2000b). Thus, the excitatory (mechanoreceptors activated by hyper and inhibited by hypotonicity) and the inhibitory systems (glial cells, taurine and neuronal glycine receptors, inhibited by hyper and activated by hypotonicity), will act in a complementary manner to modulate neuronal activity. Taken together these results demonstrate an important participation of glial cells (astrocytes and pituicytes) in the regulation of the activity of SON neurons. They serve here as sensory elements, releasing a glia-to-neuron transmitter, taurine, that acts via a ligand-gated channel, the glycine receptor, to modulate neuron excitability and hormone secretion.

107

Such a role of glial cells fits with the morphological organization of the system. In the SON, astrocytes cover the ventral part of the nucleus, separating the neuronal elements from the subarachnoid space where the cerebrospinal fluid circulates, and surround the abundant vasculature of the nucleus forming the astroglial end feet. In this way, SON astrocytes lie at the interface between neurons and extracellular fluids, an ideal location to rapidly sense any change in ionic composition and/or osmotic pressure, and signal this information to neurons. The same remark can be made concerning pituicytes in the neurohypophysis, which normally occupy most of the basal lamina surface, and therefore are well positioned to sense the composition of the blood. Of course, during high hormonal demand, the retraction of astrocytic and pituicytic processes described above should diminish this sensory role of glial cells, and should decrease the inhibitory component by limiting the access of taurine to glycine receptors. In agreement with this hypothesis, preliminary results obtained in vivo in lactating rats, a condition that is accompanied by the selective retraction of glial processes from around OT cells, indicate that strychnine application onto OT neurons has very little impact on their basal activity or burst firing, even though they express glycine receptors as indicated by the clear inhibitory effect of taurine (Hussy et al., 1996). In the same animal, strychnine systematically increases the phasic firing of VP neurons. Also, strychnine does not antagonize the hypoosmolarity-evoked inhibition of OT neurons in lactating rats (Deleuze, Moos and Hussy, unpublished observations). Thus, even though taurine is released from glial cells in these conditions, it cannot reach the glycine receptors of OT neurons. Conclusions and perspectives It has long been known that neuron–glia interactions are involved in the plastic morphological changes that characterizes the functional activation of the HNS. This type of communication induces longterm changes that result in the modulation of the classical function of glial cells, i.e. regulation of the physical and chemical environment of neurons. The expression of multiple receptors by astrocytes and pituicytes, as well as the presence of numer-

ous synaptoid contacts onto glial cells coming from neurosecretory and other axonal terminals, strongly indicate that what we presently know regarding the functional role of glial cells in the control of neuronal activity in the HNS is only the tip of the iceberg. Our data on the osmosensory role of astrocytes and pituicytes demonstrate that glial cells in this system have the capacity to provide rapid communication to neurons to regulate their electrical and secretory behavior. A parallel ought to be made with the abundant recent literature on the dynamic functional involvement of glial cells in the acute regulation of neuronal processing (see reviews by Araque et al., 1999, 2001; Carmignoto, 2000; Haydon, 2001). Astrocytes from various regions of the brain have been shown to express a plethora of receptors for neurotransmitters and peptides (Porter and McCarthy, 1997). Many of these neuropeptides can induce a transient elevation of intracellular Ca2+ in astrocytes (Verkhratsky et al., 1998). Receptor-mediated Ca2+ signaling in astrocytes can be evoked in situ by neuronal activity, following synaptically released neurotransmitters (glutamate, GABA, ACh, ATP, noradrenaline, amongst others, see references in Araque et al., 2001; Haydon, 2001). Pseudosynaptic connections between neuronal terminals and glial cells have also been described in other parts of the hypothalamus as well as in other brain areas. These include functional synaptic contacts mediated by GABA and dopamine in pituitary intermediate lobe (MudrickDonnon et al., 1993), or glutamatergic synapses onto hippocampal oligodendrocyte precursor cells (Bergles et al., 2000). Direct contacts, sometimes described as synaptoidic, between neurons and glia have also been reported in the spinal cord (Ridet et al., 1993), cortex (Aoki, 1992; Paspalas and Papadopoulos, 1996) or septum (Milner et al., 1995). The Ca2+ signaling induced in one astrocyte can propagate to other neighboring astrocytes as shown in culture (see references in Araque et al., 2001), in organotypic slice cultures (Dani et al., 1992; Harris-White et al., 1998) as well as in situ in slices from hypothalamus, thalamus, hippocampus, or retina (van den Pol et al., 1992; Newman and Zahs, 1997; Kang et al., 1998; Parri et al., 2001). Propagation principally involves ATP release from astrocytes and activation of purinergic receptors (Cotrina et al., 1998b; Guthrie et al., 1999; Newman, 2001). Gap junctions are also involved (Gi-

108

aume and Venance, 1998; Newman, 2001), although their role may be sometimes indirect (Scemes et al., 2000). Release of ATP from astrocytes is Ca2+ independent and may imply anion channels (Cotrina et al., 1998a). Ca2+ elevation in astrocytes can trigger release of other neuroactive substances like glutamate (see references in Araque et al., 2001; Haydon, 2001), most probably via vesicular secretion involving SNARE proteins (Araque et al., 2000; Pasti et al., 2001). Substances released by glial cells have been shown to activate neuronal receptors both in culture and in situ (Araque et al., 2001; Parri et al., 2001), and to modulate the strength of synaptic transmission (Araque et al., 1998; Kang et al., 1998; Robitaille, 1998). These data constitute the basis of the newly identified fast-scale bi-directional communication between neurons and glial cells, pointing to the direct, active participation of astrocytes (and possibly other types of glial cells) to rapid neuronal signaling. The large number of preparations from the nervous system where these types of interactions have been described is a strong indication that this role of glial cells is a general feature of brain functioning. It is my conviction that, in the HNS, there is far more to discover about the active participation of glial cells to the regulation and integration of information processing. The coming years promise to be exciting ones! Abbreviations ACh ANP ATP cAMP CCK cGMP CRF GABA GFAP HNS IL IP3 NO NPY OT PACAP

acetylcholine atrial natriuretic peptide adenosine triphosphate cyclic adenosine monophosphate cholecystokinin cyclic guanosine monophosphate corticotropin releasing factor γ-aminobutyric acid glial fibrillary acidic protein hypothalamo-neurohypophysial system interleukin inositol(1,4,5) trisphosphate nitric oxide neuropeptide Y oxytocin pituitary adenylate cyclase activating peptide

PVN SON VP

paraventricular nucleus supraoptic nucleus vasopressin

Acknowledgements I am grateful to Françoise C. Moos, Gérard Alonso and Vicky Tobin for critical reading of the manuscript. I also thank Gérard Alonso for providing immunohistochemical pictures and Mireille Passama for the elaboration of the figures. References Aoki, C. (1992) Beta-adrenergic receptors: astrocytic localization in the adult visual cortex and their relation to catecholamine axon terminals as revealed by electron microscopic immunocytochemistry. J. Neurosci., 12: 781–792. Araque, A., Parpura, V., Sanzgiri, R.P. and Haydon, P.G. (1998) Glutamate-dependent astrocyte modulation of synaptic transmission between cultured hippocampal neurons. Eur. J. Neurosci., 10: 2129–2142. Araque, A., Parpura, V., Sanzgiri, R.P. and Haydon, P.G. (1999) Tripartite synapses: glia, the unacknowledged partner. Trends Neurosci., 22: 208–215. Araque, A., Li, N., Doyle, R.T. and Haydon, P.G. (2000) SNARE protein-dependent glutamate release from astrocytes. J. Neurosci., 20: 666–673. Araque, A., Carmignoto, G. and Haydon, P.G. (2001) Dynamic signaling between astrocytes and neurons. Annu. Rev. Physiol., 63: 795–813. Arimura, A. and Shioda, S. (1995) Pituitary adenylate cyclase activating polypeptide (PACAP) and its receptors: neuroendocrine and endocrine interaction. Front. Neuroendocrinol., 16: 53–88. Armstrong, W.E. (1995) Morphological and electrophysiological classification of hypothalamic supraoptic neurons. Prog. Neurobiol., 47: 291–339. Beagley, G.H. and Cobbett, P. (1997) Inhibition of nitric oxide synthase induces ultrastructural changes in the neurohypophysis of dehydrated rats. Neurosci. Lett., 222: 143–146. Beagley, G.H. and Hatton, G.I. (1994) Systemic signals contribute to induced morphological changes in the hypothalmoneurohypophysial system. Brain Res. Bull., 33: 211–218. Belin, V., Moos, F. and Richard, P. (1984) Synchronization of oxytocin cells in the hypothalamic paraventricular and supraoptic nuclei in suckled rats: direct proof with paired extracellular recordings. Exp. Brain Res., 57: 201–203. Bergles, D.E., Roberts, J.D., Somogyi, P. and Jahr, C.E. (2000) Glutamatergic synapses on oligodendrocyte precursor cells in the hippocampus. Nature, 405: 187–191. Bicknell, R.J., Luckman, S.M., Inenaga, K., Mason, W.T. and Hatton, G.I. (1989) β-Adrenergic and opioid receptors on

109 pituicytes cultured from adult rat neurohypophysis: regulation of cell morphology. Brain Res. Bull., 22: 379–388. Boersma, C.J.C. and van Leeuwen, F.W. (1994) Neuron–glia interactions in the release of oxytocin and vasopressin from the rat neural lobe: the role of opioids, other neuropeptides and their receptors. Neuroscience, 62: 1003–1020. Boersma, C.J.C., Sonnemans, M.A. and van Leeuwen, F.W. (1993a) Immunocytochemical localization of neuropeptide FF (FMRFamide-like peptide) in the hypothalamoneurohypophyseal system of Wistar and Brattleboro rats by light and electron microscopy. J. Comp. Neurol., 336: 555– 570. Boersma, C.J.C., Sonnemans, M.A. and van Leeuwen, F.W. (1993b) Immunoelectron microscopic demonstration of oxytocin and vasopressin in pituicytes and in nerve terminals forming synaptoid contacts with pituicytes in the rat neural lobe. Brain Res., 611: 117–129. Boersma, C.J.C., van Leeuwen, F.W., O’Brien, W.G., Law, G.J., Mason, W.T. and Bicknell, R.J. (1993c) Dynorphin 1–17 delays the vasopressin induced mobilization of intracellular calcium in cultured astrocytes from the rat neural lobe. J. Neuroendocrinol., 5: 583–590. Bondy, C.A., Whitnall, M.H., Brady, L.S. and Gainer, H. (1989) Coexisting peptides in hypothalamic neuroendocrine systems: some functional implications. Cell. Mol. Neurobiol., 9: 427– 446. Bourque, C.W. (1998) Osmoregulation of vasopressin neurons: a synergy of intrinsic and synaptic processes. Prog. Brain Res., 119: 59–76. Bourque, C.W. and Oliet, S.H.R. (1997) Osmoreceptors in the central nervous system. Annu. Rev. Physiol., 59: 601–619. Bourque, C.W., Oliet, S.H.R. and Richard, D. (1994) Osmoreceptors, osmoreception, and osmoregulation. Front. Neuroendocrinol., 15: 231–274. Brès, V., Hurbin, A., Duvoid, A., Orcel, H., Moos, F.C., Rabie, A. and Hussy, N. (2000) Pharmacological characterization of volume-sensitive, taurine permeable anion channels in rat supraoptic glial cells. Br. J. Pharmacol., 130: 1976–1982. Buijs, R.M., van Vulpen, E.H. and Geffard, M. (1987) Ultrastructural localization of GABA in the supraoptic nucleus and neural lobe. Neuroscience, 20: 347–355. Bunn, S.J., Hanley, M.R. and Wilkin, G.P. (1986) Autoradiographic localization of peripheral benzodiazepine, dihydroalprenolol and arginine vasopressin binding sites in the pituitaries of control, stalk transected and Brattleboro rats. Neuroendocrinology, 44: 76–83. Carmignoto, G. (2000) Reciprocal communication systems between astrocytes and neurones. Prog. Neurobiol., 62: 561– 581. Chapman, D.B., Theodosis, D.T., Montagnese, C., Poulain, D.A. and Morris, J.F. (1986) Osmotic stimulation causes structural plasticity of neurone–glia relationships of the oxytocin but not vasopressin secreting neurones in the hypothalamic supraoptic nucleus. Neuroscience, 17: 679–686. Christensen, J.D., Hansen, E.W. and Fjalland, B. (1990) Influence of interleukin-1β on the secretion of oxytocin and vasopressin from the isolated rat neurohypophysis. Pharmacol. Toxicol., 67: 81–83.

Christensen, J.D., Hansen, E.W., Frederiksen, C., Molris, M. and Moesby, L. (1999) Adrenaline influences the release of interleukin-6 from murine pituicytes: role of β2 -adrenoceptors. Eur. J. Pharmacol., 378: 143–148. Cotrina, M.L., Lin, J.H., Alves-Rodrigues, A., Liu, S., Li, J., Azmi-Ghadimi, H., Kang, J., Naus, C.C. and Nedergaard, M. (1998a) Connexins regulate calcium signaling by controlling ATP release. Proc. Natl. Acad. Sci. USA, 95: 15735–15740. Cotrina, M.L., Lin, J.H. and Nedergaard, M. (1998b) Cytoskeletal assembly and ATP release regulate astrocytic calcium signaling. J. Neurosci., 18: 8794–8804. Dani, J.W., Chernjavsky, A. and Smith, S.J. (1992) Neuronal activity triggers calcium waves in hippocampal astrocyte networks. Neuron, 8: 429–440. De Laurentiis, A., Pisera, D., Lasaga, M., Diaz, M., Theas, S., Duvilanski, B. and Seilicovich, A. (2000) Effect of interleukin-6 and tumor necrosis factor-α on GABA release from mediobasal hypothalamus and posterior pituitary. Neuroimmunomodulation, 7: 77–83. Decavel, C. and Hatton, G.I. (1995) Taurine immunoreactivity in the rat supraoptic nucleus: prominent localization in glial cells. J. Comp. Neurol., 354: 13–26. Deleuze, C., Duvoid, A. and Hussy, N. (1998) Properties and glial origin of osmotic-dependent release of taurine from the rat supraoptic nucleus. J. Physiol., 507: 463–471. Deleuze, C., Duvoid, A., Moos, F.C. and Hussy, N. (2000) Tyrosine phosphorylation modulates the osmosensitivity of volume-dependent taurine efflux from glial cells in the rat supraoptic nucleus. J. Physiol., 523: 291–299. Di Scala-Guenot, D., Mouginot, D. and Strosser, M.T. (1994) Increase of intracellular calcium induced by oxytocin in hypothalamic cultured astrocytes. Glia, 11: 269–276. Diana, A., Van Dam, A.M., Winblad, B. and Schultzberg, M. (1999) Co-localization of interleukin-1 receptor type I and interleukin-1 receptor antagonist with vasopressin in magnocellular neurons of the paraventricular and supraoptic nuclei of the rat hypothalamus. Neuroscience, 89: 137–147. El Majdoubi, M., Poulain, D.A. and Theodosis, D.T. (1996) The glutamatergic innervation of oxytocin- and vasopressinsecreting neurons in the rat supraoptic nucleus and its contribution to lactation-induced synaptic plasticity. Eur. J. Neurosci., 8: 1377–1389. El Majdoubi, M., Poulain, D.A. and Theodosis, D.T. (2000) Activity-dependent morphological synaptic plasticity in an adult neurosecretory system: magnocellular oxytocin neurons of the hypothalamus. Biochem. Cell. Biol., 78: 317–327. Garten, L.L., Sofroniew, M.V. and Dyball, R.E. (1989) A direct catecholaminergic projection from the brainstem to the neurohypophysis of the rat. Neuroscience, 33: 149–155. Giaume, C. and Venance, L. (1998) Intercellular calcium signaling and gap junctional communication in astrocytes. Glia, 24: 50–64. Gies, U. and Theodosis, D.T. (1994) Synaptic plasticity in the rat supraoptic nucleus during lactation involves GABA innervation and oxytocin neurons: a quantitative immunocytochemical analysis. J. Neurosci., 14: 2861–2869. Gouzènes, L., Desarménien, M.G., Hussy, N., Richard, P. and Moos, F.C. (1998) Vasopressin regularizes the phasic firing

110 pattern of rat hypothalamic magnocellular vasopressin neurons. J. Neurosci., 18: 1879–1885. Guthrie, P.B., Knappenberger, J., Segal, M., Bennett, M.V., Charles, A.C. and Kater, S.B. (1999) ATP released from astrocytes mediates glial calcium waves. J. Neurosci., 19: 520– 528. Harris-White, M.E., Zanotti, S.A., Frautschy, S.A. and Charles, A.C. (1998) Spiral intercellular calcium waves in hippocampal slice cultures. J. Neurophysiol., 79: 1045–1052. Hatton, G.I. (1988) Pituicytes, glia and control of terminal secretion. J. Exp. Biol., 139: 67–79. Hatton, G.I. (1990) Emerging concepts of structure–function dynamics in adult brain: the hypothalamo-neurohypophysial system. Prog. Neurobiol., 34: 437–504. Hatton, G.I. (1997) Function-related plasticity in hypothalamus. Annu. Rev. Neurosci., 20: 375–397. Hatton, G.I. (1999) Astroglial modulation of neurotransmitter/peptide release from the neurohypophysis: present status. J. Chem. Neuroanat., 16: 203–221. Hatton, G.I., Luckman, S.M. and Bicknell, R.J. (1991) Adrenalin activation of β2-adrenoceptors stimulates morphological changes in astrocytes (pituicytes) cultured from adult rat neurohypophyses. Brain Res. Bull., 26: 765–769. Hatton, G.I., Bicknell, R.J., Hoyland, J., Bunting, R. and Mason, W.T. (1992) Arginine vasopressin mobilises intracellular calcium via V1-receptor activation in astrocytes (pituicytes) cultured from adult rat neural lobes. Brain Res., 588: 75–83. Haydon, P.G. (2001) Glia: listening and talking to the synapse. Nat. Rev. Neurosci., 2: 185–193. Herkenham, M., Rice, K.C., Jacobson, A.E. and Rothman, R.B. (1986) Opiate receptors in rat pituitary are confined to the neural lobe and are exclusively kappa. Brain Res., 382: 365– 371. Hussy, N., Deleuze, C., Pantaloni, A., Desarménien, M.G. and Moos, F. (1996) Regulation by taurine of the activity of rat supraoptic magnocellular neurones. J. Physiol., 495: 199P– 120P. Hussy, N., Deleuze, C., Pantaloni, A., Desarménien, M.G. and Moos, F. (1997) Agonist action of taurine on glycine receptors in rat supraoptic magnocellular neurones: possible role in osmoregulation. J. Physiol., 502: 609–621. Hussy, N., Deleuze, C., Brès, V. and Moos, F.C. (2000a) New role of taurine as an osmomediator between glial cells and neurons in the rat supraoptic nucleus. Adv. Exp. Med. Biol., 483: 227–237. Hussy, N., Deleuze, C., Desarménien, M.G. and Moos, F.C. (2000b) Osmotic regulation of neuronal activity: a new role for taurine and glial cells in a hypothalamic neuroendocrine structure. Prog. Neurobiol., 62: 113–134. Hussy, N., Brès, V., Rochette, M., Duvoid, A., Alonso, G., Dayanithi, G. and Moos, F.C. (2001) Osmoregulation of vasopressin secretion via activation of neurohypophysial nerve terminals glycine receptors by glial taurine. J. Neurosci., 21: 7110–7116. Huxtable, R.J. (1992) Physiological actions of taurine. Physiol. Rev., 72: 101–163. Kabashima, N., Shibuya, I., Ibrahim, N., Ueta, Y. and Yamashita, H. (1997) Inhibition of spontaneous EPSCs and IPSCs by

presynaptic GABAB receptors on rat supraoptic magnocellular neurons. J. Physiol., 504: 126–133. Kang, J., Jiang, L., Goldman, S.A. and Nedergaard, M. (1998) Astrocyte-mediated potentiation of inhibitory synaptic transmission. Nat. Neurosci., 1: 683–692. Kiyama, H., Sato, K. and Tohyama, M. (1993) Characteristic localization of non-NMDA type glutamate receptor subunits in the rat pituitary gland. Mol. Brain Res., 19: 262–268. Lafarga, M., Berciano, M.T., Del Olmo, E., Andres, M.A. and Pazos, A. (1992) Osmotic stimulation induces changes in the expression of β-adrenergic receptors and nuclear volume of astrocytes in supraoptic nucleus of the rat. Brain Res., 588: 311–316. Landry, M., Roche, D., Angelova, E. and Calas, A. (1997) Expression of galanin in hypothalamic magnocellular neurones of lactating rats: co-existence with vasopressin and oxytocin. J. Endocrinol., 155: 467–481. Lightman, S.L., Ninkovic, M., Hunt, S.P. and Iversen, L.L. (1983) Evidence for opiate receptors on pituicytes. Nature, 305: 235–237. Luckman, S.M. and Bicknell, R.J. (1990) Morphological plasticity that occurs in the neurohypophysis following activation of the magnocellular neurosecretory system can be mimicked in vitro by β-adrenergic stimulation. Neuroscience, 39: 701–709. Luckman, S.M. and Bicknell, R.J. (1991) Binding sites for atrial natriuretic peptide (ANP) on cultured pituicytes: lack of effect of ANP on release of neurohypophysial hormones in vitro. Neurosci. Lett., 123: 156–159. Ludwig, M. (1998) Dendritic release of vasopressin and oxytocin. J. Neuroendocrinol., 10: 881–895. Lutz-Bucher, B. and Koch, B. (1994) Evidence for an inhibitory effect of nitric oxides on neuropeptide secretion from isolated neural lobe of the rat pituitary gland. Neurosci. Lett., 165: 48–50. Lutz-Bucher, B., Monnier, D. and Koch, B. (1996) Evidence for the presence of receptors for pituitary adenylate cyclaseactivating polypeptide in the neurohypophysis that are positively coupled to cyclic AMP formation and neurohypophyseal hormone secretion. Neuroendocrinology, 64: 153–161. Martin, D.L. (1992) Synthesis and release of neuroactive substances by glial cells. Glia, 5: 81–94. Marzban, F., Tweedle, C.D. and Hatton, G.I. (1992) Reevaluation of the plasticity in the rat supraoptic nucleus after chronic dehydration using immunogold for oxytocin and vasopressin at the ultrastructural level. Brain Res. Bull., 28: 757–766. Meeker, R.B., Swanson, D.J., Greenwood, R.S. and Hayward, J.N. (1991) Ultrastructural distribution of glutamate immunoreactivity within neurosecretory endings and pituicytes of the rat neurohypophysis. Brain Res., 564: 181–193. Meister, B. (1993) Gene expression and chemical diversity in hypothalamic neurosecretory neurons. Mol. Neurobiol., 7: 87– 110. Meister, B., Villar, M.J., Schalling, M., Ehrlich, M., Greengard, P. and Hokfelt, T. (1989) Demonstration of DARPP-32 in pituicytes of the neurohypophysis-decreased expression after administration of hyperosmotic stimuli. Acta Physiol. Scand., 137: 461–462.

111 Michaloudi, H.C., el Majdoubi, M., Poulain, D.A., Papadopoulos, G.C. and Theodosis, D.T. (1997) The noradrenergic innervation of identified hypothalamic magnocellular somata and its contribution to lactation-induced synaptic plasticity. J. Neuroendocrinol., 9: 17–23. Milner, T.A., Kurucz, O.S., Veznedaroglu, E. and Pierce, J.P. (1995) Septohippocampal neurons in the rat septal complex have substantial glial coverage and receive direct contacts from noradrenaline terminals. Brain Res., 670: 121–136. Miyata, S., Matsushima, O. and Hatton, G.I. (1997) Taurine in rat posterior pituitary: localization in astrocytes and selective release by hypoosmotic stimulation. J. Comp. Neurol., 381: 513–523. Modney, B.K. and Hatton, G.I. (1989) Multiple synapse formation: a possible compensatory mechanism for increased cell size in rat supraoptic nucleus. J. Neuroendocrinol., 1: 21–27. Montagnese, C., Poulain, D.A. and Theodosis, D.T. (1990) Influence of ovarian steroids on the ultrastructural plasticity of the adult rat supraoptic nucleus induced by central administration of oxytocin. J. Neuroendocrinol., 2: 225–231. Moos, F. and Richard, P. (1989) Paraventricular and supraoptic bursting oxytocin cells in rat are locally regulated by oxytocin and functionally related. J. Physiol., 408: 1–18. Moos, F., Poulain, D.A., Rodriguez, F., Guerne, Y., Vincent, J.D. and Richard, P. (1989) Release of oxytocin within the supraoptic nucleus during the milk ejection reflex in rats. Exp. Brain Res., 76: 593–602. Moos, F., Gouzènes, L., Brown, D., Dayanithi, G., Sabatier, N., Boissin, L., Rabie, A. and Richard, P. (1998) New aspects of firing pattern autocontrol in oxytocin and vasopressin neurones. Adv. Exp. Med. Biol., 449: 153–162. Mudrick-Donnon, L.A., Williams, P.J., Pittman, Q.J. and MacVicar, B.A. (1993) Postsynaptic potentials mediated by GABA and dopamine evoked in stellate glial cells of the pituitary pars intermedia. J. Neurosci., 13: 4660–4668. Nakai, S., Furuya, K., Miyata, S. and Kiyohara, T. (1999) Intracellular Ca2+ responses to nucleotides, peptides, amines, amino acids and prostaglandins in cultured pituicytes from adult rat neurohypophysis. Neurosci. Lett., 266: 185–188. Newman, E.A. (2001) Propagation of intercellular calcium waves in retinal astrocytes and Muller cells. J. Neurosci., 21: 2215– 2223. Newman, E.A. and Zahs, K.R. (1997) Calcium waves in retinal glial cells. Science, 275: 844–847. Nilius, B., Eggermont, J., Voets, T., Buyse, G., Manolopoulos, V. and Droogmans, G. (1997) Properties of volume-regulated anion channels in mammalian cells. Prog. Biophys. Mol. Biol., 68: 69–119. Oliet, S.H.R. and Bourque, C.W. (1994) Osmoreception in magnocellular neurosecretory cells: from single channels to secretion. Trends Neurosci., 17: 340–344. Oliet, S.H.R., Piet, R. and Poulain, D.A. (2001) Control of glutamate clearance and synaptic efficacy by glial coverage of neurons. Science, 292: 923–926. Parri, H.R., Gould, T.M. and Crunelli, V. (2001) Spontaneous astrocytic Ca2+ oscillations in situ drive NMDAR-mediated neuronal excitation. Nat. Neurosci., 4: 803–812.

Pasantes-Morales, H. and Schousboe, A. (1997) Role of taurine in osmoregulation in brain cells: mechanisms and functional implications. Amino Acids, 12: 281–292. Paspalas, C.D. and Papadopoulos, G.C. (1996) Ultrastructural relationships between noradrenergic nerve fibers and nonneuronal elements in the rat cerebral cortex. Glia, 17: 133– 146. Pasti, L., Zonta, M., Pozzan, T., Vicini, S. and Carmignoto, G. (2001) Cytosolic calcium oscillations in astrocytes may regulate exocytotic release of glutamate. J. Neurosci., 21: 477–484. Pittman, Q.J., Hirasawa, M., Mouginot, D. and Kombian, S.B. (2000) Neurohypophysial peptides as retrograde transmitters in the supraoptic nucleus of the rat. Exp. Physiol., 85S: 139S– 143S. Porter, J.T. and McCarthy, K.D. (1997) Astrocytic neurotransmitter receptors in situ and in vivo. Prog. Neurobiol., 51: 439– 455. Pow, D.V. (1993) Immunocytochemistry of amino-acids in the rodent pituitary using extremely specific, very high titre antisera. J. Neuroendocrinol., 5: 349–356. Pow, D.V. and Morris, J.F. (1989) Dendrites of hypothalamic magnocellular neurons release neurohypophysial peptides by exocytosis. Neuroscience, 32: 435–439. Rambotti, M.G., Mughetti, D. and Spreca, A. (1994) Detection of particulate guanylate cyclase in rat neurohypophysis after stimulation with ANF and BNP: an ultracytochemical study. Brain Res., 644: 52–58. Rampon, C., Luppi, P.H., Fort, P., Peyron, C. and Jouvet, M. (1996) Distribution of glycine immunoreactive cell bodies and fibers in the rat brain. Neuroscience, 75: 737–755. Ramsell, K.D. and Cobbett, P. (1996) Nitric oxide induces morphological changes in cultured neurohypophysial astrocytes. J. Neuroendocrinol., 8: 235–240. Randle, J.C., Bourque, C.W. and Renaud, L.P. (1986) Characterization of spontaneous and evoked inhibitory postsynaptic potentials in rat supraoptic neurosecretory neurons in vitro. J. Neurophysiol., 56: 1703–1717. Reymond, I., Almarghini, K. and Tappaz, M. (1996) Immunocytochemical localization of cysteine sulfinate decarboxylase in astrocytes in the cerebellum and hippocampus: a quantitative double immunofluorescence study with glial fibrillary acidic protein and S-100 protein. Neuroscience, 75: 619–633. Ridet, J.L., Rajaofetra, N., Teilhac, J.R., Geffard, M. and Privat, A. (1993) Evidence for nonsynaptic serotonergic and noradrenergic innervation of the rat dorsal horn and possible involvement of neuron–glia interactions. Neuroscience, 52: 143–157. Robitaille, R. (1998) Modulation of synaptic efficacy and synaptic depression by glial cells at the frog neuromuscular junction. Neuron, 21: 847–855. Salm, A.K. (2000) Mechanisms of glial retraction in the hypothalamo-neurohypophysial system of the rat. Exp. Physiol., 85S: 197S–202S. Scemes, E., Suadicani, S.O. and Spray, D.C. (2000) Intercellular communication in spinal cord astrocytes: fine tuning between

112 gap junctions and P2 nucleotide receptors in calcium wave propagation. J. Neurosci., 20: 1435–1445. Schafer, M.K., Day, R., Ortega, M.R., Akil, H. and Watson, S.J. (1990) Proenkephalin messenger RNA is expressed both in the rat anterior and posterior pituitary. Neuroendocrinology, 51: 444–448. Schrader, L.A. and Tasker, J.G. (1997) Presynaptic modulation by metabotropic glutamate receptors of excitatory and inhibitory synaptic inputs to hypothalamic magnocellular neurons. J. Neurophysiol., 77: 527–536. Sheikh, S.P., Feldthus, N., Orkild, H., Goke, R., McGregor, G.P., Turner, D., Moller, M. and Stuenkel, E.L. (1998) Neuropeptide Y2 receptors on nerve endings from the rat neurohypophysis regulate vasopressin and oxytocin release. Neuroscience, 82: 107–115. Smithson, K.G., Suarez, I. and Hatton, G.I. (1990) Betaadrenergic stimulation decreases glial and increases neural contact with the basal lamina in rat neurointermediate lobes incubated in vitro. J. Neuroendocrinol., 2: 693–699. Stern, J.E. and Armstrong, W.E. (1998) Reorganization of the dendritic trees of oxytocin and vasopressin neurons of the rat supraoptic nucleus during lactation. J. Neurosci., 18: 841–853. Stuenkel, E.L., Dayanithi, G. and Nordmann, J.J. (1996) Arachidonic acid regulation of vasopressin release and intracellular Ca2+ in neurohypophysial nerve endings. Brain Res., 742: 129–140. Theodosis, D.T. and MacVicar, B. (1996) Neurone–glia interactions in the hypothalamus and pituitary. Trends Neurosci., 19: 363–367. Theodosis, D.T. and Poulain, D.A. (1993) Activity-dependent neuronal–glial and synaptic plasticity in the adult mammalian hypothalamus. Neuroscience, 57: 501–535. Theodosis, D.T., Chapman, D.B., Montagnese, C., Poulain, D.A. and Morris, J.F. (1986a) Structural plasticity in the hypothalamic supraoptic nucleus at lactation affects oxytocin-, but not vasopressin-secreting neurones. Neuroscience, 17: 661–678. Theodosis, D.T., Montagnese, C., Rodriguez, F., Vincent, J.D. and Poulain, D.A. (1986b) Oxytocin induces morphological plasticity in the adult hypothalamo-neurohypophysial system. Nature, 322: 738–740. Theodosis, D.T., El Majdoubi, M., Pierre, K. and Poulain, D.A. (1998) Factors governing activity-dependent structural plasticity of the hypothalamoneurohypophysial system. Cell. Mol. Neurobiol., 18: 285–298. Troadec, J.-D., Thirion, S., Nicaise, G., Lemos, J.R. and Dayanithi, G. (1998) ATP-evoked increases in [Ca2+ ]i and peptide

release from rat isolated neurohypophysial terminals via a P2X2 purinoceptor. J. Physiol., 511: 89–103. Troadec, J.-D., Thirion, S., Petturiti, D., Bohn, M.T. and Poujeol, P. (1999) ATP acting on P2Y receptors triggers calcium mobilization in primary cultures of rat neurohypophysial astrocytes (pituicytes). Pflügers Arch., 437: 745–753. Troadec, J.-D., Thirion, S., Petturiti, D. and Poujeol, P. (2000) Potassium efflux triggered by P2Y purinoceptor activation in cultured pituicytes. Pflügers Arch., 440: 770–777. Tweedle, C.D., Smithson, K.G. and Hatton, G.I. (1993) Rapid synaptic changes and bundling in the supraoptic dendritic zone of the perfused rat brain. Exp. Neurol., 124: 200–207. van den Pol, A.N., Finkbeiner, S.M. and Cornell-Bell, A.H. (1992) Calcium excitability and oscillations in suprachiasmatic nucleus neurons and glia in vitro. J. Neurosci., 12: 2648–2664. van Leeuwen, F.W., Pool, C.W. and Sluiter, A.A. (1983) Enkephalin immunoreactivity in synaptoid elements on glial cells in the rat neural lobe. Neuroscience, 8: 229–241. Verkhratsky, A., Orkand, R.K. and Kettenmann, H. (1998) Glial calcium: homeostasis and signaling function. Physiol. Rev., 78: 99–141. Voisin, D.L., Chakfe, Y. and Bourque, C.W. (1999) Coincident detection of CSF Na+ and osmotic pressure in osmoregulatory neurons of the supraoptic nucleus. Neuron, 24: 453–460. Vuillez, P., Perez, S.C. and Stoeckel, M.E. (1987) Colocalization of GABA and tyrosine hydroxylase immunoreactivities in the axons innervating the neurointermediate lobe of the rat pituitary: an ultrastructural immunogold study. Neurosci. Lett., 79: 53–58. Watt, J.A. and Hobbs, N.K. (2000) Interleukin-1β immunoreactivity in identified neurons of the rat magnocellular neurosecretory system: evidence for activity-dependent release. J. Neurosci. Res., 60: 478–489. Wittkowski, W. (1998) Tanycytes and pituicytes: morphological and functional aspects of neuroglial interaction. Microsc. Res. Tech., 41: 29–42. Wittkowski, W. and Brinkmann, H. (1974) Changes of extent of neuro-vascular contacts and number of neuro-glial synaptoid contacts in the pituitary posterior lobe of dehydrated rats. Anat. Embryol., 146: 157–165. Wuarin, J.-P. and Dudek, F.E. (1993) Patch-clamp analysis of spontaneous synaptic currents in supraoptic neuroendocrine cells of the rat hypothalamus. J. Neurosci., 13: 2323–2331. Yasin, S.A., Costa, A., Forsling, M.L. and Grossman, A. (1994) Interleukin-1β and interleukin-6 stimulate neurohypophysial hormone release in vitro. J. Neuroendocrinol., 6: 179–184.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 9

Functional synaptic plasticity in hypothalamic magnocellular neurons Jeffrey G. Tasker ∗ , Shi Di and Cherif Boudaba Department of Cell and Molecular Biology, Neurobiology Division, Tulane University, New Orleans, LA, USA

Abstract: The hypothalamic–neurohypophysial system undergoes dramatic morphological plasticity in response to physiological activation during parturition/lactation and dehydration, including somatic swelling, decreased glial coverage and increased synaptic innervation of the magnocellular neuroendocrine cells. Recent in-vitro electrophysiological studies in hypothalamic slices have demonstrated that coordinate changes in the synaptic physiology of the magnocellular neurons also occur under these conditions. Thus, the synaptic release of glutamate and GABA onto magnocellular neurons is increased during lactation and with chronic dehydration, and changes in postsynaptic glutamate and GABAA receptor expression lead to alterations of the functional properties of the glutamate and GABAA receptor channels. The presynaptic noradrenergic facilitation of glutamate release and inhibition of GABA release is also markedly enhanced following chronic dehydration. Additionally, both parturition and chronic dehydration are accompanied by an increase in the tonic activation of presynaptic metabotropic glutamate receptors due to the higher ambient glutamate concentration caused by decreased glial coverage and the resultant reduction in glutamate reuptake. Together, these electrophysiological studies reveal profound functional plasticity in the synaptic physiology of magnocellular neurons at parturition and following dehydration. The plastic changes support an increase in the excitability of magnocellular neuroendocrine cells by increasing glutamate inputs, decreasing GABA inputs, enhancing excitatory noradrenergic modulation, and reducing synaptic glutamatergic noise. Keywords: Supraoptic nucleus; Paraventricular nucleus; EPSC; IPSC; Glutamate; GABA; Norepinephrine; Noradrenaline; Metabotropic receptor

Introduction Magnocellular neuroendocrine cells are located principally in the supraoptic nucleus (SON) and paraventricular nucleus (PVN) of the hypothalamus and synthesize the neuropeptides oxytocin and vasopressin. Oxytocin and vasopressin are secreted into the bloodstream from secretory terminals in the neural lobe of the pituitary gland; oxytocin is

∗ Correspondence to: J.G. Tasker, Department of Cell and Molecular Biology, 2000 Percival Stern Hall, Tulane University, New Orleans, LA 70118, USA. E-mail: [email protected]

secreted during parturition and lactation and vasopressin under conditions of hypotension and dehydration. Release of oxytocin and vasopressin is regulated by the electrical activities of the respective hormone-secreting magnocellular neurons and is maximal with bursting patterns of activity, the cyclicity of which differs between the two groups of neuroendocrine cells. The magnocellular neuroendocrine system is highly plastic, and is subject to dramatic morphological changes with fluctuating physiological conditions; these changes are thought to alter the excitability of the neuroendocrine cells and the release of neurohormone under conditions of increased hormone demand.

114

Morphological plasticity of the magnocellular neuroendocrine system The magnocellular neuroendocrine cells and the astrocytes of the SON and PVN, as well as the neurosecretory terminals and pituicytes of the neural lobe, undergo profound, yet reversible, structural changes at parturition, under conditions of chronic dehydration, and with chronic stress. These changes include reorganization of the structural relationship between the magnocellular neurons and glial cells, changes in dendritic morphology, and an increase in the numbers of chemical and electrical synapses. Neuronal–glial structural plasticity One of the most striking morphological changes that occurs with stimulation of the system at parturition and by dehydration is the marked increase in the somatic size of the magnocellular neurons, which swell by up to 100% (Theodosis et al., 1986; Miyata et al., 1994b). This increase in somatic size is accompanied by a decrease in glial coverage of the magnocellular neuronal somata and dendrites and a corresponding increase in the extent of direct membrane apposition between neighboring magnocellular neurons (for review, see Theodosis and Poulain, 1993; Hatton, 1997). A similar decrease in pituicyte coverage of neurosecretory terminals in the neural lobe is also seen under these conditions (for review, see Hatton, 1997).

synapses (Gies and Theodosis, 1994; El Majdoubi et al., 1996, 1997; Michaloudi et al., 1997). Increased numbers of synapses have also been reported to occur with chronic dehydration (Miyata et al., 1994b) and repeated stress (Miyata et al., 1994a). There is also an increased incidence during lactation of glutamate and GABA synapses formed between single presynaptic terminals and two postsynaptic profiles, referred to as shared synapses (for review, see Hatton, 1997; El Majdoubi et al., 2000). In addition to changes in the total numbers of glutamate, GABA and norepinephrine synapses during lactation and with chronic dehydration, there is also evidence for changes in postsynaptic receptor expression. These changes include altered NMDA and metabotropic glutamate-receptor expression. Dehydration and lactation are accompanied by an upregulation of NMDA NR1 and NR2A receptors and a downregulation of NR2B receptors (Meeker et al., 1994; Decavel and Currás, 1997; CurrásCollazo and Dao, 1999; Peiris et al., 2001), and metabotropic glutamate receptor expression has been reported to increase with dehydration (Meeker et al., 1994). GABAA receptor expression also changes during parturition and lactation, characterized by an increase in the ratio of α2 to α1 receptor subunits (Brussaard et al., 1997). Functional plasticity of the synaptic regulation of magnocellular neuroendocrine cells Plasticity of fast synaptic regulation

Morphological plasticity of synaptic innervation Glutamate synapses Coincident with and perhaps triggered by the decrease in glial coverage of the magnocellular somata and dendrites at parturition and with chronic dehydration is an increase in the number of synaptic contacts impinging on the magnocellular neurons (for review, see Theodosis and Poulain, 1993; Hatton, 1997). Ultrastructural analyses have shown that the density of synapses on magnocellular neurons remains constant despite the increase in somatic surface area under stimulated conditions, indicating an increase in the numbers of synapses (El Majdoubi et al., 1997). Synapses implicated in the morphological changes in synapse numbers seen during lactation include glutamate, GABA and norepinephrine

Recent studies using whole-cell recordings in hypothalamic slices have revealed functional changes in glutamatergic synaptic inputs to magnocellular neurons that parallel the morphological changes in glutamate synapse numbers seen in the SON of lactating and chronically dehydrated rats. The frequency of glutamate-mediated miniature excitatory postsynaptic currents (EPSCs) increased over twofold in slices from lactating rats compared to those from non-lactating rats, without any increase in EPSC amplitude, indicative of an increase in glutamate release (Stern et al., 2000). The increase in miniature EPSC frequency in this study was accom-

115

panied by a decrease in paired-pulse facilitation in the lactating animals, suggesting that there was an increase in the probability of release from glutamatergic synapses during lactation. Another study found that paired-pulse facilitation was increased in SON neurons during lactation, suggesting a decrease in the probability of glutamate release (Oliet et al., 2001), such that it has not yet been established whether the increase in glutamate release during lactation is due exclusively to the increased number of glutamate synapses or to a combination of more synapses and an increased probability of release from individual synapses. Changes in glutamate release onto magnocellular neurons also have been observed with chronic dehydration, although they are less robust than those seen during lactation. Dehydration gave rise to an average increase of 27% in the frequency of EPSCs in magnocellular neurons, indicative of an increase in glutamate release (Di and Tasker, 1999). No significant change in paired-pulse facilitation was seen in these cells compared to cells from untreated control animals, suggesting that the increase in glutamate release was not due to a change in the release probability of glutamate, but rather to an increase in the number of glutamate release sites. These cells also showed a 20% increase in whole-cell capacitance with dehydration, reflecting a corresponding change in surface area, which is consistent with an increase in synapse number. Thus, the functional plasticity of glutamate synapses parallels the structural plasticity seen in morphological studies. GABA synapses As mentioned earlier, GABA synapses impinging on magnocellular neurons also increase in number at parturition and with dehydration, and the GABAA receptor subunit expression changes to increase the ratio of α2 to α1 subunits at parturition. Whole-cell recordings from magnocellular neurons in hypothalamic slices from lactating rats revealed a ∼50% increase in the frequency of GABA-mediated miniature inhibitory postsynaptic currents (IPSCs) compared to cells from non-lactating rats (Brussaard et al., 1999), indicative of an increase in GABA release that parallels the increased numbers of GABA synapses observed in morphological studies. Al-

though no apparent change in the synaptic GABAA receptor density was found (Brussaard et al., 1999), the switch in the ratio of α1 to α2 receptor subunits during lactation caused postsynaptic alteration of channel function, resulting in a decrease in the decay kinetics of GABA synaptic currents and increased charge transfer by the GABAA receptor channels (Brussaard et al., 1997). GABA release also increased in SON magnocellular neurons in slices from rats subjected to chronic dehydration with hypertonic saline, although to a lesser extent than in SON neurons from lactating rats. Whole-cell recordings in magnocellular neurons from dehydrated rats showed an average increase of 22% in the frequency of IPSCs, also suggestive of an increase in the number of GABA synapses (Di and Tasker, 1999). This increase in GABA release paralleled a 20% increase in surface area of these cells calculated from whole-cell capacitance measurements, which is consistent with an increase in synapse numbers. The electrophysiological findings in slices from dehydrated and lactating rats, therefore, provide a physiological corollary to the increased numbers of glutamate and GABA synapses on magnocellular neurons reported in morphological studies of the hypothalamic–neurohypophysial plasticity. The concomitant increases in glutamate and GABA inputs might appear at first to be counteractive, since the coordinate upregulation of both the excitatory and inhibitory drives to the magnocellular neurons would seem to cancel each other out, resulting in no net change in excitability (although the measured increase in excitatory glutamate inputs is slightly greater than the increase in inhibitory inputs, which would shift the balance toward increased excitability). The parallel increases in glutamate and GABA inputs to the magnocellular neurons implies that both synaptic excitation and synaptic inhibition are involved in the activation of magnocellular neurons under conditions of increased hormone demand, and that the increase in hormone output under these conditions is a function of subtle weighting of excitation and inhibition. Consistent with this are the findings that osmotic stimulation elicits the release of both glutamate and GABA in the SON (Neumann et al., 1995; Leng et al., 2001), and leads to a progressive, linear increase in the spiking activity of both oxy-

116

tocin and vasopressin neurons that can be modeled only if both excitatory and inhibitory synaptic inputs are included in the model (Leng et al., 2001). Synaptic modulatory mechanisms, therefore, are likely to play a critical role in fashioning subtle or gradual changes in magnocellular electrical activity in response to fluctuating synaptic inputs caused by dynamic changes in sensory stimulation. Plasticity of synaptic modulation Noradrenergic modulation of glutamate and GABA release Activation of presynaptic noradrenergic receptors has been shown with whole-cell recordings to facilitate glutamate release and to suppress GABA release onto magnocellular neurons in the SON and PVN (Boudaba et al., 1999; Wang et al., 1998), and to activate local glutamate circuits in the PVN (Daftary et al., 1998). A dramatic enhancement of the noradrenergic modulation of glutamate and GABA release onto magnocellular neurons was found in slices from chronically dehydrated rats (Di and

Tasker, 1999). Norepinephrine (NE) caused a reversible, dose-dependent increase in the frequency of glutamatergic EPSCs in magnocellular neurons recorded in slices from both untreated and dehydrated rats, but the NE-induced increase in EPSC frequency was two-fold higher in cells from dehydrated rats (∼250% and ∼500%, respectively). The NE dose responsiveness was shifted to the left after dehydration, indicating a greater sensitivity to NE. The dose-dependent decrease in the frequency of GABA-mediated IPSCs was also enhanced after dehydration, from a ∼50% decrease in IPSC frequency elicited by NE in magnocellular neurons from untreated rats to a ∼75% decrease in magnocellular neurons from dehydrated rats. The dose responsiveness of the NE effect on GABA release was also shifted toward lower concentrations of NE, suggesting a higher sensitivity to NE after dehydration. These changes in noradrenergic modulation of glutamate and GABA release following chronic dehydration indicate that there is an increase in the sensitivity of glutamate and GABA release mechanisms to NE, and suggest that noradrenergic receptors are upregulated by chronic dehydration (Fig. 1)

Fig. 1. Enhanced noradrenergic modulation of glutamate and GABA release onto magnocellular neurons following chronic dehydration. Activation of adrenergic receptors located on presynaptic glutamate somata and terminals and on presynaptic GABA terminals elicits an increase in glutamate release and a decrease in GABA release onto magnocellular neurons under unstimulated conditions (Unstimulated). Chronic dehydration leads to a two-fold increase in the noradrenergic facilitation of glutamate release and suppression of GABA release (Stimulated). These changes appear to be due to the proliferation of glutamate and GABA synapses with chronic dehydration, as well as to the upregulation of noradrenergic receptors. It is likely that the noradrenergic receptors are activated by inputs from the A1/A2 noradrenergic cells groups of the brainstem.

117

(Di and Tasker, 1999). This is consistent with an increase in the number of NE synapses seen during lactation with morphological analyses (Michaloudi et al., 1997), and suggests that a similar increase in noradrenergic synapse number occurs with chronic dehydration. The observation that NE elicited a proportionally larger increase in the release of glutamate and decrease in the release of GABA in magnocellular neurons from dehydrated rats (Di and Tasker, 1999) suggests that there is a disproportionate upregulation of presynaptic noradrenergic receptors (i.e. hypersensitization) in the SON with dehydration (Fig. 1). This coordinate upregulation of the positive modulation of glutamate release and negative modulation of GABA release by NE may tip the synaptic excitation–inhibition balance sufficiently toward excitation to provide the increase in excitability of the magnocellular neurons responsible for enhanced hormone release under stimulated conditions. It will

be interesting in the future to determine whether these plastic changes in noradrenergic modulation of glutamate and GABA release are specific to conditions of dehydration, or whether they also occur with stimulation of the hypothalamic–neurohypophysial system during parturition and lactation. Glutamate modulation of glutamate release As discussed earlier, the structural plasticity of the hypothalamic–neurohypophysial system includes retraction of astrocytic processes from around the magnocellular neurons (for review, see Theodosis and Poulain, 1993; Hatton, 1997). Because astrocytes play a major role in clearing glutamate from the synapse (Rothstein et al., 1996) and blockade of astrocytic glutamate transporters causes spillover of glutamate into extrasynaptic sites (Scanziani et al., 1997; Mitchell and Silver, 2000), the neuronal–glial

Fig. 2. Glial retraction leads to an increase in the activation of presynaptic metabotropic glutamate receptors (mGluRs). Presynaptic group III mGluRs are activated tonically at a low level by ambient extracellular glutamate under non-stimulated conditions, in which astrocytic processes provide maximal coverage of the magnocellular neurons (Unstimulated). During lactation and following chronic dehydration, retraction of astrocytic processes leads to decreased glutamate uptake resulting in increased ambient levels of extracellular glutamate, which causes increased tonic activation of presynaptic mGluRs (Stimulated). The addition of new glutamate synapses during lactation and dehydration contributes to the increase in ambient glutamate levels.

118

plasticity in the hypothalamic–neurohypophysial system would be expected to augment extracellular glutamate concentrations and increase the activation of extrasynaptic glutamate receptors. The increased numbers of glutamate synapses and increased release of glutamate under stimulated conditions could also contribute to higher ambient levels of glutamate. Glutamate and GABA terminals in the SON express group III metabotropic glutamate receptors (mGluRs), the activation of which leads to decreased glutamate and GABA release onto SON magnocellular neurons (Schrader and Tasker, 1997). The presynaptic mGluRs on glutamate terminals in the SON are activated tonically by ambient levels of extracellular glutamate, which are under the regulatory control of glutamate transporters (Schrader and Tasker, 1997; Linn and Tasker, 1999; Oliet et al., 2001). As hypothesized from morphological studies, an increase in the tonic activation of the presynaptic mGluRs was observed in slices from lactating rats (Oliet et al., 2001) and chronically dehydrated rats (Boudaba et al., 2001). The activation of presynaptic mGluRs by endogenous glutamate was less sensitive to blockade of astrocytic glutamate reuptake in slices from stimulated animals, indicative of a decrease in glutamate transporter activity. These findings suggest that extracellular glutamate levels in the SON are increased in lactating and dehydrated rats due to a decrease in glutamate clearance by glial transporters (Fig. 2), and are consistent with anatomical observations of glial retraction and reduced glial coverage of magnocellular neuronal membrane, including synapses, under these conditions. Increased tonic activation of presynaptic mGluRs would be expected to increase the signal-to-noise ratio of the system by reducing spontaneously released glutamate, thus enhancing the responsiveness of magnocellular neurons to physiologically salient inputs. This form of noise suppression is particularly important given the increased numbers of glutamate synapses and the resulting increase in basal glutamate release under stimulated conditions (Fig. 2). It is clear from these recent in vitro electrophysiological studies that the structural plasticity in the neuronal–glial and synaptic organization of the hypothalamic–neurohypophysial system seen during lactation and with dehydration results in pronounced changes in the synaptic physiology of the magnocel-

lular neuroendocrine cells. Although these changes include upregulation of both glutamate-mediated excitatory and GABA-mediated inhibitory synaptic mechanisms, parallel alterations of noradrenergic modulatory mechanisms are strongly excitatory and may contribute in a critical fashion to the increase in magnocellular neuronal excitability seen under certain physiological conditions of heightened demand for oxytocin and vasopressin secretion. Abbreviations EPSC GABA IPSC NE NMDA NR1, NR2A, NR2B mGluR PVN SON

excitatory postsynaptic current γ-aminobutyric acid inhibitory postsynaptic current norepinephrine N-methyl-D-aspartate NMDA receptor subunits 1, 2A and 2B metabotropic glutamate receptor paraventricular nucleus supraoptic nucleus

Acknowledgements We are indebted to Kriszta Szabó and Katalin Halmos for their excellent technical support. This work was supported by NIH Grants NS31187 and NS34926, and by American Heart Association Grant 96010150. References Boudaba, C., Di, S. and Tasker, J.G. (1999) Noradrenergic modulation of excitatory synaptic inputs to supraoptic and paraventricular magnocellular neurons in rat hypothalamic slices. Soc. Neurosci. Abstr., 25: 704. Brussaard, A.B., Kits, K.S., Baker, R.E., Willems, W.P., LeytingVermeulen, J.W., Voorn, P., Smit, A.B., Bicknell, R.J. and Herbison, A.E. (1997) Plasticity in fast synaptic inhibition of adult oxytocin neurons caused by switch in GABA(A) receptor subunit expression. Neuron, 19: 1103–1114. Brussaard, A.B., Devay, P., Leyting-Vermeulen, J.L. and Kits, K.S. (1999) Changes in properties and neurosteroid regulation of GABAergic synapses in the supraoptic nucleus during the mammalian female reproductive cycle. J. Physiol., 516: 513– 524. Currás-Collazo, M.C. and Dao, J. (1999) Osmotic activation of the hypothalamo–neurohypophysial system reversibly downregulates the NMDA receptor subunit, NR2B, in the supraoptic nucleus of the hypothalamus. Brain Res. Mol. Brain Res., 70: 187–196.

119 Daftary, S.S., Boudaba, C., Szabó, K. and Tasker, J.G. (1998) Noradrenergic excitation of magnocellular neurons in the rat hypothalamic paraventricular nucleus via intranuclear glutamatergic circuits. J. Neurosci., 18: 10619–10628. Decavel, C. and Currás, M.C. (1997) Increased expression of the N-methyl- D-aspartate receptor subunit, NR1, in immunohistochemically identified magnocellular hypothalamic neurons during dehydration. Neuroscience, 78: 191–202. Di, S. and Tasker, J.G. (1999) Plasticity of the noradrenergic modulation of hypothalamic magnocellular neurons during chronic dehydration. Soc. Neurosc. Abstr., 25: 705. El Majdoubi, M., Poulain, D.A. and Theodosis, D.T. (1996) The glutamatergic innervation of oxytocin- and vasopressinsecreting neurons in the rat supraoptic nucleus and its contribution to lactation-induced synaptic plasticity. Eur. J. Neurosci., 8: 1377–1389. El Majdoubi, M., Poulain, D.A. and Theodosis, D.T. (1997) Lactation-induced plasticity in the supraoptic nucleus augments axodendritic and axosomatic gabaergic and glutamatergic synapses: an ultrastructural analysis using the disector method. Neuroscience, 80: 1137–1147. El Majdoubi, M., Poulain, D.A. and Theodosis, D.T. (2000) Activity-dependent morphological synaptic plasticity in an adult neurosecretory system: magnocellular oxytocin neurons of the hypothalamus. Biochem. Cell Biol., 78: 1–11. Gies, U. and Theodosis, D.T. (1994) Synaptic plasticity in the rat supraoptic nucleus during lactation involves GABA innervation and oxytocin neurons: a quantitative immunocytochemical analysis. J. Neurosci., 14: 2861–2869. Hatton, G.I. (1997) Function-related plasticity in hypothalamus. Annu. Rev. Neurosci., 20: 375–397. Leng, G., Brown, C.H., Bull, P.M., Brown, D., Scullion, S., Currie, J., Blackburn-Munro, R.E., Feng, J., Onaka, T., Verbalis, J.G., Russell, J.A. and Ludwig, M. (2001) Responses of magnocellular neurons to osmotic stimulation involves coactivation of excitatory and inhibitory input: an experimental and theoretical analysis. J. Neurosci., 21: 6967–6977. Linn, D.M. and Tasker, J.G. (1999) Ambient glutamate levels regulate excitatory neurotransmission via presynaptic metabotropic receptors in hypothalamic neurons. Soc. Neurosci. Abstr., 25: 448. Meeker, R.B., McGinnis, S., Greenwood, R.S. and Hayward, J.N. (1994) Increased hypothalamic glutamate receptors induced by water deprivation. Neuroendocrinology, 60: 477–485. Michaloudi, H.C., El Majdoubi, M., Poulain, D.A., Papadopoulos, G.C. and Theodosis, D.T. (1997) The noradrenergic innervation of identified hypothalamic magnocellular somata and its contribution to lactation-induced synaptic plasticity. J. Neuroendocrinol., 9: 17–23. Mitchell, S.J. and Silver, R.A. (2000) Glutamate spillover suppresses inhibition by activating presynaptic mGluRs. Nature, 404: 498–502.

Miyata, S., Itoh, T., Matsushima, O., Nakashima, T. and Kiyohara, T. (1994a) Not only osmotic stress but also repeated restraint stress causes structural plasticity in the supraoptic nucleus of the rat hypothalamus. Brain Res. Bull., 33: 669–675. Miyata, S., Nakashima, T. and Kiyohara, T. (1994b) Structural dynamics of neural plasticity in the supraoptic nucleus of the rat hypothalamus during dehydration and rehydration. Brain Res. Bull., 34: 169–175. Neumann, I., Landgraf, R., Bauce, L. and Pittman, Q.J. (1995) Osmotic responsiveness and cross talk involving oxytocin, but not vasopressin or amino acids, between the supraoptic nuclei in virgin and lactating rats. J. Neurosci., 15: 3408–3417. Oliet, S.H., Piet, R. and Poulain, D.A. (2001) Control of glutamate clearance and synaptic efficacy by glial coverage of neurons. Science, 292: 923–926. Peiris, P., Payumo, A., Stivers, C., Stanley, B.G. and CurrásCollazo, M.C. (2001) Glutamate receptor subunits of the AMPA and NMDA receptor subclasses, GluR4 and NMDAR2A, are synthesized by the supraoptic and paraventricular nuclei of the rat hypothalamus. Soc. Neurosci. Abstr., 26: 179.7. Rothstein, J.D., Dykes-Hoberg, M., Pardo, C.A., Bristol, L.A., Jin, L., Kunci, R.W., Kanai, Y., Hediger, M.A., Wang, Y., Schielke, J.P. and Welty, D.F. (1996) Knockout of glutamate transporters reveals a major role for astroglial transport in excitotoxicity and clearance of glutamate. Neuron, 16: 675– 686. Scanziani, M., Salin, P.A., Vogt, K.E., Malenka, R.C. and Nicoll, R.A. (1997) Use-dependent increases in glutamate concentration activate presynaptic metabotropic glutamate receptors. Nature, 385: 630–634. Schrader, L.A. and Tasker, J.G. (1997) Presynaptic modulation by metabotropic glutamate receptors of excitatory and inhibitory synaptic inputs to hypothalamic magnocellular neurons. J. Neurophysiol., 77: 527–536. Stern, J.E., Hestrin, S. and Armstrong, W.E. (2000) Enhanced neurotransmitter release at glutamatergic synapses on oxytocin neurones during lactation in the rat. J. Physiol., 526: 109–114. Theodosis, D.T. and Poulain, D.A. (1993) Activity-dependent neuronal–glial and synaptic plasticity in the adult mammalian hypothalamus. Neuroscience, 57: 501–535. Theodosis, D.T., Chapman, D.B., Montagnese, C., Poulain, D.A. and Morris, J.F. (1986) Structural plasticity in the hypothalamic supraoptic nucleus at lactation affects oxytocin- but not vasopressin-secreting neurones. Neuroscience, 17: 661–678. Wang, Y.-F., Shibuya, I., Kabashima, N., Setiadji, V.S., Isse, T., Ueta, Y. and Yamashita, H. (1998) Inhibition of spontaneous postsynaptic currents (IPSC) by noradrenaline in rat supraoptic neurons through presynaptic α2 -adrenoceptors. Brain Res., 807: 61–69.

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 10

Postsynaptic GABAB receptors in supraoptic oxytocin and vasopressin neurons J.E. Stern ∗ , Y. Li and D.S. Richards Department of Pharmacology and Toxicology, Wright State University, 3640 Colonel Glenn Highway, Dayton, OH 45435, USA

Keywords: Supraoptic; Oxytocin; Vasopressin; GABA; Synaptic transmission

Introduction Hormone release from neurohypophysial terminals is closely related to the firing activity of magnocellular neurons in the supraoptic (SON) and paraventricular (PVN) nuclei (Poulain and Wakerley, 1982). Even though oxtytocin (OT) and vasopressin (VP) neurons are known to adopt different firing activities in response to specific stimuli (Lincoln and Wakerley, 1974; Poulain et al., 1977), the precise mechanisms involved in the generation of these different firing patterns are still not completely understood. Considering that the electrical activity of magnocellular neurons results from the integration of intrinsic as well as extrinsic factors, the expression of specific firing patterns may arise from cell type differences in their intrinsic properties and/or the actions of synaptic inputs. In the past years, we have shown that cellular factors known to affect neuronal excitability, including intrinsic membrane properties (Stern and Armstrong, 1995), dendritic structure (Stern and Armstrong, 1998) and the activity and properties of fast excitatory synaptic inputs (Stern et al., 1999)

∗ Correspondence

to: J.E. Stern, Department of Pharmacology and Toxicology, Wright State University, 3640 Colonel Glenn Highway, Dayton OH 45435, USA. Tel.: +1-937-775-3305; Fax: +1-937-775-7221; E-mail: [email protected]

differ between cell types. Furthermore, we showed that these factors undergo cell type specific plastic changes in response to particular physiological conditions (Stern and Armstrong, 1996, 1998; Stern et al., 2000), supporting their role in the adoption of firing patterns characteristic to each cell type. These studies were now extended to investigate whether SON OT and VP neurons also differ in their complement of postsynaptic GABAB receptors. GABA is the major inhibitory molecule in the hypothalamus (Decavel and Van den Pol, 1990), and has been shown to play an important role in the regulation of neuronal excitability in magnocellular neurons (Randle and Renaud, 1987; Nissen and Renaud, 1994; Moos, 1995). Whereas the involvement of the ionotropic GABAA receptor in mediating these actions has been clearly established, less is known about the expression and role of the metabotropic GABAB receptor. For example, presynaptic GABAB receptors located both in glutamatergic and GABAergic terminals within the SON have been shown to efficiently modulate neurotransmitter release from these terminals (Kombian et al., 1996; Kabashima et al., 1997; Mouginot et al., 1998). On the other hand, whether GABAB receptors are expressed postsynaptically in magnocellular neurons, and what their role is in controlling their excitability has been more controversial (see for example Voisin et al., 1996). In light of this controversy, we studied the expression and function of postsynaptic

122

GABAB receptors in magnocellular neurons, using a combination of in vitro patch clamp recordings and immunohistochemical approaches. Methods Immunohistochemical techniques were used to study the expression of GABAB receptor subunits in identified OT and VP neurons. Briefly, triple fluorescent immunohistochemical experiments were performed in 30 μm coronal hypothalamic brain slices obtained from Sprague–Dawley male and female rats. Antibodies raised against GABAB R1 and/or GABAB R2 receptor subunits were combined simultaneously with antibodies raised against OT and VP neurophysins. Co-localization of the three antibodies in individual SON and PVN neurons was assessed. Furthermore, GABAB receptor subunit immunoreactivity was quantified using an optical density analysis. The electrical activity of SON neurons was recorded using the perforated patch clamp configuration, which allows electrical access to the cell, preventing dialysis of the intracellular environment. Briefly, hypothalamic coronal slices containing the SON were obtained, and kept in a submerged recording chamber perfused with an artificial CSF main-

tained at 30–32°C (Stern et al., 1999). Neurons were visualized using a DIC-infrared videomicroscopy. All recordings were obtained in the presence of glutamate (10 μM CNQX and 100 μM ± APV) and GABAA (20 μM bicuculline) receptor antagonists. Results Fig. 1 shows low and high power confocal photomicrographs depicting GABAB R1 immunoreactivity in the SON of a female rat. As shown in Fig. 1A a strong, non-uniform, GABAB R1 immunoreactivity was found in SON neurons, characterized by a discrete and punctuate pattern, as shown at higher magnification in Fig. 1B. In order to determine whether the non-uniform GABAB receptor immunoreactivity was dependent on the phenotypic identity of SON neurons, we performed a triple immunofluorescence study, combining antibodies raised against GABAB R1, OT and VP neurophysins. Our results indicate that most of the GABAB immunoreactivity in the SON was confined to VP-positive neurons. A typical example is shown in Fig. 2. The differential GABAB R1 immunoreactivity in OT and VP neurons in the SON was further confirmed by an optical density analysis, revealing a

Fig. 1. GABAB R1 immunoreactivity in the SON. (A) Confocal photomicrograph displaying strong, non-uniform GABAB R1 in the SON. (B) At higher magnification, GABAB R1 staining appeared as a focal, punctuate pattern. Note also the absence of nuclear staining. GABAB R1 immunoreactivity was visualized by Cy3-conjugated secondary antibody.

123

Fig. 2. Triple immunofluorescence displaying stronger GABAB R1 immunoreactivity in VP neurons. (A) VP-neurophysin immunoreactivity visualized by FITC-conjugated secondary antibody. (B) GABAB R1 immunoreactivity visualized by Cy3-conjugated secondary antibody. (C) OT-neurophysin immunoreactivity visualized by AMCA-conjugated secondary antibody. Note the strong and weak GABAB R1 immunoreactivity in VP and OT neurons, respectively.

receptor agonist, efficiently and transiently inhibited the firing activity of magnocellular neurons. A typical example and the group data are shown in Fig. 4. Baclofen actions were blocked by the specific GABAB receptor antagonist CGP 55845 (1 μM) (results not shown). Discussion

Fig. 3. Immunofluorescence optical density analysis of GABAB R1 immunoreactivity in OT and VP SON neurons. The optical density analysis demonstrated that VP neurons expressed a significantly denser GABAB R1 immunoreactivity when compared to OT neurons, both in male and female rats. (* P < 0.05, n = 5/group).

significantly denser GABAB R1 immunoreactivity in VP neurons (Fig. 3). Similar results were observed when using a GABAB R2 specific antibody (results not shown). The expression of postsynaptic GABAB receptors in SON neurons was further studied using electrophysiological techniques. In the presence of glutamate and GABAA receptor antagonists, bath application of baclofen (1–10 μM), a specific GABAB

The presence and function of presynaptic GABAB receptors in both excitatory and inhibitory terminals within the SON has been previously addressed (Kombian et al., 1996; Kabashima et al., 1997; Mouginot et al., 1998). On the other hand, whether postsynaptic GABAB receptors are expressed in magnocellular neurons, and their role in controlling SON neuronal excitability is still controversial (Voisin et al., 1996; Harayama et al., 1998). Overall, our data provide further evidence for the expression of postsynaptic GABAB receptors in magnocellular SON neurons. Recent advances in the molecular structure of the GABAB receptor indicate that this is an unusual Gprotein-coupled receptor, composed of two seven transmembrane domain subunits, GABAB R1 and GABAB R2, respectively. Heteromerization of these subunits results in a functional receptor at the cell surface (see Blein et al., 2000 for review). The fact that

124

Fig. 4. Activation of postsynaptic GABAB receptors induces inhibition of firing activity in SON neurons. (A) Example of a pronounced and transient inhibition of firing activity in a SON neuron induced by transient application of baclofen (10 μM). Recordings were obtained in the presence of glutamate and GABAA receptor antagonists (see Section 2). (B) Grouped data showing that baclofen significantly inhibited the firing activity in SON neurons (* P < 0.01, n = 17).

we found both GABAB receptor subunits to be expressed in SON neurons, and that activation of these receptors induced inhibition of firing rate, argues in favor of the expression of functional postsynaptic GABAB receptors by SON magnocellular neurons. GABAB receptor immunoreactivity was significantly stronger in VP, as compared to OT neurons, suggesting that the expression of this receptor in the magnocellular neuroendocrine system is celltype dependent. Whether the differences in GABAB receptor immunoreactivity in SON neurons have a functional significance, and whether the expression and function of postsynaptic GABAB receptors are modulated in a state-dependent manner are important questions under current investigation.

Abbreviations APV CNQX CSF DIC GABA OT PVN SON VP

2-amino-5-phosphonopentanoic acid 6-cyano-7-nitroquinoxaline-2,3dione·2Na cerebrospinal fluid differential interference contrast γ-aminobutyric acid oxytocin paraventricular nucleus supraoptic nucleus vasopressin

125

Acknowledgements This study was supported by The American Heart Association Grant 0050441N and by the Ohio Board of Regents. We thank Dr. W. Zhang for outstanding technical assistant. This research has benefited from collaborations with Dr. F.J. Alvarez’s laboratory, at the Department of Anatomy, Wright State University, Dayton, OH. References Blein, S., Hawrot, E. and Barlow, P. (2000) The metabotropic GABA receptor: molecular insights and their functional consequences. Cell Mol. Life Sci., 57: 635–650. Decavel, C. and Van den Pol, A.N. (1990) GABA: a dominant neurotransmitter in the hypothalamus. J. Comp. Neurol., 302: 1019–1037. Harayama, N., Shibuya, I., Tanaka, K., Kabashima, N., Ueta, Y. and Yamashita, H. (1998) Inhibition of N- and P/Q-type calcium channels by postsynaptic GABAB receptor activation in rat supraoptic neurones. J. Physiol., 509: 371–383. Kabashima, N., Shibuya, I., Ibrahim, N., Ueta, Y. and Yamashita, H. (1997) Inhibition of spontaneous EPSCs and IPSCs by presynaptic GABAB receptors on rat supraoptic magnocellular neurons. J. Physiol., 504: 113–126. Kombian, S., Zidichousky, J. and Pittman, Q. (1996) GABA-B receptors presynaptically modulate excitatory synaptic transmission in the rat supraoptic nucleus in vitro. J. Neurophysiol., 76: 1166–1179. Lincoln, D.W. and Wakerley, J.B. (1974) Electrophysiological evidence for the activation of supraoptic neurons during the release of oxytocin. J. Physiol., 242: 533–554. Moos, F. (1995) GABA-induced facilitation of the periodic bursting activity of oxytocin neurones in suckled rats. J. Physiol., 481: 113–114.

Mouginot, D., Kombian, S.B. and Pittman, Q.J. (1998) Activation of presynaptic GABAB receptors inhibits evoked IPSCs in rat magnocellular neurons in vitro. J. Neurophysiol., 79: 1508–1517. Nissen, R. and Renaud, L. (1994) GABA receptor mediation of median preoptic nucleus inhibition of supraoptic neurosecretory neurones. J. Physiol., 479: 207–216. Poulain, D. and Wakerley, J.B. (1982) Electrophysiology of hypothalamic magnocellular neurons secreting oxytocin and vasopressin. Neuroscience, 7: 773–808. Poulain, D., Wakerley, J.B. and Dyball, R.E.J. (1977) Electrophysiological differentiation of oxytocin- and vasopressinsecreting neurones. Proc. R. Soc. Lond. B., 196: 367–384. Randle, J.C.R. and Renaud, L.P. (1987) Actions of gammaaminobutyric acid on supraoptic nucleus neurosecretory neurones in vitro. J. Physiol., 387: 629–647. Stern, J. and Armstrong, W.E. (1995) Electrophysiological differences between oxytocin and vasopressin neurons recorded from female rats in vitro. J. Physiol., 488(3): 701–708. Stern, J.E. and Armstrong, W.E. (1996) Changes in the electrical properties of supraoptic nucleus oxytocin and vasopressin neurons during lactation. J. Neurosci., 16: 4861–4871. Stern, J.E. and Armstrong, W.E. (1998) Reorganization of the dendritic trees of oxytocin and vasopressin neurons of the rat supraoptic nucleus during lactation. J. Neurosci., 18: 841–853. Stern, J.E., Galarreta, M., Foehring, R.C., Hestrin, S. and Armstrong, W.E. (1999) Differences in the properties of ionotropic glutamate synaptic currents in oxytocin and vasopressin neuroendocrine neurons. J. Neurosci., 19: 3367–3375. Stern, J.E., Hestrin, S. and Armstrong, W.E. (2000) Enhanced neurotransmitter release at glutamatergic synapses on oxytocin neurones during lactation in the rat. J. Physiol., 526: 109–114. Voisin, D.L., Herbison, A.E., Chapman, C. and Poulain, D.A. (1996) Effects of central GABAB receptor modulation upon the milk ejection reflex in the rat. Neuroendocrinology, 63: 368–376.

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 11

Neurohypophyseal hormones in the integration of physiological responses to immune challenges Krisztina J. Kovács * Laboratory of Molecular Neuroendocrinology, Institute of Experimental Medicine, Budapest, Hungary

Keywords: Paraventricular nucleus; Vasopressin; Corticotropin-releasing hormone; Cytokine; Allergy; Anaphylaxis; Brattleboro rat

The neuro-immuno-endocrine web There is a multidirectional communication between the immune, neuroendocrine and central nervous systems. Mediators released from activated immune cells may reach directly or indirectly the central nervous system and activate cell groups that are involved in autonomic, neuroendocrine and behavioral regulation. Lymphoid organs are densely innervated and immune competent cells are equipped with neurotransmitter and hormone receptors. Some hormones and neuropeptides, secreted upon immune activation, display cytokinelike effects, while cytokines may directly stimulate hormone release. Corticosterone, the end-product of the hypothalamo–pituitary–adrenocortical regulatory axis, possesses negative feedback influence not only on the central nervous system targets, but also has profound immunosuppressive effects. Dysfunction of the neuroendocrine–immune communication has been implicated as an etiologically important factor in certain forms of autoimmunity, immunosuppression and hypersensitivity. ∗ Correspondence

to: K.J. Kovács, Laboratory of Molecular Neuroendocrinology, Institute of Experimental Medicine, Szigony u. 43. H-1083. Budapest, Hungary. Tel.: +36-1-210-9952; Fax: +36-1-210-9429; E-mail: [email protected]

The paraventricular nucleus of the hypothalamus as the key integrator of physiological responses to immune stimuli The hypothalamus contains neurosecretory cells that make up the final pathway by which the brain governs several autonomic and endocrine functions. These neurons serve as cellular transducers to integrate stimulatory and inhibitory drives, which may be neural or humoral (hormonal) in nature, and provide relevant output to modulate hypophyseal hormone secretion. Much is known on the location, neurochemical phenotype and secretory capabilities of neurosecretory effector neurons, and of the organization and chemical coding of their major afferents. However, how these come to be linked to form operational circuits is still not well characterized. Functional architecture of the PVN The hypothalamic paraventricular nucleus contains three functionally distinct neuron populations that are clustered in distinct subdivisions of the nucleus (Swanson and Sawchenko, 1980). Magnocellular neurosecretory neurons release vasopressin (AVP) and oxytocin (OXY) at the posterior pituitary to serve homeostatic functions of water balance and reproduction. Parvocellular corticotropin-releasing hormone (CRH) containing neurons project to hy-

128

pophyseal portal vasculature at the external zone of the median eminence to provide the central control of the hypothalamo–pituitary–adrenocortical (HPA) axis and initiate the neuroendocrine stress cascade (Antoni et al., 1983; Swanson et al., 1983). Cells in the dorsal, ventral and lateral parvocellular subnuclei give rise long descending projections that target the periaqueductal gray, parabrachial nucleus, nucleus of the solitary tract, rostral ventrolateral medulla, as well as sympathetic and parasympathetic preganglionic cells in the medulla and spinal cord (Swanson and Sawchenko, 1980; Sawchenko and Swanson, 1982a; Saper, 1995). These neurons are involved in the coordination of autonomic functions to endocrine and behavioral responses. The neuropeptide phenotype of hypothalamic visceromotor neurons is complicated by several factors and depends on the sensitivity of the detection. An updated list of co-expressing peptides that are contained in parvo- and magnocellular neurosecretory neurons is found in a recent review of Burbach et al., 2001. In contrast to the neurosecretory neurons, autonomic projection neurons do not contain a universal identifying neuropeptide marker, subsets of these cells express OXY, AVP or CRH (Sawchenko and Swanson, 1982b). The phenotypic variation of neuropeptide gene expression has functional significance for instance co-expression, co-packing and co-secretion of synergistically acting neuropeptides under high-demand conditions. In one well studied model, withdrawal of the negative feedback signal from the HPA axis by adrenalectomy, results in upregulation of AVP expression in the parvocellular neurosecretory system (Kiss et al., 1984; Sawchenko et al., 1984; Sawchenko, 1987) where its transcription is restrained under basal conditions (Kovacs et al., 2000). Acute and chronic exposure to stressors also result in vasopressinergic phenotype in CRH-secreting parvocellular neurons (Bartanusz et al., 1993, 1994; Makino et al., 1995; Kovacs and Sawchenko, 1996a,b; Ma et al., 1997a,b). Oxytocinergic neurons became CRH immunoreactive in response to chronic salt loading (Young, 1986; Dohanics et al., 1990; Watts, 1992, 2001; Kovacs and Sawchenko, 1993), while lesion of the paraventricular nucleus provoke CRH mRNA expression in the oxytocin-containing neurons of the supraoptic nuclei (Palkovits et al., 1997). The expression of AVP and

OXY in the magnocellular neurosecretory cells is also not mutually exclusive, i.e. during lactation, coexistence of AVP and oxytocin was revealed in the SON (Mezey and Kiss, 1991; Glasgow et al., 1999). Afferent control of the hypothalamic neurosecretory neurons Activity of the hypothalamic paraventricular neurons is governed by neural and/or hormonal influences. Generally, magnocellular neurons receive neural inputs from restricted sources, while the afferent control of parvocellular neurons is much more diverse and includes direct and indirect pathways that convey relevant information from all major sensory modalities of the brain. In addition to sensory inputs, cognitive and emotionally related influences also have a great impact on the stress-related parvocellular neurons, although the anatomical pathways that may mediate these effects are not well characterized (Sawchenko et al., 1996). CRH-secreting parvocellular neurons receive significant monoaminergic innervation: adrenergic axons arise from the ventrolateral medulla (C1) and nucleus of the solitary tract (C2); noradrenergic inputs originate from A2 and A6 regions and from local hypothalamic sources (A12– A14) (Sawchenko and Swanson, 1982b; Liposits et al., 1986a,b; Liposits, 1990). 5-HT-positive fibers that make synaptic contacts with CRH neurons in the parvocellular subdivision (Liposits et al., 1987) appear to arise from three distinct serotoninergic cell groups (B7, B8 and B9) in the midbrain (Sawchenko et al., 1983). Parvocellular neurosecretory neurons receive GABAergic inputs from local interneurons (Roland and Sawchenko, 1993). GABA-containing profiles make symmetric, inhibitory synaptic contacts with CRH neurons and it is estimated that one-third of all synapses that innervate these cells is GABAergic (Miklós and Kovács, submitted). Magnocellular neurons are themselves sensitive to changes of osmotic concentration; however, both AVP and OXY neurons receive inputs from osmosensitive structures of the lamina terminalis that lies outside of the blood–brain barrier and may also confer hormonal and immune signals (Miselis, 1981, 1982; Leng et al., 1982). Descending pathways from the subfornical organ (SFO) and the vascular organ of the lamina terminalis (OVLT) provide glu-

129

tamatergic stimulatory and GABAergic inhibitory projections to the hypothalamic magnocellular neurons. Some of the inputs that originate in lamina terminalis relayed in the median preoptic nucleus, which has strongly been implicated in the inhibitory control of hypothalamic neurosecretory cells (Saper and Levisohn, 1983). It has been recently shown (Leng et al., 2001) that osmotic stimuli coactivate excitatory and inhibitory inputs to the magnocellular neurons and results in increased neuronal firing linear to the osmotic challenge. These findings are in good agreement with those of Theodosis (Theodosis and Poulain, 1993) revealing ultrastructural signs of increased GABAergic inputs under high demand physiological conditions such as lactation. The osmosensitive pathways highlighted above have also been implicated in coordination of CRH expression in parvo- and magnocellular neurons in response to chronic salt loading (Kovacs and Sawchenko, 1993). A drop in blood pressure produces a reflex release of AVP from the posterior pituitary, and it is thought that the pressor effect requires doses of vasopressin considerably higher than those needed for maximal antidiuresis (Koob et al., 1985). Baroreceptors, under basal conditions, inhibit vasopressinergic magnocellular neurons indirectly, via the A6 catecholaminergic cell group. Stimulatory influences are mediated through the nucleus of the solitary tract (NTS) and relayed in the A1 noradrenergic cell group of the caudal ventrolateral medulla that preferentially target AVP neurons (Sawchenko and Swanson, 1982b; Day and Sibbald, 1988, 1990; Day et al., 1992). NTS provides direct innervation to oxytocinergic magnocellular neurons that includes ascending peptidergic pathways and noradrenergic input from the A2 cell group (Sawchenko and Swanson, 1982a,b; Cunningham and Sawchenko, 1988; Plotsky et al., 1989). Neurosecretory cells in the PVN and SON receive histaminergic inputs from discrete cell clusters in the posterior hypothalamus/tuberomamillary region (Panula et al., 1984, 1990). Histamine has been shown to be involved in physiological functions relevant in immune–neuroendocrine crosstalk, including food-intake, thermoregulation, water balance and regulation of the HPA axis (Ookuma et al., 1993; Bealer and Abell, 1995; Kang et al., 1995; Bugajski et al., 1996; Hatton and Yang, 1996; Sakata et al., 1997).

Afferent inputs to the hypothalamic motoneurons together with their potential to express an array of functionally distinct effector molecules and supply them to autonomic, endocrine and behavioral centers constitute the hardware of the regulatory circuit. The aim of the present chapter is to reveal the organization of those operational circuits that are recruited in a situation dependent manner to mediate immune signals to the brain and integrate physiological responses to immune challenges. Immune regulation in brief (from a neuroendocrine perspective) Immune responses are regulated by antigen presenting cells and by Th (helper) lymphocyte subclasses Th1 and Th2, which secrete different sets of cytokines and promote cellular and humoral immunity, respectively (Fearon and Locksley, 1996). Any immune challenge that threatens the homeostasis should be regarded as stressor. Indeed, activation of either form of immunity, activate the HPA axis and results in a complex series of centrally mediated autonomic reaction that serve to cope with the challenge. Th1 and Th2 responses are mutually inhibitory, and hormones play a definitive role in the regulation of Th1/Th2 balance (Elenkov and Chrousos, 1999; Elenkov et al., 2000). For instance, stress has been regarded as immunosuppressive (Besedovsky et al., 1986). Indeed, corticosterone inhibits proinflammatory cytokines and suppress cellular immunity, but activation of the HPA axis boosts humoral immunity (Elenkov and Chrousos, 1999). Recent evidence indicates a CRHmast cell-histamine regulatory loop (Theoharides, 1996; Theoharides et al., 1998), and catecholamines released during sympathoadrenal activation also selectively inhibit cellular immunity and drive towards a Th2-mediated humoral immunity (Elenkov et al., 2000). Immune challenges activate parvocellular neurosecretory neurons Increased production of pro-inflammatory cytokines released by activated macrophages or lymphocytes, that are effector elements of the cellular immunity, represent an essential feature of the early events

130

of immune activation called acute phase response. Interleukin-1β (IL-1β), interleukin-6 (IL-6), tumor necrosis factor (TNFα) play a pivotal role in mediating central and peripheral host responses in that most are mimicked by exogenous administration of these cytokines or by bacterial endotoxin lipopolysaccharide (LPS) that triggers their sequential release (Chensue et al., 1991). Immune challenges commonly activate the HPA axis. Although there are reports in the literature on the direct action of cytokines and LPS on the adrenal gland and pituitary hormone secretion, since the inseminating reports from 1987 (Berkenbosch et al., 1987; Sapolsky et al., 1987) the consensus has been shifted towards acknowledgement the paraventricular nucleus as the major hypothalamic target to govern HPA axis activity in response to immune challenges. In addition to LPS, IL-1, IL-6 and TNF stimulate ACTH secretion and induce immediate-early gene markers of neural activity in the PVN (Elenkov et al., 1992; Grinevich et al., 2001). Lines of evidence supporting the CRH as the major target of IL-1 are the following: first, immunoneutralization with CRH antibody completely prevent ACTH release provoked by acute local inflammation following turpentine injections (Turnbull et al., 1998), and by IL-1 (Sapolsky et al., 1987) or IL-6 (Kageyama et al., 1995); second, IL-1 selectively depletes CRH containing terminals (Berkenbosch et al., 1987), and third, IL-1 injections induce CRH transcription in the parvocellular neurons (Suda et al., 1990; Harbuz et al., 1992; Kakucska et al., 1993; Ericsson et al., 1994; Rivest and Rivier, 1994). It is noteworthy that a single administration of IL-6 does not induce c-fos and CRF expression in the PVN (Vallieres et al., 1997), although lesioning the PVN abolishes the ACTH response to systemic IL-6 (Kovacs and Elenkov, 1995), suggesting involvement of other ACTH releasing moieties of paraventricular origin such as AVP. In line with this hypothesis, IL-6-induced ACTH response was significantly suppressed by central injection of AVP antibody (Kageyama et al., 1995). It is also possible that single exposure to IL-6 is insufficient in activating hypothalamic neurons. Recently, it has been shown, that pretreatment with LPS upregulates IL-6 receptors and amplifies the effect of IL-6 on the CRH expression in the PVN (Vallieres and Rivest, 1999).

Circulating TNFα also triggers ACTH (Kovacs and Elenkov, 1995) and corticosterone secretion (van der Meer et al., 1996), induces c-fos mRNA in the stress-related circuitry including the paraventricular nucleus, increases CRH transcription in the parvocellular neurons (Tolchard et al., 1996; Nadeau and Rivest, 1999) and stimulates CRH release in vitro (Spinedi et al., 1992). Inducible expression of vasopressin gene in the parvocellular neurosecretory neurons of the paraventricular nucleus has a pivotal regulatory role in activation of HPA axis responses to different stressors, including immune challenges (De Goeij et al., 1992a,b; Kovacs and Sawchenko, 1996a,b; Ma et al., 1997a,b; Aubry et al., 1999). Moreover, evidence has recently been accumulated supporting the view that AVP is the primary regulated variable governing HPA function during stress, which maintains the axis activity, particularly under conditions of prolonged or repeated stimulation (De Goeij et al., 1992a,b; Chowdrey et al., 1995; Makino et al., 1995; Ma et al., 1997a,b; Aubry et al., 1999; Aguilera and Rabadan-Diehl, 2000). Vasopressin, released to the hypophyseal portal circulation acts on the VP1β receptors on the pituitary corticotropes (Antoni et al., 1984) and potentiates the effect of CRH and other secretagogues on ACTH secretion (Lutz-Bucher et al., 1980; Giguere and Labrie, 1982; Gillies et al., 1982; Vale et al., 1983). Acute systemic administration of interleukin1 or LPS has been shown to deplete both AVPcontaining and AVP-deficient subtypes of CRH neurosecretory axons (Whitnall et al., 1992; Whitnall, 1993). This effect seems to be different from other systemic stressors such as immobilization, insulininduced hypoglycemia or colchicine treatment that specifically target the VP-containing CRH terminals (Whitnall, 1989). Moreover, a single cytokine injection may result in delayed and long-lasting increase of AVP stores in the external zone of the median eminence (Schmidt et al., 1995, 1996, 2001). Although IL-1 induced ACTH responses have been thought to primarily be driven by CRH, single IL-1 challenge results not only a phenotypic shift to AVP co-storing terminals, but also result in a ‘functional shift’ of the response: rats that have been pre-exposed to IL-1 show marked IL-1-induced depletion of AVP stores upon repeated exposure. These changes might be

131

due to hyperreactivity of noradrenergic nerve terminals in the PVN seen in response to IL-1 injection (Schmidt et al., 1995, 1996, 2001). Activation of magnocellular neurosecretory system by immune stimuli In addition to the parvocellular neurons, immune challenges also affect neurohypophyseal hormones secreted by the magnocellular neurons of the PVN and in the supraoptic nucleus. The situation is complicated by the fact that axons from the magnocellular neurosecretory cells ‘in passage’ through the median eminence are able to release AVP or OXY into the hypophyseal portal blood and contribute to the regulation of hormone secretion by the anterior lobe of pituitary (Holmes et al., 1986; Antoni et al., 1988). The stimulating effect of IL-1 on plasma AVP and OXY concentration seems to depend on glucocorticoid hormones. Intact animals do not display increases in circulating neurohypophyseal hormones, however in adrenalectomized rats a substantial increase was detected following IL-1 injection (Chover-Gonzalez et al., 1994; (Harbuz et al., 1996). Systemic application of IL-1 or LPS result in expression of activation markers c-fos and NGFI-B in the magnocellular neurons. Co-localization studies revealed that immunological insults seem to target oxytocinergic profiles rather than vasopressinergic neurons in the PVN and SON (Ericsson et al., 1994; Rivest and Laflamme, 1995). However, OXY and AVP heteronuclear RNA (hnRNA) levels were unchanged in the magnocellular neurons of the hypothalamus of LPS treated rats (Rivest and Laflamme, 1995). Recently, type I interleukin-1 receptor (IL-1R1) protein and IL-1 receptor antagonist were detected in the magnocellular compartment of the PVN and cells in the SON as well as in accessory magnocellular neurons. IL-1R1 was localized in vasopressin-immunoreactive neurons that provide a means by which IL-1 directly affect VP neurosecretion (Diana et al., 1999). In humans, IL-6 administration activates magnocellular AVP secreting neurons and results in a significant elevation of plasma AVP levels (Mastorakos et al., 1994). In line with these observations, highly elevated plasma AVP (and corticosterone) lev-

els were found in response to restraint stress in transgenic mice that constitutively express IL-6 under the control of GFAP (glial fibrillary acidic protein) promoter targeted to CNS astrocytes (Raber et al., 1997). Elevated plasma AVP levels in these mice are paralelled by hyperplasia of the adrenal cortex and medulla, while their ACTH response is blunted. Magnocellular neurons may also be activated secondarily by viscerosensory stimuli posed by cardiovascular responses seen in endotoxemia or after cytokine treatment (Weinberg et al., 1988; Rogausch et al., 2000; Xia and Krukoff, 2001). A unique feature of vasopressin is that this neuropeptide has a significant modulatory effect on the immune system, therefore it represent an additional means by which neuroendocrine and immune systems communicate. AVP modulates cellular immunity via its enhancement of the autologous mixed lymphocyte response (Bell et al., 1992). The response may be mediated through V1 vasopressin receptors (Bell et al., 1993). AVP and OXY are capable of replacing interleukin 2 (IL-2) requirement for T cell mitogen induction of γ-interferon in mouse spleen cultures (Johnson et al., 1982; Johnson and Torres, 1985); vasopressin potentiates primary antibody responses (Croiset et al., 1990). Increased AVP secretion may participate in susceptibility to autoimmune and inflammatory diseases. Lewis rats display blunted HPA responses to stress and immune challenges, but have augmented systemic AVP secretion that collectively potentiate Th1-mediated cellular immunity and renders this strain to be sensitive to inflammatory (adjuvant-induced arthritis, AA) and autoimmune (experimental allergic encephalomyelitis, EAE) diseases (Patchev et al., 1992; Chowdrey et al., 1995; Harbuz et al., 1997; Chikanza et al., 2000; Huitinga et al., 2000; Sternberg, 2001). A proinflammatory role of AVP in Lewis rats is supported by the fact that immunoneutralization with AVP antibodies attenuates the inflammatory responses (Patchev et al., 1993). AVP have also been implicated in human pathophysiologies, elevated circulating vasopressin levels were measured in patients with rheumatoid arthritis (Chikanza et al., 2000). Manipulation of the neuroendocrine–immune communication at this level would offer strategies to treat chronic inflammatory and autoimmune diseases.

132

Activation of autonomic-related projection neurons in the PVN in response to immune stimuli The third major visceromotor cell cluster in the paraventricular nucleus, the one that give rise to long descending projections to brainstem and spinal cord plays a crucial role in governing autonomic responses to immune challenges. Infectious and inflammatory diseases are often accompanied with changes in thermoregulation (fever), sleepiness, loss of appetite, hypotension etc. Some components of these responses may be mediated by CRH, AVP and OXY-containing neurons in the dorsal, ventral and lateral parvocellular subdivision of the PVN. Rapid, but persisting activation of these neurons was revealed by using c-fos and/or NGFI-B immediateearly gene markers following LPS (Elmquist et al., 1993, 1996; Rivest and Laflamme, 1995), IL-1 (Ericsson et al., 1994; Rivest and Rivier, 1994) and TNF-α (Nadeau and Rivest, 1999). Increased peripheral sympathetic activity that has been reported following IL-1 or LPS injections may directly be related to the activation of the this parvocellular cell group (Berkenbosch et al., 1989). Combination of retrograde tracing from different levels of the intermediolateral cell column of the spinal cord and LPS-induced c-Fos immunocytochemistry revealed the dorsal parvocellular subdivision of the PVN as the hypothalamic center mediating fever-inducing effects of endotoxin (Zhang et al., 2000). It is striking that a relatively small number of PVN neurons (100/side) may orchestrate pattern of sympathetic responses to endotoxemia such as stimulating brown adipose tissue, heart, the adrenal medulla and vascular tone in the tail artery. These responses then collectively serve to elevate body temperature, i.e. induce fever. The neurochemical phenotype of PVN neurons that mediate these effects has not been characterized. However, the distribution of PVNderived, oxytocin-containing terminals in the sympathetic preganglionic cell column of the spinal cord shows overlapping pattern with those that are activated by LPS and are involved in fever generation (Swanson and Sawchenko, 1980).

Afferent pathways that mediate immune signals to hypothalamic effector neurons Much of our recent knowledge on the activated cells and extended pathways that are responsive to a given stimulus came from application of immediate-early gene induction based functional anatomical mapping (Sagar et al., 1988; Morgan and Curran, 1989, 1991; Kovacs, 1998). Identification of these activation markers, in combination with anatomical tracing methods and in situ techniques that reveal the phenotype and transcriptional capacity of the activated profiles is a powerful tool to identify functional units in the rodent brain (Ceccatelli et al., 1989; Chan et al., 1993; Zhang et al., 2000). Interleukin-1 Because of the existence of the blood–brain barrier to circulating macromolecules such as IL-1, the manner in which blood-borne cytokines signal to the hypothalamic effector neurons remained problematic. Transduction through the circumventricular organs that lack blood–brain barrier, active transport through the barrier and local signaling from the brain vasculature have all been proposed as possible mechanism for immune-to-brain signaling (Katsuura et al., 1990; Elmquist et al., 1997). To identify central targets of circulating IL-1, distribution of IL-1 receptors (IL-1 R1) has been compared to those cell groups that display c-fos induction in response to IL-1 injections. Expression of the IL-1R1 in rats is restricted to the major barrier structures including meninges, choroid plexus, ependymal lining of the ventricles and endothelial cells of the brain capillaries (Ericsson et al., 1995). However, IL-1 responsive neurons are found in the paraventricular nucleus and in cell groups recognized as engaged with interoceptive information processing. These structures include the NTS, ventrolateral medulla, lateral parabrachial nucleus, central nucleus of amygdala and bed nucleus of stria terminalis (BNST) (Ericsson et al., 1994; Rivest and Rivier, 1994). Lesioning experiments clearly suggested that activation of CRHsecreting neurons in the PVN is dependent on the ascending catecholaminergic pathways (Ericsson et al., 1994). These findings together with localization of IL-1 receptors in the medulla are compatible with

133

the hypothesis that paracrine effects of prostaglandin PGE2, released from the perivascular cells in response to IL-1 challenge, acting through prostanoid receptor-expressing local catecholaminergic neurons that project to the PVN, contribute to the HPA axis activation by IL-1 (Ericsson et al., 1997). Involvement of the abdominal vagus nerve in mediation of cytokine signals to the central stressrelated circuitry was also proposed (Gaykema et al., 1995). Indeed, immune-related cells, that express IL1 in response to LPS challenge were found in the close proximity to the abdominal vagus (Goehler et al., 1999), and transection of the abdominal vagus has been shown to be capable of eliminating LPS-induced neural activation in the brainstem and hypothalamus (Wan et al., 1993), but exclusively after intraperitoneal, but not after intravenous administration of the endotoxin. Signaling through the circumventricular organs is supported by several reports (Katsuura et al., 1990; Xia and Krukoff, 2001). However, the concentration of IL-1 required to induce activation markers in the vascular organ of the lamina terminalis is about a magnitude higher than those for c-fos induction in the medullary catecholaminergic neurons. In line with these findings, transection of descending inputs from the circumventricular organs to the hypothalamic stress-related neurons did not prevent transcriptional activation of CRH-secreting neurons by systemic IL-1 (Ericsson et al., 1994). Tumor necrosis factor There are two cell surface TNF receptors in the rat brain. P55-TNF receptor mRNA is constitutively expressed in the circumventricular organs, choroid plexus, leptomeninges ventricular ependyma and along the brain microvasculature. However, the other form, p75 is undetectable under basal conditions. Both TNF receptors are robustly upregulated in barrier-associated structures in the brain, including capillary endothelium, in response to immune challenges. (Nadeau and Rivest, 1999). In contrast to the localization of TNF receptors to the barrier structures, cells that respond to a systemic TNF challenge are revealed throughout the brain notably along the stress-related circuitry that stereotypically induced following other challenges (Nadeau and

Rivest, 1999). These results suggest that circulating TNF activates directly the capillary endothelial cells which in turn may produce soluble molecules such as prostaglandins that act on a paracrine manner stimulating autonomic and neuroendocrine centers. Together, immune challenges in general, and proinflammatory cytokines in particular, activate the hypothalamic stress-related neurons through pathways that show similarities to those that are commonly recruited by other stressors. However, the primary sensory information originates from widespread sources including cytokine-responsive cells at the barrier structures and transduced by secondary signaling molecules. Transcriptional changes in the parvocellular neurosecretory cells in response to immune stimuli Under basal conditions, ongoing transcription from both CRH and AVP genes is undetectable in the parvocellular neurons, only a few neurons show nuclear signal when hybridized with CRH intronic probes (Herman et al., 1991, 1992; Kovacs and Sawchenko, 1996a,b). We have previously shown that ether stress-induced activation of CRH and AVP genes in the parvocellular compartment follows different time courses with peak hnRNA responses occurring at 5 min and 2 h, respectively (Kovacs and Sawchenko, 1996a,b). Based on the timing of certain coexpressing transcription factors in the parvocellular neurons, it seems likely that transcriptional activation of AVP and CRH in the same neurons involve distinct mechanisms. Rapid inducibility of CRH gene by stress is compatible with rapid phosphorylation of CREB; while delayed activation of AVP expression in the CRH cells might involve de novo synthesized transcription factors, such as AP1. It is also important to emphasize that the AVP, and not the CRF, gene is the principal target of glucocorticoid-mediated transcriptional suppression during ether stress (Kovacs et al., 2000). Upregulation of CRH transcription in the parvocellular neurosecretory neurons occurs in response to LPS, IL-1 and TNF (Kakucska et al., 1993; Ericsson et al., 1994; Rivest and Rivier, 1994; Nadeau and Rivest, 1999). Using intron-specific probes for in situ hybridization, transient induction of AVP hn-

134

RNA signal has been revealed in the parvocellular neurons following LPS treatment, while other reports did not detect significant changes in AVP mRNA levels (Juaneda et al., 1999). Particular effects of proinflammatory cytokine injections on parvocellular AVP gene expression remain to be analyzed. Notably, LPS-induced co-expression of neurotensin and CCK mRNA has been recently reported in CRH neurons, that may also contribute to the functional plasticity of the parvocellular neurosecretory system responses to inflammation (Juaneda et al., 1999, 2001). The stimulatory effect of immune challenges on CRH transcription seems to be dependent on the ascending catecholaminergic pathways originating in the nucleus of the solitary tract and ventrolateral medulla (Ericsson et al., 1994; Sawchenko et al., 2000). Although this pathway is commonly activated by physical stressors and immune challenges it is not clear if stimulus–transcription coupling recruits similar transcription factors. Inducible expression of immediate early genes, such as c-fos, and NGFIB has been detected in the parvocellular neurons following LPS, IL-1 etc. (Chan et al., 1993; Chang et al., 1993; Ericsson et al., 1994; Rivest and Rivier, 1994). It remains to be determined whether these particular transcription factors are directly involved in regulation of CRH gene expression responses during immune challenges. Signal transduction pathways and transcription factors, specific for immune challenges, such as NFκB, STAT etc. and their involvement in induction of neuropeptide gene expression in the hypothalamus have not been extensively studied (Kovalovsky et al., 2000). NFκB is a nuclear factor bound to enhancer region of the gene that encodes κ light chain of antibody molecules in B cells. NFκB is ubiquitously expressed in many cells and forms a transcriptionally inactive complex with an inhibitory protein referred to as IκB. Upon phosphorylation in response to wide range of extracellular signals, IκB dissociates from the complex and NFκB translocates to the nucleus and regulates expression of many genes, most of which play essential role in immunity and inflammation (O’Neill and Kaltschmidt, 1997). Expression of NFκB and IκB has been detected in the CNS. NFκB in astrocytes, microglia and endothelial cells of the brain capillaries may signal

inflammation, injury and viral infections, whereas neuronal NFκB has been implicated in synaptic plasticity, neuronal development and neurodegenerative diseases (van der Burg and van der Saag, 1996; O’Neill and Kaltschmidt, 1997). Systemic LPS, IL-1 and TNF has been described to induce NFκB activity at the barrier-associated structures in the brain, that is consistent with the hypothesis that immune-tobrain signaling is not mediated directly by immune mediators, such as cytokines, it rather involve reception of immune signals at the barrier structures and induction of secondary messenger molecules and activation of classic neural networks to activate effector neurons (Rivest et al., 2000). It is noteworthy that NFκB activity in the brain and in the periphery is restrained by glucocorticoids, pointing to another level of immune–endocrine interactions (van der Burg and van der Saag, 1996). Allergic reactions: another example of immune–neuroendocrine integration In contrast to infectious and inflammatory responses outlined above, hypersensitivity is predominantly mediated and governed by T helper 2 (Th2) cytokines. Allergic reactions occur immediately following the contact with innocuous allergens and clinically manifested as local reactions as hay fever, eczema, asthma, urticaria and food allergy, while anaphylaxis is a systemic allergic reaction (Ewan, 1998). Initial contact with an allergen results in IgE production by B cells, which then bound to the IgE receptors (FcεRI) on the surface of mast cells and basophils (sensitization) and IgE cross-linking results in the release of vasoactive, chemotactic mediators, such as histamine and serotonin, proteoglycans and enzymes (Wasserman, 1990). Mast cell mediators cause capillary leakage, edema, and smooth muscle contraction which collectively may lead to severe symptoms of anaphylactic shock, including hypotension, asphyxia, and respiratory arrest (Ewan, 1998). In rats, a specific syndrome can be provoked by injecting foreign proteins, which shows all symptoms of general allergic reactions and referred to as anaphylactoid reaction as it occurs after the primary contact with the allergen (Fig. 1). Experimentally, anaphylactoid reactions can be induced by i.p. or i.v. injection of egg white, ovalbumin, compound 48/80 etc.

135

Fig. 1. Anaphylactoid reaction in rats. Rats were injected intravenously with egg white (EW) or saline (CONTR) 1 h before the photographs were taken. First signs of the edema are detected around 30 min and persist up to 6–8 h post-injection. In addition to the edema in paws and legs, anaphylactoid reaction results in scratching and rubbing around the nose, head and genitalia, labored respiration, drop in blood pressure, polydipsia, and reduced exploration.

Anaphylactoid reaction results in activation of the HPA axis Egg white and 48/80 challenge results in elevation of ACTH and corticosterone plasma levels (Foldes et al., 2000); the kinetics of the response is similar to those seen in case of other systemic stress models. To define the circuitry underlying these effects, c-Fos was used as a marker to identify central neurons that are responsive to allergic challenges. Fos protein was detected throughout the PVN, including the CRH-expressing parvocellular neurons, the OXY and AVP-containing magnocellular elements as well as the autonomic projection neurons with a peak of 90–120 min after challenge (Figs. 2 and 3). Extrahy-

pothalamic sites of neural activation show striking similarities with the gross pattern of immediate-early gene induction seen in response to systemic LPS and IL-1, and included structures involved in central interoceptive information processing and autonomic regulation (Foldes et al., 2000). In addition, circumventricular structures including the vascular organ of the lamina terminalis, subfornical organ also became c-Fos-positive following egg white injection. Possible sites of immune-to-brain communication during anaphylactoid reactions Immune mediators, such as histamine and serotonin (5-HT) released from activated mast cells and ba-

136

Fig. 2. Neuronal activation in the hypothalamic paraventricular (PVN) and supraoptic nuclei (SON) in response to immune challenges. Bright field photomicrographs showing c-Fos immunoreactive cell nuclei 2 h following intravenous injection of egg white (EW) or bacterial lipopolysaccharide, LPS. These challenges can be regarded as experimental models of Th-2 mediated hypersensitivity and Th-1mediated inflammatory reactions, respectively. Abbreviations: 3V, third ventricle; OT, optic tract; mpd, dorsal medial parvocellular subdivision; mpv, ventral aspect of medial parvocellular subdivision; pm, posterior magnocellular. Scale bar: 100 μm. From Foldes et al., 2000.

sophil leukocytes, are very likely to play an essential role in immune–brain communication (Foldes et al., 2000; Matsumoto et al., 2001), since pretreatment with chromolyn (a mast cell stabilizer) attenuated the symptoms and hormonal responses seen during anaphylactoid reactions. Because neither of these mediators can passively cross the blood– brain barrier, the manner in which they reach the brain parenchyma and affect hypothalamic neurosecretory neurons remains to be established. Potential sites through which blood-borne signals transduced

into central responses include: (1) circumventricular organs, that lack blood–brain barrier (Katsuura et al., 1990); (2) stimulation of peripheral sensory nerves that may be associated with immune cells, including mast cells (Johnson and Krenger, 1992; Gaykema et al., 1995; Goehler et al., 1999); (3) interaction with the brain microvasculature, that involves release of secondary signaling molecules (Elmquist et al., 1997); and (4) involvement of central mast cells (Theoharides et al., 1995; Matsumoto et al., 2001). Whatever the signaling mechanism is, it is

137

Fig. 3. c-Fos induction in the different visceromotor parts of the paraventricular nucleus in response to immune challenge. Mean ± SEM values of c-Fos immunoreactive cell nuclei following intravenous injections of saline (control), bacterial lipopolysaccharide (LPS) and compound 48/80.

clear that anaphylactoid reactions activate all three visceromotor cell types in the PVN. Transcriptional activation of CRH and AVP genes in the neurosecretory neurons in response to allergic insults To determine if the activation of hypothalamic neurosecretory neurons is reflected in the stimulation of target gene expression, in situ hybridization techniques with intron-specific probes were used to follow the timing of CRH and vasopressin gene expression in the hypothalamus. We found a reliable upregulation of CRH transcription in the parvocellular compartment very rapidly (15 min) following anaphylactoid reaction (Figs. 4 and 5). This timing of CRH gene activation is similar to that found in response to ether stress (Kovacs and Sawchenko, 1996a,b) or to blockade of corticosteroid synthesis (Herman et al., 1992), but occurs earlier than following restraint (Ma et al., 1997a,b) or systemic LPS treatment (Rivest and Laflamme, 1995). Rapid stimulation of CRH transcription is more likely compatible with activation of preformed transcription factors (such as CREB phosphorylation) than induction of de novo synthesized transcription factors such as AP-1 (Guardiola-Diaz et al., 1994; Kovacs and

Sawchenko, 1996a,b). In agreement with previous reports, AVP hnRNA was detected in the acknowledged hypothalamic seats of vasopressin synthesis, including the supraoptic and suprachiasmatic nuclei and the magnocellular subdivision of PVN (Herman, 1995; Kovacs and Sawchenko, 1996a,b). This basal level of transcription was clearly upregulated in the magnocellular neurosecretory neurons in rats that were exposed to anaphylactoid challenge (Fig. 4). In addition, allergic insults provoke an increase of AVP expression in the parvocellular subdivision of the PVN that peaks parallel with CRH transcription (Figs. 4 and 5) It is worthy of mention, however, that this pattern of timing of parvocellular AVP induction is different from that have been detected following ether stress. Previous studies described distinct timing for transcriptional activation with peak of CRH and AVP hnRNA responses to ether at 5 and 120 min after stress, respectively. We have interpreted the delayed AVP expression as a result of stress-induced rise in circulating corticosterone levels that may restrain AVP, but not CRH transcription. Indeed, adrenalectomized rats with or without low level of corticosterone supplementation display advanced peaks of AVP hnRNA responses to ether stress (Kovacs et al., 2000). The time course of AVP hnRNA seen in response to anaphylactoid challenge shows

138

Fig. 4. Time course of CRH and AVP hnRNA responses to anaphylactoid challenge. Darkfield autoradiograms from similar rostrocaudal levels of the PVN showing nuclear hybridization signals obtained using an intron-specific cRNA probes at key time points. From low resting levels, rats show a marked increase of CRF hnRNA signal in the dorsal aspect of medial parvocellular subdivision that peaks at 15 min after egg white injection. CRH hybridization signal 2 h after challenge is not different from uninjected controls. Substantial basal levels of AVP hnRNA expression are apparent in the magnocellular subdivision which were significantly further increased up to 2 h post-challenge. Egg white induced a rapid and transient increase of AVP expression in the parvocellular subdivision that was maximal 15 min after injection.

139

Fig. 5. Effects of anaphylactoid reactions. (A) Hypotension, induced by intravenous injection of egg white. (B) Time course of transcriptional induction of CRF and AVP genes in the medial parvocellular subdivision of the paraventricular nucleus. Mean ± SEM number of hnRNA expressing cells. Note the rapid and transient activation of both genes.

similarities to that described following hypotensive hemorrhage (Chan et al., 2002), and might involve different regulatory mechanisms governing AVP expression in response to ether stress and immune challenge. It is noteworthy that aspects of steroid feedback that may delay AVP transcription are not detectable in the hemorrhage model (Thrivikraman and Plotsky, 1993). Whether this is also the case in response to anaphylactic reactions remains to be established. The regulation of AVP gene expression in the parvocellular versus magnocellular paraven-

tricular subdivisions might be different. Although AVP hnRNA levels are rapidly increase after egg white injection, the upregulation of AVP expression in the magnocellular neuron population seem to be prolonged. Anaphylactic responses are accompanied with transient hypotension that may initially trigger vasopressin expression in the magnocellular neurons. Edema in the paws and legs is another symptom seen during anaphylactoid reactions, which is maintained long after the injection of the provoking agent and might contribute to the increasing demand for pos-

140

terior pituitary hormone secretion. Accordingly, we find blunted edema formation, and more interestingly blunted ACTH and c-fos responses to anaphylactoid challenges in vasopressin-deficient Brattleboro rats. Conclusion All three functional domains of the hypothalamic paraventricular nucleus are involved in the regulation of physiological responses to immune challenges and neurohypophyseal hormones, expressed in all visceromotor cell types play a pivotal role in these effector functions (Fig. 6). Neural pathways, hormonal and transcriptional changes, which are recruited in response to activation either Th1 or Th2 immune responses, show similarities to each other and to those that are stimulated by other stressors. Dyscommunication in the neuro-immuno-endocrine web results in certain neuro- and immunopathologies, while manipulation of this regulatory system should open new therapeutic avenues in treatments of chronic inflammatory, autoimmune and allergic diseases.

Abbreviations ACTH AVP CNS CRH EAE GABA hnRNA HPA IL-1, -6 LPS NFκB NGFI-B OVLT OXY PVN SFO SON STAT Th TNF VP 5-HT

adrenocorticotropin arginine vasopressin central nervous system corticotropin-releasing hormone experimental allergic encephalomyelitis γ-aminobutyric acid heteronuclear RNA hypothalamo–pituitary–adrenocortical interleukin(-1, -6) lipopolysaccharide nuclear factor kappa B nerve growth factor-induced protein B vascular organ of the lamina terminalis oxytocin paraventricular nucleus of the hypothalamus subfornical organ supraoptic nucleus signal transducers activators of transcription T-helper tumor necrosis factor vasopressin serotonin

References

Fig. 6. Schematic summary of immune-challenge-induced physiological responses governed by distinct visceromotor parts of the paraventricular nucleus. Increased expression, synthesis and release of CRH and AVP at parvocellular compartment results in stimulation of hypothalamo–pituitary–adrenocortical axis. Corticosterone, the end-product of this regulatory system, suppress inflammatory responses, but stimulate Th-2 mediated cellular immune responses. Neurohypophyseal hormones released at the posterior pituitary are involved in the regulation of osmo- and cardiovascular responses to immune challenges. Autonomic projection neurons in the PVN orchestrate a pattern of sympathetic responses to immune challenges.

Aguilera, G. and Rabadan-Diehl, C. (2000) Vasopressinergic regulation of the hypothalamic–pituitary–adrenal axis: implications for stress adaptation. Regul. Pept., 96: 23–29. Antoni, F.A., Palkovits, M., Makara, G.B., Linton, E.A., Lowry, P.J. and Kiss, J.Z. (1983) Immunoreactive corticotropinreleasing hormone in the hypothalamoinfundibular tract. Neuroendocrinology, 36: 415–423. Antoni, F.A., Holmes, M.C., Makara, G.B., Karteszi, M. and Laszlo, F.A. (1984) Evidence that the effects of arginine-8vasopressin (AVP) on pituitary corticotropin (ACTH) release are mediated by a novel type of receptor. Peptides, 5: 519– 522. Antoni, F.A., Kovacs, K.J., Dohanits, J., Makara, G.B., Holmes, M.C. and Mazurek, M.F. (1988) Hypophysiotrophic function of vasopressin and oxytocin. Brain Res. Bull., 20: 729–736. Aubry, J.M., Bartanusz, V., Jezova, D., Belin, D. and Kiss, J.Z. (1999) Single stress induces long-lasting elevations in vasopressin mRNA levels in CRF hypophysiotrophic neurones, but repeated stress is required to modify AVP immunoreactivity. J. Neuroendocrinol., 11: 377–384. Bartanusz, V., Aubry, J.M., Jezova, D., Baffi, J. and Kiss, J.Z. (1993) Up-regulation of vasopressin mRNA in paraventricular

141 hypophysiotrophic neurons after acute immobilization stress. Neuroendocrinology, 58: 625–629. Bartanusz, V., Aubry, J.M., Steimer, T., Baffi, J. and Kiss, J.Z. (1994) Stressor-specific increase of vasopressin mRNA in paraventricular hypophysiotrophic neurons. Neurosci. Lett., 170: 35–38. Bealer, S.L. and Abell, S.O. (1995) Paraventricular nucleus histamine increases blood pressure by adrenoreceptor stimulation of vasopressin release. Am. J. Physiol., 269: H80–H85. Bell, J., Adler, M.W. and Greenstein, J.I. (1992) The effect of arginine vasopressin on the autologous mixed lymphocyte reaction. Int. J. Immunopharmacol., 14: 93–103. Bell, J., Adler, M.W., Greenstein, J.I. and Liu-Chen, L.Y. (1993) Identification and characterization of [125I]arginine vasopressin binding sites on human peripheral blood mononuclear cells. Life Sci., 52: 95–105. Berkenbosch, F., van Oers, J., del Rey, A., Tilders, F. and Besedovsky, H. (1987) Corticotropin-releasing factor-producing neurons in the rat activated by interleukin-1. Science, 238: 524–526. Berkenbosch, F., de Goeij, D.E., Rey, A.D. and Besedovsky, H.O. (1989) Neuroendocrine, sympathetic and metabolic responses induced by interleukin-1. Neuroendocrinology, 50: 570–576. Besedovsky, H., del Rey, A., Sorkin, E. and Dinarello, C.A. (1986) Immunoregulatory feedback between interleukin-1 and glucocorticoid hormones. Science, 233: 652–654. Bugajski, J., Gadek-Michalska, A., Olowska, A., Borycz, J. and Glod, R. (1996) Central histaminergic mechanisms mediate the vasopressin-induced pituitary adrenocortical stimulation. J. Physiol. Pharmacol., 47: 649–659. Burbach, J.P.H., Luckman, S.M., Murphy, D. and Gainer, H. (2001) Gene regulation in the magnocellular hypothalamoneurohypophysial system. Physiol. Rev., 81: 1197–1267. Ceccatelli, S., Villar, M.J., Goldstein, M. and Hokfelt, T. (1989) Expression of c-Fos immunoreactivity in transmittercharacterized neurons after stress. Proc. Natl. Acad. Sci. USA, 86: 9569–9573. Chan, R.K., Brown, E.R., Ericsson, A., Kovacs, K.J. and Sawchenko, P.E. (1993) A comparison of two immediate-early genes, c-fos and NGFI-B, as markers for functional activation in stress-related neuroendocrine circuitry. J. Neurosci., 13: 5126–5138. Chan, R.K.W., Rissman, R.A. and Sawchenko, P.E. (2002) Hypotensive challenge effects on hypothalamic neuropeptide gene expression and their afferent mediation. J. Comp. Neurol., in press. Chang, S.L., Ren, T. and Zadina, J.E. (1993) Interleukin-1 activation of FOS proto-oncogene protein in the rat hypothalamus. Brain Res., 617: 123–130. Chover-Gonzalez, A.J., Lightman, S.L. and Harbuz, M.S. (1994) An investigation of the effects of interleukin-1 beta on plasma arginine vasopressin in the rat: role of adrenal steroids. J. Endocrinol., 142(2): 361–366. Chensue, S.W., Terebuh, P.D., Remick, D.G., Scales, W.E. and Kunkel, S.L. (1991) In vivo biologic and immunohistochemical analysis of interleukin-1 alpha, beta and tumor necrosis

factor during experimental endotoxemia. Kinetics, Kupffer cell expression, and glucocorticoid effects. Am. J. Pathol., 138: 395–402. Chikanza, I.C., Petrou, P. and Chrousos, G. (2000) Perturbations of arginine vasopressin secretion during inflammatory stress. Pathophysiologic implications. Ann. N.Y. Acad. Sci., 917: 825– 834. Chowdrey, H.S., Larsen, P.J., Harbuz, M.S., Jessop, D.S., Aguilera, G., Eckland, D.J. and Lightman, S.L. (1995) Evidence for arginine vasopressin as the primary activator of the HPA axis during adjuvant-induced arthritis. Br. J. Pharmacol., 116: 2417–2424. Croiset, G., Heijnen, C.J. and de Wied, D. (1990) Passive avoidance behavior, vasopressin and the immune system. A link between avoidance latency and immune response. Neuroendocrinology, 51: 156–161. Cunningham Jr., 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. Day, T.A. and Sibbald, J.R. (1988) Direct catecholaminergic projection from nucleus tractus solitarii to supraoptic nucleus. Brain Res., 454: 387–392. Day, T.A. and Sibbald, J.R. (1990) Involvement of the A1 cell group in baroreceptor inhibition of neurosecretory vasopressin cells. Neurosci. Lett., 113: 156–162. Day, T.A., Sibbald, J.R. and Smith, D.W. (1992) A1 neurons and excitatory amino acid receptors in rat caudal medulla mediate vagal excitation of supraoptic vasopressin cells. Brain Res., 594: 244–252. De Goeij, D.C., Dijkstra, H. and Tilders, F.J. (1992a) Chronic psychosocial stress enhances vasopressin, but not corticotropin-releasing factor, in the external zone of the median eminence of male rats: relationship to subordinate status. Endocrinology, 131: 847–853. De Goeij, D.C., Binnekade, R. and Tilders, F.J. (1992b) Chronic stress enhances vasopressin but not corticotropin-releasing factor secretion during hypoglycemia. Am. J. Physiol., 263: E394–E399. Diana, A., Van Dam, A.M., Winblad, B. and Schultzberg, M. (1999) Co-localization of interleukin-1 receptor type I and interleukin-1 receptor antagonist with vasopressin in magnocellular neurons of the paraventricular and supraoptic nuclei of the rat hypothalamus. Neuroscience, 89: 137–147. Dohanics, J., Kovacs, K.J. and Makara, G.B. (1990) Oxytocinergic neurons in rat hypothalamus. Dexamethasone-reversible increase in their corticotropin-releasing factor-41-like immunoreactivity in response to osmotic stimulation. Neuroendocrinology, 51: 515–522. Elenkov, I.J. and Chrousos, G.P. (1999) Stress hormones, Th1/Th2 patterns, pro/anti-inflammatory cytokines and susceptibility to disease. Trends Endocrinol. Metab., 10: 359– 368. Elenkov, I.J., Kovacs, K., Kiss, J., Bertok, L. and Vizi, E.S. (1992) Lipopolysaccharide is able to bypass corticotrophinreleasing factor in affecting plasma ACTH and corticosterone

142 levels: evidence from rats with lesions of the paraventricular nucleus. J. Endocrinol., 133: 231–236. Elenkov, I.J., Wilder, R.L., Chrousos, G.P. and Vizi, E.S. (2000) The sympathetic nerve — an integrative interface between two supersystems: the brain and the immune system. Pharmacol. Rev., 52: 595–638. Elmquist, J.K., Ackermann, M.R., Register, K.B., Rimler, R.B., Ross, L.R. and Jacobson, C.D. (1993) Induction of Fos-like immunoreactivity in the rat brain following Pasteurella multocida endotoxin administration. Endocrinology, 133: 3054– 3057. Elmquist, J.K., Scammell, T.E., Jacobson, C.D. and Saper, C.B. (1996) Distribution of Fos-like immunoreactivity in the rat brain following intravenous lipopolysaccharide administration. J. Comp. Neurol., 371: 85–103. Elmquist, J.K., Scammell, T.E. and Saper, C.B. (1997) Mechanisms of CNS response to systemic immune challenge: the febrile response. Trends Neurosci., 20: 565–570. Ericsson, A., Kovacs, K.J. and Sawchenko, P.E. (1994) A functional anatomical analysis of central pathways subserving the effects of interleukin-1 on stress-related neuroendocrine neurons. J. Neurosci., 14: 897–913. Ericsson, A., Liu, C., Hart, R.P. and Sawchenko, P.E. (1995) Type 1 interleukin-1 receptor in the rat brain: distribution, regulation, and relationship to sites of IL-1-induced cellular activation. J. Comp. Neurol., 361: 681–698. Ericsson, A., Arias, C. and Sawchenko, P.E. (1997) Evidence for an intramedullary prostaglandin-dependent mechanism in the activation of stress-related neuroendocrine circuitry by intravenous interleukin-1. J. Neurosci., 17: 7166–7179. Ewan, P.W. (1998) Anaphylaxis. Br. Med. J., 316: 1442–1445. Fearon, D.T. and Locksley, R.M. (1996) The instructive role of innate immunity in the acquired immune response. Science, 272: 50–53. Foldes, A., Nemethy, Z., Szalay, O. and Kovacs, K.J. (2000) Anaphylactoid reactions activate hypothalamo–pituitary– adrenocortical axis: comparison with endotoxic reactions. Brain Res. Bull., 52: 573–579. Gaykema, R.P., Dijkstra, I. and Tilders, F.J. (1995) Subdiaphragmatic vagotomy suppresses endotoxin-induced activation of hypothalamic corticotropin-releasing hormone neurons and ACTH secretion. Endocrinology, 136: 4717–4720. Giguere, V. and Labrie, F. (1982) Vasopressin potentiates cyclic AMP accumulation and ACTH release induced by corticotropin-releasing factor (CRF) in rat anterior pituitary cells in culture. Endocrinology, 111: 1752–1754. Gillies, G.E., Linton, E.A. and Lowry, P.J. (1982) Corticotropin releasing activity of the new CRF is potentiated several times by vasopressin. Nature, 299: 355–357. Glasgow, E., Kusano, K., Chin, H., Mezey, E., Young III, W.S. and Gainer, H. (1999) Single cell reverse transcriptionpolymerase chain reaction analysis of rat supraoptic magnocellular neurons: neuropeptide phenotypes and high voltage-gated calcium channel subtypes. Endocrinology, 140: 5391–5401. Goehler, L.E., Gaykema, R.P., Nguyen, K.T., Lee, J.E., Tilders, F.J., Maier, S.F. and Watkins, L.R. (1999) Interleukin-1beta in immune cells of the abdominal vagus nerve: a link between

the immune and nervous systems?. J. Neurosci., 19: 2799– 2806. Grinevich, V., Ma, X.M., Herman, J.P., Jezova, D., Akmayev, I. and Aguilera, G. (2001) Effect of repeated lipopolysaccharide administration on tissue cytokine expression and hypothalamic–pituitary–adrenal axis activity in rats. J. Neuroendocrinol., 13: 711–723. Guardiola-Diaz, H.M., Boswell, C. and Seasholtz, A.F. (1994) The cAMP-responsive element in the corticotropin-releasing hormone gene mediates transcriptional regulation by depolarization. J. Biol. Chem., 269: 14784–14791. Harbuz, M.S., Stephanou, A., Sarlis, N. and Lightman, S.L. (1992) The effects of recombinant human interleukin (IL)-1 alpha, IL-1 beta or IL-6 on hypothalamo–pituitary–adrenal axis activation. J. Endocrinol., 133: 349–355. Harbuz, M.S., Chover-Gonzalez, A.J., Conde, G.L., Renshaw, D., Lightman, S.L. and Jessop, D.S. (1996) Interleukin-1 betainduced effects on plasma oxytocin and arginine vasopressin: role of adrenal steroids and route of administration. Neuroimmunomodulation, 3: 358–363. Harbuz, M.S., Conde, G.L., Marti, O., Lightman, S.L. and Jessop, D.S. (1997) The hypothalamic–pituitary–adrenal axis in autoimmunity. Ann. N.Y. Acad. Sci., 823: 214–224. Hatton, G.I. and Yang, Q.Z. (1996) Synaptically released histamine increases dye coupling among vasopressinergic neurons of the supraoptic nucleus: mediation by H1 receptors and cyclic nucleotides. J. Neurosci., 16: 123–129. Herman, J.P. (1995) In situ hybridization analysis of vasopressin gene transcription in the paraventricular and supraoptic nuclei of the rat: regulation by stress and glucocorticoids. J. Comp. Neurol., 363: 15–27. Herman, J.P., Schafer, M.K., Watson, S.J. and Sherman, T.G. (1991) In situ hybridization analysis of arginine vasopressin gene transcription using intron-specific probes. Mol. Endocrinol., 5: 1447–1456. Herman, J.P., Schafer, M.K., Thompson, R.C. and Watson, S.J. (1992) Rapid regulation of corticotropin-releasing hormone gene transcription in vivo. Mol. Endocrinol., 6: 1061–1069. Holmes, M.C., Antoni, F.A., Aguilera, G. and Catt, K.J. (1986) Magnocellular axons in passage through the median eminence release vasopressin. Nature, 319: 326–329. Huitinga, I., Schmidt, E.D., van der Cammen, M.J., Binnekade, R. and Tilders, F.J. (2000) Priming with interleukin-1beta suppresses experimental allergic encephalomyelitis in the Lewis rat. J. Neuroendocrinol., 12: 1186–1193. Johnson, D. and Krenger, W. (1992) Interactions of mast cells with the nervous system — recent advances. Neurochem. Res., 17: 939–951. Johnson, H.M. and Torres, B.A. (1985) Regulation of lymphokine production by arginine vasopressin and oxytocin: modulation of lymphocyte function by neurohypophyseal hormones. J. Immunol., 135: 773s–775s. Johnson, H.M., Farrar, W.L. and Torres, B.A. (1982) Vasopressin replacement of interleukin 2 requirement in gamma interferon production: lymphokine activity of a neuroendocrine hormone. J. Immunol., 129: 983–986. Juaneda, C., Lafon-Dubourg, P., Ciofi, P., Sarrieau, A., Corio,

143 M. and Tramu, G. (1999) Immune challenge-stimulated hypophysiotropic corticotropin-releasing hormone messenger RNA expression is associated with an induction of neurotensin messenger RNAs without alteration of vasopressin messenger RNAs. Neuroscience, 93: 393–400. Juaneda, C., Lafon-Dubourg, P., Ciofi, P., Sarrieau, A., Wenger, T., Tramu, G. and Corio, M. (2001) CCK mRNA expression in neuroendocrine CRH neurons is increased in rats subjected to an immune challenge. Brain Res., 901: 277–280. Kageyama, K., Watanobe, H. and Takebe, K. (1995) In vivo evidence that arginine vasopressin is involved in the adrenocorticotropin response induced by interleukin-6 but not by tumor necrosis factor-alpha in the rat. Neuroimmunomodulation, 2: 137–140. Kakucska, I., Qi, Y., Clark, B.D. and Lechan, R.M. (1993) Endotoxin-induced corticotropin-releasing hormone gene expression in the hypothalamic paraventricular nucleus is mediated centrally by interleukin-1. Endocrinology, 133: 815– 821. Kang, M., Yoshimatsu, H., Chiba, S., Kurokawa, M., Ogawa, R., Tamari, Y., Tatsukawa, M. and Sakata, T. (1995) Hypothalamic neuronal histamine modulates physiological responses induced by interleukin-1 beta. Am. J. Physiol., 269: R1308–R1313. Katsuura, G., Arimura, A., Koves, K. and Gottschall, P.E. (1990) Involvement of organum vasculosum of lamina terminalis and preoptic area in interleukin 1 beta-induced ACTH release. Am. J. Physiol., 258: E163–E171. Kiss, J.Z., Mezey, E. and Skirboll, L. (1984) Corticotropinreleasing factor-immunoreactive neurons of the paraventricular nucleus become vasopressin positive after adrenalectomy. Proc. Natl. Acad. Sci. USA, 81: 1854–1858. Koob, G.F., Lebrun, C., Martinez Jr., J.L., Dantzer, R., Le Moal, M. and Bloom, F.E. (1985) Arginine vasopressin, stress, and memory. Ann. N.Y. Acad. Sci., 444: 194–202. Kovacs, K.J. (1998) c-Fos as a transcription factor: a stressful (re)view from a functional map. Neurochem. Int., 33: 287– 297. Kovacs, K.J. and Elenkov, I.J. (1995) Differential dependence of ACTH secretion induced by various cytokines on the integrity of the paraventricular nucleus. J. Neuroendocrinol., 7: 15–23. Kovacs, K.J. and Sawchenko, P.E. (1993) Mediation of osmoregulatory influences on neuroendocrine corticotropin-releasing factor expression by the ventral lamina terminalis. Proc. Natl. Acad. Sci. USA, 90: 7681–7685. Kovacs, K.J. and Sawchenko, P.E. (1996a) Regulation of stressinduced transcriptional changes in the hypothalamic neurosecretory neurons. J. Mol. Neurosci., 7: 125–133. Kovacs, K.J. and Sawchenko, P.E. (1996b) Sequence of stressinduced alterations in indices of synaptic and transcriptional activation in parvocellular neurosecretory neurons. J. Neurosci., 16: 262–273. Kovacs, K.J., Foldes, A. and Sawchenko, P.E. (2000) Glucocorticoid negative feedback selectively targets vasopressin transcription in parvocellular neurosecretory neurons. J. Neurosci., 20: 3843–3852. Kovalovsky, D., Refojo, D., Holsboer, F. and Arzt, E. (2000) Molecular mechanisms and Th1/Th2 pathways in corticos-

teroid regulation of cytokine production. J. Neuroimmunol., 109: 23–29. Leng, G., Mason, W.T. and Dyer, R.G. (1982) The supraoptic nucleus as an osmoreceptor. Neuroendocrinology, 34: 75–82. Leng, G., Brown, C.H., Bull, P.M., Brown, D., Scullion, S., Currie, J., Blackburn-Munro, R.E., Feng, J., Onaka, T., Verbalis, J.G., Russell, J.A. and Ludwig, M. (2001) Responses of magnocellular neurons to osmotic stimulation involves coactivation of excitatory and inhibitory input: an experimental and theoretical analysis. J. Neurosci., 21: 6967–6977. Liposits, Z. (1990) Ultrastructural Immunocytochemistry of the Hypothalamic Corticotropin Releasing Hormone Synthesizing System. Gustav Fisher Verlag, Stuttgart. Liposits, Z., Phelix, C. and Paull, W.K. (1986a) Adrenergic innervation of corticotropin releasing factor (CRF)-synthesizing neurons in the hypothalamic paraventricular nucleus of the rat. A combined light and electron microscopic immunocytochemical study. Histochemistry, 84: 201–205. Liposits, Z., Phelix, C. and Paull, W.K. (1986b) Electron microscopic analysis of tyrosine hydroxylase, dopamine-betahydroxylase and phenylethanolamine-N-methyltransferase immunoreactive innervation of the hypothalamic paraventricular nucleus in the rat. Histochemistry, 84: 105–120. Liposits, Z., Phelix, C. and Paull, W.K. (1987) Synaptic interaction of serotonergic axons and corticotropin releasing factor (CRF) synthesizing neurons in the hypothalamic paraventricular nucleus of the rat. A light and electron microscopic immunocytochemical study. Histochemistry, 86: 541–549. Lutz-Bucher, B., Koch, B., Mialhe, C. and Briaud, B. (1980) Involvement of vasopressin in corticotropin-releasing effect of hypothalamic median eminence extract. Neuroendocrinology, 30: 178–182. Ma, X.M., Levy, A. and Lightman, S.L. (1997a) Emergence of an isolated arginine vasopressin (AVP) response to stress after repeated restraint: a study of both AVP and corticotropinreleasing hormone messenger ribonucleic acid (RNA) and heteronuclear RNA. Endocrinology, 138: 4351–4357. Ma, X.M., Levy, A. and Lightman, S.L. (1997b) Rapid changes in heteronuclear RNA for corticotrophin-releasing hormone and arginine vasopressin in response to acute stress. J. Endocrinol., 152: 81–89. Makino, S., Smith, M.A. and Gold, P.W. (1995) Increased expression of corticotropin-releasing hormone and vasopressin messenger ribonucleic acid (mRNA) in the hypothalamic paraventricular nucleus during repeated stress: association with reduction in glucocorticoid receptor mRNA levels. Endocrinology, 136: 3299–3309. Mastorakos, G., Weber, J.S., Magiakou, M.A., Gunn, H. and Chrousos, G.P. (1994) Hypothalamic–pituitary–adrenal axis activation and stimulation of systemic vasopressin secretion by recombinant interleukin-6 in humans: potential implications for the syndrome of inappropriate vasopressin secretion. J. Clin. Endocrinol. Metab., 79: 934–939. Matsumoto, I., Inoue, Y., Shimada, T. and Aikawa, T. (2001) Brain mast cells act as an immune gate to the hypothalamic– pituitary–adrenal axis in dogs. J. Exp. Med., 194: 71–78. Mezey, E. and Kiss, J.Z. (1991) Coexpression of vasopressin and

144 oxytocin in hypothalamic supraoptic neurons of lactating rats. Endocrinology, 129: 1814–1820. Miselis, R.R. (1981) The efferent projections of the subfornical organ of the rat: a circumventricular organ within a neural network subserving water balance. Brain Res., 230: 1–23. Miselis, R.R. (1982) The subfornical organ’s neural connections and their role in water balance. Peptides, 3: 501–502. Morgan, J.I. and Curran, T. (1989) Stimulus–transcription coupling in neurons: role of cellular immediate-early genes. Trends Neurosci., 12: 459–462. Morgan, J.I. and Curran, T. (1991) Stimulus–transcription coupling in the nervous system: involvement of the inducible proto-oncogenes fos and jun. Annu. Rev. Neurosci., 14: 421– 451. Nadeau, S. and Rivest, S. (1999) Effects of circulating tumor necrosis factor on the neuronal activity and expression of the genes encoding the tumor necrosis factor receptors (p55 and p75) in the rat brain: a view from the blood–brain barrier. Neuroscience, 93: 1449–1464. O’Neill, L.A. and Kaltschmidt, C. (1997) NF-kappa B: a crucial transcription factor for glial and neuronal cell function. Trends Neurosci., 20: 252–258. Ookuma, K., Sakata, T., Fukagawa, K., Yoshimatsu, H., Kurokawa, M., Machidori, H. and Fujimoto, K. (1993) Neuronal histamine in the hypothalamus suppresses food intake in rats. Brain Res., 628: 235–242. Palkovits, M., Kovacs, K., Young, W.S. and Makara, G.B. (1997) Corticotropin-releasing hormone expression in supraoptic neurons after bilateral lesioning of the paraventricular nucleus in rats. Neurobiology, 5: 423–430. Panula, P., Yang, H.Y. and Costa, E. (1984) Histamine-containing neurons in the rat hypothalamus. Proc. Natl. Acad. Sci. USA, 81: 2572–2676. Panula, P., Airaksinen, M.S., Pirvola, U. and Kotilainen, E. (1990) A histamine-containing neuronal system in human brain. Neuroscience, 34: 127–132. Patchev, V.K., Kalogeras, K.T., Zelazowski, P., Wilder, R.L. and Chrousos, G.P. (1992) Increased plasma concentrations, hypothalamic content, and in vitro release of arginine vasopressin in inflammatory disease-prone, hypothalamic corticotropinreleasing hormone-deficient Lewis rats. Endocrinology, 131: 1453–1457. Patchev, V.K., Mastorakos, G., Brady, L.S., Redwine, J., Wilder, R.L. and Chrousos, G.P. (1993) Increased arginine vasopressin secretion may participate in the enhanced susceptibility of Lewis rats to inflammatory disease. Neuroendocrinology, 58: 106–110. Plotsky, P.M., Cunningham Jr., E.T. and Widmaier, E.P. (1989) Catecholaminergic modulation of corticotropin-releasing factor and adrenocorticotropin secretion. Endocr. Rev., 10: 437– 458. Raber, J., O’Shea, R.D., Bloom, F.E. and Campbell, I.L. (1997) Modulation of hypothalamic–pituitary–adrenal function by transgenic expression of interleukin-6 in the CNS of mice. J. Neurosci., 17: 9473–9480. Rivest, S. and Laflamme, N. (1995) Neuronal activity and

neuropeptide gene transcription in the brains of immunechallenged rats. J. Neuroendocrinol., 7: 501–525. Rivest, S. and Rivier, C. (1994) Stress and interleukin-1 betainduced activation of c-fos, NGFI-B and CRF gene expression in the hypothalamic PVN: comparison between Sprague– Dawley, Fisher-344 and Lewis rats. J. Neuroendocrinol., 6: 101–117. Rivest, S., Lacroix, S., Vallieres, L., Nadeau, S., Zhang, J. and Laflamme, N. (2000) How the blood talks to the brain parenchyma and the paraventricular nucleus of the hypothalamus during systemic inflammatory and infectious stimuli. Proc. Soc. Exp. Biol. Med., 223: 22–38. Rogausch, H., Vo, N.T., Del Rey, A. and Besedovsky, H.O. (2000) Increased sensitivity of the baroreceptor reflex after bacterial endotoxin. Ann. N.Y. Acad. Sci., 917: 165–168. Roland, B.L. and Sawchenko, P.E. (1993) Local origins of some GABAergic projections to the paraventricular and supraoptic nuclei of the hypothalamus in the rat. J. Comp. Neurol., 332: 123–143. Sagar, S.M., Sharp, F.R. and Curran, T. (1988) Expression of c-fos protein in brain: metabolic mapping at the cellular level. Science, 240: 1328–1331. Sakata, T., Yoshimatsu, H. and Kurokawa, M. (1997) Hypothalamic neuronal histamine: implications of its homeostatic control of energy metabolism. Nutrition, 13: 403–411. Saper, C.B. (1995) Central autonomic system. In: The Rat Nervous System. Academic Press, New York. Saper, C.B. and Levisohn, D. (1983) Afferent connections of the median preoptic nucleus in the rat: anatomical evidence for a cardiovascular integrative mechanism in the anteroventral third ventricular (AV3V) region. Brain Res., 288: 21–31. Sapolsky, R., Rivier, C., Yamamoto, G., Plotsky, P. and Vale, W. (1987) Interleukin-1 stimulates the secretion of hypothalamic corticotropin-releasing factor. Science, 238: 522–524. Sawchenko, P.E. (1987) Adrenalectomy-induced enhancement of CRF and vasopressin immunoreactivity in parvocellular neurosecretory neurons: anatomic, peptide, and steroid specificity. J. Neurosci., 7: 1093–1106. Sawchenko, P.E. and Swanson, L.W. (1982a) Immunohistochemical identification of neurons in the paraventricular nucleus of the hypothalamus that project to the medulla or to the spinal cord in the rat. J. Comp. Neurol., 205: 260–272. Sawchenko, P.E. and Swanson, L.W. (1982b) The organization of noradrenergic pathways from the brainstem to the paraventricular and supraoptic nuclei in the rat. Brain Res., 257: 275– 325. Sawchenko, P.E., Swanson, L.W., Steinbusch, H.W. and Verhofstad, A.A. (1983) The distribution and cells of origin of serotonergic inputs to the paraventricular and supraoptic nuclei of the rat. Brain Res., 277: 355–360. Sawchenko, P.E., Swanson, L.W. and Vale, W.W. (1984) Coexpression of corticotropin-releasing factor and vasopressin immunoreactivity in parvocellular neurosecretory neurons of the adrenalectomized rat. Proc. Natl. Acad. Sci. USA, 81: 1883–1887. Sawchenko, P.E., Brown, E.R., Chan, R.K., Ericsson, A., Li, H.Y., Roland, B.L. and Kovacs, K.J. (1996) The paraven-

145 tricular nucleus of the hypothalamus and the functional neuroanatomy of visceromotor responses to stress. Prog. Brain Res., 107: 201–222. Sawchenko, P.E., Li, H.Y. and Ericsson, A. (2000) Circuits and mechanisms governing hypothalamic responses to stress: a tale of two paradigms. Prog. Brain Res., 122: 61–78. Schmidt, E.D., Janszen, A.W., Wouterlood, F.G. and Tilders, F.J. (1995) Interleukin-1-induced long-lasting changes in hypothalamic corticotropin-releasing hormone (CRH) — neurons and hyperresponsiveness of the hypothalamus–pituitary– adrenal axis. J. Neurosci., 15: 7417–7426. Schmidt, E.D., Binnekade, R., Janszen, A.W. and Tilders, F.J. (1996) Short stressor induced long-lasting increases of vasopressin stores in hypothalamic corticotropin-releasing hormone (CRH) neurons in adult rats. J. Neuroendocrinol., 8: 703–712. Schmidt, E.D., Schoffelmeer, A.N., De Vries, T.J., Wardeh, G., Dogterom, G., Bol, J.G., Binnekade, R. and Tilders, F.J. (2001) A single administration of interleukin-1 or amphetamine induces long-lasting increases in evoked noradrenaline release in the hypothalamus and sensitization of ACTH and corticosterone responses in rats. Eur. J. Neurosci., 13: 1923–1930. Spinedi, E., Hadid, R., Daneva, T. and Gaillard, R.C. (1992) Cytokines stimulate the CRH but not the vasopressin neuronal system: evidence for a median eminence site of interleukin-6 action. Neuroendocrinology, 56: 46–53. Sternberg, E.M. (2001) Neuroendocrine regulation of autoimmune/inflammatory disease. J. Endocrinol., 169: 429– 435. Suda, T., Tozawa, F., Ushiyama, T., Sumitomo, T., Yamada, M. and Demura, H. (1990) Interleukin-1 stimulates corticotropinreleasing factor gene expression in rat hypothalamus. Endocrinology, 126: 1223–1228. Swanson, L.W. and Sawchenko, P.E. (1980) Paraventricular nucleus: a site for the integration of neuroendocrine and autonomic mechanisms. Neuroendocrinology, 31: 410–417. 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: 165–186. Theodosis, D.T. and Poulain, D.A. (1993) Activity-dependent neuronal–glial and synaptic plasticity in the adult mammalian hypothalamus. Neuroscience, 57: 501–535. Theoharides, T.C. (1996) The mast cell: a neuroimmunoendocrine master player. Int. J. Tissue React., 18: 1–21. Theoharides, T.C., Spanos, C., Pang, X., Alferes, L., Ligris, K., Letourneau, R., Rozniecki, J.J., Webster, E. and Chrousos, G.P. (1995) Stress-induced intracranial mast cell degranulation: a corticotropin-releasing hormone-mediated effect. Endocrinology, 136: 5745–5750. Theoharides, T.C., Singh, L.K., Boucher, W., Pang, X., Letourneau, R., Webster, E. and Chrousos, G. (1998) Corticotropin-releasing hormone induces skin mast cell degranulation and increased vascular permeability, a possible explanation for its proinflammatory effects. Endocrinology, 139: 403–413. Thrivikraman, K.V. and Plotsky, P.M. (1993) Absence of gluco-

corticoid negative feedback to moderate hemorrhage in conscious rats. Am. J. Physiol., 264: E497–E503. Tolchard, S., Hare, A.S., Nutt, D.J. and Clarke, G. (1996) TNF alpha mimics the endocrine but not the thermoregulatory responses of bacterial lipopolysaccharide (LPS): correlation with FOS-expression in the brain. Neuropharmacology, 35: 243– 248. Turnbull, A.V., Lee, S. and Rivier, C. (1998) Mechanisms of hypothalamic–pituitary–adrenal axis stimulation by immune signals in the adult rat. Ann. N.Y. Acad. Sci., 840: 434–443. Vale, W., Vaughan, J., Smith, M., Yamamoto, G., Rivier, J. and Rivier, C. (1983) Effects of synthetic ovine corticotropinreleasing factor, glucocorticoids, catecholamines, neurohypophysial peptides, and other substances on cultured corticotropic cells. Endocrinology, 113: 1121–1131. Vallieres, L. and Rivest, S. (1999) Interleukin-6 is a needed proinflammatory cytokine in the prolonged neural activity and transcriptional activation of corticotropin-releasing factor during endotoxemia. Endocrinoloy, 140: 3890–3903. Vallieres, L., Lacroix, S. and Rivest, S. (1997) Influence of interleukin-6 on neural activity and transcription of the gene encoding corticotrophin-releasing factor in the rat brain: an effect depending upon the route of administration. Eur. J. Neurosci., 9: 1461–1472. van der Burg, B. and van der Saag, P.T. (1996) Nuclear factorkappa-B/steroid hormone receptor interactions as a functional basis of anti-inflammatory action of steroids in reproductive organs. Mol. Hum. Reprod., 2: 433–438. van der Meer, M.J., Sweep, C.G., Rijnkels, C.E., Pesman, G.J., Tilders, F.J., Kloppenborg, P.W. and Hermus, A.R. (1996) Acute stimulation of the hypothalamic–pituitary–adrenal axis by IL-1 beta, TNF alpha and IL-6: a dose response study. J. Endocrinol. Invest., 19: 175–182. Wan, W., Janz, L., Vriend, C.Y., Sorensen, C.M., Greenberg, A.H. and Nance, D.M. (1993) Differential induction of cFos immunoreactivity in hypothalamus and brain stem nuclei following central and peripheral administration of endotoxin. Brain Res. Bull., 32: 581–587. Wasserman, S.I. (1990) Mast cell biology. J. Allergy Clin. Immunol., 86: 590–593. Watts, A.G. (1992) Disturbance of fluid homeostasis leads to temporally and anatomically distinct responses in neuropeptide and tyrosine hydroxylase mRNA levels in the paraventricular and supraoptic nuclei of the rat. Neuroscience, 46: 859–879. Watts, A.G. (2001) Neuropeptides and the integration of motor responses to dehydration. Annu. Rev. Neurosci., 24: 357–384. Weinberg, J.R., Wright, D.J. and Guz, A. (1988) Interleukin-1 and tumour necrosis factor cause hypotension in the conscious rabbit. Clin. Sci. (Colch), 75: 251–255. Whitnall, M.H. (1989) Stress selectively activates the vasopressin-containing subset of corticotropin-releasing hormone neurons. Neuroendocrinology, 50: 702–707. Whitnall, M.H. (1993) Regulation of the hypothalamic corticotropin-releasing hormone neurosecretory system. Prog. Neurobiol., 40: 573–629. Whitnall, M.H., Perlstein, R.S., Mougey, E.H. and Neta, R. (1992) Effects of interleukin-1 on the stress-responsive and

146 -nonresponsive subtypes of corticotropin-releasing hormone neurosecretory axons. Endocrinology, 131: 37–44. Xia, Y. and Krukoff, T.L. (2001) Cardiovascular responses to subseptic doses of endotoxin contribute to differential neuronal activation in rat brain. Brain Res. Mol. Brain Res., 89: 71–85. Young III, W.S. (1986) Corticotropin-releasing factor mRNA in

the hypothalamus is affected differently by drinking saline and by dehydration. FEBS Lett., 208: 158–162. Zhang, Y.H., Lu, J., Elmquist, J.K. and Saper, C.B. (2000) Lipopolysaccharide activates specific populations of hypothalamic and brainstem neurons that project to the spinal cord. J. Neurosci., 20: 6578–6586.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 12

Involvement of the brain oxytocin system in stress coping: interactions with the hypothalamo–pituitary–adrenal axis Inga D. Neumann * Institute of Zoology, University of Regensburg, 93040 Regensburg, Germany

Abstract: In response to various ethologically relevant stressors, oxytocin is released not only from neurohypophysial terminals into the blood, but also within distinct brain regions, for example the hypothalamic supraoptic and paraventricular nuclei, the septum and the amygdala in dependence on the quality and intensity of the stressor. Thus, oxytocin secretory activity may accompany the response of the hypothalamo–pituitary–adrenal (HPA) axis to a given stressor. In the present chapter, I try to summarize our efforts to reveal the physiological significance of intracerebrally released oxytocin in rats with respect to the regulation of the HPA axis under basal and stress conditions as well as with respect to behavioural stress responses. The effects of oxytocin appear to depend on the brain region studied and the state of activity of the animal (basal versus stress). In order to reveal interactions between the oxytocin system and the HPA axis, preliminary results are presented pointing towards a differential action of glucocorticoids on intracerebral and peripheral oxytocin release. Keywords: ACTH; Anxiety; Maternal behavior; Oxytocin receptor antagonist; Retrodialysis

Introduction It has long been known that exposure to a variety of stressors triggers not only the activation of ‘classical’ neuroendocrine stress systems with the result of increased plasma levels of, for example, noradrenaline, adrenaline, prolactin, opioids and corticosterone (rodents)/cortisol (humans), but also the secretion of oxytocin (OXT) from neurohypophysial terminals into the blood. Such stressors include swimming and other physical/emotional challenges like immobilization, ether, hyperosmolality and shaker stress (Lang et al., 1983; Higuchi et al., 1988; Kasting, 1988; Landgraf et al., 1988; Lightman, 1992; Neumann et al., 1993b; Hashiguchi et al., 1997).

∗ Correspondence to: I.D. Neumann, Institute of Zoology, University of Regensburg, 93040 Regensburg, Germany. Tel.: +49-941-943-3055; Fax: +49-941-943-3052; E-mail: [email protected]

While the physiological significance of this peripherally released OXT is largely unknown, it may include, at the adrenal level, regulation of cortisol release (Legros and Ansseau, 1992). OXT can also be released within the brain, for example within the hypothalamic supraoptic (SON) and paraventricular (PVN) nuclei in a somatodendritic fashion, or within limbic and brainstem regions from OXT neurones of the PVN which project to these regions (for review see Landgraf, 1995; Ludwig, 1998). Such release has been demonstrated using intracerebral microperfusion techniques like push–pull perfusion or microdialysis in different species, including sheep (Kendrick et al., 1988), and rats (Demotes-Mainard et al., 1986; Landgraf et al., 1988; Moos et al., 1989; Neumann and Landgraf, 1989; Neumann et al., 1993a). In addition to classical, mainly reproductionrelated stimuli like parturition (Kendrick et al., 1988; Landgraf et al., 1991; Neumann et al., 1993a), and suckling in the lactating rat (Moos et al., 1989; Neumann and Landgraf, 1989), intracerebral OXT

148

release has also been monitored in response to various other stimuli. It has to be mentioned that an extensive brain OXT receptor system exists in those regions where local nonapeptide release has been described (Brinton et al., 1984; Freund-Mercier et al., 1987; Insel, 1990; Verbalis, 1999). In the present chapter I will focus on ethologically relevant stressors which trigger OXT release in the brain, and on the regulatory capacity of brain OXT regulating the main neuroendocrine stress system, the hypothalamo–pituitary–adrenal (HPA) axis. Further, if interactions between the OXT and the HPA systems are to be considered, I will give first and preliminary evidence for glucocorticoid effects on the activity of OXT neurones. Finally, the regulation of anxiety-related behaviour, aggressive/defensive and maternal behaviour under stressful conditions by brain OXT will be described. Intracerebral release of OXT in response to various stressors Combined physical and psychological stressors Swimming Swimming, being an ethologically relevant stressor, represents a combined physical and emotional stimulus due to exposure to a complex change of the environment (Abel, 1994). Thus, it combines cold stress, fear of drowning, and physical exercise during struggling and swimming. Exposure to forced swimming allows active (struggling, swimming) versus passive (floating) stress coping strategies to be monitored and, thus, conclusions about the emotional trait of the animal to be drawn. Forced swimming is therefore considered to have a predictive value for the efficacy of antidepressive treatments (Lucki, 1997). Various neuroendocrine systems are activated during swimming (Abel, 1994), including the HPA axis, with the consequence of increased plasma corticotropin (ACTH) and corticosterone concentrations. Further, exposure to cold (19–22°C), 40-cm-deep water potently increases plasma lactate concentration, which indicates increased muscular activity, and stimulates the secretion of OXT into the blood (Lang et al., 1983; Kasting, 1988; Walker et al., 1995; Neumann et al., 1998) independent of gender.

In contrast to OXT, the neurohypophysial secretion of arginine vasopressin remains unchanged (Verbalis et al., 1986; Kasting, 1988). This is the more remarkable as the plasma osmolality is significantly increased (Wotjak et al., 1998) — a classical stimulus for the vasopressinergic system (Dunn et al., 1973; Landgraf et al., 1988) – and in situ hybridization studies indicate increased expression of vasopressin mRNA in magnocellular neurones in response to swimming (Wotjak et al., 2001). It remains to be shown in detail to which extent the amino acid taurine synthesized in and released from hypothalamic glia cells and recently shown to exert an efficient inhibitory effect on vasopressin cell activity (Engelmann et al., 2001; Hussy, 2002, this volume), is involved in the inhibition of vasopressin secretion. In addition to neurohypophysial secretion, OXT is also released within the hypothalamic PVN and SON in response to forced swimming as shown in male (Wotjak et al., 1998) and female (Wigger and Neumann, 2002) rats. In the latter, intrahypothalamic release of OXT was found to be regulated by endogenous opioids with excitatory effects in virgin females and inhibitory effects in late pregnant rats (Wigger and Neumann, 2002). Activation of OXT neurons at 60 min after forced swimming (Torner et al., 2000) is also indicated by immunocytochemical and in situ hybridization studies which show increased expression of the Fos protein in OXT neurones (Thrivikraman, Plotsky, Torner and Neumann, unpublished observations) and of OXT mRNA in magnocellular neurones (Wotjak et al., 2001). Thus, swimming proves to be a robust stressor activating the OXT system including neurohypophysial and intrahypothalamic release. So far, release of OXT outside the hypothalamus has not been monitored in response to swimming. Shaker stress Another stressor, which can be considered relevant in some regions on earth, is exposure to a shaking surface. Thus, shaking is a combined physical and psychological stressor that has been shown to be accompanied by a significant response of the HPA axis in male rats (Hashiguchi et al., 1997). Furthermore, secretion of OXT into the blood as well as within the PVN was found to be increased (Hashiguchi et al., 1997; Nishioka et al., 1998).

149

Psychosocial stressors Social defeat in males The social defeat test represents a relevant psychosocial, emotional stressor for male rats. The experimental rat (intruder) is placed into the home cage of a usually larger resident rat together with a female (the presence of her litter is not obligatory). The resident rat has been trained for high aggression towards an intruder by placing a test male into the cage for a few seconds per day for at least 2 weeks before testing. Due to the high level of aggression of the resident during the experiment, the intruder rat is instantly attacked, which may result in severe injuries. Therefore, after the first attack, the intruder has to be separated behind a wired mesh barrier which allows exchange of visual, olfactory and acoustic signals between the animals, but prevents physical contact. Typically, the resident continues to display threatening postures, whereas the intruder shows signs of defensive behaviour including immobility and freezing in an upright posture as well as ultrasound vocalization. This highly stressful situation for the intruder rat results in increased plasma ACTH and corticosterone concentrations while plasma lactate remains virtually constant (sign of an emotional stressor) (Welch and Klopfer, 1961; Leshner and Politch, 1979; Haller et al., 1995; Wotjak et al., 1996). In the intruder rat, plasma OXT remains unchanged and rather tends to decrease during exposure to social defeat. In contrast, OXT release within the brain is stimulated which, however, depends on the brain region studied. For example, within the SON, there is a 2.5-fold, and within the mediolateral septum an about 2.3-fold, increase in local OXT release, whereas within the PVN, the main origin of septally released OXT, local OXT release remains unchanged as measured by intracerebral microdialysis before, during and after a 10-min social defeat exposure (Wotjak et al., 1996; Engelmann et al., 1999; Ebner et al., 2000). With regard to the social defeat paradigm as an emotional stressor, however, it would be interesting to monitor not only neuroendocrine responses including OXT release within the brain of the male intruder rat, but also of the male resident rat. The finding of OXT and vasopressin release from dendrites, perikarya and/or axons within different

brain regions, but not from neurohypophysial terminals into the blood, in response to social defeat rules out the possibility of neuropeptide diffusion from the blood compartment into distinct brain regions, thus confirming our previous results (Neumann et al., 1993b). Furthermore, it demonstrates the fascinating capacity of peptidergic neurones to differentially regulate release from different sites (dendrites, somata, axon terminals), depending on the quality and intensity of the stressor (Neumann et al., 2001). Maternal defeat in females Recently, we succeeded in establishing the maternal defence test as a relevant emotional stressor for female rats (Neumann et al., 2001). Female aggression directed against a male or female intruder that approaches the nest is exclusively related to the protection and defence of the offspring and is part of the complex behavioural patterns described as maternal behaviour (Rosenblatt et al., 1994). In our experiments, a virgin intruder rat is placed into the home cage of an early lactating resident rat (between days 2 and 8 of lactation) with her litter for 10 min (Fig. 1A). The lactating resident is naive, as it has never been exposed to an intruder before. For comparison between different reproductive states, also a lactating or pregnant rat can be used as an intruder rat. In contrast to the social defeat in males, the advantage of the maternal defence paradigm is that there is no need for separating the intruder and resident rat, as the lactating resident’s attacks never resulted in severe bites or injuries, despite being effective in preventing the intruder’s approach to the offspring. Thus, detailed behavioural recordings of both intruder and resident are possible (Neumann et al., 2001). In contrast to the significant activation of the HPA axis in the virgin (and also lactating) intruder, secretion of ACTH and corticosterone in the lactating resident is increased only slightly indicating attenuation of the response of the HPA axis in the lactating rat also to an obviously relevant stressor (Neumann et al., 2001). In virgin intruders only, the enhanced ACTH and corticosterone secretion is accompanied by an increased secretion of OXT into the blood (Fig. 2A). In contrast, OXT secretion remained unchanged in lactating intruders (Fig. 2A) as well as

150

patterns seem to parallel those of peripheral OXT secretion, in that increased local OXT release was found. In contrast, in lactating residents, intra-PVN OXT release remained virtually unchanged comparable with the non-response of OXT secretion into the blood in these animals. From these results we may conclude that there is a functional coupling of OXT release within the PVN and into the blood in response to this stressor. Preliminary studies also focus on the question of whether release of OXT can be monitored in extrahypothalamic limbic brain regions like the septum and the amygdala in response to exposure to the maternal defence test. To our knowledge, within the amygdala, successful attempts to monitor OXT release have not been reported as yet. Using microdialysis, however, our first results show that OXT release can be detected successfully even in the central amygdala with relatively high basal levels in a lactating rat compared to a virgin rat, which drop during stress exposure (Fig. 2B). Involvement of brain OXT in the regulation of the HPA axis Inhibitory effects of brain OXT on basal HPA axis activity Fig. 1. (A) The maternal defence test established for monitoring neuropeptide involvement in neuroendocrine and behavioural stress responses as well as aggressive versus defensive behaviour in female rats (and mice). R, lactating resident; I, virgin, pregnant or lactating intruder according to the scientific question to be answered, placed into the cage of the resident for 10 min. (B) Effects of an OXT receptor antagonist administered to lactating rats 10 min before exposure to a virgin intruder on offensive and maternal behaviour. Data are the mean ± SEM. * P < 0.05 versus vehicle.

lactating residents indicating unresponsiveness of the OXT system to non-reproduction-related stimuli in lactation as described before (Carter and Lightman, 1987; Lightman and Young, 1989; Walker et al., 1995; Neumann et al., 1998). Furthermore, OXT release has been monitored within distinct brain regions of both the virgin intruder and lactating resident (Fig. 2B; Krömer, Bosch and Neumann, unpublished observations). Thus, for example in the PVN of virgin intruders, OXT release

In the previous paragraphs, we described OXT release within various brain regions in response to different physical and emotional challenges. As a consequence, the question arises as to the physiological significance of such locally released neuropeptides, the more as OXT receptor expression and binding could be localized in those brain regions where neuropeptide release occurs. As release is simultaneous to activation of the HPA axis (secretion of ACTH from adenohypophysial corticotroph cells and of corticosterone from the adrenal cortex), we have focused on the capacity of brain OXT to contribute to the regulation of the activity of the HPA axis. Thus, in a first set of experiments, rats were implanted with a chronic jugular vein catheter and a guide cannula above the lateral cerebral ventricle (i.c.v.) 5 days prior to the experiments and were handled daily to avoid unspecific stress responses during the experiment including the development of a remote and stress-free method of i.c.v. infusion of a given substance (for example, an OXT recep-

151

Fig. 2. (A) Oxytocin concentration in plasma of lactating residents and virgin intruders before (basal) and 5 min after 10-min exposure to the maternal defence test (MD), as well as respective delta values. Data are the means ± SEM. ◦ P < 0.05 versus basal; ** P < 0.01 versus lactating residents. Data from Neumann et al., 2001. (B) Examples of oxytocin content in microdialysates sampled from the amygdala of a lactating resident and a virgin intruder rat under basal conditions, during (MD1) and after (MD2, MD3) exposure to a 10-min period of MD. Note the drop of high basal values in the lactating resident, and the rise of low basal values in the virgin intruder during/after MD exposure.

tor antagonist) with simultaneous blood sampling (Neumann et al., 2000a). Blockade of the receptor-mediated action of brain OXT by administration of a selective OXT receptor antagonist (des Gly–NH2 d(CH2 )5 [Tyr(Me)2 ,Thr4 ] OVT) into the lateral cerebral ventricle significantly affected basal release of ACTH and consequently, corticosterone into the blood. Compared to vehicletreated controls, the antagonist treatment significantly increased ACTH and corticosterone secretion

under resting conditions. This effect was found to be gender-independent, although the rise in hormone secretion was more pronounced in virgin female compared to male rats (Neumann et al., 2000a). In an attempt to localize the site of tonic inhibitory action of OXT within the brain under resting conditions, the OXT receptor antagonist was applied locally into the mediolateral septum, amygdala or PVN of male rats, using reversed microdialysis (Fig. 3; Neumann et al., 2000a). Thus, rats were fitted unilat-

152

Fig. 3. Effects of local administration of the OXT receptor antagonist (OXT-A) into the mediolateral septum, amygdala region and PVN via retrodialysis on basal and stress-induced ACTH secretion of male Wistar rats. Presented are the ratios of ACTH plasma concentrations after and before OXT-A under basal conditions (basal) and after and before stressor-exposure. Data are the mean ± SEM. * P < 0.05 versus vehicle.

erally (septum) or bilaterally (amygdala, PVN) with a U-shaped microdialysis probe and a jugular vein catheter 3 days prior to the experiment. Whereas ACTH and corticosterone secretion was unaffected in rats during and after dialysis of the OXT receptor antagonist into the septum and amygdala, a significant rise of HPA axis activity was found when OXT action was blocked within the PVN. This indicates that OXT, which is also released within the PVN under resting conditions, exerts a tonic inhibition on ACTH secretion possibly via inhibiting CRH neuronal activity. A quite different picture regarding the involvement of brain OXT in the regulation of the basal

activity of the HPA axis emerged in the peripartum period. In rats between days 19 and 21 of pregnancy, for example, administration of the OXT antagonist, at the same dose as that given to male and virgin rats into the lateral ventricle, failed to significantly alter basal ACTH and corticosterone secretion. In mid-lactation, the effects of blockade of brain OXT receptors was found to depend on the presence of the pups. Thus, when pups were separated from their dam at the beginning of the experiment, that is, at the time of connecting blood sampling and i.c.v. infusion devices 2 h before the first blood sample was collected, a significant effect of the OXT receptor antagonist on basal HPA axis activity was

153

undetectable (Neumann et al., 2000b). In contrast, when dams were kept under undisturbed conditions together with their litters after the connection procedure until blood sampling and i.c.v. infusion, treatment with the OXT receptor antagonist resulted in a slight though significant increase in ACTH and corticosterone secretion (Neumann et al., 2001). This suggests that the suckling stimulus provided by the pups results in an increased release of OXT within several brain regions (Moos et al., 1989; Neumann and Landgraf, 1989; Neumann et al., 1993a) which, in turn, exerts a tonic inhibition on basal HPA axis activity. Thus, OXT released, for example, within the PVN during suckling, is not only involved in the regulation of the synchronization of OXT neuronal activity and pulsatile secretion of OXT into the blood necessary for milk ejection (Lambert et al., 1993; Neumann et al., 1994), but also for the simultaneous downregulation of the activity of the brain CRH system, in particular of CRH neurones located in the PVN (Lightman and Young, 1989). An inactivated CRH system, in turn, may be a prerequisite for the sleep-like characteristics of electroencephalogram recordings of the lactating rat during suckling (Ehlers et al., 1986; Wakerley et al., 1989; Opp, 1995). Support for this hypothesis comes from human studies, although under conditions of stress. In a recent study, Heinrichs et al. (2001) demonstrated that breastfeeding 30 min prior to exposure to a psychosocial stressor — the Trier Social Stress Test — results in a blunted HPA response compared to lactating women breastfeeding their baby more than 2 h prior to the test. In addition to intracerebrally released OXT, other suckling-related factors including prolactin (Torner et al., 2001) may be involved in the attenuation of the stress response. Region-dependent effects of brain OXT on stress-induced HPA axis activity Administration of an OXT receptor antagonist into the lateral cerebral ventricle (i.c.v.) In addition to the basal activity of the HPA axis, the stress-induced secretion of ACTH and corticosterone has also been monitored under the influence of the OXT receptor antagonist applied i.c.v. In both male and virgin female rats, the stress-induced

hormonal rise was significantly more pronounced in antagonist-treated rats compared to vehicle treatment. This effect was independent of whether the i.c.v. administration was performed under undisturbed, stress-free conditions 10 min or immediately before exposure to the stressor. In the latter case, the administration procedure was used as part of the stressor as rats were removed from their cage and restrained by hand for insertion of the i.c.v. infusion cannula into the previously implanted guide cannula and i.c.v. infusion. The disinhibition of the stressinduced HPA axis response by blockade of the action of brain OXT receptors was independent of the stressor the rats were exposed to. Thus, a further rise in ACTH and corticosterone secretion was found in OXT antagonist-treated rats exposed to the elevated plus-maze (Neumann et al., 2000a), forced swimming (Neumann et al., 2000a), an elevated platform (Neumann et al., 2000c), or exposure to repeated airpuffs in the home cage (Neumann, Toschi and Douglas, unpublished observations). These results demonstrate a general inhibitory action of brain OXT on neuroendocrine stress responses. They are supported by the finding of Windle et al. (1997), who reported on reduced corticosterone responses to noise stress in ovariectomized virgin female rats chronically treated with i.c.v. synthetic OXT by means of an osmotic minipump. In an attempt to localize the inhibitory effect of OXT on the stress-induced activity of the HPA axis described above, the antagonist was applied via reversed microdialysis into the septum or amygdala during a 5 min exposure to a mild emotional stressor, the elevated platform (Fig. 3; Neumann et al., 2000c). This stressor was chosen for two reasons: (1) using a mild stressor, which results in only a slight increase in the activity of the system to be studied (HPA axis), is of advantage as a ceiling effect of the neuroendocrine response can be prevented; and (2) this stressor should best reflect the situation on the elevated plus-maze without the chance to hide in the closed and protected arm. Thus, we can exclude that differences in neuroendocrine responses after treatment are due to treatment-induced behavioural differences (for example, changes in locomotor activity, percentage of time exposed to the open, exposed arms in case of plus-maze exposure). For monitoring the local effect of the OXT antagonist

154

within the PVN, rats were exposed to forced swimming. As described below (Fig. 3), a very different picture emerged. Administration of the OXT receptor antagonist via retrodialysis into the mediolateral septum Within the septum, OXT release in response to various stimuli (Landgraf et al., 1988; Neumann and Landgraf, 1989; Ebner et al., 2000), and OXT receptor expression and binding (Freund-Mercier et al., 1987; Yoshimura et al., 1993; for review see Verbalis, 1999) have been described. Furthermore, GABAergic neurones project from the septum to the hypothalamus, thus involving septal–hypothalamic pathways in the neuroendocrine regulation. After comparing OXT receptor antagonist- and vehicle-treated male rats, a local inhibitory effect of OXT on stress-induced activity of the HPA axis was found when animals were exposed to the elevated platform, that is, the antagonist treatment further increased stress-induced ACTH and corticosterone secretion (Fig. 3). This effect was found to be stressor-dependent since it was not found when animals were exposed to forced swimming. In response to swimming, ACTH and corticosterone secretion reached high levels in vehicle-treated rats probably preventing further elevation and disinhibition by the antagonist. On the other hand, we also have to consider the possibility that the inhibitory actions of OXT on HPA axis reactivity within the mediolateral septum may depend on the quality and intensity of the stressor. Administration of the OXT receptor antagonist via retrodialysis into the central amygdala There is much evidence for an involvement of the amygdala in the regulation of neuroendocrine and behavioural stress responses including the presence of CRH immunoreactivity, CRH receptors and bidirectional connections to various hypothalamic (including the PVN) and brainstem regions (for review see Feldman and Weidenfeld, 1995; Herman and Cullinan, 1997). Further, with respect to the involvement of OXT in HPA axis regulation, OXT receptors are present in the amygdala, although direct evidence for OXT actions on neuroendocrine functions

within this limbic brain region is lacking. However, anxiolytic effects of OXT within the amygdala were recently described (see below; Neumann et al., 2000c; Bale et al., 2001). Local administration of the OXT receptor antagonist into the amygdala was found to only slightly, but not significantly, potentiate ACTH secretion stimulated by exposure to the elevated plate for 5 min (Neumann et al., 2000c) (Fig. 3). Although further studies are necessary, our results do not fully support the hypothesis that OXT released within the amygdala exerts a regulatory action on basal and/or stress-induced HPA axis activity. Administration of the OXT receptor antagonist via retrodialysis into the PVN The PVN represents the main regulatory unit of the HPA axis integrating afferents from limbic, brainstem and hypothalamic regions and regulating appropriate hormonal output, as for example, secretion of CRH into the portal blood and, as a consequence, the synthesis and secretion of ACTH and corticosterone. Hypothalamic co-expression of CRH and OXT (Sawchenko et al., 1984), and the demonstration of CRH receptors on OXT neurones (Hisano et al., 1992) point towards a close interaction of the OXT and CRH systems at PVN level. Thus, according to our hypothesis, OXT actions may be mediated via the CRH system in the brain. As mentioned above, basal HPA axis activity was found to be inhibited by OXT released within the PVN. In contrast, the stress-induced rise in ACTH and corticosterone secretion was found to be significantly reduced in antagonist-treated (via bilateral retrodialysis) compared to vehicle-treated rats (Fig. 3) demonstrating that local, endogenous OXT potentiates the HPA axis response. This rather surprising result points towards a dual mechanism of action of OXT released within the PVN. Under resting conditions, OXT acts as an endogenous inhibitor of HPA axis activity, whereas under conditions of stress and increased OXT release within the PVN it seems to support the stress-induced rise of HPA axis hormones. From here, interesting questions emerge as to the mechanisms of action depending on the concentration of OXT in the extracellular space, the onset of inhibitory/excitatory actions during a stress

155

response, or other local neuromodulators acting at the pre- or postsynaptic level. In situ hybridization studies combined with immunocytochemistry may answer such questions. Involvement of brain OXT in emotional stress responses Anxiety-related behaviour A variety of behavioural effects of brain OXT have been described including social, sexual and maternal behaviour, and various reviews focus on these functions (Argiolas and Gessa, 1991; Insel, 1992; Engelmann et al., 1996). With respect to emotionality, OXT (in most cases administered peripherally) has been shown to induce antidepressive, sedative and anxiolytic effects and to alter the behavioural stress response in rats or mice (King et al., 1985; Arletti and Bertolini, 1987; Uvnäs-Moberg et al., 1994; McCarthy et al., 1996). However, underlying mechanisms of these behavioural effects remain to be shown, as endogenous OXT cannot cross the blood–brain barrier in physiologically relevant amounts. Substantial support for effects of OXT on the behavioural stress response at brain level comes from a study by Windle et al. (1997), in which reduced anxiety-related behaviour was described after chronic i.c.v. administration of OXT into the brain of ovariectomized rats. In an attempt to block the receptor-mediated action of OXT within the brain

of virgin female, male, pregnant and lactating rats we administered vehicle or the OXT receptor antagonist i.c.v. 10 min before behavioural testing on the elevated plus-maze (Fig. 4). The exploration of the open arms of the maze, indication for anxietyrelated behaviour, was unaffected by this treatment in both virgin and male rats (with male rats being significantly more anxious), demonstrating that brain OXT is not directly involved in the regulation of anxiety under these experimental conditions (Fig. 4; Neumann et al., 2000a). In contrast, the antagonist significantly reduced the percentage of entries into the open arms in pregnant (between days 19 and 21 of pregnancy) and lactating rats and also the percentage of time spent in the open arms of the maze in pregnant rats indicating increased anxiety (Fig. 4; Neumann et al., 2000b). As a consequence, we have to conclude that the anxiolytic action of brain OXT is restricted to the peripartum period in rats, when the activity of the brain OXT system is high. With respect to the site of action of the anxiolytic effect of OXT, we found evidence for the involvement of the central amygdala, at least in late pregnant rats (Fig. 5B; Neumann et al., 2000c). Bilateral application of the OXT receptor antagonist into the central part of the amygdala by means of silicon fibres in order to guarantee minimal tissue disruption, increased anxiety-related behaviour on the elevated plus-maze indicating in a convincing manner that OXT acts as an endogenous anxiolytic at this site in the peripartum period. These results were confirmed

Fig. 4. Anxiety-related behaviour of virgin, male, pregnant and lactating rats on the elevated plus-maze treated with i.c.v. vehicle (5 μl) or oxytocin receptor antagonist (0.75 μg/5 μl) 10 min before behavioural testing. Data are the means ± SEM. * P < 0.05 versus vehicle; # P < 0.05 versus respective vehicle/virgin value if separate statistics is performed. Data from Neumann et al., 2000a,b.

156

Fig. 5. Effects of an oxytocin receptor antagonist applied (A) i.c.v. or (B) bilaterally into the central amygdala of late pregnant rats (between pregnancy days 19 and 21) 10 min before testing on anxiety-related behaviour on the elevated plus-maze (5 min) as reflected by open arm exploration. Results in B indicate that the anxiolytic effect of brain OXT is partly mediated via the central amygdala. Note that the locomotor activity (number of closed arm entries) was not treatment dependent.

recently by McCarthy’s group who demonstrated that acute administration of OXT bilaterally into the central amygdala, but not into the ventromedial nucleus of the hypothalamus of ovariectomized rats, reduced anxiety-related behaviour in the open field test, as indicated by an increased entering of the central part of the arena. In contrast, the general locomotor activity remained unchanged (Bale et al., 2001). Dopaminergic neurotransmission within the amygdala was found to be essential for this OXT receptor-mediated anxiolytic effect (Bale et al., 2001).

During pregnancy and lactation, various behavioural adaptations, including emotional changes, occur at a time of increased activity of the brain OXT system. Interestingly, reduced anxiety has been shown in lactating rats by studying the duration of freezing and response to an auditory stimulus (Hard and Hansen, 1984) as well as on the elevated plus maze (Neumann, 2001). In contrast, we recently described increasing anxiety between pregnancy days 10 and 21 (Neumann et al., 1998). Therefore, detailed studies correlating emotionality and activity of the brain OXT system

157

in relevant limbic brain regions are required to draw final conclusions. In all animals studied, the locomotor activity reflected by the number of entries into the closed arms was unchanged after OXT antagonist treatment applied either i.c.v. or into the amygdala, which indicates that endogenous OXT does not exert a general sedative effect (Fig. 5).

ternal behaviour in lactation (Fahrbach et al., 1985; McCarthy et al., 1991). Therefore, our results point towards the possibility that brain OXT might be essential for the maintenance of maternal behaviour under stressful circumstances.

Maternal aggressive behaviour

If we consider interactions between the OXT system and the HPA axis we also have to test for actions of hormones of the HPA axis on the activity of the OXT system. Fundamental effects of glucocorticoid receptor ligands on the neuronal activity within the PVN including the CRH and vasopressinergic systems have been described (Sawchenko, 1987; Herman, 1995; Akana and Dallman, 1997; Kovacs, 1998; Pinnock and Herbert, 2001). However, glucocorticoid effects on OXT neurones have not been investigated in detail. In a attempt to address this question we monitored OXT release into the blood and within the hypothalamic PVN in male sham-operated and adrenalectomized (ADX) Wistar rats under basal conditions and in response to swim stress (Torner et al., 2000). As expected, ADX resulted in high basal levels of ACTH due to the lack of corticosterone negative feedback effect; replacement of glucocorticoids by implantation of a subcutaneous corticosterone pellet at the time of ADX prevented the basal rise in ACTH secretion. With respect to the OXT system, basal secretion of OXT into the blood was not altered by ADX, but the swim-induced rise in OXT plasma levels was significantly higher after ADX and remained at a high level for 60 min, whereas in sham-operated rats, OXT concentrations already reached resting levels 60 min after stressor exposure (Fig. 6). Replacement prevented the prolonged OXT stress response (Torner et al., 2000). Thus, we may conclude that glucocorticoids not only exert a negative feedback effect on the HPA axis itself, but also on other stress-relevant neuroendocrine systems including the oxytocinergic system. Studies using acute corticosterone replacement in both sham-operated rats and in rats after ADX are currently in progress in order to further prove this hypothesis (Thrivikraman, Torner, Plotsky and Neumann, unpublished observations). The effects of glucocorticoids on the release of OXT within the PVN seem to be different from those

OXT is strongly implicated in the control of social behaviours facilitating mother–infant interactions (Pedersen and Prange, 1979), pair-bonding (Williams et al., 1994; Insel and Hulihan, 1995) and social recognition of conspecifics (Popik et al., 1992; Engelmann et al., 1998). With respect to the involvement of OXT in aggressive behaviour, results seem to depend on the species, gender and strain studied, as well as the experimental protocol used (Ferris et al., 1992; Silakov et al., 1995; Consiglio and Lucion, 1996; Winslow et al., 2000). Female aggression is part of the complex behavioural patterns described as maternal behaviour and is exclusively related to the protection and defence of the offspring (Moyer, 1968). Thus, maternal aggression has to be distinguished from irritable, territorial, sex-related, fear-induced, predatory and intermale aggression. In an attempt to study the role of endogenous brain OXT in the expression of offensive, defensive, exploratory and maternal behaviour of a lactating rat during exposure to a virgin female intruder rat in the presence of the pups (maternal defence test), lactating rats were implanted with an i.c.v. guide cannula 5 days before the experiment and, 10 min before the start of the test, infused either with an OXT antagonist or vehicle (Neumann et al., 2001). Although neuroendocrine parameters were found altered by this treatment (see above), the performance of maternal aggressive behaviour was not changed (Fig. 5B). In contrast, antagonist-treated rats showed less maternal behaviour (including nest building, carrying pups, occasional licking) during the presence of the intruder (Fig. 5B). Until now, it has been believed that brain OXT is essential for the establishment of maternal behaviour around birth as well as the initiation of maternal behaviour in non-lactating rats (Pedersen and Prange, 1979; Van Leengoed et al., 1986), rather than for the maintenance of ma-

Glucocorticoid actions on the OXT system

OXT (pg/ml)

158

20

blood

10

basal

swim +5

+60 min

OXT (pg/microdialysate)

sham ADX 7.0

PVN 5.0

2.5

basal

swim 1

swim 2

Fig. 6. Oxytocin in plasma (upper panel) and microdialysates from the PVN (lower panel) in sham-operated and adrenalectomized (ADX) rats under basal conditions and in response to forced swimming. Data are from Torner et al. (2000).

on neurohypophysial release. Whereas basal somatodendritic release of OXT within the PVN was indistinguishable between sham-operated and ADX rats, the swim-induced rise in local OXT release (10 min, 19°C) was significantly reduced in ADX compared to sham-operated rats and this effect was partly reversed in corticosterone-substituted ADX rats (Fig. 6, lower panel; Torner et al., 2000). This shows that increased concentrations of circulating glucocorticoids as a result of a stress-activated HPA axis act back on hypothalamic OXT neurones which, in turn, become excited. Neuronal excitation may result, among others, in local somatodendritic release, although such release was found to be independent of the occurrence of action potentials (Ludwig, 1998).

The result of a facilitatory effect of corticosterone on the intra-PVN release of OXT together with the finding of an excitatory effect of locally released OXT on the stress-induced reactivity of the HPA axis point towards the interesting hypothesis of an integrated circuitry involving elements of the HPA axis and the OXT system to modulate neuroendocrine responses to stressors (Fig. 7). Abbreviations ACTH ADX CRH HPA

adrenocorticotropic hormone adrenalectomy corticotropin-releasing hormone hypothalamo–pituitary–adrenal

159

Fig. 7. Hypothetical scheme demonstrating the interactions between OXT and the HPA axis at the level of the hypothalamic PVN under basal (A) and stress (B) conditions. −, inhibitory effect; +, excitatory effect; dashed line, no effect.

OXT PVN SON

oxytocin paraventricular nucleus supraoptic nucleus

Acknowledgements I would like to thank O. Bosch, A. Douglas, K. Ebner, S. Krömer, R. Landgraf, L. Torner, N. Toschi, K.V. Thrivikraman, P. Plotsky and A. Wigger for their contributions to the studies reported, and K. Moschke and M. Zimbelmann for expert technical assistance. These projects were supported by Deutsche Forschungsgemeinschaft (Ne 645). References Abel, E.L. (1994) A further analysis of physiological changes in rats in the forced swim test. Physiol. Behav., 56: 795–800. Akana, S.F. and Dallman, M.F. (1997) Chronic cold in adrenalectomized, corticosterone (B)-treated rats: facilitated corticotropin responses to acute restraint emerge as B increases. Endocrinology, 138: 3249–3258. Argiolas, A. and Gessa, G.L. (1991) Central functions of oxytocin. Neurosci. Biobehav. Rev., 15: 217–231. Arletti, R. and Bertolini, A. (1987) Oxytocin acts as an antidepressant in two animal models of depression. Life Sci., 41: 1725–1730. Bale, T.L., Davis, A.M., Auger, A.P., Dorsa, D.M. and McCarthy, M.M. (2001) CNS region-specific oxytocin receptor expression: importance in regulation of anxiety and sex behavior. J. Neurosci., 21: 2546–2552.

Brinton, R.E., Wamsley, J.K., Gee, K.W., Wan, Y.-P. and Yamamura, H.I. (1984) [3 H]Oxytocin binding sites in the rat brain demonstrated by quantitative light microscopic autoradiography. Eur. J. Pharmacol., 102: 365–367. Carter, D.A. and Lightman, S.L. (1987) Oxytocin responses to stress in lactating and hyperprolactinaemic rats. Neuroendocrinology, 46: 532–537. Consiglio, A.R. and Lucion, A.B. (1996) Lesion of hypothalamic paraventricular nucleus and maternal aggressive behavior in female rats. Physiol. Behav., 59: 591–596. Demotes-Mainard, J., Chauveau, J., Rodriguez, F., Vincent, J.D. and Poulain, D.A. (1986) Septal release of vasopressin in response to osmotic, hypovolemic and electrical stimulation in rats. Brain Res., 381: 314–321. Dunn, F.L., Brennan, T.J., Nelson, A.E. and Robertson, G.L. (1973) The role of blood osmolality and volume in regulating vasopressin secretion in the rat. J. Clin. Invest., 52: 3212– 3219. Ebner, K., Wotjak, C.T., Landgraf, R. and Engelmann, M. (2000) A single social defeat experience selectively stimulates the release of oxytocin, but not vasopressin, within the septal brain area of male rats. J. Neuroendocrinol., 872: 87–92. Ehlers, C.L., Reed, T.K. and Henriksen, S.J. (1986) Effects of corticotropin-releasing factor and growth hormone-releasing factor on sleep and activity in rats. Neuroendocrinology, 42: 467–474. Engelmann, M., Wotjak, C.T., Ludwig, M., Neumann, I.D. and Landgraf, R. (1996) Behavioral consequences of intracerebral vasopressin and oxytocin: focus on learning and memory. Neurosci. Biobehav. Rev., 20: 341–358. Engelmann, M., Ebner, K., Wotjak, C.T. and Landgraf, R. (1998) Endogenous oxytocin is involved in short-term olfactory memory in female rats. Behav. Brain Res., 19: 89–94. Engelmann, M., Ebner, K., Landgraf, R., Holsboer, F. and Wot-

160 jak, C.T. (1999) Emotional stress triggers intrahypothalamic but not peripheral release of oxytocin in male rats. J. Neuroendocrinol., 11: 867–872. Engelmann, M., Ludwig, M., Singewald, N., Ebner, K., Sabatier, N., Lubec, G., Landgraf, R. and Wotjak, C.T. (2001) Taurine modulates the secretory activity of vasopressinergic neurons in conscious rats. Eur. J. Neurosci., 14: 1047–1055. Fahrbach, S.E., Morrell, J.I. and Pfaff, D.W. (1985) Possible role for endogenous oxytocin in estrogen-facilitated maternal behavior in rats. Neuroendocrinology, 40: 526–532. Feldman, S. and Weidenfeld, J. (1995) Neural mechanisms involved in the corticosteroid feedback effects on the hypothalamo–pituitary–adrenocortical axis. Prog. Neurobiol., 45: 129–141. Feldman, S., Conforti, N. and Weidenfeld, J. (1995) Limbic pathways and hypothalamic neurotransmitters mediating adrenocortical responses to neural stimuli. Neurosci. Biobehav. Rev., 19: 235–240 Ferris, C.F., Foote, K.B., Meltser, H.M., Plenby, M.G., Smith, K.L. and Insel, T.R. (1992) Oxytocin in the amygdala facilitates maternal aggression. Ann. N.Y. Acad. Sci., 652: 456– 457. Freund-Mercier, M.J., Stoeckel, E., Palacios, J.M., Pazos, A., Reichhart, J.M., Porte, A. and Richard, P. (1987) Pharmacological characteristics and anatomical distribution of [3H]oxytocin-binding sites in the Wistar rat brain studied by autoradiography. Neuroscience, 20: 599–614. Haller, I., Barna, L. and Baranyi, M. (1995) Hormonal and metabolic responses during psychosocial stimulation in aggressive and nonaggressive rats. Psychoneuroendocrinology, 20: 65–74 Hard, E. and Hansen, S. (1984) Reduced fearfulness in the lactating rat. Physiol. Behav., 35: 641–643. Hashiguchi, H., Ye, S.H., Morris, M. and Alexander, N. (1997) Single and repeated environmental stress: effect on plasma oxytocin, corticosterone, catecholamines, and behavior. Physiol. Behav., 61: 731–736. Heinrichs, M., Meinlschmidt, G., Neumann, I., Wagner, S., Kirschbaum, C., Ehlert, U. and Hellhammer, D.H. (2001) Effects of suckling on hypothalamic–pituitary–adrenal axis responses to psychosocial stress in postpartum lactating women. J. Clin. Metab., 86: 4798–4804. Herman, J.P. (1995) In situ hybridisation analysis of vasopressin gene transcription in the paraventricular and supraoptic nuclei of the rat: regulation by stress and glucocorticoids. J. Comp. Neurol., 363: 15–27. Herman, J.P. and Cullinan, W.E. (1997) Neurocircuitry of stress: central control of the hypothalamo–pituitary–adrenocortical axia. Trends Neurosci., 20: 78–83. Higuchi, T., Honda, S., Takano, S. and Negoro, H. (1988) Reduced oxytocin response to osmotic stimulus and immobilisation stress in lactating rats. J. Endocrinol., 116: 225–230. Hisano, S., Li, S., Kagotani, Y. and Daikoku, S. (1992) Synaptic associations between oxytocin-containing magnocellular neurons and neurons containing corticotropin-releasing factor in the rat magnocellular paraventricular nucleus. Brain Res., 576: 311–318. Hussy, N. (2002) Glial cells in the hypothalamo-neurohypo-

physial system: key elements of the regulation of neuronal electrical and secretory activity. In: D. Poulain, S. Oliet and D. Theodosis (Eds.), Vasopressin and Oxytocin: From Genes to Clinical Application. Progress in Brain Research, Vol. 139. Elsevier Science, Amsterdam, pp. 95–112. Insel, T.R. (1990) Regional changes in brain OT receptors postpartum: time-course and relationship to maternal behaviour. J. Neuroendocrinol., 2: 539–545. Insel, T.R. (1992) Oxytocin — a neuropeptide for affiliation: evidence from behavioral, receptor autoradiographic, and comparative studies. Psychoneuroendocrinology, 17: 3–35. Insel, T.R. and Hulihan, T.J. (1995) A gender-specific mechanism for pair bonding: oxytocin and partner preference formation in monogamous voles. Behav. Neurosci., 109: 782–789. Kasting, N.W. (1988) Simultaneous and independent release of vasopressin and oxytocin in the rat. Can. J. Physiol. Pharmacol., 66: 22–26. Kendrick, K.M., Keverne, E.B., Chapman, C. and Baldwin, B.A. (1988) Intracranial dialysis measurement of oxytocin, monoamines and uric acid release from the olfactory bulb and substantia nigra of sheep during parturition, suckling, separation from lambs and eating. Brain Res., 439: 1–10. King, M.G., Brown, R. and Kusnecov, A. (1985) An increase in startle response in rats administered oxytocin. Peptides, 6: 567–568. Kovacs, K.J. (1998) Functional neuroanatomy of the parvocellular vasopressinergic system: transcriptional responses to stress and glucocorticoid feedback. Prog. Brain Res., 119: 31–43. Lambert, R.C., Moos, F. and Richard, P. (1993) Action of endogenous oxytocin within the paraventricular and supraoptic nuclei: a powerful link in the regulation of the bursting pattern of oxytocin neurons during the milk ejection reflex in rats. Neuroscience, 57: 1027–1038. Landgraf, R. (1995) Intracerebrally released vasopressin and oxytocin: measurement, mechanisms and behavioural consequences. J. Neuroendocrinol., 7: 243–253. Landgraf, R., Neumann, I. and Schwarzberg, H. (1988) Central and peripheral release of vasopressin and oxytocin in the conscious rat after osmotic stimulation. Brain Res., 457: 219– 225. Landgraf, R., Neumann, I. and Pittman, Q.J. (1991) Septal and hippocampal release of vasopressin and oxytocin during late pregnancy and parturition in the rat. Neuroendocrinology, 54: 378–383. Lang, R.E., Heil, J.W.E., Ganten, D., Hermann, K., Unger, T. and Rascher, W. (1983) Oxytocin unlike vasopressin is a stress hormone in the rat. Neuroendocrinology, 37: 314–316. Legros, J.J. and Ansseau, M. (1992) Neurohypophyseal peptides and psychopathology. In: D.F. Swaab, M.A. Hofman, M. Mirmiran, R. Ravid and F.W. van Leeuwen (Eds.), Progress in Brain Research, Vol. 93, Elsevier, Amsterdam, pp. 455–461. Leshner, A.I. and Politch, J.A. (1979) Hormonal control of submissiveness in mice: irrelevance of the androgens and relevance of the pituitary–adrenal hormones. Physiol. Behav., 22: 531–534. Lightman, S.L. (1992) Alterations in hypothalamic–pituitary responsiveness during lactation. In: C. Pedersen et al. (Eds.),

161 Oxytocin in Maternal, Sexual and Social Behaviors. Ann. N.Y. Acad. Sci., 652: 340–346. Lightman, S.L. and Young III, W.S. (1989) Lactation inhibits stress-mediated secretion of corticosterone and oxytocin and hypothalamic accumulation of corticotropin-releasing factor and enkephalin messenger ribonucleic acids. Endocrinology, 124: 2358–2364. Lucki, I. (1997) The forced swim test as a model for core and component behavioral effects of antidepressant drugs. Behav. Pharmacol., 8: 523–532. Ludwig, M. (1998) Dendritic release of vasopressin and oxytocin. J. Neuroendocrinol., 10: 881–895. McCarthy, M.M., Chung, S.K., Ogawa, S., Kow, L.M. and Pfaff, D.W. (1991) Behavioral effects of oxytocin: is there a unifying principle. In: S. Jard and R. Jamison (Eds.), Vasopressin, Colloque INSERM, Vol. 208, John Libbey Eurotext, pp. 195– 212. McCarthy, M.M., McDonald, C.H., Brooks, P.J. and Goldmann, D. (1996) An anxiolytic action of oxytocin is enhanced by estrogen in the mouse. Physiol. Behav., 60: 1209–1215. Moos, F., Poulain, D.A., Rodriguez, F., Guerné, Y., Vincent, J.-D. and Richard, P. (1989) Release of oxytocin within the supraoptic nucleus during the milk ejection reflex in rats. Exp. Brain Res., 76: 593–602. Moos, F. and Richard, P. (1989) Paraventricular and supraoptic bursting oxytocin cells in rat are locally regulated by oxytocin and functionally related. J. Physiol., 408: 1–18. Moyer, K.E. (1968) Kinds of aggression and their physiological basis. Commun. Behav. Biol., 2: 65–87. Neumann, I.D. (2001) Alterations in behavioral and neuroendocrine stress coping strategies in pregnant, parturient and lactating rats. In: Progress in Brain Research, Vol. 133. Elsevier Science, Amsterdam, pp. 143–152. Neumann, I. and Landgraf, R. (1989) Septal and hippocampal release of oxytocin, but not vasopressin, in the conscious lactating rat during suckling. J. Neuroendocrinol., 1: 305–308. Neumann, I., Russell, J.A. and Landgraf, R. (1993a) Oxytocin and vasopressin release within the supraoptic and paraventricular nuclei of pregnant, parturient and lactating rats: a microdialysis study. Neuroscience, 53: 65–75. Neumann, I., Ludwig, M., Engelmann, M., Pittman, Q.J. and Landgraf, R. (1993b) Simultaneous microdialysis in blood and brain: oxytocin and vasopressin release in response to central and peripheral osmotic stimulation and suckling in the rat. Neuroendocrinology, 58: 637–645. Neumann, I., Koehler, E., Landgraf, R. and Summy-Long, J.Y. (1994) An oxytocin receptor antagonist infused into the supraoptic nucleus attenuates intranuclear and peripheral release of oxytocin during suckling in conscious rats. Endocrinology, 134: 141–148. Neumann, I.D., Johnstone, H.A., Hatzinger, M., Liebsch, G., Shipston, M., Russell, J.A., Landgraf, R. and Douglas, A.J. (1998) Attenuated neuroendocrine responses to emotional and physical stressors in pregnant rats involve adenohypophyseal changes. J. Physiol., 508: 289–300. Neumann, I.D., Wigger, A., Torner, L., Holsboer, F. and Landgraf, R. (2000a) Brain oxytocin inhibits basal and stress-

induced activity of the hypothalamo–pituitary–adrenal axis in male and female rats: partial action within the paraventricular nucleus. J. Neuroendocrinol., 12: 235–243. Neumann, I., Torner, L. and Wigger, A. (2000b) Brain oxytocin: differential inhibition of neuroendocrine stress responses and anxiety-related behaviour in virgin, pregnant and lactating rats. Neuroscience, 95: 567–575. Neumann, I.D., Krömer, S.A., Toschi, N. and Ebner, K. (2000c) Brain oxytocin inhibits the (re)activity of the hypothalamo– pituitary–adrenal axis in male rats: involvement of hypothalamic and limbic brain regions. Regul. Pept., 1: 31–38. Neumann, I., Toschi, N., Ohl, F., Torner, L. and Kroemer, S. (2001) Maternal defence as an emotional stressor in female rats: correlation of neuroendocrine and behavioural parameters and involvement of brain oxytocin. Eur. J. Neurosci., 13: 1016–1024. Nishioka, T., Anselmofranci, J.A., Li, P., Callahan, M.F. and Morris, M. (1998) Stress increases oxytocin release within the hypothalamic paraventricular nucleus. Brain Res., 781: 57–61. Opp, M.R. (1995) Corticotropin-releasing hormone involvement in stressor-induced alterations in sleep and in the regulation of waking. Adv. Neuroimmunol., 5: 127–143. Popik, P., Vetulani, J. and Van Ree, J.M. (1992) Low doses of oxytocin facilitate social recognition in rats. Psychopharmacology, 106: 71–74. Pedersen, C.A. and Prange, A.J. (1979) Induction of maternal behavior in virgin rats after intracerebroventricular administration of oxytocin. Proc. Natl. Acad. Sci. USA, 76: 6661– 6665. Pinnock, S.B. and Herbert, J. (2001) Corticosterone differentially modulates expression of corticotropin releasing factor and arginine vasopressin mRNA in the hypothalamic paraventricular nucleus following either acute or repeated restraint stress. Eur. J. Neurosci., 13: 576–584. Rosenblatt, J.S., Factor, E. and Mayer, A.D. (1994) Relationship between maternal aggression and maternal care in the rat. Aggress. Behav., 20: 243–255. Sawchenko, P.E. (1987) Adrenalectomy-induced enhancement of CRF and vasopressin immunoreactivity in parvocellular neurosecretory neurons: anatomic peptide, and steroid specificity. J. Neurosci., 7: 1093–1106. Sawchenko, P.E., Swanson, L.W. and Vale, W.W. (1984) Corticotropin-releasing factor: co-expression within distinct subsets of oxytocin-, vasopressin-, and neurotensinimmunoreactive neurons in the hypothalamus of male rats. J. Neurosci., 4: 1118–1129. Silakov, V.L., Nikitin, V.S., Moiseeva, L.A. and Losev, S.S. (1995) Influence of neuropeptides on the processes of higher nervous activity in primates. Neurosci. Behav. Physiol., 25: 386–392. Torner, L., Thrivikraman, K.V., Toschi, N., Holsboer, F., Plotsky, P.M. and Neumann, I.D. (2000) Effect of adrenalectomy and corticosterone on hypothalamic and peripheral release of oxytocin and behavioral stress coping in male rats. Abstract 30th Annual Neuroscience Meeting, New Orleans, 153.12. Torner, L., Toschi, N., Pohlinger, A., Landgraf, R. and Neumann, I.D. (2001) Anxiolytic and anti-stress effects of brain pro-

162 lactin: improved efficacy of antisense targeting of the prolactin receptor by molecular modeling. J. Neurosci., 21: 3207–3214. Uvnäs-Moberg, K., Ahlenius, S., Hillegaart, V. and Alster, P. (1994) High doses of oxytocin cause sedation and low doses cause an anxiolytic-like effect in male rats. Pharmacol. Biochem. Behav., 49: 101–106. Van Leengoed, E., Kerker, E. and Swanson, H.H. (1986) Inhibition of postpartum maternal behaviour in the rat by injecting an oxytocin antagonist into the cerebral ventricles. J. Endocrinol., 112: 275–282. Verbalis, J.G. (1999) The brain oxytocin receptor(s)?. Front. Neuroendocrinol., 20: 146–156. Verbalis, J.G., McHale, C.M., Gardiner, T.W. and Stricker, E.M. (1986) Oxytocin and vasopressin secretion in response to stimuli producing learned taste aversions in rat. Behav. Neurosci., 100: 466–475. Wakerley, C., Foreman, C.T. and Ingram, C.D. (1989) Effect of centrally administered oxytocin on the association between cortical electroencephalogram and milk ejection in the rat. J. Neuroendocrinol., 1: 173–177. Walker, C.-D., Trottier, G., Rochford, J. and Lavallée, D. (1995) Dissociation between behavioral and hormonal responses to the forced swim stress in lactating rats. J. Neuroendocrinol., 7: 615–622. Welch, B.L. and Klopfer, P.H. (1961) Endocrine variability as a factor in the regulation of population density. Am. Naturalist, 95: 256–260. Wigger, A. and Neumann, I.D. (2002) Endogenous opioid regulation of stress-induced oxytocin release within the hypothalamic paraventricular nucleus is reversed in late pregnancy: a microdialysis study. Neuroscience, 112: 121–129.

Williams, J.R., Insel, T.R., Harbaugh, C.R. and Cater, C.S. (1994) Oxytocin administered centrally facilitates formation of a partner preference in female prairie voles (Microtus ochrogaster). J. Neuroendocrinol., 6: 247–250. Windle, R.J., Shanks, N., Lightman, S.L. and Ingram, C.D. (1997) Central oxytocin administration reduces stress-induced corticosterone release and anxiety behavior in rats. Endocrinology, 138: 2829–2834. Winslow, J.T., Hearn, E.F., Ferguson, J., Young, L.J., Matzuk, M.M. and Insel, T.R. (2000) Infant vocalization, adult aggression, and fear behavior of an oxytocin null mutant mouse. Horm. Behav., 37: 145–155. Wotjak, C.T., Masaharu, K., Liebsch, G., Montkowski, A., Holsboer, F., Neumann, I. and Landgraf, R. (1996) Release of vasopressin within the rat paraventricular nucleus in response to emotional stress: a novel mechanism of regulating adrenocorticotropic hormone secretion?. J. Neurosci., 16: 7725–7732. Wotjak, C.T., Ganster, J., Kohl, G., Holsboer, F., Landgraf, R. and Engelmann, M. (1998) Dissociated central and peripheral release of vasopressin, but not oxytocin, in response to repeated swim stress: new insights into the secretory capacities of peptidergic neurons. Neuroscience, 85: 1209–1222. Wotjak, C.T., Naruo, T., Muraoka, S., Simchen, R., Landgraf, R. and Engelmann, M. (2001) Swim stress stimulates the expression of vasopressin and oxytocin in magnocellular neurons of the hypothalamic paraventricular nucleus. Eur. J. Neurosci., 13: 2273–2281. Yoshimura, R., Kiyama, H., Kimura, T., Araki, T., Maeno, H. and Tanizawa, O. (1993) Localization of oxytocin receptor messenger ribonucleic acid in the rat brain. Endocrinology, 133: 1239–1246.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 13

Expression of human vasopressin and oxytocin receptors in Escherichia coli Bernard Mouillac 1,∗ , Tuhinadri Sen 1 , Thierry Durroux 1 , Gérald Gaibelet 2 and Claude Barberis 1 1

INSERM U469, 141 rue de la Cardonille, 34094 Montpellier cedex 5, France 2 AVIDIS S.A., Biopôle Clermont-Limagne, 63360 Saint-Beauzire, France

Abstract: In order to produce large amounts of human vasopressin and oxytocin receptors compatible with direct structural biology approaches such as X-ray crystallography, NMR or mass spectrometry, we have expressed these neurohypophysial hormone receptors in Escherichia coli. To facilitate the level of expression, the coding sequence for the V1a vasopressin receptor and the oxytocin receptor were first optimized for bacterial expression. The resulting ‘bacterial receptor cDNAs’ were then subcloned into pET/T7-driven prokaryotic expression vectors. Different constructs have been prepared: each cDNA was incorporated alone or in fusion with a T7 tag sequence or a glutathione-S-transferase tag sequence at the N-terminus end. Moreover, a 6 × His tag sequence has been added at the C-terminus end for one-step purification of the receptors. Screening of BL21(DE3) and BL21(DE3)pLysS bacterial strains transformed with the different constructions was achieved by Coomassie blue-stained SDS–polyacrylamide gels and by 6 × His antibody Western blotting. Several clones were selected for purification of the receptors. Expression levels of the receptors are now encouraging and will be optimized for further structural and functional studies. Moreover, at the same time, the construction of the bacterial-optimized sequence of the V2 vasopressin receptor and its expression will be performed. Keywords: G protein-coupled receptor; Escherichia coli; Heterologous expression; Inclusion body; Receptor purification

Introduction G protein-coupled receptors (GPCRs) represent the largest known protein superfamily comprising about 1–3% of the mammalian genome, and play crucial roles in many biological processes in which pharmaceutical intervention may be useful (Bockaert and Pin, 1999). High level expression and native purification of GPCRs are important steps in the biochemical and structural characterization of these proteins. Despite tremendous effort from academic and industrial

∗ Correspondence to: B. Mouillac, INSERM U469, 141 rue de la Cardonille, 34094 Montpellier cedex 5, France. Tel.: +33-4-6714-2922; Fax: +33-4-6754-2432; E-mail: [email protected]

research, only the high-resolution three-dimensional (3D) X-ray structure of the bovine rhodopsin has been obtained to date (Palczewski et al., 2000). Rhodopsin is, however, the only GPCR occurring naturally at a very high concentration. Several major obstacles have so far prevented the obtention of important structural biology data for this GPCR superfamily. The first one is the limited availability of receptor protein. Its natural abundance is usually very low, and eukaryotic and prokaryotic expression systems are often inefficient in terms of expression level. In the best cases, a few micrograms of receptor protein per milligram of membrane proteins can be produced in eukaryotic expression systems such as mammalian cells (Pawate et al., 1998; Mirzabekov et al., 1999). Moreover, eukaryotic expression of GPCRs usually results in the produc-

164

tion of heterogeneous receptors characterized with many different post-translational modifications. For example, rhodopsin once again, has been shown to be a mixture of many compounds carrying various degrees of glycosylation, phosphorylation or other secondary modifications (Whitelegge et al., 1998). Prokaryotic expression, once it can be achieved, has the advantage of resulting in a more homogeneous preparation, because none of the mentioned post-translational modifications are carried out in bacteria. A few GPCRs, which have been expressed in bacteria, have been shown to be fully active in ligand binding as well as G protein-coupling. Consequently, the post-translational modifications do not seem to be indispensable for the basic function of many (but not all) receptors (Strosberg and Marullo, 1992; Grisshammer and Tate, 1995). Arginine-vasopressin (AVP) and oxytocin (OT) receptors are typical GPCRs and have to be considered as a representative model for the study of GPCR structure/function relationships (Barberis et al., 1998). The two neurohypophysial hormones are actively involved in the regulation of many physiological processes. AVP participates, for example, in the regulation of free water reabsorption, body fluid osmolality, blood pressure, cell contraction, cell proliferation. The best-known effects of OT are stimulation of the uterine contractions during labor and of milk ejection during lactation. Potential therapeutic use of AVP and OT receptor antagonists include: (1) the blockade of V1a AVP receptors in arterial hypertension, congestive heart failure and peripheral vascular diseases; (2) the blockade of V2 AVP receptors in the syndrome of inappropriate vasopressin secretion, congestive heart failure, liver cirrhosis, nephrotic syndrome and any state of excessive retention of free water and subsequent dilutional hyponatremia; (3) the blockade of V1b AVP receptors in adrenocorticotropin-secreting tumors; and (4) the blockade of OT receptors in preterm labor and dysmenorrhea (Freidinger and Pettibone, 1997; Gimpl and Fahrenholz, 2001; Thibonnier et al., 2001). Moreover, AVP and OT agonists are also potential therapeutic agents: analogs of AVP, such as desmopressin, are currently used in the treatment of central diabetes insipidus which results from inadequate secretion of AVP and could be used as well in the treatment of nocturnal enuresis. OT is widely

used in obstetrical practice to promote labor and delivery (Barberis et al., 1999). Understanding of the molecular determinants responsible for agonist and antagonist binding to AVP/OT receptors should provide valuable information that facilitates the rational design of potential therapeutic agents and should be helpful for a better knowledge of the molecular mechanisms regulating receptor function. Different strategies have to be considered to design AVP and OT therapeutic agents: (1) the systematic or rational modifications of the ligand structure (Manning et al., 1995); (2) the random screening for new chemical compounds (Serradeil-LeGal et al., 1996); and (3) the structurebased drug design. The latter requires the knowledge of the 3D structure of both the ligand and receptor. The AVP/OT receptors have been cloned and expressed in many heterologous systems but their X-ray structure still remains to be established. In the absence of such 3D structure data, the combination of complementary approaches has been developed in order to define functional domains of these receptors, particularly the ligand-binding pocket. In several laboratories, these studies have been investigated using the combination of 3D molecular modeling, expression and site-directed mutagenesis of the receptors, molecular pharmacology and direct biochemistry techniques such as photoaffinity labeling (Chini et al., 1995; Mouillac et al., 1995; Postina et al., 1996; Mendre et al., 1997; Cotte et al., 1998; Fanelli et al., 1999; Phalipou et al., 1999; Hawtin et al., 2000). The results have demonstrated that residues located in the transmembrane domains as well as residues within extracellular domains are involved in ligand binding. The major part of the agonist-binding pocket is delimited by the ring-like arrangement of the transmembrane domains and the first extracellular loop is crucial for determining the receptor agonist binding selectivity. Photoaffinity labeling of different domains of the human AVP V1a and the OT receptors using various photoactivatable peptide antagonist ligands has suggested that agonist and peptide antagonist binding pockets might be common to all receptor subtypes of the AVP/OT family and that specific residues differentiate agonist versus antagonist binding (Cotte et al., 2000; Breton et al., 2001). Other studies indicate that the binding sites for nonpeptide antagonists could be different

165

from the agonist/peptide antagonist binding sites (Thibonnier et al., 2000). In order to produce large amounts of human AVP and OT receptors compatible with direct structural biology approaches such as X-ray crystallography, NMR or mass spectrometry, we investigated the usefulness of a bacterial expression system. For a structural analysis of the receptors, the focus is on high expression levels (mg/l of bacterial culture) and protein homogeneity. From a pharmacological point of view, insect and mammalian cell expression systems have been the most frequently employed, and seem to be the most successful in terms of both authenticity of GPCR function and reasonable expression levels. However, biophysical studies require more substantial amounts of proteins. Taking into account that point, the ease of large-scale fermentation makes Escherichia coli much more attractive. In the future, overexpression of human AVP/OT receptors at very high levels should allow their large-scale purification, which could hopefully lead to their crystallization and structure determination, as required for a full understanding of their function. Optimization of the receptor nucleotide sequence for bacterial expression Due to the degeneracy of the genetic code, most amino acids are specified by more than one codon. These codons are not used at equal frequencies, their relative frequency varying with both the gene and the organism. In E. coli, there is a strong correlation between the frequency of a codon and the abundance of its corresponding tRNA (Gribskov et al., 1984). Moreover, in E. coli, genes of highly expressed proteins, like ribosomal proteins or major outer membrane proteins, use codons corresponding to the most abundant tRNAs almost exclusively. This is thought to be due to a need for efficient translation of RNAs of abundant proteins. Because codon usage in prokaryotes and in eukaryotes like mammalian cells is different, many eukaryotic genes that have been used for expression of recombinant proteins in bacteria contain codons that are not optimal for bacterial expression. Consequently, these E. coli rare codons can prevent the expression of the corresponding heterologous proteins because of translational stalling (Kane, 1995). One method by

which this codon bias could be overcome is to reengineer the eukaryotic gene to be overexpressed so that it uses the preferred codons of E. coli. We have applied this strategy to the human AVP/OT GPCRs and we have used the polymerase chain reaction (PCR) for the rapid construction of these synthetic codon-optimized copies of the receptor cDNAs. We have created artificial receptor cDNA fragments for the human V1a AVP and the human OT receptors in a simple three-step method. The principles of this three-step method are outlined in Fig. 1 (Dillon and Rosen, 1993). Design of a synthetic gene that contains codons that are preferentially utilized in the organism to be used for protein expression can be accomplished by reverse-translating the amino acid sequence of the desired protein into a nucleic acid sequence containing codons obtained from codon usage tables that have been generated for the organism of choice (Gribskov et al., 1984). In addition to changes in codon structure, the synthetic gene can be designed to include convenient restriction sites for cloning into appropriate vectors. In the case of AVP/OT receptors, each synthetic gene contains a BamHI and an EcoRI restriction site at the 5 and 3 extremities, respectively. The ‘bacterial receptor cDNAs’ were then subcloned into the commercial pET prokaryotic expression vector series (Novagen). The pET system is the most powerful system yet developed for the cloning and expression of recombinant proteins in E. coli. Target genes, cloned in pET plasmids, are under control of strong bacteriophage T7 transcription and translation signals; expression is realized by providing a source of T7 RNA polymerase in the host cell, like BL21(DE3) or BL21(DE3)pLysS strains. Expression is induced by the addition of isopropyl-β-D-thiogalactopyranoside (IPTG) to the bacterial culture. As illustrated in Fig. 2, the synthetic gene for V1a receptor was subcloned into pET 21b+ and pET 21a+ from which the receptor is expressed alone or as a protein tagged with the short T7 epitope, respectively. In both cases, the receptor is tagged at the C-terminus end with a 6 × His sequence epitope. In the pET 21a+ version, the T7 tag corresponds to the first 11 amino acids of the N-terminal end of the T7 bacteriophage major capside protein. While the addition of the T7 tag could help to the expression of the target protein, it can be useful as well for de-

166

Fig. 1. Description of the method for constructing the synthetic AVP/OT receptor genes. (A) Diagram of design and orientation of overlapping oligonucleotides. Each oligonucleotide is constituted with around 70 nucleotides and contains overlaps that are 20 nucleotides in length. (B) The PCR starts by mixing the overlapping oligonucleotides and should be carried out with enough cycles to generate a double-stranded PCR product that spans the full length of the synthetic gene. In the case of long synthetic genes, like cDNAs for AVP/OT receptors, oligonucleotides have been first mixed 6 × 6 in four different tubes and amplified before final mixing. (C) The product of the first PCR reaction is amplified in a second PCR reaction that incorporates short flanking primers which contain restriction sites for subsequent cloning.

tecting the recombinant protein by Western blotting. Competent BL21(DE3) or BL21(DE3)pLysS bacterial strains were transformed with the different vectors. Screening of bacterial clones containing the different constructs was achieved by Coomassie bluestained SDS–polyacrylamide gels and by 6 × His antibody Western blotting. As shown in Fig. 3, in standard induction conditions and for both constructs, the human V1a receptor cannot be visualized after staining the SDS–polyacrylamide gels with Coomassie blue. The results are also negative when using the 6 × His Western blotting procedure, which is, however, a much more sensitive method than Coomassie blue (nanograms versus micrograms of proteins). Moreover, expressing the V1a protein was very toxic to BL21(DE3) cells and to a less extent to BL21(DE3)pLysS cells. In conclusion, when expression is done in standard conditions, optimizing the sequence of the V1a gene for expression in bacteria, does not allow a detectable level of production of the receptor protein. The presence of the T7 tag

at the N-terminus of the receptor does not improve the expression level. It is noteworthy that expression levels of recombinant membrane proteins in bacteria, such as receptors, are usually low, making the purification of milligram amounts of these proteins a very difficult task. The reason is probably because the expression of a membrane protein is often toxic to the bacterial cell, and therefore high levels of recombinant protein are not tolerated (Grisshammer and Tate, 1995). It has been speculated that the toxicity results from inefficient membrane insertion of the recombinant protein, that could block the machinery responsible for establishing the correct topology of membrane proteins (Kiefer et al., 1996). Mutagenesis of the N-terminal end of the receptors The T7 expression system has been used to produce substantial amounts of target protein from a wide variety of genes, both prokaryotic and eukaryotic.

167

Fig. 2. Cloning of the synthetic genes into pET prokaryotic expression vectors. The synthetic cDNA fragments coding for the V1a and OT receptors were first verified by direct dideoxynucleotide sequencing and corrected for point mutations introduced during the PCR reactions. The nucleotide sequences of both receptors were then subcloned into the prokaryotic expression vector pET series using the BamHI and EcoRI restrictions sites. In pET 21a+ and pET 21b+ vectors, the receptor cDNA sequences are in phase with a short sequence coding for a 6 × His epitope, resulting in proteins tagged at their C-terminus. Cloning the receptor sequences into the pET 21a+ version, allows production of receptors tagged at their N-terminus end with the short T7 tag. This epitope corresponds to the first 11 amino acids of the N-terminal end of the T7 bacteriophage major capside protein, and could be helpful for detection or purification of the receptors.

However, a few proteins are made in disappointingly small amounts, like the V1a receptor. As explained above, the target protein itself may interfere with gene expression or with the integrity of the cell and could become toxic. In bacteria, another factor that appears to influence target protein stability and thus protein accumulation is the amino acid immediately following the N-terminal methionine (penultimate residue). The amino acid at this position determines the removal of N-terminal fMet. Processing is catalysed by methionyl aminopeptidase and is governed by the following relationship: the degree of removal decreases as the size of the penultimate residue side chain increases (Bachmair et al., 1986; Hirel et al., 1989). In practice, little or no processing is observed when the following amino acids occupy the penulti-

mate position: His, Gln, Glu, Phe, Met, Lys, Tyr, Trp or Arg. Processing is differentially realized with the remaining amino acids. Moreover, in vivo half-life of a protein is a function of its amino-terminal end and the relationship between a protein amino terminal amino acid and its stability in bacteria has been determined (i.e. the N-end rule) (Tobias et al., 1991). It has been reported that protein half-life is only 2 min when the following amino acids were present at the amino terminus: Arg, Lys, Phe, Leu, Trp and Tyr. These are strongly destabilizing residues. In contrast, all other amino acids confer half-lives of >10 h. In the case of the human V1a receptor, the Nterminal methionine is followed with Arg and Leu residues at positions 2 and 3, respectively. Based on the N-end rule these two residues are supposed

168

Fig. 3. Screening of the bacterial clones by Coomassie blue and 6 × His Western blotting. The human V1a receptor and its version fused with the short T7 tag were expressed in BL21(DE3) or BL21(DE3)pLysS bacterial strains. The culture was performed in standard conditions with vigorous shaking at 37°C in LB medium supplemented with ampicillin 100 μg/ml (or its analog carbenicillin). At an A600 of 0.6, protein expression was induced by addition of 1 mM IPTG and growth continued for 3 h. Cells were harvested by centrifugation and the pellets resuspended in Laemmli buffer. After sonication, a few microliters of each sample were loaded onto SDS–polyacrylamide gels. After running the gels, screening of the samples was done directly by staining the proteins with Coomassie blue (A,B). In parallel, chemiluminescence detection of the receptors was performed by Western blotting using an antibody (1 : 2000 dilution) directed against the 6 × His tag and directly coupled to horseradish peroxidase (C,D). In A and B, three clones transformed with the wild-type or the T7-tagged V1a receptors were compared to a control BL21(DE3) clone transformed with the empty vector. In C, BL21(DE3) cells were transformed with or without pET 21b+ containing or not the human V1a cDNA sequence. In D, BL21(DE3)pLysS cells were transformed with or without pET 21a+ containing or not the human V1a cDNA sequence.

to be very bad, and their presence could be responsible for the poor expression level and rapid degradation of the recombinant receptor. We thus decided to mutate Arg and Leu residues with Gly and Ala residues, respectively, which are considered as two stabilizing amino acids. Introduction of the

point mutations was verified by dideoxynucleotide sequencing and the mutated receptor was expressed in BL21(DE3) or BL21(DE3)pLysS bacterial strains. As seen in the left panel of Fig. 4, the mutated receptor was easily detectable using the Western blot and chemiluminescence procedures. However, the level

169

Fig. 4. Effect of point mutations in the receptor sequence and of glucose addition in the culture medium. The human V1a receptor and its mutant combining Gly and Ala residues at positions 2 and 3 respectively, were expressed in the BL21(DE3)pLysS bacterial strain. The culture was performed in standard conditions with vigorous shaking at 37°C in LB medium supplemented with ampicillin 100 μg/ml (or carbenicillin), with or without glucose 0.4%. At an A600 of 0.6, protein expression was induced by addition of 1 mM IPTG and growth continued for 3 h. Chemiluminescence detection of the receptors was realized as described in legend to Fig. 3. In the left panel, BL21(DE3)pLysS cells were transformed with or without pET 21a+ vector containing or not the mutant receptor. In the central panel, the same bacterial clone expressing the mutant receptor was treated with glucose 0.4%. Increasing quantities of material were put into wells II and III. In the right panel, BL21(DE3) cells expressing the wild-type receptor were treated or not with glucose 0.4%. Increasing quantities of material were put into wells III and IV.

of expression was still low and did not allow direct detection of the protein with Coomassie blue (not shown). The protein band, detected using the 6 × His tag, appeared at approximately 42 kDa. This molecular mass is smaller than the one expected for the receptor which should be visualized around 50 kDa. A 42-kDa protein is thus probably the result of proteolytic degradation of the receptor. In bacterial strains like BL21(DE3) or BL21(DE3) pLysS, expression of the T7 RNA polymerase is un-

der the control of the lacUV5 promoter. Sometimes, this enzyme can be produced even in the absence of the IPTG inducer. Consequently, if the target protein, like the V1a receptor, is sufficiently toxic to E. coli, this basal level (leakiness) of T7 RNA polymerase is negative to the growth of the cells. This derepression of the lacUV5 promoter is mediated by cyclic adenosine monophosphate (cAMP) (Grossman et al., 1998). It has been demonstrated that this derepression can be effectively avoided by including glucose

170

in the culture medium. Glucose inhibits cAMP production and acts as a lac repressor. In order to verify this effect, we added 0.4% of glucose in the bacterial LB medium before IPTG induction and expressed the mutant V1a receptor (containing Gly and Ala residues at positions 2 and 3). As illustrated in the central panel of Fig. 4, the addition of glucose allowed us to visualize the receptor protein with more ease by Western blot and chemiluminescence methods. The immunoreactive band at around 42 kDa was prominent but another band at around 55 kDa could also be detected. Interestingly, this upper band roughly corresponds to the expected size of the receptor protein. The calculated molecular weight of the protein is 49 kDa including the spacer arm between the natural Cterminal amino acid and the 6 × His tag. However, it is well known that some proteins with 6 × His tags attached run more slowly on SDS gels than equivalent untagged proteins, and may appear to be several kilodaltons larger than expected. This encouraging result prompted us to check whether it could be possible to detect the wild-type receptor when adding glucose in the bacterial culture medium. As seen in the right panel of Fig. 4, the human V1a receptor was easily detectable using the Western blot and chemiluminescence procedures. Once again, the receptor appeared as a unique band at around 42 kDa. This size is compatible with that of a degraded product. In conclusion, adding glucose in the culture medium or mutating the destabilizing residues located at the N-terminus end of the V1a receptor greatly improves the level of expression of the recombinant protein in E. coli. Combining both approaches allowed us to visualize the undegraded receptor at approximately 55 kDa. Expression of the receptors as fusion proteins High-level expression of membrane proteins in bacteria, and especially GPCRs, is very difficult, probably due to the fact that the cells cannot tolerate large amounts of the foreign protein within their membranes. Besides the toxicity, the expression is extremely low as a result of degradation as well. Factors which influence the success of E. coli expression include the host, the nature of the promoter employed, the growth conditions, the origin and the nature of the recombinant protein and its intracel-

lular location. The problem of low abundance has been successfully side-stepped in the case of soluble proteins by the recombinant route. In contrast, the production of large quantities of active membrane proteins by overexpression is not routine. Therefore, the development of strategies to overcome this barrier should significantly boost membrane protein structural research. In this regard, membrane proteins expressed as fusion proteins with different partners, as well as targeted to inclusion bodies (IBs), are potentially useful. Firstly, a few GPCRs have been expressed as maltose binding protein or thioredoxin fusions and purified in microgram amounts in the presence of detergent in functional forms (Tucker and Grisshammer, 1996). Schistosomal glutathioneS-transferase (GST) is also commonly used as a fusion partner when expressing proteins in E. coli. The usefulness of GST has been demonstrated by overexpressing first the rat olfactory receptor OR5 as a fusion protein (Kiefer et al., 1996) and later with other different GPCRs (Kiefer et al., 2000). Secondly, IBs seem to form when highly expressed recombinant proteins cannot be tolerated as soluble proteins in the cell cytoplasm. These IBs are amorphous electrondense structures seen as refractile particles under phase contrast microscopy (Williams et al., 1982). Although IBs have often been considered as undesirable, dead end products of protein expression, their formation can be an advantage as their isolation from cell homogenates is a convenient and effective first purification step. Proteins in IB form are insoluble aggregates that lack functional activity, but could represent an excellent starting point for producing them in large amounts, provided that procedures can be developed to reconstitute them in vitro. Moreover, proteins accumulated in IBs are relatively pure and are protected from proteolytic degradation. In addition, toxic proteins may not inhibit cell growth when present in inactive form as IBs. Obviously, the aggregated protein has to be refolded to its native state, but there has been substantial progress in the last several years in methodologies for the refolding of several α-helical membrane proteins and two GPCRs derived from IBs (Grigorieff et al., 1996; Kiefer et al., 1996). Finally, purification of 6 × Histagged proteins is compatible with solubilization of IBs using strong denaturing agents such as urea or guanidine hydrochloride.

171

Fig. 5. Construction of the cDNA coding for the GST-V1a 6 × His fusion protein. Incorporating a MluI/SpeI fragment from the commercial pET 42b+ plasmid into the pET 21a+ plasmid cut with MluI/NheI resulted in the obtention of a new vector containing the GST-fused V1a or OT cDNA sequences. In both cases, receptors expressed from these new constructions are still tagged at their C-terminus end with the 6 × His epitope.

We thus decided to produce the human V1a /OT receptors as fusion proteins with GST being the partner. As depicted in Fig. 5, we have used a combination of the available pET 21a+ and pET 42b+ plasmids, containing the optimized gene of V1a receptor and GST, respectively, to construct a new vector in which the fusion protein GST-V1a 6 × His is under the control of the strong T7 promoter. This new fusion protein was expressed by transformation of the vector into BL21(DE3) and BL21(DE3)pLysS strains. The clones were screened by Coomassie blue-stained SDS gels as well as Western blot using chemiluminescence and antibodies against the 6 × His tag. The result is shown on Fig. 6. The GST-fused receptor was easily detectable using the Western blot and chemiluminescence procedures but the level of production was still low and did not

allow direct detection of the protein with Coomassie blue (not shown). The protein appeared as a unique intense band with a molecular mass around 62 kDa. This size is smaller than the one expected for this new construct, approximately 80 kDa. Once again, the results suggested that the fusion protein was degraded. Interestingly, considering that Western blot procedure against the 6 × His tag allows detection of receptors which are only totally translated (the 6 × His tag is located at the C-terminus of the receptor), the band at 62 kDa should contain the entire V1a receptor. In that context, the fusion protein is probably degraded in the GST part. Engineering a specific cleavage site for thrombin or factor Xa between GST and the V1a receptor should lead to complete removal of the fusion partner and generate the undegraded purified receptor for subsequent structural studies. In

172

Fig. 6. Detection of the GST-V1a 6 × His fusion protein by western blot. BL21(DE3)pLysS cells were screened for expression of the GST-V1a 6 × His fusion protein using the Western blot and chemiluminescence procedures as described in legend to Fig. 3. The culture was performed in standard conditions with vigorous shaking at 37°C in LB medium supplemented with ampicillin 100 μg/ml (or carbenicillin). At an A600 of 0.6, protein expression was induced by addition of 1 mM IPTG and growth continued for 3 h. Two representative clones are shown.

conclusion, doing a fusion with GST did not allow us to significantly improve the level of expression of the receptor, when compared to that obtained for the mutant receptor or that observed in the presence of glucose in the culture medium. However, this construction is much more promising because of the protective effect of GST against proteolysis.

Purification of the V1a receptor by immobilized metal affinity chromatography Immobilized metal affinity chromatography (Porath et al., 1975) is a powerful method for purifying proteins with engineered 6 × His tags. The purification is usually done in one-step procedure and is

173

based on the affinity between the six neighboring His residues of the tag and an immobilized metal ion (usually nickel). The nickel is held by chelation with reactive groups covalently attached to a solid support (agarose for example). The Ni-NTA resin uses nitrilotriacetic acid (NTA) as the chelator. After unbound proteins are washed away, the target protein is recovered by elution with imidazole. This system is versatile and allows proteins to be purified under gentle, non-denaturing conditions, or in the presence of either high concentrations of strong denaturing agents urea or guanidine. We have applied this purification method to the human V1a /OT receptors expressed in bacteria. A flow chart describing the steps for purification of the recombinant receptors is detailed in Fig. 7. The method

Fig. 7. Overview of the V1a /OT receptor purification procedure.

has been performed for isolating the mutant V1a receptor (containing Gly and Ala residues at positions 2 and 3, respectively) and the GST-fused V1a receptor from a 100-ml starting culture volume. Samples taken from each step of the procedure have been checked for the presence of the receptors by Western blot and chemiluminescence approaches. As illustrated in Fig. 8, the results demonstrated that both mutant and GST-fused V1a receptors could be isolated in the centrifugation pellet after cell lysis, although a low amount of the GST-fused protein was still in the supernatant. We first deduced that the receptors were not cytoplasmic and were thus localized in a particulate fraction, probably membranes or IBs. After solubilization of IBs (and membranes), incubation of the samples with Ni-NTA agarose resin and elution in

174

Fig. 8. Purification of the mutant and GST-fused V1a receptors using immobilized metal affinity chromatography. For each construct, a 100-ml culture was started in standard conditions with vigorous shaking at 37°C in LB medium supplemented with ampicillin 100 μg/ml (or carbenicillin). At an A600 of 0.6, protein expression was induced by addition of 1 mM IPTG and growth continued for 3 h. As detailed in Fig. 7, extracts from bacterial cultures, expressing the mutant or the fusion protein, were prepared for immobilized metal affinity chromatography. Samples (a few microliters) taken from different steps of the purification procedure were checked for the presence of the receptors. Chemiluminescence detection of the receptors was realized as described in the legend to Fig. 3. For each panel, the legend is the following: TPE, total protein extract; S1, supernatant of centrifugation after cell sonication (lysis); S2, supernatant of centrifugation after inclusion bodies solubilization; FT1 and FT2, flow through after incubation of the Ni-NTA resin with receptors from S1 and S2 supernatants, respectively; W1 and W2, washed material of the Ni-NTA columns from S1 and S2 samples, respectively; E1 and E2, eluates from S1 and S2 extracts, respectively.

the presence of 250 mM imidazole, we were able to purify both the mutant and the GST-fused V1a receptors. As expected, the mutant was isolated as a major band around 42 kDa and a minor band at around 55 kDa, representing a degraded product and the entire receptor, respectively. The GST-fused V1a receptor was isolated as a unique band with a molecular mass around 62 kDa, probably a degraded product as well but containing the entire V1a receptor. We estimated the quantity of purified receptors to be very low, less than 0.5 mg/l of bacterial culture. For comparison, to date, only a few receptors have been produced in E. coli at a level high enough for the purification of milligram quantities (Kiefer et al., 1996, 2000; Grisshammer and Tucker, 1997).

Concluding remarks Determination of the structure of integral membrane proteins, such as GPCRs, requires a suitable overexpression system and an efficient purification procedure. The high level expression of these seven transmembrane domain proteins is still a major problem which has not yet been solved. In the present study, we have investigated the usefulness of a bacterial expression system for the production of the AVP/OT receptors. In order to increase the level of expression, we have first optimized the nucleotide sequence of V1a and OT receptors for bacterial expression, fused these sequences with different N-terminal tags useful for expression, and added a 6 × His tag at

175

the C-terminus end for their one-step purification using immobilized metal affinity chromatography. Expressing the wild-type V1a receptor in BL21(DE3) or BL21(DE3)pLysS bacterial strains was very toxic to the cell and totally prevented the visualization of the protein either by Coomassie blue-stained gels or by Western blotting. The mutant carrying Gly and Ala residues at positions 2 and 3, respectively, was expressed and detected. The effect of the mutations could be mimicked by the addition of glucose in the culture medium. In all cases, proteins that have been detected by Western blot and chemiluminescence procedures were proteolysed. Only the addition of the GST tag to the wild-type receptor allowed to produce a protein species containing the entire receptor. Purification of the receptors using the 6 × His tag/Ni-NTA interaction was very fast and efficient. The notion that lack of expression of the wildtype receptor may be due to improper insertion of the protein in the bacterial membrane was supported by finding that the GST-fused and the mutant V1a receptors were targeted to IBs and could be better expressed. The present results confirm that insertion of overexpressed eukaryotic membrane proteins into the bacterial membrane is toxic to the cells and thus significantly decreases the level of expression. Producing the receptors directly into IBs is consequently very promising. Interestingly, these insoluble particles protect the recombinant target protein from proteolytic degradation. Moreover, although the receptors are expressed as an aggregated and misfolded state into IBs, it has been demonstrated in several cases that refolding of membrane proteins from IBs can be achieved by detergent exchange and reconstitution into liposomes. Because biophysical studies require large quantities of purified receptors, the ease of large scale fermentation still makes E. coli very attractive. However, this expression system needs substantial improvement before routine high level expression of these transmembrane proteins. The results described in this chapter, are promising, especially those obtained with the GST fusion protein. In the future, we should optimize AVP/OT receptor expression using new fusion partners that facilitate targeting of the recombinant proteins to IBs and/or protection against proteolytic degradation.

Abbreviations 3D AVP cAMP GPCR GST IBs IPTG LB medium NTA OT PCR SDS

three dimensional arginine-vasopressin adenosine 3 :5 -cyclic monophosphate G protein-coupled receptor glutathione-S-transferase inclusion bodies isopropyl-β-D-thiogalactopyranoside Luria–Bertani medium nitrilotriacetic acid oxytocin polymerase chain reaction sodium dodecyl sulfate

Acknowledgements This work was supported by grants from INSERM, the program Molécules et Cibles Thérapeutiques from INSERM and CNRS, the Action Incitative Physique et Chimie du Vivant from the CNRS, and the fellowship program Poste Vert from INSERM (to T.S.). We are grateful to Drs. J. Parello and J.-L. Banères for helpful discussions. B.M. wishes to thank N. Delaunay for the introduction to Powerpoint software. References Bachmair, A., Finley, D. and Varshavsky, A. (1986) In vivo halflife of a protein is a function of its amino-terminal residue. Science, 234: 179–186. Barberis, C., Mouillac, B. and Durroux, T. (1998) Structural bases of vasopressin/oxytocin receptor function. J. Endocrinol., 156: 223–229. Barberis, C., Morin, D., Durroux, T., Mouillac, B., Guillon, G., Seyer, R., Hibert, M., Tribollet, E. and Manning, M. (1999) Molecular pharmacology of AVP and OT receptors and therapeutic potential. Drug News Persp., 12: 279–292. Bockaert, J. and Pin, J.-P. (1999) Molecular tinkering of G protein-coupled receptors: an evolutionary success. EMBO J., 18: 1723–1729. Breton, C., Chellil, H., Kabbaj-Benmansour, M., Carnazzi, E., Seyer, R., Phalipou, S., Morin, D., Durroux, T., Zingg, H.H., Barberis, C. and Mouillac, B. (2001) Direct identification of human oxytocin receptor-binding domains using a photoactivatable cyclic peptide antagonist: comparison with the human V1a vasopressin receptor. J. Biol. Chem., 276: 26931–26941. Chini, B., Mouillac, B., Ala, Y., Balestre, M.-N., Trumpp-Kallmeyer, S., Hoflack, J., Elands, J., Hibert, M., Manning, M., Jard, S. and Barberis, C. (1995) Tyr115 is the key residue for determining agonist selectivity in the V1a vasopressin receptor. EMBO J., 14: 2176–2182.

176 Cotte, N., Balestre, M.-N., Phalipou, S., Hibert, M., Manning, M., Barberis, C. and Mouillac, B. (1998) Identification of residues responsible for the selective binding of peptide antagonists and agonists in the V2 vasopressin receptor. J. Biol. Chem., 273: 29462–29468. Cotte, N., Balestre, M.-N., Aumelas, A., Mahé, E., Phalipou, S., Morin, D., Hibert, M., Manning, M., Durroux, T., Barberis, C. and Mouillac, B. (2000) Conserved aromatic residues in the transmembrane region VI of the V1a vasopressin receptor differentiate agonist vs. antagonist ligand binding. Eur. J. Biochem., 267: 4253–4263. Dillon, P.J. and Rosen, C.A. (1993) Use of polymerase chain reaction for the rapid construction of synthetic genes. In: B.A. White (Ed.), Methods in Molecular Biology, PCR protocols: current methods and applications. Vol. 15, Humana Press, Totowa, NJ, pp. 263–268. Fanelli, F., Barbier, P., Zanchetta, D., De Benedetti, P.G. and Chini, B. (1999) Activation mechanism of human oxytocin receptor: a combined study of experimental and computersimulated mutagenesis. Mol. Pharmacol., 56: 214–225. Freidinger, R.M. and Pettibone, D.J. (1997) Small molecule ligands for oxytocin and vasopressin receptors. Med. Res. Rev., 17: 1–16. Gimpl, G. and Fahrenholz, F. (2001) The oxytocin receptor system: structure, function and regulation. Physiol. Rev., 81: 629–683. Gribskov, M., Devereux, J. and Burgess, R.R. (1984) The codon preference plot: graphic analysis of protein coding sequences and prediction of gene expression. Nucleic Acids Res., 12: 539–549. Grigorieff, N., Ceska, T.A., Downing, K.H., Baldwin, J.M. and Henderson, R. (1996) Electron-crystallographic refinement of the structure of bacteriorhodopsin. J. Mol. Biol., 259: 393– 421. Grisshammer, R. and Tate, C.G. (1995) Overexpression of integral membrane proteins for structural studies. Q. Rev. Biophys., 28: 315–422. Grisshammer, R. and Tucker, J. (1997) Quantitative evaluation of neurotensin receptor purification by immobilized metal affinity chromatography. Protein Express. Purif., 11: 53–60. Grossman, T.H., Kawasaki, E.S., Punreddy, S.R. and Osburne, M.S. (1998) Spontaneous cAMP-dependent derepression of gene expression in stationary phase plays a role in recombinant expression instability. Gene, 209: 95–103. Hawtin, S.R., Wesley, V.J., Parslow, R.A., Patel, S. and Wheatley, M. (2000) Critical role of a subdomain of the N-terminus of the V1a vasopressin receptor for binding agonists but not antagonists: functional rescue by the oxytocin receptor Nterminus. Biochemistry, 39: 13524–13533. Hirel, P.-H., Schmitter, J.-M., Dessen, P., Fayat, G. and Blanquet, S. (1989) Extent of N-terminal methionine excision from Escherichia coli proteins is governed by the side-chain length of the penultimate amino acid. Proc. Natl. Acad. Sci. USA, 86: 8247–8251. Kane, J.F. (1995) Effects of rare codon clusters on high-level expression of heterologous proteins in Escherichia coli. Curr. Opin. Biotechnol., 6: 494–500.

Kiefer, H., Krieger, J., Olszewski, J.D., Von Heijne, G., Prestwich, G.D. and Breer, H. (1996) Expression of an olfactory receptor in Escherichia coli: purification, reconstitution and ligand binding. Biochemistry, 35: 16077–16084. Kiefer, H., Vogel, R. and Maier, K. (2000) Bacterial expression of G protein-coupled receptors: prediction of expression levels from sequence. Receptors Channels, 7: 109–119. Manning, M., Cheng, L.L., Stoev, S., Sawyer, W.H., Tribollet, E., Barberis, C., Wo, N.C. and Chan, W.Y. (1995) Novel potent and selective antagonists and radioiodinated ligands for oxytocin and vasopressin receptors. In: T. Saito, K. Kurokawa and S. Yoshida (Eds.), Neurohypophysis: Recent Progress of Vasopressin and Oxytocin Research. Elsevier Science, Amsterdam, pp. 21–38. Mendre, C., Dufour, M.-N., Le Roux, S., Seyer, R., Guillou, L., Calas, B. and Guillon, G. (1997) Synthetic rat V1a vasopressin receptor fragments interfere with vasopressin binding via specific interaction with the receptor. J. Biol. Chem., 272: 21027–21036. Mirzabekov, T., Bannert, N., Farzan, M., Hofmann, W., Kolchinsky, P., Wu, L., Wyatt, R. and Sodroski, J. (1999) Enhanced expression, native purification, and characterization of CCR5, a principal HIV-1 coreceptor. J. Biol. Chem., 274: 28745– 28750. Mouillac, B., Chini, B., Balestre, M.-N., Elands, J., TrumppKallmeyer, S., Hoflack, J., Hibert, M., Jard, S. and Barberis, C. (1995) The binding site of neuropeptide vasopressin V1a receptor: evidence for a major localization within transmembrane regions. J. Biol. Chem., 270: 25771–25777. Palczewski, K., Kumasaka, T., Hori, T., Behnke, C.A., Motoshima, H., Fox, B.A., Le Trong, I., Teller, D.C., Okada, T., Stenkamp, R.E., Yamamoto, M. and Miyano, M. (2000) Crystal structure of rhodopsin: a G protein-coupled receptor. Science, 289: 739–745. Pawate, S., Schey, K.L., Meier, P., Ullian, M.E., Mais, D.E. and Halushka, P.V. (1998) Expression, characterization and purification of C-terminally hexahistidine-tagged thromboxane A2 receptors. J. Biol. Chem., 273: 22753–22760. Phalipou, S., Seyer, R., Cotte, N., Breton, C., Barberis, C., Hibert, M. and Mouillac, B. (1999) Docking of linear peptide antagonists into the human V1a vasopressin receptor: identification of binding domains by photoaffinity labeling. J. Biol. Chem., 274: 23316–23327. Porath, J., Carlsson, J., Olsson, I. and Belfrage, G. (1975) Metal chelate affinity chromatography, a new approach to protein fractionation. Nature, 258: 598–599. Postina, R., Kojro, E. and Fahrenholz, F. (1996) Separate agonist and peptide antagonist binding sites of the oxytocin receptor defined by their transfer into the V2 vasopressin receptor. J. Biol. Chem., 271: 31593–31601. Serradeil-LeGal, C., Lacour, C., Valette, G., Garcia, G., Foulon, L., Galindo, G., Bankir, L., Pouzet, B., Guillon, G., Barberis, C., Chicot, D., Jard, S., Vilain, P., Garcia, C., Marty, E., Raufaste, D., Brossard, G., Nisato, D., Maffrand, J.-P. and Le Fur, G. (1996) Characterization of SR 121463A, a highly potent and selective, orally active vasopressin V2 receptor antagonist. J. Clin. Invest., 98: 2729–2738.

177 Strosberg, A.D. and Marullo, S. (1992) Functional expression of receptors in microorganisms. Trends Pharmacol. Sci., 13: 95–98. Thibonnier, M., Coles, P., Conarty, D.M., Plesnicher, C.L. and Shoham, M. (2000) A molecular model of agonist and nonpeptide antagonist binding to the human V1 vascular vasopressin receptor. J. Pharmacol. Exp. Ther., 294: 195–203. Thibonnier, M., Coles, P., Thibonnier, A. and Shoham, M. (2001) The basic and clinical pharmacology of nonpeptide vasopressin receptor antagonists. Annu. Rev. Pharmacol. Toxicol., 41: 175–202. Tobias, J.W., Shrader, T.E., Rocap, G. and Varshavsky, A. (1991)

The N-end rule in bacteria. Science, 254: 1374–1377. Tucker, J. and Grisshammer, R. (1996) Purification of a rat neurotensin receptor expressed in Escherichia coli. Biochem. J., 317: 891–899. Whitelegge, J.P., Gundersen, C.B. and Faull, K.F. (1998) Electrospray-ionization mass spectrometry of intact intrinsic membrane proteins. Protein Sci., 7: 1423–1430. Williams, D.C., Van Frank, R.M., Muth, W.L. and Burnett, J.P. (1982) Cytoplasmic inclusion bodies in Escherichia coli producing biosynthetic human insulin proteins. Science, 215: 687–689.

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Published by Elsevier Science B.V.

CHAPTER 14

Molecular pharmacology and modeling of vasopressin receptors M. Thibonnier 1,∗ , P. Coles 1 , A. Thibonnier 2 and M. Shoham 1 1

Departments of Medicine and Biochemistry, Case Western Reserve University School of Medicine and University Hospitals of Cleveland, Cleveland, OH, USA 2 University of Michigan School of Engineering, Ann Arbor, MI, USA

Abstract: AVP receptors represent a logical target for drug development. As a new class of therapeutic agents, orally active AVP analogs could be used to treat several human pathophysiological conditions including neurogenic diabetes insipidus, the syndrome of inappropriate secretion of AVP (SIADH), congestive heart failure, arterial hypertension, liver cirrhosis, nephrotic syndrome, dysmenorrhea, and ocular hypertension. By immunoprecipitation and immunoblotting, we elucidated the phosphorylation pattern of green fluorescent protein-tagged AVP receptors and showed interactions with the specific kinases PKC and GRK5 that are agonist-, time- and receptor subtype-dependent. The tyrosine residue of the NPWIY motif present in the 7th helix of AVP receptors is rapidly and transiently phosphorylated after agonist stimulation. This phosphorylation is instrumental in the genesis of the mitogenic cascade linked to the activation of this receptor, presumably by establishing key intramolecular contacts and by participating in the creation of a scaffold of proteins that produce the activation of downstream kinases. The random screening of chemical entities and optimization of lead compounds recently resulted in the development of orally active non-peptide AVP receptor agonists and antagonists. Furthermore, the identification of the molecular determinants of receptor–ligand interactions should facilitate the development of more potent and very selective orally active compounds via the approach of structure-based drug design. We developed three-dimensional molecular docking models of peptide and non-peptide ligands to the human V1 vascular, V2 renal and V3 pituitary AVP receptors. Docking of the peptide hormone AVP to the receptor ligand binding pockets reflects its dual polar and non-polar structure, but is receptor subtype-specific. The characteristics of non-peptide AVP analogs docking to the receptors are clearly distinct from those of peptide analogs docking. Molecular modeling of the results of site-directed mutagenesis experiments performed in CHO cells stably transfected with the human AVP receptor subtypes revealed that non-peptide antagonists establish key contacts with a few amino acid residues of the receptor subtypes that are different from those involved in agonist binding. Moreover, these interactions are species-specific. These findings provide further understanding of the signal transduction pathways of AVP receptors and new leads for elucidation of drug–receptor interactions and optimization of drug design. Note to the reader: The recent cloning and molecular characterization of AVP/OT receptor subtypes call for the revision of their nomenclature. For the sake of clarity and in reference to their main site of expression, we call the V1a receptor the V1 vascular receptor, the V2 receptor the V2 renal receptor and the V1b or V3 receptor the V3 pituitary receptor in the present review. Keywords: Vasopressin; Vasopressin receptor; Nonpeptide; Antagonists; 3-D modeling; Kinase; Receptor phosphorylation

∗ Correspondence to: M. Thibonnier, Room BRB431, Division of Clinical and Molecular Endocrinology, Case Western Reserve University School of Medicine, 10900 Euclid Avenue, Cleveland, OH 44106-4951, USA. E-mail: [email protected]

180

Introduction The neurohypophysial hormone arginine vasopressin (AVP) is actively involved in the regulation of free water reabsorption, body fluid osmolality, blood volume, blood pressure, cell contraction, cell proliferation, and ACTH secretion via the stimulation of specific G protein-coupled membrane receptors (GPCRs) classified into V1 vascular (V1 R), V2 renal (V2 R), and V3 pituitary (V3 R) subtypes having distinct pharmacological profiles and intracellular second messengers (Thibonnier et al., 1998b). Following agonist stimulation, GPCRs develop a reduction in responsiveness called desensitization that involves phosphorylation of the receptor by several types of kinases including tyrosine kinases, kinases activated by second messengers (PKA, PKC), and serine/threonine G protein-coupled receptor kinases (GRKs) (Lefkowitz, 1998). The mode of interaction of AVP receptors with intracellular kinases remains to be demonstrated. The examination of the amino acid composition of AVP receptors indicates the presence of several consensus motifs for PKC and GRK. Using immunoprecipitation and immunoblotting techniques, we studied the direct interaction between green fluorescent protein (GFP)tagged AVP receptors and PKC and GRKs. AVP receptors represent a logical target for drug development. However, the potential usefulness of AVP receptor antagonists in treating human diseases still remains an unanswered question because of the lack of currently FDA-approved orally active agents. However, orally active AVP receptor antagonists could be used in the treatment of various human diseases. Potential therapeutic applications may include: • The blockade of V1 vascular AVP receptors in arterial hypertension, congestive heart failure, and peripheral vascular disease. Moreover, the blockade of the V1 vascular AVP receptors present in the non-pregnant uterus may alleviate the symptoms of primary dysmenorrhea, a major cause of lost wages in the female population. • The blockade of V2 renal AVP receptors in the syndrome of inappropriate secretion of AVP, congestive heart failure, liver cirrhosis, nephrotic syndrome and any state of excessive retention of free water and subsequent hyponatremia.

•

The blockade of V3 pituitary AVP receptors in ACTH-secreting tumors.

Phosphorylation pattern of AVP receptors Agonist activation of GPCRs triggers a cascade of events including G protein coupling, phosphorylation of intracellular domains, internalization of the occupied receptor, and activation of intracellular messengers. Indeed, AVP activation of the V1 R elicits a rapid internalization and phosphorylation of the receptor (Innamorati et al., 1998). The family of AVP receptors is interesting in that its various members are coupled to different G proteins and exert opposite effects on cell growth and proliferation: the V1 R is coupled to Gq and produces a mitogenic response whereas the V2 R is coupled to Gs and produces an antimitogenic response (Thibonnier et al., 1998a). We recently showed that in CHO cells transfected with the wild-type V1 R, AVP induced a progression through both the S and G2 –M phases of the cell cycle (Thibonnier et al., 2000b) whereas no progression through the cell cycle was observed for the wild-type V2 R stimulation. When compared to the wild-type V1 R, AVP no longer produced progression through the cell cycle for a V1 R lacking its C-terminus (V1 R359X truncated receptor). AVP produced a normal progression through the cell cycle in mutants displaying the proximal portion of the V1 R C-terminus (V1 R399X, V1 R406X, and V1 R409X truncated receptors). AVP stimulation of the V1 RS382G-R384G mutant receptor (inactivation of the GRK-specific motif) produced a progression from the G0 –G1 phase to the S phase but no further progression to the G2 –M phase. Study of the cell cycle progression in the chimeric V1 R/V2 R and V2 R/V1 R clones revealed patterns similar to those of the corresponding wild-type receptors, i.e., cell cycle progression for the V1 R/V2 R clone, but no cell cycle progression for the V2 R/V1 R chimera. These data suggest that, like for IP production and DNA synthesis, the C-terminus of the V1 R participates in the effect of AVP on cell cycle progression, but cannot elicit such response by itself. Serine/threonine phosphorylation We explored the structural elements of the human V1 R mediating the mitogenic properties of this recep-

181

Fig. 1. Structure of the C-termini of the wild-type human V1 vascular, V2 renal and V3 pituitary AVP receptors. The PKC motifs [(S/T)X(R/K)] are in enlarged italic characters and the diacidic motifs are enlarged and underlined.

tor (Berrada et al., 2000; Thibonnier et al., 2000b,c). Phosphorylation of intracytoplasmic residues of GPCRs plays an obligatory role in signal transmission (Pitcher et al., 1998). The examination of the amino acid composition of the human V1 R, V2 R and V3 R reveals the presence of several serine and threonine potential phosphorylation sites at the level of the intracytoplasmic loops of all receptors. In contrast, the V1 R and V3 R C-termini contain one proximal G protein-coupled receptor kinase (GRK) consensus motif (diacidic motif upstream of serine residues) and two to three distal protein kinase C (PKC) consensus motifs ([S/T]X[R/K]) whereas the human V2 R C-terminus contains one GRK consensus motif but no PKC motif (Fig. 1). These observations suggest that the C-terminus of AVP receptors plays a determinant role in receptor phosphorylation and signal transduction. We studied the physical association of AVP receptors with GRK and PKC in wild-types and mutated forms of the receptors fused to green fluorescent protein (Berrada et al., 2000). After agonist stimulation, PKC dissociates from the V1 R, associates with the V3 R, but does not interact with the V2 R. Furthermore, AVP activation of the receptors leads to a brief association of all the receptor subtypes with GRK5 following a time-course that varies with the receptor subtype. Exchanging the V1 R and the V2 R C-termini alters the time course of PKC and GRK5 association. Mutation of the GRK site present in the proximal portion of the V1 R Cterminus dramatically reduces the extent of receptor phosphorylation. Thus, agonist stimulation of AVP receptor subtypes triggers receptor subtype-specific interactions with GRK and PKC through specific serine-containing motifs present in the C-termini of the receptors. Our direct assessment on the physi-

cal association between the human V1 R and kinases indicates that PKC phosphorylates the unoccupied receptor whereas GRK5 is responsible for agonistinduced phosphorylation of the receptor within seconds. These findings suggest that a dynamic interaction exists between the V1 R, PKC and GRK5 in an agonist-dependent fashion. Upon agonist activation, PKC is released from the receptor C-terminus, allowing GRK5 to associate with the proximal portion of this C-terminus. The phosphorylation pattern of the V2 R is clearly different from that of the V1 R. Phosphorylation of the V2 R occurs at a slower rate, lasts longer, and involves GRK5 but not PKC. Tyrosine phosphorylation The AVP receptors share among them a NPWIY motif present in their seventh helix. An NPX2–3 Y motif is present in virtually all members of the GPCR family and is positioned to be part of the conformational changes induced by agonist binding. For instance, the NPVIY motif of the rhodopsin receptor may form a specific interaction with methionine257 of the 6th helix that stabilizes the inactive receptor conformation (Han et al., 1998). Asparagine302 and tyrosine306 of the bovine rhodopsin NPVIY motif project inside the molecule and the OH group of tyrosine306 is close to asparagine73 of the 2nd helix, facilitating hydrogen-bonding constraints between helices II and VII (Palczewski et al., 2000). The tyrosine residue present within this NPX2–3 Y motif is instrumental in several events including receptor internalization and signal transduction (Barak et al., 1994, 1995; Laporte et al., 1996). We investigated the role played by the tyrosine residue of the NPWIY motif present in the sev-

182

enth transmembrane domain of the human V1 R in ligand binding characteristics, internalization kinetics, phosphorylation pattern, and signal transduction. As we reported previously, addition of a GFP tag at the carboxy terminus of the receptor did not interfere with its membrane insertion and ligand characteristics (Berrada et al., 2000). Mutation of the tyrosine348 residue of the NPWIY motif present in the 7th helix of the V1 R did not hamper the proper insertion of the receptor inside the cell membrane and did not alter the affinity for the native ligand AVP, as K d values were similar to that of the wildtype receptor. Moreover, competition binding experiments with AVP, the non-peptide V1 R antagonist SR 49059, and the non-peptide V2 R antagonist SR 121463A indicated that the mutation of tyrosine348 did not alter the ligand subtype selectivity of the V1 R. V1 Rs internalize once they are occupied by AVP (Fishman et al., 1985; Thibonnier, 1992). As the conserved NPXn Y motif of GPCRs does not seem to be a general sequestration sequence, we explored the role of the V1 R NPWIY motif in the receptor internalization in stably transfected CHO cells expressing the V1 R Y348A mutated clone. The wildtype V1 R was quickly internalized at 37°C with an internalization half-life of 5 min and a maximal level of internalization of 89%. The extent and rate of internalization of the V1 R-Y348A mutant were similar, with an internalization half-life of 6 min and a maximal level of internalization of 91%. Thus the NPWIY motif of the V1 R does not play a role in the receptor internalization. The cellular localization of the fluorescent wild-type and mutated V1 Rs was also studied by epifluorescent microscopy before and after AVP stimulation. As reported previously, non-transfected CHO cells did not produce any fluorescence (Berrada et al., 2000). Before AVP stimulation, both the wild-type and Y348A mutant V1 Rs displayed a diffuse and homogenous fluorescent pattern at the cell surface (Fig. 2). We have previously shown that within minutes after agonist stimulation of CHO cells transfected with the wild-type AVP or OT receptors, the pattern of fluorescence changed and became granular, reflecting the internalization and aggregation of the receptors inside the cytosolic compartment (Berrada et al., 2000). A similar effect of agonist binding on the cellular distribution of the

fluorescent receptors was noted for the V1 R-Y348A mutant, thus confirming the lack of involvement of this motif in the membrane distribution and internalization process of the receptor (Fig. 3). Binding of the non-peptide V1 R antagonist SR49059 to the Y348A mutant receptor did not modify the diffuse fluorescent pattern of the receptor, thus confirming that the antagonist-occupied receptor is not internalized (Fig. 3). The V1 Rs is rapidly phosphorylated after agonist stimulation, presumably at the level of serine/ threonine residues (Innamorati et al., 1998). In fact, we mentioned above that the mutation of a GRK motif (SRR sequence downstream of a ED diacidic motif) present in the C-terminus of the V1 R prevented the association with GRK5 and reduced the extent of receptor phosphorylation (Berrada et al., 2000). However, tyrosine phosphorylation of the V1 R has not been demonstrated so far. Thus, we used immunoprecipitation and immunoblotting techniques to look at the possibility of agonistinduced tyrosine phosphorylation of the wild-type and Y348A mutant V1 Rs with a C-terminus GFP tag. We have previously shown that the addition of a GFP tag to the C-terminus of AVP/OT receptors allows their efficient and specific immunoprecipitation and immunoblotting without interfering with signal coupling (Berrada et al., 2000). On both direct immunoblotting and immunoprecipitation/immunoblotting autoradiograms, the glycosylated full length wild-type and Y348A mutated V1 R-GFP complexes both presented as a broad band of molecular mass of 105–110 kDa (Fig. 4). Minor bands of molecular mass of 65–70 kDa were detected and may represent unglycosylated or degraded forms of the receptor (Phalipou et al., 1997a). Indeed, omission of the protease inhibitors in the lysis buffer increased the intensity of these smaller bands (data not shown). Immunoprecipitation of the V1 R with an antiGFP monoclonal antibody followed by immunoblotting with an anti-phosphotyrosine antibody revealed that the wild-type V1 R underwent tyrosine phosphorylation after agonist stimulation (Fig. 5A). Tyrosine phosphorylation was rapid, occurring within 30 s after agonist stimulation, and transient, peaking within 5 min after agonist stimulation. When the same blots were reprobed with an anti-GFP polyclonal antibody, the V1 R-GFP was detected in equal amounts in all

183

Fig. 2. Cellular distribution of fluorescent wild-type and NPWIY348A mutated V1 Rs. CHO cells expressing the wild-type and NPWIY348A mutant V1 Rs fused to GFP at their C-terminus (and control non-transfected cells) were studied by fluorescent microscopy. Left panels represent direct microscopic appearance of cells, while right panels represent fluorescent patterns of the same cells.

lanes (data not shown). Immunoprecipitation and immunoblotting of extracts from cells expressing the GFP protein alone revealed no signal, thus ruling out non-specific tyrosine phosphorylation of the GFP tag (Fig. 5B). Furthermore, sequential immunoprecipitation, first with the anti-phosphotyrosine antibody, then with the anti-GFP monoclonal antibody, followed by immunoblotting with the anti-GFP poly-

clonal antibody revealed the same 105–110-kDa band (Fig. 5C), thus confirming that indeed the wildtype V1 R is tyrosine-phosphorylated in response to agonist stimulation. Conversely, immunoprecipitation with the antiphosphotyrosine antibody followed by immunoblotting with the anti-GFP antibody isolated the same band (Fig. 6). The extent of receptor phosphoryla-

184

Fig. 3. Effects of agonist and antagonist binding on the cellular distribution of the fluorescent NPWIY348A mutated V1 R. CHO cells expressing the NPWIY348A mutant V1 R fused to GFP at its C-terminus were studied by fluorescent microscopy. The distribution of the fluorescent receptors was assessed before and after agonist (1 μM AVP) or after V1 R nonpeptide antagonist (1 μM SR490590) binding for 15 min at 37°C.

tion was dramatically reduced in the V1 R-Y348A mutant, thus indicating that the tyrosine residue of the NPWIY motif of the 7th helix is the target of a tyrosine kinase (Fig. 6). As the C-terminus of the human V1 R contains also a tyrosine residue in position 388 that could be the target of tyrosine kinases after agonist stimulation, immunoblotting with an anti-phosphotyrosine antibody was performed with a truncated form of the human V1 R whose C-terminus has been deleted at position 359. This truncated V1 R underwent agonist-dependent tyrosine phosphorylation to the same degree as the wild-type receptor (data not shown), thus ruling out the involvement of the receptor C-terminus in tyrosine phosphoryla-

Fig. 4. Immunoprecipitation and immunoblotting of the GFPtagged wild-type and Y348A mutated V1 Rs. CHO cells expressing GFP-tagged wild-type and Y348A mutant V1 Rs were lyzed, immunoprecipitated (IP) with an anti-GFP monoclonal antibody and immunoblotted (IB) with an anti-GFP polyclonal antibody. Total cell lysate (TCL) samples were directly immunoblotted. A peroxidase-labeled secondary antibody was used for detection of the chemiluminescent signal.

Fig. 5. AVP-induced tyrosine phosphorylation of the GFP-tagged wild-type V1 R. Transfected cells were grown to subconfluence in 100-mm dishes then serum-starved and stimulated by AVP (1 μM at 37°C up to 10 min). The cells were lyzed and cell extracts were immunoprecipitated (IP) with an anti-GFP monoclonal antibody or an anti-phosphotyrosine monoclonal antibody. Immunoprecipitates were resolved by SDS–PAGE followed by immunoblotting (IB) with an anti-phosphotyrosine monoclonal antibody or an anti-GFP polyclonal antibody. A peroxidaselabeled secondary antibody was used for detection of the chemiluminescent signal. (A) IP and IB of CHO cells expressing GFP-tagged wild-type V1 Rs. (B) CHO cells expressing the GFP tag alone. (C) Double IP of CHO cells expressing GFP-tagged wild-type V1 Rs.

185

186

Fig. 6. Reduction of AVP-induced tyrosine phosphorylation of the GFP-tagged Y348A V1 R. CHO cells expressing GFP-tagged wild-type and Y348A mutant V1 Rs were grown to subconfluence, then serum-starved and stimulated by AVP (1 μM at 37°C up to 5 min). The cells were lyzed and cell extracts were immunoprecipitated (IP) with an anti-phosphotyrosine monoclonal antibody. Immunoprecipitates were resolved by SDS–PAGE followed by immunoblotting (IB) with an anti-GFP polyclonal antibody. A peroxidase-labeled secondary antibody was used for detection of the chemiluminescent signal.

tion. In CHO cells transfected with the pEGFP-N3 vector alone, phosphotyrosine proteins did not coimmunoprecipitate with an anti-GFP antibody (data not shown). We recently reported that AVP stimulation of the V1 R produced the simultaneous activation of several kinases, including PI 3-kinase, p42/p44 MAP kinase and calcium/calmodulin kinase II that led to a mitogenic response (Thibonnier et al., 2000b). We tested the influence of the V1 R-Y348A mutation on the activation of the kinases linked to V1 R activa-

tion. As shown in Fig. 7, the extent and duration of activation of p42/p44 MAP kinase and p70 kinase (downstream of the PI 3-kinase) were dramatically reduced in cells transfected with the mutated V1 R. The activation of calcium/calmodulin kinase II by the V1 R-Y348A mutant was not altered (data not shown). Finally, assessment of tyrosine phosphorylation of total cell lysates after AVP stimulation revealed a major reduction of the level of protein phosphorylation in the presence of the Y348A mutation, thus indicating that Tyr348 is instrumental in linking the V1 R to tyrosine kinases after AVP stimulation (Fig. 8). We investigated the role of the NPWIY motif phosphorylation in the signal transduction cascade of the V1 R. V1 Rs are coupled to phospholipase C with a subsequent increased production of inositol phosphates (IP). In CHO cells transfected with the wildtype V1 R, AVP produced a dose-dependent increase in the formation of IP (Fig. 9) that was blocked by the non-peptide V1 R antagonist SR 49059, but not by the V2 R antagonist SR 121643A (data not shown). AVP could still stimulate IP production in cells expressing the Y348A mutated V1 R, but the maximum response was only 45% of that of the wild-type V1 R ( p < 0.01). These data suggest that the NPWIY motif of the V1 R participates in the stimulation of IP production. One major cellular event resulting from the stimulation of V1 Rs is cell growth and proliferation, which can be assessed by measuring nucleic acid synthesis through [3 H]thymidine uptake. Thus, AVP-induced [3 H]thymidine uptake was measured in CHO cells transfected with the wild-type or the mutated form of the human V1 R. We have previously shown that AVP induced a dose-dependent increase in [3 H]thymidine uptake in CHO cells transfected with the wild-type V1 R (Thibonnier et al., 1998b). AVP stimulation of DNA synthesis was dramatically reduced in the Y348A mutant, whereas this mutated clone was normally responsive to stimulation by 10% FBS (Fig. 10). These data suggest that the NPWIY motif of the V1 R participates in the nucleic acid synthesis step of the mitogenic response. We have previously shown that AVP induced a dose-dependent increase of cell proliferation in CHO cells transfected with the wild-type V1 R (Thibonnier et al., 1998b). As shown in Fig. 11, AVP stimulation of cell proliferation was dramatically reduced in the

187

Wild Type V1 R

V1 R-Y348A

p42/p44 p70 βActin 0 ' 10' 20' 30' 40' 60' 90' 90' AVP stimulation (min.) 10% FBS

10% FBS

0 ' 10' 20' 30' 40' 60' 90' 90' AVP stimulation (min.)

Fig. 7. Effect of the Y348A mutation on AVP-induced kinase phosphorylation. Transfected CHO cells were grown to sub-confluence, serum-starved, then stimulated by 1 μM AVP for 0–90 min or by 10% FBS for 90 min. Cells were lysed, protein was resolved by gel electrophoresis followed by immunoblotting with p42/p44 and p70 phosphospecific antibodies. The same blots were stripped and reprobed with an anti-actin antibody to verify homogeneity of protein transfer.

Fig. 8. Effect of the Y348A mutation on AVP-induced protein phosphorylation. Transfected CHO cells were grown to sub-confluence, serum-starved, then stimulated by 1 μM AVP for 0–30 min. Cells were lyzed (TCL, total cell lysate), protein were resolved by gel electrophoresis followed by immunoblotting (IB) with an anti-phosphotyrosine monoclonal antibody. The same blots were stripped and reprobed with an anti-actin antibody to verify homogeneity of protein transfer.

Y348A mutant, thus confirming that the NPWIY motif of the V1 R is required to elicit a mitogenic response.

In conclusion, as expected for a typical GPCR, the V1 R becomes phosphorylated after agonist stimulation. We and others have shown that agonist stim-

188

IP1 Release (c.p.m.)

5000

WT-V1R V1R-Y348A

4000 3000 2000 1000 0

-12

10

-11

10

-10

10

-9

10

-8

10

-7

10

AVP Concentration (M) Fig. 9. AVP-induced inositol phosphate production by wild-type and mutated AVP receptors. CHO cells expressing the wild-type and Y348A mutant V1 Rs were grown to confluence in 12-well dishes and incubated in inositol-free DMEM buffer with myo[23 H]inositol. Formation of [3 H]inositol phosphates was measured after addition of increasing concentrations of AVP for 30 min at 37°C (n = 6 for each data point).

Fig. 10. AVP-induced [3 H]thymidine uptake in CHO cells transfected with the wild-type or mutated receptor cDNA clones. CHO cells expressing the wild-type and Y348A mutant V1 Rs were grown to sub-confluence in 24-well dishes and incubated in serum-free F12 medium for 72 h. The cells were subsequently stimulated overnight by increasing concentrations of AVP (or 10% FBS as a positive control), followed by incubation with [3 H]thymidine for 45 min, DNA precipitation and liquid scintillation counting (n = 8 for each data point).

ulation of the AVP/OT receptors led to receptor subtype-specific interactions with GRK and PKC through serine residues of specific motifs present in the C-termini of the receptors (Innamorati et al., 1998; Berrada et al., 2000). In addition to the GRK-dependent serine phosphorylation, we directly

Fig. 11. AVP-induced cell proliferation in CHO cells transfected with the wild-type or mutated receptor cDNA clones. CHO cells expressing the wild-type and Y348A mutant V1 Rs were grown to sub-confluence in 96-well dishes and incubated in serum-free F12 medium for 72 h. The cells were stimulated by AVP for 24 h, followed by addition of the dye MTS for 2 h. Absorbance was recorded at 490 nm (n = 12 for each data point).

demonstrated for the first time that the V1 R is also phosphorylated at the level of a conserved tyrosine residue present in its 7th helix. Similarly, mutation of Tyr326 of the NPLIY motif of the β2 -adrenergic receptor to an alanine residue resulted in a dramatic reduction of the receptor phosphorylation in response to agonist stimulation (Barak et al., 1995). The phosphorylation of GPCR serine/threonine residues by GRKs has been shown to trigger the binding of arrestins with subsequent interaction with clathrin, internalization and desensitization of the receptors (Laporte et al., 1999). The results presented here demonstrate that tyrosine phosphorylation of the NPWIY motif of the V1 R serves other purposes as no role in ligand binding and receptor internalization was found. As a matter of fact, prevention of Tyr348 phosphorylation in response to AVP stimulation severely hampers the mitogenic signaling cascade linked to activation of the V1 R. Similarly, alanine mutations of the residues of the NPLFY motif of the angiotensin II type 1 receptor were shown to impair the agonistinduced stimulation of IP formation (Hunyady et al., 1995). These findings suggest that agoniststimulation of the V1 R triggers at least two distinct pathways: one is governed by the GRK-dependent phosphorylation of serine/threonine residues present in the cytoplasmic C-terminus of the receptor that leads to the receptor internalization and subsequent desensitization. The other involves the tyrosine phos-

189

phorylation of the NPWIY motif of the 7th helix that plays a crucial role in the transmission of the mitogenic signal. Protein tyrosine phosphorylation regulates protein–protein interaction and enzyme activation involved in mitogenic signal transduction following activation of cytokine and growth factor receptors. Furthermore, phosphotyrosine binding (PTB) domains of Shc and insulin receptor substrate 1 preferentially bind to tyrosine-phosphorylated NPXY-containing insulin and interleukin-4 receptors. The member of the insulin receptor family leukocyte tyrosine kinase contains two NPXY motifs that equally contribute to activation of the Ras pathway and generation of mitogenic signals. Mitogenic signaling by GPCRs involves phosphorylation reactions and assembly of protein complexes. At least three types of scaffolds of GPCR– protein complexes have been identified (Luttrell et al., 1999). They include transactivated receptor tyrosine kinases (RTKs), integrin-based focal adhesions, and GPCRs themselves. Indeed, GPCRs can serve themselves as signaling platforms after agonist stimulation, recruiting GRKs, subunits of G proteins, arrestins, and other kinases (e.g. c-Src). Recruitment of β-arrestin and activation of c-Src initiates the assembly of a protein scaffolding that leads to MAPK recruitment and mitogenic signaling. The assembly of such scaffolding leads to the mitogenic response. The determination of the crystal structure of rhodopsin at a 2.8 Å resolution has identified a set of residues that mediate interaction between the helices and the cytoplasmic surface, where G protein activation occurs (Palczewski et al., 2000). Among them, the OH group of the tyrosine residue of the NPXXY motif present in the 7th helix is close to a conserved asparagine residue present in the 2nd helix, suggesting the presence of additional interhelical constraints between helices 2 and 7. We hypothesize here that the phosphorylation of the Tyr348 residue of the V1 R upon agonist stimulation produces a conformational change that activates the mitogenic cascade. In a model similar to that described for growth factor receptors, agonist activation of the V1 R triggers tyrosine phosphorylation of the NPWIY motif within the 7th helix that allows binding of proteins containing phosphotyrosine binding domains and further assembly of a scaffold of the various components of the mitogenic cascade.

Three-dimensional molecular docking of peptide AVP analogs to AVP receptors Cloning of human and animal AVP/OT receptor subtypes and their stable expression in immortalized cell lines has allowed several investigators to begin defining the molecular determinants of AVP receptor subtypes peptide ligand selectivity (Chini et al., 1995; Mouillac et al., 1995; Ufer et al., 1995; Postina et al., 1996; Phalipou et al., 1997b; Cotte et al., 1998). As shown in Fig. 12a, we built a 3-D model of the human V1 R and successfully docked AVP within the structure of this receptor (Thibonnier et al., 2000a). The 3-D model of AVP was docked onto V1 R by initially placing it in the upper portion of the transmembrane region (the expected binding pocket), and searching for the binding site with the program LIGIN within a 20×20×20 Å box around the original ligand position. In the docking of AVP, some steric overlap (1–3 residues) was allowed between the ligand and receptor. Energy minimization with program XPLOR relieved these short contacts. AVP has a polar as well as a non-polar surface. The exocyclic tripeptide Pro7 –Arg8 –Gly9 and one side of the hormone ring (Gln4 , Asn5 ) are mainly hydrophilic, whereas the other part of the ring (Cys1 , Cys6 , Tyr2 , and Phe3 ) is essentially hydrophobic in nature. This dual surface property is reflected in the nature of the binding pocket that is formed by residues from transmembrane segments (TMSs) 1, 3, 4, 5, 6, and 7, as well as the first extracellular loop (Fig. 13). The bottom of the cleft is mainly hydrophobic, closed by the aromatic and hydrophobic residues Met135, Phe136 , Phe179 , Phe307 and Ile330. The entrance to the binding pocket and one side of it contain predominantly hydrophilic residues. The Arg8 guanido group at the entrance to the cleft forms a salt bridge with Asp112 located on the first extracellular loop. Trp111 forms van der Waals contacts with the hydrophobic part of Arg8 . The ε-amino group of Lys128 forms a hydrogen bond to the amide side chain nitrogen of Asn5 . Other hydrogen bonds are formed between the side chain moieties of Gln185 and Ser182 with Gln4 , and Ser213 Oγ with Tyr2 OH. Another wall of the pocket is lined with the hydrophobic residues Ile55 and Ile330 . Some amino acid residues that are common to

190

Fig. 12. Docking of AVP to the human V1 vascular (a) and V3 pituitary (b) AVP receptors. The loops are labeled il1, il2, and il3 for the intracellular loops and el1, el2 and el3 for the extracellular loops. The transmembrane segments are labeled H1–H7. The different orientation of AVP binding to the receptor subtypes is clearly shown.

all AVP/OT receptor subtypes are important for peptide agonist binding. They are D207 , Q214 , Q218 , K308 , Q311 , Q413 , and Q620 . None of these residues is involved in peptide and non-peptide antagonist binding (Mouillac et al., 1995). The presence of a disulfide bond between two conserved cysteine residues present in exoloops 1 and 2 is required to maintain the integrity of the receptor structure. Studies performed recently with the V1 vascular, V2

renal AVP and the OT receptors from several species revealed that a few key residues determine peptide ligand selectivity for a given receptor subtype. For instance, residue Tyr115 located in the first extracellular loop is crucial for high affinity binding of peptide agonists and confers V1 vascular receptor subtype specificity (Chini et al., 1995). The use of natural small synthetic peptides mimicking segments of the V1 R revealed that the N-terminal part of the V1 R is

191

Fig. 12 (continued).

not involved in peptide agonist binding (Mendre et al., 1997). At variance, natural peptides mimicking the external loops of the V1 R, especially one peptide mimicking the 205–218 portion of the 2nd extracellular loop, were able to inhibit specific AVP binding to the V1 R. Site-directed mutagenesis experiments of the cloned bovine and porcine V2 Rs revealed that Asp103 in the first extracellular loop is responsible for high affinity binding of the V2 R peptide agonist dDAVP (Ufer et al., 1995). Similarly, residues responsible for selective binding of peptide agonists and antagonists to the V2 renal receptor were identified (Cotte et al., 1998). Residues 202 (Arg vs. Leu) in the second extracellular loop and 304 (Gly vs. Arg) in the 7th transmembrane domain are responsible for species-selective cyclic peptide antagonists binding in an independent and additive manner. Residue 100

(Lys vs. Asp) in the second transmembrane domain plays a similar role for peptide agonist discrimination. For peptide agonist binding and selectivity for the OT receptor subtype, the first three extracellular domains are most important (Postina et al., 1996). The N-terminal domain and the first extracellular loop of the OT receptor interact with the linear Cterminal tripeptidic part of the ligand OT, whereas the second extracellular loop of the OT receptor interacts with the cyclic part of OT. The molecular determinants of peptide antagonist binding to the OT receptor are different, i.e. the transmembrane helices 1, 2, and especially 7. Introduction of just seven amino acids of the upper part of the 7th TMS of the OT receptor into the V2 R sequence is sufficient to introduce high affinity binding for an OT peptide antagonist into the V2 R.

192

Fig. 13. Docking of the non-peptide antagonist OPC-21268 onto the model of the human V1 vascular AVP receptor. (a) Top view. (b) Stabilizing effect of the G337A, I224V, and I310V mutations on antagonist binding.

We created a three-dimensional molecular model of the human V3 R. As basis for our model building of the V3 R receptor we used a model of the

seven TMSs of V3 R generated by G. Vriend with the program WHATIF (Vriend, 1990) based upon the crystal structure of bacteriorhodopsin (Rodriguez

193

et al., 1998). The three extracellular and three intracellular loops of the V3 R were subsequently constructed with program Look v3.5 (Molecular Applications Group, Palo Alto, CA 94304), using the spatial constraints for the ends of each loop provided by the coordinates of the helical bundle. Look v3.5 is a protein-modeling program that segmentally builds a protein by aligning short stretches of its sequence with homologous peptides of known structure, and also performs a full energy refinement of the model. As the N-terminal and C-terminal domains of the V3 R are not involved in the binding of agonists or antagonists, they were not included in this model. A disulfide bridge exists between cysteines 107 and 186 located on the second and third extracellular loops, respectively. Disruption of this disulfide bridge is known to cause a significant drop in binding affinity of ligands. Thus, it was necessary to ascertain that these cysteine residues were close enough in the model and that the sulfhydryl groups had the proper orientation in order to be able to form the disulfide bridge. This was achieved by performing an energy refinement in program X-PLOR with the constraint of forming this particular disulfide bridge. The sulfur–sulfur distance refined to a value of 2.01 Å, consistent with the formation of a disulfide bridge. The rest of the structure was not significantly altered by this refinement procedure. The 3-D appearance of the human V3 R was compared to that of the human V1 R (Fig. 12). There is a tight overlap at the level of all TMSs with four TMSs (1st, 5th, 6th, and 7th) perpendicular to the membrane plan and three TMSs (2nd, 3rd, and 4th) slightly tilted. On the other hand, the respective extra- and intracellular loops are of different length and seem to have different conformations in these two receptors. The maximum deviation and rms values between the models we have built for the two receptor subtypes are excellent, only 3.8 and 1.4 Å for all common α carbon atoms, respectively. We docked the 3-D model of AVP that we developed previously to the human V3 R with the program LIGIN (Fig. 12b). Energy minimization with program X-PLOR relieved the short contacts. As shown in Fig. 12, it appears that the docking pattern of AVP to the human V3 R is strikingly different from its docking pattern to the human V1 R: instead of being docked in a slanted fashion like in the V1 R, AVP docked inside the

ligand binding pocket of the V3 R in a vertical fashion. The AVP hydrophilic tail is in close contact with the upper portion of the ligand binding pocket, especially the 3rd extracellular loop whereas the hydrophobic ring is buried deep inside the TMSs where it makes close contact with several residues of the 6th TMS. This unique docking pattern of AVP to the V3 R that is distinct from its docking to the V1 R but also to the V2 R, is a major finding that readily explains why ligands that are specific for the V1 R and the V2 R have a poor affinity for the V3 R and that V3 R-selective ligands have yet to be discovered. Thus, these studies suggest that the molecular determinants of peptide agonists and antagonists binding to AVP/OT receptors are distinct. Three-dimensional molecular docking of non-peptide antagonists to AVP receptors Examination of the 3-D structure of the non-peptide AVP receptor antagonists indicates that this is a rather heterogeneous class of compounds (Thibonnier et al., 2001). As there was no knowledge of the molecular determinants of AVP receptors involved in nonpeptide antagonist binding, we studied this issue by site-directed mutagenesis and molecular modeling techniques. The first non-peptide AVP V1 R antagonist found by random screening and optimization of chemical entities, OPC-21268 has an excellent affinity for the rat V1 R (25 nM), but has a poor affinity for the human V1 R (8800 nM) (Thibonnier et al., 1998b). The human and rat V1 Rs share a high degree of structural homology with 96% sequence identity. The differing residues are presumably involved in species-related variations in antagonist binding. Comparison of the human and rat V1 R sequences revealed that only 20 amino acid differences are present in the extracellular loops and the upper portions of the transmembrane segments. We reasoned that these interspecies differences in amino acid sequence modulate the receptor affinity for non-peptide compounds. Thus, we produced a series of reverse mutations in which corresponding rat amino acids were introduced by site-directed mutagenesis into the human V1 R sequence (Thibonnier et al., 2000a). The influence of these interspecies amino acid differences on nonpeptide antagonist binding was subsequently tested. The introduction of rat amino acids in positions 224,

194

310, 324, or 337 of the human V1 R sequence dramatically altered OPC-21268 affinity for the receptor whereas binding of AVP, the peptide V1 R antagonist d(CH2 )5 Tyr(Me)AVP, and the non-peptide V1 R antagonist SR49059 was not altered by these mutations. In order to gain information about the location of the OPC-21268 binding site, a model of this compound was docked onto a homology-built threedimensional model of the human V1 R (Thibonnier et al., 2001). Very little direct structural information is available for GPCRs, and for many years molecular models of these receptors have been built based upon the crystal structure of bacteriorhodopsin. Although bacteriorhodopsin consists of the seven transmembrane helices by which GPCRs are characterized, it shares very little sequence homology with any of the GPCRs. Still, the use of bacteriorhodopsin to establish the orientation of the transmembrane domains of AVP receptors is the only way to build a model based on an experimentally determined highresolution structure (Henderson et al., 1990). Coordinates of bovine rhodopsin are also available, but only for the seven TMSs without any loops. As basis for our model, building of the V1 R receptor we used a model of the seven TMS of V1 R generated by G. Vriend with the program WHATIF (Vriend, 1990) based upon the crystal structure of bacteriorhodopsin (G Protein-Coupled Receptor Data Base at http:// swift.emblheidelberg.de/7tm/htmls/consortium.html) (Rodriguez et al., 1998). The three extracellular and three intracellular loops of the V1 R were subsequently constructed with program Look v3.5 (Molecular Applications Group, Palo Alto, CA 94304), using the spatial constraints for the ends of each loop provided by the coordinates of the helical bundle. Look v3.5 is a protein-modeling program that segmentally builds a protein by aligning short stretches of its sequence with homologous peptides of known structure, and also performs a full energy refinement of the model (Levitt, 1992). As the N-terminal and C-terminal domains of the V1 R are not involved in the binding of agonists or antagonists, they were not included in this model. A disulfide bridge exists between cysteines 124 and 203 located on the second and third extracellular loops, respectively. Disruption of this disulfide bridge is known to cause a significant drop in binding

affinity of ligand. Thus, it was necessary to ascertain that these cysteine residues were close enough in the model and that the sulfhydryl groups had the proper orientation in order to be able to form the disulfide bridge. This was achieved by performing an energy refinement in program X-Plor with the constraint of forming this particular disulfide bridge. The sulfur– sulfur distance refined to a value of 2.03 Å, consistent with the formation of a disulfide bridge. The rest of the structure was not significantly altered by this refinement procedure. Models of the non-peptide AVP receptor antagonists were constructed with the program Alchemy 2000 (Tripos Inc., St. Louis, MO). First, the compounds are drawn in two dimensions, and then extended into a three-dimensional model by a 2D-to3D builder incorporated in Alchemy 2000. Conformations with the lowest energy and devoid of any short contacts were saved. Finally, the most stable conformations are subjected to an optimization using the program Gaussian 98 (Gaussian, Inc., Pittsburgh PA, 1998). Docking of the non-peptide ligand OPC-21268 to the receptors was done with the program LIGIN based on a built-in complementarity function (Sobolev et al., 1996). This function is a sum of the surface areas of atomic contacts. These contacts are weighted according to the types of atoms in contact, and another term is included to prevent short contacts. After maximizing the complementarity function, LIGIN optimizes the lengths of possible hydrogen bonds. In order to take into account possible movements of the receptor upon ligand binding, steric overlap between the ligand and a specified number of residues in the receptor can be allowed without energy penalty. The location of the bound antagonist OPC-21268 is distinct from the AVPbinding pocket with only partial overlap near the extracellular surface (Fig. 13). The hydrophobic part is embedded in the transmembrane region far deeper than AVP does, whereas the polar part is located on the surface of the extracellular side. The binding pocket is formed by residues from TMSs 4, 5, 6, 7 as well as the third extracellular loop. The 27fold increase in the affinity of the Gly337Ala mutant is explained by the formation of two van der Waals contacts of the methyl carbon with carbon atoms C22 and C28 of the bicyclic ring structure of OPC-21268

195

at the bottom of the cleft (Fig. 13b). The Glu324Asp mutant has an indirect effect. It enables the formation of a hydrogen bond of the carboxylate side chain with the amide side chain atom of Gln311 . This causes a polarization of this amide nitrogen atom and enables it in turn to form another hydrogen bond to the N57 nitrogen atom of OPC-21268. The Ile310Val mutant reduces the hydrophobicity in the vicinity of the polar oxygen atom of the antagonist. The Ile224Val mutant relieves overcrowding in a hydrophobic binding site involving the aromatic residues Trp175 , Phe179 , Phe307 and Trp304 . The smaller valine side chain allows for better positioning of the aromatic residues to interact with the bicyclic ring structure of OPC-21268. Finally, the Ile310Val mutant reduces the hydrophobicity in the vicinity of the polar oxygen atom of the antagonist. Thus, the model explains all the mutations that significantly increase the affinity towards OPC-21268. The combination of site-directed mutagenesis and 3-D modeling in our study identified key residues involved in binding of the non-peptide antagonist OPC-21268 to the V1 R. Our data clearly identified a single residue in the 7th TMS explaining the different affinities of the human and rat V1 R for OPC-21268. The docking model developed for the present study confirmed the importance of this single residue, Ala337 . Furthermore, the model predicts that a serine residue at this position should cause an even tighter binding due to the formation of a hydrogen bond between the serine Oγ atom with the quinoline oxygen atom of OPC-21268 in addition to the van der Waals interaction of the serine β-carbon with carbon atoms 22 and 28 of this antagonist. This study also suggests modifications to the antagonist to increase the affinity for the receptor. For example, elimination of the quinoline oxygen atom should stabilize the interactions with the hydrophobic pocket deep inside the transmembrane region. However, this may cause adverse solubility problems. A similar situation exists for residue 310 of the receptor and oxygen 47 of the antagonist. A hydrophobic residue in the vicinity of this polar atom is clearly unfavorable. A valine at this position, as found in the human sequence, is better than an isoleucine, the corresponding rat residue, but a threonine would be even better. Alternatively, replacement of oxygen 47 of the antagonist with a carbon atom should also

increase the affinity. With respect to residue 224, a valine at this position seems to be optimal. This residue is located in a rather crowded hydrophobic environment into which a valine seems to fit better than the bulkier isoleucine. Combination of the three mutations in positions 224, 324, and 337 did not improve further the affinity of the V1 R for OPC-21268 when compared to the two double mutations, thus suggesting that alterations of the structure of the non-peptide antagonist will be required to increase further the affinity of this compound. The field of GPCRs suffers from a lack of experimentally determined structures. Therefore, molecular modeling is a very useful tool to derive structural information for the V1 R. It provides a framework to design and test new drugs as well as site-specific mutations in a rational way. However, one has to keep in mind the limitations of molecular modeling. The approach is based on the assumption that the seven transmembrane segments are similar in structure to bacteriorhodopsin. The ‘Achilles’ heel’ of this approach are the loops connecting the helical regions as well as the N- and C-terminal non-helical segments. The former were built by sequence similarity to known protein segments from a database within the program LOOK, whereas the N- and C-terminal stretches were left out altogether from the model because they are not involved in ligand binding. The validity of the model is supported by the experimentally determined affinities for the drugs. The model explains very well all of our findings. It does not prove that the model is correct but the model is certainly consistent with the data, and it provides a tool for designing new drugs and mutants. In conclusion, our study provided for the first time the structural basis of species-selective binding of a non-peptide antagonist to the V1 R. These findings should generate new ideas for drug development of non-peptide AVP receptor antagonists and for optimizing drug–receptor interactions. Conclusions In the near future, results of ongoing clinical studies testing the new orally active non-peptide AVP receptor antagonists will tell us if these medications live up to their potential therapeutic indications. V1 R

196

antagonists may turn out to be an effective treatment of dysmenorrhea. V2 R antagonists or aquaretics will undoubtedly facilitate the treatment of hyponatremia. However, because of their potency and the risk of rapid and excessive correction of serum sodium with potential neurologic complications, their use will require caution, especially in patients who do not have free access to water. Dual V1 R/V2 R antagonists in a ratio that remains to be established may provide useful adjuvant treatment of arterial hypertension and congestive heart failure. Finally, the design of specific V3 R antagonists may offer diagnostic tools and medical treatment for ACTH-secreting tumors. Acknowledgements We would like to thank Doreen Conarty, Christine Plesnicher, Xiang Luo and Karim Berrada for their significant contribution to the work summarized in this review. This chapter was supported by grants from the National Institutes of Health to M.T. and from the National Science Foundation to M.S. The Ireland Cancer Research Center Flow Cytometry Core Facility is supported by NIH Grant P30 CA43703 and the Cystic Fibrosis Video Microscope Core Facility is supported by NIH Grant P30 DK2651. M.T. has or has had research contracts with Otsuka, Parke-Davis, Sanofi, and Wyeth-Ayerst. References Barak, L.S., Tiberi, M., Freedman, N.J., Kwatra, M.M., Jelkowitz, R.J. and Caron, M.G. (1994). J. Biol. Chem., 269: 2790– 2795. Barak, L., Ménard, L., Ferguson, S.S.G., Colapietro, A.M. and Caron, M.G. (1995). Biochemistry, 34: 15407–15414. Berrada, K., Plesnicher, C.L., Luo, X. and Thibonnier, M. (2000). J. Biol. Chem., 275: 27229–27237. Chini, B., Mouillac, B., Ala, Y., Balestre, M.N., Trumpp-Kallmeyer, S., Hoflack, J., Elands, J., Hibert, M., Manning, M., Jard, S. and Barberis, C. (1995). EMBO J., 14: 2176–2182. Cotte, N., Balestre, M.N., Phalipou, S., Hibert, M., Manning, M., Barberis, C. and Mouillac, B. (1998). J. Biol. Chem., 273: 29462–29468. Fishman, J.B., Dickey, B.F., Bucher, N.L.R. and Fine, R.E. (1985). J. Biol. Chem., 260: 12641–12646. Han, M., Smith, S.O. and Sakmar, T.P. (1998). Biochemistry, 37: 8253–8261. Henderson, R., Baldwin, J.M., Ceska, T.A., Zemlin, F., Beckmann, E. and Downing, K.H. (1990). J. Mol. Biol., 213: 899– 929.

Hunyady, L., Bor, M., Baukal, A.J., Balla, T. and Catt, K.J. (1995). J. Biol. Chem., 270: 16602–16609. Innamorati, G., Sadeghi, H. and Birnbaumer, M. (1998). J. Biol. Chem., 273: 7155–7161. Laporte, S.A., Servant, G., Richard, D.E., Escher, E., Guillemette, G. and Leduc, R. (1996). Mol. Pharmacol., 49: 89–95. Laporte, S.A., Oakley, R.H., Zhang, J., Holt, J.A., Ferguson, S.S.G., Caron, M.G. and Barak, L.S. (1999). Proc. Natl. Acad. Sci. USA, 96: 3712–3717. Lefkowitz, R.J. (1998). J. Biol. Chem., 273: 18677–18680. Levitt, M. (1992). J. Mol. Biol., 226: 507–533. Luttrell, L.M., Daaka, Y. and Lefkowitz, R.J. (1999). Curr. Opin. Cell Biol., 11: 177–183. Mendre, C., Dufour, M.N., Le Roux, S., Seyer, R., Guillou, L., Calas, B. and Guillon, G. (1997). J. Biol. Chem., 272: 21027– 21036. Mouillac, B., Chini, B., Balestre, M.N., Elands, J., TrumppKallmeyer, S., Hoflack, J., Hibert, M., Jard, S. and Barberis, C. (1995). J. Biol. Chem., 270: 25771–25777. Palczewski, K., Kumasaka, T., Hori, T., Behnke, C., Motoshima, H., Fox, B.A., Le Trong, I., Teller, D.C., Okada, T., Stenkamp, R.E., Yamamoto, M. and Miyano, M. (2000). Science, 289: 739–745. Phalipou, S., Cotte, N., Carnazzi, E., Seyer, R., Mahe, E., Jard, D., Barberis, C. and Mouillac, B. (1997a). J. Biol. Chem., 272: 26536–26544. Phalipou, S., Cotte, N., Carnazzi, E., Seyer, R., Mahe, E., Jard, D., Barberis, C. and Mouillac, B. (1997b). J. Biol. Chem., 272: 26536–26544. Pitcher, J.A., Freedman, N.J. and Lefkowitz, R.J. (1998). Annu. Rev. Biochem., 67: 653–692. Postina, R., Kojro, E. and Fahrenholz, F. (1996). J. Biol. Chem., 271: 31593–31601. Rodriguez, R., Chinea, G., Lopez, N., Pons, T. and Vriend, G. (1998). CABIOS, 14: 523–528. Sobolev, V., Wade, R.C., Vriend, G. and Edelman, M. (1996). Proteins Struct. Funct. Genet., 25: 120–129. Thibonnier, M. (1992). Regul. Pept., 38: 1–11. Thibonnier, M., Berti-Mattera, L.N., Dulin, N., Conarty, D.M. and Mattera, R. (1998a) In: I.J.A. Urban, J.P.H. Burbach and D. De Wied (Eds.), Progress in Brain Research, Vol. 119. Elsevier Science, Amsterdam, pp. 143–158. Thibonnier, M., Conarty, D.M., Preston, J.A., Wilkins, P.L., Berti-Mattera, L.N. and Mattera, R. (1998b). Adv. Exp. Med. Biol., 449: 251–276. Thibonnier, M., Coles, P., Conarty, D.M., Plesnicher, C.L. and Shoham, M. (2000a). J. Pharmacol. Exp. Ther., 294: 195–203. Thibonnier, M., Conarty, D.M. and Plesnicher, C.L. (2000b). Am. J. Physiol., 279: H2529–H2539. Thibonnier, M., Plesnicher, C., Berrada, K. and Berti-Mattera, L. (2000c). Am. J. Physiol., 281: E81–E92. Thibonnier, M., Coles, P., Thibonnier, A. and Shoham, M. (2001). Annu. Rev. Pharmacol. Toxicol., 41: 175–202. Ufer, E., Postina, R., Gorbulev, V. and Fahrenholz, F. (1995). FEBS Lett., 362: 19–23. Vriend, G. (1990). J. Mol. Graph., 8: 52–56.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 15

Nonpeptide vasopressin receptor antagonists: development of selective and orally active V1a, V2 and V1b receptor ligands C. Serradeil-Le Gal 1,∗ , J. Wagnon 2 , G. Valette 1 , G. Garcia 2 , G. Simiand, M. Pascal 1 , J.P. Maffrand 1 and G. Le Fur 3 1

2

Exploratory Research Department, Sanofi-Synthélabo Recherche, 195 Route d’Espagne 31036 Toulouse Cédex, France Exploratory Research Department, Sanofi-Synthélabo Recherche, 371 Rue du Professeur J. Blayac 34184 Montpellier Cédex 04, France 3 Sanofi-Synthélabo Recherche, 174 Avenue de France, 75635 Paris Cédex 13, France

Abstract: The involvement of vasopressin (AVP) in several pathological states has been reported recently and the selective blockade of the different AVP receptors could offer new clinical perspectives. During the past few years, various selective, orally active AVP V1a (OPC-21268, SR49059 (Relcovaptan)), V2 (OPC-31260, OPC-41061 (Tolvaptan), VPA-985 (Lixivaptan), SR121463, VP-343, FR-161282) and mixed V1a /V2 (YM-087 (Conivaptan), JTV-605, CL-385004) receptor antagonists have been intensively studied in various animal models and have reached, Phase IIb clinical trials for some of them. For many years now, our laboratory has focused on the identification of nonpeptide vasopressin antagonists with suitable oral bioavailability. Using random screening on small molecule libraries, followed by rational SAR and modelization, we identified a chemical series of 1-phenylsulfonylindolines which first yielded SR49059, a V1a receptor antagonist prototype. This compound displayed high affinity for animal and human V1a receptors and antagonized various V1a AVP-induced effects in vitro and in vivo (intracellular [Ca2+ ] increase, platelet aggregation, vascular smooth muscle cell proliferation, hypertension and coronary vasospasm). We and others have used this compound to study the role of AVP in various animal models. Recent findings from clinical trials show a potential interest for SR49059 in the treatment of dysmenorrhea and in Raynaud’s disease. Structural modifications and simplifications performed in the SR49059 chemical series yielded highly specific V2 receptor antagonists (N-arylsulfonyl-oxindoles), amongst them SR121463 which possesses powerful oral aquaretic properties in various animal species and in man. SR121463 is well-tolerated and dose-dependently increases urine output and decreases urine osmolality. It induces free water-excretion without affecting electrolyte balance in contrast to classical diuretics (e.g. furosemide and hydrochlorothiazide). Notably, in cirrhotic rats with ascites and impaired renal function, a 10-day oral treatment with SR121463 (0.5 mg/kg) totally corrected hyponatremia and restored normal urine excretion. This compound also displayed interesting new properties in a rabbit model of ocular hypertension, decreasing intraocular pressure after single or repeated instillation. Thus, V2 receptor blockade could be of interest in several water-retaining diseases such as the syndrome of inappropriate antidiuretic hormone secretion (SIADH), liver cirrhosis and congestive heart failure and deserves to be widely explored. Finally, further chemical developments in the oxindole family have led to the first specific and orally active V1b receptor antagonists (with SSR149415 as a representative), an awaited class of drugs with expected therapeutic interest mainly in ACTH-secreting tumors and various emotional diseases such as stress-related disorders, anxiety and depression. However, from the recently described tissue locali-

∗ Correspondence to: C. Serradeil-Le Gal, Exploratory Research Department, Sanofi-Synthélabo Recherche, 195 Route d’Espagne 31036 Toulouse Cédex, France. Tel.: +33-5-6116-2384; Fax: +33-5-6116-2586; E-mail: claudine.serradeil@ sanofi-synthelabo.com

198 zation for this receptor, we could also speculate on other unexpected uses. In conclusion, the development of AVP receptor antagonists is a field of intensive pharmacological and clinical investigation. Selective and orally active compounds are now available to give new insight into the pathophysiological role of AVP and to provide promising drugs. Keywords: Vasopressin; V1a receptor; V1b receptor; V2 receptor; Nonpeptide antagonist; SSR149415; SR49059; SR121463; Aquaretics; ACTH

Introduction AVP, functions and receptors Vasopressin (AVP) is a cyclic nonapeptide which exerts a variety of biological effects in mammals. The primary role of AVP involves the regulation of water and solute excretion by the kidney. However, this hormone is also actively involved in a number of other physiological functions including blood pressure control, platelet aggregation, liver glycogenolysis and neoglucogenesis, uterus contraction, cell proliferation, adrenocorticotropin (ACTH) release by the adenohypophysis, aldosterone secretion by the adrenals and clotting factor release (Barberis et al., 1999). Together with oxytocin (OT), another structurally related nonapeptide, AVP is also implicated in interneuronal communication in the central nervous system (CNS) and modulates several behavioral functions such as feeding, memory, thermoregulation and the control of adaptative, social and sexual processes (Dreifuss et al., 1991). These central and peripheral effects of AVP are based upon a local or systemic release pattern into the organism and occur via interaction with specific seven transmembrane G-protein-coupled receptors. Three AVP receptors (V1a , V1b (or V3 ) and V2 ) and one type of OT receptor have been cloned in animal species and in man and have been clearly identified by their primary structure, gene localization, mRNA distribution, pharmacology and functions (Lolait et al., 1995; Thibonnier et al., 1998). Briefly, the AVP V1 (V1a and V1b ) receptors mediate phospholipase C activation and intracellular calcium mobilization. AVP V1a receptors have a ubiquitous central and peripheral localization (liver, platelet, uterus, brain etc). The AVP V1b receptors are involved mainly in the stimulating effect of AVP on ACTH secretion in the pituitary but, as recently demonstrated, V1b receptors have a wide distribution in the rat brain and also an endocrine role

in other organs such as the pancreas and the adrenals (Lee et al., 1995; Grazzini et al., 1996). Finally, to complete the review of this field, two putative receptors for AVP (or related peptides) have been reported. First, the AVP-activated Ca2+ mobilizing (VACM-1) receptors, unlike other AVP receptors, appear to possess only one transmembrane domain and reportedly bind AVP activating a Ca2+ second messenger pathway. VACM-1 also modulates the cAMP response of V2 receptors when expressed in CHO cells. The cDNA for VACM-1 has been cloned from rabbit, rat and human tissues and mRNA is present in numerous, if not all, tissues, including brain, kidney and endothelial cells. Recently, it has been shown that the VACM-1 protein is identical to cullin-5, which belongs to a newly discovered family of proteins implicated in the cell cycle, clearly different from the AVP/OT receptor family (Burnatowska-Hledin et al., 1995, 2000). Secondly, a novel dual angiotensin II (Ang II)/AVP receptor with an AT1/V2 receptor profile has revealed specific binding for both Ang II and AVP with positive coupling to adenylyl cyclase in response to both hormones. It has been cloned in the rat and in man. This receptor localized in the renal thick ascending limb tubules and collecting ducts, may be involved in renal tubular Na+ and fluid reabsorption. Recent data have also shown a wide distribution in the rat CNS (Ruiz-Opazo et al., 1995; Hurbin et al., 2000). However, clear characterization of this 7-TM receptor using reference peptide and nonpeptide AVP/OT ligands is needed to further explore this entity. At present, the biological role of these atypical proteins, VACM-1 and the mixed AVP/Ang II receptor, remains to be defined. Design of AVP receptor antagonists In the 1980s, a number of cyclic and linear peptide receptor antagonists and agonists derived from

199

the natural hormone and exhibiting various selectivity profiles for AVP and OT receptors were designed by M. Manning in collaboration with W. Sawyer. Amongst them, d(CH2 )5 Tyr(Me)AVP, a selective AVP V1a receptor antagonist still used as a V1a reference; d(CH2 )5 [Tyr (Et)2 , Val4 , D-Arg8 ]VP, a potent V2 /V1a antagonist with aquaretic properties in the rat; d(CH2 )5 [D-Ile2 , Ile4 ]AVP, the first selective V2 peptide ligand and Aaa– D-Tyr(Et)–Phe– Val–Asn–Abu–Pro–Arg–Arg(NH2 ), the first linear V2 /V1a antagonist, represented key steps in the story of peptide AVP receptor antagonists (Manning and Sawyer, 1989, 1991; Jard, 1998). They offered the first valuable pharmacological tools for the classification of AVP/OT receptors and for the acute characterization of their role. They also represented the first generation of radiolabeled ligands for mapping the distribution of AVP and OT receptors. Of note, the team from SmithKline and Beecham Laboratories was particularly active in developing peptide V2 receptor antagonists for clinical purposes. Unfortunately, due to profound species differences well-known in the field of AVP, molecules which were potent V2 receptor antagonists in several animal models turned out to be V2 receptor agonists in humans (Allison et al., 1988). More recently, in the 1990s, besides the molecular cloning of the 4 AVP/OT receptor subtypes, the first nonpeptide AVP V1a /V2 receptor antagonists appeared and allowed, due to their oral bioavailability, clinical evaluation of their therapeutic uses. Chemical structures selective for V1a receptors or V2 receptors or exhibiting a mixed V1a /V2 antagonist profile became available and were intensively studied in various animal models and evaluated in clinical trials (Albright and Chan, 1997; Thibonnier et al., 2001). Finally, very recently in the new millennium, the first selective V1b ligands were discovered (Derick et al., 2000; Serradeil-Le Gal et al., 2002). In the present chapter, we will first review the different orally active, nonpeptide AVP receptor antagonists reported up to now and their potential therapeutic indications. Since our laboratory has focused for many years on the identification of nonpeptide AVP antagonists with suitable oral bioavailability, we will present some recent pharmacological and clinical data supporting their main therapeutic indications. For more details of the development of

orally active V1a , V2 AVP receptor antagonists in general, the reader is referred to the many excellent and accurate reports that have been published recently (Albright and Chan, 1997; Freidringer and Pettibone, 1997; Mayinger and Hensen, 1999; Thibonnier, 1999; Paranjape and Thibonnier, 2001; Thibonnier et al., 2001). The second part of this chapter focuses on the pharmacological characterization of SSR149415, the first selective V1b receptor antagonist described so far (Serradeil-Le Gal et al., 2002). This molecule constitutes a unique probe for exploring the poorly known V1b receptor and the interest of V1b blockade in several pathological states. Nonpeptide AVP V1a and/or V2 receptor antagonists and their potential clinical indications Resulting from high throughput screening (HTS) of thousands of small nonpeptide molecules (molecular weight ≈ 500) belonging to a variety of chemical libraries, the first orally active AVP receptor antagonists were reported in the 1990s and today, an impressive number of patents covers the field of nonpeptide AVP ligands. Only some of them, with different selectivity profiles (V1a , V2 , V1a and V2 ) are undergoing clinical trials (Phase IIb for the most advanced) and the results are eagerly awaited to evaluate both the benefit and the safety of this class of drugs. The term ‘Vaptan’ has been coined to officially name all the members of this new class of drugs (e.g. Relcovaptan (SR49059), Tolvaptan (OPC-41061), Lixivaptan (VPA 985) and Conivaptan (YM-087)). Nonpeptide V1a receptor antagonists The first V1a receptor antagonists, OPC-21268 and SR49059 (Relcoptan) were described in 1991 and 1993, respectively, and have been involved in various clinical trials (Yamamura et al., 1991; Serradeil-Le Gal et al., 1993). They are still the major players identified in this field. These molecules, belonging to different chemical series (quinolinone and indoline series, respectively), were obtained following chemical optimization and SAR of a lead compound found by random screening. It is important to underline that the OPC-quinolinone series was the starting point for the design of many AVP antago-

200 TABLE 1 Orally active, nonpeptide AVP V1a receptor antagonists reported in 2001 and their chemical structures

(1) (2) (3) (4) (5) (6)

Name

Company

Chemical series

Development status

OPC-21268 SR49059 (Relcovaptan) None None None None

Otsuka Sanofi-Synthélabo Eli Lilly Yamanouchi Fujisawa Yamanouchi

Quinolinone derivative N-Sulfonyl-indoline derivative Azetidinone derivatives Benzazepine derivatives Benzazepine derivatives Triazole derivatives

Phase II Japan; stopped US/Europe Phase II/stopped Preclinical Preclinical Preclinical Preclinical

(2)

(4) N

OPC-21268 -

O

O

N

F O O

(1)

N H

NH

O

NH

(5)

CH3O

O

OCH O 3

O

H3 C

(3)

OCH3

O O

(CH3 ) 3

CH 3

SR49059

OC H3 CH 3

N

O

NH2

N

CON H CONH 2

O

N

O O

SO2

N

CH 3

O

N

H

F

N

O N

HO

Cl

N

O

O

Cl

N

(6)

N

N

N N

CH3

nists reported by several competitors (Tables 1–3). Of note, structure–activity relationships among analogues of OPC-21268 yielded also derivatives with marked affinity for the human OT receptor further developed by Merck as selective OT receptor antagonists for the treatment of preterm labor (Freidringer and Pettibone, 1997). In addition, rational modifications of these selective V1a structures (OPC-21268 and SR49059) further yielded potent selective V2 , mixed V1 and V2 , and more recently, pure V1b receptor antagonists (in the case of SR49059) with oral bioavailability, as illustrated in Tables 2, 3 and 5. More recently, two other chemical series of V1a receptor ligands described in patents by Eli Lilly as azetidine derivatives and by Yamamouchi as triazole derivatives, have emerged (see illustration with compounds 3 and 6, respectively, in Table 1). However, up to now, no specific lead compounds have been identified and no pharmacological studies are reported with these new derivatives.

Clinical indications for nonpeptide V1a receptor antagonists Due to the ubiquitous localization of V1a (brain, vessels, platelet, uterus, adrenals. . . ) and to their various central and peripheral biological effects (see Section 1), V1a blockade could be of interest in a large number of diseases. AVP may be increased in several pathological situations such as hypertension, congestive heart failure, dysmenorrhea, brain edema, small cell lung cancers and various CNS disorders (depression, anxiety. . . ) (Table 4). Due to profound interspecies variability in AVP/ OT receptors (Allison et al., 1988; Pettibone et al., 1992), the affinity of OPC-21268 for the human V1a receptor was very weak and this compound failed to prevent AVP-induced contractions in various human vascular preparations in vitro (Serradeil-Le Gal et al., 1993; Burrell et al., 1994). If some clinical trials have been reported with this compound, its loss of affinity for the human V1a receptors has severely precluded clinical developments. At variance with OPC-21268, SR49059 exhibited a constant high affinity for animal and human V1a receptors and

201 TABLE 2 Orally active, non-peptide AVP V2 receptor antagonists reported in 2001 and their chemical structures

(1) (2) (3) (4) (5) (6) (7) (8) (9)

Name

Company

Chemical series

Development status

OPC-31260 OPC-41061 (Tolvaptan) SR121463 VPA-985 (Lixivaptan) WAY-140288 VP-343 and VP-339 VP-365 FR161282 None

Otsuka Otsuka Sanofi-Synthélabo Wyeth-Ayerst Wyeth-Ayerst Wakamoto Wakamoto Fujisawa Ortho-McNeil

Benzazepine derivative Benzazepine derivative N-Arylsulfonyl-oxindole derivative Benzodiazepine derivative Benzodiazepine derivative Quinoxaline derivative Benzodiazepine derivative Benzodiazepine derivative Benzothiazepine derivatives

Phase II Phase II Phase II Phase II Preclinical Preclinical Preclinical Preclinical Preclinical

H3C

N

CH3

OH

H

O

O

N O

CH3 CH2 O

N

N H

N

N

N

N O SO2

O NH

O

VP-343

HNOC

O

SR121463

OH

H3C

N

Cl

NH

NH

OCH 3

OPC-31260 H3C

O

O

VP-339

H 3C

OH

O

N

H3 C N

H3C

S

N

N

N

O

O

N

N

N

N

O

O

H3 C

NH

NH

CH3

O

O

O

F

has been intensively studied in pharmacological and clinical trials as a prototype drug (Serradeil-Le Gal et al., 1993). From several proof of concept studies performed with SR49059 — highly speculative for some of them — a number of clinical indications for AVP V1a receptor antagonists have clearly emerged. Since AVP exerts a powerful local vasoconstriction effect in several crucial vascular beds (renal, gastric, coronary. . . ), SR49059 effects were assessed in patients suffering from Raynaud’s disease. In a double-blind, placebo-controlled randomized cross-over study, at 300 mg p.o. once daily, SR49059 significantly antagonized (15%) cold-induced decrease in finger systolic blood pressure following a cold immersion test and accelerated temperature recovery (Hayoz et al., 2000). These preliminary results suggested an involvement of AVP in the abnormal vasoactive response of Raynaud’s patients

O

NH

NH

NH

WAY140288

H 3C

VPA-985

N

OC H3

O

Cl

O

OPC-41061

H

O

FR-161282 H3 C

O

O

(9)

together with a potential interest of V1a blockade in this pathology. Interesting results were also obtained with SR49059 in the treatment of dysmenorrhea. AVP plasma levels are thought to be increased in primary dysmenorrhea and this hormone is clearly involved in the development of pain by stimulating both vascular and uterine V1a receptors (Bossmar et al., 1997). In a double-blind, randomized cross-over Phase IIb study, SR49059 (100 and 300 mg) induced a dose-related positive effect in reducing pelvic pain during the first 24 h in primary dysmenorrhea, with a simultaneous decrease in the need for analgesic rescue (Brouard et al., 2000). Up to now, no convincing results are available with V1a receptor antagonists in hypertension, either in various animals models or in human, even if AVP seems to be involved in some forms of essential hypertension, in particular, in African–American

202 TABLE 3 Orally active, non-peptide AVP Va /V2 receptor antagonists reported in 2001 and their chemical structures

(1) (2) (3) (4) (5)

Name

Company

Chemical series

Development status

YM-087 (Conivaptan) YM-471 JVT-605 CL-385004 None

Yamanouchi Yamanouchi Japan Tobacco Wyeth-Ayerst Fujisawa

Benzazepine derivative Benzazepine derivative Thiazepine derivative Benzodiazepine derivative Benzazepine derivatives

Phase II Preclinical Preclinical Preclinical Preclinical

YM-471 HN

N

O

CH 3

N

CH 3 CH 3

N

N O

F

N

O

O(CH 2)3C

O

CH 3 N

N CH 3

NH

NH

O

O

N

O

O

CH 3

NH

N

N

N

S

F

CL-385004

F

NH

O

O N

JTV-605

YM-O87

O

OCH 3 N NH

(5)

O

TABLE 4 Main potential clinical indications for AVP V1a , V2 , dual V1a /V2 and V1b receptor antagonists V1a antagonists

V2 antagonists

Mixed V1a /V2 antagonists

V1b antagonists

Dysmenorrhea, preterm labor Raynaud’s disease Hypertension

Congestive heart failure SIADH Liver cirrhosis with ascites and water retention Hyponatremia Nephrotic syndrome Brain edema Glaucoma Hypertension Diabetic nephropathy Meniere’s disease

Congestive heart failure Hypertension Brain edema

Stress-related disorders, anxiety, depression ACTH-secreting tumors, Cushing’s syndrome HPA axis disorders

Congestive failure Brain edema Motion sickness Oncology (SCLC) CNS disorders

patients with elevated AVP plasma levels. An exploratory study with SR49059 (300 mg, single dose) in a situation of osmotic release of AVP (induced by a 5% hypertonic saline infusion) in black hypertensive patients failed to demonstrate a sustained blood pressure reduction (Thibonnier et al., 1999). It is

generally assumed that blockade of both V1a and V2 receptors needs to be achieved to reach a significant improvement in blood pressure alterations. A number of other clinical indications remain to be investigated, in particular CNS disorders with compounds able to cross the blood–brain barrier (Table 4).

203

Nonpeptide V2 receptor antagonists While only two AVP V1a receptor antagonists have been described, numerous selective, orally active AVP V2 receptor antagonists have been reported by several pharmaceutical companies. Ten lead compounds can be identified at various stages of investigation, with four molecules currently involved in clinical developments: OPC-31206, OPC-41061 (Tolvaptan), SR121463 and VPA-985 (Lixivaptan) (Table 2). By modifying the dihydroquinolinone chemical structure of their V1a receptor antagonist, OPC-21268, Otsuka reported in 1992 an orally active V2 compound, OPC-31260, and more recently a back-up/follow-up molecule, OPC-41061 (Tolvaptan) in the benzazepine chemical series (Yamamura et al., 1992, 1998). As shown in Table 2, the other V2 receptor antagonists reported by various pharmaceutical firms are all benzazepine derivatives, except SR121463, an oxindole derivative resulting from optimization and simplification of our V1a receptor antagonist, SR49059. In order to simplify the 1-phenylsufonyl indoline structure of SR49059 bearing three asymmetric carbons, we designed Narylsufonyl-oxindoles yielding highly specific V2 ligands among which SR121463 was chosen for further preclinical and clinical development (SerradeilLe Gal et al., 1996; Serradeil-Le Gal, 2001). Clinical indications for V2 receptor antagonists The generation of receptor-specific AVP V2 antagonists, so-called ‘aquaretics’, able to block the antidiuretic action of AVP in the collecting duct cells, and thus to specifically promote water excretion by preventing the insertion of AVP-specific water channels (aquaporin, AQP-2) into the luminal membrane, could be of high therapeutic value for the treatment of several water-retaining disorders such as SIADH, liver cirrhosis, certain stages of congestive heart failure and hypertension, nephrotic syndrome, renal failure. . . (Table 4). In most of these diseases an abnormal increase of circulating AVP plasma level, activating renal V2 receptors, seems to be the key event in water retention and subsequent hypotonic hyponatremia (Goldsmith et al., 1989; Gavras, 1991; Sorensen et al., 1995). Thus, for these pathologies, there is great clinical interest in the development

of potent V2 receptor antagonists to provide specific water diuretic/aquaretic compounds devoid of the well-known side effects of classical diuretic or saliuretic agents on the urine Na+ and/or K+ loss. According to the above rationale, several clinical trials are reported with OPC compounds and VPA-985 in CHF, cirrhosis with ascites, SIADH with subsequent hyponatremia and in congestive heart failure (Thibonnier, 1999; Paranjape and Thibonnier, 2001). Whatever the compound used, urine volume was increased, urine osmolality decreased and a normalization of serum Na+ observed. Results of repeated chronic treatments in these pathologies are eagerly awaited, but are available in some animal models. Of note, SR121463 demonstrated benefit in a model of cirrhosis (CCl4 -induced) in rats with ascites, water retention and impaired renal function after chronic treatment. Ten-day repeated oral administration of SR121463 (0.5 mg/kg) normalized serum Na+ and totally corrected hyponatremia. SR121463 restored normal urine excretion, urine osmolality and renal function since after a water overload, cirrhotic rats excreted similar urine volume as control noncirrhotic rats (Jimenez et al., 2000). In addition, due to the extrarenal localization of AVP V2 receptors (brain, endothelial, lung. . . ), other therapeutic areas deserve to be explored. For example, V2 receptor blockade could be of interest in brain edema or more surprisingly in glaucoma. As recently shown, in a rabbit model of ocular hypertension (α-chymotrypsin-induced), SR121463, after single or repeated (10 days at 1%) instillation, markedly decreased intraocular pressure with similar efficacy to the currently used α (clonidine) or β-adrenergic (timolol) treatments, demonstrating the potential benefits of V2 receptor antagonists in decreasing intraocular pressure via a mechanism of action that remains to be elucidated. Moreover, it suggested the presence of ocular V2 receptors (Lacheretz et al., 2000). Dual nonpeptide V1a , V2 receptor antagonists The class of dual V1a and V2 receptor antagonists has been extended and includes now several lead compounds reported in Table 3, but only one molecule YM-087 (Conivaptan) has been tested in humans; clinical trials are reported mainly in CHF (Norman et al., 2000; Udelson et al., 2000). Of note, all these

204

compounds are benzazepine derivatives obtained by modifying the chemical structure of the first AVP receptor antagonists (OPC-2128, OPC-31260). According to the compound, the affinity and activity ratio at V1a and V2 receptors is highly variable, generating various pharmacological profiles for these drugs. Interestingly, this V1a /V2 ratio seems a key factor when considering therapeutic purposes. Dual V1a and V2 receptor blockade is expected to modify both systemic hemodynamic and renal parameters. This strategy could be of interest in developing antihypertensive agents. Pure V1a receptor antagonists are reported to be inactive per se in various animal and human models of hypertension (even with increased AVP plasma levels). It is assumed that blockade of both V1a and V2 receptors will achieve a decrease in blood pressure by modifying both peripheral resistances and circulating blood volume. Similarly, dual V1a /V2 blockade could improve hemodynamic and fluid status in CHF. Finally, treatment of brain edema with this class of drugs needs also to be explored, based on the rationale that pure V1a and pure V2 receptor antagonists have both shown a benefit in the development of this pathology by decreasing brain water content and restoring brain Na+ content with a decrease in neurogenic inflammation. An involvement of peripheral and probably direct effect on brain vessel permeability and choroid plexus could be speculated even if the intrinsic mechanism of action is not known (Bemana et al., 1997; Laszlo et al., 1999). V1b receptor antagonists V1b receptor ligands The recently cloned V1b receptor, mainly found in the adenohypophysis, is involved in the stimulating effect of AVP on ACTH secretion (De Keyzer et al., 1994; Sugimoto et al., 1994). AVP is a direct ACTH secretagogue but also synergizes corticotropin-releasing factor (CRF)-induced ACTH release in the pituitary (Gillies et al., 1982). This receptor has a wide distribution in various tissues such as the brain, adrenals, kidney and pancreas. To date, due to the lack of selective V1b receptor ligands (agonists/antagonists) and to the absence of orally active V1b receptor antagonists, the V1b receptor is still poorly characterized and the precise role of AVP

via central and peripheral V1b receptors remains to be elucidated. Interestingly, to explore the functions of this receptor, a knockout mouse has recently been generated (Lolait et al., 2000). In contrast to the numerous potent and selective, peptide and nonpeptide V1a and V2 receptor ligands, only a few non-selective V1b peptides are available: [D-3(pyridyl)Ala2 ]AVP, a specific V1b agonist in rats turned out to be a V1a , V1b ligand at human AVP receptors; dDAVP displays both agonist V2 /V1b properties and a recent series of dDAVP analogues modified at position 2 yielded full V1b /partial V1a agonists (Derick et al., 2000). Finally, the reference peptide antagonist, [deaminopenicillamine-O-Me-Tyr,Arg]AVP, (dPen), is a dual V1b /V1a ligand. Recently, we have developed the first selective, nonpeptide and orally active V1b receptor antagonist to be described, SSR149415 (Serradeil-Le Gal et al., 2002 (Table 5)). It results from chemical optimization in the field of the indoline/oxoindole chemical series which has previously yielded nonpeptide molecules highly selective for the V1a (SR49059, Tables 1 and 6) and the for the V2 (SR121463, Tables 2 and 6) receptors. Pharmacological profile of SSR149415 As shown in Table 6, SSR149415 displays nanomolar affinity for human V1b receptors. It exhibits a TABLE 5 Orally active nonpeptide AVP V1b receptor antagonists reported in 2002 Name

Company

Chemical series

Development status

SSR149415

SanofiSynthélabo

N-Arylsulfonyloxindole derivative

Preclinical

OH OCH3 Cl

N N N

CH3

O O

SO2

OCH3 C H3O

CH3

205 TABLE 6 Comparative affinities of the nonpeptide compounds, SR49059, SR121463 and SSR149415 for human AVP (V1a , V1b , V2 ) and OT receptors K i (nM)

SR49059 SR121463 SSR149415 AVP OT

h-V1a

h-V1b

h-V2

h-OT

6.3 ± 0.6 460 ± 120 91 ± 23 1.7 64

220 ± 30 >10,000 1.5 ± 0.8 1.1 1782

275 ± 50 4.1 ± 0.8 1412 ± 314 1.1 167

320 ± 168 1213 ± 383 174 ± 35 16 0.9

Affinities of the natural hormones, AVP and OT, as references. Inhibition constants (K i ) were determined from competition experiments calculated according to the Cheng and Prussoff equation. Values are the mean ± SEM of at least three determinations.

%

highly selective profile versus V1a , V2 and OT human receptors and has no measurable affinity for a number of other receptors (n = 100). As illustrated in Fig. 1, SSR149415 dose-dependently antagonized [3 H]AVP binding to human V1b receptors with an affinity for human V1b receptors close to that of the natural hormone, AVP (K i values of 1.54 ± 0.82 and 0.80 ± 0.25 nM, respectively). SSR149415 exhibited much higher affinity than the nonselective reference

-10

-9

-8

-7

-6

Fig. 1. Inhibition of [3H]AVP specific binding to human V1b receptors by SSR149415 () and reference peptide compounds: AVP (•); dDAVP, [desamino-[ D-Arg]vasopressin] ( ); dPen, [desamino-[ D-Arg]vasopressin] (◦) and dPal, [(deamino-Cys, D3-(Pyridyl)-Ala2 -Arg8 )-vasopressin] (). Binding assays were performed for 45 min at 20°C in the presence of 30 μg/assay of CHO membranes expressing the human V1b receptors. Results represent data from a typical experiment performed in duplicate, which was repeated three times without noticeable differences.

agonist (dDAVP, [D-3(pyridyl)Ala2 ]AVP) and antagonist (dPen) V1b peptides (K i values of 20 ± 8, 12 ± 5 and 21 ± 6 nM, respectively). K i values obtained for these peptides are consistent with affinities previously reported for the human V1b receptor. In saturation binding experiments followed by Scatchard analysis, SSR149415 inhibited [3 H]AVP binding in a competitive manner with a K i value of 2.51 ± 0.45 nM (Serradeil-Le Gal et al., 2002). Earlier cellular events upstream of ACTH release, provoked by occupancy of corticotroph V1b receptors by AVP, include the activation of phospholipase C, protein kinase C and the mobilization of intracellular free Ca2+ mainly via Gq/11 G-protein recruitment. In CHO cells transfected with the human V1b receptor other intracellular pathways have also been described (e.g. cAMP production, stimulation of DNA synthesis and cell proliferation), clearly depending on the level of V1b receptor expression (Thibonnier et al., 1997). In CHO cells expressing the human V1b receptors AVP-induced cell proliferation with an EC50 value of 0.14 ± 0.12 nM (n = 4) as measured by a colorimetric method with the tetrazolium compound (MTS) according to (Thibonnier et al., 1998). In this latter model, SSR149415 (Fig. 2) dose-dependently antagonized stimulation of cell proliferation by AVP (3 nM) with a K i value of 0.43 ± 0.41 nM consistent with the SSR149415 affinity found in binding studies using the same cellular preparation. In vivo, pharmacology performed measuring ACTH secretion induced by various stimulants, such as hormones and physical stress, confirmed the full antagonist profile of SSR149415. In all these situations, SSR149415 antagonized ACTH secretion, which constitutes a critical response of the organism to stress in emotional situations (Aguilera and Rabadan-Diehl, 2000). It is important to note that SSR149415 demonstrated high potency by the oral route on the potentiation of the effect on CRF by AVP, a mechanism described as typically V1b mediated. Significant inhibition of ACTH secretion was observed from a dose of 3 mg/kg p.o., total blockade occurred at 10 mg/kg p.o., an effect that lasted for more than 4 h (Fig. 3). Conversely, the selective, orally active V1a receptor antagonist, SR49059, was unable to inhibit AVP plus CRFinduced ACTH secretion, demonstrating a specific V1b -mediated effect. Various physical stresses are

206

%

(A)

-10

-9

-8

-7

-6

-5

Fig. 2. Inhibition by SSR149415 of AVP-induced proliferation of CHO cells transfected with the human V1b receptor. CHO cells were grown for 48 h in 96-well plates (5000 cells/well). Cell proliferation was measured using the CellTiter 96 cell proliferation assay from Promega (Madison, WI). Cells were washed with 200 μl PBS and treated with AVP (3 nM) and increasing concentrations of SSR149415. After 18 h of incubation (37°C, 5% CO2 , 80% humidity), 20 μl dye solution was added to each well. The plate was incubated for 4 h and absorbance was recorded at 492 nm wavelength. Results represent data from a typical experiment which was repeated four times without noticeable change.

able to induce ACTH secretion. The stress-induced release of ACTH is believed to involve the activation of several humoral and neural pathways, including that mediated by AVP (Linton et al., 1985). As shown in Fig. 4, in rats submitted to an immobilization period of 15 min, there was a significant increase (more than 5-fold) in plasma ACTH levels. Pretreatment with SSR149415 (3–10 mg/kg i.p.) 30 min before the restraint stress period, caused dosedependent inhibition of plasma ACTH elevation in comparison with stressed animals treated with the corresponding vehicle. The regulation of ACTH secretion and consequently of the HPA axis are largely mediated by AVP and V1b receptors, and SSR149415 offers a new tool for controlling emotional or physical stress. Indeed, several neuroendocrine studies strongly suggest that dysregulation of the HPA system plays a causal role in the development and the course of diseases such as generalized anxiety, depression and addiction (Holsboer, 1999). In conclusion, SSR149415 is a potent, selective and orally active V1b receptor antagonist. Whatever the model, SSR149415 is devoid of any V1b agonist

(B)

Fig. 3. Effect of oral SSR149415 on the potentiation of exogenous CRF by AVP on ACTH secretion in conscious rats. Dose–effect (A) and time-course (B) studies. (A) SSR149415 (1–10 mg/kg) was administered by gavage 2 h prior to CRF (0.1 μg/kg i.v.) and AVP (0.03 μg/kg i.v.) injection. Inset: effect of a selective orally active V1a receptor antagonist, SR49059 (1–10 mg/kg) in this model. (B) The time-course effect of SSR149415 was studied at 10 mg/kg p.o. on plasma ACTH secretion (n = 5–10). Values are expressed as the percentage of ACTH secretion versus the CRF plus AVP control. Statistical analysis was performed using a one-way ANOVA followed by a Dunnett’s test or using the nonparametric Kruskal–Wallis test. (* P < 0.05; ** P < 0.01 versus control).

effect when tested alone. It is a unique tool for exploring the localization and the role of V1b receptors. This class of drugs exhibits a promising therapeutic profile for the treatment of stress-related disorders, anxiety and depression. However, due to the ubiquitous localization of the V1b receptor, several other

207

Fig. 4. Effect of SSR149415 on restraint stress-induced ACTH secretion in conscious rats. Rats (8–10 per group) received intraperitoneal injection of the vehicle (2 ml/kg 5% DMSO, 5% cremophor, 90% saline) or SSR149415 (3 and 10 mg/kg). Thirty minutes after the injection, rats were placed in plexiglass restrainers for 15 min. At the end of the immobilization period, they were sacrificed by decapitation and plasma ACTH levels were measured by RIA. Data are mean ± SEM. Statistical significance was assessed by an ANOVA on transformed logarithmic data followed by Dunett’s test (** P < 0.01 vs. control – unrestraint animals; ££ P < 0.01 vs. stress control – restraint animals).

therapeutic indications need to be investigated (Table 4). Conclusion AVP mediates a wide number of biological effects and may be involved in several pathological states. Selective blockade of the different AVP receptors, therefore, could offer new clinical perspectives for treating several diseases (Table 4). In 2001, four classes of orally active AVP receptor nonpeptide antagonists (V1a , V2 , V1a + V2 , and V1b ) were available, constituting the ‘Vaptan family’. In the past decade, various, selective, nonpeptide and orally active AVP V1a (OPC-21268, SR49059 (Relcovaptan)), V2 (OPC-31260, OPC-41061 (Tolvaptan), VPA985 (Lixivaptan), SR121463, VP-343, FR-161282, WAY-140288) and mixed V1a /V2 (YM-087 (Conivaptan), JTV-605, CL-385004) receptor antagonists have been intensively studied in various animal models and have reached, for some of them, Phase IIb clinical trials. It should be noted, however, that no compounds are on the market, many clinical trials and proof of concept studies are still ongoing and the results of long-term treatments are eagerly awaited to evaluate the benefit and the safety of this new class of drugs.

A wide variety of therapeutic indications could be targeted with these compounds: SR49059, a potent and selective V1a receptor antagonist may be of interest in the treatment of dysmenorrhea and in Raynaud’s disease while V2 receptor antagonists were of benefit in several water-retaining diseases (CHF, SIADH and hepatic cirrhosis) by inducing powerful aquaretic effects without modification of Na+ and/or K+ excretion, at variance with classical diuretics. They improved urine excretion, urine osmolality and renal function, and subsequently, normalized serum Na+ (partially or totally) and corrected hyponatremia. Extrarenal localization of V2 receptors has also been evidenced. Other therapeutic domains with V2 receptor antagonists remain to be explored, notably in glaucoma where SR121463, a selective V2 compound, decreased intraocular pressure by local instillations, as do the reference drugs currently used (timolol and clonidine). Finally, we have developed the first selective, orally active V1b receptor antagonist. This compound constitutes an invaluable tool for exploring this poorly characterized receptor and the precise role of AVP via central and peripheral V1b receptors. SSR149415, a representative member of this class, has potential for the treatment of anxiety, depression and stressrelated disorders. In addition, in view of the recently

208

described tissue localization of the V1b receptor protein and mRNA (brain, pancreas, adrenals. . . ), still other potential uses may be possible. Abbreviations ACTH Ang II AVP CHF CRF dDAVP dPen

adrenocorticotropin angiotensin II arginine vasopressin congestive heart failure corticotropin releasing factor desamino-[D-Arg]vasopressin [deaminopenicillamine-O-MeTyr,Arg]vasopressin HTS high throughput screening OT oxytocin SIADH syndrome of inappropriate antidiuretic hormone secretion VACM-1 receptor vasopressin-activated Ca2+ mobilizing receptor

Acknowledgements The authors would like to acknowledge the teams of Sanofi-Synthélabo Recherche from the Discovery, Preclinical, Clinical and Chemical Development Departments for their work on the Vasopressin program. Our thanks are extended to external contributors for their work and their interest in our compounds. We are thankful to Dr R. Pruss and A.J. Patacchini for helpful comments on the manuscript and M. Laborde for her skilful secretarial assistance. C.S.L.G. thanks J.L., Y. and M. Le Gal for their invaluable support and their encouragement. References Aguilera, G. and Rabadan-Diehl, C. (2000) Vasopressinergic regulation of the hypothalamic–pituitary–adrenal axis: implications for stress adaptation. Regul. Pept., 96: 23–29. Albright, J.D. and Chan, P.S. (1997) Recent advances in the discovery and development of vasopressin antagonists: Peptide and nonpeptide V-1a and V-2 receptor antagonists. Curr. Pharm. Des., 3: 615–632. Allison, N.L., Albrightson-Winslow, C.R., Brooks, D.P., Stassen, F.L., Huffman, W.F., Stote, R.M. and Kinter, L.B. (1988) Species heterogeneity and antidiuretic hormone antagonists: what are the predictors? In: J.F. Liard and D.A. Ausiello

(Eds.), Vasopressin: Cellular and Integrative functions. Raven Press, New York, pp. 207–214. Barberis, C., Morin, D., Durroux, T., Mouillac, B., Guillon, G., Seyer, R., Hibert, M., Tribollet, E. and Manning, M. (1999) Molecular pharmacology of AVP and OT receptors and therapeutic potential. In: Drug News Perspect., 12: 279–292. Bemana, I., Takahashi, E., Nakamura, T., Kuyama, H. and Nagao, S. (1997) OPC-21268, an orally effective, nonpeptide arginine vasopressin V1 receptor antagonist reduces vasogenic brain edema. Acta Neurochir. Suppl. (Wien), 70: 194–197. Bossmar, T., Brouard, R., Doberl, A. and Akerlund, M. (1997) Effects of SR 49059, an orally active V1a vasopressin receptor antagonist, on vasopressin-induced uterine contractions. Br. J. Obstet. Gynaecol., 104: 471–477. Brouard, R., Bossmar, T., Fournie-Lloret, D., Chassard, D. and Akerlund, M. (2000) Effect of SR49059, an orally active V1a vasopressin receptor antagonist, in the prevention of dysmenorrhoea. Br. J. Obstet. Gynaecol., 107: 614–619. Burnatowska-Hledin, M.A., Spielman, W.S., Smith, W.L., Shi, P., Meyer, J.M. and Dewitt, D.L. (1995) Expression cloning of an AVP-activated, calcium-mobilizing receptor from rabbit kidney medulla. Am. J. Physiol., 268: F1198–F1210. Burnatowska-Hledin, M., Zhao, P., Capps, B., Poel, A., Parmelee, K., Mungall, C., Sharangpani, A. and Listenberger, L. (2000) VACM-1, a cullin gene family member, regulates cellular signaling. Am. J. Physiol. Cell Physiol., 279: C266– C273. Burrell, L.M., Phillips, P.A., Rolls, K.A., Buxton, B.F., Johnston, C.I. and Liu, J.J. (1994) Vascular responses to vasopressin antagonists in man and rat. Clin. Sci. (Colch.), 87: 389–395. De Keyzer, Y., Auzan, F., Beldford, C., Thibonnier, M., Bertagna, X. and Clauser, E. (1994) Cloning and characterization of the human V3 pituitary vasopressin receptor. FEBS Lett., 356: 215–220. Derick, S., Andrès, M., Guillon, G., Cheng, L., Stocv, S. and Manning, M. (2000) Characterization of potent new vasopressin analogs to study vasopressin V1b receptors. J. Pept. Sci., 6: (Abstract). Dreifuss, J.J., Tribollet, E., Goumaz, M., Dubois-Dauphin, M. and Raggenbass, M. (1991) Vasopressin receptor localization and neuronal responsiveness in the rat brain. In: S. Jard and R. Jamison (Eds.), In Vasopressin. John Libbey Eurotext, Paris, pp. 159–166. Freidringer, R.M. and Pettibone, D.J. (1997) Small molecule ligands for oxytocin and vasopressin receptors. Med. Res. Rev., 17: 1–16. Gavras, H. (1991) Role of vasopressin in clinical hypertension and congestive cardiac failure: interaction with the sympathetic nervous system. Clin. Chem., 37: 1828–1830. Gillies, G.E., Linton, E.A. and Lowry, P.J. (1982) Corticotropin releasing activity of the new CRF is potentiated several times by vasopressin. Nature, 299: 355–357. Goldsmith, S.R., Dodge-Brown, D.L. and Katz, A. (1989) Alpha 2-adrenergic stimulation and vasopressin in congestive heart failure. J. Cardiovasc. Pharmacol., 14: 425–429. Grazzini, E., Lodboerer, A.M., Perez-Martin, A., Joubert, D. and Guillon, G. (1996) Molecular and functional character-

209 ization of V1b vasopressin receptor in rat adrenal medulla. Endocrinology, 137: 3906–3914. Hayoz, D., Bizzini, G., Noel, B., Depairon, M., Burnier, M., Fauveau, C., Rouillon, A., Brouard, R. and Brunner, H.R. (2000) Effect of SR 49059, a V1a vasopressin receptor antagonist, in Raynaud’s phenomenon. Rheumatology, 39: 1132–1138. Holsboer, F. (1999) The rationale for corticotropin-releasing hormone receptor (CRH-R) antagonists to treat depression and anxiety. J. Psychiatr. Res., 33: 181–214. Hurbin, A., Orcel, H., Ferraz, C., Moos, F.C. and Rabie, A. (2000) Expression of the genes encoding the vasopressinactivated calcium-mobilizing receptor and the dual angiotensin II/vasopressin receptor in the rat central nervous system. J. Neuroendocrinol., 12: 677–684. Jard, S. (1998) Vasopressin receptors. In: Zingg et al. (Eds.), Vasopressin and Oxytocin. Plenum Press, New York, pp. 1–13. Jimenez, W., Serradeil-Le Gal, C., Ros, J., Cano, C., Cejudo, P., Morales-Ruiz, M., Arroyo, V., Pascal, M., Rivera, F., Maffrand, J.P. and Rodés, J. (2000) Long-term aquaretic efficacy of a selective nonpeptide V2-vasopressin receptor antagonist, SR121463, in cirrhotic rats. J. Pharmacol. Exp. Ther., 295: 83–90. Lacheretz, F., Barbier, A., Serradeil-Le Gal, C., Elena, P.P., Maffrand, J.P. and Le Fur, G. (2000) Effect of SR121463, a selective non-peptide vasopressin V2 receptor antagonist, in a rabbit model of ocular hypertension. J. Ocul. Pharmacol. Ther., 16: 203–216. Laszlo, F.A., Varga, C. and Nakamura, S. (1999) Vasopressin receptor antagonist OPC-31260 prevents cerebral oedema after subarachnoid haemorrhage. Eur. J. Pharmacol., 364: 115–122. Lee, B., Yang, C., Chen, T.H., al Azawi, N. and Hsu, W.H. (1995) Effect of AVP and oxytocin on insulin release: involvement of V1b receptors. Am. J. Physiol., 269: E1095–E1100. Linton, E.A., Tilders, F.J., Hodgkinson, S., Berkenbosch, F., Vermes, I. and Lowry, P.J. (1985) Stress-induced secretion of adrenocorticotropin in rats is inhibited by administration of antisera to ovine corticotropin-releasing factor and vasopressin. Endocrinology, 116: 966–970. Lolait, S.J., O’Carroll, A.M. and Brownstein, M.J. (1995) Molecular biology of vasopressin receptors. Ann. N.Y. Acad. Sci., 771: 273–292. Lolait, S.J., O’Carroll, A.M., Shepard, E., Ginns, E.I. and Young, W.S. (2000) Characterization of a vasopressin V1b receptor knockout mouse. Soc. Neurosci. Abstr., 26: 2406. Manning, M. and Sawyer, W.H. (1989) Discovery, development, and some uses of vasopressin and oxytocin antagonists. J. Lab. Clin. Med., 114: 617–632. Manning, M. and Sawyer, W.H. (1991) Antagonists of vasopressin and oxytocin: current status and future perspective. In: S. Jard and R. Jamison (Eds.), Vasopressin, Vol. 208. John Libbey Eurotext, Paris, pp. 297–309. Mayinger, B. and Hensen, J. (1999) Nonpeptide vasopressin antagonists: a new group of hormone blockers entering the scene. Exp. Clin. Endocrinol. Diabetes, 107: 157–165. Norman, P., Leeson, P.A., Rabasseda, X., Castaner, J. and Castaner, R.M. (2000) Conivaptan hydrochloride. Drugs Future, 25: 1121–1130.

Paranjape, S.B. and Thibonnier, M. (2001) Development and therapeutic indications of orally-active non-peptide vasopressin receptor antagonists. Expert Opin. Investig. Drugs, 10: 825–834. Pettibone, D.J., Kishel, M.T., Woyden, C.J., Clineschmidt, B.V., Bock, M.G., Freidinger, R.M., Veber, D.F. and Williams, P.D. (1992) Radioligand binding studies reveal marked species differences in the vasopressin V1 receptor of rat, rhesus and human tissues. Life Sci., 50: 1953–1958. Ruiz-Opazo, N., Akimoto, K. and Herrera, V.L. (1995) Identification of a novel dual angiotensin II/vasopressin receptor on the basis of molecular recognition theory. Nat. Med., 1: 1074–1081. Serradeil-Le Gal, C. (2001) An overview of SR121463, a selective non-peptide vasopressin V2 receptor antagonist. Cardiovascular Drug Rev., 19: 201–214. Serradeil-Le Gal, C., Wagnon, J., Garcia, C., Lacour, C., Guiraudou, P., Christophe, B., Villanova, G., Nisato, D., Maffrand, J.P. and Le Fur, G. (1993) Biochemical and pharmacological properties of SR 49059, a new, potent, nonpeptide antagonist of rat and human vasopressin V1a receptors. J. Clin. Invest., 92: 224–231. Serradeil-Le Gal, C., Lacour, C., Valette, G., Garcia, G., Foulon, L., Galindo, G., Bankir, L., Pouzet, B., Guillon, G., Barberis, C., Chicot, D., Jard, S., Vilain, P., Garcia, C., Marty, E., Raufaste, D., Brossard, G., Nisato, D., Maffrand, J.P. and Le Fur, G. (1996) Characterization of SR 121463A, a highly potent and selective, orally active vasopressin V2 receptor antagonist. J. Clin. Invest., 98: 2729–2738. Serradeil-Le Gal, C., Wagnon, J., Simiand, J., Griebel, G., Lacour, C., Guillon, G., Barberis, C., Brossard, G., Soubrié, P., Nisato, D., Pascal, M., Pruss, R., Scatton, B., Maffrand, J.P. and Le Fur, G. (2002) Characterization of SSR149415, a selective and orally-active vasopressin V1b receptor antagonist. J. Pharmacol. Exp. Ther., 300: 1122–1130. Sorensen, J.B., Andersen, M.K. and Hansen, H.H. (1995) Syndrome of inappropriate secretion of antidiuretic hormone (SIADH) in malignant disease. J. Intern. Med., 238: 97–110. Sugimoto, T., Saito, M., Mochizuki, S., Watanabe, Y., Hashimoto, S. and Kawashima, H. (1994) Molecular cloning and functional expression of a cDNA encoding the human V1b vasopressin receptor. J. Biol. Chem., 269: 27088–27092. Thibonnier, M. (1999) Aquaretics. Vasopressin V2-renal receptor antagonists. Emerg. Drugs, 4: 223–235. Thibonnier, M., Preston, J.A., Dulin, N., Wilkins, P.L., BertiMattera, L.N. and Mattera, R. (1997) The human V3 pituitary vasopressin receptor: ligand binding profile and densitydependent signaling pathways. Endocrinology, 138: 4109– 4122. Thibonnier, M., Conarty, D.M., Preston, J.A., Wilkins, P.L., Berti-Mattera, L.N. and Mattera, R. (1998) Molecular pharmacology of human vasopressin receptors. In: Zingg et al. (Ed.), Vasopressin and Oxytocin. Plenum Press, New York, pp. 251. Thibonnier, M., Kilani, A., Rahman, M., Di Blasi, T., Warner, K., Smith, M.C., Leenhardt, A.F. and Brouard, R. (1999) Effects of the nonpeptide V(1) vasopressin receptor antagonist

210 SR49059 in hypertensive patients. Hypertension, 34: 1293– 1300. Thibonnier, M., Coles, P., Thibonnier, A. and Shoham, M. (2001) The basic and clinical pharmacology of nonpeptide vasopressin receptor antagonists. Annu. Rev. Pharmacol. Toxicol., 41: 175–202. Udelson, J.E., Smith, W.B., Hendrix, G.H., Painchaud, C.A. and Ghazzi, M.M. (2000) Hemodynamic effects of conivaptan hydrochloride (YM087, CI-1025) a combined vasopressin V1a and V2 receptor antagonist, in patients with NYHA classIII/IV heart failure. Circulation, 102: 593 (abstract). Yamamura, Y., Ogawa, H., Chihara, T., Kondo, K., Onogawa, T., Nakamura, S., Mori, T., Tominaga, M. and Yabuuchi, Y. (1991) OPC-21268, an orally effective, nonpeptide vasopressin V1 receptor antagonist. Science, 252: 572–574.

Yamamura, Y., Ogawa, H., Yamashita, H., Chihara, T., Miyamoto, H., Nakamura, S., Onogawa, T., Yamashita, T., Hosokawa, T. and Mori, T. (1992) Characterization of a novel aquaretic agent, OPC-31260, as an orally effective, nonpeptide vasopressin V2 receptor antagonist. Br. J. Pharmacol., 105: 787–791. Yamamura, Y., Nakamura, S., Itoh, S., Hirano, T., Onogawa, T., Yamashita, T., Yamada, Y., Tsujimae, K., Aoyama, M., Kotosai, K., Ogawa, H., Yamashita, H., Kondo, K., Tominaga, M., Tsujimoto, G. and Mori, T. (1998) OPC-41061, a highly potent human vasopressin V2-receptor antagonist: pharmacological profile and aquaretic effect by single and multiple oral dosing in rats. J. Pharmacol. Exp. Ther., 287: 860–867.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 16

Rat vasopressin mRNA: a model system to characterize cis-acting elements and trans-acting factors involved in dendritic mRNA sorting Evita Mohr ∗ , Iris Kächele, Carola Mullin and Dietmar Richter Universität Hamburg, Institut für Zellbiochemie und Klinische Neurobiologie, Martinistrasse 52, 20246 Hamburg, Germany

Abstract: The genes encoding the vasopressin (VP) and oxytocin (OT) precursors are expressed in magnocellular neurons of the hypothalamo-neurohypophyseal system. The neuropeptides have a dual function: (1) they are secreted from the nerve terminals into the systemic circulation to act as hormones on various peripheral target organs; and (2) VP and OT are also released from the dendrites into the central nervous system where they presumably play a role as either neurotransmitters or as modulators of the classical transmitters. Substantial amounts of VP and OT mRNAs are sorted to both axons and dendrites. Since the latter are equipped with components of the translational machinery, the peptide hormone precursors are likely to be locally synthesized in dendrites of magnocellular neurons. Evidence for axonal precursor synthesis, on the other hand, has not been obtained. Subcellular mRNA localization is a complex pathway. It is determined by sequences (cis-acting elements) within the RNA and proteins (trans-acting factors) which interact with these elements in order to guide the molecules to their ultimate destination. We have investigated the mechanisms involved in mRNA targeting in neurons by using VP mRNA as a model system. Recombinant eukaryotic expression vectors harboring the VP cDNA have been microinjected into the cell nuclei of cultured superior cervical ganglion (SCG) neurons. The subcellular distribution of the vector-expressed mRNAs was determined by non-radioactive in situ hybridization techniques. This revealed transport of VP mRNA to the dendrites, but not to the axonal compartment of SCG neurons. A complex dendritic localizer sequence (DLS) that spans part of the coding region as well as the 3 -untranslated region was identified by microinjecting constructs encoding partial sequences of the VP mRNA. In order to characterize trans-acting factors interacting with this element, protein/RNA binding experiments with radiolabeled in vitro synthesized VP RNA probes and proteins extracted from rat brain have been carried out. A protein specifically interacts with the DLS of the VP mRNA but not with sequences that obviously lack a role in subcellular RNA transport. Biochemical purification revealed that this protein is the multifunctional poly(A)-binding protein (PABP). It is well known for its ability to bind with high affinity to poly(A) tails of mRNAs, prerequisite for mRNA stabilization and stimulation of translational initiation, respectively. With lower affinities, PABP can also associate with non-poly(A) sequences. The physiological consequences of these PABP/RNA interactions include functions such as translational silencing. The translational state of mRNAs subject to dendritic sorting is most likely influenced by external stimuli. Consequently, PABP could represent one of several components necessary to regulate local synthesis of the VP precursor and possibly of other proteins. Keywords: cis-acting element; Oxytocin; Poly(A)-binding protein; RNA-binding protein; RNA localization; trans-acting factor; Vasopressin; UV-crosslinking

∗ Correspondence to: E. Mohr, Institut für Zellbiochemie und klinische Neurobiologie, Universität Hamburg, Martinistrasse 52, 20246 Hamburg, Germany. Tel.: +49-40-42803-4553; Fax: +49-40-42803-4541; E-mail: [email protected]

212

Introduction mRNA transport to cytoplasmic regions outside the cell somata is observed in various nerve cell types of the central and peripheral nervous system. Most often, defined transcripts are specifically delivered to dendrites, and the functional significance of this transport process appears to be evident: Dendrites are equipped with ribosomes and many additional components essential for translation and are, thus, competent for local protein synthesis (for review see Mohr, 1999; Tiedge et al., 1999; Steward and Worley, 2001). In fact, there is an ever-growing body of evidence indicating that on-site translation in dendrites is necessary for establishing at least particular forms of synaptic plasticity (for review see Schuman, 1999; Steward and Schuman, 2001). Hence, mRNA localization to the dendrites of the nerve cells is presumably important because synthesis of proteins involved in modulating synapse functions may be regulated in terms of both space and time. As far as mammalian nerve cells are concerned, definite proof for local protein biosynthesis in axons with the exception of the unmyelinated initial segment is still lacking. Notwithstanding, in vivo, specific delivery of distinct mRNA species to axons of nerve cells, for example primary sensory neurons projecting to the olfactory bulb and hypothalamic magnocellular neurons, has been observed (for review see Mohr and Richter, 2000), but their functional role has remained uncertain. In even rarer cases, the same mRNA species is localized to the axon as well as to dendrites, examples being the VP mRNA and transcripts encoding the Purkinje cell-specific L7 protein (for review see Mohr, 1999). By no means is subcellular mRNA localization restricted to neurons. On the contrary, this phenomenon is apparently conserved during evolution and is observed in many cell types from organisms as diverse as yeast and human. It is perhaps not surprising, however, that work on RNA sorting in postmitotic nerve cells has remained more or less descriptive for many years. Early on, the lead in this field was taken by investigations aimed at characterizing the RNA localization mechanisms in non-neuronal cells. Particularly well studied examples include developmental systems such as Xenopus oocytes, Drosophila oocytes and early embryos, but also dif-

ferentiated cell types in various organisms from yeast to mammalian species (for review see Bashirullah et al., 1998; Barbarese et al., 1999; Gonzalez et al., 1999; King et al., 1999; Lipshitz and Smibert, 2000). These studies demonstrated that RNA localization is remarkably complex and is determined by a variety of molecular entities: (1) sequence elements or motifs within the RNA molecule (cis-acting elements) that may form secondary, tertiary and, at least in some cases, quarternary structures; (2) proteins (trans-acting factors) as mediators of RNA sorting via a direct or indirect interaction with the mRNA to be transported; and (3) mechanisms of translational control including silencing and subsequent activation (or derepression). In addition, cytoskeletal elements provide the transport tracks and anchor sites of the ribonucleoproteins (for review see Bashirullah et al., 1998; Barbarese et al., 1999; Gonzalez et al., 1999; King et al., 1999; Lipshitz and Smibert, 2000). It may be assumed that mRNA localization in nerve cells takes place in a similar manner. This is underscored by the recent identification of cisacting elements, termed dendritic targeting element or dendritic localizer sequence, necessary and sufficient for dendritic targeting of mRNAs encoding the microtubule-associated protein 2 (MAP2; Blichenberg et al., 1999), the VP precursor (Prakash et al., 1997) and the α-subunit of Ca2+ /calmodulindependent protein kinase II (Mori et al., 2000; Blichenberg et al., 2001), and of the non-coding BC1 RNA (Muslimov et al., 1997). Furthermore, transacting factors have been characterized that interact, at least in vitro, with the cis-acting elements within the MAP (Rehbein et al., 2000; Monshausen et al., 2001) and the VP (Mohr et al., 2001a,b) mRNAs. In the present review, we will summarize current knowledge of mechanisms involved in VP mRNA sorting to the neurites of nerve cells in vivo and in primary cultured sympathetic neurons microinjected with eukaryotic expression vectors. In hypothalamic magnocellular neurons, VP mRNA is delivered to axons and dendrites Two different populations of magnocellular neurons (MCN) of the hypothalamic paraventricular and supraoptic nuclei express the genes encoding the VP and the structurally closely related OT precursors.

213

MCNs are remarkable since both VP and OT are released from their dendrites into the central nervous system (Morris et al., 1998) in addition to being secreted into the periphery from axon terminals that project to the posterior pituitary. Apparently, VP and OT have multiple functions. Firstly, they influence in a well characterized manner the physiological state of a variety of peripheral target organs (for review see Birnbaumer, 2000; Gimpl and Fahrenholz, 2001). Secondly, they act as neurotransmitters or as modulators of the classical transmitters in the brain even though these functions are less well understood (for review see Raggenbass, 2001). Early on, we and others have noted that both VP and OT mRNAs are sorted to the axon and to the dendrites of hypothalamic MCNs. Yet, the axonal and the dendritic transcripts exhibit different characteristics implying that they are functionally not equivalent (for review see Mohr, 1999; Mohr and Richter, 2000). Interestingly, within MCNs, VP and OT transcripts exhibit a polymorphism concerning the length of their poly(A) tails: mRNAs residing in the axonal domain have a considerably shorter poly(A) tail than those located in the cell somata and in the dendrites (which are identically sized). The functional significance of this poly(A) tail shortening that must occur prior to axonal targeting of the mRNAs is not understood to date. Moreover, axonal transcripts are invariant with respect to the length of the poly(A) tail. This is a surprising feature of the axonal mRNAs because in the cell somata, the poly(A) tails of VP and OT mRNAs can be subject to ample elongation, for instance during osmotic stimulation. We have shown that mRNA transport to the axon most likely takes place subsequent to translation, but evidence for local axonal translation of both mRNA species has not been obtained since they are not associated with ribosomes (Mohr et al., 1995). So their function continues to remain unknown. The dendritic VP and OT mRNAs, in contrast, are localized to dendrites prior to translation (Mohr et al., 1995). Dendrites differ from axons both morphologically and functionally. Immunohistochemical studies at the ultrastructural level have indicated that peptide hormone precursor synthesis might occur in dendrites, at least under certain physiological circumstances. Careful analyses by Morris et al. (1998) have shown that the dendrites of MCNs contain most of the components required

for protein synthesis such as ribosomes, polysomes, rough endoplasmic reticulum (ER), and tRNA. A currently unresolved problem is the apparent lack of Golgi-like structures. Golgi marker proteins are only detectable in parts in close apposition to the cell body. What has been noticed, however, are morphological structures, tubular/vesicular organelles with associated ribosomes, that might possibly represent the Golgi apparatus-equivalent of dendrites (Morris et al., 1998). Similar observations have been reported using other nerve cell types (Tiedge and Brosius, 1996; Torre and Steward, 1996). Even though it is likely that the majority of neurosecretory vesicles originate from the cell soma it remains an appealing assumption, given the translational capacity of dendrites, that VP (and OT) are synthesized on-site in the dendrites of hypothalamic neurons. Precisely localized precursor synthesis might provide the mutual opportunity to modulate and/or fine tune defined synaptic connections within the paraventricular and supraoptic nuclei. Molecular determinants of the mRNA localization pathway Localization of mRNAs and non-coding transcripts to distinct cytoplasmic regions is conserved during evolution and is observed in organisms as diverse as yeast and mammals. By means of mRNA localization synthesis of particular proteins may be very precisely restricted to subcellular microdomains. Initially, this phenomenon has been described in developmental systems. In Drosophila oocytes and early embryos, for example, embryonic development is specified by the asymmetrical distribution of cell fate determinants. This segregation is achieved, at least in part, by localization of the respective mRNAs (for review see Bashirullah et al., 1998). Meanwhile, it is well known that RNA compartmentalization takes also place in differentiated cells, including neurons, to allow for the establishment and maintenance of cell polarity (for review see Bashirullah et al., 1998; Hazelrigg, 1998; Mohr, 1999; Carson et al., 2001). The molecular entities specifying the entire process of RNA localization appear to be remarkably complex. First, the mRNA (or the non-coding transcript) destined for transport must be selected from a large pool of RNAs. RNA localization may even begin,

214

at least in some cases, in the cell nucleus (Schnapp, 1999; Long et al., 2001). Distinct sequences of the RNA molecule serve as targets for RNA-binding proteins, the trans-acting factors. These mediate, together with additional proteins that might associate the complex, transport of the RNAs to a defined cytoplasmic location along either microfilaments or microtubules (for review see Jansen, 1999). Anchor proteins that have very recently begun to be characterized (Zhao et al., 2001) are presumably required to attach the RNAs upon arrival at their ultimate destinations. Furthermore, the RNA localization process includes mechanisms of translational control. For instance, in Drosophila oocytes and early embryos and in Xenopus oocytes translation of localized mR-

NAs is repressed as long as they are in an incorrect subcellular environment. Only at their proper destination, this translational block is removed and protein synthesis can commence (Castagnetti et al., 2000; Wilhelm et al., 2000). Unfortunately, current knowledge as to whether or not translational control plays a role in mRNA localization in nerve cells is still scarce. Yet, there is accumulating evidence to suggest that protein synthesis within dendrites is triggered by synaptic activity implying mechanisms of translational silencing and activation of mRNAs localized to the dendritic compartment (for review see Wells et al., 2000; Steward and Schuman, 2001). Based on the available information we propose a — in many aspects still hypothetical — model (Fig. 1)

Fig. 1. Hypothetical scheme of VP mRNA transport mechanisms to axons and dendrites in hypothalamic magnocellular neurons. Two different pathways exist delivering VP mRNA molecules either to the dendrites (pathway 1) or to the axonal compartment (pathway 2). Pathway 1: a current concept proposes that mRNAs may be selected for transport in the cell nucleus. A nuclear RNA-binding protein binds to a specific sequence or structural motif and accompanies the mRNA to the cytoplasm. There, additional RNA-binding proteins join the complex and fix the cargo to cytoskeletal elements (not shown in this scheme) along which mRNA targeting to dendrites occurs. During transport, mRNAs are presumed to be translationally silenced. After delivery at the ultimate destination, the ribonucleoprotein complex is reorganized in such a way as to anchor the mRNA by still ill-defined anchor proteins. Protein synthesis is initiated by unknown mechanisms. That event is likely to be triggered by synaptic activation. The recruitment of the translational machinery presumably requires additional proteins not depicted in this scheme. Pathway 2: after entering the cytoplasm, VP mRNA is first translated at the rough endoplasmic reticulum. Gradually, poly(A) tail shortening takes place, until it exhibits a length that might no longer support an efficient precursor synthesis. Then, VP mRNA is directed toward the axon, presumably along microtubules. Axonal transport may also be mediated by RNA/protein interactions. Evidence obtained by electron spectroscopic imaging techniques suggests that the axonal transcripts are indeed organized into ribonucleoprotein particles (Mohr et al., 1993). The axonal transcripts are not translated, their functional significance has remained elusive.

215

for axonal and dendritic sorting of VP (and possibly OT) mRNA. In hypothalamic MCNs obviously two different mRNA localization pathways exist. VP mRNA destined to be localized to the dendrites (pathway 1) may be marked for transport in the cell nucleus. If true, a nuclear RNA-binding protein would interact with a specific localizer sequences and accompany the mRNA to the cytoplasm. There, additional cytoplasmic proteins join the complex and fix it to the cytoskeleton, most likely to microtubules (Mohr et al., 2001b). After delivery of VP mRNA molecules at their ultimate destination(s), the ribonucleoprotein complexes are presumably reorganized in order to allow for anchoring and active translation of the hitherto translationally dormant VP mRNA. This may occur only upon certain physiological circumstances accompanied by synaptic activation. Pathway 2 begins with nuclear export and translation in the cell soma of VP mRNA. Gradually, poly(A) tail shortening takes place until it gains a length which may no longer support an efficient precursor synthesis. At that time, VP mRNA is directed toward the axonal compartment. This localization step might be mediated by proteins as well. Earlier data have suggested the involvement of microtubules in axonal VP mRNA transport in vivo (Levy et al., 1990). The functional role of the axonal transcripts is, however, virtually unknown. Identification of cis-acting elements within VP mRNA We have employed a well established and extensively characterized primary nerve cell culture system, namely SCG neurons, to delineate cis-acting sequences that mediate VP mRNA sorting to the nerve cell processes (for review see Mohr, 1999). Eukaryotic expression vectors schematically outlined in Fig. 2A have been introduced into these cells (which lack endogenous VP gene expression) by computer-assisted nuclear microinjections. These initial experiments, summarized in Fig. 2B–D, indicate that either a sequence motif or a secondary structure of the VP mRNA must contain information essential for dendritic mRNA sorting: when injections were done with a construct driving the expression of sense RNA, substantial amounts of VP transcripts were detectable in dendrites in addition to

the cell soma (Fig. 2B). VP anti-sense RNA, in contrast, remained confined to the cell bodies (Fig. 2C). Vector-expressed α-tubulin mRNA (Fig. 2D), like its endogenously transcribed counterpart, is not localized to dendrites further underscoring specificity of dendritic VP mRNA sorting. Parenthetically, sorting to the axonal domain was not observed using this detection protocol. Thus, the determinants specifying axonal mRNA transport might be more cell-specific than those required for dendritic mRNA localization. Yet, the possibility still remains that more sensitive RNA detection procedures would reveal transport of VP mRNA to the axonal compartment of SCG neurons as well. Our earlier data (see above) have suggested VP mRNA sorting to the dendrites of MCNs takes place well before translation commences. Thus, ongoing translation should not be required for the sorting process. If true, VP mRNA should continue to be delivered to dendrites while translation is blocked. As shown in Fig. 3, culturing microinjected SCG neurons in the presence of cycloheximide dramatically reduced translation of the VP mRNA but had no recognizable effect on its transport to the dendrites, a result in line with data reported by others (Kleiman et al., 1993). Subsequent analyses specified the last 395 nucleotides of the VP mRNA as being essential for an efficient dendritic mRNA transport, because this sequence, referred to as DLS, was able to confer dendritic localization to the non-localized α-tubulin mRNA used as a reporter transcript (Mohr, 1999). Notably, the DLS is composed of coding sequences in addition to the complete 3 -untranslated region (UTR). Initially, this was regarded as an unusual feature, because previously characterized RNA localization elements were always confined to the 3 -UTR. Meanwhile, however, several localized mRNAs, in Drosophila and in yeast, have been shown to accommodate targeting information within the proteinencoding moiety of the mRNA molecule (for review see Gonzalez et al., 1999; Mohr and Richter, 2001). More recently, we have been able to show that the DLS contains several subelements, weak localizer elements on their own, only the concerted action of which mediates an efficient dendritic sorting of reporter mRNA molecules comparable to that achieved by the VP mRNA alone (Mohr, 1999).

216

Fig. 2. VP mRNA is sorted to the dendrites of primary cultured SCG neurons. (A) Schematic view of eukaryotic expression vectors microinjected into the cell nuclei of in vitro cultured SCG neurons. The expression of any inserted cDNA is driven by the cytomegalovirus (CMV) promoter. A short sequence of the bacterial β-galactosidase (β-gal) gene was included such that it forms part of the 5 -UTR of the resulting transcripts. This allows to discriminate mRNAs that are endogenously expressed in SCG neurons from those derived by transcription of the expression vector by performing in situ hybridizations with β-gal anti-sense oligodeoxyribonucleotides. Addition of a poly(A) tail is mediated by the bovine growth hormone (BGH) gene poly(A) addition sequence. (B–D) In situ hybridization analyses of cells microinjected with three different expression vector constructs. (B) Injection of a construct containing the rat VP cDNA inserted in sense orientation leads to transport of the mRNA into proximal and distal parts of the dendrites (arrows). Labeling of the axon has not been observed. (C) A cell is shown that expresses VP anti-sense transcripts. In this case, sorting out of the cell somata does not occur (arrowheads). (D) The vector-derived mRNA encoding α-tubulin is confined to the cell somata (arrowheads). Microinjected cells have been processed for in situ hybridization approximately 18 h following injections. For experimental details see (Prakash et al., 1997). From Mohr et al., 2001b. Reproduced with permission.

Mutations within the DLS impair its dendritic targeting capacity The DLS contains two motifs rich in pyrimidine nucleotides termed E1 and E2 (Fig. 4A). Short repetitive motifs are involved in the subcellular transport of some mRNAs in non-neuronal cells (Mowry and Cote, 1999). In order to determine if E1 and/or E2 have an impact on VP mRNA sorting to the nerve cell processes the elements have been deleted from the DLS either individually or together. The mutagenized DLSs were inserted as part of the 3 -UTR of α-tubulin cDNA cloned into a eukaryotic expression vector which is schematically depicted in Fig. 4B. In line with earlier observations (Mohr, 1999) the majority of SCG neurons microinjected with the construct containing the wild-type DLS (construct 1: pCMV β-gal α-tub/VP-DLS) harbor the chimeric

mRNA in the dendritic domain, predominantly in proximal and distal dendritic segments (Fig. 5A,B and Fig. 6). Deletion of element E1 (construct 2: pCMV β-gal α-tub/VP-DLSE1) reduced dendritic targeting of the recombinant mRNA substantially even though it was not completely abolished. In most cells, however, the mRNA was confined to the cell somata (Fig. 5C,D and Fig. 6). Only residual dendritic targeting was observed in the remainder of cells with localization of chimeric transcripts primarily in basal and proximal dendritic segments. Upon deletion of element E2 (construct 3: pCMV β-gal α-tub/VP-DLSE2) a somewhat weaker but still remarkable reduction of the dendritic localizer capacity was noted (Fig. 5E,F and Fig. 6). Concomitant deletion of both elements (construct 4: pCMV βgal α-tub/VP-DLSE1 + E2) had no further effect on the subcellular RNA distribution pattern which

217

Fig. 3. Dendritic VP mRNA transport occurs independently of ongoing translation. In vitro cultured SCG neurons were injected with a eukaryotic expression vector (see Fig. 2A) containing the rat VP cDNA in sense orientation. Following overnight incubation, cells were subjected to in situ hybridization with a digoxigenin-labeled VP-specific anti-sense riboprobe. Probe detection was done immunocytochemically by sequentially adding a sheep anti-digoxigenin antibody and a biotinylated donkey anti-sheep antibody, respectively. The hybridization signal was finally amplified using the Renaissance TSA-indirect system (NEN) with Neutravidin-conjugated Alexa 488 (Molecular Probes) as a fluorochrome. Concomitantly, the neurophysin II (NP II) moiety of the VP precursor was visualized using a rabbit antineurophysin II-antiserum followed by Cy3-conjugated donkey anti-rabbit immunoglobulin. Micrographs were taken using the appropriate filter sets to allow for discrimination between the two fluorochromes. (A–C) Control SCG neuron cultured in the absence of cycloheximide; (D–F) SCG neuron cultured in the presence of 40 μM cycloheximide for the entire period of the experiment. The drug was added to the culture media 1 h prior to the injection. (A,D) micrographs showing the distribution of VP mRNA; (B,E) micrographs showing the distribution of neurophysin II; (C,F) phase contrast micrographs. ICC, immunocytochemistry; ISH, in situ hybridization. Scale bar: 50 μm.

Fig. 4. Deletion of elements rich in pyrimidine nucleotides within the dendritic localizer sequence of the VP mRNA (VP-DLS) reduces its capacity to confer dendritic localization to the non-localized α-tubulin (α-tub) RNA. (A) Schematic representation of the 395-bp VP-DLS (nucleotide residues 190–584) which contains two highly homologous elements, E1 and E2, enriched in pyrimidine nucleotides. These elements were deleted either alone or together. (B) The wild-type and the mutagenized DNA fragments, respectively, were individually ligated in sense orientation to the 3 -end of the rat α-tubulin cDNA cloned into a eukaryotic expression vector. Expression of the chimeric mRNAs is driven by the cytomegalovirus (CMV) promoter. In addition, the vector contains a 50-bp fragment derived from the bacterial β-galactosidase (β-gal) gene. This sequence forms part of the 5 -untranslated region of vector-expressed transcripts. It is not relevant for the data presented here (for experimental details see Prakash et al., 1997). BGH, bovine growth hormone.

218

219

resembled that seen with pCMV β-gal α-tub/VPDLSE1 (Fig. 5G,H and Fig. 6). Obviously, deletions within the DLS are no longer compatible with a coordinated interplay of the presumably complex and multifactorial transport machinery overall resulting in a reduced dendritic transport. These results are not surprising. Cis-acting mRNA transport signals are rather complex usually consisting of several hundreds of nucleotides (reviewed in Bashirullah et al., 1998). As outlined above, the same holds true for the DLS of the VP mRNA. The DLS might contain several conserved sequence elements individually recognized by proteins of the localizing machinery. Alternatively, the interaction of trans-acting factors could primarily depend on the secondary or tertiary structure adopted by the RNA molecule. The DLS contains one copy each of the motifs E1 and E2 in its proximal and its distal half, respectively. On their own, both of these DLS-subsegments are able to direct the non-localized α-tubulin mRNA to dendrites, but only to regions proximal to the cell body (Prakash et al., 1997). According to our definition, these segments would represent weak localizer elements. Interestingly, short reiterated pyrimidine nucleotide-rich elements play a critical role in sorting Vg1 mRNA to the vegetal hemisphere in Xenopus oocytes. Upon deletion, Vg1 mRNA localization is either abolished or impaired, depending on the motif under investigation (Mowry and Cote, 1999). Deletion of E1 and E2 within the DLS likewise has a severe impact on dendritic mRNA transport, but a residual activity could still be observed. Consequently, E1 and E2 are not the sole dendritic targeting elements. Besides E1 and E2, other primary sequences might exist which harbor the residual dendritic localization capacity displayed by the mutagenized sequences. Alternatively, by deleting

these elements the DLS secondary or tertiary structure could have been altered such as to disturb the interplay with the transport machinery considerably since higher order structure of an RNA molecule is determined by its primary sequence (SenGupta et al., 1996). The data presented here do not allow to discriminate between these alternatives but they clearly show that the DLS of the VP mRNA is organized in a complex way as are the targeting elements of localized transcripts in diverse organisms. Rat poly(A)-binding protein interacts specifically with the DLS of the VP mRNA In non-neuronal systems, and probably in nerve cells as well, proteins play a pivotal role in any of the multiple events of mRNA localization (Fig. 1). By using UV-crosslinking assays with rat brain cytosolic protein extracts and radiolabeled RNA probes we have recently characterized a protein that interacts specifically with the DLS of the VP mRNA but not with its 5 -end that lacks a role in dendritic VP mRNA targeting (Mohr et al., 2001a). The identity of this protein, initially termed VP-RBP (VP mRNA-binding protein), was revealed by biochemical purification (Mohr et al., 2001b) showing that it is the well known and multifunctional PABP. As demonstrated in Fig. 7A, PABP, while specifically interacting with VP mRNA, fails to bind to a variety of other transcripts including the dendritic targeting element of the MAP2 transcript and α-tubulin mRNA. Since PABP is a ubiquitous and an extremely abundant protein (Görlach et al., 1994) it is very surprising to find the highest binding activity to the VP mRNA in brain extracts whereas peripheral tissues and non-neuronal cell lines contain very little if any binding activities (Fig. 7B). A possible

Fig. 5. Micrographs showing primary cultured SCG neurons microinjected into the cell nuclei with eukaryotic expression vector constructs containing the wild-type VP-DLS or mutagenized sequences thereof (schematically outlined in Fig. 4). (A,B) Microinjection of construct 1 (pCMV βgal αtub/VP-DLS) containing the wild-type VP-DLS. (C,D) Microinjection of construct 2 (pCMV βgal αtub/VP-DLSE1), lacking element E1. (E,F) Microinjection of construct 3 (pCMV βgal αtub/VP-DLSE2) lacking element E2. (G,H) Microinjection of construct 4 (pCMV βgal αtub/VP-DLSE1 + E2) lacking both element E1 and element E2. The subcellular distribution of the chimeric RNAs was determined by in situ hybridization with a digoxigenin-labeled in vitro synthesized VP anti-sense riboprobe. The wild-type VP-DLS mediates RNA transport to distal parts of the dendrites (arrows). In contrast, chimeric RNAs containing either of the mutagenized VP sequences are detectable, in very low amounts, in basal and proximal dendritic segments (arrows) in the majority of the injected cells (for experimental details see Prakash et al., 1997). αtub, α-tubulin; βgal, β-galactosidase; DLS, dendritic localizer sequence; VP, vasopressin. Scale bar: 25 μm.

220

Fig. 6. Evaluation of distribution patterns of chimeric transcripts in primary cultured SCG neurons microinjected with constructs schematically shown in Fig. 4. All expression vector constructs have been injected in at least 10 different experiments with embryonic cell cultures prepared from different pregnant females (800–1000 cells/construct). The relative number of cells injected with the four constructs that harbor the chimeric mRNAs in the cell body only, or in addition in basal, proximal and distal parts of the dendrites, schematically outlined in A, are graphically summarized in B.

explanation to this obvious discrepancy would be to assume that PABP’s binding specificity to VP mRNA is determined by additional parameters. For example, a covalent modification or the association with additional proteins to form a larger complex could alter PABP’s sequence selection properties. Apparently, this ‘specificity factor’, the nature of which remains to be determined, is brain specific. The literature on proteins that might play a role in mRNA transport to the neurites of nerve cells is still rather scarce. Recently, however, trans-acting factors have been characterized that interact with the

dendritic targeting element of the MAP2 mRNA, an mRNA species localized to the dendrites of various nerve cell types (for review see Kindler et al., 2001). Two proteins, termed MARTA1 (Rehbein et al., 2002) and MARTA2 (the identity is not yet known), appear to associate selectively with MAP2 transcripts because they neither bind to the rat VP mRNA nor to other transcripts known to be sorted to dendrites (Rehbein et al., 2000). The third protein found to interact with MAP2 mRNA is the rat Staufen protein (Monshausen et al., 2001). Even though Staufen binds to any RNA in vitro, its

221

species such as bicoid, oskar and prospero transcripts (for review see Bashirullah et al., 1998). Rat Staufen is expressed in many tissues, interestingly, however, in neurons it shows a somato-dendritic distribution suggesting that it might be involved in dendritic mRNA localization (for review see Kiebler and DesGrosseilliers, 2000). Taken together, the available information supports the view that the molecular determinants required for localizing different mRNA species to dendrites of nerve cells appear to be surprisingly specific for a given transcript. These findings are not completely unexpected, though, given the different temporal and spatial distribution patterns of mRNAs that have been observed in nerve cells (for review see Mohr, 1999). Proposed function of PABP in VP mRNA localization

Fig. 7. PABP binds specifically to the DLS of the VP mRNA. (A) Autoradiogram of UV-crosslinking analyses performed with 7.5 μg rat brain cytosolic protein extract and 5 fmol of the radiolabeled DLS riboprobe (lacking a poly(A) tail). Unlabeled competitor RNAs were added at a 100-fold molar excess. Lane 1, no competitor; lane 2, full size VP RNA; lane 3, dendritic targeting element of rat MAP2 mRNA; lane 4, full-size rat αtubulin RNA. The positions of molecular size marker proteins is indicated on the right. The arrow denotes the PABP/VP RNA complex. All competitor RNAs represent the sense strands; none of the transcripts possess poly(A) tails. (B) Autoradiogram of UV-crosslinking analyses performed with 7.5 μg each of various rat tissue/cell line cytosolic protein extracts and 5 fmol of the radiolabeled DLS riboprobe (lacking a poly(A) tail). Proteins were prepared from: lane 1, total brain; lane 2, hypothalamus; lane 3, heart; lane 4, lung; lane 5, spleen; lane 6, liver; lane 7, rat I cells; lane 8, PC 12 cells. The positions of molecular size marker proteins is indicated on the right. The arrow denotes the PABP/VP RNA complex. For experimental details see (Mohr et al., 2001a). comp., competitor; hypoth., hypothalamus; MAP2, microtubule-associated protein 2; PABP, poly(A)-binding protein; tot. brain, total brain; tub., tubulin; VP, vasopressin. From Mohr et al., 2001b. Reproduced with permission.

in vivo binding properties are remarkably specific. In Drosophila oocytes and neuroblasts, Staufen is involved in subcellular targeting of defined mRNA

PABP is an evolutionarily conserved RNA-binding protein. It consists of four RNA recognition motifs (RRM; each 80–100 amino acids in length) at the Nterminus of the protein and a C-terminal proline-rich auxiliary domain (for review see Burd and Dreyfuss, 1994). It exhibits high affinity binding to the poly(A) tail of mRNAs, mainly via RRMs 1 + 2, and this interaction is essential for translational efficiency. PABP enhances translation initiation by still unknown mechanisms via its interaction with initiation factors bound at the 5 -end of the mRNA (Preiss et al., 1998; Fig. 8A). Moreover, it is required for translation-dependent mRNA stabilization (Coller et al., 1998). In vitro binding assays performed with either individual RRMs or various combinations of RRMs revealed remarkable features. While single RRMs do not interact with poly(A) sequences at all, RRMs 1 + 2 bind with high affinity indistinguishable of that displayed by the full-size protein. In contrast, the affinity of RRMs 3 + 4 for poly(A) sequences is much lower. In fact, these RRMs bind more efficiently to poly(U) and poly(G) than to poly(A) sequences (Burd et al., 1991; Kühn and Pieler, 1996). Notably, PABP is essential for cell viability in yeast. Yet, critical function(s) are displayed by RRM4 together with part of the C-terminus rather than by RRMs 1 + 2 the removal of which still supported growth (Burd et al., 1991). Consequently, the RRMs of PABP are functionally diverse and its most impor-

222

Fig. 8. Functional roles of PABP in translation. (A) PABP, bound in multiple copies to the poly(A) tail of mRNAs, plays a major role in translation by stimulating the initiation of protein biosynthesis. PABP interacts with eIF4G which forms, together with the cap-binding protein eIF4E and eIF4A, an ATP-dependent RNA helicase, a ternary complex called eIF4F. Thus, the 3 -end is brought in close proximity to the 5 -end of the mRNA. This is believed to facilitate joining of the ribosomal subunits that are liberated at the stop-codon of the mRNA. (B) PABP bound to the DLS of the VP mRNA (VP-DLS) could possibly play a role in its translational silencing. By interaction of PABP bound to the DLS with those PABP molecules bound to the poly(A) tail (either directly by homophilic interactions or via an additional protein factor) translational initiation could be inhibited, because this interaction would interfere with the interaction of the poly(A) tail-bound PABP with eIF4G at the 5 -end of the VP mRNA. AUG, translational start codon; cap, cap structure at the 5 -end of the mRNA; DLS, dendritic localizer sequence; eIF, eukaryotic initiation factor; PABP, poly(A)-binding protein; poly(A), poly-adenylate; VP, vasopressin.

tant feature is clearly not the high affinity binding to poly(A) sequences. Earlier estimates have suggested PABP to be present at a concentration exceeding binding sites on poly(A) tails of mRNAs approximately three-fold (Görlach et al., 1994). Since it is able to interact with non-poly(A) sequences, at least in vitro (Görlach et al., 1994; Kühn and Pieler, 1996), PABP presumably has multiple functions in mRNA metabolism. This view is supported by recently published data. For instance, PABP expression is regulated at the translational level (Wu and Bag, 1998; Hornstein et al., 1999). Translational silencing of PABP mRNA appears to be brought about by a specific interaction of PABP itself with sequences in the 5 -UTR (De Melo Neto et al., 1995; Bag and Wu, 1996). Moreover, PABP, in concert with additional proteins, is involved in controlling the translationdependent turnover of the c-fos mRNA (Grosset et al., 2000). Thus, it is conceivable that it may also be implicated in regulating the translational state of the VP mRNA. Ever-growing evidence supports the current view that mRNAs residing in dendrites are translationally silenced until external stimuli activate protein synthesis (for review see Marin et al., 1997; Schuman, 1999; Steward and Schuman, 2001). Downregulation of dendritcally localized VP mRNA translation by PABP could, for example, be brought

about by its binding the DLS. The DLS-bound PABP could, either directly or via additional factors, interact with PABP molecules on the poly(A) tail. As a net result, this could effectively interfere with translational stimulation because it would prevent the poly(A) tail-bound PABP interaction with translational initiation factors at the 5 -end of the mRNA (Fig. 8B). Beyond doubt, we are currently still missing many of the components necessary for dendritic (and axonal) VP mRNA trafficking. Presumably, PABP is only one member of a multi-factorial complex. Yet, by applying techniques such as the yeast two hybrid system with PABP as a bait and immunoprecipitation once an anti-rat PABP antiserum is available we should be able to identify and characterize additional constituents of the sorting machinery. This will pave the way to gain a deeper understanding of the RNA localization pathways in nerve cells. Abbreviations BGH CMV DLS ER MAP2

bovine growth hormone cytomegalovirus dendritic localizer sequence endoplasmic reticulum microtubule-associated protein 2

223

MCN OT PABP RRM SCG UTR VP VP-RBP

magnocellular neuron oxytocin poly(A)-binding protein RNA recognition motif superior cervical ganglion untranslated region vasopressin VP RNA-binding protein

Acknowledgements This work is supported by the Volkswagenstiftung and the Deutsche Forschungsgemeinschaft (To D.R. and E.M.). Part of this work forms the doctoral thesis of I.K. References Bag, J. and Wu, J. (1996) Translational control of poly(A) binding protein expression. Eur. J. Biochem., 237: 143–152. Barbarese, E., Brumwell, C., Kwon, S., Cui, H. and Carson, J.H. (1999) RNA on the road to myelin. J. Neurocytol., 28: 263–270. Bashirullah, A., Cooperstock, R.L. and Lipshitz, H.D. (1998) RNA localization in development. Annu. Rev. Biochem., 67: 335–394. Birnbaumer, M. (2000) Vasopressin receptors. Trends Endocrinol. Metab., 11: 406–410. Burd, C.G. and Dreyfuss, G. (1994) Conserved structures and diversity of functions of RNA-binding proteins. Science, 265: 615–621. Burd, C.G., Matunis, E.L. and Dreyfuss, G. (1991) The multiple RNA-binding domains of the mRNA poly(A)-binding protein have different RNA-binding activities. Mol. Cell Biol., 11: 3419–3424. Blichenberg, A., Schwanke, B., Rehbein, M., Garner, C.C., Richter, D. and Kindler, S. (1999) Identification of a cis-acting dendritic targeting element in MAP2 mRNAs. J. Neurosci., 19: 8818–8829. Blichenberg, A., Rehbein, M., Muller, R., Garner, C.C., Richter, D. and Kindler, S. (2001) Identification of a cis-acting dendritic targeting element in the mRNA encoding the alpha subunit of Ca2+ /calmodulin-dependent protein kinase II. Eur. J. Neurosci., 13: 1881–1888. Carson, J.H., Cui, H., Krueger, W., Schlepchenko, B., Brumwell, C. and Barbarese, E. (2001) RNA trafficking in oligodendrocytes. In: D. Richter (Ed.), Results and Problems in Cell Differentiation: Cell Polarity and Subcellular RNA Localization, Vol. 34. Springer-Verlag, Heidelberg, pp. 69–81. Castagnetti, S., Hentze, M.W., Ephrussi, A. and Gebauer, F. (2000) Control of oskar mRNA translation by Bruno in a novel cell-free system from Drosophila ovaries. Development, 127: 1063–1068.

Coller, J.M., Gray, N.K. and Wickens, M.P. (1998) mRNA stabilization by poly(A) binding protein is independent of poly(A) tail and requires translation. Genes Dev., 12: 3226–3235. De Melo Neto, O.P., Standart, N. and de Sa, C.M. (1995) Autoregulation of poly(A)-binding protein synthesis in vitro. Nucleic Acids Res., 23: 2198–2205. Gimpl, G. and Fahrenholz, F. (2001) The oxytocin receptor system: structure, function, and regulation. Physiol. Rev., 81: 629–683. Görlach, M., Burd, C.G. and Dreyfuss, G. (1994) The mRNA poly(A)-binding protein: localization, abundance, and RNAbinding specificity. Exp. Cell Res., 211: 400–407. Gonzalez, I., Buonomo, S.B.C., Nasmyth, K. and von Ahsen, U. (1999) ASH1 mRNA localization in yeast involves multiple secondary structural elements and Ash1 protein translation. Curr. Biol., 9: 337–340. Grosset, C., Cheu, C.A., Xu, N., Sonenberg, N., JaqueminSablou, H. and Shyu, A. (2000) A mechanism for translationally coupled mRNA turnover: interaction between the poly(A) tail and a c-fos RNA coding determinant via a protein complex. Cell, 103: 29–40. Hazelrigg, T. (1998) The destinies and destinations of RNAs. Cell, 95: 451–460. Hornstein, E., Git, A., Braunstein, I., Avni, D. and Meyuhas, O. (1999) The expression of poly(A)-binding protein gene is translationally regulated in a growth-dependent fashion through a 5 terminal oligopyrimidine tract motif. J. Biol. Chem., 274: 1708–1714. Jansen, R.-P. (1999) RNA-cytoskeletal associations. FASEB J., 13: 455–466. Kiebler, M.A. and DesGrosseilliers, L. (2000) Molecular insights into mRNA transport and local translation in the mammalian nervous system. Neuron, 25: 19–28. Kindler, S., Mohr, E., Rehbein, M. and Richter, D. (2001) Extrasomatic targeting of MAP2, vasopressin and oxytocin mRNAs in mammalian neurons. In: D. Richter (Ed.), Results and Problems in Cell Differentiation: Cell Polarity and Subcellular RNA Localization, Vol. 34. Springer-Verlag, Heidelberg, pp. 83–104. King, M.L., Zhou, Y. and Bubunenko, M. (1999) Polarizing genetic information in the egg: RNA localization in the frog oocyte. BioEssays, 21: 546–557. Kleiman, R., Banker, G. and Steward, O. (1993) Inhibition of protein synthesis alters the subcellular distribution of mRNA in neurons but does not prevent dendritic transport of RNA. Proc. Natl. Acad. Sci. USA, 90: 11192–11196. Kühn, U. and Pieler, T. (1996) Xenopus poly(A) binding protein: functional domains in RNA binding and protein–protein interaction. J. Mol. Biol., 256: 20–30. Levy, A., Lightman, S.L. and Carter, D. (1990) The origin and regulation of posterior pituitary vasopressin ribonucleic acid in osmotically stimulated rats. J. Neuroendocrinol., 2: 329–334. Lipshitz, H.D. and Smibert, C.A. (2000) Mechanisms of RNA localization and translational regulation. Curr. Opin. Gen. Dev., 10: 476–488. Long, R.M., Gu, W., Meng, X., Gonsalvez, G., Singer, R.H. and Chartrand, P. (2001) An exclusively nuclear RNA-binding

224 protein affects asymmetric localization of ASH1 mRNA and Ash1p in yeast. J. Cell Biol., 153: 307–318. Marin, P., Nastiuk, K.L., Daniel, N., Girault, J.-A., Czernik, A.J., Glowinski, J., Nairn, A.C. and Prémont, J. (1997) Glutamatedependent phosphorylation of elongation factor-2 and inhibition of protein synthesis in neurons. J. Neurosci., 17: 3445– 3454. Mohr, E. (1999) Subcellular RNA compartmentalization. Prog. Neurobiol., 57: 507–525. Mohr, E. and Richter, D. (2000) Axonal mRNAs: functional significance in vertebrates and invertebrates. J. Neurocytol., 29: 787–795. Mohr, E. and Richter, D. (2001) Messenger RNA on the move: implications for cell polarity. Int. J. Biochem. Cell Biol., 33: 669–679. Mohr, E., Terjung, D., Martin, R. and Richter, D. (1993) Vasopressin mRNA in the hypothalamo-hypophyseal tract. In: P. Gross, D. Richter and G.L. Robertson (Eds.), Vasopressin. John Libbey Eurotext, Paris, pp. 119–129. Mohr, E., Morris, J.F. and Richter, D. (1995) Differential subcellular mRNA targeting: deletion of a single nucleotide prevents the transport to axons but not to dendrites of rat hypothalamic magnocellular neurons. Proc. Natl. Acad. Sci. USA, 92: 4377– 4381. Mohr, E., Fuhrmann, C. and Richter, D. (2001a) VP-RBP, a protein enriched in brain tissue, specifically interacts with the dendritic localizer sequence of the rat vasopressin mRNA. Eur. J. Neurosci., 13: 1107–1112. Mohr, E., Prakash, N., Vieluf, K., Fuhrmann, C., Buck, F. and Richter, D. (2001b) Vasopressin mRNA localization in nerve cells: characterization of cis-acting elements and trans-acting factors. Proc. Natl. Acad. Sci. USA, 98: 7072–7079. Monshausen, M., Putz, U., Rehbein, M., Schweizer, M., DesGroseillers, L., Kuhl, D., Richter, D. and Kindler, S. (2001) Two rat brain Staufen isoforms differentially bind RNA. J. Neurochem., 76: 155–165. Mori, Y., Imaizumi, K., Katayama, T., Yoneda, T. and Tohyama, M. (2000) Two cis-acting elements in the 3 untranslated region of α-CaMKII regulate its dendritic targeting. Nat. Neurosci., 3: 1079–1084. Morris, J.F., Budd, T.C., Epton, M.J., Ma, D., Pow, D.V. and Wang, H. (1998) Functions of the perikaryon and dendrites in magnocellular vasopressin-secreting neurons: new insights from ultrastructural studies. Prog. Brain Res., 119: 21–30. Mowry, K.L. and Cote, C.A. (1999) RNA sorting in Xenopus oocytes and embryos. FASEB J., 13: 435–445. Muslimov, I.A., Santi, E., Homel, P., Perini, S., Higgins, D. and Tiedge, H. (1997) RNA transport in dendrites: a cis-acting targeting element is contained within neuronal BC1 RNA. J. Neurosci., 17: 4722–4733. Prakash, N., Fehr, S., Mohr, E. and Richter, D. (1997) Dendritic localization of rat vasopressin mRNA: ultrastructural analysis and mapping of targeting elements. Eur. J. Neurosci., 9: 523– 532.

Preiss, T., Muckenthaler, M. and Hentze, M.W. (1998) Poly(A)tail-promoted translation in yeast: implications for translational control. RNA, 4: 1321–1331. Raggenbass, M. (2001) Vasopressin- and oxytocin-induced activity in the central nervous system: electrophysiological studies using in-vitro systems. Prog. Neurobiol., 64: 307–326. Rehbein, M., Kindler, S., Horke, S. and Richter, D. (2000) Two trans-acting rat-brain proteins, MARTA1 and MARTA2, interact specifically with the dendritic targeting element in MAP2 mRNAs. Mol. Brain Res., 79: 192–201. Rehbein, M., Wege, K., Buck, F., Schweizer, M., Richter, D. and Kindler, S. (2002) Molecular characterization of MARTA1, a protein interacting with the dendritic targeting element of MAP2 mRNAs. J. Neurochem., in press. Schnapp, B. (1999) RNA localization: a glimpse of the machinery. Curr. Biol., 9: R725–R727. Schuman, E. (1999) mRNA trafficking and local protein synthesis at the synapse. Neuron, 23: 645–648. SenGupta, D.J., Zhang, B., Kraemer, B., Pochart, P., Fields, S. and Wickens, M. (1996) A three-hybrid system to detect RNA–protein interactions in vivo. Proc. Natl. Acad. Sci. USA, 93: 8496–8501. Steward, O. and Schuman, E.M. (2001) Protein synthesis at synaptic sites on dendrites. Annu. Rev. Neurosci., 24: 299– 325. Steward, O. and Worley, P. (2001) Localization of mRNAs at synaptic sites on dendrites. In: D. Richter (Ed.), Results and Problems in Cell Differentiation: Cell Polarity and Subcellular RNA Localization, Vol. 34. Springer-Verlag, Heidelberg, pp. 1–26. Tiedge, H. and Brosius, J. (1996) Translational machinery in dendrites of hippocampal neurons in culture. J. Neurosci., 15: 7171–7181. Tiedge, H., Bloom, F.E. and Richter, D. (1999) RNA, whither goest thou?. Science, 283: 186–187. Torre, E.R. and Steward, O. (1996) Protein synthesis within dendrites: glycosylation of newly synthesized proteins in dendrites of hippocampal neurons in culture. J. Neurosci., 16: 5967–5978. Wells, D.G., Richter, J.D. and Fallon, J.R. (2000) Molecular mechanisms for activity-regulated protein synthesis in the synapto-dendritic compartment. Curr. Opin. Neurobiol., 10: 132–137. Wilhelm, J.E., Vale, R.D. and Hedge, R.S. (2000) Coordinate control of translation and localization of Vg1 mRNA in Xenopus oocytes. Proc. Natl. Acad. Sci. USA, 97: 13132–13137. Wu, J. and Bag, J. (1998) Negative control of the poly(A)binding protein mRNA translation is mediated by the adeninerich region of its 5 -untranslated region. J. Biol. Chem., 273: 34535–34542. Zhao, W., Jiang, C., Kroll, T.T. and Huber, P.W. (2001) A proline-rich protein binds to the localization element of Xenopus Vg1 mRNA and to ligands involved in actin polymerization. EMBO J., 20: 2315–2325.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 17

Dendritic action potentials in magnocellular neurons Jaideep S. Bains * Department of Physiology and Biophysics, Neuroscience Research Group, University of Calgary, Calgary AB, Canada

Abstract: In the central nervous system, information is traditionally thought to flow from synapses to dendrites to soma. Recent evidence, however, suggests that dendrites play more of an active role in signal processing than previously thought. This review will examine the evidence in support of dendritic spikes in magnocellular neurons. Additionally, it will shed light on a number of important roles these spikes may play in regulating the excitability of magnocellular neurons. Keywords: Oxytocin; Vasopressin; Voltage-gated channel; Calcium; Retrograde inhibition

Introduction A putative function for neuronal dendrites was first formulated by Ramon y Cajal who suggested they served to propagate impulses towards the soma. Over 50 years later this concept was formalized, stating in effect that dendrites served as cable-like structures that summed excitatory and inhibitory inputs and delivered the net sum to the soma in order to generate an action potential (Rall, 1959). Shortly after this time, the hypothesis was put forward that dendrites may also amplify or boost synaptic signals (Loronté de Nó and Coundouris, 2001). Although this was a major shift in how neuroscientists thought about the dendrite, the experimental evidence in support of this idea was still lacking. In 1961, Spencer and Kandel described an electrophysiological phenomenon in hippocampal pyramidal cells that they dubbed ‘fast pre-potentials’ (FPPs) (Spencer and Kandel, 1961). These discrete events, which preceded an action potential likely originated, according to the authors, in distant dendrites in areas they referred to as ‘trigger zones’. They surmised

∗ Correspondence to: J.S. Bains, Department of Physiology and Biophysics, Neuroscience Research Group, University of Calgary, Calgary AB, Canada.

these events were true spikes since they did not exhibit a graded response as is often observed with excitatory postsynaptic potentials, but rather were all-or-none, much like the classical Na+ action potentials. Since this seminal observation, our knowledge of dendrites has grown extensively (recent reviews see: Yuste and Tank, 1996; Magee et al., 1998; Segev and Rall, 1998; Johnston et al., 1999; Hausser et al., 2000; Segev and London, 2000), but it is safe to say that aside from the electrophysiology afficianados, relatively few neuroscientists truly grasp the fundamental role that dendrites play in dictating the way in which information is transduced in the central nervous system (CNS). In the magnocellular neurons, for example, we appreciate that the dendrites contain and indeed release the neuropeptides vasopressin (VP) and oxytocin (OT) (Pow and Morris, 1989) (for review see: Ludwig, 1998), and that these peptides have a profound effect on cellular activity (Poulain and Wakerley, 1982; Yamashita et al., 1987; Moos and Richard, 1989; Kombian et al., 1997, 2000; Moos et al., 1998; Hirasawa et al., 2001) yet we still do not fully appreciate how these interactions are initiated. The primary reason for this dearth of knowledge is that our understanding of how the nervous system works is clouded by our soma-centric perspective. Although the vast majority of the CNS is a tangle of dendrites and axons, neurophysiologists have

226

focused primarily on the cell bodies. To a large degree this stems from the experimental inaccessibility of dendrites and axons. Although in some structures, the hippocampus and cerebellum in particular, a number of groups have succeeded in the heroic feat of obtaining ‘blind’ recordings from dendrites (Spencer and Kandel, 1961; Llinas and Nicholson, 1971; Llinas and Hess, 1976; Llinas and Sugimori, 1980). As a result of recent advances in microscopy, it is now possible to obtain electrical recordings directly from the dendrites of CNS neurons on a relatively consistent basis (Stuart et al., 1993). The information gained from these experiments has confirmed what many had intuitively thought: The dendrites are dynamic structures that are far more than a simple cable linking the synapse to the cell body. Virtually all of these data has been obtained from pyramidal cells in either the neocortex or hippocampus where the architecture of these cells makes both the dendrites and the cell body readily accessible for the experimenter. Unfortunately, this is not the case with the magnocellular neurons of the hypothalamus, where dendrites are interwoven with cell bodies and axons, making intradendritic recording considerably more challenging. This means that we are only starting to scratch the surface of truly understanding just how important the role of dendrites may be in the life of a magnocellular neuron. This review will focus on one particular aspect of dendritic function, the dendritic action potential. I will cover the brief history of dendritic spikes in magnocellular neurons and then explore the potential source of these spikes. Finally, I will discuss a number of putative roles for dendritic spikes that may be unique to magnocellular neurons. I will not dwell on a number of important functions mediated by dendritic spikes, such as boosting synaptic signals or providing a substrate for a Hebbian form of plasticity. These topics have been covered extensively by several excellent reviews (Magee et al., 1998; Hausser et al., 2000; Segev and London, 2000). I will instead focus on the potential contributions of dendritic spikes to three aspects that are important to the physiology of magnocellular neurons: burst firing, dendritic release of neuromodulators (VP, OT and nitric oxide (NO)) and resistance to excitotoxicity. The recent explosion of work that has brought the dendrite to the forefront of neuroscience owes

much of its gratitude to the improvements in microscope optics. The use of an upright microscope with differential interference contrast (DIC) optics in combination with a high numerical aperture, long working distance, water immersion objective, has made it possible to make direct recordings from these structures (Stuart et al., 1993). Based on these recordings, we can now say definitively that in addition to ligand-gated receptors, dendrites are also endowed with a rich repertoire of voltage-gated conductances. Much of this work has been extensively reviewed elsewhere and the reader is referred to several excellent recent reviews on various topics related to dendritic physiology (Magee et al., 1998; Johnston et al., 1999; Hausser et al., 2000; Segev and London, 2000). Without the benefit of dendritic recordings, but with the use of a little deductive reasoning, it can be argued that dendrites in magnocellular neurons also possess a number of voltage-gated conductances. The first observation that, in hindsight, hinted at the dynamic dendrite in magnocellular neurons was the presence of TTX-insensitive action potentials (Andrew and Dudek, 1983; Mason and Leng, 1984). These spikes were not due to activation of a novel, TTX-insensitive Na+ conductance, but rather were mediated by Ca2+ ions (Mason and Leng, 1984; Bourque and Renaud, 1985). Slower, long-lasting Ca2+ -mediated depolarizing potentials had been described in fetal mouse hypothalamic neurons (Legendre et al., 1982), but their relationship to these brief, transient events was unclear. (In hindsight, it seems these events, which were observed when K+ conductances were blocked may have been prolonged dendritic depolarizations (see below).) Ca2+ spikes had been described previously in the dendrites of cerebellar Purkinje cells (Llinas and Hess, 1976) and a similar site of initiation was proposed for these events in magnocellular neurons. This was based on the observation that a dual spike waveform was present when extracellular recordings were obtained from magnocellular neurons, but not when intracellular recordings were made exclusively from the soma (Mason and Leng, 1984). Additionally, it was reported that magnocellular neurons exhibited non-synaptic depolarizing potentials (NSDPs) which were small spike-like events that were sensitive to extracellular Ca2+ (Bourque et al., 1986). The final piece of ev-

227

idence supporting a dendritic site of initiation was the demonstration that intense activation of NMDA receptors, which are located on the dendrites of magnocellular neurons, resulted in the activation of TTXinsensitive, Ca2+ -mediated spikes (Bains and Ferguson, 1999). This provided the first direct evidence that transduction of a synaptically generated EPSP to the soma may not just be a simple case of propagating an ever-decaying signal down a long, narrow tube. If we cannot record from the dendrites, how do we really know the spikes are dendritic in origin? Although the work cited above is suggestive of a dendritic origin for the Ca2+ spikes, a more rigorous set of criteria needed to be met before a somatic

site of initiation could be ruled out completely. We decided to implement a number of criteria that would allow us to safely conclude that the events we observed following NMDA receptor activation were indeed dendritic. First, in response to NMDA receptor activation, we observed rapid, inward current spikes in magnocellular neurons that were voltage clamped (Fig. 1a). Although under these conditions, the membrane potential of the cell should, theoretically, be ‘clamped’, this is rarely the case with neurons (Bains and Ferguson, 1999). Because of a cell’s dendritic tree, it is difficult to effectively clamp the voltage in the distal dendrites. Therefore, synaptic input, if sufficiently depolarizing, should activate voltage-gated channels located in these regions. Since the voltage in the soma of the neuron under examination should be clamped at the holding potential, it would be dif-

Fig. 1. Criteria for dendritic spikes. (A) NMDA agonist (1 μM, arrow) in TTX (2 μM) elicits slow inward currents accompanied by fast transients at the onset of response. These are likely unclamped Ca2+ conductances generated in the distal region of the dendrites. (B) Voltage responses to somatic depolarizing current pulses (10–50 pA) in TTX (2 μM). Large amplitude depolarization elicits a sharp spike. (C) Threshold for Ca2+ spike activation varies with stimulus. Top traces are responses of a cell to NMDA agonist application (arrow) and +10 pA depolarizing pulse. The dotted line represents spike activation threshold. Note the invariant threshold for somatic Na+ spikes. Lower traces depict the same experiment in the presence of TTX. Note the marked difference in spike thresholds between current pulses (soma) and NMDA receptor activation.

228

ficult to activate voltage-gated conductances near the recording site. Therefore, these current spikes likely indicate the activation of voltage-gated conductances in a distal region of the dendrite that was not adequately clamped (Bains and Ferguson, 1999). Similar observations were made in neocortical pyramidal cells where unclamped current spikes were observed in response to focal application of glutamate to distal segments of the apical dendrite (Schwindt and Crill, 1997, 1998). Subsequent recordings from the dendrites of these cells have confirmed the presence of dendritic Ca2+ spikes (Schiller et al., 1997, 2000). Second, the high threshold for activation following the injection of depolarizing current into the soma (Bourque and Renaud, 1985; Bains and Ferguson, 1999) suggests that the location of the spike initiation zone is some distance from the soma (Fig. 1). A pulse injected into the soma will be subjected to electrotonic filtering, so lower amplitude pulses which are sufficient to activate axonal Na+ spikes will not be sufficient to depolarize the dendrites and initiate dendritic spikes. (The smaller magnitude pulse required for Na+ spikes may also be due to a lower threshold for activation for Na+ channels as well as a requirement that fewer Na+ channels need to be recruited before a spike is initiated. This issue remains unresolved.) It is also possible that the high threshold is indicative of the opening of L-type, or high-threshold somatic Ca2+ channels, that open at potentials more depolarized than −40 mV. However, recordings from dissociated magnocellular neurons indicate the activation threshold of these currents is approximately −50 mV (Fisher and Bourque, 1995). These channels also exhibit little if any inactivation (Fisher and Bourque, 1995) indicating that a depolarizing event due to activation of these channels should be virtually the same duration as the depolarizing current pulse. This, however, is not the case. The spikes recorded in TTX were brief, transient events (Andrew and Dudek, 1983; Bourque and Renaud, 1985; Bourque et al., 1986; Bains and Ferguson, 1999), consistent with a channel that opens and inactivates quickly, much like a T-type, or lowthreshold Ca2+ channel (Fisher and Bourque, 1995). There are reports of dense distributions of T-type Ca2+ channels in dendrites. These may even be clustered at branch points, forming ‘hotspots’ to boost synaptic signaling. This would be analogous to and

in fact a confirmation of the ‘trigger zones’ proposed by Spencer and Kandel (1961). Finally, the putative dendritic spikes in magnocellular neurons exhibit a varying threshold for activation. This is in contrast to somatic Na+ spikes which exhibit an invariant threshold regardless of whether they are activated by current injection into the soma, or NMDA receptor activation at the dendrites (Fig. 1c). Again, similar observations have been made in neocortical pyramidal cells where dendritic spikes were evoked by application of glutamate directly to the apical dendrites of these cells (Schwindt and Crill, 1997, 1998). The presence of these dendritic spikes has been confirmed with direct dendritic recordings (Schiller et al., 2000). Is there more to a dendritic spike than just Ca2+ flux through voltage-sensitive calcium channels (VSCCs)? The following observations indicate that activation of VSCCs is necessary for dendritic spikes: (1) They are activated by depolarizing current pulses (Bourque and Renaud, 1985; Bains and Ferguson, 1999). This suggests a mechanism that is activated by a change in membrane voltage. (2) They are abolished by CdCl2 , a blocker of Ca2+ channels. In spite of this evidence, we should not be overly hasty in ruling out an additional, perhaps complementary mechanism — the direct influx of Ca2+ through the pore of the NMDA receptor/channel complex itself (Schiller et al., 2000). We have reasoned that activation of NMDA receptors resulted in a local depolarization that was sensed by VSCCs located close to receptors and that this depolarization was sufficient to open VSCCs resulting in a rapid influx of Ca2+ (Bains and Ferguson, 1999). This is entirely consistent with work in the neocortex (Markram and Sakmann, 1994; Schiller et al., 1997), cerebellum (Llinas and Hess, 1976) and hippocampus (Golding et al., 1999). Furthermore, direct dendritic recordings from neocortical neurons suggest that backpropagating action potentials from the soma readily activate dendritic calcium channels resulting in the generation of fast spikes (Stuart and Spruston, 1998; Larkum et al., 1999, 2001). In spite of this evidence, the recent work by Schiller et al. demonstrating that a component of the dendritic spike may result from

229

the flux of Ca2+ ions directly through the NMDA receptor itself (Schiller et al., 2000) cannot be completely ruled out. They elegantly demonstrate, using two-photon laser microscopy that the co-activation of neighboring synaptic inputs onto cortical pyramidal cells results in local, transient membrane depolarizations that are mediated to a large degree by charge movement through the NMDA receptors and may not require activation of VSCCs. A resolution of this issue in magnocellular neurons will require either a similar imaging of, or direct electrophysiological recordings from the dendrites of these cells. What influences the propagation of dendritic spikes? Although Ca2+ spikes can be generated in the dendrites of magnocellular neurons, there is still no guarantee they will reach the soma, and more importantly, that they will alter the firing behavior of the cell. In fact, it is rare that these spikes are even observed during somatic recordings in the absence of TTX. This may be largely due to the fact that a Ca2+ spike would lead to a rapid Na+ spike which would obscure the initial depolarizing event. There is, however, also evidence that the dendrites may be capable of ‘absorbing’ the dendritic spike so that it never reaches the soma. This may result either from the abundant collection of K+ conductances present on dendrites (Hoffman et al., 1997), or alternatively from a limitation imposed by the actual geometry of the dendritic tree (Vetter et al., 2001). First let us examine the influence of other dendritic voltage-gated conductances on dendritic spike propagation. In addition to Ca2+ channels, the dendrites of neurons also possess a rich diversity of K+ (Hoffman et al., 1997) and Na+ (Colbert et al., 1997; Jung et al., 1997; Golding and Spruston, 1998). It is the interaction of these channels, their distribution along the dendrite as well as the recent synaptic history of the cell that will dictate whether a dendritic spike invades the soma. A number of different K+ channels have been located on the dendrites of neurons. They can be classified, broadly, into two categories: K+ channels that buffer synaptic signals (IA and ID families) (Hoffman et al., 1997; Magee and Carruth, 1999; Migliore et al., 1999) and K+ channels that boost synaptic signals (Ih ) (Magee,

1998, 1999). For now, I will focus on the first group of channels that may effectively shunt the dendritic Ca2+ spike before it reaches the soma. In hippocampal CA1 pyramidal cells, a gradient of K+ channel distribution has been established on the dendrites, with a linear increase in IA channel density with respect to distance from the soma (Hoffman et al., 1997). This relationship is subtype specific since the delayed rectifier shows no such gradient, while the calcium activated K+ channel is present almost exclusively in the soma (Poolos and Johnston, 1999). Although the distributions of the respective K+ channels have not been mapped for magnocellular neurons, we can, for the sake of the current discussion, assume they are similar to those observed in other areas of the brain. Altering the relative contributions of these conductances to the propagation of synaptic signals can have profound effects on dendritic spiking. Pharmacological inhibition of IA can lead to an increase in dendritic spiking and a summation of consecutive spikes. The net result is a nearly constant depolarization in the dendrites, which translates into a continuous mode of firing in the soma (Hoffman et al., 1997). These observations, however, may overestimate the role of IA in regulating dendritic excitability since blocking IA with 5 mM 4-aminopyridine (4-AP) will also block another K+ current known as the delay current, ID . This current activates rapidly, like an IA , but it inactivates at a much slower rate, and perhaps of greater significance for dendritic signaling, it remains inactive for tens of seconds (Storm, 1988). It also exhibits a greater sensitivity to 4-AP (micromolar doses) (Storm, 1988) and is also sensitive to α-dendrotoxin (Rudy, 1988). Selective inhibition of ID in the dendrites of hippocampal CA1 pyramidal neurons results in the unmasking of clustered doublets of spikes in response to somatic current injection (Golding et al., 1999). In magnocellular neurons, this manipulation results in the generation of large, plateau-like depolarizations in response to NMDA receptor activation (Bains and Ferguson, 1998). This likely results from a summation of Ca2+ spikes in the dendrites leading to a depolarization that is propagated to the soma (Fig. 2). These findings suggest that under most physiological conditions, dendritic K+ effectively serve to uncouple the dendrites from the soma during intense

230

Fig. 2. Unmasking of dendritic spikes by K+ channel blockade. (A) Magnocellular neurons respond to NMDA agonist with a burst of action potentials that ride a depolarizing envelope. Following blockade of ID with 100 μM 4-AP, a similar concentration of agonist results in a plateau like depolarization. (B) The onset of the response is characterized by a broadening of spikes resulting in an eventual sustained depolarization. (C) A schematic of the presumptive interaction between calcium ions, potassium ions and dendritic spike propagation. (D) A schematic depicting the possible role of ANG in regulating potassium channels and thus influencing the propagation of dendritic spikes.

synaptic activity. This uncoupling is not observed for single EPSPs that are subthreshold for these currents. A larger depolarization such as a Ca2+ spike, however, would activate these currents and be shunted precluding it from reaching the soma. This offers a possible explanation for the observation of multipeaked waveforms in extracellular recordings (Mason and Leng, 1984). They represent a spike that is a compilation of both dendritic and somatic activity. Under certain conditions, however, dendritic spikes may invade the soma and lead to ‘intentional’ bursts. For example, if the dendrites are primed by a second transmitter that will depolarize the dendrites or inactivate K+ currents, this may provide a break in the shunt between dendrite and soma allowing dendritic depolarizations to reach the soma. The ID , in particular has a very slow recovery from inactiva-

tion. Therefore, synaptic signals that are temporally associated with the inactivation of ID can result in a relatively long (tens of seconds) depolarization. This would result in an uninterrupted run of action potentials in the soma that would be terminated only when a sufficient number of ID channels recover from inactivation to once again dampen dendritic excitability. Since this recovery is time and voltage dependent, this could occur in one of two ways: either there is a cessation of excitatory synaptic input to the dendrites for a sufficiently long time to allow recovery, or there is an increase in inhibitory synaptic input resulting in transient hyperpolarizations in the dendrites that relieve the inactivation. This pattern of sustained phasic activity has been observed in OT neurons (Poulain and Wakerley, 1982) and as yet a complete mechanism to explain this phenomenon

231

has not been put forward (but see: Jourdain et al., 1998). Retrograde inhibition of magnocellular neurons Magnocellular neurons can also release chemicals locally to modulate their own activity. The direct effects of exogenous OT and VP on the firing of magnocellular neurons have long been known (Poulain and Wakerley, 1982; Yamashita et al., 1987; Moos and Richard, 1989). There is now ample evidence that these peptides are released from the dendrites under certain physiological conditions (Ludwig, 1995, 1998). Magnocellular neurons can also synthesize and release the labile gas, nitric oxide. Blockade of NO synthesis potentiates excitatory neurotransmitter actions in PVN (Bains and Ferguson, 1994), and augments the release of both corticotrophinreleasing factor (Rivier and Shen, 1994) and VP (Yasin et al., 1993). Furthermore, competitive inhibition of NOS by NT -nitro L-arginine methyl ester (L-NAME) potentiates the membrane depolarization of magnocellular neurosecretory cells (MNCs) in the supraoptic nucleus (SON) in response to NMDA (Cui et al., 1994). Although incredibly different in physical properties, NO and the neuropeptides share common features when it comes to regulating neurohypophyseal function. According to recent studies in slice preparations, each seems to have a net inhibitory effect on the activity of these cells. NO does so by increasing GABAergic drive to the magnos (Bains and Ferguson, 1997) while OT (Kombian et al., 1997; Hirasawa et al., 2001) and VP (Kombian et al., 2000) act at presynaptic excitatory terminals to decrease the release of glutamate. This latter effect, although robust under slice conditions is at odds with other work demonstrating an increase in the activity of magnos in response to exogenous OT (Yamashita et al., 1987; Moos and Richard, 1989; Jourdain et al., 1998). Another question that remains unresolved concerns the mechanism responsible for the release of these messengers. OT can be released in response to back-propagating action potentials in the slice (Kombian et al., 1997). Although this may occur in vivo, it is likely not the stimulus that causes OT release from the dendrites since this release is insensitive to TTX (Ludwig, 1998). The release of both OT and VP, however, is sensitive to extracellular Ca2+

(Ludwig, 1998). Unlike classical neurotransmitters or neuropeptides, NO does not depend on the activation of Ca2+ sensitive machinery and exocytosis of vesicles. Since it is a gas, it can readily diffuse across the membrane. The Ca2+ -dependence of NO is one step earlier in the pathway — during production. The enzyme responsible for NO production, nitric oxide synthase (NOS), is a Ca2+ /calmodulin dependent protein which can be activated by Ca2+ influx that is associated with the activation of NMDA receptors (Garthwaite et al., 1988; Garthwaite, 1991). It can be activated by an increase in intracellular Ca2+ and its activity has been linked to increase in excitatory synaptic activity. Thus, a strong Ca2+ signal, such as a dendritic spike may be utilized as a local signal by a neuron to release vesicles of VP and OT or to manufacture NO which can then diffuse into the surrounding neuropil. The two signals may act in concert to decrease excessive excitation in specific compartments of the dendritic tree (Fig. 3).

Fig. 3. Dendritic spikes and release of retrograde messengers. The schematic depicts the potential role of dendritic spikes in regulating the release of the retrograde messengers. The influx of Ca2+ may be sufficient to cause the exocytotic release of OT or VP which can inhibit the release of glutamate. In addition, Ca2+ can drive the conversion of L-ARG to NO. Once synthesized, NO can diffuse across the membrane to potentiate the release of GABA on to the magnocellular neuron.

232

Dendritic spikes: the darker side As mentioned above, altering the relative contributions of K+ conductances can have profound effects on dendritic spiking. In magnocellular neurons, blocking ID results in a plateau-like depolarization following activation of NMDA receptors (Bains and Ferguson, 1998). This somatic response is likely a symptom of a sustained depolarization in the dendrites of these cells. These plateau depolarizations, however, are not observed when K+ channels are fully functional, suggesting that under most conditions, K+ channels in the dendrites can effectively restrict large depolarization to very selective compartments. When ID is blocked, however, we observe a profound broadening of spikes and appearance of doublets and triplets following activation of NMDA receptors which precedes a sustained depolarization (Fig. 2). This electrophysiological response can be strongly correlated with an increase in susceptibility of magnocellular neurons to the excitotoxic actions of glutamate (Bains et al., 2001). Does this suggest that dendritic spikes may be deleterious to the cell under certain conditions? This is certainly possible during periods of intense glutamatergic excitation. The interplay between Ca2+ spikes and dendritic K+ channels also raises a second interesting possibility: the rise in Ca during a dendritic spike is sufficient to activate a K+ -channel interacting proteins that can increase the conductance and inactivation time constant of Kv 4.2 in a Ca2+ -dependent manner (An et al., 2000). These proteins may be activated in response to the surge in Ca2+ that accompanies synaptic activation, specifically, but not exclusively when the NMDA receptor is involved (An et al., 2000). This would result in an increase in the conductance of K+ channels and provide a safety-valve that would protect the cell by preventing the passage of large depolarizations from the dendrites to the soma (Golding et al., 1999) following intense synaptic activity. Long-term effects of dendritic spikes Any process that requires a rapid Ca2+ flux for activation is a potential target for dendritic Ca2+ spikes. Many of the processes implicated in altering synaptic strength, i.e. changes in expression of post-synaptic

receptors, fall into this category. In magnocellular neurons, there are additional morphological changes that occur that may be sensitive to the effects of dendritic spikes. For example, there is considerable evidence that the dendrites of OT neurons may undergo pruning during lactation (Stern and Armstrong, 1998). Might this pruning be initiated by a dramatic change in synaptic activity? The synapses that are strongly activated may be strengthened, while those that exhibit a decrease in activity are pruned off. This may be a way for the neuron to increase its signal-tonoise ratio and allow it to focus more efficiently on the task at hand. Alternatively, perhaps the pruning acts to facilitate the propagation of dendritic spikes to the soma, thereby increasing the probability that any given synaptic potential will generate a somatic action potential. Summary Although our understanding of dendritic physiology has advanced greatly in the last decade, we are only beginning to scratch the surface of what these new findings mean for the release of OT and VP. While making the link between active dendritic conductances and the output of magnocellular neurons has been hindered by our inability to record faithfully from dendrites, we have still been able to make some critical advances. Most notably, we know that dendritic spikes can be activated by intense synaptic activity, perhaps mediated by NMDA receptors. We also know that the powerful dendritic K+ channels must be suppressed in order for dendritic spikes to be faithfully translated into somatic action potentials. Furthermore, if coupled with the episodic release of glutamate dendritic spikes the inactivation of K+ currents may be the substrate for burst initiation in the soma. Finally, dendritic spikes seem like the most likely candidate mechanism utilized by magnocellular neurons to release chemical messengers from their dendrites and communicate in a retrograde fashion. It is unclear whether dendritic spikes act as local signals that are restricted to selective compartments and then quickly terminated by powerful K+ conductances, or whether they increase the probability that synaptic signals that may otherwise be lost in the background are now heard loud and clear.

233

References An, W.F., Bowlby, M.R., Betty, M., Cao, J., Ling, H.P., Mendoza, G., Hinson, J.W., Mattsson, K.I., Strassle, B.W., Trimmer, J.S. and Rhodes, K.J. (2000) Modulation of A-type potassium channels by a family of calcium sensors. Nature, 403: 553–556. Andrew, R.D. and Dudek, F.E. (1983) Burst discharge in mammalian neuroendocrine cells involves an intrinsic regenerative mechanism. Science, 221: 1050–1052. Bains, J.S. and Ferguson, A.V. (1994) Angiotensin II neurotransmitter actions in paraventricular nucleus are potentiated by a nitric oxide synthase inhibitor. Regul. Pept., 50: 53–59. Bains, J.S. and Ferguson, A.V. (1997) Nitric oxide regulates NMDA-driven GABAergic inputs to type I neurones of the rat paraventricular nucleus. J. Physiol., 499(Pt 3): 733–746. Bains, J.S. and Ferguson, A.V. (1998) Hyperpolarising afterpotentials prevent long duration plateau depolarisations in rat paraventricular nucleus neurons. Eur. J. Neurosci., 10: 1412– 1421. Bains, J.S. and Ferguson, A.V. (1999) Activation of N-methylD -aspartate receptors evokes calcium spikes in the dendrites of rat hypothalamic paraventricular nucleus neurons. Neuroscience, 90: 885–891. Bains, J.S., Follwell, M.J., Latchford, K.J., Anderson, J.W. and Ferguson, A.V. (2001) Slowly-inactivating potassium conductance (ID ): a potential target for stroke therapy. Stroke, 32(11): 2624–2634. Bourque, C.W. and Renaud, L.P. (1985) Calcium-dependent action potentials in rat supraoptic neurosecretory neurones recorded in vitro. J. Physiol., 363: 419–428. Bourque, C.W., Randle, J.C. and Renaud, L.P. (1986) Nonsynaptic depolarizing potentials in rat supraoptic neurones recorded in vitro. J. Physiol., 376: 493–505. Colbert, C.M., Magee, J.C., Hoffman, D.A. and Johnston, D. (1997) Slow recovery from inactivation of Na+ channels underlies the activity-dependent attenuation of dendritic action potentials in hippocampal CA1 pyramidal neurons. J. Neurosci., 17: 6512–6521. Cui, L.N., Inenaga, K., Nagatomo, T. and Yamashita, H. (1994) Sodium nitroprusside modulates NMDA response in the rat supraoptic neurons in vitro. Brain Res. Bull., 35: 253–260. Fisher, T.E. and Bourque, C.W. (1995) Voltage-gated calcium currents in the magnocellular neurosecretory cells of the rat supraoptic nucleus. J. Physiol., 486(Pt 3): 571–580. Garthwaite, J. (1991) Glutamate, nitric oxide and cell–cell signalling in the nervous system. Trends Neurosci., 14: 60–67. Garthwaite, J., Charles, S.L. and Chess-williams, R. (1988) Endothelium-derived relaxing factor release on activation of NMDA receptors suggests role as intercellular messenger in the brain. Nature, 336: 385–388. Golding, N.L. and Spruston, N. (1998) Dendritic sodium spikes are variable triggers of axonal action potentials in hippocampal CA1 pyramidal neurons. Neuron, 21: 1189–1200. Golding, N.L., Jung, H.Y., Mickus, T. and Spruston, N. (1999) Dendritic calcium spike initiation and repolarization are controlled by distinct potassium channel subtypes in CA1 pyramidal neurons. J. Neurosci., 19: 8789–8798.

Hausser, M., Spruston, N. and Stuart, G.J. (2000) Diversity and dynamics of dendritic signaling. Science, 290: 739–744. Hirasawa, M., Kombian, S.B. and Pittman, Q.J. (2001) Oxytocin retrogradely inhibits evoked, but not miniature, EPSCs in the rat supraoptic nucleus: role. J. Physiol., 532: 595–607. Hoffman, D.A., Magee, J.C., Colbert, C.M. and Johnston, D. (1997) K+ channel regulation of signal propagation in dendrites of hippocampal pyramidal neurons. Nature, 387: 869– 875. Johnston, D., Hoffman, D.A., Colbert, C.M. and Magee, J.C. (1999) Regulation of back-propagating action potentials in hippocampal neurons. Curr. Opin. Neurobiol., 9: 288–292. Jourdain, P., Israel, J.M., Dupouy, B., Oliet, S.H., Allard, M., Vitiello, S., Theodosis, D.T. and Poulain, D.A. (1998) Evidence for a hypothalamic oxytocin-sensitive pattern-generating network governing oxytocin neurons in vitro. J. Neurosci., 18: 6641–6649. Jung, H.Y., Mickus, T. and Spruston, N. (1997) Prolonged sodium channel inactivation contributes to dendritic action potential attenuation in hippocampal pyramidal neurons. J. Neurosci., 17: 6639–6646. Kombian, S.B., Mouginot, D. and Pittman, Q.J. (1997) Dendritically released peptides act as retrograde modulators of afferent excitation in the supraoptic nucleus in vitro. Neuron, 19: 903– 912. Kombian, S.B., Mouginot, D., Hirasawa, M. and Pittman, Q.J. (2000) Vasopressin preferentially depresses excitatory over inhibitory synaptic transmission in the rat supraoptic nucleus in vitro. J. Neuroendocrinol., 12: 361–367. Larkum, M.E., Zhu, J.J. and Sakmann, B. (1999) A new cellular mechanism for coupling inputs arriving at different cortical layers. Nature, 398: 338–341. Larkum, M.E., Zhu, J.J. and Sakmann, B. (2001) Dendritic mechanisms underlying the coupling of the dendritic with the axonal action potential initiation zone of adult rat layer 5 pyramidal neurons. J. Physiol., 533: 447–466. Legendre, P., Cooke, I.M. and Vincent, J.D. (1982) Regenerative responses of long duration recorded intracellularly from dispersed cell cultures of fetal mouse hypothalamus. J. Neurophysiol., 48: 1121–1141. Llinas, R. and Hess, R. (1976) Tetrodotoxin-resistant dendritic spikes in avian Purkinje cells. Proc. Natl. Acad. Sci. USA, 73: 2520–2523. Llinas, R. and Nicholson, C. (1971) Electrophysiological properties of dendrites and somata in alligator Purkinje cells. J. Neurophysiol., 34: 532–551. Llinas, R. and Sugimori, M. (1980) Electrophysiological properties of in vitro Purkinje cell dendrites in mammalian cerebellar slices. J. Physiol., 305: 197–213. Loronté de Nó, R. and Coundouris, G.A. (2001) Decremental conduction in peripheral nerve integration of stimuli in the neuron. Proc. Natl. Acad. Sci. USA, 45: 592–617. Ludwig, M. (1995) Functional role of intrahypothalamic release of oxytocin and vasopressin: consequences and controversies. Am. J. Physiol., 268: E537–E545. Ludwig, M. (1998) Dendritic release of vasopressin and oxytocin. J. Neuroendocrinol., 10: 881–895.

234 Magee, J.C. (1998) Dendritic hyperpolarization-activated currents modify the integrative properties of hippocampal CA1 pyramidal neurons. J. Neurosci., 18: 7613–7624. Magee, J.C. (1999) Dendritic lh normalizes temporal summation in hippocampal CA1 neurons. Nat. Neurosci., 2: 508–514. Magee, J.C. and Carruth, M. (1999) Dendritic voltage-gated ion channels regulate the action potential firing mode of hippocampal CA1 pyramidal neurons. J. Neurophysiol., 82: 1895–1901. Magee, J., Hoffman, D., Colbert, C. and Johnston, D. (1998) Electrical and calcium signaling in dendrites of hippocampal pyramidal neurons. Annu. Rev. Physiol., 60: 327–346. Markram, H. and Sakmann, B. (1994) Calcium transients in dendrites of neocortical neurons evoked by single subthreshold excitatory postsynaptic potentials via low-voltage-activated calcium channels. Proc. Natl. Acad. Sci. USA, 91: 5207–5211. Mason, W.T. and Leng, G. (1984) Complex action potential waveform recorded from supraoptic and paraventricular neurones of the rat: evidence for sodium and calcium spike components at different membrane sites. Exp. Brain Res., 56: 135–143. Migliore, M., Hoffman, D.A., Magee, J.C. and Johnston, D. (1999) Role of an A-type K+ conductance in the backpropagation of action potentials in the dendrites of hippocampal pyramidal neurons. J. Comput. Neurosci., 7: 5–15. Moos, F. and Richard, P. (1989) Paraventricular and supraoptic bursting oxytocin cells in rat are locally regulated by oxytocin and functionally related. J. Physiol., 408: 1–18. Moos, F., Gouzenes, L., Brown, D., Dayanithi, G., Sabatier, N., Boissin, L., Rabie, A. and Richard, P. (1998) New aspects of firing pattern autocontrol in oxytocin and vasopressin neurones. Adv. Exp. Med. Biol., 449: 153–162. Poolos, N.P. and Johnston, D. (1999) Calcium-activated potassium conductances contribute to action potential repolarization at the soma but not the dendrites of hippocampal CA1 pyramidal neurons. J. Neurosci., 19: 5205–5212. Poulain, D.A. and Wakerley, J.B. (1982) Electrophysiology of hypothalamic magnocellular neurones secreting oxytocin and vasopressin. Neuroscience, 7: 773–808. Pow, D.V. and Morris, J.F. (1989) Dendrites of hypothalamic magnocellular neurons release neurohypophysial peptides by exocytosis. Neuroscience, 32: 435–439. Rall, W. (1959) Branching dendritic trees and motoneuron resistivity. Exp. Neurol., 1: 441–452. Rivier, C. and Shen, G.H. (1994) In the rat, endogenous nitric oxide modulates the response of the hypothalamic–pituitary– adrenal axis to interleukin-1 beta, vasopressin, and oxytocin. J. Neurosci., 14: 1985–1993.

Rudy, B. (1988) Diversity and ubiquity of K channels. Neuroscience, 25: 729–749. Schiller, J., Schiller, Y., Stuart, G. and Sakmann, B. (1997) Calcium action potentials restricted to distal apical dendrites of rat neocortical pyramidal neurons. J. Physiol., 505(3): 605–616. Schiller, J., Major, G., Koester, H.J. and Schiller, Y. (2000) NMDA spikes in basal dendrites of cortical pyramidal neurons. Nature, 404: 285–289. Schwindt, P.C. and Crill, W.E. (1997) Local and propagated dendritic action potentials evoked by glutamate iontophoresis on rat neocortical pyramidal neurons. J. Neurophysiol., 77: 2466–2483. Schwindt, P.C. and Crill, W.E. (1998) Synaptically evoked dendritic action potentials in rat neocortical pyramidal neurons. J. Neurophysiol., 79: 2432–2446. Segev, I. and London, M. (2000) Untangling dendrites with quantitative models. Science, 290: 744–750. Segev, I. and Rall, W. (1998) Excitable dendrites and spines: earlier theoretical insights elucidate recent direct observations. Trends Neurosci., 21: 453–460. Spencer, W.A. and Kandel, E.R. (1961) Electrophysiology of hippocampal neurons: IV. Fast prepotentials. J. Neurophysiol., 24: 272–285. Stern, J.E. and Armstrong, W.E. (1998) Reorganization of the dendritic trees of oxytocin and vasopressin neurons of the rat supraoptic nucleus during lactation. J. Neurosci., 18: 841–853. Storm, J.F. (1988) Temporal integration by a slowly inactivating K+ current in hippocampal neurons. Nature, 336: 379–381. Stuart, G. and Spruston, N. (1998) Determinants of voltage attenuation in neocortical pyramidal neuron dendrites. J. Neurosci., 18: 3501–3510. Stuart, G.J., Dodt, H.U. and Sakmann, B. (1993) Patch-clamp recordings from the soma and dendrites of neurons in brain slices using infrared video microscopy. Pflugers Arch., 423: 511–518. Vetter, P., Roth, A. and Hausser, M. (2001) Propagation of action potentials in dendrites depends on dendritic morphology. J. Neurophysiol., 85: 926–937. Yamashita, H., Okuya, S., Inenaga, K., Kasai, M., Uesugi, S., Kannan, H. and Kaneko, T. (1987) Oxytocin predominantly excites putative oxytocin neurons in the rat supraoptic nucleus in vitro. Brain Res., 416: 364–368. Yasin, S., Costa, A., Trainer, P., Windle, R., Forsling, M.L. and Grossman, A. (1993) Nitric oxide modulates the release of vasopressin from rat hypothalamic explants. Endocrinology, 133: 1466–1469. Yuste, R. and Tank, D.W. (1996) Dendritic integration in mammalian neurons, a century after Cajal. Neuron, 16: 701–716.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 18

Modulation of synaptic transmission by oxytocin and vasopressin in the supraoptic nucleus S.B. Kombian 1,∗ , M. Hirasawa 2, D. Mouginot 3 and Q.J. Pittman 2 2

1 Faculty of Pharmacy, Kuwait University, Kuwait Neuroscience Research Group and Department of Physiology and Biophysics, University of Calgary, Calgary, AB, Canada 3 CHUL Research Center, Neuroscience Unit, Laval University, Sainte-Foy, QC, Canada

Abstract: It is now generally accepted that magnocellular neurons of the supraoptic and paraventricular nuclei release the neuropeptides oxytocin and vasopressin from their dendrites. Peptide release from their axon terminals in the posterior pituitary and dendrites differ in dynamics suggesting that they may be independently regulated. The dendritic release of peptide within the supraoptic nucleus (SON) is an important part of its physiological function since the local peptides can regulate the electrical activity of magnocellular neurons (MCNs) which possess receptors for these peptides. This direct postsynaptic action would affect the output of peptide in the neurohypophysis. Another way that these peptides can regulate MCN activity would be to modulate afferent inputs unto themselves. Although the influence of afferent inputs (inhibitory and excitatory) on SON magnocellular neuron physiology has been extensively described in the last decade, a role for these locally released peptides on synaptic physiology of this nucleus has been difficult to show until recently, partly because of the difficulty of performing stable synaptic recordings from these cells in suitable preparations that permit extensive examination. We recently showed that under appropriate conditions, oxytocin acts as a retrograde transmitter in the SON. Oxytocin, released from the dendrites of MCNs, decreased evoked excitatory synaptic transmission by inhibiting glutamate release from the presynaptic terminals. It modulated voltage-dependent calcium channels, mainly N-type and to a lesser extent P/Q-type channels, located on glutamatergic terminals. Although evidence is less conclusive, it is possible that vasopressin has similar actions to reduce excitatory transmission. This synaptic depressant effect of oxytocin and/or vasopressin, released from dendrites, would ensure that MCNs regulate afferent input unto themselves using their own firing rate as a gauge. Alternatively, it may only be a subset of afferent terminals that are sensitive to these peptides, thereby providing a means for the MCNs to selectively filter their afferent inputs. Indeed its specificity is partly proven by our observation that oxytocin does not affect spontaneous glutamate release, or GABA release from inhibitory terminals (Brussaard et al., 1996). Thus, the dendrites of MCNs of the supraoptic nucleus serve a dual role as both recipients of afferent input and regulators of the magnitude of afferent input, allowing them to directly participate in the shaping of their output. This adds to a rapidly growing body of evidence in support of the concept of a two-way communication between presynaptic terminals and postsynaptic dendrites, and shows the potential of this nucleus as a model to study such form of synaptic transmission. Keywords: Supraoptic nucleus; Vasopressin; Oxytocin; Excitatory postsynaptic current; Retrograde transmission; Somatodendritic release; Voltage-activated calcium channel

∗ Correspondence to: S.B. Kombian, Faculty of Pharmacy, Kuwait University, Kuwait. Fax: +965-534-2807; E-mail: [email protected]

236

Introduction A long held and classic concept of neurotransmission is that dendrites of neurons only serve as recipients of synaptic inputs where they function only to integrate and relay information. Inherent in this concept is the unidirectional transfer of information from the presynaptic neuron to the postsynaptic neuron. This concept is now giving way to a more prominent role for these anatomical structures such that, in addition to receiving, integrating and relaying afferent information, they may also regulate afferent inputs unto themselves thus, helping to model the input and hence their own output (Conde, 1992; Pucak and Grace, 1994). This evolving concept implies a bidirectional mode of communication between neurons whereby information can flow from the postsynaptic neuron back to the presynaptic neuron terminal and vice versa. Thus, anterograde and retrograde synaptic transmission can occur at the synapse concurrently. In keeping with this additional role, dendrites have recently been shown to elaborate, store and release transmitters that act as retrograde messengers to affect afferent input. Conventional transmitters such as dopamine, glutamate and GABA (Cheramy et al., 1981; Klitenick et al., 1992; Glitsch et al., 1996; Zilberter et al., 1999) as well as endogenous opioids and cannabinoids (Drake et al., 1994; Kreitzer and Regehr, 2001) have all been shown to carry out such a function in several brain regions. As well, the neuropeptides oxytocin (OXT) and vasopressin (VP) have recently been shown to engage in retrograde neurotransmission where they influence afferent input onto themselves (Kombian et al., 1997; Hirasawa et al., 2001). The SON and the paraventricular nucleus (PVN) of the hypothalamus are the main sources of circulating OXT and VP in mammals (Bargmann and Scharrer, 1951). Most of this review will deal with the SON, although most of the findings here may well apply to the PVN. The SON has been shown by neurochemical, immunocytochemical and electron microscopic studies to possess all the prerequisites for engaging in retrograde neurotransmission. Immunocytochemical studies show that the soma and dendrites of magnocellular neurons have vesicles that contain the peptides (Castel and Morris, 1988). Release profiles have also been demonstrated in the

somatodendritic regions of the MCNs using electron microscopic studies (Morris and Pow, 1993). Finally, microdialysis studies reveal the presence of extracellular OXT and VP locally in this nucleus (Di Scala-Guenot et al., 1987; Ludwig and Landgraf, 1992; Neumann et al., 1993, 1996). This local release was calcium-dependent and tetrodotoxinresistant (Di Scala-Guenot et al., 1987; Ludwig and Landgraf, 1992). All the above evidence suggest that the somatodendritic region of the magnocellular neuron is equipped with the necessary components that enable them to engage in peptide release using processes similar to those that have been described at the axon terminal (Pow and Morris, 1989). As axon collaterals are very rare in the SON (Mason et al., 1984; Pow and Morris, 1988) and peptidergic afferents from other thalamic regions are scarce (Conrad and Pfaff, 1976; Saphier and Feldman, 1987), most of the local peptide must therefore arise from somatodendritic release. Not only are these peptides present in the SON and/or PVN, their receptors have been demonstrated in these regions (Brinton et al., 1984a,b; Van Leeuwen et al., 1985) and functional roles for these local, somatodendritically released peptides have been reported (Lambert et al., 1993; Neumann et al., 1996). The current issue to resolve then is: How do these peptides produce these effects? Action of OXT/VP on magnocellular neurons OXT or VP have both been shown to cause direct effects on magnocellular neurons (MCNs; the same cells that manufacture, store and secrete them) by acting on autoreceptors located on these cells, both in vivo and in vitro (Freund-Mercier and Richard, 1981; Yamashita et al., 1987). These peptides cause an increase in intracellular calcium (Lambert et al., 1994; Dayanithi et al., 1996) that may contribute to the excitation of MCNs and trigger more local and systemic peptide release. In addition to these direct effects of the peptides on postsynaptic cell, OXT has also been reported to decrease extrinsic synaptic influences on these cells by postsynaptically modulating inhibitory GABAA receptor-mediated postsynaptic currents (Brussaard et al., 1996). In this study, OXT was found to decrease spontaneous and evoked IPSC amplitude but not the frequency of occurrence of spontaneous IPSCs. These effects of

237

OXT could be mimicked by intracellular manipulation of upstream signal transduction systems such as activating or inhibiting G-proteins and mobilization of intracellular calcium in the MCN. Retrograde action of OXT/VP Although the MCNs of the hypothalamus possess several intrinsic conductances that are important to the function of SON and PVN, extrinsic synaptic inputs are very important as well (Renaud and Bourque, 1991; Armstrong, 1995). The SON receives both excitatory and inhibitory synaptic inputs from several regions including the circumventricular organs (Tribollet et al., 1985; Chaudhry et al., 1989; Leng et al., 1989; Honda et al., 1990; McKinley et al., 1992; Richard and Bourque, 1995). Both these excitatory and inhibitory inputs have recently been intensely studied as they turn to affect the excitability of the MCNs and may be involved in the generation of appropriate firing patterns to optimize peptide release in the periphery as well as locally within the nucleus. The role of these locally released peptides (OXT and VP) on these inputs is the main thrust of this review. Using an in vitro hypothalamic slice preparation containing the SON and the nystatin-patch whole-cell recording technique, we examined the role of these peptides on evoked and spontaneous synaptic events in the SON. Bath (exogenous) application of OXT, its agonist or VP all caused concentration dependent decreases in evoked glutamate-mediated excitatory postsynaptic currents (Fig. 1; Kombian et al., 1997) These synaptic depressant effects were blocked by Manning compound (MC), a non-selective OXT/VP receptor antagonist (Kruszynski et al., 1980) as well as by a selective OXT receptor antagonist. Although OXT and VP produced these effects when they were exogenously applied, a more crucial question was whether endogenous (local, dendritically released) peptides produced these same effects. To answer this question, we exploited the fact that OXT and VP, after being released, are rapidly degraded by local aminopeptidase enzymes (Burbach and Lebouille, 1983; Claybaugh and Uyehara, 1993). Therefore, we used an aminopeptidase inhibitor, amastatin and showed that this compound produced a synaptic depression similar to that produced by exogenous

peptides. This effect was specific for the aminopeptidase inhibitor as an inhibitor of endopeptidase (an enzyme present in the SON but not involved in the metabolism of OXT/VP), phosphoramidon, did not have an effect on evoked responses. Furthermore, the amastatin-induced synaptic depression was blocked by the peptide antagonist (MC) suggesting that amastatin caused synaptic depression indirectly by enhancing extracellular levels of OXT and/or VP. Thus, endogenous dendritically released peptides, when not rapidly destroyed by degradative enzymes, can decrease afferent excitation unto MCNs. The presence of these aminopeptidases in the SON suggest that there is, routinely, no tonic action of OXT and/or VP in this nucleus. This may be necessary to ensure that peptide receptors do not desensitize due to prolonged contact with agonist. The peptides may therefore produce an effect in the SON only when the enzymes are overwhelmed following massive peptide release. This may happen when MCN activity (action potential firing) is intense following large or extended peripheral demand resulting in huge/massive terminal and dendritic release. We attempted to increase intra-SON peptide level by using high-frequency afferent stimulation (HFS) that has been reported to optimize peptide release (Bicknell, 1988). HFS of afferents at 100 Hz for 1 s was followed by synaptic depression that lasted almost as long as that due to exogenous peptide application (Fig. 2A). Furthermore, the HFS induced depression was also blocked by MC indicating HFS produced a large release of peptide that overwhelmed the peptidases to produce excitatory synaptic depression (Fig. 2C). In addition to the above HFS effect, it was also possible to produce adequate release of peptide from a single MCN, enough to cause synaptic depression. This was achieved by injecting adequate positive current into MCNs to make them fire several action potentials (up to 50 Hz) for a second or more (Fig. 2B). The synaptic depression following this depolarization was also blocked by MC. All these findings indicate that peptides released from the somatodendritic region of the MCN act to decrease excitatory synaptic transmission. The consistent depression of excitatory synaptic transmission observed for OXT and VP contrasts with the effects of these peptides on GABA-mediated inhibitory transmission. While Brussaard et al., 1996

238

Fig. 1. (A) Exogenous application of OXT or AVP or endogenous enhancement of neurohypophyseal hormone levels (with appropriate enzymes) leads to a reversible decrease in evoked excitatory synaptic transmission. This effect of peptide can be blocked by Manning compound (MC), a non-selective OXT/AVP receptor antagonist. (B) Histogram showing average effects of peptides and antagonism by MC. * P < 0.05.

reported that OXT depressed the amplitude of both evoked and spontaneous IPSCs recorded in the SON, we (Kombian et al., 2000) did not observe an effect of VP on evoked IPSCs recorded in this nucleus. Given that VP can bind OXT receptors (Mühlethaler and Dreifuss, 1983), it is surprising that we did not observe an effect of VP on evoked IPSCs recorded in the SON. Furthermore, Hermes et al. (2000) reported that VP enhanced spontaneous IPSCs in the PVN indirectly, possibly through the excitation of GABAergic interneurons located in the perinuclear zone of

the PVN. Since Brussaard et al., 1996 showed that the effect of OXT was postsynaptically mediated, and different MCNs may express different peptide receptors (Hurbin et al., 1998), difference in cell sampling in the two studies may be responsible for these discordant observations. Nonetheless, they suggest that spontaneous and evoked IPSCs may selectively respond to only OXT but not VP. Taken together, these indicate that peptide modulation of inhibitory transmission in the SON is complex.

239

Fig. 2. Biophysical enhancement of peptide release is accompanied by a decrease in evoked EPSC. (A) High frequency stimulation (HFS) of afferents is accompanied by a decrease in evoked excitatory synaptic transmission. (B) Injection of positive current into a single MCN causes synaptic depression. (Insert is membrane response to current injection in current clamp mode.) (C) The effect of HFS-induced synaptic depression is blocked by MC pretreatment.

Locus of action of OXT/VP in the SON To alter synaptic transmission, neuromodulators can alter either: (1) the responsiveness of the postsynaptic cell to the same amount of transmitter; (2) the background or intrinsic currents in the postsynaptic

Fig. 3. OXT-induced synaptic depression is through a presynaptic action. (A,B) OXT-induced depression is accompanied by an increase in paired-pulse ratio. (C) OXT does not induced a change in the postsynaptic current over a wide range of voltage. * P < 0.05.

cell that will change its response to a transmitter; or (3) the amount of transmitter released from the afferent terminal. The former two are postsynaptic mechanisms while the latter is presynaptic. We used several biophysical tests to determine the locus of action of OXT and VP to decrease EPSC amplitude. Application of two successive synaptic stimuli (separated by 50 ms, Fig. 3A) was most often followed by a much larger EPSC on the second stimulus resulting in a ratio between the second and first response (2/1) of greater than unity (paired-pulse facilitation, PPF).

240

This change has been shown to be due, predominantly, to changes in presynaptic terminal calcium dynamics (Zucker, 1989). OXT-induced EPSC reduction was most often followed by a change in paired-pulse ratio (PPR; Fig. 3B) suggesting that it may be acting on presynaptic glutamate terminals to decrease glutamate release. OXT did not appear to affect postsynaptic voltage-activated intrinsic currents as measured by the membrane response to voltage steps and ramps ranging from −120 to −30 mV (Fig. 3C). Furthermore, it did not change AMPAinduced postsynaptic currents in MCNs indicating it does not affect the postsynaptic responsiveness of non-NMDA receptors that mediate evoked EPSCs. By contrast, VP appears to affect evoked EPSCs in the SON by a more complex mechanism that may involve both pre- and post-synaptic mechanisms (unpublished observation). In some cells VP decreased the evoked EPSC and this was via a presynaptic mechanism as the depression was often, but not always, followed by a change in PPR. In other cells, VP increased the evoked EPSC and this was without changing the PPR while enhancing AMPA-induced currents, suggesting a postsynaptic action to enhance evoked EPSCs (unpublished observation). These actions of VP may reflect the presence of different types of MCNs in the SON that may respond differently to VP (Stern and Armstrong, 1995; Leng et al., 1999). These dichotomous actions of VP would enable it to selectively gate certain afferent inputs into the SON by its retrograde action. The currently available evidence is that OXT and/ or VP reduce excitatory synaptic transmission by a presynaptic action where they decrease glutamate release (Kombian et al., 1997, 2000). To determine the mechanism by which OXT/VP may decrease glutamate release, we asked if the afferent terminal calcium, a critical trigger for transmitter release (Sabatini and Regehr, 1997; Wu and Saggau, 1997) was altered by the peptides. First we demonstrated that evoked EPSCs in the SON were calcium-dependent by completely abolishing them with a non-selective calcium channel blocker, Cd2+ (Fig. 4). Next we determined the types of calcium channels that mediate EPSCs in this nucleus by applying relatively selective calcium channel blockers and tested their effects on evoked EPSCs (Hirasawa et al., 2001). We found that saturating concentrations of ω-conotoxin

GVIA, an N-type calcium channel blocker, reduced the evoked EPSC by about 60% while ω-agatoxin TK, a P/Q-type channel blocker decreased it by about 40%. When both these antagonists were present, the evoked EPSC was almost completely abolished (Fig. 4). Other types of calcium channel blockers such as nickel (T- and R-type) and nicardipine (Ltype) were without effect on evoked EPSCs. Since calcium channels are present on both afferent terminals and MCNs (Fisher and Bourque, 1996), it is possible that these channel blockers may produce their effects on the evoked EPSC by altering both the preand post-synaptically located channels. A battery of tests, including PPR, kinetics of evoked currents and responses to bath applied AMPA, all indicated that these substances had very little effect on the soma of MCNs in the voltage range tested. Thus, the results from these tests were all consistent with a presynaptic action of the blockers to decrease evoked EPSC amplitude (Hirasawa et al., 2001). Now if the peptides were interacting with calcium channels, specifically the N- and P/Q-type channels, to decrease glutamate release and hence evoked EPSC amplitude, then the presence of the channel blockers should occlude the actions of OXT/VP. We therefore performed experiments where OXT was applied in the presence of ω-conotoxin or ω-agatoxin (Fig. 5). In the presence of each of these compounds, the synaptic depressant effect of OXT was significantly attenuated (Fig. 5B). This attenuation was much greater in the presence ω-conotoxin than was the case with ωagatoxin suggesting that OXT predominantly affects N-type channels and to a lesser extent P/Q-type channels in this nucleus. These occluding effects of the Nand P/Q-type channel blockers were also observed when the level of endogenous peptide (OXT/VP) was enhanced by the aminopeptidase enzyme inhibitor amastatin (Hirasawa et al., 2001). Thus, the occlusion of OXT effects by certain calcium channel blockers strongly suggest that OXT acts at the same site as the blockers to decrease synaptic transmission. More evidence in support of this fact, though indirect, is that both OXT and these channel blockers, in parallel, also do not affect tetrodotoxin-resistant, extracellular calcium-independent spontaneous EPSCs in this SON (Fig. 6). The mechanism by which these peptides may affect the calcium channels is not known right now.

241

Fig. 4. Excitatory synaptic transmission in the SON is mediated by N- and P/Q-type calcium channels. (A) A time–effect graph of a typical cell showing the effect of two calcium channel blockers on evoked EPSC amplitude. (B) Summary histograms showing the effects of several calcium channel blockers on evoked EPSC. *,** P < 0.05.

However, as most OXT effects have been shown to be G-protein-mediated (Thibonnier et al., 1998) and calcium channels, especially the N-type channels that are affected most by OXT, are known to be modulated by G-proteins (Currie and Fox, 1997), it is plausible that OXT depresses these channels via its G-protein coupling. In addition to the modulation of calcium channels to decrease transmitter release, another possible means by which OXT may affect transmitter release is through processes downstream of calcium entry. These downstream processes may control both evoked and spontaneous EPSCs, the latter in the SON being independent of action potentials and

extracellular calcium (Kabashima et al., 1997; Kombian et al., 2000). The modulation of any processes downstream of calcium entry by OXT is, however, unlikely as OXT does not affect the frequency or amplitude of spontaneous EPSCs (Fig. 6). This lack of effect of OXT on spontaneous EPSC contrasts with its inhibition of evoked EPSCs in the SON and points to the fact that spontaneous and evoked EPSCs may employ different processes to produce transmitter release. The lack of effect of OXT on spontaneous EPSCs frequency and amplitude also contrast with its reported depression of the amplitude of spontaneous IPSCs (Brussaard et al., 1996). These diverse actions of OXT on synaptic responses

242

Fig. 5. ω-CTx and ω-Aga actions occlude OXT effects on excitatory synaptic transmission. (A) Time–effect plot in a typical cell showing the effect of OXT in the presence of calcium channel blockers. (B) Summary histogram showing the effect of OXT in the presence of maximal concentrations of ω-CTx and ω-Aga. * P < 0.05.

suggests that spontaneous EPSCs and IPSCs that are action potential independent (asynchronous) and evoked EPSCs and IPSCs that are produced by the invasion of terminals by action potentials leading to synchronous release of transmitters, may perform different functions in the SON and may be differentially modulated.

Fig. 6. Calcium channel antagonists and OXT have no effect on miniature EPSCs in the SON. (A) Sample traces showing the effect of OXT and ω-CTx on evoked EPSC. (B) In these same cells OXT and ω-CTx have no effect on the amplitude or frequency of miniature EPSCs. (C) Summary histogram showing the lack of effect of OXT or ω-CTx on miniature EPSCs.

Conclusion The dendrites and soma of MCNs of the SON are endowed with peptide containing vesicles and calcium channels (Fisher and Bourque, 1996), two important cellular components that are necessary for the release of peptide into the extracellular space. Release

profiles in the form of membrane plebs have been observed in the somatodendrites of MCNs and peptides have routinely been recovered from SON using microdialysis probes. Since axon collaterals and peptidergic afferents to SON are sparse, the main source of OXT and VP in the SON is thus, through so-

243

Fig. 7. A schematic showing possible mechanisms by which OXT/AVP modulate synaptic transmission in the supraoptic nucleus. Following a burst of APs the MCN releases peptide from the soma/dendrites which cross the synaptic cleft to activate receptors located on excitatory but not inhibitory terminals. This leads to an interaction with N- and P/Q-type calcium channels causing a decrease in evoked, but not spontaneous glutamate release. Dendritically released peptides can also activate autoreceptors located on the MCN leading to a decrease in the response of the MCN to GABA (Brussard et al., 1995).

matodendritic release. Currently available evidence indicate that, at the cellular level, the excitability of MCNs is controlled both by intrinsic conductances as well as afferent excitation/inhibition. During increased peripheral demand for peptide, both afferent excitation and intrinsic conductances may cooperate to cause MCNs to switch their action potential firing pattern from routine baseline patterns and rates to those that enhance peptide release in the periphery. This switch in pattern to optimize peripheral peptide release causes MCNs to fire more action potentials. These potentials that can be initiated in the dendrites (Bains and Ferguson, 1999) or the axon hillock (that can back-propagate to the dendrites, (Stuart et al., 1997)) trigger OXT/VP release from the dendritic stores (Fig. 7). These intra-SON peptides, then act on

receptors located on excitatory, but not inhibitory afferent terminals, to inhibit evoked glutamate release. This is achieved mainly by activation of intracellular processes that affect the function of high voltage activated calcium channels, mainly the N-type and to a lesser extent the P/Q-type. This retrograde excitatory synaptic depressant effect of the peptides will ensure that excessive afferent excitation is curtailed using the MCN dendrite as a sensor. The peptides also act directly on autoreceptors located on the MCNs to increase intracellular calcium mobilization (Lambert et al., 1994; Dayanithi et al., 1997) as well as to modulate the GABAA receptor–chloride channel complex leading to a decrease in the amplitude of both evoked and spontaneous IPSCs. Both the latter two postsynaptic effects of the peptides are func-

244

tionally excitatory and will provide a counterbalance allowing the dendrite to fine tune excitation and inhibition to arrive at an appropriate level and pattern of activity. Thus, the dendrite of the MCN of the SON, in addition to its routine, classic function of reception, integration and relay of afferent information, may also play a novel and unique function of sensing excitatory and inhibitory inputs unto themselves as well as producing retrograde/auto-transmitters that help shape these inputs and hence determine their output of peptide in the periphery. This evolving concept of bidirectional communication between neurons is not unique to the SON as numerous brain regions and synapses in the CNS are now known to engage in this type of communication (Cheramy et al., 1981; Glitsch et al., 1996; Zilberter et al., 1999; Kreitzer and Regehr, 2001). However, the MCNs of the SON have the unique cytoarchitecture that would enable thorough examination of the functional roles of retrograde neurotransmission in the neurohypophysis. Numerous aspects of the processes leading up to dendritic release and the role of these peptides in modulating afferent inputs are still unknown. First of all, we know very little about how dendritic release is organized and controlled. For example it is known that OXT/VP are synthesized as precursor molecules which then undergo a maturational process en route to the posterior pituitary (Dreifuss, 1975; Brownstein et al., 1980; Robinson et al., 1989; Roberts et al., 1991). Is the same process functional in the dendrites? Are the carrier neurophysin molecules released on a one-to-one basis with the neuropeptides as is seen at the axon terminals (Brownstein et al., 1980). We know now that the distribution of calcium channel subtypes on the axon terminals and the soma are different (Fisher and Bourque, 1996). It would be interesting to know if the roles of the various subtypes in dendritic release are similar to those that have been elegantly worked out at the axon terminal (Lemos et al., 1994; Wang et al., 1997, 1999). There is evidence that the release from these two sites can be differentially controlled, (Neumann et al., 1993, 1996), suggesting that there are unique modulatory sites at the dendritic release site. Indeed it would be interesting to know if the synaptic release proteins participating in peptide release at the dendritic and axonal endings are similar. With the availability of

specific blocking molecules, genetically modified animals and perhaps more in depth anatomical studies, these questions can be answered. Abbreviations AMPA EPSC GABA HFS IPSC MC MCN NMDA OXT PPF PPR PVN SON VP

α-Amino-3-hydroxy-5-methyl isoxazolepropionic acid excitatory postsynaptic current γ-aminobutyric acid high-frequency stimulation inhibitory postsynaptic current Manning compound magnocellular nucleus N-methyl-D-aspartate oxytocin paired-pulse facilitation paired-pulse ratio paraventricular nucleus supraoptic nucleus vasopressin

Acknowledgements This work was supported by Kuwait University Grants FDT113 and FPT 116 to S.B.K., MRC/CIHR grants to Q.J.P. M.H. was supported by H&SFC and CIHR. Q.J.P. is an AHFMR Senior Scientist. References Armstrong, W.E. (1995) Morphological and electrophysiological classification of hypothalamic supraoptic neurons. Prog. Neurobiol., 47: 291–339. Bains, J.S. and Ferguson, A.V. (1999) Activation of N-methylD -aspartate receptors evokes calcium spikes in the dendrites of rat hypothalamic paraventricular nucleus neurons. Neuroscience, 90(3): 885–891. Bargmann, W. and Scharrer, E. (1951) The site of origin of hormones of the posterior pituitary. Am. Sci., 39: 255–259. Bicknell, R.J. (1988) Optimizing release from peptide hormone secretory nerve terminals. J. Exp. Biol., 139: 51–65. Brinton, R.E., Wamsley, J.K., Gee, K.W., Wan, Y.P. and Yamamura, H.I. (1984a) [3 H]Oxytocin binding sites in the rat brain demonstrated by quantitative light microscopic autoradiography. Eur. J. Pharmacol., 102: 365–367. Brinton, R.E., Gee, K.W., Wamsley, J.K., Davis, T.P. and Yamamura, H.I. (1984b) Regional distribution of putative vasopressin receptors in rat brain and pituitary by quantitative autoradiography. Proc. Natl. Acad. Sci. USA, 81: 7248–7252. Brownstein, M.J., Russell, J.T. and Gainer, H. (1980) Synthesis,

245 transport, and release of posterior pituitary hormones. Science, 207(4429): 373–378. Brussaard, A.B., Kits, K.S. and de Vlieger, T.A. (1996) Postsynaptic mechanism of depression of GABAergic synapses by oxytocin in the supraoptic nucleus in immature rat. J. Physiol., 497: 495–507. Burbach, J.P. and Lebouille, J.L. (1983) Proteolytic conversion of arginine-vasopressin and oxytocin by brain synaptic membranes: characterization of formed peptides and mechanisms of proteolysis. J. Biol. Chem., 258: 1487–1494. Castel, M. and Morris, J.F. (1988) The neurophysin-containing innervation of the forebrain of the mouse. Neuroscience, 24: 937–966. Chaudhry, M.A., Dyball, R.E., Honda, K. and Wright, N.C. (1989) The role of interconnection between supraoptic nucleus and anterior third ventricular region in osmoregulation in the rat. J. Physiol., 410: 123–135. Cheramy, A., Leviel, V. and Glowinski, J. (1981) Dendritic release of dopamine in the substantia nigra. Nature, 289: 537– 542. Claybaugh, J.R. and Uyehara, C.F. (1993) Metabolism of neurohypophysial hormones. Ann. N.Y. Acad. Sci., 689: 250–268. Conde, H. (1992) Organization and physiology of substantia nigra. Exp. Brain Res., 88: 233–248. Conrad, L.C. and Pfaff, D.W. (1976) Efferents from medial basal forebrain and hypothalamus in the rat. II. An autoradiographic study of the anterior hypothalamus. J. Comp. Neurol., 169: 221–261. Currie, K.P. and Fox, A.P. (1997) Comparison of N- and P/Qtype voltage-gated calcium channel current inhibition. J. Neurosci., 17: 4570–4579. Dayanithi, G., Widmer, H. and Richard, P. (1996) Vasopressininduced intracellular Ca2+ increase in isolated rat supraoptic cells. J. Physiol., 490: 713–727. Di Scala-Guenot, D., Strosser, M.T. and Richard, P. (1987) Electrical stimulations of perifused magnocellular nuclei in vitro elicit Ca2+ -dependent, tetrodotoxin-insensitive release of oxytocin and vasopressin. Neurosci. Lett., 76: 209–214. Drake, C.T., Terman, G.W., Simmons, M.L., Milner, T.A., Kunkel, D.D., Schwartzkroin, P.A. and Chavkin, C. (1994) Dynorphin opioids present in dentate granule cells may function as retrograde inhibitory neurotransmitters. J. Neurosci., 14: 3736–3750. Dreifuss, J.J. (1975) A review on neurosecretory granules: their contents and mechanisms of release. Ann. N.Y. Acad. Sci., 248: 184–201. Fisher, T.E. and Bourque, C.W. (1996) Calcium-channel subtypes in the somata and axon terminals of magnocellular neurosecretory cells. Trends Neurosci., 19: 440–444. Freund-Mercier, M.J. and Richard, P. (1981) Excitatory effects of intraventricular injections of oxytocin on the milk ejection reflex in the rat. Neurosci. Lett., 23: 193–198. Glitsch, M., Llano, I. and Marty, A. (1996) Glutamate as a candidate retrograde messenger in interneurone–purkinje cell synapses of rat cerebellum. J. Physiol., 497: 531–537. Hermes, M.L., Ruijter, J.M., Klop, A., Buijs, R.M. and Renaud, L.P. (2000) Vasopressin increases GABAergic inhibition of

rat hypothalamic paraventricular nucleus neurons in vitro. J. Neurophysiol., 83(2): 705–711. Hirasawa, M., Kombian, S.B. and Pittman, Q.J. (2001) Oxytocin retrogradely inhibits evoked, but not miniature EPSCs in the rat supraoptic nucleus: role of N- and P/Q-type calcium channels. J. Physiol., 532(3): 595–607. Honda, K., Negoro, H., Higuchi, T. and Tadokoro, Y. (1987) Activation of neurosecretory cells by osmotic stimulation of anteroventral third ventricle. Am. J. Physiol., 252(6 Pt 2): R1039–R1045. Honda, K., Negoro, H., Dybal, R.E.J., Higuchi, T. and Takano, S. (1990) The osmoreceptor complex in the rat: evidence for interactions between the supraoptic and other diencephalic nuclei. J. Physiol., 431: 225–241. Hurbin, A., Boissin-Agasse, L., Orcel, H., Rabie, A., Joux, N., Desarmenien, M.G., Richard, P. and Moos, F.C. (1998) The V1a and V1b, but not V2, vasopressin receptor genes are expressed in the supraoptic nucleus of the rat hypothalamus, and the transcripts are essentially colocalized in the vasopressinergic magnocellular neurons. Endocrinology, 139: 4701–4707. Kabashima, N., Shibuya, I., Ibrahim, N., Ueta, Y. and Yamashita, H. (1997) Inhibition of spontaneous EPSCs and IPSCs by presynaptic GABAB receptors on rat supraoptic magnocellular neurons. J. Physiol., 504: 113–126. Klitenick, M.A., DeWitte, P. and Kalivas, P. (1992) Regulation of somatodendritic dopamine release in the ventral tegmental area by opioids and GABA: an in vivo microdialysis study. J. Neurosci., 12(7): 2623–2632. Kombian, S.B., Mouginot, D. and Pittman, Q.J. (1997) Dendritically released peptides act as retrograde modulators of afferent excitation in the supraoptic nucleus in vitro. Neuron, 19: 903– 912. Kombian, S.B., Mouginot, D., Hirasawa, M. and Pittman, Q.J. (2000) Vasopressin preferentially depresses excitatory over inhibitory synaptic transmission in the rat supraoptic nucleus in vitro. J. Neuroendocrinol., 12: 361–367. Kreitzer, A.C. and Regehr, W.G. (2001) Retrograde inhibition of presynaptic calcium influx by endogenous cannabinoids at excitatory synapses onto Purkinje cells. Neuron, 29(3): 717– 727. Kruszynski, M., Lammek, B., Manning, M., Seto, J., Haldar, J. and Sawyer, W.H. (1980) [1-beta-Mercapto-beta,beta-cyclopentamethylenepropionic acid),2-(O-methyl)tyrosine]argininevasopressin and [1-beta-mercapto-beta,beta-cyclopentamethylenepropionic acid)]arginine-vasopressin, two highly potent antagonists of the vasopressor response to arginine-vasopressin. J. Med. Chem., 23: 364–368. Lambert, R.C., Moos, F.C. and Richard, P. (1993) Action of endogenous oxytocin within the paraventricular or supraoptic nuclei: a powerful link in the regulation of the bursting pattern of oxytocin neurons during the milk-ejection reflex in rats. Neuroscience, 57: 1027–1038. Lambert, R.C., Dayanithi, G., Moos, F.C. and Richard, P. (1994) A rise in the intracellular Ca2+ concentration of isolated rat supraoptic cells in response to oxytocin. J. Physiol., 478: 275– 287. Lemos, J.R., Wang, G., Wang, X., Stuenkel, E.L., Nordmann, J.J.

246 and Treistman, S.N. (1994) Effects of toxins on Ca2+ currents and peptide release from nerve terminals. Ann. N.Y. Acad. Sci., 710: 11–29. Leng, G., Blackbrun, R.E., Dyball, R.E.J. and Russell, J.A. (1989) Role of anterior peri-third ventricular structures in the regulation of supraoptic neuronal activity and neurohypophysial hormone secretion in the rat. J. Neuroencrinol., 1: 35–46. Leng, G., Brown, C.H. and Russell, J.A. (1999) Physiological pathways regulating the activity of magnocellular neurosecretory cells. Prog. Neurobiol., 57: 625–655. Ludwig, M. and Landgraf, R. (1992) Does the release of vasopressin within the supraoptic nucleus of the rat brain depend upon changes in osmolality and Ca2+ /K+ ?. Brain Res., 576: 231–234. Mason, W.T., Ho, Y.W. and Hatton, G.I. (1984) Axon collaterals of the supraoptic neurons: anatomical and electrophysiological evidence for their existence in the lateral hypothalamus. Neuroscience, 11: 169–182. McKinley, M.J., Bicknell, R.J., Hards, D., McAllen, R.M., Vivas, L., Weisinger, R.S. and Oldfield, B.J. (1992) Efferent neural pathways of the lamina terminalis subserving osmoregulation. Prog. Brain Res., 91: 395–402. Morris, J.F. and Pow, D.V. (1993) Anatomical insights into the inputs and outputs from hypothalamic magnocellular neurons. Ann. N.Y. Acad. Sci., 689: 16–34. Mühlethaler, M. and Dreifuss, J.J. (1983) Excitation of hippocampal neurones by posterior pituitary peptides: vasopressin and oxytocin compared. Prog. Brain Res., 60: 147–151. Neumann, I., Ludwig, M., Engelmann, M., Pittman, Q.J. and Landgraf, R. (1993) Simultaneous microdialysis in blood and brain: oxytocin and vasopressin release in response to central and peripheral osmotic stimulation and suckling in the rat. Neuroendocrinology, 58: 637–645. Neumann, I., Douglas, A.J., Pittman, Q.J., Russell, J.A. and Landgraf, R. (1996) Oxytocin released within the supraoptic nucleus of the rat brain by positive feedback action is involved in parturition-related events. J. Neuroendocrinol., 8(3): 227– 233. Pow, D.V. and Morris, J.F. (1988) Release of neuropeptides from magnocellular neurons: does anatomical compartmentation have a functional significance? In: P.T. Pickering, J.B. Wakerley and A.J.S. Summerlee (Eds.), Neurosecretion: Cellular Aspects of Production and Release of Neuropeptides. Plenum, New York, pp. 113–122. Pow, D.V. and Morris, J.F. (1989) Dendrites of hypothalamic magnocellular neurons release neurohypophysial peptides by exocytosis. Neuroscience, 32: 435–439. Pucak, M.L. and Grace, A.A. (1994) Regulation of substantia nigra dopamine neurons. Crit. Rev. Neurobiol., 9: 67–89. Renaud, L.P. and Bourque, C.W. (1991) Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog. Neurobiol., 36: 131– 169. Richard, D. and Bourque, C.W. (1995) Synaptic control of rat supraoptic neurones during osmotic stimulation of the or-

ganum vasculosum lamina terminalis in vitro. J. Physiol., 489(2): 567–577. Roberts, M.M., Robinson, A.G., Hoffman, G.E. and Fitzsimmons, M.D. (1991) Vasopressin transport regulation is coupled to the synthesis rate. Neuroendocrinology, 53(4): 416–422. Robinson, A.G., Roberts, M.M., Evron, W.A., Janocko, L.E. and Hoffman, G.E. (1989) Total translation of vasopressin and oxytocin in neurohypophysis of rats. Am. J. Physiol., 257(1 Pt 2): R109–R117. Sabatini, B.L. and Regehr, W.G. (1997) Control of neurotransmitter release by presynaptic waveform at the granular cell to Purkinje cell synapse. J. Neurosci., 17: 3425–3435. Saphier, D. and Feldman, S. (1987) Electrophysiology of supraoptico-paraventricular nucleus connections in the rat. Exp. Brain Res., 69: 60–66. Stern, J.E. and Armstrong, W.E. (1995) Electrophysiological differences between oxytocin and vasopressin neurons recorded in female rats in vitro. J. Physiol., 448: 701–708. Stuart, G., Spruston, N., Sakmann, B. and Hausser, M. (1997) Action potential initiation and back propagation in neurons of the mammalian CNS. Trends Neurosci., 20: 125–131. Thibonnier, M., Berti-Mattera, L.N., Dulin, N., Conarty, D.M. and Mattera, R. (1998) Signal transduction pathways of the huma V1-vascular, V2-renal, V3-pituitary vasopressin and oxytocin receptors. Prog. Brain Res., 119: 147–161. Tribollet, E., Armstrong, W.E., Dubois-Dauphin, M. and Dreifuss, J.J. (1985) Extra-hypothalamic afferent inputs to the supraoptic nucleus area of the rat as determined by retrograde and anterograde tracing techniques. Neuroscience, 15(1): 135– 148. Van Leeuwen, F.W., Van Heerikhuize, J., Van der Muelen, G. and Wolters, P. (1985) Light microscopic autoradiographic localization of [3 H]oxytocin binding sites in the rat brain, pituitary and mammary gland. Brain Res., 359: 320–325. Wang, G., Dayanithi, G., Kim, S., Hom, D., Nadasdi, L., Kristipati, R., Ramachandran, J., Stuenkel, E.L., Nordmann, J.J., Newcomb, R. and Lemos, J.R. (1997) Role of Q-type Ca2+ channels in vasopressin secretion from neurohypophysial terminals of the rat. J. Physiol., 502(Pt 2): 351–363. Wang, G., Dayanithi, G., Newcomb, R. and Lemos, J.R. (1999) An R-type Ca(2+) current in neurohypophysial terminals preferentially regulates oxytocin secretion. J. Neurosci., 19(21): 9235–9241. Wu, L.G. and Saggau, P. (1997) Presynaptic inhibition of elicited neurotransmitter release. Trends Neurosci., 20: 204–212. Yamashita, H., Okuya, S., Inenaga, K., Kasai, M., Uesugi, S., Kannan, H. and Kaneko, T. (1987) Oxytocin predominantly excites putative oxytocin neurons in the rat supraoptic nucleus in vitro. Brain Res., 416: 364–368. Zilberter, Y., Kaiser, K.M. and Sakmann, B. (1999) Dendritic GABA release depresses excitatory transmission between layer 2/3 pyramidal and bitufted neurons in rat neocortex. Neuron, 24: 979–988. Zucker, R.S. (1989) Short-term plasticity. Annu. Rev. Neurosci., 12: 13–31.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 19

The active role of dendrites in the regulation of magnocellular neurosecretory cell behavior Mike Ludwig 1,∗ , Nancy Sabatier 1 , Govindan Dayanithi 2, John A. Russell 1 and Gareth Leng 1 1

Department of Biomedical Sciences, University of Edinburgh Medical School, George Square, Edinburgh EH8 9XD, UK 2 Department of Neurobiology, U432, INSERM, University of Montpellier II, Place Eugene Bataillon, F-34094 Montpellier Cedex 5, France

Abstract: The interactions of the dendritically released neuropeptides vasopressin and oxytocin with co-released neuroactive substances such as opioids and nitric oxide are reviewed. Endogenous opioids regulate magnocellular neurons at the level of the supraoptic nucleus and the relationship of dendritically released peptides and co-released opioids seems to be dependent on the stimulus given and the physiological state of the animal. Nitric oxide has a prominent inhibitory action on supraoptic neurons and these actions are predominantly mediated indirectly by GABA inputs. The role of these co-released neuroactive substances in differentially regulated release of neuropeptides from dendrites versus distant axon terminals has to be determined in more detail. A picture emerges in which release of vasopressin and oxytocin from different anatomical compartments of a single neuron may arise from different intracellular secretory pools and their preparation before release. Keywords: Supraoptic nucleus; Hypothalamus; Vasopressin; Oxytocin; Nitric oxide; Opioid; Dendritic release

Introduction Classically, neurotransmission is achieved by release of transmitters from axon terminals to activate postsynaptic receptors. However, it has recently become clear that neurotransmitters are also released from dendrites to produce a local modulation of synaptic transmission. Dendritic release of neurotransmitters, such as glutamate (Cox et al., 1998), GABA (Zilberter et al., 1999), dopamine (Geffen et al., 1976), serotonin (Hery and Ternaux, 1981) and dynorphin (Simmons et al., 1995), has been shown for several neuronal populations. ∗ Correspondence

to: M. Ludwig, Department of Biomedical Sciences, University of Edinburgh Medical School, George Square, Edinburgh EH8 9XD, UK. Tel.: +44131-650-3275; Fax: +44-131-650-3711; E-mail: mike. [email protected]

The cell bodies and dendrites of the magnocellular vasopressin and oxytocin neurons are located within the hypothalamic supraoptic (SON) and paraventricular nuclei (PVN). Dendritic release of their neuropeptides has been demonstrated in vitro and in vivo (for review see Landgraf, 1995; Ludwig, 1998; Morris et al., 2000), resulting in pre- and postsynaptic regulation of the electrical activity (Brussaard et al., 1996; Kombian et al., 1997; Hermes et al., 2000), a cell-type specific receptor-mediated rise in intracellular Ca2+ (Dayanithi et al., 2000) and reorganization of the ultrastructure of the nuclei (Theodosis et al., 1986a). Several recent studies indicate striking spatial and temporal differences between the release profiles from axon terminals and dendrites (Moos et al., 1989; Wotjak et al., 1998; Engelmann et al., 1999). However, the mechanisms of this differentially regulated release from dendrites are not well understood. This dendritic release seems

248

to be of great importance in the regulation of the milk ejection reflex, since it has been shown that oxytocin release from dendrites occurs before the onset of milk ejection bursts (Moos et al., 1989). The rate of neurohypophysial peptide secretion is not only regulated by locally released vasopressin and oxytocin but also by co-localized and co-released neuroactive substances such as opioids and nitric oxide (NO). Regulation of activity by dendritic release of co-peptides has been observed for several classes of neuron, such as serotonin neurons within the raphe nuclei (Hery et al., 1982) and dopamine neurons of the substantia nigra (Cheramy et al., 1981). In the latter case, modulation of the activity of these neurons seems to depend on co-released factors since acetylcholinesterase, which is also secreted by nigrostriatal cells (Llinas and Greenfield, 1987), hyperpolarizes these cells by a mechanism independent of acetylcholine hydrolysis (Greenfield et al., 1988). Role of co-released opioids Oxytocin cells synthesize enkephalins and both vasopressin and oxytocin neurons synthesize the κ-opioid agonist, dynorphin. (Watson et al., 1982). Dynorphin is found co-packaged with vasopressin in neurosecretory vesicles in the neural lobe (Whitnall et al., 1983). κ-Receptors are expressed in both oxytocin and vasopressin cell bodies. The receptor at which dynorphin is active has been visualized in the membrane of vasopressinergic neurosecretory vesicles in the cell bodies and axon terminals (Shuster et al., 1999, 2000). Thus it seems probable that dynorphin is co-released upon dendritic exocytosis of vasopressin into the SON. We used direct osmotic stimulation via microdialysis to elicit intranuclear and systemic oxytocin release in the presence and absence of the μ-agonist morphine, or the general opioid antagonist, naloxone (Munro et al., 1994). We showed that dendritic oxytocin release in response to direct osmotic stimulation is unaffected by intravenous morphine, a μ-opioid agonist, or the opioid receptor antagonist naloxone, while oxytocin release into the systemic circulation is inhibited by the former and potentiated by the latter. The failure of the opioid antagonist to modify the intranuclear release of oxytocin in re-

sponse to direct osmotic stimulation indicates that dendritic oxytocin release is not markedly restrained by any co-released endogenous opioid peptide. However, the presence of an endogenous opioid inhibition of dendritic release has been shown in isolated SON and PVN in vitro (Ingram et al., 1996). Furthermore, local application of naloxone has been shown to increase oxytocin release measured in microdialysis samples from the SON of late pregnant rats (Douglas et al., 1995). Thus, the relationship between dendritically released peptides and co-released opioids seems to be dependent on the stimulus given and the physiological state of the animal. Oxytocin release from dendrites within the SON appears to interact with at least some of the effects of opioids on oxytocin neurons. Upon chronic administration of morphine, oxytocin neurons become both tolerant to, and dependent upon, this opiate alkaloid (Russell et al., 1995). Tolerance is seen as a reduction in the effectiveness of morphine inhibition over time, while dependence is manifested as increased activity following withdrawal of morphine. Intranuclear oxytocin release is increased following morphine withdrawal induced by intravenous administration of naloxone to morphine-dependent rats (Russell et al., 1992) and i.c.v. administration of an oxytocin antagonist reduces the magnitude of this excitation (Brown et al., 1997), indicating that dendritic release of oxytocin may contribute to the magnitude of the withdrawal excitation of oxytocin neurons. To investigate the potential importance of this intranuclear release of oxytocin in the generation of the withdrawal excitation of oxytocin neurons, we studied whether the mechanisms which generate morphine dependence reside within the oxytocin neurons themselves or whether these neurons simply follow excitation of their afferent inputs upon morphine withdrawal. We have shown that withdrawal excitation of oxytocin cells can be evoked by administration of naloxone directly into the SON of morphine-dependent rats (Ludwig et al., 1997). Thus, at least some of the mechanisms which generate morphine withdrawal excitation are localized to the SON. However, these effects may be postsynaptic upon the oxytocin neurons themselves, or pre-synaptic upon the axon terminals of their afferent inputs. Oxytocin neurons receive a prominent noradrenergic projection from the A2 cell group in

249

the nucleus of the solitary tract (NTS) (Raby and Renaud, 1989) as well as other non-noradrenergic inputs from forebrain regions (Hatton, 1990; Renaud and Bourque, 1991). However, very few of the cells that project to the SON from forebrain regions are activated by morphine withdrawal, while approximately 10% of NTS neurons are activated at this time (Murphy et al., 1997). We have shown that acute pharmacological blockade of noradrenergic transmission can reduce the magnitude of withdrawal excitation in oxytocin neurons (Brown et al., 1998a). Superficially, this provides powerful evidence that a noradrenergic drive on oxytocin neurons is essential to evoke morphine withdrawal excitation in these neurons. However, neurotoxic destruction of noradrenergic systems does not alter the magnitude of the withdrawal excitation in oxytocin neurons (Brown et al., 1998a). The most probable explanation of these apparently inconsistent results is that oxytocin neurons develop morphine dependence separately from their afferent inputs but that full expression of morphine withdrawal excitation requires excitatory synaptic inputs to maintain the underlying excitability of the oxytocin neurons. Thus, it appears that the mechanisms which generate morphine dependence in oxytocin neurons reside in the SON and that dendritic release of oxytocin may contribute to the full expression of morphine withdrawal excitation by oxytocin neurons. After demonstrating that i.v. injection of the μagonist morphine and the κ-agonist U50488 reduced the firing rate of identified oxytocin neurons and that this inhibition was completely reversed by retrodialysis of the opioid antagonist naloxone, indicating that local opioids could inhibit within the SON (Ludwig et al., 1997; Fig. 1), we focused our research on opioid effects on vasopressin cells. The dendritic neurosecretory vesicles of vasopressin neurons contain the same opioid peptide, dynorphin, as found in their neurohypophysial vesicles. Dendritically released vasopressin is predominantly inhibitory to vasopressin neurons (Ludwig and Leng, 1997; Pittman et al., 2000). We studied the contribution which the co-localized dynorphin may make to the autoregulation of vasopressin neuronal activity by dendritic release of vasopressin (Fig. 1). Retrodialysis of U50488 inhibits vasopressin neurons and retrodialysis of a κ-receptor antagonist into the SON

increases the activity of phasic vasopressin neurons by increasing the activity quotient and the intra-burst firing rate (Brown et al., 1998b). This, coupled with the recent observation that κ-receptors are localized to the membranes of vasopressin-containing vesicles within the SON (Shuster et al., 1999, 2000), indicates that dendritically released dynorphin may be a potent regulator of activity in vasopressin neurons. However, the pre- and post-synaptic actions of vasopressin seem to be independent of co-released endogenous opioid peptides, since vasopressin-induced inhibition of vasopressin neurons is not reversed by the opioid antagonist, naloxone, even at high concentrations (Ludwig and Leng, 1998). Thus, it is probable that dendritically released vasopressin and dynorphin act in parallel to curtail the activity of vasopressin neurons. κ-Opioid activation reduces the amplitude of the depolarizing after-potential in SON neurons and this may underlie the reduction in burst duration of vasopressin cells by dendritic release of an endogenous κ-agonist (Brown et al., 1999). In conclusion, it appears that the actions of dendritically released vasopressin do not rely upon co-released endogenous opioid peptides. The role of nitric oxide There is growing evidence that nitric oxide (NO) also functions as a local modulator of magnocellular neuronal activity. Neuronal NO synthase (NOS) is expressed densely in the SON and PVN (Arevalo et al., 1992), where it is colocalized with oxytocin- and vasopressin-synthesizing neurons, and its expression is functionally regulated (Sagar and Ferriero, 1987; Bredt et al., 1990; Pasqualotto et al., 1991; Vincent and Kimura, 1992). In the SON and PVN, the expression of neuronal NOS mRNA is increased in response to osmotic stimuli (Kadowaki et al., 1994; Villar et al., 1994; Ueta et al., 1995) and hypovolemia (Ueta et al., 1998), and staining for NADPH-diaphorase changes during late pregnancy and parturition (Okere and Higuchi, 1996). In rats and in humans, NO inhibits oxytocin secretion from the posterior pituitary (Chiodera et al., 1994), however direct effects of NO on dendritic release have not been studied yet. We have shown that systemic administration of NOS inhibitors led to a facilitation of oxytocin release evoked by hyperosmotic stimu-

250

Fig. 1. Local opioid inhibition of SON neurons. (A) Intravenous (i.v.) injection of morphine (MOR, 1 mg/kg) markedly reduced the spontaneous activity of an identified (excited by CCK, 20 μg/kg, i.v.) oxytocin neuron. Retrodialysis of naloxone (NLX, 2 mM) onto the supraoptic nucleus restored the spontaneous activity of this neuron and reduced the inhibitory effect of subsequent morphine injection. (B) Microdialysis administration (retrodialysis) of the κ-opioid agonist U50488 (1 mM) completely inhibits the activity of the recorded vasopressin neuron. Retrodialysis of naloxone (2 mM) onto the SON restores the spontaneous activity, indicating that the opioid inhibition occurs within the SON. (C) Retrodialysis of the κ-antagonist nor-binaltorphimine (BNI, 200 μg/ml) onto the SON increases the burst duration indicating tonic effects on vasopressin cells by endogenous κ-opioids. Adapted from Ludwig et al., 1997; Brown et al., 1998b.

lation. Direct application of the NO donor sodium nitroprusside to the supraoptic nucleus by in vivo microdialysis inhibited the electrical activity of both oxytocin neurons and vasopressin neurons, whereas direct application of an NOS inhibitor increased electrical activity, indicating that endogenous NO acts within the supraoptic nucleus to inhibit neuronal activity (Fig. 2). However, during late pregnancy, the influence of endogenous NO is dramatically downregulated, reflected by a reduced expression of nNOS mRNA in these neurons and a loss of efficacy of NOS inhibitors on stimulus-evoked oxytocin release. This down-regulation may cause the oxytocin system to become more excitable at term, resulting in

the capacity for greater release of oxytocin during parturition (Srisawat et al., 2000). Electrophysiological studies in vitro (Liu et al., 1997; Ozaki et al., 2000) and as mentioned above in vivo (Srisawat et al., 2000) have shown that the NO donor sodium nitroprusside (SNP) and the NO precursor L-arginine inhibit SON neurons, whereas the NOS inhibitor L-NAME and the NO scavenger hemoglobin excite them. One mechanism by which NO influences signaling in the central nervous system is by modulating neurotransmitter release, including, in particular, the inhibitory neurotransmitter — GABA (Segovia et al., 1994; Ohkuma et al., 1996). GABAergic synapses comprise about 40% of

251

Fig. 2. In vivo and in vitro recordings indicating inhibitory action of nitric oxide (NO) on vasopressin and oxytocin neurons via their GABAergic inputs. (A) Inhibition of a continuously firing putative vasopressin neuron during microdialysis administration (retrodialysis) of the NO precursor L-arginine (100 mM); (B) Retrodialysis of the NO synthase inhibitor, L-NNA (10 mM) induced an increase in the firing rate of this vasopressin neuron (inhibited by CCK) in vivo; (C) the NO donor SNP (50 mM) retrodialysed onto the SON also inhibits the activity of the identified (CCK, 20 μg/kg, i.v.) oxytocin neuron and (D) pre-treatment by retrodialysis of the GABAA antagonist bicuculline (2 mM) results in a reduced inhibitory effect of SNP. GABAA miniature inhibitory postsynaptic currents (mIPSCs) in a vasopressin cell in vitro before (E) and during (F) bath application of SNP (100 μM). Data modified from Srisawat et al., 2000; Stern and Ludwig, 2001.

all synaptic contacts in the SON (Theodosis et al., 1986b) and GABA plays a key role in controlling the firing activity of both oxytocin and vasopressin neurons (Wuarin and Dudek, 1993; Moos, 1995). The NO and GABA systems seem to interact strongly in the PVN (Krukoff, 1999). For instance, perfusion of the PVN with NO-containing medium by microdialysis or microinjection of SNP increases local GABA release (Horn et al., 1994). NO has also been shown to inhibit renal sympathetic outflow by modulating local GABA activity within the PVN (Zhang and Patel, 1998). Similarly, NMDA receptor activation in the PVN increased GABAergic activity on magnocellular neuroendocrine neurons, an effect mediated by local NO production (Bains and Ferguson, 1997). To study further NO–GABA interactions, and their

physiological relevance in modulating the activity of SON magnocellular neurons, we combined in vitro and in vivo electrophysiological studies on identified magnocellular neurons. The results suggest that nitric oxide inhibition of neuronal excitability in oxytocin and vasopressin neurons involves preand post-synaptic potentiation of GABAergic synaptic activity in the SON (Stern and Ludwig, 2001; Fig. 2). The mechanisms of opioid dependence and withdrawal excitation in magnocellular neurons are not well understood. One possible adaptation involves attenuation of inhibitory mechanisms: NO generation is a candidate. To investigate the involvement of endogenous NO in morphine dependence of SON oxytocin neurons we completed a series of blood

252

sampling and electrophysiology experiments. The data indicate that NO mechanisms restraining oxytocin neurons are unchanged in morphine dependence. During withdrawal excitation, endogenous NO restrains oxytocin neurons; however, the restraint may be predominantly post-synaptic rather than on GABA terminals, as in naïve rats. Although there is a GABA inhibition of oxytocin neurons during withdrawal excitation, this loss of NO action on GABA terminals may contribute to excitation (Ludwig et al., 2001). Vasopressin and oxytocin It is now well accepted that both peptides are released from the dendrites of the magnocellular neurons. However, depending upon the stimulus given, neurons within the magnocellular nuclei and other brain areas are apparently capable of regulating their local dendritic and distant axonal terminal release of

neuroactive substances in either a co-ordinated or independent manner. Differential release of neurotransmitters from different compartments of a single neuron requires subtle regulatory mechanisms and the contributions of nitric oxide and endogenous opioids to the regulation of these mechanisms is currently not well understood. Part of these differences may be explained by local feedback mechanisms induced by the peptides themselves. Autoregulation of magnocellular neurons and a number of post- and, recently, pre-synaptic effects have been reported (Fig. 3). For dendritically released oxytocin modulation of glutamatergic inputs seems to be pre-synaptic, acting to suppress glutamate-mediated EPSPs on oxytocin neurons. Interestingly, vasopressin acts mainly post-synaptically, enhancing EPSPs in oxytocin neurons and depressing them in vasopressin cells (Kombian et al., 1997; Pittman et al., 2000). The in vitro effect of vasopressin is consistent with recent in vivo reports showing

Fig. 3. Dendritic release of neuroactive substances; oxytocin, opioids and nitric oxide. (1) Dendritic oxytocin release results, at least in part, from a receptor-mediated (OTR), G protein-coupled positive autoregulatory action of oxytocin. (2) Dendritic exocytosis of oxytocin contributes to the coordination of the electrical activity within the four nuclei during birth and lactation. (3) Oxytocin also modulates pre- and post-synaptically afferent GABAergic and glutamatergic inputs. (4) Furthermore, oxytocin induces changes in the intracellular calcium concentration resulting in vesicle priming. (5) In combination with other factors, such as steroids, oxytocin induces morphological changes (e.g. retraction of glial cells between the neurons, increases dendrodendritic interactions) within the nuclei which favors synchronization of oxytocin neurons during reproduction. Pre- and post-synaptic regulation of the electrical activity of the magnocellular neurons also occurs through dendritically released endogenous opioids (enkephalin, dynorphin) co-packaged with oxytocin, and which bind to specific μ- and κ-receptors (OR), respectively. Finally, nitric oxide released from dendrites/cell bodies acts as a local autoregulatory signal.

253

that vasopressin predominantly suppresses the firing rate of vasopressin neurons (Ludwig and Leng, 1997). Interestingly, Gouzenes et al. have shown that vasopressin inhibits or excites vasopressin neurons depending on their ongoing electrical activity, so that fast-firing neurons are slowed, and slow-firing neurons are excited (Gouzenes et al., 1998). This may be due to activation of different intracellular secondmessenger pathways, since intracellular mechanisms activated by binding of vasopressin to its receptors increases intracellular Ca2+ concentrations via both adenylate cyclase- and phospholipase C-coupled intracellular transduction pathways (Sabatier et al., 1998). An alternative explanation for the excitatory effects of vasopressin could be in the modulation of different synaptic inputs. We have recently shown that dendritically released vasopressin facilitates noradrenaline release from terminals in the supraoptic nucleus (Ludwig et al., 2000). Alternatively, differences in peptide release from dendrites versus axon terminals may be due to: (1) independent peptide biosynthesis and storage of neurosecretory granules in dendrites; (2) site-specific synaptic contacts on dendrites or soma; (3) compartmentalization of receptor populations and/or second-messenger systems between dendrites and soma; (4) mediation of changes in intracellular Ca2+ via second-messenger systems in dendrites and by voltage-dependent channels in the soma; and (5) dissociation of electrical activity in axons and dendrites. We have focused our research on the possibility that dendritic release is dissociated from the electrical activity of magnocellular neurons and mediated by large changes in intracellular Ca2+ . Antidromic activation of magnocellular SON neurons has shown that even intense spike activity does not lead to vasopressin and oxytocin release from dendrites. Therefore we are currently focusing on the effects of intracellular calcium changes on dendritic peptide release. Researchers in Montpellier have shown that application of oxytocin or vasopressin to isolated magnocellular neurons produces a cell-type specific rise in [Ca2+ ]i . Oxytocin mobilizes Ca2+ mainly from thapsigargin-sensitive intracellular Ca2+ stores, whereas the response induced by vasopressin requires an influx of external Ca2+ as well as mobilization of thapsigargin-sensitive intracellular Ca2+ stores (for review see Dayanithi et al., 2000). The

consequences of peptide-induced changes in [Ca2+ ]i in magnocellular neurons are not known, but may have rapid effects on both systemic peptide secretion (by altering electrical activity) and dendritic peptide secretion (Moos et al., 1984; Wotjak et al., 1994). A rise in [Ca2+ ]i may also have sustained effects, e.g. inducing the accumulation of peptidecontaining vesicles beneath the plasma membrane and subsequently priming release (Ashery et al., 2000). Priming is an essential and rate-limiting step in secretion from neurons and neuroendocrine cells (Sudhof, 1995; Brose et al., 2000) and may be required before action potentials that propagate back into the dendrites can trigger substantial peptide release. Ongoing experiments are investigating this hypothesis. Abbreviations GABA NO NOS NTS PVN SON

γ-aminobutyric acid nitric oxide nitric oxide synthase nucleus of the solitary tract paraventricular nucleus supraoptic nucleus

References Arevalo, R., Sanchez, F., Alonso, J.R., Carretero, J., Vazquez, R. and Aijon, J. (1992) NADPH-diaphorase activity in the hypothalamic magnocellular neurosecretory nuclei of the rat. Brain Res. Bull., 28: 599–603. Ashery, U., Varoqueaux, F., Voets, T., Betz, A., Thakur, P., Koch, H., Neher, E., Brose, N. and Rettig, J. (2000) Munc13-1 acts as a priming factor for large dense-core vesicles in bovine chromaffin cells. EMBO J., 19: 3586–3596. Bains, J.S. and Ferguson, A.V. (1997) Nitric oxide depolarizes type II paraventricular nucleus neurons in vitro. Neuroscience, 79: 149–159. Bredt, D.S., Hwang, P.M. and Snyder, S.H. (1990) Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature, 347: 768–770. Brose, N., Rosenmund, C. and Rettig, J. (2000) Regulation of transmitter release by Unc-13 and its homologues. Curr. Opin. Neurobiol., 10: 303–311. Brown, C.H., Munro, G., Johnstone, L.E., Robson, A.C., Landgraf, R. and Russell, J.A. (1997) Oxytocin neurone autoexcitation during morphine withdrawal in anaesthetized rats. NeuroReport, 8: 951–955. Brown, C.H., Murphy, N.P., Munro, G., Ludwig, M., Bull, P.M., Leng, G. and Russell, J.A. (1998a) Interruption of central

254 noradrenergic pathways and morphine withdrawal excitation of oxytocin neurones in the rat. J. Physiol., 507: 831–842. Brown, C.H., Ludwig, M. and Leng, G. (1998b) κ-Opioid regulation of neuronal activity in the rat supraoptic nucleus in vivo. J. Neurosci., 18: 9480–9488. Brown, C.H., Ghamari-Langroudi, M., Leng, G. and Bourque, C.W. (1999) κ-Opioid receptor activation inhibits post-spike depolarizing after-potentials in rat supraoptic nucleus neurones in vitro. J. Neuroendocrinol., 11: 825–828. Brussaard, A.B., Kits, K.S. and de Vlieger, T.A. (1996) Postsynaptic mechanism of depression of GABAergic synapses by oxytocin in the supraoptic nucleus of immature rat. J. Physiol., 497: 495–507. Cheramy, A., Leviel, V. and Glowinski, J. (1981) Dendritic release of dopamine in the substantia nigra. Nature, 289: 537– 542. Chiodera, P., Volpi, R. and Coiro, V. (1994) Inhibitory control of nitric oxide on the arginine-vasopressin and oxytocin response to hypoglycemia in normal men. NeuroReport, 5: 1822–1824. Cox, C.L., Zhou, Q. and Sherman, S.M. (1998) Glutamate locally activates dendritic outputs of thalamic interneurons. Nature, 394: 478–482. Dayanithi, G., Sabatier, N. and Widmer, H. (2000) Intracellular calcium signaling in magnocellular neurones of the rat supraoptic nucleus: understanding the autoregulatory mechanisms. Exp. Physiol., 85: 75S–84S. Douglas, A.J., Neumann, I., Meeren, H.K., Leng, G., Johnstone, L.E., Munro, G. and Russell, J.A. (1995) Central endogenous opioid inhibition of supraoptic oxytocin neurons in pregnant rats. J. Neurosci., 15: 5049–5057. Engelmann, M., Ebner, K., Landgraf, R., Holsboer, F. and Wotjak, C.T. (1999) Emotional stress triggers intrahypothalamic but not peripheral release of oxytocin in male rats. J. Neuroendocrinol., 11: 867–872. Geffen, L.B., Jessell, T.M., Cuello, A.C. and Iversen, L.L. (1976) Release of dopamine from dendrites in rat substantia nigra. Nature, 260: 258–260. Gouzenes, L., Desarménien, M.G., Hussy, N., Richard, P. and Moos, F.C. (1998) Vasopressin regularizes the phasic firing pattern of rat hypothalamic magnocellular neurons. J. Neurosci., 18: 1879–1885. Greenfield, S.A., Jack, J.J., Last, A.T. and French, M. (1988) An electrophysiological action of acetylcholinesterase independent of its catalytic site. Exp. Brain Res., 70: 441–444. Hatton, G.I. (1990) Emerging concepts of structure–function dynamics in adult brain: the hypothalamo-neurohypophysial system. Prog. Neurobiol., 34: 437–504. Hermes, M.L.H.J., Ruijter, J.M., Klop, A., Buijs, R.M. and Renaud, L.P. (2000) Vasopressin increases GABAergic inhibition of rat hypothalamic paraventricular nucleus neurons in vitro. J. Neurophysiol., 83: 705–711. Hery, F. and Ternaux, J.P. (1981) Regulation of release processes in central serotoninergic neurons. J. Physiol. (Paris), 77: 287– 301. Hery, F., Faudon, M. and Ternaux, J.P. (1982) In vivo release of serotonin in two raphe nuclei (raphe dorsalis and magnus) of the cat. Brain Res. Bull., 8: 123–129.

Horn, T., Smith, P.M., McLaughlin, B.E., Bauce, L., Marks, G.S., Pittman, Q.J. and Ferguson, A.V. (1994) Nitric oxide actions in paraventricular nucleus: cardiovascular and neurochemical implications. Am. J. Physiol., 266: R306–R313. Ingram, C.D., Kavadas, V., Thomas, M.R.M. and Threapleton, J.D. (1996) Endogenous opioid control of somatodendritic oxytocin release from the hypothalamic supraoptic and paraventricular nuclei in vitro. Neurosci. Res., 25: 17–24. Kadowaki, K., Kishimoto, J., Leng, G. and Emson, P.C. (1994) Up-regulation of nitric oxide synthase (NOS) gene expression together with NOS activity in the rat hypothalamo-hypophysial system after chronic salt loading: evidence of a neuromodulatory role of nitric oxide in arginine vasopressin and oxytocin secretion. Endocrinology, 134: 1011–1017. Kombian, S.B., Mouginot, D. and Pittman, Q.J. (1997) Dendritic released peptides act as retrograde modulators of afferent excitation in the supraoptic nucleus in vitro. Neuron, 19: 903– 912. Krukoff, T.L. (1999) Central actions of nitric oxide in regulation of autonomic functions. Brain Res. Rev., 30: 52–65. Landgraf, R. (1995) Intracerebrally released vasopressin and oxytocin: measurement, mechanisms and behavioral consequences. J. Neuroendocrinol., 7: 243–253. Liu, Q.S., Jia, Y.S. and Ju, G. (1997) Nitric oxide inhibits neuronal activity in the supraoptic nucleus of the rat hypothalamic slices. Brain Res. Bull., 43: 121–125. Llinas, R.R. and Greenfield, S.A. (1987) On-line visualization of dendritic release of acetylcholinesterase from mammalian substantia nigra neurons. Proc. Natl. Acad. Sci. USA, 84: 3047–3050. Ludwig, M. (1998) Dendritic release of vasopressin and oxytocin. J. Neuroendocrinol., 10: 881–895. Ludwig, M. and Leng, G. (1997) Autoinhibition of supraoptic nucleus vasopressin neurons in vivo — a combined retrodialysis/electrophysiological study in rats. Eur. J. Neurosci., 9: 2532–2540. Ludwig, M. and Leng, G. (1998) Intrahypothalamic vasopressin release. An inhibitor of systemic vasopressin secretion?. Adv. Exp. Med. Biol., 449: 163–173. Ludwig, M., Brown, C.H., Russell, J.A. and Leng, G. (1997) Local opioid inhibition and morphine dependence of supraoptic nucleus oxytocin neurones in the rat in vivo. J. Physiol., 505: 145–152. Ludwig, M., Onaka, T. and Yagi, K. (2000) Vasopressin regulation of noradrenaline release within the supraoptic nucleus. J. Neuroendocrinol., 12: 477–479. Ludwig, M., Blackburn-Munro, G., Cohen, H.D., Brown, C.H., Bull, P.M., Leng, G. and Russell, J.A. (2001) The role of nitric oxide in morphine withdrawal excitation of oxytocin neurones in rats. Br. Neurosci. Assoc., 16: Abstract 30.17–86. Moos, F.C. (1995) GABA-induced facilitation of the periodic bursting activity of oxytocin neurones in suckled rats. J. Physiol., 488: 103–114. Moos, F., Freund-Mercier, M.J., Guerne, Y., Guerne, J.M., Stoeckel, M.E. and Richard, P. (1984) Release of oxytocin and vasopressin by magnocellular nuclei in vitro: specific fa-

255 cilitatory effect of oxytocin on its own release. J. Endocrinol., 102: 63–72. Moos, F., Poulain, D.A., Rodriguez, F., Guerne, Y., Vincent, J.D. and Richard, P. (1989) Release of oxytocin within the supraoptic nucleus during the milk ejection reflex in rats. Exp. Brain Res., 76: 593–602. Morris, J.F., Christian, H., Ma, D. and Wang, H. (2000) Dendritic secretion of peptides from hypothalamic magnocellular neurosecretory neurones: a local dynamic control system and its functions. Exp. Physiol., 85: 131S–138S. Munro, G., Ludwig, M., Landgraf, R. and Russell, J.A. (1994) Opioids influence neurohypophysial but not central oxytocin release following direct hyperosmotic stimulation of the supraoptic nucleus in urethane-anaesthetised rats. Neuropeptides, 27: 121–127. Murphy, N.P., Onaka, T., Brown, C.H. and Leng, G. (1997) The role of afferent inputs to supraoptic nucleus oxytocin neurons during naloxone-precipitated morphine withdrawal in the rat. Neuroscience, 80: 567–577. Ohkuma, S., Katsura, M., Chen, D.Z., Narihara, H. and Kuriyama, K. (1996) Nitric oxide-evoked [3H]gammaaminobutyric acid release is mediated by two distinct release mechanisms. Mol. Brain Res., 36: 137–144. Okere, C.O. and Higuchi, T. (1996) Down-regulation of endogenous nitric oxide synthase in late-pregnancy and parturition in the rat hypothalamic magnocellular neurons and neurohypophysis. Neurosci. Lett., 220: 133–136. Ozaki, M., Shibuya, I., Kabashima, N., Isse, T., Noguchi, J., Ueta, Y., Inoue, Y., Shigematsu, A. and Yamashita, H. (2000) Preferential potentiation by nitric oxide of spontaneous inhibitory postsynaptic currents in rat supraoptic neurones. J. Neuroendocrinol., 12: 273–282. Pasqualotto, B.A., Hope, B.T. and Vincent, S.R. (1991) Citrulline in the rat brain: immunohistochemistry and coexistence with NADPH-diaphorase. Neurosci. Lett., 128: 155–160. Pittman, Q.J., Hirasawa, M., Mouginot, D. and Kombian, S.B. (2000) Neurohypophysial peptides as retrograde transmitters in the supraoptic nucleus of the rat. Exp. Physiol., 85: 139S– 143S. 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. Renaud, L.P. and Bourque, C.W. (1991) Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog. Neurobiol., 36: 131– 169. Russell, J.A., Neumann, I. and Landgraf, R. (1992) Oxytocin and vasopressin release in discrete brain areas after naloxone in morphine-tolerant and -dependent anesthetized rats: push– pull perfusion study. J. Neurosci., 12: 1024–1032. Russell, J.A., Leng, G. and Bicknell, R.J. (1995) Opioid tolerance and dependence in the magnocellular oxytocin system: a physiological mechanism?. Exp. Physiol., 80: 307–340. Sabatier, N., Richard, P. and Dayanithi, G. (1998) Activation of multiple intracellular transduction signals by vasopressin in

vasopressin-sensitive neurones of the rat supraoptic nucleus. J. Physiol., 513: 699–710. Sagar, S.M. and Ferriero, D.M. (1987) NADPH diaphorase activity in the posterior pituitary: relation to neuronal function. Brain Res., 400: 348–352. Segovia, G., Porras, A. and Mora, F. (1994) Effects of a nitric oxide donor on glutamate and GABA release in striatum and hippocampus of the conscious rat. NeuroReport, 5: 1937– 1940. Shuster, S.J., Riedl, M., Li, X., Vulchanova, L. and Elde, R. (1999) Stimulus-dependent translocation of k opioid receptors to the plasma membrane. J. Neurosci., 19: 2658–2664. Shuster, S.J., Riedl, M., Li, X., Vulchanova, L. and Elde, R. (2000) The kappa opioid receptor and dynorphin co-localize in vasopressin magnocellular neurosecretory neurons in guineapig hypothalamus. Neuroscience, 96: 373–383. Simmons, M.L., Terman, G.W., Gibbs, S.M. and Chavkin, C. (1995) L-type calcium channels mediate dynorphin neuropeptide release from dendrites but not axons of hippocampal granule cells. Neuron, 14: 1265–1272. Srisawat, R., Ludwig, M., Bull, P.M., Douglas, A.J., Russell, J.A. and Leng, G. (2000) Nitric oxide and the oxytocin system in pregnancy. J. Neurosci., 20: 6721–6727. Stern, J.E. and Ludwig, M. (2001) NO inhibits supraoptic oxytocin and vasopressin neurons via activation of GABAergic synaptic inputs. Am. J. Physiol., 280: R1815–R1822. Sudhof, T.C. (1995) The synaptic vesicle cycle: a cascade of protein–protein interactions. Nature, 375: 645–653. Theodosis, D.T., Montagnese, C., Rodriguez, F., Vincent, J.D. and Poulain, D.A. (1986a) Oxytocin induces morphological plasticity in the adult hypothalamo-neurohypophysial system. Nature, 322: 738–740. Theodosis, D.T., Paut, L. and Tappaz, M.L. (1986b) Immunocytochemical analysis of the GABAergic innervation of oxytocin- and vasopressin-secreting neurons in the rat supraoptic nucleus. Neuroscience, 19: 207–222. Ueta, Y., Levy, A., Chowdrey, H.S. and Lightman, S.L. (1995) Water deprivation in the rat induces nitric oxide synthase (NOS) gene expression in the hypothalamic paraventricular and supraoptic nuclei. Neurosci. Res., 23: 317–319. Ueta, Y., Levy, A., Lightman, S.L., Hara, Y., Serino, R., Nomura, M., Shibuya, I., Hattori, Y. and Yamashita, H. (1998) Hypovolemia upregulates the expression of neuronal nitric oxide synthase gene in the paraventricular and supraoptic nuclei of rats. Brain Res., 790: 25–32. Villar, M.J., Ceccatelli, S., Ronnqvist, M. and Hokfelt, T. (1994) Nitric oxide synthase increases in hypothalamic magnocellular neurons after salt loading in the rat. An immunohistochemical and in situ hybridization study. Brain Res., 644: 273–281. Vincent, S.R. and Kimura, H. (1992) Histochemical mapping of nitric oxide synthase in the rat brain. Neuroscience, 46: 755– 784. Watson, S.J., Akil, H., Fischli, W., Goldstein, A., Zimmerman, E., Nilaver, G. and van Wimersma Greidanus, T.B. (1982) Dynorphin and vasopressin: common localization in magnocellular neurons. Science, 216: 85–87. Whitnall, M.H., Gainer, H., Cox, B.M. and Molineaux, C.J.

256 (1983) Dynorphin-A-(1–8) is contained within vasopressin neurosecretory vesicles in rat pituitary. Science, 222: 1137– 1139. Wotjak, C.T., Ludwig, M. and Landgraf, R. (1994) Vasopressin facilitates its own release within the rat supraoptic nucleus in vivo. NeuroReport, 5: 1181–1184. Wotjak, C.T., Ganster, J., Kohl, G., Holsboer, F., Landgraf, R. and Engelmann, M. (1998) Dissociated central and peripheral release of vasopressin, but not oxytocin, in response to repeated swim stress: new insights into the secretory capacities of peptidergic neurons. Neuroscience, 85: 1209–1222.

Wuarin, J.-P. and Dudek, F.E. (1993) Patch-clamp analysis of spontaneous synaptic currents in supraoptic neuroendocrine cells of the rat hypothalamus. J. Neurosci., 13: 2323–2331. Zhang, K. and Patel, K.P. (1998) Effect of nitric oxide within the paraventricular nucleus on renal sympathetic nerve discharge: role of GABA. Am. J. Physiol., 275: R728–R734. Zilberter, Y., Kaiser, K.M.M. and Sakmann, B. (1999) Dendritic GABA release depresses excitatory transmission between layer 2/3 pyramidal and bitufted neurons in rat neocortex. Neuron, 24: 979–988.

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 20

Cardiovascular regulation of supraoptic vasopressin neurons J. Thomas Cunningham ∗ , Stacy B. Bruno, Regina R. Grindstaff, Ryan J. Grindstaff, Karen H.R. Higgs, Danilo Mazzella and Margaret J. Sullivan Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri – Columbia, Research Park, Columbia, MO 65211, USA

Abstract: A number of laboratories have identified several key areas in the central nervous system that relay information from arterial baroreceptors to the supraoptic nucleus. Two of these regions are the diagonal band of Broca and the perinuclear zone of the supraoptic nucleus. Recent findings suggest that the inhibition of vasopressin neurons in the SON by caval–atrial stretch may also involve the perinuclear zone. Using Fos immunocytochemistry in combination with volume expansion in unanesthetized rats, we observed that volume expansion activates a number of regions in the CNS including the area postrema, the nucleus of the solitary tract, the caudal ventrolateral medulla, the paraventricular nucleus, the perinuclear zone and oxytocin neurons in the supraoptic nucleus. Further experiments using pericardial catheters demonstrate that the activation of the nucleus of the solitary tract, the ventrolateral medulla, the paraventricular nucleus and the perinuclear zone by volume expansion is dependent on cardiac afferents. However, the Fos in the area postrema and oxytocin neurons of the supraoptic nucleus is not affected by removal of cardiac afferents. Similarly, electrophysiological experiments show that stimulation of cardiac receptors in the caval–atrial junction inhibits supraoptic vasopressin neurons but does not significantly affect the activity of supraoptic oxytocin neurons. These experiments suggest that while the inhibition of supraoptic vasopressin neurons during volume expansion is mediated by cardiac afferents, the activation of supraoptic oxytocin is independent of cardiac afferents and may be mediated by other visceral afferents or humoral factors. Additional electrophysiological experiments examined the importance of the perinuclear zone in cardiopulmonary regulation of vasopressin. Excitotoxin lesions of the perinuclear zone region block the inhibitory effects of caval–atrial stretch on supraoptic vasopressin neurons. This lesion has previously been shown to block the inhibitory effects of arterial baroreceptor stimulation on supraoptic vasopressin neurons. Thus, the neural pathways that inhibit vasopressin release in response to an increase in blood pressure and an increase in blood volume may overlap at the perinuclear zone of the supraoptic nucleus. Also while the inhibition of supraoptic vasopressin neurons during volume expansion is mediated by cardiac afferents, the activation of supraoptic oxytocin neurons is independent of cardiac afferents and may be mediated by other visceral afferents or hormonal factors. Keywords: Baroreceptors; Cardiopulmonary receptors; Diagonal band of Broca; Oxytocin; Volume expansion; Congestive heart failure

∗ Correspondence to: J.T. Cunningham, Dalton Cardiovascular Research Center, Research Park, Columbia, MO 65211, USA. Tel.: +1-573-884-7229; Fax: +1-573-884-4232; E-mail: [email protected]

258

Introduction The supraoptic nuclei (SON) in the rat are located on the ventral surface of the hypothalamus lateral and dorsal to the optic chiasm and optic tracts. The magnocellular neurons located in the SON project to the posterior pituitary along with magnocellular neurons from the paraventricular (PVN) nuclei of the hypothalamus and accessory regions in the hypothalamus (Armstrong, 1985). These neurons release peptide hormones into the systemic circulation. The two primary secretory products of these cells are vasopressin and oxytocin. Vasopressin, one of the major hormonal factors involved in body fluid homeostasis, is a potent vasoconstrictor, increases water reabsorption by the collecting duct of the kidneys and alters arterial baroreceptor and cardiopulmonary reflex control of the sympathetic nervous system (Cunningham and Sawchenko, 1991). Oxytocin stimulates lactation and parturition (Cunningham and Sawchenko, 1991) and has been suggested to have natriuretic effects. Previous studies have demonstrated that vasopressin and oxytocin are synthesized (Xi et al., 1999) and released by separate populations of neurosecretory cells (Cunningham and Sawchenko, 1991; Renaud and Bourque, 1991). Vasopressin neurons can be distinguished from oxytocin neurosecretory neurons based on their spontaneous activity and their responsiveness to various peripheral stimuli (Bourque and Renaud, 1990; Renaud and Bourque, 1991). For example, vasopressin neurons in the SON have either phasic or continuous patterns of spontaneous activity (Bourque and Renaud, 1990; Renaud and Bourque, 1991) and selectively inhibited by baroreceptor afferent stimulation, and are activated by chemoreceptor afferent input (Harris, 1979). Oxytocin neurons, on the other hand, are selectively activated by gastric distension and peripherally injected CCK (Renaud et al., 1987; Renaud and Bourque, 1991). Neuroanatomical experiments have detailed the afferent input to the SON and the magnocellular PVN. Regions projecting to the hypothalamic magnocellular neurosecretory cells include the caudal ventrolateral medulla (CVL), the nucleus of the solitary tract (NTS), the raphe nuclei, the A14 and A15 dopamine cell groups, the tuberomammillary nuclei, the arcuate nucleus, the subfornical organ, the median preoptic nucleus, the organum vascu-

losum of the lamina terminalis, and the olfactory bulb (Miselis, 1981; Silverman et al., 1981; Tribollet and Dreifuss, 1981; Sawchenko and Swanson, 1982; Tribollet et al., 1985; Wiess et al., 1989; Wilkin et al., 1989). The supraoptic nucleus also receives direct retinal hypothalamic projections (Levine et al., 1994). In addition, the SON is also innervated by the perinuclear zone (PNZ) of the SON, which lies in the lateral hypothalamus dorsal to the SON (Tribollet et al., 1985; Jhamandas et al., 1989a,b). The putative neurotransmitters used by many of these neuronal projections have been described. For example, the hypothalamic neurosecretory neurons receive prominent noradrenergic (Sawchenko and Swanson, 1982), dopaminergic (Van Vulpen et al., 1999), serotinergic (Sawchenko et al., 1983), histaminergic (Panula et al., 1984), peptidergic (Jhamandas et al., 1989a,b; Sawchenko et al., 1990), and GABAergic projections (Tappaz et al., 1983; Van Den Pol, 1988; Theodosis et al., 1986). Classic studies done in the 1970s demonstrated that the electrical activity of magnocellular neurons in the SON and the PVN is strongly related to the plasma levels of vasopressin and oxytocin (Dyball, 1971; Dyball and Dyer, 1971; Dutton and Dyball, 1979; Poulain and Wakerly, 1982; Renaud and Bourque, 1991; Armstrong, 1995; Leng et al., 1999). Both the frequency of activity and the patterning of action potentials have been related to hormone release (Poulain and Theodosis, 1988). The patterning of actions potentials in magnocellular neurons is determined by the currents expressed by the cells and these properties can be influenced by the physiological state of the animal (Armstrong and Stern, 1998). Thus, the synaptic input to SON neurons interacts with the membrane properties of the cells to determine the circulation levels of vasopressin and oxytocin. There is also evidence that many of these pathways may interact with each other in the SON through presynaptic mechanisms and the dendritic release of vasopressin and oxytocin (Ludwig, 1998; Pittman et al., 1998, 2000). Traditionally, vasopressin release has been described as being controlled by osmotic and nonosmotic signals. Studies on the osmotic regulation of vasopressin release have focused on the intrinsic osmosensitivity of magnocellular neurons themselves (Bourque and Oliet, 1997; Bourque, 1998) and osmotically sensitive circumventricular organs in the

259

forebrain (Bourque, 1998) and the hindbrain (Curtis et al., 1999). Peripheral sodium receptors may also contribute to the osmotic regulation of vasopressin secretion (Huang et al., 2000a,b). Non-osmotic signals that regulate vasopressin release include all other factors that alter vasopressin release such as changes in blood pressure and blood volume and stress (Cunningham and Sawchenko, 1991; Share, 1996; Leng et al., 1999). The neural projections to the magnocellular neurons in the hypothalamus provide the framework for these different signals to reach the neurosecretory cells, influence their activity and consequently alter circulating levels of vasopressin and oxytocin. While earlier reviews have examined general aspects of the neurophysiology and pharmacology of the magnocellular system (Armstrong, 1995; Hatton and Li, 1998; Leng et al., 1999), this review will focus on arterial baroreceptor and cardiac receptor pathways that inhibit vasopressin SON neurons. Arterial baroreceptor pathways to vasopressin neurons in the SON Arterial baroreceptor neurons are primary sensory cells that convert the distention of major arteries into electrochemical signals that are relayed to the central nervous system. These cells have their sensory nerve endings embedded in the adventitia of either the aortic arch or the carotid sinuses while their primary central target is the nucleus of the solitary tract (NTS). Baroreceptor neurons, their role in reflex control of the autonomic nervous system and their interaction with NTS neurons have been the topic of many recent review articles (Andresen and Kunze, 1994; Andresen and Mendelowitz, 1996; Drummond et al., 2001). Another function of arterial baroreceptors is their role in regulating vasopressin release and the baroreceptor-mediated inhibition of SON neurons (Share, 1988, 1996; Kumada et al., 1990; Cowely, 1992; Dampney, 1994). The influence of peripheral baroreceptors on vasopressin release has been described in whole animal experiments that examined the effects of peripheral nerve section on circulating vasopressin levels. Share and Levy (1962) demonstrated that circulating levels of vasopressin in the dog are elevated by vagotomy and carotid occlusion. Similar experiments also in

the dog have demonstrated that plasma vasopressin can be elevated by the removal of carotid, aortic, and cardiopulmonary baroreceptors (Thames and Schmid, 1979; Share, 1988). In the rat, baroreceptor denervation produces an acute elevation in circulating levels of vasopressin (Alexander and Morris, 1986; Morris and Alexander, 1989) and oxytocin (Morris and Alexander, 1989). The results of these studies indicate that baroreceptors tonically inhibit vasopressin release (Share, 1988). Electrophysiological studies also have demonstrated that changes in blood pressure, atrial pressure and the stimulation of peripheral baroreceptor afferents directly influence the spontaneous activity of magnocellular neurons. Stimulation of the carotid sinus nerve in the cat produces excitatory and inhibitory effects on supraoptic neurons that are intensity-dependent (Yamashita, 1977). Yamashita suggested that the excitatory effects of carotid sinus nerve stimulation were due to chemoreceptor stimulation while the inhibitory effects were produced by the activation of baroreceptor nerves. Unfortunately, this study could not determine if the effect of carotid nerve stimulation was selective for vasopressin neurons or common to both types of magnocellular neurons. In the rat, Harris (1979) observed that an increase in arterial pressure or stretch of the isolated carotid sinus interrupts the spontaneous activity of phasic SON neurons. He also showed that chemoreceptor stimulation with CO2 -saturated saline increases the activity of phasic neurons. Based on these results, Harris suggested that the activation of arterial baroreceptors inhibits the activity of vasopressin neurons in the SON. Since the publication of Yamashita and Harris’ work it has been established that phasic activity is a property of vasopressin neurons in the rat (Cobbett et al., 1986; Yamashita et al., 1983; Renaud and Bourque, 1991). Subsequently, Day and Sibbald (1993a) have shown that aortic depressor nerve stimulation inhibits the activity of vasopressin neurons in the rat. These data suggest that baroreceptor activation inhibits the activity of vasopressin neurons in the supraoptic nucleus. An early study by McAllen and Harris (1988) indicated that this pathway is probably polysynaptic. Subsequent studies by Renaud and colleagues (Renaud et al., 1993; Renaud, 1996), Day and Sibbald (1990, 1993a,b) and now in this laboratory by

260

Grindstaff et al. (2000a,b), have concentrated on the central pathways that bring arterial baroreceptor information to the SON. The results of these studies have described a complex polysynaptic pathway as predicted by McAllen and Harris (1988). A number of studies have provided evidence that the neurotransmitter GABA mediates the baroreceptor inhibition of SON neurons. Pretreating either the SON (Jhamandas and Renaud, 1986a, 1987) or the PVN (Kasai et al., 1987) with the GABA-A receptor antagonist bicuculline prevents the inhibition of vasopressin neuron associated with the stimulation of peripheral baroreceptors. Increased blood pressure produced by phenylephrine increases extracellular GABA in the SON region (Voisin et al., 1994). Results from in vitro studies demonstrate that magnocellular neurons are postsynaptically inhibited by GABA acting via GABA-A receptors (Renaud and Bourque, 1991). The SON receives a dense GABA input (Tappaz et al., 1983; Theodosis et al., 1986) and much of this appears to be of local origin (Meyer et al., 1980). The diagonal band of Broca (DBB), which is located in the forebrain anterior to the lamina terminalis region, has been identified as a critical component of the pathway that brings baroreceptor information to the SON (Jhamandas and Renaud, 1986a,b). The DBB contains a sizable population of GABAergic neurons and electrical stimulation of the DBB selectively inhibits the vasopressin neurons of the SON (Jhamandas and Renaud, 1986b). Electrical stimulation of the DBB also increases the extracellular concentrations of GABA in the SON (Voisin et al., 1994). Ibotenic acid lesions of the DBB prevent the baroreceptor-mediated inhibition of SON vasopressin neurons (Cunningham et al., 1992a). However, anatomical studies indicate that the DBB does not project directly to the SON but rather to the PNZ (Tribollet et al., 1985; Jhamandas et al., 1989a,b). Therefore, it was suggested that the DBB exerted its inhibitory influence on the SON not directly but through these local GABAergic interneurons (Jhamandas et al., 1989a,b). The PNZ region has previously been shown to contain GABA neurons (Tappaz et al., 1983; Theodosis et al., 1986) although this finding is not universally supported (Roland and Sawchenko, 1993). Functional evidence for the existence of the PNZ

interneurons has been provided by a series of experiments that have taken advantage of the resistance of SON neurons to excitotoxins (Hastings et al., 1985; Herman and Wiegand, 1986). In vitro studies have demonstrated that the electrophysiological properties of SON neurons are not affected by ibotenic acid injections into the SON-PNZ region that destroy the surrounding neuropile including the GABAergic neurons in the PNZ (Hu et al., 1992). Lateral hypothalamic lesions involving the DBB terminal field in the PNZ significantly decreased the number of vasopressin SON neurons that are inhibited by increases in arterial pressure (Nissen et al., 1993). Moreover, the DBB-evoked inhibition of the same vasopressin SON neurons was also abolished in rats with PNZ lesions (Nissen et al., 1993). However, these lesions did not block the effects of either electrical stimulation of the median preoptic nucleus or peripherally injected angiotensin II on vasopressin SON neurons indicating that fibers of passage remained intact. These results demonstrate that both the baroreceptor-mediated inhibition and the DBBevoked inhibition of vasopressin SON neurons are dependent on the integrity of PNZ neurons located in the lateral hypothalamus. This supports the hypothesis that PNZ neurons are a necessary component of the neural system mediating the baroreceptorinduced inhibition of vasopressin SON neurons. The source of the baroreceptor information to the DBB has also been the focus of recent investigation. Several studies suggest that the noradrenergic innervation of the DBB is involved in the baroreceptormediated inhibition of vasopressin releasing neurons in the SON. First, catecholamine depletion of the DBB blocks the effects of baroreceptor stimulation on vasopressin neurons in the SON (Cunningham et al., 1992b). Also, norepinephrine injected into the DBB inhibits the activity of vasopressin neurons in the SON without significantly influencing the activity of oxytocin neurons (Cunningham et al., 1993). Finally, Bealer (1997) has shown that increases in blood pressure produced with a peripherally acting vasoconstrictor are associated with an increase in the norepinephrine content of the DBB. Neuroanatomical studies of the DBB suggest that it receives afferents from the locus coeruleus (LC), the parabrachial nucleus and the A1 region of the ventrolateral medulla (Jones and Moore, 1977; Lind-

261

vall and Stenevi, 1978; Saper and Loewy, 1980; Vertes, 1988; Zaborszky and Cullinan, 1996; Senatorov and Renaud, 1999). The existence of a direct pathway from the NTS to the DBB is less established (Ricardo and Koh, 1978; Vertes, 1988; Senatorov and Renaud, 1999). Several studies suggest that the parabrachial nucleus (Ohman et al., 1990; Jhamandas et al., 1991a,b) and the LC region (Kannan et al., 1981; Sved, 1986) are involved in regulating the activity of SON neurons. The A1 region of the ventrolateral medulla has been shown to mediate vasopressin release produced by hemorrhage (Blessing and Willoughby, 1985; Head et al., 1989). Electrophysiological studies of LC neurons suggest that a majority of its cells are activated by decreases in blood pressure and inhibited by increases in blood pressure (Elam et al., 1985; Olpe et al., 1985), although one study did report that a subpopulation of LC neurons was activated by increased blood pressure (Olpe et al., 1985). However, a recent study showed that electrical stimulation of the aortic depressor nerve produces excitatory and inhibitory effects on the LC neurons (Murase et al., 1994), and another study demonstrated Fos activation in the LC after aortic depressor nerve stimulation (McKitrick et al., 1992). Banks and Harris (1984) showed that lesions of the LC would disrupt the baroreceptormediated inhibition of SON vasopressin neurons. Yet, studies by Day suggest that the A1 region of the ventrolateral medulla (Day and Sibbald, 1990), the parabrachial nucleus (Day and Sibbald, 1993a), the LC (Day and Sibbald, 1993b) are not involved in transmitting baroreceptor information to the SON. Recently, we have addressed this issue by combining retrograde tract tracing with Fos immunocytochemistry (Grindstaff et al., 2000b). Fos is the protein product of the early active gene c-fos and its expression has been widely used as an indicator of neuronal activation (Morgan and Curran, 1991; Curran and Morgan, 1995). This approach has several limitations. First, it has been shown that not all cells in the CNS express Fos after they are stimulated (Dampney et al., 1995). Second, Fos will only be expressed in neurons that are activated and not by cells that are inhibited (Dampney et al., 1995). Finally, immunocytochemical detection of Fos requires a one to two hour time period for expression of the Fos protein to reach detectable levels (Morgan and Curran,

1991; Dampney et al., 1995). On the positive side, the Fos technique allows us to perform experiments on unanesthetized rats free from the confounding effects of anesthetics. It can be combined with retrograde tract tracing and immunocytochemical staining for other proteins that are expressed in the CNS. Previous studies have successfully used Fos immunocytochemistry to determine which regions of the CNS are activated by baroreceptor stimulation produced by infusing peripheral vasoconstrictors (Badoer et al., 1994; Li and Dampney, 1994; Graham et al., 1995; Potts et al., 1997). Initially, we looked at Fos staining in the major noradrenergic cell groups that are reported to project to the DBB after baroreceptor stimulation produced by phenylephrine infusions (Fig. 1; Grindstaff et al., 2000b). Our results indicated that of these regions only the LC contained a significant population of the DBH-positive cells that stained for Fos after phenylephrine infusion. Sino-aortic denervation blocked this increase in Fos demonstrating that it was mediated by arterial baroreceptors (Grindstaff et al., 2000b). In addition, we injected retrograde tract tracers into the DBB and found that LC neurons that were activated by baroreceptor stimulation did project to the DBB (Grindstaff et al., 2000b). Finally, ibotenic acid lesions of the LC blocked the baroreceptor-mediated inhibition in a significant number of phasic vasopressin neurons in the SON (Grindstaff et al., 2000). These data indicate that the LC may provide the DBB with baroreceptor information. How baroreceptor reaches the LC remains controversial. Although there is evidence that the LC receives projections from the NTS (Cedarbaum and Aghajanian, 1978; Ter Horst et al., 1989), another study found no find evidence for such a projection (Aston- Jones et al., 1986). It is possible that information from peripheral baroreceptors may reach the LC via alternative projections from the ventrolateral medulla (Aston-Jones et al., 1986; Pieribone and Aston-Jones, 1991; Valentino et al., 1992). We have also conducted experiments to further characterize the role of the DBB in the regulation of vasopressin release. Electrophysiological experiments have described several different cell types in DBB (Griffith and Matthews, 1986; Griffith, 1988; Markram and Segal, 1990; Matthews and Lee, 1991). One cell type that has broad action potentials and

262

Fig. 1. Changes in the mean number of Fos and Fos + DBH cells in sham SAD rats (top) and SAD rats (bottom) after either saline or phenylephrine infusions which significantly increased blood pressure. In sham SAD rats, phenylephrine significantly increases Fos above control levels in each region presented (* P < 0.05 from saline control). Only in the locus coeruleus did the baroreceptor stimulation produced by phenylephrine increase the number of Fos + DBH neurons. In SAD rats the phenylephrine infusions did not significantly increase Fos staining in any of the regions examined. This suggests that activation of arterial baroreceptors activated noradrenergic neurons in the locus coeruleus. (From Grindstaff et al., 2000a.)

large after-hyperpolarizations has been characterized as cholinergic (Griffith and Matthews, 1986; Griffith, 1988; Markram and Segal, 1990; Matthews and Lee, 1991). We have used retrograde tract tracing from the PNZ and immunocytochemistry for choline acetyltransferase (ChAT) and acetylcholine-specific immunotoxin lesions to evaluate the contribution of cholinergic DBB neurons to the baroreceptormediated inhibition of vasopressin SON neurons (Grindstaff et al., 2000a). The results of this study

demonstrated that cholinergic neurons do not contribute to this system. DBB neurons that were retrogradely labeled from the PNZ were not positive for ChAT immunofluorescence. Similarly, injections of a cholinergic-specific immunotoxin into the DBB, which significantly depleted the cholinergic neurons of the DBB did not significantly affect the inhibition of vasopressin SON neurons by baroreceptor stimulation. This indicates that a different population of DBB neurons project to the PNZ. The DBB

263

contains a number of other cell types including neurons that produce luteinizing hormone releasing hormone, gonadotropin-releasing hormone, calretinin, vasopressin, substance P, neurotensin, and other neuropeptides (Caffe et al., 1989; Kiss et al., 1997; Ebling et al., 1998; Gonzalo-Ruiz et al., 1999). In addition, the DBB contains glutamatergic neurons (Gonzalo-Ruiz et al., 1999). These cells may be the most likely source of DBB innervation to the PNZ. The DBB has also been studied to determine its role in autonomic regulation. Kirouac and Ciriello (1997) report that glutamate injections into the horizontal limb of the DBB of anesthetized rats produces sympatho-inhibition. Abdelmalek et al. (1999) report that injections of beta-adrenergic agonists into the DBB in conscious rats produced a depressor response. In this study the authors also observed a pressor response to alpha2-adrenoceptor stimulation of the DBB and the magnitude of the pressor response was affected by anesthesia. Injections of colchicines into the DBB also influence blood pressure (Abdelmalek et al., 1994). The colchicine lesions of the DBB result in hypotension without affecting plasma vasopressin or the vasopressin response to hemorrhage. The authors suggest that the lesion of the DBB produces a sympatho-inhibition that is not compensated for by vasopressin release. Additional experiments have used ibotenic acid lesions of the DBB to examine drinking behavior (Sullivan et al., 1991), cardiovascular parameters, and vasopressin release (Mazzella et al., 2000). Rats with ibotenate lesions of the DBB drank significantly more water following angiotensin II or polyethylene glycol injections than sham-lesioned rats (Sullivan et al., 1991). The lesions do not significantly affect their drinking responses to hypertonic saline and water deprivation, indicating that the DBB may be selectively involved in drinking behavior related to extracellular dehydration. The resting blood pressure of DBB lesioned rats is not significantly different from rats with control injections (Mazzella et al., 2000). The rats with DBB lesions did have resting heart rates that were significantly lower compared to controls and basal plasma vasopressin levels were significantly higher than controls (Fig. 2; Mazzella et al., 2000). Our preliminary data suggest that at the time point when we measured the plasma vasopressin the DBB-lesioned rats have a normal fluid

intake, food intake and water balance (Fig. 2). Exogenously administered vasopressin has been shown to act at the area postrema to suppress baroreflex function (Hasser et al., 1997). However, this effect has not been consistently observed in the rat. The decrease in resting heart rate that was observed in the DBB-lesioned rats could be due to an inhibition of the sympathetic nervous system produced by the increase in circulating vasopressin. The magnitude of the change in circulating vasopressin may not be sufficient to produce the change in heart rate that we observed in the DBB-lesioned animals. Nevertheless, this hypothesis could be easily tested by infusing DBB-lesioned rats with a vasopressin antagonist. These data and the results from other laboratories suggest that the DBB is involved in more than just vasopressin release and its role in body fluid balance and autonomic regulations should be explored in greater detail. Cardiopulmonary baroreceptor pathways to vasopressin neurons in the SON Fewer studies have addressed the role of cardiac receptors or atrial receptors in the control of neurons in the SON (Schmid et al., 1984). Activation of cardiac receptors by atrial distension inhibits the activity of magnocellular neurons in the SON of the dog (Koizumi and Yamashita, 1978) and the cat (Koizumi and Yamashita, 1978; Menninger, 1979). In the rat, stimulation of stretch receptors in the right atria inhibits drinking behavior (Kaufman, 1984), vasopressin release (Bennett et al., 1983; Kaufman, 1987) and renal sympathetic nerve activity (Hines et al., 1994; Hines and Mifflin, 1995). Similarly, isotonic increases in plasma volume, which could also stimulate cardiopulmonary baroreceptors, inhibit the activity of vasopressin neurons while oxytocin neurons show an initial excitation followed by inhibition (Pendlebury et al., 1992). These data suggest that increased plasma volume inhibits the release of both vasopressin and oxytocin. Other data suggest that volume expansion may increase vasopressin and oxytocin release. Using Fos immunocytochemistry in urethane-anesthetized rats, Naraveaz et al. (1993) suggested that volume expansion activates vasopressin neurons in the SON. In this study the anesthetized rats were injected with either

264

Fig. 2. Basal plasma AVP levels are significantly increased in rats with ibotenic acid lesions of the DBB (DBX, n = 6) compared to rats injected with vehicle (sham, n = 6). Basal heart rate is significantly decreased in rats with DBB lesions (DBX, n = 14) as compared to control (n = 14). These observations were made 5–8 days following lesion. In a separate group of rats, metabolism cage studies were used to examine the water balance of DBB-lesioned rats (DBX, n = 11) and control (CON, n = 9). Immediately after surgery (S) both groups showed a significant change from baseline that has recovered to control levels by day 10.

0.5 or 5 ml of saline i.v. Both protocols increased Fos in vasopressin neurons in the SON by approximately 25%, which they attribute to the activation of atrial receptors. These results appear to contradict the earlier studies cited above which indicate that volume expansion and the stimulation of atrial receptors inhibits vasopressin release. A series of studies by Antunes-Rodrigues et al. (Antunes-Rodrigues et al., 1993; Reis et al., 1994; Haanwinckel et al., 1995)

indicate that volume expansion is associated with a significant increase in circulating oxytocin. They suggest that volume expansion activates baroreceptors and renal afferents to stimulate oxytocin release from the hypothalamus via the LC and by cholinergic and ANP neurons in the lamina terminalis region (Haanwinckel et al., 1995). Oxytocin release stimulated by volume expansion is hypothesized to increase sodium excretion by acting directly on the

265

kidney and by facilitating ANP release from the atria. Thus, it appears that the role of atrial receptors in the regulation of vasopressin and oxytocin is perhaps not as clear as earlier work in this field suggested. In earlier electrophysiological studies that examined the influence of cardiac receptors on the activity of SON neurons, the authors did not determine whether the effects that they observed were selective to vasopressin or oxytocin neurons (Koizumi and Yamashita, 1978; Menninger, 1979). In part this was due to the fact that these studies were completed before differences in firing pattern and differential responses to physiological stimuli among vasopressin and oxytocin neurons had been characterized. Another factor is the species used. Vasopressin magnocellular neurons in the cat do not have a regenerative component of their whole-cell calcium current that is necessary for phasic activity (Fagan and Andrew, 1991). Therefore it would not have been possible for Koizumi and Yamashita (1978) and Menninger (1979) to discriminate vasopressin neurons from oxytocin neurons based on their patterns of activity. Although Pendlebury et al. (1992) did characterize their cells as vasopressin and oxytocin, the nature of the stimulus that they used and its time course were more consistent with determining the effects of humoral mechanisms on the activity of SON neurons. In order to better characterize the specific effects of cardiac receptor stimulation on magnocellular neurons in the SON, we used a technique that was first used to study drinking behavior. As initially described by Kaufman (1984), a latex balloon catheter was placed at the junction of the right atrium and the vena cava of anesthetized rats (Grindstaff et al., 2000). This catheter was used to stimulate cardiac receptors while extracellular recordings were obtained from characterized SON neurons. Short periods of stimulation were used to avoid any confounding effects of hormone release (Garcia et al., 1987). In these experiments, 63/84 vasopressin neurons were inhibited by stimulation of the caval–atrial junction and this inhibitory effect was blocked by bilateral vagotomy (Fig. 3). Oxytocin neurons, on the other hand, were not significantly affected by caval–atrial stretch. Stimulation of cardiopulmonary baroreceptors by atrial distention influences the activity of vasopressin

releasing neurons in the SON. However, there is little information to describe how cardiopulmonary baroreceptor information reaches the SON. Electrophysiological studies indicate that neurons in the NTS (Hines et al., 1994; Hines and Mifflin, 1995) and the LC (Jhamandas et al., 1988) are activated by atrial distention but it has not been demonstrated that these neurons project to the hypothalamus. In Fos experiments with isotonic volume expansion in unanesthetized rats, we have observed significant Fos activation in the PNZ, the CVL, the area postrema and the NTS (Randolph et al., 1998). In this experiment, we also observed by isotonic volume expansion selectively activated oxytocin neurons in the SON. In order to determine whether these responses were dependent on cardiac receptors, rats were instrumented with intrapericardial catheters (Cunningham et al., 2000). This technique permits substances to be injected into the pericardial space of unanesthetized animals (Bell et al., 1993). We used it to inject procaine into the pericardial space to block cardiac afferents prior to isotonic volume expansion. Pericardial pretreatment with procaine blocked the increase in Fos staining seen in the PNZ, CVL and NTS following volume expansion. A significant increase in Fos staining was observed in the area postrema and the oxytocin neurons of the SON in the volume-expanded group that received pericardial procaine. Thus, the activation of neurons in the PNZ, CVL and NTS following volume expansion depends on the integrity of cardiac afferents while the increased Fos staining in the area postrema and the SON oxytocin neurons do not require cardiac afferent input. This suggests that a humoral factor (Garcia et al., 1987; Chenault et al., 1992) or a non-cardiac afferent is primarily responsible for activating the area postrema and the SON oxytocin neurons during volume expansion. Functionally, the PNZ plays a major role in the arterial baroreceptor-mediated inhibition of vasopressin neurons in the SON. Our Fos data suggest that the PNZ could play a similar role in the inhibitory effects of atrial distention on vasopressin releasing neurons in the SON. Previously, excitotoxic lesions of the PNZ were used to test its involvement in the baroreceptor-mediated inhibition of vasopressin releasing neurons in the SON (Nissen et al., 1993). This technique has also been employed

266

Fig. 3. Continuous rate meter recording from a phasic vasopressin neuron in the SON of a normal rat instrumented with a caval–atrial balloon catheter showing the effects of bilateral cervical vagotomy of baroreceptor and cardiac receptor inhibition. Inflation of the balloon (Balloon) and i.v. injection of phenylephrine (Phe) both inhibit the cell prior to vagotomy. After sectioning of the left (L) and right (R) vagi the inhibitory effects of the balloon are blocked.

to evaluate the contribution of the PNZ to the inhibition of vasopressin neurons by caval–atrial stretch (Grindstaff and Cunningham, 2001). The results of this study showed that PNZ lesions that blocked the baroreceptor-mediated inhibition of vasopressin SON neurons also blocked the inhibitory effects of caval–atrial stretch (Fig. 4). This indicates that these two pathways overlap in the PNZ, but there is insufficient information to speculate whether or not the same population of PNZ neurons mediates baroreceptor and cardiac receptor inhibition of vasopressin neurons. In the Fos experiments with volume expansion, we did not observe an increase in Fos in the DBB. This finding suggests that while the PNZ is activated by volume expansion the DBB is not. Although the failure to observe Fos expression in a region does not

necessarily mean that the cells were not activated, we hypothesize that the differences in Fos expression observed in the DBB and PNZ do represent differences in neural activation associated with volume expansion. Thus, the pathways for baroreceptor-mediated inhibition of vasopressin neurons may overlap only at the level of the PNZ in the forebrain. This topic will be the focus of future investigations. Summary The maintenance of body fluid and electrolyte balance requires the coordination of physiological, behavioral and hormonal systems, which include vasopressin and oxytocin release. This organization necessarily involves a complex neural network to sense changes in hydromineral balance and initiate the ap-

267

Fig. 4. (A) Rate meter recordings of phasic vasopressin neurons from rats injected into the PNZ with either vehicle (Con) or ibotenic acid (PNZX). Note that in the rate meter record from the control animal acute increases in blood pressure produced by i.v. phenylephrine (filled boxes) and caval–atrial stretch (open bars) inhibit the activity of the cell. In a rat with a lesion of the PNZ neither stimulus interrupts the activity of the cell. (B) An example of a PNZ lesion (gray shaded area) that blocked both arterial baroreceptor and cardiac receptors inhibition of phasically active vasopressin neurons.

268

propriate short-term and long-term adjustments to maintain body fluid homeostasis. While this area has been the subject of considerable study (Abboud and Thames, 1982; Share, 1988; Kumada et al., 1990; Cowely, 1992; Dampney, 1994), much of the central network involved in the maintenance of body fluid balance remains to be defined. Recent studies by Schrier indicate that vasopressin may play a significant role in congestive heart failure (Schrier et al., 1998). Specifically, non-osmotic regulation of vasopressin secretion may contribute to high circulating levels of vasopressin in end-stage heart failure (Martin and Schrier, 1997; Schrier et al., 1998). Norsk (1996) has proposed that changes in arterial baroreceptor and atrial receptor control of vasopressin release may contribute to changes in extracellular fluid volume observed during space flight and prolonged bed rest. Alterations in control of vasopressin release have also been observed during pregnancy (Lindheimer et al., 1995). Understanding the neural pathways involved in the cardiovascular regulation of vasopressin release is an important first step that is needed to address the role of vasopressin in each of these altered physiological states. Research in the neurohypophysial system is unique in that it provides the opportunity to study fundamental neuroscience within an important integrative physiological context. Abbreviations ChAT CO2 CNS CVL DBB GABA LC NTS PNZ PVN SON

choline acetyl transferase carbon dioxide central nervous system caudal ventrolateral medulla diagonal band of Broca gamma-amino butyric acid locus coeruleus nucleus of the solitary tract perinuclear zone of the supraoptic nucleus paraventricular nucleus of the hypothalamus supraoptic nucleus

Acknowledgements This work was supported by National Heart, Lung and Blood Institute Grants R29-HL55692 (JTC),

R01-HL62579 (JTC) and K02-HL03620 (JTC), and NIDDK grant R01-DK57822 (MJS), a Scientist Development award from the American Heart Association (MJS) and a Grant-in-Aid from the Missouri Affiliate of the American Heart Association (JTC). The authors thank Drs. Qian Li and Kathleen S. Curtis for their participation in the work reviewed here. The authors also thank the Organizing Committee for the 2001 WCNH Conference. MJS passed away on June 2, 2001. References Abboud, F.M. and Thames, M.D. (1982) Interaction of cardiovascular reflexes in circulatory control. In: J.T. Shepherd and F.M. Abboud (Eds.), Handbook of Physiology — Cardiovascular System III. American Physiological Society, Bethesda, MD, pp. 675–753. Abdelmalek, A., Forsling, M.L. and Thornton, S.N. (1994) Colchicine injected into the anterior forebrain in rats decreases blood pressure without changing the responses to hemorrhage. Brain Res. Bull., 34: 313–317. Abdelmalek, A., Ayad, G. and Thornton, S.N. (1999) Cardiovascular effects of catecholamines injected into the DBB of rats, influence of urethane anesthesia and local colchicines. Brain Res., 821: 50–59. Alexander, N. and Morris, M. (1986) Increased plasma vasopressin in sinoaortic denervated rats. Neuroendocrinology, 42: 361–367. Andresen, M.C. and Kunze, D.L. (1994) Nucleus tractus solitarius — gateway to neural circulatory control. Annu. Rev. Physiol., 56: 93–116. Andresen, M.C. and Mendelowitz, D. (1996) Sensory afferent neurotransmission in caudal nucleus tractus solitarius — common denominators. Chem. Sens., 21: 387–395. Antunes-Rodrigues, J., Marubayashi, U., Favaretto, A.L.V., Gutkowska, J. and McCann, S.M. (1993) Essential role of the hypothalamic muscarinic and alpha-adrenergic receptors in atrial natriuretic peptide release induced by blood volume expansion. Proc. Natl. Acad. Sci. USA, 90: 10240–10244. Armstrong, W.E. (1985) Hypothalamic supraoptic and paraventricular nuclei. In: G. Paxinos (Ed.), The Rat Nervous System, Vol. 1. Forebrain and Midbrain. Academic Press Australia, Sydney. Armstrong, W.E. (1995) Morphological and electrophysiological classification of hypothalamic supraoptic neurons. Prog. Neurobiol., 47: 291–339. Armstrong, W.E. and Stern, J.E. (1998) Phenotypic and statedependent expression of the electrical and morphological properties of oxytocin and vasopressin neurons. Prog. Brain Res., 119: 101–113. Aston-Jones, G., Ennis, M., Pieribone, V.A., Nickell, W.T. and Shipley, M.T. (1986) The brain nucleus coeruleus: restricted afferent control of a broad efferent network. Science, 234: 734–737.

269 Badoer, E., McKinley, M.J., Oldfield, B.J. and McAllen, R.M. (1994) Localization of barosensitive neurons in the caudal ventrolateral medulla which project to the rostral ventrolateral medulla. Brain Res., 657: 258–268. Banks, D. and Harris, M.C. (1984) Lesions of the locus coeruleus abolish baroreceptor-induced depression of supraoptic neurones in the rat. J. Physiol., 355: 383–398. Bealer, S.L. (1997) Acute hypertension increases norepinephrine release in the diagonal band of Broca. Brain Res., 745: 313– 316. Bell, L.B., Quandt, L.M., O’Hagan, K.P., Mittelstadt, S.W. and Clifford, P.S. (1993) Respiratory alterations with intrapericardial procaine in the conscious rabbit. Clin. Exp. Pharmacol. Physiol., 20: 753–762. Bennett, K.L., Linden, R.J. and Mary, D.A. (1983) The effect of stimulation of atrial receptors on the plasma concentration of vasopressin. Q. J. Exp. Physiol., 68: 579–589. Blessing, W.W. and Willoughby, J.O. (1985) Inhibiting the rabbit caudal ventrolateral medulla prevents baroreceptor-initiated secretion of vasopressin. J. Physiol., 367: 253–266. Bourque, C.W. (1989) Ionic basis for the intrinsic activation of rat supraoptic neurons by hyperosmotic stimuli. J. Physiol. London, 417: 263–277. Bourque, C.W. and Renaud, L.P. (1990) Electrophysiology of mammalian magnocellular vasopressin and oxytocin neurosecretory neurons. Frontiers Neuroendocrinol., 11: 183–212. Bourque, C.W. and Oliet, S.H.R. (1997) Osmoreceptors in the central nervous system, Review. Ann. Rev. Physiol., 59: 601– 619. Caffe, A.R., Vart 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. Comp. Neurol., 287: 302–325. 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. Chenault, V.M., Morris, M., Lynch, C.D., Maultsby, S.J. and Hutchins, P.M. (1992) Hormonal and electrolyte responses to acute ischemic volume expansion in unanesthetized rats. Proc. Exp. Biol. Med., 202: 331–337. Cobbett, P., Smithson, K.G. and Hatton, G.I. (1986) Immunoreactivity to vasopressin- but not oxytocin-associated neurophysin antiserum in phasic neurons of rat hypothalamic paraventricular nucleus. Brain Res., 362: 7–16. Cowely Jr., A.W. (1992) Long-term control of arterial blood pressure. Physiol. Rev., 72: 231–300. Cunningham Jr., E.T. and Sawchenko, P.E. (1991) Reflex control of magnocellular vasopressin and oxytocin secretion. TINS, 14: 406–411. Cunningham, J.T., Nissen, R. and Renaud, L.P. (1992a) Ibotenate lesions of the diagonal band of Broca attenuate baroreceptor sensitivity of rat supraoptic vasopressin neurons. J. Neuroendocrinol., 4: 303–309. Cunningham, J.T., Nissen, R. and Renaud, L.P. (1992b) Catecholamine depletion of the diagonal band reduces baroreflex inhibition of supraoptic neurons. Am. J. Physiol., 263: R363– 367.

Cunningham, J.T., Nissen, R. and Renaud, L.P. (1993) Noradrenaline injections in the diagonal band selectively reduce supraoptic vasopressin neural activity. Brain Res., 610: 152– 155. Cunningham, J.T., Bruno, S.B., Higgs, K.A.N. and Sullivan, M.J. (2000) Intrapericardial procaine effects volume expansioninduced Fos expression in conscious rats. Exper. Neurol., 174: 181–192. Curran, T. and Morgan, J.I. (1995) Fos: An immediate-early transcription factor in neurons. J. Neurobiol., 26: 403–412. Curtis, K.S., Huang, W., Sved, A.F., Verbalis, J.G. and Stricker, E.M. (1999) Impaired osmoregulatory responses in rats with area postrema lesions. Am. J. Physiol., 277: R209–R219. Dampney, R.A. (1994) Functional organization of central pathways regulating the cardiovascular system. Physiol. Rev., 745: 323–364. Dampney, R.A., Li, Y.W., Hirooka, Y., Potts, P. and Polson, J.W. (1995) Use of c-fos functional mapping to identify the central baroreceptor reflex pathway: advantages and limitations. Clin. Exp. Hypertens. (N.Y.), 17: 197–208. Day, T.A. and Sibbald, J.R. (1990) Involvement of the A1 cell group in baroreceptor inhibition of neurosecretory vasopressin cells. Neurosci. Lett., 113: 156–161. Day, T.A. and Sibbald, J.R. (1993a) Locus coeruleus effects on baroreceptor responsiveness and activity of neurosecretory vasopressin cells. JANS, 42: 259–264. Day, T.A. and Sibbald, J.R. (1993b) Differing effects of electrical and chemical parabrachial nucleus stimulation on supraoptic vasopressin cells. JANS, 45: 175–179. Drummond, H.A., Welsh, M.J. and Abboud, F.M. (2001) EnaC subunits are molecular component of the arterial baroreceptor complex. Ann. N.Y. Acad. Sci., 940: 42–47. Dutton, A. and Dyball, R.E.J. (1979) Phasic firing enhances vasopressin release from the rat neurohypophysis. J. Physiol., 290: 433–440. Dyball, R.E.J. (1971) Oxytocin and ADH secretion in relation to electrical activity in antidromically identified supraoptic and paraventricular units. J. Physiol., 214: 245–256. Dyball, R.E.J. and Dyer, R.G. (1971) Plasma oxytocin concentration and paraventricular neurone activity in rats with diencephalic islands and intact brains. J. Physiol., 216: 227– 235. Ebling, F.J.P., Cronin, A.S. and Hastings, M.H. (1998) Resistance of gonadotropin-releasing hormone neurons to glutamatergic neurotoxicity. Brain Res. Bull., 47: 575–584. Elam, M., Svensson, T.H. and Thoren, P. (1985) Differentiated cardiovascular afferent regulation of locus coeruleus neurons and sympathetic nerves. Brain Res., 358: 77–84. Fagan, M. and Andrew, R.D. (1991) Intracellular study of calcium-related events in cat magnocellular neuroendocrine cells. J. Physiol. (London), 434: 337–349. Garcia, R., Cantin, M. and Thibault, G. (1987) Role of right and left atria in natriuresis and atrial natriuretic factor release during blood volume changes in the conscious rat. Circ. Res., 61: 99–106. Gonzalo-Ruiz, A., Morte, L., Flecha, J.M. and Sanz, J.M. (1999) Neurotransmitter characteristics of neurons projecting to the

270 supramammillary nucleus of the rat. Anat. Embryol., 200: 377–392. Graham, J.C., Hoffman, G.E. and Sved, A.F. (1995) C-Fos expression in brain in response to hypotension and hypertension in conscious rats. J. Auton. Nerv. Syst., 55: 92–104. Griffith, W.H. (1988) Membrane properties of cell types within guinea pig basal forebrain nuclei in vitro. J. Neurophysiol., 59: 1590–1612. Griffith, W.H. and Matthews, R.T. (1986) Electrophysiology of AChE-positive neurons in basal forebrain slices. Neurosci. Lett., 71: 169–174. Grindstaff, R.R. and Cunningham, J.T. (2001) Lesion of the perinuclear zone attenuates cardiac sensitivity of vasopressinergic supraoptic neurons. Am. J. Physiol., 280: R630–R638. Grindstaff, R.J., Grindstaff, R.R. Cunningham, J.T. (2000a) Baroreceptor sensitivity of rat supraoptic vasopressin neurons involves noncholinergic neurons in the DBB. Am. J. Physiol., 279: R1934–R1943. Grindstaff, R.J., Grindstaff, R.R., Sullivan, M.J. and Cunningham, J.T. (2000b) The role of the locus coeruleus in the baroreceptor regulation of vasopressin supraoptic neurons. Am. J. Physiol., 279: R306–R319. Grindstaff, R.R., Grindstaff, R.J. and Cunningham, J.T. (2000) Effects of atrial stretch on vasopressin and oxytocin supraoptic neurons in the rat. Am. J. Physiol., 278: R1605–R1615. Haanwinckel, M.A., Elias, L.K., Favaretto, A.L.V., Gutkowska, J., McCann, S.M. and Antunes-Rodrigues, J. (1995) Oxytocin mediates atrial natriuretic peptide after volume expansion. Proc. Natl. Acad. Sci. USA, 92: 7902–7906. Harris, M.C. (1979) Effects of chemoreceptor and baroreceptor stimulation on the discharge of hypothalamic supraoptic neurones in rats. J. Endocrinol., 82: 115–125. Hasser, E.M., Bishop, V.S. and Hay, M. (1997) Interactions between vasopressin and baroreflex control of the sympathetic nervous system. Clin. Exp. Pharmacol. Physiol., 24: 102–108. Hastings, H.M., Winn, P. and Dunnett, S.B. (1985) Neurotoxic amino acid lesions of the lateral hypothalamus: a parametric comparison of the effects of ibotenate, N-methyl- D,L-aspartate and quisqualate in the rat. Brain Res., 360: 248–256. Hatton, G.I. and Li, Z.H. (1998) Neurophysiology of magnocellular neuroendocrine cells: recent advances. Prog. Brain Res., 119: 77–99. Head, G.A., Quail, A.W. and Woods, R.L. (1989) Lesions of the A1 noradrenergic cells affect AVP release and heart rate during hemorrhage. Am. J. Physiol., 253: H1012–H1017. Herman, J.P. and Wiegand, S.J. (1986) Ibotenate-induced cell death in the hypothalamic paraventricular nucleus: differential susceptibility of magnocellular and parvocellular neurons. Brain Res., 383: 367–372. Hines, T. and Mifflin, S.W. (1995) Gestational effects on volumesensitive cardiopulmonary receptor reflexes in the rat. Am. J. Physiol., 268: R735–R743. Hines, T., Toney, G.M. and Mifflin, S.W. (1994) Responses of neurons in the nucleus of the tractus solitarius to stimulation of heart and lung receptors in the rat. Circ. Res., 74: 1188– 1196. Huang, W., Sved, A.F. and Striker, E.M. (2000a) Water inges-

tion provides an early signal inhibiting osmotically stimulated vasopressin secretion in rats. Am. J Physiol., 279: R756–R760. Huang, W., Sved, A.F. and Striker, E.M. (2000b) Vasopressin and oxytocin release evoked by NaCl loads are selectively blunted by area postrema lesions. Am. J. Physiol., 278: R732–R740. Hu, B., Cunningham, J.T., Nissen, R., Renaud, L.P. and Bourque, C.W. (1992) Rat supraoptic nucleus is histologically and physiologically resistant to glutamate neurotoxicity. NeuroReport, 3: 87–90. Jhamandas, J.H. and Renaud, L.P. (1986a) A gamma-aminobutyric acid-mediated baroreceptor input to supraoptic vasopressin neurons in the rat. J. Physiol., 381: 595–606. Jhamandas, J.H. and Renaud, L.P. (1986b) Diagonal band neurons may mediate arterial baroreceptor input to hypothalamic vasopressin secreting neurons. Neurosci. Lett., 65: 214–218. Jhamandas, J.H. and Renaud, L.P. (1987) Bicuculline blocks an inhibitory baroreflex input to supraoptic vasopressin neurons. Am. J. Physiol., 256: R947–R952. Jhamandas, J.H., Kaufman, S. and Reffenstein, R.J. (1983) Locus coeruleus neurons are activated by right atrial stretch receptors. Soc. Neurosci. Abstr., 14: 193. Jhamandas, J.H., Raby, W., Rogers, J., Buijs, R.M. and Renaud, L.P. (1989a) Diagonal band projection towards the hypothalamic supraoptic nucleus: light and electron microscopic observations in the rat. J. Comp. Neurol., 282: 15–23. Jhamandas, J.H., Lind, R.W. and Renaud, L.P. (1989b) Angiotensin II may mediate excitatory neurotransmission from the subfornical organ to the hypothalamic supraoptic nucleus: an anatomical and electrophysiological study in the rat. Brain Res., 487: 52–61. Jhamandas, J.H., Aippersach, S.E. and Harris, K.H. (1991a) Cardiovascular influences on rat parabrachial nucleus: an electrophysiological study. Am. J. Physiol., 260: R225–R231. Jhamandas, J.H., Harris, K.H. and Krukoff, T.L. (1991b) Parabrachial nucleus projection towards the hypothalamic supraoptic nucleus: electrophysiological and anatomical observations in the rat. J. Comp. Neurol., 308: 42–50. Jones, B.E. and Moore, R.Y. (1977) Ascending projections of the locus coeruleus in the rat, II. Autoradiographic study. Brain Res., 127: 23–53. Kannan, H., Yagi, K. and Sawaki, Y. (1981) Pontine neurons: electrophysiological evidence of mediating carotid baroreceptor inputs to supraoptic neurons in rats. Exp. Brain Res., 42: 362–370. Kasai, M., Osaka, T., Inenaga, K., Kannan, H. and Yamashita, H. (1987) Gamma-aminobutyric acid antagonist blocks baroreceptor-activated inhibition of neurosecretory cells in the hypothalamic paraventricular nucleus of rats. Neurosci. Lett., 81: 319–324. Kaufman, S. (1984) Role of right atrial receptors in the control of drinking in the rat. J. Physiol. London, 349: 389–396. Kaufman, S. (1987) Influence of right atrial stretch on plasma AVP activity in the conscious rat. Can. J. Physiol. Pharmacol., 64: 255–259. Kirouac, G.J. and Ciriello, J. (1997) Cardiovascular responses to glutamate stimulation of diagonal band of Broca. Am. J. Physiol., 273: H540–H545.

271 Kiss, J., Magloczky, Z., Somogyi, J. and Freund, T.F. (1997) Distribution of calretinin-containing neurons relative to other neurochemically identified cell types in the medial septum of the rat. Neuroscience, 78: 399–410. Koizumi, K. and Yamashita, H. (1978) Influence of stretch receptors on hypothalamic neurosecretory neurones. J. Physiol. London, 285: 341–358. Kumada, M., Terui, N. and Kuwaki, T. (1990) Arterial baroreceptor reflex: its central and peripheral neural mechanisms. Prog. Neurobiol., 35: 331–361. Leng, G., Brown, C.H. and Russel, J.A. (1999) Physiological pathways regulating the activity of magnocellular neurosecretory cells. Prog. Neurobiol., 57: 625–655. Levine, J.D., Zhao, X. and Miselis, R.R. (1994) Direct and indirect retinohypothalamic projections to the supraoptic nucleus in the female albino rat. J. Comp. Neurol., 341: 214–224. Li, Y.W. and Dampney, R.A.L. (1994) Expression of Fos-like protein in brain following sustained hypertension and hypotension in conscious rabbits. Neuroscience, 61: 613–634. Lindheimer, M.D. and Davison, J.M. (1995) Osmoregulation, the secretion of arginine vasopressin and its metabolism during pregnancy. Euro. J. Endocrol., 132: 133–143. Lindvall, O. and Stenevi, U. (1978) Dopamine and noradrenaline neurons projecting to the septal area in the rat. Cell Tiss. Res., 190: 383–407. Ludwig, M. (1998) Dendritic release of vasopressin and oxytocin. J. Neuroendocrinol., 10: 881–895. Markram, H. and Segal, M. (1990) Electrophysiological characteristics of cholinergic and non-cholinergic neurons in the medial; septum-diagonal band complex. Brain Res., 513: 171– 174. Martin, P.Y. and Schrier, R.W. (1997) Sodium and water retention in heart failure — pathogenesis and treatment. Kidney Int., 59: S57–S61. Matthews, R.T. and Lee, W.L. (1991) A comparison of extracellular and intracellular recordings from medial septum/diagonal band neurons in vitro. Neuroscience, 42: 451–462. Mazzella, D., Cunningham, J.T. and Sullivan, M.J. (2000) The effects of diagonal band lesions (DBB) on vasopressin release. FASEB J., 14: A272. McAllen, R.M. and Harris, M.C. (1988) Long latency baroreceptor inhibition of supraoptic neurons in the cat. Neurosci. Lett., 84: 287–290. McKitrick, D.J., Krukoff, T.L. and Calaresu, F.R. (1992) Expression of c-fos in rat brain after electrical stimulation of the aortic depressor nerve. Brain Res., 599: 215–222. Menninger, R.P. (1979) Effects of carotid occlusion and left atrial stretch on supraoptic neurosecretory cells. Am. J. Physiol., 237: R63–R67. Meyer, D.K., Oertel, W.H. and Brownstein, M.J. (1980) Deafferentation studies on the glutamic acid decarboxylase content of the supraoptic nucleus of the rat. Brain Res., 200: 165–168. Miselis, R.R. (1981) The efferent projections of the subfornical organ of the rat: a circumventricular organ with a neural network subserving body fluid balance. Brain Res., 230: 1–23. Morgan, J.I. and Curran, T. (1991) Stimulus–transcription cou-

pling in the nervous system: involvement of the inducible protooncogenes fos and jun. Annu. Rev. Neurosci., 14: 421–451. Morris, M. and Alexander, N. (1989) Baroreceptor influences on oxytocin and vasopressin secretion. Hypertension, 13: 110– 114. Murase, S., Inui, K. and Nosaka, S. (1994) Baroreceptor inhibition of the locus coeruleus noradrenergic neurons. Neuroscience, 61: 635–643. Naraveaz, J.A., Covenas, R., de Leon, M., Aguirre, J.A., Cintra, A., Goldstein, M. and Fuxe, K. (1993) Induction of cfos immunoreactivity in tyrosine hydroxylase and phenylethanolamine-N-methyltransferase immunoreactive neurons in the medulla oblongata of the rat after phosphate-buffered saline load in the urethane-anaesthetized rat. Brain Res., 602: 342– 349. Nissen, R., Cunningham, J.T. and Renaud, L.P. (1993) Lateral hypothalamic lesions alter baroreceptor-evoked inhibition of rat supraoptic neurones. J. Physiol. (London), 470: 751–766. Norsk, P. (1996) Role of arginine vasopressin in the regulation of extracellular fluid volume. Med. Sci. Sports Exerc., 28: S36– S41. Ohman, L.E., Shade, R.E. and Haywood, J.R. (1990) Parabrachial modulation of vasopressin release. Am. J. Physiol., 258: R358–R364. Olpe, H.R., Berecek, K., Jones, R.S.G., Steinmann, M.W., Sonnenburg, Ch. and Hofbauer, K.G. (1985) Reduced activity of locus coeruleus neurons in hypertensive rats. Neurosci. Lett., 61: 25–29. Panula, P., Yang, H.Y. and Costa, E. (1984) Histamine-containing neurons in the rat hypothalamus. Proceedings of the National Academy of Sciences of the United States of America, 81(8): 2572–2576. Pendlebury, S.T., Dyball, R.E.J. and Honda, K. (1992) Expansion of plasma volume by intragastric isotonic saline inhibits supraoptic neurones in rats. J. Endocrinol., 135: 527–533. Pieribone, V.A. and Aston-Jones, G. (1991) Adrenergic innervation of the rat nucleus locus coeruleus arises predominantly from the C1 adrenergic cell group in the rostral medulla. Neuroscience, 41: 525–542. Pittman, Q.J., Mouginot, D. and Kombian, S.B. (1998) GABAB receptors and supraoptic neural activity. Adv. Exp. Med. Biol., 449: 107–115. Pittman, Q.J., Hirasawa, M., Mouginot, D. and Kombian, S.B. (2000) Neurohypophysial peptides as retrograde transmitters in the supraoptic nucleus of the rat. Exp. Physiol., 85S: 139S– 143S. Potts, P.D., Polson, J.W., Hirooka, Y. and Dampney, R.A.L. (1997) Effects of sinoaortic denervation on Fos expression in the brain evoked by hypertension and hypotension in conscious rabbits. Neuroscience, 77: 503–520. Poulain, D.A. and Theodosis, D.T. (1988) Electrical activity of peptidergic neurones and its relation to hormone release. In: B.T. Pickering, J.B. Wakerly and A.J.S. Summerlee (Eds.), Neurosecretion: Cellular Aspects of the Production and Release of Neuropeptides. Plenum Press, New York. Poulain, D.A. and Wakerly, J.B. (1982) Electrophysiology of

272 hypothalamic magnocellular neurones secreting oxytocin and vasopressin. Neuroscience, 7: 773–808. Randolph, R.R., Li, Q., Curtis, K.S., Sullivan, M.J. and Cunningham, J.T. (1998) Fos expression following isotonic volume expansion in unanesthetized male rat. Am. J. Physiol., 274: R1345–R1352. Reis, L.C., Ramalho, M.J., Favaretto, A.L.V., Gutkowska, J., McCann, S.M. and Antunes-Rodrigues, J. (1994) Participation of the ascending serotonergic system in the stimulation of atrial natriuretic peptide release. Proc. Natl. Acad. Sci. USA, 91: 12022–12026. Renaud, L.P. (1996) CNS pathways mediating cardiovascular regulation of vasopressin. Clin. Exp. Pharmacol. Physiol., 23: 157–160. Renaud, L.P. and Bourque, C.W. (1991) Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog. Neurobiol., 36: 131– 169. Renaud, L.P., Tand, M., McCann, M.J., Stricker, E.M. and Veralis, J.G. (1987) Cholecystokinin and gastric distention activate oxytocinergic cell in rat hypothalamus. Am. J. Physiol., 253: R661–R665. Renaud, L.P., Cunningham, J.T., Nissen, R., Yang, C.R. (1993) Electrophysiology of central pathways controlling release of neurohypophysial hormones. Focus on the lamina terminalis and diagonal band inputs to the supraoptic nucleus. Ann. N.Y. Acad. Sci., 689: 122–132. Ricardo, J. and Koh, E.T. (1978) Anatomical evidence of direct projections from the nucleus of the solitary tract to the hypothalamus, amygdala, and other forebrain structures in the rat. Brain Res., 153: 1–26. Roland, B.L. and Sawchenko, P.E. (1993) Local origins of some GABAergic projections of the paraventricular and supraoptic nuclei of the hypothalamus in the rat. J. Comp. Neurol., 332: 123–143. Saper, C.B. and Loewy, A.D. (1980) Efferent connections of the parabrachial nucleus in the rat. Brain Res., 197: 291–317. Sawchenko, P.E. and Swanson, L.W. (1982) The organization of noradrenergic pathways from the brainstem to the paraventricular and the supraoptic nuclei in the rat. Brain Res. Rev., 4: 275–325. Sawchenko, P.E., Swanson, L.W., Steinbusch, H.W.M. and Verhofstad, N.A.J. (1983) The distribution and cells of origin of serotinergic inputs to the paraventricular and supraoptic nuclei of the rat. Brain Res., 277: 355–360. Sawchenko, P.E., Arias, C. and Bittencourt, J.C. (1990) Inhibin beta- somatostatin- and enkephalin- immunoreactivities coexist in caudal medullary neurons that project to the paraventricular nucleus of the hypothalamus. J. Comp. Neurol., 291: 269–280. Schmid, P.G., Sharbi, F.M., Guo, G.B., Abboud, F.M. and Thames, M.D. (1984) Vasopressin and oxytocin in the neural control of the circulation. Fed. Proc., 43: 97–102. Schrier, R.W., Fassett, R.G., Ohara, M. and Martin, P.Y. (1998) Vasopressin release, water channels, and vasopressin antagonism in cardiac failure, cirrhosis and pregnancy. Proc. Assoc. Am. Phys., 110: 407–411. Senatorov, V.V. and Renaud, L.P. (1999) Projections of medullary

and pontine noradrenergic neurons to the horizontal limb of the nucleus diagonal band of the rat. Neuroscience, 88: 939– 947. Share, L. (1988) Role of vasopressin in cardiovascular regulation. Physiol. Rev., 68: 1248–1273. Share, L. (1996) Control of vasopressin release — an old but continuing story. NIPS, 11: 7–13. Share, L. and Levy, M.N. (1962) Cardiovascular receptors and blood titer of antidiuretic hormone. Am. J. Physiol., 203: 425– 428. Silverman, A.J., Hoffman, D.L. and Zimmerman, E.A. (1981) The descending afferent connections of the paraventricular nucleus of the hypothalamus (PVN). Brain Res. Bull., 6: 47–61. Sullivan, M.J., Cunningham, J.T., Nissen, R., Allen, A.M., Coderre, E. and Renaud, L.P. (1991) Ibotenic acid lesions of the diagonal band of Broca results in exaggerated polyethylene glycol-induced drinking behavior. Soc. Neurosci. Abstr., 17: 885. Sved, A.F. (1986) Pontine pressor regions which release vasopressin. Brain Res., 369: 143–150. Tappaz, M.L., Wassef, M., Oertel, W.H., Paut, L. and Pujol, J.F. (1983) Light- and electron-microscopic immunocytochemistry of glutamic acid decarboxylase (GAD) in the basal hypothalamus: morphological evidence for neuroendocrine gammaamino butyrate (GABA). Neuroscience, 9: 271–287. Ter Horst, G.J., De Boer, P., Luiten, P.G.M. and Van Willigen, J.D. (1989) Ascending projections from the solitary tract nucleus to the hypothalamus: a phaseolus vulgaris lectin tracing study in the rat. Neuroscience, 31: 785–797. Thames, M.D. and Schmid, P.G. (1979) Cardiopulmonary receptors with vagal afferents tonically inhibit ADH release in the dog. Am. J. Physiol., 237: H299–H304. Theodosis, D.T., Paut, L. and Tappaz, M.L. (1986) Immunocytochemical analysis of the GABAergic innervation of oxytocinand vasopressin-secreting neurons in the rat supraoptic nucleus. Neuroscience, 19: 207–222. Tribollet, E. and Dreifuss, J.J. (1981) Localization of neurones projecting to the hypothalamic paraventricular nucleus area of the rat: a horseradish peroxidase study. Neuroscience, 6: 1315–1328. Tribollet, E., Armstrong, W.E., Dubois-Dauphin, M. and Dreifuss, J.J. (1985) Extra-hypothalamic afferent inputs to the supraoptic nucleus area of the rat as determined by retrograde and anterograde tracing techniques. Neuroscience, 15: 135–148. Valentino, R.J., Page, M., Van Bockstaele, E. and Aston-Jones, G. (1992) Corticotrophin-releasing factor innervation of the locus coeruleus region: distribution of fiber and sources of input. Neuroscience, 48: 689–705. Van Den Pol, A. (1988) Dual ultrastructural localization of two neurotransmitter-related agents: colloidal gold-labeled neurophysin-immunoreactive supraoptic neurons receive peroxidase labeled glutamate decarboxylase- or gold-labeled synapses. J. Neurosci., 5: 2940–2954. Van Vulpen, E.H., Yang, C.R., Nissen, R. and Renaud, L.P. (1999) Hypothalamic A14 and A15 catecholamine cells provide the dopaminergic innervation to the supraoptic nucleus in rat: a combined retrograde tracer and immunohistochemical study. Neuroscience, 93: 675–680.

273 Vertes, R.P. (1988) Brainstem afferents to the basal forebrain in the rat. Neuroscience, 24: 907–935. Voisin, D.L., Chapman, C., Poulain, D.A. and Herbison, A.E. (1994) Extracellular GABA concentrations in rat supraoptic nucleus during lactation and following haemodynamic changes: an in vivo microdialysis study. Neuroscience, 63: 547–558. Wiess, M.L., Yang, Q.Z. and Hatton, G.I. (1989) Magnocellular tuberomammillary nucleus input to the supraoptic nucleus in the rat: anatomical and in vitro electrophysiological investigations. Neuroscience, 31: 581–586. Wilkin, L.D., Mitchell, L.D., Ganten, D. and Johnson, A.K. (1989) The supraoptic nucleus: afferents from areas involved in control of body fluid homeostasis. Neuroscience, 28: 573– 584.

Xi, D., Kusano, K. and Gainer, H. (1999) Quantitative analysis of oxytocin and vasopressin messenger ribonucleic acids in single magnocellular neurons isolated from supraoptic nucleus of rat hypothalamus. Endocrinology, 140: 4677–4682. Yamashita, H. (1977) Effect of baro- and chemoreceptor activation on supraoptic nuclei neurons in the hypothalamus. Brain Res., 126: 551–556. Yamashita, H., Inega, K., Kawata, M. and Sano, Y. (1983) Phasically firing neurons in the supraoptic nucleus of the rat hypothalamus: Immunocytochemical and electrophysiological studies. Neurosci. Lett., 37: 87–92. Zaborszky, L. and Cullinan, W.E. (1996) Direct catecholaminergic–cholinergic interaction in the basal forebrain. I. Dopamine-beta-hydroxylase and tyrosine hydroxylase input to cholinergic neurons. J. Comp. Neurol., 374: 535–554.

This Page Intentionally Left Blank

D. Poulain, S. Oliet and D. Theodosis (Eds.) Progress in Brain Research, Vol. 139 © 2002 Elsevier Science B.V. All rights reserved

CHAPTER 21

The central vasopressinergic system in experimental left ventricular hypertrophy and dysfunction Frank Muders 1,∗ , Günter A.J. Riegger 1 , Udo Bahner 2 and Miklos Palkovits 3 1

Klinik und Poliklinik für Innere Medizin II, University of Regensburg, Regensburg, Germany 2 Medizinische Universitätklinik, Würzburg, Germany 3 Laboratory of Neuromorphology, Semmelweis University, Budapest, Hungary

Introduction While the pathophysiological role of compensatory neurohumoral mechanisms in heart failure has been widely investigated, the significance of the centrally acting cardiovascular neuropeptide systems remains largely unclear. Measurements of metabolic markers such as hexokinase, cFos and NADPH-diaphorase suggest increased neuron activity in hypothalamic areas of the brain, the locus coeruleus and in other nuclei of the brain involved in central circulatory regulation in rats 6 weeks after myocardial infarction (Patel et al., 1993, 2000; Zhang et al., 1998). There have been only isolated investigations of the functional state of central neuropeptide systems involved in cardiocirculatory regulation in heart failure. With regard to the central renin–angiotensin system, there have been two experimental studies suggesting stimulation. Intracerebroventricular administration of the AT1 receptor antagonist, losartan, in sheep with heart failure due to rapid ventricular pacing led to hemodynamic changes, including a fall in blood pressure, which did not occur in healthy animals (Rademaker

∗ Correspondence

to: F. Muders, Klinik und Poliklinik für Innere Medizin II, Universitätsklinikum, FranzJosef-Strauss-Allee 11, 93053 Regensburg, Germany. Tel.: +49-941-9447211; Fax: +49-941-9447213; E-mail: [email protected]

et al., 1995). In rats with an aortocaval fistula, there was increased AT1 receptor expression in the nucleus paraventricularis, nucleus tractus solitarius and subfornical organ (Yoshimura et al., 2000). Investigations of the central vasopressinergic system in heart failure have not been conducted to date. Our aim, therefore, was to investigate the central vasopressinergic system in a model of myocardial hypertrophy and left ventricular dysfunction (supravalvular aortic stenosis model). Banding of the ascending aorta in 8-weeks-old Wistar rats produces severe ventricular hypertrophy after 12 weeks of pressure overload. At this stage, this model is characterized by a transition of LV hypertrophy into cardiac failure, as suggested by functional and molecular studies, as well as by an increased mortality (Weinberg et al., 1994; Bruckschlegel et al., 1995). Neurohumoral vasoconstrictor systems were stimulated in this model in order to maintain blood pressure in the chronic aortic stenosis with a reduced cardiac index. In comparison to healthy control animals, vasopressin plasma levels and plasma renin activities were significantly increased, while levels of plasma norepinephrine and 24-h urinary excretion were unchanged (Muders et al., 1995), parameters both indicating the activity of the sympathetic nervous system (Goldstein et al., 1983). Increased levels of plasma vasopressin can be attributed to non-osmotic stimuli in this model since plasma osmolality was unchanged in comparison to

276

control animals. In an earlier investigation of the same animal model, at a later time (after 15 weeks), enhanced vasopressin plasma levels, not related to plasma osmolality, were found in animals which had developed a marked left ventricular dysfunction (Riegger et al., 1988). Our studies support the hypothesis that non-osmotic stimuli play an important role in the regulation of vasopressin in cardiovascular disorders. In the model of supravalvular aortic stenosis, baroreceptor stimulation and/or angiotensin IImediated stimulation of vasopressin release is possible: the supravalvular position of the silver clip leads to a weakened signal to baroreceptors in the aorta and carotid artery and, consequently, to stimulation of vasopressin secretion. In contrast, cardiac baroreceptors are exposed to increased left ventricular filling pressure and, consequently, counteract increased vasopressin release. However, cardiopulmonary receptors play a subordinate role in the regulation of vasopressin release under physiological conditions in the presence of intact arterial baroreceptors (Chen et al., 1991). In order to test the hypothesis that levels of vasopressin change centrally as well as peripherally, levels of vasopressin were measured in 20 different areas of the brain involved in central cardiocirculatory regulation. The brain areas were obtained by the micropunch technique of Palkovits (Palkovits, 1973). In the hypothalamus, different changes were found in the areas producing vasopressin. While vasopressin content in the paraventricular and suprachiasmatic nuclei was significantly raised in rats with aortic stenosis, no changes were found in the supraoptic nucleus, compared to healthy control animals. This aspect is of particular interest, since the paraventricular and suprachiasmatic nuclei are involved in central cardiocirculatory regulation through their vasopressinergic connections to the brainstem while the supraoptic nucleus does not have these connections. These different changes in vasopressin content in these hypothalamic brain areas indicate an alteration in the central vasopressinergic cardiocirculatory control in animals with supravalvular aortic stenosis. In addition, there was a significantly increased content of vasopressin in the median eminence, an area which is part of the hypothalamo-hypophysial axis. This implies that stimulation of the axis in animals with aortic stenosis consequently raises plasma levels of vasopressin.

The vasopressin content was also significantly increased in four other ‘extrahypothalamic’ brain areas involved in central cardiocirculatory control. Interestingly, vasopressin content in the locus coeruleus was markedly diminished compared to control animals. In association with the unchanged norepinephrine plasma levels and urinary excretion in rats with aortic stenosis, the results suggest an important central circulatory regulating mechanism, in which a reduced vasopressin concentration in the locus coeruleus may counteract activation of the sympathetic nervous system. In summary, our studies of the central vasopressinergic system in experimental left ventricular hypertrophy and dysfunction show that, in addition to stimulation of peripheral neurohumoral factors, a central neuropeptide system is altered which, in rats, is involved with supravalvular aortic stenosis through its cardiovascular regulatory effects (Muders et al., 1995). Modulation of the central vasopressinergic system by blockade of the renin–angiotensin system In the 1980s, an association was found between the renin–angiotensin system and vasopressin in patients with heart failure. In these patients, who had plasma levels that were increased inappropriately relative to the plasma osmolality, chronic treatment with the ACE inhibitor, Captopril, led to a significant reduction in the vasopressin plasma levels and to a restoration of the relationship between plasma osmolality and vasopressin (Lee and Packer, 1986; Riegger and Kochsiek, 1986). Nowadays, blockade of the renin– angiotensin system by ACE inhibitors and AT1 receptor antagonists is established in the treatment of heart failure, left ventricular hypertrophy and hypertension. The therapeutic efficacy of these medications, including a reduction in the mortality of these cardiovascular disorders, has been confirmed in animal models as well as in patients (The SOLVD Investigators, 1992; The Acute Infarction Ramipril Efficacy Study Investigators, 1993; Remme, 1995). This also includes investigations that were performed using the model of supravalvular aortic stenosis in the rat. Despite persisting strain on the left ventricle, treatment with an ACE inhibitor and AT1 receptor antagonist, carried out from the 6th postoperative week, led to a signif-

277

icant regression of left ventricular hypertrophy and to a reduction in mortality, in both treatment groups (Bruckschlegel et al., 1995). It is not known whether chronic treatment with an ACE inhibitor and AT1 receptor antagonist, respectively, also produces modulation of central neuropeptide systems involved in cardiocirculatory regulation, like the vasopressinergic system. In order to pursue this question, we measured vasopressin in plasma and in individual brain areas in rats with supravalvular aortic stenosis, which were treated chronically either with an ACE inhibitor or with an AT1 receptor antagonist. Twelve weeks after placement of the aortic clip, plasma vasopressin levels were reduced in both treatment groups compared to untreated aortic stenosis animals and showed no significant differences with healthy animals. Treatment with the ACE inhibitor and AT1 receptor antagonist also led to a reduction in vasopressin concentrations in specific areas of the hypothalamus and brainstem. These effects were different in some brain areas, which can be attributed to a different distribution of ACE and angiotensin II receptors in these areas (Muders et al., 1999). The fact that treatment of rats with supravalvular aortic stenosis with an ACE inhibitor and AT1 receptor antagonist, respectively, led not to activation but to suppression of the vasopressinergic system, despite a further fall in blood pressure, is of particular interest pathophysiologically and can be explained by (1) modulation of the baroreceptor reflex, and (2) direct peripheral or central inhibition of vasopressin synthesis and release. As explained above, the dysregulation of baroreceptors with consequent restriction of sympathoadrenal and humoral inhibition is a pathophysiological characteristic of chronic heart failure, which includes increased vasopressin synthesis and release (Zucker and Gilmore, 1985; Hirsch et al., 1987). The ability of the baroreceptors to shift their baseline in the direction of the surrounding pressure (reflex resetting) is absent and the sensitivity of the baroreceptors is reduced, as we also found in the model of supravalvular aortic stenosis (unpublished data). As has already been shown in rabbits with experimental heart failure and in patients, treatment with an ACE inhibitor or AT1 receptor antagonist leads to an improvement in the sensitivity of the arterial baroreceptors and restores the ability of the pressure

discharge curve to be readjusted (Murakami et al., 1996; Grassi et al., 1997). Whether chronic blockade of the renin–angiotensin system in the model of supravalvular aortic stenosis leads to an improvement in the baroreceptor reflex is unknown and should be investigated further. In addition to these effects mediated by the baroreceptor reflex, a ‘direct’ effect of the ACE inhibitor and AT1 receptor antagonist, respectively, on the vasopressinergic system is possible, in which vasopressin synthesis and release are suppressed either through angiotensin II receptors of the circumventricular organs or through direct inhibition of central ACE or angiotensin II receptors in specific areas of the brain. Central action of ACE inhibitors and AT1 receptor antagonists: in vitro autoradiography of the brain The existence of a central renin–angiotensin system and the significance of central angiotensin II in cardiocirculatory regulation are recognized today. In addition to synaptic inhibition of the baroreceptor reflex, central angiotensin II causes stimulation of sympathetic nervous activity and vasopressin synthesis and release, which can be suppressed by prior administration of an AT1 receptor antagonist (Unger et al., 1988; Wright and Harding, 1995). On the other hand, the central efficacy of systemically administered ACE inhibitors and AT1 receptor antagonists is controversial. Since chronic oral treatment with an ACE inhibitor and AT1 receptor antagonist in rats with supravalvular aortic stenosis produces suppression of central vasopressin and may be attributed to a central effect of these drugs, we studied the central efficacy of an ACE inhibitor and AT1 receptor antagonist, using in vitro autoradiography after systemic administration for 2 and 4 weeks, respectively (Muders et al., 1997, 2001). We were able to show that both medications caused significant ACE inhibition and angiotensin II receptor blockade in specific brain areas. Moreover, we recorded diminished ACE activities and angiotensin II receptor densities in brain areas that have no blood–brain barrier (and were thus accessible to the drug circulating in the blood) as well as in areas situated inside the brain.

278

A precondition for the central efficacy of an orally administered medication in specific brain areas is penetration of the blood–brain barrier. Studies of the distribution of orally administered radioactively marked drugs often show an absence of accumulation of the ACE inhibitor and AT1 receptor antagonists after a single dose. However, autoradiographic investigations that employed chronic administration as in this investigation, revealed a central blockade of the renin–angiotensin system. Chronic therapy with an ACE inhibitor or AT1 receptor antagonist (as is usual in the treatment of cardiovascular diseases) therefore includes central effects (Gohlke et al., 1989; Song et al., 1991; Polidori et al., 1996). There has been evidence of a central action, even for treatment with Captopril, the most hydrophilic drug of all the ACE inhibitors. Neurophysiological investigations showed for Captopril a marked improvement in cognitive abilities in comparison with placebo-treated patients (Zubenko and Nixon, 1984). In summary, mechanisms of the blood–brain barrier are complex and the penetration of a drug is dependent on many factors, including the duration of use, lipophilia of individual drugs and dosage. The central efficacy of ACE inhibitors and AT1 receptor antagonists is of interest insofar as the modulation of central cardiocirculatory mechanisms contributes to the therapeutic effects of these medications in the treatment of cardiovascular diseases.

centrations in the locus coeruleus, an important regulatory area of sympathetic nervous activity, suggest a central regulatory mechanism through which stimulation of the sympathetic nervous activity can be prevented. Our investigations showed that non-osmotic factors like the baroreceptor reflex and angiotensin II, are important stimuli of the vasopressinergic system. We were also able to show that the central vasopressinergic system in rats with experimental heart failure and myocardial hypertrophy is inhibited by treatment with an ACE inhibitor and AT1 receptor antagonist. As seen with autoradiography, this effect is mediated by a central effect of the drugs. Research into central regulatory mechanisms in cardiovascular diseases is, on the one hand, of crucial importance to our understanding of complex pathophysiological processes, and on the other hand, it serves the development of new therapeutic approaches with the goal of influencing these mechanisms directly pharmacologically and for the elucidation of central, currently unknown effects of cardiovascular drugs. Abbreviations ACE AT LV NADPH

angiotensin converting enzyme angiotensin left ventricular nicotinamide adenine dinucleotide phosphate

Summary References In the course of cardiac diseases, various neurohormonal systems in the plasma are activated. So far there have been only isolated results of investigations about the functional state of central neuropeptide systems in cardiac diseases and, in particular, in heart failure. We investigated, therefore, the central vasopressinergic system, an important neuropeptide system in cardiocirculatory regulation in a model of myocardial hypertrophy and le